1 / 14

CS 152b Final Report

CS 152b Final Report. Group 6. Background.

zuwena
Télécharger la présentation

CS 152b Final Report

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript


  1. CS 152b Final Report Group 6

  2. Background Our goal was to design and implement a functional 16-bit RISC processor using the Xilinx software. The constrains on our chip design was that it must be able to run on a Xess board and must adhere to the standards specified by our primary customer Young Cho Enterprises out of UCLA. Group 6 staff The R2-Yu2 processor • Randy Grant – Technical lead • Robert Johnson – VHDL master • Anthony (moo) Yu – Datapath guru • George Yu – Software designer

  3. Instruction Set Architecture • R-Type Instructions • No Op Instruction • I-Type Instructions • J-Type Instructions

  4. Pipeline Instruction Read Instruction Decode ALU Memory Register Write Back

  5. Instruction Cache • 32 Word Direct Mapped Cache • Instructions are read-only, so no write-back or write-through schemes are needed • Multiplexor selects proper cache entry for controller • Implemented in structural and behavioral VHDL • Each cache entry has 1 bit valid flag, 11 bits tag, 16 bits data • Cache Controller processes hits and misses • Demuxliplexor sets write signal for each cache entry (Max one bit high) • 28 bit registers store cache entries

  6. Instruction Cache • Cache integration with datapath • Memory accesses and cache lookups done in parallel On a hit, the next instruction comes from the cache On a miss, the next instruction comes from memory and is also pulled into the cache Given more time, we would further modify the integration of the cache with the datapath to increase the benefit of the cache addition

  7. Instruction Cache Stalling the Pipeline • Hold the PC value by feeding a zero increment        • Allow any JMP or BEQ instructions to alter it-send NOPs through the pipeline • Resume normal execution once hit goes high

  8. Typical Compiler • ScannerTakes in input file and tokenizes the input • Parser Transforms token stream into a grammatical phrases • Code Generation Creates assembly code from the grammar • Linker Turns the assembly code into the native bit code of the computer

  9. Our Compiler Solution • Implemented a stack for embedded structures • Combined parsing and scanning into a single array of struct (tokenlist), which stored all the tokens

  10. Implementing a Stack Initial state of the memory Push register sw r1, 51(r0) sw r2, 52(r0) sw r3, 53(r0) sw r4, 54(r0) sw r5, 55(r0) sw r6, 56(r0) sw r7, 57(r0) Heapstart = Heapstart + 7 Pop Register lw r7, 57(r0) lw r6, 56(r0) lw r5, 55(r0) lw r4, 54(r0) lw r3, 53(r0) lw r2, 52(r0) lw r1, 51(r0) Heapstart = Heapstart - 7

  11. Our Compiler Design L6: lw r7, 57(r0) lw r6, 56(r0) lw r5, 55(r0) lw r2, 52(r0) lw r1, 51(r0) beq r2, r1, L7 add r1, r1, r3 jmp L3 L7: hlt lw r7, 72(r0) L5: lw r7, 65(r0) lw r6, 71(r0) lw r6, 64(r0) lw r5, 70(r0) lw r5, 63(r0) lw r4, 69(r0) lw r4, 62(r0) lw r3, 68(r0) lw r3, 61(r0) lw r2, 67(r0) lw r2, 60(r0) lw r1, 66(r0) lw r1, 59(r0) beq r2, r1, L6 jmp L4 add r1, r1, r3 L3: sw r1, 51(r0) sw r2, 52(r0) sw r3, 53(r0) sw r4, 54(r0) sw r5, 55(r0) sw r6, 56(r0) sw r7, 57(r0) L4: sw r1, 59(r0) sw r2, 60(r0) sw r3, 61(r0) sw r4, 62(r0) sw r5, 63(r0) sw r6, 64(r0) sw r7, 65(r0) sw r1, 66(r0) sw r2, 67(r0) sw r3, 68(r0) sw r4, 69(r0) sw r5, 70(r0) sw r6, 71(r0) sw r7, 72(r0) lw r4, 51(r0) add r1, r0, r4 addi r2, r0, 49 addi r3, r0, 1 lw r5, 0(r3) lw r6, 0(r4) slt r7, r6, r5 beq r7, r0, L5 sw r5, 50(r0) add r5, r6, r0 sw r5, 0(r3) lw r6, 50(r0) sw r6, 0(r4) addi r1, r0, 0 addi r2, r0, 48 addi r3, r0, 1

  12. Conclusion In Conclusion the Group 6 16-bit RISC Processor Meets and Exceeds Design Requirements!

  13. References • Cache ReferenceComputer Organization & Design The Hardware/Software Interface by Patterson & Hennessy • VHDL Referencehttp://www.ee.ucla.edu/~young/csm152b/vhdl_comp.pdf • Compiler Referencehttp://cs.wisc.edu/~bodik/cs536/Notes/1.Overview.html

  14. Thank You • For viewing this presentation you qualify for Group 6’s special offer • For a limited time only you can get a copy of the Group 6 processor schematic for only $999.95 1-800-BUY-RISC 1-800-289-7472

More Related