Bug 444996

Summary:	glibc cacoshl() miscompiled by gcc-4.3
Product:	[Fedora] Fedora	Reporter:	Bill Nottingham <notting>
Component:	glibc	Assignee:	Jakub Jelinek <jakub>
Status:	CLOSED CURRENTRELEASE	QA Contact:	Fedora Extras Quality Assurance <extras-qa>
Severity:	low	Docs Contact:
Priority:	low
Version:	9	CC:	drepper, fweimer, rvokal, wwoods
Target Milestone:	---
Target Release:	---
Hardware:	powerpc
OS:	Linux
Whiteboard:
Fixed In Version:		Doc Type:	Bug Fix
Doc Text:		Story Points:	---
Clone Of:		Environment:
Last Closed:	2008-12-30 17:08:50 UTC	Type:	---
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:	235705

Description Bill Nottingham 2008-05-02 16:59:09 UTC

Description of problem:

See: http://gcc.gnu.org/bugzilla/show_bug.cgi?id=36090

Version-Release number of selected component (if applicable):

glibc-2.8-2

Comment 1 Will Woods 2008-05-05 15:05:24 UTC

Should be fixed in glibc-2.8-3. Build available here:
http://koji.fedoraproject.org/koji/buildinfo?buildID=48111

Comment 2 Will Woods 2008-05-08 16:37:17 UTC

Can someone verify the fix, or give me a suggestion about how to verify it?

Comment 3 Jakub Jelinek 2008-05-08 16:51:32 UTC

It wasn't fixed, just worked around.  The real fix is being checked into GCC
today, so too late for F9.
Anyway, to verify cacoshl etc., just look for
build-ppc64-linuxnptl/math/test-ldouble.out
and
build-ppc64-linuxnptl/math/test-ildoubl.out
in TESTING DETAILS in glibc 2.8-2 vs. 2.8-3 ppc64 build.log.
In 2.8-2 you will see a lot of errors, in 2.8-3 it isn't present (because the
test succeeded).
On ppc/ppc64 you need to ignore all the power6/ failures, because the lame koji
buildboxes don't have Altivec support, to get a clean testing you need to build
on Power6 or at least some Power box with Altivec ISA (e.g. G5).

Comment 4 Will Woods 2008-05-09 20:25:53 UTC

Thanks for the info, Jakub. Since we have a workaround in F9 I'm going to move
this off the blocker list.

I'm moving it to F9Target, under the assumption that the real fix will show up
as a F9 update.

Comment 5 Bug Zapper 2008-05-14 10:32:11 UTC

Changing version to '9' as part of upcoming Fedora 9 GA.
More information and reason for this action is here:
http://fedoraproject.org/wiki/BugZappers/HouseKeeping

Comment 6 Ulrich Drepper 2008-12-30 17:08:50 UTC

The new gcc has long been deployed.  Closing.