Re: [PATCH 0/2] jump label: 2.6.38 updates

[Mike Frysinger]

On Mon, Feb 14, 2011 at 12:38, Peter Zijlstra wrote:
> On Mon, 2011-02-14 at 12:29 -0500, Mike Frysinger wrote:
>> On Mon, Feb 14, 2011 at 12:27, Peter Zijlstra wrote:
>> > On Mon, 2011-02-14 at 12:18 -0500, Steven Rostedt wrote:
>> >> blackfin, seems to be doing quite a lot. Not sure if it is required, but
>> >> that may need a bit of investigating to understand why it does the
>> >> raw_uncached thing.
>> >
>> > From what I can tell its flushing its write cache, invalidating its
>> > d-cache and then issue the read, something which is _way_ overboard.
>>
>> not when the cores in a SMP system lack cache coherency
>
> But atomic_read() is completely unordered, so even on a non-coherent
> system a regular read should suffice, any old value is correct.

the words you use seem to form a line of reasoning that makes sense to
me.  we'll have to play first though to double check.

> The only problem would be when you could get cache aliasing and read
> something totally unrelated.

being a nommu arch, there shouldnt be any cache aliasing issues.
we're just trying to make sure that what another core has pushed out
isnt stale in another core's cache when the other core does the read.
-mike
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/