Commit 4cd7bfcc authored by Herbert Xu's avatar Herbert Xu Committed by David S. Miller

[IPSEC]: Proper percpu handling in flow cache.

parent 80020f5d
...@@ -15,6 +15,9 @@ ...@@ -15,6 +15,9 @@
#include <linux/interrupt.h> #include <linux/interrupt.h>
#include <linux/smp.h> #include <linux/smp.h>
#include <linux/completion.h> #include <linux/completion.h>
#include <linux/percpu.h>
#include <linux/bitops.h>
#include <linux/notifier.h>
#include <net/flow.h> #include <net/flow.h>
#include <asm/atomic.h> #include <asm/atomic.h>
#include <asm/semaphore.h> #include <asm/semaphore.h>
...@@ -33,7 +36,10 @@ atomic_t flow_cache_genid = ATOMIC_INIT(0); ...@@ -33,7 +36,10 @@ atomic_t flow_cache_genid = ATOMIC_INIT(0);
static u32 flow_hash_shift; static u32 flow_hash_shift;
#define flow_hash_size (1 << flow_hash_shift) #define flow_hash_size (1 << flow_hash_shift)
static struct flow_cache_entry **flow_table; static DEFINE_PER_CPU(struct flow_cache_entry **, flow_tables) = { NULL };
#define flow_table(cpu) (per_cpu(flow_tables, cpu))
static kmem_cache_t *flow_cachep; static kmem_cache_t *flow_cachep;
static int flow_lwm, flow_hwm; static int flow_lwm, flow_hwm;
...@@ -43,11 +49,14 @@ struct flow_percpu_info { ...@@ -43,11 +49,14 @@ struct flow_percpu_info {
u32 hash_rnd; u32 hash_rnd;
int count; int count;
} ____cacheline_aligned; } ____cacheline_aligned;
static struct flow_percpu_info flow_hash_info[NR_CPUS]; static DEFINE_PER_CPU(struct flow_percpu_info, flow_hash_info) = { 0 };
#define flow_hash_rnd_recalc(cpu) (flow_hash_info[cpu].hash_rnd_recalc) #define flow_hash_rnd_recalc(cpu) \
#define flow_hash_rnd(cpu) (flow_hash_info[cpu].hash_rnd) (per_cpu(flow_hash_info, cpu).hash_rnd_recalc)
#define flow_count(cpu) (flow_hash_info[cpu].count) #define flow_hash_rnd(cpu) \
(per_cpu(flow_hash_info, cpu).hash_rnd)
#define flow_count(cpu) \
(per_cpu(flow_hash_info, cpu).count)
static struct timer_list flow_hash_rnd_timer; static struct timer_list flow_hash_rnd_timer;
...@@ -55,17 +64,24 @@ static struct timer_list flow_hash_rnd_timer; ...@@ -55,17 +64,24 @@ static struct timer_list flow_hash_rnd_timer;
struct flow_flush_info { struct flow_flush_info {
atomic_t cpuleft; atomic_t cpuleft;
unsigned long cpumap;
struct completion completion; struct completion completion;
}; };
static struct tasklet_struct flow_flush_tasklets[NR_CPUS]; static DEFINE_PER_CPU(struct tasklet_struct, flow_flush_tasklets) = { NULL };
static DECLARE_MUTEX(flow_flush_sem);
#define flow_flush_tasklet(cpu) (&per_cpu(flow_flush_tasklets, cpu))
static DECLARE_MUTEX(flow_cache_cpu_sem);
static unsigned long flow_cache_cpu_map;
static unsigned int flow_cache_cpu_count;
static void flow_cache_new_hashrnd(unsigned long arg) static void flow_cache_new_hashrnd(unsigned long arg)
{ {
int i; int i;
for (i = 0; i < NR_CPUS; i++) for (i = 0; i < NR_CPUS; i++)
flow_hash_rnd_recalc(i) = 1; if (test_bit(i, &flow_cache_cpu_map))
flow_hash_rnd_recalc(i) = 1;
flow_hash_rnd_timer.expires = jiffies + FLOW_HASH_RND_PERIOD; flow_hash_rnd_timer.expires = jiffies + FLOW_HASH_RND_PERIOD;
add_timer(&flow_hash_rnd_timer); add_timer(&flow_hash_rnd_timer);
...@@ -79,7 +95,7 @@ static void __flow_cache_shrink(int cpu, int shrink_to) ...@@ -79,7 +95,7 @@ static void __flow_cache_shrink(int cpu, int shrink_to)
for (i = 0; i < flow_hash_size; i++) { for (i = 0; i < flow_hash_size; i++) {
int k = 0; int k = 0;
flp = &flow_table[cpu*flow_hash_size+i]; flp = &flow_table(cpu)[i];
while ((fle = *flp) != NULL && k < shrink_to) { while ((fle = *flp) != NULL && k < shrink_to) {
k++; k++;
flp = &fle->next; flp = &fle->next;
...@@ -159,11 +175,16 @@ void *flow_cache_lookup(struct flowi *key, u16 family, u8 dir, ...@@ -159,11 +175,16 @@ void *flow_cache_lookup(struct flowi *key, u16 family, u8 dir,
local_bh_disable(); local_bh_disable();
cpu = smp_processor_id(); cpu = smp_processor_id();
fle = NULL;
if (!test_bit(cpu, &flow_cache_cpu_map))
goto nocache;
if (flow_hash_rnd_recalc(cpu)) if (flow_hash_rnd_recalc(cpu))
flow_new_hash_rnd(cpu); flow_new_hash_rnd(cpu);
hash = flow_hash_code(key, cpu); hash = flow_hash_code(key, cpu);
head = &flow_table[(cpu << flow_hash_shift) + hash]; head = &flow_table(cpu)[hash];
for (fle = *head; fle; fle = fle->next) { for (fle = *head; fle; fle = fle->next) {
if (fle->family == family && if (fle->family == family &&
fle->dir == dir && fle->dir == dir &&
...@@ -197,6 +218,7 @@ void *flow_cache_lookup(struct flowi *key, u16 family, u8 dir, ...@@ -197,6 +218,7 @@ void *flow_cache_lookup(struct flowi *key, u16 family, u8 dir,
} }
} }
nocache:
{ {
void *obj; void *obj;
atomic_t *obj_ref; atomic_t *obj_ref;
...@@ -230,7 +252,7 @@ static void flow_cache_flush_tasklet(unsigned long data) ...@@ -230,7 +252,7 @@ static void flow_cache_flush_tasklet(unsigned long data)
for (i = 0; i < flow_hash_size; i++) { for (i = 0; i < flow_hash_size; i++) {
struct flow_cache_entry *fle; struct flow_cache_entry *fle;
fle = flow_table[(cpu << flow_hash_shift) + i]; fle = flow_table(cpu)[i];
for (; fle; fle = fle->next) { for (; fle; fle = fle->next) {
unsigned genid = atomic_read(&flow_cache_genid); unsigned genid = atomic_read(&flow_cache_genid);
...@@ -246,6 +268,7 @@ static void flow_cache_flush_tasklet(unsigned long data) ...@@ -246,6 +268,7 @@ static void flow_cache_flush_tasklet(unsigned long data)
complete(&info->completion); complete(&info->completion);
} }
static void flow_cache_flush_per_cpu(void *) __attribute__((__unused__));
static void flow_cache_flush_per_cpu(void *data) static void flow_cache_flush_per_cpu(void *data)
{ {
struct flow_flush_info *info = data; struct flow_flush_info *info = data;
...@@ -253,23 +276,31 @@ static void flow_cache_flush_per_cpu(void *data) ...@@ -253,23 +276,31 @@ static void flow_cache_flush_per_cpu(void *data)
struct tasklet_struct *tasklet; struct tasklet_struct *tasklet;
cpu = smp_processor_id(); cpu = smp_processor_id();
tasklet = &flow_flush_tasklets[cpu]; if (!test_bit(cpu, &info->cpumap))
tasklet_init(tasklet, flow_cache_flush_tasklet, (unsigned long)info); return;
tasklet = flow_flush_tasklet(cpu);
tasklet->data = (unsigned long)info;
tasklet_schedule(tasklet); tasklet_schedule(tasklet);
} }
void flow_cache_flush(void) void flow_cache_flush(void)
{ {
struct flow_flush_info info; struct flow_flush_info info;
static DECLARE_MUTEX(flow_flush_sem);
down(&flow_cache_cpu_sem);
info.cpumap = flow_cache_cpu_map;
atomic_set(&info.cpuleft, flow_cache_cpu_count);
up(&flow_cache_cpu_sem);
atomic_set(&info.cpuleft, num_online_cpus());
init_completion(&info.completion); init_completion(&info.completion);
down(&flow_flush_sem); down(&flow_flush_sem);
local_bh_disable(); local_bh_disable();
smp_call_function(flow_cache_flush_per_cpu, &info, 1, 0); smp_call_function(flow_cache_flush_per_cpu, &info, 1, 0);
flow_cache_flush_per_cpu(&info); flow_cache_flush_tasklet((unsigned long)&info);
local_bh_enable(); local_bh_enable();
wait_for_completion(&info.completion); wait_for_completion(&info.completion);
...@@ -277,11 +308,51 @@ void flow_cache_flush(void) ...@@ -277,11 +308,51 @@ void flow_cache_flush(void)
up(&flow_flush_sem); up(&flow_flush_sem);
} }
static int __init flow_cache_init(void) static void __devinit flow_cache_cpu_online(int cpu)
{ {
struct tasklet_struct *tasklet;
unsigned long order; unsigned long order;
int i;
flow_hash_rnd_recalc(cpu) = 1;
for (order = 0;
(PAGE_SIZE << order) <
(sizeof(struct flow_cache_entry *)*flow_hash_size);
order++)
/* NOTHING */;
flow_table(cpu) = (struct flow_cache_entry **)
__get_free_pages(GFP_KERNEL, order);
memset(flow_table(cpu), 0, PAGE_SIZE << order);
tasklet = flow_flush_tasklet(cpu);
tasklet_init(tasklet, flow_cache_flush_tasklet, 0);
down(&flow_cache_cpu_sem);
set_bit(cpu, &flow_cache_cpu_map);
flow_cache_cpu_count++;
up(&flow_cache_cpu_sem);
}
static int __devinit flow_cache_cpu_notify(struct notifier_block *self,
unsigned long action, void *hcpu)
{
unsigned long cpu = (unsigned long)cpu;
switch (action) {
case CPU_UP_PREPARE:
flow_cache_cpu_online(cpu);
break;
}
return NOTIFY_OK;
}
static struct notifier_block __devinitdata flow_cache_cpu_nb = {
.notifier_call = flow_cache_cpu_notify,
};
static int __init flow_cache_init(void)
{
flow_cachep = kmem_cache_create("flow_cache", flow_cachep = kmem_cache_create("flow_cache",
sizeof(struct flow_cache_entry), sizeof(struct flow_cache_entry),
0, SLAB_HWCACHE_ALIGN, 0, SLAB_HWCACHE_ALIGN,
...@@ -294,27 +365,13 @@ static int __init flow_cache_init(void) ...@@ -294,27 +365,13 @@ static int __init flow_cache_init(void)
flow_lwm = 2 * flow_hash_size; flow_lwm = 2 * flow_hash_size;
flow_hwm = 4 * flow_hash_size; flow_hwm = 4 * flow_hash_size;
for (i = 0; i < NR_CPUS; i++)
flow_hash_rnd_recalc(i) = 1;
init_timer(&flow_hash_rnd_timer); init_timer(&flow_hash_rnd_timer);
flow_hash_rnd_timer.function = flow_cache_new_hashrnd; flow_hash_rnd_timer.function = flow_cache_new_hashrnd;
flow_hash_rnd_timer.expires = jiffies + FLOW_HASH_RND_PERIOD; flow_hash_rnd_timer.expires = jiffies + FLOW_HASH_RND_PERIOD;
add_timer(&flow_hash_rnd_timer); add_timer(&flow_hash_rnd_timer);
for (order = 0; flow_cache_cpu_online(smp_processor_id());
(PAGE_SIZE << order) < register_cpu_notifier(&flow_cache_cpu_nb);
(NR_CPUS*sizeof(struct flow_entry *)*flow_hash_size);
order++)
/* NOTHING */;
flow_table = (struct flow_cache_entry **)
__get_free_pages(GFP_ATOMIC, order);
if (!flow_table)
panic("Failed to allocate flow cache hash table\n");
memset(flow_table, 0, PAGE_SIZE << order);
return 0; return 0;
} }
......
Markdown is supported
0%
or
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment