@ void func(uint8_t *block, const uint8_t *pixels, int line_size, int h)
@ block = word aligned, pixles = unaligned
pld [r1]
- stmfd sp!, {r4-r11, lr} @ R14 is also called LR
+ push {r4-r11, lr}
JMP_ALIGN r5, r1
1:
- ldmia r1, {r4-r7}
+ ldm r1, {r4-r7}
add r1, r1, r2
- stmia r0, {r4-r7}
+ stm r0, {r4-r7}
pld [r1]
subs r3, r3, #1
add r0, r0, r2
bne 1b
- ldmfd sp!, {r4-r11, pc}
+ pop {r4-r11, pc}
.align 5
2:
- ldmia r1, {r4-r8}
+ ldm r1, {r4-r8}
add r1, r1, r2
ADJ_ALIGN_QUADWORD_D 1, r9, r10, r11, r12, r4, r5, r6, r7, r8
pld [r1]
subs r3, r3, #1
- stmia r0, {r9-r12}
+ stm r0, {r9-r12}
add r0, r0, r2
bne 2b
- ldmfd sp!, {r4-r11, pc}
+ pop {r4-r11, pc}
.align 5
3:
- ldmia r1, {r4-r8}
+ ldm r1, {r4-r8}
add r1, r1, r2
ADJ_ALIGN_QUADWORD_D 2, r9, r10, r11, r12, r4, r5, r6, r7, r8
pld [r1]
subs r3, r3, #1
- stmia r0, {r9-r12}
+ stm r0, {r9-r12}
add r0, r0, r2
bne 3b
- ldmfd sp!, {r4-r11, pc}
+ pop {r4-r11, pc}
.align 5
4:
- ldmia r1, {r4-r8}
+ ldm r1, {r4-r8}
add r1, r1, r2
ADJ_ALIGN_QUADWORD_D 3, r9, r10, r11, r12, r4, r5, r6, r7, r8
pld [r1]
subs r3, r3, #1
- stmia r0, {r9-r12}
+ stm r0, {r9-r12}
add r0, r0, r2
bne 4b
- ldmfd sp!, {r4-r11,pc}
+ pop {r4-r11,pc}
.endfunc
@ ----------------------------------------------------------------
@ void func(uint8_t *block, const uint8_t *pixels, int line_size, int h)
@ block = word aligned, pixles = unaligned
pld [r1]
- stmfd sp!, {r4-r5,lr} @ R14 is also called LR
+ push {r4-r5,lr}
JMP_ALIGN r5, r1
1:
- ldmia r1, {r4-r5}
+ ldm r1, {r4-r5}
add r1, r1, r2
subs r3, r3, #1
pld [r1]
- stmia r0, {r4-r5}
+ stm r0, {r4-r5}
add r0, r0, r2
bne 1b
- ldmfd sp!, {r4-r5,pc}
+ pop {r4-r5,pc}
.align 5
2:
- ldmia r1, {r4-r5, r12}
+ ldm r1, {r4-r5, r12}
add r1, r1, r2
ADJ_ALIGN_DOUBLEWORD 1, r4, r5, r12
pld [r1]
subs r3, r3, #1
- stmia r0, {r4-r5}
+ stm r0, {r4-r5}
add r0, r0, r2
bne 2b
- ldmfd sp!, {r4-r5,pc}
+ pop {r4-r5,pc}
.align 5
3:
- ldmia r1, {r4-r5, r12}
+ ldm r1, {r4-r5, r12}
add r1, r1, r2
ADJ_ALIGN_DOUBLEWORD 2, r4, r5, r12
pld [r1]
subs r3, r3, #1
- stmia r0, {r4-r5}
+ stm r0, {r4-r5}
add r0, r0, r2
bne 3b
- ldmfd sp!, {r4-r5,pc}
+ pop {r4-r5,pc}
.align 5
4:
- ldmia r1, {r4-r5, r12}
+ ldm r1, {r4-r5, r12}
add r1, r1, r2
ADJ_ALIGN_DOUBLEWORD 3, r4, r5, r12
pld [r1]
subs r3, r3, #1
- stmia r0, {r4-r5}
+ stm r0, {r4-r5}
add r0, r0, r2
bne 4b
- ldmfd sp!, {r4-r5,pc}
+ pop {r4-r5,pc}
.endfunc
@ ----------------------------------------------------------------
@ void func(uint8_t *block, const uint8_t *pixels, int line_size, int h)
@ block = word aligned, pixles = unaligned
pld [r1]
- stmfd sp!, {r4-r10,lr} @ R14 is also called LR
+ push {r4-r10,lr}
ldr r12, =0xfefefefe
JMP_ALIGN r5, r1
1:
- ldmia r1, {r4-r5, r10}
+ ldm r1, {r4-r5, r10}
add r1, r1, r2
ADJ_ALIGN_DOUBLEWORD_D 1, r6, r7, r4, r5, r10
pld [r1]
RND_AVG32 r8, r9, r4, r5, r6, r7, r12
subs r3, r3, #1
- stmia r0, {r8-r9}
+ stm r0, {r8-r9}
add r0, r0, r2
bne 1b
- ldmfd sp!, {r4-r10,pc}
+ pop {r4-r10,pc}
.align 5
2:
- ldmia r1, {r4-r5, r10}
+ ldm r1, {r4-r5, r10}
add r1, r1, r2
ADJ_ALIGN_DOUBLEWORD_D 1, r6, r7, r4, r5, r10
ADJ_ALIGN_DOUBLEWORD_D 2, r8, r9, r4, r5, r10
pld [r1]
RND_AVG32 r4, r5, r6, r7, r8, r9, r12
subs r3, r3, #1
- stmia r0, {r4-r5}
+ stm r0, {r4-r5}
add r0, r0, r2
bne 2b
- ldmfd sp!, {r4-r10,pc}
+ pop {r4-r10,pc}
.align 5
3:
- ldmia r1, {r4-r5, r10}
+ ldm r1, {r4-r5, r10}
add r1, r1, r2
ADJ_ALIGN_DOUBLEWORD_D 2, r6, r7, r4, r5, r10
ADJ_ALIGN_DOUBLEWORD_D 3, r8, r9, r4, r5, r10
pld [r1]
RND_AVG32 r4, r5, r6, r7, r8, r9, r12
subs r3, r3, #1
- stmia r0, {r4-r5}
+ stm r0, {r4-r5}
add r0, r0, r2
bne 3b
- ldmfd sp!, {r4-r10,pc}
+ pop {r4-r10,pc}
.align 5
4:
- ldmia r1, {r4-r5, r10}
+ ldm r1, {r4-r5, r10}
add r1, r1, r2
ADJ_ALIGN_DOUBLEWORD_D 3, r6, r7, r4, r5, r10
pld [r1]
RND_AVG32 r8, r9, r6, r7, r5, r10, r12
subs r3, r3, #1
- stmia r0, {r8-r9}
+ stm r0, {r8-r9}
add r0, r0, r2
bne 4b
- ldmfd sp!, {r4-r10,pc} @@ update PC with LR content.
+ pop {r4-r10,pc}
.endfunc
.align 5
@ void func(uint8_t *block, const uint8_t *pixels, int line_size, int h)
@ block = word aligned, pixles = unaligned
pld [r1]
- stmfd sp!, {r4-r10,lr} @ R14 is also called LR
+ push {r4-r10,lr}
ldr r12, =0xfefefefe
JMP_ALIGN r5, r1
1:
- ldmia r1, {r4-r5, r10}
+ ldm r1, {r4-r5, r10}
add r1, r1, r2
ADJ_ALIGN_DOUBLEWORD_D 1, r6, r7, r4, r5, r10
pld [r1]
NO_RND_AVG32 r8, r9, r4, r5, r6, r7, r12
subs r3, r3, #1
- stmia r0, {r8-r9}
+ stm r0, {r8-r9}
add r0, r0, r2
bne 1b
- ldmfd sp!, {r4-r10,pc}
+ pop {r4-r10,pc}
.align 5
2:
- ldmia r1, {r4-r5, r10}
+ ldm r1, {r4-r5, r10}
add r1, r1, r2
ADJ_ALIGN_DOUBLEWORD_D 1, r6, r7, r4, r5, r10
ADJ_ALIGN_DOUBLEWORD_D 2, r8, r9, r4, r5, r10
pld [r1]
NO_RND_AVG32 r4, r5, r6, r7, r8, r9, r12
subs r3, r3, #1
- stmia r0, {r4-r5}
+ stm r0, {r4-r5}
add r0, r0, r2
bne 2b
- ldmfd sp!, {r4-r10,pc}
+ pop {r4-r10,pc}
.align 5
3:
- ldmia r1, {r4-r5, r10}
+ ldm r1, {r4-r5, r10}
add r1, r1, r2
ADJ_ALIGN_DOUBLEWORD_D 2, r6, r7, r4, r5, r10
ADJ_ALIGN_DOUBLEWORD_D 3, r8, r9, r4, r5, r10
pld [r1]
NO_RND_AVG32 r4, r5, r6, r7, r8, r9, r12
subs r3, r3, #1
- stmia r0, {r4-r5}
+ stm r0, {r4-r5}
add r0, r0, r2
bne 3b
- ldmfd sp!, {r4-r10,pc}
+ pop {r4-r10,pc}
.align 5
4:
- ldmia r1, {r4-r5, r10}
+ ldm r1, {r4-r5, r10}
add r1, r1, r2
ADJ_ALIGN_DOUBLEWORD_D 3, r6, r7, r4, r5, r10
pld [r1]
NO_RND_AVG32 r8, r9, r6, r7, r5, r10, r12
subs r3, r3, #1
- stmia r0, {r8-r9}
+ stm r0, {r8-r9}
add r0, r0, r2
bne 4b
- ldmfd sp!, {r4-r10,pc} @@ update PC with LR content.
+ pop {r4-r10,pc}
.endfunc
@ void func(uint8_t *block, const uint8_t *pixels, int line_size, int h)
@ block = word aligned, pixles = unaligned
pld [r1]
- stmfd sp!, {r4-r11,lr} @ R14 is also called LR
+ push {r4-r11,lr}
mov r3, r3, lsr #1
ldr r12, =0xfefefefe
JMP_ALIGN r5, r1
1:
- ldmia r1, {r4-r5}
+ ldm r1, {r4-r5}
add r1, r1, r2
-6: ldmia r1, {r6-r7}
+6: ldm r1, {r6-r7}
add r1, r1, r2
pld [r1]
RND_AVG32 r8, r9, r4, r5, r6, r7, r12
- ldmia r1, {r4-r5}
+ ldm r1, {r4-r5}
add r1, r1, r2
- stmia r0, {r8-r9}
+ stm r0, {r8-r9}
add r0, r0, r2
pld [r1]
RND_AVG32 r8, r9, r6, r7, r4, r5, r12
subs r3, r3, #1
- stmia r0, {r8-r9}
+ stm r0, {r8-r9}
add r0, r0, r2
bne 6b
- ldmfd sp!, {r4-r11,pc}
+ pop {r4-r11,pc}
.align 5
2:
- ldmia r1, {r4-r6}
+ ldm r1, {r4-r6}
add r1, r1, r2
pld [r1]
ADJ_ALIGN_DOUBLEWORD 1, r4, r5, r6
-6: ldmia r1, {r7-r9}
+6: ldm r1, {r7-r9}
add r1, r1, r2
pld [r1]
ADJ_ALIGN_DOUBLEWORD 1, r7, r8, r9
RND_AVG32 r10, r11, r4, r5, r7, r8, r12
- stmia r0, {r10-r11}
+ stm r0, {r10-r11}
add r0, r0, r2
- ldmia r1, {r4-r6}
+ ldm r1, {r4-r6}
add r1, r1, r2
pld [r1]
ADJ_ALIGN_DOUBLEWORD 1, r4, r5, r6
subs r3, r3, #1
RND_AVG32 r10, r11, r7, r8, r4, r5, r12
- stmia r0, {r10-r11}
+ stm r0, {r10-r11}
add r0, r0, r2
bne 6b
- ldmfd sp!, {r4-r11,pc}
+ pop {r4-r11,pc}
.align 5
3:
- ldmia r1, {r4-r6}
+ ldm r1, {r4-r6}
add r1, r1, r2
pld [r1]
ADJ_ALIGN_DOUBLEWORD 2, r4, r5, r6
-6: ldmia r1, {r7-r9}
+6: ldm r1, {r7-r9}
add r1, r1, r2
pld [r1]
ADJ_ALIGN_DOUBLEWORD 2, r7, r8, r9
RND_AVG32 r10, r11, r4, r5, r7, r8, r12
- stmia r0, {r10-r11}
+ stm r0, {r10-r11}
add r0, r0, r2
- ldmia r1, {r4-r6}
+ ldm r1, {r4-r6}
add r1, r1, r2
pld [r1]
ADJ_ALIGN_DOUBLEWORD 2, r4, r5, r6
subs r3, r3, #1
RND_AVG32 r10, r11, r7, r8, r4, r5, r12
- stmia r0, {r10-r11}
+ stm r0, {r10-r11}
add r0, r0, r2
bne 6b
- ldmfd sp!, {r4-r11,pc}
+ pop {r4-r11,pc}
.align 5
4:
- ldmia r1, {r4-r6}
+ ldm r1, {r4-r6}
add r1, r1, r2
pld [r1]
ADJ_ALIGN_DOUBLEWORD 3, r4, r5, r6
-6: ldmia r1, {r7-r9}
+6: ldm r1, {r7-r9}
add r1, r1, r2
pld [r1]
ADJ_ALIGN_DOUBLEWORD 3, r7, r8, r9
RND_AVG32 r10, r11, r4, r5, r7, r8, r12
- stmia r0, {r10-r11}
+ stm r0, {r10-r11}
add r0, r0, r2
- ldmia r1, {r4-r6}
+ ldm r1, {r4-r6}
add r1, r1, r2
pld [r1]
ADJ_ALIGN_DOUBLEWORD 3, r4, r5, r6
subs r3, r3, #1
RND_AVG32 r10, r11, r7, r8, r4, r5, r12
- stmia r0, {r10-r11}
+ stm r0, {r10-r11}
add r0, r0, r2
bne 6b
- ldmfd sp!, {r4-r11,pc}
+ pop {r4-r11,pc}
.endfunc
.align 5
@ void func(uint8_t *block, const uint8_t *pixels, int line_size, int h)
@ block = word aligned, pixles = unaligned
pld [r1]
- stmfd sp!, {r4-r11,lr} @ R14 is also called LR
+ push {r4-r11,lr}
mov r3, r3, lsr #1
ldr r12, =0xfefefefe
JMP_ALIGN r5, r1
1:
- ldmia r1, {r4-r5}
+ ldm r1, {r4-r5}
add r1, r1, r2
-6: ldmia r1, {r6-r7}
+6: ldm r1, {r6-r7}
add r1, r1, r2
pld [r1]
NO_RND_AVG32 r8, r9, r4, r5, r6, r7, r12
- ldmia r1, {r4-r5}
+ ldm r1, {r4-r5}
add r1, r1, r2
- stmia r0, {r8-r9}
+ stm r0, {r8-r9}
add r0, r0, r2
pld [r1]
NO_RND_AVG32 r8, r9, r6, r7, r4, r5, r12
subs r3, r3, #1
- stmia r0, {r8-r9}
+ stm r0, {r8-r9}
add r0, r0, r2
bne 6b
- ldmfd sp!, {r4-r11,pc}
+ pop {r4-r11,pc}
.align 5
2:
- ldmia r1, {r4-r6}
+ ldm r1, {r4-r6}
add r1, r1, r2
pld [r1]
ADJ_ALIGN_DOUBLEWORD 1, r4, r5, r6
-6: ldmia r1, {r7-r9}
+6: ldm r1, {r7-r9}
add r1, r1, r2
pld [r1]
ADJ_ALIGN_DOUBLEWORD 1, r7, r8, r9
NO_RND_AVG32 r10, r11, r4, r5, r7, r8, r12
- stmia r0, {r10-r11}
+ stm r0, {r10-r11}
add r0, r0, r2
- ldmia r1, {r4-r6}
+ ldm r1, {r4-r6}
add r1, r1, r2
pld [r1]
ADJ_ALIGN_DOUBLEWORD 1, r4, r5, r6
subs r3, r3, #1
NO_RND_AVG32 r10, r11, r7, r8, r4, r5, r12
- stmia r0, {r10-r11}
+ stm r0, {r10-r11}
add r0, r0, r2
bne 6b
- ldmfd sp!, {r4-r11,pc}
+ pop {r4-r11,pc}
.align 5
3:
- ldmia r1, {r4-r6}
+ ldm r1, {r4-r6}
add r1, r1, r2
pld [r1]
ADJ_ALIGN_DOUBLEWORD 2, r4, r5, r6
-6: ldmia r1, {r7-r9}
+6: ldm r1, {r7-r9}
add r1, r1, r2
pld [r1]
ADJ_ALIGN_DOUBLEWORD 2, r7, r8, r9
NO_RND_AVG32 r10, r11, r4, r5, r7, r8, r12
- stmia r0, {r10-r11}
+ stm r0, {r10-r11}
add r0, r0, r2
- ldmia r1, {r4-r6}
+ ldm r1, {r4-r6}
add r1, r1, r2
pld [r1]
ADJ_ALIGN_DOUBLEWORD 2, r4, r5, r6
subs r3, r3, #1
NO_RND_AVG32 r10, r11, r7, r8, r4, r5, r12
- stmia r0, {r10-r11}
+ stm r0, {r10-r11}
add r0, r0, r2
bne 6b
- ldmfd sp!, {r4-r11,pc}
+ pop {r4-r11,pc}
.align 5
4:
- ldmia r1, {r4-r6}
+ ldm r1, {r4-r6}
add r1, r1, r2
pld [r1]
ADJ_ALIGN_DOUBLEWORD 3, r4, r5, r6
-6: ldmia r1, {r7-r9}
+6: ldm r1, {r7-r9}
add r1, r1, r2
pld [r1]
ADJ_ALIGN_DOUBLEWORD 3, r7, r8, r9
NO_RND_AVG32 r10, r11, r4, r5, r7, r8, r12
- stmia r0, {r10-r11}
+ stm r0, {r10-r11}
add r0, r0, r2
- ldmia r1, {r4-r6}
+ ldm r1, {r4-r6}
add r1, r1, r2
pld [r1]
ADJ_ALIGN_DOUBLEWORD 3, r4, r5, r6
subs r3, r3, #1
NO_RND_AVG32 r10, r11, r7, r8, r4, r5, r12
- stmia r0, {r10-r11}
+ stm r0, {r10-r11}
add r0, r0, r2
bne 6b
- ldmfd sp!, {r4-r11,pc}
+ pop {r4-r11,pc}
.endfunc
.ltorg
@ l1= (a & 0x03030303) + (b & 0x03030303) ?(+ 0x02020202)
@ h1= ((a & 0xFCFCFCFCUL) >> 2) + ((b & 0xFCFCFCFCUL) >> 2)
.if \align == 0
- ldmia r1, {r6-r8}
+ ldm r1, {r6-r8}
.elseif \align == 3
- ldmia r1, {r5-r7}
+ ldm r1, {r5-r7}
.else
- ldmia r1, {r8-r10}
+ ldm r1, {r8-r10}
.endif
add r1, r1, r2
pld [r1]
.macro RND_XY2_EXPAND align, rnd
RND_XY2_IT \align, \rnd
-6: stmfd sp!, {r8-r11}
+6: push {r8-r11}
RND_XY2_IT \align, \rnd
- ldmfd sp!, {r4-r7}
+ pop {r4-r7}
add r4, r4, r8
add r5, r5, r9
ldr r14, =0x0f0f0f0f
and r5, r14, r5, lsr #2
add r4, r4, r6
add r5, r5, r7
- stmia r0, {r4-r5}
+ stm r0, {r4-r5}
add r0, r0, r2
bge 6b
- ldmfd sp!, {r4-r11,pc}
+ pop {r4-r11,pc}
.endm
.align 5
@ void func(uint8_t *block, const uint8_t *pixels, int line_size, int h)
@ block = word aligned, pixles = unaligned
pld [r1]
- stmfd sp!, {r4-r11,lr} @ R14 is also called LR
+ push {r4-r11,lr} @ R14 is also called LR
JMP_ALIGN r5, r1
1:
RND_XY2_EXPAND 0, lsl
@ void func(uint8_t *block, const uint8_t *pixels, int line_size, int h)
@ block = word aligned, pixles = unaligned
pld [r1]
- stmfd sp!, {r4-r11,lr} @ R14 is also called LR
+ push {r4-r11,lr}
JMP_ALIGN r5, r1
1:
RND_XY2_EXPAND 0, lsr