linux-kernel - Re: [PATCH 2/2] test_firmware: silence underflow warning in test_dev_config_update

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 for Android: free password hash cracker in your pocket

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <20190221191855.GB1737@kadam>
Date:   Thu, 21 Feb 2019 22:18:56 +0300
From:   Dan Carpenter <dan.carpenter@...cle.com>
To:     Andrew Morton <akpm@...ux-foundation.org>
Cc:     "Luis R. Rodriguez" <mcgrof@...nel.org>,
        Randy Dunlap <rdunlap@...radead.org>,
        linux-kernel@...r.kernel.org, kernel-janitors@...r.kernel.org,
        Greg Kroah-Hartman <gregkh@...uxfoundation.org>
Subject: Re: [PATCH 2/2] test_firmware: silence underflow warning in
 test_dev_config_update_u8()

On Thu, Feb 21, 2019 at 10:54:58AM -0800, Andrew Morton wrote:
> On Thu, 21 Feb 2019 21:38:26 +0300 Dan Carpenter <dan.carpenter@...cle.com> wrote:
> 
> > We put an upper bound on "new" but we don't check for negatives.
> 
> U8_MAX has unsigned type, so `if (new > U8_MAX)' does check for negative.
> 

No, doesn't work in this case.

#define U8_MAX          ((u8)~0U)

It would need to unsigned long for the type promotion to prevent
negatives, but it starts as unsigned int, then unsigned char, which is
type promoted to int.

It would be more clear to just write it as:

#define U8_MAX 0xff

> > In
> > this case the underflow doesn't matter very much, but we may as well
> > make the static checker happy.
> > 
> > ...
> >
> > --- a/lib/test_firmware.c
> > +++ b/lib/test_firmware.c
> > @@ -326,15 +326,12 @@ static ssize_t test_dev_config_show_int(char *buf, int cfg)
> >  static int test_dev_config_update_u8(const char *buf, size_t size, u8 *cfg)
> >  {
> >  	int ret;
> > -	long new;
> > +	u8 new;
> >  
> > -	ret = kstrtol(buf, 10, &new);
> > +	ret = kstrtou8(buf, 10, &new);
> >  	if (ret)
> >  		return ret;
> >  
> > -	if (new > U8_MAX)
> > -		return -EINVAL;
> > -
> >  	mutex_lock(&test_fw_mutex);
> >  	*(u8 *)cfg = new;
> >  	mutex_unlock(&test_fw_mutex);
> 
> if *buf=="257",
> 
> previous behavior: -EINVAL
> new behavior: *cfg = 1
> 
> yes?

No.  The kstrtou8() check the limit so it will return -ERANGE.  I should
have probably spelled that out better in the commit message.  I'll
resend tomorrow.

regard,
dan carpenter