Chris Wilson [Mon, 13 Feb 2017 17:15:37 +0000 (17:15 +0000)]
drm/i915: Add a live dmabuf selftest
Though we have good coverage of our dmabuf interface through the mock
tests, we also want to check the heavy module unload paths of the live
i915 driver.
Chris Wilson [Mon, 13 Feb 2017 17:15:35 +0000 (17:15 +0000)]
drm/i915: Sanity check all registers for matching fw domains
Add a late selftest that walks over all forcewake registers (those below
0x40000) and uses the mmio debug register to check to see if any are
unclaimed. This is possible if we fail to wake the appropriate
powerwells for the register.
Chris Wilson [Mon, 13 Feb 2017 17:15:34 +0000 (17:15 +0000)]
drm/i915: Test all fw tables during mock selftests
In addition to just testing the fw table we load, during the initial
mock testing we can test that all tables are valid (so the testing is
not limited to just the platforms that load that particular table).
Chris Wilson [Mon, 13 Feb 2017 17:15:28 +0000 (17:15 +0000)]
drm/i915: Add selftests for object allocation, phys
The phys object is a rarely used device (only very old machines require
a chunk of physically contiguous pages for a few hardware interactions).
As such, it is not exercised by CI and to combat that we want to add a
test that exercises the phys object on all platforms.
v2: Always set err on error paths and not rely on inheriting the err.
Chris Wilson [Mon, 13 Feb 2017 17:15:21 +0000 (17:15 +0000)]
drm/i915: Add selftests for i915_gem_request
Simple starting point for adding seltests for i915_gem_request, first
mock a device (with engines and contexts) that allows us to construct
and execute a request, along with waiting for the request to complete.
Chris Wilson [Mon, 13 Feb 2017 17:15:20 +0000 (17:15 +0000)]
drm/i915: Create a fake object for testing huge allocations
We would like to be able to exercise huge allocations even on memory
constrained devices. To do this we create an object that allocates only
a few pages and remaps them across its whole range - each page is reused
multiple times. We can therefore pretend we are rendering into a much
larger object.
Chris Wilson [Mon, 13 Feb 2017 17:15:13 +0000 (17:15 +0000)]
drm/i915: Add some selftests for sg_table manipulation
Start exercising the scattergather lists, especially looking at
iteration after coalescing.
v2: Comment on the peculiarity of table construction (i.e. why this
sg_table might be interesting).
v3: Added one __func__ to identify expect_pfn_sg()
v4: Loop until we have crossed the chain boundary (forcing sg_table to
do multiple allocations) before squelching a potential ENOMEM from oom.
Chris Wilson [Mon, 13 Feb 2017 17:15:12 +0000 (17:15 +0000)]
drm/i915: Provide a hook for selftests
Some pieces of code are independent of hardware but are very tricky to
exercise through the normal userspace ABI or via debugfs hooks. Being
able to create mock unit tests and execute them through CI is vital.
Start by adding a central point where we can execute unit tests and
a parameter to enable them. This is disabled by default as the
expectation is that these tests will occasionally explode.
To facilitate integration with igt, any parameter beginning with
i915.igt__ is interpreted as a subtest executable independently via
igt/drv_selftest.
Two classes of selftests are recognised: mock unit tests and integration
tests. Mock unit tests are run as soon as the module is loaded, before
the device is probed. At that point there is no driver instantiated and
all hw interactions must be "mocked". This is very useful for writing
universal tests to exercise code not typically run on a broad range of
architectures. Alternatively, you can hook into the live selftests and
run when the device has been instantiated - hw interactions are real.
v2: Add a macro for compiling conditional code for mock objects inside
real objects.
v3: Differentiate between mock unit tests and late integration test.
v4: List the tests in natural order, use igt to sort after modparam.
v5: s/late/live/
v6: s/unsigned long/unsigned int/
v7: Use igt_ prefixes for long helpers.
v8: Deobfuscate macros overriding functions, stop using -I$(src)
Chris Wilson [Sun, 12 Feb 2017 17:20:02 +0000 (17:20 +0000)]
drm/i915: Clear the last_retired_context following a hang/reset
Following a hang and reset, we know that the engine is idle and all
context state has been saved or lost. Consequently, we know that the
engine is no longer referencing the last context and we can relinquish
our tracking.
Chris Wilson [Sun, 12 Feb 2017 17:20:01 +0000 (17:20 +0000)]
drm/i915: Park the breadcrumbs signaler across a GPU reset
The signal threads may be running concurrently with the GPU reset. The
completion from the GPU run asynchronous with the reset and two threads
may see different snapshots of the state, and the signaler may mark a
request as complete as we try to reset it. We don't tolerate 2 different
views of the same state and complain if we try to mark a request as
failed if it is already complete. Disable the signal threads during
reset to prevent this conflict (even though the conflict implies that
the state we resetting to is invalid, we have already made our
decision!).
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=99733
References: https://bugs.freedesktop.org/show_bug.cgi?id=99671 Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> Cc: Mika Kuoppala <mika.kuoppala@intel.com> Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com> Link: http://patchwork.freedesktop.org/patch/msgid/20170212172002.23072-4-chris@chris-wilson.co.uk Reviewed-by: Mika Kuoppala <mika.kuoppala@intel.com>
Chris Wilson [Sun, 12 Feb 2017 17:20:00 +0000 (17:20 +0000)]
drm/i915: Kill the tasklet then disable
Disabling the tasklet leaves it if scheduled on the ready to run list
until it is re-enabled. This will leave the ksoftird thread spinning
until satisfied. To prevent this situation on starting the GPU reset, we
want to kill the tasklet first and then disable. The same problem will
arise when a tasklet is scheduled from another device, so a better
solution is required for the general case.
Reported-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com> Fixes: 1f7b847d72c3 ("drm/i915: Disable engine->irq_tasklet around resets") Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com> Cc: Mika Kuoppala <mika.kuoppala@intel.com> Link: http://patchwork.freedesktop.org/patch/msgid/20170212172002.23072-3-chris@chris-wilson.co.uk Reviewed-by: Mika Kuoppala <mika.kuoppala@intel.com>
Chris Wilson [Sun, 12 Feb 2017 17:19:59 +0000 (17:19 +0000)]
drm/i915: Always call i915_gem_reset_finish() following i915_gem_reset_prepare()
As i915_gem_reset_finish() undoes the steps from
i915_gem_reset_prepare() to leave the system in a fully-working state,
e.g. to be able to free the breadcrumb signal threads, make sure that we
always call it even on the error path.
Chris Wilson [Tue, 7 Feb 2017 10:23:19 +0000 (10:23 +0000)]
drm/i915: Rename conditional GEM execution macros
After a brief discussion, we settled on a naming convention for the
conditional GEM debugging data that should be clearer to the casual
user: GEM_DEBUG
Chris Wilson [Fri, 10 Feb 2017 18:52:14 +0000 (18:52 +0000)]
drm/i915: Move the irq_barrier for reset earlier into reset_prepare
When updating the bookkeeping following the reset, we need the seqno to
be coherent on the CPU prior to trusting its result for deciding whether
any request is completed. We need the irq_barrier before we start making
these decisions, i.e. in reset_prepare.
Chris Wilson [Fri, 10 Feb 2017 16:35:23 +0000 (16:35 +0000)]
drm/i915: Flush the freed object queue on device release
As dmabufs may live beyond the PCI device removal, we need to flush the
freed object worker on device release, and include a warning in case
there is a leak.
Chris Wilson [Fri, 10 Feb 2017 16:35:22 +0000 (16:35 +0000)]
drm/i915: Unbind any residual objects/vma from the Global GTT on shutdown
We may unload the PCI device before all users (such as dma-buf) are
completely shutdown. This may leave VMA in the global GTT which we want
to revoke, whilst keeping the objects themselves around to service the
dma-buf.
Chris Wilson [Fri, 10 Feb 2017 16:35:21 +0000 (16:35 +0000)]
drm/i915: Split device release from unload
We may need to keep our memory management alive after we have unloaded
the physical pci device. For example, if we have exported an object via
dmabuf, that will keep the device around but the pci device may be
removed before the dmabuf itself is released, use of the pci hardware
will be revoked, but the memory and object management needs to persist
for the dmabuf.
Lyude [Sat, 4 Feb 2017 02:18:25 +0000 (21:18 -0500)]
drm/i915/debugfs: Add i915_hpd_storm_ctl
This adds a file in i915's debugfs directory that allows userspace to
manually control HPD storm detection. This is mainly for hotplugging
tests, where we might want to test HPD storm functionality or disable
storm detection to speed up hotplugging tests without breaking anything.
Changes since v1:
- Make HPD storm interval configurable
- Misc code cleanup
Ville Syrjälä [Wed, 8 Feb 2017 17:52:54 +0000 (19:52 +0200)]
drm/i915: Avoid spurious WARNs about the wrong pipe in the PPS code
Until recently vlv_steal_power_sequencer() wasn't being called for
normal DP ports, and hence it could assert that it should only be
called for pipe A and B (since pipe C doesn't support eDP). However
that changed when we started to consider normal DP ports as well when
choosing a PPS. So we will now get spurious warnings when
vlv_steal_power_sequencer() does get called for pipe C. Avoid this by
moving the WARN down into vlv_detach_power_sequencer() where this
assertion should still hold.
Cc: Imre Deak <imre.deak@intel.com> Cc: stable@vger.kernel.org Fixes: 9f2bdb006a7e ("drm/i915: Prevent PPS stealing from a normal DP port on VLV/CHV")
References: https://bugs.freedesktop.org/show_bug.cgi?id=95287 Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com> Link: http://patchwork.freedesktop.org/patch/msgid/20170208175254.10958-1-ville.syrjala@linux.intel.com Reviewed-by: Imre Deak <imre.deak@intel.com>
Chris Wilson [Fri, 10 Feb 2017 10:14:22 +0000 (10:14 +0000)]
drm/i915: Always pin contexts into the high GGTT
Now that we have fast top-down insertion into the drm_mm, we can use it
for frequent runtime operations like insertion of the context object,
whereas before we limited it to the one-off insertion of the pinned
kernel context. Keeping the active context objects out of the mappable
region of the global GTT (except under memory pressure) improves our
ability to allocate mappable aperture region without triggering a GPU
stall.
drm/i915: Remove unused function intel_ddi_get_link_dpll()
The function intel_ddi_get_link_dpll() was added in f169660ed4e5
("drm/i915/dp: Add a standalone function to obtain shared dpll for
HSW/BDW/SKL/BXT") to "allow for the implementation of a platform
neutral upfront link training function", but such implementation
never landed.
So remove that function and clean up the exported shared DPLL interface.
Fixes: f169660ed4e5 ("drm/i915/dp: Add a standalone function to obtain shared dpll for HSW/BDW/SKL/BXT") Cc: Durgadoss R <durgadoss.r@intel.com> Cc: Manasi Navare <manasi.d.navare@intel.com> Cc: Jim Bride <jim.bride@linux.intel.com> Cc: Rodrigo Vivi <rodrigo.vivi@intel.com> Cc: Daniel Vetter <daniel.vetter@intel.com> Cc: Jani Nikula <jani.nikula@linux.intel.com> Cc: intel-gfx@lists.freedesktop.org Signed-off-by: Ander Conselvan de Oliveira <ander.conselvan.de.oliveira@intel.com> Acked-by: Jani Nikula <jani.nikula@intel.com> Link: http://patchwork.freedesktop.org/patch/msgid/1484310032-1863-1-git-send-email-ander.conselvan.de.oliveira@intel.com
There are currently 30 power domains, which puts us pretty close to the
limit with 32 bit masks. Prepare for the future and increase the limit
to 64 bit.
v2: Rebase
v3: s/unsigned long long/u64/ (Joonas)
Allow the 64th bit of the mask to be used. (Joonas)
Cc: Jani Nikula <jani.nikula@linux.intel.com> Cc: Joonas Lahtinen <joonas.lahtinen@linux.intel.com> Cc: Rodrigo Vivi <rodrigo.vivi@intel.com> Cc: Ville Syrjälä <ville.syrjala@linux.intel.com> Signed-off-by: Ander Conselvan de Oliveira <ander.conselvan.de.oliveira@intel.com> Reviewed-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com> Link: http://patchwork.freedesktop.org/patch/msgid/20170209093121.24410-1-ander.conselvan.de.oliveira@intel.com
Dave Airlie [Fri, 10 Feb 2017 00:13:30 +0000 (10:13 +1000)]
Merge branch 'drm-next-4.11' of git://people.freedesktop.org/~agd5f/linux into drm-next
Some additional fixes for 4.11. Delayed a bit due to Chinese New Year. Highlights:
- Powerplay fixes
- VCE and UVD powergating fixes
- Clean up amdgpu SI gfx code to match CI and VI
- Misc bug fixes
* 'drm-next-4.11' of git://people.freedesktop.org/~agd5f/linux: (30 commits)
drm/amdgpu: report the number of bytes moved at buffer creation
drm/amdgpu: fix a potential deadlock in amdgpu_bo_create_restricted()
drm/amdgpu: add support for new smc firmware on polaris
drm/amd/powerplay: refine code to avoid potential bug that the memory not cleared.
drm/amdgpu: shut up #warning for compile testing
drm/amdgpu/virt: fix double kfree on bo_va
drm/radeon: remove some dead code
drm/radeon: avoid kernel segfault in vce when gpu fails to resume
drm/amd/powerplay: set fan speed to max in profile peak mode only.
drm/amd/gfx6: update gb_addr_config
drm/amdgpu: update HAINAN_GB_ADDR_CONFIG_GOLDEN
drm/amdgpu: update VERDE_GB_ADDR_CONFIG_GOLDEN
drm/amdgpu: refine si_read_register
drm/amdgpu/gfx6: clean up spi configuration
drm/amdgpu/gfx6: clean up cu configuration
drm/amdgpu/gfx6: clean up rb configuration
drm/amdgpu: refine vce3.0 code and related powerplay pg code.
drm/amdgpu: move subfunctions to the front of vce_v2_0.c.
drm/amdgpu: enable vce pg feature on Kv.
drm/amdgpu: refine code for VCE2.0 and related dpm code.
...
Dave Airlie [Fri, 10 Feb 2017 00:12:56 +0000 (10:12 +1000)]
Merge tag 'drm-fsl-dcu-for-v4.11' of http://git.agner.ch/git/linux-drm-fsl-dcu into drm-next
two minor fixes.
* tag 'drm-fsl-dcu-for-v4.11' of http://git.agner.ch/git/linux-drm-fsl-dcu:
drm/fsl-dcu: check for clk_prepare_enable() error
drm/fsl-dcu: remove unneeded 'ret' assignment
Dave Airlie [Fri, 10 Feb 2017 00:12:00 +0000 (10:12 +1000)]
Merge tag 'drm-misc-next-fixes-2017-02-09' of git://anongit.freedesktop.org/git/drm-misc into drm-next
Just 3 bugfixes for 4.11 merge window:
- fbdev module unload oops fix from Chris
- patch from Dan that look really dangers, better safe than sorry
* tag 'drm-misc-next-fixes-2017-02-09' of git://anongit.freedesktop.org/git/drm-misc:
drm/atomic: fix an error code in mode_fixup()
drm: Cancel drm_fb_helper_resume_work on unload
drm: Cancel drm_fb_helper_dirty_work on unload
Chris Wilson [Thu, 9 Feb 2017 14:40:36 +0000 (14:40 +0000)]
drm/i915: Use the size/type of address space to make decisions
Once the address space has been created (using 3 or 4 levels of page
tables), we should use that to program the appropriate type into the
contexts. This gives us the flexibility to handle different types of
address spaces at runtime.
Samuel Pitoiset [Thu, 9 Feb 2017 10:33:37 +0000 (11:33 +0100)]
drm/amdgpu: report the number of bytes moved at buffer creation
Like ttm_bo_validate(), ttm_bo_init() might need to move BO and
the number of bytes moved by TTM should be reported. This can help
the throttle buffer migration mechanism to make a better decision.
v2: fix computation
Reviewed-by: Christian König <christian.koenig@amd.com> Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Samuel Pitoiset [Thu, 9 Feb 2017 10:33:36 +0000 (11:33 +0100)]
drm/amdgpu: fix a potential deadlock in amdgpu_bo_create_restricted()
When ttm_bo_init() fails, the reservation mutex should be unlocked.
In debug build, the kernel reported "possible recursive locking
detected" in this codepath. For debugging purposes, I also added
a "WARN_ON(ww_mutex_is_locked())" when ttm_bo_init() fails and the
mutex was locked as expected.
This should fix (random) GPU hangs. The easy way to reproduce the
issue is to change the "Super Sampling" option from 1.0 to 2.0 in
Hitman. It will create a huge buffer, evict a bunch of buffers
(around ~5k) and deadlock.
This regression has been introduced pretty recently.
v2: only release the mutex if resv is NULL
Fixes: 12a852219583 ("drm/amdgpu: improve AMDGPU_GEM_CREATE_VRAM_CLEARED handling (v2)") Reviewed-by: Christian König <christian.koenig@amd.com> Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
drm/i915: Enable atomic support by default on supported platforms.
i915 is pretty much feature complete. Support for atomic i915-specific
connector properties is still missing; those properties can (for now)
only be set through the legacy ioctl.
ILK style watermarks and gen9+ watermarks are handled atomically,
and nonblocking modesets work. FBC has also been made to work with
atomic.
gen4x- and vlv/chv watermarks still need to be fixed, so disable atomic
by default there for now.
Flip the switch!!
Signed-off-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com> Cc: Ander Conselvan de Oliveira <ander.conselvan.de.oliveira@intel.com> Cc: Chris Wilson <chris@chris-wilson.co.uk> Cc: Daniel Vetter <daniel.vetter@ffwll.ch> Cc: Lyude <cpaul@redhat.com> Cc: Matt Roper <matthew.d.roper@intel.com> Cc: Paulo Zanoni <paulo.r.zanoni@intel.com> Cc: Rodrigo Vivi <rodrigo.vivi@intel.com> Cc: Ville Syrjälä <ville.syrjala@linux.intel.com> Link: http://patchwork.freedesktop.org/patch/msgid/1486021302-24910-1-git-send-email-maarten.lankhorst@linux.intel.com
[mlankhorst: Fix checkpatch warning about extra space in match_info] Acked-by: Daniel Stone <daniel@fooishbar.org> Reviewed-by: Lyude <lyude@redhat.com>
Chris Wilson [Thu, 9 Feb 2017 11:19:33 +0000 (11:19 +0000)]
drm/i915: Assert that we never create a vma for the aliasing_ppgtt
The aliasing_ppgtt is just a container for the HW context that mirrors
the global gtt. It should never be used directly, so assert if we make
the mistake of trying to allocate a VMA for it.
Chris Wilson [Wed, 8 Feb 2017 18:12:38 +0000 (18:12 +0000)]
drm/i915: Check for timeout completion when waiting for the rq to submitted
We first wait for a request to be submitted to hw and assigned a seqno,
before we can wait for the hw to signal completion (otherwise we don't
know the hw id we need to wait upon). Whilst waiting for the request to
be submitted, we may exceed the user's timeout and need to propagate the
error back.
v2: Make ETIME into an error from wait_for_execute for consistent exit
handling.
Reported-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com> Fixes: 4680816be336 ("drm/i915: Wait first for submission, before waiting for request completion")
Testcase: igt/gem_wait/basic-await Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com> Cc: Joonas Lahtinen <joonas.lahtinen@linux.intel.com> Cc: <drm-intel-fixes@lists.freedesktop.org> # v4.10-rc1+ Cc: stable@vger.kernel.org Link: http://patchwork.freedesktop.org/patch/msgid/20170208181238.7232-1-chris@chris-wilson.co.uk Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Arnd Bergmann [Wed, 1 Feb 2017 15:59:21 +0000 (16:59 +0100)]
drm/amdgpu: shut up #warning for compile testing
My randconfig tests on linux-next showed a newly introduced warning:
drivers/gpu/drm/amd/amdgpu/amdgpu_object.c: In function 'amdgpu_bo_create_restricted':
drivers/gpu/drm/amd/amdgpu/amdgpu_object.c:377:2: error: #warning Please enable CONFIG_MTRR and CONFIG_X86_PAT for better performance thanks to write-combining [-Werror=cpp]
Generally speaking, warnings about bad kernel configuration are not particularly
helpful. We could enforce the selection of X86_PAT through Kconfig, so the driver
cannot even be used unless it is enabled, or we could just rely on the runtime
warning that is also there.
In this version, I'm making the warning conditional on CONFIG_COMPILE_TEST, which
shuts it up for me, but not people that may actually want to run the kernel
as a compromize.
Fixes: a2e2f29970aa ("drm/amdgpu: Bring bo creation in line with radeon driver (v2)") Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> Signed-off-by: Arnd Bergmann <arnd@arndb.de> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Colin Ian King [Fri, 3 Feb 2017 20:23:42 +0000 (20:23 +0000)]
drm/amdgpu/virt: fix double kfree on bo_va
bo_va is being kfree'd twice, once in the call to amdgpu_vm_bo_rmv
and then a short while later. Fix this double free by removing
the 2nd kfree.
Detected by CoverityScan, CID#1399524 ("Double Free")
Reviewed-by: Monk Liu <monk.liu@amd.com> Signed-off-by: Colin Ian King <colin.king@canonical.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Dan Carpenter [Tue, 7 Feb 2017 13:16:04 +0000 (16:16 +0300)]
drm/radeon: remove some dead code
If "rdev->bios" is NULL then we don't need to free it.
Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Jérôme Glisse [Mon, 6 Feb 2017 20:13:18 +0000 (15:13 -0500)]
drm/radeon: avoid kernel segfault in vce when gpu fails to resume
When GPU fails to resume we can not trust that value we write to GPU
memory will post and we might get garbage (more like 0xffffffff on
x86) when reading them back. This trigger out of range memory access
in the kernel inside the vce resume code path.
This patch use canonical value to compute offset instead of reading
back value from GPU memory.
Reviewed-by: Christian König <christian.koenig@amd.com> Signed-off-by: Jérôme Glisse <jglisse@redhat.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Rex Zhu [Thu, 26 Jan 2017 02:47:00 +0000 (10:47 +0800)]
drm/amdgpu: refine vce3.0 code and related powerplay pg code.
1. not start vce3.0 when hw_init
2. stop vce3.0 when vce idle.
3. pg mask used to ctrl power down/up vce.
4. change cg pg sequence in powerplay.
Signed-off-by: Rex Zhu <Rex.Zhu@amd.com> Acked-by: Christian König <christian.koenig@amd.com> Reviewed-by: Alex Deucher <alexander.deucher@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Rex Zhu [Thu, 26 Jan 2017 08:46:22 +0000 (16:46 +0800)]
drm/amdgpu: move subfunctions to the front of vce_v2_0.c.
Signed-off-by: Rex Zhu <Rex.Zhu@amd.com> Acked-by: Christian König <christian.koenig@amd.com> Reviewed-by: Alex Deucher <alexander.deucher@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Rex Zhu [Wed, 25 Jan 2017 08:50:15 +0000 (16:50 +0800)]
drm/amdgpu: enable vce pg feature on Kv.
Signed-off-by: Rex Zhu <Rex.Zhu@amd.com> Acked-by: Christian König <christian.koenig@amd.com> Reviewed-by: Alex Deucher <alexander.deucher@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Rex Zhu [Thu, 26 Jan 2017 08:25:05 +0000 (16:25 +0800)]
drm/amdgpu: refine code for VCE2.0 and related dpm code.
v2: clean up vce cg function.
use sw cg when vce stoped.
1. implement vce_stop function.
2. not start vce when hw_init.
3. refine vce cg/pg code.
4. delete bypass mode.
Signed-off-by: Rex Zhu <Rex.Zhu@amd.com> Acked-by: Christian König <christian.koenig@amd.com> Reviewed-by: Alex Deucher <alexander.deucher@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Rex Zhu [Wed, 25 Jan 2017 09:35:14 +0000 (17:35 +0800)]
drm/amdgpu: when dpm disabled, also need to stop/start vce.
Signed-off-by: Rex Zhu <Rex.Zhu@amd.com> Reviewed-by: Alex Deucher <alexander.deucher@amd.com> Acked-by: Christian König <christian.koenig@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Rex Zhu [Fri, 20 Jan 2017 09:46:34 +0000 (17:46 +0800)]
drm/amdgpu: refine uvd5.0/6.0 code.
1. delete redundant cg pg mask check.
pg mask use to ctrl power on/down uvd.
not start/stop uvd.
cg mask will be check when enable mgcg.
2. no need to start uvd when initializ.
when ring test/ib test/encode, uvd was enabled.
when uvd idle, uvd was stopped.
3. chang cg pg sequence in powerplay.
Signed-off-by: Rex Zhu <Rex.Zhu@amd.com> Reviewed-by: Alex Deucher <alexander.deucher@amd.com> Acked-by: Christian König <christian.koenig@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Rex Zhu [Fri, 20 Jan 2017 07:56:45 +0000 (15:56 +0800)]
drm/amdgpu: refine uvd4.2 init/stop code.
1. set uvd_status busy before uvd_start.
2. clear uvd_status to 0 after uvd stop.
smu firmware may check uvd_status.
3. wait uvd idle before stop uvd.
4. not start uvd when hw_init.
Signed-off-by: Rex Zhu <Rex.Zhu@amd.com> Acked-by: Alex Deucher <alexander.deucher@amd.com> Acked-by: Christian König <christian.koenig@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Rex Zhu [Fri, 20 Jan 2017 06:34:43 +0000 (14:34 +0800)]
drm/amdgpu: refine uvd pg code in kv_dpm.c
1. no need to set cg as use hw dynamic cg.
2. when uvd idle, stop uvd. encode, start uvd.
3. if pg feature enabled, power on/down uvd by smu.
4. drm/amdgpu: dpm do not set uvd pg status.
Signed-off-by: Rex Zhu <Rex.Zhu@amd.com> Acked-by: Christian König <christian.koenig@amd.com> Reviewed-by: Alex Deucher <alexander.deucher@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Rex Zhu [Fri, 20 Jan 2017 07:07:47 +0000 (15:07 +0800)]
drm/amdgpu: power down/up uvd4 when smu disabled.
Signed-off-by: Rex Zhu <Rex.Zhu@amd.com> Acked-by: Christian König <christian.koenig@amd.com> Reviewed-by: Alex Deucher <alexander.deucher@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Rex Zhu [Wed, 25 Jan 2017 04:17:59 +0000 (12:17 +0800)]
drm/amdgpu: add current_pg_status register define for smu7.1
Signed-off-by: Rex Zhu <Rex.Zhu@amd.com> Acked-by: Christian König <christian.koenig@amd.com> Reviewed-by: Alex Deucher <alexander.deucher@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Rex Zhu [Fri, 20 Jan 2017 04:06:05 +0000 (12:06 +0800)]
drm/amdgpu: when dpm disabled, also can enable uvd cg/pg.
Signed-off-by: Rex Zhu <Rex.Zhu@amd.com> Acked-by: Christian König <christian.koenig@amd.com> Reviewed-by: Alex Deucher <alexander.deucher@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Chris Wilson [Wed, 8 Feb 2017 14:30:33 +0000 (14:30 +0000)]
drm/i915: Disable engine->irq_tasklet around resets
When we restart the engines, and we have active requests, a request on
the first engine may complete and queue a request to the second engine
before we try to restart the second engine. That queueing of the
request may race with the engine to restart, and so may corrupt the
current state. Disabling the engine->irq_tasklet prevents the two paths
from writing into ELSP simultaneously (and modifyin the execlists_port[]
at the same time).
Fixes: 821ed7df6e2a ("drm/i915: Update reset path to fix incomplete requests")
Testcase: igt/gem_exec_fence/await-hang Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com> Cc: Mika Kuoppala <mika.kuoppala@intel.com> Reviewed-by: Mika Kuoppala <mika.kuoppala@intel.com> Link: http://patchwork.freedesktop.org/patch/msgid/20170208143033.11651-3-chris@chris-wilson.co.uk
Chris Wilson [Wed, 8 Feb 2017 14:30:32 +0000 (14:30 +0000)]
drm/i915: Split GEM resetting into 3 phases
Currently we do a reset prepare/finish around the call to reset the GPU,
but it looks like we need a later stage after the hw has been
reinitialised to allow GEM to restart itself. Start by splitting the 2
GEM phases into 3:
prepare - before the reset, check if GEM recovered, then stop GEM
reset - after the reset, update GEM bookkeeping
finish - after the re-initialisation following the reset, restart GEM
Chris Wilson [Wed, 8 Feb 2017 14:30:31 +0000 (14:30 +0000)]
drm/i915: Move calling engine->init_hw() to its own function
Just a simple refactor to move a loop and some support code out of
i915_gem_init_hw(). This is in preparation for avoiding a race between
the tasklet writing to ELSP whilst simultaneously being written by
engine->init_hw() following a GPU reset.
Chris Wilson [Wed, 8 Feb 2017 10:47:10 +0000 (10:47 +0000)]
drm/i915: i915_gem_shrink_all() needs an awake device
Since to unbind an object, we may need a powered up device to access the
GTT entries, we only shrink bound objects if awake. Callers to
i915_gem_shrink_all() had to take this into account and take the rpm
wakeref, but we can move this wakeref into the shrink_all itself for
convenience and making the function live up to its name.
Ville Syrjälä [Thu, 26 Jan 2017 19:52:01 +0000 (21:52 +0200)]
drm/i915: Replace the .modeset_commit_cdclk() hook with a more direct .set_cdclk() hook
With the cdclk state, all the .modeset_commit_cdclk() hooks are
now pointless wrappers. Let's replace them with just a .set_cdclk()
function pointer. However let's wrap that in a small helper that
does the state comparison and prints a unified debug message across
all platforms. We didn't even have the debug print on all platforms
previously. This reduces the clutter in intel_atomic_commit_tail() a
little bit.
v2: Wrap .set_cdclk() in intel_set_cdclk()
v3: Add kernel-docs
v4: Deal with IS_GEN9_BC()
Ville Syrjälä [Fri, 20 Jan 2017 18:22:03 +0000 (20:22 +0200)]
drm/i915: Nuke the VLV/CHV PFI programming power domain workaround
The hack to grab the pipe A power domain around VLV/CHV cdclk
programming has surely outlived its usefulness. We should be
holding sufficient power domains during any modeset, so let's
just nuke this hack.
Ville Syrjälä [Fri, 20 Jan 2017 18:21:59 +0000 (20:21 +0200)]
drm/i915: Track full cdclk state for the logical and actual cdclk frequencies
The current dev_cdclk vs. cdclk vs. atomic_cdclk_freq is quite a mess.
So here I'm introducing the "actual" and "logical" naming for our
cdclk state. "actual" is what we'll bash into the hardware and "logical"
is what everyone should use for state computaion/checking and whatnot.
We'll track both using the intel_cdclk_state as both will need other
differing parameters than just the actual cdclk frequency.
While doing that we can at the same time unify the appearance of the
.modeset_calc_cdclk() implementations a little bit.
v2: Commit dev_priv->cdclk.actual since that already has the
new state by the time .modeset_commit_cdclk() is called.
v3: s/locical/logical/ and improve the docs a bit
Ville Syrjälä [Tue, 7 Feb 2017 18:33:45 +0000 (20:33 +0200)]
drm/i915: Start moving the cdclk stuff into a distinct state structure
Introduce intel_cdclk state which for now will track the cdclk
frequency, the vco frequency and the reference frequency (not sure we
want the last one, but I put it there anyway). We'll also make the
.get_cdclk() function fill out this state structure rather than
just returning the current cdclk frequency.
One immediate benefit is that calling .get_cdclk() will no longer
clobber state stored under dev_priv unless ex[plicitly told to do
so. Previously it clobbered the vco and reference clocks stored
there on some platforms.
We'll expand the use of this structure to actually precomputing the
state and whatnot later.
v2: Constify intel_cdclk_state_compare()
v3: Document intel_cdclk_state_compare()
v4: Deal with i945gm_get_cdclk()
Ville Syrjälä [Tue, 7 Feb 2017 18:33:05 +0000 (20:33 +0200)]
drm/i915: Move most cdclk/rawclk related code to intel_cdclk.c
Let's try to shrink intel_display.c a bit by moving the cdclk/rawclk
stuff to a new file. It's all reasonably self contained so we don't
even have to add that many non-static symbols.
We'll also take the opportunity to shuffle around the functions a bit
to get things in a more consistent order based on the platform.
v2: Add kernel-docs (Ander)
v3: Deal with IS_GEN9_BC()
v4: Deal with i945gm_get_cdclk()
Ville Syrjälä [Tue, 7 Feb 2017 18:32:26 +0000 (20:32 +0200)]
drm/i915: Clean up the .get_cdclk() assignment if ladder
Let's clean up the mess we have in the if ladder that assigns the
.get_cdclk() hooks. The grouping of the platforms by the function
results in a thing that's not really legible, so let's do it the
other way around and order the if ladder by platform and duplicate
whatever assignments we need.
To further avoid confusion with the function names let's rename
them to just fixed_<freq>_get_cdclk(). The other option would
be to duplicate the functions entirely but it seems quite
pointless to do that since each one just returns a fixed value.
Ville Syrjälä [Fri, 20 Jan 2017 18:21:53 +0000 (20:21 +0200)]
drm/i915: Nuke intel_mode_max_pixclk()
ilk_max_pixel_rate() will now give the "correct" pixel rate for all
platforms, so let's rename it to intel_max_pixel_rate() and kill
off intel_mode_max_pixclk().
projects / mirror_ubuntu-focal-kernel.git / log