See racy trace_event: http://crbug.com/1091723 The trace event macros were racy because of this code: INTERNAL_TRACE_EVENT_GET_CATEGORY_INFO(...) The macro is using a static variable which is not thread-safe. The design of tracing was racy (by design) and rely on the fact that addTraceEvent(...) is threadsafe internally via TraceLog::lock_ This CL is replacing the static variable with a scope (C++11) pattern. The static variable is guaranted (C++) to be atomic and it is implemented efficiently. see: §6.7 [stmt.dcl] p4 " If control enters the declaration concurrently while the variable is being initialized, the concurrent execution shall wait for completion of the initialization. " Bug to track follow-up migration: http://anglebug.com/4702 Bug: chromium:1091259, chromium:1091723 Change-Id: If67b501e6e826ccf603eb2349c3f0aa6272c9a52 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2233410 Commit-Queue: Jamie Madill <jmadill@chromium.org> Reviewed-by:Jamie Madill <jmadill@chromium.org>
| Name |
Last commit
|
Last update |
|---|---|---|
| .. | ||
| android_system_settings | Loading commit data... | |
| common | Loading commit data... | |
| compiler | Loading commit data... | |
| feature_support_util | Loading commit data... | |
| gpu_info_util | Loading commit data... | |
| image_util | Loading commit data... | |
| libANGLE | Loading commit data... | |
| libEGL | Loading commit data... | |
| libGL | Loading commit data... | |
| libGLESv1_CM | Loading commit data... | |
| libGLESv2 | Loading commit data... | |
| tests | Loading commit data... | |
| third_party | Loading commit data... | |
| commit_id.py | Loading commit data... | |
| compiler.gni | Loading commit data... | |
| copy_compiler_dll.bat | Loading commit data... | |
| libGLESv2.gni | Loading commit data... |