summaryrefslogtreecommitdiff
path: root/fs/cifs/cifsglob.h
diff options
context:
space:
mode:
Diffstat (limited to 'fs/cifs/cifsglob.h')
-rw-r--r--fs/cifs/cifsglob.h61
1 files changed, 61 insertions, 0 deletions
diff --git a/fs/cifs/cifsglob.h b/fs/cifs/cifsglob.h
index abb83101903..e5cb1941e25 100644
--- a/fs/cifs/cifsglob.h
+++ b/fs/cifs/cifsglob.h
@@ -213,6 +213,10 @@ struct smb_version_operations {
bool (*need_neg)(struct TCP_Server_Info *);
/* negotiate to the server */
int (*negotiate)(const unsigned int, struct cifs_ses *);
+ /* set negotiated write size */
+ unsigned int (*negotiate_wsize)(struct cifs_tcon *, struct smb_vol *);
+ /* set negotiated read size */
+ unsigned int (*negotiate_rsize)(struct cifs_tcon *, struct smb_vol *);
/* setup smb sessionn */
int (*sess_setup)(const unsigned int, struct cifs_ses *,
const struct nls_table *);
@@ -516,6 +520,63 @@ get_next_mid(struct TCP_Server_Info *server)
}
/*
+ * When the server supports very large reads and writes via POSIX extensions,
+ * we can allow up to 2^24-1, minus the size of a READ/WRITE_AND_X header, not
+ * including the RFC1001 length.
+ *
+ * Note that this might make for "interesting" allocation problems during
+ * writeback however as we have to allocate an array of pointers for the
+ * pages. A 16M write means ~32kb page array with PAGE_CACHE_SIZE == 4096.
+ *
+ * For reads, there is a similar problem as we need to allocate an array
+ * of kvecs to handle the receive, though that should only need to be done
+ * once.
+ */
+#define CIFS_MAX_WSIZE ((1<<24) - 1 - sizeof(WRITE_REQ) + 4)
+#define CIFS_MAX_RSIZE ((1<<24) - sizeof(READ_RSP) + 4)
+
+/*
+ * When the server doesn't allow large posix writes, only allow a rsize/wsize
+ * of 2^17-1 minus the size of the call header. That allows for a read or
+ * write up to the maximum size described by RFC1002.
+ */
+#define CIFS_MAX_RFC1002_WSIZE ((1<<17) - 1 - sizeof(WRITE_REQ) + 4)
+#define CIFS_MAX_RFC1002_RSIZE ((1<<17) - 1 - sizeof(READ_RSP) + 4)
+
+/*
+ * The default wsize is 1M. find_get_pages seems to return a maximum of 256
+ * pages in a single call. With PAGE_CACHE_SIZE == 4k, this means we can fill
+ * a single wsize request with a single call.
+ */
+#define CIFS_DEFAULT_IOSIZE (1024 * 1024)
+
+/*
+ * Windows only supports a max of 60kb reads and 65535 byte writes. Default to
+ * those values when posix extensions aren't in force. In actuality here, we
+ * use 65536 to allow for a write that is a multiple of 4k. Most servers seem
+ * to be ok with the extra byte even though Windows doesn't send writes that
+ * are that large.
+ *
+ * Citation:
+ *
+ * http://blogs.msdn.com/b/openspecification/archive/2009/04/10/smb-maximum-transmit-buffer-size-and-performance-tuning.aspx
+ */
+#define CIFS_DEFAULT_NON_POSIX_RSIZE (60 * 1024)
+#define CIFS_DEFAULT_NON_POSIX_WSIZE (65536)
+
+/*
+ * On hosts with high memory, we can't currently support wsize/rsize that are
+ * larger than we can kmap at once. Cap the rsize/wsize at
+ * LAST_PKMAP * PAGE_SIZE. We'll never be able to fill a read or write request
+ * larger than that anyway.
+ */
+#ifdef CONFIG_HIGHMEM
+#define CIFS_KMAP_SIZE_LIMIT (LAST_PKMAP * PAGE_CACHE_SIZE)
+#else /* CONFIG_HIGHMEM */
+#define CIFS_KMAP_SIZE_LIMIT (1<<24)
+#endif /* CONFIG_HIGHMEM */
+
+/*
* Macros to allow the TCP_Server_Info->net field and related code to drop out
* when CONFIG_NET_NS isn't set.
*/