From nobody Sun Nov 9 17:53:26 2025 Delivered-To: importer@patchew.org Received-SPF: pass (zoho.com: domain of gnu.org designates 209.51.188.17 as permitted sender) client-ip=209.51.188.17; envelope-from=qemu-devel-bounces+importer=patchew.org@nongnu.org; helo=lists.gnu.org; Authentication-Results: mx.zohomail.com; spf=pass (zoho.com: domain of gnu.org designates 209.51.188.17 as permitted sender) smtp.mailfrom=qemu-devel-bounces+importer=patchew.org@nongnu.org Return-Path: Received: from lists.gnu.org (lists.gnu.org [209.51.188.17]) by mx.zohomail.com with SMTPS id 1551446055261350.62661647226685; Fri, 1 Mar 2019 05:14:15 -0800 (PST) Received: from localhost ([127.0.0.1]:37747 helo=lists.gnu.org) by lists.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1gzhzM-0001X9-Ac for importer@patchew.org; Fri, 01 Mar 2019 08:14:08 -0500 Received: from eggs.gnu.org ([209.51.188.92]:54318) by lists.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1gzhvD-0006wA-HU for qemu-devel@nongnu.org; Fri, 01 Mar 2019 08:09:52 -0500 Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71) (envelope-from ) id 1gzhvC-0004GP-C9 for qemu-devel@nongnu.org; Fri, 01 Mar 2019 08:09:51 -0500 Received: from mx2.rt-rk.com ([89.216.37.149]:58448 helo=mail.rt-rk.com) by eggs.gnu.org with esmtps (TLS1.0:DHE_RSA_AES_256_CBC_SHA1:32) (Exim 4.71) (envelope-from ) id 1gzhvC-0003oG-0i for qemu-devel@nongnu.org; Fri, 01 Mar 2019 08:09:50 -0500 Received: from localhost (localhost [127.0.0.1]) by mail.rt-rk.com (Postfix) with ESMTP id 0681E1A222C; Fri, 1 Mar 2019 14:08:47 +0100 (CET) Received: from rtrkw310-lin.domain.local (rtrkw310-lin.domain.local [10.10.13.57]) by mail.rt-rk.com (Postfix) with ESMTPSA id C486D1A21FB; Fri, 1 Mar 2019 14:08:46 +0100 (CET) X-Virus-Scanned: amavisd-new at rt-rk.com From: Mateja Marjanovic To: qemu-devel@nongnu.org Date: Fri, 1 Mar 2019 14:08:37 +0100 Message-Id: <1551445720-19622-2-git-send-email-mateja.marjanovic@rt-rk.com> X-Mailer: git-send-email 2.7.4 In-Reply-To: <1551445720-19622-1-git-send-email-mateja.marjanovic@rt-rk.com> References: <1551445720-19622-1-git-send-email-mateja.marjanovic@rt-rk.com> MIME-Version: 1.0 Content-Transfer-Encoding: quoted-printable X-detected-operating-system: by eggs.gnu.org: GNU/Linux 3.x X-Received-From: 89.216.37.149 Subject: [Qemu-devel] [PATCH v2 1/4] target/mips: Optimize support for MSA instructions ILVEV. X-BeenThere: qemu-devel@nongnu.org X-Mailman-Version: 2.1.21 Precedence: list List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Cc: arikalo@wavecomp.com, amarkovic@wavecomp.com, aurelien@aurel32.net Errors-To: qemu-devel-bounces+importer=patchew.org@nongnu.org Sender: "Qemu-devel" Content-Type: text/plain; charset="utf-8" From: Mateja Marjanovic Optimize support for MSA instructions ILVEV.B, ILVEV.H, ILVEV.W, and ILVEV.D. Optimization is done by eliminating loops, and explicitly assigning desired values to individual data elements. Performance measurement is done by executing the instructions large number of times on a computer with Intel Core i7-3770 CPU @ 3.40GHz=C3=978. Measured time before optimization: ILVEV.B: 119.02 ms ILVEV.H: 94.16 ms ILVEV.W: 120.97 ms ILVEV.D: 42.99 ms Measured time after optimization: ILVEV.B: 61.81 ms ILVEV.H: 42.78 ms ILVEV.W: 39.47 ms ILVEV.D: 39.11 ms Signed-off-by: Mateja Marjanovic --- target/mips/msa_helper.c | 60 +++++++++++++++++++++++++++++++++++++++++---= ---- 1 file changed, 52 insertions(+), 8 deletions(-) diff --git a/target/mips/msa_helper.c b/target/mips/msa_helper.c index c74e3cd..49332e6 100644 --- a/target/mips/msa_helper.c +++ b/target/mips/msa_helper.c @@ -1198,14 +1198,6 @@ MSA_FN_DF(ilvl_df) MSA_FN_DF(ilvr_df) #undef MSA_DO =20 -#define MSA_DO(DF) \ - do { \ - pwx->DF[2*i] =3D pwt->DF[2*i]; \ - pwx->DF[2*i+1] =3D pws->DF[2*i]; \ - } while (0) -MSA_FN_DF(ilvev_df) -#undef MSA_DO - #define MSA_DO(DF) \ do { \ pwx->DF[2*i] =3D pwt->DF[2*i+1]; \ @@ -1230,6 +1222,58 @@ MSA_FN_DF(vshf_df) #undef MSA_LOOP_COND #undef MSA_FN_DF =20 + +void helper_msa_ilvev_df(CPUMIPSState *env, uint32_t df, uint32_t wd, + uint32_t ws, uint32_t wt) +{ + wr_t *pwd =3D &(env->active_fpu.fpr[wd].wr); + wr_t *pws =3D &(env->active_fpu.fpr[ws].wr); + wr_t *pwt =3D &(env->active_fpu.fpr[wt].wr); + + switch (df) { + case DF_BYTE: + pwd->b[15] =3D pws->b[14]; + pwd->b[14] =3D pwt->b[14]; + pwd->b[13] =3D pws->b[12]; + pwd->b[12] =3D pwt->b[12]; + pwd->b[11] =3D pws->b[10]; + pwd->b[10] =3D pwt->b[10]; + pwd->b[9] =3D pws->b[8]; + pwd->b[8] =3D pwt->b[8]; + pwd->b[7] =3D pws->b[6]; + pwd->b[6] =3D pwt->b[6]; + pwd->b[5] =3D pws->b[4]; + pwd->b[4] =3D pwt->b[4]; + pwd->b[3] =3D pws->b[2]; + pwd->b[2] =3D pwt->b[2]; + pwd->b[1] =3D pws->b[0]; + pwd->b[0] =3D pwt->b[0]; + break; + case DF_HALF: + pwd->h[7] =3D pws->h[6]; + pwd->h[6] =3D pwt->h[6]; + pwd->h[5] =3D pws->h[4]; + pwd->h[4] =3D pwt->h[4]; + pwd->h[3] =3D pws->h[2]; + pwd->h[2] =3D pwt->h[2]; + pwd->h[1] =3D pws->h[0]; + pwd->h[0] =3D pwt->h[0]; + break; + case DF_WORD: + pwd->w[3] =3D pws->w[2]; + pwd->w[2] =3D pwt->w[2]; + pwd->w[1] =3D pws->w[0]; + pwd->w[0] =3D pwt->w[0]; + break; + case DF_DOUBLE: + pwd->d[1] =3D pws->d[0]; + pwd->d[0] =3D pwt->d[0]; + break; + default: + assert(0); + } +} + void helper_msa_sldi_df(CPUMIPSState *env, uint32_t df, uint32_t wd, uint32_t ws, uint32_t n) { --=20 2.7.4