Revision 9f141d6ef6258a3a37a045842d9ba7e68f368956 authored by Jan Kara on 19 October 2016, 12:34:31 UTC, committed by Dan Williams on 27 December 2016, 04:29:25 UTC
Currently ->iomap_begin() handler is called with entry lock held. If the
filesystem held any locks between ->iomap_begin() and ->iomap_end()
(such as ext4 which will want to hold transaction open), this would cause
lock inversion with the iomap_apply() from standard IO path which first
calls ->iomap_begin() and only then calls ->actor() callback which grabs
entry locks for DAX (if it faults when copying from/to user provided
buffers).

Fix the problem by nesting grabbing of entry lock inside ->iomap_begin()
- ->iomap_end() pair.

Reviewed-by: Ross Zwisler <ross.zwisler@linux.intel.com>
Signed-off-by: Jan Kara <jack@suse.cz>
Signed-off-by: Dan Williams <dan.j.williams@intel.com>
1 parent f449b93
History
File Mode Size
ad1816a
ad1848
cs423x
es1688
galaxy
gus
msnd
opti9xx
sb
wavefront
wss
Kconfig -rw-r--r-- 12.5 KB
Makefile -rw-r--r-- 857 bytes
adlib.c -rw-r--r-- 2.6 KB
als100.c -rw-r--r-- 10.3 KB
azt2320.c -rw-r--r-- 9.6 KB
cmi8328.c -rw-r--r-- 13.3 KB
cmi8330.c -rw-r--r-- 22.5 KB
es18xx.c -rw-r--r-- 69.3 KB
opl3sa2.c -rw-r--r-- 28.9 KB
sc6000.c -rw-r--r-- 17.7 KB
sscape.c -rw-r--r-- 32.6 KB

back to top