Bug 230423

Summary: ptrace causing SIGSEGV in debugged process when sending SIGKILL to it.
Product: [Fedora] Fedora Reporter: Kevin Van Brunt <kvanbrunt>
Component: kernelAssignee: Roland McGrath <roland>
Status: CLOSED CURRENTRELEASE QA Contact: Brian Brock <bbrock>
Severity: medium Docs Contact:
Priority: medium    
Version: 6CC: cebbert, davej, wtogami
Target Milestone: ---   
Target Release: ---   
Hardware: i386   
OS: Linux   
Whiteboard:
Fixed In Version: 2.6.20-1.2925.fc6 Doc Type: Bug Fix
Doc Text:
Story Points: ---
Clone Of: Environment:
Last Closed: 2007-03-20 17:27:07 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:
Attachments:
Description Flags
A test c program that demonstrates the bug. none

Description Kevin Van Brunt 2007-02-28 19:37:56 UTC 
From Bugzilla Helper:
User-Agent: Mozilla/4.0 (compatible; MSIE 7.0; Windows NT 5.1; .NET CLR 1.1.4322; .NET CLR 2.0.50727; .NET CLR 3.0.04506.30)

Description of problem:
I have written a small program to demonstrate this bug.  It causes ptrace to trace /bin/ls.  After the first waitpid, the parent sends a SIGKILL to ls and ls performs a segmentation fault.  /bin/ls is only used as an example.  Any traced process has the same problem.  I have traced xterm, firefox, and gnome-cd as well and seen the bug.

The code will be attached to this bug report.

Version-Release number of selected component (if applicable):
kernel-2.6.19-1.2911.fc6

How reproducible:
Always


Steps to Reproduce:
1. Compile and run the attached code.

Actual Results:
The message "GOT SIGSEGV" should appear because the traced program performs a segmentation fault.

Expected Results:
The traced program should not perform a segmentation fault and no message should appear.

Additional info:
Comment 1 Kevin Van Brunt 2007-02-28 19:41:19 UTC 
Created attachment 148952 [details]
A test c program that demonstrates the bug.
Comment 2 Albert Cahalan 2007-03-07 04:31:25 UTC 
A pure 2.6.20.1 kernel does not show the problem at all. I even tried adding
sched_yield in various places to see if scheduling order would affect things.

This is probably a utrace bug.
Comment 3 Roland McGrath 2007-03-19 23:03:55 UTC 
I reproduced this on 2.6.19-1.2911.fc6 but not on 2.6.20-1.2925.fc6.
Please try the latest fc6 update kernel and verify that the problem is gone for you.
Comment 4 Kevin Van Brunt 2007-03-20 17:20:41 UTC 
It looks to be fixed in 2.6.20-1.2925.fc6.