Bug 464104

Summary:	sys-apps/portage-2.2.0_alpha169: sys-libs/glibc-2.17 upgrade - /usr/bin/python2.7: relocation error: /lib64/libresolv.so.2: symbol __sendmmsg, version GLIBC_PRIVATE not defined in file libc.so.6 with link time reference
Product:	Gentoo Linux	Reporter:	Martin von Gagern <Martin.vGagern>
Component:	[OLD] Core system	Assignee:	Portage team <dev-portage>
Status:	RESOLVED OBSOLETE
Severity:	normal	CC:	flyser42, grimm26, stefan.tell, toralf, wschlich
Priority:	Normal
Version:	unspecified
Hardware:	All
OS:	Linux
See Also:	https://bugs.gentoo.org/show_bug.cgi?id=448858
Whiteboard:
Package list:		Runtime testing required:	---
Attachments:	emerge --info AMD64 emerge --info x86 emerge --info

Description Martin von Gagern 2013-04-01 17:25:10 UTC

Created attachment 343974 [details]
emerge --info

After updating glibc from 2.16.0 to 2.17, my world update emerge process terminated unexpectedly:

>>> Installing (1 of 121) sys-libs/glibc-2.17
 * Defaulting /etc/host.conf:multi to on
 * >>> SetUID: [chmod go-r] /usr/lib64/misc/glibc/pt_chown ... [ ok ]
 * >>> SetUID: [chmod go-r] /usr/lib32/misc/glibc/pt_chown ... [ ok ]
 * Generating locale-archive: forcing # of jobs to 1
 * Generating 27 locales (this might take a while) with 1 jobs
 *  ( 1/27) Generating en_US.ISO-8859-1 ... [ ok ]
[…]
 *  (27/27) Generating ja_JP.EUC-JP ... [ ok ]
 * Generation complete
/usr/bin/python2.7: relocation error: /lib64/libresolv.so.2: symbol __sendmmsg, version GLIBC_PRIVATE not defined in file libc.so.6 with link time reference

This appears closely related to bug #125868, but the actual function in questin is a different one this time, as are the versions involved and the possible causes, nptl stuff in particular.

I very much hope that my glibc install is complete despite this error. qcheck mentions no missing files, and the locale generation looks fine, too. I guess I'll re-emerge glibc again nevertheless, just to be sure.

Comment 1 Tom Wijsman (TomWij) (RETIRED) gentoo-dev

2013-04-01 23:19:05 UTC

Please attach the complete build log.

Comment 2 SpanKY gentoo-dev

2013-04-02 01:09:27 UTC


*** This bug has been marked as a duplicate of bug 448858 ***

Comment 3 Zac Medico gentoo-dev

2013-04-03 05:55:18 UTC

I went through this upgrade myself yesterday, but I didn't experience a relocation error. However, I running emerge with python3.3 instead of python2.7 (portage is installed with USE=python3), so maybe that makes a difference.

Comment 4 SpanKY gentoo-dev

2013-04-03 06:30:06 UTC

i use python-2.7 myself but haven't seen this

Comment 5 SpanKY gentoo-dev

2014-01-04 13:21:56 UTC

*** Bug 496952 has been marked as a duplicate of this bug. ***

Comment 6 Agostino Sarubbo gentoo-dev

2014-01-04 18:27:04 UTC

(In reply to SpanKY from comment #4)
> i use python-2.7 myself but haven't seen this

me too.

Comment 7 Martin von Gagern 2014-01-04 19:33:47 UTC

(In reply to Toralf Förster from bug #496952 comment #1)
> FWIW I do use tmpfs

So do I. Coincidence? I guess so, but I'm reporting this just in case it is not.

Comment 8 Toralf Förster gentoo-dev

2014-01-04 19:54:24 UTC

(In reply to Martin von Gagern from comment #7)
> (In reply to Toralf Förster from bug #496952 comment #1)
> > FWIW I do use tmpfs
> 
> So do I. Coincidence? I guess so, but I'm reporting this just in case it is
> not.

When I run the first time into this issue (I found a similar bug I filed about this a year ago : #bug 448858 with glibc-2.17) I used a tmpfs file system too - in a more complex environemnr.

I do chrooted into a ~x86 user mode linux image and I mounted /var/tmp/portage onto a mount point of a ramdisk (tmpfs) before to speed up things and b/c the due to the fact that UML image doesn't have enough space in it for compilation.

Comment 9 Mark Keisler 2014-01-07 05:28:19 UTC

FYI, I got this error tonight, then ran the emerge again with keepwork and it ran without error the second time :)

Comment 10 Sebastian Schocke 2014-01-31 08:50:06 UTC

Regarding this bug, I just hit it on two of our machines at the office, one running x86 and Python 3.3, the other running AMD64 and Python 3.3.

No other side effects from the bug, and glibc-2.17 seems to be installed correctly, but I had to restart the emerge -aDNuv world after crashing

Comment 11 Sebastian Schocke 2014-01-31 08:52:06 UTC

Created attachment 369186 [details]
AMD64 emerge --info

Comment 12 Sebastian Schocke 2014-01-31 08:54:17 UTC

Created attachment 369188 [details]
x86 emerge --info

Comment 13 Andreas K. Hüttel archtester

2017-09-10 09:00:35 UTC

Given that 2.17 is masked and the oldest in the tree, this is probably obsolete now.