From: Bill Davidsen <davidsen@tmr.com>
To: Dan Williams <dan.j.williams@intel.com>
Cc: "BERTRAND Joël" <joel.bertrand@systella.fr>,
linux-raid@vger.kernel.org, sparclinux@vger.kernel.org
Subject: Re: [BUG] Raid5 trouble
Date: Thu, 18 Oct 2007 22:55:46 -0400 [thread overview]
Message-ID: <47181CB2.1060602@tmr.com> (raw)
In-Reply-To: <e9c3a7c20710170840u2ed8d6a9x26523eec6700ad11@mail.gmail.com>
Dan Williams wrote:
> I found a problem which may lead to the operations count dropping
> below zero. If ops_complete_biofill() gets preempted in between the
> following calls:
>
> raid5.c:554> clear_bit(STRIPE_OP_BIOFILL, &sh->ops.ack);
> raid5.c:555> clear_bit(STRIPE_OP_BIOFILL, &sh->ops.pending);
>
> ...then get_stripe_work() can recount/re-acknowledge STRIPE_OP_BIOFILL
> causing the assertion. In fact, the 'pending' bit should always be
> cleared first, but the other cases are protected by
> spin_lock(&sh->lock). Patch attached.
>
Once this patch has been vetted, can it be offered to -stable for
2.6.23? Or to be pedantic, it *can*, will you make that happen?
--
bill davidsen <davidsen@tmr.com>
CTO TMR Associates, Inc
Doing interesting things with small computers since 1979
WARNING: multiple messages have this Message-ID (diff)
From: Bill Davidsen <davidsen@tmr.com>
To: Dan Williams <dan.j.williams@intel.com>
Cc: "BERTRAND Joël" <joel.bertrand@systella.fr>,
linux-raid@vger.kernel.org, sparclinux@vger.kernel.org
Subject: Re: [BUG] Raid5 trouble
Date: Fri, 19 Oct 2007 02:55:46 +0000 [thread overview]
Message-ID: <47181CB2.1060602@tmr.com> (raw)
In-Reply-To: <e9c3a7c20710170840u2ed8d6a9x26523eec6700ad11@mail.gmail.com>
Dan Williams wrote:
> I found a problem which may lead to the operations count dropping
> below zero. If ops_complete_biofill() gets preempted in between the
> following calls:
>
> raid5.c:554> clear_bit(STRIPE_OP_BIOFILL, &sh->ops.ack);
> raid5.c:555> clear_bit(STRIPE_OP_BIOFILL, &sh->ops.pending);
>
> ...then get_stripe_work() can recount/re-acknowledge STRIPE_OP_BIOFILL
> causing the assertion. In fact, the 'pending' bit should always be
> cleared first, but the other cases are protected by
> spin_lock(&sh->lock). Patch attached.
>
Once this patch has been vetted, can it be offered to -stable for
2.6.23? Or to be pedantic, it *can*, will you make that happen?
--
bill davidsen <davidsen@tmr.com>
CTO TMR Associates, Inc
Doing interesting things with small computers since 1979
next prev parent reply other threads:[~2007-10-19 2:55 UTC|newest]
Thread overview: 72+ messages / expand[flat|nested] mbox.gz Atom feed top
2007-10-16 13:24 [BUG] Raid5 trouble BERTRAND Joël
2007-10-16 13:24 ` BERTRAND Joël
2007-10-17 14:32 ` BERTRAND Joël
2007-10-17 14:32 ` BERTRAND Joël
2007-10-17 14:58 ` Dan Williams
2007-10-17 14:58 ` Dan Williams
2007-10-17 15:40 ` Dan Williams
2007-10-17 15:40 ` Dan Williams
2007-10-17 16:44 ` BERTRAND Joël
2007-10-17 16:44 ` BERTRAND Joël
2007-10-18 0:46 ` Dan Williams
2007-10-18 0:46 ` Dan Williams
2007-10-18 8:29 ` BERTRAND Joël
2007-10-18 8:29 ` BERTRAND Joël
2007-10-19 2:55 ` Bill Davidsen [this message]
2007-10-19 2:55 ` Bill Davidsen
2007-10-19 8:04 ` BERTRAND Joël
2007-10-19 8:04 ` BERTRAND Joël
2007-10-19 15:51 ` Dan Williams
2007-10-19 15:51 ` Dan Williams
2007-10-19 16:03 ` BERTRAND Joël
2007-10-19 16:03 ` BERTRAND Joël
[not found] ` <4718DE66.8000905@tmr.com>
2007-10-19 20:42 ` BERTRAND Joël
2007-10-19 20:42 ` BERTRAND Joël
2007-10-19 20:49 ` [BUG] Raid1/5 over iSCSI trouble BERTRAND Joël
2007-10-19 20:49 ` BERTRAND Joël
2007-10-19 21:02 ` [Iscsitarget-devel] " Ross S. W. Walker
2007-10-19 21:02 ` Ross S. W. Walker
2007-10-19 21:06 ` BERTRAND Joël
2007-10-19 21:06 ` [Iscsitarget-devel] " BERTRAND Joël
2007-10-19 21:10 ` Ross S. W. Walker
2007-10-19 21:10 ` [Iscsitarget-devel] " Ross S. W. Walker
2007-10-20 7:45 ` BERTRAND Joël
2007-10-20 7:45 ` [Iscsitarget-devel] " BERTRAND Joël
2007-10-19 21:11 ` Scott Kaelin
2007-10-19 21:11 ` Scott Kaelin
2007-10-19 21:04 ` BERTRAND Joël
2007-10-19 21:04 ` BERTRAND Joël
2007-10-19 21:08 ` Ross S. W. Walker
2007-10-19 21:08 ` [Iscsitarget-devel] " Ross S. W. Walker
2007-10-19 21:12 ` Dan Williams
2007-10-19 21:12 ` Dan Williams
2007-10-20 8:05 ` BERTRAND Joël
2007-10-20 8:05 ` BERTRAND Joël
2007-10-24 7:12 ` BERTRAND Joël
2007-10-24 7:12 ` BERTRAND Joël
2007-10-24 20:10 ` Bill Davidsen
2007-10-24 20:10 ` Bill Davidsen
2007-10-24 23:49 ` Dan Williams
2007-10-24 23:49 ` Dan Williams
2007-10-25 0:03 ` David Miller
2007-10-25 0:03 ` David Miller
2007-10-27 13:29 ` BERTRAND Joël
2007-10-27 13:29 ` BERTRAND Joël
2007-10-27 18:27 ` Dan Williams
2007-10-27 18:27 ` Dan Williams
2007-10-27 19:35 ` BERTRAND Joël
2007-10-27 19:35 ` BERTRAND Joël
2007-10-27 21:13 ` Ming Zhang
2007-10-27 21:13 ` Ming Zhang
2007-10-29 10:40 ` BERTRAND Joël
2007-10-29 10:40 ` BERTRAND Joël
2007-10-19 21:19 ` Ming Zhang
2007-10-19 21:19 ` [Iscsitarget-devel] " Ming Zhang
2007-10-19 23:50 ` Bill Davidsen
2007-10-19 23:50 ` Bill Davidsen
2007-10-19 23:58 ` Bill Davidsen
2007-10-19 23:58 ` Bill Davidsen
2007-10-20 7:52 ` BERTRAND Joël
2007-10-20 7:52 ` BERTRAND Joël
2007-10-17 16:07 ` [BUG] Raid5 trouble BERTRAND Joël
2007-10-17 16:07 ` BERTRAND Joël
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=47181CB2.1060602@tmr.com \
--to=davidsen@tmr.com \
--cc=dan.j.williams@intel.com \
--cc=joel.bertrand@systella.fr \
--cc=linux-raid@vger.kernel.org \
--cc=sparclinux@vger.kernel.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.