linux-kernel - Re: [PATCH] x86: suppress sparse warning in copy_to

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: <1475570946.5324.37.camel@sipsolutions.net>
Date:   Tue, 04 Oct 2016 10:49:06 +0200
From:   Johannes Berg <johannes@...solutions.net>
To:     Jan Beulich <JBeulich@...e.com>
Cc:     Ingo Molnar <mingo@...nel.org>, x86@...nel.org,
        Thomas Gleixner <tglx@...utronix.de>,
        linux-kernel@...r.kernel.org, "H. Peter Anvin" <hpa@...or.com>
Subject: Re: [PATCH] x86: suppress sparse warning in copy_to_user()

> > > If that was the case, everyone should have seen such warnings
> > > from the day the original patch got introduced. 
> > 
> > Only if they run sparse. Clearly people don't, or we wouldn't have
> > a history of a ton of such problems, e.g.
> 
> No - you say "which gcc declares with (void *, int type) prototype".
> If that was the case, there would need to be a warning.

There would need to be a warning when?

> > > And the compiler warnings
> > > I get when testing with all four combinations of const and
> > > volatile
> > > also supports this by saying "expected 'const void *' but ..." 
> > 
> > It's not a compiler warning though that I'm getting.
> > 
> > What tool are you using to get such a warning?
> 
> I'm talking about gcc and the warning surfacing when I additonally
> add volatile.

Oh, sorry. If you get the warning, it prints "expected 'const void *'"
... yeah, I see.

> > https://gcc.gnu.org/onlinedocs/gcc/Object-Size-Checking.html 
> 
> Perhaps it's just the documentation which is imprecise here?

Hmm, yeah, that could be right, or maybe it changed at some point?

If it were defined the way the documentation says, you should have
gotten a compiler warning ("passing argument 1 ... discards ‘const’
qualifier from pointer target type") with the code as it is (without my
patch), since you can't pass a const pointer to a function that expects
a non-const pointer. Clearly that didn't happen.

That does indicate that the prototype is indeed with the const, I guess
I'll go fix sparse instead.

Sorry I misread your earlier explanation entirely!

johannes