2012-02-28 19:46:07

by Thomas Gleixner

Subject: [patch 0/4] time/timekeeping: Simplify vsyscall locking

vsyscall updates have a separate seqlock which nests into
xtime_lock. There is no point to grab another lock for nothing. All it
needs is to update the seqcount which is user space visible.

The following series cleans that up.

Thanks,

tglx

2012-02-29 20:13:12

by Andrew Lutomirski

Subject: Re: [patch 0/4] time/timekeeping: Simplify vsyscall locking

On 02/28/2012 11:46 AM, Thomas Gleixner wrote:
> vsyscall updates have a separate seqlock which nests into
> xtime_lock. There is no point to grab another lock for nothing. All it
> needs is to update the seqcount which is user space visible.
>
> The following series cleans that up.

FWIW, patches 2 and 3 are
Reviewed-by: Andy Lutomirski <[email protected]>

Can you let me know when these hit a git tree somewhere? I have two
more vdso clock speedups I want to send for 3.4, and I'll rebase them on
top of these.

Thanks,
Andy