Bug 2074663

Summary:	darktable build fails on ppc64le only
Product:	Red Hat Enterprise Linux 8	Reporter:	Germano Massullo <germano.massullo>
Component:	gcc-toolset-11	Assignee:	Marek Polacek <mpolacek>
Status:	CLOSED NOTABUG	QA Contact:	qe-baseos-tools-bugs
Severity:	unspecified	Docs Contact:
Priority:	unspecified
Version:	8.5	CC:	dan, sipoyare
Target Milestone:	rc
Target Release:	---
Hardware:	Unspecified
OS:	Unspecified
Whiteboard:
Fixed In Version:		Doc Type:	If docs needed, set a value
Doc Text:		Story Points:	---
Clone Of:		Environment:
Last Closed:	2022-04-13 13:03:48 UTC	Type:	Bug
Regression:	---	Mount Type:	---
Documentation:	---	CRM:
Verified Versions:		Category:	---
oVirt Team:	---	RHEL 7.3 requirements from Atomic Host:
Cloudforms Team:	---	Target Upstream Version:
Embargoed:
Bug Depends On:
Bug Blocks:	2074700

Description Germano Massullo 2022-04-12 18:11:29 UTC

Good day.
Following darktable build fails on ppc64le only. In past a similar problem was caused by a bug in OpenMP / GCC, so I wanted to report this problem here.

epel8-next https://koji.fedoraproject.org/koji/taskinfo?taskID=85563318
epel8 https://koji.fedoraproject.org/koji/taskinfo?taskID=85568505

the compiler seems to fail on file darktable/src/common/fast_guided_filter.h
Best regards

Comment 1 Marek Polacek 2022-04-12 18:24:55 UTC

So this looks like a dup of https://bugzilla.redhat.com/show_bug.cgi?id=2038839

What is different this time?  What gcc is used to build darktable?

Comment 2 Germano Massullo 2022-04-12 19:53:33 UTC

gcc-toolset-11 has been now used. Instead in 
https://bugzilla.redhat.com/show_bug.cgi?id=2038839
it was GCC 8.5

Comment 3 Marek Polacek 2022-04-12 20:21:53 UTC

I see.  Argh, I thought this worked: https://bugzilla.redhat.com/show_bug.cgi?id=2038839#c1
So no clue why it doesn't work now.  Will have to try to reproduce...

Comment 4 Germano Massullo 2022-04-12 20:26:49 UTC

If you need any help, let me know

Comment 5 Germano Massullo 2022-04-13 07:48:44 UTC

In devel.org I have been told that a similar problem is being solved in
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=102059

Comment 6 Dan Horák 2022-04-13 09:18:02 UTC

Seems the darktable build isn't using gcc 11 from the toolset, but the system one (8.5). The toolset's gcc isn't enabled correctly, a PR will follow.

Comment 7 Germano Massullo 2022-04-13 10:50:22 UTC

Hello, this bugreport can be closed as "not a bug" since I did not correctly enable GCC toolset 11, but this has now been fixed by Dan Horák in
https://src.fedoraproject.org/rpms/darktable/pull-request/6
and
https://src.fedoraproject.org/rpms/darktable/pull-request/7

and the package is correctly building
https://koji.fedoraproject.org/koji/buildinfo?buildID=1948510

Cheers

Comment 8 Marek Polacek 2022-04-13 13:03:48 UTC

(In reply to Germano Massullo from comment #7)
> Hello, this bugreport can be closed as "not a bug" since I did not correctly
> enable GCC toolset 11, but this has now been fixed by Dan Horák in
> https://src.fedoraproject.org/rpms/darktable/pull-request/6
> and
> https://src.fedoraproject.org/rpms/darktable/pull-request/7
> 
> and the package is correctly building
> https://koji.fedoraproject.org/koji/buildinfo?buildID=1948510
> 
> Cheers

Great.

Closing then.