[RFA] confine _TIMEVAL_DEFINED guard in libc/include/sys/time.h

[Ralf Corsepius]

On 11/12/2012 05:04 PM, Christopher Faylor wrote:

> CVS spelunking would show that these particular assumptions have only
> been in place since July.

RTEMS-4.11 newlib is pretty closely tracking newlib-CVS (Last synced ca. 
2012-10-11).

Though some upstream newlib patches have been reverted (and others 
added), your previous _TIMEVAL_DEFINED patch is part of the RTEMS newlib 
patches.

> But, yes, the whole point of my attempting to be cautious was obviously
> so that I didn't break code which relied on legacy behavior.
I don't see how your new patch would change RTEMS behaviour, because 
RTEMS hasn't had defined _TIMEVAL_DEFINED before and therefore had 
relied upon receiving struct timeval and struct timezone from sys/time.h

That said (withouth having tested it), your patch seems OK with me.

Ralf