Daniel Vetter [Sun, 4 Dec 2011 11:08:59 +0000 (12:08 +0100)]
tests: add gem_reloc_vs_gpu
Tests whether the kernel properly waits for the gpu before
applying a reloc.
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Daniel Vetter [Fri, 2 Dec 2011 19:40:03 +0000 (20:40 +0100)]
Revert some accidental things
... somehow slipped in.
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Daniel Vetter [Fri, 2 Dec 2011 19:36:06 +0000 (20:36 +0100)]
tests: add gem_exec_faulting_reloc
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Daniel Vetter [Fri, 2 Dec 2011 17:41:34 +0000 (18:41 +0100)]
tests: add gem_exec_bad_domains
To check whether the kernel properly rejects non-gpu domains in
relocs.
Signed-Off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Daniel Vetter [Fri, 2 Dec 2011 09:55:56 +0000 (10:55 +0100)]
tests: add gem_tiled_pread_pwrite
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Daniel Vetter [Tue, 1 Nov 2011 13:33:53 +0000 (14:33 +0100)]
improve README
Signed-Off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Daniel Vetter [Wed, 30 Nov 2011 11:36:29 +0000 (12:36 +0100)]
tests/gem_gtt_speed: Add modes that force that to the gtt domain
No sane userspace doesn't upload data to the gpu without actually
using it there ...
Signed-Off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Paul Berry [Wed, 30 Nov 2011 17:47:56 +0000 (09:47 -0800)]
Add missing executables to .gitignore.
Daniel Vetter [Sun, 27 Nov 2011 01:39:14 +0000 (02:39 +0100)]
tests: implement snb+ XY_SETUP_CLIP_BLT workaround
Better safe than sorry.
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Daniel Vetter [Sun, 27 Nov 2011 00:55:42 +0000 (01:55 +0100)]
tests/sysfs_edit_reader: posixify
Debian's default sh complains ...
Signed-Off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Daniel Vetter [Sun, 27 Nov 2011 00:37:59 +0000 (01:37 +0100)]
tests: properly terminate tests when values mismatch
Not much use running them in the test rig otherwise.
Signed-Off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Eugeni Dodonov [Fri, 11 Nov 2011 13:17:59 +0000 (11:17 -0200)]
tests/gem_hang: fix copy-paste problem
I suspect that we should do different things for different pipes..
Spotted by Thomas Jarosh on #intel-gfx freenode.
Signed-off-by: Eugeni Dodonov <eugeni.dodonov@intel.com>
Eugeni Dodonov [Thu, 10 Nov 2011 13:27:27 +0000 (11:27 -0200)]
tests: add a test for checking edid reading delays
With base on EDID timing testing, when we take more than 1s to run
xrandr command, something is wrong.. So add this test for testing the time
we take to read the status of all the connectors from sysfs. It should do
us an average picture of how long we'd take to run xrandr (roughtly 2x
that value).
Signed-off-by: Eugeni Dodonov <eugeni.dodonov@intel.com>
Eugeni Dodonov [Thu, 10 Nov 2011 13:12:10 +0000 (11:12 -0200)]
tests: add test for kernel segmentation fault
As asked by Daniel Vetter, this is a tech which checks if we can cause
division by zero in kernel by reading the i915_emon_status debugfs
entry repeatably.
Signed-off-by: Eugeni Dodonov <eugeni.dodonov@intel.com>
Eugeni Dodonov [Tue, 8 Nov 2011 17:47:03 +0000 (15:47 -0200)]
intel_gpu_top: use debugfs interface for forcewake
Using debugfs facilities for forcewake-related stuff.
Acked-by: Ben Widawsky <ben@bwidawsk.net>
Signed-off-by: Eugeni Dodonov <eugeni.dodonov@intel.com>
Daniel Vetter [Tue, 8 Nov 2011 11:36:24 +0000 (12:36 +0100)]
tests: disable pipe_control on i965g
That hw is broken.
Reported-by: Dave Airlie <airlied@gmail.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Chris Wilson [Fri, 4 Nov 2011 17:10:06 +0000 (17:10 +0000)]
Call Arrandale gen5 not gen4.
Chris Wilson [Fri, 4 Nov 2011 13:26:13 +0000 (13:26 +0000)]
testdisplay: Round tiled allocations up to pot stride and fence size
Be simple and use the strictest requirements from gen2/3 with old
kernels so that this simply works everywhere.
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=42585
Jesse Barnes [Wed, 2 Nov 2011 19:57:40 +0000 (12:57 -0700)]
testdisplay: find planes for each pipe
Fixes a bug where we were masking against the wrong value.
Jesse Barnes [Wed, 2 Nov 2011 19:57:20 +0000 (12:57 -0700)]
testdisplay: update to new addfb2 definition
Daniel Vetter [Tue, 1 Nov 2011 16:20:01 +0000 (17:20 +0100)]
cargo-cult the build system into shape
Signed-Off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Daniel Vetter [Tue, 1 Nov 2011 16:04:44 +0000 (17:04 +0100)]
always require libdrm
... and also add the missing files to lib/Makefile.am
Signed-Off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Daniel Vetter [Tue, 1 Nov 2011 11:10:35 +0000 (12:10 +0100)]
tests/gem_tiled_pread: fixup detiler for gen2
Signed-Off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Daniel Vetter [Tue, 1 Nov 2011 10:52:53 +0000 (11:52 +0100)]
tests/gem_largeobject: accomodate for puny apertures
... like on my i855gm.
Signed-Off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Ben Widawsky [Tue, 1 Nov 2011 00:19:54 +0000 (17:19 -0700)]
i-g-t/debugger - make the debugger work with distcheck
Signed-off-by: Ben Widawsky <ben@bwidawsk.net>
Ben Widawsky [Mon, 31 Oct 2011 20:36:50 +0000 (13:36 -0700)]
i-g-t/debugger: Only allow the shader debugger if the user has specified a gen4asm
Signed-off-by: Ben Widawsky <ben@bwidawsk.net>
Daniel Vetter [Mon, 31 Oct 2011 22:19:08 +0000 (23:19 +0100)]
tests: mark gem_bad_address as HANG
Kills my little i855gm.
Signed-Off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Daniel Vetter [Mon, 31 Oct 2011 16:40:37 +0000 (17:40 +0100)]
tests/debugfs_reader: fight bashism
debian's dash doesn't like this.
Signed-Off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Daniel Vetter [Mon, 31 Oct 2011 15:08:02 +0000 (16:08 +0100)]
tests: actually add the drm client check script
Ooops.
Signed-Off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Daniel Vetter [Mon, 31 Oct 2011 15:04:34 +0000 (16:04 +0100)]
tests: check for other drm clients
Signed-Off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Daniel Vetter [Sun, 30 Oct 2011 22:20:12 +0000 (23:20 +0100)]
tests/gem_bad_length: disable broken batchbuffer
Submitting random crap hangs the gpu. Suprise!
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Daniel Vetter [Sun, 30 Oct 2011 21:27:19 +0000 (22:27 +0100)]
tests: don't run kernel check from make check
Add a new target test for that.
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Daniel Vetter [Sun, 30 Oct 2011 21:04:39 +0000 (22:04 +0100)]
tools: script to gather bug report data
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Daniel Vetter [Sun, 30 Oct 2011 20:24:12 +0000 (21:24 +0100)]
make distcheck fixes
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Daniel Vetter [Sun, 30 Oct 2011 19:51:03 +0000 (20:51 +0100)]
tests: add test to read all debugfs files
Catches a bunch of forcewake warnings on snb+
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Daniel Vetter [Sun, 30 Oct 2011 19:45:42 +0000 (20:45 +0100)]
testdisplay: compile fix for !DRM_PLANES
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Daniel Vetter [Fri, 28 Oct 2011 15:39:14 +0000 (17:39 +0200)]
tests: add gem_unref_active_buffers
The crucial trick to reproduce the bug is that we need to have
a decent pile of active bos to retire. Because we unref the bo
after having it moved off the active list, our recursion depth
in fdo bug #42180 is limited by the number of active objects that
can retire at the same time.
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=42180
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Jesse Barnes [Fri, 28 Oct 2011 03:25:53 +0000 (05:25 +0200)]
testdisplay: allow overlay plane to be moved & scaled at runtime
For additional testing.
Signed-off-by: Jesse Barnes <jbarnes@virtuousgeek.org>
Jesse Barnes [Fri, 28 Oct 2011 15:53:48 +0000 (08:53 -0700)]
intel_reg_dumper: handle 3 pipe configs when dumping HDMI config
Could be on pipe A, B, or C.
Signed-off-by: Jesse Barnes <jbarnes@virtuousgeek.org>
Jesse Barnes [Fri, 28 Oct 2011 00:47:58 +0000 (17:47 -0700)]
testdisplay: add stub disable_planes() function
To fix the compile with non-plane libdrms.
Signed-off-by: Jesse Barnes <jbarnes@virtuousgeek.org>
Daniel Vetter [Tue, 25 Oct 2011 10:24:17 +0000 (12:24 +0200)]
tools/decode: decode XY_SCANLINES_BLT
Daniel Vetter [Tue, 25 Oct 2011 10:15:27 +0000 (12:15 +0200)]
tools/decode: decode XY_SETUP_* cmds
And extract decode code for br00 and br01.
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Jesse Barnes [Mon, 24 Oct 2011 15:15:34 +0000 (17:15 +0200)]
testdisplay: add sprite scaling and location support
Allow testdisplay to locate the sprite on the screen and potentially
scale it using different crtc width/height vs the source width/height
(determined by the resolution on the target pipe).
Also fix exit, making sure we properly disable all the planes.
Signed-off-by: Jesse Barnes <jbarnes@virtuougseek.org>
Daniel Vetter [Fri, 21 Oct 2011 17:06:08 +0000 (19:06 +0200)]
tools/intel_error_decode: decode gen4+ fences
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Daniel Vetter [Fri, 21 Oct 2011 08:09:14 +0000 (10:09 +0200)]
tools/decode: properly decode gen6+ PIPE_CONTROL flags
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Chris Wilson [Tue, 18 Oct 2011 13:05:54 +0000 (14:05 +0100)]
top: Avoid FPE by removing meaningless metric
Eric Anholt [Fri, 23 Sep 2011 21:47:44 +0000 (14:47 -0700)]
intel_reg_checker: New tool to accumulate checks for HW workarounds.
The goal of this tool is to be usable as a test for whether something
(suspend/resume, GPU reset, bugs, whatever) may have lost various
required hardware setup specified in the docs.
Reviewed-By: Eugeni Dodonov <eugeni.dodonov@intel.com>
Daniel Vetter [Tue, 18 Oct 2011 14:30:56 +0000 (16:30 +0200)]
intel/decode: print out chipset gen
... instead of i965+ for almost everything that Intel is still selling.
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Daniel Vetter [Tue, 18 Oct 2011 13:54:31 +0000 (15:54 +0200)]
tools/decode: don't move around the display register section
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Daniel Vetter [Tue, 18 Oct 2011 13:49:41 +0000 (15:49 +0200)]
tools/decode: don't forget to print the name of the last ring
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Eugeni Dodonov [Mon, 17 Oct 2011 19:48:18 +0000 (17:48 -0200)]
Allow intel_dump_decode to read data from stdin.
Signed-off-by: Eugeni Dodonov <eugeni.dodonov@intel.com>
Daniel Vetter [Mon, 17 Oct 2011 17:41:42 +0000 (19:41 +0200)]
tools/intel_dump_decode: make devid parsing more useful
We use hex for these!
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Daniel Vetter [Tue, 11 Oct 2011 11:14:02 +0000 (13:14 +0200)]
tools/decode: retain the ring name
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Daniel Vetter [Sun, 16 Oct 2011 19:13:34 +0000 (21:13 +0200)]
tests: add gem_hangcheck_forcewake
Provokes the forcewake warning when the hangcheck runs and no
one waits for the gpu (and hence holds the dev->struct_mutex).
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Daniel Vetter [Sat, 15 Oct 2011 11:46:19 +0000 (13:46 +0200)]
tests/gem_fence_trash: some retuning
We actually want to test lru behaviour, so do a bit of work with
the fence before yielding to the next thread (we use twice as many
fences as there are, so yielding always is pretty bad, no matter how
clever our fence stealing).
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Daniel Vetter [Sat, 15 Oct 2011 11:09:36 +0000 (13:09 +0200)]
tests/gem_*_blits: reduce buffer count to not trash swap
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Eugeni Dodonov [Fri, 14 Oct 2011 18:26:13 +0000 (15:26 -0300)]
Allow to output statistics to stdout.
This allows intel_gpu_top to run both in statistics-collecting mode
(collecting the per-ring statistics in gnuplot-friendly format) and
ncurses top-like mode at the same time.
It also allows to output the statistics directly to stdout, by using "-o
-", so the results can be parsed directly via a popen() parsing.
If you are using intel_gpu_top as previously (without any command-line
arguments), it should change nothing for you. If you were using its
logging facilities (e.g., the '-o file'), note that the logging will keep
running, but the detailed top-like interface will be on the screen at the
same time.
Signed-off-by: Eugeni Dodonov <eugeni.dodonov@intel.com>
Adam Jackson [Tue, 11 Oct 2011 22:25:26 +0000 (18:25 -0400)]
bios_reader: Allow forcing the device ID with DEVICE=0xNNNN
As a pleasant side effect you can now feed it an opregion dump and it'll
decode without crashing.
Signed-off-by: Adam Jackson <ajax@redhat.com>
Jesse Barnes [Tue, 11 Oct 2011 18:37:25 +0000 (11:37 -0700)]
intel_reg_dumper: use none instead of NULL if no port is assigned to TRANS_DP
Makes the output a little more readable.
Jesse Barnes [Mon, 10 Oct 2011 18:24:42 +0000 (11:24 -0700)]
Treat PPT like CPT as we do in the kernel
Set the pch type to CPT if we detect a PPT as well.
Daniel Vetter [Wed, 28 Sep 2011 08:05:56 +0000 (10:05 +0200)]
tests: disable storedw tests on !render
According to docs, this should work with ppgtt and just convert to
noops otherwise. In reality, it hangs the machine.
So just disable these for the moment.
Signed-Off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Daniel Vetter [Wed, 28 Sep 2011 08:02:57 +0000 (10:02 +0200)]
tools/decode: add missing shift to MI_SEM_MBOX decoder
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Paulo Zanoni [Fri, 16 Sep 2011 21:06:27 +0000 (18:06 -0300)]
testdisplay: return 0 when successful
We use the "ret" variable to check function return values, and these
values are not necessarily 0 on the successful cases, so we need to
reset "ret" after we finish using it.
Signed-off-by: Paulo Zanoni <paulo.r.zanoni@intel.com>
Signed-Off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Paulo Zanoni [Thu, 15 Sep 2011 18:18:40 +0000 (15:18 -0300)]
getstats: remove useless assertion
getstats.c:48:2: warning: comparison of unsigned expression >= 0 is
always true
Signed-off-by: Paulo Zanoni <paulo.r.zanoni@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Paulo Zanoni [Thu, 15 Sep 2011 18:17:46 +0000 (15:17 -0300)]
drm_open_any: explain why we are abort()ing
Signed-off-by: Paulo Zanoni <paulo.r.zanoni@intel.com>
[danvet: pimped error msg]
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Paulo Zanoni [Thu, 15 Sep 2011 18:16:33 +0000 (15:16 -0300)]
testdisplay: remove wrong return statement
It seems that the kernel patches required by the TEST_PLANES feature are
still not upstream, so I didn't test this patch.
Signed-off-by: Paulo Zanoni <paulo.r.zanoni@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Paulo Zanoni [Thu, 15 Sep 2011 18:14:18 +0000 (15:14 -0300)]
testdisplay: don't enter the main loop if dump_info or test_all_modes
If calls inside update_display fail, the function returns and we don't
quit the program if dump_info or test_all_modes. So we enter the main
loop and keep waiting for user input, even on cases where we are not
supposed to require user input. To fix this, we move the check to
outside the update_display function. As a side effect, we also do
cleanup instead of just exit(0) and we return 1 in case update_display
failed.
Signed-off-by: Paulo Zanoni <paulo.r.zanoni@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Paulo Zanoni [Thu, 15 Sep 2011 18:13:41 +0000 (15:13 -0300)]
testdisplay: test all modes if no option is provided
Previously, when called without any arguments, the application required
user input to finish. However, testdisplay is ran by "make check", and
it is not a good idea to run a program that requires user input in "make
check". So we change the default behavior to something that doesn't
require user input. You can still access the previous behavior using the
"-m" argument.
Signed-off-by: Paulo Zanoni <paulo.r.zanoni@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Paulo Zanoni [Thu, 15 Sep 2011 18:12:48 +0000 (15:12 -0300)]
Add missing files to .gitignore
Signed-off-by: Paulo Zanoni <paulo.r.zanoni@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Paulo Zanoni [Thu, 15 Sep 2011 18:10:52 +0000 (15:10 -0300)]
testdisplay: fix compiler warnings
testdisplay.c:117:1: warning: comparison between signed and unsigned
integer expressions
testdisplay.c:125:1: warning: comparison between signed and unsigned
integer expressions
testdisplay.c:145:1: warning: comparison between signed and unsigned
integer expressions
testdisplay.c:1060:16: warning: comparison between signed and unsigned
integer expressions
Signed-off-by: Paulo Zanoni <paulo.r.zanoni@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Daniel Vetter [Thu, 15 Sep 2011 16:47:21 +0000 (18:47 +0200)]
tests/gem_mmap_gtt: also test gtt pwrite paths
This needs a properly pre-faulted dst bo.
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Daniel Vetter [Wed, 14 Sep 2011 12:58:55 +0000 (14:58 +0200)]
tests: build all tests by default
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Daniel Vetter [Wed, 14 Sep 2011 12:48:45 +0000 (14:48 +0200)]
tests: separate testcases in gem_bad_length
Creating a zero-length bo should fail, so check that. Still run
the minimal batchbuffer, but without the zero-length reloc.
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Daniel Vetter [Wed, 14 Sep 2011 09:06:03 +0000 (11:06 +0200)]
tests: disable gem_bad_blit
Hangs gen3 and simply writes garbage into the unmappable part of
gtt on gen4+, which might cause issues later on.
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Daniel Vetter [Tue, 13 Sep 2011 18:57:04 +0000 (20:57 +0200)]
tests/gem_bad_length: dont throw garbage at the gpu
It doesn't like it. Really.
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Ben Widawsky [Tue, 13 Sep 2011 15:54:33 +0000 (08:54 -0700)]
intel: we don't do distcheck for the debugger
Daniel Vetter [Tue, 13 Sep 2011 15:20:12 +0000 (17:20 +0200)]
intel_gpu_dump: really kill it
This is getting a bit hilarious ...
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Daniel Vetter [Tue, 13 Sep 2011 10:05:58 +0000 (12:05 +0200)]
intel_gpu_dump: also kill manpage
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Daniel Vetter [Tue, 13 Sep 2011 09:43:06 +0000 (11:43 +0200)]
tools: remove intel_gpu_dump
Superseded by the kernel's i915_error_state dumping facility.
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Daniel Vetter [Tue, 13 Sep 2011 09:41:42 +0000 (11:41 +0200)]
tools: beef up intel_dump_decode
- now also handles ascii dumps like i915_ringbuffer_data
- getoptified
- totally lazy binary detector
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Daniel Vetter [Tue, 13 Sep 2011 09:05:13 +0000 (11:05 +0200)]
tests: exit(77) to properly mark tests as SKIP
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Daniel Vetter [Mon, 12 Sep 2011 18:56:13 +0000 (20:56 +0200)]
tests: mark gen3-only tests as such
... instead of hanging the gpu.
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Daniel Vetter [Mon, 12 Sep 2011 18:45:06 +0000 (20:45 +0200)]
tests: add basic test for missed irqs
Didn't catch any known bug, but can't hurt.
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Daniel Vetter [Wed, 7 Sep 2011 07:31:00 +0000 (09:31 +0200)]
tests: check pipe_control coherency
v2: proper support for gen6+
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Daniel Vetter [Mon, 12 Sep 2011 18:33:58 +0000 (20:33 +0200)]
tools: add intel_gen
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Daniel Vetter [Fri, 9 Sep 2011 18:44:27 +0000 (20:44 +0200)]
tests: basic ring<->cpu and ring<->ring tests
Using a dummy reloc that doesn't matter to trick the kernel into
synchroizing the rings.
v2: properly apply MI_NOOP workaround to MI_FLUSH_DW and
switch to MI_COND_BATCH_BUFFER_END as a dummy command on the
render ring to avoid PIPE_CONTROL errata.
v3: somebody clever decided that in C, you cound from 1,
i.e. I915_EXEC_RENDER == 1. It works now ...
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Daniel Vetter [Thu, 8 Sep 2011 08:50:37 +0000 (10:50 +0200)]
tests: split up gem_storedw_loop into ring specific variations
Also start to shortly explain testcases with an easily-greppable
header like this:
/*
* Testcase:
*
* [Possible further explanation.]
*
*/
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Daniel Vetter [Wed, 7 Sep 2011 08:18:06 +0000 (10:18 +0200)]
tests/gem_storedw: disable on pre-gen6
Doesn't work.
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Daniel Vetter [Wed, 7 Sep 2011 07:20:36 +0000 (09:20 +0200)]
lib/batch: hackish interface to run batches on specific rings
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Daniel Vetter [Tue, 6 Sep 2011 19:16:21 +0000 (21:16 +0200)]
test/gem_storedw_*: tune some more
MI_*/PIPE_CONTROL writes need to be in DOMAIN_INSTRUCTION, because
that is what mesa uses and I plan to use this to work around a
gen6 ppgtt issue.
Also testing with intentionally b0rked GFX_MODE on my snb shows that
we need to increase the loop counter a bit to reliably hit the tlb
invalidation problem. Test still completes within a few seconds.
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Daniel Vetter [Fri, 9 Sep 2011 15:22:28 +0000 (17:22 +0200)]
tools: decode some more MI_ ops
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Daniel Vetter [Fri, 9 Sep 2011 08:10:07 +0000 (10:10 +0200)]
tools: decode MI_SEMAPHORE_MBOX
Juggle the length check up to avoid duplication.
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Daniel Vetter [Tue, 6 Sep 2011 18:54:16 +0000 (20:54 +0200)]
tools/intel_gpu_dump: add hackish --devid parsing
For bugreports and whatnotelse.
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Paulo Zanoni [Wed, 24 Aug 2011 17:57:31 +0000 (14:57 -0300)]
gem_vmap_blits: fix compilation if I915_PARAM_HAS_VMAP is not defined
Previously, "make check" failed because the main() function was not
defined.
Signed-off-by: Paulo Zanoni <paulo.r.zanoni@intel.com>
Tested-By: Eugeni Dodonov <eugeni.dodonov@intel.com>
Signed-Off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Eugeni Dodonov [Mon, 5 Sep 2011 23:04:17 +0000 (20:04 -0300)]
intel_gpu_top: access hardware before running profiling command
Without this patch, intel_gpu_top will try to get access to the hardware
after running the profiling command in the background. In case such access
fails, the main process quits, but the background one will continue
running.
So let's move this test up, so it is done before we fork.
Signed-off-by: Eugeni Dodonov <eugeni.dodonov@intel.com>
Eugeni Dodonov [Mon, 5 Sep 2011 22:59:40 +0000 (19:59 -0300)]
intel_gpu_tools: describe -e option in usage screen
This patch adds a short description of '-e' option to the usage screen.
Signed-off-by: Eugeni Dodonov <eugeni.dodonov@intel.com>
Eugeni Dodonov [Mon, 5 Sep 2011 22:50:52 +0000 (19:50 -0300)]
intel_gpu_top: adopt to kernel coding style
This patch fixes adopts the code to the kernel coding_style.
Eugeni Dodonov [Mon, 5 Sep 2011 22:41:24 +0000 (19:41 -0300)]
intel_gpu_top: support profiling user-specified commands
This patch adds support for running intel_gpu_top to profile specific
commands. The required command will be carried out in separate process,
and main intel_gpu_top will leave when the child process will exit.
Signed-off-by: Eugeni Dodonov <eugeni.dodonov@intel.com>
Eugeni Dodonov [Mon, 5 Sep 2011 22:39:58 +0000 (19:39 -0300)]
intel_gpu_top: support non-interactive mode
This patch adds support for non-interactive mode, invoked by running with
'-o output' switch. In this case, no interactive output is being
performed, but the execution statistics are being saved into the output
file.
The output file is generated in both human and gnuplot-readable format.
Unlike interactive mode, where non-supported pipes and non-active
registers are skipped, the content of such pipes and registers is recorded
into the log file to simplify parsing and standardize the list of columns.
Also, unlike interactive mode, the registers are not sorted according to
the usage - this way, their variation over time can be analysed offline.
Signed-off-by: Eugeni Dodonov <eugeni.dodonov@intel.com>
Eugeni Dodonov [Mon, 5 Sep 2011 22:35:44 +0000 (19:35 -0300)]
This patch initializes the last_stats[] for registers prior to starting
the monitoring itself. This way, the first measure will already contain
the difference from the previous value instead of non-initialized value.
Signed-off-by: Eugeni Dodonov <eugeni.dodonov@intel.com>