Re: 2.6.1-rc1 [resend]

From: Paolo Ornati (ornati_at_lycos.it)
Date: 12/31/03

  • Next message: Robert Gadsdon: "2.6.1-rc1-mm1 - ieee1394 broken again?"
    To: William Lee Irwin III <wli@holomorphy.com>
    Date:	Wed, 31 Dec 2003 16:45:23 +0100
    
    

    On Wednesday 31 December 2003 16:20, you wrote:
    > On Wednesday 31 December 2003 16:06, you wrote:
    > >> What io scheduler are you using? Or, could you post /var/log/dmesg?
    >
    > On Wed, Dec 31, 2003 at 04:19:27PM +0100, Paolo Ornati wrote:
    > > "dmesg" and "config" attached.
    >
    > Could you try this with elevator=deadline?

    ok, I have just tried...
    I don't see any big difference.

    The output of this script:
    _________________________________________________
    #!/bin/bash

    echo "HD test for linux `uname -r`"
    echo

    ra=8
    for i in `seq 7`; do
        echo "READAHEAD = $ra";
        hdparm -a $ra /dev/hda;
        for j in `seq 3`; do
            hdparm -t /dev/hda;
        done;
        ra=$(($ra*2));
    done
    _________________________________________________

    is:
    _________________________________________________
    HD test for linux 2.6.1-rc1

    READAHEAD = 8

    /dev/hda:
     setting fs readahead to 8
     readahead = 8 (on)

    /dev/hda:
     Timing buffered disk reads: 64 MB in 4.79 seconds = 13.37 MB/sec

    /dev/hda:
     Timing buffered disk reads: 64 MB in 4.79 seconds = 13.36 MB/sec

    /dev/hda:
     Timing buffered disk reads: 64 MB in 4.79 seconds = 13.37 MB/sec
    READAHEAD = 16

    /dev/hda:
     setting fs readahead to 16
     readahead = 16 (on)

    /dev/hda:
     Timing buffered disk reads: 64 MB in 2.48 seconds = 25.80 MB/sec

    /dev/hda:
     Timing buffered disk reads: 64 MB in 2.45 seconds = 26.14 MB/sec

    /dev/hda:
     Timing buffered disk reads: 64 MB in 2.43 seconds = 26.30 MB/sec
    READAHEAD = 32

    /dev/hda:
     setting fs readahead to 32
     readahead = 32 (on)

    /dev/hda:
     Timing buffered disk reads: 64 MB in 2.44 seconds = 26.23 MB/sec

    /dev/hda:
     Timing buffered disk reads: 64 MB in 2.43 seconds = 26.32 MB/sec

    /dev/hda:
     Timing buffered disk reads: 64 MB in 2.44 seconds = 26.27 MB/sec
    READAHEAD = 64

    /dev/hda:
     setting fs readahead to 64
     readahead = 64 (on)

    /dev/hda:
     Timing buffered disk reads: 64 MB in 2.43 seconds = 26.38 MB/sec

    /dev/hda:
     Timing buffered disk reads: 64 MB in 2.44 seconds = 26.27 MB/sec

    /dev/hda:
     Timing buffered disk reads: 64 MB in 2.43 seconds = 26.36 MB/sec
    READAHEAD = 128

    /dev/hda:
     setting fs readahead to 128
     readahead = 128 (on)

    /dev/hda:
     Timing buffered disk reads: 64 MB in 2.43 seconds = 26.33 MB/sec

    /dev/hda:
     Timing buffered disk reads: 64 MB in 2.44 seconds = 26.20 MB/sec

    /dev/hda:
     Timing buffered disk reads: 64 MB in 2.42 seconds = 26.40 MB/sec
    READAHEAD = 256

    /dev/hda:
     setting fs readahead to 256
     readahead = 256 (on)

    /dev/hda:
     Timing buffered disk reads: 64 MB in 2.41 seconds = 26.56 MB/sec

    /dev/hda:
     Timing buffered disk reads: 64 MB in 2.41 seconds = 26.55 MB/sec

    /dev/hda:
     Timing buffered disk reads: 64 MB in 2.41 seconds = 26.54 MB/sec
    READAHEAD = 512

    /dev/hda:
     setting fs readahead to 512
     readahead = 512 (on)

    /dev/hda:
     Timing buffered disk reads: 64 MB in 2.41 seconds = 26.52 MB/sec

    /dev/hda:
     Timing buffered disk reads: 64 MB in 2.40 seconds = 26.67 MB/sec

    /dev/hda:
     Timing buffered disk reads: 64 MB in 2.41 seconds = 26.54 MB/sec
    _________________________________________________

    >
    >
    > -- wli

    -- 
    	Paolo Ornati
    	Linux v2.6.1-rc1
    -
    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/
    

  • Next message: Robert Gadsdon: "2.6.1-rc1-mm1 - ieee1394 broken again?"

    Relevant Pages

    • Re: readahead logic and I/O errors
      ... And the reason is the readahead logic, ... request comes to the device, according to readahead logic, ... linux tries to get more data from the drive than requested. ... possible to FIRE the system this way (with certain drives ...
      (Linux-Kernel)
    • Re: sha1sum problems :-(
      ... After burning it to two dvd's on two different machines (one Windows, ... This uses the isoinfo command to determine how big the ISO file is, then runs a dd command that attempts to read exactly that much off the drive. ... I'm currently seeing a return of the old mediacheck problem (linux readahead bug). ... The old solution was to turn off readahead, I tried that recently, but there really *was* a problem in my case. ...
      (Fedora)
    • Re: poor sata performance on 2.6
      ... > have 32 bit DMA support, and will use bounce buffers quite a lot of the ... > again with no himem support at all and see if it helps. ... setting fs readahead to 8129 ... send the line "unsubscribe linux-kernel" in ...
      (Linux-Kernel)
    • Re: poor sata performance on 2.6
      ... and dmesg... ... >> again with no himem support at all and see if it helps. ... > setting fs readahead to 8129 ... send the line "unsubscribe linux-kernel" in ...
      (Linux-Kernel)
    • Re: [PATCH][2.6-mm] Readahead issues and AIO read speedup
      ... Doing it in readahead one more time causes ... Is it mainly for small reads that the performance hit due to the ... extra pagecache lookup is significant compared to the cycles spent ... send the line "unsubscribe linux-kernel" in ...
      (Linux-Kernel)