Re: [PATCH v2 4/6] mm: Only IPI CPUs to drain local pages if they exist

[PATCH v2 0/6] Reduce cross CPU IPI interference · Gilad Ben-Yossef <gilad@benyossef.com> · 2011-10-23
[PATCH v2 1/6] smp: Introduce a generic on_each_cpu_mask function · Gilad Ben-Yossef <gilad@benyossef.com> · 2011-10-23
[PATCH v2 2/6] arm: Move arm over to generic on_each_cpu_mask · Gilad Ben-Yossef <gilad@benyossef.com> · 2011-10-23
[PATCH v2 3/6] tile: Move tile to use generic on_each_cpu_mask · Gilad Ben-Yossef <gilad@benyossef.com> · 2011-10-23
[PATCH v2 4/6] mm: Only IPI CPUs to drain local pages if they exist · Gilad Ben-Yossef <gilad@benyossef.com> · 2011-10-23
Re: [PATCH v2 4/6] mm: Only IPI CPUs to drain local pages if they exist · Rik van Riel <hidden> · 2011-10-28
Re: [PATCH v2 4/6] mm: Only IPI CPUs to drain local pages if they exist · Gilad Ben-Yossef <gilad@benyossef.com> · 2011-10-29
Re: [PATCH v2 4/6] mm: Only IPI CPUs to drain local pages if they exist · Christoph Lameter <cl@gentwo.org> · 2011-11-02
Re: [PATCH v2 4/6] mm: Only IPI CPUs to drain local pages if they exist · Gilad Ben-Yossef <gilad@benyossef.com> · 2011-11-10
[PATCH v2 5/6] slub: Only IPI CPUs that have per cpu obj to flush · Gilad Ben-Yossef <gilad@benyossef.com> · 2011-10-23
Re: [PATCH v2 5/6] slub: Only IPI CPUs that have per cpu obj to flush · Christoph Lameter <cl@gentwo.org> · 2011-10-28
Re: [PATCH v2 5/6] slub: Only IPI CPUs that have per cpu obj to flush · Gilad Ben-Yossef <gilad@benyossef.com> · 2011-10-28
[PATCH v2 6/6] slub: only preallocate cpus_with_slabs if offstack · Gilad Ben-Yossef <gilad@benyossef.com> · 2011-10-23
Re: [PATCH v2 6/6] slub: only preallocate cpus_with_slabs if offstack · Christoph Lameter <cl@gentwo.org> · 2011-10-28
Re: [PATCH v2 6/6] slub: only preallocate cpus_with_slabs if offstack · Gilad Ben-Yossef <gilad@benyossef.com> · 2011-10-28
Re: [PATCH v2 6/6] slub: only preallocate cpus_with_slabs if offstack · Christoph Lameter <cl@gentwo.org> · 2011-11-02
Re: [PATCH v2 6/6] slub: only preallocate cpus_with_slabs if offstack · Pekka Enberg <penberg@kernel.org> · 2011-11-06
Re: [PATCH v2 6/6] slub: only preallocate cpus_with_slabs if offstack · Gilad Ben-Yossef <gilad@benyossef.com> · 2011-11-13

From: Christoph Lameter <cl@gentwo.org>
Date: 2011-11-02 08:53:47
Also in: lkml

On Sat, 29 Oct 2011, Gilad Ben-Yossef wrote:

quoted

+/* Which CPUs have per cpu pages  */
+cpumask_var_t cpus_with_pcp;
+static DEFINE_PER_CPU(unsigned long, total_cpu_pcp_count);

Does the flushing happen so frequently that it is worth keeping this
state on a per-cpu basis, or would it be better to check each CPU's
pcp info and assemble a cpumask at flush time like done in patch 5?

No, I don't  believe it is frequent at all. I will try to re-work the
patch as suggested.

The draining of the pcp pages is done from the vmstat callback which
occurs every second. Only if there is something to clean in the caches
will the flush happen.

`h`	back out one level
`j`	next message in thread
`k`	previous message in thread
`l`	drill in
`Esc`	close help / fold thread tree
`?`	toggle this help