Merge remote-tracking branch 'qatar/master'

* qatar/master:
  x86: fix build with nasm 2.08
  x86: use nop cpu directives only if supported
  x86: fix rNmp macros with nasm
  build: add trailing / to yasm/nasm -I flags
  x86: use 32-bit source registers with movd instruction
  x86: add colons after labels

Conflicts:
	Makefile
	libavutil/x86/x86inc.asm

Merged-by: Michael Niedermayer <michaelni@gmx.at>

common.mak

@@ -29,7 +29,8 @@ CFLAGS     += $(ECFLAGS)
 CCFLAGS     = $(CPPFLAGS) $(CFLAGS)
 ASFLAGS    := $(CPPFLAGS) $(ASFLAGS)
 CXXFLAGS   += $(CPPFLAGS) $(CFLAGS)
-YASMFLAGS  += $(IFLAGS) -I$(SRC_PATH)/libavutil/x86/ -Pconfig.asm
+YASMFLAGS  += $(IFLAGS:%=%/) -I$(SRC_PATH)/libavutil/x86/ -Pconfig.asm
+
 HOSTCCFLAGS = $(IFLAGS) $(HOSTCFLAGS)
 LDFLAGS    := $(ALLFFLIBS:%=-Llib%) $(LDFLAGS)
 

configure

@@ -1220,6 +1220,7 @@ HAVE_LIST="
     closesocket
     cmov
     cpuid
+    cpunop
     dcbzl
     dev_bktr_ioctl_bt848_h
     dev_bktr_ioctl_meteor_h
@@ -3229,6 +3230,7 @@ EOF
             die "yasm not found, use --disable-yasm for a crippled build"
         check_yasm "vextractf128 xmm0, ymm0, 0" || disable avx
         check_yasm "vfmaddps ymm0, ymm1, ymm2, ymm3" || disable fma4
+        check_yasm "CPU amdnop" && enable cpunop
     fi
 
     case "$cpu" in

libavcodec/x86/deinterlace.asm

@@ -39,7 +39,7 @@ cglobal deinterlace_line_mmx,         7,7,7, dst, lum_m4, lum_m3, lum_m2, lum_m1
 %endif
     pxor  mm7, mm7
     movq  mm6, [pw_4]
-.nextrow
+.nextrow:
     movd  mm0, [lum_m4q]
     movd  mm1, [lum_m3q]
     movd  mm2, [lum_m2q]

libavcodec/x86/dsputil_yasm.asm

@@ -1143,7 +1143,7 @@ VECTOR_CLIP_INT32 6, 1, 0, 0
 cglobal vector_fmul_reverse, 4,4,2, dst, src0, src1, len
     lea       lenq, [lend*4 - 2*mmsize]
 ALIGN 16
-.loop
+.loop:
 %if cpuflag(avx)
     vmovaps     xmm0, [src1q + 16]
     vinsertf128 m0, m0, [src1q], 1
@@ -1182,7 +1182,7 @@ VECTOR_FMUL_REVERSE
 cglobal vector_fmul_add, 5,5,2, dst, src0, src1, src2, len
     lea       lenq, [lend*4 - 2*mmsize]
 ALIGN 16
-.loop
+.loop:
     mova    m0,   [src0q + lenq]
     mova    m1,   [src0q + lenq + mmsize]
     mulps   m0, m0, [src1q + lenq]
@@ -1313,7 +1313,7 @@ cglobal bswap32_buf, 3,4,5
     add      r0, 4
     dec      r2
     jnz      .loop2
-.end
+.end:
     RET
 
 ; %1 = aligned/unaligned

libavcodec/x86/dsputilenc_yasm.asm

@@ -184,7 +184,7 @@ cglobal hadamard8_diff16_%1, 5, 6, %2
     call hadamard8x8_diff_%1
     add            r5d, eax
 
-.done
+.done:
     mov            eax, r5d
 %ifndef m8
     ADD            rsp, pad
@@ -288,7 +288,7 @@ cglobal sse16_sse2, 5, 5, 8
     pxor      m0, m0         ; mm0 = 0
     pxor      m7, m7         ; mm7 holds the sum
 
-.next2lines ; FIXME why are these unaligned movs? pix1[] is aligned
+.next2lines: ; FIXME why are these unaligned movs? pix1[] is aligned
     movu      m1, [r1   ]    ; mm1 = pix1[0][0-15]
     movu      m2, [r2   ]    ; mm2 = pix2[0][0-15]
     movu      m3, [r1+r3]    ; mm3 = pix1[1][0-15]

libavcodec/x86/fft_mmx.asm

@@ -608,7 +608,7 @@ cglobal fft_calc, 2,5,8
     add     rcx, 3
     shl     r2, cl
     sub     r4, r2
-.loop
+.loop:
 %if mmsize == 8
     PSWAPD  m0, [r4 + r2 + 4]
     mova [r4 + r2 + 4], m0

libavcodec/x86/fmtconvert.asm

@@ -404,7 +404,7 @@ cglobal float_interleave2_%1, 3,4,%2, dst, src, len, src1
     mov     src1q, [srcq+gprsize]
     mov      srcq, [srcq        ]
     sub     src1q, srcq
-.loop
+.loop:
     MOVPS      m0, [srcq             ]
     MOVPS      m1, [srcq+src1q       ]
     MOVPS      m3, [srcq      +mmsize]

libavcodec/x86/h264_chromamc.asm

@@ -69,7 +69,7 @@ SECTION .text
 
 %macro mv0_pixels_mc8 0
     lea           r4, [r2*2 ]
-.next4rows
+.next4rows:
     movq         mm0, [r1   ]
     movq         mm1, [r1+r2]
     add           r1, r4
@@ -117,7 +117,7 @@ cglobal %1_%2_chroma_mc8_%3, 6, 7 + extra_regs, 0
     mv0_pixels_mc8
     REP_RET
 
-.at_least_one_non_zero
+.at_least_one_non_zero:
 %ifidn %2, rv40
 %if ARCH_X86_64
     mov           r7, r5
@@ -145,7 +145,7 @@ cglobal %1_%2_chroma_mc8_%3, 6, 7 + extra_regs, 0
     test         r4d, r4d
     mov           r6, r2        ; dxy = x ? 1 : stride
     jne .both_non_zero
-.my_is_zero
+.my_is_zero:
     ; mx == 0 XOR my == 0 - 1 dimensional filter only
     or           r4d, r5d       ; x + y
 
@@ -166,7 +166,7 @@ cglobal %1_%2_chroma_mc8_%3, 6, 7 + extra_regs, 0
     pxor          m7, m7
     psubw         m4, m5        ; mm4 = A = 8-x
 
-.next1drow
+.next1drow:
     movq          m0, [r1   ]   ; mm0 = src[0..7]
     movq          m2, [r1+r6]   ; mm1 = src[1..8]
 
@@ -197,7 +197,7 @@ cglobal %1_%2_chroma_mc8_%3, 6, 7 + extra_regs, 0
     jne .next1drow
     REP_RET
 
-.both_non_zero ; general case, bilinear
+.both_non_zero: ; general case, bilinear
     movd          m4, r4d         ; x
     movd          m6, r5d         ; y
 %ifidn %2, rv40
@@ -232,7 +232,7 @@ cglobal %1_%2_chroma_mc8_%3, 6, 7 + extra_regs, 0
 
     movq          m0, [r1  ]      ; mm0 = src[0..7]
     movq          m1, [r1+1]      ; mm1 = src[1..8]
-.next2drow
+.next2drow:
     add           r1, r2
 
     movq          m2, m0
@@ -330,7 +330,7 @@ cglobal %1_%2_chroma_mc4_%3, 6, 6 + extra_regs, 0
     pmullw        m6, m2
     paddw         m6, m0
 
-.next2rows
+.next2rows:
     movd          m0, [r1  ]
     movd          m1, [r1+1]
     add           r1, r2
@@ -397,7 +397,7 @@ cglobal %1_%2_chroma_mc2_%3, 6, 7, 0
     punpcklbw     m2, m7
     pshufw        m2, m2, 0x94    ; mm0 = src[0,1,1,2]
 
-.nextrow
+.nextrow:
     add           r1, r2
     movq          m1, m2
     pmaddwd       m1, m5          ; mm1 = A * src[0,1] + B * src[1,2]
@@ -474,7 +474,7 @@ cglobal %1_%2_chroma_mc8_%3, 6, 7, 8
     mv0_pixels_mc8
     REP_RET
 
-.at_least_one_non_zero
+.at_least_one_non_zero:
     test         r5d, r5d
     je .my_is_zero
     test         r4d, r4d
@@ -501,7 +501,7 @@ cglobal %1_%2_chroma_mc8_%3, 6, 7, 8
     movlhps       m7, m7
     movlhps       m6, m6
 
-.next2rows
+.next2rows:
     movq          m1, [r1+r2*1   ]
     movq          m2, [r1+r2*1+1]
     movq          m3, [r1+r2*2  ]
@@ -535,7 +535,7 @@ cglobal %1_%2_chroma_mc8_%3, 6, 7, 8
     jg .next2rows
     REP_RET
 
-.my_is_zero
+.my_is_zero:
     mov          r5d, r4d
     shl          r4d, 8
     add           r4, 8
@@ -545,7 +545,7 @@ cglobal %1_%2_chroma_mc8_%3, 6, 7, 8
     pshuflw       m7, m7, 0
     movlhps       m7, m7
 
-.next2xrows
+.next2xrows:
     movq          m0, [r1     ]
     movq          m1, [r1   +1]
     movq          m2, [r1+r2  ]
@@ -572,7 +572,7 @@ cglobal %1_%2_chroma_mc8_%3, 6, 7, 8
     jg .next2xrows
     REP_RET
 
-.mx_is_zero
+.mx_is_zero:
     mov          r4d, r5d
     shl          r5d, 8
     add           r5, 8
@@ -582,7 +582,7 @@ cglobal %1_%2_chroma_mc8_%3, 6, 7, 8
     pshuflw       m7, m7, 0
     movlhps       m7, m7
 
-.next2yrows
+.next2yrows:
     movq          m0, [r1     ]
     movq          m1, [r1+r2  ]
     movdqa        m2, m1
@@ -632,7 +632,7 @@ cglobal %1_%2_chroma_mc4_%3, 6, 7, 0
     punpcklbw     m0, [r1+1]
     pshufw        m6, m6, 0
 
-.next2rows
+.next2rows:
     movd          m1, [r1+r2*1  ]
     movd          m3, [r1+r2*2  ]
     punpcklbw     m1, [r1+r2*1+1]

libavcodec/x86/h264_chromamc_10bit.asm

@@ -38,7 +38,7 @@ SECTION .text
 %macro MV0_PIXELS_MC8 0
     lea           r4, [r2*3   ]
     lea           r5, [r2*4   ]
-.next4rows
+.next4rows:
     movu          m0, [r1     ]
     movu          m1, [r1+r2  ]
     CHROMAMC_AVG  m0, [r0     ]
@@ -72,14 +72,14 @@ cglobal %1_h264_chroma_mc8_10, 6,7,8
     MV0_PIXELS_MC8
     REP_RET
 
-.at_least_one_non_zero
+.at_least_one_non_zero:
     mov          r6d, 2
     test         r5d, r5d
     je .x_interpolation
     mov           r6, r2        ; dxy = x ? 1 : stride
     test         r4d, r4d
     jne .xy_interpolation
-.x_interpolation
+.x_interpolation:
     ; mx == 0 XOR my == 0 - 1 dimensional filter only
     or           r4d, r5d       ; x + y
     movd          m5, r4d
@@ -88,7 +88,7 @@ cglobal %1_h264_chroma_mc8_10, 6,7,8
     SPLATW        m5, m5        ; mm5 = B = x
     psubw         m4, m5        ; mm4 = A = 8-x
 
-.next1drow
+.next1drow:
     movu          m0, [r1   ]   ; mm0 = src[0..7]
     movu          m2, [r1+r6]   ; mm2 = src[1..8]
 
@@ -107,7 +107,7 @@ cglobal %1_h264_chroma_mc8_10, 6,7,8
     jne .next1drow
     REP_RET
 
-.xy_interpolation ; general case, bilinear
+.xy_interpolation: ; general case, bilinear
     movd          m4, r4m         ; x
     movd          m6, r5m         ; y
 
@@ -125,7 +125,7 @@ cglobal %1_h264_chroma_mc8_10, 6,7,8
 
     movu          m0, [r1  ]      ; mm0 = src[0..7]
     movu          m1, [r1+2]      ; mm1 = src[1..8]
-.next2drow
+.next2drow:
     add           r1, r2
 
     pmullw        m2, m0, m4
@@ -192,7 +192,7 @@ cglobal %1_h264_chroma_mc4_10, 6,6,7
     pmullw        m6, m2
     paddw         m6, m0
 
-.next2rows
+.next2rows:
     MC4_OP m0, m6
     MC4_OP m6, m0
     sub   r3d, 2
@@ -221,7 +221,7 @@ cglobal %1_h264_chroma_mc2_10, 6,7
     pxor          m7, m7
     pshufw        m2, [r1], 0x94    ; mm0 = src[0,1,1,2]
 
-.nextrow
+.nextrow:
     add           r1, r2
     movq          m1, m2
     pmaddwd       m1, m5          ; mm1 = A * src[0,1] + B * src[1,2]

libavcodec/x86/h264_deblock_10bit.asm

@@ -625,7 +625,7 @@ cglobal deblock_v_luma_intra_10, 4,7,16
     shl    r2d, 2
     shl    r3d, 2
     LOAD_AB aa, bb, r2d, r3d
-.loop
+.loop:
     mova    p2, [r4+r1]
     mova    p1, [r4+2*r1]
     mova    p0, [r4+r5]
@@ -676,7 +676,7 @@ cglobal deblock_h_luma_intra_10, 4,7,16
     mova    m0, [pw_2]
     shl    r2d, 2
     shl    r3d, 2
-.loop
+.loop:
     movu    q3, [r0-8]
     movu    q2, [r0+r1-8]
     movu    q1, [r0+r1*2-8]