From patchwork Thu Jun 24 10:55:09 2021 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: LIU Zhiwei X-Patchwork-Id: 12341967 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org X-Spam-Level: X-Spam-Status: No, score=-16.7 required=3.0 tests=BAYES_00, HEADER_FROM_DIFFERENT_DOMAINS,INCLUDES_CR_TRAILER,INCLUDES_PATCH, MAILING_LIST_MULTI,SPF_HELO_NONE,SPF_PASS,UNPARSEABLE_RELAY,URIBL_BLOCKED, USER_AGENT_GIT autolearn=ham autolearn_force=no version=3.4.0 Received: from mail.kernel.org (mail.kernel.org [198.145.29.99]) by smtp.lore.kernel.org (Postfix) with ESMTP id ED0C5C48BDF for ; Thu, 24 Jun 2021 11:40:30 +0000 (UTC) Received: from lists.gnu.org (lists.gnu.org [209.51.188.17]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by mail.kernel.org (Postfix) with ESMTPS id 7B2E361374 for ; Thu, 24 Jun 2021 11:40:30 +0000 (UTC) DMARC-Filter: OpenDMARC Filter v1.3.2 mail.kernel.org 7B2E361374 Authentication-Results: mail.kernel.org; dmarc=none (p=none dis=none) header.from=c-sky.com Authentication-Results: mail.kernel.org; spf=pass smtp.mailfrom=qemu-devel-bounces+qemu-devel=archiver.kernel.org@nongnu.org Received: from localhost ([::1]:49506 helo=lists1p.gnu.org) by lists.gnu.org with esmtp (Exim 4.90_1) (envelope-from ) id 1lwNif-00014k-L4 for qemu-devel@archiver.kernel.org; Thu, 24 Jun 2021 07:40:29 -0400 Received: from eggs.gnu.org ([2001:470:142:3::10]:44542) by lists.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256) (Exim 4.90_1) (envelope-from ) id 1lwNEa-0007Ad-31; Thu, 24 Jun 2021 07:09:24 -0400 Received: from mail142-29.mail.alibaba.com ([198.11.142.29]:16622) by eggs.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256) (Exim 4.90_1) (envelope-from ) id 1lwNEW-0004o2-UL; Thu, 24 Jun 2021 07:09:23 -0400 X-Alimail-AntiSpam: AC=CONTINUE; BC=0.07436283|-1; CH=blue; DM=|OVERLOAD|false|; DS=CONTINUE|ham_system_inform|0.524607-0.00375863-0.471634; FP=0|0|0|0|0|-1|-1|-1; HT=ay29a033018047207; MF=zhiwei_liu@c-sky.com; NM=1; PH=DS; RN=6; RT=6; SR=0; TI=SMTPD_---.KXK1QJp_1624532949; Received: from roman-VirtualBox.hz.ali.com(mailfrom:zhiwei_liu@c-sky.com fp:SMTPD_---.KXK1QJp_1624532949) by smtp.aliyun-inc.com(10.147.42.197); Thu, 24 Jun 2021 19:09:09 +0800 From: LIU Zhiwei To: qemu-devel@nongnu.org, qemu-riscv@nongnu.org Subject: [PATCH v3 25/37] target/riscv: Non-SIMD Q31 saturation ALU Instructions Date: Thu, 24 Jun 2021 18:55:09 +0800 Message-Id: <20210624105521.3964-26-zhiwei_liu@c-sky.com> X-Mailer: git-send-email 2.17.1 In-Reply-To: <20210624105521.3964-1-zhiwei_liu@c-sky.com> References: <20210624105521.3964-1-zhiwei_liu@c-sky.com> Received-SPF: none client-ip=198.11.142.29; envelope-from=zhiwei_liu@c-sky.com; helo=mail142-29.mail.alibaba.com X-Spam_score_int: -25 X-Spam_score: -2.6 X-Spam_bar: -- X-Spam_report: (-2.6 / 5.0 requ) BAYES_00=-1.9, RCVD_IN_DNSWL_LOW=-0.7, SPF_HELO_NONE=0.001, SPF_NONE=0.001, UNPARSEABLE_RELAY=0.001 autolearn=ham autolearn_force=no X-Spam_action: no action X-BeenThere: qemu-devel@nongnu.org X-Mailman-Version: 2.1.23 Precedence: list List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Cc: palmer@dabbelt.com, bin.meng@windriver.com, Alistair.Francis@wdc.com, LIU Zhiwei Errors-To: qemu-devel-bounces+qemu-devel=archiver.kernel.org@nongnu.org Sender: "Qemu-devel" Q31 saturation is to limit the result to the range [INT32_MIN, INT32_MAX]. Signed-off-by: LIU Zhiwei --- target/riscv/helper.h | 15 ++ target/riscv/insn32.decode | 16 ++ target/riscv/insn_trans/trans_rvp.c.inc | 17 ++ target/riscv/packed_helper.c | 214 ++++++++++++++++++++++++ 4 files changed, 262 insertions(+) diff --git a/target/riscv/helper.h b/target/riscv/helper.h index 6ce22a186e..b3485f95a2 100644 --- a/target/riscv/helper.h +++ b/target/riscv/helper.h @@ -1369,3 +1369,18 @@ DEF_HELPER_3(khmbt, tl, env, tl, tl) DEF_HELPER_3(khmtt, tl, env, tl, tl) DEF_HELPER_3(ukaddh, tl, env, tl, tl) DEF_HELPER_3(uksubh, tl, env, tl, tl) + +DEF_HELPER_3(kaddw, tl, env, tl, tl) +DEF_HELPER_3(ukaddw, tl, env, tl, tl) +DEF_HELPER_3(ksubw, tl, env, tl, tl) +DEF_HELPER_3(uksubw, tl, env, tl, tl) +DEF_HELPER_3(kdmbb, tl, env, tl, tl) +DEF_HELPER_3(kdmbt, tl, env, tl, tl) +DEF_HELPER_3(kdmtt, tl, env, tl, tl) +DEF_HELPER_3(kslraw, tl, env, tl, tl) +DEF_HELPER_3(kslraw_u, tl, env, tl, tl) +DEF_HELPER_3(ksllw, tl, env, tl, tl) +DEF_HELPER_4(kdmabb, tl, env, tl, tl, tl) +DEF_HELPER_4(kdmabt, tl, env, tl, tl, tl) +DEF_HELPER_4(kdmatt, tl, env, tl, tl, tl) +DEF_HELPER_2(kabsw, tl, env, tl) diff --git a/target/riscv/insn32.decode b/target/riscv/insn32.decode index f465851f03..a25294baab 100644 --- a/target/riscv/insn32.decode +++ b/target/riscv/insn32.decode @@ -972,3 +972,19 @@ khmbt 0001110 ..... ..... 001 ..... 1110111 @r khmtt 0010110 ..... ..... 001 ..... 1110111 @r ukaddh 0001010 ..... ..... 001 ..... 1110111 @r uksubh 0001011 ..... ..... 001 ..... 1110111 @r + +kaddw 0000000 ..... ..... 001 ..... 1110111 @r +ukaddw 0001000 ..... ..... 001 ..... 1110111 @r +ksubw 0000001 ..... ..... 001 ..... 1110111 @r +uksubw 0001001 ..... ..... 001 ..... 1110111 @r +kdmbb 0000101 ..... ..... 001 ..... 1110111 @r +kdmbt 0001101 ..... ..... 001 ..... 1110111 @r +kdmtt 0010101 ..... ..... 001 ..... 1110111 @r +kslraw 0110111 ..... ..... 001 ..... 1110111 @r +kslraw_u 0111111 ..... ..... 001 ..... 1110111 @r +ksllw 0010011 ..... ..... 001 ..... 1110111 @r +kslliw 0011011 ..... ..... 001 ..... 1110111 @sh5 +kdmabb 1101001 ..... ..... 001 ..... 1110111 @r +kdmabt 1110001 ..... ..... 001 ..... 1110111 @r +kdmatt 1111001 ..... ..... 001 ..... 1110111 @r +kabsw 1010110 10100 ..... 000 ..... 1110111 @r2 diff --git a/target/riscv/insn_trans/trans_rvp.c.inc b/target/riscv/insn_trans/trans_rvp.c.inc index 48eb190bc6..d2c7ab1440 100644 --- a/target/riscv/insn_trans/trans_rvp.c.inc +++ b/target/riscv/insn_trans/trans_rvp.c.inc @@ -681,3 +681,20 @@ GEN_RVP_R_OOL(khmbt); GEN_RVP_R_OOL(khmtt); GEN_RVP_R_OOL(ukaddh); GEN_RVP_R_OOL(uksubh); + +/* Non-SIMD Q31 saturation ALU Instructions */ +GEN_RVP_R_OOL(kaddw); +GEN_RVP_R_OOL(ukaddw); +GEN_RVP_R_OOL(ksubw); +GEN_RVP_R_OOL(uksubw); +GEN_RVP_R_OOL(kdmbb); +GEN_RVP_R_OOL(kdmbt); +GEN_RVP_R_OOL(kdmtt); +GEN_RVP_R_OOL(kslraw); +GEN_RVP_R_OOL(kslraw_u); +GEN_RVP_R_OOL(ksllw); +GEN_RVP_SHIFTI(kslliw, NULL, gen_helper_ksllw); +GEN_RVP_R_ACC_OOL(kdmabb); +GEN_RVP_R_ACC_OOL(kdmabt); +GEN_RVP_R_ACC_OOL(kdmatt); +GEN_RVP_R2_OOL(kabsw); diff --git a/target/riscv/packed_helper.c b/target/riscv/packed_helper.c index 171f88face..89d203730d 100644 --- a/target/riscv/packed_helper.c +++ b/target/riscv/packed_helper.c @@ -2604,3 +2604,217 @@ static inline void do_uksubh(CPURISCVState *env, void *vd, void *va, } RVPR(uksubh, 2, 4); + +/* Q31 saturation Instructions */ +static inline void do_kaddw(CPURISCVState *env, void *vd, void *va, + void *vb, uint8_t i) +{ + target_long *d = vd; + int32_t *a = va, *b = vb; + + *d = sadd32(env, 0, a[H4(i)], b[H4(i)]); +} + +RVPR(kaddw, 2, 4); + +static inline void do_ukaddw(CPURISCVState *env, void *vd, void *va, + void *vb, uint8_t i) +{ + target_long *d = vd; + uint32_t *a = va, *b = vb; + + *d = (int32_t)saddu32(env, 0, a[H4(i)], b[H4(i)]); +} + +RVPR(ukaddw, 2, 4); + +static inline void do_ksubw(CPURISCVState *env, void *vd, void *va, + void *vb, uint8_t i) +{ + target_long *d = vd; + int32_t *a = va, *b = vb; + + *d = ssub32(env, 0, a[H4(i)], b[H4(i)]); +} + +RVPR(ksubw, 2, 4); + +static inline void do_uksubw(CPURISCVState *env, void *vd, void *va, + void *vb, uint8_t i) +{ + target_long *d = vd; + uint32_t *a = va, *b = vb; + + *d = (int32_t)ssubu32(env, 0, a[H4(i)], b[H4(i)]); +} + +RVPR(uksubw, 2, 4); + +static inline void do_kdmbb(CPURISCVState *env, void *vd, void *va, + void *vb, uint8_t i) +{ + target_long *d = vd; + int16_t *a = va, *b = vb; + + if (a[H2(i)] == INT16_MIN && b[H2(i)] == INT16_MIN) { + *d = INT32_MAX; + env->vxsat = 0x1; + } else { + *d = (int64_t)a[H2(i)] * b[H2(i)] << 1; + } +} + +RVPR(kdmbb, 4, 2); + +static inline void do_kdmbt(CPURISCVState *env, void *vd, void *va, + void *vb, uint8_t i) +{ + target_long *d = vd; + int16_t *a = va, *b = vb; + + if (a[H2(i)] == INT16_MIN && b[H2(i + 1)] == INT16_MIN) { + *d = INT32_MAX; + env->vxsat = 0x1; + } else { + *d = (int64_t)a[H2(i)] * b[H2(i + 1)] << 1; + } +} + +RVPR(kdmbt, 4, 2); + +static inline void do_kdmtt(CPURISCVState *env, void *vd, void *va, + void *vb, uint8_t i) +{ + target_long *d = vd; + int16_t *a = va, *b = vb; + + if (a[H2(i + 1)] == INT16_MIN && b[H2(i + 1)] == INT16_MIN) { + *d = INT32_MAX; + env->vxsat = 0x1; + } else { + *d = (int64_t)a[H2(i + 1)] * b[H2(i + 1)] << 1; + } +} + +RVPR(kdmtt, 4, 2); + +static inline void do_kslraw(CPURISCVState *env, void *vd, void *va, + void *vb, uint8_t i) +{ + target_long *d = vd; + int32_t *a = va; + int32_t shift = sextract32((*(uint32_t *)vb), 0, 6); + + if (shift >= 0) { + *d = (int32_t)sat64(env, (int64_t)a[H4(i)] << shift, 31); + } else { + shift = -shift; + shift = (shift == 32) ? 31 : shift; + *d = a[H4(i)] >> shift; + } +} + +RVPR(kslraw, 2, 4); + +static inline void do_kslraw_u(CPURISCVState *env, void *vd, void *va, + void *vb, uint8_t i) +{ + target_long *d = vd; + int32_t *a = va; + int32_t shift = sextract32((*(uint32_t *)vb), 0, 6); + + if (shift >= 0) { + *d = (int32_t)sat64(env, (int64_t)a[H4(i)] << shift, 31); + } else { + shift = -shift; + shift = (shift == 32) ? 31 : shift; + *d = vssra32(env, 0, a[H4(i)], shift); + } +} + +RVPR(kslraw_u, 2, 4); + +static inline void do_ksllw(CPURISCVState *env, void *vd, void *va, + void *vb, uint8_t i) +{ + target_long *d = vd; + int32_t *a = va; + uint8_t shift = *(uint8_t *)vb & 0x1f; + + *d = (int32_t)sat64(env, (int64_t)a[H4(i)] << shift, 31); +} + +RVPR(ksllw, 2, 4); + +static inline void do_kdmabb(CPURISCVState *env, void *vd, void *va, + void *vb, void *vc, uint8_t i) + +{ + target_long *d = vd; + int16_t *a = va, *b = vb; + int32_t *c = vc, m0; + + if (a[H2(i)] == INT16_MIN && b[H2(i)] == INT16_MIN) { + m0 = INT32_MAX; + env->vxsat = 0x1; + } else { + m0 = (int32_t)a[H2(i)] * b[H2(i)] << 1; + } + *d = sadd32(env, 0, c[H4(i)], m0); +} + +RVPR_ACC(kdmabb, 4, 2); + +static inline void do_kdmabt(CPURISCVState *env, void *vd, void *va, + void *vb, void *vc, uint8_t i) + +{ + target_long *d = vd; + int16_t *a = va, *b = vb; + int32_t *c = vc, m0; + + if (a[H2(i)] == INT16_MIN && b[H2(i + 1)] == INT16_MIN) { + m0 = INT32_MAX; + env->vxsat = 0x1; + } else { + m0 = (int32_t)a[H2(i)] * b[H2(i + 1)] << 1; + } + *d = sadd32(env, 0, c[H4(i)], m0); +} + +RVPR_ACC(kdmabt, 4, 2); + +static inline void do_kdmatt(CPURISCVState *env, void *vd, void *va, + void *vb, void *vc, uint8_t i) + +{ + target_long *d = vd; + int16_t *a = va, *b = vb; + int32_t *c = vc, m0; + + if (a[H2(i + 1)] == INT16_MIN && b[H2(i + 1)] == INT16_MIN) { + m0 = INT32_MAX; + env->vxsat = 0x1; + } else { + m0 = (int32_t)a[H2(i + 1)] * b[H2(i + 1)] << 1; + } + *d = sadd32(env, 0, c[H4(i)], m0); +} + +RVPR_ACC(kdmatt, 4, 2); + +static inline void do_kabsw(CPURISCVState *env, void *vd, void *va, uint8_t i) + +{ + target_long *d = vd; + int32_t *a = va; + + if (a[H4(i)] == INT32_MIN) { + *d = INT32_MAX; + env->vxsat = 0x1; + } else { + *d = (int32_t)abs(a[H4(i)]); + } +} + +RVPR2(kabsw, 2, 4);