linux-kernel - Re: [PATCH v6 2/4] leds: Add driver for Qualcomm LPG

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]

Date:   Thu, 29 Apr 2021 15:54:36 -0500
From:   Bjorn Andersson <bjorn.andersson@...aro.org>
To:     Marijn Suijten <marijn.suijten@...ainline.org>
Cc:     Pavel Machek <pavel@....cz>, Dan Murphy <dmurphy@...com>,
        Rob Herring <robh+dt@...nel.org>,
        Andy Gross <agross@...nel.org>,
        Thierry Reding <thierry.reding@...il.com>,
        Uwe Kleine-K?nig <u.kleine-koenig@...gutronix.de>,
        Lee Jones <lee.jones@...aro.org>,
        Martin Botka <martin.botka1@...il.com>,
        linux-leds@...r.kernel.org, devicetree@...r.kernel.org,
        linux-kernel@...r.kernel.org, linux-arm-msm@...r.kernel.org,
        linux-pwm@...r.kernel.org
Subject: Re: [PATCH v6 2/4] leds: Add driver for Qualcomm LPG

On Thu 29 Apr 14:31 CDT 2021, Marijn Suijten wrote:

> On 4/29/21 12:39 AM, Bjorn Andersson wrote:
> > On Sun 18 Apr 16:54 CDT 2021, Marijn Suijten wrote:
[..]
> > > > +	ret = lpg_init_lut(lpg);
> > > > +	if (ret < 0)
> > > > +		return ret;
> > > 
> > > 
> > > How about turning these returns into dev_err_probe?  I'm not sure if that's
> > > the expected way to go nowadays, but having some form of logging when a
> > > driver fails to probe is always good to have.
> > > 
> > 
> > The intention is that each code path through these functions will either
> > pass or spit out an error in the log. I looked through them again and
> > think I cover all paths...
> 
> 
> That is true, all the errors not covered are extremely unlikely like
> -ENOMEM.  I vaguely recall having to insert extra logging to get through
> initial probe, but that might have been something inside lpg_add_led as
> well.  Fine to leave this as it is.
> 

When kzalloc et al returns -ENOMEM it will be done with an error print,
so that does not need an additional print. That said, another pass
through lpg_add_led() made me spot that if you get a parse error on
the "color" property we would silently return -EINVAL. I've corrected
this.

Thanks,
Bjorn