Merge remote-tracking branch 'qatar/master'

[ffmpeg] / libavcodec / arm / vp8_armv6.S

diff --git a/libavcodec/arm/vp8_armv6.S b/libavcodec/arm/vp8_armv6.S
index 93f4dd664bfdc357d7aa67b06e67bd0135b7d321..b995360e0c6eed308ae50e2384384af3b184e067 100644 (file)
--- a/libavcodec/arm/vp8_armv6.S
+++ b/libavcodec/arm/vp8_armv6.S
@@ -25,13 +25,18 @@
         lsl             \cw, \cw, \t0
         lsl             \t0, \h,  \t0
         rsb             \h,  \pr, #256
+        it              cs
         ldrhcs          \t1, [\buf], #2
         smlabb          \h,  \t0, \pr, \h
+T       itttt           cs
         rev16cs         \t1, \t1
-        orrcs           \cw, \cw, \t1, lsl \bs
+A       orrcs           \cw, \cw, \t1, lsl \bs
+T       lslcs           \t1, \t1, \bs
+T       orrcs           \cw, \cw, \t1
         subcs           \bs, \bs, #16
         lsr             \h,  \h,  #8
         cmp             \cw, \h,  lsl #16
+        itt             ge
         subge           \cw, \cw, \h,  lsl #16
         subge           \h,  \t0, \h
 .endm
@@ -40,14 +45,20 @@
         adds            \bs, \bs, \t0
         lsl             \cw, \cw, \t0
         lsl             \t0, \h,  \t0
+        it              cs
         ldrhcs          \t1, [\buf], #2
         mov             \h,  #128
+        it              cs
         rev16cs         \t1, \t1
         add             \h,  \h,  \t0, lsl #7
-        orrcs           \cw, \cw, \t1, lsl \bs
+A       orrcs           \cw, \cw, \t1, lsl \bs
+T       ittt            cs
+T       lslcs           \t1, \t1, \bs
+T       orrcs           \cw, \cw, \t1
         subcs           \bs, \bs, #16
         lsr             \h,  \h,  #8
         cmp             \cw, \h,  lsl #16
+        itt             ge
         subge           \cw, \cw, \h,  lsl #16
         subge           \h,  \t0, \h
 .endm
@@ -59,6 +70,7 @@ function ff_decode_block_coeffs_armv6, export=1
         cmp             r3,  #0
         ldr             r11, [r5]
         ldm             r0,  {r5-r7}                    @ high, bits, buf
+        it              ne
         pkhtbne         r11, r11, r11, asr #16
         ldr             r8,  [r0, #16]                  @ code_word
 0:
@@ -80,19 +92,26 @@ function ff_decode_block_coeffs_armv6, export=1
         adds            r6,  r6,  r9
         add             r4,  r4,  #11
         lsl             r8,  r8,  r9
+        it              cs
         ldrhcs          r10, [r7], #2
         lsl             r9,  r5,  r9
         mov             r5,  #128
+        it              cs
         rev16cs         r10, r10
         add             r5,  r5,  r9,  lsl #7
-        orrcs           r8,  r8,  r10, lsl r6
+T       ittt            cs
+T       lslcs           r10, r10, r6
+T       orrcs           r8,  r8,  r10
+A       orrcs           r8,  r8,  r10, lsl r6
         subcs           r6,  r6,  #16
         lsr             r5,  r5,  #8
         cmp             r8,  r5,  lsl #16
         movrel          r10, zigzag_scan-1
+        itt             ge
         subge           r8,  r8,  r5,  lsl #16
         subge           r5,  r9,  r5
         ldrb            r10, [r10, r3]
+        it              ge
         rsbge           r12, r12, #0
         cmp             r3,  #16
         strh            r12, [r1, r10]
@@ -108,6 +127,7 @@ function ff_decode_block_coeffs_armv6, export=1
         ldr             r0,  [sp]
         ldr             r9,  [r0, #12]
         cmp             r7,  r9
+        it              hi
         movhi           r7,  r9
         stm             r0,  {r5-r7}                    @ high, bits, buf
         str             r8,  [r0, #16]                  @ code_word
@@ -131,11 +151,13 @@ function ff_decode_block_coeffs_armv6, export=1
         mov             r12, #2
         ldrb            r0,  [r4, #4]
         rac_get_prob    r5,  r6,  r7,  r8,  r0,  r9,  r10
+        it              ge
         addge           r12, #1
         ldrb            r9,  [lr, r5]
         blt             4f
         ldrb            r0,  [r4, #5]
         rac_get_prob    r5,  r6,  r7,  r8,  r0,  r9,  r10
+        it              ge
         addge           r12, #1
         ldrb            r9,  [lr, r5]
         b               4f
@@ -153,6 +175,7 @@ function ff_decode_block_coeffs_armv6, export=1
         mov             r12, #5
         mov             r0,  #159
         rac_get_prob    r5,  r6,  r7,  r8,  r0,  r9,  r10
+        it              ge
         addge           r12, r12, #1
         ldrb            r9,  [lr, r5]
         b               4f
@@ -160,23 +183,28 @@ function ff_decode_block_coeffs_armv6, export=1
         mov             r12, #7
         mov             r0,  #165
         rac_get_prob    r5,  r6,  r7,  r8,  r0,  r9,  r10
+        it              ge
         addge           r12, r12, #2
         ldrb            r9,  [lr, r5]
         mov             r0,  #145
         rac_get_prob    r5,  r6,  r7,  r8,  r0,  r9,  r10
+        it              ge
         addge           r12, r12, #1
         ldrb            r9,  [lr, r5]
         b               4f
 3:
         ldrb            r0,  [r4, #8]
         rac_get_prob    r5,  r6,  r7,  r8,  r0,  r9,  r10
+        it              ge
         addge           r4,  r4,  #1
         ldrb            r9,  [lr, r5]
+        ite             ge
         movge           r12, #2
         movlt           r12, #0
         ldrb            r0,  [r4, #9]
         rac_get_prob    r5,  r6,  r7,  r8,  r0,  r9,  r10
         mov             r9,  #8
+        it              ge
         addge           r12, r12, #1
         movrel          r4,  X(ff_vp8_dct_cat_prob)
         lsl             r9,  r9,  r12
@@ -189,6 +217,7 @@ function ff_decode_block_coeffs_armv6, export=1
         lsl             r1,  r1,  #1
         rac_get_prob    r5,  r6,  r7,  r8,  r0,  r9,  r10
         ldrb            r0,  [r4], #1
+        it              ge
         addge           r1,  r1,  #1
         cmp             r0,  #0
         bne             1b
@@ -200,6 +229,7 @@ function ff_decode_block_coeffs_armv6, export=1
         add             r4,  r2,  r4
         add             r4,  r4,  #22
         rac_get_128     r5,  r6,  r7,  r8,  r9,  r10
+        it              ge
         rsbge           r12, r12, #0
         smulbb          r12, r12, r11
         movrel          r9,  zigzag_scan-1