All of lore.kernel.org
 help / color / mirror / Atom feed
* Re: [PATCH 0/5][-next] TTM bo reservation and fencing improvements
@ 2010-11-17 16:40 Sedat Dilek
  2010-11-17 17:44 ` Thomas Hellstrom
  0 siblings, 1 reply; 6+ messages in thread
From: Sedat Dilek @ 2010-11-17 16:40 UTC (permalink / raw)
  To: Thomas Hellstrom, DRI

Against which GIT tree is this patchset, against latest drm-next GIT they fail?
Missing a patch or patchset?

- Sedat -

$ cd /mnt/sdb5/linux-kernel/drm-2.6

$ git branch
* drm-next
  master

$ git describe
v2.6.34-rc5-32040-g01d5a1f

$ cat patches/series
drm-ttm-next/1-5-drm-ttm-Optimize-ttm_eu_backoff_reservation.patch
drm-ttm-next/2-5-drm-ttm-Don-t-deadlock-on-recursive-multi-bo-reservations.patch
drm-ttm-next/3-5-drm-ttm-radeon-nouveau-Kill-the-bo-lock-in-favour-of-a-bo-device-fence_lock.patch
drm-ttm-next/4-5-drm-ttm-Improved-fencing-of-buffer-object-lists.patch
drm-ttm-next/5-5-drm-ttm-vmwgfx-Have-TTM-manage-the-validation-sequence..patch
drm-ttm-next/1-1-drm-radeon-Use-the-ttm-execbuf-utilities.patch

$ quilt -f push
Wende Patch drm-ttm-next/1-5-drm-ttm-Optimize-ttm_eu_backoff_reservation.patch
an
patching file drivers/gpu/drm/ttm/ttm_execbuf_util.c
Hunk #1 succeeded at 35 (offset -64 lines).

Jetzt in Patch drm-ttm-next/1-5-drm-ttm-Optimize-ttm_eu_backoff_reservation.patch
$ quilt -f push
Wende Patch drm-ttm-next/2-5-drm-ttm-Don-t-deadlock-on-recursive-multi-bo-reservations.patch
an
patching file drivers/gpu/drm/ttm/ttm_bo.c
Hunk #1 FAILED at 223.
1 out of 1 hunk FAILED -- saving rejects to file
drivers/gpu/drm/ttm/ttm_bo.c.rej
patching file include/drm/ttm/ttm_bo_driver.h
Hunk #2 FAILED at 871.
1 out of 2 hunks FAILED -- saving rejects to file
include/drm/ttm/ttm_bo_driver.h.rej
Patch drm-ttm-next/2-5-drm-ttm-Don-t-deadlock-on-recursive-multi-bo-reservations.patch
angewandt (erzwungen, muss aktualisiert werden (Refresh))

^ permalink raw reply	[flat|nested] 6+ messages in thread
* Re: [PATCH 0/5][-next] TTM bo reservation and fencing improvements
  2010-11-17 16:40 [PATCH 0/5][-next] TTM bo reservation and fencing improvements Sedat Dilek
@ 2010-11-17 17:44 ` Thomas Hellstrom
  2010-11-17 20:17   ` Sedat Dilek
  2010-11-17 22:42   ` Sedat Dilek
  0 siblings, 2 replies; 6+ messages in thread
From: Thomas Hellstrom @ 2010-11-17 17:44 UTC (permalink / raw)
  To: sedat.dilek; +Cc: Sedat Dilek, DRI

On 11/17/2010 05:40 PM, Sedat Dilek wrote:
> Against which GIT tree is this patchset, against latest drm-next GIT they fail?
> Missing a patch or patchset?
>    
On top of drm-next you first need to apply

drm/ttm:  Fix up a theoretical deadlock.
drm/ttm: Add a bo list reserve fastpath (need v2 which was recently posted)
kref: Add a kref_sub function
drm/ttm: Use kref_sub instead of repeatedly...

/Thomas

> - Sedat -
>
> $ cd /mnt/sdb5/linux-kernel/drm-2.6
>
> $ git branch
> * drm-next
>    master
>
> $ git describe
> v2.6.34-rc5-32040-g01d5a1f
>
> $ cat patches/series
> drm-ttm-next/1-5-drm-ttm-Optimize-ttm_eu_backoff_reservation.patch
> drm-ttm-next/2-5-drm-ttm-Don-t-deadlock-on-recursive-multi-bo-reservations.patch
> drm-ttm-next/3-5-drm-ttm-radeon-nouveau-Kill-the-bo-lock-in-favour-of-a-bo-device-fence_lock.patch
> drm-ttm-next/4-5-drm-ttm-Improved-fencing-of-buffer-object-lists.patch
> drm-ttm-next/5-5-drm-ttm-vmwgfx-Have-TTM-manage-the-validation-sequence..patch
> drm-ttm-next/1-1-drm-radeon-Use-the-ttm-execbuf-utilities.patch
>
> $ quilt -f push
> Wende Patch drm-ttm-next/1-5-drm-ttm-Optimize-ttm_eu_backoff_reservation.patch
> an
> patching file drivers/gpu/drm/ttm/ttm_execbuf_util.c
> Hunk #1 succeeded at 35 (offset -64 lines).
>
> Jetzt in Patch drm-ttm-next/1-5-drm-ttm-Optimize-ttm_eu_backoff_reservation.patch
> $ quilt -f push
> Wende Patch drm-ttm-next/2-5-drm-ttm-Don-t-deadlock-on-recursive-multi-bo-reservations.patch
> an
> patching file drivers/gpu/drm/ttm/ttm_bo.c
> Hunk #1 FAILED at 223.
> 1 out of 1 hunk FAILED -- saving rejects to file
> drivers/gpu/drm/ttm/ttm_bo.c.rej
> patching file include/drm/ttm/ttm_bo_driver.h
> Hunk #2 FAILED at 871.
> 1 out of 2 hunks FAILED -- saving rejects to file
> include/drm/ttm/ttm_bo_driver.h.rej
> Patch drm-ttm-next/2-5-drm-ttm-Don-t-deadlock-on-recursive-multi-bo-reservations.patch
> angewandt (erzwungen, muss aktualisiert werden (Refresh))
>    

^ permalink raw reply	[flat|nested] 6+ messages in thread
* Re: [PATCH 0/5][-next] TTM bo reservation and fencing improvements
  2010-11-17 17:44 ` Thomas Hellstrom
@ 2010-11-17 20:17   ` Sedat Dilek
  2010-11-17 22:42   ` Sedat Dilek
  1 sibling, 0 replies; 6+ messages in thread
From: Sedat Dilek @ 2010-11-17 20:17 UTC (permalink / raw)
  To: Thomas Hellstrom; +Cc: DRI

On Wed, Nov 17, 2010 at 6:44 PM, Thomas Hellstrom <thellstrom@vmware.com> wrote:
> On 11/17/2010 05:40 PM, Sedat Dilek wrote:
>>
>> Against which GIT tree is this patchset, against latest drm-next GIT they
>> fail?
>> Missing a patch or patchset?
>>
>
> On top of drm-next you first need to apply
>
> drm/ttm:  Fix up a theoretical deadlock.
> drm/ttm: Add a bo list reserve fastpath (need v2 which was recently posted)
> kref: Add a kref_sub function
> drm/ttm: Use kref_sub instead of repeatedly...
>
> /Thomas
>

Thanks for the quick help, I have totally overseen
"drm-ttm-Fix-up-a-theoretical-deadlock.patch" in my series.

- Sedat -

  (+) OK   drm-ttm/drm-ttm-Fix-up-a-theoretical-deadlock.patch
  (+) OK   drm-ttm/v2-drm-ttm-Add-a-bo-list-reserve-fastpath.patch
  (+) OK   drm-ttm/1-2-kref-Add-a-kref_sub-function.patch
  (+) OK   drm-ttm/2-2-drm-ttm-Use-kref_sub-instead-of-repeatedly-calling-kref_put.patch
  (+) OK   drm-ttm-next/1-5-drm-ttm-Optimize-ttm_eu_backoff_reservation.patch
  (+) OK   drm-ttm-next/2-5-drm-ttm-Don-t-deadlock-on-recursive-multi-bo-reservations.patch
  (+) OK   drm-ttm-next/3-5-drm-ttm-radeon-nouveau-Kill-the-bo-lock-in-favour-of-a-bo-device-fence_lock.patch
  (+) OK   drm-ttm-next/4-5-drm-ttm-Improved-fencing-of-buffer-object-lists.patch
  (+) OK   drm-ttm-next/5-5-drm-ttm-vmwgfx-Have-TTM-manage-the-validation-sequence.patch
  (+) OK   drm-ttm-next/1-1-drm-radeon-Use-the-ttm-execbuf-utilities.patch

>> - Sedat -
>>
>> $ cd /mnt/sdb5/linux-kernel/drm-2.6
>>
>> $ git branch
>> * drm-next
>>   master
>>
>> $ git describe
>> v2.6.34-rc5-32040-g01d5a1f
>>
>> $ cat patches/series
>> drm-ttm-next/1-5-drm-ttm-Optimize-ttm_eu_backoff_reservation.patch
>>
>> drm-ttm-next/2-5-drm-ttm-Don-t-deadlock-on-recursive-multi-bo-reservations.patch
>>
>> drm-ttm-next/3-5-drm-ttm-radeon-nouveau-Kill-the-bo-lock-in-favour-of-a-bo-device-fence_lock.patch
>> drm-ttm-next/4-5-drm-ttm-Improved-fencing-of-buffer-object-lists.patch
>>
>> drm-ttm-next/5-5-drm-ttm-vmwgfx-Have-TTM-manage-the-validation-sequence..patch
>> drm-ttm-next/1-1-drm-radeon-Use-the-ttm-execbuf-utilities.patch
>>
>> $ quilt -f push
>> Wende Patch
>> drm-ttm-next/1-5-drm-ttm-Optimize-ttm_eu_backoff_reservation.patch
>> an
>> patching file drivers/gpu/drm/ttm/ttm_execbuf_util.c
>> Hunk #1 succeeded at 35 (offset -64 lines).
>>
>> Jetzt in Patch
>> drm-ttm-next/1-5-drm-ttm-Optimize-ttm_eu_backoff_reservation.patch
>> $ quilt -f push
>> Wende Patch
>> drm-ttm-next/2-5-drm-ttm-Don-t-deadlock-on-recursive-multi-bo-reservations.patch
>> an
>> patching file drivers/gpu/drm/ttm/ttm_bo.c
>> Hunk #1 FAILED at 223.
>> 1 out of 1 hunk FAILED -- saving rejects to file
>> drivers/gpu/drm/ttm/ttm_bo.c.rej
>> patching file include/drm/ttm/ttm_bo_driver.h
>> Hunk #2 FAILED at 871.
>> 1 out of 2 hunks FAILED -- saving rejects to file
>> include/drm/ttm/ttm_bo_driver.h.rej
>> Patch
>> drm-ttm-next/2-5-drm-ttm-Don-t-deadlock-on-recursive-multi-bo-reservations.patch
>> angewandt (erzwungen, muss aktualisiert werden (Refresh))
>>
>
>
_______________________________________________
dri-devel mailing list
dri-devel@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/dri-devel

^ permalink raw reply	[flat|nested] 6+ messages in thread
* Re: [PATCH 0/5][-next] TTM bo reservation and fencing improvements
  2010-11-17 17:44 ` Thomas Hellstrom
  2010-11-17 20:17   ` Sedat Dilek
@ 2010-11-17 22:42   ` Sedat Dilek
  1 sibling, 0 replies; 6+ messages in thread
From: Sedat Dilek @ 2010-11-17 22:42 UTC (permalink / raw)
  To: Thomas Hellstrom; +Cc: DRI

Hi,

I have tested Thomas drm-ttm(-next) patch-sets in combination with
Daniel Vetter's patch-sets (series see below).
No bad - no good in OpenArena-benchmark.

( It is definitely pageflip-support patches causing a slow-down of 4-5
fps... 13-14 vs. 17-18 fps on a radeon RV250 is a real difference. )

Kind Regards,
- Sedat -

$ cd ~/src/mesa/

sd@tbox:~/src/mesa$ ./scripts/run_openarena-benchmark.sh
840 frames 49.0 seconds 17.1 fps 10.0/58.4/198.0/19.3 ms
sd@tbox:~/src/mesa$ ./scripts/run_openarena-benchmark.sh
840 frames 47.6 seconds 17.7 fps 10.0/56.7/146.0/18.8 ms
sd@tbox:~/src/mesa$ ./scripts/run_openarena-benchmark.sh
840 frames 50.2 seconds 16.7 fps 11.0/59.8/147.0/19.0 ms
sd@tbox:~/src/mesa$ ./scripts/run_openarena-benchmark.sh
840 frames 47.5 seconds 17.7 fps 11.0/56.5/136.0/19.3 ms
sd@tbox:~/src/mesa$ ./scripts/run_openarena-benchmark.sh
840 frames 45.3 seconds 18.6 fps 10.0/53.9/146.0/19.1 ms

$ cd ~/src/linux-2.6$ grep OK setup_linux-2.6_git2.sd.2.log
  (+) OK   upstream/patch-2.6.37-rc2-git2
  (+) OK   upstream/revert-extraversion.diff
  (+) OK   kbuild-2.6-for-next/kbuild-2.6-for-next.patch
  (+) OK   bkl-config/0001-staging-stradis-mark-as-depends-on-BKL.patch
  (+) OK   bkl-config/0002-drm-i810-remove-the-BKL.patch
  (+) OK   bkl-config/0003-BKL-remove-extraneous-include-smp_lock.h.patch
  (+) OK   bkl-config/0004-BKL-remove-references-to-lock_kernel-from-comments.patch
  (+) OK   bkl-config/0005-BKL-disable-by-default.patch
  (+) OK   bkl-config/0006-BKL-mark-lock_kernel-as-deprecated.patch
  (+) OK   bkl-config/0007-BKL-move-CONFIG_BKL-to-staging.patch
  (+) OK   danvet-drm-for-sedat-dilek/0001-drm-nouveau-don-t-munge-in-drm_mm-internals.patch
  (+) OK   danvet-drm-for-sedat-dilek/0002-drm_mm-add-support-for-range-restricted-fair-lru-sca.patch
  (+) OK   danvet-drm-for-sedat-dilek/0003-drm-mm-track-free-areas-implicitly.patch
  (+) OK   danvet-drm-for-sedat-dilek/0004-drm-mm-extract-node-insert-helper-functions.patch
  (+) OK   danvet-drm-for-sedat-dilek/0005-drm-mm-add-api-for-embedding-struct-drm_mm_node.patch
  (+) OK   danvet-drm-for-sedat-dilek/0006-drm-mm-add-helper-to-unwind-scan-state.patch
  (+) OK   danvet-embed-drm_gem_object-into-radeon_bo/1-3-drm-radeon-embed-struct-drm_gem_object.patch
  (+) OK   danvet-embed-drm_gem_object-into-radeon_bo/2-3-drm-radeon-introduce-gem_to_radeon_bo-helper.patch
  (+) OK   danvet-embed-drm_gem_object-into-radeon_bo/3-3-drm-radeon-kill-radeon_bo--gobj-pointer.patch
  (+) OK   drm-vblank-timestamping/0001-drm-vblank-Add-support-for-precise-vblank-timestampi.patch
  (+) OK   drm-vblank-timestamping/0002-drm-radeon-Add-support-for-precise-vblank-timestampi.patch
  (+) OK   for-drm-radeon-testing/drm-radeon-kms-enable-writeback-on-radeon-AGP-boards.patch
  (+) OK   backlight-type/Backlight-Add-backlight-type-v2.patch
  (+) OK   backlight-type/radeon-Expose-backlight-class-device-for-legacy-LVDS-encoder.patch
  (+) OK   drm-ttm/drm-ttm-Fix-up-a-theoretical-deadlock.patch
  (+) OK   drm-ttm/v2-drm-ttm-Add-a-bo-list-reserve-fastpath.patch
  (+) OK   drm-ttm/1-2-kref-Add-a-kref_sub-function.patch
  (+) OK   drm-ttm/2-2-drm-ttm-Use-kref_sub-instead-of-repeatedly-calling-kref_put.patch
  (+) OK   drm-ttm-next/1-5-drm-ttm-Optimize-ttm_eu_backoff_reservation.patch
  (+) OK   drm-ttm-next/2-5-drm-ttm-Don-t-deadlock-on-recursive-multi-bo-reservations.patch
  (+) OK   drm-ttm-next/3-5-drm-ttm-radeon-nouveau-Kill-the-bo-lock-in-favour-of-a-bo-device-fence_lock.patch
  (+) OK   drm-ttm-next/4-5-drm-ttm-Improved-fencing-of-buffer-object-lists.patch
  (+) OK   drm-ttm-next/5-5-drm-ttm-vmwgfx-Have-TTM-manage-the-validation-sequence.patch
  (+) OK   drm-ttm-next/1-1-drm-radeon-Use-the-ttm-execbuf-utilities-for-danvet.patch
  (+) OK   for-drm-fixes/fixes-drm-vmwgfx-Fix-up-an-error-path-during-bo-creation.patch
  (+) OK   mm-fixes/cgroup-Avoid-a-memset-by-using-vzalloc.patch
  (+) OK   mm-fixes/kmemleak-remove-memset-by-using-kzalloc.patch
  (+) OK   tpm-fix/tpm-Autodetect-itpm-devices.patch
  (+) OK   debian/version.patch
  (+) OK   debian/kernelvariables-2.6.37.patch
  (+) OK   debian/doc-build-parallel.patch
  (+) OK   bugfix/ia64/hardcode-arch-script-output.patch
  (+) OK   bugfix/mips/disable-advansys.patch
  (+) OK   bugfix/arm/disable-scsi_acard.patch
  (+) OK   debian/mips-disable-werror.patch
  (+) OK   bugfix/powerpc/lpar-console.patch
  (+) OK   features/all/i915-autoload-without-CONFIG_DRM_I915_KMS.patch
  (+) OK   debian/arch-sh4-fix-uimage-build.patch
  (+) OK   bugfix/mips/mips-ide-flush-dcache.patch
  (+) OK   bugfix/all/qla4xxx-Fix-build-on-some-architectures-lacking-64-bit-I-O.patch
  (+) OK   bugfix/x86/Skip-looking-for-ioapic-overrides-when-ioapics-are-not-present.patch

$ diff -Naur 1-1-drm-radeon-Use-the-ttm-execbuf-utilities.patch
1-1-drm-radeon-Use-the-ttm-execbuf-utilities-for-danvet.patch
--- 1-1-drm-radeon-Use-the-ttm-execbuf-utilities.patch  2010-11-17
16:52:20.483483002 +0100
+++ 1-1-drm-radeon-Use-the-ttm-execbuf-utilities-for-danvet.patch
 2010-11-17 21:10:14.332144001 +0100
@@ -57,7 +57,7 @@
 +++ b/drivers/gpu/drm/radeon/radeon_cs.c
 @@ -77,13 +77,13 @@ int radeon_cs_parser_relocs(struct radeon_cs_parser *p)
                        p->relocs_ptr[i] = &p->relocs[i];
-                       p->relocs[i].robj = p->relocs[i].gobj->driver_private;
+                       p->relocs[i].robj = gem_to_radeon_bo(p->relocs[i].gobj);
                        p->relocs[i].lobj.bo = p->relocs[i].robj;
 -                      p->relocs[i].lobj.rdomain = r->read_domains;
                        p->relocs[i].lobj.wdomain = r->write_domain;

^ permalink raw reply	[flat|nested] 6+ messages in thread
* Re: [PATCH 0/5][-next] TTM bo reservation and fencing improvements
  2010-11-17 12:28 Thomas Hellstrom
@ 2010-11-17 19:07 ` Jerome Glisse
  0 siblings, 0 replies; 6+ messages in thread
From: Jerome Glisse @ 2010-11-17 19:07 UTC (permalink / raw)
  To: Thomas Hellstrom; +Cc: airlied, dri-devel

On Wed, Nov 17, 2010 at 7:28 AM, Thomas Hellstrom <thellstrom@vmware.com> wrote:
> This patch series improves on the performance and correctness of the
> reserving / fencing sequence.
>

Reviewed-by: Jerome Glisse <jglisse@redhat.com>

^ permalink raw reply	[flat|nested] 6+ messages in thread
* [PATCH 0/5][-next] TTM bo reservation and fencing improvements
@ 2010-11-17 12:28 Thomas Hellstrom
  2010-11-17 19:07 ` Jerome Glisse
  0 siblings, 1 reply; 6+ messages in thread
From: Thomas Hellstrom @ 2010-11-17 12:28 UTC (permalink / raw)
  To: airlied; +Cc: dri-devel

This patch series improves on the performance and correctness of the
reserving / fencing sequence.

^ permalink raw reply	[flat|nested] 6+ messages in thread
end of thread, other threads:[~2010-11-17 22:42 UTC | newest]

Thread overview: 6+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2010-11-17 16:40 [PATCH 0/5][-next] TTM bo reservation and fencing improvements Sedat Dilek
2010-11-17 17:44 ` Thomas Hellstrom
2010-11-17 20:17   ` Sedat Dilek
2010-11-17 22:42   ` Sedat Dilek
  -- strict thread matches above, loose matches on Subject: below --
2010-11-17 12:28 Thomas Hellstrom
2010-11-17 19:07 ` Jerome Glisse

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.