platform/upstream/libdrm.git
16 years agoradeon: add some more r100 support to test AGP
Dave Airlie [Sun, 2 Nov 2008 23:23:19 +0000 (09:23 +1000)]
radeon: add some more r100 support to test AGP

16 years agoMerge branch 'modesetting-gem' of ssh://git.freedesktop.org/git/mesa/drm into modeset...
Jerome Glisse [Wed, 29 Oct 2008 22:41:48 +0000 (23:41 +0100)]
Merge branch 'modesetting-gem' of ssh://git.freedesktop.org/git/mesa/drm into modesetting-gem

16 years agoradeon: libdrm_radeon updates bo & cs interfaces
Jerome Glisse [Wed, 29 Oct 2008 22:40:20 +0000 (23:40 +0100)]
radeon: libdrm_radeon updates bo & cs interfaces

16 years agolibdrm: fix ABI change in drm_mode_get_encoder
Jesse Barnes [Wed, 29 Oct 2008 19:17:39 +0000 (12:17 -0700)]
libdrm: fix ABI change in drm_mode_get_encoder

The encoder_type field should be at the top to match the kernel.

16 years agoradeon: reloc are backend dependant
Jerome Glisse [Mon, 27 Oct 2008 22:26:15 +0000 (23:26 +0100)]
radeon: reloc are backend dependant

16 years agoMerge branch 'modesetting-gem' of ssh://git.freedesktop.org/git/mesa/drm into modeset...
Jerome Glisse [Mon, 27 Oct 2008 18:33:56 +0000 (19:33 +0100)]
Merge branch 'modesetting-gem' of ssh://git.freedesktop.org/git/mesa/drm into modesetting-gem

16 years agoradeon: radeon util library
Jerome Glisse [Mon, 27 Oct 2008 18:27:15 +0000 (19:27 +0100)]
radeon: radeon util library

16 years agodrm: make handles 32-bits again not sure why they changed
Dave Airlie [Mon, 27 Oct 2008 07:06:23 +0000 (17:06 +1000)]
drm: make handles 32-bits again not sure why they changed

16 years agomode: Try to settle on a standard for struct fields
Jakob Bornecrantz [Fri, 24 Oct 2008 16:43:55 +0000 (18:43 +0200)]
mode: Try to settle on a standard for struct fields

16 years agomode: Indent defines
Jakob Bornecrantz [Fri, 24 Oct 2008 16:23:39 +0000 (18:23 +0200)]
mode: Indent defines

16 years agolink libdrm_intel properly
Dave Airlie [Thu, 16 Oct 2008 20:40:38 +0000 (06:40 +1000)]
link libdrm_intel properly

libdrm_intel needs symbols from libdrm, so link against it.
(cherry picked from commit d9c2f65dd8e50736a33e97a55c257ef6843e1ce7)

Conflicts:

libdrm/Makefile.am

16 years agolibdrm: don't depend or link to libdrm_intel
Dave Airlie [Sun, 12 Oct 2008 21:16:33 +0000 (07:16 +1000)]
libdrm: don't depend or link to libdrm_intel

16 years agointel: Protect bufmgr objects with a pthread mutex.
Eric Anholt [Fri, 13 Jun 2008 06:22:26 +0000 (23:22 -0700)]
intel: Protect bufmgr objects with a pthread mutex.

We want to be able to use the bufmgr from multiple threads for GL, and thus
we need to protect the internal structures.

The pthread-stubs package is used so that programs not linked against
pthreads get weak symbols to stubs and don't eat most of the cost.

16 years agoradeon: add initial agp support.
Dave Airlie [Thu, 16 Oct 2008 00:59:31 +0000 (10:59 +1000)]
radeon: add initial agp support.

This add agpmode command line option.

16 years agoradeon: add CS support for r100/r200 in 2D driver
Dave Airlie [Thu, 16 Oct 2008 00:57:31 +0000 (10:57 +1000)]
radeon: add CS support for r100/r200 in 2D driver

16 years agoradeon: fixup interrupt suspend/resume
Dave Airlie [Thu, 16 Oct 2008 00:55:24 +0000 (10:55 +1000)]
radeon: fixup interrupt suspend/resume

16 years agoradeon: fixup suspend/resume bus master enable
Dave Airlie [Thu, 16 Oct 2008 00:53:55 +0000 (10:53 +1000)]
radeon: fixup suspend/resume bus master enable

16 years agoradeon: re-enable hw blits for copying from VRAM
Dave Airlie [Thu, 16 Oct 2008 00:53:26 +0000 (10:53 +1000)]
radeon: re-enable hw blits for copying from VRAM

16 years agoradeon: fix buffer copying for VRAM->TT
Dave Airlie [Thu, 16 Oct 2008 00:52:53 +0000 (10:52 +1000)]
radeon: fix buffer copying for VRAM->TT

16 years agoradeon: move memcpy until after CP is stopped
Dave Airlie [Thu, 16 Oct 2008 00:52:28 +0000 (10:52 +1000)]
radeon: move memcpy until after CP is stopped

16 years agodrm: remove stray debug code
Dave Airlie [Thu, 16 Oct 2008 00:51:56 +0000 (10:51 +1000)]
drm: remove stray debug code

16 years agoradeon: use discardable flags on no backing store objects
Dave Airlie [Thu, 16 Oct 2008 00:51:31 +0000 (10:51 +1000)]
radeon: use discardable flags on no backing store objects

16 years agodrm: add discardable flag.
Dave Airlie [Thu, 16 Oct 2008 00:50:31 +0000 (10:50 +1000)]
drm: add discardable flag.

This discards memory contents on suspend/resume with the
hope the upper layers know something we don't.

16 years agodrm/radeon: initial suspend/resume fix.
Dave Airlie [Thu, 16 Oct 2008 00:49:58 +0000 (10:49 +1000)]
drm/radeon: initial suspend/resume fix.

This enables the evict code and also sets radeon up
to allow evict from VRAM to LOCAL

16 years agoradeon: pull in recent fixes from ddx
Alex Deucher [Tue, 7 Oct 2008 18:10:39 +0000 (14:10 -0400)]
radeon: pull in recent fixes from ddx

- fixup atom digital encoder setup
- pull in add get edid (currently disabled due to
  lack of support for atom fb/scratch space)

16 years agoradeon: fix alignment so Xv works again
Dave Airlie [Mon, 6 Oct 2008 06:40:45 +0000 (16:40 +1000)]
radeon: fix alignment so Xv works again

16 years agoradeon: fix pin ioctl interface to mesa can find offset for pinned buffers
Dave Airlie [Mon, 6 Oct 2008 06:40:20 +0000 (16:40 +1000)]
radeon: fix pin ioctl interface to mesa can find offset for pinned buffers

16 years agodrm/radeon: fixup clean flag handling
Dave Airlie [Mon, 6 Oct 2008 06:39:25 +0000 (16:39 +1000)]
drm/radeon: fixup clean flag handling

16 years agointel bufmgr: reinstate buffer handle tracking
Jesse Barnes [Tue, 30 Sep 2008 23:35:26 +0000 (16:35 -0700)]
intel bufmgr: reinstate buffer handle tracking

We need a way of getting at the underlying handle for use with mode
setting.  We can either export it in the dri_bo object or provide a new
callback to get it.

16 years agoMerge remote branch 'origin/modesetting-gem' into modesetting-gem
Dave Airlie [Tue, 30 Sep 2008 04:17:27 +0000 (14:17 +1000)]
Merge remote branch 'origin/modesetting-gem' into modesetting-gem

16 years agoMerge remote branch 'origin/master' into modesetting-gem
Dave Airlie [Fri, 26 Sep 2008 05:37:21 +0000 (15:37 +1000)]
Merge remote branch 'origin/master' into modesetting-gem

Conflicts:

libdrm/Makefile.am
libdrm/dri_bufmgr.h
linux-core/drm_irq.c
linux-core/drm_sysfs.c
linux-core/drm_ttm.c
shared-core/i915_dma.c
shared-core/i915_irq.c
shared-core/nouveau_drv.h
shared-core/radeon_cp.c

16 years agoradeon: use atom for ext tmds on r4xx
Alex Deucher [Fri, 26 Sep 2008 21:32:15 +0000 (17:32 -0400)]
radeon: use atom for ext tmds on r4xx

16 years agoradeon: make atom on r4xx a module option
Alex Deucher [Fri, 26 Sep 2008 21:20:04 +0000 (17:20 -0400)]
radeon: make atom on r4xx a module option

default is legacy modesetting.  pass module option r4xx_atom
to try using atom on r4xx.

16 years agoMerge branch 'modesetting-gem' of git+ssh://agd5f@git.freedesktop.org/git/mesa/drm...
Alex Deucher [Thu, 25 Sep 2008 22:47:47 +0000 (18:47 -0400)]
Merge branch 'modesetting-gem' of git+ssh://agd5f@git.freedesktop.org/git/mesa/drm into modesetting-gem

16 years agoradeon: first pass at using atombios on r4xx hw
Alex Deucher [Thu, 25 Sep 2008 22:45:07 +0000 (18:45 -0400)]
radeon: first pass at using atombios on r4xx hw

16 years agoSeperate modesetting userspace bits into drm_mode.h
Jakob Bornecrantz [Thu, 25 Sep 2008 21:12:07 +0000 (23:12 +0200)]
Seperate modesetting userspace bits into drm_mode.h

16 years agoradeon: add r600 modesetting registers writes
Dave Airlie [Wed, 24 Sep 2008 05:25:35 +0000 (15:25 +1000)]
radeon: add r600 modesetting registers writes

16 years agoradeon: add parsing for r6xx object tables
Dave Airlie [Wed, 24 Sep 2008 03:58:26 +0000 (13:58 +1000)]
radeon: add parsing for r6xx object tables

16 years agointel: Allow up to 15 seconds chewing on one buffer before acknowledging -EBUSY.
Eric Anholt [Wed, 24 Sep 2008 00:06:01 +0000 (17:06 -0700)]
intel: Allow up to 15 seconds chewing on one buffer before acknowledging -EBUSY.

The gltestperf demo in some cases took over seven seconds to make it through
one batchbuffer on a GM965.

Bug #17004.

16 years agointel: Replace wraparound test logic in bufmgr_fake. Again.
Eric Anholt [Tue, 23 Sep 2008 17:48:39 +0000 (10:48 -0700)]
intel: Replace wraparound test logic in bufmgr_fake.  Again.

I'd swapped the operands, so if we weren't in lockstep with the hardware we
said the sequence was always passed.  Additionally, a race was available that
we might have failed at recovering from.  Instead, I've replaced the logic
with new stuff that should be more robust and not rely on all the parties in
userland following the same IRQ_EMIT() == 1 protocol.  Also, in a radical
departure from past efforts, include a long comment describing the failure
modes and how we're working around them.

Thanks to haihao for catching the original issue.

16 years agointel: Do strerror on errno, not on the -1 return value from ioctl.
Eric Anholt [Tue, 23 Sep 2008 17:47:21 +0000 (10:47 -0700)]
intel: Do strerror on errno, not on the -1 return value from ioctl.

16 years agoUpdate intel modesetting to use mm_private instead of mm_handle.
Kristian Høgsberg [Tue, 23 Sep 2008 06:52:06 +0000 (16:52 +1000)]
Update intel modesetting to use mm_private instead of mm_handle.

16 years agoradeon: fix minor cursor issues
Dave Airlie [Tue, 23 Sep 2008 06:50:39 +0000 (16:50 +1000)]
radeon: fix minor cursor issues

16 years agoradeon: Fix type in check for tmds type.
Kristian Høgsberg [Tue, 23 Sep 2008 06:50:22 +0000 (16:50 +1000)]
radeon: Fix type in check for tmds type.

16 years agoStore the buffer object backing the fb as a void pointer, not a handle.
Kristian Høgsberg [Tue, 23 Sep 2008 06:47:34 +0000 (16:47 +1000)]
Store the buffer object backing the fb as a void pointer, not a handle.

This lets us defer handle creation until userspace acutally asks for
one, at which point we also have a drm_file to associate it with.

16 years agodrm: fix brace placement
Dave Airlie [Thu, 28 Aug 2008 11:20:19 +0000 (21:20 +1000)]
drm: fix brace placement

16 years agointel: Fix driver-supplied argument to exec function (fd.o bug #17653).
Xiang, Haihao [Mon, 22 Sep 2008 02:16:19 +0000 (10:16 +0800)]
intel: Fix driver-supplied argument to exec function (fd.o bug #17653).

16 years agoradeon: rmx_fixup() fixes for legacy chips
Alex Deucher [Fri, 19 Sep 2008 15:19:00 +0000 (11:19 -0400)]
radeon: rmx_fixup() fixes for legacy chips

16 years agoradeon: fix voverplus calculation
Dave Airlie [Fri, 19 Sep 2008 14:48:11 +0000 (00:48 +1000)]
radeon: fix voverplus calculation

16 years agoradeon: tmds bracket failure found by krh.
Dave Airlie [Thu, 18 Sep 2008 22:03:34 +0000 (08:03 +1000)]
radeon: tmds bracket failure found by krh.

16 years agoradeon: pll and interlace updates from the ddx
Alex Deucher [Thu, 18 Sep 2008 21:27:00 +0000 (17:27 -0400)]
radeon: pll and interlace updates from the ddx

also some formatting cleanup in radeon_reg.h

16 years agoradeon: Add functions to set mem/eng clocks
Alex Deucher [Thu, 18 Sep 2008 20:42:22 +0000 (16:42 -0400)]
radeon: Add functions to set mem/eng clocks

16 years agoradeon: add function to configure PCIE lanes
Alex Deucher [Thu, 18 Sep 2008 20:07:41 +0000 (16:07 -0400)]
radeon: add function to configure PCIE lanes

16 years agoRadeon: restructure PLL data
Alex Deucher [Thu, 18 Sep 2008 19:11:48 +0000 (15:11 -0400)]
Radeon: restructure PLL data

- store pixel clocks, core clock, and memory clocks separately
- grab all pll limits from bios tables

16 years agoExport drm_put_minor
Alex Deucher [Thu, 18 Sep 2008 18:30:05 +0000 (14:30 -0400)]
Export drm_put_minor

16 years ago[FreeBSD] Convert to using cdevpriv for file_priv tracking
Robert Noland [Thu, 18 Sep 2008 03:10:15 +0000 (23:10 -0400)]
[FreeBSD] Convert to using cdevpriv for file_priv tracking

16 years agofixup radeon stuff - need to checkout irqs
Dave Airlie [Thu, 18 Sep 2008 00:28:42 +0000 (10:28 +1000)]
fixup radeon stuff - need to checkout irqs

16 years agoradeon: fixup buffer and cs bits
Dave Airlie [Thu, 18 Sep 2008 00:22:23 +0000 (10:22 +1000)]
radeon: fixup buffer and cs bits

16 years ago[PATCH] radeon: fixup GEM domain setting - allows more userspace paths
Dave Airlie [Thu, 18 Sep 2008 00:20:29 +0000 (10:20 +1000)]
[PATCH] radeon: fixup GEM domain setting - allows more userspace paths

also dirty buffer on validate

16 years ago[PATCH] radeon: disable blit moves
Dave Airlie [Thu, 18 Sep 2008 00:19:53 +0000 (10:19 +1000)]
[PATCH] radeon: disable blit moves

16 years agoradeon: fail properly if we can't create the ring.
Dave Airlie [Thu, 18 Sep 2008 00:19:08 +0000 (10:19 +1000)]
radeon: fail properly if we can't create the ring.

Normally this will be due to an AGP driver needing updating

16 years agoradeon: do proper memory controller init and setup
Dave Airlie [Thu, 18 Sep 2008 00:17:27 +0000 (10:17 +1000)]
radeon: do proper memory controller init and setup

16 years agoradeon: fix return value
Dave Airlie [Thu, 18 Sep 2008 00:16:41 +0000 (10:16 +1000)]
radeon: fix return value

16 years agoradeon: fixup reference counting properly
Dave Airlie [Thu, 18 Sep 2008 00:16:09 +0000 (10:16 +1000)]
radeon: fixup reference counting properly

16 years agoradeon: sort out atom vs combios tables for r400 cards
Dave Airlie [Thu, 18 Sep 2008 00:15:43 +0000 (10:15 +1000)]
radeon: sort out atom vs combios tables for r400 cards

16 years agoradeon: remove unneeded debugging
Dave Airlie [Thu, 18 Sep 2008 00:14:32 +0000 (10:14 +1000)]
radeon: remove unneeded debugging

16 years agoradeon: don't do full edid for detection purposes
Dave Airlie [Thu, 18 Sep 2008 00:13:30 +0000 (10:13 +1000)]
radeon: don't do full edid for detection purposes

16 years agomake text reserve 256k
Dave Airlie [Thu, 18 Sep 2008 00:11:43 +0000 (10:11 +1000)]
make text reserve 256k

16 years agoradeon: add initial suspend/resume support
Dave Airlie [Thu, 18 Sep 2008 00:05:59 +0000 (10:05 +1000)]
radeon: add initial suspend/resume support

plus a bunch of fixes

16 years agomodesetting: Add helper to force restore modes on crtcs at resume time
Dave Airlie [Wed, 17 Sep 2008 23:59:08 +0000 (09:59 +1000)]
modesetting: Add helper to force restore modes on crtcs at resume time

16 years agoradeon: fix legacy LVDS
Alex Deucher [Wed, 17 Sep 2008 22:16:25 +0000 (18:16 -0400)]
radeon: fix legacy LVDS

16 years agonv50: add initial context for chipset 0xaa
Ben Skeggs [Wed, 17 Sep 2008 12:18:03 +0000 (22:18 +1000)]
nv50: add initial context for chipset 0xaa

This just doesn't look right..

16 years agonv50: add initial context to match ctxprog for chipset 0x50
Ben Skeggs [Wed, 17 Sep 2008 12:03:38 +0000 (22:03 +1000)]
nv50: add initial context to match ctxprog for chipset 0x50

16 years agonv50: add ctxprog for chipset 0x50
Ben Skeggs [Wed, 17 Sep 2008 05:13:27 +0000 (15:13 +1000)]
nv50: add ctxprog for chipset 0x50

16 years agonv50: add ctxprog for chipset 0xaa
Ben Skeggs [Wed, 17 Sep 2008 05:02:54 +0000 (15:02 +1000)]
nv50: add ctxprog for chipset 0xaa

16 years agonv50: add support for chipset 0x92
Ben Skeggs [Wed, 17 Sep 2008 04:49:04 +0000 (14:49 +1000)]
nv50: add support for chipset 0x92

16 years agoRemove incomplete and obsolete free/net/open code.
vehemens [Tue, 9 Sep 2008 05:06:09 +0000 (22:06 -0700)]
Remove incomplete and obsolete free/net/open code.

Signed-off-by: Robert Noland <rnoland@2hip.net>
16 years agoAdd missing \ to libdrm_la_SOURCES
Eric Anholt [Thu, 11 Sep 2008 03:04:42 +0000 (20:04 -0700)]
Add missing \ to libdrm_la_SOURCES

Reported by jcristau.

16 years agointel: don't forget to include config.h in bufmgr code.
Eric Anholt [Wed, 10 Sep 2008 20:54:34 +0000 (13:54 -0700)]
intel: don't forget to include config.h in bufmgr code.

Thanks to airlied for catching this.

16 years agointel: move drm calls to exec buffers to libdrm_intel.
Eric Anholt [Mon, 8 Sep 2008 15:51:40 +0000 (08:51 -0700)]
intel: move drm calls to exec buffers to libdrm_intel.

This avoids duplicating the effort in 3 places.  Also, added emit/wait fence
callbacks back in bufmgr_fake since we need it for non-drm 2d.  Sigh.

16 years agointel: Move IRQ emit/wait from callbacks into the bufmgr.
Eric Anholt [Sat, 6 Sep 2008 02:07:41 +0000 (03:07 +0100)]
intel: Move IRQ emit/wait from callbacks into the bufmgr.

In the process, work around the glaring bugs of the kernel irq wait function.

16 years agoMove intel libdrm stuff to libdrm_intel.so
Eric Anholt [Fri, 5 Sep 2008 09:35:32 +0000 (10:35 +0100)]
Move intel libdrm stuff to libdrm_intel.so

dri_bufmgr.h is replaced by intel_bufmgr.h, and several functions are renamed,
though the structures and many functions remain dri_bufmgr_* and dri_bo_*

16 years agodrm: Add tests for GEM_FLINK ioctl.
Eric Anholt [Wed, 3 Sep 2008 22:05:16 +0000 (23:05 +0100)]
drm: Add tests for GEM_FLINK ioctl.

16 years ago[FreeBSD] We need to call drm_detach before we free dev->driver.
Robert Noland [Mon, 8 Sep 2008 20:40:52 +0000 (16:40 -0400)]
[FreeBSD] We need to call drm_detach before we free dev->driver.

The driver is in control of the show, so when you try and unload a module
the driver detach routine is called first.  It is what drives the whole
unload process and so lots of panics occur if dev->driver is already
free.

16 years ago[FreeBSD] Implement drm_ioremap_wc() to set a range of mem to write-combining
Robert Noland [Sun, 7 Sep 2008 16:44:02 +0000 (12:44 -0400)]
[FreeBSD] Implement drm_ioremap_wc() to set a range of mem to write-combining

16 years ago[FreeBSD] IGP gart needs to be un-cached.
Robert Noland [Sun, 7 Sep 2008 01:08:33 +0000 (21:08 -0400)]
[FreeBSD] IGP gart needs to be un-cached.

Airlied inadvertently discovered that the IGP gart needs to be un-cached
for radeon rs485 and rs690 to work.  Initial tests by placing a wbinvd()
after allocating the gart were successful.  This is an attempt at a more
appropriate method of achieving success.

16 years agoFree temp_pagelist on error. Free in reverse order. Noticed by open.
vehemens [Sun, 7 Sep 2008 01:02:31 +0000 (18:02 -0700)]
Free temp_pagelist on error. Free in reverse order. Noticed by open.

Signed-off-by: Robert Noland <rnoland@2hip.net>
16 years agoPass lock data like linux and open.
vehemens [Sat, 6 Sep 2008 22:33:47 +0000 (15:33 -0700)]
Pass lock data like linux and open.

Signed-off-by: Robert Noland <rnoland@2hip.net>
16 years agoMove order to end like linux.
vehemens [Sat, 6 Sep 2008 21:19:32 +0000 (14:19 -0700)]
Move order to end like linux.

Signed-off-by: Robert Noland <rnoland@2hip.net>
16 years agoStyle white space cleanup part 2.
vehemens [Sat, 6 Sep 2008 21:16:51 +0000 (14:16 -0700)]
Style white space cleanup part 2.

Signed-off-by: Robert Noland <rnoland@2hip.net>
16 years ago[FreeBSD] Ensure that drm_pci_alloc is never called while locks are held.
Robert Noland [Sat, 6 Sep 2008 22:37:06 +0000 (18:37 -0400)]
[FreeBSD] Ensure that drm_pci_alloc is never called while locks are held.

16 years agoradeon: change interface from headers add pin_domain into padding
Dave Airlie [Sat, 6 Sep 2008 22:09:24 +0000 (08:09 +1000)]
radeon: change interface from headers add pin_domain into padding

16 years agoReorder lock functions like linux.
vehemens [Tue, 2 Sep 2008 20:56:35 +0000 (13:56 -0700)]
Reorder lock functions like linux.

Signed-off-by: Robert Noland <rnoland@2hip.net>
16 years agoStyle white space cleanup.
vehemens [Tue, 2 Sep 2008 11:57:36 +0000 (04:57 -0700)]
Style white space cleanup.

Signed-off-by: Robert Noland <rnoland@2hip.net>
16 years agoNeed M_NOWAIT for malloc.
vehemens [Tue, 2 Sep 2008 09:43:19 +0000 (02:43 -0700)]
Need M_NOWAIT for malloc.

Signed-off-by: Robert Noland <rnoland@2hip.net>
16 years agoi915: fix i915_ring_validate()
Robert Noland [Fri, 5 Sep 2008 15:52:53 +0000 (11:52 -0400)]
i915: fix i915_ring_validate()

16 years agoradeon: rs690 GART tables need to be in uncached memory.
Dave Airlie [Fri, 5 Sep 2008 00:56:18 +0000 (10:56 +1000)]
radeon: rs690 GART tables need to be in uncached memory.

Allocate the rs480/690 tables from uncached memory.

16 years agonouveau: 8200 cards are 0xA0 family.
Stephane Marchesin [Thu, 4 Sep 2008 22:17:52 +0000 (00:17 +0200)]
nouveau: 8200 cards are 0xA0 family.

16 years agoradeon: fixup a number of avivo checks for rs690
Dave Airlie [Thu, 4 Sep 2008 01:57:00 +0000 (11:57 +1000)]
radeon: fixup a number of avivo checks for rs690

16 years agodrm: fix sysfs error path.
Dave Airlie [Tue, 2 Sep 2008 00:07:14 +0000 (10:07 +1000)]
drm: fix sysfs error path.

Pointed out by Roel Kluin on dri-devel.

Signed-off-by: Dave Airlie <airlied@redhat.com>