unix_sigpipe: Add test for SIGPIPE disposition in child processes#121573
unix_sigpipe: Add test for SIGPIPE disposition in child processes#121573bors merged 2 commits intorust-lang:masterfrom
Conversation
|
rustbot has assigned @Mark-Simulacrum. Use r? to explicitly pick a reviewer |
9dfde0f to
527be25
Compare
|
Is there a reason this is a run-make test? It seems like it should be just a run-pass test like this one https://github.com/rust-lang/rust/blob/master/tests/ui/panics/abort-on-panic.rs#L1-L3 (or plenty of others with run-pass and revisions:) |
|
@Mark-Simulacrum I don't think run-pass tests supports building the separate program that this test needs? The example you linked to runs itself as a child process, but my test can't do that, since the Rust runtime would mess with |
|
Ah. I think it might be possible with auxiliary test files, but it's probably not worth the extra work then. @bors r+ |
…ark-Simulacrum unix_sigpipe: Add test for SIGPIPE disposition in child processes To make it clearer what the impact would be to stop using `SIG_IGN` and instead use a noop handler, like suggested [here](rust-lang#62569 (comment)) and implemented [here](rust-lang#121578). Part of rust-lang#97889
…iaskrgr Rollup of 5 pull requests Successful merges: - rust-lang#121573 (unix_sigpipe: Add test for SIGPIPE disposition in child processes) - rust-lang#121754 ([bootstrap] Move the `split-debuginfo` setting to the per-target section) - rust-lang#122205 (ensure that sysroot is properly set for cross-targets) - rust-lang#122257 (mir-opt tests: don't run a different set on --bless; enable --keep-stage-std) - rust-lang#122272 (Subtree update of `rust-analyzer`) Failed merges: - rust-lang#122108 (Add `target.*.runner` configuration for targets) r? `@ghost` `@rustbot` modify labels: rollup
…ark-Simulacrum unix_sigpipe: Add test for SIGPIPE disposition in child processes To make it clearer what the impact would be to stop using `SIG_IGN` and instead use a noop handler, like suggested [here](rust-lang#62569 (comment)) and implemented [here](rust-lang#121578). Part of rust-lang#97889
Rollup of 8 pull requests Successful merges: - rust-lang#121148 (Add slice::try_range) - rust-lang#121573 (unix_sigpipe: Add test for SIGPIPE disposition in child processes) - rust-lang#121633 (Win10: Use `GetSystemTimePreciseAsFileTime` directly) - rust-lang#121840 (Expose the Freeze trait again (unstably) and forbid implementing it manually) - rust-lang#121907 (skip sanity check for non-host targets in `check` builds) - rust-lang#122002 (std::threads: revisit stack address calculation on netbsd.) - rust-lang#122108 (Add `target.*.runner` configuration for targets) - rust-lang#122298 (RawVec::into_box: avoid unnecessary intermediate reference) r? `@ghost` `@rustbot` modify labels: rollup
|
Failed in #122327: #122327 (comment) @bors r- |
527be25 to
65eb51c
Compare
compiletest: Add support for `//@ aux-bin: foo.rs` Which enables ui tests to use auxiliary binaries. See the added self-test for an example. This is an enabler for the test in rust-lang#121573.
Rollup merge of rust-lang#122634 - Enselic:aux-bin, r=oli-obk compiletest: Add support for `//@ aux-bin: foo.rs` Which enables ui tests to use auxiliary binaries. See the added self-test for an example. This is an enabler for the test in rust-lang#121573.
|
☔ The latest upstream changes (presumably #122735) made this pull request unmergeable. Please resolve the merge conflicts. |
0dc26de to
3095794
Compare
|
@Mark-Simulacrum After rebasing on #122634 the only commit left in this PR is the test itself (which is now a regular run-pass ui test) which you more or less already approved, so would be great if you could take another look. |
Mark-Simulacrum
left a comment
There was a problem hiding this comment.
r=me if we really need the extra ignore
| @@ -0,0 +1,56 @@ | |||
| //@ revisions: default sig_dfl sig_ign inherit | |||
| //@ ignore-cross-compile because we run the compiled code | |||
There was a problem hiding this comment.
This isn't present in most run-pass tests, so what makes this one special?
There was a problem hiding this comment.
Turns out it is not run-pass that is the problem here, because the following run-pass and aux-build test passes in cross-compilation context:
src/ci/docker/run.sh --dev armhf-gnu
../x test --target arm-unknown-linux-gnueabihf tests/ui/cross-crate/reexported-static-methods-cross-crate.rsInstead, it is aux-bin that does not support cross compilation yet. For example, remote-test-client does not yet know how to upload auxiliary binaries to the target device. That it something that would be nice to solve, but it does not block this PR. I updated the motivation for the ignore.
I would be happy to @bors r=Mark-Simulacrum the latest commit but you'll need to @bors delegate=Enselic first.
For robustness, also test the disposition in our own process even if other tests in `tests/ui/attributes/unix_sigpipe` already covers it.
3095794 to
71eb763
Compare
|
@bors r+ rollup |
…ark-Simulacrum unix_sigpipe: Add test for SIGPIPE disposition in child processes To make it clearer what the impact would be to stop using `SIG_IGN` and instead use a noop handler, like suggested [here](rust-lang#62569 (comment)) and implemented [here](rust-lang#121578). Part of rust-lang#97889
…iaskrgr Rollup of 5 pull requests Successful merges: - rust-lang#121573 (unix_sigpipe: Add test for SIGPIPE disposition in child processes) - rust-lang#123170 (Replace regions in const canonical vars' types with `'static` in next-solver canonicalizer) - rust-lang#123200 (KCFI: Require -C panic=abort) - rust-lang#123201 (Improve wording in std::any explanation) - rust-lang#123224 (compiletest: print reason for failing to read tests) r? `@ghost` `@rustbot` modify labels: rollup
Rollup merge of rust-lang#121573 - Enselic:sigpipe-child-process, r=Mark-Simulacrum unix_sigpipe: Add test for SIGPIPE disposition in child processes To make it clearer what the impact would be to stop using `SIG_IGN` and instead use a noop handler, like suggested [here](rust-lang#62569 (comment)) and implemented [here](rust-lang#121578). Part of rust-lang#97889
To make it clearer what the impact would be to stop using
SIG_IGNand instead use a noop handler, like suggested here and implemented here.Part of #97889