lkml.org 
[lkml]   [2012]   [Jan]   [20]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
    /
    Date
    SubjectRe: [PATCH] Avoid mask based num_possible_cpus and num_online_cpus
    From
    On Thu, Jan 19, 2012 at 12:43 PM, Srivatsa S. Bhat
    <srivatsa.bhat@linux.vnet.ibm.com> wrote:
    > Hi,
    >
    > On 01/20/2012 01:31 AM, Venkatesh Pallipadi wrote:
    >
    >> Does this look better? Will send separate patch to fix code
    >> using num_possible_cpus() when they actually need nr_cpu_ids.
    >>
    >>
    >>
    >> Kernel's notion of possible cpus (from include/linux/cpumask.h)
    >>  *     cpu_possible_mask- has bit 'cpu' set iff cpu is populatable
    >>
    >>  *  The cpu_possible_mask is fixed at boot time, as the set of CPU id's
    >>  *  that it is possible might ever be plugged in at anytime during the
    >>  *  life of that system boot.
    >>
    >>  #define num_possible_cpus()     cpumask_weight(cpu_possible_mask)
    >>
    >> and on x86 cpumask_weight() calls hweight64 and hweight64 (on older kernels
    >> and systems with !X86_FEATURE_POPCNT) or a popcnt based alternative.
    >>
    >> i.e, We needlessly go through this mask based calculation everytime
    >> num_possible_cpus() is called.
    >>
    >> The problem is there with cpu_online_mask() as well, which is fixed value at
    >> boot time in !CONFIG_HOTPLUG_CPU case and should not change that often even
    >> in HOTPLUG case.
    >>
    >> Though most of the callers of these two routines are init time (with few
    >> exceptions of runtime calls), it is cleaner to use variables
    >> and not go through this repeated mask based calculation.
    >>
    >> Signed-off-by: Venkatesh Pallipadi <venki@google.com>
    >> ---
    >>  include/linux/cpumask.h |   10 ++++++++--
    >>  kernel/cpu.c            |    5 +++++
    >>  kernel/smp.c            |    4 ++++
    >>  3 files changed, 17 insertions(+), 2 deletions(-)
    >>
    >> diff --git a/include/linux/cpumask.h b/include/linux/cpumask.h
    >> index 4f7a632..ac3113b 100644
    >> --- a/include/linux/cpumask.h
    >> +++ b/include/linux/cpumask.h
    >> @@ -23,10 +23,14 @@ typedef struct cpumask { DECLARE_BITMAP(bits, NR_CPUS); } cpumask_t;
    >>
    >>  #if NR_CPUS == 1
    >>  #define nr_cpu_ids           1
    >> +#define nr_possible_cpus     1
    >>  #else
    >>  extern int nr_cpu_ids;
    >> +extern int nr_possible_cpus;
    >>  #endif
    >>
    >> +extern int nr_online_cpus;
    >> +
    >>  #ifdef CONFIG_CPUMASK_OFFSTACK
    >>  /* Assuming NR_CPUS is huge, a runtime limit is more efficient.  Also,
    >>   * not all bits may be allocated. */
    >> @@ -81,8 +85,10 @@ extern const struct cpumask *const cpu_present_mask;
    >>  extern const struct cpumask *const cpu_active_mask;
    >>
    >>  #if NR_CPUS > 1
    >> -#define num_online_cpus()    cpumask_weight(cpu_online_mask)
    >> -#define num_possible_cpus()  cpumask_weight(cpu_possible_mask)
    >> +
    >> +#define num_online_cpus()    (nr_online_cpus)
    >> +#define num_possible_cpus()  (nr_possible_cpus)
    >> +
    >>  #define num_present_cpus()   cpumask_weight(cpu_present_mask)
    >>  #define num_active_cpus()    cpumask_weight(cpu_active_mask)
    >>  #define cpu_online(cpu)              cpumask_test_cpu((cpu), cpu_online_mask)
    >> diff --git a/kernel/cpu.c b/kernel/cpu.c
    >> index 2060c6e..f179baa 100644
    >> --- a/kernel/cpu.c
    >> +++ b/kernel/cpu.c
    >> @@ -622,6 +622,9 @@ static DECLARE_BITMAP(cpu_active_bits, CONFIG_NR_CPUS) __read_mostly;
    >>  const struct cpumask *const cpu_active_mask = to_cpumask(cpu_active_bits);
    >>  EXPORT_SYMBOL(cpu_active_mask);
    >>
    >> +int nr_online_cpus __read_mostly;
    >> +EXPORT_SYMBOL(nr_online_cpus);
    >> +
    >>  void set_cpu_possible(unsigned int cpu, bool possible)
    >>  {
    >>       if (possible)
    >
    >
    > Did you forget to add:
    >
    >        nr_possible_cpus = cpumask_weight(cpu_possible_mask);
    >
    > inside set_cpu_possible() ?

    No. That was intentional as I have that coupled with nr_cpu_ids and
    set once after all the bits are set in setup_nr_cpu_ids() instead of
    doing for each bit set.

    >
    >
    >> @@ -644,6 +647,8 @@ void set_cpu_online(unsigned int cpu, bool online)
    >>               cpumask_set_cpu(cpu, to_cpumask(cpu_online_bits));
    >>       else
    >>               cpumask_clear_cpu(cpu, to_cpumask(cpu_online_bits));
    >> +
    >> +     nr_online_cpus = cpumask_weight(cpu_online_mask);
    >>  }
    >>
    >>  void set_cpu_active(unsigned int cpu, bool active)
    >> diff --git a/kernel/smp.c b/kernel/smp.c
    >> index db197d6..106e519 100644
    >> --- a/kernel/smp.c
    >> +++ b/kernel/smp.c
    >> @@ -658,10 +658,14 @@ early_param("maxcpus", maxcpus);
    >>  int nr_cpu_ids __read_mostly = NR_CPUS;
    >>  EXPORT_SYMBOL(nr_cpu_ids);
    >>
    >> +int nr_possible_cpus __read_mostly = NR_CPUS;
    >> +EXPORT_SYMBOL(nr_possible_cpus);
    >> +
    >>  /* An arch may set nr_cpu_ids earlier if needed, so this would be redundant */
    >>  void __init setup_nr_cpu_ids(void)
    >>  {
    >>       nr_cpu_ids = find_last_bit(cpumask_bits(cpu_possible_mask),NR_CPUS) + 1;
    >> +     nr_possible_cpus = cpumask_weight(cpu_possible_mask);
    >>  }
    >>
    >>  /* Called by boot processor to activate the rest. */
    >
    >
    > Regards,
    > Srivatsa S. Bhat
    >
    --
    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: 2012-01-21 00:11    [W:0.037 / U:60.936 seconds]
    ©2003-2016 Jasper Spaans. hosted at Digital OceanAdvertise on this site