summaryrefslogtreecommitdiff
path: root/arch/x86/mm/tlb.c
diff options
context:
space:
mode:
Diffstat (limited to 'arch/x86/mm/tlb.c')
-rw-r--r--arch/x86/mm/tlb.c24
1 files changed, 12 insertions, 12 deletions
diff --git a/arch/x86/mm/tlb.c b/arch/x86/mm/tlb.c
index 90a9e4740913..6cf881a942bb 100644
--- a/arch/x86/mm/tlb.c
+++ b/arch/x86/mm/tlb.c
@@ -607,18 +607,15 @@ void switch_mm_irqs_off(struct mm_struct *unused, struct mm_struct *next,
cond_mitigation(tsk);
/*
- * Stop remote flushes for the previous mm.
- * Skip kernel threads; we never send init_mm TLB flushing IPIs,
- * but the bitmap manipulation can cause cache line contention.
+ * Leave this CPU in prev's mm_cpumask. Atomic writes to
+ * mm_cpumask can be expensive under contention. The CPU
+ * will be removed lazily at TLB flush time.
*/
- if (prev != &init_mm) {
- VM_WARN_ON_ONCE(!cpumask_test_cpu(cpu,
- mm_cpumask(prev)));
- cpumask_clear_cpu(cpu, mm_cpumask(prev));
- }
+ VM_WARN_ON_ONCE(prev != &init_mm && !cpumask_test_cpu(cpu,
+ mm_cpumask(prev)));
/* Start receiving IPIs and then read tlb_gen (and LAM below) */
- if (next != &init_mm)
+ if (next != &init_mm && !cpumask_test_cpu(cpu, mm_cpumask(next)))
cpumask_set_cpu(cpu, mm_cpumask(next));
next_tlb_gen = atomic64_read(&next->context.tlb_gen);
@@ -760,10 +757,13 @@ static void flush_tlb_func(void *info)
if (!local) {
inc_irq_stat(irq_tlb_count);
count_vm_tlb_event(NR_TLB_REMOTE_FLUSH_RECEIVED);
+ }
- /* Can only happen on remote CPUs */
- if (f->mm && f->mm != loaded_mm)
- return;
+ /* The CPU was left in the mm_cpumask of the target mm. Clear it. */
+ if (f->mm && f->mm != loaded_mm) {
+ cpumask_clear_cpu(raw_smp_processor_id(), mm_cpumask(f->mm));
+ trace_tlb_flush(TLB_REMOTE_WRONG_CPU, 0);
+ return;
}
if (unlikely(loaded_mm == &init_mm))