Speculative fix for mem watchdog crash

We keep hitting the memory watchdog ~60s after
the end of a trace session (note that 60s is
precisely a watchdog window). I am suspecting this
is heap churn created by ReadBuffers which accumulates
in scudo fragmentation.
Force a M_PURGE call on every ReadBuffers().
If this makes the problem go away we should throttle this
a bit more. If not we should remove it.

Bug: 195145848
Change-Id: I7554fbe812ca16e08451234c9bccff752c3856cf
2 files changed
tree: c092016a125fa29cda09c244c1e6fd19449150d5
  1. .github/
  2. bazel/
  3. build_overrides/
  4. buildtools/
  5. debian/
  6. docs/
  7. examples/
  8. gn/
  9. include/
  10. infra/
  11. protos/
  12. src/
  13. test/
  14. tools/
  15. ui/
  16. .clang-format
  17. .clang-tidy
  18. .gitattributes
  19. .gitignore
  20. .gn
  21. .style.yapf
  22. Android.bp
  23. Android.bp.extras
  24. BUILD
  25. BUILD.extras
  26. BUILD.gn
  27. CHANGELOG
  28. codereview.settings
  29. DIR_METADATA
  30. heapprofd.rc
  31. LICENSE
  32. meson.build
  33. METADATA
  34. MODULE_LICENSE_APACHE2
  35. OWNERS
  36. perfetto.rc
  37. PerfettoIntegrationTests.xml
  38. PRESUBMIT.py
  39. README.chromium
  40. README.md
  41. TEST_MAPPING
  42. traced_perf.rc
  43. WORKSPACE
README.md

Perfetto - System profiling, app tracing and trace analysis

Perfetto is a production-grade open-source stack for performance instrumentation and trace analysis. It offers services and libraries and for recording system-level and app-level traces, native + java heap profiling, a library for analyzing traces using SQL and a web-based UI to visualize and explore multi-GB traces.

See https://perfetto.dev/docs or the /docs/ directory for documentation.