Re: LEAF vkernel support update 13-May-2008

[Aggelos Economopoulos <aoiko@xxxxxxxxxxxxxx>]

On Wednesday 14 May 2008, Matthew Dillon wrote:
[...]
>     In the mean time, if you want to mess around with the vkernel
>     support there are two scripts in /usr/local/bin on leaf that will
>     get you started, called 'initvkernel' and 'runvkernel'.

Great, used it to reproduce a bug I'm hitting on my box. Just run initvkernel,
and then (pasting from my scrollback buffer)

leaf:/home/aggelos> gdb ~/build/vkernel/kernel
GNU gdb 6.7.1
Copyright (C) 2007 Free Software Foundation, Inc.
License GPLv3+: GNU GPL version 3 or later <http://gnu.org/licenses/gpl.html>
This is free software: you are free to change and redistribute it.
There is NO WARRANTY, to the extent permitted by law.  Type "show copying"
and "show warranty" for details.
This GDB was configured as "i386-dragonfly"...
(gdb) b sbflush
Breakpoint 1 at 0x80e5723: file /usr/src/sys/kern/uipc_sockbuf.c, line 442.
(gdb) handle SIGSEGV noprint
Signal        Stop      Print   Pass to program Description
SIGSEGV       No        No      Yes             Segmentation fault
(gdb) handle SIGUSR1 noprint
Signal        Stop      Print   Pass to program Description
SIGUSR1       No        No      Yes             User defined signal 1
(gdb) run  -m 64m -i ~/build/vkernel/memimg.01 -r ~/build/vkernel/rootimg.01 -r ~/build/vkernel/swapimg.01
Starting program: /build/home/aggelos/vkernel/kernel -m 64m -i ~/build/vkernel/memimg.01 -r ~/build/vkernel/rootimg.01 -r ~/build/vkernel/swapimg.01
Using memory file: /home/aggelos/build/vkernel/memimg.01
KVM mapped at 0x40000000-0x80000000
Copyright (c) 2003-2008 The DragonFly Project.
Copyright (c) 1992-2003 The FreeBSD Project.
Copyright (c) 1979, 1980, 1983, 1986, 1988, 1989, 1991, 1992, 1993, 1994
        The Regents of the University of California. All rights reserved.
DragonFly 1.13.0-DEVELOPMENT #0: Tue May 13 15:30:51 PDT 2008
    root@leaf.dragonflybsd.org:/usr/obj/usr/src/sys/VKLEAF
real memory = 67108864 (65536K bytes)
avail memory = 62414848 (60952K bytes)
DragonFly/MP: Multiprocessor
 cpu0 (BSP)
 cpu1 (AP)
initclocks
SMP: AP CPU #1 Launched!
objcache(exec-args): too small for ncpus, adjusting cluster_limit 16->32
md0: Malloc disk

The vkernel gets stuck here. Really stuck.

leaf:/home/aggelos> ps auxHlww | grep kernel
aggelos 74113     0  0.0  2.3 24480 23464  p4  IL    5:19AM   0:00.36 gdb /home/aggelo  2042 74113 73958   0 152   0 24480 23464 wait   IL    p4    0:00.36 gdb /home/aggelos/build/vkernel/kernel
aggelos 74216     3  0.0  0.4 1052872 4480  p4  TXL+  5:20AM   0:00.00 /build/home/agge  2042 74216 74113   4 154   0 1052872 4480 umtxsl TXL+  p4    0:00.00 /build/home/aggelos/vkernel/kernel -m 64m -i /home/aggelos/build/vkernel/memimg.01 -r /home/aggelos/build/vkernel/rootimg.01 -r /home/aggelos/build/vkernel/swapimg.01
aggelos 74216     2  0.0  0.4 1052872 4480  p4  TXL+  5:20AM   0:00.00 /build/home/agge  2042 74216 74113   2 153   0 1052872 4480 umtxsl TXL+  p4    0:00.00 /build/home/aggelos/vkernel/kernel -m 64m -i /home/aggelos/build/vkernel/memimg.01 -r /home/aggelos/build/vkernel/rootimg.01 -r /home/aggelos/build/vkernel/swapimg.01
aggelos 74216     1  0.0  0.4 1052872 4480  p4  TXL+  5:20AM   0:00.01 /build/home/agge  2042 74216 74113   0 152   0 1052872 4480 umtxsl TXL+  p4    0:00.01 /build/home/aggelos/vkernel/kernel -m 64m -i /home/aggelos/build/vkernel/memimg.01 -r /home/aggelos/build/vkernel/rootimg.01 -r /home/aggelos/build/vkernel/swapimg.01
aggelos 74216     0  0.0  0.4 1052872 4480  p4  TXL+  5:20AM   0:00.25 /build/home/agge  2042 74216 74113   0 152   0 1052872 4480 stop   TXL+  p4    0:00.25 /build/home/aggelos/vkernel/kernel -m 64m -i /home/aggelos/build/vkernel/memimg.01 -r /home/aggelos/build/vkernel/rootimg.01 -r /home/aggelos/build/vkernel/swapimg.01
leaf:/home/aggelos> kill 74216
leaf:/home/aggelos> kill -9 74216
leaf:/home/aggelos> kill -9 74216

(notice, no ESRCH)

Is this a known bug?

TIA,
Aggelos