Theodore Ts'o [Tue, 20 Dec 2011 22:06:08 +0000 (17:06 -0500)]
ext4: fix potential deadlock with setuid files and EXT4_IOC_MOVE_EXT
file_remove_suid() must be called with i_mutex down, since it calls
notify_change(). In addition, we really want to remove the suid file
*before* we modify the donor file, to avoid someone from trying to
exploit a race.
Ville Syrjälä [Mon, 19 Dec 2011 22:33:24 +0000 (00:33 +0200)]
drm: Add drm_format_num_planes() utility function
This function returns the number of planes used by a specific pixel
format.
Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com> Reviewed-by: Rob Clark <rob.clark@linaro.org> Signed-off-by: Dave Airlie <airlied@redhat.com>
Rob Clark [Thu, 15 Dec 2011 20:53:24 +0000 (14:53 -0600)]
drm: call connector dpms fxn, when setting config
Call connector->funcs->dpms(DPMS_ON) rather than just setting
connector->dpms = DPMS_ON. This ensures that if the connector
has something to do to enable the output (rather than just using
drm_helper_connector_dpms helper directly), that this happens
at bootup. This solves an issue with connectors not getting
enabled from fbcon_init() when the driver is loaded.
Signed-off-by: Rob Clark <rob@ti.com> Reviewed-by: Adam Jackson <ajax@redhat.com> Signed-off-by: Dave Airlie <airlied@redhat.com>
Dave Airlie [Tue, 20 Dec 2011 11:44:30 +0000 (11:44 +0000)]
drm/radeon/kms: don't use 0 bpc for adjusting hdmi clock
If the bpc is set from the connector is 0, we then use it later to adjust
in a special case the HDMI pixel clock, however if the bpc is 0, we end up
passing a 0 pixel clock into the code.
I'm not sure if this is the correct answer or if we should avoid the HDMI
clock adjustment for 0 values.
This fixes a divide by 0 on my Llano system with a HDMI monitor and hdmi
audio enabled.
Reviewed-by: Alex Deucher <alexdeucher@gmail.com> Signed-off-by: Dave Airlie <airlied@redhat.com>
Dave Airlie [Tue, 20 Dec 2011 19:53:44 +0000 (19:53 +0000)]
Merge branch 'drm-radeon-next' of ../drm-radeon-next into drm-core-next
* 'drm-radeon-next' of ../drm-radeon-next:
drm/radeon: introduce a sub allocator and convert ib pool to it v4
drm/radeon/kms: add support for per-ring fence interrupts
drm/radeon/kms: add cayman specific fence_ring_emit
drm/radeon/kms: add some new ring params to better handle other ring types
drm/radeon: improve radeon_test_syncing function
drm/radeon: precompute fence cpu/gpu addr once v3
drm/radeon: move ring debugfs into radeon_ring.c
drm/radeon: rename struct radeon_cp to radeon_ring
drm/radeon: disable compute rings on cayman for now
drm/radeon: add radeon_fence_count_emited function
drm/radeon: make some asic pointers per ring
drm/radeon: Add radeon_test_syncing function v2
drm/radeon: make cp variable an array
drm/radeon: make ring rptr and wptr register offsets variable
drm/radeon: make all functions work with multiple rings.
drm/radeon/kms: add support for semaphores v3
drm/radeon/kms: add support for multiple fence queues v2
drm/radeon: fix a spelling mistake
drm/radeon: no need to check all relocs for duplicates
drm/radeon: fix debugfs handling v3
Jerome Glisse [Tue, 15 Nov 2011 16:48:34 +0000 (11:48 -0500)]
drm/radeon: introduce a sub allocator and convert ib pool to it v4
Somewhat specializaed sub-allocator designed to perform sub-allocation
for command buffer not only for current cs ioctl but for future command
submission ioctl as well. Patch also convert current ib pool to use
the sub allocator. Idea is that ib poll buffer can be share with other
command buffer submission not having 64K granularity.
v2 Harmonize pool handling and add suspend/resume callback to pin/unpin
sa bo (tested on rv280, rv370, r420, rv515, rv610, rv710, redwood, cayman,
rs480, rs690, rs880)
v3 Simplify allocator
v4 Fix radeon_ib_get error path to properly free fence
Signed-off-by: Jerome Glisse <jglisse@redhat.com> Signed-off-by: Dave Airlie <airlied@redhat.com>
Alex Deucher [Fri, 18 Nov 2011 01:13:28 +0000 (20:13 -0500)]
drm/radeon/kms: add support for per-ring fence interrupts
Signed-off-by: Alex Deucher <alexander.deucher@amd.com> Signed-off-by: Christian König <deathsimple@vodafone.de> Signed-off-by: Dave Airlie <airlied@redhat.com>
Alex Deucher [Thu, 17 Nov 2011 19:57:50 +0000 (14:57 -0500)]
drm/radeon/kms: add cayman specific fence_ring_emit
cayman is wb only and doesn't have a VC.
Signed-off-by: Alex Deucher <alexander.deucher@amd.com> Signed-off-by: Christian König <deathsimple@vodafone.de> Signed-off-by: Dave Airlie <airlied@redhat.com>
Alex Deucher [Thu, 17 Nov 2011 19:25:56 +0000 (14:25 -0500)]
drm/radeon/kms: add some new ring params to better handle other ring types
Some rptr/wptrs fields have different offsets and not all rings are pm4
so add a new nop field.
Signed-off-by: Alex Deucher <alexander.deucher@amd.com> Signed-off-by: Christian König <deathsimple@vodafone.de> Signed-off-by: Dave Airlie <airlied@redhat.com>
Christian König [Thu, 17 Nov 2011 14:22:44 +0000 (15:22 +0100)]
drm/radeon: improve radeon_test_syncing function
Also test multiple waits on the same semaphore.
Signed-off-by: Christian König <deathsimple@vodafone.de> Reviewed-by: Jerome Glisse <jglisse@redhat.com> Signed-off-by: Dave Airlie <airlied@redhat.com>
Jerome Glisse [Sun, 20 Nov 2011 20:45:34 +0000 (20:45 +0000)]
drm/radeon: precompute fence cpu/gpu addr once v3
Add a start fence driver helper function which will be call
once for each ring and will compute cpu/gpu addr for fence
depending on wether to use wb buffer or scratch reg.
This patch replace initialize fence driver separately which
was broken in regard of GPU lockup. The fence list for created,
emited, signaled must be initialize once and only from the
asic init callback not from the startup call back which is
call from the gpu reset.
v2: With this in place we no longer need to know the number of
rings in fence_driver_init, also writing to the scratch reg
before knowing its offset is a bad idea.
v3: rebase on top of change to previous patch in the serie
Signed-off-by: Christian König <deathsimple@vodafone.de> Signed-off-by: Jerome Glisse <jglisse@redhat.com> Signed-off-by: Dave Airlie <airlied@redhat.com>
Christian König [Mon, 24 Oct 2011 15:08:44 +0000 (17:08 +0200)]
drm/radeon: move ring debugfs into radeon_ring.c
Those debugfs files aren't r600 specific, so they
shouldn't be in r600.c. Move them to radeon_ring.c
and also add functionality to dump CP1 & CP2 ring
informations.
Signed-off-by: Christian König <deathsimple@vodafone.de> Reviewed-by: Jerome Glisse <jglisse@redhat.com> Signed-off-by: Dave Airlie <airlied@redhat.com>
Christian König [Sun, 23 Oct 2011 10:56:27 +0000 (12:56 +0200)]
drm/radeon: rename struct radeon_cp to radeon_ring
That naming seems to make more sense, since we not
only want to run PM4 rings with it.
Signed-off-by: Christian König <deathsimple@vodafone.de> Reviewed-by: Jerome Glisse <jglisse@redhat.com> Signed-off-by: Dave Airlie <airlied@redhat.com>
Christian König [Thu, 20 Oct 2011 10:47:13 +0000 (12:47 +0200)]
drm/radeon: disable compute rings on cayman for now
Disable the additional compute rings on cayman
until their setup is fully implemented.
Signed-off-by: Christian König <deathsimple@vodafone.de> Reviewed-by: Jerome Glisse <jglisse@redhat.com> Signed-off-by: Dave Airlie <airlied@redhat.com>
Christian König [Thu, 20 Oct 2011 10:38:09 +0000 (12:38 +0200)]
drm/radeon: add radeon_fence_count_emited function
Split counting of emited fences out of power
management into a seperate function.
Signed-off-by: Christian König <deathsimple@vodafone.de> Reviewed-by: Jerome Glisse <jglisse@redhat.com> Signed-off-by: Dave Airlie <airlied@redhat.com>
Christian König [Wed, 19 Oct 2011 17:02:21 +0000 (19:02 +0200)]
drm/radeon: make some asic pointers per ring
Emitting fences, semaphores and ib works differently
on different ring, so its is easier to maintain
separate functions for each ring.
Signed-off-by: Christian König <deathsimple@vodafone.de> Reviewed-by: Jerome Glisse <jglisse@redhat.com> Signed-off-by: Dave Airlie <airlied@redhat.com>
Christian König [Tue, 27 Sep 2011 10:31:00 +0000 (12:31 +0200)]
drm/radeon: Add radeon_test_syncing function v2
Tests syncing between all rings by using
semaphores and fences.
v2: use radeon_testing as a bit flag rather than on/off switch
this allow to test for one thing at a time (bo_move or semaphore
test). It kind of break the usage if user wheren't using 1
for bo move test but as it's a test feature i believe it's ok.
Signed-off-by: Christian König <deathsimple@vodafone.de> Reviewed-by: Jerome Glisse <jglisse@redhat.com> Signed-off-by: Dave Airlie <airlied@redhat.com>
Christian König [Thu, 13 Oct 2011 11:19:22 +0000 (13:19 +0200)]
drm/radeon: make cp variable an array
Replace cp, cp1 and cp2 members with just an array
of radeon_cp structs.
Signed-off-by: Christian König <deathsimple@vodafone.de> Reviewed-by: Jerome Glisse <jglisse@redhat.com> Signed-off-by: Dave Airlie <airlied@redhat.com>
Christian König [Thu, 13 Oct 2011 10:48:45 +0000 (12:48 +0200)]
drm/radeon: make ring rptr and wptr register offsets variable
Every ring seems to have the concept of read and
write pointers. Make the register offset variable
so we can use the functions for different types of rings.
Signed-off-by: Christian König <deathsimple@vodafone.de> Reviewed-by: Jerome Glisse <jglisse@redhat.com> Signed-off-by: Dave Airlie <airlied@redhat.com>
Christian König [Fri, 23 Sep 2011 13:11:23 +0000 (15:11 +0200)]
drm/radeon: make all functions work with multiple rings.
Give all asic and radeon_ring_* functions a
radeon_cp parameter, so they know the ring to work with.
Signed-off-by: Christian König <deathsimple@vodafone.de> Reviewed-by: Jerome Glisse <jglisse@redhat.com> Signed-off-by: Dave Airlie <airlied@redhat.com>
Christian König [Thu, 15 Sep 2011 17:02:22 +0000 (19:02 +0200)]
drm/radeon/kms: add support for semaphores v3
They are used to sync between rings, while fences
sync between a ring and the cpu.
v2 Fix radeon_semaphore_driver_fini when no semaphore were
allocated.
v3 Initialize list early on to avoid issue in case or early
error
Signed-off-by: Christian König <deathsimple@vodafone.de> Signed-off-by: Jerome Glisse <jglisse@redhat.com> Signed-off-by: Dave Airlie <airlied@redhat.com>
Alex Deucher [Thu, 25 Aug 2011 17:39:48 +0000 (13:39 -0400)]
drm/radeon/kms: add support for multiple fence queues v2
For supporting multiple CP ring buffers, async DMA
engines and UVD. We still need a way to synchronize
between engines.
v2 initialize unused fence driver ring to avoid issue in
suspend/unload
Signed-off-by: Alex Deucher <alexander.deucher@amd.com> Signed-off-by: Christian König <deathsimple@vodafone.de> Reviewed-by: Jerome Glisse <jglisse@redhat.com> Signed-off-by: Dave Airlie <airlied@redhat.com>
Christian König [Mon, 24 Oct 2011 13:05:29 +0000 (15:05 +0200)]
drm/radeon: fix a spelling mistake
Better fix it before this obvious typo spreads even more.
Signed-off-by: Christian König <deathsimple@vodafone.de> Reviewed-by: Alex Deucher <alexander.deucher@amd.com> Reviewed-by: Jerome Glisse <jglisse@redhat.com> Signed-off-by: Dave Airlie <airlied@redhat.com>
Christian König [Mon, 24 Oct 2011 12:59:17 +0000 (14:59 +0200)]
drm/radeon: no need to check all relocs for duplicates
Only check the previously checked relocs for
duplicates. Also leaving the handle uninitialized
isn't such a good idea.
Signed-off-by: Christian König <deathsimple@vodafone.de> Reviewed-by: Jerome Glisse <jglisse@redhat.com> Signed-off-by: Dave Airlie <airlied@redhat.com>
projects / karo-tx-linux.git / log