Path: utzoo!utgpu!jarvis.csri.toronto.edu!mailrus!tut.cis.ohio-state.edu!cs.utexas.edu!uunet!mcsun!tuvie!inst182
From: inst182@tuvie (Inst.f.Techn.Informatik)
Newsgroups: gnu.gcc
Subject: Re: short jumps
Keywords: port
Message-ID: <1051@tuvie>
Date: 4 Feb 90 12:32:04 GMT
References: <1049@tuvie> <25CA909C.21229@paris.ics.uci.edu>
Reply-To: mike@vlsivie.uucp (Michael K. Gschwind)
Organization: Technical University of Vienna, EDP-Center
Lines: 23

In article <25CA909C.21229@paris.ics.uci.edu> rfg@ics.uci.edu (Ron Guilmette) writes:
>>This jump has completely different semantics than a far jump. How can I
>>check if I can use the short jump instead of a far jump. I cannot delay
>>the decision till assembly time, so I *MUST* have gcc emit the right 
>>instruction. *ANY* hints will be appreciated.
>
>Why can't you "dealy the decision till assembly time"?  I believe that at
>least some assemblers (e.g. for ns32k) optimize long jumps into short ones
>where they find that it is possible and legal to do.
>
>// rfg

The problem is that the jumps have *COMPLETELY* different semantics.
This means that I have to clobber a register because far jumps are
always register indirect. Now I could say, "OK, every jump clobbers a
register", but this I do not want to do.

       ____  ____
      /   / / / /   Michael K. Gschwind             mike@vlsivie.at
     /   / / / /    Institute for VLSI-Design       mike@vlsivie.uucp
     ---/           Technical University, Vienna 
       / 
   ___/