lkml.org 
[lkml]   [2010]   [Jan]   [14]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
    /
    SubjectRe: [RFC PATCH] introduce sys_membarrier(): process-wide memory barrier (v5)
    From
    Date
    On Thu, 2010-01-14 at 12:54 -0500, Mathieu Desnoyers wrote:
    > * Peter Zijlstra (peterz@infradead.org) wrote:
    > > On Thu, 2010-01-14 at 11:26 -0500, Mathieu Desnoyers wrote:
    > >
    > > > It's this scenario that is causing problem. Let's consider this
    > > > execution:
    > > >
    >
    > (slightly augmented)
    >
    > CPU 0 (membarrier) CPU 1 (another mm -> our mm)
    > <user-space>
    > <kernel-space>
    > switch_mm()
    > smp_mb()
    > clear_mm_cpumask()
    > set_mm_cpumask()
    > smp_mb() (by load_cr3() on x86)
    > switch_to()
    > memory access before membarrier
    > <call sys_membarrier()>
    > smp_mb()
    > mm_cpumask includes CPU 1
    > rcu_read_lock()
    > if (CPU 1 mm != our mm)

    But here, CPU 1 updated its mm already and did a
    smp_mb, won't that make us send the smp_mb
    anyway?

    -- Steve

    > skip CPU 1.
    > rcu_read_unlock()
    > smp_mb()
    > <return to user-space>
    > current = next (1)
    > <switch back to user-space>
    > urcu read lock()
    > read gp
    > store local gp (2)
    > barrier()
    > access critical section data (3)
    > memory access after membarrier




    \
     
     \ /
      Last update: 2010-01-14 19:53    [W:0.035 / U:120.400 seconds]
    ©2003-2016 Jasper Spaans. hosted at Digital OceanAdvertise on this site