lkml.org 
[lkml]   [2003]   [Sep]   [2]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
    /
    SubjectRe: x86, ARM, PARISC, PPC, MIPS and Sparc folks please run this
    From
    Date
    On Mon, 2003-09-01 at 12:08, Jamie Lokier wrote:
    > Kars de Jong wrote:
    > > On Mon, 2003-09-01 at 10:34, Geert Uytterhoeven wrote:
    > > > BTW, probably you want us to run your test program on other m68k boxes? Mine
    > > > got a 68040, that leaves us with:
    > > > - 68020+68551
    > > > - 68060
    > >
    > > I can run it on these boxes if no-one else has done it yet before I come
    > > home tonight. I'm sure there are more people with a 68060 out there, not
    > > too sure about the 68020+68851.
    >
    > I would prefer that you run the attached program. It fixes a bug in
    > the function which tests whether the problem is in the L1 cache or
    > store buffer. The bug probably didn't affect the test, but it might
    > have.
    >
    > Ideally you could run the program Geert linked to as well?
    > Please remember to compile both with optimisation.

    OK, here are my results (I'll skip the 68060 because Roman has already
    run the program on that one):

    This is on a Plessey PME 68-22. It's sooooo fast... Sam, is there a Sun
    slower than this?

    Original program:

    fikkie:/tmp# ./jamie_test
    Test separation: 4096 bytes: pass
    Test separation: 8192 bytes: pass
    Test separation: 16384 bytes: pass
    Test separation: 32768 bytes: pass
    Test separation: 65536 bytes: pass
    Test separation: 131072 bytes: pass
    Test separation: 262144 bytes: pass
    Test separation: 524288 bytes: pass
    Test separation: 1048576 bytes: pass
    Test separation: 2097152 bytes: pass
    Test separation: 4194304 bytes: pass
    Test separation: 8388608 bytes: pass
    Test separation: 16777216 bytes: pass
    VM page alias coherency test: all sizes passed

    New program:

    fikkie:/tmp# time ./jamie_test2
    (2048) [10000,10000,0] Test separation: 4096 bytes: pass
    (2048) [10000,10000,0] Test separation: 8192 bytes: pass
    (2048) [10000,10000,0] Test separation: 16384 bytes: pass
    (2048) [10000,10000,0] Test separation: 32768 bytes: pass
    (2048) [10000,10000,0] Test separation: 65536 bytes: pass
    (2048) [10000,10000,0] Test separation: 131072 bytes: pass
    (2048) [10000,10000,0] Test separation: 262144 bytes: pass
    (2048) [10000,10000,0] Test separation: 524288 bytes: pass
    (2048) [10000,10000,0] Test separation: 1048576 bytes: pass
    (2048) [10000,10000,0] Test separation: 2097152 bytes: pass
    (2048) [10000,10000,0] Test separation: 4194304 bytes: pass
    (2048) [10000,10000,0] Test separation: 8388608 bytes: pass
    (2048) [10000,10000,0] Test separation: 16777216 bytes: pass
    VM page alias coherency test: all sizes passed

    real 1m51.210s
    user 1m44.950s
    sys 0m4.930s
    fikkie:/tmp# cat /proc/cpuinfo
    CPU: 68020
    MMU: 68851
    FPU: 68881
    Clocking: 15.6MHz
    BogoMips: 3.90
    Calibration: 19520 loops
    fikkie:/tmp#

    And no, this board has no way of getting a better time resolution than
    the 100 Hz tick timer either ;-)

    Regards,

    Kars.

    -
    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/
    On Mon, 2003-09-01 at 12:08, Jamie Lokier wrote:
    > Kars de Jong wrote:
    > > On Mon, 2003-09-01 at 10:34, Geert Uytterhoeven wrote:
    > > > BTW, probably you want us to run your test program on other m68k boxes? Mine
    > > > got a 68040, that leaves us with:
    > > > - 68020+68551
    > > > - 68060
    > >
    > > I can run it on these boxes if no-one else has done it yet before I come
    > > home tonight. I'm sure there are more people with a 68060 out there, not
    > > too sure about the 68020+68851.
    >
    > I would prefer that you run the attached program. It fixes a bug in
    > the function which tests whether the problem is in the L1 cache or
    > store buffer. The bug probably didn't affect the test, but it might
    > have.
    >
    > Ideally you could run the program Geert linked to as well?
    > Please remember to compile both with optimisation.

    OK, here are my results (I'll skip the 68060 because Roman has already
    run the program on that one):

    This is on a Plessey PME 68-22. It's sooooo fast... Sam, is there a Sun
    slower than this?

    Original program:

    fikkie:/tmp# ./jamie_test
    Test separation: 4096 bytes: pass
    Test separation: 8192 bytes: pass
    Test separation: 16384 bytes: pass
    Test separation: 32768 bytes: pass
    Test separation: 65536 bytes: pass
    Test separation: 131072 bytes: pass
    Test separation: 262144 bytes: pass
    Test separation: 524288 bytes: pass
    Test separation: 1048576 bytes: pass
    Test separation: 2097152 bytes: pass
    Test separation: 4194304 bytes: pass
    Test separation: 8388608 bytes: pass
    Test separation: 16777216 bytes: pass
    VM page alias coherency test: all sizes passed

    New program:

    fikkie:/tmp# time ./jamie_test2
    (2048) [10000,10000,0] Test separation: 4096 bytes: pass
    (2048) [10000,10000,0] Test separation: 8192 bytes: pass
    (2048) [10000,10000,0] Test separation: 16384 bytes: pass
    (2048) [10000,10000,0] Test separation: 32768 bytes: pass
    (2048) [10000,10000,0] Test separation: 65536 bytes: pass
    (2048) [10000,10000,0] Test separation: 131072 bytes: pass
    (2048) [10000,10000,0] Test separation: 262144 bytes: pass
    (2048) [10000,10000,0] Test separation: 524288 bytes: pass
    (2048) [10000,10000,0] Test separation: 1048576 bytes: pass
    (2048) [10000,10000,0] Test separation: 2097152 bytes: pass
    (2048) [10000,10000,0] Test separation: 4194304 bytes: pass
    (2048) [10000,10000,0] Test separation: 8388608 bytes: pass
    (2048) [10000,10000,0] Test separation: 16777216 bytes: pass
    VM page alias coherency test: all sizes passed

    real 1m51.210s
    user 1m44.950s
    sys 0m4.930s
    fikkie:/tmp# cat /proc/cpuinfo
    CPU: 68020
    MMU: 68851
    FPU: 68881
    Clocking: 15.6MHz
    BogoMips: 3.90
    Calibration: 19520 loops
    fikkie:/tmp#

    And no, this board has no way of getting a better time resolution than
    the 100 Hz tick timer either ;-)

    Regards,

    Kars.

    -
    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: 2005-03-22 13:48    [W:0.033 / U:32.748 seconds]
    ©2003-2016 Jasper Spaans. hosted at Digital OceanAdvertise on this site