On 02/19/2016 10:24 AM, Peter Maydell wrote: > On 6 February 2016 at 13:43, Peter Maydell <peter.mayd...@linaro.org> wrote: >> On 6 February 2016 at 00:51, Richard Henderson <r...@twiddle.net> wrote: >>> On 01/27/2016 05:17 AM, Peter Maydell wrote: >>>> diff --git a/tcg/i386/tcg-target.c b/tcg/i386/tcg-target.c >>>> index 9187d34..d90636c 100644 >>>> --- a/tcg/i386/tcg-target.c >>>> +++ b/tcg/i386/tcg-target.c >>>> @@ -22,6 +22,7 @@ >>>> * THE SOFTWARE. >>>> */ >>>> >>>> +#include "qemu/osdep.h" >>>> #include "tcg-be-ldst.h" >>>> >>>> #ifndef NDEBUG >>> >>> >>> Nack to these, and the others like them. >>> These files are not standalone, they are >>> included into tcg.c, so we ought not be >>> re-including qemu/osdep.h here. >> >> Mmm, but I preferred to retain the invariant that "all .c >> files include osdep.h first" rather than special casing >> these; the re-include is harmless. >> >> Alternatively we could rename these tcg-target.c >> files to some other extension that makes it clearer that >> they're not standalone source files. > > How do you feel about renaming the tcg-target.c files to > tcg-target.inc.c ? Then I can make clean-includes know that > *.inc.c are not to have the cleaning rules applied to them.
I suppose. Though is the pattern "*.inc.c" really any better than "tcg-target.c"? r~