qemu-nbd tool (Anthony Liguori)

git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@4596 c046a42c-6fe2-441c-8c8c-71466251a162

qemu-nbd tool (Anthony Liguori)
git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@4596 c046a42c-6fe2-441c-8c8c-71466251a162
bellard
1 parent e00c1e71
Showing 7 changed files with 881 additions and 2 deletions
Makefile
configure
exec-all.h
nbd.c
nbd.h
qemu-nbd.c
qemu-nbd.texi
@@ -17,7 +17,7 @@ ifdef CONFIG_STATIC
 LDFLAGS += -static
 endif
 ifdef BUILD_DOCS
-DOCS=qemu-doc.html qemu-tech.html qemu.1 qemu-img.1
+DOCS=qemu-doc.html qemu-tech.html qemu.1 qemu-img.1 qemu-nbd.8
 else
 DOCS=
 endif
@@ -159,6 +159,10 @@ qemu-img-%.o: %.c
 %.o: %.c
 	$(CC) $(CFLAGS) $(CPPFLAGS) -c -o $@ $<
+qemu-nbd$(EXESUF):  qemu-nbd.o nbd.o qemu-img-block.o \
+		    $(QEMU_IMG_BLOCK_OBJS)
+	$(CC) $(LDFLAGS) -o $@ $^ -lz $(LIBS)
+
 # dyngen host tool
 dyngen$(EXESUF): dyngen.c
 	$(HOST_CC) $(CFLAGS) $(CPPFLAGS) -o $@ $^
@@ -191,6 +195,8 @@ install-doc: $(DOCS)
 ifndef CONFIG_WIN32
 	mkdir -p "$(DESTDIR)$(mandir)/man1"
 	$(INSTALL) qemu.1 qemu-img.1 "$(DESTDIR)$(mandir)/man1"
+	mkdir -p "$(DESTDIR)$(mandir)/man8"
+	$(INSTALL) qemu-nbd.8 "$(DESTDIR)$(mandir)/man8"
 endif
 install: all $(if $(BUILD_DOCS),install-doc)
@@ -244,6 +250,10 @@ qemu-img.1: qemu-img.texi
 	$(SRC_PATH)/texi2pod.pl $< qemu-img.pod
 	pod2man --section=1 --center=" " --release=" " qemu-img.pod > $@
+qemu-nbd.8: qemu-nbd.texi
+	$(SRC_PATH)/texi2pod.pl $< qemu-nbd.pod
+	pod2man --section=8 --center=" " --release=" " qemu-nbd.pod > $@
+
 info: qemu-doc.info qemu-tech.info
 dvi: qemu-doc.dvi qemu-tech.dvi
@@ -296,6 +306,7 @@ tarbin:
         $(bindir)/qemu-sh4eb \
         $(bindir)/qemu-cris \
         $(bindir)/qemu-img \
+        $(bindir)/qemu-nbd \
 	$(datadir)/bios.bin \
 	$(datadir)/vgabios.bin \
 	$(datadir)/vgabios-cirrus.bin \
@@ -309,6 +320,7 @@ tarbin:
 	$(docdir)/qemu-doc.html \
 	$(docdir)/qemu-tech.html \
 	$(mandir)/man1/qemu.1 $(mandir)/man1/qemu-img.1
+	$(mandir)/man8/qemu-nbd.8
 # Include automatically generated dependency files
 -include $(wildcard *.d audio/*.d slirp/*.d)
@@ -1105,6 +1105,9 @@ echo &quot;#define CONFIG_UNAME_RELEASE \&quot;$uname_release\&quot;&quot; &gt;&gt; $config_h
 tools=
 if test `expr "$target_list" : ".*softmmu.*"` != 0 ; then
   tools="qemu-img\$(EXESUF) $tools"
+  if [ "$linux" = "yes" ] ; then
+      tools="qemu-nbd\$(EXESUF) $tools"
+  fi
 fi
 echo "TOOLS=$tools" >> $config_mak
@@ -126,7 +126,7 @@ static inline int tlb_set_page(CPUState *env1, target_ulong vaddr,
 #define CODE_GEN_BUFFER_SIZE     (6 * 1024 * 1024)
 #else
 /* XXX: make it dynamic on x86 */
-#define CODE_GEN_BUFFER_SIZE     (16 * 1024 * 1024)
+#define CODE_GEN_BUFFER_SIZE     (64 * 1024 * 1024)
 #endif
 //#define CODE_GEN_BUFFER_SIZE     (128 * 1024)
+/*\
+ *  Copyright (C) 2005  Anthony Liguori <anthony@codemonkey.ws>
+ *
+ *  Network Block Device
+ *
+ *  This program is free software; you can redistribute it and/or modify
+ *  it under the terms of the GNU General Public License as published by
+ *  the Free Software Foundation; under version 2 of the License.
+ *
+ *  This program is distributed in the hope that it will be useful,
+ *  but WITHOUT ANY WARRANTY; without even the implied warranty of
+ *  MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ *  GNU General Public License for more details.
+ *
+ *  You should have received a copy of the GNU General Public License
+ *  along with this program; if not, write to the Free Software
+ *  Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307  USA
+\*/
+
+#include "nbd.h"
+
+#include <errno.h>
+#include <string.h>
+#include <sys/ioctl.h>
+#include <ctype.h>
+#include <inttypes.h>
+#include <sys/socket.h>
+#include <netinet/in.h>
+#include <netinet/tcp.h>
+#include <arpa/inet.h>
+#include <netdb.h>
+
+extern int verbose;
+
+#define LOG(msg, ...) do { \
+    fprintf(stderr, "%s:%s():L%d: " msg "\n", \
+            __FILE__, __FUNCTION__, __LINE__, ## __VA_ARGS__); \
+} while(0)
+
+#define TRACE(msg, ...) do { \
+    if (verbose) LOG(msg, ## __VA_ARGS__); \
+} while(0)
+
+/* This is all part of the "official" NBD API */
+
+#define NBD_REQUEST_MAGIC       0x25609513
+#define NBD_REPLY_MAGIC         0x67446698
+
+#define NBD_SET_SOCK            _IO(0xab, 0)
+#define NBD_SET_BLKSIZE         _IO(0xab, 1)
+#define NBD_SET_SIZE            _IO(0xab, 2)
+#define NBD_DO_IT               _IO(0xab, 3)
+#define NBD_CLEAR_SOCK          _IO(0xab, 4)
+#define NBD_CLEAR_QUE           _IO(0xab, 5)
+#define NBD_PRINT_DEBUG	        _IO(0xab, 6)
+#define NBD_SET_SIZE_BLOCKS	_IO(0xab, 7)
+#define NBD_DISCONNECT          _IO(0xab, 8)
+
+/* That's all folks */
+
+#define read_sync(fd, buffer, size) wr_sync(fd, buffer, size, true)
+#define write_sync(fd, buffer, size) wr_sync(fd, buffer, size, false)
+
+static size_t wr_sync(int fd, void *buffer, size_t size, bool do_read)
+{
+    size_t offset = 0;
+
+    while (offset < size) {
+        ssize_t len;
+
+        if (do_read) {
+            len = read(fd, buffer + offset, size - offset);
+        } else {
+            len = write(fd, buffer + offset, size - offset);
+        }
+
+        /* recoverable error */
+        if (len == -1 && errno == EAGAIN) {
+            continue;
+        }
+
+        /* eof */
+        if (len == 0) {
+            break;
+        }
+
+        /* unrecoverable error */
+        if (len == -1) {
+            return 0;
+        }
+
+        offset += len;
+    }
+
+    return offset;
+}
+
+static int tcp_socket_outgoing(const char *address, uint16_t port)
+{
+    int s;
+    struct in_addr in;
+    struct sockaddr_in addr;
+    int serrno;
+
+    s = socket(PF_INET, SOCK_STREAM, 0);
+    if (s == -1) {
+        return -1;
+    }
+
+    if (inet_aton(address, &in) == 0) {
+        struct hostent *ent;
+
+        ent = gethostbyname(address);
+        if (ent == NULL) {
+            goto error;
+        }
+
+        memcpy(&in, ent->h_addr, sizeof(in));
+    }
+
+    addr.sin_family = AF_INET;
+    addr.sin_port = htons(port);
+    memcpy(&addr.sin_addr.s_addr, &in, sizeof(in));
+
+    if (connect(s, (struct sockaddr *)&addr, sizeof(addr)) == -1) {
+        goto error;
+    }
+
+    return s;
+error:
+    serrno = errno;
+    close(s);
+    errno = serrno;
+    return -1;
+}
+
+int tcp_socket_incoming(const char *address, uint16_t port)
+{
+    int s;
+    struct in_addr in;
+    struct sockaddr_in addr;
+    int serrno;
+    int opt;
+
+    s = socket(PF_INET, SOCK_STREAM, 0);
+    if (s == -1) {
+        return -1;
+    }
+
+    if (inet_aton(address, &in) == 0) {
+        struct hostent *ent;
+
+        ent = gethostbyname(address);
+        if (ent == NULL) {
+            goto error;
+        }
+
+        memcpy(&in, ent->h_addr, sizeof(in));
+    }
+
+    addr.sin_family = AF_INET;
+    addr.sin_port = htons(port);
+    memcpy(&addr.sin_addr.s_addr, &in, sizeof(in));
+
+    opt = 1;
+    if (setsockopt(s, SOL_SOCKET, SO_REUSEADDR, &opt, sizeof(opt)) == -1) {
+        goto error;
+    }
+
+    if (bind(s, (struct sockaddr *)&addr, sizeof(addr)) == -1) {
+        goto error;
+    }
+
+    if (listen(s, 128) == -1) {
+        goto error;
+    }
+
+    return s;
+error:
+    serrno = errno;
+    close(s);
+    errno = serrno;
+    return -1;
+}
+
+/* Basic flow
+
+   Server         Client
+
+   Negotiate
+                  Request
+   Response
+                  Request
+   Response
+                  ...
+   ...
+                  Request (type == 2)
+*/
+
+int nbd_negotiate(BlockDriverState *bs, int csock, off_t size)
+{
+	char buf[8 + 8 + 8 + 128];
+
+	/* Negotiate
+	   [ 0 ..   7]   passwd   ("NBDMAGIC")
+	   [ 8 ..  15]   magic    (0x00420281861253)
+	   [16 ..  23]   size
+	   [24 .. 151]   reserved (0)
+	 */
+
+	TRACE("Beginning negotiation.");
+	memcpy(buf, "NBDMAGIC", 8);
+	cpu_to_be64w((uint64_t*)(buf + 8), 0x00420281861253LL);
+	cpu_to_be64w((uint64_t*)(buf + 16), size);
+	memset(buf + 24, 0, 128);
+
+	if (write_sync(csock, buf, sizeof(buf)) != sizeof(buf)) {
+		LOG("write failed");
+		errno = EINVAL;
+		return -1;
+	}
+
+	TRACE("Negotation succeeded.");
+
+	return 0;
+}
+
+int nbd_receive_negotiate(int fd, int csock)
+{
+	char buf[8 + 8 + 8 + 128];
+	uint64_t magic;
+	off_t size;
+	size_t blocksize;
+
+	TRACE("Receiving negotation.");
+
+	if (read_sync(csock, buf, sizeof(buf)) != sizeof(buf)) {
+		LOG("read failed");
+		errno = EINVAL;
+		return -1;
+	}
+
+	magic = be64_to_cpup((uint64_t*)(buf + 8));
+	size = be64_to_cpup((uint64_t*)(buf + 16));
+	blocksize = 1024;
+
+	TRACE("Magic is %c%c%c%c%c%c%c%c",
+	      isprint(buf[0]) ? buf[0] : '.',
+	      isprint(buf[1]) ? buf[1] : '.',
+	      isprint(buf[2]) ? buf[2] : '.',
+	      isprint(buf[3]) ? buf[3] : '.',
+	      isprint(buf[4]) ? buf[4] : '.',
+	      isprint(buf[5]) ? buf[5] : '.',
+	      isprint(buf[6]) ? buf[6] : '.',
+	      isprint(buf[7]) ? buf[7] : '.');
+	TRACE("Magic is 0x%" PRIx64, magic);
+	TRACE("Size is %" PRIu64, size);
+
+	if (memcmp(buf, "NBDMAGIC", 8) != 0) {
+		LOG("Invalid magic received");
+		errno = EINVAL;
+		return -1;
+	}
+
+	TRACE("Checking magic");
+
+	if (magic != 0x00420281861253LL) {
+		LOG("Bad magic received");
+		errno = EINVAL;
+		return -1;
+	}
+
+	TRACE("Setting block size to %lu", (unsigned long)blocksize);
+
+	if (ioctl(fd, NBD_SET_BLKSIZE, blocksize) == -1) {
+		int serrno = errno;
+		LOG("Failed setting NBD block size");
+		errno = serrno;
+		return -1;
+	}
+
+	TRACE("Setting size to %llu block(s)",
+	      (unsigned long long)(size / blocksize));
+
+	if (ioctl(fd, NBD_SET_SIZE_BLOCKS, size / blocksize) == -1) {
+		int serrno = errno;
+		LOG("Failed setting size (in blocks)");
+		errno = serrno;
+		return -1;
+	}
+
+	TRACE("Clearing NBD socket");
+
+	if (ioctl(fd, NBD_CLEAR_SOCK) == -1) {
+		int serrno = errno;
+		LOG("Failed clearing NBD socket");
+		errno = serrno;
+		return -1;
+	}
+
+	TRACE("Setting NBD socket");
+
+	if (ioctl(fd, NBD_SET_SOCK, csock) == -1) {
+		int serrno = errno;
+		LOG("Failed to set NBD socket");
+		errno = serrno;
+		return -1;
+	}
+
+	TRACE("Negotiation ended");
+
+	return 0;
+}
+
+int nbd_disconnect(int fd)
+{
+	ioctl(fd, NBD_CLEAR_QUE);
+	ioctl(fd, NBD_DISCONNECT);
+	ioctl(fd, NBD_CLEAR_SOCK);
+	return 0;
+}
+
+int nbd_client(int fd, int csock)
+{
+	int ret;
+	int serrno;
+
+	TRACE("Doing NBD loop");
+
+	ret = ioctl(fd, NBD_DO_IT);
+	serrno = errno;
+
+	TRACE("NBD loop returned %d: %s", ret, strerror(serrno));
+
+	TRACE("Clearing NBD queue");
+	ioctl(fd, NBD_CLEAR_QUE);
+
+	TRACE("Clearing NBD socket");
+	ioctl(fd, NBD_CLEAR_SOCK);
+
+	errno = serrno;
+	return ret;
+}
+
+int nbd_trip(BlockDriverState *bs, int csock, off_t size, uint64_t dev_offset, off_t *offset, bool readonly)
+{
+#ifndef _REENTRANT
+	static uint8_t data[1024 * 1024]; // keep this off of the stack
+#else
+	uint8_t data[1024 * 1024];
+#endif
+	uint8_t buf[4 + 4 + 8 + 8 + 4];
+	uint32_t magic;
+	uint32_t type;
+	uint64_t from;
+	uint32_t len;
+
+	TRACE("Reading request.");
+
+	if (read_sync(csock, buf, sizeof(buf)) != sizeof(buf)) {
+		LOG("read failed");
+		errno = EINVAL;
+		return -1;
+	}
+
+	/* Request
+	  [ 0 ..  3]   magic   (NBD_REQUEST_MAGIC)
+	  [ 4 ..  7]   type    (0 == READ, 1 == WRITE)
+	  [ 8 .. 15]   handle
+	  [16 .. 23]   from
+	  [24 .. 27]   len
+	 */
+
+	magic = be32_to_cpup((uint32_t*)buf);
+	type  = be32_to_cpup((uint32_t*)(buf + 4));
+	from  = be64_to_cpup((uint64_t*)(buf + 16));
+	len   = be32_to_cpup((uint32_t*)(buf + 24));
+
+	TRACE("Got request: "
+	      "{ magic = 0x%x, .type = %d, from = %" PRIu64" , len = %u }",
+	      magic, type, from, len);
+
+
+	if (magic != NBD_REQUEST_MAGIC) {
+		LOG("invalid magic (got 0x%x)", magic);
+		errno = EINVAL;
+		return -1;
+	}
+
+	if (len > sizeof(data)) {
+		LOG("len (%u) is larger than max len (%lu)",
+		    len, sizeof(data));
+		errno = EINVAL;
+		return -1;
+	}
+
+	if ((from + len) < from) {
+		LOG("integer overflow detected! "
+		    "you're probably being attacked");
+		errno = EINVAL;
+		return -1;
+	}
+
+	if ((from + len) > size) {
+	        LOG("From: %" PRIu64 ", Len: %u, Size: %" PRIu64
+		    ", Offset: %" PRIu64 "\n",
+		     from, len, size, dev_offset);
+		LOG("requested operation past EOF--bad client?");
+		errno = EINVAL;
+		return -1;
+	}
+
+	/* Reply
+	 [ 0 ..  3]    magic   (NBD_REPLY_MAGIC)
+	 [ 4 ..  7]    error   (0 == no error)
+         [ 7 .. 15]    handle
+	 */
+	cpu_to_be32w((uint32_t*)buf, NBD_REPLY_MAGIC);
+	cpu_to_be32w((uint32_t*)(buf + 4), 0);
+
+	TRACE("Decoding type");
+
+	switch (type) {
+	case 0:
+		TRACE("Request type is READ");
+
+		if (bdrv_read(bs, (from + dev_offset) / 512, data, len / 512) == -1) {
+			LOG("reading from file failed");
+			errno = EINVAL;
+			return -1;
+		}
+		*offset += len;
+
+		TRACE("Read %u byte(s)", len);
+
+		TRACE("Sending OK response");
+
+		if (write_sync(csock, buf, 16) != 16) {
+			LOG("writing to socket failed");
+			errno = EINVAL;
+			return -1;
+		}
+
+		TRACE("Sending data to client");
+
+		if (write_sync(csock, data, len) != len) {
+			LOG("writing to socket failed");
+			errno = EINVAL;
+			return -1;
+		}
+		break;
+	case 1:
+		TRACE("Request type is WRITE");
+
+		TRACE("Reading %u byte(s)", len);
+
+		if (read_sync(csock, data, len) != len) {
+			LOG("reading from socket failed");
+			errno = EINVAL;
+			return -1;
+		}
+
+		if (readonly) {
+			TRACE("Server is read-only, return error");
+
+			cpu_to_be32w((uint32_t*)(buf + 4), 1);
+		} else {
+			TRACE("Writing to device");
+
+			if (bdrv_write(bs, (from + dev_offset) / 512, data, len / 512) == -1) {
+				LOG("writing to file failed");
+				errno = EINVAL;
+				return -1;
+			}
+
+			*offset += len;
+		}
+
+		TRACE("Sending response to client");
+
+		if (write_sync(csock, buf, 16) != 16) {
+			LOG("writing to socket failed");
+			errno = EINVAL;
+			return -1;
+		}
+		break;
+	case 2:
+		TRACE("Request type is DISCONNECT");
+		errno = 0;
+		return 1;
+	default:
+		LOG("invalid request type (%u) received", type);
+		errno = EINVAL;
+		return -1;
+	}
+
+	TRACE("Request/Reply complete");
+
+	return 0;
+}
+/*\
+ *  Copyright (C) 2005  Anthony Liguori <anthony@codemonkey.ws>
+ *
+ *  Network Block Device
+ *
+ *  This program is free software; you can redistribute it and/or modify
+ *  it under the terms of the GNU General Public License as published by
+ *  the Free Software Foundation; under version 2 of the License.
+ *
+ *  This program is distributed in the hope that it will be useful,
+ *  but WITHOUT ANY WARRANTY; without even the implied warranty of
+ *  MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ *  GNU General Public License for more details.
+ *
+ *  You should have received a copy of the GNU General Public License
+ *  along with this program; if not, write to the Free Software
+ *  Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307  USA
+\*/
+
+#ifndef NBD_H
+#define NBD_H
+
+#include <sys/types.h>
+#include <stdbool.h>
+
+#include <qemu-common.h>
+#include "block_int.h"
+
+int tcp_socket_incoming(const char *address, uint16_t port);
+
+int nbd_negotiate(BlockDriverState *bs, int csock, off_t size);
+int nbd_receive_negotiate(int fd, int csock);
+int nbd_trip(BlockDriverState *bs, int csock, off_t size, uint64_t dev_offset, off_t *offset, bool readonly);
+int nbd_client(int fd, int csock);
+int nbd_disconnect(int fd);
+
+#endif
+/*\
+ *  Copyright (C) 2005  Anthony Liguori <anthony@codemonkey.ws>
+ *
+ *  Network Block Device
+ *
+ *  This program is free software; you can redistribute it and/or modify
+ *  it under the terms of the GNU General Public License as published by
+ *  the Free Software Foundation; under version 2 of the License.
+ *
+ *  This program is distributed in the hope that it will be useful,
+ *  but WITHOUT ANY WARRANTY; without even the implied warranty of
+ *  MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ *  GNU General Public License for more details.
+ *
+ *  You should have received a copy of the GNU General Public License
+ *  along with this program; if not, write to the Free Software
+ *  Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307  USA
+ */
+
+#include <qemu-common.h>
+#include "block_int.h"
+#include "nbd.h"
+
+#include <malloc.h>
+#include <stdarg.h>
+#include <stdio.h>
+#include <getopt.h>
+#include <err.h>
+#include <sys/socket.h>
+#include <netinet/in.h>
+#include <netinet/tcp.h>
+#include <arpa/inet.h>
+
+int verbose;
+
+static void usage(const char *name)
+{
+    printf(
+"Usage: %s [OPTIONS] FILE\n"
+"QEMU Disk Network Block Device Server\n"
+"\n"
+"  -p, --port=PORT      port to listen on (default `1024')\n"
+"  -o, --offset=OFFSET  offset into the image\n"
+"  -b, --bind=IFACE     interface to bind to (default `0.0.0.0')\n"
+"  -r, --read-only      export read-only\n"
+"  -P, --partition=NUM  only expose partition NUM\n"
+"  -v, --verbose        display extra debugging information\n"
+"  -h, --help           display this help and exit\n"
+"  -V, --version        output version information and exit\n"
+"\n"
+"Report bugs to <anthony@codemonkey.ws>\n"
+    , name);
+}
+
+static void version(const char *name)
+{
+    printf(
+"qemu-nbd version 0.0.1\n"
+"Written by Anthony Liguori.\n"
+"\n"
+"Copyright (C) 2006 Anthony Liguori <anthony@codemonkey.ws>.\n"
+"This is free software; see the source for copying conditions.  There is NO\n"
+"warranty; not even for MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.\n"
+    );
+}
+
+struct partition_record
+{
+    uint8_t bootable;
+    uint8_t start_head;
+    uint32_t start_cylinder;
+    uint8_t start_sector;
+    uint8_t system;
+    uint8_t end_head;
+    uint8_t end_cylinder;
+    uint8_t end_sector;
+    uint32_t start_sector_abs;
+    uint32_t nb_sectors_abs;
+};
+
+static void read_partition(uint8_t *p, struct partition_record *r)
+{
+    r->bootable = p[0];
+    r->start_head = p[1];
+    r->start_cylinder = p[3] | ((p[2] << 2) & 0x0300);
+    r->start_sector = p[2] & 0x3f;
+    r->system = p[4];
+    r->end_head = p[5];
+    r->end_cylinder = p[7] | ((p[6] << 2) & 0x300);
+    r->end_sector = p[6] & 0x3f;
+    r->start_sector_abs = p[8] | p[9] << 8 | p[10] << 16 | p[11] << 24;
+    r->nb_sectors_abs = p[12] | p[13] << 8 | p[14] << 16 | p[15] << 24;
+}
+
+static int find_partition(BlockDriverState *bs, int partition,
+                          off_t *offset, off_t *size)
+{
+    struct partition_record mbr[4];
+    uint8_t data[512];
+    int i;
+    int ext_partnum = 4;
+
+    if (bdrv_read(bs, 0, data, 1))
+        errx(EINVAL, "error while reading");
+
+    if (data[510] != 0x55 || data[511] != 0xaa) {
+        errno = -EINVAL;
+        return -1;
+    }
+
+    for (i = 0; i < 4; i++) {
+        read_partition(&data[446 + 16 * i], &mbr[i]);
+
+        if (!mbr[i].nb_sectors_abs)
+            continue;
+
+        if (mbr[i].system == 0xF || mbr[i].system == 0x5) {
+            struct partition_record ext[4];
+            uint8_t data1[512];
+            int j;
+
+            if (bdrv_read(bs, mbr[i].start_sector_abs, data1, 1))
+                errx(EINVAL, "error while reading");
+
+            for (j = 0; j < 4; j++) {
+                read_partition(&data1[446 + 16 * j], &ext[j]);
+                if (!ext[j].nb_sectors_abs)
+                    continue;
+
+                if ((ext_partnum + j + 1) == partition) {
+                    *offset = (uint64_t)ext[j].start_sector_abs << 9;
+                    *size = (uint64_t)ext[j].nb_sectors_abs << 9;
+                    return 0;
+                }
+            }
+            ext_partnum += 4;
+        } else if ((i + 1) == partition) {
+            *offset = (uint64_t)mbr[i].start_sector_abs << 9;
+            *size = (uint64_t)mbr[i].nb_sectors_abs << 9;
+            return 0;
+        }
+    }
+
+    errno = -ENOENT;
+    return -1;
+}
+
+int main(int argc, char **argv)
+{
+    BlockDriverState *bs;
+    off_t dev_offset = 0;
+    off_t offset = 0;
+    bool readonly = false;
+    const char *bindto = "0.0.0.0";
+    int port = 1024;
+    int sock, csock;
+    struct sockaddr_in addr;
+    socklen_t addr_len = sizeof(addr);
+    off_t fd_size;
+    const char *sopt = "hVbo:p:rsP:v";
+    struct option lopt[] = {
+        { "help", 0, 0, 'h' },
+        { "version", 0, 0, 'V' },
+        { "bind", 1, 0, 'b' },
+        { "port", 1, 0, 'p' },
+        { "offset", 1, 0, 'o' },
+        { "read-only", 0, 0, 'r' },
+        { "partition", 1, 0, 'P' },
+        { "snapshot", 0, 0, 's' },
+        { "verbose", 0, 0, 'v' },
+    };
+    int ch;
+    int opt_ind = 0;
+    int li;
+    char *end;
+    bool snapshot = false;
+    int partition = -1;
+
+    while ((ch = getopt_long(argc, argv, sopt, lopt, &opt_ind)) != -1) {
+        switch (ch) {
+        case 's':
+            snapshot = true;
+            break;
+        case 'b':
+            bindto = optarg;
+            break;
+        case 'p':
+            li = strtol(optarg, &end, 0);
+            if (*end) {
+                errx(EINVAL, "Invalid port `%s'", optarg);
+            }
+            if (li < 1 || li > 65535) {
+                errx(EINVAL, "Port out of range `%s'", optarg);
+            }
+            port = (uint16_t)li;
+            break;
+        case 'o':
+                dev_offset = strtoll (optarg, &end, 0);
+            if (*end) {
+                errx(EINVAL, "Invalid offset `%s'", optarg);
+            }
+            if (dev_offset < 0) {
+                errx(EINVAL, "Offset must be positive `%s'", optarg);
+            }
+            break;
+        case 'r':
+            readonly = true;
+            break;
+        case 'P':
+            partition = strtol(optarg, &end, 0);
+            if (*end)
+                errx(EINVAL, "Invalid partition `%s'", optarg);
+            if (partition < 1 || partition > 8)
+                errx(EINVAL, "Invalid partition %d", partition);
+            break;
+        case 'v':
+            verbose = 1;
+            break;
+        case 'V':
+            version(argv[0]);
+            exit(0);
+            break;
+        case 'h':
+            usage(argv[0]);
+            exit(0);
+            break;
+        case '?':
+            errx(EINVAL, "Try `%s --help' for more information.",
+                 argv[0]);
+        }
+    }
+
+    if ((argc - optind) != 1) {
+        errx(EINVAL, "Invalid number of argument.\n"
+             "Try `%s --help' for more information.",
+             argv[0]);
+    }
+
+    bdrv_init();
+
+    bs = bdrv_new("hda");
+    if (bs == NULL)
+        return 1;
+
+    if (bdrv_open(bs, argv[optind], snapshot) == -1)
+        return 1;
+
+    fd_size = bs->total_sectors * 512;
+
+    if (partition != -1 &&
+        find_partition(bs, partition, &dev_offset, &fd_size))
+        errx(errno, "Could not find partition %d", partition);
+
+    sock = tcp_socket_incoming(bindto, port);
+    if (sock == -1)
+        return 1;
+
+    csock = accept(sock,
+               (struct sockaddr *)&addr,
+               &addr_len);
+    if (csock == -1)
+        return 1;
+
+    /* new fd_size is calculated by find_partition */
+    if (nbd_negotiate(bs, csock, fd_size) == -1)
+        return 1;
+
+    while (nbd_trip(bs, csock, fd_size, dev_offset, &offset, readonly) == 0);
+
+    close(csock);
+    close(sock);
+    bdrv_close(bs);
+
+    return 0;
+}
+@example
+@c man begin SYNOPSIS
+usage: qemu-nbd [OPTION]...  @var{filename}
+@c man end
+@end example
+
+@c man begin DESCRIPTION
+
+Export Qemu disk image using NBD protocol.
+
+@c man end
+
+@c man begin OPTIONS
+@table @var
+@item filename
+ is a disk image filename
+@item -p, --port=PORT
+  port to listen on (default `1024')
+@item -o, --offset=OFFSET
+  offset into the image
+@item -b, --bind=IFACE
+  interface to bind to (default `0.0.0.0')
+@item -r, --read-only
+  export read-only
+@item -P, --partition=NUM
+  only expose partition NUM
+@item -v, --verbose
+  display extra debugging information
+@item -h, --help
+  display this help and exit
+@item -V, --version
+  output version information and exit
+@end table
+
+@c man end
+
+@ignore
+
+@setfilename qemu-nbd
+@settitle QEMU Disk Network Block Device Server
+
+@c man begin AUTHOR
+Copyright (C) 2006 Anthony Liguori <anthony@codemonkey.ws>.
+This is free software; see the source for copying conditions.  There is NO
+warranty; not even for MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.
+@c man end
+
+@c man begin SEEALSO
+qemu-img(1)
+@c man end
+
+@end ignore