ARM: h264dsp_neon cosmetics

- Replace 'ip' with 'r12'.
- Use correct size designators for vld1/vst1.
- Whitespace fixes.

Signed-off-by: Mans Rullgard <mans@mansr.com>
This commit is contained in:
Mans Rullgard 2011-12-02 16:45:00 +00:00
parent a760f530bb
commit 59807fee6d
1 changed files with 313 additions and 313 deletions

View File

@ -36,7 +36,7 @@ A muls r7, r4, r5
T mul r7, r4, r5
T cmp r7, #0
rsb r6, r7, r5, lsl #3
rsb ip, r7, r4, lsl #3
rsb r12, r7, r4, lsl #3
sub r4, r7, r4, lsl #3
sub r4, r4, r5, lsl #3
add r4, r4, #64
@ -47,10 +47,10 @@ T cmp r7, #0
vdup.8 d0, r4
lsl r4, r2, #1
vdup.8 d1, ip
vld1.64 {d4, d5}, [r1], r4
vdup.8 d1, r12
vld1.8 {d4, d5}, [r1], r4
vdup.8 d2, r6
vld1.64 {d6, d7}, [r5], r4
vld1.8 {d6, d7}, [r5], r4
vdup.8 d3, r7
vext.8 d5, d4, d5, #1
@ -59,7 +59,7 @@ T cmp r7, #0
1: pld [r5]
vmull.u8 q8, d4, d0
vmlal.u8 q8, d5, d1
vld1.64 {d4, d5}, [r1], r4
vld1.8 {d4, d5}, [r1], r4
vmlal.u8 q8, d6, d2
vext.8 d5, d4, d5, #1
vmlal.u8 q8, d7, d3
@ -69,57 +69,57 @@ T cmp r7, #0
vmlal.u8 q9, d4, d2
vmlal.u8 q9, d5, d3
vrshrn.u16 d16, q8, #6
vld1.64 {d6, d7}, [r5], r4
vld1.8 {d6, d7}, [r5], r4
pld [r1]
vrshrn.u16 d17, q9, #6
.ifc \type,avg
vld1.64 {d20}, [lr,:64], r2
vld1.64 {d21}, [lr,:64], r2
vld1.8 {d20}, [lr,:64], r2
vld1.8 {d21}, [lr,:64], r2
vrhadd.u8 q8, q8, q10
.endif
vext.8 d7, d6, d7, #1
vst1.64 {d16}, [r0,:64], r2
vst1.64 {d17}, [r0,:64], r2
vst1.8 {d16}, [r0,:64], r2
vst1.8 {d17}, [r0,:64], r2
bgt 1b
pop {r4-r7, pc}
2: tst r6, r6
add ip, ip, r6
add r12, r12, r6
vdup.8 d0, r4
vdup.8 d1, ip
vdup.8 d1, r12
beq 4f
add r5, r1, r2
lsl r4, r2, #1
vld1.64 {d4}, [r1], r4
vld1.64 {d6}, [r5], r4
vld1.8 {d4}, [r1], r4
vld1.8 {d6}, [r5], r4
3: pld [r5]
vmull.u8 q8, d4, d0
vmlal.u8 q8, d6, d1
vld1.64 {d4}, [r1], r4
vld1.8 {d4}, [r1], r4
vmull.u8 q9, d6, d0
vmlal.u8 q9, d4, d1
vld1.64 {d6}, [r5], r4
vld1.8 {d6}, [r5], r4
vrshrn.u16 d16, q8, #6
vrshrn.u16 d17, q9, #6
.ifc \type,avg
vld1.64 {d20}, [lr,:64], r2
vld1.64 {d21}, [lr,:64], r2
vld1.8 {d20}, [lr,:64], r2
vld1.8 {d21}, [lr,:64], r2
vrhadd.u8 q8, q8, q10
.endif
subs r3, r3, #2
pld [r1]
vst1.64 {d16}, [r0,:64], r2
vst1.64 {d17}, [r0,:64], r2
vst1.8 {d16}, [r0,:64], r2
vst1.8 {d17}, [r0,:64], r2
bgt 3b
pop {r4-r7, pc}
4: vld1.64 {d4, d5}, [r1], r2
vld1.64 {d6, d7}, [r1], r2
4: vld1.8 {d4, d5}, [r1], r2
vld1.8 {d6, d7}, [r1], r2
vext.8 d5, d4, d5, #1
vext.8 d7, d6, d7, #1
@ -127,7 +127,7 @@ T cmp r7, #0
subs r3, r3, #2
vmull.u8 q8, d4, d0
vmlal.u8 q8, d5, d1
vld1.64 {d4, d5}, [r1], r2
vld1.8 {d4, d5}, [r1], r2
vmull.u8 q9, d6, d0
vmlal.u8 q9, d7, d1
pld [r1]
@ -135,14 +135,14 @@ T cmp r7, #0
vrshrn.u16 d16, q8, #6
vrshrn.u16 d17, q9, #6
.ifc \type,avg
vld1.64 {d20}, [lr,:64], r2
vld1.64 {d21}, [lr,:64], r2
vld1.8 {d20}, [lr,:64], r2
vld1.8 {d21}, [lr,:64], r2
vrhadd.u8 q8, q8, q10
.endif
vld1.64 {d6, d7}, [r1], r2
vld1.8 {d6, d7}, [r1], r2
vext.8 d7, d6, d7, #1
vst1.64 {d16}, [r0,:64], r2
vst1.64 {d17}, [r0,:64], r2
vst1.8 {d16}, [r0,:64], r2
vst1.8 {d17}, [r0,:64], r2
bgt 5b
pop {r4-r7, pc}
@ -164,7 +164,7 @@ A muls r7, r4, r5
T mul r7, r4, r5
T cmp r7, #0
rsb r6, r7, r5, lsl #3
rsb ip, r7, r4, lsl #3
rsb r12, r7, r4, lsl #3
sub r4, r7, r4, lsl #3
sub r4, r4, r5, lsl #3
add r4, r4, #64
@ -175,10 +175,10 @@ T cmp r7, #0
vdup.8 d0, r4
lsl r4, r2, #1
vdup.8 d1, ip
vld1.64 {d4}, [r1], r4
vdup.8 d1, r12
vld1.8 {d4}, [r1], r4
vdup.8 d2, r6
vld1.64 {d6}, [r5], r4
vld1.8 {d6}, [r5], r4
vdup.8 d3, r7
vext.8 d5, d4, d5, #1
@ -192,12 +192,12 @@ T cmp r7, #0
1: pld [r5]
vmull.u8 q8, d4, d0
vmlal.u8 q8, d6, d2
vld1.64 {d4}, [r1], r4
vld1.8 {d4}, [r1], r4
vext.8 d5, d4, d5, #1
vtrn.32 d4, d5
vmull.u8 q9, d6, d0
vmlal.u8 q9, d4, d2
vld1.64 {d6}, [r5], r4
vld1.8 {d6}, [r5], r4
vadd.i16 d16, d16, d17
vadd.i16 d17, d18, d19
vrshrn.u16 d16, q8, #6
@ -217,9 +217,9 @@ T cmp r7, #0
pop {r4-r7, pc}
2: tst r6, r6
add ip, ip, r6
add r12, r12, r6
vdup.8 d0, r4
vdup.8 d1, ip
vdup.8 d1, r12
vtrn.32 d0, d1
beq 4f
@ -251,8 +251,8 @@ T cmp r7, #0
pop {r4-r7, pc}
4: vld1.64 {d4}, [r1], r2
vld1.64 {d6}, [r1], r2
4: vld1.8 {d4}, [r1], r2
vld1.8 {d6}, [r1], r2
vext.8 d5, d4, d5, #1
vext.8 d7, d6, d7, #1
vtrn.32 d4, d5
@ -261,7 +261,7 @@ T cmp r7, #0
5: vmull.u8 q8, d4, d0
vmull.u8 q9, d6, d0
subs r3, r3, #2
vld1.64 {d4}, [r1], r2
vld1.8 {d4}, [r1], r2
vext.8 d5, d4, d5, #1
vtrn.32 d4, d5
vadd.i16 d16, d16, d17
@ -273,7 +273,7 @@ T cmp r7, #0
vld1.32 {d20[1]}, [lr,:32], r2
vrhadd.u8 d16, d16, d20
.endif
vld1.64 {d6}, [r1], r2
vld1.8 {d6}, [r1], r2
vext.8 d7, d6, d7, #1
vtrn.32 d6, d7
pld [r1]
@ -363,16 +363,16 @@ endfunc
/* H.264 loop filter */
.macro h264_loop_filter_start
ldr ip, [sp]
ldr r12, [sp]
tst r2, r2
ldr ip, [ip]
ldr r12, [r12]
it ne
tstne r3, r3
vmov.32 d24[0], ip
and ip, ip, ip, lsl #16
vmov.32 d24[0], r12
and r12, r12, r12, lsl #16
it eq
bxeq lr
ands ip, ip, ip, lsl #8
ands r12, r12, r12, lsl #8
it lt
bxlt lr
.endm
@ -448,24 +448,24 @@ endfunc
function ff_h264_v_loop_filter_luma_neon, export=1
h264_loop_filter_start
vld1.64 {d0, d1}, [r0,:128], r1
vld1.64 {d2, d3}, [r0,:128], r1
vld1.64 {d4, d5}, [r0,:128], r1
vld1.8 {d0, d1}, [r0,:128], r1
vld1.8 {d2, d3}, [r0,:128], r1
vld1.8 {d4, d5}, [r0,:128], r1
sub r0, r0, r1, lsl #2
sub r0, r0, r1, lsl #1
vld1.64 {d20,d21}, [r0,:128], r1
vld1.64 {d18,d19}, [r0,:128], r1
vld1.64 {d16,d17}, [r0,:128], r1
vld1.8 {d20,d21}, [r0,:128], r1
vld1.8 {d18,d19}, [r0,:128], r1
vld1.8 {d16,d17}, [r0,:128], r1
vpush {d8-d15}
h264_loop_filter_luma
sub r0, r0, r1, lsl #1
vst1.64 {d8, d9}, [r0,:128], r1
vst1.64 {d16,d17}, [r0,:128], r1
vst1.64 {d0, d1}, [r0,:128], r1
vst1.64 {d10,d11}, [r0,:128]
vst1.8 {d8, d9}, [r0,:128], r1
vst1.8 {d16,d17}, [r0,:128], r1
vst1.8 {d0, d1}, [r0,:128], r1
vst1.8 {d10,d11}, [r0,:128]
vpop {d8-d15}
bx lr
@ -475,22 +475,22 @@ function ff_h264_h_loop_filter_luma_neon, export=1
h264_loop_filter_start
sub r0, r0, #4
vld1.64 {d6}, [r0], r1
vld1.64 {d20}, [r0], r1
vld1.64 {d18}, [r0], r1
vld1.64 {d16}, [r0], r1
vld1.64 {d0}, [r0], r1
vld1.64 {d2}, [r0], r1
vld1.64 {d4}, [r0], r1
vld1.64 {d26}, [r0], r1
vld1.64 {d7}, [r0], r1
vld1.64 {d21}, [r0], r1
vld1.64 {d19}, [r0], r1
vld1.64 {d17}, [r0], r1
vld1.64 {d1}, [r0], r1
vld1.64 {d3}, [r0], r1
vld1.64 {d5}, [r0], r1
vld1.64 {d27}, [r0], r1
vld1.8 {d6}, [r0], r1
vld1.8 {d20}, [r0], r1
vld1.8 {d18}, [r0], r1
vld1.8 {d16}, [r0], r1
vld1.8 {d0}, [r0], r1
vld1.8 {d2}, [r0], r1
vld1.8 {d4}, [r0], r1
vld1.8 {d26}, [r0], r1
vld1.8 {d7}, [r0], r1
vld1.8 {d21}, [r0], r1
vld1.8 {d19}, [r0], r1
vld1.8 {d17}, [r0], r1
vld1.8 {d1}, [r0], r1
vld1.8 {d3}, [r0], r1
vld1.8 {d5}, [r0], r1
vld1.8 {d27}, [r0], r1
transpose_8x8 q3, q10, q9, q8, q0, q1, q2, q13
@ -558,16 +558,16 @@ function ff_h264_v_loop_filter_chroma_neon, export=1
h264_loop_filter_start
sub r0, r0, r1, lsl #1
vld1.64 {d18}, [r0,:64], r1
vld1.64 {d16}, [r0,:64], r1
vld1.64 {d0}, [r0,:64], r1
vld1.64 {d2}, [r0,:64]
vld1.8 {d18}, [r0,:64], r1
vld1.8 {d16}, [r0,:64], r1
vld1.8 {d0}, [r0,:64], r1
vld1.8 {d2}, [r0,:64]
h264_loop_filter_chroma
sub r0, r0, r1, lsl #1
vst1.64 {d16}, [r0,:64], r1
vst1.64 {d0}, [r0,:64], r1
vst1.8 {d16}, [r0,:64], r1
vst1.8 {d0}, [r0,:64], r1
bx lr
endfunc
@ -715,12 +715,12 @@ endfunc
function put_h264_qpel16_h_lowpass_neon_packed
mov r4, lr
mov ip, #16
mov r12, #16
mov r3, #8
bl put_h264_qpel8_h_lowpass_neon
sub r1, r1, r2, lsl #4
add r1, r1, #8
mov ip, #16
mov r12, #16
mov lr, r4
b put_h264_qpel8_h_lowpass_neon
endfunc
@ -728,20 +728,20 @@ endfunc
.macro h264_qpel_h_lowpass type
function \type\()_h264_qpel16_h_lowpass_neon
push {lr}
mov ip, #16
mov r12, #16
bl \type\()_h264_qpel8_h_lowpass_neon
sub r0, r0, r3, lsl #4
sub r1, r1, r2, lsl #4
add r0, r0, #8
add r1, r1, #8
mov ip, #16
mov r12, #16
pop {lr}
endfunc
function \type\()_h264_qpel8_h_lowpass_neon
1: vld1.64 {d0, d1}, [r1], r2
vld1.64 {d16,d17}, [r1], r2
subs ip, ip, #2
1: vld1.8 {d0, d1}, [r1], r2
vld1.8 {d16,d17}, [r1], r2
subs r12, r12, #2
lowpass_8 d0, d1, d16, d17, d0, d16
.ifc \type,avg
vld1.8 {d2}, [r0,:64], r3
@ -750,8 +750,8 @@ function \type\()_h264_qpel8_h_lowpass_neon
vrhadd.u8 d16, d16, d3
sub r0, r0, r3
.endif
vst1.64 {d0}, [r0,:64], r3
vst1.64 {d16}, [r0,:64], r3
vst1.8 {d0}, [r0,:64], r3
vst1.8 {d16}, [r0,:64], r3
bne 1b
bx lr
endfunc
@ -763,7 +763,7 @@ endfunc
.macro h264_qpel_h_lowpass_l2 type
function \type\()_h264_qpel16_h_lowpass_l2_neon
push {lr}
mov ip, #16
mov r12, #16
bl \type\()_h264_qpel8_h_lowpass_l2_neon
sub r0, r0, r2, lsl #4
sub r1, r1, r2, lsl #4
@ -771,16 +771,16 @@ function \type\()_h264_qpel16_h_lowpass_l2_neon
add r0, r0, #8
add r1, r1, #8
add r3, r3, #8
mov ip, #16
mov r12, #16
pop {lr}
endfunc
function \type\()_h264_qpel8_h_lowpass_l2_neon
1: vld1.64 {d0, d1}, [r1], r2
vld1.64 {d16,d17}, [r1], r2
vld1.64 {d28}, [r3], r2
vld1.64 {d29}, [r3], r2
subs ip, ip, #2
1: vld1.8 {d0, d1}, [r1], r2
vld1.8 {d16,d17}, [r1], r2
vld1.8 {d28}, [r3], r2
vld1.8 {d29}, [r3], r2
subs r12, r12, #2
lowpass_8 d0, d1, d16, d17, d0, d1
vrhadd.u8 q0, q0, q14
.ifc \type,avg
@ -790,8 +790,8 @@ function \type\()_h264_qpel8_h_lowpass_l2_neon
vrhadd.u8 d1, d1, d3
sub r0, r0, r2
.endif
vst1.64 {d0}, [r0,:64], r2
vst1.64 {d1}, [r0,:64], r2
vst1.8 {d0}, [r0,:64], r2
vst1.8 {d1}, [r0,:64], r2
bne 1b
bx lr
endfunc
@ -832,19 +832,19 @@ function \type\()_h264_qpel16_v_lowpass_neon
endfunc
function \type\()_h264_qpel8_v_lowpass_neon
vld1.64 {d8}, [r1], r3
vld1.64 {d10}, [r1], r3
vld1.64 {d12}, [r1], r3
vld1.64 {d14}, [r1], r3
vld1.64 {d22}, [r1], r3
vld1.64 {d24}, [r1], r3
vld1.64 {d26}, [r1], r3
vld1.64 {d28}, [r1], r3
vld1.64 {d9}, [r1], r3
vld1.64 {d11}, [r1], r3
vld1.64 {d13}, [r1], r3
vld1.64 {d15}, [r1], r3
vld1.64 {d23}, [r1]
vld1.8 {d8}, [r1], r3
vld1.8 {d10}, [r1], r3
vld1.8 {d12}, [r1], r3
vld1.8 {d14}, [r1], r3
vld1.8 {d22}, [r1], r3
vld1.8 {d24}, [r1], r3
vld1.8 {d26}, [r1], r3
vld1.8 {d28}, [r1], r3
vld1.8 {d9}, [r1], r3
vld1.8 {d11}, [r1], r3
vld1.8 {d13}, [r1], r3
vld1.8 {d15}, [r1], r3
vld1.8 {d23}, [r1]
transpose_8x8 q4, q5, q6, q7, q11, q12, q13, q14
lowpass_8 d8, d9, d10, d11, d8, d10
@ -873,14 +873,14 @@ function \type\()_h264_qpel8_v_lowpass_neon
sub r0, r0, r2, lsl #3
.endif
vst1.64 {d8}, [r0,:64], r2
vst1.64 {d10}, [r0,:64], r2
vst1.64 {d12}, [r0,:64], r2
vst1.64 {d14}, [r0,:64], r2
vst1.64 {d22}, [r0,:64], r2
vst1.64 {d24}, [r0,:64], r2
vst1.64 {d26}, [r0,:64], r2
vst1.64 {d28}, [r0,:64], r2
vst1.8 {d8}, [r0,:64], r2
vst1.8 {d10}, [r0,:64], r2
vst1.8 {d12}, [r0,:64], r2
vst1.8 {d14}, [r0,:64], r2
vst1.8 {d22}, [r0,:64], r2
vst1.8 {d24}, [r0,:64], r2
vst1.8 {d26}, [r0,:64], r2
vst1.8 {d28}, [r0,:64], r2
bx lr
endfunc
@ -896,9 +896,9 @@ function \type\()_h264_qpel16_v_lowpass_l2_neon
sub r1, r1, r3, lsl #2
bl \type\()_h264_qpel8_v_lowpass_l2_neon
sub r0, r0, r3, lsl #4
sub ip, ip, r2, lsl #4
sub r12, r12, r2, lsl #4
add r0, r0, #8
add ip, ip, #8
add r12, r12, #8
sub r1, r1, r3, lsl #4
sub r1, r1, r3, lsl #2
add r1, r1, #8
@ -908,19 +908,19 @@ function \type\()_h264_qpel16_v_lowpass_l2_neon
endfunc
function \type\()_h264_qpel8_v_lowpass_l2_neon
vld1.64 {d8}, [r1], r3
vld1.64 {d10}, [r1], r3
vld1.64 {d12}, [r1], r3
vld1.64 {d14}, [r1], r3
vld1.64 {d22}, [r1], r3
vld1.64 {d24}, [r1], r3
vld1.64 {d26}, [r1], r3
vld1.64 {d28}, [r1], r3
vld1.64 {d9}, [r1], r3
vld1.64 {d11}, [r1], r3
vld1.64 {d13}, [r1], r3
vld1.64 {d15}, [r1], r3
vld1.64 {d23}, [r1]
vld1.8 {d8}, [r1], r3
vld1.8 {d10}, [r1], r3
vld1.8 {d12}, [r1], r3
vld1.8 {d14}, [r1], r3
vld1.8 {d22}, [r1], r3
vld1.8 {d24}, [r1], r3
vld1.8 {d26}, [r1], r3
vld1.8 {d28}, [r1], r3
vld1.8 {d9}, [r1], r3
vld1.8 {d11}, [r1], r3
vld1.8 {d13}, [r1], r3
vld1.8 {d15}, [r1], r3
vld1.8 {d23}, [r1]
transpose_8x8 q4, q5, q6, q7, q11, q12, q13, q14
lowpass_8 d8, d9, d10, d11, d8, d9
@ -929,17 +929,17 @@ function \type\()_h264_qpel8_v_lowpass_l2_neon
lowpass_8 d26, d27, d28, d29, d26, d27
transpose_8x8 d8, d9, d12, d13, d22, d23, d26, d27
vld1.64 {d0}, [ip], r2
vld1.64 {d1}, [ip], r2
vld1.64 {d2}, [ip], r2
vld1.64 {d3}, [ip], r2
vld1.64 {d4}, [ip], r2
vld1.8 {d0}, [r12], r2
vld1.8 {d1}, [r12], r2
vld1.8 {d2}, [r12], r2
vld1.8 {d3}, [r12], r2
vld1.8 {d4}, [r12], r2
vrhadd.u8 q0, q0, q4
vld1.64 {d5}, [ip], r2
vld1.8 {d5}, [r12], r2
vrhadd.u8 q1, q1, q6
vld1.64 {d10}, [ip], r2
vld1.8 {d10}, [r12], r2
vrhadd.u8 q2, q2, q11
vld1.64 {d11}, [ip], r2
vld1.8 {d11}, [r12], r2
vrhadd.u8 q5, q5, q13
.ifc \type,avg
@ -962,14 +962,14 @@ function \type\()_h264_qpel8_v_lowpass_l2_neon
sub r0, r0, r3, lsl #3
.endif
vst1.64 {d0}, [r0,:64], r3
vst1.64 {d1}, [r0,:64], r3
vst1.64 {d2}, [r0,:64], r3
vst1.64 {d3}, [r0,:64], r3
vst1.64 {d4}, [r0,:64], r3
vst1.64 {d5}, [r0,:64], r3
vst1.64 {d10}, [r0,:64], r3
vst1.64 {d11}, [r0,:64], r3
vst1.8 {d0}, [r0,:64], r3
vst1.8 {d1}, [r0,:64], r3
vst1.8 {d2}, [r0,:64], r3
vst1.8 {d3}, [r0,:64], r3
vst1.8 {d4}, [r0,:64], r3
vst1.8 {d5}, [r0,:64], r3
vst1.8 {d10}, [r0,:64], r3
vst1.8 {d11}, [r0,:64], r3
bx lr
endfunc
@ -979,32 +979,32 @@ endfunc
h264_qpel_v_lowpass_l2 avg
function put_h264_qpel8_hv_lowpass_neon_top
lowpass_const ip
mov ip, #12
1: vld1.64 {d0, d1}, [r1], r3
vld1.64 {d16,d17}, [r1], r3
subs ip, ip, #2
lowpass_const r12
mov r12, #12
1: vld1.8 {d0, d1}, [r1], r3
vld1.8 {d16,d17}, [r1], r3
subs r12, r12, #2
lowpass_8 d0, d1, d16, d17, q11, q12, narrow=0
vst1.64 {d22-d25}, [r4,:128]!
vst1.8 {d22-d25}, [r4,:128]!
bne 1b
vld1.64 {d0, d1}, [r1]
vld1.8 {d0, d1}, [r1]
lowpass_8_1 d0, d1, q12, narrow=0
mov ip, #-16
add r4, r4, ip
vld1.64 {d30,d31}, [r4,:128], ip
vld1.64 {d20,d21}, [r4,:128], ip
vld1.64 {d18,d19}, [r4,:128], ip
vld1.64 {d16,d17}, [r4,:128], ip
vld1.64 {d14,d15}, [r4,:128], ip
vld1.64 {d12,d13}, [r4,:128], ip
vld1.64 {d10,d11}, [r4,:128], ip
vld1.64 {d8, d9}, [r4,:128], ip
vld1.64 {d6, d7}, [r4,:128], ip
vld1.64 {d4, d5}, [r4,:128], ip
vld1.64 {d2, d3}, [r4,:128], ip
vld1.64 {d0, d1}, [r4,:128]
mov r12, #-16
add r4, r4, r12
vld1.8 {d30,d31}, [r4,:128], r12
vld1.8 {d20,d21}, [r4,:128], r12
vld1.8 {d18,d19}, [r4,:128], r12
vld1.8 {d16,d17}, [r4,:128], r12
vld1.8 {d14,d15}, [r4,:128], r12
vld1.8 {d12,d13}, [r4,:128], r12
vld1.8 {d10,d11}, [r4,:128], r12
vld1.8 {d8, d9}, [r4,:128], r12
vld1.8 {d6, d7}, [r4,:128], r12
vld1.8 {d4, d5}, [r4,:128], r12
vld1.8 {d2, d3}, [r4,:128], r12
vld1.8 {d0, d1}, [r4,:128]
swap4 d1, d3, d5, d7, d8, d10, d12, d14
transpose16_4x4 q0, q1, q2, q3, q4, q5, q6, q7
@ -1012,31 +1012,31 @@ function put_h264_qpel8_hv_lowpass_neon_top
swap4 d17, d19, d21, d31, d24, d26, d28, d22
transpose16_4x4 q8, q9, q10, q15, q12, q13, q14, q11
vst1.64 {d30,d31}, [r4,:128]!
vst1.64 {d6, d7}, [r4,:128]!
vst1.64 {d20,d21}, [r4,:128]!
vst1.64 {d4, d5}, [r4,:128]!
vst1.64 {d18,d19}, [r4,:128]!
vst1.64 {d2, d3}, [r4,:128]!
vst1.64 {d16,d17}, [r4,:128]!
vst1.64 {d0, d1}, [r4,:128]
vst1.8 {d30,d31}, [r4,:128]!
vst1.8 {d6, d7}, [r4,:128]!
vst1.8 {d20,d21}, [r4,:128]!
vst1.8 {d4, d5}, [r4,:128]!
vst1.8 {d18,d19}, [r4,:128]!
vst1.8 {d2, d3}, [r4,:128]!
vst1.8 {d16,d17}, [r4,:128]!
vst1.8 {d0, d1}, [r4,:128]
lowpass_8.16 q4, q12, d8, d9, d24, d25, d8
lowpass_8.16 q5, q13, d10, d11, d26, d27, d9
lowpass_8.16 q6, q14, d12, d13, d28, d29, d10
lowpass_8.16 q7, q11, d14, d15, d22, d23, d11
vld1.64 {d16,d17}, [r4,:128], ip
vld1.64 {d30,d31}, [r4,:128], ip
vld1.8 {d16,d17}, [r4,:128], r12
vld1.8 {d30,d31}, [r4,:128], r12
lowpass_8.16 q8, q15, d16, d17, d30, d31, d12
vld1.64 {d16,d17}, [r4,:128], ip
vld1.64 {d30,d31}, [r4,:128], ip
vld1.8 {d16,d17}, [r4,:128], r12
vld1.8 {d30,d31}, [r4,:128], r12
lowpass_8.16 q8, q15, d16, d17, d30, d31, d13
vld1.64 {d16,d17}, [r4,:128], ip
vld1.64 {d30,d31}, [r4,:128], ip
vld1.8 {d16,d17}, [r4,:128], r12
vld1.8 {d30,d31}, [r4,:128], r12
lowpass_8.16 q8, q15, d16, d17, d30, d31, d14
vld1.64 {d16,d17}, [r4,:128], ip
vld1.64 {d30,d31}, [r4,:128]
vld1.8 {d16,d17}, [r4,:128], r12
vld1.8 {d30,d31}, [r4,:128]
lowpass_8.16 q8, q15, d16, d17, d30, d31, d15
transpose_8x8 d12, d13, d14, d15, d8, d9, d10, d11
@ -1068,14 +1068,14 @@ function \type\()_h264_qpel8_hv_lowpass_neon
sub r0, r0, r2, lsl #3
.endif
vst1.64 {d12}, [r0,:64], r2
vst1.64 {d13}, [r0,:64], r2
vst1.64 {d14}, [r0,:64], r2
vst1.64 {d15}, [r0,:64], r2
vst1.64 {d8}, [r0,:64], r2
vst1.64 {d9}, [r0,:64], r2
vst1.64 {d10}, [r0,:64], r2
vst1.64 {d11}, [r0,:64], r2
vst1.8 {d12}, [r0,:64], r2
vst1.8 {d13}, [r0,:64], r2
vst1.8 {d14}, [r0,:64], r2
vst1.8 {d15}, [r0,:64], r2
vst1.8 {d8}, [r0,:64], r2
vst1.8 {d9}, [r0,:64], r2
vst1.8 {d10}, [r0,:64], r2
vst1.8 {d11}, [r0,:64], r2
mov lr, r10
bx lr
@ -1090,12 +1090,12 @@ function \type\()_h264_qpel8_hv_lowpass_l2_neon
mov r10, lr
bl put_h264_qpel8_hv_lowpass_neon_top
vld1.64 {d0, d1}, [r2,:128]!
vld1.64 {d2, d3}, [r2,:128]!
vld1.8 {d0, d1}, [r2,:128]!
vld1.8 {d2, d3}, [r2,:128]!
vrhadd.u8 q0, q0, q6
vld1.64 {d4, d5}, [r2,:128]!
vld1.8 {d4, d5}, [r2,:128]!
vrhadd.u8 q1, q1, q7
vld1.64 {d6, d7}, [r2,:128]!
vld1.8 {d6, d7}, [r2,:128]!
vrhadd.u8 q2, q2, q4
vrhadd.u8 q3, q3, q5
.ifc \type,avg
@ -1117,14 +1117,14 @@ function \type\()_h264_qpel8_hv_lowpass_l2_neon
vrhadd.u8 d7, d7, d23
sub r0, r0, r3, lsl #3
.endif
vst1.64 {d0}, [r0,:64], r3
vst1.64 {d1}, [r0,:64], r3
vst1.64 {d2}, [r0,:64], r3
vst1.64 {d3}, [r0,:64], r3
vst1.64 {d4}, [r0,:64], r3
vst1.64 {d5}, [r0,:64], r3
vst1.64 {d6}, [r0,:64], r3
vst1.64 {d7}, [r0,:64], r3
vst1.8 {d0}, [r0,:64], r3
vst1.8 {d1}, [r0,:64], r3
vst1.8 {d2}, [r0,:64], r3
vst1.8 {d3}, [r0,:64], r3
vst1.8 {d4}, [r0,:64], r3
vst1.8 {d5}, [r0,:64], r3
vst1.8 {d6}, [r0,:64], r3
vst1.8 {d7}, [r0,:64], r3
mov lr, r10
bx lr
@ -1177,7 +1177,7 @@ function ff_\type\()_h264_qpel8_mc10_neon, export=1
lowpass_const r3
mov r3, r1
sub r1, r1, #2
mov ip, #8
mov r12, #8
b \type\()_h264_qpel8_h_lowpass_l2_neon
endfunc
@ -1185,7 +1185,7 @@ function ff_\type\()_h264_qpel8_mc20_neon, export=1
lowpass_const r3
sub r1, r1, #2
mov r3, r2
mov ip, #8
mov r12, #8
b \type\()_h264_qpel8_h_lowpass_neon
endfunc
@ -1193,13 +1193,13 @@ function ff_\type\()_h264_qpel8_mc30_neon, export=1
lowpass_const r3
add r3, r1, #1
sub r1, r1, #2
mov ip, #8
mov r12, #8
b \type\()_h264_qpel8_h_lowpass_l2_neon
endfunc
function ff_\type\()_h264_qpel8_mc01_neon, export=1
push {lr}
mov ip, r1
mov r12, r1
\type\()_h264_qpel8_mc01:
lowpass_const r3
mov r3, r2
@ -1222,12 +1222,12 @@ T mov sp, r0
mov r0, sp
sub r1, r1, #2
mov r3, #8
mov ip, #8
mov r12, #8
vpush {d8-d15}
bl put_h264_qpel8_h_lowpass_neon
ldrd r0, [r11], #8
mov r3, r2
add ip, sp, #64
add r12, sp, #64
sub r1, r1, r2, lsl #1
mov r2, #8
bl \type\()_h264_qpel8_v_lowpass_l2_neon
@ -1248,7 +1248,7 @@ T mov sp, r0
sub r1, r1, #2
mov r3, #8
mov r0, sp
mov ip, #8
mov r12, #8
vpush {d8-d15}
bl put_h264_qpel8_h_lowpass_neon
mov r4, r0
@ -1333,7 +1333,7 @@ endfunc
function ff_\type\()_h264_qpel8_mc03_neon, export=1
push {lr}
add ip, r1, r2
add r12, r1, r2
b \type\()_h264_qpel8_mc01
endfunc
@ -1385,7 +1385,7 @@ endfunc
function ff_\type\()_h264_qpel16_mc01_neon, export=1
push {r4, lr}
mov ip, r1
mov r12, r1
\type\()_h264_qpel16_mc01:
lowpass_const r3
mov r3, r2
@ -1412,7 +1412,7 @@ T mov sp, r0
bl put_h264_qpel16_h_lowpass_neon
ldrd r0, [r11], #8
mov r3, r2
add ip, sp, #64
add r12, sp, #64
sub r1, r1, r2, lsl #1
mov r2, #16
bl \type\()_h264_qpel16_v_lowpass_l2_neon
@ -1515,7 +1515,7 @@ endfunc
function ff_\type\()_h264_qpel16_mc03_neon, export=1
push {r4, lr}
add ip, r1, r2
add r12, r1, r2
b \type\()_h264_qpel16_mc01
endfunc