linux-kernel - Re: [PATCH] rtc: snvs: make sure clock is enabled for interrupt handle

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: <20190111143141.GC2547@piout.net>
Date:   Fri, 11 Jan 2019 15:31:41 +0100
From:   Alexandre Belloni <alexandre.belloni@...tlin.com>
To:     Anson Huang <anson.huang@....com>
Cc:     "a.zummo@...ertech.it" <a.zummo@...ertech.it>,
        "linux-rtc@...r.kernel.org" <linux-rtc@...r.kernel.org>,
        "linux-kernel@...r.kernel.org" <linux-kernel@...r.kernel.org>,
        dl-linux-imx <linux-imx@....com>
Subject: Re: [PATCH] rtc: snvs: make sure clock is enabled for interrupt
 handle

Hi,

On 11/01/2019 07:09:02+0000, Anson Huang wrote:
> During system suspend, the SNVS RTC's clock will be disabled in
> noirq suspend phase, but SNVS RTC's alarm interrupt could still
> arrive, system will hang if SNVS RTC driver tries to access register
> without clock enabled, this patch fixes the issue of this scenario.
> 

Are you sure this is the real issue? I don't think the handler can be
called before the resume_noirq callback. Isn't the issue that your clock
driver has not yet resumed by the time you call the rtc driver
resume_noirq callback ?

> Signed-off-by: Anson Huang <Anson.Huang@....com>
> ---
>  drivers/rtc/rtc-snvs.c | 6 ++++++
>  1 file changed, 6 insertions(+)
> 
> diff --git a/drivers/rtc/rtc-snvs.c b/drivers/rtc/rtc-snvs.c
> index b2483a7..0b9eff1 100644
> --- a/drivers/rtc/rtc-snvs.c
> +++ b/drivers/rtc/rtc-snvs.c
> @@ -239,6 +239,9 @@ static irqreturn_t snvs_rtc_irq_handler(int irq, void *dev_id)
>  	u32 lpsr;
>  	u32 events = 0;
>  
> +	if (data->clk)
> +		clk_enable(data->clk);
> +

Anyway, won't that need a clk_prepare_enable because it has been
clk_disable_unprepare in suspend_noirq? And this is something you can
not do in atomic context

-- 
Alexandre Belloni, Bootlin
Embedded Linux and Kernel engineering
https://bootlin.com