netdev - Re: [PATCH] be2net: fix adapter->big_page

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:   Mon, 22 Jul 2019 17:13:16 -0400
From:   Qian Cai <cai@....pw>
To:     David Miller <davem@...emloft.net>
Cc:     morbo@...gle.com, ndesaulniers@...gle.com, jyknight@...gle.com,
        sathya.perla@...adcom.com, ajit.khaparde@...adcom.com,
        sriharsha.basavapatna@...adcom.com, somnath.kotur@...adcom.com,
        arnd@...db.de, dhowells@...hat.com, hpa@...or.com,
        netdev@...r.kernel.org, linux-arch@...r.kernel.org,
        linux-kernel@...r.kernel.org, natechancellor@...il.com,
        Jakub Jelinek <jakub@...hat.com>
Subject: Re: [PATCH] be2net: fix adapter->big_page_size miscaculation

On Fri, 2019-07-19 at 17:47 -0400, Qian Cai wrote:
> On Thu, 2019-07-18 at 16:29 -0700, David Miller wrote:
> > From: Qian Cai <cai@....pw>
> > Date: Thu, 18 Jul 2019 19:26:47 -0400
> > 
> > >  
> > >  
> > > > On Jul 18, 2019, at 5:21 PM, Bill Wendling <morbo@...gle.com> wrote:
> > > >  
> > > > [My previous response was marked as spam...]
> > > >  
> > > > Top-of-tree clang says that it's const:
> > > >  
> > > > $ gcc a.c -O2 && ./a.out
> > > > a is a const.
> > > >  
> > > > $ clang a.c -O2 && ./a.out
> > > > a is a const.
> > > 
> > >  
> > >  
> > > I used clang-7.0.1. So, this is getting worse where both GCC and clang
> > > will
> > 
> > start to suffer the
> > > same problem.
> > 
> > Then rewrite the module parameter macros such that the non-constness
> > is evident to all compilers regardless of version.
> > 
> > That is the place to fix this, otherwise we will just be adding hacks
> > all over the place rather than in just one spot.
> 
> The problem is that when the compiler is compiling be_main.o, it has no
> knowledge about what is going to happen in load_module().  The compiler can
> only
> see that a "const struct kernel_param_ops" "__param_ops_rx_frag_size" at the
> time with
> 
> __param_ops_rx_frag_size.arg = &rx_frag_size
> 
> but only in load_module()->parse_args()->parse_one()->param_set_ushort(), it
> changes "__param_ops_rx_frag_size.arg" which in-turn changes the value
> of "rx_frag_size".

Even for an obvious case, the compilers still go ahead optimizing a variable as
a constant. Maybe it is best to revert the commit d66acc39c7ce ("bitops:
Optimise get_order()") unless some compiler experts could improve the situation.

#include <stdio.h>

int a = 1;

int main(void)
{
        int *p;

        p = &a;
        *p = 2;

        if (__builtin_constant_p(a))
                printf("a is a const.\n");

        printf("a = %d\n", a);

        return 0;
}

# gcc -O2 const.c -o const

# ./const
a is a const.
a = 2