| Age | Commit message (Collapse) | Author |
|---|
|
Signed-off-by: Petri Latvala <petri.latvala@intel.com>
Acked-by: Daniel Vetter <daniel.vetter@ffwll.ch>
|
|
Signed-off-by: Petri Latvala <petri.latvala@intel.com>
Acked-by: Daniel Vetter <daniel.vetter@ffwll.ch>
|
|
media_fill.c has been unused from around 2014 when its contents moved
to intel_batchbuffer.c.
Signed-off-by: Petri Latvala <petri.latvala@intel.com>
Acked-by: Daniel Vetter <daniel.vetter@ffwll.ch>
|
|
demos/Makefile.am was left behind. Nuke it.
Signed-off-by: Petri Latvala <petri.latvala@intel.com>
Acked-by: Daniel Vetter <daniel.vetter@ffwll.ch>
|
|
Signed-off-by: Petri Latvala <petri.latvala@intel.com>
Acked-by: Daniel Vetter <daniel.vetter@ffwll.ch>
|
|
Signed-off-by: Petri Latvala <petri.latvala@intel.com>
Acked-by: Daniel Vetter <daniel.vetter@ffwll.ch>
|
|
Most of these tests have no reason to look at those members,
so try other ways of getting the information.
Signed-off-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com>
Reviewed-by: Mika Kahola <mika.kahola@intel.com>
|
|
This test was flipping the primary plane instead of the sprite plane.
Flip the correct plane to make the test pass properly.
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=102691
Signed-off-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com>
Reviewed-by: Mika Kahola <mika.kahola@intel.com>
Tested-by: Marta Lofstedt <marta.lofstedt@intel.com>
Acked-by: Marta Lofstedt <marta.lofstedt@intel.com>
|
|
Rename kms_pipe_color to kms_color, and rename the invalid tests to pipe-invalid.
To prepare for adding support for plane color management.
Signed-off-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com>
Reviewed-by: Mika Kahola <mika.kahola@intel.com>
|
|
Hide the initial setup cost of a new batch by performing it before we
start the clock for measuring the execute-wait latency.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
|
|
Adding LIBDRM_INTEL_BENCHMARKS into benchmarks_PROGRAMS had a tab
before the assignment, making the line be inserted into the Makefile
as is instead of being evaluated by automake.
Signed-off-by: Petri Latvala <petri.latvala@intel.com>
Reviewed-by: Arkadiusz Hiler <arkadiusz.hiler@intel.com>
|
|
Printf family functions return a number of bytes, not tokens,
printed so the existing check (== 1) was wrong.
Signed-off-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk>
|
|
Measure how many high priority batches we can execute whilst running a
bg spinner.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
|
|
Check and measure how well we can submit a second high priority task
when the engine is already busy with a low priority task and see how
long it takes to complete (and wake up the client).
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
|
|
Using vgem as our cork for building the request queue limits us to 10s
of setup (or else the fence autoexpires and we start executing too
early). Add timeouts to the setup loops and SKIP if we cannot establish
the workload within 10s, the machine and driver is too slow to evaluate
the expected results.
To avoid the artificial limit of 10s requires lifting the dependency on
vgem and switching to sw_sync and explicit fencing. Just a matter of
plumbing!
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
|
|
Signed-off-by: Petri Latvala <petri.latvala@intel.com>
Reviewed-by: Daniel Vetter <daniel.vetter@ffwll.ch>
|
|
Either source or build directory is required as a command line
parameter.
Also use mkdir -p when creating the build directory to avoid errors
when it already exists.
Signed-off-by: Petri Latvala <petri.latvala@intel.com>
Reviewed-by: Daniel Vetter <daniel.vetter@ffwll.ch>
|
|
We have no means to check write only registers as
this would need access through context image. For now we
know that cnl has a one such register, 0xe5f0 which is used
to set WaForceContextSaveRestoreNonCoherent:cnl. By inspecting
the context image without and with workaround applied:
0x0000a840: 0x0000e5f0 0xffff0800
0x0000a840: 0x0000e5f0 0xffff0820
we can conclude that the workaround setup is working right
in this particular case.
Add a write only list and add register 0xe5f0 into this list.
This is a temporary solution until a more capable context image
checker emerges.
v2: add code comment about adhocness (Petri)
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=102943
Cc: Oscar Mateo <oscar.mateo@intel.com>
Cc: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Rodrigo Vivi <rodrigo.vivi@intel.com>
Cc: Petri Latvala <petri.latvala@intel.com>
Signed-off-by: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Petri Latvala <petri.latvala@intel.com>
Acked-by: Oscar Mateo <oscar.mateo@intel.com>
|
|
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Matthew Auld <matthew.auld@intel.com>
Reviewed-by: Matthew Auld <matthew.auld@intel.com>
|
|
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
|
|
Michal wants to limit machines that can do preemption, which means that
we no longer can assume that if we have a scheduler for execbuf, that
implies we have preemption.
v2: Try a capability mask instead
v3: Pretty print the caps.
v4: Not so pretty print the caps.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
|
|
Dump debugfs/i915_error_state to the debug channel if we detect an ERROR
uevent. This poses a few problems, not least that it is the auxiliary
process doing the dumping (so the output may be interleaved with the
test, but considering a hang occurred it is likely the test is blocked)
and the average error state is around 60k, which may prove unwieldy!
On the other hand, it may prove invaluable in debugging those impossible
to reproduce bugs.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Petri Latvala <petri.latvala@intel.com>
|
|
When plugging the device, we need to submit batches at highest priority
so that they cannot be gazumped by the queued requests. On older kernels
that do not support the user changing context priority, all contexts
therefore have the same default priority and we can ignore the error.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Michał Winiarski <michal.winiarski@intel.com>
|
|
[[ a != b ]] is a bashism. As it's just comparing $1 to an empty
string, use -n with a normal [ ].
/bin/sh is dash in CI.
v2: Also change the script to #!/bin/bash to avoid having to babysit
bashism later.
Fixes: f0243a761f1b ("tests/igt_command_line.sh: Allow testing individual tests")
CC: Daniel Vetter <daniel.vetter@ffwll.ch>
Signed-off-by: Petri Latvala <petri.latvala@intel.com>
Acked-by: Daniel Vetter <daniel.vetter@ffwll.ch>
|
|
Run ./meson.sh once, then you have
$ make
and
$ make test
available in the normal src root.
v2:
Add
$ make reconfigure
which is the meson equivalent to rerunning ./configure. Also takes
some arguments if needed. Start out with --help, as usual.
v3: Use ninja -C (Chris).
v4: Catch more automake targets and point out what's happening
(Petri).
v5: Use EOF as heredoc marker (Arek)
Cc: Petri Latvala <petri.latvala@intel.com>
Cc: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Arkadiusz Hiler <arkadiusz.hiler@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
|
|
No complex check of reordered writes, just throwing lots of random
priority batches with long dependency chains at the scheduler. This
should help to smoketest preemption.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
|
|
Like wide(), reorder-wide() didn't check the ring size before flooding
and so would exhaust its available space and block (causing a GPU hang
to recover), thus failing the test. Also, since we use a new context for
each iteration of the test, the available ring space is reduced (due to
the overhead in setting up that context via the ring), so make sure that
when we measure the available ring space we also use a new context.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
|
|
Replace instances of -MAX_PRIO with MIN_PRIO for readability.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
|
|
Instead of redirecting output to pipes and forking, redirect after
forking to avoid having to carefully unredirect before logging
anything.
igt@tools_test@sysfs_l3_parity had a racy condition where it prints
the output of intel_l3_parity prepended by [cmd], but that ended up
being printed again prepended by [cmd] because output was redirected,
causing outputs to appear multiple times. This patch fixes that.
CC: Abdiel Janulgue <abdiel.janulgue@linux.intel.com>
Signed-off-by: Petri Latvala <petri.latvala@intel.com>
Reviewed-by: Abdiel Janulgue <abdiel.janulgue@linux.intel.com>
|
|
Multiple misunderstandings of the expectations of the test and some
missed parts of the shell-to-c conversion caused a couple of issues to
remain.
First, we need to actually disable a subbank before we check that a
subbank is disabled (invoke the tool with -d).
Second, the original pipe to wc -l was to check that the tool prints
only one line, not that it prints "at least" a line. Modify the last
check to verify that an "is disabled" text is _not_ printed.
v2: Add a TODO comment
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=101650
Signed-off-by: Petri Latvala <petri.latvala@intel.com>
Reviewed-by: Abdiel Janulgue <abdiel.janulgue@linux.intel.com>
|
|
Remove the local recursive spinner in favour of igt_spin_t.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
|
|
Switched to strncmp.
Signed-off-by: Dhinakaran Pandiyan <dhinakaran.pandiyan@gmail.com>
Reviewed-by: Radhakrishna Sripada <radhakrishna.sripada@intel.com>
Signed-off-by: Rodrigo Vivi <rodrigo.vivi@intel.com>
|
|
The substring to be matched is "DRRS Supported: Yes" and not
"DRRS Support: No". The kernel prints the DRRS status for all active
CRTC's. Matching for No support gives a false positive when multiple
CRTC's are active.
Fixes: 33355210a43e (igt/kms_psr_sink_crc: Add psr_drrs subtest)
Cc: C. Ramalingam <ramalingam.c@intel.com>
Cc: Lohith BS <lohith.bs@intel.com>
Cc: Rodrigo Vivi <rodrigo.vivi@intel.com>
Cc: Dhinakaran Pandiyan <dhinakaran.pandiyan@intel.com>
Signed-off-by: Radhakrishna Sripada <radhakrishna.sripada@intel.com>
Reviewed-by: Rodrigo Vivi <rodrigo.vivi@intel.com>
Signed-off-by: Rodrigo Vivi <rodrigo.vivi@intel.com>
|
|
Switch to using igt_cairo_image_surface_create_from_png_file() over the
raw cairo version so that the test can actually find the image file
no matter where we run it from.
Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Reviewed-by: Petri Latvala <petri.latvala@intel.com>
|
|
Raw usage of cairo_image_surface_create_from_png() doesn't work
since it doesn't know about IGT_DATADIR and IGT_SRCDIR. Let's extract
the helper from igt_paint_image() that uses igt_fopen_data() +
cairo_image_surface_create_from_png_stream() and call it
igt_cairo_image_surface_create_from_png().
Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Reviewed-by: Petri Latvala <petri.latvala@intel.com>
|
|
Added the missing IGT_TEST_DESCRIPTION and some subtest
descriptions.
v2: Removed duplication, addressed comments, cc'd test author
v3: Only comment abstract code, change some igt_info to igt_debug.
Changed description to reflect this is a patch, not an RFC.
Cc: Michał Winiarski <michal.winiarski@intel.com>
Cc: Eric Anholt <eric@anholt.net>
Cc: Arkadiusz Hiler <arkadiusz.hiler@intel.com>
Cc: Daniel Vetter <daniel.vetter@intel.com>
Signed-off-by: Vinay Belgaumkar <vinay.belgaumkar@intel.com>
Reviewed-by: Radoslaw Szwichtenberg <radoslaw.szwichtenberg@intel.com>
|
|
Signed-off-by: Petri Latvala <petri.latvala@intel.com>
Cc: Tomi Sarvela <tomi.p.sarvela@intel.com>
Cc: Arkadiusz Hiler <arkadiusz.hiler@intel.com>
Reviewed-by: Arkadiusz Hiler <arkadiusz.hiler@intel.com>
Tested-by: Arkadiusz Hiler <arkadiusz.hiler@intel.com>
|
|
Piglit expects test-list*.txt files to be in the same directory as the
test binaries. Installing non-executable files to libexecdir requires
going around some of autotools' sanity checks.
This change should make 'make install':ed IGT runnable with piglit.
Signed-off-by: Petri Latvala <petri.latvala@intel.com>
Cc: Tomi Sarvela <tomi.p.sarvela@intel.com>
Cc: Arkadiusz Hiler <arkadiusz.hiler@intel.com>
Reviewed-by: Arkadiusz Hiler <arkadiusz.hiler@intel.com>
Tested-by: Arkadiusz Hiler <arkadiusz.hiler@intel.com>
|
|
In sync with 41693fd52373 ("drm/i915/kbl: Change a KBL pci id
to GT2 from GT1.5")
"See Mesa commit 9c588ff"
v2: s/DT/Mobile
Cc: Anuj Phogat <anuj.phogat@gmail.com>
Signed-off-by: Rodrigo Vivi <rodrigo.vivi@intel.com>
Reviewed-by: Anuj Phogat <anuj.phogat@gmail.com>
|
|
The batch is now on the other end of the list.
Fixes issues where modern Mesa and modern kernels together start putting
the batch at the front of the list, and intel_aubdump looks for it at
the end of the list, causing it to interpret some other buffer as the
batch. Then AubLoad or aubinator see bogus data like 1.0 float as a
GPU command and get grumpy.
|
|
We don't expect every machine to be able to pass the WC/GTT coherency
check, see
kernel commit 3b5724d702ef24ee41ca008a1fab1cf94f3d31b5
Author: Chris Wilson <chris@chris-wilson.co.uk>
Date: Thu Aug 18 17:16:49 2016 +0100
drm/i915: Wait for writes through the GTT to land before reading back
If we quickly switch from writing through the GTT to a read of the
physical page directly with the CPU (e.g. performing relocations through
the GTT and then running the command parser), we can observe that the
writes are not visible to the CPU. It is not a coherency problem, as
extensive investigations with clflush have demonstrated, but a mere
timing issue - we have to wait for the GTT to complete it's write before
we start our read from the CPU.
The issue can be illustrated in userspace with:
gtt = gem_mmap__gtt(fd, handle, 0, OBJECT_SIZE, PROT_READ | PROT_WRITE);
cpu = gem_mmap__cpu(fd, handle, 0, OBJECT_SIZE, PROT_READ | PROT_WRITE);
gem_set_domain(fd, handle, I915_GEM_DOMAIN_GTT, I915_GEM_DOMAIN_GTT);
for (i = 0; i < OBJECT_SIZE / 64; i++) {
int x = 16*i + (i%16);
gtt[x] = i;
clflush(&cpu[x], sizeof(cpu[x]));
assert(cpu[x] == i);
}
Experimenting with that shows that this behaviour is indeed limited to
recent Atom-class hardware.
so split out the interleave coherency check from the basic
interopability check.
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=102577
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Michał Winiarski <michal.winiarski@intel.com>
|
|
Suggested by Jani. And rename from config_h to plain config, to make
it's multi-use character a bit more obvious.
Cc: Jani Nikula <jani.nikula@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@intel.com>
Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
|
|
- I forgot the chamelium tests
- Order tests the same way in both build systems. Since testdisplay is
special, it's easier to put that at the end in meson, so adjusted
automake to suit.
With this you can diff the 2 test lists and end up with 0 differences,
which will be useful to CI meson vs. automake.
Signed-off-by: Daniel Vetter <daniel.vetter@intel.com>
Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
|
|
Collecting CRC may force a modeset, which is a bad idea after we just
forced a hang. The hang is intended to make sure the page flip doesn't
complete before the cursor, making sure that works.
Signed-off-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com>
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=102259
Cc: Marta Lofstedt <marta.lofstedt@intel.com>
Tested-by: Marta Lofstedt <marta.lofstedt@intel.com>
Reviewed-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
[mlankhorst: Prevent overflow when collecting crcs, based on Villes feedback]
|
|
Break up request coalescing by submitting a queue of requests from
different contexts. This makes sure that we mark all ready but not yet
submitted-to-hw requests as completed/EIO upon wedging.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Michał Winiarski <michal.winiarski@intel.com>
Reviewed-by: Michał Winiarski <michal.winiarski@intel.com>
|
|
Pull the two subtests together into the same subtest group as the
fixtures were repeated for both and so we can consolidate down to one.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Reviewed-by: Mika Kuoppala <mika.kuoppala@linux.intel.com>
|
|
Some overlap with gem_exec_fence, but confirm that light for waiting on
a fence (both native and external), that a request queued is also
flagged as EIO upon wedging.
v2: check all ABI engines
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Michał Winiarski <michal.winiarski@intel.com>
|
|
If we wedged one engine with unready requests to a second engine
(blocked by waiting on requests from the first, using a dma-fence),
check that we propagate the -EIO to those in-flight requests.
v2: check all ABI engines
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Michał Winiarski <michal.winiarski@intel.com>
|
|
As our hangcheck may exceed 10s to declare the device wedged, we need to
hold the plugging fence indefinitely. This makes using vgem as our input
fence unusable, so resort to using sw_sync. At the same time, we can
then check that the async result is also -EIO.
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=102616
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Michał Winiarski <michal.winiarski@intel.com>
|
|
Platforms do not support psr and drrs simultaneously.
Adding a subtest to make the check.
Cc: Rodrigo Vivi <rodrigo.vivi@intel.com>
Cc: Daniel Vetter <daniel.vetter@ffwll.ch>
Signed-off-by: Radhakrishna Sripada <radhakrishna.sripada@intel.com>
|
