Messages in this thread | | | Date | Sat, 22 Jan 2011 11:36:28 +0100 (CET) | From | Thomas Gleixner <> | Subject | Re: [PATCH 02/18] provide xtime_update() which does not require holding xtime_lock like do_timer() |
| |
On Sat, 22 Jan 2011, Torben Hohn wrote:
> some arch code failed to lock the xtime_lock. > and some code looks like its using the xtime_lock to protect > other stuff.
That's not a good argument for creating xtime_update().
The point is that do_timer() needs to write lock xtime_lock and we want to avoid the duplicated code all over the place.
The fact that some of the architectures have other code in the xtime_lock protected region is completely irrelevant for this change. That needs to be addressed by the arch specific patches.
> + > +/* xtime_update - updates the timer infrastructure.
Please use proper kernel doc style. Also it does not update the timer infrastructure, it's the timekeeping update.
The comment should also document, that this code needs to be called with interrupts disabled.
Thanks,
tglx
| |