[ANNOUNCE] intel-gpu-tools 1.3

Daniel Vetter daniel.vetter at ffwll.ch
Mon Aug 27 01:26:16 PDT 2012

Hi all,

New intel-gpu-tools release with too much stuff changed. Biggest part is
improvements to the testsuit, but also some other neat things:
- dpf tool for handling l3 remapping on gen7 (Ben)
- robustify against wc gtt mappings (Chris)
- improvements to the reg_dumper and register read/write tools
- haswell support

Cheers, Daniel

Alan Coopersmith (4):
      Fall back to CLOCK_MONOTONIC on systems without CLOCK_MONOTONIC_RAW
      Rename NOPID to NO_PID to avoid conflict with Solaris NOPID
      flip_test: add cast to pacify Solaris Studio compiler
      Add --disable-nouveau option for platforms without nouveau support

Ben Widawsky (30):
      tests: use common code for scripts
      tests: add wedging tests
      build: make sure we have asprintf
      drm/i915: extract card getting
      tests: rc6 residency test
      chipset updates
      chipset: accidentally left the old IS_GEN7 macro
      intel_infoframe: fix assertion off by 1
      tests/wait render timeout test
      tools/dpf: Tool to read and write l3 remap registers.
      tests/dpf: simple dpf test
      context: update for new execbuf2 element
      context tests not requiring libdrm
      ctx_bad_destroy: a little more description is nice
      ctx_bad_destroy - ioctl is 2e, not 23
      ctx_bad_destroy: danvet changed to ENOENT
      lib: add a gen7 rendercopy
      gem_stress: use the gen7 rendercopy
      intel: update libdrm requirement
      context: libdrm wrappers
      context_basic: the most stressful context test to date.
      rendercopy_gen7: add support for haswell
      ctx_basic: we stopped using rendercopy
      gem_ctx_basic: fixes
      reg_read: basic register read ioctl test
      reg_read: skip older kernels
      reg_read: use good offset, and errno
      intel_reg_read: support -c option to read multiple dwords
      intel_reg_read: use strtol instead of atoi
      Converted a display debug script for use by designers

Chris Wilson (16):
      tests/gem_ringfill: Exercise all rings
      lib: Downgrade error for failing to determine swap interface to a warning
      lib: Downgrade error for unknown 'swap' not 'ram'
      configure: Add AC_SYS_LARGEFILE for mmap64 support on 64-bit
      decode: Use the correct start mask for gen3 fence registers
      intel_gpu_top: Trust the register access is safe
      Make the assertions guarding syscalls (drmIoctl in particular) verbose
      test: Exercise concurrent GPU read/write with CPU domain access
      test/gem_cpu_concurrent_blit: Limit memory usage to half aperture
      Use mmap64() for performing the GTT mapping
      lib: Fix render copy on gen2
      intel_error_decode: Be more lax for whitespace around parsing PCI-ID
      tests/gem_tiled_pread_pwrite fix progname for pretty print progress
      test/gem_gtt_speed: Add a baseline test for the performance of a CPU mmap
      intel_gtt: Harden against changes to kernel mappings of the GTT
      mmio: Limit the uc- mapping to only map the registers

Damien Lespiau (1):
      build: Make autoreconf honour ACLOCAL_FLAGS

Daniel Vetter (82):
      prepend 'intel_' to installed programms
      lib/drmtest: fixup copy&paste printf format issue
      remove old .c files
      test/gem_tiled_pread_pwrite: add progress indicator
      tests: fixup storedw tests
      tests/intel_reg_dumper: s/pipe/disp_pipe/ to shut off the compiler
      tests/gem_ringfill: shut up the compiler
      demos/intel_sprite_on: build warning fixes
      tests: add tiled partial pwrite/pread test
      tests/gem_tiled_partial_pread/write: fixes
      tests: some more tuning on the tiled partial pwrite/pread test
      intel_sprite_on: remove copy&pasted gem functions
      intel_sprite_on: kill overtly verbose and totally content-free comments
      update .gitingore
      tests: add gem_mmap_offset_exhaustion
      tests: don't just bail out when there's not enough RAM
      tests: adapt storedw tests to ppgtt
      tests: add gem_tiled_after_untiled_blt
      tests/gem_exec_faulting_reloc: kill superflous mmap_bo
      tests/gem_fence_thrash: make some good use of drmtest helpers
      tests: use gem_create and gem_close more
      tests: use / as path separator instead of _
      tests: fixup drm_lib.sh sourcing
      tests: explicit use bash
      tests/sysfs_rc6_residency: add some sleep before testing
      tests: add sysfs_rc6_residency to .gitignore
      tests: add gem_gtt_concurrent_blit
      tests: skip rc6_residency test if not supported
      tests: pimp tiled_after_untiled_blt
      tests: rename gem_tiled_after_untiled_blt into gem_set_tiling_vs_blt
      tests/gem_set_tiling_vs_blt: fix on gen4+
      tests: add gem_set_tiling_vs_gtt
      tests/gem_set_tiling_vs_gtt: update copyright header
      tests: add gem_set_tiling_vs_pwrite
      tests: re-enable commented-out testcase
      tests/gem_ringfill: add progress indicator
      tests: add gem_unfence_active_buffers
      test: fix up hangman script and enable by default
      test/ZZ_hangman: adjust to new error_state reset semantics
      tests/gtt_speed: try to outwit gcc
      tests/flip_test: rip out funny pictures
      tests: add flip_test to .gitignore
      tests/flip_test: rip out usage
      tests: use drm_open_any more
      lib: extract kmstest_create_fb
      test/flip_test: use kmstest_create_fb
      lib: extract kmstest_dump_mode
      test/flip_test: fail harder
      tests/hangman: fix when run with absolute path
      tests: fixup type in module_reload
      tests/gem_wait_rendering_timeout: make it compile
      tests/gem_wait_rendering_timeout: properly skip test
      tests/gem_wait_rendering_timeout: some fixes for the blt function
      tests/get_wait_rendering_timeout: fixup the auto-tuner
      tests/gem_wait_rendering_timout: adjust .gitignore
      tests/gem_wait_rendering_timeout: fix up the autotuner
      tests/gem_wait_rendering_timeout: test the timeout case
      lib: fix gen5 workaround emission
      lib/gem_wait_rendering_timeout: sprinkle busy_ioctl checks
      tests/gem_wait_rendering_timeout: check polling behaviour
      cleanups after l3_parity merge
      test/sysfs_l3_parity: fixup logic
      test/sysfs_l3_parity: skip test if the sysfs interface doesn't exist
      intel_reg_dumper: dump pre-pch-split blc regs
      tests/debgufs_emon_crash: make it quiet
      tests/gem_wait_rendering_timeout: needs more fudge
      tests: improve concurrent blit tests
      shut up the compiler a bit
      tests: s/drm_reg_read/gem_reg_read
      tests: add basic re-importing prime test
      tests: add missing stuff to .gitignore
      tests/prime: more paranoia
      tests: fixup prime test paranoia
      test/flip_test: don't fail if we can't find a crtc
      tests/gem_reg_read: actually wait a bit between reading TIMESTAMP reg
      tools/reg_dumper: dump pipe C regs
      tools/reg_dumper: really dump pipe C regs
      tests: add gem_gtt_cpu_tlb
      tests/prime_nv: don't enable buffer reuse
      Revert "tests: disable nouveau if unavailable"
      bump to 1.3 release

Eric Anholt (5):
      instdone: Add support for one of the IVB instdone regs.
      instdone: Fix fallthroughs to invalid instdone bits on newer hardware.
      reg_dumper: Refactor duplicated reg-dumping code.
      intel_reg_dumper: Add dumping of GPU turbo regs.
      gem_ringfill: Make this actually test that all the batches executed.

Eugeni Dodonov (7):
      intel_reg_read: add support for getopt
      intel_reg_read: support reading multiple registers
      intel_reg_read: add a flag to simplify bit decoding
      tools: add Haswell registers into intel_reg_dumper
      intel_gpu_top: allow to work on Gen7
      intel_reg_dumper: dump more PM registers
      intel_reg_dumper: use intel_register_access_init/fini

Jesse Barnes (5):
      testdisplay: remove plane test support
      testdisplay: remove unnecessary passing of width, height, stride, depth
      tests: add simple flip test
      tests/flip_test: test on each CRTC for each detected output
      add VLV PCI ID

Kenneth Graunke (1):
      Remove pointless Android ifdefs for sys/fcntl.h.

Maarten Lankhorst (8):
      Add intel <-> nouveau prime tests
      satisfy danvet's OC
      fix crash with insufficient permissions
      And add another corner case skip
      Fixup previous commit
      tests/prime_nv_pcopy: kill debug code that's not upstream
      tests/prime_nv_test: add missing call for test8
      tests: disable nouveau if unavailable

Paulo Zanoni (9):
      reg_dumper: add interlaced and rotation information to PIPECONF
      reg_dumper: add interlaced information to TRANSCONF
      reg_dumper: change and fix behavior when using dump files
      Add intel_panel_fitter tool
      Add tools/intel_infoframes
      intel_infoframes: AVI InfoFrame version should be 2
      intel_infoframes: add support for Gen 4
      intel_infoframes: options -p and -P require an argument
      lib: add more Haswell PCI IDs

Sateesh Kavuri (1):
      Added support for Android

Vijay Purushothaman (1):
      tools: Added intel_dpio_read and intel_dpio_write

Wang Xingchao (2):
      intel_audio_dump: fix wrong port definition
      intel_audio_dump: add Haswell audio dump support

Yi Sun (3):
      tests/testdisplay.c: Add a option '-r'.
      tests/testdisplay.c Remove an uncomfortable error output
      Update the image file pass.png with higher correction level

root (2):
      tests: add module reload testcase
      tests: add some help to module_reload

git tag: intel-gpu-tools-1.3

MD5:  67facd6241e26e2c68614728e3a932e9  intel-gpu-tools-1.3.tar.bz2
SHA1: 35e4907c4dbc48ad37921e44d73f0fe15a989aa2  intel-gpu-tools-1.3.tar.bz2
SHA256: 51d22fdb3d415a1b3b7d0a172c1bb24dec6f16116e80a9ce49873f44527f20a0  intel-gpu-tools-1.3.tar.bz2

MD5:  e4b2d38a178673672527dc7bbd8cd245  intel-gpu-tools-1.3.tar.gz
SHA1: d96b5aa8ce8c5089eefaa13dc1c8594915f832ac  intel-gpu-tools-1.3.tar.gz
SHA256: 14f76a45c434cf30828599e79097d3bbaa0803ac235a5e12b4f0f5e1f85bc45e  intel-gpu-tools-1.3.tar.gz

Daniel Vetter
Mail: daniel at ffwll.ch
Mobile: +41 (0)79 365 57 48

More information about the xorg-announce mailing list