From patchwork Tue Nov 8 13:25:28 2016 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Laurent Vivier X-Patchwork-Id: 692295 Return-Path: X-Original-To: incoming@patchwork.ozlabs.org Delivered-To: patchwork-incoming@bilbo.ozlabs.org Received: from lists.gnu.org (lists.gnu.org [IPv6:2001:4830:134:3::11]) (using TLSv1 with cipher AES256-SHA (256/256 bits)) (No client certificate requested) by ozlabs.org (Postfix) with ESMTPS id 3tCqq0037Pz9t0w for ; Wed, 9 Nov 2016 00:27:08 +1100 (AEDT) Received: from localhost ([::1]:32976 helo=lists.gnu.org) by lists.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1c46Qb-0002uz-Nf for incoming@patchwork.ozlabs.org; Tue, 08 Nov 2016 08:27:05 -0500 Received: from eggs.gnu.org ([2001:4830:134:3::10]:36934) by lists.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1c46PR-0001ro-QL for qemu-devel@nongnu.org; Tue, 08 Nov 2016 08:25:55 -0500 Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71) (envelope-from ) id 1c46PL-0007gV-JZ for qemu-devel@nongnu.org; Tue, 08 Nov 2016 08:25:53 -0500 Received: from mout.kundenserver.de ([212.227.126.187]:64604) by eggs.gnu.org with esmtps (TLS1.0:DHE_RSA_AES_256_CBC_SHA1:32) (Exim 4.71) (envelope-from ) id 1c46PK-0007fN-TC for qemu-devel@nongnu.org; Tue, 08 Nov 2016 08:25:47 -0500 Received: from Quad.localdomain ([78.238.229.36]) by mrelayeu.kundenserver.de (mreue003) with ESMTPSA (Nemesis) id 0MTtH3-1cCaqc0F8r-00QgK4; Tue, 08 Nov 2016 14:25:35 +0100 From: Laurent Vivier To: qemu-devel@nongnu.org Date: Tue, 8 Nov 2016 14:25:28 +0100 Message-Id: <1478611528-24643-3-git-send-email-laurent@vivier.eu> X-Mailer: git-send-email 2.7.4 In-Reply-To: <1478611528-24643-1-git-send-email-laurent@vivier.eu> References: <1478611528-24643-1-git-send-email-laurent@vivier.eu> X-Provags-ID: V03:K0:Lt5mzCG0I24F4Gv/XsSjEEjyFjmVwERyHc8yc+vcaq5X6z/wHJE yDqSSdIZusNUCUniRdKp2L3khknr+zsUTBmWDF370Ioisy7fzNCTKtK5gnObrFE6YNuEtd9 AeFqiYwVHKmNK9kwiX9qgr2n+I1ox0val3s2Py72DPxBmDYkn79TOZF2raPgUI61TzN+r9J H8F/+80RvsVCw50LJq3dA== X-UI-Out-Filterresults: notjunk:1; V01:K0:V+5/JUOooIo=:tSVu86KhSreArXQMutvT8z b4E3J4SLwMiI167b5EkBGgv/OqvEdYcfM9sLhoqLTF37Vy/OsM8Zs5vjuD5w7+OyaPOfIT7Ye cMBcD2SBsEGJFTkYS1iR51UUO3kJ9mrS2fxQPszs60EXk2N380grAvykpgPHT+xM/ScMZDZeb P+dhc3oZYyR1jauur6GFIOFzZ9od0y252FDTPl5Mk20FycN2KQpB3+67FfImNAVZ416iDXUL+ oO1sRiQjshnwGPATGJaVYyadHXhDmfR2dzEn+PwAtxT3hNx4aWhSeATnJSFUd7IGPnxDVKEtr 7y0gdbjjuahgD8+1hisFoIWWy7FC7PdSaOPZp0F1V5ytlWko8ayI6mVAKQzxvgt2d+g3/xTsy guSDSEvI85bA0rSiznhki/bckRoFcaIbMyvZk69GBjtZBZLs2SLAjRQl+AUA7xS2hBGwNiXlk Sbdxtf0Jfi0ds6ZXHYMDX8ZuxiLBiN12Tgb+sEyUGWZR9/oDtk940P39gpoQuDJwPwikPFK0E WcjQj2D9RJ0caWz7uolHnRB5qVDZmIlvE0ZybYrkzUDgdXBDo1DkxSZk5HuJ/vhNwe30uKTm+ s4WyOD0pEF7DPuML/e87iykvM5gMbuq2/AFC7VZE7G09JagnUedwI7h3xZGpXpDNVwauaWHhn VXptUAWcjqJ890dD2F50hBbxJu7QH9abEhPf6ScGGRY8V7ufPJEDL51NrowF41phtWvs= X-detected-operating-system: by eggs.gnu.org: GNU/Linux 2.2.x-3.x [generic] X-Received-From: 212.227.126.187 Subject: [Qemu-devel] [PATCH v4 2/2] target-m68k: add cas/cas2 ops 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: Laurent Vivier , gerg@uclinux.org, schwab@linux-m68k.org, agraf@suse.de, Richard Henderson Errors-To: qemu-devel-bounces+incoming=patchwork.ozlabs.org@nongnu.org Sender: "Qemu-devel" Implement CAS using cmpxchg. Implement CAS2 using helper and either cmpxchg when the 32bit addresses are consecutive, or with parallel_cpus+cpu_loop_exit_atomic() otherwise. Suggested-by: Richard Henderson Signed-off-by: Laurent Vivier Reviewed-by: Richard Henderson --- target-m68k/helper.h | 2 + target-m68k/op_helper.c | 109 ++++++++++++++++++++++++++++++++++ target-m68k/translate.c | 153 ++++++++++++++++++++++++++++++++++++++++++++++++ 3 files changed, 264 insertions(+) diff --git a/target-m68k/helper.h b/target-m68k/helper.h index d863e55..17ec342 100644 --- a/target-m68k/helper.h +++ b/target-m68k/helper.h @@ -9,6 +9,8 @@ DEF_HELPER_4(divull, void, env, int, int, i32) DEF_HELPER_4(divsll, void, env, int, int, s32) DEF_HELPER_2(set_sr, void, env, i32) DEF_HELPER_3(movec, void, env, i32, i32) +DEF_HELPER_4(cas2w, void, env, i32, i32, i32) +DEF_HELPER_4(cas2l, void, env, i32, i32, i32) DEF_HELPER_2(f64_to_i32, f32, env, f64) DEF_HELPER_2(f64_to_f32, f32, env, f64) diff --git a/target-m68k/op_helper.c b/target-m68k/op_helper.c index a4bfa4e..52edd42 100644 --- a/target-m68k/op_helper.c +++ b/target-m68k/op_helper.c @@ -359,3 +359,112 @@ void HELPER(divsll)(CPUM68KState *env, int numr, int regr, int32_t den) env->dregs[regr] = rem; env->dregs[numr] = quot; } + +void HELPER(cas2w)(CPUM68KState *env, uint32_t regs, uint32_t a1, uint32_t a2) +{ + uint32_t Dc1 = extract32(regs, 9, 3); + uint32_t Dc2 = extract32(regs, 6, 3); + uint32_t Du1 = extract32(regs, 3, 3); + uint32_t Du2 = extract32(regs, 0, 3); + int16_t c1 = env->dregs[Dc1]; + int16_t c2 = env->dregs[Dc2]; + int16_t u1 = env->dregs[Du1]; + int16_t u2 = env->dregs[Du2]; + int16_t l1, l2; + uintptr_t ra = GETPC(); + + if (parallel_cpus) { + /* Tell the main loop we need to serialize this insn. */ + cpu_loop_exit_atomic(ENV_GET_CPU(env), ra); + } else { + /* We're executing in a serial context -- no need to be atomic. */ + l1 = cpu_lduw_data_ra(env, a1, ra); + l2 = cpu_lduw_data_ra(env, a2, ra); + if (l1 == c1 && l2 == c2) { + cpu_stw_data_ra(env, a1, u1, ra); + cpu_stw_data_ra(env, a2, u2, ra); + } + } + + if (c1 != l1) { + env->cc_n = l1; + env->cc_v = c1; + } else { + env->cc_n = l2; + env->cc_v = c2; + } + env->cc_op = CC_OP_CMPW; + env->dregs[Dc1] = deposit32(env->dregs[Dc1], 0, 16, l1); + env->dregs[Dc2] = deposit32(env->dregs[Dc2], 0, 16, l2); +} + +void HELPER(cas2l)(CPUM68KState *env, uint32_t regs, uint32_t a1, uint32_t a2) +{ + uint32_t Dc1 = extract32(regs, 9, 3); + uint32_t Dc2 = extract32(regs, 6, 3); + uint32_t Du1 = extract32(regs, 3, 3); + uint32_t Du2 = extract32(regs, 0, 3); + uint32_t c1 = env->dregs[Dc1]; + uint32_t c2 = env->dregs[Dc2]; + uint32_t u1 = env->dregs[Du1]; + uint32_t u2 = env->dregs[Du2]; + uint32_t l1, l2; + uint64_t c, u, l; + uintptr_t ra = GETPC(); +#if defined(CONFIG_ATOMIC64) && !defined(CONFIG_USER_ONLY) + int mmu_idx = cpu_mmu_index(env, 0); + TCGMemOpIdx oi; +#endif + + if (parallel_cpus) { + /* We're executing in a parallel context -- must be atomic. */ +#ifdef CONFIG_ATOMIC64 + if ((a1 & 7) == 0 && a2 == a1 + 4) { + c = deposit64(c2, 32, 32, c1); + u = deposit64(u2, 32, 32, u1); +#ifdef CONFIG_USER_ONLY + l = helper_atomic_cmpxchgq_be(env, a1, c, u); +#else + oi = make_memop_idx(MO_BEQ, mmu_idx); + l = helper_atomic_cmpxchgq_be_mmu(env, a1, c, u, oi, ra); +#endif + l1 = l >> 32; + l2 = l; + } else if ((a2 & 7) == 0 && a1 == a2 + 4) { + c = deposit64(c1, 32, 32, c2); + u = deposit64(u1, 32, 32, u2); +#ifdef CONFIG_USER_ONLY + l = helper_atomic_cmpxchgq_be(env, a2, c, u); +#else + oi = make_memop_idx(MO_BEQ, mmu_idx); + l = helper_atomic_cmpxchgq_be_mmu(env, a2, c, u, oi, ra); +#endif + l2 = l >> 32; + l1 = l; + } else +#endif + { + /* Tell the main loop we need to serialize this insn. */ + cpu_loop_exit_atomic(ENV_GET_CPU(env), ra); + } + } else { + /* We're executing in a serial context -- no need to be atomic. */ + l1 = cpu_ldl_data_ra(env, a1, ra); + l2 = cpu_ldl_data_ra(env, a2, ra); + if (l1 == c1 && l2 == c2) { + cpu_stl_data_ra(env, a1, u1, ra); + cpu_stl_data_ra(env, a2, u2, ra); + } + } + + if (c1 != l1) { + env->cc_n = l1; + env->cc_v = c1; + } else { + env->cc_n = l2; + env->cc_v = c2; + } + env->cc_op = CC_OP_CMPL; + env->dregs[Dc1] = l1; + env->dregs[Dc2] = l2; +} diff --git a/target-m68k/translate.c b/target-m68k/translate.c index f4eba90..5bd6d9e 100644 --- a/target-m68k/translate.c +++ b/target-m68k/translate.c @@ -1878,6 +1878,154 @@ DISAS_INSN(arith_im) tcg_temp_free(dest); } +DISAS_INSN(cas) +{ + int opsize; + TCGv addr; + uint16_t ext; + TCGv load; + TCGv cmp; + TCGMemOp opc; + + switch ((insn >> 9) & 3) { + case 1: + opsize = OS_BYTE; + opc = MO_UB; + break; + case 2: + opsize = OS_WORD; + opc = MO_TEUW; + break; + case 3: + opsize = OS_LONG; + opc = MO_TEUL; + break; + default: + g_assert_not_reached(); + } + opc |= MO_ALIGN; + + ext = read_im16(env, s); + + /* cas Dc,Du, */ + + addr = gen_lea(env, s, insn, opsize); + if (IS_NULL_QREG(addr)) { + gen_addr_fault(s); + return; + } + + cmp = gen_extend(DREG(ext, 0), opsize, 0); + + /* if == Dc then + * = Du + * Dc = (because == Dc) + * else + * Dc = + */ + + load = tcg_temp_new(); + tcg_gen_atomic_cmpxchg_i32(load, addr, cmp, DREG(ext, 6), + IS_USER(s), opc); + gen_partset_reg(opsize, DREG(ext, 0), load); + + gen_update_cc_cmp(s, load, cmp, opsize); + tcg_temp_free(load); +} + +DISAS_INSN(cas2w) +{ + uint16_t ext1, ext2; + TCGv addr1, addr2; + TCGv regs; + + /* cas2 Dc1:Dc2,Du1:Du2,(Rn1):(Rn2) */ + + ext1 = read_im16(env, s); + + if (ext1 & 0x8000) { + /* Address Register */ + addr1 = AREG(ext1, 12); + } else { + /* Data Register */ + addr1 = DREG(ext1, 12); + } + + ext2 = read_im16(env, s); + if (ext2 & 0x8000) { + /* Address Register */ + addr2 = AREG(ext2, 12); + } else { + /* Data Register */ + addr2 = DREG(ext2, 12); + } + + /* if (R1) == Dc1 && (R2) == Dc2 then + * (R1) = Du1 + * (R2) = Du2 + * else + * Dc1 = (R1) + * Dc2 = (R2) + */ + + regs = tcg_const_i32(REG(ext2, 6) | + (REG(ext1, 6) << 3) | + (REG(ext2, 0) << 6) | + (REG(ext1, 0) << 9)); + gen_helper_cas2w(cpu_env, regs, addr1, addr2); + tcg_temp_free(regs); + + /* Note that cas2w also assigned to env->cc_op. */ + s->cc_op = CC_OP_CMPW; + s->cc_op_synced = 1; +} + +DISAS_INSN(cas2l) +{ + uint16_t ext1, ext2; + TCGv addr1, addr2, regs; + + /* cas2 Dc1:Dc2,Du1:Du2,(Rn1):(Rn2) */ + + ext1 = read_im16(env, s); + + if (ext1 & 0x8000) { + /* Address Register */ + addr1 = AREG(ext1, 12); + } else { + /* Data Register */ + addr1 = DREG(ext1, 12); + } + + ext2 = read_im16(env, s); + if (ext2 & 0x8000) { + /* Address Register */ + addr2 = AREG(ext2, 12); + } else { + /* Data Register */ + addr2 = DREG(ext2, 12); + } + + /* if (R1) == Dc1 && (R2) == Dc2 then + * (R1) = Du1 + * (R2) = Du2 + * else + * Dc1 = (R1) + * Dc2 = (R2) + */ + + regs = tcg_const_i32(REG(ext2, 6) | + (REG(ext1, 6) << 3) | + (REG(ext2, 0) << 6) | + (REG(ext1, 0) << 9)); + gen_helper_cas2l(cpu_env, regs, addr1, addr2); + tcg_temp_free(regs); + + /* Note that cas2l also assigned to env->cc_op. */ + s->cc_op = CC_OP_CMPL; + s->cc_op_synced = 1; +} + DISAS_INSN(byterev) { TCGv reg; @@ -3872,6 +4020,11 @@ void register_m68k_insns (CPUM68KState *env) INSN(arith_im, 0680, fff8, CF_ISA_A); INSN(arith_im, 0c00, ff38, CF_ISA_A); INSN(arith_im, 0c00, ff00, M68000); + INSN(cas, 0ac0, ffc0, CAS); + INSN(cas, 0cc0, ffc0, CAS); + INSN(cas, 0ec0, ffc0, CAS); + INSN(cas2w, 0cfc, ffff, CAS); + INSN(cas2l, 0efc, ffff, CAS); BASE(bitop_im, 0800, ffc0); BASE(bitop_im, 0840, ffc0); BASE(bitop_im, 0880, ffc0);