Merge branch 'master' of github.com:taddeus/peephole

Conflicts:
	src/optimize.py

.gitignore

@@ -2,7 +2,13 @@
 *.pdf
 *.pyc
 *~
+*.aux
+*.log
+*.out
+*.toc
 .coverage
+parser.out
+parsetab.py
 coverage/
 build/
 src/Makefile_old

Makefile

@@ -1,4 +1,5 @@
 BUILD=build/
+CLEAN=src/*.pyc src/optimize/*.pyc
 
 # Fix pdflatex search path
 TGT_DIR :=

QUESTIONS

@@ -1 +0,0 @@
-

TODO CSE.txt

@@ -0,0 +1,8 @@
+Common subexpression elimination
+
+Loop through statements of each block
+    for each binary operator, look back for usage of rs and rt. If rs or rt are
+    assigned, break. If exact same operator is found, add it to the list of
+    common subexpressions. If you reach the end of the block, or rs or rt are
+    assigned, make new destination address, and change each occurence of this
+    expression with a move from the new register address.

benchmarks/.gitignore

@@ -0,0 +1,6 @@
+acron
+clinpack
+dhrystone
+pi
+slalom
+whet

benchmarks/build/hello.s

@@ -6,33 +6,65 @@
  # -mgas -mgpOPT
 
  # Cc1 arguments (-G value = 8, Cpu = default, ISA = 1):
- # -quiet -dumpbase -o
+ # -quiet -dumpbase -O0 -o
 
 gcc2_compiled.:
 __gnu_compiled_c:
+	.sdata
+	.align	2
+$LC2:
+	.ascii	"e: %d\n\000"
+	.align	3
+$LC0:
+	.word	0x00000000		# 2
+	.word	0x40000000
+	.align	3
+$LC1:
+	.word	0x00000000		# 3.5
+	.word	0x400c0000
 	.text
 	.align	2
 	.globl	main
 
 	.text
 
-	.loc	1 2
+	.loc	1 3
 	.ent	main
 main:
-	.frame	$fp,24,$31		# vars= 0, regs= 2/0, args= 16, extra= 0
+	.frame	$fp,64,$31		# vars= 40, regs= 2/0, args= 16, extra= 0
 	.mask	0xc0000000,-4
 	.fmask	0x00000000,0
-	subu	$sp,$sp,24
-	sw	$31,20($sp)
-	sw	$fp,16($sp)
+	subu	$sp,$sp,64
+	sw	$31,60($sp)
+	sw	$fp,56($sp)
 	move	$fp,$sp
 	jal	__main
+	li	$2,0x00000001		# 1
+	sw	$2,16($fp)
+	li	$2,0x00000005		# 5
+	sw	$2,20($fp)
+	lw	$2,16($fp)
+	lw	$3,20($fp)
+	addu	$2,$2,$3
+	sw	$2,24($fp)
+	lw	$2,16($fp)
+	addu	$3,$2,10
+	sw	$3,28($fp)
+	l.d	$f0,$LC0
+	s.d	$f0,32($fp)
+	l.d	$f0,$LC1
+	s.d	$f0,40($fp)
+	li	$2,0x00000061		# 97
+	sb	$2,48($fp)
+	la	$4,$LC2
+	lw	$5,28($fp)
+	jal	printf
 	move	$2,$0
 	j	$L1
 $L1:
 	move	$sp,$fp			# sp not trusted here
-	lw	$31,20($sp)
-	lw	$fp,16($sp)
-	addu	$sp,$sp,24
+	lw	$31,60($sp)
+	lw	$fp,56($sp)
+	addu	$sp,$sp,64
 	j	$31
 	.end	main

benchmarks/hello.c

@@ -1,4 +1,11 @@
-int main(void)
-{
+#include <stdio.h>
+
+int main(void) {
+    int x = 1, b = 5, d = x + b, e = x + 10;
+    double y = 2., z = 3.5;
+    char c = 'a';
+
+	printf("e: %d\n", e);  // 11
+
     return 0;
 }

benchmarks/optimized/.gitignore

@@ -1 +0,0 @@
-*.s

benchmarks/optimized/acron.s

@@ -0,0 +1,413 @@
+	.file	1 "acron.c"
+# GNU C 2.7.2.3 [AL 1.1, MM 40, tma 0.1] SimpleScalar running sstrix compiled by GNU C
+# Cc1 defaults:
+# -mgas -mgpOPT
+# Cc1 arguments (-G value = 8, Cpu = default, ISA = 1):
+# -quiet -dumpbase -O0 -o
+gcc2_compiled.:
+__gnu_compiled_c:
+	.globl	w
+	.data
+	.align	2
+w:
+	.word	$LC0
+	.word	$LC1
+	.word	$LC2
+	.word	$LC3
+	.word	$LC4
+	.word	$LC5
+	.sdata
+	.align	2
+$LC5:
+	.ascii	"Eephole\000"
+	.rdata
+	.align	2
+$LC4:
+	.ascii	"Peephole\000"
+	.align	2
+$LC3:
+	.ascii	"Optimization\000"
+	.align	2
+$LC2:
+	.ascii	"Practicum\000"
+	.align	2
+$LC1:
+	.ascii	"Ertalerbouw\000"
+	.align	2
+$LC0:
+	.ascii	"Vertalerbouw\000"
+	.text
+	.align	2
+	.globl	is_vowel
+	.sdata
+	.align	2
+$LC6:
+	.ascii	"%s\000"
+	.align	2
+$LC7:
+	.ascii	" %s\000"
+	.align	2
+$LC8:
+	.ascii	"\n\000"
+	.text
+	.align	2
+	.globl	do_perm
+	.align	2
+	.globl	main
+	.comm	acron,12
+	.comm	command,100
+	.comm	done,24
+	.comm	pindex,28
+	.text
+	.loc	1 10
+	.ent	is_vowel
+is_vowel:
+	.frame	$fp,16,$31		# vars= 8, regs= 1/0, args= 0, extra= 0
+	.mask	0x40000000,-8
+	.fmask	0x00000000,0
+	subu	$sp,$sp,16
+	sw	$fp,8($sp)
+	move	$fp,$sp
+	move	$3,$4
+	sb	$3,0($fp)
+	move	$2,$0
+	lb	$4,0($fp)
+	li	$5,0x00000041		# 65
+	beq	$4,$5,$L3
+	lb	$4,0($fp)
+	li	$5,0x00000045		# 69
+	beq	$4,$5,$L3
+	lb	$4,0($fp)
+	li	$5,0x00000049		# 73
+	beq	$4,$5,$L3
+	lb	$4,0($fp)
+	li	$5,0x0000004f		# 79
+	beq	$4,$5,$L3
+	lb	$4,0($fp)
+	li	$5,0x00000055		# 85
+	beq	$4,$5,$L3
+	lb	$4,0($fp)
+	li	$5,0x00000059		# 89
+	bne	$4,$5,$L2
+$L3:
+	li	$2,0x00000001		# 1
+$L2:
+	j	$L1
+$L1:
+	move	$sp,$fp			# sp not trusted here
+	lw	$fp,8($sp)
+	addu	$sp,$sp,16
+	j	$31
+	.end	is_vowel
+	.loc	1 15
+	.ent	do_perm
+do_perm:
+	.frame	$fp,56,$31		# vars= 24, regs= 4/0, args= 16, extra= 0
+	.mask	0xc0030000,-4
+	.fmask	0x00000000,0
+	subu	$sp,$sp,56
+	sw	$31,52($sp)
+	sw	$fp,48($sp)
+	sw	$17,44($sp)
+	sw	$16,40($sp)
+	move	$fp,$sp
+	sw	$4,56($fp)
+	sw	$5,60($fp)
+	sw	$6,64($fp)
+	sw	$7,68($fp)
+	sw	$0,24($fp)
+	lw	$2,64($fp)
+	li	$3,0x00000001		# 1
+	bne	$2,$3,$L5
+	lw	$2,pindex
+	move	$3,$2
+	sll	$2,$3,2
+	la	$3,w
+	addu	$2,$2,$3
+	lw	$3,0($2)
+	lb	$4,0($3)
+	jal	is_vowel
+	bne	$2,$0,$L5
+	lw	$2,56($fp)
+	move	$3,$2
+	sll	$2,$3,2
+	la	$3,w
+	addu	$2,$2,$3
+	lw	$3,0($2)
+	lb	$4,0($3)
+	jal	is_vowel
+	beq	$2,$0,$L4
+$L5:
+	lw	$2,64($fp)
+	slt	$3,$2,2
+	bne	$3,$0,$L6
+	lw	$2,64($fp)
+	move	$3,$2
+	sll	$2,$3,2
+	la	$3,pindex-8
+	addu	$2,$2,$3
+	lw	$3,0($2)
+	move	$2,$3
+	sll	$3,$2,2
+	la	$4,w
+	addu	$2,$3,$4
+	lw	$3,0($2)
+	lb	$4,0($3)
+	jal	is_vowel
+	move	$16,$2
+	lw	$2,64($fp)
+	move	$3,$2
+	sll	$2,$3,2
+	la	$3,pindex-4
+	addu	$2,$2,$3
+	lw	$3,0($2)
+	move	$2,$3
+	sll	$3,$2,2
+	la	$4,w
+	addu	$2,$3,$4
+	lw	$3,0($2)
+	lb	$4,0($3)
+	jal	is_vowel
+	move	$17,$2
+	lw	$2,56($fp)
+	move	$3,$2
+	sll	$2,$3,2
+	la	$3,w
+	addu	$2,$2,$3
+	lw	$3,0($2)
+	lb	$4,0($3)
+	jal	is_vowel
+	addu	$3,$16,$17
+	addu	$2,$3,$2
+	sw	$2,24($fp)
+	lw	$2,24($fp)
+	beq	$2,$0,$L8
+	lw	$2,24($fp)
+	li	$3,0x00000003		# 3
+	bne	$2,$3,$L7
+$L8:
+	j	$L4
+$L7:
+$L6:
+	lw	$2,64($fp)
+	addu	$3,$2,1
+	sw	$3,64($fp)
+	sll	$3,$2,2
+	la	$4,pindex
+	addu	$2,$3,$4
+	lw	$3,56($fp)
+	sw	$3,0($2)
+	lw	$2,64($fp)
+	slt	$3,$2,6
+	beq	$3,$0,$L9
+	lw	$3,68($fp)
+	subu	$2,$3,1
+	move	$3,$2
+	sw	$3,68($fp)
+	beq	$3,$0,$L9
+	sw	$0,16($fp)
+$L10:
+	lw	$2,16($fp)
+	slt	$3,$2,6
+	beq	$3,$0,$L11
+$L13:
+	lw	$2,16($fp)
+	move	$3,$2
+	sll	$2,$3,2
+	lw	$3,60($fp)
+	addu	$2,$2,$3
+	lw	$3,0($2)
+	bne	$3,$0,$L14
+	lw	$2,16($fp)
+	move	$3,$2
+	sll	$2,$3,2
+	lw	$3,60($fp)
+	addu	$2,$2,$3
+	li	$3,0x00000001		# 1
+	sw	$3,0($2)
+	lw	$4,16($fp)
+	lw	$5,60($fp)
+	lw	$6,64($fp)
+	lw	$7,68($fp)
+	jal	do_perm
+	lw	$2,16($fp)
+	move	$3,$2
+	sll	$2,$3,2
+	lw	$3,60($fp)
+	addu	$2,$2,$3
+	sw	$0,0($2)
+$L14:
+$L12:
+	lw	$3,16($fp)
+	addu	$2,$3,1
+	move	$3,$2
+	sw	$3,16($fp)
+	j	$L10
+$L11:
+	j	$L15
+$L9:
+	sw	$0,28($fp)
+	sw	$0,20($fp)
+$L16:
+	lw	$2,20($fp)
+	lw	$3,64($fp)
+	slt	$2,$2,$3
+	beq	$2,$0,$L17
+$L19:
+	sw	$0,32($fp)
+$L20:
+	lw	$2,20($fp)
+	move	$3,$2
+	sll	$2,$3,2
+	la	$3,pindex
+	addu	$2,$2,$3
+	lw	$3,0($2)
+	move	$2,$3
+	sll	$3,$2,2
+	la	$4,w
+	addu	$2,$3,$4
+	lw	$3,0($2)
+	lw	$4,32($fp)
+	addu	$2,$3,$4
+	lb	$4,0($2)
+	jal	isupper
+	beq	$2,$0,$L21
+$L22:
+	lw	$2,28($fp)
+	addu	$3,$2,1
+	sw	$3,28($fp)
+	lw	$3,20($fp)
+	move	$4,$3
+	sll	$3,$4,2
+	la	$4,pindex
+	addu	$3,$3,$4
+	lw	$4,0($3)
+	move	$3,$4
+	sll	$4,$3,2
+	la	$5,w
+	addu	$3,$4,$5
+	lw	$4,32($fp)
+	addu	$5,$4,1
+	sw	$5,32($fp)
+	lw	$5,0($3)
+	addu	$3,$4,$5
+	lbu	$4,0($3)
+	sb	$4,acron($2)
+	j	$L20
+$L21:
+$L18:
+	lw	$3,20($fp)
+	addu	$2,$3,1
+	move	$3,$2
+	sw	$3,20($fp)
+	j	$L16
+$L17:
+	lw	$2,28($fp)
+	la	$3,acron
+	addu	$2,$2,$3
+	sb	$0,0($2)
+	la	$4,$LC6
+	la	$5,acron
+	jal	printf
+	sw	$0,20($fp)
+$L23:
+	lw	$2,20($fp)
+	lw	$3,64($fp)
+	slt	$2,$2,$3
+	beq	$2,$0,$L24
+$L26:
+	lw	$2,20($fp)
+	move	$3,$2
+	sll	$2,$3,2
+	la	$3,pindex
+	addu	$2,$2,$3
+	lw	$3,0($2)
+	move	$2,$3
+	sll	$3,$2,2
+	la	$4,w
+	addu	$2,$3,$4
+	la	$4,$LC7
+	lw	$5,0($2)
+	jal	printf
+$L25:
+	lw	$3,20($fp)
+	addu	$2,$3,1
+	move	$3,$2
+	sw	$3,20($fp)
+	j	$L23
+$L24:
+	la	$4,$LC8
+	jal	printf
+$L15:
+$L4:
+	move	$sp,$fp			# sp not trusted here
+	lw	$31,52($sp)
+	lw	$fp,48($sp)
+	lw	$17,44($sp)
+	lw	$16,40($sp)
+	addu	$sp,$sp,56
+	j	$31
+	.end	do_perm
+	.loc	1 53
+	.ent	main
+main:
+	.frame	$fp,32,$31		# vars= 8, regs= 2/0, args= 16, extra= 0
+	.mask	0xc0000000,-4
+	.fmask	0x00000000,0
+	subu	$sp,$sp,32
+	sw	$31,28($sp)
+	sw	$fp,24($sp)
+	move	$fp,$sp
+	jal	__main
+	li	$2,0x00000004		# 4
+	sw	$2,20($fp)
+$L28:
+	lw	$2,20($fp)
+	slt	$3,$2,7
+	beq	$3,$0,$L29
+$L31:
+	sw	$0,16($fp)
+$L32:
+	lw	$2,16($fp)
+	slt	$3,$2,6
+	beq	$3,$0,$L33
+$L35:
+	lw	$2,16($fp)
+	move	$3,$2
+	sll	$2,$3,2
+	la	$3,done
+	addu	$2,$2,$3
+	li	$3,0x00000001		# 1
+	sw	$3,0($2)
+	lw	$4,16($fp)
+	la	$5,done
+	move	$6,$0
+	lw	$7,20($fp)
+	jal	do_perm
+	lw	$2,16($fp)
+	move	$3,$2
+	sll	$2,$3,2
+	la	$3,done
+	addu	$2,$2,$3
+	sw	$0,0($2)
+$L34:
+	lw	$3,16($fp)
+	addu	$2,$3,1
+	move	$3,$2
+	sw	$3,16($fp)
+	j	$L32
+$L33:
+$L30:
+	lw	$3,20($fp)
+	addu	$2,$3,1
+	move	$3,$2
+	sw	$3,20($fp)
+	j	$L28
+$L29:
+$L27:
+	move	$sp,$fp			# sp not trusted here
+	lw	$31,28($sp)
+	lw	$fp,24($sp)
+	addu	$sp,$sp,32
+	j	$31
+	.end	main

benchmarks/optimized/clinpack.s

@@ -0,0 +1,3740 @@
+	.file	1 "clinpack.c"
+# GNU C 2.7.2.3 [AL 1.1, MM 40, tma 0.1] SimpleScalar running sstrix compiled by GNU C
+# Cc1 defaults:
+# -mgas -mgpOPT
+# Cc1 arguments (-G value = 8, Cpu = default, ISA = 1):
+# -quiet -dumpbase -O0 -o
+gcc2_compiled.:
+__gnu_compiled_c:
+	.lcomm	aa.2,320000
+	.lcomm	a.3,321600
+	.lcomm	b.4,1600
+	.lcomm	x.5,1600
+	.lcomm	ipvt.6,800
+	.lcomm	n.7,4
+	.lcomm	i.8,4
+	.lcomm	ntimes.9,4
+	.lcomm	info.10,4
+	.lcomm	lda.11,4
+	.lcomm	ldaa.12,4
+	.lcomm	kflops.13,4
+	.rdata
+	.align	2
+$LC1:
+	.ascii	"Unrolled \000"
+	.sdata
+	.align	2
+$LC2:
+	.ascii	"Double \000"
+	.rdata
+	.align	2
+$LC3:
+	.ascii	"Precision Linpack\n\n\000"
+	.align	2
+$LC6:
+	.ascii	"   norm. resid      resid           machep\000"
+	.align	2
+$LC7:
+	.ascii	"         x[0]-1        x[n-1]-1\n\000"
+	.align	2
+$LC8:
+	.ascii	"%8.1f      %16.8e%16.8e%16.8e%16.8e\n\000"
+	.align	2
+$LC9:
+	.ascii	" times are reported for matrices of order %5d\n\000"
+	.align	2
+$LC10:
+	.ascii	"      dgefa      dgesl      total       kflops     unit\000"
+	.align	2
+$LC11:
+	.ascii	"      ratio\n\000"
+	.align	2
+$LC14:
+	.ascii	" times for array with leading dimension of%5d\n\000"
+	.align	2
+$LC16:
+	.ascii	" times for array with leading dimension of%4d\n\000"
+	.align	2
+$LC17:
+	.ascii	" Precision %5d Kflops ; %d Reps \n\000"
+	.sdata
+	.align	3
+$LC0:
+	.word	0x3126e979		# 0.056000000000000001166
+	.word	0x3facac08
+	.align	3
+$LC4:
+	.word	0x00000000		# 3
+	.word	0x40080000
+	.align	3
+$LC5:
+	.word	0x00000000		# 1
+	.word	0x3ff00000
+	.align	3
+$LC12:
+	.word	0x00000000		# 1000
+	.word	0x408f4000
+	.align	3
+$LC13:
+	.word	0x00000000		# 2000
+	.word	0x409f4000
+	.align	3
+$LC15:
+	.word	0x00000000		# 0.5
+	.word	0x3fe00000
+	.text
+	.align	2
+	.globl	main
+	.rdata
+	.align	2
+$LC18:
+	.ascii	"%11.2f%11.2f%11.2f%11.0f%11.2f%11.2f\n\000"
+	.text
+	.align	2
+	.globl	print_time
+	.sdata
+	.align	3
+$LC19:
+	.word	0x00000000		# 32768
+	.word	0x40e00000
+	.align	3
+$LC20:
+	.word	0x00000000		# 16384
+	.word	0x40d00000
+	.text
+	.align	2
+	.globl	matgen
+	.sdata
+	.align	3
+$LC21:
+	.word	0x00000000		# -1
+	.word	0xbff00000
+	.text
+	.align	2
+	.globl	dgefa
+	.align	2
+	.globl	dgesl
+	.align	2
+	.globl	daxpy
+	.align	2
+	.globl	ddot
+	.align	2
+	.globl	dscal
+	.align	2
+	.globl	idamax
+	.sdata
+	.align	3
+$LC22:
+	.word	0x55555555		# 1.3333333333333332593
+	.word	0x3ff55555
+	.align	3
+$LC23:
+	.word	0x00000000		# 1
+	.word	0x3ff00000
+	.text
+	.align	2
+	.globl	epslon
+	.align	2
+	.globl	dmxpy
+	.sdata
+	.align	3
+$LC24:
+	.word	0xa0b5ed8d		# 9.9999999999999995475e-07
+	.word	0x3eb0c6f7
+	.text
+	.align	2
+	.globl	dtime
+	.lcomm	st,384
+	.comm	rusage,72
+	.text
+	.loc	1 72
+	.ent	main
+main:
+	.frame	$fp,168,$31		# vars= 96, regs= 2/2, args= 48, extra= 0
+	.mask	0xc0000000,-20
+	.fmask	0x00f00000,-4
+	subu	$sp,$sp,168
+	sw	$31,148($sp)
+	sw	$fp,144($sp)
+	s.d	$f22,160($sp)
+	s.d	$f20,152($sp)
+	move	$fp,$sp
+	jal	__main
+	li	$2,0x000000c9		# 201
+	sw	$2,lda.11
+	li	$2,0x000000c8		# 200
+	sw	$2,ldaa.12
+	l.d	$f0,$LC0
+	s.d	$f0,48($fp)
+	li	$2,0x00000019		# 25
+	sw	$2,n.7
+	la	$4,$LC1
+	jal	printf
+	la	$4,$LC2
+	jal	printf
+	la	$4,$LC3
+	jal	printf
+	lw	$2,n.7
+	lw	$3,n.7
+	mult	$2,$3
+	mflo	$2
+	lw	$3,n.7
+	mult	$2,$3
+	mflo	$2
+	mtc1	$2,$f20
+	#nop
+	cvt.d.w	$f20,$f20
+	add.d	$f0,$f20,$f20
+	l.d	$f2,$LC4
+	div.d	$f0,$f0,$f2
+	lw	$2,n.7
+	lw	$3,n.7
+	mult	$2,$3
+	mflo	$2
+	mtc1	$2,$f22
+	#nop
+	cvt.d.w	$f22,$f22
+	add.d	$f2,$f22,$f22
+	add.d	$f0,$f0,$f2
+	s.d	$f0,56($fp)
+	addu	$2,$fp,72
+	sw	$2,16($sp)
+	la	$4,a.3
+	lw	$5,lda.11
+	lw	$6,n.7
+	la	$7,b.4
+	jal	matgen
+	jal	dtime
+	s.d	$f0,120($fp)
+	la	$2,info.10
+	sw	$2,16($sp)
+	la	$4,a.3
+	lw	$5,lda.11
+	lw	$6,n.7
+	la	$7,ipvt.6
+	jal	dgefa
+	jal	dtime
+	l.d	$f2,120($fp)
+	sub.d	$f0,$f0,$f2
+	s.d	$f0,st
+	jal	dtime
+	s.d	$f0,120($fp)
+	la	$2,b.4
+	sw	$2,16($sp)
+	sw	$0,20($sp)
+	la	$4,a.3
+	lw	$5,lda.11
+	lw	$6,n.7
+	la	$7,ipvt.6
+	jal	dgesl
+	jal	dtime
+	l.d	$f2,120($fp)
+	sub.d	$f0,$f0,$f2
+	s.d	$f0,st+48
+	l.d	$f0,st
+	l.d	$f2,st+48
+	add.d	$f0,$f0,$f2
+	s.d	$f0,64($fp)
+	sw	$0,i.8
+$L2:
+	lw	$2,i.8
+	lw	$3,n.7
+	slt	$2,$2,$3
+	bne	$2,$0,$L5
+	j	$L3
+$L5:
+	lw	$2,i.8
+	move	$3,$2
+	sll	$2,$3,3
+	la	$3,x.5
+	addu	$2,$2,$3
+	lw	$3,i.8
+	move	$4,$3
+	sll	$3,$4,3
+	la	$4,b.4
+	addu	$3,$3,$4
+	l.d	$f0,0($3)
+	s.d	$f0,0($2)
+$L4:
+	lw	$3,i.8
+	addu	$2,$3,1
+	move	$3,$2
+	sw	$3,i.8
+	j	$L2
+$L3:
+	addu	$2,$fp,72
+	sw	$2,16($sp)
+	la	$4,a.3
+	lw	$5,lda.11
+	lw	$6,n.7
+	la	$7,b.4
+	jal	matgen
+	sw	$0,i.8
+$L6:
+	lw	$2,i.8
+	lw	$3,n.7
+	slt	$2,$2,$3
+	bne	$2,$0,$L9
+	j	$L7
+$L9:
+	lw	$2,i.8
+	move	$3,$2
+	sll	$2,$3,3
+	la	$3,b.4
+	addu	$2,$2,$3
+	lw	$3,i.8
+	move	$4,$3
+	sll	$3,$4,3
+	la	$4,b.4
+	addu	$3,$3,$4
+	l.d	$f0,0($3)
+	neg.d	$f2,$f0
+	s.d	$f2,0($2)
+$L8:
+	lw	$3,i.8
+	addu	$2,$3,1
+	move	$3,$2
+	sw	$3,i.8
+	j	$L6
+$L7:
+	la	$2,x.5
+	sw	$2,16($sp)
+	la	$2,a.3
+	sw	$2,20($sp)
+	lw	$4,n.7
+	la	$5,b.4
+	lw	$6,n.7
+	lw	$7,lda.11
+	jal	dmxpy
+	dsz	88($fp)
+	dsz	80($fp)
+	sw	$0,i.8
+$L10:
+	lw	$2,i.8
+	lw	$3,n.7
+	slt	$2,$2,$3
+	bne	$2,$0,$L13
+	j	$L11
+$L13:
+	lw	$2,i.8
+	move	$3,$2
+	sll	$2,$3,3
+	la	$3,b.4
+	addu	$2,$2,$3
+	l.d	$f2,0($2)
+	abs.d	$f0,$f2
+	l.d	$f2,88($fp)
+	c.lt.d	$f0,$f2
+	bc1f	$L14
+	l.d	$f0,88($fp)
+	j	$L15
+$L14:
+	lw	$2,i.8
+	move	$3,$2
+	sll	$2,$3,3
+	la	$3,b.4
+	addu	$2,$2,$3
+	l.d	$f2,0($2)
+	abs.d	$f0,$f2
+$L15:
+	s.d	$f0,88($fp)
+	lw	$2,i.8
+	move	$3,$2
+	sll	$2,$3,3
+	la	$3,x.5
+	addu	$2,$2,$3
+	l.d	$f2,0($2)
+	abs.d	$f0,$f2
+	l.d	$f2,80($fp)
+	c.lt.d	$f0,$f2
+	bc1f	$L16
+	l.d	$f0,80($fp)
+	j	$L17
+$L16:
+	lw	$2,i.8
+	move	$3,$2
+	sll	$2,$3,3
+	la	$3,x.5
+	addu	$2,$2,$3
+	l.d	$f2,0($2)
+	abs.d	$f0,$f2
+$L17:
+	s.d	$f0,80($fp)
+$L12:
+	lw	$3,i.8
+	addu	$2,$3,1
+	move	$3,$2
+	sw	$3,i.8
+	j	$L10
+$L11:
+	l.d	$f12,$LC5
+	jal	epslon
+	s.d	$f0,104($fp)
+	l.s	$f0,n.7
+	#nop
+	cvt.d.w	$f0,$f0
+	l.d	$f2,72($fp)
+	mul.d	$f0,$f0,$f2
+	l.d	$f2,80($fp)
+	mul.d	$f0,$f0,$f2
+	l.d	$f2,104($fp)
+	mul.d	$f0,$f0,$f2
+	l.d	$f2,88($fp)
+	div.d	$f0,$f2,$f0
+	s.d	$f0,96($fp)
+	la	$4,$LC6
+	jal	printf
+	la	$4,$LC7
+	jal	printf
+	l.d	$f0,88($fp)
+	s.d	$f0,16($sp)
+	l.d	$f0,104($fp)
+	s.d	$f0,24($sp)
+	l.d	$f0,x.5
+	l.d	$f2,$LC5
+	sub.d	$f0,$f0,$f2
+	s.d	$f0,32($sp)
+	lw	$2,n.7
+	move	$3,$2
+	sll	$2,$3,3
+	addu	$2,$2,$3
+	l.d	$f0,0($2)
+	l.d	$f2,$LC5
+	sub.d	$f0,$f0,$f2
+	s.d	$f0,40($sp)
+	la	$4,$LC8
+	dlw	$6,96($fp)
+	jal	printf
+	la	$4,$LC9
+	lw	$5,n.7
+	jal	printf
+	la	$4,$LC10
+	jal	printf
+	la	$4,$LC11
+	jal	printf
+	l.d	$f0,64($fp)
+	s.d	$f0,st+96
+	l.d	$f0,64($fp)
+	l.d	$f2,$LC12
+	mul.d	$f0,$f0,$f2
+	l.d	$f2,56($fp)
+	div.d	$f0,$f2,$f0
+	s.d	$f0,st+144
+	l.d	$f0,$LC13
+	l.d	$f2,st+144
+	div.d	$f0,$f0,$f2
+	s.d	$f0,st+192
+	l.d	$f0,64($fp)
+	l.d	$f2,48($fp)
+	div.d	$f0,$f0,$f2
+	s.d	$f0,st+240
+	la	$4,$LC14
+	lw	$5,lda.11
+	jal	printf
+	move	$4,$0
+	jal	print_time
+	addu	$2,$fp,72
+	sw	$2,16($sp)
+	la	$4,a.3
+	lw	$5,lda.11
+	lw	$6,n.7
+	la	$7,b.4
+	jal	matgen
+	jal	dtime
+	s.d	$f0,120($fp)
+	la	$2,info.10
+	sw	$2,16($sp)
+	la	$4,a.3
+	lw	$5,lda.11
+	lw	$6,n.7
+	la	$7,ipvt.6
+	jal	dgefa
+	jal	dtime
+	l.d	$f2,120($fp)
+	sub.d	$f0,$f0,$f2
+	s.d	$f0,st+8
+	jal	dtime
+	s.d	$f0,120($fp)
+	la	$2,b.4
+	sw	$2,16($sp)
+	sw	$0,20($sp)
+	la	$4,a.3
+	lw	$5,lda.11
+	lw	$6,n.7
+	la	$7,ipvt.6
+	jal	dgesl
+	jal	dtime
+	l.d	$f2,120($fp)
+	sub.d	$f0,$f0,$f2
+	s.d	$f0,st+56
+	l.d	$f0,st+8
+	l.d	$f2,st+56
+	add.d	$f0,$f0,$f2
+	s.d	$f0,64($fp)
+	l.d	$f0,64($fp)
+	s.d	$f0,st+104
+	l.d	$f0,64($fp)
+	l.d	$f2,$LC12
+	mul.d	$f0,$f0,$f2
+	l.d	$f2,56($fp)
+	div.d	$f0,$f2,$f0
+	s.d	$f0,st+152
+	l.d	$f0,$LC13
+	l.d	$f2,st+152
+	div.d	$f0,$f0,$f2
+	s.d	$f0,st+200
+	l.d	$f0,64($fp)
+	l.d	$f2,48($fp)
+	div.d	$f0,$f0,$f2
+	s.d	$f0,st+248
+	addu	$2,$fp,72
+	sw	$2,16($sp)
+	la	$4,a.3
+	lw	$5,lda.11
+	lw	$6,n.7
+	la	$7,b.4
+	jal	matgen
+	jal	dtime
+	s.d	$f0,120($fp)
+	la	$2,info.10
+	sw	$2,16($sp)
+	la	$4,a.3
+	lw	$5,lda.11
+	lw	$6,n.7
+	la	$7,ipvt.6
+	jal	dgefa
+	jal	dtime
+	l.d	$f2,120($fp)
+	sub.d	$f0,$f0,$f2
+	s.d	$f0,st+16
+	jal	dtime
+	s.d	$f0,120($fp)
+	la	$2,b.4
+	sw	$2,16($sp)
+	sw	$0,20($sp)
+	la	$4,a.3
+	lw	$5,lda.11
+	lw	$6,n.7
+	la	$7,ipvt.6
+	jal	dgesl
+	jal	dtime
+	l.d	$f2,120($fp)
+	sub.d	$f0,$f0,$f2
+	s.d	$f0,st+64
+	l.d	$f0,st+16
+	l.d	$f2,st+64
+	add.d	$f0,$f0,$f2
+	s.d	$f0,64($fp)
+	l.d	$f0,64($fp)
+	s.d	$f0,st+112
+	l.d	$f0,64($fp)
+	l.d	$f2,$LC12
+	mul.d	$f0,$f0,$f2
+	l.d	$f2,56($fp)
+	div.d	$f0,$f2,$f0
+	s.d	$f0,st+160
+	l.d	$f0,$LC13
+	l.d	$f2,st+160
+	div.d	$f0,$f0,$f2
+	s.d	$f0,st+208
+	l.d	$f0,64($fp)
+	l.d	$f2,48($fp)
+	div.d	$f0,$f0,$f2
+	s.d	$f0,st+256
+	li	$2,0x00000001		# 1
+	sw	$2,ntimes.9
+	dsz	136($fp)
+	jal	dtime
+	s.d	$f0,120($fp)
+	sw	$0,i.8
+$L18:
+	lw	$2,i.8
+	lw	$3,ntimes.9
+	slt	$2,$2,$3
+	bne	$2,$0,$L21
+	j	$L19
+$L21:
+	jal	dtime
+	s.d	$f0,128($fp)
+	addu	$2,$fp,72
+	sw	$2,16($sp)
+	la	$4,a.3
+	lw	$5,lda.11
+	lw	$6,n.7
+	la	$7,b.4
+	jal	matgen
+	jal	dtime
+	l.d	$f2,136($fp)
+	add.d	$f0,$f0,$f2
+	l.d	$f2,128($fp)
+	sub.d	$f0,$f0,$f2
+	s.d	$f0,136($fp)
+	la	$2,info.10
+	sw	$2,16($sp)
+	la	$4,a.3
+	lw	$5,lda.11
+	lw	$6,n.7
+	la	$7,ipvt.6
+	jal	dgefa
+$L20:
+	lw	$3,i.8
+	addu	$2,$3,1
+	move	$3,$2
+	sw	$3,i.8
+	j	$L18
+$L19:
+	jal	dtime
+	l.d	$f2,120($fp)
+	sub.d	$f0,$f0,$f2
+	l.d	$f2,136($fp)
+	sub.d	$f0,$f0,$f2
+	l.s	$f2,ntimes.9
+	#nop
+	cvt.d.w	$f2,$f2
+	div.d	$f0,$f0,$f2
+	s.d	$f0,st+24
+	jal	dtime
+	s.d	$f0,120($fp)
+	sw	$0,i.8
+$L22:
+	lw	$2,i.8
+	lw	$3,ntimes.9
+	slt	$2,$2,$3
+	bne	$2,$0,$L25
+	j	$L23
+$L25:
+	la	$2,b.4
+	sw	$2,16($sp)
+	sw	$0,20($sp)
+	la	$4,a.3
+	lw	$5,lda.11
+	lw	$6,n.7
+	la	$7,ipvt.6
+	jal	dgesl
+$L24:
+	lw	$3,i.8
+	addu	$2,$3,1
+	move	$3,$2
+	sw	$3,i.8
+	j	$L22
+$L23:
+	jal	dtime
+	l.d	$f2,120($fp)
+	sub.d	$f0,$f0,$f2
+	l.s	$f2,ntimes.9
+	#nop
+	cvt.d.w	$f2,$f2
+	div.d	$f0,$f0,$f2
+	s.d	$f0,st+72
+	l.d	$f0,st+24
+	l.d	$f2,st+72
+	add.d	$f0,$f0,$f2
+	s.d	$f0,64($fp)
+	l.d	$f0,64($fp)
+	s.d	$f0,st+120
+	l.d	$f0,64($fp)
+	l.d	$f2,$LC12
+	mul.d	$f0,$f0,$f2
+	l.d	$f2,56($fp)
+	div.d	$f0,$f2,$f0
+	s.d	$f0,st+168
+	l.d	$f0,$LC13
+	l.d	$f2,st+168
+	div.d	$f0,$f0,$f2
+	s.d	$f0,st+216
+	l.d	$f0,64($fp)
+	l.d	$f2,48($fp)
+	div.d	$f0,$f0,$f2
+	s.d	$f0,st+264
+	li	$4,0x00000001		# 1
+	jal	print_time
+	li	$4,0x00000002		# 2
+	jal	print_time
+	li	$4,0x00000003		# 3
+	jal	print_time
+	addu	$2,$fp,72
+	sw	$2,16($sp)
+	la	$4,aa.2
+	lw	$5,ldaa.12
+	lw	$6,n.7
+	la	$7,b.4
+	jal	matgen
+	jal	dtime
+	s.d	$f0,120($fp)
+	la	$2,info.10
+	sw	$2,16($sp)
+	la	$4,aa.2
+	lw	$5,ldaa.12
+	lw	$6,n.7
+	la	$7,ipvt.6
+	jal	dgefa
+	jal	dtime
+	l.d	$f2,120($fp)
+	sub.d	$f0,$f0,$f2
+	s.d	$f0,st+32
+	jal	dtime
+	s.d	$f0,120($fp)
+	la	$2,b.4
+	sw	$2,16($sp)
+	sw	$0,20($sp)
+	la	$4,aa.2
+	lw	$5,ldaa.12
+	lw	$6,n.7
+	la	$7,ipvt.6
+	jal	dgesl
+	jal	dtime
+	l.d	$f2,120($fp)
+	sub.d	$f0,$f0,$f2
+	s.d	$f0,st+80
+	l.d	$f0,st+32
+	l.d	$f2,st+80
+	add.d	$f0,$f0,$f2
+	s.d	$f0,64($fp)
+	l.d	$f0,64($fp)
+	s.d	$f0,st+128
+	l.d	$f0,64($fp)
+	l.d	$f2,$LC12
+	mul.d	$f0,$f0,$f2
+	l.d	$f2,56($fp)
+	div.d	$f0,$f2,$f0
+	s.d	$f0,st+176
+	l.d	$f0,$LC13
+	l.d	$f2,st+176
+	div.d	$f0,$f0,$f2
+	s.d	$f0,st+224
+	l.d	$f0,64($fp)
+	l.d	$f2,48($fp)
+	div.d	$f0,$f0,$f2
+	s.d	$f0,st+272
+	addu	$2,$fp,72
+	sw	$2,16($sp)
+	la	$4,aa.2
+	lw	$5,ldaa.12
+	lw	$6,n.7
+	la	$7,b.4
+	jal	matgen
+	jal	dtime
+	s.d	$f0,120($fp)
+	la	$2,info.10
+	sw	$2,16($sp)
+	la	$4,aa.2
+	lw	$5,ldaa.12
+	lw	$6,n.7
+	la	$7,ipvt.6
+	jal	dgefa
+	jal	dtime
+	l.d	$f2,120($fp)
+	sub.d	$f0,$f0,$f2
+	s.d	$f0,st+40
+	jal	dtime
+	s.d	$f0,120($fp)
+	la	$2,b.4
+	sw	$2,16($sp)
+	sw	$0,20($sp)
+	la	$4,aa.2
+	lw	$5,ldaa.12
+	lw	$6,n.7
+	la	$7,ipvt.6
+	jal	dgesl
+	jal	dtime
+	l.d	$f2,120($fp)
+	sub.d	$f0,$f0,$f2
+	s.d	$f0,st+88
+	l.d	$f0,st+40
+	l.d	$f2,st+88
+	add.d	$f0,$f0,$f2
+	s.d	$f0,64($fp)
+	l.d	$f0,64($fp)
+	s.d	$f0,st+136
+	l.d	$f0,64($fp)
+	l.d	$f2,$LC12
+	mul.d	$f0,$f0,$f2
+	l.d	$f2,56($fp)
+	div.d	$f0,$f2,$f0
+	s.d	$f0,st+184
+	l.d	$f0,$LC13
+	l.d	$f2,st+184
+	div.d	$f0,$f0,$f2
+	s.d	$f0,st+232
+	l.d	$f0,64($fp)
+	l.d	$f2,48($fp)
+	div.d	$f0,$f0,$f2
+	s.d	$f0,st+280
+	addu	$2,$fp,72
+	sw	$2,16($sp)
+	la	$4,aa.2
+	lw	$5,ldaa.12
+	lw	$6,n.7
+	la	$7,b.4
+	jal	matgen
+	jal	dtime
+	s.d	$f0,120($fp)
+	la	$2,info.10
+	sw	$2,16($sp)
+	la	$4,aa.2
+	lw	$5,ldaa.12
+	lw	$6,n.7
+	la	$7,ipvt.6
+	jal	dgefa
+	jal	dtime
+	l.d	$f2,120($fp)
+	sub.d	$f0,$f0,$f2
+	s.d	$f0,st+48
+	jal	dtime
+	s.d	$f0,120($fp)
+	la	$2,b.4
+	sw	$2,16($sp)
+	sw	$0,20($sp)
+	la	$4,aa.2
+	lw	$5,ldaa.12
+	lw	$6,n.7
+	la	$7,ipvt.6
+	jal	dgesl
+	jal	dtime
+	l.d	$f2,120($fp)
+	sub.d	$f0,$f0,$f2
+	s.d	$f0,st+96
+	l.d	$f0,st+48
+	l.d	$f2,st+96
+	add.d	$f0,$f0,$f2
+	s.d	$f0,64($fp)
+	l.d	$f0,64($fp)
+	s.d	$f0,st+144
+	l.d	$f0,64($fp)
+	l.d	$f2,$LC12
+	mul.d	$f0,$f0,$f2
+	l.d	$f2,56($fp)
+	div.d	$f0,$f2,$f0
+	s.d	$f0,st+192
+	l.d	$f0,$LC13
+	l.d	$f2,st+192
+	div.d	$f0,$f0,$f2
+	s.d	$f0,st+240
+	l.d	$f0,64($fp)
+	l.d	$f2,48($fp)
+	div.d	$f0,$f0,$f2
+	s.d	$f0,st+288
+	li	$2,0x00000001		# 1
+	sw	$2,ntimes.9
+	dsz	136($fp)
+	jal	dtime
+	s.d	$f0,120($fp)
+	sw	$0,i.8
+$L26:
+	lw	$2,i.8
+	lw	$3,ntimes.9
+	slt	$2,$2,$3
+	bne	$2,$0,$L29
+	j	$L27
+$L29:
+	jal	dtime
+	s.d	$f0,128($fp)
+	addu	$2,$fp,72
+	sw	$2,16($sp)
+	la	$4,aa.2
+	lw	$5,ldaa.12
+	lw	$6,n.7
+	la	$7,b.4
+	jal	matgen
+	jal	dtime
+	l.d	$f2,136($fp)
+	add.d	$f0,$f0,$f2
+	l.d	$f2,128($fp)
+	sub.d	$f0,$f0,$f2
+	s.d	$f0,136($fp)
+	la	$2,info.10
+	sw	$2,16($sp)
+	la	$4,aa.2
+	lw	$5,ldaa.12
+	lw	$6,n.7
+	la	$7,ipvt.6
+	jal	dgefa
+$L28:
+	lw	$3,i.8
+	addu	$2,$3,1
+	move	$3,$2
+	sw	$3,i.8
+	j	$L26
+$L27:
+	jal	dtime
+	l.d	$f2,120($fp)
+	sub.d	$f0,$f0,$f2
+	l.d	$f2,136($fp)
+	sub.d	$f0,$f0,$f2
+	l.s	$f2,ntimes.9
+	#nop
+	cvt.d.w	$f2,$f2
+	div.d	$f0,$f0,$f2
+	s.d	$f0,st+56
+	jal	dtime
+	s.d	$f0,120($fp)
+	sw	$0,i.8
+$L30:
+	lw	$2,i.8
+	lw	$3,ntimes.9
+	slt	$2,$2,$3
+	bne	$2,$0,$L33
+	j	$L31
+$L33:
+	la	$2,b.4
+	sw	$2,16($sp)
+	sw	$0,20($sp)
+	la	$4,aa.2
+	lw	$5,ldaa.12
+	lw	$6,n.7
+	la	$7,ipvt.6
+	jal	dgesl
+$L32:
+	lw	$3,i.8
+	addu	$2,$3,1
+	move	$3,$2
+	sw	$3,i.8
+	j	$L30
+$L31:
+	jal	dtime
+	l.d	$f2,120($fp)
+	sub.d	$f0,$f0,$f2
+	l.s	$f2,ntimes.9
+	#nop
+	cvt.d.w	$f2,$f2
+	div.d	$f0,$f0,$f2
+	s.d	$f0,st+104
+	l.d	$f0,st+56
+	l.d	$f2,st+104
+	add.d	$f0,$f0,$f2
+	s.d	$f0,64($fp)
+	l.d	$f0,64($fp)
+	s.d	$f0,st+152
+	l.d	$f0,64($fp)
+	l.d	$f2,$LC12
+	mul.d	$f0,$f0,$f2
+	l.d	$f2,56($fp)
+	div.d	$f0,$f2,$f0
+	s.d	$f0,st+200
+	l.d	$f0,$LC13
+	l.d	$f2,st+200
+	div.d	$f0,$f0,$f2
+	s.d	$f0,st+248
+	l.d	$f0,64($fp)
+	l.d	$f2,48($fp)
+	div.d	$f0,$f0,$f2
+	s.d	$f0,st+296
+	l.d	$f0,st+168
+	mtc1	$0,$f2
+	mtc1	$0,$f3
+	c.lt.d	$f0,$f2
+	bc1f	$L34
+	dsz	st+168
+$L34:
+	l.d	$f0,st+200
+	mtc1	$0,$f2
+	mtc1	$0,$f3
+	c.lt.d	$f0,$f2
+	bc1f	$L35
+	dsz	st+200
+$L35:
+	l.d	$f0,st+168
+	s.d	$f0,112($fp)
+	l.d	$f0,st+200
+	l.d	$f2,st+168
+	c.lt.d	$f0,$f2
+	bc1f	$L36
+	l.d	$f0,st+200
+	s.d	$f0,112($fp)
+$L36:
+	l.d	$f0,112($fp)
+	l.d	$f2,$LC15
+	add.d	$f0,$f0,$f2
+	trunc.w.d $f4,$f0,$8
+	s.s	$f4,kflops.13
+	la	$4,$LC16
+	lw	$5,ldaa.12
+	jal	printf
+	li	$4,0x00000004		# 4
+	jal	print_time
+	li	$4,0x00000005		# 5
+	jal	print_time
+	li	$4,0x00000006		# 6
+	jal	print_time
+	li	$4,0x00000007		# 7
+	jal	print_time
+	la	$4,$LC1
+	jal	printf
+	la	$4,$LC2
+	jal	printf
+	la	$4,$LC17
+	lw	$5,kflops.13
+	li	$6,0x00000001		# 1
+	jal	printf
+$L1:
+	move	$sp,$fp			# sp not trusted here
+	lw	$31,148($sp)
+	lw	$fp,144($sp)
+	l.d	$f22,160($sp)
+	l.d	$f20,152($sp)
+	addu	$sp,$sp,168
+	j	$31
+	.end	main
+	.loc	1 300
+	.ent	print_time
+print_time:
+	.frame	$fp,64,$31		# vars= 0, regs= 2/0, args= 56, extra= 0
+	.mask	0xc0000000,-4
+	.fmask	0x00000000,0
+	subu	$sp,$sp,64
+	sw	$31,60($sp)
+	sw	$fp,56($sp)
+	move	$fp,$sp
+	sw	$4,64($fp)
+	lw	$2,64($fp)
+	move	$3,$2
+	sll	$2,$3,3
+	la	$3,st
+	addu	$2,$2,$3
+	lw	$3,64($fp)
+	move	$4,$3
+	sll	$3,$4,3
+	la	$4,st+48
+	addu	$3,$3,$4
+	l.d	$f0,0($3)
+	s.d	$f0,16($sp)
+	lw	$3,64($fp)
+	move	$4,$3
+	sll	$3,$4,3
+	la	$4,st+96
+	addu	$3,$3,$4
+	l.d	$f0,0($3)
+	s.d	$f0,24($sp)
+	lw	$3,64($fp)
+	move	$4,$3
+	sll	$3,$4,3
+	la	$4,st+144
+	addu	$3,$3,$4
+	l.d	$f0,0($3)
+	s.d	$f0,32($sp)
+	lw	$3,64($fp)
+	move	$4,$3
+	sll	$3,$4,3
+	la	$4,st+192
+	addu	$3,$3,$4
+	l.d	$f0,0($3)
+	s.d	$f0,40($sp)
+	lw	$3,64($fp)
+	move	$4,$3
+	sll	$3,$4,3
+	la	$4,st+240
+	addu	$3,$3,$4
+	l.d	$f0,0($3)
+	s.d	$f0,48($sp)
+	la	$4,$LC18
+	dlw	$6,0($2)
+	jal	printf
+$L37:
+	move	$sp,$fp			# sp not trusted here
+	lw	$31,60($sp)
+	lw	$fp,56($sp)
+	addu	$sp,$sp,64
+	j	$31
+	.end	print_time
+	.loc	1 309
+	.ent	matgen
+matgen:
+	.frame	$fp,24,$31		# vars= 16, regs= 1/0, args= 0, extra= 0
+	.mask	0x40000000,-8
+	.fmask	0x00000000,0
+	subu	$sp,$sp,24
+	sw	$fp,16($sp)
+	move	$fp,$sp
+	sw	$4,24($fp)
+	sw	$5,28($fp)
+	sw	$6,32($fp)
+	sw	$7,36($fp)
+	li	$2,0x0000052d		# 1325
+	sw	$2,0($fp)
+	lw	$2,40($fp)
+	dsz	0($2)
+	sw	$0,8($fp)
+$L39:
+	lw	$2,8($fp)
+	lw	$3,32($fp)
+	slt	$2,$2,$3
+	bne	$2,$0,$L42
+	j	$L40
+$L42:
+	sw	$0,4($fp)
+$L43:
+	lw	$2,4($fp)
+	lw	$3,32($fp)
+	slt	$2,$2,$3
+	bne	$2,$0,$L46
+	j	$L44
+$L46:
+	lw	$3,0($fp)
+	move	$4,$3
+	sll	$2,$4,1
+	addu	$2,$2,$3
+	sll	$4,$2,6
+	addu	$2,$2,$4
+	sll	$4,$2,2
+	addu	$4,$4,$3
+	sll	$2,$4,2
+	addu	$2,$2,$3
+	move	$3,$2
+	bgez	$3,$L47
+	li	$4,0x0000ffff		# 65535
+	addu	$3,$3,$4
+$L47:
+	sra	$3,$3,16
+	move	$4,$3
+	sll	$3,$4,16
+	subu	$2,$2,$3
+	sw	$2,0($fp)
+	lw	$2,28($fp)
+	lw	$3,8($fp)
+	mult	$2,$3
+	mflo	$2
+	lw	$3,4($fp)
+	addu	$2,$2,$3
+	move	$3,$2
+	sll	$2,$3,3
+	lw	$3,24($fp)
+	addu	$2,$2,$3
+	l.s	$f0,0($fp)
+	#nop
+	cvt.d.w	$f0,$f0
+	l.d	$f2,$LC19
+	sub.d	$f0,$f0,$f2
+	l.d	$f2,$LC20
+	div.d	$f0,$f0,$f2
+	s.d	$f0,0($2)
+	lw	$2,40($fp)
+	lw	$3,28($fp)
+	lw	$4,8($fp)
+	mult	$3,$4
+	mflo	$3
+	lw	$4,4($fp)
+	addu	$3,$3,$4
+	move	$4,$3
+	sll	$3,$4,3
+	lw	$4,24($fp)
+	addu	$3,$3,$4
+	lw	$4,40($fp)
+	l.d	$f0,0($3)
+	l.d	$f2,0($4)
+	c.lt.d	$f2,$f0
+	bc1f	$L48
+	lw	$3,28($fp)
+	lw	$4,8($fp)
+	mult	$3,$4
+	mflo	$3
+	lw	$4,4($fp)
+	addu	$3,$3,$4
+	move	$4,$3
+	sll	$3,$4,3
+	lw	$4,24($fp)
+	addu	$3,$3,$4
+	l.d	$f0,0($3)
+	j	$L49
+$L48:
+	lw	$3,40($fp)
+	l.d	$f0,0($3)
+$L49:
+	s.d	$f0,0($2)
+$L45:
+	lw	$3,4($fp)
+	addu	$2,$3,1
+	move	$3,$2
+	sw	$3,4($fp)
+	j	$L43
+$L44:
+$L41:
+	lw	$3,8($fp)
+	addu	$2,$3,1
+	move	$3,$2
+	sw	$3,8($fp)
+	j	$L39
+$L40:
+	.set	noreorder
+	nop
+	.set	reorder
+	sw	$0,4($fp)
+$L50:
+	lw	$2,4($fp)
+	lw	$3,32($fp)
+	slt	$2,$2,$3
+	bne	$2,$0,$L53
+	j	$L51
+$L53:
+	lw	$2,4($fp)
+	move	$3,$2
+	sll	$2,$3,3
+	lw	$3,36($fp)
+	addu	$2,$2,$3
+	dsz	0($2)
+$L52:
+	lw	$3,4($fp)
+	addu	$2,$3,1
+	move	$3,$2
+	sw	$3,4($fp)
+	j	$L50
+$L51:
+	.set	noreorder
+	nop
+	.set	reorder
+	sw	$0,8($fp)
+$L54:
+	lw	$2,8($fp)
+	lw	$3,32($fp)
+	slt	$2,$2,$3
+	bne	$2,$0,$L57
+	j	$L55
+$L57:
+	sw	$0,4($fp)
+$L58:
+	lw	$2,4($fp)
+	lw	$3,32($fp)
+	slt	$2,$2,$3
+	bne	$2,$0,$L61
+	j	$L59
+$L61:
+	lw	$2,4($fp)
+	move	$3,$2
+	sll	$2,$3,3
+	lw	$3,36($fp)
+	addu	$2,$2,$3
+	lw	$3,4($fp)
+	move	$4,$3
+	sll	$3,$4,3
+	lw	$4,36($fp)
+	addu	$3,$3,$4
+	lw	$4,28($fp)
+	lw	$5,8($fp)
+	mult	$4,$5
+	mflo	$4
+	lw	$5,4($fp)
+	addu	$4,$4,$5
+	move	$5,$4
+	sll	$4,$5,3
+	lw	$5,24($fp)
+	addu	$4,$4,$5
+	l.d	$f0,0($3)
+	l.d	$f2,0($4)
+	add.d	$f0,$f0,$f2
+	s.d	$f0,0($2)
+$L60:
+	lw	$3,4($fp)
+	addu	$2,$3,1
+	move	$3,$2
+	sw	$3,4($fp)
+	j	$L58
+$L59:
+$L56:
+	lw	$3,8($fp)
+	addu	$2,$3,1
+	move	$3,$2
+	sw	$3,8($fp)
+	j	$L54
+$L55:
+$L38:
+	move	$sp,$fp			# sp not trusted here
+	lw	$fp,16($sp)
+	addu	$sp,$sp,24
+	j	$31
+	.end	matgen
+	.loc	1 339
+	.ent	dgefa
+dgefa:
+	.frame	$fp,72,$31		# vars= 32, regs= 2/0, args= 32, extra= 0
+	.mask	0xc0000000,-4
+	.fmask	0x00000000,0
+	subu	$sp,$sp,72
+	sw	$31,68($sp)
+	sw	$fp,64($sp)
+	move	$fp,$sp
+	sw	$4,72($fp)
+	sw	$5,76($fp)
+	sw	$6,80($fp)
+	sw	$7,84($fp)
+	lw	$2,88($fp)
+	sw	$0,0($2)
+	lw	$2,80($fp)
+	subu	$3,$2,1
+	sw	$3,56($fp)
+	lw	$2,56($fp)
+	bltz	$2,$L63
+	sw	$0,44($fp)
+$L64:
+	lw	$2,44($fp)
+	lw	$3,56($fp)
+	slt	$2,$2,$3
+	bne	$2,$0,$L67
+	j	$L65
+$L67:
+	lw	$2,44($fp)
+	addu	$3,$2,1
+	sw	$3,48($fp)
+	lw	$2,80($fp)
+	lw	$3,44($fp)
+	subu	$2,$2,$3
+	lw	$3,76($fp)
+	lw	$4,44($fp)
+	mult	$3,$4
+	mflo	$3
+	lw	$4,44($fp)
+	addu	$3,$3,$4
+	move	$4,$3
+	sll	$3,$4,3
+	lw	$4,72($fp)
+	addu	$3,$3,$4
+	move	$4,$2
+	move	$5,$3
+	li	$6,0x00000001		# 1
+	jal	idamax
+	lw	$3,44($fp)
+	addu	$2,$2,$3
+	sw	$2,52($fp)
+	lw	$2,44($fp)
+	move	$3,$2
+	sll	$2,$3,2
+	lw	$3,84($fp)
+	addu	$2,$2,$3
+	lw	$3,52($fp)
+	sw	$3,0($2)
+	lw	$2,76($fp)
+	lw	$3,44($fp)
+	mult	$2,$3
+	mflo	$2
+	lw	$3,52($fp)
+	addu	$2,$2,$3
+	move	$3,$2
+	sll	$2,$3,3
+	lw	$3,72($fp)
+	addu	$2,$2,$3
+	l.d	$f0,0($2)
+	mtc1	$0,$f2
+	mtc1	$0,$f3
+	c.eq.d	$f0,$f2
+	bc1t	$L68
+	lw	$2,52($fp)
+	lw	$3,44($fp)
+	beq	$2,$3,$L69
+	lw	$2,76($fp)
+	lw	$3,44($fp)
+	mult	$2,$3
+	mflo	$2
+	lw	$3,52($fp)
+	addu	$2,$2,$3
+	move	$3,$2
+	sll	$2,$3,3
+	lw	$3,72($fp)
+	addu	$2,$2,$3
+	l.d	$f0,0($2)
+	s.d	$f0,32($fp)
+	lw	$2,76($fp)
+	lw	$3,44($fp)
+	mult	$2,$3
+	mflo	$2
+	lw	$3,52($fp)
+	addu	$2,$2,$3
+	move	$3,$2
+	sll	$2,$3,3
+	lw	$3,72($fp)
+	addu	$2,$2,$3
+	lw	$3,76($fp)
+	lw	$4,44($fp)
+	mult	$3,$4
+	mflo	$3
+	lw	$4,44($fp)
+	addu	$3,$3,$4
+	move	$4,$3
+	sll	$3,$4,3
+	lw	$4,72($fp)
+	addu	$3,$3,$4
+	l.d	$f0,0($3)
+	s.d	$f0,0($2)
+	lw	$2,76($fp)
+	lw	$3,44($fp)
+	mult	$2,$3
+	mflo	$2
+	lw	$3,44($fp)
+	addu	$2,$2,$3
+	move	$3,$2
+	sll	$2,$3,3
+	lw	$3,72($fp)
+	addu	$2,$2,$3
+	l.d	$f0,32($fp)
+	s.d	$f0,0($2)
+$L69:
+	lw	$2,76($fp)
+	lw	$3,44($fp)
+	mult	$2,$3
+	mflo	$2
+	lw	$3,44($fp)
+	addu	$2,$2,$3
+	move	$3,$2
+	sll	$2,$3,3
+	lw	$3,72($fp)
+	addu	$2,$2,$3
+	l.d	$f0,$LC21
+	l.d	$f2,0($2)
+	div.d	$f0,$f0,$f2
+	s.d	$f0,32($fp)
+	lw	$3,80($fp)
+	subu	$2,$3,1
+	lw	$3,44($fp)
+	subu	$2,$2,$3
+	lw	$3,76($fp)
+	lw	$4,44($fp)
+	mult	$3,$4
+	mflo	$3
+	lw	$4,44($fp)
+	addu	$3,$3,$4
+	move	$4,$3
+	sll	$3,$4,3
+	addu	$4,$3,8
+	lw	$3,72($fp)
+	addu	$4,$3,$4
+	sw	$4,16($sp)
+	li	$3,0x00000001		# 1
+	sw	$3,20($sp)
+	move	$4,$2
+	dlw	$6,32($fp)
+	jal	dscal
+	lw	$2,48($fp)
+	sw	$2,40($fp)
+$L70:
+	lw	$2,40($fp)
+	lw	$3,80($fp)
+	slt	$2,$2,$3
+	bne	$2,$0,$L73
+	j	$L71
+$L73:
+	lw	$2,76($fp)
+	lw	$3,40($fp)
+	mult	$2,$3
+	mflo	$2
+	lw	$3,52($fp)
+	addu	$2,$2,$3
+	move	$3,$2
+	sll	$2,$3,3
+	lw	$3,72($fp)
+	addu	$2,$2,$3
+	l.d	$f0,0($2)
+	s.d	$f0,32($fp)
+	lw	$2,52($fp)
+	lw	$3,44($fp)
+	beq	$2,$3,$L74
+	lw	$2,76($fp)
+	lw	$3,40($fp)
+	mult	$2,$3
+	mflo	$2
+	lw	$3,52($fp)
+	addu	$2,$2,$3
+	move	$3,$2
+	sll	$2,$3,3
+	lw	$3,72($fp)
+	addu	$2,$2,$3
+	lw	$3,76($fp)
+	lw	$4,40($fp)
+	mult	$3,$4
+	mflo	$3
+	lw	$4,44($fp)
+	addu	$3,$3,$4
+	move	$4,$3
+	sll	$3,$4,3
+	lw	$4,72($fp)
+	addu	$3,$3,$4
+	l.d	$f0,0($3)
+	s.d	$f0,0($2)
+	lw	$2,76($fp)
+	lw	$3,40($fp)
+	mult	$2,$3
+	mflo	$2
+	lw	$3,44($fp)
+	addu	$2,$2,$3
+	move	$3,$2
+	sll	$2,$3,3
+	lw	$3,72($fp)
+	addu	$2,$2,$3
+	l.d	$f0,32($fp)
+	s.d	$f0,0($2)
+$L74:
+	lw	$3,80($fp)
+	subu	$2,$3,1
+	lw	$3,44($fp)
+	subu	$2,$2,$3
+	lw	$3,76($fp)
+	lw	$4,44($fp)
+	mult	$3,$4
+	mflo	$3
+	lw	$4,44($fp)
+	addu	$3,$3,$4
+	move	$4,$3
+	sll	$3,$4,3
+	addu	$4,$3,8
+	lw	$3,72($fp)
+	addu	$4,$3,$4
+	sw	$4,16($sp)
+	li	$3,0x00000001		# 1
+	sw	$3,20($sp)
+	lw	$3,76($fp)
+	lw	$4,40($fp)
+	mult	$3,$4
+	mflo	$3
+	lw	$4,44($fp)
+	addu	$3,$3,$4
+	move	$4,$3
+	sll	$3,$4,3
+	addu	$4,$3,8
+	lw	$3,72($fp)
+	addu	$4,$3,$4
+	sw	$4,24($sp)
+	li	$3,0x00000001		# 1
+	sw	$3,28($sp)
+	move	$4,$2
+	dlw	$6,32($fp)
+	jal	daxpy
+$L72:
+	lw	$3,40($fp)
+	addu	$2,$3,1
+	move	$3,$2
+	sw	$3,40($fp)
+	j	$L70
+$L71:
+	j	$L75
+$L68:
+	lw	$2,88($fp)
+	lw	$3,44($fp)
+	sw	$3,0($2)
+$L75:
+$L66:
+	lw	$3,44($fp)
+	addu	$2,$3,1
+	move	$3,$2
+	sw	$3,44($fp)
+	j	$L64
+$L65:
+$L63:
+	lw	$2,80($fp)
+	move	$3,$2
+	sll	$2,$3,2
+	lw	$4,84($fp)
+	addu	$3,$2,$4
+	subu	$2,$3,4
+	lw	$3,80($fp)
+	subu	$4,$3,1
+	sw	$4,0($2)
+	lw	$3,80($fp)
+	subu	$2,$3,1
+	lw	$3,76($fp)
+	mult	$2,$3
+	mflo	$2
+	lw	$3,80($fp)
+	addu	$2,$2,$3
+	move	$3,$2
+	sll	$2,$3,3
+	lw	$3,72($fp)
+	addu	$2,$2,$3
+	subu	$3,$2,8
+	l.d	$f0,0($3)
+	mtc1	$0,$f2
+	mtc1	$0,$f3
+	c.eq.d	$f0,$f2
+	bc1f	$L76
+	lw	$2,88($fp)
+	lw	$3,80($fp)
+	subu	$4,$3,1
+	sw	$4,0($2)
+$L76:
+$L62:
+	move	$sp,$fp			# sp not trusted here
+	lw	$31,68($sp)
+	lw	$fp,64($sp)
+	addu	$sp,$sp,72
+	j	$31
+	.end	dgefa
+	.loc	1 455
+	.ent	dgesl
+dgesl:
+	.frame	$fp,64,$31		# vars= 24, regs= 2/0, args= 32, extra= 0
+	.mask	0xc0000000,-4
+	.fmask	0x00000000,0
+	subu	$sp,$sp,64
+	sw	$31,60($sp)
+	sw	$fp,56($sp)
+	move	$fp,$sp
+	sw	$4,64($fp)
+	sw	$5,68($fp)
+	sw	$6,72($fp)
+	sw	$7,76($fp)
+	lw	$2,72($fp)
+	subu	$3,$2,1
+	sw	$3,52($fp)
+	lw	$2,84($fp)
+	bne	$2,$0,$L78
+	lw	$2,52($fp)
+	blez	$2,$L79
+	sw	$0,40($fp)
+$L80:
+	lw	$2,40($fp)
+	lw	$3,52($fp)
+	slt	$2,$2,$3
+	bne	$2,$0,$L83
+	j	$L81
+$L83:
+	lw	$2,40($fp)
+	move	$3,$2
+	sll	$2,$3,2
+	lw	$3,76($fp)
+	addu	$2,$2,$3
+	lw	$3,0($2)
+	sw	$3,48($fp)
+	lw	$2,48($fp)
+	move	$3,$2
+	sll	$2,$3,3
+	lw	$3,80($fp)
+	addu	$2,$2,$3
+	l.d	$f0,0($2)
+	s.d	$f0,32($fp)
+	lw	$2,48($fp)
+	lw	$3,40($fp)
+	beq	$2,$3,$L84
+	lw	$2,48($fp)
+	move	$3,$2
+	sll	$2,$3,3
+	lw	$3,80($fp)
+	addu	$2,$2,$3
+	lw	$3,40($fp)
+	move	$4,$3
+	sll	$3,$4,3
+	lw	$4,80($fp)
+	addu	$3,$3,$4
+	l.d	$f0,0($3)
+	s.d	$f0,0($2)
+	lw	$2,40($fp)
+	move	$3,$2
+	sll	$2,$3,3
+	lw	$3,80($fp)
+	addu	$2,$2,$3
+	l.d	$f0,32($fp)
+	s.d	$f0,0($2)
+$L84:
+	lw	$3,72($fp)
+	subu	$2,$3,1
+	lw	$3,40($fp)
+	subu	$2,$2,$3
+	lw	$3,68($fp)
+	lw	$4,40($fp)
+	mult	$3,$4
+	mflo	$3
+	lw	$4,40($fp)
+	addu	$3,$3,$4
+	move	$4,$3
+	sll	$3,$4,3
+	addu	$4,$3,8
+	lw	$3,64($fp)
+	addu	$4,$3,$4
+	sw	$4,16($sp)
+	li	$3,0x00000001		# 1
+	sw	$3,20($sp)
+	lw	$3,40($fp)
+	move	$4,$3
+	sll	$3,$4,3
+	addu	$4,$3,8
+	lw	$3,80($fp)
+	addu	$4,$3,$4
+	sw	$4,24($sp)
+	li	$3,0x00000001		# 1
+	sw	$3,28($sp)
+	move	$4,$2
+	dlw	$6,32($fp)
+	jal	daxpy
+$L82:
+	lw	$3,40($fp)
+	addu	$2,$3,1
+	move	$3,$2
+	sw	$3,40($fp)
+	j	$L80
+$L81:
+$L79:
+	.set	noreorder
+	nop
+	.set	reorder
+	sw	$0,44($fp)
+$L85:
+	lw	$2,44($fp)
+	lw	$3,72($fp)
+	slt	$2,$2,$3
+	bne	$2,$0,$L88
+	j	$L86
+$L88:
+	lw	$3,72($fp)
+	subu	$2,$3,1
+	lw	$3,44($fp)
+	subu	$2,$2,$3
+	sw	$2,40($fp)
+	lw	$2,40($fp)
+	move	$3,$2
+	sll	$2,$3,3
+	lw	$3,80($fp)
+	addu	$2,$2,$3
+	lw	$3,40($fp)
+	move	$4,$3
+	sll	$3,$4,3
+	lw	$4,80($fp)
+	addu	$3,$3,$4
+	lw	$4,68($fp)
+	lw	$5,40($fp)
+	mult	$4,$5
+	mflo	$4
+	lw	$5,40($fp)
+	addu	$4,$4,$5
+	move	$5,$4
+	sll	$4,$5,3
+	lw	$5,64($fp)
+	addu	$4,$4,$5
+	l.d	$f0,0($3)
+	l.d	$f2,0($4)
+	div.d	$f0,$f0,$f2
+	s.d	$f0,0($2)
+	lw	$2,40($fp)
+	move	$3,$2
+	sll	$2,$3,3
+	lw	$3,80($fp)
+	addu	$2,$2,$3
+	l.d	$f0,0($2)
+	neg.d	$f2,$f0
+	s.d	$f2,32($fp)
+	lw	$2,68($fp)
+	lw	$3,40($fp)
+	mult	$2,$3
+	mflo	$2
+	move	$3,$2
+	sll	$2,$3,3
+	lw	$3,64($fp)
+	addu	$2,$3,$2
+	sw	$2,16($sp)
+	li	$2,0x00000001		# 1
+	sw	$2,20($sp)
+	lw	$2,80($fp)
+	sw	$2,24($sp)
+	li	$2,0x00000001		# 1
+	sw	$2,28($sp)
+	lw	$4,40($fp)
+	dlw	$6,32($fp)
+	jal	daxpy
+$L87:
+	lw	$3,44($fp)
+	addu	$2,$3,1
+	move	$3,$2
+	sw	$3,44($fp)
+	j	$L85
+$L86:
+	j	$L89
+$L78:
+	sw	$0,40($fp)
+$L90:
+	lw	$2,40($fp)
+	lw	$3,72($fp)
+	slt	$2,$2,$3
+	bne	$2,$0,$L93
+	j	$L91
+$L93:
+	lw	$2,68($fp)
+	lw	$3,40($fp)
+	mult	$2,$3
+	mflo	$2
+	move	$3,$2
+	sll	$2,$3,3
+	lw	$3,64($fp)
+	addu	$2,$2,$3
+	li	$3,0x00000001		# 1
+	sw	$3,16($sp)
+	lw	$4,40($fp)
+	move	$5,$2
+	li	$6,0x00000001		# 1
+	lw	$7,80($fp)
+	jal	ddot
+	s.d	$f0,32($fp)
+	lw	$2,40($fp)
+	move	$3,$2
+	sll	$2,$3,3
+	lw	$3,80($fp)
+	addu	$2,$2,$3
+	lw	$3,40($fp)
+	move	$4,$3
+	sll	$3,$4,3
+	lw	$4,80($fp)
+	addu	$3,$3,$4
+	l.d	$f0,0($3)
+	l.d	$f2,32($fp)
+	sub.d	$f0,$f0,$f2
+	lw	$3,68($fp)
+	lw	$4,40($fp)
+	mult	$3,$4
+	mflo	$3
+	lw	$4,40($fp)
+	addu	$3,$3,$4
+	move	$4,$3
+	sll	$3,$4,3
+	lw	$4,64($fp)
+	addu	$3,$3,$4
+	l.d	$f2,0($3)
+	div.d	$f0,$f0,$f2
+	s.d	$f0,0($2)
+$L92:
+	lw	$3,40($fp)
+	addu	$2,$3,1
+	move	$3,$2
+	sw	$3,40($fp)
+	j	$L90
+$L91:
+	lw	$2,52($fp)
+	blez	$2,$L94
+	li	$2,0x00000001		# 1
+	sw	$2,44($fp)
+$L95:
+	lw	$2,44($fp)
+	lw	$3,52($fp)
+	slt	$2,$2,$3
+	bne	$2,$0,$L98
+	j	$L96
+$L98:
+	lw	$3,72($fp)
+	subu	$2,$3,1
+	lw	$3,44($fp)
+	subu	$2,$2,$3
+	sw	$2,40($fp)
+	lw	$3,72($fp)
+	subu	$2,$3,1
+	lw	$3,40($fp)
+	subu	$2,$2,$3
+	lw	$3,68($fp)
+	lw	$4,40($fp)
+	mult	$3,$4
+	mflo	$3
+	lw	$4,40($fp)
+	addu	$3,$3,$4
+	move	$4,$3
+	sll	$3,$4,3
+	addu	$4,$3,8
+	lw	$5,64($fp)
+	addu	$3,$4,$5
+	lw	$4,40($fp)
+	move	$5,$4
+	sll	$4,$5,3
+	addu	$5,$4,8
+	lw	$4,80($fp)
+	addu	$7,$5,$4
+	li	$4,0x00000001		# 1
+	sw	$4,16($sp)
+	move	$4,$2
+	move	$5,$3
+	li	$6,0x00000001		# 1
+	jal	ddot
+	lw	$2,40($fp)
+	move	$3,$2
+	sll	$2,$3,3
+	lw	$3,80($fp)
+	addu	$2,$2,$3
+	lw	$3,40($fp)
+	move	$4,$3
+	sll	$3,$4,3
+	lw	$4,80($fp)
+	addu	$3,$3,$4
+	l.d	$f2,0($3)
+	add.d	$f0,$f2,$f0
+	s.d	$f0,0($2)
+	lw	$2,40($fp)
+	move	$3,$2
+	sll	$2,$3,2
+	lw	$3,76($fp)
+	addu	$2,$2,$3
+	lw	$3,0($2)
+	sw	$3,48($fp)
+	lw	$2,48($fp)
+	lw	$3,40($fp)
+	beq	$2,$3,$L99
+	lw	$2,48($fp)
+	move	$3,$2
+	sll	$2,$3,3
+	lw	$3,80($fp)
+	addu	$2,$2,$3
+	l.d	$f0,0($2)
+	s.d	$f0,32($fp)
+	lw	$2,48($fp)
+	move	$3,$2
+	sll	$2,$3,3
+	lw	$3,80($fp)
+	addu	$2,$2,$3
+	lw	$3,40($fp)
+	move	$4,$3
+	sll	$3,$4,3
+	lw	$4,80($fp)
+	addu	$3,$3,$4
+	l.d	$f0,0($3)
+	s.d	$f0,0($2)
+	lw	$2,40($fp)
+	move	$3,$2
+	sll	$2,$3,3
+	lw	$3,80($fp)
+	addu	$2,$2,$3
+	l.d	$f0,32($fp)
+	s.d	$f0,0($2)
+$L99:
+$L97:
+	lw	$3,44($fp)
+	addu	$2,$3,1
+	move	$3,$2
+	sw	$3,44($fp)
+	j	$L95
+$L96:
+$L94:
+$L89:
+$L77:
+	move	$sp,$fp			# sp not trusted here
+	lw	$31,60($sp)
+	lw	$fp,56($sp)
+	addu	$sp,$sp,64
+	j	$31
+	.end	dgesl
+	.loc	1 585
+	.ent	daxpy
+daxpy:
+	.frame	$fp,32,$31		# vars= 24, regs= 1/0, args= 0, extra= 0
+	.mask	0x40000000,-8
+	.fmask	0x00000000,0
+	subu	$sp,$sp,32
+	sw	$fp,24($sp)
+	move	$fp,$sp
+	sw	$4,32($fp)
+	dsw	$6,40($fp)
+	lw	$2,32($fp)
+	bgtz	$2,$L101
+	j	$L100
+$L101:
+	l.d	$f0,40($fp)
+	mtc1	$0,$f2
+	mtc1	$0,$f3
+	c.eq.d	$f0,$f2
+	bc1f	$L102
+	j	$L100
+$L102:
+	lw	$2,52($fp)
+	li	$3,0x00000001		# 1
+	bne	$2,$3,$L104
+	lw	$2,60($fp)
+	li	$3,0x00000001		# 1
+	bne	$2,$3,$L104
+	j	$L103
+$L104:
+	li	$2,0x00000001		# 1
+	sw	$2,4($fp)
+	li	$2,0x00000001		# 1
+	sw	$2,8($fp)
+	lw	$2,52($fp)
+	bgez	$2,$L105
+	lw	$3,32($fp)
+	subu	$2,$0,$3
+	addu	$3,$2,1
+	lw	$4,52($fp)
+	mult	$3,$4
+	mflo	$2
+	addu	$3,$2,1
+	sw	$3,4($fp)
+$L105:
+	lw	$2,60($fp)
+	bgez	$2,$L106
+	lw	$3,32($fp)
+	subu	$2,$0,$3
+	addu	$3,$2,1
+	lw	$4,60($fp)
+	mult	$3,$4
+	mflo	$2
+	addu	$3,$2,1
+	sw	$3,8($fp)
+$L106:
+	.set	noreorder
+	nop
+	.set	reorder
+	sw	$0,0($fp)
+$L107:
+	lw	$2,0($fp)
+	lw	$3,32($fp)
+	slt	$2,$2,$3
+	bne	$2,$0,$L110
+	j	$L108
+$L110:
+	lw	$2,8($fp)
+	move	$3,$2
+	sll	$2,$3,3
+	lw	$3,56($fp)
+	addu	$2,$2,$3
+	lw	$3,8($fp)
+	move	$4,$3
+	sll	$3,$4,3
+	lw	$4,56($fp)
+	addu	$3,$3,$4
+	lw	$4,4($fp)
+	move	$5,$4
+	sll	$4,$5,3
+	lw	$5,48($fp)
+	addu	$4,$4,$5
+	l.d	$f0,40($fp)
+	l.d	$f2,0($4)
+	mul.d	$f0,$f0,$f2
+	l.d	$f2,0($3)
+	add.d	$f0,$f2,$f0
+	s.d	$f0,0($2)
+	lw	$2,4($fp)
+	lw	$3,52($fp)
+	addu	$2,$2,$3
+	sw	$2,4($fp)
+	lw	$2,8($fp)
+	lw	$3,60($fp)
+	addu	$2,$2,$3
+	sw	$2,8($fp)
+$L109:
+	lw	$3,0($fp)
+	addu	$2,$3,1
+	move	$3,$2
+	sw	$3,0($fp)
+	j	$L107
+$L108:
+	j	$L100
+$L103:
+	lw	$2,32($fp)
+	move	$3,$2
+	bgez	$3,$L111
+	addu	$3,$3,3
+$L111:
+	sra	$3,$3,2
+	move	$4,$3
+	sll	$3,$4,2
+	subu	$2,$2,$3
+	sw	$2,12($fp)
+	lw	$2,12($fp)
+	beq	$2,$0,$L112
+	sw	$0,0($fp)
+$L113:
+	lw	$2,0($fp)
+	lw	$3,12($fp)
+	slt	$2,$2,$3
+	bne	$2,$0,$L116
+	j	$L114
+$L116:
+	lw	$2,0($fp)
+	move	$3,$2
+	sll	$2,$3,3
+	lw	$3,56($fp)
+	addu	$2,$2,$3
+	lw	$3,0($fp)
+	move	$4,$3
+	sll	$3,$4,3
+	lw	$4,56($fp)
+	addu	$3,$3,$4
+	lw	$4,0($fp)
+	move	$5,$4
+	sll	$4,$5,3
+	lw	$5,48($fp)
+	addu	$4,$4,$5
+	l.d	$f0,40($fp)
+	l.d	$f2,0($4)
+	mul.d	$f0,$f0,$f2
+	l.d	$f2,0($3)
+	add.d	$f0,$f2,$f0
+	s.d	$f0,0($2)
+$L115:
+	lw	$3,0($fp)
+	addu	$2,$3,1
+	move	$3,$2
+	sw	$3,0($fp)
+	j	$L113
+$L114:
+	lw	$2,32($fp)
+	slt	$3,$2,4
+	bne	$3,$0,$L100
+$L117:
+$L112:
+	.set	noreorder
+	nop
+	.set	reorder
+	lw	$2,12($fp)
+	sw	$2,0($fp)
+$L118:
+	lw	$2,0($fp)
+	lw	$3,32($fp)
+	slt	$2,$2,$3
+	bne	$2,$0,$L121
+	j	$L119
+$L121:
+	lw	$2,0($fp)
+	move	$3,$2
+	sll	$2,$3,3
+	lw	$3,56($fp)
+	addu	$2,$2,$3
+	lw	$3,0($fp)
+	move	$4,$3
+	sll	$3,$4,3
+	lw	$4,56($fp)
+	addu	$3,$3,$4
+	lw	$4,0($fp)
+	move	$5,$4
+	sll	$4,$5,3
+	lw	$5,48($fp)
+	addu	$4,$4,$5
+	l.d	$f0,40($fp)
+	l.d	$f2,0($4)
+	mul.d	$f0,$f0,$f2
+	l.d	$f2,0($3)
+	add.d	$f0,$f2,$f0
+	s.d	$f0,0($2)
+	lw	$2,0($fp)
+	move	$3,$2
+	sll	$2,$3,3
+	lw	$4,56($fp)
+	addu	$3,$2,$4
+	addu	$2,$3,8
+	lw	$3,0($fp)
+	move	$4,$3
+	sll	$3,$4,3
+	lw	$5,56($fp)
+	addu	$4,$3,$5
+	addu	$3,$4,8
+	lw	$4,0($fp)
+	move	$5,$4
+	sll	$4,$5,3
+	lw	$5,48($fp)
+	addu	$4,$4,$5
+	addu	$5,$4,8
+	l.d	$f0,40($fp)
+	l.d	$f2,0($5)
+	mul.d	$f0,$f0,$f2
+	l.d	$f2,0($3)
+	add.d	$f0,$f2,$f0
+	s.d	$f0,0($2)
+	lw	$2,0($fp)
+	move	$3,$2
+	sll	$2,$3,3
+	lw	$4,56($fp)
+	addu	$3,$2,$4
+	addu	$2,$3,16
+	lw	$3,0($fp)
+	move	$4,$3
+	sll	$3,$4,3
+	lw	$5,56($fp)
+	addu	$4,$3,$5
+	addu	$3,$4,16
+	lw	$4,0($fp)
+	move	$5,$4
+	sll	$4,$5,3
+	lw	$5,48($fp)
+	addu	$4,$4,$5
+	addu	$5,$4,16
+	l.d	$f0,40($fp)
+	l.d	$f2,0($5)
+	mul.d	$f0,$f0,$f2
+	l.d	$f2,0($3)
+	add.d	$f0,$f2,$f0
+	s.d	$f0,0($2)
+	lw	$2,0($fp)
+	move	$3,$2
+	sll	$2,$3,3
+	lw	$4,56($fp)
+	addu	$3,$2,$4
+	addu	$2,$3,24
+	lw	$3,0($fp)
+	move	$4,$3
+	sll	$3,$4,3
+	lw	$5,56($fp)
+	addu	$4,$3,$5
+	addu	$3,$4,24
+	lw	$4,0($fp)
+	move	$5,$4
+	sll	$4,$5,3
+	lw	$5,48($fp)
+	addu	$4,$4,$5
+	addu	$5,$4,24
+	l.d	$f0,40($fp)
+	l.d	$f2,0($5)
+	mul.d	$f0,$f0,$f2
+	l.d	$f2,0($3)
+	add.d	$f0,$f2,$f0
+	s.d	$f0,0($2)
+$L120:
+	lw	$2,0($fp)
+	addu	$3,$2,4
+	sw	$3,0($fp)
+	j	$L118
+$L119:
+$L100:
+	move	$sp,$fp			# sp not trusted here
+	lw	$fp,24($sp)
+	addu	$sp,$sp,32
+	j	$31
+	.end	daxpy
+	.loc	1 641
+	.ent	ddot
+ddot:
+	.frame	$fp,40,$31		# vars= 32, regs= 1/0, args= 0, extra= 0
+	.mask	0x40000000,-8
+	.fmask	0x00000000,0
+	subu	$sp,$sp,40
+	sw	$fp,32($sp)
+	move	$fp,$sp
+	sw	$4,40($fp)
+	sw	$5,44($fp)
+	sw	$6,48($fp)
+	sw	$7,52($fp)
+	dsz	0($fp)
+	lw	$2,40($fp)
+	bgtz	$2,$L123
+	mtc1	$0,$f0
+	mtc1	$0,$f1
+	j	$L122
+$L123:
+	lw	$2,48($fp)
+	li	$3,0x00000001		# 1
+	bne	$2,$3,$L125
+	lw	$2,56($fp)
+	li	$3,0x00000001		# 1
+	bne	$2,$3,$L125
+	j	$L124
+$L125:
+	sw	$0,12($fp)
+	sw	$0,16($fp)
+	lw	$2,48($fp)
+	bgez	$2,$L126
+	lw	$3,40($fp)
+	subu	$2,$0,$3
+	addu	$3,$2,1
+	lw	$2,48($fp)
+	mult	$3,$2
+	mflo	$3
+	sw	$3,12($fp)
+$L126:
+	lw	$2,56($fp)
+	bgez	$2,$L127
+	lw	$3,40($fp)
+	subu	$2,$0,$3
+	addu	$3,$2,1
+	lw	$2,56($fp)
+	mult	$3,$2
+	mflo	$3
+	sw	$3,16($fp)
+$L127:
+	.set	noreorder
+	nop
+	.set	reorder
+	sw	$0,8($fp)
+$L128:
+	lw	$2,8($fp)
+	lw	$3,40($fp)
+	slt	$2,$2,$3
+	bne	$2,$0,$L131
+	j	$L129
+$L131:
+	lw	$2,12($fp)
+	move	$3,$2
+	sll	$2,$3,3
+	lw	$3,44($fp)
+	addu	$2,$2,$3
+	lw	$3,16($fp)
+	move	$4,$3
+	sll	$3,$4,3
+	lw	$4,52($fp)
+	addu	$3,$3,$4
+	l.d	$f0,0($2)
+	l.d	$f2,0($3)
+	mul.d	$f0,$f0,$f2
+	l.d	$f2,0($fp)
+	add.d	$f0,$f2,$f0
+	s.d	$f0,0($fp)
+	lw	$2,12($fp)
+	lw	$3,48($fp)
+	addu	$2,$2,$3
+	sw	$2,12($fp)
+	lw	$2,16($fp)
+	lw	$3,56($fp)
+	addu	$2,$2,$3
+	sw	$2,16($fp)
+$L130:
+	lw	$3,8($fp)
+	addu	$2,$3,1
+	move	$3,$2
+	sw	$3,8($fp)
+	j	$L128
+$L129:
+	l.d	$f0,0($fp)
+	j	$L122
+$L124:
+	lw	$2,40($fp)
+	li	$6,0x66666667		# 1717986919
+	mult	$2,$6
+	mfhi	$5
+	mflo	$4
+	srl	$6,$5,0
+	move	$7,$0
+	sra	$3,$6,1
+	sra	$4,$2,31
+	subu	$3,$3,$4
+	move	$5,$3
+	sll	$4,$5,2
+	addu	$4,$4,$3
+	subu	$2,$2,$4
+	sw	$2,20($fp)
+	lw	$2,20($fp)
+	beq	$2,$0,$L132
+	sw	$0,8($fp)
+$L133:
+	lw	$2,8($fp)
+	lw	$3,20($fp)
+	slt	$2,$2,$3
+	bne	$2,$0,$L136
+	j	$L134
+$L136:
+	lw	$2,8($fp)
+	move	$3,$2
+	sll	$2,$3,3
+	lw	$3,44($fp)
+	addu	$2,$2,$3
+	lw	$3,8($fp)
+	move	$4,$3
+	sll	$3,$4,3
+	lw	$4,52($fp)
+	addu	$3,$3,$4
+	l.d	$f0,0($2)
+	l.d	$f2,0($3)
+	mul.d	$f0,$f0,$f2
+	l.d	$f2,0($fp)
+	add.d	$f0,$f2,$f0
+	s.d	$f0,0($fp)
+$L135:
+	lw	$3,8($fp)
+	addu	$2,$3,1
+	move	$3,$2
+	sw	$3,8($fp)
+	j	$L133
+$L134:
+	lw	$2,40($fp)
+	slt	$3,$2,5
+	beq	$3,$0,$L137
+	l.d	$f0,0($fp)
+	j	$L122
+$L137:
+$L132:
+	.set	noreorder
+	nop
+	.set	reorder
+	lw	$2,20($fp)
+	sw	$2,8($fp)
+$L138:
+	lw	$2,8($fp)
+	lw	$3,40($fp)
+	slt	$2,$2,$3
+	bne	$2,$0,$L141
+	j	$L139
+$L141:
+	lw	$2,8($fp)
+	move	$3,$2
+	sll	$2,$3,3
+	lw	$3,44($fp)
+	addu	$2,$2,$3
+	lw	$3,8($fp)
+	move	$4,$3
+	sll	$3,$4,3
+	lw	$4,52($fp)
+	addu	$3,$3,$4
+	l.d	$f0,0($2)
+	l.d	$f2,0($3)
+	mul.d	$f0,$f0,$f2
+	l.d	$f2,0($fp)
+	add.d	$f0,$f0,$f2
+	lw	$2,8($fp)
+	move	$3,$2
+	sll	$2,$3,3
+	lw	$4,44($fp)
+	addu	$3,$2,$4
+	addu	$2,$3,8
+	lw	$3,8($fp)
+	move	$4,$3
+	sll	$3,$4,3
+	lw	$4,52($fp)
+	addu	$3,$3,$4
+	addu	$4,$3,8
+	l.d	$f2,0($2)
+	l.d	$f4,0($4)
+	mul.d	$f2,$f2,$f4
+	add.d	$f0,$f0,$f2
+	lw	$2,8($fp)
+	move	$3,$2
+	sll	$2,$3,3
+	lw	$4,44($fp)
+	addu	$3,$2,$4
+	addu	$2,$3,16
+	lw	$3,8($fp)
+	move	$4,$3
+	sll	$3,$4,3
+	lw	$4,52($fp)
+	addu	$3,$3,$4
+	addu	$4,$3,16
+	l.d	$f2,0($2)
+	l.d	$f4,0($4)
+	mul.d	$f2,$f2,$f4
+	add.d	$f0,$f0,$f2
+	lw	$2,8($fp)
+	move	$3,$2
+	sll	$2,$3,3
+	lw	$4,44($fp)
+	addu	$3,$2,$4
+	addu	$2,$3,24
+	lw	$3,8($fp)
+	move	$4,$3
+	sll	$3,$4,3
+	lw	$4,52($fp)
+	addu	$3,$3,$4
+	addu	$4,$3,24
+	l.d	$f2,0($2)
+	l.d	$f4,0($4)
+	mul.d	$f2,$f2,$f4
+	add.d	$f0,$f0,$f2
+	lw	$2,8($fp)
+	move	$3,$2
+	sll	$2,$3,3
+	lw	$4,44($fp)
+	addu	$3,$2,$4
+	addu	$2,$3,32
+	lw	$3,8($fp)
+	move	$4,$3
+	sll	$3,$4,3
+	lw	$4,52($fp)
+	addu	$3,$3,$4
+	addu	$4,$3,32
+	l.d	$f2,0($2)
+	l.d	$f4,0($4)
+	mul.d	$f2,$f2,$f4
+	add.d	$f0,$f0,$f2
+	s.d	$f0,0($fp)
+$L140:
+	lw	$2,8($fp)
+	addu	$3,$2,5
+	sw	$3,8($fp)
+	j	$L138
+$L139:
+	l.d	$f0,0($fp)
+	j	$L122
+$L122:
+	move	$sp,$fp			# sp not trusted here
+	lw	$fp,32($sp)
+	addu	$sp,$sp,40
+	j	$31
+	.end	ddot
+	.loc	1 699
+	.ent	dscal
+dscal:
+	.frame	$fp,24,$31		# vars= 16, regs= 1/0, args= 0, extra= 0
+	.mask	0x40000000,-8
+	.fmask	0x00000000,0
+	subu	$sp,$sp,24
+	sw	$fp,16($sp)
+	move	$fp,$sp
+	sw	$4,24($fp)
+	dsw	$6,32($fp)
+	lw	$2,24($fp)
+	bgtz	$2,$L143
+	j	$L142
+$L143:
+	lw	$2,44($fp)
+	li	$3,0x00000001		# 1
+	beq	$2,$3,$L144
+	lw	$2,24($fp)
+	lw	$3,44($fp)
+	mult	$2,$3
+	mflo	$2
+	sw	$2,12($fp)
+	sw	$0,0($fp)
+$L145:
+	lw	$2,0($fp)
+	lw	$3,12($fp)
+	slt	$2,$2,$3
+	bne	$2,$0,$L148
+	j	$L146
+$L148:
+	lw	$2,0($fp)
+	move	$3,$2
+	sll	$2,$3,3
+	lw	$3,40($fp)
+	addu	$2,$2,$3
+	lw	$3,0($fp)
+	move	$4,$3
+	sll	$3,$4,3
+	lw	$4,40($fp)
+	addu	$3,$3,$4
+	l.d	$f0,32($fp)
+	l.d	$f2,0($3)
+	mul.d	$f0,$f0,$f2
+	s.d	$f0,0($2)
+$L147:
+	lw	$2,0($fp)
+	lw	$3,44($fp)
+	addu	$2,$2,$3
+	sw	$2,0($fp)
+	j	$L145
+$L146:
+	j	$L142
+$L144:
+	lw	$2,24($fp)
+	li	$6,0x66666667		# 1717986919
+	mult	$2,$6
+	mfhi	$5
+	mflo	$4
+	srl	$6,$5,0
+	move	$7,$0
+	sra	$3,$6,1
+	sra	$4,$2,31
+	subu	$3,$3,$4
+	move	$5,$3
+	sll	$4,$5,2
+	addu	$4,$4,$3
+	subu	$2,$2,$4
+	sw	$2,4($fp)
+	lw	$2,4($fp)
+	beq	$2,$0,$L149
+	sw	$0,0($fp)
+$L150:
+	lw	$2,0($fp)
+	lw	$3,4($fp)
+	slt	$2,$2,$3
+	bne	$2,$0,$L153
+	j	$L151
+$L153:
+	lw	$2,0($fp)
+	move	$3,$2
+	sll	$2,$3,3
+	lw	$3,40($fp)
+	addu	$2,$2,$3
+	lw	$3,0($fp)
+	move	$4,$3
+	sll	$3,$4,3
+	lw	$4,40($fp)
+	addu	$3,$3,$4
+	l.d	$f0,32($fp)
+	l.d	$f2,0($3)
+	mul.d	$f0,$f0,$f2
+	s.d	$f0,0($2)
+$L152:
+	lw	$3,0($fp)
+	addu	$2,$3,1
+	move	$3,$2
+	sw	$3,0($fp)
+	j	$L150
+$L151:
+	lw	$2,24($fp)
+	slt	$3,$2,5
+	bne	$3,$0,$L142
+$L154:
+$L149:
+	.set	noreorder
+	nop
+	.set	reorder
+	lw	$2,4($fp)
+	sw	$2,0($fp)
+$L155:
+	lw	$2,0($fp)
+	lw	$3,24($fp)
+	slt	$2,$2,$3
+	bne	$2,$0,$L158
+	j	$L156
+$L158:
+	lw	$2,0($fp)
+	move	$3,$2
+	sll	$2,$3,3
+	lw	$3,40($fp)
+	addu	$2,$2,$3
+	lw	$3,0($fp)
+	move	$4,$3
+	sll	$3,$4,3
+	lw	$4,40($fp)
+	addu	$3,$3,$4
+	l.d	$f0,32($fp)
+	l.d	$f2,0($3)
+	mul.d	$f0,$f0,$f2
+	s.d	$f0,0($2)
+	lw	$2,0($fp)
+	move	$3,$2
+	sll	$2,$3,3
+	lw	$4,40($fp)
+	addu	$3,$2,$4
+	addu	$2,$3,8
+	lw	$3,0($fp)
+	move	$4,$3
+	sll	$3,$4,3
+	lw	$4,40($fp)
+	addu	$3,$3,$4
+	addu	$4,$3,8
+	l.d	$f0,32($fp)
+	l.d	$f2,0($4)
+	mul.d	$f0,$f0,$f2
+	s.d	$f0,0($2)
+	lw	$2,0($fp)
+	move	$3,$2
+	sll	$2,$3,3
+	lw	$4,40($fp)
+	addu	$3,$2,$4
+	addu	$2,$3,16
+	lw	$3,0($fp)
+	move	$4,$3
+	sll	$3,$4,3
+	lw	$4,40($fp)
+	addu	$3,$3,$4
+	addu	$4,$3,16
+	l.d	$f0,32($fp)
+	l.d	$f2,0($4)
+	mul.d	$f0,$f0,$f2
+	s.d	$f0,0($2)
+	lw	$2,0($fp)
+	move	$3,$2
+	sll	$2,$3,3
+	lw	$4,40($fp)
+	addu	$3,$2,$4
+	addu	$2,$3,24
+	lw	$3,0($fp)
+	move	$4,$3
+	sll	$3,$4,3
+	lw	$4,40($fp)
+	addu	$3,$3,$4
+	addu	$4,$3,24
+	l.d	$f0,32($fp)
+	l.d	$f2,0($4)
+	mul.d	$f0,$f0,$f2
+	s.d	$f0,0($2)
+	lw	$2,0($fp)
+	move	$3,$2
+	sll	$2,$3,3
+	lw	$4,40($fp)
+	addu	$3,$2,$4
+	addu	$2,$3,32
+	lw	$3,0($fp)
+	move	$4,$3
+	sll	$3,$4,3
+	lw	$4,40($fp)
+	addu	$3,$3,$4
+	addu	$4,$3,32
+	l.d	$f0,32($fp)
+	l.d	$f2,0($4)
+	mul.d	$f0,$f0,$f2
+	s.d	$f0,0($2)
+$L157:
+	lw	$2,0($fp)
+	addu	$3,$2,5
+	sw	$3,0($fp)
+	j	$L155
+$L156:
+$L142:
+	move	$sp,$fp			# sp not trusted here
+	lw	$fp,16($sp)
+	addu	$sp,$sp,24
+	j	$31
+	.end	dscal
+	.loc	1 748
+	.ent	idamax
+idamax:
+	.frame	$fp,32,$31		# vars= 24, regs= 1/0, args= 0, extra= 0
+	.mask	0x40000000,-8
+	.fmask	0x00000000,0
+	subu	$sp,$sp,32
+	sw	$fp,24($sp)
+	move	$fp,$sp
+	sw	$4,32($fp)
+	sw	$5,36($fp)
+	sw	$6,40($fp)
+	lw	$2,32($fp)
+	bgtz	$2,$L160
+	li	$2,-1				# 0xffffffff
+	j	$L159
+$L160:
+	lw	$2,32($fp)
+	li	$3,0x00000001		# 1
+	bne	$2,$3,$L161
+	move	$2,$0
+	j	$L159
+$L161:
+	lw	$2,40($fp)
+	li	$3,0x00000001		# 1
+	beq	$2,$3,$L162
+	li	$2,0x00000001		# 1
+	sw	$2,12($fp)
+	lw	$2,36($fp)
+	l.d	$f0,0($2)
+	abs.d	$f2,$f0
+	s.d	$f2,0($fp)
+	lw	$2,12($fp)
+	lw	$3,40($fp)
+	addu	$2,$2,$3
+	sw	$2,12($fp)
+	li	$2,0x00000001		# 1
+	sw	$2,8($fp)
+$L163:
+	lw	$2,8($fp)
+	lw	$3,32($fp)
+	slt	$2,$2,$3
+	bne	$2,$0,$L166
+	j	$L164
+$L166:
+	lw	$2,12($fp)
+	move	$3,$2
+	sll	$2,$3,3
+	lw	$3,36($fp)
+	addu	$2,$2,$3
+	l.d	$f2,0($2)
+	abs.d	$f0,$f2
+	l.d	$f2,0($fp)
+	c.lt.d	$f2,$f0
+	bc1f	$L167
+	lw	$2,8($fp)
+	sw	$2,16($fp)
+	lw	$2,12($fp)
+	move	$3,$2
+	sll	$2,$3,3
+	lw	$3,36($fp)
+	addu	$2,$2,$3
+	l.d	$f0,0($2)
+	abs.d	$f2,$f0
+	s.d	$f2,0($fp)
+$L167:
+	lw	$2,12($fp)
+	lw	$3,40($fp)
+	addu	$2,$2,$3
+	sw	$2,12($fp)
+$L165:
+	lw	$3,8($fp)
+	addu	$2,$3,1
+	move	$3,$2
+	sw	$3,8($fp)
+	j	$L163
+$L164:
+	j	$L168
+$L162:
+	sw	$0,16($fp)
+	lw	$2,36($fp)
+	l.d	$f0,0($2)
+	abs.d	$f2,$f0
+	s.d	$f2,0($fp)
+	li	$2,0x00000001		# 1
+	sw	$2,8($fp)
+$L169:
+	lw	$2,8($fp)
+	lw	$3,32($fp)
+	slt	$2,$2,$3
+	bne	$2,$0,$L172
+	j	$L170
+$L172:
+	lw	$2,8($fp)
+	move	$3,$2
+	sll	$2,$3,3
+	lw	$3,36($fp)
+	addu	$2,$2,$3
+	l.d	$f2,0($2)
+	abs.d	$f0,$f2
+	l.d	$f2,0($fp)
+	c.lt.d	$f2,$f0
+	bc1f	$L173
+	lw	$2,8($fp)
+	sw	$2,16($fp)
+	lw	$2,8($fp)
+	move	$3,$2
+	sll	$2,$3,3
+	lw	$3,36($fp)
+	addu	$2,$2,$3
+	l.d	$f0,0($2)
+	abs.d	$f2,$f0
+	s.d	$f2,0($fp)
+$L173:
+$L171:
+	lw	$3,8($fp)
+	addu	$2,$3,1
+	move	$3,$2
+	sw	$3,8($fp)
+	j	$L169
+$L170:
+$L168:
+	lw	$2,16($fp)
+	j	$L159
+$L159:
+	move	$sp,$fp			# sp not trusted here
+	lw	$fp,24($sp)
+	addu	$sp,$sp,32
+	j	$31
+	.end	idamax
+	.loc	1 789
+	.ent	epslon
+epslon:
+	.frame	$fp,40,$31		# vars= 32, regs= 1/0, args= 0, extra= 0
+	.mask	0x40000000,-8
+	.fmask	0x00000000,0
+	subu	$sp,$sp,40
+	sw	$fp,32($sp)
+	move	$fp,$sp
+	s.d	$f12,40($fp)
+	l.d	$f0,$LC22
+	s.d	$f0,0($fp)
+	dsz	24($fp)
+$L175:
+	l.d	$f0,24($fp)
+	mtc1	$0,$f2
+	mtc1	$0,$f3
+	c.eq.d	$f0,$f2
+	bc1t	$L177
+	j	$L176
+$L177:
+	l.d	$f0,0($fp)
+	l.d	$f2,$LC23
+	sub.d	$f0,$f0,$f2
+	s.d	$f0,8($fp)
+	l.d	$f0,8($fp)
+	l.d	$f2,8($fp)
+	add.d	$f0,$f0,$f2
+	l.d	$f2,8($fp)
+	add.d	$f0,$f0,$f2
+	s.d	$f0,16($fp)
+	l.d	$f0,16($fp)
+	l.d	$f2,$LC23
+	sub.d	$f0,$f0,$f2
+	abs.d	$f2,$f0
+	s.d	$f2,24($fp)
+	j	$L175
+$L176:
+	l.d	$f2,40($fp)
+	abs.d	$f0,$f2
+	l.d	$f4,24($fp)
+	mul.d	$f2,$f0,$f4
+	mov.d	$f0,$f2
+	j	$L174
+$L174:
+	move	$sp,$fp			# sp not trusted here
+	lw	$fp,32($sp)
+	addu	$sp,$sp,40
+	j	$31
+	.end	epslon
+	.loc	1 836
+	.ent	dmxpy
+dmxpy:
+	.frame	$fp,24,$31		# vars= 16, regs= 1/0, args= 0, extra= 0
+	.mask	0x40000000,-8
+	.fmask	0x00000000,0
+	subu	$sp,$sp,24
+	sw	$fp,16($sp)
+	move	$fp,$sp
+	sw	$4,24($fp)
+	sw	$5,28($fp)
+	sw	$6,32($fp)
+	sw	$7,36($fp)
+	lw	$2,32($fp)
+	sra	$3,$2,31
+	srl	$4,$3,31
+	addu	$5,$2,$4
+	sra	$3,$5,1
+	move	$4,$3
+	sll	$3,$4,1
+	subu	$2,$2,$3
+	sw	$2,0($fp)
+	lw	$2,0($fp)
+	blez	$2,$L179
+	lw	$2,0($fp)
+	subu	$3,$2,1
+	sw	$3,0($fp)
+	sw	$0,4($fp)
+$L180:
+	lw	$2,4($fp)
+	lw	$3,24($fp)
+	slt	$2,$2,$3
+	bne	$2,$0,$L183
+	j	$L181
+$L183:
+	lw	$2,4($fp)
+	move	$3,$2
+	sll	$2,$3,3
+	lw	$3,28($fp)
+	addu	$2,$2,$3
+	lw	$3,4($fp)
+	move	$4,$3
+	sll	$3,$4,3
+	lw	$4,28($fp)
+	addu	$3,$3,$4
+	lw	$4,0($fp)
+	move	$5,$4
+	sll	$4,$5,3
+	lw	$5,40($fp)
+	addu	$4,$4,$5
+	lw	$5,36($fp)
+	lw	$6,0($fp)
+	mult	$5,$6
+	mflo	$5
+	lw	$6,4($fp)
+	addu	$5,$5,$6
+	move	$6,$5
+	sll	$5,$6,3
+	lw	$6,44($fp)
+	addu	$5,$5,$6
+	l.d	$f0,0($4)
+	l.d	$f2,0($5)
+	mul.d	$f0,$f0,$f2
+	l.d	$f2,0($3)
+	add.d	$f0,$f2,$f0
+	s.d	$f0,0($2)
+$L182:
+	lw	$3,4($fp)
+	addu	$2,$3,1
+	move	$3,$2
+	sw	$3,4($fp)
+	j	$L180
+$L181:
+$L179:
+	lw	$2,32($fp)
+	move	$3,$2
+	bgez	$3,$L184
+	addu	$3,$3,3
+$L184:
+	sra	$3,$3,2
+	move	$4,$3
+	sll	$3,$4,2
+	subu	$2,$2,$3
+	sw	$2,0($fp)
+	lw	$2,0($fp)
+	slt	$3,$2,2
+	bne	$3,$0,$L185
+	lw	$2,0($fp)
+	subu	$3,$2,1
+	sw	$3,0($fp)
+	sw	$0,4($fp)
+$L186:
+	lw	$2,4($fp)
+	lw	$3,24($fp)
+	slt	$2,$2,$3
+	bne	$2,$0,$L189
+	j	$L187
+$L189:
+	lw	$2,4($fp)
+	move	$3,$2
+	sll	$2,$3,3
+	lw	$3,28($fp)
+	addu	$2,$2,$3
+	lw	$3,4($fp)
+	move	$4,$3
+	sll	$3,$4,3
+	lw	$4,28($fp)
+	addu	$3,$3,$4
+	lw	$4,0($fp)
+	move	$5,$4
+	sll	$4,$5,3
+	lw	$6,40($fp)
+	addu	$5,$4,$6
+	subu	$4,$5,8
+	lw	$6,0($fp)
+	subu	$5,$6,1
+	lw	$6,36($fp)
+	mult	$5,$6
+	mflo	$5
+	lw	$6,4($fp)
+	addu	$5,$5,$6
+	move	$6,$5
+	sll	$5,$6,3
+	lw	$6,44($fp)
+	addu	$5,$5,$6
+	l.d	$f0,0($4)
+	l.d	$f2,0($5)
+	mul.d	$f0,$f0,$f2
+	l.d	$f2,0($3)
+	add.d	$f0,$f0,$f2
+	lw	$3,0($fp)
+	move	$4,$3
+	sll	$3,$4,3
+	lw	$4,40($fp)
+	addu	$3,$3,$4
+	lw	$4,36($fp)
+	lw	$5,0($fp)
+	mult	$4,$5
+	mflo	$4
+	lw	$5,4($fp)
+	addu	$4,$4,$5
+	move	$5,$4
+	sll	$4,$5,3
+	lw	$5,44($fp)
+	addu	$4,$4,$5
+	l.d	$f2,0($3)
+	l.d	$f4,0($4)
+	mul.d	$f2,$f2,$f4
+	add.d	$f0,$f0,$f2
+	s.d	$f0,0($2)
+$L188:
+	lw	$3,4($fp)
+	addu	$2,$3,1
+	move	$3,$2
+	sw	$3,4($fp)
+	j	$L186
+$L187:
+$L185:
+	lw	$2,32($fp)
+	move	$3,$2
+	bgez	$3,$L190
+	addu	$3,$3,7
+$L190:
+	sra	$3,$3,3
+	move	$4,$3
+	sll	$3,$4,3
+	subu	$2,$2,$3
+	sw	$2,0($fp)
+	lw	$2,0($fp)
+	slt	$3,$2,4
+	bne	$3,$0,$L191
+	lw	$2,0($fp)
+	subu	$3,$2,1
+	sw	$3,0($fp)
+	sw	$0,4($fp)
+$L192:
+	lw	$2,4($fp)
+	lw	$3,24($fp)
+	slt	$2,$2,$3
+	bne	$2,$0,$L195
+	j	$L193
+$L195:
+	lw	$2,4($fp)
+	move	$3,$2
+	sll	$2,$3,3
+	lw	$3,28($fp)
+	addu	$2,$2,$3
+	lw	$3,4($fp)
+	move	$4,$3
+	sll	$3,$4,3
+	lw	$4,28($fp)
+	addu	$3,$3,$4
+	lw	$4,0($fp)
+	move	$5,$4
+	sll	$4,$5,3
+	lw	$6,40($fp)
+	addu	$5,$4,$6
+	subu	$4,$5,24
+	lw	$6,0($fp)
+	subu	$5,$6,3
+	lw	$6,36($fp)
+	mult	$5,$6
+	mflo	$5
+	lw	$6,4($fp)
+	addu	$5,$5,$6
+	move	$6,$5
+	sll	$5,$6,3
+	lw	$6,44($fp)
+	addu	$5,$5,$6
+	l.d	$f0,0($4)
+	l.d	$f2,0($5)
+	mul.d	$f0,$f0,$f2
+	l.d	$f2,0($3)
+	add.d	$f0,$f0,$f2
+	lw	$3,0($fp)
+	move	$4,$3
+	sll	$3,$4,3
+	lw	$5,40($fp)
+	addu	$4,$3,$5
+	subu	$3,$4,16
+	lw	$5,0($fp)
+	subu	$4,$5,2
+	lw	$5,36($fp)
+	mult	$4,$5
+	mflo	$4
+	lw	$5,4($fp)
+	addu	$4,$4,$5
+	move	$5,$4
+	sll	$4,$5,3
+	lw	$5,44($fp)
+	addu	$4,$4,$5
+	l.d	$f2,0($3)
+	l.d	$f4,0($4)
+	mul.d	$f2,$f2,$f4
+	add.d	$f0,$f0,$f2
+	lw	$3,0($fp)
+	move	$4,$3
+	sll	$3,$4,3
+	lw	$5,40($fp)
+	addu	$4,$3,$5
+	subu	$3,$4,8
+	lw	$5,0($fp)
+	subu	$4,$5,1
+	lw	$5,36($fp)
+	mult	$4,$5
+	mflo	$4
+	lw	$5,4($fp)
+	addu	$4,$4,$5
+	move	$5,$4
+	sll	$4,$5,3
+	lw	$5,44($fp)
+	addu	$4,$4,$5
+	l.d	$f2,0($3)
+	l.d	$f4,0($4)
+	mul.d	$f2,$f2,$f4
+	add.d	$f0,$f0,$f2
+	lw	$3,0($fp)
+	move	$4,$3
+	sll	$3,$4,3
+	lw	$4,40($fp)
+	addu	$3,$3,$4
+	lw	$4,36($fp)
+	lw	$5,0($fp)
+	mult	$4,$5
+	mflo	$4
+	lw	$5,4($fp)
+	addu	$4,$4,$5
+	move	$5,$4
+	sll	$4,$5,3
+	lw	$5,44($fp)
+	addu	$4,$4,$5
+	l.d	$f2,0($3)
+	l.d	$f4,0($4)
+	mul.d	$f2,$f2,$f4
+	add.d	$f0,$f0,$f2
+	s.d	$f0,0($2)
+$L194:
+	lw	$3,4($fp)
+	addu	$2,$3,1
+	move	$3,$2
+	sw	$3,4($fp)
+	j	$L192
+$L193:
+$L191:
+	lw	$2,32($fp)
+	move	$3,$2
+	bgez	$3,$L196
+	addu	$3,$3,15
+$L196:
+	sra	$3,$3,4
+	move	$4,$3
+	sll	$3,$4,4
+	subu	$2,$2,$3
+	sw	$2,0($fp)
+	lw	$2,0($fp)
+	slt	$3,$2,8
+	bne	$3,$0,$L197
+	lw	$2,0($fp)
+	subu	$3,$2,1
+	sw	$3,0($fp)
+	sw	$0,4($fp)
+$L198:
+	lw	$2,4($fp)
+	lw	$3,24($fp)
+	slt	$2,$2,$3
+	bne	$2,$0,$L201
+	j	$L199
+$L201:
+	lw	$2,4($fp)
+	move	$3,$2
+	sll	$2,$3,3
+	lw	$3,28($fp)
+	addu	$2,$2,$3
+	lw	$3,4($fp)
+	move	$4,$3
+	sll	$3,$4,3
+	lw	$4,28($fp)
+	addu	$3,$3,$4
+	lw	$4,0($fp)
+	move	$5,$4
+	sll	$4,$5,3
+	lw	$6,40($fp)
+	addu	$5,$4,$6
+	subu	$4,$5,56
+	lw	$6,0($fp)
+	subu	$5,$6,7
+	lw	$6,36($fp)
+	mult	$5,$6
+	mflo	$5
+	lw	$6,4($fp)
+	addu	$5,$5,$6
+	move	$6,$5
+	sll	$5,$6,3
+	lw	$6,44($fp)
+	addu	$5,$5,$6
+	l.d	$f0,0($4)
+	l.d	$f2,0($5)
+	mul.d	$f0,$f0,$f2
+	l.d	$f2,0($3)
+	add.d	$f0,$f0,$f2
+	lw	$3,0($fp)
+	move	$4,$3
+	sll	$3,$4,3
+	lw	$5,40($fp)
+	addu	$4,$3,$5
+	subu	$3,$4,48
+	lw	$5,0($fp)
+	subu	$4,$5,6
+	lw	$5,36($fp)
+	mult	$4,$5
+	mflo	$4
+	lw	$5,4($fp)
+	addu	$4,$4,$5
+	move	$5,$4
+	sll	$4,$5,3
+	lw	$5,44($fp)
+	addu	$4,$4,$5
+	l.d	$f2,0($3)
+	l.d	$f4,0($4)
+	mul.d	$f2,$f2,$f4
+	add.d	$f0,$f0,$f2
+	lw	$3,0($fp)
+	move	$4,$3
+	sll	$3,$4,3
+	lw	$5,40($fp)
+	addu	$4,$3,$5
+	subu	$3,$4,40
+	lw	$5,0($fp)
+	subu	$4,$5,5
+	lw	$5,36($fp)
+	mult	$4,$5
+	mflo	$4
+	lw	$5,4($fp)
+	addu	$4,$4,$5
+	move	$5,$4
+	sll	$4,$5,3
+	lw	$5,44($fp)
+	addu	$4,$4,$5
+	l.d	$f2,0($3)
+	l.d	$f4,0($4)
+	mul.d	$f2,$f2,$f4
+	add.d	$f0,$f0,$f2
+	lw	$3,0($fp)
+	move	$4,$3
+	sll	$3,$4,3
+	lw	$5,40($fp)
+	addu	$4,$3,$5
+	subu	$3,$4,32
+	lw	$5,0($fp)
+	subu	$4,$5,4
+	lw	$5,36($fp)
+	mult	$4,$5
+	mflo	$4
+	lw	$5,4($fp)
+	addu	$4,$4,$5
+	move	$5,$4
+	sll	$4,$5,3
+	lw	$5,44($fp)
+	addu	$4,$4,$5
+	l.d	$f2,0($3)
+	l.d	$f4,0($4)
+	mul.d	$f2,$f2,$f4
+	add.d	$f0,$f0,$f2
+	lw	$3,0($fp)
+	move	$4,$3
+	sll	$3,$4,3
+	lw	$5,40($fp)
+	addu	$4,$3,$5
+	subu	$3,$4,24
+	lw	$5,0($fp)
+	subu	$4,$5,3
+	lw	$5,36($fp)
+	mult	$4,$5
+	mflo	$4
+	lw	$5,4($fp)
+	addu	$4,$4,$5
+	move	$5,$4
+	sll	$4,$5,3
+	lw	$5,44($fp)
+	addu	$4,$4,$5
+	l.d	$f2,0($3)
+	l.d	$f4,0($4)
+	mul.d	$f2,$f2,$f4
+	add.d	$f0,$f0,$f2
+	lw	$3,0($fp)
+	move	$4,$3
+	sll	$3,$4,3
+	lw	$5,40($fp)
+	addu	$4,$3,$5
+	subu	$3,$4,16
+	lw	$5,0($fp)
+	subu	$4,$5,2
+	lw	$5,36($fp)
+	mult	$4,$5
+	mflo	$4
+	lw	$5,4($fp)
+	addu	$4,$4,$5
+	move	$5,$4
+	sll	$4,$5,3
+	lw	$5,44($fp)
+	addu	$4,$4,$5
+	l.d	$f2,0($3)
+	l.d	$f4,0($4)
+	mul.d	$f2,$f2,$f4
+	add.d	$f0,$f0,$f2
+	lw	$3,0($fp)
+	move	$4,$3
+	sll	$3,$4,3
+	lw	$5,40($fp)
+	addu	$4,$3,$5
+	subu	$3,$4,8
+	lw	$5,0($fp)
+	subu	$4,$5,1
+	lw	$5,36($fp)
+	mult	$4,$5
+	mflo	$4
+	lw	$5,4($fp)
+	addu	$4,$4,$5
+	move	$5,$4
+	sll	$4,$5,3
+	lw	$5,44($fp)
+	addu	$4,$4,$5
+	l.d	$f2,0($3)
+	l.d	$f4,0($4)
+	mul.d	$f2,$f2,$f4
+	add.d	$f0,$f0,$f2
+	lw	$3,0($fp)
+	move	$4,$3
+	sll	$3,$4,3
+	lw	$4,40($fp)
+	addu	$3,$3,$4
+	lw	$4,36($fp)
+	lw	$5,0($fp)
+	mult	$4,$5
+	mflo	$4
+	lw	$5,4($fp)
+	addu	$4,$4,$5
+	move	$5,$4
+	sll	$4,$5,3
+	lw	$5,44($fp)
+	addu	$4,$4,$5
+	l.d	$f2,0($3)
+	l.d	$f4,0($4)
+	mul.d	$f2,$f2,$f4
+	add.d	$f0,$f0,$f2
+	s.d	$f0,0($2)
+$L200:
+	lw	$3,4($fp)
+	addu	$2,$3,1
+	move	$3,$2
+	sw	$3,4($fp)
+	j	$L198
+$L199:
+$L197:
+	lw	$2,32($fp)
+	move	$3,$2
+	bgez	$3,$L202
+	addu	$3,$3,15
+$L202:
+	sra	$3,$3,4
+	move	$4,$3
+	sll	$3,$4,4
+	subu	$2,$2,$3
+	addu	$3,$2,16
+	sw	$3,8($fp)
+	lw	$2,8($fp)
+	subu	$3,$2,1
+	sw	$3,0($fp)
+$L203:
+	lw	$2,0($fp)
+	lw	$3,32($fp)
+	slt	$2,$2,$3
+	bne	$2,$0,$L206
+	j	$L204
+$L206:
+	sw	$0,4($fp)
+$L207:
+	lw	$2,4($fp)
+	lw	$3,24($fp)
+	slt	$2,$2,$3
+	bne	$2,$0,$L210
+	j	$L208
+$L210:
+	lw	$2,4($fp)
+	move	$3,$2
+	sll	$2,$3,3
+	lw	$3,28($fp)
+	addu	$2,$2,$3
+	lw	$3,4($fp)
+	move	$4,$3
+	sll	$3,$4,3
+	lw	$4,28($fp)
+	addu	$3,$3,$4
+	lw	$4,0($fp)
+	move	$5,$4
+	sll	$4,$5,3
+	lw	$6,40($fp)
+	addu	$5,$4,$6
+	subu	$4,$5,120
+	lw	$6,0($fp)
+	subu	$5,$6,15
+	lw	$6,36($fp)
+	mult	$5,$6
+	mflo	$5
+	lw	$6,4($fp)
+	addu	$5,$5,$6
+	move	$6,$5
+	sll	$5,$6,3
+	lw	$6,44($fp)
+	addu	$5,$5,$6
+	l.d	$f0,0($4)
+	l.d	$f2,0($5)
+	mul.d	$f0,$f0,$f2
+	l.d	$f2,0($3)
+	add.d	$f0,$f0,$f2
+	lw	$3,0($fp)
+	move	$4,$3
+	sll	$3,$4,3
+	lw	$5,40($fp)
+	addu	$4,$3,$5
+	subu	$3,$4,112
+	lw	$5,0($fp)
+	subu	$4,$5,14
+	lw	$5,36($fp)
+	mult	$4,$5
+	mflo	$4
+	lw	$5,4($fp)
+	addu	$4,$4,$5
+	move	$5,$4
+	sll	$4,$5,3
+	lw	$5,44($fp)
+	addu	$4,$4,$5
+	l.d	$f2,0($3)
+	l.d	$f4,0($4)
+	mul.d	$f2,$f2,$f4
+	add.d	$f0,$f0,$f2
+	lw	$3,0($fp)
+	move	$4,$3
+	sll	$3,$4,3
+	lw	$5,40($fp)
+	addu	$4,$3,$5
+	subu	$3,$4,104
+	lw	$5,0($fp)
+	subu	$4,$5,13
+	lw	$5,36($fp)
+	mult	$4,$5
+	mflo	$4
+	lw	$5,4($fp)
+	addu	$4,$4,$5
+	move	$5,$4
+	sll	$4,$5,3
+	lw	$5,44($fp)
+	addu	$4,$4,$5
+	l.d	$f2,0($3)
+	l.d	$f4,0($4)
+	mul.d	$f2,$f2,$f4
+	add.d	$f0,$f0,$f2
+	lw	$3,0($fp)
+	move	$4,$3
+	sll	$3,$4,3
+	lw	$5,40($fp)
+	addu	$4,$3,$5
+	subu	$3,$4,96
+	lw	$5,0($fp)
+	subu	$4,$5,12
+	lw	$5,36($fp)
+	mult	$4,$5
+	mflo	$4
+	lw	$5,4($fp)
+	addu	$4,$4,$5
+	move	$5,$4
+	sll	$4,$5,3
+	lw	$5,44($fp)
+	addu	$4,$4,$5
+	l.d	$f2,0($3)
+	l.d	$f4,0($4)
+	mul.d	$f2,$f2,$f4
+	add.d	$f0,$f0,$f2
+	lw	$3,0($fp)
+	move	$4,$3
+	sll	$3,$4,3
+	lw	$5,40($fp)
+	addu	$4,$3,$5
+	subu	$3,$4,88
+	lw	$5,0($fp)
+	subu	$4,$5,11
+	lw	$5,36($fp)
+	mult	$4,$5
+	mflo	$4
+	lw	$5,4($fp)
+	addu	$4,$4,$5
+	move	$5,$4
+	sll	$4,$5,3
+	lw	$5,44($fp)
+	addu	$4,$4,$5
+	l.d	$f2,0($3)
+	l.d	$f4,0($4)
+	mul.d	$f2,$f2,$f4
+	add.d	$f0,$f0,$f2
+	lw	$3,0($fp)
+	move	$4,$3
+	sll	$3,$4,3
+	lw	$5,40($fp)
+	addu	$4,$3,$5
+	subu	$3,$4,80
+	lw	$5,0($fp)
+	subu	$4,$5,10
+	lw	$5,36($fp)
+	mult	$4,$5
+	mflo	$4
+	lw	$5,4($fp)
+	addu	$4,$4,$5
+	move	$5,$4
+	sll	$4,$5,3
+	lw	$5,44($fp)
+	addu	$4,$4,$5
+	l.d	$f2,0($3)
+	l.d	$f4,0($4)
+	mul.d	$f2,$f2,$f4
+	add.d	$f0,$f0,$f2
+	lw	$3,0($fp)
+	move	$4,$3
+	sll	$3,$4,3
+	lw	$5,40($fp)
+	addu	$4,$3,$5
+	subu	$3,$4,72
+	lw	$5,0($fp)
+	subu	$4,$5,9
+	lw	$5,36($fp)
+	mult	$4,$5
+	mflo	$4
+	lw	$5,4($fp)
+	addu	$4,$4,$5
+	move	$5,$4
+	sll	$4,$5,3
+	lw	$5,44($fp)
+	addu	$4,$4,$5
+	l.d	$f2,0($3)
+	l.d	$f4,0($4)
+	mul.d	$f2,$f2,$f4
+	add.d	$f0,$f0,$f2
+	lw	$3,0($fp)
+	move	$4,$3
+	sll	$3,$4,3
+	lw	$5,40($fp)
+	addu	$4,$3,$5
+	subu	$3,$4,64
+	lw	$5,0($fp)
+	subu	$4,$5,8
+	lw	$5,36($fp)
+	mult	$4,$5
+	mflo	$4
+	lw	$5,4($fp)
+	addu	$4,$4,$5
+	move	$5,$4
+	sll	$4,$5,3
+	lw	$5,44($fp)
+	addu	$4,$4,$5
+	l.d	$f2,0($3)
+	l.d	$f4,0($4)
+	mul.d	$f2,$f2,$f4
+	add.d	$f0,$f0,$f2
+	lw	$3,0($fp)
+	move	$4,$3
+	sll	$3,$4,3
+	lw	$5,40($fp)
+	addu	$4,$3,$5
+	subu	$3,$4,56
+	lw	$5,0($fp)
+	subu	$4,$5,7
+	lw	$5,36($fp)
+	mult	$4,$5
+	mflo	$4
+	lw	$5,4($fp)
+	addu	$4,$4,$5
+	move	$5,$4
+	sll	$4,$5,3
+	lw	$5,44($fp)
+	addu	$4,$4,$5
+	l.d	$f2,0($3)
+	l.d	$f4,0($4)
+	mul.d	$f2,$f2,$f4
+	add.d	$f0,$f0,$f2
+	lw	$3,0($fp)
+	move	$4,$3
+	sll	$3,$4,3
+	lw	$5,40($fp)
+	addu	$4,$3,$5
+	subu	$3,$4,48
+	lw	$5,0($fp)
+	subu	$4,$5,6
+	lw	$5,36($fp)
+	mult	$4,$5
+	mflo	$4
+	lw	$5,4($fp)
+	addu	$4,$4,$5
+	move	$5,$4
+	sll	$4,$5,3
+	lw	$5,44($fp)
+	addu	$4,$4,$5
+	l.d	$f2,0($3)
+	l.d	$f4,0($4)
+	mul.d	$f2,$f2,$f4
+	add.d	$f0,$f0,$f2
+	lw	$3,0($fp)
+	move	$4,$3
+	sll	$3,$4,3
+	lw	$5,40($fp)
+	addu	$4,$3,$5
+	subu	$3,$4,40
+	lw	$5,0($fp)
+	subu	$4,$5,5
+	lw	$5,36($fp)
+	mult	$4,$5
+	mflo	$4
+	lw	$5,4($fp)
+	addu	$4,$4,$5
+	move	$5,$4
+	sll	$4,$5,3
+	lw	$5,44($fp)
+	addu	$4,$4,$5
+	l.d	$f2,0($3)
+	l.d	$f4,0($4)
+	mul.d	$f2,$f2,$f4
+	add.d	$f0,$f0,$f2
+	lw	$3,0($fp)
+	move	$4,$3
+	sll	$3,$4,3
+	lw	$5,40($fp)
+	addu	$4,$3,$5
+	subu	$3,$4,32
+	lw	$5,0($fp)
+	subu	$4,$5,4
+	lw	$5,36($fp)
+	mult	$4,$5
+	mflo	$4
+	lw	$5,4($fp)
+	addu	$4,$4,$5
+	move	$5,$4
+	sll	$4,$5,3
+	lw	$5,44($fp)
+	addu	$4,$4,$5
+	l.d	$f2,0($3)
+	l.d	$f4,0($4)
+	mul.d	$f2,$f2,$f4
+	add.d	$f0,$f0,$f2
+	lw	$3,0($fp)
+	move	$4,$3
+	sll	$3,$4,3
+	lw	$5,40($fp)
+	addu	$4,$3,$5
+	subu	$3,$4,24
+	lw	$5,0($fp)
+	subu	$4,$5,3
+	lw	$5,36($fp)
+	mult	$4,$5
+	mflo	$4
+	lw	$5,4($fp)
+	addu	$4,$4,$5
+	move	$5,$4
+	sll	$4,$5,3
+	lw	$5,44($fp)
+	addu	$4,$4,$5
+	l.d	$f2,0($3)
+	l.d	$f4,0($4)
+	mul.d	$f2,$f2,$f4
+	add.d	$f0,$f0,$f2
+	lw	$3,0($fp)
+	move	$4,$3
+	sll	$3,$4,3
+	lw	$5,40($fp)
+	addu	$4,$3,$5
+	subu	$3,$4,16
+	lw	$5,0($fp)
+	subu	$4,$5,2
+	lw	$5,36($fp)
+	mult	$4,$5
+	mflo	$4
+	lw	$5,4($fp)
+	addu	$4,$4,$5
+	move	$5,$4
+	sll	$4,$5,3
+	lw	$5,44($fp)
+	addu	$4,$4,$5
+	l.d	$f2,0($3)
+	l.d	$f4,0($4)
+	mul.d	$f2,$f2,$f4
+	add.d	$f0,$f0,$f2
+	lw	$3,0($fp)
+	move	$4,$3
+	sll	$3,$4,3
+	lw	$5,40($fp)
+	addu	$4,$3,$5
+	subu	$3,$4,8
+	lw	$5,0($fp)
+	subu	$4,$5,1
+	lw	$5,36($fp)
+	mult	$4,$5
+	mflo	$4
+	lw	$5,4($fp)
+	addu	$4,$4,$5
+	move	$5,$4
+	sll	$4,$5,3
+	lw	$5,44($fp)
+	addu	$4,$4,$5
+	l.d	$f2,0($3)
+	l.d	$f4,0($4)
+	mul.d	$f2,$f2,$f4
+	add.d	$f0,$f0,$f2
+	lw	$3,0($fp)
+	move	$4,$3
+	sll	$3,$4,3
+	lw	$4,40($fp)
+	addu	$3,$3,$4
+	lw	$4,36($fp)
+	lw	$5,0($fp)
+	mult	$4,$5
+	mflo	$4
+	lw	$5,4($fp)
+	addu	$4,$4,$5
+	move	$5,$4
+	sll	$4,$5,3
+	lw	$5,44($fp)
+	addu	$4,$4,$5
+	l.d	$f2,0($3)
+	l.d	$f4,0($4)
+	mul.d	$f2,$f2,$f4
+	add.d	$f0,$f0,$f2
+	s.d	$f0,0($2)
+$L209:
+	lw	$3,4($fp)
+	addu	$2,$3,1
+	move	$3,$2
+	sw	$3,4($fp)
+	j	$L207
+$L208:
+$L205:
+	lw	$2,0($fp)
+	addu	$3,$2,16
+	sw	$3,0($fp)
+	j	$L203
+$L204:
+$L178:
+	move	$sp,$fp			# sp not trusted here
+	lw	$fp,16($sp)
+	addu	$sp,$sp,24
+	j	$31
+	.end	dmxpy
+	.loc	1 995
+	.ent	dtime
+dtime:
+	.frame	$fp,32,$31		# vars= 8, regs= 2/0, args= 16, extra= 0
+	.mask	0xc0000000,-4
+	.fmask	0x00000000,0
+	subu	$sp,$sp,32
+	sw	$31,28($sp)
+	sw	$fp,24($sp)
+	move	$fp,$sp
+	move	$4,$0
+	la	$5,rusage
+	jal	getrusage
+	l.s	$f0,rusage
+	#nop
+	cvt.d.w	$f0,$f0
+	s.d	$f0,16($fp)
+	l.s	$f0,rusage+4
+	#nop
+	cvt.d.w	$f0,$f0
+	l.d	$f2,$LC24
+	mul.d	$f0,$f0,$f2
+	l.d	$f2,16($fp)
+	add.d	$f0,$f2,$f0
+	s.d	$f0,16($fp)
+	l.d	$f0,16($fp)
+	j	$L211
+$L211:
+	move	$sp,$fp			# sp not trusted here
+	lw	$31,28($sp)
+	lw	$fp,24($sp)
+	addu	$sp,$sp,32
+	j	$31
+	.end	dtime

benchmarks/optimized/dhrystone.s

@@ -0,0 +1,886 @@
+	.file	1 "dhrystone.c"
+# GNU C 2.7.2.3 [AL 1.1, MM 40, tma 0.1] SimpleScalar running sstrix compiled by GNU C
+# Cc1 defaults:
+# -mgas -mgpOPT
+# Cc1 arguments (-G value = 8, Cpu = default, ISA = 1):
+# -quiet -dumpbase -O0 -o
+gcc2_compiled.:
+__gnu_compiled_c:
+	.globl	Version
+	.sdata
+	.align	2
+Version:
+	.ascii	"1.1\000"
+	.text
+	.align	2
+	.globl	main
+	.rdata
+	.align	2
+$LC0:
+	.ascii	"DHRYSTONE PROGRAM, SOME STRING\000"
+	.align	2
+$LC1:
+	.ascii	"DHRYSTONE PROGRAM, 1'ST STRING\000"
+	.align	2
+$LC2:
+	.ascii	"DHRYSTONE PROGRAM, 2'ND STRING\000"
+	.align	2
+$LC3:
+	.ascii	"Dhrystone(%s) time for %ld passes = %ld\n\000"
+	.align	2
+$LC4:
+	.ascii	"This machine benchmarks at %ld dhrystones/second\n\000"
+	.text
+	.align	2
+	.globl	Proc0
+	.align	2
+	.globl	Proc1
+	.align	2
+	.globl	Proc2
+	.align	2
+	.globl	Proc3
+	.align	2
+	.globl	Proc4
+	.align	2
+	.globl	Proc5
+	.align	2
+	.globl	Proc6
+	.align	2
+	.globl	Proc7
+	.align	2
+	.globl	Proc8
+	.align	2
+	.globl	Func1
+	.align	2
+	.globl	Func2
+	.align	2
+	.globl	Func3
+	.comm	IntGlob,4
+	.comm	BoolGlob,4
+	.comm	Char1Glob,1
+	.comm	Char2Glob,1
+	.comm	Array1Glob,204
+	.comm	Array2Glob,10404
+	.comm	PtrGlb,4
+	.comm	PtrGlbNext,4
+	.text
+	.loc	1 414
+	.ent	main
+main:
+	.frame	$fp,24,$31		# vars= 0, regs= 2/0, args= 16, extra= 0
+	.mask	0xc0000000,-4
+	.fmask	0x00000000,0
+	subu	$sp,$sp,24
+	sw	$31,20($sp)
+	sw	$fp,16($sp)
+	move	$fp,$sp
+	jal	__main
+	jal	Proc0
+	move	$4,$0
+	jal	exit
+$L1:
+	move	$sp,$fp			# sp not trusted here
+	lw	$31,20($sp)
+	lw	$fp,16($sp)
+	addu	$sp,$sp,24
+	j	$31
+	.end	main
+	.loc	1 432
+	.ent	Proc0
+Proc0:
+	.frame	$fp,152,$31		# vars= 120, regs= 3/0, args= 16, extra= 0
+	.mask	0xc0010000,-8
+	.fmask	0x00000000,0
+	subu	$sp,$sp,152
+	sw	$31,144($sp)
+	sw	$fp,140($sp)
+	sw	$16,136($sp)
+	move	$fp,$sp
+	addu	$4,$fp,120
+	jal	times
+	lw	$2,120($fp)
+	sw	$2,104($fp)
+	move	$16,$0
+$L3:
+	sltu	$2,$16,5000
+	beq	$2,$0,$L4
+$L6:
+$L5:
+	addu	$16,$16,1
+	j	$L3
+$L4:
+	addu	$4,$fp,120
+	jal	times
+	lw	$2,120($fp)
+	lw	$3,104($fp)
+	subu	$2,$2,$3
+	sw	$2,112($fp)
+	li	$4,0x00000030		# 48
+	jal	malloc
+	sw	$2,PtrGlbNext
+	li	$4,0x00000030		# 48
+	jal	malloc
+	sw	$2,PtrGlb
+	lw	$2,PtrGlb
+	lw	$3,PtrGlbNext
+	sw	$3,0($2)
+	lw	$2,PtrGlb
+	sw	$0,4($2)
+	lw	$2,PtrGlb
+	li	$3,0x00000002		# 2
+	sw	$3,8($2)
+	lw	$2,PtrGlb
+	li	$3,0x00000028		# 40
+	sw	$3,12($2)
+	lw	$3,PtrGlb
+	addu	$2,$3,16
+	move	$4,$2
+	la	$5,$LC0
+	jal	strcpy
+	addu	$2,$fp,40
+	move	$4,$2
+	la	$5,$LC1
+	jal	strcpy
+	li	$2,0x0000000a		# 10
+	sw	$2,Array2Glob+1660
+	addu	$4,$fp,120
+	jal	times
+	lw	$2,120($fp)
+	sw	$2,104($fp)
+	move	$16,$0
+$L7:
+	sltu	$2,$16,5000
+	beq	$2,$0,$L8
+$L10:
+	jal	Proc5
+	jal	Proc4
+	li	$2,0x00000002		# 2
+	sw	$2,16($fp)
+	li	$2,0x00000003		# 3
+	sw	$2,20($fp)
+	addu	$2,$fp,72
+	move	$4,$2
+	la	$5,$LC2
+	jal	strcpy
+	li	$2,0x00000001		# 1
+	sw	$2,32($fp)
+	addu	$2,$fp,40
+	addu	$3,$fp,72
+	move	$4,$2
+	move	$5,$3
+	jal	Func2
+	xori	$3,$2,0x0000
+	sltu	$2,$3,1
+	sw	$2,BoolGlob
+$L11:
+	lw	$2,16($fp)
+	lw	$3,20($fp)
+	slt	$2,$2,$3
+	beq	$2,$0,$L12
+$L13:
+	lw	$2,16($fp)
+	move	$4,$2
+	sll	$3,$4,2
+	addu	$3,$3,$2
+	lw	$2,20($fp)
+	subu	$3,$3,$2
+	sw	$3,24($fp)
+	addu	$2,$fp,24
+	lw	$4,16($fp)
+	lw	$5,20($fp)
+	move	$6,$2
+	jal	Proc7
+	lw	$3,16($fp)
+	addu	$2,$3,1
+	move	$3,$2
+	sw	$3,16($fp)
+	j	$L11
+$L12:
+	la	$4,Array1Glob
+	la	$5,Array2Glob
+	lw	$6,16($fp)
+	lw	$7,24($fp)
+	jal	Proc8
+	lw	$4,PtrGlb
+	jal	Proc1
+	li	$2,0x00000041		# 65
+	sb	$2,29($fp)
+$L14:
+	lb	$2,29($fp)
+	lb	$3,Char2Glob
+	slt	$2,$3,$2
+	bne	$2,$0,$L15
+$L17:
+	lb	$2,29($fp)
+	move	$4,$2
+	li	$5,0x00000043		# 67
+	jal	Func1
+	lw	$3,32($fp)
+	bne	$3,$2,$L18
+	addu	$2,$fp,32
+	move	$4,$0
+	move	$5,$2
+	jal	Proc6
+$L18:
+$L16:
+	lbu	$2,29($fp)
+	addu	$3,$2,1
+	move	$2,$3
+	sb	$2,29($fp)
+	j	$L14
+$L15:
+	lw	$2,20($fp)
+	lw	$3,16($fp)
+	mult	$2,$3
+	mflo	$2
+	sw	$2,24($fp)
+	lw	$2,24($fp)
+	lw	$3,16($fp)
+	div	$2,$2,$3
+	sw	$2,20($fp)
+	lw	$2,24($fp)
+	lw	$3,20($fp)
+	subu	$2,$2,$3
+	move	$4,$2
+	sll	$3,$4,3
+	subu	$3,$3,$2
+	lw	$2,16($fp)
+	subu	$3,$3,$2
+	sw	$3,20($fp)
+	addu	$4,$fp,16
+	jal	Proc2
+$L9:
+	addu	$16,$16,1
+	j	$L7
+$L8:
+	addu	$4,$fp,120
+	jal	times
+	lw	$2,120($fp)
+	lw	$3,104($fp)
+	subu	$2,$2,$3
+	lw	$3,112($fp)
+	subu	$2,$2,$3
+	sw	$2,108($fp)
+	lw	$2,108($fp)
+	li	$6,0x51eb851f		# 1374389535
+	mult	$2,$6
+	mfhi	$5
+	mflo	$4
+	srl	$6,$5,0
+	move	$7,$0
+	sra	$3,$6,5
+	sra	$4,$2,31
+	subu	$2,$3,$4
+	la	$4,$LC3
+	la	$5,Version
+	li	$6,0x00001388		# 5000
+	move	$7,$2
+	jal	printf
+	li	$2,0x0007a120		# 500000
+	lw	$3,108($fp)
+	div	$2,$2,$3
+	la	$4,$LC4
+	move	$5,$2
+	jal	printf
+$L2:
+	move	$sp,$fp			# sp not trusted here
+	lw	$31,144($sp)
+	lw	$fp,140($sp)
+	lw	$16,136($sp)
+	addu	$sp,$sp,152
+	j	$31
+	.end	Proc0
+	.loc	1 540
+	.ent	Proc1
+Proc1:
+	.frame	$fp,24,$31		# vars= 0, regs= 2/0, args= 16, extra= 0
+	.mask	0xc0000000,-4
+	.fmask	0x00000000,0
+	subu	$sp,$sp,24
+	sw	$31,20($sp)
+	sw	$fp,16($sp)
+	move	$fp,$sp
+	sw	$4,24($fp)
+	lw	$2,24($fp)
+	lw	$3,0($2)
+	lw	$2,PtrGlb
+	move	$4,$3
+	move	$5,$2
+	li	$6,0x00000030		# 48
+	jal	memcpy
+	lw	$2,24($fp)
+	li	$3,0x00000005		# 5
+	sw	$3,12($2)
+	lw	$3,24($fp)
+	lw	$2,0($3)
+	lw	$3,24($fp)
+	lw	$4,12($3)
+	sw	$4,12($2)
+	lw	$3,24($fp)
+	lw	$2,0($3)
+	lw	$3,24($fp)
+	lw	$4,0($3)
+	sw	$4,0($2)
+	lw	$2,24($fp)
+	lw	$3,0($2)
+	lw	$4,0($3)
+	jal	Proc3
+	lw	$2,24($fp)
+	lw	$3,0($2)
+	lw	$2,4($3)
+	bne	$2,$0,$L20
+	lw	$2,24($fp)
+	lw	$3,0($2)
+	li	$2,0x00000006		# 6
+	sw	$2,12($3)
+	lw	$2,24($fp)
+	lw	$3,24($fp)
+	lw	$4,0($3)
+	addu	$3,$4,8
+	lw	$4,8($2)
+	move	$5,$3
+	jal	Proc6
+	lw	$3,24($fp)
+	lw	$2,0($3)
+	lw	$3,PtrGlb
+	lw	$4,0($3)
+	sw	$4,0($2)
+	lw	$3,24($fp)
+	lw	$2,0($3)
+	lw	$3,24($fp)
+	lw	$4,0($3)
+	addu	$3,$4,12
+	lw	$4,12($2)
+	li	$5,0x0000000a		# 10
+	move	$6,$3
+	jal	Proc7
+	j	$L21
+$L20:
+	lw	$2,24($fp)
+	lw	$3,24($fp)
+	lw	$4,0($3)
+	move	$3,$4
+	move	$4,$2
+	move	$5,$3
+	li	$6,0x00000030		# 48
+	jal	memcpy
+$L21:
+$L19:
+	move	$sp,$fp			# sp not trusted here
+	lw	$31,20($sp)
+	lw	$fp,16($sp)
+	addu	$sp,$sp,24
+	j	$31
+	.end	Proc1
+	.loc	1 563
+	.ent	Proc2
+Proc2:
+	.frame	$fp,16,$31		# vars= 8, regs= 1/0, args= 0, extra= 0
+	.mask	0x40000000,-8
+	.fmask	0x00000000,0
+	subu	$sp,$sp,16
+	sw	$fp,8($sp)
+	move	$fp,$sp
+	sw	$4,16($fp)
+	lw	$2,16($fp)
+	lw	$3,0($2)
+	addu	$2,$3,10
+	sw	$2,0($fp)
+$L23:
+	lb	$2,Char1Glob
+	li	$3,0x00000041		# 65
+	bne	$2,$3,$L26
+	lw	$3,0($fp)
+	subu	$2,$3,1
+	move	$3,$2
+	sw	$3,0($fp)
+	lw	$2,16($fp)
+	lw	$3,0($fp)
+	lw	$4,IntGlob
+	subu	$3,$3,$4
+	sw	$3,0($2)
+	sw	$0,4($fp)
+$L26:
+	lw	$2,4($fp)
+	beq	$2,$0,$L24
+$L27:
+$L25:
+	j	$L23
+$L24:
+$L22:
+	move	$sp,$fp			# sp not trusted here
+	lw	$fp,8($sp)
+	addu	$sp,$sp,16
+	j	$31
+	.end	Proc2
+	.loc	1 583
+	.ent	Proc3
+Proc3:
+	.frame	$fp,24,$31		# vars= 0, regs= 2/0, args= 16, extra= 0
+	.mask	0xc0000000,-4
+	.fmask	0x00000000,0
+	subu	$sp,$sp,24
+	sw	$31,20($sp)
+	sw	$fp,16($sp)
+	move	$fp,$sp
+	sw	$4,24($fp)
+	lw	$2,PtrGlb
+	beq	$2,$0,$L29
+	lw	$2,24($fp)
+	lw	$3,PtrGlb
+	lw	$4,0($3)
+	sw	$4,0($2)
+	j	$L30
+$L29:
+	li	$2,0x00000064		# 100
+	sw	$2,IntGlob
+$L30:
+	lw	$3,PtrGlb
+	addu	$2,$3,12
+	li	$4,0x0000000a		# 10
+	lw	$5,IntGlob
+	move	$6,$2
+	jal	Proc7
+$L28:
+	move	$sp,$fp			# sp not trusted here
+	lw	$31,20($sp)
+	lw	$fp,16($sp)
+	addu	$sp,$sp,24
+	j	$31
+	.end	Proc3
+	.loc	1 593
+	.ent	Proc4
+Proc4:
+	.frame	$fp,16,$31		# vars= 8, regs= 1/0, args= 0, extra= 0
+	.mask	0x40000000,-8
+	.fmask	0x00000000,0
+	subu	$sp,$sp,16
+	sw	$fp,8($sp)
+	move	$fp,$sp
+	lb	$2,Char1Glob
+	xori	$3,$2,0x0041
+	sltu	$2,$3,1
+	sw	$2,0($fp)
+	lw	$2,0($fp)
+	lw	$3,BoolGlob
+	or	$2,$2,$3
+	sw	$2,0($fp)
+	li	$2,0x00000042		# 66
+	sb	$2,Char2Glob
+$L31:
+	move	$sp,$fp			# sp not trusted here
+	lw	$fp,8($sp)
+	addu	$sp,$sp,16
+	j	$31
+	.end	Proc4
+	.loc	1 602
+	.ent	Proc5
+Proc5:
+	.frame	$fp,8,$31		# vars= 0, regs= 1/0, args= 0, extra= 0
+	.mask	0x40000000,-8
+	.fmask	0x00000000,0
+	subu	$sp,$sp,8
+	sw	$fp,0($sp)
+	move	$fp,$sp
+	li	$2,0x00000041		# 65
+	sb	$2,Char1Glob
+	sw	$0,BoolGlob
+$L32:
+	move	$sp,$fp			# sp not trusted here
+	lw	$fp,0($sp)
+	addu	$sp,$sp,8
+	j	$31
+	.end	Proc5
+	.loc	1 610
+	.ent	Proc6
+Proc6:
+	.frame	$fp,24,$31		# vars= 0, regs= 2/0, args= 16, extra= 0
+	.mask	0xc0000000,-4
+	.fmask	0x00000000,0
+	subu	$sp,$sp,24
+	sw	$31,20($sp)
+	sw	$fp,16($sp)
+	move	$fp,$sp
+	sw	$4,24($fp)
+	sw	$5,28($fp)
+	lw	$2,28($fp)
+	lw	$3,24($fp)
+	sw	$3,0($2)
+	lw	$4,24($fp)
+	jal	Func3
+	bne	$2,$0,$L34
+	lw	$2,28($fp)
+	li	$3,0x00000003		# 3
+	sw	$3,0($2)
+$L34:
+	lw	$2,24($fp)
+	sltu	$3,$2,5
+	beq	$3,$0,$L44
+	lw	$2,24($fp)
+	move	$3,$2
+	sll	$2,$3,2
+	la	$3,$L43
+	addu	$2,$2,$3
+	lw	$3,0($2)
+	j	$3
+	.rdata
+	.align	3
+	.align	2
+$L43:
+	.word	$L36
+	.word	$L37
+	.word	$L40
+	.word	$L41
+	.word	$L42
+	.text
+$L36:
+	lw	$2,28($fp)
+	sw	$0,0($2)
+	j	$L35
+$L37:
+	lw	$2,IntGlob
+	slt	$3,$2,101
+	bne	$3,$0,$L38
+	lw	$2,28($fp)
+	sw	$0,0($2)
+	j	$L39
+$L38:
+	lw	$2,28($fp)
+	li	$3,0x00000003		# 3
+	sw	$3,0($2)
+$L39:
+	j	$L35
+$L40:
+	lw	$2,28($fp)
+	li	$3,0x00000001		# 1
+	sw	$3,0($2)
+	j	$L35
+$L41:
+	j	$L35
+$L42:
+	lw	$2,28($fp)
+	li	$3,0x00000002		# 2
+	sw	$3,0($2)
+$L44:
+$L35:
+$L33:
+	move	$sp,$fp			# sp not trusted here
+	lw	$31,20($sp)
+	lw	$fp,16($sp)
+	addu	$sp,$sp,24
+	j	$31
+	.end	Proc6
+	.loc	1 629
+	.ent	Proc7
+Proc7:
+	.frame	$fp,16,$31		# vars= 8, regs= 1/0, args= 0, extra= 0
+	.mask	0x40000000,-8
+	.fmask	0x00000000,0
+	subu	$sp,$sp,16
+	sw	$fp,8($sp)
+	move	$fp,$sp
+	sw	$4,16($fp)
+	sw	$5,20($fp)
+	sw	$6,24($fp)
+	lw	$2,16($fp)
+	addu	$3,$2,2
+	sw	$3,0($fp)
+	lw	$2,24($fp)
+	lw	$3,20($fp)
+	lw	$4,0($fp)
+	addu	$3,$3,$4
+	sw	$3,0($2)
+$L45:
+	move	$sp,$fp			# sp not trusted here
+	lw	$fp,8($sp)
+	addu	$sp,$sp,16
+	j	$31
+	.end	Proc7
+	.loc	1 640
+	.ent	Proc8
+Proc8:
+	.frame	$fp,16,$31		# vars= 8, regs= 1/0, args= 0, extra= 0
+	.mask	0x40000000,-8
+	.fmask	0x00000000,0
+	subu	$sp,$sp,16
+	sw	$fp,8($sp)
+	move	$fp,$sp
+	sw	$4,16($fp)
+	sw	$5,20($fp)
+	sw	$6,24($fp)
+	sw	$7,28($fp)
+	lw	$2,24($fp)
+	addu	$3,$2,5
+	sw	$3,0($fp)
+	lw	$2,0($fp)
+	move	$3,$2
+	sll	$2,$3,2
+	lw	$3,16($fp)
+	addu	$2,$2,$3
+	lw	$3,28($fp)
+	sw	$3,0($2)
+	lw	$2,0($fp)
+	move	$3,$2
+	sll	$2,$3,2
+	lw	$4,16($fp)
+	addu	$3,$2,$4
+	addu	$2,$3,4
+	lw	$3,0($fp)
+	move	$4,$3
+	sll	$3,$4,2
+	lw	$4,16($fp)
+	addu	$3,$3,$4
+	lw	$4,0($3)
+	sw	$4,0($2)
+	lw	$2,0($fp)
+	move	$3,$2
+	sll	$2,$3,2
+	lw	$3,16($fp)
+	addu	$2,$2,$3
+	addu	$3,$2,120
+	lw	$2,0($fp)
+	sw	$2,0($3)
+	lw	$2,0($fp)
+	sw	$2,4($fp)
+$L47:
+	lw	$3,0($fp)
+	addu	$2,$3,1
+	lw	$3,4($fp)
+	slt	$2,$2,$3
+	bne	$2,$0,$L48
+$L50:
+	lw	$3,0($fp)
+	lw	$2,4($fp)
+	move	$4,$2
+	sll	$2,$4,2
+	move	$5,$3
+	sll	$4,$5,1
+	addu	$4,$4,$3
+	sll	$3,$4,4
+	addu	$4,$4,$3
+	sll	$3,$4,2
+	lw	$4,20($fp)
+	addu	$3,$3,$4
+	addu	$2,$2,$3
+	lw	$3,0($fp)
+	sw	$3,0($2)
+$L49:
+	lw	$3,4($fp)
+	addu	$2,$3,1
+	move	$3,$2
+	sw	$3,4($fp)
+	j	$L47
+$L48:
+	lw	$3,0($fp)
+	lw	$2,0($fp)
+	move	$4,$2
+	sll	$2,$4,2
+	move	$5,$3
+	sll	$4,$5,1
+	addu	$4,$4,$3
+	sll	$3,$4,4
+	addu	$4,$4,$3
+	sll	$3,$4,2
+	lw	$4,20($fp)
+	addu	$3,$3,$4
+	addu	$4,$2,$3
+	subu	$2,$4,4
+	lw	$3,0($fp)
+	lw	$2,0($fp)
+	move	$4,$2
+	sll	$2,$4,2
+	move	$5,$3
+	sll	$4,$5,1
+	addu	$4,$4,$3
+	sll	$3,$4,4
+	addu	$4,$4,$3
+	sll	$3,$4,2
+	lw	$4,20($fp)
+	addu	$3,$3,$4
+	addu	$4,$2,$3
+	subu	$2,$4,4
+	lw	$4,0($fp)
+	lw	$3,0($fp)
+	move	$5,$3
+	sll	$3,$5,2
+	move	$6,$4
+	sll	$5,$6,1
+	addu	$5,$5,$4
+	sll	$4,$5,4
+	addu	$5,$5,$4
+	sll	$4,$5,2
+	lw	$5,20($fp)
+	addu	$4,$4,$5
+	addu	$3,$3,$4
+	subu	$4,$3,4
+	lw	$5,0($4)
+	addu	$3,$5,1
+	move	$4,$3
+	sw	$4,0($2)
+	lw	$3,0($fp)
+	lw	$2,0($fp)
+	move	$4,$2
+	sll	$2,$4,2
+	move	$5,$3
+	sll	$4,$5,1
+	addu	$4,$4,$3
+	sll	$3,$4,4
+	addu	$4,$4,$3
+	sll	$3,$4,2
+	lw	$4,20($fp)
+	addu	$3,$3,$4
+	addu	$4,$2,$3
+	addu	$2,$4,4080
+	lw	$3,0($fp)
+	move	$4,$3
+	sll	$3,$4,2
+	lw	$4,16($fp)
+	addu	$3,$3,$4
+	lw	$4,0($3)
+	sw	$4,0($2)
+	li	$2,0x00000005		# 5
+	sw	$2,IntGlob
+$L46:
+	move	$sp,$fp			# sp not trusted here
+	lw	$fp,8($sp)
+	addu	$sp,$sp,16
+	j	$31
+	.end	Proc8
+	.loc	1 660
+	.ent	Func1
+Func1:
+	.frame	$fp,16,$31		# vars= 8, regs= 1/0, args= 0, extra= 0
+	.mask	0x40000000,-8
+	.fmask	0x00000000,0
+	subu	$sp,$sp,16
+	sw	$fp,8($sp)
+	move	$fp,$sp
+	move	$3,$4
+	move	$4,$5
+	sb	$3,0($fp)
+	sb	$4,1($fp)
+	lbu	$2,0($fp)
+	sb	$2,2($fp)
+	lbu	$2,2($fp)
+	sb	$2,3($fp)
+	lb	$2,3($fp)
+	lb	$5,1($fp)
+	beq	$2,$5,$L52
+	move	$2,$0
+	j	$L51
+	j	$L53
+$L52:
+	li	$2,0x00000001		# 1
+	j	$L51
+$L53:
+$L51:
+	move	$sp,$fp			# sp not trusted here
+	lw	$fp,8($sp)
+	addu	$sp,$sp,16
+	j	$31
+	.end	Func1
+	.loc	1 675
+	.ent	Func2
+Func2:
+	.frame	$fp,32,$31		# vars= 8, regs= 2/0, args= 16, extra= 0
+	.mask	0xc0000000,-4
+	.fmask	0x00000000,0
+	subu	$sp,$sp,32
+	sw	$31,28($sp)
+	sw	$fp,24($sp)
+	move	$fp,$sp
+	sw	$4,32($fp)
+	sw	$5,36($fp)
+	li	$2,0x00000001		# 1
+	sw	$2,16($fp)
+$L55:
+	lw	$2,16($fp)
+	slt	$3,$2,2
+	beq	$3,$0,$L56
+$L57:
+	lw	$2,32($fp)
+	lw	$4,16($fp)
+	addu	$3,$2,$4
+	lb	$2,0($3)
+	lw	$3,16($fp)
+	lw	$4,36($fp)
+	addu	$3,$3,$4
+	addu	$4,$3,1
+	lb	$3,0($4)
+	move	$4,$2
+	move	$5,$3
+	jal	Func1
+	bne	$2,$0,$L58
+	li	$2,0x00000041		# 65
+	sb	$2,20($fp)
+	lw	$3,16($fp)
+	addu	$2,$3,1
+	move	$3,$2
+	sw	$3,16($fp)
+$L58:
+	j	$L55
+$L56:
+	lb	$2,20($fp)
+	slt	$3,$2,87
+	bne	$3,$0,$L59
+	lb	$2,20($fp)
+	slt	$3,$2,91
+	beq	$3,$0,$L59
+	li	$2,0x00000007		# 7
+	sw	$2,16($fp)
+$L59:
+	lb	$2,20($fp)
+	li	$3,0x00000058		# 88
+	bne	$2,$3,$L60
+	li	$2,0x00000001		# 1
+	j	$L54
+	j	$L61
+$L60:
+	lw	$4,32($fp)
+	lw	$5,36($fp)
+	jal	strcmp
+	blez	$2,$L62
+	lw	$2,16($fp)
+	addu	$3,$2,7
+	sw	$3,16($fp)
+	li	$2,0x00000001		# 1
+	j	$L54
+	j	$L63
+$L62:
+	move	$2,$0
+	j	$L54
+$L63:
+$L61:
+$L54:
+	move	$sp,$fp			# sp not trusted here
+	lw	$31,28($sp)
+	lw	$fp,24($sp)
+	addu	$sp,$sp,32
+	j	$31
+	.end	Func2
+	.loc	1 705
+	.ent	Func3
+Func3:
+	.frame	$fp,16,$31		# vars= 8, regs= 1/0, args= 0, extra= 0
+	.mask	0x40000000,-8
+	.fmask	0x00000000,0
+	subu	$sp,$sp,16
+	sw	$fp,8($sp)
+	move	$fp,$sp
+	sw	$4,16($fp)
+	lw	$2,16($fp)
+	sw	$2,0($fp)
+	lw	$2,0($fp)
+	li	$3,0x00000002		# 2
+	bne	$2,$3,$L65
+	li	$2,0x00000001		# 1
+	j	$L64
+$L65:
+	move	$2,$0
+	j	$L64
+$L64:
+	move	$sp,$fp			# sp not trusted here
+	lw	$fp,8($sp)
+	addu	$sp,$sp,16
+	j	$31
+	.end	Func3

benchmarks/optimized/pi.s

@@ -0,0 +1,130 @@
+	.file	1 "pi.c"
+# GNU C 2.7.2.3 [AL 1.1, MM 40, tma 0.1] SimpleScalar running sstrix compiled by GNU C
+# Cc1 defaults:
+# -mgas -mgpOPT
+# Cc1 arguments (-G value = 8, Cpu = default, ISA = 1):
+# -quiet -dumpbase -o
+gcc2_compiled.:
+__gnu_compiled_c:
+	.rdata
+	.align	2
+$LC0:
+	.ascii	"Usage: %s <iterations>\n\000"
+	.sdata
+	.align	2
+$LC3:
+	.ascii	"%.10f\n\000"
+	.align	3
+$LC1:
+	.word	0xffc00000		# 2147483647
+	.word	0x41dfffff
+	.align	3
+$LC2:
+	.word	0x00000000		# 1
+	.word	0x3ff00000
+	.align	3
+$LC4:
+	.word	0x00000000		# 4
+	.word	0x40100000
+	.text
+	.align	2
+	.globl	main
+	.extern	stderr, 4
+	.text
+	.loc	1 5
+	.ent	main
+main:
+	.frame	$fp,56,$31		# vars= 32, regs= 2/0, args= 16, extra= 0
+	.mask	0xc0000000,-4
+	.fmask	0x00000000,0
+	subu	$sp,$sp,56
+	sw	$31,52($sp)
+	sw	$fp,48($sp)
+	move	$fp,$sp
+	sw	$4,56($fp)
+	sw	$5,60($fp)
+	jal	__main
+	sw	$0,24($fp)
+	lw	$2,56($fp)
+	li	$3,0x00000002		# 2
+	beq	$2,$3,$L2
+	lw	$2,60($fp)
+	lw	$4,stderr
+	la	$5,$LC0
+	lw	$6,0($2)
+	jal	fprintf
+	move	$4,$0
+	jal	exit
+$L2:
+	lw	$3,60($fp)
+	addu	$2,$3,4
+	lw	$4,0($2)
+	jal	atoi
+	sw	$2,20($fp)
+	li	$4,0x00000001		# 1
+	jal	srandom
+	sw	$0,16($fp)
+$L3:
+	lw	$2,16($fp)
+	lw	$3,20($fp)
+	slt	$2,$2,$3
+	bne	$2,$0,$L6
+	j	$L4
+$L6:
+	jal	random
+	mtc1	$2,$f0
+	#nop
+	cvt.d.w	$f0,$f0
+	l.d	$f2,$LC1
+	div.d	$f0,$f0,$f2
+	s.d	$f0,32($fp)
+	jal	random
+	mtc1	$2,$f0
+	#nop
+	cvt.d.w	$f0,$f0
+	l.d	$f2,$LC1
+	div.d	$f0,$f0,$f2
+	s.d	$f0,40($fp)
+	l.d	$f0,32($fp)
+	l.d	$f2,32($fp)
+	mul.d	$f0,$f0,$f2
+	l.d	$f2,40($fp)
+	l.d	$f4,40($fp)
+	mul.d	$f2,$f2,$f4
+	add.d	$f0,$f0,$f2
+	l.d	$f2,$LC2
+	c.le.d	$f0,$f2
+	bc1f	$L7
+	lw	$3,24($fp)
+	addu	$2,$3,1
+	move	$3,$2
+	sw	$3,24($fp)
+$L7:
+$L5:
+	lw	$3,16($fp)
+	addu	$2,$3,1
+	move	$3,$2
+	sw	$3,16($fp)
+	j	$L3
+$L4:
+	l.s	$f0,24($fp)
+	#nop
+	cvt.d.w	$f0,$f0
+	l.s	$f2,20($fp)
+	#nop
+	cvt.d.w	$f2,$f2
+	div.d	$f0,$f0,$f2
+	l.d	$f2,$LC4
+	mul.d	$f0,$f0,$f2
+	la	$4,$LC3
+	dmfc1	$6,$f0
+	jal	printf
+	li	$2,0x00000001		# 1
+	j	$L1
+$L1:
+	move	$sp,$fp			# sp not trusted here
+	lw	$31,52($sp)
+	lw	$fp,48($sp)
+	addu	$sp,$sp,56
+	j	$31
+	.end	main

benchmarks/optimized/slalom.s

@@ -0,0 +1,4630 @@
+	.file	1 "slalom.c"
+# GNU C 2.7.2.3 [AL 1.1, MM 40, tma 0.1] SimpleScalar running sstrix compiled by GNU C
+# Cc1 defaults:
+# -mgas -mgpOPT
+# Cc1 arguments (-G value = 8, Cpu = default, ISA = 1):
+# -quiet -dumpbase -O0 -o
+gcc2_compiled.:
+__gnu_compiled_c:
+	.rdata
+	.align	2
+$LC0:
+	.ascii	"Enter the number of seconds that is the goal: \000"
+	.sdata
+	.align	2
+$LC1:
+	.ascii	"%lg\000"
+	.rdata
+	.align	2
+$LC2:
+	.ascii	"Enter a lower bound for n: \000"
+	.sdata
+	.align	2
+$LC3:
+	.ascii	"%d\000"
+	.rdata
+	.align	2
+$LC4:
+	.ascii	"Must take less than %g seconds.  Took %g.\n\000"
+	.align	2
+$LC5:
+	.ascii	"Enter an upper bound for n: \000"
+	.align	2
+$LC6:
+	.ascii	"Must take at least %g seconds.  Took %g.\n\000"
+	.align	2
+$LC7:
+	.ascii	"New interval: [%d,%d]\n\000"
+	.text
+	.align	2
+	.globl	main
+	.data
+	.align	2
+info.4:
+	.word	$LC8
+	.word	$LC9
+	.word	$LC10
+	.word	$LC11
+	.word	$LC12
+	.word	$LC13
+	.word	$LC14
+	.word	$LC15
+	.word	$LC16
+	.word	0
+	.rdata
+	.align	2
+$LC16:
+	.ascii	"Options:  -O                   Date:       23 May 1990\000"
+	.align	2
+$LC15:
+	.ascii	"Compiler: cc                   Run by:     M. Carter\000"
+	.align	2
+$LC14:
+	.ascii	"Language: C                    Alone:      yes\000"
+	.align	2
+$LC13:
+	.ascii	"OS:       SUNOS 4.0.3          Timer:      Wall, gettime"
+	.ascii	"ofday()\000"
+	.align	2
+$LC12:
+	.ascii	"Disk:     .3GB SCSI+.7GB SMD   Node name:  amssun2\000"
+	.align	2
+$LC11:
+	.ascii	"NMAX:     512                  Clock:      25 MHz\000"
+	.align	2
+$LC10:
+	.ascii	"Cache:    128 KB               # used:     1\000"
+	.align	2
+$LC9:
+	.ascii	"Memory:   32 MB                # of procs: 1\000"
+	.align	2
+$LC8:
+	.ascii	"Machine:  SUN 4/370GX          Processor:  SPARC\000"
+	.sdata
+	.align	2
+$LC17:
+	.ascii	"\n\000"
+	.rdata
+	.align	2
+$LC18:
+	.ascii	"M ops:    %-13lg        Time:       %-.3lf seconds\n\000"
+	.align	2
+$LC20:
+	.ascii	"n:        %-6d               MFLOPS:     %-.5lg\n\000"
+	.align	2
+$LC21:
+	.ascii	"Approximate data memory use: %d bytes.\n\000"
+	.sdata
+	.align	3
+$LC19:
+	.word	0xa0b5ed8d		# 9.9999999999999995475e-07
+	.word	0x3eb0c6f7
+	.text
+	.align	2
+	.globl	What
+	.lcomm	area.7,16384
+	.lcomm	box.8,56
+	.lcomm	coeff.9,33554432
+	.lcomm	diag.10,49152
+	.lcomm	emiss.11,144
+	.lcomm	place.12,49152
+	.lcomm	result.13,49152
+	.lcomm	rho.14,144
+	.lcomm	rhs.15,49152
+	.lcomm	size.16,32768
+	.lcomm	loop.17,48
+	.data
+	.align	2
+tasks.18:
+	.word	$LC22
+	.word	$LC23
+	.word	$LC24
+	.word	$LC25
+	.word	$LC26
+	.word	$LC27
+	.word	$LC28
+	.sdata
+	.align	2
+$LC28:
+	.ascii	"Storer\000"
+	.align	2
+$LC27:
+	.ascii	"Solver\000"
+	.align	2
+$LC26:
+	.ascii	"SetUp3\000"
+	.align	2
+$LC25:
+	.ascii	"SetUp2\000"
+	.align	2
+$LC24:
+	.ascii	"SetUp1\000"
+	.align	2
+$LC23:
+	.ascii	"Region\000"
+	.align	2
+$LC22:
+	.ascii	"Reader\000"
+	.rdata
+	.align	2
+$LC29:
+	.ascii	"%6.6s%8.3f%17.0f%14.6f%10.1f %%\n\000"
+	.sdata
+	.align	2
+format.19:
+	.word	$LC29
+	.rdata
+	.align	2
+$LC30:
+	.ascii	"Must be at least 6, the number of faces.\n\000"
+	.align	2
+$LC31:
+	.ascii	"Exceeds %d = maximum for this system.\n\000"
+	.sdata
+	.align	2
+$LC32:
+	.ascii	"answer\000"
+	.rdata
+	.align	2
+$LC50:
+	.ascii	"%d patches:\n\000"
+	.align	2
+$LC51:
+	.ascii	" Task  Seconds       Operations        MFLOPS    %% of T"
+	.ascii	"ime\n\000"
+	.sdata
+	.align	2
+$LC55:
+	.ascii	"TOTALS\000"
+	.align	3
+$LC33:
+	.word	0x00000000		# 0.5
+	.word	0x3fe00000
+	.align	3
+$LC34:
+	.word	0x00000000		# 258
+	.word	0x40702000
+	.align	3
+$LC35:
+	.word	0x00000000		# 8
+	.word	0x40200000
+	.align	3
+$LC36:
+	.word	0x00000000		# 154
+	.word	0x40634000
+	.align	3
+$LC37:
+	.word	0x00000000		# 532
+	.word	0x4080a000
+	.align	3
+$LC38:
+	.word	0x00000000		# 6
+	.word	0x40180000
+	.align	3
+$LC39:
+	.word	0x00000000		# 370
+	.word	0x40772000
+	.align	3
+$LC40:
+	.word	0x00000000		# 2
+	.word	0x40000000
+	.align	3
+$LC41:
+	.word	0x00000000		# 9
+	.word	0x40220000
+	.align	3
+$LC42:
+	.word	0x00000000		# 72
+	.word	0x40520000
+	.align	3
+$LC43:
+	.word	0x00000000		# 7.5
+	.word	0x401e0000
+	.align	3
+$LC44:
+	.word	0x00000000		# 2.5
+	.word	0x40040000
+	.align	3
+$LC45:
+	.word	0x00000000		# 21
+	.word	0x40350000
+	.align	3
+$LC46:
+	.word	0x00000000		# 1
+	.word	0x3ff00000
+	.align	3
+$LC47:
+	.word	0x00000000		# 16.5
+	.word	0x40308000
+	.align	3
+$LC48:
+	.word	0x00000000		# 35.5
+	.word	0x4041c000
+	.align	3
+$LC49:
+	.word	0x00000000		# 3
+	.word	0x40080000
+	.align	3
+$LC52:
+	.word	0xd2f1a9fc		# 0.0010000000000000000208
+	.word	0x3f50624d
+	.align	3
+$LC53:
+	.word	0xa0b5ed8d		# 9.9999999999999995475e-07
+	.word	0x3eb0c6f7
+	.align	3
+$LC54:
+	.word	0x00000000		# 100
+	.word	0x40590000
+	.text
+	.align	2
+	.globl	Meter
+	.sdata
+	.align	3
+$LC56:
+	.word	0xa0b5ed8d		# 9.9999999999999995475e-07
+	.word	0x3eb0c6f7
+	.text
+	.align	2
+	.globl	When
+	.sdata
+	.align	2
+$LC57:
+	.ascii	"geom\000"
+	.align	2
+$LC58:
+	.ascii	"r\000"
+	.rdata
+	.align	2
+$LC59:
+	.ascii	"slalom:  'geom' geometry file not found.\n\000"
+	.align	2
+$LC60:
+	.ascii	"Must specify exactly 3 box coordinates.\n\000"
+	.align	2
+$LC61:
+	.ascii	"Must specify exactly 18 box coordinates.\n\000"
+	.align	2
+$LC64:
+	.ascii	"Box dimensions must be between 1 and 100.\n\000"
+	.align	2
+$LC66:
+	.ascii	"Reflectivities must be between .000 and .999.\n\000"
+	.align	2
+$LC67:
+	.ascii	"Emissivity cannot be negative.\n\000"
+	.align	2
+$LC68:
+	.ascii	"Emissivities are zero.  Problem is trivial.\n\000"
+	.sdata
+	.align	3
+$LC62:
+	.word	0x00000000		# 1
+	.word	0x3ff00000
+	.align	3
+$LC63:
+	.word	0x00000000		# 100
+	.word	0x40590000
+	.align	3
+$LC65:
+	.word	0xd916872b		# 0.99899999999999999911
+	.word	0x3feff7ce
+	.text
+	.align	2
+	.globl	Reader
+	.rdata
+	.align	2
+$LC70:
+	.ascii	"Eccentric box requires more patches.\n\000"
+	.sdata
+	.align	3
+$LC69:
+	.word	0x00000000		# 0.5
+	.word	0x3fe00000
+	.align	3
+$LC71:
+	.word	0x54442d18		# 25.132741228718344928
+	.word	0x403921fb
+	.text
+	.align	2
+	.globl	Region
+	.sdata
+	.align	3
+$LC72:
+	.word	0x00000000		# 1
+	.word	0x3ff00000
+	.align	3
+$LC73:
+	.word	0x00000000		# -4
+	.word	0xc0100000
+	.text
+	.align	2
+	.globl	SetUp1
+	.sdata
+	.align	3
+$LC74:
+	.word	0xb7f87a0f		# 1.0000000000000000079e-35
+	.word	0x38aa95a5
+	.align	3
+$LC75:
+	.word	0x00000000		# 1
+	.word	0x3ff00000
+	.align	3
+$LC76:
+	.word	0x00000000		# 4
+	.word	0x40100000
+	.text
+	.align	2
+	.globl	SetUp2
+	.rdata
+	.align	2
+$LC79:
+	.ascii	"Total form factor is too far from unity.\n\000"
+	.sdata
+	.align	3
+$LC77:
+	.word	0x00000000		# 1
+	.word	0x3ff00000
+	.align	3
+$LC78:
+	.word	0xe826d695		# 5.0000000000000003114e-10
+	.word	0x3e012e0b
+	.text
+	.align	2
+	.globl	SetUp3
+	.sdata
+	.align	3
+$LC80:
+	.word	0x00000000		# 1
+	.word	0x3ff00000
+	.text
+	.align	2
+	.globl	Solver
+	.sdata
+	.align	2
+$LC81:
+	.ascii	"w\000"
+	.rdata
+	.align	2
+$LC82:
+	.ascii	"Unable to open 'answer' file.\n\000"
+	.align	2
+$LC83:
+	.ascii	" Patch  Face       Position in w, h, d              Widt"
+	.ascii	"h     Height\n\000"
+	.align	2
+$LC84:
+	.ascii	"%5d   %4d%11.5lf%11.5lf%11.5lf  %11.5lf%11.5lf\n\000"
+	.align	2
+$LC85:
+	.ascii	"\n"
+	.ascii	" Patch  Face  Radiosities\n\000"
+	.align	2
+$LC86:
+	.ascii	"%5d   %4d%12.8lf%12.8lf%12.8lf\n\000"
+	.text
+	.align	2
+	.globl	Storer
+	.rdata
+	.align	2
+$LC88:
+	.ascii	"Residual is too large: %lg\n\000"
+	.sdata
+	.align	3
+$LC87:
+	.word	0x29a4692c		# 1.5000000000000001968e-08
+	.word	0x3e501b2b
+	.text
+	.align	2
+	.globl	Verify
+	.align	2
+	.globl	Ddot
+	.comm	goal,8
+	.comm	timing,8
+	.comm	work,8
+	.comm	mean,4
+	.comm	n,4
+	.comm	nupper,4
+	.text
+	.loc	1 179
+	.ent	main
+main:
+	.frame	$fp,40,$31		# vars= 8, regs= 2/0, args= 24, extra= 0
+	.mask	0xc0000000,-4
+	.fmask	0x00000000,0
+	subu	$sp,$sp,40
+	sw	$31,36($sp)
+	sw	$fp,32($sp)
+	move	$fp,$sp
+	jal	__main
+	la	$4,$LC0
+	jal	printf
+	la	$4,$LC1
+	la	$5,goal
+	jal	scanf
+$L2:
+	la	$4,$LC2
+	jal	printf
+	la	$4,$LC3
+	la	$5,n
+	jal	scanf
+	lw	$2,n
+	bgtz	$2,$L5
+	move	$4,$0
+	jal	exit
+$L5:
+	lw	$4,n
+	la	$5,timing
+	la	$6,work
+	jal	Meter
+	sw	$2,24($fp)
+	l.d	$f0,timing
+	l.d	$f2,goal
+	c.le.d	$f2,$f0
+	bc1f	$L6
+	l.d	$f0,timing
+	s.d	$f0,16($sp)
+	la	$4,$LC4
+	dlw	$6,goal
+	jal	printf
+$L6:
+$L4:
+	lw	$2,24($fp)
+	beq	$2,$0,$L7
+	l.d	$f0,timing
+	l.d	$f2,goal
+	c.le.d	$f2,$f0
+	bc1t	$L7
+	j	$L3
+$L7:
+	j	$L2
+$L3:
+	.set	noreorder
+	nop
+	.set	reorder
+$L8:
+	la	$4,$LC5
+	jal	printf
+	la	$4,$LC3
+	la	$5,nupper
+	jal	scanf
+	lw	$2,nupper
+	bgtz	$2,$L11
+	move	$4,$0
+	jal	exit
+$L11:
+	lw	$4,nupper
+	la	$5,timing
+	la	$6,work
+	jal	Meter
+	sw	$2,24($fp)
+	l.d	$f0,timing
+	l.d	$f2,goal
+	c.lt.d	$f0,$f2
+	bc1f	$L12
+	l.d	$f0,timing
+	s.d	$f0,16($sp)
+	la	$4,$LC6
+	dlw	$6,goal
+	jal	printf
+	lw	$2,nupper
+	lw	$3,n
+	slt	$4,$2,$3
+	beq	$4,$0,$L13
+	lw	$2,n
+$L13:
+	sw	$2,n
+$L12:
+$L10:
+	lw	$2,24($fp)
+	beq	$2,$0,$L14
+	l.d	$f0,timing
+	l.d	$f2,goal
+	c.lt.d	$f0,$f2
+	bc1t	$L14
+	j	$L9
+$L14:
+	j	$L8
+$L9:
+	.set	noreorder
+	nop
+	.set	reorder
+$L15:
+	lw	$2,nupper
+	lw	$3,n
+	subu	$2,$2,$3
+	slt	$3,$2,2
+	bne	$3,$0,$L16
+$L17:
+	lw	$2,n
+	lw	$3,nupper
+	addu	$2,$2,$3
+	sra	$3,$2,31
+	srl	$4,$3,31
+	addu	$2,$2,$4
+	sra	$3,$2,1
+	sw	$3,mean
+	lw	$4,mean
+	la	$5,timing
+	la	$6,work
+	jal	Meter
+	sw	$2,24($fp)
+	l.d	$f0,timing
+	l.d	$f2,goal
+	c.lt.d	$f0,$f2
+	bc1f	$L18
+	lw	$2,mean
+	sw	$2,n
+	j	$L19
+$L18:
+	lw	$2,mean
+	sw	$2,nupper
+$L19:
+	la	$4,$LC7
+	lw	$5,n
+	lw	$6,nupper
+	jal	printf
+	j	$L15
+$L16:
+	lw	$4,n
+	la	$5,timing
+	la	$6,work
+	jal	Meter
+	sw	$2,24($fp)
+	l.d	$f0,work
+	s.d	$f0,16($sp)
+	lw	$4,n
+	dlw	$6,timing
+	jal	What
+$L1:
+	move	$sp,$fp			# sp not trusted here
+	lw	$31,36($sp)
+	lw	$fp,32($sp)
+	addu	$sp,$sp,40
+	j	$31
+	.end	main
+	.loc	1 236
+	.ent	What
+What:
+	.frame	$fp,40,$31		# vars= 8, regs= 2/0, args= 24, extra= 0
+	.mask	0xc0000000,-4
+	.fmask	0x00000000,0
+	subu	$sp,$sp,40
+	sw	$31,36($sp)
+	sw	$fp,32($sp)
+	move	$fp,$sp
+	sw	$4,40($fp)
+	dsw	$6,48($fp)
+	la	$4,$LC17
+	jal	printf
+	sw	$0,24($fp)
+$L21:
+	lw	$2,24($fp)
+	move	$3,$2
+	sll	$2,$3,2
+	la	$3,info.4
+	addu	$2,$2,$3
+	lw	$3,0($2)
+	bne	$3,$0,$L24
+	j	$L22
+$L24:
+	lw	$2,24($fp)
+	move	$3,$2
+	sll	$2,$3,2
+	la	$3,info.4
+	addu	$2,$2,$3
+	lw	$4,0($2)
+	jal	puts
+$L23:
+	lw	$3,24($fp)
+	addu	$2,$3,1
+	move	$3,$2
+	sw	$3,24($fp)
+	j	$L21
+$L22:
+	l.d	$f0,56($fp)
+	l.d	$f2,$LC19
+	mul.d	$f0,$f0,$f2
+	l.d	$f2,48($fp)
+	s.d	$f2,16($sp)
+	la	$4,$LC18
+	dmfc1	$6,$f0
+	jal	printf
+	l.d	$f0,56($fp)
+	l.d	$f2,48($fp)
+	div.d	$f0,$f0,$f2
+	l.d	$f2,$LC19
+	mul.d	$f0,$f0,$f2
+	la	$4,$LC20
+	lw	$5,40($fp)
+	dmfc1	$6,$f0
+	jal	printf
+	lw	$2,40($fp)
+	move	$3,$2
+	sll	$2,$3,3
+	lw	$3,40($fp)
+	mult	$2,$3
+	mflo	$2
+	lw	$3,40($fp)
+	move	$5,$3
+	sll	$4,$5,4
+	subu	$4,$4,$3
+	sll	$3,$4,3
+	addu	$2,$2,$3
+	addu	$3,$2,800
+	la	$4,$LC21
+	move	$5,$3
+	jal	printf
+$L20:
+	move	$sp,$fp			# sp not trusted here
+	lw	$31,36($sp)
+	lw	$fp,32($sp)
+	addu	$sp,$sp,40
+	j	$31
+	.end	What
+	.loc	1 282
+	.ent	Meter
+Meter:
+	.frame	$fp,280,$31		# vars= 208, regs= 2/3, args= 40, extra= 0
+	.mask	0xc0000000,-28
+	.fmask	0x03f00000,-4
+	subu	$sp,$sp,280
+	sw	$31,252($sp)
+	sw	$fp,248($sp)
+	s.d	$f24,272($sp)
+	s.d	$f22,264($sp)
+	s.d	$f20,256($sp)
+	move	$fp,$sp
+	sw	$4,280($fp)
+	sw	$5,284($fp)
+	sw	$6,288($fp)
+	lw	$2,280($fp)
+	slt	$3,$2,6
+	beq	$3,$0,$L26
+	la	$4,$LC30
+	jal	printf
+	move	$2,$0
+	j	$L25
+	j	$L27
+$L26:
+	lw	$2,280($fp)
+	slt	$3,$2,2049
+	bne	$3,$0,$L28
+	la	$4,$LC31
+	li	$5,0x00000800		# 2048
+	jal	printf
+	move	$2,$0
+	j	$L25
+$L28:
+$L27:
+	la	$4,$LC32
+	jal	unlink
+	jal	When
+	s.d	$f0,152($fp)
+	la	$4,box.8
+	la	$5,rho.14
+	la	$6,emiss.11
+	jal	Reader
+	bne	$2,$0,$L29
+	move	$2,$0
+	j	$L25
+$L29:
+	jal	When
+	s.d	$f0,160($fp)
+	la	$2,size.16
+	sw	$2,16($sp)
+	la	$2,area.7
+	sw	$2,20($sp)
+	lw	$4,280($fp)
+	la	$5,loop.17
+	la	$6,box.8
+	la	$7,place.12
+	jal	Region
+	bne	$2,$0,$L30
+	move	$2,$0
+	j	$L25
+$L30:
+	jal	When
+	s.d	$f0,168($fp)
+	la	$2,size.16
+	sw	$2,16($sp)
+	lw	$4,280($fp)
+	la	$5,loop.17
+	la	$6,coeff.9
+	la	$7,place.12
+	jal	SetUp1
+	jal	When
+	s.d	$f0,176($fp)
+	la	$2,size.16
+	sw	$2,16($sp)
+	lw	$4,280($fp)
+	la	$5,loop.17
+	la	$6,coeff.9
+	la	$7,place.12
+	jal	SetUp2
+	jal	When
+	s.d	$f0,184($fp)
+	la	$2,emiss.11
+	sw	$2,16($sp)
+	la	$2,coeff.9
+	sw	$2,20($sp)
+	la	$2,diag.10
+	sw	$2,24($sp)
+	la	$2,rhs.15
+	sw	$2,28($sp)
+	lw	$4,280($fp)
+	la	$5,loop.17
+	la	$6,area.7
+	la	$7,rho.14
+	jal	SetUp3
+	bne	$2,$0,$L31
+	move	$2,$0
+	j	$L25
+$L31:
+	jal	When
+	s.d	$f0,192($fp)
+	lw	$2,loop.17+8
+	sw	$2,240($fp)
+	la	$2,rhs.15
+	sw	$2,16($sp)
+	la	$2,result.13
+	sw	$2,20($sp)
+	lw	$4,280($fp)
+	lw	$5,240($fp)
+	la	$6,coeff.9
+	la	$7,diag.10
+	jal	Solver
+	jal	When
+	s.d	$f0,200($fp)
+	la	$2,result.13
+	sw	$2,16($sp)
+	lw	$4,280($fp)
+	la	$5,loop.17
+	la	$6,place.12
+	la	$7,size.16
+	jal	Storer
+	jal	When
+	s.d	$f0,208($fp)
+	lw	$2,284($fp)
+	l.d	$f0,208($fp)
+	l.d	$f2,152($fp)
+	sub.d	$f0,$f0,$f2
+	s.d	$f0,0($2)
+	sw	$0,232($fp)
+$L32:
+	lw	$2,232($fp)
+	slt	$3,$2,7
+	bne	$3,$0,$L35
+	j	$L33
+$L35:
+	lw	$2,232($fp)
+	move	$3,$2
+	sll	$2,$3,3
+	addu	$4,$fp,40
+	addu	$3,$2,$4
+	addu	$2,$3,112
+	addu	$3,$fp,160
+	lw	$4,232($fp)
+	move	$5,$4
+	sll	$4,$5,3
+	addu	$3,$3,$4
+	lw	$4,232($fp)
+	move	$5,$4
+	sll	$4,$5,3
+	addu	$5,$fp,40
+	addu	$4,$4,$5
+	addu	$5,$4,112
+	l.d	$f0,0($3)
+	l.d	$f2,0($5)
+	sub.d	$f0,$f0,$f2
+	s.d	$f0,0($2)
+$L34:
+	lw	$3,232($fp)
+	addu	$2,$3,1
+	move	$3,$2
+	sw	$3,232($fp)
+	j	$L32
+$L33:
+	sw	$0,236($fp)
+	dsz	216($fp)
+	sw	$0,232($fp)
+$L36:
+	lw	$2,232($fp)
+	slt	$3,$2,6
+	bne	$3,$0,$L39
+	j	$L37
+$L39:
+	lw	$2,232($fp)
+	move	$3,$2
+	sll	$2,$3,3
+	addu	$4,$fp,40
+	addu	$3,$2,$4
+	addu	$2,$3,64
+	lw	$3,232($fp)
+	move	$4,$3
+	sll	$3,$4,3
+	lw	$4,232($fp)
+	move	$5,$4
+	sll	$4,$5,3
+	lw	$3,loop.17+4($3)
+	lw	$4,loop.17($4)
+	subu	$3,$3,$4
+	addu	$4,$3,1
+	mtc1	$4,$f0
+	#nop
+	cvt.d.w	$f0,$f0
+	s.d	$f0,0($2)
+	lw	$2,232($fp)
+	move	$3,$2
+	sll	$2,$3,3
+	addu	$4,$fp,40
+	addu	$3,$2,$4
+	addu	$2,$3,64
+	lw	$3,232($fp)
+	move	$4,$3
+	sll	$3,$4,3
+	addu	$4,$fp,40
+	addu	$3,$3,$4
+	addu	$4,$3,64
+	l.d	$f0,0($2)
+	l.d	$f2,0($4)
+	mul.d	$f0,$f0,$f2
+	l.d	$f2,216($fp)
+	add.d	$f0,$f2,$f0
+	s.d	$f0,216($fp)
+	l.s	$f24,236($fp)
+	#nop
+	cvt.d.w	$f24,$f24
+	lw	$2,232($fp)
+	move	$3,$2
+	sll	$2,$3,3
+	addu	$4,$fp,40
+	addu	$3,$2,$4
+	addu	$2,$3,64
+	lw	$3,232($fp)
+	move	$4,$3
+	sll	$3,$4,3
+	la	$4,box.8
+	addu	$3,$3,$4
+	l.d	$f0,0($2)
+	l.d	$f2,0($3)
+	mul.d	$f0,$f0,$f2
+	lw	$2,232($fp)
+	move	$3,$2
+	sll	$2,$3,3
+	la	$3,box.8+8
+	addu	$2,$2,$3
+	l.d	$f2,0($2)
+	div.d	$f0,$f0,$f2
+	mov.d	$f12,$f0
+	jal	sqrt
+	l.d	$f2,$LC33
+	add.d	$f0,$f0,$f2
+	add.d	$f2,$f24,$f0
+	trunc.w.d $f10,$f2,$8
+	s.s	$f10,236($fp)
+$L38:
+	lw	$3,232($fp)
+	addu	$2,$3,1
+	move	$3,$2
+	sw	$3,232($fp)
+	j	$L36
+$L37:
+	l.d	$f0,104($fp)
+	l.d	$f2,128($fp)
+	mul.d	$f0,$f0,$f2
+	l.d	$f2,112($fp)
+	l.d	$f4,136($fp)
+	mul.d	$f2,$f2,$f4
+	add.d	$f0,$f0,$f2
+	l.d	$f2,120($fp)
+	l.d	$f4,144($fp)
+	mul.d	$f2,$f2,$f4
+	add.d	$f0,$f0,$f2
+	s.d	$f0,224($fp)
+	l.d	$f0,$LC34
+	s.d	$f0,40($fp)
+	l.s	$f0,236($fp)
+	#nop
+	cvt.d.w	$f0,$f0
+	l.d	$f2,$LC35
+	mul.d	$f0,$f0,$f2
+	l.d	$f2,$LC36
+	add.d	$f0,$f0,$f2
+	l.s	$f2,280($fp)
+	#nop
+	cvt.d.w	$f2,$f2
+	add.d	$f0,$f0,$f2
+	s.d	$f0,48($fp)
+	l.d	$f0,224($fp)
+	l.d	$f2,$LC37
+	mul.d	$f0,$f0,$f2
+	l.d	$f2,$LC38
+	add.d	$f0,$f0,$f2
+	s.d	$f0,56($fp)
+	lw	$2,280($fp)
+	move	$3,$2
+	sll	$2,$3,3
+	mtc1	$2,$f0
+	#nop
+	cvt.d.w	$f0,$f0
+	l.s	$f2,280($fp)
+	#nop
+	cvt.d.w	$f2,$f2
+	l.s	$f4,280($fp)
+	#nop
+	cvt.d.w	$f4,$f4
+	mul.d	$f2,$f2,$f4
+	l.d	$f4,216($fp)
+	sub.d	$f2,$f2,$f4
+	l.d	$f20,224($fp)
+	add.d	$f4,$f20,$f20
+	sub.d	$f2,$f2,$f4
+	l.d	$f4,$LC39
+	mul.d	$f2,$f2,$f4
+	l.d	$f4,$LC40
+	div.d	$f2,$f2,$f4
+	add.d	$f0,$f0,$f2
+	s.d	$f0,64($fp)
+	l.s	$f0,280($fp)
+	#nop
+	cvt.d.w	$f0,$f0
+	l.d	$f2,$LC41
+	mul.d	$f0,$f0,$f2
+	l.d	$f2,$LC42
+	add.d	$f0,$f0,$f2
+	l.s	$f2,280($fp)
+	#nop
+	cvt.d.w	$f2,$f2
+	l.s	$f4,280($fp)
+	#nop
+	cvt.d.w	$f4,$f4
+	mul.d	$f2,$f2,$f4
+	add.d	$f0,$f0,$f2
+	l.d	$f2,216($fp)
+	sub.d	$f0,$f0,$f2
+	s.d	$f0,72($fp)
+	l.s	$f0,280($fp)
+	#nop
+	cvt.d.w	$f0,$f0
+	l.s	$f2,280($fp)
+	#nop
+	cvt.d.w	$f2,$f2
+	l.s	$f4,280($fp)
+	#nop
+	cvt.d.w	$f4,$f4
+	l.d	$f6,$LC43
+	add.d	$f4,$f4,$f6
+	mul.d	$f2,$f2,$f4
+	l.d	$f4,$LC44
+	sub.d	$f2,$f2,$f4
+	mul.d	$f0,$f0,$f2
+	l.d	$f2,$LC45
+	sub.d	$f0,$f0,$f2
+	lw	$3,240($fp)
+	addu	$2,$3,1
+	mtc1	$2,$f2
+	#nop
+	cvt.d.w	$f2,$f2
+	lw	$3,240($fp)
+	addu	$2,$3,1
+	mtc1	$2,$f4
+	#nop
+	cvt.d.w	$f4,$f4
+	l.s	$f22,240($fp)
+	#nop
+	cvt.d.w	$f22,$f22
+	l.d	$f6,$LC46
+	add.d	$f22,$f22,$f6
+	add.d	$f6,$f22,$f22
+	l.d	$f8,$LC47
+	sub.d	$f6,$f6,$f8
+	mul.d	$f4,$f4,$f6
+	l.d	$f6,$LC48
+	add.d	$f4,$f4,$f6
+	mul.d	$f2,$f2,$f4
+	add.d	$f0,$f0,$f2
+	lw	$3,240($fp)
+	addu	$2,$3,1
+	lw	$3,280($fp)
+	mult	$2,$3
+	mflo	$2
+	mtc1	$2,$f2
+	#nop
+	cvt.d.w	$f2,$f2
+	l.s	$f4,240($fp)
+	#nop
+	cvt.d.w	$f4,$f4
+	l.d	$f6,$LC46
+	add.d	$f4,$f4,$f6
+	l.d	$f6,$LC49
+	mul.d	$f4,$f4,$f6
+	l.d	$f6,$LC41
+	sub.d	$f4,$f6,$f4
+	mul.d	$f2,$f2,$f4
+	add.d	$f0,$f0,$f2
+	s.d	$f0,80($fp)
+	lw	$2,280($fp)
+	move	$4,$2
+	sll	$3,$4,1
+	addu	$3,$3,$2
+	sll	$2,$3,4
+	mtc1	$2,$f0
+	#nop
+	cvt.d.w	$f0,$f0
+	s.d	$f0,88($fp)
+	lw	$2,288($fp)
+	l.d	$f0,40($fp)
+	l.d	$f2,48($fp)
+	add.d	$f0,$f0,$f2
+	l.d	$f2,56($fp)
+	add.d	$f0,$f0,$f2
+	l.d	$f2,64($fp)
+	add.d	$f0,$f0,$f2
+	l.d	$f2,72($fp)
+	add.d	$f0,$f0,$f2
+	l.d	$f2,80($fp)
+	add.d	$f0,$f0,$f2
+	l.d	$f2,88($fp)
+	add.d	$f0,$f0,$f2
+	s.d	$f0,0($2)
+	la	$4,$LC50
+	lw	$5,280($fp)
+	jal	printf
+	la	$4,$LC51
+	jal	printf
+	sw	$0,232($fp)
+$L40:
+	lw	$2,232($fp)
+	slt	$3,$2,7
+	bne	$3,$0,$L43
+	j	$L41
+$L43:
+	lw	$2,232($fp)
+	move	$3,$2
+	sll	$2,$3,3
+	addu	$3,$fp,40
+	addu	$2,$2,$3
+	addu	$3,$2,112
+	l.d	$f0,0($3)
+	mtc1	$0,$f2
+	mtc1	$0,$f3
+	c.eq.d	$f0,$f2
+	bc1f	$L44
+	lw	$2,232($fp)
+	move	$3,$2
+	sll	$2,$3,3
+	addu	$3,$fp,40
+	addu	$2,$2,$3
+	addu	$3,$2,112
+	l.d	$f0,$LC52
+	s.d	$f0,0($3)
+$L44:
+	lw	$2,232($fp)
+	move	$3,$2
+	sll	$2,$3,2
+	la	$3,tasks.18
+	addu	$2,$2,$3
+	lw	$3,232($fp)
+	move	$4,$3
+	sll	$3,$4,3
+	addu	$5,$fp,40
+	addu	$4,$3,$5
+	addu	$3,$4,112
+	lw	$4,232($fp)
+	move	$5,$4
+	sll	$4,$5,3
+	addu	$5,$fp,40
+	addu	$4,$4,$5
+	l.d	$f0,0($4)
+	s.d	$f0,16($sp)
+	lw	$4,232($fp)
+	move	$5,$4
+	sll	$4,$5,3
+	addu	$5,$fp,40
+	addu	$4,$4,$5
+	lw	$5,232($fp)
+	move	$6,$5
+	sll	$5,$6,3
+	addu	$6,$fp,40
+	addu	$5,$5,$6
+	addu	$6,$5,112
+	l.d	$f0,0($4)
+	l.d	$f2,0($6)
+	div.d	$f0,$f0,$f2
+	l.d	$f2,$LC53
+	mul.d	$f0,$f0,$f2
+	s.d	$f0,24($sp)
+	lw	$4,232($fp)
+	move	$5,$4
+	sll	$4,$5,3
+	addu	$5,$fp,40
+	addu	$4,$4,$5
+	addu	$5,$4,112
+	l.d	$f0,0($5)
+	l.d	$f2,$LC54
+	mul.d	$f0,$f0,$f2
+	lw	$4,284($fp)
+	l.d	$f2,0($4)
+	div.d	$f0,$f0,$f2
+	s.d	$f0,32($sp)
+	lw	$4,format.19
+	lw	$5,0($2)
+	dlw	$6,0($3)
+	jal	printf
+$L42:
+	lw	$3,232($fp)
+	addu	$2,$3,1
+	move	$3,$2
+	sw	$3,232($fp)
+	j	$L40
+$L41:
+	lw	$2,284($fp)
+	lw	$3,288($fp)
+	l.d	$f0,0($3)
+	s.d	$f0,16($sp)
+	lw	$3,288($fp)
+	lw	$4,284($fp)
+	l.d	$f0,0($3)
+	l.d	$f2,0($4)
+	div.d	$f0,$f0,$f2
+	l.d	$f2,$LC53
+	mul.d	$f0,$f0,$f2
+	s.d	$f0,24($sp)
+	l.d	$f0,$LC54
+	s.d	$f0,32($sp)
+	lw	$4,format.19
+	la	$5,$LC55
+	dlw	$6,0($2)
+	jal	printf
+	la	$2,result.13
+	sw	$2,16($sp)
+	lw	$4,280($fp)
+	la	$5,coeff.9
+	la	$6,diag.10
+	la	$7,rhs.15
+	jal	Verify
+	li	$2,0x00000001		# 1
+	j	$L25
+$L25:
+	move	$sp,$fp			# sp not trusted here
+	lw	$31,252($sp)
+	lw	$fp,248($sp)
+	l.d	$f24,272($sp)
+	l.d	$f22,264($sp)
+	l.d	$f20,256($sp)
+	addu	$sp,$sp,280
+	j	$31
+	.end	Meter
+	.loc	1 396
+	.ent	When
+When:
+	.frame	$fp,40,$31		# vars= 16, regs= 2/0, args= 16, extra= 0
+	.mask	0xc0000000,-4
+	.fmask	0x00000000,0
+	subu	$sp,$sp,40
+	sw	$31,36($sp)
+	sw	$fp,32($sp)
+	move	$fp,$sp
+	addu	$2,$fp,24
+	addu	$4,$fp,16
+	move	$5,$2
+	jal	gettimeofday
+	l.s	$f0,16($fp)
+	#nop
+	cvt.d.w	$f0,$f0
+	l.s	$f2,20($fp)
+	#nop
+	cvt.d.w	$f2,$f2
+	l.d	$f4,$LC56
+	mul.d	$f2,$f2,$f4
+	add.d	$f4,$f0,$f2
+	mov.d	$f0,$f4
+	j	$L45
+$L45:
+	move	$sp,$fp			# sp not trusted here
+	lw	$31,36($sp)
+	lw	$fp,32($sp)
+	addu	$sp,$sp,40
+	j	$31
+	.end	When
+	.loc	1 409
+	.ent	Reader
+Reader:
+	.frame	$fp,144,$31		# vars= 120, regs= 2/0, args= 16, extra= 0
+	.mask	0xc0000000,-4
+	.fmask	0x00000000,0
+	subu	$sp,$sp,144
+	sw	$31,140($sp)
+	sw	$fp,136($sp)
+	move	$fp,$sp
+	sw	$4,144($fp)
+	sw	$5,148($fp)
+	sw	$6,152($fp)
+	la	$4,$LC57
+	la	$5,$LC58
+	jal	fopen
+	sw	$2,40($fp)
+	lw	$2,40($fp)
+	bne	$2,$0,$L47
+	la	$4,$LC59
+	jal	printf
+	li	$4,0x00000001		# 1
+	jal	exit
+$L47:
+	sw	$0,24($fp)
+	sw	$0,16($fp)
+$L48:
+	lw	$2,16($fp)
+	slt	$3,$2,3
+	bne	$3,$0,$L51
+	j	$L49
+$L51:
+	lw	$2,16($fp)
+	move	$3,$2
+	sll	$2,$3,3
+	lw	$3,144($fp)
+	addu	$2,$2,$3
+	lw	$4,40($fp)
+	la	$5,$LC1
+	move	$6,$2
+	jal	fscanf
+	lw	$3,24($fp)
+	addu	$2,$3,$2
+	sw	$2,24($fp)
+$L50:
+	lw	$3,16($fp)
+	addu	$2,$3,1
+	move	$3,$2
+	sw	$3,16($fp)
+	j	$L48
+$L49:
+	addu	$2,$fp,48
+	move	$4,$2
+	li	$5,0x00000050		# 80
+	lw	$6,40($fp)
+	jal	fgets
+	lw	$2,24($fp)
+	li	$3,0x00000003		# 3
+	beq	$2,$3,$L52
+	la	$4,$LC60
+	jal	printf
+	li	$4,0x00000001		# 1
+	jal	exit
+$L52:
+	sw	$0,24($fp)
+	sw	$0,20($fp)
+$L53:
+	lw	$2,20($fp)
+	slt	$3,$2,3
+	bne	$3,$0,$L56
+	j	$L54
+$L56:
+	sw	$0,16($fp)
+$L57:
+	lw	$2,16($fp)
+	slt	$3,$2,6
+	bne	$3,$0,$L60
+	j	$L58
+$L60:
+	lw	$2,16($fp)
+	move	$4,$2
+	sll	$3,$4,1
+	addu	$3,$3,$2
+	sll	$2,$3,3
+	lw	$3,148($fp)
+	addu	$2,$2,$3
+	lw	$3,20($fp)
+	move	$4,$3
+	sll	$3,$4,3
+	addu	$2,$2,$3
+	lw	$4,40($fp)
+	la	$5,$LC1
+	move	$6,$2
+	jal	fscanf
+	lw	$3,24($fp)
+	addu	$2,$3,$2
+	sw	$2,24($fp)
+$L59:
+	lw	$3,16($fp)
+	addu	$2,$3,1
+	move	$3,$2
+	sw	$3,16($fp)
+	j	$L57
+$L58:
+$L55:
+	lw	$3,20($fp)
+	addu	$2,$3,1
+	move	$3,$2
+	sw	$3,20($fp)
+	j	$L53
+$L54:
+	addu	$2,$fp,48
+	move	$4,$2
+	li	$5,0x00000050		# 80
+	lw	$6,40($fp)
+	jal	fgets
+	lw	$2,24($fp)
+	li	$3,0x00000012		# 18
+	beq	$2,$3,$L61
+	la	$4,$LC61
+	jal	printf
+	li	$4,0x00000001		# 1
+	jal	exit
+$L61:
+	sw	$0,24($fp)
+	sw	$0,20($fp)
+$L62:
+	lw	$2,20($fp)
+	slt	$3,$2,3
+	bne	$3,$0,$L65
+	j	$L63
+$L65:
+	sw	$0,16($fp)
+$L66:
+	lw	$2,16($fp)
+	slt	$3,$2,6
+	bne	$3,$0,$L69
+	j	$L67
+$L69:
+	lw	$2,16($fp)
+	move	$4,$2
+	sll	$3,$4,1
+	addu	$3,$3,$2
+	sll	$2,$3,3
+	lw	$3,152($fp)
+	addu	$2,$2,$3
+	lw	$3,20($fp)
+	move	$4,$3
+	sll	$3,$4,3
+	addu	$2,$2,$3
+	lw	$4,40($fp)
+	la	$5,$LC1
+	move	$6,$2
+	jal	fscanf
+	lw	$3,24($fp)
+	addu	$2,$3,$2
+	sw	$2,24($fp)
+$L68:
+	lw	$3,16($fp)
+	addu	$2,$3,1
+	move	$3,$2
+	sw	$3,16($fp)
+	j	$L66
+$L67:
+$L64:
+	lw	$3,20($fp)
+	addu	$2,$3,1
+	move	$3,$2
+	sw	$3,20($fp)
+	j	$L62
+$L63:
+	addu	$2,$fp,48
+	move	$4,$2
+	li	$5,0x00000050		# 80
+	lw	$6,40($fp)
+	jal	fgets
+	lw	$2,24($fp)
+	li	$3,0x00000012		# 18
+	beq	$2,$3,$L70
+	la	$4,$LC61
+	jal	printf
+	li	$4,0x00000001		# 1
+	jal	exit
+$L70:
+	lw	$4,40($fp)
+	jal	fclose
+	sw	$0,20($fp)
+$L71:
+	lw	$2,20($fp)
+	slt	$3,$2,3
+	bne	$3,$0,$L74
+	j	$L72
+$L74:
+	lw	$2,20($fp)
+	move	$3,$2
+	sll	$2,$3,3
+	lw	$3,144($fp)
+	addu	$2,$2,$3
+	l.d	$f0,0($2)
+	l.d	$f2,$LC62
+	c.lt.d	$f0,$f2
+	bc1t	$L76
+	lw	$2,20($fp)
+	move	$3,$2
+	sll	$2,$3,3
+	lw	$3,144($fp)
+	addu	$2,$2,$3
+	l.d	$f0,0($2)
+	l.d	$f2,$LC63
+	c.le.d	$f2,$f0
+	bc1t	$L76
+	j	$L75
+$L76:
+	la	$4,$LC64
+	jal	printf
+	move	$2,$0
+	j	$L46
+$L75:
+	lw	$2,20($fp)
+	move	$3,$2
+	sll	$2,$3,3
+	lw	$4,144($fp)
+	addu	$3,$2,$4
+	addu	$2,$3,24
+	lw	$3,20($fp)
+	move	$4,$3
+	sll	$3,$4,3
+	lw	$4,144($fp)
+	addu	$3,$3,$4
+	l.d	$f0,0($3)
+	s.d	$f0,0($2)
+	dsz	32($fp)
+	sw	$0,16($fp)
+$L77:
+	lw	$2,16($fp)
+	slt	$3,$2,6
+	bne	$3,$0,$L80
+	j	$L78
+$L80:
+	lw	$3,16($fp)
+	lw	$2,20($fp)
+	move	$4,$2
+	sll	$2,$4,3
+	move	$5,$3
+	sll	$4,$5,1
+	addu	$4,$4,$3
+	sll	$3,$4,3
+	lw	$4,148($fp)
+	addu	$3,$3,$4
+	addu	$2,$2,$3
+	l.d	$f0,0($2)
+	mtc1	$0,$f2
+	mtc1	$0,$f3
+	c.lt.d	$f0,$f2
+	bc1t	$L82
+	lw	$3,16($fp)
+	lw	$2,20($fp)
+	move	$4,$2
+	sll	$2,$4,3
+	move	$5,$3
+	sll	$4,$5,1
+	addu	$4,$4,$3
+	sll	$3,$4,3
+	lw	$4,148($fp)
+	addu	$3,$3,$4
+	addu	$2,$2,$3
+	l.d	$f0,0($2)
+	l.d	$f2,$LC65
+	c.lt.d	$f2,$f0
+	bc1t	$L82
+	j	$L81
+$L82:
+	la	$4,$LC66
+	jal	printf
+	move	$2,$0
+	j	$L46
+$L81:
+	lw	$3,16($fp)
+	lw	$2,20($fp)
+	move	$4,$2
+	sll	$2,$4,3
+	move	$5,$3
+	sll	$4,$5,1
+	addu	$4,$4,$3
+	sll	$3,$4,3
+	lw	$4,152($fp)
+	addu	$3,$3,$4
+	addu	$2,$2,$3
+	l.d	$f0,0($2)
+	mtc1	$0,$f2
+	mtc1	$0,$f3
+	c.lt.d	$f0,$f2
+	bc1f	$L83
+	la	$4,$LC67
+	jal	printf
+	move	$2,$0
+	j	$L46
+$L83:
+	lw	$3,16($fp)
+	lw	$2,20($fp)
+	move	$4,$2
+	sll	$2,$4,3
+	move	$5,$3
+	sll	$4,$5,1
+	addu	$4,$4,$3
+	sll	$3,$4,3
+	lw	$4,152($fp)
+	addu	$3,$3,$4
+	addu	$2,$2,$3
+	l.d	$f0,32($fp)
+	l.d	$f2,0($2)
+	c.lt.d	$f0,$f2
+	bc1f	$L84
+	lw	$3,16($fp)
+	lw	$2,20($fp)
+	move	$4,$2
+	sll	$2,$4,3
+	move	$5,$3
+	sll	$4,$5,1
+	addu	$4,$4,$3
+	sll	$3,$4,3
+	lw	$4,152($fp)
+	addu	$3,$3,$4
+	addu	$2,$2,$3
+	l.d	$f0,0($2)
+	s.d	$f0,32($fp)
+$L84:
+$L79:
+	lw	$3,16($fp)
+	addu	$2,$3,1
+	move	$3,$2
+	sw	$3,16($fp)
+	j	$L77
+$L78:
+	l.d	$f0,32($fp)
+	mtc1	$0,$f2
+	mtc1	$0,$f3
+	c.eq.d	$f0,$f2
+	bc1f	$L85
+	la	$4,$LC68
+	jal	printf
+	move	$2,$0
+	j	$L46
+$L85:
+$L73:
+	lw	$3,20($fp)
+	addu	$2,$3,1
+	move	$3,$2
+	sw	$3,20($fp)
+	j	$L71
+$L72:
+	lw	$3,144($fp)
+	addu	$2,$3,48
+	lw	$4,144($fp)
+	addu	$3,$4,24
+	l.d	$f0,0($3)
+	s.d	$f0,0($2)
+	li	$2,0x00000001		# 1
+	j	$L46
+$L46:
+	move	$sp,$fp			# sp not trusted here
+	lw	$31,140($sp)
+	lw	$fp,136($sp)
+	addu	$sp,$sp,144
+	j	$31
+	.end	Reader
+	.loc	1 504
+	.ent	Region
+Region:
+	.frame	$fp,120,$31		# vars= 88, regs= 2/1, args= 16, extra= 0
+	.mask	0xc0000000,-12
+	.fmask	0x00300000,-4
+	subu	$sp,$sp,120
+	sw	$31,108($sp)
+	sw	$fp,104($sp)
+	s.d	$f20,112($sp)
+	move	$fp,$sp
+	sw	$4,120($fp)
+	sw	$5,124($fp)
+	sw	$6,128($fp)
+	sw	$7,132($fp)
+	lw	$2,128($fp)
+	lw	$4,128($fp)
+	addu	$3,$4,8
+	l.d	$f0,0($2)
+	l.d	$f2,0($3)
+	mul.d	$f20,$f0,$f2
+	lw	$3,128($fp)
+	addu	$2,$3,8
+	lw	$4,128($fp)
+	addu	$3,$4,16
+	l.d	$f0,0($2)
+	l.d	$f2,0($3)
+	mul.d	$f0,$f0,$f2
+	add.d	$f20,$f20,$f0
+	lw	$3,128($fp)
+	addu	$2,$3,16
+	lw	$3,128($fp)
+	l.d	$f0,0($2)
+	l.d	$f2,0($3)
+	mul.d	$f0,$f0,$f2
+	add.d	$f20,$f20,$f0
+	add.d	$f0,$f20,$f20
+	s.d	$f0,64($fp)
+	dsz	72($fp)
+	l.s	$f0,120($fp)
+	#nop
+	cvt.d.w	$f0,$f0
+	s.d	$f0,80($fp)
+	lw	$2,124($fp)
+	sw	$0,0($2)
+	sw	$0,24($fp)
+$L87:
+	lw	$2,24($fp)
+	slt	$3,$2,5
+	bne	$3,$0,$L90
+	j	$L88
+$L90:
+	lw	$2,24($fp)
+	move	$3,$2
+	sll	$2,$3,3
+	lw	$3,128($fp)
+	addu	$2,$2,$3
+	lw	$3,24($fp)
+	move	$4,$3
+	sll	$3,$4,3
+	lw	$4,128($fp)
+	addu	$3,$3,$4
+	addu	$4,$3,8
+	l.d	$f0,0($2)
+	l.d	$f2,0($4)
+	mul.d	$f0,$f0,$f2
+	l.d	$f2,72($fp)
+	add.d	$f0,$f2,$f0
+	s.d	$f0,72($fp)
+	lw	$2,24($fp)
+	move	$3,$2
+	sll	$2,$3,3
+	lw	$4,124($fp)
+	addu	$3,$2,$4
+	addu	$2,$3,4
+	l.d	$f0,80($fp)
+	l.d	$f2,72($fp)
+	mul.d	$f0,$f0,$f2
+	l.d	$f2,64($fp)
+	div.d	$f0,$f0,$f2
+	l.d	$f2,$LC69
+	add.d	$f0,$f0,$f2
+	trunc.w.d $f4,$f0,$8
+	mfc1	$3,$f4
+	subu	$4,$3,1
+	sw	$4,0($2)
+	lw	$2,24($fp)
+	move	$3,$2
+	sll	$2,$3,3
+	lw	$4,124($fp)
+	addu	$3,$2,$4
+	addu	$2,$3,8
+	lw	$3,24($fp)
+	move	$4,$3
+	sll	$3,$4,3
+	lw	$4,124($fp)
+	addu	$3,$3,$4
+	addu	$4,$3,4
+	lw	$3,0($4)
+	addu	$4,$3,1
+	sw	$4,0($2)
+$L89:
+	lw	$3,24($fp)
+	addu	$2,$3,1
+	move	$3,$2
+	sw	$3,24($fp)
+	j	$L87
+$L88:
+	lw	$3,124($fp)
+	addu	$2,$3,44
+	lw	$3,120($fp)
+	subu	$4,$3,1
+	sw	$4,0($2)
+	sw	$0,24($fp)
+$L91:
+	lw	$2,24($fp)
+	slt	$3,$2,6
+	bne	$3,$0,$L94
+	j	$L92
+$L94:
+	lw	$2,24($fp)
+	move	$3,$2
+	sll	$2,$3,3
+	lw	$4,124($fp)
+	addu	$3,$2,$4
+	addu	$2,$3,4
+	lw	$3,24($fp)
+	move	$4,$3
+	sll	$3,$4,3
+	lw	$4,124($fp)
+	addu	$3,$3,$4
+	lw	$2,0($2)
+	lw	$3,0($3)
+	subu	$2,$2,$3
+	addu	$3,$2,1
+	sw	$3,48($fp)
+	dsz	80($fp)
+	lw	$2,24($fp)
+	slt	$3,$2,3
+	bne	$3,$0,$L95
+	lw	$2,24($fp)
+	move	$3,$2
+	sll	$2,$3,3
+	lw	$3,128($fp)
+	addu	$2,$2,$3
+	subu	$3,$2,8
+	l.d	$f0,0($3)
+	s.d	$f0,80($fp)
+$L95:
+	l.s	$f0,48($fp)
+	#nop
+	cvt.d.w	$f0,$f0
+	lw	$2,24($fp)
+	move	$3,$2
+	sll	$2,$3,3
+	lw	$3,128($fp)
+	addu	$2,$2,$3
+	l.d	$f2,0($2)
+	mul.d	$f0,$f0,$f2
+	lw	$2,24($fp)
+	move	$3,$2
+	sll	$2,$3,3
+	lw	$3,128($fp)
+	addu	$2,$2,$3
+	addu	$3,$2,8
+	l.d	$f2,0($3)
+	div.d	$f0,$f0,$f2
+	mov.d	$f12,$f0
+	jal	sqrt
+	l.d	$f2,$LC69
+	add.d	$f0,$f0,$f2
+	trunc.w.d $f4,$f0,$8
+	s.s	$f4,44($fp)
+	lw	$2,44($fp)
+	lw	$3,48($fp)
+	slt	$2,$3,$2
+	beq	$2,$0,$L96
+	lw	$2,48($fp)
+	sw	$2,44($fp)
+$L96:
+	lw	$2,44($fp)
+	bne	$2,$0,$L97
+	li	$2,0x00000001		# 1
+	sw	$2,44($fp)
+$L97:
+	lw	$2,24($fp)
+	move	$3,$2
+	sll	$2,$3,3
+	lw	$3,128($fp)
+	addu	$2,$2,$3
+	l.s	$f0,44($fp)
+	#nop
+	cvt.d.w	$f0,$f0
+	l.d	$f2,0($2)
+	div.d	$f0,$f2,$f0
+	s.d	$f0,96($fp)
+	lw	$2,44($fp)
+	subu	$3,$2,1
+	sw	$3,28($fp)
+	dsz	64($fp)
+	sw	$0,16($fp)
+$L98:
+	lw	$2,16($fp)
+	lw	$3,44($fp)
+	slt	$2,$2,$3
+	bne	$2,$0,$L101
+	j	$L99
+$L101:
+	lw	$2,28($fp)
+	lw	$3,44($fp)
+	div	$2,$2,$3
+	sw	$2,32($fp)
+	lw	$2,28($fp)
+	lw	$3,48($fp)
+	addu	$2,$2,$3
+	lw	$3,44($fp)
+	div	$2,$2,$3
+	lw	$3,32($fp)
+	subu	$2,$2,$3
+	sw	$2,52($fp)
+	lw	$2,52($fp)
+	bne	$2,$0,$L102
+	la	$4,$LC70
+	jal	printf
+	move	$2,$0
+	j	$L86
+$L102:
+	lw	$2,24($fp)
+	move	$3,$2
+	sll	$2,$3,3
+	lw	$3,128($fp)
+	addu	$2,$2,$3
+	addu	$3,$2,8
+	l.s	$f0,52($fp)
+	#nop
+	cvt.d.w	$f0,$f0
+	l.d	$f2,0($3)
+	div.d	$f0,$f2,$f0
+	s.d	$f0,56($fp)
+	dsz	72($fp)
+	l.d	$f0,96($fp)
+	l.d	$f2,56($fp)
+	mul.d	$f0,$f0,$f2
+	l.d	$f2,$LC71
+	mul.d	$f0,$f0,$f2
+	s.d	$f0,88($fp)
+	lw	$2,24($fp)
+	move	$3,$2
+	sll	$2,$3,3
+	lw	$3,124($fp)
+	addu	$2,$2,$3
+	lw	$3,0($2)
+	lw	$2,32($fp)
+	addu	$3,$3,$2
+	sw	$3,40($fp)
+	lw	$2,40($fp)
+	lw	$3,52($fp)
+	addu	$2,$2,$3
+	sw	$2,36($fp)
+	lw	$2,40($fp)
+	sw	$2,20($fp)
+$L103:
+	lw	$2,20($fp)
+	lw	$3,36($fp)
+	slt	$2,$2,$3
+	bne	$2,$0,$L106
+	j	$L104
+$L106:
+	lw	$2,20($fp)
+	move	$3,$2
+	sll	$2,$3,3
+	lw	$3,136($fp)
+	addu	$2,$2,$3
+	l.d	$f0,96($fp)
+	s.d	$f0,0($2)
+	lw	$2,20($fp)
+	move	$3,$2
+	sll	$2,$3,3
+	lw	$3,136($fp)
+	addu	$2,$2,$3
+	addu	$3,$2,16384
+	l.d	$f0,56($fp)
+	s.d	$f0,0($3)
+	lw	$2,20($fp)
+	move	$3,$2
+	sll	$2,$3,3
+	lw	$3,132($fp)
+	addu	$2,$2,$3
+	l.d	$f0,64($fp)
+	s.d	$f0,0($2)
+	lw	$2,20($fp)
+	move	$3,$2
+	sll	$2,$3,3
+	lw	$3,132($fp)
+	addu	$2,$2,$3
+	addu	$3,$2,16384
+	l.d	$f0,72($fp)
+	s.d	$f0,0($3)
+	lw	$2,20($fp)
+	move	$3,$2
+	sll	$2,$3,3
+	lw	$3,132($fp)
+	addu	$2,$2,$3
+	li	$3,0x00008000		# 32768
+	addu	$2,$2,$3
+	l.d	$f0,80($fp)
+	s.d	$f0,0($2)
+	lw	$2,20($fp)
+	move	$3,$2
+	sll	$2,$3,3
+	lw	$3,140($fp)
+	addu	$2,$2,$3
+	l.d	$f0,88($fp)
+	s.d	$f0,0($2)
+	l.d	$f0,72($fp)
+	l.d	$f2,56($fp)
+	add.d	$f0,$f0,$f2
+	s.d	$f0,72($fp)
+$L105:
+	lw	$3,20($fp)
+	addu	$2,$3,1
+	move	$3,$2
+	sw	$3,20($fp)
+	j	$L103
+$L104:
+	l.d	$f0,64($fp)
+	l.d	$f2,96($fp)
+	add.d	$f0,$f0,$f2
+	s.d	$f0,64($fp)
+	lw	$2,28($fp)
+	lw	$3,48($fp)
+	addu	$2,$2,$3
+	sw	$2,28($fp)
+$L100:
+	lw	$3,16($fp)
+	addu	$2,$3,1
+	move	$3,$2
+	sw	$3,16($fp)
+	j	$L98
+$L99:
+$L93:
+	lw	$3,24($fp)
+	addu	$2,$3,1
+	move	$3,$2
+	sw	$3,24($fp)
+	j	$L91
+$L92:
+	li	$2,0x00000001		# 1
+	j	$L86
+$L86:
+	move	$sp,$fp			# sp not trusted here
+	lw	$31,108($sp)
+	lw	$fp,104($sp)
+	l.d	$f20,112($sp)
+	addu	$sp,$sp,120
+	j	$31
+	.end	Region
+	.loc	1 590
+	.ent	SetUp1
+SetUp1:
+	.frame	$fp,256,$31		# vars= 232, regs= 2/0, args= 16, extra= 0
+	.mask	0xc0000000,-4
+	.fmask	0x00000000,0
+	subu	$sp,$sp,256
+	sw	$31,252($sp)
+	sw	$fp,248($sp)
+	move	$fp,$sp
+	sw	$4,256($fp)
+	sw	$5,260($fp)
+	sw	$6,264($fp)
+	sw	$7,268($fp)
+	sw	$0,36($fp)
+$L108:
+	lw	$2,36($fp)
+	slt	$3,$2,3
+	bne	$3,$0,$L111
+	j	$L109
+$L111:
+	lw	$2,36($fp)
+	addu	$3,$2,3
+	sw	$3,44($fp)
+	lw	$2,44($fp)
+	move	$3,$2
+	sll	$2,$3,3
+	lw	$3,260($fp)
+	addu	$2,$2,$3
+	lw	$3,0($2)
+	move	$2,$3
+	sll	$3,$2,3
+	lw	$4,268($fp)
+	addu	$2,$3,$4
+	li	$3,0x00008000		# 32768
+	addu	$2,$2,$3
+	lw	$3,44($fp)
+	move	$4,$3
+	sll	$3,$4,3
+	lw	$4,260($fp)
+	addu	$3,$3,$4
+	lw	$4,0($3)
+	move	$3,$4
+	sll	$4,$3,3
+	lw	$5,268($fp)
+	addu	$3,$4,$5
+	li	$4,0x00008000		# 32768
+	addu	$3,$3,$4
+	l.d	$f0,0($2)
+	l.d	$f2,0($3)
+	mul.d	$f0,$f0,$f2
+	s.d	$f0,184($fp)
+	l.d	$f0,184($fp)
+	l.d	$f2,184($fp)
+	add.d	$f0,$f0,$f2
+	s.d	$f0,224($fp)
+	lw	$2,36($fp)
+	move	$3,$2
+	sll	$2,$3,3
+	lw	$3,260($fp)
+	addu	$2,$2,$3
+	lw	$3,0($2)
+	sw	$3,40($fp)
+$L112:
+	lw	$2,36($fp)
+	move	$3,$2
+	sll	$2,$3,3
+	lw	$3,260($fp)
+	addu	$2,$2,$3
+	addu	$3,$2,4
+	lw	$2,40($fp)
+	lw	$3,0($3)
+	slt	$2,$3,$2
+	bne	$2,$0,$L113
+$L115:
+	lw	$2,44($fp)
+	move	$3,$2
+	sll	$2,$3,3
+	lw	$3,260($fp)
+	addu	$2,$2,$3
+	lw	$3,0($2)
+	sw	$3,48($fp)
+$L116:
+	lw	$2,44($fp)
+	move	$3,$2
+	sll	$2,$3,3
+	lw	$3,260($fp)
+	addu	$2,$2,$3
+	addu	$3,$2,4
+	lw	$2,48($fp)
+	lw	$3,0($3)
+	slt	$2,$3,$2
+	bne	$2,$0,$L117
+$L119:
+	sw	$0,20($fp)
+$L120:
+	lw	$2,20($fp)
+	slt	$3,$2,2
+	bne	$3,$0,$L123
+	j	$L121
+$L123:
+	lw	$2,20($fp)
+	move	$3,$2
+	sll	$2,$3,3
+	addu	$4,$fp,16
+	addu	$3,$2,$4
+	addu	$2,$3,40
+	lw	$4,20($fp)
+	lw	$3,48($fp)
+	move	$5,$3
+	sll	$3,$5,3
+	sll	$5,$4,14
+	lw	$6,268($fp)
+	addu	$4,$5,$6
+	addu	$3,$3,$4
+	lw	$5,20($fp)
+	lw	$4,40($fp)
+	move	$6,$4
+	sll	$4,$6,3
+	sll	$6,$5,14
+	lw	$7,268($fp)
+	addu	$5,$6,$7
+	addu	$4,$4,$5
+	l.d	$f0,0($3)
+	l.d	$f2,0($4)
+	sub.d	$f0,$f0,$f2
+	s.d	$f0,0($2)
+	lw	$2,20($fp)
+	move	$3,$2
+	sll	$2,$3,3
+	addu	$4,$fp,16
+	addu	$3,$2,$4
+	addu	$2,$3,72
+	lw	$3,20($fp)
+	move	$4,$3
+	sll	$3,$4,3
+	addu	$5,$fp,16
+	addu	$4,$3,$5
+	addu	$3,$4,40
+	lw	$5,20($fp)
+	lw	$4,48($fp)
+	move	$6,$4
+	sll	$4,$6,3
+	sll	$6,$5,14
+	lw	$7,272($fp)
+	addu	$5,$6,$7
+	addu	$4,$4,$5
+	l.d	$f0,0($3)
+	l.d	$f2,0($4)
+	add.d	$f0,$f0,$f2
+	s.d	$f0,0($2)
+	lw	$2,20($fp)
+	move	$3,$2
+	sll	$2,$3,3
+	addu	$4,$fp,16
+	addu	$3,$2,$4
+	addu	$2,$3,56
+	lw	$3,20($fp)
+	move	$4,$3
+	sll	$3,$4,3
+	addu	$5,$fp,16
+	addu	$4,$3,$5
+	addu	$3,$4,40
+	lw	$5,20($fp)
+	lw	$4,40($fp)
+	move	$6,$4
+	sll	$4,$6,3
+	sll	$6,$5,14
+	lw	$7,272($fp)
+	addu	$5,$6,$7
+	addu	$4,$4,$5
+	l.d	$f0,0($3)
+	l.d	$f2,0($4)
+	sub.d	$f0,$f0,$f2
+	s.d	$f0,0($2)
+	lw	$2,20($fp)
+	move	$3,$2
+	sll	$2,$3,3
+	addu	$4,$fp,16
+	addu	$3,$2,$4
+	addu	$2,$3,88
+	lw	$3,20($fp)
+	move	$4,$3
+	sll	$3,$4,3
+	addu	$5,$fp,16
+	addu	$4,$3,$5
+	addu	$3,$4,72
+	lw	$5,20($fp)
+	lw	$4,40($fp)
+	move	$6,$4
+	sll	$4,$6,3
+	sll	$6,$5,14
+	lw	$7,272($fp)
+	addu	$5,$6,$7
+	addu	$4,$4,$5
+	l.d	$f0,0($3)
+	l.d	$f2,0($4)
+	sub.d	$f0,$f0,$f2
+	s.d	$f0,0($2)
+	lw	$2,20($fp)
+	move	$3,$2
+	sll	$2,$3,3
+	addu	$4,$fp,16
+	addu	$3,$2,$4
+	addu	$2,$3,104
+	lw	$3,20($fp)
+	move	$4,$3
+	sll	$3,$4,3
+	addu	$5,$fp,16
+	addu	$4,$3,$5
+	addu	$3,$4,40
+	lw	$4,20($fp)
+	move	$5,$4
+	sll	$4,$5,3
+	addu	$5,$fp,16
+	addu	$4,$4,$5
+	addu	$5,$4,40
+	l.d	$f0,0($3)
+	l.d	$f2,0($5)
+	mul.d	$f0,$f0,$f2
+	s.d	$f0,0($2)
+	lw	$2,20($fp)
+	move	$3,$2
+	sll	$2,$3,3
+	addu	$4,$fp,16
+	addu	$3,$2,$4
+	addu	$2,$3,136
+	lw	$3,20($fp)
+	move	$4,$3
+	sll	$3,$4,3
+	addu	$5,$fp,16
+	addu	$4,$3,$5
+	addu	$3,$4,72
+	lw	$4,20($fp)
+	move	$5,$4
+	sll	$4,$5,3
+	addu	$5,$fp,16
+	addu	$4,$4,$5
+	addu	$5,$4,72
+	l.d	$f0,0($3)
+	l.d	$f2,0($5)
+	mul.d	$f0,$f0,$f2
+	s.d	$f0,0($2)
+	lw	$2,20($fp)
+	move	$3,$2
+	sll	$2,$3,3
+	addu	$4,$fp,16
+	addu	$3,$2,$4
+	addu	$2,$3,120
+	lw	$3,20($fp)
+	move	$4,$3
+	sll	$3,$4,3
+	addu	$5,$fp,16
+	addu	$4,$3,$5
+	addu	$3,$4,56
+	lw	$4,20($fp)
+	move	$5,$4
+	sll	$4,$5,3
+	addu	$5,$fp,16
+	addu	$4,$4,$5
+	addu	$5,$4,56
+	l.d	$f0,0($3)
+	l.d	$f2,0($5)
+	mul.d	$f0,$f0,$f2
+	s.d	$f0,0($2)
+	lw	$2,20($fp)
+	move	$3,$2
+	sll	$2,$3,3
+	addu	$4,$fp,16
+	addu	$3,$2,$4
+	addu	$2,$3,152
+	lw	$3,20($fp)
+	move	$4,$3
+	sll	$3,$4,3
+	addu	$5,$fp,16
+	addu	$4,$3,$5
+	addu	$3,$4,88
+	lw	$4,20($fp)
+	move	$5,$4
+	sll	$4,$5,3
+	addu	$5,$fp,16
+	addu	$4,$4,$5
+	addu	$5,$4,88
+	l.d	$f0,0($3)
+	l.d	$f2,0($5)
+	mul.d	$f0,$f0,$f2
+	s.d	$f0,0($2)
+$L122:
+	lw	$3,20($fp)
+	addu	$2,$3,1
+	move	$3,$2
+	sw	$3,20($fp)
+	j	$L120
+$L121:
+	dsz	192($fp)
+	sw	$0,28($fp)
+$L124:
+	lw	$2,28($fp)
+	slt	$3,$2,2
+	bne	$3,$0,$L127
+	j	$L125
+$L127:
+	sw	$0,16($fp)
+$L128:
+	lw	$2,16($fp)
+	slt	$3,$2,2
+	bne	$3,$0,$L131
+	j	$L129
+$L131:
+	lw	$2,16($fp)
+	move	$3,$2
+	sll	$2,$3,4
+	lw	$3,28($fp)
+	move	$4,$3
+	sll	$3,$4,5
+	addu	$2,$2,$3
+	addu	$3,$fp,120
+	addu	$2,$3,$2
+	l.d	$f0,8($2)
+	l.d	$f2,184($fp)
+	add.d	$f0,$f0,$f2
+	s.d	$f0,200($fp)
+	l.d	$f12,200($fp)
+	jal	sqrt
+	s.d	$f0,208($fp)
+	l.d	$f0,$LC72
+	l.d	$f2,208($fp)
+	div.d	$f0,$f0,$f2
+	s.d	$f0,216($fp)
+	dsz	240($fp)
+	sw	$0,24($fp)
+$L132:
+	lw	$2,24($fp)
+	slt	$3,$2,2
+	bne	$3,$0,$L135
+	j	$L133
+$L135:
+	sw	$0,32($fp)
+$L136:
+	lw	$2,32($fp)
+	slt	$3,$2,2
+	bne	$3,$0,$L139
+	j	$L137
+$L139:
+	lw	$2,32($fp)
+	move	$3,$2
+	sll	$2,$3,4
+	lw	$3,24($fp)
+	move	$4,$3
+	sll	$3,$4,5
+	addu	$2,$2,$3
+	addu	$3,$fp,56
+	addu	$2,$3,$2
+	l.d	$f0,0($2)
+	s.d	$f0,232($fp)
+	l.d	$f0,232($fp)
+	l.d	$f2,216($fp)
+	mul.d	$f0,$f0,$f2
+	mov.d	$f12,$f0
+	jal	atan
+	l.d	$f4,232($fp)
+	neg.d	$f2,$f4
+	mul.d	$f0,$f2,$f0
+	l.d	$f2,240($fp)
+	sub.d	$f0,$f0,$f2
+	s.d	$f0,240($fp)
+$L138:
+	lw	$3,32($fp)
+	addu	$2,$3,1
+	move	$3,$2
+	sw	$3,32($fp)
+	j	$L136
+$L137:
+	l.d	$f0,240($fp)
+	neg.d	$f2,$f0
+	s.d	$f2,240($fp)
+$L134:
+	lw	$3,24($fp)
+	addu	$2,$3,1
+	move	$3,$2
+	sw	$3,24($fp)
+	j	$L132
+$L133:
+	l.d	$f0,152($fp)
+	l.d	$f2,200($fp)
+	add.d	$f0,$f0,$f2
+	l.d	$f2,136($fp)
+	l.d	$f4,200($fp)
+	add.d	$f2,$f2,$f4
+	mul.d	$f0,$f0,$f2
+	l.d	$f2,120($fp)
+	l.d	$f4,200($fp)
+	add.d	$f2,$f2,$f4
+	l.d	$f4,168($fp)
+	l.d	$f6,200($fp)
+	add.d	$f4,$f4,$f6
+	mul.d	$f2,$f2,$f4
+	div.d	$f0,$f0,$f2
+	mov.d	$f12,$f0
+	jal	log
+	l.d	$f2,208($fp)
+	l.d	$f4,$LC73
+	mul.d	$f2,$f2,$f4
+	l.d	$f4,240($fp)
+	mul.d	$f2,$f2,$f4
+	l.d	$f4,192($fp)
+	sub.d	$f2,$f2,$f4
+	l.d	$f4,224($fp)
+	mul.d	$f0,$f0,$f4
+	sub.d	$f2,$f2,$f0
+	s.d	$f2,192($fp)
+$L130:
+	lw	$3,16($fp)
+	addu	$2,$3,1
+	move	$3,$2
+	sw	$3,16($fp)
+	j	$L128
+$L129:
+	l.d	$f0,192($fp)
+	neg.d	$f2,$f0
+	s.d	$f2,192($fp)
+$L126:
+	lw	$3,28($fp)
+	addu	$2,$3,1
+	move	$3,$2
+	sw	$3,28($fp)
+	j	$L124
+$L125:
+	.set	noreorder
+	nop
+	.set	reorder
+	sw	$0,28($fp)
+$L140:
+	lw	$2,28($fp)
+	slt	$3,$2,2
+	bne	$3,$0,$L143
+	j	$L141
+$L143:
+	sw	$0,16($fp)
+$L144:
+	lw	$2,16($fp)
+	slt	$3,$2,2
+	bne	$3,$0,$L147
+	j	$L145
+$L147:
+	lw	$2,16($fp)
+	move	$3,$2
+	sll	$2,$3,4
+	lw	$3,28($fp)
+	move	$4,$3
+	sll	$3,$4,5
+	addu	$2,$2,$3
+	addu	$3,$fp,120
+	addu	$2,$3,$2
+	l.d	$f0,0($2)
+	l.d	$f2,184($fp)
+	add.d	$f0,$f0,$f2
+	mov.d	$f12,$f0
+	jal	sqrt
+	s.d	$f0,208($fp)
+	l.d	$f0,$LC72
+	l.d	$f2,208($fp)
+	div.d	$f0,$f0,$f2
+	s.d	$f0,216($fp)
+	dsz	240($fp)
+	sw	$0,24($fp)
+$L148:
+	lw	$2,24($fp)
+	slt	$3,$2,2
+	bne	$3,$0,$L151
+	j	$L149
+$L151:
+	sw	$0,32($fp)
+$L152:
+	lw	$2,32($fp)
+	slt	$3,$2,2
+	bne	$3,$0,$L155
+	j	$L153
+$L155:
+	lw	$2,32($fp)
+	move	$3,$2
+	sll	$2,$3,4
+	lw	$3,24($fp)
+	move	$4,$3
+	sll	$3,$4,5
+	addu	$2,$2,$3
+	addu	$3,$fp,56
+	addu	$2,$3,$2
+	l.d	$f0,8($2)
+	s.d	$f0,232($fp)
+	l.d	$f0,232($fp)
+	l.d	$f2,216($fp)
+	mul.d	$f0,$f0,$f2
+	mov.d	$f12,$f0
+	jal	atan
+	l.d	$f4,232($fp)
+	neg.d	$f2,$f4
+	mul.d	$f0,$f2,$f0
+	l.d	$f2,240($fp)
+	sub.d	$f0,$f0,$f2
+	s.d	$f0,240($fp)
+$L154:
+	lw	$3,32($fp)
+	addu	$2,$3,1
+	move	$3,$2
+	sw	$3,32($fp)
+	j	$L152
+$L153:
+	l.d	$f0,240($fp)
+	neg.d	$f2,$f0
+	s.d	$f2,240($fp)
+$L150:
+	lw	$3,24($fp)
+	addu	$2,$3,1
+	move	$3,$2
+	sw	$3,24($fp)
+	j	$L148
+$L149:
+	l.d	$f0,208($fp)
+	l.d	$f2,$LC73
+	mul.d	$f0,$f0,$f2
+	l.d	$f2,240($fp)
+	mul.d	$f0,$f0,$f2
+	l.d	$f2,192($fp)
+	sub.d	$f0,$f0,$f2
+	s.d	$f0,192($fp)
+$L146:
+	lw	$3,16($fp)
+	addu	$2,$3,1
+	move	$3,$2
+	sw	$3,16($fp)
+	j	$L144
+$L145:
+	l.d	$f0,192($fp)
+	neg.d	$f2,$f0
+	s.d	$f2,192($fp)
+$L142:
+	lw	$3,28($fp)
+	addu	$2,$3,1
+	move	$3,$2
+	sw	$3,28($fp)
+	j	$L140
+$L141:
+	lw	$3,40($fp)
+	lw	$2,48($fp)
+	move	$4,$2
+	sll	$2,$4,3
+	sll	$4,$3,14
+	lw	$5,264($fp)
+	addu	$3,$4,$5
+	addu	$2,$2,$3
+	l.d	$f0,192($fp)
+	s.d	$f0,0($2)
+	lw	$3,48($fp)
+	lw	$2,40($fp)
+	move	$4,$2
+	sll	$2,$4,3
+	sll	$4,$3,14
+	lw	$5,264($fp)
+	addu	$3,$4,$5
+	addu	$2,$2,$3
+	l.d	$f0,192($fp)
+	s.d	$f0,0($2)
+$L118:
+	lw	$3,48($fp)
+	addu	$2,$3,1
+	move	$3,$2
+	sw	$3,48($fp)
+	j	$L116
+$L117:
+$L114:
+	lw	$3,40($fp)
+	addu	$2,$3,1
+	move	$3,$2
+	sw	$3,40($fp)
+	j	$L112
+$L113:
+$L110:
+	lw	$3,36($fp)
+	addu	$2,$3,1
+	move	$3,$2
+	sw	$3,36($fp)
+	j	$L108
+$L109:
+$L107:
+	move	$sp,$fp			# sp not trusted here
+	lw	$31,252($sp)
+	lw	$fp,248($sp)
+	addu	$sp,$sp,256
+	j	$31
+	.end	SetUp1
+	.loc	1 673
+	.ent	SetUp2
+SetUp2:
+	.frame	$fp,512,$31		# vars= 440, regs= 2/6, args= 16, extra= 0
+	.mask	0xc0000000,-52
+	.fmask	0xfff00000,-4
+	subu	$sp,$sp,512
+	sw	$31,460($sp)
+	sw	$fp,456($sp)
+	s.d	$f30,504($sp)
+	s.d	$f28,496($sp)
+	s.d	$f26,488($sp)
+	s.d	$f24,480($sp)
+	s.d	$f22,472($sp)
+	s.d	$f20,464($sp)
+	move	$fp,$sp
+	sw	$4,512($fp)
+	sw	$5,516($fp)
+	sw	$6,520($fp)
+	sw	$7,524($fp)
+	sw	$0,20($fp)
+$L157:
+	lw	$2,20($fp)
+	slt	$3,$2,6
+	bne	$3,$0,$L160
+	j	$L158
+$L160:
+	sw	$0,16($fp)
+$L161:
+	lw	$2,16($fp)
+	slt	$3,$2,2
+	bne	$3,$0,$L164
+	j	$L162
+$L164:
+	lw	$2,20($fp)
+	lw	$4,16($fp)
+	addu	$3,$2,$4
+	addu	$2,$3,1
+	li	$6,0x2aaaaaab		# 715827883
+	mult	$2,$6
+	mfhi	$5
+	mflo	$4
+	srl	$6,$5,0
+	move	$7,$0
+	sra	$4,$2,31
+	subu	$3,$6,$4
+	move	$5,$3
+	sll	$4,$5,1
+	addu	$4,$4,$3
+	sll	$3,$4,1
+	subu	$2,$2,$3
+	sw	$2,28($fp)
+	lw	$2,20($fp)
+	move	$3,$2
+	sll	$2,$3,3
+	lw	$3,516($fp)
+	addu	$2,$2,$3
+	lw	$3,0($2)
+	sw	$3,24($fp)
+$L165:
+	lw	$2,20($fp)
+	move	$3,$2
+	sll	$2,$3,3
+	lw	$3,516($fp)
+	addu	$2,$2,$3
+	addu	$3,$2,4
+	lw	$2,24($fp)
+	lw	$3,0($3)
+	slt	$2,$3,$2
+	bne	$2,$0,$L166
+$L168:
+	lw	$3,16($fp)
+	lw	$2,24($fp)
+	move	$4,$2
+	sll	$2,$4,3
+	sll	$4,$3,14
+	lw	$5,524($fp)
+	addu	$3,$4,$5
+	addu	$2,$2,$3
+	lw	$3,28($fp)
+	move	$4,$3
+	sll	$3,$4,3
+	lw	$4,516($fp)
+	addu	$3,$3,$4
+	lw	$4,0($3)
+	move	$3,$4
+	sll	$4,$3,3
+	lw	$5,524($fp)
+	addu	$3,$4,$5
+	li	$4,0x00008000		# 32768
+	addu	$3,$3,$4
+	l.d	$f0,0($2)
+	l.d	$f2,0($3)
+	sub.d	$f0,$f0,$f2
+	s.d	$f0,216($fp)
+	lw	$3,16($fp)
+	lw	$2,24($fp)
+	move	$4,$2
+	sll	$2,$4,3
+	sll	$4,$3,14
+	lw	$5,528($fp)
+	addu	$3,$4,$5
+	addu	$2,$2,$3
+	l.d	$f0,216($fp)
+	l.d	$f2,0($2)
+	add.d	$f0,$f0,$f2
+	s.d	$f0,224($fp)
+	l.d	$f0,216($fp)
+	l.d	$f2,216($fp)
+	mul.d	$f0,$f0,$f2
+	s.d	$f0,216($fp)
+	l.d	$f0,224($fp)
+	l.d	$f2,224($fp)
+	mul.d	$f0,$f0,$f2
+	s.d	$f0,224($fp)
+	lw	$2,28($fp)
+	move	$3,$2
+	sll	$2,$3,3
+	lw	$3,516($fp)
+	addu	$2,$2,$3
+	lw	$3,0($2)
+	sw	$3,32($fp)
+$L169:
+	lw	$2,28($fp)
+	move	$3,$2
+	sll	$2,$3,3
+	lw	$3,516($fp)
+	addu	$2,$2,$3
+	addu	$3,$2,4
+	lw	$2,32($fp)
+	lw	$3,0($3)
+	slt	$2,$3,$2
+	bne	$2,$0,$L170
+$L172:
+	lw	$3,16($fp)
+	lw	$2,32($fp)
+	move	$4,$2
+	sll	$2,$4,3
+	sll	$4,$3,14
+	lw	$5,524($fp)
+	addu	$3,$4,$5
+	addu	$2,$2,$3
+	li	$3,0x00000001		# 1
+	lw	$5,16($fp)
+	subu	$4,$3,$5
+	lw	$3,24($fp)
+	move	$5,$3
+	sll	$3,$5,3
+	sll	$5,$4,14
+	lw	$6,524($fp)
+	addu	$4,$5,$6
+	addu	$3,$3,$4
+	l.d	$f0,0($2)
+	l.d	$f2,0($3)
+	sub.d	$f0,$f0,$f2
+	s.d	$f0,56($fp)
+	lw	$3,16($fp)
+	lw	$2,32($fp)
+	move	$4,$2
+	sll	$2,$4,3
+	sll	$4,$3,14
+	lw	$5,528($fp)
+	addu	$3,$4,$5
+	addu	$2,$2,$3
+	l.d	$f0,56($fp)
+	l.d	$f2,0($2)
+	add.d	$f0,$f0,$f2
+	s.d	$f0,64($fp)
+	li	$2,0x00000001		# 1
+	lw	$4,16($fp)
+	subu	$3,$2,$4
+	lw	$2,24($fp)
+	move	$4,$2
+	sll	$2,$4,3
+	sll	$4,$3,14
+	lw	$5,528($fp)
+	addu	$3,$4,$5
+	addu	$2,$2,$3
+	l.d	$f0,56($fp)
+	l.d	$f2,0($2)
+	sub.d	$f0,$f0,$f2
+	s.d	$f0,72($fp)
+	li	$2,0x00000001		# 1
+	lw	$4,16($fp)
+	subu	$3,$2,$4
+	lw	$2,24($fp)
+	move	$4,$2
+	sll	$2,$4,3
+	sll	$4,$3,14
+	lw	$5,528($fp)
+	addu	$3,$4,$5
+	addu	$2,$2,$3
+	l.d	$f0,64($fp)
+	l.d	$f2,0($2)
+	sub.d	$f0,$f0,$f2
+	s.d	$f0,80($fp)
+	l.d	$f0,56($fp)
+	l.d	$f2,56($fp)
+	mul.d	$f0,$f0,$f2
+	s.d	$f0,88($fp)
+	l.d	$f0,64($fp)
+	l.d	$f2,64($fp)
+	mul.d	$f0,$f0,$f2
+	s.d	$f0,96($fp)
+	l.d	$f0,72($fp)
+	l.d	$f2,72($fp)
+	mul.d	$f0,$f0,$f2
+	s.d	$f0,104($fp)
+	l.d	$f0,80($fp)
+	l.d	$f2,80($fp)
+	mul.d	$f0,$f0,$f2
+	s.d	$f0,112($fp)
+	li	$2,0x00000001		# 1
+	lw	$4,16($fp)
+	subu	$3,$2,$4
+	lw	$2,32($fp)
+	move	$4,$2
+	sll	$2,$4,3
+	sll	$4,$3,14
+	lw	$5,524($fp)
+	addu	$3,$4,$5
+	addu	$2,$2,$3
+	lw	$3,24($fp)
+	move	$4,$3
+	sll	$3,$4,3
+	lw	$4,524($fp)
+	addu	$3,$3,$4
+	li	$4,0x00008000		# 32768
+	addu	$3,$3,$4
+	l.d	$f0,0($2)
+	l.d	$f2,0($3)
+	sub.d	$f0,$f0,$f2
+	s.d	$f0,232($fp)
+	li	$2,0x00000001		# 1
+	lw	$4,16($fp)
+	subu	$3,$2,$4
+	lw	$2,32($fp)
+	move	$4,$2
+	sll	$2,$4,3
+	sll	$4,$3,14
+	lw	$5,528($fp)
+	addu	$3,$4,$5
+	addu	$2,$2,$3
+	l.d	$f0,232($fp)
+	l.d	$f2,0($2)
+	add.d	$f0,$f0,$f2
+	s.d	$f0,240($fp)
+	l.d	$f0,232($fp)
+	l.d	$f2,232($fp)
+	mul.d	$f0,$f0,$f2
+	l.d	$f2,$LC74
+	add.d	$f0,$f0,$f2
+	s.d	$f0,232($fp)
+	l.d	$f0,240($fp)
+	l.d	$f2,240($fp)
+	mul.d	$f0,$f0,$f2
+	l.d	$f2,$LC74
+	add.d	$f0,$f0,$f2
+	s.d	$f0,240($fp)
+	l.d	$f0,216($fp)
+	l.d	$f2,232($fp)
+	add.d	$f0,$f0,$f2
+	s.d	$f0,152($fp)
+	l.d	$f0,216($fp)
+	l.d	$f2,240($fp)
+	add.d	$f0,$f0,$f2
+	s.d	$f0,160($fp)
+	l.d	$f0,224($fp)
+	l.d	$f2,232($fp)
+	add.d	$f0,$f0,$f2
+	s.d	$f0,168($fp)
+	l.d	$f0,224($fp)
+	l.d	$f2,240($fp)
+	add.d	$f0,$f0,$f2
+	s.d	$f0,176($fp)
+	l.d	$f12,152($fp)
+	jal	sqrt
+	s.d	$f0,120($fp)
+	l.d	$f12,160($fp)
+	jal	sqrt
+	s.d	$f0,128($fp)
+	l.d	$f12,168($fp)
+	jal	sqrt
+	s.d	$f0,136($fp)
+	l.d	$f12,176($fp)
+	jal	sqrt
+	s.d	$f0,144($fp)
+	l.d	$f0,$LC75
+	l.d	$f2,120($fp)
+	div.d	$f0,$f0,$f2
+	s.d	$f0,184($fp)
+	l.d	$f0,$LC75
+	l.d	$f2,128($fp)
+	div.d	$f0,$f0,$f2
+	s.d	$f0,192($fp)
+	l.d	$f0,$LC75
+	l.d	$f2,136($fp)
+	div.d	$f0,$f0,$f2
+	s.d	$f0,200($fp)
+	l.d	$f0,$LC75
+	l.d	$f2,144($fp)
+	div.d	$f0,$f0,$f2
+	s.d	$f0,208($fp)
+	l.d	$f0,56($fp)
+	l.d	$f2,184($fp)
+	mul.d	$f0,$f0,$f2
+	mov.d	$f12,$f0
+	jal	atan
+	mov.d	$f20,$f0
+	l.d	$f0,64($fp)
+	l.d	$f2,184($fp)
+	mul.d	$f0,$f0,$f2
+	mov.d	$f12,$f0
+	jal	atan
+	mov.d	$f22,$f0
+	l.d	$f0,72($fp)
+	l.d	$f2,184($fp)
+	mul.d	$f0,$f0,$f2
+	mov.d	$f12,$f0
+	jal	atan
+	mov.d	$f24,$f0
+	l.d	$f0,80($fp)
+	l.d	$f2,184($fp)
+	mul.d	$f0,$f0,$f2
+	mov.d	$f12,$f0
+	jal	atan
+	mov.d	$f26,$f0
+	l.d	$f0,56($fp)
+	l.d	$f2,192($fp)
+	mul.d	$f0,$f0,$f2
+	mov.d	$f12,$f0
+	jal	atan
+	mov.d	$f28,$f0
+	l.d	$f0,64($fp)
+	l.d	$f2,192($fp)
+	mul.d	$f0,$f0,$f2
+	mov.d	$f12,$f0
+	jal	atan
+	mov.d	$f30,$f0
+	l.d	$f0,72($fp)
+	l.d	$f2,192($fp)
+	mul.d	$f0,$f0,$f2
+	mov.d	$f12,$f0
+	jal	atan
+	s.d	$f0,376($fp)
+	l.d	$f0,80($fp)
+	l.d	$f2,192($fp)
+	mul.d	$f0,$f0,$f2
+	mov.d	$f12,$f0
+	jal	atan
+	s.d	$f0,384($fp)
+	l.d	$f0,56($fp)
+	l.d	$f2,200($fp)
+	mul.d	$f0,$f0,$f2
+	mov.d	$f12,$f0
+	jal	atan
+	s.d	$f0,392($fp)
+	l.d	$f0,64($fp)
+	l.d	$f2,200($fp)
+	mul.d	$f0,$f0,$f2
+	mov.d	$f12,$f0
+	jal	atan
+	s.d	$f0,400($fp)
+	l.d	$f0,72($fp)
+	l.d	$f2,200($fp)
+	mul.d	$f0,$f0,$f2
+	mov.d	$f12,$f0
+	jal	atan
+	s.d	$f0,408($fp)
+	l.d	$f0,80($fp)
+	l.d	$f2,200($fp)
+	mul.d	$f0,$f0,$f2
+	mov.d	$f12,$f0
+	jal	atan
+	s.d	$f0,416($fp)
+	l.d	$f0,56($fp)
+	l.d	$f2,208($fp)
+	mul.d	$f0,$f0,$f2
+	mov.d	$f12,$f0
+	jal	atan
+	s.d	$f0,424($fp)
+	l.d	$f0,64($fp)
+	l.d	$f2,208($fp)
+	mul.d	$f0,$f0,$f2
+	mov.d	$f12,$f0
+	jal	atan
+	s.d	$f0,432($fp)
+	l.d	$f0,72($fp)
+	l.d	$f2,208($fp)
+	mul.d	$f0,$f0,$f2
+	mov.d	$f12,$f0
+	jal	atan
+	s.d	$f0,440($fp)
+	l.d	$f0,80($fp)
+	l.d	$f2,208($fp)
+	mul.d	$f0,$f0,$f2
+	mov.d	$f12,$f0
+	jal	atan
+	l.d	$f4,56($fp)
+	mul.d	$f2,$f20,$f4
+	l.d	$f6,64($fp)
+	mul.d	$f4,$f22,$f6
+	sub.d	$f2,$f2,$f4
+	l.d	$f6,72($fp)
+	mul.d	$f4,$f24,$f6
+	sub.d	$f2,$f2,$f4
+	l.d	$f6,80($fp)
+	mul.d	$f4,$f26,$f6
+	add.d	$f2,$f2,$f4
+	l.d	$f4,120($fp)
+	mul.d	$f2,$f2,$f4
+	l.d	$f4,56($fp)
+	neg.d	$f6,$f4
+	mul.d	$f4,$f6,$f28
+	l.d	$f8,64($fp)
+	mul.d	$f6,$f30,$f8
+	add.d	$f4,$f4,$f6
+	l.d	$f8,72($fp)
+	l.d	$f10,376($fp)
+	mul.d	$f6,$f10,$f8
+	add.d	$f4,$f4,$f6
+	l.d	$f8,80($fp)
+	l.d	$f10,384($fp)
+	mul.d	$f6,$f10,$f8
+	sub.d	$f4,$f4,$f6
+	l.d	$f6,128($fp)
+	mul.d	$f4,$f4,$f6
+	add.d	$f2,$f2,$f4
+	l.d	$f4,56($fp)
+	neg.d	$f6,$f4
+	l.d	$f10,392($fp)
+	mul.d	$f4,$f6,$f10
+	l.d	$f8,64($fp)
+	l.d	$f10,400($fp)
+	mul.d	$f6,$f10,$f8
+	add.d	$f4,$f4,$f6
+	l.d	$f8,72($fp)
+	l.d	$f10,408($fp)
+	mul.d	$f6,$f10,$f8
+	add.d	$f4,$f4,$f6
+	l.d	$f8,80($fp)
+	l.d	$f10,416($fp)
+	mul.d	$f6,$f10,$f8
+	sub.d	$f4,$f4,$f6
+	l.d	$f6,136($fp)
+	mul.d	$f4,$f4,$f6
+	add.d	$f2,$f2,$f4
+	l.d	$f6,56($fp)
+	l.d	$f10,424($fp)
+	mul.d	$f4,$f10,$f6
+	l.d	$f8,64($fp)
+	l.d	$f10,432($fp)
+	mul.d	$f6,$f10,$f8
+	sub.d	$f4,$f4,$f6
+	l.d	$f8,72($fp)
+	l.d	$f10,440($fp)
+	mul.d	$f6,$f10,$f8
+	sub.d	$f4,$f4,$f6
+	l.d	$f6,80($fp)
+	mul.d	$f0,$f0,$f6
+	add.d	$f4,$f4,$f0
+	l.d	$f6,144($fp)
+	mul.d	$f0,$f4,$f6
+	add.d	$f2,$f2,$f0
+	s.d	$f2,48($fp)
+	l.d	$f0,88($fp)
+	l.d	$f2,152($fp)
+	add.d	$f0,$f0,$f2
+	s.d	$f0,248($fp)
+	l.d	$f0,96($fp)
+	l.d	$f2,152($fp)
+	add.d	$f0,$f0,$f2
+	s.d	$f0,256($fp)
+	l.d	$f0,104($fp)
+	l.d	$f2,152($fp)
+	add.d	$f0,$f0,$f2
+	s.d	$f0,264($fp)
+	l.d	$f0,112($fp)
+	l.d	$f2,152($fp)
+	add.d	$f0,$f0,$f2
+	s.d	$f0,272($fp)
+	l.d	$f0,88($fp)
+	l.d	$f2,160($fp)
+	add.d	$f0,$f0,$f2
+	s.d	$f0,280($fp)
+	l.d	$f0,96($fp)
+	l.d	$f2,160($fp)
+	add.d	$f0,$f0,$f2
+	s.d	$f0,288($fp)
+	l.d	$f0,104($fp)
+	l.d	$f2,160($fp)
+	add.d	$f0,$f0,$f2
+	s.d	$f0,296($fp)
+	l.d	$f0,112($fp)
+	l.d	$f2,160($fp)
+	add.d	$f0,$f0,$f2
+	s.d	$f0,304($fp)
+	l.d	$f0,88($fp)
+	l.d	$f2,168($fp)
+	add.d	$f0,$f0,$f2
+	s.d	$f0,312($fp)
+	l.d	$f0,96($fp)
+	l.d	$f2,168($fp)
+	add.d	$f0,$f0,$f2
+	s.d	$f0,320($fp)
+	l.d	$f0,104($fp)
+	l.d	$f2,168($fp)
+	add.d	$f0,$f0,$f2
+	s.d	$f0,328($fp)
+	l.d	$f0,112($fp)
+	l.d	$f2,168($fp)
+	add.d	$f0,$f0,$f2
+	s.d	$f0,336($fp)
+	l.d	$f0,88($fp)
+	l.d	$f2,176($fp)
+	add.d	$f0,$f0,$f2
+	s.d	$f0,344($fp)
+	l.d	$f0,96($fp)
+	l.d	$f2,176($fp)
+	add.d	$f0,$f0,$f2
+	s.d	$f0,352($fp)
+	l.d	$f0,104($fp)
+	l.d	$f2,176($fp)
+	add.d	$f0,$f0,$f2
+	s.d	$f0,360($fp)
+	l.d	$f0,112($fp)
+	l.d	$f2,176($fp)
+	add.d	$f0,$f0,$f2
+	s.d	$f0,368($fp)
+	l.d	$f0,248($fp)
+	l.d	$f2,344($fp)
+	mul.d	$f0,$f0,$f2
+	l.d	$f2,280($fp)
+	l.d	$f4,312($fp)
+	mul.d	$f2,$f2,$f4
+	div.d	$f0,$f0,$f2
+	mov.d	$f12,$f0
+	jal	log
+	mov.d	$f20,$f0
+	l.d	$f0,256($fp)
+	l.d	$f2,352($fp)
+	mul.d	$f0,$f0,$f2
+	l.d	$f2,288($fp)
+	l.d	$f4,320($fp)
+	mul.d	$f2,$f2,$f4
+	div.d	$f0,$f0,$f2
+	mov.d	$f12,$f0
+	jal	log
+	mov.d	$f22,$f0
+	l.d	$f0,264($fp)
+	l.d	$f2,360($fp)
+	mul.d	$f0,$f0,$f2
+	l.d	$f2,296($fp)
+	l.d	$f4,328($fp)
+	mul.d	$f2,$f2,$f4
+	div.d	$f0,$f0,$f2
+	mov.d	$f12,$f0
+	jal	log
+	mov.d	$f24,$f0
+	l.d	$f0,272($fp)
+	l.d	$f2,368($fp)
+	mul.d	$f0,$f0,$f2
+	l.d	$f2,304($fp)
+	l.d	$f4,336($fp)
+	mul.d	$f2,$f2,$f4
+	div.d	$f0,$f0,$f2
+	mov.d	$f12,$f0
+	jal	log
+	mov.d	$f26,$f0
+	l.d	$f0,248($fp)
+	l.d	$f2,272($fp)
+	mul.d	$f0,$f0,$f2
+	l.d	$f2,256($fp)
+	l.d	$f4,264($fp)
+	mul.d	$f2,$f2,$f4
+	div.d	$f0,$f0,$f2
+	mov.d	$f12,$f0
+	jal	log
+	mov.d	$f28,$f0
+	l.d	$f0,280($fp)
+	l.d	$f2,304($fp)
+	mul.d	$f0,$f0,$f2
+	l.d	$f2,288($fp)
+	l.d	$f4,296($fp)
+	mul.d	$f2,$f2,$f4
+	div.d	$f0,$f0,$f2
+	mov.d	$f12,$f0
+	jal	log
+	mov.d	$f30,$f0
+	l.d	$f0,312($fp)
+	l.d	$f2,336($fp)
+	mul.d	$f0,$f0,$f2
+	l.d	$f2,320($fp)
+	l.d	$f4,328($fp)
+	mul.d	$f2,$f2,$f4
+	div.d	$f0,$f0,$f2
+	mov.d	$f12,$f0
+	jal	log
+	s.d	$f0,448($fp)
+	l.d	$f0,344($fp)
+	l.d	$f2,368($fp)
+	mul.d	$f0,$f0,$f2
+	l.d	$f2,352($fp)
+	l.d	$f4,360($fp)
+	mul.d	$f2,$f2,$f4
+	div.d	$f0,$f0,$f2
+	mov.d	$f12,$f0
+	jal	log
+	l.d	$f4,88($fp)
+	mul.d	$f2,$f20,$f4
+	l.d	$f6,96($fp)
+	mul.d	$f4,$f22,$f6
+	sub.d	$f2,$f2,$f4
+	l.d	$f6,104($fp)
+	mul.d	$f4,$f24,$f6
+	sub.d	$f2,$f2,$f4
+	l.d	$f6,112($fp)
+	mul.d	$f4,$f26,$f6
+	add.d	$f2,$f2,$f4
+	l.d	$f6,152($fp)
+	mul.d	$f4,$f28,$f6
+	sub.d	$f2,$f2,$f4
+	l.d	$f6,160($fp)
+	mul.d	$f4,$f30,$f6
+	add.d	$f2,$f2,$f4
+	l.d	$f6,168($fp)
+	l.d	$f10,448($fp)
+	mul.d	$f4,$f10,$f6
+	add.d	$f2,$f2,$f4
+	l.d	$f4,176($fp)
+	mul.d	$f0,$f0,$f4
+	sub.d	$f2,$f2,$f0
+	s.d	$f2,40($fp)
+	lw	$3,24($fp)
+	lw	$2,32($fp)
+	move	$4,$2
+	sll	$2,$4,3
+	sll	$4,$3,14
+	lw	$5,520($fp)
+	addu	$3,$4,$5
+	addu	$2,$2,$3
+	l.d	$f0,48($fp)
+	l.d	$f2,$LC76
+	mul.d	$f0,$f0,$f2
+	l.d	$f2,40($fp)
+	add.d	$f0,$f0,$f2
+	abs.d	$f2,$f0
+	s.d	$f2,0($2)
+	lw	$3,32($fp)
+	lw	$2,24($fp)
+	move	$4,$2
+	sll	$2,$4,3
+	sll	$4,$3,14
+	lw	$5,520($fp)
+	addu	$3,$4,$5
+	addu	$2,$2,$3
+	lw	$4,24($fp)
+	lw	$3,32($fp)
+	move	$5,$3
+	sll	$3,$5,3
+	sll	$5,$4,14
+	lw	$6,520($fp)
+	addu	$4,$5,$6
+	addu	$3,$3,$4
+	l.d	$f0,0($3)
+	s.d	$f0,0($2)
+$L171:
+	lw	$3,32($fp)
+	addu	$2,$3,1
+	move	$3,$2
+	sw	$3,32($fp)
+	j	$L169
+$L170:
+$L167:
+	lw	$3,24($fp)
+	addu	$2,$3,1
+	move	$3,$2
+	sw	$3,24($fp)
+	j	$L165
+$L166:
+$L163:
+	lw	$3,16($fp)
+	addu	$2,$3,1
+	move	$3,$2
+	sw	$3,16($fp)
+	j	$L161
+$L162:
+$L159:
+	lw	$3,20($fp)
+	addu	$2,$3,1
+	move	$3,$2
+	sw	$3,20($fp)
+	j	$L157
+$L158:
+$L156:
+	move	$sp,$fp			# sp not trusted here
+	lw	$31,460($sp)
+	lw	$fp,456($sp)
+	l.d	$f30,504($sp)
+	l.d	$f28,496($sp)
+	l.d	$f26,488($sp)
+	l.d	$f24,480($sp)
+	l.d	$f22,472($sp)
+	l.d	$f20,464($sp)
+	addu	$sp,$sp,512
+	j	$31
+	.end	SetUp2
+	.loc	1 787
+	.ent	SetUp3
+SetUp3:
+	.frame	$fp,96,$31		# vars= 72, regs= 2/0, args= 16, extra= 0
+	.mask	0xc0000000,-4
+	.fmask	0x00000000,0
+	subu	$sp,$sp,96
+	sw	$31,92($sp)
+	sw	$fp,88($sp)
+	move	$fp,$sp
+	sw	$4,96($fp)
+	sw	$5,100($fp)
+	sw	$6,104($fp)
+	sw	$7,108($fp)
+	sw	$0,20($fp)
+$L174:
+	lw	$2,20($fp)
+	slt	$3,$2,6
+	bne	$3,$0,$L177
+	j	$L175
+$L177:
+	sw	$0,16($fp)
+$L178:
+	lw	$2,16($fp)
+	slt	$3,$2,3
+	bne	$3,$0,$L181
+	j	$L179
+$L181:
+	lw	$2,16($fp)
+	move	$3,$2
+	sll	$2,$3,3
+	addu	$4,$fp,16
+	addu	$3,$2,$4
+	addu	$2,$3,24
+	lw	$4,20($fp)
+	lw	$3,16($fp)
+	move	$5,$3
+	sll	$3,$5,3
+	move	$6,$4
+	sll	$5,$6,1
+	addu	$5,$5,$4
+	sll	$4,$5,3
+	lw	$5,108($fp)
+	addu	$4,$4,$5
+	addu	$3,$3,$4
+	l.d	$f0,$LC77
+	l.d	$f2,0($3)
+	div.d	$f0,$f0,$f2
+	s.d	$f0,0($2)
+	lw	$2,16($fp)
+	move	$3,$2
+	sll	$2,$3,3
+	addu	$4,$fp,16
+	addu	$3,$2,$4
+	addu	$2,$3,48
+	lw	$4,20($fp)
+	lw	$3,16($fp)
+	move	$5,$3
+	sll	$3,$5,3
+	move	$6,$4
+	sll	$5,$6,1
+	addu	$5,$5,$4
+	sll	$4,$5,3
+	lw	$5,112($fp)
+	addu	$4,$4,$5
+	addu	$3,$3,$4
+	lw	$4,16($fp)
+	move	$5,$4
+	sll	$4,$5,3
+	addu	$5,$fp,16
+	addu	$4,$4,$5
+	addu	$5,$4,24
+	l.d	$f0,0($3)
+	l.d	$f2,0($5)
+	mul.d	$f0,$f0,$f2
+	s.d	$f0,0($2)
+$L180:
+	lw	$3,16($fp)
+	addu	$2,$3,1
+	move	$3,$2
+	sw	$3,16($fp)
+	j	$L178
+$L179:
+	.set	noreorder
+	nop
+	.set	reorder
+	lw	$2,20($fp)
+	move	$3,$2
+	sll	$2,$3,3
+	lw	$3,100($fp)
+	addu	$2,$2,$3
+	lw	$3,0($2)
+	sw	$3,24($fp)
+$L182:
+	lw	$2,20($fp)
+	move	$3,$2
+	sll	$2,$3,3
+	lw	$3,100($fp)
+	addu	$2,$2,$3
+	addu	$3,$2,4
+	lw	$2,24($fp)
+	lw	$3,0($3)
+	slt	$2,$3,$2
+	bne	$2,$0,$L183
+$L185:
+	dsz	32($fp)
+	sw	$0,28($fp)
+$L186:
+	lw	$2,20($fp)
+	move	$3,$2
+	sll	$2,$3,3
+	lw	$3,100($fp)
+	addu	$2,$2,$3
+	lw	$3,28($fp)
+	lw	$2,0($2)
+	slt	$3,$3,$2
+	bne	$3,$0,$L189
+	j	$L187
+$L189:
+	lw	$3,24($fp)
+	lw	$2,28($fp)
+	move	$4,$2
+	sll	$2,$4,3
+	sll	$4,$3,14
+	lw	$5,116($fp)
+	addu	$3,$4,$5
+	addu	$2,$2,$3
+	l.d	$f0,32($fp)
+	l.d	$f2,0($2)
+	add.d	$f0,$f0,$f2
+	s.d	$f0,32($fp)
+$L188:
+	lw	$3,28($fp)
+	addu	$2,$3,1
+	move	$3,$2
+	sw	$3,28($fp)
+	j	$L186
+$L187:
+	.set	noreorder
+	nop
+	.set	reorder
+	lw	$2,20($fp)
+	move	$3,$2
+	sll	$2,$3,3
+	lw	$3,100($fp)
+	addu	$2,$2,$3
+	addu	$3,$2,4
+	lw	$2,0($3)
+	addu	$3,$2,1
+	sw	$3,28($fp)
+$L190:
+	lw	$2,28($fp)
+	lw	$3,96($fp)
+	slt	$2,$2,$3
+	bne	$2,$0,$L193
+	j	$L191
+$L193:
+	lw	$3,24($fp)
+	lw	$2,28($fp)
+	move	$4,$2
+	sll	$2,$4,3
+	sll	$4,$3,14
+	lw	$5,116($fp)
+	addu	$3,$4,$5
+	addu	$2,$2,$3
+	l.d	$f0,32($fp)
+	l.d	$f2,0($2)
+	add.d	$f0,$f0,$f2
+	s.d	$f0,32($fp)
+$L192:
+	lw	$3,28($fp)
+	addu	$2,$3,1
+	move	$3,$2
+	sw	$3,28($fp)
+	j	$L190
+$L191:
+	lw	$2,24($fp)
+	move	$3,$2
+	sll	$2,$3,3
+	lw	$3,104($fp)
+	addu	$2,$2,$3
+	l.d	$f0,32($fp)
+	l.d	$f4,0($2)
+	sub.d	$f2,$f0,$f4
+	abs.d	$f0,$f2
+	l.d	$f2,32($fp)
+	l.d	$f4,$LC78
+	mul.d	$f2,$f2,$f4
+	c.lt.d	$f2,$f0
+	bc1f	$L194
+	la	$4,$LC79
+	jal	printf
+	move	$2,$0
+	j	$L173
+$L194:
+	l.d	$f0,32($fp)
+	neg.d	$f2,$f0
+	s.d	$f2,32($fp)
+	lw	$2,20($fp)
+	move	$3,$2
+	sll	$2,$3,3
+	lw	$3,100($fp)
+	addu	$2,$2,$3
+	lw	$3,0($2)
+	sw	$3,28($fp)
+$L195:
+	lw	$2,20($fp)
+	move	$3,$2
+	sll	$2,$3,3
+	lw	$3,100($fp)
+	addu	$2,$2,$3
+	addu	$3,$2,4
+	lw	$2,28($fp)
+	lw	$3,0($3)
+	slt	$2,$3,$2
+	bne	$2,$0,$L196
+$L198:
+	lw	$3,24($fp)
+	lw	$2,28($fp)
+	move	$4,$2
+	sll	$2,$4,3
+	sll	$4,$3,14
+	lw	$5,116($fp)
+	addu	$3,$4,$5
+	addu	$2,$2,$3
+	dsz	0($2)
+$L197:
+	lw	$3,28($fp)
+	addu	$2,$3,1
+	move	$3,$2
+	sw	$3,28($fp)
+	j	$L195
+$L196:
+	.set	noreorder
+	nop
+	.set	reorder
+	sw	$0,16($fp)
+$L199:
+	lw	$2,16($fp)
+	slt	$3,$2,3
+	bne	$3,$0,$L202
+	j	$L200
+$L202:
+	lw	$3,16($fp)
+	lw	$2,24($fp)
+	move	$4,$2
+	sll	$2,$4,3
+	sll	$4,$3,14
+	lw	$5,120($fp)
+	addu	$3,$4,$5
+	addu	$2,$2,$3
+	lw	$3,16($fp)
+	move	$4,$3
+	sll	$3,$4,3
+	addu	$4,$fp,16
+	addu	$3,$3,$4
+	addu	$4,$3,24
+	l.d	$f0,0($4)
+	l.d	$f2,32($fp)
+	mul.d	$f0,$f0,$f2
+	s.d	$f0,0($2)
+	lw	$3,16($fp)
+	lw	$2,24($fp)
+	move	$4,$2
+	sll	$2,$4,3
+	sll	$4,$3,14
+	lw	$5,124($fp)
+	addu	$3,$4,$5
+	addu	$2,$2,$3
+	lw	$3,16($fp)
+	move	$4,$3
+	sll	$3,$4,3
+	addu	$4,$fp,16
+	addu	$3,$3,$4
+	addu	$4,$3,48
+	l.d	$f0,0($4)
+	l.d	$f2,32($fp)
+	mul.d	$f0,$f0,$f2
+	s.d	$f0,0($2)
+$L201:
+	lw	$3,16($fp)
+	addu	$2,$3,1
+	move	$3,$2
+	sw	$3,16($fp)
+	j	$L199
+$L200:
+$L184:
+	lw	$3,24($fp)
+	addu	$2,$3,1
+	move	$3,$2
+	sw	$3,24($fp)
+	j	$L182
+$L183:
+$L176:
+	lw	$3,20($fp)
+	addu	$2,$3,1
+	move	$3,$2
+	sw	$3,20($fp)
+	j	$L174
+$L175:
+	li	$2,0x00000001		# 1
+	j	$L173
+$L173:
+	move	$sp,$fp			# sp not trusted here
+	lw	$31,92($sp)
+	lw	$fp,88($sp)
+	addu	$sp,$sp,96
+	j	$31
+	.end	SetUp3
+	.loc	1 855
+	.ent	Solver
+Solver:
+	.frame	$fp,56,$31		# vars= 24, regs= 2/0, args= 24, extra= 0
+	.mask	0xc0000000,-4
+	.fmask	0x00000000,0
+	subu	$sp,$sp,56
+	sw	$31,52($sp)
+	sw	$fp,48($sp)
+	move	$fp,$sp
+	sw	$4,56($fp)
+	sw	$5,60($fp)
+	sw	$6,64($fp)
+	sw	$7,68($fp)
+	sw	$0,36($fp)
+$L204:
+	lw	$2,36($fp)
+	slt	$3,$2,3
+	bne	$3,$0,$L207
+	j	$L205
+$L207:
+	lw	$2,60($fp)
+	sw	$2,24($fp)
+$L208:
+	lw	$2,24($fp)
+	lw	$3,56($fp)
+	slt	$2,$2,$3
+	bne	$2,$0,$L211
+	j	$L209
+$L211:
+	lw	$3,24($fp)
+	lw	$2,24($fp)
+	move	$4,$2
+	sll	$2,$4,3
+	sll	$4,$3,14
+	lw	$5,64($fp)
+	addu	$3,$4,$5
+	addu	$2,$2,$3
+	lw	$4,36($fp)
+	lw	$3,24($fp)
+	move	$5,$3
+	sll	$3,$5,3
+	sll	$5,$4,14
+	lw	$6,68($fp)
+	addu	$4,$5,$6
+	addu	$3,$3,$4
+	l.d	$f0,0($3)
+	s.d	$f0,0($2)
+	lw	$3,36($fp)
+	lw	$2,24($fp)
+	move	$4,$2
+	sll	$2,$4,3
+	sll	$4,$3,14
+	lw	$5,76($fp)
+	addu	$3,$4,$5
+	addu	$2,$2,$3
+	lw	$4,36($fp)
+	lw	$3,24($fp)
+	move	$5,$3
+	sll	$3,$5,3
+	sll	$5,$4,14
+	lw	$6,72($fp)
+	addu	$4,$5,$6
+	addu	$3,$3,$4
+	l.d	$f0,0($3)
+	s.d	$f0,0($2)
+	sw	$0,28($fp)
+$L212:
+	lw	$2,28($fp)
+	lw	$3,24($fp)
+	slt	$2,$2,$3
+	bne	$2,$0,$L215
+	j	$L213
+$L215:
+	lw	$3,24($fp)
+	lw	$2,28($fp)
+	move	$4,$2
+	sll	$2,$4,3
+	sll	$4,$3,14
+	lw	$5,64($fp)
+	addu	$3,$4,$5
+	addu	$2,$2,$3
+	lw	$4,28($fp)
+	lw	$3,24($fp)
+	move	$5,$3
+	sll	$3,$5,3
+	sll	$5,$4,14
+	lw	$6,64($fp)
+	addu	$4,$5,$6
+	addu	$3,$3,$4
+	l.d	$f0,0($3)
+	s.d	$f0,0($2)
+$L214:
+	lw	$3,28($fp)
+	addu	$2,$3,1
+	move	$3,$2
+	sw	$3,28($fp)
+	j	$L212
+$L213:
+$L210:
+	lw	$3,24($fp)
+	addu	$2,$3,1
+	move	$3,$2
+	sw	$3,24($fp)
+	j	$L208
+$L209:
+	.set	noreorder
+	nop
+	.set	reorder
+	sw	$0,28($fp)
+$L216:
+	lw	$2,28($fp)
+	lw	$3,60($fp)
+	slt	$2,$2,$3
+	bne	$2,$0,$L219
+	j	$L217
+$L219:
+	lw	$3,28($fp)
+	lw	$2,28($fp)
+	move	$4,$2
+	sll	$2,$4,3
+	sll	$4,$3,14
+	lw	$5,64($fp)
+	addu	$3,$4,$5
+	addu	$2,$2,$3
+	lw	$4,36($fp)
+	lw	$3,28($fp)
+	move	$5,$3
+	sll	$3,$5,3
+	sll	$5,$4,14
+	lw	$6,68($fp)
+	addu	$4,$5,$6
+	addu	$3,$3,$4
+	l.d	$f0,$LC80
+	l.d	$f2,0($3)
+	div.d	$f0,$f0,$f2
+	s.d	$f0,0($2)
+	lw	$3,36($fp)
+	lw	$2,28($fp)
+	move	$4,$2
+	sll	$2,$4,3
+	sll	$4,$3,14
+	lw	$5,76($fp)
+	addu	$3,$4,$5
+	addu	$2,$2,$3
+	lw	$4,36($fp)
+	lw	$3,28($fp)
+	move	$5,$3
+	sll	$3,$5,3
+	sll	$5,$4,14
+	lw	$6,72($fp)
+	addu	$4,$5,$6
+	addu	$3,$3,$4
+	l.d	$f0,0($3)
+	s.d	$f0,0($2)
+$L218:
+	lw	$3,28($fp)
+	addu	$2,$3,1
+	move	$3,$2
+	sw	$3,28($fp)
+	j	$L216
+$L217:
+	.set	noreorder
+	nop
+	.set	reorder
+	lw	$2,60($fp)
+	sw	$2,28($fp)
+$L220:
+	lw	$2,28($fp)
+	lw	$3,56($fp)
+	slt	$2,$2,$3
+	bne	$2,$0,$L223
+	j	$L221
+$L223:
+	lw	$2,60($fp)
+	sw	$2,32($fp)
+$L224:
+	lw	$2,32($fp)
+	lw	$3,28($fp)
+	slt	$2,$2,$3
+	bne	$2,$0,$L227
+	j	$L225
+$L227:
+	lw	$2,32($fp)
+	move	$3,$2
+	sll	$2,$3,14
+	lw	$3,64($fp)
+	addu	$2,$2,$3
+	lw	$3,28($fp)
+	move	$4,$3
+	sll	$3,$4,14
+	lw	$4,64($fp)
+	addu	$3,$3,$4
+	li	$4,0x00000001		# 1
+	sw	$4,16($sp)
+	lw	$4,32($fp)
+	move	$5,$2
+	li	$6,0x00000001		# 1
+	move	$7,$3
+	jal	Ddot
+	lw	$3,28($fp)
+	lw	$2,32($fp)
+	move	$4,$2
+	sll	$2,$4,3
+	sll	$4,$3,14
+	lw	$5,64($fp)
+	addu	$3,$4,$5
+	addu	$2,$2,$3
+	lw	$4,28($fp)
+	lw	$3,32($fp)
+	move	$5,$3
+	sll	$3,$5,3
+	sll	$5,$4,14
+	lw	$6,64($fp)
+	addu	$4,$5,$6
+	addu	$3,$3,$4
+	l.d	$f2,0($3)
+	sub.d	$f0,$f2,$f0
+	s.d	$f0,0($2)
+$L226:
+	lw	$3,32($fp)
+	addu	$2,$3,1
+	move	$3,$2
+	sw	$3,32($fp)
+	j	$L224
+$L225:
+	.set	noreorder
+	nop
+	.set	reorder
+	sw	$0,32($fp)
+$L228:
+	lw	$2,32($fp)
+	lw	$3,28($fp)
+	slt	$2,$2,$3
+	bne	$2,$0,$L231
+	j	$L229
+$L231:
+	lw	$3,28($fp)
+	lw	$2,32($fp)
+	move	$4,$2
+	sll	$2,$4,3
+	sll	$4,$3,14
+	lw	$5,64($fp)
+	addu	$3,$4,$5
+	addu	$2,$2,$3
+	l.d	$f0,0($2)
+	s.d	$f0,40($fp)
+	lw	$3,28($fp)
+	lw	$2,32($fp)
+	move	$4,$2
+	sll	$2,$4,3
+	sll	$4,$3,14
+	lw	$5,64($fp)
+	addu	$3,$4,$5
+	addu	$2,$2,$3
+	lw	$4,32($fp)
+	lw	$3,32($fp)
+	move	$5,$3
+	sll	$3,$5,3
+	sll	$5,$4,14
+	lw	$6,64($fp)
+	addu	$4,$5,$6
+	addu	$3,$3,$4
+	l.d	$f0,40($fp)
+	l.d	$f2,0($3)
+	mul.d	$f0,$f0,$f2
+	s.d	$f0,0($2)
+	lw	$3,28($fp)
+	lw	$2,28($fp)
+	move	$4,$2
+	sll	$2,$4,3
+	sll	$4,$3,14
+	lw	$5,64($fp)
+	addu	$3,$4,$5
+	addu	$2,$2,$3
+	lw	$4,28($fp)
+	lw	$3,28($fp)
+	move	$5,$3
+	sll	$3,$5,3
+	sll	$5,$4,14
+	lw	$6,64($fp)
+	addu	$4,$5,$6
+	addu	$3,$3,$4
+	lw	$5,28($fp)
+	lw	$4,32($fp)
+	move	$6,$4
+	sll	$4,$6,3
+	sll	$6,$5,14
+	lw	$7,64($fp)
+	addu	$5,$6,$7
+	addu	$4,$4,$5
+	l.d	$f0,40($fp)
+	l.d	$f2,0($4)
+	mul.d	$f0,$f0,$f2
+	l.d	$f2,0($3)
+	sub.d	$f0,$f2,$f0
+	s.d	$f0,0($2)
+$L230:
+	lw	$3,32($fp)
+	addu	$2,$3,1
+	move	$3,$2
+	sw	$3,32($fp)
+	j	$L228
+$L229:
+	lw	$3,28($fp)
+	lw	$2,28($fp)
+	move	$4,$2
+	sll	$2,$4,3
+	sll	$4,$3,14
+	lw	$5,64($fp)
+	addu	$3,$4,$5
+	addu	$2,$2,$3
+	lw	$4,28($fp)
+	lw	$3,28($fp)
+	move	$5,$3
+	sll	$3,$5,3
+	sll	$5,$4,14
+	lw	$6,64($fp)
+	addu	$4,$5,$6
+	addu	$3,$3,$4
+	l.d	$f0,$LC80
+	l.d	$f2,0($3)
+	div.d	$f0,$f0,$f2
+	s.d	$f0,0($2)
+$L222:
+	lw	$3,28($fp)
+	addu	$2,$3,1
+	move	$3,$2
+	sw	$3,28($fp)
+	j	$L220
+$L221:
+	.set	noreorder
+	nop
+	.set	reorder
+	lw	$2,60($fp)
+	sw	$2,32($fp)
+$L232:
+	lw	$2,32($fp)
+	lw	$3,56($fp)
+	slt	$2,$2,$3
+	bne	$2,$0,$L235
+	j	$L233
+$L235:
+	lw	$2,36($fp)
+	move	$3,$2
+	sll	$2,$3,14
+	lw	$3,76($fp)
+	addu	$2,$2,$3
+	lw	$3,32($fp)
+	move	$4,$3
+	sll	$3,$4,14
+	lw	$4,64($fp)
+	addu	$3,$3,$4
+	li	$4,0x00000001		# 1
+	sw	$4,16($sp)
+	lw	$4,32($fp)
+	move	$5,$2
+	li	$6,0x00000001		# 1
+	move	$7,$3
+	jal	Ddot
+	lw	$3,36($fp)
+	lw	$2,32($fp)
+	move	$4,$2
+	sll	$2,$4,3
+	sll	$4,$3,14
+	lw	$5,76($fp)
+	addu	$3,$4,$5
+	addu	$2,$2,$3
+	lw	$4,36($fp)
+	lw	$3,32($fp)
+	move	$5,$3
+	sll	$3,$5,3
+	sll	$5,$4,14
+	lw	$6,76($fp)
+	addu	$4,$5,$6
+	addu	$3,$3,$4
+	l.d	$f2,0($3)
+	sub.d	$f0,$f2,$f0
+	s.d	$f0,0($2)
+$L234:
+	lw	$3,32($fp)
+	addu	$2,$3,1
+	move	$3,$2
+	sw	$3,32($fp)
+	j	$L232
+$L233:
+	.set	noreorder
+	nop
+	.set	reorder
+	sw	$0,32($fp)
+$L236:
+	lw	$2,32($fp)
+	lw	$3,56($fp)
+	slt	$2,$2,$3
+	bne	$2,$0,$L239
+	j	$L237
+$L239:
+	lw	$3,36($fp)
+	lw	$2,32($fp)
+	move	$4,$2
+	sll	$2,$4,3
+	sll	$4,$3,14
+	lw	$5,76($fp)
+	addu	$3,$4,$5
+	addu	$2,$2,$3
+	lw	$4,36($fp)
+	lw	$3,32($fp)
+	move	$5,$3
+	sll	$3,$5,3
+	sll	$5,$4,14
+	lw	$6,76($fp)
+	addu	$4,$5,$6
+	addu	$3,$3,$4
+	lw	$5,32($fp)
+	lw	$4,32($fp)
+	move	$6,$4
+	sll	$4,$6,3
+	sll	$6,$5,14
+	lw	$7,64($fp)
+	addu	$5,$6,$7
+	addu	$4,$4,$5
+	l.d	$f0,0($3)
+	l.d	$f2,0($4)
+	mul.d	$f0,$f0,$f2
+	s.d	$f0,0($2)
+$L238:
+	lw	$3,32($fp)
+	addu	$2,$3,1
+	move	$3,$2
+	sw	$3,32($fp)
+	j	$L236
+$L237:
+	.set	noreorder
+	nop
+	.set	reorder
+	lw	$2,56($fp)
+	subu	$3,$2,2
+	sw	$3,32($fp)
+$L240:
+	lw	$2,32($fp)
+	lw	$3,60($fp)
+	slt	$2,$2,$3
+	bne	$2,$0,$L241
+$L243:
+	lw	$3,56($fp)
+	subu	$2,$3,1
+	lw	$3,32($fp)
+	subu	$2,$2,$3
+	lw	$3,36($fp)
+	move	$4,$3
+	sll	$3,$4,14
+	lw	$4,76($fp)
+	addu	$3,$3,$4
+	lw	$4,32($fp)
+	move	$5,$4
+	sll	$4,$5,3
+	addu	$5,$4,8
+	addu	$3,$3,$5
+	lw	$4,32($fp)
+	move	$5,$4
+	sll	$4,$5,14
+	addu	$5,$4,16384
+	lw	$6,64($fp)
+	addu	$4,$5,$6
+	lw	$5,32($fp)
+	move	$6,$5
+	sll	$5,$6,3
+	addu	$7,$4,$5
+	li	$4,0x00000800		# 2048
+	sw	$4,16($sp)
+	move	$4,$2
+	move	$5,$3
+	li	$6,0x00000001		# 1
+	jal	Ddot
+	lw	$3,36($fp)
+	lw	$2,32($fp)
+	move	$4,$2
+	sll	$2,$4,3
+	sll	$4,$3,14
+	lw	$5,76($fp)
+	addu	$3,$4,$5
+	addu	$2,$2,$3
+	lw	$4,36($fp)
+	lw	$3,32($fp)
+	move	$5,$3
+	sll	$3,$5,3
+	sll	$5,$4,14
+	lw	$6,76($fp)
+	addu	$4,$5,$6
+	addu	$3,$3,$4
+	l.d	$f2,0($3)
+	sub.d	$f0,$f2,$f0
+	s.d	$f0,0($2)
+$L242:
+	lw	$3,32($fp)
+	subu	$2,$3,1
+	move	$3,$2
+	sw	$3,32($fp)
+	j	$L240
+$L241:
+	.set	noreorder
+	nop
+	.set	reorder
+	lw	$2,60($fp)
+	subu	$3,$2,1
+	sw	$3,32($fp)
+$L244:
+	lw	$2,32($fp)
+	bgez	$2,$L247
+	j	$L245
+$L247:
+	lw	$2,56($fp)
+	lw	$3,60($fp)
+	subu	$2,$2,$3
+	lw	$3,36($fp)
+	move	$4,$3
+	sll	$3,$4,14
+	lw	$4,76($fp)
+	addu	$3,$3,$4
+	lw	$4,60($fp)
+	move	$5,$4
+	sll	$4,$5,3
+	addu	$3,$3,$4
+	lw	$4,60($fp)
+	move	$5,$4
+	sll	$4,$5,14
+	lw	$5,64($fp)
+	addu	$4,$4,$5
+	lw	$5,32($fp)
+	move	$6,$5
+	sll	$5,$6,3
+	addu	$7,$4,$5
+	li	$4,0x00000800		# 2048
+	sw	$4,16($sp)
+	move	$4,$2
+	move	$5,$3
+	li	$6,0x00000001		# 1
+	jal	Ddot
+	lw	$3,36($fp)
+	lw	$2,32($fp)
+	move	$4,$2
+	sll	$2,$4,3
+	sll	$4,$3,14
+	lw	$5,76($fp)
+	addu	$3,$4,$5
+	addu	$2,$2,$3
+	lw	$4,36($fp)
+	lw	$3,32($fp)
+	move	$5,$3
+	sll	$3,$5,3
+	sll	$5,$4,14
+	lw	$6,76($fp)
+	addu	$4,$5,$6
+	addu	$3,$3,$4
+	l.d	$f2,0($3)
+	sub.d	$f0,$f2,$f0
+	s.d	$f0,0($2)
+$L246:
+	lw	$3,32($fp)
+	subu	$2,$3,1
+	move	$3,$2
+	sw	$3,32($fp)
+	j	$L244
+$L245:
+$L206:
+	lw	$3,36($fp)
+	addu	$2,$3,1
+	move	$3,$2
+	sw	$3,36($fp)
+	j	$L204
+$L205:
+$L203:
+	move	$sp,$fp			# sp not trusted here
+	lw	$31,52($sp)
+	lw	$fp,48($sp)
+	addu	$sp,$sp,56
+	j	$31
+	.end	Solver
+	.loc	1 919
+	.ent	Storer
+Storer:
+	.frame	$fp,80,$31		# vars= 16, regs= 2/0, args= 56, extra= 0
+	.mask	0xc0000000,-4
+	.fmask	0x00000000,0
+	subu	$sp,$sp,80
+	sw	$31,76($sp)
+	sw	$fp,72($sp)
+	move	$fp,$sp
+	sw	$4,80($fp)
+	sw	$5,84($fp)
+	sw	$6,88($fp)
+	sw	$7,92($fp)
+	la	$4,$LC32
+	la	$5,$LC81
+	jal	fopen
+	sw	$2,68($fp)
+	lw	$2,68($fp)
+	bne	$2,$0,$L249
+	la	$4,$LC82
+	jal	printf
+	li	$4,0x00000001		# 1
+	jal	exit
+$L249:
+	lw	$4,68($fp)
+	la	$5,$LC50
+	lw	$6,80($fp)
+	jal	fprintf
+	lw	$4,68($fp)
+	la	$5,$LC83
+	jal	fprintf
+	sw	$0,60($fp)
+$L250:
+	lw	$2,60($fp)
+	slt	$3,$2,6
+	bne	$3,$0,$L253
+	j	$L251
+$L253:
+	lw	$2,60($fp)
+	move	$3,$2
+	sll	$2,$3,3
+	lw	$3,84($fp)
+	addu	$2,$2,$3
+	lw	$3,0($2)
+	sw	$3,64($fp)
+$L254:
+	lw	$2,60($fp)
+	move	$3,$2
+	sll	$2,$3,3
+	lw	$3,84($fp)
+	addu	$2,$2,$3
+	addu	$3,$2,4
+	lw	$2,64($fp)
+	lw	$3,0($3)
+	slt	$2,$3,$2
+	bne	$2,$0,$L255
+$L257:
+	lw	$3,64($fp)
+	addu	$2,$3,1
+	lw	$4,60($fp)
+	addu	$3,$4,1
+	lw	$4,64($fp)
+	move	$5,$4
+	sll	$4,$5,3
+	lw	$5,88($fp)
+	addu	$4,$4,$5
+	l.d	$f0,0($4)
+	s.d	$f0,16($sp)
+	lw	$4,64($fp)
+	move	$5,$4
+	sll	$4,$5,3
+	lw	$5,88($fp)
+	addu	$4,$4,$5
+	addu	$5,$4,16384
+	l.d	$f0,0($5)
+	s.d	$f0,24($sp)
+	lw	$4,64($fp)
+	move	$5,$4
+	sll	$4,$5,3
+	lw	$5,88($fp)
+	addu	$4,$4,$5
+	li	$5,0x00008000		# 32768
+	addu	$4,$4,$5
+	l.d	$f0,0($4)
+	s.d	$f0,32($sp)
+	lw	$4,64($fp)
+	move	$5,$4
+	sll	$4,$5,3
+	lw	$5,92($fp)
+	addu	$4,$4,$5
+	l.d	$f0,0($4)
+	s.d	$f0,40($sp)
+	lw	$4,64($fp)
+	move	$5,$4
+	sll	$4,$5,3
+	lw	$5,92($fp)
+	addu	$4,$4,$5
+	addu	$5,$4,16384
+	l.d	$f0,0($5)
+	s.d	$f0,48($sp)
+	lw	$4,68($fp)
+	la	$5,$LC84
+	move	$6,$2
+	move	$7,$3
+	jal	fprintf
+$L256:
+	lw	$3,64($fp)
+	addu	$2,$3,1
+	move	$3,$2
+	sw	$3,64($fp)
+	j	$L254
+$L255:
+$L252:
+	lw	$3,60($fp)
+	addu	$2,$3,1
+	move	$3,$2
+	sw	$3,60($fp)
+	j	$L250
+$L251:
+	lw	$4,68($fp)
+	la	$5,$LC85
+	jal	fprintf
+	sw	$0,60($fp)
+$L258:
+	lw	$2,60($fp)
+	slt	$3,$2,6
+	bne	$3,$0,$L261
+	j	$L259
+$L261:
+	lw	$2,60($fp)
+	move	$3,$2
+	sll	$2,$3,3
+	lw	$3,84($fp)
+	addu	$2,$2,$3
+	lw	$3,0($2)
+	sw	$3,64($fp)
+$L262:
+	lw	$2,60($fp)
+	move	$3,$2
+	sll	$2,$3,3
+	lw	$3,84($fp)
+	addu	$2,$2,$3
+	addu	$3,$2,4
+	lw	$2,64($fp)
+	lw	$3,0($3)
+	slt	$2,$3,$2
+	bne	$2,$0,$L263
+$L265:
+	lw	$3,64($fp)
+	addu	$2,$3,1
+	lw	$4,60($fp)
+	addu	$3,$4,1
+	lw	$4,64($fp)
+	move	$5,$4
+	sll	$4,$5,3
+	lw	$5,96($fp)
+	addu	$4,$4,$5
+	l.d	$f0,0($4)
+	s.d	$f0,16($sp)
+	lw	$4,64($fp)
+	move	$5,$4
+	sll	$4,$5,3
+	lw	$5,96($fp)
+	addu	$4,$4,$5
+	addu	$5,$4,16384
+	l.d	$f0,0($5)
+	s.d	$f0,24($sp)
+	lw	$4,64($fp)
+	move	$5,$4
+	sll	$4,$5,3
+	lw	$5,96($fp)
+	addu	$4,$4,$5
+	li	$5,0x00008000		# 32768
+	addu	$4,$4,$5
+	l.d	$f0,0($4)
+	s.d	$f0,32($sp)
+	lw	$4,68($fp)
+	la	$5,$LC86
+	move	$6,$2
+	move	$7,$3
+	jal	fprintf
+$L264:
+	lw	$3,64($fp)
+	addu	$2,$3,1
+	move	$3,$2
+	sw	$3,64($fp)
+	j	$L262
+$L263:
+$L260:
+	lw	$3,60($fp)
+	addu	$2,$3,1
+	move	$3,$2
+	sw	$3,60($fp)
+	j	$L258
+$L259:
+	lw	$4,68($fp)
+	jal	fclose
+$L248:
+	move	$sp,$fp			# sp not trusted here
+	lw	$31,76($sp)
+	lw	$fp,72($sp)
+	addu	$sp,$sp,80
+	j	$31
+	.end	Storer
+	.loc	1 972
+	.ent	Verify
+Verify:
+	.frame	$fp,72,$31		# vars= 48, regs= 2/0, args= 16, extra= 0
+	.mask	0xc0000000,-4
+	.fmask	0x00000000,0
+	subu	$sp,$sp,72
+	sw	$31,68($sp)
+	sw	$fp,64($sp)
+	move	$fp,$sp
+	sw	$4,72($fp)
+	sw	$5,76($fp)
+	sw	$6,80($fp)
+	sw	$7,84($fp)
+	dsz	16($fp)
+	sw	$0,56($fp)
+$L267:
+	lw	$2,56($fp)
+	slt	$3,$2,3
+	bne	$3,$0,$L270
+	j	$L268
+$L270:
+	sw	$0,48($fp)
+$L271:
+	lw	$2,48($fp)
+	lw	$3,72($fp)
+	slt	$2,$2,$3
+	bne	$2,$0,$L274
+	j	$L272
+$L274:
+	lw	$3,48($fp)
+	lw	$2,48($fp)
+	move	$4,$2
+	sll	$2,$4,3
+	sll	$4,$3,14
+	lw	$5,76($fp)
+	addu	$3,$4,$5
+	addu	$2,$2,$3
+	lw	$4,56($fp)
+	lw	$3,48($fp)
+	move	$5,$3
+	sll	$3,$5,3
+	sll	$5,$4,14
+	lw	$6,80($fp)
+	addu	$4,$5,$6
+	addu	$3,$3,$4
+	l.d	$f0,0($3)
+	s.d	$f0,0($2)
+	sw	$0,52($fp)
+$L275:
+	lw	$2,52($fp)
+	lw	$3,48($fp)
+	slt	$2,$2,$3
+	bne	$2,$0,$L278
+	j	$L276
+$L278:
+	lw	$3,48($fp)
+	lw	$2,52($fp)
+	move	$4,$2
+	sll	$2,$4,3
+	sll	$4,$3,14
+	lw	$5,76($fp)
+	addu	$3,$4,$5
+	addu	$2,$2,$3
+	lw	$4,52($fp)
+	lw	$3,48($fp)
+	move	$5,$3
+	sll	$3,$5,3
+	sll	$5,$4,14
+	lw	$6,76($fp)
+	addu	$4,$5,$6
+	addu	$3,$3,$4
+	l.d	$f0,0($3)
+	s.d	$f0,0($2)
+$L277:
+	lw	$3,52($fp)
+	addu	$2,$3,1
+	move	$3,$2
+	sw	$3,52($fp)
+	j	$L275
+$L276:
+$L273:
+	lw	$3,48($fp)
+	addu	$2,$3,1
+	move	$3,$2
+	sw	$3,48($fp)
+	j	$L271
+$L272:
+	mtc1	$0,$f0
+	mtc1	$0,$f1
+	s.d	$f0,40($fp)
+	s.d	$f0,32($fp)
+	sw	$0,52($fp)
+$L279:
+	lw	$2,52($fp)
+	lw	$3,72($fp)
+	slt	$2,$2,$3
+	bne	$2,$0,$L282
+	j	$L280
+$L282:
+	lw	$3,56($fp)
+	lw	$2,52($fp)
+	move	$4,$2
+	sll	$2,$4,3
+	sll	$4,$3,14
+	lw	$5,84($fp)
+	addu	$3,$4,$5
+	addu	$2,$2,$3
+	l.d	$f0,0($2)
+	s.d	$f0,24($fp)
+	sw	$0,48($fp)
+$L283:
+	lw	$2,48($fp)
+	lw	$3,72($fp)
+	slt	$2,$2,$3
+	bne	$2,$0,$L286
+	j	$L284
+$L286:
+	lw	$3,52($fp)
+	lw	$2,48($fp)
+	move	$4,$2
+	sll	$2,$4,3
+	sll	$4,$3,14
+	lw	$5,76($fp)
+	addu	$3,$4,$5
+	addu	$2,$2,$3
+	lw	$4,56($fp)
+	lw	$3,48($fp)
+	move	$5,$3
+	sll	$3,$5,3
+	sll	$5,$4,14
+	lw	$6,88($fp)
+	addu	$4,$5,$6
+	addu	$3,$3,$4
+	l.d	$f0,0($2)
+	l.d	$f2,0($3)
+	mul.d	$f0,$f0,$f2
+	l.d	$f2,24($fp)
+	sub.d	$f0,$f2,$f0
+	s.d	$f0,24($fp)
+	lw	$3,52($fp)
+	lw	$2,48($fp)
+	move	$4,$2
+	sll	$2,$4,3
+	sll	$4,$3,14
+	lw	$5,76($fp)
+	addu	$3,$4,$5
+	addu	$2,$2,$3
+	l.d	$f2,0($2)
+	abs.d	$f0,$f2
+	l.d	$f2,32($fp)
+	c.lt.d	$f0,$f2
+	bc1f	$L287
+	l.d	$f0,32($fp)
+	j	$L288
+$L287:
+	lw	$3,52($fp)
+	lw	$2,48($fp)
+	move	$4,$2
+	sll	$2,$4,3
+	sll	$4,$3,14
+	lw	$5,76($fp)
+	addu	$3,$4,$5
+	addu	$2,$2,$3
+	l.d	$f2,0($2)
+	abs.d	$f0,$f2
+$L288:
+	s.d	$f0,32($fp)
+$L285:
+	lw	$3,48($fp)
+	addu	$2,$3,1
+	move	$3,$2
+	sw	$3,48($fp)
+	j	$L283
+$L284:
+	lw	$3,56($fp)
+	lw	$2,52($fp)
+	move	$4,$2
+	sll	$2,$4,3
+	sll	$4,$3,14
+	lw	$5,88($fp)
+	addu	$3,$4,$5
+	addu	$2,$2,$3
+	l.d	$f2,0($2)
+	abs.d	$f0,$f2
+	l.d	$f2,40($fp)
+	c.lt.d	$f0,$f2
+	bc1f	$L289
+	l.d	$f0,40($fp)
+	j	$L290
+$L289:
+	lw	$3,56($fp)
+	lw	$2,52($fp)
+	move	$4,$2
+	sll	$2,$4,3
+	sll	$4,$3,14
+	lw	$5,88($fp)
+	addu	$3,$4,$5
+	addu	$2,$2,$3
+	l.d	$f2,0($2)
+	abs.d	$f0,$f2
+$L290:
+	s.d	$f0,40($fp)
+	l.d	$f2,24($fp)
+	abs.d	$f0,$f2
+	l.d	$f2,16($fp)
+	add.d	$f0,$f2,$f0
+	s.d	$f0,16($fp)
+$L281:
+	lw	$3,52($fp)
+	addu	$2,$3,1
+	move	$3,$2
+	sw	$3,52($fp)
+	j	$L279
+$L280:
+$L269:
+	lw	$3,56($fp)
+	addu	$2,$3,1
+	move	$3,$2
+	sw	$3,56($fp)
+	j	$L267
+$L268:
+	l.d	$f0,32($fp)
+	l.d	$f2,40($fp)
+	mul.d	$f0,$f0,$f2
+	l.d	$f2,16($fp)
+	div.d	$f0,$f2,$f0
+	s.d	$f0,16($fp)
+	l.d	$f0,16($fp)
+	l.d	$f2,$LC87
+	c.lt.d	$f2,$f0
+	bc1f	$L291
+	la	$4,$LC88
+	dlw	$6,16($fp)
+	jal	printf
+	move	$2,$0
+	j	$L266
+$L291:
+	li	$2,0x00000001		# 1
+	j	$L266
+$L266:
+	move	$sp,$fp			# sp not trusted here
+	lw	$31,68($sp)
+	lw	$fp,64($sp)
+	addu	$sp,$sp,72
+	j	$31
+	.end	Verify
+	.loc	1 1120
+	.ent	Ddot
+Ddot:
+	.frame	$fp,8,$31		# vars= 0, regs= 1/0, args= 0, extra= 0
+	.mask	0x40000000,-8
+	.fmask	0x00000000,0
+	subu	$sp,$sp,8
+	sw	$fp,0($sp)
+	move	$fp,$sp
+	move	$2,$6
+	move	$6,$7
+	lw	$3,24($fp)
+	mtc1	$0,$f2
+	mtc1	$0,$f3
+$L293:
+	subu	$4,$4,1
+	li	$7,-1				# 0xffffffff
+	bne	$4,$7,$L295
+	j	$L294
+$L295:
+	l.d	$f0,0($5)
+	l.d	$f4,0($6)
+	mul.d	$f0,$f0,$f4
+	add.d	$f2,$f2,$f0
+	move	$7,$2
+	sll	$8,$7,3
+	addu	$5,$5,$8
+	move	$7,$3
+	sll	$8,$7,3
+	addu	$6,$6,$8
+	j	$L293
+$L294:
+	mov.d	$f0,$f2
+	j	$L292
+$L292:
+	move	$sp,$fp			# sp not trusted here
+	lw	$fp,0($sp)
+	addu	$sp,$sp,8
+	j	$31
+	.end	Ddot

src/main.py → main.py

@@ -1,7 +1,7 @@
 #!/usr/bin/python
-from parser import parse_file
-from optimize import optimize
-from writer import write_statements
+from src.parser import parse_file
+from src.optimize import optimize
+from src.writer import write_statements
 
 if __name__ == '__main__':
     from sys import argv, exit

report/Makefile

@@ -0,0 +1,10 @@
+RM=rm -rf
+
+all: report.pdf
+
+%.pdf: %.tex
+	pdflatex $^
+	pdflatex $^
+
+clean:
+	$(RM) *.pdf *.aux *.log *.out *.toc *.snm *.nav

report/report.tex

@@ -11,39 +11,230 @@
 \usepackage{hyperref}
 
 \title{Peephole Optimizer}
-\author{Jayke Meijer (6049885), Richard Torenvliet (6138861), Taddeus Kroes (6054129)}
+\author{Jayke Meijer (6049885), Richard Torenvliet (6138861), Tadde\"us Kroes
+    (6054129)}
 
 \begin{document}
 \maketitle
+\tableofcontents
+\pagebreak
 
 \section{Introduction}
-The goal of the assignment is to implement the optimization stage of the compiler. To reach this goal the parser part of the compiler has to be implemented.
 
-The output of the gcc cross compiler on a c program is our input, the output of the gcc cross compiler is in the form of Assembly code, but not optimized. Our assignment includes a number of c programs, an important part of the assignment is parsing the data. Parsing the data is done with lex and yacc. The lexer is a program that finds keywords that meets the regular expression provided in the lexer. After the lexer, the yaccer takes over. Yaccer can turn the keywords in to an action.
+The goal of the assignment is to implement the optimization stage of the
+compiler. To reach this goal the parser and the optimizer part of the compiler
+have to be implemented.
+
+The output of the xgcc cross compiler on a C program is our input. The output
+of the xgcc cross compiler is in the form of Assembly code, but not optimized.
+Our assignment includes a number of C programs. An important part of the
+assignment is parsing the data. Parsing the data is done with Lex and Yacc. The
+Lexer is a program that finds keywords that meets the regular expression
+provided in the Lexer. After the Lexer, the Yaccer takes over. Yacc can turn
+the keywords in to an action.
+
+\section{Design}
+
+There are two general types of of optimizations of the assembly code, global
+optimizations and optimizations on a so-called basic block. These optimizations
+will be discussed separately
+
+\subsection{Global optimizations}
+
+We only perform one global optimization, which is optimizing branch-jump
+statements. The unoptimized Assembly code contains sequences of code of the
+following structure:
+\begin{verbatim}
+    beq ...,$Lx
+    j $Ly
+$Lx:   ...
+\end{verbatim}
+This is inefficient, since there is a jump to a label that follows this code.
+It would be more efficient to replace the branch statement with a \texttt{bne}
+(the opposite case) to the label used in the jump statement. This way the jump
+statement can be eliminated, since the next label follows anyway. The same can
+of course be done for the opposite case, where a \texttt{bne} is changed into a
+\texttt{beq}.
+
+Since this optimization is done between two series of codes with jumps and
+labels, we can not perform this code during the basic block optimizations. The
+reason for this will become clearer in the following section.
+
+\subsection{Basic Block Optimizations}
+
+Optimizations on basic blocks are a more important part of the optimizer.
+First, what is a basic block? A basic block is a sequence of statements
+guaranteed to be executed in that order, and that order alone. This is the case
+for a piece of code not containing any branches or jumps.
+
+To create a basic block, you need to define what is the leader of a basic
+block. We call a statement a leader if it is either a jump/branch statement, or
+the target of such a statement. Then a basic block runs from one leader until
+the next leader.
+
+There are quite a few optimizations we perform on these basic blocks, so we
+will describe the types of optimizations here in stead of each optimization.
+
+\subsubsection*{Standard peephole optimizations}
+
+These are optimizations that simply look for a certain statement or pattern of
+statements, and optimize these. For example,
+\begin{verbatim}
+mov $regA,$regB
+instr $regA, $regA,... 
+\end{verbatim}
+can be optimized into
+\begin{verbatim}
+instr $regA, $regB,...
+\end{verbatim}
+since the register \texttt{\$regA} gets overwritten by the second instruction
+anyway, and the instruction can easily use \texttt{\$regB} in stead of
+\texttt{\$regA}. There are a few more of these cases, which are the same as
+those described on the practicum page
+\footnote{\url{http://staff.science.uva.nl/~andy/compiler/prac.html}} and in
+Appendix \ref{opt}.
+
+\subsubsection*{Common subexpression elimination}
+
+A more advanced optimization is common subexpression elimination. This means
+that expensive operations as a multiplication or addition are performed only
+once and the result is then `copied' into variables where needed.
+
+A standard method for doing this is the creation of a DAG or Directed Acyclic
+Graph. However, this requires a fairly advanced implementation. Our
+implementation is a slightly less fancy, but easier to implement.
+We search from the end of the block up for instructions that are eligible for
+CSE. If we find one, we check further up in the code for the same instruction,
+and add that to a temporary storage list. This is done until the beginning of
+the block or until one of the arguments of this expression is assigned.
+
+We now add the instruction above the first use, and write the result in a new
+variable. Then all occurrences of this expression can be replaced by a move of
+from new variable into the original destination variable of the instruction.
+
+This is a less efficient method then the dag, but because the basic blocks are
+in general not very large and the execution time of the optimizer is not a
+primary concern, this is not a big problem.
+
+\section{Implementation}
+
+We decided to implement the optimization in Python. We chose this programming
+language because Python is an easy language to manipulate strings, work
+object-oriented etc.
+It turns out that a Lex and Yacc are also available as a Python module,
+named PLY(Python Lex-Yacc). This allows us to use one language, Python, instead
+of two, i.e. C and Python. Also no debugging is needed in C, only in Python
+which makes our assignment more feasible.
+
+The program has three steps, parsing the Assembly code into a datastructure we
+can use, the so-called Intermediate Representation, performing optimizations on
+this IR and writing the IR back to Assembly.
+
+\subsection{Parsing}
+
+The parsing is done with PLY, which allows us to perform Lex-Yacc tasks in
+Python by using a Lex-Yacc like syntax. This way there is no need to combine
+languages like we should do otherwise since Lex and Yacc are coupled with C.
+
+The decision was made to not recognize exactly every possible instruction in
+the parser, but only if something is for example a command, a comment or a gcc
+directive. We then transform per line to a object called a Statement. A
+statement has a type, a name and optionally a list of arguments. These
+statements together form a statement list, which is placed in another object
+called a Block. In the beginning there is one block for the entire program, but
+after global optimizations this will be separated in several blocks that are
+the basic blocks.
+
+\subsection{Optimizations}
+
+The optimizations are done in two different steps. First the global
+optimizations are performed, which are only the optimizations on branch-jump
+constructions. This is done repeatedly until there are no more changes.
+
+After all possible global optimizations are done, the program is seperated into
+basic blocks. The algorithm to do this is described earlier, and means all
+jump and branch instructions are called leaders, as are their targets. A basic
+block then goes from leader to leader.
+
+After the division in basic blocks, optimizations are performed on each of
+these basic blocks. This is also done repeatedly, since some times several
+steps can be done to optimize something.
+
+\subsection{Writing}
+
+Once all the optimizations have been done, the IR needs to be rewritten into
+Assembly code, so the xgcc crosscompiler can make binary code out of it.
+
+The writer expects a list of statements, so first the blocks have to be
+concatenated again into a list. After this is done, the list is passed on to
+the writer, which writes the instructions back to Assembly and saves the file
+so we can let xgcc compile it.
 
-\section{Design \& Implementation}
-We decided to implement the optimization in python. We chose this programming language because python is an easy language to manipulate strings, work objective ori\"ented etc.
-It turns out that a lex and yacc are also implemented in a python version, named PLY(Python Lex-Yacc). This allows us to use one language, Python, instead of two i.e. C and Python. Also no debugging is needed in C, only in Python which makes our assignment more feasible.
+\section{Results}
 
-\subsection{Design}
+\subsection{pi.c}
 
+\subsection{acron.c}
 
-\subsection*{Implementation}
-This 
+\subsection{whet.c}
 
-\subsubsection*{PLY}
+\subsection{slalom.c}
+
+\subsection{clinpack.c}
+
+\section{Conclusion}
+
+\appendix
+
+\section{List of all optimizations}
+
+\label{opt}
+
+\textbf{Global optimizations}
+
+\begin{verbatim}
+    beq ...,$Lx             bne ...,$Ly
+    j $Ly               ->  $Lx:   ...
+$Lx:   ...
+
+
+    bne ...,$Lx             beq ...,$Ly
+    j $Ly               ->  $Lx:   ...
+$Lx:   ...
+\end{verbatim}
+\textbf{Standard basic block optimizations}
+
+\begin{verbatim}
+mov $regA,$regA         ->  --- // remove it
+
+
+mov $regA,$regB         ->  instr $regA, $regB,...
+instr $regA, $regA,...
+
+
+instr $regA,...             instr $4,...
+mov [$4-$7], $regA      ->  jal XXX
+jal  XXX
 
-\section{Results}
 
-\subsection*{pi.c}
+sw $regA,XXX            ->  sw $regA, XXX
+ld $regA,XXX
 
-\subsection*{arcron.c}
 
-\subsection*{whet.c}
+shift $regA,$regA,0     ->  --- // remove it
 
-\subsection*{slalom.c}
 
-\subsection*{clinpack.c}
+add $regA,$regA,X       ->  lw ...,X($regA)
+lw ...,0($regA)
+\end{verbatim}
+\textbf{Advanced basic block optimizations}
 
-\section{conclusion}
-\end{document}
+\begin{verbatim}
+# Common subexpression elimination
+addu $regA, $regB, 4        addu $regD, $regB, 4
+...                         move $regA, $regD
+Code not writing $regB  ->  ...
+...                         ...
+addu $regC, $regB, 4        move $regC, $regD
+\end{verbatim}
+\end{document}

src/dataflow.py

@@ -1,4 +1,4 @@
-from copy import copy
+#from copy import copy
 
 from statement import Block
 
@@ -122,55 +122,102 @@ def generate_flow_graph(blocks):
             b.add_edge_to(blocks[i + 1])
 
 
-def generate_dominator_tree(nodes):
-    """Add dominator administration to the given flow graph nodes."""
-    # Dominator of the start node is the start itself
-    nodes[0].dom = set([nodes[0]])
-
-    # For all other nodes, set all nodes as the dominators
-    for n in nodes[1:]:
-        n.dom = set(copy(nodes))
-
-    def pred(n, known=[]):
-        """Recursively find all predecessors of a node."""
-        direct = filter(lambda x: x not in known, n.edges_from)
-        p = copy(direct)
-
-        for ancestor in direct:
-            p += pred(ancestor, direct)
-
-        return p
-
-    # Iteratively eliminate nodes that are not dominators
-    changed = True
-
-    while changed:
-        changed = False
-
-        for n in nodes[1:]:
-            old_dom = n.dom
-            intersection = lambda p1, p2: p1.dom & p2.dom
-            n.dom = set([n]) | reduce(intersection, pred(n), set([]))
-
-            if n.dom != old_dom:
-                changed = True
-
-    def idom(d, n):
-        """Check if d immediately dominates n."""
-        for b in n.dom:
-            if b != d and b != n and b in n.dom:
-                return False
-
-        return True
-
-    # Build tree using immediate dominators
-    for n in nodes:
-        for d in n.dom:
-            if idom(d, n):
-                d.set_dominates(n)
-                break
-
-# statements = parse_file(...)
-# b = find_basic_blocks(statements)
-# generate_flow_graph(b)  # nodes now have edges
-# generate_dominator_tree(b)  # nodes now have dominators
+#def generate_dominator_tree(nodes):
+#    """Add dominator administration to the given flow graph nodes."""
+#    # Dominator of the start node is the start itself
+#    nodes[0].dom = set([nodes[0]])
+#
+#    # For all other nodes, set all nodes as the dominators
+#    for n in nodes[1:]:
+#        n.dom = set(copy(nodes))
+#
+#    def pred(n, known=[]):
+#        """Recursively find all predecessors of a node."""
+#        direct = filter(lambda x: x not in known, n.edges_from)
+#        p = copy(direct)
+#
+#        for ancestor in direct:
+#            p += pred(ancestor, direct)
+#
+#        return p
+#
+#    # Iteratively eliminate nodes that are not dominators
+#    changed = True
+#
+#    while changed:
+#        changed = False
+#
+#        for n in nodes[1:]:
+#            old_dom = n.dom
+#            intersection = lambda p1, p2: p1.dom & p2.dom
+#            n.dom = set([n]) | reduce(intersection, pred(n), set([]))
+#
+#            if n.dom != old_dom:
+#                changed = True
+#
+#    def idom(d, n):
+#        """Check if d immediately dominates n."""
+#        for b in n.dom:
+#            if b != d and b != n and b in n.dom:
+#                return False
+#
+#        return True
+#
+#    # Build tree using immediate dominators
+#    for n in nodes:
+#        for d in n.dom:
+#            if idom(d, n):
+#                d.set_dominates(n)
+#                break
+
+
+class Dag:
+    def __init__(self, block):
+        """Create the Directed Acyclic Graph of all binary operations in a
+        basic block."""
+        self.nodes = []
+
+        for s in block:
+            if s.is_command('move') or s.is_monop():
+                rd, rs = s
+                y = self.find_reg_node(rs)
+                self.find_op_node(s.name, rd, y)
+            elif s.is_binop():
+                rd, rs, rt = s
+                y = self.find_reg_node(rs)
+                z = self.find_reg_node(rt)
+                self.find_op_node(s.name, rd, y, z)
+
+    def find_reg_node(self, reg):
+        for n in self.nodes:
+            if reg in n.reg:
+                return n
+
+        node = DagLeaf(reg)
+        self.nodes.append(node)
+
+        return node
+
+    def find_op_node(self, op, rd, *args):
+        for n in self.nodes:
+            if not isinstance(n, DagLeaf) and n.op == op and n.nodes == args:
+                n.labels.append(rd)
+
+                return n
+
+        node = DagNode(op, rd, *args)
+        self.nodes.append(node)
+
+        return node
+
+
+class DagNode:
+    def __init__(self, op, label, *args):
+        self.op = op
+        self.labels = [label]
+        self.nodes = args
+
+
+class DagLeaf:
+    def __init__(self, reg):
+        self.reg = reg

src/optimize.py

@@ -1,152 +0,0 @@
-import re
-
-from dataflow import find_basic_blocks
-
-
-def optimize_global(statements):
-    """Optimize statement sequences on a global level."""
-    old_len = -1
-
-    while old_len != len(statements):
-        old_len = len(statements)
-
-        while not statements.end():
-            s = statements.read()
-
-            # mov $regA, $regA          ->  --- remove it
-            if s.is_command('move') and s[0] == s[1]:
-                statements.replace(1, [])
-                continue
-
-            # mov $regA, $regB          ->  instr $regA, $regB, ...
-            # instr $regA, $regA, ...
-            if s.is_command('move'):
-                ins = statements.peek()
-
-                if ins and len(ins) >= 2 and ins[0] == s[0] and ins[1] == s[0]:
-                    ins[1] = s[1]
-                    statements.replace(2, [ins])
-                    continue
-
-            # instr $regA, ...          ->  instr $4, ...
-            # mov $4, $regA                 jal XX
-            # jal XX
-            if s.is_command() and len(s):
-                following = statements.peek(2)
-
-                if len(following) == 2:
-                    mov, jal = following
-
-                    if mov.is_command('move') and mov[1] == s[0] \
-                            and re.match('^\$[4-7]$', mov[0]) \
-                            and jal.is_command('jal'):
-                        s[0] = mov[0]
-                        statements.replace(2, [s])
-                        continue
-
-            # sw $regA, XX              ->  sw $regA, XX
-            # ld $regA, XX
-            if s.is_command('sw'):
-                ld = statements.peek()
-
-                if ld.is_command('ld') and ld[0] == s[0]:
-                    statements.replace(2, [s])
-                    continue
-
-            # shift $regA, $regA, 0     ->  --- remove it
-            if s.is_shift() and s[0] == s[1] and s[2] == 0:
-                statements.replace(1, [])
-                continue
-
-            # add $regA, $regA, X       ->  lw ..., X($regA)
-            # lw ..., 0($regA)
-            if s.is_command('add') and s[0] == s[1]:
-                lw = statements.peek()
-
-                if lw.is_command('lw') and lw[-1] == '0(%s)' % s[0]:
-                    lw[-1] = str(s[2]) + lw[-1][1:]
-                    statements.replace(2, [lw])
-                    continue
-
-            #     beq ..., $Lx          ->      bne ..., $Ly
-            #     j $Ly                     $Lx:
-            # $Lx:
-            if s.is_command('beq'):
-                following = statements.peek(2)
-
-                if len(following) == 2:
-                    j, label = following
-
-                    if j.is_command('j') and label.is_label(s[2]):
-                        s.name = 'bne'
-                        s[2] = j[0]
-                        statements.replace(3, [s, label])
-
-
-def optimize_blocks(blocks):
-    """Call the optimizer for each basic block. Do this several times until
-    no more optimizations are achieved."""
-    changed = True
-
-    while changed:
-        changed = False
-        optimized = []
-
-        for block in blocks:
-            block_changed, b = optimize_block(block)
-            optimized.append(b)
-
-            if block_changed:
-                changed = True
-
-        blocks = optimized
-
-    
-    
-
-    return reduce(lambda a, b: a + b, blocks, [])
-
-
-def optimize_block(statements):
-    """Optimize a basic block."""
-    changed = False
-    output_statements = []
-
-    for statement in statements:
-        new_statement = statement
-
-        output_statements.append(new_statement)
-
-    return changed, output_statements
-
-
-def optimize(statements, verbose=0):
-    """optimization wrapper function, calls global and basic-block level
-    optimization functions."""
-    # Optimize on a global level
-    o = len(statements)
-    optimize_global(statements)
-    g = len(statements)
-
-    # Optimize basic blocks
-    basic_blocks = find_basic_blocks(statements)
-    blocks = optimize_blocks(basic_blocks)
-    block_statements = map(lambda b: b.statements, blocks)
-    opt_blocks = reduce(lambda a, b: a + b, block_statements)
-    b = len(opt_blocks)
-
-    # - Common subexpression elimination
-    # - Constant folding
-    # - Copy propagation
-    # - Dead-code elimination
-    # - Temporary variable renaming
-    # - Interchange of independent statements
-
-    if verbose:
-        print 'Original statements:             %d' % o
-        print 'After global optimization:       %d' % g
-        print 'After basic blocks optimization: %d' % b
-        print 'Speedup:                         %d (%d%%)' \
-                % (b - o, int((b - o) / o * 100))
-
-    return opt_blocks

src/optimize/__init__.py

@@ -0,0 +1,94 @@
+from src.dataflow import find_basic_blocks
+
+from standard import redundant_move_1, redundant_move_2, \
+        redundant_move_3, redundant_move_4, redundant_load, \
+        redundant_shift, redundant_add
+from advanced import eliminate_common_subexpressions, fold_constants, \
+    copy_propagation
+
+
+def optimize_global(statements):
+    """Optimize statement sequences on a global level."""
+    old_len = -1
+
+    while old_len != len(statements):
+        old_len = len(statements)
+
+        while not statements.end():
+            s = statements.read()
+
+            #     beq/bne ..., $Lx      ->      bne/beq ..., $Ly
+            #     j $Ly                     $Lx:
+            # $Lx:
+            if s.is_command('beq', 'bne'):
+                following = statements.peek(2)
+
+                if len(following) == 2:
+                    j, label = following
+
+                    if j.is_command('j') and label.is_label(s[2]):
+                        s.name = 'bne' if s.is_command('beq') else 'beq'
+                        s[2] = j[0]
+                        statements.replace(3, [s, label])
+
+
+def optimize_block(block):
+    """Optimize a basic block."""
+    standard = [redundant_move_1, redundant_move_2, redundant_move_3, \
+                redundant_move_4, redundant_load, redundant_shift, \
+                redundant_add]
+    old_len = -1
+
+    # Standard optimizations
+    while old_len != len(block):
+        old_len = len(block)
+
+        while not block.end():
+            s = block.read()
+
+            for callback in standard:
+                if callback(s, block):
+                    break
+
+    # Advanced optimizations
+    #changed = True
+
+    #while changed:
+    #    changed = eliminate_common_subexpressions(block) \
+    #              or fold_constants(block)
+
+    while eliminate_common_subexpressions(block) \
+            | fold_constants(block) \
+            | copy_propagation(block):
+        pass
+
+def optimize(statements, verbose=0):
+    """Optimization wrapper function, calls global and basic-block level
+    optimization functions."""
+    # Optimize on a global level
+    o = len(statements)
+    optimize_global(statements)
+    g = len(statements)
+
+    # Optimize basic blocks
+    blocks = find_basic_blocks(statements)
+    map(optimize_block, blocks)
+    block_statements = map(lambda b: b.statements, blocks)
+    opt_blocks = reduce(lambda a, b: a + b, block_statements)
+    b = len(opt_blocks)
+
+    # - Common subexpression elimination
+    # - Constant folding
+    # - Copy propagation
+    # - Dead-code elimination
+    # - Temporary variable renaming
+    # - Interchange of independent statements
+
+    if verbose:
+        print 'Original statements:             %d' % o
+        print 'After global optimization:       %d' % g
+        print 'After basic blocks optimization: %d' % b
+        print 'Optimization:                    %d (%d%%)' \
+                % (b - o, int((b - o) / float(o) * 100))
+
+    return opt_blocks

src/optimize/advanced.py

@@ -0,0 +1,202 @@
+from src.statement import Statement as S
+
+
+def create_variable():
+    return '$15'
+
+
+def eliminate_common_subexpressions(block):
+    """
+    Common subexpression elimination:
+    - Traverse through the statements in reverse order.
+    - If the statement can be possibly be eliminated, walk further collecting
+      all other occurrences of the expression until one of the arguments is
+      assigned in a statement, or the start of the block has been reached.
+    - If one or more occurrences were found, insert the expression with a new
+      destination address before the last found occurrence and change all
+      occurrences to a move instruction from that address.
+    """
+    found = False
+    block.reverse_statements()
+
+    while not block.end():
+        s = block.read()
+
+        if s.is_arith():
+            pointer = block.pointer
+            last = False
+            new_reg = False
+            args = s[1:]
+
+            # Collect similar statements
+            while not block.end():
+                s2 = block.read()
+
+                # Stop if one of the arguments is assigned
+                if len(s2) and s2[0] in args:
+                    break
+
+                # Replace a similar expression by a move instruction
+                if s2.name == s.name and s2[1:] == args:
+                    if not new_reg:
+                        new_reg = create_variable()
+
+                    block.replace(1, [S('command', 'move', s2[0], new_reg)])
+                    last = block.pointer
+
+            # Insert an additional expression with a new destination address
+            if last:
+                block.insert(S('command', s.name, [new_reg] + args), last)
+                found = True
+
+            # Reset pointer to and continue from the original statement
+            block.pointer = pointer
+
+    block.reverse_statements()
+
+    return found
+
+
+def to_hex(value):
+    """Create the hexadecimal string of an integer."""
+    return '0x%08x' % value
+
+
+def fold_constants(block):
+    """
+    Constant folding:
+    - An immidiate load defines a register value:
+        li $reg, XX     ->  register[$reg] = XX
+    - Integer variable definition is of the following form:
+        li $reg, XX     ->  constants[VAR] = XX
+        sw $reg, VAR    ->  register[$reg] = XX
+    - When a variable is used, the following happens:
+        lw $reg, VAR    ->  register[$reg] = constants[VAR]
+    """
+    found = False
+
+    # Variable values
+    constants = {}
+
+    # Current known values in register
+    register = {}
+
+    while not block.end():
+        s = block.read()
+
+        if not s.is_command():
+            continue
+
+        if s.name == 'li':
+            # Save value in register
+            register[s[0]] = int(s[1], 16)
+        elif s.name == 'move' and s[0] in register:
+            reg_to, reg_from = s
+
+            if reg_from in register:
+                # Other value is also known, copy its value
+                register[reg_to] = register[reg_to]
+            else:
+                # Other value is unknown, delete the value
+                del register[reg_to]
+        elif s.name == 'sw' and s[0] in register:
+            # Constant variable definition, e.g. 'int a = 1;'
+            constants[s[1]] = register[s[0]]
+        elif s.name == 'lw' and s[1] in constants:
+            # Usage of variable with constant value
+            register[s[0]] = constants[s[1]]
+        elif s.name in ['addu', 'subu', 'mult', 'div']:
+            # Calculation with constants
+            rd, rs, rt = s
+            rs_known = rs in register
+            rt_known = rt in register
+
+            if rs_known and rt_known:
+                # a = 5         ->  b = 15
+                # c = 10
+                # b = a + c
+                rs_val = register[rs]
+                rt_val = register[rt]
+
+                if s.name == 'addu':
+                    result = to_hex(rs_val + rt_val)
+
+                if s.name == 'subu':
+                    result = to_hex(rs_val - rt_val)
+
+                if s.name == 'mult':
+                    result = to_hex(rs_val * rt_val)
+
+                if s.name == 'div':
+                    result = to_hex(rs_val / rt_val)
+
+                block.replace(1, [S('command', 'li', result)])
+                register[rd] = result
+                found = True
+            elif rt_known:
+                # c = 10        ->  b = a + 10
+                # b = c + a
+                s[2] = register[rt]
+                found = True
+            elif rs_known and s.name in ['addu', 'mult']:
+                # a = 10        ->  b = c + 10
+                # b = c + a
+                s[1] = rt
+                s[2] = register[rs]
+                found = True
+        elif len(s) and s[0] in register:
+            # Known register is overwritten, remove its value
+            del register[s[0]]
+
+    return found
+
+
+def copy_propagation(block):
+    """
+    Rename values that were copied to there original, so the copy statement
+    might be useless, allowing it to be removed by dead code elimination.
+    """
+    moves_from = []
+    moves_to = []
+
+    while not block.end():
+        s = block.read()
+        
+        if len(s) == 3:
+            print "s[0] = ", s[0]
+            print "s[1] = ", s[1]
+            print "s[2] = ", s[2]
+            
+            if moves_from:
+                print "fr: ", moves_from
+                print "to: ", moves_to
+
+        if s.is_command('move') and s[0] not in moves_to:
+            moves_from.append(s[1])
+            moves_to.append(s[0])
+            print "Added move to list."
+        elif s.is_command('move'):
+            for i in xrange(len(moves_to)):
+                if moves_to[i] == s[0]:
+                    moves_from[i] = s[1]
+        elif len(s) == 3 and s[0] in moves_to:
+            print len(s)
+            print len(moves_to)
+            for i in xrange(len(moves_to)):
+                if moves_to[i] == s[0]:
+                    del moves_to[i]
+                    del moves_from[i]
+                    "Removed move from list."
+        elif len(s) == 3 and (s[1] in moves_to or s[2] in moves_to):
+            print "Have to propagate."
+            for i in xrange(len(moves_to)):
+                if s[1] == moves_to[i]:
+                    s[1] = moves_from[i]
+                    print "Propagated"
+                
+                if s[2] == moves_to[i]:
+                    s[2] = moves_from[i]
+                    print "Propagated"
+                    
+        print ""       
+    return False

src/optimize/standard.py

@@ -0,0 +1,101 @@
+import re
+
+
+def redundant_move_1(mov, statements):
+    """
+    mov $regA, $regA          ->  --- remove it
+    """
+    if mov.is_command('move') and mov[0] == mov[1]:
+        statements.replace(1, [])
+
+        return True
+
+
+def redundant_move_2(mov, statements):
+    """
+    mov $regA, $regB          ->  instr $regA, $regB, ...
+    instr $regA, $regA, ...
+    """
+    if mov.is_command('move'):
+        ins = statements.peek()
+
+        if ins and len(ins) >= 2 and ins[0] == mov[0] and ins[1] == mov[0]:
+            ins[1] = mov[1]
+            statements.replace(2, [ins])
+
+            return True
+
+
+def redundant_move_3(ins, statements):
+    """
+    instr $regA, ...          ->  instr $4, ...
+    mov $4, $regA                 jal XX
+    jal XX
+    """
+    if ins.is_command() and len(ins):
+        following = statements.peek(2)
+
+        if len(following) == 2:
+            mov, jal = following
+
+            if mov.is_command('move') and mov[1] == ins[0] \
+                    and re.match('^\$[4-7]$', mov[0]) \
+                    and jal.is_command('jal'):
+                ins[0] = mov[0]
+                statements.replace(2, [ins])
+
+                return True
+
+
+def redundant_move_4(mov1, statements):
+    """
+    mov $RegA, $RegB         ->  move $RegA, $RegB
+    mov $RegB, $RegA
+    """
+    if mov1.is_command('move'):
+        mov2 = statements.peek()
+
+        if mov2.is_command('move') and mov2[0] == mov1[1] and \
+                mov2[1] == mov1[0]:
+            statements.replace(2, [mov1])
+
+            return True
+
+
+def redundant_load(sw, statements):
+    """
+    sw $regA, XX              ->  sw $regA, XX
+    ld $regA, XX
+    """
+    if sw.is_command('sw'):
+        ld = statements.peek()
+
+        if ld.is_command('lw') and ld.args == sw.args:
+            statements.replace(2, [sw])
+
+            return True
+
+
+def redundant_shift(shift, statements):
+    """
+    shift $regA, $regA, 0     ->  --- remove it
+    """
+    if shift.is_shift() and shift[0] == shift[1] and shift[2] == 0:
+        statements.replace(1, [])
+
+        return True
+
+
+def redundant_add(add, statements):
+    """
+    add $regA, $regA, X       ->  lw ..., X($regA)
+    lw ..., 0($regA)
+    """
+    if add.is_command('addu') and add[0] == add[1] and isinstance(add[2], int):
+        lw = statements.peek()
+
+        if lw.is_load() and lw[-1] == '0(%s)' % add[0]:
+            lw[-1] = '%s(%s)' % (add[2], add[0])
+            statements.replace(2, [lw])
+
+            return True

src/parser.py

@@ -3,11 +3,13 @@ import ply.yacc as yacc
 
 from statement import Statement as S, Block
 
+
 # Global statements administration
 statements = []
 
 tokens = ('NEWLINE', 'WORD', 'COMMENT', 'DIRECTIVE', 'COMMA', 'COLON')
 
+
 # Tokens
 def t_NEWLINE(t):
     r'\n+'
@@ -32,7 +34,7 @@ def t_DIRECTIVE(t):
     return t
 
 def t_hex_word(t):
-    r'0x[0-9a-fA-F]{8}'
+    r'0x([0-9a-fA-F]{8}|[0-9a-fA-F]{4})'
     t.type = 'WORD'
     return t
 
@@ -47,9 +49,10 @@ def t_int(t):
     return t
 
 def t_WORD(t):
-    r'[a-zA-Z0-9$_.+()]+'
+    r'[a-zA-Z0-9$_.+()-]+'
     return t
 
+
 # Ignore whitespaces
 t_ignore = ' \t'
 
@@ -57,9 +60,11 @@ def t_error(t):
     print('Illegal character "%s"' % t.value[0])
     t.lexer.skip(1)
 
+
 # Build the lexer
 lexer = lex.lex()
 
+
 # Parsing rules
 start = 'input'
 
@@ -102,9 +107,14 @@ def p_command(p):
 def p_error(p):
     print 'Syntax error at "%s" on line %d' % (p.value, lexer.lineno)
 
+
+# Build YACC
 yacc.yacc()
 
+
 def parse_file(filename):
+    """Parse a given Assembly file, return a Block with Statement objects
+    containing the parsed instructions."""
     global statements
 
     statements = []

src/statement.py

@@ -21,7 +21,7 @@ class Statement:
         arguments."""
         return self.stype == other.stype and self.name == other.name \
                and self.args == other.args
-               
+
     def __len__(self):
         return len(self.args)
 
@@ -45,9 +45,8 @@ class Statement:
         return self.stype == 'label' if name == None \
                else self.stype == 'label' and self.name == name
 
-    def is_command(self, name=None):
-        return self.stype == 'command' if name == None \
-               else self.stype == 'command' and self.name == name
+    def is_command(self, *args):
+        return self.stype == 'command' and (not len(args) or self.name in args)
 
     def is_jump(self):
         """Check if the statement is a jump."""
@@ -74,6 +73,14 @@ class Statement:
         return self.is_command() \
                and re.match('^(add|sub|mult|div|abs|neg)(u|\.d)?$', self.name)
 
+    def is_monop(self):
+        """Check if the statement is an unary operation."""
+        return len(self) == 2 and self.is_arith()
+
+    def is_binop(self):
+        """Check if the statement is an binary operation."""
+        return self.is_command() and len(self) == 3 and not self.is_jump()
+
     def jump_target(self):
         """Get the jump target of this statement."""
         if not self.is_jump():
@@ -111,6 +118,9 @@ class Block:
     def peek(self, count=1):
         """Read the statements until an offset from the current pointer
         position."""
+        if self.end():
+            return Statement('empty', '') if count == 1 else []
+
         return self.statements[self.pointer] if count == 1 \
                else self.statements[self.pointer:self.pointer + count]
 
@@ -120,7 +130,7 @@ class Block:
         replacement."""
         if self.pointer == 0:
             raise Exception('No statement have been read yet.')
-        
+
         if start == None:
             start = self.pointer - 1
 
@@ -129,7 +139,18 @@ class Block:
         self.statements = before + replacement + after
         self.pointer = start + len(replacement)
 
+    def insert(self, statement, index=None):
+        if index == None:
+            index = self.pointer
+
+        self.statements.insert(index, statement)
+
     def apply_filter(self, callback):
         """Apply a filter to the statement list. If the callback returns True,
         the statement will remain in the list.."""
         self.statements = filter(callback, self.statements)
+
+    def reverse_statements(self):
+        """Reverse the statement list and reset the pointer."""
+        self.statements = self.statements[::-1]
+        self.pointer = 0

src/writer.py

@@ -1,5 +1,6 @@
 from math import ceil
 
+
 def write_statements(statements):
     """Write a list of statements to valid assembly code."""
     out = ''
@@ -12,13 +13,15 @@ def write_statements(statements):
         if s.is_label():
             line = s.name + ':'
             indent_level = 1
-        elif s.is_inline_comment():
-            line = '#' + s.name
-            l = len(prevline.expandtabs(4))
-            tabs = int(ceil((24 - l) / 4.)) + 1
-            newline = '\t' * tabs
         elif s.is_comment():
-            line = '\t' * indent_level + line
+            line = '#' + s.name
+
+            if s.is_inline_comment():
+                l = len(prevline.expandtabs(4))
+                tabs = int(ceil((24 - l) / 4.)) + 1
+                newline = '\t' * tabs
+            else:
+                line = '\t' * indent_level + line
         elif s.is_directive():
             line = '\t' + s.name
         elif s.is_command():
@@ -37,6 +40,9 @@ def write_statements(statements):
         out += newline + line
         prevline = line
 
+    # Add newline at end of file
+    out += '\n'
+
     return out
 
 def write_to_file(filename, statements):

tests/test_dataflow.py

@@ -2,7 +2,7 @@ import unittest
 
 from src.statement import Statement as S
 from src.dataflow import BasicBlock as B, find_leaders, find_basic_blocks, \
-        generate_flow_graph
+        generate_flow_graph, Dag, DagNode, DagLeaf
 
 
 class TestDataflow(unittest.TestCase):
@@ -73,3 +73,54 @@ class TestDataflow(unittest.TestCase):
         self.assertEqual(b2.edges_to, [b3])
         self.assertIn(b1, b3.edges_from)
         self.assertIn(b2, b3.edges_from)
+
+    def test_dag_unary(self):
+        dag = Dag(B([S('command', 'neg.d', '$rd', '$rs')]))
+        expect = Dag([])
+        expect.nodes = [DagLeaf('$rs'), DagNode('neg.d', '$rd', \
+                        DagLeaf('$rs'))]
+
+        self.assertEqualDag(dag, expect)
+
+    def test_dag_binary(self):
+        dag = Dag(B([S('command', 'addu', '$rd', '$r1', '$r2')]))
+        expect = Dag([])
+        expect.nodes = [DagLeaf('$r1'),
+                        DagLeaf('$r2'),
+                        DagNode('addu', '$rd', DagLeaf('$r1'), DagLeaf('$r2'))]
+
+        self.assertEqualDag(dag, expect)
+
+#    def test_dag_combinednode(self):
+#        dag = Dag(B([S('command', 'mult', '$rd1', '$r1', '$r2'),
+#                     S('command', 'mult', '$rd2', '$r1', '$r2')]))
+#        expect = Dag([])
+#        multnode = DagNode('mult',
+#                           DagLeaf('$r1'),
+#                           DagLeaf('$r2'))
+#        multnode.labels = ['$rd1', '$rd2']
+#        expect.nodes = [DagLeaf('$r1'),
+#                        DagLeaf('$r2'),
+#                        multnode]
+#
+#        self.assertEqualDag(dag, expect)
+
+    def assertEqualDag(self, dag1, dag2):
+        self.assertEqual(len(dag1.nodes), len(dag2.nodes))
+
+        for node1, node2 in zip(dag1.nodes, dag2.nodes):
+            self.assertEqualNodes(node1, node2)
+
+    def assertEqualNodes(self, node1, node2):
+        if isinstance(node1, DagLeaf):
+            self.assertIsInstance(node2, DagLeaf)
+            self.assertEqual(node1.reg, node2.reg)
+        elif isinstance(node2, DagLeaf):
+            raise AssertionError
+        else:
+            self.assertEqual(node1.op, node2.op)
+            self.assertEqual(node1.labels, node2.labels)
+            self.assertEqual(len(node1.nodes), len(node2.nodes))
+
+            for child1, child2 in zip(node1.nodes, node2.nodes):
+                self.assertEqualNodes(child1, child2)

tests/test_optimize.py

@@ -1,218 +1,225 @@
 import unittest
 
-from src.optimize import optimize_global
+from src.optimize import optimize_global, optimize_block
 from src.statement import Statement as S, Block as B
 
 
 class TestOptimize(unittest.TestCase):
 
     def setUp(self):
-        pass
+        self.foo = S('command', 'foo')
+        self.bar = S('command', 'bar')
 
-    def test_optimize_global_movaa(self):
-        foo = S('command', 'foo')
-        bar = S('command', 'bar')
-        block = B([foo,
+    def tearDown(self):
+        del self.foo
+        del self.bar
+
+    def test_optimize_block_movaa(self):
+        block = B([self.foo,
                    S('command', 'move', '$regA', '$regA'),
-                   bar])
-        optimize_global(block)
-        self.assertEquals(block.statements, [foo, bar])
-        
-    def test_optimize_global_movab(self):
-        foo = S('command', 'foo')
+                   self.bar])
+        optimize_block(block)
+        self.assertEquals(block.statements, [self.foo, self.bar])
+
+    def test_optimize_block_movab(self):
         move = S('command', 'move', '$regA', '$regB')
-        bar = S('command', 'bar')
-        block = B([foo,
+        block = B([self.foo,
                    move,
-                   bar])
-        optimize_global(block)
-        self.assertEquals(block.statements, [foo, move, bar])
-        
-    def test_optimize_global_movinst_true(self):
-        foo = S('command', 'foo')
-        bar = S('command', 'bar')
-        
-        block = B([foo,
+                   self.bar])
+        optimize_block(block)
+        self.assertEquals(block.statements, [self.foo, move, self.bar])
+
+    def test_optimize_block_movinst_true(self):
+        block = B([self.foo,
                    S('command', 'move', '$regA', '$regB'),
                    S('command', 'addu', '$regA', '$regA', 2),
-                   bar])
-        optimize_global(block)
-        self.assertEquals(block.statements, [foo,                     
-                   S('command', 'addu', '$regA', '$regB', 2),      
-                   bar])
-                   
-    def test_optimize_global_movinst_false(self):
-        foo = S('command', 'foo')
-        bar = S('command', 'bar')
-        statements = [foo, \
+                   self.bar])
+        optimize_block(block)
+        self.assertEquals(block.statements, [self.foo,
+                   S('command', 'addu', '$regA', '$regB', 2),
+                   self.bar])
+
+    def test_optimize_block_movinst_false(self):
+        statements = [self.foo, \
                       S('command', 'move', '$regA', '$regB'), \
                       S('command', 'addu', '$regA', '$regC', 2), \
-                      bar]
-        
+                      self.bar]
+
         block = B(statements)
-        optimize_global(block)
+        optimize_block(block)
         self.assertEquals(block.statements, statements)
-                
-    def test_optimize_global_instr_mov_jal_true(self):
-        foo = S('command', 'foo')
-        bar = S('command', 'bar')
-        
-        block = B([foo,
+
+    def test_optimize_block_instr_mov_jal_true(self):
+        block = B([self.foo,
                    S('command', 'addu', '$regA', '$regC', 2),
                    S('command', 'move', '$4', '$regA'),
                    S('command', 'jal', 'L1'),
-                   bar])
-        optimize_global(block)
-        
-        self.assertEquals(block.statements, [foo,                     
+                   self.bar])
+        optimize_block(block)
+
+        self.assertEquals(block.statements, [self.foo,
                    S('command', 'addu', '$4', '$regC', 2),
-                   S('command', 'jal', 'L1'),      
-                   bar])
-        
-    def test_optimize_global_instr_mov_jal_false(self):
-        foo = S('command', 'foo')
-        bar = S('command', 'bar')
-        
-        arguments = [foo, \
-                      S('command', 'addu', '$regA', '$regC', 2), \
-                      S('command', 'move', '$3', '$regA'), \
-                      S('command', 'jal', 'L1'), \
-                      bar]
+                   S('command', 'jal', 'L1'),
+                   self.bar])
+
+    def test_optimize_block_instr_mov_jal_false(self):
+        arguments = [self.foo, \
+                     S('command', 'addu', '$regA', '$regC', 2), \
+                     S('command', 'move', '$3', '$regA'), \
+                     S('command', 'jal', 'L1'), \
+                     self.bar]
         block = B(arguments)
-        optimize_global(block)
-        
+        optimize_block(block)
+
         self.assertEquals(block.statements, arguments)
-        
-    def test_optimize_global_sw_ld_true(self):
-        foo = S('command', 'foo')
-        bar = S('command', 'bar')
-        
-        block = B([foo,
+
+    def test_optimize_block_sw_ld_true(self):
+        block = B([self.foo,
                    S('command', 'sw', '$regA', '$regB'),
-                   S('command', 'ld', '$regA', '$regC'),
-                   bar])
-        optimize_global(block)
-        
-        self.assertEquals(block.statements, [foo,
+                   S('command', 'lw', '$regA', '$regB'),
+                   self.bar])
+        optimize_block(block)
+
+        self.assertEquals(block.statements, [self.foo,
                    S('command', 'sw', '$regA', '$regB'),
-                   bar])
-                   
-    def test_optimize_global_sw_ld_false(self):
-        foo = S('command', 'foo')
-        bar = S('command', 'bar')
-        
-        arguments = [foo, \
+                   self.bar])
+
+    def test_optimize_block_sw_ld_false(self):
+        arguments = [self.foo, \
                      S('command', 'sw', '$regA', '$regB'), \
-                     S('command', 'ld', '$regD', '$regC'), \
-                     bar]
+                     S('command', 'lw', '$regD', '$regC'), \
+                     self.bar]
         block = B(arguments)
-        optimize_global(block)
-        
+        optimize_block(block)
+
         self.assertEquals(block.statements, arguments)
 
-    def test_optimize_global_shift_true(self):
-        foo = S('command', 'foo')
-        bar = S('command', 'bar')
-        
-        block = B([foo,
+    def test_optimize_block_shift_true(self):
+        block = B([self.foo,
                    S('command', 'sll', '$regA', '$regA', 0),
-                   bar])
-        optimize_global(block)
-        
-        self.assertEquals(block.statements, [foo,
-                   bar])
-                   
-    def test_optimize_global_shift_false(self):
-        foo = S('command', 'foo')
-        bar = S('command', 'bar')
-        
-        arguments = [foo, \
+                   self.bar])
+        optimize_block(block)
+
+        self.assertEquals(block.statements, [self.foo, self.bar])
+
+    def test_optimize_block_shift_false(self):
+        arguments = [self.foo, \
                      S('command', 'sll', '$regA', '$regB', 0), \
-                     bar]
+                     self.bar]
         block = B(arguments)
-        optimize_global(block)
-        
+        optimize_block(block)
+
         self.assertEquals(block.statements, arguments)
-        
-        arguments2 = [foo, \
+
+        arguments2 = [self.foo, \
                      S('command', 'sll', '$regA', '$regA', 1), \
-                     bar]
+                     self.bar]
         block2 = B(arguments2)
-        optimize_global(block2)
-        
+        optimize_block(block2)
+
         self.assertEquals(block2.statements, arguments2)
-    
-    def test_optimize_global_add_lw_true(self):
-        foo = S('command', 'foo')
-        bar = S('command', 'bar')
-        
-        block = B([foo,
-                   S('command', 'add', '$regA', '$regA', 10),
+
+    def test_optimize_block_add_lw_true(self):
+        block = B([self.foo,
+                   S('command', 'addu', '$regA', '$regA', 10),
                    S('command', 'lw', '$regB', '0($regA)'),
-                   bar])
-        optimize_global(block)
-        
-        self.assertEquals(block.statements, [foo,
+                   self.bar])
+        optimize_block(block)
+
+        self.assertEquals(block.statements, [self.foo,
                    S('command', 'lw', '$regB', '10($regA)'),
-                   bar])
-                   
-    def test_optimize_global_add_lw_false(self):
-        foo = S('command', 'foo')
-        bar = S('command', 'bar')
-        
-        arguments = [foo, \
-                     S('command', 'add', '$regA', '$regA', 10), \
+                   self.bar])
+
+    def test_optimize_block_add_lw_false(self):
+        arguments = [self.foo, \
+                     S('command', 'addu', '$regA', '$regA', 10), \
                      S('command', 'lw', '$regB', '0($regC)'), \
-                     bar]
+                     self.bar]
         block = B(arguments)
-        optimize_global(block)
-        
-        arguments2 = [foo, \
-                     S('command', 'add', '$regA', '$regB', 10), \
+        optimize_block(block)
+
+        arguments2 = [self.foo, \
+                     S('command', 'addu', '$regA', '$regB', 10), \
                      S('command', 'lw', '$regB', '0($regA)'), \
-                     bar]
+                     self.bar]
         block2 = B(arguments2)
-        
-        arguments3 = [foo, \
-                     S('command', 'add', '$regA', '$regA', 10), \
+
+        arguments3 = [self.foo, \
+                     S('command', 'addu', '$regA', '$regA', 10), \
                      S('command', 'lw', '$regB', '1($regA)'), \
-                     bar]
+                     self.bar]
         block3 = B(arguments3)
-        optimize_global(block3)
-        
+        optimize_block(block3)
+
         self.assertEquals(block.statements, arguments)
         self.assertEquals(block2.statements, arguments2)
         self.assertEquals(block3.statements, arguments3)
-        
-    #     beq ..., $Lx          ->      bne ..., $Ly
-    #     j $Ly                     $Lx:
-    # $Lx:
+
     def test_optimize_global_beq_j_true(self):
-        foo = S('command', 'foo')
-        bar = S('command', 'bar')
-        
-        block = B([foo,
+        block = B([self.foo,
                    S('command', 'beq', '$regA', '$regB', '$Lx'),
                    S('command', 'j', '$Ly'),
                    S('label', '$Lx'),
-                   bar])
+                   self.bar])
         optimize_global(block)
-        
-        self.assertEquals(block.statements, [foo,
+
+        self.assertEquals(block.statements, [self.foo,
                    S('command', 'bne', '$regA', '$regB', '$Ly'),
                    S('label', '$Lx'),
-                   bar])
-                   
+                   self.bar])
+
     def test_optimize_global_beq_j_false(self):
-        foo = S('command', 'foo')
-        bar = S('command', 'bar')
-        
-        arguments = [foo, \
+        arguments = [self.foo, \
                      S('command', 'beq', '$regA', '$regB', '$Lz'), \
                      S('command', 'j', '$Ly'), \
                      S('label', '$Lx'), \
-                     bar]
+                     self.bar]
         block = B(arguments)
         optimize_global(block)
-        
+
+        self.assertEquals(block.statements, arguments)
+
+    def test_optimize_global_bne_j_true(self):
+        block = B([self.foo,
+                   S('command', 'bne', '$regA', '$regB', '$Lx'),
+                   S('command', 'j', '$Ly'),
+                   S('label', '$Lx'),
+                   self.bar])
+        optimize_global(block)
+
+        self.assertEquals(block.statements, [self.foo,
+                   S('command', 'beq', '$regA', '$regB', '$Ly'),
+                   S('label', '$Lx'),
+                   self.bar])
+
+    def test_optimize_global_bne_j_false(self):
+        arguments = [self.foo, \
+                     S('command', 'bne', '$regA', '$regB', '$Lz'), \
+                     S('command', 'j', '$Ly'), \
+                     S('label', '$Lx'), \
+                     self.bar]
+        block = B(arguments)
+        optimize_global(block)
+
+        self.assertEquals(block.statements, arguments)
+
+    def test_optimize_block_move_move_true(self):
+        block = B([self.foo,
+                   S('command', 'move', '$regA', '$regB'),
+                   S('command', 'move', '$regB', '$regA'),
+                   self.bar])
+        optimize_block(block)
+
+        self.assertEquals(block.statements, [self.foo,
+                   S('command', 'move', '$regA', '$regB'),
+                   self.bar])
+
+    def test_optimize_block_mov_mov_false(self):
+        arguments = [self.foo, \
+                     S('command', 'move', '$regA', '$regB'), \
+                     S('command', 'move', '$regB', '$regC'), \
+                     self.bar]
+        block = B(arguments)
+        optimize_block(block)
+
         self.assertEquals(block.statements, arguments)

tests/test_optimize_advanced.py

@@ -0,0 +1,43 @@
+import unittest
+
+from src.optimize.advanced import eliminate_common_subexpressions, \
+        fold_constants, copy_propagation
+from src.statement import Statement as S, Block as B
+
+
+class TestOptimizeAdvanced(unittest.TestCase):
+
+    def setUp(self):
+        self.foo = S('command', 'foo')
+        self.bar = S('command', 'bar')
+
+    def test_eliminate_common_subexpressions(self):
+        pass
+        
+    def test_copy_propagation_true(self):        
+        print "testing true"
+        block = B([self.foo,
+                   S('command', 'move', '$1', '$2'),
+                   self.foo,
+                   S('command', 'addu', '$3', '$1', '$4'),
+                   self.bar])
+                   
+        copy_propagation(block)
+        self.assertEqual(block.statements, [self.foo,
+                   S('command', 'move', '$1', '$2'),
+                   self.foo,
+                   S('command', 'addu', '$3', '$2', '$4'),
+                   self.bar])
+        print "Test true succesfull"
+                   
+#    def test_copy_propagation_false(self):
+#        print "Testing false"
+#        arguments = [self.foo,
+#                   S('command', 'move', '$1', '$2'),
+#                   S('command', 'move', '$10', '$20'),
+#                   S('command', 'addu', '$1', '$5', 1),
+#                   S('command', 'addu', '$3', '$1', '$4'),
+#                   self.bar]
+#        block = B(arguments)
+#        copy_propagation(block)
+#        self.assertEqual(block.statements, arguments)