On Fri, Oct 31, 2014 at 10:57:06AM +0100, Maxime Ripard wrote:

On a 3.18-rc2 kernel:

$ git grep kmap -- crypto/
crypto/ahash.c:                         walk->data = kmap(walk->pg);
crypto/ahash.c:                         walk->data = kmap_atomic(walk->pg);
crypto/async_tx/async_memcpy.c:         dest_buf = kmap_atomic(dest) + dest_offset;
crypto/async_tx/async_memcpy.c:         src_buf = kmap_atomic(src) + src_offset;
crypto/scatterwalk.c:                   return kmap_atomic(scatterwalk_page(walk)) +
crypto/shash.c:                         data = kmap_atomic(sg_page(sg));
crypto/shash.c:                         data = kmap_atomic(sg_page(sg));

None of the drivers are.

What do you mean? It's precisely because the page can be in highmem
that we are mapping it.  If it's not in highmem it'll be a noop.

Admittedly I haven't tested highmem since moving over to x86-64
some years ago, but it definitely used to work on x86-32.

Cheers,
-- 
Email: Herbert Xu [off-list ref]
Home Page: http://gondor.apana.org.au/~herbert/
PGP Key: http://gondor.apana.org.au/~herbert/pubkey.txt