Article 149611 of comp.os.vms: In article <31C1B34B.C3E@interramp.com>, Ray Cadmus writes: >I'm having problems building PERL5 on a DEC Alpha. Peter Prymmer >pointed me to the source for the good VMS files @upenn. With these >I was able to get v5.002 up on my VAX systems. I've been trying for weeks >(off & on) to get it up on an AXP. Keep getting things like: > >******************************************************************************** >RCADMUS(ZORRO) $ mmk /descrip=[.vms] /Macro="__AXP__=1" >CC /Define=(DEBUGGING)/Include=[]/Standard=Relaxed_ANSI/Prefix=All/Obj=.obj/NoLi >st MINIPERLMAIN.C > ># pragma __message disable (__MISALGNDSTRCT) >.............................................^ >%CC-W-UNKMSGID, Unknown message id or group "MISALGNDSTRCT" is ignored. >at line number 313 in file DISK$ZORROD1:[PERL.PERL5_002]VMSISH.H;1 > ># pragma __message disable (__MISALGNDMEM) >...........................................^ >%CC-W-UNKMSGID, Unknown message id or group "MISALGNDMEM" is ignored. >at line number 314 in file DISK$ZORROD1:[PERL.PERL5_002]VMSISH.H;1 >%MMK-F-ERRUPD, error status %X10B91260 occurred when updating target MINIPERLMAI >N.OBJ >*********************************************************************************** >**** > >I get an eror and die after the creation of each object file, but it's there. > >If I keep stepping through, I build a libperl.olb but get the err there also. >I geep going and get lots of undefined symbols and end up with garbage. > >HELP! Please! > >Ray Ray - this looks tough. Most VMS_AXP people forget the /Macro="__AXP__=1" but you did not. Hmm... I know that I have built 5.002_01 on AXP - using MMS rather than MMK. MMK is supposed to do the DECC thing properly when on AXP - but maybe it isn't working? For right now I would play with combinations such as $ mmk /descrip=[.vms]/macro=("decc=1") /Macro="__AXP__=1" or even this (?) $ mmk /descrip=[.vms]/macro=("decc=1", "cc=CC/DECC") /Macro="__AXP__=1" Actually I think the "decc=1" _is_ required. I will send this along to the VMSPerl mailing list as well. Peter Prymmer pvhp@lns62.lns.cornell.edu