MIPS I-IV instruction set standards

Maciej W. Rozycki macro at linux-mips.org
Tue Mar 7 17:42:42 CST 2017

On Tue, 28 Feb 2017, Angelo Papenhoff via cctech wrote:

> I'm wondering where the MIPS I-IV standards that are referenced
> everywhere are defined. I was able to actually find what seems to be the
> IV standard [1] but found no such thing for I-III. I didn't even find
> any bibliographic references to them. Did they only exist as printed
> books and nobody bothered to scan them? Or are they under copyright?
> Would be nice to have them accessible somewhere.
> [1]
> http://www.cs.cmu.edu/afs/cs/academic/class/15740-f97/public/doc/mips-isa.pdf

 A differently formatted copy of the same document used to be available at 
though `techpubs.sgi.com' has been recently taken offline I'm told (and I 
haven't checked if `archive.org' has a copy).  In either version of the 
document you can see how the ISA has progressed in the opcode tables 
towards the end of each appendix; this might be the best original ISA 
reference readily available.

 As to the earlier ISA levels I believe their definitions were only buried 
as parts of the respective MIPSCO processor manuals, i.e. the R2000/R3000, 
the R6000 and the R4000/R4400.

 The latter is available online in several places; originally at 
<http://www.mips.com/Documentation/R4400_Uman_book_Ed2.pdf> and 

 The final version of the R6000 manual may not have been made.  A very 
early revision was found and has been scanned and published at 
<ftp://ftp.linux-mips.org/pub/mips/docs/r6000/> as tarballs of TIFF images 
(I have made a PDF conversion and placed the result at 
<ftp://ftp.linux-mips.org/pub/linux/mips/people/macro/MIPS/>).  It is 
inaccurate and lacks opcode encodings.

 I have seen a copy of the original R2000 manual once, in the form of a 
collection of pages in a ring binder.  That copy may have been lost since.  
Silicon manufacturers like IDT, LSI Logic, Performance Semiconductor or 
Siemens published their R2000/R3000 (and R2010/R3010 FPA) implementation 
manuals though that may serve as a reference; you should be able to track 
down scanned copies online.



More information about the cctech mailing list