This is the mail archive of the newlib@sourceware.org mailing list for the newlib project.
Index Nav: | [Date Index] [Subject Index] [Author Index] [Thread Index] | |
---|---|---|
Message Nav: | [Date Prev] [Date Next] | [Thread Prev] [Thread Next] |
Other format: | [Raw text] |
On Apr 15 12:57, Sebastian Huber wrote: > > > On 15/04/16 12:08, Corinna Vinschen wrote: > >On Apr 14 13:54, Sebastian Huber wrote: > >>Signed-off-by: Sebastian Huber <sebastian.huber@embedded-brains.de> > >>--- > >> newlib/libc/include/sys/_timespec.h | 3 ++- > >> newlib/libc/include/sys/_timeval.h | 3 ++- > >> newlib/libc/include/sys/_types.h | 4 +++- > >> newlib/libc/include/sys/timeb.h | 7 ++++--- > >> newlib/libc/include/sys/types.h | 7 ++++--- > >> newlib/libc/sys/sparc64/sys/_timeval.h | 7 ++++--- > >> 6 files changed, 19 insertions(+), 12 deletions(-) > >> > >>diff --git a/newlib/libc/include/sys/_timespec.h b/newlib/libc/include/sys/_timespec.h > >>index 4a7aabc..7609e4a 100644 > >>--- a/newlib/libc/include/sys/_timespec.h > >>+++ b/newlib/libc/include/sys/_timespec.h > >>@@ -36,9 +36,10 @@ > >> #include <sys/_types.h> > >>-#ifndef __time_t_defined > >>+#if !defined(__time_t_defined) && !defined(_TIME_T_DECLARED) > >> typedef _TIME_T_ time_t; > >> #define __time_t_defined > >>+#define _TIME_T_DECLARED > >> #endif > >>[...] > >Would it make sense to outsource the definition of time_t to it's > >own file, e.g. sys/_time.h and just include it everywhere it's > >needed? > > It is used in FreeBSD like this: > > ./sys/sys/types.h:#ifndef _TIME_T_DECLARED > ./sys/sys/timeb.h:#ifndef _TIME_T_DECLARED > ./sys/sys/_timespec.h:#ifndef _TIME_T_DECLARED > ./sys/sys/sem.h:#ifndef _TIME_T_DECLARED > ./sys/sys/_timeval.h:#ifndef _TIME_T_DECLARED > ./sys/sys/shm.h:#ifndef _TIME_T_DECLARED > ./sys/sys/msg.h:#ifndef _TIME_T_DECLARED > ./include/pwd.h:#ifndef _TIME_T_DECLARED > ./include/time.h:#ifndef _TIME_T_DECLARED > ./include/timeconv.h:#ifndef _TIME_T_DECLARED > ./include/utime.h:#ifndef _TIME_T_DECLARED > > FreeBSD uses the _XYZ_T_DECLARED mechanism, glibc the __need_xyz, and there > are surely other ways to deal with the <a.h> and <b.h> both need type xyz_t, > but should not make visible other types problem. Ok. Other than 02, 04 and 17 the patches look good. As for 04, only if it's not asked too much. I can also apply a followup patch removing the Cygwin definition. Thanks, Corinna -- Corinna Vinschen Cygwin Maintainer Red Hat
Attachment:
signature.asc
Description: PGP signature
Index Nav: | [Date Index] [Subject Index] [Author Index] [Thread Index] | |
---|---|---|
Message Nav: | [Date Prev] [Date Next] | [Thread Prev] [Thread Next] |