Google Groups no longer supports new Usenet posts or subscriptions. Historical content remains viewable.
Dismiss

PR/49862 CVS commit: src/sys/external/bsd/drm2/ttm

5 views
Skip to first unread message

Taylor R Campbell

unread,
Jul 27, 2015, 6:30:13 PM7/27/15
to
The following reply was made to PR kern/49862; it has been noted by GNATS.

From: "Taylor R Campbell" <rias...@netbsd.org>
To: gnats...@gnats.NetBSD.org
Cc:
Subject: PR/49862 CVS commit: src/sys/external/bsd/drm2/ttm
Date: Mon, 27 Jul 2015 22:26:51 +0000

Module Name: src
Committed By: riastradh
Date: Mon Jul 27 22:26:51 UTC 2015

Modified Files:
src/sys/external/bsd/drm2/ttm: ttm_bo_vm.c

Log Message:
Fix error branch: ttm_bo_unreserve on failure after ttm_bo_reserve.

Tiny chance this will fix PR kern/49862 by not leaking a ww_mutex
lock.


To generate a diff of this commit:
cvs rdiff -u -r1.8 -r1.9 src/sys/external/bsd/drm2/ttm/ttm_bo_vm.c

Please note that diffs are not public domain; they are subject to the
copyright notices on the relevant files.


--
Posted automagically by a mail2news gateway at muc.de e.V.
Please direct questions, flames, donations, etc. to news-...@muc.de

Thomas Klausner

unread,
Jul 27, 2015, 8:05:34 PM7/27/15
to
The following reply was made to PR kern/49862; it has been noted by GNATS.

From: Thomas Klausner <w...@NetBSD.org>
To: gnats...@NetBSD.org
Cc:
Subject: Re: PR/49862 CVS commit: src/sys/external/bsd/drm2/ttm
Date: Tue, 28 Jul 2015 02:04:09 +0200

On Mon, Jul 27, 2015 at 10:30:01PM +0000, Taylor R Campbell wrote:
> The following reply was made to PR kern/49862; it has been noted by GNATS.
>
> From: "Taylor R Campbell" <rias...@netbsd.org>
> To: gnats...@gnats.NetBSD.org
> Cc:
> Subject: PR/49862 CVS commit: src/sys/external/bsd/drm2/ttm
> Date: Mon, 27 Jul 2015 22:26:51 +0000
>
> Module Name: src
> Committed By: riastradh
> Date: Mon Jul 27 22:26:51 UTC 2015
>
> Modified Files:
> src/sys/external/bsd/drm2/ttm: ttm_bo_vm.c
>
> Log Message:
> Fix error branch: ttm_bo_unreserve on failure after ttm_bo_reserve.
>
> Tiny chance this will fix PR kern/49862 by not leaking a ww_mutex
> lock.

I've just retried my simple:

libreoffice *.doc

test (in a directory with about 20 doc files) and paniced.
SYSCALL 1 5 EXIT 0 7
WARNING: SPL NOT LOWERED ON SYSCALL 100 4 EXIT 0 6
radeon_ttm_fault() at WARNING: SPL NOT LOWERED ON SYSCALL 1 5 EXIT 0 7
WARNING: SPL NOT LOWERED ON SYSCALL 122 4 EXIT 0 6
WARNING: SPL NOT LOWERED ON SYSCALL 1 5 EXIT 0 7
WARNING: SPL NOT LOWERED ON SYSCALL 100 4 EXIT 0 6
WARNING: SPL NOT LOWERED ON SYSCALL 1 5 EXIT 0 7
WARNING: SPL NOT LOWERED ON SYSCALL 100 4 EXIT 0 6
netbsd:radeon_ttm_fault+0x6a
WARNING: SPL NOT LOWERED ON SYSCALL 1 5 EXIT 0 7
WARNING: SPL NOT LOWERED ON SYSCALL 129 4 EXIT 0 6
WARNING: SPL NOT LOWERED ON SYSCALL 1 5 EXIT 0 7
WARNING: SPL NOT LOWERED ON SYSCALL 100 4 EXIT 0 6
WARNING: SPL NOT LOWERED ON SYSCALL 1 5 EXIT 0 7
WARNING: SPL NOT LOWERED ON SYSCALL 100 4 EXIT 0 6
WARNING: SPL NOT LOWERED ON SYSCALL 1 5 EXIT 0 7
WARNING: SPL NOT LOWERED ON SYSCALL 100 4 EXIT 0 6
WARNING: SPL NOT LOWERED ON SYSCALL 1 5 EXIT 0 7
WARNING: SPL NOT LOWERED ON SYSCALL 100 4 EXIT 0 6
WARNING: SPL NOT LOWERED ON SYSCALL 1 5 EXIT 0 7
WARNING: SPL NOT LOWERED ON SYSCALL 100 4 EXIT 0 6
WARNING: SPL NOT LOWERED ON SYSCALL 1 5 EXIT 0 7
uvm_WfAaRuNlItN_Gi:n tSePrLn aNlO(T) LaOtW ERED ON SYSCALL 101 4 EXIT 0 6
WARNING: SPL NOT LOWERED ON SYSCALL 1 5 EXIT 0 7
WARNING: SPL NOT LOWERED ON SYSCALL 123 4 EXIT 1 6
WARNING: SPL NOT LOWERED ON SYSCALL 1 5 EXIT 0 7
netbsd:uvm_fault_internal+0x828
WARNING: SPL NOT LOWERED ON SYSCALL 132 4 EXIT 0 6
WARNING: SPL NOT LOWERED ON SYSCALL 1 5 EXIT 0 7
WARNING: SPL NOT LOWERED ON SYSCALL 100 4 EXIT 0 6
WARNING: SPL NOT LOWERED ON SYSCALL 1 5 EXIT 0 7
WARNING: SPL NOT LOWERED ON SYSCALL 100 4 EXIT 0 6
WARNING: SPL NOT LOWERED ON SYSCALL 1 5 EXIT 0 7
WARNING: SPL NOT LOWERED ON SYSCALL 100 4 EXIT 0 6
WARNING: SPL NOT LOWERED ON SYSCALL 1 5 EXIT 0 7
WARNING: SPL NOT LOWERED ON SYSCALL 100 4 EXIT 0 6
WARNING: SPL NOT LOWERED ON SYSCALL 1 5 EXIT 0 7
WARNING: SPL NOT LOWERED ON SYSCALL 100 4 EXIT 0 6
WARNING: SPL NOT LOWERED ON SYSCALL 1 5 EXIT 0 7
WARNING: SPL NOT LOWERED ON SYSCALL 100 4 EXIT 0 6
trap() at WARNING: SPL NOT LOWERED ON SYSCALL 1 5 EXIT 0 7
WARNING: SPL NOT LOWERED ON SYSCALL 110 4 EXIT 0 6
WARNING: SPL NOT LOWERED ON SYSCALL 1 5 EXIT 0 7
WARNING: SPL NOT LOWERED ON SYSCALL 100 4 EXIT 0 6
netbsd:trap+0x346
WARNING: SPL NOT LOWERED ON SYSCALL 1 5 EXIT 0 7
--- trap (number 6) ---
WARNING: SPL NOT LOWERED ON SYSCALL 142 4 EXIT 0 6
WARNING: SPL NOT LOWERED ON SYSCALL 1 5 EXIT 0 7
WARNING: SPL NOT LOWERED ON SYSCALL 100 4 EXIT 0 6
7f7fefe2563d:
WARNING: SPL NOT LOWERED ON SYSCALL 1 5 EXIT 0 7
cpu5: End traceback...
WARNING: SPL NOT LOWERED ON SYSCALL 137 4 EXIT 0 6

savecore: reboot after panic: panic: kernel diagnostic assertion "((mutex->wwm_state != WW_OWNED) || (mutex->wwm_u.owner != curlwp))" failed: file "/archive/foreign/src/sys/external/bsd/drm2/linux/linux_ww_mutex.c", line 760 locking 0xfffffe87Wf6ce7aA2R0N IaNgGa:i nSsPtL myself: 0xfffffe881ef51060
NOT LOWERED ON SYSCALL 207 4 EXIT 0 6
savecore: system went down at Tue Jul 28 01:56:40 2015

savecore: writing compressed core to /var/crash/netbsd.80.core.gz

Core and netbsd.gdb file (/netbsd.7.99.20c.gdb) available on request.
Thomas

Taylor R Campbell

unread,
Jul 27, 2015, 9:30:12 PM7/27/15
to
The following reply was made to PR kern/49862; it has been noted by GNATS.

From: "Taylor R Campbell" <rias...@netbsd.org>
To: gnats...@gnats.NetBSD.org
Cc:
Subject: PR/49862 CVS commit: src/sys/external/bsd/drm2/ttm
Date: Tue, 28 Jul 2015 01:25:00 +0000

Module Name: src
Committed By: riastradh
Date: Tue Jul 28 01:25:00 UTC 2015

Modified Files:
src/sys/external/bsd/drm2/ttm: ttm_bo_vm.c

Log Message:
ttm_bo_unreserve in case of ttm_bo_uvm_fault_idle restart.

Better chance of fixing PR kern/49862 by avoiding leaking a buffer
ww_mutex lock.


To generate a diff of this commit:
cvs rdiff -u -r1.9 -r1.10 src/sys/external/bsd/drm2/ttm/ttm_bo_vm.c

Please note that diffs are not public domain; they are subject to the
copyright notices on the relevant files.


dhol...@netbsd.org

unread,
Jul 28, 2015, 1:01:14 AM7/28/15
to
Synopsis: drm-induced kernel panic

State-Changed-From-To: open->feedback
State-Changed-By: dhol...@NetBSD.org
State-Changed-When: Tue, 28 Jul 2015 05:01:02 +0000
State-Changed-Why:
after I found the second of these, my radeon no longer explodes with similar
symptoms, so there's a good chance we've found this one.

Thomas Klausner

unread,
Jul 28, 2015, 4:30:13 AM7/28/15
to
The following reply was made to PR kern/49862; it has been noted by GNATS.

From: Thomas Klausner <w...@NetBSD.org>
To: gnats...@NetBSD.org
Cc:
Subject: Re: kern/49862 (drm-induced kernel panic)
Date: Tue, 28 Jul 2015 10:27:31 +0200

On Tue, Jul 28, 2015 at 05:01:02AM +0000, David Holland wrote:
> Synopsis: drm-induced kernel panic
>
> State-Changed-From-To: open->feedback
> State-Changed-By: dhol...@NetBSD.org
> State-Changed-When: Tue, 28 Jul 2015 05:01:02 +0000
> State-Changed-Why:
> after I found the second of these, my radeon no longer explodes with similar
> symptoms, so there's a good chance we've found this one.

Well, kind of.
soffice from libreoffice4 doesn't complete its startup for me any longer.
(xterm, firefox, gmpc still work).
Thomas

dhol...@netbsd.org

unread,
Jul 28, 2015, 4:31:51 AM7/28/15
to
Synopsis: drm-induced kernel panic

State-Changed-From-To: feedback->open
State-Changed-By: dhol...@NetBSD.org
State-Changed-When: Tue, 28 Jul 2015 08:31:45 +0000
State-Changed-Why:
alas, although maybe it's progress

Thomas Klausner

unread,
Jul 28, 2015, 4:45:16 AM7/28/15
to
The following reply was made to PR kern/49862; it has been noted by GNATS.

From: Thomas Klausner <w...@NetBSD.org>
To: NetBSD bugtracking <gnats...@NetBSD.org>
Cc:
Subject: Re: kern/49862 (drm-induced kernel panic)
Date: Tue, 28 Jul 2015 10:40:31 +0200

On Tue, Jul 28, 2015 at 08:30:00AM +0000, Thomas Klausner wrote:
> soffice from libreoffice4 doesn't complete its startup for me any longer.
> (xterm, firefox, gmpc still work).

More details:
The splash screen appears and gets to about 30 percent, then disappears.
Before the patch, after the splash screen finishes its progress bar,
the documents open, instead I get a core dump:

wiz@yt:~> gdb /usr/pkg/libreoffice4-4.4.4.3/program/soffice.bin soffice.bin.core
GNU gdb (GDB) 7.7.1
Copyright (C) 2014 Free Software Foundation, Inc.
License GPLv3+: GNU GPL version 3 or later <http://gnu.org/licenses/gpl.html>
This is free software: you are free to change and redistribute it.
There is NO WARRANTY, to the extent permitted by law. Type "show copying"
and "show warranty" for details.
This GDB was configured as "x86_64--netbsd".
Type "show configuration" for configuration details.
For bug reporting instructions, please see:
<http://www.gnu.org/software/gdb/bugs/>.
Find the GDB manual and other documentation resources online at:
<http://www.gnu.org/software/gdb/documentation/>.
For help, type "help".
Type "apropos word" to search for commands related to "word"...
Reading symbols from /usr/pkg/libreoffice4-4.4.4.3/program/soffice.bin...(no debugging symbols found)...done.
[New process 1]
[New process 4]
[New process 3]
[New process 2]
Core was generated by `soffice.bin'.
Program terminated with signal SIGSEGV, Segmentation fault.
#0 0x00007f7fe3c18eb9 in FcPatternObjectPosition () from /usr/pkg/lib/libfontconfig.so.1
(gdb) bt
#0 0x00007f7fe3c18eb9 in FcPatternObjectPosition () from /usr/pkg/lib/libfontconfig.so.1
#1 0x00007f7fe3c194aa in FcPatternObjectFindElt () from /usr/pkg/lib/libfontconfig.so.1
#2 0x00007f7fe3c1a087 in FcPatternObjectGet () from /usr/pkg/lib/libfontconfig.so.1
#3 0x00007f7fe3c1a216 in FcPatternObjectGetString () from /usr/pkg/lib/libfontconfig.so.1
#4 0x00007f7fe3c0b521 in FcConfigAddCache () from /usr/pkg/lib/libfontconfig.so.1
#5 0x00007f7fe3c0b668 in FcConfigAddDirList () from /usr/pkg/lib/libfontconfig.so.1
#6 0x00007f7fe3c0b7d1 in IA__FcConfigAppFontAddDir () from /usr/pkg/lib/libfontconfig.so.1
#7 0x00007f7ff08044c5 in psp::PrintFontManager::addFontconfigDir(rtl::OString const&) ()
from /usr/pkg/libreoffice4-4.4.4.3/program/libvcllo.so
#8 0x00007f7ff080d65b in psp::PrintFontManager::initialize() () from /usr/pkg/libreoffice4-4.4.4.3/program/libvcllo.so
#9 0x00007f7ff080f221 in psp::PrintFontManager::get() () from /usr/pkg/libreoffice4-4.4.4.3/program/libvcllo.so
#10 0x00007f7fda44882a in GtkSalGraphics::updateSettings(AllSettings&) () from /usr/pkg/libreoffice4-4.4.4.3/program/libvclplug_gtklo.so
#11 0x00007f7ff054ef24 in vcl::Window::ImplUpdateGlobalSettings(AllSettings&, bool) () from /usr/pkg/libreoffice4-4.4.4.3/program/libvcllo.so
#12 0x00007f7ff05e0b41 in vcl::Window::ImplInit(vcl::Window*, long, SystemParentData*) ()
from /usr/pkg/libreoffice4-4.4.4.3/program/libvcllo.so
#13 0x00007f7ff057666e in Dialog::ImplInit(vcl::Window*, long) () from /usr/pkg/libreoffice4-4.4.4.3/program/libvcllo.so
#14 0x00007f7ff0576b1a in Dialog::Dialog(vcl::Window*, long) () from /usr/pkg/libreoffice4-4.4.4.3/program/libvcllo.so
#15 0x00007f7ff0589bd8 in MessageDialog::MessageDialog(vcl::Window*, rtl::OUString const&, VclMessageType, VclButtonsType, long) ()
from /usr/pkg/libreoffice4-4.4.4.3/program/libvcllo.so
#16 0x00007f7ff70317b6 in desktop::Lockfile_execWarning(desktop::Lockfile*) () from /usr/pkg/libreoffice4-4.4.4.3/program/libsofficeapp.so
#17 0x00007f7ff3c1a970 in desktop::Lockfile::check(bool (*)(desktop::Lockfile*)) ()
from /usr/pkg/libreoffice4-4.4.4.3/program/libdeploymentmisclo.so
#18 0x00007f7ff7020140 in desktop::Desktop::Main() () from /usr/pkg/libreoffice4-4.4.4.3/program/libsofficeapp.so
#19 0x00007f7ff07802a3 in ImplSVMain() () from /usr/pkg/libreoffice4-4.4.4.3/program/libvcllo.so
#20 0x00007f7ff07802c3 in SVMain() () from /usr/pkg/libreoffice4-4.4.4.3/program/libvcllo.so
#21 0x00007f7ff7036fcf in soffice_main () from /usr/pkg/libreoffice4-4.4.4.3/program/libsofficeapp.so
#22 0x0000000000400aab in main ()
(gdb)

Taylor R Campbell

unread,
Jul 28, 2015, 11:37:37 AM7/28/15
to
Date: Tue, 28 Jul 2015 10:40:31 +0200
From: Thomas Klausner <w...@NetBSD.org>

More details:
The splash screen appears and gets to about 30 percent, then disappears.
Before the patch, after the splash screen finishes its progress bar,
the documents open, instead I get a core dump:

wiz@yt:~> gdb /usr/pkg/libreoffice4-4.4.4.3/program/soffice.bin soffice.bin.core
[...]
(gdb) bt
#0 0x00007f7fe3c18eb9 in FcPatternObjectPosition () from /usr/pkg/lib/libfontconfig.so.1
#1 0x00007f7fe3c194aa in FcPatternObjectFindElt () from /usr/pkg/lib/libfontconfig.so.1

Doesn't really look like a kernel or X bug to me -- looks like a
vanilla bug in fontconfig, unless FcPatternObjectPosition is working
on a graphics buffer or something.

I don't suppose you can rebuild with debug symbols and try again, can
you?

Taylor R Campbell

unread,
Jul 28, 2015, 11:40:12 AM7/28/15
to
The following reply was made to PR kern/49862; it has been noted by GNATS.

From: Taylor R Campbell <rias...@NetBSD.org>
To: gnats...@NetBSD.org
Cc: gnats...@netbsd.org, netbs...@netbsd.org, Thomas Klausner <w...@NetBSD.org>
Subject: Re: kern/49862 (drm-induced kernel panic)
Date: Tue, 28 Jul 2015 15:37:18 +0000

Date: Tue, 28 Jul 2015 10:40:31 +0200
From: Thomas Klausner <w...@NetBSD.org>

More details:
The splash screen appears and gets to about 30 percent, then disappears.
Before the patch, after the splash screen finishes its progress bar,
the documents open, instead I get a core dump:

wiz@yt:~> gdb /usr/pkg/libreoffice4-4.4.4.3/program/soffice.bin soffice.=
bin.core
[...]
(gdb) bt
#0 0x00007f7fe3c18eb9 in FcPatternObjectPosition () from /usr/pkg/lib/l=
ibfontconfig.so.1
#1 0x00007f7fe3c194aa in FcPatternObjectFindElt () from /usr/pkg/lib/li=

matthew green

unread,
Jul 28, 2015, 6:21:22 PM7/28/15
to

> From: Thomas Klausner <w...@NetBSD.org>
> To: gnats...@NetBSD.org
> Cc:
> Subject: Re: kern/49862 (drm-induced kernel panic)
> Date: Tue, 28 Jul 2015 10:27:31 +0200
>
> On Tue, Jul 28, 2015 at 05:01:02AM +0000, David Holland wrote:
> > Synopsis: drm-induced kernel panic
> >
> > State-Changed-From-To: open->feedback
> > State-Changed-By: dhol...@NetBSD.org
> > State-Changed-When: Tue, 28 Jul 2015 05:01:02 +0000
> > State-Changed-Why:
> > after I found the second of these, my radeon no longer explodes with similar
> > symptoms, so there's a good chance we've found this one.
>
> Well, kind of.
> soffice from libreoffice4 doesn't complete its startup for me any longer.
> (xterm, firefox, gmpc still work).

i think we should move this problem into another (new) PR as it
seems like the drm problem is solved.

matthew green

unread,
Jul 28, 2015, 6:25:13 PM7/28/15
to
The following reply was made to PR kern/49862; it has been noted by GNATS.

From: matthew green <m...@eterna.com.au>
To: gnats...@NetBSD.org
Cc: rias...@NetBSD.org, gnats...@netbsd.org, netbs...@netbsd.org,
Thomas Klausner <w...@NetBSD.org>
0 new messages