mirror of https://lore.kernel.org/lkml/
 help / color / mirror / Atom feed
From: Thomas Gleixner <tglx@linutronix.de>
To: Oleg Nesterov <oleg@redhat.com>
Cc: LKML <linux-kernel@vger.kernel.org>,
	x86@kernel.org, "Eric W. Biederman" <ebiederm@xmission.com>,
	Frederic Weisbecker <frederic@kernel.org>,
	John Stultz <john.stultz@linaro.org>,
	Paolo Bonzini <pbonzini@redhat.com>
Subject: Re: [patch V2 3/5] posix-cpu-timers: Provide mechanisms to defer timer handling to task_work
Date: Fri, 17 Jul 2020 20:35:46 +0200	[thread overview]
Message-ID: <87zh7yq9a5.fsf@nanos.tec.linutronix.de> (raw)
In-Reply-To: <20200717172627.GC6067@redhat.com>

Oleg,

Oleg Nesterov <oleg@redhat.com> writes:
> Looks correct to me, but I forgot everything about posix-timers.c

that's not a problem because this is about posix-cpu-timers.c :)

> this obviously means that the expired timer won't fire until the
> task returns to user-mode but probably we don't care.

If the signal goes to the task itself it does not matter at all because
it's going to be delivered when the task goes out to user space.

If the signal goes to a supervisor process, then it will be slightly
delayed but I could not find a problem with that at all.

I'll add more reasoning to the changelog on V3.

>> +#ifdef CONFIG_POSIX_CPU_TIMERS_TASK_WORK
>> +void posix_cpu_timers_work(struct callback_head *work);
>> +
>> +static inline void posix_cputimer_init_work(struct posix_cputimers *pct)
>> +{
>> +	pct->task_work.func = posix_cpu_timers_work;
>
> init_task_work() ?

Yeah.

>> +}
>> +#else
>> +static inline void posix_cputimer_init_work(struct posix_cputimers *pct) { }
>> +#endif
>> +
>>  static inline void posix_cputimers_init(struct posix_cputimers *pct)
>>  {
>>  	memset(pct, 0, sizeof(*pct));
>>  	pct->bases[0].nextevt = U64_MAX;
>>  	pct->bases[1].nextevt = U64_MAX;
>>  	pct->bases[2].nextevt = U64_MAX;
>> +	posix_cputimer_init_work(pct);
>>  }
>
> And I can't resist. I know this is a common practice, please ignore, but to me
>
> 	static inline void posix_cputimers_init(struct posix_cputimers *pct)
> 	{
> 		memset(pct, 0, sizeof(*pct));
> 		pct->bases[0].nextevt = U64_MAX;
> 		pct->bases[1].nextevt = U64_MAX;
> 		pct->bases[2].nextevt = U64_MAX;
> 	#ifdef CONFIG_POSIX_CPU_TIMERS_TASK_WORK
> 		init_task_work(&pct->task_work, posix_cpu_timers_work);
> 	#endif
> 	}
>
> looks better than 2 posix_cputimer_init_work() definitions above.

Gah, I hate ifdefs in the middle of the code :)

> Note also that signal_struct->posix_cputimers.task_work is never used, perhaps
> it would be better to move this task_work into task_struct? This way we do not
> even need to change posix_cputimers_init(), we call simply initialize
> init_task.posix_task_work.

Let me look into that.

Thanks,

        tglx

  reply	other threads:[~2020-07-17 18:35 UTC|newest]

Thread overview: 21+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2020-07-16 20:19 [patch V2 0/5] posix-cpu-timers: Move expiry into task work context Thomas Gleixner
2020-07-16 20:19 ` [patch V2 1/5] posix-cpu-timers: Split run_posix_cpu_timers() Thomas Gleixner
2020-07-16 20:19 ` [patch V2 2/5] posix-cpu-timers: Convert the flags to a bitmap Thomas Gleixner
2020-07-21 12:34   ` Frederic Weisbecker
2020-07-21 16:10     ` Thomas Gleixner
2020-07-21 16:23       ` David Laight
2020-07-21 18:30         ` Thomas Gleixner
2020-07-16 20:19 ` [patch V2 3/5] posix-cpu-timers: Provide mechanisms to defer timer handling to task_work Thomas Gleixner
2020-07-16 22:50   ` Peter Zijlstra
2020-07-17 18:37     ` Thomas Gleixner
2020-07-23  1:03     ` Frederic Weisbecker
2020-07-23  8:32       ` Thomas Gleixner
2020-07-23 12:15         ` Frederic Weisbecker
2020-07-16 22:54   ` Peter Zijlstra
2020-07-17 18:38     ` Thomas Gleixner
2020-07-19 19:33       ` Thomas Gleixner
2020-07-21 18:50         ` Thomas Gleixner
2020-07-17 17:26   ` Oleg Nesterov
2020-07-17 18:35     ` Thomas Gleixner [this message]
2020-07-16 20:19 ` [patch V2 4/5] posix-cpu-timers: Expiry timers directly when in task work context Thomas Gleixner
2020-07-16 20:19 ` [patch V2 5/5] x86: Select POSIX_CPU_TIMERS_TASK_WORK Thomas Gleixner

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=87zh7yq9a5.fsf@nanos.tec.linutronix.de \
    --to=tglx@linutronix.de \
    --cc=ebiederm@xmission.com \
    --cc=frederic@kernel.org \
    --cc=john.stultz@linaro.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=oleg@redhat.com \
    --cc=pbonzini@redhat.com \
    --cc=x86@kernel.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox

all inboxes | Powered by JetHome®