[igt-dev] [ANNOUNCE] intel-gpu-tools 1.22

[igt-dev] [ANNOUNCE] intel-gpu-tools 1.22

[Petri Latvala]

[-- Attachment #1.1: Type: text/plain, Size: 14146 bytes --]


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

General changes:

- Libudev has been made a mandatory dependency. (Antonio Argenziano)

- Documentation changed to refer to the new igt-dev mailing
  list. (Rhys Kidd)

Library changes:

- Refactored timer usage for smaller code size. (Chris Wilson)

- Various fixes to support planar framebuffers. (Maarten Lankhorst)

- Added support for fetching the most recent CRC without waiting for a
  vblank, along with flushing the queue of already collected
  CRCs. (Maarten Lankhorst)

- Added a helper to mark BOs purgeable on vc4. (Boris Brezillon)

- Moved handling of a "cork" BO into lib from various tests.
  (Daniele Ceraolo Spurio)

- Added support for looping over physical i915 engines, as opposed to
  uABI engines that can alias. (Chris Wilson)

- Added an accelerated method for reading from WC buffers.
  (Chris Wilson)

Tools changes:

- Improved intel_vbt_decode output, updated the data from current
  kernel. (Jani Nikula)

- intel_reg can now read/write registers using a given engine.
  (Mika Kuoppala)

- Aubdump can now simulate enhanced execlist submission, for
  gen11+. (Scott D Phillips)





And many other bug fixes, improvements, cleanups and new tests.


And the full changelog follows:

Adam Jackson (1):
      intel_vbt_decode: Typo fixes

Antonio Argenziano (9):
      tests/gem_reset_stats: Fix retrieval of hangcheck stats expectation
      igt/gem_exec_params: Drop drm master privileges only on drm master fds
      lib: Move __gem_context_create to common ioctl wrapper library.
      lib/igt_aux: Add function to swap int64 in array
      tests/gem_ctx_param: Update invalid param
      tests/gem_eio: use igt_cork
      tests/gem_busy: Use intel_measure_ring_size
      igt: Use lib gem_execbuf where possible
      igt: Make libudev mandatory

Arkadiusz Hiler (5):
      meson: Use bufmgr stubs headers if libdrm_intel is not found
      meson: Make cairo mandatory
      meson: Get rid of lib_headers
      tests/meson: Build gem_ctx_isolation
      meson: Force libdrm version also for the subpackages

Boris Brezillon (3):
      igt/vc4: Add a helper function to mark BOs purgeable
      igt: Add igt_vc4_get_param() helper
      igt: Add VC4 purgeable BO tests

Carlos Santa (2):
      lib/igt_kms: Add doc for kmstest_ and igt_ functions
      tests/kms_plane_scaling: Use existing igt_kms apis to paint the fb

Chris Wilson (52):
      igt/gem_linear_blits: Compute GTT size using 4G limit
      igt/gem_tiled_fence_blits: Allocate bo array
      igt/pm_rps: Increase load for waitboosting
      igt/pm_rps: Include related kernel RPS info
      igt/kms_frontbuffer_tracking: Show FBC status during the wait
      lib: Refactor igt_wait() to use library timers
      igt/prime_mmap_coherency: Close dmabuf after use
      tools: Clear unused fields in register spec
      igt/pm_rc6_residency: Check debugfs existence before reading
      igt/perf_pmu: Semaphores do not exist before gen6
      igt/gem_exec_schedule: Limit smoketest to the desired engines
      igt/gem_exec_capture: MI_STORE_DWORD requires EXEC_SECURE + DRM_MASTER on ctg/ilk
      lib: force a reset on an uncooperative i915 device
      lib/kms: Clear unused fields for getproperty ioctl
      igt/kms_frontbuffer_tracking: Disable FBC testing for -ENODEV
      igt/perf_pmu: Retain original GTT offset when resubmitting the spinner
      igt/gem_sync: Exercise and measure idle requests
      lib: Cache the debugfs mountpoint
      lib: Always set mismatching index for igt_find_crc_mismatch
      igt/gem_exec_flush: Silence old compiler warning
      igt/syncobj: Tidy ye olde compiler warnings
      lib: Remove overzealous assertion on gem_set_caching()
      igt/gem_fenced_exec_thrash: Use fixed durations
      igt/kms_frontbuffer_tracking: Wait for PSR to be disabled
      igt/gem_exec_schedule: Trim max number of contexts used
      igt/gem_exec_schedule: Dump the engine info prior to sync on preempt_other
      igt/gem_exec_schedule: Replace constant 16 with its magic macro
      igt/gem_eio: Use slow spinners to inject hangs
      igt/gem_ctx_switch: Do a warmup pass over all contexts
      igt/gem_ctx_isolation: Check isolation of registers between contexts
      igt/kms_fbcon_fbt: Handle ENODEV when checking i915_fbc_info for chipset support
      igt/perf_pmu: Use a self-correcting busy pwm
      igt/perf_pmu: Fix 64b printf-isms
      igt/gem_ctx_isolation: Fix checking for context support
      Iterate over physical engines
      igt/gem_busy: Fix extended-bsd aliasing checks
      igt/kms_force_connector_basic: Clear any previous connector override
      igt/gem_softpin: Only expect EINVAL for color-overlaps for user objects
      lib: Export kmsg()
      igt/debugfs_tests: Record which file is being opened in kmsg
      lib/dummyload: Avoid assertions in lowlevel spin constructor
      lib: Provide an accelerated routine for readback from WC
      igt/gem_exec_capture: Exercise readback of userptr
      igt/gem_ctx_switch: Exercise all engines at once
      lib: Fix MI_BATCH_BUFFER_START for hang injection
      tests/perf_pmu: Handle CPU hotplug failures better
      igt/gem_spin_batch: Avoid waiting when running concurrently
      igt/gen7_forcewake_mt: Mark the mmio register as volatile
      igt: Remove gen7_forcewake_mt
      Bump measure_ring_size() timer interval
      lib: Use C99 initialisers to clear context parameters
      igt/drv_hangman: Check that the error state does hold the expected state

Daniele Ceraolo Spurio (9):
      igt/gem_exec_fence: Test that the in-fence is not overwritten
      lib/igt_dummyload: add igt_cork
      lib/igt_gt: add intel_measure_ring_size
      tests/gem_exec_schedule: use new common functions
      tests/gem_exec_fence: use new common functions
      tests/gem_exec_latency: use new common functions
      tests/gem_wait: use igt_cork
      tests/gem_exec_await: use intel_measure_ring_size
      tests/gem_ringfill: use intel_measure_ring_size

Jani Nikula (4):
      tools/intel_vbt_decode: update vbt defs from kernel
      tools/intel_vbt_decode: print child device count
      tools/intel_vbt_decode: add --header option to only print header
      tools/intel_vbt_decode: add --describe option

Jordan Justen (3):
      tools/aubdump: Support alignment of BO in execbuffer2
      tools/aubdump: Note pinned BO in verbose output
      tools/aubdump: Signal drm sync objects when device override is used

Jyoti Yadav (3):
      tests/kms_plane_scaling: test scaling with tiling rotation and pixel formats, v3.
      tests/kms_plane_scaling: test scaler with clipping clamping, v3.
      tests/kms_plane_scaling: test for multi pipe with scaling, v3.

Lionel Landwerlin (10):
      tests/perf: fix report validity check on gen10+
      overlay: add missing foreground option
      overlay: reindent tracepoint parser
      overlay: fix invalid pointer access
      tests/pm_sseu: dump the bit of sysfs
      tests/perf: make oa-exponents subtest more reliable
      test/perf: fixup inverted condition
      tests/perf: simplify enable-disable subtest
      tests/perf: simplify buffer-fill subtest
      tests/pm_sseu: adapt debugfs parsing for newer kernels

Lofstedt, Marta (2):
      igt/kms_fbcon_fbt: Handle ENODEV when checking i915_edp_psr_status for chipset support
      igt/kms_fronbuffer_tracking: Handle ENODEV when checking i915_edp_psr_status for chipset support

Lohith BS (1):
      tests/kms_frontbuffer_tracking: Including DRRS test coverage

Maarten Lankhorst (41):
      tests/kms_plane_scaling: Move the actual test to its own function.
      tests/kms_plane_scaling: Convert from simple test to full test
      tests/kms_plane_scaling: Move get_num_scalers to a function, v2.
      tests/kms_plane_scaling: Clean up tests to work better with igt_kms, v2.
      tests/kms_flip: Remove blt/rcs flip tests.
      kms_vblank: Reorganize subtests by pipe, v2.
      tests/kms_flip: Move kms_flip.vblank-vs-hang to kms_vblank, v4.
      kms_vblank: Add tests implemented in kms_flip
      kms_flip: Remove redundant vblank tests.
      kms_vblank: Remove teardown code from cleanup_crtc
      tests: Remove kms_render
      lib/igt_fb: Add igt_put_cairo_ctx as counter to igt_get_cairo_ctx
      lib/igt_fb: Pass format to igt_calc_fb_size
      lib/fb: Handle planar formats in igt_calc_fb_size and create_bo_for_fb
      lib/intel_batchbuffer: Add delta argument to igt_blitter_fast_copy__raw, v2.
      lib/intel_batchbuffer: Add src/dst delta arguments to igt_blitter_fast_copy too
      lib/fb: Add support for creating planar framebuffers, v3.
      lib/igt_fb: Add support for NV12 format through conversion, v2.
      tests/kms_vblank: Fix spurious test failure.
      tests/kms_panel_fitting: Remove dead code
      tests/kms_atomic: Add the test for CRTC_ID/FB_ID mismatch.
      tests/kms_plane_scaling: Test all pixel formats in pipe-*-scaler-with-rotation
      tests/kms_plane_scaling: Test all pixel formats with clamping and clipping too
      lib/igt_fb: Make igt_remove_fb more robust
      tests: Always call igt_remove_fb without checking.
      lib/igt_fb: Add igt_fb_supported_format()
      lib/igt_fb: Remove igt_get_all_cairo_formats()
      lib/igt_debugfs: Add igt_pipe_crc_get_single and igt_pipe_crc_drain, v4.
      tests/kms_rotation_crc: Fix bad-tiling testcase.
      tests/kms_rotation_crc: Move bad_format parameter to test_plane_rotation
      tests/kms_rotation_crc: Always run the flip tests when available.
      tests/kms_rotation_crc: Remove primary-rotation-90-Y-tiled.
      tests/kms_rotation_crc: Perform lazy cleanup and require atomic.
      tests/kms_rotation_crc: Clean up exhaust-fences subtest
      tests/kms_rotation_crc: Test all pixel formats on all planes.
      lib/igt_debugfs: Fix igt_pipe_get_crcs.
      lib/igt_kms: Fix igt_display_drop_events() to work as intended
      tests/kms_chv_cursor_fail: Use igt_display_reset() and igt_pipe_crc_drain()
      lib/igt_kms: Fix docbook warnings
      lib/igt_kms: Remove remaining docbook warnings mentioning igt_kms files.
      lib/igt_draw: Fix bo leak in gpu draw routines

Mahesh Kumar (1):
      tests/kms_plane_scaling: Fix basic scaling test, v3.

Mika Kahola (1):
      tests/kms_plane_lowres: Drain pipe before reading CRC

Mika Kuoppala (1):
      tools/intel_reg: Add reading and writing registers through engine

Petri Latvala (7):
      meson: Refactor get_option() calls for directories
      meson: Use absolute path for IGT_DATADIR
      meson: Don't build docs when cross-compiling
      configure.ac: Bump libdrm_amdgpu version requirement
      Update MAINTAINERS
      configure.ac: Properly bump libdrm_amdgpu version requirement
      Update NEWS, bump version to 1.22.

Rhys Kidd (5):
      doc: Fix typos in CONTRIBUTING
      lib/igt.cocci: Fix typo
      doc: Correct sentence in README
      doc: Update CONTRIBUTING for new igt-dev mailing list
      man: Update for new igt-dev mailing list

Rodrigo Vivi (1):
      lib/i915_pciids.h: Add Cannonlake PCI IDs for another SKU.

Sagar Arun Kamble (1):
      tools/intel_guc_logger: Send GuC log level in new i915 expected format

Scott D Phillips (1):
      tools/intel_aubdump: Simulate "enhanced execlist" submission for gen11+

Sean Paul (1):
      CONTRIBUTING: Fix spelling mistake and line length

Thierry Reding (2):
      tests/testdisplay: Explicitly use GLIB_CFLAGS
      tools/intel_dp_compliance: Add missing GLIB_CFLAGS

Tvrtko Ursulin (23):
      tests/perf_pmu: Tighten busy measurement
      tests/perf_pmu: More busy measurement tightening
      tests/perf_pmu: Use measured sleep in all time based tests
      tests/perf_pmu: Convert to flags
      tests/perf_pmu: Add trailing edge idle test variants
      tests/perf_pmu: PMU enable race test
      tests/perf_pmu: Always skip missing engines
      tests/perf_pmu: Explicitly test for engine availability in init tests
      tests/perf_pmu: Use short batches from hotplug test
      tests/perf_pmu: Use perf timestamps when calculating average frequency
      tests/perf_pmu: Test RC6 during runtime suspend
      tests/perf_pmu: Use perf timestamps in a few more places
      tests/perf_pmu: Handle thermally throttled devices
      tests/perf_pmu: Give sampling more time
      tests/perf_pmu: Log perf timestamp in semaphore wait tests
      tests/perf_pmu: Verify engine busyness accuracy
      perf_pmu: Fix some compile warnings with old compilers / 32-bit builds
      tests/perf_pmu: Skip hotplug test on Broxton
      lib/igt_perf: Find active perf CPU
      tests/kms_frontbuffer_tracking: Fix build warning
      tests/perf: Fix build warning
      tests/perf_pmu: Test busyness reporting in face of GPU hangs
      lib/igt_pm: Restore runtime pm state on test exit

Ville Syrjälä (6):
      tests/kms_ccs: Grab the CRC when the plane is actually enabled
      tests/kms_atomic_transition: Don't abuse the HSKEW flag to force a modeset
      tests/kms_ccs: Move fbs out from data
      tests/kms_ccs: Don't skip the entire subtest if one plane can't do CCS
      lib/sysfs: Fix fbcon unbind, again
      lib/sysfs: s/kick_fbcon/bind_fbcon/

git tag: intel-gpu-tools-1.22

https://xorg.freedesktop.org/archive/individual/app/intel-gpu-tools-1.22.tar.xz
MD5:  965c591b23a132084113c2a0604f537a  intel-gpu-tools-1.22.tar.xz
SHA1: 5d624efb14eca16237e1ae2141aeed6ac51df6e3  intel-gpu-tools-1.22.tar.xz
SHA256: 3d66c1dc5110712ca4d22199b3ce9853f261be1690064edf87e69e5392e39a5c  intel-gpu-tools-1.22.tar.xz
SHA512: af7964a3782bbc335c50c1c9c42f090306932423e76b9968d6919a2b633f3c11837e7ba572d352632c2b4db79bc98218d9d425c3b62e69abad5aa83d733dea8a  intel-gpu-tools-1.22.tar.xz
PGP:  https://xorg.freedesktop.org/archive/individual/app/intel-gpu-tools-1.22.tar.xz.sig


[-- Attachment #1.2: signature.asc --]
[-- Type: application/pgp-signature, Size: 833 bytes --]

[-- Attachment #2: Type: text/plain, Size: 154 bytes --]

_______________________________________________
igt-dev mailing list
igt-dev@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/igt-dev

Re: [igt-dev] [ANNOUNCE] intel-gpu-tools 1.22

[Matt Turner]

On Fri, Mar 9, 2018 at 6:28 AM, Petri Latvala <petri.latvala@intel.com> wrote:
>
> A new intel-gpu-tools quarterly release is available with the
> following changes:
>

Since the Meson build system has no configuration options it's not yet
a replacement distros can use. Please ship the 1.23 tarball with the
autotools files generated by make dist.
_______________________________________________
igt-dev mailing list
igt-dev@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/igt-dev

Re: [igt-dev] [ANNOUNCE] intel-gpu-tools 1.22

[Daniel Vetter]

On Wed, Jun 13, 2018 at 02:54:52PM -0700, Matt Turner wrote:
> On Fri, Mar 9, 2018 at 6:28 AM, Petri Latvala <petri.latvala@intel.com> wrote:
> >
> > A new intel-gpu-tools quarterly release is available with the
> > following changes:
> >
> 
> Since the Meson build system has no configuration options it's not yet
> a replacement distros can use. Please ship the 1.23 tarball with the
> autotools files generated by make dist.

Meson upstream seems to think that autoconfiguring with defaults Is Better
(tm). What would distro's want from the meson build (since I think it's
useful to cut down a bit from the all the options, mostly they didn't
really work after a short while anyway)?
-Daniel
-- 
Daniel Vetter
Software Engineer, Intel Corporation
http://blog.ffwll.ch
_______________________________________________
igt-dev mailing list
igt-dev@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/igt-dev

Re: [igt-dev] [ANNOUNCE] intel-gpu-tools 1.22

[Matt Turner]

On Mon, Jun 18, 2018 at 8:51 AM Daniel Vetter <daniel@ffwll.ch> wrote:
>
> On Wed, Jun 13, 2018 at 02:54:52PM -0700, Matt Turner wrote:
> > On Fri, Mar 9, 2018 at 6:28 AM, Petri Latvala <petri.latvala@intel.com> wrote:
> > >
> > > A new intel-gpu-tools quarterly release is available with the
> > > following changes:
> > >
> >
> > Since the Meson build system has no configuration options it's not yet
> > a replacement distros can use. Please ship the 1.23 tarball with the
> > autotools files generated by make dist.
>
> Meson upstream seems to think that autoconfiguring with defaults Is Better
> (tm). What would distro's want from the meson build (since I think it's
> useful to cut down a bit from the all the options, mostly they didn't
> really work after a short while anyway)?

That works fine for developers building it out of their ~/projects
directory, but not so well for distros. Source distros in particular
don't work well with so-called "automagic" dependencies, i.e., those
that are enabled if detected by the configure script without the
ability to explicitly disable them.

Gentoo's dependency tracking works by saying "Depend on package XYZ
iff support for XYZ is requested". Automagic deps break this by
assumption by linking with XYZ if it happens to be installed even if
support isn't requested.

If we just want to make everything required, then... I guess that will
work. But because autotools already has configuration support, the
source code is already ready. FWIW, this came to my attention recently
because I worked with the Chrome team at Intel to add v1.22 to Gentoo.
I suggested switching to Meson and then realized it wasn't
configurable at all.
_______________________________________________
igt-dev mailing list
igt-dev@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/igt-dev

Re: [igt-dev] [ANNOUNCE] intel-gpu-tools 1.22

[Matthieu Herrb]

[-- Attachment #1.1: Type: text/plain, Size: 1545 bytes --]

On Mon, Jun 18, 2018 at 10:23:03AM -0700, Matt Turner wrote:
> On Mon, Jun 18, 2018 at 8:51 AM Daniel Vetter <daniel@ffwll.ch> wrote:
> >
> > On Wed, Jun 13, 2018 at 02:54:52PM -0700, Matt Turner wrote:
> > > On Fri, Mar 9, 2018 at 6:28 AM, Petri Latvala <petri.latvala@intel.com> wrote:
> > > >
> > > > A new intel-gpu-tools quarterly release is available with the
> > > > following changes:
> > > >
> > >
> > > Since the Meson build system has no configuration options it's not yet
> > > a replacement distros can use. Please ship the 1.23 tarball with the
> > > autotools files generated by make dist.
> >
> > Meson upstream seems to think that autoconfiguring with defaults Is Better
> > (tm). What would distro's want from the meson build (since I think it's
> > useful to cut down a bit from the all the options, mostly they didn't
> > really work after a short while anyway)?
> 
> That works fine for developers building it out of their ~/projects
> directory, but not so well for distros. Source distros in particular
> don't work well with so-called "automagic" dependencies, i.e., those
> that are enabled if detected by the configure script without the
> ability to explicitly disable them.
> 
> Gentoo's dependency tracking works by saying "Depend on package XYZ
> iff support for XYZ is requested". Automagic deps break this by
> assumption by linking with XYZ if it happens to be installed even if
> support isn't requested.

+1.  same kind of issues on build system for *BSD.

-- 
Matthieu Herrb

[-- Attachment #1.2: signature.asc --]
[-- Type: application/pgp-signature, Size: 833 bytes --]

[-- Attachment #2: Type: text/plain, Size: 154 bytes --]

_______________________________________________
igt-dev mailing list
igt-dev@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/igt-dev

Re: [igt-dev] [ANNOUNCE] intel-gpu-tools 1.22

[Matt Turner]

On Mon, Jun 18, 2018 at 3:14 PM Ross Vandegrift <ross@kallisti.us> wrote:
>
> On Mon, Jun 18, 2018 at 10:23:03AM -0700, Matt Turner wrote:
> > On Mon, Jun 18, 2018 at 8:51 AM Daniel Vetter <daniel@ffwll.ch> wrote:
> > > On Wed, Jun 13, 2018 at 02:54:52PM -0700, Matt Turner wrote:
> > > > Since the Meson build system has no configuration options it's not yet
> > > > a replacement distros can use. Please ship the 1.23 tarball with the
> > > > autotools files generated by make dist.
> > >
> > > Meson upstream seems to think that autoconfiguring with defaults Is Better
> > > (tm). What would distro's want from the meson build (since I think it's
> > > useful to cut down a bit from the all the options, mostly they didn't
> > > really work after a short while anyway)?
> >
> [snip]
> >
> > source code is already ready. FWIW, this came to my attention recently
> > because I worked with the Chrome team at Intel to add v1.22 to Gentoo.
> > I suggested switching to Meson and then realized it wasn't
> > configurable at all.
>
> Maybe I've missed some context, but I don't think this is a meson limitation:
> https://github.com/mesonbuild/meson/blob/master/docs/markdown/Build-options.md

No, of course not. I meant that i-g-t's meson_options.txt is nearly
empty and its meson.build is only wired to enable deps based on
whether they're present.

> Here is an example option that disables wayland support by default, but
> allows enabling it at build time by specifying "-Dwayland=true":
> https://git.enlightenment.org/core/enlightenment.git/tree/meson_options.txt#n48

Yep, I just finished transitioning Gentoo's Mesa ebuild from autotools
to Meson last week :)
_______________________________________________
igt-dev mailing list
igt-dev@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/igt-dev

Re: [igt-dev] [ANNOUNCE] intel-gpu-tools 1.22

[Daniel Vetter]

On Mon, Jun 18, 2018 at 10:23:03AM -0700, Matt Turner wrote:
> On Mon, Jun 18, 2018 at 8:51 AM Daniel Vetter <daniel@ffwll.ch> wrote:
> >
> > On Wed, Jun 13, 2018 at 02:54:52PM -0700, Matt Turner wrote:
> > > On Fri, Mar 9, 2018 at 6:28 AM, Petri Latvala <petri.latvala@intel.com> wrote:
> > > >
> > > > A new intel-gpu-tools quarterly release is available with the
> > > > following changes:
> > > >
> > >
> > > Since the Meson build system has no configuration options it's not yet
> > > a replacement distros can use. Please ship the 1.23 tarball with the
> > > autotools files generated by make dist.
> >
> > Meson upstream seems to think that autoconfiguring with defaults Is Better
> > (tm). What would distro's want from the meson build (since I think it's
> > useful to cut down a bit from the all the options, mostly they didn't
> > really work after a short while anyway)?
> 
> That works fine for developers building it out of their ~/projects
> directory, but not so well for distros. Source distros in particular
> don't work well with so-called "automagic" dependencies, i.e., those
> that are enabled if detected by the configure script without the
> ability to explicitly disable them.
> 
> Gentoo's dependency tracking works by saying "Depend on package XYZ
> iff support for XYZ is requested". Automagic deps break this by
> assumption by linking with XYZ if it happens to be installed even if
> support isn't requested.
> 
> If we just want to make everything required, then... I guess that will
> work. But because autotools already has configuration support, the
> source code is already ready. FWIW, this came to my attention recently
> because I worked with the Chrome team at Intel to add v1.22 to Gentoo.
> I suggested switching to Meson and then realized it wasn't
> configurable at all.

Let me rephrase my questions: Which specific options do distros want? All
of the ones we have as optional dependencies already, or just a specific
subset, or something else?

In the past we had options like "only build the tools" because someone
asked for that, and that also doens't exist anymore iirc.
-Daniel
-- 
Daniel Vetter
Software Engineer, Intel Corporation
http://blog.ffwll.ch
_______________________________________________
igt-dev mailing list
igt-dev@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/igt-dev