On Thursday 25 October 2007 14:11, Andrew Morton wrote:
quoted text
> On Wed, 24 Oct 2007 08:24:57 -0400 Matthew Wilcox <matthew@wil.cx> wrote:
> > and associated infrastructure such as sync_page_killable and
> > fatal_signal_pending.  Use lock_page_killable in
> > do_generic_mapping_read() to allow us to kill `cat' of a file on an
> > NFS-mounted filesystem.
>
> whoa, big change.
>
> What exactly are the semantics here?  If the process has actually been
> killed (ie: we know that userspace won't be running again) then we break
> out of a lock_page() and allow the process to exit?  ie: it's basically
> invisible to userspace?

The actual conversions should also be relatively useful groundwork
if we ever want to make more things become generally interruptible.


quoted text
> If so, it sounds OK.  I guess.  We're still screwed if the process is doing
> a synchronous write and lots of other scenarios.

I don't think it will matter in too many situations. If the process is
doing a synchronous write, nothing is guaranteed until the syscall
returns success...
-
unsubscribe notice
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/