Jaromír Dolecek wrote:
> 
> Scott Bartram wrote:
> > "Jarom_r Dolecek" wrote:
> > > * compat/ibcs2/ibcs2_misc.c:ibcs2_sys_time() checked for emul_ibcs2_xout;
> > >   I've disabled that check for now, since there is now only single
> > >   emul_ibcs2. What is the right thing to do here ?
> >
> > So the answer is that the emul code needs to be able to query the original
> > object format otherwise each format has to become a unique emulation.
> 
> Seems so. Though the thing in ibcs2_sys_time() seems more like abuse
> of the exec info than the right thing - most probably, it's just
> compatibility quirk for "really old" SCO time(2) syscall. (That makes
> me wonder why they did not version it).

Yes, I suppose it could be considered "abuse". I'd be glad to consider
alternatives. If we had a generic flags field in the emul_ibcs2 structure
that could be set and tested.

I'm sure there are other places our emulation is not quite correct for
old XENIX binaries. That just happens to be one that someone stumbled on
and I figured out through disassembly since I didn't have any XENIX docs.