From owner-linux-arm-kernel@lists.arm.linux.org.uk Tue Aug 01 00:02:19 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13JOZD-0003QU-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Tue, 01 Aug 2000 00:01:51 +0100 Received: from [161.114.1.208] (helo=ztxmail04.ztx.compaq.com) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13JOZC-0003Pl-00 for linux-arm-kernel@lists.arm.linux.org.uk; Tue, 01 Aug 2000 00:01:51 +0100 Received: by ztxmail04.ztx.compaq.com (Postfix, from userid 12345) id CB27C4D1; Mon, 31 Jul 2000 18:01:29 -0500 (CDT) Received: from oils.ozy.dec.com (oils.ozy.dec.com [16.153.192.1]) by ztxmail04.ztx.compaq.com (Postfix) with ESMTP id B7E13754; Mon, 31 Jul 2000 18:01:27 -0500 (CDT) Received: from ozy.dec.com by oils.ozy.dec.com (8.8.8/1.1.10.5/12Oct99-0148PM) id JAA0000029541; Tue, 1 Aug 2000 09:01:21 +1000 (EST) Message-ID: <39860529.E0452BF0@ozy.dec.com> Date: Tue, 01 Aug 2000 09:00:57 +1000 From: Charlie Flynn X-Mailer: Mozilla 4.7 [en] (WinNT; I) X-Accept-Language: en MIME-Version: 1.0 To: samcconn@cotw.com Cc: linux-arm-kernel@lists.arm.linux.org.uk Subject: Re: Touch Screen Drivers References: <3985F480.549D0D99@cotw.com> Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk Hi Scott, > > The kernel used by the iPAQ had three touch screens: > Actually there may be three touch screen files but the iPAQ uses only the h3650_ts.c. The other two drivers are for other platforms and interface to different touch screen hardware. There is a move under foot to make all touch screendrivers have a generic interface ( as far as X is concerned ) you can find the current (draft) spec at: http://handhelds.org/projects/ts_spec.html > > h3650_ts.c sa1100_ts.c ucb1200_ts.c (the last two being nearly the > same.) I was not able to determin how the init code was called for > h365_ts.c. I am not exactly sure what you mean. There is a function in h3650_ts called init_module() which initialises the driver. If you mean what inits the iPAQ touch screen hardware then the iPAQ driver doesn't talk to a touch screen device directly but to a microcontroller which ( in turn ) talks to the touch screen device. Hope this helps -Charlie > unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Tue Aug 01 03:08:04 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13JRSl-0005yb-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Tue, 01 Aug 2000 03:07:23 +0100 Received: from [210.184.5.125] (helo=pikachu.idthk.com) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13JRSi-0005y5-00 for linux-arm-kernel@lists.arm.linux.org.uk; Tue, 01 Aug 2000 03:07:21 +0100 Received: from localhost (thomas@localhost) by pikachu.idthk.com (8.9.3/8.8.7) with ESMTP id KAA12043; Tue, 1 Aug 2000 10:08:54 +0800 Date: Tue, 1 Aug 2000 10:08:54 +0800 (CST) From: Tak-Shing Chan To: Charlie Flynn cc: linux-arm-kernel@lists.arm.linux.org.uk Subject: Re: Touch Screen Drivers In-Reply-To: <39860529.E0452BF0@ozy.dec.com> Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk Is there already someone working on the generic touchscreen driver for the Assabet? Tak-Shing Chan On Tue, 1 Aug 2000, Charlie Flynn wrote: > There is a move under foot to make all touch screendrivers have a generic interface ( as > far as X is concerned ) you can find the current (draft) spec at: > > http://handhelds.org/projects/ts_spec.html unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Tue Aug 01 04:40:49 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13JSux-0007OH-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Tue, 01 Aug 2000 04:40:35 +0100 Received: from [161.114.1.209] (helo=ztxmail05.ztx.compaq.com) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13JSuw-0007NV-00 for linux-arm-kernel@lists.arm.linux.org.uk; Tue, 01 Aug 2000 04:40:34 +0100 Received: by ztxmail05.ztx.compaq.com (Postfix, from userid 12345) id 58B034DB; Mon, 31 Jul 2000 22:40:13 -0500 (CDT) Received: from oils.ozy.dec.com (oils.ozy.dec.com [16.153.192.1]) by ztxmail05.ztx.compaq.com (Postfix) with ESMTP id 34BB5428; Mon, 31 Jul 2000 22:40:07 -0500 (CDT) Received: from ozy.dec.com by oils.ozy.dec.com (8.8.8/1.1.10.5/12Oct99-0148PM) id NAA0000024499; Tue, 1 Aug 2000 13:40:02 +1000 (EST) Message-ID: <39864678.56F1EDE9@ozy.dec.com> Date: Tue, 01 Aug 2000 13:39:37 +1000 From: Charlie Flynn X-Mailer: Mozilla 4.7 [en] (WinNT; I) X-Accept-Language: en MIME-Version: 1.0 To: Tak-Shing Chan Cc: linux-arm-kernel@lists.arm.linux.org.uk Subject: Re: Touch Screen Drivers References: Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk Tak-Shing Chan wrote: > Is there already someone working on the generic touchscreen driver > for the Assabet? We are still in the draft spec phase. When it is finalised individual driver writers will be able to modify their interfaces. > > > Tak-Shing Chan > > On Tue, 1 Aug 2000, Charlie Flynn wrote: > > > There is a move under foot to make all touch screendrivers have a generic interface ( as > > far as X is concerned ) you can find the current (draft) spec at: > > > > http://handhelds.org/projects/ts_spec.html unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Tue Aug 01 09:11:53 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13JX8r-0005DG-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Tue, 01 Aug 2000 09:11:13 +0100 Received: from [203.73.94.214] (helo=sinamail.com) by www.linux.org.uk with smtp (Exim 3.13 #1) id 13JX8p-0005D9-00 for linux-arm-kernel@lists.arm.linux.org.uk; Tue, 01 Aug 2000 09:11:12 +0100 Received: (qmail 12422 invoked by uid 99); 1 Aug 2000 07:52:33 -0000 Message-ID: <20000801075233.12419.qmail@sinamail.com> X-Mailer: SinaMail 3.0Beta (FireToad) From: yjchang2k To: linux-arm-kernel@lists.arm.linux.org.uk Cc: stevenc@dvm.tatung.com.tw Subject: system utilities of ARM File system Date: Tue, 01 Aug 2000 07:52:33 +0800 Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk Dear all: Do somebody know where to get the source code of the utilities program ( ls, cat vi ... etc). Then build the image like the ramdisk for SA1110 or SA1100. Regards, ______________________________________ =================================================================== ·Q¾Ö¦³¥v¤W¶W±jªº50MB¶W¤jEmail? http://www.sinamail.com ½Ð¤W·s®öºôhttp://www.sina.com.tw unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Tue Aug 01 11:32:48 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13JZK9-0007Lk-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Tue, 01 Aug 2000 11:31:01 +0100 Received: from [193.67.108.1] (helo=white.plant.nl) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13JZK8-0007LX-00 for linux-arm-kernel@lists.arm.linux.org.uk; Tue, 01 Aug 2000 11:31:00 +0100 Received: from [194.134.11.59] by white.plant.nl (NTMail 4.30.0013/NT3720.00.67541ba1) with ESMTP id lbrigaaa for ; Tue, 1 Aug 2000 12:28:48 +0200 Message-ID: <004e01bffba2$eb885f40$3c04a8c0@EDWIN> From: "Daven Baptist" To: References: <20000801075233.12419.qmail@sinamail.com> Subject: how to boot linux after angel on a assabet? Date: Tue, 1 Aug 2000 12:26:13 +0200 MIME-Version: 1.0 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: 7bit X-Priority: 3 X-MSMail-Priority: Normal X-Mailer: Microsoft Outlook Express 5.00.2919.5600 X-MimeOLE: Produced By Microsoft MimeOLE V5.00.2919.5600 Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk hello, can someone tell me how to boot linux automaticaly after turning on the assabet. curently i have to upload the kernel and ram disk before it starts. if my device driver goes wrong, i have to reaload the kernel how can i place the kernel in rom/flash to speed up this process. thanks, Daven Baptist unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Tue Aug 01 11:49:44 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13JZbt-0007cT-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Tue, 01 Aug 2000 11:49:21 +0100 Received: from [193.145.240.2] (helo=tid.tid.es) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13JZbr-0007cN-00 for linux-arm-kernel@lists.arm.linux.org.uk; Tue, 01 Aug 2000 11:49:19 +0100 Received: from tid.es ([172.17.1.9]) by tid.tid.es (Netscape Messaging Server 4.15) with ESMTP id FYM0QE02.I4W for ; Tue, 1 Aug 2000 12:49:26 +0200 Message-ID: <3986AA12.5FCC008B@tid.es> Date: Tue, 01 Aug 2000 12:44:34 +0200 From: David Aparicio Campos X-Mailer: Mozilla 4.7 [en] (Win98; I) X-Accept-Language: en MIME-Version: 1.0 To: linux-arm-kernel@lists.arm.linux.org.uk Subject: CATS motherboard & ISA compatible registers Content-Type: multipart/mixed; boundary="------------435F040E50B4704129AAF827" Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk This is a multi-part message in MIME format. --------------435F040E50B4704129AAF827 Content-Type: multipart/alternative; boundary="------------1E631CAFB69191BFF11DCB77" --------------1E631CAFB69191BFF11DCB77 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Hello all: I'm still trying to build a firmware of my own in order to boot Linux from a CATS motherboard. ----------------------------------------------- This one is similar to an EBSA285, but it has some important differences: - Has a SouthBridge, this is: a PCI/ISA bridge, I/O controller, IDE controller, USB and PMU funcionality. - Has a RTC (a Dallas 1687) connected to that SouthBridge. Less important: - Flash is an 256 KB AMD compatible (not an 4MB Intel one like EBSA and NetWinder). - It's an ATX form factor, and has standard PC connectors, like KBD, IDE and so on. A NetWinder machine is closer in design, but it uses Winbond bridges, I couldn't get useful information from "nettrom" code (firmware code of NetWinder machines). ----------------------------------------------- I sent a previous mail about an "IRQ7" problem when I started Linux, something like: eter> call 0x8000 boot no options Linux version 2.2.13-rmk2 (root@lobos) (gcc version 2.95.1 19990816 (release)) #5 Thu Jul 20 07:14:04 CEST 2000 initrd extends beyond end of memory (0xea000c22 > 0xc1000000) - disabling initrd Calibrating delay loop... <3>IRQ LOCK: IRQ7 is locking the system, disabled IRQ LOCK: IRQ7 is locking the system, disabled IRQ LOCK: IRQ7 is locking the system, disabled ... Russell King suggested a problem with PIC controller (IRQ7 is something related to bad inicialization of some ISA-related thing), and Jamey Hicks told me to revise external interrupts from RTC (I'm almost sure that this is the source of the deadly interrupt). Hey guys, thank you both!! I tried to change my code, leaving CATS in a similar state that original firmware does, but this didn't changed anything. (If you wonder why I am working in something that is already done, you have not seen CATS firmware running :-). On the other hand, that is not free sofware). I sweeped throught Linux kernel and tried to find some clue about this problem. What I have seen is that, If I comment some lines in init/main.c (start_kernel)... // sti() // calibrate_delay() ...Linux does all inicialization and tries to mount root FS. So what is happening is that when kernel enables interrupts, timer interrupt floods the system (Russell and Jamey were totally right). Surely, its related with timer driving (CATS architecture uses RTC periodic interrupt to handle tick beats, routed thought ISA IRQ subsystem to the StrongArm). The problem by now is that datasheet for my southbridge is very obscure about "ISA compatible registers", those related to PIC (0x20, 0x21, 0xA0 & 0xA1) and RTC (0x70 and 0x71). I have some doubts about this specific inicialization. Seems like some register inicialization should be removed from Linux kernel and put in firmware/loader code (e.g bus master handling in arch/arm/kernel/hw-ebsa285.c and so on). I'm trying to summarize this, if there is somebody interested... Does anybody have some more information about this ISA-PC-compatible inicialization issue?? TYIA David Aparicio --------------1E631CAFB69191BFF11DCB77 Content-Type: text/html; charset=us-ascii Content-Transfer-Encoding: 7bit Hello all:

I'm still trying to build a firmware of my own in order to boot Linux from a
CATS motherboard.
-----------------------------------------------
This one is similar to an EBSA285, but it has some important differences:

- Has a SouthBridge, this is: a PCI/ISA bridge, I/O controller, IDE controller,
    USB and PMU funcionality.
- Has a RTC (a Dallas 1687) connected to that SouthBridge.

Less important:
- Flash is an 256 KB AMD compatible (not an 4MB Intel one like EBSA
    and NetWinder).
- It's an ATX form factor, and has standard PC connectors, like KBD, IDE
    and so on.

A NetWinder machine is closer in design, but it uses Winbond bridges, I
couldn't get useful information from "nettrom" code (firmware code of
NetWinder machines).
-----------------------------------------------
I sent a previous mail about an "IRQ7" problem when I started Linux,
something like:

eter> call 0x8000
boot no options
Linux version 2.2.13-rmk2 (root@lobos) (gcc version 2.95.1 19990816 (release)) #5 Thu Jul 20 07:14:04
CEST 2000
initrd extends beyond end of memory (0xea000c22 > 0xc1000000) - disabling initrd
Calibrating delay loop... <3>IRQ LOCK: IRQ7 is locking the system, disabled
IRQ LOCK: IRQ7 is locking the system, disabled
IRQ LOCK: IRQ7 is locking the system, disabled
...

Russell King suggested a problem with PIC controller (IRQ7 is something
related to bad inicialization of some ISA-related thing), and Jamey Hicks
told me to revise external interrupts from RTC (I'm almost sure that
this is the source of the deadly interrupt). Hey guys, thank you both!!

I tried to change my code, leaving CATS in a similar state that original
firmware does, but this didn't changed anything.
(If you wonder why I am working in something that is already done, you
have not seen CATS firmware running :-). On the other hand, that is not free
sofware).
I sweeped throught Linux kernel and tried to find some clue about this problem.
What I have seen is that, If I comment some lines in init/main.c (start_kernel)...

// sti()
// calibrate_delay()

...Linux does all inicialization and tries to mount root FS. So what is happening
is that when kernel enables interrupts, timer interrupt floods the system (Russell
and Jamey were totally right). Surely, its related with timer driving (CATS
architecture uses RTC periodic interrupt to handle tick beats, routed thought ISA
IRQ subsystem to the StrongArm).

The problem by now is that datasheet for my southbridge is very obscure about
"ISA compatible registers", those related to PIC (0x20, 0x21, 0xA0 & 0xA1)
and RTC (0x70 and 0x71). I have some doubts about this specific inicialization.

Seems like some register inicialization should be removed from Linux kernel
and put in firmware/loader code (e.g bus master handling in
arch/arm/kernel/hw-ebsa285.c and so on). I'm trying to summarize this, if there
is somebody interested...

Does anybody have some more information about this ISA-PC-compatible
inicialization issue??
TYIA

David Aparicio --------------1E631CAFB69191BFF11DCB77-- --------------435F040E50B4704129AAF827 Content-Type: text/x-vcard; charset=us-ascii; name="daparic.vcf" Content-Transfer-Encoding: 7bit Content-Description: Card for David Aparicio Campos Content-Disposition: attachment; filename="daparic.vcf" begin:vcard n:Aparicio Campos;David tel;fax:+34-91-3374502 tel;work:+34-91-3374178 x-mozilla-html:TRUE org:Telefónica I+D;Gerencia 1200 adr:;;Emilio Vargas, 6;Madrid;;28043;Spain version:2.1 email;internet:daparic@tid.es title:R&D Engineer end:vcard --------------435F040E50B4704129AAF827-- unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Tue Aug 01 12:15:17 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13Ja0q-00084B-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Tue, 01 Aug 2000 12:15:08 +0100 Received: from [130.161.42.1] (helo=duteinh.et.tudelft.nl) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13Ja0p-000843-00 for linux-arm-kernel@lists.arm.linux.org.uk; Tue, 01 Aug 2000 12:15:07 +0100 Received: from dutein75 by duteinh.et.tudelft.nl (8.9.3/BB-2.0) id NAA29325; Tue, 1 Aug 2000 13:15:14 +0200 (MET DST) Message-Id: <200008011115.NAA29325@duteinh.et.tudelft.nl> From: Erik Mouw To: linux@arm.linux.org.uk Cc: linux-arm-kernel@lists.arm.linux.org.uk Subject: Re: Compiler errors when using linux/interrupt.h In-Reply-To: <200007312018.VAA17566@flint.arm.linux.org.uk> Organization: Eric conspiracy secret labs Date: Tue 01 Aug 2000 13:15:13 +0100 (MDT) X-Face: "!Z|1XjFg!3Iw)BwM^M\k)J5KHiPddz}#/Zw%?$&Q!xqob"OA|})3_@&BTTiQERo?/^g[Z<9P4R8:5C:aqq.cw!hotLc=+u-"+1rq@9$V^8PbOUlodyi_sejt+D`iqkLpx$>.aXl%vnfPIt"LK|/%>94_Zrm8jy?`_[RKt'TuoxUxfk(gMhY(n|DBBl~c$dhO>G%6}R4+DGX3|idBm4oU&UM/XUP4I9aoN{]`ax%B?rgD(I3OJ<62Y6L X-Mailer: XCmail 1.2devel - with PGP support, PGP engine version 0.5 (IRIX64) X-Mailerorigin: http://www.fsai.fh-trier.de/~schmitzj/Xclasses/XCmail/ X-Eric-Conspiracy: There is no conspiracy! X-Loop: erik@it.et.tudelft.nl MIME-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: 8bit Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk On Mon, 31 Jul 2000 21:18:20 +0100 (BST), Russell King - ARM Linux Admin wrote: > Erik Mouw writes: >> include/asm/hardirq.h:10: warning: `local_irq_count' redefined >> include/linux/modules/armksyms.ver:12: warning: this is the location of the > previous definition >> include/asm/hardirq.h:11: warning: `local_bh_count' redefined >> include/linux/modules/armksyms.ver:10: warning: this is the location of the > previous definition > > Ah darn! Guess I'll have to grab some changes from Keith Owens little > "improvement". Is it already generally approved by the port maintainers? I didn't see much objection on linux-kernel, but the patch is not yet in 2.4.0-test6-pre1. Erik -- Aug 01 13:15:11 linux kernel: lp0 printer error Aug 01 13:15:11 linux kernel: lp0 on fire -- Linux kernel log unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Tue Aug 01 13:52:51 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13JbWp-00015o-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Tue, 01 Aug 2000 13:52:15 +0100 Received: from [132.233.247.11] (helo=thalia.fm.intel.com) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13JbWo-00015a-00 for linux-arm-kernel@lists.arm.linux.org.uk; Tue, 01 Aug 2000 13:52:14 +0100 Received: from SMTP (fmsmsxvs04-1.fm.intel.com [132.233.42.204]) by thalia.fm.intel.com (8.9.1a+p1/8.9.1/d: relay.m4,v 1.30 2000/06/08 18:25:35 dmccart Exp $) with SMTP id MAA24789; Tue, 1 Aug 2000 12:53:18 GMT Received: from fmsmsx26.fm.intel.com ([132.233.48.26]) by 132.233.48.204 (Norton AntiVirus for Internet Email Gateways 1.0) ; Tue, 01 Aug 2000 12:52:20 0000 (GMT) Received: by fmsmsx26.fm.intel.com with Internet Mail Service (5.5.2650.21) id <3X9RNGJM>; Tue, 1 Aug 2000 05:52:19 -0700 Message-ID: From: "Seger, Justin M" To: "'Daven Baptist'" , linux-arm-kernel@lists.arm.linux.org.uk Subject: RE: how to boot linux after angel on a assabet? Date: Tue, 1 Aug 2000 05:52:15 -0700 MIME-Version: 1.0 X-Mailer: Internet Mail Service (5.5.2650.21) Content-Type: text/plain; charset="iso-8859-1" Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk > -----Original Message----- > hello, > can someone tell me how to boot linux automaticaly after > turning on the assabet. > > curently i have to upload the kernel and ram disk before it starts. > if my device driver goes wrong, i have to reaload the kernel > > how can i place the kernel in rom/flash to speed up this process. > Angel does not support doing thing. You will need to install another bootloader. blob (http://www.lart.tudelft.nl/lartware/blob/) version 1.0.8-pre2 supports Assabet. You will need to make sure that you're using a very recent kernel (2.4.0-test4-rmk1-np1 or later, IIRC) and you need to disable the support for the "Angel Debug Monitor" in the kernel config. Good luck, Justin Seger unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Tue Aug 01 14:50:20 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13JcQo-0001xu-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Tue, 01 Aug 2000 14:50:06 +0100 Received: from [193.81.246.11] (helo=eins.siemens.at) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13JcQn-0001xM-00 for linux-arm-kernel@lists.arm.linux.org.uk; Tue, 01 Aug 2000 14:50:05 +0100 Received: from scesie13.sie.siemens.at (forix-hme0 [10.1.140.2]) by eins.siemens.at with ESMTP id PAA05809; Tue, 1 Aug 2000 15:41:48 +0200 (MET DST) Received: (from smap@localhost) by scesie13.sie.siemens.at (8.9.3/8.9.3) id PAA01315; Tue, 1 Aug 2000 15:49:51 +0200 (MET DST) Received: from vies141a.sie.siemens.at(195.1.196.56) by scesie13 via smap (V2.0beta) id xma029705; Tue, 1 Aug 00 15:48:45 +0200 Received: by vies141a.sie.siemens.at with Internet Mail Service (5.5.2650.21) id ; Tue, 1 Aug 2000 15:48:42 +0200 Message-ID: From: Ritter Thomas To: "'linux-arm-kernel@lists.arm.linux.org.uk'" , "'john+@cs.cmu.edu'" Subject: PCMCIA with Brutus Date: Tue, 1 Aug 2000 15:48:41 +0200 MIME-Version: 1.0 X-Mailer: Internet Mail Service (5.5.2650.21) Content-Type: text/plain; charset="iso-8859-1" Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk Hi, After successfully starting Linux with the default configuration I'm trying now to make the networking stuff (via a PCMCIA card) working. In the PCMCIA doc (from Documentation/arm/SA1100) I found a promising chapter : -------- snipp --------------- Intel SA-1100 Evaluation Platform (http://developer.intel.com/design/strong/) Brutus I/O Pins and Chipselect Register pcmcia-brutus.c, by Ivo Clarysse (What's the official reference for this info?) ..... -------- snipp --------------- Well, I didn't find the mentioned pcmcia-brutus.c file in the kernel source tree. What patch am I missing / where can I get this file and maybe more PCMCIA stuff ? My kernel version is V2.4.0-test5-rmk1-np1. regards Thomas ------------------------------- Thomas RITTER Phone: ++43-51707/35127 SIEMENS AG PRO CDA3 Fax : ++43-51707/55169 Erdbergerlaende 26 mailto://thomas.ritter@siemens.at A - 1030 VIENNA http://wwwpro.siemens.at ---------- PGP Fingerprint ---- 1795 DD10 124B 2E39 1DCA 51CF D981 2F17 EE65 F15C unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Tue Aug 01 16:26:07 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13JdvE-0003UH-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Tue, 01 Aug 2000 16:25:36 +0100 Received: from [12.38.17.9] (helo=mail.aeptec.com) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13JdvD-0003U8-00 for linux-arm-kernel@lists.arm.linux.org.uk; Tue, 01 Aug 2000 16:25:35 +0100 Received: by MAIL with Internet Mail Service (5.5.2650.21) id ; Tue, 1 Aug 2000 11:16:11 -0400 Message-ID: <32CC5B62AF0BD2119E4C00A0C9663E221F3BF8@MAIL> From: "Sun, Lei" To: "'linux-arm-kernel@lists.arm.linux.org.uk'" Subject: PCMCIA driver compilation! Date: Tue, 1 Aug 2000 11:16:03 -0400 MIME-Version: 1.0 X-Mailer: Internet Mail Service (5.5.2650.21) Content-Type: text/plain; charset="iso-8859-1" Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk Hi all: I am heading to write a device driver for a PCMCIA based data card. I tried to compile serial_cs.c and dummy_cs.c, just to see if I can compile them as module. However I got tons of warning and errors. like this: /home/linuxsrc/linux/include/asm/semaphore.h:17: parse error before `wait_queue /home/arm_backup/arm/lib/gcc-lib/arm-linux/2.95.2/../../../../arm-linux/sys- incl ude/bits/types.h:95: warning: `__NFDBITS' redefined serial_cs.c:249: `KERN_NOTICE' undeclared (first use in this function) serial_cs.c:379: `link' undeclared (first use in this function) bla bla..... My Makefile looks like this: DEBUG=y CC=arm-linux-gcc LD=arm-linux-ld INCLUDEDIR = /home/linuxsrc/linux/include ifeq ($(DEBUG),y) DEBFLAGS = -O -g -DPCMCIA_DEBUG else DEBFLAGS = -O2 endif CFLAGS = -D__kernel__ -DMODULE -Wall -I$(INCLUDEDIR) $(DEBFLAGS) OBJS = serial_cs.o all: $(OBJS) It looks like all the required header are sit in there. How comes it could't find the symbol (or some symbols are redifined as WARNING). Anybody can tell me what's going on , please! thank you! Lei Sun unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Tue Aug 01 16:29:07 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13JdyP-0003XM-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Tue, 01 Aug 2000 16:28:53 +0100 Received: from [192.58.206.2] (helo=crl.dec.com) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13JdyN-0003XC-00 for linux-arm-kernel@lists.arm.linux.org.uk; Tue, 01 Aug 2000 16:28:51 +0100 Received: from ims.crl.dec.com (ims.crl.dec.com [16.11.0.11]) by crl.dec.com (8.8.8/RWD-1.2) with ESMTP id LAA09668; Tue, 1 Aug 2000 11:29:00 -0400 (EDT) Received: by ims.crl.dec.com with Internet Mail Service (5.5.2650.21) id ; Tue, 1 Aug 2000 11:34:30 -0400 Message-ID: From: George France To: "'David Aparicio Campos'" , linux-arm-kernel@lists.arm.linux.org.uk Subject: RE: CATS motherboard & ISA compatible registers Date: Tue, 1 Aug 2000 11:29:04 -0400 MIME-Version: 1.0 X-Mailer: Internet Mail Service (5.5.2650.21) Content-Type: multipart/alternative; boundary="----_=_NextPart_001_01BFFBCD.FB423DE5" Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk This message is in MIME format. Since your mail reader does not understand this format, some or all of this message may not be legible. ------_=_NextPart_001_01BFFBCD.FB423DE5 Content-Type: text/plain; charset="iso-8859-1" Hello David; Why don't you use the boot loader from handhelds.org? It already works on the CATS motherboard. --George George France, france@crl.dec.com Cambridge Research Laboratory, Compaq Computer Corporation One Cambridgecenter MS: CRL Cambridge, MA 02142 USA -----Original Message----- From: David Aparicio Campos [mailto:daparic@tid.es] Sent: Tuesday, August 01, 2000 6:45 AM To: linux-arm-kernel@lists.arm.linux.org.uk Subject: CATS motherboard & ISA compatible registers Hello all: I'm still trying to build a firmware of my own in order to boot Linux from a CATS motherboard. ----------------------------------------------- This one is similar to an EBSA285, but it has some important differences: - Has a SouthBridge, this is: a PCI/ISA bridge, I/O controller, IDE controller, USB and PMU funcionality. - Has a RTC (a Dallas 1687) connected to that SouthBridge. Less important: - Flash is an 256 KB AMD compatible (not an 4MB Intel one like EBSA and NetWinder). - It's an ATX form factor, and has standard PC connectors, like KBD, IDE and so on. A NetWinder machine is closer in design, but it uses Winbond bridges, I couldn't get useful information from "nettrom" code (firmware code of NetWinder machines). ----------------------------------------------- I sent a previous mail about an "IRQ7" problem when I started Linux, something like: eter> call 0x8000 boot no options Linux version 2.2.13-rmk2 (root@lobos) (gcc version 2.95.1 19990816 (release)) #5 Thu Jul 20 07:14:04 CEST 2000 initrd extends beyond end of memory (0xea000c22 > 0xc1000000) - disabling initrd Calibrating delay loop... <3>IRQ LOCK: IRQ7 is locking the system, disabled IRQ LOCK: IRQ7 is locking the system, disabled IRQ LOCK: IRQ7 is locking the system, disabled ... Russell King suggested a problem with PIC controller (IRQ7 is something related to bad inicialization of some ISA-related thing), and Jamey Hicks told me to revise external interrupts from RTC (I'm almost sure that this is the source of the deadly interrupt). Hey guys, thank you both!! I tried to change my code, leaving CATS in a similar state that original firmware does, but this didn't changed anything. (If you wonder why I am working in something that is already done, you have not seen CATS firmware running :-). On the other hand, that is not free sofware). I sweeped throught Linux kernel and tried to find some clue about this problem. What I have seen is that, If I comment some lines in init/main.c (start_kernel)... // sti() // calibrate_delay() ...Linux does all inicialization and tries to mount root FS. So what is happening is that when kernel enables interrupts, timer interrupt floods the system (Russell and Jamey were totally right). Surely, its related with timer driving (CATS architecture uses RTC periodic interrupt to handle tick beats, routed thought ISA IRQ subsystem to the StrongArm). The problem by now is that datasheet for my southbridge is very obscure about "ISA compatible registers", those related to PIC (0x20, 0x21, 0xA0 & 0xA1) and RTC (0x70 and 0x71). I have some doubts about this specific inicialization. Seems like some register inicialization should be removed from Linux kernel and put in firmware/loader code (e.g bus master handling in arch/arm/kernel/hw-ebsa285.c and so on). I'm trying to summarize this, if there is somebody interested... Does anybody have some more information about this ISA-PC-compatible inicialization issue?? TYIA David Aparicio ------_=_NextPart_001_01BFFBCD.FB423DE5 Content-Type: text/html; charset="iso-8859-1"

Hello David;
 
Why don't you use the boot loader from handhelds.org? It already works on the CATS motherboard.
 
--George
 

George France,      france@crl.dec.com
Cambridge Research Laboratory, Compaq Computer Corporation
One Cambridgecenter    MS: CRL
Cambridge, MA 02142 USA

-----Original Message-----
From: David Aparicio Campos [mailto:daparic@tid.es]
Sent: Tuesday, August 01, 2000 6:45 AM
To: linux-arm-kernel@lists.arm.linux.org.uk
Subject: CATS motherboard & ISA compatible registers

Hello all:

I'm still trying to build a firmware of my own in order to boot Linux from a
CATS motherboard.
-----------------------------------------------
This one is similar to an EBSA285, but it has some important differences:

- Has a SouthBridge, this is: a PCI/ISA bridge, I/O controller, IDE controller,
    USB and PMU funcionality.
- Has a RTC (a Dallas 1687) connected to that SouthBridge.

Less important:
- Flash is an 256 KB AMD compatible (not an 4MB Intel one like EBSA
    and NetWinder).
- It's an ATX form factor, and has standard PC connectors, like KBD, IDE
    and so on.

A NetWinder machine is closer in design, but it uses Winbond bridges, I
couldn't get useful information from "nettrom" code (firmware code of
NetWinder machines).
-----------------------------------------------
I sent a previous mail about an "IRQ7" problem when I started Linux,
something like:

eter> call 0x8000
boot no options
Linux version 2.2.13-rmk2 (root@lobos) (gcc version 2.95.1 19990816 (release)) #5 Thu Jul 20 07:14:04
CEST 2000
initrd extends beyond end of memory (0xea000c22 > 0xc1000000) - disabling initrd
Calibrating delay loop... <3>IRQ LOCK: IRQ7 is locking the system, disabled
IRQ LOCK: IRQ7 is locking the system, disabled
IRQ LOCK: IRQ7 is locking the system, disabled
...

Russell King suggested a problem with PIC controller (IRQ7 is something
related to bad inicialization of some ISA-related thing), and Jamey Hicks
told me to revise external interrupts from RTC (I'm almost sure that
this is the source of the deadly interrupt). Hey guys, thank you both!!

I tried to change my code, leaving CATS in a similar state that original
firmware does, but this didn't changed anything.
(If you wonder why I am working in something that is already done, you
have not seen CATS firmware running :-). On the other hand, that is not free
sofware).
I sweeped throught Linux kernel and tried to find some clue about this problem.
What I have seen is that, If I comment some lines in init/main.c (start_kernel)...

// sti()
// calibrate_delay()

...Linux does all inicialization and tries to mount root FS. So what is happening
is that when kernel enables interrupts, timer interrupt floods the system (Russell
and Jamey were totally right). Surely, its related with timer driving (CATS
architecture uses RTC periodic interrupt to handle tick beats, routed thought ISA
IRQ subsystem to the StrongArm).

The problem by now is that datasheet for my southbridge is very obscure about
"ISA compatible registers", those related to PIC (0x20, 0x21, 0xA0 & 0xA1)
and RTC (0x70 and 0x71). I have some doubts about this specific inicialization.

Seems like some register inicialization should be removed from Linux kernel
and put in firmware/loader code (e.g bus master handling in
arch/arm/kernel/hw-ebsa285.c and so on). I'm trying to summarize this, if there
is somebody interested...

Does anybody have some more information about this ISA-PC-compatible
inicialization issue??
TYIA

David Aparicio

------_=_NextPart_001_01BFFBCD.FB423DE5-- unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Tue Aug 01 16:33:40 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13Je2s-0003bW-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Tue, 01 Aug 2000 16:33:30 +0100 Received: from [192.58.206.2] (helo=crl.dec.com) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13Je2r-0003bP-00 for linux-arm-kernel@lists.arm.linux.org.uk; Tue, 01 Aug 2000 16:33:29 +0100 Received: from crlwebmail.crl.dec.com (crlwebmail.crl.dec.com [16.11.0.2]) by crl.dec.com (8.8.8/RWD-1.2) with ESMTP id LAA06955; Tue, 1 Aug 2000 11:33:32 -0400 (EDT) Received: by crlwebmail.crl.dec.com with Internet Mail Service (5.5.2650.21) id ; Tue, 1 Aug 2000 11:38:38 -0400 Message-ID: From: George France To: "'yjchang2k'" , linux-arm-kernel@lists.arm.linux.org.uk Cc: stevenc@dvm.tatung.com.tw Subject: RE: system utilities of ARM File system Date: Tue, 1 Aug 2000 11:33:36 -0400 MIME-Version: 1.0 X-Mailer: Internet Mail Service (5.5.2650.21) Content-Type: text/plain; charset="iso-8859-1" Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk You can get ramdisk and the fileutils rpms from http://handhelds.org. The fileutils are at: ftp://ftp.handhelds.org/pub/linux/arm/carleton-winder-rpms There are many ramdisk at: ftp://ftp.handhelds.org/pub/linux/compaq/ipaq/v0.?? Have fun, --George George France, france@crl.dec.com Cambridge Research Laboratory, Compaq Computer Corporation One Cambridgecenter MS: CRL Cambridge, MA 02142 USA > unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Tue Aug 01 16:34:32 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13Je3r-0003ce-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Tue, 01 Aug 2000 16:34:31 +0100 Received: from [193.145.240.2] (helo=tid.tid.es) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13Je3q-0003cU-00 for linux-arm-kernel@lists.arm.linux.org.uk; Tue, 01 Aug 2000 16:34:30 +0100 Received: from tid.es ([172.17.1.9]) by tid.tid.es (Netscape Messaging Server 4.15) with ESMTP id FYMDXQ02.770 for ; Tue, 1 Aug 2000 17:34:38 +0200 Message-ID: <3986ECE7.AE288008@tid.es> Date: Tue, 01 Aug 2000 17:29:43 +0200 From: David Aparicio Campos X-Mailer: Mozilla 4.7 [en] (Win98; I) X-Accept-Language: en MIME-Version: 1.0 To: linux-arm-kernel@lists.arm.linux.org.uk Subject: Re: CATS motherboard & ISA compatible registers References: <3986AA12.5FCC008B@tid.es> Content-Type: multipart/mixed; boundary="------------B5AD4726CA8B0C6D9BF230F6" Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk This is a multi-part message in MIME format. --------------B5AD4726CA8B0C6D9BF230F6 Content-Type: multipart/alternative; boundary="------------1BE50F80159A7D4C871FECED" --------------1BE50F80159A7D4C871FECED Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Hello all again: I did some more tricks. BTW, I'm working with kernel 2.2.13. Yes, I know of 2.3.x and 2.4.0-preX, but they are so big to me, that I prefer to do some training with this smaller version (oh, nice 1.0.9 :-)). What I did was to modify: include/asm/arch/irq.h, function "irq_init_irq": Force "isa_irq" to -1, for a MACH_TYPE_CATS (so PIC is not used) include/asm/arch/time.h, function "setup_timer": Changed to use IRQ_TIMER1 instead of IRQ_ISA_TIMER (because ISA IRQ system is now disabled). Now Linux says: boot no options Linux version 2.2.13-rmk2 (root@lobos) (gcc version 2.95.1 19990816 (release)) #16 Tue Aug 1 15:52:00 CEST 2000 Usando TIMER1... Calibrating delay loop... 214.63 BogoMIPS Memory: 31788k/32M available (476k code, 20k reserved, 480k data, 4k init) POSIX conformance testing by UNIFIX PCI: Probing PCI hardware PCI: device 00:a0 has unknown irq line b Linux NET4.0 for Linux 2.2 Based upon Swansea University Computer Society NET3.039 Starting kswapd v 1.5 Real Time Clock Driver v1.09 rtc: IRQ 8 is not free. RAM disk driver initialized: 16 RAM disks of 4096K size VFS: Cannot open root device 00:ff Kernel panic: VFS: Unable to mount root fs on 00:ff Sure, I didn't compile any network driver, so couldn't mount NFS, but... ...It worked!!. Two minor problems, most interesting being "PCI device 00:a0 has unknown irq line b", probably because some mistake on FW side, (sig). Some clues? I downloaded some (obscure to me) 8259 datasheets, and it seems like somebody (BIOS?) has to inicialize this to 8086 operating mode, by writing a certain sequence, before proper operation can proceed. Can anybody explain me better this aspect about 8259? I have an ALi M1543 and a OKI MSM82C59A datasheet, but it still confuses me. Thank you! David Aparicio Campos wrote: > Hello all: > > I'm still trying to build a firmware of my own in order to boot Linux > from a > CATS motherboard. > ... > ----------------------------------------------- > I sent a previous mail about an "IRQ7" problem when I started Linux, > something like: > > eter> call 0x8000 > boot no options > Linux version 2.2.13-rmk2 (root@lobos) (gcc version 2.95.1 19990816 > (release)) #5 Thu Jul 20 07:14:04 > CEST 2000 > initrd extends beyond end of memory (0xea000c22 > 0xc1000000) - > disabling initrd > Calibrating delay loop... <3>IRQ LOCK: IRQ7 is locking the system, > disabled > IRQ LOCK: IRQ7 is locking the system, disabled > IRQ LOCK: IRQ7 is locking the system, disabled > ... > > Russell King suggested a problem with PIC controller (IRQ7 is > something > related to bad inicialization of some ISA-related thing), and Jamey > Hicks > told me to revise external interrupts from RTC (I'm almost sure that > this is the source of the deadly interrupt). Hey guys, thank you > both!! > > I tried to change my code, leaving CATS in a similar state that > original > firmware does, but this didn't changed anything. ... --------------1BE50F80159A7D4C871FECED Content-Type: text/html; charset=us-ascii Content-Transfer-Encoding: 7bit  
Hello all again:

I did some more tricks. BTW, I'm working with kernel 2.2.13. Yes, I know
of 2.3.x and 2.4.0-preX, but they are so big to me, that I prefer to do some
training with this smaller version (oh, nice 1.0.9 :-)).
What I did was to modify:

include/asm/arch/irq.h, function "irq_init_irq":
Force "isa_irq" to -1, for a MACH_TYPE_CATS (so PIC is not used)

include/asm/arch/time.h, function "setup_timer":
Changed to use IRQ_TIMER1 instead of IRQ_ISA_TIMER (because ISA
IRQ system is now disabled).

Now Linux says:

boot no options
Linux version 2.2.13-rmk2 (root@lobos) (gcc version 2.95.1 19990816 (release)) #16 Tue Aug 1 15:52:00 CEST 2000
Usando TIMER1...
Calibrating delay loop... 214.63 BogoMIPS
Memory: 31788k/32M available (476k code, 20k reserved, 480k data, 4k init)
POSIX conformance testing by UNIFIX
PCI: Probing PCI hardware
PCI: device 00:a0 has unknown irq line b
Linux NET4.0 for Linux 2.2
Based upon Swansea University Computer Society NET3.039
Starting kswapd v 1.5
Real Time Clock Driver v1.09
rtc: IRQ 8 is not free.
RAM disk driver initialized:  16 RAM disks of 4096K size
VFS: Cannot open root device 00:ff
Kernel panic: VFS: Unable to mount root fs on 00:ff

Sure, I didn't compile any network driver, so couldn't mount NFS, but...
...It worked!!.
Two minor problems, most interesting being "PCI device 00:a0 has unknown
irq line b", probably because some mistake on FW side, (sig). Some clues?

I downloaded some (obscure to me) 8259 datasheets, and it seems like
somebody (BIOS?) has to inicialize this to 8086 operating mode, by writing a
certain sequence, before proper operation can proceed.
Can anybody explain me better this aspect about 8259? I have an ALi M1543
and a OKI MSM82C59A datasheet, but it still confuses me.

Thank you!

David Aparicio Campos wrote:

Hello all:

I'm still trying to build a firmware of my own in order to boot Linux from a
CATS motherboard.
...
-----------------------------------------------
I sent a previous mail about an "IRQ7" problem when I started Linux,
something like:

eter> call 0x8000
boot no options
Linux version 2.2.13-rmk2 (root@lobos) (gcc version 2.95.1 19990816 (release)) #5 Thu Jul 20 07:14:04
CEST 2000
initrd extends beyond end of memory (0xea000c22 > 0xc1000000) - disabling initrd
Calibrating delay loop... <3>IRQ LOCK: IRQ7 is locking the system, disabled
IRQ LOCK: IRQ7 is locking the system, disabled
IRQ LOCK: IRQ7 is locking the system, disabled
...

Russell King suggested a problem with PIC controller (IRQ7 is something
related to bad inicialization of some ISA-related thing), and Jamey Hicks
told me to revise external interrupts from RTC (I'm almost sure that
this is the source of the deadly interrupt). Hey guys, thank you both!!

I tried to change my code, leaving CATS in a similar state that original
firmware does, but this didn't changed anything.

...
  --------------1BE50F80159A7D4C871FECED-- --------------B5AD4726CA8B0C6D9BF230F6 Content-Type: text/x-vcard; charset=us-ascii; name="daparic.vcf" Content-Transfer-Encoding: 7bit Content-Description: Card for David Aparicio Campos Content-Disposition: attachment; filename="daparic.vcf" begin:vcard n:Aparicio Campos;David tel;fax:+34-91-3374502 tel;work:+34-91-3374178 x-mozilla-html:TRUE org:Telefónica I+D;Gerencia 1200 adr:;;Emilio Vargas, 6;Madrid;;28043;Spain version:2.1 email;internet:daparic@tid.es title:R&D Engineer end:vcard --------------B5AD4726CA8B0C6D9BF230F6-- unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Tue Aug 01 18:02:06 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13JfQD-00051J-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Tue, 01 Aug 2000 18:01:41 +0100 Received: from [63.192.134.89] (helo=tsuenwan.causality.com) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13JfQB-00051C-00 for linux-arm-kernel@lists.arm.linux.org.uk; Tue, 01 Aug 2000 18:01:39 +0100 Received: from causality.com (localhost [127.0.0.1]) by tsuenwan.causality.com (8.9.3/8.9.2) with ESMTP id JAA01560; Tue, 1 Aug 2000 09:51:08 -0700 (PDT) (envelope-from neil@causality.com) Message-ID: <3986FFFC.592ABEC@causality.com> Date: Tue, 01 Aug 2000 09:51:08 -0700 From: "Neil A. Carson" Organization: Causality Limited X-Mailer: Mozilla 4.6 [en] (X11; I; FreeBSD 3.2-RELEASE i386) X-Accept-Language: en-GB MIME-Version: 1.0 To: David Aparicio Campos CC: linux-arm-kernel@lists.arm.linux.org.uk Subject: Re: CATS motherboard & ISA compatible registers References: <3986AA12.5FCC008B@tid.es> Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk The latest CATS firmware from Simtec handles Linux booting natively. You should be able to get an upgrade from them. Otherwise you'll probably get into all sorts of problems with gfx cards etc. Phil did the u/g work for you? Neil unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Tue Aug 01 19:12:06 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13JgW2-00067e-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Tue, 01 Aug 2000 19:11:46 +0100 Received: from [12.38.17.9] (helo=mail.aeptec.com) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13JgW1-00067S-00 for linux-arm-kernel@lists.arm.linux.org.uk; Tue, 01 Aug 2000 19:11:45 +0100 Received: by MAIL with Internet Mail Service (5.5.2650.21) id ; Tue, 1 Aug 2000 14:02:20 -0400 Message-ID: <32CC5B62AF0BD2119E4C00A0C9663E221F3BFA@MAIL> From: "Sun, Lei" To: "'linux-arm-kernel@lists.arm.linux.org.uk'" Subject: PCMCIA driver Date: Tue, 1 Aug 2000 14:02:20 -0400 MIME-Version: 1.0 X-Mailer: Internet Mail Service (5.5.2650.21) Content-Type: text/plain; charset="iso-8859-1" Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk Hi: Forget about my question I posted! I forget to define -D__KERNEL__, too stupid! Sorry for the trouble! thanks! Lei Sun unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Wed Aug 02 06:47:23 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13JrM0-0002f3-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Wed, 02 Aug 2000 06:46:08 +0100 Received: from [193.145.240.2] (helo=tid.tid.es) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13JrLu-0002ek-00 for linux-arm-kernel@lists.arm.linux.org.uk; Wed, 02 Aug 2000 06:46:07 +0100 Received: from tid.es ([172.17.1.9]) by tid.tid.es (Netscape Messaging Server 4.15) with ESMTP id FYNHCK02.S8S; Wed, 2 Aug 2000 07:45:56 +0200 Message-ID: <3987B468.B70F0BE1@tid.es> Date: Wed, 02 Aug 2000 07:40:56 +0200 From: David Aparicio Campos X-Mailer: Mozilla 4.7 [en] (Win98; I) X-Accept-Language: en MIME-Version: 1.0 To: "Neil A. Carson" CC: linux-arm-kernel@lists.arm.linux.org.uk Subject: Re: CATS motherboard & ISA compatible registers References: <3986AA12.5FCC008B@tid.es> <3986FFFC.592ABEC@causality.com> Content-Type: multipart/mixed; boundary="------------F81237DF2C51050F637C2F49" Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk This is a multi-part message in MIME format. --------------F81237DF2C51050F637C2F49 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: 8bit "Neil A. Carson" wrote: > The latest CATS firmware from Simtec handles Linux booting natively. You > should be able to get an upgrade from them. Otherwise you'll probably > get into all sorts of problems with gfx cards etc. > > Phil did the u/g work for you? > > Neil I purchased some more boards lately, FW version is 0.5 in those ones... I loaded same kernel that I use in my FW development (but zImage in CATS versus a bzipped Image in my own version), and I don't know what I'm doing wrong... When I try to "boot" it complains "no entry point". When I "load", then "call 0x10000000", it prints some garbage to screen. I keep 38400N8 for serial speed. You know, docs on these boards are not dense, neither clear... They lack a documented command set... awful. Still lost with this, any idea about what I'm doing bad? On the other hand, I succeed when downloaded a special image from Chalice WEB site, but I couldn´t run any compiled on my own. I need also to boot from an AMD PCnet network card, but I read than only NE2000 clones works. I also wrote Chalice about a development license, without response... seems like developing is not so bad idea :-)) And, yes, I can run on problems if I use a VGA card, but is not a point for my requirements, so it doesn't mind... BTW, some other people wrote: >> New firmware for the CATS? Cool! We have a couple of these, but haven't >> run Linux on them because it's such a headache. New firmware for them would >> ... "...Seems I'm not alone in being alone..." ;-) Cheers! --------------F81237DF2C51050F637C2F49 Content-Type: text/x-vcard; charset=us-ascii; name="daparic.vcf" Content-Transfer-Encoding: 7bit Content-Description: Card for David Aparicio Campos Content-Disposition: attachment; filename="daparic.vcf" begin:vcard n:Aparicio Campos;David tel;fax:+34-91-3374502 tel;work:+34-91-3374178 x-mozilla-html:TRUE org:Telefónica I+D;Gerencia 1200 adr:;;Emilio Vargas, 6;Madrid;;28043;Spain version:2.1 email;internet:daparic@tid.es title:R&D Engineer end:vcard --------------F81237DF2C51050F637C2F49-- unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Wed Aug 02 10:07:10 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13JuTa-0005wG-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Wed, 02 Aug 2000 10:06:10 +0100 Received: from [195.64.48.18] (helo=imail.knoware.nl) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13JuTa-0005wA-00 for linux-arm-kernel@lists.arm.linux.org.uk; Wed, 02 Aug 2000 10:06:10 +0100 Received: from mail.knoware.nl (mail.knoware.nl [195.64.48.17]) by imail.knoware.nl (Postfix) with ESMTP id 467DFBD709 for ; Wed, 2 Aug 2000 11:06:10 +0200 (CEST) Received: from avxsrv04.cypres.nl (cypr-45-228.cypres.nl [195.64.45.228]) by mail.knoware.nl (Postfix) with ESMTP id A653BA6CEC for ; Wed, 2 Aug 2000 11:06:08 +0200 (CEST) Received: by avxsrv04.cypres.nl with Internet Mail Service (5.5.2650.21) id <38B7LBVP>; Wed, 2 Aug 2000 11:06:53 +0200 Message-ID: <8BD00E18F88DD3118EC8005004258990DCEB7E@avxsrv04.cypres.nl> From: "Melis, A. (Angelo)" To: "'linux-arm-kernel@lists.arm.linux.org.uk'" Subject: Screen mode change Date: Wed, 2 Aug 2000 11:06:51 +0200 MIME-Version: 1.0 X-Mailer: Internet Mail Service (5.5.2650.21) Content-Type: text/plain; charset="iso-8859-1" Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk Hi, I finally compiled a working kernel (2.4.0-test5-rmk2) for my RISC PC 700 (still testing, but looks stable). Thanks to everyone who replied to my questions. But I noticed that the screenmode is much lower then kernel 2.0.36. Is it possible to set the screenmod higher for the new kernel ? And if so ... how ? Thanks in advance, Angelo Angelo Melis Technical Specialist AINO N.V. email: a.melis@aino.nl unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Wed Aug 02 10:14:59 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13Jubt-00064a-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Wed, 02 Aug 2000 10:14:45 +0100 Received: from [62.253.162.43] (helo=mta03-svc.ntlworld.com) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13Jubt-00064U-00 for linux-arm-kernel@lists.arm.linux.org.uk; Wed, 02 Aug 2000 10:14:45 +0100 Received: from moo ([62.253.144.87]) by mta03-svc.ntlworld.com (InterMail vM.4.01.02.27 201-229-119-110) with SMTP id <20000802091445.QPKF16423.mta03-svc.ntlworld.com@moo> for ; Wed, 2 Aug 2000 10:14:45 +0100 Date: Wed, 02 Aug 2000 10:14:27 +0100 From: Peter Naulls To: linux-arm-kernel@lists.arm.linux.org.uk Subject: Re: Screen mode change Message-ID: <3c6fafe749%peter@erble.freeserve.co.uk> In-Reply-To: <8BD00E18F88DD3118EC8005004258990DCEB7E@avxsrv04.cypres.nl> X-Mailer: Messenger v1.40f for RISC OS MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii X-Posting-Agent: RISC OS Newsbase 0.60m User-Agent: POPstar/2.02 Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk In message <8BD00E18F88DD3118EC8005004258990DCEB7E@avxsrv04.cypres.nl> "Melis, A. (Angelo)" wrote: > I finally compiled a working kernel (2.4.0-test5-rmk2) for my RISC PC 700 > (still testing, but looks stable). Thanks to everyone who replied to my > questions. > > But I noticed that the screenmode is much lower then kernel 2.0.36. Is it > possible to set the screenmod higher for the new kernel ? And if so ... how You want the fbset program - if you're using Debian, then this is packaged. -- ------------------------------------------------------------------------ Peter Naulls - peter@erble.freeserve.co.uk Homepage - http://free.prohosting.com/~chocky/ Java for RISC OS and ARM - http://free.prohosting.com/~chocky/java/ Debian Linux on RiscPCs - http://www.erble.freeserve.co.uk/ ------------------------------------------------------------------------ unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Wed Aug 02 15:24:35 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13JzOX-0002M5-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Wed, 02 Aug 2000 15:21:17 +0100 Received: from [208.242.241.39] (helo=cvsftp.cotw.com) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13JzOV-0002Lu-00 for linux-arm-kernel@lists.arm.linux.org.uk; Wed, 02 Aug 2000 15:21:16 +0100 Received: from cotw.com (IDENT:samcconn@ptecdev2.ptec.com [192.168.10.4]) by cvsftp.cotw.com (8.9.3/8.9.3) with ESMTP id JAA30972; Wed, 2 Aug 2000 09:20:36 -0500 Message-ID: <39882ED0.A3E8A390@cotw.com> Date: Wed, 02 Aug 2000 09:23:12 -0500 From: S A McConnell Reply-To: samcconn@cotw.com X-Mailer: Mozilla 4.7 [en] (X11; I; Linux 2.2.14 i686) X-Accept-Language: en MIME-Version: 1.0 To: Charlie Flynn , linux-arm-kernel@lists.arm.linux.org.uk Subject: Re: Touch Screen Drivers References: <3985F480.549D0D99@cotw.com> <39860529.E0452BF0@ozy.dec.com> <3986CCFD.DC7BA506@cotw.com> <398773D0.D138E40B@ozy.dec.com> Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk Charlie Flynn wrote: > > > > - Where in the kernel should the device driver reside? (/drivers/char?) > > There are different classes of devices > character > block > sound > network > Each is kept in it's own class directory > /drivers/char > /drivers/block > /drivers/sound > /drivers/net > There appears to be a sub class to /drivers/char called "misc_device" Mouse/pointing devices appear to be located (See /drivers/char/misc.c and /include/linux/miscdevice.h) They have a different mechanism for registering the file ops. The MIPS folks used this "API?" over the "char API". misc_register(&vr41xx_tpanel); -vs- register_chrdev(0,"sa1100-ts",&sa1100_ts_fops); Has that interface been replaced? I think we need to ask what is a touch screen? -It appears to me to be a pointing device closely related to mice. As such shouldn't we use the misc_device approach? (This is where you find major 10 devices.) If it does not matter shouldn't we stop using one of them to gain more common set of device driver API's? BTW, do not forget about the hidden API/grammer in the Read/Write routines. When we read the device the data it returns has some meaning. The meaning of that data needs to be well specified. (The API appears to be doing that) Applications (X, Microwindows,...) ================================================== Module/non-Module) API | File Ops API | ioctl API --------------------------------------------------- Serial Interface Bus -------------------------------------- UCB1200 (Touch Screen controller) | File ops API Touch Screen | Touch Screen filtering | ------------ | ioctl API Touch Screen: -Knows how to read an write the registers and knows how to initialize the hardware. Touch Screen Filtering: -Knows how to apply filtering algorithms, has an API to provide the cooked data, might have an API that allows the parameters used by the filtering algorithm to be changed. File Ops API: - How applications get access to the data. ioctl API: (I may not have a valid understanding of this) -Methods to control the settings of the filters and hardware We should try to hammer out the layers then move fuctionality up and down those layers until we reach some level of agreement ;-). For instance, if we find most applications are performing the same fuctionality that implies an additional layer. Scott unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Wed Aug 02 18:48:36 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13K2be-0005lU-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Wed, 02 Aug 2000 18:47:02 +0100 Received: from [192.58.206.2] (helo=crl.dec.com) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13K2bd-0005lF-00 for linux-arm-kernel@lists.arm.linux.org.uk; Wed, 02 Aug 2000 18:47:01 +0100 Received: from crlwebmail.crl.dec.com (crlwebmail.crl.dec.com [16.11.0.2]) by crl.dec.com (8.8.8/RWD-1.2) with ESMTP id NAA00037; Wed, 2 Aug 2000 13:46:59 -0400 (EDT) Received: by crlwebmail.crl.dec.com with Internet Mail Service (5.5.2650.21) id ; Wed, 2 Aug 2000 13:52:06 -0400 Message-ID: From: George France To: "'samcconn@cotw.com'" , Charlie Flynn , linux-arm-kernel@lists.arm.linux.org.uk Subject: RE: Touch Screen Drivers Date: Wed, 2 Aug 2000 13:47:03 -0400 MIME-Version: 1.0 X-Mailer: Internet Mail Service (5.5.2650.21) Content-Type: text/plain; charset="iso-8859-1" Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk > > I think we need to ask what is a touch screen? > -It appears to me to be a pointing device closely related to mice. As > such shouldn't we use the misc_device approach? (This is > where you find > major 10 devices.) > The touch screen is not like a mouse. Please see the either the arm-linux archives or the handhelds.org mail archives, where this has been extensively discussed. --George George France, france@crl.dec.com Cambridge Research Laboratory, Compaq Computer Corporation One Cambridgecenter MS: CRL Cambridge, MA 02142 USA unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Wed Aug 02 20:49:03 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13K4Uu-0008NI-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Wed, 02 Aug 2000 20:48:12 +0100 Received: from [216.18.11.193] (helo=duke.electric.net) by www.linux.org.uk with smtp (Exim 3.13 #1) id 13K4Us-0008N9-00 for linux-arm-kernel@lists.arm.linux.org.uk; Wed, 02 Aug 2000 20:48:11 +0100 Received: (qmail 10822 invoked from network); 2 Aug 2000 19:48:06 -0000 Received: from localhost (HELO relay.electric.net) (@127.0.0.1) by duke.electric.net with SMTP; 2 Aug 2000 19:48:06 -0000 Received: (from smtpd@localhost) by relay.electric.net (8.8.8/8.8.6) id MAA10774 for linux-arm-kernel@lists.arm.linux.org.uk; Wed, 2 Aug 2000 12:48:06 -0700 (PDT) Received: from cobra.intrinsyc.com(64.16.10.158), claiming to be "dp.intrinsyc.com" via SMTP by relay.electric.net, id smtpdI10759; Wed Aug 2 12:48:05 2000 Received: from ileong (helo=localhost) by dp.intrinsyc.com with local-esmtp (Exim 3.13 #1 (Debian)) id 13K4Um-0000Mi-00 for ; Wed, 02 Aug 2000 12:48:04 -0700 Date: Wed, 2 Aug 2000 12:48:04 -0700 (PDT) From: I-Gene Leong To: linux-arm-kernel@lists.arm.linux.org.uk Subject: SA1100 UCB1200 audio drivers? Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk I see some in the kernel (arch/arm/special), but they don't seem to run as modules, and they don't ever seem to be included in the kernel. Are they deprecated, and if not, how do I use them? Thanks. -- I-Gene Leong ileong@intrinsyc.com / 604-646-6962 "We see Linux as being much of a fad as the Internet was in 1995." - Irving Wladawsky-Berger, IBM VP of technology and strategy unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Wed Aug 02 21:08:08 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13K4nm-0000Iq-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Wed, 02 Aug 2000 21:07:42 +0100 Received: from [212.250.233.16] (helo=caramon.arm.linux.org.uk ident=root) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13K4nj-0000IY-00 for linux-arm-kernel@lists.arm.linux.org.uk; Wed, 02 Aug 2000 21:07:39 +0100 Received: from flint.arm.linux.org.uk (root@flint [192.168.0.4]) by caramon.arm.linux.org.uk (8.9.3/8.9.3) with ESMTP id VAA09070 for ; Wed, 2 Aug 2000 21:07:38 +0100 Received: (from linux@localhost) by flint.arm.linux.org.uk (8.9.3/8.9.3) id VAA26376 for linux-arm-kernel@lists.arm.linux.org.uk; Wed, 2 Aug 2000 21:05:40 +0100 From: Russell King - ARM Linux Admin Message-Id: <200008022005.VAA26376@flint.arm.linux.org.uk> Subject: Important - Is someone selling your hard work out? To: linux-arm-kernel@lists.arm.linux.org.uk Date: Wed, 2 Aug 2000 21:05:40 +0100 (BST) X-Mailer: ELM [version 2.5 PL1] MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk Referencing: http://www.flatpanels.com/LinuxPR.html The above web address contains a page which is about the release of a product. Specifically, I wish to draw your attention to the following paragraph containing: The ARM port was developed by members of the SA-1100 workgroup of the Open Source Developers Community, and the derived kernel runs on a number of different ARM-based hardware platforms. Everyone who has contributed to non-SA1100 ports of Linux should see this comment for what it is - a lie. Did the SA1100 guys port Linux to the A5000, RiscPC, EBSA110, EBSA285, NetWinder, CATS, CLPS7110, NEXUSPCI, Integrator, Prospector, Brutus, etc? No. Therefore, this comment is 100% wrong. Maybe some companies who have contributed code to the ARM Linux kernel may wish to take this matter up with AppliedData.Com and have that changed to be the TRUTH (hint hint)? Personally, I would like to see that page removed from the site until these incorrect comments are able to be corrected to everyones satisfaction. _____ |_____| ------------------------------------------------- ---+---+- | | Russell King linux@arm.linux.org.uk --- --- | | | | http://www.arm.linux.org.uk/ / / | | +-+-+ --- -+- / | THE developer of ARM Linux |+| /|\ / | | | --- | +-+-+ ------------------------------------------------- /\\\ | unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Wed Aug 02 21:13:23 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13K4so-0000Np-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Wed, 02 Aug 2000 21:12:54 +0100 Received: from [212.250.233.16] (helo=caramon.arm.linux.org.uk ident=root) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13K4sm-0000Ne-00 for linux-arm-kernel@lists.arm.linux.org.uk; Wed, 02 Aug 2000 21:12:53 +0100 Received: from flint.arm.linux.org.uk (root@flint [192.168.0.4]) by caramon.arm.linux.org.uk (8.9.3/8.9.3) with ESMTP id VAA09109; Wed, 2 Aug 2000 21:12:22 +0100 Received: (from linux@localhost) by flint.arm.linux.org.uk (8.9.3/8.9.3) id VAA26413; Wed, 2 Aug 2000 21:10:24 +0100 From: Russell King - ARM Linux Admin Message-Id: <200008022010.VAA26413@flint.arm.linux.org.uk> Subject: Re: Screen mode change To: a.melis@aino.nl ("Melis, A. (Angelo)") Date: Wed, 2 Aug 2000 21:10:24 +0100 (BST) Cc: linux-arm-kernel@lists.arm.linux.org.uk ('linux-arm-kernel@lists.arm.linux.org.uk') In-Reply-To: <8BD00E18F88DD3118EC8005004258990DCEB7E@avxsrv04.cypres.nl> from "Melis, A. (Angelo)" at Aug 02, 2000 11:06:51 AM X-Mailer: ELM [version 2.5 PL1] MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk "Melis, A. (Angelo)" writes: > I finally compiled a working kernel (2.4.0-test5-rmk2) for my RISC PC 700 > (still testing, but looks stable). Thanks to everyone who replied to my > questions. > > But I noticed that the screenmode is much lower then kernel 2.0.36. Is it > possible to set the screenmod higher for the new kernel ? And if so ... how You can set the boot-time mode by supplying (eg) "video=1024x768@60". However, your monitor must be able to cope with it (and you must tell Linux what your monitor is using "video=mon:hmin:hmax:vmin:vmax" You can change it at run time using the "fbset" utility. _____ |_____| ------------------------------------------------- ---+---+- | | Russell King linux@arm.linux.org.uk --- --- | | | | http://www.arm.linux.org.uk/ / / | | +-+-+ --- -+- / | THE developer of ARM Linux |+| /|\ / | | | --- | +-+-+ ------------------------------------------------- /\\\ | unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Wed Aug 02 21:48:45 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13K5Qw-00014p-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Wed, 02 Aug 2000 21:48:10 +0100 Received: from [130.161.42.1] (helo=duteinh.et.tudelft.nl) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13K5Qv-00014j-00 for linux-arm-kernel@lists.arm.linux.org.uk; Wed, 02 Aug 2000 21:48:09 +0100 Received: from dutein75 by duteinh.et.tudelft.nl (8.9.3/BB-2.0) id WAA05975; Wed, 2 Aug 2000 22:45:01 +0200 (MET DST) Message-Id: <200008022045.WAA05975@duteinh.et.tudelft.nl> From: Erik Mouw To: ileong@intrinsyc.com Cc: linux-arm-kernel@lists.arm.linux.org.uk Subject: Re: SA1100 UCB1200 audio drivers? In-Reply-To: Organization: Eric conspiracy secret labs Date: Wed 02 Aug 2000 22:44:18 +0100 (MDT) X-Face: "!Z|1XjFg!3Iw)BwM^M\k)J5KHiPddz}#/Zw%?$&Q!xqob"OA|})3_@&BTTiQERo?/^g[Z<9P4R8:5C:aqq.cw!hotLc=+u-"+1rq@9$V^8PbOUlodyi_sejt+D`iqkLpx$>.aXl%vnfPIt"LK|/%>94_Zrm8jy?`_[RKt'TuoxUxfk(gMhY(n|DBBl~c$dhO>G%6}R4+DGX3|idBm4oU&UM/XUP4I9aoN{]`ax%B?rgD(I3OJ<62Y6L X-Mailer: XCmail 1.2devel - with PGP support, PGP engine version 0.5 (IRIX64) X-Mailerorigin: http://www.fsai.fh-trier.de/~schmitzj/Xclasses/XCmail/ X-Eric-Conspiracy: There is no conspiracy! X-Loop: erik@it.et.tudelft.nl MIME-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: 8bit Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk On Wed, 2 Aug 2000 12:48:04 -0700 (PDT), I-Gene Leong wrote: > I see some in the kernel (arch/arm/special), but they don't seem to run as > modules, and they don't ever seem to be included in the kernel. Are they > deprecated, and if not, how do I use them? Thanks. I ported the driver from linux-2.2.x to linux-2.3.35 just to get the audio working at all. I only did the basic things to get it compiled, nothing more. The driver needs a major rewrite: - make it modular - remove the old Itsy power management stuff and replace it with the new linux-2.4 power management interface - use DMA instead of polled IO - remove the audio mixing from the kernel (policy belongs to userland) When I have some more time in the future I will update the driver, but I have other things to do as well. It is correct that it isn't included in most SA1100 implementations, it is only enabled for some platforms (Brutus, LART, and Tifon). If you want to enable it for your platform, included it in drivers/char/mem.c. Note that Nicolas Pitre's latest patch forgot to enable it. Get my latest patch against Nico's stuff at http://www.lart.tudelft.nl/lartware/port/diff-2.4.0-test5-rmk1-np1-em1.gz Erik -- J.A.K. (Erik) Mouw, Information and Communication Theory Group, Department of Electrical Engineering, Faculty of Information Technology and Systems, Delft University of Technology, PO BOX 5031, 2600 GA Delft, The Netherlands Phone: +31-15-2783635 Fax: +31-15-2781843 Email: J.A.K.Mouw@its.tudelft.nl WWW: http://www-ict.its.tudelft.nl/~erik/ unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Wed Aug 02 22:03:39 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13K5fW-0001Pz-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Wed, 02 Aug 2000 22:03:14 +0100 Received: from [212.250.233.16] (helo=caramon.arm.linux.org.uk ident=root) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13K5fU-0001Ps-00 for linux-arm-kernel@lists.arm.linux.org.uk; Wed, 02 Aug 2000 22:03:13 +0100 Received: from flint.arm.linux.org.uk (root@flint [192.168.0.4]) by caramon.arm.linux.org.uk (8.9.3/8.9.3) with ESMTP id WAA09405 for ; Wed, 2 Aug 2000 22:03:12 +0100 Received: (from linux@localhost) by flint.arm.linux.org.uk (8.9.3/8.9.3) id WAA26671 for linux-arm-kernel@lists.arm.linux.org.uk; Wed, 2 Aug 2000 22:01:14 +0100 From: Russell King - ARM Linux Admin Message-Id: <200008022101.WAA26671@flint.arm.linux.org.uk> Subject: Re: Important - Is someone selling your hard work out? To: linux-arm-kernel@lists.arm.linux.org.uk Date: Wed, 2 Aug 2000 22:01:14 +0100 (BST) In-Reply-To: <200008022005.VAA26376@flint.arm.linux.org.uk> from "Russell King - ARM Linux Admin" at Aug 02, 2000 09:05:40 PM X-Mailer: ELM [version 2.5 PL1] MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk Russell King - ARM Linux Admin writes: > The ARM port was developed by members of the SA-1100 workgroup of > the Open Source Developers Community, and the derived kernel runs > on a number of different ARM-based hardware platforms. Ok, they are going to update the page to reflect the true situation. _____ |_____| ------------------------------------------------- ---+---+- | | Russell King linux@arm.linux.org.uk --- --- | | | | http://www.arm.linux.org.uk/ / / | | +-+-+ --- -+- / | THE developer of ARM Linux |+| /|\ / | | | --- | +-+-+ ------------------------------------------------- /\\\ | unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Wed Aug 02 23:26:18 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13K6xF-00032r-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Wed, 02 Aug 2000 23:25:37 +0100 Received: from [192.58.206.2] (helo=crl.dec.com) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13K6xE-00032j-00 for linux-arm-kernel@lists.arm.linux.org.uk; Wed, 02 Aug 2000 23:25:36 +0100 Received: from ims.crl.dec.com (ims.crl.dec.com [16.11.0.11]) by crl.dec.com (8.8.8/RWD-1.2) with ESMTP id SAA06257; Wed, 2 Aug 2000 18:25:34 -0400 (EDT) Received: by ims.crl.dec.com with Internet Mail Service (5.5.2650.21) id ; Wed, 2 Aug 2000 18:31:05 -0400 Message-ID: From: George France To: "'Russell King - ARM Linux Admin'" , linux-arm-kernel@lists.arm.linux.org.uk Subject: RE: Important - Is someone selling your hard work out? Date: Wed, 2 Aug 2000 18:25:38 -0400 MIME-Version: 1.0 X-Mailer: Internet Mail Service (5.5.2650.21) Content-Type: text/plain; charset="iso-8859-1" Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk Thank you Russell. Well Done. --George George France, france@crl.dec.com Cambridge Research Laboratory, Compaq Computer Corporation One Cambridgecenter MS: CRL Cambridge, MA 02142 USA > -----Original Message----- > From: Russell King - ARM Linux Admin [mailto:linux@arm.linux.org.uk] > Sent: Wednesday, August 02, 2000 5:01 PM > To: linux-arm-kernel@lists.arm.linux.org.uk > Subject: Re: Important - Is someone selling your hard work out? > > > Russell King - ARM Linux Admin writes: > > The ARM port was developed by members of the SA-1100 workgroup of > > the Open Source Developers Community, and the derived kernel runs > > on a number of different ARM-based hardware platforms. > > Ok, they are going to update the page to reflect the true situation. > _____ > |_____| ------------------------------------------------- ---+---+- > | | Russell King linux@arm.linux.org.uk --- --- > | | | | http://www.arm.linux.org.uk/ / / | > | +-+-+ --- -+- > / | THE developer of ARM Linux |+| /|\ > / | | | --- | > +-+-+ ------------------------------------------------- /\\\ | > > > unsubscribe: body of `unsubscribe linux-arm-kernel' to > majordomo@lists.arm.linux.org.uk > unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Thu Aug 03 14:25:27 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13KKx8-0003rP-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Thu, 03 Aug 2000 14:22:26 +0100 Received: from [203.73.94.205] (helo=sinamail.com) by www.linux.org.uk with smtp (Exim 3.13 #1) id 13KKx6-0003rH-00 for linux-arm-kernel@lists.arm.linux.org.uk; Thu, 03 Aug 2000 14:22:25 +0100 Received: (qmail 5752 invoked by uid 99); 3 Aug 2000 00:35:22 -0000 Message-ID: <20000803003522.5751.qmail@sinamail.com> X-Mailer: SinaMail 3.0Beta (FireToad) From: yjchang2k To: George@sinamail.com, France@sinamail.com, linux-arm-kernel@lists.arm.linux.org.uk Subject: Date: Thu, 03 Aug 2000 00:35:22 +0800 Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk Hi, all: May somebody tell me the source code location for system utilities, for examples the bash, tar, cardinfo,...etc. Thanks a lot! Steve ______________________________________ ********************************* Steven Chang ********************************* =================================================================== ·Q¾Ö¦³¥v¤W¶W±jªº50MB¶W¤jEmail? http://www.sinamail.com ½Ð¤W·s®öºôhttp://www.sina.com.tw unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Thu Aug 03 14:25:44 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13KKyp-0003sx-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Thu, 03 Aug 2000 14:24:11 +0100 Received: from [203.73.94.205] (helo=sinamail.com) by www.linux.org.uk with smtp (Exim 3.13 #1) id 13KKyp-0003sp-00 for linux-arm-kernel@lists.arm.linux.org.uk; Thu, 03 Aug 2000 14:24:11 +0100 Received: (qmail 19476 invoked by uid 99); 3 Aug 2000 00:36:50 -0000 Message-ID: <20000803003650.19475.qmail@sinamail.com> X-Mailer: SinaMail 3.0Beta (FireToad) From: yjchang2k To: linux-arm-kernel@lists.arm.linux.org.uk Cc: stevenc@dvm.tatung.com.tw Subject: system utilities of ARM File system Date: Thu, 03 Aug 2000 00:36:49 +0800 Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk Dear all: Do somebody know where to get the source code of the utilities program ( ls, cat vi ... etc). Then build the image like the ramdisk for SA1110 or SA1100. Regards, ______________________________________ =================================================================== ·Q¾Ö¦³¥v¤W¶W±jªº50MB¶W¤jEmail? http://www.sinamail.com ½Ð¤W·s®öºôhttp://www.sina.com.tw =================================================================== ·Q¾Ö¦³¥v¤W¶W±jªº50MB¶W¤jEmail? http://www.sinamail.com ½Ð¤W·s®öºôhttp://www.sina.com.tw unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Fri Aug 04 01:38:15 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13KVTM-0008Ti-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Fri, 04 Aug 2000 01:36:24 +0100 Received: from [158.152.145.93] (helo=willow.armlinux.org) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13KVTL-0008TY-00 for linux-arm-kernel@lists.arm.linux.org.uk; Fri, 04 Aug 2000 01:36:23 +0100 Received: from localhost (localhost [127.0.0.1]) by willow.armlinux.org (Postfix) with SMTP id 8F030FC8BF for ; Thu, 3 Aug 2000 23:20:33 +0100 (BST) Date: Thu, 3 Aug 2000 23:20:32 +0100 From: Chris Rutter To: linux-arm-kernel@lists.arm.linux.org.uk Subject: Non-portability in kernel Message-ID: <20000803232032.A3541@willow> Reply-To: chris@fluff.org Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 X-Mailer: Balsa 0.8.0 Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk Um, does no one care that code like this: struct disc_record { unsigned char log2secsize; ... unsigned short zone_spare; unsigned long root; unsigned long disc_size; unsigned short disc_id; unsigned char disc_name[10]; ... isn't portable to anything which doesn't have identical type sizes, endianness and packing to gcc 2.95-arml? It is assumed no one will want to ever mount a FileCore disc in anything but Acorn hardware? c. unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Fri Aug 04 04:03:43 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13KXlC-0002ZX-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Fri, 04 Aug 2000 04:02:58 +0100 Received: from [140.96.158.1] (helo=extmx.itri.org.tw) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13KXl3-0002Xc-00 for linux-arm-kernel@lists.arm.linux.org.uk; Fri, 04 Aug 2000 04:02:52 +0100 Received: from nti.itri.org.tw (nti.itri.org.tw [140.96.157.2]) by extmx.itri.org.tw (8.8.8/8.8.8) with ESMTP id LAA14363 for ; Fri, 4 Aug 2000 11:04:39 +0800 (CST) Received: from itri.org.tw (pc120163.ccl.itri.org.tw [140.96.120.163]) by nti.itri.org.tw (8.8.8/8.8.8) with ESMTP id LAA07517 for ; Fri, 4 Aug 2000 11:02:22 +0800 (CST) Message-ID: <398A312D.FFF9243A@itri.org.tw> Date: Fri, 04 Aug 2000 10:57:49 +0800 From: Yung-Li Ji (=?iso-8859-1?Q?=A9u=A5=C3=A5=DF?=) Organization: ITRI X-Mailer: Mozilla 4.7 [en] (WinNT; I) X-Accept-Language: zh,zh-TW,en MIME-Version: 1.0 To: linux-arm-kernel@lists.arm.linux.org.uk Subject: Re: system utilities of ARM File system References: Content-Type: multipart/mixed; boundary="------------50AC129D37473750E6AA3887" Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk This is a multi-part message in MIME format. --------------50AC129D37473750E6AA3887 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: quoted-printable X-MIME-Autoconverted: from 8bit to quoted-printable by extmx.itri.org.tw id LAA14363 There are two places where you can go. First, there are almost everythi= ng in www.redhat.com. At least, I have gotten the source code of telnet, ls,= cp in linux. Second, ftp://sunsite.unc.edu/pub/Linux/utils/ is another famous source code ftp site. Joe yjchang2k wrote: > Dear all: > Do somebody know where to get the source code of the utilities prog= ram > ( ls, cat vi ... etc). Then build the image like the ramdisk for SA1110= or > SA1100. > > Regards, > > ______________________________________ > > =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D > =B7Q=BE=D6=A6=B3=A5v=A4W=B6W=B1j=AA=BA50MB=B6W=A4jEmail? http://www.sin= amail.com > =BD=D0=A4W=B7s=AE=F6=BA=F4http://www.sina.com.tw > > =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D > =B7Q=BE=D6=A6=B3=A5v=A4W=B6W=B1j=AA=BA50MB=B6W=A4jEmail? http://www.sin= amail.com > =BD=D0=A4W=B7s=AE=F6=BA=F4http://www.sina.com.tw > > unsubscribe: body of `unsubscribe linux-arm-kernel' to > majordomo@lists.arm.linux.org.uk --------------50AC129D37473750E6AA3887 Content-Type: text/x-vcard; charset=us-ascii; name="joe.vcf" Content-Description: Card for Yung-Li Ji (©u¥Ã¥ß) Content-Disposition: attachment; filename="joe.vcf" Content-Transfer-Encoding: quoted-printable X-MIME-Autoconverted: from 8bit to quoted-printable by extmx.itri.org.tw id LAA14363 begin:vcard=20 n:Ji;Joe tel;fax:03-5820042 tel;work:03-5917381 x-mozilla-html:FALSE org:Computer Systems Development Dept. (X100);Computer & Communications R= esearch Laboratories, ITRI version:2.1 email;internet:joe@itri.org.tw title;quoted-printable:=3DA4u=3DAC=3DE3=3DB0|=3DB9q=3DB3q=3DA9=3DD2=3DB9q= =3DB8=3DA3=3DA8t=3DB2=3DCE=3DB6}=3DB5o=3DB3=3DA1 =3DB3]=3DADp=3DA4u=3DB5{= =3DAEv adr;quoted-printable:;;=3DB7s=3DA6=3DCB=3DBF=3DA4=3DA6=3DCB=3DAAF=3DC2=3D= ED=3DA4=3DA4=3DBF=3DB3=3DB8=3DF44=3DACq195 - 11=3DB8=3DB951=3DC0] X100=3D= 0D=3D0A=3D3B X100 CCL/ITRI Bldg.51, 195-11 Sec.4, =3D0D=3D0AChung Hsing R= d. Chutung, Hsinchu, =3D0D=3D0ATaiwan 310 R.O.C.;;;; fn:Yung-Li Ji(=A9u=A5=C3=A5=DF)=20 end:vcard --------------50AC129D37473750E6AA3887-- unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Fri Aug 04 06:47:30 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13KaJk-0005MP-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Fri, 04 Aug 2000 06:46:48 +0100 Received: from [203.34.97.3] (helo=mail.ocs.com.au) by www.linux.org.uk with smtp (Exim 3.13 #1) id 13KaJh-0005LZ-00 for linux-arm-kernel@lists.arm.linux.org.uk; Fri, 04 Aug 2000 06:46:46 +0100 Received: (qmail 12265 invoked from network); 4 Aug 2000 05:46:40 -0000 Received: from ocs3.ocs-net (192.168.255.3) by mail.ocs.com.au with SMTP; 4 Aug 2000 05:46:40 -0000 X-Mailer: exmh version 2.1.1 10/15/1999 From: Keith Owens To: linux-arm-kernel@lists.arm.linux.org.uk, linuxppc-dev@lists.linuxppc.org, linux-mips@fnet.fr, sparclinux@vger.rutgers.edu, ultralinux@vger.rutgers.edu, linux-alpha@vger.rutgers.edu, linux-kernel@vger.rutgers.edu, linux-ia64@linuxia64.org, linux-vm@vm.marist.edu Subject: [RFC] 2.4.0-test6-pre2 Merge softirq, local_irq_count, local_bh_count Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Date: Fri, 04 Aug 2000 15:46:39 +1000 Message-ID: <7892.965367999@ocs3.ocs-net> Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk This patch hits every arch. It is being cross mailed to every arch mailing list that has not already confirmed that the patch works. Please trim replies to the relevant mailing list. Also please cc: kaos@ocs.com.au on replies, I am not on every list. --- At the moment (2.4.0-test6-pre2) all architectures have a common definition for softirq_state, it occupies its own cache line with a lot of unused padding. The definitions for local_irq_count and local_bh_count are all over the place, some architectures use a cache aligned structure, some use arrays of integers (lets you play cache ping/pong), some even use atomic_t. Each arch defines its own set of mapping macros to get to local_irq_count and local_bh_count. To add insult to injury, not all architectures use the mapping macros, there are bits of code scattered around that use hard coded array lookups which are not optimized for SMP vs non-SMP. ftp://ftp.ocs.com.au/pub/2.4.0-test6-pre2-softirq-bh-merge.gz is a merge of softirq, local_irq_count and local_bh_count into one cache aligned structure. The aim was :- * Put all these fields in a per-cpu cache line. They are hit by the same code paths and are only ever updated from one cpu. Softirq was already cache aligned but the other fields were not, in most cases we were using multiple cache lines for this data. On most archs this will save a cache line. In the worst case we use the same number of lines but still get a code cleanup. * Replace all the explicit references like __local_irq_count[cpu] with wrapper macros. Some arch's had already done this but there was quite a bit of crud left. * Replace multiple arch definitions and export of irq_stat with a common one, optimized for SMP or non-SMP. I know that gcc should do this, but proving it will always optimize was too messy so I went for the safe case and hand optimized the definitions. Besides, s390 is completely different (the story of my life ;). * Create a standard definition for the wrappers that can be used by almost all architectures. Only s390 and sparc64 are special cases but every arch had its own slightly incompatible definition. * Add a per-cpu syscall_count, ready for performance reporting on soft interrupts as well as hard interrupts. This field is not being set at the moment but will be easy to update once the above patch is accepted. I have updated all architectures, but have only been able to test this patch on ix86 and IA64. Could the other arch maintainers try this patch, it is almost guaranteed to contain some Assembler errors. If your arch supports SMP, please try both UP and SMP. I want to know if the patch works or not on each arch so I can send the patch to Linus. So I would appreciate feedback on tests, whether successful or not. m68k and super-h have already replied, nobody else has confirmed one way or the other. All architectures except s390 and sparc64 use the same definitions, see include/linux/irq_cpustat.h. For all but s390, sparc64 and m68k, please check the Assembler changes, the common code should be fine. S390 No irq_stat structure, the equivalent fields are stored in S390_lowcore. That moved cpu_data, ipl_device and the SMP info area up by 16 bytes. There was no need for local_bh_count and local_irq_count to be atomic. Please check the changes to __LC_ in lowcore.h. Also the use of lowcore instead of irq_stat[NR_CPUS] means that asm-s390/hardirq.h is quite different from the other archs, it has its own unique definitions for all the wrapper macros. sparc64 For UP, it uses the common code, local_irq_count is an int in irq_stat. For SMP, local_irq_count is a brlock in cpu_data. asm-sparc64/hardirq.h defines a special SMP mapping for local_irq_count. m68k The change to arch/m68k/atari/ataints.c is a complete guess. There were no examples of how to reference SYMBOL+8 from asm in atari. Files patched: arch/alpha/kernel/alpha_ksyms.c arch/alpha/kernel/entry.S arch/alpha/kernel/irq_alpha.c arch/alpha/kernel/irq_smp.c arch/alpha/kernel/smp.c arch/arm/kernel/armksyms.c arch/arm/kernel/entry-common.S arch/arm/kernel/irq.c arch/i386/kernel/entry.S arch/i386/kernel/i386_ksyms.c arch/i386/kernel/io_apic.c arch/i386/kernel/irq.c arch/i386/kernel/traps.c arch/ia64/kernel/entry.S arch/ia64/kernel/ia64_ksyms.c arch/ia64/kernel/irq.c arch/m68k/atari/ataints.c arch/m68k/kernel/entry.S arch/m68k/kernel/ints.c arch/m68k/kernel/m68k_ksyms.c arch/mips/baget/irq.c arch/mips/dec/irq.c arch/mips/kernel/entry.S arch/mips/kernel/irq.c arch/mips/kernel/mips_ksyms.c arch/mips/kernel/scall_o32.S arch/mips/orion/irq.c arch/mips/sgi/kernel/indy_int.c arch/mips64/kernel/entry.S arch/mips64/kernel/scall_64.S arch/mips64/kernel/scall_o32.S arch/mips64/sgi-ip22/ip22-int.c arch/mips64/sgi-ip27/ip27-irq.c arch/ppc/kernel/entry.S arch/ppc/kernel/irq.c arch/ppc/kernel/local_irq.h arch/ppc/kernel/ppc_ksyms.c arch/s390/kernel/entry.S arch/s390/kernel/irq.c arch/s390/kernel/process.c arch/s390/kernel/s390io.c arch/s390/mm/fault.c arch/sh/kernel/entry.S arch/sh/kernel/irq.c arch/sparc/kernel/irq.c arch/sparc/kernel/rtrap.S arch/sparc/kernel/sparc_ksyms.c arch/sparc64/kernel/irq.c arch/sparc64/kernel/rtrap.S arch/sparc64/kernel/smp.c arch/sparc64/kernel/sparc64_ksyms.c include/asm-alpha/hardirq.h include/asm-alpha/smp.h include/asm-alpha/softirq.h include/asm-arm/hardirq.h include/asm-i386/hardirq.h include/asm-ia64/hardirq.h include/asm-m68k/hardirq.h include/asm-m68k/softirq.h include/asm-m68k/system.h include/asm-mips/hardirq.h include/asm-mips64/hardirq.h include/asm-mips64/processor.h include/asm-mips64/smp.h include/asm-ppc/hardirq.h include/asm-ppc/softirq.h include/asm-s390/hardirq.h include/asm-s390/irq.h include/asm-s390/lowcore.h include/asm-s390/softirq.h include/asm-sh/hardirq.h include/asm-sh/softirq.h include/asm-sparc/hardirq.h include/asm-sparc/irq.h include/asm-sparc/softirq.h include/asm-sparc64/hardirq.h include/asm-sparc64/smp.h include/asm-sparc64/softirq.h include/linux/interrupt.h include/linux/irq_cpustat.h kernel/ksyms.c kernel/sched.c kernel/softirq.c net/netsyms.c unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Fri Aug 04 10:47:15 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13Ke3Z-0000lt-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Fri, 04 Aug 2000 10:46:21 +0100 Received: from [212.1.130.10] (helo=www.uklinux.net ident=root) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13Ke3Y-0000ln-00 for linux-arm-kernel@lists.arm.linux.org.uk; Fri, 04 Aug 2000 10:46:20 +0100 Received: from caramon.arm.linux.org.uk (root@ppp-2-9.cvx2.telinco.net [212.1.141.9]) by www.uklinux.net (8.9.3/8.8.7) with ESMTP id KAA08060; Fri, 4 Aug 2000 10:50:01 +0100 Received: from flint.arm.linux.org.uk (root@flint [192.168.0.4]) by caramon.arm.linux.org.uk (8.9.3/8.9.3) with ESMTP id JAA03439; Fri, 4 Aug 2000 09:36:14 +0100 Received: (from linux@localhost) by flint.arm.linux.org.uk (8.9.3/8.9.3) id JAA03577; Fri, 4 Aug 2000 09:35:35 +0100 From: Russell King - ARM Linux Admin Message-Id: <200008040835.JAA03577@flint.arm.linux.org.uk> Subject: Re: Non-portability in kernel To: chris@fluff.org Date: Fri, 4 Aug 2000 09:35:35 +0100 (BST) Cc: linux-arm-kernel@lists.arm.linux.org.uk In-Reply-To: <20000803232032.A3541@willow> from "Chris Rutter" at Aug 03, 2000 11:20:32 PM X-Mailer: ELM [version 2.5 PL1] MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk Chris Rutter writes: > Um, does no one care that code like this: That's very a harsh way of putting it. Yes, people do care, however no one has reported this specific area as a problem before now, and therefore it hasn't been fixed. What would've been more useful is to have received a patch to correct this. _____ |_____| ------------------------------------------------- ---+---+- | | Russell King linux@arm.linux.org.uk --- --- | | | | http://www.arm.linux.org.uk/ / / | | +-+-+ --- -+- / | THE developer of ARM Linux |+| /|\ / | | | --- | +-+-+ ------------------------------------------------- /\\\ | unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Fri Aug 04 11:22:49 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13Kecd-0001Nw-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Fri, 04 Aug 2000 11:22:35 +0100 Received: from [158.152.145.93] (helo=willow.armlinux.org) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13Kecc-0001No-00 for linux-arm-kernel@lists.arm.linux.org.uk; Fri, 04 Aug 2000 11:22:34 +0100 Received: from localhost (localhost [127.0.0.1]) by willow.armlinux.org (Postfix) with SMTP id 761B7FC8BF for ; Fri, 4 Aug 2000 11:05:49 +0100 (BST) Date: Fri, 4 Aug 2000 11:05:48 +0100 From: Chris Rutter To: linux-arm-kernel@lists.arm.linux.org.uk Subject: Sector size question Message-ID: <20000804110548.D3541@willow> Reply-To: chris@fluff.org Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 X-Mailer: Balsa 0.8.0 Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk [I know this is off-topic for the list; kill me if you want.] Are the sizes of the sectors as in add_gd_partition() the same as what get_ptable_blocksize() returns, or something else? (See `linux/fs/partitions/check.c'.) c. unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Fri Aug 04 11:52:01 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13Kf4s-0001oJ-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Fri, 04 Aug 2000 11:51:46 +0100 Received: from [138.38.32.2] (helo=pat.bath.ac.uk) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13Kf4s-0001o9-00 for linux-arm-kernel@lists.arm.linux.org.uk; Fri, 04 Aug 2000 11:51:46 +0100 Received: from mary.bath.ac.uk ([138.38.32.14] helo=bath.ac.uk ident=mmdf) by pat.bath.ac.uk with smtp (Exim 3.12 #1) id 13Kf4u-0003Ib-00 for linux-arm-kernel@lists.arm.linux.org.uk; Fri, 04 Aug 2000 11:51:48 +0100 Date: Fri, 4 Aug 2000 11:51:47 +0100 (BST) From: James Alford To: linux-arm-kernel@lists.arm.linux.org.uk Subject: SA1100 booting kernel problems. Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk Hi, I am trying to get the Intel StrongARM SA1100 Multimedia Development Board working with a linux kernel. I can get the 2.2.14-rmk5-np17 kernel up but it crashes at "POSIX conformance teting by UNIFIX" This is probably due to the fact that the kernel is built for a Brutus. I have not yet started to edit the arch specfic bits yet. My problem is that with 2.4.0-test5-rmk1-np1. I can't get the kerenl to start. The head stuff starts ok and decompresses the kernel outputing the line "Uncompressing Linux........................ done, botting the kernel." It just stops there it does not seem to step into the uncompressed kernel. I think that the 2.5-test5 head stuff is working as it will boot the 2.2.14 kernel. I would prefer to modify the 2.4 tree to work with the SA1100 MM devel board rather than the 2.2 tree. If anybody could point me in the right direction to sort this out I would be most gratefull. James Alford unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Fri Aug 04 12:03:33 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13KfFu-00024s-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Fri, 04 Aug 2000 12:03:10 +0100 Received: from [130.161.42.1] (helo=duteinh.et.tudelft.nl) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13KfFt-00024l-00 for linux-arm-kernel@lists.arm.linux.org.uk; Fri, 04 Aug 2000 12:03:09 +0100 Received: from dutein75 by duteinh.et.tudelft.nl (8.9.3/BB-2.0) id NAA12816; Fri, 4 Aug 2000 13:02:57 +0200 (MET DST) Message-Id: <200008041102.NAA12816@duteinh.et.tudelft.nl> From: Erik Mouw To: linux@arm.linux.org.uk Cc: chris@fluff.org, linux-arm-kernel@lists.arm.linux.org.uk Subject: Re: Non-portability in kernel In-Reply-To: <200008040835.JAA03577@flint.arm.linux.org.uk> Organization: Eric conspiracy secret labs Date: Fri 04 Aug 2000 13:02:57 +0100 (MDT) X-Face: "!Z|1XjFg!3Iw)BwM^M\k)J5KHiPddz}#/Zw%?$&Q!xqob"OA|})3_@&BTTiQERo?/^g[Z<9P4R8:5C:aqq.cw!hotLc=+u-"+1rq@9$V^8PbOUlodyi_sejt+D`iqkLpx$>.aXl%vnfPIt"LK|/%>94_Zrm8jy?`_[RKt'TuoxUxfk(gMhY(n|DBBl~c$dhO>G%6}R4+DGX3|idBm4oU&UM/XUP4I9aoN{]`ax%B?rgD(I3OJ<62Y6L X-Mailer: XCmail 1.2devel - with PGP support, PGP engine version 0.5 (IRIX64) X-Mailerorigin: http://www.fsai.fh-trier.de/~schmitzj/Xclasses/XCmail/ X-Eric-Conspiracy: There is no conspiracy! X-Loop: erik@it.et.tudelft.nl MIME-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: 8bit Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk On Fri, 4 Aug 2000 09:35:35 +0100 (BST), Russell King - ARM Linux Admin wrote: > Chris Rutter writes: >> Um, does no one care that code like this: > > That's very a harsh way of putting it. Yes, people do care, however > no one has reported this specific area as a problem before now, and > therefore it hasn't been fixed. > > What would've been more useful is to have received a patch to correct > this. Or at least document that it is non-portable and what should be done to make it that way. Erik -- "I'm just this guy you know?" -- Zaphod Beeblebrox in "The Hitchhikers Guide to the Galaxy" by Douglas Adams unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Fri Aug 04 12:41:41 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13Kfr4-0002eE-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Fri, 04 Aug 2000 12:41:34 +0100 Received: from [130.161.42.1] (helo=duteinh.et.tudelft.nl) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13Kfr3-0002e4-00 for linux-arm-kernel@lists.arm.linux.org.uk; Fri, 04 Aug 2000 12:41:33 +0100 Received: from dutein75 by duteinh.et.tudelft.nl (8.9.3/BB-2.0) id NAA12928; Fri, 4 Aug 2000 13:36:27 +0200 (MET DST) Message-Id: <200008041136.NAA12928@duteinh.et.tudelft.nl> From: Erik Mouw To: ee7joa@bath.ac.uk Cc: linux-arm-kernel@lists.arm.linux.org.uk Subject: Re: SA1100 booting kernel problems. In-Reply-To: Organization: Eric conspiracy secret labs Date: Fri 04 Aug 2000 13:36:27 +0100 (MDT) X-Face: "!Z|1XjFg!3Iw)BwM^M\k)J5KHiPddz}#/Zw%?$&Q!xqob"OA|})3_@&BTTiQERo?/^g[Z<9P4R8:5C:aqq.cw!hotLc=+u-"+1rq@9$V^8PbOUlodyi_sejt+D`iqkLpx$>.aXl%vnfPIt"LK|/%>94_Zrm8jy?`_[RKt'TuoxUxfk(gMhY(n|DBBl~c$dhO>G%6}R4+DGX3|idBm4oU&UM/XUP4I9aoN{]`ax%B?rgD(I3OJ<62Y6L X-Mailer: XCmail 1.2devel - with PGP support, PGP engine version 0.5 (IRIX64) X-Mailerorigin: http://www.fsai.fh-trier.de/~schmitzj/Xclasses/XCmail/ X-Eric-Conspiracy: There is no conspiracy! X-Loop: erik@it.et.tudelft.nl MIME-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: 8bit Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk On Fri, 4 Aug 2000 11:51:47 +0100 (BST), James Alford wrote: > I am trying to get the Intel StrongARM SA1100 Multimedia Development Board > working with a linux kernel. That's a Brutus. Are you sure that it is not the SA1110 kit (the Assabet)? > I can get the 2.2.14-rmk5-np17 kernel up but it crashes at > "POSIX conformance teting by UNIFIX" > This is probably due to the fact that the kernel is built for a Brutus. I > have not yet started to edit the arch specfic bits yet. So, you don't have a Brutus, otherwise you wouldn't have to edit the arch specific bits. 2.2.14 has Brutus support, but is obsolete w.r.t. Assabet. > My problem is that with 2.4.0-test5-rmk1-np1. I can't get the kerenl to > start. The head stuff starts ok and decompresses the kernel outputing the > line > > "Uncompressing Linux........................ done, botting the kernel." > > It just stops there it does not seem to step into the uncompressed kernel. Sounds like inproper DRAM timing to me. > I think that the 2.5-test5 head stuff is working as it will boot the > 2.2.14 kernel. > > I would prefer to modify the 2.4 tree to work with the SA1100 MM devel > board rather than the 2.2 tree. > > If anybody could point me in the right direction to sort this out I would > be most gratefull. If you really have an Assabet (SA1110 CPU), have a look at http://www.cs.cmu.edu/~wearable/software/assabet.html Oh, and subscribe to the sa1100-linux mailing list. The LART documentation page has information: http://www.lart.tudelft.nl/doc.php3 Erik PS: note that I don't have a Brutus or Assabet -- J.A.K. (Erik) Mouw, Information and Communication Theory Group, Department of Electrical Engineering, Faculty of Information Technology and Systems, Delft University of Technology, PO BOX 5031, 2600 GA Delft, The Netherlands Phone: +31-15-2783635 Fax: +31-15-2781843 Email: J.A.K.Mouw@its.tudelft.nl WWW: http://www-ict.its.tudelft.nl/~erik/ unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Fri Aug 04 13:10:12 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13KgIQ-00035M-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Fri, 04 Aug 2000 13:09:50 +0100 Received: from [212.1.130.10] (helo=www.uklinux.net ident=root) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13KgIO-00035F-00 for linux-arm-kernel@lists.arm.linux.org.uk; Fri, 04 Aug 2000 13:09:48 +0100 Received: from caramon.arm.linux.org.uk (root@ppp-2-3.cvx6.telinco.net [212.1.135.3]) by www.uklinux.net (8.9.3/8.8.7) with ESMTP id NAA15718; Fri, 4 Aug 2000 13:13:34 +0100 Received: from flint.arm.linux.org.uk (root@flint [192.168.0.4]) by caramon.arm.linux.org.uk (8.9.3/8.9.3) with ESMTP id NAA04292; Fri, 4 Aug 2000 13:09:48 +0100 Received: (from linux@localhost) by flint.arm.linux.org.uk (8.9.3/8.9.3) id NAA07445; Fri, 4 Aug 2000 13:09:09 +0100 From: Russell King - ARM Linux Admin Message-Id: <200008041209.NAA07445@flint.arm.linux.org.uk> Subject: Re: Sector size question To: chris@fluff.org Date: Fri, 4 Aug 2000 13:09:09 +0100 (BST) Cc: linux-arm-kernel@lists.arm.linux.org.uk In-Reply-To: <20000804110548.D3541@willow> from "Chris Rutter" at Aug 04, 2000 11:05:48 AM X-Mailer: ELM [version 2.5 PL1] MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk Chris Rutter writes: > Are the sizes of the sectors as in add_gd_partition() the same as what > get_ptable_blocksize() returns, or something else? They are in units of 512-byte sectors. This is because the block device driver layer works with a lowest common denominator of 512-byte sectors. _____ |_____| ------------------------------------------------- ---+---+- | | Russell King linux@arm.linux.org.uk --- --- | | | | http://www.arm.linux.org.uk/ / / | | +-+-+ --- -+- / | THE developer of ARM Linux |+| /|\ / | | | --- | +-+-+ ------------------------------------------------- /\\\ | unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Fri Aug 04 13:17:29 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13KgPQ-0003By-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Fri, 04 Aug 2000 13:17:04 +0100 Received: from [212.1.130.10] (helo=www.uklinux.net ident=root) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13KgPP-0003Bs-00 for linux-arm-kernel@lists.arm.linux.org.uk; Fri, 04 Aug 2000 13:17:03 +0100 Received: from caramon.arm.linux.org.uk (root@ppp-2-3.cvx6.telinco.net [212.1.135.3]) by www.uklinux.net (8.9.3/8.8.7) with ESMTP id NAA16191; Fri, 4 Aug 2000 13:20:50 +0100 Received: from flint.arm.linux.org.uk (root@flint [192.168.0.4]) by caramon.arm.linux.org.uk (8.9.3/8.9.3) with ESMTP id NAA04319; Fri, 4 Aug 2000 13:17:04 +0100 Received: (from linux@localhost) by flint.arm.linux.org.uk (8.9.3/8.9.3) id NAA07474; Fri, 4 Aug 2000 13:16:25 +0100 From: Russell King - ARM Linux Admin Message-Id: <200008041216.NAA07474@flint.arm.linux.org.uk> Subject: Re: Sector size question To: chris@fluff.org Date: Fri, 4 Aug 2000 13:16:25 +0100 (BST) Cc: linux-arm-kernel@lists.arm.linux.org.uk In-Reply-To: <20000804110548.D3541@willow> from "Chris Rutter" at Aug 04, 2000 11:05:48 AM X-Mailer: ELM [version 2.5 PL1] MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk Chris Rutter writes: > Are the sizes of the sectors as in add_gd_partition() the same as what > get_ptable_blocksize() returns, or something else? I'm assuming that you're seeing a problem around here - could you enlighten us as to what the problem is please? _____ |_____| ------------------------------------------------- ---+---+- | | Russell King linux@arm.linux.org.uk --- --- | | | | http://www.arm.linux.org.uk/ / / | | +-+-+ --- -+- / | THE developer of ARM Linux |+| /|\ / | | | --- | +-+-+ ------------------------------------------------- /\\\ | unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Fri Aug 04 14:05:49 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13KhAO-00043r-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Fri, 04 Aug 2000 14:05:36 +0100 Received: from [193.145.240.2] (helo=tid.tid.es) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13KhAH-00043c-00 for linux-arm-kernel@lists.arm.linux.org.uk; Fri, 04 Aug 2000 14:05:34 +0100 Received: from tid.es ([172.17.1.9]) by tid.tid.es (Netscape Messaging Server 4.15) with ESMTP id FYRR0U00.H02; Fri, 4 Aug 2000 15:05:18 +0200 Message-ID: <398ABE4D.2F39E358@tid.es> Date: Fri, 04 Aug 2000 14:59:57 +0200 From: David Aparicio Campos X-Mailer: Mozilla 4.7 [en] (Win98; I) X-Accept-Language: en MIME-Version: 1.0 To: linux-arm-kernel@lists.arm.linux.org.uk Subject: More IRQ questions References: <200007192019.VAA25864@flint.arm.linux.org.uk> Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk Hello all: I hope you don't hate me for hammering again and again with CATS specific questions... ;-) --------part 1: some tricks Russell King - ARM Linux Admin wrote: > David Aparicio Campos writes: > > IRQ LOCK: IRQ7 is locking the system, disabled > > This is the symptom of a stuck active IRQ line. I'm just guessing, > but its probably related to the Southbridge. Most PC PICs (programable > interrupt controllers) return IRQ7 when they don't know what the source > of the interrupt is (hence why a LOT of applications don't use IRQs with > LPTs by default). I did a work-around for some problems relating to launch Linux with a firmware developed on my own. The broken thing was a IRQ generated by "southbridge" devices (i.e ISA bus and so on). I did some experimental changes (kernel 2.2.13-rmk2) to these files: include/asm/arch-ebsa285/irq.h include/asm/arch-ebsa-285/time.h in order to ignore interrupts from those devices (CATS motherboard becoming more like a standard EBSA285). I use TIMER0 as system tick, like EBSA285, not the ISA one. That way, Linux starts up with no problems... yes, it's not the best solution, but it's a begining... ---------part 2: questions about IRQ mapping Russell King - ARM Linux Admin wrote: > > > Is IRQ7 related to a "PC compatible" architecture > > Yes. IRQs 0 through 15 are as per your normal PC. IRQs 16 upwards refer > to the 21285 IRQs. Check include/asm-arm/arch-ebsa285/irq*.h for more > information about these. > _____ OK. I was somehow confused about this. CATS original firmware does some strange (to me, but any hint is welcomed) things about asigning IRQ numbers. It gives IRQ 1 thru 3 for PCI add-on cards (there are only three slots), and IRQ 128+x for onboard southbridge devices. When Linux starts, it looks those numbers and maps them to 21285 IRQs this way: IRQ 1 -> IRQ_PCI (INTA) number 28 (FootBridge 12) IRQ 2 -> IRQ_IN0 (INTB) number 21 (FootBridge 5) IRQ 3 -> IRQ_IN1 (INTC) number 22 (FootBridge 6) IRQ 4 -> IRQ_IN3 (INTD) where is slot 4?? PCI/ISA bridge?? IRQ128+x -> IRQ 16+x (Footbridge x???) SouthBridge common IRQ pin is routed to IRQ_IN2, so number 23 (FootBridge 7), that is the one related to that "IRQ7 lock" message, isn't it?... I then revised other firmware sources, but no new clues. Original firmware from Chalice doesn't work for me, because I need to load with BOOTP/TFTP from an unsupported NIC. Is anybody there related with Chalice in order to give me information about source code licensing? "support" mailbox seems to be out of order... So, questions: * Why Linux for CATS doen't assign IRQ from onboard devices to ISA range (less than 16)?? For example, USB is 139, so Linux gives IRQ 27 (Footbridge 11). CATS information says something about giving ISA IRQ-11 to USB... That is in "arch/arm/kernel/hw-footbridge.c", function "cats_irqval". * Same source code, question about "IRQ 4". Where is slot 4?? It seems like some mistake when derived from EBSA285 (which has 4 slots). * In order to develop my own firmare, should I use "compatible" IRQ assignment (and resource mapping) or should I propose a new Linux architecture? (First seems good at first, but I have almost no information about original firmware, and reverse engineering is very limited because of write-only registers). A new Linux architecture seems reasonable if more driver support could be added by people (volunteers?) to this firmware, becoming then better than original one. If there is any interest, and as soon as it becomes stable, I will release a snapshot...could evolve to a general EBSA285/CATS one... * There is a problem with message consistency, related to IRQ debugging: 1) arm subsystem prints original IRQ numbers from PCI CSR space (1..4 and 128..144), but 2) main Linux code shows "compatible", mapped numbers (range 0..31) Can we change messages in arch/arm/kernel in order to show same numbers than main Linux development? (maybe two numbers, with both "compatible" and PCI assignments). Thank you for your patience... David Aparicio BTW: I have some problems related with my link to internet, so I couldn't download a 2.4.0-test kernel by the time of writing this, I'm not checked if those questions have already been fixed... I will check and post another mail if I find same problems there... BTW: it's 3rd time I try to post you this :-(( my email server is on fire also... unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Fri Aug 04 17:24:26 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13KkFv-0007Ko-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Fri, 04 Aug 2000 17:23:31 +0100 Received: from [130.161.40.91] (helo=thorgal.et.tudelft.nl) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13KkFv-0007Kd-00 for linux-arm-kernel@lists.arm.linux.org.uk; Fri, 04 Aug 2000 17:23:31 +0100 Received: from [130.161.115.44] (jdbakker.et.tudelft.nl [130.161.115.44]) by thorgal.et.tudelft.nl (8.8.7/8.8.7) with ESMTP id SAA21115; Fri, 4 Aug 2000 18:10:56 +0200 Message-Id: In-Reply-To: <200008041136.NAA12928@duteinh.et.tudelft.nl> References: Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Date: Fri, 4 Aug 2000 18:09:38 +0200 To: ee7joa@bath.ac.uk From: "J.D. Bakker" Subject: Re: SA1100 booting kernel problems. Cc: Erik Mouw , linux-arm-kernel@lists.arm.linux.org.uk Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk >On Fri, 4 Aug 2000 11:51:47 +0100 (BST), James Alford wrote: >> My problem is that with 2.4.0-test5-rmk1-np1. I can't get the kerenl to >> start. The head stuff starts ok and decompresses the kernel outputing the >> line >> >> "Uncompressing Linux........................ done, botting the kernel." >> >> It just stops there it does not seem to step into the uncompressed kernel. The kernel expects to be called with specific values in registers r0 and r1; the correct value for IIRC r1 has changed in 2.3.late. Supplying the wrong values can lead to the symptoms you describe. See linux/arch/arm/kernel/head-armv.S for additional clues. What bootloader are you using ? JDB. -- LART. 250 MIPS under one Watt. Free hardware design files. http://www.lart.tudelft.nl/ unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Fri Aug 04 18:22:14 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13KlAL-0008T2-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Fri, 04 Aug 2000 18:21:49 +0100 Received: from [212.1.130.10] (helo=www.uklinux.net ident=root) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13KlAK-0008Su-00 for linux-arm-kernel@lists.arm.linux.org.uk; Fri, 04 Aug 2000 18:21:48 +0100 Received: from caramon.arm.linux.org.uk (root@ppp-2-104.cvx4.telinco.net [212.1.149.104]) by www.uklinux.net (8.9.3/8.8.7) with ESMTP id SAA02166; Fri, 4 Aug 2000 18:25:36 +0100 Received: from flint.arm.linux.org.uk (root@flint [192.168.0.4]) by caramon.arm.linux.org.uk (8.9.3/8.9.3) with ESMTP id RAA05190; Fri, 4 Aug 2000 17:31:39 +0100 Received: (from linux@localhost) by flint.arm.linux.org.uk (8.9.3/8.9.3) id RAA08879; Fri, 4 Aug 2000 17:31:00 +0100 From: Russell King - ARM Linux Admin Message-Id: <200008041631.RAA08879@flint.arm.linux.org.uk> Subject: Re: More IRQ questions To: daparic@tid.es (David Aparicio Campos) Date: Fri, 4 Aug 2000 17:30:59 +0100 (BST) Cc: linux-arm-kernel@lists.arm.linux.org.uk In-Reply-To: <398ABE4D.2F39E358@tid.es> from "David Aparicio Campos" at Aug 04, 2000 02:59:57 PM X-Mailer: ELM [version 2.5 PL1] MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk David Aparicio Campos writes: > I hope you don't hate me for hammering again and again with CATS > specific questions... ;-) Its fine - its kernel stuff on the kernel list. > SouthBridge common IRQ pin is routed to IRQ_IN2, so number 23 > (FootBridge 7), that is the one related to that "IRQ7 lock" > message, isn't it?... No. "IRQ7 lock" means Linux IRQ7, which translates to ISA IRQ7. Therefore, what I believe is happening is that for whatever reason, the Southbridge is signalling that it has a pending IRQ via, as you say IRQ23. When IRQ23 is detected, we issue a PCI IACK (interrupt acknowledge) cycle and the southbridge places into the PCI bus the IRQ number of the interrupt. At this point, the southbridge says "Hey, I don't have an interrupt pending, so I'm going to send you the value 7". This translates through to IRQ7. > * Why Linux for CATS doen't assign IRQ from onboard devices to ISA > range (less than 16)?? >From your description, the PCI interrupts are not connected to the southbridge, which handles IRQs 0 through 15. > * Same source code, question about "IRQ 4". Where is slot 4?? It seems > like some mistake when derived from EBSA285 (which has 4 slots). Its a harmless mistake. Phil? > * In order to develop my own firmare, should I use "compatible" IRQ > assignment (and resource mapping) or should I propose a new Linux > architecture? (First seems good at first, but I have almost no > information about original firmware, and reverse engineering is very > limited because of write-only registers). A new Linux architecture > seems reasonable if more driver support could be added by people > (volunteers?) to this firmware, becoming then better than original > one. > If there is any interest, and as soon as it becomes stable, I will > release a snapshot...could evolve to a general EBSA285/CATS one... I don't believe this is necessary. > * There is a problem with message consistency, related to IRQ debugging: > 1) arm subsystem prints original IRQ numbers from PCI CSR space (1..4 > and 128..144), but > 2) main Linux code shows "compatible", mapped numbers (range 0..31) > Can we change messages in arch/arm/kernel in order to show same > numbers than main Linux development? (maybe two numbers, with both > "compatible" and PCI assignments). Please explain some more. What/where/how are you getting these IRQ numbers? _____ |_____| ------------------------------------------------- ---+---+- | | Russell King linux@arm.linux.org.uk --- --- | | | | http://www.arm.linux.org.uk/ / / | | +-+-+ --- -+- / | THE developer of ARM Linux |+| /|\ / | | | --- | +-+-+ ------------------------------------------------- /\\\ | unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Sat Aug 05 05:42:09 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13Kvj7-0003qn-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Sat, 05 Aug 2000 05:38:25 +0100 Received: from [64.4.15.84] (helo=hotmail.com) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13Kvj5-0003qL-00 for linux-arm-kernel@lists.arm.linux.org.uk; Sat, 05 Aug 2000 05:38:24 +0100 Received: from mail pickup service by hotmail.com with Microsoft SMTPSVC; Fri, 4 Aug 2000 21:37:57 -0700 Received: from 212.38.224.118 by lw10fd.law10.hotmail.msn.com with HTTP; Sat, 05 Aug 2000 GMT X-Originating-IP: [212.38.224.118] From: "Jan Jenssen" To: linux-arm-kernel@lists.arm.linux.org.uk Subject: pcmcia for older kernels Date: Sat, 05 Aug 2000 04:37:57 GMT Mime-Version: 1.0 Content-Type: text/plain; format=flowed Message-ID: X-OriginalArrivalTime: 05 Aug 2000 04:37:57.0414 (UTC) FILETIME=[ED946C60:01BFFE96] Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk I am NOT using the latest kernel with my Assabet. My current kernel version is 2.3.99-pre6-rmk1-np1. Now I am adding the pcmcia support to it. I have read the http://www.cs.cmu.edu/~wearable/software/pcmcia-arm.html document tryint to find out the correct versions of the Standard Card Service distribution and the patch files. Could you tell me (ask more info if something is missing) are the right components as follows, For the kernel patching: linux-pcmcia-3.patch For the CS patching: pcmcia-cs-3.1.14-arm5.patch & arm6.patch And is the STD CS version 3.1.14 the best for this kernel? ________________________________________________________________________ Get Your Private, Free E-mail from MSN Hotmail at http://www.hotmail.com unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Sat Aug 05 09:16:42 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13Kz75-00072D-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Sat, 05 Aug 2000 09:15:23 +0100 Received: from [203.73.94.214] (helo=sinamail.com) by www.linux.org.uk with smtp (Exim 3.13 #1) id 13Kz73-000726-00 for linux-arm-kernel@lists.arm.linux.org.uk; Sat, 05 Aug 2000 09:15:22 +0100 Received: (qmail 26150 invoked by uid 99); 4 Aug 2000 00:30:10 -0000 Message-ID: <20000804003010.26148.qmail@sinamail.com> X-Mailer: SinaMail 3.0Beta (FireToad) From: yjchang2k To: linux-arm-kernel@lists.arm.linux.org.uk Subject: system utilities of ARM File system Date: Fri, 04 Aug 2000 00:30:10 +0800 Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk Dear all: Do somebody know where to get the source code of the utilities program ( ls, cat vi ... etc). Then build the image like the ramdisk for SA1110 or SA1100. Regards, ______________________________________ =================================================================== ·Q¾Ö¦³¥v¤W¶W±jªº50MB¶W¤jEmail? http://www.sinamail.com ½Ð¤W·s®öºôhttp://www.sina.com.tw =================================================================== ·Q¾Ö¦³¥v¤W¶W±jªº50MB¶W¤jEmail? http://www.sinamail.com ½Ð¤W·s®öºôhttp://www.sina.com.tw unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Mon Aug 07 07:15:27 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13Lg6Q-0007Et-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Mon, 07 Aug 2000 07:09:34 +0100 Received: from [193.145.240.2] (helo=tid.tid.es) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13Lg6O-0007En-00 for linux-arm-kernel@lists.arm.linux.org.uk; Mon, 07 Aug 2000 07:09:32 +0100 Received: from tid.es ([172.17.1.9]) by tid.tid.es (Netscape Messaging Server 4.15) with ESMTP id FYWRRO00.L0T; Mon, 7 Aug 2000 08:09:24 +0200 Message-ID: <398E5136.E110BCF5@tid.es> Date: Mon, 07 Aug 2000 08:03:35 +0200 From: David Aparicio Campos X-Mailer: Mozilla 4.7 [en] (Win98; I) X-Accept-Language: en MIME-Version: 1.0 To: Russell King - ARM Linux Admin CC: linux-arm-kernel@lists.arm.linux.org.uk Subject: Re: More IRQ questions References: <200008041631.RAA08879@flint.arm.linux.org.uk> Content-Type: multipart/mixed; boundary="------------39F0C07B2DEFB60E90A45760" Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk This is a multi-part message in MIME format. --------------39F0C07B2DEFB60E90A45760 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Russell King - ARM Linux Admin wrote: > > * Why Linux for CATS doen't assign IRQ from onboard devices to ISA > > range (less than 16)?? > > From your description, the PCI interrupts are not connected to the > southbridge, which handles IRQs 0 through 15. Right. PCI IRQ INTA..D are directly connected to FootBridge pins, but also routed to Southbridge "as they would be in a PC, and can be assigned ISA interrupts" (according to some sheets from Chalice). What confuses me is asignation of on-board southbridge devices, like USB, which is given a 139 IRQ, then Linux assigns IRQ (139-128)+16 -> 27 that would be a IRQ_DMA2. This makes no sense to me. In this same sheets Chalice says it assigns USB an ISA IRQ, number 11. This would lead to a IRQ (139-128) -> 11, seems righter. All this is related to BIOS assigned IRQ > 128, which Linux translates to FootBridge range, above 16: ("arch/arm/kernel/hw-footbridge.c" function "cats_irqval"). Has anybody played with USB driver for CATS?? > > * In order to develop my own firmare, should I use "compatible" IRQ > > assignment (and resource mapping) or should I propose a new Linux > > architecture? (First seems good at first, but I have almost no > > information about original firmware, and reverse engineering is very > > limited because of write-only registers). A new Linux architecture > > seems reasonable if more driver support could be added by people > > (volunteers?) to this firmware, becoming then better than original > > one. > > If there is any interest, and as soon as it becomes stable, I will > > release a snapshot...could evolve to a general EBSA285/CATS one... > > I don't believe this is necessary. OK, I will try to be "compatible" with former Chalice BIOS (sig). > > * There is a problem with message consistency, related to IRQ debugging: > > 1) arm subsystem prints original IRQ numbers from PCI CSR space (1..4 > > and 128..144), but > > 2) main Linux code shows "compatible", mapped numbers (range 0..31) > > Can we change messages in arch/arm/kernel in order to show same > > numbers than main Linux development? (maybe two numbers, with both > > "compatible" and PCI assignments). > > Please explain some more. What/where/how are you getting these IRQ > numbers? I have revised all "printk" calls in arch/arm, and I found only one example. Usually, Linux prints standard PC numbering. When some driver (in CATS architecture) finds a PCI device, prints an IRQ assignment above IRQ 16 (which is right). In "arch/arm/kernel/hw-footbridge.c", same "cats_irqval": irq printed is that assigned by BIOS, not a Linux number (which can not be done because there is no possible assignment according to the table). My complaining is about labelling only: Message like this "PCI: device AA:BB has unknown irq line X" could be some like "PCI: device AA:BB has unknown BIOS INT number X" Maybe too much words for an improbable error, it's my fault! (I have seen this message because I assigned wrong numbers in my BIOS, but besides Chalice radically change its firmware, this message is never going to be seen). As I'm new to arm development, I only wanted to express my first impression about being confused by "prink" information with my IRQ problems... but as I write, it seems less important to me, about to delete all this paragraph... The idea was because x86 mixes PCI and ISA interrupts in range 0..15 (the "classic" concept about IRQ that we see in messages). So ARM does in a different way, it has: 1) IRQ "compatible numbers" 2) INT native & PCI numbers, (compatible with #1) 3) BIOS numeration, (incompatible with #2 and #3) My proposal was to maintain labels like "IRQ n" for case 1, but change 2 and 3: 2) Arquitecture Interrupt Number x : "AIRQ x" 3) Bios Interrupt Number x: "BIRQ x" Maybe #2 being a crazy idea and #3 a practical one... and I don't know about other architectures, if it makes sense or not... David --------------39F0C07B2DEFB60E90A45760 Content-Type: text/x-vcard; charset=us-ascii; name="daparic.vcf" Content-Transfer-Encoding: 7bit Content-Description: Card for David Aparicio Campos Content-Disposition: attachment; filename="daparic.vcf" begin:vcard n:Aparicio Campos;David tel;fax:+34-91-3374502 tel;work:+34-91-3374178 x-mozilla-html:TRUE org:Telefónica I+D;Gerencia 1200 adr:;;Emilio Vargas, 6;Madrid;;28043;Spain version:2.1 email;internet:daparic@tid.es title:R&D Engineer end:vcard --------------39F0C07B2DEFB60E90A45760-- unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Mon Aug 07 18:56:03 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13Lr6y-0003xc-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Mon, 07 Aug 2000 18:54:52 +0100 Received: from [144.92.9.40] (helo=mail1.doit.wisc.edu) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13Lr6x-0003xW-00 for linux-arm-kernel@lists.arm.linux.org.uk; Mon, 07 Aug 2000 18:54:51 +0100 Received: from [144.92.81.37] by mail1.doit.wisc.edu id MAA279662 (8.9.1/50); Mon, 7 Aug 2000 12:54:52 -0500 Message-ID: <398EF7BB.77A95242@facstaff.wisc.edu> Date: Mon, 07 Aug 2000 12:54:03 -0500 From: Russell Poyner X-Mailer: Mozilla 4.7 [en] (X11; I; Linux 2.2.16 i686) X-Accept-Language: en MIME-Version: 1.0 To: linux-arm-kernel@lists.arm.linux.org.uk Subject: advise on kernel for cl7500fe Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk Hello, I am the lucky owner of three set top boxes built on cirrus logic's 7500fe chip. Two newcom webpals and one daewoo inet.top.box I am trying to hack linux onto them and would like advise/ tips on which arm kernel to use. I see from the archives that there has been some work on a cl7500fe specific port of the kernel, and I wonder what the status of that is. Alternatively it might be more prudent for now to stick with an older kernel while I work through issues of hacking the box itself. There has been one report at linux-hacker.net of a webpal booting what appeared to be an arm6 kernel. Thanks in advance for your help. Russell Poyner rpoyner@facstaff.wisc.edu unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Mon Aug 07 22:04:44 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13Lu3P-0007W2-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Mon, 07 Aug 2000 22:03:23 +0100 Received: from [212.1.130.10] (helo=www.uklinux.net ident=root) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13Lu3O-0007Vu-00 for linux-arm-kernel@lists.arm.linux.org.uk; Mon, 07 Aug 2000 22:03:22 +0100 Received: from caramon.arm.linux.org.uk (root@ppp-1-182.cvx5.telinco.net [212.1.152.182]) by www.uklinux.net (8.9.3/8.8.7) with ESMTP id WAA03864; Mon, 7 Aug 2000 22:07:11 +0100 Received: from flint.arm.linux.org.uk (root@flint [192.168.0.4]) by caramon.arm.linux.org.uk (8.9.3/8.9.3) with ESMTP id WAA04785; Mon, 7 Aug 2000 22:03:20 +0100 Received: (from linux@localhost) by flint.arm.linux.org.uk (8.9.3/8.9.3) id WAA04073; Mon, 7 Aug 2000 22:01:29 +0100 From: Russell King - ARM Linux Admin Message-Id: <200008072101.WAA04073@flint.arm.linux.org.uk> Subject: Re: More IRQ questions To: daparic@tid.es (David Aparicio Campos) Date: Mon, 7 Aug 2000 22:01:28 +0100 (BST) Cc: linux-arm-kernel@lists.arm.linux.org.uk In-Reply-To: <398E5136.E110BCF5@tid.es> from "David Aparicio Campos" at Aug 07, 2000 08:03:35 AM X-Mailer: ELM [version 2.5 PL1] MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk David Aparicio Campos writes: > 1) IRQ "compatible numbers" > 2) INT native & PCI numbers, (compatible with #1) > 3) BIOS numeration, (incompatible with #2 and #3) > > My proposal was to maintain labels like "IRQ n" for case 1, but change 2 and > 3: > 2) Arquitecture Interrupt Number x : "AIRQ x" > 3) Bios Interrupt Number x: "BIRQ x" > > Maybe #2 being a crazy idea and #3 a practical one... and I don't know about > other architectures, if it makes sense or not... It does seem sensible, especially as it's only the CATS which has this weird and wonderful mapping scheme. Feel free to submit a patch, but note that you're working on a non-development kernel. I don't know if Phil has any comments on this problem (its his code after all). _____ |_____| ------------------------------------------------- ---+---+- | | Russell King linux@arm.linux.org.uk --- --- | | | | http://www.arm.linux.org.uk/ / / | | +-+-+ --- -+- / | THE developer of ARM Linux |+| /|\ / | | | --- | +-+-+ ------------------------------------------------- /\\\ | unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Tue Aug 08 02:21:55 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13Ly52-0003dk-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Tue, 08 Aug 2000 02:21:20 +0100 Received: from [207.69.200.157] (helo=tisch.mail.mindspring.net) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13Ly51-0003dX-00 for linux-arm-kernel@lists.arm.linux.org.uk; Tue, 08 Aug 2000 02:21:19 +0100 Received: from cementship.extenex.net (user-33qth5k.dialup.mindspring.com [199.174.196.180]) by tisch.mail.mindspring.net (8.9.3/8.8.5) with ESMTP id VAA30815; Mon, 7 Aug 2000 21:21:15 -0400 (EDT) Received: from blowhole.extenex.com (blowhole.extenex.net [192.168.100.100]) by cementship.extenex.net (8.9.3/8.8.7) with ESMTP id SAA06771; Mon, 7 Aug 2000 18:19:59 -0700 Message-Id: <4.3.2.7.2.20000807175115.00b1b5d8@willats.com> X-Sender: willats.com/extenex@willats.com X-Mailer: QUALCOMM Windows Eudora Version 4.3.2 Date: Mon, 07 Aug 2000 18:20:12 -0700 To: linux-arm-kernel@lists.arm.linux.org.uk From: Ward Willats Subject: serial_sa1100 Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii"; format=flowed Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk Hello, New to the list and arm-linux and still finding my way around. Good to meet you. Have what I hope is an easy question.... If I take 2.4.0-test5 sources and apply rmk2 (or rmk1) patches and configure for Assabet, an attempt is made to link in serial_sa1100.c which is not part of the rmk patches. Looks like I can go over to netwinder and get this from Nico Pitre (thank you!) -- but I guess I'm wondering if this is what I'm "supposed" to do? That is, does a driver like this "live" with the author/maintainer? Or is it something that eventually migrates over to the rmk tree and then the main kernel sources? As the list traffic goes by and I poke around in the archives the landscape is slowly being revealed, but it takes a while to figure out who is doing what where and the other secret handshakes of your mysterious tribe! :-) Any pointers to keep me on the One True Path(tm) are appreciated. Thanks, -- Ward unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Tue Aug 08 06:54:48 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13M2Kt-0000UI-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Tue, 08 Aug 2000 06:53:59 +0100 Received: from [193.145.240.2] (helo=tid.tid.es) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13M2Kn-0000U9-00 for linux-arm-kernel@lists.arm.linux.org.uk; Tue, 08 Aug 2000 06:53:58 +0100 Received: from tid.es ([172.17.1.9]) by tid.tid.es (Netscape Messaging Server 4.15) with ESMTP id FYYLPS01.J5H; Tue, 8 Aug 2000 07:53:52 +0200 Message-ID: <398F9F07.30701B08@tid.es> Date: Tue, 08 Aug 2000 07:47:51 +0200 From: David Aparicio Campos X-Mailer: Mozilla 4.7 [en] (Win98; I) X-Accept-Language: en MIME-Version: 1.0 To: Russell King - ARM Linux Admin CC: linux-arm-kernel@lists.arm.linux.org.uk Subject: Re: More IRQ questions References: <200008072101.WAA04073@flint.arm.linux.org.uk> Content-Type: multipart/mixed; boundary="------------F3DAD6998242A4A78BA942E1" Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk This is a multi-part message in MIME format. --------------F3DAD6998242A4A78BA942E1 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Russell King - ARM Linux Admin wrote: > David Aparicio Campos writes: > > 1) IRQ "compatible numbers" > > 2) INT native & PCI numbers, (compatible with #1) > > 3) BIOS numeration, (incompatible with #2 and #3) > > > > My proposal was to maintain labels like "IRQ n" for case 1, but change 2 and > > 3: > > 2) Arquitecture Interrupt Number x : "AIRQ x" > > 3) Bios Interrupt Number x: "BIRQ x" > > > > Maybe #2 being a crazy idea and #3 a practical one... and I don't know about > > other architectures, if it makes sense or not... > > It does seem sensible, especially as it's only the CATS which has this > weird and wonderful mapping scheme. Feel free to submit a patch, but > note that you're working on a non-development kernel. I will download 2.2.15 and 2.4.0-preX, and try to send patches for both. Still guessing If I should move on 2.4.0... I will give it a try. Please note that label (3) is for CATS (I think there are no messages about BIOS in other architectures) , but (2) is a general ARM question. From your mail, I will only patch CATS part. > I don't know if Phil has any comments on this problem (its his code after > all). Please, Phil, tell me also about removing slot 4... > That is in "arch/arm/kernel/hw-footbridge.c", function "cats_irqval". > > * Same source code, question about "IRQ 4". Where is slot 4?? It seems > like some mistake when derived from EBSA285 (which has 4 slots). Also, please tell me about on-board IRQs (above 128). I still think that USB assignment is wrong (at least I don't understand it)... but so I don't have USB devices (neither drivers knowledge) I cannot test this. > * Why Linux for CATS doen't assign IRQ from onboard devices to ISA range > (less than 16)?? For example, USB is 139, so Linux gives IRQ 27 > (Footbridge 11). CATS information says something about giving ISA > IRQ-11 to USB... > That is in "arch/arm/kernel/hw-footbridge.c", function "cats_irqval". Regards David Aparicio --------------F3DAD6998242A4A78BA942E1 Content-Type: text/x-vcard; charset=us-ascii; name="daparic.vcf" Content-Transfer-Encoding: 7bit Content-Description: Card for David Aparicio Campos Content-Disposition: attachment; filename="daparic.vcf" begin:vcard n:Aparicio Campos;David tel;fax:+34-91-3374502 tel;work:+34-91-3374178 x-mozilla-html:TRUE org:Telefónica I+D;Gerencia 1200 adr:;;Emilio Vargas, 6;Madrid;;28043;Spain version:2.1 email;internet:daparic@tid.es title:R&D Engineer end:vcard --------------F3DAD6998242A4A78BA942E1-- unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Tue Aug 08 20:34:30 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13MF6q-0006fX-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Tue, 08 Aug 2000 20:32:20 +0100 Received: from [216.18.11.193] (helo=duke.electric.net) by www.linux.org.uk with smtp (Exim 3.13 #1) id 13MF6p-0006fR-00 for linux-arm-kernel@lists.arm.linux.org.uk; Tue, 08 Aug 2000 20:32:19 +0100 Received: (qmail 47676 invoked from network); 8 Aug 2000 19:32:21 -0000 Received: from localhost (HELO relay.electric.net) (@127.0.0.1) by duke.electric.net with SMTP; 8 Aug 2000 19:32:21 -0000 Received: (from smtpd@localhost) by relay.electric.net (8.8.8/8.8.6) id MAA47653 for linux-arm-kernel@lists.arm.linux.org.uk; Tue, 8 Aug 2000 12:32:20 -0700 (PDT) Received: from cobra.intrinsyc.com(64.16.10.158), claiming to be "ccarey" via SMTP by relay.electric.net, id smtpdX47619; Tue Aug 8 12:32:20 2000 From: Chester Carey Organization: Intrinsyc To: linux-arm-kernel@lists.arm.linux.org.uk Subject: Xwindows Date: Tue, 8 Aug 2000 12:25:24 -0700 X-Mailer: KMail [version 1.0.29] Content-Type: text/plain MIME-Version: 1.0 Message-Id: <00080812340708.22983@ccarey> Content-Transfer-Encoding: 8bit Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk Hi all, We are currently trying to get X up and running on our ARM based board and have hit a bit of a snag. X looks for a keyboard and we do not have a keyboard attached to the board. We currently interface with the board through a serial connection at ttyS0. Is there some way to make X think that ttyS0 is the keyboard or has the default keyboard info? Or alternatively is there a way to make X run without a keyboard? Chester Carey Inrinsyc Software unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Wed Aug 09 16:01:24 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13MXLf-0000xq-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Wed, 09 Aug 2000 16:00:51 +0100 Received: from [208.225.80.11] (helo=ice.ar51.net) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13MXLe-0000xi-00 for linux-arm-kernel@lists.arm.linux.org.uk; Wed, 09 Aug 2000 16:00:51 +0100 Received: from server01.asi (gnede.ar51.net [208.225.80.76] (may be forged)) by ice.ar51.net (8.9.3/8.8.7) with ESMTP id LAA08563; Wed, 9 Aug 2000 11:00:17 -0400 Received: by SERVER01 with Internet Mail Service (5.5.2448.0) id ; Wed, 9 Aug 2000 11:02:06 -0400 Message-ID: From: Jeff Sutherland To: "'Ward Willats'" , "'linux-arm-kernel@lists.arm.linux.org.uk'" Subject: RE: serial_sa1100 Date: Tue, 8 Aug 2000 09:18:20 -0400 MIME-Version: 1.0 X-Mailer: Internet Mail Service (5.5.2448.0) Content-Type: text/plain Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk > Hello, > > New to the list and arm-linux and still finding my way > around. Good to meet > you. Have what I hope is an easy question.... > > If I take 2.4.0-test5 sources and apply rmk2 (or rmk1) patches and > configure for Assabet, an attempt is made to link in > serial_sa1100.c which > is not part of the rmk patches. Looks like I can go over to > netwinder and > get this from Nico Pitre (thank you!) -- but I guess I'm > wondering if this > is what I'm "supposed" to do? That is, does a driver like > this "live" with > the author/maintainer? Or is it something that eventually > migrates over to > the rmk tree and then the main kernel sources? > > As the list traffic goes by and I poke around in the archives > the landscape > is slowly being revealed, but it takes a while to figure out > who is doing > what where and the other secret handshakes of your mysterious > tribe! :-) > Any pointers to keep me on the One True Path(tm) are appreciated. > > Thanks, > > -- Ward For the SA1100 specific support, particularly for the Assabet, after applying Russell King's patch, you must apply Nicolas Pitre's patch. These are found at ftp.netwinder.org/users/n/nico and are mirrored at other places. Nico also has a ramdisk image which you will need after you get your kernel to boot in order to do anything useful. After patching, please see the ARM specific section under linux/Documentation in your kernel source. //Jeff unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Wed Aug 09 16:01:24 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13MXKm-0000vK-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Wed, 09 Aug 2000 15:59:57 +0100 Received: from [208.225.80.11] (helo=ice.ar51.net) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13MXKl-0000v7-00 for linux-arm-kernel@lists.arm.linux.org.uk; Wed, 09 Aug 2000 15:59:55 +0100 Received: from server01.asi (gnede.ar51.net [208.225.80.76] (may be forged)) by ice.ar51.net (8.9.3/8.8.7) with ESMTP id KAA08478; Wed, 9 Aug 2000 10:59:48 -0400 Received: by SERVER01 with Internet Mail Service (5.5.2448.0) id ; Wed, 9 Aug 2000 11:02:04 -0400 Message-ID: From: Jeff Sutherland To: "Linux ARM kernel mailing list (E-mail)" Cc: "Nicolas Pitre (E-mail)" Subject: Patch for the SA1100 frame buffer driver for Assabet users Date: Mon, 7 Aug 2000 13:47:19 -0400 MIME-Version: 1.0 X-Mailer: Internet Mail Service (5.5.2448.0) Content-Type: text/plain; charset="iso-8859-1" Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk Since Nicolas Pitre is still on vacation, I'm posting this here for Assabet users of recent kernel versions (2.4.0-testx series) if you wish to experiment with this. Tak-Shing Chan & I resolved an issue caused by a change made to the Assabet's PLD earlier this year which broke the frame buffer driver for newer Phase 4 Assabets. The patch also includes a fix to a bug which showed up in -test4-rmk1-np1 I believe. Some other parameters were changed to optimize for the Sharp display. The vertical alignment is now correct, and the frame rate is approximately 60Hz. The display's pixel clock divisor is now determined at runtime so feel free to experiment with your cpu speed settings without fear of messing up your display :) This patch was made against the SA1100 frame buffer driver in 2.4.0-test5-rmk1-np1: --- sa1100fb.c_orig Thu Aug 3 15:25:49 2000 +++ sa1100fb.c Mon Aug 7 12:27:44 2000 @@ -357,8 +357,8 @@ var->transp.offset = 0; } else { var->red.length = 5; - var->blue.length = 6; - var->green.length = 5; + var->blue.length = 5; + var->green.length = 6; var->transp.length = 0; var->red.offset = 11; var->green.offset = 5; @@ -641,6 +641,7 @@ init_var.blue.length = 5; init_var.grayscale = 0; init_var.sync = 0; + init_var.pixclock = 171521; } else if (machine_is_bitsy()) { current_par.max_xres = 320; current_par.max_yres = 240; @@ -813,6 +814,12 @@ /* the last multiplication by 1.2 is to handle */ /* sync problems */ } + if (machine_is_assabet()) { + pcd = frequency[PPCR & 0xf] / 1000; + pcd *= pixclock / 1000; + pcd = pcd / 1000000; + pcd++; /* make up for integer math truncations */ + } return pcd; } @@ -850,15 +857,15 @@ LCCR0_LDM + LCCR0_BAM + LCCR0_ERM + LCCR0_Act + LCCR0_LtlEnd + LCCR0_DMADel(0); lcd_shadow.lccr1 = - LCCR1_DisWdth(var->xres) + LCCR1_HorSnchWdth(4) + - LCCR1_BegLnDel(30) + LCCR1_EndLnDel(30); + LCCR1_DisWdth(var->xres) + LCCR1_HorSnchWdth(6) + + LCCR1_BegLnDel(61) + LCCR1_EndLnDel(9); lcd_shadow.lccr2 = LCCR2_DisHght(var->yres) + LCCR2_VrtSnchWdth(1) + - LCCR2_BegFrmDel(0) + LCCR2_EndFrmDel(0); + LCCR2_BegFrmDel(3) + LCCR2_EndFrmDel(0); lcd_shadow.lccr3 = LCCR3_OutEnH + LCCR3_PixFlEdg + LCCR3_VrtSnchH + LCCR3_HorSnchH + LCCR3_ACBsCntOff + - LCCR3_ACBsDiv(2) + LCCR3_PixClkDiv(28); + LCCR3_ACBsDiv(2) + LCCR3_PixClkDiv(pcd); /* Set board control register to handle new color depth */ sa1100fb_assabet_set_truecolor(var->bits_per_pixel >= 16); Regards, //Jeff Sutherland Accelent Systems, Inc. unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Wed Aug 09 16:01:24 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13MXL7-0000wU-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Wed, 09 Aug 2000 16:00:17 +0100 Received: from [208.225.80.11] (helo=ice.ar51.net) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13MXL6-0000wJ-00 for linux-arm-kernel@lists.arm.linux.org.uk; Wed, 09 Aug 2000 16:00:16 +0100 Received: from server01.asi (gnede.ar51.net [208.225.80.76] (may be forged)) by ice.ar51.net (8.9.3/8.8.7) with ESMTP id LAA08555; Wed, 9 Aug 2000 11:00:14 -0400 Received: by SERVER01 with Internet Mail Service (5.5.2448.0) id ; Wed, 9 Aug 2000 11:02:06 -0400 Message-ID: From: Jeff Sutherland To: "'Linux ARM kernel mailing list (E-mail)'" Cc: "'Nicolas Pitre (E-mail)'" Subject: Patch for the SA1100 frame buffer driver for Assabet users Date: Tue, 8 Aug 2000 09:13:09 -0400 MIME-Version: 1.0 X-Mailer: Internet Mail Service (5.5.2448.0) Content-Type: text/plain; charset="iso-8859-1" Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk Since Nicolas Pitre is still on vacation, I'm posting this here for Assabet users of recent kernel versions (2.4.0-testx series) if you wish to experiment with this. Tak-Shing Chan & I resolved an issue caused by a change made to the Assabet's PLD earlier this year which broke the frame buffer driver for newer Phase 4 Assabets. The patch also includes a fix to a bug which showed up in -test4-rmk1-np1 I believe. Some other parameters were changed to optimize for the Sharp display. The vertical alignment is now correct, and the frame rate is approximately 60Hz. The display's pixel clock divisor is now determined at runtime so feel free to experiment with your cpu speed settings without fear of messing up your display :) This patch was made against the SA1100 frame buffer driver in 2.4.0-test5-rmk1-np1: --- sa1100fb.c_orig Thu Aug 3 15:25:49 2000 +++ sa1100fb.c Mon Aug 7 12:27:44 2000 @@ -357,8 +357,8 @@ var->transp.offset = 0; } else { var->red.length = 5; - var->blue.length = 6; - var->green.length = 5; + var->blue.length = 5; + var->green.length = 6; var->transp.length = 0; var->red.offset = 11; var->green.offset = 5; @@ -641,6 +641,7 @@ init_var.blue.length = 5; init_var.grayscale = 0; init_var.sync = 0; + init_var.pixclock = 171521; } else if (machine_is_bitsy()) { current_par.max_xres = 320; current_par.max_yres = 240; @@ -813,6 +814,12 @@ /* the last multiplication by 1.2 is to handle */ /* sync problems */ } + if (machine_is_assabet()) { + pcd = frequency[PPCR & 0xf] / 1000; + pcd *= pixclock / 1000; + pcd = pcd / 1000000; + pcd++; /* make up for integer math truncations */ + } return pcd; } @@ -850,15 +857,15 @@ LCCR0_LDM + LCCR0_BAM + LCCR0_ERM + LCCR0_Act + LCCR0_LtlEnd + LCCR0_DMADel(0); lcd_shadow.lccr1 = - LCCR1_DisWdth(var->xres) + LCCR1_HorSnchWdth(4) + - LCCR1_BegLnDel(30) + LCCR1_EndLnDel(30); + LCCR1_DisWdth(var->xres) + LCCR1_HorSnchWdth(6) + + LCCR1_BegLnDel(61) + LCCR1_EndLnDel(9); lcd_shadow.lccr2 = LCCR2_DisHght(var->yres) + LCCR2_VrtSnchWdth(1) + - LCCR2_BegFrmDel(0) + LCCR2_EndFrmDel(0); + LCCR2_BegFrmDel(3) + LCCR2_EndFrmDel(0); lcd_shadow.lccr3 = LCCR3_OutEnH + LCCR3_PixFlEdg + LCCR3_VrtSnchH + LCCR3_HorSnchH + LCCR3_ACBsCntOff + - LCCR3_ACBsDiv(2) + LCCR3_PixClkDiv(28); + LCCR3_ACBsDiv(2) + LCCR3_PixClkDiv(pcd); /* Set board control register to handle new color depth */ sa1100fb_assabet_set_truecolor(var->bits_per_pixel >= 16); Regards, //Jeff Sutherland Accelent Systems, Inc. unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Wed Aug 09 16:09:28 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13MXTk-0001CB-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Wed, 09 Aug 2000 16:09:12 +0100 Received: from [208.225.80.11] (helo=ice.ar51.net) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13MXTj-0001C3-00 for linux-arm-kernel@lists.arm.linux.org.uk; Wed, 09 Aug 2000 16:09:11 +0100 Received: from jSutherland (gnede.ar51.net [208.225.80.76] (may be forged)) by ice.ar51.net (8.9.3/8.8.7) with SMTP id LAA09344 for ; Wed, 9 Aug 2000 11:08:28 -0400 Reply-To: From: "Jeff Sutherland" To: "Linux ARM kernel mailing list \(E-mail\)" Subject: test- ignore Date: Wed, 9 Aug 2000 11:10:13 -0400 Message-ID: MIME-Version: 1.0 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: 7bit X-Priority: 3 (Normal) X-MSMail-Priority: Normal X-Mailer: Microsoft Outlook CWS, Build 9.0.2416 (9.0.2910.0) Importance: Normal X-MimeOLE: Produced By Microsoft MimeOLE V5.00.2919.6700 Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk test, please ignore unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Wed Aug 09 21:33:43 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13McV2-0007fn-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Wed, 09 Aug 2000 21:30:52 +0100 Received: from [193.145.240.2] (helo=tid.tid.es) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13McV1-0007fh-00 for linux-arm-kernel@lists.arm.linux.org.uk; Wed, 09 Aug 2000 21:30:51 +0100 Received: from tid.es ([172.17.1.9]) by tid.tid.es (Netscape Messaging Server 4.15) with ESMTP id FZ12AW01.A6B; Wed, 9 Aug 2000 15:47:21 +0200 Message-ID: <39915F67.1F911AE@tid.es> Date: Wed, 09 Aug 2000 15:40:55 +0200 From: David Aparicio X-Mailer: Mozilla 4.7 [en] (Win98; I) X-Accept-Language: en MIME-Version: 1.0 To: Russell King - ARM Linux Admin , linux-arm-kernel@lists.arm.linux.org.uk Subject: Re: More IRQ questions References: <200008072101.WAA04073@flint.arm.linux.org.uk> <398F9F07.30701B08@tid.es> Content-Type: multipart/alternative; boundary="------------FDB2E81EE407EA3D7C566117" Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk --------------FDB2E81EE407EA3D7C566117 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Hello all again: David Aparicio Campos wrote: > Russell King - ARM Linux Admin wrote: > > >> It does seem sensible, especially as it's only the CATS which has this > > weird and wonderful mapping scheme. Feel free to submit a patch, but > > note that you're working on a non-development kernel. > > I will download 2.2.15 and 2.4.0-preX, and try to send patches for both. Still > guessing If I should move on 2.4.0... I will give it a try. > Please note that label (3) is for CATS (I think there are no messages about BIOS > in other architectures) , but (2) is a general ARM question. From your mail, I will > only patch CATS part. I downloaded 2.2.16 (rmk2) and 2-4.0test5 (rmk2). See below. > > I don't know if Phil has any comments on this problem (its his code after > > all). > > Please, Phil, tell me also about removing slot 4... I suppose it doesn't matter to change or leave it, what I would really want to know is if "slot 4" is any on-board function or if its an add-on card (as I guess). > > > That is in "arch/arm/kernel/hw-footbridge.c", function "cats_irqval". > > > > * Same source code, question about "IRQ 4". Where is slot 4?? It seems > > like some mistake when derived from EBSA285 (which has 4 slots). > > Also, please tell me about on-board IRQs (above 128). I still think that USB > assignment is wrong (at least I don't understand it)... but so I don't have USB > devices (neither drivers knowledge) I cannot test this. Sorry for the last one, I saw that 2.4.0 (somewhere in 2.3.xx I suppose) fixed it. I'm far from sending any patch, I had to install a new toolchain (I had incompatible binutils). Also 2.2.16 needed to be touched by hand (some illegal assembler intructions, there was a previous mail in this list about that). What happened is that my FW is broken with this new toolchain, I am checking all the thing again :-(. I used a "frozen" image of my FW with 2.4.0-test5. It hangs also (IRQ still beating somewhere), but it gives some more information. It's related to a semaphore inside irq lock handling... What is IRQ_32 (arrrggg) ?? Regards! boot no options Linux version 2.4.0-test5-rmk2 (root@lobos) (gcc version 2.95.2 20000516 (release) [Rebel.com]) #5 Tue Aug 8 18:32:04 CEST 2000 Processor: Intel StrongARM-110 revision 3 Architecture: Chalice-CATS On node 0 totalpages: 8192 zone(0): 8192 pages. zone(1): 0 pages. zone(2): 0 pages. Kernel command line: root=/dev/hda1 ro mem=32M parport=0x378,7 ide0=autotune ide_setup: ide0=autotune Calibrating delay loop... <2>Bad mode in interrupt handler detected: mode IRQ_32 Vectors: 00000000: ef9f0000 ea0000cd e59ff3d0 ea0000bb ea00009a ea0000ea ea000078 ea0000e7 00000020: f9000000 50000000 00100000 000000d1 00000000 00000000 00000000 00000000 Stubs: 00000200: e59fd1dc e24ee004 e58de000 e10fe000 e58de004 e3a0d093 e161f00d e20ee00f 00000220: e79fe10e e1b0f00e c00329a0 c0032688 c0032688 c00327a0 c0032688 c0032688 00000240: c0032688 c0032688 c0032688 c0032688 c0032688 c0032688 c0032688 c0032688 00000260: c0032688 c0032688 00000000 00000000 00000000 00000000 00000000 00000000 00000280: e59fd164 e24ee008 e58de000 e10fe000 e58de004 e3a0d093 e161f00d e20ee00f 000002a0: e79fe10e e1b0f00e c0032960 c0032674 c0032674 c0032740 c0032674 c0032674 000002c0: c0032674 c0032674 c0032674 c0032674 c0032674 c0032674 c0032674 c0032674 000002e0: c0032674 c0032674 00000000 00000000 00000000 00000000 00000000 00000000 00000300: e59fd0e4 e24ee004 e58de000 e10fe000 e58de004 e3a0d093 e161f00d e21ee00f 00000320: 059fe004 159fe004 e1b0f00e c0032b40 c0032660 00000000 00000000 00000000 00000340: e59fd0a0 e58de000 e10fe000 e58de004 e3a0d093 e161f00d e20ee00f e79fe10e 00000360: e1b0f00e c0032ae0 c003269c c003269c c00328e0 c003269c c003269c c003269c 00000380: c003269c c003269c c003269c c003269c c003269c c003269c c003269c c003269c 000003a0: c003269c 00000000 00000000 00000000 00000000 00000000 00000000 00000000 000003c0: e25ef004 eafffffe 00000000 00000000 00000000 00000000 00000000 00000000 000003e0: c0032c40 c0124a64 c0124a70 c0124a7c 00000000 00000000 00000000 00000000 00000400: 00000000 00000000 00000000 00000000 00000000 00000000 00000000 00000000 00000420: 00000000 00000000 00000000 00000000 00000000 00000000 00000000 00000000 00000440: 00000000 00000000 00000000 00000000 00000000 00000000 00000000 00000000 00000460: 00000000 00000000 00000000 00000000 00000000 00000000 00000000 00000000 00000480: 00000000 00000000 00000000 00000000 00000000 00000000 00000000 00000000 000004a0: 00000000 00000000 00000000 00000000 00000000 00000000 Internal error: Oops: 0 CPU: 0 pc : [] lr : [] sp : c00418c0 ip : c00083e8 fp : c0123fc0 r10: c0123fe4 r9 : 00015040 r8 : 4401a103 r7 : c0124a88 r6 : c013e150 r5 : c013e0e8 r4 : c013f538 r3 : c01248c4 r2 : 0000001a r1 : 80000053 r0 : c003e900 Flags: nZCv IRQs off FIQs off Mode IRQ_32 Segment kernel Control: 517D Table: 0000517D DAC: 0000001D Process swapper (pid: 0, stackpage=c0123000) Stack: invalid kernel stack pointer c0123fc0 Backtrace: invalid frame pointer c0123fc0 Code: pc not in code space Kernel panic: Attempted to kill the idle task! In idle task - not syncing .... --------------FDB2E81EE407EA3D7C566117 Content-Type: text/html; charset=us-ascii Content-Transfer-Encoding: 7bit Hello all again:

David Aparicio Campos wrote:

Russell King - ARM Linux Admin wrote:

>> It does seem sensible, especially as it's only the CATS which has this
> weird and wonderful mapping scheme.  Feel free to submit a patch, but
> note that you're working on a non-development kernel.

I will download 2.2.15 and 2.4.0-preX, and try to send patches for both. Still
guessing If I should move on 2.4.0... I will give it a try.
Please note that label (3) is for CATS (I think there are no messages about BIOS
in other architectures) , but (2) is a general ARM question. From your mail, I will
only patch CATS part.

I downloaded 2.2.16 (rmk2) and 2-4.0test5 (rmk2). See below.
> I don't know if Phil has any comments on this problem (its his code after
> all).

Please, Phil,  tell me also about removing slot 4...

I suppose it doesn't matter to change or leave it, what I would really want
to know is if "slot 4" is any on-board function or if its an add-on card (as
I guess).
 
>    That is in "arch/arm/kernel/hw-footbridge.c", function "cats_irqval".
>
>    * Same source code, question about "IRQ 4". Where is slot 4?? It seems
>      like some mistake when derived from EBSA285 (which has 4 slots).

Also, please tell me about on-board IRQs (above 128). I still think that USB
assignment is wrong (at least I don't understand it)... but so I don't have USB
devices (neither drivers knowledge) I cannot test this.

Sorry for the last one, I saw that 2.4.0 (somewhere in 2.3.xx I suppose) fixed it.

I'm far from sending any patch, I had to install a new toolchain (I had incompatible
binutils). Also 2.2.16 needed to be touched by hand (some illegal assembler intructions,
there was a previous mail in this list about that).
What happened is that my FW is broken with this new toolchain, I am checking all
the thing again :-(.
I used a "frozen" image of my FW with 2.4.0-test5. It hangs also (IRQ still beating
somewhere), but it gives some more information. It's related to a semaphore inside
irq lock handling... What is IRQ_32 (arrrggg) ??

Regards!

boot no options
Linux version 2.4.0-test5-rmk2 (root@lobos) (gcc version 2.95.2 20000516 (release) [Rebel.com]) #5 Tue Aug 8 18:32:04 CEST 2000
Processor: Intel StrongARM-110 revision 3
Architecture: Chalice-CATS
On node 0 totalpages: 8192
zone(0): 8192 pages.
zone(1): 0 pages.
zone(2): 0 pages.
Kernel command line: root=/dev/hda1 ro mem=32M parport=0x378,7 ide0=autotune
ide_setup: ide0=autotune
Calibrating delay loop... <2>Bad mode in interrupt handler detected: mode IRQ_32
Vectors:
00000000: ef9f0000 ea0000cd e59ff3d0 ea0000bb  ea00009a ea0000ea ea000078 ea0000e7
00000020: f9000000 50000000 00100000 000000d1  00000000 00000000 00000000 00000000
Stubs:
00000200: e59fd1dc e24ee004 e58de000 e10fe000  e58de004 e3a0d093 e161f00d e20ee00f
00000220: e79fe10e e1b0f00e c00329a0 c0032688  c0032688 c00327a0 c0032688 c0032688
00000240: c0032688 c0032688 c0032688 c0032688  c0032688 c0032688 c0032688 c0032688
00000260: c0032688 c0032688 00000000 00000000  00000000 00000000 00000000 00000000
00000280: e59fd164 e24ee008 e58de000 e10fe000  e58de004 e3a0d093 e161f00d e20ee00f
000002a0: e79fe10e e1b0f00e c0032960 c0032674  c0032674 c0032740 c0032674 c0032674
000002c0: c0032674 c0032674 c0032674 c0032674  c0032674 c0032674 c0032674 c0032674
000002e0: c0032674 c0032674 00000000 00000000  00000000 00000000 00000000 00000000
00000300: e59fd0e4 e24ee004 e58de000 e10fe000  e58de004 e3a0d093 e161f00d e21ee00f
00000320: 059fe004 159fe004 e1b0f00e c0032b40  c0032660 00000000 00000000 00000000
00000340: e59fd0a0 e58de000 e10fe000 e58de004  e3a0d093 e161f00d e20ee00f e79fe10e
00000360: e1b0f00e c0032ae0 c003269c c003269c  c00328e0 c003269c c003269c c003269c
00000380: c003269c c003269c c003269c c003269c  c003269c c003269c c003269c c003269c
000003a0: c003269c 00000000 00000000 00000000  00000000 00000000 00000000 00000000
000003c0: e25ef004 eafffffe 00000000 00000000  00000000 00000000 00000000 00000000
000003e0: c0032c40 c0124a64 c0124a70 c0124a7c  00000000 00000000 00000000 00000000
00000400: 00000000 00000000 00000000 00000000  00000000 00000000 00000000 00000000
00000420: 00000000 00000000 00000000 00000000  00000000 00000000 00000000 00000000
00000440: 00000000 00000000 00000000 00000000  00000000 00000000 00000000 00000000
00000460: 00000000 00000000 00000000 00000000  00000000 00000000 00000000 00000000
00000480: 00000000 00000000 00000000 00000000  00000000 00000000 00000000 00000000
000004a0: 00000000 00000000 00000000 00000000  00000000 00000000
Internal error: Oops: 0
CPU: 0
pc : [<c000840c>]    lr : [<c00fdda4>]
sp : c00418c0  ip : c00083e8  fp : c0123fc0
r10: c0123fe4  r9 : 00015040  r8 : 4401a103
r7 : c0124a88  r6 : c013e150  r5 : c013e0e8  r4 : c013f538
r3 : c01248c4  r2 : 0000001a  r1 : 80000053  r0 : c003e900
Flags: nZCv  IRQs off  FIQs off  Mode IRQ_32  Segment kernel
Control: 517D  Table: 0000517D  DAC: 0000001D
Process swapper (pid: 0, stackpage=c0123000)
Stack: invalid kernel stack pointer c0123fc0
Backtrace: invalid frame pointer c0123fc0
Code: pc not in code space
Kernel panic: Attempted to kill the idle task!
In idle task - not syncing
.... --------------FDB2E81EE407EA3D7C566117-- unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Fri Aug 11 02:50:37 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13N3ud-0007HQ-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Fri, 11 Aug 2000 02:47:07 +0100 Received: from [24.200.137.248] (helo=xanadu.gn.com) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13N3ub-0007HI-00 for linux-arm-kernel@lists.arm.linux.org.uk; Fri, 11 Aug 2000 02:47:05 +0100 Received: from localhost (nico@localhost) by xanadu.gn.com (8.9.3/8.9.3) with ESMTP id VAA04019; Thu, 10 Aug 2000 21:46:15 -0400 X-Authentication-Warning: xanadu.gn.com: nico owned process doing -bs Date: Thu, 10 Aug 2000 21:46:15 -0400 (EDT) From: Nicolas Pitre X-Sender: nico@xanadu.gn.com To: Jeff Sutherland cc: "Linux ARM kernel mailing list (E-mail)" Subject: Re: Patch for the SA1100 frame buffer driver for Assabet users In-Reply-To: Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk On Mon, 7 Aug 2000, Jeff Sutherland wrote: > Since Nicolas Pitre is still on vacation... I'm back into civilization! ... with plenty of emails to process... I should resurface soon. Nicolas unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Fri Aug 11 11:11:20 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13NBle-0002ht-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Fri, 11 Aug 2000 11:10:22 +0100 Received: from [210.200.129.221] (helo=ms1.tomail.com.tw) by www.linux.org.uk with smtp (Exim 3.13 #1) id 13NBlc-0002hj-00 for linux-arm-kernel@lists.arm.linux.org.uk; Fri, 11 Aug 2000 11:10:20 +0100 Received: (qmail 28278 invoked from network); 11 Aug 2000 10:10:13 -0000 Received: from ns.avision.com.tw (HELO cschiu) (203.73.132.5) by ms1.tomail.com.tw with SMTP; 11 Aug 2000 10:10:13 -0000 Message-ID: <002401c0037c$09e674a0$4314010a@avision.com.tw> From: "cschiu" To: "linux-arm-kernel" Subject: The different between arm linux and ucLinux Date: Fri, 11 Aug 2000 18:07:30 +0800 MIME-Version: 1.0 Content-Type: text/plain; charset="big5" Content-Transfer-Encoding: 7bit X-Priority: 3 X-MSMail-Priority: Normal X-Mailer: Microsoft Outlook Express 5.00.2615.200 X-MimeOLE: Produced By Microsoft MimeOLE V5.00.2615.200 Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk Hello, Could somebody tell me that the different between official arm linux(from Russel King) kernel and ucLinux in ARM7. If I have ARM7 evaluation board and no MMU. If I want to port Linux on it ,which one is better? Thanks __________________________________________________________ PC home §K¶O¹q¤l«H½c¡A¥Ó½Ð½Ð¦Ü: http://www.pchome.com.tw ·| ­û ²Ä ¤@ ¡A ¥x ÆW ³Ì ¤j ªº ¤J ¤f ºô ¯¸ unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Fri Aug 11 11:53:43 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13NCRJ-0003K5-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Fri, 11 Aug 2000 11:53:25 +0100 Received: from [130.161.42.1] (helo=duteinh.et.tudelft.nl) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13NCRI-0003Jz-00 for linux-arm-kernel@lists.arm.linux.org.uk; Fri, 11 Aug 2000 11:53:24 +0100 Received: from dutein75 by duteinh.et.tudelft.nl (8.9.3/BB-2.0) id MAA12079; Fri, 11 Aug 2000 12:52:56 +0200 (MET DST) Message-Id: <200008111052.MAA12079@duteinh.et.tudelft.nl> From: Erik Mouw To: cschiu@ms1.tomail.com.tw Cc: linux-arm-kernel@lists.arm.linux.org.uk Subject: Re: The different between arm linux and ucLinux In-Reply-To: <002401c0037c$09e674a0$4314010a@avision.com.tw> Organization: Eric conspiracy secret labs Date: Fri 11 Aug 2000 12:52:55 +0100 (MDT) X-Face: "!Z|1XjFg!3Iw)BwM^M\k)J5KHiPddz}#/Zw%?$&Q!xqob"OA|})3_@&BTTiQERo?/^g[Z<9P4R8:5C:aqq.cw!hotLc=+u-"+1rq@9$V^8PbOUlodyi_sejt+D`iqkLpx$>.aXl%vnfPIt"LK|/%>94_Zrm8jy?`_[RKt'TuoxUxfk(gMhY(n|DBBl~c$dhO>G%6}R4+DGX3|idBm4oU&UM/XUP4I9aoN{]`ax%B?rgD(I3OJ<62Y6L X-Mailer: XCmail 1.2devel - with PGP support, PGP engine version 0.5 (IRIX64) X-Mailerorigin: http://www.fsai.fh-trier.de/~schmitzj/Xclasses/XCmail/ X-Eric-Conspiracy: There is no conspiracy! X-Loop: erik@it.et.tudelft.nl MIME-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: 8bit Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk On Fri, 11 Aug 2000 18:07:30 +0800, cschiu wrote: > Could somebody tell me that the different between official arm linux(from > Russel King) kernel and ucLinux in ARM7. > > If I have ARM7 evaluation board and no MMU. > If I want to port Linux on it ,which one is better? ARM linux is for ARM CPUs with an MMU, ucLinux is for CPUs without MMU. Erik -- J.A.K. (Erik) Mouw, Information and Communication Theory Group, Department of Electrical Engineering, Faculty of Information Technology and Systems, Delft University of Technology, PO BOX 5031, 2600 GA Delft, The Netherlands Phone: +31-15-2783635 Fax: +31-15-2781843 Email: J.A.K.Mouw@its.tudelft.nl WWW: http://www-ict.its.tudelft.nl/~erik/ unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Fri Aug 11 12:29:46 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13ND0B-0003vk-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Fri, 11 Aug 2000 12:29:27 +0100 Received: from [212.27.32.5] (helo=smtp1.free.fr ident=root) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13ND0B-0003ve-00 for linux-arm-kernel@lists.arm.linux.org.uk; Fri, 11 Aug 2000 12:29:27 +0100 Received: from imp3.free.fr (imp3.free.fr [212.27.32.25]) by smtp1.free.fr (8.9.3/8.9.3/Debian/GNU) with ESMTP id NAA04692 for ; Fri, 11 Aug 2000 13:29:27 +0200 From: chris.bernadine@free.fr Received: (from www-data@localhost) by imp3.free.fr (8.9.3/8.9.3/Debian 8.9.3-6) id NAA14507 for linux-arm-kernel@lists.arm.linux.org.uk; Fri, 11 Aug 2000 13:29:27 +0200 X-Authentication-Warning: imp3.free.fr: www-data set sender to chris.bernadine@free.fr using -f To: linux-arm-kernel@lists.arm.linux.org.uk Subject: Linux for ARM compilation remarks Message-ID: <965993367.3993e39768b06@imp.free.fr> Date: Fri, 11 Aug 2000 13:29:27 +0200 (MEST) MIME-Version: 1.0 Content-Type: text/plain Content-Transfer-Encoding: 8bit User-Agent: IMP/PHP IMAP webmail program 2.2.0 X-Originating-IP: 212.27.38.36 Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk Hello, I am not sure I am writing to the write person.However.I would like to report some bugs,.. probably. Using your last kernel version : 2.4.0-test6 with the correspondant patch.I am compiling Linux for an 'asm-arm' architecture on a 'arch-rpc' machine. *During compilation a SUBDIR called 'special' is asked to be compiled in /arch/arm/Makefile, SUBDIR which is not present in the kernel. -> solution to suppress line 169 of the Makefile : arch/arm/special line 172 " " : DRIVERS += arch/arm/special/special.a -> or develop the 'special' SUBDIR *Also in file /include/asm-arm/arch-rpc/io.h, the compiler doesn't like the DECLARE_IO stuff with methods __inX(----------). Do know why I have just wrote the three methods and it gone through. ->unsigned char(--------) ->unsigned short(--------) ->unsigned int(--------) *Moreover I had a error in the file /arch/arm/kernel/ecard.c line 892: proc_bus is not declared. -> I think there is a define problem with CONFIG_PROC_FS which appears in /include/linux/proc_fs.h. I have enabled the /proc file system and it worked *Finally I have other failure which I can't manage to solve. Here are the compiler output : drivers/acorn/char/acorn-char.o: In function `handle_rawcode': drivers/acorn/char/acorn-char.o(.text+0x2bc): undefined reference to `handle_scancode' drivers/acorn/char/acorn-char.o: In function `ps2kbd_rx': drivers/acorn/char/acorn-char.o(.text+0x418): undefined reference to `kbd_pt_regs' drivers/acorn/char/acorn-char.o(.text+0x41c): undefined reference to `keyboard_tasklet' drivers/acorn/char/acorn-char.o: In function `pcf8583_attach': drivers/acorn/char/acorn-char.o(.text+0x948): undefined reference to `i2c_transfer' drivers/acorn/char/acorn-char.o(.text+0x95c): undefined reference to `i2c_attach_client' drivers/acorn/char/acorn-char.o: In function `pcf8583_probe': drivers/acorn/char/acorn-char.o(.text+0x980): undefined reference to `i2c_probe' drivers/acorn/char/acorn-char.o: In function `pcf8583_detach': drivers/acorn/char/acorn-char.o(.text+0x99c): undefined reference to `i2c_detach_client' drivers/acorn/char/acorn-char.o: In function `pcf8583_get_datetime': drivers/acorn/char/acorn-char.o(.text+0xa5c): undefined reference to `i2c_transfer' drivers/acorn/char/acorn-char.o: In function `pcf8583_set_datetime': drivers/acorn/char/acorn-char.o(.text+0xcc0): undefined reference to `i2c_master_send' drivers/acorn/char/acorn-char.o(.text+0xcdc): undefined reference to `i2c_master_send' drivers/acorn/char/acorn-char.o: In function `pcf8583_set_ctrl': drivers/acorn/char/acorn-char.o(.text+0xd38): undefined reference to i2c_master_send' rivers/acorn/char/acorn-char.o: In function `pcf8583_read_mem': drivers/acorn/char/acorn-char.o(.text+0xe04): undefined reference to `i2c_transfer' drivers/acorn/char/acorn-char.o: In function `pcf8583_write_mem': drivers/acorn/char/acorn-char.o(.text+0xedc): undefined reference to `i2c_transfer' drivers/acorn/char/acorn-char.o: In function `i2c_ioc_init': drivers/acorn/char/acorn-char.o(.text.init+0x110): undefined reference to `i2c_bit_add_bus' drivers/acorn/char/acorn-char.o: In function `pcf8583_init': drivers/acorn/char/acorn-char.o(.text.init+0x130): undefined reference to `i2c_add_driver' make: *** [vmlinux] Error 1 If you could help me, I am still a newbie. In advance, Thank you. Chris. (chris.bernadine@free.fr) unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Fri Aug 11 14:07:48 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13NEWx-0005HY-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Fri, 11 Aug 2000 14:07:23 +0100 Received: from dyn-33.linux.theplanet.co.uk ([195.92.244.33] helo=caramon.arm.linux.org.uk) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13NEWv-0005HS-00; Fri, 11 Aug 2000 14:07:21 +0100 Received: from flint.arm.linux.org.uk (root@flint [192.168.0.4]) by caramon.arm.linux.org.uk (8.9.3/8.9.3) with ESMTP id OAA22485; Fri, 11 Aug 2000 14:07:16 +0100 Received: (from linux@localhost) by flint.arm.linux.org.uk (8.9.3/8.9.3) id OAA01563; Fri, 11 Aug 2000 14:05:17 +0100 From: Russell King - ARM Linux Admin Message-Id: <200008111305.OAA01563@flint.arm.linux.org.uk> Subject: Re: Linux for ARM compilation remarks To: chris.bernadine@free.fr Date: Fri, 11 Aug 2000 14:05:16 +0100 (BST) Cc: linux-arm-kernel@lists.arm.linux.org.uk In-Reply-To: <965993367.3993e39768b06@imp.free.fr> from "chris.bernadine@free.fr" at Aug 11, 2000 01:29:27 PM X-Mailer: ELM [version 2.5 PL1] MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk chris.bernadine@free.fr writes: > I am not sure I am writing to the write person. However, I would like to > report some bugs,.. probably. > Using your last kernel version : 2.4.0-test6 with the correspondant patch. > I am compiling Linux for an 'asm-arm' architecture on a 'arch-rpc' machine. The ARM patch for test6 hasn't been released yet, and it doesn't compile directly from Linus' sources. > *During compilation a SUBDIR called 'special' is asked to be compiled > in /arch/arm/Makefile, SUBDIR which is not present in the kernel. > -> solution to suppress line 169 of the Makefile : arch/arm/special > line 172 " " : DRIVERS += arch/arm/special/special.a > -> or develop the 'special' SUBDIR As of the to be test6 patch, this directory is now removed so no longer an issue. > *Also in file /include/asm-arm/arch-rpc/io.h, the compiler doesn't like the > DECLARE_IO stuff with methods __inX(----------). Do know why > I have just wrote the three methods and it gone through. > ->unsigned char(--------) > ->unsigned short(--------) > ->unsigned int(--------) No changes there; works for me. Please supply the compiler errors/warnings. > *Moreover I had a error in the file /arch/arm/kernel/ecard.c > line 892: proc_bus is not declared. > -> I think there is a define problem with CONFIG_PROC_FS which appears in > /include/linux/proc_fs.h. > I have enabled the /proc file system and it worked Ok, must fix that then, but /proc should be compiled in (except if you aren't going to be using stuff like the ps command and have severe memory limitations). > *Finally I have other failure which I can't manage to solve. Can you supply your .config file please? (in private mail). Thanks. _____ |_____| ------------------------------------------------- ---+---+- | | Russell King linux@arm.linux.org.uk --- --- | | | | http://www.arm.linux.org.uk/ / / | | +-+-+ --- -+- / | THE developer of ARM Linux |+| /|\ / | | | --- | +-+-+ ------------------------------------------------- /\\\ | unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Fri Aug 11 14:32:26 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13NEv0-0005e9-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Fri, 11 Aug 2000 14:32:14 +0100 Received: from [139.175.250.58] (helo=mail2000.com.tw) by www.linux.org.uk with smtp (Exim 3.13 #1) id 13NEuz-0005dh-00 for linux-arm-kernel@lists.arm.linux.org.uk; Fri, 11 Aug 2000 14:32:14 +0100 Received: By OpenMail Mailer;Fri, 11 Aug 2000 21:31:32 +0800 (CST) From: "Chien-Min Lee" Reply-To: cm_lee@mail2000.com.tw Subject: Re: The different between arm linux and ucLinux Message-ID: <966000692.17438.cm_lee@mail2000.com.tw> To: "linux-arm-kernel" Date: Fri, 11 Aug 2000 21:31:32 +0800 (CST) MIME-Version: 1.0 Content-Type: text/plain Content-Transfer-Encoding: 8bit Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk > > Could somebody tell me that the different between official arm linux(from > > Russel King) kernel and ucLinux in ARM7. > > If I have ARM7 evaluation board and no MMU. > > If I want to port Linux on it ,which one is better? > ARM linux is for ARM CPUs with an MMU, ucLinux is for CPUs without MMU. There is one port of ucLinux for ARM7TDMI at following address: http://aplionet.aplio.fr/ucLinux/page3.htm Current ucLinux is based on Linux 2.0.38 Have a nice day! kirk ----== Mailed via Openfind ==----- http://mail2000.com.tw/ ´£¨Ñ§K¶O¹q¤l¶l¥ó«H½c unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Fri Aug 11 19:49:41 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13NJqs-0002XB-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Fri, 11 Aug 2000 19:48:18 +0100 Received: from [24.200.137.248] (helo=xanadu.gn.com) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13NJqr-0002X4-00 for linux-arm-kernel@lists.arm.linux.org.uk; Fri, 11 Aug 2000 19:48:17 +0100 Received: from localhost (nico@localhost) by xanadu.gn.com (8.9.3/8.9.3) with ESMTP id OAA04723; Fri, 11 Aug 2000 14:43:24 -0400 X-Authentication-Warning: xanadu.gn.com: nico owned process doing -bs Date: Fri, 11 Aug 2000 14:43:24 -0400 (EDT) From: Nicolas Pitre X-Sender: nico@xanadu.gn.com To: James Alford cc: SA-1100/Linux , linux-arm-kernel@lists.arm.linux.org.uk Subject: Re: SA1100 kernel booting problem. In-Reply-To: Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk On Fri, 4 Aug 2000, James Alford wrote: > I am trying to get the Intel StrongARM SA1100 Multimedia Development Board > working with a linux kernel. It is unfortunatly not a Brutus (or Assabet). > > I can get the 2.2.14-rmk5-np17 kernel up but it crashes at > "POSIX conformance teting by UNIFIX" > This is probably due to the fact that the kernel is built for a Brutus. I > have not yet started to edit the arch specfic bits yet. This is most likely because the memory declaration is wrong for your board. > My problem is that with 2.4.0-test5-rmk1-np1. I can't get the kernel to > start. The head stuff starts ok and decompresses the kernel outputing the > line > > "Uncompressing Linux........................ done, botting the kernel." > > It just stops there, it does not seem to step into the uncompressed > kernel. I think that the 2.5-test5 head stuff is working as it will boot > the 2.2.14 kernel. The uncompressed kernel expects some specific values to be passed in r0/r1 registers by the bootloader. In the 2.2.x kernels those values were bolted into the code because they were the same for all SA1100 designs. See linux/Documentation/arm/README.txt (near the end) for details. Nicolas unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Fri Aug 11 20:18:53 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13NKKI-00031V-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Fri, 11 Aug 2000 20:18:42 +0100 Received: from [24.200.137.248] (helo=xanadu.gn.com) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13NKKH-00031N-00 for linux-arm-kernel@lists.arm.linux.org.uk; Fri, 11 Aug 2000 20:18:41 +0100 Received: from localhost (nico@localhost) by xanadu.gn.com (8.9.3/8.9.3) with ESMTP id PAA04757; Fri, 11 Aug 2000 15:17:39 -0400 X-Authentication-Warning: xanadu.gn.com: nico owned process doing -bs Date: Fri, 11 Aug 2000 15:17:39 -0400 (EDT) From: Nicolas Pitre X-Sender: nico@xanadu.gn.com To: Ritter Thomas cc: "'linux-arm-kernel@lists.arm.linux.org.uk'" , "'john+@cs.cmu.edu'" Subject: Re: PCMCIA with Brutus In-Reply-To: Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk On Tue, 1 Aug 2000, Ritter Thomas wrote: [...] > Well, I didn't find the mentioned pcmcia-brutus.c file in the kernel source > tree. > > What patch am I missing / where can I get this file and maybe more PCMCIA > stuff ? > > My kernel version is V2.4.0-test5-rmk1-np1. PCMCIA support for Brutus simply hasn't been written yet. It only lacks the Brutus' specific interface. All the other implementations i.e. sa1100-*.c files should give a good example to the volunteer who wants to write the missing bits. Nicolas unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Fri Aug 11 20:35:11 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13NKa4-0003H5-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Fri, 11 Aug 2000 20:35:00 +0100 Received: from [216.18.11.193] (helo=duke.electric.net) by www.linux.org.uk with smtp (Exim 3.13 #1) id 13NKa3-0003Gn-00 for linux-arm-kernel@lists.arm.linux.org.uk; Fri, 11 Aug 2000 20:34:59 +0100 Received: (qmail 26717 invoked from network); 11 Aug 2000 19:34:58 -0000 Received: from localhost (HELO relay.electric.net) (@127.0.0.1) by duke.electric.net with SMTP; 11 Aug 2000 19:34:58 -0000 Received: (from smtpd@localhost) by relay.electric.net (8.8.8/8.8.6) id MAA26706 for linux-arm-kernel@lists.arm.linux.org.uk; Fri, 11 Aug 2000 12:34:58 -0700 (PDT) Received: from cobra.intrinsyc.com(64.16.10.158), claiming to be "ccarey" via SMTP by relay.electric.net, id smtpdW26665; Fri Aug 11 12:34:57 2000 From: Chester Carey Organization: Intrinsyc To: linux-arm-kernel@lists.arm.linux.org.uk Subject: PCMCIA and the test5 kernel Date: Fri, 11 Aug 2000 12:36:03 -0700 X-Mailer: KMail [version 1.0.29] Content-Type: text/plain MIME-Version: 1.0 Message-Id: <00081112364104.09447@ccarey> Content-Transfer-Encoding: 8bit Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk We get a kernel a panic on a PCMCIA interupt. Is PCMCIA broken for2.4.0-test5-rmk1-np1? Things work fine for the test4 version of the same atched kernel. Chester Carey Inrinsyc Software unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Sat Aug 12 05:23:55 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13NSn3-0003rf-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Sat, 12 Aug 2000 05:20:57 +0100 Received: from [139.175.250.58] (helo=mail2000.com.tw) by www.linux.org.uk with smtp (Exim 3.13 #1) id 13NSn2-0003r4-00 for linux-arm-kernel@lists.arm.linux.org.uk; Sat, 12 Aug 2000 05:20:56 +0100 Received: By OpenMail Mailer;Sat, 12 Aug 2000 12:19:51 +0800 (CST) From: "Chien-Min Lee" Reply-To: cm_lee@mail2000.com.tw Subject: Re: PCMCIA and the test5 kernel Message-ID: <966053991.26911.cm_lee@mail2000.com.tw> To: "ccarey" Cc: "linux-arm-kernel" Date: Sat, 12 Aug 2000 12:19:51 +0800 (CST) MIME-Version: 1.0 Content-Type: text/plain Content-Transfer-Encoding: 8bit Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk Hi: > We get a kernel a panic on a PCMCIA interupt. Is PCMCIA broken for2.4.0-test5-rmk1-np1? > Things work fine for the test4 version of the same atched kernel. I have tried the PCMCIA parts for 2.4.0-test5-rmk1-np1 with my Assabet. And it works well. HTH kirk ----== Mailed via Openfind ==----- http://mail2000.com.tw/ ´£¨Ñ§K¶O¹q¤l¶l¥ó«H½c unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Sat Aug 12 07:04:36 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13NUP0-0005HG-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Sat, 12 Aug 2000 07:04:14 +0100 Received: from [140.109.13.42] (helo=linux.org.tw ident=root) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13NUOy-0005HA-00 for linux-arm-kernel@lists.arm.linux.org.uk; Sat, 12 Aug 2000 07:04:12 +0100 Received: from localhost (chester@localhost) by linux.org.tw (8.9.3/8.9.3) with ESMTP id OAA12859; Sat, 12 Aug 2000 14:06:45 +0800 Date: Sat, 12 Aug 2000 14:06:45 +0800 (CST) From: To: Chester Carey cc: linux-arm-kernel@lists.arm.linux.org.uk Subject: Re: PCMCIA and the test5 kernel In-Reply-To: <00081112364104.09447@ccarey> Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk No,it works fine with SA-1110 board. Don't choose debug message when config PCMCIA/CardBus/CF enable Good luck, Chester On Fri, 11 Aug 2000, Chester Carey wrote: > We get a kernel a panic on a PCMCIA interupt. Is PCMCIA broken for2.4.0-test5-rmk1-np1? > Things work fine for the test4 version of the same atched kernel. > > Chester Carey > Inrinsyc Software > > > unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk > unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Sat Aug 12 09:54:55 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13NX3n-0000nB-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Sat, 12 Aug 2000 09:54:31 +0100 Received: from [140.109.13.42] (helo=linux.org.tw ident=root) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13NUOy-0005HA-00 for linux-arm-kernel@lists.arm.linux.org.uk; Sat, 12 Aug 2000 07:04:12 +0100 Received: from localhost (chester@localhost) by linux.org.tw (8.9.3/8.9.3) with ESMTP id OAA12859; Sat, 12 Aug 2000 14:06:45 +0800 Date: Sat, 12 Aug 2000 14:06:45 +0800 (CST) From: To: Chester Carey cc: linux-arm-kernel@lists.arm.linux.org.uk Subject: Re: PCMCIA and the test5 kernel In-Reply-To: <00081112364104.09447@ccarey> Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk No,it works fine with SA-1110 board. Don't choose debug message when config PCMCIA/CardBus/CF enable Good luck, Chester On Fri, 11 Aug 2000, Chester Carey wrote: > We get a kernel a panic on a PCMCIA interupt. Is PCMCIA broken for2.4.0-test5-rmk1-np1? > Things work fine for the test4 version of the same atched kernel. > > Chester Carey > Inrinsyc Software > > > unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk > unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Sun Aug 13 17:37:34 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13O0jw-0006BN-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Sun, 13 Aug 2000 17:36:00 +0100 Received: from [158.152.220.239] (helo=tazenda.demon.co.uk ident=mail) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13O0ju-0006BE-00 for linux-arm-kernel@lists.arm.linux.org.uk; Sun, 13 Aug 2000 17:35:59 +0100 Received: from kings-cross.london.uk.eu.org [::ffff:192.168.2.83] by tazenda.demon.co.uk with esmtp (Exim 3.11 #1 (Debian)) id 13O0mg-0006Ea-00; Sun, 13 Aug 2000 17:38:50 +0100 Received: from localhost ([::ffff:127.0.0.1] helo=tazenda.demon.co.uk ident=pb) by kings-cross.london.uk.eu.org with esmtp (Exim 3.12 #1 (Debian)) id 13O0kj-0004qi-00; Sun, 13 Aug 2000 17:36:49 +0100 X-Mailer: exmh version 2.1.1 10/15/1999 (debian) To: David Aparicio Campos Cc: linux-arm-kernel@lists.arm.linux.org.uk Subject: Re: More IRQ questions Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Date: Sun, 13 Aug 2000 17:36:48 +0100 From: Philip Blundell Message-Id: Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk >What confuses me is asignation of on-board southbridge devices, like USB, >which is given a 139 IRQ, then Linux assigns IRQ (139-128)+16 -> 27 >that would be a IRQ_DMA2. This makes no sense to me. In this same sheets >Chalice says it assigns USB an ISA IRQ, number 11. This would lead to >a IRQ (139-128) -> 11, seems righter. The code you are looking at is out of date. On footbridge platforms the ISA irqs used to be in the range 16..31. I don't recall exactly when this was changed, but I don't really expect 2.2 kernels to work very well on the CATS without some extra patches. You would be better off with 2.4. >3) Bios Interrupt Number x: "BIRQ x" I don't think this would do any harm, though I can't see that it buys you very much either. >Please, Phil, tell me also about removing slot 4... To be honest I don't really remember. It seems most likely that this is indeed just a mistake, but it won't do any harm since that slot number will never crop up. I'll check the code when I get time to make sure. >What is IRQ_32 The 32-bit "IRQ" mode (not to be confused with IRQ 32...). The processor is only in this mode for a very few instructions and interrupts should be disabled. p. unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Mon Aug 14 09:41:18 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13OFm2-0005ka-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Mon, 14 Aug 2000 09:39:10 +0100 Received: from [193.145.240.2] (helo=tid.tid.es) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13OFlm-0005kG-00 for linux-arm-kernel@lists.arm.linux.org.uk; Mon, 14 Aug 2000 09:39:09 +0100 Received: from tid.es ([172.17.1.9]) by tid.tid.es (Netscape Messaging Server 4.15) with ESMTP id FZ9XCS00.X0F; Mon, 14 Aug 2000 10:38:52 +0200 Message-ID: <3997AF5B.90C0886B@tid.es> Date: Mon, 14 Aug 2000 10:35:39 +0200 From: David Aparicio X-Mailer: Mozilla 4.7 [en] (Win98; I) X-Accept-Language: en MIME-Version: 1.0 To: Philip Blundell CC: linux-arm-kernel@lists.arm.linux.org.uk Subject: Re: More IRQ questions References: Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk Philip Blundell wrote: > >What confuses me is asignation of on-board southbridge devices, like USB, > >which is given a 139 IRQ, then Linux assigns IRQ (139-128)+16 -> 27 > >that would be a IRQ_DMA2. This makes no sense to me. In this same sheets > >Chalice says it assigns USB an ISA IRQ, number 11. This would lead to > >a IRQ (139-128) -> 11, seems righter. > > The code you are looking at is out of date. On footbridge platforms the ISA > irqs used to be in the range 16..31. I don't recall exactly when this was > changed, but I don't really expect 2.2 kernels to work very well on the CATS > without some extra patches. You would be better off with 2.4. I just have a space proble, 2.4 doen't fit in a 512 KB FLASH, so I will need to download every time I boot (disk and net boot from FW still unusable... I only have a 38400 serial link :-(). That was the reason I was using 2.2, but it seems it's time to change... > >3) Bios Interrupt Number x: "BIRQ x" > > I don't think this would do any harm, though I can't see that it buys you very > much either. You are right :-) > >Please, Phil, tell me also about removing slot 4... > > To be honest I don't really remember. It seems most likely that this is > indeed just a mistake, but it won't do any harm since that slot number will > never crop up. I'll check the code when I get time to make sure. > > >What is IRQ_32 > > The 32-bit "IRQ" mode (not to be confused with IRQ 32...). The processor is > only in this mode for a very few instructions and interrupts should be > disabled. Ok. What I plan to do is to support interrupt handling on FW (althought I don't need for drivers), unmask them manually (special command), and see which one and why is hammering... That way I will avoid to download 500 Kbytes of code for every test, (It would break my nerves for sure...). After that Linux 2.4.x should start up fine... I will be on holidays for the next weeks, so I will check this on early September... Happy hacking!! unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Wed Aug 16 15:41:14 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13P4KP-0003Li-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Wed, 16 Aug 2000 15:38:01 +0100 Received: from [211.32.117.44] (helo=s-mail12.hanmail.net) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13P4KO-0003LV-00 for linux-arm-kernel@lists.arm.linux.org.uk; Wed, 16 Aug 2000 15:38:00 +0100 Received: from www7.hanmail.net (www7.hanmail.net [211.32.117.27]) by s-mail12.hanmail.net (8.10.0/8.9.1) with ESMTP id e7GDOHX05578; Wed, 16 Aug 2000 22:24:17 +0900 Received: (from hanadmin@localhost) by www7.hanmail.net (8.9.1/8.9.1) id XAA12972; Wed, 16 Aug 2000 23:27:31 +0900 (KST) X-Originating-IP: [210.221.126.89] From: "¹éº´±â" Reply-To: "¹éº´±â" To: Russell King - ARM Linux Admin Cc: linux-arm-kernel@lists.arm.linux.org.uk Subject: More help : [Question] About SA1100 Memory map X-Mailer: Daum Web Mailer 1.0 Date: Wed, 16 Aug 2000 23:27:31 KST Message-Id: <20000816232731.HM.60000000000lHsT@www7.hanmail.net> MIME-Version: 1.0 Content-Type: text/plain; charset=euc-kr Content-Transfer-Encoding: quoted-printable X-MIME-Autoconverted: from 8bit to quoted-printable by s-mail12.hanmail.net id e7GDOHX05578 Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk Yes I modified mm-sa1100.c like this : { 0xf8000000, 0x80000000, 0x02000000, DOMAIN_IO, 0, 1 }, \ { 0xfa000000, 0x90000000, 0x02000000, DOMAIN_IO, 0, 1 }, \ { 0xfc000000, 0xa0000000, 0x02000000, DOMAIN_IO, 0, 1 }, \ { 0xfe000000, 0xb0000000, 0x02000000, DOMAIN_IO, 0, 1 }, \ /* under line is attached by me for ISA area(0x18000000) */ { 0xd1800000, 0x18000000, 0x00800000, DOMAIN_IO, 0, 1 }, { 0xe0000000, 0x20000000, 0x04000000, DOMAIN_IO, 0, 1 }, /* PCMC= IA0 IO */ \ ............. and then I modified the base address of inb/outb related macro to 0xd1800000 in hardware.h and io.h. The kernel didn't give me Oops or panic. But I couldn't access the right ISA device(for me RTL8019 lan card). Is it need any more modification of kernel ? Thank you=20 ByungGi Baek. =20 ---------> =B9=DE=C0=BA =B8=DE=C0=CF =B3=BB=BF=EB <---------- >From : Russell King - ARM Linux Admin >>=B9=E9=BA=B4=B1=E2 writes: >>> After linux kernel loaded and during run, I can't find any memory ma= p >>> about static memory area(0x0 - 0x20000000) at hardware.h and mm-init.= h >>> and related files. >> >>That is correct - this region is "user space". >> >>> When I accessed 0x18000000 using inb(), the kernel gives me an Oops. >>> And when I accessed 0x18000000 using memory access, the kernel panic = too. >> >>Also correct. >> >>> I think that because the kernel doesn't have any memory map and any >>> virt_to_phys/phys_to_virt features about static memory area. >> >>virt_to_phys and friends are only for use with the kernel direct mapped >>memory region (0xc0000000 to 0xd0000000). >> >>> How can I access my ISA region? Simple insertion of new map of ISA >>> region is enough ? >> >>You need to add a mapping into the map_desc for your system in >>arch/arm/mm/mm-sa1100.c. Note that the virtual address must NOT be >>in the 0x00000000 to 0xd0000000 region, and must be inserted in virtual >>address order, and must not overlap an existing region. >> >>Oh, and please post to the correct forum - it helps to get answers. >>(this is a question about kernel stuff, therefore >>linux-arm-kernel@lists.arm.linux.org.uk is the correct mailing list >>to post to). >> _____ >> |_____| ------------------------------------------------- ---+---+- >> | | Russell King linux@arm.linux.org.uk --- --- >> | | | | http://www.arm.linux.org.uk/ / / | >> | +-+-+ --- -+- >> / | THE developer of ARM Linux |+| /|\ >> / | | | --- | >> +-+-+ ------------------------------------------------- /\\\ | >> =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D ByungGi Baek. weapon100@hanmail.net=20 =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D =BF=EC=B8=AE =C0=CE=C5=CD=B3=DD, Daum =C6=F2=BB=FD =BE=B2=B4=C2 =B9=AB=B7=E1 E-mail =C1=D6=BC=D2 =C7=D1=B8=DE=C0= =CF=B3=DD =C1=F6=B1=B8=C3=CC =C7=D1=B1=DB =B0=CB=BB=F6=BC=AD=BA=F1=BD=BA Daum FIREB= ALL http://www.daum.net unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Wed Aug 16 16:11:56 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13P4qm-0003yk-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Wed, 16 Aug 2000 16:11:28 +0100 Received: from [207.97.93.2] (helo=blazie.com) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13P4ql-0003yc-00 for linux-arm-kernel@lists.arm.linux.org.uk; Wed, 16 Aug 2000 16:11:27 +0100 Received: from localhost (localhost [127.0.0.1]) by blazie.com (8.9.2/8.8.7) with ESMTP id LAA03453 for ; Wed, 16 Aug 2000 11:11:36 -0400 (EDT) Date: Wed, 16 Aug 2000 11:11:36 -0400 (EDT) From: Chris Blazie To: linux-arm-kernel@lists.arm.linux.org.uk Subject: slower decompress? Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk Hi, I just upgraded the kernel for our SA-1110 (custom) board from 2.3.99-pre3 to 2.4.0-test6-rmk2-np1. I've noticed that decompressing takes much longer than it used to (5-10 times longer). Is there a change in mmu or icache setup I'm missing? Thanks, Chris unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Wed Aug 16 16:15:59 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13P4uu-000434-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Wed, 16 Aug 2000 16:15:44 +0100 Received: from [205.250.170.220] (helo=xanadu.vipswitch.com) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13P4ut-00042v-00 for linux-arm-kernel@lists.arm.linux.org.uk; Wed, 16 Aug 2000 16:15:43 +0100 Date: Wed, 16 Aug 2000 11:14:48 -0400 (EDT) From: Nicolas Pitre To: =?iso-8859-1?B?uem6tLHi?= cc: Russell King - ARM Linux Admin , linux-arm-kernel@lists.arm.linux.org.uk Subject: Re: More help : [Question] About SA1100 Memory map In-Reply-To: <20000816232731.HM.60000000000lHsT@www7.hanmail.net> Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=iso-8859-1 Content-Transfer-Encoding: QUOTED-PRINTABLE Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk On Wed, 16 Aug 2000, =B9=E9=BA=B4=B1=E2 wrote: > Yes I modified mm-sa1100.c like this : >=20 > { 0xf8000000, 0x80000000, 0x02000000, DOMAIN_IO, 0, 1 }, \ > { 0xfa000000, 0x90000000, 0x02000000, DOMAIN_IO, 0, 1 }, \ > { 0xfc000000, 0xa0000000, 0x02000000, DOMAIN_IO, 0, 1 }, \ > { 0xfe000000, 0xb0000000, 0x02000000, DOMAIN_IO, 0, 1 }, \ > /* under line is attached by me for ISA area(0x18000000) */ > =09{ 0xd1800000, 0x18000000, 0x00800000, DOMAIN_IO, 0, 1 }, >=20 > { 0xe0000000, 0x20000000, 0x04000000, DOMAIN_IO, 0, 1 }, /* PCMC= IA0 IO */ \ > =09............. >=20 > and then I modified the base address of inb/outb related macro to > 0xd1800000 in hardware.h and io.h. Are you using 2.4.x? If so please don't do that. Instead use 0xd1800000 as your IO base adress with inb/outb. If you're using 2.2.x, may I strongly suggest you use 2.4.x instead since the SA1100 support is way better and things like specific architecture mapping like above are much cleanly done. Nicolas unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Wed Aug 16 16:37:27 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13P5Fh-0004Mg-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Wed, 16 Aug 2000 16:37:13 +0100 Received: from [205.250.170.220] (helo=xanadu.vipswitch.com) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13P5Fg-0004MX-00 for linux-arm-kernel@lists.arm.linux.org.uk; Wed, 16 Aug 2000 16:37:12 +0100 Date: Wed, 16 Aug 2000 11:36:05 -0400 (EDT) From: Nicolas Pitre To: Chris Blazie cc: linux-arm-kernel@lists.arm.linux.org.uk Subject: Re: slower decompress? In-Reply-To: Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk On Wed, 16 Aug 2000, Chris Blazie wrote: > Hi, > I just upgraded the kernel for our SA-1110 (custom) board from > 2.3.99-pre3 to 2.4.0-test6-rmk2-np1. I've noticed that decompressing > takes much longer than it used to (5-10 times longer). Is there a change > in mmu or icache setup I'm missing? See the end of linux/arch/arm/boot/compressed/head-sa1100.S: /* * Pause for a short time so that we give enough time * for the host to start a terminal up. */ mov r0, #0x00200000 1: subs r0, r0, #1 bne 1b Try removing that code completely. Nicolas unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Wed Aug 16 18:45:46 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13P7FF-0006UL-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Wed, 16 Aug 2000 18:44:53 +0100 Received: from dyn-33.linux.theplanet.co.uk ([195.92.244.33] helo=caramon.arm.linux.org.uk) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13P7FC-0006UB-00; Wed, 16 Aug 2000 18:44:51 +0100 Received: from flint.arm.linux.org.uk (root@flint [192.168.0.4]) by caramon.arm.linux.org.uk (8.9.3/8.9.3) with ESMTP id SAA03431; Wed, 16 Aug 2000 18:33:01 +0100 Received: (from linux@localhost) by flint.arm.linux.org.uk (8.9.3/8.9.3) id SAA09879; Wed, 16 Aug 2000 18:30:20 +0100 From: Russell King - ARM Linux Admin Message-Id: <200008161730.SAA09879@flint.arm.linux.org.uk> Subject: Re: slower decompress? To: chris@blazie.com (Chris Blazie) Date: Wed, 16 Aug 2000 18:30:19 +0100 (BST) Cc: linux-arm-kernel@lists.arm.linux.org.uk In-Reply-To: from "Chris Blazie" at Aug 16, 2000 11:11:36 AM X-Mailer: ELM [version 2.5 PL1] MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk Chris Blazie writes: > I just upgraded the kernel for our SA-1110 (custom) board from > 2.3.99-pre3 to 2.4.0-test6-rmk2-np1. I've noticed that decompressing > takes much longer than it used to (5-10 times longer). Is there a change > in mmu or icache setup I'm missing? There is currently code at the bottom of arch/arm/boot/compressed/head-sa1100.S to add a 1 second delay before decompression starts. This is so that the boot messages are actually readable when you use angelboot+minicom. _____ |_____| ------------------------------------------------- ---+---+- | | Russell King linux@arm.linux.org.uk --- --- | | | | http://www.arm.linux.org.uk/ / / | | +-+-+ --- -+- / | THE developer of ARM Linux |+| /|\ / | | | --- | +-+-+ ------------------------------------------------- /\\\ | unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Wed Aug 16 20:13:47 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13P8cd-0007sj-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Wed, 16 Aug 2000 20:13:07 +0100 Received: from [207.97.93.2] (helo=blazie.com) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13P8cb-0007sa-00 for linux-arm-kernel@lists.arm.linux.org.uk; Wed, 16 Aug 2000 20:13:05 +0100 Received: from localhost (localhost [127.0.0.1]) by blazie.com (8.9.2/8.8.7) with ESMTP id PAA06132; Wed, 16 Aug 2000 15:13:11 -0400 (EDT) Date: Wed, 16 Aug 2000 15:13:11 -0400 (EDT) From: Chris Blazie To: Russell King - ARM Linux Admin cc: linux-arm-kernel@lists.arm.linux.org.uk Subject: Re: slower decompress? In-Reply-To: <200008161730.SAA09879@flint.arm.linux.org.uk> Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk On Wed, 16 Aug 2000, Russell King - ARM Linux Admin wrote: > There is currently code at the bottom of arch/arm/boot/compressed/head-sa1100.S > to add a 1 second delay before decompression starts. This is so that the > boot messages are actually readable when you use angelboot+minicom. I saw that and removed it as Nicolas suggested. But the actual decompression itself (i.e. after Uncompressing Linux......) is where the real problem is. I just re-checked with an older kernel, and it decompresses in less than a second. The 2.4.0-test6 kernel takes 10-15 seconds. Something is definitely not getting set up right in my case. I'll try to pin down the changes from pre3. Thanks, Chris unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Wed Aug 16 21:02:04 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13P9NY-0000AP-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Wed, 16 Aug 2000 21:01:36 +0100 Received: from [208.225.80.11] (helo=ice.ar51.net) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13P9NW-0000AE-00 for linux-arm-kernel@lists.arm.linux.org.uk; Wed, 16 Aug 2000 21:01:34 +0100 Received: from jSutherland (akron-80-169.ar51.net [208.225.80.169]) by ice.ar51.net (8.9.3/8.8.7) with SMTP id QAA22495; Wed, 16 Aug 2000 16:01:37 -0400 Reply-To: From: "Jeff Sutherland" To: "'Chris Blazie'" Cc: Subject: RE: slower decompress? Date: Wed, 16 Aug 2000 16:03:53 -0400 Message-ID: MIME-Version: 1.0 Content-Type: text/plain; charset="US-ASCII" Content-Transfer-Encoding: 7bit X-Priority: 3 (Normal) X-MSMail-Priority: Normal X-Mailer: Microsoft Outlook CWS, Build 9.0.2416 (9.0.2910.0) In-Reply-To: X-MimeOLE: Produced By Microsoft MimeOLE V5.00.2919.6700 Importance: Normal Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk > On Wed, 16 Aug 2000, Russell King - ARM Linux Admin wrote: > > > There is currently code at the bottom of > arch/arm/boot/compressed/head-sa1100.S > > to add a 1 second delay before decompression starts. This > is so that the > > boot messages are actually readable when you use angelboot+minicom. > > I saw that and removed it as Nicolas suggested. But the actual > decompression itself (i.e. after Uncompressing Linux......) > is where the > real problem is. I just re-checked with an older kernel, and it > decompresses in less than a second. The 2.4.0-test6 kernel > takes 10-15 > seconds. Something is definitely not getting set up right in my case. > I'll try to pin down the changes from pre3. > > Thanks, > Chris What CPU clock speed are you running? Sounds like the MMU and the Icache aren't getting turned on. Having DEBUG defined in head.S could trip you up. //Jeff unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Wed Aug 16 21:46:23 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13PA35-0000qq-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Wed, 16 Aug 2000 21:44:31 +0100 Received: from [205.250.170.220] (helo=xanadu.vipswitch.com) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13PA33-0000qa-00 for linux-arm-kernel@lists.arm.linux.org.uk; Wed, 16 Aug 2000 21:44:29 +0100 Date: Wed, 16 Aug 2000 16:43:02 -0400 (EDT) From: Nicolas Pitre To: Chris Blazie cc: Russell King - ARM Linux Admin , linux-arm-kernel@lists.arm.linux.org.uk Subject: Re: slower decompress? In-Reply-To: Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk On Wed, 16 Aug 2000, Chris Blazie wrote: > On Wed, 16 Aug 2000, Russell King - ARM Linux Admin wrote: > > > There is currently code at the bottom of arch/arm/boot/compressed/head-sa1100.S > > to add a 1 second delay before decompression starts. This is so that the > > boot messages are actually readable when you use angelboot+minicom. > > I saw that and removed it as Nicolas suggested. But the actual > decompression itself (i.e. after Uncompressing Linux......) is where the > real problem is. I just re-checked with an older kernel, and it > decompresses in less than a second. The 2.4.0-test6 kernel takes 10-15 > seconds. Something is definitely not getting set up right in my case. Are the dots from the "Uncompressing Linux ..." banner actually appearing at a slow pace, or is the boot process stalling for a while after all dots appeared? Nicolas unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Thu Aug 17 14:04:15 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13PPGB-0006hA-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Thu, 17 Aug 2000 13:59:03 +0100 Received: from [207.97.93.2] (helo=blazie.com) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13PPGA-0006h4-00 for linux-arm-kernel@lists.arm.linux.org.uk; Thu, 17 Aug 2000 13:59:02 +0100 Received: from localhost (localhost [127.0.0.1]) by blazie.com (8.9.2/8.8.7) with ESMTP id IAA11969; Thu, 17 Aug 2000 08:59:08 -0400 (EDT) Date: Thu, 17 Aug 2000 08:59:08 -0400 (EDT) From: Chris Blazie To: Nicolas Pitre cc: linux-arm-kernel@lists.arm.linux.org.uk Subject: Re: slower decompress? In-Reply-To: Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk > Are the dots from the "Uncompressing Linux ..." banner actually appearing > at a slow pace, or is the boot process stalling for a while after all dots > appeared? The dots are appearing at a slow pace. What should the status of the mmu and icache be when decompress_kernel is called? Chris unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Thu Aug 17 16:13:07 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13PRJr-0000JC-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Thu, 17 Aug 2000 16:10:59 +0100 Received: from [211.32.117.45] (helo=s-mail13.hanmail.net) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13PRJq-0000J6-00 for linux-arm-kernel@lists.arm.linux.org.uk; Thu, 17 Aug 2000 16:10:58 +0100 Received: from www7.hanmail.net (www7.hanmail.net [211.32.117.27]) by s-mail13.hanmail.net (8.10.0/8.9.1) with ESMTP id e7HEM5n29381; Thu, 17 Aug 2000 23:22:10 +0900 Received: (from hanadmin@localhost) by www7.hanmail.net (8.9.1/8.9.1) id XAA22355; Thu, 17 Aug 2000 23:27:44 +0900 (KST) X-Originating-IP: [210.221.126.89] From: "¹éº´±â" Reply-To: "¹éº´±â" To: Russell King - ARM Linux Admin Cc: linux-arm-kernel@lists.arm.linux.org.uk Subject: [Questions] SA1100 : 2.4.x kernel archtecture type ??? X-Mailer: Daum Web Mailer 1.0 Date: Thu, 17 Aug 2000 23:27:44 KST Message-Id: <20000817232744.HM.60000000000lU79@www7.hanmail.net> MIME-Version: 1.0 Content-Type: text/plain; charset=euc-kr Content-Transfer-Encoding: quoted-printable X-MIME-Autoconverted: from 8bit to quoted-printable by s-mail13.hanmail.net id e7HEM5n29381 Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk Hello... Now I'm trying to port 2.4.x kernel to my board. I modified some basic files (head-sa1100.S, setup-sa1100.S). When I powered on, my system gave me under messages : -------------------------------------------------------- Uncompressing Linux.............................. done, booting the kerne= l. Error: a. -------------------------------------------------------- This messages from head-armv.S at under code=20 and I guess it's because wrong archtecture type. __entry: teq r0, #0 @ wrong register = vals? movne r0, #'i' @ yes, error 'i' bne __error bl __lookup_processor_type teq r10, #0 @ invalid process= or? moveq r0, #'p' @ yes, error 'p' beq __error bl __lookup_architecture_type teq r7, #0 @ invalid archite= cture? moveq r0, #'a' @ yes, error 'a' beq __error = =20 Could you give a piece of advice for it(me) ? Where should I check ? Thanks.. ByungGi Baek. =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D ByungGi Baek. weapon100@hanmail.net=20 =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D =BF=EC=B8=AE =C0=CE=C5=CD=B3=DD, Daum =C6=F2=BB=FD =BE=B2=B4=C2 =B9=AB=B7=E1 E-mail =C1=D6=BC=D2 =C7=D1=B8=DE=C0= =CF=B3=DD =C1=F6=B1=B8=C3=CC =C7=D1=B1=DB =B0=CB=BB=F6=BC=AD=BA=F1=BD=BA Daum FIREB= ALL http://www.daum.net unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Thu Aug 17 19:00:03 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13PTv9-0002cn-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Thu, 17 Aug 2000 18:57:39 +0100 Received: from [130.161.42.1] (helo=duteinh.et.tudelft.nl) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13PTv8-0002ch-00 for linux-arm-kernel@lists.arm.linux.org.uk; Thu, 17 Aug 2000 18:57:38 +0100 Received: from dutein75 by duteinh.et.tudelft.nl (8.9.3/BB-2.0) id TAA00921; Thu, 17 Aug 2000 19:57:36 +0200 (MET DST) Message-Id: <200008171757.TAA00921@duteinh.et.tudelft.nl> From: Erik Mouw To: weapon100@hanmail.net Cc: linux@arm.linux.org.uk, linux-arm-kernel@lists.arm.linux.org.uk Subject: Re: [Questions] SA1100 : 2.4.x kernel archtecture type ??? In-Reply-To: <20000817232744.HM.60000000000lU79@www7.hanmail.net> Organization: Eric conspiracy secret labs Date: Thu 17 Aug 2000 19:57:35 +0100 (MDT) X-Face: "!Z|1XjFg!3Iw)BwM^M\k)J5KHiPddz}#/Zw%?$&Q!xqob"OA|})3_@&BTTiQERo?/^g[Z<9P4R8:5C:aqq.cw!hotLc=+u-"+1rq@9$V^8PbOUlodyi_sejt+D`iqkLpx$>.aXl%vnfPIt"LK|/%>94_Zrm8jy?`_[RKt'TuoxUxfk(gMhY(n|DBBl~c$dhO>G%6}R4+DGX3|idBm4oU&UM/XUP4I9aoN{]`ax%B?rgD(I3OJ<62Y6L X-Mailer: XCmail 1.2devel - with PGP support, PGP engine version 0.5 (IRIX64) X-Mailerorigin: http://www.fsai.fh-trier.de/~schmitzj/Xclasses/XCmail/ X-Eric-Conspiracy: There is no conspiracy! X-Loop: erik@it.et.tudelft.nl MIME-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: quoted-printable X-MIME-Autoconverted: from 8bit to quoted-printable by duteinh.et.tudelft.nl id TAA00921 Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk On Thu, 17 Aug 2000 23:27:44 KST, =B9=E9=BA=B4=B1=E2 wrote: > Now I'm trying to port 2.4.x kernel to my board. >=20 > I modified some basic files (head-sa1100.S, setup-sa1100.S). >=20 > When I powered on, my system gave me under messages : > -------------------------------------------------------- > Uncompressing Linux.............................. done, booting the ker= nel. > Error: a. > -------------------------------------------------------- >=20 >=20 > This messages from head-armv.S at under code=20 > and I guess it's because wrong archtecture type. >=20 > __entry: > teq r0, #0 @ wrong registe= r vals? > movne r0, #'i' @ yes, error 'i= ' > bne __error > bl __lookup_processor_type > teq r10, #0 @ invalid proce= ssor? > moveq r0, #'p' @ yes, error 'p= ' > beq __error > bl __lookup_architecture_type > teq r7, #0 @ invalid archi= tecture? > moveq r0, #'a' @ yes, error 'a= ' > beq __error = =20 >=20 >=20 > Could you give a piece of advice for it(me) ? > Where should I check ? Request an architecture number from Russell King. See Documentation/arm/README in the Linux source tree. Erik --=20 Real Programmers consider "what you see is what you get" to be just as=20 bad a concept in Text Editors as it is in women. No, the Real Programmer wants a "you asked for it, you got it" text editor -- complicated,=20 cryptic, powerful, unforgiving, dangerous. unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Thu Aug 17 19:00:05 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13PTvA-0002cu-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Thu, 17 Aug 2000 18:57:40 +0100 Received: from [205.250.170.220] (helo=xanadu.vipswitch.com) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13PTv9-0002ca-00 for linux-arm-kernel@lists.arm.linux.org.uk; Thu, 17 Aug 2000 18:57:39 +0100 Date: Thu, 17 Aug 2000 13:56:55 -0400 (EDT) From: Nicolas Pitre To: =?iso-8859-1?B?uem6tLHi?= cc: linux-arm-kernel@lists.arm.linux.org.uk Subject: Re: [Questions] SA1100 : 2.4.x kernel archtecture type ??? In-Reply-To: <20000817232744.HM.60000000000lU79@www7.hanmail.net> Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=iso-8859-1 Content-Transfer-Encoding: QUOTED-PRINTABLE Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk On Thu, 17 Aug 2000, =B9=E9=BA=B4=B1=E2 wrote: > Hello... >=20 > Now I'm trying to port 2.4.x kernel to my board. >=20 > I modified some basic files (head-sa1100.S, setup-sa1100.S). >=20 > When I powered on, my system gave me under messages : > -------------------------------------------------------- > Uncompressing Linux.............................. done, booting the kerne= l. > Error: a. > -------------------------------------------------------- >=20 In linux/arch/arm/kernel/arch.[ch] you'll see usage of some macros that defines many parameters for all architectures. Those macros create a table at link time which is looked up against the architecture number passed in r1 by the bootloader. If there is no match then you get the message above. Be sure you have the proper definitions in linux/arch/arm/kernel/arch.c and the correct value in r1 from your bootloader. Nicolas unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Thu Aug 17 20:10:49 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13PV3G-0003ly-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Thu, 17 Aug 2000 20:10:06 +0100 Received: from dyn-33.linux.theplanet.co.uk ([195.92.244.33] helo=caramon.arm.linux.org.uk) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13PV3B-0003lU-00; Thu, 17 Aug 2000 20:10:02 +0100 Received: from flint.arm.linux.org.uk (root@flint [192.168.0.4]) by caramon.arm.linux.org.uk (8.9.3/8.9.3) with ESMTP id UAA07768; Thu, 17 Aug 2000 20:07:45 +0100 Received: (from linux@localhost) by flint.arm.linux.org.uk (8.9.3/8.9.3) id UAA15653; Thu, 17 Aug 2000 20:04:59 +0100 From: Russell King - ARM Linux Admin Message-Id: <200008171904.UAA15653@flint.arm.linux.org.uk> Subject: Re: [Questions] SA1100 : 2.4.x kernel archtecture type ??? To: weapon100@hanmail.net Date: Thu, 17 Aug 2000 20:04:59 +0100 (BST) Cc: linux-arm-kernel@lists.arm.linux.org.uk In-Reply-To: <20000817232744.HM.60000000000lU79@www7.hanmail.net> from "¹éº´±â" at Aug 17, 2000 11:27:44 PM X-Mailer: ELM [version 2.5 PL1] MIME-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: 8bit Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk ¹éº´±â writes: > Now I'm trying to port 2.4.x kernel to my board. > I modified some basic files (head-sa1100.S, setup-sa1100.S). > When I powered on, my system gave me under messages : > Error: a. > This messages from head-armv.S at under code > and I guess it's because wrong archtecture type. You need to get an architecture number allocated as per the comments above that code at the entry point. You then need to arrange for your loader or whatever to set r1 to that allocated value before calling the kernel code. _____ |_____| ------------------------------------------------- ---+---+- | | Russell King linux@arm.linux.org.uk --- --- | | | | http://www.arm.linux.org.uk/ / / | | +-+-+ --- -+- / | THE developer of ARM Linux |+| /|\ / | | | --- | +-+-+ ------------------------------------------------- /\\\ | unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Fri Aug 18 09:16:22 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13PhIg-0007L9-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Fri, 18 Aug 2000 09:14:50 +0100 Received: from [140.109.13.42] (helo=linux.org.tw ident=root) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13PhIe-0007L0-00 for linux-arm-kernel@lists.arm.linux.org.uk; Fri, 18 Aug 2000 09:14:48 +0100 Received: from localhost (chester@localhost) by linux.org.tw (8.9.3/8.9.3) with ESMTP id QAA21129 for ; Fri, 18 Aug 2000 16:17:40 +0800 Date: Fri, 18 Aug 2000 16:17:40 +0800 (CST) From: To: linux-arm-kernel@lists.arm.linux.org.uk Subject: Kernel panic Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk Hi, i think it's kernel related. I have rewrited some TS part of code,and want to test calibration program. But i got some trouble when i touch LCD. ---------- Unable to handle kernel paging request at virtual address 00004604 pgd = c0004000 *pgd = c01a4001, *pmd = c01a4001, *pte = 00000000, *ppte = 00000000 Internal error: Oops: 2 CPU: 0 pc : [] lr : [<00004601>] sp : c0167cb8 ip : 20000013 fp : c0167ce8 r10: 01000000 r9 : 00000004 r8 : c0167d0c r7 : 00000002 r6 : 00000000 r5 : e5942000 r4 : 00004601 r3 : 00000001 r2 : 00000004 r1 : c017c764 r0 : 00004601 Flags: Nzcv IRQs on FIQs on Mode SVC_32 Segment kernel Control: C166517F Table: C166517F DAC: 0000001D Process swapper (pid: 0, stackpage=c0167000) Stack: .............. 78 Backtrace: Function entered at [] from [] Function entered at [] from [] r7 = C017C688 r6 = 0000001D r5 = C0167D40 r4 = FFFFFFFF Function entered at [] from [] r6 = C019ACB0 r5 = C019B2E0 r4 = C019ACB0 Function entered at [] from [] Function entered at [] from [] r5 = 0000002C r4 = C019ACB0 Function entered at [] from [] r8 = C017C108 r7 = C017C688 r6 = 04000000 r5 = 0000002C r4 = C0BAAC00 Function entered at [] from [] Function entered at [] from [] Function entered at [] from [] Function entered at [] from [] r6 = C018F018 r5 = C018F028 r4 = C019ACB0 Function entered at [] from [] r4 = 00000000 Function entered at [] from [] r4 = 00000002 --- And i find that entry function of 0xc0020514 from System.map is do_alignment. How do i debug it. Thanks, Chester unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Fri Aug 18 10:02:13 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13Pi1e-00086N-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Fri, 18 Aug 2000 10:01:18 +0100 Received: from dyn-33.linux.theplanet.co.uk ([195.92.244.33] helo=caramon.arm.linux.org.uk) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13Pi1b-00086D-00; Fri, 18 Aug 2000 10:01:16 +0100 Received: from flint.arm.linux.org.uk (root@flint [192.168.0.4]) by caramon.arm.linux.org.uk (8.9.3/8.9.3) with ESMTP id KAA10614; Fri, 18 Aug 2000 10:01:23 +0100 Received: (from linux@localhost) by flint.arm.linux.org.uk (8.9.3/8.9.3) id JAA18087; Fri, 18 Aug 2000 09:58:34 +0100 From: Russell King - ARM Linux Admin Message-Id: <200008180858.JAA18087@flint.arm.linux.org.uk> Subject: Re: Kernel panic To: chester@linux.org.tw Date: Fri, 18 Aug 2000 09:58:34 +0100 (BST) Cc: linux-arm-kernel@lists.arm.linux.org.uk In-Reply-To: from "chester@linux.org.tw" at Aug 18, 2000 04:17:40 PM X-Mailer: ELM [version 2.5 PL1] MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk chester@linux.org.tw writes: > Hi, i think it's kernel related. > > I have rewrited some TS part of code,and want to test calibration > program. > But i got some trouble when i touch LCD. > And i find that entry function of 0xc0020514 from System.map > is do_alignment. > > How do i debug it. Try looking a few more of those up in System.map. (yes, we must get a ksymoops running...). _____ |_____| ------------------------------------------------- ---+---+- | | Russell King linux@arm.linux.org.uk --- --- | | | | http://www.arm.linux.org.uk/ / / | | +-+-+ --- -+- / | THE developer of ARM Linux |+| /|\ / | | | --- | +-+-+ ------------------------------------------------- /\\\ | unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Fri Aug 18 12:17:23 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13Pk8M-0001tF-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Fri, 18 Aug 2000 12:16:22 +0100 Received: from [62.192.10.197] (helo=ganja.hemp.org) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13Pk8L-0001sk-00 for linux-arm-kernel@lists.arm.linux.org.uk; Fri, 18 Aug 2000 12:16:21 +0100 Received: from infomatec.de (IDENT:savages@localhost.localdomain [127.0.0.1]) by ganja.hemp.org (8.9.3/8.9.3) with ESMTP id IAA00899 for ; Fri, 18 Aug 2000 08:24:18 +0200 Message-ID: <399CD692.27FCFC3B@infomatec.de> Date: Fri, 18 Aug 2000 08:24:18 +0200 From: Shaun Savage X-Mailer: Mozilla 4.72 [en] (X11; U; Linux 2.4.0-test6 i586) X-Accept-Language: en MIME-Version: 1.0 To: linux-arm-kernel@lists.arm.linux.org.uk Subject: CATS again Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk I have seen many post to the maillist about cats but none about kernel 2.4. Here is hat I have done: booted linux using bootp and tftp using the downloaded kernel from chalice. It works sort of. the disk locks up alot and I get tons of NFS error I am using the titan-VI for nfs root, I had to add development RPM's to compile I downloaded the linux-cats kernel source also I do a 'make oldconfig' using the config in the linux-cats to configure the new kernel, then get rid of alot of things I am trying to compile 2.4.0t6-rmk4 tes6 + patch rmk4 issues: mach-types.h is not made early and I get error uncompress.h needs machine_is_netwinder() defined by adding a .h file alim15x3.c 647,665,672 needs 'hw.' to get it to compile NOw I have a compressed kernel based on 2.4.0t6rmk4. do a tftp of this new kernel , nothing Questions?? 1> how does the cats-bootstrap.S used in the old kernel and new kernel. 2> how does the bootloader do wd0: booting Shaun Savage unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Fri Aug 18 14:15:05 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13Plwb-00045i-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Fri, 18 Aug 2000 14:12:21 +0100 Received: from dyn-33.linux.theplanet.co.uk ([195.92.244.33] helo=caramon.arm.linux.org.uk) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13PlwZ-00045V-00; Fri, 18 Aug 2000 14:12:20 +0100 Received: from flint.arm.linux.org.uk (root@flint [192.168.0.4]) by caramon.arm.linux.org.uk (8.9.3/8.9.3) with ESMTP id NAA11301; Fri, 18 Aug 2000 13:13:09 +0100 Received: (from linux@localhost) by flint.arm.linux.org.uk (8.9.3/8.9.3) id NAA18604; Fri, 18 Aug 2000 13:10:19 +0100 From: Russell King - ARM Linux Admin Message-Id: <200008181210.NAA18604@flint.arm.linux.org.uk> Subject: Re: Kernel bootsequence (was: Unable to write registers) To: Arjan.Opmeer@emn.ericsson.se ("Arjan Opmeer (EMN)") Date: Fri, 18 Aug 2000 13:10:19 +0100 (BST) Cc: nico@cam.org, linux-arm-kernel@lists.arm.linux.org.uk In-Reply-To: from "Arjan Opmeer (EMN)" at Aug 18, 2000 12:36:23 PM X-Mailer: ELM [version 2.5 PL1] MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk Please note that I've changed the CC: to point at the correct list. "Arjan Opmeer (EMN)" writes: > I arrived at the same conclusion as Nicholas and fixed it the same > way by calling the init function from init/main.c. Now, I can fully > initialise the USB controller and have it detect 1 port connected > to the controller. Please DON'T do this. Please try out my latest 2.4.0-test6 kernel and report back. I'd like some feedback from you guys, so that I know that the fix which is in the test7 pre patches works for you. If not, we need to try again before test7 comes out for real. > It is an ugly solution indeed. And as such, it is not a solution that should even be considered. Its great as a "quick hack" to get the kernel running, but nothing more than that. > Okay, are you trying to get this changed in the base kernel? No - for a couple of reasons: 1. its "ugly". Basically, if there is a bug, and the fix is more ugly than the bug itself, the bug will remain. If the fix is less ugly than the bug, then the fix goes in. 2. We believe that we've fixed your problem in the latest test6-rmk patch, and test7 pre patches. Therefore, the more correct fix that I proposed is already in, but still needs you guys to say "it fixed it". _____ |_____| ------------------------------------------------- ---+---+- | | Russell King linux@arm.linux.org.uk --- --- | | | | http://www.arm.linux.org.uk/ / / | | +-+-+ --- -+- / | THE developer of ARM Linux |+| /|\ / | | | --- | +-+-+ ------------------------------------------------- /\\\ | unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Fri Aug 18 15:17:10 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13PmwD-0005ZN-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Fri, 18 Aug 2000 15:16:01 +0100 Received: from [205.250.170.220] (helo=xanadu.vipswitch.com) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13PmwC-0005Z7-00 for linux-arm-kernel@lists.arm.linux.org.uk; Fri, 18 Aug 2000 15:16:00 +0100 Date: Fri, 18 Aug 2000 10:15:10 -0400 (EDT) From: Nicolas Pitre To: Russell King - ARM Linux Admin cc: "Arjan Opmeer (EMN)" , linux-arm-kernel@lists.arm.linux.org.uk Subject: Re: Kernel bootsequence (was: Unable to write registers) In-Reply-To: <200008181210.NAA18604@flint.arm.linux.org.uk> Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk On Fri, 18 Aug 2000, Russell King - ARM Linux Admin wrote: > Please note that I've changed the CC: to point at the correct list. > > "Arjan Opmeer (EMN)" writes: > > I arrived at the same conclusion as Nicholas and fixed it the same > > way by calling the init function from init/main.c. Now, I can fully > > initialise the USB controller and have it detect 1 port connected > > to the controller. > > Please DON'T do this. Please try out my latest 2.4.0-test6 kernel and > report back. I still have to make a SA1100 patch against your latest one. Hopefully tonight... Nicolas unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Fri Aug 18 15:22:38 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13Pn26-0005if-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Fri, 18 Aug 2000 15:22:06 +0100 Received: from [194.237.142.116] (helo=albatross-ext.wise.edt.ericsson.se) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13Pn25-0005iZ-00 for linux-arm-kernel@lists.arm.linux.org.uk; Fri, 18 Aug 2000 15:22:05 +0100 Received: from esealnt406.al.sw.ericsson.se (esealnt406.al.sw.ericsson.se [153.88.251.29]) by albatross.wise.edt.ericsson.se (8.11.0/8.11.0/WIREfire-1.3) with ESMTP id e7IEMGp15517 for ; Fri, 18 Aug 2000 16:22:16 +0200 (MEST) Received: from esealnt406.al.sw.ericsson.se ([153.88.251.29]) by esealnt406.al.sw.ericsson.se with Microsoft SMTPSVC(5.0.2195.1600); Fri, 18 Aug 2000 16:22:16 +0200 Received: from esealnt743.al.sw.ericsson.se ([153.88.251.13]) by esealnt406.al.sw.ericsson.se (NAVIEG 2.1 bld 61) with SMTP id M2000081816221613400 ; Fri, 18 Aug 2000 16:22:16 +0200 Received: by esealnt743.al.sw.ericsson.se with Internet Mail Service (5.5.2651.58) id ; Fri, 18 Aug 2000 16:22:16 +0200 Message-ID: From: "Arjan Opmeer (EMN)" To: "'Russell King - ARM Linux Admin'" Cc: nico@cam.org, linux-arm-kernel@lists.arm.linux.org.uk Subject: RE: Kernel bootsequence (was: Unable to write registers) Date: Fri, 18 Aug 2000 16:22:13 +0200 MIME-Version: 1.0 X-Mailer: Internet Mail Service (5.5.2651.58) Content-Type: text/plain; charset="iso-8859-1" X-OriginalArrivalTime: 18 Aug 2000 14:22:16.0492 (UTC) FILETIME=[B5C9EEC0:01C0091F] Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk > Please note that I've changed the CC: to point at the correct list. Great! I didn't know it had changed... > > It is an ugly solution indeed. > > And as such, it is not a solution that should even be considered. Its > great as a "quick hack" to get the kernel running, but nothing more > than that. That's about my feeling towards it too. > > Okay, are you trying to get this changed in the base kernel? > > No - for a couple of reasons: [snip good reasons] Sorry, I didn't make myself more clear, but I was talking about getting the correct fix in the base kernel, not about the quick hack. > Therefore, the more correct fix that I proposed is already > in, but still needs you guys to say "it fixed it". Okay, will do. But at the moment I am still running linux-2.4.0-test6-rmk2-np1 with the quick hack method in it. I expect Nicholas to release a patch against your work in a couple of days. I will try again when it is available. At the moment I am more or less stuck with getting the USB port working because I don't understand the whole USB interrupt mumbojumbo. I am waiting and hoping someone can explain that somewhat to me. Arjan unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Fri Aug 18 17:36:10 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13Pp4b-00080C-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Fri, 18 Aug 2000 17:32:49 +0100 Received: from [202.140.96.55] (helo=smtp1.ismart.net) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13Pp4a-0007zm-00 for linux-arm-kernel@lists.arm.linux.org.uk; Fri, 18 Aug 2000 17:32:49 +0100 Received: from ismart.net ([202.140.113.172]) by smtp1.ismart.net (Sun Internet Mail Server sims.3.5.1999.07.30.00.05.p8) with ESMTP id <0FZH00DO8XV6MC@smtp1.ismart.net> for linux-arm-kernel@lists.arm.linux.org.uk; Sat, 19 Aug 2000 00:30:43 +0800 (HKT) Date: Sat, 19 Aug 2000 00:36:31 +0800 From: Ah Yin Subject: Error to run init To: arm linux Message-id: <399D660F.E2052B51@ismart.net> Organization: Home MIME-version: 1.0 X-Mailer: Mozilla 4.7 [en] (X11; I; Linux 2.2.5-15 i686) Content-type: text/plain; charset=big5 Content-transfer-encoding: 7bit X-Accept-Language: en Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk Hi, I am trying to upgrade my arm linux from 2.2.1 to 2.4. After the kernel mount the ramdisk and try to open console and run init, it said "No such device" when open /dev/console and "Exec format error" (errno=8) when exec init. The ramdisk is what I used for a long time. What's wrong about that? Yours, Jacky unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Fri Aug 18 18:11:01 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13Ppen-0000H1-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Fri, 18 Aug 2000 18:10:13 +0100 Received: from dyn-33.linux.theplanet.co.uk ([195.92.244.33] helo=caramon.arm.linux.org.uk) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13Ppek-0000Gr-00; Fri, 18 Aug 2000 18:10:11 +0100 Received: from flint.arm.linux.org.uk (root@flint [192.168.0.4]) by caramon.arm.linux.org.uk (8.9.3/8.9.3) with ESMTP id SAA12555; Fri, 18 Aug 2000 18:10:19 +0100 Received: (from linux@localhost) by flint.arm.linux.org.uk (8.9.3/8.9.3) id SAA19631; Fri, 18 Aug 2000 18:07:28 +0100 From: Russell King - ARM Linux Admin Message-Id: <200008181707.SAA19631@flint.arm.linux.org.uk> Subject: Re: Error to run init To: snakie@ismart.net (Ah Yin) Date: Fri, 18 Aug 2000 18:07:28 +0100 (BST) Cc: linux-arm-kernel@lists.arm.linux.org.uk (arm linux) In-Reply-To: <399D660F.E2052B51@ismart.net> from "Ah Yin" at Aug 19, 2000 12:36:31 AM X-Mailer: ELM [version 2.5 PL1] MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk Ah Yin writes: > I am trying to upgrade my arm linux from 2.2.1 to 2.4. After the kernel > mount the ramdisk and try to open console and run init, it said "No such > device" when open /dev/console and "Exec format error" (errno=8) when > exec init. > > The ramdisk is what I used for a long time. What's wrong about that? Nico's old ramdisk used the illegal major/minors for /dev/console. It used the serial port major/minor IDs which is just not on. Since the major/minors for the 21285 serial port has now been officially allocated, they have changed. Also, Nico has fixed his ramdisk to use the correct major/minors now. Hence, please upgrade your ramdisk image. _____ |_____| ------------------------------------------------- ---+---+- | | Russell King linux@arm.linux.org.uk --- --- | | | | http://www.arm.linux.org.uk/ / / | | +-+-+ --- -+- / | THE developer of ARM Linux |+| /|\ / | | | --- | +-+-+ ------------------------------------------------- /\\\ | unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Fri Aug 18 18:56:35 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13PqMv-0000yL-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Fri, 18 Aug 2000 18:55:49 +0100 Received: from [216.136.129.38] (helo=web9205.mail.yahoo.com) by www.linux.org.uk with smtp (Exim 3.13 #1) id 13PqMu-0000xq-00 for linux-arm-kernel@lists.arm.linux.org.uk; Fri, 18 Aug 2000 18:55:48 +0100 Message-ID: <20000818175529.8193.qmail@web9205.mail.yahoo.com> Received: from [63.198.94.3] by web9205.mail.yahoo.com; Fri, 18 Aug 2000 10:55:29 PDT Date: Fri, 18 Aug 2000 10:55:29 -0700 (PDT) From: Al Kam Subject: bootpImage problem To: linux-arm-kernel@lists.arm.linux.org.uk MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk Am trying to build a zImage + initrd using "make bootpImage" with linux-2.4.0-test6 kernel. HW target is EBSA-285. Kernel build fails due to missing bootp.lds file. Can anyone provide an example of a bootp.lds file or explain its contents. Thanks. Al __________________________________________________ Do You Yahoo!? Send instant messages & get email alerts with Yahoo! Messenger. http://im.yahoo.com/ unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Fri Aug 18 19:26:43 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13Pqpz-0001S2-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Fri, 18 Aug 2000 19:25:51 +0100 Received: from dyn-33.linux.theplanet.co.uk ([195.92.244.33] helo=caramon.arm.linux.org.uk) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13Pqpw-0001Rw-00; Fri, 18 Aug 2000 19:25:49 +0100 Received: from flint.arm.linux.org.uk (root@flint [192.168.0.4]) by caramon.arm.linux.org.uk (8.9.3/8.9.3) with ESMTP id TAA12936; Fri, 18 Aug 2000 19:25:58 +0100 Received: (from linux@localhost) by flint.arm.linux.org.uk (8.9.3/8.9.3) id TAA19918; Fri, 18 Aug 2000 19:23:06 +0100 From: Russell King - ARM Linux Admin Message-Id: <200008181823.TAA19918@flint.arm.linux.org.uk> Subject: Re: bootpImage problem To: alkam77@yahoo.com (Al Kam) Date: Fri, 18 Aug 2000 19:23:06 +0100 (BST) Cc: linux-arm-kernel@lists.arm.linux.org.uk In-Reply-To: <20000818175529.8193.qmail@web9205.mail.yahoo.com> from "Al Kam" at Aug 18, 2000 10:55:29 AM X-Mailer: ELM [version 2.5 PL1] MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk Al Kam writes: > Am trying to build a zImage + initrd using "make > bootpImage" with linux-2.4.0-test6 kernel. HW target > is EBSA-285. Kernel build fails due to missing > bootp.lds file. Can anyone provide an example of a > bootp.lds file or explain its contents. Thanks. It is actually there, but make clean removes it ;( _____ |_____| ------------------------------------------------- ---+---+- | | Russell King linux@arm.linux.org.uk --- --- | | | | http://www.arm.linux.org.uk/ / / | | +-+-+ --- -+- / | THE developer of ARM Linux |+| /|\ / | | | --- | +-+-+ ------------------------------------------------- /\\\ | unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Sat Aug 19 16:52:05 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13QAta-0004Lj-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Sat, 19 Aug 2000 16:50:54 +0100 Received: from [24.200.137.248] (helo=xanadu.gn.com) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13QAtZ-0004Ld-00 for linux-arm-kernel@lists.arm.linux.org.uk; Sat, 19 Aug 2000 16:50:53 +0100 Received: from localhost (nico@localhost) by xanadu.gn.com (8.9.3/8.9.3) with ESMTP id LAA24118; Sat, 19 Aug 2000 11:49:59 -0400 X-Authentication-Warning: xanadu.gn.com: nico owned process doing -bs Date: Sat, 19 Aug 2000 11:49:59 -0400 (EDT) From: Nicolas Pitre X-Sender: nico@xanadu.gn.com To: Chris Blazie cc: linux-arm-kernel@lists.arm.linux.org.uk, Nicolas Pitre Subject: Re: slower decompress? In-Reply-To: Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk On Thu, 17 Aug 2000, Chris Blazie wrote: > > > > Are the dots from the "Uncompressing Linux ..." banner actually appearing > > at a slow pace, or is the boot process stalling for a while after all dots > > appeared? > > The dots are appearing at a slow pace. What should the status of > the mmu and icache be when decompress_kernel is called? MMU, i-cache and d-cache are actually enabled for decompression speed in linux/arch/arm/compressed/head.S. Just tried it again on my Assabet and everything looks OK here. What bogomips number do you get? Nicolas unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Sun Aug 20 01:46:05 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13QJDz-0005RC-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Sun, 20 Aug 2000 01:44:31 +0100 Received: from [24.147.212.45] (helo=compaq.parker.boston.ma.us) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13QJDy-0005R2-00 for linux-arm-kernel@lists.arm.linux.org.uk; Sun, 20 Aug 2000 01:44:30 +0100 Received: from p2.parker.boston.ma.us (p2 [192.245.5.16]) by compaq.parker.boston.ma.us (8.8.5/8.8.5) with ESMTP id UAA14687; Sat, 19 Aug 2000 20:44:30 -0400 Received: from p2.parker.boston.ma.us (localhost [127.0.0.1]) by p2.parker.boston.ma.us (8.9.3/8.8.8) with ESMTP id UAA35962; Sat, 19 Aug 2000 20:44:29 -0400 (EDT) (envelope-from brad@p2.parker.boston.ma.us) Message-Id: <200008200044.UAA35962@p2.parker.boston.ma.us> To: Russell King - ARM Linux Admin cc: Arjan.Opmeer@emn.ericsson.se ("Arjan Opmeer (EMN)"), nico@cam.org, linux-arm-kernel@lists.arm.linux.org.uk Subject: Re: Kernel bootsequence (was: Unable to write registers) In-reply-to: Your message of "Fri, 18 Aug 2000 13:10:19 BST." <200008181210.NAA18604@flint.arm.linux.org.uk> Date: Sat, 19 Aug 2000 20:44:29 -0400 From: Brad Parker Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk I never saw the original message which is (partially) quoted here and I wonder if it ever came across the list (I don't think it did). Could someone repost it? it appears to have to do with a USB controller (SA1111 perhaps?) which is of some interest to several people on this list. thanks! -brad Russell King - ARM Linux Admin wrote: >Please note that I've changed the CC: to point at the correct list. > >"Arjan Opmeer (EMN)" writes: >> I arrived at the same conclusion as Nicholas and fixed it the same >> way by calling the init function from init/main.c. Now, I can fully >> initialise the USB controller and have it detect 1 port connected >> to the controller. > >Please DON'T do this. Please try out my latest 2.4.0-test6 kernel and >report back. I'd like some feedback from you guys, so that I know that >the fix which is in the test7 pre patches works for you. If not, we >need to try again before test7 comes out for real. > >> It is an ugly solution indeed. > >And as such, it is not a solution that should even be considered. Its >great as a "quick hack" to get the kernel running, but nothing more than >that. > >> Okay, are you trying to get this changed in the base kernel? > >No - for a couple of reasons: > >1. its "ugly". Basically, if there is a bug, and the fix is more ugly > than the bug itself, the bug will remain. If the fix is less ugly > than the bug, then the fix goes in. > >2. We believe that we've fixed your problem in the latest test6-rmk > patch, and test7 pre patches. > >Therefore, the more correct fix that I proposed is already in, but still >needs you guys to say "it fixed it". > _____ > |_____| ------------------------------------------------- ---+---+- > | | Russell King linux@arm.linux.org.uk --- --- > | | | | http://www.arm.linux.org.uk/ / / | > | +-+-+ --- -+- > / | THE developer of ARM Linux |+| /|\ > / | | | --- | > +-+-+ ------------------------------------------------- /\\\ | > > >unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.lin >ux.org.uk unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Sun Aug 20 22:27:06 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13QcaJ-00087Y-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Sun, 20 Aug 2000 22:24:51 +0100 Received: from [194.154.183.10] (helo=www.freecheck.co.uk) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13QcaH-00087S-00 for linux-arm-kernel@lists.arm.linux.org.uk; Sun, 20 Aug 2000 22:24:49 +0100 Received: from atm26 (cust34.freecheck.co.uk [194.154.183.34]) by www.freecheck.co.uk (8.9.3/8.8.7) with SMTP id WAA09376 for ; Sun, 20 Aug 2000 22:14:52 GMT From: theo@markettos.org.uk (Theo Markettos) To: linux-arm-kernel@lists.arm.linux.org.uk Subject: Archimedes page tables in 3/4 MEMC1 system Date: Sun, 20 Aug 2000 21:55:43 +0100 Message-ID: <20000820.215543.70@markettos.org.uk> X-Mailer: TTFN version 0.43 (Acorn RISC OS) X-Editor: Zap, using ZapEmail 0.22 (08 Oct 1998) X-Posting-Agent: RISC OS Newsbase 0.60m User-Agent: POPstar/2.02 Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk Does Linux work on the A540 with 12 or 16MB? The reason I ask is I'm trying to figure out how that memory system works, but don't have a technical reference manual. How are the logical-physical address translators programmed if there's more than one slave MEMC? The MEMC databook (ARM FDM) only seems to cover the case where there's one slave MEMC, in which case A[7] is used to choose between master and slave, and is don't care otherwise. But there are no other don't cares in the address translator entry, so where does the bit to choose between MEMCs 0/1 and MEMCs 2/3 come in? Or have Acorn done something odd in their memory system to fudge this (but how do they squeeze the translators into the memory map?) Does anyone have an A540 schematic handy? Thanks, Theo -- Theo Markettos theo.markettos@cai.cam.ac.uk Liphook theo@markettos.org.uk Hampshire UK http://www.markettos.org.uk/ unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Sun Aug 20 23:13:48 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13QdL4-0000Rd-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Sun, 20 Aug 2000 23:13:10 +0100 Received: from dyn-33.linux.theplanet.co.uk ([195.92.244.33] helo=caramon.arm.linux.org.uk) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13QdL1-0000RT-00; Sun, 20 Aug 2000 23:13:08 +0100 Received: from flint.arm.linux.org.uk (root@flint [192.168.0.4]) by caramon.arm.linux.org.uk (8.9.3/8.9.3) with ESMTP id XAA22141; Sun, 20 Aug 2000 23:00:44 +0100 Received: (from linux@localhost) by flint.arm.linux.org.uk (8.9.3/8.9.3) id WAA02329; Sun, 20 Aug 2000 22:57:41 +0100 From: Russell King - ARM Linux Admin Message-Id: <200008202157.WAA02329@flint.arm.linux.org.uk> Subject: Re: Archimedes page tables in 3/4 MEMC1 system To: theo@markettos.org.uk (Theo Markettos) Date: Sun, 20 Aug 2000 22:57:41 +0100 (BST) Cc: linux-arm-kernel@lists.arm.linux.org.uk In-Reply-To: <20000820.215543.70@markettos.org.uk> from "Theo Markettos" at Aug 20, 2000 09:55:43 PM X-Mailer: ELM [version 2.5 PL1] MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk Theo Markettos writes: > Does Linux work on the A540 with 12 or 16MB? The reason I ask is I'm trying > to figure out how that memory system works, but don't have a technical > reference manual. How are the logical-physical address translators > programmed if there's more than one slave MEMC? None of the MEMCs are in slave mode - they're all in master mode. There is however a PLD which controls them, and the accesses to them. It may have something to do with the unused bits while writing to the MEMC (14:12), but a machine to fiddle at assembly level would be needed to work this out. The PLD controls A22 and A7 to each MEMC, and takes A25 through A22, A14 through A12 and A7. > and MEMCs 2/3 come in? Or have Acorn done something odd in their memory > system to fudge this (but how do they squeeze the translators into the memory > map?) Yes, its very odd. Maybe one way would be to scope the PLD, and have the assembly code go around writing a constant code in a loop (and monitor all its outputs). Then, change the code its writing (thereby changing the PLDs input) and repeat. Then build up a truth table of whats going on... The PLD is IC39, a 20 pin device, with the following pinout: 1 A25 2 A24 3 A23 4 A22 5 A14 6 A13 7 A12 8 A7 9 nR/W 10 GND 11 unused 12 A7 to memc in slot 3 13 A7 to memc in slot 2 14 A7 to memc in slot 1 15 A7 to memc onboard 16 A22 to memc in slot 3 17 A22 to memc in slot 2 18 A22 to memc in slot 1 19 A22 to memc onboard 20 VCC Feel free to discuss your findings with me. _____ |_____| ------------------------------------------------- ---+---+- | | Russell King linux@arm.linux.org.uk --- --- | | | | http://www.arm.linux.org.uk/ / / | | +-+-+ --- -+- / | THE developer of ARM Linux |+| /|\ / | | | --- | +-+-+ ------------------------------------------------- /\\\ | unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Mon Aug 21 10:17:52 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13Qngi-0002D2-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Mon, 21 Aug 2000 10:16:12 +0100 Received: from [194.237.142.116] (helo=albatross-ext.wise.edt.ericsson.se) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13Qngh-0002Cw-00 for linux-arm-kernel@lists.arm.linux.org.uk; Mon, 21 Aug 2000 10:16:11 +0100 Received: from esealnt406.al.sw.ericsson.se (esealnt406.al.sw.ericsson.se [153.88.251.29]) by albatross.wise.edt.ericsson.se (8.11.0/8.11.0/WIREfire-1.3) with ESMTP id e7L9GDp28232 for ; Mon, 21 Aug 2000 11:16:14 +0200 (MEST) Received: from esealnt406.al.sw.ericsson.se ([153.88.251.29]) by esealnt406.al.sw.ericsson.se with Microsoft SMTPSVC(5.0.2195.1600); Mon, 21 Aug 2000 11:16:13 +0200 Received: from esealnt743.al.sw.ericsson.se ([153.88.251.13]) by esealnt406.al.sw.ericsson.se (NAVIEG 2.1 bld 61) with SMTP id M2000082111161321533 ; Mon, 21 Aug 2000 11:16:13 +0200 Received: by esealnt743.al.sw.ericsson.se with Internet Mail Service (5.5.2651.58) id ; Mon, 21 Aug 2000 11:14:51 +0200 Message-ID: From: "Arjan Opmeer (EMN)" To: "'Brad Parker'" Cc: "'linux-arm-kernel@lists.arm.linux.org.uk'" Subject: USB port (was: Kernel bootsequence (was: Unable to write register s)) Date: Mon, 21 Aug 2000 11:16:10 +0200 MIME-Version: 1.0 X-Mailer: Internet Mail Service (5.5.2651.58) Content-Type: text/plain; charset="iso-8859-1" X-OriginalArrivalTime: 21 Aug 2000 09:16:13.0947 (UTC) FILETIME=[74192CB0:01C00B50] Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk > I never saw the original message which is (partially) quoted here > and I wonder if it ever came across the list (I don't think it did). No, I started this discussion on the general arm-linux mailing list. At one time it became quite kernel specific so Russell moved it quite rightly to this mailing list. > it appears to have to do with a USB controller (SA1111 perhaps?) > which is of some interest to several people on this list. Yes. I am talking about the USB controller on the SA1111 board. I am trying to get the USB port working. At the moment I have the ohci driver modified enough to have it recognize the USB controller and detect one port. But I'll have to look into the interrupt handling stuff as it doesn't seem to detect devices connected to the port. I could use some help with this, so if you know anyone who could offer me some advice, I would appreciate it. Arjan unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Mon Aug 21 11:28:04 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13QonZ-0003JQ-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Mon, 21 Aug 2000 11:27:21 +0100 Received: from [24.147.212.45] (helo=compaq.parker.boston.ma.us) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13QonX-0003JI-00 for linux-arm-kernel@lists.arm.linux.org.uk; Mon, 21 Aug 2000 11:27:19 +0100 Received: from p2.parker.boston.ma.us (p2 [192.245.5.16]) by compaq.parker.boston.ma.us (8.8.5/8.8.5) with ESMTP id GAA17303; Mon, 21 Aug 2000 06:27:21 -0400 Received: from p2.parker.boston.ma.us (localhost [127.0.0.1]) by p2.parker.boston.ma.us (8.9.3/8.8.8) with ESMTP id GAA38608; Mon, 21 Aug 2000 06:27:21 -0400 (EDT) (envelope-from brad@p2.parker.boston.ma.us) Message-Id: <200008211027.GAA38608@p2.parker.boston.ma.us> To: "Arjan Opmeer (EMN)" cc: "'Brad Parker'" , "'linux-arm-kernel@lists.arm.linux.org.uk'" Subject: Re: USB port (was: Kernel bootsequence (was: Unable to write register s)) In-reply-to: Your message of "Mon, 21 Aug 2000 11:16:10 +0200." Date: Mon, 21 Aug 2000 06:27:21 -0400 From: Brad Parker Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk "Arjan Opmeer (EMN)" wrote: ... >Yes. I am talking about the USB controller on the SA1111 board. I am >trying to get the USB port working. At the moment I have the ohci >driver modified enough to have it recognize the USB controller and >detect one port. But I'll have to look into the interrupt handling >stuff as it doesn't seem to detect devices connected to the port. Hi Arjan, Over the past few months several people have indicated they are also working on this. As far as I can tell no one has gotten it to work. (please chime in if anyone has it working!) I think I am very close but can't get sa1111 interrupts to work... they did work on one kernel but stopped on a different rev. Some non-trivial modifications to the OHCI driver (usb-ohci.c) are required. Most of the changes have to do with cache coherency. However, there seems to be some discussion going on about the base ohci driver itself and how to interpret the ohci specification. Someone posted some controversial changed to the base driver recently which seem to improve things for me. I should have interrupts fixed today and I'll post my patched up driver if you like... -brad unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Mon Aug 21 11:43:21 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13Qp2p-0003Z4-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Mon, 21 Aug 2000 11:43:07 +0100 Received: from [194.237.142.116] (helo=albatross-ext.wise.edt.ericsson.se) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13Qp2o-0003Yq-00 for linux-arm-kernel@lists.arm.linux.org.uk; Mon, 21 Aug 2000 11:43:06 +0100 Received: from esealnt409.al.sw.ericsson.se (esealnt409.al.sw.ericsson.se [153.88.251.32]) by albatross.wise.edt.ericsson.se (8.11.0/8.11.0/WIREfire-1.3) with ESMTP id e7LAh8p10433 for ; Mon, 21 Aug 2000 12:43:08 +0200 (MEST) Received: from esealnt409 ([153.88.251.32]) by esealnt409.al.sw.ericsson.se with Microsoft SMTPSVC(5.0.2172.1); Mon, 21 Aug 2000 12:42:51 +0200 Received: from esealnt743.al.sw.ericsson.se ([153.88.251.13]) by esealnt409 (NAVIEG 2.1 bld 61) with SMTP id M2000082112425126178 ; Mon, 21 Aug 2000 12:42:51 +0200 Received: by esealnt743.al.sw.ericsson.se with Internet Mail Service (5.5.2651.58) id ; Mon, 21 Aug 2000 12:41:42 +0200 Message-ID: From: "Arjan Opmeer (EMN)" To: "'Brad Parker'" , "Arjan Opmeer (EMN)" Cc: "'linux-arm-kernel@lists.arm.linux.org.uk'" Subject: RE: USB port (was: Kernel bootsequence (was: Unable to write regi ster s)) Date: Mon, 21 Aug 2000 12:43:02 +0200 MIME-Version: 1.0 X-Mailer: Internet Mail Service (5.5.2651.58) Content-Type: text/plain; charset="iso-8859-1" X-OriginalArrivalTime: 21 Aug 2000 10:42:51.0437 (UTC) FILETIME=[8E0B35D0:01C00B5C] Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk Hi brad, > I should have interrupts fixed today and I'll post my patched > up driver if you like... Oh yes, I would very much like you to do that! I am really only interested in _using_ the USB port on the SA1111 board for a project here at work. But because nobody seemed to be working on it (or so I thought) I tried my hand at creating a driver myself. If I just can use your driver it would save me lots of time which I could use for other stuff like the Ethernet controller on that board or writing a driver for a piece of hardware I wish to connect to the development board. So, please let me know when you have something working and where I might get it. Thank in advance, Arjan unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Mon Aug 21 15:05:23 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13QsBs-0006pm-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Mon, 21 Aug 2000 15:04:40 +0100 Received: from [210.230.34.106] (helo=so-fs.musicpod.co.jp) by www.linux.org.uk with smtp (Exim 3.13 #1) id 13QsBq-0006pd-00 for linux-arm-kernel@lists.arm.linux.org.uk; Mon, 21 Aug 2000 15:04:39 +0100 Received: (qmail 10933 invoked from network); 21 Aug 2000 14:04:39 -0000 Received: from bak.side7.zetabits.com (HELO bak.zetabits.com) (imai@192.168.36.68) by so-fs.musicpod.co.jp with SMTP; 21 Aug 2000 14:04:39 -0000 Date: Mon, 21 Aug 2000 23:04:38 +0900 Message-ID: From: Kunihiko IMAI To: linux-arm-kernel@lists.arm.linux.org.uk Subject: RE: USB port (was: Kernel bootsequence (was: Unable to write regi ster s)) In-Reply-To: In your message of "Mon, 21 Aug 2000 12:43:02 +0200" References: User-Agent: Wanderlust/1.1.1 (Purple Rain) WEMI/1.13.7 (Shimada) CLIME/1.13.6 (=?ISO-2022-JP?B?GyRCQ2YlTj4xGyhC?=) MULE XEmacs/21.1 (patch 6) (Big Bend) (i386-vine-linux) MIME-Version: 1.0 (generated by WEMI 1.13.7 - "Shimada") Content-Type: text/plain; charset=US-ASCII Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk Hello, At Mon, 21 Aug 2000 12:43:02 +0200, Arjan Opmeer (EMN) wrote: > So, please let me know when you have something working and where > I might get it. It is mainly duscussed at linux-usb-devel mailing list. At SA1111, there are two problems: 1. cache coherency and 2. no PCI bus You can get driver of Jamey's version from: At Mon, 31 Jul 2000 08:50:08 -0400, Jamey Hicks wrote: > [3] > http://www.handhelds.org/cgi-bin/viewcvs.cgi/linux/kernel/drivers/usb/usb-oh > ci.c > > [4] > http://www.handhelds.org/cgi-bin/viewcvs.cgi/linux/kernel/drivers/usb/usb-oh > ci.h This driver solves problem 1 above. I wrote an another version of driver: http://www.d2.dion.ne.jp/~bak/linux/SA1111.tar.gz This driver works on SA1111. But not faces cache problem, only escape from it ( with silly way ). This driver only shows an idea of non-PCI implementation. At Mon, 21 Aug 2000 11:16:10 +0200, Arjan Opmeer (EMN) wrote: > But I'll have to look into the interrupt handling > stuff as it doesn't seem to detect devices connected to the port. Which interrupt do you hook? Use nIrqHciM (SA1111 #44). At the latest arm-linux, this interrupt seems as IRQ #93. Latest arm-linux has a handler of SA1111 interrupts. You don't have to special IRQ handler, I think. Or, which kind of machine do you use? Assabet + neponset? Or original arch? If your machine is an original architecture, you should better to check its schematics. I heard from a circuit designer that Intel's reference schematics had a bug of a DMA line with SA1111. (Now it was fixed.) If this bug is not fixed, DMA bus master cannot work, and SA1111 USB cannot work. # I had suffered from this bug for one month! Thanks. _._. __._ _.. . .___ ._. _____ _... ._ _._ _.._. .____ .__. ... . _._ Kunihiko IMAI unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Mon Aug 21 15:09:49 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13QsGi-0006w5-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Mon, 21 Aug 2000 15:09:40 +0100 Received: from [211.44.193.250] (helo=www.palmpalm.com) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13QsGg-0006vw-00 for linux-arm-kernel@lists.arm.linux.org.uk; Mon, 21 Aug 2000 15:09:39 +0100 Received: from alice ([192.168.1.13]) by www.palmpalm.com (8.9.3/8.8.7) with SMTP id XAA01135; Mon, 21 Aug 2000 23:19:24 +0900 Message-ID: <001701c00b79$de918140$0d01a8c0@palmpalm.com> From: "Hyun-Mo Chung" To: "'Brad Parker'" Cc: References: <200008211027.GAA38608@p2.parker.boston.ma.us> Subject: Re: USB port (was: Kernel bootsequence (was: Unable to write register s)) Date: Mon, 21 Aug 2000 23:12:41 +0900 MIME-Version: 1.0 Content-Type: text/plain; charset="Windows-1252" Content-Transfer-Encoding: 7bit X-Priority: 3 X-MSMail-Priority: Normal X-Mailer: Microsoft Outlook Express 5.00.2919.6600 X-MimeOLE: Produced By Microsoft MimeOLE V5.00.2919.6600 Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk Hi, I'm also interested in this subject. > Over the past few months several people have indicated they are also > working on this. As far as I can tell no one has gotten it to work. > (please chime in if anyone has it working!) I agree. There seems to be no ohci driver working on SA1111 yet. I remember Kunihiko IMAI reported some weeks ago that his USB driver was working happily on SA1111. But I guess his driver may have a few problems as well. (oops problem mentioned below, for example.) As for my work, Logitech USB mouse is now working well with SA1111, except the chronic oops when unplugging the mouse. The revision I'm working on is test1-ac7-rmk1-np6. > Someone posted some controversial changed to the base driver recently > which seem to improve things for me. Would you give me a direction as to where I can find it ? Thank you. Hyun-Mo unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Mon Aug 21 21:18:19 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13Qy0d-000510-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Mon, 21 Aug 2000 21:17:27 +0100 Received: from [195.40.1.44] (helo=chalfont.mail.uk.easynet.net) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13Qy0c-00050q-00 for linux-arm-kernel@lists.arm.linux.org.uk; Mon, 21 Aug 2000 21:17:26 +0100 Received: from atm26 (tnt-17-25.easynet.co.uk [212.134.30.25]) by chalfont.mail.uk.easynet.net (Postfix) with SMTP id 6817FF83E0; Mon, 21 Aug 2000 21:17:21 +0100 (BST) To: linux@arm.linux.org.uk Cc: linux-arm-kernel@lists.arm.linux.org.uk From: theo@markettos.org.uk (Theo Markettos) Subject: Re: Archimedes page tables in 3/4 MEMC1 system In-Reply-To: <200008202157.WAA02329@flint.arm.linux.org.uk> Date: Mon, 21 Aug 2000 20:04:34 +0100 Message-ID: <20000821.200434.52@markettos.org.uk> X-Mailer: TTFN version 0.43 (Acorn RISC OS) X-Editor: Zap, using ZapEmail 0.22 (08 Oct 1998) X-Posting-Agent: RISC OS Newsbase 0.60m User-Agent: POPstar/2.02 Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk On Sun, 20 Aug 2000 22:57:41 +0100 (BST), linux@arm.linux.org.uk said: > None of the MEMCs are in slave mode - they're all in master mode. There is > however a PLD which controls them, and the accesses to them. It may have > something to do with the unused bits while writing to the MEMC (14:12), but > a machine to fiddle at assembly level would be needed to work this out. > The PLD controls A22 and A7 to each MEMC, and takes A25 through A22, A14 > through A12 and A7. OK... it sounds like they're using A7 as an enable line when writing page tables. As to what they're doing with A22, that's bit 7 of the logical page number, so I have no idea why they would want to mess with that... I don't actually have a machine to play with - my random thought was to synthesize in a PLD or FPGA as much of a MEMC as was required to do the page tables and DRAM refresh, and make a 16MB RAM upgrade for my A4. Having seen how tight space is inside the A4, and having to piggyback both MEMC and VIDC to get the signals, it's probably not easy mechanically if it was doable electrically. Still, it's an interesting thought experiment... Thanks, Theo -- Theo Markettos theo.markettos@cai.cam.ac.uk Liphook theo@markettos.org.uk Hampshire UK http://www.markettos.org.uk/ unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Tue Aug 22 14:13:04 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13RDoT-0004hZ-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Tue, 22 Aug 2000 14:09:57 +0100 Received: from [216.85.232.66] (helo=blazieresearch.com) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13RDoS-0004hT-00 for linux-arm-kernel@lists.arm.linux.org.uk; Tue, 22 Aug 2000 14:09:56 +0100 Received: from localhost (chris@localhost) by blazieresearch.com (8.9.3/8.8.7) with SMTP id JAA01854; Tue, 22 Aug 2000 09:09:57 -0400 Date: Tue, 22 Aug 2000 09:09:57 -0400 (EDT) From: Chris Blazie To: Nicolas Pitre cc: linux-arm-kernel@lists.arm.linux.org.uk, Nicolas Pitre Subject: Re: slower decompress? In-Reply-To: Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk > > The dots are appearing at a slow pace. What should the status of > > the mmu and icache be when decompress_kernel is called? > > MMU, i-cache and d-cache are actually enabled for decompression speed in > linux/arch/arm/compressed/head.S. > > Just tried it again on my Assabet and everything looks OK here. > What bogomips number do you get? I'm getting 124.52 bogomips. Sound about right? Chris unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Tue Aug 22 14:46:00 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13RELf-0005Ft-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Tue, 22 Aug 2000 14:44:15 +0100 Received: from [208.225.80.11] (helo=ice.ar51.net) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13RELe-0005Fm-00 for linux-arm-kernel@lists.arm.linux.org.uk; Tue, 22 Aug 2000 14:44:14 +0100 Received: from jSutherland (akron-80-174.ar51.net [208.225.80.174]) by ice.ar51.net (8.9.3/8.8.7) with SMTP id JAA24837; Tue, 22 Aug 2000 09:44:16 -0400 Reply-To: From: "Jeff Sutherland" To: "'Chris Blazie'" Cc: Subject: RE: slower decompress? Date: Tue, 22 Aug 2000 09:46:46 -0400 Message-ID: MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit X-Priority: 3 (Normal) X-MSMail-Priority: Normal X-Mailer: Microsoft Outlook CWS, Build 9.0.2416 (9.0.2910.0) In-reply-to: X-MimeOLE: Produced By Microsoft MimeOLE V5.00.2919.6700 Importance: Normal Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk > > > The dots are appearing at a slow pace. What should the > status of > > > the mmu and icache be when decompress_kernel is called? > > > > MMU, i-cache and d-cache are actually enabled for > decompression speed in > > linux/arch/arm/compressed/head.S. > > > > Just tried it again on my Assabet and everything looks OK here. > > What bogomips number do you get? > > I'm getting 124.52 bogomips. Sound about right? > > Chris Looks a little on the low side to me. Which version of the Assabet do you have? Early versions up to phase 4 were designed to run at 147.5MHz if you want the 1341 audio codec to work properly. I ran mine all the way up to 221MHz at one point (works as long as the SDRAM clock is running at CPU/4, not CPU/2). A bug in the PLD for the lcd mandates running SDRAM at CPU/4 anyway. But, even at 124.52Bogomips (CPU should be somewhere around 132.7MHz) it still moves along pretty well. It should take maybe 1.5 seconds to decompress at this frequency if the Icache and buffering are enabled. If you are fooling around with the mmu and the buffer settings beware that there are only a few valid combinations. Turning on debugging in the low level assembler routines turns off the mmu and buffering, which gives a drastic reduction in performance. //Jeff unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Tue Aug 22 14:47:24 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13RENC-0005HM-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Tue, 22 Aug 2000 14:45:50 +0100 Received: from [205.250.170.220] (helo=xanadu.vipswitch.com) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13RENB-0005HG-00 for linux-arm-kernel@lists.arm.linux.org.uk; Tue, 22 Aug 2000 14:45:49 +0100 Date: Tue, 22 Aug 2000 09:43:47 -0400 (EDT) From: Nicolas Pitre To: Chris Blazie cc: linux-arm-kernel@lists.arm.linux.org.uk, Nicolas Pitre Subject: Re: slower decompress? In-Reply-To: Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk On Tue, 22 Aug 2000, Chris Blazie wrote: > > > > The dots are appearing at a slow pace. What should the status of > > > the mmu and icache be when decompress_kernel is called? > > > > MMU, i-cache and d-cache are actually enabled for decompression speed in > > linux/arch/arm/compressed/head.S. > > > > Just tried it again on my Assabet and everything looks OK here. > > What bogomips number do you get? > > I'm getting 124.52 bogomips. Sound about right? For a 133MHz part, yes. Nicolas unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Tue Aug 22 15:28:00 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13RF0e-0005wc-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Tue, 22 Aug 2000 15:26:36 +0100 Received: from [216.85.232.66] (helo=blazieresearch.com) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13RF0c-0005wU-00 for linux-arm-kernel@lists.arm.linux.org.uk; Tue, 22 Aug 2000 15:26:35 +0100 Received: from localhost (chris@localhost) by blazieresearch.com (8.9.3/8.8.7) with SMTP id KAA03401; Tue, 22 Aug 2000 10:26:36 -0400 Date: Tue, 22 Aug 2000 10:26:36 -0400 (EDT) From: Chris Blazie Reply-To: Chris Blazie To: Jeff Sutherland cc: linux-arm-kernel@lists.arm.linux.org.uk Subject: RE: slower decompress? In-Reply-To: Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk > > I'm getting 124.52 bogomips. Sound about right? > > > > Chris > > Looks a little on the low side to me. Which version of the Assabet do you have? Early versions up to phase 4 were designed to run It's not an Assabet, just a custom development board (SA-1110) which we're clocking at 133mhz and SDRAM divisor is CPU/4. But again, with the 2.3.99-pre3 kernel it decompressed in about 1 second. I have debugging turned off, and just verified that immediately before decompress_kernel is called, both icache and mmu are turned on. What else could have changed here? Could the page table setup in head.S affect this? Thanks, Chris unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Tue Aug 22 16:31:20 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13RG11-0006zv-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Tue, 22 Aug 2000 16:31:03 +0100 Received: from [205.250.170.220] (helo=xanadu.vipswitch.com) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13RG0y-0006zg-00 for linux-arm-kernel@lists.arm.linux.org.uk; Tue, 22 Aug 2000 16:31:00 +0100 Date: Tue, 22 Aug 2000 11:30:00 -0400 (EDT) From: Nicolas Pitre To: Chris Blazie cc: Jeff Sutherland , linux-arm-kernel@lists.arm.linux.org.uk Subject: RE: slower decompress? In-Reply-To: Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk On Tue, 22 Aug 2000, Chris Blazie wrote: > I have debugging turned off, and just verified that immediately > before decompress_kernel is called, both icache and mmu are turned on. > > What else could have changed here? Could the page table setup > in head.S affect this? Are you executing zImage from RAM or directly from Flash? Nicolas unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Tue Aug 22 16:31:26 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13RG0P-0006zH-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Tue, 22 Aug 2000 16:30:25 +0100 Received: from [216.85.232.66] (helo=blazieresearch.com) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13RG0N-0006z7-00 for linux-arm-kernel@lists.arm.linux.org.uk; Tue, 22 Aug 2000 16:30:23 +0100 Received: from localhost (chris@localhost) by blazieresearch.com (8.9.3/8.8.7) with SMTP id LAA04489 for ; Tue, 22 Aug 2000 11:29:52 -0400 Date: Tue, 22 Aug 2000 11:29:52 -0400 (EDT) From: Chris Blazie Reply-To: Chris Blazie To: linux-arm-kernel@lists.arm.linux.org.uk Subject: RE: slower decompress? -- got it In-Reply-To: Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk Okay, I answered my own question (partially); my problem is in the page table setup changes since pre3. I made the following changes to cache_on in head.S: /* * Initialise the page tables */ mov r0, r3 mov r8, r0, lsr #18 mov r8, r8, lsl #18 @ start of RAM add r9, r8, #0x20000000 @ the maximum RAM size mov r1, #0x1e // CHANGED THIS from 0x12 orr r1, r1, #3 << 10 add r2, r3, #16384 // Commented out next 4 lines 1: @cmp r1, r8 @ if virt > start of RAM @orrge r1, r1, #0x0c @ set cacheable, bufferable @cmp r1, r9 @ if virt > end of RAM @bicge r1, r1, #0x0c @ clear cacheable, bufferable str r1, [r0], #4 @ 1:1 mapping add r1, r1, #1048576 teq r0, r2 bne 1b And it decompresses normally now (about a second). So it seems that the test for virt > start of RAM is failing, and cachable/bufferable is not getting set at all. Can someone explain what these setup changes are about? I'm guessing one of the entry registers is incorrectly set in my case, but not really sure right now. Thanks, Chris unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Tue Aug 22 16:48:55 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13RGI1-0007OO-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Tue, 22 Aug 2000 16:48:37 +0100 Received: from [128.2.10.83] (helo=smtp3.andrew.cmu.edu) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13RGHz-0007Np-00 for linux-arm-kernel@lists.arm.linux.org.uk; Tue, 22 Aug 2000 16:48:36 +0100 Received: from cs.cmu.edu (NINESIXTEEN.REM.CMU.EDU [128.2.81.157]) by smtp3.andrew.cmu.edu (8.9.3/8.9.3) with ESMTP id LAA18428; Tue, 22 Aug 2000 11:48:22 -0400 (EDT) Message-ID: <39A2A06F.6B6B9FA5@cs.cmu.edu> Date: Tue, 22 Aug 2000 11:46:56 -0400 From: John G Dorsey Organization: Carnegie Mellon University X-Mailer: Mozilla 4.72 [en] (X11; U; Linux 2.2.14-5.0 i686) X-Accept-Language: en MIME-Version: 1.0 To: "Chagas, Jason" CC: linux-arm-kernel@lists.arm.linux.org.uk Subject: Re: Assabet / Simultaneous audio recording & play ?? References: Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk [cross-posting pared down] "Chagas, Jason" wrote: > I'm wondering if the audio driver for the [SA-1111 Serial Audio Controller] > Philips UDA1341 codec supports simultaneous audio play and recording. Has > anyone actually tested it? I haven't tested it. (Off the top of my head, it seems like it should. Historically, however, that's a near guarantee that it won't. =) That driver still has a ways to go, in several respects. First, the DMA code hasn't been tweaked to acknowledge the documented bugs in SAC DMA transfers from SDRAM. Second, the UDA1341 driver in general doesn't really implement the OSS API. Third, there are still weird audio artifacts on Neponset that I'm not sure are explained by the DMA bugs from the errata. My next few days are spoken for already, but after that I'll probably begin whittling away at this code again. -jd unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Tue Aug 22 17:10:52 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13RGc4-0007oY-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Tue, 22 Aug 2000 17:09:20 +0100 Received: from [216.85.232.66] (helo=blazieresearch.com) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13RGc3-0007oP-00 for linux-arm-kernel@lists.arm.linux.org.uk; Tue, 22 Aug 2000 17:09:20 +0100 Received: from localhost (chris@localhost) by blazieresearch.com (8.9.3/8.8.7) with SMTP id MAA05329; Tue, 22 Aug 2000 12:09:21 -0400 Date: Tue, 22 Aug 2000 12:09:21 -0400 (EDT) From: Chris Blazie To: Nicolas Pitre cc: linux-arm-kernel@lists.arm.linux.org.uk Subject: RE: slower decompress? In-Reply-To: Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk > Are you executing zImage from RAM or directly from Flash? It's executing directly from flash (0x00050000 to be exact). Chris unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Tue Aug 22 17:36:49 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13RH2I-0008PR-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Tue, 22 Aug 2000 17:36:26 +0100 Received: from [205.250.170.220] (helo=xanadu.vipswitch.com) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13RH2H-0008PI-00 for linux-arm-kernel@lists.arm.linux.org.uk; Tue, 22 Aug 2000 17:36:25 +0100 Date: Tue, 22 Aug 2000 12:35:25 -0400 (EDT) From: Nicolas Pitre To: Chris Blazie cc: linux-arm-kernel@lists.arm.linux.org.uk Subject: RE: slower decompress? In-Reply-To: Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk On Tue, 22 Aug 2000, Chris Blazie wrote: > > > Are you executing zImage from RAM or directly from Flash? > > It's executing directly from flash (0x00050000 to be exact). Ha! That's the reason. The code in head.S is only activating the cache from the start of RAM. The solution would be also to activate the cache for 2 MB from the run-time address of "start". If the kernel is started from RAM, the cache will simply be enabled twice. Nicolas unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Tue Aug 22 18:05:34 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13RHUE-0000mS-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Tue, 22 Aug 2000 18:05:18 +0100 Received: from [205.250.170.220] (helo=xanadu.vipswitch.com) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13RHUD-0000mK-00 for linux-arm-kernel@lists.arm.linux.org.uk; Tue, 22 Aug 2000 18:05:17 +0100 Date: Tue, 22 Aug 2000 13:04:48 -0400 (EDT) From: Nicolas Pitre To: John G Dorsey cc: "Chagas, Jason" , linux-arm-kernel@lists.arm.linux.org.uk Subject: Re: Assabet / Simultaneous audio recording & play ?? In-Reply-To: <39A2A06F.6B6B9FA5@cs.cmu.edu> Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk On Tue, 22 Aug 2000, John G Dorsey wrote: > [cross-posting pared down] > > "Chagas, Jason" wrote: > > > I'm wondering if the audio driver for the [SA-1111 Serial Audio Controller] > > Philips UDA1341 codec supports simultaneous audio play and recording. Has > > anyone actually tested it? > > I haven't tested it. (Off the top of my head, it seems like it should. > Historically, however, that's a near guarantee that it won't. =) Don't know for the UDA1341 on the SA1111, but the SA1110 actually does. Nicolas unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Tue Aug 22 19:55:58 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13RJCK-0003Xx-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Tue, 22 Aug 2000 19:54:56 +0100 Received: from [216.18.11.193] (helo=duke.electric.net) by www.linux.org.uk with smtp (Exim 3.13 #1) id 13RJCI-0003Xn-00 for linux-arm-kernel@lists.arm.linux.org.uk; Tue, 22 Aug 2000 19:54:55 +0100 Received: (qmail 13500 invoked from network); 22 Aug 2000 18:54:57 -0000 Received: from localhost (HELO relay.electric.net) (@127.0.0.1) by duke.electric.net with SMTP; 22 Aug 2000 18:54:57 -0000 Received: (from smtpd@localhost) by relay.electric.net (8.8.8/8.8.6) id LAA13455 for linux-arm-kernel@lists.arm.linux.org.uk; Tue, 22 Aug 2000 11:54:57 -0700 (PDT) Received: from cobra.intrinsyc.com(64.16.10.158), claiming to be "dp.intrinsyc.com" via SMTP by relay.electric.net, id smtpdQ13435; Tue Aug 22 11:54:55 2000 Received: from ileong (helo=localhost) by dp.intrinsyc.com with local-esmtp (Exim 3.13 #1 (Debian)) id 13RJCI-0002e1-00 for ; Tue, 22 Aug 2000 11:54:54 -0700 Date: Tue, 22 Aug 2000 11:54:54 -0700 (PDT) From: I-Gene Leong To: linux-arm-kernel@lists.arm.linux.org.uk Subject: Recent kernels and the Cerf... Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk Seems to me that: test4-rmk1-np1 was fine test5-rmk1-np1 broke PCMCIA test6-rmk2-np1 fixed PCMCIA but broke something so that start-stop-daemon causes recoverable kernel panics test6-rmk5-np1 doesn't boot at all It's that last one that worries me the most. What's changed between the two? I don't even see a single bootup message. -- I-Gene Leong ileong@intrinsyc.com / 604-646-6962 "We see Linux as being much of a fad as the Internet was in 1995." - Irving Wladawsky-Berger, IBM VP of technology and strategy unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Tue Aug 22 20:45:04 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13RJxG-0004GN-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Tue, 22 Aug 2000 20:43:26 +0100 Received: from [205.250.170.220] (helo=xanadu.vipswitch.com) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13RJxF-0004GD-00 for linux-arm-kernel@lists.arm.linux.org.uk; Tue, 22 Aug 2000 20:43:25 +0100 Date: Tue, 22 Aug 2000 15:42:51 -0400 (EDT) From: Nicolas Pitre To: I-Gene Leong cc: linux-arm-kernel@lists.arm.linux.org.uk Subject: Re: Recent kernels and the Cerf... In-Reply-To: Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk On Tue, 22 Aug 2000, I-Gene Leong wrote: > Seems to me that: > > test4-rmk1-np1 was fine > test5-rmk1-np1 broke PCMCIA > test6-rmk2-np1 fixed PCMCIA but broke something so that start-stop-daemon > causes recoverable kernel panics I experiment the same at the moment. > test6-rmk5-np1 doesn't boot at all It boots for me on Assabet. Unfortunately I don't have a Cerf to test on. > It's that last one that worries me the most. What's changed between the > two? Quite a lot. You can produce a diff between versions to see the details. > I don't even see a single bootup message. That's weird. Not much changed in the initial boot code... Nicolas unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Tue Aug 22 22:17:16 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13RLMe-0005d0-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Tue, 22 Aug 2000 22:13:44 +0100 Received: from [216.18.11.193] (helo=duke.electric.net) by www.linux.org.uk with smtp (Exim 3.13 #1) id 13RLMc-0005cc-00 for linux-arm-kernel@lists.arm.linux.org.uk; Tue, 22 Aug 2000 22:13:43 +0100 Received: (qmail 56831 invoked from network); 22 Aug 2000 21:13:41 -0000 Received: from localhost (HELO relay.electric.net) (@127.0.0.1) by duke.electric.net with SMTP; 22 Aug 2000 21:13:41 -0000 Received: (from smtpd@localhost) by relay.electric.net (8.8.8/8.8.6) id OAA56822; Tue, 22 Aug 2000 14:13:41 -0700 (PDT) Received: from cobra.intrinsyc.com(64.16.10.158), claiming to be "dp.intrinsyc.com" via SMTP by relay.electric.net, id smtpdS56741; Tue Aug 22 14:13:40 2000 Received: from ileong (helo=localhost) by dp.intrinsyc.com with local-esmtp (Exim 3.13 #1 (Debian)) id 13RLMW-0001xB-00; Tue, 22 Aug 2000 14:13:36 -0700 Date: Tue, 22 Aug 2000 14:13:36 -0700 (PDT) From: I-Gene Leong To: Nicolas Pitre cc: linux-arm-kernel@lists.arm.linux.org.uk Subject: Re: Recent kernels and the Cerf... In-Reply-To: Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk On Tue, 22 Aug 2000, Nicolas Pitre wrote: > On Tue, 22 Aug 2000, I-Gene Leong wrote: > > Seems to me that: > > test4-rmk1-np1 was fine > > test5-rmk1-np1 broke PCMCIA > > test6-rmk2-np1 fixed PCMCIA but broke something so that start-stop-daemon > > causes recoverable kernel panics > I experiment the same at the moment. Good to know it's not just here that's messed up. :) > > test6-rmk5-np1 doesn't boot at all > It boots for me on Assabet. Unfortunately I don't have a Cerf to test on. I reconfigured it without a lot of stuff and to use a ramdisk as root and it booted up fine. I enabled PCMCIA support and I got the same kernel panics as test6-rmk2-np1. Interesting... well, I'm going to compile in some other stuff and see if everything else comes up ok. -- I-Gene Leong ileong@intrinsyc.com / 604-646-6962 "We see Linux as being much of a fad as the Internet was in 1995." - Irving Wladawsky-Berger, IBM VP of technology and strategy unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Tue Aug 22 22:22:04 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13RLTB-0005j5-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Tue, 22 Aug 2000 22:20:29 +0100 Received: from dyn-33.linux.theplanet.co.uk ([195.92.244.33] helo=caramon.arm.linux.org.uk) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13RLTA-0005io-00; Tue, 22 Aug 2000 22:20:29 +0100 Received: from flint.arm.linux.org.uk (root@flint [192.168.0.4]) by caramon.arm.linux.org.uk (8.9.3/8.9.3) with ESMTP id TAA30414; Tue, 22 Aug 2000 19:24:46 +0100 Received: (from linux@localhost) by flint.arm.linux.org.uk (8.9.3/8.9.3) id TAA12632; Tue, 22 Aug 2000 19:21:34 +0100 From: Russell King - ARM Linux Admin Message-Id: <200008221821.TAA12632@flint.arm.linux.org.uk> Subject: Re: slower decompress? -- got it To: chris@blazie.com Date: Tue, 22 Aug 2000 19:21:34 +0100 (BST) Cc: linux-arm-kernel@lists.arm.linux.org.uk In-Reply-To: from "Chris Blazie" at Aug 22, 2000 11:29:52 AM X-Mailer: ELM [version 2.5 PL1] MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk Chris Blazie writes: > And it decompresses normally now (about a second). So it seems > that the test for virt > start of RAM is failing, and cachable/bufferable > is not getting set at all. Can someone explain what these setup changes > are about? I'm guessing one of the entry registers is incorrectly set in > my case, but not really sure right now. Basically, its to ensure that only the RAM is marked cacheable, and not the IO. If the IO is marked cacheable/bufferable, then Bad Things (tm) happen. _____ |_____| ------------------------------------------------- ---+---+- | | Russell King linux@arm.linux.org.uk --- --- | | | | http://www.arm.linux.org.uk/ / / | | +-+-+ --- -+- / | THE developer of ARM Linux |+| /|\ / | | | --- | +-+-+ ------------------------------------------------- /\\\ | unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Tue Aug 22 22:59:32 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13RM4E-0006Jd-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Tue, 22 Aug 2000 22:58:46 +0100 Received: from [205.250.170.220] (helo=xanadu.vipswitch.com) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13RM4D-0006JS-00 for linux-arm-kernel@lists.arm.linux.org.uk; Tue, 22 Aug 2000 22:58:45 +0100 Date: Tue, 22 Aug 2000 17:57:35 -0400 (EDT) From: Nicolas Pitre To: Chris Blazie cc: linux-arm-kernel@lists.arm.linux.org.uk Subject: RE: slower decompress? In-Reply-To: Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk On Tue, 22 Aug 2000, Chris Blazie wrote: > > > Ha! That's the reason. > > > > The code in head.S is only activating the cache from the start of > > RAM. The solution would be also to activate the cache for 2 MB from > > the run-time address of "start". > > I'm not sure what you mean by "start" here -- you mean the > execution address in Flash or the start address in RAM (0xc0008000) or > something else? Also is there a way to implement a generic solution here? > I'm surely not the only one executing zImage from Flash. Here's what could be done. Beware that I didn't test it yet. --- orig/linux/arch/arm/boot/compressed/head.S Tue Aug 15 15:02:07 2000 +++ linux/arch/arm/boot/compressed/head.S Tue Aug 22 17:12:47 2000 @@ -198,7 +198,6 @@ 1: sub r3, r4, #16384 @ Page directory size bic r3, r3, #0xff @ Align the pointer - bic r3, r3, #0x3f /* * Initialise the page tables */ @@ -217,6 +216,20 @@ add r1, r1, #1048576 teq r0, r2 bne 1b +/* + * If ever we are running from Flash, then we surely want the cache + * to be enabled also for our execution instance... We map 2MB of it + * so there is no overlap problem for up to 1 MB compressed kernel. + * If the execution is in RAM then we'll only duplicate the above. + */ + mov r1, #0x1e + orr r1, r1, #3 << 10 + mov r2, pc, lsr #20 + orr r1, r1, r2, lsl #20 + add r0, r3, r2, lsl #2 + str r1, [r0], #4 + add r1, r1, #1048576 + str r1, [r0] mov r0, #0 mcr p15, 0, r0, c7, c10, 4 @ drain write buffer unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Tue Aug 22 23:42:15 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13RMjp-0006wT-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Tue, 22 Aug 2000 23:41:45 +0100 Received: from dyn-33.linux.theplanet.co.uk ([195.92.244.33] helo=caramon.arm.linux.org.uk) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13RMjj-0006wG-00; Tue, 22 Aug 2000 23:41:40 +0100 Received: from flint.arm.linux.org.uk (root@flint [192.168.0.4]) by caramon.arm.linux.org.uk (8.9.3/8.9.3) with ESMTP id XAA31200; Tue, 22 Aug 2000 23:41:42 +0100 Received: (from linux@localhost) by flint.arm.linux.org.uk (8.9.3/8.9.3) id XAA13443; Tue, 22 Aug 2000 23:38:29 +0100 From: Russell King - ARM Linux Admin Message-Id: <200008222238.XAA13443@flint.arm.linux.org.uk> Subject: Re: slower decompress? To: nico@cam.org (Nicolas Pitre) Date: Tue, 22 Aug 2000 23:38:29 +0100 (BST) Cc: chris@blazie.com (Chris Blazie), linux-arm-kernel@lists.arm.linux.org.uk In-Reply-To: from "Nicolas Pitre" at Aug 22, 2000 05:57:35 PM X-Mailer: ELM [version 2.5 PL1] MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk Nicolas Pitre writes: > Here's what could be done. Beware that I didn't test it yet. > > --- orig/linux/arch/arm/boot/compressed/head.S Tue Aug 15 15:02:07 2000 > +++ linux/arch/arm/boot/compressed/head.S Tue Aug 22 17:12:47 2000 > @@ -198,7 +198,6 @@ > 1: > sub r3, r4, #16384 @ Page directory size > bic r3, r3, #0xff @ Align the pointer > - bic r3, r3, #0x3f Oops, there's a couple of '0's missing off the end of that line, so it should probably be: - bic r3, r3, #0x3f + bic r3, r3, #0x3f00 _____ |_____| ------------------------------------------------- ---+---+- | | Russell King linux@arm.linux.org.uk --- --- | | | | http://www.arm.linux.org.uk/ / / | | +-+-+ --- -+- / | THE developer of ARM Linux |+| /|\ / | | | --- | +-+-+ ------------------------------------------------- /\\\ | unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Wed Aug 23 13:46:50 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13RZsz-0000OT-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Wed, 23 Aug 2000 13:44:05 +0100 Received: from [211.44.193.250] (helo=www.palmpalm.com) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13RZsy-0000OF-00 for linux-arm-kernel@lists.arm.linux.org.uk; Wed, 23 Aug 2000 13:44:04 +0100 Received: from alice ([192.168.1.13]) by www.palmpalm.com (8.9.3/8.8.7) with SMTP id VAA06202 for ; Wed, 23 Aug 2000 21:06:58 +0900 Message-ID: <00e901c00cf9$78a8eae0$0d01a8c0@palmpalm.com> From: "Hyun-Mo Chung" To: References: <200007132127.WAA08910@flint.arm.linux.org.uk> Subject: proc-arm720.S: Apostrophe in comment makes assember unhappy Date: Wed, 23 Aug 2000 20:58:36 +0900 MIME-Version: 1.0 Content-Type: text/plain; charset="Windows-1252" Content-Transfer-Encoding: 7bit X-Priority: 3 X-MSMail-Priority: Normal X-Mailer: Microsoft Outlook Express 5.00.2919.6600 X-MimeOLE: Produced By Microsoft MimeOLE V5.00.2919.6600 Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk Hi, During a compile of test6-rmk5, I observed arch/arm/mm/proc-arm720.S causing an assembler error as follows : proc-arm720.S:165: unterminated character constant The line in problem is : beq Ldata_simple @ don't have to do anything if zero Assembler became quiet after the apostrophe in had been ommited. Please let me know what degree of test proc-arm720.S has taken. I need to know it in order to decide how much I can reply on the code. Thank you. Hyun-Mo unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Wed Aug 23 14:17:31 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13RaNl-00012g-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Wed, 23 Aug 2000 14:15:53 +0100 Received: from [139.175.250.58] (helo=mail2000.com.tw) by www.linux.org.uk with smtp (Exim 3.13 #1) id 13RaNj-00012E-00 for linux-arm-kernel@lists.arm.linux.org.uk; Wed, 23 Aug 2000 14:15:52 +0100 Received: By OpenMail Mailer;Wed, 23 Aug 2000 17:31:55 +0800 (CST) From: "Chien-Min Lee" Reply-To: cm_lee@mail2000.com.tw Subject: What does "external abort on non-linefetch" mean? Message-ID: <967023115.87257.cm_lee@mail2000.com.tw> To: "linux-arm-kernel" Date: Wed, 23 Aug 2000 17:31:55 +0800 (CST) MIME-Version: 1.0 Content-Type: text/plain Content-Transfer-Encoding: 8bit Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk Dear All: I got the following error when testing a driver on arm-linux. [root@Linux /root]$Internal error: external abort on non-linefetch: 28 CPU: 0 pc : [] lr : [] sp : c012ff68 ip : 00000000 fp : c012ffc0 r10: c001307c r9 : ffffffff r8 : 80000013 r7 : c00203ac r6 : c0019b1c r5 : c012ffb0 r4 : fa050000 r3 : c01431c4 r2 : 00000000 r1 : 00000000 r0 : 00000003 Flags: Nzcv IRQs off FIQs on Mode SVC_32 Segment kernel Control: C165917F Table: C165917F DAC: 0000001D Process swapper (pid: 0, stackpage=c012f000) Stack: [skip .....] However, I do not understand this error message. Could someone tell me what it talks about?? Thanks kirk ----== Mailed via Openfind ==----- http://mail2000.com.tw/ ´£¨Ñ§K¶O¹q¤l¶l¥ó«H½c unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Wed Aug 23 14:53:30 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13Raxg-0001hW-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Wed, 23 Aug 2000 14:53:00 +0100 Received: from [211.62.252.68] (helo=s-mail8.hanmail.net ident=root) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13Raxf-0001hM-00 for linux-arm-kernel@lists.arm.linux.org.uk; Wed, 23 Aug 2000 14:52:59 +0100 Received: from www37.hanmail.net (www37.hanmail.net [211.32.117.217]) by s-mail8.hanmail.net (8.10.0/8.9.1) with ESMTP id e7NDpsT15862; Wed, 23 Aug 2000 22:51:54 +0900 Received: (from hanadmin@localhost) by www37.hanmail.net (8.10.0/8.9.1) id e7NDsXx03978; Wed, 23 Aug 2000 22:54:33 +0900 (KST) X-Originating-IP: [211.44.193.196] From: "Byung-Soo Kim" Reply-To: "Byung-Soo Kim" Organization: To: linux-arm-kernel@lists.arm.linux.org.uk Subject: Source of Informaion on devfs X-Mailer: Daum Web Mailer 1.0 Date: Wed, 23 Aug 2000 22:54:33 KST Message-Id: <20000823225433.HM.a00000000001QZ8@www37.hanmail.net> MIME-Version: 1.0 Content-Type: text/plain; charset=iso-2022-kr Content-Transfer-Encoding: 7bit Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk Dear experts, Sorry this question is not arm-linux specific, but the majordomo of linux-kernel seems to be broken. I'm a kernel newbie who is about to write a device driver. I have read a few books on kernel programming and device driver writing, but found that "real" device drivers in -test6 kernel have a lot of features not described in any book. I believe there exists an official web site maintaining up-to-date information - or history log - on how to use the kernel services such as devfs, kernel timer, wait queue, and so on. Where did authors of the device drivers in -test6 learn the usage of devfs ? Thank you in advance. Byung-Soo unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Wed Aug 23 15:08:34 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13RbCO-0001wN-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Wed, 23 Aug 2000 15:08:12 +0100 Received: from [130.161.42.1] (helo=duteinh.et.tudelft.nl) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13RbCN-0001wH-00 for linux-arm-kernel@lists.arm.linux.org.uk; Wed, 23 Aug 2000 15:08:11 +0100 Received: from dutein75 by duteinh.et.tudelft.nl (8.9.3/BB-2.0) id QAA27389; Wed, 23 Aug 2000 16:07:52 +0200 (MET DST) Message-Id: <200008231407.QAA27389@duteinh.et.tudelft.nl> From: Erik Mouw To: linuxmama@hanmail.net Cc: linux-arm-kernel@lists.arm.linux.org.uk Subject: Re: Source of Informaion on devfs In-Reply-To: <20000823225433.HM.a00000000001QZ8@www37.hanmail.net> Organization: Eric conspiracy secret labs Date: Wed 23 Aug 2000 16:07:50 +0100 (MDT) X-Face: "!Z|1XjFg!3Iw)BwM^M\k)J5KHiPddz}#/Zw%?$&Q!xqob"OA|})3_@&BTTiQERo?/^g[Z<9P4R8:5C:aqq.cw!hotLc=+u-"+1rq@9$V^8PbOUlodyi_sejt+D`iqkLpx$>.aXl%vnfPIt"LK|/%>94_Zrm8jy?`_[RKt'TuoxUxfk(gMhY(n|DBBl~c$dhO>G%6}R4+DGX3|idBm4oU&UM/XUP4I9aoN{]`ax%B?rgD(I3OJ<62Y6L X-Mailer: XCmail 1.2devel - with PGP support, PGP engine version 0.5 (IRIX64) X-Mailerorigin: http://www.fsai.fh-trier.de/~schmitzj/Xclasses/XCmail/ X-Eric-Conspiracy: There is no conspiracy! X-Loop: erik@it.et.tudelft.nl MIME-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: 8bit Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk On Wed, 23 Aug 2000 22:54:33 KST, Byung-Soo Kim wrote: > Sorry this question is not arm-linux specific, > but the majordomo of linux-kernel seems to be broken. Correct, vger.rutgers.edu died last week. Use vger.kernel.org instead. > I'm a kernel newbie who is about to write a device driver. > I have read a few books on kernel programming > and device driver writing, but found that > "real" device drivers in -test6 kernel have > a lot of features not described in any book. > I believe there exists an official web site > maintaining up-to-date information - or history log - > on how to use the kernel services such as devfs, > kernel timer, wait queue, and so on. AFAIK, there is no such web site. I always use the source of the other drivers for inspiration. How to use devfs will be the least of your worries, getting your first driver to work at all will be much more difficult. > Where did authors of the device drivers in -test6 > learn the usage of devfs ? Most of them didn't. Richard Gooch (the author of devfs) changed quite a lot of drivers. There is also quite a lot of information in Documentation/filesystems/devfs/ in the 2.4.0-test6 kernel. Erik -- J.A.K. (Erik) Mouw, Information and Communication Theory Group, Department of Electrical Engineering, Faculty of Information Technology and Systems, Delft University of Technology, PO BOX 5031, 2600 GA Delft, The Netherlands Phone: +31-15-2783635 Fax: +31-15-2781843 Email: J.A.K.Mouw@its.tudelft.nl WWW: http://www-ict.its.tudelft.nl/~erik/ unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Wed Aug 23 15:33:55 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13Rbar-0002RC-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Wed, 23 Aug 2000 15:33:29 +0100 Received: from [216.85.232.66] (helo=blazieresearch.com) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13Rbaq-0002R5-00 for linux-arm-kernel@lists.arm.linux.org.uk; Wed, 23 Aug 2000 15:33:28 +0100 Received: from localhost (chris@localhost) by blazieresearch.com (8.9.3/8.8.7) with SMTP id KAA25793; Wed, 23 Aug 2000 10:33:25 -0400 Date: Wed, 23 Aug 2000 10:33:25 -0400 (EDT) From: Chris Blazie To: Nicolas Pitre cc: linux-arm-kernel@lists.arm.linux.org.uk Subject: RE: slower decompress? In-Reply-To: Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk > Here's what could be done. Beware that I didn't test it yet. [snip] I tested your changes and it works great, booting both from Flash and RAM. Thanks Nicolas! Chris unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Wed Aug 23 15:49:39 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13Rbq3-0002jW-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Wed, 23 Aug 2000 15:49:11 +0100 Received: from [140.109.13.42] (helo=linux.org.tw ident=root) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13Rbq1-0002jQ-00 for linux-arm-kernel@lists.arm.linux.org.uk; Wed, 23 Aug 2000 15:49:10 +0100 Received: from localhost (chester@localhost) by linux.org.tw (8.9.3/8.9.3) with ESMTP id WAA28450; Wed, 23 Aug 2000 22:51:53 +0800 Date: Wed, 23 Aug 2000 22:51:53 +0800 (CST) From: To: Chien-Min Lee cc: linux-arm-kernel Subject: Re: What does "external abort on non-linefetch" mean? In-Reply-To: <967023115.87257.cm_lee@mail2000.com.tw> Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=X-UNKNOWN Content-Transfer-Encoding: QUOTED-PRINTABLE Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk What's procedure do you do ?? could give a brief?? =09=09=09=09Chester On Wed, 23 Aug 2000, Chien-Min Lee wrote: > Dear All: >=20 > I got the following error when testing a driver on arm-linux. >=20 > [root@Linux /root]$Internal error: external abort on non-linefetch: 28 > CPU: 0 > pc : [] lr : [] > sp : c012ff68 ip : 00000000 fp : c012ffc0 > r10: c001307c r9 : ffffffff r8 : 80000013 > r7 : c00203ac r6 : c0019b1c r5 : c012ffb0 r4 : fa050000 > r3 : c01431c4 r2 : 00000000 r1 : 00000000 r0 : 00000003 > Flags: Nzcv IRQs off FIQs on Mode SVC_32 Segment kernel > Control: C165917F Table: C165917F DAC: 0000001D > Process swapper (pid: 0, stackpage=3Dc012f000) > Stack: > [skip .....] >=20 > However, I do not understand this error message. > Could someone tell me what it talks about?? >=20 > Thanks >=20 > kirk >=20 > =20 >=20 > ----=3D=3D Mailed via Openfind =3D=3D----- > http://mail2000.com.tw/ =B4=A3=A8=D1=A7K=B6O=B9q=A4l=B6l=A5=F3=ABH=BDc >=20 >=20 > unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.ar= m.linux.org.uk >=20 unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Wed Aug 23 16:43:53 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13Rcfv-0003c7-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Wed, 23 Aug 2000 16:42:47 +0100 Received: from [139.175.250.58] (helo=mail2000.com.tw) by www.linux.org.uk with smtp (Exim 3.13 #1) id 13Rcfu-0003bZ-00 for linux-arm-kernel@lists.arm.linux.org.uk; Wed, 23 Aug 2000 16:42:46 +0100 Received: By OpenMail Mailer;Wed, 23 Aug 2000 23:41:40 +0800 (CST) From: "Chien-Min Lee" Reply-To: cm_lee@mail2000.com.tw Subject: Re: What does "external abort on non-linefetch" mean? Message-ID: <967045300.6405.cm_lee@mail2000.com.tw> To: "" Cc: "linux-arm-kernel" Date: Wed, 23 Aug 2000 23:41:40 +0800 (CST) MIME-Version: 1.0 Content-Type: text/plain Content-Transfer-Encoding: 8bit Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk Hi: > What's procedure do you do ?? > could give a brief?? I am not sure exactly which code causing this error. I use DMA to transfer data from peripheral device of sa1110 to certain memory space. Due to the fifo part of that peripheral device, I have to drain data in fifo out to following address. That error occurred when I try to move data from fifo to memory. kirk > > [root@Linux /root]$Internal error: external abort on non-linefetch: 28 > > CPU: 0 > > pc : [] lr : [] > > sp : c012ff68 ip : 00000000 fp : c012ffc0 > > r10: c001307c r9 : ffffffff r8 : 80000013 > > r7 : c00203ac r6 : c0019b1c r5 : c012ffb0 r4 : fa050000 > > r3 : c01431c4 r2 : 00000000 r1 : 00000000 r0 : 00000003 > > Flags: Nzcv IRQs off FIQs on Mode SVC_32 Segment kernel > > Control: C165917F Table: C165917F DAC: 0000001D > > Process swapper (pid: 0, stackpage=c012f000) > > Stack: > > [skip .....] ----== Mailed via Openfind ==----- http://mail2000.com.tw/ ´£¨Ñ§K¶O¹q¤l¶l¥ó«H½c unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Wed Aug 23 17:28:42 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13RdNn-0004Zy-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Wed, 23 Aug 2000 17:28:07 +0100 Received: from [205.250.170.220] (helo=xanadu.vipswitch.com) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13RdNl-0004Zk-00 for linux-arm-kernel@lists.arm.linux.org.uk; Wed, 23 Aug 2000 17:28:06 +0100 Date: Wed, 23 Aug 2000 12:27:33 -0400 (EDT) From: Nicolas Pitre To: Chien-Min Lee cc: , linux-arm-kernel Subject: Re: What does "external abort on non-linefetch" mean? In-Reply-To: <967045300.6405.cm_lee@mail2000.com.tw> Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk On Wed, 23 Aug 2000, Chien-Min Lee wrote: > Hi: > > What's procedure do you do ?? > > could give a brief?? > > I am not sure exactly which code causing this error. > I use DMA to transfer data from peripheral device of sa1110 to certain memory space. > Due to the fifo part of that peripheral device, I have to drain data in fifo out to following address. > That error occurred when I try to move data from fifo to memory. This means that you passed an invalid address to the DMA controller. Nicolas unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Wed Aug 23 19:29:04 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13RfFy-0007GR-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Wed, 23 Aug 2000 19:28:10 +0100 Received: from [192.58.206.2] (helo=crl.dec.com) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13RfFx-0007GL-00 for linux-arm-kernel@lists.arm.linux.org.uk; Wed, 23 Aug 2000 19:28:09 +0100 Received: from crlwebmail.crl.dec.com (crlwebmail.crl.dec.com [16.11.0.2]) by crl.dec.com (8.8.8/RWD-1.2) with ESMTP id OAA32566; Wed, 23 Aug 2000 14:27:56 -0400 (EDT) Received: by crlwebmail.crl.dec.com with Internet Mail Service (5.5.2650.21) id ; Wed, 23 Aug 2000 14:33:16 -0400 Message-ID: From: George France To: "'Erik Mouw'" , linuxmama@hanmail.net Cc: linux-arm-kernel@lists.arm.linux.org.uk Subject: RE: Source of Informaion on devfs Date: Wed, 23 Aug 2000 14:27:57 -0400 MIME-Version: 1.0 X-Mailer: Internet Mail Service (5.5.2650.21) Content-Type: text/plain; charset="iso-8859-1" Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk Richard Gooch has a web page on devfs at: http://www.atnf.csiro.au/~rgooch/linux/docs/devfs.html I think it has all the data that you need. Have fun, --George George France, france@crl.dec.com Cambridge Research Laboratory, Compaq Computer Corporation One Cambridgecenter MS: CRL Cambridge, MA 02142 USA > -----Original Message----- > From: Erik Mouw [mailto:J.A.K.Mouw@its.tudelft.nl] > Sent: Wednesday, August 23, 2000 11:08 AM > To: linuxmama@hanmail.net > Cc: linux-arm-kernel@lists.arm.linux.org.uk > Subject: Re: Source of Informaion on devfs > > > On Wed, 23 Aug 2000 22:54:33 KST, Byung-Soo Kim wrote: > > Sorry this question is not arm-linux specific, > > but the majordomo of linux-kernel seems to be broken. > > Correct, vger.rutgers.edu died last week. Use vger.kernel.org instead. > > > I'm a kernel newbie who is about to write a device driver. > > I have read a few books on kernel programming > > and device driver writing, but found that > > "real" device drivers in -test6 kernel have > > a lot of features not described in any book. > > I believe there exists an official web site > > maintaining up-to-date information - or history log - > > on how to use the kernel services such as devfs, > > kernel timer, wait queue, and so on. > > AFAIK, there is no such web site. I always use the source of the other > drivers for inspiration. How to use devfs will be the least of your > worries, getting your first driver to work at all will be much more > difficult. > > > Where did authors of the device drivers in -test6 > > learn the usage of devfs ? > > Most of them didn't. Richard Gooch (the author of devfs) > changed quite a > lot of drivers. There is also quite a lot of information in > Documentation/filesystems/devfs/ in the 2.4.0-test6 kernel. > > > Erik > > -- > J.A.K. (Erik) Mouw, Information and Communication Theory > Group, Department > of Electrical Engineering, Faculty of Information Technology > and Systems, > Delft University of Technology, PO BOX 5031, 2600 GA Delft, > The Netherlands > Phone: +31-15-2783635 Fax: +31-15-2781843 Email: > J.A.K.Mouw@its.tudelft.nl > WWW: http://www-ict.its.tudelft.nl/~erik/ > > > > > > unsubscribe: body of `unsubscribe linux-arm-kernel' to > majordomo@lists.arm.linux.org.uk > unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Wed Aug 23 20:06:06 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13RfqH-0007tG-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Wed, 23 Aug 2000 20:05:41 +0100 Received: from dyn-33.linux.theplanet.co.uk ([195.92.244.33] helo=caramon.arm.linux.org.uk) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13RfqF-0007t6-00; Wed, 23 Aug 2000 20:05:40 +0100 Received: from flint.arm.linux.org.uk (root@flint [192.168.0.4]) by caramon.arm.linux.org.uk (8.9.3/8.9.3) with ESMTP id TAA02128; Wed, 23 Aug 2000 19:22:03 +0100 Received: (from linux@localhost) by flint.arm.linux.org.uk (8.9.3/8.9.3) id TAA16471; Wed, 23 Aug 2000 19:18:46 +0100 From: Russell King - ARM Linux Admin Message-Id: <200008231818.TAA16471@flint.arm.linux.org.uk> Subject: Re: What does "external abort on non-linefetch" mean? To: cm_lee@mail2000.com.tw Date: Wed, 23 Aug 2000 19:18:46 +0100 (BST) Cc: linux-arm-kernel@lists.arm.linux.org.uk (linux-arm-kernel) In-Reply-To: <967023115.87257.cm_lee@mail2000.com.tw> from "Chien-Min Lee" at Aug 23, 2000 05:31:55 PM X-Mailer: ELM [version 2.5 PL1] MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk Chien-Min Lee writes: > Internal error: external abort on non-linefetch: 28 > However, I do not understand this error message. > Could someone tell me what it talks about?? You can find all the gory details in the ARM architecture reference manual (see the section on data aborts at the back). Its probably not a kernel problem, but a hardware problem. _____ |_____| ------------------------------------------------- ---+---+- | | Russell King linux@arm.linux.org.uk --- --- | | | | http://www.arm.linux.org.uk/ / / | | +-+-+ --- -+- / | THE developer of ARM Linux |+| /|\ / | | | --- | +-+-+ ------------------------------------------------- /\\\ | unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Wed Aug 23 20:08:43 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13RfqK-0007tN-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Wed, 23 Aug 2000 20:05:44 +0100 Received: from dyn-33.linux.theplanet.co.uk ([195.92.244.33] helo=caramon.arm.linux.org.uk) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13RfqI-0007t6-00; Wed, 23 Aug 2000 20:05:43 +0100 Received: from flint.arm.linux.org.uk (root@flint [192.168.0.4]) by caramon.arm.linux.org.uk (8.9.3/8.9.3) with ESMTP id TAA02136; Wed, 23 Aug 2000 19:24:21 +0100 Received: (from linux@localhost) by flint.arm.linux.org.uk (8.9.3/8.9.3) id TAA16492; Wed, 23 Aug 2000 19:21:05 +0100 From: Russell King - ARM Linux Admin Message-Id: <200008231821.TAA16492@flint.arm.linux.org.uk> Subject: Re: What does "external abort on non-linefetch" mean? To: nico@cam.org (Nicolas Pitre) Date: Wed, 23 Aug 2000 19:21:05 +0100 (BST) Cc: cm_lee@mail2000.com.tw (Chien-Min Lee), chester@linux.org.tw, linux-arm-kernel@lists.arm.linux.org.uk (linux-arm-kernel) In-Reply-To: from "Nicolas Pitre" at Aug 23, 2000 12:27:33 PM X-Mailer: ELM [version 2.5 PL1] MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk Nicolas Pitre writes: > On Wed, 23 Aug 2000, Chien-Min Lee wrote: > > I am not sure exactly which code causing this error. > > I use DMA to transfer data from peripheral device of sa1110 to certain memory space. > > Due to the fifo part of that peripheral device, I have to drain data in fifo out to following address. > > That error occurred when I try to move data from fifo to memory. > > This means that you passed an invalid address to the DMA controller. Surely not. If this is true, the SA1110 does not follow the ARM ARM. Data aborts are not supposed to be spuriously raised. _____ |_____| ------------------------------------------------- ---+---+- | | Russell King linux@arm.linux.org.uk --- --- | | | | http://www.arm.linux.org.uk/ / / | | +-+-+ --- -+- / | THE developer of ARM Linux |+| /|\ / | | | --- | +-+-+ ------------------------------------------------- /\\\ | unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Wed Aug 23 22:15:31 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13Rhpt-0001dg-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Wed, 23 Aug 2000 22:13:25 +0100 Received: from [216.18.11.193] (helo=duke.electric.net) by www.linux.org.uk with smtp (Exim 3.13 #1) id 13Rhpr-0001dS-00 for linux-arm-kernel@lists.arm.linux.org.uk; Wed, 23 Aug 2000 22:13:24 +0100 Received: (qmail 25360 invoked from network); 23 Aug 2000 21:13:13 -0000 Received: from localhost (HELO relay.electric.net) (@127.0.0.1) by duke.electric.net with SMTP; 23 Aug 2000 21:13:13 -0000 Received: (from smtpd@localhost) by relay.electric.net (8.8.8/8.8.6) id OAA25339 for linux-arm-kernel@lists.arm.linux.org.uk; Wed, 23 Aug 2000 14:13:13 -0700 (PDT) Received: from cobra.intrinsyc.com(64.16.10.158), claiming to be "dp.intrinsyc.com" via SMTP by relay.electric.net, id smtpdj25256; Wed Aug 23 14:13:10 2000 Received: from ileong (helo=localhost) by dp.intrinsyc.com with local-esmtp (Exim 3.13 #1 (Debian)) id 13Rhpe-0003pT-00 for ; Wed, 23 Aug 2000 14:13:10 -0700 Date: Wed, 23 Aug 2000 14:13:10 -0700 (PDT) From: I-Gene Leong To: linux-arm-kernel@lists.arm.linux.org.uk Subject: test6-rmk5-np1 PCMCIA / UCB1200 audio? Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk Is PCMCIA just fundamentally broken? It works, I just still get that start-stop-daemon oops, and if I get rid of PCMCIA I have no problems at all. And is anyone working on a proper UCB1200 audio driver? I'm looking at the new UCB1200 touchscreen driver and trying to decide whether a second UCB driver will interfere with it (I'm wondering this because of the unified UCB interface in arch/arm/special)... and if nobody is writing a driver, I'll do it, although the question of a unified interface should probably be resolved before I start. -- I-Gene Leong ileong@intrinsyc.com / 604-646-6962 "We see Linux as being much of a fad as the Internet was in 1995." - Irving Wladawsky-Berger, IBM VP of technology and strategy unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Wed Aug 23 22:34:55 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13RiAH-0001zj-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Wed, 23 Aug 2000 22:34:29 +0100 Received: from [205.250.170.220] (helo=xanadu.vipswitch.com) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13RiAG-0001zd-00 for linux-arm-kernel@lists.arm.linux.org.uk; Wed, 23 Aug 2000 22:34:28 +0100 Date: Wed, 23 Aug 2000 17:33:57 -0400 (EDT) From: Nicolas Pitre To: I-Gene Leong cc: linux-arm-kernel@lists.arm.linux.org.uk Subject: Re: test6-rmk5-np1 PCMCIA / UCB1200 audio? In-Reply-To: Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk On Wed, 23 Aug 2000, I-Gene Leong wrote: > Is PCMCIA just fundamentally broken? It works, I just still get that > start-stop-daemon oops, and if I get rid of PCMCIA I have no problems at > all. It worked before. I apparently let a change slip into the code that broke it somehow. Now we just need to revert that change or fix it appropriately. I'm sure the first person with some free time among us will fix it. > And is anyone working on a proper UCB1200 audio driver? I'm looking at the > new UCB1200 touchscreen driver and trying to decide whether a second UCB > driver will interfere with it (I'm wondering this because of the unified > UCB interface in arch/arm/special)... and if nobody is writing a driver, > I'll do it, although the question of a unified interface should probably > be resolved before I start. We definitely need a master UCB1200 driver that will arbiter the access from all the slave drivers providing services, like touchscreen, audio, etc. I kept the old special/ files for reference only. They should be reintegrated/rewritten to support the master/slave structure as it was done before. Nicolas unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Wed Aug 23 22:41:20 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13RiFV-000266-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Wed, 23 Aug 2000 22:39:53 +0100 Received: from [216.18.11.193] (helo=duke.electric.net) by www.linux.org.uk with smtp (Exim 3.13 #1) id 13RiFU-00025z-00 for linux-arm-kernel@lists.arm.linux.org.uk; Wed, 23 Aug 2000 22:39:52 +0100 Received: (qmail 63545 invoked from network); 23 Aug 2000 21:39:52 -0000 Received: from localhost (HELO relay.electric.net) (@127.0.0.1) by duke.electric.net with SMTP; 23 Aug 2000 21:39:52 -0000 Received: (from smtpd@localhost) by relay.electric.net (8.8.8/8.8.6) id OAA63513; Wed, 23 Aug 2000 14:39:52 -0700 (PDT) Received: from cobra.intrinsyc.com(64.16.10.158), claiming to be "dp.intrinsyc.com" via SMTP by relay.electric.net, id smtpda63473; Wed Aug 23 14:39:51 2000 Received: from ileong (helo=localhost) by dp.intrinsyc.com with local-esmtp (Exim 3.13 #1 (Debian)) id 13RiFN-0004cL-00; Wed, 23 Aug 2000 14:39:45 -0700 Date: Wed, 23 Aug 2000 14:39:45 -0700 (PDT) From: I-Gene Leong To: Nicolas Pitre cc: linux-arm-kernel@lists.arm.linux.org.uk Subject: Re: test6-rmk5-np1 PCMCIA / UCB1200 audio? In-Reply-To: Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk On Wed, 23 Aug 2000, Nicolas Pitre wrote: > On Wed, 23 Aug 2000, I-Gene Leong wrote: > > Is PCMCIA just fundamentally broken? It works, I just still get that > > start-stop-daemon oops, and if I get rid of PCMCIA I have no problems at > > all. > It worked before. I apparently let a change slip into the code that broke > it somehow. Now we just need to revert that change or fix it > appropriately. I'm sure the first person with some free time among us > will fix it. test4-rmk1-np1 was fine, test5-rmk1-np1 wasn't... I have a diff between the two and am going to start looking. > > And is anyone working on a proper UCB1200 audio driver? I'm looking at the > > new UCB1200 touchscreen driver and trying to decide whether a second UCB > > driver will interfere with it (I'm wondering this because of the unified > > UCB interface in arch/arm/special)... and if nobody is writing a driver, > > I'll do it, although the question of a unified interface should probably > > be resolved before I start. > We definitely need a master UCB1200 driver that will arbiter the access > from all the slave drivers providing services, like touchscreen, audio, > etc. I kept the old special/ files for reference only. They should be > reintegrated/rewritten to support the master/slave structure as it was > done before. Sure, can do... yeah, I looked through and realized that we'll have conflicts for anything that wants the UCB interrupt, so I volunteer to do this driver. -- I-Gene Leong ileong@intrinsyc.com / 604-646-6962 "We see Linux as being much of a fad as the Internet was in 1995." - Irving Wladawsky-Berger, IBM VP of technology and strategy unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Wed Aug 23 23:19:17 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13RirP-0002jv-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Wed, 23 Aug 2000 23:19:03 +0100 Received: from dyn-33.linux.theplanet.co.uk ([195.92.244.33] helo=caramon.arm.linux.org.uk) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13RirN-0002jf-00; Wed, 23 Aug 2000 23:19:02 +0100 Received: from flint.arm.linux.org.uk (root@flint [192.168.0.4]) by caramon.arm.linux.org.uk (8.9.3/8.9.3) with ESMTP id XAA03348; Wed, 23 Aug 2000 23:19:00 +0100 Received: (from linux@localhost) by flint.arm.linux.org.uk (8.9.3/8.9.3) id XAA17571; Wed, 23 Aug 2000 23:15:43 +0100 From: Russell King - ARM Linux Admin Message-Id: <200008232215.XAA17571@flint.arm.linux.org.uk> Subject: Re: test6-rmk5-np1 PCMCIA / UCB1200 audio? To: ileong@intrinsyc.com (I-Gene Leong) Date: Wed, 23 Aug 2000 23:15:43 +0100 (BST) Cc: linux-arm-kernel@lists.arm.linux.org.uk In-Reply-To: from "I-Gene Leong" at Aug 23, 2000 02:13:10 PM X-Mailer: ELM [version 2.5 PL1] MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk I-Gene Leong writes: > Is PCMCIA just fundamentally broken? It works, I just still get that > start-stop-daemon oops, and if I get rid of PCMCIA I have no problems at > all. This is a well-known trademark of wrong flags passed into kernel_thread() (most notably '0' passed in for the flags). This ends up with an invalid pointer to a couple of structures in the task struct, which procfs then dereferences when start-stop-daemon searches through all PIDs. _____ |_____| ------------------------------------------------- ---+---+- | | Russell King linux@arm.linux.org.uk --- --- | | | | http://www.arm.linux.org.uk/ / / | | +-+-+ --- -+- / | THE developer of ARM Linux |+| /|\ / | | | --- | +-+-+ ------------------------------------------------- /\\\ | unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Wed Aug 23 23:20:25 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13RisR-0002lL-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Wed, 23 Aug 2000 23:20:07 +0100 Received: from [216.18.11.193] (helo=duke.electric.net) by www.linux.org.uk with smtp (Exim 3.13 #1) id 13RisQ-0002lC-00 for linux-arm-kernel@lists.arm.linux.org.uk; Wed, 23 Aug 2000 23:20:07 +0100 Received: (qmail 16752 invoked from network); 23 Aug 2000 22:20:02 -0000 Received: from localhost (HELO relay.electric.net) (@127.0.0.1) by duke.electric.net with SMTP; 23 Aug 2000 22:20:02 -0000 Received: (from smtpd@localhost) by relay.electric.net (8.8.8/8.8.6) id PAA16737; Wed, 23 Aug 2000 15:20:02 -0700 (PDT) Received: from cobra.intrinsyc.com(64.16.10.158), claiming to be "dp.intrinsyc.com" via SMTP by relay.electric.net, id smtpdF16654; Wed Aug 23 15:19:59 2000 Received: from ileong (helo=localhost) by dp.intrinsyc.com with local-esmtp (Exim 3.13 #1 (Debian)) id 13RisJ-0004ju-00; Wed, 23 Aug 2000 15:19:59 -0700 Date: Wed, 23 Aug 2000 15:19:58 -0700 (PDT) From: I-Gene Leong To: Nicolas Pitre cc: linux-arm-kernel@lists.arm.linux.org.uk Subject: test5+ PCMCIA In-Reply-To: Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk Looks to me as though some major changes happened inside the PCMCIA driver itself between test4 and test5. All this is beyond my comprehension, so I can't really do anything about fixing it... -- I-Gene Leong ileong@intrinsyc.com / 604-646-6962 "We see Linux as being much of a fad as the Internet was in 1995." - Irving Wladawsky-Berger, IBM VP of technology and strategy unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Wed Aug 23 23:35:47 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13Rj7R-00030m-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Wed, 23 Aug 2000 23:35:37 +0100 Received: from [216.18.11.193] (helo=duke.electric.net) by www.linux.org.uk with smtp (Exim 3.13 #1) id 13Rj7Q-00030g-00 for linux-arm-kernel@lists.arm.linux.org.uk; Wed, 23 Aug 2000 23:35:37 +0100 Received: (qmail 36368 invoked from network); 23 Aug 2000 22:35:36 -0000 Received: from localhost (HELO relay.electric.net) (@127.0.0.1) by duke.electric.net with SMTP; 23 Aug 2000 22:35:36 -0000 Received: (from smtpd@localhost) by relay.electric.net (8.8.8/8.8.6) id PAA36348; Wed, 23 Aug 2000 15:35:34 -0700 (PDT) Received: from cobra.intrinsyc.com(64.16.10.158), claiming to be "dp.intrinsyc.com" via SMTP by relay.electric.net, id smtpdJ36267; Wed Aug 23 15:35:29 2000 Received: from ileong (helo=localhost) by dp.intrinsyc.com with local-esmtp (Exim 3.13 #1 (Debian)) id 13Rj7I-00053z-00; Wed, 23 Aug 2000 15:35:28 -0700 Date: Wed, 23 Aug 2000 15:35:28 -0700 (PDT) From: I-Gene Leong To: Russell King - ARM Linux Admin cc: linux-arm-kernel@lists.arm.linux.org.uk Subject: Re: test6-rmk5-np1 PCMCIA / UCB1200 audio? In-Reply-To: <200008232215.XAA17571@flint.arm.linux.org.uk> Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk On Wed, 23 Aug 2000, Russell King - ARM Linux Admin wrote: > I-Gene Leong writes: > > Is PCMCIA just fundamentally broken? It works, I just still get that > > start-stop-daemon oops, and if I get rid of PCMCIA I have no problems at > > all. > This is a well-known trademark of wrong flags passed into kernel_thread() > (most notably '0' passed in for the flags). This ends up with an invalid > pointer to a couple of structures in the task struct, which procfs then > dereferences when start-stop-daemon searches through all PIDs. Quick grep reveals sa1100_generic.c with this; test4 doesn't have a single mention of it. I have no idea what I'm doing here, but yenta.c also has a kernel_thread call with CLONE_FS | CLONE_FILES | CLONE_SIGHAND, so I'm going to wander out on a limb and give that a shot. People who know more about this than I do (i.e. probably just about everyone) feel free to flame me in the right direction. :) -- I-Gene Leong ileong@intrinsyc.com / 604-646-6962 "We see Linux as being much of a fad as the Internet was in 1995." - Irving Wladawsky-Berger, IBM VP of technology and strategy unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Wed Aug 23 23:51:24 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13RjMK-0003Ef-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Wed, 23 Aug 2000 23:51:00 +0100 Received: from dyn-33.linux.theplanet.co.uk ([195.92.244.33] helo=caramon.arm.linux.org.uk) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13RjMI-0003EZ-00; Wed, 23 Aug 2000 23:50:59 +0100 Received: from flint.arm.linux.org.uk (root@flint [192.168.0.4]) by caramon.arm.linux.org.uk (8.9.3/8.9.3) with ESMTP id XAA03512; Wed, 23 Aug 2000 23:50:57 +0100 Received: (from linux@localhost) by flint.arm.linux.org.uk (8.9.3/8.9.3) id XAA17730; Wed, 23 Aug 2000 23:47:40 +0100 From: Russell King - ARM Linux Admin Message-Id: <200008232247.XAA17730@flint.arm.linux.org.uk> Subject: Re: test6-rmk5-np1 PCMCIA / UCB1200 audio? To: ileong@intrinsyc.com (I-Gene Leong) Date: Wed, 23 Aug 2000 23:47:40 +0100 (BST) Cc: linux-arm-kernel@lists.arm.linux.org.uk In-Reply-To: from "I-Gene Leong" at Aug 23, 2000 03:35:28 PM X-Mailer: ELM [version 2.5 PL1] MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk I-Gene Leong writes: > Quick grep reveals sa1100_generic.c with this; test4 doesn't have a single > mention of it. > > I have no idea what I'm doing here, but yenta.c also has a kernel_thread > call with CLONE_FS | CLONE_FILES | CLONE_SIGHAND That's correct. It is also arguable that it should have SIGCHLD ored in there as well, but this will not affect your problem. (see the usage in fs/buffer.c) _____ |_____| ------------------------------------------------- ---+---+- | | Russell King linux@arm.linux.org.uk --- --- | | | | http://www.arm.linux.org.uk/ / / | | +-+-+ --- -+- / | THE developer of ARM Linux |+| /|\ / | | | --- | +-+-+ ------------------------------------------------- /\\\ | unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Thu Aug 24 04:39:04 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13Rnp0-0004iX-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Thu, 24 Aug 2000 04:36:54 +0100 Received: from [216.136.129.36] (helo=web9203.mail.yahoo.com) by www.linux.org.uk with smtp (Exim 3.13 #1) id 13Rnoz-0004i5-00 for linux-arm-kernel@lists.arm.linux.org.uk; Thu, 24 Aug 2000 04:36:53 +0100 Message-ID: <20000824033619.29342.qmail@web9203.mail.yahoo.com> Received: from [24.15.114.165] by web9203.mail.yahoo.com; Wed, 23 Aug 2000 20:36:19 PDT Date: Wed, 23 Aug 2000 20:36:19 -0700 (PDT) From: Al Kam Subject: bootpimage problem To: linux-arm-kernel@lists.arm.linux.org.uk MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk I am having trouble with a downloaded zImage+initrd image using bootp (make bootpImage). Currently using 2.4.0-test6-rmk5 and nico's ramdisk-img on an EBSA-285. Downloading just zImage kernel works fine. Here are the last kernel messages before the box hangs: > Got BOOTP answer from eagle (192.168.220.1), my address is 192.168.220.2 > TFTPing /tftpboot/linux.arm... Ok - 2172KB > Now booting image... > .ran out of input data > > out of input data What am I doing wrong? Al __________________________________________________ Do You Yahoo!? Yahoo! Mail - Free email you can access from anywhere! http://mail.yahoo.com/ unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Thu Aug 24 05:36:10 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13RokA-0005Xw-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Thu, 24 Aug 2000 05:35:58 +0100 Received: from [210.184.5.125] (helo=pikachu.idthk.com) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13Rok6-0005XI-00 for linux-arm-kernel@lists.arm.linux.org.uk; Thu, 24 Aug 2000 05:35:56 +0100 Received: from localhost (thomas@localhost) by pikachu.idthk.com (8.9.3/8.8.7) with ESMTP id MAA16874; Thu, 24 Aug 2000 12:36:21 +0800 Date: Thu, 24 Aug 2000 12:36:21 +0800 (CST) From: Tak-Shing Chan To: I-Gene Leong cc: linux-arm-kernel@lists.arm.linux.org.uk Subject: Re: test6-rmk5-np1 PCMCIA / UCB1200 audio? In-Reply-To: Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk On Wed, 23 Aug 2000, I-Gene Leong wrote: >>> And is anyone working on a proper UCB1200 audio driver? I'm looking at the Erik Mouw's LART drivers (branched out from Compaq's Itsy drivers) already have all the master/slave functionality you're looking for. I was working on integrating the touchscreen driver with the LART drivers as well as upgrading the calibration stuffs as per http://www.handhelds.org/projects/ts_spec.html. It will be ready when the final COMPAQ touchscreen specs are released. Tak-Shing Chan unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Thu Aug 24 07:07:56 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13RqAz-0006qw-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Thu, 24 Aug 2000 07:07:45 +0100 Received: from [140.109.13.42] (helo=linux.org.tw ident=root) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13RqAx-0006qm-00 for linux-arm-kernel@lists.arm.linux.org.uk; Thu, 24 Aug 2000 07:07:43 +0100 Received: from localhost (chester@localhost) by linux.org.tw (8.9.3/8.9.3) with ESMTP id OAA07182; Thu, 24 Aug 2000 14:09:51 +0800 Date: Thu, 24 Aug 2000 14:09:51 +0800 (CST) From: To: Tak-Shing Chan cc: I-Gene Leong , linux-arm-kernel@lists.arm.linux.org.uk Subject: Re: test6-rmk5-np1 PCMCIA / UCB1200 audio? In-Reply-To: Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk > Erik Mouw's LART drivers (branched out from Compaq's Itsy > drivers) already have all the master/slave functionality you're > looking for. I was working on integrating the touchscreen driver > with the LART drivers as well as upgrading the calibration stuffs > as per http://www.handhelds.org/projects/ts_spec.html. It will > be ready when the final COMPAQ touchscreen specs are released. > > Tak-Shing Chan Thanks for your intergration,but it seems that your struct TS_EVENT is unbound data, typedef struct { short pressure; int x; int y; int millisecs }TS_EVENT I also make a TS driver that based Tifon's driver, ftp://ftp.netwinder.org/users/c/chester/TS/ and i hope can intergrate with others. BTW,please don't use Major=11 ,it will conflict when uses keyboard. Maybe dynamic assign or use a reserved number,(0 or 60-63). Chester unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Thu Aug 24 09:17:46 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13RsC0-0000Ls-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Thu, 24 Aug 2000 09:16:56 +0100 Received: from [210.184.5.125] (helo=pikachu.idthk.com) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13RsBx-0000LH-00 for linux-arm-kernel@lists.arm.linux.org.uk; Thu, 24 Aug 2000 09:16:54 +0100 Received: from localhost (thomas@localhost) by pikachu.idthk.com (8.9.3/8.8.7) with ESMTP id QAA16632; Thu, 24 Aug 2000 16:17:01 +0800 Date: Thu, 24 Aug 2000 16:17:01 +0800 (CST) From: Tak-Shing Chan To: chester@linux.org.tw cc: linux-arm-kernel@lists.arm.linux.org.uk Subject: Re: test6-rmk5-np1 PCMCIA / UCB1200 audio? In-Reply-To: Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk On Thu, 24 Aug 2000 chester@linux.org.tw wrote: > Thanks for your intergration,but it seems that your struct TS_EVENT > is unbound data, I can't get your point--please enlighten me. By the way, TS_EVENT is taken verbatim from Compaq's touchscreen specs version 0.2. > I also make a TS driver that based Tifon's driver, > ftp://ftp.netwinder.org/users/c/chester/TS/ > and i hope can intergrate with others. Just have a glance at your code. Looks like there have been a lot of work (like powermgr stuffs), perhaps we should just mix and match the two drivers to avoid duplicated efforts. > BTW,please don't use Major=11 ,it will conflict when uses keyboard. > Maybe dynamic assign or use a reserved number,(0 or 60-63). The specs suggest a major of 10, but I change it to 11 because I don't yet want to switch to misc_register() until the specs are finalized, and 11 because I bet no one is using SCSI CDROMs on the SA11x0. It's just temporary. I'll change it to the correct values as soon as the specs are ready. By the way, how come your keyboard uses SCSI_CDROM_MAJOR? Regards, Tak-Shing Chan unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Thu Aug 24 10:03:11 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13RsuX-000182-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Thu, 24 Aug 2000 10:02:57 +0100 Received: from [210.184.5.125] (helo=pikachu.idthk.com) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13RsuU-00017W-00 for linux-arm-kernel@lists.arm.linux.org.uk; Thu, 24 Aug 2000 10:02:55 +0100 Received: from localhost (thomas@localhost) by pikachu.idthk.com (8.9.3/8.8.7) with ESMTP id RAA16832; Thu, 24 Aug 2000 17:03:07 +0800 Date: Thu, 24 Aug 2000 17:03:07 +0800 (CST) From: Tak-Shing Chan To: Erik Mouw cc: linux-arm-kernel@lists.arm.linux.org.uk Subject: arch/arm/special Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk Hi Erik, Do you have plans to move the UCB1200 code, for example mcp*.h to include/linux, audio-sa1100-mcp.c to drivers/sound and mcp_common.c to drivers/char? In addition, do you think it's a good idea to take the codec routines out of audio-sa1100-mcp.c and revert it to a header file (perhaps rename it as something like linux/ucb1200_codec.h) such that other drivers could reuse the inline codec code? Tak-Shing Chan unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Thu Aug 24 10:31:01 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13RtK4-0001Wp-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Thu, 24 Aug 2000 10:29:20 +0100 Received: from [139.175.250.58] (helo=mail2000.com.tw) by www.linux.org.uk with smtp (Exim 3.13 #1) id 13RtK2-0001WH-00 for linux-arm-kernel@lists.arm.linux.org.uk; Thu, 24 Aug 2000 10:29:19 +0100 Received: By OpenMail Mailer;Thu, 24 Aug 2000 17:28:44 +0800 (CST) From: "Chien-Min Lee" Reply-To: cm_lee@mail2000.com.tw Subject: Question about usbc_h3600_dma.c from iPAQ distribution Message-ID: <967109324.51519.cm_lee@mail2000.com.tw> To: "linux-arm-kernel" Date: Thu, 24 Aug 2000 17:28:44 +0800 (CST) MIME-Version: 1.0 Content-Type: text/plain Content-Transfer-Encoding: 8bit Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk Dear All: I am not sure is it suitable to ask such question here. I found the following code in usbc_h3600_dma.c int disable_dma_sa1100 (dmach_t channel) { int status; dma_regs_t * pDMA; if ( channel >= MAX_USB_CHANNELS ) { printk("enable_dma:Bad channel\n"); return -1; } pDMA = DMA[channel]; status = pDMA->RdDCSR; if( !(status & (DCSR_DONEA|DCSR_DONEB)) ) printk("WARNING : DMA DONEA/B not SET\n"); /* need to do this for buffers fragments 0 < x < 64 */ pDMA->ClrDCSR = 0x7f; return 0; } I just want to ask what "need to do this for buffer fragments 0 To: Tak-Shing Chan cc: chester@linux.org.tw, linux-arm-kernel@lists.arm.linux.org.uk Subject: Re: test6-rmk5-np1 PCMCIA / UCB1200 audio? In-Reply-To: Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk On Thu, 24 Aug 2000, Tak-Shing Chan wrote: > On Thu, 24 Aug 2000 chester@linux.org.tw wrote: > > > Thanks for your intergration,but it seems that your struct TS_EVENT > > is unbound data, > > I can't get your point--please enlighten me. By the way, > TS_EVENT is taken verbatim from Compaq's touchscreen specs > version 0.2. > Ooop,their should be old on web site. look at their some code about h3650_ts.h > > I also make a TS driver that based Tifon's driver, > > ftp://ftp.netwinder.org/users/c/chester/TS/ > > and i hope can intergrate with others. > > Just have a glance at your code. Looks like there have been > a lot of work (like powermgr stuffs), perhaps we should just mix > and match the two drivers to avoid duplicated efforts. > > > BTW,please don't use Major=11 ,it will conflict when uses keyboard. > > Maybe dynamic assign or use a reserved number,(0 or 60-63). > > The specs suggest a major of 10, but I change it to 11 > because I don't yet want to switch to misc_register() until the > specs are finalized, and 11 because I bet no one is using SCSI > CDROMs on the SA11x0. It's just temporary. I'll change it to > the correct values as soon as the specs are ready. > > By the way, how come your keyboard uses SCSI_CDROM_MAJOR? >From devices.txt 11 char Raw keyboard device You can choose 60-63 for temporary(flash driver use 60/block), But maybe we should register a number to linux Chester unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Thu Aug 24 11:33:11 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13RuJQ-0002XP-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Thu, 24 Aug 2000 11:32:44 +0100 Received: from [210.184.5.125] (helo=pikachu.idthk.com) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13RuJI-0002Wm-00 for linux-arm-kernel@lists.arm.linux.org.uk; Thu, 24 Aug 2000 11:32:39 +0100 Received: from localhost (thomas@localhost) by pikachu.idthk.com (8.9.3/8.8.7) with ESMTP id SAA19925; Thu, 24 Aug 2000 18:32:09 +0800 Date: Thu, 24 Aug 2000 18:32:09 +0800 (CST) From: Tak-Shing Chan To: chester@linux.org.tw cc: linux-arm-kernel@lists.arm.linux.org.uk Subject: Re: test6-rmk5-np1 PCMCIA / UCB1200 audio? In-Reply-To: Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk Chester, Oh I've made a BIG mistake. Thanks for correcting me!!! :) I was too obsessed with the absolute numbers and forget about the fundamental differences of block and char devices... Tak-Shing Chan On Thu, 24 Aug 2000 chester@linux.org.tw wrote: > 11 char Raw keyboard device > > You can choose 60-63 for temporary(flash driver use 60/block), > But maybe we should register a number to linux > > Chester unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Thu Aug 24 14:17:57 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13RwrS-00058E-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Thu, 24 Aug 2000 14:16:02 +0100 Received: from [208.242.241.39] (helo=cvsftp.cotw.com ident=root) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13RwrQ-000580-00 for linux-arm-kernel@lists.arm.linux.org.uk; Thu, 24 Aug 2000 14:16:01 +0100 Received: from cotw.com (IDENT:samcconn@ptecdev2.ptec.com [192.168.10.4]) by cvsftp.cotw.com (8.9.3/8.9.3) with ESMTP id IAA18722; Thu, 24 Aug 2000 08:15:24 -0500 Message-ID: <39A520A0.110B0F16@cotw.com> Date: Thu, 24 Aug 2000 08:18:24 -0500 From: S A McConnell Reply-To: samcconn@cotw.com X-Mailer: Mozilla 4.7 [en] (X11; I; Linux 2.2.14 i686) X-Accept-Language: en MIME-Version: 1.0 To: chester@linux.org.tw CC: Tak-Shing Chan , linux-arm-kernel@lists.arm.linux.org.uk Subject: Re: test6-rmk5-np1 PCMCIA / UCB1200 audio? References: Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk I am interested in a "generic" interface to the UCB1200 that can be used to support touch screen, audio and telephony drivers. I have been working on a touch screen based upon sa1100_ts.c and ucb1100_ts.c. Have I made a mistake looking at those drivers? I found that the above drivers had a fair amount of duplicated code. (measure_x, measure_y, measure_pressure...) Which I have reduced. I also created several macro's that significantly reduce the complexity of the UCB register reads and writes. I also disliked the TS knowing about the register formats. I tried to move all that knowledge into the *.h files as well as the code used to read and write the UCB registers. I have not been able to find out who wrote the sa1100_ts.c and ucb1100_ts.c code. I have questions about how the touch screen should detect events. (interrupts vs timer based polling) I am currently using an Interrupt to detect touch events. Should the driver buffer touch events? How many? If data is buffered what should be done on overruns? (Toss oldest? Toss youngest?) I have mine mapped to Major 10, minor 24 and I am using the misc_device approach to register the driver with the kernel. Did I make a misake? You discussion on char vs block is very confusing. This is a character device right? Scott unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Thu Aug 24 15:42:11 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13RyC0-0006R3-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Thu, 24 Aug 2000 15:41:20 +0100 Received: from [211.62.252.66] (helo=s-mail6.hanmail.net) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13RyBy-0006Qx-00 for linux-arm-kernel@lists.arm.linux.org.uk; Thu, 24 Aug 2000 15:41:19 +0100 Received: from www37.hanmail.net (www37.hanmail.net [211.32.117.217]) by s-mail6.hanmail.net (8.10.0/8.9.1) with ESMTP id e7OEcJ514205; Thu, 24 Aug 2000 23:38:19 +0900 Received: (from hanadmin@localhost) by www37.hanmail.net (8.10.0/8.9.1) id e7OEbn117469; Thu, 24 Aug 2000 23:37:49 +0900 (KST) X-Originating-IP: [211.44.193.194] From: "Byung-SooKim" Reply-To: "Byung-SooKim" Organization: To: linux-arm-kernel@lists.arm.linux.org.uk Subject: Three macros in entry-armv.S X-Mailer: Daum Web Mailer 1.0 Date: Thu, 24 Aug 2000 23:37:49 KST Message-Id: <20000824233749.HM.a00000000001pEQ@www37.hanmail.net> MIME-Version: 1.0 Content-Type: text/plain; charset=euc-kr Content-Transfer-Encoding: quoted-printable X-MIME-Autoconverted: from 8bit to quoted-printable by s-mail6.hanmail.net id e7OEcJ514205 Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk Dear experts, Thank you for the answers to my previous question. I could succeed in my first device driver with devfs support. (a simple ISA toy composed of several LEDs and buttons) Thank you again ! Now I'm planning to port test6-rmk5 kernel to KS17C4000 architecture. (ARM-based microcontroller made by SAMSUNG) While reading the source code, I found that arch/arm/kernel/entry-armv.S provides three assembly macros per each architecture : .macro disable_fiq .macro get_irqnr_and_base, irqnr, irqstat, base .macro irq_prio_table I would like to know : 1. why most architectures leave the disable_fiq empty. (dummy macro) Can I just follow others and define my disable_fiq to be an empty macro ? 2. how to compose irq_prio_table. Forgive me if these are silly -or bothering- questions to you experts. I tried my best to find out some source of information on these, but I couldn't. Thank you. - Byungsoo p.s. It looks quite strange to me that there exists no official web site containing these kind of information. :-( =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D =BF=EC=B8=AE =C0=CE=C5=CD=B3=DD, Daum =C6=F2=BB=FD =BE=B2=B4=C2 =B9=AB=B7=E1 E-mail =C1=D6=BC=D2 =C7=D1=B8=DE=C0= =CF=B3=DD =C1=F6=B1=B8=C3=CC =C7=D1=B1=DB =B0=CB=BB=F6=BC=AD=BA=F1=BD=BA Daum FIREB= ALL http://www.daum.net unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Thu Aug 24 15:57:43 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13RyRY-0006hb-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Thu, 24 Aug 2000 15:57:24 +0100 Received: from [205.250.170.220] (helo=xanadu.vipswitch.com) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13RyRX-0006hM-00 for linux-arm-kernel@lists.arm.linux.org.uk; Thu, 24 Aug 2000 15:57:23 +0100 Date: Thu, 24 Aug 2000 10:56:49 -0400 (EDT) From: Nicolas Pitre To: chester@linux.org.tw cc: Tak-Shing Chan , linux-arm-kernel@lists.arm.linux.org.uk Subject: Re: test6-rmk5-np1 PCMCIA / UCB1200 audio? In-Reply-To: Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk On Thu, 24 Aug 2000 chester@linux.org.tw wrote: > You can choose 60-63 for temporary(flash driver use 60/block), > But maybe we should register a number to linux No because the flash driver is destined to die i.e. merged with the MTD stuff. Nicolas unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Thu Aug 24 16:45:38 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13RzBU-0007X1-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Thu, 24 Aug 2000 16:44:52 +0100 Received: from dyn-33.linux.theplanet.co.uk ([195.92.244.33] helo=caramon.arm.linux.org.uk) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13RzBS-0007Wt-00; Thu, 24 Aug 2000 16:44:50 +0100 Received: from flint.arm.linux.org.uk (root@flint [192.168.0.4]) by caramon.arm.linux.org.uk (8.9.3/8.9.3) with ESMTP id PAA06118; Thu, 24 Aug 2000 15:29:28 +0100 Received: (from linux@localhost) by flint.arm.linux.org.uk (8.9.3/8.9.3) id PAA20143; Thu, 24 Aug 2000 15:26:07 +0100 From: Russell King - ARM Linux Admin Message-Id: <200008241426.PAA20143@flint.arm.linux.org.uk> Subject: Re: arch/arm/special To: tchan.rd@idthk.com (Tak-Shing Chan) Date: Thu, 24 Aug 2000 15:26:07 +0100 (BST) Cc: J.A.K.Mouw@its.tudelft.nl (Erik Mouw), linux-arm-kernel@lists.arm.linux.org.uk In-Reply-To: from "Tak-Shing Chan" at Aug 24, 2000 05:03:07 PM X-Mailer: ELM [version 2.5 PL1] MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk Tak-Shing Chan writes: > Do you have plans to move the UCB1200 code, for example > mcp*.h to include/linux, audio-sa1100-mcp.c to drivers/sound and > mcp_common.c to drivers/char? In addition, do you think it's a > good idea to take the codec routines out of audio-sa1100-mcp.c > and revert it to a header file (perhaps rename it as something > like linux/ucb1200_codec.h) such that other drivers could reuse > the inline codec code? There is/was recently a lot of discussion about where things should go. At the moment, the drivers/char directory (among others) are becoming very bulky, and indistinct. IMHO, moving such things into there would be a mistake at the moment. Its probably better to leave it as is for the time being (until 2.5). _____ |_____| ------------------------------------------------- ---+---+- | | Russell King linux@arm.linux.org.uk --- --- | | | | http://www.arm.linux.org.uk/ / / | | +-+-+ --- -+- / | THE developer of ARM Linux |+| /|\ / | | | --- | +-+-+ ------------------------------------------------- /\\\ | unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Thu Aug 24 16:51:16 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13RzHD-0007cs-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Thu, 24 Aug 2000 16:50:47 +0100 Received: from [205.250.170.220] (helo=xanadu.vipswitch.com) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13RzHC-0007ce-00 for linux-arm-kernel@lists.arm.linux.org.uk; Thu, 24 Aug 2000 16:50:46 +0100 Date: Thu, 24 Aug 2000 11:49:58 -0400 (EDT) From: Nicolas Pitre To: S A McConnell cc: chester@linux.org.tw, Tak-Shing Chan , linux-arm-kernel@lists.arm.linux.org.uk Subject: Re: test6-rmk5-np1 PCMCIA / UCB1200 audio? In-Reply-To: <39A520A0.110B0F16@cotw.com> Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk On Thu, 24 Aug 2000, S A McConnell wrote: > I am interested in a "generic" interface to the UCB1200 that can be used > to support touch screen, audio and telephony drivers. > > I have been working on a touch screen based upon sa1100_ts.c and > ucb1100_ts.c. Have I made a mistake looking at those drivers? > > I found that the above drivers had a fair amount of duplicated code. > (measure_x, measure_y, measure_pressure...) Which I have reduced. > > I also created several macro's that significantly reduce the complexity > of the UCB register reads and writes. > > I also disliked the TS knowing about the register formats. I tried to > move all that knowledge into the *.h files as well as the code used to > read and write the UCB registers. > > I have not been able to find out who wrote the sa1100_ts.c and > ucb1100_ts.c code. > > I have questions about how the touch screen should detect events. > (interrupts vs timer based polling) I am currently using an Interrupt > to detect touch events. Interrupts is always the best. Like the iPAQ touchscreen interface, your events only need a timestamp to measure velocity. > Should the driver buffer touch events? How many? Probably, but not much. > If data is buffered what should be done on overruns? > (Toss oldest? Toss youngest?) Probably the oldest since the newer pen position is more relevant than the oldest which is not current anymore. > I have mine mapped to Major 10, minor 24 and I am using the misc_device > approach to register the driver with the kernel. > Did I make a misake? Probably not. > You discussion on char vs block is very confusing. This is a character > device right? Yes. Nicolas unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Thu Aug 24 17:03:14 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13RzSx-0007p5-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Thu, 24 Aug 2000 17:02:55 +0100 Received: from [205.250.170.220] (helo=xanadu.vipswitch.com) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13RzSw-0007oz-00 for linux-arm-kernel@lists.arm.linux.org.uk; Thu, 24 Aug 2000 17:02:54 +0100 Date: Thu, 24 Aug 2000 12:01:50 -0400 (EDT) From: Nicolas Pitre To: Russell King - ARM Linux Admin cc: Tak-Shing Chan , Erik Mouw , linux-arm-kernel@lists.arm.linux.org.uk Subject: Re: arch/arm/special In-Reply-To: <200008241426.PAA20143@flint.arm.linux.org.uk> Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk On Thu, 24 Aug 2000, Russell King - ARM Linux Admin wrote: > Tak-Shing Chan writes: > > Do you have plans to move the UCB1200 code, for example > > mcp*.h to include/linux, audio-sa1100-mcp.c to drivers/sound and > > mcp_common.c to drivers/char? In addition, do you think it's a > > good idea to take the codec routines out of audio-sa1100-mcp.c > > and revert it to a header file (perhaps rename it as something > > like linux/ucb1200_codec.h) such that other drivers could reuse > > the inline codec code? > > There is/was recently a lot of discussion about where things should > go. At the moment, the drivers/char directory (among others) are > becoming very bulky, and indistinct. > > IMHO, moving such things into there would be a mistake at the moment. > Its probably better to leave it as is for the time being (until 2.5). At the moment, all that stuff is currently in the now disappeared arch/arm/special directory... and IMHO it needs a big restructuration before it goes anywhere. Nicolas unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Thu Aug 24 17:08:35 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13RzY9-0007wZ-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Thu, 24 Aug 2000 17:08:17 +0100 Received: from dyn-33.linux.theplanet.co.uk ([195.92.244.33] helo=caramon.arm.linux.org.uk) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13RzY8-0007wP-00; Thu, 24 Aug 2000 17:08:16 +0100 Received: from flint.arm.linux.org.uk (root@flint [192.168.0.4]) by caramon.arm.linux.org.uk (8.9.3/8.9.3) with ESMTP id RAA06526; Thu, 24 Aug 2000 17:08:16 +0100 Received: (from linux@localhost) by flint.arm.linux.org.uk (8.9.3/8.9.3) id RAA20474; Thu, 24 Aug 2000 17:04:54 +0100 From: Russell King - ARM Linux Admin Message-Id: <200008241604.RAA20474@flint.arm.linux.org.uk> Subject: Re: arch/arm/special To: nico@cam.org (Nicolas Pitre) Date: Thu, 24 Aug 2000 17:04:54 +0100 (BST) Cc: tchan.rd@idthk.com (Tak-Shing Chan), J.A.K.Mouw@its.tudelft.nl (Erik Mouw), linux-arm-kernel@lists.arm.linux.org.uk In-Reply-To: from "Nicolas Pitre" at Aug 24, 2000 12:01:50 PM X-Mailer: ELM [version 2.5 PL1] MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk Nicolas Pitre writes: > At the moment, all that stuff is currently in the now disappeared > arch/arm/special directory... and IMHO it needs a big restructuration > before it goes anywhere. Well, before anyone starts thinking, may I suggest a visit to your nearest linux-kernel mailing list archives for the discussions from this week/last week (the thread about moving input drivers). Also, I'd also suggest having a look at my test7 when it comes out this weekend. But, before anything happens, it probably ought to be discussed on this list first. _____ |_____| ------------------------------------------------- ---+---+- | | Russell King linux@arm.linux.org.uk --- --- | | | | http://www.arm.linux.org.uk/ / / | | +-+-+ --- -+- / | THE developer of ARM Linux |+| /|\ / | | | --- | +-+-+ ------------------------------------------------- /\\\ | unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Thu Aug 24 17:18:43 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13Rzhr-00089Y-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Thu, 24 Aug 2000 17:18:19 +0100 Received: from [140.109.13.42] (helo=linux.org.tw ident=root) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13Rzhp-00089O-00 for linux-arm-kernel@lists.arm.linux.org.uk; Thu, 24 Aug 2000 17:18:17 +0100 Received: from localhost (chester@localhost) by linux.org.tw (8.9.3/8.9.3) with ESMTP id AAA16725; Fri, 25 Aug 2000 00:20:46 +0800 Date: Fri, 25 Aug 2000 00:20:45 +0800 (CST) From: To: Nicolas Pitre cc: S A McConnell , chester@linux.org.tw, Tak-Shing Chan , linux-arm-kernel@lists.arm.linux.org.uk Subject: Re: test6-rmk5-np1 PCMCIA / UCB1200 audio? In-Reply-To: Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk > > to support touch screen, audio and telephony drivers. Go,do it. > > I found that the above drivers had a fair amount of duplicated code. > > (measure_x, measure_y, measure_pressure...) Which I have reduced. > > > > I have questions about how the touch screen should detect events. > > (interrupts vs timer based polling) I am currently using an Interrupt > > to detect touch events. > > Interrupts is always the best. Like the iPAQ touchscreen interface, your > events only need a timestamp to measure velocity. Agreed,that's why i don't like original sa1100_ts.c,it use polling mode. For handwriting recognization,it's a bad soluation for asia glyhps. Chester unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Thu Aug 24 17:24:47 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13Rznh-0008G2-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Thu, 24 Aug 2000 17:24:21 +0100 Received: from [205.250.170.220] (helo=xanadu.vipswitch.com) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13Rzng-0008Ft-00 for linux-arm-kernel@lists.arm.linux.org.uk; Thu, 24 Aug 2000 17:24:20 +0100 Date: Thu, 24 Aug 2000 12:23:36 -0400 (EDT) From: Nicolas Pitre To: Russell King - ARM Linux Admin cc: Tak-Shing Chan , Erik Mouw , linux-arm-kernel@lists.arm.linux.org.uk Subject: Re: arch/arm/special In-Reply-To: <200008241604.RAA20474@flint.arm.linux.org.uk> Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk On Thu, 24 Aug 2000, Russell King - ARM Linux Admin wrote: > Nicolas Pitre writes: > > At the moment, all that stuff is currently in the now disappeared > > arch/arm/special directory... and IMHO it needs a big restructuration > > before it goes anywhere. > > Well, before anyone starts thinking, may I suggest a visit to your > nearest linux-kernel mailing list archives for the discussions > from this week/last week (the thread about moving input drivers). The point is: let's write a proper, modularized, UCB1200 driver. When it's done then we'll bother about the proper location for it. Since this chip provides audio, telephony, touchscreen, and other general purpose IO lines, there might be a submodule in /drivers/sound, another in /drivers/telephony, another in /drivers/char, and the master module in which all submodules need to register with which location is yet not so obvious. All this on the same SPI port and the same interrupt. But again, let's write the code and argue on the location when it's ready. Nicolas unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Thu Aug 24 19:01:17 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13S1GI-0001LI-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Thu, 24 Aug 2000 18:57:58 +0100 Received: from [216.18.11.193] (helo=duke.electric.net) by www.linux.org.uk with smtp (Exim 3.13 #1) id 13S1GH-0001LC-00 for linux-arm-kernel@lists.arm.linux.org.uk; Thu, 24 Aug 2000 18:57:57 +0100 Received: (qmail 21091 invoked from network); 24 Aug 2000 17:57:57 -0000 Received: from localhost (HELO relay.electric.net) (@127.0.0.1) by duke.electric.net with SMTP; 24 Aug 2000 17:57:57 -0000 Received: (from smtpd@localhost) by relay.electric.net (8.8.8/8.8.6) id KAA21077 for linux-arm-kernel@lists.arm.linux.org.uk; Thu, 24 Aug 2000 10:57:56 -0700 (PDT) Received: from cobra.intrinsyc.com(64.16.10.158), claiming to be "dp.intrinsyc.com" via SMTP by relay.electric.net, id smtpdE20995; Thu Aug 24 10:57:54 2000 Received: from ileong (helo=localhost) by dp.intrinsyc.com with local-esmtp (Exim 3.13 #1 (Debian)) id 13S1GD-00008S-00 for ; Thu, 24 Aug 2000 10:57:53 -0700 Date: Thu, 24 Aug 2000 10:57:53 -0700 (PDT) From: I-Gene Leong To: linux-arm-kernel@lists.arm.linux.org.uk Subject: test6-rmk5-np2 and start-stop-daemon (yet again) Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk That PCMCIA fix now causes start-stop-daemon to _hang_. :) Additionally, looks to me as though some heavy work is going on in there; subbing in the test4-rmk1-np1 sa1100_generic.c still causes it to oops. Looks to me as though it's a memory mapping problem, though: cardmgr[33]: starting, version is 3.1.11 cardmgr[33]: watching 2 sockets cardmgr[33]: initializing socket 1 cardmgr[33]: socket 1: ATA/IDE Fixed Disk Unable to handle kernel paging request at virtual address c4000006 pgd = c1dac000 *pgd = 00000000, *pmd = 00000000 Internal error: Oops: 0 CPU: 0 pc : [] lr : [] sp : c1d455e4 ip : a0000013 fp : c1d45604 r10: 60000013 r9 : c01e8a60 r8 : c0149b04 r7 : 000000ec r6 : c01499d0 r5 : c0149b04 r4 : c0149a20 r3 : 000000a0 r2 : c4000006 r1 : c1d455a0 r0 : 00000000 Flags: nZCv IRQs on FIQs on Mode SVC_32 Segment user Control: C1DAD17F Table: C1DAD17F DAC: 00000015 Process cardmgr (pid: 33, stackpage=c1d45000) Stack: c1d455c0: c00203d4 c0092448 60000013 c1d455e0: ffffffff c0149a20 c0149b04 c01499d0 00000000 c0149b04 c1d45630 c1d45608 c1d45600: c0092ac0 c00923e4 c1d45634 c01499d0 00000000 c1d45634 c01499d0 c0128b04 c1d45620: 00000000 c1d45664 c1d45634 c00934e8 c00929d4 00000001 00000001 00000001 c1d45640: 00000001 c01499d0 00000000 c013ad84 c01499d0 c01499d0 c1d45674 c1d45668 c1d45660: c008e948 c0093480 c1d4569c c1d45678 c008f2ec c008e930 c4000008 c1d456a0 c1d45680: e400000e 0000002b e4000000 e400000e c1d45704 c1d456a0 c008f390 c008f1cc c1d456a0: c4000000 c4000001 c4000002 c4000003 c4000004 c4000005 c4000006 c4000007 c1d456c0: c400000e 00000000 0000002b 00000000 00000000 00000000 00000000 00000000 c1d456e0: 00000000 00000000 c1d45b54 00000000 00000000 c01e8b60 c1d45b9c c1d45708 c1d45700: c009502c c008f338 00000000 00000000 ffffffff 00000000 c01e8b60 21095901 c1d45720: 00000201 00000027 0007a120 0006ddd0 00086470 c1ff1800 00000002 000c3500 c1d45740: ec000000 c1d45900 c1d45918 c009b994 c009fb9c c1d45940 c1d45986 c1ff1800 c1d45760: c1ff1800 00000000 c1ff1b74 00000005 c1ff1800 00000017 c1ff1800 ec000000 c1d45780: c1d45984 00000000 c009b994 00000000 00000016 00000000 c1ff1800 c1ff0164 c1d457a0: 00000000 00000010 00000006 00000018 c1d459d0 c1d459e7 c01e8660 c1d45b0c c1d457c0: c1d45ccc 00000000 bffff85c c1d45c80 c1d459bc c1d459a4 c009ccd0 c009cba4 c1d457e0: c1d45b0f c1d45b0e c1d45ad0 c1d459cc c1d459c0 c009db6c c009cc80 c1d45b08 c1d45800: c1d459d0 c1ff1800 00000002 c1ff1bab 00000016 000000b4 c1d45a28 c1d45a1c c1d45820: 000000f0 0000ffff c00f00cc 00000800 00000000 00000000 00000001 c1d45b74 c1d45840: 000000b4 00000001 c1ff1800 c1d45a50 c1d45a20 c009c5d4 c009c390 c1d45a28 c1d45860: c1d45a44 c1d4ffff c1d45b74 c01e8a60 c01e8a60 00000001 c01e8b8c 00000004 c1d45880: c1d45a70 c1d45a54 c009c378 c009c500 c1ff0300 00000000 00000021 00000021 c1d458a0: 00000021 c0037ed0 00000000 c1d458d0 c1d458bc c00a47a8 c00a45c8 c01e8260 c1d458c0: c012888c c1d458ec c1d458d4 c008bdf8 c00a4694 c01e8360 c012888c c1ff0300 c1d458e0: 00005901 00000200 00000021 00050910 0006ddd0 00086470 c1ff1800 00000002 c1d45900: 0006ddd0 ec000000 c1d45900 c1d45918 c009b994 c009fb9c c1d45940 c1d45986 c1d45920: c1ff1800 c1ff1800 00000000 c1ff1b74 00000005 c1ff1800 00000017 c1ff1800 c1d45940: ec000000 c1d45984 00000000 c009b994 00000000 00000016 00000000 c1ff1800 c1d45960: c1ff0064 00000000 00000010 00000006 00000018 c1d459d0 c1d459e7 c01e8660 c1d45980: c1d45b0c c1d45ccc 00000000 bffff85c c1d45c80 c1d459bc c1d459a4 c009ccd0 c1d459a0: c009cba4 c1d45b0f c1d45b0e c1d45ad0 c1d459cc c1d459c0 c009db6c c009cc80 c1d459c0: c1d45b08 c1d459d0 c1ff1800 00000002 c1ff1bab 00000016 000000b4 c1d45a28 c1d459e0: c1d45a1c 00000000 0000ffff c00f00cc 00000800 00000000 00000000 00000001 c1d45a00: c1d45b74 000000b4 00000001 c1ff1800 c1d45a50 c1d45a20 c009c5d4 c009c390 c1d45a20: c1d45a28 c1d45a44 c1d4ffff c1d45b74 c01e8a60 c01e8a60 00000001 c01e8b8c c1d45a40: 00000004 c1d45a70 c1d45a54 c009c378 c009c500 b5210101 755d4d1e fffff064 c1d45a60: 00000021 c1d45bac c1d45a74 c009dc28 c009c284 c1d45ac0 c1d45ae2 c1ff1800 c1d45a80: c1ff1800 00000002 c1ff1bab 00000016 000000b4 c1d45ae0 c1d45ad4 c1d45aa8 c1d45aa0: c009c120 c00f14cc c1d45ae0 c1d45c80 000000b4 00000001 c1d45c80 000000b4 c1d45ac0: 00000001 c1ff1800 c1d45b08 c1d45ad8 c009c5d4 c009c390 c1d45ae0 00000000 c1d45ae0: 0000ffff 00000000 c1d45b0c c01e8660 c1d45ccc 00000000 c1d45c80 c1d45cc4 c1d45b00: c1d45b0c c009ddcc c009c518 00030104 530b0c08 00000001 c1e5d000 00000000 c1d45b20: c013cca0 00000000 c011c080 c013cca0 c1d45c00 c1d45b54 c1d45b44 c0089828 c1d45b40: c00208d0 00000000 c1d45b6c c1d45b58 c00897b8 00000000 c013c91b 00000110 c1d45b60: 00000000 0000006e ff00061b c0089706 c1d45a54 c1ff1800 c01e8a60 c01e8a9c c1d45b80: 00000001 c01e8b8c c014a5b0 00000004 c1d45bac c1d45ba0 c0095280 c0094c68 c1d45ba0: c1d45c20 c1d45bb0 c0099f64 c00951f8 c0027ed0 00000000 00000011 20000000 c1d45bc0: c012bf38 c1d45bfc c1d45bd4 c0015b90 c0027ce8 fa050000 c1d45c48 00000002 c1d45be0: c009bfd8 80000013 ffffffff 00000001 c1d45c70 c1d45c00 c01e8b8c c1d45c44 c1d45c00: c0128c50 c0128c30 c00951ec 02011eb8 c01e84a4 c1d45c40 c1d45c24 c009b614 c1d45c20: c0099d30 c1ff1bab c01f78e0 c01dc260 c01e8b60 c1d45c94 c1d45c44 c0094b68 c1d45c40: c009b3c4 c0128c30 00000012 0000660c c00951ec c1d45ccc 000000b4 00000001 c1d45c60: 00000000 c1d45c98 c01e8b60 c1d45c94 00000210 c01f78e0 c01dc260 00000000 c1d45c80: c1d45ccc 00000001 c1d45cc4 c1d45c98 c00971c8 c0094a48 c1000001 c01dc260 c1d45ca0: c050643c c01e84a4 00000000 00000000 00000001 00000050 c1d45f88 c1d45cc8 c1d45cc0: c0097e00 c0097060 80000000 5f656469 00007363 00000000 00000000 00000000 c1d45ce0: 00000000 00000000 00000000 00000000 00000000 00000000 00000000 00000000 c1d45d00: 00000000 00000000 00000000 00000000 00000000 00000000 00000000 c1d45d28 c1d45d20: c003be10 c0039978 c1d45d58 c1d45d38 c003becc c003bde8 c000001f 00000000 c1d45d40: c1fbb300 c1fb95a0 c1dab454 c1d45d7c c1d45d5c c002f1e4 c002f200 c1dab454 c1d45d60: c1fb95a0 c1fbb300 00000001 c1dab454 c1d45da4 c1d45d80 c002f29c c002f19c c1d45d80: 40015000 c1dab454 c1d44000 c1fb95a0 c1dad000 40015000 c1d45de0 c1d45da8 c1d45da0: c002f484 c002f254 c1dab454 00000000 00000001 00000001 c1fb95a0 c1d44000 c1d45dc0: ffffffff 40015000 00000000 00000000 c1fbb300 c1d45e08 c1d45de4 c001a55c c1d45de0: c002f360 c1fbb31c c1d44000 c1d45ee4 40015000 c1fbb300 00000400 c1d45ec0 c1d45e00: c1d45e0c c001a700 c001a4f4 c0132550 00001104 00000000 c1d45e24 c0053604 c1d45e20: c00208d0 00000000 00000000 c01dbfd8 00000000 00000000 00000000 c1d45e4c c1d45e40: c0112750 00000000 00000020 c0053930 c0053520 00000000 c01dd800 c1fea120 c1d45e60: 00001104 c1d22c60 ffffffea 00000001 00000000 c1d45e9c c1d45e84 c0056eb0 c1d45e80: c0053868 c1fea120 c1d266a0 c1fea120 00000000 00000000 00000000 00000000 c1d45ea0: 00000000 c1d1301b ffffffff 60000013 00000e88 c01c1c18 c01c1c00 c0123544 c1d45ec0: 00000000 c1dad004 00115000 c1dab458 c1d45eec c1d45ee0 c0039b80 c0039450 c1d45ee0: c1d45f04 c1d45ef0 c003a228 c0039b40 c1d100bf 00000001 c1d45f50 c1d45f08 c1d45f00: c002da7c c003a178 40115000 c1dad004 00000001 00000001 c1fb95a0 40015000 c1d45f20: c1d45f50 c1d45f30 c00302dc c0037d9c c1d44000 c1fb95a0 c1d44000 c1d44000 c1d45f40: 00000000 c1d45f94 c1d45f54 c00307e8 c00303e8 c1fb9760 40016000 00000000 c1d45f60: 02011eb8 c0012720 c050643c ffffffe7 00000001 00000036 00000000 c1d45fac c1d45f80: c1d45f8c c004e008 c0097880 00000001 02011eb8 20000010 00000000 c00150a4 c1d45fa0: 00000000 c1d45fb0 c0014f20 c004ddc8 00000001 c001ae74 00000001 c050643c c1d45fc0: 02011eb8 00000000 00000001 02011eb8 02010fd8 00000000 020224a8 02011014 c1d45fe0: 00000000 bffffb4c 02010ef0 bffffb1c 02002aa0 400b3b54 20000010 00000001 Backtrace: Function entered at [do_probe] from [] r8 = C0149B04 r7 = 00000000 r6 = C01499D0 r5 = C0149B04 r4 = C0149A20 Function entered at [probe_hwif] from [] Function entered at [ideprobe_init] from [] r8 = C01499D0 r7 = C01499D0 r6 = C013AD84 r5 = 00000000 r4 = C01499D0 Function entered at [ide_probe_module] from [] Function entered at [ide_register_hw] from [] Function entered at [ide_register] from [] r7 = C01E8B60 r6 = 00000000 r5 = 00000000 r4 = C1D45B54 Function entered at [ide_config] from [] Function entered at [ide_event] from [] Function entered at [pcmcia_register_client] from [] Function entered at [CardServices] from [] r4 = C01E8B60 Function entered at [ide_attach] from [] r8 = 00000001 r7 = C1D45CCC r6 = 00000000 r5 = C01DC260 r4 = C01F78E0 Function entered at [bind_request] from [] Function entered at [ds_ioctl] from [] Function entered at [sys_ioctl] from [] r8 = C00150A4 r7 = 00000000 r6 = 20000010 r5 = 02011EB8 r4 = 00000001 Code: e5962020 e5d430a8 (e5c23000) ebfff626 e59430cc [yes, I did sub in the function names from System.map] -- I-Gene Leong ileong@intrinsyc.com / 604-646-6962 "We see Linux as being much of a fad as the Internet was in 1995." - Irving Wladawsky-Berger, IBM VP of technology and strategy unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Thu Aug 24 20:30:30 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13S2aL-0003yC-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Thu, 24 Aug 2000 20:22:45 +0100 Received: from [216.18.11.193] (helo=duke.electric.net) by www.linux.org.uk with smtp (Exim 3.13 #1) id 13S2aJ-0003xu-00 for linux-arm-kernel@lists.arm.linux.org.uk; Thu, 24 Aug 2000 20:22:43 +0100 Received: (qmail 55674 invoked from network); 24 Aug 2000 19:22:34 -0000 Received: from localhost (HELO relay.electric.net) (@127.0.0.1) by duke.electric.net with SMTP; 24 Aug 2000 19:22:34 -0000 Received: (from smtpd@localhost) by relay.electric.net (8.8.8/8.8.6) id MAA55638 for linux-arm-kernel@lists.arm.linux.org.uk; Thu, 24 Aug 2000 12:22:34 -0700 (PDT) Received: from cobra.intrinsyc.com(64.16.10.158), claiming to be "dp.intrinsyc.com" via SMTP by relay.electric.net, id smtpdW55534; Thu Aug 24 12:22:30 2000 Received: from ileong (helo=localhost) by dp.intrinsyc.com with local-esmtp (Exim 3.13 #1 (Debian)) id 13S2a5-0000Jz-00 for ; Thu, 24 Aug 2000 12:22:29 -0700 Date: Thu, 24 Aug 2000 12:22:29 -0700 (PDT) From: I-Gene Leong To: linux-arm-kernel@lists.arm.linux.org.uk Subject: Poking around in PCMCIA, t6-rmk5-np2 Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk I saw Erik Mouw's changes to include/asm-arm/arch-sa1100/ide.h and stuffed in a printk to see what was being passed in. The t4-rmk1-np1 ide.h just used the port that was sent in, while the t6-rmk5-np2 ide.h added a base value to it that corresponds to PCMCIA slot 0. What I'm talking about: --- t4-rmk1-np1 static __inline__ void ide_init_hwif_ports(hw_regs_t *hw, int data_port, int ctrl_port, int *irq) { ide_ioreg_t reg; int i; int ioshift = 0; /* The Empeg board has the first two address lines unused */ if (machine_is_empeg()) ioshift = 2; memset(hw, 0, sizeof(*hw)); reg = (ide_ioreg_t) (data_port << ioshift); for (i = IDE_DATA_OFFSET; i <= IDE_STATUS_OFFSET; i++) { hw->io_ports[i] = reg; reg += (1 << ioshift); } --- --- t6-rmk5-np2 #define PCMCIA_IO_0_BASE 0xe0000000 #define PCMCIA_IO_1_BASE 0xe4000000 /* * Set up a hw structure for a specified data port, control port and IRQ. * This should follow whatever the default interface uses. */ static __inline__ void ide_init_hwif_ports(hw_regs_t *hw, int data_port, int ctrl_port, int *irq) { ide_ioreg_t reg; int i; int ioshift = 0; int base = PCMCIA_IO_0_BASE; /* The Empeg board has the first two address lines unused */ if (machine_is_empeg()) ioshift = 2; /* The LART doesn't use A0 for IDE */ if (machine_is_lart()) ioshift = 1; memset(hw, 0, sizeof(*hw)); reg = (ide_ioreg_t) (base + (data_port << ioshift)); for (i = IDE_DATA_OFFSET; i <= IDE_STATUS_OFFSET; i++) { hw->io_ports[i] = reg; reg += (1 << ioshift); } --- Now, the ide/pcmcia driver code passes in the actual absolute address for the dataport. I'm guessing that dataport is meant to be relative, since it's shifted by ioshift. Anyway, for the Cerf, which has one CF+ slot, data_port is passed in as e4000000 (PCMCIA slot 1). For t4-rmk1-np1 it stays as e4000000, while t6-rmk5-np2 adds PCMCIA_IO_0_BASE (e0000000) to it, giving me c4000000. So there's where my kernel oops is coming from. The question is, how should it be fixed? I can think of a quick fix: editing ide.h not to add base on. But it doesn't look to me as though that's the intention because data_port is shifted, which makes no sense for an absolute address. The alternative is to go digging through the pcmcia code to find out where this port is set... in which case, how will I indicate that it's PCMCIA slot 1 I want, and not PCMCIA slot 0? -- I-Gene Leong ileong@intrinsyc.com / 604-646-6962 "We see Linux as being much of a fad as the Internet was in 1995." - Irving Wladawsky-Berger, IBM VP of technology and strategy unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Thu Aug 24 21:12:16 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13S3Ks-0004u1-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Thu, 24 Aug 2000 21:10:50 +0100 Received: from [216.18.11.193] (helo=duke.electric.net) by www.linux.org.uk with smtp (Exim 3.13 #1) id 13S3Kq-0004ti-00 for linux-arm-kernel@lists.arm.linux.org.uk; Thu, 24 Aug 2000 21:10:48 +0100 Received: (qmail 26182 invoked from network); 24 Aug 2000 20:10:41 -0000 Received: from localhost (HELO relay.electric.net) (@127.0.0.1) by duke.electric.net with SMTP; 24 Aug 2000 20:10:41 -0000 Received: (from smtpd@localhost) by relay.electric.net (8.8.8/8.8.6) id NAA26172; Thu, 24 Aug 2000 13:10:40 -0700 (PDT) Received: from cobra.intrinsyc.com(64.16.10.158), claiming to be "dp.intrinsyc.com" via SMTP by relay.electric.net, id smtpdu25948; Thu Aug 24 13:10:36 2000 Received: from ileong (helo=localhost) by dp.intrinsyc.com with local-esmtp (Exim 3.13 #1 (Debian)) id 13S3Kd-0000Oc-00; Thu, 24 Aug 2000 13:10:35 -0700 Date: Thu, 24 Aug 2000 13:10:35 -0700 (PDT) From: I-Gene Leong To: linux-arm-kernel@lists.arm.linux.org.uk cc: linux@arm.linux.org.uk, nico@cam.org Subject: More PCMCIA thoughts Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk I wondered about changing the PCMCIA driver to accomodate the ide.h in test6-rmk5-np2... but then realized that would break everything else that depended on PCMCIA handing out absolute addresses for I/O (probably nearly everything). So I made a little change to ide.h that will probably break a lot of stuff... this is tested to work on a Cerf, and will probably work for PCMCIA IDE, but again, this is a horrible, horrible way to go. Oh, and this fixes the start-stop-daemon problem, too... wonder why... --- old/include/asm-arm/arch-sa1100/ide.h Thu Aug 24 10:25:11 2000 +++ linuxwork/include/asm-arm/arch-sa1100/ide.h Thu Aug 24 13:03:28 2000 @@ -29,7 +29,10 @@ ide_ioreg_t reg; int i; int ioshift = 0; - int base = PCMCIA_IO_0_BASE; +/* int base = PCMCIA_IO_0_BASE; */ + int base = 0; /* nasty hack; can't figure + this out, someone else + do it */ /* The Empeg board has the first two address lines unused */ if (machine_is_empeg()) -- I-Gene Leong ileong@intrinsyc.com / 604-646-6962 "We see Linux as being much of a fad as the Internet was in 1995." - Irving Wladawsky-Berger, IBM VP of technology and strategy unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Thu Aug 24 21:39:52 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13S3ld-0005Sb-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Thu, 24 Aug 2000 21:38:29 +0100 Received: from [130.161.42.1] (helo=duteinh.et.tudelft.nl) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13S3lc-0005ST-00 for linux-arm-kernel@lists.arm.linux.org.uk; Thu, 24 Aug 2000 21:38:29 +0100 Received: from dutein75 by duteinh.et.tudelft.nl (8.9.3/BB-2.0) id WAA06422; Thu, 24 Aug 2000 22:38:18 +0200 (MET DST) Message-Id: <200008242038.WAA06422@duteinh.et.tudelft.nl> From: Erik Mouw To: ileong@intrinsyc.com Cc: linux-arm-kernel@lists.arm.linux.org.uk, nico@cam.org Subject: Re: Poking around in PCMCIA, t6-rmk5-np2 In-Reply-To: Organization: Eric conspiracy secret labs Date: Thu 24 Aug 2000 22:38:17 +0100 (MDT) X-Face: "!Z|1XjFg!3Iw)BwM^M\k)J5KHiPddz}#/Zw%?$&Q!xqob"OA|})3_@&BTTiQERo?/^g[Z<9P4R8:5C:aqq.cw!hotLc=+u-"+1rq@9$V^8PbOUlodyi_sejt+D`iqkLpx$>.aXl%vnfPIt"LK|/%>94_Zrm8jy?`_[RKt'TuoxUxfk(gMhY(n|DBBl~c$dhO>G%6}R4+DGX3|idBm4oU&UM/XUP4I9aoN{]`ax%B?rgD(I3OJ<62Y6L X-Mailer: XCmail 1.2devel - with PGP support, PGP engine version 0.5 (IRIX64) X-Mailerorigin: http://www.fsai.fh-trier.de/~schmitzj/Xclasses/XCmail/ X-Eric-Conspiracy: There is no conspiracy! X-Loop: erik@it.et.tudelft.nl MIME-Version: 1.0 Content-Type: multipart/mixed; boundary="boundary_XCmail_1.2devel_39A587B9_dutein75" Content-Transfer-Encoding: 8bit Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk This is the preamble area of a multipart message. Mail readers that understand multipart format should ignore this preamble. If you are reading this text, you might want to consider changing to a mail reader that understands how to properly display multipart messages. (like the cool XCmail) --boundary_XCmail_1.2devel_39A587B9_dutein75 Content-Type: text/plain Content-Description: Main text Content-Transfer-Encoding: 8bit On Thu, 24 Aug 2000 12:22:29 -0700 (PDT), I-Gene Leong wrote: > I saw Erik Mouw's changes to include/asm-arm/arch-sa1100/ide.h and stuffed > in a printk to see what was being passed in. The t4-rmk1-np1 ide.h just > used the port that was sent in, while the t6-rmk5-np2 ide.h added a base > value to it that corresponds to PCMCIA slot 0. > > What I'm talking about: [...] > > --- t6-rmk5-np2 [...] > --- > > Now, the ide/pcmcia driver code passes in the actual absolute address for > the dataport. > > I'm guessing that dataport is meant to be relative, since it's shifted by > ioshift. Yes, it was meant to be relative. > Anyway, for the Cerf, which has one CF+ slot, data_port is passed > in as e4000000 (PCMCIA slot 1). For t4-rmk1-np1 it stays as e4000000, > while t6-rmk5-np2 adds PCMCIA_IO_0_BASE (e0000000) to it, giving me > c4000000. So there's where my kernel oops is coming from. I see. Well, Cerf was not mentioned in the ide.h file, so I didn't think of it. I also didn't realize that the PCMCIA code used this function, sorry. > The question is, how should it be fixed? I can think of a quick > fix: editing ide.h not to add base on. But it doesn't look to me as though > that's the intention because data_port is shifted, which makes no sense > for an absolute address. The alternative is to go digging through the > pcmcia code to find out where this port is set... in which case, how will > I indicate that it's PCMCIA slot 1 I want, and not PCMCIA slot 0? Try my attached patch, it makes the address absolute again. I tested it with the LART (it boots, ship it ;-) ), it should also work for the Cerf, Empeg and the Victor. Nico, please apply this patch. Erik -- J.A.K. (Erik) Mouw, Information and Communication Theory Group, Department of Electrical Engineering, Faculty of Information Technology and Systems, Delft University of Technology, PO BOX 5031, 2600 GA Delft, The Netherlands Phone: +31-15-2783635 Fax: +31-15-2781843 Email: J.A.K.Mouw@its.tudelft.nl WWW: http://www-ict.its.tudelft.nl/~erik/ --boundary_XCmail_1.2devel_39A587B9_dutein75 Content-Type: text/plain; name=diff-2.4.0-test6-rmk5-np2-em1-ide.diff Content-Description: File attachment Content-Transfer-Encoding: quoted-printable --- linux-2.4.0-test6-rmk5-np2-vanilla/include/asm-arm/arch-sa1100/ide.h Th= u Aug 24 13:49:09 2000 +++ linux-2.4.0-test6-rmk5-np2/include/asm-arm/arch-sa1100/ide.h Thu Aug 24= 22:32:30 2000 @@ -28,27 +28,26 @@ { ide_ioreg_t reg; int i; - int ioshift =3D 0; - int base =3D PCMCIA_IO_0_BASE; - + int regincr =3D 1; +=09 /* The Empeg board has the first two address lines unused */ if (machine_is_empeg()) - ioshift =3D 2; + regincr =3D 1 << 2; =20 /* The LART doesn't use A0 for IDE */ if (machine_is_lart()) - ioshift =3D 1; + regincr =3D 1 << 1; =20 memset(hw, 0, sizeof(*hw)); =20 - reg =3D (ide_ioreg_t) (base + (data_port << ioshift)); + reg =3D (ide_ioreg_t)data_port; + for (i =3D IDE_DATA_OFFSET; i <=3D IDE_STATUS_OFFSET; i++) { hw->io_ports[i] =3D reg; - reg +=3D (1 << ioshift); + reg +=3D regincr; } =09 - hw->io_ports[IDE_CONTROL_OFFSET] =3D=20 - (ide_ioreg_t) (base + (ctrl_port << ioshift)); + hw->io_ports[IDE_CONTROL_OFFSET] =3D (ide_ioreg_t) ctrl_port; =09 if (irq) *irq =3D 0; @@ -87,10 +86,10 @@ /* MAC 23/4/1999, swap these round so that the left hand hard disk is hda when viewed from the front. This doesn't match the silkscreen however. */ - ide_init_hwif_ports(&hw, 0x10, 0x1e, NULL); + ide_init_hwif_ports(&hw, PCMCIA_IO_0_BASE + 0x40, PCMCIA_IO_0_BASE + 0x78= , NULL); hw.irq =3D EMPEG_IRQ_IDE2; ide_register_hw(&hw, NULL); - ide_init_hwif_ports(&hw, 0x00, 0x0e, NULL); + ide_init_hwif_ports(&hw, PCMCIA_IO_0_BASE + 0x00, PCMCIA_IO_0_BASE + 0x38= , NULL); hw.irq =3D ,EMPEG_IRQ_IDE1; ide_register_hw(&hw, NULL); #endif @@ -107,7 +106,7 @@ /* set the pcmcia interface timing */ MECR =3D 0x00060006; =20 - ide_init_hwif_ports(&hw, 0x1f0, 0x3f6, NULL); + ide_init_hwif_ports(&hw, PCMCIA_IO_0_BASE + 0x1f0, PCMCIA_IO_0_BASE + 0x3= f6, NULL); hw.irq =3D IRQ_GPIO7; ide_register_hw(&hw, NULL); #endif @@ -124,7 +123,7 @@ MECR =3D 0x00060006; =20 /* init the interface */ - ide_init_hwif_ports(&hw, 0x0000, 0x0800, NULL); + ide_init_hwif_ports(&hw, PCMCIA_IO_0_BASE + 0x0000, PCMCIA_IO_0_BASE + 0x= 1000, NULL); hw.irq =3D IRQ_GPIO1; ide_register_hw(&hw, NULL); #endif --boundary_XCmail_1.2devel_39A587B9_dutein75-- unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Thu Aug 24 21:55:25 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13S3y9-0005gd-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Thu, 24 Aug 2000 21:51:25 +0100 Received: from [216.18.11.193] (helo=duke.electric.net) by www.linux.org.uk with smtp (Exim 3.13 #1) id 13S3y8-0005gW-00 for linux-arm-kernel@lists.arm.linux.org.uk; Thu, 24 Aug 2000 21:51:25 +0100 Received: (qmail 87531 invoked from network); 24 Aug 2000 20:51:26 -0000 Received: from localhost (HELO relay.electric.net) (@127.0.0.1) by duke.electric.net with SMTP; 24 Aug 2000 20:51:26 -0000 Received: (from smtpd@localhost) by relay.electric.net (8.8.8/8.8.6) id NAA87510; Thu, 24 Aug 2000 13:51:25 -0700 (PDT) Received: from cobra.intrinsyc.com(64.16.10.158), claiming to be "dp.intrinsyc.com" via SMTP by relay.electric.net, id smtpdC87457; Thu Aug 24 13:51:24 2000 Received: from ileong (helo=localhost) by dp.intrinsyc.com with local-esmtp (Exim 3.13 #1 (Debian)) id 13S3y7-0000Ug-00; Thu, 24 Aug 2000 13:51:23 -0700 Date: Thu, 24 Aug 2000 13:51:23 -0700 (PDT) From: I-Gene Leong To: Erik Mouw cc: linux-arm-kernel@lists.arm.linux.org.uk, nico@cam.org Subject: Re: Poking around in PCMCIA, t6-rmk5-np2 In-Reply-To: <200008242038.WAA06422@duteinh.et.tudelft.nl> Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk Your patch works here, thanks. Disregard my pathetic-excuse-for-a-patch, everyone. :) -- I-Gene Leong ileong@intrinsyc.com / 604-646-6962 "We see Linux as being much of a fad as the Internet was in 1995." - Irving Wladawsky-Berger, IBM VP of technology and strategy unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Thu Aug 24 22:42:14 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13S4kX-0006WF-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Thu, 24 Aug 2000 22:41:25 +0100 Received: from [195.70.253.46] (helo=smtp1.utaiop.at) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13S4kW-0006W7-00 for linux-arm-kernel@lists.arm.linux.org.uk; Thu, 24 Aug 2000 22:41:24 +0100 Received: from multi1.utaiop.at ([10.1.1.195]) by smtp1.utaiop.at (Netscape Messaging Server 4.15) with ESMTP id FZTG7V01.Z2D for ; Thu, 24 Aug 2000 23:40:43 +0200 Received: from wshome ([212.152.138.229]) by multi1.utaiop.at (Netscape Messaging Server 4.15) with SMTP id FZTG7V00.932 for ; Thu, 24 Aug 2000 23:40:43 +0200 From: "Thomas Ritter" To: "linux-arm-kernel" Subject: AW: What does "external abort on non-linefetch" mean? Date: Thu, 24 Aug 2000 23:44:49 +0200 Message-ID: <000a01c00e14$873ab400$e58a98d4@wshome> MIME-Version: 1.0 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: 8bit X-Priority: 3 (Normal) X-MSMail-Priority: Normal X-Mailer: Microsoft Outlook 8.5, Build 4.71.2173.0 In-Reply-To: <967023115.87257.cm_lee@mail2000.com.tw> X-MimeOLE: Produced By Microsoft MimeOLE V5.00.2919.6700 Importance: Normal Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk Hi, I got this message when trying to access a reserved memory region. e.g. addr >= 0x50000000 ... ;-) See ARM architecture reference manual. tom > -----Ursprüngliche Nachricht----- > Von: owner-linux-arm-kernel@lists.arm.linux.org.uk > [mailto:owner-linux-arm-kernel@lists.arm.linux.org.uk]Im Auftrag von > Chien-Min Lee > Gesendet am: Mittwoch, 23. August 2000 11:32 > An: linux-arm-kernel > Betreff: What does "external abort on non-linefetch" mean? > > Dear All: > > I got the following error when testing a driver on arm-linux. > > [root@Linux /root]$Internal error: external abort on non-linefetch: 28 > CPU: 0 > pc : [] lr : [] > sp : c012ff68 ip : 00000000 fp : c012ffc0 > r10: c001307c r9 : ffffffff r8 : 80000013 > r7 : c00203ac r6 : c0019b1c r5 : c012ffb0 r4 : fa050000 > r3 : c01431c4 r2 : 00000000 r1 : 00000000 r0 : 00000003 > Flags: Nzcv IRQs off FIQs on Mode SVC_32 Segment kernel > Control: C165917F Table: C165917F DAC: 0000001D > Process swapper (pid: 0, stackpage=c012f000) > Stack: > [skip .....] > > However, I do not understand this error message. > Could someone tell me what it talks about?? > > Thanks > > kirk > > > > ----== Mailed via Openfind ==----- > http://mail2000.com.tw/ ´£¨Ñ§K¶O¹q¤l¶l¥ó«H½c > > > unsubscribe: body of `unsubscribe linux-arm-kernel' to > majordomo@lists.arm.linux.org.uk > unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Fri Aug 25 04:16:38 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13S9x3-00039J-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Fri, 25 Aug 2000 04:14:41 +0100 Received: from [210.184.5.125] (helo=pikachu.idthk.com) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13S9wx-00038V-00 for linux-arm-kernel@lists.arm.linux.org.uk; Fri, 25 Aug 2000 04:14:38 +0100 Received: from localhost (thomas@localhost) by pikachu.idthk.com (8.9.3/8.8.7) with ESMTP id LAA21470 for ; Fri, 25 Aug 2000 11:15:15 +0800 Date: Fri, 25 Aug 2000 11:15:15 +0800 (CST) From: Tak-Shing Chan To: linux-arm-kernel@lists.arm.linux.org.uk Subject: Re: test6-rmk5-np1 PCMCIA / UCB1200 audio? In-Reply-To: <39A520A0.110B0F16@cotw.com> Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk On Thu, 24 Aug 2000, S A McConnell wrote: > I am interested in a "generic" interface to the UCB1200 that can be used > to support touch screen, audio and telephony drivers. We can perhaps reuse arch/arm/special/mcp_common.c. > I have been working on a touch screen based upon sa1100_ts.c and > ucb1100_ts.c. Have I made a mistake looking at those drivers? Which version? There are no less than 3 different UCB1200 touchscreen drivers in ARM Linux's history (the Itsy port, Tifon, ADS/Microwindows and various others). > I also created several macro's that significantly reduce the complexity > of the UCB register reads and writes. The sa1100_ts.h in test6 is derived from Itsy and it does inlined codec_read() and codec_write(), similar to macros. A better solution is perhaps to use the original Itsy codec header file that is already integrated with mcp_common.c. (Itsy drivers are written for an obsolete version of ARM Linux; some heavy portings are needed. Erik Mouw has already done a lot of such portings.) > I have not been able to find out who wrote the sa1100_ts.c and > ucb1100_ts.c code. The ones before test6 are originally by Century Software (ADS/Microwindows); the ones since test6 use interrupts and are originally by Peter Danielsson (Tifon). Both have been patched heavily. > I have questions about how the touch screen should detect events. > (interrupts vs timer based polling) I am currently using an Interrupt > to detect touch events. Interrupts can detect pen downs only; you need a timer for continuous sampling during the pen down period or else you can't drag the ``mouse''. > You discussion on char vs block is very confusing. This is a character > device right? Sorry about the confusion. This is certainly a character device, of course. :) In early August, I was modifying the ucb1200_ts.c driver such that it uses interrupts and that it can emulate an iPAQ. The result is that I've pasted in the Tifon driver's code and deleted most of the original code, and added various stuffs from various sources. The touchscreen specs said the major should be 10 but as I'm too lazy to switch to misc_register() just yet, I use 11,0 instead thinking that it is the unused /dev/scd0 (unfortunately it isn't). In the meantime I've already switched to misc_register() with a minor of 240 and have preliminary integration with mcp_common.c (derived from Chester's code) but the touchscreen specs and the code layout in the linux directories are still not sorted out yet... (I'll post my patch here somewhere in September when the issues are resolved.) Tak-Shing Chan unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Fri Aug 25 08:26:35 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13SDqT-0006eP-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Fri, 25 Aug 2000 08:24:09 +0100 Received: from [210.184.5.125] (helo=pikachu.idthk.com) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13SDqQ-0006dk-00 for linux-arm-kernel@lists.arm.linux.org.uk; Fri, 25 Aug 2000 08:24:07 +0100 Received: from localhost (thomas@localhost) by pikachu.idthk.com (8.9.3/8.8.7) with ESMTP id PAA22469; Fri, 25 Aug 2000 15:24:44 +0800 Date: Fri, 25 Aug 2000 15:24:44 +0800 (CST) From: Tak-Shing Chan To: Nicolas Pitre cc: linux-arm-kernel@lists.arm.linux.org.uk Subject: Re: arch/arm/special In-Reply-To: Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk On Thu, 24 Aug 2000, Nicolas Pitre wrote: > But again, let's write the code and argue on the location when it's ready. And before anyone writes or modifies anything, should we have a who's-doing-what list in order to minimize duplicated efforts? Tak-Shing Chan unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Fri Aug 25 16:55:41 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13SLnS-0005tf-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Fri, 25 Aug 2000 16:53:34 +0100 Received: from [205.250.170.220] (helo=xanadu.vipswitch.com) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13SLnR-0005tZ-00 for linux-arm-kernel@lists.arm.linux.org.uk; Fri, 25 Aug 2000 16:53:33 +0100 Date: Fri, 25 Aug 2000 11:53:11 -0400 (EDT) From: Nicolas Pitre To: Tak-Shing Chan cc: linux-arm-kernel@lists.arm.linux.org.uk, sa1100-linux@pa.dec.com Subject: Re: test6-rmk5-np1 PCMCIA / UCB1200 audio? In-Reply-To: Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk On Fri, 25 Aug 2000, Tak-Shing Chan wrote: > On Thu, 24 Aug 2000, S A McConnell wrote: > > > I am interested in a "generic" interface to the UCB1200 that can be used > > to support touch screen, audio and telephony drivers. > We can perhaps reuse arch/arm/special/mcp_common.c. Yes, and rename it to ucb1200-generic.c, or better yet sa1100-ucb1200.c for more clarity. This could become ucb1200-xyz.c if ever some other architecture xyz uses the UCB1200 chip. The slave drivers could be ucb1200-ts for touchscreen, ucb1200-audio for audio, etc. > And before anyone writes or modifies anything, should we > have a who's-doing-what list in order to minimize duplicated > efforts? I guess volunteers might identify themselves on this list. You are one? :-) Nicolas unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Fri Aug 25 20:32:07 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13SPBL-0000em-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Fri, 25 Aug 2000 20:30:27 +0100 Received: from [208.242.241.39] (helo=cvsftp.cotw.com ident=root) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13SPBK-0000eg-00 for linux-arm-kernel@lists.arm.linux.org.uk; Fri, 25 Aug 2000 20:30:26 +0100 Received: from cotw.com (IDENT:samcconn@ptecdev2.ptec.com [192.168.10.4]) by cvsftp.cotw.com (8.9.3/8.9.3) with ESMTP id OAA24023; Fri, 25 Aug 2000 14:29:44 -0500 Message-ID: <39A6C9D5.606F5F82@cotw.com> Date: Fri, 25 Aug 2000 14:32:37 -0500 From: S A McConnell Reply-To: samcconn@cotw.com X-Mailer: Mozilla 4.7 [en] (X11; I; Linux 2.2.14 i686) X-Accept-Language: en MIME-Version: 1.0 To: Nicolas Pitre CC: Tak-Shing Chan , linux-arm-kernel@lists.arm.linux.org.uk, sa1100-linux@pa.dec.com Subject: Re: test6-rmk5-np1 PCMCIA / UCB1200 audio? References: Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk Nicolas Pitre wrote: > > Yes, and rename it to ucb1200-generic.c, or better yet sa1100-ucb1200.c > for more clarity. This could become ucb1200-xyz.c if ever some > other architecture xyz uses the UCB1200 chip. Lets use the generic name. The LinkUp L7200 & L7205 uses the UCB1200 (ARM720) > > The slave drivers could be ucb1200-ts for touchscreen, ucb1200-audio for > audio, etc. > > > And before anyone writes or modifies anything, should we > > have a who's-doing-what list in order to minimize duplicated > > efforts? > > I guess volunteers might identify themselves on this list. > You are one? :-) I am interested in the TS at this point. Scott unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Fri Aug 25 21:56:55 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13SQWd-0001y0-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Fri, 25 Aug 2000 21:56:31 +0100 Received: from [208.242.241.39] (helo=cvsftp.cotw.com ident=root) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13SQWc-0001xq-00 for linux-arm-kernel@lists.arm.linux.org.uk; Fri, 25 Aug 2000 21:56:31 +0100 Received: from cotw.com (IDENT:samcconn@ptecdev2.ptec.com [192.168.10.4]) by cvsftp.cotw.com (8.9.3/8.9.3) with ESMTP id PAA24436; Fri, 25 Aug 2000 15:56:10 -0500 Message-ID: <39A6DE17.E3DFB87A@cotw.com> Date: Fri, 25 Aug 2000 15:59:03 -0500 From: S A McConnell Reply-To: samcconn@cotw.com X-Mailer: Mozilla 4.7 [en] (X11; I; Linux 2.2.14 i686) X-Accept-Language: en MIME-Version: 1.0 To: Tak-Shing Chan CC: linux-arm-kernel@lists.arm.linux.org.uk Subject: Re: test6-rmk5-np1 PCMCIA / UCB1200 audio? References: Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk Tak-Shing Chan wrote: > > > > I have questions about how the touch screen should detect events. > > (interrupts vs timer based polling) I am currently using an Interrupt > > to detect touch events. > Interrupts can detect pen downs only; you need a timer for > continuous sampling during the pen down period or else you > can't drag the ``mouse''. > (I am a novice at this...) Using thresholds, pen ups can be detected by interrupts. With the hacked up code I have been playing with the last interrupt is always x=-1, y=-1 (Pen Up) I do not believe it is the job of the driver to detect "drags". Rather it is the job of the driver to provide enough information to allow userland app's to detect drags. If there is data available the pen is down. If there is no data available the pen is up. If the current position is more than z units different from the last measurement it moved. X, Y with timestamp will allow momentum and angle to be calculated. (I do not believe that is the job of the driver) The question I have is: will an "interrupt" based approach generate events at a rate greater than they can be sampled by user land app's? Is this a problem? How has this problem been solved in the past? (keyboards?) This is where the timer based approach could help resolve this problem. Any other thoughts. Scott unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Sat Aug 26 18:07:56 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13SjNv-000441-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Sat, 26 Aug 2000 18:04:47 +0100 Received: from [195.92.244.33] (helo=caramon.arm.linux.org.uk) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13SjNn-00043m-00 for linux-arm-kernel@lists.arm.linux.org.uk; Sat, 26 Aug 2000 18:04:39 +0100 Received: from flint.arm.linux.org.uk (root@flint [192.168.0.4]) by caramon.arm.linux.org.uk (8.9.3/8.9.3) with ESMTP id SAA01273 for ; Sat, 26 Aug 2000 18:04:43 +0100 Received: (from linux@localhost) by flint.arm.linux.org.uk (8.9.3/8.9.3) id SAA04379 for linux-arm-kernel@lists.arm.linux.org.uk; Sat, 26 Aug 2000 18:03:56 +0100 From: Russell King - ARM Linux Admin Message-Id: <200008261703.SAA04379@flint.arm.linux.org.uk> Subject: Kernel 2.4.0-test6-rmk6 released To: linux-arm-kernel@lists.arm.linux.org.uk Date: Sat, 26 Aug 2000 18:03:55 +0100 (BST) X-Mailer: ELM [version 2.5 PL1] MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk Hi, Linux Kernel version 2.4.0-test6-rmk6 is now on the ARM Linux FTP site, which can be found at: ftp://ftp.arm.linux.org.uk/pub/armlinux/source/kernel-patches/v2.4/patch-2.4.0-test6-rmk6.gz Pre-requsists: -------------- To use this patch, you will need to obtain the following: * Linus' 2.4.0-test6 kernel source ftp://ftp.xx.kernel.org/pub/linux/kernel/v2.4/ Replace 'xx' with your two-character country code. Things to watch out for: ------------------------ * There have been reports on the main linux-kernel mailing list about possible minor corruption of ext2fs filesystems. The symptoms are that fsck complains that the counts are wrong. Although these appear to have been one-offs, if you do experience this, please let the mailing lists know. Changelog: ---------- 2.4.0-test6-rmk6 * Create arch/arm/mach-sa1100 and friends, move some [me] of the architecture stuff into there * Pass EXTRAVERSION into any install.sh script [me] * Move various header files into include/asm-arm/mach [me] * Kill processor flushing exports (I don't believe they [me] are used) - please let me know ASAP if this causes you a problem * Make DEC21142 sleep a "fixup" [me] * Drop out the machine-type PCI stuff into relevent [me] arch/arm/mach-* directory * Make DMA and Acorn-related SCSI stuff use PCI cache [me] coherency macros instead of going directly for CPU stuff * Fix problem in ptrace implementation [me] * Clean up various make files to use new style [me] * Most of the CPU functions got renamed for clarity [me] * Reserve special DMA memory on A5000/Archimedes boot [me] * Don't map the flash in on footbridge-based platforms [me] * Fix section names in vmlinux-armo.lds.in [me] * Fix ether1_writebuffer/ether1_readbuffer [me] * Make nwflash ioremap the flash (untested) [me] * Update ICSIDE wrt DMA changes [me] * Fix acornfb screen size on old machines [me] * Fix old machine locks [me] * Kill dma_cache_* functions [me] * Patch 393/1 * Patch 395/1 2.4.0-test6-rmk5 * Includes bootp.lds! (and a fix to the makefile to stop [me] it being deleted) * Fix mach-types.h generation so its present before we [me] build anything * Change ARM stuff to use __virt_to_bus instead of the [me] virt_to_bus * Update footbridge mm initialisation [me] * Update sa1100 mm initialisation [me] * Clean up drivers/acorn/char/Makefile [me] * Fix alim15x3 ide driver [Shaun Savage] * Use higher-resolution palette in cyber2000fb [Pete Zaitcev] 2.4.0-test6-rmk4 * Fix memzero implementation [me/dwmw2/prumpf] * Split pagetable_init functionality into memory table [me] initialisation and io table initialisations. * Allocate resources for prefetchable and [me] non-prefetchable PCI regions * Fix A5000 floppy DMA [me] * Various task struct optimisations [me] * Various string function optimisations [me] * USB 'input.o' fix [me] * Make cyberpro option depend on CONFIG_PCI [me] 2.4.0-test6-rmk3 * Split out machine types from asm-arm/system.h [me] (see arch/arm/tools/mach-types for the list) * Brought semaphores inline with x86 [me/Matthew Wilcox] * Fix Oops on RiscPC with Sysrq-M [me] (it will still report wrong values though) * Clean up initrd freeing to use generic free region code [me] * Kill PageSkip/PG_skip [me] * Make USB use initcalls [me] * NetFilter fixes [Paul `Rusty' Russell] * Patch 391/1 * Patch 388/1 * Patch 389/1 * Patch 390/1 2.4.0-test6-rmk2 * PCI resource setup fix [Linus Torvalds] * Fix pci_update_resource to complain on failure and [me] to handle ROM cases * Fix compressed bootup infinite loop on NetWinder [me] * Fix Netwinder IDE [me] 2.4.0-test6-rmk1 * arch/arm/special has been sent to the great directory [me] pool in the sky. * Added 'make bootpImage' to create a zImage+Initrd [me] (Only RiscPC and EBSA285, not tested yet) * move #ifdef architecture to head-armo.S [me] * Fix IDE such that it detects W83C553 again correctly [me] * Patch 380/1 * Patch 383/1 * Patch 384/1 2.4.0-test5-rmk3 * More cleaning up of assembler rules [me/Michael Elizabeth Chastain] * Issue warning when obsolete code is used [me] * Fix warning in dma-rpc.c [me] * Fix occasional oops on reboot on RiscPC [me] * Make RED led turn on when halted on NetWinder [me] * Fix FAS216 reset timeout to be 250ms [me] * Fix opening of busmice [me] * Make cyber2000fb available to everyone [me] * Wrap fs/partitions/acorn.c sensibly and clean it up [me] * Fix architecture incompatabilities in [me/Chris Rutter] fs/partitions/acorn.h * Add Pleb machine type macros [me] * Fix fbcon "insert char" bug [me] * Patch 377/1 * Patch 379/1 2.4.0-test5-rmk2 * More moving of assembler rules [me] * Pull small bit of Angel boot code into head.S and [me] reallocate the registers in the architecture code * Kill local_bh_count and local_irq_count in favour of [me] irq_stat * More ARM2/ARM3 fixes [Dave Gilbert] * Optimise ret_from_syscall [me/Phil Blundell] * Make various strings/arrays const [Phil Blundell] * Make arch/arm/kernel/signal.c more efficient [Phil Blundell] * Update SA1100 IrDA driver [me] * Make IrDA discovery random time slot actually appear [me] to be random * Cleanup IrDA initialisation [me] * Make the irias "Device" "DeviceName" object contain the [me] correct value after a sysctl write to /proc/sys/net/irda/devname 2.4.0-test5-rmk1 * Reduced inclusion of files from include/linux back into [me] include/asm-arm * Added definitions for clf() and stf() [me/Phil Blundell] * Fixed resource limits [Phil Blundell] * Fix get_user and friends to return -EFAULT not -ENOTBLK [me/Phil Blundell] * Old ARM2/ARM3 semaphore asm [Dave Gilbert] * Remove redundant get_sp() [me/Phil Blundell] * Various L7200 updates [Steve Hill] * Various CL-PS7500FE updates [Jason Gunthorpe] * Fix EBSA285 PAGE_OFFSET/PHYS_OFFSET [me] * Couple of "dirty" USB fixes [me] * CS8900 driver [Jason Gunthorpe] * Fix Promise PDC202xx driver [Andre Hedrick] * Remove 50baud to 150baud that the 21285 can't handle [me] * Fix ASM makefile rules [me] * Optimise user string functions [Phil Blundell] * Fix no_dma warning [me] 2.4.0-test4-rmk1 * Added 'ConfigVars' to document EBSA285-based config [me] definitions * Combined A5000 and Archimedes ports more [me] * Various config.in cleanups [me] * Cleanup DMA support [me] * Allow larger ram sizes on NetWinder [Ralph Siemsen] * Correct EBSA285 IRQ mappings [me] * Make sure BUG() isn't recursive [me] * Add half-word data abort support to ARM720 [Cliff Brake] * Support baud rates above 200baud in 21285 driver [me] * Fix return types in SA1100 framebuffer driver [me] * Fix Winbond 83977 IrDA driver [me] * Fix compiler error in arch-ebsa285/irq.h [me/Bill Arbaugh] * Fix missing definition of have_isa_bridge [me/Bill Arbaugh] 2.4.0-test3-pre8-rmk1 * Cleanup dma support [me] 2.4.0-test3-pre7-rmk1 * Cleanup make of arch/arm/lib [me] * Fix inclusion of linux/config.h [me] * Make oldlatches.c compatable with A5000 [me] * Made fd1772.c initialisation conditional on [me] machine_is_arc() * Made serial_21285.c properly modular [me] * Cyber2000 DMPS support [Ralph Siemsen] * Added architecture number for Itsy [me] * unaligned.h now handles 64-bit quantities [me] * Allow unix domain sockets to pass up device ioctls [me] 2.4.0-test2-ac1-rmk3 * More Configure.help entries [George France] * Temporarily disable omission of frame pointer during [me] debug * SA1100 stuff [Nico Pitre] * Add basic elements of PM support [me] * Move SA1100 architecture Victor power off routine [me] to arch.c * Fix oops caused by kecardd when debian runs [me] start-stop-daemon * Add "keepinitrd" option to preserve the initrd memory [me] while kernel is running. * Fix bug in misc.c preventing misc devices being opened [me] * Fix warnings in nwbutton.c [me] * Add initial SA1100 FIR irda driver (SIR only atm) [me] * Stop NetWinder IR driver running >115200baud [me] (needs investigating why it doesn't work) * Kill arch_power_off() [me] * Fix warnings in keyboard.c caused by SA1100 keyboard [me] header file. * Fix various warnings in IrDA code [me] * Re-architect skb handling in IrLAP code [me] * Temporary structure size and alignment fixes (will be [me] removed in next version) * Fix IrDA parameters infinite loop [me] 2.4.0-test2-ac1-rmk2 * SA1100 LART Configure.help entry [J.D. Bakker] * elevator (block device queue) fixes [Jens Axboe] * read() functionality restored to Cyberpro grabber [me] * Kill unnecessary MOD_*_USE_COUNT in videodev [me] 2.4.0-test2-ac1-rmk1 * Assabet and SA1100 framebuffer Configure.help entries [Tak-Shing Chan] added * Make CLPS7500 v3 architecture [Phil Blundell] * Kill redundant check in binfmt_aout.c [me] 2.4.0-test1-ac22-rmk2 * Added various Configure.help entries [Jamey Hicks] * Change "Bitsy" to "Compaq iPAQ 3600" [Jamey Hicks] * Cleaned up include files in arch/arm/{kernel,mm} [me] * Removed reporting of expansion cards in boot messages [me] (look under /proc/bus/ecard) * Update ptrace [me/Phil Blundell] * Make old 21285 major/minor compatability a config [me] option * Revert to "traditional" way of IDE interrupt handling [me] * Fix SA1100 FB driver prototypes [me] * Fix SA1100 IRQ header file [me] * Fix Shark __ASSEMBLER__ macro [me] * Kill ARM2/3 xchg for shorts [me] 2.4.0-test1-ac22-rmk1 * Cleaned up xchg_str/__bad_xchg [me] * Cleanup asm/system.h, asm/dma.h, asm/irq.h includes [me] * Removed debugging strings in hw-footbridge when not [me] debugging * Removed memparse routine from ARM specific files [me] * Fixed leak in SIGSEGV siginfo [me] * Make it possible for baddataabort to continue [me] * Move defkeymap-acorn.c to drivers/acorn/char [me] * Move console.c to drivers/char [me] * Back out some changes in the floppy driver [me] * Make NFSD cache entries time out after 2 minutes [me] (fixes problems with NFS serving root-NFS booting machines without clocks) * Pull out virt_to_* and friends into asm/memory.h [me] (this file should not be included by non-asm/*.h files) * Backed out change to bootmem.c [me] 2.4.0-test1-ac19-rmk1 * Register optimised ARM720 processor support [me] * am79c961 received byte accounting [me] * More configuration cleanups [me] * Use actual footbridge mode value rather than [me] configuration-option derived version * Always display CPU type [Nicolas Pitre] * SA1100 updates [Nicolas Pitre] 2.4.0-test1-ac14-rmk1 * Don't scan the PCI bus if we're not in central [me] function mode * Remove lock/unlock_8390_module from etherh [me] * Revert kernel/sched.c del_timer_sync() change [me] 2.4.0-test1-ac7-rmk1 * Removed redundant adr in head.S [me] * Move isapnp_disable to arch.c for NetWinder [me] * Move DC21285 ISA port registration into dec21285.c [me] * Don't disable floppy on SA1100 architectures [me] * Added some SA1100 bits [me] * Don't program sl82c105 hardware for PIO when we're [me] using DMA * Make sure bdflush and kupdate are running (clean+ [me] reliable method) * Make SA1100's serial port for decompression messages [me] run-time selectable by architecture number 2.4.0-test1-ac6-rmk1 * Fix bug in EBSA110 driver causing tx queue to be put [me] to sleep * Kill some msr spsr_c, #value [me] * Fix portability issues in netfilter code [me] * Fix no-ide-dma problem in sl82c105.c [me] * Make sl82c105 use 32bit IO and unmask irqs in PIO mode [me] * CLPS7500 bits [Jason Gunthorpe] * Make sure EBSA110 has Timer LED when LEDs configured [me] * serial_21285.c initialisation cleanup [me] * Added missing machine_is_graphicsclient() [me] _____ |_____| ------------------------------------------------- ---+---+- | | Russell King linux@arm.linux.org.uk --- --- | | | | http://www.arm.linux.org.uk/~rmk/armlinux.html / / | | +-+-+ --- -+- / | THE developer of ARM Linux |+| /|\ / | | | --- | +-+-+ ------------------------------------------------- /\\\ | unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Sun Aug 27 02:28:47 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13SrDB-0000VH-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Sun, 27 Aug 2000 02:26:13 +0100 Received: from [63.195.182.101] (helo=barbados.bluemug.com) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13SrDA-0000VA-00 for linux-arm-kernel@lists.arm.linux.org.uk; Sun, 27 Aug 2000 02:26:12 +0100 Received: from fiji.bluemug.com [192.168.50.32] by barbados.bluemug.com with esmtp (Exim 2.05 #1 (Debian)) id 13SrDG-0007IR-00; Sat, 26 Aug 2000 18:26:18 -0700 Received: from miket by fiji.bluemug.com with local (Exim 3.12 #1 (Debian)) id 13SrDG-0005Ac-00; Sat, 26 Aug 2000 18:26:18 -0700 Date: Sat, 26 Aug 2000 18:26:18 -0700 From: Mike Touloumtzis To: linux-arm-kernel@lists.arm.linux.org.uk Subject: need help booting EDB7211 eval board under 2.4.0-test* Message-ID: <20000826182618.A15112@fiji.bluemug.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk Hi, I'm trying to get a Cirrus Logic EDB7211 evaluation board booting the 2.4.0 kernels. I have the basic architecture files and a serial driver written/ported, everything compiles and links, and now I'm trying to get it booting to userspace. My current kernel is patched against 2.4.0-test3-pre7-rmk1, so it's a little out of date. The EDB7211 has non-contiguous physical memory, so like the SA-1100 I'm using the NUMA support (two nodes, each with one bank, and using a similar bit-shifting trick with __virt_to_phys and __phys_to_virt). #define __virt_to_phys(x) (((x) & 0xff7fffff) | ((x) & 0x00800000) << 1) #define __phys_to_virt(x) (((x) & 0xfeffffff) | ((x) & 0x01000000) >> 1) DRAM part of the memory map is: Physical Virtual Size NUMA Node -------- ------- ---- --------- 0xc0000000 0xc0000000 8 MB 0 0xc1000000 0xc0800000 8 MB 1 (the amount of DRAM is fixed, so I chose to make DRAM virtually contiguous). A problem arises when arch/arm/mm/init.c:paging_init() tries to set up the second node. The boot process hangs (the kernel's probably caught in some trap handler somewhere; I haven't figured out where) when it tries to set up the mem_map array for the second node. The stack looks like this: alloc_bootmem_core() mm/bootmem.c alloc_bootmem_node() mm/bootmem.c free_area_init_core() mm/page_alloc.c free_area_init_node() mm/numa.c paging_init() arch/arm/mm/init.c ... more stuff ... When I investigated, it turns out that free_area_init_core() wants to allocate each node's mem_map within the node itself, so it passes the nid to alloc_bootmem_node(). The hang occurs when alloc_bootmem_core() tries to zero the newly allocated bootmem area; access to the 0xc0800000 region (second bank) doesn't work. Upon further checking, I couldn't access 0xc0400000 (halfway through the 1st bank) either. I added serial output to pagetable_init in mm-armv.c (now called memtable_init) and it's handling both nodes as expected; physical and virtual addresses all check out. It's like the new page tables don't 'take', and I'm still going through the boot-time 4MB page directory. My questions are: -- Is there some point after [page|mem]table_init at which the values are actually 'kicked in', or do they take effect right away? -- Does pagetable_init modify the swapper_pgdir directory, and is the swapper_pgdir the main pgdir for kernel access from task 0? -- Does anyone know what the heck is going on :-)? Also, what is the intended purpose of FLUSH_BASE and FLUSH_BASE_PHYS, and how should they be chosen? I'm a little unclear on what they are used for, and they're suspicously proximate to the failure point. I can boot beyond this point (to other memory related problems outside the scope of this mail :-) if I remove the second node (leaving 8MB of contiguous memory) from arch/arm/kernel/arch.c. I'm just learning both the ARM architecture and the guts of the kernel, so I thought I'd ask to see if anyone here had insight into the problem. miket unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Sun Aug 27 08:36:47 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13Swyt-0003Zf-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Sun, 27 Aug 2000 08:35:51 +0100 Received: from [194.134.0.10] (helo=pop1.euronet.nl) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13Swys-0003ZZ-00 for linux-arm-kernel@lists.arm.linux.org.uk; Sun, 27 Aug 2000 08:35:50 +0100 Received: from angelo.home.nl (i0184.vwr.euronet.nl [194.134.208.185]) by pop1.euronet.nl (Postfix) with SMTP id AE7AE4F6D7; Sun, 27 Aug 2000 09:35:47 +0200 (MET DST) Date: Sun, 27 Aug 2000 09:32:49 +0100 (BST) From: Angelo Subject: switched to strongARM and 160 Mb and Debian 2.2 doesn't boot anymore. Why ? To: debian-arm@lists.debian.org Cc: linux-arm-kernel@lists.arm.linux.org.uk Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; CHARSET=US-ASCII X-Organization: Organisation name, location. Telephone/Fax? X-Mailer: ANT RISCOS Marcel [ver 1.53] Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk Hi, I have debian 2.2 running for a few weeks on my RISC PC 700, with an ARM 7 and 64 Mb (+2 Mb VRAM). This worked rather well (did some recompiling and stuff). Yesterday I recieved my memoryupgrade and StrongARM and I installed it. The memory went up from 64 (2x32Mb) to 160 (128 + 32Mb) and I put the StrongARM in. But when I tried to reboot, debian hang at the 'Starting INIT 2.78' line. Could someone please tell me why ? Is there something which I can do to make it work again ? It should work on StrongARM, doesn't it ? Or is it a memory problem ? I am running the kernel which is included in the distribution (the 2.2.13, which worked on the ARM 7). Thanks in advance, Angelo -- Angelo Melis @ home random quote, generated by my computer: If practice makes perfect, and nobody's perfect, why practice. unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Sun Aug 27 08:54:48 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13SxG3-0003iQ-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Sun, 27 Aug 2000 08:53:35 +0100 Received: from [211.44.193.250] (helo=www.palmpalm.com) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13SxG2-0003iI-00 for linux-arm-kernel@lists.arm.linux.org.uk; Sun, 27 Aug 2000 08:53:34 +0100 Received: from alice ([192.168.1.13]) by www.palmpalm.com (8.9.3/8.8.7) with SMTP id RAA09865; Sun, 27 Aug 2000 17:04:39 +0900 Message-ID: <001a01c00ffc$5ef40700$0d01a8c0@palmpalm.com> From: "Hyun-Mo Chung" To: "Russell King - ARM Linux Admin" , References: <200008261703.SAA04379@flint.arm.linux.org.uk> Subject: Re: Kernel 2.4.0-test6-rmk6 released Date: Sun, 27 Aug 2000 16:56:56 +0900 MIME-Version: 1.0 Content-Type: text/plain; charset="Windows-1252" Content-Transfer-Encoding: 7bit X-Priority: 3 X-MSMail-Priority: Normal X-Mailer: Microsoft Outlook Express 5.00.2919.6600 X-MimeOLE: Produced By Microsoft MimeOLE V5.00.2919.6600 Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk Hi, It looks like arch/arm/mm/proc-arm720.S has a typo. The following two symbols cause linker fail : cpu_arm720_invalidate_icache_range cpu_arm720_invalidate_icache_page The linker can get happy by changing them into cpu_arm720_icache_invalidate_range cpu_arm720_icache_invalidate_page Please let me know if I'm getting into the correct path. Thank you. -Hyunmo unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Sun Aug 27 22:01:26 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13T9WD-0001zN-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Sun, 27 Aug 2000 21:59:05 +0100 Received: from [158.152.220.239] (helo=tazenda.demon.co.uk ident=mail) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13T9WB-0001zH-00 for linux-arm-kernel@lists.arm.linux.org.uk; Sun, 27 Aug 2000 21:59:04 +0100 Received: from kings-cross.london.uk.eu.org [::ffff:192.168.2.83] by tazenda.demon.co.uk with esmtp (Exim 3.11 #1 (Debian)) id 13T9f8-0000dH-00; Sun, 27 Aug 2000 22:08:18 +0100 Received: from localhost ([::ffff:127.0.0.1] helo=tazenda.demon.co.uk ident=pb) by kings-cross.london.uk.eu.org with esmtp (Exim 3.12 #1 (Debian)) id 13T9Yt-0002E6-00 for ; Sun, 27 Aug 2000 22:01:51 +0100 X-Mailer: exmh version 2.1.1 10/15/1999 (debian) To: linux-arm-kernel@lists.arm.linux.org.uk Subject: Re: kernel crash with munmap In-Reply-To: Message from Philip Blundell of "Tue, 25 Jul 2000 19:30:17 BST." Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Date: Sun, 27 Aug 2000 22:01:50 +0100 From: Philip Blundell Message-Id: Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk The bug below still seems to be present in 2.4.0-test6. Is anybody else able to reproduce the problem? p. >When I run the program below on my CATS under 2.4.0-test2, the machine locks >up. It seems to be OK on my i386 with 2.4.0-test5. Any ideas what might be >wrong? > >p. > >-- > >#include >#include >#include > >#define NULL 0 > >int main(int argc, char *argv) >{ > char *s = tmpnam(NULL); > int fd = open(s, O_CREAT | O_RDWR); > int fd2 = open("/bin/ls", O_RDONLY); > ftruncate(fd, 640*1024); > unlink(s); > mmap(0x8000, 640*1024, PROT_READ | PROT_WRITE | PROT_EXEC, MAP_PRIVATE | MAP >_FIXED, fd, 0); > read(fd2, 0x8000, 640*1024); > munmap(0x8000, 640*1024); > exit(0); >} > > unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Mon Aug 28 00:39:53 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13TC1C-0003N8-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Mon, 28 Aug 2000 00:39:14 +0100 Received: from dyn-33.linux.theplanet.co.uk ([195.92.244.33] helo=caramon.arm.linux.org.uk) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13TC1A-0003N2-00; Mon, 28 Aug 2000 00:39:13 +0100 Received: from flint.arm.linux.org.uk (root@flint [192.168.0.4]) by caramon.arm.linux.org.uk (8.9.3/8.9.3) with ESMTP id AAA06689; Mon, 28 Aug 2000 00:39:19 +0100 Received: (from linux@localhost) by flint.arm.linux.org.uk (8.9.3/8.9.3) id AAA15049; Mon, 28 Aug 2000 00:38:24 +0100 From: Russell King - ARM Linux Admin Message-Id: <200008272338.AAA15049@flint.arm.linux.org.uk> Subject: Re: Kernel 2.4.0-test6-rmk6 released To: hmjung@palmpalm.co.kr (Hyun-Mo Chung) Date: Mon, 28 Aug 2000 00:38:24 +0100 (BST) Cc: linux-arm-kernel@lists.arm.linux.org.uk In-Reply-To: <001a01c00ffc$5ef40700$0d01a8c0@palmpalm.com> from "Hyun-Mo Chung" at Aug 27, 2000 04:56:56 PM X-Mailer: ELM [version 2.5 PL1] MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk Hyun-Mo Chung writes: > It looks like arch/arm/mm/proc-arm720.S has a typo. > The following two symbols cause linker fail : > cpu_arm720_invalidate_icache_range > cpu_arm720_invalidate_icache_page > The linker can get happy by changing them into > cpu_arm720_icache_invalidate_range > cpu_arm720_icache_invalidate_page > Please let me know if I'm getting into the correct path. > Thank you. Yes, its correct. _____ |_____| ------------------------------------------------- ---+---+- | | Russell King linux@arm.linux.org.uk --- --- | | | | http://www.arm.linux.org.uk/ / / | | +-+-+ --- -+- / | THE developer of ARM Linux |+| /|\ / | | | --- | +-+-+ ------------------------------------------------- /\\\ | unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Mon Aug 28 02:41:52 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13TDtz-0004Ox-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Mon, 28 Aug 2000 02:39:55 +0100 Received: from [210.184.5.125] (helo=pikachu.idthk.com) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13TDtx-0004Oc-00 for linux-arm-kernel@lists.arm.linux.org.uk; Mon, 28 Aug 2000 02:39:54 +0100 Received: from localhost (thomas@localhost) by pikachu.idthk.com (8.9.3/8.8.7) with ESMTP id JAA31181; Mon, 28 Aug 2000 09:39:42 +0800 Date: Mon, 28 Aug 2000 09:39:42 +0800 (CST) From: Tak-Shing Chan To: Nicolas Pitre cc: linux-arm-kernel@lists.arm.linux.org.uk, sa1100-linux@pa.dec.com Subject: Re: test6-rmk5-np1 PCMCIA / UCB1200 audio? In-Reply-To: Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk On Fri, 25 Aug 2000, Nicolas Pitre wrote: > I guess volunteers might identify themselves on this list. > You are one? :-) Sure. I plan to port the latest Itsy drivers (Y2K version) from 2.0.30 to 2.4.0-test6-rmk5-np1 (quite a tedious task but hopefully I can start from Erik Mouw's port of an older version), and rename them to the ones you proposed... I'll also make their TS driver iPAQ-compatible. This seems to solve all the problems at once. Is there anyone else already doing this? Tak-Shing Chan unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Mon Aug 28 03:44:34 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13TEu2-0004ww-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Mon, 28 Aug 2000 03:44:02 +0100 Received: from [210.184.5.125] (helo=pikachu.idthk.com) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13TEtx-0004wb-00 for linux-arm-kernel@lists.arm.linux.org.uk; Mon, 28 Aug 2000 03:44:00 +0100 Received: from localhost (thomas@localhost) by pikachu.idthk.com (8.9.3/8.8.7) with ESMTP id KAA31344; Mon, 28 Aug 2000 10:43:48 +0800 Date: Mon, 28 Aug 2000 10:43:48 +0800 (CST) From: Tak-Shing Chan To: S A McConnell cc: linux-arm-kernel@lists.arm.linux.org.uk Subject: Re: test6-rmk5-np1 PCMCIA / UCB1200 audio? In-Reply-To: <39A6DE17.E3DFB87A@cotw.com> Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk On Fri, 25 Aug 2000, S A McConnell wrote: > Using thresholds, pen ups can be detected by interrupts. With the hacked > up code I have been playing with the last interrupt is always x=-1, y=-1 > (Pen Up) You must be talking about the polling driver from Century Software. The interrupt-based driver in test6-rmk5-np1 has no thresholds nor the x=-1, y=-1 thing. > I do not believe it is the job of the driver to detect "drags". Rather > it is the job of the driver to provide enough information to allow > userland app's to detect drags. I didn't say drivers detect drags. What I meant is that with a pure interrupt-based driver there will not be ``enough information to allow userland app's to detect drags''. Let me clarify how the touchscreen interrupt actually works. As per the UCB1300 specification [1], ``a touch screen interrupt is generated either when the touch screen is pressed (falling edge enabled) or when the touch screen is released (rising edge enabled). It can be used to activate the system around the UCB1300 to start a touch screen read-out sequence.'' The above paragraph is the most important part of the touchscreen specification. It says that pen movements during pen down will NOT be detected by the interrupts. Therefore you do need a timer for the continuous sampling during pen down, and ideally the rate should be software-selectable. Tak-Shing Chan [1] Philips Semiconductor. UCB1300 Product Specfication (2000 Jun 26), p. 19, available on the web at: http://www.semiconductors.com/acrobat/datasheets/UCB1300_2.pdf unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Mon Aug 28 09:32:15 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13TKJp-0007z1-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Mon, 28 Aug 2000 09:31:01 +0100 Received: from [195.64.48.18] (helo=imail.knoware.nl) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13TKJo-0007yv-00 for linux-arm-kernel@lists.arm.linux.org.uk; Mon, 28 Aug 2000 09:31:00 +0100 Received: from mail.knoware.nl (mail.knoware.nl [195.64.48.17]) by imail.knoware.nl (Postfix) with ESMTP id AC965BD72A; Mon, 28 Aug 2000 10:31:09 +0200 (CEST) Received: from avxsrv04.cypres.nl (cypr-45-228.cypres.nl [195.64.45.228]) by mail.knoware.nl (Postfix) with ESMTP id 998B9A6CE6; Mon, 28 Aug 2000 10:31:08 +0200 (CEST) Received: by avxsrv04.cypres.nl with Internet Mail Service (5.5.2650.21) id ; Mon, 28 Aug 2000 10:32:48 +0200 Message-ID: <8BD00E18F88DD3118EC8005004258990DCEBF4@avxsrv04.cypres.nl> From: "Melis, A. (Angelo)" To: "'ccarey@intrinsyc.com'" Cc: "'linux-arm-kernel@lists.arm.linux.org.uk'" , "'debian-arm@lists.debian.org'" Subject: FW: switched to strongARM and 160 Mb and Debian 2.2 doesn't boot anymore. Why ? Date: Mon, 28 Aug 2000 10:32:44 +0200 MIME-Version: 1.0 X-Mailer: Internet Mail Service (5.5.2650.21) Content-Type: text/plain; charset="iso-8859-1" Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk Hi, Sorry if this message arrives twice, but I got a message that sending failed and I don't know if it did get throu, so I resend it again. -----Oorspronkelijk bericht----- Hi Chester, (and everyone on the list) > > Did you set the architecture to StrongArm? Eh ? No. How do I do that ? I did get it to boot btw, by swapping the 128MB and 32 MB SIMM (now the 128MB SIMM is in bank 0). But I do get an error with e2fsck runs, that symbol GLIB2.0 not found with runtime linking. Does anyone know what that means ? (I did not yet install all packages, so I could imagine that I miss a package.) Bye, Angelo > > Quoting Angelo : > > > Hi, > > > > I have debian 2.2 running for a few weeks on my RISC > > PC 700, with an ARM 7 > > and 64 Mb (+2 Mb VRAM). This worked rather well (did > > some recompiling and > > stuff). > > > > Yesterday I recieved my memoryupgrade and StrongARM > > and I installed it. The > > memory went up from 64 (2x32Mb) to 160 (128 + 32Mb) > > and I put the StrongARM > > in. But when I tried to reboot, debian hang at the > > 'Starting INIT 2.78' line. > > Could someone please tell me why ? Is there something > > which I can do to make > > it work again ? It should work on StrongARM, doesn't > > it ? Or is it a memory > > problem ? > > > > I am running the kernel which is included in the > > distribution (the 2.2.13, > > which worked on the ARM 7). > > > > Thanks in advance, > > Angelo > > > > -- > > Angelo Melis @ home > > random quote, generated by my computer: > > If practice makes perfect, and nobody's perfect, why > > practice. > > > > > > > > unsubscribe: body of `unsubscribe linux-arm-kernel' to > > majordomo@lists.arm.linux.org.uk > > > unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Mon Aug 28 10:58:12 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13TLfk-0000I1-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Mon, 28 Aug 2000 10:57:44 +0100 Received: from [158.152.220.239] (helo=tazenda.demon.co.uk ident=mail) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13TLfi-0000Hr-00 for linux-arm-kernel@lists.arm.linux.org.uk; Mon, 28 Aug 2000 10:57:43 +0100 Received: from kings-cross.london.uk.eu.org [::ffff:192.168.2.83] by tazenda.demon.co.uk with esmtp (Exim 3.11 #1 (Debian)) id 13TLBZ-0000yh-00; Mon, 28 Aug 2000 10:26:33 +0100 Received: from localhost ([::ffff:127.0.0.1] helo=tazenda.demon.co.uk ident=pb) by kings-cross.london.uk.eu.org with esmtp (Exim 3.12 #1 (Debian)) id 13TL5A-0003wL-00; Mon, 28 Aug 2000 10:19:56 +0100 X-Mailer: exmh version 2.1.1 10/15/1999 (debian) To: "Melis, A. (Angelo)" cc: "'ccarey@intrinsyc.com'" , "'linux-arm-kernel@lists.arm.linux.org.uk'" , "'debian-arm@lists.debian.org'" Subject: Re: FW: switched to strongARM and 160 Mb and Debian 2.2 doesn't boot anymore. Why ? In-Reply-To: Message from "Melis, A. (Angelo)" of "Mon, 28 Aug 2000 10:32:44 +0200." <8BD00E18F88DD3118EC8005004258990DCEBF4@avxsrv04.cypres.nl> References: <8BD00E18F88DD3118EC8005004258990DCEBF4@avxsrv04.cypres.nl> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Date: Mon, 28 Aug 2000 10:19:56 +0100 From: Philip Blundell Message-Id: Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk >and 32 MB SIMM (now the 128MB SIMM is in bank 0). But I do get an error with >e2fsck runs, that symbol GLIB2.0 not found with runtime linking. Does anyone >know what that means ? (I did not yet install all packages, so I could >imagine that I miss a package.) It's not related to your hardware changes -- something is wrong with the version of glibc that you have. What's the exact message? p. unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Mon Aug 28 14:37:40 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13TP68-0002RW-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Mon, 28 Aug 2000 14:37:12 +0100 Received: from [207.97.93.2] (helo=blazie.com) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13TP67-0002RQ-00 for linux-arm-kernel@lists.arm.linux.org.uk; Mon, 28 Aug 2000 14:37:11 +0100 Received: from localhost (localhost [127.0.0.1]) by blazie.com (8.9.2/8.8.7) with ESMTP id JAA21311; Mon, 28 Aug 2000 09:37:06 -0400 (EDT) Date: Mon, 28 Aug 2000 09:37:06 -0400 (EDT) From: Chris Blazie To: Angelo cc: linux-arm-kernel@lists.arm.linux.org.uk Subject: Re: switched to strongARM and 160 Mb and Debian 2.2 doesn't boot anymore. Why ? In-Reply-To: Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk > Yesterday I recieved my memoryupgrade and StrongARM and I installed it. The > memory went up from 64 (2x32Mb) to 160 (128 + 32Mb) and I put the StrongARM > in. But when I tried to reboot, debian hang at the 'Starting INIT 2.78' line. > Could someone please tell me why ? Is there something which I can do to make As far as I know, memory sizes greater than 32MB are not supported in the Strongarm kernel right now. You can verify this by modifying arch/arm/arch.c and arch/arm/setup.c to declare only 32MB and see if it works. Our hardware works when "detuned" like that; at some point I'm planning to take a look at the problem if it's still there. Regards, Chris unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Mon Aug 28 14:42:03 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13TPAa-0002Ug-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Mon, 28 Aug 2000 14:41:48 +0100 Received: from dyn-33.linux.theplanet.co.uk ([195.92.244.33] helo=caramon.arm.linux.org.uk) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13TPAY-0002Ua-00; Mon, 28 Aug 2000 14:41:47 +0100 Received: from flint.arm.linux.org.uk (root@flint [192.168.0.4]) by caramon.arm.linux.org.uk (8.9.3/8.9.3) with ESMTP id OAA09270; Mon, 28 Aug 2000 14:41:53 +0100 Received: (from linux@localhost) by flint.arm.linux.org.uk (8.9.3/8.9.3) id OAA17412; Mon, 28 Aug 2000 14:40:56 +0100 From: Russell King - ARM Linux Admin Message-Id: <200008281340.OAA17412@flint.arm.linux.org.uk> Subject: Re: switched to strongARM and 160 Mb and Debian 2.2 doesn't boot To: chris@blazie.com (Chris Blazie) Date: Mon, 28 Aug 2000 14:40:56 +0100 (BST) Cc: a.melis@aino.nl (Angelo), linux-arm-kernel@lists.arm.linux.org.uk In-Reply-To: from "Chris Blazie" at Aug 28, 2000 09:37:06 AM X-Mailer: ELM [version 2.5 PL1] MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk Chris Blazie writes: > As far as I know, memory sizes greater than 32MB are not supported > in the Strongarm kernel right now. You can verify this by modifying > arch/arm/arch.c and arch/arm/setup.c to declare only 32MB and see if it > works. Our hardware works when "detuned" like that; at some point I'm > planning to take a look at the problem if it's still there. Huh? _____ |_____| ------------------------------------------------- ---+---+- | | Russell King linux@arm.linux.org.uk --- --- | | | | http://www.arm.linux.org.uk/ / / | | +-+-+ --- -+- / | THE developer of ARM Linux |+| /|\ / | | | --- | +-+-+ ------------------------------------------------- /\\\ | unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Mon Aug 28 15:00:37 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13TPSK-0002i2-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Mon, 28 Aug 2000 15:00:08 +0100 Received: from [207.97.93.2] (helo=blazie.com) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13TPSJ-0002hw-00 for linux-arm-kernel@lists.arm.linux.org.uk; Mon, 28 Aug 2000 15:00:07 +0100 Received: from localhost (localhost [127.0.0.1]) by blazie.com (8.9.2/8.8.7) with ESMTP id JAA21421; Mon, 28 Aug 2000 09:59:54 -0400 (EDT) Date: Mon, 28 Aug 2000 09:59:54 -0400 (EDT) From: Chris Blazie To: Russell King - ARM Linux Admin cc: Angelo , linux-arm-kernel@lists.arm.linux.org.uk Subject: Re: switched to strongARM and 160 Mb and Debian 2.2 doesn't boot In-Reply-To: <200008281340.OAA17412@flint.arm.linux.org.uk> Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk > > As far as I know, memory sizes greater than 32MB are not supported > > in the Strongarm kernel right now. You can verify this by modifying > > arch/arm/arch.c and arch/arm/setup.c to declare only 32MB and see if it > > works. Our hardware works when "detuned" like that; at some point I'm > > planning to take a look at the problem if it's still there. > > Huh? A couple of months ago when I first brought up our hardware, declared memory sizes greater than 32MB generated an oops on bootup. Based on the function backtraces, it definitely appeared that something in the kernel was choking on the larger memory size. Based on my messages about it, Nicolas tenatively agreed that something may be wrong. I suppose it could be our hardware, but the memory checks out fine with every memory test routine I ran in the bootloader. With plenty of other things to work on, I left the problem alone. Maybe it's time I checked it again. Regards, Chris unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Mon Aug 28 15:20:20 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13TPlf-0002w6-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Mon, 28 Aug 2000 15:20:07 +0100 Received: from [130.161.40.91] (helo=thorgal.et.tudelft.nl) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13TPle-0002vz-00 for linux-arm-kernel@lists.arm.linux.org.uk; Mon, 28 Aug 2000 15:20:06 +0100 Received: from [130.161.40.82] (tvs082.et.tudelft.nl [130.161.40.82]) by thorgal.et.tudelft.nl (8.8.7/8.8.7) with ESMTP id QAA21802; Mon, 28 Aug 2000 16:20:07 +0200 Message-Id: In-Reply-To: References: <200008281340.OAA17412@flint.arm.linux.org.uk> Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Date: Mon, 28 Aug 2000 16:19:56 +0200 To: Chris Blazie From: "J.D. Bakker" Subject: Re: switched to strongARM and 160 Mb and Debian 2.2 doesn't boot Cc: Russell King - ARM Linux Admin , Angelo , linux-arm-kernel@lists.arm.linux.org.uk Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk At 15:59 +0200 28-08-2000, Chris Blazie wrote: >> > As far as I know, memory sizes greater than 32MB are not supported >> > in the Strongarm kernel right now. > >> Huh? > > A couple of months ago when I first brought up our hardware, >declared memory sizes greater than 32MB generated an oops on bootup. That should be > 32MB *per bank* on the SA-11[01]0, right ? We have 64MB expansion boards for LART which work just fine when we use only 32MB (for a total of 64MB); attempts to use more memory result in the problems you describe. According to Nico it's an assumption in the kernel that a bank can never be >32MB. Fixing this is on our Todo-list, albeit not very high. JDB. -- Jan-Derk Bakker, bakker@mmc.et.tudelft.nl The lazy man's proverb: 'There's no business like slow business !' unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Mon Aug 28 15:29:42 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13TPug-00034i-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Mon, 28 Aug 2000 15:29:26 +0100 Received: from [194.237.142.110] (helo=penguin-ext.wise.edt.ericsson.se) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13TPud-00034c-00 for linux-arm-kernel@lists.arm.linux.org.uk; Mon, 28 Aug 2000 15:29:25 +0100 Received: from esealnt409.al.sw.ericsson.se (esealnt409.al.sw.ericsson.se [153.88.251.32]) by penguin.wise.edt.ericsson.se (8.11.0/8.10.1/WIREfire-1.3) with ESMTP id e7SETWw04327 for ; Mon, 28 Aug 2000 16:29:32 +0200 (MEST) Received: from esealnt409 ([153.88.251.32]) by esealnt409.al.sw.ericsson.se with Microsoft SMTPSVC(5.0.2172.1); Mon, 28 Aug 2000 16:29:05 +0200 Received: from esealnt743.al.sw.ericsson.se ([153.88.251.13]) by esealnt409 (NAVIEG 2.1 bld 61) with SMTP id M2000082816290128380 ; Mon, 28 Aug 2000 16:29:01 +0200 Received: by esealnt743.al.sw.ericsson.se with Internet Mail Service (5.5.2651.58) id ; Mon, 28 Aug 2000 16:28:04 +0200 Message-ID: From: "Arjan Opmeer (EMN)" To: "'linux-arm-kernel@lists.arm.linux.org.uk'" Cc: "'imai@vasara.co.jp'" Subject: Kunihiko Imais USB ohci driver doesn't work? Date: Mon, 28 Aug 2000 16:29:19 +0200 MIME-Version: 1.0 X-Mailer: Internet Mail Service (5.5.2651.58) Content-Type: text/plain; charset="iso-8859-1" X-OriginalArrivalTime: 28 Aug 2000 14:29:05.0484 (UTC) FILETIME=[51B2A8C0:01C010FC] Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk Hi, I just tested Kunihiko Imais ohci driver for the SA1111 board that he recently mentioned here on the list, but it doesn't seem to work. I tested it against kernel 2.4.0-test6-rmk5-np2. It does detect the ohci hub, but not the USB mouse I connected to the port. The driver also detects an overcurrent condition on the port, with, but also without, the mouse attached. Did anyone else have success trying this driver? He also mentioned a possible SA1111 DMA bug which would prevent USB from working altogether. Does someone know whether Neponset can suffer from this bug? Thanks, Arjan unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Mon Aug 28 16:16:57 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13TQeP-0003dB-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Mon, 28 Aug 2000 16:16:41 +0100 Received: from [128.2.10.83] (helo=smtp3.andrew.cmu.edu) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13TQeO-0003d3-00 for linux-arm-kernel@lists.arm.linux.org.uk; Mon, 28 Aug 2000 16:16:40 +0100 Received: from unix10.andrew.cmu.edu (UNIX10.ANDREW.CMU.EDU [128.2.15.14]) by smtp3.andrew.cmu.edu (8.9.3/8.9.3) with SMTP id LAA21074; Mon, 28 Aug 2000 11:16:43 -0400 (EDT) Date: Mon, 28 Aug 2000 11:16:43 -0400 (EDT) From: John G Dorsey To: "Arjan Opmeer (EMN)" cc: "'linux-arm-kernel@lists.arm.linux.org.uk'" Subject: Re: Kunihiko Imais USB ohci driver doesn't work? In-Reply-To: Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk On Mon, 28 Aug 2000, Arjan Opmeer (EMN) wrote: > He also mentioned a possible SA1111 DMA bug which would prevent USB > from working altogether. Does someone know whether Neponset can > suffer from this bug? The bug is with the SA-1111; it would appear in any implementation. See the SA-1111 specification update (June 2000), errata #7. (Disclaimer: I haven't looked into this yet, and wouldn't be surprised if the following feature already exists in the kernel.) One possible solution would be to add a constraints parameter to the uncached buffer request service. There are at least two constraints that a viable SA-1111 DMA buffer must meet: 1. Must reside in whichever dynamic memory bank is wired for SA-1111 access. (This is jumper-selectable on Neponset, for example.) 2. If the buffer is placed above the 1MB mark in its respective bank, it must reside in a particular range / be addressable with a particular mask / &c. This is necessary for my SAC patch as well, so if you'd like to run with it, please keep in touch. =) -jd unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Mon Aug 28 16:19:09 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13TQga-0003eo-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Mon, 28 Aug 2000 16:18:56 +0100 Received: from [205.250.170.220] (helo=xanadu.vipswitch.com) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13TQgZ-0003eh-00 for linux-arm-kernel@lists.arm.linux.org.uk; Mon, 28 Aug 2000 16:18:55 +0100 Date: Mon, 28 Aug 2000 11:17:43 -0400 (EDT) From: Nicolas Pitre To: Russell King - ARM Linux Admin cc: Chris Blazie , Angelo , linux-arm-kernel@lists.arm.linux.org.uk Subject: Re: switched to strongARM and 160 Mb and Debian 2.2 doesn't boot In-Reply-To: <200008281340.OAA17412@flint.arm.linux.org.uk> Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk On Mon, 28 Aug 2000, Russell King - ARM Linux Admin wrote: > Chris Blazie writes: > > As far as I know, memory sizes greater than 32MB are not supported > > in the Strongarm kernel right now. You can verify this by modifying > > arch/arm/arch.c and arch/arm/setup.c to declare only 32MB and see if it > > works. Our hardware works when "detuned" like that; at some point I'm > > planning to take a look at the problem if it's still there. > > Huh? This restriction is only SA11x0 related, and the limit is actually 32 MB for each of the four memory banks. This is due to my original and incorrect design which I plan to fix once I'll be sure to do it the right way. Since the generic discontigous memory support moved a bit in latest kernels, I wanted to wait a bit more. In short, this doesn't concern other architectures than SA1100 based. Nicolas unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Mon Aug 28 16:44:22 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13TQzv-0003um-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Mon, 28 Aug 2000 16:38:55 +0100 Received: from [194.237.142.110] (helo=penguin-ext.wise.edt.ericsson.se) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13TQzv-0003ue-00 for linux-arm-kernel@lists.arm.linux.org.uk; Mon, 28 Aug 2000 16:38:55 +0100 Received: from esealnt409.al.sw.ericsson.se (esealnt409.al.sw.ericsson.se [153.88.251.32]) by penguin.wise.edt.ericsson.se (8.11.0/8.10.1/WIREfire-1.3) with ESMTP id e7SFd4w29610 for ; Mon, 28 Aug 2000 17:39:04 +0200 (MEST) Received: from esealnt409 ([153.88.251.32]) by esealnt409.al.sw.ericsson.se with Microsoft SMTPSVC(5.0.2172.1); Mon, 28 Aug 2000 17:38:37 +0200 Received: from esealnt743.al.sw.ericsson.se ([153.88.251.13]) by esealnt409 (NAVIEG 2.1 bld 61) with SMTP id M2000082817383727370 ; Mon, 28 Aug 2000 17:38:37 +0200 Received: by esealnt743.al.sw.ericsson.se with Internet Mail Service (5.5.2651.58) id ; Mon, 28 Aug 2000 17:39:03 +0200 Message-ID: From: "Arjan Opmeer (EMN)" To: "'John G Dorsey'" , "Arjan Opmeer (EMN)" Cc: "'linux-arm-kernel@lists.arm.linux.org.uk'" Subject: RE: Kunihiko Imais USB ohci driver doesn't work? Date: Mon, 28 Aug 2000 17:38:58 +0200 MIME-Version: 1.0 X-Mailer: Internet Mail Service (5.5.2651.58) Content-Type: text/plain; charset="iso-8859-1" X-OriginalArrivalTime: 28 Aug 2000 15:38:37.0125 (UTC) FILETIME=[08308F50:01C01106] Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk Hi, > > He also mentioned a possible SA1111 DMA bug which would prevent USB > > from working altogether. Does someone know whether Neponset can > > suffer from this bug? > > The bug is with the SA-1111; it would appear in any implementation. > > See the SA-1111 specification update (June 2000), errata #7. [ snip above 1MB memory DMA transfer problem ] Yes, I heard about this one. But Kunihiko Imai seems to have spoken to a circuit designer who told him that there was a problem with a SA1111 DMA line that "now" (?) should be fixed. The bug seems to interfere with DMA bus mastering. He also says that he suffered from the bug for a month... (and then probably had his hardware fixed?). I am curious whether there really is a hardware problem with the Neponset which would interfere with the USB driver. And if so which revisions do have this symptom and how you can get your board fixed. > This is necessary for my SAC patch as well, so if you'd like > to run with it, please keep in touch. =) I am interested in anything that helps me use the USB port or the Ethernet port on the SA1111 board. So keep me informed... ;) Arjan unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Mon Aug 28 17:07:31 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13TRRJ-0004Ce-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Mon, 28 Aug 2000 17:07:13 +0100 Received: from dyn-33.linux.theplanet.co.uk ([195.92.244.33] helo=caramon.arm.linux.org.uk) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13TRRG-0004CW-00; Mon, 28 Aug 2000 17:07:11 +0100 Received: from flint.arm.linux.org.uk (root@flint [192.168.0.4]) by caramon.arm.linux.org.uk (8.9.3/8.9.3) with ESMTP id RAA09855; Mon, 28 Aug 2000 17:07:17 +0100 Received: (from linux@localhost) by flint.arm.linux.org.uk (8.9.3/8.9.3) id RAA18542; Mon, 28 Aug 2000 17:06:20 +0100 From: Russell King - ARM Linux Admin Message-Id: <200008281606.RAA18542@flint.arm.linux.org.uk> Subject: Re: switched to strongARM and 160 Mb and Debian 2.2 doesn't boot To: nico@cam.org (Nicolas Pitre) Date: Mon, 28 Aug 2000 17:06:20 +0100 (BST) Cc: chris@blazie.com (Chris Blazie), a.melis@aino.nl (Angelo), linux-arm-kernel@lists.arm.linux.org.uk In-Reply-To: from "Nicolas Pitre" at Aug 28, 2000 11:17:43 AM X-Mailer: ELM [version 2.5 PL1] MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk Nicolas Pitre writes: > > > > On Mon, 28 Aug 2000, Russell King - ARM Linux Admin wrote: > > > Chris Blazie writes: > > > As far as I know, memory sizes greater than 32MB are not supported > > > in the Strongarm kernel right now. You can verify this by modifying > > > arch/arm/arch.c and arch/arm/setup.c to declare only 32MB and see if it > > > works. Our hardware works when "detuned" like that; at some point I'm > > > planning to take a look at the problem if it's still there. > > > > Huh? > > This restriction is only SA11x0 related, and the limit is actually 32 MB > for each of the four memory banks. This is due to my original and > incorrect design which I plan to fix once I'll be sure to do it the > right way. Since the generic discontigous memory support moved a bit in > latest kernels, I wanted to wait a bit more. > > In short, this doesn't concern other architectures than SA1100 based. Err, this has nothing to do with the thread - it was concerning >32MB on a RiscPC, not SA11x0. _____ |_____| ------------------------------------------------- ---+---+- | | Russell King linux@arm.linux.org.uk --- --- | | | | http://www.arm.linux.org.uk/ / / | | +-+-+ --- -+- / | THE developer of ARM Linux |+| /|\ / | | | --- | +-+-+ ------------------------------------------------- /\\\ | unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Mon Aug 28 17:47:21 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13TS3M-0004cf-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Mon, 28 Aug 2000 17:46:32 +0100 Received: from [12.38.17.9] (helo=mail.aeptec.com) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13TS3L-0004cZ-00 for linux-arm-kernel@lists.arm.linux.org.uk; Mon, 28 Aug 2000 17:46:32 +0100 Received: by MAIL with Internet Mail Service (5.5.2650.21) id ; Mon, 28 Aug 2000 12:36:16 -0400 Message-ID: <32CC5B62AF0BD2119E4C00A0C9663E221F3C1A@MAIL> From: "Sun, Lei" To: "'linux-arm-kernel@lists.arm.linux.org.uk'" Subject: usb based ethernet card in StrongArm linux Date: Mon, 28 Aug 2000 12:36:13 -0400 MIME-Version: 1.0 X-Mailer: Internet Mail Service (5.5.2650.21) Content-Type: text/plain; charset="iso-8859-1" Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk Hi all: Does anybody know if there is usb based ethernet card which has linux driver. so I can have two ethernet on SA1110, don't need the neponset anymore? Thank you! Lei Sun unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Mon Aug 28 18:00:46 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13TSGx-0004le-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Mon, 28 Aug 2000 18:00:35 +0100 Received: from [192.58.206.2] (helo=crl.dec.com) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13TSGx-0004lW-00 for linux-arm-kernel@lists.arm.linux.org.uk; Mon, 28 Aug 2000 18:00:35 +0100 Received: from crlwebmail.crl.dec.com (crlwebmail.crl.dec.com [16.11.0.2]) by crl.dec.com (8.8.8/RWD-1.2) with ESMTP id NAA07730; Mon, 28 Aug 2000 13:00:29 -0400 (EDT) Received: by crlwebmail.crl.dec.com with Internet Mail Service (5.5.2650.21) id ; Mon, 28 Aug 2000 13:05:52 -0400 Message-ID: From: George France To: "'Nicolas Pitre'" , Russell King - ARM Linux Admin Cc: Chris Blazie , Angelo , linux-arm-kernel@lists.arm.linux.org.uk Subject: RE: switched to strongARM and 160 Mb and Debian 2.2 doesn't boot Date: Mon, 28 Aug 2000 13:00:31 -0400 MIME-Version: 1.0 X-Mailer: Internet Mail Service (5.5.2650.21) Content-Type: text/plain; charset="iso-8859-1" Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk Hello Nico; We are having the same problem. On Friday, We removed my 32 MB chips from my only bank on the iPAQ and soldered on some 64 MB chips. Best Regards, --George > This restriction is only SA11x0 related, and the limit is > actually 32 MB > for each of the four memory banks. This is due to my original and > incorrect design which I plan to fix once I'll be sure to do it the > right way. Since the generic discontigous memory support > moved a bit in > latest kernels, I wanted to wait a bit more. > > In short, this doesn't concern other architectures than SA1100 based. > > > Nicolas > > > > unsubscribe: body of `unsubscribe linux-arm-kernel' to > majordomo@lists.arm.linux.org.uk > unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Mon Aug 28 18:23:17 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13TScb-0004zE-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Mon, 28 Aug 2000 18:22:57 +0100 Received: from [130.161.42.1] (helo=duteinh.et.tudelft.nl) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13TSca-0004z8-00 for linux-arm-kernel@lists.arm.linux.org.uk; Mon, 28 Aug 2000 18:22:56 +0100 Received: from dutein75 by duteinh.et.tudelft.nl (8.9.3/BB-2.0) id TAA14407; Mon, 28 Aug 2000 19:22:32 +0200 (MET DST) Message-Id: <200008281722.TAA14407@duteinh.et.tudelft.nl> From: Erik Mouw To: france@crl.dec.com Cc: nico@cam.org, linux@arm.linux.org.uk, chris@blazie.com, a.melis@aino.nl, linux-arm-kernel@lists.arm.linux.org.uk Subject: RE: switched to strongARM and 160 Mb and Debian 2.2 doesn't boot In-Reply-To: Organization: Eric conspiracy secret labs Date: Mon 28 Aug 2000 19:22:31 +0100 (MDT) X-Face: "!Z|1XjFg!3Iw)BwM^M\k)J5KHiPddz}#/Zw%?$&Q!xqob"OA|})3_@&BTTiQERo?/^g[Z<9P4R8:5C:aqq.cw!hotLc=+u-"+1rq@9$V^8PbOUlodyi_sejt+D`iqkLpx$>.aXl%vnfPIt"LK|/%>94_Zrm8jy?`_[RKt'TuoxUxfk(gMhY(n|DBBl~c$dhO>G%6}R4+DGX3|idBm4oU&UM/XUP4I9aoN{]`ax%B?rgD(I3OJ<62Y6L X-Mailer: XCmail 1.2devel - with PGP support, PGP engine version 0.5 (IRIX64) X-Mailerorigin: http://www.fsai.fh-trier.de/~schmitzj/Xclasses/XCmail/ X-Eric-Conspiracy: There is no conspiracy! X-Loop: erik@it.et.tudelft.nl MIME-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: 8bit Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk On Mon, 28 Aug 2000 13:00:31 -0400, George France wrote: > We are having the same problem. On Friday, We removed my 32 MB chips from my > only bank on the iPAQ and soldered on some 64 MB chips. You know this will void the warranty of your iPaq? ;-) But same over here on the LART with an extra 64MB board: I can only use the first 32MB of the 64MB RAM expansion. Nico, is there an easy way to fix the 32MB per bank limit? Can you point us where to look in the code? Erik -- J.A.K. (Erik) Mouw, Information and Communication Theory Group, Department of Electrical Engineering, Faculty of Information Technology and Systems, Delft University of Technology, PO BOX 5031, 2600 GA Delft, The Netherlands Phone: +31-15-2783635 Fax: +31-15-2781843 Email: J.A.K.Mouw@its.tudelft.nl WWW: http://www-ict.its.tudelft.nl/~erik/ unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Mon Aug 28 18:24:48 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13TSeG-00050H-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Mon, 28 Aug 2000 18:24:40 +0100 Received: from [130.161.42.1] (helo=duteinh.et.tudelft.nl) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13TSeF-000509-00 for linux-arm-kernel@lists.arm.linux.org.uk; Mon, 28 Aug 2000 18:24:39 +0100 Received: from dutein75 by duteinh.et.tudelft.nl (8.9.3/BB-2.0) id TAA14418; Mon, 28 Aug 2000 19:24:47 +0200 (MET DST) Message-Id: <200008281724.TAA14418@duteinh.et.tudelft.nl> From: Erik Mouw To: Sun@AEPTEC.COM Cc: linux-arm-kernel@lists.arm.linux.org.uk Subject: Re: usb based ethernet card in StrongArm linux In-Reply-To: <32CC5B62AF0BD2119E4C00A0C9663E221F3C1A@MAIL> Organization: Eric conspiracy secret labs Date: Mon 28 Aug 2000 19:24:47 +0100 (MDT) X-Face: "!Z|1XjFg!3Iw)BwM^M\k)J5KHiPddz}#/Zw%?$&Q!xqob"OA|})3_@&BTTiQERo?/^g[Z<9P4R8:5C:aqq.cw!hotLc=+u-"+1rq@9$V^8PbOUlodyi_sejt+D`iqkLpx$>.aXl%vnfPIt"LK|/%>94_Zrm8jy?`_[RKt'TuoxUxfk(gMhY(n|DBBl~c$dhO>G%6}R4+DGX3|idBm4oU&UM/XUP4I9aoN{]`ax%B?rgD(I3OJ<62Y6L X-Mailer: XCmail 1.2devel - with PGP support, PGP engine version 0.5 (IRIX64) X-Mailerorigin: http://www.fsai.fh-trier.de/~schmitzj/Xclasses/XCmail/ X-Eric-Conspiracy: There is no conspiracy! X-Loop: erik@it.et.tudelft.nl MIME-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: 8bit Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk On Mon, 28 Aug 2000 12:36:13 -0400, Sun, Lei wrote: > Does anybody know if there is usb based ethernet card which has linux > driver. so I can have two ethernet on SA1110, don't need the neponset > anymore? If you find an USB based ethernet card, you really need the Neponset. The SA1110 alone is USB _client_, not an USB _master_. Erik -- Real Programmers consider "what you see is what you get" to be just as bad a concept in Text Editors as it is in women. No, the Real Programmer wants a "you asked for it, you got it" text editor -- complicated, cryptic, powerful, unforgiving, dangerous. unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Mon Aug 28 20:19:16 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13TUQX-0006Mb-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Mon, 28 Aug 2000 20:18:37 +0100 Received: from [205.250.170.220] (helo=xanadu.vipswitch.com) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13TUQV-0006MR-00 for linux-arm-kernel@lists.arm.linux.org.uk; Mon, 28 Aug 2000 20:18:36 +0100 Date: Mon, 28 Aug 2000 15:17:24 -0400 (EDT) From: Nicolas Pitre To: Erik Mouw cc: france@crl.dec.com, linux@arm.linux.org.uk, chris@blazie.com, linux-arm-kernel@lists.arm.linux.org.uk Subject: Re: over 32MB per bank on the SA11x0 In-Reply-To: <200008281722.TAA14407@duteinh.et.tudelft.nl> Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk On Mon, 28 Aug 2000, Erik Mouw wrote: > Nico, is there an easy way to fix the 32MB per bank limit? Can you point > us where to look in the code? It's kind of scattered. As a prerequisite, we need to change the PCMCIA code to ioremap() the PCMCIA address spaces. We need to get rid of most of the PCMCIA region from 0xe0000000 to 0xf7ffffff. They are probably excessively wide and the PCMCIA code must know how large it wish them to be. Nicolas unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Mon Aug 28 20:27:19 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13TUYi-0006SN-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Mon, 28 Aug 2000 20:27:04 +0100 Received: from [130.161.42.1] (helo=duteinh.et.tudelft.nl) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13TUYi-0006SG-00 for linux-arm-kernel@lists.arm.linux.org.uk; Mon, 28 Aug 2000 20:27:04 +0100 Received: from dutein75 by duteinh.et.tudelft.nl (8.9.3/BB-2.0) id VAA15049; Mon, 28 Aug 2000 21:26:58 +0200 (MET DST) Message-Id: <200008281926.VAA15049@duteinh.et.tudelft.nl> From: Erik Mouw To: nico@cam.org Cc: france@crl.dec.com, linux@arm.linux.org.uk, chris@blazie.com, linux-arm-kernel@lists.arm.linux.org.uk Subject: Re: over 32MB per bank on the SA11x0 In-Reply-To: Organization: Eric conspiracy secret labs Date: Mon 28 Aug 2000 21:26:57 +0100 (MDT) X-Face: "!Z|1XjFg!3Iw)BwM^M\k)J5KHiPddz}#/Zw%?$&Q!xqob"OA|})3_@&BTTiQERo?/^g[Z<9P4R8:5C:aqq.cw!hotLc=+u-"+1rq@9$V^8PbOUlodyi_sejt+D`iqkLpx$>.aXl%vnfPIt"LK|/%>94_Zrm8jy?`_[RKt'TuoxUxfk(gMhY(n|DBBl~c$dhO>G%6}R4+DGX3|idBm4oU&UM/XUP4I9aoN{]`ax%B?rgD(I3OJ<62Y6L X-Mailer: XCmail 1.2devel - with PGP support, PGP engine version 0.5 (IRIX64) X-Mailerorigin: http://www.fsai.fh-trier.de/~schmitzj/Xclasses/XCmail/ X-Eric-Conspiracy: There is no conspiracy! X-Loop: erik@it.et.tudelft.nl MIME-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: 8bit Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk On Mon, 28 Aug 2000 15:17:24 -0400 (EDT), Nicolas Pitre wrote: > On Mon, 28 Aug 2000, Erik Mouw wrote: >> Nico, is there an easy way to fix the 32MB per bank limit? Can you point >> us where to look in the code? > > It's kind of scattered. > > As a prerequisite, we need to change the PCMCIA code to ioremap() the > PCMCIA address spaces. We need to get rid of most of the PCMCIA region > from 0xe0000000 to 0xf7ffffff. They are probably excessively wide and the > PCMCIA code must know how large it wish them to be. Why? From 0xc0000000 to 0xdfffffff is 512MB of memory, which is the maximum physical memory on an SA11x0 CPU. No need to move the PCMCIA space. However, there are some SA11x0 boards that use 0xd0000000 to map the flash memory. So either we limit the total DRAM to 256MB, or we move the flash region. Erik -- I'm not a lawyer. I don't even play one on TV. -- Linus Torvalds on the gcc mailing list unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Mon Aug 28 21:10:58 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13TVEu-0006uN-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Mon, 28 Aug 2000 21:10:40 +0100 Received: from dyn-33.linux.theplanet.co.uk ([195.92.244.33] helo=caramon.arm.linux.org.uk) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13TVEs-0006u9-00; Mon, 28 Aug 2000 21:10:38 +0100 Received: from flint.arm.linux.org.uk (root@flint [192.168.0.4]) by caramon.arm.linux.org.uk (8.9.3/8.9.3) with ESMTP id VAA10820; Mon, 28 Aug 2000 21:08:13 +0100 Received: (from linux@localhost) by flint.arm.linux.org.uk (8.9.3/8.9.3) id VAA19461; Mon, 28 Aug 2000 21:07:15 +0100 From: Russell King - ARM Linux Admin Message-Id: <200008282007.VAA19461@flint.arm.linux.org.uk> Subject: Re: switched to strongARM and 160 Mb and Debian 2.2 doesn't boot To: J.A.K.Mouw@its.tudelft.nl (Erik Mouw) Date: Mon, 28 Aug 2000 21:07:15 +0100 (BST) Cc: france@crl.dec.com, nico@cam.org, chris@blazie.com, a.melis@aino.nl, linux-arm-kernel@lists.arm.linux.org.uk In-Reply-To: <200008281722.TAA14407@duteinh.et.tudelft.nl> from "Erik Mouw" at Aug 28, 2000 07:22:31 PM X-Mailer: ELM [version 2.5 PL1] MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk Erik Mouw writes: > Nico, is there an easy way to fix the 32MB per bank limit? Can you point > us where to look in the code? See linux/include/asm-arm/arch-sa1100/memory.h - its to do with the virtual to physical memory translation "compression" that is done. Note, however, that there are several problems with modifing that: 1. It may cause machines with only 4MB in the first bank to fail to boot. 2. Effectively, you make 0xc0000000 through to 0xdfffffff available for the direct mapped region; unfortunately, 0xd0000000 upwards is used for IO (typically flash, SCR, BCR and the SA1111 on Assabet). See linux/arch/arm/mm/mm-sa1100.c for that. In all, getting >32MB per bank will probably not be trivial. _____ |_____| ------------------------------------------------- ---+---+- | | Russell King linux@arm.linux.org.uk --- --- | | | | http://www.arm.linux.org.uk/ / / | | +-+-+ --- -+- / | THE developer of ARM Linux |+| /|\ / | | | --- | +-+-+ ------------------------------------------------- /\\\ | unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Mon Aug 28 21:12:06 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13TVGA-0006vk-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Mon, 28 Aug 2000 21:11:58 +0100 Received: from dyn-33.linux.theplanet.co.uk ([195.92.244.33] helo=caramon.arm.linux.org.uk) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13TVG8-0006vb-00; Mon, 28 Aug 2000 21:11:57 +0100 Received: from flint.arm.linux.org.uk (root@flint [192.168.0.4]) by caramon.arm.linux.org.uk (8.9.3/8.9.3) with ESMTP id VAA10925; Mon, 28 Aug 2000 21:12:03 +0100 Received: (from linux@localhost) by flint.arm.linux.org.uk (8.9.3/8.9.3) id VAA19501; Mon, 28 Aug 2000 21:11:05 +0100 From: Russell King - ARM Linux Admin Message-Id: <200008282011.VAA19501@flint.arm.linux.org.uk> Subject: Re: over 32MB per bank on the SA11x0 To: J.A.K.Mouw@its.tudelft.nl (Erik Mouw) Date: Mon, 28 Aug 2000 21:11:05 +0100 (BST) Cc: nico@cam.org, france@crl.dec.com, chris@blazie.com, linux-arm-kernel@lists.arm.linux.org.uk In-Reply-To: <200008281926.VAA15049@duteinh.et.tudelft.nl> from "Erik Mouw" at Aug 28, 2000 09:26:57 PM X-Mailer: ELM [version 2.5 PL1] MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk Erik Mouw writes: > Why? From 0xc0000000 to 0xdfffffff is 512MB of memory, which is the > maximum physical memory on an SA11x0 CPU. No need to move the PCMCIA > space. However, there are some SA11x0 boards that use 0xd0000000 to map > the flash memory. So either we limit the total DRAM to 256MB, or we move > the flash region. If 0xdfffffff is the top of your memory, and 0xe0000000 is where IO-type stuff starts, which nibble does the vmalloc space fit in between these? Yes, vmalloc sits above the kernel direct mapped memory, so 0xe0000000 upwards has to change as well. _____ |_____| ------------------------------------------------- ---+---+- | | Russell King linux@arm.linux.org.uk --- --- | | | | http://www.arm.linux.org.uk/ / / | | +-+-+ --- -+- / | THE developer of ARM Linux |+| /|\ / | | | --- | +-+-+ ------------------------------------------------- /\\\ | unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Mon Aug 28 21:32:51 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13TVZq-0007C3-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Mon, 28 Aug 2000 21:32:18 +0100 Received: from [130.161.40.91] (helo=thorgal.et.tudelft.nl) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13TVZp-0007Bw-00 for linux-arm-kernel@lists.arm.linux.org.uk; Mon, 28 Aug 2000 21:32:17 +0100 Received: from [130.161.40.82] (tvs082.et.tudelft.nl [130.161.40.82]) by thorgal.et.tudelft.nl (8.8.7/8.8.7) with ESMTP id WAA23352; Mon, 28 Aug 2000 22:32:01 +0200 Message-Id: In-Reply-To: <200008282007.VAA19461@flint.arm.linux.org.uk> References: <200008281722.TAA14407@duteinh.et.tudelft.nl> from "Erik Mouw" at Aug 28, 2000 07:22:31 PM Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Date: Mon, 28 Aug 2000 22:30:38 +0200 To: Russell King - ARM Linux Admin From: "J.D. Bakker" Subject: Re: switched to strongARM and 160 Mb and Debian 2.2 doesn't boot Cc: J.A.K.Mouw@its.tudelft.nl (Erik Mouw), france@crl.dec.com, nico@cam.org, chris@blazie.com, a.melis@aino.nl, linux-arm-kernel@lists.arm.linux.org.uk Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk At 22:07 +0200 28-08-2000, Russell King - ARM Linux Admin wrote: >2. Effectively, you make 0xc0000000 through to 0xdfffffff available for > the direct mapped region; unfortunately, 0xd0000000 upwards is used > for IO (typically flash, SCR, BCR and the SA1111 on Assabet). > See linux/arch/arm/mm/mm-sa1100.c for that. Maybe I'm not seeing the bigger picture, but wouldn't changing #define __virt_to_phys(x) (((x) & 0xf9ffffff) | ((x) & 0x06000000) << 2) #define __phys_to_virt(x) (((x) & 0xe7ffffff) | ((x) & 0x18000000) >> 2) to #define __virt_to_phys(x) (((x) & 0xf9ffffff) | ((x) & 0x06000000) << 1) #define __phys_to_virt(x) (((x) & 0xe7ffffff) | ((x) & 0x18000000) >> 1) yield 64M banks, while using only the 0xCxxxxxxx region for DRAM ? I admit I don't know too much about the new memory mappings, but 0xC0000000 - 0xCFFFFFFF seems otherwise empty, right ? JDB [who should maybe stick to hardware, but...] -- LART. 250 MIPS under one Watt. Free hardware design files. http://www.lart.tudelft.nl/ unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Mon Aug 28 21:43:57 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13TVks-0007MD-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Mon, 28 Aug 2000 21:43:42 +0100 Received: from dyn-33.linux.theplanet.co.uk ([195.92.244.33] helo=caramon.arm.linux.org.uk) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13TVkq-0007M4-00; Mon, 28 Aug 2000 21:43:40 +0100 Received: from flint.arm.linux.org.uk (root@flint [192.168.0.4]) by caramon.arm.linux.org.uk (8.9.3/8.9.3) with ESMTP id VAA11098; Mon, 28 Aug 2000 21:43:46 +0100 Received: (from linux@localhost) by flint.arm.linux.org.uk (8.9.3/8.9.3) id VAA19654; Mon, 28 Aug 2000 21:42:48 +0100 From: Russell King - ARM Linux Admin Message-Id: <200008282042.VAA19654@flint.arm.linux.org.uk> Subject: Re: switched to strongARM and 160 Mb and Debian 2.2 doesn't boot To: bakker@thorgal.et.tudelft.nl (J.D. Bakker) Date: Mon, 28 Aug 2000 21:42:48 +0100 (BST) Cc: J.A.K.Mouw@its.tudelft.nl (Erik Mouw), france@crl.dec.com, nico@cam.org, chris@blazie.com, a.melis@aino.nl, linux-arm-kernel@lists.arm.linux.org.uk In-Reply-To: from "J.D. Bakker" at Aug 28, 2000 10:30:38 PM X-Mailer: ELM [version 2.5 PL1] MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk J.D. Bakker writes: > #define __virt_to_phys(x) (((x) & 0xf9ffffff) | ((x) & 0x06000000) << 2) > #define __phys_to_virt(x) (((x) & 0xe7ffffff) | ((x) & 0x18000000) >> 2) > > to > > #define __virt_to_phys(x) (((x) & 0xf9ffffff) | ((x) & 0x06000000) << 1) > #define __phys_to_virt(x) (((x) & 0xe7ffffff) | ((x) & 0x18000000) >> 1) > > yield 64M banks, while using only the 0xCxxxxxxx region for DRAM ? I admit > I don't know too much about the new memory mappings, but 0xC0000000 - > 0xCFFFFFFF seems otherwise empty, right ? Not quite correct, but: #define __virt_to_phys(x) (((x) & 0xf3ffffff) | ((x) & 0x0c000000) << 1) #define __phys_to_virt(x) (((x) & 0xe7ffffff) | ((x) & 0x18000000) >> 1) is probably what you were meaning, and yes, it still fits in the 0xc0000000 to 0xcfffffff space. But what about the vmalloc area, when you've got all the banks used and full (ie 256MB of memory)? You need to move stuff at 0xd0000000 out of the way at least. _____ |_____| ------------------------------------------------- ---+---+- | | Russell King linux@arm.linux.org.uk --- --- | | | | http://www.arm.linux.org.uk/ / / | | +-+-+ --- -+- / | THE developer of ARM Linux |+| /|\ / | | | --- | +-+-+ ------------------------------------------------- /\\\ | unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Mon Aug 28 22:04:52 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13TW3u-0007dz-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Mon, 28 Aug 2000 22:03:22 +0100 Received: from [130.161.40.91] (helo=thorgal.et.tudelft.nl) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13TW3t-0007dt-00 for linux-arm-kernel@lists.arm.linux.org.uk; Mon, 28 Aug 2000 22:03:21 +0100 Received: from [130.161.40.82] (tvs082.et.tudelft.nl [130.161.40.82]) by thorgal.et.tudelft.nl (8.8.7/8.8.7) with ESMTP id XAA23458; Mon, 28 Aug 2000 23:03:16 +0200 Message-Id: In-Reply-To: <200008282042.VAA19654@flint.arm.linux.org.uk> References: from "J.D. Bakker" at Aug 28, 2000 10:30:38 PM Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Date: Mon, 28 Aug 2000 23:03:07 +0200 To: Russell King - ARM Linux Admin From: "J.D. Bakker" Subject: Re: switched to strongARM and 160 Mb and Debian 2.2 doesn't boot Cc: J.A.K.Mouw@its.tudelft.nl (Erik Mouw), france@crl.dec.com, nico@cam.org, chris@blazie.com, a.melis@aino.nl, linux-arm-kernel@lists.arm.linux.org.uk Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk At 22:42 +0200 28-08-2000, Russell King - ARM Linux Admin wrote: >#define __virt_to_phys(x) (((x) & 0xf3ffffff) | ((x) & 0x0c000000) << 1) >#define __phys_to_virt(x) (((x) & 0xe7ffffff) | ((x) & 0x18000000) >> 1) > >is probably what you were meaning, Duh. Yes, you're right; shows me once again to test theories before posting them. OTOH, a LART with two 16M banks and this change (and the obvious one to mmzone.h) boots and runs happily ;-) > and yes, it still fits in the 0xc0000000 >to 0xcfffffff space. But what about the vmalloc area, when you've got all >the banks used and full (ie 256MB of memory)? That's not likely to be the normal mode of operation on SA boards (famous last words). With increasing SDRAM densities and tough SDRAM timing requirements on a loaded bus I expect the norm to be 1 or 2 banks filled. That being said it would be silly to have a default config that breaks when, say, bank 3 has 64M in it yes. >You need to move stuff at 0xd0000000 out of the way at least. Agreed. BTW, why isn't DRAM mapped contiguously any more ? And how do other platforms handle this when faced with, say, 1GB DRAM and the standard 3/1 split ? JDB. [who hasn't touched these code paths since 2.2.x] -- Jan-Derk Bakker, bakker@mmc.et.tudelft.nl The lazy man's proverb: 'There's no business like slow business !' unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Mon Aug 28 22:09:13 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13TW8o-0007iG-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Mon, 28 Aug 2000 22:08:26 +0100 Received: from dyn-33.linux.theplanet.co.uk ([195.92.244.33] helo=caramon.arm.linux.org.uk) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13TW8m-0007i5-00; Mon, 28 Aug 2000 22:08:25 +0100 Received: from flint.arm.linux.org.uk (root@flint [192.168.0.4]) by caramon.arm.linux.org.uk (8.9.3/8.9.3) with ESMTP id WAA11270; Mon, 28 Aug 2000 22:08:31 +0100 Received: (from linux@localhost) by flint.arm.linux.org.uk (8.9.3/8.9.3) id WAA19802; Mon, 28 Aug 2000 22:07:33 +0100 From: Russell King - ARM Linux Admin Message-Id: <200008282107.WAA19802@flint.arm.linux.org.uk> Subject: Re: switched to strongARM and 160 Mb and Debian 2.2 doesn't boot To: bakker@thorgal.et.tudelft.nl (J.D. Bakker) Date: Mon, 28 Aug 2000 22:07:32 +0100 (BST) Cc: J.A.K.Mouw@its.tudelft.nl (Erik Mouw), france@crl.dec.com, nico@cam.org, chris@blazie.com, a.melis@aino.nl, linux-arm-kernel@lists.arm.linux.org.uk In-Reply-To: from "J.D. Bakker" at Aug 28, 2000 11:03:07 PM X-Mailer: ELM [version 2.5 PL1] MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk J.D. Bakker writes: > At 22:42 +0200 28-08-2000, Russell King - ARM Linux Admin wrote: > >#define __virt_to_phys(x) (((x) & 0xf3ffffff) | ((x) & 0x0c000000) << 1) > >#define __phys_to_virt(x) (((x) & 0xe7ffffff) | ((x) & 0x18000000) >> 1) > Agreed. BTW, why isn't DRAM mapped contiguously any more ? And how do other > platforms handle this when faced with, say, 1GB DRAM and the standard 3/1 > split ? The problem is all to do with those two macros above. As they stand, they should compile down to: bic %0, x, #0x0c000000 and %1, x, #0x0c000000 orr %0, %0, %1, lsl #1 (but I imagine that as always GCC will produce far worse code than that) ;( These macros are used extensively within the memory management system, and, if they are slow, you have a slow machine. Therefore, we do everything we can to make these as fast as possible. Normally, I'd like them to be a pure phys = virt - offset type arrangement, but the SA1100 platforms are the only ones which don't allow this. _____ |_____| ------------------------------------------------- ---+---+- | | Russell King linux@arm.linux.org.uk --- --- | | | | http://www.arm.linux.org.uk/ / / | | +-+-+ --- -+- / | THE developer of ARM Linux |+| /|\ / | | | --- | +-+-+ ------------------------------------------------- /\\\ | unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Mon Aug 28 22:22:25 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13TWLR-0007rW-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Mon, 28 Aug 2000 22:21:29 +0100 Received: from [130.161.42.1] (helo=duteinh.et.tudelft.nl) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13TWLQ-0007rO-00 for linux-arm-kernel@lists.arm.linux.org.uk; Mon, 28 Aug 2000 22:21:28 +0100 Received: from dutein75 by duteinh.et.tudelft.nl (8.9.3/BB-2.0) id XAA15507; Mon, 28 Aug 2000 23:21:23 +0200 (MET DST) Message-Id: <200008282121.XAA15507@duteinh.et.tudelft.nl> From: Erik Mouw To: linux@arm.linux.org.uk Cc: bakker@thorgal.et.tudelft.nl, france@crl.dec.com, nico@cam.org, chris@blazie.com, a.melis@aino.nl, linux-arm-kernel@lists.arm.linux.org.uk Subject: Re: switched to strongARM and 160 Mb and Debian 2.2 doesn't boot In-Reply-To: <200008282042.VAA19654@flint.arm.linux.org.uk> Organization: Eric conspiracy secret labs Date: Mon 28 Aug 2000 23:21:22 +0100 (MDT) X-Face: "!Z|1XjFg!3Iw)BwM^M\k)J5KHiPddz}#/Zw%?$&Q!xqob"OA|})3_@&BTTiQERo?/^g[Z<9P4R8:5C:aqq.cw!hotLc=+u-"+1rq@9$V^8PbOUlodyi_sejt+D`iqkLpx$>.aXl%vnfPIt"LK|/%>94_Zrm8jy?`_[RKt'TuoxUxfk(gMhY(n|DBBl~c$dhO>G%6}R4+DGX3|idBm4oU&UM/XUP4I9aoN{]`ax%B?rgD(I3OJ<62Y6L X-Mailer: XCmail 1.2devel - with PGP support, PGP engine version 0.5 (IRIX64) X-Mailerorigin: http://www.fsai.fh-trier.de/~schmitzj/Xclasses/XCmail/ X-Eric-Conspiracy: There is no conspiracy! X-Loop: erik@it.et.tudelft.nl MIME-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: 8bit Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk On Mon, 28 Aug 2000 21:42:48 +0100 (BST), Russell King - ARM Linux Admin wrote: > J.D. Bakker writes: > Not quite correct, but: > > #define __virt_to_phys(x) (((x) & 0xf3ffffff) | ((x) & 0x0c000000) << 1) > #define __phys_to_virt(x) (((x) & 0xe7ffffff) | ((x) & 0x18000000) >> 1) > > is probably what you were meaning, and yes, it still fits in the 0xc0000000 > to 0xcfffffff space. But what about the vmalloc area, when you've got all > the banks used and full (ie 256MB of memory)? This patch allows 64MB/bank; tested on the LART, YMMV. Oh, and #include . -------------------------------------------------------------------- diff -urN linux-2.4.0-test6-rmk5-np2-vanilla/include/asm-arm/arch-sa1100/memory.h linux-2.4.0-test6-rmk5-np2/include/asm-arm/arch-sa1100/memory.h --- linux-2.4.0-test6-rmk5-np2-vanilla/include/asm-arm/arch-sa1100/memory.h Thu Jan 13 22:30:31 2000 +++ linux-2.4.0-test6-rmk5-np2/include/asm-arm/arch-sa1100/memory.h Mon Aug 28 23:02:36 2000 @@ -29,10 +29,11 @@ #define __phys_to_virt__is_a_macro /* - * The following gives a maximum memory size of 128MB (32MB in each bank). + * The following gives a maximum memory size of 256MB (64MB in each bank). */ -#define __virt_to_phys(x) (((x) & 0xf9ffffff) | ((x) & 0x06000000) << 2) -#define __phys_to_virt(x) (((x) & 0xe7ffffff) | ((x) & 0x18000000) >> 2) +#define __virt_to_phys(x) (((x) & 0xf3ffffff) | ((x) & 0x0c000000) << 1) +#define __phys_to_virt(x) (((x) & 0xe7ffffff) | ((x) & 0x18000000) >> 1) + /* * Virtual view <-> DMA view memory address translations diff -urN linux-2.4.0-test6-rmk5-np2-vanilla/include/asm-arm/arch-sa1100/mmzone.h linux-2.4.0-test6-rmk5-np2/include/asm-arm/arch-sa1100/mmzone.h --- linux-2.4.0-test6-rmk5-np2-vanilla/include/asm-arm/arch-sa1100/mmzone.h Thu Aug 24 12:52:38 2000 +++ linux-2.4.0-test6-rmk5-np2/include/asm-arm/arch-sa1100/mmzone.h Mon Aug 28 23:02:37 2000 @@ -20,9 +20,9 @@ extern pg_data_t sa1100_node_data[]; /* - * 32MB max in each bank, must fit with __virt_to_phys() & __phys_to_virt() + * 64MB max in each bank, must fit with __virt_to_phys() & __phys_to_virt() */ -#define NODE_MAX_MEM_SHIFT 25 +#define NODE_MAX_MEM_SHIFT 26 #define NODE_MAX_MEM_SIZE (1< You need to move stuff at 0xd0000000 out of the way at least. OK, so where is a good place? (Is the memory map on ARM linux somewhere defined?) Erik -- J.A.K. (Erik) Mouw, Information and Communication Theory Group, Department of Electrical Engineering, Faculty of Information Technology and Systems, Delft University of Technology, PO BOX 5031, 2600 GA Delft, The Netherlands Phone: +31-15-2783635 Fax: +31-15-2781843 Email: J.A.K.Mouw@its.tudelft.nl WWW: http://www-ict.its.tudelft.nl/~erik/ unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Mon Aug 28 22:36:59 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13TWZj-000831-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Mon, 28 Aug 2000 22:36:15 +0100 Received: from dyn-33.linux.theplanet.co.uk ([195.92.244.33] helo=caramon.arm.linux.org.uk) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13TWZg-00082u-00; Mon, 28 Aug 2000 22:36:13 +0100 Received: from flint.arm.linux.org.uk (root@flint [192.168.0.4]) by caramon.arm.linux.org.uk (8.9.3/8.9.3) with ESMTP id WAA11373; Mon, 28 Aug 2000 22:36:19 +0100 Received: (from linux@localhost) by flint.arm.linux.org.uk (8.9.3/8.9.3) id WAA19901; Mon, 28 Aug 2000 22:35:21 +0100 From: Russell King - ARM Linux Admin Message-Id: <200008282135.WAA19901@flint.arm.linux.org.uk> Subject: Re: switched to strongARM and 160 Mb and Debian 2.2 doesn't boot To: J.A.K.Mouw@its.tudelft.nl (Erik Mouw) Date: Mon, 28 Aug 2000 22:35:21 +0100 (BST) Cc: bakker@thorgal.et.tudelft.nl, france@crl.dec.com, nico@cam.org, chris@blazie.com, a.melis@aino.nl, linux-arm-kernel@lists.arm.linux.org.uk In-Reply-To: <200008282121.XAA15507@duteinh.et.tudelft.nl> from "Erik Mouw" at Aug 28, 2000 11:21:22 PM X-Mailer: ELM [version 2.5 PL1] MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk Erik Mouw writes: > This patch allows 64MB/bank; tested on the LART, YMMV. Ok, it looks mostly sane, but you do want to look at include/asm-arm/arch-sa1100/vmalloc.h as well - it defines the area for vmalloc/ioremap stuff. > > You need to move stuff at 0xd0000000 out of the way at least. > > OK, so where is a good place? That is up to Nico/individual platform people, but there are a large number of SA1100-based platforms that have stuff down at 0xd0000000. > (Is the memory map on ARM linux somewhere defined?) It's architecture dependent, so there is no "one true reference". However, in the spirit of Linux, it's in the source. arch/arm/mm/mm-sa1100.c defines the memory mapping above the so called "direct mapped" kernel memory that virt_to_phys and friends work on. In test6-rmk5 (?) and higher, its defined by two tables - a common "standard_io_desc" which is for all SA1100 devices, and the individual platform tables (eg, "assabet_io_desc"). Glue the relevent tables together, and sort by virtual address and you have the memory map. (Note that that is not the whole story, but as a generalisation, 0x0000-0x0fff are for vectory-type things, 0x1000 to 0xbfffffff is userspace, 0xc0000000 upwards is kernel direct mapped memory, and on top of that is IO). _____ |_____| ------------------------------------------------- ---+---+- | | Russell King linux@arm.linux.org.uk --- --- | | | | http://www.arm.linux.org.uk/ / / | | +-+-+ --- -+- / | THE developer of ARM Linux |+| /|\ / | | | --- | +-+-+ ------------------------------------------------- /\\\ | unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Mon Aug 28 22:46:05 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13TWid-0008Ak-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Mon, 28 Aug 2000 22:45:27 +0100 Received: from [208.162.98.43] (helo=lhs.com) by www.linux.org.uk with smtp (Exim 3.13 #1) id 13TWid-0008Ac-00 for linux-arm-kernel@lists.arm.linux.org.uk; Mon, 28 Aug 2000 22:45:27 +0100 Received: by lhs.com(Lotus SMTP MTA v4.6.6 (890.1 7-16-1999)) id 85256949.007880E7 ; Mon, 28 Aug 2000 17:56:13 -0400 X-Lotus-FromDomain: LHS From: ebunce@lhsl.com To: Russell King - ARM Linux Admin cc: bakker@thorgal.et.tudelft.nl (J.D. Bakker), J.A.K.Mouw@its.tudelft.nl (Erik Mouw), france@crl.dec.com, nico@cam.org, chris@blazie.com, a.melis@aino.nl, linux-arm-kernel@lists.arm.linux.org.uk Message-ID: <85256949.00788094.00@lhs.com> Date: Mon, 28 Aug 2000 17:43:50 -0400 Subject: Re: switched to strongARM and 160 Mb and Debian 2.2 doesn't boot Mime-Version: 1.0 Content-type: text/plain; charset=us-ascii Content-Disposition: inline Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk Russell King wrote: > is probably what you were meaning, and yes, it still fits in the 0xc0000000 > to 0xcfffffff space. But what about the vmalloc area, when you've got all > the banks used and full (ie 256MB of memory)? > Just for the record, According to the SA1110 Advance Developer Manual each of the four SDRAM banks in an SA1110 can contain 128 MB of RAM (using 32M x 8, or 64M x 16, or 64M x 4 components) for a total of 512MB of memory. I'm not saying you'd want to do this, but that it is technically possible using commercially available parts. For example: Using 4 Samsung K4S560832A 32Mx8 component Yields 128MB bank w/32 bit bus Using 4 Samsung K4S560432A 64Mx4 component Enjoy, Erik unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Mon Aug 28 22:50:06 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13TWmW-0008DX-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Mon, 28 Aug 2000 22:49:28 +0100 Received: from dyn-33.linux.theplanet.co.uk ([195.92.244.33] helo=caramon.arm.linux.org.uk) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13TWmU-0008DM-00; Mon, 28 Aug 2000 22:49:27 +0100 Received: from flint.arm.linux.org.uk (root@flint [192.168.0.4]) by caramon.arm.linux.org.uk (8.9.3/8.9.3) with ESMTP id WAA11433; Mon, 28 Aug 2000 22:49:33 +0100 Received: (from linux@localhost) by flint.arm.linux.org.uk (8.9.3/8.9.3) id WAA19958; Mon, 28 Aug 2000 22:48:34 +0100 From: Russell King - ARM Linux Admin Message-Id: <200008282148.WAA19958@flint.arm.linux.org.uk> Subject: SA1100 DRAM sizes (was Re: switched to strongARM and 160 Mb and Debian 2.2 doesn't boot) To: ebunce@lhsl.com Date: Mon, 28 Aug 2000 22:48:34 +0100 (BST) Cc: bakker@thorgal.et.tudelft.nl (J.D. Bakker), J.A.K.Mouw@its.tudelft.nl (Erik Mouw), france@crl.dec.com, nico@cam.org, chris@blazie.com, a.melis@aino.nl, linux-arm-kernel@lists.arm.linux.org.uk In-Reply-To: <85256949.00788094.00@lhs.com> from "ebunce@lhsl.com" at Aug 28, 2000 05:43:50 PM X-Mailer: ELM [version 2.5 PL1] MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk ebunce@lhsl.com writes: > Just for the record, According to the SA1110 Advance Developer Manual each of > the four SDRAM banks in an SA1110 can contain 128 MB of RAM (using 32M x 8, or > 64M x 16, or 64M x 4 components) for a total of 512MB of memory. I'm not saying > you'd want to do this, but that it is technically possible using commercially > available parts. > > For example: > Using 4 Samsung K4S560832A 32Mx8 component Yields 128MB bank w/32 bit bus > Using 4 Samsung K4S560432A 64Mx4 component > Enjoy, > Erik Yep, full size is 512MB, but with those macros modified as described, you can't use all of that. _____ |_____| ------------------------------------------------- ---+---+- | | Russell King linux@arm.linux.org.uk --- --- | | | | http://www.arm.linux.org.uk/ / / | | +-+-+ --- -+- / | THE developer of ARM Linux |+| /|\ / | | | --- | +-+-+ ------------------------------------------------- /\\\ | unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Mon Aug 28 22:50:32 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13TWn2-0008Dx-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Mon, 28 Aug 2000 22:50:00 +0100 Received: from [208.162.98.43] (helo=lhs.com) by www.linux.org.uk with smtp (Exim 3.13 #1) id 13TWn2-0008Dq-00 for linux-arm-kernel@lists.arm.linux.org.uk; Mon, 28 Aug 2000 22:50:00 +0100 Received: by lhs.com(Lotus SMTP MTA v4.6.6 (890.1 7-16-1999)) id 85256949.00790962 ; Mon, 28 Aug 2000 18:02:02 -0400 X-Lotus-FromDomain: LHS From: ebunce@lhsl.com To: Russell King - ARM Linux Admin cc: bakker@thorgal.et.tudelft.nl (J.D. Bakker), J.A.K.Mouw@its.tudelft.nl (Erik Mouw), france@crl.dec.com, nico@cam.org, chris@blazie.com, a.melis@aino.nl, linux-arm-kernel@lists.arm.linux.org.uk Message-ID: <85256949.00790785.00@lhs.com> Date: Mon, 28 Aug 2000 17:49:36 -0400 Subject: Re: switched to strongARM and 160 Mb and Debian 2.2 doesn't boot Mime-Version: 1.0 Content-type: text/plain; charset=us-ascii Content-Disposition: inline Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk Russell King wrote: > is probably what you were meaning, and yes, it still fits in the 0xc0000000 > to 0xcfffffff space. But what about the vmalloc area, when you've got all > the banks used and full (ie 256MB of memory)? > Just for the record, According to the SA1110 Advance Developer Manual each of the four SDRAM banks in an SA1110 can contain 128 MB of RAM (using 32M x 8, or 64M x 16, or 64M x 4 components) for a total of 512MB of memory. I'm not saying you'd want to do this, but that it is technically possible using commercially available parts. For example: Using 4 Samsung K4S560832A 32Mx8 components yields 128MB bank w/32 bit bus Using 4 Samsung K4S560432A 64Mx4 components yields 128MB bank w/16 bit bus Enjoy, Erik unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Mon Aug 28 23:01:01 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13TWx0-0008PG-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Mon, 28 Aug 2000 23:00:18 +0100 Received: from [130.161.42.1] (helo=duteinh.et.tudelft.nl) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13TWwz-0008P9-00 for linux-arm-kernel@lists.arm.linux.org.uk; Mon, 28 Aug 2000 23:00:17 +0100 Received: from dutein75 by duteinh.et.tudelft.nl (8.9.3/BB-2.0) id AAA15655; Tue, 29 Aug 2000 00:00:08 +0200 (MET DST) Message-Id: <200008282200.AAA15655@duteinh.et.tudelft.nl> From: Erik Mouw To: linux@arm.linux.org.uk Cc: ebunce@lhsl.com, bakker@thorgal.et.tudelft.nl, france@crl.dec.com, nico@cam.org, chris@blazie.com, a.melis@aino.nl, linux-arm-kernel@lists.arm.linux.org.uk Subject: Re: SA1100 DRAM sizes In-Reply-To: <200008282148.WAA19958@flint.arm.linux.org.uk> Organization: Eric conspiracy secret labs Date: Tue 29 Aug 2000 00:00:08 +0100 (MDT) X-Face: "!Z|1XjFg!3Iw)BwM^M\k)J5KHiPddz}#/Zw%?$&Q!xqob"OA|})3_@&BTTiQERo?/^g[Z<9P4R8:5C:aqq.cw!hotLc=+u-"+1rq@9$V^8PbOUlodyi_sejt+D`iqkLpx$>.aXl%vnfPIt"LK|/%>94_Zrm8jy?`_[RKt'TuoxUxfk(gMhY(n|DBBl~c$dhO>G%6}R4+DGX3|idBm4oU&UM/XUP4I9aoN{]`ax%B?rgD(I3OJ<62Y6L X-Mailer: XCmail 1.2devel - with PGP support, PGP engine version 0.5 (IRIX64) X-Mailerorigin: http://www.fsai.fh-trier.de/~schmitzj/Xclasses/XCmail/ X-Eric-Conspiracy: There is no conspiracy! X-Loop: erik@it.et.tudelft.nl MIME-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: 8bit Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk On Mon, 28 Aug 2000 22:48:34 +0100 (BST), Russell King - ARM Linux Admin wrote: > ebunce@lhsl.com writes: >> For example: >> Using 4 Samsung K4S560832A 32Mx8 component Yields 128MB bank w/32 bit bus >> Using 4 Samsung K4S560432A 64Mx4 component >> Enjoy, >> Erik > > Yep, full size is 512MB, but with those macros modified as described, you > can't use all of that. Correct. But if we want to support the full 512MB, we need 0xc0000000 to 0xdfffffff. Vmalloc space should go on top of that at 0xe0000000, but that's already occupied by the SA1100 I/O space (PCMCIA slots and SA1100 registers). Can we just move the memory down to (say) 0xb0000000 0xcfffffff and put vmalloc space at 0xd0000000? Erik [oops, time to leave this building...] -- J.A.K. (Erik) Mouw, Information and Communication Theory Group, Department of Electrical Engineering, Faculty of Information Technology and Systems, Delft University of Technology, PO BOX 5031, 2600 GA Delft, The Netherlands Phone: +31-15-2783635 Fax: +31-15-2781843 Email: J.A.K.Mouw@its.tudelft.nl WWW: http://www-ict.its.tudelft.nl/~erik/ unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Mon Aug 28 23:09:57 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13TX5Z-00004r-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Mon, 28 Aug 2000 23:09:09 +0100 Received: from [63.195.182.101] (helo=barbados.bluemug.com) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13TX5Y-0008WQ-00 for linux-arm-kernel@lists.arm.linux.org.uk; Mon, 28 Aug 2000 23:09:09 +0100 Received: from fiji.bluemug.com [192.168.50.32] by barbados.bluemug.com with esmtp (Exim 2.05 #1 (Debian)) id 13TX5f-0000b6-00; Mon, 28 Aug 2000 15:09:15 -0700 Received: from miket by fiji.bluemug.com with local (Exim 3.12 #1 (Debian)) id 13TX5f-0003og-00; Mon, 28 Aug 2000 15:09:15 -0700 Date: Mon, 28 Aug 2000 15:09:15 -0700 From: Mike Touloumtzis To: Russell King - ARM Linux Admin Cc: linux-arm-kernel@lists.arm.linux.org.uk Subject: Re: switched to strongARM and 160 Mb and Debian 2.2 doesn't boot Message-ID: <20000828150915.B12563@fiji.bluemug.com> References: <200008282107.WAA19802@flint.arm.linux.org.uk> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii In-Reply-To: <200008282107.WAA19802@flint.arm.linux.org.uk>; from Russell King - ARM Linux Admin on Mon, Aug 28, 2000 at 10:07:32PM +0100 Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk On Mon, Aug 28, 2000 at 10:07:32PM +0100, Russell King - ARM Linux Admin wrote: > > These macros are used extensively within the memory management system, and, > if they are slow, you have a slow machine. Therefore, we do everything we > can to make these as fast as possible. Normally, I'd like them to be a > pure phys = virt - offset type arrangement, but the SA1100 platforms are > the only ones which don't allow this. The Cirrus Logic 7110 and descendants have the same issue; I'm currently using the same technique as SA1100 with the EP7211, so I'm following this discussion with interest. The original Psion port used a table lookup for virt_to_phys and phys_to_virt, but I imagine that's a performance nightmare. The EP7211 has two DRAM banks (each 256MB, so they're unlikely to be contiguous), and to add insult to injury, the EDB7211 eval board has not-contiguous RAM within the first bank! The recent kernels also have VALID_PAGE(), which looks like another potential problem. The SA1100 now has: #define VALID_PAGE(page) ((page - mem_map) < max_mapnr) but from looking at the NUMA code and the other macros in mmzone.h, it looks like there could be per-node mem_map arrays, so the above code wouldn't work. Am I missing something? miket unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Mon Aug 28 23:21:04 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13TXGQ-0000DJ-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Mon, 28 Aug 2000 23:20:22 +0100 Received: from dyn-33.linux.theplanet.co.uk ([195.92.244.33] helo=caramon.arm.linux.org.uk) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13TXGN-0000D6-00; Mon, 28 Aug 2000 23:20:20 +0100 Received: from flint.arm.linux.org.uk (root@flint [192.168.0.4]) by caramon.arm.linux.org.uk (8.9.3/8.9.3) with ESMTP id XAA11585; Mon, 28 Aug 2000 23:20:26 +0100 Received: (from linux@localhost) by flint.arm.linux.org.uk (8.9.3/8.9.3) id XAA20092; Mon, 28 Aug 2000 23:19:27 +0100 From: Russell King - ARM Linux Admin Message-Id: <200008282219.XAA20092@flint.arm.linux.org.uk> Subject: Re: SA1100 DRAM sizes To: J.A.K.Mouw@its.tudelft.nl (Erik Mouw) Date: Mon, 28 Aug 2000 23:19:27 +0100 (BST) Cc: ebunce@lhsl.com, bakker@thorgal.et.tudelft.nl, france@crl.dec.com, nico@cam.org, chris@blazie.com, a.melis@aino.nl, linux-arm-kernel@lists.arm.linux.org.uk In-Reply-To: <200008282200.AAA15655@duteinh.et.tudelft.nl> from "Erik Mouw" at Aug 29, 2000 12:00:08 AM X-Mailer: ELM [version 2.5 PL1] MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk Erik Mouw writes: > Correct. But if we want to support the full 512MB, we need 0xc0000000 to > 0xdfffffff. Vmalloc space should go on top of that at 0xe0000000, but > that's already occupied by the SA1100 I/O space (PCMCIA slots and SA1100 > registers). Can we just move the memory down to (say) 0xb0000000 > 0xcfffffff and put vmalloc space at 0xd0000000? You'll need to change both TASK_SIZE and PAGE_OFFSET (TASK_SIZE is the "top" of the 32-bit task space, PAGE_OFFSET is the virtual address of the start of the kernel direct mapped memory). However, I suspect that you don't actually need to keep everything mapped. Eg, On the assabet: 1. The flash should be mapped in as needed by the MTD drivers 2. The scr on the Assabet can be unmapped once its been read. 3. The BCR and SA1111 can be mapped at 0xde000000 and df000000 respectively. With those changes, you've just gained 224MB of extra space. If you also consider that PCMCIA IO normally doesn't need more than 1MB, then that's another 126MB of space. So now we're at 0xe6000000 being the lower end of the mapped IO region. So, we've just gained a total of 350MB, which will cover the 256MB of extra ram space. That leaves you 94MB for vmalloc/ioremap. This may be enough without ioremap, but once you start using it, you really want to free up some more (by making the pcmcia drivers ioremap the minimal space required on an as-needed basis). _____ |_____| ------------------------------------------------- ---+---+- | | Russell King linux@arm.linux.org.uk --- --- | | | | http://www.arm.linux.org.uk/ / / | | +-+-+ --- -+- / | THE developer of ARM Linux |+| /|\ / | | | --- | +-+-+ ------------------------------------------------- /\\\ | unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Mon Aug 28 23:24:37 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13TXJq-0000HQ-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Mon, 28 Aug 2000 23:23:54 +0100 Received: from dyn-33.linux.theplanet.co.uk ([195.92.244.33] helo=caramon.arm.linux.org.uk) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13TXJp-0000HK-00; Mon, 28 Aug 2000 23:23:54 +0100 Received: from flint.arm.linux.org.uk (root@flint [192.168.0.4]) by caramon.arm.linux.org.uk (8.9.3/8.9.3) with ESMTP id XAA11610; Mon, 28 Aug 2000 23:24:01 +0100 Received: (from linux@localhost) by flint.arm.linux.org.uk (8.9.3/8.9.3) id XAA20119; Mon, 28 Aug 2000 23:23:03 +0100 From: Russell King - ARM Linux Admin Message-Id: <200008282223.XAA20119@flint.arm.linux.org.uk> Subject: Re: SA1100 DRAM sizes (was: Re: switched to strongARM and 160 Mb and Debian 2.2 doesn't boot) To: miket@bluemug.com (Mike Touloumtzis) Date: Mon, 28 Aug 2000 23:23:03 +0100 (BST) Cc: linux-arm-kernel@lists.arm.linux.org.uk In-Reply-To: <20000828150915.B12563@fiji.bluemug.com> from "Mike Touloumtzis" at Aug 28, 2000 03:09:15 PM X-Mailer: ELM [version 2.5 PL1] MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk Mike Touloumtzis writes: > The recent kernels also have VALID_PAGE(), which looks like another > potential problem. The SA1100 now has: Yes, it is another problem, but one which (hopefully) the numa people are tackling at this present time (they were looking at it - I've got their private discussion archived here). _____ |_____| ------------------------------------------------- ---+---+- | | Russell King linux@arm.linux.org.uk --- --- | | | | http://www.arm.linux.org.uk/ / / | | +-+-+ --- -+- / | THE developer of ARM Linux |+| /|\ / | | | --- | +-+-+ ------------------------------------------------- /\\\ | unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Mon Aug 28 23:29:09 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13TXOJ-0000KY-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Mon, 28 Aug 2000 23:28:31 +0100 Received: from [192.58.206.2] (helo=crl.dec.com) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13TXOI-0000KS-00 for linux-arm-kernel@lists.arm.linux.org.uk; Mon, 28 Aug 2000 23:28:30 +0100 Received: from ims.crl.dec.com (ims.crl.dec.com [16.11.0.11]) by crl.dec.com (8.8.8/RWD-1.2) with ESMTP id SAA14090; Mon, 28 Aug 2000 18:26:41 -0400 (EDT) Received: by ims.crl.dec.com with Internet Mail Service (5.5.2650.21) id ; Mon, 28 Aug 2000 18:32:33 -0400 Message-ID: From: George France To: "'Erik Mouw'" , linux@arm.linux.org.uk Cc: bakker@thorgal.et.tudelft.nl, George France , nico@cam.org, chris@blazie.com, a.melis@aino.nl, linux-arm-kernel@lists.arm.linux.org.uk Subject: RE: switched to strongARM and 160 Mb and Debian 2.2 doesn't boot Date: Mon, 28 Aug 2000 18:26:36 -0400 MIME-Version: 1.0 X-Mailer: Internet Mail Service (5.5.2650.21) Content-Type: text/plain; charset="iso-8859-1" Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk Thank you, Erik, The patch works. :-) Best Regards, --George George France, france@crl.dec.com Cambridge Research Laboratory, Compaq Computer Corporation One Cambridgecenter MS: CRL Cambridge, MA 02142 USA > -----Original Message----- > From: Erik Mouw [mailto:J.A.K.Mouw@its.tudelft.nl] > Sent: Monday, August 28, 2000 6:21 PM > To: linux@arm.linux.org.uk > Cc: bakker@thorgal.et.tudelft.nl; france@crl.dec.com; nico@cam.org; > chris@blazie.com; a.melis@aino.nl; > linux-arm-kernel@lists.arm.linux.org.uk > Subject: Re: switched to strongARM and 160 Mb and Debian 2.2 doesn't > boot > > > On Mon, 28 Aug 2000 21:42:48 +0100 (BST), Russell King - ARM > Linux Admin wrote: > > J.D. Bakker writes: > > Not quite correct, but: > > > > #define __virt_to_phys(x) (((x) & 0xf3ffffff) | ((x) & > 0x0c000000) << 1) > > #define __phys_to_virt(x) (((x) & 0xe7ffffff) | ((x) & > 0x18000000) >> 1) > > > > is probably what you were meaning, and yes, it still fits > in the 0xc0000000 > > to 0xcfffffff space. But what about the vmalloc area, when > you've got all > > the banks used and full (ie 256MB of memory)? > > This patch allows 64MB/bank; tested on the LART, YMMV. > Oh, and #include . > > -------------------------------------------------------------------- > diff -urN > linux-2.4.0-test6-rmk5-np2-vanilla/include/asm-arm/arch-sa1100 > /memory.h > linux-2.4.0-test6-rmk5-np2/include/asm-arm/arch-sa1100/memory.h > --- > linux-2.4.0-test6-rmk5-np2-vanilla/include/asm-arm/arch-sa1100 > /memory.h Thu Jan 13 22:30:31 2000 > +++ > linux-2.4.0-test6-rmk5-np2/include/asm-arm/arch-sa1100/memory. > h Mon Aug 28 23:02:36 2000 > @@ -29,10 +29,11 @@ > #define __phys_to_virt__is_a_macro > > /* > - * The following gives a maximum memory size of 128MB (32MB > in each bank). > + * The following gives a maximum memory size of 256MB (64MB > in each bank). > */ > -#define __virt_to_phys(x) (((x) & 0xf9ffffff) | ((x) & > 0x06000000) << 2) > -#define __phys_to_virt(x) (((x) & 0xe7ffffff) | ((x) & > 0x18000000) >> 2) > +#define __virt_to_phys(x) (((x) & 0xf3ffffff) | ((x) & > 0x0c000000) << 1) > +#define __phys_to_virt(x) (((x) & 0xe7ffffff) | ((x) & > 0x18000000) >> 1) > + > > /* > * Virtual view <-> DMA view memory address translations > diff -urN > linux-2.4.0-test6-rmk5-np2-vanilla/include/asm-arm/arch-sa1100 > /mmzone.h > linux-2.4.0-test6-rmk5-np2/include/asm-arm/arch-sa1100/mmzone.h > --- > linux-2.4.0-test6-rmk5-np2-vanilla/include/asm-arm/arch-sa1100 > /mmzone.h Thu Aug 24 12:52:38 2000 > +++ > linux-2.4.0-test6-rmk5-np2/include/asm-arm/arch-sa1100/mmzone. > h Mon Aug 28 23:02:37 2000 > @@ -20,9 +20,9 @@ > extern pg_data_t sa1100_node_data[]; > > /* > - * 32MB max in each bank, must fit with __virt_to_phys() & > __phys_to_virt() > + * 64MB max in each bank, must fit with __virt_to_phys() & > __phys_to_virt() > */ > -#define NODE_MAX_MEM_SHIFT 25 > +#define NODE_MAX_MEM_SHIFT 26 > #define NODE_MAX_MEM_SIZE (1< > /* > -------------------------------------------------------------------- > > > You need to move stuff at 0xd0000000 out of the way at least. > > OK, so where is a good place? (Is the memory map on ARM linux > somewhere > defined?) > > > Erik > > -- > J.A.K. (Erik) Mouw, Information and Communication Theory > Group, Department > of Electrical Engineering, Faculty of Information Technology > and Systems, > Delft University of Technology, PO BOX 5031, 2600 GA Delft, > The Netherlands > Phone: +31-15-2783635 Fax: +31-15-2781843 Email: > J.A.K.Mouw@its.tudelft.nl > WWW: http://www-ict.its.tudelft.nl/~erik/ > > > > > > unsubscribe: body of `unsubscribe linux-arm-kernel' to > majordomo@lists.arm.linux.org.uk > unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Tue Aug 29 00:34:12 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13TYOs-00015I-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Tue, 29 Aug 2000 00:33:10 +0100 Received: from [63.195.182.101] (helo=barbados.bluemug.com) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13TYOr-00015C-00 for linux-arm-kernel@lists.arm.linux.org.uk; Tue, 29 Aug 2000 00:33:09 +0100 Received: from fiji.bluemug.com [192.168.50.32] by barbados.bluemug.com with esmtp (Exim 2.05 #1 (Debian)) id 13TYP1-0000dB-00; Mon, 28 Aug 2000 16:33:19 -0700 Received: from miket by fiji.bluemug.com with local (Exim 3.12 #1 (Debian)) id 13TYOt-00044c-00; Mon, 28 Aug 2000 16:33:11 -0700 Date: Mon, 28 Aug 2000 16:33:11 -0700 From: Mike Touloumtzis To: linux-arm-kernel@lists.arm.linux.org.uk Subject: FLUSH_BASE and FLUSH_BASE_PHYS Message-ID: <20000828163311.C12563@fiji.bluemug.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk Hi, Could anyone tell me the purpose of the mapping which uses FLUSH_BASE and FLUSH_BASE_PHYS? I'd very much appreciate it. thanks, miket unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Tue Aug 29 02:52:49 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13TaYE-0002cf-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Tue, 29 Aug 2000 02:50:58 +0100 Received: from [205.250.170.220] (helo=xanadu.vipswitch.com) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13TaYB-0002cW-00 for linux-arm-kernel@lists.arm.linux.org.uk; Tue, 29 Aug 2000 02:50:56 +0100 Date: Mon, 28 Aug 2000 21:46:31 -0400 (EDT) From: Nicolas Pitre To: Russell King - ARM Linux Admin cc: John G Dorsey , Erik Mouw , ebunce@lhsl.com, bakker@thorgal.et.tudelft.nl, france@crl.dec.com, chris@blazie.com, a.melis@aino.nl, linux-arm-kernel@lists.arm.linux.org.uk Subject: Re: SA1100 DRAM sizes In-Reply-To: <200008282219.XAA20092@flint.arm.linux.org.uk> Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk [To John G. Dorsey: If you're listening, I would like your coments on the PCMCIA stuff below.] Woohoooo! I got my inbox flodded... ;) I wanted to get home and have a look at the code before suggesting the requested quick hack... but apparently Erik got it first. On Mon, 28 Aug 2000, Russell King - ARM Linux Admin wrote: > Erik Mouw writes: > > Correct. But if we want to support the full 512MB, we need 0xc0000000 to > > 0xdfffffff. Vmalloc space should go on top of that at 0xe0000000, but > > that's already occupied by the SA1100 I/O space (PCMCIA slots and SA1100 > > registers). Can we just move the memory down to (say) 0xb0000000 > > 0xcfffffff and put vmalloc space at 0xd0000000? > > You'll need to change both TASK_SIZE and PAGE_OFFSET (TASK_SIZE is the > "top" of the 32-bit task space, PAGE_OFFSET is the virtual address of > the start of the kernel direct mapped memory). IMHO that's not a good idea. > However, I suspect that you don't actually need to keep everything > mapped. Eg, On the assabet: > > 1. The flash should be mapped in as needed by the MTD drivers Indeed. > 2. The scr on the Assabet can be unmapped once its been read. The scr is in reality IO pins which aren't in the way... and not desirable to unmap at all. > 3. The BCR and SA1111 can be mapped at 0xde000000 and df000000 > respectively. The cache invalidation area is already there... and this doesn't solve the 512MB possibility either. Well... here's the plan: virt_to_phys()/phys_to_virt() become: #define __virt_to_phys(x) (x) #define __phys_to_virt(x) (x) That gcc knows how to optimize! ;-) Next the PCMCIA 0 and 1 (IO only) are mapped at 0xf6000000 and 0xf7000000. This should cover all designs with hardwired devices i.e. straight IDE interface, ehternet, etc. (16MB for each should be more than enough, especially since this usually uses ISA component which are used to a very tiny address space). The integrated peripherals mapping from 0xf8000000 stays unchanged. Cache invalidation area goes to 0xf5000000 and 0xf5800000. And... all guizmos like flash, SA-1111, CPLDs, etc go between 0xf0000000 to 0xf4ffffff... and it could possibly be grown downwards if needed. RAM is now from 0xc0000000 to 0xdfffffff, the same as physically. This leaves 0xe0000000 to 0xefffffff for vm areas. That's it! **HOWEVER** this requires the generic SA1100 PCMCIA driver to be modified so it actually ioremap() its memory and attribute space (even IO could be done that way too for more clarity). This change must occur before all the rest otherwise PCMCIA will be broken and people will step on my toes. John --> You have code that did just that in the past. Could it be done quickly? In the mean time I'll include the 64MB change... but hopefully the real solution will come sooner than later, otherwise someone will certainly come by with a 128MB per bank design... Any comments? Nicolas unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Tue Aug 29 03:02:15 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13TaiV-0002kA-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Tue, 29 Aug 2000 03:01:35 +0100 Received: from [24.200.137.248] (helo=xanadu.gn.com) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13TaiT-0002k2-00 for linux-arm-kernel@lists.arm.linux.org.uk; Tue, 29 Aug 2000 03:01:34 +0100 Received: from localhost (nico@localhost) by xanadu.gn.com (8.9.3/8.9.3) with ESMTP id WAA29298; Mon, 28 Aug 2000 22:00:41 -0400 X-Authentication-Warning: xanadu.gn.com: nico owned process doing -bs Date: Mon, 28 Aug 2000 22:00:41 -0400 (EDT) From: Nicolas Pitre X-Sender: nico@xanadu.gn.com To: Mike Touloumtzis cc: linux-arm-kernel@lists.arm.linux.org.uk Subject: Re: FLUSH_BASE and FLUSH_BASE_PHYS In-Reply-To: <20000828163311.C12563@fiji.bluemug.com> Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk On Mon, 28 Aug 2000, Mike Touloumtzis wrote: > > Hi, > > Could anyone tell me the purpose of the mapping which uses > FLUSH_BASE and FLUSH_BASE_PHYS? I'd very much appreciate > it. See linux/arch/arm/mm/proc-*.S. It is used to fill the cache with junk so actual dirty data is evicted from the cache. Yes, that's the only way to write back the whole cache on the StrongARM. Nicolas unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Tue Aug 29 03:29:59 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13Tb9P-00038V-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Tue, 29 Aug 2000 03:29:23 +0100 Received: from [128.2.10.81] (helo=smtp1.andrew.cmu.edu) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13Tb9O-000385-00 for linux-arm-kernel@lists.arm.linux.org.uk; Tue, 29 Aug 2000 03:29:22 +0100 Received: from cs.cmu.edu (NINESIXTEEN.REM.CMU.EDU [128.2.81.157]) by smtp1.andrew.cmu.edu (8.9.3/8.9.3) with ESMTP id WAA29177; Mon, 28 Aug 2000 22:24:07 -0400 (EDT) Message-ID: <39AB1E6F.29A4055@cs.cmu.edu> Date: Mon, 28 Aug 2000 22:22:39 -0400 From: John G Dorsey Organization: Carnegie Mellon University X-Mailer: Mozilla 4.72 [en] (X11; U; Linux 2.2.14-5.0 i686) X-Accept-Language: en MIME-Version: 1.0 To: Nicolas Pitre CC: Russell King - ARM Linux Admin , Erik Mouw , ebunce@lhsl.com, bakker@thorgal.et.tudelft.nl, france@crl.dec.com, chris@blazie.com, a.melis@aino.nl, linux-arm-kernel@lists.arm.linux.org.uk Subject: Re: SA1100 DRAM sizes References: Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk Nicolas Pitre wrote: > **HOWEVER** this requires the generic SA1100 PCMCIA driver to be modified > so it actually ioremap() its memory and attribute space (even IO could be > done that way too for more clarity). This change must occur before all > the rest otherwise PCMCIA will be broken and people will step on my toes. > > John --> You have code that did just that in the past. Could it be done > quickly? I'm on it. (Might as well just do all three per-slot spaces on the fly, as Card Services is already issuing the setup calls for them. We just configure timings now and no-op the actual allocation request.) > In the mean time I'll include the 64MB change... but hopefully the real > solution will come sooner than later, otherwise someone will certainly > come by with a 128MB per bank design... Someone (me) will come by in late October with such a design. It passed final engineering review last week. =) -jd unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Tue Aug 29 05:54:50 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13TdO5-0004j7-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Tue, 29 Aug 2000 05:52:41 +0100 Received: from [210.230.34.106] (helo=so-fs.musicpod.co.jp) by www.linux.org.uk with smtp (Exim 3.13 #1) id 13TdO4-0004iy-00 for linux-arm-kernel@lists.arm.linux.org.uk; Tue, 29 Aug 2000 05:52:41 +0100 Received: (qmail 23673 invoked from network); 29 Aug 2000 04:52:43 -0000 Received: from bak.side7.zetabits.com (imai@192.168.36.68) by so-fs.musicpod.co.jp with SMTP; 29 Aug 2000 04:52:43 -0000 Date: Tue, 29 Aug 2000 13:52:42 +0900 Message-ID: From: Kunihiko IMAI To: linux-arm-kernel@lists.arm.linux.org.uk Subject: RE: Kunihiko Imais USB ohci driver doesn't work? In-Reply-To: In your message of "Mon, 28 Aug 2000 17:38:58 +0200" References: User-Agent: Wanderlust/1.1.1 (Purple Rain) WEMI/1.13.7 (Shimada) CLIME/1.13.6 (=?ISO-2022-JP?B?GyRCQ2YlTj4xGyhC?=) MULE XEmacs/21.1 (patch 6) (Big Bend) (i386-vine-linux) MIME-Version: 1.0 (generated by WEMI 1.13.7 - "Shimada") Content-Type: text/plain; charset=US-ASCII Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk Hello, At Mon, 28 Aug 2000 17:38:58 +0200, Arjan Opmeer (EMN) wrote: > [ snip above 1MB memory DMA transfer problem ] > > Yes, I heard about this one. But Kunihiko Imai seems to have spoken to > a circuit designer who told him that there was a problem with a SA1111 > DMA line that "now" (?) should be fixed. The bug seems to interfere > with DMA bus mastering. He also says that he suffered from the bug for > a month... (and then probably had his hardware fixed?). The problem which I wrote before is another problem. It is interconnection between SA1110 and SA1111. The 6/30/00 errata of SA1111 has not fixed yet. This bug exists internal of SA1111 chip. But the errata only points the POSSIBILITY of lost data. So some extent, DMA of SA1111 can work. By the way do you fix usb-ohci-sa1111.h of my patch? If you use with neponset, this part: #ifdef CONFIG_SA1100_XP860 #define PwrSensePolLow 1 #define PwrCtrlPolLow 1 #else #define PwrSensePolLow 0 #define PwrCtrlPolLow 0 #endif should be #if defined(CONFIG_SA1100_XP860) || defined(CONFIG_ASSABET_NEPONSET) #define PwrSensePolLow 1 #define PwrCtrlPolLow 1 #else #define PwrSensePolLow 0 #define PwrCtrlPolLow 0 #endif I compared schematics of XP860 and neponset, they has same polarity of these pins. # The descriptions of these pins on SA1111 manual confuse us, I think. # It should be written more clear. Thanks. _._. __._ _.. . .___ ._. _____ _... ._ _._ _.._. .____ .__. ... . _._ Kunihiko IMAI unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Tue Aug 29 05:55:56 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13TdQ8-0004k7-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Tue, 29 Aug 2000 05:54:48 +0100 Received: from [207.245.35.202] (helo=kei.netwinder.org) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13TdQ7-0004jz-00 for linux-arm-kernel@lists.arm.linux.org.uk; Tue, 29 Aug 2000 05:54:47 +0100 Received: from netbackup ([10.1.16.35]) by kei.netwinder.org (8.9.3/8.9.3) with SMTP id AAA29088; Tue, 29 Aug 2000 00:54:43 -0400 Message-ID: <010b01c01175$3d8758e0$LocalHost@netbackup> From: "Scott Bambrough" To: "Chris Blazie" , "Russell King - ARM Linux Admin" Cc: "Angelo" , References: Subject: Re: switched to strongARM and 160 Mb and Debian 2.2 doesn't boot Date: Tue, 29 Aug 2000 00:54:38 -0400 MIME-Version: 1.0 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: 7bit X-Priority: 3 X-MSMail-Priority: Normal X-Mailer: Microsoft Outlook Express 5.00.2314.1300 X-MimeOLE: Produced By Microsoft MimeOLE V5.00.2314.1300 Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk > > > As far as I know, memory sizes greater than 32MB are not supported > > > in the Strongarm kernel right now. You can verify this by modifying > > > arch/arm/arch.c and arch/arm/setup.c to declare only 32MB and see if it > > > works. Our hardware works when "detuned" like that; at some point I'm > > > planning to take a look at the problem if it's still there. Huh? We have had NetWinder's running with 256 MB of RAM. We regularly ship machines with 128 MB. It's StrongARM based. Scott unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Tue Aug 29 07:10:56 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13Teal-0005YI-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Tue, 29 Aug 2000 07:09:51 +0100 Received: from [24.200.137.248] (helo=xanadu.gn.com) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13Teak-0005YC-00 for linux-arm-kernel@lists.arm.linux.org.uk; Tue, 29 Aug 2000 07:09:50 +0100 Received: from localhost (nico@localhost) by xanadu.gn.com (8.9.3/8.9.3) with ESMTP id CAA31451 for ; Tue, 29 Aug 2000 02:09:09 -0400 X-Authentication-Warning: xanadu.gn.com: nico owned process doing -bs X-Received: from localhost (IDENT:nico@localhost [127.0.0.1]) by xanadu.gn.com (8.9.3/8.9.3) with ESMTP id CAA31440 for ; Tue, 29 Aug 2000 02:03:50 -0400 X-Received: from hydro.cam.org by localhost with POP3 (fetchmail-5.0.0) for nico@localhost (single-drop); Tue, 29 Aug 2000 02:03:54 -0400 (EDT) X-Received: from atom.kaist.ac.kr (smslab.kaist.ac.kr [143.248.147.115]) by cogito.cam.org (8.11.0/8.11.0.Beta1) with ESMTP id e7T46xs15978 for ; Tue, 29 Aug 2000 00:06:59 -0400 X-Received: from titan (titan.kaist.ac.kr [143.248.159.120]) by atom.kaist.ac.kr (8.9.3/8.9.3) with SMTP id NAA03867 for ; Tue, 29 Aug 2000 13:05:00 +0900 (KST) Message-ID: <002101c0116e$a3ebaca0$789ff88f@titan> From: À̿Ϻ¹ To: Subject: linux porting problem to EBSA-285 Date: Tue, 29 Aug 2000 13:07:25 +0900 MIME-Version: 1.0 Content-Type: multipart/alternative; boundary="----=_NextPart_000_001E_01C011BA.13AD2F00" X-Priority: 3 X-MSMail-Priority: Normal X-Mailer: Microsoft Outlook Express 5.00.2919.6700 X-MimeOLE: Produced By Microsoft MimeOLE V5.00.2919.6700 X-UIDL: :b,"!MjQ!!olL!![=D!! Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk This is a multi-part message in MIME format. ------=_NextPart_000_001E_01C011BA.13AD2F00 Content-Type: text/plain; charset="ks_c_5601-1987" Content-Transfer-Encoding: base64 SGksIA0KSSBoYXZlIGEgRUJTQS0yODUgYm9hcmQsIG9udG8gaXQgSSB0cnllZCB0byBwb3J0IHRo ZSBsaW51eCAyLjIuMTMsIHlldCBmYWlsZWQgbXkgcG9ydGluZyB0YXNrLg0KU2luY2UsIEkgaGF2 ZSBvbmx5IHRoZSBib2FyZCBFQlNBLTI4NSB3aXRob3V0IGl0cyBiYWNrcGxhbmUgYm9hcmQsIEkg d2FudCB0byBiZSBjb21maXJtZWQNCndoZXRoZXIgSSBjb3VsZCBsb2FkIHRoZSBsaW51eCBrZXJu ZWwgYW5kIHJ1biBvbiBpdCB3aXRob3V0IHRoZSBiYWNrcGxhbmUgYm9hcmQuDQpJZiB5b3UgcmVw bHkgdGhlIGFuc3dlciB0byBtZSwgSSB3aWxsIGJlIGdyZWF0bHkgYXByZWNpYXRlZC4NCg0KSSBq dXN0IHNldHRlZCB0aGUgRUJTQS0yODUgYm9hcmQgYXMgaG9zdCBtb2RlLCBhbmQgcHV0IGl0IGlu dG8gdGhlIFBDSSBzbG90IG9mIG15IHBlbnRpdW1tLUlJIGNsYXNzIFBDLiBUaGUgRUJTQS0yODUg d2FzIHN1cHBsaWVkIGl0cyBwb3dlciBhbmQgUENJIGNsb2NrIGZyb20gdGhlIFBDLg0KQWZ0ZXIg Y3Jvc3MgY29tcGlsaW5nIHRoZSBsaW51eCBrZXJuZWwsIHRoZSBkb3dubG9hZGluZyBwcm9jZXNz IHdhcyBjb21wbGV0ZWQgc3VjY2Vzc2Z1bHkgdXNpbmcNCnRoZSBhbmdlbGJvb3QgcGFja2FnZSBm cm9tIGEgbGludXggaG9zdCBtYWNoaW5lLg0KQnV0LCBubyBib290aW5nIHVwIG1lc3NhZ2Ugd2Fz IHNob3duLiBJIGFtIHdvbmRlcmluZyBub3cgaWYgaXQgaXMgcG9zc2libGUgdG8gYm9vdCB1cCB0 aGUgRUJTQS0yODUgYm9hcmQgYWxvbmUgYXQgYWxsLiANClBsZWFzZSBoZWxwIG1lLiBnaXZlIG1l IGEgcmVzcG9uc2UgcGxlYXNlLg0KWW91ciB0cml2aWFsIGhlbHAgd291bGQgYmUgYSBncmVhdCBh c3Npc3RhbmNlIHRvIG1lLg0KDQpzaW5jZXJlbHkgd2FpdGluZyB5b3VyIHJlcGx5LCANCldhbmJv ayBMZWUsIGZyb20gU291dGgtS29yZWENCndibGVlQHNtc2xhYi5rYWlzdC5hYy5rcg0KDQpQLlMu IEkgYWxzbyB3YW50IHRvIGtub3cgaG93IHRvIGNhdGNoIHRoZSBjb25zb2xlIG1lc3NhZ2Ugb2Yg dGhlIGxpbnV4IG9uIEVCU0EtMjg1Lg0KDQo= ------=_NextPart_000_001E_01C011BA.13AD2F00 Content-Type: text/html; charset="ks_c_5601-1987" Content-Transfer-Encoding: base64 PCFET0NUWVBFIEhUTUwgUFVCTElDICItLy9XM0MvL0RURCBIVE1MIDQuMCBUcmFuc2l0aW9uYWwv L0VOIj4NCjxIVE1MPjxIRUFEPg0KPE1FVEEgY29udGVudD0idGV4dC9odG1sOyBjaGFyc2V0PWtz X2NfNTYwMS0xOTg3IiBodHRwLWVxdWl2PUNvbnRlbnQtVHlwZT4NCjxNRVRBIGNvbnRlbnQ9Ik1T SFRNTCA1LjAwLjI5MjAuMCIgbmFtZT1HRU5FUkFUT1I+DQo8U1RZTEU+PC9TVFlMRT4NCjwvSEVB RD4NCjxCT0RZIGJnQ29sb3I9I2Q0ZDBjOD4NCjxESVY+PEZPTlQgc2l6ZT0yPjxGT05UIHNpemU9 Mj4NCjxESVY+PEZPTlQgc2l6ZT0yPkhpLCA8L0ZPTlQ+PC9ESVY+DQo8RElWPjxGT05UIHNpemU9 Mj5JIGhhdmUgYSBFQlNBLTI4NSBib2FyZCwgb250byBpdCBJIHRyeWVkIHRvIHBvcnQgdGhlIGxp bnV4IA0KMi4yLjEzLCB5ZXQgZmFpbGVkIG15IHBvcnRpbmcgdGFzay48L0ZPTlQ+PC9ESVY+DQo8 RElWPjxGT05UIHNpemU9Mj5TaW5jZSwgSSBoYXZlIG9ubHkgdGhlIGJvYXJkIEVCU0EtMjg1IHdp dGhvdXQgaXRzIGJhY2twbGFuZSANCmJvYXJkLCBJIHdhbnQgdG8gYmUgY29tZmlybWVkPC9GT05U PjwvRElWPg0KPERJVj48Rk9OVCBzaXplPTI+d2hldGhlciBJIGNvdWxkIGxvYWQgdGhlIGxpbnV4 IGtlcm5lbCBhbmQgcnVuIG9uIGl0IHdpdGhvdXQgDQp0aGUgYmFja3BsYW5lIGJvYXJkLjwvRk9O VD48L0RJVj4NCjxESVY+PEZPTlQgc2l6ZT0yPklmIHlvdSByZXBseSB0aGUgYW5zd2VyIHRvIG1l LCBJIHdpbGwgYmUgZ3JlYXRseSANCmFwcmVjaWF0ZWQuPC9GT05UPjwvRElWPg0KPERJVj4mbmJz cDs8L0RJVj4NCjxESVY+PEZPTlQgc2l6ZT0yPkkganVzdCBzZXR0ZWQgdGhlIEVCU0EtMjg1IGJv YXJkIGFzIGhvc3QgbW9kZSwgYW5kIHB1dCBpdCBpbnRvIA0KdGhlIFBDSSBzbG90IG9mIG15IHBl bnRpdW1tLUlJIDwvRk9OVD48Rk9OVCBzaXplPTI+Y2xhc3MgUEMuIDwvRk9OVD48Rk9OVCANCnNp emU9Mj5UaGUgRUJTQS0yODUgd2FzIHN1cHBsaWVkIGl0cyBwb3dlciBhbmQgUENJIGNsb2NrIGZy b20gdGhlIA0KUEMuPC9GT05UPjwvRElWPg0KPERJVj48Rk9OVCBzaXplPTI+QWZ0ZXIgY3Jvc3Mg Y29tcGlsaW5nIHRoZSBsaW51eCBrZXJuZWwsIHRoZSBkb3dubG9hZGluZyANCnByb2Nlc3Mgd2Fz IGNvbXBsZXRlZCBzdWNjZXNzZnVseSB1c2luZzwvRk9OVD48L0RJVj4NCjxESVY+PEZPTlQgc2l6 ZT0yPnRoZSBhbmdlbGJvb3QgcGFja2FnZSBmcm9tIGEgbGludXggaG9zdCBtYWNoaW5lLjwvRk9O VD48L0RJVj4NCjxESVY+PEZPTlQgc2l6ZT0yPkJ1dCwgbm8gYm9vdGluZyB1cCBtZXNzYWdlIHdh cyBzaG93bi4gSSBhbSB3b25kZXJpbmcgbm93IGlmIGl0IA0KaXMgcG9zc2libGUgdG8gYm9vdCB1 cCB0aGUgRUJTQS0yODUgYm9hcmQgYWxvbmUgYXQgYWxsLiA8L0ZPTlQ+PC9ESVY+DQo8RElWPjxG T05UIHNpemU9Mj5QbGVhc2UgaGVscCBtZS4gZ2l2ZSBtZSBhIHJlc3BvbnNlIHBsZWFzZS48L0ZP TlQ+PC9ESVY+DQo8RElWPjxGT05UIHNpemU9Mj5Zb3VyIHRyaXZpYWwgaGVscCB3b3VsZCBiZSBh IGdyZWF0IGFzc2lzdGFuY2UgdG8gDQptZS48L0ZPTlQ+PC9ESVY+DQo8RElWPiZuYnNwOzwvRElW Pg0KPERJVj48Rk9OVCBzaXplPTI+c2luY2VyZWx5IHdhaXRpbmcgeW91ciByZXBseSwgPC9GT05U PjwvRElWPg0KPERJVj48Rk9OVCBzaXplPTI+V2FuYm9rIExlZSwgZnJvbSBTb3V0aC1Lb3JlYTwv Rk9OVD48L0RJVj4NCjxESVY+PEZPTlQgc2l6ZT0yPjxBIA0KaHJlZj0ibWFpbHRvOndibGVlQHNt c2xhYi5rYWlzdC5hYy5rciI+d2JsZWVAc21zbGFiLmthaXN0LmFjLmtyPC9BPjwvRk9OVD48L0RJ Vj4NCjxESVY+Jm5ic3A7PC9ESVY+DQo8RElWPjxGT05UIHNpemU9Mj5QLlMuIEkgYWxzbyB3YW50 IHRvIGtub3cgaG93IHRvIGNhdGNoIHRoZSBjb25zb2xlIG1lc3NhZ2Ugb2YgDQp0aGUgbGludXgg b24gRUJTQS0yODUuPC9GT05UPjwvRElWPg0KPERJVj4mbmJzcDs8L0RJVj48L0ZPTlQ+PC9GT05U PjwvRElWPjwvQk9EWT48L0hUTUw+DQo= ------=_NextPart_000_001E_01C011BA.13AD2F00-- unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Tue Aug 29 08:15:43 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13TfbP-0006F4-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Tue, 29 Aug 2000 08:14:35 +0100 Received: from [202.130.151.66] (helo=gw-hk1.philips.com) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13TfbN-0006Eu-00 for linux-arm-kernel@lists.arm.linux.org.uk; Tue, 29 Aug 2000 08:14:34 +0100 Received: from smtprelay-hk1.philips.com (localhost.philips.com [127.0.0.1]) by gw-hk1.philips.com with ESMTP id PAA18953 for ; Tue, 29 Aug 2000 15:14:39 +0800 (HKT) (envelope-from peter.mc.chen@philips.com) From: peter.mc.chen@philips.com Received: from smtprelay-asp1.philips.com(130.147.65.5) by gw-hk1.philips.com via mwrap (4.0a) id xma018951; Tue, 29 Aug 00 15:14:39 +0800 Received: from APLMS01.DIAMOND.PHILIPS.COM (aplms01sv1.diamond.philips.com [130.147.79.213]) by smtprelay-hk1.philips.com (8.9.3/8.8.5-1.2.2m-19990317) with ESMTP id PAA20787 for ; Tue, 29 Aug 2000 15:14:38 +0800 (HKT) Received: by APLMS01.DIAMOND.PHILIPS.COM (Soft-Switch LMS 4.0) with snapi via APAC id 0056920007226016; Tue, 29 Aug 2000 15:13:43 +0800 To: Subject: porting linux to SA1110 Assabet? Message-ID: <0056920007226016000002L262*@MHS> Date: Tue, 29 Aug 2000 15:13:43 +0800 MIME-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1; name="MEMO 08/29/00 15:13:29" Content-Transfer-Encoding: quoted-printable Content-Disposition: inline Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk Hi : Many people base on 2.4.0-test2-ac1-rmk2-np1 kernel, Can I ask a very = simple question. What is the meaning of this version name? (I mean ac1,= rmk2, or np1) And where can I get one(sorry I have spend time finding = such kernel, but till now only=20 2.4.0-test1-ac7-rmk1-np6-hh1 I can find) Cheers Peter Chen.= unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Tue Aug 29 10:45:26 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13Thum-0007nn-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Tue, 29 Aug 2000 10:42:44 +0100 Received: from dyn-33.linux.theplanet.co.uk ([195.92.244.33] helo=caramon.arm.linux.org.uk) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13Thuf-0007nV-00; Tue, 29 Aug 2000 10:42:37 +0100 Received: from flint.arm.linux.org.uk (root@flint [192.168.0.4]) by caramon.arm.linux.org.uk (8.9.3/8.9.3) with ESMTP id JAA13392; Tue, 29 Aug 2000 09:49:41 +0100 Received: (from linux@localhost) by flint.arm.linux.org.uk (8.9.3/8.9.3) id JAA21825; Tue, 29 Aug 2000 09:48:40 +0100 From: Russell King - ARM Linux Admin Message-Id: <200008290848.JAA21825@flint.arm.linux.org.uk> Subject: Improvements? (was Re: switched to strongARM and 160 Mb and Debian 2.2 doesn't boot) To: scottb@netwinder.org (Scott Bambrough) Date: Tue, 29 Aug 2000 09:48:40 +0100 (BST) Cc: linux-arm-kernel@lists.arm.linux.org.uk In-Reply-To: <010b01c01175$3d8758e0$LocalHost@netbackup> from "Scott Bambrough" at Aug 29, 2000 12:54:38 AM X-Mailer: ELM [version 2.5 PL1] MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk Scott Bambrough writes: > Huh? We have had NetWinder's running with 256 MB of RAM. We regularly ship > machines with 128 MB. It's StrongARM based. The original mail was about a RiscPC: | I have debian 2.2 running for a few weeks on my RISC PC 700, with an ARM 7 | and 64 Mb (+2 Mb VRAM). This worked rather well (did some recompiling and | stuff). Can people please remember that, with us supporting some 40 different types of platform, with about 20 of them being "StrongARM" based, that when someone talks about a processor type (ie, ARM7, StrongARM, ARM920T) that they are not necessarily talking about your particular platform. It may help also if the platform is mentioned in the subject line? Ideas for improvement of quality welcome. _____ |_____| ------------------------------------------------- ---+---+- | | Russell King linux@arm.linux.org.uk --- --- | | | | http://www.arm.linux.org.uk/ / / | | +-+-+ --- -+- / | THE developer of ARM Linux |+| /|\ / | | | --- | +-+-+ ------------------------------------------------- /\\\ | unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Tue Aug 29 11:09:26 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13TiJX-00085P-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Tue, 29 Aug 2000 11:08:19 +0100 Received: from [195.64.48.18] (helo=imail.knoware.nl) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13TiJW-00085J-00 for linux-arm-kernel@lists.arm.linux.org.uk; Tue, 29 Aug 2000 11:08:18 +0100 Received: from mail.knoware.nl (mail.knoware.nl [195.64.48.17]) by imail.knoware.nl (Postfix) with ESMTP id 675C9BD677 for ; Tue, 29 Aug 2000 12:08:29 +0200 (CEST) Received: from avxsrv04.cypres.nl (cypr-45-228.cypres.nl [195.64.45.228]) by mail.knoware.nl (Postfix) with ESMTP id C6DD7A6CF0 for ; Tue, 29 Aug 2000 12:08:15 +0200 (CEST) Received: by avxsrv04.cypres.nl with Internet Mail Service (5.5.2650.21) id ; Tue, 29 Aug 2000 12:09:51 +0200 Message-ID: <8BD00E18F88DD3118EC8005004258990DCEC07@avxsrv04.cypres.nl> From: "Melis, A. (Angelo)" To: "'linux-arm-kernel@lists.arm.linux.org.uk'" Subject: RE: Improvements? (was Re: switched to strongARM and 160 Mb and D ebian 2.2 doesn't boot) Date: Tue, 29 Aug 2000 12:09:51 +0200 MIME-Version: 1.0 X-Mailer: Internet Mail Service (5.5.2650.21) Content-Type: text/plain; charset="iso-8859-1" Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk Hi, > > Scott Bambrough writes: > > Huh? We have had NetWinder's running with 256 MB of RAM. We regularly ship > > machines with 128 MB. It's StrongARM based. > > The original mail was about a RiscPC: Correct. I send the original message :). The platform is an RISC PC 700 with a StrongARM 200 Mhz and 128+32MB RAM and 2 MB VRAM. > > | I have debian 2.2 running for a few weeks on my RISC PC 700, with an ARM 7 > | and 64 Mb (+2 Mb VRAM). This worked rather well (did some recompiling and > | stuff). > Status at the moment is that I did get it to boot debian 2.2 the day before yesterday. What I did was swapping the SIMMS around. I put the 128MB (2 banks of 64Mb) in bank 0 and the 32MB (2 banks of 16MB) in bank 1. In this configuration the system booted. But I did get an error with fsck which hanged after 88.8% of the check. After booting my other partition I got to the login prompt and tried to do the fsck process by hand which gave the error that the runtime linking of something failed. The computer hanged and I resetted and didn't look at at further. But yesterday I tried to reboot into the partition and ofcourse fsck started on that partition also. It did finish the process without any trouble, but after that there where a lot of errors about malloc/free failing and core dumps etc. So there is still something very wrong. I will try to reinstall everything again and see what happens ... Bye and thanks for all the help from everyone, Angelo unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Tue Aug 29 13:45:35 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13Tkkj-0001Iq-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Tue, 29 Aug 2000 13:44:33 +0100 Received: from dyn-33.linux.theplanet.co.uk ([195.92.244.33] helo=caramon.arm.linux.org.uk) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13Tkkh-0001IT-00; Tue, 29 Aug 2000 13:44:32 +0100 Received: from flint.arm.linux.org.uk (root@flint [192.168.0.4]) by caramon.arm.linux.org.uk (8.9.3/8.9.3) with ESMTP id LAA13767; Tue, 29 Aug 2000 11:01:31 +0100 Received: (from linux@localhost) by flint.arm.linux.org.uk (8.9.3/8.9.3) id LAA22151; Tue, 29 Aug 2000 11:00:30 +0100 From: Russell King - ARM Linux Admin Message-Id: <200008291000.LAA22151@flint.arm.linux.org.uk> Subject: Re: porting linux to SA1110 Assabet? To: peter.mc.chen@philips.com Date: Tue, 29 Aug 2000 11:00:30 +0100 (BST) Cc: linux-arm-kernel@lists.arm.linux.org.uk In-Reply-To: <0056920007226016000002L262*@MHS> from "peter.mc.chen@philips.com" at Aug 29, 2000 03:13:43 PM X-Mailer: ELM [version 2.5 PL1] MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk peter.mc.chen@philips.com writes: > Many people base on 2.4.0-test2-ac1-rmk2-np1 kernel, Can I ask a > very simple question. What is the meaning of this version name? (I mean > ac1, rmk2, or np1) And where can I get one(sorry I have spend time finding > such kernel, but till now only 2.4.0-test1-ac7-rmk1-np6-hh1 I can find) Its a potential (real) 2.4.0 kernel, test2 candidate. On top of that, it contains Alan Cox first patch (ac1). On top of that, there is my second patch for that (2.4.0-test2-ac1) version, and Nico Pitres first patch on top of that (2.4.0-test2-ac1-rmk2) version. So, from the other version you describe: 2.4.0 potential 2.4.0 kernel test1 test1 candidate ac7 Alan Cox 7th patch rmk1 My first patch np6 Nico's 6th patch hh1 HH's first patch (who's HH?) Normally, people only produce patches that are either: 1. between theirs and Linus' kernel release (ie, Nico may produce a patch between 2.4.0-test1 and 2.4.0-test1-ac7-rmk1-np6 which would contain in there the -ac7 and -rmk1 patches). 2. between theirs and the patch directly below them (ie, Nico produces a patch beween 2.4.0-test1-ac7-rmk1 and 2.4.0-test1-ac7-rmk1-np6). 3. between their current and previous versions (eg, 2.4.0-test1-ac7-rmk1-np5 and 2.4.0-test1-ac7-rmk1-np6). When you mean "where can I get one" do you mean kernel, patch, or version number? The -rmk? stuff isn't distributed as kernel tar balls because bandwidth is expensive around here (and uploading 20MB per is just too huge to even contemplate). They're distributed as patches ontop of Linus kernel, with or without Alans patch applied. _____ |_____| ------------------------------------------------- ---+---+- | | Russell King linux@arm.linux.org.uk --- --- | | | | http://www.arm.linux.org.uk/ / / | | +-+-+ --- -+- / | THE developer of ARM Linux |+| /|\ / | | | --- | +-+-+ ------------------------------------------------- /\\\ | unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Tue Aug 29 14:28:16 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13TlQH-0001ny-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Tue, 29 Aug 2000 14:27:29 +0100 Received: from [192.58.206.2] (helo=crl.dec.com) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13TlQG-0001nm-00 for linux-arm-kernel@lists.arm.linux.org.uk; Tue, 29 Aug 2000 14:27:28 +0100 Received: from ims.crl.dec.com (ims.crl.dec.com [16.11.0.11]) by crl.dec.com (8.8.8/RWD-1.2) with ESMTP id JAA27282; Tue, 29 Aug 2000 09:27:16 -0400 (EDT) Received: by ims.crl.dec.com with Internet Mail Service (5.5.2650.21) id ; Tue, 29 Aug 2000 09:33:09 -0400 Message-ID: From: Jamey Hicks To: "'Russell King - ARM Linux Admin'" , peter.mc.chen@philips.com Cc: linux-arm-kernel@lists.arm.linux.org.uk Subject: RE: porting linux to SA1110 Assabet? Date: Tue, 29 Aug 2000 09:27:20 -0400 MIME-Version: 1.0 X-Mailer: Internet Mail Service (5.5.2650.21) Content-Type: text/plain Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk > 2.4.0 potential 2.4.0 kernel > test1 test1 candidate > ac7 Alan Cox 7th patch > rmk1 My first patch > np6 Nico's 6th patch > hh1 HH's first patch (who's HH?) HH is for handhelds.org -- contains patches from several of us working on Linux on the iPAQ handheld computer. -Jamey unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Tue Aug 29 15:12:22 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13Tm6m-0002Hj-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Tue, 29 Aug 2000 15:11:24 +0100 Received: from [139.175.250.58] (helo=mail2000.com.tw) by www.linux.org.uk with smtp (Exim 3.13 #1) id 13Tm6l-0002HN-00 for linux-arm-kernel@lists.arm.linux.org.uk; Tue, 29 Aug 2000 15:11:23 +0100 Received: By OpenMail Mailer;Tue, 29 Aug 2000 22:10:22 +0800 (CST) From: "Chien-Min Lee" Reply-To: cm_lee@mail2000.com.tw Subject: Re: What does "external abort on non-linefetch" mean? Message-ID: <967558222.6847.cm_lee@mail2000.com.tw> To: "" Cc: "" , "linux-arm-kernel" , "" Date: Tue, 29 Aug 2000 22:10:22 +0800 (CST) MIME-Version: 1.0 Content-Type: text/plain Content-Transfer-Encoding: 8bit Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk Hello: > > This means that you passed an invalid address to the DMA controller. > Surely not. If this is true, the SA1110 does not follow the ARM ARM. Data > aborts are not supposed to be spuriously raised. Thanks for your replies. The problem seems from virt_to_phys(). I jus pass the physical address translated by this macro to DMA DBSA/DBSB, and start DMA transfer to get data from device. Then I translated the physical address in DBSA/DBSB to virtual address, and drain the left data in device's fifo to following memory address. In above sequence, sometimes it just works well, but sometimes it crash with "external abort on non-linefetch" error. Now, I use another variable to keep the required physical address, and void using virt_to_phys()/phys_to_virt(). At least the "external..." error does not come again. I still use test5-rmk1-np1 patch... best regards kirk ----== Mailed via Openfind ==----- http://mail2000.com.tw/ ´£¨Ñ§K¶O¹q¤l¶l¥ó«H½c unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Tue Aug 29 15:26:03 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13TmKH-0002SH-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Tue, 29 Aug 2000 15:25:21 +0100 Received: from [130.161.42.1] (helo=duteinh.et.tudelft.nl) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13TmKG-0002SB-00 for linux-arm-kernel@lists.arm.linux.org.uk; Tue, 29 Aug 2000 15:25:20 +0100 Received: from dutein75 by duteinh.et.tudelft.nl (8.9.3/BB-2.0) id QAA20110; Tue, 29 Aug 2000 16:25:31 +0200 (MET DST) Message-Id: <200008291425.QAA20110@duteinh.et.tudelft.nl> From: Erik Mouw To: linux-arm-kernel@lists.arm.linux.org.uk Subject: crystal driver and irq #0 Organization: Eric conspiracy secret labs Date: Tue 29 Aug 2000 16:25:28 +0100 (MDT) X-Face: "!Z|1XjFg!3Iw)BwM^M\k)J5KHiPddz}#/Zw%?$&Q!xqob"OA|})3_@&BTTiQERo?/^g[Z<9P4R8:5C:aqq.cw!hotLc=+u-"+1rq@9$V^8PbOUlodyi_sejt+D`iqkLpx$>.aXl%vnfPIt"LK|/%>94_Zrm8jy?`_[RKt'TuoxUxfk(gMhY(n|DBBl~c$dhO>G%6}R4+DGX3|idBm4oU&UM/XUP4I9aoN{]`ax%B?rgD(I3OJ<62Y6L X-Mailer: XCmail 1.2devel - with PGP support, PGP engine version 0.5 (IRIX64) X-Mailerorigin: http://www.fsai.fh-trier.de/~schmitzj/Xclasses/XCmail/ X-Eric-Conspiracy: There is no conspiracy! X-Loop: erik@it.et.tudelft.nl MIME-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: 8bit Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk Hi all, I'm hacking on the Crystal CS8900 ethernet driver (cs89x0.c) and I'm currently at a point where it can send packets (I can see arp broadcasts), but somehow there are no interrupts processed. When I try to ping another host, I can see the interrupts from the response packets in /proc/interrupts, but the driver doesn't respond. I'm using interrupt #0 (GPIO0 interrupt on the SA1100), is there somehow a big no-no on that interrupt? Erik -- J.A.K. (Erik) Mouw, Information and Communication Theory Group, Department of Electrical Engineering, Faculty of Information Technology and Systems, Delft University of Technology, PO BOX 5031, 2600 GA Delft, The Netherlands Phone: +31-15-2783635 Fax: +31-15-2781843 Email: J.A.K.Mouw@its.tudelft.nl WWW: http://www-ict.its.tudelft.nl/~erik/ unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Tue Aug 29 15:43:59 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13TmZH-0002do-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Tue, 29 Aug 2000 15:40:51 +0100 Received: from [24.147.212.45] (helo=compaq.parker.boston.ma.us) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13TmZG-0002dh-00 for linux-arm-kernel@lists.arm.linux.org.uk; Tue, 29 Aug 2000 15:40:50 +0100 Received: from p2.parker.boston.ma.us (p2 [192.245.5.16]) by compaq.parker.boston.ma.us (8.8.5/8.8.5) with ESMTP id KAA00509; Tue, 29 Aug 2000 10:40:49 -0400 Received: from p2.parker.boston.ma.us (localhost [127.0.0.1]) by p2.parker.boston.ma.us (8.9.3/8.8.8) with ESMTP id KAA61441; Tue, 29 Aug 2000 10:40:49 -0400 (EDT) (envelope-from brad@p2.parker.boston.ma.us) Message-Id: <200008291440.KAA61441@p2.parker.boston.ma.us> To: "Arjan Opmeer (EMN)" cc: "'linux-arm-kernel@lists.arm.linux.org.uk'" , "'imai@vasara.co.jp'" Subject: Re: Kunihiko Imais USB ohci driver doesn't work? In-reply-to: Your message of "Mon, 28 Aug 2000 16:29:19 +0200." Date: Tue, 29 Aug 2000 10:40:48 -0400 From: Brad Parker Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk "Arjan Opmeer (EMN)" wrote: > >Hi, > >I just tested Kunihiko Imais ohci driver for the SA1111 board that he >recently mentioned here on the list, but it doesn't seem to work. >I tested it against kernel 2.4.0-test6-rmk5-np2. It does detect the >ohci hub, but not the USB mouse I connected to the port. The driver >also detects an overcurrent condition on the port, with, but also >without, the mouse attached. > >Did anyone else have success trying this driver? > >He also mentioned a possible SA1111 DMA bug which would prevent USB >from working altogether. Does someone know whether Neponset can >suffer from this bug? I've been working on this on and off for some time. I have several variants of the driver, none of which work correctly. My current problem is that if I run the Intel diag problem first, and then run linux, the usb controller magically springs to life and works pretty well. If I power cycle the h/w it stops working. So, obviously something is not being initialized correctly in linux. I thought it was the MBREQ/MBGNT gpio lines, but that didn't fix it. I'm trying to figure out what is missing. The built in hub seems to work but the usb controller itself seems half dead. Perhaps a missing clock, I'm not sure... You can see this if you turn on SOF interrupts and notice that then never occur. Also, the frame # is bogus and never increments. Root hub change interrupts do occur. If you run the diags and then reload linux SOF interrupts start happening and you get sof interrupts... The fixes which move the hcca, td's & ed's into 'consistent' memory seem to help. So does turning off the specilative ed removal (EARLY_ED_UNLINK). There is a bug, however, which looks like dl_del_list is called twice when a device is removed; I'm not sure what's causing it but the symptom is an ed is free'd twice. The way the new ed allocation code was written it reuses the hwINFO field and this causes a crash in dl_del_list. I believe simply moving the link field will make the problem go away but I plan to track down the real cause... (once I get the controller to work consistently from power up) I have not seen any evidence of a dma problem, but one may well exist... -brad unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Tue Aug 29 16:11:47 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13Tn1N-0002zA-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Tue, 29 Aug 2000 16:09:53 +0100 Received: from [205.250.170.220] (helo=xanadu.vipswitch.com) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13Tn1L-0002z3-00 for linux-arm-kernel@lists.arm.linux.org.uk; Tue, 29 Aug 2000 16:09:51 +0100 Date: Tue, 29 Aug 2000 11:09:28 -0400 (EDT) From: Nicolas Pitre To: Chien-Min Lee cc: , , linux-arm-kernel Subject: Re: What does "external abort on non-linefetch" mean? In-Reply-To: <967558222.6847.cm_lee@mail2000.com.tw> Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk On Tue, 29 Aug 2000, Chien-Min Lee wrote: > The problem seems from virt_to_phys(). > I jus pass the physical address translated by this macro to DMA DBSA/DBSB, > and start DMA transfer to get data from device. > Then I translated the physical address in DBSA/DBSB to virtual address, > and drain the left data in device's fifo to following memory address. May I suggest you use the generic SA1100 DMA API in my latest patch? The next patch will also include documentation. Nicolas unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Tue Aug 29 16:14:03 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13Tn4n-00033K-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Tue, 29 Aug 2000 16:13:25 +0100 Received: from [205.250.170.220] (helo=xanadu.vipswitch.com) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13Tn4m-00033D-00 for linux-arm-kernel@lists.arm.linux.org.uk; Tue, 29 Aug 2000 16:13:24 +0100 Date: Tue, 29 Aug 2000 11:13:19 -0400 (EDT) From: Nicolas Pitre To: Erik Mouw cc: linux-arm-kernel@lists.arm.linux.org.uk Subject: Re: crystal driver and irq #0 In-Reply-To: <200008291425.QAA20110@duteinh.et.tudelft.nl> Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk On Tue, 29 Aug 2000, Erik Mouw wrote: > Hi all, > > I'm hacking on the Crystal CS8900 ethernet driver (cs89x0.c) and I'm > currently at a point where it can send packets (I can see arp broadcasts), > but somehow there are no interrupts processed. When I try to ping another > host, I can see the interrupts from the response packets in > /proc/interrupts, but the driver doesn't respond. I'm using interrupt #0 > (GPIO0 interrupt on the SA1100), is there somehow a big no-no on that > interrupt? No problem here. Some drivers might think that IRQ autoprobing failed when the returned IRQ is 0, but that's easy to fix, and that's probably not the problem you're experimenting. Nicolas unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Tue Aug 29 16:28:26 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13TnId-0003Hq-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Tue, 29 Aug 2000 16:27:43 +0100 Received: from [130.161.42.1] (helo=duteinh.et.tudelft.nl) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13TnIc-0003Hh-00 for linux-arm-kernel@lists.arm.linux.org.uk; Tue, 29 Aug 2000 16:27:42 +0100 Received: from dutein75 by duteinh.et.tudelft.nl (8.9.3/BB-2.0) id RAA20568; Tue, 29 Aug 2000 17:27:48 +0200 (MET DST) Message-Id: <200008291527.RAA20568@duteinh.et.tudelft.nl> From: Erik Mouw To: nico@cam.org Cc: linux-arm-kernel@lists.arm.linux.org.uk Subject: Re: crystal driver and irq #0 In-Reply-To: Organization: Eric conspiracy secret labs Date: Tue 29 Aug 2000 17:27:45 +0100 (MDT) X-Face: "!Z|1XjFg!3Iw)BwM^M\k)J5KHiPddz}#/Zw%?$&Q!xqob"OA|})3_@&BTTiQERo?/^g[Z<9P4R8:5C:aqq.cw!hotLc=+u-"+1rq@9$V^8PbOUlodyi_sejt+D`iqkLpx$>.aXl%vnfPIt"LK|/%>94_Zrm8jy?`_[RKt'TuoxUxfk(gMhY(n|DBBl~c$dhO>G%6}R4+DGX3|idBm4oU&UM/XUP4I9aoN{]`ax%B?rgD(I3OJ<62Y6L X-Mailer: XCmail 1.2devel - with PGP support, PGP engine version 0.5 (IRIX64) X-Mailerorigin: http://www.fsai.fh-trier.de/~schmitzj/Xclasses/XCmail/ X-Eric-Conspiracy: There is no conspiracy! X-Loop: erik@it.et.tudelft.nl MIME-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: 8bit Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk On Tue, 29 Aug 2000 11:13:19 -0400 (EDT), Nicolas Pitre wrote: > On Tue, 29 Aug 2000, Erik Mouw wrote: >> I'm hacking on the Crystal CS8900 ethernet driver (cs89x0.c) and I'm >> currently at a point where it can send packets (I can see arp broadcasts), >> but somehow there are no interrupts processed. When I try to ping another >> host, I can see the interrupts from the response packets in >> /proc/interrupts, but the driver doesn't respond. I'm using interrupt #0 >> (GPIO0 interrupt on the SA1100), is there somehow a big no-no on that >> interrupt? > > No problem here. Some drivers might think that IRQ autoprobing failed > when the returned IRQ is 0, but that's easy to fix, and that's probably > not the problem you're experimenting. What does ifconfig say about the interrupt number in that case? Erik -- J.A.K. (Erik) Mouw, Information and Communication Theory Group, Department of Electrical Engineering, Faculty of Information Technology and Systems, Delft University of Technology, PO BOX 5031, 2600 GA Delft, The Netherlands Phone: +31-15-2783635 Fax: +31-15-2781843 Email: J.A.K.Mouw@its.tudelft.nl WWW: http://www-ict.its.tudelft.nl/~erik/ unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Tue Aug 29 16:36:36 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13TnQe-0003OH-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Tue, 29 Aug 2000 16:36:00 +0100 Received: from dyn-33.linux.theplanet.co.uk ([195.92.244.33] helo=caramon.arm.linux.org.uk) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13TnQb-0003O6-00; Tue, 29 Aug 2000 16:35:58 +0100 Received: from flint.arm.linux.org.uk (root@flint [192.168.0.4]) by caramon.arm.linux.org.uk (8.9.3/8.9.3) with ESMTP id QAA14961; Tue, 29 Aug 2000 16:36:04 +0100 Received: (from linux@localhost) by flint.arm.linux.org.uk (8.9.3/8.9.3) id QAA24253; Tue, 29 Aug 2000 16:35:01 +0100 From: Russell King - ARM Linux Admin Message-Id: <200008291535.QAA24253@flint.arm.linux.org.uk> Subject: Re: crystal driver and irq #0 To: J.A.K.Mouw@its.tudelft.nl (Erik Mouw) Date: Tue, 29 Aug 2000 16:35:01 +0100 (BST) Cc: nico@cam.org, linux-arm-kernel@lists.arm.linux.org.uk In-Reply-To: <200008291527.RAA20568@duteinh.et.tudelft.nl> from "Erik Mouw" at Aug 29, 2000 05:27:45 PM X-Mailer: ELM [version 2.5 PL1] MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk Erik Mouw writes: > On Tue, 29 Aug 2000 11:13:19 -0400 (EDT), Nicolas Pitre wrote: > > On Tue, 29 Aug 2000, Erik Mouw wrote: > >> I'm hacking on the Crystal CS8900 ethernet driver (cs89x0.c) and I'm > >> currently at a point where it can send packets (I can see arp broadcasts), > >> but somehow there are no interrupts processed. When I try to ping another > >> host, I can see the interrupts from the response packets in > >> /proc/interrupts, but the driver doesn't respond. I'm using interrupt #0 > >> (GPIO0 interrupt on the SA1100), is there somehow a big no-no on that > >> interrupt? > > > > No problem here. Some drivers might think that IRQ autoprobing failed > > when the returned IRQ is 0, but that's easy to fix, and that's probably > > not the problem you're experimenting. > > What does ifconfig say about the interrupt number in that case? Really, irq0 is an in-grained brokenness in the x86 Linux people. IRQ0 as far as they are concerned is a universal "this interrupt is does not exist", rather than "-1" which is incidentally specified as "no interrupt" in the PCI specs... _____ |_____| ------------------------------------------------- ---+---+- | | Russell King linux@arm.linux.org.uk --- --- | | | | http://www.arm.linux.org.uk/ / / | | +-+-+ --- -+- / | THE developer of ARM Linux |+| /|\ / | | | --- | +-+-+ ------------------------------------------------- /\\\ | unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Tue Aug 29 16:37:39 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13TnRn-0003PV-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Tue, 29 Aug 2000 16:37:11 +0100 Received: from [205.250.170.220] (helo=xanadu.vipswitch.com) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13TnRn-0003PL-00 for linux-arm-kernel@lists.arm.linux.org.uk; Tue, 29 Aug 2000 16:37:11 +0100 Date: Tue, 29 Aug 2000 11:36:57 -0400 (EDT) From: Nicolas Pitre To: Erik Mouw cc: linux-arm-kernel@lists.arm.linux.org.uk Subject: Re: crystal driver and irq #0 In-Reply-To: <200008291527.RAA20568@duteinh.et.tudelft.nl> Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk On Tue, 29 Aug 2000, Erik Mouw wrote: > On Tue, 29 Aug 2000 11:13:19 -0400 (EDT), Nicolas Pitre wrote: > > On Tue, 29 Aug 2000, Erik Mouw wrote: > >> I'm hacking on the Crystal CS8900 ethernet driver (cs89x0.c) and I'm > >> currently at a point where it can send packets (I can see arp broadcasts), > >> but somehow there are no interrupts processed. When I try to ping another > >> host, I can see the interrupts from the response packets in > >> /proc/interrupts, but the driver doesn't respond. I'm using interrupt #0 > >> (GPIO0 interrupt on the SA1100), is there somehow a big no-no on that > >> interrupt? > > > > No problem here. Some drivers might think that IRQ autoprobing failed > > when the returned IRQ is 0, but that's easy to fix, and that's probably > > not the problem you're experimenting. > > What does ifconfig say about the interrupt number in that case? Well... if it's 0 it should say 0. Or maybe it is broken and thinks there is no interrupt? I don't remember how it behaved... See 'cat /proc/interrupts' for the ultimate reference. Nicolas unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Tue Aug 29 16:52:05 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13TnfU-0003b9-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Tue, 29 Aug 2000 16:51:20 +0100 Received: from [151.190.254.100] (helo=gate1.itt.com) by www.linux.org.uk with smtp (Exim 3.13 #1) id 13TnfT-0003ao-00 for linux-arm-kernel@lists.arm.linux.org.uk; Tue, 29 Aug 2000 16:51:19 +0100 Received: from fwemail2.de.ittind.com by gate1.itt.com via smtpd (for parcelfarce.linux.theplanet.co.uk [195.92.249.252]) with SMTP; 29 Aug 2000 15:51:31 UT Received: by fwemail2.de.ittind.com with Internet Mail Service (5.5.2650.21) id ; Tue, 29 Aug 2000 10:51:26 -0500 Message-ID: From: "Latham, Steve" To: "'sa1100-linux@pa.dec.com'" , linux-arm-kernel@lists.arm.linux.org.uk Cc: "'jason.chagas@intel.com'" Subject: Consecutive Telnet sessions on assabet don't work Date: Tue, 29 Aug 2000 10:48:17 -0500 MIME-Version: 1.0 X-Mailer: Internet Mail Service (5.5.2650.21) Content-Type: text/plain; charset="iso-8859-1" Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk I downloaded and used blob-kit-assabet-1.3.tar.gz that uses 2.4.0-test5-rmk1-np1 and telnet works great but only the first time. Subsequent sessions hang at "connected", to local or remote clients. I've had this problem on earlier versions of kernel/ramdisk combinations as well. Is anyone else experiencing this problem? Is it possible that sockets aren't cleaned up? Steve ************************************ If this email is not intended for you, or you are not responsible for the delivery of this message to the addressee, please note that this message may contain ITT Privileged/Proprietary Information. In such a case, you may not copy or deliver this message to anyone. You should destroy this message and kindly notify the sender by reply email. Information contained in this message that does not relate to the business of ITT is neither endorsed by nor attributable to ITT. ************************************ unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Tue Aug 29 19:09:19 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13Tpo8-00057n-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Tue, 29 Aug 2000 19:08:24 +0100 Received: from [63.195.182.101] (helo=barbados.bluemug.com) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13Tpo7-00057h-00 for linux-arm-kernel@lists.arm.linux.org.uk; Tue, 29 Aug 2000 19:08:23 +0100 Received: from fiji.bluemug.com [192.168.50.32] by barbados.bluemug.com with esmtp (Exim 2.05 #1 (Debian)) id 13TpoH-0000yn-00; Tue, 29 Aug 2000 11:08:33 -0700 Received: from miket by fiji.bluemug.com with local (Exim 3.12 #1 (Debian)) id 13TpoH-0006a4-00; Tue, 29 Aug 2000 11:08:33 -0700 Date: Tue, 29 Aug 2000 11:08:33 -0700 From: Mike Touloumtzis To: linux-arm-kernel@lists.arm.linux.org.uk Subject: EP7211 pgdir problem -- solved (+ARM kernel question) Message-ID: <20000829110833.J12563@fiji.bluemug.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk Hello world, Yesterday I resolved the EP7211 porting issue that I described on Saturday (memory maps apparently not being updated by paging_init()). The problem turned out to be that the initial 4MB identity mapping and swapper_pg_dir don't overlap on the EP7211 because I'm leaving the first 128MB of DRAM reserved for video (this is a hardware-imposed constraint). The identity mapping pgdir address is computed in head-armv.S assembly language as: physical RAM start + SWAPPER_PGDIR_OFFSET where SWAPPER_PGDIR_OFFSET is 0x4000. The swapper_pg_dir mapping pgdir address is computed in a #define in the same file as: TEXTADDR - 0x8000 + SWAPPER_PGDIR_OFFSET All existing ARM architectures start the kernel image at 0x8000 past the beginning of physical DRAM, so these computations produce the same result. It's not practical for EP7211 because of the aforementioned video memory mapping issue. Given this, can anyone tell me which of the following is correct: (a) There is no assumption that the head-*.S identity mapping must lie at the same place as swapper_pg_dir. There is a missing set_pgdir() to switch between them, but this hasn't come up because they currently happen to overlap, so the set_pgdir() would be a no-op. (b) The identity mapping and swapper_pg_dir must always overlap, because the overwriting in-place of the identity mapping by paging_init() is intended. Architectures which can't put TEXTADDR at (start of RAM + 0x8000) must modify one of the computations listed above to force an overlap. (c) None of the above :-). thanks, miket unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Tue Aug 29 19:25:13 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13Tq3o-0005JO-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Tue, 29 Aug 2000 19:24:36 +0100 Received: from dyn-33.linux.theplanet.co.uk ([195.92.244.33] helo=caramon.arm.linux.org.uk) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13Tq3m-0005JH-00; Tue, 29 Aug 2000 19:24:35 +0100 Received: from flint.arm.linux.org.uk (root@flint [192.168.0.4]) by caramon.arm.linux.org.uk (8.9.3/8.9.3) with ESMTP id TAA15678; Tue, 29 Aug 2000 19:24:44 +0100 Received: (from linux@localhost) by flint.arm.linux.org.uk (8.9.3/8.9.3) id TAA24801; Tue, 29 Aug 2000 19:23:41 +0100 From: Russell King - ARM Linux Admin Message-Id: <200008291823.TAA24801@flint.arm.linux.org.uk> Subject: Re: EP7211 pgdir problem -- solved (+ARM kernel question) To: miket@bluemug.com (Mike Touloumtzis) Date: Tue, 29 Aug 2000 19:23:40 +0100 (BST) Cc: linux-arm-kernel@lists.arm.linux.org.uk In-Reply-To: <20000829110833.J12563@fiji.bluemug.com> from "Mike Touloumtzis" at Aug 29, 2000 11:08:33 AM X-Mailer: ELM [version 2.5 PL1] MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk Mike Touloumtzis writes: > The problem turned out to be that the initial 4MB identity mapping and > swapper_pg_dir don't overlap on the EP7211 because I'm leaving the first > 128MB of DRAM reserved for video (this is a hardware-imposed constraint). Where precisely are you loading the kernel, and what are the start addresses of physical and virtual RAM? > (b) The identity mapping and swapper_pg_dir must always overlap, because > the overwriting in-place of the identity mapping by paging_init() > is intended. Architectures which can't put TEXTADDR at (start of > RAM + 0x8000) must modify one of the computations listed above to > force an overlap. Correct. The table should be located at TEXTADDR - 0x4000, so the one that is broken is in the page table code. _____ |_____| ------------------------------------------------- ---+---+- | | Russell King linux@arm.linux.org.uk --- --- | | | | http://www.arm.linux.org.uk/ / / | | +-+-+ --- -+- / | THE developer of ARM Linux |+| /|\ / | | | --- | +-+-+ ------------------------------------------------- /\\\ | unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Tue Aug 29 19:44:37 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13TqMS-0005Tt-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Tue, 29 Aug 2000 19:43:52 +0100 Received: from [207.183.249.4] (helo=ultra2.quiknet.com) by www.linux.org.uk with smtp (Exim 3.13 #1) id 13TqMR-0005Tl-00 for linux-arm-kernel@lists.arm.linux.org.uk; Tue, 29 Aug 2000 19:43:51 +0100 Received: (qmail 6251 invoked from network); 29 Aug 2000 18:43:58 -0000 Received: from unknown (HELO quiknet.com) (root@207.231.75.21) by ultra2.quiknet.com with SMTP; 29 Aug 2000 18:43:58 -0000 Message-ID: <39AC040E.93BDE5B7@quiknet.com> Date: Tue, 29 Aug 2000 11:42:22 -0700 From: pecondon X-Mailer: Mozilla 4.72 [en] (X11; U; Linux 2.2.14-5.0 i686) X-Accept-Language: en MIME-Version: 1.0 To: linux-arm-kernel@lists.arm.linux.org.uk Subject: Re: Improvements? (was Re: switched to strongARM and 160 Mb and Debian 2.2 doesn't boot) References: <200008290848.JAA21825@flint.arm.linux.org.uk> Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk I'm following this list in the hope that I can learn about the various ARM platforms that are available and supported. Eventually, I hope to make a (mildly) rational selection of platform for a device I want to build. Is there a catalog of platforms giving a short description of each and an official name for each? Is there a database of features that are, or are not, included in each platform? (i.e. the sort check list that marketers like to use to flog their products.) Russell King - ARM Linux Admin wrote: > Scott Bambrough writes: > > Huh? We have had NetWinder's running with 256 MB of RAM. We regularly ship > > machines with 128 MB. It's StrongARM based. > > The original mail was about a RiscPC: > > | I have debian 2.2 running for a few weeks on my RISC PC 700, with an ARM 7 > | and 64 Mb (+2 Mb VRAM). This worked rather well (did some recompiling and > | stuff). > > Can people please remember that, with us supporting some 40 different > types of platform, with about 20 of them being "StrongARM" based, that > when someone talks about a processor type (ie, ARM7, StrongARM, ARM920T) > that they are not necessarily talking about your particular platform. > It may help also if the platform is mentioned in the subject line? > > Ideas for improvement of quality welcome. > _____ > |_____| ------------------------------------------------- ---+---+- > | | Russell King linux@arm.linux.org.uk --- --- > | | | | http://www.arm.linux.org.uk/ / / | > | +-+-+ --- -+- > / | THE developer of ARM Linux |+| /|\ > / | | | --- | > +-+-+ ------------------------------------------------- /\\\ | > > unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Tue Aug 29 19:58:09 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13TqZh-0005dM-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Tue, 29 Aug 2000 19:57:33 +0100 Received: from dyn-33.linux.theplanet.co.uk ([195.92.244.33] helo=caramon.arm.linux.org.uk) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13TqZf-0005dF-00; Tue, 29 Aug 2000 19:57:32 +0100 Received: from flint.arm.linux.org.uk (root@flint [192.168.0.4]) by caramon.arm.linux.org.uk (8.9.3/8.9.3) with ESMTP id TAA15773; Tue, 29 Aug 2000 19:57:41 +0100 Received: (from linux@localhost) by flint.arm.linux.org.uk (8.9.3/8.9.3) id TAA24905; Tue, 29 Aug 2000 19:56:38 +0100 From: Russell King - ARM Linux Admin Message-Id: <200008291856.TAA24905@flint.arm.linux.org.uk> Subject: Re: Improvements? (was Re: switched to strongARM and 160 Mb and Debian To: pecondon@quiknet.com (pecondon) Date: Tue, 29 Aug 2000 19:56:38 +0100 (BST) Cc: linux-arm-kernel@lists.arm.linux.org.uk In-Reply-To: <39AC040E.93BDE5B7@quiknet.com> from "pecondon" at Aug 29, 2000 11:42:22 AM X-Mailer: ELM [version 2.5 PL1] MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk pecondon writes: > I'm following this list in the hope that I can learn about the various ARM platforms that > are available and supported. Eventually, I hope to make a (mildly) rational selection of > platform for a device I want to build. Is there a catalog of platforms giving a short > description of each and an official name for each? Is there a database of features that > are, or are not, included in each platform? (i.e. the sort check list that marketers like > to use to flog their products.) If you'd like to suggest some initial categories of the types of things you'd like to see, I can probably track down a large percentage of the maintainers to get them to provide the relevent information. They should actually all see this mail as well... _____ |_____| ------------------------------------------------- ---+---+- | | Russell King linux@arm.linux.org.uk --- --- | | | | http://www.arm.linux.org.uk/ / / | | +-+-+ --- -+- / | THE developer of ARM Linux |+| /|\ / | | | --- | +-+-+ ------------------------------------------------- /\\\ | unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Tue Aug 29 20:15:33 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13Tqp5-0005qL-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Tue, 29 Aug 2000 20:13:27 +0100 Received: from [63.195.182.101] (helo=barbados.bluemug.com) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13Tqp4-0005qF-00 for linux-arm-kernel@lists.arm.linux.org.uk; Tue, 29 Aug 2000 20:13:27 +0100 Received: from fiji.bluemug.com [192.168.50.32] by barbados.bluemug.com with esmtp (Exim 2.05 #1 (Debian)) id 13Tqp9-00010S-00; Tue, 29 Aug 2000 12:13:31 -0700 Received: from miket by fiji.bluemug.com with local (Exim 3.12 #1 (Debian)) id 13Tqp9-0006rk-00; Tue, 29 Aug 2000 12:13:31 -0700 Date: Tue, 29 Aug 2000 12:13:31 -0700 From: Mike Touloumtzis To: Russell King - ARM Linux Admin Cc: linux-arm-kernel@lists.arm.linux.org.uk Subject: Re: EP7211 pgdir problem -- solved (+ARM kernel question) Message-ID: <20000829121331.K12563@fiji.bluemug.com> References: <20000829110833.J12563@fiji.bluemug.com> <200008291823.TAA24801@flint.arm.linux.org.uk> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii In-Reply-To: <200008291823.TAA24801@flint.arm.linux.org.uk>; from Russell King - ARM Linux Admin on Tue, Aug 29, 2000 at 07:23:40PM +0100 Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk On Tue, Aug 29, 2000 at 07:23:40PM +0100, Russell King - ARM Linux Admin wrote: > Mike Touloumtzis writes: > > The problem turned out to be that the initial 4MB identity mapping and > > swapper_pg_dir don't overlap on the EP7211 because I'm leaving the first > > 128MB of DRAM reserved for video (this is a hardware-imposed constraint). > > Where precisely are you loading the kernel, and what are the start > addresses of physical and virtual RAM? Start of virtual RAM and physical RAM are both 0xc0000000. Video RAM reserved area is 128K: 0xc0000000:0xc0020000. The LCD controller in the 7211 can only use the very beginning of a memory bank as video RAM. Param block is 0xc0020000 -- right after video RAM. TEXTADDR is 0xc0038000 (legacy of 2.2.1/Psion port; I'm thinking of moving to 0xc0028000). So right now, swapper_pg_dir is at 0xc0034000. > > > (b) The identity mapping and swapper_pg_dir must always overlap, because > > the overwriting in-place of the identity mapping by paging_init() > > is intended. Architectures which can't put TEXTADDR at (start of > > RAM + 0x8000) must modify one of the computations listed above to > > force an overlap. > > Correct. The table should be located at TEXTADDR - 0x4000, so the one > that is broken is in the page table code. To get things working for me, I made the following change in head-armv.S (r5 is physical address of start or RAM): __create_page_tables: # XXX miket: changing pgdir address from following # add r4, r5, #SWAPPER_PGDIR_OFFSET # in order to handle the video RAM offset. # This is a quick hack fix, not a real solution. add r4, r5, #0x00034000 miket unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Tue Aug 29 20:38:05 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13TrBz-00068c-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Tue, 29 Aug 2000 20:37:07 +0100 Received: from [63.195.182.101] (helo=barbados.bluemug.com) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13TrBy-00068W-00 for linux-arm-kernel@lists.arm.linux.org.uk; Tue, 29 Aug 2000 20:37:06 +0100 Received: from fiji.bluemug.com [192.168.50.32] by barbados.bluemug.com with esmtp (Exim 2.05 #1 (Debian)) id 13TrC6-000118-00; Tue, 29 Aug 2000 12:37:14 -0700 Received: from miket by fiji.bluemug.com with local (Exim 3.12 #1 (Debian)) id 13TrC5-0006xp-00; Tue, 29 Aug 2000 12:37:13 -0700 Date: Tue, 29 Aug 2000 12:37:13 -0700 From: Mike Touloumtzis To: Mike Touloumtzis Cc: Russell King - ARM Linux Admin , linux-arm-kernel@lists.arm.linux.org.uk Subject: Re: EP7211 pgdir problem -- solved (+ARM kernel question) Message-ID: <20000829123713.M12563@fiji.bluemug.com> References: <20000829110833.J12563@fiji.bluemug.com> <200008291823.TAA24801@flint.arm.linux.org.uk> <20000829121331.K12563@fiji.bluemug.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii In-Reply-To: <20000829121331.K12563@fiji.bluemug.com>; from Mike Touloumtzis on Tue, Aug 29, 2000 at 12:13:31PM -0700 Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk Just a suggestion about the TEXTADDR thing; if instead of (or in addition to) TEXTADDR, we had a TEXT_OFFSET (which would tend to be an acceptable ARM immediate value), the pgdir setup could become something like: add r4, r5, #TEXT_OFFSET sub r4, r4, #SWAPPER_PGDIR_OFFSET miket unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Tue Aug 29 22:25:25 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13TsqW-0000K7-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Tue, 29 Aug 2000 22:23:04 +0100 Received: from [130.161.42.1] (helo=duteinh.et.tudelft.nl) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13TsqV-0000Jz-00 for linux-arm-kernel@lists.arm.linux.org.uk; Tue, 29 Aug 2000 22:23:03 +0100 Received: from dutein75 by duteinh.et.tudelft.nl (8.9.3/BB-2.0) id XAA22247; Tue, 29 Aug 2000 23:22:23 +0200 (MET DST) Message-Id: <200008292122.XAA22247@duteinh.et.tudelft.nl> From: Erik Mouw To: john+@cs.cmu.edu Cc: nico@cam.org, linux@arm.linux.org.uk, ebunce@lhsl.com, bakker@thorgal.et.tudelft.nl, france@crl.dec.com, chris@blazie.com, a.melis@aino.nl, linux-arm-kernel@lists.arm.linux.org.uk Subject: Re: SA1100 DRAM sizes In-Reply-To: <39AC25A6.1AF7B535@cs.cmu.edu> Organization: Eric conspiracy secret labs Date: Tue 29 Aug 2000 23:22:22 +0100 (MDT) X-Face: "!Z|1XjFg!3Iw)BwM^M\k)J5KHiPddz}#/Zw%?$&Q!xqob"OA|})3_@&BTTiQERo?/^g[Z<9P4R8:5C:aqq.cw!hotLc=+u-"+1rq@9$V^8PbOUlodyi_sejt+D`iqkLpx$>.aXl%vnfPIt"LK|/%>94_Zrm8jy?`_[RKt'TuoxUxfk(gMhY(n|DBBl~c$dhO>G%6}R4+DGX3|idBm4oU&UM/XUP4I9aoN{]`ax%B?rgD(I3OJ<62Y6L X-Mailer: XCmail 1.2devel - with PGP support, PGP engine version 0.5 (IRIX64) X-Mailerorigin: http://www.fsai.fh-trier.de/~schmitzj/Xclasses/XCmail/ X-Eric-Conspiracy: There is no conspiracy! X-Loop: erik@it.et.tudelft.nl MIME-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: 8bit Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk On Tue, 29 Aug 2000 17:05:42 -0400, John G Dorsey wrote: > I've tried it with pcnet_cs on Assabet and Neponset, and ide_cs on Assabet. I > was surprised when I found: > > #define PCMCIA_IO_0_BASE 0xe0000000 > #define PCMCIA_IO_1_BASE 0xe4000000 > > ...in asm/arch/ide.h. Owners of this driver for devices that don't go through > Card Services should probably tell Nicolas what they want this code to look > like now. That's my stuff (well, I recently cleaned it up), although it is in the kernel since linux-2.2.x (or maybe longer). The Empeg, LART, and Victor have an IDE interface mapped into PCMCIA IO space 0. It is independent of the card services because it doesn't use card services, the drive is just always there. What's your suggestion how it should look like? Erik -- J.A.K. (Erik) Mouw, Information and Communication Theory Group, Department of Electrical Engineering, Faculty of Information Technology and Systems, Delft University of Technology, PO BOX 5031, 2600 GA Delft, The Netherlands Phone: +31-15-2783635 Fax: +31-15-2781843 Email: J.A.K.Mouw@its.tudelft.nl WWW: http://www-ict.its.tudelft.nl/~erik/ unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Tue Aug 29 23:57:48 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13TuJ2-0003RX-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Tue, 29 Aug 2000 23:56:36 +0100 Received: from [128.2.10.81] (helo=smtp1.andrew.cmu.edu) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13TuJ1-0003RR-00 for linux-arm-kernel@lists.arm.linux.org.uk; Tue, 29 Aug 2000 23:56:35 +0100 Received: from cs.cmu.edu (NINESIXTEEN.REM.CMU.EDU [128.2.81.157]) by smtp1.andrew.cmu.edu (8.9.3/8.9.3) with ESMTP id SAA26192; Tue, 29 Aug 2000 18:51:35 -0400 (EDT) Message-ID: <39AC3E1D.67023D7A@cs.cmu.edu> Date: Tue, 29 Aug 2000 18:50:05 -0400 From: John G Dorsey Organization: Carnegie Mellon University X-Mailer: Mozilla 4.72 [en] (X11; U; Linux 2.2.14-5.0 i686) X-Accept-Language: en MIME-Version: 1.0 To: Erik Mouw CC: nico@cam.org, linux@arm.linux.org.uk, ebunce@lhsl.com, bakker@thorgal.et.tudelft.nl, france@crl.dec.com, chris@blazie.com, a.melis@aino.nl, linux-arm-kernel@lists.arm.linux.org.uk Subject: Re: SA1100 DRAM sizes References: <200008292122.XAA22247@duteinh.et.tudelft.nl> Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk Erik Mouw wrote: > That's my stuff (well, I recently cleaned it up), although it is in the > kernel since linux-2.2.x (or maybe longer). I haven't tested ide_cs with any regularity since Erik Bunce first got it running (2.3.99), but removing the hardwired base address (now in ide_init_hwif_ports()) was part of that conversion. The code as it is currently written should fail when run against CS because of the address arithmetic being performed. > The Empeg, LART, and Victor > have an IDE interface mapped into PCMCIA IO space 0. It is independent of > the card services because it doesn't use card services, the drive is just > always there. What's your suggestion how it should look like? Once the static mapping is removed from the official tree, this code simply won't work. Constructing the mapping on demand, as the sa1100 socket driver now does, seems appropriate. I'm not sure where I'd do this in the code, though. I know ide_init_hwif_ports() is inline on other architectures, but it's becoming portly on SA-1100. (Also, on SPARC, there's a warning in the code against using ioremap() in this routine, as it can be called with degenerate parameters.) I can also imagine a simple implementation of the low-level socket driver operations that might allow these boards to just use Card Services. (Advantage: consolidated resource management code. Disadvantage: needs ramdisk to load ide_cs at boot.) -jd unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Wed Aug 30 00:12:23 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13TuWn-0003bp-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Wed, 30 Aug 2000 00:10:49 +0100 Received: from [128.2.205.86] (helo=post.ices.cmu.edu) by www.linux.org.uk with smtp (Exim 3.13 #1) id 13TuWm-0003bc-00 for linux-arm-kernel@lists.arm.linux.org.uk; Wed, 30 Aug 2000 00:10:48 +0100 Received: from VANCOUVER.ICES.CMU.EDU by post.ices.cmu.edu id aa26471; 29 Aug 2000 17:07 EDT Message-ID: <39AC25A6.1AF7B535@cs.cmu.edu> Date: Tue, 29 Aug 2000 17:05:42 -0400 From: John G Dorsey Organization: Carnegie Mellon University X-Mailer: Mozilla 4.74 [en] (X11; U; Linux 2.2.16-17smp i686) X-Accept-Language: en MIME-Version: 1.0 To: Nicolas Pitre CC: Russell King - ARM Linux Admin , Erik Mouw , ebunce@lhsl.com, bakker@thorgal.et.tudelft.nl, france@crl.dec.com, chris@blazie.com, a.melis@aino.nl, linux-arm-kernel@lists.arm.linux.org.uk Subject: Re: SA1100 DRAM sizes References: Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk Nicolas Pitre wrote: > **HOWEVER** this requires the generic SA1100 PCMCIA driver to be modified > so it actually ioremap() its memory and attribute space (even IO could be > done that way too for more clarity). This change must occur before all > the rest otherwise PCMCIA will be broken and people will step on my toes. Done. See linux-2.4.0-test6-rmk5-np2-jd1.patch.gz on http://www.cs.cmu.edu/~wearable/software/assabet.html (under "Bleeding Edge"). I've tried it with pcnet_cs on Assabet and Neponset, and ide_cs on Assabet. I was surprised when I found: #define PCMCIA_IO_0_BASE 0xe0000000 #define PCMCIA_IO_1_BASE 0xe4000000 ...in asm/arch/ide.h. Owners of this driver for devices that don't go through Card Services should probably tell Nicolas what they want this code to look like now. -jd unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Wed Aug 30 00:25:20 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13Tuke-0003oQ-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Wed, 30 Aug 2000 00:25:08 +0100 Received: from [63.195.182.101] (helo=barbados.bluemug.com) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13Tukd-0003oH-00 for linux-arm-kernel@lists.arm.linux.org.uk; Wed, 30 Aug 2000 00:25:07 +0100 Received: from fiji.bluemug.com [192.168.50.32] by barbados.bluemug.com with esmtp (Exim 2.05 #1 (Debian)) id 13TujS-000165-00; Tue, 29 Aug 2000 16:23:54 -0700 Received: from miket by fiji.bluemug.com with local (Exim 3.12 #1 (Debian)) id 13TujR-0007nY-00; Tue, 29 Aug 2000 16:23:53 -0700 Date: Tue, 29 Aug 2000 16:23:53 -0700 From: Mike Touloumtzis To: Nicolas Pitre Cc: Russell King - ARM Linux Admin , John G Dorsey , Erik Mouw , ebunce@lhsl.com, bakker@thorgal.et.tudelft.nl, france@crl.dec.com, chris@blazie.com, a.melis@aino.nl, linux-arm-kernel@lists.arm.linux.org.uk Subject: Re: SA1100 DRAM sizes Message-ID: <20000829162353.A29614@fiji.bluemug.com> References: <200008282219.XAA20092@flint.arm.linux.org.uk> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii In-Reply-To: ; from Nicolas Pitre on Mon, Aug 28, 2000 at 09:46:31PM -0400 Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk On Mon, Aug 28, 2000 at 09:46:31PM -0400, Nicolas Pitre wrote: > > Well... here's the plan: > > virt_to_phys()/phys_to_virt() become: > > #define __virt_to_phys(x) (x) > #define __phys_to_virt(x) (x) > > That gcc knows how to optimize! ;-) > Are you planning on using multiple NUMA nodes, or one node with multiple, discontiguous banks? From looking at the code, gaps in mem_map get returned to the free page pool, so I'm not sure what the advantage of having multiple NUMA nodes is; can anyone explain? miket unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Wed Aug 30 00:26:30 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13Tulv-0003qC-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Wed, 30 Aug 2000 00:26:27 +0100 Received: from [24.200.137.248] (helo=xanadu.gn.com) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13Tulu-0003q2-00 for linux-arm-kernel@lists.arm.linux.org.uk; Wed, 30 Aug 2000 00:26:27 +0100 Received: from localhost (nico@localhost) by xanadu.gn.com (8.9.3/8.9.3) with ESMTP id TAA31828; Tue, 29 Aug 2000 19:24:02 -0400 X-Authentication-Warning: xanadu.gn.com: nico owned process doing -bs Date: Tue, 29 Aug 2000 19:24:02 -0400 (EDT) From: Nicolas Pitre X-Sender: nico@xanadu.gn.com To: Erik Mouw cc: john+@cs.cmu.edu, linux@arm.linux.org.uk, ebunce@lhsl.com, bakker@thorgal.et.tudelft.nl, france@crl.dec.com, chris@blazie.com, a.melis@aino.nl, linux-arm-kernel@lists.arm.linux.org.uk Subject: Re: SA1100 DRAM sizes In-Reply-To: <200008292122.XAA22247@duteinh.et.tudelft.nl> Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk On Tue, 29 Aug 2000, Erik Mouw wrote: > On Tue, 29 Aug 2000 17:05:42 -0400, John G Dorsey wrote: > > I've tried it with pcnet_cs on Assabet and Neponset, and ide_cs on Assabet. I > > was surprised when I found: > > > > #define PCMCIA_IO_0_BASE 0xe0000000 > > #define PCMCIA_IO_1_BASE 0xe4000000 > > > > ...in asm/arch/ide.h. Owners of this driver for devices that don't go through > > Card Services should probably tell Nicolas what they want this code to look > > like now. > > That's my stuff (well, I recently cleaned it up), although it is in the > kernel since linux-2.2.x (or maybe longer). The Empeg, LART, and Victor > have an IDE interface mapped into PCMCIA IO space 0. It is independent of > the card services because it doesn't use card services, the drive is just > always there. What's your suggestion how it should look like? It should stay there, but only the address will change. Those are used for designs that actually have an hardwired IDE interface on the PCMCIA space, just like if it was an ISA bus. It shouldn't conflict with real PCMCIA implementations. Nicolas unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Wed Aug 30 00:31:00 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13Tupu-0003u4-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Wed, 30 Aug 2000 00:30:34 +0100 Received: from [24.200.137.248] (helo=xanadu.gn.com) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13Tupt-0003tw-00 for linux-arm-kernel@lists.arm.linux.org.uk; Wed, 30 Aug 2000 00:30:33 +0100 Received: from localhost (nico@localhost) by xanadu.gn.com (8.9.3/8.9.3) with ESMTP id TAA31833; Tue, 29 Aug 2000 19:29:21 -0400 X-Authentication-Warning: xanadu.gn.com: nico owned process doing -bs Date: Tue, 29 Aug 2000 19:29:21 -0400 (EDT) From: Nicolas Pitre X-Sender: nico@xanadu.gn.com To: John G Dorsey cc: Erik Mouw , linux@arm.linux.org.uk, ebunce@lhsl.com, bakker@thorgal.et.tudelft.nl, france@crl.dec.com, chris@blazie.com, a.melis@aino.nl, linux-arm-kernel@lists.arm.linux.org.uk Subject: Re: SA1100 DRAM sizes In-Reply-To: <39AC3E1D.67023D7A@cs.cmu.edu> Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk On Tue, 29 Aug 2000, John G Dorsey wrote: > Erik Mouw wrote: > > > That's my stuff (well, I recently cleaned it up), although it is in the > > kernel since linux-2.2.x (or maybe longer). > > I haven't tested ide_cs with any regularity since Erik Bunce first got it running > (2.3.99), but removing the hardwired base address (now in ide_init_hwif_ports()) > was part of that conversion. The code as it is currently written should fail when > run against CS because of the address arithmetic being performed. > > > The Empeg, LART, and Victor > > have an IDE interface mapped into PCMCIA IO space 0. It is independent of > > the card services because it doesn't use card services, the drive is just > > always there. What's your suggestion how it should look like? > > Once the static mapping is removed from the official tree, this code simply won't > work. I wanted to leave a small mapped area for those. This might be considered like a restricted ISA bus and some boards, like the empeg, use it to stick more than one device. Nicolas unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Wed Aug 30 01:06:49 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13TvOo-0004Ij-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Wed, 30 Aug 2000 01:06:38 +0100 Received: from [63.195.182.101] (helo=barbados.bluemug.com) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13TvOn-0004Id-00 for linux-arm-kernel@lists.arm.linux.org.uk; Wed, 30 Aug 2000 01:06:37 +0100 Received: from fiji.bluemug.com [192.168.50.32] by barbados.bluemug.com with esmtp (Exim 2.05 #1 (Debian)) id 13TvOA-00016o-00; Tue, 29 Aug 2000 17:05:58 -0700 Received: from miket by fiji.bluemug.com with local (Exim 3.12 #1 (Debian)) id 13TvOA-0007vM-00; Tue, 29 Aug 2000 17:05:58 -0700 Date: Tue, 29 Aug 2000 17:05:58 -0700 From: Mike Touloumtzis To: Mike Touloumtzis Cc: Nicolas Pitre , Russell King - ARM Linux Admin , John G Dorsey , Erik Mouw , ebunce@lhsl.com, bakker@thorgal.et.tudelft.nl, france@crl.dec.com, chris@blazie.com, a.melis@aino.nl, linux-arm-kernel@lists.arm.linux.org.uk Subject: Re: SA1100 DRAM sizes Message-ID: <20000829170558.B29614@fiji.bluemug.com> References: <200008282219.XAA20092@flint.arm.linux.org.uk> <20000829162353.A29614@fiji.bluemug.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii In-Reply-To: <20000829162353.A29614@fiji.bluemug.com>; from Mike Touloumtzis on Tue, Aug 29, 2000 at 04:23:53PM -0700 Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk On Tue, Aug 29, 2000 at 04:23:53PM -0700, Mike Touloumtzis wrote: > On Mon, Aug 28, 2000 at 09:46:31PM -0400, Nicolas Pitre wrote: > > > > Well... here's the plan: > > > > virt_to_phys()/phys_to_virt() become: > > > > #define __virt_to_phys(x) (x) > > #define __phys_to_virt(x) (x) > > > > That gcc knows how to optimize! ;-) > > > > Are you planning on using multiple NUMA nodes, or one node with multiple, > discontiguous banks? From looking at the code, gaps in mem_map get > returned to the free page pool, so I'm not sure what the advantage of > having multiple NUMA nodes is; can anyone explain? > Apologies for responding to my own post; after further thought, I realized that the SA1100 has much larger gaps between physical memory banks than the hardware I'm dealing with (two banks with only an 8MB gap between them); a mem_map array able to address all of this space as one node would be impractically large. I'm currently in the process of retooling my EP7211 port (whose memory handling was based on SA1100) with physical == virtual and two NUMA nodes, each with a local mem_map; I'll see if I can get that configuration working... miket unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Wed Aug 30 08:49:54 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13U2Zp-0003Af-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Wed, 30 Aug 2000 08:46:29 +0100 Received: from [130.161.42.1] (helo=duteinh.et.tudelft.nl) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13U2Zo-0003AZ-00 for linux-arm-kernel@lists.arm.linux.org.uk; Wed, 30 Aug 2000 08:46:28 +0100 Received: from dutein75 by duteinh.et.tudelft.nl (8.9.3/BB-2.0) id JAA23993; Wed, 30 Aug 2000 09:45:31 +0200 (MET DST) Message-Id: <200008300745.JAA23993@duteinh.et.tudelft.nl> From: Erik Mouw To: john+@cs.cmu.edu Cc: nico@cam.org, linux@arm.linux.org.uk, ebunce@lhsl.com, bakker@thorgal.et.tudelft.nl, france@crl.dec.com, chris@blazie.com, a.melis@aino.nl, linux-arm-kernel@lists.arm.linux.org.uk Subject: Re: SA1100 DRAM sizes In-Reply-To: <39AC3E1D.67023D7A@cs.cmu.edu> Organization: Eric conspiracy secret labs Date: Wed 30 Aug 2000 09:45:30 +0100 (MDT) X-Face: "!Z|1XjFg!3Iw)BwM^M\k)J5KHiPddz}#/Zw%?$&Q!xqob"OA|})3_@&BTTiQERo?/^g[Z<9P4R8:5C:aqq.cw!hotLc=+u-"+1rq@9$V^8PbOUlodyi_sejt+D`iqkLpx$>.aXl%vnfPIt"LK|/%>94_Zrm8jy?`_[RKt'TuoxUxfk(gMhY(n|DBBl~c$dhO>G%6}R4+DGX3|idBm4oU&UM/XUP4I9aoN{]`ax%B?rgD(I3OJ<62Y6L X-Mailer: XCmail 1.2devel - with PGP support, PGP engine version 0.5 (IRIX64) X-Mailerorigin: http://www.fsai.fh-trier.de/~schmitzj/Xclasses/XCmail/ X-Eric-Conspiracy: There is no conspiracy! X-Loop: erik@it.et.tudelft.nl MIME-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: 8bit Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk On Tue, 29 Aug 2000 18:50:05 -0400, John G Dorsey wrote: > Erik Mouw wrote: > >> That's my stuff (well, I recently cleaned it up), although it is in the >> kernel since linux-2.2.x (or maybe longer). > > I haven't tested ide_cs with any regularity since Erik Bunce first got it > running > (2.3.99), but removing the hardwired base address (now in > ide_init_hwif_ports()) > was part of that conversion. The code as it is currently written should fail > when > run against CS because of the address arithmetic being performed. The base address is no longer in ide_init_hwif_ports(). I send a patch to this list (or was it sa1100-linux?) that fixes it. I assume Nico will put it in his next patch. >> The Empeg, LART, and Victor >> have an IDE interface mapped into PCMCIA IO space 0. It is independent of >> the card services because it doesn't use card services, the drive is just >> always there. What's your suggestion how it should look like? > > Once the static mapping is removed from the official tree, this code simply > won't > work. Constructing the mapping on demand, as the sa1100 socket driver now does, > seems appropriate. I'm not sure where I'd do this in the code, though. I know > ide_init_hwif_ports() is inline on other architectures, but it's becoming > portly > on SA-1100. (Also, on SPARC, there's a warning in the code against using > ioremap() > in this routine, as it can be called with degenerate parameters.) Yes, some old stuff in the IDE code simply calls this function with "all the world is a PC" in mind. > I can also imagine a simple implementation of the low-level socket driver > operations that might allow these boards to just use Card Services. (Advantage: > consolidated resource management code. Disadvantage: needs ramdisk to load > ide_cs at boot.) That's not a clean solution, IMHO. The LART doesn't need Card Services, so why should the IDE driver depend on it? The idea is that it emulates ISA I/O space which is not larger than 64k. A small static mapping for this kind of devices would be nicer. Erik -- J.A.K. (Erik) Mouw, Information and Communication Theory Group, Department of Electrical Engineering, Faculty of Information Technology and Systems, Delft University of Technology, PO BOX 5031, 2600 GA Delft, The Netherlands Phone: +31-15-2783635 Fax: +31-15-2781843 Email: J.A.K.Mouw@its.tudelft.nl WWW: http://www-ict.its.tudelft.nl/~erik/ unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Wed Aug 30 09:12:26 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13U2wC-0003Qv-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Wed, 30 Aug 2000 09:09:36 +0100 Received: from [130.161.42.1] (helo=duteinh.et.tudelft.nl) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13U2wB-0003Qp-00 for linux-arm-kernel@lists.arm.linux.org.uk; Wed, 30 Aug 2000 09:09:35 +0100 Received: from dutein75 by duteinh.et.tudelft.nl (8.9.3/BB-2.0) id KAA24194; Wed, 30 Aug 2000 10:09:12 +0200 (MET DST) Message-Id: <200008300809.KAA24194@duteinh.et.tudelft.nl> From: Erik Mouw To: miket@bluemug.com Cc: nico@cam.org, linux@arm.linux.org.uk, john+@cs.cmu.edu, ebunce@lhsl.com, bakker@thorgal.et.tudelft.nl, france@crl.dec.com, chris@blazie.com, a.melis@aino.nl, linux-arm-kernel@lists.arm.linux.org.uk Subject: Re: SA1100 DRAM sizes In-Reply-To: <20000829170558.B29614@fiji.bluemug.com> Organization: Eric conspiracy secret labs Date: Wed 30 Aug 2000 10:09:11 +0100 (MDT) X-Face: "!Z|1XjFg!3Iw)BwM^M\k)J5KHiPddz}#/Zw%?$&Q!xqob"OA|})3_@&BTTiQERo?/^g[Z<9P4R8:5C:aqq.cw!hotLc=+u-"+1rq@9$V^8PbOUlodyi_sejt+D`iqkLpx$>.aXl%vnfPIt"LK|/%>94_Zrm8jy?`_[RKt'TuoxUxfk(gMhY(n|DBBl~c$dhO>G%6}R4+DGX3|idBm4oU&UM/XUP4I9aoN{]`ax%B?rgD(I3OJ<62Y6L X-Mailer: XCmail 1.2devel - with PGP support, PGP engine version 0.5 (IRIX64) X-Mailerorigin: http://www.fsai.fh-trier.de/~schmitzj/Xclasses/XCmail/ X-Eric-Conspiracy: There is no conspiracy! X-Loop: erik@it.et.tudelft.nl MIME-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: 8bit Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk On Tue, 29 Aug 2000 17:05:58 -0700, Mike Touloumtzis wrote: > Apologies for responding to my own post; after further thought, > I realized that the SA1100 has much larger gaps between physical > memory banks than the hardware I'm dealing with (two banks with > only an 8MB gap between them); a mem_map array able to address > all of this space as one node would be impractically large. Oh, but the current implementation would still do. The LART also has an 8MB gap between two 8MB regions in a single bank, and the current implementation works well. The trick is to mark the 8MB gap as reserved. Have a look at arch/arm/kernel/arch.c and search for LART. Erik -- J.A.K. (Erik) Mouw, Information and Communication Theory Group, Department of Electrical Engineering, Faculty of Information Technology and Systems, Delft University of Technology, PO BOX 5031, 2600 GA Delft, The Netherlands Phone: +31-15-2783635 Fax: +31-15-2781843 Email: J.A.K.Mouw@its.tudelft.nl WWW: http://www-ict.its.tudelft.nl/~erik/ unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Wed Aug 30 12:49:53 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13U6Mr-0005ew-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Wed, 30 Aug 2000 12:49:21 +0100 Received: from [210.184.5.125] (helo=pikachu.idthk.com) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13U6Mp-0005eR-00 for linux-arm-kernel@lists.arm.linux.org.uk; Wed, 30 Aug 2000 12:49:20 +0100 Received: from localhost (thomas@localhost) by pikachu.idthk.com (8.9.3/8.8.7) with ESMTP id TAA12917; Wed, 30 Aug 2000 19:49:44 +0800 Date: Wed, 30 Aug 2000 19:49:43 +0800 (CST) From: Tak-Shing Chan To: Tak-Shing Chan cc: linux-arm-kernel@lists.arm.linux.org.uk, sa1100-linux@pa.dec.com Subject: Re: test6-rmk5-np1 PCMCIA / UCB1200 audio? In-Reply-To: Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk Dear ARM Linux people, I apologize for replying to my own post--my plan of porting has to be canceled as I have to finish up some urgent tasks before I leave my present job. I guess I need to pull my name off the volunteers list for now... In the past six months I have learned a great deal about ARM Linux just by reading your discussions. Thanks for the advices and knowledge given to me during my stay! Cheers and farewell, Tak-Shing Chan On Mon, 28 Aug 2000, Tak-Shing Chan wrote: > Sure. I plan to port the latest Itsy drivers (Y2K version) > from 2.0.30 to 2.4.0-test6-rmk5-np1 (quite a tedious task but > hopefully I can start from Erik Mouw's port of an older version), > and rename them to the ones you proposed... I'll also make their > TS driver iPAQ-compatible. This seems to solve all the problems > at once. Is there anyone else already doing this? unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Wed Aug 30 17:41:46 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13UAvH-0000jG-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Wed, 30 Aug 2000 17:41:11 +0100 Received: from [206.207.108.63] (helo=ridgerun-lx.ridgerun.cxm) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13UAvF-0000j4-00 for linux-arm-kernel@lists.arm.linux.org.uk; Wed, 30 Aug 2000 17:41:10 +0100 Received: from ridgerun.com (IDENT:gmcnutt@gmcnutt-lx.ridgerun.cxm [192.168.1.17]) by ridgerun-lx.ridgerun.cxm (8.10.2/8.10.2) with ESMTP id e7UGePx07550 for ; Wed, 30 Aug 2000 10:40:25 -0600 Message-ID: <39AD38F9.E8733B66@ridgerun.com> Date: Wed, 30 Aug 2000 10:40:25 -0600 From: Gordon McNutt X-Mailer: Mozilla 4.75 [en] (X11; U; Linux 2.4.0-test5 i686) X-Accept-Language: en MIME-Version: 1.0 To: linux-arm-kernel@lists.arm.linux.org.uk Subject: FAQ or HOWTO Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk I'm a newbie trying to get a Linux kernel dowloaded and booting on an EBSA285. I've checked the FAQ at the www.arm.linux.org.uk web site and this mailing list's archives. Is there a FAQ or a HOWTO somewhere that explicitly describes the steps necessary to do what I want? Thanks, Gordon unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Thu Aug 31 04:44:38 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13ULFv-0008Mj-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Thu, 31 Aug 2000 04:43:11 +0100 Received: from [202.130.151.66] (helo=gw-hk1.philips.com) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13ULFs-0008Ma-00 for linux-arm-kernel@lists.arm.linux.org.uk; Thu, 31 Aug 2000 04:43:09 +0100 Received: from smtprelay-hk1.philips.com (localhost.philips.com [127.0.0.1]) by gw-hk1.philips.com with ESMTP id LAA12344 for ; Thu, 31 Aug 2000 11:43:03 +0800 (HKT) (envelope-from peter.mc.chen@philips.com) From: peter.mc.chen@philips.com Received: from smtprelay-asp1.philips.com(130.147.65.5) by gw-hk1.philips.com via mwrap (4.0a) id xma012342; Thu, 31 Aug 00 11:43:03 +0800 Received: from APLMS01.DIAMOND.PHILIPS.COM (aplms01sv1.diamond.philips.com [130.147.79.213]) by smtprelay-hk1.philips.com (8.9.3/8.8.5-1.2.2m-19990317) with ESMTP id LAA06670 for ; Thu, 31 Aug 2000 11:43:02 +0800 (HKT) Received: by APLMS01.DIAMOND.PHILIPS.COM (Soft-Switch LMS 4.0) with snapi via APAC id 0056920007262295; Thu, 31 Aug 2000 11:42:06 +0800 To: Message-ID: <0056920007262295000002L252*@MHS> Date: Thu, 31 Aug 2000 11:42:06 +0800 MIME-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1; name="MEMO 08/31/00 11:42:29" Content-Transfer-Encoding: quoted-printable Content-Disposition: inline Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk Hi: Now I am working on porting linux to SA1110 Assabet. I had done follow= ing: (1) download kernel and patch files to generate linux-2.4.0-test6-rmk5-= np2(advised from Erik, thanks) (2) using the binary toolchain "arm-linux-cross.tar.bz2" from "http://w= ww.lart.tudelft.nl/lartware/compile-tools/" to compile kernel. (3) follow make config_assabet, make coldconfig, make dep, ... to gerer= ate zImage. (4) download angelboot and minicom and jflash from "http://www.cs.cmu.e= du/~wearable/software/assabet.html" (5) use jflas to download angel.bin boundle with jflash. (6) setup minicom with "9600bps, 8 data bits, no parity, 1 stop bit, an= d no flow control", and reset Assabet. I can see some message like: "...(some characters strange) | Angel Debug Monitor for DB1110 : Seria= l/FIQ : MMU on, ...(I can't see because character wrapping problem in m= y minicom) (6) i try to use angelboot to download zImage. but how can i use? i try= many ways but the message never show on the minicom, also the image si= ze is zero. another question is how to set the r0 r1 register? Need i m= odify the kernel source code? Thanks a lot!=20 Peter Chen.= unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Thu Aug 31 16:41:02 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13UWRO-0008AI-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Thu, 31 Aug 2000 16:39:46 +0100 Received: from [128.2.10.81] (helo=smtp1.andrew.cmu.edu) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13UWRI-0008A4-00 for linux-arm-kernel@lists.arm.linux.org.uk; Thu, 31 Aug 2000 16:39:44 +0100 Received: from unix10.andrew.cmu.edu (UNIX10.ANDREW.CMU.EDU [128.2.15.14]) by smtp1.andrew.cmu.edu (8.9.3/8.9.3) with SMTP id LAA25079; Thu, 31 Aug 2000 11:39:18 -0400 (EDT) Date: Thu, 31 Aug 2000 11:39:18 -0400 (EDT) From: John G Dorsey Reply-To: John G Dorsey To: peter.mc.chen@philips.com cc: linux-arm-kernel@lists.arm.linux.org.uk Subject: Re: (questions about angelboot) In-Reply-To: <0056920007262295000002L252*@MHS> Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk On Thu, 31 Aug 2000 peter.mc.chen@philips.com wrote: > Now I am working on porting linux to SA1110 Assabet. Today is your lucky day! We've already ported it for you. =) > (6) setup minicom with "9600bps, 8 data bits, no parity, 1 stop bit, and > no flow control", and reset Assabet. I can see some message like: > "...(some characters strange) | Angel Debug Monitor for DB1110 : > Serial/FIQ : MMU on, ...(I can't see because character wrapping problem > in my minicom) After you reset your board, run angelboot first. The config file that appears on the Assabet page ("Software Setup" part 3, "Loading the Kernel") includes a line ("exec minicom") to start your terminal package after the kernel and ramdisk transfers complete. > (6) i try to use angelboot to download zImage. but how can i use? i try > many ways but the message never show on the minicom, also the image size > is zero. another question is how to set the r0 r1 register? Need i > modify the kernel source code? Again, the .angelrc on the Assabet page includes lines that begin "r0" and "r1" that cause Angel to load whatever values you specify into the named registers. No kernel tweaking needed. You mention that "the image size is zero." Do you mean the zImage file on your development workstation, after building (`make zImage`)? -jd unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk From owner-linux-arm-kernel@lists.arm.linux.org.uk Thu Aug 31 22:12:55 2000 Received: from majordomo by www.linux.org.uk with local (Exim 3.13 #1) id 13Ubcr-00024m-00 for linux-arm-kernel-outgoingx@www.linux.org.uk; Thu, 31 Aug 2000 22:11:57 +0100 Received: from [206.207.108.63] (helo=ridgerun-lx.ridgerun.cxm) by www.linux.org.uk with esmtp (Exim 3.13 #1) id 13Ubco-00024c-00 for linux-arm-kernel@lists.arm.linux.org.uk; Thu, 31 Aug 2000 22:11:56 +0100 Received: from ridgerun.com (IDENT:gmcnutt@gmcnutt-lx.ridgerun.cxm [192.168.1.17]) by ridgerun-lx.ridgerun.cxm (8.10.2/8.10.2) with ESMTP id e7VLAwx16331 for ; Thu, 31 Aug 2000 15:10:58 -0600 Message-ID: <39AEC9E1.BA0BEE30@ridgerun.com> Date: Thu, 31 Aug 2000 15:10:57 -0600 From: Gordon McNutt X-Mailer: Mozilla 4.75 [en] (X11; U; Linux 2.4.0-test5 i686) X-Accept-Language: en MIME-Version: 1.0 To: linux-arm-kernel@lists.arm.linux.org.uk Subject: angelboot & EBSA285 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk I'm trying to use angelboot with the SA-110 development board, the EBSA-285. If I run minicom I see what I expect to see ("Angel Debug Monitor..." etc). If I run angelboot instead I see it get through the "Negotiating..." message, after which it repeatedly prints "Error: received 0 when expecting start packet character". My hunch is that the only characters angelboot is reading back over the serial port are the welcome characters being printed (the ones I see on minicom), because eventually angelboot hangs waiting on serial_getc. My EBSA285 is running in Host Bridge mode on a stand-alone PCI backplane. The diagnostics report no errors. My angel options file looks like this: base 0x410000c0 entry 0x410000c0 r0 0x00000000 r1 0x00000004 device /dev/ttyS0 options "9600 8N1" baud 115200 otherfile ramdisk-img.gz otherbase 0x00008010 I'm sure the addresses here are wrong but that shouldn't prevent the serial connection from working. Any help will be greatly appreciated. Thanks, Gordon unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk