gfs2: Fix end-of-file handling in gfs2_page_mkwrite

Browse source

When the filesystem block size is smaller than the page size, the last
page may contain blocks that lie entirely beyond the end of the file.
Make sure to only allocate blocks that lie at least partially in the
file.  Allocating blocks beyond that isn't useful, and what's more, they
will not be zeroed out and may end up containing random data.

With that change in place, make sure we'll still always unstuff stuffed
inodes: iomap_writepage and iomap_writepages currently can't handle
stuffed files.

In addition, simplify and move the end-of-file check further to the top
in gfs2_page_mkwrite to avoid weird side effects like unstuffing when
we're not.

Fixes xfstest generic/263.

Signed-off-by: Andreas Gruenbacher <agruenba@redhat.com>

...

This commit is contained in:

Andreas Gruenbacher 2019-11-06 14:09:25 +00:00

parent f53056c430

commit 184b4e6085

1 changed files with 26 additions and 14 deletions

Show all changes Ignore whitespace when comparing lines Ignore changes in amount of whitespace Ignore changes in whitespace at EOL

Download patch file Download diff file Expand all files Collapse all files

40

fs/gfs2/file.c

View file