home *** CD-ROM | disk | FTP | other *** search
- Path: sparky!uunet!mcsun!sunic!dkuug!dkuug!diku!terra
- From: terra@diku.dk (Morten Welinder)
- Newsgroups: comp.sys.intel
- Subject: Re: Undocumented instructions in x86 CPUs
- Message-ID: <1992Nov7.154508.11802@odin.diku.dk>
- Date: 7 Nov 92 15:45:08 GMT
- References: <mostert.51@cs.sun.ac.za> <1992Nov5.101717.26632@umr.edu> <mostert.54@cs.sun.ac.za>
- Sender: terra@rimfaxe.diku.dk
- Organization: Department of Computer Science, U of Copenhagen
- Lines: 33
-
- mostert@cs.sun.ac.za (A. Mostert) writes:
-
- >In article <1992Nov5.101717.26632@umr.edu> mcastle@mcs213d.cs.umr.edu (Michael R Castle) writes:
- >>In article <mostert.51@cs.sun.ac.za> mostert@cs.sun.ac.za (A. Mostert) writes:
- >>>I just became aware of an undocumented instruction that (I think) works on
- >>>all members of the x86 family. The opcode is D4 10 and it does this:
- > ^^^^
- >>> AH = AL >> 4
- >>> AL = AL & 0xF
-
- >>hmm.... listed in my Microprocessors and Peripheral Handbook... AAM
-
- >AAM is opcode D4 0A
- >AAD is opcode D5 0A
-
- >The second byte is actually a number specifying the base, as various people
- >pointed out to me since my original posting. However, only the form with 0A
- >as base is officialy documented by Intel.
-
- >Can anyone tell me if this works on the NEC v20/V30 processors ?
-
- >A. Mostert
- >mostert@cs.sun.ac.za
-
- From memory: AAM does, AAD does not.
-
- Morten Welinder
- terra@diku.dk
- --
- -----------------------------------------------------------
- Signature file under construction -- please drive carefully
- -----------------------------------------------------------
- Morten Welinder, terra@diku.dk
-