| Message ID | 1720169301-21002-1-git-send-email-ssengar@linux.microsoft.com (mailing list archive) |
|---|---|
| State | New |
| Headers | show |
| Series | mm/vmstat: Defer the refresh_zone_stat_thresholds after all CPUs bringup | expand |
On Fri, 5 Jul 2024 01:48:21 -0700 Saurabh Sengar <ssengar@linux.microsoft.com> wrote: > refresh_zone_stat_thresholds function has two loops which is expensive for > higher number of CPUs and NUMA nodes. > > Below is the rough estimation of total iterations done by these loops > based on number of NUMA and CPUs. > > Total number of iterations: nCPU * 2 * Numa * mCPU > Where: > nCPU = total number of CPUs > Numa = total number of NUMA nodes > mCPU = mean value of total CPUs (e.g., 512 for 1024 total CPUs) > > For the system under test with 16 NUMA nodes and 1024 CPUs, this > results in a substantial increase in the number of loop iterations > during boot-up when NUMA is enabled: > > No NUMA = 1024*2*1*512 = 1,048,576 : Here refresh_zone_stat_thresholds > takes around 224 ms total for all the CPUs in the system under test. > 16 NUMA = 1024*2*16*512 = 16,777,216 : Here refresh_zone_stat_thresholds > takes around 4.5 seconds total for all the CPUs in the system under test. Did you measure the overall before-and-after times? IOW, how much of that 4.5s do we reclaim? > Calling this for each CPU is expensive when there are large number > of CPUs along with multiple NUMAs. Fix this by deferring > refresh_zone_stat_thresholds to be called later at once when all the > secondary CPUs are up. Also, register the DYN hooks to keep the > existing hotplug functionality intact. > Seems risky - we'll now have online CPUs which have unintialized data, yes? What assurance do we have that this data won't be accessed? Another approach might be to make the code a bit smarter - instead of calculating thresholds for the whole world, we make incremental changes to the existing thresholds on behalf of the new resource which just became available?
diff --git a/mm/vmstat.c b/mm/vmstat.c index 6e3347789eb2..5c0df62238d9 100644 --- a/mm/vmstat.c +++ b/mm/vmstat.c @@ -31,6 +31,7 @@ #include "internal.h" +static int vmstat_late_init_done; #ifdef CONFIG_NUMA int sysctl_vm_numa_stat = ENABLE_NUMA_STAT; @@ -2107,7 +2108,8 @@ static void __init init_cpu_node_state(void) static int vmstat_cpu_online(unsigned int cpu) { - refresh_zone_stat_thresholds(); + if (vmstat_late_init_done) + refresh_zone_stat_thresholds(); if (!node_state(cpu_to_node(cpu), N_CPU)) { node_set_state(cpu_to_node(cpu), N_CPU); @@ -2139,6 +2141,14 @@ static int vmstat_cpu_dead(unsigned int cpu) return 0; } +static int __init vmstat_late_init(void) +{ + refresh_zone_stat_thresholds(); + vmstat_late_init_done = 1; + + return 0; +} +late_initcall(vmstat_late_init); #endif struct workqueue_struct *mm_percpu_wq;
refresh_zone_stat_thresholds function has two loops which is expensive for higher number of CPUs and NUMA nodes. Below is the rough estimation of total iterations done by these loops based on number of NUMA and CPUs. Total number of iterations: nCPU * 2 * Numa * mCPU Where: nCPU = total number of CPUs Numa = total number of NUMA nodes mCPU = mean value of total CPUs (e.g., 512 for 1024 total CPUs) For the system under test with 16 NUMA nodes and 1024 CPUs, this results in a substantial increase in the number of loop iterations during boot-up when NUMA is enabled: No NUMA = 1024*2*1*512 = 1,048,576 : Here refresh_zone_stat_thresholds takes around 224 ms total for all the CPUs in the system under test. 16 NUMA = 1024*2*16*512 = 16,777,216 : Here refresh_zone_stat_thresholds takes around 4.5 seconds total for all the CPUs in the system under test. Calling this for each CPU is expensive when there are large number of CPUs along with multiple NUMAs. Fix this by deferring refresh_zone_stat_thresholds to be called later at once when all the secondary CPUs are up. Also, register the DYN hooks to keep the existing hotplug functionality intact. Signed-off-by: Saurabh Sengar <ssengar@linux.microsoft.com> --- mm/vmstat.c | 12 +++++++++++- 1 file changed, 11 insertions(+), 1 deletion(-)