GeezerVisonThe World Wide Web, restored to 800 pixels of glory. | ||||||||||||||||
| Home|About|Gaming cabinet|Archives|How it works|Starter sites | ||||||||||||||||
GeezerVison is showing github.com — Opcode Opcode
hyqneuron edited this page
Note: information here is not organised. Many parts may be unclear and confusing. Information may be inaccurate. Currently 2 types of opcodes found: 4-byte and 8-byte. For instruction names and their identifier bits (na, nb), see Instructions, nanb. General format: The numbering of bit fields below, such as re0 3:5, start from the MSB, but the displayed binary digits start from the LSB. For instance, when pr is 0101, then "pr 2" will be 1 and "pr 0:2" will be 101. These two bit fields specify the instruction name. Invalid bits result in invalid opcode. Predicate bits
Note: Predicate register 7 , P7, is actually recognised by cuobjdump as "pt", which is similar to "RZ" in the sense that it possibly means "no predicate" (1110 means no predicate used). Also, P4, P5 and P6 have not been tested. Use of those predicate registers may trigger error. re0 is most often used for as the destination register. However, for SETP instructions re0 0:2 is the destination predicate register(p), and re0 3:5 is destination predicate for NOT p(~p). re1 is often the second register operand. It is also used to store the Rxx in For instructions with the suffix I32, the entirety of the field is used to store a 32-bit immediate value. For most other instructions that use this field, imme is divided into two fields: immea(imme10:31) and immeb(imme0:9). immea can be used to indicate a 20-bit immediate value, a constant memory address, or a register. immea is most often used by a composite operand
immeb is sometimes used to store modifier bits. General format:
Clone this wiki locally | ||||||||||||||||
| You are visitor 01402503 — GeezerVison does not store the pages you look up. Best viewed with Netscape Navigator 4.7 or Internet Explorer 6 at 800×600. No cookies required. |