Re: kern/52598 (/proc/self/exe misbehaves if it's on a null mount)

[Christos Zoulas <christos%zoulas.com@localhost>]

there are many more changes needed I think in the exec code. can you reproduce the crash on head or 9?

christos



> On Jan 13, 2020, at 5:35 PM, Frédéric Fauberteau <triaxx%netbsd.org@localhost> wrote:
> 
> The following reply was made to PR kern/52598; it has been noted by GNATS.
> 
> From: =?UTF-8?Q?Fr=C3=A9d=C3=A9ric_Fauberteau?= <triaxx%NetBSD.org@localhost>
> To: gnats-bugs%NetBSD.org@localhost
> Cc: Pierre Pronchery <khorben%defora.org@localhost>, maya%NetBSD.org@localhost,
> christos%NetBSD.org@localhost
> Subject: Re: kern/52598 (/proc/self/exe misbehaves if it's on a null mount)
> Date: Mon, 13 Jan 2020 23:23:27 +0100
> 
> I tried to apply the changes made by christos to introduce p_path in 
> struct proc to netbsd-8-0-RELEASE:
> cvs rdiff -u -r1.197.2.2 -r1.200 src/sys/miscfs/procfs/procfs_vnops.c
> cvs rdiff -u -r1.340.6.1 -r1.343 src/sys/sys/proc.h
> 
> Now I get the following behavior:
> # ls -l /proc/curproc
> lr-xr-xr-x  1 root  wheel  2 Jan 13 23:09 /proc/curproc -> 49
> # ls -l /proc/curproc/
> causes a fatal page fault.
> 
> The backtrace I got using crash(1):
> # crash -M netbsd.9.core
> Crash version 8.0, image version 8.0.
> System panicked: trap
> Backtrace from time of crash is available.
> crash> bt
> _KERNEL_OPT_NARCNET() at 0
> ?() at ffff800064978000
> vpanic() at vpanic+0x166
> snprintf() at snprintf
> trap() at trap+0xa00
> --- trap (number 6) ---
> strlen() at strlen+0x20
> VOP_GETATTR() at VOP_GETATTR+0x4d
> vn_stat() at vn_stat+0x3d
> do_sys_statat() at do_sys_statat+0x89
> sys___lstat50() at sys___lstat50+0x25
> syscall() at syscall+0x1ec
> --- syscall (number 441) ---
> 7f5df9ca17ba:
>