uri@UUNET.UU.NET (01/15/90)
Hello, Sorry if I'm reporting an old, known and fixed bug. Machine: 386 At clone. System: ESIX Rev.C (SysV.3.2). GCC: standard 1.36 (not modified). It compiled OK, stage1 and stage2, but when I ran `cmp' aftr that (tail +10 ... of course) - one file failed the comparison. It was `regclass.o'. I'll provide more details, if you tell me what you need to fix it (or just to diagnose :-). Regards, Uri. uunet!bywater!scifi!angmar!uri
rob@phavl.UUCP (Robert Ransbottom) (01/15/90)
In article <9001150202.AA18124@scifi> bywater!scifi!angmar!uri@UUNET.UU.NET writes: > Machine: 386 At clone. > System: ESIX Rev.C (SysV.3.2). > GCC: standard 1.36 (not modified). > > It compiled OK, stage1 and stage2, but when I ran `cmp' > aftr that (tail +10 ... of course) - one file failed > the comparison. It was `regclass.o'. Most likely there is no problem. Except that the COFF timestamp bytes you're trying to avoid have generated a newline in one file. Do this: tail -NN regclass.o > foo1 # NN should be the total lines in regclass.o - 3 tail -NN stageN/regclass.o > foo2 cmp foo1 foo2 Now if I could just get g++ going on 386/ix.