OpenE2K
/
qemu-e2k
13
4
Fork 0
Code Issues 4 Pull Requests Projects 2 Releases Activity

nbd patches for 2019-02-04 

- deprecate 'qemu-nbd --partition'
 - preparation for NBD reconnect, including better logging of read errors
 -----BEGIN PGP SIGNATURE-----
 
 iQEcBAABCAAGBQJcWQggAAoJEKeha0olJ0NqC70H/2hH64TdfVQzEviOjCfB/8cD
 srM0M98dZyy/r5M0lA289z5edgEgURMX2dPR4U/LeKXvQGNWkLjIUTxeM9Wh/znS
 yIlGhvbl3CH4i6QHh0ZOTo3l5FSqaTag+YoLQEEeSRGHXKPnlfZ28kudazJgTsvw
 82TaCZqZsWLXEk02PsViVb0rCMF4D3CEjJL4yCI9M0NAu9YgLD+4lsqCE72N3QfN
 4tIgg7nSo5HgbIsAcBcJXOomc9dxi9XZzVDb2GTvsf0N019GuGXmWHh32xvZohhh
 IuYe2B4M+FvooTbDEPXoeTe40XQC6wk5zLeRHcN7I/NF9zW6paxIpSK5on+D+Ag=
 =YUga
 -----END PGP SIGNATURE-----

Merge remote-tracking branch 'remotes/ericb/tags/pull-nbd-2019-02-04' into staging

nbd patches for 2019-02-04

- deprecate 'qemu-nbd --partition'
- preparation for NBD reconnect, including better logging of read errors

# gpg: Signature made Tue 05 Feb 2019 03:50:56 GMT
# gpg:                using RSA key A7A16B4A2527436A
# gpg: Good signature from "Eric Blake <eblake@redhat.com>" [full]
# gpg:                 aka "Eric Blake (Free Software Programmer) <ebb9@byu.net>" [full]
# gpg:                 aka "[jpeg image of size 6874]" [full]
# Primary key fingerprint: 71C2 CC22 B1C4 6029 27D2  F3AA A7A1 6B4A 2527 436A

* remotes/ericb/tags/pull-nbd-2019-02-04:
  block/nbd-client: rename read_reply_co to connection_co
  block/nbd-client: don't check ioc
  block/nbd-client: fix nbd_reply_chunk_iter_receive
  block/nbd-client: split connection from initialization
  block/nbd: move connection code from block/nbd to block/nbd-client
  block/nbd-client: split channel errors from export errors
  nbd: generalize usage of nbd_read
  qemu-nbd: Deprecate qemu-nbd --partition

Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
This commit is contained in:
Peter Maydell 2019-02-05 12:46:18 +00:00
parent 9ae805637a bc5a03350c
commit 1c3d45df5e
10 changed files with 239 additions and 196 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

199
block/nbd-client.c
View File

@ -53,15 +53,13 @@ static void nbd_teardown_connection(BlockDriverState *bs)
{
NBDClientSession *client = nbd_get_client_session(bs);
if (!client->ioc) { /* Already closed */
return;
}
assert(client->ioc);
/* finish any pending coroutines */
qio_channel_shutdown(client->ioc,
QIO_CHANNEL_SHUTDOWN_BOTH,
NULL);
BDRV_POLL_WHILE(bs, client->read_reply_co);
BDRV_POLL_WHILE(bs, client->connection_co);
nbd_client_detach_aio_context(bs);
object_unref(OBJECT(client->sioc));
@ -70,7 +68,7 @@ static void nbd_teardown_connection(BlockDriverState *bs)
client->ioc = NULL;
}
static coroutine_fn void nbd_read_reply_entry(void *opaque)
static coroutine_fn void nbd_connection_entry(void *opaque)
{
NBDClientSession *s = opaque;
uint64_t i;
@ -102,14 +100,14 @@ static coroutine_fn void nbd_read_reply_entry(void *opaque)
}
/* We're woken up again by the request itself. Note that there
* is no race between yielding and reentering read_reply_co. This
* is no race between yielding and reentering connection_co. This
* is because:
*
* - if the request runs on the same AioContext, it is only
* entered after we yield
*
* - if the request runs on a different AioContext, reentering
* read_reply_co happens through a bottom half, which can only
* connection_co happens through a bottom half, which can only
* run after we yield.
*/
aio_co_wake(s->requests[i].coroutine);
@ -118,7 +116,7 @@ static coroutine_fn void nbd_read_reply_entry(void *opaque)
s->quit = true;
nbd_recv_coroutines_wake_all(s);
s->read_reply_co = NULL;
s->connection_co = NULL;
aio_wait_kick();
}
@ -154,10 +152,7 @@ static int nbd_co_send_request(BlockDriverState *bs,
rc = -EIO;
goto err;
}
if (!s->ioc) {
rc = -EPIPE;
goto err;
}
assert(s->ioc);
if (qiov) {
qio_channel_set_cork(s->ioc, true);
@ -338,10 +333,9 @@ static int nbd_co_receive_offset_data_payload(NBDClientSession *s,
return -EINVAL;
}
if (nbd_read(s->ioc, &offset, sizeof(offset), errp) < 0) {
if (nbd_read64(s->ioc, &offset, "OFFSET_DATA offset", errp) < 0) {
return -EIO;
}
be64_to_cpus(&offset);
data_size = chunk->length - sizeof(offset);
assert(data_size);
@ -388,7 +382,7 @@ static coroutine_fn int nbd_co_receive_structured_payload(
}
*payload = g_new(char, len);
ret = nbd_read(s->ioc, *payload, len, errp);
ret = nbd_read(s->ioc, *payload, len, "structured payload", errp);
if (ret < 0) {
g_free(*payload);
*payload = NULL;
@ -426,14 +420,15 @@ static coroutine_fn int nbd_co_do_receive_one_chunk(
}
*request_ret = 0;
/* Wait until we're woken up by nbd_read_reply_entry. */
/* Wait until we're woken up by nbd_connection_entry. */
s->requests[i].receiving = true;
qemu_coroutine_yield();
s->requests[i].receiving = false;
if (!s->ioc || s->quit) {
if (s->quit) {
error_setg(errp, "Connection closed");
return -EIO;
}
assert(s->ioc);
assert(s->reply.handle == handle);
@ -500,30 +495,29 @@ static coroutine_fn int nbd_co_do_receive_one_chunk(
}
/* nbd_co_receive_one_chunk
* Read reply, wake up read_reply_co and set s->quit if needed.
* Read reply, wake up connection_co and set s->quit if needed.
* Return value is a fatal error code or normal nbd reply error code
*/
static coroutine_fn int nbd_co_receive_one_chunk(
NBDClientSession *s, uint64_t handle, bool only_structured,
QEMUIOVector *qiov, NBDReply *reply, void **payload, Error **errp)
int *request_ret, QEMUIOVector *qiov, NBDReply *reply, void **payload,
Error **errp)
{
int request_ret;
int ret = nbd_co_do_receive_one_chunk(s, handle, only_structured,
&request_ret, qiov, payload, errp);
request_ret, qiov, payload, errp);
if (ret < 0) {
s->quit = true;
} else {
/* For assert at loop start in nbd_read_reply_entry */
/* For assert at loop start in nbd_connection_entry */
if (reply) {
*reply = s->reply;
}
s->reply.handle = 0;
ret = request_ret;
}
if (s->read_reply_co) {
aio_co_wake(s->read_reply_co);
if (s->connection_co) {
aio_co_wake(s->connection_co);
}
return ret;
@ -531,22 +525,17 @@ static coroutine_fn int nbd_co_receive_one_chunk(
typedef struct NBDReplyChunkIter {
int ret;
bool fatal;
int request_ret;
Error *err;
bool done, only_structured;
} NBDReplyChunkIter;
static void nbd_iter_error(NBDReplyChunkIter *iter, bool fatal,
int ret, Error **local_err)
static void nbd_iter_channel_error(NBDReplyChunkIter *iter,
int ret, Error **local_err)
{
assert(ret < 0);
if ((fatal && !iter->fatal) || iter->ret == 0) {
if (iter->ret != 0) {
error_free(iter->err);
iter->err = NULL;
}
iter->fatal = fatal;
if (!iter->ret) {
iter->ret = ret;
error_propagate(&iter->err, *local_err);
} else {
@ -556,6 +545,15 @@ static void nbd_iter_error(NBDReplyChunkIter *iter, bool fatal,
*local_err = NULL;
}
static void nbd_iter_request_error(NBDReplyChunkIter *iter, int ret)
{
assert(ret < 0);
if (!iter->request_ret) {
iter->request_ret = ret;
}
}
/* NBD_FOREACH_REPLY_CHUNK
*/
#define NBD_FOREACH_REPLY_CHUNK(s, iter, handle, structured, \
@ -571,13 +569,13 @@ static bool nbd_reply_chunk_iter_receive(NBDClientSession *s,
QEMUIOVector *qiov, NBDReply *reply,
void **payload)
{
int ret;
int ret, request_ret;
NBDReply local_reply;
NBDStructuredReplyChunk *chunk;
Error *local_err = NULL;
if (s->quit) {
error_setg(&local_err, "Connection closed");
nbd_iter_error(iter, true, -EIO, &local_err);
nbd_iter_channel_error(iter, -EIO, &local_err);
goto break_loop;
}
@ -591,14 +589,16 @@ static bool nbd_reply_chunk_iter_receive(NBDClientSession *s,
}
ret = nbd_co_receive_one_chunk(s, handle, iter->only_structured,
qiov, reply, payload, &local_err);
&request_ret, qiov, reply, payload,
&local_err);
if (ret < 0) {
/* If it is a fatal error s->quit is set by nbd_co_receive_one_chunk */
nbd_iter_error(iter, s->quit, ret, &local_err);
nbd_iter_channel_error(iter, ret, &local_err);
} else if (request_ret < 0) {
nbd_iter_request_error(iter, request_ret);
}
/* Do not execute the body of NBD_FOREACH_REPLY_CHUNK for simple reply. */
if (nbd_reply_is_simple(&s->reply) || s->quit) {
if (nbd_reply_is_simple(reply) || s->quit) {
goto break_loop;
}
@ -631,7 +631,7 @@ break_loop:
}
static int nbd_co_receive_return_code(NBDClientSession *s, uint64_t handle,
Error **errp)
int *request_ret, Error **errp)
{
NBDReplyChunkIter iter;
@ -640,12 +640,13 @@ static int nbd_co_receive_return_code(NBDClientSession *s, uint64_t handle,
}
error_propagate(errp, iter.err);
*request_ret = iter.request_ret;
return iter.ret;
}
static int nbd_co_receive_cmdread_reply(NBDClientSession *s, uint64_t handle,
uint64_t offset, QEMUIOVector *qiov,
Error **errp)
int *request_ret, Error **errp)
{
NBDReplyChunkIter iter;
NBDReply reply;
@ -670,7 +671,7 @@ static int nbd_co_receive_cmdread_reply(NBDClientSession *s, uint64_t handle,
offset, qiov, &local_err);
if (ret < 0) {
s->quit = true;
nbd_iter_error(&iter, true, ret, &local_err);
nbd_iter_channel_error(&iter, ret, &local_err);
}
break;
default:
@ -680,7 +681,7 @@ static int nbd_co_receive_cmdread_reply(NBDClientSession *s, uint64_t handle,
error_setg(&local_err,
"Unexpected reply type: %d (%s) for CMD_READ",
chunk->type, nbd_reply_type_lookup(chunk->type));
nbd_iter_error(&iter, true, -EINVAL, &local_err);
nbd_iter_channel_error(&iter, -EINVAL, &local_err);
}
}
@ -689,12 +690,14 @@ static int nbd_co_receive_cmdread_reply(NBDClientSession *s, uint64_t handle,
}
error_propagate(errp, iter.err);
*request_ret = iter.request_ret;
return iter.ret;
}
static int nbd_co_receive_blockstatus_reply(NBDClientSession *s,
uint64_t handle, uint64_t length,
NBDExtent *extent, Error **errp)
NBDExtent *extent,
int *request_ret, Error **errp)
{
NBDReplyChunkIter iter;
NBDReply reply;
@ -716,7 +719,7 @@ static int nbd_co_receive_blockstatus_reply(NBDClientSession *s,
if (received) {
s->quit = true;
error_setg(&local_err, "Several BLOCK_STATUS chunks in reply");
nbd_iter_error(&iter, true, -EINVAL, &local_err);
nbd_iter_channel_error(&iter, -EINVAL, &local_err);
}
received = true;
@ -725,7 +728,7 @@ static int nbd_co_receive_blockstatus_reply(NBDClientSession *s,
&local_err);
if (ret < 0) {
s->quit = true;
nbd_iter_error(&iter, true, ret, &local_err);
nbd_iter_channel_error(&iter, ret, &local_err);
}
break;
default:
@ -735,7 +738,7 @@ static int nbd_co_receive_blockstatus_reply(NBDClientSession *s,
"Unexpected reply type: %d (%s) "
"for CMD_BLOCK_STATUS",
chunk->type, nbd_reply_type_lookup(chunk->type));
nbd_iter_error(&iter, true, -EINVAL, &local_err);
nbd_iter_channel_error(&iter, -EINVAL, &local_err);
}
}
@ -750,14 +753,16 @@ static int nbd_co_receive_blockstatus_reply(NBDClientSession *s,
iter.ret = -EIO;
}
}
error_propagate(errp, iter.err);
*request_ret = iter.request_ret;
return iter.ret;
}
static int nbd_co_request(BlockDriverState *bs, NBDRequest *request,
QEMUIOVector *write_qiov)
{
int ret;
int ret, request_ret;
Error *local_err = NULL;
NBDClientSession *client = nbd_get_client_session(bs);
@ -773,7 +778,8 @@ static int nbd_co_request(BlockDriverState *bs, NBDRequest *request,
return ret;
}
ret = nbd_co_receive_return_code(client, request->handle, &local_err);
ret = nbd_co_receive_return_code(client, request->handle,
&request_ret, &local_err);
if (local_err) {
trace_nbd_co_request_fail(request->from, request->len, request->handle,
request->flags, request->type,
@ -781,13 +787,13 @@ static int nbd_co_request(BlockDriverState *bs, NBDRequest *request,
ret, error_get_pretty(local_err));
error_free(local_err);
}
return ret;
return ret ? ret : request_ret;
}
int nbd_client_co_preadv(BlockDriverState *bs, uint64_t offset,
uint64_t bytes, QEMUIOVector *qiov, int flags)
{
int ret;
int ret, request_ret;
Error *local_err = NULL;
NBDClientSession *client = nbd_get_client_session(bs);
NBDRequest request = {
@ -808,7 +814,7 @@ int nbd_client_co_preadv(BlockDriverState *bs, uint64_t offset,
}
ret = nbd_co_receive_cmdread_reply(client, request.handle, offset, qiov,
&local_err);
&request_ret, &local_err);
if (local_err) {
trace_nbd_co_request_fail(request.from, request.len, request.handle,
request.flags, request.type,
@ -816,7 +822,7 @@ int nbd_client_co_preadv(BlockDriverState *bs, uint64_t offset,
ret, error_get_pretty(local_err));
error_free(local_err);
}
return ret;
return ret ? ret : request_ret;
}
int nbd_client_co_pwritev(BlockDriverState *bs, uint64_t offset,
@ -910,7 +916,7 @@ int coroutine_fn nbd_client_co_block_status(BlockDriverState *bs,
int64_t *pnum, int64_t *map,
BlockDriverState **file)
{
int64_t ret;
int ret, request_ret;
NBDExtent extent = { 0 };
NBDClientSession *client = nbd_get_client_session(bs);
Error *local_err = NULL;
@ -935,7 +941,7 @@ int coroutine_fn nbd_client_co_block_status(BlockDriverState *bs,
}
ret = nbd_co_receive_blockstatus_reply(client, request.handle, bytes,
&extent, &local_err);
&extent, &request_ret, &local_err);
if (local_err) {
trace_nbd_co_request_fail(request.from, request.len, request.handle,
request.flags, request.type,
@ -943,8 +949,8 @@ int coroutine_fn nbd_client_co_block_status(BlockDriverState *bs,
ret, error_get_pretty(local_err));
error_free(local_err);
}
if (ret < 0) {
return ret;
if (ret < 0 || request_ret < 0) {
return ret ? ret : request_ret;
}
assert(extent.length);
@ -964,7 +970,7 @@ void nbd_client_attach_aio_context(BlockDriverState *bs,
{
NBDClientSession *client = nbd_get_client_session(bs);
qio_channel_attach_aio_context(QIO_CHANNEL(client->ioc), new_context);
aio_co_schedule(new_context, client->read_reply_co);
aio_co_schedule(new_context, client->connection_co);
}
void nbd_client_close(BlockDriverState *bs)
@ -972,26 +978,55 @@ void nbd_client_close(BlockDriverState *bs)
NBDClientSession *client = nbd_get_client_session(bs);
NBDRequest request = { .type = NBD_CMD_DISC };
if (client->ioc == NULL) {
return;
}
assert(client->ioc);
nbd_send_request(client->ioc, &request);
nbd_teardown_connection(bs);
}
int nbd_client_init(BlockDriverState *bs,
QIOChannelSocket *sioc,
const char *export,
QCryptoTLSCreds *tlscreds,
const char *hostname,
const char *x_dirty_bitmap,
Error **errp)
static QIOChannelSocket *nbd_establish_connection(SocketAddress *saddr,
Error **errp)
{
QIOChannelSocket *sioc;
Error *local_err = NULL;
sioc = qio_channel_socket_new();
qio_channel_set_name(QIO_CHANNEL(sioc), "nbd-client");
qio_channel_socket_connect_sync(sioc, saddr, &local_err);
if (local_err) {
object_unref(OBJECT(sioc));
error_propagate(errp, local_err);
return NULL;
}
qio_channel_set_delay(QIO_CHANNEL(sioc), false);
return sioc;
}
static int nbd_client_connect(BlockDriverState *bs,
SocketAddress *saddr,
const char *export,
QCryptoTLSCreds *tlscreds,
const char *hostname,
const char *x_dirty_bitmap,
Error **errp)
{
NBDClientSession *client = nbd_get_client_session(bs);
int ret;
/*
* establish TCP connection, return error if it fails
* TODO: Configurable retry-until-timeout behaviour.
*/
QIOChannelSocket *sioc = nbd_establish_connection(saddr, errp);
if (!sioc) {
return -ECONNREFUSED;
}
/* NBD handshake */
logout("session init %s\n", export);
qio_channel_set_blocking(QIO_CHANNEL(sioc), true, NULL);
@ -1007,6 +1042,7 @@ int nbd_client_init(BlockDriverState *bs,
g_free(client->info.name);
if (ret < 0) {
logout("Failed to negotiate with the NBD server\n");
object_unref(OBJECT(sioc));
return ret;
}
if (x_dirty_bitmap && !client->info.base_allocation) {
@ -1029,10 +1065,7 @@ int nbd_client_init(BlockDriverState *bs,
bs->supported_zero_flags |= BDRV_REQ_MAY_UNMAP;
}
qemu_co_mutex_init(&client->send_mutex);
qemu_co_queue_init(&client->free_sema);
client->sioc = sioc;
object_ref(OBJECT(client->sioc));
if (!client->ioc) {
client->ioc = QIO_CHANNEL(sioc);
@ -1042,7 +1075,7 @@ int nbd_client_init(BlockDriverState *bs,
/* Now that we're connected, set the socket to be non-blocking and
* kick the reply mechanism. */
qio_channel_set_blocking(QIO_CHANNEL(sioc), false, NULL);
client->read_reply_co = qemu_coroutine_create(nbd_read_reply_entry, client);
client->connection_co = qemu_coroutine_create(nbd_connection_entry, client);
nbd_client_attach_aio_context(bs, bdrv_get_aio_context(bs));
logout("Established connection with NBD server\n");
@ -1058,6 +1091,26 @@ int nbd_client_init(BlockDriverState *bs,
NBDRequest request = { .type = NBD_CMD_DISC };
nbd_send_request(client->ioc ?: QIO_CHANNEL(sioc), &request);
object_unref(OBJECT(sioc));
return ret;
}
}
int nbd_client_init(BlockDriverState *bs,
SocketAddress *saddr,
const char *export,
QCryptoTLSCreds *tlscreds,
const char *hostname,
const char *x_dirty_bitmap,
Error **errp)
{
NBDClientSession *client = nbd_get_client_session(bs);
qemu_co_mutex_init(&client->send_mutex);
qemu_co_queue_init(&client->free_sema);
return nbd_client_connect(bs, saddr, export, tlscreds, hostname,
x_dirty_bitmap, errp);
}

6
block/nbd-client.h
View File

@ -20,7 +20,7 @@
typedef struct {
Coroutine *coroutine;
uint64_t offset; /* original offset of the request */
bool receiving; /* waiting for read_reply_co? */
bool receiving; /* waiting for connection_co? */
} NBDClientRequest;
typedef struct NBDClientSession {
@ -30,7 +30,7 @@ typedef struct NBDClientSession {
CoMutex send_mutex;
CoQueue free_sema;
Coroutine *read_reply_co;
Coroutine *connection_co;
int in_flight;
NBDClientRequest requests[MAX_NBD_REQUESTS];
@ -41,7 +41,7 @@ typedef struct NBDClientSession {
NBDClientSession *nbd_get_client_session(BlockDriverState *bs);
int nbd_client_init(BlockDriverState *bs,
QIOChannelSocket *sock,
SocketAddress *saddr,
const char *export_name,
QCryptoTLSCreds *tlscreds,
const char *hostname,

40
block/nbd.c
View File

@ -295,30 +295,6 @@ NBDClientSession *nbd_get_client_session(BlockDriverState *bs)
return &s->client;
}
static QIOChannelSocket *nbd_establish_connection(SocketAddress *saddr,
Error **errp)
{
QIOChannelSocket *sioc;
Error *local_err = NULL;
sioc = qio_channel_socket_new();
qio_channel_set_name(QIO_CHANNEL(sioc), "nbd-client");
qio_channel_socket_connect_sync(sioc,
saddr,
&local_err);
if (local_err) {
object_unref(OBJECT(sioc));
error_propagate(errp, local_err);
return NULL;
}
qio_channel_set_delay(QIO_CHANNEL(sioc), false);
return sioc;
}
static QCryptoTLSCreds *nbd_get_tls_creds(const char *id, Error **errp)
{
Object *obj;
@ -394,7 +370,6 @@ static int nbd_open(BlockDriverState *bs, QDict *options, int flags,
BDRVNBDState *s = bs->opaque;
QemuOpts *opts = NULL;
Error *local_err = NULL;
QIOChannelSocket *sioc = NULL;
QCryptoTLSCreds *tlscreds = NULL;
const char *hostname = NULL;
int ret = -EINVAL;
@ -434,22 +409,11 @@ static int nbd_open(BlockDriverState *bs, QDict *options, int flags,
hostname = s->saddr->u.inet.host;
}
/* establish TCP connection, return error if it fails
* TODO: Configurable retry-until-timeout behaviour.
*/
sioc = nbd_establish_connection(s->saddr, errp);
if (!sioc) {
ret = -ECONNREFUSED;
goto error;
}
/* NBD handshake */
ret = nbd_client_init(bs, sioc, s->export, tlscreds, hostname,
ret = nbd_client_init(bs, s->saddr, s->export, tlscreds, hostname,
qemu_opt_get(opts, "x-dirty-bitmap"), errp);
error:
if (sioc) {
object_unref(OBJECT(sioc));
}
if (tlscreds) {
object_unref(OBJECT(tlscreds));
}

32
include/block/nbd.h
View File

@ -23,6 +23,7 @@
#include "qapi/qapi-types-block.h"
#include "io/channel-socket.h"
#include "crypto/tlscreds.h"
#include "qapi/error.h"
/* Handshake phase structs - this struct is passed on the wire */
@ -336,11 +337,38 @@ void nbd_server_start(SocketAddress *addr, const char *tls_creds,
* Reads @size bytes from @ioc. Returns 0 on success.
*/
static inline int nbd_read(QIOChannel *ioc, void *buffer, size_t size,
Error **errp)
const char *desc, Error **errp)
{
return qio_channel_read_all(ioc, buffer, size, errp) < 0 ? -EIO : 0;
int ret = qio_channel_read_all(ioc, buffer, size, errp) < 0 ? -EIO : 0;
if (ret < 0) {
if (desc) {
error_prepend(errp, "Failed to read %s: ", desc);
}
return -1;
}
return 0;
}
#define DEF_NBD_READ_N(bits) \
static inline int nbd_read##bits(QIOChannel *ioc, \
uint##bits##_t *val, \
const char *desc, Error **errp) \
{ \
if (nbd_read(ioc, val, sizeof(*val), desc, errp) < 0) { \
return -1; \
} \
*val = be##bits##_to_cpu(*val); \
return 0; \
}
DEF_NBD_READ_N(16) /* Defines nbd_read16(). */
DEF_NBD_READ_N(32) /* Defines nbd_read32(). */
DEF_NBD_READ_N(64) /* Defines nbd_read64(). */
#undef DEF_NBD_READ_N
static inline bool nbd_reply_is_simple(NBDReply *reply)
{
return reply->magic == NBD_SIMPLE_REPLY_MAGIC;

88
nbd/client.c
View File

@ -113,8 +113,7 @@ static int nbd_receive_option_reply(QIOChannel *ioc, uint32_t opt,
NBDOptionReply *reply, Error **errp)
{
QEMU_BUILD_BUG_ON(sizeof(*reply) != 20);
if (nbd_read(ioc, reply, sizeof(*reply), errp) < 0) {
error_prepend(errp, "failed to read option reply: ");
if (nbd_read(ioc, reply, sizeof(*reply), "option reply", errp) < 0) {
nbd_send_opt_abort(ioc);
return -1;
}
@ -166,8 +165,8 @@ static int nbd_handle_reply_err(QIOChannel *ioc, NBDOptionReply *reply,
goto cleanup;
}
msg = g_malloc(reply->length + 1);
if (nbd_read(ioc, msg, reply->length, errp) < 0) {
error_prepend(errp, "failed to read option error %" PRIu32
if (nbd_read(ioc, msg, reply->length, NULL, errp) < 0) {
error_prepend(errp, "Failed to read option error %" PRIu32
" (%s) message: ",
reply->type, nbd_rep_lookup(reply->type));
goto cleanup;
@ -284,12 +283,10 @@ static int nbd_receive_list(QIOChannel *ioc, char **name, char **description,
nbd_send_opt_abort(ioc);
return -1;
}
if (nbd_read(ioc, &namelen, sizeof(namelen), errp) < 0) {
error_prepend(errp, "failed to read option name length: ");
if (nbd_read32(ioc, &namelen, "option name length", errp) < 0) {
nbd_send_opt_abort(ioc);
return -1;
}
namelen = be32_to_cpu(namelen);
len -= sizeof(namelen);
if (len < namelen) {
error_setg(errp, "incorrect option name length");
@ -298,8 +295,7 @@ static int nbd_receive_list(QIOChannel *ioc, char **name, char **description,
}
local_name = g_malloc(namelen + 1);
if (nbd_read(ioc, local_name, namelen, errp) < 0) {
error_prepend(errp, "failed to read export name: ");
if (nbd_read(ioc, local_name, namelen, "export name", errp) < 0) {
nbd_send_opt_abort(ioc);
goto out;
}
@ -307,8 +303,7 @@ static int nbd_receive_list(QIOChannel *ioc, char **name, char **description,
len -= namelen;
if (len) {
local_desc = g_malloc(len + 1);
if (nbd_read(ioc, local_desc, len, errp) < 0) {
error_prepend(errp, "failed to read export description: ");
if (nbd_read(ioc, local_desc, len, "export description", errp) < 0) {
nbd_send_opt_abort(ioc);
goto out;
}
@ -410,13 +405,11 @@ static int nbd_opt_info_or_go(QIOChannel *ioc, uint32_t opt,
nbd_send_opt_abort(ioc);
return -1;
}
if (nbd_read(ioc, &type, sizeof(type), errp) < 0) {
error_prepend(errp, "failed to read info type: ");
if (nbd_read16(ioc, &type, "info type", errp) < 0) {
nbd_send_opt_abort(ioc);
return -1;
}
len -= sizeof(type);
type = be16_to_cpu(type);
switch (type) {
case NBD_INFO_EXPORT:
if (len != sizeof(info->size) + sizeof(info->flags)) {
@ -425,18 +418,14 @@ static int nbd_opt_info_or_go(QIOChannel *ioc, uint32_t opt,
nbd_send_opt_abort(ioc);
return -1;
}
if (nbd_read(ioc, &info->size, sizeof(info->size), errp) < 0) {
error_prepend(errp, "failed to read info size: ");
if (nbd_read64(ioc, &info->size, "info size", errp) < 0) {
nbd_send_opt_abort(ioc);
return -1;
}
info->size = be64_to_cpu(info->size);
if (nbd_read(ioc, &info->flags, sizeof(info->flags), errp) < 0) {
error_prepend(errp, "failed to read info flags: ");
if (nbd_read16(ioc, &info->flags, "info flags", errp) < 0) {
nbd_send_opt_abort(ioc);
return -1;
}
info->flags = be16_to_cpu(info->flags);
trace_nbd_receive_negotiate_size_flags(info->size, info->flags);
break;
@ -447,27 +436,23 @@ static int nbd_opt_info_or_go(QIOChannel *ioc, uint32_t opt,
nbd_send_opt_abort(ioc);
return -1;
}
if (nbd_read(ioc, &info->min_block, sizeof(info->min_block),
errp) < 0) {
error_prepend(errp, "failed to read info minimum block size: ");
if (nbd_read32(ioc, &info->min_block, "info minimum block size",
errp) < 0) {
nbd_send_opt_abort(ioc);
return -1;
}
info->min_block = be32_to_cpu(info->min_block);
if (!is_power_of_2(info->min_block)) {
error_setg(errp, "server minimum block size %" PRIu32
" is not a power of two", info->min_block);
nbd_send_opt_abort(ioc);
return -1;
}
if (nbd_read(ioc, &info->opt_block, sizeof(info->opt_block),
errp) < 0) {
error_prepend(errp,
"failed to read info preferred block size: ");
if (nbd_read32(ioc, &info->opt_block, "info preferred block size",
errp) < 0)
{
nbd_send_opt_abort(ioc);
return -1;
}
info->opt_block = be32_to_cpu(info->opt_block);
if (!is_power_of_2(info->opt_block) ||
info->opt_block < info->min_block) {
error_setg(errp, "server preferred block size %" PRIu32
@ -475,13 +460,12 @@ static int nbd_opt_info_or_go(QIOChannel *ioc, uint32_t opt,
nbd_send_opt_abort(ioc);
return -1;
}
if (nbd_read(ioc, &info->max_block, sizeof(info->max_block),
errp) < 0) {
error_prepend(errp, "failed to read info maximum block size: ");
if (nbd_read32(ioc, &info->max_block, "info maximum block size",
errp) < 0)
{
nbd_send_opt_abort(ioc);
return -1;
}
info->max_block = be32_to_cpu(info->max_block);
if (info->max_block < info->min_block) {
error_setg(errp, "server maximum block size %" PRIu32
" is not valid", info->max_block);
@ -731,14 +715,13 @@ static int nbd_receive_one_meta_context(QIOChannel *ioc,
return -1;
}
if (nbd_read(ioc, &local_id, sizeof(local_id), errp) < 0) {
if (nbd_read32(ioc, &local_id, "context id", errp) < 0) {
return -1;
}
local_id = be32_to_cpu(local_id);
reply.length -= sizeof(local_id);
local_name = g_malloc(reply.length + 1);
if (nbd_read(ioc, local_name, reply.length, errp) < 0) {
if (nbd_read(ioc, local_name, reply.length, "context name", errp) < 0) {
g_free(local_name);
return -1;
}
@ -896,11 +879,9 @@ static int nbd_start_negotiate(QIOChannel *ioc, QCryptoTLSCreds *tlscreds,
return -EINVAL;
}
if (nbd_read(ioc, &magic, sizeof(magic), errp) < 0) {
error_prepend(errp, "Failed to read initial magic: ");
if (nbd_read64(ioc, &magic, "initial magic", errp) < 0) {
return -EINVAL;
}
magic = be64_to_cpu(magic);
trace_nbd_receive_negotiate_magic(magic);
if (magic != NBD_INIT_MAGIC) {
@ -908,11 +889,9 @@ static int nbd_start_negotiate(QIOChannel *ioc, QCryptoTLSCreds *tlscreds,
return -EINVAL;
}
if (nbd_read(ioc, &magic, sizeof(magic), errp) < 0) {
error_prepend(errp, "Failed to read server magic: ");
if (nbd_read64(ioc, &magic, "server magic", errp) < 0) {
return -EINVAL;
}
magic = be64_to_cpu(magic);
trace_nbd_receive_negotiate_magic(magic);
if (magic == NBD_OPTS_MAGIC) {
@ -920,11 +899,9 @@ static int nbd_start_negotiate(QIOChannel *ioc, QCryptoTLSCreds *tlscreds,
uint16_t globalflags;
bool fixedNewStyle = false;
if (nbd_read(ioc, &globalflags, sizeof(globalflags), errp) < 0) {
error_prepend(errp, "Failed to read server flags: ");
if (nbd_read16(ioc, &globalflags, "server flags", errp) < 0) {
return -EINVAL;
}
globalflags = be16_to_cpu(globalflags);
trace_nbd_receive_negotiate_server_flags(globalflags);
if (globalflags & NBD_FLAG_FIXED_NEWSTYLE) {
fixedNewStyle = true;
@ -992,17 +969,13 @@ static int nbd_negotiate_finish_oldstyle(QIOChannel *ioc, NBDExportInfo *info,
{
uint32_t oldflags;
if (nbd_read(ioc, &info->size, sizeof(info->size), errp) < 0) {
error_prepend(errp, "Failed to read export length: ");
if (nbd_read64(ioc, &info->size, "export length", errp) < 0) {
return -EINVAL;
}
info->size = be64_to_cpu(info->size);
if (nbd_read(ioc, &oldflags, sizeof(oldflags), errp) < 0) {
error_prepend(errp, "Failed to read export flags: ");
if (nbd_read32(ioc, &oldflags, "export flags", errp) < 0) {
return -EINVAL;
}
oldflags = be32_to_cpu(oldflags);
if (oldflags & ~0xffff) {
error_setg(errp, "Unexpected export flags %0x" PRIx32, oldflags);
return -EINVAL;
@ -1079,17 +1052,13 @@ int nbd_receive_negotiate(QIOChannel *ioc, QCryptoTLSCreds *tlscreds,
}
/* Read the response */
if (nbd_read(ioc, &info->size, sizeof(info->size), errp) < 0) {
error_prepend(errp, "Failed to read export length: ");
if (nbd_read64(ioc, &info->size, "export length", errp) < 0) {
return -EINVAL;
}
info->size = be64_to_cpu(info->size);
if (nbd_read(ioc, &info->flags, sizeof(info->flags), errp) < 0) {
error_prepend(errp, "Failed to read export flags: ");
if (nbd_read16(ioc, &info->flags, "export flags", errp) < 0) {
return -EINVAL;
}
info->flags = be16_to_cpu(info->flags);
break;
case 0: /* oldstyle, parse length and flags */
if (*info->name) {
@ -1379,7 +1348,7 @@ static int nbd_receive_simple_reply(QIOChannel *ioc, NBDSimpleReply *reply,
assert(reply->magic == NBD_SIMPLE_REPLY_MAGIC);
ret = nbd_read(ioc, (uint8_t *)reply + sizeof(reply->magic),
sizeof(*reply) - sizeof(reply->magic), errp);
sizeof(*reply) - sizeof(reply->magic), "reply", errp);
if (ret < 0) {
return ret;
}
@ -1404,7 +1373,8 @@ static int nbd_receive_structured_reply_chunk(QIOChannel *ioc,
assert(chunk->magic == NBD_STRUCTURED_REPLY_MAGIC);
ret = nbd_read(ioc, (uint8_t *)chunk + sizeof(chunk->magic),
sizeof(*chunk) - sizeof(chunk->magic), errp);
sizeof(*chunk) - sizeof(chunk->magic), "structured chunk",
errp);
if (ret < 0) {
return ret;
}

2
nbd/common.c
View File

@ -31,7 +31,7 @@ int nbd_drop(QIOChannel *ioc, size_t size, Error **errp)
buffer = sizeof(small) >= size ? small : g_malloc(MIN(65536, size));
while (size > 0) {
ssize_t count = MIN(65536, size);
ret = nbd_read(ioc, buffer, MIN(65536, size), errp);
ret = nbd_read(ioc, buffer, MIN(65536, size), NULL, errp);
if (ret < 0) {
goto cleanup;

27
nbd/server.c
View File

@ -438,8 +438,7 @@ static int nbd_negotiate_handle_export_name(NBDClient *client,
error_setg(errp, "Bad length received");
return -EINVAL;
}
if (nbd_read(client->ioc, name, client->optlen, errp) < 0) {
error_prepend(errp, "read failed: ");
if (nbd_read(client->ioc, name, client->optlen, "export name", errp) < 0) {
return -EIO;
}
name[client->optlen] = '\0';
@ -1046,11 +1045,9 @@ static int nbd_negotiate_options(NBDClient *client, uint16_t myflags,
... Rest of request
*/
if (nbd_read(client->ioc, &flags, sizeof(flags), errp) < 0) {
error_prepend(errp, "read failed: ");
if (nbd_read32(client->ioc, &flags, "flags", errp) < 0) {
return -EIO;
}
flags = be32_to_cpu(flags);
trace_nbd_negotiate_options_flags(flags);
if (flags & NBD_FLAG_C_FIXED_NEWSTYLE) {
fixedNewstyle = true;
@ -1070,30 +1067,23 @@ static int nbd_negotiate_options(NBDClient *client, uint16_t myflags,
uint32_t option, length;
uint64_t magic;
if (nbd_read(client->ioc, &magic, sizeof(magic), errp) < 0) {
error_prepend(errp, "read failed: ");
if (nbd_read64(client->ioc, &magic, "opts magic", errp) < 0) {
return -EINVAL;
}
magic = be64_to_cpu(magic);
trace_nbd_negotiate_options_check_magic(magic);
if (magic != NBD_OPTS_MAGIC) {
error_setg(errp, "Bad magic received");
return -EINVAL;
}
if (nbd_read(client->ioc, &option,
sizeof(option), errp) < 0) {
error_prepend(errp, "read failed: ");
if (nbd_read32(client->ioc, &option, "option", errp) < 0) {
return -EINVAL;
}
option = be32_to_cpu(option);
client->opt = option;
if (nbd_read(client->ioc, &length, sizeof(length), errp) < 0) {
error_prepend(errp, "read failed: ");
if (nbd_read32(client->ioc, &length, "option length", errp) < 0) {
return -EINVAL;
}
length = be32_to_cpu(length);
assert(!client->optlen);
client->optlen = length;
@ -1306,7 +1296,7 @@ static int nbd_receive_request(QIOChannel *ioc, NBDRequest *request,
uint32_t magic;
int ret;
ret = nbd_read(ioc, buf, sizeof(buf), errp);
ret = nbd_read(ioc, buf, sizeof(buf), "request", errp);
if (ret < 0) {
return ret;
}
@ -2111,8 +2101,9 @@ static int nbd_co_receive_request(NBDRequestData *req, NBDRequest *request,
}
}
if (request->type == NBD_CMD_WRITE) {
if (nbd_read(client->ioc, req->data, request->len, errp) < 0) {
error_prepend(errp, "reading from socket failed: ");
if (nbd_read(client->ioc, req->data, request->len, "CMD_WRITE data",
errp) < 0)
{
return -EIO;
}
req->complete = true;

33
qemu-deprecated.texi
View File

@ -160,3 +160,36 @@ Example of legacy encoding:
The above, converted to the current supported format:
@code{json:@{"file.driver":"rbd", "file.pool":"rbd", "file.image":"name"@}}
@section Related binaries
@subsection qemu-nbd --partition (since 4.0.0)
The ``qemu-nbd --partition $digit'' code (also spelled @option{-P})
can only handle MBR partitions, and has never correctly handled
logical partitions beyond partition 5. If you know the offset and
length of the partition (perhaps by using @code{sfdisk} within the
guest), you can achieve the effect of exporting just that subset of
the disk by use of the @option{--image-opts} option with a raw
blockdev using the @code{offset} and @code{size} parameters layered on
top of any other existing blockdev. For example, if partition 1 is
100MiB long starting at 1MiB, the old command:
@code{qemu-nbd -t -P 1 -f qcow2 file.qcow2}
can be rewritten as:
@code{qemu-nbd -t --image-opts driver=raw,offset=1M,size=100M,file.driver=qcow2,file.backing.driver=file,file.backing.filename=file.qcow2}
Alternatively, the @code{nbdkit} project provides a more powerful
partition filter on top of its nbd plugin, which can be used to select
an arbitrary MBR or GPT partition on top of any other full-image NBD
export. Using this to rewrite the above example results in:
@code{qemu-nbd -t -k /tmp/sock -f qcow2 file.qcow2 &}
@code{nbdkit -f --filter=partition nbd socket=/tmp/sock partition=1}
Note that if you are exposing the export via /dev/nbd0, it is easier
to just export the entire image and then mount only /dev/nbd0p1 than
it is to reinvoke @command{qemu-nbd -c /dev/nbd0} limited to just a
subset of the image.

2
qemu-nbd.c
View File

@ -787,6 +787,8 @@ int main(int argc, char **argv)
flags &= ~BDRV_O_RDWR;
break;
case 'P':
warn_report("The '-P' option is deprecated; use --image-opts with "
"a raw device wrapper for subset exports instead");
if (qemu_strtoi(optarg, NULL, 0, &partition) < 0 ||
partition < 1 || partition > 8) {
error_report("Invalid partition '%s'", optarg);

6
qemu-nbd.texi
View File

@ -56,8 +56,10 @@ auto-detecting.
@item -r, --read-only
Export the disk as read-only.
@item -P, --partition=@var{num}
Only expose MBR partition @var{num}. Understands physical partitions
1-4 and logical partitions 5-8.
Deprecated: Only expose MBR partition @var{num}. Understands physical
partitions 1-4 and logical partition 5. New code should instead use
@option{--image-opts} with the raw driver wrapping a subset of the
original image.
@item -B, --bitmap=@var{name}
If @var{filename} has a qcow2 persistent bitmap @var{name}, expose
that bitmap via the ``qemu:dirty-bitmap:@var{name}'' context