The power metter was not showing up due to an erroneous check for a
failure to open the i915 perf interface.
Reported-by: Marius Vlad <marius.c.vlad@intel.com>
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
In this subtest, as a first step, MAX_FENCES+1 number of framebuffers are
created backed up by objects that have multiple GGTT views (normal and
rotated). Next, we have the i915 driver instantiate a normal view followed
by a rotated view. We continue doing the above MAX_FENCES + 1 times.
v2:
- Add a igt_require() to check if there is enough GTT space left for
MAX_FENCES+1 framebuffers. (Tvrtko)
- Make data2 local to test_plane_rotation_exhaust_fences(). (Tvrtko)
- If there is a failure, deallocate all the previously allocated
framebuffers before asserting.
v3: Close the gem handle if set_tiling or addfb fails. (Tvrtko)
Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Signed-off-by: Vivek Kasireddy <vivek.kasireddy@intel.com>
Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Signed-off-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
We don't allow ARGB8888 anymore on primary planes on most platforms,
so use XRGB8888 instead as the format.
Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Would be nice to see how many stereo modes we managed to extract from
the EDID if it doesn't match the expected 13. So use igt_assert_eq()
which prints the real count on failure.
Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
The setcrtc ioctl ignores the fb_id when there's no mode specified.
So passing -1 doens't make much sense. When there is a more, -1 means
to preserve the current fb.
Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
To make more multi-pipe tests run on IVB, do the modesets in the reverse
order (ie. pipe C first, pipe A last). This way pipe B can't reserve the
2 shared FDI lanes before pipe C is set up.
Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Currently kms_flip leaks the state of the pipes from one subtest to the
next. Meaning a single pipe test can actually have two or more pipes
actually up and running, and similarly a two pipe test can have three
pipes running.
This is particularly nasty on IVB since one of the pipes still running
but not actually part of the test maybe have reserved the shared FDI
lanes, thus preventing one of the pipes taking part in the test from
being enabled.
To avoid such problems explicitly disable all pipes before each
subtests.
v2: Use kmstest_unset_all_crtcs() (Paulo)
Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Reduced the Sleep period to 200mS and reduced the repetition count to 7
to decrease the test run time significantly.
v2: Changed uS to us
v3: removed the output formatting change as the issue will be addressed
in a seperate patch from Thomas Wood.
v4: mS -> ms
Signed-off-by: Derek Morton <derek.j.morton@intel.com>
Signed-off-by: Thomas Wood <thomas.wood@intel.com>
There is a MST encoder for each crtc, and each MST connector
will be connected to the encoder bound to that crtc.
This breaks the kms_setmode assertion that is only 1 encoder per
connector, so make an exception to that rule for displayport.
Signed-off-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com>
When extending the batch for multiple copies, we need to remember to
flag it as being in the CPU write domain so that the new values get
flushed out to main memory before execution. We also have to be careful
not to specify NO_RELOC for the extended batch as the execobjects will
have been updated but we write the wrong presumed offsets. Subsequent
iterations will be correct and we can tell the kernel then to skip the
relocations entirely.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Older generations are more limited in how much they can fence, and the
limits is enforced in the SET_TILING ioctl. So if it reports an EINVAL,
we cannot perform the tiled test and may just skip it instead.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Disable output of terminal control characters and progress meters when
IGT_PLAIN_OUTPUT is set in the environment.
Cc: Derek Morton <derek.j.morton@intel.com>
Signed-off-by: Thomas Wood <thomas.wood@intel.com>
Initialization was included in commit a976d7e (tests/kms_fbc_crc:
refactor context handling code), but won't be executed since it is
declared before the first label within a switch statement.
kms_fbc_crc.c:178:2: warning: ‘context’ may be used uninitialized in this function [-Wmaybe-uninitialized]
rendercopy(batch, context,
^
kms_fbc_crc.c:271:22: note: ‘context’ was declared here
drm_intel_context *context = NULL;
^
Cc: Paulo Zanoni <paulo.r.zanoni@intel.com>
Signed-off-by: Thomas Wood <thomas.wood@intel.com>
Reviewed-by: Paulo Zanoni <paulo.r.zanoni@intel.com>
This prevents the simulator from barfing when it sees commands from another
ring. I've been using this locally for a very long time.
Signed-off-by: Ben Widawsky <benjamin.widawsky@intel.com>
Reviewed-by: Kristian Høgsberg <krh@bitplanet.net>
Add tests for KMS atomic modesetting, to exercise the basic interface
and test failure/corner cases. Should ensure coherency between the
legacy and atomic interfaces.
v2: New patch.
v3: Disable connector checking for now, as it was causing GPU hangs on
newer kernels.
v4: Rebase.
v5: Use do_ioctl or do_ioctl_err consistently. Use igt_assert_*()
helper macros rather than igt_assert() directly.
Move assertions into helper/check functions. Define atomic commit
helper.
v6: Use do_ioctl_err, and define macros to move errors to
actual callsite, rather than helper functions.
Co-authored-by: Micah Fedke <micah.fedke@collabora.com>
Signed-off-by: Daniel Stone <daniels@collabora.com>
This should hit the bug fixed in:
commit 8731b269f01e16193390c7276e70530366b8d626
Author: Maneet Singh <mmaneetsingh@nvidia.com>
Date: Thu Oct 8 10:10:24 2015 -0400
drm: Correct arguments to list_tail_add in create blob ioctl
Arguments passed to list_add_tail were reversed resulting in deletion
of old blob property everytime the new one is added.
Fixes
commit e2f5d2ea479b9b2619965d43db70939589afe43a
Author: Daniel Stone <daniels@collabora.com>
Date: Fri May 22 13:34:51 2015 +0100
drm/mode: Add user blob-creation ioctl
Signed-off-by: Maneet Singh <mmaneetsingh@nvidia.com>
[seanpaul tweaked commit subject a little]
Signed-off-by: Sean Paul <seanpaul@chromium.org>
Cc: stable@kernel.org # v4.2
Reviewed-by: Daniel Stone <daniels@collabora.com>
Reviewed-by: Jani Nikula <jani.nikula@intel.com>
Signed-off-by: Dave Airlie <airlied@gmail.com>
which was introduced with the initial blob support in:
commit e2f5d2ea479b9b2619965d43db70939589afe43a
Author: Daniel Stone <daniels@collabora.com>
Date: Fri May 22 13:34:51 2015 +0100
drm/mode: Add user blob-creation ioctl
Add an ioctl which allows users to create blob properties from supplied
data. Currently this only supports modes, creating a drm_display_mode from
the userspace drm_mode_modeinfo.
v2: Removed size/type checks.
Rebased on new patches to allow error propagation from create_blob,
as well as avoiding double-allocation.
Signed-off-by: Daniel Stone <daniels@collabora.com>
Reviewed-by: Maarten Lankhorst <maarten.lankhorst@intel.com>
Tested-by: Sean Paul <seanpaul@chromium.org>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Signed-off-by: Daniel Stone <daniels@collabora.com>
Since we allocate 2 times count number of objects, that is the number of
objects we should be using when checking for sufficient RAM to use for
the test.
References: https://bugs.freedesktop.org/show_bug.cgi?id=92799
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Apparently the pthread shim on Android doesn't have pthread cancellation,
so use the plain old volatile to terminate the CPU hogs.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
One scenario under recent discussion is that of having a thundering herd
in i915_wait_request - where the overhead of waking up every waiter for
every batchbuffer was significantly impacting customer throughput. This
benchmark tries to replicate something to that effect by having a large
number of consumers generating a busy load (a large copy followed by
lots of small copies to generate lots of interrupts) and tries to wait
upon all the consumers concurrenctly (to reproduce the thundering herd
effect). To measure the overhead, we have a bunch of cpu hogs - less
kernel overhead in waiting should allow more CPU throughput.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
The main goal of this subtest is to trigger the following warning in
the function i915_gem_object_get_fence():
if (WARN_ON(!obj->map_and_fenceable))
To trigger this warning, the subtest first creates a Y-tiled object and
an associated framebuffer with the Y-fb modifier. Furthermore, to
prevent the map_and_fenceable from being set, we make sure that
the object does not have a normal VMA by refraining from rendering to the
object and by setting the rotation property upfront before calling commit.
v2: Do not call paint_squares and just use one output.
v3: Convert an if condition to igt_require and move the plane rotation
requirement further up before the fb allocation.
Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Signed-off-by: Vivek Kasireddy <vivek.kasireddy@intel.com>
Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
That option is not needed anymore since:
commit 982934625ac67234c6d85c6cf29a5a487e54d4f0
Author: Thomas Wood <thomas.wood@intel.com>
Date: Wed Sep 16 14:36:24 2015 +0100
lib: allow wildcard matching when specifying subtests
In fact, using "--run-subtest 'fbc-*'" is better than using --fbc-only
due to how SKIPs are handled. In the former, only the tests matching
the expression are tried, so the number of SKIPs only contains the
number of tests on the specified pattern that were skipped. If you
used --fbc-only, all the non-fbc tests would count as SKIPs, so it
would be harder to know which of the tests marked as skipped were
actual FBC tests.
Signed-off-by: Paulo Zanoni <paulo.r.zanoni@intel.com>
This is a corner case not exercised by the other subtests. The test is
expected to pass both with the current Kernel tree and with the
patches that are on the mailing list.
The patches currently on the mailing list change how the CFB is
allocated, and this subtest is designed to make sure everything still
works as expected.
Signed-off-by: Paulo Zanoni <paulo.r.zanoni@intel.com>
This fixes the failures for cases where you use --run-subtest to run
single subtests that don't use any drawing patterns.
Signed-off-by: Paulo Zanoni <paulo.r.zanoni@intel.com>
Don't use the rounded vrefresh info to predict the frame duration.
Instead calculate if from the clock.
Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
