linux-kernel - Re: your mail

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]

Message-ID: <20140709075636.GD30458@sirena.org.uk>
Date:	Wed, 9 Jul 2014 09:56:36 +0200
From:	Mark Brown <broonie@...nel.org>
To:	James Ban <james.ban.opensource@...semi.com>
Cc:	Liam Girdwood <lgirdwood@...il.com>,
	Support Opensource <support.opensource@...semi.com>,
	LKML <linux-kernel@...r.kernel.org>,
	David Dajun Chen <david.chen@...semi.com>
Subject: Re: your mail

On Wed, Jul 09, 2014 at 10:03:32AM +0900, James Ban wrote:

> > > +	ret = regmap_read(chip->regmap, DA9211_REG_EVENT_B, &reg_val);
> > > +	if (ret < 0)
> > > +		goto error_i2c;

> > > +	if (reg_val & DA9211_E_OV_CURR_A) {

> > > +	if (reg_val & DA9211_E_OV_CURR_B) {

> > > +	return IRQ_HANDLED;

> > This is buggy - the driver should only return IRQ_HANDLED if it handled the
> > interrupt somehow, otherwise it should return IRQ_NONE and let the interrupt
> > core handle things.  This is especially important since the device appears to
> > require that interrupts are explicitly acknoweldged so if something is flagged
> > but not handled the interrupt will just sit constantly asserted.

> Basically all interrupts are masked when the chip wakes up. 
> Only two interrupts are unmasked at the start of driver like below.

I know that's the intention but the code should still be written
robustly - something might go wrong somewhere which causes another
interrupt to be enabled, or we might even gain support for shared
threaded interrupts in the interrupt core and someone could then
try to use that in a system.

Download attachment "signature.asc" of type "application/pgp-signature" (837 bytes)