Messages in this thread | | | Date | Thu, 02 Oct 2014 17:22:41 +0300 | From | Tanya Brokhman <> | Subject | Re: [PATCH 4/4] UBI: Fastmap: Ensure that only one fastmap work is scheduled |
| |
On 9/30/2014 10:44 AM, Richard Weinberger wrote: > Am 30.09.2014 09:39, schrieb Bityutskiy, Artem: >> On Tue, 2014-09-30 at 08:59 +0200, Richard Weinberger wrote: >>> Am 30.09.2014 08:45, schrieb Bityutskiy, Artem: >>>> On Tue, 2014-09-30 at 00:20 +0200, Richard Weinberger wrote: >>>>> + spin_lock(&ubi->wl_lock); >>>>> + ubi->fm_work_scheduled = 0; >>>>> + spin_unlock(&ubi->wl_lock); >>>> >>>> Andrew Morton once said me that if I am protecting an integer change >>>> like this with a spinlock, I have a problem in my locking design. He was >>>> right for my particular case. >>>> >>>> Integer is changes atomic. The only other thing spinlock adds are the >>>> barriers. >>> >>> I've added the spinlock to have a barrier in any case. >> >> Examples of any? > > You mean a case where the compiler would reorder code and the barrier is needed? > I don't have one, but I'm not that creative as a modern C compiler. > If you say that no barrier is needed I'll trust you. :-)
we just implemented the same thing :) It's being tested.... Why not use atomic_t fm_work_scheduled and save the spin_lock?
> > Thanks, > //richard > > > ______________________________________________________ > Linux MTD discussion mailing list > http://lists.infradead.org/mailman/listinfo/linux-mtd/ >
-- Employee of Qualcomm Innovation Center, Inc. Qualcomm Innovation Center, Inc. is a member of Code Aurora Forum, hosted by The Linux Foundation
| |