/*
- * QEMU Block driver for NBD
+ * QEMU Block driver for NBD
*
* Copyright (c) 2019 Virtuozzo International GmbH.
- * Copyright (C) 2016 Red Hat, Inc.
+ * Copyright Red Hat
* Copyright (C) 2008 Bull S.A.S.
* Author: Laurent Vivier <Laurent.Vivier@bull.net>
*
#include "qemu/option.h"
#include "qemu/cutils.h"
#include "qemu/main-loop.h"
-#include "qemu/atomic.h"
#include "qapi/qapi-visit-sockets.h"
#include "qapi/qmp/qstring.h"
#define EN_OPTSTR ":exportname="
#define MAX_NBD_REQUESTS 16
-#define HANDLE_TO_INDEX(bs, handle) ((handle) ^ (uint64_t)(intptr_t)(bs))
-#define INDEX_TO_HANDLE(bs, index) ((index) ^ (uint64_t)(intptr_t)(bs))
+#define COOKIE_TO_INDEX(cookie) ((cookie) - 1)
+#define INDEX_TO_COOKIE(index) ((index) + 1)
typedef struct {
Coroutine *coroutine;
uint64_t offset; /* original offset of the request */
bool receiving; /* sleeping in the yield in nbd_receive_replies */
- bool reply_possible; /* reply header not yet received */
} NBDClientRequest;
typedef enum NBDClientState {
QIOChannel *ioc; /* The current I/O channel */
NBDExportInfo info;
- CoMutex send_mutex;
+ /*
+ * Protects state, free_sema, in_flight, requests[].coroutine,
+ * reconnect_delay_timer.
+ */
+ QemuMutex requests_lock;
+ NBDClientState state;
CoQueue free_sema;
+ unsigned in_flight;
+ NBDClientRequest requests[MAX_NBD_REQUESTS];
+ QEMUTimer *reconnect_delay_timer;
+
+ /* Protects sending data on the socket. */
+ CoMutex send_mutex;
+ /*
+ * Protects receiving reply headers from the socket, as well as the
+ * fields reply and requests[].receiving
+ */
CoMutex receive_mutex;
- int in_flight;
- NBDClientState state;
+ NBDReply reply;
- QEMUTimer *reconnect_delay_timer;
QEMUTimer *open_timer;
- NBDClientRequest requests[MAX_NBD_REQUESTS];
- NBDReply reply;
BlockDriverState *bs;
/* Connection parameters */
s->x_dirty_bitmap = NULL;
}
-static bool nbd_client_connected(BDRVNBDState *s)
-{
- return qatomic_load_acquire(&s->state) == NBD_CLIENT_CONNECTED;
-}
-
-static bool nbd_recv_coroutine_wake_one(NBDClientRequest *req)
+/* Called with s->receive_mutex taken. */
+static bool coroutine_fn nbd_recv_coroutine_wake_one(NBDClientRequest *req)
{
if (req->receiving) {
req->receiving = false;
return false;
}
-static void nbd_recv_coroutines_wake(BDRVNBDState *s, bool all)
+static void coroutine_fn nbd_recv_coroutines_wake(BDRVNBDState *s)
{
int i;
+ QEMU_LOCK_GUARD(&s->receive_mutex);
for (i = 0; i < MAX_NBD_REQUESTS; i++) {
- if (nbd_recv_coroutine_wake_one(&s->requests[i]) && !all) {
+ if (nbd_recv_coroutine_wake_one(&s->requests[i])) {
return;
}
}
}
-static void nbd_channel_error(BDRVNBDState *s, int ret)
+/* Called with s->requests_lock held. */
+static void coroutine_fn nbd_channel_error_locked(BDRVNBDState *s, int ret)
{
- if (nbd_client_connected(s)) {
+ if (s->state == NBD_CLIENT_CONNECTED) {
qio_channel_shutdown(s->ioc, QIO_CHANNEL_SHUTDOWN_BOTH, NULL);
}
if (ret == -EIO) {
- if (nbd_client_connected(s)) {
+ if (s->state == NBD_CLIENT_CONNECTED) {
s->state = s->reconnect_delay ? NBD_CLIENT_CONNECTING_WAIT :
NBD_CLIENT_CONNECTING_NOWAIT;
}
} else {
s->state = NBD_CLIENT_QUIT;
}
+}
- nbd_recv_coroutines_wake(s, true);
+static void coroutine_fn nbd_channel_error(BDRVNBDState *s, int ret)
+{
+ QEMU_LOCK_GUARD(&s->requests_lock);
+ nbd_channel_error_locked(s, ret);
}
static void reconnect_delay_timer_del(BDRVNBDState *s)
{
BDRVNBDState *s = opaque;
- if (qatomic_load_acquire(&s->state) == NBD_CLIENT_CONNECTING_WAIT) {
- s->state = NBD_CLIENT_CONNECTING_NOWAIT;
- nbd_co_establish_connection_cancel(s->conn);
- while (qemu_co_enter_next(&s->free_sema, NULL)) {
- /* Resume all queued requests */
+ reconnect_delay_timer_del(s);
+ WITH_QEMU_LOCK_GUARD(&s->requests_lock) {
+ if (s->state != NBD_CLIENT_CONNECTING_WAIT) {
+ return;
}
+ s->state = NBD_CLIENT_CONNECTING_NOWAIT;
}
-
- reconnect_delay_timer_del(s);
+ nbd_co_establish_connection_cancel(s->conn);
}
static void reconnect_delay_timer_init(BDRVNBDState *s, uint64_t expire_time_ns)
{
- if (qatomic_load_acquire(&s->state) != NBD_CLIENT_CONNECTING_WAIT) {
- return;
- }
-
assert(!s->reconnect_delay_timer);
s->reconnect_delay_timer = aio_timer_new(bdrv_get_aio_context(s->bs),
QEMU_CLOCK_REALTIME,
s->ioc = NULL;
}
- s->state = NBD_CLIENT_QUIT;
+ WITH_QEMU_LOCK_GUARD(&s->requests_lock) {
+ s->state = NBD_CLIENT_QUIT;
+ }
}
static void open_timer_del(BDRVNBDState *s)
timer_mod(s->open_timer, expire_time_ns);
}
-static bool nbd_client_connecting(BDRVNBDState *s)
-{
- NBDClientState state = qatomic_load_acquire(&s->state);
- return state == NBD_CLIENT_CONNECTING_WAIT ||
- state == NBD_CLIENT_CONNECTING_NOWAIT;
-}
-
-static bool nbd_client_connecting_wait(BDRVNBDState *s)
+static bool nbd_client_will_reconnect(BDRVNBDState *s)
{
- return qatomic_load_acquire(&s->state) == NBD_CLIENT_CONNECTING_WAIT;
+ /*
+ * Called only after a socket error, so this is not performance sensitive.
+ */
+ QEMU_LOCK_GUARD(&s->requests_lock);
+ return s->state == NBD_CLIENT_CONNECTING_WAIT;
}
/*
* Return failure if the server's advertised options are incompatible with the
* client's needs.
*/
-static int nbd_handle_updated_info(BlockDriverState *bs, Error **errp)
+static int coroutine_fn GRAPH_RDLOCK
+nbd_handle_updated_info(BlockDriverState *bs, Error **errp)
{
BDRVNBDState *s = (BDRVNBDState *)bs->opaque;
int ret;
}
int coroutine_fn nbd_co_do_establish_connection(BlockDriverState *bs,
- Error **errp)
+ bool blocking, Error **errp)
{
BDRVNBDState *s = (BDRVNBDState *)bs->opaque;
int ret;
- bool blocking = nbd_client_connecting_wait(s);
IO_CODE();
+ assert_bdrv_graph_readable();
assert(!s->ioc);
s->ioc = nbd_co_establish_connection(s->conn, &s->info, blocking, errp);
* We have connected, but must fail for other reasons.
* Send NBD_CMD_DISC as a courtesy to the server.
*/
- NBDRequest request = { .type = NBD_CMD_DISC };
+ NBDRequest request = { .type = NBD_CMD_DISC, .mode = s->info.mode };
nbd_send_request(s->ioc, &request);
}
qio_channel_set_blocking(s->ioc, false, NULL);
- qio_channel_attach_aio_context(s->ioc, bdrv_get_aio_context(bs));
+ qio_channel_set_follow_coroutine_ctx(s->ioc, true);
/* successfully connected */
- s->state = NBD_CLIENT_CONNECTED;
- qemu_co_queue_restart_all(&s->free_sema);
+ WITH_QEMU_LOCK_GUARD(&s->requests_lock) {
+ s->state = NBD_CLIENT_CONNECTED;
+ }
return 0;
}
-/* called under s->send_mutex */
-static coroutine_fn void nbd_reconnect_attempt(BDRVNBDState *s)
+/* Called with s->requests_lock held. */
+static bool nbd_client_connecting(BDRVNBDState *s)
+{
+ return s->state == NBD_CLIENT_CONNECTING_WAIT ||
+ s->state == NBD_CLIENT_CONNECTING_NOWAIT;
+}
+
+/* Called with s->requests_lock taken. */
+static void coroutine_fn GRAPH_RDLOCK nbd_reconnect_attempt(BDRVNBDState *s)
{
+ int ret;
+ bool blocking = s->state == NBD_CLIENT_CONNECTING_WAIT;
+
+ /*
+ * Now we are sure that nobody is accessing the channel, and no one will
+ * try until we set the state to CONNECTED.
+ */
assert(nbd_client_connecting(s));
- assert(s->in_flight == 0);
+ assert(s->in_flight == 1);
- if (nbd_client_connecting_wait(s) && s->reconnect_delay &&
- !s->reconnect_delay_timer)
- {
+ trace_nbd_reconnect_attempt(s->bs->in_flight);
+
+ if (blocking && !s->reconnect_delay_timer) {
/*
- * It's first reconnect attempt after switching to
+ * It's the first reconnect attempt after switching to
* NBD_CLIENT_CONNECTING_WAIT
*/
+ g_assert(s->reconnect_delay);
reconnect_delay_timer_init(s,
qemu_clock_get_ns(QEMU_CLOCK_REALTIME) +
s->reconnect_delay * NANOSECONDS_PER_SECOND);
}
- /*
- * Now we are sure that nobody is accessing the channel, and no one will
- * try until we set the state to CONNECTED.
- */
-
/* Finalize previous connection if any */
if (s->ioc) {
- qio_channel_detach_aio_context(QIO_CHANNEL(s->ioc));
yank_unregister_function(BLOCKDEV_YANK_INSTANCE(s->bs->node_name),
nbd_yank, s->bs);
object_unref(OBJECT(s->ioc));
s->ioc = NULL;
}
- nbd_co_do_establish_connection(s->bs, NULL);
+ qemu_mutex_unlock(&s->requests_lock);
+ 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);
/*
* The reconnect attempt is done (maybe successfully, maybe not), so
reconnect_delay_timer_del(s);
}
-static coroutine_fn int nbd_receive_replies(BDRVNBDState *s, uint64_t handle)
+static coroutine_fn int nbd_receive_replies(BDRVNBDState *s, uint64_t cookie,
+ Error **errp)
{
int ret;
- uint64_t ind = HANDLE_TO_INDEX(s, handle), ind2;
+ uint64_t ind = COOKIE_TO_INDEX(cookie), ind2;
QEMU_LOCK_GUARD(&s->receive_mutex);
while (true) {
- if (s->reply.handle == handle) {
+ if (s->reply.cookie == cookie) {
/* We are done */
return 0;
}
- if (!nbd_client_connected(s)) {
- return -EIO;
- }
-
- if (s->reply.handle != 0) {
+ if (s->reply.cookie != 0) {
/*
* Some other request is being handled now. It should already be
- * woken by whoever set s->reply.handle (or never wait in this
+ * woken by whoever set s->reply.cookie (or never wait in this
* yield). So, we should not wake it here.
*/
- ind2 = HANDLE_TO_INDEX(s, s->reply.handle);
+ ind2 = COOKIE_TO_INDEX(s->reply.cookie);
assert(!s->requests[ind2].receiving);
s->requests[ind].receiving = true;
qemu_coroutine_yield();
/*
- * We may be woken for 3 reasons:
+ * We may be woken for 2 reasons:
* 1. From this function, executing in parallel coroutine, when our
- * handle is received.
- * 2. From nbd_channel_error(), when connection is lost.
- * 3. From nbd_co_receive_one_chunk(), when previous request is
- * finished and s->reply.handle set to 0.
+ * cookie is received.
+ * 2. From nbd_co_receive_one_chunk(), when previous request is
+ * finished and s->reply.cookie set to 0.
* Anyway, it's OK to lock the mutex and go to the next iteration.
*/
continue;
}
- /* We are under mutex and handle is 0. We have to do the dirty work. */
- assert(s->reply.handle == 0);
- ret = nbd_receive_reply(s->bs, s->ioc, &s->reply, NULL);
- if (ret <= 0) {
- ret = ret ? ret : -EIO;
+ /* We are under mutex and cookie is 0. We have to do the dirty work. */
+ assert(s->reply.cookie == 0);
+ ret = nbd_receive_reply(s->bs, s->ioc, &s->reply, s->info.mode, errp);
+ if (ret == 0) {
+ ret = -EIO;
+ error_setg(errp, "server dropped connection");
+ }
+ if (ret < 0) {
nbd_channel_error(s, ret);
return ret;
}
- if (nbd_reply_is_structured(&s->reply) && !s->info.structured_reply) {
+ if (nbd_reply_is_structured(&s->reply) &&
+ s->info.mode < NBD_MODE_STRUCTURED) {
nbd_channel_error(s, -EINVAL);
+ error_setg(errp, "unexpected structured reply");
return -EINVAL;
}
- if (s->reply.handle == handle) {
- /* We are done */
- return 0;
- }
- ind2 = HANDLE_TO_INDEX(s, s->reply.handle);
- if (ind2 >= MAX_NBD_REQUESTS || !s->requests[ind2].reply_possible) {
+ ind2 = COOKIE_TO_INDEX(s->reply.cookie);
+ if (ind2 >= MAX_NBD_REQUESTS || !s->requests[ind2].coroutine) {
nbd_channel_error(s, -EINVAL);
+ error_setg(errp, "unexpected cookie value");
return -EINVAL;
}
+ if (s->reply.cookie == cookie) {
+ /* We are done */
+ return 0;
+ }
nbd_recv_coroutine_wake_one(&s->requests[ind2]);
}
}
-static int nbd_co_send_request(BlockDriverState *bs,
- NBDRequest *request,
- QEMUIOVector *qiov)
+static int coroutine_fn GRAPH_RDLOCK
+nbd_co_send_request(BlockDriverState *bs, NBDRequest *request,
+ QEMUIOVector *qiov)
{
BDRVNBDState *s = (BDRVNBDState *)bs->opaque;
int rc, i = -1;
- qemu_co_mutex_lock(&s->send_mutex);
-
+ qemu_mutex_lock(&s->requests_lock);
while (s->in_flight == MAX_NBD_REQUESTS ||
- (!nbd_client_connected(s) && s->in_flight > 0))
- {
- qemu_co_queue_wait(&s->free_sema, &s->send_mutex);
- }
-
- if (nbd_client_connecting(s)) {
- nbd_reconnect_attempt(s);
- }
-
- if (!nbd_client_connected(s)) {
- rc = -EIO;
- goto err;
+ (s->state != NBD_CLIENT_CONNECTED && s->in_flight > 0)) {
+ qemu_co_queue_wait(&s->free_sema, &s->requests_lock);
}
s->in_flight++;
+ if (s->state != NBD_CLIENT_CONNECTED) {
+ if (nbd_client_connecting(s)) {
+ nbd_reconnect_attempt(s);
+ qemu_co_queue_restart_all(&s->free_sema);
+ }
+ if (s->state != NBD_CLIENT_CONNECTED) {
+ rc = -EIO;
+ goto err;
+ }
+ }
for (i = 0; i < MAX_NBD_REQUESTS; i++) {
if (s->requests[i].coroutine == NULL) {
}
}
- g_assert(qemu_in_coroutine());
assert(i < MAX_NBD_REQUESTS);
-
s->requests[i].coroutine = qemu_coroutine_self();
s->requests[i].offset = request->from;
s->requests[i].receiving = false;
- s->requests[i].reply_possible = true;
+ qemu_mutex_unlock(&s->requests_lock);
- request->handle = INDEX_TO_HANDLE(s, i);
+ qemu_co_mutex_lock(&s->send_mutex);
+ request->cookie = INDEX_TO_COOKIE(i);
+ request->mode = s->info.mode;
assert(s->ioc);
if (qiov) {
qio_channel_set_cork(s->ioc, true);
rc = nbd_send_request(s->ioc, request);
- if (nbd_client_connected(s) && rc >= 0) {
- if (qio_channel_writev_all(s->ioc, qiov->iov, qiov->niov,
- NULL) < 0) {
- rc = -EIO;
- }
- } else if (rc >= 0) {
+ if (rc >= 0 && qio_channel_writev_all(s->ioc, qiov->iov, qiov->niov,
+ NULL) < 0) {
rc = -EIO;
}
qio_channel_set_cork(s->ioc, false);
} else {
rc = nbd_send_request(s->ioc, request);
}
+ qemu_co_mutex_unlock(&s->send_mutex);
-err:
if (rc < 0) {
- nbd_channel_error(s, rc);
+ qemu_mutex_lock(&s->requests_lock);
+err:
+ nbd_channel_error_locked(s, rc);
if (i != -1) {
s->requests[i].coroutine = NULL;
- s->in_flight--;
}
+ s->in_flight--;
qemu_co_queue_next(&s->free_sema);
+ qemu_mutex_unlock(&s->requests_lock);
}
- qemu_co_mutex_unlock(&s->send_mutex);
return rc;
}
*/
static int nbd_parse_blockstatus_payload(BDRVNBDState *s,
NBDStructuredReplyChunk *chunk,
- uint8_t *payload, uint64_t orig_length,
- NBDExtent *extent, Error **errp)
+ uint8_t *payload, bool wide,
+ uint64_t orig_length,
+ NBDExtent64 *extent, Error **errp)
{
uint32_t context_id;
+ uint32_t count;
+ size_t ext_len = wide ? sizeof(*extent) : sizeof(NBDExtent32);
+ size_t pay_len = sizeof(context_id) + wide * sizeof(count) + ext_len;
/* The server succeeded, so it must have sent [at least] one extent */
- if (chunk->length < sizeof(context_id) + sizeof(*extent)) {
+ if (chunk->length < pay_len) {
error_setg(errp, "Protocol error: invalid payload for "
"NBD_REPLY_TYPE_BLOCK_STATUS");
return -EINVAL;
return -EINVAL;
}
- extent->length = payload_advance32(&payload);
- extent->flags = payload_advance32(&payload);
+ if (wide) {
+ count = payload_advance32(&payload);
+ extent->length = payload_advance64(&payload);
+ extent->flags = payload_advance64(&payload);
+ } else {
+ count = 0;
+ extent->length = payload_advance32(&payload);
+ extent->flags = payload_advance32(&payload);
+ }
if (extent->length == 0) {
error_setg(errp, "Protocol error: server sent status chunk with "
* (always a safe status, even if it loses information).
*/
if (s->info.min_block && !QEMU_IS_ALIGNED(extent->length,
- s->info.min_block)) {
+ s->info.min_block)) {
trace_nbd_parse_blockstatus_compliance("extent length is unaligned");
if (extent->length > s->info.min_block) {
extent->length = QEMU_ALIGN_DOWN(extent->length,
/*
* We used NBD_CMD_FLAG_REQ_ONE, so the server should not have
* sent us any more than one extent, nor should it have included
- * status beyond our request in that extent. However, it's easy
- * enough to ignore the server's noncompliance without killing the
+ * status beyond our request in that extent. Furthermore, a wide
+ * server should have replied with an accurate count (we left
+ * count at 0 for a narrow server). However, it's easy enough to
+ * ignore the server's noncompliance without killing the
* connection; just ignore trailing extents, and clamp things to
* the length of our request.
*/
- if (chunk->length > sizeof(context_id) + sizeof(*extent)) {
- trace_nbd_parse_blockstatus_compliance("more than one extent");
+ if (count != wide || chunk->length > pay_len) {
+ trace_nbd_parse_blockstatus_compliance("unexpected extent count");
}
if (extent->length > orig_length) {
extent->length = orig_length;
return 0;
}
-static int nbd_co_receive_offset_data_payload(BDRVNBDState *s,
- uint64_t orig_offset,
- QEMUIOVector *qiov, Error **errp)
+static int coroutine_fn
+nbd_co_receive_offset_data_payload(BDRVNBDState *s, uint64_t orig_offset,
+ QEMUIOVector *qiov, Error **errp)
{
QEMUIOVector sub_qiov;
uint64_t offset;
* corresponding to the server's error reply), and errp is unchanged.
*/
static coroutine_fn int nbd_co_do_receive_one_chunk(
- BDRVNBDState *s, uint64_t handle, bool only_structured,
+ BDRVNBDState *s, uint64_t cookie, bool only_structured,
int *request_ret, QEMUIOVector *qiov, void **payload, Error **errp)
{
int ret;
- int i = HANDLE_TO_INDEX(s, handle);
+ int i = COOKIE_TO_INDEX(cookie);
void *local_payload = NULL;
NBDStructuredReplyChunk *chunk;
}
*request_ret = 0;
- nbd_receive_replies(s, handle);
- if (!nbd_client_connected(s)) {
- error_setg(errp, "Connection closed");
+ ret = nbd_receive_replies(s, cookie, errp);
+ if (ret < 0) {
+ error_prepend(errp, "Connection closed: ");
return -EIO;
}
assert(s->ioc);
- assert(s->reply.handle == handle);
+ assert(s->reply.cookie == cookie);
if (nbd_reply_is_simple(&s->reply)) {
if (only_structured) {
}
/* handle structured reply chunk */
- assert(s->info.structured_reply);
+ assert(s->info.mode >= NBD_MODE_STRUCTURED);
chunk = &s->reply.structured;
if (chunk->type == NBD_REPLY_TYPE_NONE) {
* Return value is a fatal error code or normal nbd reply error code
*/
static coroutine_fn int nbd_co_receive_one_chunk(
- BDRVNBDState *s, uint64_t handle, bool only_structured,
+ BDRVNBDState *s, uint64_t cookie, bool only_structured,
int *request_ret, QEMUIOVector *qiov, NBDReply *reply, void **payload,
Error **errp)
{
- int ret = nbd_co_do_receive_one_chunk(s, handle, only_structured,
+ int ret = nbd_co_do_receive_one_chunk(s, cookie, only_structured,
request_ret, qiov, payload, errp);
if (ret < 0) {
/* For assert at loop start in nbd_connection_entry */
*reply = s->reply;
}
- s->reply.handle = 0;
+ s->reply.cookie = 0;
- nbd_recv_coroutines_wake(s, false);
+ nbd_recv_coroutines_wake(s);
return ret;
}
* NBD_FOREACH_REPLY_CHUNK
* The pointer stored in @payload requires g_free() to free it.
*/
-#define NBD_FOREACH_REPLY_CHUNK(s, iter, handle, structured, \
+#define NBD_FOREACH_REPLY_CHUNK(s, iter, cookie, structured, \
qiov, reply, payload) \
for (iter = (NBDReplyChunkIter) { .only_structured = structured }; \
- nbd_reply_chunk_iter_receive(s, &iter, handle, qiov, reply, payload);)
+ nbd_reply_chunk_iter_receive(s, &iter, cookie, qiov, reply, payload);)
/*
* nbd_reply_chunk_iter_receive
* The pointer stored in @payload requires g_free() to free it.
*/
-static bool nbd_reply_chunk_iter_receive(BDRVNBDState *s,
- NBDReplyChunkIter *iter,
- uint64_t handle,
- QEMUIOVector *qiov, NBDReply *reply,
- void **payload)
+static bool coroutine_fn nbd_reply_chunk_iter_receive(BDRVNBDState *s,
+ NBDReplyChunkIter *iter,
+ uint64_t cookie,
+ QEMUIOVector *qiov,
+ NBDReply *reply,
+ void **payload)
{
int ret, request_ret;
NBDReply local_reply;
NBDStructuredReplyChunk *chunk;
Error *local_err = NULL;
- if (!nbd_client_connected(s)) {
- error_setg(&local_err, "Connection closed");
- nbd_iter_channel_error(iter, -EIO, &local_err);
- goto break_loop;
- }
if (iter->done) {
/* Previous iteration was last. */
reply = &local_reply;
}
- ret = nbd_co_receive_one_chunk(s, handle, iter->only_structured,
+ ret = nbd_co_receive_one_chunk(s, cookie, iter->only_structured,
&request_ret, qiov, reply, payload,
&local_err);
if (ret < 0) {
}
/* Do not execute the body of NBD_FOREACH_REPLY_CHUNK for simple reply. */
- if (nbd_reply_is_simple(reply) || !nbd_client_connected(s)) {
+ if (nbd_reply_is_simple(reply) || iter->ret < 0) {
goto break_loop;
}
return true;
break_loop:
- s->requests[HANDLE_TO_INDEX(s, handle)].coroutine = NULL;
-
- qemu_co_mutex_lock(&s->send_mutex);
+ qemu_mutex_lock(&s->requests_lock);
+ s->requests[COOKIE_TO_INDEX(cookie)].coroutine = NULL;
s->in_flight--;
qemu_co_queue_next(&s->free_sema);
- qemu_co_mutex_unlock(&s->send_mutex);
+ qemu_mutex_unlock(&s->requests_lock);
return false;
}
-static int nbd_co_receive_return_code(BDRVNBDState *s, uint64_t handle,
- int *request_ret, Error **errp)
+static int coroutine_fn
+nbd_co_receive_return_code(BDRVNBDState *s, uint64_t cookie,
+ int *request_ret, Error **errp)
{
NBDReplyChunkIter iter;
- NBD_FOREACH_REPLY_CHUNK(s, iter, handle, false, NULL, NULL, NULL) {
+ NBD_FOREACH_REPLY_CHUNK(s, iter, cookie, false, NULL, NULL, NULL) {
/* nbd_reply_chunk_iter_receive does all the work */
}
return iter.ret;
}
-static int nbd_co_receive_cmdread_reply(BDRVNBDState *s, uint64_t handle,
- uint64_t offset, QEMUIOVector *qiov,
- int *request_ret, Error **errp)
+static int coroutine_fn
+nbd_co_receive_cmdread_reply(BDRVNBDState *s, uint64_t cookie,
+ uint64_t offset, QEMUIOVector *qiov,
+ int *request_ret, Error **errp)
{
NBDReplyChunkIter iter;
NBDReply reply;
void *payload = NULL;
Error *local_err = NULL;
- NBD_FOREACH_REPLY_CHUNK(s, iter, handle, s->info.structured_reply,
+ NBD_FOREACH_REPLY_CHUNK(s, iter, cookie,
+ s->info.mode >= NBD_MODE_STRUCTURED,
qiov, &reply, &payload)
{
int ret;
return iter.ret;
}
-static int nbd_co_receive_blockstatus_reply(BDRVNBDState *s,
- uint64_t handle, uint64_t length,
- NBDExtent *extent,
- int *request_ret, Error **errp)
+static int coroutine_fn
+nbd_co_receive_blockstatus_reply(BDRVNBDState *s, uint64_t cookie,
+ uint64_t length, NBDExtent64 *extent,
+ int *request_ret, Error **errp)
{
NBDReplyChunkIter iter;
NBDReply reply;
bool received = false;
assert(!extent->length);
- NBD_FOREACH_REPLY_CHUNK(s, iter, handle, false, NULL, &reply, &payload) {
+ NBD_FOREACH_REPLY_CHUNK(s, iter, cookie, false, NULL, &reply, &payload) {
int ret;
NBDStructuredReplyChunk *chunk = &reply.structured;
+ bool wide;
assert(nbd_reply_is_structured(&reply));
switch (chunk->type) {
+ case NBD_REPLY_TYPE_BLOCK_STATUS_EXT:
case NBD_REPLY_TYPE_BLOCK_STATUS:
+ wide = chunk->type == NBD_REPLY_TYPE_BLOCK_STATUS_EXT;
+ if ((s->info.mode >= NBD_MODE_EXTENDED) != wide) {
+ trace_nbd_extended_headers_compliance("block_status");
+ }
if (received) {
nbd_channel_error(s, -EINVAL);
error_setg(&local_err, "Several BLOCK_STATUS chunks in reply");
}
received = true;
- ret = nbd_parse_blockstatus_payload(s, &reply.structured,
- payload, length, extent,
- &local_err);
+ ret = nbd_parse_blockstatus_payload(
+ s, &reply.structured, payload, wide,
+ length, extent, &local_err);
if (ret < 0) {
nbd_channel_error(s, ret);
nbd_iter_channel_error(&iter, ret, &local_err);
return iter.ret;
}
-static int nbd_co_request(BlockDriverState *bs, NBDRequest *request,
- QEMUIOVector *write_qiov)
+static int coroutine_fn GRAPH_RDLOCK
+nbd_co_request(BlockDriverState *bs, NBDRequest *request,
+ QEMUIOVector *write_qiov)
{
int ret, request_ret;
Error *local_err = NULL;
continue;
}
- ret = nbd_co_receive_return_code(s, request->handle,
+ ret = nbd_co_receive_return_code(s, request->cookie,
&request_ret, &local_err);
if (local_err) {
trace_nbd_co_request_fail(request->from, request->len,
- request->handle, request->flags,
+ request->cookie, request->flags,
request->type,
nbd_cmd_lookup(request->type),
ret, error_get_pretty(local_err));
error_free(local_err);
local_err = NULL;
}
- } while (ret < 0 && nbd_client_connecting_wait(s));
+ } while (ret < 0 && nbd_client_will_reconnect(s));
return ret ? ret : request_ret;
}
-static int nbd_client_co_preadv(BlockDriverState *bs, int64_t offset,
- int64_t bytes, QEMUIOVector *qiov,
- BdrvRequestFlags flags)
+static int coroutine_fn GRAPH_RDLOCK
+nbd_client_co_preadv(BlockDriverState *bs, int64_t offset, int64_t bytes,
+ QEMUIOVector *qiov, BdrvRequestFlags flags)
{
int ret, request_ret;
Error *local_err = NULL;
};
assert(bytes <= NBD_MAX_BUFFER_SIZE);
- assert(!flags);
if (!bytes) {
return 0;
continue;
}
- ret = nbd_co_receive_cmdread_reply(s, request.handle, offset, qiov,
+ ret = nbd_co_receive_cmdread_reply(s, request.cookie, offset, qiov,
&request_ret, &local_err);
if (local_err) {
- trace_nbd_co_request_fail(request.from, request.len, request.handle,
+ trace_nbd_co_request_fail(request.from, request.len, request.cookie,
request.flags, request.type,
nbd_cmd_lookup(request.type),
ret, error_get_pretty(local_err));
error_free(local_err);
local_err = NULL;
}
- } while (ret < 0 && nbd_client_connecting_wait(s));
+ } while (ret < 0 && nbd_client_will_reconnect(s));
return ret ? ret : request_ret;
}
-static int nbd_client_co_pwritev(BlockDriverState *bs, int64_t offset,
- int64_t bytes, QEMUIOVector *qiov,
- BdrvRequestFlags flags)
+static int coroutine_fn GRAPH_RDLOCK
+nbd_client_co_pwritev(BlockDriverState *bs, int64_t offset, int64_t bytes,
+ QEMUIOVector *qiov, BdrvRequestFlags flags)
{
BDRVNBDState *s = (BDRVNBDState *)bs->opaque;
NBDRequest request = {
return nbd_co_request(bs, &request, qiov);
}
-static int nbd_client_co_pwrite_zeroes(BlockDriverState *bs, int64_t offset,
- int64_t bytes, BdrvRequestFlags flags)
+static int coroutine_fn GRAPH_RDLOCK
+nbd_client_co_pwrite_zeroes(BlockDriverState *bs, int64_t offset, int64_t bytes,
+ BdrvRequestFlags flags)
{
BDRVNBDState *s = (BDRVNBDState *)bs->opaque;
NBDRequest request = {
.type = NBD_CMD_WRITE_ZEROES,
.from = offset,
- .len = bytes, /* .len is uint32_t actually */
+ .len = bytes,
};
- assert(bytes <= UINT32_MAX); /* rely on max_pwrite_zeroes */
+ /* rely on max_pwrite_zeroes */
+ assert(bytes <= UINT32_MAX || s->info.mode >= NBD_MODE_EXTENDED);
assert(!(s->info.flags & NBD_FLAG_READ_ONLY));
if (!(s->info.flags & NBD_FLAG_SEND_WRITE_ZEROES)) {
return nbd_co_request(bs, &request, NULL);
}
-static int nbd_client_co_flush(BlockDriverState *bs)
+static int coroutine_fn GRAPH_RDLOCK nbd_client_co_flush(BlockDriverState *bs)
{
BDRVNBDState *s = (BDRVNBDState *)bs->opaque;
NBDRequest request = { .type = NBD_CMD_FLUSH };
return nbd_co_request(bs, &request, NULL);
}
-static int nbd_client_co_pdiscard(BlockDriverState *bs, int64_t offset,
- int64_t bytes)
+static int coroutine_fn GRAPH_RDLOCK
+nbd_client_co_pdiscard(BlockDriverState *bs, int64_t offset, int64_t bytes)
{
BDRVNBDState *s = (BDRVNBDState *)bs->opaque;
NBDRequest request = {
.type = NBD_CMD_TRIM,
.from = offset,
- .len = bytes, /* len is uint32_t */
+ .len = bytes,
};
- assert(bytes <= UINT32_MAX); /* rely on max_pdiscard */
+ /* rely on max_pdiscard */
+ assert(bytes <= UINT32_MAX || s->info.mode >= NBD_MODE_EXTENDED);
assert(!(s->info.flags & NBD_FLAG_READ_ONLY));
if (!(s->info.flags & NBD_FLAG_SEND_TRIM) || !bytes) {
return nbd_co_request(bs, &request, NULL);
}
-static int coroutine_fn nbd_client_co_block_status(
+static int coroutine_fn GRAPH_RDLOCK nbd_client_co_block_status(
BlockDriverState *bs, bool want_zero, int64_t offset, int64_t bytes,
int64_t *pnum, int64_t *map, BlockDriverState **file)
{
int ret, request_ret;
- NBDExtent extent = { 0 };
+ NBDExtent64 extent = { 0 };
BDRVNBDState *s = (BDRVNBDState *)bs->opaque;
Error *local_err = NULL;
NBDRequest request = {
.type = NBD_CMD_BLOCK_STATUS,
.from = offset,
- .len = MIN(QEMU_ALIGN_DOWN(INT_MAX, bs->bl.request_alignment),
- MIN(bytes, s->info.size - offset)),
+ .len = MIN(bytes, s->info.size - offset),
.flags = NBD_CMD_FLAG_REQ_ONE,
};
*file = bs;
return BDRV_BLOCK_DATA | BDRV_BLOCK_OFFSET_VALID;
}
+ if (s->info.mode < NBD_MODE_EXTENDED) {
+ request.len = MIN(QEMU_ALIGN_DOWN(INT_MAX, bs->bl.request_alignment),
+ request.len);
+ }
/*
* Work around the fact that the block layer doesn't do
continue;
}
- ret = nbd_co_receive_blockstatus_reply(s, request.handle, bytes,
+ ret = nbd_co_receive_blockstatus_reply(s, request.cookie, bytes,
&extent, &request_ret,
&local_err);
if (local_err) {
- trace_nbd_co_request_fail(request.from, request.len, request.handle,
+ trace_nbd_co_request_fail(request.from, request.len, request.cookie,
request.flags, request.type,
nbd_cmd_lookup(request.type),
ret, error_get_pretty(local_err));
error_free(local_err);
local_err = NULL;
}
- } while (ret < 0 && nbd_client_connecting_wait(s));
+ } while (ret < 0 && nbd_client_will_reconnect(s));
if (ret < 0 || request_ret < 0) {
return ret ? ret : request_ret;
BlockDriverState *bs = opaque;
BDRVNBDState *s = (BDRVNBDState *)bs->opaque;
- qatomic_store_release(&s->state, NBD_CLIENT_QUIT);
- qio_channel_shutdown(QIO_CHANNEL(s->ioc), QIO_CHANNEL_SHUTDOWN_BOTH, NULL);
+ QEMU_LOCK_GUARD(&s->requests_lock);
+ qio_channel_shutdown(s->ioc, QIO_CHANNEL_SHUTDOWN_BOTH, NULL);
+ s->state = NBD_CLIENT_QUIT;
}
static void nbd_client_close(BlockDriverState *bs)
{
BDRVNBDState *s = (BDRVNBDState *)bs->opaque;
- NBDRequest request = { .type = NBD_CMD_DISC };
+ NBDRequest request = { .type = NBD_CMD_DISC, .mode = s->info.mode };
if (s->ioc) {
nbd_send_request(s->ioc, &request);
BDRVNBDState *s = (BDRVNBDState *)bs->opaque;
s->bs = bs;
- qemu_co_mutex_init(&s->send_mutex);
+ qemu_mutex_init(&s->requests_lock);
qemu_co_queue_init(&s->free_sema);
+ qemu_co_mutex_init(&s->send_mutex);
qemu_co_mutex_init(&s->receive_mutex);
if (!yank_register_instance(BLOCKDEV_YANK_INSTANCE(bs->node_name), errp)) {
}
s->state = NBD_CLIENT_CONNECTING_WAIT;
- ret = nbd_do_establish_connection(bs, errp);
+ ret = nbd_do_establish_connection(bs, true, errp);
if (ret < 0) {
goto fail;
}
return ret;
}
-static int nbd_co_flush(BlockDriverState *bs)
-{
- return nbd_client_co_flush(bs);
-}
-
static void nbd_refresh_limits(BlockDriverState *bs, Error **errp)
{
BDRVNBDState *s = (BDRVNBDState *)bs->opaque;
bs->bl.max_pwrite_zeroes = max;
bs->bl.max_transfer = max;
+ /*
+ * Assume that if the server supports extended headers, it also
+ * supports unlimited size zero and trim commands.
+ */
+ if (s->info.mode >= NBD_MODE_EXTENDED) {
+ bs->bl.max_pdiscard = bs->bl.max_pwrite_zeroes = 0;
+ }
+
if (s->info.opt_block &&
s->info.opt_block > bs->bl.opt_transfer) {
bs->bl.opt_transfer = s->info.opt_block;
return 0;
}
-static int64_t nbd_getlength(BlockDriverState *bs)
+static int64_t coroutine_fn nbd_co_getlength(BlockDriverState *bs)
{
BDRVNBDState *s = bs->opaque;
reconnect_delay_timer_del(s);
+ qemu_mutex_lock(&s->requests_lock);
if (s->state == NBD_CLIENT_CONNECTING_WAIT) {
s->state = NBD_CLIENT_CONNECTING_NOWAIT;
- qemu_co_queue_restart_all(&s->free_sema);
}
+ qemu_mutex_unlock(&s->requests_lock);
nbd_co_establish_connection_cancel(s->conn);
}
* the reconnect_delay_timer cannot be active here.
*/
assert(!s->reconnect_delay_timer);
-
- if (s->ioc) {
- qio_channel_attach_aio_context(s->ioc, new_context);
- }
}
static void nbd_detach_aio_context(BlockDriverState *bs)
assert(!s->open_timer);
assert(!s->reconnect_delay_timer);
-
- if (s->ioc) {
- qio_channel_detach_aio_context(s->ioc);
- }
}
static BlockDriver bdrv_nbd = {
.bdrv_co_pwritev = nbd_client_co_pwritev,
.bdrv_co_pwrite_zeroes = nbd_client_co_pwrite_zeroes,
.bdrv_close = nbd_close,
- .bdrv_co_flush_to_os = nbd_co_flush,
+ .bdrv_co_flush_to_os = nbd_client_co_flush,
.bdrv_co_pdiscard = nbd_client_co_pdiscard,
.bdrv_refresh_limits = nbd_refresh_limits,
.bdrv_co_truncate = nbd_co_truncate,
- .bdrv_getlength = nbd_getlength,
+ .bdrv_co_getlength = nbd_co_getlength,
.bdrv_refresh_filename = nbd_refresh_filename,
.bdrv_co_block_status = nbd_client_co_block_status,
.bdrv_dirname = nbd_dirname,
.bdrv_co_pwritev = nbd_client_co_pwritev,
.bdrv_co_pwrite_zeroes = nbd_client_co_pwrite_zeroes,
.bdrv_close = nbd_close,
- .bdrv_co_flush_to_os = nbd_co_flush,
+ .bdrv_co_flush_to_os = nbd_client_co_flush,
.bdrv_co_pdiscard = nbd_client_co_pdiscard,
.bdrv_refresh_limits = nbd_refresh_limits,
.bdrv_co_truncate = nbd_co_truncate,
- .bdrv_getlength = nbd_getlength,
+ .bdrv_co_getlength = nbd_co_getlength,
.bdrv_refresh_filename = nbd_refresh_filename,
.bdrv_co_block_status = nbd_client_co_block_status,
.bdrv_dirname = nbd_dirname,
.bdrv_co_pwritev = nbd_client_co_pwritev,
.bdrv_co_pwrite_zeroes = nbd_client_co_pwrite_zeroes,
.bdrv_close = nbd_close,
- .bdrv_co_flush_to_os = nbd_co_flush,
+ .bdrv_co_flush_to_os = nbd_client_co_flush,
.bdrv_co_pdiscard = nbd_client_co_pdiscard,
.bdrv_refresh_limits = nbd_refresh_limits,
.bdrv_co_truncate = nbd_co_truncate,
- .bdrv_getlength = nbd_getlength,
+ .bdrv_co_getlength = nbd_co_getlength,
.bdrv_refresh_filename = nbd_refresh_filename,
.bdrv_co_block_status = nbd_client_co_block_status,
.bdrv_dirname = nbd_dirname,
projects / mirror_qemu.git / blobdiff