Imre Deak [Mon, 22 Oct 2012 17:40:07 +0000 (20:40 +0300)]
flip_test: add comment about the race between flip vs. vblank events
Signed-off-by: Imre Deak <imre.deak@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Imre Deak [Mon, 22 Oct 2012 17:40:06 +0000 (20:40 +0300)]
flip_test: check if the vblank and flip states correlate
Signed-off-by: Imre Deak <imre.deak@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Imre Deak [Mon, 22 Oct 2012 17:40:05 +0000 (20:40 +0300)]
flip_test: fixup zero timestamp for premature vblanks
Signed-off-by: Imre Deak <imre.deak@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Imre Deak [Mon, 22 Oct 2012 17:40:04 +0000 (20:40 +0300)]
flip_test: increase duration for vbl+flip tests with ts check
We need a big enough duration, so the hotplug event's affect doesn't
distort our final frame count too much.
Without this these tests will exit with a "dropped frames" error.
Signed-off-by: Imre Deak <imre.deak@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Daniel Vetter [Thu, 18 Oct 2012 20:37:20 +0000 (22:37 +0200)]
tests/gem_cs_tlb: make it more robust
... using a fine roll of duct-tape. Dunno what exactly leaks out from
the previous testcase, but the bo gets moved about a bit and makes the
self-check of the test fail.
Daniel Vetter [Wed, 17 Oct 2012 09:09:13 +0000 (11:09 +0200)]
flip_test: check whether wait_vblank fails when it should
Daniel Vetter [Tue, 16 Oct 2012 22:35:43 +0000 (00:35 +0200)]
flip_test: drop vblank vs. pan
pageflip vs. pan is interesting since we wait for outstanding flips
before updating the fb base address. For vblank absolutely nothing
interesting happens.
Daniel Vetter [Tue, 16 Oct 2012 22:34:43 +0000 (00:34 +0200)]
flip_test: tune test-durations a bit
15 seconds for timestamp checkers (to at least hit the dreaded hotplug
once), 5 seconds for fast tests that check every vblank something, 30
for slow test (dpms/modeset cycles).
Imre Deak [Tue, 16 Oct 2012 02:43:19 +0000 (05:43 +0300)]
flip_test: add wait-for-vblank tests
Signed-off-by: Imre Deak <imre.deak@intel.com>
[danvet: add {} for multi-line but single-statement if block.]
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Imre Deak [Tue, 16 Oct 2012 01:11:43 +0000 (04:11 +0300)]
flip_test: add check to see if any event has occured
At the moment select() would time out in case we don't get any event.
When we add vblank events in a later patch, it's possible that we
receive one type of events but not the other type. In this case select()
doesn't time out and we need another way to detect this.
Signed-off-by: Imre Deak <imre.deak@intel.com>
Imre Deak [Tue, 16 Oct 2012 01:09:43 +0000 (04:09 +0300)]
flip_test: add event sequence number tracking
Signed-off-by: Imre Deak <imre.deak@intel.com>
Imre Deak [Mon, 15 Oct 2012 22:47:15 +0000 (01:47 +0300)]
flip_test: add logic to track pending events
When we add the vblank tests later, it can happen that one type of event
is pending while we are servicing another. Prepare for that by
maintaining a pending/completed event mask.
No functional change.
Signed-off-by: Imre Deak <imre.deak@intel.com>
Imre Deak [Mon, 15 Oct 2012 22:47:15 +0000 (01:47 +0300)]
flip_test: make page flip tests conditional
Since later we want to add vblank tests where we don't want to flip,
make the flips and corresponding state checks conditional.
No functional change.
Signed-off-by: Imre Deak <imre.deak@intel.com>
Imre Deak [Mon, 15 Oct 2012 18:18:40 +0000 (21:18 +0300)]
flip_test: unify the name of the current test in status messages
Signed-off-by: Imre Deak <imre.deak@intel.com>
Imre Deak [Mon, 15 Oct 2012 18:15:18 +0000 (21:15 +0300)]
flip_test: store the pipe id in the test context obj
For DRM_IOCTL_WAIT_VBLANK we'll have to pass the pipe id - as opposed to
the crtc id required by the rest of the IOCTLs we use.
Thanks for Chris Wilson for pointing this out.
No functional change.
Signed-off-by: Imre Deak <imre.deak@intel.com>
Imre Deak [Tue, 16 Oct 2012 13:18:08 +0000 (16:18 +0300)]
flip_test: don't skip checks for sequence #1
So far we skipped some tests for seq#0 and #1, since at that point we
were missing the 'last' state against which we could compare the current
state. Since in the previous patches we fixed the ordering in the test
loop and moved the increment of count to the update state phase, we have
a proper 'last' state for seq#1, so enable the tests already at that
point.
Signed-off-by: Imre Deak <imre.deak@intel.com>
Imre Deak [Mon, 15 Oct 2012 17:37:01 +0000 (20:37 +0300)]
flip_test: factor out the event state
Factor out the status vars that are specific to the flip events into a
separate structure. This will allow us tracking the vblank events with
the same code.
We move the increment of the flip event counter to the end of the test
loop, but this shouldn't affect anything. This is the only functional
change.
Signed-off-by: Imre Deak <imre.deak@intel.com>
Imre Deak [Mon, 15 Oct 2012 23:41:02 +0000 (02:41 +0300)]
flip_test: swap the order of check state/run test step
At the moment we first check the state then run the test step in our
test loop. Swapping the order makes the starting state of each iteration
better defined, allowing an easier extension of these steps in the
future.
Since now it's guaranteed that we exit the event loop with no pending
flips, we can also get rid of the final flushing of events.
We don't want the first initializing flip to affect the test loop other
than setting an initial FB, so before starting the test loop wait for it
to complete by calling wait_for_events() and leave the flip event
counter at zero.
Signed-off-by: Imre Deak <imre.deak@intel.com>
Imre Deak [Mon, 15 Oct 2012 23:37:15 +0000 (02:37 +0300)]
flip_test: split the flip handler into logical parts
The handler consits of handle_event/run_test/check_state/update_state
logical steps, split the function accordingly. This is needed by the
following patches that need to do part of these steps for both flip and
vblank events.
No functional change.
Signed-off-by: Imre Deak <imre.deak@intel.com>
Imre Deak [Mon, 15 Oct 2012 23:26:35 +0000 (02:26 +0300)]
flip_test: store current flip/received timestamps in the context obj
This is needed by the next patch that splits the flip handler function
into logical parts. Make the timestamps accesible to these parts.
No functional change.
Signed-off-by: Imre Deak <imre.deak@intel.com>
Imre Deak [Tue, 16 Oct 2012 02:50:22 +0000 (05:50 +0300)]
flip_test: factor out the final state check
Needed by an upcoming patch where we want to make a final state check
for both the flip and vblank events.
No functional change.
Signed-off-by: Imre Deak <imre.deak@intel.com>
Imre Deak [Fri, 12 Oct 2012 10:31:25 +0000 (13:31 +0300)]
flip_test: factor out the event loop/wait for event logic
Needed by an upcoming patch where we want to wait for an event without
starting a new round of test run.
No functional change.
Signed-off-by: Imre Deak <imre.deak@intel.com>
Imre Deak [Wed, 10 Oct 2012 11:47:51 +0000 (14:47 +0300)]
flip_test: move output panning inside the flip_handler
Move the panning to a more logical place where the rest of the test
steps are performed. As Daniel Vetter pointed it out, the proper place
is _after_ the flip command is sent, so that fb_set_base correctly waits
for any outstanding flip. So move the function between the flip and the
dpms/modeset off calls.
Signed-off-by: Imre Deak <imre.deak@intel.com>
Imre Deak [Wed, 10 Oct 2012 09:01:28 +0000 (12:01 +0300)]
flip_test: store fb width, height in test context object
We will need these in event handlers, so store them where the handlers
have access to them.
No functional change.
Signed-off-by: Imre Deak <imre.deak@intel.com>
Imre Deak [Tue, 16 Oct 2012 14:34:41 +0000 (17:34 +0300)]
flip_test: factor out drmModePageFlip
For better readability and to prepare for the upcoming patch marking
pending flip events with a flag.
No functional change.
Signed-off-by: Imre Deak <imre.deak@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Imre Deak [Tue, 16 Oct 2012 14:34:39 +0000 (17:34 +0300)]
test_flip: fix checking for delayed event reception
The intent for the time limit seems to be 2ms, but the current condition
will result in a 1s limit and makes the check against tv_usec redundant.
Fix the condition to check for a 2ms limit.
Signed-off-by: Imre Deak <imre.deak@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Imre Deak [Tue, 16 Oct 2012 14:34:38 +0000 (17:34 +0300)]
flip_test: check drmHandleEvents()' return value
Signed-off-by: Imre Deak <imre.deak@intel.com>
[danvet: s/assert/do_or_die]
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Imre Deak [Tue, 16 Oct 2012 14:34:37 +0000 (17:34 +0300)]
flip_test: reset the state for each test run
Each test run needs a clean state.
Signed-off-by: Imre Deak <imre.deak@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Imre Deak [Tue, 16 Oct 2012 14:34:36 +0000 (17:34 +0300)]
flip_test: free FBs after each test run
Though the FBs will be freed implicitly at process exit, with the
growing number of test cases it's probably better not to let them
accumulate and free them as soon as we can.
Signed-off-by: Imre Deak <imre.deak@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Imre Deak [Tue, 16 Oct 2012 14:34:35 +0000 (17:34 +0300)]
drmtest: add function to remove an DRM FB
Signed-off-by: Imre Deak <imre.deak@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Ben Widawsky [Mon, 15 Oct 2012 17:05:56 +0000 (10:05 -0700)]
sysfs_rps: Add negative test case for bad values
Signed-off-by: Ben Widawsky <ben@bwidawsk.net>
Daniel Vetter [Fri, 12 Oct 2012 08:45:52 +0000 (10:45 +0200)]
lib/mmio: gen2 has a bigger register window, too
It's split up into two parts, with the gtt pte range in between.
Without this I couldn't read the ADPA register ...
Imre Deak [Wed, 10 Oct 2012 13:04:45 +0000 (16:04 +0300)]
fix warn in testdisplay: ignoring return value of 'chdir'
Signed-off-by: Imre Deak <imre.deak@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Imre Deak [Wed, 10 Oct 2012 13:04:44 +0000 (16:04 +0300)]
fix warn in prime_nv_*: ignoring return value of ‘fgets’
Signed-off-by: Imre Deak <imre.deak@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Imre Deak [Wed, 10 Oct 2012 13:04:43 +0000 (16:04 +0300)]
fix warn in gem_reg_read: 'read' shadows a global declaration
Signed-off-by: Imre Deak <imre.deak@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Imre Deak [Wed, 10 Oct 2012 13:04:42 +0000 (16:04 +0300)]
fix warn in sysfs_{rc6, rps}*: ignoring return value of 'fscanf'
Signed-off-by: Imre Deak <imre.deak@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Imre Deak [Wed, 10 Oct 2012 13:04:41 +0000 (16:04 +0300)]
fix warn in gem_readwrite: 'read/write' shadows a global declaration
Signed-off-by: Imre Deak <imre.deak@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Imre Deak [Wed, 10 Oct 2012 13:04:40 +0000 (16:04 +0300)]
fix warn in intel_reg_snapshot: ignoring return value of 'write'
Signed-off-by: Imre Deak <imre.deak@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Imre Deak [Wed, 10 Oct 2012 13:04:39 +0000 (16:04 +0300)]
fix warn in intel_error_decode: ignoring return value of 'asprintf'
Signed-off-by: Imre Deak <imre.deak@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Imre Deak [Wed, 10 Oct 2012 13:04:38 +0000 (16:04 +0300)]
fix warn: 'div' shadows a global declaration
Signed-off-by: Imre Deak <imre.deak@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Daniel Vetter [Tue, 9 Oct 2012 20:47:59 +0000 (22:47 +0200)]
tests/gem_exec_big: fixups
- don't emit a self-modifying batch, the kernel doesn't like those
- also change the buffer size, to exercise all corner-cases of the
sg_table stuff
Daniel Vetter [Tue, 9 Oct 2012 17:43:28 +0000 (19:43 +0200)]
tests/gem_exec_big: add a reloc at the end of the big batch
Since that's the other place where the broken get_page blows up.
Actually it seems to only blow up on specific pages ...
Daniel Vetter [Tue, 9 Oct 2012 17:29:02 +0000 (19:29 +0200)]
tests: add big batchbuffer test
... and use it in the hangman to stress the error_state capture a bit more.
Uncovered a bug in the obj->pages ->obj->sg_list rework.
Daniel Vetter [Tue, 9 Oct 2012 17:27:41 +0000 (19:27 +0200)]
lib: remove auto-quiescenting of the gpu
In retrospective, this is an ugly idea. Any tests that needs this
can call it themselves.
Daniel Vetter [Tue, 9 Oct 2012 15:50:40 +0000 (17:50 +0200)]
lib: make drmtest_progress more robust
Daniel Vetter [Tue, 9 Oct 2012 15:32:00 +0000 (17:32 +0200)]
tests: update .gitignore
I've thought I've fixed this, but looks like I've failed to commit.
Daniel Vetter [Tue, 9 Oct 2012 15:30:59 +0000 (17:30 +0200)]
tests: add gem_cs_tlb
Should test tlb coherency, but somehow it fails to hit the bug I want
it to hit ... needs more work.
Daniel Vetter [Tue, 9 Oct 2012 13:51:13 +0000 (15:51 +0200)]
tests/gem_cs_prefetch: clarify what it tests
I've forgotten ;-)
Daniel Vetter [Thu, 4 Oct 2012 19:40:54 +0000 (21:40 +0200)]
tests/flip_test: check -EBUSY and -EINVAL behaviour
We want -EBUSY for a pending flip and -EINVAL if the pipe is off
(either dpms off or completely off). With the small exception that
someone thought it would be funny to return -EBUSY when the crtc is
fully off.
Daniel Vetter [Thu, 4 Oct 2012 11:22:16 +0000 (13:22 +0200)]
tests/*storedw*: add subcases with more outstanding batches
Just in case this is better at hitting seqno signalling races ...
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Daniel Vetter [Wed, 3 Oct 2012 21:16:18 +0000 (23:16 +0200)]
flip_test: fix timestamp order
When grabbing the timestamp after a flip, it has a decent chance to
actually be after the flip. Surprising.
Daniel Vetter [Wed, 3 Oct 2012 16:57:40 +0000 (18:57 +0200)]
tests/gem_caching: only disable on vintage i965g/gm
Hooray for confusing macros!
Acked-by: Chris Wilson <chris@chris-wilson.co.uk>
Daniel Vetter [Wed, 3 Oct 2012 11:44:30 +0000 (13:44 +0200)]
tests: add drm_threaded_access_tiled
This catches parallel access to bo->virtual causing sigbus
Signed-off-by: Mika Kuoppala <mika.kuoppala@intel.com>
[danvet: applied some bikeshed
- changed prefix from drm_ to gem_, it's a kernel gem test
- added autohell magic to link with pthreads
- .gitignore entry
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Daniel Vetter [Tue, 2 Oct 2012 19:29:57 +0000 (21:29 +0200)]
flip_test: add an explicit flag for timing checks
We'll need this for the wait_vblank test cases.
Chris Wilson [Tue, 2 Oct 2012 20:40:46 +0000 (21:40 +0100)]
gem_cpu_reloc: Do another pass explicitly moving to the CPU write domain
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Tue, 2 Oct 2012 20:04:36 +0000 (21:04 +0100)]
gem_cpu_reloc: Use the mappable aperture size!
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Tue, 2 Oct 2012 20:02:22 +0000 (21:02 +0100)]
gem_cpu_reloc: And run the test in reverse to check with already bound batches
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Tue, 2 Oct 2012 19:50:17 +0000 (20:50 +0100)]
gem_cpu_reloc: Fix for running on SNB+
I work with these everyday and I still made a basic mistake.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Daniel Vetter [Tue, 2 Oct 2012 17:42:11 +0000 (19:42 +0200)]
flip_test: increase runtime of the plain flip to 15s
This ensures that we'll hit the stupid hpd poll handler in every run.
Also don't yell around about jitter for tv outputs ...
Daniel Vetter [Tue, 2 Oct 2012 17:33:30 +0000 (19:33 +0200)]
tests/flip_test: measure inter-frame ts jitter
Yell if it's wrong.
For some odd reason this blows up on my snb. And always on the same
o->count frame on the 2nd crtc ... And we have to thank the hpd poll
helper for that. Comment explaining this added, also made the error
non-fatal.
Daniel Vetter [Tue, 2 Oct 2012 15:38:04 +0000 (17:38 +0200)]
flip_test: race against modesets, too
Even slower than dpms switching ;-)
Chris Wilson [Tue, 2 Oct 2012 18:22:37 +0000 (19:22 +0100)]
gem_wait_render_timeout: Fix SKIP return code
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Tue, 2 Oct 2012 16:27:16 +0000 (17:27 +0100)]
tests: Add gem_cpu_reloc
Attempt to stress test performing relocations whilst the batch is in the
CPU domain.
A freshly allocated buffer starts in the CPU domain, and the pwrite
should also be performed whilst in the CPU domain and so we should
execute the relocations within the CPU domain. If for any reason one of
those steps should land it in the GTT domain, we take the secondary
precaution of filling the mappable portion of the GATT.
In order to detect whether a relocation fails, we first fill a target
buffer with a sequence of invalid commands that would cause the GPU to
immediate hang, and then attempt to overwrite them with a legal, if
short, batchbuffer using a BLT. Then we come to execute the bo, if the
relocation fail and we either copy across all zeros or garbage, then the
GPU will hang.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Daniel Vetter [Tue, 2 Oct 2012 15:22:09 +0000 (17:22 +0200)]
tests/flip_test: also test pannning with dummy load
Needs a bit more relaxed select timeout to work (which is not
required when testing dummy_load vs. dpms, since the dpms forces the
sync, not the select timeout).
Daniel Vetter [Tue, 2 Oct 2012 15:17:40 +0000 (17:17 +0200)]
tests/flip_test: test interaction with panning
Since panning with set_crtc is synchronous, we need to stall for any
outstanding pageflips. This new testcase exercise that code. Unfortunately
we still need eyes to check whether we don't loose the offset :(
Daniel Vetter [Tue, 2 Oct 2012 15:01:06 +0000 (17:01 +0200)]
tests/flip_test: robustifications
- don't yell around about dropped frames on tv connectors (and explain
why in a comment)
- wait a bit when using a test config that checks for dropped frames before
starting. Also allow for 1% of fudge, this makes it reliably work
- make the dummy load more variable, took too long on older machines.
Daniel Vetter [Tue, 2 Oct 2012 14:09:00 +0000 (16:09 +0200)]
tests/flip_test: don't complain about timestamps in the future
those can happen due to the scanline vblank timestamp adjuster ...
Daniel Vetter [Tue, 2 Oct 2012 13:59:58 +0000 (15:59 +0200)]
tests/flip_test: actually try to check what I want to check
/me sucks
Daniel Vetter [Tue, 2 Oct 2012 13:42:34 +0000 (15:42 +0200)]
tests/flip_test: check whether timestamp isn't in the past
This is mostly useful for when we add support for other clock sources,
e.g. clock_monotic, to check whether the kernel picked the right
clock.
Daniel Vetter [Tue, 2 Oct 2012 13:37:24 +0000 (15:37 +0200)]
tests/flip_test: check whether the pageflip event took too long
We're suspecting that something is fishy with the event deliver/vblank
timestamp handling on gmch platforms. Unfortunately, this isn't it.
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Chris Wilson [Tue, 2 Oct 2012 11:59:16 +0000 (12:59 +0100)]
flip_test: count the number of flips and check we don't drop any
In the basic case of just flipping between two buffers with no
additional work, we should neve drop a frame.
Daniel Vetter [Tue, 2 Oct 2012 09:50:20 +0000 (11:50 +0200)]
tests/flip_test: re-enable disabled testcases
oops
Daniel Vetter [Tue, 2 Oct 2012 08:33:13 +0000 (10:33 +0200)]
tests/flip_test: new testcase with a busy load before the flip
I've hoped that this would be good enough to hang pch platforms, but
it looks like those just complete the flip immediately on disabled pipes.
Daniel Vetter [Tue, 2 Oct 2012 08:32:25 +0000 (10:32 +0200)]
fixup .gitignore for newly add sysfs_rps test
Daniel Vetter [Tue, 2 Oct 2012 08:32:10 +0000 (10:32 +0200)]
lib: helper to convert gem_hanlde to drm_intel_bo
Daniel Vetter [Mon, 1 Oct 2012 09:34:10 +0000 (11:34 +0200)]
drmtest: use do_or_die
Chris Wilson [Fri, 28 Sep 2012 12:03:02 +0000 (13:03 +0100)]
tests/flip_test: Revamp to demonstrate race with modeset/dpms and pageflips
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Damien Lespiau [Wed, 26 Sep 2012 17:17:53 +0000 (18:17 +0100)]
intel_infoframes: Dump HDMI vendor infoframes
Those infoframes are programmed when using stereo 3D modes.
Signed-off-by: Damien Lespiau <damien.lespiau@intel.com>
Reviewed-by: Rodrigo Vivi <rodrigo.vivi@gmail.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Yi Sun [Tue, 25 Sep 2012 09:13:17 +0000 (17:13 +0800)]
Add option -o <number>, which can test only one specified mode.
Each mode line has a number just like '[i]'. So we can only test the specified mode with giving the number of mode to '-o' parameter.
Signed-off-by: Yi Sun <yi.sun@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Chris Wilson [Tue, 18 Sep 2012 13:04:58 +0000 (14:04 +0100)]
tests/module_reload: Silence warnings when unloaded shared modules
So long as we spit out critical errors during i915 reload, or detect any
oops the test will do its job.
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=55040
Ben Widawsky [Sat, 8 Sep 2012 01:12:07 +0000 (18:12 -0700)]
sysfs rps test added
Signed-off-by: Ben Widawsky <ben@bwidawsk.net>
Chris Wilson [Thu, 13 Sep 2012 14:05:34 +0000 (15:05 +0100)]
gem_gtt_speed: Add a test that caches the mmapping across iterations
Repeat the memset streaming performance test on the same mapping so that
we can factor out the overhead of establishing the GTT/CPU mmaps.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Thu, 13 Sep 2012 13:55:53 +0000 (14:55 +0100)]
gem_gtt_speed: Use a memset() to test streaming performance
This make the reasonable assumption that the libc code for memset() can
saturate the memory bandwidth -- at any rate it should do better than
the copy.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Ben Widawsky [Thu, 13 Sep 2012 00:48:21 +0000 (17:48 -0700)]
intel_sprite_on: plane needs 16.16 width and height
This only ever worked because we used to have a bug in our driver which
was fixed months ago by:
commit
b4db1e35ac59c144965f517bc575a0d75b60b03f
Author: Jesse Barnes <jbarnes@virtuousgeek.org>
Date: Tue Mar 20 10:59:09 2012 -0700
drm/i915: treat src w & h as fixed point in sprite handling code
Reported-by: Armin Reese <armin.c.reese@intel.com>
Signed-off-by: Ben Widawsky <ben@bwidawsk.net>
Ben Widawsky [Thu, 13 Sep 2012 00:27:08 +0000 (17:27 -0700)]
intel_sprite_on: coding style fixes
It was driving me nuts...
Signed-off-by: Ben Widawsky <ben@bwidawsk.net>
Ben Widawsky [Sat, 8 Sep 2012 20:40:04 +0000 (13:40 -0700)]
intel_sprite_on: Idententation + Copyright fix
Signed-off-by: Ben Widawsky <ben@bwidawsk.net>
Damien Lespiau [Fri, 7 Sep 2012 10:42:20 +0000 (11:42 +0100)]
tools: Update gitignore with intel_dpio_{read, write}
Signed-off-by: Damien Lespiau <damien.lespiau@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Damien Lespiau [Mon, 3 Sep 2012 15:16:29 +0000 (16:16 +0100)]
intel_reg_dumper: Add more information when dumping single registers
Now that we can dump registers giving a partial name, adding more
information about the dumped registers seems useful.
Signed-off-by: Damien Lespiau <damien.lespiau@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Damien Lespiau [Mon, 3 Sep 2012 15:16:28 +0000 (16:16 +0100)]
intel_reg_dumper: Allow partial register names on the command line
Let people give just a part of the register name. Handy when not
remembering the exact name or if the register is defined with a
different name than the one in the spec being looked at.
Signed-off-by: Damien Lespiau <damien.lespiau@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Damien Lespiau [Mon, 3 Sep 2012 15:16:27 +0000 (16:16 +0100)]
intel_reg_dumper: Also decode registers given by address
One can now give an address instead of a register name to decode a
single register.
Signed-off-by: Damien Lespiau <damien.lespiau@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Damien Lespiau [Mon, 3 Sep 2012 15:16:26 +0000 (16:16 +0100)]
intel_reg_dumper: Add a single register decode mode
From time to time, one would like to decode a register value that have
been captured at a certain point in time (and say printed out with a
printk). intel_reg_dumper has all the knowledge to do that and this
patch adds a way to ask it to decode a value.
Example usage:
$ ./tools/intel_reg_dumper PCH_PP_CONTROL 0xabcd0002
PCH_PP_CONTROL: 0xabcd0002 (blacklight disabled, power...
v2: friendlier invocation (Chris Wilson)
v3: remove unecessary casts and use strcasecmp (Jani Nikula)
Signed-off-by: Damien Lespiau <damien.lespiau@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Ben Widawsky [Thu, 30 Aug 2012 21:17:01 +0000 (14:17 -0700)]
clang: Fix static analysis warnings from clang
Signed-off-by: Ben Widawsky <ben@bwidawsk.net>
Ben Widawsky [Thu, 30 Aug 2012 21:03:54 +0000 (14:03 -0700)]
clang: Fix warnings found through clang.
Signed-off-by: Ben Widawsky <ben@bwidawsk.net>
Chris Wilson [Mon, 16 Jul 2012 15:41:16 +0000 (16:41 +0100)]
tests/gem_cacheing: Exercise snoop coherency
This is based on tests/gem_partial_pwrite_pread which aims to detect
incoherency between CPU reads and writes to a bo whilst using it as a
source and target for GPU writes.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Chris Wilson [Mon, 16 Jul 2012 15:39:57 +0000 (16:39 +0100)]
Add bare-metal interface to adjust cacheing (i.e. snoop status) of a bo
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Daniel Vetter [Mon, 27 Aug 2012 07:37:01 +0000 (09:37 +0200)]
bump to 1.3 release
Chris Wilson [Sun, 26 Aug 2012 20:22:15 +0000 (21:22 +0100)]
mmio: Limit the uc- mapping to only map the registers
In the future, we may like to enable wc mapping of at least the GATT,
and so causing a conflict if we attempt to map the entire bar as uc-
here. Obviously we need a better fallback plan, but for the moment only
attempt to map the portion of the pci space that we use for register
access.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Chris Wilson [Sun, 26 Aug 2012 20:22:14 +0000 (21:22 +0100)]
intel_gtt: Harden against changes to kernel mappings of the GTT
Rather than use the common mmio segment which will be in future
restricted to just the registers and so exclude the GTT portion on all
architectures, explicitly mmap the GTT ourselves. Repeat this mmapping
with a couple of flags until we matching the existing kernel mapping.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Daniel Vetter [Fri, 24 Aug 2012 22:01:27 +0000 (00:01 +0200)]
s/NO_PID/NOP_ID
Alan typo'ed it, I've failed to notice :(
Alan Coopersmith [Fri, 24 Aug 2012 21:02:04 +0000 (14:02 -0700)]
Add --disable-nouveau option for platforms without nouveau support
Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>