- tests with subtests need to be in the _M target for correct
enumeration.
- No need for igt_exit when using the igt_(simple_)main wrappers.
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
This test demonstrates the performance cliff clients face when they
unwittingly use too many fenced surfaces in a looped upload.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
v2: Wait psr enable with timeout and more subtest added.
v3: Add wait for v_blank leeting test more reliable and preparing to
add Baytrail per-pipe tests.
v4: Call busy_ioctl on mmap_gtt to match the real usage and remove the need
of inactivate on set_domain, what was semantically wrong.
v5: Adding more test cases to cover mmap_gtt with and without followed by
busy ioctl and also without busy and waiting 10 seconds between
set_domain and actual write.
v6: rebase after kms_plane added and debugfs api changed.
Plus adding more test cases.
v7: Some rework and sprite test added
Signed-off-by: Rodrigo Vivi <rodrigo.vivi@gmail.com>
This reverts commit 9a70e29f8380c2fc6f9b43757c68e24bf2619061.
Patch needs polish and after a quick irc discussion Ben&I agreed that
the revert is in order.
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Switch to XY_COLOR_BLT from COLOR_BLT and use the appropriate
macros to make the code work on BDW.
Also make the blit 8bpp instead if 16bpp. 8bpp is what it was
supposed to use all along.
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=76307
Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
We actually allocate twice as many bo as needed to fill the aperture to
cause ENOSPC, but the entire payload has to still fit into RAM.
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=75138
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Move rc6_residency_check to subtest, add new rc6_residency_counter subtest
for pm_rc6_residency IGT case.
Test results run on platforms show as below:
On HSW
---------------------------------------
[root@x-hswu opt]# ./pm_rc6_residency
IGT-Version: 1.6-g35b31df (x86_64) (Linux: 3.15.0-rc3_drm-intel-nightly_0791a3_20140520+ x86_64)
Subtest rc6-residency-check: SUCCESS
This machine doesn't support rc6pp
This machine doesn't support rc6p
This machine entry rc6 status.
The residency counter : 0.999667
Subtest rc6-residency-counter: SUCCESS
On IVB
----------------------------------------
[root@IVB tests]# ./pm_rc6_residency
IGT-Version: 1.6-g35b31df (x86_64) (Linux: 3.13.6_20140318+ x86_64)
Subtest rc6-residency-check: SUCCESS
This machine entry rc6p status.
The residency counter : 0.997000
Subtest rc6-residency-counter: SUCCESS
On BYT
----------------------------------------
root@x-byt:/opt# ./pm_rc6_residency
IGT-Version: 1.6-g0d39021 (x86_64) (Linux: 3.14.0_kcloud_ceabbb_20140521+ x86_64)
Subtest rc6-residency-check: SUCCESS
This machine doesn't support rc6pp
This machine doesn't support rc6p
The residency counter : 1.144333
Test assertion failure function rc6_residency_counter, file pm_rc6_residency.c:131:
Last errno: 0, Success
Failed assertion: counter_result <=1
Debug files must be wrong,
Subtest rc6-residency-counter: FAIL
On BDW
---------------------------------------
[root@x-bdw01 opt]# ./pm_rc6_residency
IGT-Version: 1.6-g0d39021 (x86_64) (Linux: 3.15.0-rc5_drm-intel-nightly_367653_20140521+ x86_64)
Subtest rc6-residency-check: SUCCESS
This machine doesn't support rc6pp
This machine doesn't support rc6p
The residency counter : 0.994333
This machine entry rc6 state.
Subtest rc6-residency-counter: SUCCESS
Signed-off-by: Liu, Lei A <lei.a.liu@intel.com>
Signed-off-by: Wendy Wang <wendy.wang@intel.com>
Signed-off-by: Ben Widawsky <benjamin.widawsky@intel.com>
The goal is to only fill the aperture, not all of memory, so fine-tune
the computed number of surfaces, hopefully avoiding an integer overflow
in the process.
References: https://bugs.freedesktop.org/show_bug.cgi?id=79573
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
The EBUSY checking is very fragile currently in case there's any kind
extra delay in the test loop. At least the flip-vs-rmfb fails reliably
on my IVB.
So to make the test less fragile remove the EBUSY check from all the
current flip tests, and instead add a specific busy-flip test. To
better guarantee that we get the EBUSY add a blocking vblank wait just
before we issue the first flip. This maximizes the time we have to
submit two flips during one frame.
Reviewed-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
In Makefile.sources, kms_fence_pin_leak was in the "Multi" test list,
ie tests with subtests, whereas it is actually simple/single test.
This was confusing the Android script for enumerating tests.
Signed-off-by: Tim Gore <tim.gore@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
kms_render.c included testdisplay.h but did not need it.
This was preventing it from building on Android due to the
lack of a Glib port. So I have removed this #include and
changed Android.mk so that kms_render is built if we have
cairo.
Signed-off-by: Tim Gore <tim.gore@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
The missing macro that was preventing the gem_exec_params
test from building is now in i915_drm.h, in ABT at least,
and this test can now build. So I have removed it from the
skip list in Android.mk
For Gmin I have added a patch for i915_drm.h to the Wiki
Signed-off-by: Tim Gore <tim.gore@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
If we move the assertion from out of the callback, we can get a much
more useful error message.
References: https://bugs.freedesktop.org/show_bug.cgi?id=79573
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Now that we support cursor changes other than 64x64, a bug was found
where the size change was only applied at cursor enable time, rather
than at every update. Add a testcase for that.
Signed-off-by: Antti Koskipaa <antti.koskipaa@linux.intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Make sure DSPSURF will change during the panning operation
in flip-vs-panning-vs-hang.
This will now test agains bugs between the kernel's mmio vs.
CS flip race handling and GPU resets. If the kernel is buggy
if will fail to notice that the panning operation changed the
base address before the GPU reset had a chance to deal with the
pending page flips, and so the flip would never complete due to
DSPSURFLIVE not matching the expected value.
Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
The kms_mmio_vs_cs_flip test uses igt_kms.c which in turn
uses cairo. So in Android.mk add this test to the skip list
if we dont have cairo
Signed-off-by: Tim Gore <tim.gore@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Guarantees that error capture works at a very basic level.
v2: Also check that the ring object contains a reloc with MI_BB_START
for the presumed batch object's address.
v3: Chris review comments:
- Move variables to local scope.
- Do not assume there is only one request.
- Some gen encode flags into the BB start address.
Also, use igt_set/get_stop_rings as suggested by Mika Kuoppala.
v4: Make as a subtest of drv_hangman.
v5: Rebase
Signed-off-by: Oscar Mateo <oscar.mateo@intel.com> <v4>
Signed-off-by: Mika Kuoppala <mika.kuoppala@intel.com>
Mixing script and standlone tests didn't mix well with the
strict i915_ring_stop flags handling. Also squash drv_missed_irq_hang
to the new test.
v2: - Remove missed irq test (Daniel Vetter)
- gitignore fixed (Oscar Mateo)
- fix check_other_clients to handle dangling fd's
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=78322
Suggested-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Reviewed-by: Oscar Mateo <oscar.mateo@intel.com> <v1>
Signed-off-by: Mika Kuoppala <mika.kuoppala@intel.com>
When the patch was merged, the ioctl numbers had to be adjusted to leave
no holes. Also there was a final piece of munging of the API to
downgrade unsynced userptr for export over dma-buf.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
kms_mmio_vs_cs_flip has two subtests:
- setplane_vs_cs_flip tests the interaction between
fullscreen sprites and CS flips
- setcrtc_vs_cs_flip tests the interaction between
primary plane panning and CS flips
v2: Skip sprite test when there are no sprites
Reduce busy_bo to 64MB (now works on my gen2)
Handle pipe vs. port incompatibility
v3: Handle missing auto pipe crc
Reviewed-by Rodrigo Vivi <rodrigo.vivi@gmail.com>
Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
This reverts commit 3005ac3ee8d7aede73a3e63d2068a6074156ebe5.
QA has done the testing we've wanted.
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
- Some debug logging for the ts continuity checks.
- Add a plain vblank-vs-suspend test where kms_flip doesn't switch off
the displays first.
- Ensure that we do at least 2 rounds, the suspend tests bailed out
after 1 round and so didn't test anything.
- Frob the testnames a bit.
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
This reverts commit f00efff326610fdba92dbc91d951790a3320052e.
This is a temporary revert since I want QA to first test with the
original testcase whether it got faster again. This is to test the
effects of
commit 227f782e4667fc622810bce8be8ccdeee45f89c2
Author: Chris Wilson <chris@chris-wilson.co.uk>
Date: Thu May 15 10:41:42 2014 +0100
drm/i915: Retire requests before creating a new one
We should revert this revert again as soon as QA has completed
testing.
References: https://bugs.freedesktop.org/show_bug.cgi?id=78024
Cc: Jesse Barnes <jbarnes@virtuousgeek.org>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
These tests are really for catching TLB or GTT mapping failures due to
bad programming in the kernel driver. We've never needed more than a
few pages worth of data write to actually see those.
One proposed solution only fixes up the SNA behaviour, so reduce the
test case to probe only that particular pathology.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
If it's a simple test, it needs to be in the simple lists. Tests with
subtests go into the _M tests.
Without that test enumeration is all screwed up.
Cc: Rodrigo Vivi <rodrigo.vivi@gmail.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
This is a "review by igt test" for a bug located in
i915_gem_object_pin_to_display_plane and fixed by:
commit 392013bdd4b6128795e33c84bd6d6d3fd66ff0a3
Author: Oscar Mateo <oscar.mateo@intel.com>
Date: Fri May 16 11:23:12 2014 +0100
drm/i915: Gracefully handle obj not bound to GGTT in is_pin_display
Otherwise, we do a NULL pointer dereference.
I've seen this happen while handling an error in
i915_gem_object_pin_to_display_plane():
If i915_gem_object_set_cache_level() fails, we call is_pin_display()
to handle the error. At this point, the object is still not pinned
to GGTT and maybe not even bound, so we have to check before we
dereference its GGTT vma.
v2: Chris Wilson says restoring the old value is easier, but that
is_pin_display is useful as a theory of operation. Take the solomonic
decision: at least this way is_pin_display is a little more robust
(until Chris can kill it off).
v2: Avoid code duplication by using igt_create_fb_with_bo_size() as
requested by Ville Syrjälä (original author of the "too big" test idea).
Signed-off-by: Oscar Mateo <oscar.mateo@intel.com>
Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Plus naming convention for runtime pm tests to have "rpm" somewhere in
their tests. Note that all the pc8-specific tests (for e.g. residency
or similar) already have pc8 in their subtest names, so we don't lose
any information here.
Cc: "Yang, Guang A" <guang.a.yang@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
So I accidentally fixed a check when converting to the new macros and
the kernel never returned -ENOENT for invalid regs. Adjust the test.
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
These check whether everything is still ok wrt vblank handling after
runtime pm and system suspend-resume.
In addition to the usual checks they also ensure that the vblank frame
counter isn't totally ridiculous, something Keith complained about
aeons ago. With Ville's drm_vblank_on/off rework this should now be
fixed and solid.
v2:
- Ignore seq_step, vblanks completely immediately when the crtc goes off
- Only run system suspend/resume tests once.
Cc: Keith Packard <keithp@keithp.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
The kms_fence_pin_leak test uses igt_kms.c which in turn
uses cairo. So in Android.mk add this test to the skip list
if we dont have cairo
Issue: VIZ-3894
Signed-off-by: Tim Gore <tim.gore@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
The bug doesn't seem to occur on SNB, so we can skip the workaround and
hence we do not expect the kernel to prevent invalid relocated offsets.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
This test feeds a batch containing self-references into the kernel and
checks that the relocation offsets remain as valid GTT addresses. This
is to exercise SNA passing in negative relocation deltas which can hang
the GPU if they wrap around.
References: https://bugs.freedesktop.org/show_bug.cgi?id=78533
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
