[PATCH 0/2] Optimization of CMA allocation and reclaiming

[PATCH 0/2] Optimization of CMA allocation and reclaiming

[zhaoyang.huang]

From: Zhaoyang Huang <zhaoyang.huang@unisoc.com>

These two commits help to solve bellowing OOM issue which all free pages are
composed of CMA.

04166 < 4> [   36.172486] [03-19 10:05:52.172] ActivityManager: page allocation failure: order:0, mode:0xc00(GFP_NOIO), nodemask=(null),cpuset=foreground,mems_allowed=0
0419C < 4> [   36.189447] [03-19 10:05:52.189] DMA32: 0*4kB 447*8kB (C) 217*16kB (C) 124*32kB (C) 136*64kB (C) 70*128kB (C) 22*256kB (C) 3*512kB (C) 0*1024kB 0*2048kB 0*4096kB = 35848kB
0419D < 4> [   36.193125] [03-19 10:05:52.193] Normal: 231*4kB (UMEH) 49*8kB (MEH) 14*16kB (H) 13*32kB (H) 8*64kB (H) 2*128kB (H) 0*256kB 1*512kB (H) 0*1024kB 0*2048kB 0*4096kB = 3236kB
	......
041EA < 4> [   36.234447] [03-19 10:05:52.234] SLUB: Unable to allocate memory on node -1, gfp=0xa20(GFP_ATOMIC)
041EB < 4> [   36.234455] [03-19 10:05:52.234] cache: ext4_io_end, object size: 64, buffer size: 64, default order: 0, min order: 0
041EC < 4> [   36.234459] [03-19 10:05:52.234] node 0: slabs: 53,objs: 3392, free: 0

Zhaoyang Huang (2):
  mm: optimization on page allocation when CMA enabled
  mm: skip CMA pages when they are not available

 mm/page_alloc.c | 46 ++++++++++++++++++++++++++++++++++++++++------
 mm/vmscan.c     | 15 +++++++++++++--
 2 files changed, 53 insertions(+), 8 deletions(-)

-- 
1.9.1

[PATCH 1/2] mm: optimization on page allocation when CMA enabled

[zhaoyang.huang]

From: Zhaoyang Huang <zhaoyang.huang@unisoc.com>

Let us look at the series of scenarios below with WMARK_LOW=25MB,WMARK_MIN=5MB
(managed pages 1.9GB). We can know that current 'fixed 1/2 ratio' start to use
CMA since C which actually has caused U&R lower than WMARK_LOW (this should be
deemed as against current memory policy, that is, UNMOVABLE & RECLAIMABLE should
either stay around WATERMARK_LOW when no allocation or do reclaim via entering
slowpath)

-- Free_pages
|
|
-- WMARK_LOW
|
-- Free_CMA
|
|
--

Free_CMA/Free_pages(MB)      A(12/30)     B(12/25)     C(12/20)
fixed 1/2 ratio                 N             N           Y
this commit                     Y             Y           Y

Signed-off-by: Zhaoyang Huang <zhaoyang.huang@unisoc.com>
---
v2: do proportion check when zone_watermark_ok, update commit message
v3: update coding style and simplify the logic when zone_watermark_ok
---
---
 mm/page_alloc.c | 46 ++++++++++++++++++++++++++++++++++++++++------
 1 file changed, 40 insertions(+), 6 deletions(-)

diff --git a/mm/page_alloc.c b/mm/page_alloc.c
index 0745aed..7aca49d 100644
--- a/mm/page_alloc.c
+++ b/mm/page_alloc.c
@@ -3071,6 +3071,41 @@ static bool unreserve_highatomic_pageblock(const struct alloc_context *ac,
 
 }
 
+#ifdef CONFIG_CMA
+/*
+ * GFP_MOVABLE allocation could drain UNMOVABLE & RECLAIMABLE page blocks via
+ * the help of CMA which makes GFP_KERNEL failed. Checking if zone_watermark_ok
+ * again without ALLOC_CMA to see if to use CMA first.
+ */
+static bool __if_use_cma_first(struct zone *zone, unsigned int order, unsigned int alloc_flags)
+{
+	unsigned long watermark;
+	bool cma_first = false;
+
+	watermark = wmark_pages(zone, alloc_flags & ALLOC_WMARK_MASK);
+	/* check if GFP_MOVABLE pass previous zone_watermark_ok via the help of CMA */
+	if (!zone_watermark_ok(zone, order, watermark, 0, alloc_flags & (~ALLOC_CMA)))
+		/*
+		 * watermark failed means UNMOVABLE & RECLAIMBLE is not enough
+		 * now, we should use cma first to keep them stay around the
+		 * corresponding watermark
+		 */
+		cma_first = true;
+	else
+		/*
+		 * remain previous fixed 1/2 logic when watermark ok as we have
+		 * above protection now
+		 */
+		cma_first = (zone_page_state(zone, NR_FREE_CMA_PAGES) >
+				zone_page_state(zone, NR_FREE_PAGES) / 2);
+	return cma_first;
+}
+#else
+static bool __if_use_cma_first(struct zone *zone, unsigned int order, unsigned int alloc_flags)
+{
+	return false;
+}
+#endif
 /*
  * Do the hard work of removing an element from the buddy allocator.
  * Call me with the zone->lock already held.
@@ -3084,13 +3119,12 @@ static bool unreserve_highatomic_pageblock(const struct alloc_context *ac,
 	if (IS_ENABLED(CONFIG_CMA)) {
 		/*
 		 * Balance movable allocations between regular and CMA areas by
-		 * allocating from CMA when over half of the zone's free memory
-		 * is in the CMA area.
+		 * allocating from CMA base on judging zone_watermark_ok again
+		 * to see if the latest check got pass via the help of CMA
 		 */
-		if (alloc_flags & ALLOC_CMA &&
-		    zone_page_state(zone, NR_FREE_CMA_PAGES) >
-		    zone_page_state(zone, NR_FREE_PAGES) / 2) {
-			page = __rmqueue_cma_fallback(zone, order);
+		if (migratetype == MIGRATE_MOVABLE) {
+			page = __if_use_cma_first(zone, order, alloc_flags) ?
+				__rmqueue_cma_fallback(zone, order) : NULL;
 			if (page)
 				return page;
 		}
-- 
1.9.1

[PATCH 2/2] mm: skip CMA pages when they are not available

[zhaoyang.huang]

From: Zhaoyang Huang <zhaoyang.huang@unisoc.com>

This patch fixes unproductive reclaiming of CMA pages by skipping them when they
are not available for current context. It is arise from bellowing OOM issue, which
caused by large proportion of MIGRATE_CMA pages among free pages. There has been
commit(168676649) to fix it by trying CMA pages first instead of fallback in
rmqueue.

04166 < 4> [   36.172486] [03-19 10:05:52.172] ActivityManager: page allocation failure: order:0, mode:0xc00(GFP_NOIO), nodemask=(null),cpuset=foreground,mems_allowed=0
0419C < 4> [   36.189447] [03-19 10:05:52.189] DMA32: 0*4kB 447*8kB (C) 217*16kB (C) 124*32kB (C) 136*64kB (C) 70*128kB (C) 22*256kB (C) 3*512kB (C) 0*1024kB 0*2048kB 0*4096kB = 35848kB
0419D < 4> [   36.193125] [03-19 10:05:52.193] Normal: 231*4kB (UMEH) 49*8kB (MEH) 14*16kB (H) 13*32kB (H) 8*64kB (H) 2*128kB (H) 0*256kB 1*512kB (H) 0*1024kB 0*2048kB 0*4096kB = 3236kB
	......
041EA < 4> [   36.234447] [03-19 10:05:52.234] SLUB: Unable to allocate memory on node -1, gfp=0xa20(GFP_ATOMIC)
041EB < 4> [   36.234455] [03-19 10:05:52.234] cache: ext4_io_end, object size: 64, buffer size: 64, default order: 0, min order: 0
041EC < 4> [   36.234459] [03-19 10:05:52.234] node 0: slabs: 53,objs: 3392, free: 0

Signed-off-by: Zhaoyang Huang <zhaoyang.huang@unisoc.com>
---
 mm/vmscan.c | 15 +++++++++++++--
 1 file changed, 13 insertions(+), 2 deletions(-)

diff --git a/mm/vmscan.c b/mm/vmscan.c
index bd6637f..19fb445 100644
--- a/mm/vmscan.c
+++ b/mm/vmscan.c
@@ -2225,10 +2225,16 @@ static unsigned long isolate_lru_folios(unsigned long nr_to_scan,
 	unsigned long nr_skipped[MAX_NR_ZONES] = { 0, };
 	unsigned long skipped = 0;
 	unsigned long scan, total_scan, nr_pages;
+	bool cma_cap = true;
+	struct page *page;
 	LIST_HEAD(folios_skipped);
 
 	total_scan = 0;
 	scan = 0;
+	if ((IS_ENABLED(CONFIG_CMA)) && !current_is_kswapd()
+		&& (gfp_migratetype(sc->gfp_mask) != MIGRATE_MOVABLE))
+		cma_cap = false;
+
 	while (scan < nr_to_scan && !list_empty(src)) {
 		struct list_head *move_to = src;
 		struct folio *folio;
@@ -2239,12 +2245,17 @@ static unsigned long isolate_lru_folios(unsigned long nr_to_scan,
 		nr_pages = folio_nr_pages(folio);
 		total_scan += nr_pages;
 
-		if (folio_zonenum(folio) > sc->reclaim_idx) {
+		page = &folio->page;
+
+		if ((folio_zonenum(folio) > sc->reclaim_idx)
+#ifdef CONFIG_CMA
+			|| (get_pageblock_migratetype(page) == MIGRATE_CMA && !cma_cap)
+#endif
+		) {
 			nr_skipped[folio_zonenum(folio)] += nr_pages;
 			move_to = &folios_skipped;
 			goto move;
 		}
-
 		/*
 		 * Do not count skipped folios because that makes the function
 		 * return with no isolated folios if the LRU mostly contains
-- 
1.9.1

Re: [PATCH 2/2] mm: skip CMA pages when they are not available

[Zhaoyang Huang]

+Matthew Wilcox

On Mon, May 8, 2023 at 5:35 PM zhaoyang.huang <zhaoyang.huang@unisoc.com> wrote:
>
> From: Zhaoyang Huang <zhaoyang.huang@unisoc.com>
>
> This patch fixes unproductive reclaiming of CMA pages by skipping them when they
> are not available for current context. It is arise from bellowing OOM issue, which
> caused by large proportion of MIGRATE_CMA pages among free pages. There has been
> commit(168676649) to fix it by trying CMA pages first instead of fallback in
> rmqueue.
>
> 04166 < 4> [   36.172486] [03-19 10:05:52.172] ActivityManager: page allocation failure: order:0, mode:0xc00(GFP_NOIO), nodemask=(null),cpuset=foreground,mems_allowed=0
> 0419C < 4> [   36.189447] [03-19 10:05:52.189] DMA32: 0*4kB 447*8kB (C) 217*16kB (C) 124*32kB (C) 136*64kB (C) 70*128kB (C) 22*256kB (C) 3*512kB (C) 0*1024kB 0*2048kB 0*4096kB = 35848kB
> 0419D < 4> [   36.193125] [03-19 10:05:52.193] Normal: 231*4kB (UMEH) 49*8kB (MEH) 14*16kB (H) 13*32kB (H) 8*64kB (H) 2*128kB (H) 0*256kB 1*512kB (H) 0*1024kB 0*2048kB 0*4096kB = 3236kB
>         ......
> 041EA < 4> [   36.234447] [03-19 10:05:52.234] SLUB: Unable to allocate memory on node -1, gfp=0xa20(GFP_ATOMIC)
> 041EB < 4> [   36.234455] [03-19 10:05:52.234] cache: ext4_io_end, object size: 64, buffer size: 64, default order: 0, min order: 0
> 041EC < 4> [   36.234459] [03-19 10:05:52.234] node 0: slabs: 53,objs: 3392, free: 0
>
> Signed-off-by: Zhaoyang Huang <zhaoyang.huang@unisoc.com>
> ---
>  mm/vmscan.c | 15 +++++++++++++--
>  1 file changed, 13 insertions(+), 2 deletions(-)
>
> diff --git a/mm/vmscan.c b/mm/vmscan.c
> index bd6637f..19fb445 100644
> --- a/mm/vmscan.c
> +++ b/mm/vmscan.c
> @@ -2225,10 +2225,16 @@ static unsigned long isolate_lru_folios(unsigned long nr_to_scan,
>         unsigned long nr_skipped[MAX_NR_ZONES] = { 0, };
>         unsigned long skipped = 0;
>         unsigned long scan, total_scan, nr_pages;
> +       bool cma_cap = true;
> +       struct page *page;
>         LIST_HEAD(folios_skipped);
>
>         total_scan = 0;
>         scan = 0;
> +       if ((IS_ENABLED(CONFIG_CMA)) && !current_is_kswapd()
> +               && (gfp_migratetype(sc->gfp_mask) != MIGRATE_MOVABLE))
> +               cma_cap = false;
> +
>         while (scan < nr_to_scan && !list_empty(src)) {
>                 struct list_head *move_to = src;
>                 struct folio *folio;
> @@ -2239,12 +2245,17 @@ static unsigned long isolate_lru_folios(unsigned long nr_to_scan,
>                 nr_pages = folio_nr_pages(folio);
>                 total_scan += nr_pages;
>
> -               if (folio_zonenum(folio) > sc->reclaim_idx) {
> +               page = &folio->page;
> +
> +               if ((folio_zonenum(folio) > sc->reclaim_idx)
> +#ifdef CONFIG_CMA
> +                       || (get_pageblock_migratetype(page) == MIGRATE_CMA && !cma_cap)
> +#endif
> +               ) {
>                         nr_skipped[folio_zonenum(folio)] += nr_pages;
>                         move_to = &folios_skipped;
>                         goto move;
>                 }
> -
>                 /*
>                  * Do not count skipped folios because that makes the function
>                  * return with no isolated folios if the LRU mostly contains
> --
> 1.9.1
>