Re: [Qemu-devel] [PATCH][v3] Introduce NBD named exports.

[Laurent Vivier]

>On 08/25/2010 03:48 PM, Laurent Vivier wrote:
>> This patch allows to connect Qemu using NBD protocol to an nbd-server
>> using named exports.
>>    
>
>Does using a named export make the NBD protocol incompatible with a 
>server that doesn't support named exports?  IOW, it's essentially a new 
>protocol layered underneath of NBD?

Named exports is fully compatible with old nbd-server protocol.

All is done in the "negotiate" part: there is a new magic number which is used 
when nbd-server uses named exports, this magic number allows the client to send 
a name to select the exported disk. Then client receives disk size and data as 
usual.

I have tested this patch with current "qemu-nbd" (witch tcp and unix socket) 
and it is working fine.

Regards,
Laurent

>Regards,
>
>Anthony Liguori
>
>> For instance, if on the host "isoserver", in /etc/nbd-server/config, you 
>have:
>>
>> [generic]
>> [debian-500-ppc-netinst]
>>          exportname = /ISO/debian-500-powerpc-netinst.iso
>> [Fedora-10-ppc-netinst]
>>          exportname = /ISO/Fedora-10-ppc-netinst.iso
>>
>> You can connect to it, using:
>>
>>      qemu -cdrom nbd:isoserver:exportname=debian-500-ppc-netinst
>>      qemu -cdrom nbd:isoserver:exportname=Fedora-10-ppc-netinst
>>
>> NOTE: you need at least nbd-server 2.9.18
>>
>> Signed-off-by: Laurent Vivier<address@hidden>
>> ---
>> v3: change sizeof() by strlen()
>> v2: address Kevin comments (CODING STYLE)
>>
>>   block/nbd.c   |   68 +++++++++++++++++++++++---------
>>   nbd.c         |  118 
>+++++++++++++++++++++++++++++++++++++++++++++++++++------
>>   nbd.h         |    5 ++-
>>   qemu-doc.texi |    7 +++
>>   qemu-nbd.c    |    4 +-
>>   5 files changed, 169 insertions(+), 33 deletions(-)
>>
>> diff --git a/block/nbd.c b/block/nbd.c
>> index a1ec123..5e9d6cb 100644
>> --- a/block/nbd.c
>> +++ b/block/nbd.c
>> @@ -33,6 +33,8 @@
>>   #include<sys/types.h>
>>   #include<unistd.h>
>>
>> +#define EN_OPTSTR ":exportname="
>> +
>>   typedef struct BDRVNBDState {
>>       int sock;
>>       off_t size;
>> @@ -42,55 +44,83 @@ typedef struct BDRVNBDState {
>>   static int nbd_open(BlockDriverState *bs, const char* filename, int 
>flags)
>>   {
>>       BDRVNBDState *s = bs->opaque;
>> +    uint32_t nbdflags;
>> +
>> +    char *file;
>> +    char *name;
>>       const char *host;
>>       const char *unixpath;
>>       int sock;
>>       off_t size;
>>       size_t blocksize;
>>       int ret;
>> +    int err = -EINVAL;
>> +
>> +    file = qemu_strdup(filename);
>>
>> -    if (!strstart(filename, "nbd:",&host))
>> -        return -EINVAL;
>> +    name = strstr(file, EN_OPTSTR);
>> +    if (name) {
>> +        if (name[strlen(EN_OPTSTR)] == 0) {
>> +            goto out;
>> +        }
>> +        name[0] = 0;
>> +        name += strlen(EN_OPTSTR);
>> +    }
>> +
>> +    if (!strstart(file, "nbd:",&host)) {
>> +        goto out;
>> +    }
>>
>>       if (strstart(host, "unix:",&unixpath)) {
>>
>> -        if (unixpath[0] != '/')
>> -            return -EINVAL;
>> +        if (unixpath[0] != '/') {
>> +            goto out;
>> +        }
>>
>>           sock = unix_socket_outgoing(unixpath);
>>
>>       } else {
>> -        uint16_t port;
>> +        uint16_t port = NBD_DEFAULT_PORT;
>>           char *p, *r;
>>           char hostname[128];
>>
>>           pstrcpy(hostname, 128, host);
>>
>>           p = strchr(hostname, ':');
>> -        if (p == NULL)
>> -            return -EINVAL;
>> +        if (p != NULL) {
>> +            *p = '\0';
>> +            p++;
>> +
>> +            port = strtol(p,&r, 0);
>> +            if (r == p) {
>> +                goto out;
>> +            }
>> +        } else if (name == NULL) {
>> +            goto out;
>> +        }
>>
>> -        *p = '\0';
>> -        p++;
>> -
>> -        port = strtol(p,&r, 0);
>> -        if (r == p)
>> -            return -EINVAL;
>>           sock = tcp_socket_outgoing(hostname, port);
>>       }
>>
>> -    if (sock == -1)
>> -        return -errno;
>> +    if (sock == -1) {
>> +        err = -errno;
>> +        goto out;
>> +    }
>>
>> -    ret = nbd_receive_negotiate(sock,&size,&blocksize);
>> -    if (ret == -1)
>> -        return -errno;
>> +    ret = nbd_receive_negotiate(sock, name,&nbdflags,&size,&blocksize);
>> +    if (ret == -1) {
>> +        err = -errno;
>> +        goto out;
>> +    }
>>
>>       s->sock = sock;
>>       s->size = size;
>>       s->blocksize = blocksize;
>> +    err = 0;
>>
>> -    return 0;
>> +out:
>> +    qemu_free(file);
>> +    return err;
>>   }
>>
>>   static int nbd_read(BlockDriverState *bs, int64_t sector_num,
>> diff --git a/nbd.c b/nbd.c
>> index a9f295f..30dec8d 100644
>> --- a/nbd.c
>> +++ b/nbd.c
>> @@ -62,6 +62,8 @@
>>   #define NBD_SET_SIZE_BLOCKS        _IO(0xab, 7)
>>   #define NBD_DISCONNECT          _IO(0xab, 8)
>>
>> +#define NBD_OPT_EXPORT_NAME (1<<  0)
>> +
>>   /* That's all folks */
>>
>>   #define read_sync(fd, buffer, size) nbd_wr_sync(fd, buffer, size, true)
>> @@ -296,22 +298,27 @@ int nbd_negotiate(int csock, off_t size)
>>      return 0;
>>   }
>>
>> -int nbd_receive_negotiate(int csock, off_t *size, size_t *blocksize)
>> +int nbd_receive_negotiate(int csock, const char *name, uint32_t *flags,
>> +                          off_t *size, size_t *blocksize)
>>   {
>> -    char buf[8 + 8 + 8 + 128];
>> -    uint64_t magic;
>> +    char buf[256];
>> +    uint64_t magic, s;
>> +    uint16_t tmp;
>>
>>      TRACE("Receiving negotation.");
>>
>> -    if (read_sync(csock, buf, sizeof(buf)) != sizeof(buf)) {
>> +    if (read_sync(csock, buf, 8) != 8) {
>>              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;
>> +    buf[8] = '\0';
>> +    if (strlen(buf) == 0) {
>> +            LOG("server connection closed");
>> +            errno = EINVAL;
>> +            return -1;
>> +    }
>>
>>      TRACE("Magic is %c%c%c%c%c%c%c%c",
>>            qemu_isprint(buf[0]) ? buf[0] : '.',
>> @@ -322,8 +329,6 @@ int nbd_receive_negotiate(int csock, off_t *size, size_t 
>*blocksize)
>>            qemu_isprint(buf[5]) ? buf[5] : '.',
>>            qemu_isprint(buf[6]) ? buf[6] : '.',
>>            qemu_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");
>> @@ -331,10 +336,99 @@ int nbd_receive_negotiate(int csock, off_t *size, 
>size_t *blocksize)
>>              return -1;
>>      }
>>
>> -    TRACE("Checking magic");
>> +    if (read_sync(csock,&magic, sizeof(magic)) != sizeof(magic)) {
>> +            LOG("read failed");
>> +            errno = EINVAL;
>> +            return -1;
>> +    }
>> +    magic = be64_to_cpu(magic);
>> +    TRACE("Magic is 0x%" PRIx64, magic);
>> +
>> +    if (name) {
>> +            uint32_t reserved = 0;
>> +            uint32_t opt;
>> +            uint32_t namesize;
>> +
>> +            TRACE("Checking magic (opts_magic)");
>> +            if (magic != 0x49484156454F5054LL) {
>> +                    LOG("Bad magic received");
>> +                    errno = EINVAL;
>> +                    return -1;
>> +            }
>> +            if (read_sync(csock,&tmp, sizeof(tmp)) != sizeof(tmp)) {
>> +                    LOG("flags read failed");
>> +                    errno = EINVAL;
>> +                    return -1;
>> +            }
>> +            *flags = be16_to_cpu(tmp)<<  16;
>> +            /* reserved for future use */
>> +            if (write_sync(csock,&reserved, sizeof(reserved)) !=
>> +                sizeof(reserved)) {
>> +                    LOG("write failed (reserved)");
>> +                    errno = EINVAL;
>> +                    return -1;
>> +            }
>> +            /* write the export name */
>> +            magic = cpu_to_be64(magic);
>> +            if (write_sync(csock,&magic, sizeof(magic)) != sizeof(magic)) {
>> +                    LOG("write failed (magic)");
>> +                    errno = EINVAL;
>> +                    return -1;
>> +            }
>> +            opt = cpu_to_be32(NBD_OPT_EXPORT_NAME);
>> +            if (write_sync(csock,&opt, sizeof(opt)) != sizeof(opt)) {
>> +                    LOG("write failed (opt)");
>> +                    errno = EINVAL;
>> +                    return -1;
>> +            }
>> +            namesize = cpu_to_be32(strlen(name));
>> +            if (write_sync(csock,&namesize, sizeof(namesize)) !=
>> +                sizeof(namesize)) {
>> +                    LOG("write failed (namesize)");
>> +                    errno = EINVAL;
>> +                    return -1;
>> +            }
>> +            if (write_sync(csock, (char*)name, strlen(name)) != 
>> strlen(name)) {
>> +                    LOG("write failed (name)");
>> +                    errno = EINVAL;
>> +                    return -1;
>> +            }
>> +    } else {
>> +            TRACE("Checking magic (cli_magic)");
>> +
>> +            if (magic != 0x00420281861253LL) {
>> +                    LOG("Bad magic received");
>> +                    errno = EINVAL;
>> +                    return -1;
>> +            }
>> +    }
>> +
>> +    if (read_sync(csock,&s, sizeof(s)) != sizeof(s)) {
>> +            LOG("read failed");
>> +            errno = EINVAL;
>> +            return -1;
>> +    }
>> +    *size = be64_to_cpu(s);
>> +    *blocksize = 1024;
>> +    TRACE("Size is %" PRIu64, *size);
>>
>> -    if (magic != 0x00420281861253LL) {
>> -            LOG("Bad magic received");
>> +    if (!name) {
>> +            if (read_sync(csock, flags, sizeof(*flags)) != sizeof(*flags)) {
>> +                    LOG("read failed (flags)");
>> +                    errno = EINVAL;
>> +                    return -1;
>> +            }
>> +            *flags = be32_to_cpup(flags);
>> +    } else {
>> +            if (read_sync(csock,&tmp, sizeof(tmp)) != sizeof(tmp)) {
>> +                    LOG("read failed (tmp)");
>> +                    errno = EINVAL;
>> +                    return -1;
>> +            }
>> +            *flags |= be32_to_cpu(tmp);
>> +    }
>> +    if (read_sync(csock,&buf, 124) != 124) {
>> +            LOG("read failed (buf)");
>>              errno = EINVAL;
>>              return -1;
>>      }
>> diff --git a/nbd.h b/nbd.h
>> index 5a1fbdf..8ff65a1 100644
>> --- a/nbd.h
>> +++ b/nbd.h
>> @@ -42,6 +42,8 @@ enum {
>>       NBD_CMD_DISC = 2
>>   };
>>
>> +#define NBD_DEFAULT_PORT    10809
>> +
>>   size_t nbd_wr_sync(int fd, void *buffer, size_t size, bool do_read);
>>   int tcp_socket_outgoing(const char *address, uint16_t port);
>>   int tcp_socket_incoming(const char *address, uint16_t port);
>> @@ -49,7 +51,8 @@ int unix_socket_outgoing(const char *path);
>>   int unix_socket_incoming(const char *path);
>>
>>   int nbd_negotiate(int csock, off_t size);
>> -int nbd_receive_negotiate(int csock, off_t *size, size_t *blocksize);
>> +int nbd_receive_negotiate(int csock, const char *name, uint32_t *flags,
>> +                          off_t *size, size_t *blocksize);
>>   int nbd_init(int fd, int csock, off_t size, size_t blocksize);
>>   int nbd_send_request(int csock, struct nbd_request *request);
>>   int nbd_receive_reply(int csock, struct nbd_reply *reply);
>> diff --git a/qemu-doc.texi b/qemu-doc.texi
>> index 55a966f..d7d760f 100644
>> --- a/qemu-doc.texi
>> +++ b/qemu-doc.texi
>> @@ -620,6 +620,13 @@ qemu linux1.img -hdb nbd:unix:/tmp/my_socket
>>   qemu linux2.img -hdb nbd:unix:/tmp/my_socket
>>   @end example
>>
>> +If the nbd-server uses named exports (since NBD 2.9.18), you must use the
>> +"exportname" option:
>> address@hidden
>> +qemu -cdrom nbd:localhost:exportname=debian-500-ppc-netinst
>> +qemu -cdrom nbd:localhost:exportname=openSUSE-11.1-ppc-netinst
>> address@hidden example
>> +
>>   @node pcsys_network
>>   @section Network emulation
>>
>> diff --git a/qemu-nbd.c b/qemu-nbd.c
>> index 4e607cf..41e5c5d 100644
>> --- a/qemu-nbd.c
>> +++ b/qemu-nbd.c
>> @@ -230,6 +230,7 @@ int main(int argc, char **argv)
>>       int nb_fds = 0;
>>       int max_fd;
>>       int persistent = 0;
>> +    uint32_t nbdflags;
>>
>>       while ((ch = getopt_long(argc, argv, sopt, lopt,&opt_ind)) != -1) {
>>           switch (ch) {
>> @@ -398,7 +399,8 @@ int main(int argc, char **argv)
>>                   goto out;
>>               }
>>
>> -            ret = nbd_receive_negotiate(sock,&size,&blocksize);
>> +            ret = nbd_receive_negotiate(sock, NULL,&nbdflags,
>> +                                    &size,&blocksize);
>>               if (ret == -1) {
>>                   ret = 1;
>>                   goto out;
>>    
>
>



-- 
--------------------- address@hidden  ---------------------
"Tout ce qui est impossible reste à accomplir"    Jules Verne
"Things are only impossible until they're not" Jean-Luc Picard