From owner-linux-arm-kernel@lists.arm.linux.org.uk Sat Nov 6 20:37:48 1999 Received: (from majordomo@localhost) by zen.ics.uwe.ac.uk (8.9.3/8.8.7) id UAA15944 for linux-arm-kernel-outgoing; Sat, 6 Nov 1999 20:37:48 GMT Received: from po6.andrew.cmu.edu (PO6.ANDREW.CMU.EDU [128.2.10.106]) by zen.ics.uwe.ac.uk (8.9.3/8.8.7) with ESMTP id UAA15940 for ; Sat, 6 Nov 1999 20:37:45 GMT Received: (from postman@localhost) by po6.andrew.cmu.edu (8.9.3/8.9.3) id PAA08872 for linux-arm-kernel@lists.arm.linux.org.uk; Sat, 6 Nov 1999 15:37:43 -0500 (EST) Received: via switchmail; Sat, 6 Nov 1999 15:37:43 -0500 (EST) Received: from unix14.andrew.cmu.edu via qmail ID ; Sat, 6 Nov 1999 15:37:10 -0500 (EST) Received: from unix14.andrew.cmu.edu via qmail ID ; Sat, 6 Nov 1999 15:37:10 -0500 (EST) Received: from mms.4.60.Jun.27.1996.03.02.53.sun4.51.EzMail.2.0.CUILIB.3.45.SNAP.NOT.LINKED.unix14.andrew.cmu.edu.sun4m.54 via MS.5.6.unix14.andrew.cmu.edu.sun4_51; Sat, 6 Nov 1999 15:37:10 -0500 (EST) Message-ID: Date: Sat, 6 Nov 1999 15:37:10 -0500 (EST) From: David Sean Friedman To: linux-arm-kernel@lists.arm.linux.org.uk Subject: bug in 2.2 arch/arm/Makefile Cc: Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk The file arch/arm/Makefile incorrectly sets HOSTCFLAGS to $(CFLAGS:-fomit-frame-pointer=), instead of $(HOSTCFLAGS:-fomit-frame-pointer=) In my case it broke the kernel compile since the cross compiler was gcc 2.95, which supports -fno-strict-aliasing, while the host compiler was gcc 2.7.2.3 which does not. This seems to be present in all 2.2 kernels. A patch follows. David Friedman --- diff -urN linux-orig/arch/arm/Makefile linux/arch/arm/Makefile --- linux-orig/arch/arm/Makefile Sat Nov 6 15:25:31 1999 +++ linux/arch/arm/Makefile Sat Nov 6 15:15:03 1999 @@ -128,7 +128,7 @@ CPP = $(CC) -E ARCHCC := $(word 1,$(CC)) GCCLIB := $(shell $(CC) $(CFLAGS) --print-libgcc-file-name) -HOSTCFLAGS := $(CFLAGS:-fomit-frame-pointer=) +HOSTCFLAGS := $(HOSTCFLAGS:-fomit-frame-pointer=) ifeq ($(CONFIG_FRAME_POINTER),y) CFLAGS := $(CFLAGS:-fomit-frame-pointer=) endif 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 Nov 8 09:02:01 1999 Received: (from majordomo@localhost) by zen.ics.uwe.ac.uk (8.9.3/8.8.7) id JAA18318 for linux-arm-kernel-outgoing; Mon, 8 Nov 1999 09:02:01 GMT Received: from rulhmpc56.leidenuniv.nl (devel@rulhmpc56.LeidenUniv.nl [132.229.96.156]) by zen.ics.uwe.ac.uk (8.9.3/8.8.7) with ESMTP id JAA18314 for ; Mon, 8 Nov 1999 09:01:57 GMT Received: from localhost (devel@localhost) by rulhmpc56.leidenuniv.nl (8.8.5/8.8.5) with SMTP id KAA03140; Mon, 8 Nov 1999 10:58:14 +0100 Date: Mon, 8 Nov 1999 10:58:14 +0100 (MET) From: kjdevel Reply-To: kjdevel To: linux-arm-kernel@lists.arm.linux.org.uk cc: Philip Blundell Subject: Re: arm-linux cross-gcc build 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 Hi, I'm trying to build the latest version of the gnu cross-compiling toolchain for arm-linux targets (in my case, on a i686-linux host, for a Psion Series 5 target) The files/versions I intend to use: binutils-2.9.5.0.19 gcc-2.95.2 glibc-2.1.2 Now my question (as first asked on linux-arm) is: > > which linux kernel versions are compatible with the above versions of > > the tool-chain? > > (in particular, how about linux-2.2.1, since it seems to be the > > de facto standard for the Psion) On Sun, 7 Nov 1999, Philip Blundell wrote: > The kernel experts can probably answer this more definitively (try the > linux-arm-kernel list), but I'm fairly sure that version will not work. > Both the 2.2 and 2.3 kernels in CVS seem to be OK though. Could anyone on this list give a more definitive answer? Another question is: >From the ftp site, the most recent version seems to be linux-2.2.13, with patch-2.2.13-rmk1.gz. Is that sufficient, or do I need other patches in addition? Thanks in advance! Klaasjan van Druten 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 Nov 10 16:45:56 1999 Received: (from majordomo@localhost) by zen.ics.uwe.ac.uk (8.9.3/8.8.7) id QAA10069 for linux-arm-kernel-outgoing; Wed, 10 Nov 1999 16:45:56 GMT Received: from uq.net.au (root@fox.uq.net.au [203.101.255.1]) by zen.ics.uwe.ac.uk (8.9.3/8.8.7) with ESMTP id QAA10063 for ; Wed, 10 Nov 1999 16:45:51 GMT Received: from bart.linux.bogus (dyn-10-165.dialin.uq.net.au [203.100.10.165]) by uq.net.au (8.9.3/8.9.3) with ESMTP id CAA06982 for ; Thu, 11 Nov 1999 02:45:43 +1000 (GMT+1000) Received: from localhost (benw@localhost) by bart.linux.bogus (8.9.3/8.9.3) with ESMTP id CAA15627 for ; Thu, 11 Nov 1999 02:46:10 +1000 X-Authentication-Warning: bart.linux.bogus: benw owned process doing -bs Date: Thu, 11 Nov 1999 02:46:10 +1000 (EST) From: Ben Williamson X-Sender: benw@bart.linux.bogus To: linux-arm-kernel@lists.arm.linux.org.uk Subject: ARM7s can be ARMv4 too 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 Disclaimer: I'm still back in 2.2.1 land, so if this already exists in newer kernels/patches, please beat me with a stick. I spent the last few days hacking on Werner's crash+burn-26 to make it boot on an EP7211. (More info posted on linux-7110.) This part contains an ARM720T core, ie an ARM7 with architecture v4 (and Thumb). Various spots in the kernel seem to assert that v4 implies StrongARM, which is not the case here. I tried to do what looked reasonable and it seems to work so far; my patch is now available for your viewing pleasure: ftp://ftp.netwinder.org/users/b/benw/ep7211/cb26-benw1.gz I'd welcome any feedback on what I've done, I'm learning a lot as I go along and I'm sure that some of this should be done differently to coexist happily with the RPC-that-might-have-a-SA110-card world. I'm particularly uncertain of putting #ifdef CONFIG_ARCH_ARMV4 in proc-arm6,7.S to deal with the MMU differences. Suggestions? Phil: You've mentioned playing with support for Thumb user-space binaries. Do you have any hardware to do that on yet? I looked in the latest philb and rmk patches and couldn't find any mention of ARM720T, ARM7v4 etc. I'm behind a !@#$% modem now so I haven't managed to download a 2.3.x kernel yet. *sniff* I miss DSL. Cheers, - Ben. ------------------------------------------------------------------- Ben Williamson benw@pobox.com http://www.pobox.com/~benw/ 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 Nov 10 19:12:22 1999 Received: (from majordomo@localhost) by zen.ics.uwe.ac.uk (8.9.3/8.8.7) id TAA17079 for linux-arm-kernel-outgoing; Wed, 10 Nov 1999 19:12:22 GMT Received: from caramon.arm.linux.org.uk (root@p16-cordelia-gui.tch.enablis.net [212.250.233.16]) by zen.ics.uwe.ac.uk (8.9.3/8.8.7) with ESMTP id TAA17073 for ; Wed, 10 Nov 1999 19:12:17 GMT Received: from raistlin.arm.linux.org.uk (linux@raistlin [192.168.0.3]) by caramon.arm.linux.org.uk (8.9.3/8.9.3) with ESMTP id TAA00622; Wed, 10 Nov 1999 19:12:14 GMT From: Russell King - ARM Linux Admin Received: (from linux@localhost) by raistlin.arm.linux.org.uk (8.7.4/8.7.3) id TAA12265; Wed, 10 Nov 1999 19:11:15 GMT Message-Id: <199911101911.TAA12265@raistlin.arm.linux.org.uk> Subject: Re: ARM7s can be ARMv4 too To: benw@pobox.com (Ben Williamson) Date: Wed, 10 Nov 1999 19:11:14 +0000 (GMT) Cc: linux-arm-kernel@lists.arm.linux.org.uk In-Reply-To: from "Ben Williamson" at Nov 11, 99 02:46:10 am X-Mailer: ELM [version 2.4 PL24] 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 Ben Williamson writes: > I spent the last few days hacking on Werner's crash+burn-26 to make it > boot on an EP7211. (More info posted on linux-7110.) This part contains > an ARM720T core, ie an ARM7 with architecture v4 (and Thumb). Oh dear. I suppose something crazy like this was bound to happen some day. > Various spots in the kernel seem to assert that v4 implies StrongARM, > which is not the case here. Which spots are these? As far as I am aware, the only places which v3/v4 matter are in arch/arm/Makefile and arch/arm/mm/proc-*.S. The correct way to handle this problem would be to create another proc-*.S, rather than breaking the ARM710 support (which, incidentally is v3). > with the MMU differences. Suggestions? See above. In later kernels (can't remember which version atm but I think it's v2.3), the proc-*.S files are selectively compiled in, so you don't have to have all the different processors compiled in. _____ |_____| ------------------------------------------------- ---+---+- | | 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 Wed Nov 10 19:33:23 1999 Received: (from majordomo@localhost) by zen.ics.uwe.ac.uk (8.9.3/8.8.7) id TAA18036 for linux-arm-kernel-outgoing; Wed, 10 Nov 1999 19:33:23 GMT Received: from inkvine.fluff.org (postfix@inkvine.fluff.org [212.219.8.163]) by zen.ics.uwe.ac.uk (8.9.3/8.8.7) with ESMTP id TAA18032 for ; Wed, 10 Nov 1999 19:33:21 GMT Received: from localhost (localhost [127.0.0.1]) by inkvine.fluff.org (Postfix) with ESMTP id E2A154681B for ; Wed, 10 Nov 1999 19:33:22 +0000 (GMT) Date: Wed, 10 Nov 1999 19:33:22 +0000 (GMT) From: Chris Rutter Reply-To: chris@fluff.org To: linux-arm-kernel@lists.arm.linux.org.uk Subject: Curtis-Mathes UniView 7500FE-based box 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 Curtis-Mathes UniView is a box best forgotten: it contains an ARM 7500FE, and enough standard hardware that RISC OS seems to run without much modification on it. What would be required to get Linux up and running on this box? I see a 7500 arch in the 2.3 tree, but that seems to have numbered IRQs and references to ISA and so forth, so it looks like it has a more PCish bus and wouldn't be suitable for this very Acorn-ish machine. On that subject, what exactly constitutes an architecture (as in `arch-rpc', `arc-cl7500', and so on)? I see these various architecutres which then contain, often, #ifdefs to do individual things based on processor type, so it's not exactly clear *when* a new architecture directory should be created (consider these two cases in point: Riscstation and UniView). Ta. -- Chris ( http://www.fluff.org/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 Nov 10 21:24:15 1999 Received: (from majordomo@localhost) by zen.ics.uwe.ac.uk (8.9.3/8.8.7) id VAA22447 for linux-arm-kernel-outgoing; Wed, 10 Nov 1999 21:24:15 GMT Received: from oracle.clara.net (oracle.clara.net [195.8.69.94]) by zen.ics.uwe.ac.uk (8.9.3/8.8.7) with ESMTP id VAA22443 for ; Wed, 10 Nov 1999 21:24:11 GMT Received: from [212.126.134.109] (helo=tardis) by oracle.clara.net with esmtp (Exim 2.12 #2) id 11lfDq-0003bQ-00; Wed, 10 Nov 1999 21:24:07 +0000 Date: Wed, 10 Nov 1999 21:18:47 +0000 (GMT) From: Dave Gilbert X-Sender: gilbertd@tardis.home.dave To: Chris Rutter cc: linux-arm-kernel@lists.arm.linux.org.uk Subject: Re: Curtis-Mathes UniView 7500FE-based box 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, 10 Nov 1999, Chris Rutter wrote: > The Curtis-Mathes UniView is a box best forgotten: it contains an ARM > 7500FE, and enough standard hardware that RISC OS seems to run without > much modification on it. Define much modification :-) > On that subject, what exactly constitutes an architecture (as in > `arch-rpc', `arc-cl7500', and so on)? I see these various architecutres > which then contain, often, #ifdefs to do individual things based on > processor type, so it's not exactly clear *when* a new architecture > directory should be created (consider these two cases in point: > Riscstation and UniView). Well each architecture may consist of one or more machines which are mostly alike; small differences can be taken care of within the architectures. Ideally a kernel built for a particular architecture should boot on any machine included in that architecture. So if you Curtis thingamy box is basically the same as an existing architecture with a couple of tweaks (preferably that you can figure out at run time) then stay with that architecture. Dave -- -------------------------------------------------------------------- / Dr. David Alan Gilbert | Running Linux on | Happy \ \ gro.gilbert @ treblig.org | Alpha, x86, ARM and SPARC | In Hex / ____________________________|___ http://www.treblig.clara.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 Nov 11 00:24:42 1999 Received: (from majordomo@localhost) by zen.ics.uwe.ac.uk (8.9.3/8.8.7) id AAA30493 for linux-arm-kernel-outgoing; Thu, 11 Nov 1999 00:24:42 GMT Received: from caramon.arm.linux.org.uk (root@p39-magpie-gui.tch.enablis.net [194.168.180.39]) by zen.ics.uwe.ac.uk (8.9.3/8.8.7) with ESMTP id AAA30489 for ; Thu, 11 Nov 1999 00:24:39 GMT Received: from raistlin.arm.linux.org.uk (linux@raistlin [192.168.0.3]) by caramon.arm.linux.org.uk (8.9.3/8.9.3) with ESMTP id AAA01736; Thu, 11 Nov 1999 00:24:36 GMT From: Russell King - ARM Linux Admin Received: (from linux@localhost) by raistlin.arm.linux.org.uk (8.7.4/8.7.3) id AAA14086; Thu, 11 Nov 1999 00:23:36 GMT Message-Id: <199911110023.AAA14086@raistlin.arm.linux.org.uk> Subject: Re: Curtis-Mathes UniView 7500FE-based box To: chris@fluff.org Date: Thu, 11 Nov 1999 00:23:35 +0000 (GMT) Cc: linux-arm-kernel@lists.arm.linux.org.uk In-Reply-To: from "Chris Rutter" at Nov 10, 99 07:33:22 pm X-Mailer: ELM [version 2.4 PL24] 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: > What would be required to get Linux up and running on this box? I see > a 7500 arch in the 2.3 tree, but that seems to have numbered IRQs and > references to ISA and so forth, so it looks like it has a more PCish > bus and wouldn't be suitable for this very Acorn-ish machine. There is no 7500 stuff in the 2.3 tree. > On that subject, what exactly constitutes an architecture (as in > `arch-rpc', `arc-cl7500', and so on)? I see these various architecutres > which then contain, often, #ifdefs to do individual things based on > processor type, so it's not exactly clear *when* a new architecture > directory should be created (consider these two cases in point: > Riscstation and UniView). In general, if an architecture requires a load of #ifdefs in its architecture specific directories, and there are a lot of them, then it does not warrant living together. With a "correct" design, there should be zero #ifdefs in the architecture code since we have a very efficient optimised way of causing GCC to "comment out" code using the following: if (machine_is_xxx()) { } At compile time, if support for machine xxx is not requested, it will be evaluated as zero, and GCC will optimise out the unreachable code. If, however, it is the only machine type selected, it will be evaluated as non-zero, and GCC will optimise the condition away and include the code. If multiple machines are selected, GCC will leave the code in along with the condition. There are other factors that will affect your choice. In general, I do not like seeing the virt_to_phys() macros be anything other than constant mathematical expressions (which will be a single instruction) since page table walking uses these heavily. Lookup tables and functions occupy many instructions, and will impact heavily the speed at which page tables can be traversed (they have to convert physical->virtual for every second-level table). Hope this helps. _____ |_____| ------------------------------------------------- ---+---+- | | 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 Thu Nov 11 11:17:06 1999 Received: (from majordomo@localhost) by zen.ics.uwe.ac.uk (8.9.3/8.8.7) id LAA24203 for linux-arm-kernel-outgoing; Thu, 11 Nov 1999 11:17:06 GMT Received: from inkvine.fluff.org (postfix@inkvine.fluff.org [212.219.8.163]) by zen.ics.uwe.ac.uk (8.9.3/8.8.7) with ESMTP id LAA24179 for ; Thu, 11 Nov 1999 11:17:03 GMT Received: from localhost (localhost [127.0.0.1]) by inkvine.fluff.org (Postfix) with ESMTP id C61504681B for ; Thu, 11 Nov 1999 11:17:03 +0000 (GMT) Date: Thu, 11 Nov 1999 11:17:03 +0000 (GMT) From: Chris Rutter Reply-To: chris@fluff.org To: linux-arm-kernel@lists.arm.linux.org.uk Subject: ADFS patches to add ",xxx" typing 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 In a fit of desperation (I needed to get an ADFS image looking like normal ADFS over NFS) I added some code to the 2.2 ADFS stuff to make it append ",xxx" to indicate file types to files. I have no idea whether I've done this well, or optimally, but it does appear to work. The code hinges around adding the extensions in readdir(), and then catching look-up of dentry structs with the ",xxx" extension, where it shaves off the extension before generating the inode. This way, accessing both "File" and "File,xxx" works. It adds an options bitfield to the ADFS superblock, and it turns on support when it comes across "xxx" in the options. Hence use `-o xxx' as a parameter to mount to turn it on. The patch is at: http://www.inkvine.fluff.org/~chris/ftp/linux-2.2.12-adfs-xxx.patch ftp.inkvine.fluff.org:/users/chris/linux-2.2.12-adfs-xxx.patch (It also adds `unsigned' to a couple of chars, which means that the map-validation code doesn't fail on i386. However, I still have many errors about lost fragments and hole-y directories -- perhaps there's some other porting issue to be addressed in the code. Anyway, it now appears to work basically fine on my i386 Linux box.) Problems: * `,' seems to be a valid FileCore leafname character, thus: * Files that end with ",xxx" where all `x' are valid hex digits will be truncated before a look-up ever gets performed. The kludge for this is to try and access a file called "File,xxx,yyy", where the ",yyy" will be stripped off, leaving the look-up routine with the original filename. Normally, of course, you would only access files that you see, and so the kludge would happen automatically, but if you have a symlink lying around, an you just happened to have mounted your image with "xxx" today... * Directories aren't masked out in the look-up routine, and so munging will (incorrectly) be done; same kludge as above applies. * It doesn't cope with un-typed files, but then again, does anyone really use them these days? A possible fix to the first problem would be to prevent further exten- sion if the file name already ends in a valid ",xxx" contruction, possibly with the extra clause that the "xxx" must match the file's type. This would allow NFS-alike files on ADFS partitions to appear correctly over NFS. These files would then need to be flagged as such in the dentry, or alternatively if the first look-up fails, a second with the preivously-removed extension could be performed. This, however, falls over when a valid file exists. Consider: `foo', `foo,xxx' and `foo,xxx,yyy' (the final one being the extended one). -- Chris ( http://www.fluff.org/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 Nov 11 13:14:47 1999 Received: (from majordomo@localhost) by zen.ics.uwe.ac.uk (8.9.3/8.8.7) id NAA32182 for linux-arm-kernel-outgoing; Thu, 11 Nov 1999 13:14:47 GMT Received: from aleph.locsoft.com (root@[212.240.65.10]) by zen.ics.uwe.ac.uk (8.9.3/8.8.7) with ESMTP id NAA32178 for ; Thu, 11 Nov 1999 13:14:38 GMT Received: from uphill ([212.240.65.50]) by aleph.locsoft.com (8.8.8/8.8.7) with SMTP id NAA22712 for ; Thu, 11 Nov 1999 13:37:12 GMT Message-ID: <008a01bf2c46$29dee6b0$3241f0d4@uphill.locsoft.com> From: "Paul Robertson" To: Subject: booting kernel on EBSA285 Date: Thu, 11 Nov 1999 13:10:44 -0000 MIME-Version: 1.0 Content-Type: multipart/alternative; boundary="----=_NextPart_000_0087_01BF2C46.29A7F830" X-Priority: 3 X-MSMail-Priority: Normal X-Mailer: Microsoft Outlook Express 4.72.3110.5 X-MimeOLE: Produced By Microsoft MimeOLE V4.72.3110.3 Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk This is a multi-part message in MIME format. ------=_NextPart_000_0087_01BF2C46.29A7F830 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable I have a zImage configured for EBSA-285, and some boot code which loads = the zImage at 0x8000, then executes it. decompress_kernel() outputs its messages: Uncompressing Linux...done. Now booting the kernel After that nothing. Any ideas how I could move on from here? The boot code is Russell King's Bios 1.05, but hacked so that it doesn't = use any peripheral devices, simply copies the zImage from Flash into = RAM. Paul Robertson LocSoft Limited "The key to your embedded solutions" www.locsoft.com Phone +44 (0) 1793 432557=20 Fax +44 (0) 1793 432727 ------=_NextPart_000_0087_01BF2C46.29A7F830 Content-Type: text/html; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable
I have a zImage configured for = EBSA-285, and=20 some boot code which loads the zImage at 0x8000, then executes = it.
decompress_kernel() outputs its = messages:
 
Uncompressing Linux...done.
Now booting the kernel
 
After that nothing.
Any ideas how I could move on from = here?
 
The boot code is Russell King's Bios = 1.05, but=20 hacked so that it doesn't use any peripheral devices, simply copies the = zImage=20 from Flash into RAM.
 
Paul Robertson
LocSoft = Limited
"The=20 key to your embedded solutions"
www.locsoft.com
Phone +44 (0) = 1793 432557=20
Fax     +44 (0) 1793=20 432727
------=_NextPart_000_0087_01BF2C46.29A7F830-- 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 Nov 12 09:03:27 1999 Received: (from majordomo@localhost) by zen.ics.uwe.ac.uk (8.9.3/8.8.7) id JAA08210 for linux-arm-kernel-outgoing; Fri, 12 Nov 1999 09:03:27 GMT Received: from sparkle.Generation.NET (sparkle.Generation.NET [205.205.119.4]) by zen.ics.uwe.ac.uk (8.9.3/8.8.7) with ESMTP id JAA08203 for ; Fri, 12 Nov 1999 09:03:24 GMT Received: from dual (generic253.197.205.205.IN-ADDR.ARPA [205.205.197.253] (may be forged)) by sparkle.Generation.NET (8.9.3/8.9.3) with SMTP id KAA25424 for ; Thu, 11 Nov 1999 10:46:10 -0500 (EST) From: Francois Desloges To: linux-arm-kernel@lists.arm.linux.org.uk Subject: Fwd: Re: map file etc... Date: Thu, 11 Nov 1999 10:41:25 -0500 Content-Type: text/plain MIME-Version: 1.0 Message-Id: <99111110431201.00742@dual> Content-Transfer-Encoding: 8bit Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk Ok, I've build that loadmap thing for i386, hack it a little bit in order for it to create a map on a mounted drive partition which is gone to be / once transfered to the EBSA. This partition has nothing else then a /boot with my arm-linux kernel and a /etc with a boot.conf. Now, the boot sector that loadmap seems to write is a copy of the boot.map map file. Well, the BIOS expects to see two things in this boot sector: 0x80 at address 0x1be, and the address of the first sector of the boot.map map file. I don't think this is what loadmap write in the boot sector, but this is definitely what mkboot (from the BIOS) can write in the boot sector, provided that I can pass it as an argument, the sector number of the first sector of boot.map. Therefore: Anybody knows if there is an application I can run on a shell or anything else that can give me the sector number of the first sector of a file? Thanks! -- François Desloges fd@vipswitch.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 Nov 12 09:16:18 1999 Received: (from majordomo@localhost) by zen.ics.uwe.ac.uk (8.9.3/8.8.7) id JAA08828 for linux-arm-kernel-outgoing; Fri, 12 Nov 1999 09:16:18 GMT Received: from uq.net.au (root@fox.uq.net.au [203.101.255.1]) by zen.ics.uwe.ac.uk (8.9.3/8.8.7) with ESMTP id JAA08819 for ; Fri, 12 Nov 1999 09:16:13 GMT Received: from bart.linux.bogus (dyn-12-102.dialin.uq.net.au [203.100.12.102]) by uq.net.au (8.9.3/8.9.3) with ESMTP id KAA13841 for ; Fri, 12 Nov 1999 10:23:33 +1000 (GMT+1000) Received: from localhost (benw@localhost) by bart.linux.bogus (8.9.3/8.9.3) with ESMTP id KAA01373 for ; Fri, 12 Nov 1999 10:24:02 +1000 X-Authentication-Warning: bart.linux.bogus: benw owned process doing -bs Date: Fri, 12 Nov 1999 10:24:02 +1000 (EST) From: Ben Williamson X-Sender: benw@bart.linux.bogus To: linux-arm-kernel@lists.arm.linux.org.uk Subject: Re: ARM7s can be ARMv4 too In-Reply-To: <199911101911.TAA12265@raistlin.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, 10 Nov 1999, Russell King - ARM Linux Admin wrote: > The correct way to handle this problem would be to create another > proc-*.S, rather than breaking the ARM710 support (which, incidentally > is v3). Right. To this end I put #ifdef CONFIG_ARCH_ARMV4 around my changes in proc-arm6,7.S. Here's a relevant bit of patch: --- ep7211/linux/arch/arm/mm/proc-arm6,7.S:1.1.1.1 Sat Nov 6 10:10:18 1999 +++ ep7211/linux/arch/arm/mm/proc-arm6,7.S Tue Nov 9 13:02:58 1999 @@ -22,7 +22,7 @@ */ _arm6_7_flush_cache: mov r0, #0 - mcr p15, 0, r0, c7, c0, 0 @ flush cache + mcr p15, 0, r0, c7, c7, 0 @ flush cache (v3 or v4) _arm6_7_null: mov pc, lr @@ -33,7 +33,11 @@ */ _arm6_7_flush_tlb_all: mov r0, #0 - mcr p15, 0, r0, c5, c0, 0 @ flush TLB +#ifdef CONFIG_ARCH_ARMV4 + mcr p15, 0, r0, c8, c7, 0 @ flush TLB (v4) +#else + mcr p15, 0, r0, c5, c0, 0 @ flush TLB (v3) +#endif mov pc, lr Does this look sufficient, or do you want it in a seperate file so that one kernel could support both ARM7v3 and ARM7v4? > See above. In later kernels (can't remember which version atm but I think it's > v2.3), the proc-*.S files are selectively compiled in, so you don't have to > have all the different processors compiled in. Cool, then I won't mess around with this too much until I get onto 2.3. Thanks, - Ben. ------------------------------------------------------------------- Ben Williamson benw@pobox.com http://www.pobox.com/~benw/ 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 Nov 12 09:26:45 1999 Received: (from majordomo@localhost) by zen.ics.uwe.ac.uk (8.9.3/8.8.7) id JAA09254 for linux-arm-kernel-outgoing; Fri, 12 Nov 1999 09:26:45 GMT Received: from caramon.arm.linux.org.uk (root@p18-robin-gui.tch.enablis.net [194.168.180.78]) by zen.ics.uwe.ac.uk (8.9.3/8.8.7) with ESMTP id JAA09249 for ; Fri, 12 Nov 1999 09:26:37 GMT Received: from raistlin.arm.linux.org.uk (linux@raistlin [192.168.0.3]) by caramon.arm.linux.org.uk (8.9.3/8.9.3) with ESMTP id UAA05148 for ; Thu, 11 Nov 1999 20:32:34 GMT From: Russell King - ARM Linux Admin Received: (from linux@localhost) by raistlin.arm.linux.org.uk (8.7.4/8.7.3) id UAA17312 for linux-arm-kernel@lists.arm.linux.org.uk; Thu, 11 Nov 1999 20:31:31 GMT Message-Id: <199911112031.UAA17312@raistlin.arm.linux.org.uk> Subject: Development on 2.2.xx To: linux-arm-kernel@lists.arm.linux.org.uk Date: Thu, 11 Nov 1999 20:31:30 +0000 (GMT) X-Mailer: ELM [version 2.4 PL24] 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, I notice that most people seem to be developing on the 2.2.xx platform. While this is a stable kernel base, it will leave you wide open for duplication of effort and extra effort if you want your code to work with 2.3 and eventually 2.4. 2.3 is now stable enough to develop and test new code against. Please bear this in mind when developing new code. _____ |_____| ------------------------------------------------- ---+---+- | | 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 Fri Nov 12 09:26:42 1999 Received: (from majordomo@localhost) by zen.ics.uwe.ac.uk (8.9.3/8.8.7) id JAA09259 for linux-arm-kernel-outgoing; Fri, 12 Nov 1999 09:26:42 GMT Received: from caramon.arm.linux.org.uk (root@p18-robin-gui.tch.enablis.net [194.168.180.78]) by zen.ics.uwe.ac.uk (8.9.3/8.8.7) with ESMTP id JAA09253 for ; Fri, 12 Nov 1999 09:26:39 GMT Received: from raistlin.arm.linux.org.uk (linux@raistlin [192.168.0.3]) by caramon.arm.linux.org.uk (8.9.3/8.9.3) with ESMTP id TAA04996; Thu, 11 Nov 1999 19:52:42 GMT From: Russell King - ARM Linux Admin Received: (from linux@localhost) by raistlin.arm.linux.org.uk (8.7.4/8.7.3) id TAA17077; Thu, 11 Nov 1999 19:51:39 GMT Message-Id: <199911111951.TAA17077@raistlin.arm.linux.org.uk> Subject: Re: ADFS patches to add ",xxx" typing To: chris@fluff.org Date: Thu, 11 Nov 1999 19:51:39 +0000 (GMT) Cc: linux-arm-kernel@lists.arm.linux.org.uk In-Reply-To: from "Chris Rutter" at Nov 11, 99 11:17:03 am X-Mailer: ELM [version 2.4 PL24] 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: > (It also adds `unsigned' to a couple of chars, which means that the > map-validation code doesn't fail on i386. However, I still have many > errors about lost fragments and hole-y directories -- perhaps there's > some other porting issue to be addressed in the code. Anyway, it now > appears to work basically fine on my i386 Linux box.) Upgrade to 2.2.13. It's been fixed for several months now. If you still get lost fragments and hole-y directories, then check your map. Also, don't develop ADFS on 2.2 - it's outdated wrt 2.3 by a long way. 2.3 has preliminary F+ and write support (which it has had it for about 4 months). _____ |_____| ------------------------------------------------- ---+---+- | | 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 Fri Nov 12 09:26:49 1999 Received: (from majordomo@localhost) by zen.ics.uwe.ac.uk (8.9.3/8.8.7) id JAA09267 for linux-arm-kernel-outgoing; Fri, 12 Nov 1999 09:26:49 GMT Received: from caramon.arm.linux.org.uk (root@p18-robin-gui.tch.enablis.net [194.168.180.78]) by zen.ics.uwe.ac.uk (8.9.3/8.8.7) with ESMTP id JAA09258 for ; Fri, 12 Nov 1999 09:26:41 GMT Received: from raistlin.arm.linux.org.uk (linux@raistlin [192.168.0.3]) by caramon.arm.linux.org.uk (8.9.3/8.9.3) with ESMTP id TAA05006; Thu, 11 Nov 1999 19:55:34 GMT From: Russell King - ARM Linux Admin Received: (from linux@localhost) by raistlin.arm.linux.org.uk (8.7.4/8.7.3) id TAA17115; Thu, 11 Nov 1999 19:54:31 GMT Message-Id: <199911111954.TAA17115@raistlin.arm.linux.org.uk> Subject: Re: booting kernel on EBSA285 To: paul.robertson@locsoft.com (Paul Robertson) Date: Thu, 11 Nov 1999 19:54:30 +0000 (GMT) Cc: linux-arm-kernel@lists.arm.linux.org.uk In-Reply-To: <008a01bf2c46$29dee6b0$3241f0d4@uphill.locsoft.com> from "Paul Robertson" at Nov 11, 99 01:10:44 pm X-Mailer: ELM [version 2.4 PL24] 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 Paul Robertson writes: > I have a zImage configured for EBSA-285, and some boot code which loads > the zImage at 0x8000, then executes it. Is it an EBSA285 board that you're trying to get it running on, or an EBSA285-alike board? > Uncompressing Linux...done. > Now booting the kernel > > After that nothing. > Any ideas how I could move on from here? >From what's been discussed previously when this happens, and it's running on an EBSA285-alike board, it normally turns out to be missing pull-up resistors on the PCI arbiter signals (iirc). > The boot code is Russell King's Bios 1.05, but hacked so that it doesn't > use any peripheral devices, simply copies the zImage from Flash into > RAM. I do hope that it initialises the DC21285 still. If not, this is probably your problem. _____ |_____| ------------------------------------------------- ---+---+- | | 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 Fri Nov 12 19:00:21 1999 Received: (from majordomo@localhost) by zen.ics.uwe.ac.uk (8.9.3/8.8.7) id TAA05722 for linux-arm-kernel-outgoing; Fri, 12 Nov 1999 19:00:21 GMT Received: from sparkle.Generation.NET (sparkle.Generation.NET [205.205.119.4]) by zen.ics.uwe.ac.uk (8.9.3/8.8.7) with ESMTP id TAA05718 for ; Fri, 12 Nov 1999 19:00:18 GMT Received: from dual (generic253.197.205.205.IN-ADDR.ARPA [205.205.197.253] (may be forged)) by sparkle.Generation.NET (8.9.3/8.9.3) with SMTP id NAA26351; Fri, 12 Nov 1999 13:59:49 -0500 (EST) From: Francois Desloges To: Russell King - ARM Linux Admin Subject: Re: booting kernel on EBSA285 Date: Fri, 12 Nov 1999 13:38:21 -0500 Content-Type: text/plain References: <199911111954.TAA17115@raistlin.arm.linux.org.uk> In-Reply-To: <199911111954.TAA17115@raistlin.arm.linux.org.uk> Cc: linux-arm-kernel@lists.arm.linux.org.uk MIME-Version: 1.0 Message-Id: <99111213565400.04589@dual> Content-Transfer-Encoding: 8bit Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk On Thu, 11 Nov 1999, Russell King - ARM Linux Admin wrote: > Paul Robertson writes: > > I have a zImage configured for EBSA-285, and some boot code which loads > > the zImage at 0x8000, then executes it. > > > > Uncompressing Linux...done. > > Now booting the kernel > > > > After that nothing. > > Any ideas how I could move on from here? Wow! I've got the exact same situation here! On our two different systems. First system is an EBSA plugged in its backplane (which provide the Arbiter) Second system is a one board solution including the PCI slots. Since we don't need the X-Bus, we have plugged 21285 PCI Bus Master Request signal (req_l) directly into 21285 Bus Master Grant signal (gnt_l), so that 21285 is the Arbiter. > From what's been discussed previously when this happens, and it's running > on an EBSA285-alike board, it normally turns out to be missing pull-up > resistors on the PCI arbiter signals (iirc). If a system has the 21285 as the Arbiter, is the kernel configured for EBSA able to deal with this or I must configure it otherwise (CATS?) > > The boot code is Russell King's Bios 1.05, but hacked so that it doesn't > > use any peripheral devices, simply copies the zImage from Flash into > > RAM. > > I do hope that it initialises the DC21285 still. If not, this is probably > your problem. I load the kernel from a RawPartition device on an ide drive (in a hacked version of ide_load() in ide.c), so I guess the BIOS normal PCI initialization has been done at this point? Where is this basic 21285 initialization performed in the BIOS ? -- François Desloges fd@vipswitch.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 Nov 12 21:55:10 1999 Received: (from majordomo@localhost) by zen.ics.uwe.ac.uk (8.9.3/8.8.7) id VAA13358 for linux-arm-kernel-outgoing; Fri, 12 Nov 1999 21:55:10 GMT Received: from caramon.arm.linux.org.uk (root@p11-cordelia-gui.tch.enablis.net [212.250.233.11]) by zen.ics.uwe.ac.uk (8.9.3/8.8.7) with ESMTP id VAA13323 for ; Fri, 12 Nov 1999 21:55:02 GMT Received: from raistlin.arm.linux.org.uk (linux@raistlin [192.168.0.3]) by caramon.arm.linux.org.uk (8.9.3/8.9.3) with ESMTP id VAA10295; Fri, 12 Nov 1999 21:54:54 GMT From: Russell King - ARM Linux Admin Received: (from linux@localhost) by raistlin.arm.linux.org.uk (8.7.4/8.7.3) id VAA22355; Fri, 12 Nov 1999 21:53:46 GMT Message-Id: <199911122153.VAA22355@raistlin.arm.linux.org.uk> Subject: Re: booting kernel on EBSA285 To: fd@vipswitch.com (Francois Desloges) Date: Fri, 12 Nov 1999 21:53:46 +0000 (GMT) Cc: linux-arm-kernel@lists.arm.linux.org.uk In-Reply-To: <99111213565400.04589@dual> from "Francois Desloges" at Nov 12, 99 01:38:21 pm X-Mailer: ELM [version 2.4 PL24] 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 Francois Desloges writes: > First system is an EBSA plugged in its backplane (which provide the Arbiter) > Second system is a one board solution including the PCI slots. Since we don't > need the X-Bus, we have plugged 21285 PCI Bus Master Request signal (req_l) > directly into 21285 Bus Master Grant signal (gnt_l), so that 21285 is the > Arbiter. Ok, here is a "generic" way of finding out what's going wrong... Do a make config (or menuconfig or xconfig depending on your taste). You should find an option called CONFIG_DEBUG_LL (Kernel low-level debugging functions). Turn this on. Also, turn on CONFIG_DEBUG_DC21285_PORT (Kernel low-level debugging messages via DC21285 port). Then, add into kernel/printk.c a "printascii(buf+3) as shown below: va_start(args, fmt); i = vsprintf(buf + 3, fmt, args); /* hopefully i < sizeof(buf)-4 */ buf_end = buf + 3 + i; va_end(args); printascii(buf+3); for (p = buf + 3; p < buf_end; p++) { You can add other printascii(), printhex() etc calls in start_kernel if you so desire to find out where it's going wrong. The above will write messages passed to printk() to the DC21285's serial port before the kernel initialises any console. > Where is this basic 21285 initialization performed in the BIOS ? drivers/pci.c _____ |_____| ------------------------------------------------- ---+---+- | | 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 Fri Nov 12 22:20:25 1999 Received: (from majordomo@localhost) by zen.ics.uwe.ac.uk (8.9.3/8.8.7) id WAA14223 for linux-arm-kernel-outgoing; Fri, 12 Nov 1999 22:20:25 GMT Received: from po6.andrew.cmu.edu (PO6.ANDREW.CMU.EDU [128.2.10.106]) by zen.ics.uwe.ac.uk (8.9.3/8.8.7) with ESMTP id WAA14219 for ; Fri, 12 Nov 1999 22:20:22 GMT Received: (from postman@localhost) by po6.andrew.cmu.edu (8.9.3/8.9.3) id RAA12523 for linux-arm-kernel@lists.arm.linux.org.uk; Fri, 12 Nov 1999 17:20:20 -0500 (EST) Received: via switchmail; Fri, 12 Nov 1999 17:20:19 -0500 (EST) Received: from unix7.andrew.cmu.edu via qmail ID ; Fri, 12 Nov 1999 17:18:46 -0500 (EST) Received: from unix7.andrew.cmu.edu via qmail ID ; Fri, 12 Nov 1999 17:18:42 -0500 (EST) Received: from mms.4.60.Jun.27.1996.03.02.53.sun4.51.EzMail.2.0.CUILIB.3.45.SNAP.NOT.LINKED.unix7.andrew.cmu.edu.sun4m.54 via MS.5.6.unix7.andrew.cmu.edu.sun4_51; Fri, 12 Nov 1999 17:18:42 -0500 (EST) Message-ID: Date: Fri, 12 Nov 1999 17:18:42 -0500 (EST) From: David Sean Friedman To: linux-arm-kernel@lists.arm.linux.org.uk Subject: 2.2.13 problems Cc: Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk I've been trying to get 2.2.13 to boot on an ebsa285 for the past few days. I've been using 2.2.10-rmk2 without any problems, booting from a ramdisk in the flash ROM. It seems that the image is not properly uncompressed, resulting in ext2 errors. After a lot of debugging, I've noticed that the first call to block_write() by gunzip() only writes 2k not all 32k. This appears to be caused by stack corruption. A variable in block_write() is overwritten with the address of nr_free_pages. Since the arm linux patch doesn't touch any of the code in block_write(), I assumed the corruption was caused by an interrupt or an exception. So I turned on the debug messages for exception handling and I put cli/sti around the section where the corruption occurred. The board is being used in add-in mode, with only the patches from Mark van Doesburg's for loading the arm kernel into the ebsa's RAM from the host. The board still boots with an old 2.2.10-rmk2 kernel. I am out of ideas on what else could cause this problem. If anyone has any suggestions, please let me know. Thanks, David Friedman 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 Nov 12 23:19:31 1999 Received: (from majordomo@localhost) by zen.ics.uwe.ac.uk (8.9.3/8.8.7) id XAA16535 for linux-arm-kernel-outgoing; Fri, 12 Nov 1999 23:19:31 GMT Received: from caramon.arm.linux.org.uk (root@p38-cordelia-gui.tch.enablis.net [212.250.233.38]) by zen.ics.uwe.ac.uk (8.9.3/8.8.7) with ESMTP id XAA16531 for ; Fri, 12 Nov 1999 23:19:28 GMT Received: from raistlin.arm.linux.org.uk (linux@raistlin [192.168.0.3]) by caramon.arm.linux.org.uk (8.9.3/8.9.3) with ESMTP id XAA10666; Fri, 12 Nov 1999 23:19:15 GMT From: Russell King - ARM Linux Admin Received: (from linux@localhost) by raistlin.arm.linux.org.uk (8.7.4/8.7.3) id XAA22907; Fri, 12 Nov 1999 23:18:07 GMT Message-Id: <199911122318.XAA22907@raistlin.arm.linux.org.uk> Subject: Re: 2.2.13 problems To: dsf+@andrew.cmu.edu (David Sean Friedman) Date: Fri, 12 Nov 1999 23:18:07 +0000 (GMT) Cc: linux-arm-kernel@lists.arm.linux.org.uk In-Reply-To: from "David Sean Friedman" at Nov 12, 99 05:18:42 pm X-Mailer: ELM [version 2.4 PL24] 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 Sean Friedman writes: > It seems that the image is not properly uncompressed, resulting in ext2 > errors. After a lot of debugging, I've noticed that the first call to > block_write() by gunzip() only writes 2k not all 32k. This appears to be > caused by stack corruption. A variable in block_write() is overwritten > with the address of nr_free_pages. I've just asked Alan about this, and he says he hasn't received any bug reports from the rest of the Linux masses, so I guess that this isn't a generic kernel bug. Can you give details of exactly how you built the kernel (ie, which make commands from unpacking, patching, configuring), and the versions of your ARM toolchain, and what host you are building the kernel on? _____ |_____| ------------------------------------------------- ---+---+- | | 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 Sat Nov 13 00:25:05 1999 Received: (from majordomo@localhost) by zen.ics.uwe.ac.uk (8.9.3/8.8.7) id AAA18680 for linux-arm-kernel-outgoing; Sat, 13 Nov 1999 00:25:05 GMT Received: from po7.andrew.cmu.edu (PO7.ANDREW.CMU.EDU [128.2.10.107]) by zen.ics.uwe.ac.uk (8.9.3/8.8.7) with ESMTP id AAA18656 for ; Sat, 13 Nov 1999 00:25:01 GMT Received: (from postman@localhost) by po7.andrew.cmu.edu (8.9.3/8.9.3) id TAA24541 for linux-arm-kernel@lists.arm.linux.org.uk; Fri, 12 Nov 1999 19:24:58 -0500 (EST) Received: via switchmail; Fri, 12 Nov 1999 19:24:57 -0500 (EST) Received: from unix7.andrew.cmu.edu via qmail ID ; Fri, 12 Nov 1999 19:24:04 -0500 (EST) Received: from unix7.andrew.cmu.edu via qmail ID ; Fri, 12 Nov 1999 19:24:03 -0500 (EST) Received: from mms.4.60.Jun.27.1996.03.02.53.sun4.51.EzMail.2.0.CUILIB.3.45.SNAP.NOT.LINKED.unix7.andrew.cmu.edu.sun4m.54 via MS.5.6.unix7.andrew.cmu.edu.sun4_51; Fri, 12 Nov 1999 19:24:03 -0500 (EST) Message-ID: Date: Fri, 12 Nov 1999 19:24:03 -0500 (EST) From: David Sean Friedman To: linux-arm-kernel@lists.arm.linux.org.uk Subject: Re: 2.2.13 problems Cc: In-Reply-To: <199911122318.XAA22907@raistlin.arm.linux.org.uk> References: <199911122318.XAA22907@raistlin.arm.linux.org.uk> Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk Excerpts from arm-linux: 12-Nov-99 Re: 2.2.13 problems by Russell A. L. Admin@arm. > I've just asked Alan about this, and he says he hasn't received any bug > reports from the rest of the Linux masses, so I guess that this isn't a > generic kernel bug. Can you give details of exactly how you built the > kernel (ie, which make commands from unpacking, patching, configuring), > and the versions of your ARM toolchain, and what host you are building > the kernel on? I got the kernel from kernel.org, and I applied the patch-2.2.13-rmk1 like this: patch -p0 < patch-2.2.13-rmk1 Since I was building on x86, I changed the Makefile so that it would compile for arm, instead of the host arch. Then I applied the fix for HOSTCFLAGS that I posted here a few days ago. Next I applied a modified patch from Mark's co285 patch. This just adds the symbols boot_command_line and boot_memory_end. It looks like this: --- ../new_kernel/linux-2.2.10-orig/arch/arm/kernel/head-armv.S Wed Jul 7 15:30 :09 1999 +++ linux/arch/arm/kernel/head-armv.S Mon Jul 12 18:32:33 1999 @@ -10,6 +10,10 @@ .globl SYMBOL_NAME(swapper_pg_dir) .equ SYMBOL_NAME(swapper_pg_dir), TEXTADDR - 0x4000 + .globl SYMBOL_NAME(boot_command_line) + .equ SYMBOL_NAME(boot_command_line), TEXTADDR - 0x5000 + .globl SYMBOL_NAME(boot_memory_end) + .equ SYMBOL_NAME(boot_memory_end), TEXTADDR - 0x4004 .section ".text.init",#alloc,#execinstr ENTRY(stext) I applied this in the same way as patch-2.2.13-rmk1. Next I patched arch/arm/kernel/setup.c to set up the ramdisk. This includes setting the ROOT_DEV and providing the correct arguments for setup_ram() and setup_initrd(). BTW, check_initrd does not seem to handle ramdisks loaded from ROM correctly, since the ROM is located at an address beyond the end of memory. Then I configured the kernel with only what I need to have it boot: Footbridge, not host mode, sa110 Binfmt-elf Ramdisk, initrd non-standard serial, console on 21285 serial procfs, ext2fs frame pointer, and verbose kernel and user messages Then I compile it with gcc-2.95.2 and binutils 2.9.5.0.16. These were built without any patches. To boot the kernel, I use Mark's co285 modules on the host machine: insmod arm_run.o echo -n 'debug' > /dev/arm0_config cat zImage > /dev/arm0 The kernel will start to boot and then panic after it receives an error in ext2_check_descriptors. Let me know if you need more detail. Thanks, David Friedman 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 Nov 14 13:29:35 1999 Received: (from majordomo@localhost) by zen.ics.uwe.ac.uk (8.9.3/8.8.7) id NAA25073 for linux-arm-kernel-outgoing; Sun, 14 Nov 1999 13:29:35 GMT Received: from kings-cross.london.uk.eu.org (exim@tazenda.demon.co.uk [158.152.220.239]) by zen.ics.uwe.ac.uk (8.9.3/8.8.7) with ESMTP id NAA25069 for ; Sun, 14 Nov 1999 13:29:31 GMT Received: from localhost ([127.0.0.1] helo=kings-cross.london.uk.eu.org ident=phil) by kings-cross.london.uk.eu.org with esmtp (Exim 2.05 #1) id 11mzic-0008ME-00 for linux-arm-kernel@lists.arm.linux.org.uk; Sun, 14 Nov 1999 13:29:22 +0000 X-Mailer: exmh version 2.0.2 2/24/98 (debian) To: linux-arm-kernel@lists.arm.linux.org.uk Subject: bios32.h Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Date: Sun, 14 Nov 1999 13:29:22 +0000 From: Philip Blundell Message-Id: Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk Recent versions of arch/arm/kernel/bios32.c seem to call for a `bios32.h' header of which I can find no trace. I don't know whether my source tree is incomplete or this inclusion is spurious. Should this file exist? 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 Nov 15 00:08:51 1999 Received: (from majordomo@localhost) by zen.ics.uwe.ac.uk (8.9.3/8.8.7) id AAA11831 for linux-arm-kernel-outgoing; Mon, 15 Nov 1999 00:08:51 GMT Received: from xanadu.gn.com (modemcable233.137-200-24.mtl.mc.videotron.net [24.200.137.233]) by zen.ics.uwe.ac.uk (8.9.3/8.8.7) with ESMTP id AAA11821 for ; Mon, 15 Nov 1999 00:08:47 GMT Received: from localhost (nico@localhost) by xanadu.gn.com (8.9.3/8.9.3) with ESMTP id TAA25220 for ; Sun, 14 Nov 1999 19:09:15 -0500 X-Authentication-Warning: xanadu.gn.com: nico owned process doing -bs Date: Sun, 14 Nov 1999 19:09:15 -0500 (EST) From: Nicolas Pitre X-Sender: nico@xanadu.gn.com To: linux-arm-kernel@lists.arm.linux.org.uk Subject: advice required for 2.3.26 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 working on the SA1100 support in 2.3.26 and I'm somewhat comfused by the new memory management framework. First, what is the definition and relation between those macros: PAGE_OFFSET PHYS_OFFSET __pa() __va() virt_to_phys() phys_to_virt() >From what I understood, PAGE_OFFSET is the virtual start address for all the ram the kernel can see, and PHYS_OFFSET is the physical address of the first ram page. On the SA1100, both of these are then equal to 0xC0000000. In arch/arm/kernel/setup.c, there are multiple calls to reserve_bootmem()/free_bootmem() wich an address argument which seems to be a physical address relative to the physical start of the ram. The __pa() macro apparently convert a virtual address into that relative physical address. The problem is that __pa() is defined like this in include/asm-arm/page.h: #define __pa(x) ((unsigned long)(x) - PAGE_OFFSET + PHYS_OFFSET) although the definition from include/asm-i386/page.h is: #define __pa(x) ((unsigned long)(x) - PAGE_OFFSET) Since PHYS_OFFSET on the SA1100 isn't 0 like it is for EBSA285, this lead to inconsistency when the following call is made from setup.c: reserve_bootmem(0, PAGE_SIZE); reserve_bootmem(__pa(swapper_pg_dir), PTRS_PER_PGD * sizeof(void*)); Also, if __pa() is really relative to the start of physical ram, the call: bootmem_end = __pa(PAGE_ALIGN((unsigned long)&_end)); bootmem_end += init_bootmem(bootmem_end >> PAGE_SHIFT, end_pfn, PHYS_OFFSET ); implies that phys_to_virt() is also assuming relative physical addresses since init_bootmem() uses phys_to_virt() on the argument passed as bootmem_end. The i386 architecture defines phys_to_virt(x) as __va(x), wich isn't the case for ARM... Some code in include/asm-arm/proc-armv/pgtable.h using __virt_to_phys() implies that this macro must return an absolute physical address, not a relative one. The problems on the SA1100 occur because the physical ram address is 0xc0000000, not 0 like for the EBSA285 on which relative and absolute physical addresses start at 0. So... Any enlightment on how to sort out all this on the SA1100 would be greatly appreciated. I'm about to try some ideas, but an "official" statement might help. 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 Nov 15 13:13:23 1999 Received: (from majordomo@localhost) by zen.ics.uwe.ac.uk (8.9.3/8.8.7) id NAA13421 for linux-arm-kernel-outgoing; Mon, 15 Nov 1999 13:13:23 GMT Received: from aleph.locsoft.com (root@[212.240.65.10]) by zen.ics.uwe.ac.uk (8.9.3/8.8.7) with ESMTP id NAA13414 for ; Mon, 15 Nov 1999 13:12:50 GMT Received: from uphill ([212.240.65.50]) by aleph.locsoft.com (8.8.8/8.8.7) with SMTP id NAA29103; Mon, 15 Nov 1999 13:35:44 GMT Message-ID: <004a01bf2f6a$8af76090$3241f0d4@uphill.locsoft.com> From: "Paul Robertson" To: "Russell King - ARM Linux Admin" Cc: Subject: Re: booting kernel on EBSA285 Date: Mon, 15 Nov 1999 13:08:41 -0000 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 4.72.3110.5 X-MimeOLE: Produced By Microsoft MimeOLE V4.72.3110.3 Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk Thanks once again, The kernel is failing with: ... RAM disk driver initialized: 16 RAM disks of 4096K size <0>Kernel panic: VFS: Unable to mount root fs on 01:00 ... I'm afraid I need more help with this. Can the kernel load my compressed ramdisk image from EBSA285 Flash. If so, then what kernel parameters to set? Otherwise, should I load the compressed ramdisk image to RAM ? If so, then what RAM address, and what kernel parameters? Paul Robertson LocSoft Limited "The key to your embedded solutions" www.locsoft.com Phone +44 (0) 1793 432557 Fax +44 (0) 1793 432727 -----Original Message----- From: Russell King - ARM Linux Admin To: Paul Robertson Cc: linux-arm-kernel@lists.arm.linux.org.uk Date: Friday, November 12, 1999 9:58 AM Subject: Re: booting kernel on EBSA285 Ok, here is a "generic" way of finding out what's going wrong... Do a make config (or menuconfig or xconfig depending on your taste). You should find an option called CONFIG_DEBUG_LL (Kernel low-level debugging functions). Turn this on. Also, turn on CONFIG_DEBUG_DC21285_PORT (Kernel low-level debugging messages via DC21285 port). Then, add into kernel/printk.c a "printascii(buf+3) as shown below: va_start(args, fmt); i = vsprintf(buf + 3, fmt, args); /* hopefully i < sizeof(buf)-4 */ buf_end = buf + 3 + i; va_end(args); printascii(buf+3); for (p = buf + 3; p < buf_end; p++) { You can add other printascii(), printhex() etc calls in start_kernel if you so desire to find out where it's going wrong. The above will write messages passed to printk() to the DC21285's serial port before the kernel initialises any console. 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 Nov 16 16:09:51 1999 Received: (from majordomo@localhost) by zen.ics.uwe.ac.uk (8.9.3/8.8.7) id QAA23906 for linux-arm-kernel-outgoing; Tue, 16 Nov 1999 16:09:51 GMT Received: from sparkle.Generation.NET (sparkle.Generation.NET [205.205.119.4]) by zen.ics.uwe.ac.uk (8.9.3/8.8.7) with ESMTP id QAA23902 for ; Tue, 16 Nov 1999 16:09:28 GMT Received: from dual (generic253.197.205.205.IN-ADDR.ARPA [205.205.197.253] (may be forged)) by sparkle.Generation.NET (8.9.3/8.9.3) with SMTP id LAA09616 for ; Tue, 16 Nov 1999 11:09:13 -0500 (EST) From: Francois Desloges To: linux-arm-kernel@lists.arm.linux.org.uk Subject: Re: 3c905B NIC + EBSA? Date: Tue, 16 Nov 1999 11:06:14 -0500 Content-Type: text/plain MIME-Version: 1.0 Message-Id: <99111611063102.04589@dual> Content-Transfer-Encoding: 8bit Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk On Tue, 19 Oct 1999, Philip Blundell wrote: > >I'v read somewhere in the july Mailing List Archive that people were having > >problems with 3Com 3c905B NIC on the EBSA285? > >Anybody knows the recent status of success/failure with 3c905B + EBSA285 ? > > I think the cache coherency stuff was missing at the time. It seems to be > present now so I guess it should work, though I haven't tried it myself. > > Are you having trouble? Well it's not kernel trouble, yet! It's BIOS trouble. With either 3c905-TX (device == 9050) or 3c905B (device == 9055) I can of course easily have the BIOS probe the cards right by adding appropriate > if (!pci_config_base) > pci_config_base = pci_lookup_vendor_device(0x10b7, 0x905x); statements. When it tries to do_bootp() however, it reports me the right MAC address of the cards, but nothing gets out of the NIC (we listen at the other end with tcpdump). I guess the 3c59x family and the 3c905 family are different enough for the BIOS not to work? -- François Desloges fd@vipswitch.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 Tue Nov 16 17:53:26 1999 Received: (from majordomo@localhost) by zen.ics.uwe.ac.uk (8.9.3/8.8.7) id RAA27132 for linux-arm-kernel-outgoing; Tue, 16 Nov 1999 17:53:26 GMT Received: from caramon.arm.linux.org.uk (root@p35-cordelia-gui.tch.enablis.net [212.250.233.35]) by zen.ics.uwe.ac.uk (8.9.3/8.8.7) with ESMTP id RAA27126 for ; Tue, 16 Nov 1999 17:53:16 GMT Received: from raistlin.arm.linux.org.uk (linux@raistlin [192.168.0.3]) by caramon.arm.linux.org.uk (8.9.3/8.9.3) with ESMTP id RAA27400; Tue, 16 Nov 1999 17:53:04 GMT From: Russell King - ARM Linux Admin Received: (from linux@localhost) by raistlin.arm.linux.org.uk (8.7.4/8.7.3) id RAA22181; Tue, 16 Nov 1999 17:51:39 GMT Message-Id: <199911161751.RAA22181@raistlin.arm.linux.org.uk> Subject: Re: 3c905B NIC + EBSA? To: fd@vipswitch.com (Francois Desloges) Date: Tue, 16 Nov 1999 17:51:37 +0000 (GMT) Cc: linux-arm-kernel@lists.arm.linux.org.uk In-Reply-To: <99111611063102.04589@dual> from "Francois Desloges" at Nov 16, 99 11:06:14 am X-Mailer: ELM [version 2.4 PL24] 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 Francois Desloges writes: > I guess the 3c59x family and the 3c905 family are different enough for the BIOS > not to work? I guess so. I don't have any info on the 3c905 family, apart from the Linux source. And before someone starts talking about cache coherency, please forget it - the BIOS is not that inteligent, and it hand-loads the packets on to and off of the card. Are you using 10bT? If so, do you get the heartbeat on the link? (check the LEDs on the card and hub). _____ |_____| ------------------------------------------------- ---+---+- | | 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 Wed Nov 17 11:47:03 1999 Received: (from majordomo@localhost) by zen.ics.uwe.ac.uk (8.9.3/8.8.7) id LAA02437 for linux-arm-kernel-outgoing; Wed, 17 Nov 1999 11:47:03 GMT Received: from aleph.locsoft.com (root@[212.240.65.10]) by zen.ics.uwe.ac.uk (8.9.3/8.8.7) with ESMTP id LAA02432 for ; Wed, 17 Nov 1999 11:46:55 GMT Received: from uphill ([212.240.65.50]) by aleph.locsoft.com (8.8.8/8.8.7) with SMTP id MAA03828 for ; Wed, 17 Nov 1999 12:10:17 GMT Message-ID: <008f01bf30f0$e55ee8f0$3241f0d4@uphill.locsoft.com> From: "Paul Robertson" To: Subject: Re: Adress of rd_start/initrd_start on EBSA285 Date: Wed, 17 Nov 1999 11:42:58 -0000 MIME-Version: 1.0 Content-Type: multipart/alternative; boundary="----=_NextPart_000_008C_01BF30F0.E52AE0A0" X-Priority: 3 X-MSMail-Priority: Normal X-Mailer: Microsoft Outlook Express 4.72.3110.5 X-MimeOLE: Produced By Microsoft MimeOLE V4.72.3110.3 Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk This is a multi-part message in MIME format. ------=_NextPart_000_008C_01BF30F0.E52AE0A0 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable Apologies for cross-posting - I guess this follow-up really belongs here Gaixia wrote (in linux-arm list): >Hi, > >I am trying to get a ramdisk for ARM-Linux on EBSA285. I ran into the >following >questions: What is the virtual address I should give to rd_start in the >param page? >and What is the physical address I should load the ramdisk image? > >How about initrd_start? > >I can only find 4200 0000 mapped to fe00 0000 and some others in >hardware.h, >Where could I get the related information for ramdisk? I have the same problem, plus some others... Where did you get your ramdisk image? I can't remember where I got mine, only that I tried ftp://ssdl-redstone.stanford.edu/pub/ramdisk.gz unsuccessfully My ARM-Linux kernel successfully mounts this ramdisk image as a NFS on another machine, but then crashes when it tries to execute /sbin/init Paul Robertson LocSoft Limited "The key to your embedded solutions" www.locsoft.com Phone +44 (0) 1793 432557 Fax +44 (0) 1793 432727 ------=_NextPart_000_008C_01BF30F0.E52AE0A0 Content-Type: text/html; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable
Apologies for cross-posting - = I guess this=20 follow-up really belongs here
 
Gaixia wrote (in linux-arm list):

>Hi,
>
>I = am trying=20 to get a ramdisk for ARM-Linux on EBSA285. I ran into=20 the
>following
>questions: What is the virtual address I = should give=20 to rd_start in the
>param page?
>and What is the physical = address I=20 should load the ramdisk image?
>
>How about=20 initrd_start?
>
>I can only find 4200 0000 mapped to fe00 = 0000 and=20 some others in
>hardware.h,
>Where could I get the related=20 information for ramdisk?


I have the same problem, plus some=20 others...

Where did you get your ramdisk image?

I can't = remember=20 where I got mine, only that I tried
ftp://ssdl-redst= one.stanford.edu/pub/ramdisk.gz
unsuccessfully

My=20 ARM-Linux kernel successfully mounts this ramdisk image as a NFS = on
another=20 machine, but then crashes when it tries to execute = /sbin/init

Paul=20 Robertson
LocSoft Limited
"The key to your embedded=20 solutions"
www.locsoft.com
Phone=20 +44 (0) 1793 432557
Fax     +44 (0) 1793=20 432727
------=_NextPart_000_008C_01BF30F0.E52AE0A0-- 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 Nov 17 15:20:54 1999 Received: (from majordomo@localhost) by zen.ics.uwe.ac.uk (8.9.3/8.8.7) id PAA15281 for linux-arm-kernel-outgoing; Wed, 17 Nov 1999 15:20:54 GMT Received: from caramon.arm.linux.org.uk (root@p48-magpie-gui.tch.enablis.net [194.168.180.48]) by zen.ics.uwe.ac.uk (8.9.3/8.8.7) with ESMTP id PAA15271 for ; Wed, 17 Nov 1999 15:20:47 GMT Received: from raistlin.arm.linux.org.uk (linux@raistlin [192.168.0.3]) by caramon.arm.linux.org.uk (8.9.3/8.9.3) with ESMTP id LAA30476; Wed, 17 Nov 1999 11:00:35 GMT From: Russell King - ARM Linux Admin Received: (from linux@localhost) by raistlin.arm.linux.org.uk (8.7.4/8.7.3) id LAA23471; Wed, 17 Nov 1999 11:00:31 GMT Message-Id: <199911171100.LAA23471@raistlin.arm.linux.org.uk> Subject: Re: advice required for 2.3.26 To: nico@CAM.ORG (Nicolas Pitre) Date: Wed, 17 Nov 1999 11:00:30 +0000 (GMT) Cc: linux-arm-kernel@lists.arm.linux.org.uk In-Reply-To: from "Nicolas Pitre" at Nov 14, 99 07:09:15 pm X-Mailer: ELM [version 2.4 PL24] 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: > First, what is the definition and relation between those macros: > > PAGE_OFFSET The start address of kernel virtual RAM. > PHYS_OFFSET The physical start address of RAM. > __pa() Fast conversion to a physical address from a virtual address > __va() Fast conversion from a virtual address to a physical address > virt_to_phys() (Slower) conversion ala __pa() > phys_to_virt() (Slower) conversion ala __va() > In arch/arm/kernel/setup.c, there are multiple calls to > reserve_bootmem()/free_bootmem() which an address argument which seems to > be a physical address relative to the physical start of the ram. The whole bootmem stuff is a mess on the ARM at the moment. It's probably best to stick with .24 for the moment (except for EBSA285/NetWinder which are the only two I have been testing this stuff on so far). Currently, there are a couple of "more important" problems which need sorting before this one gets cleaned up. To this end, I will not be releasing a 2.3.28 tree (my build trees are too messed up with various "other" patches to make this worth while), hopefully the required changes to the generic code will be in Linus' 2.3.29 to allow an ARM release. _____ |_____| ------------------------------------------------- ---+---+- | | 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 Mon Nov 22 16:10:47 1999 Received: (from majordomo@localhost) by zen.ics.uwe.ac.uk (8.9.3/8.8.7) id QAA10804 for linux-arm-kernel-outgoing; Mon, 22 Nov 1999 16:10:47 GMT Received: from mail4.wlv.netzero.net (mail4.wlv.netzero.net [209.247.163.54]) by zen.ics.uwe.ac.uk (8.9.3/8.8.7) with SMTP id QAA10800 for ; Mon, 22 Nov 1999 16:10:39 GMT Received: (qmail 6562 invoked by uid 0); 22 Nov 1999 00:10:43 -0000 Received: from pppa8-minneapolisb22-3r5050.saturn.bbn.com (HELO netzero.net) (4.12.5.198) by mail4.wlv.netzero.net with SMTP; 22 Nov 1999 00:10:43 -0000 Message-ID: <38388A35.9FC4E76E@netzero.net> Date: Sun, 21 Nov 1999 18:11:33 -0600 From: Gaixia Zhang X-Mailer: Mozilla 4.61 [en] (Win98; I) X-Accept-Language: en MIME-Version: 1.0 To: linux-arm-kernel@lists.arm.linux.org.uk, Paul Robertson Subject: Re: Adress of rd_start/initrd_start on EBSA285 References: <007101bf30f0$28b99b00$3241f0d4@uphill.locsoft.com> Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk Maybe you can try this one: ftp://ftp.netwinder.org/users/n/nico/ But I didn't see big difference. Actually, the one you are using works fine on my EBSA285 as a NFS root, I am thnking about building my own ramdisk based on Netwinder's 2.2 disk image (huge one), but I am not sure I will have time, Best wishes, Gaixia Paul Robertson wrote: > Gaixia wrote: > > >Hi, > > > >I am trying to get a ramdisk for ARM-Linux on EBSA285. I ran into the > >following > >questions: What is the virtual address I should give to rd_start in the > >param page? > >and What is the physical address I should load the ramdisk image? > > > >How about initrd_start? > > > >I can only find 4200 0000 mapped to fe00 0000 and some others in > >hardware.h, > >Where could I get the related information for ramdisk? > > I have the same problem, plus some others... > > Where did you get your ramdisk image? > > I can't remember where I got mine, only that I tried > ftp://ssdl-redstone.stanford.edu/pub/ramdisk.gz > unsuccessfully > > My ARM-Linux kernel successfully mounts this ramdisk image as a NFS on > another machine, but then crashes when it tries to execute /sbin/init > > Paul Robertson > LocSoft Limited > "The key to your embedded solutions" > www.locsoft.com > Phone +44 (0) 1793 432557 > Fax +44 (0) 1793 432727 __________________________________________ NetZero - Defenders of the Free World Get your FREE Internet Access and Email at http://www.netzero.net/download/index.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 Nov 23 16:52:39 1999 Received: (from majordomo@localhost) by zen.ics.uwe.ac.uk (8.9.3/8.8.7) id QAA13590 for linux-arm-kernel-outgoing; Tue, 23 Nov 1999 16:52:39 GMT Received: from aleph.locsoft.com (root@[212.240.65.10]) by zen.ics.uwe.ac.uk (8.9.3/8.8.7) with ESMTP id QAA13582 for ; Tue, 23 Nov 1999 16:52:31 GMT Received: from uphill ([212.240.65.50]) by aleph.locsoft.com (8.8.8/8.8.7) with SMTP id LAA16803 for ; Tue, 23 Nov 1999 11:51:27 GMT Message-ID: <004701bf35a5$2382e4c0$3241f0d4@uphill.locsoft.com> From: "Paul Robertson" To: "LAK list" Subject: HOWTO: load & mount ramdisk on EBSA-285 Date: Tue, 23 Nov 1999 11:23:15 -0000 MIME-Version: 1.0 Content-Type: multipart/alternative; boundary="----=_NextPart_000_0044_01BF35A5.228B3E50" X-Priority: 3 X-MSMail-Priority: Normal X-Mailer: Microsoft Outlook Express 4.72.3110.5 X-MimeOLE: Produced By Microsoft MimeOLE V4.72.3110.3 Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk This is a multi-part message in MIME format. ------=_NextPart_000_0044_01BF35A5.228B3E50 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable Well, this is how I did it... 1. (In boot loader) copy the compressed kernel from FLASH to RAM at = 0x8000 2. Copy the compressed ramdisk image to RAM at 0x800000 3. Set up the following kernel parameters in param_struct: root_dev =3D 0x0100 root_flags =3D FLAG_RDLOAD initrd_start =3D 0xc0800000 /* NB virtual address */ initrd_size =3D 2188926 /* size of compressed image in = bytes */ 4. Jump to 0x8000 This works, but it isn't very clever in the way it enough RAM for the = compressed ramdisk is available at a fixed address and wont be = overwritten by the initialising kernel. I don't know how best to improve this. I'd like to add a routine to = arch/arm/kernel/setup.c to copy the ramdisk image from FLASH to RAM if = initrd_start is a FLASH address, but I don't know if that would be the = right thing to do. Suggestions? Paul Robertson LocSoft Limited "The key to your embedded solutions" www.locsoft.com Phone +44 (0) 1793 432557=20 Fax +44 (0) 1793 432727 ------=_NextPart_000_0044_01BF35A5.228B3E50 Content-Type: text/html; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable
Well, this is how I did = it...
 
1. (In boot loader) copy the compressed kernel from = FLASH to=20 RAM at 0x8000
 
2. Copy the compressed ramdisk image to RAM at=20 0x800000
 
3. Set up the following kernel parameters in=20 param_struct:
    root_dev=20 =3D 0x0100
    root_flags =3D=20 FLAG_RDLOAD
    initrd_start =3D=20 0xc0800000        /* NB virtual = address=20 */
    initrd_size =3D=20 2188926           =   =20 /* size of compressed image in bytes */
 
4. Jump to 0x8000
 
This works, but it isn't very clever in the way it = enough RAM=20 for the compressed ramdisk is available at a fixed address and wont be=20 overwritten by the initialising kernel.
 
I don't know how best to improve this. I'd like to = add a=20 routine to arch/arm/kernel/setup.c to copy the ramdisk image from FLASH = to RAM=20 if initrd_start is a FLASH address, but I don't know if that would be = the right=20 thing to do.
 
Suggestions?
 
Paul Robertson
LocSoft = Limited
"The=20 key to your embedded solutions"
www.locsoft.com
Phone +44 (0) = 1793 432557=20
Fax     +44 (0) 1793=20 432727
------=_NextPart_000_0044_01BF35A5.228B3E50-- 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 Nov 23 16:52:33 1999 Received: (from majordomo@localhost) by zen.ics.uwe.ac.uk (8.9.3/8.8.7) id QAA13586 for linux-arm-kernel-outgoing; Tue, 23 Nov 1999 16:52:33 GMT Received: from aleph.locsoft.com (root@[212.240.65.10]) by zen.ics.uwe.ac.uk (8.9.3/8.8.7) with ESMTP id QAA13579 for ; Tue, 23 Nov 1999 16:52:30 GMT Received: from uphill ([212.240.65.50]) by aleph.locsoft.com (8.8.8/8.8.7) with SMTP id MAA17148 for ; Tue, 23 Nov 1999 12:59:14 GMT Message-ID: <005501bf35ae$9ac274c0$3241f0d4@uphill.locsoft.com> From: "Paul Robertson" To: "LAK list" Subject: insmod wont Date: Tue, 23 Nov 1999 12:31:02 -0000 MIME-Version: 1.0 Content-Type: multipart/alternative; boundary="----=_NextPart_000_0052_01BF35AE.9A8E6C70" X-Priority: 3 X-MSMail-Priority: Normal X-Mailer: Microsoft Outlook Express 4.72.3110.5 X-MimeOLE: Produced By Microsoft MimeOLE V4.72.3110.3 Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk This is a multi-part message in MIME format. ------=_NextPart_000_0052_01BF35AE.9A8E6C70 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable I can't get insmod to insert a module into my ARM linux kernel. The kernel is 2.2.13 built with CONFIG_MODULES and CONFIG_MODVERSIONS = defined. The module is a do-nothing "hello world" module which implements = init_module and cleanup_module calling printk. (The printk version = matches the kernel). Insmod is version 2.0.0 Insmod reports "/lib/modules/misc/hello.o not an object file" file /lib/modules/misc/hello.o reports ELF 32-bit LSB relocatable, ARM, = version 1, not stripped Can anyone help? Thanks, Paul Robertson LocSoft Limited "The key to your embedded solutions" www.locsoft.com Phone +44 (0) 1793 432557=20 Fax +44 (0) 1793 432727 ------=_NextPart_000_0052_01BF35AE.9A8E6C70 Content-Type: text/html; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable
I can't get insmod to insert a = module into my=20 ARM linux kernel.
 
The kernel is 2.2.13 built with CONFIG_MODULES and=20 CONFIG_MODVERSIONS defined.
The module is a do-nothing "hello world" = module=20 which implements init_module and cleanup_module calling=20 printk. (The printk version matches the kernel).
Insmod is version 2.0.0
 
Insmod reports "/lib/modules/misc/hello.o not = an object=20 file"
file=20 /lib/modules/misc/hello.o reports ELF 32-bit LSB relocatable, ARM, = version 1,=20 not stripped
 
Can anyone help?
 
Thanks,
 
Paul Robertson
LocSoft = Limited
"The=20 key to your embedded solutions"
www.locsoft.com
Phone +44 (0) = 1793 432557=20
Fax     +44 (0) 1793=20 432727
------=_NextPart_000_0052_01BF35AE.9A8E6C70-- 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 Nov 23 16:59:14 1999 Received: (from majordomo@localhost) by zen.ics.uwe.ac.uk (8.9.3/8.8.7) id QAA13766 for linux-arm-kernel-outgoing; Tue, 23 Nov 1999 16:59:14 GMT Received: from oskar.dev.nanoteq.co.za (oskar.dev.nanoteq.co.za [196.7.114.5]) by zen.ics.uwe.ac.uk (8.9.3/8.8.7) with ESMTP id QAA13761 for ; Tue, 23 Nov 1999 16:59:08 GMT Received: (from chris@localhost) by oskar.dev.nanoteq.co.za (8.9.3/8.9.0) id UAA13296; Mon, 22 Nov 1999 20:02:24 +0200 (SAT) Date: Mon, 22 Nov 1999 20:02:21 +0200 (SAT) From: Chris Bohme To: linux-arm-kernel@lists.arm.linux.org.uk Subject: StrongARM cache coherency and EtherExpress NICs 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 all, I am trying to get Intel EtherExpress (82558/9) cards to work with the EBSA285. From the driver code it does not seem as if cache coherency is addressed. Has anyone implemented these cache fixups for the eepro100 driver ? And another question, if the cache is to be off, where should this be done ? I tried turning it off in kernel/head-armv.S, but that did not seem to have any effect (is it re-enabled later on ?) 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 Nov 24 11:57:10 1999 Received: (from majordomo@localhost) by zen.ics.uwe.ac.uk (8.9.3/8.8.7) id LAA02023 for linux-arm-kernel-outgoing; Wed, 24 Nov 1999 11:57:10 GMT Received: from caramon.arm.linux.org.uk (root@p23-robin-gui.tch.enablis.net [194.168.180.83]) by zen.ics.uwe.ac.uk (8.9.3/8.8.7) with ESMTP id LAA02019 for ; Wed, 24 Nov 1999 11:57:07 GMT Received: from raistlin.arm.linux.org.uk (linux@raistlin [192.168.0.3]) by caramon.arm.linux.org.uk (8.9.3/8.9.3) with ESMTP id KAA08396; Wed, 24 Nov 1999 10:00:11 GMT From: Russell King - ARM Linux Admin Received: (from linux@localhost) by raistlin.arm.linux.org.uk (8.7.4/8.7.3) id JAA16556; Wed, 24 Nov 1999 09:59:18 GMT Message-Id: <199911240959.JAA16556@raistlin.arm.linux.org.uk> Subject: Re: HOWTO: load & mount ramdisk on EBSA-285 To: paul.robertson@locsoft.com (Paul Robertson) Date: Wed, 24 Nov 1999 09:59:17 +0000 (GMT) Cc: linux-arm-kernel@lists.arm.linux.org.uk In-Reply-To: <004701bf35a5$2382e4c0$3241f0d4@uphill.locsoft.com> from "Paul Robertson" at Nov 23, 99 11:23:15 am X-Mailer: ELM [version 2.4 PL24] 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 Paul Robertson writes: > This works, but it isn't very clever in the way it enough RAM for the > compressed ramdisk is available at a fixed address and wont be > overwritten by the initialising kernel. If the initrd image is overwritten, then the kernel will flag it in the boot messages. Unfortunately, with 2.2 and older kernels this is the way it is on all architectures. The problem has been fixed in the 2.3 kernels with the advent of a "bootmem" allocator, which is told regions that it must not allocate from, including the initrd image. _____ |_____| ------------------------------------------------- ---+---+- | | 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 Wed Nov 24 11:57:37 1999 Received: (from majordomo@localhost) by zen.ics.uwe.ac.uk (8.9.3/8.8.7) id LAA02046 for linux-arm-kernel-outgoing; Wed, 24 Nov 1999 11:57:37 GMT Received: from caramon.arm.linux.org.uk (root@p23-robin-gui.tch.enablis.net [194.168.180.83]) by zen.ics.uwe.ac.uk (8.9.3/8.8.7) with ESMTP id LAA02033 for ; Wed, 24 Nov 1999 11:57:24 GMT Received: from raistlin.arm.linux.org.uk (linux@raistlin [192.168.0.3]) by caramon.arm.linux.org.uk (8.9.3/8.9.3) with ESMTP id KAA08404; Wed, 24 Nov 1999 10:02:55 GMT From: Russell King - ARM Linux Admin Received: (from linux@localhost) by raistlin.arm.linux.org.uk (8.7.4/8.7.3) id KAA16619; Wed, 24 Nov 1999 10:02:01 GMT Message-Id: <199911241002.KAA16619@raistlin.arm.linux.org.uk> Subject: Re: insmod wont To: paul.robertson@locsoft.com (Paul Robertson) Date: Wed, 24 Nov 1999 10:02:01 +0000 (GMT) Cc: linux-arm-kernel@lists.arm.linux.org.uk In-Reply-To: <005501bf35ae$9ac274c0$3241f0d4@uphill.locsoft.com> from "Paul Robertson" at Nov 23, 99 12:31:02 pm X-Mailer: ELM [version 2.4 PL24] 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 Paul Robertson writes: > The kernel is 2.2.13 built with CONFIG_MODULES and CONFIG_MODVERSIONS = > defined. > The module is a do-nothing "hello world" module which implements = > init_module and cleanup_module calling printk. (The printk version = > matches the kernel). > Insmod is version 2.0.0 > > Insmod reports "/lib/modules/misc/hello.o not an object file" > file /lib/modules/misc/hello.o reports ELF 32-bit LSB relocatable, ARM, = > version 1, not stripped Have you updated your module utilities to a version which supports ELF? The original insmod utilities do not understand the ELF binary format. _____ |_____| ------------------------------------------------- ---+---+- | | 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 Wed Nov 24 20:47:57 1999 Received: (from majordomo@localhost) by zen.ics.uwe.ac.uk (8.9.3/8.8.7) id UAA00665 for linux-arm-kernel-outgoing; Wed, 24 Nov 1999 20:47:57 GMT Received: from po8.andrew.cmu.edu (PO8.ANDREW.CMU.EDU [128.2.10.108]) by zen.ics.uwe.ac.uk (8.9.3/8.8.7) with ESMTP id UAA00661 for ; Wed, 24 Nov 1999 20:47:46 GMT Received: (from postman@localhost) by po8.andrew.cmu.edu (8.9.3/8.9.3) id MAA17907; Wed, 24 Nov 1999 12:33:54 -0500 (EST) Received: via switchmail; Wed, 24 Nov 1999 12:33:54 -0500 (EST) Received: from unix6.andrew.cmu.edu via qmail ID ; Wed, 24 Nov 1999 12:32:08 -0500 (EST) Received: from unix6.andrew.cmu.edu via qmail ID ; Wed, 24 Nov 1999 12:32:08 -0500 (EST) Received: from mms.4.60.Jun.27.1996.03.02.53.sun4.51.EzMail.2.0.CUILIB.3.45.SNAP.NOT.LINKED.unix6.andrew.cmu.edu.sun4m.54 via MS.5.6.unix6.andrew.cmu.edu.sun4_51; Wed, 24 Nov 1999 12:32:07 -0500 (EST) Message-ID: Date: Wed, 24 Nov 1999 12:32:07 -0500 (EST) From: David Sean Friedman To: linux-arm-kernel@lists.arm.linux.org.uk, Paul Robertson , Gaixia Zhang Subject: Re: Adress of rd_start/initrd_start on EBSA285 Cc: In-Reply-To: <38388A35.9FC4E76E@netzero.net> References: <007101bf30f0$28b99b00$3241f0d4@uphill.locsoft.com> <38388A35.9FC4E76E@netzero.net> Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk Excerpts from arm-linux: 21-Nov-99 Re: Adress of rd_start/init.. by Gaixia Zhang@netzero.net > I am thnking about building my own ramdisk based on Netwinder's > 2.2 disk image (huge one), but I am not sure I will have time, I built a ramdisk from the netwinder distro a few months ago. I planned to use the full disk image, but I found a rescue image on the netwinder site which had almost everything I needed. Try http://www.netwinder.org/allmisc.html, and look for files named rescue (I don't remember which one I used). David Friedman Carnegie Mellon 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 Nov 24 20:52:55 1999 Received: (from majordomo@localhost) by zen.ics.uwe.ac.uk (8.9.3/8.8.7) id UAA00832 for linux-arm-kernel-outgoing; Wed, 24 Nov 1999 20:52:55 GMT Received: from po6.andrew.cmu.edu (PO6.ANDREW.CMU.EDU [128.2.10.106]) by zen.ics.uwe.ac.uk (8.9.3/8.8.7) with ESMTP id UAA00828 for ; Wed, 24 Nov 1999 20:52:53 GMT Received: (from postman@localhost) by po6.andrew.cmu.edu (8.9.3/8.9.3) id MAA03428; Wed, 24 Nov 1999 12:23:01 -0500 (EST) Received: via switchmail; Wed, 24 Nov 1999 12:23:01 -0500 (EST) Received: from unix6.andrew.cmu.edu via qmail ID ; Wed, 24 Nov 1999 12:21:23 -0500 (EST) Received: from unix6.andrew.cmu.edu via qmail ID ; Wed, 24 Nov 1999 12:21:22 -0500 (EST) Received: from mms.4.60.Jun.27.1996.03.02.53.sun4.51.EzMail.2.0.CUILIB.3.45.SNAP.NOT.LINKED.unix6.andrew.cmu.edu.sun4m.54 via MS.5.6.unix6.andrew.cmu.edu.sun4_51; Wed, 24 Nov 1999 12:21:22 -0500 (EST) Message-ID: Date: Wed, 24 Nov 1999 12:21:22 -0500 (EST) From: David Sean Friedman To: linux-arm-kernel@lists.arm.linux.org.uk, Chris Bohme Subject: Re: StrongARM cache coherency and EtherExpress NICs Cc: In-Reply-To: References: Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk Excerpts from arm-linux: 22-Nov-99 StrongARM cache coherency a.. by Chris Bohme@oskar.dev.na > I am trying to get Intel EtherExpress (82558/9) cards to work with the > EBSA285. From the driver code it does not seem as if cache coherency is > addressed. > > Has anyone implemented these cache fixups for the eepro100 driver ? My group has a patch for this, but the driver has a few problems, which probably are not ARM related. Check the mailing list archives for linux-eepro100 at tux.org for a full discussion. Supposedly, the latest driver fixes these problems for most people. > > And another question, if the cache is to be off, where should this be > done ? I tried turning it off in kernel/head-armv.S, but that did not > seem to have any effect (is it re-enabled later on ?) This is the way we used to disable the cache. I don't have a tree with Russell's patch in front of me, but we used to change a line that looked like this in the label .Larmv4_flush_early: orr r0, r0, #0x003d @ I...S..DPWC.M so that it would disable the data cache and write buffers (I think those are the C and W bits). If you are interested in our patch, let me know. We used the eepro100 with an ebsa285 and intel 82557/8. David Friedman Carnegie Mellon 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 Nov 24 21:57:56 1999 Received: (from majordomo@localhost) by zen.ics.uwe.ac.uk (8.9.3/8.8.7) id VAA02058 for linux-arm-kernel-outgoing; Wed, 24 Nov 1999 21:57:56 GMT Received: from mail.quack.com (IDENT:root@mail.quack.com [216.50.211.108]) by zen.ics.uwe.ac.uk (8.9.3/8.8.7) with ESMTP id VAA02054 for ; Wed, 24 Nov 1999 21:57:52 GMT Received: from quack.com (IDENT:jon@quack.com [216.50.211.108]) by mail.quack.com (8.9.3/8.9.3) with ESMTP id NAA15523; Wed, 24 Nov 1999 13:57:50 -0800 Message-ID: <383C5F5D.38A897F1@quack.com> Date: Wed, 24 Nov 1999 13:57:49 -0800 From: Jonathan Kliegman Organization: Quackware X-Mailer: Mozilla 4.61 [en] (X11; U; Linux 2.2.12 i686) X-Accept-Language: en MIME-Version: 1.0 CC: linux-arm-kernel@lists.arm.linux.org.uk, Chris Bohme Subject: Re: StrongARM cache coherency and EtherExpress NICs References: Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk David Sean Friedman wrote: > > > And another question, if the cache is to be off, where should this be > > done ? I tried turning it off in kernel/head-armv.S, but that did not > > seem to have any effect (is it re-enabled later on ?) > This is the way we used to disable the cache. I don't have a tree with > Russell's patch in front of me, but we used to change a line that looked > like this in the label .Larmv4_flush_early: > orr r0, r0, #0x003d @ I...S..DPWC.M > > so that it would disable the data cache and write buffers (I think those > are the C and W bits). > > If you are interested in our patch, let me know. We used the eepro100 > with an ebsa285 and intel 82557/8. > > David Friedman I remember that there was a second problem with disabling the cache. If 'CONFIG_KERNEL_ALIGNMENT_TRAP' (or something similar, I don't have a tree in front of me) is turned on then there is a variable (whose name escapes me) which stores the user mode state. On context switches the MMU is updated based in part on this variable. If you search for the tree for the MCR and MRC instructions you should be able to find it. Sorry I'm not being more helpful - I haven't had time to look/play with a kernel tree in a while... -Jon 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 Nov 24 22:06:45 1999 Received: (from majordomo@localhost) by zen.ics.uwe.ac.uk (8.9.3/8.8.7) id WAA02380 for linux-arm-kernel-outgoing; Wed, 24 Nov 1999 22:06:45 GMT Received: from xanadu.vipswitch.com (generic199.197.205.205.IN-ADDR.ARPA [205.205.197.199]) by zen.ics.uwe.ac.uk (8.9.3/8.8.7) with ESMTP id WAA02374 for ; Wed, 24 Nov 1999 22:06:42 GMT Date: Wed, 24 Nov 1999 15:01:59 -0500 (EST) From: Nicolas Pitre To: Russell King - ARM Linux Admin cc: linux-arm-kernel@lists.arm.linux.org.uk Subject: Re: Next EBSA285 BIOS release In-Reply-To: <199911241152.LAA16969@raistlin.arm.linux.org.uk> Message-ID: MIME-Version: 1.0 Content-Type: MULTIPART/MIXED; BOUNDARY="-842151995-1655952973-943473719=:4172" Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk This message is in MIME format. The first part should be readable text, while the remaining parts are likely unreadable without MIME-aware tools. Send mail to mime@docserver.cac.washington.edu for more info. ---842151995-1655952973-943473719=:4172 Content-Type: TEXT/PLAIN; charset=US-ASCII On Wed, 24 Nov 1999, Russell King - ARM Linux Admin wrote: > Hi, > > I'm going to be releasing BIOS 1.06 soon. If anyone has any patches > against 1.05, could they email them to me by the weekend? Thanks. Here is one, attached. It contains code to boot a kernel by uploading it with ymodem over the serial line (pretty handy when everything else fail). Also fixes for the linker script from Francois Desloges. Nicolas ---842151995-1655952973-943473719=:4172 Content-Type: APPLICATION/octet-stream; name="diff-bios-1.05.gz" Content-Transfer-Encoding: BASE64 Content-ID: Content-Description: Content-Disposition: attachment; filename="diff-bios-1.05.gz" H4sICB0/PDgCA2RpZmYtYmlvcy0xLjA1AJ0aa3faRvaz+BW3TjcW5mFJgB8Q Z4NtnLqxTQ6QttluDkdIA9ZGSKweDm7r/773zoweyAJnS2yjuXPfr9HMxHbm c2jEwR3oTa1zOHP88PDi3vQWLIT8oNJoNJ6jKJOYwa35CEYH9E63pXVbR6Cf np5WarXaBr3yK7Phzn8Aow16u9s6RWSB+e4dNPR6C2p6XW/Du3cVEmPA6PYD qMiWuBNetQIVgAO4ctbIaRX4M5ctwfMjsFnErMjxFrBkSz94hND5g4HlBwGC 3cdKjf4lNkU+t6JLMMDPVWB6lu+EcMlC1yeMN3P73YOzCr85kXXftPzl20pN acCcy3Ud7ysLILQCZxVxeIAyH3AmYHMWMM8SMqJ7Bv3rK7hnpo3488BfghVE WnOcCL5zLN81Q/joRAGDNx4O31nmsukHCyHQtG1k+9hY+jYaGq9c37Rh5qPB Sxbd+zYyskuCd2t+ZXPHZbAxKoQvASvj2IOfYxegTfEzMCp6IX4p6mYA9aOu fpwF0NDqx1DjfymC/ZGinP2oXoyG4/H0Ynj78fpmUDWDCowno+uPJXNhhA7F CDcGN1eK0j2DH9Xwnrkuft9cVqHxC/wFi4CtYJ+585ZhBst9eAuHNns49GJE e/0amHXvw2O1UnuRBVIjl60MMM+cOfsvqD+qyKpaJ9DF1U3//VhRamfQWJor K2y0DHyyVvEZqu57C+Sp6xo0LpGkNC524DywIDy07s2gEKTSqULESnGUyX0M YzSJYnLU7Zx2OwbF5CQLXzldIZatjVjq9Q4VY4dHsgI300l/9H4wIZcSl6ZZ adxMh+c/j7mXHxZm06/UihAIHhFMrvQsN7YZRmEy/Hh5PaoejmKXhc0l6vKy o5CLBeXgXQ6ieW7kJbOAp3RX0/AHH45OtzgnpdnqGK2ukWMMnSd57fAAK/kA RsxiyAZ4PLFp3MPn2+Hl4FYAIuwvIXYG6liRjyVPNJzuzo9YF2Byj83HwhqH OMTecYKaRehnZn0N4yV8u3ese1hiD6RON6O2tlyZkTNLZHFW8hP6S0ZyFoG5 RJ6u6SypK2I7CuPVyg+iRLNEqbCZakNGHur6IRq+szV1CVu59iKGUiL0luMh /8H5uG+cdPZDOL8ejmHuB9gCbuHG8eJ1KuNQdOJXSUa84TEII9vxm/dvSyYC 1L5sJnKWrAxOzXEbHAudT3ENbKwCj8F4MDrvjwegrduGpmn6kZbNDT9AbjQY jYYjUBt6NYNd9W+INgNMRp8GoGfji/4dqPu/7b/WWsc5uvHwpzzWePIbGDlJ wwm0s2H/4gMcZcO7PqpldHIyr28Hw08TVM3IiRgNJqPPt/3fsKdnwF8viDfa oFWFG8IIE8kCF7sXnD9GuP6xedRL4RhYOHe/uszLYFQoEDJrFs9/1zWj/SWb ir3QWXi0RBK/g5UfE6+cnAffsSFaq5yxBajEn7QSYoZjLqsHqvqASRfRgBAO qmoSnhoGSD+pVuE1nFR7RLMdGbGwSyHSU04yoQRrlb4odVCzRDhXhbNM1ScQ W68czHzkxLzIQXtDVEKSUq1oWk8s41x5NYf25iwhrv4p1nlnDir8sKHyhqxy Q/WjKkgGgM4622lyL0EMWBQHnrQI4Im++B85IbPlmXtcyg8en1UcoR2g8kAf rOpclpe4ywNc/1o98fj27RkYvZwXPOzHVSGZhx+7LNe7kBnVVbVWK4kQNi5L vi5J8Z65ZInogC2cMMI5qZkMQJq3KEpbH8/Fh2u1QhBxQLfq+IvtxMNcBpVg ddANmUsUn4OV1DvPzox8V10JJOm/4YcSvb9ZC5Z3WrDG6nimNs802dXrSPOf Oq4PQYivmL0t5onsRMaRFQdBHVgQ+EHIJ6i/qmJMdvfkHLxJKz8B1WrSsjSZ yF5VikZirIsOFc0ZNSLEVRBRAVn4OE9V3uOgBa5fgKZSzJWnPLuU2xk1tzIm xgkREXGXzyRWCQWMqpCQrnpkE4dwZVPkGqgcW+iLXZXaqiKtU4Q+3DUrzhfD wD2NA9nHoNGg8VviT6RcT6XMJygEvalVxbw0fua4C9YToFRXrAjMH57rkgO3 hjykbPN4nvsm5xyFmopopJyxM3Czuf2cXGZm4qOEx1M+RswNWTFQuBYQvSTn a0NvFz4uZjl8vh5KfK57N6GV7ZwbCT+cJS0n61G4AuAqVs23qALT5wUWrFde Wl4rr7y4ZEfANyIWTaOYa5QXJjVTqZtmNUvsuKJaWaFYZDp3TpbVyBNXZZmy uezlgeSiNwqkxGNPUjXJ5qX2EqxBLTWZgu5RO8maRMg8m/xUB2uWrN1EJ1Fl ZUGKhwB0UNZUer2iG1DPBDkLYq5+M1eK94IMiRyYIZ7JSiYtsCOnqZw5Vs7y xYEqbJa0j6wrv016yj+Th+7GqwuSJSuY0Cb1QzXXUDJ+jbN0PinI4jIgO1HO YRi2Xml5bRib2Joz8xmPHfQ7ky6XLi+owdNuS+EKkvK5p9JElDudNBu3pAwp kRStTIoNXWYBM78W04ReKbUvaWPL25jh5d8Ripzhr7+SYtkq7XvaTcgCx3Sn dNyiipei/NtiwEJZUivcnkSo096d/42nR24H+MgPbZrN5r+9PRAVQW9EGPhn b0MkZ2radiA7QchrSrpalbRclAp7k2Q3+Y+wjlAWhmB6j/CVPZKkOiS+oo/K eZ2hC6tYLnu0cXRZxPawZPbmJqpqJ6phAVvqRicSVm7xjB0vl48F10g6LaMK YisCufPi31NBjub9mam5J4B79QyUl/LdYPLiFmTUqLflrIy5c/puurY87MwB CscLuRl+6NlfBWAcg3bc1U+6ulE4NMtjv3DWYpzSmUJL46cthwfg2vJ0kfbt dEjCd9LnyI2OPw8rDVxNP36aTK+Go9v+RN3j52ENM1hi+LcNMEqlVK4TRS57 iRYkbX908ZOKIIQM7vA1U53itieg07Hx4GJyPbwbVzC0DYBpxNaU682eoiho 0oSGpkdnpKbd8D33EWwzMkWoyCSAJifpitRIB99HzqOeyaykFXCgck7VDcjc WcerTdDCi5vfzMBzvMXmROCToGqWV01q3TfX7+/UdrLOTVnB2gFq6s+BQ6kV O76XFSWqy9tQ4+8QcipuemJnk49ST12mbin5pJ7a4JDaKixtJBBUUSiZYm8o yaFlSkqf/P9UvDVz3abhH0g6vv7XYHgl9epJ42dhOMWGFCRuQ9bn4zHsspmS C8nS3JLPyncRC2uKUjf8hpObbiPz0VxEhTRF8uOnCv6UdSN5aHW4cfoF26cK HaoUp9ir9K6hFXpVOd3OrmUc1dtQw79H8owYU5Zh++eLhNhnq5bvhRGk5wh0 JzON+BePZk2S8FWQb7XzBMiE6qsCr9B1znyLu5zoUNypQBHwzDXpjLzx8ID8 oHXb7a5x8swhGfamGwxs9Z1c89br6JWazps4OSI907sc9W+nd6P+aNT/PFba xRkOH15dKdpaw08bf8klCdKn/mhyOcplorbmZ5MZxnn/0+V01J8MppfXv1yP h6NpS9M4qnGktXbhGW2B2MqfIj5HOz0SaCfaLiz91OBo7Z1YrRMhUz/dhdVB iRxrp/q63uEij3YiGSecV2enyxLnblGYn12CwlcQSvF32MRwtSPQwvVnriJW vwqI726locwcSwkwG8DF/derFqIug1AJNNyErMIAhxwBh/T7CoM6R5gfBHmY zcnCQCESgiJf1w6UEGvod8781ckX0ufwgF+HimtKft84Gt4mN6BLcyUy9PSo rut0MXdSx5cUylFl6T8oQhgNrEBZ6Z06SA0serZQikEdV+DqQllK0hrXhSBn 48+358Ob6V3/dqBOxeIBYtYozjKxCkviVhkx9nyk17sCRSNbjS8oF0tHwWYg QboANZQwnoXcJfSLkJqAEG/65WSzRaToM3JV3mTRvgx+W3nckdeVysxjHDcz WE8E6ySYaA3tS94jlM3Z6FkSPSPXj77k+WvPEYwv2/vc0nS85CYsByjrc2Jm 42b3uKsdbelzEnt3u29jo6/hX+Et2bnLXvE9hrvwXQiOzXYjhFbo9NLVYfs2 gtYQkHuTAtYBP7bqbZ2WD+HvX/hehCxs0VJWax1LC185c+wacDG8u7p+Pz0f DifTu8EEI/Y6sbKeLk01CRRqIVy5+3Rzg2s8aSiyTdyNY9LJ/9+Ae3bKStrQ OrRD41c3OKKcU5amF+PWhXjyPdkruUnnR5UObgJz0pp8kypOA4EfByopHR0M SBVB4TyQuRlHfsoaStPNdWaF6+k8pJBw+an8yoqvGh2t2ypuizbQX9gXtWlb JPyVv4FGFoULaCt6XLGmn57I4qN8i/HhIRRbZ3xMH5am6/pW4c56FxN6O3mJ 19+56CZvcN4WFMYlXhYT33GnXUB+6TK71ZJ32eImeezQKYF4IatyyOzxhf+q Ii+Q+T02/sQhpi9u0eZ+7NENF/rRpqDd12EWRwh2bSbh8g2XTpE8/1tz1x1x ehVM/+idsVIjHeHZW2PxnJS/X67Tc08+9Ngif38HKlFSEe3DvtzsIYCfQ8r7 oWS+sV8860JeyLyhZ0ejKaXc04j6zXD1XlHy69dcd7qX2Nf20+EbHJ4+E7iG A7qV6WVjfgPBxVahQRxKDruQSpr9VPkf3OzHLfMlAAA= ---842151995-1655952973-943473719=:4172-- 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 Nov 24 22:07:25 1999 Received: (from majordomo@localhost) by zen.ics.uwe.ac.uk (8.9.3/8.8.7) id WAA02397 for linux-arm-kernel-outgoing; Wed, 24 Nov 1999 22:07:25 GMT Received: from caramon.arm.linux.org.uk (root@p51-magpie-gui.tch.enablis.net [194.168.180.51]) by zen.ics.uwe.ac.uk (8.9.3/8.8.7) with ESMTP id WAA02393 for ; Wed, 24 Nov 1999 22:07:21 GMT Received: from raistlin.arm.linux.org.uk (linux@raistlin [192.168.0.3]) by caramon.arm.linux.org.uk (8.9.3/8.9.3) with ESMTP id WAA11131; Wed, 24 Nov 1999 22:06:18 GMT From: Russell King - ARM Linux Admin Received: (from linux@localhost) by raistlin.arm.linux.org.uk (8.7.4/8.7.3) id WAA19454; Wed, 24 Nov 1999 22:05:21 GMT Message-Id: <199911242205.WAA19454@raistlin.arm.linux.org.uk> Subject: Re: StrongARM cache coherency and EtherExpress NICs To: jon@quack.com (Jonathan Kliegman) Date: Wed, 24 Nov 1999 22:05:20 +0000 (GMT) Cc: linux-arm-kernel@lists.arm.linux.org.uk, chris@oskar.dev.nanoteq.co.za In-Reply-To: <383C5F5D.38A897F1@quack.com> from "Jonathan Kliegman" at Nov 24, 99 01:57:49 pm X-Mailer: ELM [version 2.4 PL24] 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 Jonathan Kliegman writes: > I remember that there was a second problem with disabling the cache. If > 'CONFIG_KERNEL_ALIGNMENT_TRAP' (or something similar, I don't have a > tree in front of me) is turned on then there is a variable (whose name > escapes me) which stores the user mode state. On context switches the > MMU is updated based in part on this variable. If you search for the > tree for the MCR and MRC instructions you should be able to find it. It's called cr_alignment, which is initialised from the mov instruction in arch/arm/mm/proc-*.S (via head-armv.S). Therefore, disabling the cache in arch/arm/mm/proc-*.S will disable the cache while Linux is running in total. (It's done this way since you can't guarantee that you can read- modify-write the control register, and the control register value is processor specific). _____ |_____| ------------------------------------------------- ---+---+- | | 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 Thu Nov 25 16:23:29 1999 Received: (from majordomo@localhost) by zen.ics.uwe.ac.uk (8.9.3/8.8.7) id QAA05732 for linux-arm-kernel-outgoing; Thu, 25 Nov 1999 16:23:29 GMT Received: from xanadu.vipswitch.com (generic199.197.205.205.IN-ADDR.ARPA [205.205.197.199] (may be forged)) by zen.ics.uwe.ac.uk (8.9.3/8.8.7) with ESMTP id QAA05727 for ; Thu, 25 Nov 1999 16:23:15 GMT Date: Thu, 25 Nov 1999 11:20:33 -0500 (EST) From: Nicolas Pitre To: linux-arm-kernel@lists.arm.linux.org.uk Subject: turning dcache off Message-ID: MIME-Version: 1.0 Content-Type: MULTIPART/MIXED; BOUNDARY="-842151995-2023707852-943546833=:1352" Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk This message is in MIME format. The first part should be readable text, while the remaining parts are likely unreadable without MIME-aware tools. Send mail to mime@docserver.cac.washington.edu for more info. ---842151995-2023707852-943546833=:1352 Content-Type: TEXT/PLAIN; charset=US-ASCII For those testing cache coherency problems, here is a little module that turns dcache off when inserted and turn it back on when removed. This is handy for real time testing. See the source in attachment. To compile: gcc -O2 -c dcache_off.c To disable data cache: insmod ./dcache_off.o To reenable data cache: rmmod dcache_off Nicolas P.S.: I'm experiencing cache coherency problems with 3C905 network cards at the moment (either with a 3c905B Cyclone and a 3c905 Boomerang). This is with 2.2.12. Didn't try 2.3.26 yet. Anybody else played with those cards? ---842151995-2023707852-943546833=:1352 Content-Type: TEXT/PLAIN; charset=US-ASCII; name="dcache_off.c" Content-Transfer-Encoding: BASE64 Content-ID: Content-Description: Content-Disposition: attachment; filename="dcache_off.c" I2RlZmluZSBfX0tFUk5FTF9fDQojZGVmaW5lIE1PRFVMRQ0KDQojaW5jbHVk ZSA8bGludXgvbW9kdWxlLmg+DQojaW5jbHVkZSA8bGludXgva2VybmVsLmg+ DQojaW5jbHVkZSA8YXNtL3N5c3RlbS5oPg0KI2luY2x1ZGUgPGxpbnV4L3Zt YWxsb2MuaD4NCg0KaW50IGluaXRfbW9kdWxlKHZvaWQpDQp7DQoJdW5zaWdu ZWQgaW50IGZsYWdzLCB2YWwxLCB2YWwyOw0KDQoJc2F2ZV9mbGFnc19jbGko IGZsYWdzICk7DQoJZmx1c2hfY2FjaGVfYWxsKCk7DQoJYXNtKCAiDQoJbXJj CXAxNSwgMCwgJTAsIGMxLCBjMCwgMAlAIGN0cmwgcmVnaXN0ZXINCgliaWMJ JTEsICUwLCAjMHgwQwkJQCBjbGVhciBXIGFuZCBDDQoJbWNyCXAxNSwgMCwg JTEsIGMxLCBjMCwgMA0KCSIgOiAiPXIiICh2YWwxKSwgIj1yIiAodmFsMikg KTsNCglyZXN0b3JlX2ZsYWdzKCBmbGFncyApOw0KCXByaW50ayggIkRhdGEg Q2FjaGUgdHVybmVkIG9mZiAoY3RybCByZWc6IDB4JTA4WCAtLT4gMHglMDhY KVxuIiwgDQoJCXZhbDEsIHZhbDIgKTsNCglyZXR1cm4gMDsNCn0JCQ0KDQp2 b2lkIGNsZWFudXBfbW9kdWxlKCB2b2lkICkNCnsNCgl1bnNpZ25lZCBpbnQg ZmxhZ3MsIHZhbDEsIHZhbDI7DQoNCglzYXZlX2ZsYWdzX2NsaSggZmxhZ3Mg KTsNCglmbHVzaF9jYWNoZV9hbGwoKTsNCglhc20oICINCgltcmMJcDE1LCAw LCAlMCwgYzEsIGMwLCAwCUAgY3RybCByZWdpc3Rlcg0KCW9ycgklMSwgJTAs ICMweDBDCQlAIHNldCBXIGFuZCBDDQoJbWNyCXAxNSwgMCwgJTEsIGMxLCBj MCwgMA0KCSIgOiAiPXIiICh2YWwxKSwgIj1yIiAodmFsMikgKTsNCglyZXN0 b3JlX2ZsYWdzKCBmbGFncyApOw0KCXByaW50ayggIkRhdGEgQ2FjaGUgdHVy bmVkIG9uIChjdHJsIHJlZzogMHglMDhYKSAtLT4gMHglMDhYXG4iLCANCgkJ dmFsMSwgdmFsMiApOw0KfQ0KDQo= ---842151995-2023707852-943546833=:1352-- 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 Nov 25 17:18:05 1999 Received: (from majordomo@localhost) by zen.ics.uwe.ac.uk (8.9.3/8.8.7) id RAA07805 for linux-arm-kernel-outgoing; Thu, 25 Nov 1999 17:18:05 GMT Received: from andrew.cmu.edu (ANDREW.CMU.EDU [128.2.10.101]) by zen.ics.uwe.ac.uk (8.9.3/8.8.7) with ESMTP id RAA07797 for ; Thu, 25 Nov 1999 17:17:53 GMT Received: (from postman@localhost) by andrew.cmu.edu (8.9.3/8.9.3) id MAA18329; Thu, 25 Nov 1999 12:17:50 -0500 (EST) Received: via switchmail; Thu, 25 Nov 1999 12:17:50 -0500 (EST) Received: from unix4.andrew.cmu.edu via qmail ID ; Thu, 25 Nov 1999 12:16:41 -0500 (EST) Received: from unix4.andrew.cmu.edu via qmail ID ; Thu, 25 Nov 1999 12:16:40 -0500 (EST) Received: from mms.4.60.Jun.27.1996.03.02.53.sun4.51.EzMail.2.0.CUILIB.3.45.SNAP.NOT.LINKED.unix4.andrew.cmu.edu.sun4m.54 via MS.5.6.unix4.andrew.cmu.edu.sun4_51; Thu, 25 Nov 1999 12:16:40 -0500 (EST) Message-ID: Date: Thu, 25 Nov 1999 12:16:40 -0500 (EST) From: David Sean Friedman To: linux-arm-kernel@lists.arm.linux.org.uk, Nicolas Pitre Subject: Re: turning dcache off Cc: In-Reply-To: References: Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk Excerpts from arm-linux: 25-Nov-99 turning dcache off by Nicolas Pitre@CAM.ORG > P.S.: I'm experiencing cache coherency problems with 3C905 network cards > at the moment (either with a 3c905B Cyclone and a 3c905 Boomerang). This > is with 2.2.12. Didn't try 2.3.26 yet. Anybody else played with those > cards? Check the mailing list archives. I think there was a thread on this a few weeks ago (or maybe on linux-arm). David Friedman Carnegie Mellon 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 Nov 26 11:33:45 1999 Received: (from majordomo@localhost) by zen.ics.uwe.ac.uk (8.9.3/8.8.7) id LAA03874 for linux-arm-kernel-outgoing; Fri, 26 Nov 1999 11:33:45 GMT Received: from funai-tky.co.jp (ns.funai-tky.co.jp [210.154.37.2]) by zen.ics.uwe.ac.uk (8.9.3/8.8.7) with ESMTP id LAA03741 for ; Fri, 26 Nov 1999 11:33:36 GMT Received: from hiropu ([192.168.1.19]) by funai-tky.co.jp (8.9.1/3.7Wpr7-98092916) with SMTP id LAA16324 for ; Fri, 26 Nov 1999 11:45:43 +0900 Date: Fri, 26 Nov 1999 11:45:42 +0900 From: Hiroshi Ishii To: linux-arm-kernel@lists.arm.linux.org.uk Subject: MMU trable Message-Id: <383DF45633E.EE80HIROSHI@ns.funai-tky.co.jp> MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit X-Mailer: Becky! ver 1.25.07 Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk Hello. I'm trying to port to SA-1110, and I have trouble. In arch/arm/kernel/head-armv.S mcr p15, 0, r0, c1, c0 @ Enable MMU & caches. In 3 instructions When enable MMU, system halt. So, can't 'M' bit enable. What trouble in it? silicon bugs? Can MMU on later(not on this stage)? Regards; ---- ++ Funai Electric R & D Co.,LTD ++ ++ Hiroshi Ishii ++ 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 Nov 26 11:33:37 1999 Received: (from majordomo@localhost) by zen.ics.uwe.ac.uk (8.9.3/8.8.7) id LAA03762 for linux-arm-kernel-outgoing; Fri, 26 Nov 1999 11:33:37 GMT Received: from funai-tky.co.jp (ns.funai-tky.co.jp [210.154.37.2]) by zen.ics.uwe.ac.uk (8.9.3/8.8.7) with ESMTP id LAA03633 for ; Fri, 26 Nov 1999 11:33:25 GMT Received: from hiropu ([192.168.1.19]) by funai-tky.co.jp (8.9.1/3.7Wpr7-98092916) with SMTP id LAA16220 for ; Fri, 26 Nov 1999 11:33:47 +0900 Date: Fri, 26 Nov 1999 11:33:46 +0900 From: Hiroshi Ishii To: linux-arm-kernel@zen.ics.uwe.ac.uk Subject: MMU trable Message-Id: <383DF18A64.EE7FHIROSHI@ns.funai-tky.co.jp> MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit X-Mailer: Becky! ver 1.25.07 Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk Hello. I'm trying to port to SA-1110, and I have trouble. In arch/arm/kernel/head-armv.S mcr p15, 0, r0, c1, c0 @ Enable MMU & caches. In 3 instructions When enable MMU, system halt. So, can't 'M' bit enable. What trouble in it? silicon bugs? Can MMU on later(not on this stage)? Regards; ---- ++ Funai Electric R & D Co.,LTD ++ ++ Hiroshi Ishii ++ 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 Nov 26 15:26:23 1999 Received: (from majordomo@localhost) by zen.ics.uwe.ac.uk (8.9.3/8.8.7) id PAA01103 for linux-arm-kernel-outgoing; Fri, 26 Nov 1999 15:26:23 GMT Received: from xanadu.vipswitch.com (generic199.197.205.205.IN-ADDR.ARPA [205.205.197.199]) by zen.ics.uwe.ac.uk (8.9.3/8.8.7) with ESMTP id PAA01099 for ; Fri, 26 Nov 1999 15:26:19 GMT Date: Fri, 26 Nov 1999 10:22:43 -0500 (EST) From: Nicolas Pitre To: Hiroshi Ishii cc: linux-arm-kernel@lists.arm.linux.org.uk Subject: Re: MMU trable In-Reply-To: <383DF45633E.EE80HIROSHI@ns.funai-tky.co.jp> 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, 26 Nov 1999, Hiroshi Ishii wrote: > Hello. > > I'm trying to port to SA-1110, and I have trouble. > > In arch/arm/kernel/head-armv.S > > mcr p15, 0, r0, c1, c0 @ Enable MMU & caches. In 3 instructions > > When enable MMU, system halt. So, can't 'M' bit enable. > What trouble in it? silicon bugs? Did you use SA1100 configuration? If so, are you sure the physical ram is correctly located? 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 Sat Nov 27 03:29:50 1999 Received: (from majordomo@localhost) by zen.ics.uwe.ac.uk (8.9.3/8.8.7) id DAA08508 for linux-arm-kernel-outgoing; Sat, 27 Nov 1999 03:29:50 GMT Received: from funai-tky.co.jp (ns.funai-tky.co.jp [210.154.37.2]) by zen.ics.uwe.ac.uk (8.9.3/8.8.7) with ESMTP id DAA08504 for ; Sat, 27 Nov 1999 03:29:46 GMT Received: from pekepon.com ([192.168.1.52]) by funai-tky.co.jp (8.9.1/3.7Wpr7-98092916) with ESMTP id MAA28149; Sat, 27 Nov 1999 12:29:35 +0900 Message-ID: <383F5033.563D6D5C@pekepon.com> Date: Sat, 27 Nov 1999 12:29:55 +0900 From: Hiroshi Ishii X-Mailer: Mozilla 4.04 [ja] (Win95; I) MIME-Version: 1.0 To: Nicolas Pitre CC: "linux-arm-kernel@lists.arm.linux.org.uk" Subject: Re: MMU trable References: Content-Type: text/plain; charset=iso-2022-jp Content-Transfer-Encoding: 7bit Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk Nicolas Pitre wrote: > > > > mcr p15, 0, r0, c1, c0 @ Enable MMU & caches. In 3 instructions > > > > When enable MMU, system halt. So, can't 'M' bit enable. > > What trouble in it? silicon bugs? > > Did you use SA1100 configuration? If so, are you sure the physical ram is > correctly located? > I use SA1100 configuration, and I probably located physical ram correctly. Because angel and Intel's initialize application can enable MMU. Intel's initialize application is "SA-1110 Initialization Code Example " (see http://developer.intel.com/design/strong/swsup/). I use angelboot. Is there other problem? Thanks; 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 Nov 27 09:15:20 1999 Received: (from majordomo@localhost) by zen.ics.uwe.ac.uk (8.9.3/8.8.7) id JAA28067 for linux-arm-kernel-outgoing; Sat, 27 Nov 1999 09:15:20 GMT Received: from caramon.arm.linux.org.uk (root@p27-magpie-gui.tch.enablis.net [194.168.180.27]) by zen.ics.uwe.ac.uk (8.9.3/8.8.7) with ESMTP id JAA28063 for ; Sat, 27 Nov 1999 09:15:15 GMT 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 IAA02385; Sat, 27 Nov 1999 08:44:07 GMT Received: (from linux@localhost) by flint.arm.linux.org.uk (8.9.3/8.9.3) id IAA04961; Sat, 27 Nov 1999 08:42:43 GMT From: Russell King - ARM Linux Admin Message-Id: <199911270842.IAA04961@flint.arm.linux.org.uk> Subject: Re: MMU trable To: hiroshi@pekepon.com (Hiroshi Ishii) Date: Sat, 27 Nov 1999 08:42:43 +0000 (GMT) Cc: nico@CAM.ORG, linux-arm-kernel@lists.arm.linux.org.uk In-Reply-To: <383F5033.563D6D5C@pekepon.com> from "Hiroshi Ishii" at Nov 27, 99 12:29:55 pm X-Mailer: ELM [version 2.4 PL24] 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 Hiroshi Ishii writes: > Nicolas Pitre wrote: > > > mcr p15, 0, r0, c1, c0 @ Enable MMU & caches. In 3 instructions > > > > > > When enable MMU, system halt. So, can't 'M' bit enable. > > > What trouble in it? silicon bugs? > > > > Did you use SA1100 configuration? If so, are you sure the physical ram is > > correctly located? > > I use SA1100 configuration, and I probably located physical ram correctly. > Because angel and Intel's initialize application can enable MMU. Intel's initialize > application > is "SA-1110 Initialization Code Example " (see > http://developer.intel.com/design/strong/swsup/). > I use angelboot. > > Is there other problem? Did you not get my email about this problem? _____ |_____| ------------------------------------------------- ---+---+- | | 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 Sat Nov 27 15:50:23 1999 Received: (from majordomo@localhost) by zen.ics.uwe.ac.uk (8.9.3/8.8.7) id PAA06893 for linux-arm-kernel-outgoing; Sat, 27 Nov 1999 15:50:23 GMT Received: from funai-tky.co.jp (ns.funai-tky.co.jp [210.154.37.2]) by zen.ics.uwe.ac.uk (8.9.3/8.8.7) with ESMTP id PAA06755 for ; Sat, 27 Nov 1999 15:49:24 GMT Received: from funai-tky.co.jp ([192.168.1.52]) by funai-tky.co.jp (8.9.1/3.7Wpr7-98092916) with ESMTP id AAA30649; Sun, 28 Nov 1999 00:47:55 +0900 Message-ID: <383FFD30.6464F573@funai-tky.co.jp> Date: Sun, 28 Nov 1999 00:48:01 +0900 From: hiroshi X-Mailer: Mozilla 4.08 [ja_JP.EUC] (X11; I; FreeBSD 3.3-RELEASE i386) MIME-Version: 1.0 To: Russell King - ARM Linux Admin CC: "linux-arm-kernel@lists.arm.linux.org.uk" Subject: Re: MMU trable References: <199911270842.IAA04961@flint.arm.linux.org.uk> Content-Type: text/plain; charset=iso-2022-jp Content-Transfer-Encoding: 7bit Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk Russell King - ARM Linux Admin wrote: > > > I use SA1100 configuration, and I probably located physical ram correctly. > > Because angel and Intel's initialize application can enable MMU. Intel's initialize > > application > > is "SA-1110 Initialization Code Example " (see > > http://developer.intel.com/design/strong/swsup/). > > I use angelboot. > > > > Is there other problem? > > Did you not get my email about this problem? > No. I didn't get it. -- ---- ++ Funai Electric R & D Co.,LTD ++ ++ Hiroshi Ishii ++ 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 Nov 27 20:14:48 1999 Received: (from majordomo@localhost) by zen.ics.uwe.ac.uk (8.9.3/8.8.7) id UAA16585 for linux-arm-kernel-outgoing; Sat, 27 Nov 1999 20:14:48 GMT Received: from caramon.arm.linux.org.uk (root@p39-magpie-gui.tch.enablis.net [194.168.180.39]) by zen.ics.uwe.ac.uk (8.9.3/8.8.7) with ESMTP id UAA16581 for ; Sat, 27 Nov 1999 20:14:37 GMT Received: from raistlin.arm.linux.org.uk (root@raistlin [192.168.0.3]) by caramon.arm.linux.org.uk (8.9.3/8.9.3) with ESMTP id UAA04482; Sat, 27 Nov 1999 20:14:30 GMT From: Russell King - ARM Linux Admin Received: (from linux@localhost) by raistlin.arm.linux.org.uk (8.7.4/8.7.3) id UAA00615; Fri, 26 Nov 1999 20:20:07 GMT Message-Id: <199911262020.UAA00615@raistlin.arm.linux.org.uk> Subject: Re: MMU trable To: hiroshi@funai-tky.co.jp (Hiroshi Ishii) Date: Fri, 26 Nov 1999 20:20:06 +0000 (GMT) Cc: linux-arm-kernel@lists.arm.linux.org.uk In-Reply-To: <383DF45633E.EE80HIROSHI@ns.funai-tky.co.jp> from "Hiroshi Ishii" at Nov 26, 99 11:45:42 am X-Mailer: ELM [version 2.4 PL24] 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 Hiroshi Ishii writes: > In arch/arm/kernel/head-armv.S > > mcr p15, 0, r0, c1, c0 @ Enable MMU & caches. In 3 instructions > > When enable MMU, system halt. So, can't 'M' bit enable. > What trouble in it? silicon bugs? I think you're probably trying to use Angel to step through the code. This will not work since as soon as the MMU is enabled, Angel will disappear from the memory map. _____ |_____| ------------------------------------------------- ---+---+- | | 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 Mon Nov 29 01:51:32 1999 Received: (from majordomo@localhost) by zen.ics.uwe.ac.uk (8.9.3/8.8.7) id BAA08655 for linux-arm-kernel-outgoing; Mon, 29 Nov 1999 01:51:32 GMT Received: from xanadu.gn.com (modemcable233.137-200-24.mtl.mc.videotron.net [24.200.137.233]) by zen.ics.uwe.ac.uk (8.9.3/8.8.7) with ESMTP id BAA08651 for ; Mon, 29 Nov 1999 01:51:29 GMT Received: from localhost (nico@localhost) by xanadu.gn.com (8.9.3/8.9.3) with ESMTP id UAA10697; Sun, 28 Nov 1999 20:45:47 -0500 X-Authentication-Warning: xanadu.gn.com: nico owned process doing -bs Date: Sun, 28 Nov 1999 20:45:47 -0500 (EST) From: Nicolas Pitre X-Sender: nico@xanadu.gn.com To: Hiroshi Ishii cc: "linux-arm-kernel@lists.arm.linux.org.uk" Subject: Re: MMU trable In-Reply-To: <383F5033.563D6D5C@pekepon.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 Sat, 27 Nov 1999, Hiroshi Ishii wrote: > Nicolas Pitre wrote: > > > > > > > mcr p15, 0, r0, c1, c0 @ Enable MMU & caches. In 3 instructions > > > > > > When enable MMU, system halt. So, can't 'M' bit enable. > > > What trouble in it? silicon bugs? > > > > Did you use SA1100 configuration? If so, are you sure the physical ram is > > correctly located? > > I use SA1100 configuration, and I probably located physical ram correctly. What I mean is that the current SA1100 kernel won't work for the SA1110 if the physical ram isn't located at 0xc0000000 like it is on the SA1100. Some changes are required if this is different. 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 Nov 29 07:06:48 1999 Received: (from majordomo@localhost) by zen.ics.uwe.ac.uk (8.9.3/8.8.7) id HAA17662 for linux-arm-kernel-outgoing; Mon, 29 Nov 1999 07:06:48 GMT Received: from funai-tky.co.jp (ns.funai-tky.co.jp [210.154.37.2]) by zen.ics.uwe.ac.uk (8.9.3/8.8.7) with ESMTP id HAA17658 for ; Mon, 29 Nov 1999 07:06:43 GMT Received: from hiropu ([192.168.1.19]) by funai-tky.co.jp (8.9.1/3.7Wpr7-98092916) with SMTP id QAA05019; Mon, 29 Nov 1999 16:06:34 +0900 Date: Mon, 29 Nov 1999 16:06:34 +0900 From: Hiroshi Ishii To: Nicolas Pitre Cc: linux-arm-kernel Subject: Re: MMU trable In-Reply-To: References: <383F5033.563D6D5C@pekepon.com> Message-Id: <384225FA33E.E9F9HIROSHI@ns.funai-tky.co.jp> MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit X-Mailer: Becky! ver 1.25.07 Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk On Sun, 28 Nov 1999 20:45:47 -0500 (EST) Nicolas Pitre wrote: > What I mean is that the current SA1100 kernel won't work for the SA1110 if > the physical ram isn't located at 0xc0000000 like it is on the SA1100. > Some changes are required if this is different. > Compare memory map sa1100 and sa1110 (refer to reference manual), physical ram seems to be located same address (0xc0000000). Am I mistake? ---- ++ Funai Electric R & D Co.,LTD ++ ++ Hiroshi Ishii ++ 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 Nov 29 08:26:46 1999 Received: (from majordomo@localhost) by zen.ics.uwe.ac.uk (8.9.3/8.8.7) id IAA20591 for linux-arm-kernel-outgoing; Mon, 29 Nov 1999 08:26:46 GMT Received: from cwcsun41.cwc.nus.edu.sg (cwcsun41.cwc.nus.edu.sg [137.132.163.102]) by zen.ics.uwe.ac.uk (8.9.3/8.8.7) with ESMTP id IAA20587 for ; Mon, 29 Nov 1999 08:26:16 GMT Received: from konggf.nus.edu.sg ([137.132.163.221]) by cwcsun41.cwc.nus.edu.sg (8.9.3/8.9.3) with SMTP id QAA08403 for ; Mon, 29 Nov 1999 16:25:19 +0800 (SGT) Message-Id: <3.0.32.19991129162230.00a038ac@postman.cwc.nus.edu.sg> X-Sender: konggf@postman.cwc.nus.edu.sg X-Mailer: Windows Eudora Pro Version 3.0 (32) Date: Mon, 29 Nov 1999 16:22:31 +0800 To: linux-arm-kernel@lists.arm.linux.org.uk From: Kong Gang Feng Subject: system call from Kernel timer 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'd like to implement a function button to sync buffer cache. I hook this button handler to the kernel timer. Once this button is pressed more than an interval, say, 2 secs. The button handler will call 'sys_sync()'. But this give me coredump everytime. I am aware that this may be the context problem, when 'sys_sync()' is called from BH of kernel timer. Is there any simple way to work around it? Thanks Kong G.F. 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 Nov 30 05:01:21 1999 Received: (from majordomo@localhost) by zen.ics.uwe.ac.uk (8.9.3/8.8.7) id FAA05104 for linux-arm-kernel-outgoing; Tue, 30 Nov 1999 05:01:21 GMT Received: from funai-tky.co.jp (ns.funai-tky.co.jp [210.154.37.2]) by zen.ics.uwe.ac.uk (8.9.3/8.8.7) with ESMTP id FAA05096 for ; Tue, 30 Nov 1999 05:01:15 GMT Received: from hiropu ([192.168.1.19]) by funai-tky.co.jp (8.9.1/3.7Wpr7-98092916) with SMTP id OAA11276 for ; Tue, 30 Nov 1999 14:01:06 +0900 Date: Tue, 30 Nov 1999 14:01:05 +0900 From: Hiroshi Ishii To: linux-arm-kernel Subject: Re: MMU trable -> solved!!! In-Reply-To: <384225FA33E.E9F9HIROSHI@ns.funai-tky.co.jp> References: <384225FA33E.E9F9HIROSHI@ns.funai-tky.co.jp> Message-Id: <38435A1124E.1956HIROSHI@ns.funai-tky.co.jp> MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit X-Mailer: Becky! ver 1.25.07 Sender: owner-linux-arm-kernel@lists.arm.linux.org.uk Precedence: bulk Solved my problem!!! I try linux-2.3.24-rmk1-np1. Working good on SA1110!!! Thanks everyone. ---- ++ Funai Electric R & D Co.,LTD ++ ++ Hiroshi Ishii ++ unsubscribe: body of `unsubscribe linux-arm-kernel' to majordomo@lists.arm.linux.org.uk