linux.git/fs/afs/write.c, branch v4.14

fs: convert a pile of fsync routines to errseq_t based reporting

2017-08-01T12:39:29+00:00

This patch converts most of the in-kernel filesystems that do writeback
out of the pagecache to report errors using the errseq_t-based
infrastructure that was recently added. This allows them to report
errors once for each open file description.

Most filesystems have a fairly straightforward fsync operation. They
call filemap_write_and_wait_range to write back all of the data and
wait on it, and then (sometimes) sync out the metadata.

For those filesystems this is a straightforward conversion from calling
filemap_write_and_wait_range in their fsync operation to calling
file_write_and_wait_range.

Acked-by: Jan Kara 
Acked-by: Dave Kleikamp 
Signed-off-by: Jeff Layton

afs: Don't wait for page writeback with the page lock held

2017-03-16T16:29:30+00:00

Drop the page lock before waiting for page writeback.

Signed-off-by: David Howells

afs: ->writepage() shouldn't call clear_page_dirty_for_io()

2017-03-16T16:29:30+00:00

The ->writepage() op shouldn't call clear_page_dirty_for_io() as that has
already been called by the caller.

Fix afs_writepage() by moving the call out of
afs_write_back_from_locked_page() to afs_writepages_region() where it is
needed.

Signed-off-by: David Howells

afs: Fix afs_kill_pages()

2017-03-16T16:29:30+00:00

Fix afs_kill_pages() in two ways:

 (1) If a writeback has been partially flushed, then if we try and kill the
     pages it contains, some of them may no longer be undergoing writeback
     and end_page_writeback() will assert.

     Fix this by checking to see whether the page in question is actually
     undergoing writeback before ending that writeback.

 (2) The loop that scans for pages to kill doesn't increase the first page
     index, and so the loop may not terminate, but it will try to process
     the same pages over and over again.

     Fix this by increasing the first page index to one after the last page
     we processed.

Signed-off-by: David Howells

afs: Fix page leak in afs_write_begin()

2017-03-16T16:27:48+00:00

afs_write_begin() leaks a ref and a lock on a page if afs_fill_page()
fails.  Fix the leak by unlocking and releasing the page in the error path.

Signed-off-by: David Howells

afs: Flush outstanding writes when an fd is closed

2017-03-16T16:27:45+00:00

Flush outstanding writes in afs when an fd is closed.  This is what NFS and
CIFS do.

Reported-by: Marc Dionne 
Signed-off-by: David Howells

afs: Handle a short write to an AFS page

2017-03-16T16:27:44+00:00

Handle the situation where afs_write_begin() is told to expect that a
full-page write will be made, but this doesn't happen (EFAULT, CTRL-C,
etc.), and so afs_write_end() sees a partial write took place.  Currently,
no attempt is to deal with the discrepency.

Fix this by loading the gap from the server.

Reported-by: Al Viro 
Signed-off-by: David Howells

afs: Fix page overput in afs_fill_page()

2017-03-16T16:27:43+00:00

afs_fill_page() loads the page it wants to fill into the afs_read request
without incrementing its refcount - but then calls afs_put_read() to clean
up afterwards, which then releases a ref on the page.

Fix this by getting a ref on the page before calling
afs_vnode_fetch_data().

This causes sync after a write to hang in afs_writepages_region() because
find_get_pages_tag() gets confused and doesn't return.

Fixes: 196ee9cd2d04 ("afs: Make afs_fs_fetch_data() take a list of pages")
Reported-by: Marc Dionne 
Signed-off-by: David Howells 
Tested-by: Marc Dionne

afs: Fix missing put_page()

2017-03-16T16:27:43+00:00

In afs_writepages_region(), inside the loop where we find dirty pages to
deal with, one of the if-statements is missing a put_page().

Signed-off-by: David Howells

afs: Make afs_fs_fetch_data() take a list of pages

2017-01-06T16:54:41+00:00

Make afs_fs_fetch_data() take a list of pages for bulk data transfer.  This
will allow afs_readpages() to be made more efficient.

Signed-off-by: David Howells