|
7181239d
|
2025-04-04T14:13:58
|
|
Add long ANGLE traces feature
Enables very long Android captures by swapping binary
data chunked buffers to/from disk.
Bug: b/425728227
Change-Id: I539f72590eece03cfc69d42fc34be9825a9ff1fe
Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/6476924
Reviewed-by: Cody Northrop <cnorthrop@google.com>
Commit-Queue: Mark Łobodziński <mark@lunarg.com>
|
|
c0d806b4
|
2025-02-10T23:01:33
|
|
CL: OpenCL support for ANGLE Capture/Replay
Implementation of OpenCL Capture/Replay tool in ANGLE.
Brief notes about the change:
- Most meaningful changes for the capture process are
made in src/libANGLE/capture/
- Most meaningful changes for replay are made in
util/capture/ and src/tests/perf_tests/
- Many autogenerated files are changed/added to allow
the capture of OpenCL objects & calls
- The following applications were captured/replayed:
benchmark_model, GeekBench Compute, GeekBench ML,
AI-Benchmark, various OCL CTS tests
- End2end test added to capture_tests.
CapturedTestCL.MultiFrameCL/ES3_Vulkan
Bug: angleproject:383841335
Change-Id: I55fdaa6cd6c7ba740aaa2351e4d29050059d6d1d
Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/6102105
Commit-Queue: Cody Northrop <cnorthrop@google.com>
Reviewed-by: Roman Lavrov <romanl@google.com>
Reviewed-by: Cody Northrop <cnorthrop@google.com>
|
|
e19377af
|
2024-08-21T14:52:09
|
|
Tests: Support multiple users on Android
On most Android platforms, the userId is '0' and many of our defaults
just work. But on some, the default is non-zero, and that complicates
things.
To support this, we have to specify which user is correct, which can
be determined with:
adb shell am get-current-user
Update everywhere we were using sdcard directly and use:
Before: /sdcard
After: /storage/emulated/<userId>
Incorporate the userId to find the application's home directory:
Before: /data/data/com.android.angle.test
After: /data/user/<userId>/com.android.angle.test
This CL:
- Starts using a user flag in our Android scripts
- Updates where tests look for test data and binaries
- Updates where scripts use temp and external storage
- Updates script to not use sdcard at all when root
- Disables run-as when user is not default
Bug: b/361388557
Change-Id: I0fb556ecb2d0f6bc6569a858c290b127217d4638
Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/5825122
Commit-Queue: Cody Northrop <cnorthrop@google.com>
Reviewed-by: Roman Lavrov <romanl@google.com>
|
|
c9c7c4d7
|
2024-03-11T12:17:11
|
|
Trace tests: save debug files when gz decompression fails
Input and output bytes saved to CAS outputs,
for futher manual inspection
Bug: b/296921272
Change-Id: I57c88090bdb41bf9ef48466f38c7da90572ce467
Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/5361574
Commit-Queue: Roman Lavrov <romanl@google.com>
Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org>
Auto-Submit: Roman Lavrov <romanl@google.com>
|
|
|
d85de0e9
|
2023-08-09T14:15:56
|
|
Capture/Replay: Add optional replay of trimmed resources
This CL:
- Adds '--include-inactive-resources' option to
angle_trace_tests
- Removes the 'trim-enabled' option
- Outputs all previously trimmed shaders/programs to trace file
in a new Setup function, SetupReplayContextSharedInactive()
which is executed only if the new option is specified
- Modifies CaptureTest to add inactive resources, but does not
set the include-inactive-resources flag
Bug: b/296055694
Change-Id: I33b18d5da727d55c90c2012c2bf64b1413521429
Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/4781552
Reviewed-by: Cody Northrop <cnorthrop@google.com>
Commit-Queue: Mark Łobodziński <mark@lunarg.com>
Reviewed-by: Roman Lavrov <romanl@google.com>
|
|
|
52152933
|
2023-06-06T15:36:36
|
|
Add trace_interface: functions and callbacks for traces
Defines the interface between the test suite
(or an other TraceLibrary class user) and trace libraries.
TraceFunctions defines entry points for calls suite->trace, such as
SetupReplay() or SetBinaryDataDir().
TraceCallbacks defines entry points for calls trace->suite, for example
for loading .angledata.gz files.
These are set up via the exported SetupEntryPoints() call. Functions
like SetupReplay etc no longer need to be exported from the trace
library.
TraceInfo (parsed representation of the trace json) is moved to
trace_interface as is. This is convenient for further changes to the
fixture that will allow to easily move some of the captured parameters
to json.
This also moves Decompress functionality (and memory ownership) to test
suite entirely, which avoids Decompress/Delete callbacks - the trace
just calls LoadBinaryData via TraceCallbacks and TraceLibrary releases
the memory either on FinishReplay or in its destructor.
This should also take care of the memory leak described in
https://crrev.com/c/3858185
Bug: b/286072760
Change-Id: Ibc6f6f64156ad805b1917c8fc41a3b0d2c0d6375
Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/4594445
Reviewed-by: Amirali Abdolrashidi <abdolrashidi@google.com>
Commit-Queue: Roman Lavrov <romanl@google.com>
Reviewed-by: Cody Northrop <cnorthrop@google.com>
|
|
|
4e073356
|
2023-04-13T15:16:53
|
|
Trace Interpreter: gzip cpp and run gold tests.
New test step:
angle_restricted_trace_gold_interpreted_tests
similar to existing
angle_restricted_trace_gold_tests
but runs a subset of newest traces that have the right format.
Currently there are 8 of them and the filter is hard-coded in
test_suites.pyl. rise_of_empires can also be run interpreted but
happened to flake during testing so it's not included for now.
Trace cpp files are concatenated and gzipped
as a build step producing gen/tracegz_{trace}.gz
and these gz files are then distributed as part of the build
--trace-interpreter flag changes from boolean to string to support:
--trace-interpreter=gz:
* this uses the gz file above
* gold tests added in this CL use this option
--trace-interpreter=c:
* using uncompressed c/cpp file
* existing angle_trace_interpreter_tests runs a retrace
saving .c files and uses this option
Bug: b/276742336
Change-Id: I69544f25bda873af191978195d02ffbdd34363c0
Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/4424690
Reviewed-by: Cody Northrop <cnorthrop@google.com>
Commit-Queue: Roman Lavrov <romanl@google.com>
|
|
|
2c7f31af
|
2023-04-13T11:47:34
|
|
Trace Interpreter: Move to a shared library.
Based on https://crrev.com/c/4178024 but does only one thing:
makes trace interpreter a shared library (as if it was a compiled trace)
I am not sure what impact this actually has on the TracePerfTest
callbacks (mentioned in the CL above) as I see onEglMakeCurrent callback
getting called without this CL as well. Anyways, this makes things a bit
more consistent.
* Gets rid of TraceReplayInterface
* TraceInterpreter is now an implementation detail
* Need to additionally pass list of trace files to TraceLibrary
(new SetTraceInfo replay export)
* GetResourceIDMapValue is just moved as is to the right lib
Bug: b/276742336
Change-Id: I67ea9fbcb4f7db999ab71c8443ea91c5631df942
Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/4423869
Commit-Queue: Roman Lavrov <romanl@google.com>
Reviewed-by: Cody Northrop <cnorthrop@google.com>
|
|
|
b0739436
|
2023-02-13T14:28:17
|
|
Traces: Add key frame support
Allow specifying a key frame in the trace JSON, i.e:
"KeyFrames": [ 20 ],
This allows our infra to use a frame other than 1 when
taking screenshots for quality comparison.
Adds new flag `--run-to-key-frame`, which will stop the
trace once key frame has been reached. If no key frame in
JSON, frame 1 will be used.
Note the name in JSON is plural, but we only support one
key frame for now. Multiple key frame support can come
in the future.
This CL also updates the code to allow ending traces
early with `--max-steps-performed` which has been broken
since http://crrev/c/4008998
It also removes `--one-frame-only` which is superseded by
`--run-to-key-frame`, and can be replicated using
`--max-steps-performed 1`.
Test: angle_trace_tests --gtest_filter="*tmnt_shredders_revenge*"
Bug: angleproject:8035
Bug: b/270426257
Change-Id: Ib02ef60d887ae5efb0288f5a9b8c2914dafc6efc
Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/4284637
Commit-Queue: Cody Northrop <cnorthrop@google.com>
Reviewed-by: Yuly Novikov <ynovikov@chromium.org>
|
|
5ca36d3c
|
2023-01-05T15:16:35
|
|
Trace Interpreter: Properly support resource ID maps.
Previously we were using the key value as the resource value.
In some cases these values would become mismatched when running
a replay, which would cause interpreter errors.
Bug: angleproject:7887
Change-Id: I7f8e687986e8282a6df066bdae5ace07e4510b47
Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/4140258
Reviewed-by: Cody Northrop <cnorthrop@google.com>
Commit-Queue: Jamie Madill <jmadill@chromium.org>
|
|
|
e5e78ab4
|
2022-12-29T13:16:13
|
|
Trace Interpreter: Auto-gen ReplayCustomFunctionCall.
This makes it easier to add or remove fixture functions.
Bug: angleproject:7775
Change-Id: I1c56465cfa0580414715f562cadc916c4baf3d11
Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/4128718
Commit-Queue: Jamie Madill <jmadill@chromium.org>
Reviewed-by: Cody Northrop <cnorthrop@google.com>
|
|
f3e047b5
|
2022-10-12T16:28:32
|
|
Add new retrace command get_min_reqs
`retrace_restricted_traces.py get_min_reqs` finds the minimum
required GLES version and extensions and record them to a trace's
json file. TracePerfTest has also been modified to use the new
"RequiredExtensions" json field to determine if it should run a
test or not. This allows several lines to be removed from
TracePerfTest's constructor.
get_min_reqs uses a divide-and-conquer strategy to find the extensions.
This approach should be faster for traces where the number of required
extensions is less than 20.
This adds a new command to retrace_restricted_traces.py called
get_min_reqs. This runs an existing trace multiple times to find what
the minimum required GLES version is, then records that version to the
trace's json file.
Bug: angleproject:7639
Change-Id: I42530e438054efec1ad79ababebe06d221d5c695
Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/3884175
Reviewed-by: Geoff Lang <geofflang@chromium.org>
Reviewed-by: Cody Northrop <cnorthrop@google.com>
Commit-Queue: Sean Risser <srisser@google.com>
|
|
|
dc62b3ee
|
2022-10-10T21:00:16
|
|
Capture/Replay: Add trace interpreter.
Also adds a self-test using the retrace script.
Bug: angleproject:7752
Change-Id: I1985b47250bef99726d2ca2d90bef859208e357e
Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/3965128
Reviewed-by: Cody Northrop <cnorthrop@google.com>
Commit-Queue: Jamie Madill <jmadill@chromium.org>
Reviewed-by: Yuxin Hu <yuxinhu@google.com>
|
|
|
6193274a
|
2022-10-10T21:00:12
|
|
Capture/Replay: Redesign in-memory call capture replay.
This will allow the replay to use the call captures returned
by the interpreter's parser.
Bug: angleproject:7752
Change-Id: If1b281d9ce7ccfbdc23bea615e1e2258c8a029f9
Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/3963367
Reviewed-by: Cody Northrop <cnorthrop@google.com>
Commit-Queue: Jamie Madill <jmadill@chromium.org>
Reviewed-by: Yuxin Hu <yuxinhu@google.com>
|
|
|
4b33633a
|
2022-08-26T16:16:50
|
|
Capture/Replay: Delete binary data in FinishReplay via callback
Delete callback fixes b/179188489 as delete[] will get called from the
same module where allocation happens. When decompress/delete callbacks
not provided, new[] and delete[] are called from the fixture.
AFAICT, gBinaryData == nullptr every time, even with a single test and
--gtest_repeat=N, so we never delete the uncompressed data. When
running multiple tests locally, I see RES usage grow significantly
on Linux which this CL fixes. I don't think this matters for bot runs as
we run tests one by one there.
Bug: b/179188489
Change-Id: Iee3bccb9d6d72c315c4358b70ec9c3c0ac963258
Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/3858185
Reviewed-by: Jamie Madill <jmadill@chromium.org>
Reviewed-by: Cody Northrop <cnorthrop@google.com>
Commit-Queue: Roman Lavrov <romanl@google.com>
|
|
|
153d8db9
|
2022-06-06T13:49:40
|
|
Turn missing {trace_name}.json trace files into test failures.
Before this change, a missing or broken trace json file results in all
trace tests "disappear" from the list with only ERR in logs. After this
change, a missing or broken trace json will result in that specific test
FAIL status and error in logs.
Note that this also allows for inconsistency between
restricted_traces.json and trace json files on the device because we're
no longer requiring for trace json to be loaded unless the test is
actually being run.
Bug: angleproject:7410
Change-Id: I07532dc0bf85a0d2fcaf425ce8d4a589473a898d
Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/3690734
Commit-Queue: Roman Lavrov <romanl@google.com>
Reviewed-by: Cody Northrop <cnorthrop@google.com>
Reviewed-by: Jamie Madill <jmadill@chromium.org>
|
|
2ce60b52
|
2022-05-05T08:23:07
|
|
Fixes for C++20 support.
* The compiler grew smart enough to warn about value changes due to
casting to float inside floatToNormalized(). Make it smart enough to
realize this isn't a problem by using constexpr if.
* Types on both sides of a comparison operator should be the same.
* Structs with user-declared constructors are no longer aggregates.
Provide a constructor and call it.
* std::result_of<F(x)> is gone, use std::invoke_result<F, x> instead.
Bug: chromium:1284275
Change-Id: I6487bb18c65837a6d7d2661f65e097dc6a7605b6
Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/3630478
Reviewed-by: Jamie Madill <jmadill@chromium.org>
Commit-Queue: Jamie Madill <jmadill@chromium.org>
Auto-Submit: Peter Kasting <pkasting@chromium.org>
|
|
|
2f3e4db8
|
2021-09-02T10:31:13
|
|
Capture/Replay: Serialize trace metadata to a JSON file.
This will allow us to remove more code auto-generation for the trace
tests. The trace info now can be loaded directly from JSON instead of
from the autogenerated information.
Bug: angleproject:5133
Change-Id: I04e22b9279b19282df274bc8defcd363d0449111
Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/3140218
Commit-Queue: Jamie Madill <jmadill@chromium.org>
Reviewed-by: Tim Van Patten <timvp@google.com>
Reviewed-by: Cody Northrop <cnorthrop@google.com>
|