Skip to content

Various fixes for posix_spawn refactoring #213

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
merged 8 commits into from
Jul 20, 2021

Conversation

bgamari
Copy link
Contributor

@bgamari bgamari commented Jul 19, 2021

Sadly, the refactor in #208 had a few issues which my earlier testing hadn't uncovered:

  • the posix_spawn path didn't correctly handle the case where the platform does not support POSIX_SPAWN_SETPGROUP
  • some compilers warned of missing (but in principle unreachable) cases, breaking the -Werror build in some cases
  • fork_exec's setup_std_handle_fork failed to return a value in some cases
  • the configure script failed to declare its dependency on <spawn.h>, causing it to fail to find posix_spawn in some cases
  • GCC 6.3 threw spurious warnings in find_executable.c, breaking the -Werror build
  • the output expected by the GHC testsuite had a spurious platform dependence due to the new posix_spawn implementation

I have fixed all of these and tested on Linux, Darwin, and FreeBSD.

Fixes #210.

@bgamari bgamari changed the title Various fixes Various fixes due to posix_spawn refactoring Jul 19, 2021
@bgamari bgamari changed the title Various fixes due to posix_spawn refactoring Various fixes for posix_spawn refactoring Jul 19, 2021
@snoyberg snoyberg merged commit 83b9ef8 into haskell:master Jul 20, 2021
@snoyberg
Copy link
Collaborator

Awesome, thanks! Want me to make a release?

@bgamari
Copy link
Contributor Author

bgamari commented Jul 20, 2021

Thanks @snoyberg! Let's hold off on the release just a tad longer; I've seen something in further testing that has me a bit concerned.

@bgamari
Copy link
Contributor Author

bgamari commented Jul 21, 2021

A brief update: I am increasingly certain that the issue that I mentioned earlier is actually a (rather nasty) GHC bug (namely https://gitlab.haskell.org/ghc/ghc/-/issues/20137) but it would likely still be prudent to wait until we have a fix and verification before doing a release. I'm hopeful that we'll have a fix tomorrow..

@bgamari
Copy link
Contributor Author

bgamari commented Jul 25, 2021

Alright, at this point I have eliminated this patch as the source for any of the issues we have seen in process in CI. Feel free to cut a release when you have time, @snoyberg. Cheers!

@snoyberg
Copy link
Collaborator

Upload complete!

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

GHC build/testsuite failures with process 1.6.13.0
2 participants