| lists.openwall.net | lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC | |
|
Open Source and information security mailing list archives
| ||
|
Message-ID: <51FB7B36.8030909@linux.vnet.ibm.com>
Date: Fri, 02 Aug 2013 14:56:14 +0530
From: Preeti U Murthy <preeti@...ux.vnet.ibm.com>
To: ±èÁؼö <iamjoonsoo.kim@....com>
CC: "'Ingo Molnar'" <mingo@...hat.com>,
"'Peter Zijlstra'" <peterz@...radead.org>,
linux-kernel@...r.kernel.org, "'Mike Galbraith'" <efault@....de>,
"'Paul Turner'" <pjt@...gle.com>,
"'Alex Shi'" <alex.shi@...el.com>,
"'Vincent Guittot'" <vincent.guittot@...aro.org>,
"'Morten Rasmussen'" <morten.rasmussen@....com>,
"'Namhyung Kim'" <namhyung@...nel.org>,
"'Joonsoo Kim'" <js1304@...il.com>
Subject: Re: [PATCH v2 2/3] sched: factor out code to should_we_balance()
On 08/02/2013 02:35 PM, ±èÁؼö wrote:
>
>
>> -----Original Message-----
>> From: Preeti U Murthy [mailto:preeti@...ux.vnet.ibm.com]
>> Sent: Friday, August 02, 2013 1:23 PM
>> To: Joonsoo Kim
>> Cc: Ingo Molnar; Peter Zijlstra; linux-kernel@...r.kernel.org; Mike
>> Galbraith; Paul Turner; Alex Shi; Vincent Guittot; Morten Rasmussen;
>> Namhyung Kim; Joonsoo Kim
>> Subject: Re: [PATCH v2 2/3] sched: factor out code to should_we_balance()
>>
>> Hi Joonsoo,
>
> Hello, Preeti.
>
>>
>> On 08/02/2013 07:20 AM, Joonsoo Kim wrote:
>>> Now checking whether this cpu is appropriate to balance or not is
>>> embedded into update_sg_lb_stats() and this checking has no direct
>>> relationship to this function. There is not enough reason to place
>>> this checking at update_sg_lb_stats(), except saving one iteration for
>>> sched_group_cpus.
>>>
>>> In this patch, I factor out this checking to should_we_balance()
>> function.
>>> And before doing actual work for load_balancing, check whether this
>>> cpu is appropriate to balance via should_we_balance(). If this cpu is
>>> not a candidate for balancing, it quit the work immediately.
>>>
>>> With this change, we can save two memset cost and can expect better
>>> compiler optimization.
>>>
>>> Below is result of this patch.
>>>
>>> * Vanilla *
>>> text data bss dec hex filename
>>> 34499 1136 116 35751 8ba7 kernel/sched/fair.o
>>>
>>> * Patched *
>>> text data bss dec hex filename
>>> 34243 1136 116 35495 8aa7 kernel/sched/fair.o
>>>
>>> In addition, rename @balance to @should_balance in order to represent
>>> its purpose more clearly.
>>>
>>> Signed-off-by: Joonsoo Kim <iamjoonsoo.kim@....com>
>>>
>>> diff --git a/kernel/sched/fair.c b/kernel/sched/fair.c index
>>> eaae77e..7f51b8c 100644
>>> --- a/kernel/sched/fair.c
>>> +++ b/kernel/sched/fair.c
>>> @@ -4426,22 +4426,17 @@ fix_small_capacity(struct sched_domain *sd,
>> struct sched_group *group)
>>> * @group: sched_group whose statistics are to be updated.
>>> * @load_idx: Load index of sched_domain of this_cpu for load calc.
>>> * @local_group: Does group contain this_cpu.
>>> - * @balance: Should we balance.
>>> * @sgs: variable to hold the statistics for this group.
>>> */
>>> static inline void update_sg_lb_stats(struct lb_env *env,
>>> struct sched_group *group, int load_idx,
>>> - int local_group, int *balance, struct sg_lb_stats
> *sgs)
>>> + int local_group, struct sg_lb_stats *sgs)
>>> {
>>> unsigned long nr_running, max_nr_running, min_nr_running;
>>> unsigned long load, max_cpu_load, min_cpu_load;
>>> - unsigned int balance_cpu = -1, first_idle_cpu = 0;
>>> unsigned long avg_load_per_task = 0;
>>> int i;
>>>
>>> - if (local_group)
>>> - balance_cpu = group_balance_cpu(group);
>>> -
>>> /* Tally up the load of all CPUs in the group */
>>> max_cpu_load = 0;
>>> min_cpu_load = ~0UL;
>>> @@ -4454,15 +4449,9 @@ static inline void update_sg_lb_stats(struct
>> lb_env *env,
>>> nr_running = rq->nr_running;
>>>
>>> /* Bias balancing toward cpus of our domain */
>>> - if (local_group) {
>>> - if (idle_cpu(i) && !first_idle_cpu &&
>>> - cpumask_test_cpu(i,
>> sched_group_mask(group))) {
>>> - first_idle_cpu = 1;
>>> - balance_cpu = i;
>>> - }
>>> -
>>> + if (local_group)
>>> load = target_load(i, load_idx);
>>> - } else {
>>> + else {
>>> load = source_load(i, load_idx);
>>> if (load > max_cpu_load)
>>> max_cpu_load = load;
>>> @@ -4482,22 +4471,9 @@ static inline void update_sg_lb_stats(struct
>> lb_env *env,
>>> sgs->idle_cpus++;
>>> }
>>>
>>> - /*
>>> - * First idle cpu or the first cpu(busiest) in this sched group
>>> - * is eligible for doing load balancing at this and above
>>> - * domains. In the newly idle case, we will allow all the cpu's
>>> - * to do the newly idle load balance.
>>> - */
>>> - if (local_group) {
>>> - if (env->idle != CPU_NEWLY_IDLE) {
>>> - if (balance_cpu != env->dst_cpu) {
>>> - *balance = 0;
>>> - return;
>>> - }
>>> - update_group_power(env->sd, env->dst_cpu);
>>> - } else if (time_after_eq(jiffies, group->sgp->next_update))
>>> - update_group_power(env->sd, env->dst_cpu);
>>> - }
>>
>> Observe what is happening in the above code which checks if we should
>> balance on the env->dst_cpu.
>>
>> Only if the env->dst_cpu "belongs" to the group considered in
>> update_sg_lb_stats(), which means local_group = 1,should the above checks
>> be carried out.
>>
>> Meaning, if there is a better CPU in the same group to which
>> env->dst_cpu belongs, to carry out load balancing for the system (in the
>> above case, balance_cpu), cancel the current iteration of load balancing
>> on env->dst_cpu. Wait for the right cpu in this group to do the load
>> balancing.
>>
>> Keeping this in mind see the below comments around should_we_balance().
>
> Okay.
>
>>
>>> @@ -5001,13 +4964,47 @@ static int need_active_balance(struct lb_env
>>> *env)
>>>
>>> static int active_load_balance_cpu_stop(void *data);
>>>
>>> +static int should_we_balance(struct lb_env *env) {
>>> + struct sched_group *sg = env->sd->groups;
>>> + struct cpumask *sg_cpus, *sg_mask;
>>> + int cpu, balance_cpu = -1;
>>> +
>>> + /*
>>> + * In the newly idle case, we will allow all the cpu's
>>> + * to do the newly idle load balance.
>>> + */
>>> + if (env->idle == CPU_NEWLY_IDLE)
>>> + return 1;
>>> +
>>> + sg_cpus = sched_group_cpus(sg);
>>> + sg_mask = sched_group_mask(sg);
>>> + /* Try to find first idle cpu */
>>> + for_each_cpu_and(cpu, sg_cpus, env->cpus) {
>>> + if (!cpumask_test_cpu(cpu, sg_mask) || idle_cpu(cpu))
>>> + continue;
>>> +
>>> + balance_cpu = cpu;
>>> + break;
>>> + }
>>
>> You need to iterate over all the groups of the sched domain env->sd and
>> not just the first group of env->sd like you are doing above. This is to
>
> I don't think so.
> IIRC, env->sd->groups always means local group,
> so we don't need to find our group by iterating over all the groups.
Take a look at update_sd_lb_stats(). That should clarify this. There is
an exclusive
local_group check there.
sd->groups points to the first group in the list of groups under this sd.
Regards
Preeti U Murthy
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
Powered by blists - more mailing lists