diff options
Diffstat (limited to 'block')
-rw-r--r-- | block/block-copy.c | 33 | ||||
-rw-r--r-- | block/copy-before-write.c | 110 | ||||
-rw-r--r-- | block/mirror.c | 2 | ||||
-rw-r--r-- | block/nbd.c | 8 | ||||
-rw-r--r-- | block/trace-events | 2 |
5 files changed, 115 insertions, 40 deletions
diff --git a/block/block-copy.c b/block/block-copy.c index ec46775ea5..bb947afdda 100644 --- a/block/block-copy.c +++ b/block/block-copy.c @@ -883,23 +883,42 @@ static int coroutine_fn block_copy_common(BlockCopyCallState *call_state) return ret; } +static void coroutine_fn block_copy_async_co_entry(void *opaque) +{ + block_copy_common(opaque); +} + int coroutine_fn block_copy(BlockCopyState *s, int64_t start, int64_t bytes, - bool ignore_ratelimit) + bool ignore_ratelimit, uint64_t timeout_ns, + BlockCopyAsyncCallbackFunc cb, + void *cb_opaque) { - BlockCopyCallState call_state = { + int ret; + BlockCopyCallState *call_state = g_new(BlockCopyCallState, 1); + + *call_state = (BlockCopyCallState) { .s = s, .offset = start, .bytes = bytes, .ignore_ratelimit = ignore_ratelimit, .max_workers = BLOCK_COPY_MAX_WORKERS, + .cb = cb, + .cb_opaque = cb_opaque, }; - return block_copy_common(&call_state); -} + ret = qemu_co_timeout(block_copy_async_co_entry, call_state, timeout_ns, + g_free); + if (ret < 0) { + assert(ret == -ETIMEDOUT); + block_copy_call_cancel(call_state); + /* call_state will be freed by running coroutine. */ + return ret; + } -static void coroutine_fn block_copy_async_co_entry(void *opaque) -{ - block_copy_common(opaque); + ret = call_state->ret; + g_free(call_state); + + return ret; } BlockCopyCallState *block_copy_async(BlockCopyState *s, diff --git a/block/copy-before-write.c b/block/copy-before-write.c index a8a06fdc09..c24b8dd117 100644 --- a/block/copy-before-write.c +++ b/block/copy-before-write.c @@ -24,6 +24,7 @@ */ #include "qemu/osdep.h" +#include "qapi/qmp/qjson.h" #include "sysemu/block-backend.h" #include "qemu/cutils.h" @@ -40,6 +41,8 @@ typedef struct BDRVCopyBeforeWriteState { BlockCopyState *bcs; BdrvChild *target; + OnCbwError on_cbw_error; + uint32_t cbw_timeout_ns; /* * @lock: protects access to @access_bitmap, @done_bitmap and @@ -64,6 +67,14 @@ typedef struct BDRVCopyBeforeWriteState { * node. These areas must not be rewritten by guest. */ BlockReqList frozen_read_reqs; + + /* + * @snapshot_error is normally zero. But on first copy-before-write failure + * when @on_cbw_error == ON_CBW_ERROR_BREAK_SNAPSHOT, @snapshot_error takes + * value of this error (<0). After that all in-flight and further + * snapshot-API requests will fail with that error. + */ + int snapshot_error; } BDRVCopyBeforeWriteState; static coroutine_fn int cbw_co_preadv( @@ -73,6 +84,13 @@ static coroutine_fn int cbw_co_preadv( return bdrv_co_preadv(bs->file, offset, bytes, qiov, flags); } +static void block_copy_cb(void *opaque) +{ + BlockDriverState *bs = opaque; + + bdrv_dec_in_flight(bs); +} + /* * Do copy-before-write operation. * @@ -94,16 +112,36 @@ static coroutine_fn int cbw_do_copy_before_write(BlockDriverState *bs, return 0; } + if (s->snapshot_error) { + return 0; + } + off = QEMU_ALIGN_DOWN(offset, cluster_size); end = QEMU_ALIGN_UP(offset + bytes, cluster_size); - ret = block_copy(s->bcs, off, end - off, true); - if (ret < 0) { + /* + * Increase in_flight, so that in case of timed-out block-copy, the + * remaining background block_copy() request (which can't be immediately + * cancelled by timeout) is presented in bs->in_flight. This way we are + * sure that on bs close() we'll previously wait for all timed-out but yet + * running block_copy calls. + */ + bdrv_inc_in_flight(bs); + ret = block_copy(s->bcs, off, end - off, true, s->cbw_timeout_ns, + block_copy_cb, bs); + if (ret < 0 && s->on_cbw_error == ON_CBW_ERROR_BREAK_GUEST_WRITE) { return ret; } WITH_QEMU_LOCK_GUARD(&s->lock) { - bdrv_set_dirty_bitmap(s->done_bitmap, off, end - off); + if (ret < 0) { + assert(s->on_cbw_error == ON_CBW_ERROR_BREAK_SNAPSHOT); + if (!s->snapshot_error) { + s->snapshot_error = ret; + } + } else { + bdrv_set_dirty_bitmap(s->done_bitmap, off, end - off); + } reqlist_wait_all(&s->frozen_read_reqs, off, end - off, &s->lock); } @@ -175,6 +213,11 @@ static BlockReq *cbw_snapshot_read_lock(BlockDriverState *bs, QEMU_LOCK_GUARD(&s->lock); + if (s->snapshot_error) { + g_free(req); + return NULL; + } + if (bdrv_dirty_bitmap_next_zero(s->access_bitmap, offset, bytes) != -1) { g_free(req); return NULL; @@ -328,46 +371,36 @@ static void cbw_child_perm(BlockDriverState *bs, BdrvChild *c, } } -static bool cbw_parse_bitmap_option(QDict *options, BdrvDirtyBitmap **bitmap, - Error **errp) +static BlockdevOptions *cbw_parse_options(QDict *options, Error **errp) { - QDict *bitmap_qdict = NULL; - BlockDirtyBitmap *bmp_param = NULL; + BlockdevOptions *opts = NULL; Visitor *v = NULL; - bool ret = false; - *bitmap = NULL; + qdict_put_str(options, "driver", "copy-before-write"); - 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); + v = qobject_input_visitor_new_flat_confused(options, 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) { + visit_type_BlockdevOptions(v, NULL, &opts, errp); + if (!opts) { goto out; } - ret = true; + /* + * Delete options which we are going to parse through BlockdevOptions + * object for original options. + */ + qdict_extract_subqdict(options, NULL, "bitmap"); + qdict_del(options, "on-cbw-error"); + qdict_del(options, "cbw-timeout"); out: - qapi_free_BlockDirtyBitmap(bmp_param); visit_free(v); - qobject_unref(bitmap_qdict); + qdict_del(options, "driver"); - return ret; + return opts; } static int cbw_open(BlockDriverState *bs, QDict *options, int flags, @@ -376,6 +409,15 @@ static int cbw_open(BlockDriverState *bs, QDict *options, int flags, BDRVCopyBeforeWriteState *s = bs->opaque; BdrvDirtyBitmap *bitmap = NULL; int64_t cluster_size; + g_autoptr(BlockdevOptions) full_opts = NULL; + BlockdevOptionsCbw *opts; + + full_opts = cbw_parse_options(options, errp); + if (!full_opts) { + return -EINVAL; + } + assert(full_opts->driver == BLOCKDEV_DRIVER_COPY_BEFORE_WRITE); + opts = &full_opts->u.copy_before_write; bs->file = bdrv_open_child(NULL, options, "file", bs, &child_of_bds, BDRV_CHILD_FILTERED | BDRV_CHILD_PRIMARY, @@ -390,9 +432,17 @@ static int cbw_open(BlockDriverState *bs, QDict *options, int flags, return -EINVAL; } - if (!cbw_parse_bitmap_option(options, &bitmap, errp)) { - return -EINVAL; + if (opts->has_bitmap) { + bitmap = block_dirty_bitmap_lookup(opts->bitmap->node, + opts->bitmap->name, NULL, errp); + if (!bitmap) { + return -EINVAL; + } } + s->on_cbw_error = opts->has_on_cbw_error ? opts->on_cbw_error : + ON_CBW_ERROR_BREAK_GUEST_WRITE; + s->cbw_timeout_ns = opts->has_cbw_timeout ? + opts->cbw_timeout * NANOSECONDS_PER_SECOND : 0; bs->total_sectors = bs->file->bs->total_sectors; bs->supported_write_flags = BDRV_REQ_WRITE_UNCHANGED | diff --git a/block/mirror.c b/block/mirror.c index d8ecb9efa2..3c4ab1159d 100644 --- a/block/mirror.c +++ b/block/mirror.c @@ -73,7 +73,7 @@ typedef struct MirrorBlockJob { uint64_t last_pause_ns; unsigned long *in_flight_bitmap; - int in_flight; + unsigned in_flight; int64_t bytes_in_flight; QTAILQ_HEAD(, MirrorOp) ops_in_flight; int ret; diff --git a/block/nbd.c b/block/nbd.c index 7f5f50ec46..97683cce27 100644 --- a/block/nbd.c +++ b/block/nbd.c @@ -77,7 +77,7 @@ typedef struct BDRVNBDState { QemuMutex requests_lock; NBDClientState state; CoQueue free_sema; - int in_flight; + unsigned in_flight; NBDClientRequest requests[MAX_NBD_REQUESTS]; QEMUTimer *reconnect_delay_timer; @@ -371,6 +371,7 @@ static bool nbd_client_connecting(BDRVNBDState *s) /* Called with s->requests_lock taken. */ static coroutine_fn void nbd_reconnect_attempt(BDRVNBDState *s) { + int ret; bool blocking = s->state == NBD_CLIENT_CONNECTING_WAIT; /* @@ -380,6 +381,8 @@ static coroutine_fn void nbd_reconnect_attempt(BDRVNBDState *s) assert(nbd_client_connecting(s)); assert(s->in_flight == 1); + trace_nbd_reconnect_attempt(s->bs->in_flight); + if (blocking && !s->reconnect_delay_timer) { /* * It's the first reconnect attempt after switching to @@ -401,7 +404,8 @@ static coroutine_fn void nbd_reconnect_attempt(BDRVNBDState *s) } qemu_mutex_unlock(&s->requests_lock); - nbd_co_do_establish_connection(s->bs, blocking, NULL); + ret = nbd_co_do_establish_connection(s->bs, blocking, NULL); + trace_nbd_reconnect_attempt_result(ret, s->bs->in_flight); qemu_mutex_lock(&s->requests_lock); /* diff --git a/block/trace-events b/block/trace-events index 549090d453..48dbf10c66 100644 --- a/block/trace-events +++ b/block/trace-events @@ -172,6 +172,8 @@ nbd_read_reply_entry_fail(int ret, const char *err) "ret = %d, err: %s" nbd_co_request_fail(uint64_t from, uint32_t len, uint64_t handle, uint16_t flags, uint16_t type, const char *name, int ret, const char *err) "Request failed { .from = %" PRIu64", .len = %" PRIu32 ", .handle = %" PRIu64 ", .flags = 0x%" PRIx16 ", .type = %" PRIu16 " (%s) } ret = %d, err: %s" nbd_client_handshake(const char *export_name) "export '%s'" nbd_client_handshake_success(const char *export_name) "export '%s'" +nbd_reconnect_attempt(unsigned in_flight) "in_flight %u" +nbd_reconnect_attempt_result(int ret, unsigned in_flight) "ret %d in_flight %u" # ssh.c ssh_restart_coroutine(void *co) "co=%p" |