summaryrefslogtreecommitdiff
path: root/libavutil/mips
diff options
context:
space:
mode:
authorJames Cowgill <james410@cowgill.org.uk>2015-03-05 17:40:15 +0000
committerMichael Niedermayer <michaelni@gmx.at>2015-03-06 14:37:44 +0100
commit157d6f0d5b95261bf053566d8a1c2652387da30c (patch)
treeca481834165a1f3c103eea9dc50fd5e877bbcbde /libavutil/mips
parenteae13eae9d6499e81cd77bcd9e13f7bcb31b5577 (diff)
downloadffmpeg-157d6f0d5b95261bf053566d8a1c2652387da30c.tar.gz
mips: port optimizations to mips n64
This mainly consists of replacing all the pointer arithmatic 'addiu' instructions with PTR_ADDIU which will handle the differences in pointer sizes when compiled on 64 bit mips systems. The header asmdefs.h contains the PTR_ macros which expend to the correct mips instructions to manipulate registers containing pointers. Signed-off-by: James Cowgill <james410@cowgill.org.uk> Reviewed-by: Nedeljko Babic <Nedeljko.Babic@imgtec.com> Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
Diffstat (limited to 'libavutil/mips')
-rw-r--r--libavutil/mips/asmdefs.h48
-rw-r--r--libavutil/mips/float_dsp_mips.c21
2 files changed, 59 insertions, 10 deletions
diff --git a/libavutil/mips/asmdefs.h b/libavutil/mips/asmdefs.h
new file mode 100644
index 0000000000..4d2922c951
--- /dev/null
+++ b/libavutil/mips/asmdefs.h
@@ -0,0 +1,48 @@
+/*
+ * Copyright (c) 2015 Imagination Technologies Ltd
+ *
+ * This file is part of FFmpeg.
+ *
+ * FFmpeg is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU Lesser General Public
+ * License as published by the Free Software Foundation; either
+ * version 2.1 of the License, or (at your option) any later version.
+ *
+ * FFmpeg is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
+ * Lesser General Public License for more details.
+ *
+ * You should have received a copy of the GNU Lesser General Public
+ * License along with FFmpeg; if not, write to the Free Software
+ * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
+ */
+
+/**
+ * @file
+ * MIPS assembly defines from sys/asm.h but rewritten for use with C inline
+ * assembly (rather than from within .s files).
+ */
+
+#ifndef AVCODEC_MIPS_ASMDEFS_H
+#define AVCODEC_MIPS_ASMDEFS_H
+
+#include <sgidefs.h>
+
+#if _MIPS_SIM == _ABI64
+# define PTRSIZE " 8 "
+# define PTRLOG " 3 "
+# define PTR_ADDU "daddu "
+# define PTR_ADDIU "daddiu "
+# define PTR_SUBU "dsubu "
+# define PTR_L "ld "
+#else
+# define PTRSIZE " 4 "
+# define PTRLOG " 2 "
+# define PTR_ADDU "addu "
+# define PTR_ADDIU "addiu "
+# define PTR_SUBU "subu "
+# define PTR_L "lw "
+#endif
+
+#endif /* AVCODEC_MIPS_ASMDEFS_H */
diff --git a/libavutil/mips/float_dsp_mips.c b/libavutil/mips/float_dsp_mips.c
index a54f959329..a45568778a 100644
--- a/libavutil/mips/float_dsp_mips.c
+++ b/libavutil/mips/float_dsp_mips.c
@@ -53,6 +53,7 @@
#include "config.h"
#include "libavutil/float_dsp.h"
+#include "libavutil/mips/asmdefs.h"
#if HAVE_INLINE_ASM && HAVE_MIPSFPU
static void vector_fmul_mips(float *dst, const float *src0, const float *src1,
@@ -90,9 +91,9 @@ static void vector_fmul_mips(float *dst, const float *src0, const float *src1,
"swc1 %[src0_1], 4(%[d]) \n\t"
"swc1 %[src0_2], 8(%[d]) \n\t"
"swc1 %[src0_3], 12(%[d]) \n\t"
- "addiu %[s0], %[s0], 16 \n\t"
- "addiu %[s1], %[s1], 16 \n\t"
- "addiu %[d], %[d], 16 \n\t"
+ PTR_ADDIU "%[s0], %[s0], 16 \n\t"
+ PTR_ADDIU "%[s1], %[s1], 16 \n\t"
+ PTR_ADDIU "%[d], %[d], 16 \n\t"
"bne %[d], %[d_end], 1b \n\t"
: [src0_0]"=&f"(src0_0), [src0_1]"=&f"(src0_1),
@@ -122,12 +123,12 @@ static void vector_fmul_scalar_mips(float *dst, const float *src, float mul,
"lwc1 %[temp1], 4(%[src]) \n\t"
"lwc1 %[temp2], 8(%[src]) \n\t"
"lwc1 %[temp3], 12(%[src]) \n\t"
- "addiu %[dst], %[dst], 16 \n\t"
+ PTR_ADDIU "%[dst], %[dst], 16 \n\t"
"mul.s %[temp0], %[temp0], %[mul] \n\t"
"mul.s %[temp1], %[temp1], %[mul] \n\t"
"mul.s %[temp2], %[temp2], %[mul] \n\t"
"mul.s %[temp3], %[temp3], %[mul] \n\t"
- "addiu %[src], %[src], 16 \n\t"
+ PTR_ADDIU "%[src], %[src], 16 \n\t"
"swc1 %[temp0], -16(%[dst]) \n\t"
"swc1 %[temp1], -12(%[dst]) \n\t"
"swc1 %[temp2], -8(%[dst]) \n\t"
@@ -251,8 +252,8 @@ static void butterflies_float_mips(float *av_restrict v1, float *av_restrict v2,
"add.s %[temp13], %[temp2], %[temp6] \n\t"
"sub.s %[temp14], %[temp3], %[temp7] \n\t"
"add.s %[temp15], %[temp3], %[temp7] \n\t"
- "addiu %[v1], %[v1], 16 \n\t"
- "addiu %[v2], %[v2], 16 \n\t"
+ PTR_ADDIU "%[v1], %[v1], 16 \n\t"
+ PTR_ADDIU "%[v2], %[v2], 16 \n\t"
"addiu %[pom], %[pom], -1 \n\t"
"lwc1 %[temp0], 0(%[v1]) \n\t"
"lwc1 %[temp1], 4(%[v1]) \n\t"
@@ -321,9 +322,9 @@ static void vector_fmul_reverse_mips(float *dst, const float *src0, const float
"mul.s %[temp2], %[temp3], %[temp2] \n\t"
"mul.s %[temp4], %[temp5], %[temp4] \n\t"
"mul.s %[temp6], %[temp7], %[temp6] \n\t"
- "addiu %[src0], %[src0], 16 \n\t"
- "addiu %[src1], %[src1], -16 \n\t"
- "addiu %[dst], %[dst], 16 \n\t"
+ PTR_ADDIU "%[src0], %[src0], 16 \n\t"
+ PTR_ADDIU "%[src1], %[src1], -16 \n\t"
+ PTR_ADDIU "%[dst], %[dst], 16 \n\t"
"swc1 %[temp0], -16(%[dst]) \n\t"
"swc1 %[temp2], -12(%[dst]) \n\t"
"swc1 %[temp4], -8(%[dst]) \n\t"