[Lsb-infrastructure] a couple more small bumps with the new headers

[Wichmann, Mats D]

Denis Silakov wrote:
> Wichmann, Mats D wrote:
>> the prototype for glBitmap needed a const qualifer
>> on the final argument.  I'm checking this in now.
>> One of the parts of the build has something with
>> treat-warnings-as-errors turned on so it died.
>> I did turn on the GL headers today by request, or
>> we would not be seeing this.
> 
> update Parameter set Ptype=16546 where Pint=2794 and Ppos=6;

I have already handled this one.

>> _POSIX_VERSION has vanished from the unistd.h hdr.
>> Or to be more accurate, it's LSB_VERSION < 21 only.
>> this kills another piece of the libbat build (which
>> is a cornerstone for a number of other things).
> 
> Simply apply the patch from bug 2039 - in addition to h_errno
> discussed earlier, it will turn back _POSIX_VERSION, _POSIX2_VERSION
> and some other constants.

Okay.  I'm trying to catch up, honest :-)

I've also integrated a bunch of devchk-reported size changes,
but there are some still to go in the X11 area.  I know
there's another round to go as in some cases devchk tells
us we must create a zero-size entry first, then next time
we get reports on what the sizes should be for each arch.
Unfortunately I didn't get all of these in in time for today's
spin, so they'll get some new numbers after the next one.

Something I wasn't sure about was the large number of proposed
ArchType entries in the X11 area where it was only complaining 
about the three 32-bit architectures, and a rather smaller 
number of cases where it was complaining only about the 64-bit 
architectures. These come from the recent imports, as far as
I can tell. Maybe these will work out after a few cycles; 
but as I was surprised, I didn't apply yet.  These are
entries like IcePointer, IceConn, IceAuthDataEntry,
IceAuthFileEntry, etc.