linux-kernel - Re: [PATCH] sched/core: Avoid selecting the task that is throttled to run when core-sched enable

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

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <20230320124701.GL2194297@hirez.programming.kicks-ass.net>
Date:   Mon, 20 Mar 2023 13:47:01 +0100
From:   Peter Zijlstra <peterz@...radead.org>
To:     Hao Jia <jiahao.os@...edance.com>
Cc:     mingo@...hat.com, mingo@...nel.org, juri.lelli@...hat.com,
        vincent.guittot@...aro.org, dietmar.eggemann@....com,
        rostedt@...dmis.org, bsegall@...gle.com, mgorman@...e.de,
        bristot@...hat.com, vschneid@...hat.com,
        mgorman@...hsingularity.net, linux-kernel@...r.kernel.org,
        Cruz Zhao <CruzZhao@...ux.alibaba.com>
Subject: Re: [PATCH] sched/core: Avoid selecting the task that is throttled
 to run when core-sched enable

On Thu, Mar 16, 2023 at 04:18:06PM +0800, Hao Jia wrote:

> diff --git a/kernel/sched/core.c b/kernel/sched/core.c
> index af017e038b48..27cb06e19b12 100644
> --- a/kernel/sched/core.c
> +++ b/kernel/sched/core.c
> @@ -261,36 +261,51 @@ void sched_core_dequeue(struct rq *rq, struct task_struct *p, int flags)
>  		resched_curr(rq);
>  }
>  
> +static int sched_task_is_throttled(struct task_struct *p, int cpu)
>  {
> +	if (p->sched_class->task_is_throttled)
> +		return p->sched_class->task_is_throttled(p, cpu);
>  
> +	return 0;
>  }
>  
>  static struct task_struct *sched_core_next(struct task_struct *p, unsigned long cookie)
>  {
>  	struct rb_node *node = &p->core_node;
> +	int cpu = task_cpu(p);
> +
> +	do {
> +		node = rb_next(node);
> +		if (!node)
> +			return NULL;
> +
> +		p = container_of(node, struct task_struct, core_node);

I've changed this to __node_2_sc() to match the rest. It looks to have
been randomly not using it.

> +		if (p->core_cookie != cookie)
> +			return NULL;
> +	} while (sched_task_is_throttled(p, cpu));
> +
> +	return p;
> +}
>  
> +/*
> + * Find left-most (aka, highest priority) and unthrottled task matching @cookie.
> + * If no suitable task is found, NULL will be returned.
> + */
> +static struct task_struct *sched_core_find(struct rq *rq, unsigned long cookie)
> +{
> +	struct task_struct *p;
> +	struct rb_node *node;
> +
> +	node = rb_find_first((void *)cookie, &rq->core_tree, rb_sched_core_cmp);
>  	if (!node)
>  		return NULL;
>  
> +	p = __node_2_sc(node);
> +	if (!sched_task_is_throttled(p, rq->cpu))
> +		return p;
> +
> +	return sched_core_next(p, cookie);
>  }
>  
>  /*

OK, fair enough.