linux-kernel - Re: [PATCH 15/46] selinux: One check and function call less in genfs

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, 23 Mar 2017 18:05:29 -0400
From:   Paul Moore <paul@...l-moore.com>
To:     Casey Schaufler <casey@...aufler-ca.com>
Cc:     SF Markus Elfring <elfring@...rs.sourceforge.net>,
        linux-security-module@...r.kernel.org, selinux@...ho.nsa.gov,
        Eric Paris <eparis@...isplace.org>,
        James Morris <james.l.morris@...cle.com>,
        "Serge E. Hallyn" <serge@...lyn.com>,
        Stephen Smalley <sds@...ho.nsa.gov>,
        William Roberts <william.c.roberts@...el.com>,
        LKML <linux-kernel@...r.kernel.org>,
        kernel-janitors@...r.kernel.org
Subject: Re: [PATCH 15/46] selinux: One check and function call less in
 genfs_read() after error detection

On Tue, Jan 17, 2017 at 12:53 PM, Casey Schaufler
<casey@...aufler-ca.com> wrote:
> On 1/17/2017 8:37 AM, SF Markus Elfring wrote:
>>>> @@ -2015,7 +2015,7 @@ static int genfs_read(struct policydb *p, void *fp)
>>>>             newgenfs = kzalloc(sizeof(*newgenfs), GFP_KERNEL);
>>>>             if (!newgenfs) {
>>>>                     rc = -ENOMEM;
>>>> -                   goto out;
>>>> +                   goto exit;
>>>>             }
>>>>
>>>>             rc = str_read(&newgenfs->fstype, GFP_KERNEL, fp, len);
>>>> @@ -2101,7 +2101,7 @@ static int genfs_read(struct policydb *p, void *fp)
>>>>             kfree(newgenfs);
>>>>     }
>>>>     ocontext_destroy(newc, OCON_FSUSE);
>>>> -
>>>> +exit:
>>>>     return rc;
>>> Why not replace the "goto out" with "return rc" rather
>>> than add a target?
>> Would you accept to use the statement "return -ENOMEM;" there instead?
>
> That would be even better.

I *hate* code that does a jump to a label only to then do a
return/exit.  That said, see my earlier comments about not worrying
too much about performance of the error path and in this case I like
the "feel good" nature of the code where ever failure in the loop goes
to "out".

-- 
paul moore
www.paul-moore.com