From patchwork Thu Sep 16 15:53:59 2021 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Alexander Graf X-Patchwork-Id: 12499597 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=-16.8 required=3.0 tests=BAYES_00, HEADER_FROM_DIFFERENT_DOMAINS,INCLUDES_CR_TRAILER,INCLUDES_PATCH, MAILING_LIST_MULTI,SPF_HELO_NONE,SPF_PASS,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 D143FC433FE for ; Thu, 16 Sep 2021 16:10:34 +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 6733361390 for ; Thu, 16 Sep 2021 16:10:34 +0000 (UTC) DMARC-Filter: OpenDMARC Filter v1.4.1 mail.kernel.org 6733361390 Authentication-Results: mail.kernel.org; dmarc=none (p=none dis=none) header.from=csgraf.de Authentication-Results: mail.kernel.org; spf=pass smtp.mailfrom=nongnu.org Received: from localhost ([::1]:47030 helo=lists1p.gnu.org) by lists.gnu.org with esmtp (Exim 4.90_1) (envelope-from ) id 1mQty5-0001OT-IU for qemu-devel@archiver.kernel.org; Thu, 16 Sep 2021 12:10:33 -0400 Received: from eggs.gnu.org ([2001:470:142:3::10]:46664) by lists.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256) (Exim 4.90_1) (envelope-from ) id 1mQtiO-00034d-BW; Thu, 16 Sep 2021 11:54:20 -0400 Received: from mail.csgraf.de ([85.25.223.15]:51372 helo=zulu616.server4you.de) by eggs.gnu.org with esmtp (Exim 4.90_1) (envelope-from ) id 1mQtiH-00058G-8M; Thu, 16 Sep 2021 11:54:20 -0400 Received: from localhost.localdomain (dynamic-077-009-016-098.77.9.pool.telefonica.de [77.9.16.98]) by csgraf.de (Postfix) with ESMTPSA id 31E0660806CB; Thu, 16 Sep 2021 17:54:08 +0200 (CEST) From: Alexander Graf To: QEMU Developers Subject: [PATCH v12 05/10] arm/hvf: Add a WFI handler Date: Thu, 16 Sep 2021 17:53:59 +0200 Message-Id: <20210916155404.86958-6-agraf@csgraf.de> X-Mailer: git-send-email 2.30.1 (Apple Git-130) In-Reply-To: <20210916155404.86958-1-agraf@csgraf.de> References: <20210916155404.86958-1-agraf@csgraf.de> MIME-Version: 1.0 Received-SPF: pass client-ip=85.25.223.15; envelope-from=agraf@csgraf.de; helo=zulu616.server4you.de X-Spam_score_int: -18 X-Spam_score: -1.9 X-Spam_bar: - X-Spam_report: (-1.9 / 5.0 requ) BAYES_00=-1.9, SPF_HELO_NONE=0.001, SPF_PASS=-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: Peter Maydell , Eduardo Habkost , Sergio Lopez , Peter Collingbourne , Richard Henderson , Cameron Esfahani , Roman Bolshakov , qemu-arm , Frank Yang , Paolo Bonzini , =?utf-8?q?Philippe_Mathieu-Daud=C3=A9?= Errors-To: qemu-devel-bounces+qemu-devel=archiver.kernel.org@nongnu.org Sender: "Qemu-devel" From: Peter Collingbourne Sleep on WFI until the VTIMER is due but allow ourselves to be woken up on IPI. In this implementation IPI is blocked on the CPU thread at startup and pselect() is used to atomically unblock the signal and begin sleeping. The signal is sent unconditionally so there's no need to worry about races between actually sleeping and the "we think we're sleeping" state. It may lead to an extra wakeup but that's better than missing it entirely. Signed-off-by: Peter Collingbourne [agraf: Remove unused 'set' variable, always advance PC on WFX trap, support vm stop / continue operations and cntv offsets] Signed-off-by: Alexander Graf Acked-by: Roman Bolshakov Reviewed-by: Sergio Lopez --- v6 -> v7: - Move WFI into function - Improve comment wording v8 -> v9: - Add support for cntv offsets - Improve code readability v11 -> v12: - Use muldiv64 - Declare ts at start of block --- accel/hvf/hvf-accel-ops.c | 5 +-- include/sysemu/hvf_int.h | 1 + target/arm/hvf/hvf.c | 79 +++++++++++++++++++++++++++++++++++++++ 3 files changed, 82 insertions(+), 3 deletions(-) diff --git a/accel/hvf/hvf-accel-ops.c b/accel/hvf/hvf-accel-ops.c index 4f75927a8e..93976f4ece 100644 --- a/accel/hvf/hvf-accel-ops.c +++ b/accel/hvf/hvf-accel-ops.c @@ -370,15 +370,14 @@ static int hvf_init_vcpu(CPUState *cpu) cpu->hvf = g_malloc0(sizeof(*cpu->hvf)); /* init cpu signals */ - sigset_t set; struct sigaction sigact; memset(&sigact, 0, sizeof(sigact)); sigact.sa_handler = dummy_signal; sigaction(SIG_IPI, &sigact, NULL); - pthread_sigmask(SIG_BLOCK, NULL, &set); - sigdelset(&set, SIG_IPI); + pthread_sigmask(SIG_BLOCK, NULL, &cpu->hvf->unblock_ipi_mask); + sigdelset(&cpu->hvf->unblock_ipi_mask, SIG_IPI); #ifdef __aarch64__ r = hv_vcpu_create(&cpu->hvf->fd, (hv_vcpu_exit_t **)&cpu->hvf->exit, NULL); diff --git a/include/sysemu/hvf_int.h b/include/sysemu/hvf_int.h index 7c245c7b11..6545f7cd61 100644 --- a/include/sysemu/hvf_int.h +++ b/include/sysemu/hvf_int.h @@ -52,6 +52,7 @@ struct hvf_vcpu_state { uint64_t fd; void *exit; bool vtimer_masked; + sigset_t unblock_ipi_mask; }; void assert_hvf_ok(hv_return_t ret); diff --git a/target/arm/hvf/hvf.c b/target/arm/hvf/hvf.c index f3b4023030..110a9d547d 100644 --- a/target/arm/hvf/hvf.c +++ b/target/arm/hvf/hvf.c @@ -2,6 +2,7 @@ * QEMU Hypervisor.framework support for Apple Silicon * Copyright 2020 Alexander Graf + * Copyright 2020 Google LLC * * This work is licensed under the terms of the GNU GPL, version 2 or later. * See the COPYING file in the top-level directory. @@ -490,6 +491,7 @@ int hvf_arch_init_vcpu(CPUState *cpu) void hvf_kick_vcpu_thread(CPUState *cpu) { + cpus_kick_thread(cpu); hv_vcpus_exit(&cpu->hvf->fd, 1); } @@ -608,6 +610,80 @@ static uint64_t hvf_vtimer_val_raw(void) return mach_absolute_time() - hvf_state->vtimer_offset; } +static uint64_t hvf_vtimer_val(void) +{ + if (!runstate_is_running()) { + /* VM is paused, the vtimer value is in vtimer.vtimer_val */ + return vtimer.vtimer_val; + } + + return hvf_vtimer_val_raw(); +} + +static void hvf_wait_for_ipi(CPUState *cpu, struct timespec *ts) +{ + /* + * Use pselect to sleep so that other threads can IPI us while we're + * sleeping. + */ + qatomic_mb_set(&cpu->thread_kicked, false); + qemu_mutex_unlock_iothread(); + pselect(0, 0, 0, 0, ts, &cpu->hvf->unblock_ipi_mask); + qemu_mutex_lock_iothread(); +} + +static void hvf_wfi(CPUState *cpu) +{ + ARMCPU *arm_cpu = ARM_CPU(cpu); + struct timespec ts; + hv_return_t r; + uint64_t ctl; + uint64_t cval; + int64_t ticks_to_sleep; + uint64_t seconds; + uint64_t nanos; + uint32_t cntfrq; + + if (cpu->interrupt_request & (CPU_INTERRUPT_HARD | CPU_INTERRUPT_FIQ)) { + /* Interrupt pending, no need to wait */ + return; + } + + r = hv_vcpu_get_sys_reg(cpu->hvf->fd, HV_SYS_REG_CNTV_CTL_EL0, &ctl); + assert_hvf_ok(r); + + if (!(ctl & 1) || (ctl & 2)) { + /* Timer disabled or masked, just wait for an IPI. */ + hvf_wait_for_ipi(cpu, NULL); + return; + } + + r = hv_vcpu_get_sys_reg(cpu->hvf->fd, HV_SYS_REG_CNTV_CVAL_EL0, &cval); + assert_hvf_ok(r); + + ticks_to_sleep = cval - hvf_vtimer_val(); + if (ticks_to_sleep < 0) { + return; + } + + cntfrq = gt_cntfrq_period_ns(arm_cpu); + seconds = muldiv64(ticks_to_sleep, cntfrq, NANOSECONDS_PER_SECOND); + ticks_to_sleep -= muldiv64(seconds, NANOSECONDS_PER_SECOND, cntfrq); + nanos = ticks_to_sleep * cntfrq; + + /* + * Don't sleep for less than the time a context switch would take, + * so that we can satisfy fast timer requests on the same CPU. + * Measurements on M1 show the sweet spot to be ~2ms. + */ + if (!seconds && nanos < (2 * SCALE_MS)) { + return; + } + + ts = (struct timespec) { seconds, nanos }; + hvf_wait_for_ipi(cpu, &ts); +} + static void hvf_sync_vtimer(CPUState *cpu) { ARMCPU *arm_cpu = ARM_CPU(cpu); @@ -728,6 +804,9 @@ int hvf_vcpu_exec(CPUState *cpu) } case EC_WFX_TRAP: advance_pc = true; + if (!(syndrome & WFX_IS_WFE)) { + hvf_wfi(cpu); + } break; case EC_AA64_HVC: cpu_synchronize_state(cpu);