Merge branch 'for-linus' of git://oss.sgi.com/xfs/xfs
[sfrench/cifs-2.6.git] / fs / cifs / file.c
index d843631c028d5939f57992701e95943318fe14ab..0de17c1db608fc1af7da8b1e43ea1a97572238fe 100644 (file)
@@ -287,6 +287,7 @@ void cifsFileInfo_put(struct cifsFileInfo *cifs_file)
        struct inode *inode = cifs_file->dentry->d_inode;
        struct cifsTconInfo *tcon = tlink_tcon(cifs_file->tlink);
        struct cifsInodeInfo *cifsi = CIFS_I(inode);
+       struct cifs_sb_info *cifs_sb = CIFS_SB(inode->i_sb);
        struct cifsLockInfo *li, *tmp;
 
        spin_lock(&cifs_file_list_lock);
@@ -302,6 +303,13 @@ void cifsFileInfo_put(struct cifsFileInfo *cifs_file)
        if (list_empty(&cifsi->openFileList)) {
                cFYI(1, "closing last open instance for inode %p",
                        cifs_file->dentry->d_inode);
+
+               /* in strict cache mode we need invalidate mapping on the last
+                  close  because it may cause a error when we open this file
+                  again and get at least level II oplock */
+               if (cifs_sb->mnt_cifs_flags & CIFS_MOUNT_STRICT_IO)
+                       CIFS_I(inode)->invalid_mapping = true;
+
                cifs_set_oplock_level(cifsi, 0);
        }
        spin_unlock(&cifs_file_list_lock);
@@ -726,12 +734,12 @@ int cifs_lock(struct file *file, int cmd, struct file_lock *pfLock)
 
                /* BB we could chain these into one lock request BB */
                rc = CIFSSMBLock(xid, tcon, netfid, length, pfLock->fl_start,
-                                0, 1, lockType, 0 /* wait flag */ );
+                                0, 1, lockType, 0 /* wait flag */, 0);
                if (rc == 0) {
                        rc = CIFSSMBLock(xid, tcon, netfid, length,
                                         pfLock->fl_start, 1 /* numUnlock */ ,
                                         0 /* numLock */ , lockType,
-                                        0 /* wait flag */ );
+                                        0 /* wait flag */, 0);
                        pfLock->fl_type = F_UNLCK;
                        if (rc != 0)
                                cERROR(1, "Error unlocking previously locked "
@@ -748,13 +756,13 @@ int cifs_lock(struct file *file, int cmd, struct file_lock *pfLock)
                                rc = CIFSSMBLock(xid, tcon, netfid, length,
                                        pfLock->fl_start, 0, 1,
                                        lockType | LOCKING_ANDX_SHARED_LOCK,
-                                       0 /* wait flag */);
+                                       0 /* wait flag */, 0);
                                if (rc == 0) {
                                        rc = CIFSSMBLock(xid, tcon, netfid,
                                                length, pfLock->fl_start, 1, 0,
                                                lockType |
                                                LOCKING_ANDX_SHARED_LOCK,
-                                               0 /* wait flag */);
+                                               0 /* wait flag */, 0);
                                        pfLock->fl_type = F_RDLCK;
                                        if (rc != 0)
                                                cERROR(1, "Error unlocking "
@@ -797,8 +805,8 @@ int cifs_lock(struct file *file, int cmd, struct file_lock *pfLock)
 
                if (numLock) {
                        rc = CIFSSMBLock(xid, tcon, netfid, length,
-                                       pfLock->fl_start,
-                                       0, numLock, lockType, wait_flag);
+                                        pfLock->fl_start, 0, numLock, lockType,
+                                        wait_flag, 0);
 
                        if (rc == 0) {
                                /* For Windows locks we must store them. */
@@ -818,9 +826,9 @@ int cifs_lock(struct file *file, int cmd, struct file_lock *pfLock)
                                                (pfLock->fl_start + length) >=
                                                (li->offset + li->length)) {
                                        stored_rc = CIFSSMBLock(xid, tcon,
-                                                       netfid,
-                                                       li->length, li->offset,
-                                                       1, 0, li->type, false);
+                                                       netfid, li->length,
+                                                       li->offset, 1, 0,
+                                                       li->type, false, 0);
                                        if (stored_rc)
                                                rc = stored_rc;
                                        else {
@@ -839,31 +847,8 @@ int cifs_lock(struct file *file, int cmd, struct file_lock *pfLock)
        return rc;
 }
 
-/*
- * Set the timeout on write requests past EOF. For some servers (Windows)
- * these calls can be very long.
- *
- * If we're writing >10M past the EOF we give a 180s timeout. Anything less
- * than that gets a 45s timeout. Writes not past EOF get 15s timeouts.
- * The 10M cutoff is totally arbitrary. A better scheme for this would be
- * welcome if someone wants to suggest one.
- *
- * We may be able to do a better job with this if there were some way to
- * declare that a file should be sparse.
- */
-static int
-cifs_write_timeout(struct cifsInodeInfo *cifsi, loff_t offset)
-{
-       if (offset <= cifsi->server_eof)
-               return CIFS_STD_OP;
-       else if (offset > (cifsi->server_eof + (10 * 1024 * 1024)))
-               return CIFS_VLONG_OP;
-       else
-               return CIFS_LONG_OP;
-}
-
 /* update the file size (if needed) after a write */
-static void
+void
 cifs_update_eof(struct cifsInodeInfo *cifsi, loff_t offset,
                      unsigned int bytes_written)
 {
@@ -882,7 +867,7 @@ ssize_t cifs_user_write(struct file *file, const char __user *write_data,
        unsigned int total_written;
        struct cifs_sb_info *cifs_sb;
        struct cifsTconInfo *pTcon;
-       int xid, long_op;
+       int xid;
        struct cifsFileInfo *open_file;
        struct cifsInodeInfo *cifsi = CIFS_I(inode);
 
@@ -903,7 +888,6 @@ ssize_t cifs_user_write(struct file *file, const char __user *write_data,
 
        xid = GetXid();
 
-       long_op = cifs_write_timeout(cifsi, *poffset);
        for (total_written = 0; write_size > total_written;
             total_written += bytes_written) {
                rc = -EAGAIN;
@@ -931,7 +915,7 @@ ssize_t cifs_user_write(struct file *file, const char __user *write_data,
                                min_t(const int, cifs_sb->wsize,
                                      write_size - total_written),
                                *poffset, &bytes_written,
-                               NULL, write_data + total_written, long_op);
+                               NULL, write_data + total_written, 0);
                }
                if (rc || (bytes_written == 0)) {
                        if (total_written)
@@ -944,8 +928,6 @@ ssize_t cifs_user_write(struct file *file, const char __user *write_data,
                        cifs_update_eof(cifsi, *poffset, bytes_written);
                        *poffset += bytes_written;
                }
-               long_op = CIFS_STD_OP; /* subsequent writes fast -
-                                   15 seconds is plenty */
        }
 
        cifs_stats_bytes_written(pTcon, total_written);
@@ -974,7 +956,7 @@ static ssize_t cifs_write(struct cifsFileInfo *open_file,
        unsigned int total_written;
        struct cifs_sb_info *cifs_sb;
        struct cifsTconInfo *pTcon;
-       int xid, long_op;
+       int xid;
        struct dentry *dentry = open_file->dentry;
        struct cifsInodeInfo *cifsi = CIFS_I(dentry->d_inode);
 
@@ -987,7 +969,6 @@ static ssize_t cifs_write(struct cifsFileInfo *open_file,
 
        xid = GetXid();
 
-       long_op = cifs_write_timeout(cifsi, *poffset);
        for (total_written = 0; write_size > total_written;
             total_written += bytes_written) {
                rc = -EAGAIN;
@@ -1017,7 +998,7 @@ static ssize_t cifs_write(struct cifsFileInfo *open_file,
                                rc = CIFSSMBWrite2(xid, pTcon,
                                                open_file->netfid, len,
                                                *poffset, &bytes_written,
-                                               iov, 1, long_op);
+                                               iov, 1, 0);
                        } else
                                rc = CIFSSMBWrite(xid, pTcon,
                                         open_file->netfid,
@@ -1025,7 +1006,7 @@ static ssize_t cifs_write(struct cifsFileInfo *open_file,
                                               write_size - total_written),
                                         *poffset, &bytes_written,
                                         write_data + total_written,
-                                        NULL, long_op);
+                                        NULL, 0);
                }
                if (rc || (bytes_written == 0)) {
                        if (total_written)
@@ -1038,8 +1019,6 @@ static ssize_t cifs_write(struct cifsFileInfo *open_file,
                        cifs_update_eof(cifsi, *poffset, bytes_written);
                        *poffset += bytes_written;
                }
-               long_op = CIFS_STD_OP; /* subsequent writes fast -
-                                   15 seconds is plenty */
        }
 
        cifs_stats_bytes_written(pTcon, total_written);
@@ -1239,7 +1218,7 @@ static int cifs_writepages(struct address_space *mapping,
        struct pagevec pvec;
        int rc = 0;
        int scanned = 0;
-       int xid, long_op;
+       int xid;
 
        cifs_sb = CIFS_SB(mapping->host->i_sb);
 
@@ -1377,43 +1356,67 @@ retry:
                                break;
                }
                if (n_iov) {
+retry_write:
                        open_file = find_writable_file(CIFS_I(mapping->host),
                                                        false);
                        if (!open_file) {
                                cERROR(1, "No writable handles for inode");
                                rc = -EBADF;
                        } else {
-                               long_op = cifs_write_timeout(cifsi, offset);
                                rc = CIFSSMBWrite2(xid, tcon, open_file->netfid,
                                                   bytes_to_write, offset,
                                                   &bytes_written, iov, n_iov,
-                                                  long_op);
+                                                  0);
                                cifsFileInfo_put(open_file);
-                               cifs_update_eof(cifsi, offset, bytes_written);
                        }
 
-                       if (rc || bytes_written < bytes_to_write) {
-                               cERROR(1, "Write2 ret %d, wrote %d",
-                                         rc, bytes_written);
-                               mapping_set_error(mapping, rc);
-                       } else {
+                       cFYI(1, "Write2 rc=%d, wrote=%u", rc, bytes_written);
+
+                       /*
+                        * For now, treat a short write as if nothing got
+                        * written. A zero length write however indicates
+                        * ENOSPC or EFBIG. We have no way to know which
+                        * though, so call it ENOSPC for now. EFBIG would
+                        * get translated to AS_EIO anyway.
+                        *
+                        * FIXME: make it take into account the data that did
+                        *        get written
+                        */
+                       if (rc == 0) {
+                               if (bytes_written == 0)
+                                       rc = -ENOSPC;
+                               else if (bytes_written < bytes_to_write)
+                                       rc = -EAGAIN;
+                       }
+
+                       /* retry on data-integrity flush */
+                       if (wbc->sync_mode == WB_SYNC_ALL && rc == -EAGAIN)
+                               goto retry_write;
+
+                       /* fix the stats and EOF */
+                       if (bytes_written > 0) {
                                cifs_stats_bytes_written(tcon, bytes_written);
+                               cifs_update_eof(cifsi, offset, bytes_written);
                        }
 
                        for (i = 0; i < n_iov; i++) {
                                page = pvec.pages[first + i];
-                               /* Should we also set page error on
-                               success rc but too little data written? */
-                               /* BB investigate retry logic on temporary
-                               server crash cases and how recovery works
-                               when page marked as error */
-                               if (rc)
+                               /* on retryable write error, redirty page */
+                               if (rc == -EAGAIN)
+                                       redirty_page_for_writepage(wbc, page);
+                               else if (rc != 0)
                                        SetPageError(page);
                                kunmap(page);
                                unlock_page(page);
                                end_page_writeback(page);
                                page_cache_release(page);
                        }
+
+                       if (rc != -EAGAIN)
+                               mapping_set_error(mapping, rc);
+                       else
+                               rc = 0;
+
                        if ((wbc->nr_to_write -= n_iov) <= 0)
                                done = 1;
                        index = next;
@@ -1525,27 +1528,47 @@ static int cifs_write_end(struct file *file, struct address_space *mapping,
        return rc;
 }
 
-int cifs_fsync(struct file *file, int datasync)
+int cifs_strict_fsync(struct file *file, int datasync)
 {
        int xid;
        int rc = 0;
        struct cifsTconInfo *tcon;
        struct cifsFileInfo *smbfile = file->private_data;
        struct inode *inode = file->f_path.dentry->d_inode;
+       struct cifs_sb_info *cifs_sb = CIFS_SB(inode->i_sb);
 
        xid = GetXid();
 
        cFYI(1, "Sync file - name: %s datasync: 0x%x",
                file->f_path.dentry->d_name.name, datasync);
 
-       rc = filemap_write_and_wait(inode->i_mapping);
-       if (rc == 0) {
-               struct cifs_sb_info *cifs_sb = CIFS_SB(inode->i_sb);
+       if (!CIFS_I(inode)->clientCanCacheRead)
+               cifs_invalidate_mapping(inode);
 
-               tcon = tlink_tcon(smbfile->tlink);
-               if (!(cifs_sb->mnt_cifs_flags & CIFS_MOUNT_NOSSYNC))
-                       rc = CIFSSMBFlush(xid, tcon, smbfile->netfid);
-       }
+       tcon = tlink_tcon(smbfile->tlink);
+       if (!(cifs_sb->mnt_cifs_flags & CIFS_MOUNT_NOSSYNC))
+               rc = CIFSSMBFlush(xid, tcon, smbfile->netfid);
+
+       FreeXid(xid);
+       return rc;
+}
+
+int cifs_fsync(struct file *file, int datasync)
+{
+       int xid;
+       int rc = 0;
+       struct cifsTconInfo *tcon;
+       struct cifsFileInfo *smbfile = file->private_data;
+       struct cifs_sb_info *cifs_sb = CIFS_SB(file->f_path.dentry->d_sb);
+
+       xid = GetXid();
+
+       cFYI(1, "Sync file - name: %s datasync: 0x%x",
+               file->f_path.dentry->d_name.name, datasync);
+
+       tcon = tlink_tcon(smbfile->tlink);
+       if (!(cifs_sb->mnt_cifs_flags & CIFS_MOUNT_NOSSYNC))
+               rc = CIFSSMBFlush(xid, tcon, smbfile->netfid);
 
        FreeXid(xid);
        return rc;
@@ -1596,42 +1619,242 @@ int cifs_flush(struct file *file, fl_owner_t id)
        return rc;
 }
 
-ssize_t cifs_user_read(struct file *file, char __user *read_data,
-       size_t read_size, loff_t *poffset)
+static int
+cifs_write_allocate_pages(struct page **pages, unsigned long num_pages)
 {
-       int rc = -EACCES;
-       unsigned int bytes_read = 0;
-       unsigned int total_read = 0;
-       unsigned int current_read_size;
-       struct cifs_sb_info *cifs_sb;
+       int rc = 0;
+       unsigned long i;
+
+       for (i = 0; i < num_pages; i++) {
+               pages[i] = alloc_page(__GFP_HIGHMEM);
+               if (!pages[i]) {
+                       /*
+                        * save number of pages we have already allocated and
+                        * return with ENOMEM error
+                        */
+                       num_pages = i;
+                       rc = -ENOMEM;
+                       goto error;
+               }
+       }
+
+       return rc;
+
+error:
+       for (i = 0; i < num_pages; i++)
+               put_page(pages[i]);
+       return rc;
+}
+
+static inline
+size_t get_numpages(const size_t wsize, const size_t len, size_t *cur_len)
+{
+       size_t num_pages;
+       size_t clen;
+
+       clen = min_t(const size_t, len, wsize);
+       num_pages = clen / PAGE_CACHE_SIZE;
+       if (clen % PAGE_CACHE_SIZE)
+               num_pages++;
+
+       if (cur_len)
+               *cur_len = clen;
+
+       return num_pages;
+}
+
+static ssize_t
+cifs_iovec_write(struct file *file, const struct iovec *iov,
+                unsigned long nr_segs, loff_t *poffset)
+{
+       size_t total_written = 0, written = 0;
+       unsigned long num_pages, npages;
+       size_t copied, len, cur_len, i;
+       struct kvec *to_send;
+       struct page **pages;
+       struct iov_iter it;
+       struct inode *inode;
+       struct cifsFileInfo *open_file;
        struct cifsTconInfo *pTcon;
+       struct cifs_sb_info *cifs_sb;
+       int xid, rc;
+
+       len = iov_length(iov, nr_segs);
+       if (!len)
+               return 0;
+
+       rc = generic_write_checks(file, poffset, &len, 0);
+       if (rc)
+               return rc;
+
+       cifs_sb = CIFS_SB(file->f_path.dentry->d_sb);
+       num_pages = get_numpages(cifs_sb->wsize, len, &cur_len);
+
+       pages = kmalloc(sizeof(struct pages *)*num_pages, GFP_KERNEL);
+       if (!pages)
+               return -ENOMEM;
+
+       to_send = kmalloc(sizeof(struct kvec)*(num_pages + 1), GFP_KERNEL);
+       if (!to_send) {
+               kfree(pages);
+               return -ENOMEM;
+       }
+
+       rc = cifs_write_allocate_pages(pages, num_pages);
+       if (rc) {
+               kfree(pages);
+               kfree(to_send);
+               return rc;
+       }
+
+       xid = GetXid();
+       open_file = file->private_data;
+       pTcon = tlink_tcon(open_file->tlink);
+       inode = file->f_path.dentry->d_inode;
+
+       iov_iter_init(&it, iov, nr_segs, len, 0);
+       npages = num_pages;
+
+       do {
+               size_t save_len = cur_len;
+               for (i = 0; i < npages; i++) {
+                       copied = min_t(const size_t, cur_len, PAGE_CACHE_SIZE);
+                       copied = iov_iter_copy_from_user(pages[i], &it, 0,
+                                                        copied);
+                       cur_len -= copied;
+                       iov_iter_advance(&it, copied);
+                       to_send[i+1].iov_base = kmap(pages[i]);
+                       to_send[i+1].iov_len = copied;
+               }
+
+               cur_len = save_len - cur_len;
+
+               do {
+                       if (open_file->invalidHandle) {
+                               rc = cifs_reopen_file(open_file, false);
+                               if (rc != 0)
+                                       break;
+                       }
+                       rc = CIFSSMBWrite2(xid, pTcon, open_file->netfid,
+                                          cur_len, *poffset, &written,
+                                          to_send, npages, 0);
+               } while (rc == -EAGAIN);
+
+               for (i = 0; i < npages; i++)
+                       kunmap(pages[i]);
+
+               if (written) {
+                       len -= written;
+                       total_written += written;
+                       cifs_update_eof(CIFS_I(inode), *poffset, written);
+                       *poffset += written;
+               } else if (rc < 0) {
+                       if (!total_written)
+                               total_written = rc;
+                       break;
+               }
+
+               /* get length and number of kvecs of the next write */
+               npages = get_numpages(cifs_sb->wsize, len, &cur_len);
+       } while (len > 0);
+
+       if (total_written > 0) {
+               spin_lock(&inode->i_lock);
+               if (*poffset > inode->i_size)
+                       i_size_write(inode, *poffset);
+               spin_unlock(&inode->i_lock);
+       }
+
+       cifs_stats_bytes_written(pTcon, total_written);
+       mark_inode_dirty_sync(inode);
+
+       for (i = 0; i < num_pages; i++)
+               put_page(pages[i]);
+       kfree(to_send);
+       kfree(pages);
+       FreeXid(xid);
+       return total_written;
+}
+
+static ssize_t cifs_user_writev(struct kiocb *iocb, const struct iovec *iov,
+                               unsigned long nr_segs, loff_t pos)
+{
+       ssize_t written;
+       struct inode *inode;
+
+       inode = iocb->ki_filp->f_path.dentry->d_inode;
+
+       /*
+        * BB - optimize the way when signing is disabled. We can drop this
+        * extra memory-to-memory copying and use iovec buffers for constructing
+        * write request.
+        */
+
+       written = cifs_iovec_write(iocb->ki_filp, iov, nr_segs, &pos);
+       if (written > 0) {
+               CIFS_I(inode)->invalid_mapping = true;
+               iocb->ki_pos = pos;
+       }
+
+       return written;
+}
+
+ssize_t cifs_strict_writev(struct kiocb *iocb, const struct iovec *iov,
+                          unsigned long nr_segs, loff_t pos)
+{
+       struct inode *inode;
+
+       inode = iocb->ki_filp->f_path.dentry->d_inode;
+
+       if (CIFS_I(inode)->clientCanCacheAll)
+               return generic_file_aio_write(iocb, iov, nr_segs, pos);
+
+       /*
+        * In strict cache mode we need to write the data to the server exactly
+        * from the pos to pos+len-1 rather than flush all affected pages
+        * because it may cause a error with mandatory locks on these pages but
+        * not on the region from pos to ppos+len-1.
+        */
+
+       return cifs_user_writev(iocb, iov, nr_segs, pos);
+}
+
+static ssize_t
+cifs_iovec_read(struct file *file, const struct iovec *iov,
+                unsigned long nr_segs, loff_t *poffset)
+{
+       int rc;
        int xid;
+       unsigned int total_read, bytes_read = 0;
+       size_t len, cur_len;
+       int iov_offset = 0;
+       struct cifs_sb_info *cifs_sb;
+       struct cifsTconInfo *pTcon;
        struct cifsFileInfo *open_file;
-       char *smb_read_data;
-       char __user *current_offset;
        struct smb_com_read_rsp *pSMBr;
+       char *read_data;
+
+       if (!nr_segs)
+               return 0;
+
+       len = iov_length(iov, nr_segs);
+       if (!len)
+               return 0;
 
        xid = GetXid();
        cifs_sb = CIFS_SB(file->f_path.dentry->d_sb);
 
-       if (file->private_data == NULL) {
-               rc = -EBADF;
-               FreeXid(xid);
-               return rc;
-       }
        open_file = file->private_data;
        pTcon = tlink_tcon(open_file->tlink);
 
        if ((file->f_flags & O_ACCMODE) == O_WRONLY)
                cFYI(1, "attempting read on write only file instance");
 
-       for (total_read = 0, current_offset = read_data;
-            read_size > total_read;
-            total_read += bytes_read, current_offset += bytes_read) {
-               current_read_size = min_t(const int, read_size - total_read,
-                                         cifs_sb->rsize);
+       for (total_read = 0; total_read < len; total_read += bytes_read) {
+               cur_len = min_t(const size_t, len - total_read, cifs_sb->rsize);
                rc = -EAGAIN;
-               smb_read_data = NULL;
+               read_data = NULL;
+
                while (rc == -EAGAIN) {
                        int buf_type = CIFS_NO_BUFFER;
                        if (open_file->invalidHandle) {
@@ -1639,27 +1862,25 @@ ssize_t cifs_user_read(struct file *file, char __user *read_data,
                                if (rc != 0)
                                        break;
                        }
-                       rc = CIFSSMBRead(xid, pTcon,
-                                        open_file->netfid,
-                                        current_read_size, *poffset,
-                                        &bytes_read, &smb_read_data,
-                                        &buf_type);
-                       pSMBr = (struct smb_com_read_rsp *)smb_read_data;
-                       if (smb_read_data) {
-                               if (copy_to_user(current_offset,
-                                               smb_read_data +
-                                               4 /* RFC1001 length field */ +
-                                               le16_to_cpu(pSMBr->DataOffset),
-                                               bytes_read))
+                       rc = CIFSSMBRead(xid, pTcon, open_file->netfid,
+                                        cur_len, *poffset, &bytes_read,
+                                        &read_data, &buf_type);
+                       pSMBr = (struct smb_com_read_rsp *)read_data;
+                       if (read_data) {
+                               char *data_offset = read_data + 4 +
+                                               le16_to_cpu(pSMBr->DataOffset);
+                               if (memcpy_toiovecend(iov, data_offset,
+                                                     iov_offset, bytes_read))
                                        rc = -EFAULT;
-
                                if (buf_type == CIFS_SMALL_BUFFER)
-                                       cifs_small_buf_release(smb_read_data);
+                                       cifs_small_buf_release(read_data);
                                else if (buf_type == CIFS_LARGE_BUFFER)
-                                       cifs_buf_release(smb_read_data);
-                               smb_read_data = NULL;
+                                       cifs_buf_release(read_data);
+                               read_data = NULL;
+                               iov_offset += bytes_read;
                        }
                }
+
                if (rc || (bytes_read == 0)) {
                        if (total_read) {
                                break;
@@ -1672,13 +1893,57 @@ ssize_t cifs_user_read(struct file *file, char __user *read_data,
                        *poffset += bytes_read;
                }
        }
+
        FreeXid(xid);
        return total_read;
 }
 
+ssize_t cifs_user_read(struct file *file, char __user *read_data,
+                      size_t read_size, loff_t *poffset)
+{
+       struct iovec iov;
+       iov.iov_base = read_data;
+       iov.iov_len = read_size;
+
+       return cifs_iovec_read(file, &iov, 1, poffset);
+}
+
+static ssize_t cifs_user_readv(struct kiocb *iocb, const struct iovec *iov,
+                              unsigned long nr_segs, loff_t pos)
+{
+       ssize_t read;
+
+       read = cifs_iovec_read(iocb->ki_filp, iov, nr_segs, &pos);
+       if (read > 0)
+               iocb->ki_pos = pos;
+
+       return read;
+}
+
+ssize_t cifs_strict_readv(struct kiocb *iocb, const struct iovec *iov,
+                         unsigned long nr_segs, loff_t pos)
+{
+       struct inode *inode;
+
+       inode = iocb->ki_filp->f_path.dentry->d_inode;
+
+       if (CIFS_I(inode)->clientCanCacheRead)
+               return generic_file_aio_read(iocb, iov, nr_segs, pos);
+
+       /*
+        * In strict cache mode we need to read from the server all the time
+        * if we don't have level II oplock because the server can delay mtime
+        * change - so we can't make a decision about inode invalidating.
+        * And we can also fail with pagereading if there are mandatory locks
+        * on pages affected by this read but not on the region from pos to
+        * pos+len-1.
+        */
+
+       return cifs_user_readv(iocb, iov, nr_segs, pos);
+}
 
 static ssize_t cifs_read(struct file *file, char *read_data, size_t read_size,
-       loff_t *poffset)
+                        loff_t *poffset)
 {
        int rc = -EACCES;
        unsigned int bytes_read = 0;
@@ -1746,6 +2011,21 @@ static ssize_t cifs_read(struct file *file, char *read_data, size_t read_size,
        return total_read;
 }
 
+int cifs_file_strict_mmap(struct file *file, struct vm_area_struct *vma)
+{
+       int rc, xid;
+       struct inode *inode = file->f_path.dentry->d_inode;
+
+       xid = GetXid();
+
+       if (!CIFS_I(inode)->clientCanCacheRead)
+               cifs_invalidate_mapping(inode);
+
+       rc = generic_file_mmap(file, vma);
+       FreeXid(xid);
+       return rc;
+}
+
 int cifs_file_mmap(struct file *file, struct vm_area_struct *vma)
 {
        int rc, xid;
@@ -2192,7 +2472,8 @@ void cifs_oplock_break(struct work_struct *work)
         */
        if (!cfile->oplock_break_cancelled) {
                rc = CIFSSMBLock(0, tlink_tcon(cfile->tlink), cfile->netfid, 0,
-                                0, 0, 0, LOCKING_ANDX_OPLOCK_RELEASE, false);
+                                0, 0, 0, LOCKING_ANDX_OPLOCK_RELEASE, false,
+                                cinode->clientCanCacheRead ? 1 : 0);
                cFYI(1, "Oplock release rc = %d", rc);
        }