SB1250 locking up in init on current 2.6.16 kernel

SB1250 locking up in init on current 2.6.16 kernel

[Larry Stefani]

Hi,

I've been trying to upgrade from 2.6.16.18 to
2.6.16.60, but am seeing a hard lockup right before
"INIT: version 2.78 booting" on my SB1250-based board.

I found a related discussion on the Debian mailing
list:

http://groups.google.com/group/linux.debian.bugs.dist/browse_thread/thread/b7159ee25106c7f9

However, after applying Thiemo's patch to mark pages
tainted by PIO IDE as dirty, the lockup still occurs.

I narrowed the file changes to

     arch/mips/mm/c-sb1.c
     arch/mips/mm/cache.c
     arch/mips/mm/init.c
     include/asm-mips/cache-flush.h
     include/asm-mips/page.h

between 2.6.16.27 and 2.6.16.29.  There was no
2.6.16.28 tarball posted on linux-mips.org, so I
basically brought .27 to .29 until I found the
offending files.

Is anyone running a 2.6.16 kernel (after 2.6.16.27) on
a SB1250-based board?

Thanks,
Larry Stefani
lstefani@yahoo.com


      ____________________________________________________________________________________
Be a better friend, newshound, and 
know-it-all with Yahoo! Mobile.  Try it now.  http://mobile.yahoo.com/;_ylt=Ahu06i62sR8HDtDypao8Wcj9tAcJ

Re: SB1250 locking up in init on current 2.6.16 kernel

[Ralf Baechle]

On Mon, Mar 24, 2008 at 07:00:15AM -0700, Larry Stefani wrote:

> I've been trying to upgrade from 2.6.16.18 to
> 2.6.16.60, but am seeing a hard lockup right before
> "INIT: version 2.78 booting" on my SB1250-based board.
> 
> I found a related discussion on the Debian mailing
> list:
> 
> http://groups.google.com/group/linux.debian.bugs.dist/browse_thread/thread/b7159ee25106c7f9
> 
> However, after applying Thiemo's patch to mark pages
> tainted by PIO IDE as dirty, the lockup still occurs.

It's a bug which should be fixed but nevertheless I can highly recommend
something like a SiliconImage SATA controller - the onboard PIO PATA
controller is so slow.

> I narrowed the file changes to
> 
>      arch/mips/mm/c-sb1.c
>      arch/mips/mm/cache.c
>      arch/mips/mm/init.c
>      include/asm-mips/cache-flush.h
>      include/asm-mips/page.h
> 
> between 2.6.16.27 and 2.6.16.29.  There was no
> 2.6.16.28 tarball posted on linux-mips.org, so I
> basically brought .27 to .29 until I found the
> offending files.

I've pushed the tag again so now there is a tarball.

If you need to track something like this you're probably best with
git bisect which should bring you right to the offending commit.

> Is anyone running a 2.6.16 kernel (after 2.6.16.27) on
> a SB1250-based board?

Later kernels do run on bcm1480 which is close enough.

  Ralf

Re: SB1250 locking up in init on current 2.6.16 kernel

[Larry Stefani]

Hello Ralf,

Thank you very much for the quick response.  I know
you must be terribly busy with the server move, so I
won't take up too much more of your time.

I did find an old (possibly related) discussion you
were involved in:

http://www.linux-mips.org/archives/linux-mips/2006-09/msg00194.html

I can't tell from the thread whether it was a problem
seen on 2.6.16.29, but that might have been tip at the
time.

> It's a bug which should be fixed but nevertheless I
> can highly recommend
> something like a SiliconImage SATA controller - the
> onboard PIO PATA
> controller is so slow.

I understand, but changing that is not an option for
me today.

> I've pushed the tag again so now there is a tarball.

Thanks.  I thought something was terribly wrong with
.28 for it to be skipped.  

> If you need to track something like this you're
> probably best with
> git bisect which should bring you right to the
> offending commit.

I probably should have used that approach instead of
diffing .27 and .29 and narrowing the file list by
hand.  As it was, there were changes to non-MIPS
platforms and devices I'm not using so those were easy
to apply to .27.  Also, there were many file changes
for MT SMP support, and *most* (but not all) of those
changes were wrapped with conditional compiles, so
those were also easy to apply.  I knew once I got to
these five files I was in some interesting code that
could point to the problems I'm seeing.

> Later kernels do run on bcm1480 which is close
> enough.

By "later kernels", do you mean 2.6.16.60 or different
kernel branches?

Perhaps, but I'm seeing identical failures in .29 and
.60, although that could be misleading.  I do see
significant SMP-related changes in c-sb1.c between .29
and .60, and I am running in SMP.  I wish there was an
easy way to know whether it's in the same code.

Anyway, in the interest of time I may revert to .27
which appears to work.  I don't need any of the MT SMP
related changes that followed, and perhaps I can live
without the enhancements between .27 and .60 for now.

Thanks,
Larry Stefani
lstefani@yahoo.com



      ____________________________________________________________________________________
Never miss a thing.  Make Yahoo your home page. 
http://www.yahoo.com/r/hs

Re: SB1250 locking up in init on current 2.6.16 kernel

[Thiemo Seufer]

Larry Stefani wrote:
> Hi,
> 
> I've been trying to upgrade from 2.6.16.18 to
> 2.6.16.60, but am seeing a hard lockup right before
> "INIT: version 2.78 booting" on my SB1250-based board.
> 
> I found a related discussion on the Debian mailing
> list:
> 
> http://groups.google.com/group/linux.debian.bugs.dist/browse_thread/thread/b7159ee25106c7f9
> 
> However, after applying Thiemo's patch to mark pages
> tainted by PIO IDE as dirty, the lockup still occurs.
> 
> I narrowed the file changes to
> 
>      arch/mips/mm/c-sb1.c
>      arch/mips/mm/cache.c
>      arch/mips/mm/init.c
>      include/asm-mips/cache-flush.h
>      include/asm-mips/page.h
> 
> between 2.6.16.27 and 2.6.16.29.  There was no
> 2.6.16.28 tarball posted on linux-mips.org, so I
> basically brought .27 to .29 until I found the
> offending files.

Please learn about git bisect (git-bisect start; git-bisect good;
git-bisect bad) and use it to isolate the offending commit.


Thiemo

Re: SB1250 locking up in init on current 2.6.16 kernel

[Thiemo Seufer]

Larry Stefani wrote:
> Hi,
> 
> I've been trying to upgrade from 2.6.16.18 to
> 2.6.16.60, but am seeing a hard lockup right before
> "INIT: version 2.78 booting" on my SB1250-based board.
> 
> I found a related discussion on the Debian mailing
> list:
> 
> http://groups.google.com/group/linux.debian.bugs.dist/browse_thread/thread/b7159ee25106c7f9
> 
> However, after applying Thiemo's patch to mark pages
> tainted by PIO IDE as dirty, the lockup still occurs.
> 
> I narrowed the file changes to
> 
>      arch/mips/mm/c-sb1.c
>      arch/mips/mm/cache.c
>      arch/mips/mm/init.c
>      include/asm-mips/cache-flush.h
>      include/asm-mips/page.h
> 
> between 2.6.16.27 and 2.6.16.29.  There was no
> 2.6.16.28 tarball posted on linux-mips.org, so I
> basically brought .27 to .29 until I found the
> offending files.

Please learn about git bisect (git-bisect start; git-bisect good;
git-bisect bad) and use it to isolate the offending commit.


Thiemo
From root@linux-mips.net Wed Mar 26 17:09:36 2008
Received: with ECARTIS (v1.0.0; list linux-mips); Wed, 26 Mar 2008 17:09:39 +0100 (CET)
Received: (root@lappi.linux-mips.net) by lappi.linux-mips.net
	id S1100670AbYCZMiY (ORCPT <rfc822;linux-mips@linux-mips.org>);
	Wed, 26 Mar 2008 13:38:24 +0100
Resent-From: root@lappi.linux-mips.net
Resent-Date: Wed, 26 Mar 2008 13:38:23 +0100
Resent-Message-ID: <20080326123823.GA2581@lappi.linux-mips.net>
Resent-To: linux-mips@linux-mips.org
Received: from oss.sgi.com ([192.48.170.157]:64171 "EHLO oss.sgi.com")
	by lappi.linux-mips.net with ESMTP id S1096320AbYCZLhr (ORCPT
	<rfc822;linux-mips@linux-mips.org>); Wed, 26 Mar 2008 12:37:47 +0100
Received: from p549F5321.dip.t-dialin.net (localhost [127.0.0.1])
	by oss.sgi.com (8.12.11.20060308/8.12.11/SuSE Linux 0.7) with ESMTP id m2QBVUZi016060
	for <linux-mips@linux-mips.org>; Wed, 26 Mar 2008 04:37:07 -0700
Received: from web38815.mail.mud.yahoo.com ([209.191.125.106]:52659 "HELO
	web38815.mail.mud.yahoo.com") by lappi.linux-mips.net with SMTP
	id S1101128AbYCYNFi (ORCPT <rfc822;linux-mips@linux-mips.org>);
	Tue, 25 Mar 2008 14:05:38 +0100
Received: (qmail 44851 invoked by uid 60001); 25 Mar 2008 13:05:35 -0000
DomainKey-Signature: a=rsa-sha1; q=dns; c=nofws;
  s=s1024; d=yahoo.com;
  h=X-YMail-OSG:Received:Date:From:Subject:To:Cc:In-Reply-To:MIME-Version:Content-Type:Content-Transfer-Encoding:Message-ID;
  b=xCBNhWpKinYwWrVvE1Uft9KGjTtZxt3g4fLPn/rjRmDA6K+sN1agoijgRQYs+RIgIIhruDSUXidESuHIwJJ8q05/w71YSGb0eQBTMPTuA0gXFkoooie/8tjkvSWhoLRoN5h3DU151KFKXteS111lV/m/mKh9Gz3fOpXMhQFNPIQ=;
X-YMail-OSG: n7Waq3QVM1l.9KDtAO5iZnkVuqvXgNAognoUnsO_4moP6ISOuUdkSnqDuQ3FVPWwTqPhmQrTxytPdaJySzD.AJtfVyUvj5fRjj_mMOBdPFB.ETj0Xmw-
Received: from [68.236.82.170] by web38815.mail.mud.yahoo.com via HTTP; Tue, 25 Mar 2008 06:05:35 PDT
Date:	Tue, 25 Mar 2008 06:05:35 -0700 (PDT)
From:	Larry Stefani <lstefani@yahoo.com>
Subject: Re: SB1250 locking up in init on current 2.6.16 kernel
To:	Ralf Baechle <ralf@linux-mips.org>
Cc:	linux-mips@linux-mips.org
In-Reply-To: <20080324203311.GB15294@linux-mips.org>
MIME-Version: 1.0
Content-Type:	text/plain; charset=US-ASCII
Content-Transfer-Encoding: 7BIT
Message-ID: <64859.43533.qm@web38815.mail.mud.yahoo.com>
X-Virus-Scanned: ClamAV 0.91.2/6021/Wed Feb 27 15:55:48 2008 on oss.sgi.com
X-Virus-Status:	Clean
Return-Path: <root@linux-mips.net>
X-Envelope-To: <"|/home/ecartis/ecartis -s linux-mips"> (uid 0)
X-Orcpt: rfc822;linux-mips@linux-mips.org
Original-Recipient: rfc822;linux-mips@linux-mips.org
X-archive-position: 18546
X-ecartis-version: Ecartis v1.0.0
Sender: linux-mips-bounce@linux-mips.org
Errors-to: linux-mips-bounce@linux-mips.org
X-original-sender: root@lappi.linux-mips.net
Precedence: bulk
X-list: linux-mips
Content-Length: 2270
Lines: 70

Hello Ralf,

Thank you very much for the quick response.  I know
you must be terribly busy with the server move, so I
won't take up too much more of your time.

I did find an old (possibly related) discussion you
were involved in:

http://www.linux-mips.org/archives/linux-mips/2006-09/msg00194.html

I can't tell from the thread whether it was a problem
seen on 2.6.16.29, but that might have been tip at the
time.

> It's a bug which should be fixed but nevertheless I
> can highly recommend
> something like a SiliconImage SATA controller - the
> onboard PIO PATA
> controller is so slow.

I understand, but changing that is not an option for
me today.

> I've pushed the tag again so now there is a tarball.

Thanks.  I thought something was terribly wrong with
.28 for it to be skipped.  

> If you need to track something like this you're
> probably best with
> git bisect which should bring you right to the
> offending commit.

I probably should have used that approach instead of
diffing .27 and .29 and narrowing the file list by
hand.  As it was, there were changes to non-MIPS
platforms and devices I'm not using so those were easy
to apply to .27.  Also, there were many file changes
for MT SMP support, and *most* (but not all) of those
changes were wrapped with conditional compiles, so
those were also easy to apply.  I knew once I got to
these five files I was in some interesting code that
could point to the problems I'm seeing.

> Later kernels do run on bcm1480 which is close
> enough.

By "later kernels", do you mean 2.6.16.60 or different
kernel branches?

Perhaps, but I'm seeing identical failures in .29 and
.60, although that could be misleading.  I do see
significant SMP-related changes in c-sb1.c between .29
and .60, and I am running in SMP.  I wish there was an
easy way to know whether it's in the same code.

Anyway, in the interest of time I may revert to .27
which appears to work.  I don't need any of the MT SMP
related changes that followed, and perhaps I can live
without the enhancements between .27 and .60 for now.

Thanks,
Larry Stefani
lstefani@yahoo.com



      ____________________________________________________________________________________
Never miss a thing.  Make Yahoo your home page. 
http://www.yahoo.com/r/hs
From rkota@broadcom.com Wed Mar 26 17:28:37 2008
Received: with ECARTIS (v1.0.0; list linux-mips); Wed, 26 Mar 2008 17:28:44 +0100 (CET)
Received: from oss.sgi.com ([192.48.170.157]:54696 "EHLO oss.sgi.com")
	by lappi.linux-mips.net with ESMTP id S1100635AbYCZNEj (ORCPT
	<rfc822;linux-mips@linux-mips.org>); Wed, 26 Mar 2008 14:04:39 +0100
Received: from p549F5321.dip.t-dialin.net (localhost [127.0.0.1])
	by oss.sgi.com (8.12.11.20060308/8.12.11/SuSE Linux 0.7) with ESMTP id m2QBVULO016060
	for <linux-mips@linux-mips.org>; Wed, 26 Mar 2008 06:03:58 -0700
Received: from mms2.broadcom.com ([216.31.210.18]:18962 "EHLO
	mms2.broadcom.com") by lappi.linux-mips.net with ESMTP
	id S1102827AbYCZFV0 convert rfc822-to-8bit (ORCPT
	<rfc822;linux-mips@linux-mips.org>); Wed, 26 Mar 2008 06:21:26 +0100
Received: from [10.10.64.154] by mms2.broadcom.com with ESMTP (Broadcom
 SMTP Relay (Email Firewall v6.3.2)); Tue, 25 Mar 2008 22:20:45 -0700
X-Server-Uuid: D3C04415-6FA8-4F2C-93C1-920E106A2031
Received: by mail-irva-10.broadcom.com (Postfix, from userid 47) id
 B72B12B1; Tue, 25 Mar 2008 22:20:45 -0700 (PDT)
Received: from mail-irva-8.broadcom.com (mail-irva-8 [10.10.64.221]) by
 mail-irva-10.broadcom.com (Postfix) with ESMTP id A39182B0 for
 <linux-mips@linux-mips.org>; Tue, 25 Mar 2008 22:20:45 -0700 (PDT)
Received: from mail-sj1-12.sj.broadcom.com (mail-sj1-12.sj.broadcom.com
 [10.16.128.215]) by mail-irva-8.broadcom.com (MOS 3.7.5a-GA) with ESMTP
 id GRH15356; Tue, 25 Mar 2008 22:20:44 -0700 (PDT)
Received: from NT-SJCA-0752.brcm.ad.broadcom.com (nt-sjca-0752
 [10.16.192.222]) by mail-sj1-12.sj.broadcom.com (Postfix) with ESMTP id
 CFACF20501 for <linux-mips@linux-mips.org>; Tue, 25 Mar 2008 22:20:44
 -0700 (PDT)
X-MimeOLE: Produced By Microsoft Exchange V6.5
Content-class: urn:content-classes:message
MIME-Version: 1.0
Subject: MTD Partitions Permissions at Run Time.
Date:	Tue, 25 Mar 2008 22:20:43 -0700
Message-ID: <E06E3B7BBC07864CADE892DAF1EB0FBD0641257D@NT-SJCA-0752.brcm.ad.broadcom.com>
In-Reply-To: <47E7EBFF.6020406@ru.mvista.com>
Thread-Topic: MTD Partitions Permissions at Run Time.
Thread-Index: AciN2LlE4bF7ypqjS4Cg9fHtifqyCABJ9L3g
References: <47E559B6.9010001@ru.mvista.com>
 <47E7EBFF.6020406@ru.mvista.com>
From:	"Ramgopal Kota" <rkota@broadcom.com>
To:	linux-mips@linux-mips.org
X-WSS-ID: 6BF702A741K18759677-01-01
Content-Type: text/plain;
 charset=us-ascii
Content-Transfer-Encoding: 8BIT
X-Virus-Scanned: ClamAV 0.91.2/6021/Wed Feb 27 15:55:48 2008 on oss.sgi.com
X-Virus-Status:	Clean
Return-Path: <rkota@broadcom.com>
X-Envelope-To: <"|/home/ecartis/ecartis -s linux-mips"> (uid 0)
X-Orcpt: rfc822;linux-mips@linux-mips.org
Original-Recipient: rfc822;linux-mips@linux-mips.org
X-archive-position: 18567
X-ecartis-version: Ecartis v1.0.0
Sender: linux-mips-bounce@linux-mips.org
Errors-to: linux-mips-bounce@linux-mips.org
X-original-sender: rkota@broadcom.com
Precedence: bulk
X-list: linux-mips
Content-Length: 602
Lines: 18

Hi ,

Ours is a embedded product with pre-defined flash partitions say "Boot"
, "Kernel" , "Rootfs" & "ProductData".
The kernel which is in the box (linux 2.6.14)has Boot partition as "RO"
(Read Only). We have a requirement in the field that the boot code
should be upgraded also.
Is there any possibility of changing the partition permissions at
run-time? If so how ? 
Our solution was to create a Kernel image Boot as RW, upgrade the kernel
& reboot. Later change the Boot code.

But we are seeing if we can avoid such situation by remounting the MTD
partition with a RW permission.

Ramgopal Kota

Re: SB1250 locking up in init on current 2.6.16 kernel

[Thiemo Seufer]

Larry Stefani wrote:
> Hi,
> 
> I've been trying to upgrade from 2.6.16.18 to
> 2.6.16.60, but am seeing a hard lockup right before
> "INIT: version 2.78 booting" on my SB1250-based board.
> 
> I found a related discussion on the Debian mailing
> list:
> 
> http://groups.google.com/group/linux.debian.bugs.dist/browse_thread/thread/b7159ee25106c7f9
> 
> However, after applying Thiemo's patch to mark pages
> tainted by PIO IDE as dirty, the lockup still occurs.
> 
> I narrowed the file changes to
> 
>      arch/mips/mm/c-sb1.c
>      arch/mips/mm/cache.c
>      arch/mips/mm/init.c
>      include/asm-mips/cache-flush.h
>      include/asm-mips/page.h
> 
> between 2.6.16.27 and 2.6.16.29.  There was no
> 2.6.16.28 tarball posted on linux-mips.org, so I
> basically brought .27 to .29 until I found the
> offending files.

Please learn about git bisect (git-bisect start; git-bisect good;
git-bisect bad) and use it to isolate the offending commit.


Thiemo

Re: SB1250 locking up in init on current 2.6.16 kernel

[Larry Stefani]

Hi Ralf,

I used git bisect and narrowed the lockup to the
"[MIPS] Retire flush_icache_page from mm use." patch
(see git results below).  This is consistent with my
earlier testing and what Thiemo reported March 3 on
the linux.debian.kernel list.  I tried his patch (mark
pages tainted by PIO IDE as dirty) on 2.6.16.60, but
it didn't prevent the lockup.

Regards,
Larry Stefani
lstefani@yahoo.com

======================================================
/testkernels/linux> git bisect bad
Bisecting: 0 revisions left to test after this
[3a57f2ad7436d27dfa90717921b921fc3a168504] [MIPS]
Retire flush_icache_page from mm use.
/testkernels/linux> git bisect bad
3a57f2ad7436d27dfa90717921b921fc3a168504 is first bad
commit
commit 3a57f2ad7436d27dfa90717921b921fc3a168504
Author: Ralf Baechle <ralf@linux-mips.org>
Date:   Sat Aug 12 16:40:08 2006 +0100

    [MIPS] Retire flush_icache_page from mm use.

    On the 34K the redundant cache operations were
causing excessive stalls
    resulting in realtime code running on the second
VPE missing its deadline.
    For all other platforms this patch is just a
significant performance
    improvment as illustrated by below benchmark
numbers.

    Processor, Processes - times in microseconds -
smaller is better
   
------------------------------------------------------------------------------
    Host                 OS  Mhz null null      open
slct sig  sig  fork exec sh
                                 call  I/O stat clos
TCP  inst hndl proc proc proc
    --------- ------------- ---- ---- ---- ---- ----
---- ---- ---- ---- ---- ----
    25Kf      2.6.18-rc4     533 0.49 1.16 7.57 33.4
30.5 1.34 12.4 5497 17.K 54.K
    25Kf      2.6.18-rc4-p   533 0.49 1.16 6.68 23.0
30.7 1.36 8.55 5030 16.K 48.K
    4Kc       2.6.18-rc4      80 4.21 15.0 131. 289.
261. 16.5 258. 18.K 70.K 227K
    4Kc       2.6.18-rc4-p    80 4.34 13.1 128. 285.
262. 18.2 258. 12.K 52.K 176K
    34Kc      2.6.18-rc4      40 5.01 14.0 61.6 90.0
477. 17.9 94.7 29.K 108K 342K
    34Kc      2.6.18-rc4-p    40 4.98 13.9 61.2 89.7
475. 17.6 93.7 8758 44.K 158K
    BCM1480   2.6.18-rc4     700 0.28 0.60 3.68 5.92
16.0 0.78 5.08 931. 3163 15.K
    BCM1480   2.6.18-rc4-p   700 0.28 0.61 3.65 5.85
16.0 0.79 5.20 395. 1464 8385
    TX49-16K  2.6.18-rc3     197 0.73 2.41 19.0 37.8
82.9 2.94 17.5 4438 14.K 56.K
    TX49-16K  2.6.18-rc3-p   197 0.73 2.40 19.9 36.3
82.9 2.94 23.4 2577 9103 38.K
    TX49-32K  2.6.18-rc3     396 0.36 1.19 6.80 11.8
41.0 1.46 8.17 2738 8465 32.K
    TX49-32K  2.6.18-rc3-p   396 0.36 1.19 6.82 10.2
41.0 1.46 8.18 1330 4638 18.K

    Original patch by me with enhancements by Atsushi
Nemoto.

    Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
    Signed-off-by: Atsushi Nemoto
<anemo@mba.ocn.ne.jp>
    (cherry picked from
4bbd62a93a1ab4b7d8a5b402b0c78ac265b35661 commit)

:040000 040000
6b4aeaca51e1f8974394f5c85c36cfcbf40984c9
2af41ed07f6c83962f9a4871812adb6f7c0dbee7 M      arch
:040000 040000
c2fb4be66f03aed2a1746d30f0e17922086eefdd
ab8cb20f7da9b97e72742946adfb5ab927dc777a M     
include
=====================================================



--- Ralf Baechle <ralf@linux-mips.org> wrote:

> On Mon, Mar 24, 2008 at 07:00:15AM -0700, Larry
> Stefani wrote:
> 
> > I've been trying to upgrade from 2.6.16.18 to
> > 2.6.16.60, but am seeing a hard lockup right
> before
> > "INIT: version 2.78 booting" on my SB1250-based
> board.
> > 
> > I found a related discussion on the Debian mailing
> > list:
> > 
> >
>
http://groups.google.com/group/linux.debian.bugs.dist/browse_thread/thread/b7159ee25106c7f9
> > 
> > However, after applying Thiemo's patch to mark
> pages
> > tainted by PIO IDE as dirty, the lockup still
> occurs.
> 
> It's a bug which should be fixed but nevertheless I
> can highly recommend
> something like a SiliconImage SATA controller - the
> onboard PIO PATA
> controller is so slow.
> 
> > I narrowed the file changes to
> > 
> >      arch/mips/mm/c-sb1.c
> >      arch/mips/mm/cache.c
> >      arch/mips/mm/init.c
> >      include/asm-mips/cache-flush.h
> >      include/asm-mips/page.h
> > 
> > between 2.6.16.27 and 2.6.16.29.  There was no
> > 2.6.16.28 tarball posted on linux-mips.org, so I
> > basically brought .27 to .29 until I found the
> > offending files.
> 
> I've pushed the tag again so now there is a tarball.
> 
> If you need to track something like this you're
> probably best with
> git bisect which should bring you right to the
> offending commit.
> 
> > Is anyone running a 2.6.16 kernel (after
> 2.6.16.27) on
> > a SB1250-based board?
> 
> Later kernels do run on bcm1480 which is close
> enough.
> 
>   Ralf
> 




      ____________________________________________________________________________________
Be a better friend, newshound, and 
know-it-all with Yahoo! Mobile.  Try it now.  http://mobile.yahoo.com/;_ylt=Ahu06i62sR8HDtDypao8Wcj9tAcJ

Re: SB1250 locking up in init on current 2.6.16 kernel

[Thiemo Seufer]

Larry Stefani wrote:
> Hi Ralf,
> 
> I used git bisect and narrowed the lockup to the
> "[MIPS] Retire flush_icache_page from mm use." patch
> (see git results below).  This is consistent with my
> earlier testing and what Thiemo reported March 3 on
> the linux.debian.kernel list.  I tried his patch (mark
> pages tainted by PIO IDE as dirty) on 2.6.16.60, but
> it didn't prevent the lockup.

ISTR I got 2.6.16.60 to work by always enabling the cache flush in
ide.h (it is currently only run to clean out aliases). Keep in mind
that this is a crude workaround on top of other cache code hacks for
the SB-1.


Thiemo

Re: SB1250 locking up in init on current 2.6.16 kernel

[Larry Stefani]

Hi Thiemo,

I applied your patch (from
http://www.linux-mips.org/archives/linux-mips/2008-03/msg00001.html)
on 2.6.16.60, and also patched arch/mips/mm/c-sb1.c to
remove:

          local_flush_data_cache_page = (void *)
sb1_nop;

in order to compile after your changes to cache.c and
cacheflush.h.  However, this did not work on my board,
and I experienced the same lockup as before.

>>Keep in mind that this is a crude workaround on top
of other cache code hacks for the SB-1.

What other "cache code hacks for SB-1"?  Are there
additional changes required to 2.6.16.60 to make SB1
work properly?  Did you post those hacks somewhere?

Thanks,
Larry


--- Thiemo Seufer <ths@networkno.de> wrote:

> Larry Stefani wrote:
> > Hi Ralf,
> > 
> > I used git bisect and narrowed the lockup to the
> > "[MIPS] Retire flush_icache_page from mm use."
> patch
> > (see git results below).  This is consistent with
> my
> > earlier testing and what Thiemo reported March 3
> on
> > the linux.debian.kernel list.  I tried his patch
> (mark
> > pages tainted by PIO IDE as dirty) on 2.6.16.60,
> but
> > it didn't prevent the lockup.
> 
> ISTR I got 2.6.16.60 to work by always enabling the
> cache flush in
> ide.h (it is currently only run to clean out
> aliases). Keep in mind
> that this is a crude workaround on top of other
> cache code hacks for
> the SB-1.
> 
> 
> Thiemo
> 



      ____________________________________________________________________________________
Never miss a thing.  Make Yahoo your home page. 
http://www.yahoo.com/r/hs

Re: SB1250 locking up in init on current 2.6.16 kernel

[Thiemo Seufer]

Larry Stefani wrote:
> Hi Thiemo,
> 
> I applied your patch (from
> http://www.linux-mips.org/archives/linux-mips/2008-03/msg00001.html)
> on 2.6.16.60, and also patched arch/mips/mm/c-sb1.c to
> remove:
> 
>           local_flush_data_cache_page = (void *)
> sb1_nop;
> 
> in order to compile after your changes to cache.c and
> cacheflush.h.  However, this did not work on my board,
> and I experienced the same lockup as before.

Stick with the original 2.6.16.60 code but try to remove the

   if (cpu_has_dc_aliases) {

conditional in ide.h _without_ using my patch. This is what made
it boot for me.

> >>Keep in mind that this is a crude workaround on top
> of other cache code hacks for the SB-1.
> 
> What other "cache code hacks for SB-1"?  Are there
> additional changes required to 2.6.16.60 to make SB1
> work properly?  Did you post those hacks somewhere?

No, what I meant to say is that the old sb-1 cache code isn't quite
the most trustibe code, it had some holes which were papered over
by doing more cache flushes than necessary.


Thiemo

Re: SB1250 locking up in init on current 2.6.16 kernel

[Larry Stefani]

Hi Thiemo,

I went back to 2.6.16.60, reverted your patch changes,
then removed the

  if (cpu_has_dc_aliases) {

conditional expression in three places in
include/asm-mips/mach-generic/ide.h.

Unfortunately, I get the same lockup as before.

- Larry

--- Thiemo Seufer <ths@networkno.de> wrote:

> Larry Stefani wrote:
> > Hi Thiemo,
> > 
> > I applied your patch (from
> >
>
http://www.linux-mips.org/archives/linux-mips/2008-03/msg00001.html)
> > on 2.6.16.60, and also patched
> arch/mips/mm/c-sb1.c to
> > remove:
> > 
> >           local_flush_data_cache_page = (void *)
> > sb1_nop;
> > 
> > in order to compile after your changes to cache.c
> and
> > cacheflush.h.  However, this did not work on my
> board,
> > and I experienced the same lockup as before.
> 
> Stick with the original 2.6.16.60 code but try to
> remove the
> 
>    if (cpu_has_dc_aliases) {
> 
> conditional in ide.h _without_ using my patch. This
> is what made
> it boot for me.
> 
> > >>Keep in mind that this is a crude workaround on
> top
> > of other cache code hacks for the SB-1.
> > 
> > What other "cache code hacks for SB-1"?  Are there
> > additional changes required to 2.6.16.60 to make
> SB1
> > work properly?  Did you post those hacks
> somewhere?
> 
> No, what I meant to say is that the old sb-1 cache
> code isn't quite
> the most trustibe code, it had some holes which were
> papered over
> by doing more cache flushes than necessary.
> 
> 
> Thiemo
> 




      ____________________________________________________________________________________
Be a better friend, newshound, and 
know-it-all with Yahoo! Mobile.  Try it now.  http://mobile.yahoo.com/;_ylt=Ahu06i62sR8HDtDypao8Wcj9tAcJ