On Thu, Sep 22, 2016 at 02:45:46PM +0900, Joonsoo Kim wrote:

On Wed, Sep 21, 2016 at 11:20:11AM +0200, Vlastimil Babka wrote:

quoted

On 08/29/2016 07:07 AM, js1304@gmail.com wrote:

quoted

From: Joonsoo Kim <redacted>

Until now, reserved pages for CMA are managed in the ordinary zones
where page's pfn are belong to. This approach has numorous problems
and fixing them isn't easy. (It is mentioned on previous patch.)
To fix this situation, ZONE_CMA is introduced in previous patch, but,
not yet populated. This patch implement population of ZONE_CMA
by stealing reserved pages from the ordinary zones.

Unlike previous implementation that kernel allocation request with
__GFP_MOVABLE could be serviced from CMA region, allocation request only
with GFP_HIGHUSER_MOVABLE can be serviced from CMA region in the new
approach. This is an inevitable design decision to use the zone
implementation because ZONE_CMA could contain highmem. Due to this
decision, ZONE_CMA will work like as ZONE_HIGHMEM or ZONE_MOVABLE.

I don't think it would be a problem because most of file cache pages
and anonymous pages are requested with GFP_HIGHUSER_MOVABLE. It could
be proved by the fact that there are many systems with ZONE_HIGHMEM and
they work fine. Notable disadvantage is that we cannot use these pages
for blockdev file cache page, because it usually has __GFP_MOVABLE but
not __GFP_HIGHMEM and __GFP_USER. But, in this case, there is pros and
cons. In my experience, blockdev file cache pages are one of the top
reason that causes cma_alloc() to fail temporarily. So, we can get more
guarantee of cma_alloc() success by discarding that case.

Implementation itself is very easy to understand. Steal when cma area is
initialized and recalculate various per zone stat/threshold.

Signed-off-by: Joonsoo Kim <redacted>

...

quoted

@@ -145,6 +145,28 @@ err:

static int __init cma_init_reserved_areas(void)
{
	int i;
+	struct zone *zone;
+	unsigned long start_pfn = UINT_MAX, end_pfn = 0;
+
+	if (!cma_area_count)
+		return 0;
+
+	for (i = 0; i < cma_area_count; i++) {
+		if (start_pfn > cma_areas[i].base_pfn)
+			start_pfn = cma_areas[i].base_pfn;
+		if (end_pfn < cma_areas[i].base_pfn + cma_areas[i].count)
+			end_pfn = cma_areas[i].base_pfn + cma_areas[i].count;
+	}
+
+	for_each_zone(zone) {
+		if (!is_zone_cma(zone))
+			continue;
+
+		/* ZONE_CMA doesn't need to exceed CMA region */
+		zone->zone_start_pfn = max(zone->zone_start_pfn, start_pfn);
+		zone->spanned_pages = min(zone_end_pfn(zone), end_pfn) -
+					zone->zone_start_pfn;
+	}

Hmm, so what happens on a system with multiple nodes? Each will have
its own ZONE_CMA, and all will have the same start pfn and spanned
pages?

Each of zone_start_pfn and spanned_pages are initialized in
calculate_node_totalpages() which considers node boundary. So, they will
have not the same start pfn and spanned pages. However, each would
contain unnecessary holes.

quoted

quoted

/* Free whole pageblock and set its migration type to MIGRATE_CMA. */
void __init init_cma_reserved_pageblock(struct page *page)
{
	unsigned i = pageblock_nr_pages;
+	unsigned long pfn = page_to_pfn(page);
	struct page *p = page;
+	int nid = page_to_nid(page);
+
+	/*
+	 * ZONE_CMA will steal present pages from other zones by changing
+	 * page links so page_zone() is changed. Before that,
+	 * we need to adjust previous zone's page count first.
+	 */
+	adjust_present_page_count(page, -pageblock_nr_pages);

	do {
		__ClearPageReserved(p);
		set_page_count(p, 0);
-	} while (++p, --i);
+
+		/* Steal pages from other zones */
+		set_page_links(p, ZONE_CMA, nid, pfn);
+	} while (++p, ++pfn, --i);
+
+	adjust_present_page_count(page, pageblock_nr_pages);

This seems to assign pages to ZONE_CMA on the proper node, which is
good. But then ZONE_CMA on multiple nodes will have unnecessary
holes in the spanned pages, as each will contain only a subset.

True, I will fix it and respin the series.

I now realize that it's too late to send full series for next
merge window. I will send full series after next merge window is closed.

Anyway, I'd like to confirm that following incremental patch will solve
your concern.

Thanks.


------>8--------------
 mm/cma.c | 25 ++++++++++++++++---------
 1 file changed, 16 insertions(+), 9 deletions(-)

diff --git a/mm/cma.c b/mm/cma.c
index d69bdf7..8375554 100644
--- a/mm/cma.c
+++ b/mm/cma.c

@@ -146,22 +146,29 @@ static int __init cma_init_reserved_areas(void)
 {
        int i;
        struct zone *zone;
-       unsigned long start_pfn = UINT_MAX, end_pfn = 0;
+       pg_data_t *pgdat;
 
        if (!cma_area_count)
                return 0;
 
-       for (i = 0; i < cma_area_count; i++) {
-               if (start_pfn > cma_areas[i].base_pfn)
-                       start_pfn = cma_areas[i].base_pfn;
-               if (end_pfn < cma_areas[i].base_pfn + cma_areas[i].count)
-                       end_pfn = cma_areas[i].base_pfn + cma_areas[i].count;
-       }
+       for_each_online_pgdat(pgdat) {
+               unsigned long start_pfn = UINT_MAX, end_pfn = 0;
 
-       for_each_zone(zone) {
-               if (!is_zone_cma(zone))
+               for (i = 0; i < cma_area_count; i++) {
+                       if (page_to_nid(pfn_to_page(cma_areas[i].base_pfn)) !=
+                               pgdat->node_id)
+                               continue;
+
+                       start_pfn = min(start_pfn, cma_areas[i].base_pfn);
+                       end_pfn = max(end_pfn, cma_areas[i].base_pfn +
+                                               cma_areas[i].count);
+               }
+
+               if (!end_pfn)
                        continue;
 
+               zone = &pgdat->node_zones[ZONE_CMA];
+
                /* ZONE_CMA doesn't need to exceed CMA region */
                zone->zone_start_pfn = max(zone->zone_start_pfn, start_pfn);
                zone->spanned_pages = min(zone_end_pfn(zone), end_pfn) -

--

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>