Re: include/linux/fortify-string.h:52:29: warning: '__builtin_strcpy' source argument is the same as destination

[Yujie Liu]

On Wed, Dec 20, 2023 at 02:20:52PM +0800, Yujie Liu wrote:
> Hi Kees,
> 
> On Thu, Nov 30, 2023 at 02:11:54PM -0800, Kees Cook wrote:
> > On Thu, Nov 30, 2023 at 12:02:50PM +0800, kernel test robot wrote:
> > > tree:   https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git master
> > > head:   3b47bc037bd44f142ac09848e8d3ecccc726be99
> > > commit: ba38961a069b0d8d03b53218a6c29d737577d448 um: Enable FORTIFY_SOURCE
> > > date:   1 year, 3 months ago
> > > config: um-randconfig-r034-20230830 (https://download.01.org/0day-ci/archive/20231130/202311301039.7i51bZCz-lkp@xxxxxxxxx/config)
> > > compiler: gcc-9 (Debian 9.3.0-22) 9.3.0
> > > reproduce (this is a W=1 build): (https://download.01.org/0day-ci/archive/20231130/202311301039.7i51bZCz-lkp@xxxxxxxxx/reproduce)
> > > 
> > > If you fix the issue in a separate patch/commit (i.e. not just a new version of
> > > the same patch/commit), kindly add following tags
> > > | Reported-by: kernel test robot <lkp@xxxxxxxxx>
> > > | Closes: https://lore.kernel.org/oe-kbuild-all/202311301039.7i51bZCz-lkp@xxxxxxxxx/
> > > 
> > > [...]
> > >    kernel/kallsyms.c: In function '__sprint_symbol.isra.0':
> > > >> include/linux/fortify-string.h:52:29: warning: '__builtin_strcpy' source argument is the same as destination [-Wrestrict]
> > >       52 | #define __underlying_strcpy __builtin_strcpy
> > >          |                             ^
> > >    include/linux/fortify-string.h:567:10: note: in expansion of macro '__underlying_strcpy'
> > >      567 |   return __underlying_strcpy(p, q);
> > >          |          ^~~~~~~~~~~~~~~~~~~
> > 
> > The only strcpy() in __sprint_symbol() is:
> > 
> > 	if (name != buffer)
> > 		strcpy(buffer, name);
> > 
> > Which is explicitly not the same address...
> > 
> > This appears to be a GCC 9 false positive, maybe?
> 
> Sorry for our late reply. We tested this with various versions of gcc,
> and this error disappeared when building with GCC 10 and 11, so it is
> very likely to be a GCC 9 false positive. We will configure the bot to
> use compilers newer than GCC 9 to test um arch.

Sorry, we will keep testing um with GCC 9 and configure the bot to
ignore this specific warning.