ARM: SIMD assembly optimization for function BlitRGBtoRGBPixelAlpha

Much of the heavy lifting of this optimization is lifted from the Pixman
project, which is distributed under an MIT-style license. As far as possible,
these elements have been relicensed to the zlib license.
Ben Avison 2019-10-24 21:13:05 -04:00
parent 6a6a05289e
commit 57723b83e8
4 changed files with 1259 additions and 0 deletions

View File

@ -389,6 +389,23 @@ BlitRGBtoRGBPixelAlphaMMX(SDL_BlitInfo * info)
#endif /* __MMX__ */ #endif /* __MMX__ */
#if SDL_ARM_SIMD_BLITTERS
void BlitRGBtoRGBPixelAlphaARMSIMDAsm(int32_t w, int32_t h, uint32_t *dst, int32_t dst_stride, uint32_t *src, int32_t src_stride);
static void
BlitRGBtoRGBPixelAlphaARMSIMD(SDL_BlitInfo * info)
{
int32_t width = info->dst_w;
int32_t height = info->dst_h;
uint32_t *dstp = (uint32_t *)info->dst;
int32_t dststride = width + (info->dst_skip >> 2);
uint32_t *srcp = (uint32_t *)info->src;
int32_t srcstride = width + (info->src_skip >> 2);
BlitRGBtoRGBPixelAlphaARMSIMDAsm(width, height, dstp, dststride, srcp, srcstride);
}
#endif
/* fast RGB888->(A)RGB888 blending with surface alpha=128 special case */ /* fast RGB888->(A)RGB888 blending with surface alpha=128 special case */
static void static void
BlitRGBtoRGBSurfaceAlpha128(SDL_BlitInfo * info) BlitRGBtoRGBSurfaceAlpha128(SDL_BlitInfo * info)
@ -1315,6 +1332,10 @@ SDL_CalculateBlitA(SDL_Surface * surface)
} }
#endif /* __MMX__ || __3dNOW__ */ #endif /* __MMX__ || __3dNOW__ */
if (sf->Amask == 0xff000000) { if (sf->Amask == 0xff000000) {
#if SDL_ARM_SIMD_BLITTERS
if (SDL_HasARMSIMD())
return BlitRGBtoRGBPixelAlphaARMSIMD;
#endif
return BlitRGBtoRGBPixelAlpha; return BlitRGBtoRGBPixelAlpha;
} }
} }

View File

@ -0,0 +1,36 @@
/*
* Copyright © 2010 Nokia Corporation
*
* Permission to use, copy, modify, distribute, and sell this software and its
* documentation for any purpose is hereby granted without fee, provided that
* the above copyright notice appear in all copies and that both that
* copyright notice and this permission notice appear in supporting
* documentation, and that the name of Mozilla Corporation not be used in
* advertising or publicity pertaining to distribution of the software without
* specific, written prior permission. Mozilla Corporation makes no
* representations about the suitability of this software for any purpose. It
* is provided "as is" without express or implied warranty.
*
* THE COPYRIGHT HOLDERS DISCLAIM ALL WARRANTIES WITH REGARD TO THIS
* SOFTWARE, INCLUDING ALL IMPLIED WARRANTIES OF MERCHANTABILITY AND
* FITNESS, IN NO EVENT SHALL THE COPYRIGHT HOLDERS BE LIABLE FOR ANY
* SPECIAL, INDIRECT OR CONSEQUENTIAL DAMAGES OR ANY DAMAGES
* WHATSOEVER RESULTING FROM LOSS OF USE, DATA OR PROFITS, WHETHER IN
* AN ACTION OF CONTRACT, NEGLIGENCE OR OTHER TORTIOUS ACTION, ARISING
* OUT OF OR IN CONNECTION WITH THE USE OR PERFORMANCE OF THIS
* SOFTWARE.
*
* Author: Siarhei Siamashka (siarhei.siamashka@nokia.com)
*
*/
/* Supplementary macro for setting function attributes */
.macro pixman_asm_function fname
.func fname
.global fname
#ifdef __ELF__
.hidden fname
.type fname, %function
#endif
fname:
.endm

View File

@ -0,0 +1,168 @@
/*
* Copyright (c) 2016 RISC OS Open Ltd
*
* This software is provided 'as-is', without any express or implied
* warranty. In no event will the authors be held liable for any damages
* arising from the use of this software.
*
* Permission is granted to anyone to use this software for any purpose,
* including commercial applications, and to alter it and redistribute it
* freely, subject to the following restrictions:
*
* 1. The origin of this software must not be misrepresented; you must not
* claim that you wrote the original software. If you use this software
* in a product, an acknowledgment in the product documentation would be
* appreciated but is not required.
* 2. Altered source versions must be plainly marked as such, and must not be
* misrepresented as being the original software.
* 3. This notice may not be removed or altered from any source distribution.
*/
/* Prevent the stack from becoming executable */
#if defined(__linux__) && defined(__ELF__)
.section .note.GNU-stack,"",%progbits
#endif
.text
.arch armv6
.object_arch armv4
.arm
.altmacro
.p2align 2
#include "pixman-arm-asm.h"
#include "pixman-arm-simd-asm.h"
/* A head macro should do all processing which results in an output of up to
* 16 bytes, as far as the final load instruction. The corresponding tail macro
* should complete the processing of the up-to-16 bytes. The calling macro will
* sometimes choose to insert a preload or a decrement of X between them.
* cond ARM condition code for code block
* numbytes Number of output bytes that should be generated this time
* firstreg First WK register in which to place output
* unaligned_src Whether to use non-wordaligned loads of source image
* unaligned_mask Whether to use non-wordaligned loads of mask image
* preload If outputting 16 bytes causes 64 bytes to be read, whether an extra preload should be output
*/
/******************************************************************************/
/* This differs from the over_8888_8888 routine in Pixman in that the destination
* alpha component is always left unchanged, and RGB components are not
* premultiplied by alpha. It differs from BlitRGBtoRGBPixelAlpha in that
* renormalisation is done by multiplying by 257/256 (with rounding) rather than
* simply shifting right by 8 bits - removing the need to special-case alpha=0xff.
*/
.macro RGBtoRGBPixelAlpha_init
line_saved_regs STRIDE_S, ORIG_W
mov MASK, #0x80
.endm
.macro RGBtoRGBPixelAlpha_1pixel_translucent s, d, tmp0, tmp1, tmp2, tmp3, half
uxtb tmp3, s
uxtb tmp0, d
sub tmp0, tmp3, tmp0
uxtb tmp3, s, ror #16
uxtb tmp1, d, ror #16
sub tmp1, tmp3, tmp1
uxtb tmp3, s, ror #8
mov s, s, lsr #24
uxtb tmp2, d, ror #8
sub tmp2, tmp3, tmp2
smlabb tmp0, tmp0, s, half
smlabb tmp1, tmp1, s, half
smlabb tmp2, tmp2, s, half
add tmp0, tmp0, asr #8
add tmp1, tmp1, asr #8
add tmp2, tmp2, asr #8
pkhbt tmp0, tmp0, tmp1, lsl #16
and tmp2, tmp2, #0xff00
uxtb16 tmp0, tmp0, ror #8
orr tmp0, tmp0, tmp2
uadd8 d, d, tmp0
.endm
.macro RGBtoRGBPixelAlpha_1pixel_opaque s, d
and d, d, #0xff000000
bic s, s, #0xff000000
orr d, d, s
.endm
.macro RGBtoRGBPixelAlpha_process_head cond, numbytes, firstreg, unaligned_src, unaligned_mask, preload
.if numbytes == 16
ldm SRC!, {WK0, WK1}
ldm SRC!, {STRIDE_S, STRIDE_M}
ldrd WK2, WK3, [DST], #16
orr SCRATCH, WK0, WK1
and ORIG_W, WK0, WK1
orr SCRATCH, SCRATCH, STRIDE_S
and ORIG_W, ORIG_W, STRIDE_S
orr SCRATCH, SCRATCH, STRIDE_M
and ORIG_W, ORIG_W, STRIDE_M
tst SCRATCH, #0xff000000
.elseif numbytes == 8
ldm SRC!, {WK0, WK1}
ldm DST!, {WK2, WK3}
orr SCRATCH, WK0, WK1
and ORIG_W, WK0, WK1
tst SCRATCH, #0xff000000
.else // numbytes == 4
ldr WK0, [SRC], #4
ldr WK2, [DST], #4
tst WK0, #0xff000000
.endif
.endm
.macro RGBtoRGBPixelAlpha_process_tail cond, numbytes, firstreg
beq 20f @ all transparent
.if numbytes == 16
cmp ORIG_W, #0xff000000
bhs 10f @ all opaque
RGBtoRGBPixelAlpha_1pixel_translucent WK0, WK2, STRIDE_S, STRIDE_M, SCRATCH, ORIG_W, MASK
RGBtoRGBPixelAlpha_1pixel_translucent WK1, WK3, STRIDE_S, STRIDE_M, SCRATCH, ORIG_W, MASK
strd WK2, WK3, [DST, #-16]
ldrd WK0, WK1, [SRC, #-8]
ldrd WK2, WK3, [DST, #-8]
RGBtoRGBPixelAlpha_1pixel_translucent WK0, WK2, STRIDE_S, STRIDE_M, SCRATCH, ORIG_W, MASK
RGBtoRGBPixelAlpha_1pixel_translucent WK1, WK3, STRIDE_S, STRIDE_M, SCRATCH, ORIG_W, MASK
b 19f
10: RGBtoRGBPixelAlpha_1pixel_opaque WK0, WK2
RGBtoRGBPixelAlpha_1pixel_opaque WK1, WK3
strd WK2, WK3, [DST, #-16]
ldrd WK0, WK1, [SRC, #-8]
ldrd WK2, WK3, [DST, #-8]
RGBtoRGBPixelAlpha_1pixel_opaque WK0, WK2
RGBtoRGBPixelAlpha_1pixel_opaque WK1, WK3
19: strd WK2, WK3, [DST, #-8]
.elseif numbytes == 8
cmp ORIG_W, #0xff000000
bhs 10f @ all opaque
RGBtoRGBPixelAlpha_1pixel_translucent WK0, WK2, STRIDE_S, STRIDE_M, SCRATCH, ORIG_W, MASK
RGBtoRGBPixelAlpha_1pixel_translucent WK1, WK3, STRIDE_S, STRIDE_M, SCRATCH, ORIG_W, MASK
b 19f
10: RGBtoRGBPixelAlpha_1pixel_opaque WK0, WK2
RGBtoRGBPixelAlpha_1pixel_opaque WK1, WK3
19: strd WK2, WK3, [DST, #-8]
.else // numbytes == 4
cmp WK0, #0xff000000
bhs 10f @ opaque
RGBtoRGBPixelAlpha_1pixel_translucent WK0, WK2, STRIDE_S, STRIDE_M, SCRATCH, ORIG_W, MASK
b 19f
10: RGBtoRGBPixelAlpha_1pixel_opaque WK0, WK2
19: str WK2, [DST, #-4]
.endif
20:
.endm
generate_composite_function \
BlitRGBtoRGBPixelAlphaARMSIMDAsm, 32, 0, 32, \
FLAG_DST_READWRITE | FLAG_BRANCH_OVER | FLAG_PROCESS_CORRUPTS_PSR | FLAG_PROCESS_DOES_STORE | FLAG_SPILL_LINE_VARS | FLAG_PROCESS_CORRUPTS_WK0, \
2, /* prefetch distance */ \
RGBtoRGBPixelAlpha_init, \
nop_macro, /* newline */ \
nop_macro, /* cleanup */ \
RGBtoRGBPixelAlpha_process_head, \
RGBtoRGBPixelAlpha_process_tail
/******************************************************************************/

File diff suppressed because it is too large Load Diff