NUMA balancing: optimize memory placement for memory tiering system
Subsystems (10)
cover NUMA balancing: optimize memory placement for memory tiering system
The changes since the last post are as follows,
- Rebased on v5.16-rc1
- Revise error processing for [1/6] (promotion counter) per Yang's comments
- Add sysctl document for [2/6] (optimize page placement)
- Reset threshold adjustment state when disable/enable tiering mode
- Reset threshold when workload transition is detected.
--
With the advent of various new memory types, some machines will have multiple types of memory, e.g. DRAM and PMEM (persistent memory). The memory subsystem of these machines can be called memory tiering system, because the performance of the different types of memory are different.
After commit c221c0b0308f ("device-dax: "Hotplug" persistent memory for use like normal RAM"), the PMEM could be used as the cost-effective volatile memory in separate NUMA nodes. In a typical memory tiering system, there are CPUs, DRAM and PMEM in each physical NUMA node. The CPUs and the DRAM will be put in one logical node, while the PMEM will be put in another (faked) logical node.
To optimize the system overall performance, the hot pages should be placed in DRAM node. To do that, we need to identify the hot pages in the PMEM node and migrate them to DRAM node via NUMA migration.
In the original NUMA balancing, there are already a set of existing mechanisms to identify the pages recently accessed by the CPUs in a node and migrate the pages to the node. So we can reuse these mechanisms to build the mechanisms to optimize the page placement in the memory tiering system. This is implemented in this patchset.
At the other hand, the cold pages should be placed in PMEM node. So, we also need to identify the cold pages in the DRAM node and migrate them to PMEM node.
In commit 26aa2d199d6f ("mm/migrate: demote pages during reclaim"), a mechanism to demote the cold DRAM pages to PMEM node under memory pressure is implemented. Based on that, the cold DRAM pages can be demoted to PMEM node proactively to free some memory space on DRAM node to accommodate the promoted hot PMEM pages. This is implemented in this patchset too.
We have tested the solution with the pmbench memory accessing benchmark with the 80:20 read/write ratio and the normal access address distribution on a 2 socket Intel server with Optane DC Persistent Memory Model. The test results of the base kernel and step by step optimizations are as follows,
Throughput Promotion DRAM bandwidth
access/s MB/s MB/s
----------- ---------- --------------
Base 69263986.8 1830.2
Patch 2 135691921.4 385.6 11315.9 Patch 3 133239016.8 384.7 11065.2 Patch 4 151310868.9 197.6 11397.0 Patch 5 142311252.8 99.3 9580.8 Patch 6 149044263.9 65.5 9922.8
The whole patchset improves the benchmark score up to 115.2%. The basic NUMA balancing based optimization solution (patch 2), the hot page selection algorithm (patch 4), and the threshold automatic adjustment algorithms (patch 6) improves the performance or reduce the overhead (promotion MB/s) greatly.
Changelog:
v10:
- Rebased on v5.16-rc1
- Revise error processing for [1/6] (promotion counter) per Yang's comments
- Add sysctl document for [2/6] (optimize page placement)
- Reset threshold adjustment state when disable/enable tiering mode
- Reset threshold when workload transition is detected.
v9:
- Rebased on v5.15-rc4
- Make "add promotion counter" the first patch per Yang's comments
v8:
- Rebased on v5.15-rc1
- Make user-specified threshold take effect sooner
v7:
- Rebased on the mmots tree of 2021-07-15.
- Some minor fixes.
v6:
- Rebased on the latest page demotion patchset. (which bases on v5.11)
v5:
- Rebased on the latest page demotion patchset. (which bases on v5.10)
v4:
- Rebased on the latest page demotion patchset. (which bases on v5.9-rc6)
- Add page promotion counter.
v3:
- Move the rate limit control as late as possible per Mel Gorman's comments.
- Revise the hot page selection implementation to store page scan time in struct page.
- Code cleanup.
- Rebased on the latest page demotion patchset.
v2:
- Addressed comments for V1.
- Rebased on v5.5.
Best Regards,
Huang, Ying
PATCH RESEND 1/6 NUMA Balancing: add page promotion counter
In a system with multiple memory types, e.g. DRAM and PMEM, the CPU and DRAM in one socket will be put in one NUMA node as before, while the PMEM will be put in another NUMA node as described in the description of the commit c221c0b0308f ("device-dax: "Hotplug" persistent memory for use like normal RAM"). So, the NUMA balancing mechanism will identify all PMEM accesses as remote access and try to promote the PMEM pages to DRAM.
To distinguish the number of the inter-type promoted pages from that of the inter-socket migrated pages. A new vmstat count is added. The counter is per-node (count in the target node). So this can be used to identify promotion imbalance among the NUMA nodes.
--- include/linux/mmzone.h | 3 +++ include/linux/node.h | 5 +++++ mm/migrate.c | 13 ++++++++++--- mm/vmstat.c | 3 +++ 4 files changed, 21 insertions(+), 3 deletions(-)
include/linux/mmzone.h +3 -0
@@ -210,6 +210,9 @@ enum node_stat_item { NR_PAGETABLE, /* used for pagetables */#ifdef CONFIG_SWAP NR_SWAPCACHE, +#endif +#ifdef CONFIG_NUMA_BALANCING +PGPROMOTE_SUCCESS ,/* promote successfully */ #endifNR_VM_NODE_STAT_ITEMS };
include/linux/node.h +5 -0
@@ -181,4 +181,9 @@ static inline void register_hugetlbfs_with_node(node_registration_func_t reg,#define to_node (device) container_of(device, struct node, dev) +static inline bool node_is_toptier (int node) +{ +return node_state (node,N_CPU ); +} +#endif /* _LINUX_NODE_H_ */
mm/migrate.c +10 -3
@@ -2141,6 +2141,7 @@ int migrate_misplaced_page(struct page *page, struct vm_area_struct *vma,pg_data_t *pgdat =NODE_DATA (node);int isolated;int nr_remaining; +int nr_succeeded;LIST_HEAD (migratepages);new_page_t *new;bool compound;
@@ -2179,7 +2180,8 @@ int migrate_misplaced_page(struct page *page, struct vm_area_struct *vma,list_add (&page->lru, &migratepages); nr_remaining =migrate_pages (&migratepages, *new,NULL , node, -MIGRATE_ASYNC ,MR_NUMA_MISPLACED ,NULL ); +MIGRATE_ASYNC ,MR_NUMA_MISPLACED , + &nr_succeeded);if (nr_remaining) {if (!list_empty (&migratepages)) {list_del (&page->lru);
@@ -2188,8 +2190,13 @@ int migrate_misplaced_page(struct page *page, struct vm_area_struct *vma,putback_lru_page (page); } isolated =0 ; - }else -count_vm_numa_events (NUMA_PAGE_MIGRATE , nr_pages); + } +if (nr_succeeded) { +count_vm_numa_events (NUMA_PAGE_MIGRATE , nr_succeeded); +if (!node_is_toptier (page_to_nid (page)) &&node_is_toptier (node)) +mod_node_page_state (NODE_DATA (node),PGPROMOTE_SUCCESS , + nr_succeeded); + }BUG_ON (!list_empty (&migratepages));return isolated;
mm/vmstat.c +3 -0
@@ -1242,6 +1242,9 @@ const char * const vmstat_text[] = {#ifdef CONFIG_SWAP "nr_swapcached",#endif +#ifdef CONFIG_NUMA_BALANCING +"pgpromote_success" , +#endif/* enum writeback_stat_item counters */ "nr_dirty_threshold",
Signature · 1 line
2.30.2
Re: [PATCH -V10 RESEND 1/6] NUMA Balancing: add page promotion counter
Hi Huang,
include/linux/mmzone.h enum node_stat_item { · line 215
Quoted 2 lines
include/linux/mmzone.h +2
-0+#ifdef CONFIG_NUMA_BALANCING + PGPROMOTE_SUCCESS, /* promote successfully */
I find a breakdown of Anon and File page promotion can often be useful to understand an application's behavior (i.e. what kind of pages are moved to remote node and later being promoted). What do you think about adding counters for such a breakdown?
What's your thought on adding counters for failures on different reasons?
Best,
Hasan
Re: [PATCH -V10 RESEND 1/6] NUMA Balancing: add page promotion counter
Quoted 11 lines · Hasan Al Maruf
Hi Huang,
include/linux/mmzone.h enum node_stat_item { · line 215
Quoted 2 lines
include/linux/mmzone.h +2
-0+#ifdef CONFIG_NUMA_BALANCING + PGPROMOTE_SUCCESS, /* promote successfully */I find a breakdown of Anon and File page promotion can often be useful to understand an application's behavior (i.e. what kind of pages are moved to remote node and later being promoted). What do you think about adding counters for such a breakdown?
What's your thought on adding counters for failures on different reasons?
I think that all these provide helpful information. But I think that we can add them in separate patches. That will make reviewing simpler.
Best Regards,
Huang, Ying
Re: [PATCH -V10 RESEND 1/6] NUMA Balancing: add page promotion counter
In reply to Huang YingQuoted 28 lines · Huang Ying
In a system with multiple memory types, e.g. DRAM and PMEM, the CPU and DRAM in one socket will be put in one NUMA node as before, while the PMEM will be put in another NUMA node as described in the description of the commit
c221c0b0308f("device-dax: "Hotplug" persistent memory for use like normal RAM"). So, the NUMA balancing mechanism will identify all PMEM accesses as remote access and try to promote the PMEM pages to DRAM.To distinguish the number of the inter-type promoted pages from that of the inter-socket migrated pages. A new vmstat count is added. The counter is per-node (count in the target node). So this can be used to identify promotion imbalance among the NUMA nodes.
Signed-off-by "Huang, Ying" <[email protected]>Reviewed-by Yang Shi <[email protected]>Cc Andrew Morton <[email protected]>Cc Michal Hocko <[email protected]>Cc Rik van Riel <[email protected]>Cc Mel Gorman <[email protected]>Cc Peter Zijlstra <[email protected]>Cc Dave Hansen <[email protected]>Cc Zi Yan <[email protected]>Cc Wei Xu <[email protected]>Cc osalvador <[email protected]>Cc Shakeel Butt <[email protected]>Cc Hasan Al Maruf <[email protected]>
Quoted 1 line
---
Tested on my tiered memory system, and works well. Please feel free to add:
Quoted 86 lines
include/linux/mmzone.h | 3 +++ include/linux/node.h | 5 +++++ mm/migrate.c | 13 ++++++++++--- mm/vmstat.c | 3 +++ 4 files changed, 21 insertions(+), 3 deletions(-)include/linux/mmzone.h +3
-0@@ -210,6 +210,9 @@ enum node_stat_item { NR_PAGETABLE, /* used for pagetables */#ifdef CONFIG_SWAP NR_SWAPCACHE, +#endif +#ifdef CONFIG_NUMA_BALANCING +PGPROMOTE_SUCCESS ,/* promote successfully */ #endifNR_VM_NODE_STAT_ITEMS };include/linux/node.h +5
-0@@ -181,4 +181,9 @@ static inline void register_hugetlbfs_with_node(node_registration_func_t reg,#define to_node (device) container_of(device, struct node, dev) +static inline bool node_is_toptier (int node) +{ +return node_state (node,N_CPU ); +} +#endif /* _LINUX_NODE_H_ */ mm/migrate.c +10
-3@@ -2141,6 +2141,7 @@ int migrate_misplaced_page(struct page *page, struct vm_area_struct *vma,pg_data_t *pgdat =NODE_DATA (node);int isolated;int nr_remaining; +int nr_succeeded;LIST_HEAD (migratepages);new_page_t *new;bool compound;@@ -2179,7 +2180,8 @@ int migrate_misplaced_page(struct page *page, struct vm_area_struct *vma,list_add (&page->lru, &migratepages); nr_remaining =migrate_pages (&migratepages, *new,NULL , node, -MIGRATE_ASYNC ,MR_NUMA_MISPLACED ,NULL ); +MIGRATE_ASYNC ,MR_NUMA_MISPLACED , + &nr_succeeded);if (nr_remaining) {if (!list_empty (&migratepages)) {list_del (&page->lru);@@ -2188,8 +2190,13 @@ int migrate_misplaced_page(struct page *page, struct vm_area_struct *vma,putback_lru_page (page); } isolated =0 ; - }else -count_vm_numa_events (NUMA_PAGE_MIGRATE , nr_pages); + } +if (nr_succeeded) { +count_vm_numa_events (NUMA_PAGE_MIGRATE , nr_succeeded); +if (!node_is_toptier (page_to_nid (page)) &&node_is_toptier (node)) +mod_node_page_state (NODE_DATA (node),PGPROMOTE_SUCCESS , + nr_succeeded); + }BUG_ON (!list_empty (&migratepages));return isolated;mm/vmstat.c +3
-0@@ -1242,6 +1242,9 @@ const char * const vmstat_text[] = {#ifdef CONFIG_SWAP "nr_swapcached",#endif +#ifdef CONFIG_NUMA_BALANCING +"pgpromote_success" , +#endif/* enum writeback_stat_item counters */ "nr_dirty_threshold",
PATCH RESEND 2/6 NUMA balancing: optimize page placement for memory tiering system
In reply to Huang YingWith the advent of various new memory types, some machines will have multiple types of memory, e.g. DRAM and PMEM (persistent memory). The memory subsystem of these machines can be called memory tiering system, because the performance of the different types of memory are usually different.
In such system, because of the memory accessing pattern changing etc, some pages in the slow memory may become hot globally. So in this patch, the NUMA balancing mechanism is enhanced to optimize the page placement among the different memory types according to hot/cold dynamically.
In a typical memory tiering system, there are CPUs, fast memory and slow memory in each physical NUMA node. The CPUs and the fast memory will be put in one logical node (called fast memory node), while the slow memory will be put in another (faked) logical node (called slow memory node). That is, the fast memory is regarded as local while the slow memory is regarded as remote. So it's possible for the recently accessed pages in the slow memory node to be promoted to the fast memory node via the existing NUMA balancing mechanism.
The original NUMA balancing mechanism will stop to migrate pages if the free memory of the target node will become below the high watermark. This is a reasonable policy if there's only one memory type. But this makes the original NUMA balancing mechanism almost not work to optimize page placement among different memory types. Details are as follows.
It's the common cases that the working-set size of the workload is larger than the size of the fast memory nodes. Otherwise, it's unnecessary to use the slow memory at all. So in the common cases, there are almost always no enough free pages in the fast memory nodes, so that the globally hot pages in the slow memory node cannot be promoted to the fast memory node. To solve the issue, we have 2 choices as follows,
a. Ignore the free pages watermark checking when promoting hot pages
from the slow memory node to the fast memory node. This will create some memory pressure in the fast memory node, thus trigger the memory reclaiming. So that, the cold pages in the fast memory node will be demoted to the slow memory node.
b. Make kswapd of the fast memory node to reclaim pages until the free
pages are a little more (about 10MB) than the high watermark. Then, if the free pages of the fast memory node reaches high watermark, and some hot pages need to be promoted, kswapd of the fast memory node will be waken up to demote some cold pages in the fast memory node to the slow memory node. This will free some extra space in the fast memory node, so the hot pages in the slow memory node can be promoted to the fast memory node.
The choice "a" will create the memory pressure in the fast memory node. If the memory pressure of the workload is high, the memory pressure may become so high that the memory allocation latency of the workload is influenced, e.g. the direct reclaiming may be triggered.
The choice "b" works much better at this aspect. If the memory pressure of the workload is high, the hot pages promotion will stop earlier because its allocation watermark is higher than that of the normal memory allocation. So in this patch, choice "b" is implemented.
In addition to the original page placement optimization among sockets, the NUMA balancing mechanism is extended to be used to optimize page placement according to hot/cold among different memory types. So the sysctl user space interface (numa_balancing) is extended in a backward compatible way as follow, so that the users can enable/disable these functionality individually.
The sysctl is converted from a Boolean value to a bits field. The definition of the flags is,
- 0x0: NUMA_BALANCING_DISABLED
- 0x1: NUMA_BALANCING_NORMAL
- 0x2: NUMA_BALANCING_MEMORY_TIERING
--- Documentation/admin-guide/sysctl/kernel.rst | 29 ++++++++++++++------- include/linux/sched/sysctl.h | 10 +++++++ kernel/sched/core.c | 21 ++++++++++++--- kernel/sysctl.c | 3 ++- mm/migrate.c | 19 ++++++++++++-- mm/vmscan.c | 16 ++++++++++++ 6 files changed, 82 insertions(+), 16 deletions(-)
Documentation/admin-guide/sysctl/kernel.rst +20 -9
@@ -595,16 +595,23 @@ Documentation/admin-guide/kernel-parameters.rst). numa_balancing ============== -Enables/disables automatic page fault based NUMA memory -balancing. Memory is moved automatically to nodes -that access it often. +Enables/disables and configure automatic page fault based NUMA memory +balancing. Memory is moved automatically to nodes that access it +often. The value to set can be the result to OR the following, -Enables/disables automatic NUMA memory balancing. On NUMA machines, there -is a performance penalty if remote memory is accessed by a CPU. When this -feature is enabled the kernel samples what task thread is accessing memory -by periodically unmapping pages and later trapping a page fault. At the -time of the page fault, it is determined if the data being accessed should -be migrated to a local memory node. += ================================= +0x0 NUMA_BALANCING_DISABLED +0x1 NUMA_BALANCING_NORMAL +0x2 NUMA_BALANCING_MEMORY_TIERING += ================================= + +Or NUMA_BALANCING_NORMAL to optimize page placement among different +NUMA nodes to reduce remote accessing. On NUMA machines, there is a +performance penalty if remote memory is accessed by a CPU. When this +feature is enabled the kernel samples what task thread is accessing +memory by periodically unmapping pages and later trapping a page +fault. At the time of the page fault, it is determined if the data +being accessed should be migrated to a local memory node. The unmapping of pages and trapping faults incur additional overhead that ideally is offset by improved memory locality but there is no universal
@@ -615,6 +622,10 @@ faults may be controlled by the `numa_balancing_scan_period_min_ms, numa_balancing_scan_delay_ms, numa_balancing_scan_period_max_ms, numa_balancing_scan_size_mb`_, and numa_balancing_settle_count sysctls. +Or NUMA_BALANCING_MEMORY_TIERING to optimize page placement among +different types of memory (represented as different NUMA nodes) to +place the hot pages in the fast memory. This is implemented based on +unmapping and page fault too. numa_balancing_scan_period_min_ms, numa_balancing_scan_delay_ms, numa_balancing_scan_period_max_ms, numa_balancing_scan_size_mb ===============================================================================================================================
include/linux/sched/sysctl.h +10 -0
@@ -35,6 +35,16 @@ enum sched_tunable_scaling { SCHED_TUNABLESCALING_END, }; +#define NUMA_BALANCING_DISABLED 0x0 +#define NUMA_BALANCING_NORMAL 0x1 +#define NUMA_BALANCING_MEMORY_TIERING 0x2 + +#ifdef CONFIG_NUMA_BALANCING +extern int sysctl_numa_balancing_mode; +#else +#define sysctl_numa_balancing_mode 0 +#endif +/* * control realtime throttling: *
kernel/sched/core.c +17 -4
@@ -4265,7 +4265,9 @@ DEFINE_STATIC_KEY_FALSE(sched_numa_balancing);#ifdef CONFIG_NUMA_BALANCING -void set_numabalancing_state (bool enabled) +int sysctl_numa_balancing_mode; + +static void __set_numabalancing_state (bool enabled) {if (enabled)static_branch_enable (&sched_numa_balancing);
@@ -4273,13 +4275,22 @@ void set_numabalancing_state(bool enabled)static_branch_disable (&sched_numa_balancing); } +void set_numabalancing_state (bool enabled) +{ +if (enabled) + sysctl_numa_balancing_mode =NUMA_BALANCING_NORMAL ; +else + sysctl_numa_balancing_mode =NUMA_BALANCING_DISABLED ; +__set_numabalancing_state (enabled); +} +#ifdef CONFIG_PROC_SYSCTL int sysctl_numa_balancing (struct ctl_table *table,int write,void *buffer,size_t *lenp,loff_t *ppos) {struct ctl_table t;int err; -int state =static_branch_likely (&sched_numa_balancing); +int state = sysctl_numa_balancing_mode;if (write && !capable (CAP_SYS_ADMIN ))return -EPERM ;
@@ -4289,8 +4300,10 @@ int sysctl_numa_balancing(struct ctl_table *table, int write, err =proc_dointvec_minmax (&t, write, buffer, lenp, ppos);if (err <0 )return err; -if (write) -set_numabalancing_state (state); +if (write) { + sysctl_numa_balancing_mode = state; +__set_numabalancing_state (state); + } return err; }#endif
kernel/sysctl.c +2 -1
@@ -115,6 +115,7 @@ static int sixty = 60;static int __maybe_unused neg_one =-1 ;static int __maybe_unused two =2 ; +static int __maybe_unused three =3 ;static int __maybe_unused four =4 ;static unsigned long zero_ul;static unsigned long one_ul =1 ;
@@ -1808,7 +1809,7 @@ static struct ctl_table kern_table[] = { .mode = 0644, .proc_handler = sysctl_numa_balancing, .extra1 = SYSCTL_ZERO, - .extra2 = SYSCTL_ONE, + .extra2 = &three, },#endif /* CONFIG_NUMA_BALANCING */ {
mm/migrate.c +17 -2
@@ -50,6 +50,7 @@#include <linux/ptrace.h> #include <linux/oom.h> #include <linux/memory.h> +#include <linux/sched/sysctl.h> #include <asm/tlbflush.h>
@@ -2103,16 +2104,30 @@ static int numamigrate_isolate_page(pg_data_t *pgdat, struct page *page) {int page_lru;int nr_pages =thp_nr_pages (page); +int order =compound_order (page); -VM_BUG_ON_PAGE (compound_order (page) && !PageTransHuge (page), page); +VM_BUG_ON_PAGE (order && !PageTransHuge (page), page);/* Do not migrate THP mapped by multiple processes */ if (PageTransHuge (page) &&total_mapcount (page) >1 )return 0 ;/* Avoid migrating to a node that is nearly full */ -if (!migrate_balanced_pgdat (pgdat, nr_pages)) +if (!migrate_balanced_pgdat (pgdat, nr_pages)) { +int z; + +if (!(sysctl_numa_balancing_mode &NUMA_BALANCING_MEMORY_TIERING ) || + !numa_demotion_enabled) +return 0 ; +if (next_demotion_node (pgdat->node_id) ==NUMA_NO_NODE ) +return 0 ; +for (z = pgdat->nr_zones -1 ; z >=0 ; z--) { +if (populated_zone (pgdat->node_zones + z)) +break ; + } +wakeup_kswapd (pgdat->node_zones + z,0 , order,ZONE_MOVABLE );return 0 ; + }if (isolate_lru_page (page))return 0 ;
mm/vmscan.c +16 -0
@@ -56,6 +56,7 @@#include <linux/swapops.h> #include <linux/balloon_compaction.h> +#include <linux/sched/sysctl.h> #include "internal.h"
@@ -3919,6 +3920,12 @@ static bool pgdat_watermark_boosted(pg_data_t *pgdat, int highest_zoneidx)return false; } +/* +* Keep the free pages on fast memory node a little more than the high +* watermark to accommodate the promoted pages. +*/ +#define NUMA_BALANCING_PROMOTE_WATERMARK (10UL * 1024 * 1024 >> PAGE_SHIFT) +/* * Returns true if there is an eligible zone balanced for the request order * and highest_zoneidx
@@ -3940,6 +3947,15 @@ static bool pgdat_balanced(pg_data_t *pgdat, int order, int highest_zoneidx)continue ; mark =high_wmark_pages (zone); +if (sysctl_numa_balancing_mode &NUMA_BALANCING_MEMORY_TIERING && + numa_demotion_enabled && +next_demotion_node (pgdat->node_id) !=NUMA_NO_NODE ) { +unsigned long promote_mark; + + promote_mark =min (NUMA_BALANCING_PROMOTE_WATERMARK , + pgdat->node_present_pages >>6 ); + mark += promote_mark; + }if (zone_watermark_ok_safe (zone, order, mark, highest_zoneidx))return true; }
Signature · 1 line
2.30.2
Re: [PATCH -V10 RESEND 2/6] NUMA balancing: optimize page placement for memory tiering system
Hi Huang,
kernel/sched/core.c void set_numabalancing_state(bool enabled) · line 4286
Quoted 9 lines
kernel/sched/core.c +9
-0+void set_numabalancing_state (bool enabled) +{ +if (enabled) + sysctl_numa_balancing_mode =NUMA_BALANCING_NORMAL ; +else + sysctl_numa_balancing_mode =NUMA_BALANCING_DISABLED ; +__set_numabalancing_state (enabled); +} +
One of the properties of optimized NUMA Balancing for tiered memory is we are not going to scan top-tier nodes as promotion doesn't make sense there (implemented in the next patch [3/6]). However, if a system has only single memory node with CPU, does it make sense to run NUMA_BALANCING_NORMAL mode there? What do you think about downgrading to NUMA_BALANCING_MEMORY_TIERING mode if a user setup NUMA Balancing on the default mode of NUMA_BALANCING_NORMAL on a single toptier memory node?
Quoted 31 lines
mm/vmscan.c +9
-0@@ -56,6 +56,7 @@#include <linux/swapops.h> #include <linux/balloon_compaction.h> +#include <linux/sched/sysctl.h> #include "internal.h" @@ -3919,6 +3920,12 @@ static bool pgdat_watermark_boosted(pg_data_t *pgdat, int highest_zoneidx)return false; } +/* +* Keep the free pages on fast memory node a little more than the high +* watermark to accommodate the promoted pages. +*/ +#define NUMA_BALANCING_PROMOTE_WATERMARK (10UL * 1024 * 1024 >> PAGE_SHIFT) +/* * Returns true if there is an eligible zone balanced for the request order * and highest_zoneidx @@ -3940,6 +3947,15 @@ static bool pgdat_balanced(pg_data_t *pgdat, int order, int highest_zoneidx)continue ; mark =high_wmark_pages (zone); + if (sysctl_numa_balancing_mode & NUMA_BALANCING_MEMORY_TIERING && + numa_demotion_enabled &&
mm/vmscan.c static bool pgdat_balanced(pg_data_t *pgdat, int order, int highest_zoneidx) · line 3961
Quoted 10 lines
mm/vmscan.c +7
-0+ next_demotion_node(pgdat->node_id) != NUMA_NO_NODE) { +unsigned long promote_mark; + + promote_mark =min (NUMA_BALANCING_PROMOTE_WATERMARK , + pgdat->node_present_pages >>6 ); + mark += promote_mark; + }if (zone_watermark_ok_safe (zone, order, mark, highest_zoneidx))return true; }
This can be moved to a different patch. I think, this patch [2/6] can be splitted into two basic patches -- 1. NUMA Balancing interface for tiered memory and 2. maintaining a headroom for promotion.
Instead of having a static value for NUMA_BALANCING_PROMOTE_WATERMARK what about decoupling the allocation and reclamation and add a user-space interface for controling them?
Do you think patch [2/5] and [3/5] of this series can be merged to your current patchset?
https://lore.kernel.org/all/[email protected]/
Best,
Hasan
Re: [PATCH -V10 RESEND 2/6] NUMA balancing: optimize page placement for memory tiering system
Quoted 12 lines · Hasan Al Maruf
Hi Huang,
kernel/sched/core.c void set_numabalancing_state(bool enabled) · line 4286
Quoted 9 lines
kernel/sched/core.c +9
-0+void set_numabalancing_state (bool enabled) +{ +if (enabled) + sysctl_numa_balancing_mode =NUMA_BALANCING_NORMAL ; +else + sysctl_numa_balancing_mode =NUMA_BALANCING_DISABLED ; +__set_numabalancing_state (enabled); +} +
Quoted 8 lines
One of the properties of optimized NUMA Balancing for tiered memory is we are not going to scan top-tier nodes as promotion doesn't make sense there (implemented in the next patch [3/6]). However, if a system has only single memory node with CPU, does it make sense to run
NUMA_BALANCING_NORMALmode there? What do you think about downgrading toNUMA_BALANCING_MEMORY_TIERINGmode if a user setup NUMA Balancing on the default mode ofNUMA_BALANCING_NORMALon a single toptier memory node?
Consider a system with only 1 NUMA node and no PMEM, should we refuse NUMA balancing to be enabled at all?
Per my understanding, the philosophy behind is to keep thing as small as possible instead of as smart as possible. Do you agree?
Quoted 37 lines
Quoted 37 lines
mm/vmscan.c +15
-0@@ -56,6 +56,7 @@#include <linux/swapops.h> #include <linux/balloon_compaction.h> +#include <linux/sched/sysctl.h> #include "internal.h" @@ -3919,6 +3920,12 @@ static bool pgdat_watermark_boosted(pg_data_t *pgdat, int highest_zoneidx)return false; } +/* +* Keep the free pages on fast memory node a little more than the high +* watermark to accommodate the promoted pages. +*/ +#define NUMA_BALANCING_PROMOTE_WATERMARK (10UL * 1024 * 1024 >> PAGE_SHIFT) +/* * Returns true if there is an eligible zone balanced for the request order * and highest_zoneidx @@ -3940,6 +3947,15 @@ static bool pgdat_balanced(pg_data_t *pgdat, int order, int highest_zoneidx)continue ; mark =high_wmark_pages (zone); +if (sysctl_numa_balancing_mode &NUMA_BALANCING_MEMORY_TIERING && + numa_demotion_enabled && +next_demotion_node (pgdat->node_id) !=NUMA_NO_NODE ) { +unsigned long promote_mark; + + promote_mark =min (NUMA_BALANCING_PROMOTE_WATERMARK , + pgdat->node_present_pages >>6 ); + mark += promote_mark;
Quoted 8 lines
mm/vmscan.c static bool pgdat_balanced(pg_data_t *pgdat, int order, int highest_zoneidx) · line 3961
Quoted 4 lines
mm/vmscan.c +1
-0+ }if (zone_watermark_ok_safe (zone, order, mark, highest_zoneidx))return true; }This can be moved to a different patch. I think, this patch [2/6] can be splitted into two basic patches -- 1. NUMA Balancing interface for tiered memory and 2. maintaining a headroom for promotion.
Johannes has taught me that, if we introduce a new function, variable, or interface, it's better to introduce its user together. So that we can determine whether it's necessary to do that, whether the definition is suitable, etc. I think that makes sense. So I try to do that in this patchset too.
As in [2/5] of your patchset below, another possibility is to make
1. NUMA balancing interface for tiered memory and 2. skip scanning top tier memory in NUMA balancing in one patch. One concern is that although this is an optimization, there's almost no measurable performance difference. This makes it hard to justify to extend the user space interface. Do you have better data to support this?
Quoted 3 lines
Instead of having a static value for
NUMA_BALANCING_PROMOTE_WATERMARKwhat about decoupling the allocation and reclamation and add a user-space interface for controling them?
This means to add a new user space ABI. Because we may need to support the new ABI forever, we should have strong justification to add it. I am not against to add an ABI to adjust promotion watermark in general. I think that the path could be,
- Have a simplest solution that works without introducing new ABI, like something in this patch, or revised.
- Then try to add a new ABI in a separate patch with enough justification, for example, with much improved performance data.
Do you agree?
Quoted 4 lines
Do you think patch [2/5] and [3/5] of this series can be merged to your current patchset?
Best Regards,
Huang, Ying
Re: [PATCH -V10 RESEND 2/6] NUMA balancing: optimize page placement for memory tiering system
In reply to Huang YingQuoted 299 lines · Huang Ying
With the advent of various new memory types, some machines will have multiple types of memory, e.g. DRAM and PMEM (persistent memory). The memory subsystem of these machines can be called memory tiering system, because the performance of the different types of memory are usually different.
In such system, because of the memory accessing pattern changing etc, some pages in the slow memory may become hot globally. So in this patch, the NUMA balancing mechanism is enhanced to optimize the page placement among the different memory types according to hot/cold dynamically.
In a typical memory tiering system, there are CPUs, fast memory and slow memory in each physical NUMA node. The CPUs and the fast memory will be put in one logical node (called fast memory node), while the slow memory will be put in another (faked) logical node (called slow memory node). That is, the fast memory is regarded as local while the slow memory is regarded as remote. So it's possible for the recently accessed pages in the slow memory node to be promoted to the fast memory node via the existing NUMA balancing mechanism.
The original NUMA balancing mechanism will stop to migrate pages if the free memory of the target node will become below the high watermark. This is a reasonable policy if there's only one memory type. But this makes the original NUMA balancing mechanism almost not work to optimize page placement among different memory types. Details are as follows.
It's the common cases that the working-set size of the workload is larger than the size of the fast memory nodes. Otherwise, it's unnecessary to use the slow memory at all. So in the common cases, there are almost always no enough free pages in the fast memory nodes, so that the globally hot pages in the slow memory node cannot be promoted to the fast memory node. To solve the issue, we have 2 choices as follows,
a. Ignore the free pages watermark checking when promoting hot pages
from the slow memory node to the fast memory node. This will create some memory pressure in the fast memory node, thus trigger the memory reclaiming. So that, the cold pages in the fast memory node will be demoted to the slow memory node.b. Make kswapd of the fast memory node to reclaim pages until the free
pages are a little more (about 10MB) than the high watermark. Then, if the free pages of the fast memory node reaches high watermark, and some hot pages need to be promoted, kswapd of the fast memory node will be waken up to demote some cold pages in the fast memory node to the slow memory node. This will free some extra space in the fast memory node, so the hot pages in the slow memory node can be promoted to the fast memory node.The choice "a" will create the memory pressure in the fast memory node. If the memory pressure of the workload is high, the memory pressure may become so high that the memory allocation latency of the workload is influenced, e.g. the direct reclaiming may be triggered.
The choice "b" works much better at this aspect. If the memory pressure of the workload is high, the hot pages promotion will stop earlier because its allocation watermark is higher than that of the normal memory allocation. So in this patch, choice "b" is implemented.
In addition to the original page placement optimization among sockets, the NUMA balancing mechanism is extended to be used to optimize page placement according to hot/cold among different memory types. So the sysctl user space interface (
numa_balancing) is extended in a backward compatible way as follow, so that the users can enable/disable these functionality individually.The sysctl is converted from a Boolean value to a bits field. The definition of the flags is,
-
0x0:NUMA_BALANCING_DISABLED
-0x1:NUMA_BALANCING_NORMAL
-0x2:NUMA_BALANCING_MEMORY_TIERINGSigned-off-by "Huang, Ying" <[email protected]>Cc Andrew Morton <[email protected]>Cc Michal Hocko <[email protected]>Cc Rik van Riel <[email protected]>Cc Mel Gorman <[email protected]>Cc Peter Zijlstra <[email protected]>Cc Dave Hansen <[email protected]>Cc Yang Shi <[email protected]>Cc Zi Yan <[email protected]>Cc Wei Xu <[email protected]>Cc osalvador <[email protected]>Cc Shakeel Butt <[email protected]>Cc Hasan Al Maruf <[email protected]>--- Documentation/admin-guide/sysctl/kernel.rst | 29 ++++++++++++++------- include/linux/sched/sysctl.h | 10 +++++++ kernel/sched/core.c | 21 ++++++++++++--- kernel/sysctl.c | 3 ++- mm/migrate.c | 19 ++++++++++++-- mm/vmscan.c | 16 ++++++++++++ 6 files changed, 82 insertions(+), 16 deletions(-)Documentation/admin-guide/sysctl/kernel.rst +20
-9@@ -595,16 +595,23 @@ Documentation/admin-guide/kernel-parameters.rst). numa_balancing ============== -Enables/disables automatic page fault based NUMA memory -balancing. Memory is moved automatically to nodes -that access it often. +Enables/disables and configure automatic page fault based NUMA memory +balancing. Memory is moved automatically to nodes that access it +often. The value to set can be the result to OR the following, -Enables/disables automatic NUMA memory balancing. On NUMA machines, there -is a performance penalty if remote memory is accessed by a CPU. When this -feature is enabled the kernel samples what task thread is accessing memory -by periodically unmapping pages and later trapping a page fault. At the -time of the page fault, it is determined if the data being accessed should -be migrated to a local memory node. += ================================= +0x0 NUMA_BALANCING_DISABLED +0x1 NUMA_BALANCING_NORMAL +0x2 NUMA_BALANCING_MEMORY_TIERING += ================================= + +Or NUMA_BALANCING_NORMAL to optimize page placement among different +NUMA nodes to reduce remote accessing. On NUMA machines, there is a +performance penalty if remote memory is accessed by a CPU. When this +feature is enabled the kernel samples what task thread is accessing +memory by periodically unmapping pages and later trapping a page +fault. At the time of the page fault, it is determined if the data +being accessed should be migrated to a local memory node. The unmapping of pages and trapping faults incur additional overhead that ideally is offset by improved memory locality but there is no universal@@ -615,6 +622,10 @@ faults may be controlled by the `numa_balancing_scan_period_min_ms, numa_balancing_scan_delay_ms, numa_balancing_scan_period_max_ms, numa_balancing_scan_size_mb`_, and numa_balancing_settle_count sysctls. +Or NUMA_BALANCING_MEMORY_TIERING to optimize page placement among +different types of memory (represented as different NUMA nodes) to +place the hot pages in the fast memory. This is implemented based on +unmapping and page fault too. numa_balancing_scan_period_min_ms, numa_balancing_scan_delay_ms, numa_balancing_scan_period_max_ms, numa_balancing_scan_size_mb ===============================================================================================================================include/linux/sched/sysctl.h +10
-0@@ -35,6 +35,16 @@ enum sched_tunable_scaling { SCHED_TUNABLESCALING_END, }; +#define NUMA_BALANCING_DISABLED 0x0 +#define NUMA_BALANCING_NORMAL 0x1 +#define NUMA_BALANCING_MEMORY_TIERING 0x2 + +#ifdef CONFIG_NUMA_BALANCING +extern int sysctl_numa_balancing_mode; +#else +#define sysctl_numa_balancing_mode 0 +#endif +/* * control realtime throttling: * kernel/sched/core.c +17
-4@@ -4265,7 +4265,9 @@ DEFINE_STATIC_KEY_FALSE(sched_numa_balancing);#ifdef CONFIG_NUMA_BALANCING -void set_numabalancing_state (bool enabled) +int sysctl_numa_balancing_mode; + +static void __set_numabalancing_state (bool enabled) {if (enabled)static_branch_enable (&sched_numa_balancing);@@ -4273,13 +4275,22 @@ void set_numabalancing_state(bool enabled)static_branch_disable (&sched_numa_balancing); } +void set_numabalancing_state (bool enabled) +{ +if (enabled) + sysctl_numa_balancing_mode =NUMA_BALANCING_NORMAL ; +else + sysctl_numa_balancing_mode =NUMA_BALANCING_DISABLED ; +__set_numabalancing_state (enabled); +} +#ifdef CONFIG_PROC_SYSCTL int sysctl_numa_balancing (struct ctl_table *table,int write,void *buffer,size_t *lenp,loff_t *ppos) {struct ctl_table t;int err; -int state =static_branch_likely (&sched_numa_balancing); +int state = sysctl_numa_balancing_mode;if (write && !capable (CAP_SYS_ADMIN ))return -EPERM ;@@ -4289,8 +4300,10 @@ int sysctl_numa_balancing(struct ctl_table *table, int write, err =proc_dointvec_minmax (&t, write, buffer, lenp, ppos);if (err <0 )return err; -if (write) -set_numabalancing_state (state); +if (write) { + sysctl_numa_balancing_mode = state; +__set_numabalancing_state (state); + } return err; }#endif kernel/sysctl.c +2
-1@@ -115,6 +115,7 @@ static int sixty = 60;static int __maybe_unused neg_one =-1 ;static int __maybe_unused two =2 ; +static int __maybe_unused three =3 ;static int __maybe_unused four =4 ;static unsigned long zero_ul;static unsigned long one_ul =1 ;@@ -1808,7 +1809,7 @@ static struct ctl_table kern_table[] = { .mode = 0644, .proc_handler = sysctl_numa_balancing, .extra1 = SYSCTL_ZERO, - .extra2 = SYSCTL_ONE, + .extra2 = &three, },#endif /* CONFIG_NUMA_BALANCING */ {mm/migrate.c +17
-2@@ -50,6 +50,7 @@#include <linux/ptrace.h> #include <linux/oom.h> #include <linux/memory.h> +#include <linux/sched/sysctl.h> #include <asm/tlbflush.h> @@ -2103,16 +2104,30 @@ static int numamigrate_isolate_page(pg_data_t *pgdat, struct page *page) {int page_lru;int nr_pages =thp_nr_pages (page); +int order =compound_order (page); -VM_BUG_ON_PAGE (compound_order (page) && !PageTransHuge (page), page); +VM_BUG_ON_PAGE (order && !PageTransHuge (page), page);/* Do not migrate THP mapped by multiple processes */ if (PageTransHuge (page) &&total_mapcount (page) >1 )return 0 ;/* Avoid migrating to a node that is nearly full */ -if (!migrate_balanced_pgdat (pgdat, nr_pages)) +if (!migrate_balanced_pgdat (pgdat, nr_pages)) { +int z; + +if (!(sysctl_numa_balancing_mode &NUMA_BALANCING_MEMORY_TIERING ) || + !numa_demotion_enabled) +return 0 ; +if (next_demotion_node (pgdat->node_id) ==NUMA_NO_NODE ) +return 0 ; +for (z = pgdat->nr_zones -1 ; z >=0 ; z--) { +if (populated_zone (pgdat->node_zones + z)) +break ; + } +wakeup_kswapd (pgdat->node_zones + z,0 , order,ZONE_MOVABLE );return 0 ; + }if (isolate_lru_page (page))return 0 ;mm/vmscan.c +1
-0@@ -56,6 +56,7 @@#include <linux/swapops.h> #include <linux/balloon_compaction.h> +#include <linux/sched/sysctl.h> #include "internal.h"
mm/vmscan.c static bool pgdat_watermark_boosted(pg_data_t *pgdat, int highest_zoneidx) · line 3927
Quoted 8 lines
mm/vmscan.c +5
-0@@ -3919,6 +3920,12 @@ static bool pgdat_watermark_boosted(pg_data_t *pgdat, int highest_zoneidx)return false; } +/* +* Keep the free pages on fast memory node a little more than the high +* watermark to accommodate the promoted pages. +*/ +#define NUMA_BALANCING_PROMOTE_WATERMARK (10UL * 1024 * 1024 >> PAGE_SHIFT)
From our testing the fixed promote wartermark is not suitable for all scenarios, but as you said, I also agree with that we can start from a simplest solution that works. So please feel free to add:
Quoted 20 lines
mm/vmscan.c +10
-0+/* * Returns true if there is an eligible zone balanced for the request order * and highest_zoneidx @@ -3940,6 +3947,15 @@ static bool pgdat_balanced(pg_data_t *pgdat, int order, int highest_zoneidx)continue ; mark =high_wmark_pages (zone); +if (sysctl_numa_balancing_mode &NUMA_BALANCING_MEMORY_TIERING && + numa_demotion_enabled && +next_demotion_node (pgdat->node_id) !=NUMA_NO_NODE ) { +unsigned long promote_mark; + + promote_mark =min (NUMA_BALANCING_PROMOTE_WATERMARK , + pgdat->node_present_pages >>6 ); + mark += promote_mark; + }if (zone_watermark_ok_safe (zone, order, mark, highest_zoneidx))return true; }
PATCH RESEND 3/6 memory tiering: skip to scan fast memory
In reply to Huang YingIf the NUMA balancing isn't used to optimize the page placement among sockets but only among memory types, the hot pages in the fast memory node couldn't be migrated (promoted) to anywhere. So it's unnecessary to scan the pages in the fast memory node via changing their PTE/PMD mapping to be PROT_NONE. So that the page faults could be avoided too.
In the test, if only the memory tiering NUMA balancing mode is enabled, the number of the NUMA balancing hint faults for the DRAM node is reduced to almost 0 with the patch. While the benchmark score doesn't change visibly.
--- mm/huge_memory.c | 30 +++++++++++++++++++++--------- mm/mprotect.c | 13 ++++++++++++- 2 files changed, 33 insertions(+), 10 deletions(-)
mm/huge_memory.c +21 -9
@@ -34,6 +34,7 @@#include <linux/oom.h> #include <linux/numa.h> #include <linux/page_owner.h> +#include <linux/sched/sysctl.h> #include <asm/tlb.h> #include <asm/pgalloc.h>
@@ -1766,17 +1767,28 @@ int change_huge_pmd(struct vm_area_struct *vma, pmd_t *pmd, }#endif -/* -* Avoid trapping faults against the zero page. The read-only -* data is likely to be read-cached on the local CPU and -* local/remote hits to the zero page are not interesting. -*/ -if (prot_numa &&is_huge_zero_pmd (*pmd)) - goto unlock; +if (prot_numa) { +struct page *page; +/* +* Avoid trapping faults against the zero page. The read-only +* data is likely to be read-cached on the local CPU and +* local/remote hits to the zero page are not interesting. +*/ +if (is_huge_zero_pmd (*pmd)) + goto unlock; -if (prot_numa &&pmd_protnone (*pmd)) - goto unlock; +if (pmd_protnone (*pmd)) + goto unlock; + page =pmd_page (*pmd); +/* +* Skip scanning top tier node if normal numa +* balancing is disabled +*/ +if (!(sysctl_numa_balancing_mode &NUMA_BALANCING_NORMAL ) && +node_is_toptier (page_to_nid (page))) + goto unlock; + }/* * In case prot_numa, we are under mmap_read_lock(mm). It's critical * to not clear pmd intermittently to avoid race with MADV_DONTNEED
mm/mprotect.c +12 -1
@@ -29,6 +29,7 @@#include <linux/uaccess.h> #include <linux/mm_inline.h> #include <linux/pgtable.h> +#include <linux/sched/sysctl.h> #include <asm/cacheflush.h> #include <asm/mmu_context.h> #include <asm/tlbflush.h>
@@ -83,6 +84,7 @@ static unsigned long change_pte_range(struct vm_area_struct *vma, pmd_t *pmd,*/ if (prot_numa) {struct page *page; +int nid;/* Avoid TLB flush if possible */ if (pte_protnone (oldpte))
@@ -109,7 +111,16 @@ static unsigned long change_pte_range(struct vm_area_struct *vma, pmd_t *pmd,* Don't mess with PTEs if page is already on the node * a single-threaded process is running on. */ -if (target_node ==page_to_nid (page)) + nid =page_to_nid (page); +if (target_node == nid) +continue ; + +/* +* Skip scanning top tier node if normal numa +* balancing is disabled +*/ +if (!(sysctl_numa_balancing_mode &NUMA_BALANCING_NORMAL ) && +node_is_toptier (nid))continue ; }
Signature · 1 line
2.30.2
Re: [PATCH -V10 RESEND 3/6] memory tiering: skip to scan fast memory
Quoted 27 lines · Huang Ying
If the NUMA balancing isn't used to optimize the page placement among sockets but only among memory types, the hot pages in the fast memory node couldn't be migrated (promoted) to anywhere. So it's unnecessary to scan the pages in the fast memory node via changing their PTE/PMD mapping to be
PROT_NONE. So that the page faults could be avoided too.In the test, if only the memory tiering NUMA balancing mode is enabled, the number of the NUMA balancing hint faults for the DRAM node is reduced to almost 0 with the patch. While the benchmark score doesn't change visibly.
Signed-off-by "Huang, Ying" <[email protected]>Suggested-by Dave Hansen <[email protected]>Cc Andrew Morton <[email protected]>Cc Michal Hocko <[email protected]>Cc Rik van Riel <[email protected]>Cc Mel Gorman <[email protected]>Cc Peter Zijlstra <[email protected]>Cc Yang Shi <[email protected]>Cc Zi Yan <[email protected]>Cc Wei Xu <[email protected]>Cc osalvador <[email protected]>Cc Shakeel Butt <[email protected]>Cc Hasan Al Maruf <[email protected]>
Quoted 1 line
---
LGTM. Please feel free to add:
PATCH RESEND 4/6 memory tiering: hot page selection with hint page fault latency
In reply to Huang YingTo optimize page placement in a memory tiering system with NUMA balancing, the hot pages in the slow memory node need to be identified. Essentially, the original NUMA balancing implementation selects the mostly recently accessed (MRU) pages as the hot pages. But this isn't a very good algorithm to identify the hot pages.
So, in this patch we implemented a better hot page selection algorithm. Which is based on NUMA balancing page table scanning and hint page fault as follows,
- When the page tables of the processes are scanned to change PTE/PMD to be PROT_NONE, the current time is recorded in struct page as scan time.
- When the page is accessed, hint page fault will occur. The scan time is gotten from the struct page. And The hint page fault latency is defined as
hint page fault time - scan time
The shorter the hint page fault latency of a page is, the higher the probability of their access frequency to be higher. So the hint page fault latency is a good estimation of the page hot/cold.
But it's hard to find some extra space in struct page to hold the scan time. Fortunately, we can reuse some bits used by the original NUMA balancing.
NUMA balancing uses some bits in struct page to store the page accessing CPU and PID (referring to page_cpupid_xchg_last()). Which is used by the multi-stage node selection algorithm to avoid to migrate pages shared accessed by the NUMA nodes back and forth. But for pages in the slow memory node, even if they are shared accessed by multiple NUMA nodes, as long as the pages are hot, they need to be promoted to the fast memory node. So the accessing CPU and PID information are unnecessary for the slow memory pages. We can reuse these bits in struct page to record the scan time for them. For the fast memory pages, these bits are used as before.
The remaining problem is how to determine the hot threshold. It's not easy to be done automatically. So we provide a sysctl knob:
kernel.numa_balancing_hot_threshold_ms. All pages with hint page fault latency < the threshold will be considered hot. The system administrator can determine the hot threshold via various information, such as PMEM bandwidth limit, the average number of the pages pass the hot threshold, etc. The default hot threshold is 1 second, which works well in our performance test.
The downside of the patch is that the response time to the workload hot spot changing may be much longer. For example,
- A previous cold memory area becomes hot
- The hint page fault will be triggered. But the hint page fault latency isn't shorter than the hot threshold. So the pages will not be promoted.
- When the memory area is scanned again, maybe after a scan period, the hint page fault latency measured will be shorter than the hot threshold and the pages will be promoted.
To mitigate this,
- If there are enough free space in the fast memory node, the hot threshold will not be used, all pages will be promoted upon the hint page fault for fast response.
- If fast response is more important for system performance, the administrator can set a higher hot threshold.
--- include/linux/mm.h | 29 ++++++++++++++++ include/linux/sched/sysctl.h | 1 + kernel/sched/fair.c | 67 ++++++++++++++++++++++++++++++++++++ kernel/sysctl.c | 7 ++++ mm/huge_memory.c | 13 +++++-- mm/memory.c | 11 +++++- mm/migrate.c | 12 +++++++ mm/mmzone.c | 17 +++++++++ mm/mprotect.c | 8 ++++- 9 files changed, 160 insertions(+), 5 deletions(-)
include/linux/mm.h +29 -0
@@ -1393,6 +1393,18 @@ static inline int folio_nid(const struct folio *folio) }#ifdef CONFIG_NUMA_BALANCING +/* page access time bits needs to hold at least 4 seconds */ +#define PAGE_ACCESS_TIME_MIN_BITS 12 +#if LAST_CPUPID_SHIFT <PAGE_ACCESS_TIME_MIN_BITS +#define PAGE_ACCESS_TIME_BUCKETS \ + (PAGE_ACCESS_TIME_MIN_BITS - LAST_CPUPID_SHIFT) +#else +#define PAGE_ACCESS_TIME_BUCKETS 0 +#endif + +#define PAGE_ACCESS_TIME_MASK \ + (LAST_CPUPID_MASK << PAGE_ACCESS_TIME_BUCKETS) +static inline int cpu_pid_to_cpupid (int cpu,int pid) {return ((cpu &LAST__CPU_MASK ) <<LAST__PID_SHIFT ) | (pid &LAST__PID_MASK );
@@ -1435,6 +1447,16 @@ static inline int page_cpupid_xchg_last(struct page *page, int cpupid)return xchg (&page->_last_cpupid, cpupid &LAST_CPUPID_MASK ); } +static inline unsigned int xchg_page_access_time (struct page *page, +unsigned int time) +{ +unsigned int last_time; + + last_time =xchg (&page->_last_cpupid, + (time >>PAGE_ACCESS_TIME_BUCKETS ) &LAST_CPUPID_MASK ); +return last_time <<PAGE_ACCESS_TIME_BUCKETS ; +} +static inline int page_cpupid_last (struct page *page) {return page->_last_cpupid;
@@ -1450,6 +1472,7 @@ static inline int page_cpupid_last(struct page *page) }extern int page_cpupid_xchg_last (struct page *page,int cpupid); +extern unsigned int xchg_page_access_time (struct page *page,unsigned int time);static inline void page_cpupid_reset_last (struct page *page) {
@@ -1462,6 +1485,12 @@ static inline int page_cpupid_xchg_last(struct page *page, int cpupid)return page_to_nid (page);/* XXX */ } +static inline unsigned int xchg_page_access_time (struct page *page, +unsigned int time) +{ +return 0 ; +} +static inline int page_cpupid_last (struct page *page) {return page_to_nid (page);/* XXX */
include/linux/sched/sysctl.h +1 -0
@@ -41,6 +41,7 @@ enum sched_tunable_scaling {#ifdef CONFIG_NUMA_BALANCING extern int sysctl_numa_balancing_mode; +extern unsigned int sysctl_numa_balancing_hot_threshold;#else #define sysctl_numa_balancing_mode 0#endif
kernel/sched/fair.c +67 -0
97 lines not on this page · raw message
kernel/sysctl.c +7 -0
@@ -1811,6 +1811,13 @@ static struct ctl_table kern_table[] = { .extra1 = SYSCTL_ZERO, .extra2 = &three, }, + { + .procname = "numa_balancing_hot_threshold_ms", + .data = &sysctl_numa_balancing_hot_threshold, + .maxlen = sizeof(unsigned int), + .mode = 0644, + .proc_handler = proc_dointvec, + },#endif /* CONFIG_NUMA_BALANCING */ { .procname = "sched_rt_period_us",
mm/huge_memory.c +10 -3
@@ -1430,7 +1430,7 @@ vm_fault_t do_huge_pmd_numa_page(struct vm_fault *vmf)struct page *page;unsigned long haddr = vmf->address &HPAGE_PMD_MASK ;int page_nid =NUMA_NO_NODE ; -int target_nid, last_cpupid =-1 ; +int target_nid, last_cpupid = (-1 &LAST_CPUPID_MASK );bool migrated = false;bool was_writable =pmd_savedwrite (oldpmd);int flags =0 ;
@@ -1451,7 +1451,8 @@ vm_fault_t do_huge_pmd_numa_page(struct vm_fault *vmf) flags |=TNF_NO_GROUP ; page_nid =page_to_nid (page); - last_cpupid =page_cpupid_last (page); +if (node_is_toptier (page_nid)) + last_cpupid =page_cpupid_last (page); target_nid =numa_migrate_prep (page, vma, haddr, page_nid, &flags);
@@ -1769,6 +1770,7 @@ int change_huge_pmd(struct vm_area_struct *vma, pmd_t *pmd,if (prot_numa) {struct page *page; +bool toptier;/* * Avoid trapping faults against the zero page. The read-only * data is likely to be read-cached on the local CPU and
@@ -1781,13 +1783,18 @@ int change_huge_pmd(struct vm_area_struct *vma, pmd_t *pmd, goto unlock; page =pmd_page (*pmd); + toptier =node_is_toptier (page_to_nid (page));/* * Skip scanning top tier node if normal numa * balancing is disabled */ if (!(sysctl_numa_balancing_mode &NUMA_BALANCING_NORMAL ) && -node_is_toptier (page_to_nid (page))) + toptier) goto unlock; + +if (sysctl_numa_balancing_mode &NUMA_BALANCING_MEMORY_TIERING && + !toptier) +xchg_page_access_time (page,jiffies_to_msecs (jiffies)); }/* * In case prot_numa, we are under mmap_read_lock(mm). It's critical
mm/memory.c +10 -1
@@ -73,6 +73,7 @@#include <linux/perf_event.h> #include <linux/ptrace.h> #include <linux/vmalloc.h> +#include <linux/sched/sysctl.h> #include <trace/events/kmem.h>
@@ -4379,8 +4380,16 @@ static vm_fault_t do_numa_page(struct vm_fault *vmf)if (page_mapcount (page) >1 && (vma->vm_flags &VM_SHARED )) flags |=TNF_SHARED ; - last_cpupid =page_cpupid_last (page); page_nid =page_to_nid (page); +/* +* In memory tiering mode, cpupid of slow memory page is used +* to record page access time. So use default value. +*/ +if ((sysctl_numa_balancing_mode &NUMA_BALANCING_MEMORY_TIERING ) && + !node_is_toptier (page_nid)) + last_cpupid = (-1 &LAST_CPUPID_MASK ); +else + last_cpupid =page_cpupid_last (page); target_nid =numa_migrate_prep (page, vma, vmf->address, page_nid, &flags);if (target_nid ==NUMA_NO_NODE ) {
mm/migrate.c +12 -0
@@ -572,6 +572,18 @@ void folio_migrate_flags(struct folio *newfolio, struct folio *folio)* future migrations of this same page. */ cpupid =page_cpupid_xchg_last (&folio->page,-1 ); +/* +* If migrate between slow and fast memory node, reset cpupid, +* because that is used to record page access time in slow +* memory node +*/ +if (sysctl_numa_balancing_mode &NUMA_BALANCING_MEMORY_TIERING ) { +bool f_toptier =node_is_toptier (page_to_nid (&folio->page)); +bool t_toptier =node_is_toptier (page_to_nid (&newfolio->page)); + +if (f_toptier != t_toptier) + cpupid =-1 ; + }page_cpupid_xchg_last (&newfolio->page, cpupid);folio_migrate_ksm (newfolio, folio);
mm/mmzone.c +17 -0
@@ -99,4 +99,21 @@ int page_cpupid_xchg_last(struct page *page, int cpupid) return last_cpupid; } + +unsigned int xchg_page_access_time (struct page *page,unsigned int time) +{ +unsigned long old_flags, flags; +unsigned int last_time; + + time >>=PAGE_ACCESS_TIME_BUCKETS ; +do { + old_flags = flags = page->flags; + last_time = (flags >>LAST_CPUPID_PGSHIFT ) &LAST_CPUPID_MASK ; + + flags &= ~(LAST_CPUPID_MASK <<LAST_CPUPID_PGSHIFT ); + flags |= (time &LAST_CPUPID_MASK ) <<LAST_CPUPID_PGSHIFT ; + }while (unlikely (cmpxchg (&page->flags, old_flags, flags) != old_flags)); + +return last_time <<PAGE_ACCESS_TIME_BUCKETS ; +}#endif
mm/mprotect.c +7 -1
@@ -85,6 +85,7 @@ static unsigned long change_pte_range(struct vm_area_struct *vma, pmd_t *pmd,if (prot_numa) {struct page *page;int nid; +bool toptier;/* Avoid TLB flush if possible */ if (pte_protnone (oldpte))
@@ -114,14 +115,19 @@ static unsigned long change_pte_range(struct vm_area_struct *vma, pmd_t *pmd, nid =page_to_nid (page);if (target_node == nid)continue ; + toptier =node_is_toptier (nid);/* * Skip scanning top tier node if normal numa * balancing is disabled */ if (!(sysctl_numa_balancing_mode &NUMA_BALANCING_NORMAL ) && -node_is_toptier (nid)) + toptier)continue ; +if (sysctl_numa_balancing_mode &NUMA_BALANCING_MEMORY_TIERING && + !toptier) +xchg_page_access_time (page, +jiffies_to_msecs (jiffies)); } oldpte =ptep_modify_prot_start (vma, addr, pte);
Signature · 1 line
2.30.2
PATCH RESEND 5/6 memory tiering: rate limit NUMA migration throughput
In reply to Huang YingIn NUMA balancing memory tiering mode, the hot slow memory pages could be promoted to the fast memory node via NUMA balancing. But this incurs some overhead too. So that sometimes the workload performance may be hurt. To avoid too much disturbing to the workload in these situations, we should make it possible to rate limit the promotion throughput.
So, in this patch, we implement a simple rate limit algorithm as follows. The number of the candidate pages to be promoted to the fast memory node via NUMA balancing is counted, if the count exceeds the limit specified by the users, the NUMA balancing promotion will be stopped until the next second.
A new sysctl knob kernel.numa_balancing_rate_limit_mbps is added for the users to specify the limit.
TODO: Add ABI document for new sysctl knob.
--- include/linux/mmzone.h | 5 +++++ include/linux/sched/sysctl.h | 1 + kernel/sched/fair.c | 29 +++++++++++++++++++++++++++-- kernel/sysctl.c | 8 ++++++++ mm/vmstat.c | 1 + 5 files changed, 42 insertions(+), 2 deletions(-)
include/linux/mmzone.h +5 -0
@@ -213,6 +213,7 @@ enum node_stat_item {#endif #ifdef CONFIG_NUMA_BALANCING PGPROMOTE_SUCCESS ,/* promote successfully */ +PGPROMOTE_CANDIDATE ,/* candidate pages to promote */ #endifNR_VM_NODE_STAT_ITEMS };
@@ -902,6 +903,10 @@ typedef struct pglist_data {struct deferred_split deferred_split_queue;#endif +#ifdef CONFIG_NUMA_BALANCING +unsigned long numa_ts; +unsigned long numa_nr_candidate; +#endif /* Fields commonly accessed by the page reclaim scanner */ /*
include/linux/sched/sysctl.h +1 -0
@@ -42,6 +42,7 @@ enum sched_tunable_scaling {#ifdef CONFIG_NUMA_BALANCING extern int sysctl_numa_balancing_mode;extern unsigned int sysctl_numa_balancing_hot_threshold; +extern unsigned int sysctl_numa_balancing_rate_limit;#else #define sysctl_numa_balancing_mode 0#endif
kernel/sched/fair.c +27 -2
@@ -1028,6 +1028,11 @@ unsigned int sysctl_numa_balancing_scan_delay = 1000;/* The page with hint page fault latency < threshold in ms is considered hot */ unsigned int sysctl_numa_balancing_hot_threshold =1000 ; +/* +* Restrict the NUMA migration per second in MB for each target node +* if no enough free space in target node +*/ +unsigned int sysctl_numa_balancing_rate_limit =65536 ;struct numa_group {refcount_t refcount;
@@ -1401,6 +1406,23 @@ static int numa_hint_fault_latency(struct page *page)return (time - last_time) &PAGE_ACCESS_TIME_MASK ; } +static bool numa_migration_check_rate_limit (struct pglist_data *pgdat, +unsigned long rate_limit,int nr) +{ +unsigned long nr_candidate; +unsigned long now = jiffies, last_ts; + +mod_node_page_state (pgdat,PGPROMOTE_CANDIDATE , nr); + nr_candidate =node_page_state (pgdat,PGPROMOTE_CANDIDATE ); + last_ts = pgdat->numa_ts; +if (now > last_ts +HZ && +cmpxchg (&pgdat->numa_ts, last_ts, now) == last_ts) + pgdat->numa_nr_candidate = nr_candidate; +if (nr_candidate - pgdat->numa_nr_candidate > rate_limit) +return false; +return true; +} +bool should_numa_migrate_memory (struct task_struct *p,struct page * page,int src_nid,int dst_cpu) {
@@ -1415,7 +1437,7 @@ bool should_numa_migrate_memory(struct task_struct *p, struct page * page,if (sysctl_numa_balancing_mode &NUMA_BALANCING_MEMORY_TIERING && !node_is_toptier (src_nid)) {struct pglist_data *pgdat; -unsigned long latency, th; +unsigned long rate_limit, latency, th; pgdat =NODE_DATA (dst_nid);if (pgdat_free_space_enough (pgdat))
@@ -1426,7 +1448,10 @@ bool should_numa_migrate_memory(struct task_struct *p, struct page * page,if (latency > th)return false; -return true; + rate_limit = + sysctl_numa_balancing_rate_limit << (20 -PAGE_SHIFT ); +return numa_migration_check_rate_limit (pgdat, rate_limit, +thp_nr_pages (page)); } this_cpupid =cpu_pid_to_cpupid (dst_cpu, current->pid);
kernel/sysctl.c +8 -0
@@ -1818,6 +1818,14 @@ static struct ctl_table kern_table[] = { .mode = 0644, .proc_handler = proc_dointvec, }, + { + .procname = "numa_balancing_rate_limit_mbps", + .data = &sysctl_numa_balancing_rate_limit, + .maxlen = sizeof(unsigned int), + .mode = 0644, + .proc_handler = proc_dointvec_minmax, + .extra1 = SYSCTL_ZERO, + },#endif /* CONFIG_NUMA_BALANCING */ { .procname = "sched_rt_period_us",
mm/vmstat.c +1 -0
@@ -1244,6 +1244,7 @@ const char * const vmstat_text[] = {#endif #ifdef CONFIG_NUMA_BALANCING "pgpromote_success", + "pgpromote_candidate",#endif /* enum writeback_stat_item counters */
Signature · 1 line
2.30.2
PATCH RESEND 6/6 memory tiering: adjust hot threshold automatically
In reply to Huang YingIt isn't easy for the administrator to determine the hot threshold. So in this patch, a method to adjust the hot threshold automatically is implemented. The basic idea is to control the number of the candidate promotion pages to match the promotion rate limit. If the hint page fault latency of a page is less than the hot threshold, we will try to promote the page, and the page is called the candidate promotion page.
If the number of the candidate promotion pages in the statistics interval is much more than the promotion rate limit, the hot threshold will be decreased to reduce the number of the candidate promotion pages. Otherwise, the hot threshold will be increased to increase the number of the candidate promotion pages.
To make the above method works, in each statistics interval, the total number of the pages to check (on which the hint page faults occur) and the hot/cold distribution need to be stable. Because the page tables are scanned linearly in NUMA balancing, but the hot/cold distribution isn't uniform along the address, the statistics interval should be larger than the NUMA balancing scan period. So in the patch, the max scan period is used as statistics interval and it works well in our tests.
The sysctl knob kernel.numa_balancing_hot_threshold_ms becomes the initial value and max value of the hot threshold.
--- include/linux/mmzone.h | 3 ++ include/linux/sched/sysctl.h | 2 ++ kernel/sched/core.c | 15 +++++++++ kernel/sched/fair.c | 64 +++++++++++++++++++++++++++++++++--- kernel/sysctl.c | 3 +- 5 files changed, 81 insertions(+), 6 deletions(-)
include/linux/mmzone.h +3 -0
@@ -906,6 +906,9 @@ typedef struct pglist_data {#ifdef CONFIG_NUMA_BALANCING unsigned long numa_ts;unsigned long numa_nr_candidate; +unsigned long numa_threshold_ts; +unsigned long numa_threshold_nr_candidate; +unsigned long numa_threshold;#endif /* Fields commonly accessed by the page reclaim scanner */
include/linux/sched/sysctl.h +2 -0
@@ -84,6 +84,8 @@ int sysctl_sched_uclamp_handler(struct ctl_table *table, int write, void *buffer, size_t *lenp, loff_t *ppos);int sysctl_numa_balancing (struct ctl_table *table,int write,void *buffer,size_t *lenp,loff_t *ppos); +int sysctl_numa_balancing_threshold (struct ctl_table *table,int write,void *buffer, +size_t *lenp,loff_t *ppos);int sysctl_schedstats (struct ctl_table *table,int write,void *buffer,size_t *lenp,loff_t *ppos);
kernel/sched/core.c +15 -0
@@ -4285,6 +4285,18 @@ void set_numabalancing_state(bool enabled) }#ifdef CONFIG_PROC_SYSCTL +static void reset_memory_tiering (void ) +{ +struct pglist_data *pgdat; + +for_each_online_pgdat (pgdat) { + pgdat->numa_threshold =0 ; + pgdat->numa_threshold_nr_candidate = +node_page_state (pgdat,PGPROMOTE_CANDIDATE ); + pgdat->numa_threshold_ts = jiffies; + } +} +int sysctl_numa_balancing (struct ctl_table *table,int write,void *buffer,size_t *lenp,loff_t *ppos) {
@@ -4301,6 +4313,9 @@ int sysctl_numa_balancing(struct ctl_table *table, int write,if (err <0 )return err;if (write) { +if (!(sysctl_numa_balancing_mode &NUMA_BALANCING_MEMORY_TIERING ) && + (state &NUMA_BALANCING_MEMORY_TIERING )) +reset_memory_tiering (); sysctl_numa_balancing_mode = state;__set_numabalancing_state (state); }
kernel/sched/fair.c +59 -5
84 lines not on this page · raw message
kernel/sysctl.c +2 -1
@@ -1816,7 +1816,8 @@ static struct ctl_table kern_table[] = { .data = &sysctl_numa_balancing_hot_threshold, .maxlen = sizeof(unsigned int), .mode = 0644, - .proc_handler = proc_dointvec, + .proc_handler = sysctl_numa_balancing_threshold, + .extra1 = SYSCTL_ZERO, }, { .procname = "numa_balancing_rate_limit_mbps",
Signature · 1 line
2.30.2
Peter Zijlstra
maintainerQuoted 6 lines · Huang Ying
After commit
c221c0b0308f("device-dax: "Hotplug" persistent memory for use like normal RAM"), the PMEM could be used as the cost-effective volatile memory in separate NUMA nodes. In a typical memory tiering system, there are CPUs, DRAM and PMEM in each physical NUMA node. The CPUs and the DRAM will be put in one logical node, while the PMEM will be put in another (faked) logical node.
So what does a system like that actually look like, SLIT table wise, and how does that affect init_numa_topology_type() ?
Hi, Peter,
Quoted 10 lines · Peter Zijlstra
Quoted 6 lines · Huang Ying
After commit
c221c0b0308f("device-dax: "Hotplug" persistent memory for use like normal RAM"), the PMEM could be used as the cost-effective volatile memory in separate NUMA nodes. In a typical memory tiering system, there are CPUs, DRAM and PMEM in each physical NUMA node. The CPUs and the DRAM will be put in one logical node, while the PMEM will be put in another (faked) logical node.So what does a system like that actually look like, SLIT table wise, and how does that affect
init_numa_topology_type()?
The SLIT table is as follows,
[000h 0000 4] Signature : "SLIT" [System Locality Information Table] [004h 0004 4] Table Length : 0000042C [008h 0008 1] Revision : 01 [009h 0009 1] Checksum : 59 [00Ah 0010 6] Oem ID : "INTEL " [010h 0016 8] Oem Table ID : "S2600WF " [018h 0024 4] Oem Revision : 00000001 [01Ch 0028 4] Asl Compiler ID : "INTL" [020h 0032 4] Asl Compiler Revision : 20091013
[024h 0036 8] Localities : 0000000000000004 [02Ch 0044 4] Locality 0 : 0A 15 11 1C [030h 0048 4] Locality 1 : 15 0A 1C 11 [034h 0052 4] Locality 2 : 11 1C 0A 1C [038h 0056 4] Locality 3 : 1C 11 1C 0A
The numactl -H output is as follows,
available: 4 nodes (0-3)
node 0 cpus: 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 node 0 size: 64136 MB
node 0 free: 5981 MB
node 1 cpus: 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 node 1 size: 64466 MB
node 1 free: 10415 MB
node 2 cpus:
node 2 size: 253952 MB
node 2 free: 253920 MB
node 3 cpus:
node 3 size: 253952 MB
node 3 free: 253951 MB
node distances:
node 0 1 2 3
0: 10 21 17 28 1: 21 10 28 17 2: 17 28 10 28 3: 28 17 28 10
init_numa_topology_type() set sched_numa_topology_type to NUMA_DIRECT.
The node 0 and node 1 are onlined during boot. While the PMEM node, that is, node 2 and node 3 are onlined later. As in the following dmesg snippet.
[ 2.252573][ T0] ACPI: SRAT: Node 0 PXM 0 [mem 0x00000000-0x7fffffff] [ 2.259224][ T0] ACPI: SRAT: Node 0 PXM 0 [mem 0x100000000-0x107fffffff] [ 2.266139][ T0] ACPI: SRAT: Node 2 PXM 2 [mem 0x1080000000-0x4f7fffffff] non-volatile [ 2.274267][ T0] ACPI: SRAT: Node 1 PXM 1 [mem 0x4f80000000-0x5f7fffffff] [ 2.281271][ T0] ACPI: SRAT: Node 3 PXM 3 [mem 0x5f80000000-0x9e7fffffff] non-volatile [ 2.289403][ T0] NUMA: Initialized distance table, cnt=4 [ 2.294934][ T0] NUMA: Node 0 [mem 0x00000000-0x7fffffff] + [mem 0x100000000-0x107fffffff] -> [mem 0x00000000-0x107fffffff] [ 2.306266][ T0] NODE_DATA(0) allocated [mem 0x107ffd5000-0x107fffffff] [ 2.313115][ T0] NODE_DATA(1) allocated [mem 0x5f7ffd0000-0x5f7fffafff]
[ 5.391151][ T1] smp: Brought up 2 nodes, 96 CPUs
Full dmesg is attached.
Best Regards,
Huang, Ying
Peter Zijlstra
maintainerQuoted 55 lines · Ying
Hi, Peter,
Quoted 10 lines · Peter Zijlstra
Quoted 6 lines · Huang Ying
After commit
c221c0b0308f("device-dax: "Hotplug" persistent memory for use like normal RAM"), the PMEM could be used as the cost-effective volatile memory in separate NUMA nodes. In a typical memory tiering system, there are CPUs, DRAM and PMEM in each physical NUMA node. The CPUs and the DRAM will be put in one logical node, while the PMEM will be put in another (faked) logical node.So what does a system like that actually look like, SLIT table wise, and how does that affect
init_numa_topology_type()?The SLIT table is as follows,
[000h 0000 4] Signature : "SLIT" [System Locality Information Table] [004h 0004 4] Table Length : 0000042C [008h 0008 1] Revision : 01 [009h 0009 1] Checksum : 59 [00Ah 0010 6] Oem ID : "INTEL " [010h 0016 8] Oem Table ID : "S2600WF " [018h 0024 4] Oem Revision : 00000001 [01Ch 0028 4] Asl Compiler ID : "INTL" [020h 0032 4] Asl Compiler Revision : 20091013[024h 0036 8] Localities : 0000000000000004 [02Ch 0044 4] Locality 0 : 0A 15 11 1C [030h 0048 4] Locality 1 : 15 0A 1C 11 [034h 0052 4] Locality 2 : 11 1C 0A 1C [038h 0056 4] Locality 3 : 1C 11 1C 0AThe
numactl -Houtput is as follows,available: 4 nodes (0-3)
node 0 cpus: 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 node 0 size: 64136 MB
node 0 free: 5981 MB
node 1 cpus: 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 node 1 size: 64466 MB
node 1 free: 10415 MB
node 2 cpus:
node 2 size: 253952 MB
node 2 free: 253920 MB
node 3 cpus:
node 3 size: 253952 MB
node 3 free: 253951 MB
node distances:
node 0 1 2 30: 10 21 17 28 1: 21 10 28 17 2: 17 28 10 28 3: 28 17 28 10
Quoted 5 lines
init_numa_topology_type()setsched_numa_topology_typetoNUMA_DIRECT.The node 0 and node 1 are onlined during boot. While the PMEM node, that is, node 2 and node 3 are onlined later. As in the following dmesg snippet.
But how? sched_init_numa() scans the *whole* SLIT table to determine nr_levels / sched_domains_numa_levels, even offline nodes. Therefore it should find 4 distinct distance values and end up not selecting NUMA_DIRECT.
Similarly for the other types it uses for_each_online_node(), which would include the pmem nodes once they've been onlined, but I'm thinking we explicitly want to skip CPU-less nodes in that iteration.
Quoted 62 lines · Peter Zijlstra
Quoted 55 lines · Ying
Hi, Peter,
Quoted 10 lines · Peter Zijlstra
Quoted 6 lines · Huang Ying
After commit
c221c0b0308f("device-dax: "Hotplug" persistent memory for use like normal RAM"), the PMEM could be used as the cost-effective volatile memory in separate NUMA nodes. In a typical memory tiering system, there are CPUs, DRAM and PMEM in each physical NUMA node. The CPUs and the DRAM will be put in one logical node, while the PMEM will be put in another (faked) logical node.So what does a system like that actually look like, SLIT table wise, and how does that affect
init_numa_topology_type()?The SLIT table is as follows,
[000h 0000 4] Signature : "SLIT" [System Locality Information Table] [004h 0004 4] Table Length : 0000042C [008h 0008 1] Revision : 01 [009h 0009 1] Checksum : 59 [00Ah 0010 6] Oem ID : "INTEL " [010h 0016 8] Oem Table ID : "S2600WF " [018h 0024 4] Oem Revision : 00000001 [01Ch 0028 4] Asl Compiler ID : "INTL" [020h 0032 4] Asl Compiler Revision : 20091013[024h 0036 8] Localities : 0000000000000004 [02Ch 0044 4] Locality 0 : 0A 15 11 1C [030h 0048 4] Locality 1 : 15 0A 1C 11 [034h 0052 4] Locality 2 : 11 1C 0A 1C [038h 0056 4] Locality 3 : 1C 11 1C 0AThe
numactl -Houtput is as follows,available: 4 nodes (0-3)
node 0 cpus: 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 node 0 size: 64136 MB
node 0 free: 5981 MB
node 1 cpus: 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 node 1 size: 64466 MB
node 1 free: 10415 MB
node 2 cpus:
node 2 size: 253952 MB
node 2 free: 253920 MB
node 3 cpus:
node 3 size: 253952 MB
node 3 free: 253951 MB
node distances:
node 0 1 2 30: 10 21 17 28 1: 21 10 28 17 2: 17 28 10 28 3: 28 17 28 10Quoted 5 lines
init_numa_topology_type()setsched_numa_topology_typetoNUMA_DIRECT.The node 0 and node 1 are onlined during boot. While the PMEM node, that is, node 2 and node 3 are onlined later. As in the following dmesg snippet.
Quoted 8 lines
But how?
sched_init_numa()scans the *whole* SLIT table to determinenr_levels/sched_domains_numa_levels, even offline nodes. Therefore it should find 4 distinct distance values and end up not selectingNUMA_DIRECT.Similarly for the other types it uses
for_each_online_node(), which would include the pmem nodes once they've been onlined, but I'm thinking we explicitly want to skip CPU-less nodes in that iteration.
I used the debug patch as below, and get the log in dmesg as follows,
[ 5.394577][ T1] sched_numa_topology_type: 0, levels: 4, max_distance: 28
I found that I forget another caller of init_numa_topology_type() run during hotplug. I will add another printk() to show it. Sorry about that.
Best Regards,
Huang, Ying
-------------------------------8<------------------------------------ From 11cea4be2db6220333d84f5b168174f534ac0933 Mon Sep 17 00:00:00 2001
From: Huang Ying <[email protected]>
Date: Thu, 13 Jan 2022 09:53:15 +0800
Subject: [PATCH] dbg: show sched_numa_topology_type
--- kernel/sched/topology.c | 4 ++++ 1 file changed, 4 insertions(+)
kernel/sched/topology.c +4 -0
@@ -1914,6 +1914,10 @@ void sched_init_numa(void)init_numa_topology_type (); +pr_info ("sched_numa_topology_type: %d, levels: %d, max_distance: %d\n" , + sched_numa_topology_type, sched_domains_numa_levels, + sched_max_numa_distance); + sched_numa_onlined_nodes =bitmap_alloc (nr_node_ids,GFP_KERNEL );if (!sched_numa_onlined_nodes)return ;
Signature · 1 line
2.30.2
Peter Zijlstra
maintainerQuoted 16 lines · Ying
Quoted 15 lines · Peter Zijlstra
Quoted 14 lines · Ying
Quoted 11 lines · Peter Zijlstra
Quoted 6 lines · Huang Ying
After commit
c221c0b0308f("device-dax: "Hotplug" persistent memory for use like normal RAM"), the PMEM could be used as the cost-effective volatile memory in separate NUMA nodes. In a typical memory tiering system, there are CPUs, DRAM and PMEM in each physical NUMA node. The CPUs and the DRAM will be put in one logical node, while the PMEM will be put in another (faked) logical node.So what does a system like that actually look like, SLIT table wise, and how does that affect
init_numa_topology_type()?The SLIT table is as follows,
<snip>
Quoted 22 lines
Quoted 13 lines
Quoted 7 lines
node distances:
node 0 1 2 30: 10 21 17 28 1: 21 10 28 17 2: 17 28 10 28 3: 28 17 28 10Quoted 5 lines
init_numa_topology_type()setsched_numa_topology_typetoNUMA_DIRECT.The node 0 and node 1 are onlined during boot. While the PMEM node, that is, node 2 and node 3 are onlined later. As in the following dmesg snippet.
Quoted 8 lines
But how?
sched_init_numa()scans the *whole* SLIT table to determinenr_levels/sched_domains_numa_levels, even offline nodes. Therefore it should find 4 distinct distance values and end up not selectingNUMA_DIRECT.Similarly for the other types it uses
for_each_online_node(), which would include the pmem nodes once they've been onlined, but I'm thinking we explicitly want to skip CPU-less nodes in that iteration.
Quoted 7 lines
I used the debug patch as below, and get the log in dmesg as follows,
[ 5.394577][ T1] sched_numa_topology_type: 0, levels: 4, max_distance: 28I found that I forget another caller of
init_numa_topology_type()run during hotplug. I will add anotherprintk()to show it. Sorry about that.
Can you try with this on?
I'm suspecting there's a problem with init_numa_topology_type(); it will never find the max distance due to the _online_ clause in the iteration, since you said the pmem nodes are not online yet.
---
kernel/sched/topology.c +2 -0
@@ -1756,6 +1756,8 @@ static void init_numa_topology_type(void)return ; } } + +WARN (1 ,"no NUMA type determined" ); }
Quoted 62 lines · Peter Zijlstra
Quoted 16 lines · Ying
Quoted 15 lines · Peter Zijlstra
Quoted 14 lines · Ying
Quoted 11 lines · Peter Zijlstra
Quoted 6 lines · Huang Ying
After commit
c221c0b0308f("device-dax: "Hotplug" persistent memory for use like normal RAM"), the PMEM could be used as the cost-effective volatile memory in separate NUMA nodes. In a typical memory tiering system, there are CPUs, DRAM and PMEM in each physical NUMA node. The CPUs and the DRAM will be put in one logical node, while the PMEM will be put in another (faked) logical node.So what does a system like that actually look like, SLIT table wise, and how does that affect
init_numa_topology_type()?The SLIT table is as follows,
<snip>
Quoted 22 lines
Quoted 13 lines
Quoted 7 lines
node distances:
node 0 1 2 30: 10 21 17 28 1: 21 10 28 17 2: 17 28 10 28 3: 28 17 28 10Quoted 5 lines
init_numa_topology_type()setsched_numa_topology_typetoNUMA_DIRECT.The node 0 and node 1 are onlined during boot. While the PMEM node, that is, node 2 and node 3 are onlined later. As in the following dmesg snippet.
Quoted 8 lines
But how?
sched_init_numa()scans the *whole* SLIT table to determinenr_levels/sched_domains_numa_levels, even offline nodes. Therefore it should find 4 distinct distance values and end up not selectingNUMA_DIRECT.Similarly for the other types it uses
for_each_online_node(), which would include the pmem nodes once they've been onlined, but I'm thinking we explicitly want to skip CPU-less nodes in that iteration.Quoted 7 lines
I used the debug patch as below, and get the log in dmesg as follows,
[ 5.394577][ T1] sched_numa_topology_type: 0, levels: 4, max_distance: 28I found that I forget another caller of
init_numa_topology_type()run during hotplug. I will add anotherprintk()to show it. Sorry about that.Can you try with this on?
I'm suspecting there's a problem with
init_numa_topology_type(); it will never find the max distance due to the _online_ clause in the iteration, since you said the pmem nodes are not online yet.---
kernel/sched/topology.c +0
-0
kernel/sched/topology.c static void init_numa_topology_type(void) · line 1761
Quoted 6 lines
kernel/sched/topology.c +2
-0@@ -1756,6 +1756,8 @@ static void init_numa_topology_type(void)return ; } } + +WARN (1 ,"no NUMA type determined" ); }
Sure. Will do this.
Best Regards,
Huang, Ying
Quoted 62 lines · Peter Zijlstra
Quoted 16 lines · Ying
Quoted 15 lines · Peter Zijlstra
Quoted 14 lines · Ying
Quoted 11 lines · Peter Zijlstra
Quoted 6 lines · Huang Ying
After commit
c221c0b0308f("device-dax: "Hotplug" persistent memory for use like normal RAM"), the PMEM could be used as the cost-effective volatile memory in separate NUMA nodes. In a typical memory tiering system, there are CPUs, DRAM and PMEM in each physical NUMA node. The CPUs and the DRAM will be put in one logical node, while the PMEM will be put in another (faked) logical node.So what does a system like that actually look like, SLIT table wise, and how does that affect
init_numa_topology_type()?The SLIT table is as follows,
<snip>
Quoted 22 lines
Quoted 13 lines
Quoted 7 lines
node distances:
node 0 1 2 30: 10 21 17 28 1: 21 10 28 17 2: 17 28 10 28 3: 28 17 28 10Quoted 5 lines
init_numa_topology_type()setsched_numa_topology_typetoNUMA_DIRECT.The node 0 and node 1 are onlined during boot. While the PMEM node, that is, node 2 and node 3 are onlined later. As in the following dmesg snippet.
Quoted 8 lines
But how?
sched_init_numa()scans the *whole* SLIT table to determinenr_levels/sched_domains_numa_levels, even offline nodes. Therefore it should find 4 distinct distance values and end up not selectingNUMA_DIRECT.Similarly for the other types it uses
for_each_online_node(), which would include the pmem nodes once they've been onlined, but I'm thinking we explicitly want to skip CPU-less nodes in that iteration.Quoted 7 lines
I used the debug patch as below, and get the log in dmesg as follows,
[ 5.394577][ T1] sched_numa_topology_type: 0, levels: 4, max_distance: 28I found that I forget another caller of
init_numa_topology_type()run during hotplug. I will add anotherprintk()to show it. Sorry about that.Can you try with this on?
I'm suspecting there's a problem with
init_numa_topology_type(); it will never find the max distance due to the _online_ clause in the iteration, since you said the pmem nodes are not online yet.---
kernel/sched/topology.c +0
-0
kernel/sched/topology.c static void init_numa_topology_type(void) · line 1761
Quoted 6 lines
kernel/sched/topology.c +2
-0@@ -1756,6 +1756,8 @@ static void init_numa_topology_type(void)return ; } } + +WARN (1 ,"no NUMA type determined" ); }
Hi, Peter,
I have run the test, the warning is triggered in the dmesg as follows. I will continue to debug hotplug tomorrow.
[ 5.400923][ T1] ------------[ cut here ]------------ [ 5.401917][ T1] no NUMA type determined [ 5.401921][ T1] WARNING: CPU: 0 PID: 1 at kernel/sched/topology.c:1760 init_numa_topology_type+0x199/0x1c0 [ 5.403918][ T1] Modules linked in: [ 5.404917][ T1] CPU: 0 PID: 1 Comm: swapper/0 Not tainted 5.16.0-rc8-00053-gbe30433a13c0 #1 [ 5.405917][ T1] Hardware name: Intel Corporation S2600WFD/S2600WFD, BIOS SE5C620.86B.0D.01.0286.011120190816 01/11/2019 [ 5.406917][ T1] RIP: 0010:init_numa_topology_type+0x199/0x1c0 [ 5.407917][ T1] Code: de 82 41 89 dc e8 07 4f 4e 00 3d 00 04 00 00 44 0f 4e e0 3d ff 03 00 00 0f 8e ca fe ff ff 48 c7 c7 a7 88 55 82 e8 0c e5 b3 00 <0f> 0b e9 74 ff ff ff 66 66 2e 0f 1f 84 00 00 00 00 00 66 66 2e 0f [ 5.408917][ T1] RSP: 0000:ffffc900000b7e00 EFLAGS: 00010286 [ 5.409917][ T1] RAX: 0000000000000000 RBX: 0000000000000400 RCX: c0000000ffff7fff [ 5.410917][ T1] RDX: ffffc900000b7c28 RSI: 00000000ffff7fff RDI: 0000000000000000 [ 5.411917][ T1] RBP: 000000000000001c R08: 0000000000000000 R09: ffffc900000b7c20 [ 5.412917][ T1] R10: 0000000000000001 R11: 0000000000000001 R12: 0000000000000400 [ 5.413917][ T1] R13: 0000000000000400 R14: 0000000000000400 R15: 000000000000000c [ 5.414917][ T1] FS: 0000000000000000(0000) GS:ffff88903f600000(0000) knlGS:0000000000000000 [ 5.415917][ T1] CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 [ 5.416917][ T1] CR2: ffff88df7fc01000 CR3: 0000005f7ec0a001 CR4: 00000000007706f0 [ 5.417917][ T1] DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000 [ 5.418917][ T1] DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400 [ 5.419917][ T1] PKRU: 55555554 [ 5.420917][ T1] Call Trace: [ 5.421919][ T1] <TASK> [ 5.422919][ T1] sched_init_numa+0x4a7/0x5c0 [ 5.423918][ T1] sched_init_smp+0x18/0x79 [ 5.424918][ T1] kernel_init_freeable+0x136/0x276 [ 5.425918][ T1] ? rest_init+0x100/0x100 [ 5.426917][ T1] kernel_init+0x16/0x140 [ 5.427917][ T1] ret_from_fork+0x1f/0x30 [ 5.428918][ T1] </TASK> [ 5.429919][ T1] ---[ end trace aa5563c4363f1ba3 ]--- [ 5.430917][ T1] sched_numa_topology_type: 0, levels: 4, max_distance: 28
Best Regards,
Huang, Ying
Quoted 73 lines · Ying
Quoted 62 lines · Peter Zijlstra
Quoted 16 lines · Ying
Quoted 15 lines · Peter Zijlstra
Quoted 14 lines · Ying
Quoted 11 lines · Peter Zijlstra
Quoted 6 lines · Huang Ying
After commit
c221c0b0308f("device-dax: "Hotplug" persistent memory for use like normal RAM"), the PMEM could be used as the cost-effective volatile memory in separate NUMA nodes. In a typical memory tiering system, there are CPUs, DRAM and PMEM in each physical NUMA node. The CPUs and the DRAM will be put in one logical node, while the PMEM will be put in another (faked) logical node.So what does a system like that actually look like, SLIT table wise, and how does that affect
init_numa_topology_type()?The SLIT table is as follows,
<snip>
Quoted 22 lines
Quoted 13 lines
Quoted 7 lines
node distances:
node 0 1 2 30: 10 21 17 28 1: 21 10 28 17 2: 17 28 10 28 3: 28 17 28 10Quoted 5 lines
init_numa_topology_type()setsched_numa_topology_typetoNUMA_DIRECT.The node 0 and node 1 are onlined during boot. While the PMEM node, that is, node 2 and node 3 are onlined later. As in the following dmesg snippet.
Quoted 8 lines
But how?
sched_init_numa()scans the *whole* SLIT table to determinenr_levels/sched_domains_numa_levels, even offline nodes. Therefore it should find 4 distinct distance values and end up not selectingNUMA_DIRECT.Similarly for the other types it uses
for_each_online_node(), which would include the pmem nodes once they've been onlined, but I'm thinking we explicitly want to skip CPU-less nodes in that iteration.Quoted 7 lines
I used the debug patch as below, and get the log in dmesg as follows,
[ 5.394577][ T1] sched_numa_topology_type: 0, levels: 4, max_distance: 28I found that I forget another caller of
init_numa_topology_type()run during hotplug. I will add anotherprintk()to show it. Sorry about that.Can you try with this on?
I'm suspecting there's a problem with
init_numa_topology_type(); it will never find the max distance due to the _online_ clause in the iteration, since you said the pmem nodes are not online yet.---
kernel/sched/topology.c +0
-0kernel/sched/topology.c static void init_numa_topology_type(void) · line 1761
Quoted 6 lines
kernel/sched/topology.c +2
-0@@ -1756,6 +1756,8 @@ static void init_numa_topology_type(void)return ; } } + +WARN (1 ,"no NUMA type determined" ); }Hi, Peter,
Quoted 2 lines
I have run the test, the warning is triggered in the dmesg as follows. I will continue to debug hotplug tomorrow.
I did more experiments and found that init_numa_topology_type() will not be called during PMEM nodes plugging in. Because it will only be called when a CPU of a never-onlined-before node is onlined. There's no CPU on the PMEM nodes (2/3). So, when the PMEM node is onlined, init_numa_topology_type() will not be called. And sched_numa_topology_type will not be changed.
Best Regards,
Huang, Ying