|
From: Christoph B. <bar...@gm...> - 2006-09-05 18:56:13
Attachments:
trace
|
Hi,
I could write a small test programm to reproduce the error I reported earlier
today.
The end of the trace is attached.
Here is the programm:
extern "C" {
#include <unistd.h>
}
int main() {
while (true) {
sleep(1);
}
}
1. Compile it with g++ -g <programmname>
2. Start valgrind:
valgrind --tool=callgrind --dump-instr=yes --collect-jumps=yes \
--simulate-cache=yes --instr-atstart=no --num-callers=32 --error-limit=no \
--time-stamp=yes --log-file=vg --trace-children=no --trace-syscalls=yes \
programmname
3. Start instrumentation: callgrind_control -i on <PID>
4. Dump: callgrind_control -d <PID>
5. Valgrind crashed.
The error does not occur on i386 machines only on x86_64 ones. Maybe a 64bit
problem.
If you need more information, just ask.
Greetings
Christoph
|
|
From: Christoph B. <bar...@gm...> - 2006-09-07 18:45:36
|
Hi, Here is another crash of the same problem: sleep BB# 418 Callgrind: callstack.c:211 (vgCallgrind_push_call_stack): Assertion 'current_entry->cxt != 0' failed. ==21768== at 0x380189DA: report_and_quit (m_libcassert.c:136) ==21768== by 0x38018D3D: vgPlain_assert_fail (m_libcassert.c:200) ==21768== by 0x38010CAB: vgCallgrind_push_call_stack (callstack.c:211) ==21768== by 0x38006864: vgCallgrind_setup_bbcc (bbcc.c:529) ==21768== by 0x4028D5181: ??? ==21768== by 0x38020A87: mkFreeBlock (m_mallocfree.c:916) ==21768== by 0x4028818FF: ??? ==21768== by 0x4021354AF: ??? ==21768== by 0x10: ??? ==21768== by 0xCBB4F: ??? ==21768== by 0x4005B5: main (debug-prog2.C:8) Here is also a minimal set of options necessary for the problem: valgrind --tool=callgrind ./a.out but one then has to interate callgrind_control -i off callgrind_control -i on callgrind_control -d till valgrind crashes. Greetings, Christoph Bartoschek |
|
From: Josef W. <Jos...@gm...> - 2006-09-08 09:18:05
|
Hi Christoph, thanks. I can reproduce it. I just hope that I have enough time in the next days to look at it. Josef On Thursday 07 September 2006 20:45, Christoph Bartoschek wrote: > Hi, > > Here is another crash of the same problem: > > sleep > BB# 418 > > Callgrind: callstack.c:211 (vgCallgrind_push_call_stack): > Assertion 'current_entry->cxt != 0' failed. > ==21768== at 0x380189DA: report_and_quit (m_libcassert.c:136) > ==21768== by 0x38018D3D: vgPlain_assert_fail (m_libcassert.c:200) > ==21768== by 0x38010CAB: vgCallgrind_push_call_stack (callstack.c:211) > ==21768== by 0x38006864: vgCallgrind_setup_bbcc (bbcc.c:529) > ==21768== by 0x4028D5181: ??? > ==21768== by 0x38020A87: mkFreeBlock (m_mallocfree.c:916) > ==21768== by 0x4028818FF: ??? > ==21768== by 0x4021354AF: ??? > ==21768== by 0x10: ??? > ==21768== by 0xCBB4F: ??? > ==21768== by 0x4005B5: main (debug-prog2.C:8) > > Here is also a minimal set of options necessary for the problem: > > valgrind --tool=callgrind ./a.out > > but one then has to interate > > callgrind_control -i off > callgrind_control -i on > callgrind_control -d > > till valgrind crashes. > > Greetings, > Christoph Bartoschek > > ------------------------------------------------------------------------- > Using Tomcat but need to do more? Need to support web services, security? > Get stuff done quickly with pre-integrated technology to make your job easier > Download IBM WebSphere Application Server v.1.0.1 based on Apache Geronimo > http://sel.as-us.falkag.net/sel?cmd=lnk&kid=120709&bid=263057&dat=121642 > _______________________________________________ > Valgrind-developers mailing list > Val...@li... > https://lists.sourceforge.net/lists/listinfo/valgrind-developers > > |
|
From: Josef W. <Jos...@gm...> - 2006-09-10 22:52:01
|
On Thursday 07 September 2006 20:45, Christoph Bartoschek wrote: > Here is another crash of the same problem: > > sleep > BB# 418 > > Callgrind: callstack.c:211 (vgCallgrind_push_call_stack): > Assertion 'current_entry->cxt != 0' failed. Hmmm... this one should be fixed now. However, I do not think that is the same problem as the first one you reported in this thread. There, the trace shows: valgrind: the 'impossible' happened: Killed by fatal signal ==00:00:00:28.625 2095== at 0x38014799: fprint_bbcc (global.h:691) ==00:00:00:28.625 2095== by 0x380158CD: print_bbccs_of_thread (dump.c:1550) ==00:00:00:28.625 2095== by 0x38015BCA: vgCallgrind_dump_profile (dump.c:1579) This is a SEGFAULT when dumping the data. I still have to try to reproduce it, as I currently do not have a amd64 machine around. Josef |