SHA256
1
0
forked from jengelh/ffmpeg-4
ffmpeg-4/ffmpeg4_swscale_fix_altivec_vsx_recent_gcc.patch

51 lines
2.1 KiB
Diff
Raw Normal View History

From daniel at octaforge.org Tue Aug 6 18:55:37 2019
From: daniel at octaforge.org (Daniel Kolesa)
Date: Tue, 6 Aug 2019 17:55:37 +0200
Subject: [FFmpeg-devel] [PATCH 2/2] swscale: Fix AltiVec/VSX build with
recent GCC
In-Reply-To: <20190806155537.25535-1-daniel@octaforge.org>
References: <20190806155537.25535-1-daniel@octaforge.org>
Message-ID: <20190806155537.25535-3-daniel@octaforge.org>
The argument to vec_splat_u16 must be a literal. By making the
function always inline and marking the arguments const, gcc can
turn those into literals, and avoid build errors like:
swscale_vsx.c:165:53: error: argument 1 must be a 5-bit signed literal
Signed-off-by: Daniel Kolesa <daniel at octaforge.org>
---
libswscale/ppc/swscale_vsx.c | 12 ++++++++----
1 file changed, 8 insertions(+), 4 deletions(-)
Index: ffmpeg-4.2/libswscale/ppc/swscale_vsx.c
===================================================================
--- ffmpeg-4.2.orig/libswscale/ppc/swscale_vsx.c
+++ ffmpeg-4.2/libswscale/ppc/swscale_vsx.c
@@ -154,8 +154,10 @@ static void yuv2plane1_nbps_u(const int1
}
}
-static void yuv2plane1_nbps_vsx(const int16_t *src, uint16_t *dest, int dstW,
- int big_endian, int output_bits)
+static av_always_inline void yuv2plane1_nbps_vsx(const int16_t *src,
+ uint16_t *dest, int dstW,
+ const int big_endian,
+ const int output_bits)
{
const int dst_u = -(uintptr_t)dest & 7;
const int shift = 15 - output_bits;
@@ -273,8 +275,10 @@ static void yuv2plane1_16_u(const int32_
}
}
-static void yuv2plane1_16_vsx(const int32_t *src, uint16_t *dest, int dstW,
- int big_endian, int output_bits)
+static av_always_inline void yuv2plane1_16_vsx(const int32_t *src,
+ uint16_t *dest, int dstW,
+ const int big_endian,
+ int output_bits)
{
const int dst_u = -(uintptr_t)dest & 7;
const int shift = 3;