Changeset 533 in xtideuniversalbios for trunk/XTIDE_Universal_BIOS/Src/Handlers

Timestamp:

Apr 3, 2013, 4:12:34 PM (12 years ago)

Author:

aitotat@…

google:author:

aitotat@gmail.com

Message:

Changes to XTIDE Universal BIOS:

• P-Cylinders returned by AH=48h are now calculated from L-CHS capacity when RESERVE_DIAGNOSTIC_CYLINDER is defined.
• Total sector count returned by AH=48h is now always P-Cylinders * P-Heads * P-Sectors per track for drives with less than or equal 15,482,880 LBA sectors.

Location:

trunk/XTIDE_Universal_BIOS/Src/Handlers

Files:

• Int13h.asm (modified) (1 diff)
• Int13h/EBIOS/AH48h_GetExtendedDriveParameters.asm (modified) (1 diff)

trunk/XTIDE_Universal_BIOS/Src/Handlers/Int13h.asm

@@ -251 +251 @@
 ;   Parameters:
 ;       AH:     BIOS Error code
-;       DS:     RAMVARS segment
 ;       SS:BP:  Ptr to IDEPACK
 ;   Returns:

trunk/XTIDE_Universal_BIOS/Src/Handlers/Int13h/EBIOS/AH48h_GetExtendedDriveParameters.asm

@@ -23 +23 @@
 ;--------------------------------------------------------------------
 ; Int 13h function AH=48h, Get Extended Drive Parameters.
+;
+; It is not completely clear what this function should return as total sector count in some cases.
+; What is certain is that:
+; A) Phoenix Enhanced Disk Drive Specification v3.0 says that P-CHS values
+;    are never returned for drives with more than 15,482,880 sectors (16384*15*63).
+;    For those drives we can simply return total sector count from
+;    ATA ID WORDs 60 and 61 (LBA28) or WORDs 100-103 (LBA48).
+; B) IBM PC DOS 7.1 fdisk32 displays warning message if P-CHS values multiplied
+;    together are different than total sector count. Therefore for drives with less
+;    than or equal 15,482,880 sectors we MUST NOT return total sector count from
+;    ATA ID WORDs 60 and 61.
+;
+;    Lets take an example. 6 GB Hitachi microdrive reports following values in ATA ID:
+;    Sector count from WORDs 60 and 61  : 12,000,556
+;    Cylinders                          : 11905
+;    Heads                              :    16
+;    Sectors per track                  :    63
+;
+;    When multiplying C*H*S we get      : 12,000,240
+;    So the CHS size is a little bit less than LBA size. But we must use
+;    the smaller value since C*H*S must equal total sector count!
+;
+; Now we get to the uncertain area where I could not find any information.
+; Award BIOS from 1997 Pentium motherboard returns following values:
+;    AH=08h L-CHS:   745, 255, 63 (exactly the same as what we return)
+;    => Total Sector Count:       745*255*63 = 11,968,425
+;    AH=48h P-CHS: 11873,  16, 63
+;    AH=48h Total Sector Count: 11873* 16*63 = 11,967,984
+;
+; Notice how AH=48h returns lesser total sector count than AH=8h! The only
+; way I could think of to get 11873 cylinders is to divide AH=08h sector
+; count with P-CHS heads and sectors: (745*255*63) / (16*63) = 11873
+;
+; I have no idea what is the reasoning behind it but at least there is one
+; BIOS that does just that.
+;
+; Since I don't have any better knowledge, I decided that when RESERVE_DIAGNOSTIC_CYLINDER
+; is defined, we do what the Award BIOS does. When it is not defined, we multiply
+; P-CHS values together and use that as total sector count.
+;
 ;
 ; AH48h_GetExtendedDriveParameters