On Tue, Feb 2, 2021 at 4:57 PM Emil Velikov <emil.l.velikov@xxxxxxxxx> wrote:
On Tue, 2 Feb 2021 at 23:25, James Park <james.park@xxxxxxxxxxxxxxxx> wrote:
>
> On Tue, Feb 2, 2021 at 2:47 PM Emil Velikov <emil.l.velikov@xxxxxxxxx> wrote:
> >
> > Currently, the drm_fourcc.h header depends on drm.h for __u32 and __u64.
> > At the same time drm.h pulls a lot of unneeded symbols.
> >
> > Add new guard DRM_FOURCC_STANDALONE, which when set will use local
> > declaration of said symbols.
> >
> > When used on linux - this is a trivial but only when building in strict c99
> > mode. One is welcome to ignore the warning, silence it or use c11. If neither
> > of the three is an option, then do _not_ set the new guard.
> >
> > Cc: James Park <james.park@xxxxxxxxxxxxxxxx>
> > Cc: Pekka Paalanen <pekka.paalanen@xxxxxxxxxxxxx>
> > Cc: Simon Ser <contact@xxxxxxxxxxx>
> > Signed-off-by: Emil Velikov <emil.l.velikov@xxxxxxxxx>
> > ---
> > As mentioned before - there's little point in having yet another header
> > since keeping those in sync has been a PITA in the past.
> > ---
> > include/uapi/drm/drm_fourcc.h | 19 +++++++++++++++++++
> > 1 file changed, 19 insertions(+)
> >
> > diff --git a/include/uapi/drm/drm_fourcc.h b/include/uapi/drm/drm_fourcc.h
> > index 6f0628eb13a6..c1522902f6c9 100644
> > --- a/include/uapi/drm/drm_fourcc.h
> > +++ b/include/uapi/drm/drm_fourcc.h
> > @@ -24,7 +24,26 @@
> > #ifndef DRM_FOURCC_H
> > #define DRM_FOURCC_H
> >
> > +/*
> > + * Define DRM_FOURCC_STANDALONE you're interested only FOURCC and do not want
> > + * to pull drm.h into your application.
> > + */
> > +#ifdef DRM_FOURCC_STANDALONE
> > +#if defined(__linux__)
> > +
> > +#include <linux/types.h>
> > +
> > +#else /* One of the BSDs */
> > +
> > +#include <stdint.h>
> > +typedef uint32_t __u32;
> > +typedef uint64_t __u64;
> > +
> > +#endif /* __linux __ */
> > +
> > +#else
> > #include "drm.h"
> > +#endif /* DRM_FOURCC_STANDALONE */
> >
> > #if defined(__cplusplus)
> > extern "C" {
> > --
> > 2.30.0
> >
>
> One of my earlier solutions similarly would have forced people to deal
> with duplicate typedefs, and we arrived at the current solution
> because we didn't want to burden anyone with that.
As summed in the commit message the burden is only applicable when all
of the following are set:
- non-linux
- force DRM_FOURCC_STANDALONE
- c99 -pedantic
Even then, we're talking about a compilation warning. So yeah - let's
keep things short and sweet.
Side note: AFAICT MSVC will not trigger a warning so your logs should be clean.
-Emil
I'm having trouble reading your commit message, this sentence in particular: "When used on linux - this is a trivial but only when building in strict c99 mode."
This asymmetric copy/paste grosses me out so much. I don't think a patch like this should be opinionated about someone's build settings. Am I alone? Doesn't this bother anyone else?
- James
_______________________________________________ dri-devel mailing list dri-devel@xxxxxxxxxxxxxxxxxxxxx https://lists.freedesktop.org/mailman/listinfo/dri-devel
