aio_*(2): mention ENOSYS under ERRORS
ClosedPublic
Actions

Authored by ngie on Mar 25 2025, 10:00 PM.

Details

Reviewers

ziaee

Commits

rG9620fa7091a2: aio_*(2): mention ENOSYS under ERRORS
rGd97e44784bb5: aio_*(2): mention ENOSYS under ERRORS

Summary

ENOSYS can occur if aio(4) is not loaded in the kernel. Document this
behavior so consumers on FreeBSD can better understand that this is a
possible scenario.

Clean up the manpages slightly while here:

Sort ERRORS by errno(3).
Use .Fx instead of FreeBSD.

PR: 190942

Test Plan

igor/manlint pass with the changes (ignore the .Dd complaints):

% for i in aio_cancel.2 aio_error.2 aio_fsync.2 aio_mlock.2 aio_read.2 aio_return.2 aio_suspend.2 aio_waitcomplete.2 aio_write.2 ; do mandoc -Tlint  $PWD/$i; done
% for i in aio_cancel.2 aio_error.2 aio_fsync.2 aio_mlock.2 aio_read.2 aio_return.2 aio_suspend.2 aio_waitcomplete.2 aio_write.2 ; do igor $PWD/$i; done
aio_cancel.2:25:date not today, March 25, 2025:.Dd [January 19, 2000]
aio_error.2:25:date not today, March 25, 2025:.Dd [January 2, 2021]
aio_fsync.2:25:date not today, March 25, 2025:.Dd [November 15, 2023]
aio_mlock.2:25:date not today, March 25, 2025:.Dd [August 19, 2016]
aio_read.2:25:date not today, March 25, 2025:.Dd [February 1, 2024]
aio_return.2:25:date not today, March 25, 2025:.Dd [January 11, 2021]
aio_suspend.2:25:date not today, March 25, 2025:.Dd [October 23, 2017]
aio_waitcomplete.2:25:date not today, March 25, 2025:.Dd [March 21, 2016]
aio_write.2:25:date not today, March 25, 2025:.Dd [February 1, 2024]

The rendered output looks good after the change.

Diff Detail

Repository

rG FreeBSD src repository

Lint

Lint Skipped

Unit

Tests Skipped

Build Status

Buildable 63119
Build 60003: arc lint + arc unit

Event Timeline

ngie created this revision.Mar 25 2025, 10:00 PM

Herald added subscribers: ziaee, imp. · View Herald TranscriptMar 25 2025, 10:00 PM

ngie requested review of this revision.Mar 25 2025, 10:00 PM

Harbormaster completed remote builds in B63119: Diff 152670.Mar 25 2025, 10:00 PM

ngie edited the test plan for this revision. (Show Details)Mar 25 2025, 10:03 PM

ngie added a reviewer: ziaee.

ngie edited subscribers, added: linimon; removed: ziaee.

Thank you so much for doing this!

This revision is now accepted and ready to land.Mar 25 2025, 10:20 PM

Do you need to mention the SIGSYS issue? Unless you have blocked SIGSYS, any system call that returns ENOSYS will generate that signal.

In D49502#1128871, @imp wrote:

Do you need to mention the SIGSYS issue? Unless you have blocked SIGSYS, any system call that returns ENOSYS will generate that signal.

Good question. It looks like ENOSYS without SIGSYS is the standard in libsys:

% rg -l SIGSYS lib/libsys/*.2 | wc -l
       1
% rg -l ENOSYS lib/libsys/*.2 | wc -l
      24

In D49502#1128876, @ngie wrote:
In D49502#1128871, @imp wrote:

Do you need to mention the SIGSYS issue? Unless you have blocked SIGSYS, any system call that returns ENOSYS will generate that signal.

Good question. It looks like ENOSYS without SIGSYS is the standard in libsys:
% rg -l SIGSYS lib/libsys/*.2 | wc -l
       1
% rg -l ENOSYS lib/libsys/*.2 | wc -l
      24

Ok looks like intro covers it

In D49502#1128880, @imp wrote:
In D49502#1128876, @ngie wrote:
In D49502#1128871, @imp wrote:

Do you need to mention the SIGSYS issue? Unless you have blocked SIGSYS, any system call that returns ENOSYS will generate that signal.

Good question. It looks like ENOSYS without SIGSYS is the standard in libsys:
% rg -l SIGSYS lib/libsys/*.2 | wc -l
       1
% rg -l ENOSYS lib/libsys/*.2 | wc -l
      24
Ok looks like intro covers it