[pmuldoon@localhost bindir]$ gcc -g ~/segfault.c -o ~/segfault -lpthread
[pmuldoon@localhost bindir]$ fcatch ~/segfault
Print Entry Thread 1
Print Entry Thread 2
SIGSEGV detected: dumping stack trace
#0 0x804851f in main () from: /home/pmuldoon/segfault.c#36
#1 0x4da7ff2c in __libc_start_main ()
#2 0x8048411 in _start ()
fcatch does not exit, but seems to sit waiting hanging at the console.
System is Fedora Core 6
[pmuldoon@localhost frysk_bin]$ uname -a
Linux localhost.localdomain 2.6.19-1.2895.fc6 #1 SMP Wed Jan 10 19:28:18 EST
2007 i686 i686 i386 GNU/Linux
Created attachment 1524 [details]
The rather evil little segfaulter, added as an attachment.
2007-02-23 Mike Cvet <email@example.com>
* FCatch.java (Blocker): Added. TaskObserver.Instruction blocking
(signalObserver): Added. Add the same SignalObserver to each of the
(generateStackTrace): Decrement the numTasks counter.
(handleTaskBlock): Added. Switches on the signal value to appropriate
update the stack trace StringBuffer. If all the tasks of the process
have passed through, prints the StringBuffer and unblocks all the tasks.
(CatchObserver.updateAttached): Initialize the global SignalObserver if
it is null, and add to each task that passes through.
(CatchObserver.updateTerminating): Removed eventloop stopping code.
(CatchObserver.updateTerminated): Requests a stop from the eventloop if
all the Tasks have completed stack tracing and are unblocked and
terminated. Fixes #3932.
(SignalObserver.updateSignaled): Assign some global variables from the
signaled Task. Moved switch statement and stack trace calls to
handleTaskBlock(). Block each of the threads of the process with an
InstructionObserver. Fixes #4092.