From patchwork Wed Jun 15 17:14:04 2011 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Marcelo Tosatti X-Patchwork-Id: 100564 Return-Path: X-Original-To: incoming@patchwork.ozlabs.org Delivered-To: patchwork-incoming@bilbo.ozlabs.org Received: from lists.gnu.org (lists.gnu.org [140.186.70.17]) (using TLSv1 with cipher AES256-SHA (256/256 bits)) (Client did not present a certificate) by ozlabs.org (Postfix) with ESMTPS id 96FD2B6F8F for ; Thu, 16 Jun 2011 05:02:40 +1000 (EST) Received: from localhost ([::1]:51113 helo=lists.gnu.org) by lists.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1QWvMJ-0002Ls-Ky for incoming@patchwork.ozlabs.org; Wed, 15 Jun 2011 15:02:35 -0400 Received: from eggs.gnu.org ([140.186.70.92]:46942) by lists.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1QWu5r-0004Eb-A6 for qemu-devel@nongnu.org; Wed, 15 Jun 2011 13:41:33 -0400 Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71) (envelope-from ) id 1QWu5o-00062Q-G8 for qemu-devel@nongnu.org; Wed, 15 Jun 2011 13:41:30 -0400 Received: from mx1.redhat.com ([209.132.183.28]:45109) by eggs.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1QWu5n-00062D-Rz for qemu-devel@nongnu.org; Wed, 15 Jun 2011 13:41:28 -0400 Received: from int-mx09.intmail.prod.int.phx2.redhat.com (int-mx09.intmail.prod.int.phx2.redhat.com [10.5.11.22]) by mx1.redhat.com (8.14.4/8.14.4) with ESMTP id p5FHfQi4003471 (version=TLSv1/SSLv3 cipher=DHE-RSA-AES256-SHA bits=256 verify=OK); Wed, 15 Jun 2011 13:41:26 -0400 Received: from ns3.rdu.redhat.com (ns3.rdu.redhat.com [10.11.255.199]) by int-mx09.intmail.prod.int.phx2.redhat.com (8.14.4/8.14.4) with ESMTP id p5FHfPu6008322; Wed, 15 Jun 2011 13:41:26 -0400 Received: from amt.cnet (vpn-10-33.rdu.redhat.com [10.11.10.33]) by ns3.rdu.redhat.com (8.13.8/8.13.8) with ESMTP id p5FHfOHN023069; Wed, 15 Jun 2011 13:41:24 -0400 Received: from amt.cnet (amt.cnet [127.0.0.1]) by amt.cnet (Postfix) with ESMTP id 23BEE6560E8; Wed, 15 Jun 2011 14:41:06 -0300 (BRT) Received: (from marcelo@localhost) by amt.cnet (8.14.5/8.14.5/Submit) id p5FHf4IA023920; Wed, 15 Jun 2011 14:41:04 -0300 Message-Id: <20110615174008.350658294@amt.cnet> User-Agent: quilt/0.48-1 Date: Wed, 15 Jun 2011 14:14:04 -0300 From: Marcelo Tosatti To: qemu-devel@nongnu.org References: <20110615171403.158790293@amt.cnet> Content-Disposition: inline; filename=blkmirror X-Scanned-By: MIMEDefang 2.68 on 10.5.11.22 X-detected-operating-system: by eggs.gnu.org: Genre and OS details not recognized. X-Received-From: 209.132.183.28 Cc: kwolf@redhat.com, Marcelo Tosatti , Jes.Sorensen@redhat.com, dlaor@redhat.com, avi@redhat.com, stefanha@gmail.com, jdenemar@redhat.com Subject: [Qemu-devel] [patch 1/4] Add blkmirror block driver 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 Mirrored writes are used by live block copy. Signed-off-by: Marcelo Tosatti Index: qemu-block-copy/block/blkmirror.c =================================================================== --- /dev/null +++ qemu-block-copy/block/blkmirror.c @@ -0,0 +1,282 @@ +/* + * Block driver for mirrored writes. + * + * Copyright (C) 2011 Red Hat, Inc. + * + * This work is licensed under the terms of the GNU GPL, version 2 or later. + * See the COPYING file in the top-level directory. + */ + +#include +#include "block_int.h" + +typedef struct { + BlockDriverState *bs[2]; +} BdrvMirrorState; + +typedef struct DupAIOCB DupAIOCB; + +typedef struct SingleAIOCB { + BlockDriverAIOCB *aiocb; + int finished; + DupAIOCB *parent; +} SingleAIOCB; + +struct DupAIOCB { + BlockDriverAIOCB common; + int count; + + BlockDriverCompletionFunc *cb; + SingleAIOCB aios[2]; + int ret; +}; + +/* Valid blkmirror filenames look like + * blkmirror:path/to/image1:path/to/image2 */ +static int blkmirror_open(BlockDriverState *bs, const char *filename, int flags) +{ + BdrvMirrorState *m = bs->opaque; + int ret, escape, i, n; + char *filename2; + + /* Parse the blkmirror: prefix */ + if (strncmp(filename, "blkmirror:", strlen("blkmirror:"))) { + return -EINVAL; + } + filename += strlen("blkmirror:"); + + /* Parse the raw image filename */ + filename2 = qemu_malloc(strlen(filename)+1); + escape = 0; + for (i = n = 0; i < strlen(filename); i++) { + if (!escape && filename[i] == ':') { + break; + } + if (!escape && filename[i] == '\\') { + escape = 1; + } else { + escape = 0; + } + + if (!escape) { + filename2[n++] = filename[i]; + } + } + filename2[n] = '\0'; + + m->bs[0] = bdrv_new(""); + if (m->bs[0] == NULL) { + free(filename2); + return -ENOMEM; + } + ret = bdrv_open(m->bs[0], filename2, flags, NULL); + free(filename2); + if (ret < 0) { + return ret; + } + filename += i + 1; + + m->bs[1] = bdrv_new(""); + if (m->bs[1] == NULL) { + bdrv_delete(m->bs[0]); + return -ENOMEM; + } + ret = bdrv_open(m->bs[1], filename, flags, NULL); + if (ret < 0) { + bdrv_delete(m->bs[0]); + return ret; + } + + return 0; +} + +static void blkmirror_close(BlockDriverState *bs) +{ + BdrvMirrorState *m = bs->opaque; + int i; + + for (i = 0; i < 2; i++) { + bdrv_delete(m->bs[i]); + m->bs[i] = NULL; + } +} + +static int blkmirror_flush(BlockDriverState *bs) +{ + BdrvMirrorState *m = bs->opaque; + int ret; + + ret = bdrv_flush(m->bs[0]); + if (ret < 0) { + return ret; + } + + return bdrv_flush(m->bs[1]); +} + +static int64_t blkmirror_getlength(BlockDriverState *bs) +{ + BdrvMirrorState *m = bs->opaque; + + return bdrv_getlength(m->bs[0]); +} + +static BlockDriverAIOCB *blkmirror_aio_readv(BlockDriverState *bs, + int64_t sector_num, + QEMUIOVector *qiov, + int nb_sectors, + BlockDriverCompletionFunc *cb, + void *opaque) +{ + BdrvMirrorState *m = bs->opaque; + return bdrv_aio_readv(m->bs[0], sector_num, qiov, nb_sectors, cb, opaque); +} + +static void dup_aio_cancel(BlockDriverAIOCB *blockacb) +{ + DupAIOCB *acb = container_of(blockacb, DupAIOCB, common); + int i; + + for (i = 0 ; i < 2; i++) { + if (!acb->aios[i].finished) { + bdrv_aio_cancel(acb->aios[i].aiocb); + } + } + qemu_aio_release(acb); +} + +static AIOPool dup_aio_pool = { + .aiocb_size = sizeof(DupAIOCB), + .cancel = dup_aio_cancel, +}; + +static void blkmirror_aio_cb(void *opaque, int ret) +{ + SingleAIOCB *scb = opaque; + DupAIOCB *dcb = scb->parent; + + scb->finished = 1; + dcb->count--; + assert(dcb->count >= 0); + if (ret < 0) { + dcb->ret = ret; + } + if (dcb->count == 0) { + dcb->common.cb(dcb->common.opaque, dcb->ret); + qemu_aio_release(dcb); + } +} + +static DupAIOCB *dup_aio_get(BlockDriverState *bs, + BlockDriverCompletionFunc *cb, + void *opaque) +{ + DupAIOCB *dcb; + int i; + + dcb = qemu_aio_get(&dup_aio_pool, bs, cb, opaque); + if (!dcb) { + return NULL; + } + dcb->count = 2; + for (i = 0; i < 2; i++) { + dcb->aios[i].parent = dcb; + dcb->aios[i].finished = 0; + } + dcb->ret = 0; + + return dcb; +} + +static BlockDriverAIOCB *blkmirror_aio_writev(BlockDriverState *bs, + int64_t sector_num, + QEMUIOVector *qiov, + int nb_sectors, + BlockDriverCompletionFunc *cb, + void *opaque) +{ + BdrvMirrorState *m = bs->opaque; + DupAIOCB *dcb = dup_aio_get(bs, cb, opaque); + int i; + + for (i = 0; i < 2; i++) { + dcb->aios[i].aiocb = bdrv_aio_writev(m->bs[i], sector_num, qiov, + nb_sectors, &blkmirror_aio_cb, + &dcb->aios[i]); + if (!dcb->aios[i].aiocb) { + int a; + + for (a = 0; a < i; a++) { + bdrv_aio_cancel(dcb->aios[i].aiocb); + } + qemu_aio_release(dcb); + return NULL; + } + } + + return &dcb->common; +} + +static BlockDriverAIOCB *blkmirror_aio_flush(BlockDriverState *bs, + BlockDriverCompletionFunc *cb, + void *opaque) +{ + BdrvMirrorState *m = bs->opaque; + DupAIOCB *dcb = dup_aio_get(bs, cb, opaque); + int i; + + for (i = 0; i < 2; i++) { + dcb->aios[i].aiocb = bdrv_aio_flush(m->bs[i], &blkmirror_aio_cb, + &dcb->aios[i]); + if (!dcb->aios[i].aiocb) { + int a; + + for (a = 0; a < i; a++) { + bdrv_aio_cancel(dcb->aios[i].aiocb); + } + qemu_aio_release(dcb); + return NULL; + } + } + + return &dcb->common; +} + +static int blkmirror_discard(BlockDriverState *bs, int64_t sector_num, + int nb_sectors) +{ + BdrvMirrorState *m = bs->opaque; + int ret; + + ret = bdrv_discard(m->bs[0], sector_num, nb_sectors); + if (ret < 0) { + return ret; + } + + return bdrv_discard(m->bs[1], sector_num, nb_sectors); +} + + +static BlockDriver bdrv_blkmirror = { + .format_name = "blkmirror", + .protocol_name = "blkmirror", + .instance_size = sizeof(BdrvMirrorState), + + .bdrv_getlength = blkmirror_getlength, + + .bdrv_file_open = blkmirror_open, + .bdrv_close = blkmirror_close, + .bdrv_flush = blkmirror_flush, + .bdrv_discard = blkmirror_discard, + + .bdrv_aio_readv = blkmirror_aio_readv, + .bdrv_aio_writev = blkmirror_aio_writev, + .bdrv_aio_flush = blkmirror_aio_flush, +}; + +static void bdrv_blkmirror_init(void) +{ + bdrv_register(&bdrv_blkmirror); +} + +block_init(bdrv_blkmirror_init); Index: qemu-block-copy/Makefile.objs =================================================================== --- qemu-block-copy.orig/Makefile.objs +++ qemu-block-copy/Makefile.objs @@ -22,7 +22,7 @@ block-nested-y += raw.o cow.o qcow.o vdi block-nested-y += qcow2.o qcow2-refcount.o qcow2-cluster.o qcow2-snapshot.o qcow2-cache.o block-nested-y += qed.o qed-gencb.o qed-l2-cache.o qed-table.o qed-cluster.o block-nested-y += qed-check.o -block-nested-y += parallels.o nbd.o blkdebug.o sheepdog.o blkverify.o +block-nested-y += parallels.o nbd.o blkdebug.o sheepdog.o blkverify.o blkmirror.o block-nested-$(CONFIG_WIN32) += raw-win32.o block-nested-$(CONFIG_POSIX) += raw-posix.o block-nested-$(CONFIG_CURL) += curl.o Index: qemu-block-copy/docs/blkmirror.txt =================================================================== --- /dev/null +++ qemu-block-copy/docs/blkmirror.txt @@ -0,0 +1,15 @@ +Block mirror driver +------------------- + +This driver will mirror writes to two distinct images. +Its used internally by live block copy. + +Format +------ + +blkmirror:/image1.img:/image2.img + +'\' (backslash) can be used to escape colon processing +as a separator, in the first image filename. + +