Fragmentation avoidance improvements v2
Subsystems (9)
cover Fragmentation avoidance improvements v2
The 1-socket machine is different to the one used in v1 so some of the results are changed on that basis. The baseline has changed to 4.20-rc1 so the __GFP_THISNODE removal for THP is in effect which alters the behaviour on 2-socket in particular. The biggest changes are in the fourth patch, both in terms of functional changes and the fact it adds a vmstat and tracepoint for measuring stall latency.
Changelog since v1
o Rebase to v4.20-rc1 for the THP __GFP_THISNODE patch in particular
o Add tracepoint to record fragmentation stall durations
o Add vmstat event to record that a fragmentation stall occurred
o Stalls now alter watermark boosting
o Stalls occur only when the allocation is about to fail
It has been noted before that fragmentation avoidance (aka anti-fragmentation) is not perfect. Given sufficient time or an adverse workload, memory gets fragmented and the long-term success of high-order allocations degrades. This series defines an adverse workload, a definition of external fragmentation events (including serious) ones and a series that reduces the level of those fragmentation events.
The details of the workload and the consequences are described in more detail in the changelogs. However, from patch 1, this is a high-level summary of the adverse workload. The exact details are found in the mmtests implementation.
The broad details of the workload are as follows;
1. Create an XFS filesystem (not specified in the configuration but done as part of the testing for this patch)
2. Start 4 fio threads that write a number of 64K files inefficiently. Inefficiently means that files are created on first access and not created in advance (fio parameterr create_on_open=1) and fallocate is not used (fallocate=none). With multiple IO issuers this creates a mix of slab and page cache allocations over time. The total size of the files is 150% physical memory so that the slabs and page cache pages get mixed
3. Warm up a number of fio read-only threads accessing the same files created in step 2. This part runs for the same length of time it took to create the files. It'll fault back in old data and further interleave slab and page cache allocations. As it's now low on memory due to step 2, fragmentation occurs as pageblocks get stolen.
4. While step 3 is still running, start a process that tries to allocate 75% of memory as huge pages with a number of threads. The number of threads is based on a (NR_CPUS_SOCKET - NR_FIO_THREADS)/4 to avoid THP threads contending with fio, any other threads or forcing cross-NUMA scheduling. Note that the test has not been used on a machine with less than 8 cores. The benchmark records whether huge pages were allocated and what the fault latency was in microseconds
5. Measure the number of events potentially causing external fragmentation, the fault latency and the huge page allocation success rate.
6. Cleanup
Overall the series reduces external fragmentation causing events by over 95% on 1 and 2 socket machines, which in turn impacts high-order allocation success rates over the long term. There are differences in latencies and high-order allocation success rates. Latencies are a mixed bag as they are vulnerable to exact system state and whether allocations succeeded so they are treated as a secondary metric.
Patch 1 uses lower zones if they are populated and have free memory
instead of fragmenting a higher zone. It's special cased to handle a Normal->DMA32 fallback with the reasons explained in the changelog.
Patch 2+3 boosts watermarks temporarily when an external fragmentation
event occurs. kswapd wakes to reclaim a small amount of old memory and then wakes kcompactd on completion to recover the system slightly. This introduces some overhead in the slowpath. The level of boosting can be tuned or disabled depending on the tolerance for fragmentation vs allocation latency.
Patch 4 is more heavy handed. In the event of a movable allocation
request that can stall, it'll wake kswapd as in patch 3. However, if the expected fragmentation event is serious then the request will stall briefly on pfmemalloc_wait until kswapd completes light reclaim work and retry the allocation without stalling. This can avoid the fragmentation event entirely in some cases. The definition of a serious fragmentation event can be tuned or disabled.
Patch 5 is the hardest to prove it's a real benefit. In the event
that fragmentation was unavoidable, it'll queue a pageblock for kcompactd to clean. It's a fixed-length queue that is neither guaranteed to have a slot available or successfully clean a pageblock.
Patches 4 and 5 can be treated independently or dropped if necessary. This is particularly true of patch 5 as the benefit is difficult to detect given the impact of the first 4 patches. The bulk of the improvement in fragmentation avoidance is from patches 1-3 (94-97% reduction in fragmentation events for an adverse workload on both a 1-socket and 2-socket machine). The primary benefit of patch 4 is the increase in THP success rates and the fact it reduces fragmentation events to almost negligible levels with the option of eliminating them.
Documentation/sysctl/vm.txt | 42 +++++++ include/linux/compaction.h | 4 + include/linux/migrate.h | 7 +- include/linux/mm.h | 2 + include/linux/mmzone.h | 18 ++- include/linux/vm_event_item.h | 1 + include/trace/events/compaction.h | 62 ++++++++++ include/trace/events/kmem.h | 21 ++++ kernel/sysctl.c | 18 +++ mm/compaction.c | 147 +++++++++++++++++++++-- mm/internal.h | 14 ++- mm/migrate.c | 6 +- mm/page_alloc.c | 246 ++++++++++++++++++++++++++++++++++---- mm/vmscan.c | 123 +++++++++++++++++-- mm/vmstat.c | 1 + 15 files changed, 661 insertions(+), 51 deletions(-)
Signature · 1 line
2.16.4
PATCH 1/5 mm, page_alloc: Spread allocations across zones before introducing fragmentation
The page allocator zone lists are iterated based on the watermarks of each zone which does not take anti-fragmentation into account. On x86, node 0 may have multiple zones while other nodes have one zone. A consequence is that tasks running on node 0 may fragment ZONE_NORMAL even though ZONE_DMA32 has plenty of free memory. This patch special cases the allocator fast path such that it'll try an allocation from a lower local zone before fragmenting a higher zone. In this case, stealing of pageblocks or orders larger than a pageblock are still allowed in the fast path as they are uninteresting from a fragmentation point of view.
This was evaluated using a benchmark designed to fragment memory before attempting THPs. It's implemented in mmtests as the following configurations
configs/config-global-dhp__workload_thpfioscale configs/config-global-dhp__workload_thpfioscale-defrag configs/config-global-dhp__workload_thpfioscale-madvhugepage
e.g. from mmtests
./run-mmtests.sh --run-monitor --config configs/config-global-dhp__workload_thpfioscale test-run-1
The broad details of the workload are as follows;
1. Create an XFS filesystem (not specified in the configuration but done as part of the testing for this patch)
2. Start 4 fio threads that write a number of 64K files inefficiently. Inefficiently means that files are created on first access and not created in advance (fio parameterr create_on_open=1) and fallocate is not used (fallocate=none). With multiple IO issuers this creates a mix of slab and page cache allocations over time. The total size of the files is 150% physical memory so that the slabs and page cache pages get mixed
3. Warm up a number of fio read-only threads accessing the same files created in step 2. This part runs for the same length of time it took to create the files. It'll fault back in old data and further interleave slab and page cache allocations. As it's now low on memory due to step 2, fragmentation occurs as pageblocks get stolen.
4. While step 3 is still running, start a process that tries to allocate 75% of memory as huge pages with a number of threads. The number of threads is based on a (NR_CPUS_SOCKET - NR_FIO_THREADS)/4 to avoid THP threads contending with fio, any other threads or forcing cross-NUMA scheduling. Note that the test has not been used on a machine with less than 8 cores. The benchmark records whether huge pages were allocated and what the fault latency was in microseconds
5. Measure the number of events potentially causing external fragmentation, the fault latency and the huge page allocation success rate.
6. Cleanup
Note that due to the use of IO and page cache that this benchmark is not suitable for running on large machines where the time to fragment memory may be excessive. Also note that while this is one mix that generates fragmentation that it's not the only mix that generates fragmentation. Differences in workload that are more slab-intensive or whether SLUB is used with high-order pages may yield different results.
When the page allocator fragments memory, it records the event using the mm_page_alloc_extfrag event. If the fallback_order is smaller than a pageblock order (order-9 on 64-bit x86) then it's considered an event that may cause external fragmentation issues in the future. Hence, the primary metric here is the number of external fragmentation events that occur with order < 9. The secondary metric is allocation latency and huge page allocation success rates but note that differences in latencies and what the success rate also can affect the number of external fragmentation event which is why it's a secondary metric.
1-socket Skylake machine config-global-dhp__workload_thpfioscale XFS (no special madvise) 4 fio threads, 1 THP allocating thread
--------------------------------------
4.20-rc1 extfrag events < order 9: 1023463
4.20-rc1+patch: 358574 (65% reduction)
thpfioscale Fault Latencies
4.20.0-rc1 4.20.0-rc1
vanilla lowzone-v2r4
Min fault-base-1 588.00 ( 0.00%) 557.00 ( 5.27%)
Min fault-huge-1 0.00 ( 0.00%) 0.00 ( 0.00%)
Amean fault-base-1 663.58 ( 0.00%) 663.65 ( -0.01%)
Amean fault-huge-1 0.00 ( 0.00%) 0.00 ( 0.00%) 4.20.0-rc1 4.20.0-rc1
vanilla lowzone-v2r4Percentage huge-1 0.00 ( 0.00%) 0.00 ( 0.00%)
Fault latencies are reduced while allocation success rates remain at zero asthis configuration does not make any heavy effort to allocate THP and fio is heavily active at the time and filling memory. However, a 65% reduction of serious fragmentation events reduces the changes of external fragmentation being a problem in the future.
1-socket Skylake machine global-dhp__workload_thpfioscale-madvhugepage-xfs (MADV_HUGEPAGE)
-----------------------------------------------------------------
4.20-rc1 extfrag events < order 9: 342549
4.20-rc1+patch: 337890 (1% reduction)
thpfioscale Fault Latencies
4.20.0-rc1 4.20.0-rc1
vanilla lowzone-v2r4
Amean fault-base-1 1517.06 ( 0.00%) 1531.37 ( -0.94%)
Amean fault-huge-1 1129.50 ( 0.00%) 1160.95 ( -2.78%)thpfioscale Percentage Faults Huge
4.20.0-rc1 4.20.0-rc1
vanilla lowzone-v2r4Percentage huge-1 78.01 ( 0.00%) 78.97 ( 1.23%)
Nothing dramatic. Fragmentation events were only reduced slightly which is very different to what was reported in V1. A big difference with V1 is the relative size of Normal to the DMA32 zone. This machine has double the memory so the impact of using a small zone to avoid fragmentation events is much lower.
2-socket Haswell machine config-global-dhp__workload_thpfioscale XFS (no special madvise) 4 fio threads, 5 THP allocating threads
----------------------------------------------------------------
4.20-rc1 extfrag events < order 9: 209820
4.20-rc1+patch: 185923 (11% reduction)
thpfioscale Fault Latencies
4.20.0-rc1 4.20.0-rc1
vanilla lowzone-v2r4
Amean fault-base-5 1324.93 ( 0.00%) 1334.99 ( -0.76%)
Amean fault-huge-5 4681.71 ( 0.00%) 2428.43 ( 48.13%) 4.20.0-rc1 4.20.0-rc1
vanilla lowzone-v2r4Percentage huge-5 1.05 ( 0.00%) 1.13 ( 7.94%)
The reduction of external fragmentation events is expected. A careful reader may spot that the reduction is lower than it was on v1. This is due to the removal of __GFP_THISNODE in commit ac5b2c18911f ("mm: thp: relax __GFP_THISNODE for MADV_HUGEPAGE mappings") as THP allocations can now spill over to remote nodes instead of fragmenting local memory. This reduces the impact of the use of a lower zone to avoid fragmentation. It's also worth noting relative to v1 that the allocation success rate is slightly higher.
2-socket Haswell machine global-dhp__workload_thpfioscale-madvhugepage-xfs (MADV_HUGEPAGE)
-----------------------------------------------------------------
4.20-rc1 extfrag events < order 9: 167464
4.20-rc1+patch: 130081 (22% reduction)
thpfioscale Fault Latencies
4.20.0-rc1 4.20.0-rc1
vanilla lowzone-v2r4
Amean fault-base-5 7721.82 ( 0.00%) 6652.67 ( 13.85%)
Amean fault-huge-5 3896.10 ( 0.00%) 2486.89 * 36.17%*thpfioscale Percentage Faults Huge
4.20.0-rc1 4.20.0-rc1
vanilla lowzone-v2r4Percentage huge-5 95.02 ( 0.00%) 94.49 ( -0.56%)
In this case, there was both a reduction in the external fragmentation causing events and the huge page allocation success latency with little change in the allocation success rates which were already high. A careful reader will note that V1 had very different outcomes both in terms of the number of fragmentation events and the allocation success rates. In this case, it's due to the baseline including the THP __GFP_THISNODE removaal patch.
Overall, the patch significantly reduces the number of external fragmentation causing events so the success of THP over long periods of time would be improved for this adverse workload. While there are large differences compared to how V1 behaved, this is almost entirely accounted for by ac5b2c18911f ("mm: thp: relax __GFP_THISNODE for MADV_HUGEPAGE mappings").
--- mm/internal.h | 13 +++++--- mm/page_alloc.c | 100 +++++++++++++++++++++++++++++++++++++++++++++++++------- 2 files changed, 98 insertions(+), 15 deletions(-)
mm/internal.h +9 -4
@@ -480,10 +480,15 @@ unsigned long reclaim_clean_pages_from_list(struct zone *zone,#define ALLOC_OOM ALLOC_NO_WATERMARKS#endif -#define ALLOC_HARDER 0x10/* try to alloc harder */ -#define ALLOC_HIGH 0x20/* __GFP_HIGH set */ -#define ALLOC_CPUSET 0x40/* check for correct cpuset */ -#define ALLOC_CMA 0x80/* allow allocations from CMA areas */ +#define ALLOC_HARDER 0x10/* try to alloc harder */ +#define ALLOC_HIGH 0x20/* __GFP_HIGH set */ +#define ALLOC_CPUSET 0x40/* check for correct cpuset */ +#define ALLOC_CMA 0x80/* allow allocations from CMA areas */ +#ifdef CONFIG_ZONE_DMA32 +#define ALLOC_NOFRAGMENT 0x100/* avoid mixing pageblock types */ +#else +#define ALLOC_NOFRAGMENT 0x0 +#endif enum ttu_flags ;struct tlbflush_unmap_batch ;
mm/page_alloc.c +89 -11
212 lines not on this page · raw message
Signature · 1 line
2.16.4
PATCH 2/5 mm: Move zone watermark accesses behind an accessor
In reply to Mel GormanThis is a preparation patch only, no functional change.
--- include/linux/mmzone.h | 9 +++++---- mm/compaction.c | 2 +- mm/page_alloc.c | 12 ++++++------ 3 files changed, 12 insertions(+), 11 deletions(-)
include/linux/mmzone.h +5 -4
@@ -269,9 +269,10 @@ enum zone_watermarks {NR_WMARK }; -#define min_wmark_pages (z) (z->watermark[WMARK_MIN]) -#define low_wmark_pages (z) (z->watermark[WMARK_LOW]) -#define high_wmark_pages (z) (z->watermark[WMARK_HIGH]) +#define min_wmark_pages (z) (z->_watermark[WMARK_MIN]) +#define low_wmark_pages (z) (z->_watermark[WMARK_LOW]) +#define high_wmark_pages (z) (z->_watermark[WMARK_HIGH]) +#define wmark_pages (z, i) (z->_watermark[i])struct per_cpu_pages {int count;/* number of pages in the list */
@@ -362,7 +363,7 @@ struct zone {/* Read-mostly fields */ /* zone watermarks, access with *_wmark_pages(zone) macros */ -unsigned long watermark[NR_WMARK ]; +unsigned long _watermark[NR_WMARK ];unsigned long nr_reserved_highatomic;
mm/compaction.c +1 -1
@@ -1431,7 +1431,7 @@ static enum compact_result __compaction_suitable(struct zone *zone, int order,if (is_via_compact_memory (order))return COMPACT_CONTINUE ; - watermark = zone->watermark[alloc_flags &ALLOC_WMARK_MASK ]; + watermark =wmark_pages (zone, alloc_flags &ALLOC_WMARK_MASK );/* * If watermarks for high-order allocation are already met, there * should be no need for compaction at all.
mm/page_alloc.c +6 -6
@@ -3370,7 +3370,7 @@ get_page_from_freelist(gfp_t gfp_mask, unsigned int order, int alloc_flags, } } - mark = zone->watermark[alloc_flags &ALLOC_WMARK_MASK ]; + mark =wmark_pages (zone, alloc_flags &ALLOC_WMARK_MASK );if (!zone_watermark_fast (zone, order, mark,ac_classzone_idx (ac), alloc_flags)) {int ret;
@@ -4792,7 +4792,7 @@ long si_mem_available(void) pages[lru] =global_node_page_state (NR_LRU_BASE + lru);for_each_zone (zone) - wmark_low += zone->watermark[WMARK_LOW ]; + wmark_low +=low_wmark_pages (zone);/* * Estimate the amount of memory available for userspace allocations,
@@ -7418,13 +7418,13 @@ static void __setup_per_zone_wmarks(void) min_pages = zone->managed_pages /1024 ; min_pages =clamp (min_pages,SWAP_CLUSTER_MAX ,128UL ); - zone->watermark[WMARK_MIN ] = min_pages; + zone->_watermark[WMARK_MIN ] = min_pages; }else {/* * If it's a lowmem zone, reserve a number of pages * proportionate to the zone's size. */ - zone->watermark[WMARK_MIN ] = tmp; + zone->_watermark[WMARK_MIN ] = tmp; }/*
@@ -7436,8 +7436,8 @@ static void __setup_per_zone_wmarks(void)mult_frac (zone->managed_pages, watermark_scale_factor, 10000)); - zone->watermark[WMARK_LOW ] =min_wmark_pages (zone) + tmp; - zone->watermark[WMARK_HIGH ] =min_wmark_pages (zone) + tmp *2 ; + zone->_watermark[WMARK_LOW ] =min_wmark_pages (zone) + tmp; + zone->_watermark[WMARK_HIGH ] =min_wmark_pages (zone) + tmp *2 ;spin_unlock_irqrestore (&zone->lock, flags); }
Signature · 1 line
2.16.4
PATCH 3/5 mm: Reclaim small amounts of memory when an external fragmentation event occurs
In reply to Mel GormanAn external fragmentation event was previously described as
When the page allocator fragments memory, it records the event using the mm_page_alloc_extfrag event. If the fallback_order is smaller than a pageblock order (order-9 on 64-bit x86) then it's considered an event that will cause external fragmentation issues in the future.
The kernel reduces the probability of such events by increasing the watermark sizes by calling set_recommended_min_free_kbytes early in the lifetime of the system. This works reasonably well in general but if there are enough sparsely populated pageblocks then the problem can still occur as enough memory is free overall and kswapd stays asleep.
This patch introduces a watermark_boost_factor sysctl that allows a zone watermark to be temporarily boosted when an external fragmentation causing events occurs. The boosting will stall allocations that would decrease free memory below the boosted low watermark and kswapd is woken unconditionally to reclaim an amount of memory relative to the size of the high watermark and the watermark_boost_factor until the boost is cleared. When kswapd finishes, it wakes kcompactd at the pageblock order to clean some of the pageblocks that may have been affected by the fragmentation event. kswapd avoids any writeback or swap from reclaim context during this operation to avoid excessive system disruption in the name of fragmentation avoidance. Care is taken so that kswapd will do normal reclaim work if the system is really low on memory.
This was evaluated using the same workloads as "mm, page_alloc: Spread allocations across zones before introducing fragmentation".
1-socket Skylake machine config-global-dhp__workload_thpfioscale XFS (no special madvise) 4 fio threads, 1 THP allocating thread
--------------------------------------
4.20-rc1 extfrag events < order 9: 1023463
4.20-rc1+patch: 358574 (65% reduction)
4.20-rc1+patch1-3: 19274 (98% reduction)
4.20.0-rc1 4.20.0-rc1
lowzone-v2r4 boost-v2r4
Amean fault-base-1 663.65 ( 0.00%) 659.85 * 0.57%*
Amean fault-huge-1 0.00 ( 0.00%) 172.19 * -99.00%* 4.20.0-rc1 4.20.0-rc1
lowzone-v2r4 boost-v2r4Percentage huge-1 0.00 ( 0.00%) 1.68 ( 100.00%)
Note that external fragmentation causing events are massively reduced by this path whether in comparison to the previous kernel or the vanilla kernel. The fault latency for huge pages appears to be increased but that is only because THP allocations were successful with the patch applied.
1-socket Skylake machine global-dhp__workload_thpfioscale-madvhugepage-xfs (MADV_HUGEPAGE)
-----------------------------------------------------------------
4.20-rc1 extfrag events < order 9: 342549
4.20-rc1+patch: 337890 ( 1% reduction)
4.20-rc1+patch1-3: 12801 (96% reduction)
thpfioscale Fault Latencies
thpfioscale Fault Latencies
4.20.0-rc1 4.20.0-rc1
lowzone-v2r4 boost-v2r4
Amean fault-base-1 1531.37 ( 0.00%) 1578.91 ( -3.10%)
Amean fault-huge-1 1160.95 ( 0.00%) 1090.23 * 6.09%* 4.20.0-rc1 4.20.0-rc1
lowzone-v2r4 boost-v2r4Percentage huge-1 78.97 ( 0.00%) 82.59 ( 4.58%)
As before, massive reduction in external fragmentation events, some jitter on latencies and an increase in THP allocation success rates.
2-socket Haswell machine config-global-dhp__workload_thpfioscale XFS (no special madvise) 4 fio threads, 5 THP allocating threads
----------------------------------------------------------------
4.20-rc1 extfrag events < order 9: 209820
4.20-rc1+patch: 185923 (11% reduction)
4.20-rc1+patch1-3: 11240 (95% reduction)
4.20.0-rc1 4.20.0-rc1
lowzone-v2r4 boost-v2r4
Amean fault-base-5 1334.99 ( 0.00%) 1395.28 ( -4.52%)
Amean fault-huge-5 2428.43 ( 0.00%) 539.69 ( 77.78%) 4.20.0-rc1 4.20.0-rc1
lowzone-v2r4 boost-v2r4Percentage huge-5 1.13 ( 0.00%) 0.53 ( -52.94%)
This is an illustration of why latencies are not the primary metric. There is a 95% reduction in fragmentation causing events but the huge page latencies look fantastic until you account for the fact it might be because the success rate was lower. Given how low it was initially, this is partially down to luck.
2-socket Haswell machine global-dhp__workload_thpfioscale-madvhugepage-xfs (MADV_HUGEPAGE)
-----------------------------------------------------------------
4.20-rc1 extfrag events < order 9: 167464
4.20-rc1+patch: 130081 (22% reduction)
4.20-rc1+patch1-3: 12057 (92% reduction)
thpfioscale Fault Latencies
4.20.0-rc1 4.20.0-rc1
lowzone-v2r4 boost-v2r4
Amean fault-base-5 6652.67 ( 0.00%) 8691.83 * -30.65%*
Amean fault-huge-5 2486.89 ( 0.00%) 2899.83 * -16.60%* 4.20.0-rc1 4.20.0-rc1
lowzone-v2r4 boost-v2r4Percentage huge-5 94.49 ( 0.00%) 95.55 ( 1.13%)
There is a large reduction in fragmentation events with a very slightly higher THP allocation success rate. The latencies look bad but a closer look at the data seems to indicate the problem is at the tails. Given the high THP allocation success rate, the system is under quite some pressure.
--- Documentation/sysctl/vm.txt | 19 +++++++ include/linux/mm.h | 1 + include/linux/mmzone.h | 11 ++-- kernel/sysctl.c | 8 +++ mm/page_alloc.c | 53 +++++++++++++++++-- mm/vmscan.c | 123 ++++++++++++++++++++++++++++++++++++++++---- 6 files changed, 199 insertions(+), 16 deletions(-)
Documentation/sysctl/vm.txt +19 -0
@@ -63,6 +63,7 @@ files can be found in mm/swap.c. - swappiness - user_reserve_kbytes - vfs_cache_pressure +- watermark_boost_factor - watermark_scale_factor - zone_reclaim_mode
@@ -856,6 +857,24 @@ ten times more freeable objects than there are. ============================================================= +watermark_boost_factor: + +This factor controls the level of reclaim when memory is being fragmented. +It defines the percentage of the low watermark of a zone that will be +reclaimed if pages of different mobility are being mixed within pageblocks. +The intent is so that compaction has less work to do and increase the +success rate of future high-order allocations such as SLUB allocations, +THP and hugetlbfs pages. + +To make it sensible with respect to the watermark_scale_factor parameter, +the unit is in fractions of 10,000. The default value of 15000 means +that 150% of the high watermark will be reclaimed in the event of a +pageblock being mixed due to fragmentation. If this value is smaller +than a pageblock then a pageblocks worth of pages will be reclaimed (e.g. +2MB on 64-bit x86). A boost factor of 0 will disable the feature. + +============================================================= + watermark_scale_factor: This factor controls the aggressiveness of kswapd. It defines the
include/linux/mm.h +1 -0
@@ -2194,6 +2194,7 @@ extern void zone_pcp_reset(struct zone *zone);/* page_alloc.c */ extern int min_free_kbytes; +extern int watermark_boost_factor;extern int watermark_scale_factor;/* nommu.c */
include/linux/mmzone.h +7 -4
@@ -269,10 +269,10 @@ enum zone_watermarks {NR_WMARK }; -#define min_wmark_pages (z) (z->_watermark[WMARK_MIN]) -#define low_wmark_pages (z) (z->_watermark[WMARK_LOW]) -#define high_wmark_pages (z) (z->_watermark[WMARK_HIGH]) -#define wmark_pages (z, i) (z->_watermark[i]) +#define min_wmark_pages (z) (z->_watermark[WMARK_MIN] + z->watermark_boost) +#define low_wmark_pages (z) (z->_watermark[WMARK_LOW] + z->watermark_boost) +#define high_wmark_pages (z) (z->_watermark[WMARK_HIGH] + z->watermark_boost) +#define wmark_pages (z, i) (z->_watermark[i] + z->watermark_boost)struct per_cpu_pages {int count;/* number of pages in the list */
@@ -364,6 +364,7 @@ struct zone {/* zone watermarks, access with *_wmark_pages(zone) macros */ unsigned long _watermark[NR_WMARK ]; +unsigned long watermark_boost;unsigned long nr_reserved_highatomic;
@@ -885,6 +886,8 @@ static inline int is_highmem(struct zone *zone)struct ctl_table ;int min_free_kbytes_sysctl_handler (struct ctl_table *,int ,void __user *,size_t *,loff_t *); +int watermark_boost_factor_sysctl_handler (struct ctl_table *,int , +void __user *,size_t *,loff_t *);int watermark_scale_factor_sysctl_handler (struct ctl_table *,int ,void __user *,size_t *,loff_t *);extern int sysctl_lowmem_reserve_ratio[MAX_NR_ZONES ];
kernel/sysctl.c +8 -0
@@ -1462,6 +1462,14 @@ static struct ctl_table vm_table[] = { .proc_handler = min_free_kbytes_sysctl_handler, .extra1 = &zero, }, + { + .procname = "watermark_boost_factor", + .data = &watermark_boost_factor, + .maxlen = sizeof(watermark_boost_factor), + .mode = 0644, + .proc_handler = watermark_boost_factor_sysctl_handler, + .extra1 = &zero, + }, { .procname = "watermark_scale_factor", .data = &watermark_scale_factor,
mm/page_alloc.c +50 -3
105 lines not on this page · raw message
mm/vmscan.c +114 -9
184 lines not on this page · raw message
Signature · 1 line
2.16.4
PATCH 4/5 mm: Stall movable allocations until kswapd progresses during serious external fragmentation event
In reply to Mel GormanAn event that potentially causes external fragmentation problems has already been described but there are degrees of severity. A "serious" event is defined as one that steals a contiguous range of pages of an order lower than fragment_stall_order (PAGE_ALLOC_COSTLY_ORDER by default). If a movable allocation request that is allowed to sleep needs to steal a small block then it schedules until kswapd makes progress or a timeout passes. The watermarks are also boosted slightly faster so that kswapd makes greater effort to reclaim enough pages to avoid the fragmentation event.
This stall is not guaranteed to avoid serious fragmentation events. If memory pressure is high enough, the pages freed by kswapd may be reallocated or the free pages may not be in pageblocks that contain only movable pages. Furthermore an allocation request that cannot stall (e.g. atomic allocations) or unmovable/reclaimable allocations will still proceed without stalling.
The worst-case scenario for stalling is a combination of both high memory pressure where kswapd is having trouble keeping free pages over the pfmemalloc_reserve and movable allocations are fragmenting memory. In this case, an allocation request may sleep for longer. There are both vmstats to identify stalls are happening and a tracepoint to quantify what the stall durations are. Note that the granularity of the stall detection is a jiffy so the delay accounting is not precise.
1-socket Skylake machine config-global-dhp__workload_thpfioscale XFS (no special madvise) 4 fio threads, 1 THP allocating thread
--------------------------------------
4.20-rc1 extfrag events < order 9: 1023463
4.20-rc1+patch: 358574 (65% reduction)
4.20-rc1+patch1-3: 19274 (98% reduction)
4.20-rc1+patch1-4: 1351 (99.9% reduction)
4.20.0-rc1 4.20.0-rc1
boost-v2r4 stall-v2r6
Amean fault-base-1 659.85 ( 0.00%) 648.66 * 1.70%*
Amean fault-huge-1 172.19 ( 0.00%) 167.79 ( 2.56%)thpfioscale Percentage Faults Huge
4.20.0-rc1 4.20.0-rc1
boost-v2r4 stall-v2r6Percentage huge-1 1.68 ( 0.00%) 1.16 ( -30.69%)
Fragmentation events are now reduced to negligible levels.
The latencies and allocation success rates are roughly similar. Over the course of 16 minutes, there were 100 stalls due to fragmentation avoidance with a total stall time of 0.4 seconds.
1-socket Skylake machine global-dhp__workload_thpfioscale-madvhugepage-xfs (MADV_HUGEPAGE)
-----------------------------------------------------------------
4.20-rc1 extfrag events < order 9: 342549
4.20-rc1+patch: 337890 ( 1% reduction)
4.20-rc1+patch1-3: 12801 (96% reduction)
4.20-rc1+patch1-4: 1117 (99.7% reduction)
4.20.0-rc1 4.20.0-rc1
boost-v2r4 stall-v2r6
Amean fault-base-1 1578.91 ( 0.00%) 43404.60 (-2649.02%)
Amean fault-huge-1 1090.23 ( 0.00%) 1424.32 * -30.64%* 4.20.0-rc1 4.20.0-rc1
boost-v2r4 stall-v2r6Percentage huge-1 82.59 ( 0.00%) 99.92 ( 20.97%)
The fragmentation events were reduced but the latencies went a bit crazy. The "problem" is that the allocation success rates were very high and forward progress was being made. This put the system under further pressure and while compactions were succeeding, the latencies were high in cases where compaction failed. The THP allocation vm stats are illustrative in this case
4.20.0-rc1 4.20.0-rc1
boost-v2r4 stall-v2r6
THP fault alloc 4974 6016
THP fault fallback 1048 5
THP collapse alloc 65 56
THP collapse fail 4 4
THP split 0 3719
THP split failed 0 224Note the THP fault alloc stats where they almost all succeeded relative to the baseline. While the latencies are much higher, it is the case that the application specifically requested THP while the system was under heavy memory pressure.
There were 314 stalls over the course of 16 minutes for a total stall time of roughly 11 seconds. The distribution of stalls is as follows
205 4000
1 8000
1 20000
1 32000
1 36000
6 40000
1 56000
98 100000This is showing that 98 of the stalls waited until the timeout expired at 25 jiffies which 100000 microseconds on this particular configuration. If this is considered problematic, the timeout can be reduced to tradeoff fault times against fragmentation avoidance.
2-socket Haswell machine config-global-dhp__workload_thpfioscale XFS (no special madvise) 4 fio threads, 5 THP allocating threads
----------------------------------------------------------------
4.20-rc1 extfrag events < order 9: 209820
4.20-rc1+patch: 185923 (11% reduction)
4.20-rc1+patch1-3: 11240 (95% reduction)
4.20-rc1+patch1-4: 13241 (93% reduction)
4.20.0-rc1 4.20.0-rc1
boost-v2r4 stall-v2r6
Amean fault-base-5 1395.28 ( 0.00%) 1508.94 * -8.15%*
Amean fault-huge-5 539.69 ( 0.00%) 614.88 * -13.93%* 4.20.0-rc1 4.20.0-rc1
boost-v2r4 stall-v2r6Percentage huge-5 0.53 ( 0.00%) 3.38 ( 534.38%
There is a slight increase in fragmentation events but given that it's already heavily reduced, there are elements of luck. There is a small increase in latencies which is partially offset by a slight increase in THP allocation success rates. There were 65 stalls over the course of 87 minutes with stall time of a total of roughly 0.4 milliseconds.
2-socket Haswell machine global-dhp__workload_thpfioscale-madvhugepage-xfs (MADV_HUGEPAGE)
-----------------------------------------------------------------
4.20-rc1 extfrag events < order 9: 167464
4.20-rc1+patch: 130081 (22% reduction)
4.20-rc1+patch1-3: 12057 (92% reduction)
4.20-rc1+patch1-4: 11060 (93% reduction)
thpfioscale Fault Latencies
4.20.0-rc1 4.20.0-rc1
boost-v2r4 stall-v2r6
Amean fault-base-5 8691.83 ( 0.00%) 9363.89 ( -7.73%)
Amean fault-huge-5 2899.83 ( 0.00%) 3638.29 * -25.47%* 4.20.0-rc1 4.20.0-rc1
boost-v2r4 stall-v2r6Percentage huge-5 95.55 ( 0.00%) 99.27 ( 3.89%)
The fragmentation events are reduced and while there is some hit on the latency, the success rate is near 100% while under heavy pressure. There were 2486 stalls over the course of 85 minutes with a total stall time of roughly 12 seconds.
This patch does reduce fragmentation rates overall but it's not free as some allocataions can stall for short periods of time and there are knock-on effects to latency when THP allocation success rates are higher. While it's within acceptable limits for the adverse test case, there may be other workloads that cannot tolerate the stalls. If this occurs, it can be tuned to disable the feature or more ideally, the test case is made available for analysis to see if the stall behaviour can be reduced while still limiting the fragmentation events. On the flip-side, it has been checked that setting the fragment_stall_order to 9 eliminated fragmentation events entirely.
--- Documentation/sysctl/vm.txt | 23 +++++++++++ include/linux/mm.h | 1 + include/linux/mmzone.h | 2 + include/linux/vm_event_item.h | 1 + include/trace/events/kmem.h | 21 ++++++++++ kernel/sysctl.c | 10 +++++ mm/internal.h | 1 + mm/page_alloc.c | 94 +++++++++++++++++++++++++++++++++++++------ mm/vmstat.c | 1 + 9 files changed, 142 insertions(+), 12 deletions(-)
Documentation/sysctl/vm.txt +23 -0
@@ -31,6 +31,7 @@ files can be found in mm/swap.c. - dirty_writeback_centisecs - drop_caches - extfrag_threshold +- fragment_stall_order - hugetlb_shm_group - laptop_mode - legacy_va_layout
@@ -275,6 +276,28 @@ any throttling. ============================================================== +fragment_stall_order + +External fragmentation control is managed on a pageblock level where the +page allocator tries to avoid mixing pages of different mobility within page +blocks (e.g. order 9 on 64-bit x86). If external fragmentation is perfectly +controlled then a THP allocation will often succeed up to the number of +movable pageblocks in the system as reported by /proc/pagetypeinfo. + +When memory is low, the system may have to mix pageblocks and will wake +kswapd to try control future fragmentation. fragment_stall_order controls if +the allocating task will stall if possible until kswapd makes some progress +in preference to fragmenting the system. This incurs a small stall penalty +in exchange for future success at allocating huge pages. If the stalls +are undesirable and high-order allocations are irrelevant then this can +be disabled by writing 0 to the tunable. Writing the pageblock order will +strongly (but not perfectly) control external fragmentation. + +The default will stall for fragmenting allocations smaller than the +PAGE_ALLOC_COSTLY_ORDER (defined as order-3 at the time of writing). + +============================================================== + hugetlb_shm_group hugetlb_shm_group contains group id that is allowed to create SysV
include/linux/mm.h +1 -0
@@ -2196,6 +2196,7 @@ extern void zone_pcp_reset(struct zone *zone);extern int min_free_kbytes;extern int watermark_boost_factor;extern int watermark_scale_factor; +extern int fragment_stall_order;/* nommu.c */ extern atomic_long_t mmap_pages_allocated;
include/linux/mmzone.h +2 -0
@@ -890,6 +890,8 @@ int watermark_boost_factor_sysctl_handler(struct ctl_table *, int, void __user *, size_t *, loff_t *);int watermark_scale_factor_sysctl_handler (struct ctl_table *,int ,void __user *,size_t *,loff_t *); +int fragment_stall_order_sysctl_handler (struct ctl_table *,int , +void __user *,size_t *,loff_t *);extern int sysctl_lowmem_reserve_ratio[MAX_NR_ZONES ];int lowmem_reserve_ratio_sysctl_handler (struct ctl_table *,int ,void __user *,size_t *,loff_t *);
include/linux/vm_event_item.h +1 -0
@@ -43,6 +43,7 @@ enum vm_event_item { PGPGIN, PGPGOUT, PSWPIN, PSWPOUT,PAGEOUTRUN ,PGROTATED ,DROP_PAGECACHE ,DROP_SLAB ,OOM_KILL , +FRAGMENTSTALL , #ifdef CONFIG_NUMA_BALANCINGNUMA_PTE_UPDATES , NUMA_HUGE_PTE_UPDATES,
include/trace/events/kmem.h +21 -0
@@ -315,6 +315,27 @@ TRACE_EVENT(mm_page_alloc_extfrag, __entry->change_ownership) ); +TRACE_EVENT (mm_fragmentation_stall, + + TP_PROTO(int nid, unsigned long duration), + +TP_ARGS (nid, duration), + +TP_STRUCT__entry ( + __field( int, nid ) + __field( unsigned long, duration ) + ), + +TP_fast_assign ( + __entry->nid = nid; + __entry->duration = duration + ), + +TP_printk ("nid=%d duration=%lu" , + __entry->nid, + __entry->duration) +); +#endif /* _TRACE_KMEM_H */ /* This part must be outside protection */
kernel/sysctl.c +10 -0
@@ -126,6 +126,7 @@ static int zero;static int __maybe_unused one =1 ;static int __maybe_unused two =2 ;static int __maybe_unused four =4 ; +static int __maybe_unused max_order =MAX_ORDER ;static unsigned long one_ul =1 ;static int one_hundred =100 ;static int one_thousand =1000 ;
@@ -1479,6 +1480,15 @@ static struct ctl_table vm_table[] = { .extra1 = &one, .extra2 = &one_thousand, }, + { + .procname = "fragment_stall_order", + .data = &fragment_stall_order, + .maxlen = sizeof(fragment_stall_order), + .mode = 0644, + .proc_handler = fragment_stall_order_sysctl_handler, + .extra1 = &zero, + .extra2 = &max_order, + }, { .procname = "percpu_pagelist_fraction", .data = &percpu_pagelist_fraction,
mm/internal.h +1 -0
@@ -489,6 +489,7 @@ unsigned long reclaim_clean_pages_from_list(struct zone *zone,#else #define ALLOC_NOFRAGMENT 0x0#endif +#define ALLOC_FRAGMENT_STALL 0x200/* stall if fragmenting heavily */ enum ttu_flags ;struct tlbflush_unmap_batch ;
mm/page_alloc.c +82 -12
204 lines not on this page · raw message
mm/vmstat.c +1 -0
@@ -1211,6 +1211,7 @@ const char * const vmstat_text[] = {"drop_pagecache" ,"drop_slab" ,"oom_kill" , +"fragment_stall" , #ifdef CONFIG_NUMA_BALANCING "numa_pte_updates",
Signature · 1 line
2.16.4
PATCH 5/5 mm: Target compaction on pageblocks that were recently fragmented
In reply to Mel GormanDespite the earlier patches, external fragmentation events are still inevitable as not all callers can stall or are appropriate to stall. In the event the result is a mixed pageblock, it is desirable to move all movable pages from that block so that unmovable/unreclaimable allocations do not further pollute the address space. This patch queues such pageblocks for early compaction and relies on kswapd to wake kcompactd when some pages are reclaimed. Waking kcompactd after kswapd makes progress is so that the compaction is more likely to have a suitable migration destination.
This patch may be controversial as there are multiple other design decisions that can be made. We could refuse to change pageblock ownership in some cases but great care would need to be taken to avoid premature OOMs or a livelock. Similarly, we could tag pageblocks as mixed and search for them but that would increase scanning costs. Finally, there is a corner case that a mixed pageblock that is after the point where a free scanner can operate may fail to clean the pageblock but addressing that would require a fundamental alteration to how compaction works.
Unlike the previous patches, the benefit here is harder to quantify as any work that is queued may or may not help an allocation request in the future. The timing of the allocation stream is critical and detecting differences in latency may be within the noise. Hence, the potential benefit of this patch is more conceptual than quantitive even though there are some positive results.
1-socket Skylake machine config-global-dhp__workload_thpfioscale XFS (no special madvise) 4 fio threads, 1 THP allocating thread
--------------------------------------
4.20-rc1 extfrag events < order 9: 1023463
4.20-rc1+patch: 358574 (65% reduction)
4.20-rc1+patch1-3: 19274 (98% reduction)
4.20-rc1+patch1-4: 1351 (99.9% reduction)
4.20-rc1+patch1-5: 2554 (99.8% reduction)
4.20.0-rc1 4.20.0-rc1
stall-v2r6 proactive-v2r6
Amean fault-base-1 648.66 ( 0.00%) 655.18 * -1.00%*
Amean fault-huge-1 167.79 ( 0.00%) 163.00 ( 2.85%) 4.20.0-rc1 4.20.0-rc1
stall-v2r6 proactive-v2r6Percentage huge-1 1.16 ( 0.00%) 0.03 ( -97.14%)
The performance is similar but not necessarily indicative that the patch had any effect. There was no reported compaction activity so essentially the patch was a no-op.
1-socket Skylake machine global-dhp__workload_thpfioscale-madvhugepage-xfs (MADV_HUGEPAGE)
-----------------------------------------------------------------
4.20-rc1 extfrag events < order 9: 342549
4.20-rc1+patch: 337890 ( 1% reduction)
4.20-rc1+patch1-3: 12801 (96% reduction)
4.20-rc1+patch1-4: 1511 (99.7% reduction)
4.20.0-rc1 4.20.0-rc1
stall-v2r6 proactive-v2r6
Amean fault-base-1 43404.60 ( 0.00%) 0.00 ( 100.00%)
Amean fault-huge-1 1424.32 ( 0.00%) 540.99 * 62.02%* 4.20.0-rc1 4.20.0-rc1
stall-v2r6 proactive-v2r6Percentage huge-1 99.92 ( 0.00%) 100.00 ( 0.08%)
Slight increase in fragmentation events but the latency was improved and THP allocations had a 100% success rate.
2-socket Haswell machine config-global-dhp__workload_thpfioscale XFS (no special madvise) 4 fio threads, 5 THP allocating threads
----------------------------------------------------------------
4.20-rc1 extfrag events < order 9: 209820
4.20-rc1+patch: 185923 (11% reduction)
4.20-rc1+patch1-3: 11240 (95% reduction) 4.20-rc1+patch1-4: 13241 (93% reduction) 4.20-rc1+patch1-5: 11916 (94% reduction)
thpfioscale Fault Latencies
4.20.0-rc1 4.20.0-rc1
stall-v2r6 proactive-v2r6
Amean fault-base-5 1508.94 ( 0.00%) 1545.56 ( -2.43%)
Amean fault-huge-5 614.88 ( 0.00%) 557.46 * 9.34%* 4.20.0-rc1 4.20.0-rc1
stall-v2r6 proactive-v2r6Percentage huge-5 3.38 ( 0.00%) 4.53 ( 33.99%)
Fragmentation-causing events are slightly reduced and there is a slight improvement in THP allocation latencies and success rates. Remember that no special effort is being made to allocate THP in this workload.
2-socket Haswell machine global-dhp__workload_thpfioscale-madvhugepage-xfs (MADV_HUGEPAGE)
-----------------------------------------------------------------
4.20-rc1 extfrag events < order 9: 167464
4.20-rc1+patch: 130081 (22% reduction)
4.20-rc1+patch1-3: 12057 (92% reduction)
4.20-rc1+patch1-4: 11060 (93% reduction)
4.20-rc1+patch1-5: 8903 (95% reduction)
4.20.0-rc1 4.20.0-rc1
stall-v2r6 proactive-v2r6
Amean fault-base-5 9363.89 ( 0.00%) 9067.00 ( 3.17%)
Amean fault-huge-5 3638.29 ( 0.00%) 1509.51 * 58.51%*thpfioscale Percentage Faults Huge
4.20.0-rc1 4.20.0-rc1
stall-v2r6 proactive-v2r6Percentage huge-5 99.27 ( 0.00%) 99.93 ( 0.67%)
There is a small decrease in fragmentation events but the most notable part is the decrease in latency with a similarly high THP allocation success rate.
It is less obvious whether this is a universal win as fragmentation-causing events were already low and in the case of MADV_HUGEPAGE, the allocation success rates were already high. However, it's encouraging that the THP allocation latencies were improved.
--- include/linux/compaction.h | 4 ++ include/linux/migrate.h | 7 +- include/linux/mmzone.h | 4 ++ include/trace/events/compaction.h | 62 ++++++++++++++++ mm/compaction.c | 145 +++++++++++++++++++++++++++++++++++--- mm/migrate.c | 6 +- mm/page_alloc.c | 7 ++ 7 files changed, 224 insertions(+), 11 deletions(-)
include/linux/compaction.h +4 -0
@@ -177,6 +177,7 @@ 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 kcompactd_queue_migration (struct zone *zone,struct page *page);#else static inline void reset_isolation_suitable (pg_data_t *pgdat)
@@ -225,6 +226,9 @@ static inline void wakeup_kcompactd(pg_data_t *pgdat, int order, int classzone_i { } +static inline void kcompactd_queue_migration (struct zone *zone,struct page *page) +{ +}#endif /* CONFIG_COMPACTION */ #if defined(CONFIG_COMPACTION ) && defined(CONFIG_SYSFS ) && defined(CONFIG_NUMA )
include/linux/migrate.h +5 -2
@@ -61,7 +61,7 @@ static inline struct page *new_page_nodemask(struct page *page,#ifdef CONFIG_MIGRATION -extern void putback_movable_pages (struct list_head *l); +extern unsigned int putback_movable_pages (struct list_head *l);extern int migrate_page (struct address_space *mapping,struct page *newpage,struct page *page,enum migrate_mode mode);
@@ -82,7 +82,10 @@ extern int migrate_page_move_mapping(struct address_space *mapping, int extra_count);#else -static inline void putback_movable_pages (struct list_head *l) {} +static inline unsigned int putback_movable_pages (struct list_head *l) +{ +return 0 ; +}static inline int migrate_pages (struct list_head *l,new_page_t new,free_page_t free,unsigned long private,enum migrate_mode mode,int reason)
include/linux/mmzone.h +4 -0
@@ -497,6 +497,10 @@ struct zone {unsigned int compact_considered;unsigned int compact_defer_shift;int compact_order_failed; + +#define COMPACT_QUEUE_LENGTH 16 +unsigned long compact_queue[COMPACT_QUEUE_LENGTH ]; +int nr_compact;#endif #if definedCONFIG_COMPACTION || definedCONFIG_CMA
include/trace/events/compaction.h +62 -0
68 lines not on this page · raw message
mm/compaction.c +137 -8
193 lines not on this page · raw message
mm/migrate.c +5 -1
@@ -164,12 +164,14 @@ void putback_movable_page(struct page *page)* built from lru, balloon, hugetlbfs page. See isolate_migratepages_range() * and isolate_huge_page(). */ -void putback_movable_pages (struct list_head *l) +unsigned int putback_movable_pages (struct list_head *l) {struct page *page;struct page *page2; +unsigned int nr_putback =0 ;list_for_each_entry_safe (page, page2, l, lru) { + nr_putback++;if (unlikely (PageHuge (page))) {putback_active_hugepage (page);continue ;
@@ -195,6 +197,8 @@ void putback_movable_pages(struct list_head *l)putback_lru_page (page); } } + +return nr_putback; }/*
mm/page_alloc.c +7 -0
@@ -2214,6 +2214,9 @@ static bool steal_suitable_fallback(struct zone *zone, struct page *page,boost_watermark (zone, false);wakeup_kswapd (zone,0 ,0 ,zone_idx (zone)); +if (start_type ==MIGRATE_MOVABLE || old_block_type ==MIGRATE_MOVABLE ) +kcompactd_queue_migration (zone, page); +if ((alloc_flags &ALLOC_FRAGMENT_STALL ) && current_order < fragment_stall_order) {return false;
@@ -6457,7 +6460,11 @@ static void pgdat_init_split_queue(struct pglist_data *pgdat) {}#ifdef CONFIG_COMPACTION static void pgdat_init_kcompactd (struct pglist_data *pgdat) { +int i; +init_waitqueue_head (&pgdat->kcompactd_wait); +for (i =0 ; i <MAX_NR_ZONES ; i++) + pgdat->node_zones[i].nr_compact =0 ; }#else static void pgdat_init_kcompactd (struct pglist_data *pgdat) {}
Signature · 1 line
2.16.4