[lkml]   [2004]   [Dec]   [21]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
Messages in this thread
    SubjectIncrease page fault rate by prezeroing V1 [0/3]: Overview

    The patches increasing the page fault rate (introduction of atomic pte operations
    and anticipatory prefaulting) do so by reducing the locking overhead and are
    therefore mainly of interest for applications running in SMP systems with a high
    number of cpus. The single thread performance does just show minor increases.
    Only the performance of multi-threaded applications increase significantly.

    The most expensive operation in the page fault handler is (apart of SMP
    locking overhead) the zeroing of the page that is also done in the page fault
    handler. Others have seen this too and have tried provide a way to provide
    zeroed pages to the page fault handler:

    The problem so far has been that simple zeroing of pages simply shifts
    the time spend somewhere else. Plus one would not want to zero hot

    This patch addresses those issues by making it more effective to zero pages by:

    1. Aggregating zeroing operations to mainly apply to larger order pages
    which results in many later order 0 pages to be zeroed in one go.
    For that purpose a new achitecture specific function zero_page(page, order)
    is introduced.

    2. Hardware support for offloading zeroing from the cpu. This avoids
    the invalidation of the cpu caches by extensive zeroing operations.

    The result is a significant increase of the page fault performance even for
    single threaded applications:

    w/o patch:
    Gb Rep Threads User System Wall flt/cpu/s fault/wsec
    4 3 1 0.146s 11.155s 11.030s 69584.896 69566.852

    Gb Rep Threads User System Wall flt/cpu/s fault/wsec
    1 1 1 0.014s 0.110s 0.012s524292.194 517665.538

    This is a performance increase by a factor 8!

    The performance can only be upheld if enough zeroed pages are available.
    In a heavy memory intensive benchmark the system will run out of these very
    fast but the efficient algorithm for page zeroing still makes this a winner
    (8 way system with 6 GB RAM, no hardware zeroing support):

    w/o patch:

    Gb Rep Threads User System Wall flt/cpu/s fault/wsec
    4 3 1 0.146s 11.155s 11.030s 69584.896 69566.852
    4 3 2 0.170s 14.909s 7.097s 52150.369 98643.687
    4 3 4 0.181s 16.597s 5.079s 46869.167 135642.420
    4 3 8 0.166s 23.239s 4.037s 33599.215 179791.120

    Gb Rep Threads User System Wall flt/cpu/s fault/wsec
    4 3 1 0.183s 2.750s 2.093s268077.996 267952.890
    4 3 2 0.185s 4.876s 2.097s155344.562 263967.292
    4 3 4 0.150s 6.617s 2.097s116205.793 264774.080
    4 3 8 0.186s 13.693s 3.054s 56659.819 221701.073

    The patch is composed of 3 parts:

    [1/3] Introduce __GFP_ZERO
    Modifies the page allocator to be able to take the __GFP_ZERO flag
    and returns zeroed memory on request. Modifies locations throughout
    the linux sources that retrieve a page and then zeroe it to request
    a zeroed page.
    Adds new low level zero_page functions for i386, ia64 and x86_64.
    (x64_64 untested)

    [2/3] Page Zeroing
    Adds management of ZEROED and NOT_ZEROED pages and a background daemon
    called scrubd. scrubd is disable by default but can be enabled
    by writing an order number to /proc/sys/vm/scrub_start. If a page
    is coalesced of that order then the scrub daemon will start zeroing
    until all pages of order /proc/sys/vm/scrub_stop and higher are

    [3/3] SGI Altix Block Transfer Engine Support
    Implements a driver to shift the zeroing off the cpu into hardware.
    With hardware support there will be minimal impact of zeroing
    on the performance of the system.

    To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
    the body of a message to
    More majordomo info at
    Please read the FAQ at

     \ /
      Last update: 2005-03-22 14:08    [W:0.024 / U:107.988 seconds]
    ©2003-2016 Jasper Spaans. hosted at Digital OceanAdvertise on this site