* [PATCH 0/3] softirq: possible speedup and neatenings
@ 2013-11-17 9:55 Joe Perches
2013-11-17 9:55 ` [PATCH 1/3] softirq: Use ffs in __do_softirq Joe Perches
` (3 more replies)
0 siblings, 4 replies; 10+ messages in thread
From: Joe Perches @ 2013-11-17 9:55 UTC (permalink / raw)
To: Frederic Weisbecker; +Cc: Andrew Morton, linux-kernel
The first one boots, but it's barely tested.
Joe Perches (3):
softirq: Use ffs in __do_softirq
softirq: Convert printks to pr_<level>
softirq: Use const char * const for softirq_to_name, whitespace neatening
include/linux/interrupt.h | 2 +-
kernel/softirq.c | 75 +++++++++++++++++++++--------------------------
2 files changed, 35 insertions(+), 42 deletions(-)
--
1.8.1.2.459.gbcd45b4.dirty
^ permalink raw reply [flat|nested] 10+ messages in thread* [PATCH 1/3] softirq: Use ffs in __do_softirq 2013-11-17 9:55 [PATCH 0/3] softirq: possible speedup and neatenings Joe Perches @ 2013-11-17 9:55 ` Joe Perches 2013-12-09 18:44 ` Frederic Weisbecker 2013-11-17 9:55 ` [PATCH 2/3] softirq: Convert printks to pr_<level> Joe Perches ` (2 subsequent siblings) 3 siblings, 1 reply; 10+ messages in thread From: Joe Perches @ 2013-11-17 9:55 UTC (permalink / raw) To: Frederic Weisbecker; +Cc: Andrew Morton, linux-kernel Possible speed improvement of the __do_softirq function by using ffs instead of using a while loop with an & 1 test then single bit shift. Signed-off-by: Joe Perches <joe@perches.com> --- kernel/softirq.c | 43 ++++++++++++++++++++++--------------------- 1 file changed, 22 insertions(+), 21 deletions(-) diff --git a/kernel/softirq.c b/kernel/softirq.c index 11025cc..7be95d7 100644 --- a/kernel/softirq.c +++ b/kernel/softirq.c @@ -219,6 +219,7 @@ asmlinkage void __do_softirq(void) int cpu; unsigned long old_flags = current->flags; int max_restart = MAX_SOFTIRQ_RESTART; + int softirq_bit; /* * Mask out PF_MEMALLOC s current task context is borrowed for the @@ -242,30 +243,30 @@ restart: h = softirq_vec; - do { - if (pending & 1) { - unsigned int vec_nr = h - softirq_vec; - int prev_count = preempt_count(); - - kstat_incr_softirqs_this_cpu(vec_nr); - - trace_softirq_entry(vec_nr); - h->action(h); - trace_softirq_exit(vec_nr); - if (unlikely(prev_count != preempt_count())) { - printk(KERN_ERR "huh, entered softirq %u %s %p" - "with preempt_count %08x," - " exited with %08x?\n", vec_nr, - softirq_to_name[vec_nr], h->action, - prev_count, preempt_count()); - preempt_count_set(prev_count); - } + while ((softirq_bit = ffs(pending))) { + unsigned int vec_nr; + int prev_count; + + h += softirq_bit - 1; + + vec_nr = h - softirq_vec; + prev_count = preempt_count(); - rcu_bh_qs(cpu); + kstat_incr_softirqs_this_cpu(vec_nr); + + trace_softirq_entry(vec_nr); + h->action(h); + trace_softirq_exit(vec_nr); + if (unlikely(prev_count != preempt_count())) { + printk(KERN_ERR "huh, entered softirq %u %s %p with preempt_count %08x, exited with %08x?\n", + vec_nr, softirq_to_name[vec_nr], h->action, + prev_count, preempt_count()); + preempt_count_set(prev_count); } + rcu_bh_qs(cpu); h++; - pending >>= 1; - } while (pending); + pending >>= softirq_bit; + } local_irq_disable(); -- 1.8.1.2.459.gbcd45b4.dirty ^ permalink raw reply related [flat|nested] 10+ messages in thread
* Re: [PATCH 1/3] softirq: Use ffs in __do_softirq 2013-11-17 9:55 ` [PATCH 1/3] softirq: Use ffs in __do_softirq Joe Perches @ 2013-12-09 18:44 ` Frederic Weisbecker 2013-12-09 18:56 ` Joe Perches 0 siblings, 1 reply; 10+ messages in thread From: Frederic Weisbecker @ 2013-12-09 18:44 UTC (permalink / raw) To: Joe Perches; +Cc: Andrew Morton, linux-kernel On Sun, Nov 17, 2013 at 01:55:10AM -0800, Joe Perches wrote: > Possible speed improvement of the __do_softirq function by using ffs > instead of using a while loop with an & 1 test then single bit shift. > > Signed-off-by: Joe Perches <joe@perches.com> > --- > kernel/softirq.c | 43 ++++++++++++++++++++++--------------------- > 1 file changed, 22 insertions(+), 21 deletions(-) > > diff --git a/kernel/softirq.c b/kernel/softirq.c > index 11025cc..7be95d7 100644 > --- a/kernel/softirq.c > +++ b/kernel/softirq.c > @@ -219,6 +219,7 @@ asmlinkage void __do_softirq(void) > int cpu; > unsigned long old_flags = current->flags; > int max_restart = MAX_SOFTIRQ_RESTART; > + int softirq_bit; > > /* > * Mask out PF_MEMALLOC s current task context is borrowed for the > @@ -242,30 +243,30 @@ restart: > > h = softirq_vec; > > - do { > - if (pending & 1) { > - unsigned int vec_nr = h - softirq_vec; > - int prev_count = preempt_count(); > - > - kstat_incr_softirqs_this_cpu(vec_nr); > - > - trace_softirq_entry(vec_nr); > - h->action(h); > - trace_softirq_exit(vec_nr); > - if (unlikely(prev_count != preempt_count())) { > - printk(KERN_ERR "huh, entered softirq %u %s %p" > - "with preempt_count %08x," > - " exited with %08x?\n", vec_nr, > - softirq_to_name[vec_nr], h->action, > - prev_count, preempt_count()); > - preempt_count_set(prev_count); > - } > + while ((softirq_bit = ffs(pending))) { > + unsigned int vec_nr; > + int prev_count; > + > + h += softirq_bit - 1; Perhaps using for_each_set_bit() would simplify that more? > + > + vec_nr = h - softirq_vec; > + prev_count = preempt_count(); > > - rcu_bh_qs(cpu); > + kstat_incr_softirqs_this_cpu(vec_nr); > + > + trace_softirq_entry(vec_nr); > + h->action(h); > + trace_softirq_exit(vec_nr); > + if (unlikely(prev_count != preempt_count())) { > + printk(KERN_ERR "huh, entered softirq %u %s %p with preempt_count %08x, exited with %08x?\n", > + vec_nr, softirq_to_name[vec_nr], h->action, > + prev_count, preempt_count()); > + preempt_count_set(prev_count); > } > + rcu_bh_qs(cpu); > h++; > - pending >>= 1; > - } while (pending); > + pending >>= softirq_bit; > + } > > local_irq_disable(); > > -- > 1.8.1.2.459.gbcd45b4.dirty > ^ permalink raw reply [flat|nested] 10+ messages in thread
* Re: [PATCH 1/3] softirq: Use ffs in __do_softirq 2013-12-09 18:44 ` Frederic Weisbecker @ 2013-12-09 18:56 ` Joe Perches 2013-12-09 19:13 ` Frederic Weisbecker 0 siblings, 1 reply; 10+ messages in thread From: Joe Perches @ 2013-12-09 18:56 UTC (permalink / raw) To: Frederic Weisbecker; +Cc: Andrew Morton, linux-kernel On Mon, 2013-12-09 at 19:44 +0100, Frederic Weisbecker wrote: > On Sun, Nov 17, 2013 at 01:55:10AM -0800, Joe Perches wrote: > > Possible speed improvement of the __do_softirq function by using ffs > > instead of using a while loop with an & 1 test then single bit shift. [] > Perhaps using for_each_set_bit() would simplify that more? It might simplify the appearance of the code but it would/could expand the amount of generated code because for_each_set_bit uses an address_of(unsigned long) and the value tested is an unsigned int. extra dereferences, can't be in a register, etc... ^ permalink raw reply [flat|nested] 10+ messages in thread
* Re: [PATCH 1/3] softirq: Use ffs in __do_softirq 2013-12-09 18:56 ` Joe Perches @ 2013-12-09 19:13 ` Frederic Weisbecker 0 siblings, 0 replies; 10+ messages in thread From: Frederic Weisbecker @ 2013-12-09 19:13 UTC (permalink / raw) To: Joe Perches; +Cc: Andrew Morton, linux-kernel On Mon, Dec 09, 2013 at 10:56:46AM -0800, Joe Perches wrote: > On Mon, 2013-12-09 at 19:44 +0100, Frederic Weisbecker wrote: > > On Sun, Nov 17, 2013 at 01:55:10AM -0800, Joe Perches wrote: > > > Possible speed improvement of the __do_softirq function by using ffs > > > instead of using a while loop with an & 1 test then single bit shift. > [] > > Perhaps using for_each_set_bit() would simplify that more? > > It might simplify the appearance of the code but it > would/could expand the amount of generated code because > for_each_set_bit uses an address_of(unsigned long) and > the value tested is an unsigned int. > > extra dereferences, can't be in a register, etc... I'm not sure that would matter that much. But yeah it appears that find_first_bit/find_next_bit aren't even overriden in x86. So they are function calls. Although I guess that most of the time only one softirq is pending at a time. But anyway perhaps we want that path to stay very optimized, so you're patch look ok. Thanks. ^ permalink raw reply [flat|nested] 10+ messages in thread
* [PATCH 2/3] softirq: Convert printks to pr_<level> 2013-11-17 9:55 [PATCH 0/3] softirq: possible speedup and neatenings Joe Perches 2013-11-17 9:55 ` [PATCH 1/3] softirq: Use ffs in __do_softirq Joe Perches @ 2013-11-17 9:55 ` Joe Perches 2013-11-17 9:55 ` [PATCH 3/3] softirq: Use const char * const for softirq_to_name, whitespace neatening Joe Perches 2013-12-09 17:22 ` [PATCH 0/3] softirq: possible speedup and neatenings Joe Perches 3 siblings, 0 replies; 10+ messages in thread From: Joe Perches @ 2013-11-17 9:55 UTC (permalink / raw) To: Frederic Weisbecker; +Cc: Andrew Morton, linux-kernel Use a more current logging style. Signed-off-by: Joe Perches <joe@perches.com> --- kernel/softirq.c | 6 ++++-- 1 file changed, 4 insertions(+), 2 deletions(-) diff --git a/kernel/softirq.c b/kernel/softirq.c index 7be95d7..49a44cb 100644 --- a/kernel/softirq.c +++ b/kernel/softirq.c @@ -8,6 +8,8 @@ * Rewritten. Old one was good in 2.2, but in 2.3 it was immoral. --ANK (990903) */ +#define pr_fmt(fmt) KBUILD_MODNAME ": " fmt + #include <linux/export.h> #include <linux/kernel_stat.h> #include <linux/interrupt.h> @@ -258,7 +260,7 @@ restart: h->action(h); trace_softirq_exit(vec_nr); if (unlikely(prev_count != preempt_count())) { - printk(KERN_ERR "huh, entered softirq %u %s %p with preempt_count %08x, exited with %08x?\n", + pr_err("huh, entered softirq %u %s %p with preempt_count %08x, exited with %08x?\n", vec_nr, softirq_to_name[vec_nr], h->action, prev_count, preempt_count()); preempt_count_set(prev_count); @@ -562,7 +564,7 @@ EXPORT_SYMBOL(tasklet_init); void tasklet_kill(struct tasklet_struct *t) { if (in_interrupt()) - printk("Attempt to kill tasklet from interrupt\n"); + pr_notice("Attempt to kill tasklet from interrupt\n"); while (test_and_set_bit(TASKLET_STATE_SCHED, &t->state)) { do { -- 1.8.1.2.459.gbcd45b4.dirty ^ permalink raw reply related [flat|nested] 10+ messages in thread
* [PATCH 3/3] softirq: Use const char * const for softirq_to_name, whitespace neatening 2013-11-17 9:55 [PATCH 0/3] softirq: possible speedup and neatenings Joe Perches 2013-11-17 9:55 ` [PATCH 1/3] softirq: Use ffs in __do_softirq Joe Perches 2013-11-17 9:55 ` [PATCH 2/3] softirq: Convert printks to pr_<level> Joe Perches @ 2013-11-17 9:55 ` Joe Perches 2013-12-24 15:19 ` Wang YanQing 2013-12-09 17:22 ` [PATCH 0/3] softirq: possible speedup and neatenings Joe Perches 3 siblings, 1 reply; 10+ messages in thread From: Joe Perches @ 2013-11-17 9:55 UTC (permalink / raw) To: Frederic Weisbecker; +Cc: Andrew Morton, linux-kernel Reduce data size a little. Reduce checkpatch noise. $ size kernel/softirq.o* text data bss dec hex filename 11554 6013 4008 21575 5447 kernel/softirq.o.new 11474 6093 4008 21575 5447 kernel/softirq.o.old Signed-off-by: Joe Perches <joe@perches.com> --- include/linux/interrupt.h | 2 +- kernel/softirq.c | 28 +++++++++------------------- 2 files changed, 10 insertions(+), 20 deletions(-) diff --git a/include/linux/interrupt.h b/include/linux/interrupt.h index db43b58..0053add 100644 --- a/include/linux/interrupt.h +++ b/include/linux/interrupt.h @@ -360,7 +360,7 @@ enum /* map softirq index to softirq name. update 'softirq_to_name' in * kernel/softirq.c when adding a new softirq. */ -extern char *softirq_to_name[NR_SOFTIRQS]; +extern const char * const softirq_to_name[NR_SOFTIRQS]; /* softirq mask and active fields moved to irq_cpustat_t in * asm/hardirq.h to get better cache usage. KAO diff --git a/kernel/softirq.c b/kernel/softirq.c index 49a44cb..c828df5 100644 --- a/kernel/softirq.c +++ b/kernel/softirq.c @@ -56,7 +56,7 @@ static struct softirq_action softirq_vec[NR_SOFTIRQS] __cacheline_aligned_in_smp DEFINE_PER_CPU(struct task_struct *, ksoftirqd); -char *softirq_to_name[NR_SOFTIRQS] = { +const char * const softirq_to_name[NR_SOFTIRQS] = { "HI", "TIMER", "NET_TX", "NET_RX", "BLOCK", "BLOCK_IOPOLL", "TASKLET", "SCHED", "HRTIMER", "RCU" }; @@ -128,7 +128,6 @@ void local_bh_disable(void) { __local_bh_disable(_RET_IP_, SOFTIRQ_DISABLE_OFFSET); } - EXPORT_SYMBOL(local_bh_disable); static void __local_bh_enable(unsigned int cnt) @@ -150,7 +149,6 @@ void _local_bh_enable(void) WARN_ON_ONCE(in_irq()); __local_bh_enable(SOFTIRQ_DISABLE_OFFSET); } - EXPORT_SYMBOL(_local_bh_enable); static inline void _local_bh_enable_ip(unsigned long ip) @@ -167,7 +165,7 @@ static inline void _local_bh_enable_ip(unsigned long ip) /* * Keep preemption disabled until we are done with * softirq processing: - */ + */ preempt_count_sub(SOFTIRQ_DISABLE_OFFSET - 1); if (unlikely(!in_interrupt() && local_softirq_pending())) { @@ -289,8 +287,6 @@ restart: tsk_restore_flags(current, old_flags, PF_MEMALLOC); } - - asmlinkage void do_softirq(void) { __u32 pending; @@ -430,8 +426,7 @@ void open_softirq(int nr, void (*action)(struct softirq_action *)) /* * Tasklets */ -struct tasklet_head -{ +struct tasklet_head { struct tasklet_struct *head; struct tasklet_struct **tail; }; @@ -450,7 +445,6 @@ void __tasklet_schedule(struct tasklet_struct *t) raise_softirq_irqoff(TASKLET_SOFTIRQ); local_irq_restore(flags); } - EXPORT_SYMBOL(__tasklet_schedule); void __tasklet_hi_schedule(struct tasklet_struct *t) @@ -464,7 +458,6 @@ void __tasklet_hi_schedule(struct tasklet_struct *t) raise_softirq_irqoff(HI_SOFTIRQ); local_irq_restore(flags); } - EXPORT_SYMBOL(__tasklet_hi_schedule); void __tasklet_hi_schedule_first(struct tasklet_struct *t) @@ -475,7 +468,6 @@ void __tasklet_hi_schedule_first(struct tasklet_struct *t) __this_cpu_write(tasklet_hi_vec.head, t); __raise_softirq_irqoff(HI_SOFTIRQ); } - EXPORT_SYMBOL(__tasklet_hi_schedule_first); static void tasklet_action(struct softirq_action *a) @@ -495,7 +487,8 @@ static void tasklet_action(struct softirq_action *a) if (tasklet_trylock(t)) { if (!atomic_read(&t->count)) { - if (!test_and_clear_bit(TASKLET_STATE_SCHED, &t->state)) + if (!test_and_clear_bit(TASKLET_STATE_SCHED, + &t->state)) BUG(); t->func(t->data); tasklet_unlock(t); @@ -530,7 +523,8 @@ static void tasklet_hi_action(struct softirq_action *a) if (tasklet_trylock(t)) { if (!atomic_read(&t->count)) { - if (!test_and_clear_bit(TASKLET_STATE_SCHED, &t->state)) + if (!test_and_clear_bit(TASKLET_STATE_SCHED, + &t->state)) BUG(); t->func(t->data); tasklet_unlock(t); @@ -548,7 +542,6 @@ static void tasklet_hi_action(struct softirq_action *a) } } - void tasklet_init(struct tasklet_struct *t, void (*func)(unsigned long), unsigned long data) { @@ -558,7 +551,6 @@ void tasklet_init(struct tasklet_struct *t, t->func = func; t->data = data; } - EXPORT_SYMBOL(tasklet_init); void tasklet_kill(struct tasklet_struct *t) @@ -574,7 +566,6 @@ void tasklet_kill(struct tasklet_struct *t) tasklet_unlock_wait(t); clear_bit(TASKLET_STATE_SCHED, &t->state); } - EXPORT_SYMBOL(tasklet_kill); /* @@ -724,9 +715,8 @@ static void takeover_tasklets(unsigned int cpu) } #endif /* CONFIG_HOTPLUG_CPU */ -static int cpu_callback(struct notifier_block *nfb, - unsigned long action, - void *hcpu) +static int cpu_callback(struct notifier_block *nfb, unsigned long action, + void *hcpu) { switch (action) { #ifdef CONFIG_HOTPLUG_CPU -- 1.8.1.2.459.gbcd45b4.dirty ^ permalink raw reply related [flat|nested] 10+ messages in thread
* Re: [PATCH 3/3] softirq: Use const char * const for softirq_to_name, whitespace neatening 2013-11-17 9:55 ` [PATCH 3/3] softirq: Use const char * const for softirq_to_name, whitespace neatening Joe Perches @ 2013-12-24 15:19 ` Wang YanQing 2013-12-24 15:27 ` Joe Perches 0 siblings, 1 reply; 10+ messages in thread From: Wang YanQing @ 2013-12-24 15:19 UTC (permalink / raw) To: Joe Perches; +Cc: Frederic Weisbecker, Andrew Morton, linux-kernel On Sun, Nov 17, 2013 at 01:55:12AM -0800, Joe Perches wrote: > Reduce data size a little. > Reduce checkpatch noise. > > $ size kernel/softirq.o* > text data bss dec hex filename > 11554 6013 4008 21575 5447 kernel/softirq.o.new > 11474 6093 4008 21575 5447 kernel/softirq.o.old Hi, could you tell me why this patch could reduce data size? Thanks. ^ permalink raw reply [flat|nested] 10+ messages in thread
* Re: [PATCH 3/3] softirq: Use const char * const for softirq_to_name, whitespace neatening 2013-12-24 15:19 ` Wang YanQing @ 2013-12-24 15:27 ` Joe Perches 0 siblings, 0 replies; 10+ messages in thread From: Joe Perches @ 2013-12-24 15:27 UTC (permalink / raw) To: Wang YanQing; +Cc: Frederic Weisbecker, Andrew Morton, linux-kernel On Tue, 2013-12-24 at 23:19 +0800, Wang YanQing wrote: > On Sun, Nov 17, 2013 at 01:55:12AM -0800, Joe Perches wrote: > > Reduce data size a little. > > Reduce checkpatch noise. > > > > $ size kernel/softirq.o* > > text data bss dec hex filename > > 11554 6013 4008 21575 5447 kernel/softirq.o.new > > 11474 6093 4008 21575 5447 kernel/softirq.o.old > > Hi, could you tell me why this patch could reduce data size? It moves the softirq_to_name array of 10 char * from data (non-const) to text (const). -char *softirq_to_name[NR_SOFTIRQS] = { +const char * const softirq_to_name[NR_SOFTIRQS] = { "HI", "TIMER", "NET_TX", "NET_RX", "BLOCK", "BLOCK_IOPOLL", "TASKLET", "SCHED", "HRTIMER", "RCU" }; Use objdump or "make kernel/softirq.lst" and inspect the object code produced to see for yourself. cheers, Joe ^ permalink raw reply [flat|nested] 10+ messages in thread
* Re: [PATCH 0/3] softirq: possible speedup and neatenings 2013-11-17 9:55 [PATCH 0/3] softirq: possible speedup and neatenings Joe Perches ` (2 preceding siblings ...) 2013-11-17 9:55 ` [PATCH 3/3] softirq: Use const char * const for softirq_to_name, whitespace neatening Joe Perches @ 2013-12-09 17:22 ` Joe Perches 3 siblings, 0 replies; 10+ messages in thread From: Joe Perches @ 2013-12-09 17:22 UTC (permalink / raw) To: Frederic Weisbecker; +Cc: Andrew Morton, linux-kernel On Sun, 2013-11-17 at 01:55 -0800, Joe Perches wrote: > The first one boots, but it's barely tested. > > Joe Perches (3): > softirq: Use ffs in __do_softirq > softirq: Convert printks to pr_<level> > softirq: Use const char * const for softirq_to_name, whitespace neatening > > include/linux/interrupt.h | 2 +- > kernel/softirq.c | 75 +++++++++++++++++++++-------------------------- > 2 files changed, 35 insertions(+), 42 deletions(-) ping? ^ permalink raw reply [flat|nested] 10+ messages in thread
end of thread, other threads:[~2013-12-24 15:27 UTC | newest] Thread overview: 10+ messages (download: mbox.gz follow: Atom feed -- links below jump to the message on this page -- 2013-11-17 9:55 [PATCH 0/3] softirq: possible speedup and neatenings Joe Perches 2013-11-17 9:55 ` [PATCH 1/3] softirq: Use ffs in __do_softirq Joe Perches 2013-12-09 18:44 ` Frederic Weisbecker 2013-12-09 18:56 ` Joe Perches 2013-12-09 19:13 ` Frederic Weisbecker 2013-11-17 9:55 ` [PATCH 2/3] softirq: Convert printks to pr_<level> Joe Perches 2013-11-17 9:55 ` [PATCH 3/3] softirq: Use const char * const for softirq_to_name, whitespace neatening Joe Perches 2013-12-24 15:19 ` Wang YanQing 2013-12-24 15:27 ` Joe Perches 2013-12-09 17:22 ` [PATCH 0/3] softirq: possible speedup and neatenings Joe Perches
This is a public inbox, see mirroring instructions for how to clone and mirror all data and code used for this inbox