nmbd dies very often with a core dump (PR#1477)

[Andrew Tridgell]

> We're using samba release 1.9.17p4 on a Sparc 10, SunOS 4.1.3. 
> Only this samba version causes the nmbd to die very often with a
> core dump. Here is the output of the debugger:
> 
>>Core was generated by `nmbd'.
>>Program terminated with signal 11, Segmentation fault.
>>Reading symbols from /usr/lib/libc.so.1.8...done.
>>Reading symbols from /usr/lib/libdl.so.1.0...done.
>>#0  0x4aa0 in putip ()
>>(gdb) where
>>#0  0x4aa0 in putip ()
>>#1  0x9018 in interpret_addr ()
>>#2  0x1ce3c in dns_name_search ()
>>#3  0x1fce8 in reply_name_query ()
>>#4  0x16308 in process_nmb ()
>>#5  0x163ac in run_packet_queue ()
>>#6  0x2be4 in process ()
>>#7  0x3d9c in main ()

This sounds like a bug I have seen before. At the time we concluded that
SunOS 4.1.3 has a broken gethostbyname() call that can't handle it when a 
DNS entry has a large number of IP addresses and may set the h_addr field
incorrecty.

Does anyone know if this is a known SunOS bug? Is there a fix? I'd like to
include it in the FAQ. Or perhaps there is a workaround?

Cheers, Andrew

PS: A likely quick fix is to use "dns proxy = no" in smb.conf