Attaching to a Running Process
Unfortunately, some UNIXes (in particular some recent Linux kernels)
refuse to dump a core file if the task has changed UID (which smbd
does often). To debug with this sort of system, you could try to attach
to the running process using
, where you get
to continue and try to cause the core dump
using the client. The debugger should catch the fault and tell you
where it occurred.
Sometimes it is necessary to build Samba binary files that have debugging
symbols so as to make it possible to capture enough information from a crashed
operation to permit the Samba Team to fix the problem.
-g to ensure you have symbols in place.
Add the following line to the
smb.conf file global section:
panic action = "/bin/sleep 90000"
to catch any panics. If
seems to be frozen, look for any sleep
processes. If it is not, and appears to be spinning, find the PID
of the spinning process and type:
root# gdb /usr/local/samba/sbin/smbd
then “attach `pid'” (of the spinning process), then type “bt” to
get a backtrace to see where the smbd is in the call path.