From patchwork Tue Jul 11 16:07:30 2017 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Max Reitz X-Patchwork-Id: 9835129 Return-Path: Received: from mail.wl.linuxfoundation.org (pdx-wl-mail.web.codeaurora.org [172.30.200.125]) by pdx-korg-patchwork.web.codeaurora.org (Postfix) with ESMTP id D58F860363 for ; Tue, 11 Jul 2017 16:42:16 +0000 (UTC) Received: from mail.wl.linuxfoundation.org (localhost [127.0.0.1]) by mail.wl.linuxfoundation.org (Postfix) with ESMTP id C334F28511 for ; Tue, 11 Jul 2017 16:42:16 +0000 (UTC) Received: by mail.wl.linuxfoundation.org (Postfix, from userid 486) id B7AAF28543; Tue, 11 Jul 2017 16:42:16 +0000 (UTC) X-Spam-Checker-Version: SpamAssassin 3.3.1 (2010-03-16) on pdx-wl-mail.web.codeaurora.org X-Spam-Level: X-Spam-Status: No, score=-6.9 required=2.0 tests=BAYES_00,RCVD_IN_DNSWL_HI autolearn=ham version=3.3.1 Received: from lists.gnu.org (lists.gnu.org [208.118.235.17]) (using TLSv1 with cipher AES256-SHA (256/256 bits)) (No client certificate requested) by mail.wl.linuxfoundation.org (Postfix) with ESMTPS id A0BD428511 for ; Tue, 11 Jul 2017 16:42:15 +0000 (UTC) Received: from localhost ([::1]:47499 helo=lists.gnu.org) by lists.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1dUyEo-0002na-LJ for patchwork-qemu-devel@patchwork.kernel.org; Tue, 11 Jul 2017 12:42:14 -0400 Received: from eggs.gnu.org ([2001:4830:134:3::10]:45954) by lists.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1dUxkq-0000dH-LN for qemu-devel@nongnu.org; Tue, 11 Jul 2017 12:11:20 -0400 Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71) (envelope-from ) id 1dUxko-0006Z7-Id for qemu-devel@nongnu.org; Tue, 11 Jul 2017 12:11:16 -0400 Received: from mx1.redhat.com ([209.132.183.28]:52596) by eggs.gnu.org with esmtps (TLS1.0:DHE_RSA_AES_256_CBC_SHA1:32) (Exim 4.71) (envelope-from ) id 1dUxkk-0006Xp-3g; Tue, 11 Jul 2017 12:11:10 -0400 Received: from smtp.corp.redhat.com (int-mx04.intmail.prod.int.phx2.redhat.com [10.5.11.14]) (using TLSv1.2 with cipher AECDH-AES256-SHA (256/256 bits)) (No client certificate requested) by mx1.redhat.com (Postfix) with ESMTPS id 6DC7B3D94F; Tue, 11 Jul 2017 16:11:08 +0000 (UTC) DMARC-Filter: OpenDMARC Filter v1.3.2 mx1.redhat.com 6DC7B3D94F Authentication-Results: ext-mx06.extmail.prod.ext.phx2.redhat.com; dmarc=none (p=none dis=none) header.from=redhat.com Authentication-Results: ext-mx06.extmail.prod.ext.phx2.redhat.com; spf=pass smtp.mailfrom=mreitz@redhat.com DKIM-Filter: OpenDKIM Filter v2.11.0 mx1.redhat.com 6DC7B3D94F Received: from localhost (ovpn-204-123.brq.redhat.com [10.40.204.123]) by smtp.corp.redhat.com (Postfix) with ESMTPS id 736D65D9C8; Tue, 11 Jul 2017 16:11:06 +0000 (UTC) From: Max Reitz To: qemu-block@nongnu.org Date: Tue, 11 Jul 2017 18:07:30 +0200 Message-Id: <20170711160814.20941-42-mreitz@redhat.com> In-Reply-To: <20170711160814.20941-1-mreitz@redhat.com> References: <20170711160814.20941-1-mreitz@redhat.com> X-Scanned-By: MIMEDefang 2.79 on 10.5.11.14 X-Greylist: Sender IP whitelisted, not delayed by milter-greylist-4.5.16 (mx1.redhat.com [10.5.110.30]); Tue, 11 Jul 2017 16:11:09 +0000 (UTC) X-detected-operating-system: by eggs.gnu.org: GNU/Linux 2.2.x-3.x [generic] [fuzzy] X-Received-From: 209.132.183.28 Subject: [Qemu-devel] [PULL 41/85] qcow2: add persistent dirty bitmaps support X-BeenThere: qemu-devel@nongnu.org X-Mailman-Version: 2.1.21 Precedence: list List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Cc: qemu-devel@nongnu.org, Max Reitz Errors-To: qemu-devel-bounces+patchwork-qemu-devel=patchwork.kernel.org@nongnu.org Sender: "Qemu-devel" X-Virus-Scanned: ClamAV using ClamSMTP From: Vladimir Sementsov-Ogievskiy Store persistent dirty bitmaps in qcow2 image. Signed-off-by: Vladimir Sementsov-Ogievskiy Reviewed-by: Max Reitz Message-id: 20170628120530.31251-20-vsementsov@virtuozzo.com [mreitz: Always assign ret in store_bitmap() in case of an error] Signed-off-by: Max Reitz --- block/qcow2.h | 1 + block/qcow2-bitmap.c | 476 +++++++++++++++++++++++++++++++++++++++++++++++++++ block/qcow2.c | 9 + 3 files changed, 486 insertions(+) diff --git a/block/qcow2.h b/block/qcow2.h index 633bd43..308cefd 100644 --- a/block/qcow2.h +++ b/block/qcow2.h @@ -647,5 +647,6 @@ int qcow2_check_bitmaps_refcounts(BlockDriverState *bs, BdrvCheckResult *res, int64_t *refcount_table_size); bool qcow2_load_autoloading_dirty_bitmaps(BlockDriverState *bs, Error **errp); int qcow2_reopen_bitmaps_rw(BlockDriverState *bs, Error **errp); +void qcow2_store_persistent_dirty_bitmaps(BlockDriverState *bs, Error **errp); #endif diff --git a/block/qcow2-bitmap.c b/block/qcow2-bitmap.c index 52e4616..81c49ae 100644 --- a/block/qcow2-bitmap.c +++ b/block/qcow2-bitmap.c @@ -27,6 +27,7 @@ #include "qemu/osdep.h" #include "qapi/error.h" +#include "qemu/cutils.h" #include "block/block_int.h" #include "block/qcow2.h" @@ -42,6 +43,10 @@ #define BME_MIN_GRANULARITY_BITS 9 #define BME_MAX_NAME_SIZE 1023 +#if BME_MAX_TABLE_SIZE * 8ULL > INT_MAX +#error In the code bitmap table physical size assumed to fit into int +#endif + /* Bitmap directory entry flags */ #define BME_RESERVED_FLAGS 0xfffffffcU #define BME_FLAG_IN_USE (1U << 0) @@ -72,6 +77,8 @@ typedef struct Qcow2BitmapTable { uint32_t size; /* number of 64bit entries */ QSIMPLEQ_ENTRY(Qcow2BitmapTable) entry; } Qcow2BitmapTable; +typedef QSIMPLEQ_HEAD(Qcow2BitmapTableList, Qcow2BitmapTable) + Qcow2BitmapTableList; typedef struct Qcow2Bitmap { Qcow2BitmapTable table; @@ -79,6 +86,8 @@ typedef struct Qcow2Bitmap { uint8_t granularity_bits; char *name; + BdrvDirtyBitmap *dirty_bitmap; + QSIMPLEQ_ENTRY(Qcow2Bitmap) entry; } Qcow2Bitmap; typedef QSIMPLEQ_HEAD(Qcow2BitmapList, Qcow2Bitmap) Qcow2BitmapList; @@ -104,6 +113,15 @@ static int update_header_sync(BlockDriverState *bs) return bdrv_flush(bs); } +static inline void bitmap_table_to_be(uint64_t *bitmap_table, size_t size) +{ + size_t i; + + for (i = 0; i < size; ++i) { + cpu_to_be64s(&bitmap_table[i]); + } +} + static int check_table_entry(uint64_t entry, int cluster_size) { uint64_t offset; @@ -127,6 +145,70 @@ static int check_table_entry(uint64_t entry, int cluster_size) return 0; } +static int check_constraints_on_bitmap(BlockDriverState *bs, + const char *name, + uint32_t granularity, + Error **errp) +{ + BDRVQcow2State *s = bs->opaque; + int granularity_bits = ctz32(granularity); + int64_t len = bdrv_getlength(bs); + + assert(granularity > 0); + assert((granularity & (granularity - 1)) == 0); + + if (len < 0) { + error_setg_errno(errp, -len, "Failed to get size of '%s'", + bdrv_get_device_or_node_name(bs)); + return len; + } + + if (granularity_bits > BME_MAX_GRANULARITY_BITS) { + error_setg(errp, "Granularity exceeds maximum (%llu bytes)", + 1ULL << BME_MAX_GRANULARITY_BITS); + return -EINVAL; + } + if (granularity_bits < BME_MIN_GRANULARITY_BITS) { + error_setg(errp, "Granularity is under minimum (%llu bytes)", + 1ULL << BME_MIN_GRANULARITY_BITS); + return -EINVAL; + } + + if ((len > (uint64_t)BME_MAX_PHYS_SIZE << granularity_bits) || + (len > (uint64_t)BME_MAX_TABLE_SIZE * s->cluster_size << + granularity_bits)) + { + error_setg(errp, "Too much space will be occupied by the bitmap. " + "Use larger granularity"); + return -EINVAL; + } + + if (strlen(name) > BME_MAX_NAME_SIZE) { + error_setg(errp, "Name length exceeds maximum (%u characters)", + BME_MAX_NAME_SIZE); + return -EINVAL; + } + + return 0; +} + +static void clear_bitmap_table(BlockDriverState *bs, uint64_t *bitmap_table, + uint32_t bitmap_table_size) +{ + BDRVQcow2State *s = bs->opaque; + int i; + + for (i = 0; i < bitmap_table_size; ++i) { + uint64_t addr = bitmap_table[i] & BME_TABLE_ENTRY_OFFSET_MASK; + if (!addr) { + continue; + } + + qcow2_free_clusters(bs, addr, s->cluster_size, QCOW2_DISCARD_OTHER); + bitmap_table[i] = 0; + } +} + static int bitmap_table_load(BlockDriverState *bs, Qcow2BitmapTable *tb, uint64_t **bitmap_table) { @@ -165,6 +247,28 @@ fail: return ret; } +static int free_bitmap_clusters(BlockDriverState *bs, Qcow2BitmapTable *tb) +{ + int ret; + uint64_t *bitmap_table; + + ret = bitmap_table_load(bs, tb, &bitmap_table); + if (ret < 0) { + assert(bitmap_table == NULL); + return ret; + } + + clear_bitmap_table(bs, bitmap_table, tb->size); + qcow2_free_clusters(bs, tb->offset, tb->size * sizeof(uint64_t), + QCOW2_DISCARD_OTHER); + g_free(bitmap_table); + + tb->offset = 0; + tb->size = 0; + + return 0; +} + /* This function returns the number of disk sectors covered by a single qcow2 * cluster of bitmap data. */ static uint64_t sectors_covered_by_bitmap_cluster(const BDRVQcow2State *s, @@ -748,6 +852,69 @@ static int update_ext_header_and_dir_in_place(BlockDriverState *bs, */ } +static int update_ext_header_and_dir(BlockDriverState *bs, + Qcow2BitmapList *bm_list) +{ + BDRVQcow2State *s = bs->opaque; + int ret; + uint64_t new_offset = 0; + uint64_t new_size = 0; + uint32_t new_nb_bitmaps = 0; + uint64_t old_offset = s->bitmap_directory_offset; + uint64_t old_size = s->bitmap_directory_size; + uint32_t old_nb_bitmaps = s->nb_bitmaps; + uint64_t old_autocl = s->autoclear_features; + + if (bm_list != NULL && !QSIMPLEQ_EMPTY(bm_list)) { + new_nb_bitmaps = bitmap_list_count(bm_list); + + if (new_nb_bitmaps > QCOW2_MAX_BITMAPS) { + return -EINVAL; + } + + ret = bitmap_list_store(bs, bm_list, &new_offset, &new_size, false); + if (ret < 0) { + return ret; + } + + ret = bdrv_flush(bs->file->bs); + if (ret < 0) { + goto fail; + } + + s->autoclear_features |= QCOW2_AUTOCLEAR_BITMAPS; + } else { + s->autoclear_features &= ~(uint64_t)QCOW2_AUTOCLEAR_BITMAPS; + } + + s->bitmap_directory_offset = new_offset; + s->bitmap_directory_size = new_size; + s->nb_bitmaps = new_nb_bitmaps; + + ret = update_header_sync(bs); + if (ret < 0) { + goto fail; + } + + if (old_size > 0) { + qcow2_free_clusters(bs, old_offset, old_size, QCOW2_DISCARD_OTHER); + } + + return 0; + +fail: + if (new_offset > 0) { + qcow2_free_clusters(bs, new_offset, new_size, QCOW2_DISCARD_OTHER); + } + + s->bitmap_directory_offset = old_offset; + s->bitmap_directory_size = old_size; + s->nb_bitmaps = old_nb_bitmaps; + s->autoclear_features = old_autocl; + + return ret; +} + /* for g_slist_foreach for GSList of BdrvDirtyBitmap* elements */ static void release_dirty_bitmap_helper(gpointer bitmap, gpointer bs) @@ -890,3 +1057,312 @@ out: return ret; } + +/* store_bitmap_data() + * Store bitmap to image, filling bitmap table accordingly. + */ +static uint64_t *store_bitmap_data(BlockDriverState *bs, + BdrvDirtyBitmap *bitmap, + uint32_t *bitmap_table_size, Error **errp) +{ + int ret; + BDRVQcow2State *s = bs->opaque; + int64_t sector; + uint64_t sbc; + uint64_t bm_size = bdrv_dirty_bitmap_size(bitmap); + const char *bm_name = bdrv_dirty_bitmap_name(bitmap); + uint8_t *buf = NULL; + BdrvDirtyBitmapIter *dbi; + uint64_t *tb; + uint64_t tb_size = + size_to_clusters(s, + bdrv_dirty_bitmap_serialization_size(bitmap, 0, bm_size)); + + if (tb_size > BME_MAX_TABLE_SIZE || + tb_size * s->cluster_size > BME_MAX_PHYS_SIZE) + { + error_setg(errp, "Bitmap '%s' is too big", bm_name); + return NULL; + } + + tb = g_try_new0(uint64_t, tb_size); + if (tb == NULL) { + error_setg(errp, "No memory"); + return NULL; + } + + dbi = bdrv_dirty_iter_new(bitmap, 0); + buf = g_malloc(s->cluster_size); + sbc = sectors_covered_by_bitmap_cluster(s, bitmap); + assert(DIV_ROUND_UP(bm_size, sbc) == tb_size); + + while ((sector = bdrv_dirty_iter_next(dbi)) != -1) { + uint64_t cluster = sector / sbc; + uint64_t end, write_size; + int64_t off; + + sector = cluster * sbc; + end = MIN(bm_size, sector + sbc); + write_size = + bdrv_dirty_bitmap_serialization_size(bitmap, sector, end - sector); + assert(write_size <= s->cluster_size); + + off = qcow2_alloc_clusters(bs, s->cluster_size); + if (off < 0) { + error_setg_errno(errp, -off, + "Failed to allocate clusters for bitmap '%s'", + bm_name); + goto fail; + } + tb[cluster] = off; + + bdrv_dirty_bitmap_serialize_part(bitmap, buf, sector, end - sector); + if (write_size < s->cluster_size) { + memset(buf + write_size, 0, s->cluster_size - write_size); + } + + ret = qcow2_pre_write_overlap_check(bs, 0, off, s->cluster_size); + if (ret < 0) { + error_setg_errno(errp, -ret, "Qcow2 overlap check failed"); + goto fail; + } + + ret = bdrv_pwrite(bs->file, off, buf, s->cluster_size); + if (ret < 0) { + error_setg_errno(errp, -ret, "Failed to write bitmap '%s' to file", + bm_name); + goto fail; + } + + if (end >= bm_size) { + break; + } + + bdrv_set_dirty_iter(dbi, end); + } + + *bitmap_table_size = tb_size; + g_free(buf); + bdrv_dirty_iter_free(dbi); + + return tb; + +fail: + clear_bitmap_table(bs, tb, tb_size); + g_free(buf); + bdrv_dirty_iter_free(dbi); + g_free(tb); + + return NULL; +} + +/* store_bitmap() + * Store bm->dirty_bitmap to qcow2. + * Set bm->table_offset and bm->table_size accordingly. + */ +static int store_bitmap(BlockDriverState *bs, Qcow2Bitmap *bm, Error **errp) +{ + int ret; + uint64_t *tb; + int64_t tb_offset; + uint32_t tb_size; + BdrvDirtyBitmap *bitmap = bm->dirty_bitmap; + const char *bm_name; + + assert(bitmap != NULL); + + bm_name = bdrv_dirty_bitmap_name(bitmap); + + tb = store_bitmap_data(bs, bitmap, &tb_size, errp); + if (tb == NULL) { + return -EINVAL; + } + + assert(tb_size <= BME_MAX_TABLE_SIZE); + tb_offset = qcow2_alloc_clusters(bs, tb_size * sizeof(tb[0])); + if (tb_offset < 0) { + error_setg_errno(errp, -tb_offset, + "Failed to allocate clusters for bitmap '%s'", + bm_name); + ret = tb_offset; + goto fail; + } + + ret = qcow2_pre_write_overlap_check(bs, 0, tb_offset, + tb_size * sizeof(tb[0])); + if (ret < 0) { + error_setg_errno(errp, -ret, "Qcow2 overlap check failed"); + goto fail; + } + + bitmap_table_to_be(tb, tb_size); + ret = bdrv_pwrite(bs->file, tb_offset, tb, tb_size * sizeof(tb[0])); + if (ret < 0) { + error_setg_errno(errp, -ret, "Failed to write bitmap '%s' to file", + bm_name); + goto fail; + } + + g_free(tb); + + bm->table.offset = tb_offset; + bm->table.size = tb_size; + + return 0; + +fail: + clear_bitmap_table(bs, tb, tb_size); + + if (tb_offset > 0) { + qcow2_free_clusters(bs, tb_offset, tb_size * sizeof(tb[0]), + QCOW2_DISCARD_OTHER); + } + + g_free(tb); + + return ret; +} + +static Qcow2Bitmap *find_bitmap_by_name(Qcow2BitmapList *bm_list, + const char *name) +{ + Qcow2Bitmap *bm; + + QSIMPLEQ_FOREACH(bm, bm_list, entry) { + if (strcmp(name, bm->name) == 0) { + return bm; + } + } + + return NULL; +} + +void qcow2_store_persistent_dirty_bitmaps(BlockDriverState *bs, Error **errp) +{ + BdrvDirtyBitmap *bitmap; + BDRVQcow2State *s = bs->opaque; + uint32_t new_nb_bitmaps = s->nb_bitmaps; + uint64_t new_dir_size = s->bitmap_directory_size; + int ret; + Qcow2BitmapList *bm_list; + Qcow2Bitmap *bm; + Qcow2BitmapTableList drop_tables; + Qcow2BitmapTable *tb, *tb_next; + + if (!bdrv_has_changed_persistent_bitmaps(bs)) { + /* nothing to do */ + return; + } + + if (!can_write(bs)) { + error_setg(errp, "No write access"); + return; + } + + QSIMPLEQ_INIT(&drop_tables); + + if (s->nb_bitmaps == 0) { + bm_list = bitmap_list_new(); + } else { + bm_list = bitmap_list_load(bs, s->bitmap_directory_offset, + s->bitmap_directory_size, errp); + if (bm_list == NULL) { + return; + } + } + + /* check constraints and names */ + for (bitmap = bdrv_dirty_bitmap_next(bs, NULL); bitmap != NULL; + bitmap = bdrv_dirty_bitmap_next(bs, bitmap)) + { + const char *name = bdrv_dirty_bitmap_name(bitmap); + uint32_t granularity = bdrv_dirty_bitmap_granularity(bitmap); + Qcow2Bitmap *bm; + + if (!bdrv_dirty_bitmap_get_persistance(bitmap) || + bdrv_dirty_bitmap_readonly(bitmap)) + { + continue; + } + + if (check_constraints_on_bitmap(bs, name, granularity, errp) < 0) { + error_prepend(errp, "Bitmap '%s' doesn't satisfy the constraints: ", + name); + goto fail; + } + + bm = find_bitmap_by_name(bm_list, name); + if (bm == NULL) { + if (++new_nb_bitmaps > QCOW2_MAX_BITMAPS) { + error_setg(errp, "Too many persistent bitmaps"); + goto fail; + } + + new_dir_size += calc_dir_entry_size(strlen(name), 0); + if (new_dir_size > QCOW2_MAX_BITMAP_DIRECTORY_SIZE) { + error_setg(errp, "Bitmap directory is too large"); + goto fail; + } + + bm = g_new0(Qcow2Bitmap, 1); + bm->name = g_strdup(name); + QSIMPLEQ_INSERT_TAIL(bm_list, bm, entry); + } else { + if (!(bm->flags & BME_FLAG_IN_USE)) { + error_setg(errp, "Bitmap '%s' already exists in the image", + name); + goto fail; + } + tb = g_memdup(&bm->table, sizeof(bm->table)); + bm->table.offset = 0; + bm->table.size = 0; + QSIMPLEQ_INSERT_TAIL(&drop_tables, tb, entry); + } + bm->flags = bdrv_dirty_bitmap_get_autoload(bitmap) ? BME_FLAG_AUTO : 0; + bm->granularity_bits = ctz32(bdrv_dirty_bitmap_granularity(bitmap)); + bm->dirty_bitmap = bitmap; + } + + /* allocate clusters and store bitmaps */ + QSIMPLEQ_FOREACH(bm, bm_list, entry) { + if (bm->dirty_bitmap == NULL) { + continue; + } + + ret = store_bitmap(bs, bm, errp); + if (ret < 0) { + goto fail; + } + } + + ret = update_ext_header_and_dir(bs, bm_list); + if (ret < 0) { + error_setg_errno(errp, -ret, "Failed to update bitmap extension"); + goto fail; + } + + /* Bitmap directory was successfully updated, so, old data can be dropped. + * TODO it is better to reuse these clusters */ + QSIMPLEQ_FOREACH_SAFE(tb, &drop_tables, entry, tb_next) { + free_bitmap_clusters(bs, tb); + g_free(tb); + } + + bitmap_list_free(bm_list); + return; + +fail: + QSIMPLEQ_FOREACH(bm, bm_list, entry) { + if (bm->dirty_bitmap == NULL || bm->table.offset == 0) { + continue; + } + + free_bitmap_clusters(bs, &bm->table); + } + + QSIMPLEQ_FOREACH_SAFE(tb, &drop_tables, entry, tb_next) { + g_free(tb); + } + + bitmap_list_free(bm_list); +} diff --git a/block/qcow2.c b/block/qcow2.c index 19cc49a..790b5b1 100644 --- a/block/qcow2.c +++ b/block/qcow2.c @@ -2046,6 +2046,7 @@ static int qcow2_inactivate(BlockDriverState *bs) { BDRVQcow2State *s = bs->opaque; int ret, result = 0; + Error *local_err = NULL; ret = qcow2_cache_flush(bs, s->l2_table_cache); if (ret) { @@ -2061,6 +2062,14 @@ static int qcow2_inactivate(BlockDriverState *bs) strerror(-ret)); } + qcow2_store_persistent_dirty_bitmaps(bs, &local_err); + if (local_err != NULL) { + result = -EINVAL; + error_report_err(local_err); + error_report("Persistent bitmaps are lost for node '%s'", + bdrv_get_device_or_node_name(bs)); + } + if (result == 0) { qcow2_mark_clean(bs); }