- Dec 27, 2021
-
-
Add support for local memory region (Device memory) Signed-off-by:
Ch Sai Gowtham <sai.gowtham.ch@intel.com> Cc: Zbigniew Kempczyński <zbigniew.kempczynski@intel.com> Reviewed-by:
Zbigniew Kempczyński <zbigniew.kempczynski@intel.com>
-
RPLS and ADLS use the same OA metrics. Add RPLS support for OA tests. Resolves: https://gitlab.freedesktop.org/drm/intel/-/issues/4808 Signed-off-by:
Tejas Upadhyay <tejaskumarx.surendrakumar.upadhyay@intel.com> Acked-by:
Ashutosh Dixit <ashutosh.dixit@intel.com>
-
- Dec 23, 2021
-
-
Bhanuprakash Modem authoredInstead of iterating through all the delta, break the execution when we found a match. This will optimize the time required to execute ctm-0-* tests. Cc: Karthik B S <karthik.b.s@intel.com> Cc: Uma Shankar <uma.shankar@intel.com> Signed-off-by:
Bhanuprakash Modem <bhanuprakash.modem@intel.com> Reviewed-by:
Karthik B S <karthik.b.s@intel.com>
-
Juha-Pekka Heikkilä authored
Create reference crc with same modifier as is used for testing. This will reduce non test related anomalies. Signed-off-by:
Juha-Pekka Heikkila <juhapekka.heikkila@gmail.com> Reviewed-by:
Ville Syrjälä <ville.syrjala@linux.intel.com>
-
- Dec 22, 2021
-
-
Swati2 Sharma authoredIn this patch, added upscaling tests 960x540->1920x1080. Existing downscaled tests reused for upscaling scenarios. Signed-off-by:
Swati Sharma <swati2.sharma@intel.com> Reviewed-by:
-
Swati2 Sharma authored
Signed-off-by:
-
- Dec 21, 2021
-
-
Eryk Brol authored
This commit adds a DP DSC test that checks: * Forces DSC on/off and ensures it is reset properly * Check DSC slice height property * Verify various DSC slice dimensions * Tests various combinations of link_rate + lane_count and logs if DSC enabled/disabled Tests different bpc settings and logs if DSC is enabled/disabled Change since V3: - Drop useless test Change since V2: - Remove IGT_CRTC_DSC_SLICE_HEIGHT crtc property from this commit Change since V1: - Rebase Cc: Harry Wentland <harry.wentland@amd.com> Cc: Nicholas Choi <Nicholas.Choi@amd.com> Cc: Mark Yacoub <markyacoub@chromium.org> Cc: Hayden Goodfellow <hayden.goodfellow@amd.com> Cc: Hersen Wu <hersenxs.wu@amd.com> Cc: Roman Li <roman.li@amd.com> Acked-by:
Rodrigo Siqueira <Rodrigo.Siqueira@amd.com> Signed-off-by:
Mikita Lipski <mikita.lipski@amd.com> Signed-off-by:
Eryk Brol <eryk.brol@amd.com>
-
Tejas Upadhyay authored
Adding Alder lake N platform definitions Reviewed-by:
Petri Latvala <petri.latvala@intel.com> Signed-off-by:
-
Tejas Upadhyay authored
This synchronizes with kernel commit 7e28d0b26759 ("drm/i915/adl-n: Enable ADL-N platform") to bring in ADL-N PCI IDs. Acked-by:
-
- Dec 20, 2021
-
-
Add support for local memory region (Device memory) Signed-off-by:
-
- Dec 17, 2021
-
-
Adding Raptor lake platform definitions Reviewed-by:
Raviteja Goud Talla <ravitejax.goud.talla@intel.com>
-
- Dec 16, 2021
-
-
Tejas Upadhyay authored
This synchronizes with kernel commit 52407c220c44 ("drm/i915/rpl-s: Add PCI IDS for Raptor Lake S") to bring in RPL-S PCI IDs. Acked-by: -
Petri Latvala authored
Even though the best practices for dynamic subtests is to not execute the dynamic subtest at all instead of skipping, there are a couple of tests that to the opposite, without an easy way to construct them along best practices. Accomodate this with fixing the handling of skipping dynamic subtests so that the result of the containing subtest becomes a skip also if all dynamic subtests skip, not just when there are no dynamic subtests. Signed-off-by:
-
- Dec 15, 2021
-
-
Abhinav Kumar authored
clear_yuv_buffer() checks the size of the plane_size[] to make sure its greater than the number of planes to avoid overflows. The plane_size[] is fixed to two currently. However some of the formats like YV12 indeed have more than 2 planes in the format_desc[] hence this incorrectly failing this check. Increase the size of the plane_size[] to match the correct max number of planes. changes in v3: - use drm_format->num_planes to create the plane_size - add assert checks for num of planes before memset Signed-off-by:
Abhinav Kumar <quic_abhinavk@quicinc.com> Reviewed-by:
-
Tejas Upadhyay authored
Bonding is disabled on GEN12+ platforms, except ones already supported by the i915 - TGL, RKL, and ADL-S. Currently few tests are failing for same reason, with this change they should rightly skip with proper message. Reference: https://cgit.freedesktop.org/drm-tip/commit/?id=ce7e75c7ef1bf8ea3d947da8c674d2f40fd7d734 Changes since V3 : - Initalization of bonding boolean not required - zbigniew Changes since V2 : - Solved checkpatch errors. Changes since V1 : - use function to check bonding capability so that we dont skip whole test in case bonding not supported. Acked-by:
-
Zbigniew Kempczyński authoredAs Ashutosh noticed I've handled errors from the kernel too wide packing it to same error bag and returning system memory region in this case. That's of course is wrong and can lead to return system memory on discrete if invalid arguments would be passed to the query. Return previous behavior of query memory regions handling -ENODEV path individually. For this error lets assume we got kernel which doesn't support this query yet so returning system memory region is a reasonable choice because this region exists for all of i915 gens. Signed-off-by:
-
- Dec 13, 2021
-
-
Zbigniew Kempczyński authored
For stable kernels we get -ENODEV on integrated where gem_create_ext() call tries to allocate bo in system memory. Instead of asserting this case we can fallback to gem_create() as this allocation can be resolved using legacy call. Signed-off-by:
-
Zbigniew Kempczyński authored
For stable kernels we can currently get -ENODEV on integrated gens when no CONFIG_DRM_I915_UNSTABLE_FAKE_LMEM is set. Provide system memory region stub to avoid failing on querying kernel for memory regions. v2: use offsetof() instead of manual size calculation Signed-off-by:
-
Some simple test cases to use FB_DAMAGE_CLIPS plane property. Issue a plane update with damage with a clips on different scenarios and verify the state. v2: use drm_mode_rect instead of local struct and array instead of malloc. Signed-off-by:
Deepak Rawat <drawat@vmware.com> Signed-off-by:
Jeevan B <jeevan.b@intel.com> Reviewed-by:
Jouni Högander <jouni.hogander@intel.com>
-
Change testcase design so that it avoids new subtest for each block position. Signed-off-by:
-
rename igt_debug_manual_check and patch the igt function igt_debug_wait_for_keypress() to assert if "all" is supplied. v2: calling igt_debug_wait_for_keypress() with "all" will assert. v3: Change igt_assert to igt_assert_f for adding clear log message. v4: Rebase only. Signed-off-by:
-
- Dec 10, 2021
-
-
The reason for this is that it doesn't work with the dma-resv-object locking introduced recently in the code,and it's not going to be fixed anytime soon. drm/i915: Perform execbuffer object locking as a separate step To help avoid evicting already resident buffers from the batch we're processing, perform locking as a separate step. Link:https://cgit.freedesktop.org/drm-tip/commit/?id=5cd57f676bb946a00275408f0dd0d75dbc466d25 Signed-off-by:
Mastan Katragadda <mastanx.katragadda@intel.com> Reviewed-by:
Thomas Hellström <thomas.hellstrom@linux.intel.com>
-
This test makes assumptions about the backend scheduling algorithm that a real world UMD would never assume. This test is not testing a UMD-KMD contract, rather a specific backend.It is an invalid test case thus we are skipping. Changes Since v1: -Updated commit message Cc: Matthew Brost <matthew.brost@intel.com> Signed-off-by:Matthew Brost <matthew.brost@intel.com>
-
- Dec 09, 2021
-
-
Mark Yacoub authored
[Why] flip_vs_cursor_crc and flip_vs_cursor_busy_crc use CRC in debugfs which is not supported by some drivers. [How] igt_require_pipe_crc in igt_fixture before the subtests start. Test: igt@kms_cursor_legacy@flip-vs-cursor{-busy}-crc-{atomic,legacy} on ChromeOS Jacuzzi (MTK) Signed-off-by:Mark Yacoub <markyacoub@chromium.org> Reviewed-by:
-
- Dec 07, 2021
-
-
Rob Clark authored
Tests that reads and/or writes to kernel managed buffers which should be inaccessible to userspace controlled cmdstream, are indeed inaccessible. Signed-off-by:
Rob Clark <robdclark@chromium.org> Reviewed-by:
-
Rob Clark authored
Split the readN()/writeN() helpers out into an igt_io module, so they can be re-used by tests. Signed-off-by: -
Mark Yacoub authored
[Why] Mediatek devices have a HW issue with sending their vblank IRQ at the same time interval everytime. The drift can be below or above the expected frame time, causing the timestamp to drift with a relatively larger standard deviation over a large sample. [How] Filter out the flags TEST_CHECK_TS and TEST_VBLANK_EXPIRED_SEQ from the tests flags, and restrict sequence and ts checks. Tested on Jacuzzi (MT8183) Signed-off-by:
-
Stylon Wang authored
Acked-by:
Stylon Wang <stylon.wang@amd.com>
-
Anson Jacob authoredCouple of tests were accessing pipe's without ensuring if they are enabled or not. Using for_each_pipe macro to do so. Acked-by:
Anson Jacob <Anson.Jacob@amd.com>
-
bhawanpreet lakha authored
[Why] Full screen scaling is a common usecase [How] Rework the scaling tests to add a 'fullscreen' test case. Change the test_plane function to use "dst_w/h" instead of a scale multiplier. This way we can easily add a full screen scale multiplier. Acked-by:
Bhawanpreet Lakha <Bhawanpreet.Lakha@amd.com>
-
Victor Lu authored
[Why] The MPO scale and pan subtests is valid with one display and our current CI setup only has one display. [How] Change the MPO scale and pan subtests to be single display and add "-multihead" subtests for the two for the original dual display config Acked-by:
Victor Lu <victorchengchi.lu@amd.com>
-
bhawanpreet lakha authored
[Why] Resizing MPO window is a common usecase (video in windowed mode) Our MPO feature has scaling limitations (0.25x -> 16x). So this test creates a window of various reolutions and scales them between 0.25x -> 16x [How] The test creates common video sizes and scales them from (0.30x->16x). The lower limit is 0.25 but some combination of configs fail at 0.25x so 0.30 is used as the starting point. CRTC#1 CRTC#1 +--------------------+ +---------------------+ | +-------+ | | +------------+ | | |Primary| | | |Primary | | | +-------+ | --> | | | | | | scale | | | | | | | +------------+ | |overlay | |Overlay | +--------------------+ +---------------------+ Acked-by:
-
bhawanpreet lakha authored
[Why] In a multi display MPO config the MPO window is dragged from one display to the other. In this usecase there is an instance when the overlay is moved from one display to the next. This test recreates the scenario (MPO swap) [How] Overlay should switch to the display that has >50% of the video. This test was created with this asumption. The test creates common video sizes and swaps a window of size (w/2,h) between 2 displays bezzels at the top,middle and bottom of the display. CRTC#1 CRTC#2 CRTC#1 CRTC#2 +-------------+------------+ +-------------+------------+ | | | | | | | +--------+ | | +--------+ | | |Primary | Primary | <=> | Primary |Primary | | | | | | Swap | | | | | +--------+ | | +--------+ | |overlay | | | |Overlay | +-------------+------------+ +-------------+------------+ Acked-by:
-
bhawanpreet lakha authored
[Why] Moving the MPO window is a common usecase (video window being moved). This test creates a MPO window and moves it horizontally, vertically and diagonally [How] Create common video resolution windows and move them across the displays. Test pass is determined by getting a crc of window location using a single plane, then recreating this scenario using overlay. Test panes from Start->End1, Start->End2 and Start->End3 CRTC#1 Start +--------+--------+--------+ |Primary | --> |Primary | End1 +--------+ +--------+ | | \ Overlay | | v * | +--------+ \ +--------+ |Primary | |Primary | End2 +--------+--------+--------+ End3 Acked-by: -
Victor Lu authored
[why] Older ASIC does not support multi-head MPO [how] Rename current mpo-swizzle-toggle subtest to mpo-swizzle-toggle-multihead. Change mpo-swizzle-toggle to only use one output. Acked-by:
-
Ahmad Othman authored
[Why] The test wasn't working on some setups and needs fixing [How] Added output pipe initialization Added checks for both 4k and 2160p Acked-by:
Ahmad Othman <ahmad.othman@amd.com>
-
Nicholas Kazlauskas authored
Sequence captured on ChromeOS - we experience underflow when toggling between swizzle modes on the primary plane with MPO enabled. Add a new subtest replicating that sequence, "mpo-swizzle-toggle". Converts the existing framework over to the new multi-display framework since that's the configuration where this was found. Acked-by:
Nicholas Kazlauskas <nicholas.kazlauskas@amd.com>
-
Victor Lu authored
Tests MPO using a primary plane and an overlay plane with a transparent cutout. Acked-by:
-
- Dec 03, 2021
-
-
From Haswell up until gen12, only rcs supports all the registers and MI commands required to build the wait batch; before Haswell MI does not support the required MATH operations nor the predication. However, after gen12 we may not always have rcs available and so require to run the delay on other command streamers, which requires us to use the appropriate mmio base for the alternate engine. Signed-off-by:
Chris Wilson <chris.p.wilson@intel.com> Reviewed-by:
-
- Dec 02, 2021
-
-
Reading of prev_dc counter in dc9-dpms subtest as 0, after reset comparison with 0 test gets failed. taken prev_dc counter after enterd into shallow DC state.fixes dc9-dpms subtest failures. changes since V1: - Removed new lines and warnings from checkpatch.pl - Updated Subject Signed-off-by:Jigar Bhatt <jigar.bhatt@intel.com>
-
