linux-mm.kvack.org archive mirror
 help / color / mirror / Atom feed
* [PATCH 1/2] mm/percpu.c: renew the max_contig if we merge the head and previous block.
@ 2014-03-27 11:05 Jianyu Zhan
  2014-03-27 14:57 ` Tejun Heo
  0 siblings, 1 reply; 4+ messages in thread
From: Jianyu Zhan @ 2014-03-27 11:05 UTC (permalink / raw)
  To: linux-mm; +Cc: akpm, tj, cl, linux-kernel, nasa4836

During pcpu_alloc_area(), we might merge the current head with the
previous block. Since we have calculated the max_contig using the
size of previous block before we skip it, and now we update the size
of previous block, so we should renew the max_contig.

Signed-off-by: Jianyu Zhan <nasa4836@gmail.com>
---
 mm/percpu.c | 6 ++++--
 1 file changed, 4 insertions(+), 2 deletions(-)

diff --git a/mm/percpu.c b/mm/percpu.c
index 036cfe0..cfda29c 100644
--- a/mm/percpu.c
+++ b/mm/percpu.c
@@ -506,9 +506,11 @@ static int pcpu_alloc_area(struct pcpu_chunk *chunk, int size, int align)
 		 * uncommon for percpu allocations.
 		 */
 		if (head && (head < sizeof(int) || chunk->map[i - 1] > 0)) {
-			if (chunk->map[i - 1] > 0)
+			if (chunk->map[i - 1] > 0) {
 				chunk->map[i - 1] += head;
-			else {
+				max_contig =
+					max(chunk->map[i - 1], max_contig);
+			} else {
 				chunk->map[i - 1] -= head;
 				chunk->free_size -= head;
 			}
-- 
1.8.5.3

--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@kvack.org.  For more info on Linux MM,
see: http://www.linux-mm.org/ .
Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>

^ permalink raw reply related	[flat|nested] 4+ messages in thread

* Re: [PATCH 1/2] mm/percpu.c: renew the max_contig if we merge the head and previous block.
  2014-03-27 11:05 Jianyu Zhan
@ 2014-03-27 14:57 ` Tejun Heo
  0 siblings, 0 replies; 4+ messages in thread
From: Tejun Heo @ 2014-03-27 14:57 UTC (permalink / raw)
  To: Jianyu Zhan; +Cc: linux-mm, akpm, cl, linux-kernel

On Thu, Mar 27, 2014 at 07:05:43PM +0800, Jianyu Zhan wrote:
> During pcpu_alloc_area(), we might merge the current head with the
> previous block. Since we have calculated the max_contig using the
> size of previous block before we skip it, and now we update the size
> of previous block, so we should renew the max_contig.

pcpu_alloc_area() has been reimplemented in percpu/for-3.15.  Can you
please refresh the patch against the new implementation?

Thanks.

-- 
tejun

--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@kvack.org.  For more info on Linux MM,
see: http://www.linux-mm.org/ .
Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>

^ permalink raw reply	[flat|nested] 4+ messages in thread

* [PATCH 1/2] mm/percpu.c: renew the max_contig if we merge the head and previous block
@ 2014-03-28 12:55 Jianyu Zhan
  2014-03-29 13:30 ` Tejun Heo
  0 siblings, 1 reply; 4+ messages in thread
From: Jianyu Zhan @ 2014-03-28 12:55 UTC (permalink / raw)
  To: linux-mm; +Cc: akpm, tj, cl, linux-kernel, nasa4836

Hi, tj,
I've reworked the patches on top of percpu/for-3.15.

During pcpu_alloc_area(), we might merge the current head with the
previous block. Since we have calculated the max_contig using the
size of previous block before we skip it, and now we update the size
of previous block, so we should renew the max_contig.

Signed-off-by: Jianyu Zhan <nasa4836@gmail.com>
---
 mm/percpu.c | 4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)

diff --git a/mm/percpu.c b/mm/percpu.c
index 202e104..63e24fb 100644
--- a/mm/percpu.c
+++ b/mm/percpu.c
@@ -473,9 +473,11 @@ static int pcpu_alloc_area(struct pcpu_chunk *chunk, int size, int align)
 		 * uncommon for percpu allocations.
 		 */
 		if (head && (head < sizeof(int) || !(p[-1] & 1))) {
+			*p = off += head;
 			if (p[-1] & 1)
 				chunk->free_size -= head;
-			*p = off += head;
+			else
+				max_contig = max(*p - p[-1], max_contig);
 			this_size -= head;
 			head = 0;
 		}
-- 
1.8.5.3

--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@kvack.org.  For more info on Linux MM,
see: http://www.linux-mm.org/ .
Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>

^ permalink raw reply related	[flat|nested] 4+ messages in thread

* Re: [PATCH 1/2] mm/percpu.c: renew the max_contig if we merge the head and previous block
  2014-03-28 12:55 [PATCH 1/2] mm/percpu.c: renew the max_contig if we merge the head and previous block Jianyu Zhan
@ 2014-03-29 13:30 ` Tejun Heo
  0 siblings, 0 replies; 4+ messages in thread
From: Tejun Heo @ 2014-03-29 13:30 UTC (permalink / raw)
  To: Jianyu Zhan; +Cc: linux-mm, akpm, cl, linux-kernel

On Fri, Mar 28, 2014 at 08:55:21PM +0800, Jianyu Zhan wrote:
> Hi, tj,
> I've reworked the patches on top of percpu/for-3.15.
> 
> During pcpu_alloc_area(), we might merge the current head with the
> previous block. Since we have calculated the max_contig using the
> size of previous block before we skip it, and now we update the size
> of previous block, so we should renew the max_contig.
> 
> Signed-off-by: Jianyu Zhan <nasa4836@gmail.com>

Applied to percpu/for-3.15.

Thanks!

-- 
tejun

--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@kvack.org.  For more info on Linux MM,
see: http://www.linux-mm.org/ .
Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>

^ permalink raw reply	[flat|nested] 4+ messages in thread

end of thread, other threads:[~2014-03-29 13:30 UTC | newest]

Thread overview: 4+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2014-03-28 12:55 [PATCH 1/2] mm/percpu.c: renew the max_contig if we merge the head and previous block Jianyu Zhan
2014-03-29 13:30 ` Tejun Heo
  -- strict thread matches above, loose matches on Subject: below --
2014-03-27 11:05 Jianyu Zhan
2014-03-27 14:57 ` Tejun Heo

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).