From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from ffbox0-bg.mplayerhq.hu (ffbox0-bg.ffmpeg.org [79.124.17.100]) by master.gitmailbox.com (Postfix) with ESMTP id C46A34753D for ; Sat, 11 Nov 2023 15:27:33 +0000 (UTC) Received: from [127.0.1.1] (localhost [127.0.0.1]) by ffbox0-bg.mplayerhq.hu (Postfix) with ESMTP id DBE8668CCA2; Sat, 11 Nov 2023 17:27:30 +0200 (EET) Received: from ursule.remlab.net (vps-a2bccee9.vps.ovh.net [51.75.19.47]) by ffbox0-bg.mplayerhq.hu (Postfix) with ESMTP id 40A1A68C944 for ; Sat, 11 Nov 2023 17:27:24 +0200 (EET) Received: from basile.remlab.net (localhost [IPv6:::1]) by ursule.remlab.net (Postfix) with ESMTP id C0AF5C006B for ; Sat, 11 Nov 2023 17:27:23 +0200 (EET) From: =?UTF-8?q?R=C3=A9mi=20Denis-Courmont?= To: ffmpeg-devel@ffmpeg.org Date: Sat, 11 Nov 2023 17:27:23 +0200 Message-ID: <20231111152723.41888-1-remi@remlab.net> X-Mailer: git-send-email 2.42.0 MIME-Version: 1.0 Subject: [FFmpeg-devel] [PATCH] lavc/exrdsp: unroll predictor 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 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit Errors-To: ffmpeg-devel-bounces@ffmpeg.org Sender: "ffmpeg-devel" Archived-At: List-Archive: List-Post: With explicit unrolling, we can skip half of the sign bit flips, and the compiler is then better able to optimise the scalar loop: predictor_c: 31376.0 (before) predictor_c: 23703.0 (after) --- libavcodec/exrdsp.c | 16 +++++++++++++--- 1 file changed, 13 insertions(+), 3 deletions(-) diff --git a/libavcodec/exrdsp.c b/libavcodec/exrdsp.c index 752e1eb553..248cb93c5a 100644 --- a/libavcodec/exrdsp.c +++ b/libavcodec/exrdsp.c @@ -40,10 +40,20 @@ static void reorder_pixels_scalar(uint8_t *dst, const uint8_t *src, ptrdiff_t si static void predictor_scalar(uint8_t *src, ptrdiff_t size) { - ptrdiff_t i; + /* Unrolled: `src[i + 1] += src[i] - 128;` */ + if ((size & 1) == 0) { + src[1] += src[0] ^ 0x80; + src++; + size--; + } + + for (ptrdiff_t i = 1; i < size; i += 2) { + uint8_t a = src[i] + src[i - 1]; - for (i = 1; i < size; i++) - src[i] += src[i-1] - 128; + src[i] = a; + src[i + 1] += a; + src[i] ^= 0x80; + } } av_cold void ff_exrdsp_init(ExrDSPContext *c) -- 2.42.0 _______________________________________________ ffmpeg-devel mailing list ffmpeg-devel@ffmpeg.org https://ffmpeg.org/mailman/listinfo/ffmpeg-devel To unsubscribe, visit link above, or email ffmpeg-devel-request@ffmpeg.org with subject "unsubscribe".