Bug 104368 - set_thread_area races kill/pthread_handle_sigcancel [linuxltfs]
Summary: set_thread_area races kill/pthread_handle_sigcancel [linuxltfs]
Keywords:
Status: CLOSED CURRENTRELEASE
Alias: None
Product: Red Hat Enterprise Linux 3
Classification: Red Hat
Component: glibc
Version: 3.0
Hardware: i686
OS: Linux
medium
medium
Target Milestone: ---
Assignee: Jakub Jelinek
QA Contact: Brian Brock
URL:
Whiteboard:
Depends On:
Blocks:
TreeView+ depends on / blocked
 
Reported: 2003-09-13 16:38 UTC by John Reiser
Modified: 2007-11-30 22:06 UTC (History)
1 user (show)

Fixed In Version: 2.3.2-87
Doc Type: Bug Fix
Doc Text:
Clone Of:
Environment:
Last Closed: 2005-06-30 19:20:56 UTC
Target Upstream Version:
Embargoed:


Attachments (Terms of Use)
strace -f tst-cancel2 [edited for brevity] (4.50 KB, text/plain)
2003-09-13 16:40 UTC, John Reiser
no flags Details

Description John Reiser 2003-09-13 16:38:35 UTC
From Bugzilla Helper:
User-Agent: Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.4) Gecko/20030902

Description of problem:
There is a race in the way that linuxltfs uses set_thread_area() after clone().
 With suitable delays, the new thread can get canceled before set_thread_area()
takes effect.  When this happens, the signal handler in the new thread,
pthread_handle_sigcancel(), will get a SIGSEGV because it uses segment register
%gs but the corresponding segment descriptor has not yet been setup by
set_thread_area().

Version-Release number of selected component (if applicable):
glibc-2.3.2-85

How reproducible:
Sometimes

Steps to Reproduce:
1. run testcases build-i686-linuxltfs/linuxthreads/tst-cancel2 or tst-cancel3,
but provide a slowdown of 100:1 [one hundred to one] or more
2.
3.
    

Actual Results:  Attached partial strace of running tst-cancel2 under a
memory-access checker [provides a slowdown as a side effect] shows that the
fork()ed watchdog sends SIGRT_1 to a child thread before the child thread
invokes set_thread_area().

Expected Results:  Signals whose handler uses register %gs must not be sent to a
thread that does not have set_thread_area() already setup.

Additional info:

Comment 1 John Reiser 2003-09-13 16:40:34 UTC
Created attachment 94475 [details]
strace -f tst-cancel2  [edited for brevity]

Comment 2 Jakub Jelinek 2003-09-14 20:28:13 UTC
I don't see the problem.
Before DO_SET_THREAD_AREA, %gs is what the initial INIT_THREAD_SELF (, 0)
chose and kernel tls_array in the kernel points to manager_thread.
If SIGCANCEL arrives at this moment, pthread_handle_sigcancel should catch this
(self == manager_thread) and do the right thing.
Now DO_SET_THREAD_AREA calls set_thread_area syscall, changing tls_array
from manager_thread to the actual thread descriptor. Whatever moment the signal
arrives, pthread_handle_sigcancel should either see self == manager_thread or
self == the_new_thread_descriptor, nothing else.

Comment 3 John Reiser 2003-09-15 14:05:05 UTC
OK; I did not understand the part about pthread_handle_sigcancel handling the
situation in both cases (race won, race lost).  This seems to mean that signal
handlers cannot access thread-local storage under linuxltfs (hard to setup
anyway [?], but does mean additional compatibility concerns for apps: "nptl app"
will be unreliable under linuxltfs), and that signal handlers cannot call any
pthread_* functions except pthread_handle_sigcancel.

Comment 4 Jakub Jelinek 2003-09-17 19:00:40 UTC
glibc-2.3.2-87 now does the self == manager_thread -> find self from stack, INIT_THREAD_SELF
in signal handler wrappers too.


Note You need to log in before you can comment on or make changes to this bug.