/*
+ * Copyright (C) 2016-2017 Red Hat, Inc.
*
* Network Block Device Server Side
#include "qemu/osdep.h"
#include "qapi/error.h"
+#include "trace.h"
#include "nbd-internal.h"
static int system_errno_to_nbd_errno(int err)
case EFBIG:
case ENOSPC:
return NBD_ENOSPC;
+ case ESHUTDOWN:
+ return NBD_ESHUTDOWN;
case EINVAL:
default:
return NBD_EINVAL;
/* Definitions for opaque data types */
-typedef struct NBDRequest NBDRequest;
+typedef struct NBDRequestData NBDRequestData;
-struct NBDRequest {
- QSIMPLEQ_ENTRY(NBDRequest) entry;
+struct NBDRequestData {
+ QSIMPLEQ_ENTRY(NBDRequestData) entry;
NBDClient *client;
uint8_t *data;
bool complete;
BlockBackend *blk;
char *name;
+ char *description;
off_t dev_offset;
off_t size;
- uint32_t nbdflags;
+ uint16_t nbdflags;
QTAILQ_HEAD(, NBDClient) clients;
QTAILQ_ENTRY(NBDExport) next;
AioContext *ctx;
+ BlockBackend *eject_notifier_blk;
Notifier eject_notifier;
};
struct NBDClient {
int refcount;
- void (*close)(NBDClient *client);
+ void (*close_fn)(NBDClient *client, bool negotiated);
NBDExport *exp;
QCryptoTLSCreds *tlscreds;
CoMutex send_lock;
Coroutine *send_coroutine;
- bool can_read;
-
QTAILQ_ENTRY(NBDClient) next;
int nb_requests;
bool closing;
/* That's all folks */
-static void nbd_set_handlers(NBDClient *client);
-static void nbd_unset_handlers(NBDClient *client);
-static void nbd_update_can_read(NBDClient *client);
-
-static gboolean nbd_negotiate_continue(QIOChannel *ioc,
- GIOCondition condition,
- void *opaque)
-{
- qemu_coroutine_enter(opaque);
- return TRUE;
-}
-
-static ssize_t nbd_negotiate_read(QIOChannel *ioc, void *buffer, size_t size)
-{
- ssize_t ret;
- guint watch;
-
- assert(qemu_in_coroutine());
- /* Negotiation are always in main loop. */
- watch = qio_channel_add_watch(ioc,
- G_IO_IN,
- nbd_negotiate_continue,
- qemu_coroutine_self(),
- NULL);
- ret = read_sync(ioc, buffer, size);
- g_source_remove(watch);
- return ret;
-
-}
-
-static ssize_t nbd_negotiate_write(QIOChannel *ioc, void *buffer, size_t size)
-{
- ssize_t ret;
- guint watch;
-
- assert(qemu_in_coroutine());
- /* Negotiation are always in main loop. */
- watch = qio_channel_add_watch(ioc,
- G_IO_OUT,
- nbd_negotiate_continue,
- qemu_coroutine_self(),
- NULL);
- ret = write_sync(ioc, buffer, size);
- g_source_remove(watch);
- return ret;
-}
-
-static ssize_t nbd_negotiate_drop_sync(QIOChannel *ioc, size_t size)
-{
- ssize_t ret, dropped = size;
- uint8_t *buffer = g_malloc(MIN(65536, size));
-
- while (size > 0) {
- ret = nbd_negotiate_read(ioc, buffer, MIN(65536, size));
- if (ret < 0) {
- g_free(buffer);
- return ret;
- }
-
- assert(ret <= size);
- size -= ret;
- }
-
- g_free(buffer);
- return dropped;
-}
+static void nbd_client_receive_next_request(NBDClient *client);
/* Basic flow for negotiation
*/
-static int nbd_negotiate_send_rep(QIOChannel *ioc, uint32_t type, uint32_t opt)
+/* Send a reply header, including length, but no payload.
+ * Return -errno on error, 0 on success. */
+static int nbd_negotiate_send_rep_len(QIOChannel *ioc, uint32_t type,
+ uint32_t opt, uint32_t len, Error **errp)
{
uint64_t magic;
- uint32_t len;
- TRACE("Reply opt=%" PRIx32 " type=%" PRIx32, type, opt);
+ trace_nbd_negotiate_send_rep_len(opt, nbd_opt_lookup(opt),
+ type, nbd_rep_lookup(type), len);
+ assert(len < NBD_MAX_BUFFER_SIZE);
magic = cpu_to_be64(NBD_REP_MAGIC);
- if (nbd_negotiate_write(ioc, &magic, sizeof(magic)) != sizeof(magic)) {
- LOG("write failed (rep magic)");
+ if (nbd_write(ioc, &magic, sizeof(magic), errp) < 0) {
+ error_prepend(errp, "write failed (rep magic): ");
return -EINVAL;
}
+
opt = cpu_to_be32(opt);
- if (nbd_negotiate_write(ioc, &opt, sizeof(opt)) != sizeof(opt)) {
- LOG("write failed (rep opt)");
+ if (nbd_write(ioc, &opt, sizeof(opt), errp) < 0) {
+ error_prepend(errp, "write failed (rep opt): ");
return -EINVAL;
}
+
type = cpu_to_be32(type);
- if (nbd_negotiate_write(ioc, &type, sizeof(type)) != sizeof(type)) {
- LOG("write failed (rep type)");
+ if (nbd_write(ioc, &type, sizeof(type), errp) < 0) {
+ error_prepend(errp, "write failed (rep type): ");
return -EINVAL;
}
- len = cpu_to_be32(0);
- if (nbd_negotiate_write(ioc, &len, sizeof(len)) != sizeof(len)) {
- LOG("write failed (rep data length)");
+
+ len = cpu_to_be32(len);
+ if (nbd_write(ioc, &len, sizeof(len), errp) < 0) {
+ error_prepend(errp, "write failed (rep data length): ");
return -EINVAL;
}
return 0;
}
-static int nbd_negotiate_send_rep_list(QIOChannel *ioc, NBDExport *exp)
+/* Send a reply header with default 0 length.
+ * Return -errno on error, 0 on success. */
+static int nbd_negotiate_send_rep(QIOChannel *ioc, uint32_t type, uint32_t opt,
+ Error **errp)
{
- uint64_t magic, name_len;
- uint32_t opt, type, len;
+ return nbd_negotiate_send_rep_len(ioc, type, opt, 0, errp);
+}
- TRACE("Advertising export name '%s'", exp->name ? exp->name : "");
- name_len = strlen(exp->name);
- magic = cpu_to_be64(NBD_REP_MAGIC);
- if (nbd_negotiate_write(ioc, &magic, sizeof(magic)) != sizeof(magic)) {
- LOG("write failed (magic)");
- return -EINVAL;
- }
- opt = cpu_to_be32(NBD_OPT_LIST);
- if (nbd_negotiate_write(ioc, &opt, sizeof(opt)) != sizeof(opt)) {
- LOG("write failed (opt)");
- return -EINVAL;
+/* Send an error reply.
+ * Return -errno on error, 0 on success. */
+static int GCC_FMT_ATTR(5, 6)
+nbd_negotiate_send_rep_err(QIOChannel *ioc, uint32_t type,
+ uint32_t opt, Error **errp, const char *fmt, ...)
+{
+ va_list va;
+ char *msg;
+ int ret;
+ size_t len;
+
+ va_start(va, fmt);
+ msg = g_strdup_vprintf(fmt, va);
+ va_end(va);
+ len = strlen(msg);
+ assert(len < 4096);
+ trace_nbd_negotiate_send_rep_err(msg);
+ ret = nbd_negotiate_send_rep_len(ioc, type, opt, len, errp);
+ if (ret < 0) {
+ goto out;
}
- type = cpu_to_be32(NBD_REP_SERVER);
- if (nbd_negotiate_write(ioc, &type, sizeof(type)) != sizeof(type)) {
- LOG("write failed (reply type)");
- return -EINVAL;
+ if (nbd_write(ioc, msg, len, errp) < 0) {
+ error_prepend(errp, "write failed (error message): ");
+ ret = -EIO;
+ } else {
+ ret = 0;
}
- len = cpu_to_be32(name_len + sizeof(len));
- if (nbd_negotiate_write(ioc, &len, sizeof(len)) != sizeof(len)) {
- LOG("write failed (length)");
- return -EINVAL;
+
+out:
+ g_free(msg);
+ return ret;
+}
+
+/* Send a single NBD_REP_SERVER reply to NBD_OPT_LIST, including payload.
+ * Return -errno on error, 0 on success. */
+static int nbd_negotiate_send_rep_list(QIOChannel *ioc, NBDExport *exp,
+ Error **errp)
+{
+ size_t name_len, desc_len;
+ uint32_t len;
+ const char *name = exp->name ? exp->name : "";
+ const char *desc = exp->description ? exp->description : "";
+ int ret;
+
+ trace_nbd_negotiate_send_rep_list(name, desc);
+ name_len = strlen(name);
+ desc_len = strlen(desc);
+ len = name_len + desc_len + sizeof(len);
+ ret = nbd_negotiate_send_rep_len(ioc, NBD_REP_SERVER, NBD_OPT_LIST, len,
+ errp);
+ if (ret < 0) {
+ return ret;
}
+
len = cpu_to_be32(name_len);
- if (nbd_negotiate_write(ioc, &len, sizeof(len)) != sizeof(len)) {
- LOG("write failed (length)");
+ if (nbd_write(ioc, &len, sizeof(len), errp) < 0) {
+ error_prepend(errp, "write failed (name length): ");
return -EINVAL;
}
- if (nbd_negotiate_write(ioc, exp->name, name_len) != name_len) {
- LOG("write failed (buffer)");
+
+ if (nbd_write(ioc, name, name_len, errp) < 0) {
+ error_prepend(errp, "write failed (name buffer): ");
return -EINVAL;
}
+
+ if (nbd_write(ioc, desc, desc_len, errp) < 0) {
+ error_prepend(errp, "write failed (description buffer): ");
+ return -EINVAL;
+ }
+
return 0;
}
-static int nbd_negotiate_handle_list(NBDClient *client, uint32_t length)
+/* Process the NBD_OPT_LIST command, with a potential series of replies.
+ * Return -errno on error, 0 on success. */
+static int nbd_negotiate_handle_list(NBDClient *client, uint32_t length,
+ Error **errp)
{
NBDExport *exp;
if (length) {
- if (nbd_negotiate_drop_sync(client->ioc, length) != length) {
+ if (nbd_drop(client->ioc, length, errp) < 0) {
return -EIO;
}
- return nbd_negotiate_send_rep(client->ioc,
- NBD_REP_ERR_INVALID, NBD_OPT_LIST);
+ return nbd_negotiate_send_rep_err(client->ioc,
+ NBD_REP_ERR_INVALID, NBD_OPT_LIST,
+ errp,
+ "OPT_LIST should not have length");
}
/* For each export, send a NBD_REP_SERVER reply. */
QTAILQ_FOREACH(exp, &exports, next) {
- if (nbd_negotiate_send_rep_list(client->ioc, exp)) {
+ if (nbd_negotiate_send_rep_list(client->ioc, exp, errp)) {
return -EINVAL;
}
}
/* Finish with a NBD_REP_ACK. */
- return nbd_negotiate_send_rep(client->ioc, NBD_REP_ACK, NBD_OPT_LIST);
+ return nbd_negotiate_send_rep(client->ioc, NBD_REP_ACK, NBD_OPT_LIST, errp);
}
-static int nbd_negotiate_handle_export_name(NBDClient *client, uint32_t length)
+/* Send a reply to NBD_OPT_EXPORT_NAME.
+ * Return -errno on error, 0 on success. */
+static int nbd_negotiate_handle_export_name(NBDClient *client, uint32_t length,
+ uint16_t myflags, bool no_zeroes,
+ Error **errp)
{
- int rc = -EINVAL;
char name[NBD_MAX_NAME_SIZE + 1];
+ char buf[8 + 4 + 124] = "";
+ size_t len;
+ int ret;
/* Client sends:
[20 .. xx] export name (length bytes)
*/
- TRACE("Checking length");
+ trace_nbd_negotiate_handle_export_name();
if (length >= sizeof(name)) {
- LOG("Bad length received");
- goto fail;
+ error_setg(errp, "Bad length received");
+ return -EINVAL;
}
- if (nbd_negotiate_read(client->ioc, name, length) != length) {
- LOG("read failed");
- goto fail;
+ if (nbd_read(client->ioc, name, length, errp) < 0) {
+ error_prepend(errp, "read failed: ");
+ return -EINVAL;
}
name[length] = '\0';
- TRACE("Client requested export '%s'", name);
+ trace_nbd_negotiate_handle_export_name_request(name);
client->exp = nbd_export_find(name);
if (!client->exp) {
- LOG("export not found");
- goto fail;
+ error_setg(errp, "export not found");
+ return -EINVAL;
+ }
+
+ trace_nbd_negotiate_new_style_size_flags(client->exp->size,
+ client->exp->nbdflags | myflags);
+ stq_be_p(buf, client->exp->size);
+ stw_be_p(buf + 8, client->exp->nbdflags | myflags);
+ len = no_zeroes ? 10 : sizeof(buf);
+ ret = nbd_write(client->ioc, buf, len, errp);
+ if (ret < 0) {
+ error_prepend(errp, "write failed: ");
+ return ret;
}
QTAILQ_INSERT_TAIL(&client->exp->clients, client, next);
nbd_export_get(client->exp);
- rc = 0;
-fail:
+
+ return 0;
+}
+
+/* Send a single NBD_REP_INFO, with a buffer @buf of @length bytes.
+ * The buffer does NOT include the info type prefix.
+ * Return -errno on error, 0 if ready to send more. */
+static int nbd_negotiate_send_info(NBDClient *client, uint32_t opt,
+ uint16_t info, uint32_t length, void *buf,
+ Error **errp)
+{
+ int rc;
+
+ trace_nbd_negotiate_send_info(info, nbd_info_lookup(info), length);
+ rc = nbd_negotiate_send_rep_len(client->ioc, NBD_REP_INFO, opt,
+ sizeof(info) + length, errp);
+ if (rc < 0) {
+ return rc;
+ }
+ cpu_to_be16s(&info);
+ if (nbd_write(client->ioc, &info, sizeof(info), errp) < 0) {
+ return -EIO;
+ }
+ if (nbd_write(client->ioc, buf, length, errp) < 0) {
+ return -EIO;
+ }
+ return 0;
+}
+
+/* Handle NBD_OPT_INFO and NBD_OPT_GO.
+ * Return -errno on error, 0 if ready for next option, and 1 to move
+ * into transmission phase. */
+static int nbd_negotiate_handle_info(NBDClient *client, uint32_t length,
+ uint32_t opt, uint16_t myflags,
+ Error **errp)
+{
+ int rc;
+ char name[NBD_MAX_NAME_SIZE + 1];
+ NBDExport *exp;
+ uint16_t requests;
+ uint16_t request;
+ uint32_t namelen;
+ bool sendname = false;
+ char buf[sizeof(uint64_t) + sizeof(uint16_t)];
+ const char *msg;
+
+ /* Client sends:
+ 4 bytes: L, name length (can be 0)
+ L bytes: export name
+ 2 bytes: N, number of requests (can be 0)
+ N * 2 bytes: N requests
+ */
+ if (length < sizeof(namelen) + sizeof(requests)) {
+ msg = "overall request too short";
+ goto invalid;
+ }
+ if (nbd_read(client->ioc, &namelen, sizeof(namelen), errp) < 0) {
+ return -EIO;
+ }
+ be32_to_cpus(&namelen);
+ length -= sizeof(namelen);
+ if (namelen > length - sizeof(requests) || (length - namelen) % 2) {
+ msg = "name length is incorrect";
+ goto invalid;
+ }
+ if (nbd_read(client->ioc, name, namelen, errp) < 0) {
+ return -EIO;
+ }
+ name[namelen] = '\0';
+ length -= namelen;
+ trace_nbd_negotiate_handle_export_name_request(name);
+
+ if (nbd_read(client->ioc, &requests, sizeof(requests), errp) < 0) {
+ return -EIO;
+ }
+ be16_to_cpus(&requests);
+ length -= sizeof(requests);
+ trace_nbd_negotiate_handle_info_requests(requests);
+ if (requests != length / sizeof(request)) {
+ msg = "incorrect number of requests for overall length";
+ goto invalid;
+ }
+ while (requests--) {
+ if (nbd_read(client->ioc, &request, sizeof(request), errp) < 0) {
+ return -EIO;
+ }
+ be16_to_cpus(&request);
+ length -= sizeof(request);
+ trace_nbd_negotiate_handle_info_request(request,
+ nbd_info_lookup(request));
+ /* For now, we only care about NBD_INFO_NAME; everything else
+ * is either a request we don't know or something we send
+ * regardless of request. */
+ if (request == NBD_INFO_NAME) {
+ sendname = true;
+ }
+ }
+
+ exp = nbd_export_find(name);
+ if (!exp) {
+ return nbd_negotiate_send_rep_err(client->ioc, NBD_REP_ERR_UNKNOWN,
+ opt, errp, "export '%s' not present",
+ name);
+ }
+
+ /* Don't bother sending NBD_INFO_NAME unless client requested it */
+ if (sendname) {
+ rc = nbd_negotiate_send_info(client, opt, NBD_INFO_NAME, length, name,
+ errp);
+ if (rc < 0) {
+ return rc;
+ }
+ }
+
+ /* Send NBD_INFO_DESCRIPTION only if available, regardless of
+ * client request */
+ if (exp->description) {
+ size_t len = strlen(exp->description);
+
+ rc = nbd_negotiate_send_info(client, opt, NBD_INFO_DESCRIPTION,
+ len, exp->description, errp);
+ if (rc < 0) {
+ return rc;
+ }
+ }
+
+ /* Send NBD_INFO_EXPORT always */
+ trace_nbd_negotiate_new_style_size_flags(exp->size,
+ exp->nbdflags | myflags);
+ stq_be_p(buf, exp->size);
+ stw_be_p(buf + 8, exp->nbdflags | myflags);
+ rc = nbd_negotiate_send_info(client, opt, NBD_INFO_EXPORT,
+ sizeof(buf), buf, errp);
+ if (rc < 0) {
+ return rc;
+ }
+
+ /* Final reply */
+ rc = nbd_negotiate_send_rep(client->ioc, NBD_REP_ACK, opt, errp);
+ if (rc < 0) {
+ return rc;
+ }
+
+ if (opt == NBD_OPT_GO) {
+ client->exp = exp;
+ QTAILQ_INSERT_TAIL(&client->exp->clients, client, next);
+ nbd_export_get(client->exp);
+ rc = 1;
+ }
return rc;
+
+ invalid:
+ if (nbd_drop(client->ioc, length, errp) < 0) {
+ return -EIO;
+ }
+ return nbd_negotiate_send_rep_err(client->ioc, NBD_REP_ERR_INVALID, opt,
+ errp, "%s", msg);
}
+/* Handle NBD_OPT_STARTTLS. Return NULL to drop connection, or else the
+ * new channel for all further (now-encrypted) communication. */
static QIOChannel *nbd_negotiate_handle_starttls(NBDClient *client,
- uint32_t length)
+ uint32_t length,
+ Error **errp)
{
QIOChannel *ioc;
QIOChannelTLS *tioc;
struct NBDTLSHandshakeData data = { 0 };
- TRACE("Setting up TLS");
+ trace_nbd_negotiate_handle_starttls();
ioc = client->ioc;
if (length) {
- if (nbd_negotiate_drop_sync(ioc, length) != length) {
+ if (nbd_drop(ioc, length, errp) < 0) {
return NULL;
}
- nbd_negotiate_send_rep(ioc, NBD_REP_ERR_INVALID, NBD_OPT_STARTTLS);
+ nbd_negotiate_send_rep_err(ioc, NBD_REP_ERR_INVALID, NBD_OPT_STARTTLS,
+ errp,
+ "OPT_STARTTLS should not have length");
return NULL;
}
if (nbd_negotiate_send_rep(client->ioc, NBD_REP_ACK,
- NBD_OPT_STARTTLS) < 0) {
+ NBD_OPT_STARTTLS, errp) < 0) {
return NULL;
}
tioc = qio_channel_tls_new_server(ioc,
client->tlscreds,
client->tlsaclname,
- NULL);
+ errp);
if (!tioc) {
return NULL;
}
- TRACE("Starting TLS handshake");
+ qio_channel_set_name(QIO_CHANNEL(tioc), "nbd-server-tls");
+ trace_nbd_negotiate_handle_starttls_handshake();
data.loop = g_main_loop_new(g_main_context_default(), FALSE);
qio_channel_tls_handshake(tioc,
nbd_tls_handshake,
g_main_loop_unref(data.loop);
if (data.error) {
object_unref(OBJECT(tioc));
- error_free(data.error);
+ error_propagate(errp, data.error);
return NULL;
}
return QIO_CHANNEL(tioc);
}
-
-static int nbd_negotiate_options(NBDClient *client)
+/* nbd_negotiate_options
+ * Process all NBD_OPT_* client option commands, during fixed newstyle
+ * negotiation.
+ * Return:
+ * -errno on error, errp is set
+ * 0 on successful negotiation, errp is not set
+ * 1 if client sent NBD_OPT_ABORT, i.e. on valid disconnect,
+ * errp is not set
+ */
+static int nbd_negotiate_options(NBDClient *client, uint16_t myflags,
+ Error **errp)
{
uint32_t flags;
bool fixedNewstyle = false;
+ bool no_zeroes = false;
/* Client sends:
[ 0 .. 3] client flags
+ Then we loop until NBD_OPT_EXPORT_NAME or NBD_OPT_GO:
[ 0 .. 7] NBD_OPTS_MAGIC
[ 8 .. 11] NBD option
[12 .. 15] Data length
... Rest of request
*/
- if (nbd_negotiate_read(client->ioc, &flags, sizeof(flags)) !=
- sizeof(flags)) {
- LOG("read failed");
+ if (nbd_read(client->ioc, &flags, sizeof(flags), errp) < 0) {
+ error_prepend(errp, "read failed: ");
return -EIO;
}
- TRACE("Checking client flags");
be32_to_cpus(&flags);
+ trace_nbd_negotiate_options_flags(flags);
if (flags & NBD_FLAG_C_FIXED_NEWSTYLE) {
- TRACE("Client supports fixed newstyle handshake");
fixedNewstyle = true;
flags &= ~NBD_FLAG_C_FIXED_NEWSTYLE;
}
+ if (flags & NBD_FLAG_C_NO_ZEROES) {
+ no_zeroes = true;
+ flags &= ~NBD_FLAG_C_NO_ZEROES;
+ }
if (flags != 0) {
- TRACE("Unknown client flags 0x%" PRIx32 " received", flags);
- return -EIO;
+ error_setg(errp, "Unknown client flags 0x%" PRIx32 " received", flags);
+ return -EINVAL;
}
while (1) {
int ret;
- uint32_t clientflags, length;
+ uint32_t option, length;
uint64_t magic;
- if (nbd_negotiate_read(client->ioc, &magic, sizeof(magic)) !=
- sizeof(magic)) {
- LOG("read failed");
+ if (nbd_read(client->ioc, &magic, sizeof(magic), errp) < 0) {
+ error_prepend(errp, "read failed: ");
return -EINVAL;
}
- TRACE("Checking opts magic");
- if (magic != be64_to_cpu(NBD_OPTS_MAGIC)) {
- LOG("Bad magic received");
+ 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_negotiate_read(client->ioc, &clientflags,
- sizeof(clientflags)) != sizeof(clientflags)) {
- LOG("read failed");
+ if (nbd_read(client->ioc, &option,
+ sizeof(option), errp) < 0) {
+ error_prepend(errp, "read failed: ");
return -EINVAL;
}
- clientflags = be32_to_cpu(clientflags);
+ option = be32_to_cpu(option);
- if (nbd_negotiate_read(client->ioc, &length, sizeof(length)) !=
- sizeof(length)) {
- LOG("read failed");
+ if (nbd_read(client->ioc, &length, sizeof(length), errp) < 0) {
+ error_prepend(errp, "read failed: ");
return -EINVAL;
}
length = be32_to_cpu(length);
- TRACE("Checking option 0x%" PRIx32, clientflags);
+ trace_nbd_negotiate_options_check_option(option,
+ nbd_opt_lookup(option));
if (client->tlscreds &&
client->ioc == (QIOChannel *)client->sioc) {
QIOChannel *tioc;
if (!fixedNewstyle) {
- TRACE("Unsupported option 0x%" PRIx32, clientflags);
+ error_setg(errp, "Unsupported option 0x%" PRIx32, option);
return -EINVAL;
}
- switch (clientflags) {
+ switch (option) {
case NBD_OPT_STARTTLS:
- tioc = nbd_negotiate_handle_starttls(client, length);
+ tioc = nbd_negotiate_handle_starttls(client, length, errp);
if (!tioc) {
return -EIO;
}
case NBD_OPT_EXPORT_NAME:
/* No way to return an error to client, so drop connection */
- TRACE("Option 0x%x not permitted before TLS", clientflags);
+ error_setg(errp, "Option 0x%x not permitted before TLS",
+ option);
return -EINVAL;
default:
- TRACE("Option 0x%" PRIx32 " not permitted before TLS",
- clientflags);
- if (nbd_negotiate_drop_sync(client->ioc, length) != length) {
+ if (nbd_drop(client->ioc, length, errp) < 0) {
return -EIO;
}
- ret = nbd_negotiate_send_rep(client->ioc, NBD_REP_ERR_TLS_REQD,
- clientflags);
+ ret = nbd_negotiate_send_rep_err(client->ioc,
+ NBD_REP_ERR_TLS_REQD,
+ option, errp,
+ "Option 0x%" PRIx32
+ "not permitted before TLS",
+ option);
if (ret < 0) {
return ret;
}
+ /* Let the client keep trying, unless they asked to
+ * quit. In this mode, we've already sent an error, so
+ * we can't ack the abort. */
+ if (option == NBD_OPT_ABORT) {
+ return 1;
+ }
break;
}
} else if (fixedNewstyle) {
- switch (clientflags) {
+ switch (option) {
case NBD_OPT_LIST:
- ret = nbd_negotiate_handle_list(client, length);
+ ret = nbd_negotiate_handle_list(client, length, errp);
if (ret < 0) {
return ret;
}
break;
case NBD_OPT_ABORT:
- return -EINVAL;
+ /* NBD spec says we must try to reply before
+ * disconnecting, but that we must also tolerate
+ * guests that don't wait for our reply. */
+ nbd_negotiate_send_rep(client->ioc, NBD_REP_ACK, option, NULL);
+ return 1;
case NBD_OPT_EXPORT_NAME:
- return nbd_negotiate_handle_export_name(client, length);
+ return nbd_negotiate_handle_export_name(client, length,
+ myflags, no_zeroes,
+ errp);
+
+ case NBD_OPT_INFO:
+ case NBD_OPT_GO:
+ ret = nbd_negotiate_handle_info(client, length, option,
+ myflags, errp);
+ if (ret == 1) {
+ assert(option == NBD_OPT_GO);
+ return 0;
+ }
+ if (ret) {
+ return ret;
+ }
+ break;
case NBD_OPT_STARTTLS:
- if (nbd_negotiate_drop_sync(client->ioc, length) != length) {
+ if (nbd_drop(client->ioc, length, errp) < 0) {
return -EIO;
}
if (client->tlscreds) {
- TRACE("TLS already enabled");
- ret = nbd_negotiate_send_rep(client->ioc,
- NBD_REP_ERR_INVALID,
- clientflags);
+ ret = nbd_negotiate_send_rep_err(client->ioc,
+ NBD_REP_ERR_INVALID,
+ option, errp,
+ "TLS already enabled");
} else {
- TRACE("TLS not configured");
- ret = nbd_negotiate_send_rep(client->ioc,
- NBD_REP_ERR_POLICY,
- clientflags);
+ ret = nbd_negotiate_send_rep_err(client->ioc,
+ NBD_REP_ERR_POLICY,
+ option, errp,
+ "TLS not configured");
}
if (ret < 0) {
return ret;
}
break;
default:
- TRACE("Unsupported option 0x%" PRIx32, clientflags);
- if (nbd_negotiate_drop_sync(client->ioc, length) != length) {
+ if (nbd_drop(client->ioc, length, errp) < 0) {
return -EIO;
}
- ret = nbd_negotiate_send_rep(client->ioc, NBD_REP_ERR_UNSUP,
- clientflags);
+ ret = nbd_negotiate_send_rep_err(client->ioc,
+ NBD_REP_ERR_UNSUP,
+ option, errp,
+ "Unsupported option 0x%"
+ PRIx32 " (%s)", option,
+ nbd_opt_lookup(option));
if (ret < 0) {
return ret;
}
* If broken new-style we should drop the connection
* for anything except NBD_OPT_EXPORT_NAME
*/
- switch (clientflags) {
+ switch (option) {
case NBD_OPT_EXPORT_NAME:
- return nbd_negotiate_handle_export_name(client, length);
+ return nbd_negotiate_handle_export_name(client, length,
+ myflags, no_zeroes,
+ errp);
default:
- TRACE("Unsupported option 0x%" PRIx32, clientflags);
+ error_setg(errp, "Unsupported option 0x%" PRIx32 " (%s)",
+ option, nbd_opt_lookup(option));
return -EINVAL;
}
}
}
}
-typedef struct {
- NBDClient *client;
- Coroutine *co;
-} NBDClientNewData;
-
-static coroutine_fn int nbd_negotiate(NBDClientNewData *data)
+/* nbd_negotiate
+ * Return:
+ * -errno on error, errp is set
+ * 0 on successful negotiation, errp is not set
+ * 1 if client sent NBD_OPT_ABORT, i.e. on valid disconnect,
+ * errp is not set
+ */
+static coroutine_fn int nbd_negotiate(NBDClient *client, Error **errp)
{
- NBDClient *client = data->client;
char buf[8 + 8 + 8 + 128];
- int rc;
- const int myflags = (NBD_FLAG_HAS_FLAGS | NBD_FLAG_SEND_TRIM |
- NBD_FLAG_SEND_FLUSH | NBD_FLAG_SEND_FUA);
+ int ret;
+ const uint16_t myflags = (NBD_FLAG_HAS_FLAGS | NBD_FLAG_SEND_TRIM |
+ NBD_FLAG_SEND_FLUSH | NBD_FLAG_SEND_FUA |
+ NBD_FLAG_SEND_WRITE_ZEROES);
bool oldStyle;
/* Old style negotiation header without options
[ 0 .. 7] passwd ("NBDMAGIC")
[ 8 .. 15] magic (NBD_OPTS_MAGIC)
[16 .. 17] server flags (0)
- ....options sent....
- [18 .. 25] size
- [26 .. 27] export flags
- [28 .. 151] reserved (0)
+ ....options sent, ending in NBD_OPT_EXPORT_NAME or NBD_OPT_GO....
*/
qio_channel_set_blocking(client->ioc, false, NULL);
- rc = -EINVAL;
- TRACE("Beginning negotiation.");
+ trace_nbd_negotiate_begin();
memset(buf, 0, sizeof(buf));
memcpy(buf, "NBDMAGIC", 8);
oldStyle = client->exp != NULL && !client->tlscreds;
if (oldStyle) {
- assert ((client->exp->nbdflags & ~65535) == 0);
- TRACE("advertising size %" PRIu64 " and flags %x",
- client->exp->size, client->exp->nbdflags | myflags);
+ trace_nbd_negotiate_old_style(client->exp->size,
+ client->exp->nbdflags | myflags);
stq_be_p(buf + 8, NBD_CLIENT_MAGIC);
stq_be_p(buf + 16, client->exp->size);
stw_be_p(buf + 26, client->exp->nbdflags | myflags);
- } else {
- stq_be_p(buf + 8, NBD_OPTS_MAGIC);
- stw_be_p(buf + 16, NBD_FLAG_FIXED_NEWSTYLE);
- }
- if (oldStyle) {
- if (client->tlscreds) {
- TRACE("TLS cannot be enabled with oldstyle protocol");
- goto fail;
- }
- if (nbd_negotiate_write(client->ioc, buf, sizeof(buf)) != sizeof(buf)) {
- LOG("write failed");
- goto fail;
+ if (nbd_write(client->ioc, buf, sizeof(buf), errp) < 0) {
+ error_prepend(errp, "write failed: ");
+ return -EINVAL;
}
} else {
- if (nbd_negotiate_write(client->ioc, buf, 18) != 18) {
- LOG("write failed");
- goto fail;
- }
- rc = nbd_negotiate_options(client);
- if (rc != 0) {
- LOG("option negotiation failed");
- goto fail;
- }
+ stq_be_p(buf + 8, NBD_OPTS_MAGIC);
+ stw_be_p(buf + 16, NBD_FLAG_FIXED_NEWSTYLE | NBD_FLAG_NO_ZEROES);
- assert ((client->exp->nbdflags & ~65535) == 0);
- TRACE("advertising size %" PRIu64 " and flags %x",
- client->exp->size, client->exp->nbdflags | myflags);
- stq_be_p(buf + 18, client->exp->size);
- stw_be_p(buf + 26, client->exp->nbdflags | myflags);
- if (nbd_negotiate_write(client->ioc, buf + 18, sizeof(buf) - 18) !=
- sizeof(buf) - 18) {
- LOG("write failed");
- goto fail;
+ if (nbd_write(client->ioc, buf, 18, errp) < 0) {
+ error_prepend(errp, "write failed: ");
+ return -EINVAL;
+ }
+ ret = nbd_negotiate_options(client, myflags, errp);
+ if (ret != 0) {
+ if (ret < 0) {
+ error_prepend(errp, "option negotiation failed: ");
+ }
+ return ret;
}
}
- TRACE("Negotiation succeeded.");
- rc = 0;
-fail:
- return rc;
+ trace_nbd_negotiate_success();
+
+ return 0;
}
-static ssize_t nbd_receive_request(QIOChannel *ioc, struct nbd_request *request)
+static int nbd_receive_request(QIOChannel *ioc, NBDRequest *request,
+ Error **errp)
{
uint8_t buf[NBD_REQUEST_SIZE];
uint32_t magic;
- ssize_t ret;
+ int ret;
- ret = read_sync(ioc, buf, sizeof(buf));
+ ret = nbd_read(ioc, buf, sizeof(buf), errp);
if (ret < 0) {
return ret;
}
- if (ret != sizeof(buf)) {
- LOG("read failed");
- return -EINVAL;
- }
-
/* Request
[ 0 .. 3] magic (NBD_REQUEST_MAGIC)
- [ 4 .. 7] type (0 == READ, 1 == WRITE)
+ [ 4 .. 5] flags (NBD_CMD_FLAG_FUA, ...)
+ [ 6 .. 7] type (NBD_CMD_READ, ...)
[ 8 .. 15] handle
[16 .. 23] from
[24 .. 27] len
*/
magic = ldl_be_p(buf);
- request->type = ldl_be_p(buf + 4);
+ request->flags = lduw_be_p(buf + 4);
+ request->type = lduw_be_p(buf + 6);
request->handle = ldq_be_p(buf + 8);
request->from = ldq_be_p(buf + 16);
request->len = ldl_be_p(buf + 24);
- TRACE("Got request: { magic = 0x%" PRIx32 ", .type = %" PRIx32
- ", from = %" PRIu64 " , len = %" PRIu32 " }",
- magic, request->type, request->from, request->len);
+ trace_nbd_receive_request(magic, request->flags, request->type,
+ request->from, request->len);
if (magic != NBD_REQUEST_MAGIC) {
- LOG("invalid magic (got 0x%" PRIx32 ")", magic);
+ error_setg(errp, "invalid magic (got 0x%" PRIx32 ")", magic);
return -EINVAL;
}
return 0;
}
-static ssize_t nbd_send_reply(QIOChannel *ioc, struct nbd_reply *reply)
+static int nbd_send_reply(QIOChannel *ioc, NBDReply *reply, Error **errp)
{
uint8_t buf[NBD_REPLY_SIZE];
- ssize_t ret;
reply->error = system_errno_to_nbd_errno(reply->error);
- TRACE("Sending response to client: { .error = %" PRId32
- ", handle = %" PRIu64 " }",
- reply->error, reply->handle);
+ trace_nbd_send_reply(reply->error, reply->handle);
/* Reply
[ 0 .. 3] magic (NBD_REPLY_MAGIC)
stl_be_p(buf + 4, reply->error);
stq_be_p(buf + 8, reply->handle);
- ret = write_sync(ioc, buf, sizeof(buf));
- if (ret < 0) {
- return ret;
- }
-
- if (ret != sizeof(buf)) {
- LOG("writing to socket failed");
- return -EINVAL;
- }
- return 0;
+ return nbd_write(ioc, buf, sizeof(buf), errp);
}
#define MAX_NBD_REQUESTS 16
*/
assert(client->closing);
- nbd_unset_handlers(client);
+ qio_channel_detach_aio_context(client->ioc);
object_unref(OBJECT(client->sioc));
object_unref(OBJECT(client->ioc));
if (client->tlscreds) {
}
}
-static void client_close(NBDClient *client)
+static void client_close(NBDClient *client, bool negotiated)
{
if (client->closing) {
return;
NULL);
/* Also tell the client, so that they release their reference. */
- if (client->close) {
- client->close(client);
+ if (client->close_fn) {
+ client->close_fn(client, negotiated);
}
}
-static NBDRequest *nbd_request_get(NBDClient *client)
+static NBDRequestData *nbd_request_get(NBDClient *client)
{
- NBDRequest *req;
+ NBDRequestData *req;
assert(client->nb_requests <= MAX_NBD_REQUESTS - 1);
client->nb_requests++;
- nbd_update_can_read(client);
- req = g_new0(NBDRequest, 1);
+ req = g_new0(NBDRequestData, 1);
nbd_client_get(client);
req->client = client;
return req;
}
-static void nbd_request_put(NBDRequest *req)
+static void nbd_request_put(NBDRequestData *req)
{
NBDClient *client = req->client;
g_free(req);
client->nb_requests--;
- nbd_update_can_read(client);
+ nbd_client_receive_next_request(client);
+
nbd_client_put(client);
}
NBDExport *exp = opaque;
NBDClient *client;
- TRACE("Export %s: Attaching clients to AIO context %p\n", exp->name, ctx);
+ trace_nbd_blk_aio_attached(exp->name, ctx);
exp->ctx = ctx;
QTAILQ_FOREACH(client, &exp->clients, next) {
- nbd_set_handlers(client);
+ qio_channel_attach_aio_context(client->ioc, ctx);
+ if (client->recv_coroutine) {
+ aio_co_schedule(ctx, client->recv_coroutine);
+ }
+ if (client->send_coroutine) {
+ aio_co_schedule(ctx, client->send_coroutine);
+ }
}
}
NBDExport *exp = opaque;
NBDClient *client;
- TRACE("Export %s: Detaching clients from AIO context %p\n", exp->name, exp->ctx);
+ trace_nbd_blk_aio_detach(exp->name, exp->ctx);
QTAILQ_FOREACH(client, &exp->clients, next) {
- nbd_unset_handlers(client);
+ qio_channel_detach_aio_context(client->ioc);
}
exp->ctx = NULL;
nbd_export_close(exp);
}
-NBDExport *nbd_export_new(BlockBackend *blk, off_t dev_offset, off_t size,
- uint32_t nbdflags, void (*close)(NBDExport *),
+NBDExport *nbd_export_new(BlockDriverState *bs, off_t dev_offset, off_t size,
+ uint16_t nbdflags, void (*close)(NBDExport *),
+ bool writethrough, BlockBackend *on_eject_blk,
Error **errp)
{
+ BlockBackend *blk;
NBDExport *exp = g_malloc0(sizeof(NBDExport));
+ uint64_t perm;
+ int ret;
+
+ /* Don't allow resize while the NBD server is running, otherwise we don't
+ * care what happens with the node. */
+ perm = BLK_PERM_CONSISTENT_READ;
+ if ((nbdflags & NBD_FLAG_READ_ONLY) == 0) {
+ perm |= BLK_PERM_WRITE;
+ }
+ blk = blk_new(perm, BLK_PERM_CONSISTENT_READ | BLK_PERM_WRITE_UNCHANGED |
+ BLK_PERM_WRITE | BLK_PERM_GRAPH_MOD);
+ ret = blk_insert_bs(blk, bs, errp);
+ if (ret < 0) {
+ goto fail;
+ }
+ blk_set_enable_write_cache(blk, !writethrough);
+
exp->refcount = 1;
QTAILQ_INIT(&exp->clients);
exp->blk = blk;
exp->close = close;
exp->ctx = blk_get_aio_context(blk);
- blk_ref(blk);
blk_add_aio_context_notifier(blk, blk_aio_attached, blk_aio_detach, exp);
- exp->eject_notifier.notify = nbd_eject_notifier;
- blk_add_remove_bs_notifier(blk, &exp->eject_notifier);
+ if (on_eject_blk) {
+ blk_ref(on_eject_blk);
+ exp->eject_notifier_blk = on_eject_blk;
+ exp->eject_notifier.notify = nbd_eject_notifier;
+ blk_add_remove_bs_notifier(on_eject_blk, &exp->eject_notifier);
+ }
/*
* NBD exports are used for non-shared storage migration. Make sure
return exp;
fail:
+ blk_unref(blk);
g_free(exp);
return NULL;
}
nbd_export_put(exp);
}
+void nbd_export_set_description(NBDExport *exp, const char *description)
+{
+ g_free(exp->description);
+ exp->description = g_strdup(description);
+}
+
void nbd_export_close(NBDExport *exp)
{
NBDClient *client, *next;
nbd_export_get(exp);
QTAILQ_FOREACH_SAFE(client, &exp->clients, next, next) {
- client_close(client);
+ client_close(client, true);
}
nbd_export_set_name(exp, NULL);
+ nbd_export_set_description(exp, NULL);
nbd_export_put(exp);
}
if (--exp->refcount == 0) {
assert(exp->name == NULL);
+ assert(exp->description == NULL);
if (exp->close) {
exp->close(exp);
}
if (exp->blk) {
- notifier_remove(&exp->eject_notifier);
+ if (exp->eject_notifier_blk) {
+ notifier_remove(&exp->eject_notifier);
+ blk_unref(exp->eject_notifier_blk);
+ }
blk_remove_aio_context_notifier(exp->blk, blk_aio_attached,
blk_aio_detach, exp);
blk_unref(exp->blk);
}
}
-static ssize_t nbd_co_send_reply(NBDRequest *req, struct nbd_reply *reply,
- int len)
+static int nbd_co_send_reply(NBDRequestData *req, NBDReply *reply, int len,
+ Error **errp)
{
NBDClient *client = req->client;
- ssize_t rc, ret;
+ int ret;
g_assert(qemu_in_coroutine());
+
+ trace_nbd_co_send_reply(reply->handle, reply->error, len);
+
qemu_co_mutex_lock(&client->send_lock);
client->send_coroutine = qemu_coroutine_self();
- nbd_set_handlers(client);
if (!len) {
- rc = nbd_send_reply(client->ioc, reply);
+ ret = nbd_send_reply(client->ioc, reply, errp);
} else {
qio_channel_set_cork(client->ioc, true);
- rc = nbd_send_reply(client->ioc, reply);
- if (rc >= 0) {
- ret = write_sync(client->ioc, req->data, len);
- if (ret != len) {
- rc = -EIO;
+ ret = nbd_send_reply(client->ioc, reply, errp);
+ if (ret == 0) {
+ ret = nbd_write(client->ioc, req->data, len, errp);
+ if (ret < 0) {
+ ret = -EIO;
}
}
qio_channel_set_cork(client->ioc, false);
}
client->send_coroutine = NULL;
- nbd_set_handlers(client);
qemu_co_mutex_unlock(&client->send_lock);
- return rc;
+ return ret;
}
-/* Collect a client request. Return 0 if request looks valid, -EAGAIN
- * to keep trying the collection, -EIO to drop connection right away,
- * and any other negative value to report an error to the client
- * (although the caller may still need to disconnect after reporting
- * the error). */
-static ssize_t nbd_co_receive_request(NBDRequest *req,
- struct nbd_request *request)
+/* nbd_co_receive_request
+ * Collect a client request. Return 0 if request looks valid, -EIO to drop
+ * connection right away, and any other negative value to report an error to
+ * the client (although the caller may still need to disconnect after reporting
+ * the error).
+ */
+static int nbd_co_receive_request(NBDRequestData *req, NBDRequest *request,
+ Error **errp)
{
NBDClient *client = req->client;
- uint32_t command;
- ssize_t rc;
g_assert(qemu_in_coroutine());
- client->recv_coroutine = qemu_coroutine_self();
- nbd_update_can_read(client);
-
- rc = nbd_receive_request(client->ioc, request);
- if (rc < 0) {
- if (rc != -EAGAIN) {
- rc = -EIO;
- }
- goto out;
+ assert(client->recv_coroutine == qemu_coroutine_self());
+ if (nbd_receive_request(client->ioc, request, errp) < 0) {
+ return -EIO;
}
- TRACE("Decoding type");
+ trace_nbd_co_receive_request_decode_type(request->handle, request->type,
+ nbd_cmd_lookup(request->type));
- command = request->type & NBD_CMD_MASK_COMMAND;
- if (command != NBD_CMD_WRITE) {
+ if (request->type != NBD_CMD_WRITE) {
/* No payload, we are ready to read the next request. */
req->complete = true;
}
- if (command == NBD_CMD_DISC) {
+ if (request->type == NBD_CMD_DISC) {
/* Special case: we're going to disconnect without a reply,
* whether or not flags, from, or len are bogus */
- TRACE("Request type is DISCONNECT");
- rc = -EIO;
- goto out;
+ return -EIO;
}
/* Check for sanity in the parameters, part 1. Defer as many
* checks as possible until after reading any NBD_CMD_WRITE
* payload, so we can try and keep the connection alive. */
if ((request->from + request->len) < request->from) {
- LOG("integer overflow detected, you're probably being attacked");
- rc = -EINVAL;
- goto out;
+ error_setg(errp,
+ "integer overflow detected, you're probably being attacked");
+ return -EINVAL;
}
- if (command == NBD_CMD_READ || command == NBD_CMD_WRITE) {
+ if (request->type == NBD_CMD_READ || request->type == NBD_CMD_WRITE) {
if (request->len > NBD_MAX_BUFFER_SIZE) {
- LOG("len (%" PRIu32" ) is larger than max len (%u)",
- request->len, NBD_MAX_BUFFER_SIZE);
- rc = -EINVAL;
- goto out;
+ error_setg(errp, "len (%" PRIu32" ) is larger than max len (%u)",
+ request->len, NBD_MAX_BUFFER_SIZE);
+ return -EINVAL;
}
req->data = blk_try_blockalign(client->exp->blk, request->len);
if (req->data == NULL) {
- rc = -ENOMEM;
- goto out;
+ error_setg(errp, "No memory");
+ return -ENOMEM;
}
}
- if (command == NBD_CMD_WRITE) {
- TRACE("Reading %" PRIu32 " byte(s)", request->len);
-
- if (read_sync(client->ioc, req->data, request->len) != request->len) {
- LOG("reading from socket failed");
- rc = -EIO;
- goto out;
+ if (request->type == NBD_CMD_WRITE) {
+ if (nbd_read(client->ioc, req->data, request->len, errp) < 0) {
+ error_prepend(errp, "reading from socket failed: ");
+ return -EIO;
}
req->complete = true;
+
+ trace_nbd_co_receive_request_payload_received(request->handle,
+ request->len);
}
/* Sanity checks, part 2. */
if (request->from + request->len > client->exp->size) {
- LOG("operation past EOF; From: %" PRIu64 ", Len: %" PRIu32
- ", Size: %" PRIu64, request->from, request->len,
- (uint64_t)client->exp->size);
- rc = command == NBD_CMD_WRITE ? -ENOSPC : -EINVAL;
- goto out;
+ error_setg(errp, "operation past EOF; From: %" PRIu64 ", Len: %" PRIu32
+ ", Size: %" PRIu64, request->from, request->len,
+ (uint64_t)client->exp->size);
+ return request->type == NBD_CMD_WRITE ? -ENOSPC : -EINVAL;
}
- if (request->type & ~NBD_CMD_MASK_COMMAND & ~NBD_CMD_FLAG_FUA) {
- LOG("unsupported flags (got 0x%x)",
- request->type & ~NBD_CMD_MASK_COMMAND);
+ if (request->flags & ~(NBD_CMD_FLAG_FUA | NBD_CMD_FLAG_NO_HOLE)) {
+ error_setg(errp, "unsupported flags (got 0x%x)", request->flags);
+ return -EINVAL;
+ }
+ if (request->type != NBD_CMD_WRITE_ZEROES &&
+ (request->flags & NBD_CMD_FLAG_NO_HOLE)) {
+ error_setg(errp, "unexpected flags (got 0x%x)", request->flags);
return -EINVAL;
}
- rc = 0;
-
-out:
- client->recv_coroutine = NULL;
- nbd_update_can_read(client);
-
- return rc;
+ return 0;
}
-static void nbd_trip(void *opaque)
+/* Owns a reference to the NBDClient passed as opaque. */
+static coroutine_fn void nbd_trip(void *opaque)
{
NBDClient *client = opaque;
NBDExport *exp = client->exp;
- NBDRequest *req;
- struct nbd_request request;
- struct nbd_reply reply;
- ssize_t ret;
- uint32_t command;
+ NBDRequestData *req;
+ NBDRequest request = { 0 }; /* GCC thinks it can be used uninitialized */
+ NBDReply reply;
+ int ret;
int flags;
+ int reply_data_len = 0;
+ Error *local_err = NULL;
- TRACE("Reading request.");
+ trace_nbd_trip();
if (client->closing) {
+ nbd_client_put(client);
return;
}
req = nbd_request_get(client);
- ret = nbd_co_receive_request(req, &request);
- if (ret == -EAGAIN) {
- goto done;
- }
+ ret = nbd_co_receive_request(req, &request, &local_err);
+ client->recv_coroutine = NULL;
+ nbd_client_receive_next_request(client);
if (ret == -EIO) {
- goto out;
+ goto disconnect;
}
reply.handle = request.handle;
if (ret < 0) {
reply.error = -ret;
- goto error_reply;
+ goto reply;
}
- command = request.type & NBD_CMD_MASK_COMMAND;
if (client->closing) {
/*
goto done;
}
- switch (command) {
+ switch (request.type) {
case NBD_CMD_READ:
- TRACE("Request type is READ");
-
- if (request.type & NBD_CMD_FLAG_FUA) {
+ /* XXX: NBD Protocol only documents use of FUA with WRITE */
+ if (request.flags & NBD_CMD_FLAG_FUA) {
ret = blk_co_flush(exp->blk);
if (ret < 0) {
- LOG("flush failed");
+ error_setg_errno(&local_err, -ret, "flush failed");
reply.error = -ret;
- goto error_reply;
+ break;
}
}
ret = blk_pread(exp->blk, request.from + exp->dev_offset,
req->data, request.len);
if (ret < 0) {
- LOG("reading from file failed");
+ error_setg_errno(&local_err, -ret, "reading from file failed");
reply.error = -ret;
- goto error_reply;
+ break;
}
- TRACE("Read %" PRIu32" byte(s)", request.len);
- if (nbd_co_send_reply(req, &reply, request.len) < 0)
- goto out;
+ reply_data_len = request.len;
+
break;
case NBD_CMD_WRITE:
- TRACE("Request type is WRITE");
-
if (exp->nbdflags & NBD_FLAG_READ_ONLY) {
- TRACE("Server is read-only, return error");
reply.error = EROFS;
- goto error_reply;
+ break;
}
- TRACE("Writing to device");
-
flags = 0;
- if (request.type & NBD_CMD_FLAG_FUA) {
+ if (request.flags & NBD_CMD_FLAG_FUA) {
flags |= BDRV_REQ_FUA;
}
ret = blk_pwrite(exp->blk, request.from + exp->dev_offset,
req->data, request.len, flags);
if (ret < 0) {
- LOG("writing to file failed");
+ error_setg_errno(&local_err, -ret, "writing to file failed");
reply.error = -ret;
- goto error_reply;
}
- if (nbd_co_send_reply(req, &reply, 0) < 0) {
- goto out;
- }
break;
+ case NBD_CMD_WRITE_ZEROES:
+ if (exp->nbdflags & NBD_FLAG_READ_ONLY) {
+ error_setg(&local_err, "Server is read-only, return error");
+ reply.error = EROFS;
+ break;
+ }
+ flags = 0;
+ if (request.flags & NBD_CMD_FLAG_FUA) {
+ flags |= BDRV_REQ_FUA;
+ }
+ if (!(request.flags & NBD_CMD_FLAG_NO_HOLE)) {
+ flags |= BDRV_REQ_MAY_UNMAP;
+ }
+ ret = blk_pwrite_zeroes(exp->blk, request.from + exp->dev_offset,
+ request.len, flags);
+ if (ret < 0) {
+ error_setg_errno(&local_err, -ret, "writing to file failed");
+ reply.error = -ret;
+ }
+
+ break;
case NBD_CMD_DISC:
/* unreachable, thanks to special case in nbd_co_receive_request() */
abort();
case NBD_CMD_FLUSH:
- TRACE("Request type is FLUSH");
-
ret = blk_co_flush(exp->blk);
if (ret < 0) {
- LOG("flush failed");
+ error_setg_errno(&local_err, -ret, "flush failed");
reply.error = -ret;
}
- if (nbd_co_send_reply(req, &reply, 0) < 0) {
- goto out;
- }
+
break;
case NBD_CMD_TRIM:
- TRACE("Request type is TRIM");
ret = blk_co_pdiscard(exp->blk, request.from + exp->dev_offset,
request.len);
if (ret < 0) {
- LOG("discard failed");
+ error_setg_errno(&local_err, -ret, "discard failed");
reply.error = -ret;
}
- if (nbd_co_send_reply(req, &reply, 0) < 0) {
- goto out;
- }
+
break;
default:
- LOG("invalid request type (%" PRIu32 ") received", request.type);
+ error_setg(&local_err, "invalid request type (%" PRIu32 ") received",
+ request.type);
reply.error = EINVAL;
- error_reply:
- /* We must disconnect after NBD_CMD_WRITE if we did not
- * read the payload.
- */
- if (nbd_co_send_reply(req, &reply, 0) < 0 || !req->complete) {
- goto out;
- }
- break;
}
- TRACE("Request/Reply complete");
-
-done:
- nbd_request_put(req);
- return;
-
-out:
- nbd_request_put(req);
- client_close(client);
-}
-
-static void nbd_read(void *opaque)
-{
- NBDClient *client = opaque;
-
- if (client->recv_coroutine) {
- qemu_coroutine_enter(client->recv_coroutine);
- } else {
- qemu_coroutine_enter(qemu_coroutine_create(nbd_trip, client));
+reply:
+ if (local_err) {
+ /* If we are here local_err is not fatal error, already stored in
+ * reply.error */
+ error_report_err(local_err);
+ local_err = NULL;
}
-}
-static void nbd_restart_write(void *opaque)
-{
- NBDClient *client = opaque;
-
- qemu_coroutine_enter(client->send_coroutine);
-}
+ if (nbd_co_send_reply(req, &reply, reply_data_len, &local_err) < 0) {
+ error_prepend(&local_err, "Failed to send reply: ");
+ goto disconnect;
+ }
-static void nbd_set_handlers(NBDClient *client)
-{
- if (client->exp && client->exp->ctx) {
- aio_set_fd_handler(client->exp->ctx, client->sioc->fd,
- true,
- client->can_read ? nbd_read : NULL,
- client->send_coroutine ? nbd_restart_write : NULL,
- client);
+ /* We must disconnect after NBD_CMD_WRITE if we did not
+ * read the payload.
+ */
+ if (!req->complete) {
+ error_setg(&local_err, "Request handling failed in intermediate state");
+ goto disconnect;
}
-}
-static void nbd_unset_handlers(NBDClient *client)
-{
- if (client->exp && client->exp->ctx) {
- aio_set_fd_handler(client->exp->ctx, client->sioc->fd,
- true, NULL, NULL, NULL);
+done:
+ nbd_request_put(req);
+ nbd_client_put(client);
+ return;
+
+disconnect:
+ if (local_err) {
+ error_reportf_err(local_err, "Disconnect client, due to: ");
}
+ nbd_request_put(req);
+ client_close(client, true);
+ nbd_client_put(client);
}
-static void nbd_update_can_read(NBDClient *client)
+static void nbd_client_receive_next_request(NBDClient *client)
{
- bool can_read = client->recv_coroutine ||
- client->nb_requests < MAX_NBD_REQUESTS;
-
- if (can_read != client->can_read) {
- client->can_read = can_read;
- nbd_set_handlers(client);
-
- /* There is no need to invoke aio_notify(), since aio_set_fd_handler()
- * in nbd_set_handlers() will have taken care of that */
+ if (!client->recv_coroutine && client->nb_requests < MAX_NBD_REQUESTS) {
+ nbd_client_get(client);
+ client->recv_coroutine = qemu_coroutine_create(nbd_trip, client);
+ aio_co_schedule(client->exp->ctx, client->recv_coroutine);
}
}
static coroutine_fn void nbd_co_client_start(void *opaque)
{
- NBDClientNewData *data = opaque;
- NBDClient *client = data->client;
+ NBDClient *client = opaque;
NBDExport *exp = client->exp;
+ Error *local_err = NULL;
if (exp) {
nbd_export_get(exp);
- }
- if (nbd_negotiate(data)) {
- client_close(client);
- goto out;
+ QTAILQ_INSERT_TAIL(&exp->clients, client, next);
}
qemu_co_mutex_init(&client->send_lock);
- nbd_set_handlers(client);
- if (exp) {
- QTAILQ_INSERT_TAIL(&exp->clients, client, next);
+ if (nbd_negotiate(client, &local_err)) {
+ if (local_err) {
+ error_report_err(local_err);
+ }
+ client_close(client, false);
+ return;
}
-out:
- g_free(data);
+
+ nbd_client_receive_next_request(client);
}
+/*
+ * Create a new client listener on the given export @exp, using the
+ * given channel @sioc. Begin servicing it in a coroutine. When the
+ * connection closes, call @close_fn with an indication of whether the
+ * client completed negotiation.
+ */
void nbd_client_new(NBDExport *exp,
QIOChannelSocket *sioc,
QCryptoTLSCreds *tlscreds,
const char *tlsaclname,
- void (*close_fn)(NBDClient *))
+ void (*close_fn)(NBDClient *, bool))
{
NBDClient *client;
- NBDClientNewData *data = g_new(NBDClientNewData, 1);
+ Coroutine *co;
client = g_malloc0(sizeof(NBDClient));
client->refcount = 1;
object_ref(OBJECT(client->sioc));
client->ioc = QIO_CHANNEL(sioc);
object_ref(OBJECT(client->ioc));
- client->can_read = true;
- client->close = close_fn;
+ client->close_fn = close_fn;
- data->client = client;
- data->co = qemu_coroutine_create(nbd_co_client_start, data);
- qemu_coroutine_enter(data->co);
+ co = qemu_coroutine_create(nbd_co_client_start, client);
+ qemu_coroutine_enter(co);
}
projects / qemu.git / blobdiff