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

[Wilcox, Matthew R]

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.
________________________________________
From: Rik van Riel [riel@xxxxxxxxxx]
Sent: February 18, 2014 5:28 AM
To: Linus Torvalds; Kirill A. Shutemov
Cc: Andrew Morton; Mel Gorman; Andi Kleen; Wilcox, Matthew R; Dave Hansen; Alexander Viro; Dave Chinner; linux-mm; linux-fsdevel; Linux Kernel Mailing List
Subject: Re: [RFC, PATCHv2 0/2] mm: map few pages around fault address if they are in page cache

On 02/17/2014 02:01 PM, Linus Torvalds wrote:

>  - increment the page _mapcount (iow, do "page_add_file_rmap()"
> early). This guarantees that any *subsequent* unmap activity on this
> page will walk the file mapping lists, and become serialized by the
> page table lock we hold.
>
>  - mb_after_atomic_inc() (this is generally free)
>
>  - test that the page is still unlocked and uptodate, and the page
> mapping still points to our page.
>
>  - if that is true, we're all good, we can use the page, otherwise we
> decrement the mapcount (page_remove_rmap()) and skip the page.
>
> Hmm? Doing something like this means that we would never lock the
> pages we prefault, and you can go back to your gang lookup rather than
> that "one page at a time". And the race case is basically never going
> to trigger.
>
> Comments?

What would the direct io code do when it runs into a page with
elevated mapcount, but for which a mapping cannot be found yet?

Looking at the code, it looks like the above scheme could cause
some trouble with invalidate_inode_pages2_range(), which has
the following sequence:

                        if (page_mapped(page)) {
                                ... unmap page
                        }
                        BUG_ON(page_mapped(page));

In other words, it looks like incrementing _mapcount first could
lead to an oops in the truncate and direct IO code.

The page lock is used to prevent such races.

*sigh*

--
All rights reversed
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/