deadlock bugs in tru54 51B ???

[jini_2]

here is the stack info:

Stack trace for thread 1
#0 0x3ff805c192c in __hstTransferRegisters(...) in /usr/shlib/libpthread.so
#1 0x3ff805b10d4 in __osTransferContext(...) in /usr/shlib/libpthread.so
#2 0x3ff805a2b20 in __dspDispatch(...) in /usr/shlib/libpthread.so
#3 0x3ff805aadd8 in UnknownProcedure153FromFile0(...) in /usr/shlib/libpthread.so
#4 0x3ff805ab578 in UnknownProcedure155FromFile0(...) in /usr/shlib/libpthread.so
#5 0x3ff802070b8 in UnknownProcedure3FromFile22(...) in /usr/shlib/libc.so
#6 0x3ff800d2814 in malloc(...) in /usr/shlib/libc.so
#7 0x1200016e8 in t_sig(signo=14) "t_pthread.c":9
#8 0x3ff800d5b90 in __sigtramp(...) in /usr/shlib/libc.so
#9 0x3ff80205e80 in UnknownProcedure13FromFile22(...) in /usr/shlib/libc.so
#10 0x1200019c0 in main() "t_pthread.c":101 --> call malloc()
#11 0x120001668 in __start(...) in t_p6

Stack trace for thread 2
#0 0x3ff805c192c in __hstTransferRegisters(...) in /usr/shlib/libpthread.so
#1 0x3ff805b10d4 in __osTransferContext(...) in /usr/shlib/libpthread.so
#2 0x3ff805a2b20 in __dspDispatch(...) in /usr/shlib/libpthread.so
#3 0x3ff805aadd8 in UnknownProcedure153FromFile0(...) in /usr/shlib/libpthread.so
#4 0x3ff805ab578 in UnknownProcedure155FromFile0(...) in /usr/shlib/libpthread.so
#5 0x3ff802070b8 in UnknownProcedure3FromFile22(...) in /usr/shlib/libc.so
#6 0x3ff800d2814 in malloc(...) in /usr/shlib/libc.so
#7 0x120001750 in test_func1(arg=0x0) "t_pthread.c":26
#8 0x3ff805cedc8 in __thdBase(...) in /usr/shlib/libpthread.so

it seems that the malloc() function is re-entering, and it uses a non-recursive mutex internals.

Can someone tell me how to resolved it? thanks !

[Michael Schulte zur Sur]

Hi,

what patch kit are you running?

greetings,

Michael

[jini_2]

hi:

I'm using Tru64 v5.1B 2650 alpha with T64V51BB1AS0001-20021229 OSF540 patch kit installed.

Has this problem fixed by new patchs?

[BryceX]

Greetings,

I have also experienced the same malloc hang in a threaded application. 5.1B PK4 advertises to fix this bug (patch 25003.00, 25010.00). I have verified it is fixed in 5.1B-3.

I would like to know when this bug was introduced.

Thanks in advance.

Bryce

[Derrick Tate]

Subject: deadlock bugs in tru54 51B ???

The problem I'm having is a SEGV in a application that I think is some sort of deadlock. My stack trace from ladebug looks like the one below:


#5 0x3ff802070b8 in UnknownProcedure3FromFile22(...) in /usr/shlib/libc.so
#6 0x3ff800d2814 in malloc(...) in /usr/shlib/libc.so
#7 0x120001750 in test_func1(arg=0x0) "t_pthread.c":26
#8 0x3ff805cedc8 in __thdBase(...) in /usr/shlib/libpthread.so


The line in focus is: UnknownProcedure3FromFile22()


One of the following patches is suppose to fix the bug, but it's not spelled out for me.

patch 25003.00, 25010.00

So, the question is... How do I validate installing this patch will fix this bug?

Thanks in advance,

..X..