Re: [RFC, PATCHv2 0/2] mm: map few pages around fault address if they are in page cache

[Linus Torvalds <torvalds@linux-foundation.org>]

On Tue, Feb 18, 2014 at 6:15 AM, Wilcox, Matthew R
<matthew.r.wilcox@intel.com> wrote:
> We don't really need to lock all the pages being returned to protect against truncate.  We only need to lock the one at the highest index, and check i_size while that lock is held since truncate_inode_pages_range() will block on any page that is locked.
>
> We're still vulnerable to holepunches, but there's no locking currently between holepunches and truncate, so we're no worse off now.

It's not "holepunches and truncate", it's "holepunches and page
mapping", and I do think we currently serialize the two - the whole
"check page->mapping still being non-NULL" before mapping it while
having the page locked does that.

Besides, that per-page locking should serialize against truncate too.
No, there is no "global" serialization, but there *is* exactly that
page-level serialization where both truncation and hole punching end
up making sure that the page no longer exists in the page cache and
isn't mapped.

I'm just claiming that *because* of the way rmap works for file
mappings (walking the i_mapped list and page tables), we should
actually be ok.  The anonymous rmap list is protected by the page
lock, but the file-backed rmap is protected by the pte lock (well, and
the "i_mmap_mutex" that in turn protects the i_mmap list etc).

       Linus

--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@kvack.org.  For more info on Linux MM,
see: http://www.linux-mm.org/ .
Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>