blob: 3144acc0461646c4ff03f83b1e8bd442d7f9d39f (
plain)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
|
; ***************************************************************************************
; DEPARTMENT OF ELECTRICAL AND ELECTRONIC ENGINEERING
; IMPERIAL COLLEGE LONDON
;
; EE 3.19: Real Time Digital Signal Processing
; Course by: Dr Paul Mitcheson
;
; LAB 1: Getting Started with the TI C6x DSP
;
; *********** LOAD.ASM ***********
;
; Part of the volume example. Demonstrates that assembly code can be called from C.
; Puts a dummy load on processor which can be varied
; by modiying the value of processingLoad
; (Where ProcessingLoad is a variable passed to the function in C).
; ***************************************************************************************
; Modified by D. Harvey: 24 April 2006
; Modified to include resource references 23 Jan 2008
; ***************************************************************************************
;
;
.global _load
.text
N .set 1000
;
; ********************************* _load description **********************************
; This function simulates a load on the DSP by executing N * processingLoad
; instructions, where processingLoad is the input parameter to load() from volume.c:
;
; load(processingLoad);
;
; The loop is using 8 instructions. One instruction for sub, nop and b, plus nop 5.
; (The extra nop added after sub is to make the number of instructions in the loop 8).
; By dividing N * processingLoad by 8 and using the result as the loop counter,
; N * processingLoad = the number of instruction cycles used when the function is called.
;
; See Real Time Digial signal processing by Nasser Kehtarnavaz (page 146) for more
; info on mixing C and Assembly.
; ****************************************************************************************
;
_load:
mv .s2 a4, b0 ; move parameter (processingLoad) passed from C into b0
[!b0] b .s2 lend ; end code if processingLoad is zero
mvk .s2 N,b1 ; set b1 to value of N (set above as 1000)
mpy .m2 b1,b0,b0 ; N * processingLoad -> b0
nop ; stall the pipeline
shru .s2 b0,3,b0 ; Divide b0 by 8 (b0 will be used as the loop counter)
loop: ; *************** 8 Instruction Loop ******************
sub .d2 b0,1,b0 ; b0 - 1 -> b0
nop ; added to make loop code 8 instruction cycles long
[b0] b .s2 loop ; loop back if b0 is not zero
nop 5 ; stall the pipeline
; ******************* End of Loop *********************
lend: b .s2 b3 ; branch to b3 (register b3 holds the return address)
nop 5 ; stall the pipeline
.end
|
