]> git.proxmox.com Git - mirror_ubuntu-kernels.git/commit
hfsplus: avoid deadlock on file truncation
authorErnesto A. Fernández <ernesto.mnd.fernandez@gmail.com>
Wed, 22 Aug 2018 04:59:16 +0000 (21:59 -0700)
committerLinus Torvalds <torvalds@linux-foundation.org>
Wed, 22 Aug 2018 17:52:50 +0000 (10:52 -0700)
commit31651c607151f1034cfb57e5a78678bea54c362b
treeb8868dcd8dd0330edeec238bb83716be61b3d314
parent7464726cb5998846306ed0a7d6714afb2e37b25d
hfsplus: avoid deadlock on file truncation

After an extent is removed from the extent tree, the corresponding bits
are also cleared from the block allocation file.  This is currently done
without releasing the tree lock.

The problem is that the allocation file has extents of its own; if it is
fragmented enough, some of them may be in the extent tree as well, and
hfsplus_get_block() will try to take the lock again.

To avoid deadlock, only hold the extent tree lock during the actual tree
operations.

Link: http://lkml.kernel.org/r/20180709202549.auxwkb6memlegb4a@eaf
Signed-off-by: Ernesto A. Fernández <ernesto.mnd.fernandez@gmail.com>
Reported-by: Anatoly Trosinenko <anatoly.trosinenko@gmail.com>
Cc: Viacheslav Dubeyko <slava@dubeyko.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
fs/hfsplus/extents.c