From patchwork Mon Jun 24 11:36:59 2024 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Zhao Zhili X-Patchwork-Id: 50120 Delivered-To: ffmpegpatchwork2@gmail.com Received: by 2002:a59:ae71:0:b0:482:c625:d099 with SMTP id w17csp1947598vqz; Mon, 24 Jun 2024 04:37:43 -0700 (PDT) X-Forwarded-Encrypted: i=2; AJvYcCXNKMbMgccTt3PwqkuQ7qqmwB7mw5gTerc4/dnUBaD7k3R/JAuMoWoLP0DsxtJVvcYFsUBq3Ef42qPIth4AN5Q++J8rkls8IzPqDw== X-Google-Smtp-Source: AGHT+IG3lDavniAEGKbktSMGWNEQ7xUNMyDiC7BBdwHTrzrVhnBpknDE2/LXZWq1e5iF3aydhIM4 X-Received: by 2002:a2e:8193:0:b0:2ec:2038:925d with SMTP id 38308e7fff4ca-2ec5b2c4f38mr29435711fa.1.1719229062906; Mon, 24 Jun 2024 04:37:42 -0700 (PDT) ARC-Seal: i=1; a=rsa-sha256; t=1719229062; cv=none; d=google.com; s=arc-20160816; b=F8Tx7oSvvLyX+EgnpYGYZMlyxkp69Z/3na2aF0Mn0NB2OjO23wOOABd11NqF8WBzJe gTkxATXe6aN7oA5rz6k+skc1nl+X3wiFqwJMq3Yt1cAtZeCTScVqSa8pkmaKBMGnBsw7 uxjUbP5nFhkylQU6TQ/iS2jEhUSmLiXUWWsqdgC6N3pNVHV15ggbj4fGpfRq8Z/FjyuL ImcYlf1ickjZJMyLE2hVHlf91kU2/kw9VIHPx7wHJOT6+mfMvqtEBp41A1fcehpVk/M3 o21qADsziGaDBl6mLPrEDO8Tg26P4ytKZJLWnoA/hVNca6+OeCa09fGbGs24GjlR1JJd q/6g== ARC-Message-Signature: i=1; a=rsa-sha256; c=relaxed/relaxed; d=google.com; s=arc-20160816; h=sender:errors-to:content-transfer-encoding:cc:reply-to :list-subscribe:list-help:list-post:list-archive:list-unsubscribe :list-id:precedence:subject:mime-version:date:to:from:message-id :dkim-signature:delivered-to; bh=cu6MwSTzFUlSCOBjCyv1qEKly6/g+ptBMMT3ErPRcR0=; fh=HnHYuZ9XgUo86ZRXTLWWmQxhslYEI9B9taZ5X1DLFfc=; b=dhhlxGLy1X/RK9RoX5qZcfKE2Z++1lehFb7f48F3aDVn3+udZw85sAyPGBUfoI+mdy NT9VqUxfOxt9V8gqGEvD5gE3XHLJlOu4bfAhPvQoGfBGKqsroOOmCciDLjIyeWDo8DkI MlzLPRynCqCNY4cHVlbRE4Sr/RHkmQBPSL8fYWEzPmeEu8axXOUwNclNvW7J0cgWgqzg YeKqnNv6Px1IY5jifb/29XDqywCqfxaohDNOoYVErJty9xLoWDEFUy5Ms3DK31eSeQg3 ztZX3Vk/iIRPAsCRJVzyHqx9NGrr5A86OXjQwmpqSat9yfdM4wWlYBrRlfuuHFUqfO7S pQ4A==; dara=google.com ARC-Authentication-Results: i=1; mx.google.com; dkim=neutral (body hash did not verify) header.i=@foxmail.com header.s=s201512 header.b=i6KcIOcb; spf=pass (google.com: domain of ffmpeg-devel-bounces@ffmpeg.org designates 79.124.17.100 as permitted sender) smtp.mailfrom=ffmpeg-devel-bounces@ffmpeg.org; dmarc=fail (p=NONE sp=NONE dis=NONE) header.from=foxmail.com Return-Path: Received: from ffbox0-bg.mplayerhq.hu (ffbox0-bg.ffmpeg.org. [79.124.17.100]) by mx.google.com with ESMTP id a640c23a62f3a-a6fcf565207si377001266b.615.2024.06.24.04.37.42; Mon, 24 Jun 2024 04:37:42 -0700 (PDT) Received-SPF: pass (google.com: domain of ffmpeg-devel-bounces@ffmpeg.org designates 79.124.17.100 as permitted sender) client-ip=79.124.17.100; Authentication-Results: mx.google.com; dkim=neutral (body hash did not verify) header.i=@foxmail.com header.s=s201512 header.b=i6KcIOcb; spf=pass (google.com: domain of ffmpeg-devel-bounces@ffmpeg.org designates 79.124.17.100 as permitted sender) smtp.mailfrom=ffmpeg-devel-bounces@ffmpeg.org; dmarc=fail (p=NONE sp=NONE dis=NONE) header.from=foxmail.com Received: from [127.0.1.1] (localhost [127.0.0.1]) by ffbox0-bg.mplayerhq.hu (Postfix) with ESMTP id 6761A68CE05; Mon, 24 Jun 2024 14:37:28 +0300 (EEST) X-Original-To: ffmpeg-devel@ffmpeg.org Delivered-To: ffmpeg-devel@ffmpeg.org Received: from out203-205-251-73.mail.qq.com (out203-205-251-73.mail.qq.com [203.205.251.73]) by ffbox0-bg.mplayerhq.hu (Postfix) with ESMTPS id B0A0468D5D2 for ; Mon, 24 Jun 2024 14:37:17 +0300 (EEST) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=foxmail.com; s=s201512; t=1719229025; bh=8WYXPGxnKhEGtYyFmoqHbV07jZjAb6611Iimmve4s/s=; h=From:To:Cc:Subject:Date; b=i6KcIOcbD7b6/+0Ic8f+GgDLSoSjpsZTymhA0klbqeHDjCg37M8EDYTKHSVPzUTwH s5CmFsPhSjT9diNoIeE5wU8UnvcBktoo895spgVu1oPrYMc9DGIWbQdQO3UB9P6uqs 7a9/S3dWgMev4G4qeLzUWMxh77uxJ689K24zexAY= Received: from ZHILIZHAO-MB1.tencent.com ([119.147.10.191]) by newxmesmtplogicsvrszb9-1.qq.com (NewEsmtp) with SMTP id 9438A032; Mon, 24 Jun 2024 19:37:03 +0800 X-QQ-mid: xmsmtpt1719229023t3hfsz908 Message-ID: X-QQ-XMAILINFO: MR/iVh5QLeiefIVaRaUPvAoLt3cJ0QhLVuhwWdzEQll8Tc9jLzbEpUVekLu7qq 5338KzNqppNtjOKCOraDn8uf90KjwkLw9fl6aixzD//MP2QeFjeBjAnmjutpLlDKL9P1bCdc+mle pECGjzoDHpz8xOOCObz+KpllyvqRguNaHEPPNldQy229kmboVcc/rDyuAYzZbeu7Mq5T8TLPui/S 0YjR6u04tPdFHwJyFYn83NiYqDyxBEp0Xk+Sl9q0/FsazAjvlvKz+9wiaUartcyboSm25ulQ0g/D qTNDQGJ0IjZ34zFF4c2KV/H8zbvrVubK+ARxQV+FoPqHy2XXJ72V5X0BQy8ncJaNtV3h668WcwHZ ZD6THWcGTOeNLy8NY7q76yft/sGMns3Oidd3R4ZmFERryzXwlDz4i/weJ0mAQW1Yh/ShXERH6zqB jF/dlngvMDKMSMh4uR8ROD/9a+ycToL3b//eG9YrTiiw0u/AAjIfU4vh+Xue8GKSI4LmgJDZSm3Q ml2UoDtNIdxsxYd6EvQSZdu7+LkKfzs01uecGneTXyu1hn0dla5uW/orgygl6e/qEZspad+BD/KV eCWfjt52rtyeeD4kjzPad2QtAtLaF5aVmmksDkeSa0AQVlE8KosD4FQTgY3ZddmUoIFjQ5Knsgmz q6FE7l/o2JUmgQRVwbzxzBQ9CrAkFHoIBE9l/7po+G94z2qatd9tPZaylYkK2VGHeVe5Mcdhgvgu PPD+dUXiIC8hbt6g7AzWN8EESnzizASF+W9shAOahyWcHOY+kbh8SAppQUL1OLxbYwxnRZY4HVy9 yAhC7kF2NQnflSL3BgO7JNKOxSFjEsDkbfM1SINSk+4CGqN0ovNPpP+PQjzM7LrRVc3Sd6HNUYKT 9gCUVF2wTS3J93++PcUV6/i2eOkcpRDaQnrVx/72B7n3EN39Q6xKbKFhiWhRY3Y3AG3tCWzRm0M9 ++l4D4gHlitVoMC5RPvdwzd1JsX9TrY4zX5FNCAr+VQBej5QTMJPt2JUyzVt5cwOKnc5kOMd0= X-QQ-XMRINFO: MPJ6Tf5t3I/ycC2BItcBVIA= From: Zhao Zhili To: ffmpeg-devel@ffmpeg.org Date: Mon, 24 Jun 2024 19:36:59 +0800 X-OQ-MSGID: <20240624113701.94616-1-quinkblack@foxmail.com> X-Mailer: git-send-email 2.42.0 MIME-Version: 1.0 Subject: [FFmpeg-devel] [PATCH v3 1/3] swscale/aarch64: Add bgr24 to yuv X-BeenThere: ffmpeg-devel@ffmpeg.org X-Mailman-Version: 2.1.29 Precedence: list List-Id: FFmpeg development discussions and patches List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Reply-To: FFmpeg development discussions and patches Cc: Zhao Zhili Errors-To: ffmpeg-devel-bounces@ffmpeg.org Sender: "ffmpeg-devel" X-TUID: FKX9SglEqRWZ From: Zhao Zhili Test on Apple M1 with kperf : -O3 : -O3 -fno-vectorize bgr24_to_uv_8_c : 28.5 : 52.5 bgr24_to_uv_8_neon : 54.5 : 59.7 bgr24_to_uv_128_c : 294.0 : 830.7 bgr24_to_uv_128_neon : 99.7 : 112.0 bgr24_to_uv_1080_c : 965.0 : 6624.0 bgr24_to_uv_1080_neon : 751.5 : 754.7 bgr24_to_uv_1920_c : 1693.2 : 11554.5 bgr24_to_uv_1920_neon : 1292.5 : 1307.5 bgr24_to_uv_half_8_c : 54.2 : 37.0 bgr24_to_uv_half_8_neon : 27.2 : 22.5 bgr24_to_uv_half_128_c : 127.2 : 392.5 bgr24_to_uv_half_128_neon : 63.0 : 52.0 bgr24_to_uv_half_1080_c : 880.2 : 3329.0 bgr24_to_uv_half_1080_neon : 401.5 : 390.7 bgr24_to_uv_half_1920_c : 1585.7 : 6390.7 bgr24_to_uv_half_1920_neon : 694.7 : 698.7 bgr24_to_y_8_c : 21.7 : 22.5 bgr24_to_y_8_neon : 797.2 : 25.5 bgr24_to_y_128_c : 88.0 : 280.5 bgr24_to_y_128_neon : 63.7 : 55.0 bgr24_to_y_1080_c : 616.7 : 2208.7 bgr24_to_y_1080_neon : 900.0 : 452.0 bgr24_to_y_1920_c : 1093.2 : 3894.7 bgr24_to_y_1920_neon : 777.2 : 767.5 --- libswscale/aarch64/input.S | 71 ++++++++++++++++++++++++++---------- libswscale/aarch64/swscale.c | 32 +++++++++------- 2 files changed, 71 insertions(+), 32 deletions(-) diff --git a/libswscale/aarch64/input.S b/libswscale/aarch64/input.S index 33afa34111..2cfec4cb6a 100644 --- a/libswscale/aarch64/input.S +++ b/libswscale/aarch64/input.S @@ -20,7 +20,7 @@ #include "libavutil/aarch64/asm.S" -.macro rgb24_to_yuv_load_rgb, src +.macro rgb_to_yuv_load_rgb src ld3 { v16.16b, v17.16b, v18.16b }, [\src] uxtl v19.8h, v16.8b // v19: r uxtl v20.8h, v17.8b // v20: g @@ -30,7 +30,7 @@ uxtl2 v24.8h, v18.16b // v24: b .endm -.macro rgb24_to_yuv_product, r, g, b, dst1, dst2, dst, coef0, coef1, coef2, right_shift +.macro rgb_to_yuv_product r, g, b, dst1, dst2, dst, coef0, coef1, coef2, right_shift mov \dst1\().16b, v6.16b // dst1 = const_offset mov \dst2\().16b, v6.16b // dst2 = const_offset smlal \dst1\().4s, \coef0\().4h, \r\().4h // dst1 += rx * r @@ -43,12 +43,20 @@ sqshrn2 \dst\().8h, \dst2\().4s, \right_shift // dst_higher_half = dst2 >> right_shift .endm +function ff_bgr24ToY_neon, export=1 + cmp w4, #0 // check width > 0 + ldp w12, w11, [x5] // w12: ry, w11: gy + ldr w10, [x5, #8] // w10: by + b.gt 4f + ret +endfunc + function ff_rgb24ToY_neon, export=1 cmp w4, #0 // check width > 0 ldp w10, w11, [x5] // w10: ry, w11: gy ldr w12, [x5, #8] // w12: by b.le 3f - +4: mov w9, #256 // w9 = 1 << (RGB2YUV_SHIFT - 7) movk w9, #8, lsl #16 // w9 += 32 << (RGB2YUV_SHIFT - 1) dup v6.4s, w9 // w9: const_offset @@ -59,9 +67,9 @@ function ff_rgb24ToY_neon, export=1 dup v2.8h, w12 b.lt 2f 1: - rgb24_to_yuv_load_rgb x1 - rgb24_to_yuv_product v19, v20, v21, v25, v26, v16, v0, v1, v2, #9 - rgb24_to_yuv_product v22, v23, v24, v27, v28, v17, v0, v1, v2, #9 + rgb_to_yuv_load_rgb x1 + rgb_to_yuv_product v19, v20, v21, v25, v26, v16, v0, v1, v2, #9 + rgb_to_yuv_product v22, v23, v24, v27, v28, v17, v0, v1, v2, #9 sub w4, w4, #16 // width -= 16 add x1, x1, #48 // src += 48 cmp w4, #16 // width >= 16 ? @@ -85,10 +93,7 @@ function ff_rgb24ToY_neon, export=1 ret endfunc -.macro rgb24_load_uv_coeff half - ldp w10, w11, [x6, #12] // w10: ru, w11: gu - ldp w12, w13, [x6, #20] // w12: bu, w13: rv - ldp w14, w15, [x6, #28] // w14: gv, w15: bv +.macro rgb_set_uv_coeff half .if \half mov w9, #512 movk w9, #128, lsl #16 // w9: const_offset @@ -105,12 +110,26 @@ endfunc dup v6.4s, w9 .endm +function ff_bgr24ToUV_half_neon, export=1 + cmp w5, #0 // check width > 0 + b.le 3f + + ldp w12, w11, [x6, #12] + ldp w10, w15, [x6, #20] + ldp w14, w13, [x6, #28] + b 4f +endfunc + function ff_rgb24ToUV_half_neon, export=1 cmp w5, #0 // check width > 0 b.le 3f + ldp w10, w11, [x6, #12] // w10: ru, w11: gu + ldp w12, w13, [x6, #20] // w12: bu, w13: rv + ldp w14, w15, [x6, #28] // w14: gv, w15: bv +4: cmp w5, #8 - rgb24_load_uv_coeff half=1 + rgb_set_uv_coeff half=1 b.lt 2f 1: ld3 { v16.16b, v17.16b, v18.16b }, [x3] @@ -118,8 +137,8 @@ function ff_rgb24ToUV_half_neon, export=1 uaddlp v20.8h, v17.16b // v20: g uaddlp v21.8h, v18.16b // v21: b - rgb24_to_yuv_product v19, v20, v21, v22, v23, v16, v0, v1, v2, #10 - rgb24_to_yuv_product v19, v20, v21, v24, v25, v17, v3, v4, v5, #10 + rgb_to_yuv_product v19, v20, v21, v22, v23, v16, v0, v1, v2, #10 + rgb_to_yuv_product v19, v20, v21, v24, v25, v17, v3, v4, v5, #10 sub w5, w5, #8 // width -= 8 add x3, x3, #48 // src += 48 cmp w5, #8 // width >= 8 ? @@ -158,19 +177,33 @@ function ff_rgb24ToUV_half_neon, export=1 ret endfunc +function ff_bgr24ToUV_neon, export=1 + cmp w5, #0 // check width > 0 + b.le 3f + + ldp w12, w11, [x6, #12] + ldp w10, w15, [x6, #20] + ldp w14, w13, [x6, #28] + b 4f +endfunc + function ff_rgb24ToUV_neon, export=1 cmp w5, #0 // check width > 0 b.le 3f + ldp w10, w11, [x6, #12] // w10: ru, w11: gu + ldp w12, w13, [x6, #20] // w12: bu, w13: rv + ldp w14, w15, [x6, #28] // w14: gv, w15: bv +4: cmp w5, #16 - rgb24_load_uv_coeff half=0 + rgb_set_uv_coeff half=0 b.lt 2f 1: - rgb24_to_yuv_load_rgb x3 - rgb24_to_yuv_product v19, v20, v21, v25, v26, v16, v0, v1, v2, #9 - rgb24_to_yuv_product v22, v23, v24, v27, v28, v17, v0, v1, v2, #9 - rgb24_to_yuv_product v19, v20, v21, v25, v26, v18, v3, v4, v5, #9 - rgb24_to_yuv_product v22, v23, v24, v27, v28, v19, v3, v4, v5, #9 + rgb_to_yuv_load_rgb x3 + rgb_to_yuv_product v19, v20, v21, v25, v26, v16, v0, v1, v2, #9 + rgb_to_yuv_product v22, v23, v24, v27, v28, v17, v0, v1, v2, #9 + rgb_to_yuv_product v19, v20, v21, v25, v26, v18, v3, v4, v5, #9 + rgb_to_yuv_product v22, v23, v24, v27, v28, v19, v3, v4, v5, #9 sub w5, w5, #16 add x3, x3, #48 // src += 48 cmp w5, #16 diff --git a/libswscale/aarch64/swscale.c b/libswscale/aarch64/swscale.c index e4ea3309ba..c6594944c3 100644 --- a/libswscale/aarch64/swscale.c +++ b/libswscale/aarch64/swscale.c @@ -201,19 +201,18 @@ void ff_yuv2plane1_8_neon( default: break; \ } -void ff_rgb24ToY_neon(uint8_t *_dst, const uint8_t *src, const uint8_t *unused1, - const uint8_t *unused2, int width, - uint32_t *rgb2yuv, void *opq); - -void ff_rgb24ToUV_neon(uint8_t *_dstU, uint8_t *_dstV, const uint8_t *unused0, - const uint8_t *src1, - const uint8_t *src2, int width, uint32_t *rgb2yuv, - void *opq); - -void ff_rgb24ToUV_half_neon(uint8_t *_dstU, uint8_t *_dstV, const uint8_t *unused0, - const uint8_t *src1, - const uint8_t *src2, int width, uint32_t *rgb2yuv, - void *opq); +#define NEON_INPUT(name) \ +void ff_##name##ToY_neon(uint8_t *dst, const uint8_t *src, const uint8_t *, \ + const uint8_t *, int w, uint32_t *coeffs, void *); \ +void ff_##name##ToUV_neon(uint8_t *, uint8_t *, const uint8_t *, \ + const uint8_t *, const uint8_t *, int w, \ + uint32_t *coeffs, void *); \ +void ff_##name##ToUV_half_neon(uint8_t *, uint8_t *, const uint8_t *, \ + const uint8_t *, const uint8_t *, int w, \ + uint32_t *coeffs, void *) + +NEON_INPUT(bgr24); +NEON_INPUT(rgb24); void ff_lumRangeFromJpeg_neon(int16_t *dst, int width); void ff_chrRangeFromJpeg_neon(int16_t *dstU, int16_t *dstV, int width); @@ -247,6 +246,13 @@ av_cold void ff_sws_init_swscale_aarch64(SwsContext *c) c->yuv2planeX = ff_yuv2planeX_8_neon; } switch (c->srcFormat) { + case AV_PIX_FMT_BGR24: + c->lumToYV12 = ff_bgr24ToY_neon; + if (c->chrSrcHSubSample) + c->chrToYV12 = ff_bgr24ToUV_half_neon; + else + c->chrToYV12 = ff_bgr24ToUV_neon; + break; case AV_PIX_FMT_RGB24: c->lumToYV12 = ff_rgb24ToY_neon; if (c->chrSrcHSubSample)