[roll] Roll fuchsia [trace_manager] Don't double copy buffer in streaming mode

The tracing system doesn't differentiate between setting
"rolling_data_end" to "0" to mean no data has been written, and "0", we
haven't yet committed the buffer. This is problematic in streaming mode,
as it causes us to attempt to uncondionally read the inactive buffer
after the trace ends. This happens because after we transfer the buffer,
we "reset" the rolling_data_end to "0".

In trace manager, when we copy the buffer out in streaming mode, we now
check last_wrapped_count_, which is our local copy of the last buffer we
transferred and compare it to the tracee's wrapped count before
attempting to copy the buffer.

Original-Bug: 312790394
Multiply: fuchsia-pkg://fuchsia.com/trace_manager_integration_tests#meta/streaming_test.cm
Original-Reviewed-on: https://fuchsia-review.googlesource.com/c/fuchsia/+/1266450
Original-Revision: becc20eb36e12e09539481a6565b96eff024e9e9
GitOrigin-RevId: 27dc7c857dadc44a13575c4204088dc27776accc
Change-Id: Ieaf5ed807b0c7b7b4cb9aabb9ccadf5b3ed79e26
1 file changed
tree: ba0d242c81c43b193f36bdfbef655cd19b24e233
  1. ctf/
  2. git-hooks/
  3. infra/
  4. third_party/
  5. flower
  6. jiri.lock
  7. MILESTONE
  8. minimal
  9. prebuilts
  10. README.md
  11. stem
  12. test_durations
README.md

Integration

This repository contains Fuchsia's Global Integration manifest files.

Making changes

All changes should be made to the internal version of this repository. Our infrastructure automatically updates this version when the internal one changes.

Currently all changes must be made by a Google employee. Non-Google employees wishing to make a change can ask for assistance in one of the communication channels documented at get involved.

Obtaining the source

First install Jiri.

Next run:

$ jiri init
$ jiri import minimal https://fuchsia.googlesource.com/integration
$ jiri update

Third party

Third party projects should have their own subdirectory in ./third_party.