From patchwork Tue May 12 20:30:54 2015 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Christophe Lyon X-Patchwork-Id: 48411 Return-Path: X-Original-To: linaro@patches.linaro.org Delivered-To: linaro@patches.linaro.org Received: from mail-lb0-f199.google.com (mail-lb0-f199.google.com [209.85.217.199]) by ip-10-151-82-157.ec2.internal (Postfix) with ESMTPS id 198B52121F for ; Tue, 12 May 2015 20:34:41 +0000 (UTC) Received: by lbbrr5 with SMTP id rr5sf4373598lbb.3 for ; Tue, 12 May 2015 13:34:39 -0700 (PDT) X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20130820; h=x-gm-message-state:mime-version:delivered-to:mailing-list :precedence:list-id:list-unsubscribe:list-archive:list-post :list-help:sender:delivered-to:from:to:subject:date:message-id :in-reply-to:references:x-original-sender :x-original-authentication-results; bh=B6ffQzp3MCm3Z60rFvqOTUBwUY7Zsigb1QnJG+IbTUw=; b=CHzr6a2hlvDb6LXCF3g4YklPvcZO889OfOmjVhyrhEWmf4O3scq8JwD58lfQqARA4P yHoFCmP25qExAt91U4J8v+pwxvGwq2VZprqhP8SMhKLsnJmahjvuxBxv3W0UqTKnGJm6 yn1OZN6dTazNodghfFcpXvnuuyK1KK8eHgV75F6n6hU6Ldu/uBob/6AcZKhwdgR6O3gm QLNpH5kuaKNcNbyEZXYZc7N0NrIRhj64riQ43PFwh/EXorb0omibWzm1GbfTcn8xBts/ FHkL+6XCDkdJpO0mAq5dBNcYR3R7mkdhRyMkzjlLcdQOg0xo7Mvkf2h1rqIcHxH20DNo ky6g== X-Gm-Message-State: ALoCoQnWPlpIv4IQH4u/05Dm9ox548nVBlWPo/w7ODGP8YS+TAda3xd/2Rj0RrW9vvTX6+ajbTSD X-Received: by 10.152.87.140 with SMTP id ay12mr11723984lab.8.1431462879822; Tue, 12 May 2015 13:34:39 -0700 (PDT) MIME-Version: 1.0 X-BeenThere: patchwork-forward@linaro.org Received: by 10.152.9.37 with SMTP id w5ls83790laa.77.gmail; Tue, 12 May 2015 13:34:39 -0700 (PDT) X-Received: by 10.112.205.69 with SMTP id le5mr13462067lbc.65.1431462879535; Tue, 12 May 2015 13:34:39 -0700 (PDT) Received: from mail-lb0-x235.google.com (mail-lb0-x235.google.com. [2a00:1450:4010:c04::235]) by mx.google.com with ESMTPS id ug3si11004620lbb.130.2015.05.12.13.34.39 for (version=TLSv1.2 cipher=ECDHE-RSA-AES128-GCM-SHA256 bits=128/128); Tue, 12 May 2015 13:34:39 -0700 (PDT) Received-SPF: pass (google.com: domain of patch+caf_=patchwork-forward=linaro.org@linaro.org designates 2a00:1450:4010:c04::235 as permitted sender) client-ip=2a00:1450:4010:c04::235; Received: by lbbqq2 with SMTP id qq2so14871386lbb.3 for ; Tue, 12 May 2015 13:34:39 -0700 (PDT) X-Received: by 10.152.29.161 with SMTP id l1mr13065652lah.76.1431462879374; Tue, 12 May 2015 13:34:39 -0700 (PDT) X-Forwarded-To: patchwork-forward@linaro.org X-Forwarded-For: patch@linaro.org patchwork-forward@linaro.org Delivered-To: patch@linaro.org Received: by 10.112.108.230 with SMTP id hn6csp137638lbb; Tue, 12 May 2015 13:34:37 -0700 (PDT) X-Received: by 10.68.233.71 with SMTP id tu7mr31352871pbc.14.1431462876325; Tue, 12 May 2015 13:34:36 -0700 (PDT) Received: from sourceware.org (server1.sourceware.org. [209.132.180.131]) by mx.google.com with ESMTPS id i6si24015836pdr.64.2015.05.12.13.34.35 for (version=TLSv1.2 cipher=ECDHE-RSA-AES128-GCM-SHA256 bits=128/128); Tue, 12 May 2015 13:34:36 -0700 (PDT) Received-SPF: pass (google.com: domain of gcc-patches-return-397719-patch=linaro.org@gcc.gnu.org designates 209.132.180.131 as permitted sender) client-ip=209.132.180.131; Received: (qmail 76830 invoked by alias); 12 May 2015 20:31:29 -0000 Mailing-List: list patchwork-forward@linaro.org; contact patchwork-forward+owners@linaro.org Precedence: list List-Id: List-Unsubscribe: , List-Archive: List-Post: , List-Help: , Sender: gcc-patches-owner@gcc.gnu.org Delivered-To: mailing list gcc-patches@gcc.gnu.org Received: (qmail 76703 invoked by uid 89); 12 May 2015 20:31:29 -0000 X-Virus-Found: No X-Spam-SWARE-Status: No, score=-2.6 required=5.0 tests=AWL, BAYES_00, RCVD_IN_DNSWL_LOW, SPF_PASS autolearn=ham version=3.3.2 X-HELO: mail-wg0-f54.google.com Received: from mail-wg0-f54.google.com (HELO mail-wg0-f54.google.com) (74.125.82.54) by sourceware.org (qpsmtpd/0.93/v0.84-503-g423c35a) with (AES128-GCM-SHA256 encrypted) ESMTPS; Tue, 12 May 2015 20:31:19 +0000 Received: by wggj6 with SMTP id j6so22694782wgg.3 for ; Tue, 12 May 2015 13:31:16 -0700 (PDT) X-Received: by 10.180.72.198 with SMTP id f6mr8084368wiv.21.1431462676677; Tue, 12 May 2015 13:31:16 -0700 (PDT) Received: from babel.clyon.hd.free.fr (vig38-2-82-225-222-175.fbx.proxad.net. [82.225.222.175]) by mx.google.com with ESMTPSA id u9sm24921095wju.44.2015.05.12.13.31.15 for (version=TLSv1.2 cipher=ECDHE-RSA-AES128-SHA bits=128/128); Tue, 12 May 2015 13:31:16 -0700 (PDT) From: Christophe Lyon To: gcc-patches@gcc.gnu.org Subject: [Patch ARM-AArch64/testsuite Neon intrinsics 06/13] Add vqrshl tests. Date: Tue, 12 May 2015 22:30:54 +0200 Message-Id: <1431462661-27247-7-git-send-email-christophe.lyon@linaro.org> In-Reply-To: <1431462661-27247-1-git-send-email-christophe.lyon@linaro.org> References: <1431462661-27247-1-git-send-email-christophe.lyon@linaro.org> X-IsSubscribed: yes X-Original-Sender: christophe.lyon@linaro.org X-Original-Authentication-Results: mx.google.com; spf=pass (google.com: domain of patch+caf_=patchwork-forward=linaro.org@linaro.org designates 2a00:1450:4010:c04::235 as permitted sender) smtp.mail=patch+caf_=patchwork-forward=linaro.org@linaro.org; dkim=pass header.i=@gcc.gnu.org X-Google-Group-Id: 836684582541 diff --git a/gcc/testsuite/gcc.target/aarch64/advsimd-intrinsics/vqrshl.c b/gcc/testsuite/gcc.target/aarch64/advsimd-intrinsics/vqrshl.c new file mode 100644 index 0000000..3f0cb37 --- /dev/null +++ b/gcc/testsuite/gcc.target/aarch64/advsimd-intrinsics/vqrshl.c @@ -0,0 +1,1090 @@ +#include +#include "arm-neon-ref.h" +#include "compute-ref-data.h" + +/* Expected values of cumulative_saturation flag with input=0. */ +int VECT_VAR(expected_cumulative_sat_0,int,8,8) = 0; +int VECT_VAR(expected_cumulative_sat_0,int,16,4) = 0; +int VECT_VAR(expected_cumulative_sat_0,int,32,2) = 0; +int VECT_VAR(expected_cumulative_sat_0,int,64,1) = 0; +int VECT_VAR(expected_cumulative_sat_0,uint,8,8) = 0; +int VECT_VAR(expected_cumulative_sat_0,uint,16,4) = 0; +int VECT_VAR(expected_cumulative_sat_0,uint,32,2) = 0; +int VECT_VAR(expected_cumulative_sat_0,uint,64,1) = 0; +int VECT_VAR(expected_cumulative_sat_0,int,8,16) = 0; +int VECT_VAR(expected_cumulative_sat_0,int,16,8) = 0; +int VECT_VAR(expected_cumulative_sat_0,int,32,4) = 0; +int VECT_VAR(expected_cumulative_sat_0,int,64,2) = 0; +int VECT_VAR(expected_cumulative_sat_0,uint,8,16) = 0; +int VECT_VAR(expected_cumulative_sat_0,uint,16,8) = 0; +int VECT_VAR(expected_cumulative_sat_0,uint,32,4) = 0; +int VECT_VAR(expected_cumulative_sat_0,uint,64,2) = 0; + +/* Expected results with input=0. */ +VECT_VAR_DECL(expected_0,int,8,8) [] = { 0x0, 0x0, 0x0, 0x0, + 0x0, 0x0, 0x0, 0x0 }; +VECT_VAR_DECL(expected_0,int,16,4) [] = { 0x0, 0x0, 0x0, 0x0 }; +VECT_VAR_DECL(expected_0,int,32,2) [] = { 0x0, 0x0 }; +VECT_VAR_DECL(expected_0,int,64,1) [] = { 0x0 }; +VECT_VAR_DECL(expected_0,uint,8,8) [] = { 0x0, 0x0, 0x0, 0x0, + 0x0, 0x0, 0x0, 0x0 }; +VECT_VAR_DECL(expected_0,uint,16,4) [] = { 0x0, 0x0, 0x0, 0x0 }; +VECT_VAR_DECL(expected_0,uint,32,2) [] = { 0x0, 0x0 }; +VECT_VAR_DECL(expected_0,uint,64,1) [] = { 0x0 }; +VECT_VAR_DECL(expected_0,int,8,16) [] = { 0x0, 0x0, 0x0, 0x0, + 0x0, 0x0, 0x0, 0x0, + 0x0, 0x0, 0x0, 0x0, + 0x0, 0x0, 0x0, 0x0 }; +VECT_VAR_DECL(expected_0,int,16,8) [] = { 0x0, 0x0, 0x0, 0x0, + 0x0, 0x0, 0x0, 0x0 }; +VECT_VAR_DECL(expected_0,int,32,4) [] = { 0x0, 0x0, 0x0, 0x0 }; +VECT_VAR_DECL(expected_0,int,64,2) [] = { 0x0, 0x0 }; +VECT_VAR_DECL(expected_0,uint,8,16) [] = { 0x0, 0x0, 0x0, 0x0, + 0x0, 0x0, 0x0, 0x0, + 0x0, 0x0, 0x0, 0x0, + 0x0, 0x0, 0x0, 0x0 }; +VECT_VAR_DECL(expected_0,uint,16,8) [] = { 0x0, 0x0, 0x0, 0x0, + 0x0, 0x0, 0x0, 0x0 }; +VECT_VAR_DECL(expected_0,uint,32,4) [] = { 0x0, 0x0, 0x0, 0x0 }; +VECT_VAR_DECL(expected_0,uint,64,2) [] = { 0x0, 0x0 }; + +/* Expected values of cumulative_saturation flag with input=0 and + negative shift amount. */ +int VECT_VAR(expected_cumulative_sat_0_neg,int,8,8) = 0; +int VECT_VAR(expected_cumulative_sat_0_neg,int,16,4) = 0; +int VECT_VAR(expected_cumulative_sat_0_neg,int,32,2) = 0; +int VECT_VAR(expected_cumulative_sat_0_neg,int,64,1) = 0; +int VECT_VAR(expected_cumulative_sat_0_neg,uint,8,8) = 0; +int VECT_VAR(expected_cumulative_sat_0_neg,uint,16,4) = 0; +int VECT_VAR(expected_cumulative_sat_0_neg,uint,32,2) = 0; +int VECT_VAR(expected_cumulative_sat_0_neg,uint,64,1) = 0; +int VECT_VAR(expected_cumulative_sat_0_neg,int,8,16) = 0; +int VECT_VAR(expected_cumulative_sat_0_neg,int,16,8) = 0; +int VECT_VAR(expected_cumulative_sat_0_neg,int,32,4) = 0; +int VECT_VAR(expected_cumulative_sat_0_neg,int,64,2) = 0; +int VECT_VAR(expected_cumulative_sat_0_neg,uint,8,16) = 0; +int VECT_VAR(expected_cumulative_sat_0_neg,uint,16,8) = 0; +int VECT_VAR(expected_cumulative_sat_0_neg,uint,32,4) = 0; +int VECT_VAR(expected_cumulative_sat_0_neg,uint,64,2) = 0; + +/* Expected results with input=0 and negative shift amount. */ +VECT_VAR_DECL(expected_0_neg,int,8,8) [] = { 0x0, 0x0, 0x0, 0x0, + 0x0, 0x0, 0x0, 0x0 }; +VECT_VAR_DECL(expected_0_neg,int,16,4) [] = { 0x0, 0x0, 0x0, 0x0 }; +VECT_VAR_DECL(expected_0_neg,int,32,2) [] = { 0x0, 0x0 }; +VECT_VAR_DECL(expected_0_neg,int,64,1) [] = { 0x0 }; +VECT_VAR_DECL(expected_0_neg,uint,8,8) [] = { 0x0, 0x0, 0x0, 0x0, + 0x0, 0x0, 0x0, 0x0 }; +VECT_VAR_DECL(expected_0_neg,uint,16,4) [] = { 0x0, 0x0, 0x0, 0x0 }; +VECT_VAR_DECL(expected_0_neg,uint,32,2) [] = { 0x0, 0x0 }; +VECT_VAR_DECL(expected_0_neg,uint,64,1) [] = { 0x0 }; +VECT_VAR_DECL(expected_0_neg,int,8,16) [] = { 0x0, 0x0, 0x0, 0x0, + 0x0, 0x0, 0x0, 0x0, + 0x0, 0x0, 0x0, 0x0, + 0x0, 0x0, 0x0, 0x0 }; +VECT_VAR_DECL(expected_0_neg,int,16,8) [] = { 0x0, 0x0, 0x0, 0x0, + 0x0, 0x0, 0x0, 0x0 }; +VECT_VAR_DECL(expected_0_neg,int,32,4) [] = { 0x0, 0x0, 0x0, 0x0 }; +VECT_VAR_DECL(expected_0_neg,int,64,2) [] = { 0x0, 0x0 }; +VECT_VAR_DECL(expected_0_neg,uint,8,16) [] = { 0x0, 0x0, 0x0, 0x0, + 0x0, 0x0, 0x0, 0x0, + 0x0, 0x0, 0x0, 0x0, + 0x0, 0x0, 0x0, 0x0 }; +VECT_VAR_DECL(expected_0_neg,uint,16,8) [] = { 0x0, 0x0, 0x0, 0x0, + 0x0, 0x0, 0x0, 0x0 }; +VECT_VAR_DECL(expected_0_neg,uint,32,4) [] = { 0x0, 0x0, 0x0, 0x0 }; +VECT_VAR_DECL(expected_0_neg,uint,64,2) [] = { 0x0, 0x0 }; + +/* Expected values of cumulative_saturation flag. */ +int VECT_VAR(expected_cumulative_sat,int,8,8) = 0; +int VECT_VAR(expected_cumulative_sat,int,16,4) = 0; +int VECT_VAR(expected_cumulative_sat,int,32,2) = 0; +int VECT_VAR(expected_cumulative_sat,int,64,1) = 0; +int VECT_VAR(expected_cumulative_sat,uint,8,8) = 1; +int VECT_VAR(expected_cumulative_sat,uint,16,4) = 1; +int VECT_VAR(expected_cumulative_sat,uint,32,2) = 1; +int VECT_VAR(expected_cumulative_sat,uint,64,1) = 1; +int VECT_VAR(expected_cumulative_sat,int,8,16) = 1; +int VECT_VAR(expected_cumulative_sat,int,16,8) = 1; +int VECT_VAR(expected_cumulative_sat,int,32,4) = 1; +int VECT_VAR(expected_cumulative_sat,int,64,2) = 1; +int VECT_VAR(expected_cumulative_sat,uint,8,16) = 1; +int VECT_VAR(expected_cumulative_sat,uint,16,8) = 1; +int VECT_VAR(expected_cumulative_sat,uint,32,4) = 1; +int VECT_VAR(expected_cumulative_sat,uint,64,2) = 1; + +/* Expected results. */ +VECT_VAR_DECL(expected,int,8,8) [] = { 0xe0, 0xe2, 0xe4, 0xe6, + 0xe8, 0xea, 0xec, 0xee }; +VECT_VAR_DECL(expected,int,16,4) [] = { 0xff80, 0xff88, 0xff90, 0xff98 }; +VECT_VAR_DECL(expected,int,32,2) [] = { 0xfffff000, 0xfffff100 }; +VECT_VAR_DECL(expected,int,64,1) [] = { 0xffffffffffffff80 }; +VECT_VAR_DECL(expected,uint,8,8) [] = { 0xff, 0xff, 0xff, 0xff, + 0xff, 0xff, 0xff, 0xff }; +VECT_VAR_DECL(expected,uint,16,4) [] = { 0xffff, 0xffff, 0xffff, 0xffff }; +VECT_VAR_DECL(expected,uint,32,2) [] = { 0xffffffff, 0xffffffff }; +VECT_VAR_DECL(expected,uint,64,1) [] = { 0xffffffffffffffff }; +VECT_VAR_DECL(expected,int,8,16) [] = { 0x80, 0x80, 0x80, 0x80, + 0x80, 0x80, 0x80, 0x80, + 0x80, 0x80, 0x80, 0x80, + 0x80, 0x80, 0x80, 0x80 }; +VECT_VAR_DECL(expected,int,16,8) [] = { 0x8000, 0x8000, 0x8000, 0x8000, + 0x8000, 0x8000, 0x8000, 0x8000 }; +VECT_VAR_DECL(expected,int,32,4) [] = { 0x80000000, 0x80000000, + 0x80000000, 0x80000000 }; +VECT_VAR_DECL(expected,int,64,2) [] = { 0x8000000000000000, 0x8000000000000000 }; +VECT_VAR_DECL(expected,uint,8,16) [] = { 0xff, 0xff, 0xff, 0xff, + 0xff, 0xff, 0xff, 0xff, + 0xff, 0xff, 0xff, 0xff, + 0xff, 0xff, 0xff, 0xff }; +VECT_VAR_DECL(expected,uint,16,8) [] = { 0xffff, 0xffff, 0xffff, 0xffff, + 0xffff, 0xffff, 0xffff, 0xffff }; +VECT_VAR_DECL(expected,uint,32,4) [] = { 0xffffffff, 0xffffffff, + 0xffffffff, 0xffffffff }; +VECT_VAR_DECL(expected,uint,64,2) [] = { 0xffffffffffffffff, + 0xffffffffffffffff }; + +/* Expected values of cumulative_saturation flag with negative shift + amount. */ +int VECT_VAR(expected_cumulative_sat_neg,int,8,8) = 0; +int VECT_VAR(expected_cumulative_sat_neg,int,16,4) = 0; +int VECT_VAR(expected_cumulative_sat_neg,int,32,2) = 0; +int VECT_VAR(expected_cumulative_sat_neg,int,64,1) = 0; +int VECT_VAR(expected_cumulative_sat_neg,uint,8,8) = 0; +int VECT_VAR(expected_cumulative_sat_neg,uint,16,4) = 0; +int VECT_VAR(expected_cumulative_sat_neg,uint,32,2) = 0; +int VECT_VAR(expected_cumulative_sat_neg,uint,64,1) = 0; +int VECT_VAR(expected_cumulative_sat_neg,int,8,16) = 0; +int VECT_VAR(expected_cumulative_sat_neg,int,16,8) = 0; +int VECT_VAR(expected_cumulative_sat_neg,int,32,4) = 0; +int VECT_VAR(expected_cumulative_sat_neg,int,64,2) = 0; +int VECT_VAR(expected_cumulative_sat_neg,uint,8,16) = 0; +int VECT_VAR(expected_cumulative_sat_neg,uint,16,8) = 0; +int VECT_VAR(expected_cumulative_sat_neg,uint,32,4) = 0; +int VECT_VAR(expected_cumulative_sat_neg,uint,64,2) = 0; + +/* Expected results with negative shift amount. */ +VECT_VAR_DECL(expected_neg,int,8,8) [] = { 0xfc, 0xfc, 0xfd, 0xfd, + 0xfd, 0xfd, 0xfe, 0xfe }; +VECT_VAR_DECL(expected_neg,int,16,4) [] = { 0xfffc, 0xfffc, 0xfffd, 0xfffd }; +VECT_VAR_DECL(expected_neg,int,32,2) [] = { 0xfffffffe, 0xfffffffe }; +VECT_VAR_DECL(expected_neg,int,64,1) [] = { 0xffffffffffffffff }; +VECT_VAR_DECL(expected_neg,uint,8,8) [] = { 0x3c, 0x3c, 0x3d, 0x3d, + 0x3d, 0x3d, 0x3e, 0x3e }; +VECT_VAR_DECL(expected_neg,uint,16,4) [] = { 0x3ffc, 0x3ffc, 0x3ffd, 0x3ffd }; +VECT_VAR_DECL(expected_neg,uint,32,2) [] = { 0x1ffffffe, 0x1ffffffe }; +VECT_VAR_DECL(expected_neg,uint,64,1) [] = { 0xfffffffffffffff }; +VECT_VAR_DECL(expected_neg,int,8,16) [] = { 0x0, 0x0, 0x0, 0x0, + 0x0, 0x0, 0x0, 0x0, + 0x0, 0x0, 0x0, 0x0, + 0x0, 0x0, 0x0, 0x0 }; +VECT_VAR_DECL(expected_neg,int,16,8) [] = { 0x0, 0x0, 0x0, 0x0, + 0x0, 0x0, 0x0, 0x0 }; +VECT_VAR_DECL(expected_neg,int,32,4) [] = { 0x0, 0x0, 0x0, 0x0 }; +VECT_VAR_DECL(expected_neg,int,64,2) [] = { 0x0, 0x0 }; +VECT_VAR_DECL(expected_neg,uint,8,16) [] = { 0x2, 0x2, 0x2, 0x2, + 0x2, 0x2, 0x2, 0x2, + 0x2, 0x2, 0x2, 0x2, + 0x2, 0x2, 0x2, 0x2 }; +VECT_VAR_DECL(expected_neg,uint,16,8) [] = { 0x20, 0x20, 0x20, 0x20, + 0x20, 0x20, 0x20, 0x20 }; +VECT_VAR_DECL(expected_neg,uint,32,4) [] = { 0x80000, 0x80000, + 0x80000, 0x80000 }; +VECT_VAR_DECL(expected_neg,uint,64,2) [] = { 0x100000000000, 0x100000000000 }; + +/* Expected values of cumulative_saturation flag with input=max and + shift by -1. */ +int VECT_VAR(expected_cumulative_sat_minus1,int,8,8) = 0; +int VECT_VAR(expected_cumulative_sat_minus1,int,16,4) = 0; +int VECT_VAR(expected_cumulative_sat_minus1,int,32,2) = 0; +int VECT_VAR(expected_cumulative_sat_minus1,int,64,1) = 0; +int VECT_VAR(expected_cumulative_sat_minus1,uint,8,8) = 0; +int VECT_VAR(expected_cumulative_sat_minus1,uint,16,4) = 0; +int VECT_VAR(expected_cumulative_sat_minus1,uint,32,2) = 0; +int VECT_VAR(expected_cumulative_sat_minus1,uint,64,1) = 0; +int VECT_VAR(expected_cumulative_sat_minus1,int,8,16) = 0; +int VECT_VAR(expected_cumulative_sat_minus1,int,16,8) = 0; +int VECT_VAR(expected_cumulative_sat_minus1,int,32,4) = 0; +int VECT_VAR(expected_cumulative_sat_minus1,int,64,2) = 0; +int VECT_VAR(expected_cumulative_sat_minus1,uint,8,16) = 0; +int VECT_VAR(expected_cumulative_sat_minus1,uint,16,8) = 0; +int VECT_VAR(expected_cumulative_sat_minus1,uint,32,4) = 0; +int VECT_VAR(expected_cumulative_sat_minus1,uint,64,2) = 0; + +/* Expected results with input=max and shift by -1. */ +VECT_VAR_DECL(expected_minus1,int,8,8) [] = { 0x40, 0x40, 0x40, 0x40, + 0x40, 0x40, 0x40, 0x40 }; +VECT_VAR_DECL(expected_minus1,int,16,4) [] = { 0x4000, 0x4000, 0x4000, 0x4000 }; +VECT_VAR_DECL(expected_minus1,int,32,2) [] = { 0x40000000, 0x40000000 }; +VECT_VAR_DECL(expected_minus1,int,64,1) [] = { 0x4000000000000000 }; +VECT_VAR_DECL(expected_minus1,uint,8,8) [] = { 0x80, 0x80, 0x80, 0x80, + 0x80, 0x80, 0x80, 0x80 }; +VECT_VAR_DECL(expected_minus1,uint,16,4) [] = { 0x8000, 0x8000, 0x8000, 0x8000 }; +VECT_VAR_DECL(expected_minus1,uint,32,2) [] = { 0x80000000, 0x80000000 }; +VECT_VAR_DECL(expected_minus1,uint,64,1) [] = { 0x8000000000000000 }; +VECT_VAR_DECL(expected_minus1,int,8,16) [] = { 0x40, 0x40, 0x40, 0x40, + 0x40, 0x40, 0x40, 0x40, + 0x40, 0x40, 0x40, 0x40, + 0x40, 0x40, 0x40, 0x40 }; +VECT_VAR_DECL(expected_minus1,int,16,8) [] = { 0x4000, 0x4000, 0x4000, 0x4000, + 0x4000, 0x4000, 0x4000, 0x4000 }; +VECT_VAR_DECL(expected_minus1,int,32,4) [] = { 0x40000000, 0x40000000, + 0x40000000, 0x40000000 }; +VECT_VAR_DECL(expected_minus1,int,64,2) [] = { 0x4000000000000000, + 0x4000000000000000 }; +VECT_VAR_DECL(expected_minus1,uint,8,16) [] = { 0x80, 0x80, 0x80, 0x80, + 0x80, 0x80, 0x80, 0x80, + 0x80, 0x80, 0x80, 0x80, + 0x80, 0x80, 0x80, 0x80 }; +VECT_VAR_DECL(expected_minus1,uint,16,8) [] = { 0x8000, 0x8000, 0x8000, 0x8000, + 0x8000, 0x8000, 0x8000, 0x8000 }; +VECT_VAR_DECL(expected_minus1,uint,32,4) [] = { 0x80000000, 0x80000000, + 0x80000000, 0x80000000 }; +VECT_VAR_DECL(expected_minus1,uint,64,2) [] = { 0x8000000000000000, + 0x8000000000000000 }; + +/* Expected values of cumulative_saturation flag with input=max and + shift by -3. */ +int VECT_VAR(expected_cumulative_sat_minus3,int,8,8) = 0; +int VECT_VAR(expected_cumulative_sat_minus3,int,16,4) = 0; +int VECT_VAR(expected_cumulative_sat_minus3,int,32,2) = 0; +int VECT_VAR(expected_cumulative_sat_minus3,int,64,1) = 0; +int VECT_VAR(expected_cumulative_sat_minus3,uint,8,8) = 0; +int VECT_VAR(expected_cumulative_sat_minus3,uint,16,4) = 0; +int VECT_VAR(expected_cumulative_sat_minus3,uint,32,2) = 0; +int VECT_VAR(expected_cumulative_sat_minus3,uint,64,1) = 0; +int VECT_VAR(expected_cumulative_sat_minus3,int,8,16) = 0; +int VECT_VAR(expected_cumulative_sat_minus3,int,16,8) = 0; +int VECT_VAR(expected_cumulative_sat_minus3,int,32,4) = 0; +int VECT_VAR(expected_cumulative_sat_minus3,int,64,2) = 0; +int VECT_VAR(expected_cumulative_sat_minus3,uint,8,16) = 0; +int VECT_VAR(expected_cumulative_sat_minus3,uint,16,8) = 0; +int VECT_VAR(expected_cumulative_sat_minus3,uint,32,4) = 0; +int VECT_VAR(expected_cumulative_sat_minus3,uint,64,2) = 0; + +/* Expected results with input=max and shift by -3. */ +VECT_VAR_DECL(expected_minus3,int,8,8) [] = { 0x10, 0x10, 0x10, 0x10, + 0x10, 0x10, 0x10, 0x10 }; +VECT_VAR_DECL(expected_minus3,int,16,4) [] = { 0x1000, 0x1000, 0x1000, 0x1000 }; +VECT_VAR_DECL(expected_minus3,int,32,2) [] = { 0x10000000, 0x10000000 }; +VECT_VAR_DECL(expected_minus3,int,64,1) [] = { 0x1000000000000000 }; +VECT_VAR_DECL(expected_minus3,uint,8,8) [] = { 0x20, 0x20, 0x20, 0x20, + 0x20, 0x20, 0x20, 0x20 }; +VECT_VAR_DECL(expected_minus3,uint,16,4) [] = { 0x2000, 0x2000, 0x2000, 0x2000 }; +VECT_VAR_DECL(expected_minus3,uint,32,2) [] = { 0x20000000, 0x20000000 }; +VECT_VAR_DECL(expected_minus3,uint,64,1) [] = { 0x2000000000000000 }; +VECT_VAR_DECL(expected_minus3,int,8,16) [] = { 0x10, 0x10, 0x10, 0x10, + 0x10, 0x10, 0x10, 0x10, + 0x10, 0x10, 0x10, 0x10, + 0x10, 0x10, 0x10, 0x10 }; +VECT_VAR_DECL(expected_minus3,int,16,8) [] = { 0x1000, 0x1000, 0x1000, 0x1000, + 0x1000, 0x1000, 0x1000, 0x1000 }; +VECT_VAR_DECL(expected_minus3,int,32,4) [] = { 0x10000000, 0x10000000, + 0x10000000, 0x10000000 }; +VECT_VAR_DECL(expected_minus3,int,64,2) [] = { 0x1000000000000000, + 0x1000000000000000 }; +VECT_VAR_DECL(expected_minus3,uint,8,16) [] = { 0x20, 0x20, 0x20, 0x20, + 0x20, 0x20, 0x20, 0x20, + 0x20, 0x20, 0x20, 0x20, + 0x20, 0x20, 0x20, 0x20 }; +VECT_VAR_DECL(expected_minus3,uint,16,8) [] = { 0x2000, 0x2000, 0x2000, 0x2000, + 0x2000, 0x2000, 0x2000, 0x2000 }; +VECT_VAR_DECL(expected_minus3,uint,32,4) [] = { 0x20000000, 0x20000000, + 0x20000000, 0x20000000 }; +VECT_VAR_DECL(expected_minus3,uint,64,2) [] = { 0x2000000000000000, + 0x2000000000000000 }; + +/* Expected values of cumulative_saturation flag with input=max and + large shift amount. */ +int VECT_VAR(expected_cumulative_sat_large_sh,int,8,8) = 1; +int VECT_VAR(expected_cumulative_sat_large_sh,int,16,4) = 1; +int VECT_VAR(expected_cumulative_sat_large_sh,int,32,2) = 1; +int VECT_VAR(expected_cumulative_sat_large_sh,int,64,1) = 1; +int VECT_VAR(expected_cumulative_sat_large_sh,uint,8,8) = 1; +int VECT_VAR(expected_cumulative_sat_large_sh,uint,16,4) = 1; +int VECT_VAR(expected_cumulative_sat_large_sh,uint,32,2) = 1; +int VECT_VAR(expected_cumulative_sat_large_sh,uint,64,1) = 1; +int VECT_VAR(expected_cumulative_sat_large_sh,int,8,16) = 1; +int VECT_VAR(expected_cumulative_sat_large_sh,int,16,8) = 1; +int VECT_VAR(expected_cumulative_sat_large_sh,int,32,4) = 1; +int VECT_VAR(expected_cumulative_sat_large_sh,int,64,2) = 1; +int VECT_VAR(expected_cumulative_sat_large_sh,uint,8,16) = 1; +int VECT_VAR(expected_cumulative_sat_large_sh,uint,16,8) = 1; +int VECT_VAR(expected_cumulative_sat_large_sh,uint,32,4) = 1; +int VECT_VAR(expected_cumulative_sat_large_sh,uint,64,2) = 1; + +/* Expected results with input=max and large shift amount. */ +VECT_VAR_DECL(expected_large_sh,int,8,8) [] = { 0x7f, 0x7f, 0x7f, 0x7f, + 0x7f, 0x7f, 0x7f, 0x7f }; +VECT_VAR_DECL(expected_large_sh,int,16,4) [] = { 0x7fff, 0x7fff, + 0x7fff, 0x7fff }; +VECT_VAR_DECL(expected_large_sh,int,32,2) [] = { 0x7fffffff, 0x7fffffff }; +VECT_VAR_DECL(expected_large_sh,int,64,1) [] = { 0x7fffffffffffffff }; +VECT_VAR_DECL(expected_large_sh,uint,8,8) [] = { 0xff, 0xff, 0xff, 0xff, + 0xff, 0xff, 0xff, 0xff }; +VECT_VAR_DECL(expected_large_sh,uint,16,4) [] = { 0xffff, 0xffff, + 0xffff, 0xffff }; +VECT_VAR_DECL(expected_large_sh,uint,32,2) [] = { 0xffffffff, 0xffffffff }; +VECT_VAR_DECL(expected_large_sh,uint,64,1) [] = { 0xffffffffffffffff }; +VECT_VAR_DECL(expected_large_sh,int,8,16) [] = { 0x7f, 0x7f, 0x7f, 0x7f, + 0x7f, 0x7f, 0x7f, 0x7f, + 0x7f, 0x7f, 0x7f, 0x7f, + 0x7f, 0x7f, 0x7f, 0x7f }; +VECT_VAR_DECL(expected_large_sh,int,16,8) [] = { 0x7fff, 0x7fff, + 0x7fff, 0x7fff, + 0x7fff, 0x7fff, + 0x7fff, 0x7fff }; +VECT_VAR_DECL(expected_large_sh,int,32,4) [] = { 0x7fffffff, 0x7fffffff, + 0x7fffffff, 0x7fffffff }; +VECT_VAR_DECL(expected_large_sh,int,64,2) [] = { 0x7fffffffffffffff, + 0x7fffffffffffffff }; +VECT_VAR_DECL(expected_large_sh,uint,8,16) [] = { 0xff, 0xff, 0xff, 0xff, + 0xff, 0xff, 0xff, 0xff, + 0xff, 0xff, 0xff, 0xff, + 0xff, 0xff, 0xff, 0xff }; +VECT_VAR_DECL(expected_large_sh,uint,16,8) [] = { 0xffff, 0xffff, + 0xffff, 0xffff, + 0xffff, 0xffff, + 0xffff, 0xffff }; +VECT_VAR_DECL(expected_large_sh,uint,32,4) [] = { 0xffffffff, 0xffffffff, + 0xffffffff, 0xffffffff }; +VECT_VAR_DECL(expected_large_sh,uint,64,2) [] = { 0xffffffffffffffff, + 0xffffffffffffffff }; + +/* Expected values of cumulative_saturation flag with negative input and + large shift amount. */ +int VECT_VAR(expected_cumulative_sat_neg_large_sh,int,8,8) = 1; +int VECT_VAR(expected_cumulative_sat_neg_large_sh,int,16,4) = 1; +int VECT_VAR(expected_cumulative_sat_neg_large_sh,int,32,2) = 1; +int VECT_VAR(expected_cumulative_sat_neg_large_sh,int,64,1) = 1; +int VECT_VAR(expected_cumulative_sat_neg_large_sh,uint,8,8) = 1; +int VECT_VAR(expected_cumulative_sat_neg_large_sh,uint,16,4) = 1; +int VECT_VAR(expected_cumulative_sat_neg_large_sh,uint,32,2) = 1; +int VECT_VAR(expected_cumulative_sat_neg_large_sh,uint,64,1) = 1; +int VECT_VAR(expected_cumulative_sat_neg_large_sh,int,8,16) = 1; +int VECT_VAR(expected_cumulative_sat_neg_large_sh,int,16,8) = 1; +int VECT_VAR(expected_cumulative_sat_neg_large_sh,int,32,4) = 1; +int VECT_VAR(expected_cumulative_sat_neg_large_sh,int,64,2) = 1; +int VECT_VAR(expected_cumulative_sat_neg_large_sh,uint,8,16) = 1; +int VECT_VAR(expected_cumulative_sat_neg_large_sh,uint,16,8) = 1; +int VECT_VAR(expected_cumulative_sat_neg_large_sh,uint,32,4) = 1; +int VECT_VAR(expected_cumulative_sat_neg_large_sh,uint,64,2) = 1; + +/* Expected results with negative input and large shift amount. */ +VECT_VAR_DECL(expected_neg_large_sh,int,8,8) [] = { 0x80, 0x80, 0x80, 0x80, + 0x80, 0x80, 0x80, 0x80 }; +VECT_VAR_DECL(expected_neg_large_sh,int,16,4) [] = { 0x8000, 0x8000, + 0x8000, 0x8000 }; +VECT_VAR_DECL(expected_neg_large_sh,int,32,2) [] = { 0x80000000, 0x80000000 }; +VECT_VAR_DECL(expected_neg_large_sh,int,64,1) [] = { 0x8000000000000000 }; +VECT_VAR_DECL(expected_neg_large_sh,uint,8,8) [] = { 0xff, 0xff, 0xff, 0xff, + 0xff, 0xff, 0xff, 0xff }; +VECT_VAR_DECL(expected_neg_large_sh,uint,16,4) [] = { 0xffff, 0xffff, + 0xffff, 0xffff }; +VECT_VAR_DECL(expected_neg_large_sh,uint,32,2) [] = { 0xffffffff, + 0xffffffff }; +VECT_VAR_DECL(expected_neg_large_sh,uint,64,1) [] = { 0xffffffffffffffff }; +VECT_VAR_DECL(expected_neg_large_sh,int,8,16) [] = { 0x80, 0x80, 0x80, 0x80, + 0x80, 0x80, 0x80, 0x80, + 0x80, 0x80, 0x80, 0x80, + 0x80, 0x80, 0x80, 0x80 }; +VECT_VAR_DECL(expected_neg_large_sh,int,16,8) [] = { 0x8000, 0x8000, + 0x8000, 0x8000, + 0x8000, 0x8000, + 0x8000, 0x8000 }; +VECT_VAR_DECL(expected_neg_large_sh,int,32,4) [] = { 0x80000000, 0x80000000, + 0x80000000, 0x80000000 }; +VECT_VAR_DECL(expected_neg_large_sh,int,64,2) [] = { 0x8000000000000000, + 0x8000000000000000 }; +VECT_VAR_DECL(expected_neg_large_sh,uint,8,16) [] = { 0xff, 0xff, 0xff, 0xff, + 0xff, 0xff, 0xff, 0xff, + 0xff, 0xff, 0xff, 0xff, + 0xff, 0xff, 0xff, 0xff }; +VECT_VAR_DECL(expected_neg_large_sh,uint,16,8) [] = { 0xffff, 0xffff, + 0xffff, 0xffff, + 0xffff, 0xffff, + 0xffff, 0xffff }; +VECT_VAR_DECL(expected_neg_large_sh,uint,32,4) [] = { 0xffffffff, + 0xffffffff, + 0xffffffff, + 0xffffffff }; +VECT_VAR_DECL(expected_neg_large_sh,uint,64,2) [] = { 0xffffffffffffffff, + 0xffffffffffffffff }; + +/* Expected values of cumulative_saturation flag with max/min input and + large negative shift amount. */ +int VECT_VAR(expected_cumulative_sat_large_neg_sh,int,8,8) = 0; +int VECT_VAR(expected_cumulative_sat_large_neg_sh,int,16,4) = 0; +int VECT_VAR(expected_cumulative_sat_large_neg_sh,int,32,2) = 0; +int VECT_VAR(expected_cumulative_sat_large_neg_sh,int,64,1) = 0; +int VECT_VAR(expected_cumulative_sat_large_neg_sh,uint,8,8) = 0; +int VECT_VAR(expected_cumulative_sat_large_neg_sh,uint,16,4) = 0; +int VECT_VAR(expected_cumulative_sat_large_neg_sh,uint,32,2) = 0; +int VECT_VAR(expected_cumulative_sat_large_neg_sh,uint,64,1) = 0; +int VECT_VAR(expected_cumulative_sat_large_neg_sh,int,8,16) = 0; +int VECT_VAR(expected_cumulative_sat_large_neg_sh,int,16,8) = 0; +int VECT_VAR(expected_cumulative_sat_large_neg_sh,int,32,4) = 0; +int VECT_VAR(expected_cumulative_sat_large_neg_sh,int,64,2) = 0; +int VECT_VAR(expected_cumulative_sat_large_neg_sh,uint,8,16) = 0; +int VECT_VAR(expected_cumulative_sat_large_neg_sh,uint,16,8) = 0; +int VECT_VAR(expected_cumulative_sat_large_neg_sh,uint,32,4) = 0; +int VECT_VAR(expected_cumulative_sat_large_neg_sh,uint,64,2) = 0; + +/* Expected results with max/min input and large negative shift amount. */ +VECT_VAR_DECL(expected_large_neg_sh,int,8,8) [] = { 0x0, 0x0, 0x0, 0x0, + 0x0, 0x0, 0x0, 0x0 }; +VECT_VAR_DECL(expected_large_neg_sh,int,16,4) [] = { 0x0, 0x0, 0x0, 0x0 }; +VECT_VAR_DECL(expected_large_neg_sh,int,32,2) [] = { 0x0, 0x0 }; +VECT_VAR_DECL(expected_large_neg_sh,int,64,1) [] = { 0x0 }; +VECT_VAR_DECL(expected_large_neg_sh,uint,8,8) [] = { 0x0, 0x0, 0x0, 0x0, + 0x0, 0x0, 0x0, 0x0 }; +VECT_VAR_DECL(expected_large_neg_sh,uint,16,4) [] = { 0x0, 0x0, 0x0, 0x0 }; +VECT_VAR_DECL(expected_large_neg_sh,uint,32,2) [] = { 0x0, 0x0 }; +VECT_VAR_DECL(expected_large_neg_sh,uint,64,1) [] = { 0x0 }; +VECT_VAR_DECL(expected_large_neg_sh,int,8,16) [] = { 0x0, 0x0, 0x0, 0x0, + 0x0, 0x0, 0x0, 0x0, + 0x0, 0x0, 0x0, 0x0, + 0x0, 0x0, 0x0, 0x0 }; +VECT_VAR_DECL(expected_large_neg_sh,int,16,8) [] = { 0x0, 0x0, 0x0, 0x0, + 0x0, 0x0, 0x0, 0x0 }; +VECT_VAR_DECL(expected_large_neg_sh,int,32,4) [] = { 0x0, 0x0, 0x0, 0x0 }; +VECT_VAR_DECL(expected_large_neg_sh,int,64,2) [] = { 0x0, 0x0 }; +VECT_VAR_DECL(expected_large_neg_sh,uint,8,16) [] = { 0x0, 0x0, 0x0, 0x0, + 0x0, 0x0, 0x0, 0x0, + 0x0, 0x0, 0x0, 0x0, + 0x0, 0x0, 0x0, 0x0 }; +VECT_VAR_DECL(expected_large_neg_sh,uint,16,8) [] = { 0x0, 0x0, 0x0, 0x0, + 0x0, 0x0, 0x0, 0x0 }; +VECT_VAR_DECL(expected_large_neg_sh,uint,32,4) [] = { 0x0, 0x0, 0x0, 0x0 }; +VECT_VAR_DECL(expected_large_neg_sh,uint,64,2) [] = { 0x0, 0x0 }; + +/* Expected values of cumulative_saturation flag with input=0 and + large negative shift amount. */ +int VECT_VAR(expected_cumulative_sat_0_large_neg_sh,int,8,8) = 0; +int VECT_VAR(expected_cumulative_sat_0_large_neg_sh,int,16,4) = 0; +int VECT_VAR(expected_cumulative_sat_0_large_neg_sh,int,32,2) = 0; +int VECT_VAR(expected_cumulative_sat_0_large_neg_sh,int,64,1) = 0; +int VECT_VAR(expected_cumulative_sat_0_large_neg_sh,uint,8,8) = 0; +int VECT_VAR(expected_cumulative_sat_0_large_neg_sh,uint,16,4) = 0; +int VECT_VAR(expected_cumulative_sat_0_large_neg_sh,uint,32,2) = 0; +int VECT_VAR(expected_cumulative_sat_0_large_neg_sh,uint,64,1) = 0; +int VECT_VAR(expected_cumulative_sat_0_large_neg_sh,int,8,16) = 0; +int VECT_VAR(expected_cumulative_sat_0_large_neg_sh,int,16,8) = 0; +int VECT_VAR(expected_cumulative_sat_0_large_neg_sh,int,32,4) = 0; +int VECT_VAR(expected_cumulative_sat_0_large_neg_sh,int,64,2) = 0; +int VECT_VAR(expected_cumulative_sat_0_large_neg_sh,uint,8,16) = 0; +int VECT_VAR(expected_cumulative_sat_0_large_neg_sh,uint,16,8) = 0; +int VECT_VAR(expected_cumulative_sat_0_large_neg_sh,uint,32,4) = 0; +int VECT_VAR(expected_cumulative_sat_0_large_neg_sh,uint,64,2) = 0; + +/* Expected results with input=0 and large negative shift amount. */ +VECT_VAR_DECL(expected_0_large_neg_sh,int,8,8) [] = { 0x0, 0x0, 0x0, 0x0, + 0x0, 0x0, 0x0, 0x0 }; +VECT_VAR_DECL(expected_0_large_neg_sh,int,16,4) [] = { 0x0, 0x0, 0x0, 0x0 }; +VECT_VAR_DECL(expected_0_large_neg_sh,int,32,2) [] = { 0x0, 0x0 }; +VECT_VAR_DECL(expected_0_large_neg_sh,int,64,1) [] = { 0x0 }; +VECT_VAR_DECL(expected_0_large_neg_sh,uint,8,8) [] = { 0x0, 0x0, 0x0, 0x0, + 0x0, 0x0, 0x0, 0x0 }; +VECT_VAR_DECL(expected_0_large_neg_sh,uint,16,4) [] = { 0x0, 0x0, 0x0, 0x0 }; +VECT_VAR_DECL(expected_0_large_neg_sh,uint,32,2) [] = { 0x0, 0x0 }; +VECT_VAR_DECL(expected_0_large_neg_sh,uint,64,1) [] = { 0x0 }; +VECT_VAR_DECL(expected_0_large_neg_sh,int,8,16) [] = { 0x0, 0x0, 0x0, 0x0, + 0x0, 0x0, 0x0, 0x0, + 0x0, 0x0, 0x0, 0x0, + 0x0, 0x0, 0x0, 0x0 }; +VECT_VAR_DECL(expected_0_large_neg_sh,int,16,8) [] = { 0x0, 0x0, 0x0, 0x0, + 0x0, 0x0, 0x0, 0x0 }; +VECT_VAR_DECL(expected_0_large_neg_sh,int,32,4) [] = { 0x0, 0x0, 0x0, 0x0 }; +VECT_VAR_DECL(expected_0_large_neg_sh,int,64,2) [] = { 0x0, 0x0 }; +VECT_VAR_DECL(expected_0_large_neg_sh,uint,8,16) [] = { 0x0, 0x0, 0x0, 0x0, + 0x0, 0x0, 0x0, 0x0, + 0x0, 0x0, 0x0, 0x0, + 0x0, 0x0, 0x0, 0x0 }; +VECT_VAR_DECL(expected_0_large_neg_sh,uint,16,8) [] = { 0x0, 0x0, 0x0, 0x0, + 0x0, 0x0, 0x0, 0x0 }; +VECT_VAR_DECL(expected_0_large_neg_sh,uint,32,4) [] = { 0x0, 0x0, 0x0, 0x0 }; +VECT_VAR_DECL(expected_0_large_neg_sh,uint,64,2) [] = { 0x0, 0x0 }; + +#define INSN vqrshl +#define TEST_MSG "VQRSHL/VQRSHLQ" + +#define FNNAME1(NAME) void exec_ ## NAME (void) +#define FNNAME(NAME) FNNAME1(NAME) + +FNNAME (INSN) +{ + /* Basic test: v3=vqrshl(v1,v2), then store the result. */ +#define TEST_VQRSHL2(INSN, T3, Q, T1, T2, W, N, EXPECTED_CUMULATIVE_SAT, CMT) \ + Set_Neon_Cumulative_Sat(0, VECT_VAR(vector_res, T1, W, N)); \ + VECT_VAR(vector_res, T1, W, N) = \ + INSN##Q##_##T2##W(VECT_VAR(vector, T1, W, N), \ + VECT_VAR(vector_shift, T3, W, N)); \ + vst1##Q##_##T2##W(VECT_VAR(result, T1, W, N), \ + VECT_VAR(vector_res, T1, W, N)); \ + CHECK_CUMULATIVE_SAT(TEST_MSG, T1, W, N, EXPECTED_CUMULATIVE_SAT, CMT) + + /* Two auxliary macros are necessary to expand INSN */ +#define TEST_VQRSHL1(INSN, T3, Q, T1, T2, W, N, EXPECTED_CUMULATIVE_SAT, CMT) \ + TEST_VQRSHL2(INSN, T3, Q, T1, T2, W, N, EXPECTED_CUMULATIVE_SAT, CMT) + +#define TEST_VQRSHL(T3, Q, T1, T2, W, N, EXPECTED_CUMULATIVE_SAT, CMT) \ + TEST_VQRSHL1(INSN, T3, Q, T1, T2, W, N, EXPECTED_CUMULATIVE_SAT, CMT) + + DECL_VARIABLE_ALL_VARIANTS(vector); + DECL_VARIABLE_ALL_VARIANTS(vector_res); + + DECL_VARIABLE_SIGNED_VARIANTS(vector_shift); + + clean_results (); + + /* Fill input vector with 0, to check saturation on limits. */ + VDUP(vector, , int, s, 8, 8, 0); + VDUP(vector, , int, s, 16, 4, 0); + VDUP(vector, , int, s, 32, 2, 0); + VDUP(vector, , int, s, 64, 1, 0); + VDUP(vector, , uint, u, 8, 8, 0); + VDUP(vector, , uint, u, 16, 4, 0); + VDUP(vector, , uint, u, 32, 2, 0); + VDUP(vector, , uint, u, 64, 1, 0); + VDUP(vector, q, int, s, 8, 16, 0); + VDUP(vector, q, int, s, 16, 8, 0); + VDUP(vector, q, int, s, 32, 4, 0); + VDUP(vector, q, int, s, 64, 2, 0); + VDUP(vector, q, uint, u, 8, 16, 0); + VDUP(vector, q, uint, u, 16, 8, 0); + VDUP(vector, q, uint, u, 32, 4, 0); + VDUP(vector, q, uint, u, 64, 2, 0); + + /* Choose init value arbitrarily, will be used as shift amount */ + /* Use values equal to or one-less-than the type width to check + behaviour on limits. */ + VDUP(vector_shift, , int, s, 8, 8, 7); + VDUP(vector_shift, , int, s, 16, 4, 15); + VDUP(vector_shift, , int, s, 32, 2, 31); + VDUP(vector_shift, , int, s, 64, 1, 63); + VDUP(vector_shift, q, int, s, 8, 16, 8); + VDUP(vector_shift, q, int, s, 16, 8, 16); + VDUP(vector_shift, q, int, s, 32, 4, 32); + VDUP(vector_shift, q, int, s, 64, 2, 64); + +#define CMT " (with input = 0)" + TEST_VQRSHL(int, , int, s, 8, 8, expected_cumulative_sat_0, CMT); + TEST_VQRSHL(int, , int, s, 16, 4, expected_cumulative_sat_0, CMT); + TEST_VQRSHL(int, , int, s, 32, 2, expected_cumulative_sat_0, CMT); + TEST_VQRSHL(int, , int, s, 64, 1, expected_cumulative_sat_0, CMT); + TEST_VQRSHL(int, , uint, u, 8, 8, expected_cumulative_sat_0, CMT); + TEST_VQRSHL(int, , uint, u, 16, 4, expected_cumulative_sat_0, CMT); + TEST_VQRSHL(int, , uint, u, 32, 2, expected_cumulative_sat_0, CMT); + TEST_VQRSHL(int, , uint, u, 64, 1, expected_cumulative_sat_0, CMT); + TEST_VQRSHL(int, q, int, s, 8, 16, expected_cumulative_sat_0, CMT); + TEST_VQRSHL(int, q, int, s, 16, 8, expected_cumulative_sat_0, CMT); + TEST_VQRSHL(int, q, int, s, 32, 4, expected_cumulative_sat_0, CMT); + TEST_VQRSHL(int, q, int, s, 64, 2, expected_cumulative_sat_0, CMT); + TEST_VQRSHL(int, q, uint, u, 8, 16, expected_cumulative_sat_0, CMT); + TEST_VQRSHL(int, q, uint, u, 16, 8, expected_cumulative_sat_0, CMT); + TEST_VQRSHL(int, q, uint, u, 32, 4, expected_cumulative_sat_0, CMT); + TEST_VQRSHL(int, q, uint, u, 64, 2, expected_cumulative_sat_0, CMT); + + CHECK(TEST_MSG, int, 8, 8, PRIx8, expected_0, CMT); + CHECK(TEST_MSG, int, 16, 4, PRIx16, expected_0, CMT); + CHECK(TEST_MSG, int, 32, 2, PRIx32, expected_0, CMT); + CHECK(TEST_MSG, int, 64, 1, PRIx64, expected_0, CMT); + CHECK(TEST_MSG, uint, 8, 8, PRIx8, expected_0, CMT); + CHECK(TEST_MSG, uint, 16, 4, PRIx16, expected_0, CMT); + CHECK(TEST_MSG, uint, 32, 2, PRIx32, expected_0, CMT); + CHECK(TEST_MSG, uint, 64, 1, PRIx64, expected_0, CMT); + CHECK(TEST_MSG, int, 8, 16, PRIx8, expected_0, CMT); + CHECK(TEST_MSG, int, 16, 8, PRIx16, expected_0, CMT); + CHECK(TEST_MSG, int, 32, 4, PRIx32, expected_0, CMT); + CHECK(TEST_MSG, int, 64, 2, PRIx64, expected_0, CMT); + CHECK(TEST_MSG, uint, 8, 16, PRIx8, expected_0, CMT); + CHECK(TEST_MSG, uint, 16, 8, PRIx16, expected_0, CMT); + CHECK(TEST_MSG, uint, 32, 4, PRIx32, expected_0, CMT); + CHECK(TEST_MSG, uint, 64, 2, PRIx64, expected_0, CMT); + + + /* Use negative shift amounts. */ + VDUP(vector_shift, , int, s, 8, 8, -1); + VDUP(vector_shift, , int, s, 16, 4, -2); + VDUP(vector_shift, , int, s, 32, 2, -3); + VDUP(vector_shift, , int, s, 64, 1, -4); + VDUP(vector_shift, q, int, s, 8, 16, -7); + VDUP(vector_shift, q, int, s, 16, 8, -11); + VDUP(vector_shift, q, int, s, 32, 4, -13); + VDUP(vector_shift, q, int, s, 64, 2, -20); + +#undef CMT +#define CMT " (input 0 and negative shift amount)" + TEST_VQRSHL(int, , int, s, 8, 8, expected_cumulative_sat_0_neg, CMT); + TEST_VQRSHL(int, , int, s, 16, 4, expected_cumulative_sat_0_neg, CMT); + TEST_VQRSHL(int, , int, s, 32, 2, expected_cumulative_sat_0_neg, CMT); + TEST_VQRSHL(int, , int, s, 64, 1, expected_cumulative_sat_0_neg, CMT); + TEST_VQRSHL(int, , uint, u, 8, 8, expected_cumulative_sat_0_neg, CMT); + TEST_VQRSHL(int, , uint, u, 16, 4, expected_cumulative_sat_0_neg, CMT); + TEST_VQRSHL(int, , uint, u, 32, 2, expected_cumulative_sat_0_neg, CMT); + TEST_VQRSHL(int, , uint, u, 64, 1, expected_cumulative_sat_0_neg, CMT); + TEST_VQRSHL(int, q, int, s, 8, 16, expected_cumulative_sat_0_neg, CMT); + TEST_VQRSHL(int, q, int, s, 16, 8, expected_cumulative_sat_0_neg, CMT); + TEST_VQRSHL(int, q, int, s, 32, 4, expected_cumulative_sat_0_neg, CMT); + TEST_VQRSHL(int, q, int, s, 64, 2, expected_cumulative_sat_0_neg, CMT); + TEST_VQRSHL(int, q, uint, u, 8, 16, expected_cumulative_sat_0_neg, CMT); + TEST_VQRSHL(int, q, uint, u, 16, 8, expected_cumulative_sat_0_neg, CMT); + TEST_VQRSHL(int, q, uint, u, 32, 4, expected_cumulative_sat_0_neg, CMT); + TEST_VQRSHL(int, q, uint, u, 64, 2, expected_cumulative_sat_0_neg, CMT); + + CHECK(TEST_MSG, int, 8, 8, PRIx8, expected_0_neg, CMT); + CHECK(TEST_MSG, int, 16, 4, PRIx16, expected_0_neg, CMT); + CHECK(TEST_MSG, int, 32, 2, PRIx32, expected_0_neg, CMT); + CHECK(TEST_MSG, int, 64, 1, PRIx64, expected_0_neg, CMT); + CHECK(TEST_MSG, uint, 8, 8, PRIx8, expected_0_neg, CMT); + CHECK(TEST_MSG, uint, 16, 4, PRIx16, expected_0_neg, CMT); + CHECK(TEST_MSG, uint, 32, 2, PRIx32, expected_0_neg, CMT); + CHECK(TEST_MSG, uint, 64, 1, PRIx64, expected_0_neg, CMT); + CHECK(TEST_MSG, int, 8, 16, PRIx8, expected_0_neg, CMT); + CHECK(TEST_MSG, int, 16, 8, PRIx16, expected_0_neg, CMT); + CHECK(TEST_MSG, int, 32, 4, PRIx32, expected_0_neg, CMT); + CHECK(TEST_MSG, int, 64, 2, PRIx64, expected_0_neg, CMT); + CHECK(TEST_MSG, uint, 8, 16, PRIx8, expected_0_neg, CMT); + CHECK(TEST_MSG, uint, 16, 8, PRIx16, expected_0_neg, CMT); + CHECK(TEST_MSG, uint, 32, 4, PRIx32, expected_0_neg, CMT); + CHECK(TEST_MSG, uint, 64, 2, PRIx64, expected_0_neg, CMT); + + + /* Test again, with predefined input values. */ + TEST_MACRO_ALL_VARIANTS_2_5(VLOAD, vector, buffer); + + /* Choose init value arbitrarily, will be used as shift amount. */ + VDUP(vector_shift, , int, s, 8, 8, 1); + VDUP(vector_shift, , int, s, 16, 4, 3); + VDUP(vector_shift, , int, s, 32, 2, 8); + VDUP(vector_shift, , int, s, 64, 1, 3); + VDUP(vector_shift, q, int, s, 8, 16, 10); + VDUP(vector_shift, q, int, s, 16, 8, 12); + VDUP(vector_shift, q, int, s, 32, 4, 31); + VDUP(vector_shift, q, int, s, 64, 2, 63); + +#undef CMT +#define CMT "" + TEST_VQRSHL(int, , int, s, 8, 8, expected_cumulative_sat, CMT); + TEST_VQRSHL(int, , int, s, 16, 4, expected_cumulative_sat, CMT); + TEST_VQRSHL(int, , int, s, 32, 2, expected_cumulative_sat, CMT); + TEST_VQRSHL(int, , int, s, 64, 1, expected_cumulative_sat, CMT); + TEST_VQRSHL(int, , uint, u, 8, 8, expected_cumulative_sat, CMT); + TEST_VQRSHL(int, , uint, u, 16, 4, expected_cumulative_sat, CMT); + TEST_VQRSHL(int, , uint, u, 32, 2, expected_cumulative_sat, CMT); + TEST_VQRSHL(int, , uint, u, 64, 1, expected_cumulative_sat, CMT); + TEST_VQRSHL(int, q, int, s, 8, 16, expected_cumulative_sat, CMT); + TEST_VQRSHL(int, q, int, s, 16, 8, expected_cumulative_sat, CMT); + TEST_VQRSHL(int, q, int, s, 32, 4, expected_cumulative_sat, CMT); + TEST_VQRSHL(int, q, int, s, 64, 2, expected_cumulative_sat, CMT); + TEST_VQRSHL(int, q, uint, u, 8, 16, expected_cumulative_sat, CMT); + TEST_VQRSHL(int, q, uint, u, 16, 8, expected_cumulative_sat, CMT); + TEST_VQRSHL(int, q, uint, u, 32, 4, expected_cumulative_sat, CMT); + TEST_VQRSHL(int, q, uint, u, 64, 2, expected_cumulative_sat, CMT); + + CHECK(TEST_MSG, int, 8, 8, PRIx8, expected, CMT); + CHECK(TEST_MSG, int, 16, 4, PRIx16, expected, CMT); + CHECK(TEST_MSG, int, 32, 2, PRIx32, expected, CMT); + CHECK(TEST_MSG, int, 64, 1, PRIx64, expected, CMT); + CHECK(TEST_MSG, uint, 8, 8, PRIx8, expected, CMT); + CHECK(TEST_MSG, uint, 16, 4, PRIx16, expected, CMT); + CHECK(TEST_MSG, uint, 32, 2, PRIx32, expected, CMT); + CHECK(TEST_MSG, uint, 64, 1, PRIx64, expected, CMT); + CHECK(TEST_MSG, int, 8, 16, PRIx8, expected, CMT); + CHECK(TEST_MSG, int, 16, 8, PRIx16, expected, CMT); + CHECK(TEST_MSG, int, 32, 4, PRIx32, expected, CMT); + CHECK(TEST_MSG, int, 64, 2, PRIx64, expected, CMT); + CHECK(TEST_MSG, uint, 8, 16, PRIx8, expected, CMT); + CHECK(TEST_MSG, uint, 16, 8, PRIx16, expected, CMT); + CHECK(TEST_MSG, uint, 32, 4, PRIx32, expected, CMT); + CHECK(TEST_MSG, uint, 64, 2, PRIx64, expected, CMT); + + + /* Use negative shift amounts. */ + VDUP(vector_shift, , int, s, 8, 8, -2); + VDUP(vector_shift, , int, s, 16, 4, -2); + VDUP(vector_shift, , int, s, 32, 2, -3); + VDUP(vector_shift, , int, s, 64, 1, -4); + VDUP(vector_shift, q, int, s, 8, 16, -7); + VDUP(vector_shift, q, int, s, 16, 8, -11); + VDUP(vector_shift, q, int, s, 32, 4, -13); + VDUP(vector_shift, q, int, s, 64, 2, -20); + +#undef CMT +#define CMT " (negative shift amount)" + TEST_VQRSHL(int, , int, s, 8, 8, expected_cumulative_sat_neg, CMT); + TEST_VQRSHL(int, , int, s, 16, 4, expected_cumulative_sat_neg, CMT); + TEST_VQRSHL(int, , int, s, 32, 2, expected_cumulative_sat_neg, CMT); + TEST_VQRSHL(int, , int, s, 64, 1, expected_cumulative_sat_neg, CMT); + TEST_VQRSHL(int, , uint, u, 8, 8, expected_cumulative_sat_neg, CMT); + TEST_VQRSHL(int, , uint, u, 16, 4, expected_cumulative_sat_neg, CMT); + TEST_VQRSHL(int, , uint, u, 32, 2, expected_cumulative_sat_neg, CMT); + TEST_VQRSHL(int, , uint, u, 64, 1, expected_cumulative_sat_neg, CMT); + TEST_VQRSHL(int, q, int, s, 8, 16, expected_cumulative_sat_neg, CMT); + TEST_VQRSHL(int, q, int, s, 16, 8, expected_cumulative_sat_neg, CMT); + TEST_VQRSHL(int, q, int, s, 32, 4, expected_cumulative_sat_neg, CMT); + TEST_VQRSHL(int, q, int, s, 64, 2, expected_cumulative_sat_neg, CMT); + TEST_VQRSHL(int, q, uint, u, 8, 16, expected_cumulative_sat_neg, CMT); + TEST_VQRSHL(int, q, uint, u, 16, 8, expected_cumulative_sat_neg, CMT); + TEST_VQRSHL(int, q, uint, u, 32, 4, expected_cumulative_sat_neg, CMT); + TEST_VQRSHL(int, q, uint, u, 64, 2, expected_cumulative_sat_neg, CMT); + + CHECK(TEST_MSG, int, 8, 8, PRIx8, expected_neg, CMT); + CHECK(TEST_MSG, int, 16, 4, PRIx16, expected_neg, CMT); + CHECK(TEST_MSG, int, 32, 2, PRIx32, expected_neg, CMT); + CHECK(TEST_MSG, int, 64, 1, PRIx64, expected_neg, CMT); + CHECK(TEST_MSG, uint, 8, 8, PRIx8, expected_neg, CMT); + CHECK(TEST_MSG, uint, 16, 4, PRIx16, expected_neg, CMT); + CHECK(TEST_MSG, uint, 32, 2, PRIx32, expected_neg, CMT); + CHECK(TEST_MSG, uint, 64, 1, PRIx64, expected_neg, CMT); + CHECK(TEST_MSG, int, 8, 16, PRIx8, expected_neg, CMT); + CHECK(TEST_MSG, int, 16, 8, PRIx16, expected_neg, CMT); + CHECK(TEST_MSG, int, 32, 4, PRIx32, expected_neg, CMT); + CHECK(TEST_MSG, int, 64, 2, PRIx64, expected_neg, CMT); + CHECK(TEST_MSG, uint, 8, 16, PRIx8, expected_neg, CMT); + CHECK(TEST_MSG, uint, 16, 8, PRIx16, expected_neg, CMT); + CHECK(TEST_MSG, uint, 32, 4, PRIx32, expected_neg, CMT); + CHECK(TEST_MSG, uint, 64, 2, PRIx64, expected_neg, CMT); + + + /* Fill input vector with max value, to check saturation on + limits. */ + VDUP(vector, , int, s, 8, 8, 0x7F); + VDUP(vector, , int, s, 16, 4, 0x7FFF); + VDUP(vector, , int, s, 32, 2, 0x7FFFFFFF); + VDUP(vector, , int, s, 64, 1, 0x7FFFFFFFFFFFFFFFLL); + VDUP(vector, , uint, u, 8, 8, 0xFF); + VDUP(vector, , uint, u, 16, 4, 0xFFFF); + VDUP(vector, , uint, u, 32, 2, 0xFFFFFFFF); + VDUP(vector, , uint, u, 64, 1, 0xFFFFFFFFFFFFFFFFULL); + VDUP(vector, q, int, s, 8, 16, 0x7F); + VDUP(vector, q, int, s, 16, 8, 0x7FFF); + VDUP(vector, q, int, s, 32, 4, 0x7FFFFFFF); + VDUP(vector, q, int, s, 64, 2, 0x7FFFFFFFFFFFFFFFLL); + VDUP(vector, q, uint, u, 8, 16, 0xFF); + VDUP(vector, q, uint, u, 16, 8, 0xFFFF); + VDUP(vector, q, uint, u, 32, 4, 0xFFFFFFFF); + VDUP(vector, q, uint, u, 64, 2, 0xFFFFFFFFFFFFFFFFULL); + + /* Use -1 shift amount to check cumulative saturation with + round_const. */ + VDUP(vector_shift, , int, s, 8, 8, -1); + VDUP(vector_shift, , int, s, 16, 4, -1); + VDUP(vector_shift, , int, s, 32, 2, -1); + VDUP(vector_shift, , int, s, 64, 1, -1); + VDUP(vector_shift, q, int, s, 8, 16, -1); + VDUP(vector_shift, q, int, s, 16, 8, -1); + VDUP(vector_shift, q, int, s, 32, 4, -1); + VDUP(vector_shift, q, int, s, 64, 2, -1); + +#undef CMT +#define CMT " (checking cumulative saturation: shift by -1)" + TEST_VQRSHL(int, , int, s, 8, 8, expected_cumulative_sat_minus1, CMT); + TEST_VQRSHL(int, , int, s, 16, 4, expected_cumulative_sat_minus1, CMT); + TEST_VQRSHL(int, , int, s, 32, 2, expected_cumulative_sat_minus1, CMT); + TEST_VQRSHL(int, , int, s, 64, 1, expected_cumulative_sat_minus1, CMT); + TEST_VQRSHL(int, , uint, u, 8, 8, expected_cumulative_sat_minus1, CMT); + TEST_VQRSHL(int, , uint, u, 16, 4, expected_cumulative_sat_minus1, CMT); + TEST_VQRSHL(int, , uint, u, 32, 2, expected_cumulative_sat_minus1, CMT); + TEST_VQRSHL(int, , uint, u, 64, 1, expected_cumulative_sat_minus1, CMT); + TEST_VQRSHL(int, q, int, s, 8, 16, expected_cumulative_sat_minus1, CMT); + TEST_VQRSHL(int, q, int, s, 16, 8, expected_cumulative_sat_minus1, CMT); + TEST_VQRSHL(int, q, int, s, 32, 4, expected_cumulative_sat_minus1, CMT); + TEST_VQRSHL(int, q, int, s, 64, 2, expected_cumulative_sat_minus1, CMT); + TEST_VQRSHL(int, q, uint, u, 8, 16, expected_cumulative_sat_minus1, CMT); + TEST_VQRSHL(int, q, uint, u, 16, 8, expected_cumulative_sat_minus1, CMT); + TEST_VQRSHL(int, q, uint, u, 32, 4, expected_cumulative_sat_minus1, CMT); + TEST_VQRSHL(int, q, uint, u, 64, 2, expected_cumulative_sat_minus1, CMT); + + CHECK(TEST_MSG, int, 8, 8, PRIx8, expected_minus1, CMT); + CHECK(TEST_MSG, int, 16, 4, PRIx16, expected_minus1, CMT); + CHECK(TEST_MSG, int, 32, 2, PRIx32, expected_minus1, CMT); + CHECK(TEST_MSG, int, 64, 1, PRIx64, expected_minus1, CMT); + CHECK(TEST_MSG, uint, 8, 8, PRIx8, expected_minus1, CMT); + CHECK(TEST_MSG, uint, 16, 4, PRIx16, expected_minus1, CMT); + CHECK(TEST_MSG, uint, 32, 2, PRIx32, expected_minus1, CMT); + CHECK(TEST_MSG, uint, 64, 1, PRIx64, expected_minus1, CMT); + CHECK(TEST_MSG, int, 8, 16, PRIx8, expected_minus1, CMT); + CHECK(TEST_MSG, int, 16, 8, PRIx16, expected_minus1, CMT); + CHECK(TEST_MSG, int, 32, 4, PRIx32, expected_minus1, CMT); + CHECK(TEST_MSG, int, 64, 2, PRIx64, expected_minus1, CMT); + CHECK(TEST_MSG, uint, 8, 16, PRIx8, expected_minus1, CMT); + CHECK(TEST_MSG, uint, 16, 8, PRIx16, expected_minus1, CMT); + CHECK(TEST_MSG, uint, 32, 4, PRIx32, expected_minus1, CMT); + CHECK(TEST_MSG, uint, 64, 2, PRIx64, expected_minus1, CMT); + + + /* Use -3 shift amount to check cumulative saturation with + round_const. */ + VDUP(vector_shift, , int, s, 8, 8, -3); + VDUP(vector_shift, , int, s, 16, 4, -3); + VDUP(vector_shift, , int, s, 32, 2, -3); + VDUP(vector_shift, , int, s, 64, 1, -3); + VDUP(vector_shift, q, int, s, 8, 16, -3); + VDUP(vector_shift, q, int, s, 16, 8, -3); + VDUP(vector_shift, q, int, s, 32, 4, -3); + VDUP(vector_shift, q, int, s, 64, 2, -3); + +#undef CMT +#define CMT " (checking cumulative saturation: shift by -3)" + TEST_VQRSHL(int, , int, s, 8, 8, expected_cumulative_sat_minus3, CMT); + TEST_VQRSHL(int, , int, s, 16, 4, expected_cumulative_sat_minus3, CMT); + TEST_VQRSHL(int, , int, s, 32, 2, expected_cumulative_sat_minus3, CMT); + TEST_VQRSHL(int, , int, s, 64, 1, expected_cumulative_sat_minus3, CMT); + TEST_VQRSHL(int, , uint, u, 8, 8, expected_cumulative_sat_minus3, CMT); + TEST_VQRSHL(int, , uint, u, 16, 4, expected_cumulative_sat_minus3, CMT); + TEST_VQRSHL(int, , uint, u, 32, 2, expected_cumulative_sat_minus3, CMT); + TEST_VQRSHL(int, , uint, u, 64, 1, expected_cumulative_sat_minus3, CMT); + TEST_VQRSHL(int, q, int, s, 8, 16, expected_cumulative_sat_minus3, CMT); + TEST_VQRSHL(int, q, int, s, 16, 8, expected_cumulative_sat_minus3, CMT); + TEST_VQRSHL(int, q, int, s, 32, 4, expected_cumulative_sat_minus3, CMT); + TEST_VQRSHL(int, q, int, s, 64, 2, expected_cumulative_sat_minus3, CMT); + TEST_VQRSHL(int, q, uint, u, 8, 16, expected_cumulative_sat_minus3, CMT); + TEST_VQRSHL(int, q, uint, u, 16, 8, expected_cumulative_sat_minus3, CMT); + TEST_VQRSHL(int, q, uint, u, 32, 4, expected_cumulative_sat_minus3, CMT); + TEST_VQRSHL(int, q, uint, u, 64, 2, expected_cumulative_sat_minus3, CMT); + + CHECK(TEST_MSG, int, 8, 8, PRIx8, expected_minus3, CMT); + CHECK(TEST_MSG, int, 16, 4, PRIx16, expected_minus3, CMT); + CHECK(TEST_MSG, int, 32, 2, PRIx32, expected_minus3, CMT); + CHECK(TEST_MSG, int, 64, 1, PRIx64, expected_minus3, CMT); + CHECK(TEST_MSG, uint, 8, 8, PRIx8, expected_minus3, CMT); + CHECK(TEST_MSG, uint, 16, 4, PRIx16, expected_minus3, CMT); + CHECK(TEST_MSG, uint, 32, 2, PRIx32, expected_minus3, CMT); + CHECK(TEST_MSG, uint, 64, 1, PRIx64, expected_minus3, CMT); + CHECK(TEST_MSG, int, 8, 16, PRIx8, expected_minus3, CMT); + CHECK(TEST_MSG, int, 16, 8, PRIx16, expected_minus3, CMT); + CHECK(TEST_MSG, int, 32, 4, PRIx32, expected_minus3, CMT); + CHECK(TEST_MSG, int, 64, 2, PRIx64, expected_minus3, CMT); + CHECK(TEST_MSG, uint, 8, 16, PRIx8, expected_minus3, CMT); + CHECK(TEST_MSG, uint, 16, 8, PRIx16, expected_minus3, CMT); + CHECK(TEST_MSG, uint, 32, 4, PRIx32, expected_minus3, CMT); + CHECK(TEST_MSG, uint, 64, 2, PRIx64, expected_minus3, CMT); + + + /* Use large shift amount. */ + VDUP(vector_shift, , int, s, 8, 8, 10); + VDUP(vector_shift, , int, s, 16, 4, 20); + VDUP(vector_shift, , int, s, 32, 2, 40); + VDUP(vector_shift, , int, s, 64, 1, 70); + VDUP(vector_shift, q, int, s, 8, 16, 10); + VDUP(vector_shift, q, int, s, 16, 8, 20); + VDUP(vector_shift, q, int, s, 32, 4, 40); + VDUP(vector_shift, q, int, s, 64, 2, 70); + +#undef CMT +#define CMT " (checking cumulative saturation: large shift amount)" + TEST_VQRSHL(int, , int, s, 8, 8, expected_cumulative_sat_large_sh, CMT); + TEST_VQRSHL(int, , int, s, 16, 4, expected_cumulative_sat_large_sh, CMT); + TEST_VQRSHL(int, , int, s, 32, 2, expected_cumulative_sat_large_sh, CMT); + TEST_VQRSHL(int, , int, s, 64, 1, expected_cumulative_sat_large_sh, CMT); + TEST_VQRSHL(int, , uint, u, 8, 8, expected_cumulative_sat_large_sh, CMT); + TEST_VQRSHL(int, , uint, u, 16, 4, expected_cumulative_sat_large_sh, CMT); + TEST_VQRSHL(int, , uint, u, 32, 2, expected_cumulative_sat_large_sh, CMT); + TEST_VQRSHL(int, , uint, u, 64, 1, expected_cumulative_sat_large_sh, CMT); + TEST_VQRSHL(int, q, int, s, 8, 16, expected_cumulative_sat_large_sh, CMT); + TEST_VQRSHL(int, q, int, s, 16, 8, expected_cumulative_sat_large_sh, CMT); + TEST_VQRSHL(int, q, int, s, 32, 4, expected_cumulative_sat_large_sh, CMT); + TEST_VQRSHL(int, q, int, s, 64, 2, expected_cumulative_sat_large_sh, CMT); + TEST_VQRSHL(int, q, uint, u, 8, 16, expected_cumulative_sat_large_sh, CMT); + TEST_VQRSHL(int, q, uint, u, 16, 8, expected_cumulative_sat_large_sh, CMT); + TEST_VQRSHL(int, q, uint, u, 32, 4, expected_cumulative_sat_large_sh, CMT); + TEST_VQRSHL(int, q, uint, u, 64, 2, expected_cumulative_sat_large_sh, CMT); + + CHECK(TEST_MSG, int, 8, 8, PRIx8, expected_large_sh, CMT); + CHECK(TEST_MSG, int, 16, 4, PRIx16, expected_large_sh, CMT); + CHECK(TEST_MSG, int, 32, 2, PRIx32, expected_large_sh, CMT); + CHECK(TEST_MSG, int, 64, 1, PRIx64, expected_large_sh, CMT); + CHECK(TEST_MSG, uint, 8, 8, PRIx8, expected_large_sh, CMT); + CHECK(TEST_MSG, uint, 16, 4, PRIx16, expected_large_sh, CMT); + CHECK(TEST_MSG, uint, 32, 2, PRIx32, expected_large_sh, CMT); + CHECK(TEST_MSG, uint, 64, 1, PRIx64, expected_large_sh, CMT); + CHECK(TEST_MSG, int, 8, 16, PRIx8, expected_large_sh, CMT); + CHECK(TEST_MSG, int, 16, 8, PRIx16, expected_large_sh, CMT); + CHECK(TEST_MSG, int, 32, 4, PRIx32, expected_large_sh, CMT); + CHECK(TEST_MSG, int, 64, 2, PRIx64, expected_large_sh, CMT); + CHECK(TEST_MSG, uint, 8, 16, PRIx8, expected_large_sh, CMT); + CHECK(TEST_MSG, uint, 16, 8, PRIx16, expected_large_sh, CMT); + CHECK(TEST_MSG, uint, 32, 4, PRIx32, expected_large_sh, CMT); + CHECK(TEST_MSG, uint, 64, 2, PRIx64, expected_large_sh, CMT); + + + /* Fill input vector with negative values, to check saturation on + limits. */ + VDUP(vector, , int, s, 8, 8, 0x80); + VDUP(vector, , int, s, 16, 4, 0x8000); + VDUP(vector, , int, s, 32, 2, 0x80000000); + VDUP(vector, , int, s, 64, 1, 0x8000000000000000LL); + VDUP(vector, q, int, s, 8, 16, 0x80); + VDUP(vector, q, int, s, 16, 8, 0x8000); + VDUP(vector, q, int, s, 32, 4, 0x80000000); + VDUP(vector, q, int, s, 64, 2, 0x8000000000000000LL); + + /* Use large shift amount. */ + VDUP(vector_shift, , int, s, 8, 8, 10); + VDUP(vector_shift, , int, s, 16, 4, 20); + VDUP(vector_shift, , int, s, 32, 2, 40); + VDUP(vector_shift, , int, s, 64, 1, 70); + VDUP(vector_shift, q, int, s, 8, 16, 10); + VDUP(vector_shift, q, int, s, 16, 8, 20); + VDUP(vector_shift, q, int, s, 32, 4, 40); + VDUP(vector_shift, q, int, s, 64, 2, 70); + +#undef CMT +#define CMT " (checking cumulative saturation: large shift amount with negative input)" + TEST_VQRSHL(int, , int, s, 8, 8, expected_cumulative_sat_neg_large_sh, CMT); + TEST_VQRSHL(int, , int, s, 16, 4, expected_cumulative_sat_neg_large_sh, CMT); + TEST_VQRSHL(int, , int, s, 32, 2, expected_cumulative_sat_neg_large_sh, CMT); + TEST_VQRSHL(int, , int, s, 64, 1, expected_cumulative_sat_neg_large_sh, CMT); + TEST_VQRSHL(int, , uint, u, 8, 8, expected_cumulative_sat_neg_large_sh, CMT); + TEST_VQRSHL(int, , uint, u, 16, 4, expected_cumulative_sat_neg_large_sh, CMT); + TEST_VQRSHL(int, , uint, u, 32, 2, expected_cumulative_sat_neg_large_sh, CMT); + TEST_VQRSHL(int, , uint, u, 64, 1, expected_cumulative_sat_neg_large_sh, CMT); + TEST_VQRSHL(int, q, int, s, 8, 16, expected_cumulative_sat_neg_large_sh, CMT); + TEST_VQRSHL(int, q, int, s, 16, 8, expected_cumulative_sat_neg_large_sh, CMT); + TEST_VQRSHL(int, q, int, s, 32, 4, expected_cumulative_sat_neg_large_sh, CMT); + TEST_VQRSHL(int, q, int, s, 64, 2, expected_cumulative_sat_neg_large_sh, CMT); + TEST_VQRSHL(int, q, uint, u, 8, 16, expected_cumulative_sat_neg_large_sh, CMT); + TEST_VQRSHL(int, q, uint, u, 16, 8, expected_cumulative_sat_neg_large_sh, CMT); + TEST_VQRSHL(int, q, uint, u, 32, 4, expected_cumulative_sat_neg_large_sh, CMT); + TEST_VQRSHL(int, q, uint, u, 64, 2, expected_cumulative_sat_neg_large_sh, CMT); + + CHECK(TEST_MSG, int, 8, 8, PRIx8, expected_neg_large_sh, CMT); + CHECK(TEST_MSG, int, 16, 4, PRIx16, expected_neg_large_sh, CMT); + CHECK(TEST_MSG, int, 32, 2, PRIx32, expected_neg_large_sh, CMT); + CHECK(TEST_MSG, int, 64, 1, PRIx64, expected_neg_large_sh, CMT); + CHECK(TEST_MSG, uint, 8, 8, PRIx8, expected_neg_large_sh, CMT); + CHECK(TEST_MSG, uint, 16, 4, PRIx16, expected_neg_large_sh, CMT); + CHECK(TEST_MSG, uint, 32, 2, PRIx32, expected_neg_large_sh, CMT); + CHECK(TEST_MSG, uint, 64, 1, PRIx64, expected_neg_large_sh, CMT); + CHECK(TEST_MSG, int, 8, 16, PRIx8, expected_neg_large_sh, CMT); + CHECK(TEST_MSG, int, 16, 8, PRIx16, expected_neg_large_sh, CMT); + CHECK(TEST_MSG, int, 32, 4, PRIx32, expected_neg_large_sh, CMT); + CHECK(TEST_MSG, int, 64, 2, PRIx64, expected_neg_large_sh, CMT); + CHECK(TEST_MSG, uint, 8, 16, PRIx8, expected_neg_large_sh, CMT); + CHECK(TEST_MSG, uint, 16, 8, PRIx16, expected_neg_large_sh, CMT); + CHECK(TEST_MSG, uint, 32, 4, PRIx32, expected_neg_large_sh, CMT); + CHECK(TEST_MSG, uint, 64, 2, PRIx64, expected_neg_large_sh, CMT); + + + /* Fill input vector with negative and positive values, to check + * saturation on limits */ + VDUP(vector, , int, s, 8, 8, 0x7F); + VDUP(vector, , int, s, 16, 4, 0x7FFF); + VDUP(vector, , int, s, 32, 2, 0x7FFFFFFF); + VDUP(vector, , int, s, 64, 1, 0x7FFFFFFFFFFFFFFFLL); + VDUP(vector, q, int, s, 8, 16, 0x80); + VDUP(vector, q, int, s, 16, 8, 0x8000); + VDUP(vector, q, int, s, 32, 4, 0x80000000); + VDUP(vector, q, int, s, 64, 2, 0x8000000000000000LL); + + /* Use large negative shift amount */ + VDUP(vector_shift, , int, s, 8, 8, -10); + VDUP(vector_shift, , int, s, 16, 4, -20); + VDUP(vector_shift, , int, s, 32, 2, -40); + VDUP(vector_shift, , int, s, 64, 1, -70); + VDUP(vector_shift, q, int, s, 8, 16, -10); + VDUP(vector_shift, q, int, s, 16, 8, -20); + VDUP(vector_shift, q, int, s, 32, 4, -40); + VDUP(vector_shift, q, int, s, 64, 2, -70); + +#undef CMT +#define CMT " (checking cumulative saturation: large negative shift amount)" + TEST_VQRSHL(int, , int, s, 8, 8, expected_cumulative_sat_large_neg_sh, CMT); + TEST_VQRSHL(int, , int, s, 16, 4, expected_cumulative_sat_large_neg_sh, CMT); + TEST_VQRSHL(int, , int, s, 32, 2, expected_cumulative_sat_large_neg_sh, CMT); + TEST_VQRSHL(int, , int, s, 64, 1, expected_cumulative_sat_large_neg_sh, CMT); + TEST_VQRSHL(int, , uint, u, 8, 8, expected_cumulative_sat_large_neg_sh, CMT); + TEST_VQRSHL(int, , uint, u, 16, 4, expected_cumulative_sat_large_neg_sh, CMT); + TEST_VQRSHL(int, , uint, u, 32, 2, expected_cumulative_sat_large_neg_sh, CMT); + TEST_VQRSHL(int, , uint, u, 64, 1, expected_cumulative_sat_large_neg_sh, CMT); + TEST_VQRSHL(int, q, int, s, 8, 16, expected_cumulative_sat_large_neg_sh, CMT); + TEST_VQRSHL(int, q, int, s, 16, 8, expected_cumulative_sat_large_neg_sh, CMT); + TEST_VQRSHL(int, q, int, s, 32, 4, expected_cumulative_sat_large_neg_sh, CMT); + TEST_VQRSHL(int, q, int, s, 64, 2, expected_cumulative_sat_large_neg_sh, CMT); + TEST_VQRSHL(int, q, uint, u, 8, 16, expected_cumulative_sat_large_neg_sh, CMT); + TEST_VQRSHL(int, q, uint, u, 16, 8, expected_cumulative_sat_large_neg_sh, CMT); + TEST_VQRSHL(int, q, uint, u, 32, 4, expected_cumulative_sat_large_neg_sh, CMT); + TEST_VQRSHL(int, q, uint, u, 64, 2, expected_cumulative_sat_large_neg_sh, CMT); + + CHECK(TEST_MSG, int, 8, 8, PRIx8, expected_large_neg_sh, CMT); + CHECK(TEST_MSG, int, 16, 4, PRIx16, expected_large_neg_sh, CMT); + CHECK(TEST_MSG, int, 32, 2, PRIx32, expected_large_neg_sh, CMT); + CHECK(TEST_MSG, int, 64, 1, PRIx64, expected_large_neg_sh, CMT); + CHECK(TEST_MSG, uint, 8, 8, PRIx8, expected_large_neg_sh, CMT); + CHECK(TEST_MSG, uint, 16, 4, PRIx16, expected_large_neg_sh, CMT); + CHECK(TEST_MSG, uint, 32, 2, PRIx32, expected_large_neg_sh, CMT); + CHECK(TEST_MSG, uint, 64, 1, PRIx64, expected_large_neg_sh, CMT); + CHECK(TEST_MSG, int, 8, 16, PRIx8, expected_large_neg_sh, CMT); + CHECK(TEST_MSG, int, 16, 8, PRIx16, expected_large_neg_sh, CMT); + CHECK(TEST_MSG, int, 32, 4, PRIx32, expected_large_neg_sh, CMT); + CHECK(TEST_MSG, int, 64, 2, PRIx64, expected_large_neg_sh, CMT); + CHECK(TEST_MSG, uint, 8, 16, PRIx8, expected_large_neg_sh, CMT); + CHECK(TEST_MSG, uint, 16, 8, PRIx16, expected_large_neg_sh, CMT); + CHECK(TEST_MSG, uint, 32, 4, PRIx32, expected_large_neg_sh, CMT); + CHECK(TEST_MSG, uint, 64, 2, PRIx64, expected_large_neg_sh, CMT); + + + /* Fill input vector with 0, to check saturation in case of large + * shift amount */ + VDUP(vector, , int, s, 8, 8, 0); + VDUP(vector, , int, s, 16, 4, 0); + VDUP(vector, , int, s, 32, 2, 0); + VDUP(vector, , int, s, 64, 1, 0); + VDUP(vector, q, int, s, 8, 16, 0); + VDUP(vector, q, int, s, 16, 8, 0); + VDUP(vector, q, int, s, 32, 4, 0); + VDUP(vector, q, int, s, 64, 2, 0); + + /* Use large shift amount */ + VDUP(vector_shift, , int, s, 8, 8, -10); + VDUP(vector_shift, , int, s, 16, 4, -20); + VDUP(vector_shift, , int, s, 32, 2, -40); + VDUP(vector_shift, , int, s, 64, 1, -70); + VDUP(vector_shift, q, int, s, 8, 16, -10); + VDUP(vector_shift, q, int, s, 16, 8, -20); + VDUP(vector_shift, q, int, s, 32, 4, -40); + VDUP(vector_shift, q, int, s, 64, 2, -70); + +#undef CMT +#define CMT " (checking cumulative saturation: large negative shift amount with 0 input)" + TEST_VQRSHL(int, , int, s, 8, 8, expected_cumulative_sat_large_neg_sh, CMT); + TEST_VQRSHL(int, , int, s, 16, 4, expected_cumulative_sat_large_neg_sh, CMT); + TEST_VQRSHL(int, , int, s, 32, 2, expected_cumulative_sat_large_neg_sh, CMT); + TEST_VQRSHL(int, , int, s, 64, 1, expected_cumulative_sat_large_neg_sh, CMT); + TEST_VQRSHL(int, , uint, u, 8, 8, expected_cumulative_sat_large_neg_sh, CMT); + TEST_VQRSHL(int, , uint, u, 16, 4, expected_cumulative_sat_large_neg_sh, CMT); + TEST_VQRSHL(int, , uint, u, 32, 2, expected_cumulative_sat_large_neg_sh, CMT); + TEST_VQRSHL(int, , uint, u, 64, 1, expected_cumulative_sat_large_neg_sh, CMT); + TEST_VQRSHL(int, q, int, s, 8, 16, expected_cumulative_sat_large_neg_sh, CMT); + TEST_VQRSHL(int, q, int, s, 16, 8, expected_cumulative_sat_large_neg_sh, CMT); + TEST_VQRSHL(int, q, int, s, 32, 4, expected_cumulative_sat_large_neg_sh, CMT); + TEST_VQRSHL(int, q, int, s, 64, 2, expected_cumulative_sat_large_neg_sh, CMT); + TEST_VQRSHL(int, q, uint, u, 8, 16, expected_cumulative_sat_large_neg_sh, CMT); + TEST_VQRSHL(int, q, uint, u, 16, 8, expected_cumulative_sat_large_neg_sh, CMT); + TEST_VQRSHL(int, q, uint, u, 32, 4, expected_cumulative_sat_large_neg_sh, CMT); + TEST_VQRSHL(int, q, uint, u, 64, 2, expected_cumulative_sat_large_neg_sh, CMT); + + CHECK(TEST_MSG, int, 8, 8, PRIx8, expected_large_neg_sh, CMT); + CHECK(TEST_MSG, int, 16, 4, PRIx16, expected_large_neg_sh, CMT); + CHECK(TEST_MSG, int, 32, 2, PRIx32, expected_large_neg_sh, CMT); + CHECK(TEST_MSG, int, 64, 1, PRIx64, expected_large_neg_sh, CMT); + CHECK(TEST_MSG, uint, 8, 8, PRIx8, expected_large_neg_sh, CMT); + CHECK(TEST_MSG, uint, 16, 4, PRIx16, expected_large_neg_sh, CMT); + CHECK(TEST_MSG, uint, 32, 2, PRIx32, expected_large_neg_sh, CMT); + CHECK(TEST_MSG, uint, 64, 1, PRIx64, expected_large_neg_sh, CMT); + CHECK(TEST_MSG, int, 8, 16, PRIx8, expected_large_neg_sh, CMT); + CHECK(TEST_MSG, int, 16, 8, PRIx16, expected_large_neg_sh, CMT); + CHECK(TEST_MSG, int, 32, 4, PRIx32, expected_large_neg_sh, CMT); + CHECK(TEST_MSG, int, 64, 2, PRIx64, expected_large_neg_sh, CMT); + CHECK(TEST_MSG, uint, 8, 16, PRIx8, expected_large_neg_sh, CMT); + CHECK(TEST_MSG, uint, 16, 8, PRIx16, expected_large_neg_sh, CMT); + CHECK(TEST_MSG, uint, 32, 4, PRIx32, expected_large_neg_sh, CMT); + CHECK(TEST_MSG, uint, 64, 2, PRIx64, expected_large_neg_sh, CMT); +} + +int main (void) +{ + exec_vqrshl (); + return 0; +}