[PATCH] D54889: Fiber support for thread sanitizer

[Dmitry Vyukov via Phabricator via llvm-commits]

dvyukov added a comment.

> Can't agree with this. Only fiber is visible to user. Of cause, it is the same as thread until user switches it.

Whatever we call them, Processor is for different things. It must not hold anything related to user state.

> In standalone thread we create temporary context and switch into into it. Then original thread context is posted into event loop thread, where it runs for some time together with other fibers.

Oh, I see. I guess it can have some rough edges if used incorrectly, but also probably can work if used carefully.
OK, let's leave __tsan_get_current_fiber alone.

> Currently there is no interceptor for pthread_exit(). Do you suggest to add it?

Yes.

> Isn't it enough to place _tsan_ignore_thread_begin()/_tsan_ignore_thread_end() around setjmp/longjmp calls?

These only ignore memory accesses, they don't affect interceptor operation as far as I see.


Repository:
  rCRT Compiler Runtime

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D54889/new/

https://reviews.llvm.org/D54889