From patchwork Tue Feb 11 00:40:22 2020 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Taylor Simpson X-Patchwork-Id: 1236112 Return-Path: X-Original-To: incoming@patchwork.ozlabs.org Delivered-To: patchwork-incoming@bilbo.ozlabs.org Authentication-Results: ozlabs.org; spf=pass (sender SPF authorized) smtp.mailfrom=nongnu.org (client-ip=209.51.188.17; helo=lists.gnu.org; envelope-from=qemu-devel-bounces+incoming=patchwork.ozlabs.org@nongnu.org; receiver=) Authentication-Results: ozlabs.org; dmarc=fail (p=none dis=none) header.from=quicinc.com Authentication-Results: ozlabs.org; dkim=fail reason="signature verification failed" (1024-bit key; unprotected) header.d=quicinc.com header.i=@quicinc.com header.a=rsa-sha256 header.s=qcdkim header.b=sOP5Xt3x; dkim-atps=neutral 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 ozlabs.org (Postfix) with ESMTPS id 48GlGf1lTyz9sRL for ; Tue, 11 Feb 2020 12:16:26 +1100 (AEDT) Received: from localhost ([::1]:42034 helo=lists1p.gnu.org) by lists.gnu.org with esmtp (Exim 4.90_1) (envelope-from ) id 1j1KA4-0004p4-4s for incoming@patchwork.ozlabs.org; Mon, 10 Feb 2020 20:16:24 -0500 Received: from eggs.gnu.org ([2001:470:142:3::10]:33552) by lists.gnu.org with esmtp (Exim 4.90_1) (envelope-from ) id 1j1Jce-00029u-Cx for qemu-devel@nongnu.org; Mon, 10 Feb 2020 19:41:53 -0500 Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71) (envelope-from ) id 1j1Jcc-0000n4-Kg for qemu-devel@nongnu.org; Mon, 10 Feb 2020 19:41:52 -0500 Received: from alexa-out-sd-01.qualcomm.com ([199.106.114.38]:3646) by eggs.gnu.org with esmtps (TLS1.0:RSA_AES_256_CBC_SHA1:32) (Exim 4.71) (envelope-from ) id 1j1JcZ-0004rG-0r for qemu-devel@nongnu.org; Mon, 10 Feb 2020 19:41:50 -0500 DKIM-Signature: v=1; a=rsa-sha256; c=simple/simple; d=quicinc.com; i=@quicinc.com; q=dns/txt; s=qcdkim; t=1581381707; x=1612917707; h=from:to:cc:subject:date:message-id:in-reply-to: references:mime-version:content-transfer-encoding; bh=v2Hkp+624/czX2fsXyQYAbJeOCZH24FXWJGD3s5YaQ4=; b=sOP5Xt3xXQKsF4DepaqBP3mJpKh99MeWv22rYIjsfTPSHmdzm1lKHZY6 7wuHA08BcUKmfrZ111WFp3xV7ZyFozxfxZWJq4tjbxnB7py+gkw0N84Ni tAIwb31Nsy1wURjro87daDRWn+kE+t6FEILbdAvTBdnc8dQlKMTNFSh5x 0=; Received: from unknown (HELO ironmsg03-sd.qualcomm.com) ([10.53.140.143]) by alexa-out-sd-01.qualcomm.com with ESMTP; 10 Feb 2020 16:41:02 -0800 Received: from vu-tsimpson-aus.qualcomm.com (HELO vu-tsimpson1-aus.qualcomm.com) ([10.222.150.1]) by ironmsg03-sd.qualcomm.com with ESMTP; 10 Feb 2020 16:41:01 -0800 Received: by vu-tsimpson1-aus.qualcomm.com (Postfix, from userid 47164) id 7879F1B27; Mon, 10 Feb 2020 18:41:01 -0600 (CST) From: Taylor Simpson To: qemu-devel@nongnu.org Subject: [RFC PATCH 44/66] Hexagon TCG generation - step 07 Date: Mon, 10 Feb 2020 18:40:22 -0600 Message-Id: <1581381644-13678-45-git-send-email-tsimpson@quicinc.com> X-Mailer: git-send-email 2.7.4 In-Reply-To: <1581381644-13678-1-git-send-email-tsimpson@quicinc.com> References: <1581381644-13678-1-git-send-email-tsimpson@quicinc.com> MIME-Version: 1.0 X-detected-operating-system: by eggs.gnu.org: FreeBSD 9.x [fuzzy] X-Received-From: 199.106.114.38 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: riku.voipio@iki.fi, richard.henderson@linaro.org, laurent@vivier.eu, Taylor Simpson , philmd@redhat.com, aleksandar.m.mail@gmail.com Errors-To: qemu-devel-bounces+incoming=patchwork.ozlabs.org@nongnu.org Sender: "Qemu-devel" Override dczeroa, allocframe, and return instructions Signed-off-by: Taylor Simpson --- target/hexagon/helper_overrides.h | 209 ++++++++++++++++++++++++++++++++++++++ 1 file changed, 209 insertions(+) diff --git a/target/hexagon/helper_overrides.h b/target/hexagon/helper_overrides.h index f023442..c7550d5 100644 --- a/target/hexagon/helper_overrides.h +++ b/target/hexagon/helper_overrides.h @@ -991,4 +991,213 @@ #define fWRAP_L4_ior_memoph_io(GENHLPR, SHORTCODE) \ fWRAP_MEMOP(GENHLPR, SHORTCODE, 2, tcg_gen_ori_tl(tmp, tmp, 1 << UiV)) +/* dczeroa clears the 32 byte cache line at the address given */ +#define fWRAP_Y2_dczeroa(GENHLPR, SHORTCODE) SHORTCODE + +/* We have to brute force allocframe because it has C math in the semantics */ +#define fWRAP_S2_allocframe(GENHLPR, SHORTCODE) \ + do { \ + TCGv_i64 scramble_tmp = tcg_temp_new_i64(); \ + TCGv tmp = tcg_temp_new(); \ + { fEA_RI(RxV, -8); \ + fSTORE(1, 8, EA, fFRAME_SCRAMBLE((fCAST8_8u(fREAD_LR()) << 32) | \ + fCAST4_4u(fREAD_FP()))); \ + fWRITE_FP(EA); \ + fFRAMECHECK(EA - uiV, EA); \ + tcg_gen_subi_tl(RxV, EA, uiV); \ + } \ + tcg_temp_free_i64(scramble_tmp); \ + tcg_temp_free(tmp); \ + } while (0) + +#define fWRAP_SS2_allocframe(GENHLPR, SHORTCODE) \ + do { \ + TCGv_i64 scramble_tmp = tcg_temp_new_i64(); \ + TCGv tmp = tcg_temp_new(); \ + { fEA_RI(fREAD_SP(), -8); \ + fSTORE(1, 8, EA, fFRAME_SCRAMBLE((fCAST8_8u(fREAD_LR()) << 32) | \ + fCAST4_4u(fREAD_FP()))); \ + fWRITE_FP(EA); \ + fFRAMECHECK(EA - uiV, EA); \ + tcg_gen_subi_tl(tmp, EA, uiV); \ + fWRITE_SP(tmp); \ + } \ + tcg_temp_free_i64(scramble_tmp); \ + tcg_temp_free(tmp); \ + } while (0) + +/* Also have to brute force the deallocframe variants */ +#define fWRAP_L2_deallocframe(GENHLPR, SHORTCODE) \ + do { \ + TCGv tmp = tcg_temp_new(); \ + TCGv_i64 tmp_i64 = tcg_temp_new_i64(); \ + { \ + fEA_REG(RsV); \ + fLOAD(1, 8, u, EA, tmp_i64); \ + tcg_gen_mov_i64(RddV, fFRAME_UNSCRAMBLE(tmp_i64)); \ + tcg_gen_addi_tl(tmp, EA, 8); \ + fWRITE_SP(tmp); \ + } \ + tcg_temp_free(tmp); \ + tcg_temp_free_i64(tmp_i64); \ + } while (0) + +#define fWRAP_SL2_deallocframe(GENHLPR, SHORTCODE) \ + do { \ + TCGv WORD = tcg_temp_new(); \ + TCGv tmp = tcg_temp_new(); \ + TCGv_i64 tmp_i64 = tcg_temp_new_i64(); \ + { \ + fEA_REG(fREAD_FP()); \ + fLOAD(1, 8, u, EA, tmp_i64); \ + fFRAME_UNSCRAMBLE(tmp_i64); \ + fWRITE_LR(fGETWORD(1, tmp_i64)); \ + fWRITE_FP(fGETWORD(0, tmp_i64)); \ + tcg_gen_addi_tl(tmp, EA, 8); \ + fWRITE_SP(tmp); \ + } \ + tcg_temp_free(WORD); \ + tcg_temp_free(tmp); \ + tcg_temp_free_i64(tmp_i64); \ + } while (0) + +#define fWRAP_L4_return(GENHLPR, SHORTCODE) \ + do { \ + TCGv tmp = tcg_temp_new(); \ + TCGv_i64 tmp_i64 = tcg_temp_new_i64(); \ + TCGv WORD = tcg_temp_new(); \ + { \ + fEA_REG(RsV); \ + fLOAD(1, 8, u, EA, tmp_i64); \ + tcg_gen_mov_i64(RddV, fFRAME_UNSCRAMBLE(tmp_i64)); \ + tcg_gen_addi_tl(tmp, EA, 8); \ + fWRITE_SP(tmp); \ + fJUMPR(REG_LR, fGETWORD(1, RddV), COF_TYPE_JUMPR);\ + } \ + tcg_temp_free(tmp); \ + tcg_temp_free_i64(tmp_i64); \ + tcg_temp_free(WORD); \ + } while (0) + +#define fWRAP_SL2_return(GENHLPR, SHORTCODE) \ + do { \ + TCGv tmp = tcg_temp_new(); \ + TCGv_i64 tmp_i64 = tcg_temp_new_i64(); \ + TCGv WORD = tcg_temp_new(); \ + { \ + fEA_REG(fREAD_FP()); \ + fLOAD(1, 8, u, EA, tmp_i64); \ + fFRAME_UNSCRAMBLE(tmp_i64); \ + fWRITE_LR(fGETWORD(1, tmp_i64)); \ + fWRITE_FP(fGETWORD(0, tmp_i64)); \ + tcg_gen_addi_tl(tmp, EA, 8); \ + fWRITE_SP(tmp); \ + fJUMPR(REG_LR, fGETWORD(1, tmp_i64), COF_TYPE_JUMPR);\ + } \ + tcg_temp_free(tmp); \ + tcg_temp_free_i64(tmp_i64); \ + tcg_temp_free(WORD); \ + } while (0) + +/* + * Conditional returns follow the same predicate naming convention as + * predicated loads above + */ +#define fWRAP_COND_RETURN(PRED) \ + do { \ + TCGv LSB = tcg_temp_new(); \ + TCGv_i64 LSB_i64 = tcg_temp_new_i64(); \ + TCGv zero = tcg_const_tl(0); \ + TCGv_i64 zero_i64 = tcg_const_i64(0); \ + TCGv_i64 unscramble = tcg_temp_new_i64(); \ + TCGv WORD = tcg_temp_new(); \ + TCGv SP = tcg_temp_new(); \ + TCGv_i64 tmp_i64 = tcg_temp_new_i64(); \ + TCGv tmp = tcg_temp_new(); \ + fEA_REG(RsV); \ + PRED; \ + tcg_gen_extu_i32_i64(LSB_i64, LSB); \ + fLOAD(1, 8, u, EA, tmp_i64); \ + tcg_gen_mov_i64(unscramble, fFRAME_UNSCRAMBLE(tmp_i64)); \ + READ_REG_PAIR(RddV, HEX_REG_FP); \ + tcg_gen_movcond_i64(TCG_COND_NE, RddV, LSB_i64, zero_i64, \ + unscramble, RddV); \ + tcg_gen_mov_tl(SP, hex_gpr[HEX_REG_SP]); \ + tcg_gen_addi_tl(tmp, EA, 8); \ + tcg_gen_movcond_tl(TCG_COND_NE, SP, LSB, zero, tmp, SP); \ + fWRITE_SP(SP); \ + gen_cond_return(LSB, fGETWORD(1, RddV)); \ + tcg_temp_free(LSB); \ + tcg_temp_free_i64(LSB_i64); \ + tcg_temp_free(zero); \ + tcg_temp_free_i64(zero_i64); \ + tcg_temp_free_i64(unscramble); \ + tcg_temp_free(WORD); \ + tcg_temp_free(SP); \ + tcg_temp_free_i64(tmp_i64); \ + tcg_temp_free(tmp); \ + } while (0) + +#define fWRAP_L4_return_t(GENHLPR, SHORTCODE) \ + fWRAP_COND_RETURN(fLSBOLD(PvV)) +#define fWRAP_L4_return_f(GENHLPR, SHORTCODE) \ + fWRAP_COND_RETURN(fLSBOLDNOT(PvV)) +#define fWRAP_L4_return_tnew_pt(GENHLPR, SHORTCODE) \ + fWRAP_COND_RETURN(fLSBNEW(PvN)) +#define fWRAP_L4_return_fnew_pt(GENHLPR, SHORTCODE) \ + fWRAP_COND_RETURN(fLSBNEWNOT(PvN)) +#define fWRAP_L4_return_tnew_pnt(GENHLPR, SHORTCODE) \ + fWRAP_COND_RETURN(fLSBNEW(PvN)) +#define fWRAP_L4_return_tnew_pnt(GENHLPR, SHORTCODE) \ + fWRAP_COND_RETURN(fLSBNEW(PvN)) +#define fWRAP_L4_return_fnew_pnt(GENHLPR, SHORTCODE) \ + fWRAP_COND_RETURN(fLSBNEWNOT(PvN)) + +#define fWRAP_COND_RETURN_SUBINSN(PRED) \ + do { \ + TCGv LSB = tcg_temp_new(); \ + TCGv_i64 LSB_i64 = tcg_temp_new_i64(); \ + TCGv zero = tcg_const_tl(0); \ + TCGv_i64 zero_i64 = tcg_const_i64(0); \ + TCGv_i64 unscramble = tcg_temp_new_i64(); \ + TCGv_i64 RddV = tcg_temp_new_i64(); \ + TCGv WORD = tcg_temp_new(); \ + TCGv SP = tcg_temp_new(); \ + TCGv_i64 tmp_i64 = tcg_temp_new_i64(); \ + TCGv tmp = tcg_temp_new(); \ + fEA_REG(fREAD_FP()); \ + PRED; \ + tcg_gen_extu_i32_i64(LSB_i64, LSB); \ + fLOAD(1, 8, u, EA, tmp_i64); \ + tcg_gen_mov_i64(unscramble, fFRAME_UNSCRAMBLE(tmp_i64)); \ + READ_REG_PAIR(RddV, HEX_REG_FP); \ + tcg_gen_movcond_i64(TCG_COND_NE, RddV, LSB_i64, zero_i64, \ + unscramble, RddV); \ + tcg_gen_mov_tl(SP, hex_gpr[HEX_REG_SP]); \ + tcg_gen_addi_tl(tmp, EA, 8); \ + tcg_gen_movcond_tl(TCG_COND_NE, SP, LSB, zero, tmp, SP); \ + fWRITE_SP(SP); \ + WRITE_REG_PAIR(HEX_REG_FP, RddV); \ + gen_cond_return(LSB, fGETWORD(1, RddV)); \ + tcg_temp_free(LSB); \ + tcg_temp_free_i64(LSB_i64); \ + tcg_temp_free(zero); \ + tcg_temp_free_i64(zero_i64); \ + tcg_temp_free_i64(unscramble); \ + tcg_temp_free_i64(RddV); \ + tcg_temp_free(WORD); \ + tcg_temp_free(SP); \ + tcg_temp_free_i64(tmp_i64); \ + tcg_temp_free(tmp); \ + } while (0) + +#define fWRAP_SL2_return_t(GENHLPR, SHORTCODE) \ + fWRAP_COND_RETURN_SUBINSN(fLSBOLD(fREAD_P0())) +#define fWRAP_SL2_return_f(GENHLPR, SHORTCODE) \ + fWRAP_COND_RETURN_SUBINSN(fLSBOLDNOT(fREAD_P0())) +#define fWRAP_SL2_return_tnew(GENHLPR, SHORTCODE) \ + fWRAP_COND_RETURN_SUBINSN(fLSBNEW0) +#define fWRAP_SL2_return_fnew(GENHLPR, SHORTCODE) \ + fWRAP_COND_RETURN_SUBINSN(fLSBNEW0NOT) + #endif