Just find imbalance issue on the patchset.
I write a one line program:
int main ()
{
int i;
for (i=0; i< 1; )
__asm__ __volatile__ ("nop");
}
it was compiled with name pl and start it on my 2 socket * 4 cores *
HT NUMA machine:
the cpu domain top like this:
domain 0: span 4,12 level SIBLING
groups: 4 (cpu_power = 589) 12 (cpu_power = 589)
domain 1: span 0,2,4,6,8,10,12,14 level MC
groups: 4,12 (cpu_power = 1178) 6,14 (cpu_power = 1178) 0,8
(cpu_power = 1178) 2,10 (cpu_power = 1178)
domain 2: span 0,2,4,6,8,10,12,14 level CPU
groups: 0,2,4,6,8,10,12,14 (cpu_power = 4712)
domain 3: span 0-15 level NUMA
groups: 0,2,4,6,8,10,12,14 (cpu_power = 4712) 1,3,5,7,9,11,13,15
(cpu_power = 4712)
$for ((i=0; i< I; i++)); do ./pl & done
when I = 2, they are running on cpu 0,12
I = 4, they are running on cpu 0,9,12,14
I = 8, they are running on cpu 0,4,9,10,11,12,13,14
Regards!
Alex
On Sat, Nov 17, 2012 at 12:25 AM, Ingo Molnar <mingo@xxxxxxxxxx> wrote:
> This is the split-out series of mm/ patches that got no objections
> from the latest (v15) posting of numa/core. If everyone is still
> fine with these then these will be merge candidates for v3.8.
>
> I left out the more contentious policy bits that people are still
> arguing about.
>
> The numa/base tree can also be found here:
>
> git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip.git numa/base
>
> Thanks,
>
> Ingo
>
> ------------------->
>
> Andrea Arcangeli (1):
> numa, mm: Support NUMA hinting page faults from gup/gup_fast
>
> Gerald Schaefer (1):
> sched, numa, mm, s390/thp: Implement pmd_pgprot() for s390
>
> Ingo Molnar (1):
> mm/pgprot: Move the pgprot_modify() fallback definition to mm.h
>
> Lee Schermerhorn (3):
> mm/mpol: Add MPOL_MF_NOOP
> mm/mpol: Check for misplaced page
> mm/mpol: Add MPOL_MF_LAZY
>
> Peter Zijlstra (7):
> sched, numa, mm: Make find_busiest_queue() a method
> sched, numa, mm: Describe the NUMA scheduling problem formally
> mm/thp: Preserve pgprot across huge page split
> mm/mpol: Make MPOL_LOCAL a real policy
> mm/mpol: Create special PROT_NONE infrastructure
> mm/migrate: Introduce migrate_misplaced_page()
> mm/mpol: Use special PROT_NONE to migrate pages
>
> Ralf Baechle (1):
> sched, numa, mm, MIPS/thp: Add pmd_pgprot() implementation
>
> Rik van Riel (5):
> mm/generic: Only flush the local TLB in ptep_set_access_flags()
> x86/mm: Only do a local tlb flush in ptep_set_access_flags()
> x86/mm: Introduce pte_accessible()
> mm: Only flush the TLB when clearing an accessible pte
> x86/mm: Completely drop the TLB flush from ptep_set_access_flags()
>
> Documentation/scheduler/numa-problem.txt | 230 +++++++++++++++++++++++++++++++
> arch/mips/include/asm/pgtable.h | 2 +
> arch/s390/include/asm/pgtable.h | 13 ++
> arch/x86/include/asm/pgtable.h | 7 +
> arch/x86/mm/pgtable.c | 8 +-
> include/asm-generic/pgtable.h | 4 +
> include/linux/huge_mm.h | 19 +++
> include/linux/mempolicy.h | 8 ++
> include/linux/migrate.h | 7 +
> include/linux/migrate_mode.h | 3 +
> include/linux/mm.h | 32 +++++
> include/uapi/linux/mempolicy.h | 16 ++-
> kernel/sched/fair.c | 20 +--
> mm/huge_memory.c | 174 +++++++++++++++--------
> mm/memory.c | 119 +++++++++++++++-
> mm/mempolicy.c | 143 +++++++++++++++----
> mm/migrate.c | 85 ++++++++++--
> mm/mprotect.c | 31 +++--
> mm/pgtable-generic.c | 9 +-
> 19 files changed, 807 insertions(+), 123 deletions(-)
> create mode 100644 Documentation/scheduler/numa-problem.txt
>
> --
> 1.7.11.7
>
> --
> To unsubscribe, send a message with 'unsubscribe linux-mm' in
> the body to majordomo@xxxxxxxxx. For more info on Linux MM,
> see: http://www.linux-mm.org/ .
> Don't email: <a href=mailto:"dont@xxxxxxxxx";> email@xxxxxxxxx </a>
--
Thanks
Alex
--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@xxxxxxxxx. For more info on Linux MM,
see: http://www.linux-mm.org/ .
Don't email: <a href=mailto:"dont@xxxxxxxxx";> email@xxxxxxxxx </a>