lkml.org 
[lkml]   [2006]   [Jan]   [31]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
    Patch in this message
    /
    Date
    From
    SubjectRe: [Patch] sched: new sched domain for representing multi-core
    On Tue, Jan 31, 2006 at 06:21:36PM -0800, Andrew Morton wrote:
    > I was assuming that the code really does something like:
    >
    > #ifdef CONFIG_SCHED_MC
    > some_global_thing = <expr>
    > #endif
    > #ifdef CONFIG_SCHED_SMT
    > some_global_thing = <expr>
    > #endif
    > }
    > ...
    > #ifdef CONFIG_SCHED_SMT
    > some_other_global_thing = <expr>
    > #endif
    > #ifdef CONFIG_SCHED_MC
    > some_other_global_thing = <expr>
    > #endif
    >
    > Which, looking a bit closer, was wrong (yes?)

    yes.

    >
    > It is a bit irregular that in one place we do the SMT processing first and
    > in another we do the MC processing first, but I guess it'll work OK.

    yes. It will work Ok.

    > We do need to be super-careful in the reviewing and testing here. If we
    > slip up we won't have a nice crash to tell us. Instead we'll find that
    > some machines with some configs will, under some workloads, take a few
    > percent longer than they should. We could waste people's time for years
    > until some developer stumbles across something.

    I have done testing with specJBB, kernel-compilation, specrate and we are
    doing some testing with database workload.. I will also request our
    perf team to take a stab at this.

    BTW, can you also apply this experimental only patch to -mm.

    --
    test patch for -mm.. enable CONFIG_SCHED_MC by default in Kconfig.
    on systems with no shared caches between cores, this
    should help validate domain setup and degeneration code..

    Signed-off-by: Suresh Siddha <suresh.b.siddha@intel.com>

    --- linux-2.6.16-rc1/arch/i386/Kconfig 2006-01-31 16:41:38.019406000 -0800
    +++ linux-core/arch/i386/Kconfig 2006-01-31 17:35:50.745916408 -0800
    @@ -238,6 +238,7 @@ config SCHED_SMT
    config SCHED_MC
    bool "Multi-core scheduler support"
    depends on SMP
    + default y
    help
    Multi-core scheduler support improves the CPU scheduler's decision
    making when dealing with multi-core CPU chips at a cost of slightly
    --- linux-2.6.16-rc1/arch/x86_64/Kconfig 2006-01-31 16:41:38.021405696 -0800
    +++ linux-core/arch/x86_64/Kconfig 2006-01-31 17:35:20.640493128 -0800
    @@ -249,6 +249,7 @@ config SCHED_SMT
    config SCHED_MC
    bool "Multi-core scheduler support"
    depends on SMP
    + default y
    help
    Multi-core scheduler support improves the CPU scheduler's decision
    making when dealing with multi-core CPU chips at a cost of slightly
    -
    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: 2006-02-01 03:54    [W:0.025 / U:0.324 seconds]
    ©2003-2016 Jasper Spaans. hosted at Digital OceanAdvertise on this site