From: Joonsoo Kim <[email protected]>
Hello,
This patchset is follow-up of the problem reported and discussed
two years ago [1, 2]. The problem this patchset solves is related to
the classzone_idx on the NUMA system. It causes a problem when
the lowmem reserve protection exists for some zones on a node
that do not exist on other nodes.
This problem is reported two years ago, and, at that time,
the solution got general agreements [2]. But, due to my laziness,
it's not upstreamed. Now, I tried again. This patchset is based on
next-20200317 git tree.
Thanks.
Changes on v3
- improve the commit message
- remove useless accessor, ac_classzone_idx()
[1]: http://lkml.kernel.org/r/20180102063528.GG30397@yexl-desktop
[2]: http://lkml.kernel.org/r/[email protected]
Joonsoo Kim (2):
mm/page_alloc: use ac->high_zoneidx for classzone_idx
mm/page_alloc: integrate classzone_idx and high_zoneidx
include/linux/compaction.h | 9 ++--
include/linux/mmzone.h | 12 ++---
include/trace/events/compaction.h | 22 ++++----
include/trace/events/vmscan.h | 14 +++--
mm/compaction.c | 64 +++++++++++------------
mm/internal.h | 21 +++++---
mm/memory_hotplug.c | 6 +--
mm/oom_kill.c | 4 +-
mm/page_alloc.c | 60 +++++++++++-----------
mm/slab.c | 4 +-
mm/slub.c | 4 +-
mm/vmscan.c | 105 ++++++++++++++++++++------------------
12 files changed, 175 insertions(+), 150 deletions(-)
--
2.7.4
From: Joonsoo Kim <[email protected]>
Currently, we use the zone index of preferred_zone which represents
the best matching zone for allocation, as classzone_idx. It has
a problem on NUMA systems when the lowmem reserve protection exists
for some zones on a node that do not exist on other nodes.
In NUMA system, it can be possible that each node has different populated
zones. For example, node 0 could have DMA/DMA32/NORMAL/MOVABLE zone and
node 1 could have only NORMAL zone. In this setup, allocation request
initiated on node 0 and the one on node 1 would have different
classzone_idx, 3 and 2, respectively, since their preferred_zones are
different. If the allocation is local, there is no problem. However,
if it is handled by the remote node due to memory shortage, the problem
would happen.
In the following setup, allocation initiated on node 1 will have some
precedence than allocation initiated on node 0 when former allocation is
processed on node 0 due to not enough memory on node 1. They will have
different lowmem reserve due to their different classzone_idx thus
an watermark bars are also different.
root@ubuntu:/sys/devices/system/memory# cat /proc/zoneinfo
Node 0, zone DMA
per-node stats
...
pages free 3965
min 5
low 8
high 11
spanned 4095
present 3998
managed 3977
protection: (0, 2961, 4928, 5440)
...
Node 0, zone DMA32
pages free 757955
min 1129
low 1887
high 2645
spanned 1044480
present 782303
managed 758116
protection: (0, 0, 1967, 2479)
...
Node 0, zone Normal
pages free 459806
min 750
low 1253
high 1756
spanned 524288
present 524288
managed 503620
protection: (0, 0, 0, 4096)
...
Node 0, zone Movable
pages free 130759
min 195
low 326
high 457
spanned 1966079
present 131072
managed 131072
protection: (0, 0, 0, 0)
...
Node 1, zone DMA
pages free 0
min 0
low 0
high 0
spanned 0
present 0
managed 0
protection: (0, 0, 1006, 1006)
Node 1, zone DMA32
pages free 0
min 0
low 0
high 0
spanned 0
present 0
managed 0
protection: (0, 0, 1006, 1006)
Node 1, zone Normal
per-node stats
...
pages free 233277
min 383
low 640
high 897
spanned 262144
present 262144
managed 257744
protection: (0, 0, 0, 0)
...
Node 1, zone Movable
pages free 0
min 0
low 0
high 0
spanned 262144
present 0
managed 0
protection: (0, 0, 0, 0)
min watermark for NORMAL zone on node 0
allocation initiated on node 0: 750 + 4096 = 4846
allocation initiated on node 1: 750 + 0 = 750
This watermark difference could cause too many numa_miss allocation
in some situation and then performance could be downgraded.
Recently, there was a regression report about this problem on CMA patches
since CMA memory are placed in ZONE_MOVABLE by those patches. I checked
that problem is disappeared with this fix that uses high_zoneidx
for classzone_idx.
http://lkml.kernel.org/r/20180102063528.GG30397@yexl-desktop
Using high_zoneidx for classzone_idx is more consistent way than previous
approach because system's memory layout doesn't affect anything to it.
With this patch, both classzone_idx on above example will be 3 so will
have the same min watermark.
allocation initiated on node 0: 750 + 4096 = 4846
allocation initiated on node 1: 750 + 4096 = 4846
One could wonder if there is a side effect that allocation initiated on
node 1 will use higher bar when allocation is handled on local since
classzone_idx could be higher than before. It will not happen because
the zone without managed page doesn't contributes lowmem_reserve at all.
Reported-by: Ye Xiaolong <[email protected]>
Tested-by: Ye Xiaolong <[email protected]>
Acked-by: Vlastimil Babka <[email protected]>
Signed-off-by: Joonsoo Kim <[email protected]>
---
mm/internal.h | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/mm/internal.h b/mm/internal.h
index c39c895..aebaa33 100644
--- a/mm/internal.h
+++ b/mm/internal.h
@@ -119,7 +119,7 @@ struct alloc_context {
bool spread_dirty_pages;
};
-#define ac_classzone_idx(ac) zonelist_zone_idx(ac->preferred_zoneref)
+#define ac_classzone_idx(ac) (ac->high_zoneidx)
/*
* Locate the struct page for both the matching buddy in our
--
2.7.4
From: Joonsoo Kim <[email protected]>
classzone_idx is just different name for high_zoneidx now.
So, integrate them and add some comment to struct alloc_context
in order to reduce future confusion about the meaning of this variable.
The accessor, ac_classzone_idx() is also removed since it isn't needed
after integration.
In addition to integration, this patch also renames high_zoneidx
to highest_zoneidx since it represents more precise meaning.
Signed-off-by: Joonsoo Kim <[email protected]>
---
include/linux/compaction.h | 9 ++--
include/linux/mmzone.h | 12 ++---
include/trace/events/compaction.h | 22 ++++----
include/trace/events/vmscan.h | 14 +++--
mm/compaction.c | 64 +++++++++++------------
mm/internal.h | 21 +++++---
mm/memory_hotplug.c | 6 +--
mm/oom_kill.c | 4 +-
mm/page_alloc.c | 60 +++++++++++-----------
mm/slab.c | 4 +-
mm/slub.c | 4 +-
mm/vmscan.c | 105 ++++++++++++++++++++------------------
12 files changed, 175 insertions(+), 150 deletions(-)
diff --git a/include/linux/compaction.h b/include/linux/compaction.h
index 4b898cd..3ed2f22 100644
--- a/include/linux/compaction.h
+++ b/include/linux/compaction.h
@@ -97,7 +97,7 @@ extern enum compact_result try_to_compact_pages(gfp_t gfp_mask,
struct page **page);
extern void reset_isolation_suitable(pg_data_t *pgdat);
extern enum compact_result compaction_suitable(struct zone *zone, int order,
- unsigned int alloc_flags, int classzone_idx);
+ unsigned int alloc_flags, int highest_zoneidx);
extern void defer_compaction(struct zone *zone, int order);
extern bool compaction_deferred(struct zone *zone, int order);
@@ -182,7 +182,7 @@ bool compaction_zonelist_suitable(struct alloc_context *ac, int order,
extern int kcompactd_run(int nid);
extern void kcompactd_stop(int nid);
-extern void wakeup_kcompactd(pg_data_t *pgdat, int order, int classzone_idx);
+extern void wakeup_kcompactd(pg_data_t *pgdat, int order, int highest_zoneidx);
#else
static inline void reset_isolation_suitable(pg_data_t *pgdat)
@@ -190,7 +190,7 @@ static inline void reset_isolation_suitable(pg_data_t *pgdat)
}
static inline enum compact_result compaction_suitable(struct zone *zone, int order,
- int alloc_flags, int classzone_idx)
+ int alloc_flags, int highest_zoneidx)
{
return COMPACT_SKIPPED;
}
@@ -232,7 +232,8 @@ static inline void kcompactd_stop(int nid)
{
}
-static inline void wakeup_kcompactd(pg_data_t *pgdat, int order, int classzone_idx)
+static inline void wakeup_kcompactd(pg_data_t *pgdat,
+ int order, int highest_zoneidx)
{
}
diff --git a/include/linux/mmzone.h b/include/linux/mmzone.h
index f3f2648..337b5ec 100644
--- a/include/linux/mmzone.h
+++ b/include/linux/mmzone.h
@@ -698,13 +698,13 @@ typedef struct pglist_data {
struct task_struct *kswapd; /* Protected by
mem_hotplug_begin/end() */
int kswapd_order;
- enum zone_type kswapd_classzone_idx;
+ enum zone_type kswapd_highest_zoneidx;
int kswapd_failures; /* Number of 'reclaimed == 0' runs */
#ifdef CONFIG_COMPACTION
int kcompactd_max_order;
- enum zone_type kcompactd_classzone_idx;
+ enum zone_type kcompactd_highest_zoneidx;
wait_queue_head_t kcompactd_wait;
struct task_struct *kcompactd;
#endif
@@ -782,15 +782,15 @@ static inline bool pgdat_is_empty(pg_data_t *pgdat)
void build_all_zonelists(pg_data_t *pgdat);
void wakeup_kswapd(struct zone *zone, gfp_t gfp_mask, int order,
- enum zone_type classzone_idx);
+ enum zone_type highest_zoneidx);
bool __zone_watermark_ok(struct zone *z, unsigned int order, unsigned long mark,
- int classzone_idx, unsigned int alloc_flags,
+ int highest_zoneidx, unsigned int alloc_flags,
long free_pages);
bool zone_watermark_ok(struct zone *z, unsigned int order,
- unsigned long mark, int classzone_idx,
+ unsigned long mark, int highest_zoneidx,
unsigned int alloc_flags);
bool zone_watermark_ok_safe(struct zone *z, unsigned int order,
- unsigned long mark, int classzone_idx);
+ unsigned long mark, int highest_zoneidx);
enum memmap_context {
MEMMAP_EARLY,
MEMMAP_HOTPLUG,
diff --git a/include/trace/events/compaction.h b/include/trace/events/compaction.h
index e5bf6ee..54e5bf0 100644
--- a/include/trace/events/compaction.h
+++ b/include/trace/events/compaction.h
@@ -314,40 +314,44 @@ TRACE_EVENT(mm_compaction_kcompactd_sleep,
DECLARE_EVENT_CLASS(kcompactd_wake_template,
- TP_PROTO(int nid, int order, enum zone_type classzone_idx),
+ TP_PROTO(int nid, int order, enum zone_type highest_zoneidx),
- TP_ARGS(nid, order, classzone_idx),
+ TP_ARGS(nid, order, highest_zoneidx),
TP_STRUCT__entry(
__field(int, nid)
__field(int, order)
- __field(enum zone_type, classzone_idx)
+ __field(enum zone_type, highest_zoneidx)
),
TP_fast_assign(
__entry->nid = nid;
__entry->order = order;
- __entry->classzone_idx = classzone_idx;
+ __entry->highest_zoneidx = highest_zoneidx;
),
+ /*
+ * classzone_idx is previous name of the highest_zoneidx.
+ * Reason not to change it is the ABI requirement of the tracepoint.
+ */
TP_printk("nid=%d order=%d classzone_idx=%-8s",
__entry->nid,
__entry->order,
- __print_symbolic(__entry->classzone_idx, ZONE_TYPE))
+ __print_symbolic(__entry->highest_zoneidx, ZONE_TYPE))
);
DEFINE_EVENT(kcompactd_wake_template, mm_compaction_wakeup_kcompactd,
- TP_PROTO(int nid, int order, enum zone_type classzone_idx),
+ TP_PROTO(int nid, int order, enum zone_type highest_zoneidx),
- TP_ARGS(nid, order, classzone_idx)
+ TP_ARGS(nid, order, highest_zoneidx)
);
DEFINE_EVENT(kcompactd_wake_template, mm_compaction_kcompactd_wake,
- TP_PROTO(int nid, int order, enum zone_type classzone_idx),
+ TP_PROTO(int nid, int order, enum zone_type highest_zoneidx),
- TP_ARGS(nid, order, classzone_idx)
+ TP_ARGS(nid, order, highest_zoneidx)
);
#endif
diff --git a/include/trace/events/vmscan.h b/include/trace/events/vmscan.h
index a5ab297..f2b3b9c 100644
--- a/include/trace/events/vmscan.h
+++ b/include/trace/events/vmscan.h
@@ -265,7 +265,7 @@ TRACE_EVENT(mm_shrink_slab_end,
);
TRACE_EVENT(mm_vmscan_lru_isolate,
- TP_PROTO(int classzone_idx,
+ TP_PROTO(int highest_zoneidx,
int order,
unsigned long nr_requested,
unsigned long nr_scanned,
@@ -274,10 +274,10 @@ TRACE_EVENT(mm_vmscan_lru_isolate,
isolate_mode_t isolate_mode,
int lru),
- TP_ARGS(classzone_idx, order, nr_requested, nr_scanned, nr_skipped, nr_taken, isolate_mode, lru),
+ TP_ARGS(highest_zoneidx, order, nr_requested, nr_scanned, nr_skipped, nr_taken, isolate_mode, lru),
TP_STRUCT__entry(
- __field(int, classzone_idx)
+ __field(int, highest_zoneidx)
__field(int, order)
__field(unsigned long, nr_requested)
__field(unsigned long, nr_scanned)
@@ -288,7 +288,7 @@ TRACE_EVENT(mm_vmscan_lru_isolate,
),
TP_fast_assign(
- __entry->classzone_idx = classzone_idx;
+ __entry->highest_zoneidx = highest_zoneidx;
__entry->order = order;
__entry->nr_requested = nr_requested;
__entry->nr_scanned = nr_scanned;
@@ -298,9 +298,13 @@ TRACE_EVENT(mm_vmscan_lru_isolate,
__entry->lru = lru;
),
+ /*
+ * classzone is previous name of the highest_zoneidx.
+ * Reason not to change it is the ABI requirement of the tracepoint.
+ */
TP_printk("isolate_mode=%d classzone=%d order=%d nr_requested=%lu nr_scanned=%lu nr_skipped=%lu nr_taken=%lu lru=%s",
__entry->isolate_mode,
- __entry->classzone_idx,
+ __entry->highest_zoneidx,
__entry->order,
__entry->nr_requested,
__entry->nr_scanned,
diff --git a/mm/compaction.c b/mm/compaction.c
index 827d8a2..80d6424 100644
--- a/mm/compaction.c
+++ b/mm/compaction.c
@@ -1966,7 +1966,7 @@ static enum compact_result compact_finished(struct compact_control *cc)
*/
static enum compact_result __compaction_suitable(struct zone *zone, int order,
unsigned int alloc_flags,
- int classzone_idx,
+ int highest_zoneidx,
unsigned long wmark_target)
{
unsigned long watermark;
@@ -1979,7 +1979,7 @@ static enum compact_result __compaction_suitable(struct zone *zone, int order,
* If watermarks for high-order allocation are already met, there
* should be no need for compaction at all.
*/
- if (zone_watermark_ok(zone, order, watermark, classzone_idx,
+ if (zone_watermark_ok(zone, order, watermark, highest_zoneidx,
alloc_flags))
return COMPACT_SUCCESS;
@@ -1989,9 +1989,9 @@ static enum compact_result __compaction_suitable(struct zone *zone, int order,
* watermark and alloc_flags have to match, or be more pessimistic than
* the check in __isolate_free_page(). We don't use the direct
* compactor's alloc_flags, as they are not relevant for freepage
- * isolation. We however do use the direct compactor's classzone_idx to
- * skip over zones where lowmem reserves would prevent allocation even
- * if compaction succeeds.
+ * isolation. We however do use the direct compactor's highest_zoneidx
+ * to skip over zones where lowmem reserves would prevent allocation
+ * even if compaction succeeds.
* For costly orders, we require low watermark instead of min for
* compaction to proceed to increase its chances.
* ALLOC_CMA is used, as pages in CMA pageblocks are considered
@@ -2000,7 +2000,7 @@ static enum compact_result __compaction_suitable(struct zone *zone, int order,
watermark = (order > PAGE_ALLOC_COSTLY_ORDER) ?
low_wmark_pages(zone) : min_wmark_pages(zone);
watermark += compact_gap(order);
- if (!__zone_watermark_ok(zone, 0, watermark, classzone_idx,
+ if (!__zone_watermark_ok(zone, 0, watermark, highest_zoneidx,
ALLOC_CMA, wmark_target))
return COMPACT_SKIPPED;
@@ -2009,12 +2009,12 @@ static enum compact_result __compaction_suitable(struct zone *zone, int order,
enum compact_result compaction_suitable(struct zone *zone, int order,
unsigned int alloc_flags,
- int classzone_idx)
+ int highest_zoneidx)
{
enum compact_result ret;
int fragindex;
- ret = __compaction_suitable(zone, order, alloc_flags, classzone_idx,
+ ret = __compaction_suitable(zone, order, alloc_flags, highest_zoneidx,
zone_page_state(zone, NR_FREE_PAGES));
/*
* fragmentation index determines if allocation failures are due to
@@ -2055,8 +2055,8 @@ bool compaction_zonelist_suitable(struct alloc_context *ac, int order,
* Make sure at least one zone would pass __compaction_suitable if we continue
* retrying the reclaim.
*/
- for_each_zone_zonelist_nodemask(zone, z, ac->zonelist, ac->high_zoneidx,
- ac->nodemask) {
+ for_each_zone_zonelist_nodemask(zone, z, ac->zonelist,
+ ac->highest_zoneidx, ac->nodemask) {
unsigned long available;
enum compact_result compact_result;
@@ -2069,7 +2069,7 @@ bool compaction_zonelist_suitable(struct alloc_context *ac, int order,
available = zone_reclaimable_pages(zone) / order;
available += zone_page_state_snapshot(zone, NR_FREE_PAGES);
compact_result = __compaction_suitable(zone, order, alloc_flags,
- ac_classzone_idx(ac), available);
+ ac->highest_zoneidx, available);
if (compact_result != COMPACT_SKIPPED)
return true;
}
@@ -2100,7 +2100,7 @@ compact_zone(struct compact_control *cc, struct capture_control *capc)
cc->migratetype = gfpflags_to_migratetype(cc->gfp_mask);
ret = compaction_suitable(cc->zone, cc->order, cc->alloc_flags,
- cc->classzone_idx);
+ cc->highest_zoneidx);
/* Compaction is likely to fail */
if (ret == COMPACT_SUCCESS || ret == COMPACT_SKIPPED)
return ret;
@@ -2296,7 +2296,7 @@ compact_zone(struct compact_control *cc, struct capture_control *capc)
static enum compact_result compact_zone_order(struct zone *zone, int order,
gfp_t gfp_mask, enum compact_priority prio,
- unsigned int alloc_flags, int classzone_idx,
+ unsigned int alloc_flags, int highest_zoneidx,
struct page **capture)
{
enum compact_result ret;
@@ -2308,7 +2308,7 @@ static enum compact_result compact_zone_order(struct zone *zone, int order,
.mode = (prio == COMPACT_PRIO_ASYNC) ?
MIGRATE_ASYNC : MIGRATE_SYNC_LIGHT,
.alloc_flags = alloc_flags,
- .classzone_idx = classzone_idx,
+ .highest_zoneidx = highest_zoneidx,
.direct_compaction = true,
.whole_zone = (prio == MIN_COMPACT_PRIORITY),
.ignore_skip_hint = (prio == MIN_COMPACT_PRIORITY),
@@ -2364,8 +2364,8 @@ enum compact_result try_to_compact_pages(gfp_t gfp_mask, unsigned int order,
trace_mm_compaction_try_to_compact_pages(order, gfp_mask, prio);
/* Compact each zone in the list */
- for_each_zone_zonelist_nodemask(zone, z, ac->zonelist, ac->high_zoneidx,
- ac->nodemask) {
+ for_each_zone_zonelist_nodemask(zone, z, ac->zonelist,
+ ac->highest_zoneidx, ac->nodemask) {
enum compact_result status;
if (prio > MIN_COMPACT_PRIORITY
@@ -2375,7 +2375,7 @@ enum compact_result try_to_compact_pages(gfp_t gfp_mask, unsigned int order,
}
status = compact_zone_order(zone, order, gfp_mask, prio,
- alloc_flags, ac_classzone_idx(ac), capture);
+ alloc_flags, ac->highest_zoneidx, capture);
rc = max(status, rc);
/* The allocation should succeed, stop compacting */
@@ -2510,16 +2510,16 @@ static bool kcompactd_node_suitable(pg_data_t *pgdat)
{
int zoneid;
struct zone *zone;
- enum zone_type classzone_idx = pgdat->kcompactd_classzone_idx;
+ enum zone_type highest_zoneidx = pgdat->kcompactd_highest_zoneidx;
- for (zoneid = 0; zoneid <= classzone_idx; zoneid++) {
+ for (zoneid = 0; zoneid <= highest_zoneidx; zoneid++) {
zone = &pgdat->node_zones[zoneid];
if (!populated_zone(zone))
continue;
if (compaction_suitable(zone, pgdat->kcompactd_max_order, 0,
- classzone_idx) == COMPACT_CONTINUE)
+ highest_zoneidx) == COMPACT_CONTINUE)
return true;
}
@@ -2537,16 +2537,16 @@ static void kcompactd_do_work(pg_data_t *pgdat)
struct compact_control cc = {
.order = pgdat->kcompactd_max_order,
.search_order = pgdat->kcompactd_max_order,
- .classzone_idx = pgdat->kcompactd_classzone_idx,
+ .highest_zoneidx = pgdat->kcompactd_highest_zoneidx,
.mode = MIGRATE_SYNC_LIGHT,
.ignore_skip_hint = false,
.gfp_mask = GFP_KERNEL,
};
trace_mm_compaction_kcompactd_wake(pgdat->node_id, cc.order,
- cc.classzone_idx);
+ cc.highest_zoneidx);
count_compact_event(KCOMPACTD_WAKE);
- for (zoneid = 0; zoneid <= cc.classzone_idx; zoneid++) {
+ for (zoneid = 0; zoneid <= cc.highest_zoneidx; zoneid++) {
int status;
zone = &pgdat->node_zones[zoneid];
@@ -2595,16 +2595,16 @@ static void kcompactd_do_work(pg_data_t *pgdat)
/*
* Regardless of success, we are done until woken up next. But remember
- * the requested order/classzone_idx in case it was higher/tighter than
- * our current ones
+ * the requested order/highest_zoneidx in case it was higher/tighter
+ * than our current ones
*/
if (pgdat->kcompactd_max_order <= cc.order)
pgdat->kcompactd_max_order = 0;
- if (pgdat->kcompactd_classzone_idx >= cc.classzone_idx)
- pgdat->kcompactd_classzone_idx = pgdat->nr_zones - 1;
+ if (pgdat->kcompactd_highest_zoneidx >= cc.highest_zoneidx)
+ pgdat->kcompactd_highest_zoneidx = pgdat->nr_zones - 1;
}
-void wakeup_kcompactd(pg_data_t *pgdat, int order, int classzone_idx)
+void wakeup_kcompactd(pg_data_t *pgdat, int order, int highest_zoneidx)
{
if (!order)
return;
@@ -2612,8 +2612,8 @@ void wakeup_kcompactd(pg_data_t *pgdat, int order, int classzone_idx)
if (pgdat->kcompactd_max_order < order)
pgdat->kcompactd_max_order = order;
- if (pgdat->kcompactd_classzone_idx > classzone_idx)
- pgdat->kcompactd_classzone_idx = classzone_idx;
+ if (pgdat->kcompactd_highest_zoneidx > highest_zoneidx)
+ pgdat->kcompactd_highest_zoneidx = highest_zoneidx;
/*
* Pairs with implicit barrier in wait_event_freezable()
@@ -2626,7 +2626,7 @@ void wakeup_kcompactd(pg_data_t *pgdat, int order, int classzone_idx)
return;
trace_mm_compaction_wakeup_kcompactd(pgdat->node_id, order,
- classzone_idx);
+ highest_zoneidx);
wake_up_interruptible(&pgdat->kcompactd_wait);
}
@@ -2647,7 +2647,7 @@ static int kcompactd(void *p)
set_freezable();
pgdat->kcompactd_max_order = 0;
- pgdat->kcompactd_classzone_idx = pgdat->nr_zones - 1;
+ pgdat->kcompactd_highest_zoneidx = pgdat->nr_zones - 1;
while (!kthread_should_stop()) {
unsigned long pflags;
diff --git a/mm/internal.h b/mm/internal.h
index aebaa33..d84b0d9 100644
--- a/mm/internal.h
+++ b/mm/internal.h
@@ -102,10 +102,10 @@ extern pmd_t *mm_find_pmd(struct mm_struct *mm, unsigned long address);
* between functions involved in allocations, including the alloc_pages*
* family of functions.
*
- * nodemask, migratetype and high_zoneidx are initialized only once in
+ * nodemask, migratetype and highest_zoneidx are initialized only once in
* __alloc_pages_nodemask() and then never change.
*
- * zonelist, preferred_zone and classzone_idx are set first in
+ * zonelist, preferred_zone and highest_zoneidx are set first in
* __alloc_pages_nodemask() for the fast path, and might be later changed
* in __alloc_pages_slowpath(). All other functions pass the whole strucure
* by a const pointer.
@@ -115,12 +115,21 @@ struct alloc_context {
nodemask_t *nodemask;
struct zoneref *preferred_zoneref;
int migratetype;
- enum zone_type high_zoneidx;
+
+ /*
+ * highest_zoneidx represents highest usable zone index of
+ * the allocation request. Due to the nature of the zone,
+ * memory on lower zone than the highest_zoneidx will be
+ * protected by lowmem_reserve[highest_zoneidx].
+ *
+ * highest_zoneidx is also used by reclaim/compaction to limit
+ * the target zone since higher zone than this index cannot be
+ * usable for this allocation request.
+ */
+ enum zone_type highest_zoneidx;
bool spread_dirty_pages;
};
-#define ac_classzone_idx(ac) (ac->high_zoneidx)
-
/*
* Locate the struct page for both the matching buddy in our
* pair (buddy1) and the combined O(n+1) page they form (page).
@@ -199,7 +208,7 @@ struct compact_control {
int order; /* order a direct compactor needs */
int migratetype; /* migratetype of direct compactor */
const unsigned int alloc_flags; /* alloc flags of a direct compactor */
- const int classzone_idx; /* zone index of a direct compactor */
+ const int highest_zoneidx; /* zone index of a direct compactor */
enum migrate_mode mode; /* Async or sync migration mode */
bool ignore_skip_hint; /* Scan blocks even if marked skip */
bool no_set_skip_hint; /* Don't mark blocks for skipping */
diff --git a/mm/memory_hotplug.c b/mm/memory_hotplug.c
index 8bdf484..f942969 100644
--- a/mm/memory_hotplug.c
+++ b/mm/memory_hotplug.c
@@ -880,13 +880,13 @@ static pg_data_t __ref *hotadd_new_pgdat(int nid, u64 start)
} else {
int cpu;
/*
- * Reset the nr_zones, order and classzone_idx before reuse.
- * Note that kswapd will init kswapd_classzone_idx properly
+ * Reset the nr_zones, order and highest_zoneidx before reuse.
+ * Note that kswapd will init kswapd_highest_zoneidx properly
* when it starts in the near future.
*/
pgdat->nr_zones = 0;
pgdat->kswapd_order = 0;
- pgdat->kswapd_classzone_idx = 0;
+ pgdat->kswapd_highest_zoneidx = 0;
for_each_online_cpu(cpu) {
struct per_cpu_nodestat *p;
diff --git a/mm/oom_kill.c b/mm/oom_kill.c
index dfc3576..4daedf7 100644
--- a/mm/oom_kill.c
+++ b/mm/oom_kill.c
@@ -254,7 +254,7 @@ static enum oom_constraint constrained_alloc(struct oom_control *oc)
{
struct zone *zone;
struct zoneref *z;
- enum zone_type high_zoneidx = gfp_zone(oc->gfp_mask);
+ enum zone_type highest_zoneidx = gfp_zone(oc->gfp_mask);
bool cpuset_limited = false;
int nid;
@@ -294,7 +294,7 @@ static enum oom_constraint constrained_alloc(struct oom_control *oc)
/* Check this allocation failure is caused by cpuset's wall function */
for_each_zone_zonelist_nodemask(zone, z, oc->zonelist,
- high_zoneidx, oc->nodemask)
+ highest_zoneidx, oc->nodemask)
if (!cpuset_zone_allowed(zone, oc->gfp_mask))
cpuset_limited = true;
diff --git a/mm/page_alloc.c b/mm/page_alloc.c
index cb2f896..27bacee 100644
--- a/mm/page_alloc.c
+++ b/mm/page_alloc.c
@@ -2620,7 +2620,7 @@ static bool unreserve_highatomic_pageblock(const struct alloc_context *ac,
int order;
bool ret;
- for_each_zone_zonelist_nodemask(zone, z, zonelist, ac->high_zoneidx,
+ for_each_zone_zonelist_nodemask(zone, z, zonelist, ac->highest_zoneidx,
ac->nodemask) {
/*
* Preserve at least one pageblock unless memory pressure
@@ -3488,7 +3488,7 @@ ALLOW_ERROR_INJECTION(should_fail_alloc_page, TRUE);
* to check in the allocation paths if no pages are free.
*/
bool __zone_watermark_ok(struct zone *z, unsigned int order, unsigned long mark,
- int classzone_idx, unsigned int alloc_flags,
+ int highest_zoneidx, unsigned int alloc_flags,
long free_pages)
{
long min = mark;
@@ -3533,7 +3533,7 @@ bool __zone_watermark_ok(struct zone *z, unsigned int order, unsigned long mark,
* are not met, then a high-order request also cannot go ahead
* even if a suitable page happened to be free.
*/
- if (free_pages <= min + z->lowmem_reserve[classzone_idx])
+ if (free_pages <= min + z->lowmem_reserve[highest_zoneidx])
return false;
/* If this is an order-0 request then the watermark is fine */
@@ -3566,14 +3566,15 @@ bool __zone_watermark_ok(struct zone *z, unsigned int order, unsigned long mark,
}
bool zone_watermark_ok(struct zone *z, unsigned int order, unsigned long mark,
- int classzone_idx, unsigned int alloc_flags)
+ int highest_zoneidx, unsigned int alloc_flags)
{
- return __zone_watermark_ok(z, order, mark, classzone_idx, alloc_flags,
+ return __zone_watermark_ok(z, order, mark, highest_zoneidx, alloc_flags,
zone_page_state(z, NR_FREE_PAGES));
}
static inline bool zone_watermark_fast(struct zone *z, unsigned int order,
- unsigned long mark, int classzone_idx, unsigned int alloc_flags)
+ unsigned long mark, int highest_zoneidx,
+ unsigned int alloc_flags)
{
long free_pages = zone_page_state(z, NR_FREE_PAGES);
long cma_pages = 0;
@@ -3591,22 +3592,23 @@ static inline bool zone_watermark_fast(struct zone *z, unsigned int order,
* the caller is !atomic then it'll uselessly search the free
* list. That corner case is then slower but it is harmless.
*/
- if (!order && (free_pages - cma_pages) > mark + z->lowmem_reserve[classzone_idx])
+ if (!order && (free_pages - cma_pages) >
+ mark + z->lowmem_reserve[highest_zoneidx])
return true;
- return __zone_watermark_ok(z, order, mark, classzone_idx, alloc_flags,
+ return __zone_watermark_ok(z, order, mark, highest_zoneidx, alloc_flags,
free_pages);
}
bool zone_watermark_ok_safe(struct zone *z, unsigned int order,
- unsigned long mark, int classzone_idx)
+ unsigned long mark, int highest_zoneidx)
{
long free_pages = zone_page_state(z, NR_FREE_PAGES);
if (z->percpu_drift_mark && free_pages < z->percpu_drift_mark)
free_pages = zone_page_state_snapshot(z, NR_FREE_PAGES);
- return __zone_watermark_ok(z, order, mark, classzone_idx, 0,
+ return __zone_watermark_ok(z, order, mark, highest_zoneidx, 0,
free_pages);
}
@@ -3683,8 +3685,8 @@ get_page_from_freelist(gfp_t gfp_mask, unsigned int order, int alloc_flags,
*/
no_fallback = alloc_flags & ALLOC_NOFRAGMENT;
z = ac->preferred_zoneref;
- for_next_zone_zonelist_nodemask(zone, z, ac->zonelist, ac->high_zoneidx,
- ac->nodemask) {
+ for_next_zone_zonelist_nodemask(zone, z, ac->zonelist,
+ ac->highest_zoneidx, ac->nodemask) {
struct page *page;
unsigned long mark;
@@ -3739,7 +3741,7 @@ get_page_from_freelist(gfp_t gfp_mask, unsigned int order, int alloc_flags,
mark = wmark_pages(zone, alloc_flags & ALLOC_WMARK_MASK);
if (!zone_watermark_fast(zone, order, mark,
- ac_classzone_idx(ac), alloc_flags)) {
+ ac->highest_zoneidx, alloc_flags)) {
int ret;
#ifdef CONFIG_DEFERRED_STRUCT_PAGE_INIT
@@ -3772,7 +3774,7 @@ get_page_from_freelist(gfp_t gfp_mask, unsigned int order, int alloc_flags,
default:
/* did we reclaim enough */
if (zone_watermark_ok(zone, order, mark,
- ac_classzone_idx(ac), alloc_flags))
+ ac->highest_zoneidx, alloc_flags))
goto try_this_zone;
continue;
@@ -3931,7 +3933,7 @@ __alloc_pages_may_oom(gfp_t gfp_mask, unsigned int order,
if (gfp_mask & __GFP_RETRY_MAYFAIL)
goto out;
/* The OOM killer does not needlessly kill tasks for lowmem */
- if (ac->high_zoneidx < ZONE_NORMAL)
+ if (ac->highest_zoneidx < ZONE_NORMAL)
goto out;
if (pm_suspended_storage())
goto out;
@@ -4134,10 +4136,10 @@ should_compact_retry(struct alloc_context *ac, unsigned int order, int alloc_fla
* Let's give them a good hope and keep retrying while the order-0
* watermarks are OK.
*/
- for_each_zone_zonelist_nodemask(zone, z, ac->zonelist, ac->high_zoneidx,
- ac->nodemask) {
+ for_each_zone_zonelist_nodemask(zone, z, ac->zonelist,
+ ac->highest_zoneidx, ac->nodemask) {
if (zone_watermark_ok(zone, 0, min_wmark_pages(zone),
- ac_classzone_idx(ac), alloc_flags))
+ ac->highest_zoneidx, alloc_flags))
return true;
}
return false;
@@ -4261,12 +4263,12 @@ static void wake_all_kswapds(unsigned int order, gfp_t gfp_mask,
struct zoneref *z;
struct zone *zone;
pg_data_t *last_pgdat = NULL;
- enum zone_type high_zoneidx = ac->high_zoneidx;
+ enum zone_type highest_zoneidx = ac->highest_zoneidx;
- for_each_zone_zonelist_nodemask(zone, z, ac->zonelist, high_zoneidx,
+ for_each_zone_zonelist_nodemask(zone, z, ac->zonelist, highest_zoneidx,
ac->nodemask) {
if (last_pgdat != zone->zone_pgdat)
- wakeup_kswapd(zone, gfp_mask, order, high_zoneidx);
+ wakeup_kswapd(zone, gfp_mask, order, highest_zoneidx);
last_pgdat = zone->zone_pgdat;
}
}
@@ -4401,8 +4403,8 @@ should_reclaim_retry(gfp_t gfp_mask, unsigned order,
* request even if all reclaimable pages are considered then we are
* screwed and have to go OOM.
*/
- for_each_zone_zonelist_nodemask(zone, z, ac->zonelist, ac->high_zoneidx,
- ac->nodemask) {
+ for_each_zone_zonelist_nodemask(zone, z, ac->zonelist,
+ ac->highest_zoneidx, ac->nodemask) {
unsigned long available;
unsigned long reclaimable;
unsigned long min_wmark = min_wmark_pages(zone);
@@ -4416,7 +4418,7 @@ should_reclaim_retry(gfp_t gfp_mask, unsigned order,
* reclaimable pages?
*/
wmark = __zone_watermark_ok(zone, order, min_wmark,
- ac_classzone_idx(ac), alloc_flags, available);
+ ac->highest_zoneidx, alloc_flags, available);
trace_reclaim_retry_zone(z, order, reclaimable,
available, min_wmark, *no_progress_loops, wmark);
if (wmark) {
@@ -4535,7 +4537,7 @@ __alloc_pages_slowpath(gfp_t gfp_mask, unsigned int order,
* could end up iterating over non-eligible zones endlessly.
*/
ac->preferred_zoneref = first_zones_zonelist(ac->zonelist,
- ac->high_zoneidx, ac->nodemask);
+ ac->highest_zoneidx, ac->nodemask);
if (!ac->preferred_zoneref->zone)
goto nopage;
@@ -4622,7 +4624,7 @@ __alloc_pages_slowpath(gfp_t gfp_mask, unsigned int order,
if (!(alloc_flags & ALLOC_CPUSET) || reserve_flags) {
ac->nodemask = NULL;
ac->preferred_zoneref = first_zones_zonelist(ac->zonelist,
- ac->high_zoneidx, ac->nodemask);
+ ac->highest_zoneidx, ac->nodemask);
}
/* Attempt with potentially adjusted zonelist and alloc_flags */
@@ -4756,7 +4758,7 @@ static inline bool prepare_alloc_pages(gfp_t gfp_mask, unsigned int order,
struct alloc_context *ac, gfp_t *alloc_mask,
unsigned int *alloc_flags)
{
- ac->high_zoneidx = gfp_zone(gfp_mask);
+ ac->highest_zoneidx = gfp_zone(gfp_mask);
ac->zonelist = node_zonelist(preferred_nid, gfp_mask);
ac->nodemask = nodemask;
ac->migratetype = gfpflags_to_migratetype(gfp_mask);
@@ -4795,7 +4797,7 @@ static inline void finalise_ac(gfp_t gfp_mask, struct alloc_context *ac)
* may get reset for allocations that ignore memory policies.
*/
ac->preferred_zoneref = first_zones_zonelist(ac->zonelist,
- ac->high_zoneidx, ac->nodemask);
+ ac->highest_zoneidx, ac->nodemask);
}
/*
@@ -6992,7 +6994,7 @@ void __init free_area_init_node(int nid, unsigned long *zones_size,
unsigned long end_pfn = 0;
/* pg_data_t should be reset to zero when it's allocated */
- WARN_ON(pgdat->nr_zones || pgdat->kswapd_classzone_idx);
+ WARN_ON(pgdat->nr_zones || pgdat->kswapd_highest_zoneidx);
pgdat->node_id = nid;
pgdat->node_start_pfn = node_start_pfn;
diff --git a/mm/slab.c b/mm/slab.c
index a896336..9350062 100644
--- a/mm/slab.c
+++ b/mm/slab.c
@@ -3106,7 +3106,7 @@ static void *fallback_alloc(struct kmem_cache *cache, gfp_t flags)
struct zonelist *zonelist;
struct zoneref *z;
struct zone *zone;
- enum zone_type high_zoneidx = gfp_zone(flags);
+ enum zone_type highest_zoneidx = gfp_zone(flags);
void *obj = NULL;
struct page *page;
int nid;
@@ -3124,7 +3124,7 @@ static void *fallback_alloc(struct kmem_cache *cache, gfp_t flags)
* Look through allowed nodes for objects available
* from existing per node queues.
*/
- for_each_zone_zonelist(zone, z, zonelist, high_zoneidx) {
+ for_each_zone_zonelist(zone, z, zonelist, highest_zoneidx) {
nid = zone_to_nid(zone);
if (cpuset_zone_allowed(zone, flags) &&
diff --git a/mm/slub.c b/mm/slub.c
index 1c55bf7..d220671 100644
--- a/mm/slub.c
+++ b/mm/slub.c
@@ -1909,7 +1909,7 @@ static void *get_any_partial(struct kmem_cache *s, gfp_t flags,
struct zonelist *zonelist;
struct zoneref *z;
struct zone *zone;
- enum zone_type high_zoneidx = gfp_zone(flags);
+ enum zone_type highest_zoneidx = gfp_zone(flags);
void *object;
unsigned int cpuset_mems_cookie;
@@ -1938,7 +1938,7 @@ static void *get_any_partial(struct kmem_cache *s, gfp_t flags,
do {
cpuset_mems_cookie = read_mems_allowed_begin();
zonelist = node_zonelist(mempolicy_slab_node(), flags);
- for_each_zone_zonelist(zone, z, zonelist, high_zoneidx) {
+ for_each_zone_zonelist(zone, z, zonelist, highest_zoneidx) {
struct kmem_cache_node *n;
n = get_node(s, zone_to_nid(zone));
diff --git a/mm/vmscan.c b/mm/vmscan.c
index dca623d..0616abe 100644
--- a/mm/vmscan.c
+++ b/mm/vmscan.c
@@ -3133,8 +3133,8 @@ static bool allow_direct_reclaim(pg_data_t *pgdat)
/* kswapd must be awake if processes are being throttled */
if (!wmark_ok && waitqueue_active(&pgdat->kswapd_wait)) {
- if (READ_ONCE(pgdat->kswapd_classzone_idx) > ZONE_NORMAL)
- WRITE_ONCE(pgdat->kswapd_classzone_idx, ZONE_NORMAL);
+ if (READ_ONCE(pgdat->kswapd_highest_zoneidx) > ZONE_NORMAL)
+ WRITE_ONCE(pgdat->kswapd_highest_zoneidx, ZONE_NORMAL);
wake_up_interruptible(&pgdat->kswapd_wait);
}
@@ -3387,7 +3387,7 @@ static void age_active_anon(struct pglist_data *pgdat,
} while (memcg);
}
-static bool pgdat_watermark_boosted(pg_data_t *pgdat, int classzone_idx)
+static bool pgdat_watermark_boosted(pg_data_t *pgdat, int highest_zoneidx)
{
int i;
struct zone *zone;
@@ -3399,7 +3399,7 @@ static bool pgdat_watermark_boosted(pg_data_t *pgdat, int classzone_idx)
* start prematurely when there is no boosting and a lower
* zone is balanced.
*/
- for (i = classzone_idx; i >= 0; i--) {
+ for (i = highest_zoneidx; i >= 0; i--) {
zone = pgdat->node_zones + i;
if (!managed_zone(zone))
continue;
@@ -3413,9 +3413,9 @@ static bool pgdat_watermark_boosted(pg_data_t *pgdat, int classzone_idx)
/*
* Returns true if there is an eligible zone balanced for the request order
- * and classzone_idx
+ * and highest_zoneidx
*/
-static bool pgdat_balanced(pg_data_t *pgdat, int order, int classzone_idx)
+static bool pgdat_balanced(pg_data_t *pgdat, int order, int highest_zoneidx)
{
int i;
unsigned long mark = -1;
@@ -3425,19 +3425,19 @@ static bool pgdat_balanced(pg_data_t *pgdat, int order, int classzone_idx)
* Check watermarks bottom-up as lower zones are more likely to
* meet watermarks.
*/
- for (i = 0; i <= classzone_idx; i++) {
+ for (i = 0; i <= highest_zoneidx; i++) {
zone = pgdat->node_zones + i;
if (!managed_zone(zone))
continue;
mark = high_wmark_pages(zone);
- if (zone_watermark_ok_safe(zone, order, mark, classzone_idx))
+ if (zone_watermark_ok_safe(zone, order, mark, highest_zoneidx))
return true;
}
/*
- * If a node has no populated zone within classzone_idx, it does not
+ * If a node has no populated zone within highest_zoneidx, it does not
* need balancing by definition. This can happen if a zone-restricted
* allocation tries to wake a remote kswapd.
*/
@@ -3463,7 +3463,8 @@ static void clear_pgdat_congested(pg_data_t *pgdat)
*
* Returns true if kswapd is ready to sleep
*/
-static bool prepare_kswapd_sleep(pg_data_t *pgdat, int order, int classzone_idx)
+static bool prepare_kswapd_sleep(pg_data_t *pgdat, int order,
+ int highest_zoneidx)
{
/*
* The throttled processes are normally woken up in balance_pgdat() as
@@ -3485,7 +3486,7 @@ static bool prepare_kswapd_sleep(pg_data_t *pgdat, int order, int classzone_idx)
if (pgdat->kswapd_failures >= MAX_RECLAIM_RETRIES)
return true;
- if (pgdat_balanced(pgdat, order, classzone_idx)) {
+ if (pgdat_balanced(pgdat, order, highest_zoneidx)) {
clear_pgdat_congested(pgdat);
return true;
}
@@ -3549,7 +3550,7 @@ static bool kswapd_shrink_node(pg_data_t *pgdat,
* or lower is eligible for reclaim until at least one usable zone is
* balanced.
*/
-static int balance_pgdat(pg_data_t *pgdat, int order, int classzone_idx)
+static int balance_pgdat(pg_data_t *pgdat, int order, int highest_zoneidx)
{
int i;
unsigned long nr_soft_reclaimed;
@@ -3577,7 +3578,7 @@ static int balance_pgdat(pg_data_t *pgdat, int order, int classzone_idx)
* stall or direct reclaim until kswapd is finished.
*/
nr_boost_reclaim = 0;
- for (i = 0; i <= classzone_idx; i++) {
+ for (i = 0; i <= highest_zoneidx; i++) {
zone = pgdat->node_zones + i;
if (!managed_zone(zone))
continue;
@@ -3595,7 +3596,7 @@ static int balance_pgdat(pg_data_t *pgdat, int order, int classzone_idx)
bool balanced;
bool ret;
- sc.reclaim_idx = classzone_idx;
+ sc.reclaim_idx = highest_zoneidx;
/*
* If the number of buffer_heads exceeds the maximum allowed
@@ -3625,7 +3626,7 @@ static int balance_pgdat(pg_data_t *pgdat, int order, int classzone_idx)
* on the grounds that the normal reclaim should be enough to
* re-evaluate if boosting is required when kswapd next wakes.
*/
- balanced = pgdat_balanced(pgdat, sc.order, classzone_idx);
+ balanced = pgdat_balanced(pgdat, sc.order, highest_zoneidx);
if (!balanced && nr_boost_reclaim) {
nr_boost_reclaim = 0;
goto restart;
@@ -3725,7 +3726,7 @@ static int balance_pgdat(pg_data_t *pgdat, int order, int classzone_idx)
if (boosted) {
unsigned long flags;
- for (i = 0; i <= classzone_idx; i++) {
+ for (i = 0; i <= highest_zoneidx; i++) {
if (!zone_boosts[i])
continue;
@@ -3740,7 +3741,7 @@ static int balance_pgdat(pg_data_t *pgdat, int order, int classzone_idx)
* As there is now likely space, wakeup kcompact to defragment
* pageblocks.
*/
- wakeup_kcompactd(pgdat, pageblock_order, classzone_idx);
+ wakeup_kcompactd(pgdat, pageblock_order, highest_zoneidx);
}
snapshot_refaults(NULL, pgdat);
@@ -3758,22 +3759,22 @@ static int balance_pgdat(pg_data_t *pgdat, int order, int classzone_idx)
}
/*
- * The pgdat->kswapd_classzone_idx is used to pass the highest zone index to be
- * reclaimed by kswapd from the waker. If the value is MAX_NR_ZONES which is not
- * a valid index then either kswapd runs for first time or kswapd couldn't sleep
- * after previous reclaim attempt (node is still unbalanced). In that case
- * return the zone index of the previous kswapd reclaim cycle.
+ * The pgdat->kswapd_highest_zoneidx is used to pass the highest zone index to
+ * be reclaimed by kswapd from the waker. If the value is MAX_NR_ZONES which is
+ * not a valid index then either kswapd runs for first time or kswapd couldn't
+ * sleep after previous reclaim attempt (node is still unbalanced). In that
+ * case return the zone index of the previous kswapd reclaim cycle.
*/
-static enum zone_type kswapd_classzone_idx(pg_data_t *pgdat,
- enum zone_type prev_classzone_idx)
+static enum zone_type kswapd_highest_zoneidx(pg_data_t *pgdat,
+ enum zone_type prev_highest_zoneidx)
{
- enum zone_type curr_idx = READ_ONCE(pgdat->kswapd_classzone_idx);
+ enum zone_type curr_idx = READ_ONCE(pgdat->kswapd_highest_zoneidx);
- return curr_idx == MAX_NR_ZONES ? prev_classzone_idx : curr_idx;
+ return curr_idx == MAX_NR_ZONES ? prev_highest_zoneidx : curr_idx;
}
static void kswapd_try_to_sleep(pg_data_t *pgdat, int alloc_order, int reclaim_order,
- unsigned int classzone_idx)
+ unsigned int highest_zoneidx)
{
long remaining = 0;
DEFINE_WAIT(wait);
@@ -3790,7 +3791,7 @@ static void kswapd_try_to_sleep(pg_data_t *pgdat, int alloc_order, int reclaim_o
* eligible zone balanced that it's also unlikely that compaction will
* succeed.
*/
- if (prepare_kswapd_sleep(pgdat, reclaim_order, classzone_idx)) {
+ if (prepare_kswapd_sleep(pgdat, reclaim_order, highest_zoneidx)) {
/*
* Compaction records what page blocks it recently failed to
* isolate pages from and skips them in the future scanning.
@@ -3803,18 +3804,19 @@ static void kswapd_try_to_sleep(pg_data_t *pgdat, int alloc_order, int reclaim_o
* We have freed the memory, now we should compact it to make
* allocation of the requested order possible.
*/
- wakeup_kcompactd(pgdat, alloc_order, classzone_idx);
+ wakeup_kcompactd(pgdat, alloc_order, highest_zoneidx);
remaining = schedule_timeout(HZ/10);
/*
- * If woken prematurely then reset kswapd_classzone_idx and
+ * If woken prematurely then reset kswapd_highest_zoneidx and
* order. The values will either be from a wakeup request or
* the previous request that slept prematurely.
*/
if (remaining) {
- WRITE_ONCE(pgdat->kswapd_classzone_idx,
- kswapd_classzone_idx(pgdat, classzone_idx));
+ WRITE_ONCE(pgdat->kswapd_highest_zoneidx,
+ kswapd_highest_zoneidx(pgdat,
+ highest_zoneidx));
if (READ_ONCE(pgdat->kswapd_order) < reclaim_order)
WRITE_ONCE(pgdat->kswapd_order, reclaim_order);
@@ -3829,7 +3831,7 @@ static void kswapd_try_to_sleep(pg_data_t *pgdat, int alloc_order, int reclaim_o
* go fully to sleep until explicitly woken up.
*/
if (!remaining &&
- prepare_kswapd_sleep(pgdat, reclaim_order, classzone_idx)) {
+ prepare_kswapd_sleep(pgdat, reclaim_order, highest_zoneidx)) {
trace_mm_vmscan_kswapd_sleep(pgdat->node_id);
/*
@@ -3871,7 +3873,7 @@ static void kswapd_try_to_sleep(pg_data_t *pgdat, int alloc_order, int reclaim_o
static int kswapd(void *p)
{
unsigned int alloc_order, reclaim_order;
- unsigned int classzone_idx = MAX_NR_ZONES - 1;
+ unsigned int highest_zoneidx = MAX_NR_ZONES - 1;
pg_data_t *pgdat = (pg_data_t*)p;
struct task_struct *tsk = current;
const struct cpumask *cpumask = cpumask_of_node(pgdat->node_id);
@@ -3895,22 +3897,24 @@ static int kswapd(void *p)
set_freezable();
WRITE_ONCE(pgdat->kswapd_order, 0);
- WRITE_ONCE(pgdat->kswapd_classzone_idx, MAX_NR_ZONES);
+ WRITE_ONCE(pgdat->kswapd_highest_zoneidx, MAX_NR_ZONES);
for ( ; ; ) {
bool ret;
alloc_order = reclaim_order = READ_ONCE(pgdat->kswapd_order);
- classzone_idx = kswapd_classzone_idx(pgdat, classzone_idx);
+ highest_zoneidx = kswapd_highest_zoneidx(pgdat,
+ highest_zoneidx);
kswapd_try_sleep:
kswapd_try_to_sleep(pgdat, alloc_order, reclaim_order,
- classzone_idx);
+ highest_zoneidx);
- /* Read the new order and classzone_idx */
+ /* Read the new order and highest_zoneidx */
alloc_order = reclaim_order = READ_ONCE(pgdat->kswapd_order);
- classzone_idx = kswapd_classzone_idx(pgdat, classzone_idx);
+ highest_zoneidx = kswapd_highest_zoneidx(pgdat,
+ highest_zoneidx);
WRITE_ONCE(pgdat->kswapd_order, 0);
- WRITE_ONCE(pgdat->kswapd_classzone_idx, MAX_NR_ZONES);
+ WRITE_ONCE(pgdat->kswapd_highest_zoneidx, MAX_NR_ZONES);
ret = try_to_freeze();
if (kthread_should_stop())
@@ -3931,9 +3935,10 @@ static int kswapd(void *p)
* but kcompactd is woken to compact for the original
* request (alloc_order).
*/
- trace_mm_vmscan_kswapd_wake(pgdat->node_id, classzone_idx,
+ trace_mm_vmscan_kswapd_wake(pgdat->node_id, highest_zoneidx,
alloc_order);
- reclaim_order = balance_pgdat(pgdat, alloc_order, classzone_idx);
+ reclaim_order = balance_pgdat(pgdat, alloc_order,
+ highest_zoneidx);
if (reclaim_order < alloc_order)
goto kswapd_try_sleep;
}
@@ -3951,7 +3956,7 @@ static int kswapd(void *p)
* needed.
*/
void wakeup_kswapd(struct zone *zone, gfp_t gfp_flags, int order,
- enum zone_type classzone_idx)
+ enum zone_type highest_zoneidx)
{
pg_data_t *pgdat;
enum zone_type curr_idx;
@@ -3963,10 +3968,10 @@ void wakeup_kswapd(struct zone *zone, gfp_t gfp_flags, int order,
return;
pgdat = zone->zone_pgdat;
- curr_idx = READ_ONCE(pgdat->kswapd_classzone_idx);
+ curr_idx = READ_ONCE(pgdat->kswapd_highest_zoneidx);
- if (curr_idx == MAX_NR_ZONES || curr_idx < classzone_idx)
- WRITE_ONCE(pgdat->kswapd_classzone_idx, classzone_idx);
+ if (curr_idx == MAX_NR_ZONES || curr_idx < highest_zoneidx)
+ WRITE_ONCE(pgdat->kswapd_highest_zoneidx, highest_zoneidx);
if (READ_ONCE(pgdat->kswapd_order) < order)
WRITE_ONCE(pgdat->kswapd_order, order);
@@ -3976,8 +3981,8 @@ void wakeup_kswapd(struct zone *zone, gfp_t gfp_flags, int order,
/* Hopeless node, leave it to direct reclaim if possible */
if (pgdat->kswapd_failures >= MAX_RECLAIM_RETRIES ||
- (pgdat_balanced(pgdat, order, classzone_idx) &&
- !pgdat_watermark_boosted(pgdat, classzone_idx))) {
+ (pgdat_balanced(pgdat, order, highest_zoneidx) &&
+ !pgdat_watermark_boosted(pgdat, highest_zoneidx))) {
/*
* There may be plenty of free memory available, but it's too
* fragmented for high-order allocations. Wake up kcompactd
@@ -3986,11 +3991,11 @@ void wakeup_kswapd(struct zone *zone, gfp_t gfp_flags, int order,
* ratelimit its work.
*/
if (!(gfp_flags & __GFP_DIRECT_RECLAIM))
- wakeup_kcompactd(pgdat, order, classzone_idx);
+ wakeup_kcompactd(pgdat, order, highest_zoneidx);
return;
}
- trace_mm_vmscan_wakeup_kswapd(pgdat->node_id, classzone_idx, order,
+ trace_mm_vmscan_wakeup_kswapd(pgdat->node_id, highest_zoneidx, order,
gfp_flags);
wake_up_interruptible(&pgdat->kswapd_wait);
}
--
2.7.4
On 03/20/20 at 05:32pm, [email protected] wrote:
> From: Joonsoo Kim <[email protected]>
>
> Currently, we use the zone index of preferred_zone which represents
> the best matching zone for allocation, as classzone_idx. It has
> a problem on NUMA systems when the lowmem reserve protection exists
> for some zones on a node that do not exist on other nodes.
>
> In NUMA system, it can be possible that each node has different populated
> zones. For example, node 0 could have DMA/DMA32/NORMAL/MOVABLE zone and
> node 1 could have only NORMAL zone. In this setup, allocation request
> initiated on node 0 and the one on node 1 would have different
> classzone_idx, 3 and 2, respectively, since their preferred_zones are
> different. If the allocation is local, there is no problem. However,
> if it is handled by the remote node due to memory shortage, the problem
> would happen.
Hi Joonsoo,
Not sure if adding one sentence into above paragraph would be make it
easier to understand. Assume you are only talking about the high_zoneidx
is MOVABLE_ZONE with calculation of gfp_zone(gfp_mask), since any other
case doesn't have this problem. Please correct me if I am wrong.
In NUMA system, it can be possible that each node has different populated
zones. For example, node 0 could have DMA/DMA32/NORMAL/MOVABLE zone and
node 1 could have only NORMAL zone. In this setup, if we get high_zoneidx
as 3 (namely MOVABLE zone), with gfp_zone(gfp_mask), allocation request
initiated on node 0 and the one on node 1 would have different
classzone_idx, 3 and 2, respectively, since their preferred_zones are
different. If the allocation is local, there is no problem. However,
if it is handled by the remote node due to memory shortage, the problem
would happen.
Thanks
Baoquan
>
> In the following setup, allocation initiated on node 1 will have some
> precedence than allocation initiated on node 0 when former allocation is
> processed on node 0 due to not enough memory on node 1. They will have
> different lowmem reserve due to their different classzone_idx thus
> an watermark bars are also different.
>
> root@ubuntu:/sys/devices/system/memory# cat /proc/zoneinfo
> Node 0, zone DMA
> per-node stats
> ...
> pages free 3965
> min 5
> low 8
> high 11
> spanned 4095
> present 3998
> managed 3977
> protection: (0, 2961, 4928, 5440)
> ...
> Node 0, zone DMA32
> pages free 757955
> min 1129
> low 1887
> high 2645
> spanned 1044480
> present 782303
> managed 758116
> protection: (0, 0, 1967, 2479)
> ...
> Node 0, zone Normal
> pages free 459806
> min 750
> low 1253
> high 1756
> spanned 524288
> present 524288
> managed 503620
> protection: (0, 0, 0, 4096)
> ...
> Node 0, zone Movable
> pages free 130759
> min 195
> low 326
> high 457
> spanned 1966079
> present 131072
> managed 131072
> protection: (0, 0, 0, 0)
> ...
> Node 1, zone DMA
> pages free 0
> min 0
> low 0
> high 0
> spanned 0
> present 0
> managed 0
> protection: (0, 0, 1006, 1006)
> Node 1, zone DMA32
> pages free 0
> min 0
> low 0
> high 0
> spanned 0
> present 0
> managed 0
> protection: (0, 0, 1006, 1006)
> Node 1, zone Normal
> per-node stats
> ...
> pages free 233277
> min 383
> low 640
> high 897
> spanned 262144
> present 262144
> managed 257744
> protection: (0, 0, 0, 0)
> ...
> Node 1, zone Movable
> pages free 0
> min 0
> low 0
> high 0
> spanned 262144
> present 0
> managed 0
> protection: (0, 0, 0, 0)
>
> min watermark for NORMAL zone on node 0
> allocation initiated on node 0: 750 + 4096 = 4846
> allocation initiated on node 1: 750 + 0 = 750
>
> This watermark difference could cause too many numa_miss allocation
> in some situation and then performance could be downgraded.
>
> Recently, there was a regression report about this problem on CMA patches
> since CMA memory are placed in ZONE_MOVABLE by those patches. I checked
> that problem is disappeared with this fix that uses high_zoneidx
> for classzone_idx.
>
> http://lkml.kernel.org/r/20180102063528.GG30397@yexl-desktop
>
> Using high_zoneidx for classzone_idx is more consistent way than previous
> approach because system's memory layout doesn't affect anything to it.
> With this patch, both classzone_idx on above example will be 3 so will
> have the same min watermark.
>
> allocation initiated on node 0: 750 + 4096 = 4846
> allocation initiated on node 1: 750 + 4096 = 4846
>
> One could wonder if there is a side effect that allocation initiated on
> node 1 will use higher bar when allocation is handled on local since
> classzone_idx could be higher than before. It will not happen because
> the zone without managed page doesn't contributes lowmem_reserve at all.
>
> Reported-by: Ye Xiaolong <[email protected]>
> Tested-by: Ye Xiaolong <[email protected]>
> Acked-by: Vlastimil Babka <[email protected]>
> Signed-off-by: Joonsoo Kim <[email protected]>
> ---
> mm/internal.h | 2 +-
> 1 file changed, 1 insertion(+), 1 deletion(-)
>
> diff --git a/mm/internal.h b/mm/internal.h
> index c39c895..aebaa33 100644
> --- a/mm/internal.h
> +++ b/mm/internal.h
> @@ -119,7 +119,7 @@ struct alloc_context {
> bool spread_dirty_pages;
> };
>
> -#define ac_classzone_idx(ac) zonelist_zone_idx(ac->preferred_zoneref)
> +#define ac_classzone_idx(ac) (ac->high_zoneidx)
>
> /*
> * Locate the struct page for both the matching buddy in our
> --
> 2.7.4
>
>
Hello, Baoquan.
2020년 3월 20일 (금) 오후 7:30, Baoquan He <[email protected]>님이 작성:
>
>
> On 03/20/20 at 05:32pm, [email protected] wrote:
> > From: Joonsoo Kim <[email protected]>
> >
> > Currently, we use the zone index of preferred_zone which represents
> > the best matching zone for allocation, as classzone_idx. It has
> > a problem on NUMA systems when the lowmem reserve protection exists
> > for some zones on a node that do not exist on other nodes.
> >
> > In NUMA system, it can be possible that each node has different populated
> > zones. For example, node 0 could have DMA/DMA32/NORMAL/MOVABLE zone and
> > node 1 could have only NORMAL zone. In this setup, allocation request
> > initiated on node 0 and the one on node 1 would have different
> > classzone_idx, 3 and 2, respectively, since their preferred_zones are
> > different. If the allocation is local, there is no problem. However,
> > if it is handled by the remote node due to memory shortage, the problem
> > would happen.
>
> Hi Joonsoo,
>
> Not sure if adding one sentence into above paragraph would be make it
> easier to understand. Assume you are only talking about the high_zoneidx
> is MOVABLE_ZONE with calculation of gfp_zone(gfp_mask), since any other
> case doesn't have this problem. Please correct me if I am wrong.
You're right. This example is for the allocation request with
gfp_zone(gfp_mask),
MOVABLE_ZONE.
> In NUMA system, it can be possible that each node has different populated
> zones. For example, node 0 could have DMA/DMA32/NORMAL/MOVABLE zone and
> node 1 could have only NORMAL zone. In this setup, if we get high_zoneidx
> as 3 (namely MOVABLE zone), with gfp_zone(gfp_mask), allocation request
> initiated on node 0 and the one on node 1 would have different
> classzone_idx, 3 and 2, respectively, since their preferred_zones are
> different. If the allocation is local, there is no problem. However,
> if it is handled by the remote node due to memory shortage, the problem
> would happen.
I'm okay with your change but I try again to be better. Please check the
following rewritten commit message and please let me know if it is better
than before.
Thanks.
------------------------>8-------------------------------
From 3842d99d54c02c25cc6f6077fd0e97acabc10917 Mon Sep 17 00:00:00 2001
From: Joonsoo Kim <[email protected]>
Date: Fri, 4 May 2018 09:55:21 +0900
Subject: [PATCH] mm/page_alloc: use ac->high_zoneidx for classzone_idx
Currently, we use classzone_idx to calculate lowmem reserve proetection
for an allocation request. This classzone_idx causes a problem
on NUMA systems when the lowmem reserve protection exists for some zones
on a node that do not exist on other nodes.
Before further explanation, I should first clarify how to compute
the classzone_idx and the high_zoneidx.
- ac->high_zoneidx is computed via the arcane gfp_zone(gfp_mask) and
represents the index of the highest zone the allocation can use
- classzone_idx was supposed to be the index of the highest zone on
the local node that the allocation can use, that is actually available
in the system
Think about following example. Node 0 has 4 populated zone,
DMA/DMA32/NORMAL/MOVABLE. Node 1 has 1 populated zone, NORMAL. Some zones,
such as MOVABLE, doesn't exist on node 1 and this makes following
difference.
Assume that there is an allocation request whose gfp_zone(gfp_mask) is
the zone, MOVABLE. Then, it's high_zoneidx is 3. If this allocation is
initiated on node 0, it's classzone_idx is 3 since actually
available/usable zone on local (node 0) is MOVABLE. If this allocation
is initiated on node 1, it's classzone_idx is 2 since actually
available/usable zone on local (node 1) is NORMAL.
You can see that classzone_idx of the allocation request are different
according to their starting node, even if their high_zoneidx is the same.
Think more about these two allocation requests. If they are processed
on local, there is no problem. However, if allocation is initiated
on node 1 are processed on remote, in this example, at the NORMAL zone
on node 0, due to memory shortage, problem occurs. Their different
classzone_idx leads to different lowmem reserve and then different
min watermark. See the following example.
root@ubuntu:/sys/devices/system/memory# cat /proc/zoneinfo
Node 0, zone DMA
per-node stats
...
pages free 3965
min 5
low 8
high 11
spanned 4095
present 3998
managed 3977
protection: (0, 2961, 4928, 5440)
...
Node 0, zone DMA32
pages free 757955
min 1129
low 1887
high 2645
spanned 1044480
present 782303
managed 758116
protection: (0, 0, 1967, 2479)
...
Node 0, zone Normal
pages free 459806
min 750
low 1253
high 1756
spanned 524288
present 524288
managed 503620
protection: (0, 0, 0, 4096)
...
Node 0, zone Movable
pages free 130759
min 195
low 326
high 457
spanned 1966079
present 131072
managed 131072
protection: (0, 0, 0, 0)
...
Node 1, zone DMA
pages free 0
min 0
low 0
high 0
spanned 0
present 0
managed 0
protection: (0, 0, 1006, 1006)
Node 1, zone DMA32
pages free 0
min 0
low 0
high 0
spanned 0
present 0
managed 0
protection: (0, 0, 1006, 1006)
Node 1, zone Normal
per-node stats
...
pages free 233277
min 383
low 640
high 897
spanned 262144
present 262144
managed 257744
protection: (0, 0, 0, 0)
...
Node 1, zone Movable
pages free 0
min 0
low 0
high 0
spanned 262144
present 0
managed 0
protection: (0, 0, 0, 0)
- static min watermark for the NORMAL zone on node 0 is 750.
- lowmem reserve for the request with classzone idx 3 at the NORMAL
on node 0 is 4096.
- lowmem reserve for the request with classzone idx 2 at the NORMAL
on node 0 is 0.
So, overall min watermark is:
allocation initiated on node 0 (classzone_idx 3): 750 + 4096 = 4846
allocation initiated on node 1 (classzone_idx 2): 750 + 0 = 750
allocation initiated on node 1 will have some precedence than allocation
initiated on node 0 because min watermark of the former allocation is
lower than the other. So, allocation initiated on node 1 could succeed
on node 0 when allocation initiated on node 0 could not, and, this could
cause too many numa_miss allocation. Then, performance could be
downgraded.
Recently, there was a regression report about this problem on CMA patches
since CMA memory are placed in ZONE_MOVABLE by those patches. I checked
that problem is disappeared with this fix that uses high_zoneidx
for classzone_idx.
http://lkml.kernel.org/r/20180102063528.GG30397@yexl-desktop
Using high_zoneidx for classzone_idx is more consistent way than previous
approach because system's memory layout doesn't affect anything to it.
With this patch, both classzone_idx on above example will be 3 so will
have the same min watermark.
allocation initiated on node 0: 750 + 4096 = 4846
allocation initiated on node 1: 750 + 4096 = 4846
One could wonder if there is a side effect that allocation initiated on
node 1 will use higher bar when allocation is handled on local since
classzone_idx could be higher than before. It will not happen because
the zone without managed page doesn't contributes lowmem_reserve at all.
Reported-by: Ye Xiaolong <[email protected]>
Tested-by: Ye Xiaolong <[email protected]>
Acked-by: Vlastimil Babka <[email protected]>
Signed-off-by: Joonsoo Kim <[email protected]>
---
mm/internal.h | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/mm/internal.h b/mm/internal.h
index c39c895..aebaa33 100644
--- a/mm/internal.h
+++ b/mm/internal.h
@@ -119,7 +119,7 @@ struct alloc_context {
bool spread_dirty_pages;
};
-#define ac_classzone_idx(ac) zonelist_zone_idx(ac->preferred_zoneref)
+#define ac_classzone_idx(ac) (ac->high_zoneidx)
/*
* Locate the struct page for both the matching buddy in our
--
2.7.4
On 03/23/20 at 12:50pm, Joonsoo Kim wrote:
> Hello, Baoquan.
>
> 2020년 3월 20일 (금) 오후 7:30, Baoquan He <[email protected]>님이 작성:
> >
> >
> > On 03/20/20 at 05:32pm, [email protected] wrote:
> > > From: Joonsoo Kim <[email protected]>
> > >
> > > Currently, we use the zone index of preferred_zone which represents
> > > the best matching zone for allocation, as classzone_idx. It has
> > > a problem on NUMA systems when the lowmem reserve protection exists
> > > for some zones on a node that do not exist on other nodes.
> > >
> > > In NUMA system, it can be possible that each node has different populated
> > > zones. For example, node 0 could have DMA/DMA32/NORMAL/MOVABLE zone and
> > > node 1 could have only NORMAL zone. In this setup, allocation request
> > > initiated on node 0 and the one on node 1 would have different
> > > classzone_idx, 3 and 2, respectively, since their preferred_zones are
> > > different. If the allocation is local, there is no problem. However,
> > > if it is handled by the remote node due to memory shortage, the problem
> > > would happen.
> >
> > Hi Joonsoo,
> >
> > Not sure if adding one sentence into above paragraph would be make it
> > easier to understand. Assume you are only talking about the high_zoneidx
> > is MOVABLE_ZONE with calculation of gfp_zone(gfp_mask), since any other
> > case doesn't have this problem. Please correct me if I am wrong.
>
> You're right. This example is for the allocation request with
> gfp_zone(gfp_mask),
> MOVABLE_ZONE.
>
> > In NUMA system, it can be possible that each node has different populated
> > zones. For example, node 0 could have DMA/DMA32/NORMAL/MOVABLE zone and
> > node 1 could have only NORMAL zone. In this setup, if we get high_zoneidx
> > as 3 (namely MOVABLE zone), with gfp_zone(gfp_mask), allocation request
> > initiated on node 0 and the one on node 1 would have different
> > classzone_idx, 3 and 2, respectively, since their preferred_zones are
> > different. If the allocation is local, there is no problem. However,
> > if it is handled by the remote node due to memory shortage, the problem
> > would happen.
>
> I'm okay with your change but I try again to be better. Please check the
> following rewritten commit message and please let me know if it is better
> than before.
Yeah, this new one looks very detailed, I believe anyone interested can
get what's going on. Thanks for doing this.
>
> Thanks.
>
> ------------------------>8-------------------------------
> From 3842d99d54c02c25cc6f6077fd0e97acabc10917 Mon Sep 17 00:00:00 2001
> From: Joonsoo Kim <[email protected]>
> Date: Fri, 4 May 2018 09:55:21 +0900
> Subject: [PATCH] mm/page_alloc: use ac->high_zoneidx for classzone_idx
>
> Currently, we use classzone_idx to calculate lowmem reserve proetection
> for an allocation request. This classzone_idx causes a problem
> on NUMA systems when the lowmem reserve protection exists for some zones
> on a node that do not exist on other nodes.
>
> Before further explanation, I should first clarify how to compute
> the classzone_idx and the high_zoneidx.
>
> - ac->high_zoneidx is computed via the arcane gfp_zone(gfp_mask) and
> represents the index of the highest zone the allocation can use
> - classzone_idx was supposed to be the index of the highest zone on
> the local node that the allocation can use, that is actually available
> in the system
>
> Think about following example. Node 0 has 4 populated zone,
> DMA/DMA32/NORMAL/MOVABLE. Node 1 has 1 populated zone, NORMAL. Some zones,
> such as MOVABLE, doesn't exist on node 1 and this makes following
> difference.
>
> Assume that there is an allocation request whose gfp_zone(gfp_mask) is
> the zone, MOVABLE. Then, it's high_zoneidx is 3. If this allocation is
> initiated on node 0, it's classzone_idx is 3 since actually
> available/usable zone on local (node 0) is MOVABLE. If this allocation
> is initiated on node 1, it's classzone_idx is 2 since actually
> available/usable zone on local (node 1) is NORMAL.
>
> You can see that classzone_idx of the allocation request are different
> according to their starting node, even if their high_zoneidx is the same.
>
> Think more about these two allocation requests. If they are processed
> on local, there is no problem. However, if allocation is initiated
> on node 1 are processed on remote, in this example, at the NORMAL zone
> on node 0, due to memory shortage, problem occurs. Their different
> classzone_idx leads to different lowmem reserve and then different
> min watermark. See the following example.
>
> root@ubuntu:/sys/devices/system/memory# cat /proc/zoneinfo
> Node 0, zone DMA
> per-node stats
> ...
> pages free 3965
> min 5
> low 8
> high 11
> spanned 4095
> present 3998
> managed 3977
> protection: (0, 2961, 4928, 5440)
> ...
> Node 0, zone DMA32
> pages free 757955
> min 1129
> low 1887
> high 2645
> spanned 1044480
> present 782303
> managed 758116
> protection: (0, 0, 1967, 2479)
> ...
> Node 0, zone Normal
> pages free 459806
> min 750
> low 1253
> high 1756
> spanned 524288
> present 524288
> managed 503620
> protection: (0, 0, 0, 4096)
> ...
> Node 0, zone Movable
> pages free 130759
> min 195
> low 326
> high 457
> spanned 1966079
> present 131072
> managed 131072
> protection: (0, 0, 0, 0)
> ...
> Node 1, zone DMA
> pages free 0
> min 0
> low 0
> high 0
> spanned 0
> present 0
> managed 0
> protection: (0, 0, 1006, 1006)
> Node 1, zone DMA32
> pages free 0
> min 0
> low 0
> high 0
> spanned 0
> present 0
> managed 0
> protection: (0, 0, 1006, 1006)
> Node 1, zone Normal
> per-node stats
> ...
> pages free 233277
> min 383
> low 640
> high 897
> spanned 262144
> present 262144
> managed 257744
> protection: (0, 0, 0, 0)
> ...
> Node 1, zone Movable
> pages free 0
> min 0
> low 0
> high 0
> spanned 262144
> present 0
> managed 0
> protection: (0, 0, 0, 0)
>
> - static min watermark for the NORMAL zone on node 0 is 750.
> - lowmem reserve for the request with classzone idx 3 at the NORMAL
> on node 0 is 4096.
> - lowmem reserve for the request with classzone idx 2 at the NORMAL
> on node 0 is 0.
>
> So, overall min watermark is:
> allocation initiated on node 0 (classzone_idx 3): 750 + 4096 = 4846
> allocation initiated on node 1 (classzone_idx 2): 750 + 0 = 750
>
> allocation initiated on node 1 will have some precedence than allocation
> initiated on node 0 because min watermark of the former allocation is
> lower than the other. So, allocation initiated on node 1 could succeed
> on node 0 when allocation initiated on node 0 could not, and, this could
> cause too many numa_miss allocation. Then, performance could be
> downgraded.
>
> Recently, there was a regression report about this problem on CMA patches
> since CMA memory are placed in ZONE_MOVABLE by those patches. I checked
> that problem is disappeared with this fix that uses high_zoneidx
> for classzone_idx.
>
> http://lkml.kernel.org/r/20180102063528.GG30397@yexl-desktop
>
> Using high_zoneidx for classzone_idx is more consistent way than previous
> approach because system's memory layout doesn't affect anything to it.
> With this patch, both classzone_idx on above example will be 3 so will
> have the same min watermark.
>
> allocation initiated on node 0: 750 + 4096 = 4846
> allocation initiated on node 1: 750 + 4096 = 4846
>
> One could wonder if there is a side effect that allocation initiated on
> node 1 will use higher bar when allocation is handled on local since
> classzone_idx could be higher than before. It will not happen because
> the zone without managed page doesn't contributes lowmem_reserve at all.
>
> Reported-by: Ye Xiaolong <[email protected]>
> Tested-by: Ye Xiaolong <[email protected]>
> Acked-by: Vlastimil Babka <[email protected]>
> Signed-off-by: Joonsoo Kim <[email protected]>
> ---
> mm/internal.h | 2 +-
> 1 file changed, 1 insertion(+), 1 deletion(-)
>
> diff --git a/mm/internal.h b/mm/internal.h
> index c39c895..aebaa33 100644
> --- a/mm/internal.h
> +++ b/mm/internal.h
> @@ -119,7 +119,7 @@ struct alloc_context {
> bool spread_dirty_pages;
> };
>
> -#define ac_classzone_idx(ac) zonelist_zone_idx(ac->preferred_zoneref)
> +#define ac_classzone_idx(ac) (ac->high_zoneidx)
>
> /*
> * Locate the struct page for both the matching buddy in our
> --
> 2.7.4
>
2020년 3월 23일 (월) 오후 12:59, Baoquan He <[email protected]>님이 작성:
>
> On 03/23/20 at 12:50pm, Joonsoo Kim wrote:
> > Hello, Baoquan.
> >
> > 2020년 3월 20일 (금) 오후 7:30, Baoquan He <[email protected]>님이 작성:
> > >
> > >
> > > On 03/20/20 at 05:32pm, [email protected] wrote:
> > > > From: Joonsoo Kim <[email protected]>
> > > >
> > > > Currently, we use the zone index of preferred_zone which represents
> > > > the best matching zone for allocation, as classzone_idx. It has
> > > > a problem on NUMA systems when the lowmem reserve protection exists
> > > > for some zones on a node that do not exist on other nodes.
> > > >
> > > > In NUMA system, it can be possible that each node has different populated
> > > > zones. For example, node 0 could have DMA/DMA32/NORMAL/MOVABLE zone and
> > > > node 1 could have only NORMAL zone. In this setup, allocation request
> > > > initiated on node 0 and the one on node 1 would have different
> > > > classzone_idx, 3 and 2, respectively, since their preferred_zones are
> > > > different. If the allocation is local, there is no problem. However,
> > > > if it is handled by the remote node due to memory shortage, the problem
> > > > would happen.
> > >
> > > Hi Joonsoo,
> > >
> > > Not sure if adding one sentence into above paragraph would be make it
> > > easier to understand. Assume you are only talking about the high_zoneidx
> > > is MOVABLE_ZONE with calculation of gfp_zone(gfp_mask), since any other
> > > case doesn't have this problem. Please correct me if I am wrong.
> >
> > You're right. This example is for the allocation request with
> > gfp_zone(gfp_mask),
> > MOVABLE_ZONE.
> >
> > > In NUMA system, it can be possible that each node has different populated
> > > zones. For example, node 0 could have DMA/DMA32/NORMAL/MOVABLE zone and
> > > node 1 could have only NORMAL zone. In this setup, if we get high_zoneidx
> > > as 3 (namely MOVABLE zone), with gfp_zone(gfp_mask), allocation request
> > > initiated on node 0 and the one on node 1 would have different
> > > classzone_idx, 3 and 2, respectively, since their preferred_zones are
> > > different. If the allocation is local, there is no problem. However,
> > > if it is handled by the remote node due to memory shortage, the problem
> > > would happen.
> >
> > I'm okay with your change but I try again to be better. Please check the
> > following rewritten commit message and please let me know if it is better
> > than before.
>
> Yeah, this new one looks very detailed, I believe anyone interested can
> get what's going on. Thanks for doing this.
Thanks for checking it.
I will submit v4 with including this re-written commit message.
Thanks.