Re: [patch -mm series] ia64 specific /dev/mem handlers

[Jes Sorensen]

>>>>> "Andrew" == Andrew Morton <akpm@xxxxxxxx> writes:

Andrew> jes@xxxxxxxxxxxxxxxxxx (Jes Sorensen) wrote:
>>  Hi,
>> 
>> This patch introduces ia64 specific read/write handlers for
>> /dev/mem access which is needed to avoid uncached pages to be
>> accessed through the cached kernel window which can lead to random
>> corruption. It also introduces a new page-flag PG_uncached which
>> will be used to mark the uncached pages. I assume this may be
>> useful to other architectures as well where the CPU may use
>> speculative reads which conflict with uncached access. In addition
>> I moved do_write_mem to be under ARCH_HAS_DEV_MEM as it's only ever
>> used if that is defined.

Andrew> Is it possible to avoid consuming a page flag?

Andrew> If this is an ia64-only (or 64-bit-only) thing I guess we
Andrew> could use bit 32.

Hi Andrew,

I don't think we can get away with it, I think most modern
architectures have this problem. I've been trying to avoid it for
several iterations but I kept getting pushed towards this.

>> + if (page->flags & PG_uncached)

Andrew> dude.  That ain't gonna work ;)

Pardon my lack of clue, but why not?

Maybe I should explain how I use it - in the mspec driver I pull a
chunk of pages out using alloc_pages and convert them all to uncached
(I need to do a chunk due to the speculation restrictions on ia64) and
then stick them into a special allocator and set page->PG_uncached
when doing so (the allocator is not tied to the uncached so I posted
the patch to Tony Luck). Ie. once the pages have been grabbed with
alloc_pages() and converted to uncached, they are never put back into
the generic pool.

It may be there are other places which needs to learn to respect the
flag?

Cheers,
Jes

-
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/