From: Jeff Dike <jdike@ka...> - 2002-03-08 19:49:59
> Before I spend to much time on this I would like to check if someone
> else is working on this or has done a newer patch.
Last night, there were a couple people on #kernelnewbies who expressed
interest in getting that up to date.
No progress as far as I know.
If you want to compare notes with them, go there (on opn), and look for
uberfunk and wiild.
From: Jeff Dike <jdike@ka...> - 2002-03-09 01:50:29
This is the right stack:
> #0 0xa00d5261 in nanosleep () at af_packet.c:1882
> #1 0xa0067898 in idle_sleep (secs=3D10) at time.c:108
> #2 0xa006aebb in cpu_idle () at process_kern.c:463
> #3 0xa000b482 in rest_init () at init/main.c:333
> #4 0xa0003547 in start_kernel () at init/main.c:430
> #5 0xa00694e7 in start_kernel_proc (unused=3D0x0) at um_arch.c:130
> #6 0xa00688db in signal_tramp (arg=3D0xa00694a8) at trap_user.c:83
This one is garbage:
> #0 packet_ioctl (sock=3D0xa, cmd=3D0, arg=3D2685648736) at af_packet.c=
> #1 0xa0062048 in sys_shmat (shmid=3D10, shmaddr=3D0x0, shmflg=3D-16093=
> raddr=3D0xa00655d1) at shm.c:640
> #2 0xa00655db in do_check_pgt_cache (low=3D-1609570976, high=3D0) at m=
> #3 0xa000a482 in devinet_init () at devinet.c:1159
> #4 0xa000350c in start_kernel () at init/main.c:399
> #5 0xa0063c97 in outer_tramp (arg=3D0x0) at process.c:103
> #6 0xa006308b in vsscanf (buf=3D0xa0063c58 "U\211=E5\203=EC\030=E8-=CF=
> fmt=3D0x0, args=3D0x0) at vsprintf.c:594
Mike, didn't you say you had chased down an init ordering problem that ca=
a hang around here?
> I guess I'm going to search through the diff and see what could have
> touched the pieces that UML is caling in the idle and sleep stuff.
> The nanosleep in af_packet.c is still confusing me too. =
The thing to do when this sort of thing happens is to figure out what sho=
be running, which shouldn't be hard at this stage of the boot, and figure=
why it's sleeping and not running. A stack trace of that process is usua=
what you want.
It went to sleep for some reason, and you need to figure out what is supp=
to have woken it up and why that didn't happen.