Well it happened again along the same lines..., showing
'libfuse vnode reclaim failed'
And again uninterruptable, unkillable, unrebootable system, only
power button helped.
So isn't there really any way to kill such a beast?
(
I use 6.3. The line I use to start sshfs is basically just
doas sshfs $MACHINE $MOUNTPOINT -o uid=1000 -o gid=1000
)
On 20 April 2018 at 02:34, IL Ka <kazakevichi...@gmail.com> wrote:
> First, make sure your SSH connection is stable.
I believe I can say it is stable.
(From the other-systems experience.)
But anyway, even if it were unstable, I thought the system
should never enter a state where it has to be unplugged...
> If it works, then we can try
> to investigate it:
>
>
>> libfuse vnode reclaim failed
>
> This message is from following code:
> if (fb_queue(fmp->dev, fbuf))
> printf("fusefs: libfuse vnode reclaim failed\n");
>
> "fuse_lookup.c"
>
> As we can see, "fb_queue" retruns error, but we do not know
> which error.
>
> I think we need to know this error to understand what happened.
>
> In "fb_queue" source (fusebuf.c) we can see that error is
> return code from "tsleep" or "fbuf->fb_err".
>
>
> This code is part of this commit:
> https://github.com/openbsd/src/commit/5e69e0b69176a7f878c899ef5828a6273ab7f0bb
> Written by https://github.com/natano , while file originally written by
> Sylvestre Gallon <ccna....@gmail.com>
>
> So, you have 2 options here:
>
> Debug it yourself by kernel debugging (gdb)
> or patch kernel and print this error value to kernel log
> (printf(9)) (could be good contribution to kernel btw)
>
> Or you can write to tech@ or bugs@
> about this problem and wait for developer's attention.
So I should probably write to bugs@...
Thanks
Ruda
