Bug 441406

Summary:	ptrace: Stepping out of a signal handler runs away
Product:	[Fedora] Fedora	Reporter:	Jan Kratochvil <jan.kratochvil>
Component:	kernel	Assignee:	Roland McGrath <roland>
Status:	CLOSED CURRENTRELEASE	QA Contact:	Fedora Extras Quality Assurance <extras-qa>
Severity:	low	Docs Contact:
Priority:	low
Version:	8	CC:	kernel-mgr, mjw
Target Milestone:	---
Target Release:	---
Hardware:	i686
OS:	Linux
Whiteboard:
Fixed In Version:	kernel-2.6.25.10-47.fc8.x86_64	Doc Type:	Bug Fix
Doc Text:		Story Points:	---
Clone Of:		Environment:
Last Closed:	2008-07-22 22:50:04 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:	173278, 450378

Description Jan Kratochvil 2008-04-07 21:16:56 UTC

Description of problem:
On repeated PTRACE_SINGLESTEP when a signal gets delivered the signal handler
gets traced but on the `int 0x80' of `__kernel_sigreturn' it starts to
PTRACE_CONT instead.

Version-Release number of selected component (if applicable):
FAILs kernel-2.6.24.4-64.fc8.i686
FAILs kernel-2.6.23.1-42.fc8.i686
FAILs kernel-2.6.25-0.195.rc8.git1.fc9.i686 patched by
  https://www.redhat.com/archives/utrace-devel/2008-April/msg00004.html
PASSes kernel-2.6.24.4-64.fc8.x86_64
PASSes kernel-2.6.20-1.2933.fc6.i686 (as reported by mwielaard)

How reproducible:
Always.

Steps to Reproduce:
wget -O step-through-sigret.c
'http://sources.redhat.com/cgi-bin/cvsweb.cgi/~checkout~/tests/ptrace-tests/tests/step-through-sigret.c?cvsroot=systemtap';
gcc -o step-through-sigret step-through-sigret.c -Wall -ggdb2 -D_GNU_SOURCE;
./step-through-sigret; echo $?

Actual results:
1

Expected results:
0

Additional info:

Comment 1 Mark Wielaard 2008-04-08 09:17:00 UTC

See also:
https://www.redhat.com/archives/utrace-devel/2008-April/msg00009.html

From Wenji Huang
"Verified that the test couldn't pass on > 2.6.20 i686 kernel [...]
 But passed on >= 2.6.22 (vanilla, x86_64) and 2.6.25 (utrace patched, x86_64)"

So it seems to not be utrace specific, but an upstream ptrace/signal bug since
2.6.21 on x86 only.

Comment 2 Jan Kratochvil 2008-04-09 17:53:05 UTC

->ptrace as the Bug is not utrace-related:
  https://www.redhat.com/archives/utrace-devel/2008-April/msg00011.html

Comment 3 Julien Tane 2008-05-30 16:53:39 UTC

I have the same problem on Fedora 9 with gcc :  gcc-4.3.0-8.i386,
glibc-headers-2.8-3.i386 (same version of glibc of course)
I downloaded the script given above and I also obtain 

1 
instead of:
0

This is probably related to my difficulties starting user-mode-linux kernels
from: <http://user-mode-linux.sourceforge.net/> or
<http://uml.nagafix.co.uk/kernels/>

Comment 4 Jan Kratochvil 2008-07-22 22:50:04 UTC

Thanks, verified as fixed on kernel-2.6.25.10-47.fc8.x86_64.