Re: THP-enabled filesystem vs. FALLOC_FL_PUNCH_HOLE

[Dave Chinner]

On Sun, Mar 06, 2016 at 03:30:34AM +0300, Kirill A. Shutemov wrote:
> On Sun, Mar 06, 2016 at 09:38:11AM +1100, Dave Chinner wrote:
> > On Sat, Mar 05, 2016 at 02:24:12AM +0300, Kirill A. Shutemov wrote:
> > > Would it be acceptable for fallocate(FALLOC_FL_PUNCH_HOLE) to return
> > > -EBUSY (or other errno on your choice), if we cannot split the page
> > > right away?
> > 
> > Which means THP are not transparent any more. What does an
> > application do when it gets an EBUSY, anyway?
> 
> I guess it's reasonable to expect from an application to handle EOPNOTSUPP
> as FALLOC_FL_PUNCH_HOLE is not supported by some filesystems.

Yes, but this is usually done as a check at the program
initialisation to determine whether to issue hole punches at all.
It's not suppose to be a dynamic error.

> Although, non-consistent result from the same fd can be confusing.

Exactly.

> > And it's not just hole punching that has this problem. Direct IO is
> > going to have the same issue with invalidation of the mapped ranges
> > over the IO being done. XFS already WARNs when page cache
> > invalidation fails with EBUSY in direct IO, because that is
> > indicative of an application with a potential data corruption vector
> > and there's nothing we can do in the kernel code to prevent it.
> 
> My current understanding is that for filesystems with persistent storage,
> in order to make THP any useful, we would need to implement writeback
> without splitting the huge page.

Algorithmically it is no different to filesytem block size < page
size writeback.

> At the moment, I have no idea how hard it would be..

THP support would effectively require us to remove PAGE_CACHE_SIZE
assumptions from all of the filesystem and buffer code. That's a
large chunk of work e.g.  fs/buffer.c and any filesystem that uses
bufferheads for tracking filesystem block state through the page
cache.

Cheers,

Dave.
-- 
Dave Chinner
david@xxxxxxxxxxxxx