[Intel-gfx] [ANNOUNCE] intel-gpu-tools 1.14

Marius Vlad marius.c.vlad at intel.com
Tue Mar 1 16:22:20 UTC 2016


A new intel-gpu-tools quarterly release is available with the following
changes:

- New test: gem_create validate parameters for GEM_CREATE ioctl
(Ankitprasad Sharma)

- New test: gem_softpin exercise the userptr ioctl to create shared
buffers between CPU and GPU (Vinay Belgaumkar)

- New tests: prime_mmap_coherency/kms_mmap_write_crc cache coherency
tests (Tiago Vignatti)

- New test: prime_mmap mmap() on dma-buf fds (Rob Bradford)

- New test: gem_exec_reloc sanity check of execbuf-ioctl relocations (Chris
Wilson)

- New test: gem_exec_basic sanity check of execbuf-ioctl rings (Chris Wilson)

- improved igt_hang_ring() infrastructure for generic hang injection support in
the core library (Chris Wilson)

- new igt_pm library to collect power management testing helpers (David
Weinehall)

- lig/igt_vc4: VC4 support library (Eric Anholt)

- kms_frontbuffer_tracking: included in BAT (Paulo Zanoni)

- kms_psr_sink_crc: Add BAT test for PSR active (Rodrigo Vivi)

- gem_wait: test superseded by gem_latency in benchmarks (Chris Wilson)

- igt_core: Fix logging to display extended line (Derek Morton)

- igt_core: Expand --run-subtest functionality (Derek Morton)

- kms_force_connector_basic: various fixes and included in BAT set (Daniel
Vetter)

- Many other improvements and bug fixes.

Full changelog follows:

Ankitprasad Sharma (3):
      igt/gem_stolen: Verifying extended gem_create ioctl
      igt/gem_pread: Support to verify pread/pwrite for non-shmem backed obj
      igt/gem_create: Test to validate parameters for GEM_CREATE ioctl

Bob Paauwe (2):
      igt/pm_rps: current freq < user specified min is not a fail (v3)
      igt/test/pm_rps: load GPU to force not-idle to idle transition.

Chris Wilson (134):
      benchmarks/gem_exec_nop: Flush retirement lists before executing
      igt/gem_exec_nop: Remove nop latency measurements
      igt/gem_concurrent_blit: Explicitly check for "missed interrupts"
      igt: s/basic/sanitycheck/ on prior smoketesting
      igt/gem_concurrent_all: Expand testing to cover different memory regimes
      tests/gem_pread,gem_pwrite: Fix compiler warnings
      igt/gem_read_read_speed: Tweak to show comparison against write-write
      igt/gem_exec_nop: Restore SLOW_QUICK loop terminator
      tests/gem_concurrent_all: Add a couple more sanitycheck patterns
      benchmark: Measure of latency of producers -> consumers, gem_latency
      benchmarks: Remove gem_wait
      benchmarks/gem_latency: Fix for !LLC
      benchmarks/gem_latency: Measure whole execution throughput
      benchmarks/gem_latency: Expose the workload factor
      benchmarks/gem_latency: Add nop dispatch latency measurement
      benchmarks/gem_latency: Add time control
      benchmarks/gem_latency: Split the nop/work/latency measurement
      benchmarks/gem_latency: Add output field specifier
      benchmarks/gem_latency: Tweak workload
      benchmarks/gem_latency: Rearrange thread cancellation
      benchmarks/gem_latency: Use RCS on Sandybridge
      benchmarks/gem_latency: Measure effect of using RealTime priority
      benchmarks/gem_latency: Measure CPU usage
      benchmarks/gem_latency: Guard against inferior pthreads.h
      benchmarks/gem_latency: Serialise mmio reads
      benchmarks/gem_latency: Hide spinlocks for android
      benchmarks/gem_blt: Estimate memory bandwidth to improve test runtime
      intel_error_decode: Inflate compressed error state
      intel_error_decode: Fix decode headers for HW context
      intel_error_decode: Update address parsing for 64bit offsets
      benchmarks/gem_exec_ctx: Fix fd switching between default contexts
      benchmarks/gem_exec_ctx: Run for a fixed time
      benchmarks/gem_exec_nop: Convert to running for a fixed time
      benchmarks/gem_mmap: Convert to run over a fixed period
      igt/gem_concurrent_all: Preparatory work for testing different create flags
      benchmarks/gem_latency: Allow setting an infinite time
      igt/gem_mmap_gtt: Add a test to exercise coherency between GTT/CPU
      core/sighelper: Interrupt everyone in the process group
      tests: Add gem_busy
      igt/gem_streaming_writes: Set the initial CPU write domain
      tests: Add basic gem_sync test
      gem_concurrent_blit: Don't call igt_require() outside of a subtest/fixture
      gem_concurrent_blit: Add a pread/pwrite variant to only set part of the buffer
      gem_concurrent_all: Add a "quick" tiny pass
      igt/gem_mmap_wc: Test cpu mmap vs wc mmap coherency
      igt/gem_cs_prefetch: Replace explicit sync with implicit sync
      igt/gem_storedw_loop: Remove libdrm crutches
      igt/gem_storedw_loop: Add a few more iterations
      igt/gem_ring_sync_loop: Be explicit!
      igt/gem_exec_alignment: Test all possible pot alignments
      igt/gem_exec_alignment: Test multiple objects with alignment as well
      igt/gem_exec_alignment: Remember to require 48b objects!
      igt/gem_exec_alignment: Fix off-by-one in evalation of find_last_bit()
      igt/gem_exec_alignment: Convert to subtests
      igt/gem_softpin: Remove false dependencies on esoteric features
      igt/gem_partial_pwrite_pread: Fix range computation
      lib: Always double check igt_require_hang_ring() on use
      tests: Drop the superfluous igt_require_hang_ring()
      lib: Expand igt_hang_ring() to select target context and various options
      igt/gem_softpin: Fix MI_STORE_DWORD_IMM for gen2-3
      igt/gem_concurrent_blit: Add userptr backing storage tests
      igt/gem_cs_tlb: Use softpin to remove GTT layout assumptions
      igt/gem_exec_alignment: Tweaks count for large alignments
      igt/gem_ringfill: Reduce ringfill to just filling the rings
      lib: Print memory requirements as MiB
      igt: Fix use MI_STORE_DATA_IMM on gen3
      igt/gem_softpin: Fix MI_STORE_DATA_IMM for gen3
      lib: Refactor common detection of missed interrupts
      igt/gem_userptr_blits: Exercise applying relocations to a userptr bo
      igt/gem_ringfill: Disable MI_STORE_DATA_IMM on BSD/gen6
      igt/gem_ringfill: Mark the write object as EXEC_OBJECT_WRITE
      igt/gem_concurrent_blit: Check inheritance of buffers
      igt/gem_concurrent_blit: Disable userptr+child tests
      igt/gem_concurrent_blit: Allocate a private batch cache for the child
      igt/gem_concurrent_blit: Switch to a shared mmap for userptr
      igt/gem_softpin: Exercise snoop+uncached abutting
      igt/gem_softpin: Keep last_handle around to vary hole generation
      Promote eviction memory sizes and buffer counts to uint64_t
      igt/gem_exec_alignment: Actually check aligned locations after many
      igt/gem_exec_alignment: Reduce GTT usage if !full-ppgtt
      igt/gem_concurrent_blit: Tidy blt-fill using gem wrappers
      igt/gem_concurrent_blit: Close userptr handle after importing into bufmgr
      igt/gem_concurrent_blit: Tighter scoping of buffers variable
      lib: Share common __gem_execbuf()
      lib: Query the kernel for support of a particular exec id
      Extract array of execution engines
      igt: Add gem_exec_basic
      igt/gem_storedw_loop: Use common array of rings
      igt/gem_exec_nop: Use common array of rings
      igt/gem_cs_tlb: Use common array of rings
      igt/gem_ringfill: Use common array of rings
      igt/gem_sync: Use common array of rings
      Add I915_EXEC_DEFAULT to list of known engines
      igt/gem_cs_prefetch: Check each ring
      igt/gem_concurrent_all: Pass buffer data down
      igt/gem_concurrent_blit: Disable libdrm buffer cache for child inheritance
      igt/gem_reset_stats: Convert residual calllers of gem_exec() to gem_execbuf()
      tests: Add gem_exec_reloc
      lib: Hide BSD1/BSD2 rings on hardware without BSD2
      igt/gem_ringfill: Set MI_MEM_VIRTUAL flag for gen<6
      igt: More MI_STORE_DWORD fixes for gen5
      igt/gem_cs_tlb: Increase BB start alignment to 64bytes
      igt/gem_busy: Refactor to use gem_require_ring()
      igt/gem_streaming_writes: Set bb start alignment to 64b for Ironlake
      igt/gem_ringfill: Add exercising the default-ring to basic testing
      igt/drv_hangman: Inject a true hang
      igt/drv_hangman: Make the batchbuffer check more robust
      tests/drv_hangman: Convert to using central list of engines
      igt/gem_ctx_exec: Convert from stop-rings to a real GPU hang/reset
      igt/gem_reset_stats: Convert from stop-rings to real hang injection
      igt/gem_userptr_blits: Limit amount of mlocked surfaces
      igt/gem_ctx_thrash: Rewrite to avoid extraneous allocations
      lib: Silence a common debug message when creating a context
      igt/gem_exec_alignment: Mark batch buffer as also requiring 48B
      igt/gem_exec_alignment: Fix off-by-one in buffer objects
      igt/gem_workarounds: Convert to real GPU hang injection
      igt/gem_userptr_blits: Flip sign on error check
      igt/gem_ctx_thrash: Double number of threads/proceesses
      igt/gem_ctx_thrash: Combine context thrashing with a render test
      igt: Mark gem_cs_tlb/gem_cs_prefetch as a pair of basic GTT layout tests
      igt: Report the global GTT size
      lib: Restore gem_available_aperture_size()
      igt/gem_exec_reloc: Check WC mmaps as well
      lib: Move gem_wait() to ioctl-wrappers
      igt/gem_concurrent_blit: Add dmabuf exercisers
      igt/gem_concurrent_blit: Pretend tests exist
      igt/gem_sync: Add some forked variants
      igt/gem_sync: Test across all rings simultaneously
      igt/drv_hangman: Tidy up assertion failure message
      igt/gem_sync: Enforce a timeout of 20s
      lib: Check required number of surfaces against VFS file limits
      lib: Add read/write direction support for dmabuf synchronisation
      igt/gem_concurrent_blit: Deglobalify num_buffers
      benchmarks/gem_blt: Measure the throughput of synchronous copies

Damien Lespiau (1):
      list-workarounds: Fix python 2 print statement

Daniel Stone (1):
      build: Disable x86-specific utilities on non-x86

Daniel Vetter (9):
      tests/kms_force_connector: Fixes
      lib/kms+tests: Use cached connector state
      tests/kms_force_connector: Include in BAT set
      lib: igt_fork_hang_helper must be run in fixtures
      tests/drm_lib.sh: Skip when i915 debugfs wasn't found
      tests/drv_hangman: Open drm fd before doing anything
      Revert "igt: s/basic/sanitycheck/ on prior smoketesting"
      tests/gem_eio: New ABI - no EIO even from wait_ioctl
      lib: Only compile igt_vc4 is we have it

Daniele Ceraolo Spurio (3):
      tests/gem_exec_params: test all valid execution flags
      lib: move i915_wedged_set to ig_gt.c
      tests/drv_hangman: test for acthd increasing through invalid VM space

David Weinehall (1):
      lib/igt_pm: Lib for power management

Derek Morton (5):
      gem_flink_race/prime_self_import: Improve test reliability
      igt_core: Fix logging to display extended line
      tools/Android.mk: Add zlib support
      tests/Android.mk: Make intel_residency CAIRO dependant
      lib/igt_core.c: Expand --run-subtest functionality.

Eric Anholt (10):
      igt: Disable igt_clflush_range() implementation on ARM builds
      configure.ac: Test for libdrm_vc4 and build vc4 tests based on it.
      igt: Add support for DRIVER_VC4 flags on tests.
      igt/vc4_wait_seqno: Add a test for VC4's wait-for-seqno ioctl.
      igt/vc4_wait_bo: Add a test for VC4's wait-for-BO ioctl.
      igt: Add a helper function for getting a VC4 BO that's been drawn to.
      igt: Add a helper function for mapping VC4 BOs.
      igt: Add a helper function for creating VC4 BOs.
      igt/vc4_create_bo: Test various paths in BO creation.
      igt/vc4_wait_bo: Add tests with rendering performed.

Gabriel Feceoru (1):
      igt/gem_ringfill: Allow listing subtests in gem_ringfill

Jani Nikula (11):
      intel_opregion_decode: add new ASLE fields
      intel_bios_reader: check mipi sequence block version
      intel_bios_reader: rewrite the mipi sequence block dumping
      intel_bios_reader: improve element dumpers
      intel_bios_reader: add dumping of i2c element in mipi sequence
      intel_bios_reader: pass bdb pointer around instead of having as global
      intel_bios_reader: fix size handling for 32-bit block size
      intel_bios_reader: make the VBT pointers more const
      intel_bios_reader: port find_panel_sequence_block from kernel
      intel_bios_reader: port the sequence block parsing from kernel
      intel_bios_reader: dump MIPI sequence block v3

Jesse Barnes (1):
      lib/igt_kms, tests/testdisplay: allow probing of new connector modes

Kibey, Sameer (1):
      list-workarounds: Extend the script to Mesa

Kristian Høgsberg Kristensen (3):
      aubdump: Raise SIGTRAP on failure
      aubdump: Don't use .so constructors for initializing
      aubdump: Add a get_bo() lookup function

Maarten Lankhorst (4):
      tests: Bump pageflip wait timeout to 50 ms.
      gem_ringfill: fix typo in test name
      kms_force_connector_basic: Add force-load-detect test
      tests/kms_force_connector_basic: Disable all crtc's for load-detect.

Marius Vlad (4):
      Update MAINTAINERS file.
      tests/prime_mmap: Encapsulate check_for_dma_buf_mmap() in igt_fixture.
      NEWS: Updates and bump release to 1.14.
      Bump again to 1.14, as Chris managed to push before I got chance to do a

Matt Roper (1):
      kms_atomic: drop unnecessary connector looping from plane_primary test

Michał Winiarski (2):
      lib/ioctl_wrappers: Add gem_gtt_type exposing raw HAS_ALIASING_PPGTT param
      lib/ioctl_wrappers: Add gem_has_softpin

Michel Thierry (1):
      tests/gem_softpin: Use offset addresses in canonical form

Mika Kahola (1):
      tools/intel_bios_reader: Fix typo on dump info

Mika Kuoppala (1):
      tests/gem_softpin: Fix compiler warning on 32bit systems

Paulo Zanoni (12):
      tools: add intel_residency
      lib/igt_fb: make the automatic buffer sizes/strides smaller
      kms_frontbuffer_tracking: use igt_drm_format_to_bpp()
      kms_frontbuffer_tracking: standardize the used FB sizes
      lib/igt_draw: use igt_drm_format_to_bpp()
      lib/igt_fb: fix igt_get_all_formats documentation
      tests/igt_fb: rename igt_get_all_formats to igt_get_all_cairo_formats
      tests/pm_rpm: remove POWER_DIR definition
      tests/pm_rpm: find an appropriate CRTC instead of hardcoding CRTC 0
      tools/intel_residency: use setitimer instead of {,u}alarm
      kms_frontbuffer_tracking: don't fail fbc-farfromfence on SKL
      kms_frontbuffer_tracking: add basic subtest

Rob Bradford (1):
      prime_mmap: Add new test for calling mmap() on dma-buf fds

Rodrigo Vivi (8):
      kms_frontbuffer_tracking: Increase the time we wait for PSR.
      kms_frontbuffer_tracking: Make sink crc mandatory only for PSR.
      kms_frontbuffer_tracking: Skip on unreliable CRC.
      kms_psr_sink_crc: Fix no-psr option.
      kms_psr_sink_crc: Add suspend/resume sub test.
      kms_psr_sink_crc: Reduce our time out for PSR active.
      kms_psr_sink_crc: Add BAT test for PSR active.
      kms_psr_sink_crc: Simplify debugfs reading.

Thomas Wood (4):
      tests: replace drm_open_any in gem_create and gem_stolen
      tests: add exit value constants for shell script tests
      tests: remove .gitignore and add a Makefile rule to create it
      docs: add igt_edid_template.h to IGNORE_HFILES

Tiago Vignatti (5):
      lib: Add gem_userptr and __gem_userptr helpers
      prime_mmap: Add basic tests to write in a bo using CPU
      lib: Add prime_sync_start and prime_sync_end helpers
      tests: Add kms_mmap_write_crc for cache coherency tests
      tests: Add prime_mmap_coherency for cache coherency tests

Tomeu Vizoso (1):
      lib: Fix build when vc4 headers are present

Ville Syrjälä (15):
      lib/kms: Turn base_edid into a template
      tests/kms_force_connector_basic: Add prune-stale-modes subtest
      lib: Make 'extra_long_opts' const
      lib: Use igt_assert_eq() to check for crc component count
      lib: Extract some common fb create+fill methods into helpers
      lib: Add igt_pipe_crc_new_nonblock()
      tests/kms_pipe_crc_basic: Add tests for O_NONBLOCK CRC reads
      tests/kms_chv_cursor_fail: Add a test to exercise CHV pipe C cursor fail
      Fix a bunch of printf types
      tests/gem_mmap_gtt: Deal with tile sizes on gen2/3
      tests/gem_mmap_gtt: Add progress indicators
      tests/gem_mmap_gtt: Make the small-bo tiling tests work on old platforms
      tests/kms_pipe_crc_basic: Use igt_assert_eq() to see the failing frame counts
      tests/kms_flip: Increase TEST_TS_CONT max seq difference to 150
      tests/kms_chv_cursor_fail: Skip when the pipe doesn't exist

Vinay Belgaumkar (1):
      tests/gem_softpin: New tests for softpin feature

git tag: intel-gpu-tools-1.14

http://xorg.freedesktop.org/archive/individual/app/intel-gpu-tools-1.14.tar.bz2
MD5:  b09d69526c86174007bea2228e36b2f1  intel-gpu-tools-1.14.tar.bz2
SHA1: 62527d856251fe53a30f08364332926bffc46654  intel-gpu-tools-1.14.tar.bz2
SHA256: 3596e1d36e9c1203eed5abff5a0f272d573f4eb917b5cee4f029cce5a70a0f0c  intel-gpu-tools-1.14.tar.bz2
PGP:  http://xorg.freedesktop.org/archive/individual/app/intel-gpu-tools-1.14.tar.bz2.sig

http://xorg.freedesktop.org/archive/individual/app/intel-gpu-tools-1.14.tar.gz
MD5:  ce4b9b9af4fd912b914bd9276a6cf951  intel-gpu-tools-1.14.tar.gz
SHA1: 0806b8ecb10e4153e9ac9c1a9fc5eb7fd0c00fb7  intel-gpu-tools-1.14.tar.gz
SHA256: 476136cc6f99af2500170fca7c01c9ca70fe1d04d084ddeec3b1dee2daa098e8  intel-gpu-tools-1.14.tar.gz
PGP:  http://xorg.freedesktop.org/archive/individual/app/intel-gpu-tools-1.14.tar.gz.sig

-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 473 bytes
Desc: Digital signature
URL: <https://lists.freedesktop.org/archives/intel-gfx/attachments/20160301/ce3297fb/attachment-0001.sig>


More information about the Intel-gfx mailing list