linux-kernel - Re: [PATCH v3] pwm: atmel: Enable clk when pwm already enabled in bootloader

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] [day] [month] [year] [list]

Date:   Wed, 12 Jul 2023 14:47:53 +0200
From:   Uwe Kleine-König <u.kleine-koenig@...gutronix.de>
To:     Thierry Reding <thierry.reding@...il.com>
Cc:     Guiting Shen <aarongt.shen@...il.com>,
        claudiu.beznea@...rochip.com, nicolas.ferre@...rochip.com,
        alexandre.belloni@...tlin.com,
        linux-arm-kernel@...ts.infradead.org, linux-pwm@...r.kernel.org,
        linux-kernel@...r.kernel.org
Subject: Re: [PATCH v3] pwm: atmel: Enable clk when pwm already enabled in
 bootloader

On Wed, Jul 12, 2023 at 02:31:43PM +0200, Thierry Reding wrote:
> On Wed, Jul 12, 2023 at 09:45:08AM +0800, Guiting Shen wrote:
> > On Wed, Jul 12, 2023 at 04:30:17AM GMT+8, Uwe Kleine-König wrote:
> > > Hello,
> > > 
> > > On Wed, Jul 12, 2023 at 04:09:05AM +0800, Guiting Shen wrote:
> > >> +static int atmel_pwm_enable_clk_if_on(struct atmel_pwm_chip *atmel_pwm)
> > >> +{
> > >> +	unsigned int i;
> > >> +	int err;
> > >> +	u32 sr;
> > >> +
> > >> +	sr = atmel_pwm_readl(atmel_pwm, PWM_SR);
> > >> +	if (!sr)
> > >> +		return 0;
> > >> +
> > >> +	for (i = 0; i < atmel_pwm->chip.npwm; i++) {
> > >> +		if (!(sr & (1 << i)))
> > >> +			continue;
> > >> +
> > >> +		err = clk_enable(atmel_pwm->clk);
> > >> +		if (err) {
> > >> +			dev_err(atmel_pwm->chip.dev,
> > >> +				"failed to enable clock: %pe\n", ERR_PTR(err));
> > > 
> > > Here you leak possibly a few enables. While it's not likely that the
> > > (say) third enable goes wrong, it's also not that hard to handle?!
> > 
> > The driver used the enable_count member of struct clk_core to count the
> > PWM channels(4 channels). It will enable hardware clock only when one of
> > the PWM channels becomed on from all PWM channels off which maybe return
> > error. And in second/third/fourth times to clk_enable(), it just
> > increased the enable_count of struct clk_core which would never return
> > error.
> > 
> > It maybe confused at first time to view the code.
> > Do it need to add something like that: ?
> > 
> > for (i = 0; i < atmel_pwm->chip.npwm; i++) {
> > 	if (!(sr & (1 << i)))
> > 		continue;
> > 
> > 	err = clk_enable(atmel_pwm->clk);
> > 	if (err) {
> > 		dev_err(atmel_pwm->chip.dev,
> > 			"failed to enable clock: %pe\n", ERR_PTR(err));
> > 
> > 		for (i = 0; i < cnt; i++)
> > 			clk_disable(atmel_pwm->clk);
> > 		return err;
> > 	}
> > 	cnt++;
> 
> You can also achieve this by decrementing i back to zero, that way you
> avoid the additional variable and you get a more natural unwinding of
> what you did before.
> 
> So something like:
> 
> 	while (i--)
> 		clk_disable(atmel_pwm->clk);

You'd need something like:

 	while (i--) {
		if (!(sr & (1 << i)))
			continue;

 		clk_disable(atmel_pwm->clk);
	}

Best regards
Uwe


-- 
Pengutronix e.K.                           | Uwe Kleine-König            |
Industrial Linux Solutions                 | https://www.pengutronix.de/ |

Download attachment "signature.asc" of type "application/pgp-signature" (489 bytes)