From patchwork Wed Dec 11 21:08:13 2013 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Kevin Wolf X-Patchwork-Id: 300428 Return-Path: X-Original-To: incoming@patchwork.ozlabs.org Delivered-To: patchwork-incoming@bilbo.ozlabs.org Received: from lists.gnu.org (lists.gnu.org [IPv6:2001:4830:134:3::11]) (using TLSv1 with cipher AES256-SHA (256/256 bits)) (Client did not present a certificate) by ozlabs.org (Postfix) with ESMTPS id A86552C00AE for ; Thu, 12 Dec 2013 08:19:03 +1100 (EST) Received: from localhost ([::1]:60233 helo=lists.gnu.org) by lists.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1Vqr4I-0002Gq-Sj for incoming@patchwork.ozlabs.org; Wed, 11 Dec 2013 16:11:42 -0500 Received: from eggs.gnu.org ([2001:4830:134:3::10]:43875) by lists.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1Vqr1k-0007Hr-Ra for qemu-devel@nongnu.org; Wed, 11 Dec 2013 16:09:10 -0500 Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71) (envelope-from ) id 1Vqr1e-0000i1-Nh for qemu-devel@nongnu.org; Wed, 11 Dec 2013 16:09:04 -0500 Received: from mx1.redhat.com ([209.132.183.28]:44898) by eggs.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1Vqr1e-0000hv-Fd for qemu-devel@nongnu.org; Wed, 11 Dec 2013 16:08:58 -0500 Received: from int-mx12.intmail.prod.int.phx2.redhat.com (int-mx12.intmail.prod.int.phx2.redhat.com [10.5.11.25]) by mx1.redhat.com (8.14.4/8.14.4) with ESMTP id rBBL8sNb024932 (version=TLSv1/SSLv3 cipher=DHE-RSA-AES256-SHA bits=256 verify=OK); Wed, 11 Dec 2013 16:08:54 -0500 Received: from dhcp-200-207.str.redhat.com (ovpn-116-104.ams2.redhat.com [10.36.116.104]) by int-mx12.intmail.prod.int.phx2.redhat.com (8.14.4/8.14.4) with ESMTP id rBBL8d8p028102; Wed, 11 Dec 2013 16:08:53 -0500 From: Kevin Wolf To: qemu-devel@nongnu.org Date: Wed, 11 Dec 2013 22:08:13 +0100 Message-Id: <1386796109-15264-7-git-send-email-kwolf@redhat.com> In-Reply-To: <1386796109-15264-1-git-send-email-kwolf@redhat.com> References: <1386796109-15264-1-git-send-email-kwolf@redhat.com> X-Scanned-By: MIMEDefang 2.68 on 10.5.11.25 X-detected-operating-system: by eggs.gnu.org: GNU/Linux 3.x X-Received-From: 209.132.183.28 Cc: kwolf@redhat.com, pbonzini@redhat.com, pl@kamp.de, stefanha@redhat.com Subject: [Qemu-devel] [PATCH 06/22] block: Don't use guest sector size for qemu_blockalign() X-BeenThere: qemu-devel@nongnu.org X-Mailman-Version: 2.1.14 Precedence: list List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Errors-To: qemu-devel-bounces+incoming=patchwork.ozlabs.org@nongnu.org Sender: qemu-devel-bounces+incoming=patchwork.ozlabs.org@nongnu.org bs->buffer_alignment is set by the device emulation and contains the logical block size of the guest device. This isn't something that the block layer should know, and even less something to use for determining the right alignment of buffers to be used for the host. The new BlockLimits field opt_mem_alignment tells the qemu block layer the optimal alignment to be used so that no bounce buffer must be used in the driver. This patch may change the buffer alignment from 4k to 512 for all callers that used qemu_blockalign() with the top-level image format BlockDriverState. The value was never propagated to other levels in the tree, so in particular raw-posix never required anything else than 512. While on disks with 4k sectors direct I/O requires a 4k alignment, memory may still be okay when aligned to 512 byte boundaries. This is what must have happened in practice, because otherwise this would already have failed earlier. Therefore I don't expect regressions even with this intermediate state. Later, raw-posix can implement the hook and expose a different memory alignment requirement. Signed-off-by: Kevin Wolf Reviewed-by: Wenchao Xia --- block.c | 23 ++++++++++++++++++++--- include/block/block.h | 3 +++ include/block/block_int.h | 3 +++ 3 files changed, 26 insertions(+), 3 deletions(-) diff --git a/block.c b/block.c index 62c5a75..d50a4e4 100644 --- a/block.c +++ b/block.c @@ -214,6 +214,16 @@ static void bdrv_io_limits_intercept(BlockDriverState *bs, qemu_co_queue_next(&bs->throttled_reqs[is_write]); } +size_t bdrv_opt_mem_align(BlockDriverState *bs) +{ + if (!bs || !bs->drv) { + /* 4k should be on the safe side */ + return 4096; + } + + return bs->bl.opt_mem_alignment; +} + /* check if the path starts with ":" */ static int path_has_protocol(const char *path) { @@ -493,6 +503,9 @@ int bdrv_refresh_limits(BlockDriverState *bs) if (bs->file) { bdrv_refresh_limits(bs->file); bs->bl.opt_transfer_length = bs->file->bl.opt_transfer_length; + bs->bl.opt_mem_alignment = bs->file->bl.opt_mem_alignment; + } else { + bs->bl.opt_mem_alignment = 512; } if (bs->backing_hd) { @@ -500,6 +513,9 @@ int bdrv_refresh_limits(BlockDriverState *bs) bs->bl.opt_transfer_length = MAX(bs->bl.opt_transfer_length, bs->backing_hd->bl.opt_transfer_length); + bs->bl.opt_mem_alignment = + MAX(bs->bl.opt_mem_alignment, + bs->backing_hd->bl.opt_mem_alignment); } /* Then let the driver override it */ @@ -4547,7 +4563,7 @@ void bdrv_set_buffer_alignment(BlockDriverState *bs, int align) void *qemu_blockalign(BlockDriverState *bs, size_t size) { - return qemu_memalign((bs && bs->buffer_alignment) ? bs->buffer_alignment : 512, size); + return qemu_memalign(bdrv_opt_mem_align(bs), size); } /* @@ -4556,12 +4572,13 @@ void *qemu_blockalign(BlockDriverState *bs, size_t size) bool bdrv_qiov_is_aligned(BlockDriverState *bs, QEMUIOVector *qiov) { int i; + size_t alignment = bdrv_opt_mem_align(bs); for (i = 0; i < qiov->niov; i++) { - if ((uintptr_t) qiov->iov[i].iov_base % bs->buffer_alignment) { + if ((uintptr_t) qiov->iov[i].iov_base % alignment) { return false; } - if (qiov->iov[i].iov_len % bs->buffer_alignment) { + if (qiov->iov[i].iov_len % alignment) { return false; } } diff --git a/include/block/block.h b/include/block/block.h index 64fb319..cf63ee2 100644 --- a/include/block/block.h +++ b/include/block/block.h @@ -419,6 +419,9 @@ void bdrv_img_create(const char *filename, const char *fmt, char *options, uint64_t img_size, int flags, Error **errp, bool quiet); +/* Returns the alignment in bytes that is required so that no bounce buffer + * is required throughout the stack */ +size_t bdrv_opt_mem_align(BlockDriverState *bs); void bdrv_set_buffer_alignment(BlockDriverState *bs, int align); void *qemu_blockalign(BlockDriverState *bs, size_t size); bool bdrv_qiov_is_aligned(BlockDriverState *bs, QEMUIOVector *qiov); diff --git a/include/block/block_int.h b/include/block/block_int.h index c49fa6b..6ffae7c 100644 --- a/include/block/block_int.h +++ b/include/block/block_int.h @@ -252,6 +252,9 @@ typedef struct BlockLimits { /* optimal transfer length in sectors */ int opt_transfer_length; + + /* memory alignment so that no bounce buffer is needed */ + size_t opt_mem_alignment; } BlockLimits; /*