Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::1:20 as permitted sender) client-ip=2620:137:e000::1:20;
Date:   Thu, 10 Mar 2022 08:27:24 +0000
Message-ID: <87sfrqyzyr.wl-maz@kernel.org>
From:   Marc Zyngier <maz@kernel.org>
To:     Guo Zhengkui <guozhengkui@vivo.com>
Cc:     Hector Martin <marcan@marcan.st>, Sven Peter <sven@svenpeter.dev>,
        Alyssa Rosenzweig <alyssa@rosenzweig.io>,
        Thomas Gleixner <tglx@linutronix.de>,
        linux-arm-kernel@lists.infradead.org (moderated list:ARM/APPLE MACHINE
        SUPPORT),
        linux-kernel@vger.kernel.org (open list:IRQCHIP DRIVERS),
        zhengkui_guo@outlook.com
Subject: Re: [PATCH] irqchip/apple-aic: application of sizeof() to a pointer
In-Reply-To: <20220310050238.4478-1-guozhengkui@vivo.com>
References: <20220310050238.4478-1-guozhengkui@vivo.com>
User-Agent: Wanderlust/2.15.9 (Almost Unreal) SEMI-EPG/1.14.7 (Harue)
 FLIM-LB/1.14.9 (=?UTF-8?B?R29qxY0=?=) APEL-LB/10.8 EasyPG/1.0.0 Emacs/27.1
 (x86_64-pc-linux-gnu) MULE/6.0 (HANACHIRUSATO)
MIME-Version: 1.0 (generated by SEMI-EPG 1.14.7 - "Harue")
Content-Type: text/plain; charset=US-ASCII
Precedence: bulk

On Thu, 10 Mar 2022 05:02:38 +0000,
Guo Zhengkui <guozhengkui@vivo.com> wrote:
> 
> `ic->fiq_aff[fiq]` is a pointer to the unnamed struct.
> `sizeof(ic->fiq_aff[fiq])` intends to get the size of this pointer. But
> readers may get confused. `sizeof(unsigned long)` makes more sense because
> `unsigned long` has the same size of pointer.

More sense? It really depends on who reads the code.

> 
> reference:
> https://lore.kernel.org/all/Ya%2FeUbdN1+ABFVWf@rowland.harvard.edu/
> https://lore.kernel.org/all/YbBGGI9wQenI4kP7@kroah.com/
> https://lore.kernel.org/all/20211209062441.9856-1-guozhengkui@vivo.com/
> 
> Signed-off-by: Guo Zhengkui <guozhengkui@vivo.com>
> ---
>  drivers/irqchip/irq-apple-aic.c | 2 +-
>  1 file changed, 1 insertion(+), 1 deletion(-)
> 
> diff --git a/drivers/irqchip/irq-apple-aic.c b/drivers/irqchip/irq-apple-aic.c
> index b40199c6625e..23098b469b1a 100644
> --- a/drivers/irqchip/irq-apple-aic.c
> +++ b/drivers/irqchip/irq-apple-aic.c
> @@ -810,7 +810,7 @@ static void build_fiq_affinity(struct aic_irq_chip *ic, struct device_node *aff)
>  	if (WARN_ON(n < 0))
>  		return;
>  
> -	ic->fiq_aff[fiq] = kzalloc(sizeof(ic->fiq_aff[fiq]), GFP_KERNEL);
> +	ic->fiq_aff[fiq] = kzalloc(sizeof(unsigned long), GFP_KERNEL);

And by doing that, you are making even more difficult to spot the
glaring bug that is in front of your eyes: we're not trying to
allocate a pointer, but to allocate the full data structure.

Nothing went wrong as a 64bit allocation is plenty when you need at
most 10 bits, but jeez, what a howler. I'm stashing the fixlet below
on top.

Thanks,

	M.

diff --git a/drivers/irqchip/irq-apple-aic.c b/drivers/irqchip/irq-apple-aic.c
index b40199c6625e..3f1d2f3ccb7f 100644
--- a/drivers/irqchip/irq-apple-aic.c
+++ b/drivers/irqchip/irq-apple-aic.c
@@ -810,7 +810,7 @@ static void build_fiq_affinity(struct aic_irq_chip *ic, struct device_node *aff)
 	if (WARN_ON(n < 0))
 		return;
 
-	ic->fiq_aff[fiq] = kzalloc(sizeof(ic->fiq_aff[fiq]), GFP_KERNEL);
+	ic->fiq_aff[fiq] = kzalloc(sizeof(*ic->fiq_aff[fiq]), GFP_KERNEL);
 	if (!ic->fiq_aff[fiq])
 		return;
 

-- 
Without deviation from the norm, progress is not possible.