> (gdb) p/x *kmsg
> $1 = {ikm_next = 0x4081700, ikm_prev = 0xffffff10, ikm_size = 0x100,
> ikm_marequest = 0x0, ikm_header = {msgh_bits = 0x80001200, msgh_size = 0xa0,
> msgh_remote_port = 0x0, msgh_local_port = 0x4, msgh_seqno = 0x5,
> msgh_id = 0x7788}}
If this is right (and it looks plausible), then this is an unrelated
message. That is message ID 30600 (0x7788), which is the reply message for
an exec_startup_get_info RPC. So this suggests that maybe some VM data
structures were corrupted by something earlier that made this break without
anything about this message being directly involved at all.
Does your program use device_write or device_write_inband? Try making both
it and pfinet use only device_write_inband. Using device_write involves
the generic wrapper code in oskit/ds_routines.c that uses vm_map_copyout.
I suppose that could somehow be wrong and affect it. Using only
device_write_inband avoids that code, and so if your problem disappears
we'll know something.
I will look at your output in more detail later, but I doubt I will have
much insight off the bat. You may have to slowly track down what pages got
used for what and how things got the way they are.
_______________________________________________
Bug-hurd mailing list
[EMAIL PROTECTED]
http://mail.gnu.org/mailman/listinfo/bug-hurd
