Proper Conditionals for POSIX CX and XSI

Joel Sherrill Joel.Sherrill@OARcorp.com
Mon Aug 26 08:58:00 GMT 2013


Thanks! That certainly clarifies things. Now I just need to get to coding. :)

We had a student working on adding restrict but he was working from home mostly. I need to get back with him now that he is back in school and start reviewing and submitting his patches.

Mike Frysinger <vapier@gentoo.org> wrote:


On Thursday 18 July 2013 12:01:20 Joel Sherrill wrote:
> I am trying to write the rules to give to a future volunteer which
> may or may not be me.
>
> Is it possible to get group consensus on the conditionals which reflect
> each of the columns in the API tables for SUS V4?

i think POSIX dictates the defines and their values already.  like here:
http://pubs.opengroup.org/onlinepubs/9699919799/functions/V2_chap02.html#tag_15_02_01_01

> http://www.unix.org/version4/GS5_APIs.pdf
>
> I am having trouble seeing how each API standard doesn't correspond
> to a single conditional or maybe a POSIX standard range. And if it
> was required by K&R or C89 or XXX, then it is unconditional.

iiuc, when POSIX adds things, there's no guarantees that it adds everything at
once from a specific external spec, or at the same time.  or when things get
removed.  so different spec versions can yield a different feature set.
-mike



More information about the Newlib mailing list