diff mbox series

[08/14] nbd/server: Initial support for extended headers

Message ID 20211203231539.3900865-9-eblake@redhat.com (mailing list archive)
State New, archived
Headers show
Series qemu patches for NBD_OPT_EXTENDED_HEADERS | expand

Commit Message

Eric Blake Dec. 3, 2021, 11:15 p.m. UTC
We have no reason to send NBD_REPLY_TYPE_OFFSET_HOLE_EXT since we
already cap NBD_CMD_READ to 32M.  For NBD_CMD_WRITE_ZEROES and
NBD_CMD_TRIM, the block layer already supports 64-bit operations
without any effort on our part.  For NBD_CMD_BLOCK_STATUS, the
client's length is a hint; the easiest approach is to truncate our
answer back to 32 bits, letting us delay the effort of implementing
NBD_REPLY_TYPE_BLOCK_STATUS_EXT to a separate patch.

Signed-off-by: Eric Blake <eblake@redhat.com>
---
 nbd/nbd-internal.h |   5 ++-
 nbd/server.c       | 106 ++++++++++++++++++++++++++++++++++-----------
 2 files changed, 85 insertions(+), 26 deletions(-)
diff mbox series

Patch

diff --git a/nbd/nbd-internal.h b/nbd/nbd-internal.h
index 0016793ff4b1..875b6204c28c 100644
--- a/nbd/nbd-internal.h
+++ b/nbd/nbd-internal.h
@@ -35,8 +35,11 @@ 
  * https://github.com/yoe/nbd/blob/master/doc/proto.md
  */

-/* Size of all NBD_OPT_*, without payload */
+/* Size of all compact NBD_CMD_*, without payload */
 #define NBD_REQUEST_SIZE            (4 + 2 + 2 + 8 + 8 + 4)
+/* Size of all extended NBD_CMD_*, without payload */
+#define NBD_REQUEST_EXT_SIZE        (4 + 2 + 2 + 8 + 8 + 8)
+
 /* Size of all NBD_REP_* sent in answer to most NBD_OPT_*, without payload */
 #define NBD_REPLY_SIZE              (4 + 4 + 8)
 /* Size of reply to NBD_OPT_EXPORT_NAME */
diff --git a/nbd/server.c b/nbd/server.c
index 4306fa7b426c..0e496f60ffbd 100644
--- a/nbd/server.c
+++ b/nbd/server.c
@@ -142,6 +142,7 @@  struct NBDClient {
     uint32_t check_align; /* If non-zero, check for aligned client requests */

     bool structured_reply;
+    bool extended_headers;
     NBDExportMetaContexts export_meta;

     uint32_t opt; /* Current option being negotiated */
@@ -1275,6 +1276,19 @@  static int nbd_negotiate_options(NBDClient *client, Error **errp)
                                                  errp);
                 break;

+            case NBD_OPT_EXTENDED_HEADERS:
+                if (length) {
+                    ret = nbd_reject_length(client, false, errp);
+                } else if (client->extended_headers) {
+                    ret = nbd_negotiate_send_rep_err(
+                        client, NBD_REP_ERR_INVALID, errp,
+                        "extended headers already negotiated");
+                } else {
+                    ret = nbd_negotiate_send_rep(client, NBD_REP_ACK, errp);
+                    client->extended_headers = true;
+                }
+                break;
+
             default:
                 ret = nbd_opt_drop(client, NBD_REP_ERR_UNSUP, errp,
                                    "Unsupported option %" PRIu32 " (%s)",
@@ -1410,11 +1424,13 @@  nbd_read_eof(NBDClient *client, void *buffer, size_t size, Error **errp)
 static int nbd_receive_request(NBDClient *client, NBDRequest *request,
                                Error **errp)
 {
-    uint8_t buf[NBD_REQUEST_SIZE];
-    uint32_t magic;
+    uint8_t buf[NBD_REQUEST_EXT_SIZE];
+    uint32_t magic, expect;
     int ret;
+    size_t size = client->extended_headers ? NBD_REQUEST_EXT_SIZE
+        : NBD_REQUEST_SIZE;

-    ret = nbd_read_eof(client, buf, sizeof(buf), errp);
+    ret = nbd_read_eof(client, buf, size, errp);
     if (ret < 0) {
         return ret;
     }
@@ -1422,13 +1438,21 @@  static int nbd_receive_request(NBDClient *client, NBDRequest *request,
         return -EIO;
     }

-    /* Request
-       [ 0 ..  3]   magic   (NBD_REQUEST_MAGIC)
-       [ 4 ..  5]   flags   (NBD_CMD_FLAG_FUA, ...)
-       [ 6 ..  7]   type    (NBD_CMD_READ, ...)
-       [ 8 .. 15]   handle
-       [16 .. 23]   from
-       [24 .. 27]   len
+    /*
+     * Compact request
+     *  [ 0 ..  3]   magic   (NBD_REQUEST_MAGIC)
+     *  [ 4 ..  5]   flags   (NBD_CMD_FLAG_FUA, ...)
+     *  [ 6 ..  7]   type    (NBD_CMD_READ, ...)
+     *  [ 8 .. 15]   handle
+     *  [16 .. 23]   from
+     *  [24 .. 27]   len
+     * Extended request
+     *  [ 0 ..  3]   magic   (NBD_REQUEST_EXT_MAGIC)
+     *  [ 4 ..  5]   flags   (NBD_CMD_FLAG_FUA, ...)
+     *  [ 6 ..  7]   type    (NBD_CMD_READ, ...)
+     *  [ 8 .. 15]   handle
+     *  [16 .. 23]   from
+     *  [24 .. 31]   len
      */

     magic = ldl_be_p(buf);
@@ -1436,12 +1460,18 @@  static int nbd_receive_request(NBDClient *client, NBDRequest *request,
     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); /* widen 32 to 64 bits */
+    if (client->extended_headers) {
+        request->len = ldq_be_p(buf + 24);
+        expect = NBD_REQUEST_EXT_MAGIC;
+    } else {
+        request->len = ldl_be_p(buf + 24); /* widen 32 to 64 bits */
+        expect = NBD_REQUEST_MAGIC;
+    }

     trace_nbd_receive_request(magic, request->flags, request->type,
                               request->from, request->len);

-    if (magic != NBD_REQUEST_MAGIC) {
+    if (magic != expect) {
         error_setg(errp, "invalid magic (got 0x%" PRIx32 ")", magic);
         return -EINVAL;
     }
@@ -1872,12 +1902,22 @@  static int coroutine_fn nbd_co_send_iov(NBDClient *client, struct iovec *iov,
 static inline void set_be_simple_reply(NBDClient *client, struct iovec *iov,
                                        uint64_t error, uint64_t handle)
 {
-    NBDSimpleReply *reply = iov->iov_base;
+    if (client->extended_headers) {
+        NBDSimpleReplyExt *reply = iov->iov_base;

-    iov->iov_len = sizeof(*reply);
-    stl_be_p(&reply->magic, NBD_SIMPLE_REPLY_MAGIC);
-    stl_be_p(&reply->error, error);
-    stq_be_p(&reply->handle, handle);
+        iov->iov_len = sizeof(*reply);
+        stl_be_p(&reply->magic, NBD_SIMPLE_REPLY_EXT_MAGIC);
+        stl_be_p(&reply->error, error);
+        stq_be_p(&reply->handle, handle);
+        reply->_pad1 = reply->_pad2 = 0;
+    } else {
+        NBDSimpleReply *reply = iov->iov_base;
+
+        iov->iov_len = sizeof(*reply);
+        stl_be_p(&reply->magic, NBD_SIMPLE_REPLY_MAGIC);
+        stl_be_p(&reply->error, error);
+        stq_be_p(&reply->handle, handle);
+    }
 }

 static int nbd_co_send_simple_reply(NBDClient *client,
@@ -1905,14 +1945,26 @@  static inline void set_be_chunk(NBDClient *client, struct iovec *iov,
                                 uint16_t flags, uint16_t type,
                                 uint64_t handle, uint32_t length)
 {
-    NBDStructuredReplyChunk *chunk = iov->iov_base;
+    if (client->extended_headers) {
+        NBDStructuredReplyChunkExt *chunk = iov->iov_base;

-    iov->iov_len = sizeof(*chunk);
-    stl_be_p(&chunk->magic, NBD_STRUCTURED_REPLY_MAGIC);
-    stw_be_p(&chunk->flags, flags);
-    stw_be_p(&chunk->type, type);
-    stq_be_p(&chunk->handle, handle);
-    stl_be_p(&chunk->length, length);
+        iov->iov_len = sizeof(*chunk);
+        stl_be_p(&chunk->magic, NBD_STRUCTURED_REPLY_EXT_MAGIC);
+        stw_be_p(&chunk->flags, flags);
+        stw_be_p(&chunk->type, type);
+        stq_be_p(&chunk->handle, handle);
+        stq_be_p(&chunk->length, length);
+        chunk->_pad = 0;
+    } else {
+        NBDStructuredReplyChunk *chunk = iov->iov_base;
+
+        iov->iov_len = sizeof(*chunk);
+        stl_be_p(&chunk->magic, NBD_STRUCTURED_REPLY_MAGIC);
+        stw_be_p(&chunk->flags, flags);
+        stw_be_p(&chunk->type, type);
+        stq_be_p(&chunk->handle, handle);
+        stl_be_p(&chunk->length, length);
+    }
 }

 static int coroutine_fn nbd_co_send_structured_done(NBDClient *client,
@@ -2555,7 +2607,11 @@  static coroutine_fn int nbd_handle_request(NBDClient *client,
             return nbd_send_generic_reply(client, request->handle, -EINVAL,
                                           "need non-zero length", errp);
         }
-        assert(request->len <= UINT32_MAX);
+        if (request->len > UINT32_MAX) {
+            /* For now, truncate our response to a 32 bit window */
+            request->len = QEMU_ALIGN_DOWN(BDRV_REQUEST_MAX_BYTES,
+                                           client->check_align ?: 1);
+        }
         if (client->export_meta.count) {
             bool dont_fragment = request->flags & NBD_CMD_FLAG_REQ_ONE;
             int contexts_remaining = client->export_meta.count;