Support `#[unix_sigpipe = "inherit|sig_dfl"]` on `fn main()` to prevent ignoring `SIGPIPE` #97802

Enselic · 2022-06-06T18:22:43Z

When enabled, programs don't have to explicitly handle ErrorKind::BrokenPipe any longer. Currently, the program

fn main() { loop { println!("hello world"); } }

will print an error if used with a short-lived pipe, e.g.

% ./main | head -n 1
hello world
thread 'main' panicked at 'failed printing to stdout: Broken pipe (os error 32)', library/std/src/io/stdio.rs:1016:9
note: run with `RUST_BACKTRACE=1` environment variable to display a backtrace

by enabling #[unix_sigpipe = "sig_dfl"] like this

#![feature(unix_sigpipe)]
#[unix_sigpipe = "sig_dfl"]
fn main() { loop { println!("hello world"); } }

there is no error, because SIGPIPE will not be ignored and thus the program will be killed appropriately:

% ./main | head -n 1
hello world

The current libstd behaviour of ignoring SIGPIPE before fn main() can be explicitly requested by using #[unix_sigpipe = "sig_ign"].

With #[unix_sigpipe = "inherit"], no change at all is made to SIGPIPE, which typically means the behaviour will be the same as #[unix_sigpipe = "sig_dfl"].

See #62569 and referenced issues for discussions regarding the SIGPIPE problem itself

See the this Zulip topic for more discussions, including about this PR.

Tracking issue: #97889

rust-highfive · 2022-06-06T18:22:45Z

Hey! It looks like you've submitted a new PR for the library teams!

If this PR contains changes to any rust-lang/rust public library APIs then please comment with r? rust-lang/libs-api @rustbot label +T-libs-api -T-libs to request review from a libs-api team reviewer. If you're unsure where your change falls no worries, just leave it as is and the reviewer will take a look and make a decision to forward on if necessary.

Examples of T-libs-api changes:

Stabilizing library features
Introducing insta-stable changes such as new implementations of existing stable traits on existing stable types
Introducing new or changing existing unstable library APIs (excluding permanently unstable features / features without a tracking issue)
Changing public documentation in ways that create new stability guarantees
Changing observable runtime behavior of library APIs

rust-highfive · 2022-06-06T18:22:46Z

r? @joshtriplett

(rust-highfive has picked a reviewer for you, use r? to override)

compiler/rustc_codegen_ssa/src/base.rs

library/std/src/sys/unix/mod.rs

Enselic · 2022-06-10T22:28:57Z

I consider this PR feature complete now. I'm sure there is more bike-shedding to do on what to call the attribute, and I'm sure you'll find things to comment on, but I'm happy with the implementation and the scope of the tests I added. I need to add a section to the unstable book, but I would prefer to do that after the bike-shedding is done.

Looking forward to a new code review round. If you want to me to split out some parts into separate PRs that would be fine, just let me know.

I had to do a rebase and took the opportunity to squash all my commits together, but I have already added some more commits. I plan on squashing again next time I need to rebase.

I'm also adding T-lang as requested.

@rustbot label -S-waiting-on-author +S-waiting-on-review +T-lang

compiler/rustc_codegen_cranelift/src/main_shim.rs

joshtriplett · 2022-08-31T13:46:32Z

My fault, sorry; I missed that the standard library was referring to the compiler, rather than the other way around.

I think we should just duplicate this and put a comment in each pointing to the other; it doesn't seem worth attempting to deduplicate.

RalfJung · 2022-08-31T14:08:48Z

An include! the other way around would be much less problematic, but would still make the compiler sources as distributed in the rustc-dev component incorrect. Not sure if this matters for anything except for getting RA support in projects like Clippy or Miri.

So, exposing them perma-unstable would IMO be preferable over include!.

But yeah duplicating these 4 constants might not be so bad.

Enselic · 2022-08-31T16:27:15Z

Thank you for further input, looking forward to another review round! I have addressed the comments now.

@rustbot ready

joshtriplett · 2022-08-31T19:23:39Z

New version looks good to me. @RalfJung?

Enselic · 2022-09-02T18:22:03Z

From what I can tell, the newly raised concerns have been addressed in a satisfactory way (please correct me if I'm wrong @RalfJung!), so let's try again to get this merged.

@bors r=joshtriplett

bors · 2022-09-02T18:22:06Z

📌 Commit 3810d4a has been approved by joshtriplett

It is now in the queue for this repository.

bors · 2022-09-02T21:08:12Z

⌛ Testing commit 3810d4a with merge 8c6ce6b...

bors · 2022-09-02T23:34:52Z

☀️ Test successful - checks-actions
Approved by: joshtriplett
Pushing 8c6ce6b to master...

rust-highfive · 2022-09-02T23:35:15Z

📣 Toolstate changed by #97802!

Tested on commit 8c6ce6b.
Direct link to PR: #97802

💔 miri on windows: test-fail → build-fail (cc @oli-obk @RalfJung).
💔 miri on linux: test-pass → build-fail (cc @oli-obk @RalfJung).

@oli-obk

Tested on commit rust-lang/rust@8c6ce6b. Direct link to PR: <rust-lang/rust#97802> 💔 miri on windows: test-fail → build-fail (cc @oli-obk @RalfJung). 💔 miri on linux: test-pass → build-fail (cc @oli-obk @RalfJung).

rust-timer · 2022-09-03T00:53:45Z

Finished benchmarking commit (8c6ce6b): comparison URL.

Overall result: ✅ improvements - no action needed

@rustbot label: -perf-regression

Instruction count

This is a highly reliable metric that was used to determine the overall result at the top of this comment.

	mean¹	range	count²
Regressions ❌ (primary)	-	-	0
Regressions ❌ (secondary)	-	-	0
Improvements ✅ (primary)	-	-	0
Improvements ✅ (secondary)	-0.8%	[-0.8%, -0.8%]	1
All ❌✅ (primary)	-	-	0

Max RSS (memory usage)

Results

This is a less reliable metric that may be of interest but was not used to determine the overall result at the top of this comment.

	mean¹	range	count²
Regressions ❌ (primary)	-	-	0
Regressions ❌ (secondary)	3.7%	[3.7%, 3.7%]	1
Improvements ✅ (primary)	-	-	0
Improvements ✅ (secondary)	-3.8%	[-8.3%, -2.1%]	5
All ❌✅ (primary)	-	-	0

Cycles

Results

This is a less reliable metric that may be of interest but was not used to determine the overall result at the top of this comment.

	mean¹	range	count²
Regressions ❌ (primary)	2.4%	[2.2%, 2.6%]	2
Regressions ❌ (secondary)	2.4%	[2.4%, 2.4%]	1
Improvements ✅ (primary)	-	-	0
Improvements ✅ (secondary)	-	-	0
All ❌✅ (primary)	2.4%	[2.2%, 2.6%]	2

the arithmetic mean of the percent change ↩ ↩² ↩³
number of relevant changes ↩ ↩² ↩³

Enselic · 2022-09-03T04:51:35Z

📣 Toolstate changed by #97802!

Tested on commit 8c6ce6b. Direct link to PR: #97802

💔 miri on windows: test-fail → build-fail (cc oli-obk RalfJung). 💔 miri on linux: test-pass → build-fail (cc oli-obk RalfJung).

Fixed by rust-lang/miri#2532

(I already pinged oli-obk and RalfJung, so I removed @ from the quote above)

…, r=joshtriplett Support `#[unix_sigpipe = "inherit|sig_dfl"]` on `fn main()` to prevent ignoring `SIGPIPE` When enabled, programs don't have to explicitly handle `ErrorKind::BrokenPipe` any longer. Currently, the program ```rust fn main() { loop { println!("hello world"); } } ``` will print an error if used with a short-lived pipe, e.g. % ./main | head -n 1 hello world thread 'main' panicked at 'failed printing to stdout: Broken pipe (os error 32)', library/std/src/io/stdio.rs:1016:9 note: run with `RUST_BACKTRACE=1` environment variable to display a backtrace by enabling `#[unix_sigpipe = "sig_dfl"]` like this ```rust #![feature(unix_sigpipe)] #[unix_sigpipe = "sig_dfl"] fn main() { loop { println!("hello world"); } } ``` there is no error, because `SIGPIPE` will not be ignored and thus the program will be killed appropriately: % ./main | head -n 1 hello world The current libstd behaviour of ignoring `SIGPIPE` before `fn main()` can be explicitly requested by using `#[unix_sigpipe = "sig_ign"]`. With `#[unix_sigpipe = "inherit"]`, no change at all is made to `SIGPIPE`, which typically means the behaviour will be the same as `#[unix_sigpipe = "sig_dfl"]`. See rust-lang#62569 and referenced issues for discussions regarding the `SIGPIPE` problem itself See the [this](https://rust-lang.zulipchat.com/#narrow/stream/219381-t-libs/topic/Proposal.3A.20First.20step.20towards.20solving.20the.20SIGPIPE.20problem) Zulip topic for more discussions, including about this PR. Tracking issue: rust-lang#97889

rust-highfive assigned joshtriplett Jun 6, 2022

rustbot added T-compiler Relevant to the compiler team, which will review and decide on the PR/issue. T-libs Relevant to the library team, which will review and decide on the PR/issue. labels Jun 6, 2022

rust-highfive added the S-waiting-on-review Status: Awaiting review from the assignee but also interested parties. label Jun 6, 2022

Enselic mentioned this pull request Jun 6, 2022

Should Rust still ignore SIGPIPE by default? #62569

Open

This comment has been minimized.

Sign in to view

jyn514 reviewed Jun 6, 2022

View reviewed changes

compiler/rustc_codegen_ssa/src/base.rs Outdated Show resolved Hide resolved

This comment was marked as outdated.

Sign in to view

rustbot added S-waiting-on-author Status: This is awaiting some action (such as code changes or more information) from the author. and removed S-waiting-on-review Status: Awaiting review from the assignee but also interested parties. labels Jun 6, 2022

bjorn3 reviewed Jun 7, 2022

View reviewed changes

compiler/rustc_codegen_ssa/src/base.rs Outdated Show resolved Hide resolved

Enselic mentioned this pull request Jun 8, 2022

Tracking Issue for unix_sigpipe #97889

Open

23 tasks

Enselic changed the title ~~Add feature no_ignore_sigpipe to never start ignoring SIGPIPE~~ Add attribute #![no_ignore_sigpipe] to stop libstd from ignoring SIGPIPE Jun 8, 2022

klensy reviewed Jun 9, 2022

View reviewed changes

library/std/src/sys/unix/mod.rs Outdated Show resolved Hide resolved

Enselic changed the title ~~Add attribute #![no_ignore_sigpipe] to stop libstd from ignoring SIGPIPE~~ Add attribute #![ignore_sigpipe(no)] to stop libstd from ignoring SIGPIPE Jun 9, 2022

Enselic force-pushed the add-no_ignore_sigkill-feature branch from 2cf4a23 to cf341ec Compare June 9, 2022 17:03

Enselic changed the title ~~Add attribute #![ignore_sigpipe(no)] to stop libstd from ignoring SIGPIPE~~ Add bin crate attribute #![sigpipe_handler(unchanged)] to prevent libstd from ignoring SIGPIPE Jun 10, 2022

Enselic force-pushed the add-no_ignore_sigkill-feature branch from 87ea353 to fe3f47d Compare June 10, 2022 20:14

This comment has been minimized.

Sign in to view

bjorn3 reviewed Jun 11, 2022

View reviewed changes

compiler/rustc_codegen_cranelift/src/main_shim.rs Outdated Show resolved Hide resolved

Enselic force-pushed the add-no_ignore_sigkill-feature branch from 4c97eb7 to 2f738f9 Compare June 13, 2022 04:46

Enselic changed the title ~~Add bin crate attribute #![sigpipe_handler(unchanged)] to prevent libstd from ignoring SIGPIPE~~ Support #[pipeable] on fn main() to prevent ignoring SIGPIPE Jun 13, 2022

bjorn3 added the S-waiting-on-author Status: This is awaiting some action (such as code changes or more information) from the author. label Aug 31, 2022

This comment has been minimized.

Sign in to view

Enselic added 2 commits August 31, 2022 18:20

unix_sigpipe: Inline compiler sigpipe constants in std

236903f

unix_sigpipe: Add docs for init() sigpipe param

3d1a4e4

rustbot added S-waiting-on-review Status: Awaiting review from the assignee but also interested parties. and removed S-waiting-on-author Status: This is awaiting some action (such as code changes or more information) from the author. labels Aug 31, 2022

unix_sigpipe: Make sigpipe param docs long-form

3810d4a

bors added S-waiting-on-bors Status: Waiting on bors to run and complete tests. Bors will change the label on completion. and removed S-waiting-on-review Status: Awaiting review from the assignee but also interested parties. labels Sep 2, 2022

bors added the merged-by-bors This PR was explicitly merged by bors. label Sep 2, 2022

bors merged commit 8c6ce6b into rust-lang:master Sep 2, 2022

rust-highfive mentioned this pull request Sep 2, 2022

miri no longer builds after rust-lang/rust#97802 #101352

Closed

rustbot added this to the 1.65.0 milestone Sep 2, 2022

Enselic deleted the add-no_ignore_sigkill-feature branch September 3, 2022 04:52

joshtriplett mentioned this pull request Sep 11, 2022

Change process spawning to inherit the parent's signal mask by default #101077

Merged

bjorn3 mentioned this pull request Dec 20, 2022

Verify signature of start lang item #105963

Closed

Enselic mentioned this pull request Feb 9, 2024

Stabilize #[unix_sigpipe = "sig_dfl"] on fn main() #120832

Closed

1 task

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Support `#[unix_sigpipe = "inherit|sig_dfl"]` on `fn main()` to prevent ignoring `SIGPIPE` #97802

Support `#[unix_sigpipe = "inherit|sig_dfl"]` on `fn main()` to prevent ignoring `SIGPIPE` #97802

Enselic commented Jun 6, 2022 •

edited

Loading

rust-highfive commented Jun 6, 2022

rust-highfive commented Jun 6, 2022

This comment has been minimized.

This comment was marked as outdated.

This comment has been minimized.

This comment has been minimized.

This comment has been minimized.

Enselic commented Jun 10, 2022

This comment has been minimized.

joshtriplett commented Aug 31, 2022

RalfJung commented Aug 31, 2022 •

edited

Loading

Enselic commented Aug 31, 2022

joshtriplett commented Aug 31, 2022

Enselic commented Sep 2, 2022 •

edited

Loading

bors commented Sep 2, 2022

bors commented Sep 2, 2022

bors commented Sep 2, 2022

rust-highfive commented Sep 2, 2022

rust-timer commented Sep 3, 2022

Enselic commented Sep 3, 2022

Support #[unix_sigpipe = "inherit|sig_dfl"] on fn main() to prevent ignoring SIGPIPE #97802

Support #[unix_sigpipe = "inherit|sig_dfl"] on fn main() to prevent ignoring SIGPIPE #97802

Conversation

Enselic commented Jun 6, 2022 • edited Loading

rust-highfive commented Jun 6, 2022

rust-highfive commented Jun 6, 2022

This comment has been minimized.

This comment was marked as outdated.

This comment has been minimized.

This comment has been minimized.

This comment has been minimized.

Enselic commented Jun 10, 2022

This comment has been minimized.

joshtriplett commented Aug 31, 2022

RalfJung commented Aug 31, 2022 • edited Loading

Enselic commented Aug 31, 2022

joshtriplett commented Aug 31, 2022

Enselic commented Sep 2, 2022 • edited Loading

bors commented Sep 2, 2022

bors commented Sep 2, 2022

bors commented Sep 2, 2022

rust-highfive commented Sep 2, 2022

rust-timer commented Sep 3, 2022

Overall result: ✅ improvements - no action needed

Instruction count

Max RSS (memory usage)

Cycles

Footnotes

Enselic commented Sep 3, 2022

Support `#[unix_sigpipe = "inherit|sig_dfl"]` on `fn main()` to prevent ignoring `SIGPIPE` #97802

Support `#[unix_sigpipe = "inherit|sig_dfl"]` on `fn main()` to prevent ignoring `SIGPIPE` #97802

Enselic commented Jun 6, 2022 •

edited

Loading

RalfJung commented Aug 31, 2022 •

edited

Loading

Enselic commented Sep 2, 2022 •

edited

Loading