Bug 1142317 – opiegen segfaults on OBS

Bug 1142317 - opiegen segfaults on OBS


Summary:	opiegen segfaults on OBS

Status:	NEW

Classification:	openSUSE
Product:	openSUSE Tumbleweed
Classification:	openSUSE
Component:	Basesystem
Version:	Current
Hardware:	Other Other

Priority:	P5 - None Severity: Normal (vote)
Target Milestone:	---
Assigned To:	Marcus Meissner
QA Contact:	E-mail List

URL:
Whiteboard:
Keywords:

Depends on:
Blocks:
	Show dependency tree / graph

Create test case

Clone This Bug

Reported:	2019-07-22 08:52 UTC by Luigi Baldoni
Modified:	2019-07-26 09:39 UTC (History)
CC List:	2 users (show)

See Also:
Found By:	---
Services Priority:
Business Priority:
Blocker:	---
Marketing QA Status:	---
IT Deployment:	---

Attachments
opiegen run through gdb (1.33 KB, text/x-log) 2019-07-22 08:52 UTC, Luigi Baldoni	Details
opiegen run through valgrind (1.90 KB, text/x-log) 2019-07-22 08:53 UTC, Luigi Baldoni	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 Luigi Baldoni 2019-07-22 08:52:48 UTC

Created attachment 811139 [details]
opiegen run through gdb

When having the package opie as buildtime requirement, the opiegen executable segfaults. The same occurs inside an osc chroot.

opiegen runs fine otherwise.

Comment 1 Luigi Baldoni 2019-07-22 08:53:35 UTC

Created attachment 811140 [details]
opiegen run through valgrind

Comment 2 Marcus Meissner 2019-07-26 05:57:58 UTC

I had tried this locally, but could not yet get it to crash.

Comment 3 Marcus Meissner 2019-07-26 06:19:58 UTC

ok, in osc build chroot it does.

      int i = __opiegetutmpentry(ttyname(0), &utmp);

ttyname(0) is returning NULL

Comment 4 Marcus Meissner 2019-07-26 06:20:34 UTC

mount -t devpts none /dev/pts/

in chroot and it starts to work

Comment 5 Luigi Baldoni 2019-07-26 07:08:56 UTC

Very interesting. Any idea why it would fail only on Factory but not on Leap?

Comment 6 Marcus Meissner 2019-07-26 09:39:54 UTC

might be a glibc change that no longer provides a fall for ttyname(0) if devpts is not mounted.

Not sure if we need to fiix this, I would add a safety check in the opiue code at most