Re: [PATCHSET] block: fix PIO cache coherency bug

[Tejun Heo]

James Bottomley wrote:
> On Tue, 2006-03-21 at 01:26 +0900, Tejun Heo wrote:
>> Okay by me, although I think we also need to fix flush_dcache_page()
>> such that it doesn't abruptly enable irq after itself. That's just
>> broken. I'll make some kmap wrappers such that callers don't have to try
>> too hard to get synchronization correct.
> 
> That would involve fixing all of the flush_dcache_mmap_lock/unlock()
> wrappers to take an extra flags variable (which would be unused on null
> implementations) ... it can be done, but it's a lot of work; I think,
> since all the current users aren't in atomic context, that we shouldn't
> bother unless anyone can see a real need to call it from atomic context.
> 

Hmmm... if fixing is too much work, how about adding WARN_ON() or one of
its friends if we enter flush_dcache_page() from atomic context?

-- 
tejun
-
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/