[PATCH] mm: trace tlb flush after disabling preemption in try_to_unmap_flush

[PATCH] mm: trace tlb flush after disabling preemption in try_to_unmap_flush

[Sasha Levin]

Commit "mm: send one IPI per CPU to TLB flush all entries after unmapping
pages" added a trace_tlb_flush() while preemption was still enabled. This
means that we'll access smp_processor_id() which in turn will get us quite
a few warnings.

Fix it by moving the trace to where the preemption is disabled, one line
down.

Signed-off-by: Sasha Levin <sasha.levin@oracle.com>
---

The diff is all lies: I've moved trace_tlb_flush() one line down rather
than get_cpu() a line up ;)

 mm/rmap.c |    3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/mm/rmap.c b/mm/rmap.c
index 30812e9..63ba46c 100644
--- a/mm/rmap.c
+++ b/mm/rmap.c
@@ -613,9 +613,10 @@ void try_to_unmap_flush(void)
 	if (!tlb_ubc->flush_required)
 		return;
 
+	cpu = get_cpu();
+
 	trace_tlb_flush(TLB_REMOTE_SHOOTDOWN, -1UL);
 
-	cpu = get_cpu();
 	if (cpumask_test_cpu(cpu, &tlb_ubc->cpumask))
 		percpu_flush_tlb_batch_pages(&tlb_ubc->cpumask);
 
-- 
1.7.10.4

Re: [PATCH] mm: trace tlb flush after disabling preemption in try_to_unmap_flush

[Michal Hocko]

On Thu 16-07-15 15:35:39, Sasha Levin wrote:
> Commit "mm: send one IPI per CPU to TLB flush all entries after unmapping
> pages" added a trace_tlb_flush() while preemption was still enabled. This
> means that we'll access smp_processor_id() which in turn will get us quite
> a few warnings.
> 
> Fix it by moving the trace to where the preemption is disabled, one line
> down.
> 
> Signed-off-by: Sasha Levin <sasha.levin@oracle.com>

Reviewed-by: Michal Hocko <mhocko@suse.com>

> ---
> 
> The diff is all lies: I've moved trace_tlb_flush() one line down rather
> than get_cpu() a line up ;)
> 
>  mm/rmap.c |    3 ++-
>  1 file changed, 2 insertions(+), 1 deletion(-)
> 
> diff --git a/mm/rmap.c b/mm/rmap.c
> index 30812e9..63ba46c 100644
> --- a/mm/rmap.c
> +++ b/mm/rmap.c
> @@ -613,9 +613,10 @@ void try_to_unmap_flush(void)
>  	if (!tlb_ubc->flush_required)
>  		return;
>  
> +	cpu = get_cpu();
> +
>  	trace_tlb_flush(TLB_REMOTE_SHOOTDOWN, -1UL);
>  
> -	cpu = get_cpu();
>  	if (cpumask_test_cpu(cpu, &tlb_ubc->cpumask))
>  		percpu_flush_tlb_batch_pages(&tlb_ubc->cpumask);
>  
> -- 
> 1.7.10.4
> 
> --
> 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/

-- 
Michal Hocko
SUSE Labs