From patchwork Thu Dec 13 12:07:58 2018 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Lauri Kasanen X-Patchwork-Id: 11400 Return-Path: X-Original-To: patchwork@ffaux-bg.ffmpeg.org Delivered-To: patchwork@ffaux-bg.ffmpeg.org Received: from ffbox0-bg.mplayerhq.hu (ffbox0-bg.ffmpeg.org [79.124.17.100]) by ffaux.localdomain (Postfix) with ESMTP id 645E144C5DD for ; Thu, 13 Dec 2018 14:05:37 +0200 (EET) Received: from [127.0.1.1] (localhost [127.0.0.1]) by ffbox0-bg.mplayerhq.hu (Postfix) with ESMTP id 8DD0168A918; Thu, 13 Dec 2018 14:05:37 +0200 (EET) X-Original-To: ffmpeg-devel@ffmpeg.org Delivered-To: ffmpeg-devel@ffmpeg.org Received: from mout.gmx.net (mout.gmx.net [212.227.17.21]) by ffbox0-bg.mplayerhq.hu (Postfix) with ESMTPS id CDCDC68A6A0 for ; Thu, 13 Dec 2018 14:05:31 +0200 (EET) Received: from Valinor ([84.250.81.169]) by mail.gmx.com (mrgmx103 [212.227.17.174]) with ESMTPSA (Nemesis) id 0Lg1QN-1hD7A13WcV-00pcH6 for ; Thu, 13 Dec 2018 13:05:30 +0100 Date: Thu, 13 Dec 2018 14:07:58 +0200 From: Lauri Kasanen To: FFmpeg development discussions and patches Message-Id: <20181213140758.6c8452638d9e46f52dfc183d@gmx.com> In-Reply-To: <20181213140546.68318fe2da1753bfc580805a@gmx.com> References: <20181213140546.68318fe2da1753bfc580805a@gmx.com> X-Mailer: Sylpheed 3.5.0 (GTK+ 2.18.6; x86_64-unknown-linux-gnu) Mime-Version: 1.0 X-Provags-ID: V03:K1:WbPIq+KaRJr2bs5bJlT0zKjHYFhiZXb7n7M5z0VJBbUhvpSVXlf axsEplzcyyI/nD3PeUWfY79OD7iRWjKmMkxp4/NiQzKN0phJk800fB5XCTIR438keoybrxI NxPKcWlXFdR9ia7s8SXV4/3UgJxFpi0TyR44u5gwGR6tHZbxv/cxB81erFTz/1uf+EtukTI q+25ZOpZDDxC5dw60/8kw== X-Spam-Flag: NO X-UI-Out-Filterresults: notjunk:1; V03:K0:8Vlv1Dg31fA=:8jz6+hrLE/K+K5WviG5r7B hrPfHfnXRy0GU12Oxe9lmXiEkO0dnY1ET+lAw9PoxvHYaHNjsQNBKy1ZVXgybBgYt20jrUbEZ 4zWorhmlylQHmhNEgOT/YvVEGUIG9x7G5/5pmXpASi+gNfEqE+mqgSryGlM2WyTeKVQbqg/jB dQ+cbsSlilw9Zw+5I4CDGO0Woikuv44KQu1Q9+CIstkUyW5uKFDiBf0X0CAr3cOqEzGEGIPUp SVR2ZkwLs6a3wEETeZFxBYAR4cYPTemQW3CGiLpS4rG+HLlScAyXn4OxpU+EZI70n1gFIumrf tVZLBUWAXdZCA0K7GBWninWclK92DFI94n8OWamcEoMNiD8fYcALIKut8eG4hSFBy0v0AZkEj ciBBZkbWDp0SSSO5XoqKCMCOM8jzWdNRb9Gp6ni77HlUJQE9rzOP2R9pmiAt2zUJiv3yHj92f l5UwQvz3BvK3/vsuBOmAO0O77JwDaxfSH2L2Wwh2SAD8k90SUB8h/ktxIIxfBJiO87fJq0VBf qiU2JAcF+7o199AlEVdrVBhQ80dcXDEp48sDLPPjKEtKuCbplxhe3MR9OIEc30VwZFdKoZNKN 6psfLSeshz7ZzQjShIC1edoFZgku0sX1wf0foJet+EoynRwIbCRQ81ygH8P9Bp0dccpLvBwq3 Dlzv5KFUARl/O7w3PepciKtccEI1kBP36c8Sr20jbLrJ/aiFsPXgBYHOb+Lkj1mFCwlNrK0xJ 26qi7pg5wxmj9pl/a3IogJRWi0XZ0ITnF6kT8LV6q4OkujU7EB8tkeIGVlsp76MsE0ZiyEl55 zcqI0eeGpJr6Ppt0qZ+k1GtESxkFUdHaX0OfKIuPRPm3kGHs0QXHDGgaaDD94c5ewro0Eo96E dwCCRPHt8KxCYZV/UwmCpS1HzFaKCJe8ayajDYlKe44qEZNHVo5TuO8D1UKMs2 Subject: [FFmpeg-devel] [PATCH v2] swscale/output: VSX-optimize 16-bit yuv2plane1 X-BeenThere: ffmpeg-devel@ffmpeg.org X-Mailman-Version: 2.1.20 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 Errors-To: ffmpeg-devel-bounces@ffmpeg.org Sender: "ffmpeg-devel" ./ffmpeg_g -f rawvideo -pix_fmt rgb24 -s hd1080 -i /dev/zero -pix_fmt yuv420p16le \ -f null -vframes 100 -v error -nostats - 2120 UNITS in planar1, 65393 runs, 143 skips -cpuflags 0 19157 UNITS in planar1, 65512 runs, 24 skips 9.03632 speedup, 16be similarly. Fate passes, each format tested with an image to video conversion. Signed-off-by: Lauri Kasanen --- v2: Copy-pasted rows were flipped. libswscale/ppc/swscale_vsx.c | 59 ++++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 59 insertions(+) diff --git a/libswscale/ppc/swscale_vsx.c b/libswscale/ppc/swscale_vsx.c index 6462c11..70da6ae 100644 --- a/libswscale/ppc/swscale_vsx.c +++ b/libswscale/ppc/swscale_vsx.c @@ -180,6 +180,60 @@ static void yuv2plane1_nbps_vsx(const int16_t *src, uint16_t *dest, int dstW, yuv2plane1_nbps_u(src, dest, dstW, big_endian, output_bits, i); } +#undef output_pixel + +#define output_pixel(pos, val, bias, signedness) \ + if (big_endian) { \ + AV_WB16(pos, bias + av_clip_ ## signedness ## 16(val >> shift)); \ + } else { \ + AV_WL16(pos, bias + av_clip_ ## signedness ## 16(val >> shift)); \ + } + +static void yuv2plane1_16_u(const int32_t *src, uint16_t *dest, int dstW, + int big_endian, int output_bits, int start) +{ + int i; + const int shift = 3; + + for (i = start; i < dstW; i++) { + int val = src[i] + (1 << (shift - 1)); + output_pixel(&dest[i], val, 0, uint); + } +} + +static void yuv2plane1_16_vsx(const int32_t *src, uint16_t *dest, int dstW, + int big_endian, int output_bits) +{ + const int dst_u = -(uintptr_t)dest & 7; + const int shift = 3; + const int add = (1 << (shift - 1)); + const vector uint32_t vadd = (vector uint32_t) {add, add, add, add}; + const vector uint16_t vswap = (vector uint16_t) vec_splat_u16(big_endian ? 8 : 0); + const vector uint32_t vshift = (vector uint32_t) vec_splat_u32(shift); + vector uint32_t v, v2; + vector uint16_t vd; + int i; + + yuv2plane1_16_u(src, dest, dst_u, big_endian, output_bits, 0); + + for (i = dst_u; i < dstW - 7; i += 8) { + v = vec_vsx_ld(0, (const uint32_t *) &src[i]); + v = vec_add(v, vadd); + v = vec_sr(v, vshift); + + v2 = vec_vsx_ld(0, (const uint32_t *) &src[i + 4]); + v2 = vec_add(v2, vadd); + v2 = vec_sr(v2, vshift); + + vd = vec_packsu(v, v2); + vd = vec_rl(vd, vswap); + + vec_st(vd, 0, &dest[i]); + } + + yuv2plane1_16_u(src, dest, dstW, big_endian, output_bits, i); +} + #define yuv2NBPS(bits, BE_LE, is_be, template_size, typeX_t) \ static void yuv2plane1_ ## bits ## BE_LE ## _vsx(const int16_t *src, \ uint8_t *dest, int dstW, \ @@ -197,6 +251,8 @@ yuv2NBPS(12, BE, 1, nbps, int16_t) yuv2NBPS(12, LE, 0, nbps, int16_t) yuv2NBPS(14, BE, 1, nbps, int16_t) yuv2NBPS(14, LE, 0, nbps, int16_t) +yuv2NBPS(16, BE, 1, 16, int32_t) +yuv2NBPS(16, LE, 0, 16, int32_t) #endif /* !HAVE_BIGENDIAN */ @@ -240,6 +296,9 @@ av_cold void ff_sws_init_swscale_vsx(SwsContext *c) case 14: c->yuv2plane1 = isBE(dstFormat) ? yuv2plane1_14BE_vsx : yuv2plane1_14LE_vsx; break; + case 16: + c->yuv2plane1 = isBE(dstFormat) ? yuv2plane1_16BE_vsx : yuv2plane1_16LE_vsx; + break; #endif } }