linux-kernel - Re: [PATCH] media: rzg2l-cru: Replace usleep

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: <aS_4t5q_foMuDyAl@tom-desktop>
Date: Wed, 3 Dec 2025 09:45:43 +0100
From: Tommaso Merciai <tommaso.merciai.xr@...renesas.com>
To: Laurent Pinchart <laurent.pinchart@...asonboard.com>
Cc: tomm.merciai@...il.com, linux-renesas-soc@...r.kernel.org,
	biju.das.jz@...renesas.com,
	Mauro Carvalho Chehab <mchehab@...nel.org>,
	Hans Verkuil <hverkuil+cisco@...nel.org>,
	Lad Prabhakar <prabhakar.mahadev-lad.rj@...renesas.com>,
	Daniel Scally <dan.scally+renesas@...asonboard.com>,
	Jacopo Mondi <jacopo.mondi@...asonboard.com>,
	linux-media@...r.kernel.org, linux-kernel@...r.kernel.org,
	Khai Nguyen <khai.nguyen.wx@...esas.com>,
	Hao Bui <hao.bui.yg@...esas.com>
Subject: Re: [PATCH] media: rzg2l-cru: Replace usleep_range with udelay

Hi Laurent,

Thanks for your review.

On Wed, Dec 03, 2025 at 11:35:52AM +0900, Laurent Pinchart wrote:
> Hi Tommaso,
> 
> Thank you for the patch.
> 
> On Tue, Dec 02, 2025 at 05:08:41PM +0100, Tommaso Merciai wrote:
> > `usleep_range()` should not be used in atomic contexts like between
> > spin_lock_irqsave()/spin_lock_irqrestore() pair inside function
> > rzg2l_cru_stop_image_processing(). That may cause scheduling while
> > atomic bug.
> > 
> > Signed-off-by: Khai Nguyen <khai.nguyen.wx@...esas.com>
> > Signed-off-by: Hao Bui <hao.bui.yg@...esas.com>
> > Signed-off-by: Tommaso Merciai <tommaso.merciai.xr@...renesas.com>
> > ---
> >  drivers/media/platform/renesas/rzg2l-cru/rzg2l-video.c | 4 ++--
> >  1 file changed, 2 insertions(+), 2 deletions(-)
> > 
> > diff --git a/drivers/media/platform/renesas/rzg2l-cru/rzg2l-video.c b/drivers/media/platform/renesas/rzg2l-cru/rzg2l-video.c
> > index 162e2ace6931..1355bfd186d4 100644
> > --- a/drivers/media/platform/renesas/rzg2l-cru/rzg2l-video.c
> > +++ b/drivers/media/platform/renesas/rzg2l-cru/rzg2l-video.c
> > @@ -369,7 +369,7 @@ void rzg2l_cru_stop_image_processing(struct rzg2l_cru_dev *cru)
> >  		if (cru->info->fifo_empty(cru))
> >  			break;
> >  
> > -		usleep_range(10, 20);
> > +		udelay(10);
> 
> There's an instance of msleep() earlier in this function, surrounded by
> spin_unlock_irqrestore() and spin_lock_irqsave(). I wondered if we
> should do the same here, but that lead to a second question: why does
> the driver need to cover the whole stop procedure with a spinlock in the
> first place ?

Good point :)
Mmm maybe the only critical section into the
rzg2l_cru_stop_image_processing() that needs
spin_unlock_irqrestore()/spin_lock_irqsave()
is:

spin_lock_irqsave(&cru->qlock, flags);
cru->state = RZG2L_CRU_DMA_STOPPED;
spin_unlock_irqrestore(&cru->qlock, flags);

Please correct me if I'm wrong.

Kind Regards,
Tommaso

> 
> >  	}
> >  
> >  	/* Notify that FIFO is not empty here */
> > @@ -385,7 +385,7 @@ void rzg2l_cru_stop_image_processing(struct rzg2l_cru_dev *cru)
> >  			AMnAXISTPACK_AXI_STOP_ACK)
> >  			break;
> >  
> > -		usleep_range(10, 20);
> > +		udelay(10);
> >  	}
> >  
> >  	/* Notify that AXI bus can not stop here */
> 
> -- 
> Regards,
> 
> Laurent Pinchart