All of lore.kernel.org
 help / color / mirror / Atom feed
From: Brian King <brking@us.ibm.com>
To: James.Bottomley@steeleye.com
Cc: linux-scsi@vger.kernel.org, akpm@osdl.org, dougg@torque.net
Subject: Re: PROBLEM: Oops in 2.6.3 with lots of SG_IO activity
Date: Thu, 18 Mar 2004 15:48:20 -0600	[thread overview]
Message-ID: <405A1924.10303@us.ibm.com> (raw)

[-- Attachment #1: Type: text/plain, Size: 1228 bytes --]

James,

Attached is a patch to fix an oops in sg_cmd_done. Please apply.

Thanks

-Brian


-------- Original Message --------
Subject: Re: PROBLEM: Oops in 2.6.3 with lots of SG_IO activity
Date: Wed, 10 Mar 2004 10:24:39 -0600
From: Brian King <brking@us.ibm.com>
To: dougg@torque.net
CC: James.Bottomley@steeleye.com,  tonyb@cybernetics.com
References: <40478DD3.10807@us.ibm.com> <404B1C79.4060600@torque.net> <404CD74A.1090301@us.ibm.com> 
<404F3133.7060200@torque.net>

Douglas Gilbert wrote:

 > Brian,
 > Thanks for this test code. I don't follow the "run disk exercisers" bit.
 > BTW iprinit seg faulted when the sg module wasn't loaded
 >
 > Your patch widens the srp->done window and re-orders kill_fasync()
 > and wake_up_interruptible(). Are they both needed? If not which one
 > is critical?

I need both for my testcase to run clean. sg_cmd_done cannot touch sfp
once srp->done is set.

 > Anyway I'm happy to go ahead with the patch (posted by you a little
 > while later on the lsml). Having just moved accommodation my
 > equipment still needs more setting up. I have a sym53c8xx HBA.

Thanks. Once again, here is the patch. James, please apply.


-- 
Brian King
eServer Storage I/O
IBM Linux Technology Center

[-- Attachment #2: sg_cmd_done_oops.patch --]
[-- Type: text/plain, Size: 940 bytes --]


The patch fixes a race condition in sg_cmd_done that results in an oops.


---


diff -puN drivers/scsi/sg.c~sg_cmd_done_oops drivers/scsi/sg.c
--- linux-2.6.4-rc2/drivers/scsi/sg.c~sg_cmd_done_oops	2004-03-06 22:08:45.000000000 -0600
+++ linux-2.6.4-rc2-brking/drivers/scsi/sg.c	2004-03-06 22:55:12.000000000 -0600
@@ -1256,7 +1256,6 @@ sg_cmd_done(Scsi_Cmnd * SCpnt)
 	SRpnt->sr_request->rq_disk = NULL; /* "sg" _disowns_ request blk */
 
 	srp->my_cmdp = NULL;
-	srp->done = 1;
 
 	SCSI_LOG_TIMEOUT(4, printk("sg_cmd_done: %s, pack_id=%d, res=0x%x\n",
 		sdp->disk->disk_name, srp->header.pack_id, (int) SRpnt->sr_result));
@@ -1312,8 +1311,9 @@ sg_cmd_done(Scsi_Cmnd * SCpnt)
 	}
 	if (sfp && srp) {
 		/* Now wake up any sg_read() that is waiting for this packet. */
-		wake_up_interruptible(&sfp->read_wait);
 		kill_fasync(&sfp->async_qp, SIGPOLL, POLL_IN);
+		srp->done = 1;
+		wake_up_interruptible(&sfp->read_wait);
 	}
 }
 

_


             reply	other threads:[~2004-03-18 21:49 UTC|newest]

Thread overview: 2+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2004-03-18 21:48 Brian King [this message]
  -- strict thread matches above, loose matches on Subject: below --
2004-03-04 20:13 PROBLEM: Oops in 2.6.3 with lots of SG_IO activity Brian King

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=405A1924.10303@us.ibm.com \
    --to=brking@us.ibm.com \
    --cc=James.Bottomley@steeleye.com \
    --cc=akpm@osdl.org \
    --cc=dougg@torque.net \
    --cc=linux-scsi@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.