igt-gpu-tools.git/docs, branch vm-bind

lib/i915_crc: Introduce crc32 on gpu for DG2

2022-06-13T15:28:24+00:00

Adding crc32 calculation on gpu gives us new possibility to verify data integrity without relying on trust cpu mapping is correct. Patch introduces calculating crc32 on DG2 only. On older gens ALU (MI_MATH) doesn't support bit-shifting instructions as well as multiply or divide. Emulating n-bit shifts cost hundred of instructions with predicated SRM (works on render engine only). Another limitation is lack of indexed load / store. On DG2 we can use WPARID and CS_MI_ADDRESS_OFFSET to achieve indexed operation on memory. Due to performance reasons (cpu crc32 calculation even on WC memory is still much faster than on gpu, also depends on calculated object memory region) calculation will complete in reasonable of time only for few MiB. v2: - use registers relative to engine to allow run on all engines (Chris) - use predication instead of memory access to get better performance (Chris) - add location where crc32 implementation comes from (Petri) v4: - use common crc32 table from igt_crc - add docs v5: - change BIT(n) to informative macros (Zbigniew) Signed-off-by: Zbigniew Kempczyński Acked-by: Petri Latvala

lib/igt_crc: Introduce common place for crc tables and functions

2022-06-13T15:28:24+00:00

Add crc32 table for on-cpu crc calculation function. Other tables and algorithms should be added here allowing reuse tables for in-gpu crc calculation. Signed-off-by: Zbigniew Kempczyński Reviewed-by: Petri Latvala

tests/kms_chamelium: Check if port adapters are in use

2022-05-25T09:52:51+00:00

If a DUT has Chamelium ports connected via an adapter (for example, DP on the Chamelium side -> DP-HDMI adapter -> HDMI on the DUT), this will usually cause many tests to fail. If mismatching port types are found on both sides, the tests will now be aborted with a warning. This behavior can be overridden with a new AdapterAllowed config value, which must be set in [Chamelium:PORT] blocks in .igtrc. Signed-off-by: Ryszard Knop Reviewed-by: Petri Latvala

docs/code_coverage.md: document the code coverage filter script

2022-04-14T15:19:39+00:00

Add documentation bits for the code_coverage_parse_info. While here, also drop the extensions from the scripts that were renamed. Reviewed-by: Ch Sai Gowtham Reviewed-by: Andrzej Hajda Signed-off-by: Mauro Carvalho Chehab

docs: add documentation for code coverage

2022-03-21T16:34:59+00:00

Document the IGT runner features related to code coverage data capture. Acked-by: Petri Latvala Signed-off-by: Mauro Carvalho Chehab

lib/i915_blt: Add library for blitter

2022-03-11T17:00:51+00:00

Blitter commands became complicated thus manual bitshifting is error prone and hard debugable - XY_BLOCK_COPY_BLT is the best example - in extended version (for DG2+) it takes 20 dwords of command data. To avoid mistakes and dozens of arguments for command library provides input data in more structured form. Currently supported commands: - XY_BLOCK_COPY_BLT: a) TGL/DG1 uses shorter version of command which doesn't support compression b) DG2+ command is extended and supports compression - XY_CTRL_SURF_COPY_BLT - XY_FAST_COPY_BLT Source, destination and batchbuffer are provided to blitter functions as objects (structs). This increases readability and allows use same object in many functions. Only drawback of such attitude is some fields used in one function may be ignored in another. As an example is blt_copy_object which contains a lot of information about gem object. In block-copy all of data are used but in fast-copy only some of them (fast-copy doesn't support compression). v2-v3: address review comments (Kamil) Signed-off-by: Zbigniew Kempczyński Reviewed-by: Kamil Konieczny

lib/i915/gem_create: Introduce gem-pool bo cache

2022-03-11T17:00:51+00:00

Handling batchbuffers with softpin requires tracking its state otherwise we can write to inflight batchbuffer and encounter gpu hang. Gem pool adds such tracking (similar to libdrm bo cache) and provides free and ready to use bo. If pool has no free bo new one is created what means pool can be growing during test execution. When test completes freeing buffers and memory is called from igt_core so no additional cleanup is necessary. Signed-off-by: Zbigniew Kempczyński Cc: Petri Latvala Cc: Kamil Konieczny Reviewed-by: Kamil Konieczny

igt: Split out I/O helpers

2021-12-07T22:06:42+00:00

Split the readN()/writeN() helpers out into an igt_io module, so they can be re-used by tests. Signed-off-by: Rob Clark

msm: Add helper library

2021-09-09T05:27:14+00:00

Handle some of the boilerplate for tests. v2: fix comment, drop unnecessary drm_open_driver_render() error handling, docs fixes, drop igt_msm_pipe_get_param() (for now), handle NULL in destructors Signed-off-by: Rob Clark Reviewed-by: Petri Latvala

lib: Add an intel_ctx wrapper struct and helpers (v6)

2021-06-10T16:13:20+00:00

We're trying to clean up some of our technical debt in the i915 API. In particular, context mutability and unnecessary getparam(). There's quite a bit of the introspection stuff that's not used by any userspace other than IGT. Most drivers don't care about fetching the set of engines, for instance, because they don't forget about what set of engines they asked for int the first place. Unfortunately, IGT relies heavily on context introspection for just about everything when it comes to multi-engine testing. It also likes to use ctx0 as temporary storage for whatever the current test config is. While effective at keeping IGC simple in some ways, this means we're making heavy use of context mutability. Also, passing data around with in tests isn't really what contexts are for. This patch adds a new intel_ctx_t struct which wraps a context and remembers the full context configuration. This will provide similar ease-of-use without having use ctx0 as temporary storage. v2 (Jason Ekstrand): - Make all intel_ctx_t's const v3 (Jason Ekstrand): - Fix up the docs so they build properly v4 (Jason Ekstrand): - Add an intel_ctx_create_for_engine helper v5 (Zbigniew Kempczyński): - Use SPDX license identifiers - Document default context semantics v6 (Ashutosh Dixit): - Fix SPDX in intel_ctx.h - Fix a typo in a comment v6 (Jason Ekstrand): - Add documentation about num_engines to intel_ctx_cfg_t Signed-off-by: Jason Ekstrand Reviewed-by: Zbigniew Kempczyński