deadlock or bad disk ? RELENG_8
Mike Tancsa
mike at sentex.net
Mon Jul 19 12:48:27 UTC 2010
At 12:11 AM 7/19/2010, Jeremy Chadwick wrote:
>On Sun, Jul 18, 2010 at 08:58:44PM -0700, Jeremy Chadwick wrote:
> > I took a look at the RELENG_8 code responsible for printing this
> > message: src/sys/vm/swap_pager.c
> >
> > [...]
> > 1086 static int
> > 1087 swap_pager_getpages(vm_object_t object, vm_page_t *m, int
> count, int reqpage)
> > 1088 {
> > [...]
>
>There was a change to this piece of code on May 13th. See commit
>1.312.2.4 below:
>
>http://www.freebsd.org/cgi/cvsweb.cgi/src/sys/vm/swap_pager.c
>
>Before I pull in kib@ or alc@ to help determine if what was introduced
>could cause what you're seeing, can you clarify this point?
>
> > FreeBSD 8.1-PRERELEASE #0: Tue Jul 13 09:55:48 EDT 2010
> > mdtancsa at backup3.sentex.ca:/usr/obj/usr/src/sys/backup amd64
> > [...]
> > Its never been an issue in the past 2yrs. The same box was RELENG_7
> > for some time and then in the past 3 months updated to RELENG_8.
>
>Has this system run a previous version of RELENG_8 within the past 3
Here is the recent history of what kernels the box was running
Dec 1 07:49:34 backup3 kernel: FreeBSD 7.2-STABLE #0: Mon Nov 30
23:18:16 EST 2009
Dec 1 08:06:17 backup3 kernel: FreeBSD 7.2-STABLE #0: Mon Nov 30
23:18:16 EST 2009
Mar 9 14:50:02 backup3 kernel: FreeBSD 7.3-PRERELEASE #1: Tue Mar 9
12:44:17 EST 2010
Mar 9 14:59:58 backup3 kernel: FreeBSD 7.3-PRERELEASE #1: Tue Mar 9
12:44:17 EST 2010
Apr 21 10:38:05 backup3 kernel: FreeBSD 8.0-STABLE #0: Wed Apr 21
10:19:12 EDT 2010
May 10 10:44:04 backup3 kernel: FreeBSD 8.0-STABLE #2: Mon May 10
09:30:36 EDT 2010
May 26 10:46:05 backup3 kernel: FreeBSD 8.1-PRERELEASE #0: Tue May 25
17:18:59 EDT 2010
Jul 2 13:57:32 backup3 kernel: FreeBSD 8.1-PRERELEASE #2: Fri Jul 2
12:07:22 EDT 2010
Jul 13 13:11:43 backup3 kernel: FreeBSD 8.1-PRERELEASE #0: Tue Jul 13
09:55:48 EDT 2010
Jul 18 13:46:15 backup3 kernel: FreeBSD 8.1-PRERELEASE #0: Tue Jul 13
09:55:48 EDT 2010
>months, or are you seeing this issue "as a result of upgrading from
>RELENG_7 to RELENG_8"?
The first time I have seen this was going from July 2 to the
13th. However, before drawing any conclusions, this could just be a
very odd edge case or hardware, or an actual regression. I think its
very hard to say at this point as I cant even reproduce it on demand.
---Mike
--------------------------------------------------------------------
Mike Tancsa, tel +1 519 651 3400
Sentex Communications, mike at sentex.net
Providing Internet since 1994 www.sentex.net
Cambridge, Ontario Canada www.sentex.net/mike
More information about the freebsd-stable
mailing list