Re: [Qemu-devel] [PATCH 1/9] build: disable Wstrict-prototypes

[Laszlo Ersek]

On 02/20/13 15:57, Kevin Wolf wrote:
> On Wed, Feb 20, 2013 at 03:50:11PM +0100, Laszlo Ersek wrote:
>> On 02/20/13 12:31, Kevin Wolf wrote:
>>> On Mon, Feb 18, 2013 at 05:56:57PM -0600, Anthony Liguori wrote:
>>>> GTK won't build with strict-prototypes due to gtkitemfactory.h:
>>>>
>>>>     /* We use () here to mean unspecified arguments. This is deprecated
>>>>      * as of C99, but we can't change it without breaking compatibility.
>>>>      * (Note that if we are included from a C++ program () will mean
>>>>      * (void) so an explicit cast will be needed.)
>>>>      */
>>>>     typedef        void    (*GtkItemFactoryCallback)  ();
>>>>
>>>> Signed-off-by: Anthony Liguori <address@hidden>
>>>> ---
>>>>  configure | 2 +-
>>>>  1 file changed, 1 insertion(+), 1 deletion(-)
>>>>
>>>> diff --git a/configure b/configure
>>>> index bf5970f..74d5878 100755
>>>> --- a/configure
>>>> +++ b/configure
>>>> @@ -283,7 +283,7 @@ sdl_config="${SDL_CONFIG-${cross_prefix}sdl-config}"
>>>>  # default flags for all hosts
>>>>  QEMU_CFLAGS="-fno-strict-aliasing $QEMU_CFLAGS"
>>>>  QEMU_CFLAGS="-Wall -Wundef -Wwrite-strings -Wmissing-prototypes 
>>>> $QEMU_CFLAGS"
>>>> -QEMU_CFLAGS="-Wstrict-prototypes -Wredundant-decls $QEMU_CFLAGS"
>>>> +QEMU_CFLAGS="-Wredundant-decls $QEMU_CFLAGS"
>>>>  QEMU_CFLAGS="-D_GNU_SOURCE -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE 
>>>> $QEMU_CFLAGS"
>>>>  QEMU_INCLUDES="-I. -I\$(SRC_PATH) -I\$(SRC_PATH)/include"
>>>>  if test "$debug_info" = "yes"; then
>>>
>>> Other places wrap the inclusion of problematic headers in '#pragma GCC
>>> diagnostic ...' instead of globally disabling warnings.
>>
>> Available only in gcc-4.6+:
>>
>> http://gcc.gnu.org/gcc-4.6/changes.html
> 
> I think this is mostly the push/pop part. We avoid it for this reason
> and just unconditionally make it error after including the header file.
> The pragma is also only enabled by a configure check.
> 
> Have a look at include/ui/qemu-pixman.h or coroutine-ucontext.c for an
> example.

Yes, I checked those before posting. Two points:

(a) #pragma PP_TOKEN1 PP_TOKEN2 ..., where PP_TOKEN1 is not STDC, has
implementation-defined effect. (Ie. the platform can do whatever it
wants, but has to document it). Non-STDC #pragmas that are not
recognized by the implementation must be ignored (C99 6.10.6p1
requirement). So wrapping "#pragma GCC diagnostic" is not really useful:
non-gcc compilers will surely not recognize "#pragma GCC whatever", and
an older gcc (not recognizing "#pragma GCC diagnostic") is required to
ignore it.

(b) My main point was, if you build on a gcc which understands
-Wstrict-prototypes and chokes on gtkitemfactory.h, but doesn't
understand the #pragma, then compilation will fail.

I think that on such a platform "qemu-pixman.h" / <pixman.h> from
pixman-0.16.0 would fail to parse as well.

What am I not understanding? :)

Thanks
Laszlo