changeset 19373:87dddfc314fd

Update for changes from r19378 (ASMALIGN macro).
author diego
date Sun, 13 Aug 2006 00:44:42 +0000
parents 6334c14b38eb
children acb26d9717da
files liba52/liba52_changes.diff
diffstat 1 files changed, 50 insertions(+), 55 deletions(-) [+]
line wrap: on
line diff
--- a/liba52/liba52_changes.diff	Sun Aug 13 00:21:14 2006 +0000
+++ b/liba52/liba52_changes.diff	Sun Aug 13 00:44:42 2006 +0000
@@ -210,7 +210,7 @@
  }
 --- liba52/downmix.c	2006-06-12 15:17:53.000000000 +0200
 +++ liba52/downmix.c	2006-06-05 02:23:02.000000000 +0200
-@@ -19,18 +23,47 @@
+@@ -19,18 +23,46 @@
   * You should have received a copy of the GNU General Public License
   * along with this program; if not, write to the Free Software
   * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307  USA
@@ -219,7 +219,6 @@
   */
  
  #include "config.h"
-+#include "asmalign.h"
  
  #include <string.h>
  #include <inttypes.h>
@@ -258,7 +257,7 @@
  int a52_downmix_init (int input, int flags, sample_t * level,
  		      sample_t clev, sample_t slev)
  {
-@@ -447,7 +480,7 @@
+@@ -447,7 +479,7 @@
  	samples[i] = 0;
  }
  
@@ -267,7 +266,7 @@
  		  sample_t clev, sample_t slev)
  {
      switch (CONVERT (acmod, output & A52_CHANNEL_MASK)) {
-@@ -559,7 +592,7 @@
+@@ -559,7 +591,7 @@
  	break;
  
      case CONVERT (A52_3F2R, A52_2F1R):
@@ -276,7 +275,7 @@
  	move2to1 (samples + 768, samples + 512, bias);
  	break;
  
-@@ -583,12 +616,12 @@
+@@ -583,12 +615,12 @@
  	break;
  
      case CONVERT (A52_3F1R, A52_3F2R):
@@ -291,7 +290,7 @@
  {
      switch (CONVERT (acmod, output & A52_CHANNEL_MASK)) {
  
-@@ -653,3 +686,1137 @@
+@@ -653,3 +685,1137 @@
  	goto mix_31to21;
      }
  }
@@ -303,7 +302,7 @@
 +	"movlps %2, %%xmm7		\n\t"
 +	"shufps $0x00, %%xmm7, %%xmm7	\n\t"
 +	"mov $-1024, %%"REG_S"		\n\t"
-+	ASMALIGN16
++	ASMALIGN(4)
 +	"1:				\n\t"
 +	"movaps (%0, %%"REG_S"), %%xmm0	\n\t" 
 +	"movaps 16(%0, %%"REG_S"), %%xmm1\n\t" 
@@ -326,7 +325,7 @@
 +	"movlps %1, %%xmm7		\n\t"
 +	"shufps $0x00, %%xmm7, %%xmm7	\n\t"
 +	"mov $-1024, %%"REG_S"		\n\t"
-+	ASMALIGN16
++	ASMALIGN(4)
 +	"1:				\n\t"
 +	"movaps (%0, %%"REG_S"), %%xmm0	\n\t" 
 +	"movaps 1024(%0, %%"REG_S"), %%xmm1\n\t" 
@@ -347,7 +346,7 @@
 +	"movlps %1, %%xmm7		\n\t"
 +	"shufps $0x00, %%xmm7, %%xmm7	\n\t"
 +	"mov $-1024, %%"REG_S"		\n\t"
-+	ASMALIGN16
++	ASMALIGN(4)
 +	"1:				\n\t"
 +	"movaps (%0, %%"REG_S"), %%xmm0	\n\t" 
 +	"movaps 1024(%0, %%"REG_S"), %%xmm1\n\t" 
@@ -369,7 +368,7 @@
 +	"movlps %1, %%xmm7		\n\t"
 +	"shufps $0x00, %%xmm7, %%xmm7	\n\t"
 +	"mov $-1024, %%"REG_S"		\n\t"
-+	ASMALIGN16
++	ASMALIGN(4)
 +	"1:				\n\t"
 +	"movaps (%0, %%"REG_S"), %%xmm0	\n\t" 
 +	"movaps 1024(%0, %%"REG_S"), %%xmm1\n\t" 
@@ -392,7 +391,7 @@
 +	"movlps %1, %%xmm7		\n\t"
 +	"shufps $0x00, %%xmm7, %%xmm7	\n\t"
 +	"mov $-1024, %%"REG_S"		\n\t"
-+	ASMALIGN16
++	ASMALIGN(4)
 +	"1:				\n\t"
 +	"movaps 1024(%0, %%"REG_S"), %%xmm0\n\t" 
 +	"addps %%xmm7, %%xmm0		\n\t" //common
@@ -415,7 +414,7 @@
 +		"movlps %2, %%xmm7		\n\t"
 +		"shufps $0x00, %%xmm7, %%xmm7	\n\t"
 +		"mov $-1024, %%"REG_S"		\n\t"
-+		ASMALIGN16
++		ASMALIGN(4)
 +		"1:				\n\t"
 +		"movaps 1024(%1, %%"REG_S"), %%xmm0\n\t" 
 +		"addps %%xmm7, %%xmm0		\n\t" //common
@@ -438,7 +437,7 @@
 +		"movlps %1, %%xmm7		\n\t"
 +		"shufps $0x00, %%xmm7, %%xmm7	\n\t"
 +		"mov $-1024, %%"REG_S"		\n\t"
-+		ASMALIGN16
++		ASMALIGN(4)
 +		"1:				\n\t"
 +		"movaps 2048(%0, %%"REG_S"), %%xmm0\n\t"  // surround
 +		"movaps (%0, %%"REG_S"), %%xmm1	\n\t" 
@@ -462,7 +461,7 @@
 +		"movlps %1, %%xmm7		\n\t"
 +		"shufps $0x00, %%xmm7, %%xmm7	\n\t"
 +		"mov $-1024, %%"REG_S"		\n\t"
-+		ASMALIGN16
++		ASMALIGN(4)
 +		"1:				\n\t"
 +		"movaps 1024(%0, %%"REG_S"), %%xmm0\n\t"  
 +		"addps 3072(%0, %%"REG_S"), %%xmm0\n\t"  
@@ -486,7 +485,7 @@
 +		"movlps %1, %%xmm7		\n\t"
 +		"shufps $0x00, %%xmm7, %%xmm7	\n\t"
 +		"mov $-1024, %%"REG_S"		\n\t"
-+		ASMALIGN16
++		ASMALIGN(4)
 +		"1:				\n\t"
 +		"movaps 1024(%0, %%"REG_S"), %%xmm0\n\t"  
 +		"movaps 3072(%0, %%"REG_S"), %%xmm3\n\t" // surround
@@ -512,7 +511,7 @@
 +		"movlps %1, %%xmm7		\n\t"
 +		"shufps $0x00, %%xmm7, %%xmm7	\n\t"
 +		"mov $-1024, %%"REG_S"		\n\t"
-+		ASMALIGN16
++		ASMALIGN(4)
 +		"1:				\n\t"
 +		"movaps 2048(%0, %%"REG_S"), %%xmm0\n\t"  
 +		"addps 3072(%0, %%"REG_S"), %%xmm0\n\t" // surround
@@ -537,7 +536,7 @@
 +	"movlps %1, %%xmm7		\n\t"
 +	"shufps $0x00, %%xmm7, %%xmm7	\n\t"
 +	"mov $-1024, %%"REG_S"		\n\t"
-+	ASMALIGN16
++	ASMALIGN(4)
 +	"1:				\n\t"
 +	"movaps 1024(%0, %%"REG_S"), %%xmm0\n\t" 
 +	"addps %%xmm7, %%xmm0		\n\t" // common
@@ -561,7 +560,7 @@
 +	"movlps %1, %%xmm7		\n\t"
 +	"shufps $0x00, %%xmm7, %%xmm7	\n\t"
 +	"mov $-1024, %%"REG_S"		\n\t"
-+	ASMALIGN16
++	ASMALIGN(4)
 +	"1:				\n\t"
 +	"movaps 1024(%0, %%"REG_S"), %%xmm0\n\t" 
 +	"movaps 3072(%0, %%"REG_S"), %%xmm2\n\t" 
@@ -588,7 +587,7 @@
 +		"movlps %2, %%xmm7		\n\t"
 +		"shufps $0x00, %%xmm7, %%xmm7	\n\t"
 +		"mov $-1024, %%"REG_S"		\n\t"
-+		ASMALIGN16
++		ASMALIGN(4)
 +		"1:				\n\t"
 +		"movaps (%0, %%"REG_S"), %%xmm0	\n\t"  
 +		"movaps 16(%0, %%"REG_S"), %%xmm1\n\t"  
@@ -610,7 +609,7 @@
 +	asm volatile(
 +		"mov $-1024, %%"REG_S"		\n\t"
 +		"pxor %%mm0, %%mm0		\n\t"
-+		ASMALIGN16
++		ASMALIGN(4)
 +		"1:				\n\t"
 +		"movq %%mm0, (%0, %%"REG_S")	\n\t"
 +		"movq %%mm0, 8(%0, %%"REG_S")	\n\t"
@@ -870,7 +869,7 @@
 +	"movd  %2, %%mm7	\n\t"
 +	"punpckldq %2, %%mm7	\n\t"
 +	"mov $-1024, %%"REG_S"	\n\t"
-+	ASMALIGN16
++	ASMALIGN(4)
 +	"1:			\n\t"
 +	"movq  (%0, %%"REG_S"), %%mm0	\n\t" 
 +	"movq  8(%0, %%"REG_S"), %%mm1	\n\t"
@@ -901,7 +900,7 @@
 +	"movd  %1, %%mm7	\n\t"
 +	"punpckldq %1, %%mm7	\n\t"
 +	"mov $-1024, %%"REG_S"	\n\t"
-+	ASMALIGN16
++	ASMALIGN(4)
 +	"1:			\n\t"
 +	"movq  (%0, %%"REG_S"), %%mm0	\n\t" 
 +	"movq  8(%0, %%"REG_S"), %%mm1	\n\t"
@@ -928,7 +927,7 @@
 +	"movd  %1, %%mm7	\n\t"
 +	"punpckldq %1, %%mm7	\n\t"
 +	"mov $-1024, %%"REG_S"	\n\t"
-+	ASMALIGN16
++	ASMALIGN(4)
 +	"1:			\n\t"
 +	"movq  (%0, %%"REG_S"), %%mm0	\n\t" 
 +	"movq  8(%0, %%"REG_S"), %%mm1	\n\t"
@@ -957,7 +956,7 @@
 +	"movd  %1, %%mm7	\n\t"
 +	"punpckldq %1, %%mm7	\n\t"
 +	"mov $-1024, %%"REG_S"	\n\t"
-+	ASMALIGN16
++	ASMALIGN(4)
 +	"1:			\n\t"
 +	"movq  (%0, %%"REG_S"), %%mm0	\n\t" 
 +	"movq  8(%0, %%"REG_S"), %%mm1	\n\t"
@@ -988,7 +987,7 @@
 +	"movd  %1, %%mm7	\n\t"
 +	"punpckldq %1, %%mm7	\n\t"
 +	"mov $-1024, %%"REG_S"	\n\t"
-+	ASMALIGN16
++	ASMALIGN(4)
 +	"1:			\n\t"
 +	"movq   1024(%0, %%"REG_S"), %%mm0\n\t" 
 +	"movq   1032(%0, %%"REG_S"), %%mm1\n\t"
@@ -1019,7 +1018,7 @@
 +		"movd  %2, %%mm7	\n\t"
 +		"punpckldq %2, %%mm7	\n\t"
 +		"mov $-1024, %%"REG_S"	\n\t"
-+		ASMALIGN16
++		ASMALIGN(4)
 +		"1:			\n\t"
 +		"movq  1024(%1, %%"REG_S"), %%mm0\n\t" 
 +		"movq  1032(%1, %%"REG_S"), %%mm1\n\t"
@@ -1050,7 +1049,7 @@
 +		"movd  %1, %%mm7	\n\t"
 +		"punpckldq %1, %%mm7	\n\t"
 +		"mov $-1024, %%"REG_S"	\n\t"
-+		ASMALIGN16
++		ASMALIGN(4)
 +		"1:			\n\t"
 +		"movq  2048(%0, %%"REG_S"), %%mm0\n\t"  // surround
 +		"movq  2056(%0, %%"REG_S"), %%mm1\n\t"  // surround
@@ -1083,7 +1082,7 @@
 +		"movd  %1, %%mm7	\n\t"
 +		"punpckldq %1, %%mm7	\n\t"
 +		"mov $-1024, %%"REG_S"	\n\t"
-+		ASMALIGN16
++		ASMALIGN(4)
 +		"1:			\n\t"
 +		"movq  1024(%0, %%"REG_S"), %%mm0\n\t"  
 +		"movq  1032(%0, %%"REG_S"), %%mm1\n\t"
@@ -1116,7 +1115,7 @@
 +		"movd  %1, %%mm7	\n\t"
 +		"punpckldq %1, %%mm7	\n\t"
 +		"mov $-1024, %%"REG_S"	\n\t"
-+		ASMALIGN16
++		ASMALIGN(4)
 +		"1:			\n\t"
 +		"movq   1024(%0, %%"REG_S"), %%mm0\n\t"  
 +		"movq   1032(%0, %%"REG_S"), %%mm1\n\t"
@@ -1153,7 +1152,7 @@
 +		"movd  %1, %%mm7	\n\t"
 +		"punpckldq %1, %%mm7	\n\t"
 +		"mov $-1024, %%"REG_S"	\n\t"
-+		ASMALIGN16
++		ASMALIGN(4)
 +		"1:			\n\t"
 +		"movq  2048(%0, %%"REG_S"), %%mm0\n\t"  
 +		"movq  2056(%0, %%"REG_S"), %%mm1\n\t"
@@ -1188,7 +1187,7 @@
 +	"movd  %1, %%mm7	\n\t"
 +	"punpckldq %1, %%mm7	\n\t"
 +	"mov $-1024, %%"REG_S"	\n\t"
-+	ASMALIGN16
++	ASMALIGN(4)
 +	"1:			\n\t"
 +	"movq   1024(%0, %%"REG_S"), %%mm0\n\t" 
 +	"movq   1032(%0, %%"REG_S"), %%mm1\n\t"
@@ -1220,7 +1219,7 @@
 +{
 +	asm volatile(
 +	"mov $-1024, %%"REG_S"		\n\t"
-+	ASMALIGN16
++	ASMALIGN(4)
 +	"1:			\n\t"
 +	"movd  %1, %%mm7		\n\t"
 +	"punpckldq %1, %%mm7		\n\t"
@@ -1261,7 +1260,7 @@
 +		"movd  %2, %%mm7	\n\t"
 +		"punpckldq %2, %%mm7	\n\t"
 +		"mov $-1024, %%"REG_S"	\n\t"
-+		ASMALIGN16
++		ASMALIGN(4)
 +		"1:			\n\t"
 +		"movq  (%0, %%"REG_S"), %%mm0	\n\t"  
 +		"movq  8(%0, %%"REG_S"), %%mm1	\n\t"
@@ -1431,7 +1430,7 @@
 +#endif // ARCH_X86 || ARCH_X86_64
 --- liba52/imdct.c	2006-06-12 15:18:27.000000000 +0200
 +++ liba52/imdct.c	2006-06-12 19:18:39.000000000 +0200
-@@ -22,9 +26,15 @@
+@@ -26,6 +26,11 @@
   * You should have received a copy of the GNU General Public License
   * along with this program; if not, write to the Free Software
   * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307  USA
@@ -1443,11 +1442,7 @@
   */
  
  #include "config.h"
-+#include "asmalign.h"
- 
- #include <math.h>
- #include <stdio.h>
-@@ -39,12 +49,49 @@
+@@ -39,12 +48,49 @@
  #include "a52.h"
  #include "a52_internal.h"
  #include "mm_accel.h"
@@ -1497,7 +1492,7 @@
  static uint8_t fftorder[] = {
        0,128, 64,192, 32,160,224, 96, 16,144, 80,208,240,112, 48,176,
        8,136, 72,200, 40,168,232,104,248,120, 56,184, 24,152,216, 88,
-@@ -56,6 +103,40 @@
+@@ -56,6 +102,40 @@
        6,134, 70,198, 38,166,230,102,246,118, 54,182, 22,150,214, 86
  };
  
@@ -1538,7 +1533,7 @@
  /* Root values for IFFT */
  static sample_t roots16[3];
  static sample_t roots32[7];
-@@ -241,7 +322,7 @@
+@@ -241,7 +321,7 @@
      ifft_pass (buf, roots128 - 32, 32);
  }
  
@@ -1547,7 +1542,7 @@
  {
      int i, k;
      sample_t t_r, t_i, a_r, a_i, b_r, b_i, w_1, w_2;
-@@ -285,6 +366,714 @@
+@@ -285,6 +365,714 @@
      }
  }
  
@@ -1954,7 +1949,7 @@
 +		"lea "MANGLE(bit_reverse_512)", %%"REG_a"\n\t"
 +		"mov $1008, %%"REG_D"			\n\t"
 +		"push %%"REG_BP"			\n\t" //use ebp without telling gcc
-+		ASMALIGN16
++		ASMALIGN(4)
 +		"1:					\n\t"
 +		"movlps (%0, %%"REG_S"), %%xmm0	\n\t" // XXXI
 +		"movhps 8(%0, %%"REG_D"), %%xmm0	\n\t" // RXXI
@@ -2013,7 +2008,7 @@
 +		"xorps %%xmm1, %%xmm1	\n\t"
 +		"xorps %%xmm2, %%xmm2	\n\t"
 +		"mov %0, %%"REG_S"	\n\t"
-+		ASMALIGN16
++		ASMALIGN(4)
 +		"1:			\n\t"
 +		"movlps (%%"REG_S"), %%xmm0\n\t" //buf[p]
 +		"movlps 8(%%"REG_S"), %%xmm1\n\t" //buf[q]
@@ -2034,7 +2029,7 @@
 +	asm volatile(
 +		"movaps "MANGLE(ps111_1)", %%xmm7\n\t" // 1,1,1,-1
 +		"mov %0, %%"REG_S"		\n\t"
-+		ASMALIGN16
++		ASMALIGN(4)
 +		"1:				\n\t"
 +		"movaps 16(%%"REG_S"), %%xmm2	\n\t" //r2,i2,r3,i3
 +		"shufps $0xB4, %%xmm2, %%xmm2	\n\t" //r2,i2,i3,r3
@@ -2065,7 +2060,7 @@
 +		"xorps %%xmm5, %%xmm5		\n\t"
 +		"xorps %%xmm2, %%xmm2		\n\t"
 +		"mov %0, %%"REG_S"		\n\t"
-+		ASMALIGN16
++		ASMALIGN(4)
 +		"1:				\n\t"
 +		"movaps 32(%%"REG_S"), %%xmm2	\n\t" //r4,i4,r5,i5
 +		"movaps 48(%%"REG_S"), %%xmm3	\n\t" //r6,i6,r7,i7
@@ -2106,7 +2101,7 @@
 +	buf_offset = buf+128;
 +	asm volatile(
 +		"mov %0, %%"REG_S"			\n\t"
-+		ASMALIGN16
++		ASMALIGN(4)
 +		"1:					\n\t"
 +		"xor %%"REG_D", %%"REG_D"		\n\t" // k
 +		"lea (%%"REG_S", %3), %%"REG_d"		\n\t"
@@ -2138,7 +2133,7 @@
 +    /* Post IFFT complex multiply  plus IFFT complex conjugate*/
 +	asm volatile(
 +		"mov $-1024, %%"REG_S"			\n\t"
-+		ASMALIGN16
++		ASMALIGN(4)
 +		"1:					\n\t"
 +		"movaps (%0, %%"REG_S"), %%xmm0		\n\t"
 +		"movaps (%0, %%"REG_S"), %%xmm1		\n\t"
@@ -2164,7 +2159,7 @@
 +		"xor %%"REG_S", %%"REG_S"		\n\t"  // 0
 +		"movss %3, %%xmm2			\n\t"  // bias
 +		"shufps $0x00, %%xmm2, %%xmm2		\n\t"  // bias, bias, ...
-+		ASMALIGN16
++		ASMALIGN(4)
 +		"1:					\n\t"
 +		"movlps (%0, %%"REG_S"), %%xmm0		\n\t" // ? ? A ?
 +		"movlps 8(%0, %%"REG_S"), %%xmm1	\n\t" // ? ? C ?
@@ -2191,7 +2186,7 @@
 +		"xor %%"REG_S", %%"REG_S"		\n\t"  // 0
 +		"movss %3, %%xmm2			\n\t"  // bias
 +		"shufps $0x00, %%xmm2, %%xmm2		\n\t"  // bias, bias, ...
-+		ASMALIGN16
++		ASMALIGN(4)
 +		"1:					\n\t"
 +		"movlps (%0, %%"REG_S"), %%xmm0		\n\t" // ? ? ? A
 +		"movlps 8(%0, %%"REG_S"), %%xmm1	\n\t" // ? ? ? C
@@ -2218,7 +2213,7 @@
 +	asm volatile(
 +		"xor %%"REG_D", %%"REG_D"		\n\t"  // 0
 +		"xor %%"REG_S", %%"REG_S"		\n\t"  // 0
-+		ASMALIGN16
++		ASMALIGN(4)
 +		"1:					\n\t"
 +		"movlps (%0, %%"REG_S"), %%xmm0		\n\t" // ? ? ? A
 +		"movlps 8(%0, %%"REG_S"), %%xmm1	\n\t" // ? ? ? C
@@ -2240,7 +2235,7 @@
 +	asm volatile(
 +		"mov $1024, %%"REG_D"			\n\t"  // 1024
 +		"xor %%"REG_S", %%"REG_S"		\n\t"  // 0
-+		ASMALIGN16
++		ASMALIGN(4)
 +		"1:					\n\t"
 +		"movlps (%0, %%"REG_S"), %%xmm0	\n\t" // ? ? A ?
 +		"movlps 8(%0, %%"REG_S"), %%xmm1	\n\t" // ? ? C ?
@@ -2262,7 +2257,7 @@
  void a52_imdct_256(sample_t * data, sample_t * delay, sample_t bias)
  {
      int i, k;
-@@ -364,7 +1153,7 @@
+@@ -364,7 +1152,7 @@
  
  void a52_imdct_init (uint32_t mm_accel)
  {
@@ -2271,7 +2266,7 @@
      double sum;
  
      /* compute imdct window - kaiser-bessel derived window, alpha = 5.0 */
-@@ -416,6 +1205,99 @@
+@@ -416,6 +1204,99 @@
  	post2[i].real = cos ((M_PI / 128) * (i + 0.5));
  	post2[i].imag = sin ((M_PI / 128) * (i + 0.5));
      }
@@ -2371,7 +2366,7 @@
  
  #ifdef LIBA52_DJBFFT
      if (mm_accel & MM_ACCEL_DJBFFT) {
-@@ -426,7 +1308,5 @@
+@@ -426,7 +1307,5 @@
  #endif
      {
  	fprintf (stderr, "No accelerated IMDCT transform found\n");