895080 – =dev-lang/perl-5.36.0-r1 fails to build with -Og (or other low optimization levels) w/ gcc-12 "error: inlining failed in call to 'always_inline' [...]: function not considered for inlining"

Bug 895080 - =dev-lang/perl-5.36.0-r1 fails to build with -Og (or other low optimization levels) w/ gcc-12 "error: inlining failed in call to 'always_inline' [...]: function not considered for inlining"

Summary: =dev-lang/perl-5.36.0-r1 fails to build with -Og (or other low optimization l...

Status:	UNCONFIRMED

Alias:	None

Product:	Gentoo Linux
Classification:	Unclassified
Component:	Current packages (show other bugs)
Hardware:	All Linux

Importance:	Normal normal (vote)
Assignee:	Gentoo Perl team

URL:
Whiteboard:
Keywords:	PATCH

Depends on:
Blocks:	gcc-12
	Show dependency tree

Reported:	2023-02-17 09:47 UTC by Étienne Buira
Modified:	2023-09-28 11:49 UTC (History)
CC List:	1 user (show)

See Also:	https://github.com/Perl/perl5/issues/19776 https://github.com/Perl/perl5/pull/20562 https://gcc.gnu.org/bugzilla/show_bug.cgi?id=107931 902075 896444 911727
Package list:
Runtime testing required:	---

Attachments
build.log (build.log,119.37 KB, text/x-log) 2023-02-17 09:49 UTC, Étienne Buira	Details
View All Add an attachment (proposed patch, testcase, etc.)

Note You need to log in before you can comment on or make changes to this bug.

Description Étienne Buira 2023-02-17 09:47:27 UTC

Perl fails to build with many errors like in $subject.
As a side-node, the build happens in a chroot (made like for installation).

Reproducible: Always

Steps to Reproduce:
1.emerge =dev-lang/perl-5.36.0-r1
2.
3.
Actual Results:  
C compiler errors out

Expected Results:  
Successful merge

Comment 1 Étienne Buira 2023-02-17 09:49:08 UTC

Created attachment 851994 [details]
build.log

Comment 2 Sam James archtester

2023-02-17 09:50:13 UTC

Please also include emerge --info if not already working on it.

Anyway, this is because of -Og. It's a valid bug that should be sent upstream (could you do it please?) but it's quite common when people use always_inline incorrectly.

Comment 3 Étienne Buira 2023-02-17 10:00:48 UTC

Thanks, trying with -O2
An upstream ticket already exists on this matter (#19776), and resolved by pull request #20562, didn't try this patch.

Comment 4 Sam James archtester

2023-02-17 10:14:36 UTC

(In reply to Étienne Buira from comment #3)
> Thanks, trying with -O2
> An upstream ticket already exists on this matter (#19776), and resolved by
> pull request #20562, didn't try this patch.

Comment 5 Sam James archtester

2023-02-17 10:17:01 UTC

(In reply to Sam James from comment #4)
> (In reply to Étienne Buira from comment #3)
> > Thanks, trying with -O2
> > An upstream ticket already exists on this matter (#19776), and resolved by
> > pull request #20562, didn't try this patch.

oops, didn't write anything: thank you!