Commit 9a46ad6d authored by Shaohua Li's avatar Shaohua Li Committed by Linus Torvalds

smp: make smp_call_function_many() use logic similar to smp_call_function_single()

I'm testing swapout workload in a two-socket Xeon machine.  The workload
has 10 threads, each thread sequentially accesses separate memory
region.  TLB flush overhead is very big in the workload.  For each page,
page reclaim need move it from active lru list and then unmap it.  Both
need a TLB flush.  And this is a multthread workload, TLB flush happens
in 10 CPUs.  In X86, TLB flush uses generic smp_call)function.  So this
workload stress smp_call_function_many heavily.

Without patch, perf shows:
+  24.49%  [k] generic_smp_call_function_interrupt
-  21.72%  [k] _raw_spin_lock
   - _raw_spin_lock
      + 79.80% __page_check_address
      + 6.42% generic_smp_call_function_interrupt
      + 3.31% get_swap_page
      + 2.37% free_pcppages_bulk
      + 1.75% handle_pte_fault
      + 1.54% put_super
      + 1.41% grab_super_passive
      + 1.36% __swap_duplicate
      + 0.68% blk_flush_plug_list
      + 0.62% swap_info_get
+   6.55%  [k] flush_tlb_func
+   6.46%  [k] smp_call_function_many
+   5.09%  [k] call_function_interrupt
+   4.75%  [k] default_send_IPI_mask_sequence_phys
+   2.18%  [k] find_next_bit

swapout throughput is around 1300M/s.

With the patch, perf shows:
-  27.23%  [k] _raw_spin_lock
   - _raw_spin_lock
      + 80.53% __page_check_address
      + 8.39% generic_smp_call_function_single_interrupt
      + 2.44% get_swap_page
      + 1.76% free_pcppages_bulk
      + 1.40% handle_pte_fault
      + 1.15% __swap_duplicate
      + 1.05% put_super
      + 0.98% grab_super_passive
      + 0.86% blk_flush_plug_list
      + 0.57% swap_info_get
+   8.25%  [k] default_send_IPI_mask_sequence_phys
+   7.55%  [k] call_function_interrupt
+   7.47%  [k] smp_call_function_many
+   7.25%  [k] flush_tlb_func
+   3.81%  [k] _raw_spin_lock_irqsave
+   3.78%  [k] generic_smp_call_function_single_interrupt

swapout throughput is around 1400M/s.  So there is around a 7%
improvement, and total cpu utilization doesn't change.

Without the patch, cfd_data is shared by all CPUs.
generic_smp_call_function_interrupt does read/write cfd_data several times
which will create a lot of cache ping-pong.  With the patch, the data
becomes per-cpu.  The ping-pong is avoided.  And from the perf data, this
doesn't make call_single_queue lock contend.

Next step is to remove generic_smp_call_function_interrupt() from arch
Signed-off-by: default avatarShaohua Li <>
Cc: Peter Zijlstra <>
Cc: Ingo Molnar <>
Cc: Steven Rostedt <>
Cc: Jens Axboe <>
Cc: Linus Torvalds <>
Signed-off-by: default avatarAndrew Morton <>
Signed-off-by: default avatarLinus Torvalds <>
parent 6d1c7cca
......@@ -89,7 +89,8 @@ void kick_all_cpus_sync(void);
void __init call_function_init(void);
void generic_smp_call_function_single_interrupt(void);
void generic_smp_call_function_interrupt(void);
#define generic_smp_call_function_interrupt \
static inline void call_function_init(void) { }
......@@ -16,22 +16,12 @@
#include "smpboot.h"
static struct {
struct list_head queue;
raw_spinlock_t lock;
} call_function __cacheline_aligned_in_smp =
.queue = LIST_HEAD_INIT(call_function.queue),
.lock = __RAW_SPIN_LOCK_UNLOCKED(call_function.lock),
enum {
struct call_function_data {
struct call_single_data csd;
atomic_t refs;
struct call_single_data __percpu *csd;
cpumask_var_t cpumask;
cpumask_var_t cpumask_ipi;
......@@ -60,6 +50,11 @@ hotplug_cfd(struct notifier_block *nfb, unsigned long action, void *hcpu)
if (!zalloc_cpumask_var_node(&cfd->cpumask_ipi, GFP_KERNEL,
return notifier_from_errno(-ENOMEM);
cfd->csd = alloc_percpu(struct call_single_data);
if (!cfd->csd) {
return notifier_from_errno(-ENOMEM);
......@@ -70,6 +65,7 @@ hotplug_cfd(struct notifier_block *nfb, unsigned long action, void *hcpu)
......@@ -170,85 +166,6 @@ void generic_exec_single(int cpu, struct call_single_data *data, int wait)
* Invoked by arch to handle an IPI for call function. Must be called with
* interrupts disabled.
void generic_smp_call_function_interrupt(void)
struct call_function_data *data;
int cpu = smp_processor_id();
* Shouldn't receive this interrupt on a cpu that is not yet online.
* Ensure entry is visible on call_function_queue after we have
* entered the IPI. See comment in smp_call_function_many.
* If we don't have this, then we may miss an entry on the list
* and never get another IPI to process it.
* It's ok to use list_for_each_rcu() here even though we may
* delete 'pos', since list_del_rcu() doesn't clear ->next
list_for_each_entry_rcu(data, &call_function.queue, csd.list) {
int refs;
smp_call_func_t func;
* Since we walk the list without any locks, we might
* see an entry that was completed, removed from the
* list and is in the process of being reused.
* We must check that the cpu is in the cpumask before
* checking the refs, and both must be set before
* executing the callback on this cpu.
if (!cpumask_test_cpu(cpu, data->cpumask))
if (atomic_read(&data->refs) == 0)
func = data->csd.func; /* save for later warn */
* If the cpu mask is not still set then func enabled
* interrupts (BUG), and this cpu took another smp call
* function interrupt and executed func(info) twice
* on this cpu. That nested execution decremented refs.
if (!cpumask_test_and_clear_cpu(cpu, data->cpumask)) {
WARN(1, "%pf enabled interrupts and double executed\n", func);
refs = atomic_dec_return(&data->refs);
WARN_ON(refs < 0);
if (refs)
* Invoked by arch to handle an IPI for call function single. Must be
* called from the arch with interrupts disabled.
......@@ -453,8 +370,7 @@ void smp_call_function_many(const struct cpumask *mask,
smp_call_func_t func, void *info, bool wait)
struct call_function_data *data;
unsigned long flags;
int refs, cpu, next_cpu, this_cpu = smp_processor_id();
int cpu, next_cpu, this_cpu = smp_processor_id();
* Can deadlock when called with interrupts disabled.
......@@ -486,50 +402,13 @@ void smp_call_function_many(const struct cpumask *mask,
data = &__get_cpu_var(cfd_data);
/* This BUG_ON verifies our reuse assertions and can be removed */
BUG_ON(atomic_read(&data->refs) || !cpumask_empty(data->cpumask));
* The global call function queue list add and delete are protected
* by a lock, but the list is traversed without any lock, relying
* on the rcu list add and delete to allow safe concurrent traversal.
* We reuse the call function data without waiting for any grace
* period after some other cpu removes it from the global queue.
* This means a cpu might find our data block as it is being
* filled out.
* We hold off the interrupt handler on the other cpu by
* ordering our writes to the cpu mask vs our setting of the
* refs counter. We assert only the cpu owning the data block
* will set a bit in cpumask, and each bit will only be cleared
* by the subject cpu. Each cpu must first find its bit is
* set and then check that refs is set indicating the element is
* ready to be processed, otherwise it must skip the entry.
* On the previous iteration refs was set to 0 by another cpu.
* To avoid the use of transitivity, set the counter to 0 here
* so the wmb will pair with the rmb in the interrupt handler.
atomic_set(&data->refs, 0); /* convert 3rd to 1st party write */
data->csd.func = func;
data-> = info;
/* Ensure 0 refs is visible before mask. Also orders func and info */
/* We rely on the "and" being processed before the store */
cpumask_and(data->cpumask, mask, cpu_online_mask);
cpumask_clear_cpu(this_cpu, data->cpumask);
refs = cpumask_weight(data->cpumask);
/* Some callers race with other cpus changing the passed mask */
if (unlikely(!refs)) {
if (unlikely(!cpumask_weight(data->cpumask)))
* After we put an entry into the list, data->cpumask
......@@ -537,34 +416,32 @@ void smp_call_function_many(const struct cpumask *mask,
* a SMP function call, so data->cpumask will be zero.
cpumask_copy(data->cpumask_ipi, data->cpumask);
raw_spin_lock_irqsave(&call_function.lock, flags);
* Place entry at the _HEAD_ of the list, so that any cpu still
* observing the entry in generic_smp_call_function_interrupt()
* will not miss any other list entries:
list_add_rcu(&data->csd.list, &call_function.queue);
* We rely on the wmb() in list_add_rcu to complete our writes
* to the cpumask before this write to refs, which indicates
* data is on the list and is ready to be processed.
atomic_set(&data->refs, refs);
raw_spin_unlock_irqrestore(&call_function.lock, flags);
* Make the list addition visible before sending the ipi.
* (IPIs must obey or appear to obey normal Linux cache
* coherency rules -- see comment in generic_exec_single).
for_each_cpu(cpu, data->cpumask) {
struct call_single_data *csd = per_cpu_ptr(data->csd, cpu);
struct call_single_queue *dst =
&per_cpu(call_single_queue, cpu);
unsigned long flags;
csd->func = func;
csd->info = info;
raw_spin_lock_irqsave(&dst->lock, flags);
list_add_tail(&csd->list, &dst->list);
raw_spin_unlock_irqrestore(&dst->lock, flags);
/* Send a message to all CPUs in the map */
/* Optionally wait for the CPUs to complete */
if (wait)
if (wait) {
for_each_cpu(cpu, data->cpumask) {
struct call_single_data *csd =
per_cpu_ptr(data->csd, cpu);
Markdown is supported
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment