Age | Commit message (Collapse) | Author |
|
|
|
|
|
Bug: none
Test: run simpleperf_unit_test.
Change-Id: I7d1332f36780303bb3f14adedb5b049aefbb523c
|
|
In system wide profiling, sometimes a sample record goes before all
mmap records of a process. Then thread.maps->version = 0 and
cached_map.map isn't created.
This is caught by simpleperf_unit_test.
Bug: none
Test: run simpleperf_unit_test.
Change-Id: I99254396d0fb22531f735ff3193b8d3b1f9641dd
|
|
|
|
When calculating cpu frequency and other frequencies, use time
running on cpu instead of total passed time. This is to be
consistent with linux-tools-perf.
This is reported in https://github.com/android-ndk/ndk/issues/644.
Bug: none
Test: run simpleperf_unit_test.
Change-Id: Ic385939c189fb86c6b3c648d2c0a4e9636c13c93
|
|
|
|
Bug: http://b/73875722
Test: run simpleperf manually.
Change-Id: I010c8612b18cd4034199440cb178d9d9cc8f0ff4
|
|
|
|
|
|
Bug: none
Test: run simpleperf_unit_test.
Test: run simpleperf manually.
Change-Id: I5e7229c8d5bfc4d47a6b7e403e627137c7f36b55
|
|
Increases speed (unwinds per second) by 50%.
Bug: None
Test: manual - took a profile
Test: run simpleperf_unit_test.
Change-Id: I254f70a011c1b3d31bbd4e9e4bb9e85fbf2ffbdf
|
|
|
|
Fix:
system/extras/brillo_config/Android.mk:83: error: $(file ...)
is not supported in rules.
Test: m -j32
Bug: b/73753971
Change-Id: I96099be887f404042c28d3fa1c9f4a2e12ce1222
|
|
|
|
|
|
And use BUILD_NUMBER_FROM_FILE instead of BUILD_NUMBER to reduce build
reruns.
Test: m -j32
Bug: b/70351683
Change-Id: Ic2a6e2095712828c8314fdd5fca7669f2fb88bc3
|
|
Bug: 73558129
Test: Passes unit tests.
Change-Id: I7cbe618f622b540abf486777f0d40a18f890dbef
|
|
|
|
|
|
|
|
Without flushing stdout, the tests may fail when
running `adb shell simpleperf_unit_test`.
Bug: none
Test: run `adb shell simpleperf_unit_test`.
Change-Id: I8bcf739e7672e8affc71df0a3f11264c0bd7f745
|
|
Soong_UI will handle build timestamp and export date file to Make.
Test: m -j32
Bug: b/70351683
Change-Id: I09ac0117c899afadf28e562d83e95f3132a47e20
|
|
Bug: none
Test: run simpleperf_unit_test.
Change-Id: Ida90ecd465e5c63f0f813cde63d70acfeb1281da
|
|
|
|
|
|
* bandwidth.h contains thumb mode only instructions.
Test: normal build.
Change-Id: Ib40dfda63f668949370405e7ec53f5f9828200e6
|
|
|
|
Enables cross-arch unwinding since the new unwinder supports cross-arch
unwinding.
Also change tests accordingly.
Bug: http://b/73127000
Test: run simpleperf_unit_test.
Change-Id: I3ca3f7a81b40e9f7ea2b39b9fdc951f9e7a6acc7
|
|
This commit disables "perfporfd" in PDK builds because "perfporfd"
depends on "libservices" (from "frameworks/base") which is not in PDK
builds as well.
Test: Build a target (described in http://b/72961456) with
`platform.zip` built from master FSK source tree.
Bug: 72961456
Change-Id: Ia4913435642cd09a321afb9f583331ca24a8e4be
|
|
|
|
In debug-unwind cmd:
1. Add --symfs option.
2. Add meta_info section when there isn't one.
3. Add file section when there isn't one.
Rename unwinding_result_reporter.py to debug_unwind_reporter.py,
and fix a small error in it.
Bug: http://b/72556486
Test: run simpleperf_unit_test.
Test: run debug_unwind_reporter.py manually.
Change-Id: Ibc66b7bfdc91a17f6f1c2c0e84f4769d2bcaa1e7
|
|
|
|
Support optional "--" to be compatible with linux-perf.
Test: mmma system/extras/simpleperf
Test: simpleperf_unit_test
Change-Id: I4c2627625bb97514d86c8972ef5a0a31bf518b76
|
|
|
|
Currently, we use --log debug option in record cmd to debug offline unwinding.
However, it has below disadvantages:
1. It adds extra complexity in record cmd.
2. It doesn't keep reg/stack data of samples.
3. It isn't convenient to reproduce samples in problem. Because each time
recording gets different samples.
4. It isn't very suitable for performance test of unwinding, for the same
reason as item 3.
So instead, this CL adds debug-unwind cmd focusing on debugging and testing
offline unwinding. It solves problems mentioned above.
Also change unwinding_result_reporter.py to make it work with perf.data
generated by debug-unwind cmd.
Bug: http://b/72556486
Test: run simpleperf_unit_test.
Test: run unwinding_result_reporter.py manually.
Change-Id: I11cdf1eba993f48d61ef9891ad1be54d29679fdb
|
|
|
|
When recording google.sample.tunnel app for 30s:
It took 3s to unwind samples and write unwound samples to file.
It took 0.3s to write samples containing stack/reg data to file.
The result shows recording with post unwinding consumes much
less time than unwinding samples immediately. This means we can
record with higher freq and get smaller lose rate when using
post unwinding. So make below changes:
1. Make post unwinding by default.
2. Replace --post-unwind with --no-post-unwind option.
3. Make --trace-offcpu and callchain joiner work with post unwinding.
4. Remove special operations in --log debug mode. Those will be
supported in a new command.
Bug: http://b/72556486
Test: run simpleperf_unit_test.
Test: run python test.py.
Change-Id: I9a5a5defda9d040985e674c43db19ee68e7aa305
|
|
|
|
|
|
Bug: none
Test: none
Change-Id: Ia4e6dd0e44e19379d79a08a8555ad413ba879bf2
|
|
Also split options into groups.
Bug: http://b/72495881
Test: manually.
Change-Id: Iec322fc1f1afb74ee5ee0fb19145d16926064332
|
|
|
|
Changes:
- Removes references to libunwind.
- Add enabling of global Elf cache.
- Add new ApkInspector function to get the offset from an apk by name.
- When a map name contains a '!' convert back to a name and offset.
- Initialize global elf cache in OfflineUnwinder constructor.
Bug: 65682279
Test: All unit tests pass.
Change-Id: I303b4beab8270ed2097cd17d91253c321cdc97f1
|
|
* changes:
Perfprofd: Clean up test logging helper
Perfprofd: Backport elf symbolizer config to ConfigReader
|
|
Clean up, in preparation for a potential move to libbase.
Test: mmma system/extras/perfprofd
Test: perfprofd_test
Change-Id: I1bdeffc6b69bd8c2513e1b2fce103bbae1c53dad
|
|
Add an entry to ConfigReader to initialize usage of the ELF symbolizer.
Use this to turn off on-device symbolization for live testing, as
warnings about missing DSOs will pollute the logs and make the tests
fail.
Test: mmma system/extras/perfprofd
Test: perfprofd_test
Change-Id: I3e13db9779e5d2edfb14e2fbe2b3d186ef7af0bd
|
|
|
|
|
|
Bug: http://b/72323886
Bug: http://b/72330172
Test: manually.
Change-Id: I91ece3081e0a7dcbe068154b36b35a694519af8e
|