Re: [PATCH v2 03/11] pmem: enable REQ_FUA/REQ_FLUSH handling

[Ross Zwisler <ross.zwisler@linux.intel.com>]

On Wed, Nov 18, 2015 at 11:40:55AM +0100, Jan Kara wrote:
> On Mon 16-11-15 13:09:50, Ross Zwisler wrote:
> > On Fri, Nov 13, 2015 at 06:32:40PM -0800, Dan Williams wrote:
> > > On Fri, Nov 13, 2015 at 4:43 PM, Andreas Dilger <adilger@dilger.ca> wrote:
> > > > On Nov 13, 2015, at 5:20 PM, Dan Williams <dan.j.williams@intel.com> wrote:
> > > >>
> > > >> On Fri, Nov 13, 2015 at 4:06 PM, Ross Zwisler
> > > >> <ross.zwisler@linux.intel.com> wrote:
> > > >>> Currently the PMEM driver doesn't accept REQ_FLUSH or REQ_FUA bios.  These
> > > >>> are sent down via blkdev_issue_flush() in response to a fsync() or msync()
> > > >>> and are used by filesystems to order their metadata, among other things.
> > > >>>
> > > >>> When we get an msync() or fsync() it is the responsibility of the DAX code
> > > >>> to flush all dirty pages to media.  The PMEM driver then just has issue a
> > > >>> wmb_pmem() in response to the REQ_FLUSH to ensure that before we return all
> > > >>> the flushed data has been durably stored on the media.
> > > >>>
> > > >>> Signed-off-by: Ross Zwisler <ross.zwisler@linux.intel.com>
> > > >>
> > > >> Hmm, I'm not seeing why we need this patch.  If the actual flushing of
> > > >> the cache is done by the core why does the driver need support
> > > >> REQ_FLUSH?  Especially since it's just a couple instructions.  REQ_FUA
> > > >> only makes sense if individual writes can bypass the "drive" cache,
> > > >> but no I/O submitted to the driver proper is ever cached we always
> > > >> flush it through to media.
> > > >
> > > > If the upper level filesystem gets an error when submitting a flush
> > > > request, then it assumes the underlying hardware is broken and cannot
> > > > be as aggressive in IO submission, but instead has to wait for in-flight
> > > > IO to complete.
> > > 
> > > Upper level filesystems won't get errors when the driver does not
> > > support flush.  Those requests are ended cleanly in
> > > generic_make_request_checks().  Yes, the fs still needs to wait for
> > > outstanding I/O to complete but in the case of pmem all I/O is
> > > synchronous.  There's never anything to await when flushing at the
> > > pmem driver level.
> > > 
> > > > Since FUA/FLUSH is basically a no-op for pmem devices,
> > > > it doesn't make sense _not_ to support this functionality.
> > > 
> > > Seems to be a nop either way.  Given that DAX may lead to dirty data
> > > pending to the device in the cpu cache that a REQ_FLUSH request will
> > > not touch, its better to leave it all to the mm core to handle.  I.e.
> > > it doesn't make sense to call the driver just for two instructions
> > > (sfence + pcommit) when the mm core is taking on the cache flushing.
> > > Either handle it all in the mm or the driver, not a mixture.
> > 
> > Does anyone know if ext4 and/or XFS alter their algorithms based on whether
> > the driver supports REQ_FLUSH/REQ_FUA?  Will the filesystem behave more
> > efficiently with respect to their internal I/O ordering, etc., if PMEM
> > advertises REQ_FLUSH/REQ_FUA support, even though we could do the same thing
> > at the DAX layer?
> 
> So the information whether the driver supports FLUSH / FUA is generally
> ignored by filesystems. We issue REQ_FLUSH / REQ_FUA requests to achieve
> required ordering for fs consistency and expect that block layer does the
> right thing - i.e., if the device has volatile write cache, it will be
> flushed, if it doesn't have it, the request will be ignored. So the
> difference between supporting and not supporting REQ_FLUSH / REQ_FUA is
> only in how block layer handles such requests.

Cool, thank you for the info.  Based on this I'll pull out the
REQ_FLUSH/REQ_FUA patch for v3 of this series and move the wmb_pmem() call up
to DAX as Dan suggests.  If performance data shows that we can get a benefit
from centralizing wmb_pmem() behind REQ_FUA/REQ_FLUSH, I'll add it back in
later as part of that series.