x86: replace explicit REP_RETs with RETs

From x86inc:
> On AMD cpus <=K10, an ordinary ret is slow if it immediately follows either
> a branch or a branch target. So switch to a 2-byte form of ret in that case.
> We can automatically detect "follows a branch", but not a branch target.
> (SSSE3 is a sufficient condition to know that your cpu doesn't have this problem.)

x86inc can automatically determine whether to use REP_RET rather than
REP in most of these cases, so impact is minimal. Additionally, a few
REP_RETs were used unnecessary, despite the return being nowhere near a
branch.

The only CPUs affected were AMD K10s, made between 2007 and 2011, 16
years ago and 12 years ago, respectively.

In the future, everyone involved with x86inc should consider dropping
REP_RETs altogether.

6 files changed, 16 insertions, 16 deletions

diff --git a/libswscale/x86/input.asm b/libswscale/x86/input.asm
index 6de6733faa..a197183f1f 100644
--- a/libswscale/x86/input.asm
+++ b/libswscale/x86/input.asm
@@ -207,7 +207,7 @@ cglobal %2 %+ 24ToY, 6, 6, %1, dst, src, u1, u2, w, table
     mova    [dstq+wq], m0
     add            wq, mmsize
     jl .loop
-    REP_RET
+    RET
 %endif ; ARCH_X86_64 && %0 == 3
 %endmacro
 
@@ -313,7 +313,7 @@ cglobal %2 %+ 24ToUV, 7, 7, %1, dstU, dstV, u1, src, u2, w, table
     mova   [dstVq+wq], m2
     add            wq, mmsize
     jl .loop
-    REP_RET
+    RET
 %endif ; ARCH_X86_64 && %0 == 3
 %endmacro
 
@@ -394,7 +394,7 @@ cglobal %2%3%4%5 %+ ToY, 6, 6, %1, dst, src, u1, u2, w, table
     add            wq, 2
     jl .loop2
 .end:
-    REP_RET
+    RET
 %endif ; %0 == 3
 %endmacro
 
@@ -491,7 +491,7 @@ cglobal %2%3%4%5 %+ ToUV, 7, 7, %1, dstU, dstV, u1, src, u2, w, table
     add            wq, 2
     jl .loop2
 .end:
-    REP_RET
+    RET
 %endif ; ARCH_X86_64 && %0 == 3
 %endmacro
 
@@ -543,7 +543,7 @@ RGB32_FUNCS 8, 12
     mova    [dstq+wq], m0
     add            wq, mmsize
     jl .loop_%1
-    REP_RET
+    RET
 %endmacro
 
 ; %1 = nr. of XMM registers
@@ -599,7 +599,7 @@ cglobal %2ToY, 5, 5, %1, dst, unused0, unused1, src, w
     movhps [dstVq+wq], m1
     add            wq, mmsize / 2
     jl .loop_%1
-    REP_RET
+    RET
 %endmacro
 
 ; %1 = nr. of XMM registers
@@ -657,7 +657,7 @@ cglobal %2ToUV, 4, 5, %1, dstU, dstV, unused, src, w
 %endif ; nv12/21
     add            wq, mmsize
     jl .loop_%1
-    REP_RET
+    RET
 %endmacro
 
 ; %1 = nr. of XMM registers
diff --git a/libswscale/x86/output.asm b/libswscale/x86/output.asm
index f943a27534..95ec2fa885 100644
--- a/libswscale/x86/output.asm
+++ b/libswscale/x86/output.asm
@@ -297,7 +297,7 @@ cglobal yuv2planeX_%1, %3, 8, %2, filter, fltsize, src, dst, w, dither, offset
     test          dstq, 15
     jnz .unaligned
     yuv2planeX_mainloop %1, a
-    REP_RET
+    RET
 .unaligned:
     yuv2planeX_mainloop %1, u
 %endif ; mmsize == 8/16
@@ -307,10 +307,10 @@ cglobal yuv2planeX_%1, %3, 8, %2, filter, fltsize, src, dst, w, dither, offset
     ADD             rsp, pad
     RET
 %else ; x86-64
-    REP_RET
+    RET
 %endif ; x86-32/64
 %else ; %1 == 9/10/16
-    REP_RET
+    RET
 %endif ; %1 == 8/9/10/16
 %endmacro
 
@@ -433,10 +433,10 @@ cglobal yuv2plane1_%1, %3, %3, %2, src, dst, w, dither, offset
     test          dstq, 15
     jnz .unaligned
     yuv2plane1_mainloop %1, a
-    REP_RET
+    RET
 .unaligned:
     yuv2plane1_mainloop %1, u
-    REP_RET
+    RET
 %endmacro
 
 INIT_XMM sse2
diff --git a/libswscale/x86/scale.asm b/libswscale/x86/scale.asm
index c62ae3dcc2..2e14c8c023 100644
--- a/libswscale/x86/scale.asm
+++ b/libswscale/x86/scale.asm
@@ -357,7 +357,7 @@ cglobal hscale%1to%2_%4, %5, 10, %6, pos0, dst, w, srcmem, filter, fltpos, fltsi
     add           wq, 2
 %endif ; %3 ==/!= X
     jl .loop
-    REP_RET
+    RET
 %endmacro
 
 ; SCALE_FUNCS source_width, intermediate_nbits, n_xmm
diff --git a/libswscale/x86/scale_avx2.asm b/libswscale/x86/scale_avx2.asm
index 37095e596a..179895666a 100644
--- a/libswscale/x86/scale_avx2.asm
+++ b/libswscale/x86/scale_avx2.asm
@@ -144,7 +144,7 @@ cglobal hscale8to15_%1, 7, 9, 16, pos0, dst, w, srcmem, filter, fltpos, fltsize,
     cmp countq, wq
     jl .tail_loop
 .end:
-REP_RET
+RET
 %endmacro
 
 %if ARCH_X86_64
diff --git a/libswscale/x86/yuv2yuvX.asm b/libswscale/x86/yuv2yuvX.asm
index d5b03495fd..369c850674 100644
--- a/libswscale/x86/yuv2yuvX.asm
+++ b/libswscale/x86/yuv2yuvX.asm
@@ -121,7 +121,7 @@ cglobal yuv2yuvX, 7, 7, 8, filter, filterSize, src, dest, dstW, dither, offset
     mov                  filterSizeq, filterq
     cmp                  offsetq, dstWq
     jb                  .outerloop
-    REP_RET
+    RET
 %endmacro
 
 INIT_MMX mmxext
diff --git a/libswscale/x86/yuv_2_rgb.asm b/libswscale/x86/yuv_2_rgb.asm
index c5fa3ee690..e3470fd9ad 100644
--- a/libswscale/x86/yuv_2_rgb.asm
+++ b/libswscale/x86/yuv_2_rgb.asm
@@ -354,7 +354,7 @@ add imageq, 8 * depth * time_num
 add indexq, 4 * time_num
 js .loop0
 
-REP_RET
+RET
 
 %endmacro