linux-stable.git/drivers/gpu/drm/msm, branch v3.16.40

drm/msm: protect against faults from copy_from_user() in submit ioctl

2016-11-20T01:17:19+00:00

commit d78d383ab354b0b9e1d23404ae0d9fbdeb9aa035 upstream.

An evil userspace could try to cause deadlock by passing an unfaulted-in
GEM bo as submit->bos (or submit->cmds) table.  Which will trigger
msm_gem_fault() while we already hold struct_mutex.  See:

https://github.com/freedreno/msmtest/blob/master/evilsubmittest.c

Signed-off-by: Rob Clark 
[bwh: Backported to 3.16: adjust context]
Signed-off-by: Ben Hutchings

drm/msm: fix use of copy_from_user() while holding spinlock

2016-11-20T01:17:19+00:00

commit 89f82cbb0d5c0ab768c8d02914188aa2211cd2e3 upstream.

Use instead __copy_from_user_inatomic() and fallback to slow-path where
we drop and re-aquire the lock in case of fault.

Reported-by: Vaishali Thakkar 
Signed-off-by: Rob Clark 
Signed-off-by: Ben Hutchings

drm/msm: use mutex_lock_interruptible for submit ioctl

2016-11-20T01:17:19+00:00

commit b5b4c264df4d270819676b290cef9a11d04c35f0 upstream.

Be kinder to things that do lots of signal handling (ie. Xorg)

Signed-off-by: Rob Clark 
[bwh: Backported to 3.16: adjust context]
Signed-off-by: Ben Hutchings

drm/msm: fix IOMMU cleanup for -EPROBE_DEFER

2014-06-22T12:32:10+00:00

If probe fails after IOMMU is attached, we need to detach in order to
clean up properly.  Before this change, IOMMU faults would occur if the
probe failed (-EPROBE_DEFER).

Signed-off-by: Stephane Viau 
Signed-off-by: Rob Clark

drm/msm: use PAGE_ALIGNED instead of IS_ALIGNED(PAGE_SIZE)

2014-06-22T12:32:10+00:00

use mm.h definition

Cc: David Airlie 
Cc: Rob Clark 
Signed-off-by: Fabian Frederick 
Signed-off-by: Rob Clark

drm/msm/hdmi: set hdp clock rate before prepare_enable

2014-06-22T12:32:09+00:00

The clock driver usually complains when a clock is being prepared
before setting its rate. It is the case here for "core_clk" which
needs to be set at 19.2 MHz before we attempt a prepare_enable().

Signed-off-by: Stephane Viau 
Signed-off-by: Rob Clark

drm/msm: storage class should be before const qualifier

2014-06-22T12:32:09+00:00

The C99 specification states in section 6.11.5:

The placement of a storage-class specifier other than at the beginning
of the declaration specifiers in a declaration is an obsolescent
feature.

Signed-off-by: Peter Griffin 
Cc: David Airlie 
Cc: dri-devel@lists.freedesktop.org
Cc: linux-kernel@vger.kernel.org
Signed-off-by: Rob Clark

drm/msm: Replace type of paddr to uint32_t.

2014-06-22T12:32:09+00:00

This patch helps to avoid the following build issue:

drivers/gpu/drm/msm/msm_fbdev.c:108:2: error: passing argument 3 of 'msm_gem_get_iova_locked' from incompatible pointer type [-Werror]
   msm_gem_get_iova_locked(fbdev->bo, 0, &paddr);
   ^
In file included from drivers/gpu/drm/msm/msm_fbdev.c:18:0:
drivers/gpu/drm/msm/msm_drv.h:153:5: note: expected 'uint32_t *' but argument is of type 'dma_addr_t *'
  int msm_gem_get_iova_locked(struct drm_gem_object *obj, int id,
      ^

Signed-off-by: Matwey V. Kornilov 
Signed-off-by: Rob Clark

drm: add drm_fb_helper_restore_fbdev_mode_unlocked()

2014-06-05T00:02:40+00:00

All drm_fb_helper_restore_fbdev_mode() call sites, save one, do the same
locking.  Simplify this into drm_fb_helper_restore_fbdev_mode_unlocked().

Signed-off-by: Rob Clark 
Reviewed-by: Daniel Vetter 
Signed-off-by: Dave Airlie

drm/msm: use correct gfp flag for vram allocation

2014-06-02T11:36:31+00:00

We want at least __GFP_WAIT, otherwise dma-mapping tries to use coherent
pool rather than CMA pool.

Signed-off-by: Rob Clark