Re: umount -f busted

2000-11-08 Thread Michael C . Wu
On Wed, Nov 08, 2000 at 07:43:24AM -0700, Warner Losh scribbled: | In message <[EMAIL PROTECTED]> Johan Karlsson writes: | : At Tue, 07 Nov 2000 14:54:50 MST, Warner Losh wrote: | : > In message <[EMAIL PROTECTED]> Alfred Perlstein writes: | : > : Yes, this used to work quite well for some time, I

Re: umount -f busted

2000-11-08 Thread Bill Fumerola
On Wed, Nov 08, 2000 at 07:43:24AM -0700, Warner Losh wrote: > It is a problem that I could have sworn worked before SMPNG. Negative, this occurs on releng_4 machines for me as well. It also was occuring on my -current workstation that was about 110 days old before a disk went out, so it was def

Re: umount -f busted

2000-11-08 Thread Warner Losh
In message <[EMAIL PROTECTED]> Johan Karlsson writes: : At Tue, 07 Nov 2000 14:54:50 MST, Warner Losh wrote: : > In message <[EMAIL PROTECTED]> Alfred Perlstein writes: : > : Yes, this used to work quite well for some time, I have no idea : > : who broke it. Maybe you can sprinkle some printfs in

Re: umount -f busted

2000-11-08 Thread Johan Karlsson
At Tue, 07 Nov 2000 14:54:50 MST, Warner Losh wrote: > In message <[EMAIL PROTECTED]> Alfred Perlstein writes: > : Yes, this used to work quite well for some time, I have no idea > : who broke it. Maybe you can sprinkle some printfs in the code and > : narrow it down a bit? > > I'll give it a sh

Re: umount -f busted

2000-11-07 Thread Warner Losh
In message <[EMAIL PROTECTED]> Alfred Perlstein writes: : Yes, this used to work quite well for some time, I have no idea : who broke it. Maybe you can sprinkle some printfs in the code and : narrow it down a bit? I'll give it a shot. I'm glad to see it is a "should work but is busted" rather t

Re: umount -f busted

2000-11-07 Thread Alfred Perlstein
* Warner Losh <[EMAIL PROTECTED]> [001107 13:14] wrote: > In message <[EMAIL PROTECTED]> Bill Fumerola writes: > : On Tue, Nov 07, 2000 at 01:13:41PM -0700, Warner Losh wrote: > : > : > I just tried to umount -f /home, where /home was an NFS mounted file > : > system on a network that was no long

Re: umount -f busted

2000-11-07 Thread Warner Losh
In message <[EMAIL PROTECTED]> Bill Fumerola writes: : On Tue, Nov 07, 2000 at 01:13:41PM -0700, Warner Losh wrote: : : > I just tried to umount -f /home, where /home was an NFS mounted file : > system on a network that was no longer attached to my laptop. In the : > past this has just worked, e

Re: umount -f busted

2000-11-07 Thread Bill Fumerola
On Tue, Nov 07, 2000 at 01:13:41PM -0700, Warner Losh wrote: > I just tried to umount -f /home, where /home was an NFS mounted file > system on a network that was no longer attached to my laptop. In the > past this has just worked, even if processes were hung in disk wait > state. When I tried

Re: umount -f

1999-02-11 Thread Vallo Kallaste
On Thu, Feb 11, 1999 at 08:24:52AM -0500, Mikhail Teterin wrote: > =Do you have "hard" mount or "soft" mount? I have seen such behavior > =for "hard" mounts. > > Soft. But it should not matter, should it? I don't know, I'm not a expert to say for sure. Only know that about two months ago I tr

Re: umount -f

1999-02-11 Thread Mikhail Teterin
Vallo Kallaste once stated: =On Wed, Feb 10, 1999 at 01:54:53PM -0500, Mikhail Teterin wrote: = => Nope: => => m...@xxx:/tmp (1044) umount -f -t nfs phosphorus:/phosphorus => umount: /phosphorus: Device busy => => It is not, that umount hangs, it is that it cares about the device bei

Re: umount -f

1999-02-11 Thread Vallo Kallaste
On Wed, Feb 10, 1999 at 01:54:53PM -0500, Mikhail Teterin wrote: > Nope: > > m...@xxx:/tmp (1044) umount -f -t nfs phosphorus:/phosphorus > umount: /phosphorus: Device busy > > It is not, that umount hangs, it is that it cares about the device being > busy despite `-f' flag. Or so

Re: umount -f

1999-02-11 Thread Bruce Evans
>> > Will it ever work as it appears it should? Currently I have (on 2.2.8) >> >From an email from Peter Wemm: >> >> In this situation, you need to do this: >> umount -f -t nfs phosphorus:/phosphorus >> >> This causes umount to stat("phosphorus:/phosphorus") (which fails) rather >> than

Re: umount -f

1999-02-10 Thread Mikhail Teterin
> Just to ask, have you run lsof on /phosphorus to see if it is, > indeed, busy? lsof is unable to stat /phosphorus, of course. But, in any case, this should not be relevant, because the `-f' is specified... -mi To Unsubscribe: send mail to majord...@freebsd.org with "unsubscribe freebsd

Re: umount -f

1999-02-10 Thread Mikhail Teterin
> > Will it ever work as it appears it should? Currently I have (on 2.2.8) > > > > m...@xxx:/tmp (1032) umount -f phosphorus:/phosphorus > > umount: /phosphorus: Device busy > > >From an email from Peter Wemm: > > In this situation, you need to do this: > umount -f -t nfs pho

Re: umount -f

1999-02-10 Thread David O'Brien
> Will it ever work as it appears it should? Currently I have (on 2.2.8) > > m...@xxx:/tmp (1032) umount -f phosphorus:/phosphorus > umount: /phosphorus: Device busy >From an email from Peter Wemm: In this situation, you need to do this: umount -f -t nfs phosphorus:/pho