[Devel] [PATCH rh7 10/21] ms/mm: huge_memory: use GFP_TRANSHUGE when charging huge pages

[Andrey Ryabinin]

From: Johannes Weiner <hannes at cmpxchg.org>

Transparent huge page charges prefer falling back to regular pages
rather than spending a lot of time in direct reclaim.

Desired reclaim behavior is usually declared in the gfp mask, but THP
charges use GFP_KERNEL and then rely on the fact that OOM is disabled
for THP charges, and that OOM-disabled charges don't retry reclaim.
Needless to say, this is anything but obvious and quite error prone.

Convert THP charges to use GFP_TRANSHUGE instead, which implies
__GFP_NORETRY, to indicate the low-latency requirement.

Signed-off-by: Johannes Weiner <hannes at cmpxchg.org>
Acked-by: Michal Hocko <mhocko at suse.cz>
Cc: Hugh Dickins <hughd at google.com>
Cc: Tejun Heo <tj at kernel.org>
Cc: Vladimir Davydov <vdavydov at parallels.com>
Signed-off-by: Andrew Morton <akpm at linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds at linux-foundation.org>

https://jira.sw.ru/browse/PSBM-51558
(cherry picked from commit d51d885bbb137cc8e1704e76be1846c5e0d5e8b4)
Signed-off-by: Andrey Ryabinin <aryabinin at virtuozzo.com>
---
 mm/huge_memory.c | 6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)

diff --git a/mm/huge_memory.c b/mm/huge_memory.c
index 3c23df1..daa8a144 100644
--- a/mm/huge_memory.c
+++ b/mm/huge_memory.c
@@ -716,7 +716,7 @@ static int __do_huge_pmd_anonymous_page(struct mm_struct *mm,
 
 	VM_BUG_ON_PAGE(!PageCompound(page), page);
 
-	if (unlikely(mem_cgroup_newpage_charge(page, mm, GFP_KERNEL))) {
+	if (unlikely(mem_cgroup_newpage_charge(page, mm, GFP_TRANSHUGE))) {
 		put_page(page);
 		count_vm_event(THP_FAULT_FALLBACK);
 		return VM_FAULT_FALLBACK;
@@ -1203,7 +1203,7 @@ alloc:
 		goto out;
 	}
 
-	if (unlikely(mem_cgroup_newpage_charge(new_page, mm, GFP_KERNEL))) {
+	if (unlikely(mem_cgroup_newpage_charge(new_page, mm, GFP_TRANSHUGE))) {
 		put_page(new_page);
 		if (page) {
 			split_huge_page(page);
@@ -2468,7 +2468,7 @@ static void collapse_huge_page(struct mm_struct *mm,
 	if (!new_page)
 		return;
 
-	if (unlikely(mem_cgroup_newpage_charge(new_page, mm, GFP_KERNEL)))
+	if (unlikely(mem_cgroup_newpage_charge(new_page, mm, GFP_TRANSHUGE)))
 		return;
 
 	/*
-- 
2.7.3