Jeff, thanks for review. I will try to explain what this code is going to do.
So, as you can understand this patch affects four modes:
1) strict + cacheAll;
2) strict + not cacheAll;
3) no strict + cacheAll;
4) no strict + no cacheAll.
Let's describe every one.
1) strict + cacheAll.
The client goes to cifs_file_aio_write and returns with
generic_file_aio_write. Then it appears in cifs_write_end call and
goes to the section:
+ if (cifs_sb->mnt_cifs_flags & CIFS_MOUNT_STRICT_IO) {
+ rc = copied;
+ pos += copied;
+
+ if (CIFS_I(inode)->clientCanCacheAll) {
+ SetPageUptodate(page);
+ set_page_dirty(page);
+ }
+
+ /* if we don't have an exclusive oplock the page data was
+ previously written to the server in cifs_file_aio_write,
+ so we don't need to do it again - goto exit */
+
+ goto exit;
+ }
+
Here it sets don't send anything to the server (because we can cache
the data) but sets page dirty and uptodate and exits - then this data
will be written to the server through writepages code.
2) strict + not cacheAll;
The client goes to cifs_file_aio_write call and appears in this section:
+ saved_pos = pos;
+ written = cifs_user_write(iocb->ki_filp, iov->iov_base,
+ iov->iov_len, &pos);
+
+ if (written > 0) {
+ cache_written = generic_file_aio_write(iocb, iov,
+ nr_segs, saved_pos);
+ if (cache_written != written)
+ cERROR(1, "Cache written and server written data "
+ "lengths are different");
+ } else
+ iocb->ki_pos = pos;
So, it writes the data to the server in cifs_user_write and if it was
successful (written > 0), call generic_file_aio_write that stores the
data in the filesystem cache and calls cifs_write_end. In
cifs_write_end we again appear in the section
+ if (cifs_sb->mnt_cifs_flags & CIFS_MOUNT_STRICT_IO) {
+ rc = copied;
+ pos += copied;
+
+ if (CIFS_I(inode)->clientCanCacheAll) {
+ SetPageUptodate(page);
+ set_page_dirty(page);
+ }
+
+ /* if we don't have an exclusive oplock the page data was
+ previously written to the server in cifs_file_aio_write,
+ so we don't need to do it again - goto exit */
+
+ goto exit;
+ }
+
but in this case we don't set any flags because the page isn't dirty
(we wrote it to the server and don't need to flush it again).
3) no strict + cacheAll.
The client goes to cifs_file_aio_write and returns with
generic_file_aio_write that stored the data in the filesystem cache
and call cifs_write_end. This scenario is the same as we have in the
upstream code now.
4) no strict + cacheAll.
The client goes to cifs_file_aio_write and appears in the secton:
+ if ((cifs_sb->mnt_cifs_flags & CIFS_MOUNT_STRICT_IO) == 0) {
+ written = generic_file_aio_write(iocb, iov, nr_segs, pos);
+ filemap_write_and_wait(inode->i_mapping);
+ return written;
+ }
Then it call generic_file_aio_write that stores the data in the cache
and calls cifs_write_end. This scenario again is the same as we have
now. So, nothing changes.
So, this is the description of the strict cache mode. I will answer to
your questions in the next email.
--
Best regards,
Pavel Shilovsky.
--
To unsubscribe from this list: send the line "unsubscribe linux-cifs" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
