linux-kernel - Re: [cocci] [PATCH -mm] -funsigned-char, x86: make struct p4_event

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: <alpine.DEB.2.22.394.2211030726351.2877@hadrien>
Date:   Thu, 3 Nov 2022 07:31:47 +0100 (CET)
From:   Julia Lawall <julia.lawall@...ia.fr>
To:     "Jason A. Donenfeld" <Jason@...c4.com>
cc:     Kees Cook <keescook@...omium.org>, cocci@...ia.fr,
        Linus Torvalds <torvalds@...ux-foundation.org>,
        Alexey Dobriyan <adobriyan@...il.com>,
        akpm@...ux-foundation.org, linux-kernel@...r.kernel.org,
        mm-commits@...r.kernel.org, masahiroy@...nel.org,
        gregkh@...uxfoundation.org, andriy.shevchenko@...ux.intel.com,
        Stephen Rothwell <sfr@...b.auug.org.au>
Subject: Re: [cocci] [PATCH -mm] -funsigned-char, x86: make struct
 p4_event_bind::cntr signed array



On Thu, 3 Nov 2022, Jason A. Donenfeld wrote:

> On Wed, Nov 02, 2022 at 06:17:04PM +0100, Julia Lawall wrote:
> >
> >
> > On Wed, 26 Oct 2022, Jason A. Donenfeld wrote:
> >
> > > On Wed, Oct 26, 2022 at 03:50:25AM +0200, Jason A. Donenfeld wrote:
> > > > The traditional objdump comparison does work, though. It produces a good
> > >
> > > Another thing that appears to work well is just using Coccinelle
> > > scripts. I've had some success just scrolling through the results of:
> > >
> > >     @@
> > >     char c;
> > >     expression E;
> > >     @@
> > >     (
> > >     * E > c
> > >     |
> > >     * E >= c
> > >     |
> > >     * E < c
> > >     |
> > >     * E <= c
> > >     )
> > >
> > > That also triggers on explicitly signed chars, and examining those
> > > reveals that quite a bit of code in the tree already does do the right
> > > thing, which is good.
> > >
> > > From looking at this and objdump output, it looks like most naked-char
> > > usage that isn't for strings is actually already assuming it's unsigned,
> > > using it as a byte. I'll continue to churn, and I'm sure I'll miss a few
> > > things here and there, but all and all, I don't think this is looking as
> > > terrible as I initially feared.
> > >
> > > I'm CC'ing the Coccinelle people to see if they have any nice ideas on
> > > improvements. Specifically, the thing we're trying to identify is:
> > >
> > >   - Usage of vanilla `char`, without a `signed` or `unsigned` qualifier,
> > >     where:
> >
> > Try putting
> >
> > disable optional_qualifier
> >
> > between the initial @@, to avoid the implicit matching of signed and
> > unsigned.
>
> Hmm, this doesn't quite work. Here are my rules:
>
>     @disable optional_qualifier@
>     char c;
>     expression E;
>     @@
>     (
>     * E > c
>     |
>     * E >= c
>     |
>     * E < c
>     |
>     * E <= c
>     )
>
>     @disable optional_qualifier@
>     char c;
>     @@
>     * c == -1
>
>     @disable optional_qualifier@
>     char c;
>     @@
>     * c = -1
>
> This produces, for example:
>
> diff -u -p ./sound/firewire/bebob/bebob_focusrite.c /tmp/nothing/sound/firewire/bebob/bebob_focusrite.c
> --- ./sound/firewire/bebob/bebob_focusrite.c
> +++ /tmp/nothing/sound/firewire/bebob/bebob_focusrite.c
> @@ -192,7 +192,6 @@ saffirepro_both_clk_src_get(struct snd_b
>
>         /* In a case that this driver cannot handle the value of register. */
>         value &= SAFFIREPRO_CLOCK_SOURCE_SELECT_MASK;
> -       if (value >= SAFFIREPRO_CLOCK_SOURCE_COUNT || map[value] < 0) {
>                 err = -EIO;
>                 goto end;
>         }
>
> Except map is defined as:
>
>     const signed char *map;
>
> So this would be one of those cases that I had hoped `disable
> optional_qualifier` would exclude. (I think internally coccinelle might
> be assuming `char` is signed, by the way.)

OK, I see the problem.  Coccinelle isn't taking the "disable
optional_qualifier" into account when it checks types on expressions.  It
would work if you put, eg:

char x;
... when any
* x < 0

But that would be much slower and less general.  I will fix it.

>
> > >   - It's not being used for characters; and
> > >   - It's doing something that assumes it is signed, such as various
> > >     types of comparisons or decrements.
> >
> > I took a quick look at the article, but I'm not completely sure what you
> > are getting at here.  Could you give some examples of what you do and
> > don't want to find?
> >
> > You don't want the case where c is 'x', for some x?
>
> Something I would want to find is `if (c < 0)`. Something I wouldn't
> want to find is `if (c < '9')`. IOW, I'm looking for code that assumes
> `c` is signed, and would become incorrect if `c` suddenly became
> unsigned. Most things involving actual characters are fine. But most
> things involving signed arithmetic or comparisons with numbers isn't
> find.

This seems to do what you want:

@disable optional_qualifier@
constant char cc;
expression e;
char c;
@@

(
  c < cc
|
* c < e
)

It highlights only the two return lines in:

int main () {
        char x;
        if (x < 'd')
                return x < 0;
        else return x < y;
}


julia

>
> Jason
>