Philip,
could you do the following ? (it will be very helpful).
1) Put this in your .gdbinit in your home directory
------->8-------------
define mono_backtrace
select-frame 0
set $i = 0
while ($i < $arg0)
set $foo = mono_pmip ($pc)
if ($foo == 0x00)
frame
else
printf "#%d %p in %s\n", $i, $pc, $foo
end
up-silently
set $i = $i + 1
end
end
------------------>8--------------
2) start beagled from the terminal as
beagled --fg --indexing-delay 300
then from another terminal,
3) find out the pid of beagle, $ ps ux | grep beagled
4) start gdb $ gdb --pid=<pid of beagled>
5) once gdb starts, enter at the gdb prompt
handle SIGXCPU SIG33 SIG35 SIGPWR nostop noprint
6) Enter at the gdb prompt
c or continue
7) when beagled crashes, enter this at gdb prompt
mono_backtrace 10
8) Copy the output of mono_backtrace to a file and attach it.
9) At gdb prompt, type q or quit
10) If it asks to detach, say yes.
11) Kill beagled if it is still running
Its a long process but not tedious :-). You can almost reliably reproduce the crash, so I am seeking your help. Thanks.
Philip, >8----- ------- - ------- ---->8- ------- ------
could you do the following ? (it will be very helpful).
1) Put this in your .gdbinit in your home directory
-------
define mono_backtrace
select-frame 0
set $i = 0
while ($i < $arg0)
set $foo = mono_pmip ($pc)
if ($foo == 0x00)
frame
else
printf "#%d %p in %s\n", $i, $pc, $foo
end
up-silently
set $i = $i + 1
end
end
-------
2) start beagled from the terminal as
beagled --fg --indexing-delay 300
then from another terminal,
3) find out the pid of beagle, $ ps ux | grep beagled
4) start gdb $ gdb --pid=<pid of beagled>
5) once gdb starts, enter at the gdb prompt
handle SIGXCPU SIG33 SIG35 SIGPWR nostop noprint
6) Enter at the gdb prompt
c or continue
7) when beagled crashes, enter this at gdb prompt
mono_backtrace 10
8) Copy the output of mono_backtrace to a file and attach it.
9) At gdb prompt, type q or quit
10) If it asks to detach, say yes.
11) Kill beagled if it is still running
Its a long process but not tedious :-). You can almost reliably reproduce the crash, so I am seeking your help. Thanks.