On Tue, 20 May 2008 02:52:31 -0700 Andrew Morton <akpm@xxxxxxxxxxxxxxxxxxxx> wrote: > On Tue, 20 May 2008 17:38:20 +0800 Herbert Xu <herbert@xxxxxxxxxxxxxxxxxxx> wrote: > > > On Tue, May 20, 2008 at 02:31:29AM -0700, Andrew Morton wrote: > > > > > > So here you're using it for "dma aligment" whereas crypto is using it > > > (or ARCH_SLAB_MINALIGN!) for "cpu 64-bit alignment". > > > > No the 64-bit alignment is just an example. The purpose of > > CRYPTO_MINALIGN is pretty much the same as ARCH_KMALLOC_MINALIGN, > > i.e., the minimum alignment guaranteed by kmalloc. The only > > reason it exists is because ARCH_KMALLOC_MINALIGN isn't defined > > on all platforms. > > I'm struggling to understand what you're saying here. > > The comment you have there over the CRYPTO_MINALIGN definition is quite > specific. Is it wrong? > > Whether the mapping between CRYPTO_MINALIGN and ARCH_KMALLOC_MINALIGN > is abusive is (I find) hard to say, because first one would need to be > able to say what ARCH_KMALLOC_MINALIGN is for. I expect it was for DMA > purposes. Yeah, ARCH_KMALLOC_MINALIGN is for DMA. Non coherent platforms define it. -- To unsubscribe from this list: send the line "unsubscribe linux-scsi" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
