>
> Did you have to change VFS_SYNC() to pass the forced-dismount flag? I
> don't like that, but see the problem: VFS_SYNC(XXX_WAIT) should
> normally
> just try to sync everything, but return an error if this is
> impossible,
> but for an nfs server it would take too long to determine that it is
On Tue, 31 May 2011, Rick Macklem wrote:
It is only an optimization. Any number of syncs are useless for
actually syncing the system, since sync(2) only does an async sync (it
returns without waiting for most writes to complete). As you pointed
out later in this thread, unmount(2) does a sync th
>
> It is only an optimization. Any number of syncs are useless for
> actually syncing the system, since sync(2) only does an async sync (it
> returns without waiting for most writes to complete). As you pointed
> out later in this thread, unmount(2) does a sync that works -- a sync
> sync -- befo
On Mon, 30 May 2011, Rick Macklem wrote:
On Mon, May 30, 2011 at 01:48:53PM +0100, Robert Watson wrote:
On Sun, 29 May 2011, Rick Macklem wrote:
Modify the umount(8) command so that it doesn't do
a sync(2) syscall before unmount(2) for the "-f" case.
This avoids a forced dismount from getting
>
> If it masks, for example, lateny for a synchronous RPC to the remote
> mountd to
> deregister the mountpoint, allowing a cache flush and unmount to take
> place
> concurrently, that might be a useful benefit. I'm not sure I see any
> evidence
> that is the case in the source code, however.
>
>
> No doubt. :-)
>
> If the sync(2) has actual consistency and reliability benefits, it
> should
> probably be done by the umount(2) system call, so that other future
> auto-mounters, etc, also get the same result, rather than having to
> encode it
> in every application. If it's done on blind f
On Mon, 30 May 2011, Rick Macklem wrote:
Hehe. I'm so old, I do two syncs, as required by 6th Edition.:-)
I assumed the sync() was meant to be an optimization (given the comment for
it) in the sense that it would get the writes of dirty blocks started "right
away". However, given the short p
On Mon, 30 May 2011, Kostik Belousov wrote:
On Mon, May 30, 2011 at 01:48:53PM +0100, Robert Watson wrote:
On Sun, 29 May 2011, Rick Macklem wrote:
Modify the umount(8) command so that it doesn't do a sync(2) syscall
before unmount(2) for the "-f" case. This avoids a forced dismount from
get
> On Mon, May 30, 2011 at 01:48:53PM +0100, Robert Watson wrote:
> > On Sun, 29 May 2011, Rick Macklem wrote:
> >
> > > Modify the umount(8) command so that it doesn't do
> > > a sync(2) syscall before unmount(2) for the "-f" case.
> > > This avoids a forced dismount from getting stuck for
> > > an
On Mon, May 30, 2011 at 01:48:53PM +0100, Robert Watson wrote:
> On Sun, 29 May 2011, Rick Macklem wrote:
>
> > Modify the umount(8) command so that it doesn't do
> > a sync(2) syscall before unmount(2) for the "-f" case.
> > This avoids a forced dismount from getting stuck for
> > an NFS mountpoi
On Sun, 29 May 2011, Rick Macklem wrote:
Modify the umount(8) command so that it doesn't do
a sync(2) syscall before unmount(2) for the "-f" case.
This avoids a forced dismount from getting stuck for
an NFS mountpoint in sync() when the server is not
responsive. With this commit, forced dis
Author: rmacklem
Date: Sun May 29 21:13:53 2011
New Revision: 222466
URL: http://svn.freebsd.org/changeset/base/222466
Log:
Modify the umount(8) command so that it doesn't do
a sync(2) syscall before unmount(2) for the "-f" case.
This avoids a forced dismount from getting stuck for
an NFS
12 matches
Mail list logo
