lkml.org 
[lkml]   [2011]   [Jun]   [2]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
    /
    Date
    SubjectRe: [PATCH] mm: compaction: Abort compaction if too many pages are isolated and caller is asynchronous
    From
    Hi Andrea,

    On Fri, Jun 3, 2011 at 6:40 AM, Andrea Arcangeli <aarcange@redhat.com> wrote:
    > Hello Minchan,
    >
    > On Fri, Jun 03, 2011 at 05:21:56AM +0900, Minchan Kim wrote:
    >> Isn't it rather aggressive?
    >> I think cursor page is likely to be PageTail rather than PageHead.
    >> Could we handle it simply with below code?
    >
    > It's not so likely, there is small percentage of compound pages that
    > aren't THP compared to the rest that is either regular pagecache or
    > anon regular or anon THP or regular shm. If it's THP chances are we

    I mean we have more tail pages than head pages. So I think we are likely to
    meet tail pages. Of course, compared to all pages(page cache, anon and
    so on), compound pages would be very small percentage.

    > isolated the head and it's useless to insist on more tail pages (at
    > least for large page size like on x86). Plus we've compaction so

    I can't understand your point. Could you elaborate it?

    > insisting and screwing lru ordering isn't worth it, better to be
    > permissive and abort... in fact I wouldn't dislike to remove the
    > entire lumpy logic when COMPACTION_BUILD is true, but that alters the
    > trace too...

    AFAIK, it's final destination to go as compaction will not break lru
    ordering if my patch(inorder-putback) is merged.

    >
    >> get_page(cursor_page)
    >> /* The page is freed already */
    >> if (1 == page_count(cursor_page)) {
    >>       put_page(cursor_page)
    >>       continue;
    >> }
    >> put_page(cursor_page);
    >
    > We can't call get_page on an tail page or we break split_huge_page,

    Why don't we call get_page on tail page if tail page isn't free?
    Maybe I need investigating split_huge_page.

    > only an isolated lru can be boosted, if we take the lru_lock and we
    > check the page is in lru, then we can isolate and pin it safely.
    >

    Thanks.


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

    \
     
     \ /
      Last update: 2011-06-03 00:25    [W:2.927 / U:0.044 seconds]
    ©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site