linux-kernel - Re: [PATCH -next] staging: greybus: fix a missing-check bug in gb_lights_light

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, 2 Apr 2020 17:22:37 +0300
From:   Dan Carpenter <dan.carpenter@...cle.com>
To:     Rui Miguel Silva <rmfrfs@...il.com>
Cc:     devel@...verdev.osuosl.org, elder@...nel.org,
        Chen Zhou <chenzhou10@...wei.com>, gregkh@...uxfoundation.org,
        johan@...nel.org, linux-kernel@...r.kernel.org,
        greybus-dev@...ts.linaro.org
Subject: Re: [PATCH -next] staging: greybus: fix a missing-check bug in
 gb_lights_light_config()

On Thu, Apr 02, 2020 at 02:16:18PM +0100, Rui Miguel Silva wrote:
> > > --- a/drivers/staging/greybus/light.c
> > > +++ b/drivers/staging/greybus/light.c
> > > @@ -1026,7 +1026,8 @@ static int gb_lights_light_config(struct gb_lights *glights, u8 id)
> > >  
> > >  	light->channels_count = conf.channel_count;
> > >  	light->name = kstrndup(conf.name, NAMES_MAX, GFP_KERNEL);
> > > -
> > > +	if (!light->name)
> > > +		return -ENOMEM;
> > >  	light->channels = kcalloc(light->channels_count,
> > >  				  sizeof(struct gb_channel), GFP_KERNEL);
> > >  	if (!light->channels)
> > 
> > The clean up in this function is non-existant.  :(
> 
> Yeah, this have a central point to do the cleanups, gb_lights_release,
> since we may have other lights already configured at this point, we
> could cleanup this specific one here, but than would need to make sure
> all other already configure got clean also.

Central clean up functions never work correctly.

For example, we allocate "cdev->name" in gb_lights_channel_config()
before we register the channel later in gb_lights_register_all(glights);.
Now imagine that the register fails.  Then when we're freeing it in
__gb_lights_led_unregister() we see that the ->is_registered is false
so we don't kfree(cdev->name).

That's just a small memory leak.  But there are going to be tons of
little bugs like that.

Anyway it doesn't affect this patch so it's fine.

regards,
dan carpenter