davidsen@steinmetz.ge.com (William E. Davidsen Jr) (09/27/88)
Has someone ported the GNU cc compiler to Xenix/386?
--
bill davidsen (wedu@ge-crd.arpa)
{uunet | philabs}!steinmetz!crdos1!davidsen
"Stupidity, like virtue, is its own reward" -medyer@spdcc.COM (Steve Dyer) (09/27/88)
In article <12239@steinmetz.ge.com> davidsen@crdos1.UUCP (bill davidsen) writes: > > Has someone ported the GNU cc compiler to Xenix/386? I gave it a (very brief) attempt. It blew up on several files (you know, "compiler error") and on the ones it didn't blow up on, it mentioned that it was truncating variable names to some maximum length I can't remember. I don't think the variable name length truncation is a problem, but I couldn't be sure. -- Steve Dyer dyer@harvard.harvard.edu dyer@spdcc.COM aka {harvard,husc6,linus,ima,bbn,m2c,mipseast}!spdcc!dyer
dac@f.gp.cs.cmu.edu (Daniel Christian) (09/27/88)
We put some work into porting gcc to SCO Xenix/386. We got around all the compile problems but one. Most of the problems are due to macro expansion creating lines too long for cc. We ran these through cpp and then edited them to shorten lines. The unresolved problem was that some of the string constants in gcc.c were too long. Since gcc is just a front end to cpp, cc1 and as; you could hand compile gcc.c using GNU cpp, cc1, and Xenix as. HOWEVER, the last major problem was MASM. MASM is the standard assebler for SCO Xenix (there is one other assembler, but it is worse not better). The assembler output from gcc is Unix style assembly. MASM uses a different representation for numbers and addressing modes and (especially) variable allocation. Does anyone have a Unix style 386 assembler for Xenix?????? -Dan Christian Pittsburgh Powercomputing dac@ri.cmu.edu