Date: Sun, 23 Sep 2007 22:55:11 +0900
To: just.for.lkml@googlemail.com
Cc: jens.axboe@oracle.com, linux-kernel@vger.kernel.org,
       akpm@linux-foundation.org, linux-scsi@vger.kernel.org,
       linux-ide@vger.kernel.org
Cc: fujita.tomonori@lab.ntt.co.jp
Subject: Re: What's in linux-2.6-block.git for 2.6.24
From: FUJITA Tomonori <tomof@acm.org>
In-Reply-To: <64bb37e0709230619v196cf34ev29f0c844e2a79ec9@mail.gmail.com>
References: <20070921085711.GG2367@kernel.dk>
	<64bb37e0709230619v196cf34ev29f0c844e2a79ec9@mail.gmail.com>
Mime-Version: 1.0
Content-Type: Text/Plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Message-Id: <20070923225320I.tomof@acm.org>
Sender: linux-kernel-owner@vger.kernel.org
Content-Length: 5150
Lines: 106

On Sun, 23 Sep 2007 15:19:13 +0200
"Torsten Kaiser" <just.for.lkml@googlemail.com> wrote:

> On 9/21/07, Jens Axboe <jens.axboe@oracle.com> wrote:
> > SG chaining bits:
> > - This is the bulk of the patchset. It consists of three major
> >   components:
> >
> >         - sglist-core, which add helpers for iterating sg lists and
> >           switches the block layer and SCSI to use those. Should not
> >           have any functional changes.
> >         - sglist-drivers, which converts drivers to use the sg list
> >           helpers. Again, should not contain functional changes.
> >         - sglist-arch, which adds support to most architectures and
> >           actually enables sg chaining.
> 
> Adding linux-ide and linux-scsi as CC like Andrew did with my last report.
> 
> I still have trouble with my Silicon Image, Inc. SiI 3132 Serial ATA
> Raid II Controller as reported on 2.6.23-rc4-mm1 on the new
> 2.6.23-rc6-mm1.
> 
> I'm not 100% sure if this caused by the sg chaining, but the patch
> from http://lkml.org/lkml/2007/9/10/251 which touches that chaining
> makes a difference, so it might be related.
> 
> First report: http://lkml.org/lkml/2007/9/1/92
> With patch it fails fewer times: http://lkml.org/lkml/2007/9/14/107
> 
> To update the statistik:
> prior to 2.6.23-rc4-mm1: no trouble with any drives on the SiI 3132.
> 2.6.23-rc4-mm1 without patch: 2 out of 2 bad.
> back to 2.6.23-rc3-mm1: 18x good.
> 2.6.23-rc4-mm1 with patch:  2 out of 8 bad
> after that second mail:
> 2.6.23-rc4-mm1 with patch: 1 out of 5 bad
> 2.6.23-rc6-mm1: 1 out of 2 bad

git-block.patch in 2.6.23-rc6-mm1 includes my patch that disables sg
chaining for libata but it still includes libata's sg chaining
changes. So these changes breaks libata or libata was broken after
2.6.23-rc3-mm1.

Can you try Jens's sglist-arch branch? If it works, probably libata in
-mm has bugs.

For your convenience, I put a sglist-arch branch patch against v2.6.23-rc7:

http://www.kernel.org/pub/linux/kernel/people/tomo/misc/v2.6.23-rc7-sglist-arch.diff.bz2


> switching back to 2.6.23-rc3-mm1 to rule out the hardware:
> 2.6.23-rc3-mm1: 6x good
> 
> The error messages from the failed 2.6.23-rc6-mm1:
> Sep 18 18:50:01 treogen [   33.340000] md1: bitmap initialized from
> disk: read 10/10 pages, set 0 bits
> Sep 18 18:50:01 treogen [   33.340000] created bitmap (145 pages) for device md1
> Sep 18 18:50:01 treogen [   63.440000] ata1.00: exception Emask 0x0
> SAct 0x1 SErr 0x0 action 0x6 frozen
> Sep 18 18:50:01 treogen [   63.440000] ata1.00: cmd
> 61/08:00:09:d6:42/00:00:25:00:00/40 tag 0 cdb 0x0 data 4096 out
> Sep 18 18:50:01 treogen [   63.440000]          res
> 40/00:00:00:00:00/00:00:00:00:00/00 Emask 0x4 (timeout)
> Sep 18 18:50:01 treogen [   63.440000] ata1.00: status: {DRDY }
> Sep 18 18:50:01 treogen [   63.440000] ata1: hard resetting link
> Sep 18 18:50:01 treogen [   65.740000] ata1: softreset failed (port not ready)
> Sep 18 18:50:01 treogen [   65.740000] ata1: reset failed (errno=-5),
> retrying in 8 secs
> Sep 18 18:50:01 treogen [   73.440000] ata1: hard resetting link
> Sep 18 18:50:01 treogen [   75.740000] ata1: softreset failed (port not ready)
> Sep 18 18:50:01 treogen [   75.740000] ata1: reset failed (errno=-5),
> retrying in 8 secs
> Sep 18 18:50:01 treogen [   83.440000] ata1: hard resetting link
> Sep 18 18:50:01 treogen [   85.740000] ata1: softreset failed (port not ready)
> Sep 18 18:50:01 treogen [   85.740000] ata1: reset failed (errno=-5),
> retrying in 33 secs
> Sep 18 18:50:01 treogen [  118.440000] ata1: limiting SATA link speed
> to 1.5 Gbps
> Sep 18 18:50:01 treogen [  118.440000] ata1: hard resetting link
> Sep 18 18:50:01 treogen [  120.740000] ata1: softreset failed (port not ready)
> Sep 18 18:50:01 treogen [  120.740000] ata1: reset failed, giving up
> Sep 18 18:50:01 treogen [  120.740000] ata1.00: disabled
> Sep 18 18:50:01 treogen [  120.740000] ata1: EH complete
> Sep 18 18:50:01 treogen [  120.740000] sd 0:0:0:0: [sda] Result:
> hostbyte=DID_BAD_TARGET driverbyte=DRIVER_OK,SUGGEST_OK
> Sep 18 18:50:01 treogen [  120.740000] end_request: I/O error, dev
> sda, sector 625137161
> Sep 18 18:50:01 treogen [  120.740000] md: super_written gets
> error=-5, uptodate=0
> Sep 18 18:50:01 treogen [  120.740000] raid5: Disk failure on sda2,
> disabling device. Operation continuing on 2 devices
> 
> After that many more errors like this, only differing in the sector number:
> Sep 18 18:50:01 treogen [  120.810000] sd 0:0:0:0: [sda] Result:
> hostbyte=DID_BAD_TARGET driverbyte=DRIVER_OK,SUGGEST_OK
> Sep 18 18:50:01 treogen [  120.810000] end_request: I/O error, dev
> sda, sector 19550919
> 
> Any more infos needed?
> 
> Torsten
> -
> To unsubscribe from this list: send the line "unsubscribe linux-scsi" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/