Date: Fri, 15 Nov 2002 23:04:57 -0500
To: Willard Goosey <goosey@virgo.sdc.org>
From: Rodney V Hamilton <rodneyhamilton@mpinet.net>
Subject: IDE errata and fixes

#################################
CHS test error in makedesc.bas!
#################################
I ran into the following error in the 10/30/99 version of MAKEDESC.BAS.
The CHS partition boundary test in line 505 has a typo - it does a MOD
test against "tempbig2" instead of "tempbig" which makes the program do
an unwarranted skip to the next higher cylinder.

Here is the relevant code fragment:

============================
(* CHS mode - HAS to start on 'cylsize' sector boundary (cylinder boundary)
      tempbig2=tempbig/cylsize
      IF MOD(tempbig2,cylsize)<>.0 THEN
        tempbig2=INT(tempbig2)+1.
        PRINT "Last partition did not end on an even cylinder...you will
lose ";tempbig2*cylsize-tempbig;" sectors"
      ENDIF
      partitionstart=tempbig2
      PRINT "Recommended partition start is cylinder ";partitionstart;"."
============================

The third line should instead read:

      IF MOD(tempbig,cylsize)<>.0 THEN

With this change, the program correctly detects cylinder boundaries and
also matches the corresponding LBA logic.