diff options
29 files changed, 1499 insertions, 195 deletions
diff --git a/MAINTAINERS b/MAINTAINERS index 68adaac373..4f0cc1e448 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -2515,9 +2515,12 @@ F: block/stream.c F: block/mirror.c F: qapi/job.json F: block/block-copy.c -F: include/block/block-copy.c +F: include/block/block-copy.h +F: block/reqlist.c +F: include/block/reqlist.h F: block/copy-before-write.h F: block/copy-before-write.c +F: block/snapshot-access.c F: include/block/aio_task.h F: block/aio_task.c F: util/qemu-co-shared-resource.c diff --git a/block/block-copy.c b/block/block-copy.c index ce116318b5..ef948dccec 100644 --- a/block/block-copy.c +++ b/block/block-copy.c @@ -17,6 +17,7 @@ #include "trace.h" #include "qapi/error.h" #include "block/block-copy.h" +#include "block/reqlist.h" #include "sysemu/block-backend.h" #include "qemu/units.h" #include "qemu/coroutine.h" @@ -83,7 +84,6 @@ typedef struct BlockCopyTask { */ BlockCopyState *s; BlockCopyCallState *call_state; - int64_t offset; /* * @method can also be set again in the while loop of * block_copy_dirty_clusters(), but it is never accessed concurrently @@ -94,21 +94,17 @@ typedef struct BlockCopyTask { BlockCopyMethod method; /* - * Fields whose state changes throughout the execution - * Protected by lock in BlockCopyState. - */ - CoQueue wait_queue; /* coroutines blocked on this task */ - /* - * Only protect the case of parallel read while updating @bytes - * value in block_copy_task_shrink(). + * Generally, req is protected by lock in BlockCopyState, Still req.offset + * is only set on task creation, so may be read concurrently after creation. + * req.bytes is changed at most once, and need only protecting the case of + * parallel read while updating @bytes value in block_copy_task_shrink(). */ - int64_t bytes; - QLIST_ENTRY(BlockCopyTask) list; + BlockReq req; } BlockCopyTask; static int64_t task_end(BlockCopyTask *task) { - return task->offset + task->bytes; + return task->req.offset + task->req.bytes; } typedef struct BlockCopyState { @@ -136,7 +132,7 @@ typedef struct BlockCopyState { CoMutex lock; int64_t in_flight_bytes; BlockCopyMethod method; - QLIST_HEAD(, BlockCopyTask) tasks; /* All tasks from all block-copy calls */ + BlockReqList reqs; QLIST_HEAD(, BlockCopyCallState) calls; /* * skip_unallocated: @@ -161,42 +157,6 @@ typedef struct BlockCopyState { } BlockCopyState; /* Called with lock held */ -static BlockCopyTask *find_conflicting_task(BlockCopyState *s, - int64_t offset, int64_t bytes) -{ - BlockCopyTask *t; - - QLIST_FOREACH(t, &s->tasks, list) { - if (offset + bytes > t->offset && offset < t->offset + t->bytes) { - return t; - } - } - - return NULL; -} - -/* - * If there are no intersecting tasks return false. Otherwise, wait for the - * first found intersecting tasks to finish and return true. - * - * Called with lock held. May temporary release the lock. - * Return value of 0 proves that lock was NOT released. - */ -static bool coroutine_fn block_copy_wait_one(BlockCopyState *s, int64_t offset, - int64_t bytes) -{ - BlockCopyTask *task = find_conflicting_task(s, offset, bytes); - - if (!task) { - return false; - } - - qemu_co_queue_wait(&task->wait_queue, &s->lock); - - return true; -} - -/* Called with lock held */ static int64_t block_copy_chunk_size(BlockCopyState *s) { switch (s->method) { @@ -239,7 +199,7 @@ block_copy_task_create(BlockCopyState *s, BlockCopyCallState *call_state, bytes = QEMU_ALIGN_UP(bytes, s->cluster_size); /* region is dirty, so no existent tasks possible in it */ - assert(!find_conflicting_task(s, offset, bytes)); + assert(!reqlist_find_conflict(&s->reqs, offset, bytes)); bdrv_reset_dirty_bitmap(s->copy_bitmap, offset, bytes); s->in_flight_bytes += bytes; @@ -249,12 +209,9 @@ block_copy_task_create(BlockCopyState *s, BlockCopyCallState *call_state, .task.func = block_copy_task_entry, .s = s, .call_state = call_state, - .offset = offset, - .bytes = bytes, .method = s->method, }; - qemu_co_queue_init(&task->wait_queue); - QLIST_INSERT_HEAD(&s->tasks, task, list); + reqlist_init_req(&s->reqs, &task->req, offset, bytes); return task; } @@ -270,34 +227,34 @@ static void coroutine_fn block_copy_task_shrink(BlockCopyTask *task, int64_t new_bytes) { QEMU_LOCK_GUARD(&task->s->lock); - if (new_bytes == task->bytes) { + if (new_bytes == task->req.bytes) { return; } - assert(new_bytes > 0 && new_bytes < task->bytes); + assert(new_bytes > 0 && new_bytes < task->req.bytes); - task->s->in_flight_bytes -= task->bytes - new_bytes; + task->s->in_flight_bytes -= task->req.bytes - new_bytes; bdrv_set_dirty_bitmap(task->s->copy_bitmap, - task->offset + new_bytes, task->bytes - new_bytes); + task->req.offset + new_bytes, + task->req.bytes - new_bytes); - task->bytes = new_bytes; - qemu_co_queue_restart_all(&task->wait_queue); + reqlist_shrink_req(&task->req, new_bytes); } static void coroutine_fn block_copy_task_end(BlockCopyTask *task, int ret) { QEMU_LOCK_GUARD(&task->s->lock); - task->s->in_flight_bytes -= task->bytes; + task->s->in_flight_bytes -= task->req.bytes; if (ret < 0) { - bdrv_set_dirty_bitmap(task->s->copy_bitmap, task->offset, task->bytes); + bdrv_set_dirty_bitmap(task->s->copy_bitmap, task->req.offset, + task->req.bytes); } - QLIST_REMOVE(task, list); if (task->s->progress) { progress_set_remaining(task->s->progress, bdrv_get_dirty_count(task->s->copy_bitmap) + task->s->in_flight_bytes); } - qemu_co_queue_restart_all(&task->wait_queue); + reqlist_remove_req(&task->req); } void block_copy_state_free(BlockCopyState *s) @@ -384,8 +341,10 @@ static int64_t block_copy_calculate_cluster_size(BlockDriverState *target, } BlockCopyState *block_copy_state_new(BdrvChild *source, BdrvChild *target, + const BdrvDirtyBitmap *bitmap, Error **errp) { + ERRP_GUARD(); BlockCopyState *s; int64_t cluster_size; BdrvDirtyBitmap *copy_bitmap; @@ -402,6 +361,17 @@ BlockCopyState *block_copy_state_new(BdrvChild *source, BdrvChild *target, return NULL; } bdrv_disable_dirty_bitmap(copy_bitmap); + if (bitmap) { + if (!bdrv_merge_dirty_bitmap(copy_bitmap, bitmap, NULL, errp)) { + error_prepend(errp, "Failed to merge bitmap '%s' to internal " + "copy-bitmap: ", bdrv_dirty_bitmap_name(bitmap)); + bdrv_release_dirty_bitmap(copy_bitmap); + return NULL; + } + } else { + bdrv_set_dirty_bitmap(copy_bitmap, 0, + bdrv_dirty_bitmap_size(copy_bitmap)); + } /* * If source is in backing chain of target assume that target is going to be @@ -437,7 +407,7 @@ BlockCopyState *block_copy_state_new(BdrvChild *source, BdrvChild *target, ratelimit_init(&s->rate_limit); qemu_co_mutex_init(&s->lock); - QLIST_INIT(&s->tasks); + QLIST_INIT(&s->reqs); QLIST_INIT(&s->calls); return s; @@ -470,7 +440,7 @@ static coroutine_fn int block_copy_task_run(AioTaskPool *pool, aio_task_pool_wait_slot(pool); if (aio_task_pool_status(pool) < 0) { - co_put_to_shres(task->s->mem, task->bytes); + co_put_to_shres(task->s->mem, task->req.bytes); block_copy_task_end(task, -ECANCELED); g_free(task); return -ECANCELED; @@ -583,7 +553,8 @@ static coroutine_fn int block_copy_task_entry(AioTask *task) BlockCopyMethod method = t->method; int ret; - ret = block_copy_do_copy(s, t->offset, t->bytes, &method, &error_is_read); + ret = block_copy_do_copy(s, t->req.offset, t->req.bytes, &method, + &error_is_read); WITH_QEMU_LOCK_GUARD(&s->lock) { if (s->method == t->method) { @@ -596,10 +567,10 @@ static coroutine_fn int block_copy_task_entry(AioTask *task) t->call_state->error_is_read = error_is_read; } } else if (s->progress) { - progress_work_done(s->progress, t->bytes); + progress_work_done(s->progress, t->req.bytes); } } - co_put_to_shres(s->mem, t->bytes); + co_put_to_shres(s->mem, t->req.bytes); block_copy_task_end(t, ret); return ret; @@ -679,6 +650,18 @@ static int block_copy_is_cluster_allocated(BlockCopyState *s, int64_t offset, } } +void block_copy_reset(BlockCopyState *s, int64_t offset, int64_t bytes) +{ + QEMU_LOCK_GUARD(&s->lock); + + bdrv_reset_dirty_bitmap(s->copy_bitmap, offset, bytes); + if (s->progress) { + progress_set_remaining(s->progress, + bdrv_get_dirty_count(s->copy_bitmap) + + s->in_flight_bytes); + } +} + /* * Reset bits in copy_bitmap starting at offset if they represent unallocated * data in the image. May reset subsequent contiguous bits. @@ -699,14 +682,7 @@ int64_t block_copy_reset_unallocated(BlockCopyState *s, bytes = clusters * s->cluster_size; if (!ret) { - qemu_co_mutex_lock(&s->lock); - bdrv_reset_dirty_bitmap(s->copy_bitmap, offset, bytes); - if (s->progress) { - progress_set_remaining(s->progress, - bdrv_get_dirty_count(s->copy_bitmap) + - s->in_flight_bytes); - } - qemu_co_mutex_unlock(&s->lock); + block_copy_reset(s, offset, bytes); } *count = bytes; @@ -753,22 +729,22 @@ block_copy_dirty_clusters(BlockCopyCallState *call_state) trace_block_copy_skip_range(s, offset, bytes); break; } - if (task->offset > offset) { - trace_block_copy_skip_range(s, offset, task->offset - offset); + if (task->req.offset > offset) { + trace_block_copy_skip_range(s, offset, task->req.offset - offset); } found_dirty = true; - ret = block_copy_block_status(s, task->offset, task->bytes, + ret = block_copy_block_status(s, task->req.offset, task->req.bytes, &status_bytes); assert(ret >= 0); /* never fail */ - if (status_bytes < task->bytes) { + if (status_bytes < task->req.bytes) { block_copy_task_shrink(task, status_bytes); } if (qatomic_read(&s->skip_unallocated) && !(ret & BDRV_BLOCK_ALLOCATED)) { block_copy_task_end(task, 0); - trace_block_copy_skip_range(s, task->offset, task->bytes); + trace_block_copy_skip_range(s, task->req.offset, task->req.bytes); offset = task_end(task); bytes = end - offset; g_free(task); @@ -789,11 +765,11 @@ block_copy_dirty_clusters(BlockCopyCallState *call_state) } } - ratelimit_calculate_delay(&s->rate_limit, task->bytes); + ratelimit_calculate_delay(&s->rate_limit, task->req.bytes); - trace_block_copy_process(s, task->offset); + trace_block_copy_process(s, task->req.offset); - co_get_from_shres(s->mem, task->bytes); + co_get_from_shres(s->mem, task->req.bytes); offset = task_end(task); bytes = end - offset; @@ -861,8 +837,8 @@ static int coroutine_fn block_copy_common(BlockCopyCallState *call_state) * Check that there is no task we still need to * wait to complete */ - ret = block_copy_wait_one(s, call_state->offset, - call_state->bytes); + ret = reqlist_wait_one(&s->reqs, call_state->offset, + call_state->bytes, &s->lock); if (ret == 0) { /* * No pending tasks, but check again the bitmap in this @@ -870,7 +846,7 @@ static int coroutine_fn block_copy_common(BlockCopyCallState *call_state) * between this and the critical section in * block_copy_dirty_clusters(). * - * block_copy_wait_one return value 0 also means that it + * reqlist_wait_one return value 0 also means that it * didn't release the lock. So, we are still in the same * critical section, not interrupted by any concurrent * access to state. diff --git a/block/copy-before-write.c b/block/copy-before-write.c index 80b7684dba..a8a06fdc09 100644 --- a/block/copy-before-write.c +++ b/block/copy-before-write.c @@ -33,10 +33,37 @@ #include "block/block-copy.h" #include "block/copy-before-write.h" +#include "block/reqlist.h" + +#include "qapi/qapi-visit-block-core.h" typedef struct BDRVCopyBeforeWriteState { BlockCopyState *bcs; BdrvChild *target; + + /* + * @lock: protects access to @access_bitmap, @done_bitmap and + * @frozen_read_reqs + */ + CoMutex lock; + + /* + * @access_bitmap: represents areas allowed for reading by fleecing user. + * Reading from non-dirty areas leads to -EACCES. + */ + BdrvDirtyBitmap *access_bitmap; + + /* + * @done_bitmap: represents areas that was successfully copied to @target by + * copy-before-write operations. + */ + BdrvDirtyBitmap *done_bitmap; + + /* + * @frozen_read_reqs: current read requests for fleecing user in bs->file + * node. These areas must not be rewritten by guest. + */ + BlockReqList frozen_read_reqs; } BDRVCopyBeforeWriteState; static coroutine_fn int cbw_co_preadv( @@ -46,10 +73,20 @@ static coroutine_fn int cbw_co_preadv( return bdrv_co_preadv(bs->file, offset, bytes, qiov, flags); } +/* + * Do copy-before-write operation. + * + * On failure guest request must be failed too. + * + * On success, we also wait for all in-flight fleecing read requests in source + * node, and it's guaranteed that after cbw_do_copy_before_write() successful + * return there are no such requests and they will never appear. + */ static coroutine_fn int cbw_do_copy_before_write(BlockDriverState *bs, uint64_t offset, uint64_t bytes, BdrvRequestFlags flags) { BDRVCopyBeforeWriteState *s = bs->opaque; + int ret; uint64_t off, end; int64_t cluster_size = block_copy_cluster_size(s->bcs); @@ -60,7 +97,17 @@ static coroutine_fn int cbw_do_copy_before_write(BlockDriverState *bs, off = QEMU_ALIGN_DOWN(offset, cluster_size); end = QEMU_ALIGN_UP(offset + bytes, cluster_size); - return block_copy(s->bcs, off, end - off, true); + ret = block_copy(s->bcs, off, end - off, true); + if (ret < 0) { + return ret; + } + + WITH_QEMU_LOCK_GUARD(&s->lock) { + bdrv_set_dirty_bitmap(s->done_bitmap, off, end - off); + reqlist_wait_all(&s->frozen_read_reqs, off, end - off, &s->lock); + } + + return 0; } static int coroutine_fn cbw_co_pdiscard(BlockDriverState *bs, @@ -108,6 +155,142 @@ static int coroutine_fn cbw_co_flush(BlockDriverState *bs) return bdrv_co_flush(bs->file->bs); } +/* + * If @offset not accessible - return NULL. + * + * Otherwise, set @pnum to some bytes that accessible from @file (@file is set + * to bs->file or to s->target). Return newly allocated BlockReq object that + * should be than passed to cbw_snapshot_read_unlock(). + * + * It's guaranteed that guest writes will not interact in the region until + * cbw_snapshot_read_unlock() called. + */ +static BlockReq *cbw_snapshot_read_lock(BlockDriverState *bs, + int64_t offset, int64_t bytes, + int64_t *pnum, BdrvChild **file) +{ + BDRVCopyBeforeWriteState *s = bs->opaque; + BlockReq *req = g_new(BlockReq, 1); + bool done; + + QEMU_LOCK_GUARD(&s->lock); + + if (bdrv_dirty_bitmap_next_zero(s->access_bitmap, offset, bytes) != -1) { + g_free(req); + return NULL; + } + + done = bdrv_dirty_bitmap_status(s->done_bitmap, offset, bytes, pnum); + if (done) { + /* + * Special invalid BlockReq, that is handled in + * cbw_snapshot_read_unlock(). We don't need to lock something to read + * from s->target. + */ + *req = (BlockReq) {.offset = -1, .bytes = -1}; + *file = s->target; + } else { + reqlist_init_req(&s->frozen_read_reqs, req, offset, bytes); + *file = bs->file; + } + + return req; +} + +static void cbw_snapshot_read_unlock(BlockDriverState *bs, BlockReq *req) +{ + BDRVCopyBeforeWriteState *s = bs->opaque; + + if (req->offset == -1 && req->bytes == -1) { + g_free(req); + return; + } + + QEMU_LOCK_GUARD(&s->lock); + + reqlist_remove_req(req); + g_free(req); +} + +static coroutine_fn int +cbw_co_preadv_snapshot(BlockDriverState *bs, int64_t offset, int64_t bytes, + QEMUIOVector *qiov, size_t qiov_offset) +{ + BlockReq *req; + BdrvChild *file; + int ret; + + /* TODO: upgrade to async loop using AioTask */ + while (bytes) { + int64_t cur_bytes; + + req = cbw_snapshot_read_lock(bs, offset, bytes, &cur_bytes, &file); + if (!req) { + return -EACCES; + } + + ret = bdrv_co_preadv_part(file, offset, cur_bytes, + qiov, qiov_offset, 0); + cbw_snapshot_read_unlock(bs, req); + if (ret < 0) { + return ret; + } + + bytes -= cur_bytes; + offset += cur_bytes; + qiov_offset += cur_bytes; + } + + return 0; +} + +static int coroutine_fn +cbw_co_snapshot_block_status(BlockDriverState *bs, + bool want_zero, int64_t offset, int64_t bytes, + int64_t *pnum, int64_t *map, + BlockDriverState **file) +{ + BDRVCopyBeforeWriteState *s = bs->opaque; + BlockReq *req; + int ret; + int64_t cur_bytes; + BdrvChild *child; + + req = cbw_snapshot_read_lock(bs, offset, bytes, &cur_bytes, &child); + if (!req) { + return -EACCES; + } + + ret = bdrv_block_status(child->bs, offset, cur_bytes, pnum, map, file); + if (child == s->target) { + /* + * We refer to s->target only for areas that we've written to it. + * And we can not report unallocated blocks in s->target: this will + * break generic block-status-above logic, that will go to + * copy-before-write filtered child in this case. + */ + assert(ret & BDRV_BLOCK_ALLOCATED); + } + + cbw_snapshot_read_unlock(bs, req); + + return ret; +} + +static int coroutine_fn cbw_co_pdiscard_snapshot(BlockDriverState *bs, + int64_t offset, int64_t bytes) +{ + BDRVCopyBeforeWriteState *s = bs->opaque; + + WITH_QEMU_LOCK_GUARD(&s->lock) { + bdrv_reset_dirty_bitmap(s->access_bitmap, offset, bytes); + } + + block_copy_reset(s->bcs, offset, bytes); + + return bdrv_co_pdiscard(s->target, offset, bytes); +} + static void cbw_refresh_filename(BlockDriverState *bs) { pstrcpy(bs->exact_filename, sizeof(bs->exact_filename), @@ -145,11 +328,54 @@ static void cbw_child_perm(BlockDriverState *bs, BdrvChild *c, } } +static bool cbw_parse_bitmap_option(QDict *options, BdrvDirtyBitmap **bitmap, + Error **errp) +{ + QDict *bitmap_qdict = NULL; + BlockDirtyBitmap *bmp_param = NULL; + Visitor *v = NULL; + bool ret = false; + + *bitmap = NULL; + + qdict_extract_subqdict(options, &bitmap_qdict, "bitmap."); + if (!qdict_size(bitmap_qdict)) { + ret = true; + goto out; + } + + v = qobject_input_visitor_new_flat_confused(bitmap_qdict, errp); + if (!v) { + goto out; + } + + visit_type_BlockDirtyBitmap(v, NULL, &bmp_param, errp); + if (!bmp_param) { + goto out; + } + + *bitmap = block_dirty_bitmap_lookup(bmp_param->node, bmp_param->name, NULL, + errp); + if (!*bitmap) { + goto out; + } + + ret = true; + +out: + qapi_free_BlockDirtyBitmap(bmp_param); + visit_free(v); + qobject_unref(bitmap_qdict); + + return ret; +} + static int cbw_open(BlockDriverState *bs, QDict *options, int flags, Error **errp) { BDRVCopyBeforeWriteState *s = bs->opaque; - BdrvDirtyBitmap *copy_bitmap; + BdrvDirtyBitmap *bitmap = NULL; + int64_t cluster_size; bs->file = bdrv_open_child(NULL, options, "file", bs, &child_of_bds, BDRV_CHILD_FILTERED | BDRV_CHILD_PRIMARY, @@ -164,6 +390,10 @@ static int cbw_open(BlockDriverState *bs, QDict *options, int flags, return -EINVAL; } + if (!cbw_parse_bitmap_option(options, &bitmap, errp)) { + return -EINVAL; + } + bs->total_sectors = bs->file->bs->total_sectors; bs->supported_write_flags = BDRV_REQ_WRITE_UNCHANGED | (BDRV_REQ_FUA & bs->file->bs->supported_write_flags); @@ -171,14 +401,32 @@ static int cbw_open(BlockDriverState *bs, QDict *options, int flags, ((BDRV_REQ_FUA | BDRV_REQ_MAY_UNMAP | BDRV_REQ_NO_FALLBACK) & bs->file->bs->supported_zero_flags); - s->bcs = block_copy_state_new(bs->file, s->target, errp); + s->bcs = block_copy_state_new(bs->file, s->target, bitmap, errp); if (!s->bcs) { error_prepend(errp, "Cannot create block-copy-state: "); return -EINVAL; } - copy_bitmap = block_copy_dirty_bitmap(s->bcs); - bdrv_set_dirty_bitmap(copy_bitmap, 0, bdrv_dirty_bitmap_size(copy_bitmap)); + cluster_size = block_copy_cluster_size(s->bcs); + + s->done_bitmap = bdrv_create_dirty_bitmap(bs, cluster_size, NULL, errp); + if (!s->done_bitmap) { + return -EINVAL; + } + bdrv_disable_dirty_bitmap(s->done_bitmap); + + /* s->access_bitmap starts equal to bcs bitmap */ + s->access_bitmap = bdrv_create_dirty_bitmap(bs, cluster_size, NULL, errp); + if (!s->access_bitmap) { + return -EINVAL; + } + bdrv_disable_dirty_bitmap(s->access_bitmap); + bdrv_dirty_bitmap_merge_internal(s->access_bitmap, + block_copy_dirty_bitmap(s->bcs), NULL, + true); + + qemu_co_mutex_init(&s->lock); + QLIST_INIT(&s->frozen_read_reqs); return 0; } @@ -187,6 +435,9 @@ static void cbw_close(BlockDriverState *bs) { BDRVCopyBeforeWriteState *s = bs->opaque; + bdrv_release_dirty_bitmap(s->access_bitmap); + bdrv_release_dirty_bitmap(s->done_bitmap); + block_copy_state_free(s->bcs); s->bcs = NULL; } @@ -204,6 +455,10 @@ BlockDriver bdrv_cbw_filter = { .bdrv_co_pdiscard = cbw_co_pdiscard, .bdrv_co_flush = cbw_co_flush, + .bdrv_co_preadv_snapshot = cbw_co_preadv_snapshot, + .bdrv_co_pdiscard_snapshot = cbw_co_pdiscard_snapshot, + .bdrv_co_snapshot_block_status = cbw_co_snapshot_block_status, + .bdrv_refresh_filename = cbw_refresh_filename, .bdrv_child_perm = cbw_child_perm, diff --git a/block/curl.c b/block/curl.c index 6a6cd72975..1e0f609579 100644 --- a/block/curl.c +++ b/block/curl.c @@ -458,38 +458,51 @@ static int curl_init_state(BDRVCURLState *s, CURLState *state) if (!state->curl) { return -EIO; } - curl_easy_setopt(state->curl, CURLOPT_URL, s->url); - curl_easy_setopt(state->curl, CURLOPT_SSL_VERIFYPEER, - (long) s->sslverify); - curl_easy_setopt(state->curl, CURLOPT_SSL_VERIFYHOST, - s->sslverify ? 2L : 0L); + if (curl_easy_setopt(state->curl, CURLOPT_URL, s->url) || + curl_easy_setopt(state->curl, CURLOPT_SSL_VERIFYPEER, + (long) s->sslverify) || + curl_easy_setopt(state->curl, CURLOPT_SSL_VERIFYHOST, + s->sslverify ? 2L : 0L)) { + goto err; + } if (s->cookie) { - curl_easy_setopt(state->curl, CURLOPT_COOKIE, s->cookie); + if (curl_easy_setopt(state->curl, CURLOPT_COOKIE, s->cookie)) { + goto err; + } + } + if (curl_easy_setopt(state->curl, CURLOPT_TIMEOUT, (long)s->timeout) || + curl_easy_setopt(state->curl, CURLOPT_WRITEFUNCTION, + (void *)curl_read_cb) || + curl_easy_setopt(state->curl, CURLOPT_WRITEDATA, (void *)state) || + curl_easy_setopt(state->curl, CURLOPT_PRIVATE, (void *)state) || + curl_easy_setopt(state->curl, CURLOPT_AUTOREFERER, 1) || + curl_easy_setopt(state->curl, CURLOPT_FOLLOWLOCATION, 1) || + curl_easy_setopt(state->curl, CURLOPT_NOSIGNAL, 1) || + curl_easy_setopt(state->curl, CURLOPT_ERRORBUFFER, state->errmsg) || + curl_easy_setopt(state->curl, CURLOPT_FAILONERROR, 1)) { + goto err; } - curl_easy_setopt(state->curl, CURLOPT_TIMEOUT, (long)s->timeout); - curl_easy_setopt(state->curl, CURLOPT_WRITEFUNCTION, - (void *)curl_read_cb); - curl_easy_setopt(state->curl, CURLOPT_WRITEDATA, (void *)state); - curl_easy_setopt(state->curl, CURLOPT_PRIVATE, (void *)state); - curl_easy_setopt(state->curl, CURLOPT_AUTOREFERER, 1); - curl_easy_setopt(state->curl, CURLOPT_FOLLOWLOCATION, 1); - curl_easy_setopt(state->curl, CURLOPT_NOSIGNAL, 1); - curl_easy_setopt(state->curl, CURLOPT_ERRORBUFFER, state->errmsg); - curl_easy_setopt(state->curl, CURLOPT_FAILONERROR, 1); - if (s->username) { - curl_easy_setopt(state->curl, CURLOPT_USERNAME, s->username); + if (curl_easy_setopt(state->curl, CURLOPT_USERNAME, s->username)) { + goto err; + } } if (s->password) { - curl_easy_setopt(state->curl, CURLOPT_PASSWORD, s->password); + if (curl_easy_setopt(state->curl, CURLOPT_PASSWORD, s->password)) { + goto err; + } } if (s->proxyusername) { - curl_easy_setopt(state->curl, - CURLOPT_PROXYUSERNAME, s->proxyusername); + if (curl_easy_setopt(state->curl, + CURLOPT_PROXYUSERNAME, s->proxyusername)) { + goto err; + } } if (s->proxypassword) { - curl_easy_setopt(state->curl, - CURLOPT_PROXYPASSWORD, s->proxypassword); + if (curl_easy_setopt(state->curl, + CURLOPT_PROXYPASSWORD, s->proxypassword)) { + goto err; + } } /* Restrict supported protocols to avoid security issues in the more @@ -499,18 +512,27 @@ static int curl_init_state(BDRVCURLState *s, CURLState *state) * Restricting protocols is only supported from 7.19.4 upwards. */ #if LIBCURL_VERSION_NUM >= 0x071304 - curl_easy_setopt(state->curl, CURLOPT_PROTOCOLS, PROTOCOLS); - curl_easy_setopt(state->curl, CURLOPT_REDIR_PROTOCOLS, PROTOCOLS); + if (curl_easy_setopt(state->curl, CURLOPT_PROTOCOLS, PROTOCOLS) || + curl_easy_setopt(state->curl, CURLOPT_REDIR_PROTOCOLS, PROTOCOLS)) { + goto err; + } #endif #ifdef DEBUG_VERBOSE - curl_easy_setopt(state->curl, CURLOPT_VERBOSE, 1); + if (curl_easy_setopt(state->curl, CURLOPT_VERBOSE, 1)) { + goto err; + } #endif } state->s = s; return 0; + +err: + curl_easy_cleanup(state->curl); + state->curl = NULL; + return -EIO; } /* Called with s->mutex held. */ @@ -759,14 +781,19 @@ static int curl_open(BlockDriverState *bs, QDict *options, int flags, // Get file size if (curl_init_state(s, state) < 0) { + pstrcpy(state->errmsg, CURL_ERROR_SIZE, + "curl library initialization failed."); goto out; } s->accept_range = false; - curl_easy_setopt(state->curl, CURLOPT_NOBODY, 1); - curl_easy_setopt(state->curl, CURLOPT_HEADERFUNCTION, - curl_header_cb); - curl_easy_setopt(state->curl, CURLOPT_HEADERDATA, s); + if (curl_easy_setopt(state->curl, CURLOPT_NOBODY, 1) || + curl_easy_setopt(state->curl, CURLOPT_HEADERFUNCTION, curl_header_cb) || + curl_easy_setopt(state->curl, CURLOPT_HEADERDATA, s)) { + pstrcpy(state->errmsg, CURL_ERROR_SIZE, + "curl library initialization failed."); + goto out; + } if (curl_easy_perform(state->curl)) goto out; if (curl_easy_getinfo(state->curl, CURLINFO_CONTENT_LENGTH_DOWNLOAD, &d)) { @@ -879,9 +906,8 @@ static void curl_setup_preadv(BlockDriverState *bs, CURLAIOCB *acb) snprintf(state->range, 127, "%" PRIu64 "-%" PRIu64, start, end); trace_curl_setup_preadv(acb->bytes, start, state->range); - curl_easy_setopt(state->curl, CURLOPT_RANGE, state->range); - - if (curl_multi_add_handle(s->multi, state->curl) != CURLM_OK) { + if (curl_easy_setopt(state->curl, CURLOPT_RANGE, state->range) || + curl_multi_add_handle(s->multi, state->curl) != CURLM_OK) { state->acb[0] = NULL; acb->ret = -EIO; diff --git a/block/dirty-bitmap.c b/block/dirty-bitmap.c index 0334b85805..da1b91166f 100644 --- a/block/dirty-bitmap.c +++ b/block/dirty-bitmap.c @@ -879,16 +879,25 @@ bool bdrv_dirty_bitmap_next_dirty_area(BdrvDirtyBitmap *bitmap, dirty_start, dirty_count); } +bool bdrv_dirty_bitmap_status(BdrvDirtyBitmap *bitmap, int64_t offset, + int64_t bytes, int64_t *count) +{ + return hbitmap_status(bitmap->bitmap, offset, bytes, count); +} + /** * bdrv_merge_dirty_bitmap: merge src into dest. * Ensures permissions on bitmaps are reasonable; use for public API. * * @backup: If provided, make a copy of dest here prior to merge. + * + * Returns true on success, false on failure. In case of failure bitmaps are + * untouched. */ -void bdrv_merge_dirty_bitmap(BdrvDirtyBitmap *dest, const BdrvDirtyBitmap *src, +bool bdrv_merge_dirty_bitmap(BdrvDirtyBitmap *dest, const BdrvDirtyBitmap *src, HBitmap **backup, Error **errp) { - bool ret; + bool ret = false; bdrv_dirty_bitmaps_lock(dest->bs); if (src->bs != dest->bs) { @@ -916,6 +925,8 @@ out: if (src->bs != dest->bs) { bdrv_dirty_bitmaps_unlock(src->bs); } + + return ret; } /** diff --git a/block/io.c b/block/io.c index efc011ce65..f0c8da6b9f 100644 --- a/block/io.c +++ b/block/io.c @@ -2203,6 +2203,7 @@ static int coroutine_fn bdrv_co_do_zero_pwritev(BdrvChild *child, padding = bdrv_init_padding(bs, offset, bytes, &pad); if (padding) { + assert(!(flags & BDRV_REQ_NO_WAIT)); bdrv_make_request_serialising(req, align); bdrv_padding_rmw_read(child, req, &pad, true); @@ -2339,6 +2340,7 @@ int coroutine_fn bdrv_co_pwritev_part(BdrvChild *child, * serialize the request to prevent interactions of the * widened region with other transactions. */ + assert(!(flags & BDRV_REQ_NO_WAIT)); bdrv_make_request_serialising(&req, align); bdrv_padding_rmw_read(child, &req, &pad, false); } @@ -3387,6 +3389,8 @@ static int coroutine_fn bdrv_co_copy_range_internal( /* TODO We can support BDRV_REQ_NO_FALLBACK here */ assert(!(read_flags & BDRV_REQ_NO_FALLBACK)); assert(!(write_flags & BDRV_REQ_NO_FALLBACK)); + assert(!(read_flags & BDRV_REQ_NO_WAIT)); + assert(!(write_flags & BDRV_REQ_NO_WAIT)); if (!dst || !dst->bs || !bdrv_is_inserted(dst->bs)) { return -ENOMEDIUM; @@ -3650,3 +3654,75 @@ void bdrv_cancel_in_flight(BlockDriverState *bs) bs->drv->bdrv_cancel_in_flight(bs); } } + +int coroutine_fn +bdrv_co_preadv_snapshot(BdrvChild *child, int64_t offset, int64_t bytes, + QEMUIOVector *qiov, size_t qiov_offset) +{ + BlockDriverState *bs = child->bs; + BlockDriver *drv = bs->drv; + int ret; + IO_CODE(); + + if (!drv) { + return -ENOMEDIUM; + } + + if (!drv->bdrv_co_preadv_snapshot) { + return -ENOTSUP; + } + + bdrv_inc_in_flight(bs); + ret = drv->bdrv_co_preadv_snapshot(bs, offset, bytes, qiov, qiov_offset); + bdrv_dec_in_flight(bs); + + return ret; +} + +int coroutine_fn +bdrv_co_snapshot_block_status(BlockDriverState *bs, + bool want_zero, int64_t offset, int64_t bytes, + int64_t *pnum, int64_t *map, + BlockDriverState **file) +{ + BlockDriver *drv = bs->drv; + int ret; + IO_CODE(); + + if (!drv) { + return -ENOMEDIUM; + } + + if (!drv->bdrv_co_snapshot_block_status) { + return -ENOTSUP; + } + + bdrv_inc_in_flight(bs); + ret = drv->bdrv_co_snapshot_block_status(bs, want_zero, offset, bytes, + pnum, map, file); + bdrv_dec_in_flight(bs); + + return ret; +} + +int coroutine_fn +bdrv_co_pdiscard_snapshot(BlockDriverState *bs, int64_t offset, int64_t bytes) +{ + BlockDriver *drv = bs->drv; + int ret; + IO_CODE(); + + if (!drv) { + return -ENOMEDIUM; + } + + if (!drv->bdrv_co_pdiscard_snapshot) { + return -ENOTSUP; + } + + bdrv_inc_in_flight(bs); + ret = drv->bdrv_co_pdiscard_snapshot(bs, offset, bytes); + bdrv_dec_in_flight(bs); + + return ret; +} diff --git a/block/meson.build b/block/meson.build index e42bcb58d5..0b2a60c99b 100644 --- a/block/meson.build +++ b/block/meson.build @@ -32,7 +32,9 @@ block_ss.add(files( 'qcow2.c', 'quorum.c', 'raw-format.c', + 'reqlist.c', 'snapshot.c', + 'snapshot-access.c', 'throttle-groups.c', 'throttle.c', 'vhdx-endian.c', diff --git a/block/monitor/bitmap-qmp-cmds.c b/block/monitor/bitmap-qmp-cmds.c index 972e8a0afc..8e35616c2e 100644 --- a/block/monitor/bitmap-qmp-cmds.c +++ b/block/monitor/bitmap-qmp-cmds.c @@ -263,7 +263,6 @@ BdrvDirtyBitmap *block_dirty_bitmap_merge(const char *node, const char *target, BlockDriverState *bs; BdrvDirtyBitmap *dst, *src, *anon; BlockDirtyBitmapMergeSourceList *lst; - Error *local_err = NULL; GLOBAL_STATE_CODE(); @@ -303,9 +302,7 @@ BdrvDirtyBitmap *block_dirty_bitmap_merge(const char *node, const char *target, abort(); } - bdrv_merge_dirty_bitmap(anon, src, NULL, &local_err); - if (local_err) { - error_propagate(errp, local_err); + if (!bdrv_merge_dirty_bitmap(anon, src, NULL, errp)) { dst = NULL; goto out; } diff --git a/block/preallocate.c b/block/preallocate.c index 1d4233f730..e15cb8c74a 100644 --- a/block/preallocate.c +++ b/block/preallocate.c @@ -276,6 +276,10 @@ static bool coroutine_fn handle_write(BlockDriverState *bs, int64_t offset, int64_t end = offset + bytes; int64_t prealloc_start, prealloc_end; int ret; + uint32_t file_align = bs->file->bs->bl.request_alignment; + uint32_t prealloc_align = MAX(s->opts.prealloc_align, file_align); + + assert(QEMU_IS_ALIGNED(prealloc_align, file_align)); if (!has_prealloc_perms(bs)) { /* We don't have state neither should try to recover it */ @@ -320,9 +324,14 @@ static bool coroutine_fn handle_write(BlockDriverState *bs, int64_t offset, /* Now we want new preallocation, as request writes beyond s->file_end. */ - prealloc_start = want_merge_zero ? MIN(offset, s->file_end) : s->file_end; - prealloc_end = QEMU_ALIGN_UP(end + s->opts.prealloc_size, - s->opts.prealloc_align); + prealloc_start = QEMU_ALIGN_UP( + want_merge_zero ? MIN(offset, s->file_end) : s->file_end, + file_align); + prealloc_end = QEMU_ALIGN_UP( + MAX(prealloc_start, end) + s->opts.prealloc_size, + prealloc_align); + + want_merge_zero = want_merge_zero && (prealloc_start <= offset); ret = bdrv_co_pwrite_zeroes( bs->file, prealloc_start, prealloc_end - prealloc_start, diff --git a/block/reqlist.c b/block/reqlist.c new file mode 100644 index 0000000000..08cb57cfa4 --- /dev/null +++ b/block/reqlist.c @@ -0,0 +1,85 @@ +/* + * reqlist API + * + * Copyright (C) 2013 Proxmox Server Solutions + * Copyright (c) 2021 Virtuozzo International GmbH. + * + * Authors: + * Dietmar Maurer (dietmar@proxmox.com) + * Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com> + * + * 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 "qemu/osdep.h" +#include "qemu/range.h" + +#include "block/reqlist.h" + +void reqlist_init_req(BlockReqList *reqs, BlockReq *req, int64_t offset, + int64_t bytes) +{ + assert(!reqlist_find_conflict(reqs, offset, bytes)); + + *req = (BlockReq) { + .offset = offset, + .bytes = bytes, + }; + qemu_co_queue_init(&req->wait_queue); + QLIST_INSERT_HEAD(reqs, req, list); +} + +BlockReq *reqlist_find_conflict(BlockReqList *reqs, int64_t offset, + int64_t bytes) +{ + BlockReq *r; + + QLIST_FOREACH(r, reqs, list) { + if (ranges_overlap(offset, bytes, r->offset, r->bytes)) { + return r; + } + } + + return NULL; +} + +bool coroutine_fn reqlist_wait_one(BlockReqList *reqs, int64_t offset, + int64_t bytes, CoMutex *lock) +{ + BlockReq *r = reqlist_find_conflict(reqs, offset, bytes); + + if (!r) { + return false; + } + + qemu_co_queue_wait(&r->wait_queue, lock); + + return true; +} + +void coroutine_fn reqlist_wait_all(BlockReqList *reqs, int64_t offset, + int64_t bytes, CoMutex *lock) +{ + while (reqlist_wait_one(reqs, offset, bytes, lock)) { + /* continue */ + } +} + +void coroutine_fn reqlist_shrink_req(BlockReq *req, int64_t new_bytes) +{ + if (new_bytes == req->bytes) { + return; + } + + assert(new_bytes > 0 && new_bytes < req->bytes); + + req->bytes = new_bytes; + qemu_co_queue_restart_all(&req->wait_queue); +} + +void coroutine_fn reqlist_remove_req(BlockReq *req) +{ + QLIST_REMOVE(req, list); + qemu_co_queue_restart_all(&req->wait_queue); +} diff --git a/block/snapshot-access.c b/block/snapshot-access.c new file mode 100644 index 0000000000..77b87c1946 --- /dev/null +++ b/block/snapshot-access.c @@ -0,0 +1,132 @@ +/* + * snapshot_access block driver + * + * Copyright (c) 2022 Virtuozzo International GmbH. + * + * Author: + * Sementsov-Ogievskiy Vladimir <vsementsov@virtuozzo.com> + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with this program. If not, see <http://www.gnu.org/licenses/>. + */ + +#include "qemu/osdep.h" + +#include "sysemu/block-backend.h" +#include "qemu/cutils.h" +#include "block/block_int.h" + +static coroutine_fn int +snapshot_access_co_preadv_part(BlockDriverState *bs, + int64_t offset, int64_t bytes, + QEMUIOVector *qiov, size_t qiov_offset, + BdrvRequestFlags flags) +{ + if (flags) { + return -ENOTSUP; + } + + return bdrv_co_preadv_snapshot(bs->file, offset, bytes, qiov, qiov_offset); +} + +static int coroutine_fn +snapshot_access_co_block_status(BlockDriverState *bs, + bool want_zero, int64_t offset, + int64_t bytes, int64_t *pnum, + int64_t *map, BlockDriverState **file) +{ + return bdrv_co_snapshot_block_status(bs->file->bs, want_zero, offset, + bytes, pnum, map, file); +} + +static int coroutine_fn snapshot_access_co_pdiscard(BlockDriverState *bs, + int64_t offset, int64_t bytes) +{ + return bdrv_co_pdiscard_snapshot(bs->file->bs, offset, bytes); +} + +static int coroutine_fn +snapshot_access_co_pwrite_zeroes(BlockDriverState *bs, + int64_t offset, int64_t bytes, + BdrvRequestFlags flags) +{ + return -ENOTSUP; +} + +static coroutine_fn int +snapshot_access_co_pwritev_part(BlockDriverState *bs, + int64_t offset, int64_t bytes, + QEMUIOVector *qiov, size_t qiov_offset, + BdrvRequestFlags flags) +{ + return -ENOTSUP; +} + + +static void snapshot_access_refresh_filename(BlockDriverState *bs) +{ + pstrcpy(bs->exact_filename, sizeof(bs->exact_filename), + bs->file->bs->filename); +} + +static int snapshot_access_open(BlockDriverState *bs, QDict *options, int flags, + Error **errp) +{ + bs->file = bdrv_open_child(NULL, options, "file", bs, &child_of_bds, + BDRV_CHILD_DATA | BDRV_CHILD_PRIMARY, + false, errp); + if (!bs->file) { + return -EINVAL; + } + + bs->total_sectors = bs->file->bs->total_sectors; + + return 0; +} + +static void snapshot_access_child_perm(BlockDriverState *bs, BdrvChild *c, + BdrvChildRole role, + BlockReopenQueue *reopen_queue, + uint64_t perm, uint64_t shared, + uint64_t *nperm, uint64_t *nshared) +{ + /* + * Currently, we don't need any permissions. If bs->file provides + * snapshot-access API, we can use it. + */ + *nperm = 0; + *nshared = BLK_PERM_ALL; +} + +BlockDriver bdrv_snapshot_access_drv = { + .format_name = "snapshot-access", + + .bdrv_open = snapshot_access_open, + + .bdrv_co_preadv_part = snapshot_access_co_preadv_part, + .bdrv_co_pwritev_part = snapshot_access_co_pwritev_part, + .bdrv_co_pwrite_zeroes = snapshot_access_co_pwrite_zeroes, + .bdrv_co_pdiscard = snapshot_access_co_pdiscard, + .bdrv_co_block_status = snapshot_access_co_block_status, + + .bdrv_refresh_filename = snapshot_access_refresh_filename, + + .bdrv_child_perm = snapshot_access_child_perm, +}; + +static void snapshot_access_init(void) +{ + bdrv_register(&bdrv_snapshot_access_drv); +} + +block_init(snapshot_access_init); diff --git a/hw/ide/core.c b/hw/ide/core.c index 33463d9b8f..d667d0b55e 100644 --- a/hw/ide/core.c +++ b/hw/ide/core.c @@ -434,12 +434,16 @@ static const AIOCBInfo trim_aiocb_info = { static void ide_trim_bh_cb(void *opaque) { TrimAIOCB *iocb = opaque; + BlockBackend *blk = iocb->s->blk; iocb->common.cb(iocb->common.opaque, iocb->ret); qemu_bh_delete(iocb->bh); iocb->bh = NULL; qemu_aio_unref(iocb); + + /* Paired with an increment in ide_issue_trim() */ + blk_dec_in_flight(blk); } static void ide_issue_trim_cb(void *opaque, int ret) @@ -509,6 +513,9 @@ BlockAIOCB *ide_issue_trim( IDEState *s = opaque; TrimAIOCB *iocb; + /* Paired with a decrement in ide_trim_bh_cb() */ + blk_inc_in_flight(s->blk); + iocb = blk_aio_get(&trim_aiocb_info, s->blk, cb, cb_opaque); iocb->s = s; iocb->bh = qemu_bh_new(ide_trim_bh_cb, iocb); diff --git a/include/block/block-common.h b/include/block/block-common.h index 0c5dc4a86a..fdb7306e78 100644 --- a/include/block/block-common.h +++ b/include/block/block-common.h @@ -112,7 +112,8 @@ typedef enum { /* * If we need to wait for other requests, just fail immediately. Used - * only together with BDRV_REQ_SERIALISING. + * only together with BDRV_REQ_SERIALISING. Used only with requests aligned + * to request_alignment (corresponding assertions are in block/io.c). */ BDRV_REQ_NO_WAIT = 0x400, diff --git a/include/block/block-copy.h b/include/block/block-copy.h index 99370fa38b..68bbd344b2 100644 --- a/include/block/block-copy.h +++ b/include/block/block-copy.h @@ -25,6 +25,7 @@ typedef struct BlockCopyState BlockCopyState; typedef struct BlockCopyCallState BlockCopyCallState; BlockCopyState *block_copy_state_new(BdrvChild *source, BdrvChild *target, + const BdrvDirtyBitmap *bitmap, Error **errp); /* Function should be called prior any actual copy request */ @@ -34,6 +35,7 @@ void block_copy_set_progress_meter(BlockCopyState *s, ProgressMeter *pm); void block_copy_state_free(BlockCopyState *s); +void block_copy_reset(BlockCopyState *s, int64_t offset, int64_t bytes); int64_t block_copy_reset_unallocated(BlockCopyState *s, int64_t offset, int64_t *count); diff --git a/include/block/block_int-common.h b/include/block/block_int-common.h index 5a04c778e4..8947abab76 100644 --- a/include/block/block_int-common.h +++ b/include/block/block_int-common.h @@ -598,6 +598,30 @@ struct BlockDriver { int64_t *map, BlockDriverState **file); /* + * Snapshot-access API. + * + * Block-driver may provide snapshot-access API: special functions to access + * some internal "snapshot". The functions are similar with normal + * read/block_status/discard handler, but don't have any specific handling + * in generic block-layer: no serializing, no alignment, no tracked + * requests. So, block-driver that realizes these APIs is fully responsible + * for synchronization between snapshot-access API and normal IO requests. + * + * TODO: To be able to support qcow2's internal snapshots, this API will + * need to be extended to: + * - be able to select a specific snapshot + * - receive the snapshot's actual length (which may differ from bs's + * length) + */ + int coroutine_fn (*bdrv_co_preadv_snapshot)(BlockDriverState *bs, + int64_t offset, int64_t bytes, QEMUIOVector *qiov, size_t qiov_offset); + int coroutine_fn (*bdrv_co_snapshot_block_status)(BlockDriverState *bs, + bool want_zero, int64_t offset, int64_t bytes, int64_t *pnum, + int64_t *map, BlockDriverState **file); + int coroutine_fn (*bdrv_co_pdiscard_snapshot)(BlockDriverState *bs, + int64_t offset, int64_t bytes); + + /* * Invalidate any cached meta-data. */ void coroutine_fn (*bdrv_co_invalidate_cache)(BlockDriverState *bs, diff --git a/include/block/block_int-io.h b/include/block/block_int-io.h index 3da5f01c42..bb454200e5 100644 --- a/include/block/block_int-io.h +++ b/include/block/block_int-io.h @@ -33,6 +33,15 @@ * the I/O API. */ +int coroutine_fn bdrv_co_preadv_snapshot(BdrvChild *child, + int64_t offset, int64_t bytes, QEMUIOVector *qiov, size_t qiov_offset); +int coroutine_fn bdrv_co_snapshot_block_status(BlockDriverState *bs, + bool want_zero, int64_t offset, int64_t bytes, int64_t *pnum, + int64_t *map, BlockDriverState **file); +int coroutine_fn bdrv_co_pdiscard_snapshot(BlockDriverState *bs, + int64_t offset, int64_t bytes); + + int coroutine_fn bdrv_co_preadv(BdrvChild *child, int64_t offset, int64_t bytes, QEMUIOVector *qiov, BdrvRequestFlags flags); diff --git a/include/block/dirty-bitmap.h b/include/block/dirty-bitmap.h index 40950ae3d5..6528336c4c 100644 --- a/include/block/dirty-bitmap.h +++ b/include/block/dirty-bitmap.h @@ -77,7 +77,7 @@ void bdrv_dirty_bitmap_set_persistence(BdrvDirtyBitmap *bitmap, bool persistent); void bdrv_dirty_bitmap_set_inconsistent(BdrvDirtyBitmap *bitmap); void bdrv_dirty_bitmap_set_busy(BdrvDirtyBitmap *bitmap, bool busy); -void bdrv_merge_dirty_bitmap(BdrvDirtyBitmap *dest, const BdrvDirtyBitmap *src, +bool bdrv_merge_dirty_bitmap(BdrvDirtyBitmap *dest, const BdrvDirtyBitmap *src, HBitmap **backup, Error **errp); void bdrv_dirty_bitmap_skip_store(BdrvDirtyBitmap *bitmap, bool skip); bool bdrv_dirty_bitmap_get(BdrvDirtyBitmap *bitmap, int64_t offset); @@ -115,6 +115,8 @@ int64_t bdrv_dirty_bitmap_next_zero(BdrvDirtyBitmap *bitmap, int64_t offset, bool bdrv_dirty_bitmap_next_dirty_area(BdrvDirtyBitmap *bitmap, int64_t start, int64_t end, int64_t max_dirty_count, int64_t *dirty_start, int64_t *dirty_count); +bool bdrv_dirty_bitmap_status(BdrvDirtyBitmap *bitmap, int64_t offset, + int64_t bytes, int64_t *count); BdrvDirtyBitmap *bdrv_reclaim_dirty_bitmap_locked(BdrvDirtyBitmap *bitmap, Error **errp); diff --git a/include/block/reqlist.h b/include/block/reqlist.h new file mode 100644 index 0000000000..5253497bae --- /dev/null +++ b/include/block/reqlist.h @@ -0,0 +1,75 @@ +/* + * reqlist API + * + * Copyright (C) 2013 Proxmox Server Solutions + * Copyright (c) 2021 Virtuozzo International GmbH. + * + * Authors: + * Dietmar Maurer (dietmar@proxmox.com) + * Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com> + * + * This work is licensed under the terms of the GNU GPL, version 2 or later. + * See the COPYING file in the top-level directory. + */ + +#ifndef REQLIST_H +#define REQLIST_H + +#include "qemu/coroutine.h" + +/* + * The API is not thread-safe and shouldn't be. The struct is public to be part + * of other structures and protected by third-party locks, see + * block/block-copy.c for example. + */ + +typedef struct BlockReq { + int64_t offset; + int64_t bytes; + + CoQueue wait_queue; /* coroutines blocked on this req */ + QLIST_ENTRY(BlockReq) list; +} BlockReq; + +typedef QLIST_HEAD(, BlockReq) BlockReqList; + +/* + * Initialize new request and add it to the list. Caller must be sure that + * there are no conflicting requests in the list. + */ +void reqlist_init_req(BlockReqList *reqs, BlockReq *req, int64_t offset, + int64_t bytes); +/* Search for request in the list intersecting with @offset/@bytes area. */ +BlockReq *reqlist_find_conflict(BlockReqList *reqs, int64_t offset, + int64_t bytes); + +/* + * If there are no intersecting requests return false. Otherwise, wait for the + * first found intersecting request to finish and return true. + * + * @lock is passed to qemu_co_queue_wait() + * False return value proves that lock was released at no point. + */ +bool coroutine_fn reqlist_wait_one(BlockReqList *reqs, int64_t offset, + int64_t bytes, CoMutex *lock); + +/* + * Wait for all intersecting requests. It just calls reqlist_wait_one() in a + * loop, caller is responsible to stop producing new requests in this region + * in parallel, otherwise reqlist_wait_all() may never return. + */ +void coroutine_fn reqlist_wait_all(BlockReqList *reqs, int64_t offset, + int64_t bytes, CoMutex *lock); + +/* + * Shrink request and wake all waiting coroutines (maybe some of them are not + * intersecting with shrunk request). + */ +void coroutine_fn reqlist_shrink_req(BlockReq *req, int64_t new_bytes); + +/* + * Remove request and wake all waiting coroutines. Do not release any memory. + */ +void coroutine_fn reqlist_remove_req(BlockReq *req); + +#endif /* REQLIST_H */ diff --git a/include/qemu/hbitmap.h b/include/qemu/hbitmap.h index 5e71b6d6f7..5bd986aa44 100644 --- a/include/qemu/hbitmap.h +++ b/include/qemu/hbitmap.h @@ -340,6 +340,18 @@ bool hbitmap_next_dirty_area(const HBitmap *hb, int64_t start, int64_t end, int64_t max_dirty_count, int64_t *dirty_start, int64_t *dirty_count); +/* + * bdrv_dirty_bitmap_status: + * @hb: The HBitmap to operate on + * @start: The bit to start from + * @count: Number of bits to proceed + * @pnum: Out-parameter. How many bits has same value starting from @start + * + * Returns true if bitmap is dirty at @start, false otherwise. + */ +bool hbitmap_status(const HBitmap *hb, int64_t start, int64_t count, + int64_t *pnum); + /** * hbitmap_iter_next: * @hbi: HBitmapIter to operate on. diff --git a/qapi/block-core.json b/qapi/block-core.json index 9a5a3641d0..f13b5ff942 100644 --- a/qapi/block-core.json +++ b/qapi/block-core.json @@ -2914,13 +2914,14 @@ # @blkreplay: Since 4.2 # @compress: Since 5.0 # @copy-before-write: Since 6.2 +# @snapshot-access: Since 7.0 # # Since: 2.9 ## { 'enum': 'BlockdevDriver', 'data': [ 'blkdebug', 'blklogwrites', 'blkreplay', 'blkverify', 'bochs', 'cloop', 'compress', 'copy-before-write', 'copy-on-read', 'dmg', - 'file', 'ftp', 'ftps', 'gluster', + 'file', 'snapshot-access', 'ftp', 'ftps', 'gluster', {'name': 'host_cdrom', 'if': 'HAVE_HOST_BLOCK_DEVICE' }, {'name': 'host_device', 'if': 'HAVE_HOST_BLOCK_DEVICE' }, 'http', 'https', 'iscsi', @@ -4171,11 +4172,19 @@ # # @target: The target for copy-before-write operations. # +# @bitmap: If specified, copy-before-write filter will do +# copy-before-write operations only for dirty regions of the +# bitmap. Bitmap size must be equal to length of file and +# target child of the filter. Note also, that bitmap is used +# only to initialize internal bitmap of the process, so further +# modifications (or removing) of specified bitmap doesn't +# influence the filter. (Since 7.0) +# # Since: 6.2 ## { 'struct': 'BlockdevOptionsCbw', 'base': 'BlockdevOptionsGenericFormat', - 'data': { 'target': 'BlockdevRef' } } + 'data': { 'target': 'BlockdevRef', '*bitmap': 'BlockDirtyBitmap' } } ## # @BlockdevOptions: @@ -4259,6 +4268,7 @@ 'rbd': 'BlockdevOptionsRbd', 'replication': { 'type': 'BlockdevOptionsReplication', 'if': 'CONFIG_REPLICATION' }, + 'snapshot-access': 'BlockdevOptionsGenericFormat', 'ssh': 'BlockdevOptionsSsh', 'throttle': 'BlockdevOptionsThrottle', 'vdi': 'BlockdevOptionsGenericFormat', diff --git a/tests/qemu-iotests/040 b/tests/qemu-iotests/040 index 6af5ab9e76..0e1cfd7e49 100755 --- a/tests/qemu-iotests/040 +++ b/tests/qemu-iotests/040 @@ -744,6 +744,7 @@ class TestCommitWithFilters(iotests.QMPTestCase): pattern_file) self.assertFalse('Pattern verification failed' in result) + @iotests.skip_if_unsupported(['throttle']) def setUp(self): qemu_img('create', '-f', iotests.imgfmt, self.img0, '64M') qemu_img('create', '-f', iotests.imgfmt, self.img1, '64M') diff --git a/tests/qemu-iotests/257.out b/tests/qemu-iotests/257.out index 50cbd8e882..aa76131ca9 100644 --- a/tests/qemu-iotests/257.out +++ b/tests/qemu-iotests/257.out @@ -106,6 +106,22 @@ write -P0x67 0x3fe0000 0x20000 {"return": ""} { "bitmaps": { + "backup-top": [ + { + "busy": false, + "count": 67108864, + "granularity": 65536, + "persistent": false, + "recording": false + }, + { + "busy": false, + "count": 458752, + "granularity": 65536, + "persistent": false, + "recording": false + } + ], "drive0": [ { "busy": false, @@ -566,6 +582,22 @@ write -P0x67 0x3fe0000 0x20000 {"return": ""} { "bitmaps": { + "backup-top": [ + { + "busy": false, + "count": 67108864, + "granularity": 65536, + "persistent": false, + "recording": false + }, + { + "busy": false, + "count": 458752, + "granularity": 65536, + "persistent": false, + "recording": false + } + ], "drive0": [ { "busy": false, @@ -819,6 +851,22 @@ write -P0x67 0x3fe0000 0x20000 {"return": ""} { "bitmaps": { + "backup-top": [ + { + "busy": false, + "count": 67108864, + "granularity": 65536, + "persistent": false, + "recording": false + }, + { + "busy": false, + "count": 458752, + "granularity": 65536, + "persistent": false, + "recording": false + } + ], "drive0": [ { "busy": false, @@ -1279,6 +1327,22 @@ write -P0x67 0x3fe0000 0x20000 {"return": ""} { "bitmaps": { + "backup-top": [ + { + "busy": false, + "count": 67108864, + "granularity": 65536, + "persistent": false, + "recording": false + }, + { + "busy": false, + "count": 458752, + "granularity": 65536, + "persistent": false, + "recording": false + } + ], "drive0": [ { "busy": false, @@ -1532,6 +1596,22 @@ write -P0x67 0x3fe0000 0x20000 {"return": ""} { "bitmaps": { + "backup-top": [ + { + "busy": false, + "count": 67108864, + "granularity": 65536, + "persistent": false, + "recording": false + }, + { + "busy": false, + "count": 458752, + "granularity": 65536, + "persistent": false, + "recording": false + } + ], "drive0": [ { "busy": false, @@ -1992,6 +2072,22 @@ write -P0x67 0x3fe0000 0x20000 {"return": ""} { "bitmaps": { + "backup-top": [ + { + "busy": false, + "count": 67108864, + "granularity": 65536, + "persistent": false, + "recording": false + }, + { + "busy": false, + "count": 458752, + "granularity": 65536, + "persistent": false, + "recording": false + } + ], "drive0": [ { "busy": false, @@ -2245,6 +2341,22 @@ write -P0x67 0x3fe0000 0x20000 {"return": ""} { "bitmaps": { + "backup-top": [ + { + "busy": false, + "count": 67108864, + "granularity": 65536, + "persistent": false, + "recording": false + }, + { + "busy": false, + "count": 458752, + "granularity": 65536, + "persistent": false, + "recording": false + } + ], "drive0": [ { "busy": false, @@ -2705,6 +2817,22 @@ write -P0x67 0x3fe0000 0x20000 {"return": ""} { "bitmaps": { + "backup-top": [ + { + "busy": false, + "count": 67108864, + "granularity": 65536, + "persistent": false, + "recording": false + }, + { + "busy": false, + "count": 458752, + "granularity": 65536, + "persistent": false, + "recording": false + } + ], "drive0": [ { "busy": false, @@ -2958,6 +3086,22 @@ write -P0x67 0x3fe0000 0x20000 {"return": ""} { "bitmaps": { + "backup-top": [ + { + "busy": false, + "count": 67108864, + "granularity": 65536, + "persistent": false, + "recording": false + }, + { + "busy": false, + "count": 458752, + "granularity": 65536, + "persistent": false, + "recording": false + } + ], "drive0": [ { "busy": false, @@ -3418,6 +3562,22 @@ write -P0x67 0x3fe0000 0x20000 {"return": ""} { "bitmaps": { + "backup-top": [ + { + "busy": false, + "count": 67108864, + "granularity": 65536, + "persistent": false, + "recording": false + }, + { + "busy": false, + "count": 458752, + "granularity": 65536, + "persistent": false, + "recording": false + } + ], "drive0": [ { "busy": false, @@ -3671,6 +3831,22 @@ write -P0x67 0x3fe0000 0x20000 {"return": ""} { "bitmaps": { + "backup-top": [ + { + "busy": false, + "count": 67108864, + "granularity": 65536, + "persistent": false, + "recording": false + }, + { + "busy": false, + "count": 458752, + "granularity": 65536, + "persistent": false, + "recording": false + } + ], "drive0": [ { "busy": false, @@ -4131,6 +4307,22 @@ write -P0x67 0x3fe0000 0x20000 {"return": ""} { "bitmaps": { + "backup-top": [ + { + "busy": false, + "count": 67108864, + "granularity": 65536, + "persistent": false, + "recording": false + }, + { + "busy": false, + "count": 458752, + "granularity": 65536, + "persistent": false, + "recording": false + } + ], "drive0": [ { "busy": false, @@ -4384,6 +4576,22 @@ write -P0x67 0x3fe0000 0x20000 {"return": ""} { "bitmaps": { + "backup-top": [ + { + "busy": false, + "count": 67108864, + "granularity": 65536, + "persistent": false, + "recording": false + }, + { + "busy": false, + "count": 458752, + "granularity": 65536, + "persistent": false, + "recording": false + } + ], "drive0": [ { "busy": false, @@ -4844,6 +5052,22 @@ write -P0x67 0x3fe0000 0x20000 {"return": ""} { "bitmaps": { + "backup-top": [ + { + "busy": false, + "count": 67108864, + "granularity": 65536, + "persistent": false, + "recording": false + }, + { + "busy": false, + "count": 458752, + "granularity": 65536, + "persistent": false, + "recording": false + } + ], "drive0": [ { "busy": false, diff --git a/tests/qemu-iotests/common.rc b/tests/qemu-iotests/common.rc index 3bfd94c2e0..227e0a5be9 100644 --- a/tests/qemu-iotests/common.rc +++ b/tests/qemu-iotests/common.rc @@ -20,7 +20,7 @@ # bail out, setting up .notrun file _notrun() { - echo "$*" >"$OUTPUT_DIR/$seq.notrun" + echo "$*" >"$TEST_DIR/$seq.notrun" echo "$seq not run: $*" status=0 exit @@ -739,14 +739,14 @@ _img_info() # _casenotrun() { - echo " [case not run] $*" >>"$OUTPUT_DIR/$seq.casenotrun" + echo " [case not run] $*" >>"$TEST_DIR/$seq.casenotrun" } # just plain bail out # _fail() { - echo "$*" | tee -a "$OUTPUT_DIR/$seq.full" + echo "$*" | tee -a "$TEST_DIR/$seq.full" echo "(see $seq.full for details)" status=1 exit 1 diff --git a/tests/qemu-iotests/iotests.py b/tests/qemu-iotests/iotests.py index 6027780180..508adade9e 100644 --- a/tests/qemu-iotests/iotests.py +++ b/tests/qemu-iotests/iotests.py @@ -85,7 +85,6 @@ qemu_print = os.environ.get('PRINT_QEMU', False) imgfmt = os.environ.get('IMGFMT', 'raw') imgproto = os.environ.get('IMGPROTO', 'file') -output_dir = os.environ.get('OUTPUT_DIR', '.') try: test_dir = os.environ['TEST_DIR'] @@ -279,6 +278,9 @@ def qemu_io(*args): '''Run qemu-io and return the stdout data''' return qemu_tool_pipe_and_status('qemu-io', qemu_io_wrap_args(args))[0] +def qemu_io_pipe_and_status(*args): + return qemu_tool_pipe_and_status('qemu-io', qemu_io_wrap_args(args)) + def qemu_io_log(*args): result = qemu_io(*args) log(result, filters=[filter_testfiles, filter_qemu_io]) @@ -1239,7 +1241,7 @@ def notrun(reason): # Each test in qemu-iotests has a number ("seq") seq = os.path.basename(sys.argv[0]) - with open('%s/%s.notrun' % (output_dir, seq), 'w', encoding='utf-8') \ + with open('%s/%s.notrun' % (test_dir, seq), 'w', encoding='utf-8') \ as outfile: outfile.write(reason + '\n') logger.warning("%s not run: %s", seq, reason) @@ -1254,7 +1256,7 @@ def case_notrun(reason): # Each test in qemu-iotests has a number ("seq") seq = os.path.basename(sys.argv[0]) - with open('%s/%s.casenotrun' % (output_dir, seq), 'a', encoding='utf-8') \ + with open('%s/%s.casenotrun' % (test_dir, seq), 'a', encoding='utf-8') \ as outfile: outfile.write(' [case not run] ' + reason + '\n') diff --git a/tests/qemu-iotests/testenv.py b/tests/qemu-iotests/testenv.py index 0f32897fe8..b11e943c8a 100644 --- a/tests/qemu-iotests/testenv.py +++ b/tests/qemu-iotests/testenv.py @@ -66,7 +66,7 @@ class TestEnv(ContextManager['TestEnv']): # pylint: disable=too-many-instance-attributes env_variables = ['PYTHONPATH', 'TEST_DIR', 'SOCK_DIR', 'SAMPLE_IMG_DIR', - 'OUTPUT_DIR', 'PYTHON', 'QEMU_PROG', 'QEMU_IMG_PROG', + 'PYTHON', 'QEMU_PROG', 'QEMU_IMG_PROG', 'QEMU_IO_PROG', 'QEMU_NBD_PROG', 'QSD_PROG', 'QEMU_OPTIONS', 'QEMU_IMG_OPTIONS', 'QEMU_IO_OPTIONS', 'QEMU_IO_OPTIONS_NO_FMT', @@ -106,7 +106,6 @@ class TestEnv(ContextManager['TestEnv']): TEST_DIR SOCK_DIR SAMPLE_IMG_DIR - OUTPUT_DIR """ # Path where qemu goodies live in this source tree. @@ -134,8 +133,6 @@ class TestEnv(ContextManager['TestEnv']): os.path.join(self.source_iotests, 'sample_images')) - self.output_dir = os.getcwd() # OUTPUT_DIR - def init_binaries(self) -> None: """Init binary path variables: PYTHON (for bash tests) diff --git a/tests/qemu-iotests/testrunner.py b/tests/qemu-iotests/testrunner.py index 9a94273975..41083ff9c6 100644 --- a/tests/qemu-iotests/testrunner.py +++ b/tests/qemu-iotests/testrunner.py @@ -259,9 +259,6 @@ class TestRunner(ContextManager['TestRunner']): """ f_test = Path(test) - f_bad = Path(f_test.name + '.out.bad') - f_notrun = Path(f_test.name + '.notrun') - f_casenotrun = Path(f_test.name + '.casenotrun') f_reference = Path(self.find_reference(test)) if not f_test.exists(): @@ -276,9 +273,6 @@ class TestRunner(ContextManager['TestRunner']): description='No qualified output ' f'(expected {f_reference})') - for p in (f_bad, f_notrun, f_casenotrun): - silent_unlink(p) - args = [str(f_test.resolve())] env = self.env.prepare_subprocess(args) if mp: @@ -288,6 +282,14 @@ class TestRunner(ContextManager['TestRunner']): env[d] = os.path.join(env[d], f_test.name) Path(env[d]).mkdir(parents=True, exist_ok=True) + test_dir = env['TEST_DIR'] + f_bad = Path(test_dir, f_test.name + '.out.bad') + f_notrun = Path(test_dir, f_test.name + '.notrun') + f_casenotrun = Path(test_dir, f_test.name + '.casenotrun') + + for p in (f_notrun, f_casenotrun): + silent_unlink(p) + t0 = time.time() with f_bad.open('w', encoding="utf-8") as f: with subprocess.Popen(args, cwd=str(f_test.parent), env=env, @@ -365,7 +367,10 @@ class TestRunner(ContextManager['TestRunner']): description=res.description) if res.casenotrun: - print(res.casenotrun) + if self.tap: + print('#' + res.casenotrun.replace('\n', '\n#')) + else: + print(res.casenotrun) return res diff --git a/tests/qemu-iotests/tests/image-fleecing b/tests/qemu-iotests/tests/image-fleecing index a58b5a1781..c56278639c 100755 --- a/tests/qemu-iotests/tests/image-fleecing +++ b/tests/qemu-iotests/tests/image-fleecing @@ -23,12 +23,14 @@ # Creator/Owner: John Snow <jsnow@redhat.com> import iotests -from iotests import log, qemu_img, qemu_io, qemu_io_silent +from iotests import log, qemu_img, qemu_io, qemu_io_silent, \ + qemu_io_pipe_and_status iotests.script_initialize( - supported_fmts=['qcow2', 'qcow', 'qed', 'vmdk', 'vhdx', 'raw'], + supported_fmts=['qcow2'], supported_platforms=['linux'], required_fmts=['copy-before-write'], + unsupported_imgopts=['compat'] ) patterns = [('0x5d', '0', '64k'), @@ -49,12 +51,30 @@ remainder = [('0xd5', '0x108000', '32k'), # Right-end of partial-left [1] ('0xdc', '32M', '32k'), # Left-end of partial-right [2] ('0xcd', '0x3ff0000', '64k')] # patterns[3] -def do_test(use_cbw, base_img_path, fleece_img_path, nbd_sock_path, vm): +def do_test(vm, use_cbw, use_snapshot_access_filter, base_img_path, + fleece_img_path, nbd_sock_path=None, + target_img_path=None, + bitmap=False): + push_backup = target_img_path is not None + assert (nbd_sock_path is not None) != push_backup + if push_backup: + assert use_cbw + log('--- Setting up images ---') log('') assert qemu_img('create', '-f', iotests.imgfmt, base_img_path, '64M') == 0 - assert qemu_img('create', '-f', 'qcow2', fleece_img_path, '64M') == 0 + if bitmap: + assert qemu_img('bitmap', '--add', base_img_path, 'bitmap0') == 0 + + if use_snapshot_access_filter: + assert use_cbw + assert qemu_img('create', '-f', 'raw', fleece_img_path, '64M') == 0 + else: + assert qemu_img('create', '-f', 'qcow2', fleece_img_path, '64M') == 0 + + if push_backup: + assert qemu_img('create', '-f', 'qcow2', target_img_path, '64M') == 0 for p in patterns: qemu_io('-f', iotests.imgfmt, @@ -81,27 +101,46 @@ def do_test(use_cbw, base_img_path, fleece_img_path, nbd_sock_path, vm): log('') - # create tmp_node backed by src_node - log(vm.qmp('blockdev-add', { - 'driver': 'qcow2', - 'node-name': tmp_node, - 'file': { + if use_snapshot_access_filter: + log(vm.qmp('blockdev-add', { + 'node-name': tmp_node, 'driver': 'file', 'filename': fleece_img_path, - }, - 'backing': src_node, - })) + })) + else: + # create tmp_node backed by src_node + log(vm.qmp('blockdev-add', { + 'driver': 'qcow2', + 'node-name': tmp_node, + 'file': { + 'driver': 'file', + 'filename': fleece_img_path, + }, + 'backing': src_node, + })) # Establish CBW from source to fleecing node if use_cbw: - log(vm.qmp('blockdev-add', { + fl_cbw = { 'driver': 'copy-before-write', 'node-name': 'fl-cbw', 'file': src_node, 'target': tmp_node - })) + } + + if bitmap: + fl_cbw['bitmap'] = {'node': src_node, 'name': 'bitmap0'} + + log(vm.qmp('blockdev-add', fl_cbw)) log(vm.qmp('qom-set', path=qom_path, property='drive', value='fl-cbw')) + + if use_snapshot_access_filter: + log(vm.qmp('blockdev-add', { + 'driver': 'snapshot-access', + 'node-name': 'fl-access', + 'file': 'fl-cbw', + })) else: log(vm.qmp('blockdev-backup', job_id='fleecing', @@ -109,25 +148,47 @@ def do_test(use_cbw, base_img_path, fleece_img_path, nbd_sock_path, vm): target=tmp_node, sync='none')) - log('') - log('--- Setting up NBD Export ---') - log('') + export_node = 'fl-access' if use_snapshot_access_filter else tmp_node + + if push_backup: + log('') + log('--- Starting actual backup ---') + log('') - nbd_uri = 'nbd+unix:///%s?socket=%s' % (tmp_node, nbd_sock_path) - log(vm.qmp('nbd-server-start', - {'addr': {'type': 'unix', - 'data': {'path': nbd_sock_path}}})) + log(vm.qmp('blockdev-add', **{ + 'driver': iotests.imgfmt, + 'node-name': 'target', + 'file': { + 'driver': 'file', + 'filename': target_img_path + } + })) + log(vm.qmp('blockdev-backup', device=export_node, + sync='full', target='target', + job_id='push-backup', speed=1)) + else: + log('') + log('--- Setting up NBD Export ---') + log('') - log(vm.qmp('nbd-server-add', device=tmp_node)) + nbd_uri = 'nbd+unix:///%s?socket=%s' % (export_node, nbd_sock_path) + log(vm.qmp('nbd-server-start', + {'addr': { 'type': 'unix', + 'data': { 'path': nbd_sock_path } } })) - log('') - log('--- Sanity Check ---') - log('') + log(vm.qmp('nbd-server-add', device=export_node)) - for p in patterns + zeroes: - cmd = 'read -P%s %s %s' % p - log(cmd) - assert qemu_io_silent('-r', '-f', 'raw', '-c', cmd, nbd_uri) == 0 + log('') + log('--- Sanity Check ---') + log('') + + for p in patterns + zeroes: + cmd = 'read -P%s %s %s' % p + log(cmd) + out, ret = qemu_io_pipe_and_status('-r', '-f', 'raw', '-c', cmd, + nbd_uri) + if ret != 0: + print(out) log('') log('--- Testing COW ---') @@ -138,6 +199,23 @@ def do_test(use_cbw, base_img_path, fleece_img_path, nbd_sock_path, vm): log(cmd) log(vm.hmp_qemu_io(qom_path, cmd, qdev=True)) + if push_backup: + # Check that previous operations were done during backup, not after + # If backup is already finished, it's possible that it was finished + # even before hmp qemu_io write, and we didn't actually test + # copy-before-write operation. This should not happen, as we use + # speed=1. But worth checking. + result = vm.qmp('query-block-jobs') + assert len(result['return']) == 1 + + result = vm.qmp('block-job-set-speed', device='push-backup', speed=0) + assert result == {'return': {}} + + log(vm.event_wait(name='BLOCK_JOB_COMPLETED', + match={'data': {'device': 'push-backup'}}), + filters=[iotests.filter_qmp_event]) + log(vm.qmp('blockdev-del', node_name='target')) + log('') log('--- Verifying Data ---') log('') @@ -145,13 +223,25 @@ def do_test(use_cbw, base_img_path, fleece_img_path, nbd_sock_path, vm): for p in patterns + zeroes: cmd = 'read -P%s %s %s' % p log(cmd) - assert qemu_io_silent('-r', '-f', 'raw', '-c', cmd, nbd_uri) == 0 + args = ['-r', '-c', cmd] + if push_backup: + args += [target_img_path] + else: + args += ['-f', 'raw', nbd_uri] + out, ret = qemu_io_pipe_and_status(*args) + if ret != 0: + print(out) log('') log('--- Cleanup ---') log('') + if not push_backup: + log(vm.qmp('nbd-server-stop')) + if use_cbw: + if use_snapshot_access_filter: + log(vm.qmp('blockdev-del', node_name='fl-access')) log(vm.qmp('qom-set', path=qom_path, property='drive', value=src_node)) log(vm.qmp('blockdev-del', node_name='fl-cbw')) else: @@ -160,7 +250,6 @@ def do_test(use_cbw, base_img_path, fleece_img_path, nbd_sock_path, vm): assert e is not None log(e, filters=[iotests.filter_qmp_event]) - log(vm.qmp('nbd-server-stop')) log(vm.qmp('blockdev-del', node_name=tmp_node)) vm.shutdown() @@ -177,17 +266,37 @@ def do_test(use_cbw, base_img_path, fleece_img_path, nbd_sock_path, vm): log('Done') -def test(use_cbw): +def test(use_cbw, use_snapshot_access_filter, + nbd_sock_path=None, target_img_path=None, bitmap=False): with iotests.FilePath('base.img') as base_img_path, \ iotests.FilePath('fleece.img') as fleece_img_path, \ - iotests.FilePath('nbd.sock', - base_dir=iotests.sock_dir) as nbd_sock_path, \ iotests.VM() as vm: - do_test(use_cbw, base_img_path, fleece_img_path, nbd_sock_path, vm) + do_test(vm, use_cbw, use_snapshot_access_filter, base_img_path, + fleece_img_path, nbd_sock_path, target_img_path, + bitmap=bitmap) + +def test_pull(use_cbw, use_snapshot_access_filter, bitmap=False): + with iotests.FilePath('nbd.sock', + base_dir=iotests.sock_dir) as nbd_sock_path: + test(use_cbw, use_snapshot_access_filter, nbd_sock_path, None, + bitmap=bitmap) + +def test_push(): + with iotests.FilePath('target.img') as target_img_path: + test(True, True, None, target_img_path) log('=== Test backup(sync=none) based fleecing ===\n') -test(False) +test_pull(False, False) + +log('=== Test cbw-filter based fleecing ===\n') +test_pull(True, False) + +log('=== Test fleecing-format based fleecing ===\n') +test_pull(True, True) + +log('=== Test fleecing-format based fleecing with bitmap ===\n') +test_pull(True, True, bitmap=True) -log('=== Test filter based fleecing ===\n') -test(True) +log('=== Test push backup with fleecing ===\n') +test_push() diff --git a/tests/qemu-iotests/tests/image-fleecing.out b/tests/qemu-iotests/tests/image-fleecing.out index e96d122a8b..acfc89ff0e 100644 --- a/tests/qemu-iotests/tests/image-fleecing.out +++ b/tests/qemu-iotests/tests/image-fleecing.out @@ -52,8 +52,150 @@ read -P0 0x3fe0000 64k --- Cleanup --- {"return": {}} +{"return": {}} {"data": {"device": "fleecing", "len": 67108864, "offset": 393216, "speed": 0, "type": "backup"}, "event": "BLOCK_JOB_CANCELLED", "timestamp": {"microseconds": "USECS", "seconds": "SECS"}} {"return": {}} + +--- Confirming writes --- + +read -P0xab 0 64k +read -P0xad 0x00f8000 64k +read -P0x1d 0x2008000 64k +read -P0xea 0x3fe0000 64k +read -P0xd5 0x108000 32k +read -P0xdc 32M 32k +read -P0xcd 0x3ff0000 64k + +Done +=== Test cbw-filter based fleecing === + +--- Setting up images --- + +Done + +--- Launching VM --- + +Done + +--- Setting up Fleecing Graph --- + +{"return": {}} +{"return": {}} +{"return": {}} + +--- Setting up NBD Export --- + +{"return": {}} +{"return": {}} + +--- Sanity Check --- + +read -P0x5d 0 64k +read -P0xd5 1M 64k +read -P0xdc 32M 64k +read -P0xcd 0x3ff0000 64k +read -P0 0x00f8000 32k +read -P0 0x2010000 32k +read -P0 0x3fe0000 64k + +--- Testing COW --- + +write -P0xab 0 64k +{"return": ""} +write -P0xad 0x00f8000 64k +{"return": ""} +write -P0x1d 0x2008000 64k +{"return": ""} +write -P0xea 0x3fe0000 64k +{"return": ""} + +--- Verifying Data --- + +read -P0x5d 0 64k +read -P0xd5 1M 64k +read -P0xdc 32M 64k +read -P0xcd 0x3ff0000 64k +read -P0 0x00f8000 32k +read -P0 0x2010000 32k +read -P0 0x3fe0000 64k + +--- Cleanup --- + +{"return": {}} +{"return": {}} +{"return": {}} +{"return": {}} + +--- Confirming writes --- + +read -P0xab 0 64k +read -P0xad 0x00f8000 64k +read -P0x1d 0x2008000 64k +read -P0xea 0x3fe0000 64k +read -P0xd5 0x108000 32k +read -P0xdc 32M 32k +read -P0xcd 0x3ff0000 64k + +Done +=== Test fleecing-format based fleecing === + +--- Setting up images --- + +Done + +--- Launching VM --- + +Done + +--- Setting up Fleecing Graph --- + +{"return": {}} +{"return": {}} +{"return": {}} +{"return": {}} + +--- Setting up NBD Export --- + +{"return": {}} +{"return": {}} + +--- Sanity Check --- + +read -P0x5d 0 64k +read -P0xd5 1M 64k +read -P0xdc 32M 64k +read -P0xcd 0x3ff0000 64k +read -P0 0x00f8000 32k +read -P0 0x2010000 32k +read -P0 0x3fe0000 64k + +--- Testing COW --- + +write -P0xab 0 64k +{"return": ""} +write -P0xad 0x00f8000 64k +{"return": ""} +write -P0x1d 0x2008000 64k +{"return": ""} +write -P0xea 0x3fe0000 64k +{"return": ""} + +--- Verifying Data --- + +read -P0x5d 0 64k +read -P0xd5 1M 64k +read -P0xdc 32M 64k +read -P0xcd 0x3ff0000 64k +read -P0 0x00f8000 32k +read -P0 0x2010000 32k +read -P0 0x3fe0000 64k + +--- Cleanup --- + +{"return": {}} +{"return": {}} +{"return": {}} +{"return": {}} {"return": {}} --- Confirming writes --- @@ -67,7 +209,7 @@ read -P0xdc 32M 32k read -P0xcd 0x3ff0000 64k Done -=== Test filter based fleecing === +=== Test fleecing-format based fleecing with bitmap === --- Setting up images --- @@ -82,6 +224,7 @@ Done {"return": {}} {"return": {}} {"return": {}} +{"return": {}} --- Setting up NBD Export --- @@ -95,8 +238,82 @@ read -P0xd5 1M 64k read -P0xdc 32M 64k read -P0xcd 0x3ff0000 64k read -P0 0x00f8000 32k +read failed: Invalid argument + +read -P0 0x2010000 32k +read failed: Invalid argument + +read -P0 0x3fe0000 64k +read failed: Invalid argument + + +--- Testing COW --- + +write -P0xab 0 64k +{"return": ""} +write -P0xad 0x00f8000 64k +{"return": ""} +write -P0x1d 0x2008000 64k +{"return": ""} +write -P0xea 0x3fe0000 64k +{"return": ""} + +--- Verifying Data --- + +read -P0x5d 0 64k +read -P0xd5 1M 64k +read -P0xdc 32M 64k +read -P0xcd 0x3ff0000 64k +read -P0 0x00f8000 32k +read failed: Invalid argument + read -P0 0x2010000 32k +read failed: Invalid argument + read -P0 0x3fe0000 64k +read failed: Invalid argument + + +--- Cleanup --- + +{"return": {}} +{"return": {}} +{"return": {}} +{"return": {}} +{"return": {}} + +--- Confirming writes --- + +read -P0xab 0 64k +read -P0xad 0x00f8000 64k +read -P0x1d 0x2008000 64k +read -P0xea 0x3fe0000 64k +read -P0xd5 0x108000 32k +read -P0xdc 32M 32k +read -P0xcd 0x3ff0000 64k + +Done +=== Test push backup with fleecing === + +--- Setting up images --- + +Done + +--- Launching VM --- + +Done + +--- Setting up Fleecing Graph --- + +{"return": {}} +{"return": {}} +{"return": {}} +{"return": {}} + +--- Starting actual backup --- + +{"return": {}} +{"return": {}} --- Testing COW --- @@ -108,6 +325,8 @@ write -P0x1d 0x2008000 64k {"return": ""} write -P0xea 0x3fe0000 64k {"return": ""} +{"data": {"device": "push-backup", "len": 67108864, "offset": 67108864, "speed": 0, "type": "backup"}, "event": "BLOCK_JOB_COMPLETED", "timestamp": {"microseconds": "USECS", "seconds": "SECS"}} +{"return": {}} --- Verifying Data --- diff --git a/util/hbitmap.c b/util/hbitmap.c index 305b894a63..dd0501d9a7 100644 --- a/util/hbitmap.c +++ b/util/hbitmap.c @@ -301,6 +301,39 @@ bool hbitmap_next_dirty_area(const HBitmap *hb, int64_t start, int64_t end, return true; } +bool hbitmap_status(const HBitmap *hb, int64_t start, int64_t count, + int64_t *pnum) +{ + int64_t next_dirty, next_zero; + + assert(start >= 0); + assert(count > 0); + assert(start + count <= hb->orig_size); + + next_dirty = hbitmap_next_dirty(hb, start, count); + if (next_dirty == -1) { + *pnum = count; + return false; + } + + if (next_dirty > start) { + *pnum = next_dirty - start; + return false; + } + + assert(next_dirty == start); + + next_zero = hbitmap_next_zero(hb, start, count); + if (next_zero == -1) { + *pnum = count; + return true; + } + + assert(next_zero > start); + *pnum = next_zero - start; + return false; +} + bool hbitmap_empty(const HBitmap *hb) { return hb->count == 0; |