Chris Wilson [Thu, 10 Feb 2011 10:54:39 +0000 (10:54 +0000)]
tools/intel_gpu_dump: Replace asprintf with snprintf
The maximum filename is limited so we can simply use a large buffer
rather than fail to check the error return from asprintf.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Thu, 10 Feb 2011 10:50:48 +0000 (10:50 +0000)]
tests/intel_reg_dumper: Replace asprintf with snprintf
Stops the compiler warning about not checking the potential error return
from asprintf, which was a false positive anyway.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Wed, 9 Feb 2011 10:24:39 +0000 (10:24 +0000)]
tests: Move potential machine killers away from distcheck
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Wed, 9 Feb 2011 10:21:27 +0000 (10:21 +0000)]
tests/gem_exec_blt: Compile fix
Javier Jardón [Tue, 8 Feb 2011 22:33:12 +0000 (22:33 +0000)]
build: Update autotools configuration
Replace deprecated macros and use the new libtool syntax
Also use silent-rules to make the build option less verbose.
Chris Wilson [Tue, 8 Feb 2011 12:04:17 +0000 (12:04 +0000)]
configure: Bump the libdrm dependency
Whoops, should have done this ages ago for I915_EXEC_BLT
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Sun, 6 Feb 2011 10:18:53 +0000 (10:18 +0000)]
gem_linear_blits,gem_tiled_blits: Add optional count
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Tue, 1 Feb 2011 13:35:16 +0000 (13:35 +0000)]
Fix basic blt test for SNB
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Sun, 6 Feb 2011 10:12:42 +0000 (10:12 +0000)]
Fix typo excluding Ironlake from IS_INTEL
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Sat, 5 Feb 2011 10:09:38 +0000 (10:09 +0000)]
gem_tiled_blits: Minor enhancements
Include a simple series of blits that exhaust the aperture but have the
maximum grace time between reuse.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Sat, 5 Feb 2011 10:09:38 +0000 (10:09 +0000)]
gem_linear_blits: Minor enhancements
Include a simple series of blits that exhaust the aperture but have the
maximum grace time between reuse.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Wed, 2 Feb 2011 10:32:52 +0000 (10:32 +0000)]
Simplify gem_linear_blits
By doing everything manually we can indeed exercise the full GTT.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Tue, 1 Feb 2011 19:22:32 +0000 (19:22 +0000)]
Simple test to ensure that working sets larger than memory, just work.
Based on gem_tiled_blit, but without the complication of the tiling.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Tue, 25 Jan 2011 16:34:08 +0000 (16:34 +0000)]
error decode: Search for first i915_error_state
Handy for multi-GPU systems where the IGFX may not be first.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Tue, 1 Feb 2011 13:35:36 +0000 (13:35 +0000)]
Search for the first Intel dri device.
This is vital in a multi-GPU system so that we only test the Intel card
and not the discrete GPUs.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Tue, 1 Feb 2011 11:12:40 +0000 (11:12 +0000)]
tests/gem_tiled_fence_blits: Ensure the bo count is odd
In order to avoid a self-copy, the number of bo must not be even.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Tue, 1 Feb 2011 10:53:57 +0000 (10:53 +0000)]
Exercise the whole aperture with tiled blits
After full-gtt, gem_tiled_blits doesn't allocate enough to force
eviction. So query the total aperture and accommodate.
Also introduce a similar test that utilizes fences rather than
use the BLT to perform the tiling and detiling.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Fri, 21 Jan 2011 00:43:11 +0000 (00:43 +0000)]
Time blt vs pwrite
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Thu, 20 Jan 2011 22:10:10 +0000 (22:10 +0000)]
Add a simple test to execute n nop batches
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Thu, 23 Dec 2010 12:44:51 +0000 (12:44 +0000)]
Add a GTT timing test
Using for basic benchmarking and diagnosing some mysterious slowdowns.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Sat, 29 Jan 2011 16:12:38 +0000 (16:12 +0000)]
bios: Parse SDVO panel DTDs
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Sun, 9 Jan 2011 14:01:47 +0000 (14:01 +0000)]
gpu-top: Correct computation of ring size
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Sun, 9 Jan 2011 13:26:50 +0000 (13:26 +0000)]
error: Parse ring name before gtt_offset
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Fri, 7 Jan 2011 19:59:00 +0000 (19:59 +0000)]
intel_gpu_top: Include BLT ring
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Fri, 7 Jan 2011 19:42:29 +0000 (19:42 +0000)]
intel_gpu_top: Wake the GT power well to read gen6 ring registers
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Thu, 23 Dec 2010 19:44:14 +0000 (19:44 +0000)]
error decode: print out class of chipset in the error report
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Thu, 23 Dec 2010 12:44:51 +0000 (12:44 +0000)]
Add a GTT timing test
Using for diagnosing some mysterious slowdowns. Should include a variant
for basic benchmarking...
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Jesse Barnes [Tue, 21 Dec 2010 20:06:00 +0000 (12:06 -0800)]
tools/bios_reader: try to dump more child device data
Some of our device class defines were incorrect, and we were missing
several recent ones, so fix those up and dump more child device data if
we have it.
Jesse Barnes [Tue, 21 Dec 2010 17:38:23 +0000 (09:38 -0800)]
tests: add display testing program
Just a simple program to light up all detected outputs at their native
mode and respond to hotplug events.
Eric Anholt [Fri, 17 Dec 2010 00:42:52 +0000 (16:42 -0800)]
intel_disable_clock_gating: New tool for turning off clock gating on ILK.
This is something I sometimes want to do in testing, to see if a
mystery bug (say, 29172) is due to broken clock gating. Sadly, in
this case it isn't. Note that it isn't supported on non-ILK chipsets
yet.
Eric Anholt [Tue, 16 Nov 2010 01:25:56 +0000 (09:25 +0800)]
intel_gpu_top: Print stats regs on Ironlake as well.
Chris Wilson [Fri, 10 Dec 2010 15:31:59 +0000 (15:31 +0000)]
error-decode: stdin is 0 not 1
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Sat, 4 Dec 2010 14:12:19 +0000 (14:12 +0000)]
decode: Add a couple of BLT MI.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Sat, 4 Dec 2010 11:41:30 +0000 (11:41 +0000)]
gpu-dump: Include BLT and BSD ringbuffers
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Fri, 3 Dec 2010 00:41:43 +0000 (00:41 +0000)]
decode: Add some missing protection against short buffers for gen6 instr
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Thu, 2 Dec 2010 23:05:16 +0000 (23:05 +0000)]
dump-decode: allow specification of devid, i.e. gen
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Wed, 1 Dec 2010 21:51:59 +0000 (21:51 +0000)]
error-decode: Operate as a pipe and accept input from stdin
Useful for feeding in compressed files from bugzilla:
$ bzcat /tmp/i915_error_state.bz | intel_error_decode | less
Next step would be to use gzfopen or bzfopen to automagically handle
compressed files...
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Mon, 29 Nov 2010 16:06:46 +0000 (16:06 +0000)]
decode: Improve parsing of i915 fixed-length indexed primitives
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Tue, 26 Oct 2010 09:59:18 +0000 (10:59 +0100)]
Prepare for split BLT ring on Sandybridge.
Depends on libdrm
057fab3382c02af54126ce395c43d4e6dce9439a
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=31123
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Fri, 22 Oct 2010 10:33:08 +0000 (11:33 +0100)]
Decode PGTBL_ER for i965
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Fri, 1 Oct 2010 21:22:28 +0000 (22:22 +0100)]
intel_gpu_top: Sample BSD rings as well as render
Show how busy the GPU is when decoding video as well as rendering.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Zhenyu Wang [Wed, 29 Sep 2010 01:47:46 +0000 (09:47 +0800)]
decode: fix PS sample state change bit on sandybridge
Signed-off-by: Zhenyu Wang <zhenyuw@linux.intel.com>
Zhenyu Wang [Mon, 27 Sep 2010 05:34:52 +0000 (13:34 +0800)]
intel_decode: decode 3d commands for sandybridge
Signed-off-by: Zhenyu Wang <zhenyuw@linux.intel.com>
Zhenyu Wang [Mon, 27 Sep 2010 05:24:52 +0000 (13:24 +0800)]
Add all sandybridge device ids
Signed-off-by: Zhenyu Wang <zhenyuw@linux.intel.com>
Jesse Barnes [Mon, 20 Sep 2010 21:37:07 +0000 (14:37 -0700)]
intel_reg_dumper: eDP port is on the CPU, not PCH
Made me think there was another register until I checked the offset.
Adam Jackson [Wed, 8 Sep 2010 18:39:12 +0000 (14:39 -0400)]
bios_reader: Fix to work on non-Intel machines
Actually, fix to work on _any_ machine where the ROM file doesn't match
the GPU. Just extract the device ID from the ROM itself rather than
look at the running system.
Signed-off-by: Adam Jackson <ajax@redhat.com>
Chris Wilson [Sun, 5 Sep 2010 11:05:52 +0000 (12:05 +0100)]
reg dumper: Dump ILK panel fitting control debug registers
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Sun, 5 Sep 2010 09:53:12 +0000 (10:53 +0100)]
reg dumper: Fix ilk panel fitter filter shift
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Wed, 25 Aug 2010 10:56:59 +0000 (11:56 +0100)]
tests: Fix fallout from s/EBADF/ENOENT/ in 2.6.36
After
bf79cb914dbfe848add8bb76cbb8ff89110d29ff, drm uses ENOENT to
report unknown handles buffer objects, update the tests accordingly.
Fixes:
Bug 29794 - some intel_gpu_tools cases fail
https://bugs.freedesktop.org/show_bug.cgi?id=29794
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Wed, 25 Aug 2010 10:55:25 +0000 (11:55 +0100)]
decode: Handle errors during parsing.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Wed, 25 Aug 2010 10:54:54 +0000 (11:54 +0100)]
reg dumper: Add panel backlight detailed information
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Eric Anholt [Mon, 19 Jul 2010 15:27:32 +0000 (08:27 -0700)]
intel_gpu_top: Fix the "limit printing to terminal height" again.
Eric Anholt [Fri, 9 Jul 2010 20:26:57 +0000 (13:26 -0700)]
intel_decode: Fix decode for gen6's moved VERTEX_ELEMENTS valid bit.
Chris Wilson [Thu, 15 Jul 2010 18:31:21 +0000 (19:31 +0100)]
dump decode: Add a non-installed program for decoding binary batchbuffer dumps.
Rather than add yet another decoder into the drivers, I rather just emit
a binary blob and use the decoder being developed here.
As it is a pure debugging feature, lets not package it.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Thu, 15 Jul 2010 18:18:39 +0000 (19:18 +0100)]
error: Decode i915 PGTBL errors
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Eric Anholt [Thu, 8 Jul 2010 19:13:08 +0000 (12:13 -0700)]
intel_decode: Fix whitespace for importing to Mesa.
Jesse Barnes [Wed, 30 Jun 2010 09:02:49 +0000 (02:02 -0700)]
intel_reg_dumper: add some 945 MI reg dumping
Adam Jackson [Wed, 30 Jun 2010 15:18:00 +0000 (11:18 -0400)]
Use longer string padding in intel_reg_dumper on ironlake
DIGITAL_PORT_HOTPLUG_CNTRL is 26 chars. Bump out to 30.
Signed-off-by: Adam Jackson <ajax@redhat.com>
Adam Jackson [Wed, 5 May 2010 15:46:27 +0000 (11:46 -0400)]
fix reg_dumper on systems without intel gpus
Chris Wilson [Tue, 15 Jun 2010 12:05:48 +0000 (13:05 +0100)]
intel_reg_dump: Off-by-one in calculation of i965 fence pitch
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Zhenyu Wang [Thu, 3 Jun 2010 02:31:04 +0000 (10:31 +0800)]
bios_reader: Fix SSC frequency parse for Ironlake/Sandybridge
Signed-off-by: Zhenyu Wang <zhenyuw@linux.intel.com>
Wu Fengguang [Mon, 19 Apr 2010 05:13:06 +0000 (13:13 +0800)]
intel_audio_dump: add support for IbexPeak and CougarPoint
Signed-off-by: Wu Fengguang <fengguang.wu@intel.com>
Signed-off-by: Eric Anholt <eric@anholt.net>
Adam Jackson [Thu, 22 Apr 2010 19:37:49 +0000 (15:37 -0400)]
bios reader: Decode device type from child device table
Signed-off-by: Adam Jackson <ajax@redhat.com>
Chris Wilson [Fri, 16 Apr 2010 21:23:44 +0000 (22:23 +0100)]
decode; Parse 3DSTATE_BUFFER_INFO
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Thu, 15 Apr 2010 08:53:00 +0000 (09:53 +0100)]
decode: Also override S2 for 3DPRIM_CLEAR_RECT
3DPRIM_CLEAR_RECT implicitly uses only VERTEX_XY and no texture coords.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Thu, 15 Apr 2010 08:48:00 +0000 (09:48 +0100)]
decode: Use a valid S4 mask [VERTEX_XY] for 3DPRIM_CLEAR_RECT
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Zhenyu Wang [Thu, 15 Apr 2010 15:22:06 +0000 (23:22 +0800)]
update PCH HDMI debug info
Signed-off-by: Zhenyu Wang <zhenyuw@linux.intel.com>
Zhenyu Wang [Wed, 14 Apr 2010 17:25:50 +0000 (01:25 +0800)]
reg dump update for SNB/CPT
Signed-off-by: Zhenyu Wang <zhenyuw@linux.intel.com>
Zhenyu Wang [Thu, 15 Apr 2010 14:57:02 +0000 (22:57 +0800)]
Add PCH chipset type check for Cougarpoint
Signed-off-by: Zhenyu Wang <zhenyuw@linux.intel.com>
Chris Wilson [Sun, 4 Apr 2010 23:26:35 +0000 (00:26 +0100)]
decode: There are 4 unpreserved registers on i915, not 3.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Wed, 14 Apr 2010 21:30:58 +0000 (22:30 +0100)]
decode: CLEAR_RECT uses a fixed VERTEX_XY format.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Wed, 14 Apr 2010 21:25:57 +0000 (22:25 +0100)]
decode: CLEAR_PARAMETERS has a fixed length bias of 5.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Thu, 8 Apr 2010 11:00:00 +0000 (12:00 +0100)]
gitignore: Add missing intel_reg_snapshot
Chris Wilson [Thu, 8 Apr 2010 10:56:57 +0000 (11:56 +0100)]
Enable compilation on non-Intel, non-DRM systems.
A few of the tools can be performed post-mortem from a different system,
so it is useful to be able to compile those tools on those foreign
systems. Obviously, any program to interact with the PCI device or talk
to GEM will fail on a non-Intel system.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Adam Jackson [Wed, 31 Mar 2010 21:25:23 +0000 (17:25 -0400)]
intel_reg_dumper: Add support for reading register dumps from files
Also add intel_reg_snapshot for creating such snapshots, and relevant
documentation.
Signed-off-by: Adam Jackson <ajax@redhat.com>
Chris Wilson [Sun, 4 Apr 2010 23:26:35 +0000 (00:26 +0100)]
decode: There are 4 unpreserved registers on i915, not 3.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Eric Anholt [Wed, 24 Mar 2010 19:01:00 +0000 (12:01 -0700)]
Fix INSTDONE1 bits on g4x, and use those on Ironlake too.
Eric Anholt [Wed, 24 Mar 2010 18:40:07 +0000 (11:40 -0700)]
Add Ironlake INSTDONE bits.
Eric Anholt [Tue, 9 Mar 2010 06:51:40 +0000 (22:51 -0800)]
intel_reg_dumper: Fix the i915 fence reg decode to not be pure lies.
Chris Wilson [Sun, 21 Mar 2010 19:33:45 +0000 (19:33 +0000)]
decode: Parse i915 MAP_STATE texture sizes and tiling
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Thu, 4 Mar 2010 21:56:04 +0000 (21:56 +0000)]
intel_error_decode: Pretty print i830 PGTBL_ER
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Thu, 4 Mar 2010 09:14:56 +0000 (09:14 +0000)]
decode: Fix length mask for LOAD_REGISTER_IMM.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Eric Anholt [Thu, 25 Feb 2010 20:48:00 +0000 (12:48 -0800)]
intel_gpu_top: Print stats lines even when there's no done bit on the line.
Eric Anholt [Thu, 25 Feb 2010 20:18:11 +0000 (12:18 -0800)]
intel_gpu_top: Print the vertex/primitive statistics on gen4+.
This requires that the window be much wider. I'm not sure how we
really want to lay out this interface usably, but so far giant windows
are working well enough for me.
Chris Wilson [Thu, 25 Feb 2010 18:52:15 +0000 (18:52 +0000)]
man: Add manpage for intel_error_decode
A rudimentary manpage based on intel_gpu_dump for the simple tool
for the post-processing of i915_error_state, intel_error_decode.
Eric Anholt [Wed, 24 Feb 2010 01:04:29 +0000 (17:04 -0800)]
Fix a typo in intel_gtt(1).
Eric Anholt [Wed, 24 Feb 2010 01:00:54 +0000 (17:00 -0800)]
Add manpages for a bunch of recently-added tools.
Eric Anholt [Mon, 1 Feb 2010 18:47:19 +0000 (10:47 -0800)]
Add decode for more Sandybridge packets.
Eric Anholt [Fri, 15 Jan 2010 22:28:16 +0000 (14:28 -0800)]
Add support for Sandybridge INSTDONE regs.
Eric Anholt [Fri, 15 Jan 2010 22:28:02 +0000 (14:28 -0800)]
Add support for Sandybridge mobile chipset.
Eric Anholt [Fri, 15 Jan 2010 22:27:41 +0000 (14:27 -0800)]
intel_gpu_dump: More SNB packets.
Eric Anholt [Fri, 15 Jan 2010 22:26:43 +0000 (14:26 -0800)]
intel_gpu_dump: Use the spec's names for the INSTDONE regs printed out.
Eric Anholt [Thu, 17 Dec 2009 22:21:28 +0000 (14:21 -0800)]
Add Sandybridge decode for CMD_VIEWPORT_STATE_POINTERS
Eric Anholt [Tue, 15 Dec 2009 20:38:23 +0000 (12:38 -0800)]
Add Sandybridge STATE_BASE_ADDRESS decode.
Eric Anholt [Wed, 4 Nov 2009 00:15:25 +0000 (16:15 -0800)]
intel_gpu_dump: Add support for sandybridge DEPTH_BUFFER and BINDING_TABLEs.
Eric Anholt [Thu, 22 Oct 2009 22:52:02 +0000 (15:52 -0700)]
Add some initial definitions for Sandybridge.
Chris Wilson [Thu, 25 Feb 2010 11:05:16 +0000 (11:05 +0000)]
decode: Parse MI_BATCH_BUFFER for i8xx
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Eric Anholt [Fri, 19 Feb 2010 17:23:16 +0000 (12:23 -0500)]
Add intel_bios_dumper from the 2D driver. Last of the tools, finally!
Eric Anholt [Fri, 19 Feb 2010 17:12:04 +0000 (12:12 -0500)]
Move the intel_lid tool from the 2D driver to here.
Chris Wilson [Fri, 12 Feb 2010 13:28:40 +0000 (13:28 +0000)]
Add intel_error_decode.
A simple variant of intel_gpu_dump that explicitly handles parsing of
i915_error_state.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>