Path: utzoo!utgpu!news-server.csri.toronto.edu!cs.utexas.edu!uwm.edu!bionet!agate!ucbvax!DSG4.DSE.BECKMAN.COM!a_rubin
From: a_rubin@DSG4.DSE.BECKMAN.COM (arthur rubin)
Newsgroups: comp.binaries.ibm.pc.d
Subject: Re: v11i133: ndelv00, deleted all but list (part 01/01)
Message-ID: <9104152054.AA23679@dn71.dse.beckman.com>
Date: 15 Apr 91 20:54:46 GMT
Sender: daemon@ucbvax.BERKELEY.EDU
Lines: 27

My newsserver died last Friday afternoon, but this may still be appropriate and
"timely".  I downloaded no.zip from garbo, and I discovered a serious error  If
this is not done, tsr fragmentation can cause the instruction at 26b to write
into a completely different module (it stores it at 209 offset from no's
environment segment, which could easily be in another program entirely.)


A binary patch to no.com follows (all numbers are hex, and the first column is
the byte offset in the file):

267    1E               push    ds
268    06               push    es
269    1F               pop     ds
26A    07               pop     es
26B    89 26 0209       mov     209,sp

should be:

267    89 26 0209       mov     209,sp
26B    1E               push    ds
26C    06               push    es
26D    1F               pop     ds
26E    07               pop     es

------------------------------------------------------------------------
a_rubin@dsg4.dse.beckman.com  
My opinions are my own, and do not represent those of my employer.