haunted gc test failure in ci on macOS #446
Labels
No labels
Affects/CppNix
Affects/Nightly
Affects/Only nightly
Affects/Stable
Area/build-packaging
Area/cli
Area/evaluator
Area/fetching
Area/flakes
Area/language
Area/lix ci
Area/nix-eval-jobs
Area/profiles
Area/protocol
Area/releng
Area/remote-builds
Area/repl
Area/repl/debugger
Area/store
bug
Context
contributors
Context
drive-by
Context
maintainers
Context
RFD
crash 💥
Cross Compilation
devx
docs
Downstream Dependents
E/easy
E/hard
E/help wanted
E/reproducible
E/requires rearchitecture
imported
Language/Bash
Language/C++
Language/NixLang
Language/Python
Language/Rust
Needs Langver
OS/Linux
OS/macOS
performance
regression
release-blocker
stability
Status
blocked
Status
invalid
Status
postponed
Status
wontfix
testing
testing/flakey
Topic/Large Scale Installations
ux
No milestone
No project
No assignees
3 participants
Notifications
Due date
No due date set.
Dependencies
No dependencies set.
Reference: lix-project/lix#446
Loading…
Add table
Add a link
Reference in a new issue
No description provided.
Delete branch "%!s()"
Deleting a branch is permanent. Although the deleted branch may continue to exist for a short time before it actually gets removed, it CANNOT be undone in most cases. Continue?
idk what this is but i don't think it could have possibly been because of the change in question https://buildbot.lix.systems/#/buildrequests/128100
Copying the log into here for posterity:
Oh. This was the foolish wrapper we found when we looked at the sources.
[from matrix]
proc_pidfdlist https://opensource.apple.com/source/xnu/xnu-6153.41.3/bsd/kern/proc_info.c.auto.html
userspace side:
4f43d4276f/libsyscall/wrappers/libproc/libproc.c (L101)
the part that makes no sense here is that it should not have errno zero unless....
what if you closed all the fds.
So this is now actionable, the fix to this haunting is to bypass their wrapper because it is broken and cannot distinguish there being no actual FDs from returning failure. We copy the like five lines of their wrapper except not broken, write a disappointed comment, and close this bug.
cc @nrabulinski who i think was going to test if this is in fact the bug?
Err, right, sorry, I did try it out on one of my Macs and that is indeed what can happen so you were right - if we use the wrapper there’s no way to differentiate between empty fd list and an error! How fun!
Filed FB14695751 as follows:
look-ma-no-fds.c:
fail.c:
This issue was mentioned on Gerrit on the following CLs: