On 19.07.2018 08:42, Markus Armbruster wrote: > Peter Maydell <peter.mayd...@linaro.org> writes: > >> On 17 July 2018 at 20:50, Eduardo Habkost <ehabk...@redhat.com> wrote: >>> I have been looking at patches that touch typedefs.h and >>> wondering: why do we make typedefs.h necessary at all? Why do we >>> always add typedefs for every struct and union type in QEMU? >>> >>> Why do we prefer to write this: >>> >>> ----- qemu/typedefs.h: >>> typedef struct SomeType SomeType; >>> ---------------------- >>> >>> ----- qemu/somecode.h: >>> #include <qemu/typedefs.h> >>> >>> int some_function(SomeType *a); >>> ---------------------- >>> >>> >>> ...instead of simply writing this:? >>> >>> ----- qemu/somecode.h: >>> struct SomeType; >>> int some_function(struct SomeType *a); >>> ---------------------- >>> >>> Is the maintenance burden of typedefs.h worth it? >> >> Personally I don't like typing "struct " all the time >> when I'm using the type... > > For a different point of view (which I happen to share), see > https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/tree/Documentation/process/coding-style.rst?id=024ddc0ce1049298bd3cae60ae45d9c5f0fb8b9c#n318
FWIW, as somebody who also has to write some kernel code occasionally, I'd also prefer if we could get closer to the kernel coding style here again. That would also avoid problems like this that we used to hit regularly: https://lists.nongnu.org/archive/html/qemu-devel/2018-01/msg02695.html (well, that problem will likely go away automatically since we now don't support these distros anymore, but still...) Thomas
