linux-stable.git/drivers/gpu, branch v4.1.4

drm: Stop resetting connector state to unknown

2015-08-03T16:29:10+00:00

commit 5677d67ae3949f09f57357241b88222d49b8c782 upstream.

It's causing piles of issues since we've stopped forcing full detect
cycles in the sysfs interfaces with

commit c484f02d0f02fbbfc6decc945a69aae011041a27
Author: Chris Wilson 
Date:   Fri Mar 6 12:36:42 2015 +0000

    drm: Lighten sysfs connector 'status'

The original justification for this was that the hpd handlers could
use the unknown state as a hint to force a full detection. But current
i915 code isn't doing that any more, and no one else really uses reset
on resume. So instead just keep the old state around.

References: http://article.gmane.org/gmane.comp.freedesktop.xorg.drivers.intel/62584
Bugzilla: https://bugzilla.kernel.org/show_bug.cgi?id=100641
Cc: Rui Matos 
Cc: Julien Wajsberg 
Cc: kuddel.mail@gmx.de
Cc: Lennart Poettering 
Acked-by: Rob Clark 
Tested-by: Rui Tiago Cação Matos 
Signed-off-by: Daniel Vetter 
Signed-off-by: Greg Kroah-Hartman

drm: Provide compat ioctl for addfb2.1

2015-08-03T16:29:10+00:00

commit c631d5f90e7ee246536c72f80ade86e9ef4d2f13 upstream.

Frame buffer modifiers extensions provided in;

  commit e3eb3250d84ef97b766312345774367b6a310db8
  Author: Rob Clark 
  Date:   Thu Feb 5 14:41:52 2015 +0000

      drm: add support for tiled/compressed/etc modifier in addfb2

Missed the structure packing/alignment problem where 64-bit
members were added after the odd number of 32-bit ones. This
makes the compiler produce structures of different sizes under
32- and 64-bit x86 targets and makes the ioctl need explicit
compat handling.

v2: Removed the typedef. (Daniel Vetter)

Signed-off-by: Tvrtko Ursulin 
Reviewed-by: Daniel Vetter 
Cc: dri-devel@lists.freedesktop.org
Cc: Rob Clark 
Cc: Daniel Stone 
Cc: Daniel Vetter 
[danvet: Squash in compile fix from Mika.]
Signed-off-by: Daniel Vetter 
Signed-off-by: Greg Kroah-Hartman

drm: add a check for x/y in drm_mode_setcrtc

2015-08-03T16:29:10+00:00

commit 01447e9f04ba1c49a9534ae6a5a6f26c2bb05226 upstream.

legacy setcrtc ioctl does take a 32 bit value which might indeed
overflow

the checks of crtc_req->x > INT_MAX and crtc_req->y > INT_MAX aren't
needed any more with this

v2: -polish the annotation according to Daniel's comment

Cc: Daniel Vetter 
Signed-off-by: Zhao Junwang 
Signed-off-by: Daniel Vetter 
Signed-off-by: Greg Kroah-Hartman

drm/rockchip: use drm_gem_mmap helpers

2015-08-03T16:29:10+00:00

commit 41315b793e13f884cda79389f0d5d44d027e57d1 upstream.

Rather than (incompletely [0]) re-implementing drm_gem_mmap() and
drm_gem_mmap_obj() helpers, call them directly from the rockchip mmap
routines.

Once the core functions return successfully, the rockchip mmap routines
can still use dma_mmap_attrs() to simply mmap the entire buffer.

[0] Previously, we were performing the mmap() without first taking a
reference on the underlying gem buffer.  This could leak ptes if the gem
object is destroyed while userspace is still holding the mapping.

Signed-off-by: Daniel Kurtz 
Reviewed-by: Daniel Vetter 
Signed-off-by: Daniel Vetter 
Signed-off-by: Greg Kroah-Hartman

drm/radeon/ci: silence a harmless PCC warning

2015-08-03T16:29:10+00:00

commit bda5e3e97ffe80c5a793383df5681d3581d46ac8 upstream.

This has been a source of confusion.  Make it debug only.

Signed-off-by: Alex Deucher 
Signed-off-by: Greg Kroah-Hartman

drm/radeon: fix user ptr race condition

2015-08-03T16:29:10+00:00

commit 12f1384da650bdb835fff63e66fe815ea882fc0e upstream.

Port of amdgpu patch 9298e52f8b51d1e4acd68f502832f3a97f8cf892.

Signed-off-by: Christian König 
Signed-off-by: Alex Deucher 
Signed-off-by: Greg Kroah-Hartman

drm/radeon: add a dpm quirk for Sapphire Radeon R9 270X 2GB GDDR5

2015-08-03T16:29:10+00:00

commit 5dfc71bc44d91d1620505c064fa22b0b3db58a9d upstream.

bug:
https://bugs.freedesktop.org/show_bug.cgi?id=76490

Signed-off-by: Alex Deucher 
Signed-off-by: Greg Kroah-Hartman

drm/radeon: Don't flush the GART TLB if rdev->gart.ptr == NULL

2015-08-03T16:29:10+00:00

commit 233709d2cd6bbaaeda0aeb8d11f6ca7f98563b39 upstream.

This can be the case when the GPU is powered off, e.g. via vgaswitcheroo
or runpm. When the GPU is powered up again, radeon_gart_table_vram_pin
flushes the TLB after setting rdev->gart.ptr to non-NULL.

Fixes panic on powering off R7xx GPUs.

Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=61529
Reviewed-by: Christian König 
Signed-off-by: Michel Dänzer 
Signed-off-by: Alex Deucher 
Signed-off-by: Greg Kroah-Hartman

drm/radeon: unpin cursor BOs on suspend and pin them again on resume (v2)

2015-08-03T16:29:10+00:00

commit f3cbb17bcf676a2fc6aedebe9fbebd59e550c51a upstream.

Everything is evicted from VRAM before suspend, so we need to make
sure all BOs are unpinned and re-pinned after resume. Fixes broken
mouse cursor after resume introduced by commit b9729b17.

[Michel Dänzer: Add pinning BOs on resume]

v2:
[Alex Deucher: merge cursor unpin into fb unpin loop]

Bugzilla: https://bugzilla.kernel.org/show_bug.cgi?id=100541
Reviewed-by: Christian König  (v1)
Signed-off-by: Grigori Goronzy 
Signed-off-by: Michel Dänzer 
Signed-off-by: Alex Deucher 
Signed-off-by: Greg Kroah-Hartman

drm/radeon: Clean up reference counting and pinning of the cursor BOs

2015-08-03T16:29:10+00:00

commit cd404af0c930104462aa91344f07d002cf8248ed upstream.

Take a GEM reference for and pin the new cursor BO, unpin and drop the
GEM reference for the old cursor BO in radeon_crtc_cursor_set2, and use
radeon_crtc->cursor_addr in radeon_set_cursor.

This fixes radeon_cursor_reset accidentally incrementing the cursor BO
pin count, and cleans up the code a little.

Reviewed-by: Grigori Goronzy 
Reviewed-by: Christian König 
Signed-off-by: Michel Dänzer 
Signed-off-by: Alex Deucher 
Signed-off-by: Greg Kroah-Hartman