Hi Linus, please pull from:
git://git.kernel.org/pub/scm/linux/kernel/git/nvdimm/nvdimm libnvdimm-fixes
...to receive the completion of Jan's DAX work for 4.10.
As I mentioned in the libnvdimm-for-4.10 pull request +AFs-1+AF0-, these are
some final fixes for the DAX dirty-cacheline-tracking invalidation work
that was merged through the -mm, ext4, and xfs trees in -rc1. These
patches were prepared prior to the merge window, but we waited for
4.10-rc1 to have a stable merge base after all the prerequisites were
merged.
Quoting Jan on the overall changes in these patches:
So I'd like all these 6 patches to go for rc2. The first three
patches fix invalidation of exceptional DAX entries (a bug which is
there for a long time) - without these patches data loss can occur
on power failure even though user called fsync(2). The other three
patches change locking of DAX faults so that -+AD4-iomap+AF8-begin() is
called in a more relaxed locking context and we are safe to start a
transaction there for ext4.
These have received a build success notification from the kbuild robot,
and pass the latest libnvdimm unit tests. There have not been any -next
releases since -rc1, so they have not appeared there.
+AFs-1+AF0-:+AKA-https://lists.01.org/pipermail/linux-nvdimm/2016-December/008279.h
tml
---
The following changes since commit 7ce7d89f48834cefece7804d38fc5d85382edf77:
Linux 4.10-rc1 (2016-12-25 16:13:08 -0800)
are available in the git repository at:
git://git.kernel.org/pub/scm/linux/kernel/git/nvdimm/nvdimm libnvdimm-fixes
for you to fetch changes up to 1db175428ee374489448361213e9c3b749d14900:
ext4: Simplify DAX fault path (2016-12-26 20:29:25 -0800)
----------------------------------------------------------------
Jan Kara (6):
ext2: Return BH+AF8-New buffers for zeroed blocks
mm: Invalidate DAX radix tree entries only if appropriate
dax: Avoid page invalidation races and unnecessary radix tree traversals
dax: Finish fault completely when loading holes
dax: Call -+AD4-iomap+AF8-begin without entry lock during dax fault
ext4: Simplify DAX fault path
fs/dax.c +AHw- 243 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+--------------------
fs/ext2/inode.c +AHw- 3 +--
fs/ext4/file.c +AHw- 48 +-+-+---------
include/linux/dax.h +AHw- 3 +-
mm/truncate.c +AHw- 75 +-+-+-+-+-+-+-+-+-+-+-+-+----
5 files changed, 229 insertions(+-), 143 deletions(-)
commit e568df6b84ff05a22467503afc11bee7a6ba0700
Author: Jan Kara +ADw-jack+AEA-suse.cz+AD4-
Date: Wed Aug 10 16:42:53 2016 +-0200
ext2: Return BH+AF8-New buffers for zeroed blocks
So far we did not return BH+AF8-New buffers from ext2+AF8-get+AF8-blocks() when we
allocated and zeroed-out a block for DAX inode to avoid racy zeroing in
DAX code. This zeroing is gone these days so we can remove the
workaround.
Reviewed-by: Ross Zwisler +ADw-ross.zwisler+AEA-linux.intel.com+AD4-
Reviewed-by: Christoph Hellwig +ADw-hch+AEA-lst.de+AD4-
Signed-off-by: Jan Kara +ADw-jack+AEA-suse.cz+AD4-
Signed-off-by: Dan Williams +ADw-dan.j.williams+AEA-intel.com+AD4-
commit c6dcf52c23d2d3fb5235cec42d7dd3f786b87d55
Author: Jan Kara +ADw-jack+AEA-suse.cz+AD4-
Date: Wed Aug 10 17:22:44 2016 +-0200
mm: Invalidate DAX radix tree entries only if appropriate
Currently invalidate+AF8-inode+AF8-pages2+AF8-range() and invalidate+AF8-mapping+AF8-pages()
just delete all exceptional radix tree entries they find. For DAX this
is not desirable as we track cache dirtiness in these entries and when
they are evicted, we may not flush caches although it is necessary. This
can for example manifest when we write to the same block both via mmap
and via write(2) (to different offsets) and fsync(2) then does not
properly flush CPU caches when modification via write(2) was the last
one.
Create appropriate DAX functions to handle invalidation of DAX entries
for invalidate+AF8-inode+AF8-pages2+AF8-range() and invalidate+AF8-mapping+AF8-pages() and
wire them up into the corresponding mm functions.
Acked-by: Johannes Weiner +ADw-hannes+AEA-cmpxchg.org+AD4-
Reviewed-by: Ross Zwisler +ADw-ross.zwisler+AEA-linux.intel.com+AD4-
Signed-off-by: Jan Kara +ADw-jack+AEA-suse.cz+AD4-
Signed-off-by: Dan Williams +ADw-dan.j.williams+AEA-intel.com+AD4-
commit e3fce68cdbed297d927e993b3ea7b8b1cee545da
Author: Jan Kara +ADw-jack+AEA-suse.cz+AD4-
Date: Wed Aug 10 17:10:28 2016 +-0200
dax: Avoid page invalidation races and unnecessary radix tree traversals
Currently dax+AF8-iomap+AF8-rw() takes care of invalidating page tables and
evicting hole pages from the radix tree when write(2) to the file
happens. This invalidation is only necessary when there is some block
allocation resulting from write(2). Furthermore in current place the
invalidation is racy wrt page fault instantiating a hole page just after
we have invalidated it.
So perform the page invalidation inside dax+AF8-iomap+AF8-actor() where we can
do it only when really necessary and after blocks have been allocated so
nobody will be instantiating new hole pages anymore.
Reviewed-by: Christoph Hellwig +ADw-hch+AEA-lst.de+AD4-
Reviewed-by: Ross Zwisler +ADw-ross.zwisler+AEA-linux.intel.com+AD4-
Signed-off-by: Jan Kara +ADw-jack+AEA-suse.cz+AD4-
Signed-off-by: Dan Williams +ADw-dan.j.williams+AEA-intel.com+AD4-
commit f449b936f1aff7696b24a338f493d5cee8d48d55
Author: Jan Kara +ADw-jack+AEA-suse.cz+AD4-
Date: Wed Oct 19 14:48:38 2016 +-0200
dax: Finish fault completely when loading holes
The only case when we do not finish the page fault completely is when we
are loading hole pages into a radix tree. Avoid this special case and
finish the fault in that case as well inside the DAX fault handler. It
will allow us for easier iomap handling.
Reviewed-by: Ross Zwisler +ADw-ross.zwisler+AEA-linux.intel.com+AD4-
Signed-off-by: Jan Kara +ADw-jack+AEA-suse.cz+AD4-
Signed-off-by: Dan Williams +ADw-dan.j.williams+AEA-intel.com+AD4-
commit 9f141d6ef6258a3a37a045842d9ba7e68f368956
Author: Jan Kara +ADw-jack+AEA-suse.cz+AD4-
Date: Wed Oct 19 14:34:31 2016 +-0200
dax: Call -+AD4-iomap+AF8-begin without entry lock during dax fault
Currently -+AD4-iomap+AF8-begin() handler is called with entry lock held. If the
filesystem held any locks between -+AD4-iomap+AF8-begin() and -+AD4-iomap+AF8-end()
(such as ext4 which will want to hold transaction open), this would cause
lock inversion with the iomap+AF8-apply() from standard IO path which first
calls -+AD4-iomap+AF8-begin() and only then calls -+AD4-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 -+AD4-iomap+AF8-begin()
- -+AD4-iomap+AF8-end() pair.
Reviewed-by: Ross Zwisler +ADw-ross.zwisler+AEA-linux.intel.com+AD4-
Signed-off-by: Jan Kara +ADw-jack+AEA-suse.cz+AD4-
Signed-off-by: Dan Williams +ADw-dan.j.williams+AEA-intel.com+AD4-
commit 1db175428ee374489448361213e9c3b749d14900
Author: Jan Kara +ADw-jack+AEA-suse.cz+AD4-
Date: Fri Oct 21 11:33:49 2016 +-0200
ext4: Simplify DAX fault path
Now that dax+AF8-iomap+AF8-fault() calls -+AD4-iomap+AF8-begin() without entry lock, we
can use transaction starting in ext4+AF8-iomap+AF8-begin() and thus simplify
ext4+AF8-dax+AF8-fault(). It also provides us proper retries in case of ENOSPC.
Signed-off-by: Jan Kara +ADw-jack+AEA-suse.cz+AD4-
Signed-off-by: Dan Williams +ADw-dan.j.williams+AEA-intel.com+AD4-
--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@xxxxxxxxx. For more info on Linux MM,
see: http://www.linux-mm.org/ .
Don't email: <a href