Re: [coreboot] Cbfstool compilation error with gcc 5.1

[Stefan Reinauer]

* Anatol Pomozov  [150531 08:49]:
> Hi
>
> I am trying to compile coreboot cbfstool on Arch where gcc 5.1 is used. And I
> see following compilation error. I wonder if it is coreboot or gcc issue.
>
> ==> Starting build()...
> make: Entering directory '/home/anatol/sources/archpackages/coreboot-utils-git/
> src/coreboot/util/cbfstool'
> mkdir -p ./
> cc -D_FORTIFY_SOURCE=2 -D_DEFAULT_SOURCE  -D_POSIX_C_SOURCE=200809L  -Iflashmap
> -march=x86-64 -mtune=generic -O2 -pipe -fstack-protector-strong --param=
> ssp-buffer-size=4 -g3 -std=c99 -Werror -Wall -Wextra -Wcast-qual
> -Wmissing-prototypes -Wredundant-decls -Wshadow -Wstrict-prototypes
> -Wwrite-strings -c -o rmodule.o rmodule.c
> rmodule.c: In function ‘rmodule_create’:
> rmodule.c:286:29: error: ‘phdr’ may be used uninitialized in this function
> [-Werror=maybe-uninitialized]
>         (phdr->p_vaddr + phdr->p_memsz))) {
>                              ^
> rmodule.c:203:14: note: ‘phdr’ was declared here
>   Elf64_Phdr *phdr;
>               ^
> cc1: all warnings being treated as errors
> Makefile:55: recipe for target 'rmodule.o' failed
> make: *** [rmodule.o] Error 1
> make: Leaving directory '/home/anatol/sources/archpackages/coreboot-utils-git/
> src/coreboot/util/cbfstool'
> ==> ERROR: A failure occurred in build().

Looks like a false positive. ctx-phdr only gets passed on if nelements
is 1, in which we know that phdr was set.

Stefan

Cbfstool compilation error with gcc 5.1

[Anatol Pomozov]

Hi

Sorry for a long delay

On Wed, Jun 10, 2015 at 6:38 PM, Stefan Reinauer
 wrote:
> Anatol Pomozov  [150531 08:49]:
> > Hi
> >
> > I am trying to compile coreboot cbfstool on Arch where gcc 5.1 is used. And I
> > see following compilation error. I wonder if it is coreboot or gcc issue.

That's indeed looks like a false positive from GCC 5.1. And it happens
only if "-O2" flag present.

Here is what Ron proposed and it makes GCC happy again
http://review.coreboot.org/#/c/10881/
> > ==> Starting build()...
> > make: Entering directory '/home/anatol/sources/archpackages/coreboot-utils-git/
> > src/coreboot/util/cbfstool'
> > mkdir -p ./
> > cc -D_FORTIFY_SOURCE=2 -D_DEFAULT_SOURCE  -D_POSIX_C_SOURCE=200809L  -Iflashmap
> > -march=x86-64 -mtune=generic -O2 -pipe -fstack-protector-strong --param=
> > ssp-buffer-size=4 -g3 -std=c99 -Werror -Wall -Wextra -Wcast-qual
> > -Wmissing-prototypes -Wredundant-decls -Wshadow -Wstrict-prototypes
> > -Wwrite-strings -c -o rmodule.o rmodule.c
> > rmodule.c: In function ‘rmodule_create’:
> > rmodule.c:286:29: error: ‘phdr’ may be used uninitialized in this function
> > [-Werror=maybe-uninitialized]
> >         (phdr->p_vaddr + phdr->p_memsz))) {
> >                              ^
> > rmodule.c:203:14: note: ‘phdr’ was declared here
> >   Elf64_Phdr *phdr;
> >               ^
> > cc1: all warnings being treated as errors
> > Makefile:55: recipe for target 'rmodule.o' failed
> > make: *** [rmodule.o] Error 1
> > make: Leaving directory '/home/anatol/sources/archpackages/coreboot-utils-git/
> > src/coreboot/util/cbfstool'
> > ==> ERROR: A failure occurred in build().
>
> Looks like a false positive. ctx-phdr only gets passed on if nelements
> is 1, in which we know that phdr was set.
>
> Stefan