On Tue, 20 Jul 2010, Catalin Marinas wrote: > On Tue, 2010-07-20 at 18:56 +0900, FUJITA Tomonori wrote: > > On Fri, 16 Jul 2010 15:39:17 +0100 > > Catalin Marinas <catalin.marinas@xxxxxxx> wrote: > > > I'm not sure why sg_miter doesn't call flush_dcache_page() but > > > flush_kernel_dcache_page() (I cc'ed Fujita as he seems to have added > > > this code). > > > > It is intended to handle D aliasing due to kmap. As cachetlb.txt says > > that it is assumed here that the user has no incoherent cached copies > > (it was already addressed elsewhere). So flush_dcache_page() does too > > much there. > > A related question here - does flush_kernel_dcache_page() need to do > anything for non-highmem pages? For highmem pages, on ARM, we already > handle the cache flushing in kunmap(). Well, with your proposed patch we wouldn't anymore. And semantically we shouldn't have to. Even the fact that kunmap_atomic() does flush the cache on VIVT system is an implementation issue not a semantic definition (incidentally on VIPT it is not flushed). Nicolas -- To unsubscribe from this list: send the line "unsubscribe linux-arch" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html