[kbuild-devel] Re: [PATCH] Pointer cast warnings in scripts/

2005-06-22 Thread Pierre Ossman
Bernd Petrovitsch wrote: >On Wed, 2005-06-22 at 11:21 +0200, Pierre Ossman wrote: > > >>I was referring to which of the three types is correct for str*(). >> >> > >"char" as one can read in every man-page. > > > That doesn't really make it a standard though (de facto perhaps). :) The odds

[kbuild-devel] Re: [PATCH] Pointer cast warnings in scripts/

2005-06-22 Thread Russell King
On Wed, Jun 22, 2005 at 01:12:06PM +0200, Pierre Ossman wrote: > That doesn't really make it a standard though (de facto perhaps). :) > The odds of all those man pages deviating from the standard is probably > very low. But unless someone has actually read the damn thing we won't > know for sure.

[kbuild-devel] Re: [PATCH 1/2] Fix signed char problem in scripts/basic

2005-06-22 Thread Pierre Ossman
The signed characters in scripts are causing warnings with GCC 4 on systems with proper string functions (with char*, not signed char* as parameters). All signed strings were removed because they were all directly or indirectly used in string functions. Signed-off-by: Pierre Ossman <[EMAIL PROT

[kbuild-devel] Re: [PATCH] Pointer cast warnings in scripts/

2005-06-22 Thread Bernd Petrovitsch
On Wed, 2005-06-22 at 11:21 +0200, Pierre Ossman wrote: > Bernd Petrovitsch wrote: > >The C-standard about "char", "signed char" and "unsigned char"? > >These are 3 different types. > > I was referring to which of the three types is correct for str*(). "char" as one can read in every man-page.

[kbuild-devel] [PATCH 1/2] Fix signed char problem in scripts/basic

2005-06-22 Thread Pierre Ossman
The signed characters in scripts are causing warnings with GCC 4 on systems with proper string functions (with char*, not signed char* as parameters). Some could be kept signed but most had to be reverted to normal chars. Detailed changlog: fixdep.c: - is_defined_config() just used memcmp

[kbuild-devel] Re: [PATCH] Pointer cast warnings in scripts/

2005-06-22 Thread Pierre Ossman
Bernd Petrovitsch wrote: > >The C-standard about "char", "signed char" and "unsigned char"? >These are 3 different types. > > > I was referring to which of the three types is correct for str*(). Rgds Pierre --- SF.Net email is sponsored by

[kbuild-devel] [PATCH 2/2] Fix signed char problem in scripts/kconfig

2005-06-22 Thread Pierre Ossman
The signed characters in scripts are causing warnings with GCC 4 on systems with proper string functions (with char*, not signed char* as parameters). Some could be kept signed but most had to be reverted to normal chars. Detailed changelog: mconf.c: - buf/bufptr was used in vsprintf() so

[kbuild-devel] Re: [PATCH 2/2] Fix signed char problem in scripts/kconfig

2005-06-22 Thread Roman Zippel
Hi, On Wed, 22 Jun 2005, Pierre Ossman wrote: > The signed characters in scripts are causing warnings with GCC 4 on > systems with proper string functions (with char*, not signed char* as > parameters). Some could be kept signed but most had to be reverted to > normal chars. > > Detailed changel

[kbuild-devel] Re: [PATCH 1/2] Fix signed char problem in scripts/basic

2005-06-22 Thread Roman Zippel
Hi, On Wed, 22 Jun 2005, Pierre Ossman wrote: > -void define_config(const char * name, int len) > +void define_config(const void * name, int len) Could you try to drop the remaining "signed" instead of this? bye, Roman --- SF.Net email is s