From mboxrd@z Thu Jan 1 00:00:00 1970 From: Dan Williams Subject: [md PATCH 2/4] async_pq: kill a stray dma_map() call and other cleanups Date: Tue, 20 Oct 2009 00:11:28 -0700 Message-ID: <20091020071128.6489.17320.stgit@dwillia2-linux.ch.intel.com> References: <20091020070627.6489.4006.stgit@dwillia2-linux.ch.intel.com> Mime-Version: 1.0 Content-Type: text/plain; charset="utf-8" Content-Transfer-Encoding: 7bit Return-path: In-Reply-To: <20091020070627.6489.4006.stgit@dwillia2-linux.ch.intel.com> Sender: linux-raid-owner@vger.kernel.org To: neilb@suse.de Cc: linux-raid@vger.kernel.org List-Id: linux-raid.ids - update the kernel doc for async_syndrome to indicate what NULL in the source list means - whitespace fixups Signed-off-by: Dan Williams --- crypto/async_tx/async_pq.c | 15 ++++++++------- 1 files changed, 8 insertions(+), 7 deletions(-) diff --git a/crypto/async_tx/async_pq.c b/crypto/async_tx/async_pq.c index 43b1436..6047656 100644 --- a/crypto/async_tx/async_pq.c +++ b/crypto/async_tx/async_pq.c @@ -181,10 +181,14 @@ do_sync_gen_syndrome(struct page **blocks, unsigned int offset, int disks, * blocks[disks-1] to NULL. When P or Q is omitted 'len' must be <= * PAGE_SIZE as a temporary buffer of this size is used in the * synchronous path. 'disks' always accounts for both destination - * buffers. + * buffers. If any source buffers (blocks[i] where i < disks - 2) are + * set to NULL those buffers will be replaced with the raid6_zero_page + * in the synchronous path and omitted in the hardware-asynchronous + * path. * * 'blocks' note: if submit->scribble is NULL then the contents of - * 'blocks' may be overridden + * 'blocks' may be overwritten to perform address conversions + * (dma_map_page() or page_address()). */ struct dma_async_tx_descriptor * async_gen_syndrome(struct page **blocks, unsigned int offset, int disks, @@ -283,13 +287,13 @@ async_syndrome_val(struct page **blocks, unsigned int offset, int disks, if (!P(blocks, disks)) dma_flags |= DMA_PREP_PQ_DISABLE_P; else - pq[0] = dma_map_page(dev, P(blocks,disks), + pq[0] = dma_map_page(dev, P(blocks, disks), offset, len, DMA_TO_DEVICE); if (!Q(blocks, disks)) dma_flags |= DMA_PREP_PQ_DISABLE_Q; else - pq[1] = dma_map_page(dev, Q(blocks,disks), + pq[1] = dma_map_page(dev, Q(blocks, disks), offset, len, DMA_TO_DEVICE); @@ -303,9 +307,6 @@ async_syndrome_val(struct page **blocks, unsigned int offset, int disks, coefs[src_cnt] = raid6_gfexp[i]; src_cnt++; } - pq[1] = dma_map_page(dev, Q(blocks,disks), - offset, len, - DMA_TO_DEVICE); for (;;) { tx = device->device_prep_dma_pq_val(chan, pq, dma_src,