linux-kernel - Re: [PATCH] overflow.h: Rename __ab_c_size() to __calc

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:   Fri, 5 Apr 2019 09:25:27 -0700
From:   Matthew Wilcox <willy@...radead.org>
To:     Kees Cook <keescook@...omium.org>
Cc:     Borislav Petkov <bp@...en8.de>,
        Rasmus Villemoes <linux@...musvillemoes.dk>,
        LKML <linux-kernel@...r.kernel.org>
Subject: Re: [PATCH] overflow.h: Rename __ab_c_size() to __calc_size()

On Fri, Apr 05, 2019 at 09:13:24AM -0700, Kees Cook wrote:
> On Fri, Apr 5, 2019 at 1:24 AM Borislav Petkov <bp@...en8.de> wrote:
> >
> > On Fri, Apr 05, 2019 at 10:09:31AM +0200, Rasmus Villemoes wrote:
> > > On 05/04/2019 09.52, Borislav Petkov wrote:
> > > > On Fri, Apr 05, 2019 at 08:26:45AM +0200, Rasmus Villemoes wrote:
> > >
> > > >> It computes a*b+c with overflow checking at each step. calc_size
> > > >> is way too generic and doesn't say anything at all about how the
> > > >> calc(ulation) is done.
> > > >
> > > > Ok, whatever.
> > > >
> > > > Then it would need at least a comment above it to state what it does. I
> > > > don't want to go and parse the macros each time.
> > >
> > > It's an internal helper, and struct_size is fully kernel-doc'ed. But
> > > yeah, a comment wouldn't hurt, and let's rename the parameters so they
> > > match the abc naming.
> > >
> > > diff --git a/include/linux/overflow.h b/include/linux/overflow.h
> > > index 40b48e2133cb..6534a727cadb 100644
> > > --- a/include/linux/overflow.h
> > > +++ b/include/linux/overflow.h
> > > @@ -278,11 +278,15 @@ static inline __must_check size_t
> > > array3_size(size_t a, size_t b, size_t c)
> > >         return bytes;
> > >  }
> > >
> > > -static inline __must_check size_t __ab_c_size(size_t n, size_t size,
> > > size_t c)
> > > +/*
> > > + * Compute a*b+c, returning SIZE_MAX on overflow. Internal helper for
> > > + * struct_size() below.
> > > + */
> 
> May as well make this kern-doc too?

I don't think that's a good idea; we should only document functions we
want other people to use.

> > > +static inline __must_check size_t __ab_c_size(size_t a, size_t b, size_t c)

I could also go for renaming this to __size_ab_plus_c.

For a bit of history ... the origins of this was a system which had
alloc_ab (a * b)
alloc_abc (a * b * c)
alloc_ab_c (a * b + c)
alloc_ab_cd (a * b + c * d)

In the process of getting it upstreamed, it changed from kmalloc_ab_c to
kmalloc(struct_size(...)) and we never bothered to change the name of
__ab_c_size().