linux-kernel - Re: [PATCH 2/2] clk: at91: main: do not continue if oscillators already prepared

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: <20200626210357.GX131826@piout.net>
Date:   Fri, 26 Jun 2020 23:03:57 +0200
From:   Alexandre Belloni <alexandre.belloni@...tlin.com>
To:     Claudiu Beznea <claudiu.beznea@...rochip.com>
Cc:     mturquette@...libre.com, sboyd@...nel.org,
        nicolas.ferre@...rochip.com, ludovic.desroches@...rochip.com,
        linux-kernel@...r.kernel.org, mturquette@...aro.org,
        bbrezillon@...nel.org, linux-clk@...r.kernel.org,
        linux-arm-kernel@...ts.infradead.org
Subject: Re: [PATCH 2/2] clk: at91: main: do not continue if oscillators
 already prepared

On 25/06/2020 13:09:28+0300, Claudiu Beznea wrote:
> Return in clk_main_osc_prepare()/clk_main_rc_osc_prepare() if
> oscillators are already enabled.
> 
> Fixes: 27cb1c2083373 ("clk: at91: rework main clk implementation")
> Fixes: 1bdf02326b71e ("clk: at91: make use of syscon/regmap internally")

Is this really a fix? What is the observed issue?

> Signed-off-by: Claudiu Beznea <claudiu.beznea@...rochip.com>
> ---
>  drivers/clk/at91/clk-main.c | 18 +++++++++---------
>  1 file changed, 9 insertions(+), 9 deletions(-)
> 
> diff --git a/drivers/clk/at91/clk-main.c b/drivers/clk/at91/clk-main.c
> index 37c22667e831..46b4d2131989 100644
> --- a/drivers/clk/at91/clk-main.c
> +++ b/drivers/clk/at91/clk-main.c
> @@ -74,13 +74,11 @@ static int clk_main_osc_prepare(struct clk_hw *hw)
>  	regmap_read(regmap, AT91_CKGR_MOR, &tmp);
>  	tmp &= ~MOR_KEY_MASK;
>  
> -	if (tmp & AT91_PMC_OSCBYPASS)
> +	if (tmp & (AT91_PMC_OSCBYPASS | AT91_PMC_MOSCEN))
>  		return 0;

While this seems like a good optimization, it is also not correct.
Having AT91_PMC_MOSCEN set doesn't mean that the clock is ready, you
need to at least check MOSCS once.

>  
> -	if (!(tmp & AT91_PMC_MOSCEN)) {
> -		tmp |= AT91_PMC_MOSCEN | AT91_PMC_KEY;
> -		regmap_write(regmap, AT91_CKGR_MOR, tmp);
> -	}
> +	tmp |= AT91_PMC_MOSCEN | AT91_PMC_KEY;
> +	regmap_write(regmap, AT91_CKGR_MOR, tmp);
>  
>  	while (!clk_main_osc_ready(regmap))
>  		cpu_relax();
> @@ -186,10 +184,12 @@ static int clk_main_rc_osc_prepare(struct clk_hw *hw)
>  
>  	regmap_read(regmap, AT91_CKGR_MOR, &mor);
>  
> -	if (!(mor & AT91_PMC_MOSCRCEN))
> -		regmap_update_bits(regmap, AT91_CKGR_MOR,
> -				   MOR_KEY_MASK | AT91_PMC_MOSCRCEN,
> -				   AT91_PMC_MOSCRCEN | AT91_PMC_KEY);
> +	if (mor & AT91_PMC_MOSCRCEN)
> +		return 0;
> +
> +	regmap_update_bits(regmap, AT91_CKGR_MOR,
> +			   MOR_KEY_MASK | AT91_PMC_MOSCRCEN,
> +			   AT91_PMC_MOSCRCEN | AT91_PMC_KEY);
>  
>  	while (!clk_main_rc_osc_ready(regmap))
>  		cpu_relax();
> -- 
> 2.7.4
> 

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