summaryrefslogtreecommitdiff
diff options
context:
space:
mode:
Diffstat
-rw-r--r--configure.ac6
-rw-r--r--src/libmpeg2/idct_altivec.c502
-rw-r--r--src/libmpeg2/mpeg2_internal.h9
-rw-r--r--src/libmpeg2/slice.c10
-rw-r--r--src/libmpeg2/slice_xvmc.c12
5 files changed, 34 insertions, 505 deletions
diff --git a/configure.ac b/configure.ac
index b89ec065d..f384150ce 100644
--- a/configure.ac
+++ b/configure.ac
@@ -1989,6 +1989,8 @@ case "$host_or_hostalias" in
dnl _AM_DEPENDENCIES (below), and doing the AC_SUBST elicits a warning
dnl from automake 1.6.
+ AC_CHECK_HEADER([altivec.h], , enable_altivec=no)
+
if test x$enable_altivec = xyes; then
AC_DEFINE_UNQUOTED(ENABLE_ALTIVEC,,[Define this if you want to use altivec on PowerPC CPUs])
CFLAGS="$CFLAGS -faltivec"
@@ -2000,9 +2002,13 @@ case "$host_or_hostalias" in
AS="$CC"
AC_DEFINE_UNQUOTED(FPM_PPC,,[Define to select libmad fixed point arithmetic implementation])
AC_DEFINE_UNQUOTED(ARCH_PPC,,[Define this if you're running PowerPC architecture])
+
+ AC_CHECK_HEADER([altivec.h], , enable_altivec=no)
if test x$enable_altivec = xyes; then
AC_DEFINE_UNQUOTED(ENABLE_ALTIVEC,,[Define this if you have a Motorola 74xx CPU])
+ CFLAGS="$CFLAGS -maltivec"
+ LIBMPEG2_CFLAGS="$LIBMPEG2_CFLAGS -force_cpusubtype_ALL -maltivec"
fi
;;
sparc*-*-linux*)
diff --git a/src/libmpeg2/idct_altivec.c b/src/libmpeg2/idct_altivec.c
index 2ebdd3241..ed8b58cfd 100644
--- a/src/libmpeg2/idct_altivec.c
+++ b/src/libmpeg2/idct_altivec.c
@@ -25,493 +25,13 @@
#if defined (ARCH_PPC) && defined (ENABLE_ALTIVEC)
+#include <altivec.h>
+
#include <inttypes.h>
#include "mpeg2_internal.h"
#include "xineutils.h"
-#ifndef HOST_OS_DARWIN
-
-/*
- * The asm code is generated with:
- *
- * gcc-2.95 -fvec -D__ALTIVEC__ -O9 -fomit-frame-pointer -mregnames -S
- * idct_altivec.c
- *
- * awk '{args=""; len=split ($2, arg, ",");
- * for (i=1; i<=len; i++) { a=arg[i]; if (i<len) a=a",";
- * args = args sprintf ("%-6s", a) }
- * printf ("\t\"\t%-16s%-24s\\n\"\n", $1, args) }' idct_altivec.s |
- * unexpand -a
- *
- * I then do some simple trimming on the function prolog/trailers
- */
-
-void mpeg2_idct_copy_altivec (int16_t * block, uint8_t * dest, int stride)
-{
- asm (" \n"
- "# stwu %r1, -128(%r1) \n"
- "# mflr %r0 \n"
- "# stw %r0, 132(%r1) \n"
- "# addi %r0, %r1, 128 \n"
- "# bl _savev25 \n"
-
- " addi %r9, %r3, 112 \n"
- " vspltish %v25, 4 \n"
- " vxor %v13, %v13, %v13 \n"
- " lis %r10, constants@ha \n"
- " lvx %v1, 0, %r9 \n"
- " la %r10, constants@l(%r10) \n"
- " lvx %v5, 0, %r3 \n"
- " addi %r9, %r3, 16 \n"
- " lvx %v8, 0, %r10 \n"
- " addi %r11, %r10, 32 \n"
- " lvx %v12, 0, %r9 \n"
- " lvx %v6, 0, %r11 \n"
- " addi %r8, %r3, 48 \n"
- " vslh %v1, %v1, %v25 \n"
- " addi %r9, %r3, 80 \n"
- " lvx %v11, 0, %r8 \n"
- " vslh %v5, %v5, %v25 \n"
- " lvx %v0, 0, %r9 \n"
- " addi %r11, %r10, 64 \n"
- " vsplth %v3, %v8, 2 \n"
- " lvx %v7, 0, %r11 \n"
- " addi %r9, %r3, 96 \n"
- " vslh %v12, %v12, %v25 \n"
- " vmhraddshs %v27, %v1, %v6, %v13 \n"
- " addi %r8, %r3, 32 \n"
- " vsplth %v2, %v8, 5 \n"
- " lvx %v1, 0, %r9 \n"
- " vslh %v11, %v11, %v25 \n"
- " addi %r3, %r3, 64 \n"
- " lvx %v9, 0, %r8 \n"
- " addi %r9, %r10, 48 \n"
- " vslh %v0, %v0, %v25 \n"
- " lvx %v4, 0, %r9 \n"
- " vmhraddshs %v31, %v12, %v6, %v13 \n"
- " addi %r10, %r10, 16 \n"
- " vmhraddshs %v30, %v0, %v7, %v13 \n"
- " lvx %v10, 0, %r3 \n"
- " vsplth %v19, %v8, 3 \n"
- " vmhraddshs %v15, %v11, %v7, %v13 \n"
- " lvx %v12, 0, %r10 \n"
- " vsplth %v6, %v8, 4 \n"
- " vslh %v1, %v1, %v25 \n"
- " vsplth %v11, %v8, 1 \n"
- " li %r9, 4 \n"
- " vslh %v9, %v9, %v25 \n"
- " vsplth %v7, %v8, 0 \n"
- " vmhraddshs %v18, %v1, %v4, %v13 \n"
- " vspltw %v8, %v8, 3 \n"
- " vsubshs %v0, %v13, %v27 \n"
- " vmhraddshs %v1, %v9, %v4, %v13 \n"
- " vmhraddshs %v17, %v3, %v31, %v0 \n"
- " vmhraddshs %v4, %v2, %v15, %v30 \n"
- " vslh %v10, %v10, %v25 \n"
- " vmhraddshs %v9, %v5, %v12, %v13 \n"
- " vspltish %v25, 6 \n"
- " vmhraddshs %v5, %v10, %v12, %v13 \n"
- " vmhraddshs %v28, %v19, %v30, %v15 \n"
- " vmhraddshs %v27, %v3, %v27, %v31 \n"
- " vsubshs %v0, %v13, %v18 \n"
- " vmhraddshs %v18, %v11, %v18, %v1 \n"
- " vaddshs %v30, %v17, %v4 \n"
- " vmhraddshs %v12, %v11, %v1, %v0 \n"
- " vsubshs %v4, %v17, %v4 \n"
- " vaddshs %v10, %v9, %v5 \n"
- " vsubshs %v17, %v27, %v28 \n"
- " vaddshs %v27, %v27, %v28 \n"
- " vsubshs %v1, %v9, %v5 \n"
- " vaddshs %v28, %v10, %v18 \n"
- " vsubshs %v18, %v10, %v18 \n"
- " vaddshs %v10, %v1, %v12 \n"
- " vsubshs %v1, %v1, %v12 \n"
- " vsubshs %v12, %v17, %v4 \n"
- " vaddshs %v4, %v17, %v4 \n"
- " vmhraddshs %v5, %v7, %v12, %v1 \n"
- " vmhraddshs %v26, %v6, %v4, %v10 \n"
- " vmhraddshs %v29, %v6, %v12, %v1 \n"
- " vmhraddshs %v14, %v7, %v4, %v10 \n"
- " vsubshs %v12, %v18, %v30 \n"
- " vaddshs %v9, %v28, %v27 \n"
- " vaddshs %v16, %v18, %v30 \n"
- " vsubshs %v10, %v28, %v27 \n"
- " vmrglh %v31, %v9, %v12 \n"
- " vmrglh %v30, %v5, %v26 \n"
- " vmrglh %v15, %v14, %v29 \n"
- " vmrghh %v5, %v5, %v26 \n"
- " vmrglh %v27, %v16, %v10 \n"
- " vmrghh %v9, %v9, %v12 \n"
- " vmrghh %v18, %v16, %v10 \n"
- " vmrghh %v1, %v14, %v29 \n"
- " vmrglh %v14, %v9, %v5 \n"
- " vmrglh %v16, %v31, %v30 \n"
- " vmrglh %v10, %v15, %v27 \n"
- " vmrghh %v9, %v9, %v5 \n"
- " vmrghh %v26, %v15, %v27 \n"
- " vmrglh %v27, %v16, %v10 \n"
- " vmrghh %v12, %v1, %v18 \n"
- " vmrglh %v29, %v1, %v18 \n"
- " vsubshs %v0, %v13, %v27 \n"
- " vmrghh %v5, %v31, %v30 \n"
- " vmrglh %v31, %v9, %v12 \n"
- " vmrglh %v30, %v5, %v26 \n"
- " vmrglh %v15, %v14, %v29 \n"
- " vmhraddshs %v17, %v3, %v31, %v0 \n"
- " vmrghh %v18, %v16, %v10 \n"
- " vmhraddshs %v27, %v3, %v27, %v31 \n"
- " vmhraddshs %v4, %v2, %v15, %v30 \n"
- " vmrghh %v1, %v14, %v29 \n"
- " vmhraddshs %v28, %v19, %v30, %v15 \n"
- " vmrghh %v0, %v9, %v12 \n"
- " vsubshs %v13, %v13, %v18 \n"
- " vmrghh %v5, %v5, %v26 \n"
- " vmhraddshs %v18, %v11, %v18, %v1 \n"
- " vaddshs %v9, %v0, %v8 \n"
- " vaddshs %v30, %v17, %v4 \n"
- " vmhraddshs %v12, %v11, %v1, %v13 \n"
- " vsubshs %v4, %v17, %v4 \n"
- " vaddshs %v10, %v9, %v5 \n"
- " vsubshs %v17, %v27, %v28 \n"
- " vaddshs %v27, %v27, %v28 \n"
- " vsubshs %v1, %v9, %v5 \n"
- " vaddshs %v28, %v10, %v18 \n"
- " vsubshs %v18, %v10, %v18 \n"
- " vaddshs %v10, %v1, %v12 \n"
- " vsubshs %v1, %v1, %v12 \n"
- " vsubshs %v12, %v17, %v4 \n"
- " vaddshs %v4, %v17, %v4 \n"
- " vaddshs %v9, %v28, %v27 \n"
- " vmhraddshs %v14, %v7, %v4, %v10 \n"
- " vsrah %v9, %v9, %v25 \n"
- " vmhraddshs %v5, %v7, %v12, %v1 \n"
- " vpkshus %v0, %v9, %v9 \n"
- " vmhraddshs %v29, %v6, %v12, %v1 \n"
- " stvewx %v0, 0, %r4 \n"
- " vaddshs %v16, %v18, %v30 \n"
- " vsrah %v31, %v14, %v25 \n"
- " stvewx %v0, %r9, %r4 \n"
- " add %r4, %r4, %r5 \n"
- " vsrah %v15, %v16, %v25 \n"
- " vpkshus %v0, %v31, %v31 \n"
- " vsrah %v1, %v5, %v25 \n"
- " stvewx %v0, 0, %r4 \n"
- " vsubshs %v12, %v18, %v30 \n"
- " stvewx %v0, %r9, %r4 \n"
- " vmhraddshs %v26, %v6, %v4, %v10 \n"
- " vpkshus %v0, %v1, %v1 \n"
- " add %r4, %r4, %r5 \n"
- " vsrah %v5, %v12, %v25 \n"
- " stvewx %v0, 0, %r4 \n"
- " vsrah %v30, %v29, %v25 \n"
- " stvewx %v0, %r9, %r4 \n"
- " vsubshs %v10, %v28, %v27 \n"
- " vpkshus %v0, %v15, %v15 \n"
- " add %r4, %r4, %r5 \n"
- " stvewx %v0, 0, %r4 \n"
- " vsrah %v18, %v26, %v25 \n"
- " stvewx %v0, %r9, %r4 \n"
- " vsrah %v27, %v10, %v25 \n"
- " vpkshus %v0, %v5, %v5 \n"
- " add %r4, %r4, %r5 \n"
- " stvewx %v0, 0, %r4 \n"
- " stvewx %v0, %r9, %r4 \n"
- " vpkshus %v0, %v30, %v30 \n"
- " add %r4, %r4, %r5 \n"
- " stvewx %v0, 0, %r4 \n"
- " stvewx %v0, %r9, %r4 \n"
- " vpkshus %v0, %v18, %v18 \n"
- " add %r4, %r4, %r5 \n"
- " stvewx %v0, 0, %r4 \n"
- " stvewx %v0, %r9, %r4 \n"
- " add %r4, %r4, %r5 \n"
- " vpkshus %v0, %v27, %v27 \n"
- " stvewx %v0, 0, %r4 \n"
- " stvewx %v0, %r9, %r4 \n"
-
- "# addi %r0, %r1, 128 \n"
- "# bl _restv25 \n"
- "# lwz %r0, 132(%r1) \n"
- "# mtlr %r0 \n"
- "# la %r1, 128(%r1) \n"
- " vxor %v1, %v1, %v1 \n"
- " addi %r9, %r3, 16 \n"
- " stvx %v1, 0, %r3 \n"
- " stvx %v1, 0, %r9 \n"
- " addi %r11, %r3, 32 \n"
- " stvx %v1, 0, %r11 \n"
- " addi %r9, %r3, 48 \n"
- " stvx %v1, 0, %r9 \n"
- " addi %r11, %r3, -64 \n"
- " stvx %v1, 0, %r11 \n"
- " addi %r9, %r3, -48 \n"
- " stvx %v1, 0, %r9 \n"
- " addi %r11, %r3, -32 \n"
- " stvx %v1, 0, %r11 \n"
- " addi %r3, %r3, -16 \n"
- " stvx %v1, 0, %r3 \n"
- );
-}
-
-void mpeg2_idct_add_altivec (int16_t * block, uint8_t * dest, int stride)
-{
- asm (" \n"
- "# stwu %r1, -192(%r1) \n"
- "# mflr %r0 \n"
- "# stw %r0, 196(%r1) \n"
- "# addi %r0, %r1, 192 \n"
- "# bl _savev21 \n"
-
- " addi %r9, %r3, 112 \n"
- " vspltish %v21, 4 \n"
- " vxor %v1, %v1, %v1 \n"
- " lvx %v13, 0, %r9 \n"
- " lis %r10, constants@ha \n"
- " vspltisw %v3, -1 \n"
- " la %r10, constants@l(%r10) \n"
- " lvx %v5, 0, %r3 \n"
- " addi %r9, %r3, 16 \n"
- " lvx %v8, 0, %r10 \n"
- " lvx %v12, 0, %r9 \n"
- " addi %r11, %r10, 32 \n"
- " lvx %v6, 0, %r11 \n"
- " addi %r8, %r3, 48 \n"
- " vslh %v13, %v13, %v21 \n"
- " addi %r9, %r3, 80 \n"
- " lvx %v11, 0, %r8 \n"
- " vslh %v5, %v5, %v21 \n"
- " lvx %v0, 0, %r9 \n"
- " addi %r11, %r10, 64 \n"
- " vsplth %v2, %v8, 2 \n"
- " lvx %v7, 0, %r11 \n"
- " vslh %v12, %v12, %v21 \n"
- " addi %r9, %r3, 96 \n"
- " vmhraddshs %v24, %v13, %v6, %v1 \n"
- " addi %r8, %r3, 32 \n"
- " vsplth %v17, %v8, 5 \n"
- " lvx %v13, 0, %r9 \n"
- " vslh %v11, %v11, %v21 \n"
- " addi %r3, %r3, 64 \n"
- " lvx %v10, 0, %r8 \n"
- " vslh %v0, %v0, %v21 \n"
- " addi %r9, %r10, 48 \n"
- " vmhraddshs %v31, %v12, %v6, %v1 \n"
- " lvx %v4, 0, %r9 \n"
- " addi %r10, %r10, 16 \n"
- " vmhraddshs %v26, %v0, %v7, %v1 \n"
- " lvx %v9, 0, %r3 \n"
- " vsplth %v16, %v8, 3 \n"
- " vmhraddshs %v22, %v11, %v7, %v1 \n"
- " lvx %v6, 0, %r10 \n"
- " lvsl %v19, 0, %r4 \n"
- " vsubshs %v12, %v1, %v24 \n"
- " lvsl %v0, %r5, %r4 \n"
- " vsplth %v11, %v8, 1 \n"
- " vslh %v10, %v10, %v21 \n"
- " vmrghb %v19, %v3, %v19 \n"
- " lvx %v15, 0, %r4 \n"
- " vslh %v13, %v13, %v21 \n"
- " vmrghb %v3, %v3, %v0 \n"
- " li %r9, 4 \n"
- " vmhraddshs %v14, %v2, %v31, %v12 \n"
- " vsplth %v7, %v8, 0 \n"
- " vmhraddshs %v23, %v13, %v4, %v1 \n"
- " vsplth %v18, %v8, 4 \n"
- " vmhraddshs %v27, %v10, %v4, %v1 \n"
- " vspltw %v8, %v8, 3 \n"
- " vmhraddshs %v12, %v17, %v22, %v26 \n"
- " vperm %v15, %v15, %v1, %v19 \n"
- " vslh %v9, %v9, %v21 \n"
- " vmhraddshs %v10, %v5, %v6, %v1 \n"
- " vspltish %v21, 6 \n"
- " vmhraddshs %v30, %v9, %v6, %v1 \n"
- " vmhraddshs %v26, %v16, %v26, %v22 \n"
- " vmhraddshs %v24, %v2, %v24, %v31 \n"
- " vmhraddshs %v31, %v11, %v23, %v27 \n"
- " vsubshs %v0, %v1, %v23 \n"
- " vaddshs %v23, %v14, %v12 \n"
- " vmhraddshs %v9, %v11, %v27, %v0 \n"
- " vsubshs %v12, %v14, %v12 \n"
- " vaddshs %v6, %v10, %v30 \n"
- " vsubshs %v14, %v24, %v26 \n"
- " vaddshs %v24, %v24, %v26 \n"
- " vsubshs %v13, %v10, %v30 \n"
- " vaddshs %v26, %v6, %v31 \n"
- " vsubshs %v31, %v6, %v31 \n"
- " vaddshs %v6, %v13, %v9 \n"
- " vsubshs %v13, %v13, %v9 \n"
- " vsubshs %v9, %v14, %v12 \n"
- " vaddshs %v12, %v14, %v12 \n"
- " vmhraddshs %v30, %v7, %v9, %v13 \n"
- " vmhraddshs %v25, %v18, %v12, %v6 \n"
- " vmhraddshs %v28, %v18, %v9, %v13 \n"
- " vmhraddshs %v29, %v7, %v12, %v6 \n"
- " vaddshs %v10, %v26, %v24 \n"
- " vsubshs %v5, %v31, %v23 \n"
- " vsubshs %v13, %v26, %v24 \n"
- " vaddshs %v4, %v31, %v23 \n"
- " vmrglh %v26, %v30, %v25 \n"
- " vmrglh %v31, %v10, %v5 \n"
- " vmrglh %v22, %v29, %v28 \n"
- " vmrghh %v30, %v30, %v25 \n"
- " vmrglh %v24, %v4, %v13 \n"
- " vmrghh %v10, %v10, %v5 \n"
- " vmrghh %v23, %v4, %v13 \n"
- " vmrghh %v27, %v29, %v28 \n"
- " vmrglh %v29, %v10, %v30 \n"
- " vmrglh %v4, %v31, %v26 \n"
- " vmrglh %v13, %v22, %v24 \n"
- " vmrghh %v10, %v10, %v30 \n"
- " vmrghh %v25, %v22, %v24 \n"
- " vmrglh %v24, %v4, %v13 \n"
- " vmrghh %v5, %v27, %v23 \n"
- " vmrglh %v28, %v27, %v23 \n"
- " vsubshs %v0, %v1, %v24 \n"
- " vmrghh %v30, %v31, %v26 \n"
- " vmrglh %v31, %v10, %v5 \n"
- " vmrglh %v26, %v30, %v25 \n"
- " vmrglh %v22, %v29, %v28 \n"
- " vmhraddshs %v14, %v2, %v31, %v0 \n"
- " vmrghh %v23, %v4, %v13 \n"
- " vmhraddshs %v24, %v2, %v24, %v31 \n"
- " vmhraddshs %v12, %v17, %v22, %v26 \n"
- " vmrghh %v27, %v29, %v28 \n"
- " vmhraddshs %v26, %v16, %v26, %v22 \n"
- " vmrghh %v0, %v10, %v5 \n"
- " vmhraddshs %v31, %v11, %v23, %v27 \n"
- " vmrghh %v30, %v30, %v25 \n"
- " vsubshs %v13, %v1, %v23 \n"
- " vaddshs %v10, %v0, %v8 \n"
- " vaddshs %v23, %v14, %v12 \n"
- " vsubshs %v12, %v14, %v12 \n"
- " vaddshs %v6, %v10, %v30 \n"
- " vsubshs %v14, %v24, %v26 \n"
- " vmhraddshs %v9, %v11, %v27, %v13 \n"
- " vaddshs %v24, %v24, %v26 \n"
- " vaddshs %v26, %v6, %v31 \n"
- " vsubshs %v13, %v10, %v30 \n"
- " vaddshs %v10, %v26, %v24 \n"
- " vsubshs %v31, %v6, %v31 \n"
- " vaddshs %v6, %v13, %v9 \n"
- " vsrah %v10, %v10, %v21 \n"
- " vsubshs %v13, %v13, %v9 \n"
- " vaddshs %v0, %v15, %v10 \n"
- " vsubshs %v9, %v14, %v12 \n"
- " vaddshs %v12, %v14, %v12 \n"
- " vpkshus %v15, %v0, %v0 \n"
- " stvewx %v15, 0, %r4 \n"
- " vaddshs %v4, %v31, %v23 \n"
- " vmhraddshs %v29, %v7, %v12, %v6 \n"
- " stvewx %v15, %r9, %r4 \n"
- " add %r4, %r4, %r5 \n"
- " vsubshs %v5, %v31, %v23 \n"
- " lvx %v15, 0, %r4 \n"
- " vmhraddshs %v30, %v7, %v9, %v13 \n"
- " vsrah %v22, %v4, %v21 \n"
- " vperm %v15, %v15, %v1, %v3 \n"
- " vmhraddshs %v28, %v18, %v9, %v13 \n"
- " vsrah %v31, %v29, %v21 \n"
- " vsubshs %v13, %v26, %v24 \n"
- " vaddshs %v0, %v15, %v31 \n"
- " vsrah %v27, %v30, %v21 \n"
- " vpkshus %v15, %v0, %v0 \n"
- " vsrah %v30, %v5, %v21 \n"
- " stvewx %v15, 0, %r4 \n"
- " vsrah %v26, %v28, %v21 \n"
- " stvewx %v15, %r9, %r4 \n"
- " vmhraddshs %v25, %v18, %v12, %v6 \n"
- " add %r4, %r4, %r5 \n"
- " vsrah %v24, %v13, %v21 \n"
- " lvx %v15, 0, %r4 \n"
- " vperm %v15, %v15, %v1, %v19 \n"
- " vsrah %v23, %v25, %v21 \n"
- " vaddshs %v0, %v15, %v27 \n"
- " vpkshus %v15, %v0, %v0 \n"
- " stvewx %v15, 0, %r4 \n"
- " stvewx %v15, %r9, %r4 \n"
- " add %r4, %r4, %r5 \n"
- " lvx %v15, 0, %r4 \n"
- " vperm %v15, %v15, %v1, %v3 \n"
- " vaddshs %v0, %v15, %v22 \n"
- " vpkshus %v15, %v0, %v0 \n"
- " stvewx %v15, 0, %r4 \n"
- " stvewx %v15, %r9, %r4 \n"
- " add %r4, %r4, %r5 \n"
- " lvx %v15, 0, %r4 \n"
- " vperm %v15, %v15, %v1, %v19 \n"
- " vaddshs %v0, %v15, %v30 \n"
- " vpkshus %v15, %v0, %v0 \n"
- " stvewx %v15, 0, %r4 \n"
- " stvewx %v15, %r9, %r4 \n"
- " add %r4, %r4, %r5 \n"
- " lvx %v15, 0, %r4 \n"
- " vperm %v15, %v15, %v1, %v3 \n"
- " vaddshs %v0, %v15, %v26 \n"
- " vpkshus %v15, %v0, %v0 \n"
- " stvewx %v15, 0, %r4 \n"
- " stvewx %v15, %r9, %r4 \n"
- " add %r4, %r4, %r5 \n"
- " lvx %v15, 0, %r4 \n"
- " vperm %v15, %v15, %v1, %v19 \n"
- " vaddshs %v0, %v15, %v23 \n"
- " vpkshus %v15, %v0, %v0 \n"
- " stvewx %v15, 0, %r4 \n"
- " stvewx %v15, %r9, %r4 \n"
- " add %r4, %r4, %r5 \n"
- " lvx %v15, 0, %r4 \n"
- " vperm %v15, %v15, %v1, %v3 \n"
- " vaddshs %v0, %v15, %v24 \n"
- " vpkshus %v15, %v0, %v0 \n"
- " stvewx %v15, 0, %r4 \n"
- " stvewx %v15, %r9, %r4 \n"
-
- "# addi %r0, %r1, 192 \n"
- "# bl _restv21 \n"
- "# lwz %r0, 196(%r1) \n"
- "# mtlr %r0 \n"
- "# la %r1, 192(%r1) \n"
- " addi %r9, %r3, 16 \n"
- " stvx %v1, 0, %r3 \n"
- " stvx %v1, 0, %r9 \n"
- " addi %r11, %r3, 32 \n"
- " stvx %v1, 0, %r11 \n"
- " addi %r9, %r3, 48 \n"
- " stvx %v1, 0, %r9 \n"
- " addi %r11, %r3, -64 \n"
- " stvx %v1, 0, %r11 \n"
- " addi %r9, %r3, -48 \n"
- " stvx %v1, 0, %r9 \n"
- " addi %r11, %r3, -32 \n"
- " stvx %v1, 0, %r11 \n"
- " addi %r3, %r3, -16 \n"
- " stvx %v1, 0, %r3 \n"
- );
-}
-
-static const int16_t constants[5][8] ATTR_ALIGN(16) __attribute__((used)) = {
- {23170, 13573, 6518, 21895, -23170, -21895, 32, 31},
- {16384, 22725, 21407, 19266, 16384, 19266, 21407, 22725},
- {22725, 31521, 29692, 26722, 22725, 26722, 29692, 31521},
- {21407, 29692, 27969, 25172, 21407, 25172, 27969, 29692},
- {19266, 26722, 25172, 22654, 19266, 22654, 25172, 26722}
-};
-
-void mpeg2_idct_altivec_init (void)
-{
- int i, j;
-
- /* the altivec idct uses a transposed input, so we patch scan tables */
- for (i = 0; i < 64; i++) {
- j = mpeg2_scan_norm[i];
- mpeg2_scan_norm[i] = (j >> 3) | ((j & 7) << 3);
- j = mpeg2_scan_alt[i];
- mpeg2_scan_alt[i] = (j >> 3) | ((j & 7) << 3);
- }
-}
-
-#else /* HOST_OS_DARWIN */
-
#define vector_s16_t vector signed short
#define vector_u16_t vector unsigned short
#define vector_s8_t vector signed char
@@ -622,12 +142,18 @@ void mpeg2_idct_altivec_init (void)
vx6 = vec_sra (vy6, shift); \
vx7 = vec_sra (vy7, shift);
+#if defined( __APPLE_CC__ ) && defined( __APPLE_ALTIVEC__ ) /* apple */
+#define VEC_S16(a,b,c,d,e,f,g,h) (vector_s16_t) (a, b, c, d, e, f, g, h)
+#else /* gnu */
+#define VEC_S16(a,b,c,d,e,f,g,h) (vector_s16_t) {a, b, c, d, e, f, g, h}
+#endif
+
static vector_s16_t constants[5] = {
- (vector_s16_t)(23170, 13573, 6518, 21895, -23170, -21895, 32, 31),
- (vector_s16_t)(16384, 22725, 21407, 19266, 16384, 19266, 21407, 22725),
- (vector_s16_t)(22725, 31521, 29692, 26722, 22725, 26722, 29692, 31521),
- (vector_s16_t)(21407, 29692, 27969, 25172, 21407, 25172, 27969, 29692),
- (vector_s16_t)(19266, 26722, 25172, 22654, 19266, 22654, 25172, 26722)
+ VEC_S16(23170, 13573, 6518, 21895, -23170, -21895, 32, 31),
+ VEC_S16(16384, 22725, 21407, 19266, 16384, 19266, 21407, 22725),
+ VEC_S16(22725, 31521, 29692, 26722, 22725, 26722, 29692, 31521),
+ VEC_S16(21407, 29692, 27969, 25172, 21407, 25172, 27969, 29692),
+ VEC_S16(19266, 26722, 25172, 22654, 19266, 22654, 25172, 26722)
};
void mpeg2_idct_copy_altivec (vector_s16_t * block, unsigned char * dest,
@@ -703,7 +229,5 @@ void mpeg2_idct_altivec_init (void)
}
}
-#endif /* HOST_OS_DARWIN */
-
#endif /* ARCH_PPC && ENABLED_ALTIVEC */
diff --git a/src/libmpeg2/mpeg2_internal.h b/src/libmpeg2/mpeg2_internal.h
index 915bece2c..c2ffbf909 100644
--- a/src/libmpeg2/mpeg2_internal.h
+++ b/src/libmpeg2/mpeg2_internal.h
@@ -27,6 +27,10 @@
#include "video_out.h"
#include "accel_xvmc.h"
+#ifdef ENABLE_ALTIVEC
+#include <altivec.h>
+#endif
+
/* macroblock modes */
#define MACROBLOCK_INTRA XINE_MACROBLOCK_INTRA
#define MACROBLOCK_PATTERN XINE_MACROBLOCK_PATTERN
@@ -243,10 +247,6 @@ void mpeg2_zero_block_mmx (int16_t * block);
void mpeg2_idct_mmx_init (void);
/* idct_altivec.c */
-#ifndef HOST_OS_DARWIN
-void mpeg2_idct_copy_altivec (int16_t * block, uint8_t * dest, int stride);
-void mpeg2_idct_add_altivec (int16_t * block, uint8_t * dest, int stride);
-#else /* HOST_OS_DARWIN */
# ifdef ENABLE_ALTIVEC
void mpeg2_idct_copy_altivec (vector signed short * block, unsigned char * dest,
int stride);
@@ -258,7 +258,6 @@ void mpeg2_idct_copy_altivec (signed short * block, unsigned char * dest,
void mpeg2_idct_add_altivec (signed short * block, unsigned char * dest,
int stride);
# endif /* ENABLE_ALTIVEC */
-#endif /* HOST_OS_DARWIN */
void mpeg2_idct_altivec_init (void);
/* motion_comp.c */
diff --git a/src/libmpeg2/slice.c b/src/libmpeg2/slice.c
index 50d7b4caf..f214f1f25 100644
--- a/src/libmpeg2/slice.c
+++ b/src/libmpeg2/slice.c
@@ -205,7 +205,7 @@ static inline int get_motion_delta (picture_t * picture, int f_code)
#undef bit_ptr
}
-static inline int bound_motion_vector (int vector, int f_code)
+static inline int bound_motion_vector (int vec, int f_code)
{
#if 1
unsigned int limit;
@@ -213,11 +213,11 @@ static inline int bound_motion_vector (int vector, int f_code)
limit = 16 << f_code;
- if ((unsigned int)(vector + limit) < 2 * limit)
- return vector;
+ if ((unsigned int)(vec + limit) < 2 * limit)
+ return vec;
else {
- sign = ((int32_t)vector) >> 31;
- return vector - ((2 * limit) ^ sign) + sign;
+ sign = ((int32_t)vec) >> 31;
+ return vec - ((2 * limit) ^ sign) + sign;
}
#else
return ((int32_t)vector << (27 - f_code)) >> (27 - f_code);
diff --git a/src/libmpeg2/slice_xvmc.c b/src/libmpeg2/slice_xvmc.c
index fb0fab1e9..4aa383b9a 100644
--- a/src/libmpeg2/slice_xvmc.c
+++ b/src/libmpeg2/slice_xvmc.c
@@ -256,7 +256,7 @@ static inline int get_xvmc_motion_delta (picture_t * picture, int f_code)
#undef bit_ptr
}
-static inline int bound_motion_vector (int vector, int f_code)
+static inline int bound_motion_vector (int vec, int f_code)
{
#if 1
unsigned int limit;
@@ -264,14 +264,14 @@ static inline int bound_motion_vector (int vector, int f_code)
limit = 16 << f_code;
- if ((unsigned int)(vector + limit) < 2 * limit)
- return vector;
+ if ((unsigned int)(vec + limit) < 2 * limit)
+ return vec;
else {
- sign = ((int32_t)vector) >> 31;
- return vector - ((2 * limit) ^ sign) + sign;
+ sign = ((int32_t)vec) >> 31;
+ return vec - ((2 * limit) ^ sign) + sign;
}
#else
- return ((int32_t)vector << (27 - f_code)) >> (27 - f_code);
+ return ((int32_t)vec << (27 - f_code)) >> (27 - f_code);
#endif
}
generated by cgit v1.2.3 (git 2.25.1) at 2025-04-26 12:06:29 +0000