Received-SPF: pass (google.com: domain of linux-kernel+bounces-143837-linux.lists.archive=gmail.com@vger.kernel.org designates 2604:1380:40f1:3f00::1 as permitted sender) client-ip=2604:1380:40f1:3f00::1;
Date: Sat, 13 Apr 2024 10:54:37 -0700
From: Yury Norov <yury.norov@gmail.com>
To: Dawei Li <dawei.li@shingroup.cn>
Cc: tglx@linutronix.de, akpm@linux-foundation.org, maz@kernel.org,
	florian.fainelli@broadcom.com, chenhuacai@kernel.org,
	jiaxun.yang@flygoat.com, anup@brainfault.org, palmer@dabbelt.com,
	samuel.holland@sifive.com, linux@rasmusvillemoes.dk,
	linux-kernel@vger.kernel.org
Subject: Re: [PATCH 6/6] irqchip/sifive-plic: Avoid explicit cpumask
 allocation on stack
Message-ID: <ZhrG3RLFVIKmhEbk@yury-ThinkPad>
References: <20240412105839.2896281-1-dawei.li@shingroup.cn>
 <20240412105839.2896281-7-dawei.li@shingroup.cn>
Precedence: bulk
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <20240412105839.2896281-7-dawei.li@shingroup.cn>

On Fri, Apr 12, 2024 at 06:58:39PM +0800, Dawei Li wrote:
> In general it's preferable to avoid placing cpumasks on the stack, as
> for large values of NR_CPUS these can consume significant amounts of
> stack space and make stack overflows more likely.
> 
> Use cpumask_first_and_and() to avoid the need for a temporary cpumask on
> the stack.
> 
> Signed-off-by: Dawei Li <dawei.li@shingroup.cn>
> ---
>  drivers/irqchip/irq-sifive-plic.c | 8 +++-----
>  1 file changed, 3 insertions(+), 5 deletions(-)
> 
> diff --git a/drivers/irqchip/irq-sifive-plic.c b/drivers/irqchip/irq-sifive-plic.c
> index f3d4cb9e34f7..bf5d2fc6396e 100644
> --- a/drivers/irqchip/irq-sifive-plic.c
> +++ b/drivers/irqchip/irq-sifive-plic.c
> @@ -164,15 +164,13 @@ static int plic_set_affinity(struct irq_data *d,
>  			     const struct cpumask *mask_val, bool force)
>  {
>  	unsigned int cpu;
> -	struct cpumask amask;
>  	struct plic_priv *priv = irq_data_get_irq_chip_data(d);
>  
> -	cpumask_and(&amask, &priv->lmask, mask_val);
> -
>  	if (force)
> -		cpu = cpumask_first(&amask);
> +		cpu = cpumask_first_and(&priv->lmask, mask_val);
>  	else
> -		cpu = cpumask_any_and(&amask, cpu_online_mask);
> +		cpu = cpumask_first_and_and(&priv->lmask, mask_val,
> +					    cpu_online_mask);

Don't need to split the line here. The new max length is 100 chars,
here's 85 when unsplit, and it hurts readability for nothing.

>  
>  	if (cpu >= nr_cpu_ids)
>  		return -EINVAL;
> -- 
> 2.27.0