Merge remote-tracking branch 'qatar/master'

[ffmpeg] / libavcodec / arm / simple_idct_armv5te.S

diff --git a/libavcodec/arm/simple_idct_armv5te.S b/libavcodec/arm/simple_idct_armv5te.S
index 3c4b5c06d1a6a07234d3fc5965f9ca7c63c61207..71727ceccc2f0412668b02320ede852c830f682f 100644 (file)
--- a/libavcodec/arm/simple_idct_armv5te.S
+++ b/libavcodec/arm/simple_idct_armv5te.S
@@ -49,6 +49,7 @@ function idct_row_armv5te
         ldrd   v1, [a1, #8]
         ldrd   a3, [a1]              /* a3 = row[1:0], a4 = row[3:2] */
         orrs   v1, v1, v2
+        itt    eq
         cmpeq  v1, a4
         cmpeq  v1, a3, lsr #16
         beq    row_dc_only
@@ -269,6 +270,7 @@ function idct_col_armv5te
         ldmfd  sp!, {a3, a4}
         adds   a2, a3, v1
         mov    a2, a2, lsr #20
+        it     mi
         orrmi  a2, a2, #0xf000
         add    ip, a4, v2
         mov    ip, ip, asr #20
@@ -276,6 +278,7 @@ function idct_col_armv5te
         str    a2, [a1]
         subs   a3, a3, v1
         mov    a2, a3, lsr #20
+        it     mi
         orrmi  a2, a2, #0xf000
         sub    a4, a4, v2
         mov    a4, a4, asr #20
@@ -285,6 +288,7 @@ function idct_col_armv5te
 
         subs   a2, a3, v3
         mov    a2, a2, lsr #20
+        it     mi
         orrmi  a2, a2, #0xf000
         sub    ip, a4, v4
         mov    ip, ip, asr #20
@@ -292,6 +296,7 @@ function idct_col_armv5te
         str    a2, [a1, #(16*1)]
         adds   a3, a3, v3
         mov    a2, a3, lsr #20
+        it     mi
         orrmi  a2, a2, #0xf000
         add    a4, a4, v4
         mov    a4, a4, asr #20
@@ -301,6 +306,7 @@ function idct_col_armv5te
 
         adds   a2, a3, v5
         mov    a2, a2, lsr #20
+        it     mi
         orrmi  a2, a2, #0xf000
         add    ip, a4, v6
         mov    ip, ip, asr #20
@@ -308,6 +314,7 @@ function idct_col_armv5te
         str    a2, [a1, #(16*2)]
         subs   a3, a3, v5
         mov    a2, a3, lsr #20
+        it     mi
         orrmi  a2, a2, #0xf000
         sub    a4, a4, v6
         mov    a4, a4, asr #20
@@ -317,6 +324,7 @@ function idct_col_armv5te
 
         adds   a2, a3, v7
         mov    a2, a2, lsr #20
+        it     mi
         orrmi  a2, a2, #0xf000
         add    ip, a4, fp
         mov    ip, ip, asr #20
@@ -324,6 +332,7 @@ function idct_col_armv5te
         str    a2, [a1, #(16*3)]
         subs   a3, a3, v7
         mov    a2, a3, lsr #20
+        it     mi
         orrmi  a2, a2, #0xf000
         sub    a4, a4, fp
         mov    a4, a4, asr #20
@@ -335,15 +344,19 @@ endfunc
 
 .macro  clip   dst, src:vararg
         movs   \dst, \src
+        it     mi
         movmi  \dst, #0
         cmp    \dst, #255
+        it     gt
         movgt  \dst, #255
 .endm
 
 .macro  aclip  dst, src:vararg
         adds   \dst, \src
+        it     mi
         movmi  \dst, #0
         cmp    \dst, #255
+        it     gt
         movgt  \dst, #255
 .endm
 
@@ -370,35 +383,35 @@ function idct_col_put_armv5te
         orr    a2, a3, a4, lsl #8
         rsb    v2, lr, lr, lsl #3
         ldmfd  sp!, {a3, a4}
-        strh   a2, [v2, v1]!
+        strh_pre a2, v2, v1
 
         sub    a2, a3, v3
         clip   a2, a2, asr #20
         sub    ip, a4, v4
         clip   ip, ip, asr #20
         orr    a2, a2, ip, lsl #8
-        strh   a2, [v1, lr]!
+        strh_pre a2, v1, lr
         add    a3, a3, v3
         clip   a2, a3, asr #20
         add    a4, a4, v4
         clip   a4, a4, asr #20
         orr    a2, a2, a4, lsl #8
         ldmfd  sp!, {a3, a4}
-        strh   a2, [v2, -lr]!
+        strh_dpre a2, v2, lr
 
         add    a2, a3, v5
         clip   a2, a2, asr #20
         add    ip, a4, v6
         clip   ip, ip, asr #20
         orr    a2, a2, ip, lsl #8
-        strh   a2, [v1, lr]!
+        strh_pre a2, v1, lr
         sub    a3, a3, v5
         clip   a2, a3, asr #20
         sub    a4, a4, v6
         clip   a4, a4, asr #20
         orr    a2, a2, a4, lsl #8
         ldmfd  sp!, {a3, a4}
-        strh   a2, [v2, -lr]!
+        strh_dpre a2, v2, lr
 
         add    a2, a3, v7
         clip   a2, a2, asr #20
@@ -411,7 +424,7 @@ function idct_col_put_armv5te
         sub    a4, a4, fp
         clip   a4, a4, asr #20
         orr    a2, a2, a4, lsl #8
-        strh   a2, [v2, -lr]
+        strh_dpre a2, v2, lr
 
         ldr    pc, [sp], #4
 endfunc
@@ -436,7 +449,7 @@ function idct_col_add_armv5te
         ldr    v1, [sp, #32]
         sub    a4, a4, v2
         rsb    v2, v1, v1, lsl #3
-        ldrh   ip, [v2, lr]!
+        ldrh_pre ip, v2, lr
         strh   a2, [lr]
         and    a2, ip, #255
         aclip  a3, a2, a3, asr #20
@@ -448,7 +461,7 @@ function idct_col_add_armv5te
         strh   a2, [v2]
 
         ldmfd  sp!, {a3, a4}
-        ldrh   ip, [lr, v1]!
+        ldrh_pre ip, lr, v1
         sub    a2, a3, v3
         add    a3, a3, v3
         and    v3, ip, #255
@@ -458,7 +471,7 @@ function idct_col_add_armv5te
         aclip  v3, v3, ip, lsr #8
         orr    a2, a2, v3, lsl #8
         add    a4, a4, v4
-        ldrh   ip, [v2, -v1]!
+        ldrh_dpre ip, v2, v1
         strh   a2, [lr]
         and    a2, ip, #255
         aclip  a3, a2, a3, asr #20
@@ -468,7 +481,7 @@ function idct_col_add_armv5te
         strh   a2, [v2]
 
         ldmfd  sp!, {a3, a4}
-        ldrh   ip, [lr, v1]!
+        ldrh_pre ip, lr, v1
         add    a2, a3, v5
         sub    a3, a3, v5
         and    v3, ip, #255
@@ -478,7 +491,7 @@ function idct_col_add_armv5te
         aclip  v3, v3, ip, lsr #8
         orr    a2, a2, v3, lsl #8
         sub    a4, a4, v6
-        ldrh   ip, [v2, -v1]!
+        ldrh_dpre ip, v2, v1
         strh   a2, [lr]
         and    a2, ip, #255
         aclip  a3, a2, a3, asr #20
@@ -488,7 +501,7 @@ function idct_col_add_armv5te
         strh   a2, [v2]
 
         ldmfd  sp!, {a3, a4}
-        ldrh   ip, [lr, v1]!
+        ldrh_pre ip, lr, v1
         add    a2, a3, v7
         sub    a3, a3, v7
         and    v3, ip, #255
@@ -498,7 +511,7 @@ function idct_col_add_armv5te
         aclip  v3, v3, ip, lsr #8
         orr    a2, a2, v3, lsl #8
         sub    a4, a4, fp
-        ldrh   ip, [v2, -v1]!
+        ldrh_dpre ip, v2, v1
         strh   a2, [lr]
         and    a2, ip, #255
         aclip  a3, a2, a3, asr #20