Currently, the containers are used only to build IGT. We also want to
have an option to run IGT inside a container. For this, new build stage
is added.

Signed-off-by: Oleg Vasilev <oleg.vasilev@intel.com>

Base images rarely change, there is not much sense in rebuilding it on
every commit. GitLab already has mechanism for detecting such changes.
However, it is only able to prevent rebuilding whenever there is no
changes within the same ref. Since in our CI system, git tag is created
on every series, the mechanism doesn't work.

One possible way to workaround that is to compute a checksum of a
Dockerfile, and rebuilding only if there was no image built with the
same checksum.

Signed-off-by: Oleg Vasilev <oleg.vasilev@intel.com>

No functional changes. This order is consistent with the order of
execution stages.

Signed-off-by: Oleg Vasilev <oleg.vasilev@intel.com>

Currently the runner sees

Starting subtest: pipe-C-bad-aux-stride
Test requirement not met in function test_output, file ../tests/kms_ccs.c:324:
Test requirement: valid_tests > 0
CCS not supported, skippingSubtest pipe-C-bad-aux-stride: SKIP (0.330s)

and manages to implode.

Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Petri Latvala <petri.latvala@intel.com>

If the shared legacy ringbuffer is too small to fit 64 new contexts, it
will hang and report -EIO before we are expecting it. Accommodate this
eio as it part of the ABI we are testing.

Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Matthew Auld <matthew.auld@intel.com>

Since we don't any uAPI to retrieve the downstream's device's ksv,
we can't check the real revocation through SRM.

This test just writes the facsimile SRM into the /lib/firmware/
and check the kernel parsing of it by invoking the hdcp authentication.

v2:
  typos in commit msg is fixed. [Uma]
  file mode is added. [GG]
  reset test flag. [Uma]

Signed-off-by: Ramalingam C <ramalingam.c@intel.com>
Acked-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Reviewed-by: Uma Shankar <uma.shankar@intel.com>
Link: https://patchwork.freedesktop.org/patch/315470/?series=57756&rev=10

To detect the HDCP status change we are reading the uevent sent with the
corresponding connector id and property id.

This avoids the polling of the property every mSec.

v2:
  Made it as subtest. Remaining tests will use polling. [Daniel]
v3:
  failure to receive the uevent is handled as failure [Uma]
v4:
  For UNDESIRED state no uevent is expected.

Signed-off-by: Ramalingam C <ramalingam.c@intel.com>
Acked-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Reviewed-by: Uma Shankar <uma.shankar@intel.com>
Link: https://patchwork.freedesktop.org/patch/315470/?series=57756&rev=10

Testing the content type change when the content protection already
enabled.

v2:
  Rebased.
v3:
  Type 0 -> Type 1 is included.
  Type check is removed [Uma]

Signed-off-by: Ramalingam C <ramalingam.c@intel.com>
Acked-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Reviewed-by: Uma Shankar <uma.shankar@intel.com>
Link: https://patchwork.freedesktop.org/patch/315468/?series=57756&rev=10

To validate the teardown and rebuild of the interface between
I915 and mei hdcp this subtest execute the below sequence:
	1. Test HDCP2.2 Type 1 (expected to pass)
	2. Unload the mei_hdcp
	3. Test HDCP2.2 Type 1 (expected to fail)
	2. Reload the mei_hdcp
	1. Test HDCP2.2 Type 1 (expected to pass)

v2:
  Rebased.
v3:
  Rebased.

Signed-off-by: Ramalingam C <ramalingam.c@intel.com>
Acked-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Reviewed-by: Uma Shankar <uma.shankar@intel.com>
Link: https://patchwork.freedesktop.org/patch/315466/?series=57756&rev=10

Adds a connector property called "CP_Content_Type"

Content Type takes two values which classifies the content stream:
	Type 0: Stream that can be transmitted on HDCP1.4/HDCP2.2
	Type 1: Stream that needs HDCP2.2 encryption only.

So when Type 1 is set KMD is forced to enable HDCP2.2 only.

For Type 0 request, Kernel chooses the highest capable HDCP version
(v2.2) first. If that fails, then it fall back to the next available
version(v1.4) before abondoning HDCP authentication attempts.

Please note content_type is allowed to be updated when "Content
Protection" is in UNDESIRED state.

v2:
  s/cp_content_type/content_protection_type [danvet]
v3:
  s/content_protection_type/HDCP Content Type [Pekka]
v4:
  Rebased
v5:
  Typos in the commit msg is fixed [Uma]

Signed-off-by: Ramalingam C <ramalingam.c@intel.com>
Acked-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Reviewed-by: Uma Shankar <uma.shankar@intel.com>
Link: https://patchwork.freedesktop.org/patch/315467/?series=57756&rev=10

Existing Link integrity check test is moved into dedicated subtest.
This helps to reduce the execution time of other tests by removing the
repeated Link integrity check on every other tests.

v2:
  s/reduced/reduce [Uma]

Signed-off-by: Ramalingam C <ramalingam.c@intel.com>
Acked-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Reviewed-by: Uma Shankar <uma.shankar@intel.com>
Link: https://patchwork.freedesktop.org/patch/315465/?series=57756&rev=10

Considering increase of subtests for kms_content_protection, tests are
defined through flags.

v2:
  reset the cp_tests. [Uma]

Signed-off-by: Ramalingam C <ramalingam.c@intel.com>
Acked-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Reviewed-by: Uma Shankar <uma.shankar@intel.com>
Link: https://patchwork.freedesktop.org/patch/315464/?series=57756&rev=10

Use function that is meant to get the output instead of relying on a
magic for-loop.

Cc: Maarten Lankhorst <maarten.lankhorst@linux.intel.com>
Cc: Dhinakaran Pandiyan <dhinakaran.pandiyan@intel.com>
Signed-off-by: Arkadiusz Hiler <arkadiusz.hiler@intel.com>
Reviewed-by: Petri Latvala <petri.latvala@intel.com>

If CCS is not supported, i.e. we have executed no test for it, let's
skip with a message saying so instead of mysterious requirement on
test_output().

A lot of people get confused by current output thinking that there is
something wrong with output/modes.

v2: actually run tests for bad-pixel-format

Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=110580


Cc: Maarten Lankhorst <maarten.lankhorst@linux.intel.com>
Cc: Dhinakaran Pandiyan <dhinakaran.pandiyan@intel.com>
Signed-off-by: Arkadiusz Hiler <arkadiusz.hiler@intel.com>
Reviewed-by: Petri Latvala <petri.latvala@intel.com>

Check for viability of store-dword before use.

Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Andi Shyti <andi.shyti@intel.com>

Ashutosh Dixit authored 5 years ago and Chris Wilson committed 5 years ago

i915 module parameters "enable_guc_submission" and "enable_execlists"
have been discontinued since 2017 so remove them from IGT too.

Signed-off-by: Ashutosh Dixit <ashutosh.dixit@intel.com>
Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk>

Calls to igt_pipe_crc_drain + get_single are already replaced with
igt_pipe_crc_get_current() in patch
https://patchwork.freedesktop.org/patch/240651/?series=47129&rev=1


so removing igt_pipe_crc_drain() shouldn't matter since tests are using
igt_pipe_crc_get_current() which will drain the old junk anyway. Its
double to do drain + get_current(), so calls to drain() can be removed.

Signed-off-by: Swati Sharma <swati2.sharma@intel.com>
Signed-off-by: Ramalingam C <ramalingam.c@intel.com>
Reviewed-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/319789/?series=64210&rev=1

This helps to identify which entry is wrong in the table.

v2: Add back the values and missing \n (from Daniele)

Signed-off-by: Lucas De Marchi <lucas.demarchi@intel.com>
Reviewed-by: Daniele Ceraolo Spurio <daniele.ceraolospurio@intel.com>

Doing a fresh Ubuntu 18.10 install this dependencies were also
needed to build IGT.

Reviewed-by: Petri Latvala <petri.latvala@intel.com>
Signed-off-by: José Roberto de Souza <jose.souza@intel.com>

Gen 12 and later only have global MOCS: use the new offset.

Signed-off-by: Lucas De Marchi <lucas.demarchi@intel.com>
Reviewed-by: Daniele Ceraolo Spurio <daniele.ceraolospurio@intel.com>

Some small changes compared to Ice Lake, namely:

  - Removed entries 0 and 1: they are now reserved and shouldn't be used
  - Entries 16, 17, 48..61 were previously unused and are now available

Signed-off-by: Lucas De Marchi <lucas.demarchi@intel.com>
Reviewed-by: Daniele Ceraolo Spurio <daniele.ceraolospurio@intel.com>

In the Ice Lake table we are setting all undefined entries to PTE.
Undefined entries are undefined.

Signed-off-by: Lucas De Marchi <lucas.demarchi@intel.com>
Reviewed-by: Daniele Ceraolo Spurio <daniele.ceraolospurio@intel.com>

If the legacy ringbuffer doesn't support scheduling, it doesn't set the
cap! Hypothetically if one were to enable ringbuffer scheduling, this is
part of its testing suite.

Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>

The debugfs contains "Enabled SliceX subslices:" line only if a given
slice is enabled. If we try to dbg_get_status() with everything shut
down we will get a confusing assertion failure:

CRITICAL: Test assertion failure function dbg_get_int, file ../tests/i915/i915_pm_sseu.c:120:
CRITICAL: Failed assertion: pos != NULL

To fix that let's just read the enabled subslices if the line is there,

Cc: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=110593


Signed-off-by: Arkadiusz Hiler <arkadiusz.hiler@intel.com>
Acked-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>

Connector forcing works just fine even if it already has
something physically connected. So accept any HDMI connector
for the test.

Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Reviewed-by: Arkadiusz Hiler <arkadiusz.hiler@intel.com>
Reviewed-by: Simon Ser <simon.ser@intel.com>

Connector forcing works just fine even if it already has
something physically connected. So accept any HDMI connector
for the test.

Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Reviewed-by: Arkadiusz Hiler <arkadiusz.hiler@intel.com>
Reviewed-by: Simon Ser <simon.ser@intel.com>

On pre-HSW we have separate encoders for the DP and HDMI ports,
but both can't be enabled at the same time. The test fails to
account for that and can thus fail when the kernel rejects
the modeset. We can avoid that by turning off all the ports
beforehand.

Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=105595


Reviewed-by: Arkadiusz Hiler <arkadiusz.hiler@intel.com>
Reviewed-by: Simon Ser <simon.ser@intel.com>

On pre-HSW we have separate encoders for the DP and HDMI ports,
but both can't be enabled at the same time. The test fails to
account for that and can thus fail when the kernel rejects
the modeset. We can avoid that by turning off all the ports
beforehand.

v2: rebased (Arek)

Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Cc: Simon Ser <simon.ser@intel.com>
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=105594


Signed-off-by: Arkadiusz Hiler <arkadiusz.hiler@intel.com>
Reviewed-by: Simon Ser <simon.ser@intel.com>

As we never officially write to the scratch buffer, the kernel will
leave it in the CPU read domain upon execution. Our attempt to
invalidate the CPU cache on !llc is therefore skipped as the kernel
doesn't believe the backing store has been invalidated. Use a WC mmap to
avoid the CPU cache for readback, and add an extra sanity check that the
scratch buffer is found at the same location after execution. (This
sanity check does not affect the failure rate on bsw, that is only fixed
after realising that we do not clflush for the invalidate prior to the
read).

Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=111187


Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Matthew Auld <matthew.auld@intel.com>

__for_each_physical_engine already iterates over only engines that
exist, we don't need to check again that they exist.

Removing the check also makes their respective subtest_groups not
needed.

Signed-off-by: Petri Latvala <petri.latvala@intel.com>
Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Cc: Andi Shyti <andi.shyti@intel.com>
Tested-by: Andi Shyti <andi.shyti@intel.com>
Reviewed-by: Arkadiusz Hiler <arkadiusz.hiler@intel.com>

This is a performance test to highlight significant issues on Baytail,
for example, where using the GTT was much, much slower than it was
intended to be. This is not demonstrating a driver issue, but a HW one
and so does not need to be a foundational test in BAT.

Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Acked-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>

Cometlake is yet another Skylake refresh and used the same MOCS
registers.

Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=111137


Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Tvrtko Ursulin <tvrtko.ursulin@linux.intel.com>
Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@linux.intel.com>

Mihail Atanassov authored 5 years ago and Simon Ser committed 5 years ago

Not skipping the affected tests causes all the rest to fail on
non-Intel drivers.

Signed-off-by: Mihail Atanassov <mihail.atanassov@arm.com>
Reviewed-by: Simon Ser <simon.ser@intel.com>

Dhinakaran Pandiyan authored 5 years ago and Lucas De Marchi committed 5 years ago

Fix test to check for addfb failure instead.

Cc: Ville Syrjälä <ville.syrjala@linux.intel.com>
Signed-off-by: Dhinakaran Pandiyan <dhinakaran.pandiyan@intel.com>
Signed-off-by: Lucas De Marchi <lucas.demarchi@intel.com>
Reviewed-by: Ville Syrjälä <ville.syrjala@linux.intel.com>

Cc: Rodrigo Vivi <rodrigo.vivi@intel.com>
Signed-off-by: Lucas De Marchi <lucas.demarchi@intel.com>
Reviewed-by: Arkadiusz Hiler <arkadiusz.hiler@intel.com>

Javier Villavicencio authored 5 years ago and Lucas De Marchi committed 5 years ago

Cc: Rodrigo Vivi <rodrigo.vivi@intel.com>
Signed-off-by: Javier Villavicencio <javier.villavicencio@intel.com>
Signed-off-by: Lucas De Marchi <lucas.demarchi@intel.com>
Reviewed-by: Arkadiusz Hiler <arkadiusz.hiler@intel.com>

Straight copy from the kernel file, aligned with drm-intel-next-queued
commit cb823ed9915b ("drm/i915/gt: Use intel_gt as the primary object
for handling resets")

By Stuart: Add the mobile vs non-mobile pineview platforms to device
info to match what is in the kernel.

Signed-off-by: Lucas De Marchi <lucas.demarchi@intel.com>
Signed-off-by: Stuart Summers <stuart.summers@intel.com>
Reviewed-by: Arkadiusz Hiler <arkadiusz.hiler@intel.com>

There are few short windows of opportunity when watchdogs are primed but
there is no signal handling in place, so the process may exit without
proper shutdown sequence.

This patch rearranges the existing code so that we set up the signalfd
and BLOCK the signals before setting up watchdogs and UNBLOCK only after
the watchdogs are closed properly.

If igt_runner exits due to signal, non-zero status code is returned.

v2: more error handling and minor touch ups (Simon)

Cc: Petri Latvala <petri.latvala@intel.com>
Cc: Simon Ser <simon.ser@intel.com>
Signed-off-by: Arkadiusz Hiler <arkadiusz.hiler@intel.com>
Reviewed-by: Simon Ser <simon.ser@intel.com>
Reviewed-by: Petri Latvala <petri.latvala@intel.com>

instead of being closed normally on a graceful code path

Cc: Petri Latvala <petri.latvala@intel.com>
Signed-off-by: Arkadiusz Hiler <arkadiusz.hiler@intel.com>
Reviewed-by: Simon Ser <simon.ser@intel.com>
Reviewed-by: Petri Latvala <petri.latvala@intel.com>

Setting the watchdog fd lists to NULL for extra fireworks if accessed
unintentionally.

Cc: Petri Latvala <petri.latvala@intel.com>
Signed-off-by: Arkadiusz Hiler <arkadiusz.hiler@intel.com>
Reviewed-by: Simon Ser <simon.ser@intel.com>
Reviewed-by: Petri Latvala <petri.latvala@intel.com>