linux-kernel - Re: dealing with gcc 'comparison is always false' warnings

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]

Message-ID: <a781481a0705301214h3295315esa31c40933ae4a539@mail.gmail.com>
Date:	Thu, 31 May 2007 00:44:30 +0530
From:	"Satyam Sharma" <satyam.sharma@...il.com>
To:	"Tilman Schmidt" <tilman@...p.cc>
Cc:	"Roland Dreier" <rdreier@...co.com>, openib-general@...nib.org,
	linux-kernel@...r.kernel.org
Subject: Re: dealing with gcc 'comparison is always false' warnings

Hi,

On 5/31/07, Tilman Schmidt <tilman@...p.cc> wrote:
> Am 30.05.2007 17:41 schrieb Satyam Sharma:
> > On 5/30/07, Roland Dreier <rdreier@...co.com> wrote:
> >> thanks... I'm wondering if there's a consensus among kernel hackers
> >> about changes like:
> >>
> >>  > -    if (hdr.cmd < 0 || hdr.cmd >= ARRAY_SIZE(ucma_cmd_table))
> >>  > +    if (hdr.cmd >= ARRAY_SIZE(ucma_cmd_table))
> >>  >              return -EINVAL;
> >>
> >> I understand that new gcc sees that hdr.cmd is unsigned and hence
> >> can't be < 0, and generates a warning for that, and having a build
> >> cluttered with warnings hides bugs and so on.  However the code here
> >> looks quite sensible to me -- otherwise we end up with missing range
> >> checking if hdr.cmd ever changes to a signed type.  This seems like a
> >> good way to introduce bugs: delete valid range checking code to shut
> >> up a silly gcc warning, and then change the type of a variable.
> >
> > You're *absolutely* correct about the issue that these "fixes" that remove
> > such conditions end up remove range-checking making the code more
> > flakey / less readable.
>
> I disagree. Changing the type of a variable is a significant
> modification. If someone does that, he or she *must* check every
> use of that variable, at which point he or she will also modify
> any range checks accordingly. Having checks that don't fit with
> the previous type *distracts* from that job. "Oh, did I modify
> that part already? Guess I can skip checking the rest of that
> function then." Oops.

I did not suggest the change-variable-type-from-unsigned-to-signed
thing as a "general" solution to such cases! ... in fact what I said
was that such cases do _not_ have a general solution at all, and
that shutting gcc up might not be a good idea, because a lot of
times such warnings do un-hide bugs. [ BTW when I gave the
change-type-from-unsigned-to-signed example, I had the size_t vs
ssize_t typo/bug in mind, for which changing the type is the proper
fix; and note that similar bugs can occur for non-size_t cases too. ]

> Nor is readability a suitable argument. Checking if hdr.cmd is
> less than zero gives the misleading impression that it *could*
> be less than zero, thus *impairing* readability.

Hmmm, but I tend to agree with the sentiment expressed in:
http://lkml.org/lkml/2006/11/28/206

Satyam
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/