freebsd stable - Sep 2014 - 10.1 mmap MAP_NOSYNC regression due to fast path

Hi, just wondering if the regression introduced by the changes for
'fast path' to the mmap MAP_NOSYNC behaviour will be fixed for 10.1 ?

http://lists.freebsd.org/pipermail/freebsd-hackers/2014-September/045981.html

We have been doing some testing with 10.1-BETA3 and the regression makes
10.1 completely unusable because we rely very heavily on this long standing
feature of mmap.  Our database design mmaps large files (eg. several gigabytes)
and we manually control when the sync happens by using fsync() at appropriate
times.

	Paul.

-- 
Paul Koch | Founder, CEO
AKIPS Network Monitor
http://www.akips.com
Brisbane, Australia

On Mon, Sep 29, 2014 at 8:54 PM, Paul Koch <paul.koch at akips.com> wrote:
> Hi, just wondering if the regression introduced by the changes for
> 'fast path' to the mmap MAP_NOSYNC behaviour will be fixed for 10.1
?
>
>
Yes, it will be.  I've just been overloaded with work for the past few
weeks, so kib@ and I haven't agreed on the final fix.


>
>
http://lists.freebsd.org/pipermail/freebsd-hackers/2014-September/045981.html
>
> We have been doing some testing with 10.1-BETA3 and the regression makes
> 10.1 completely unusable because we rely very heavily on this long standing
> feature of mmap.  Our database design mmaps large files (eg. several
> gigabytes)
> and we manually control when the sync happens by using fsync() at
> appropriate
> times.
>
>         Paul.
>
> --
> Paul Koch | Founder, CEO
> AKIPS Network Monitor
> http://www.akips.com
> Brisbane, Australia
> _______________________________________________
> freebsd-stable at freebsd.org mailing list
> http://lists.freebsd.org/mailman/listinfo/freebsd-stable
> To unsubscribe, send any mail to "freebsd-stable-unsubscribe at
freebsd.org"
>