|
From: Mike S. <ms...@mo...> - 2004-06-17 21:43:56
|
Hello, I'm seeing testsuite failures in valgrind on Redhat AS 2.1... I reported this as a bug a few weeks ago: http://bugs.kde.org/show_bug.cgi?id=82114 but didn't hear anything at all back from that... so I figured I'd check the user lists to see if anyone knows a way around the problem or where to start trying to figure out what is wrong. The short version is it appears SIGCHLD signal handling is messed up ... when a child exits valgrind locks up solid and requires a kill -9 to "recover". - Has anyone been using the CVS HEAD versions on Redhat AS 2.1? I tried checking out and building several cvs tags of valgrind hoping to find when this behavior was introduced. Unfortunately all 2.* versions appear to have the problem. The 1.9.5 version didn't appear to have the problem, but "regtest" didn't exist back then (when I tried to manually run the tests they worked). Much has changed between 2.0.0 and 1.9.5, so I'm looking for ideas on where to start... Valgrind locks up after reporting an assert similar to the following: === proxy 13480 for tid 1 exited status -1, res 0 got signal 17 in LWP 13473 (13473) valgrind: vg_signals.c:2015 (vg_async_signalhandler): Assertion `vgPlain_ksigismember(&uc->uc_sigmask, sigNo)' failed. ==13473== at 0xB802DE80: vgPlain_skin_assert_fail (vg_mylibc.c:1211) sched status: Thread 1: status = Runnable, associated_mx = 0x0, associated_cv = 0x0 ==13473== at 0x810D8959: wait4 (in /lib/libc-2.2.4.so) ==13473== by 0x8048E82: main (fdleak_cmsg.c:176) === |
|
From: Banibrata D. <du...@in...> - 2004-06-18 05:28:12
|
I've been using CVS HEAD (i think 2.1.2) on RH-AS2.1, but have not debugged programs handling signals in any non-default way etc. Most of the times I'd had a signal reported problem, and debugged that using Valgrind, i've faced problems, and have been told that support for signals in Valgrind is weak. So you may be running up against such a known issue. regards, bd > -----Original Message----- > From: val...@li... > [mailto:val...@li...] On Behalf > Of Mike Simons > Sent: Friday, June 18, 2004 3:14 AM > To: val...@li... > Subject: [Valgrind-users] CVS latest valgrind + AS 2.1 + > SIGCHLD == hang > > > Hello, > > I'm seeing testsuite failures in valgrind on Redhat AS 2.1... > I reported this as a bug a few weeks ago: > http://bugs.kde.org/show_bug.cgi?id=82114 > but didn't hear anything at all back from that... so I > figured I'd check the user lists to see if anyone knows a way > around the problem or where to start trying to figure out > what is wrong. > > The short version is it appears SIGCHLD signal handling is > messed up ... when a child exits valgrind locks up solid and > requires a kill -9 to "recover". > > - Has anyone been using the CVS HEAD versions on Redhat AS 2.1? > > > I tried checking out and building several cvs tags of > valgrind hoping to find when this behavior was introduced. > Unfortunately all 2.* versions appear to have the problem. > The 1.9.5 version didn't appear to have the problem, but > "regtest" didn't exist back then (when I tried to manually > run the tests they worked). > > Much has changed between 2.0.0 and 1.9.5, so I'm looking for > ideas on where to start... > > > Valgrind locks up after reporting an assert similar to the > following: === proxy 13480 for tid 1 exited status -1, res 0 > got signal 17 in LWP 13473 (13473) > > valgrind: vg_signals.c:2015 (vg_async_signalhandler): > Assertion `vgPlain_ksigismember(&uc->uc_sigmask, sigNo)' failed. > ==13473== at 0xB802DE80: vgPlain_skin_assert_fail > (vg_mylibc.c:1211) > > sched status: > > Thread 1: status = Runnable, associated_mx = 0x0, associated_cv = 0x0 > ==13473== at 0x810D8959: wait4 (in /lib/libc-2.2.4.so) > ==13473== by 0x8048E82: main (fdleak_cmsg.c:176) > === > > > ------------------------------------------------------- > This SF.Net email is sponsored by The 2004 JavaOne(SM) > Conference Learn from the experts at JavaOne(SM), Sun's > Worldwide Java Developer Conference, June 28 - July 1 at the > Moscone Center in San Francisco, CA REGISTER AND SAVE! http://java.sun.com/javaone/sf Priority Code NWMGYKND _______________________________________________ Valgrind-users mailing list Val...@li... https://lists.sourceforge.net/lists/listinfo/valgrind-users |
|
From: Mike S. <ms...@mo...> - 2004-06-18 18:35:10
|
On Fri, Jun 18, 2004 at 10:53:58AM +0530, Banibrata Dutta wrote: > I've been using CVS HEAD (i think 2.1.2) on RH-AS2.1, but have > not debugged programs handling signals in any non-default way etc. > Most of the times I'd had a signal reported problem, and debugged > that using Valgrind, i've faced problems, and have been told that > support for signals in Valgrind is weak. So you may be running up > against such a known issue. - Have you run "make regtest" in the valgrind source? - Can you confirm the testsuite hangs for you too? The same CVS version, make regtest doesn't lock up on both Debian/testing and AS 3.0. (both of those report a few dozen errors in the testsuite, but it doesn't get wedged). Thanks, Mike Simons > > -----Original Message----- > > Subject: [Valgrind-users] CVS latest valgrind + AS 2.1 + SIGCHLD == hang > > > > > > I'm seeing testsuite failures in valgrind on Redhat AS 2.1... > > I reported this as a bug a few weeks ago: > > http://bugs.kde.org/show_bug.cgi?id=82114 [...] > > The short version is it appears SIGCHLD signal handling is > > messed up ... when a child exits valgrind locks up solid and > > requires a kill -9 to "recover". |