panic: lwp_addref() in procfs_rw()

[Thomas Klausner <tk%giga.or.at@localhost>]

Hi!

I just had a kernel panic.
syslog wrote the backtrace:

Feb 10 13:39:05 yt /netbsd: lwp_addref() at netbsd:lwp_addref+0xbb
Feb 10 13:39:05 yt /netbsd: procfs_rw() at netbsd:procfs_rw+0x13e
Feb 10 13:39:05 yt /netbsd: VOP_READ() at netbsd:VOP_READ+0x47
Feb 10 13:39:05 yt /netbsd: vn_read() at netbsd:vn_read+0x96
Feb 10 13:39:05 yt /netbsd: dofileread() at netbsd:dofileread+0x79
Feb 10 13:39:05 yt /netbsd: sys_read() at netbsd:sys_read+0x46
Feb 10 13:39:05 yt /netbsd: syscall() at netbsd:syscall+0xb3
Feb 10 13:39:05 yt /netbsd: --- syscall (number 3) ---
Feb 10 13:39:05 yt /netbsd: 7f7ff683b3da:
Feb 10 13:39:05 yt /netbsd: cpu4: End traceback...

I have two procfs mounted:
procfs on /proc type procfs (local)
procfs on /usr/pkg/emul/linux32/proc type procfs (local)

My guess is that a perl program using p5-Proc-ProcessTable accesses
triggered it, but I don't know what it was doing exactly.

Kernel is 6.99.31/amd64 from early on Feb 8.
 Thomas