From patchwork Wed May 29 15:22:19 2024 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 8bit X-Patchwork-Submitter: =?utf-8?q?Alex_Benn=C3=A9e?= X-Patchwork-Id: 799852 Delivered-To: patch@linaro.org Received: by 2002:adf:e110:0:b0:35b:5a80:51b4 with SMTP id t16csp789189wrz; Wed, 29 May 2024 08:23:00 -0700 (PDT) X-Forwarded-Encrypted: i=2; AJvYcCUd4gKgBdGnSRBEgVTvDyMPzwgJGg/LSKt7Fes1y/2IvqK7fem4+77yaNVwKJVDr7yxrN9Wo5/d3XuvUo5oXbIX X-Google-Smtp-Source: AGHT+IEv5hFD/AjfbI2LF9xiZY/ovJs158n5xhoto/JIWKFJxNxi9rVE4S8pX8nx2Y+N2nzQinji X-Received: by 2002:ac8:7dcc:0:b0:43a:cdb9:a8e2 with SMTP id d75a77b69052e-43fb0f1dcafmr159236261cf.65.1716996180401; Wed, 29 May 2024 08:23:00 -0700 (PDT) ARC-Seal: i=1; a=rsa-sha256; t=1716996180; cv=none; d=google.com; s=arc-20160816; b=c7403YrHN0aePAzznx4uDVdOZFEeQKJ6wNTvo4Nn/wbaMUmnXjAUXMdI3wqF9+Un1X R/MS+I6HCWaKZzXTf18jdwVmDeDYzGduxMc2qy4GeAjyKxFsKQ4ilno/wHd6ujFTJkwy XPFaK70x+Q0b8o0yu7MYe7d5FPUPYwaroLVkZzNypZ3K1I0vjkhe9D2iTYY78ehiuTbd /HeoCOJ7JeVNB5wp3EKBpxNSaUvyjtJUbv9DXbGgDUvAuMLwqdKd3SvK9trGYbrOsi+Z vx5YEVRvklgfaw/CkZNnDbjWUHJUiaRWZpP3wmD/3Ggfg9kHDcty4swRAWy4DCblsJzd 7MAg== ARC-Message-Signature: i=1; a=rsa-sha256; c=relaxed/relaxed; d=google.com; s=arc-20160816; h=sender:errors-to:list-subscribe:list-help:list-post:list-archive :list-unsubscribe:list-id:precedence:content-transfer-encoding :mime-version:message-id:date:subject:cc:to:from:dkim-signature; bh=fTksdLPLiWK2it8kbCH0j2Wz6G5Oz9kxuWwnHiWElfQ=; fh=0ILNp12mujKgID/C7NNVpT4Y1ShxdYIlYDpVBx3JrPs=; b=p7HitlWSOehiXPGYm8+YOLr4rMd54/9E72kcIixR8lmV/ivc9fJ1d7rFoW83ZQL2wG tVxQy3R11ifn6Boj20csT5UJ/yKthD6Trifzp1ciCkC5Rg4KFInud1TRGpGTRpBTriK6 Am81yB48isnP26x2qALPr/L4wkaDQkl+knS+e9JzOhCIHUM9Hfc5/JgZUn93MekCbzqA A86WVYpwjJdLa+lHQE/4p7TzgfUDv7cJCFk8axeF5+8tUZgrTETJjZnObkXJsVriNjHQ Fh0/pk0fhb3mal55EQtzwUMy1qcBrLrrmqF3TrgKqmgXrjcUWugnxozApTzErFF17Oa4 lNfg==; dara=google.com ARC-Authentication-Results: i=1; mx.google.com; dkim=pass header.i=@linaro.org header.s=google header.b=L4XcBMv4; spf=pass (google.com: domain of qemu-devel-bounces+patch=linaro.org@nongnu.org designates 209.51.188.17 as permitted sender) smtp.mailfrom="qemu-devel-bounces+patch=linaro.org@nongnu.org"; dmarc=pass (p=NONE sp=NONE dis=NONE) header.from=linaro.org Return-Path: Received: from lists.gnu.org (lists.gnu.org. [209.51.188.17]) by mx.google.com with ESMTPS id d75a77b69052e-43fb17baacfsi144694611cf.58.2024.05.29.08.23.00 for (version=TLS1_2 cipher=ECDHE-ECDSA-CHACHA20-POLY1305 bits=256/256); Wed, 29 May 2024 08:23:00 -0700 (PDT) Received-SPF: pass (google.com: domain of qemu-devel-bounces+patch=linaro.org@nongnu.org designates 209.51.188.17 as permitted sender) client-ip=209.51.188.17; Authentication-Results: mx.google.com; dkim=pass header.i=@linaro.org header.s=google header.b=L4XcBMv4; spf=pass (google.com: domain of qemu-devel-bounces+patch=linaro.org@nongnu.org designates 209.51.188.17 as permitted sender) smtp.mailfrom="qemu-devel-bounces+patch=linaro.org@nongnu.org"; dmarc=pass (p=NONE sp=NONE dis=NONE) header.from=linaro.org Received: from localhost ([::1] helo=lists1p.gnu.org) by lists.gnu.org with esmtp (Exim 4.90_1) (envelope-from ) id 1sCL8M-0001de-7q; Wed, 29 May 2024 11:22:34 -0400 Received: from eggs.gnu.org ([2001:470:142:3::10]) by lists.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256) (Exim 4.90_1) (envelope-from ) id 1sCL8L-0001dS-8T for qemu-devel@nongnu.org; Wed, 29 May 2024 11:22:33 -0400 Received: from mail-lj1-x231.google.com ([2a00:1450:4864:20::231]) by eggs.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_128_GCM_SHA256:128) (Exim 4.90_1) (envelope-from ) id 1sCL8J-0001nS-6M for qemu-devel@nongnu.org; Wed, 29 May 2024 11:22:32 -0400 Received: by mail-lj1-x231.google.com with SMTP id 38308e7fff4ca-2e9819a630fso5280941fa.1 for ; Wed, 29 May 2024 08:22:30 -0700 (PDT) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=linaro.org; s=google; t=1716996149; x=1717600949; darn=nongnu.org; h=content-transfer-encoding:mime-version:message-id:date:subject:cc :to:from:from:to:cc:subject:date:message-id:reply-to; bh=fTksdLPLiWK2it8kbCH0j2Wz6G5Oz9kxuWwnHiWElfQ=; b=L4XcBMv4gzLkB6c6XgYzQJuxc66ORCjFJ2aS0Gdu68Mc6OVDlsi8fnwXQ35S0iQNNA jR6XzRmsdsiSjavWooXvWxw+g0I0EQm+v5EF0NWIh8mEHdzXB6OUK3cHs1Z3Xn81PPcf pDYzgcBdx8YInuMcWvYoUtJam7PCnT2lWaT9L0pNE67pxkHry+EhgqXTLb+Vk822nLV6 4PLWe8uk2lxcPMk+t00R7FS/Nd1G0qzwBjOXHf5JieZlTonsTsbxsI3A1J7iY0BuPSUk NApVllyHSKq4C7f2cGa1fh6KAY6sRMHqOplAg/JEhAA6SJfooGwZ/mLl0bhjzrhRqCZ/ zQBQ== X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20230601; t=1716996149; x=1717600949; h=content-transfer-encoding:mime-version:message-id:date:subject:cc :to:from:x-gm-message-state:from:to:cc:subject:date:message-id :reply-to; bh=fTksdLPLiWK2it8kbCH0j2Wz6G5Oz9kxuWwnHiWElfQ=; b=SxtsX1YWTlMEuWH47ex5Elcrr9V3VTUNNabXptXZElx4jb9rpcbDe/dlNDPO2vP98u zz8EoLfTycAdTvetxjfiYLaqmS8N3OTjcC7di1lZ/Etrd96uUSCkLPGyxdOxkIxUx3IS Q6YMVEuQz+9/clkuNvBh4DvBjkb0m8XUJ6EjTK5s1f6UiibdjPtCyVHbdBKKyA5l9GoJ ViL4O/yQJfTYDbOJTYeAVGgrIdIpSJBDzH3BCKT5v8w+fbrc5gvzIe98n1es6NGkH++I EYnTee6sOLvCYZJwrUT00/mXVOGpbDJ+OhZxmKTIx0YL8bS5ciDgGmA5BPbC+pOOq50G fN9A== X-Gm-Message-State: AOJu0YzmkADj2YjFE95tZbvKq5fHRgDnRHThLQlq3f6x6D1iXZqJallX OPJ9XYYXQ4hHcngZNrFozTof6yMVosOLQMlbzv8yf7iy+ziKVcRVPm3i/q8LYb0= X-Received: by 2002:a2e:804b:0:b0:2ea:8163:5f4f with SMTP id 38308e7fff4ca-2ea816360bcmr4005311fa.41.1716996148999; Wed, 29 May 2024 08:22:28 -0700 (PDT) Received: from draig.lan ([85.9.250.243]) by smtp.gmail.com with ESMTPSA id a640c23a62f3a-a626cda31b4sm729528466b.191.2024.05.29.08.22.28 (version=TLS1_3 cipher=TLS_AES_256_GCM_SHA384 bits=256/256); Wed, 29 May 2024 08:22:28 -0700 (PDT) Received: from draig.lan (localhost [IPv6:::1]) by draig.lan (Postfix) with ESMTP id 081345F760; Wed, 29 May 2024 16:22:28 +0100 (BST) From: =?utf-8?q?Alex_Benn=C3=A9e?= To: qemu-devel@nongnu.org Cc: =?utf-8?q?Alex_Benn=C3=A9e?= , Pierrick Bouvier , =?utf-8?q?Philippe_Mathieu-?= =?utf-8?q?Daud=C3=A9?= , Riku Voipio , Richard Henderson , Paolo Bonzini , Eduardo Habkost , Marcel Apfelbaum , Yanan Wang , Alexandre Iooss , Mahmoud Mandour Subject: [RFC PATCH] cpus: split qemu_init_vcpu and delay vCPU thread creation Date: Wed, 29 May 2024 16:22:19 +0100 Message-Id: <20240529152219.825680-1-alex.bennee@linaro.org> X-Mailer: git-send-email 2.39.2 MIME-Version: 1.0 Received-SPF: pass client-ip=2a00:1450:4864:20::231; envelope-from=alex.bennee@linaro.org; helo=mail-lj1-x231.google.com X-Spam_score_int: -20 X-Spam_score: -2.1 X-Spam_bar: -- X-Spam_report: (-2.1 / 5.0 requ) BAYES_00=-1.9, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, DKIM_VALID_EF=-0.1, RCVD_IN_DNSWL_NONE=-0.0001, SPF_HELO_NONE=0.001, SPF_PASS=-0.001, T_SCC_BODY_TEXT_LINE=-0.01 autolearn=ham autolearn_force=no X-Spam_action: no action X-BeenThere: qemu-devel@nongnu.org X-Mailman-Version: 2.1.29 Precedence: list List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Errors-To: qemu-devel-bounces+patch=linaro.org@nongnu.org Sender: qemu-devel-bounces+patch=linaro.org@nongnu.org This ensures we don't start the thread until cpu_common_realizefn has finished. This ensures that plugins will always run qemu_plugin_vcpu_init__async first before any other states. It doesn't totally eliminate the race that plugin_cpu_update__locked has to work around though. I found this while reviewing the ips plugin which makes heavy use of the vcpu phase callbacks. An alternative might be to move the explicit creation of vCPU threads to qdev_machine_creation_done()? It doesn't affect user-mode which already has a thread to execute in and ensures the QOM object has completed creation in cpu_create() before continuing. Signed-off-by: Alex Bennée Cc: Pierrick Bouvier Cc: Philippe Mathieu-Daudé --- include/hw/core/cpu.h | 8 ++++++++ accel/tcg/user-exec-stub.c | 5 +++++ hw/core/cpu-common.c | 7 ++++++- plugins/core.c | 5 +++++ system/cpus.c | 15 ++++++++++----- 5 files changed, 34 insertions(+), 6 deletions(-) diff --git a/include/hw/core/cpu.h b/include/hw/core/cpu.h index bb398e8237..6920699585 100644 --- a/include/hw/core/cpu.h +++ b/include/hw/core/cpu.h @@ -1041,6 +1041,14 @@ void end_exclusive(void); */ void qemu_init_vcpu(CPUState *cpu); +/** + * qemu_start_vcpu: + * @cpu: The vCPU to start. + * + * Create the vCPU thread and start it running. + */ +void qemu_start_vcpu(CPUState *cpu); + #define SSTEP_ENABLE 0x1 /* Enable simulated HW single stepping */ #define SSTEP_NOIRQ 0x2 /* Do not use IRQ while single stepping */ #define SSTEP_NOTIMER 0x4 /* Do not Timers while single stepping */ diff --git a/accel/tcg/user-exec-stub.c b/accel/tcg/user-exec-stub.c index 4fbe2dbdc8..162bb72bbe 100644 --- a/accel/tcg/user-exec-stub.c +++ b/accel/tcg/user-exec-stub.c @@ -18,6 +18,11 @@ void cpu_exec_reset_hold(CPUState *cpu) { } +void qemu_start_vcpu(CPUState *cpu) +{ + /* NOP for user-mode, we already have a thread */ +} + /* User mode emulation does not support record/replay yet. */ bool replay_exception(void) diff --git a/hw/core/cpu-common.c b/hw/core/cpu-common.c index 0f0a247f56..68895ddd59 100644 --- a/hw/core/cpu-common.c +++ b/hw/core/cpu-common.c @@ -230,7 +230,12 @@ static void cpu_common_realizefn(DeviceState *dev, Error **errp) } #endif - /* NOTE: latest generic point where the cpu is fully realized */ + /* + * With everything set up we can finally start the vCPU thread. + * This is a NOP for linux-user. + * NOTE: latest generic point where the cpu is fully realized + */ + qemu_start_vcpu(cpu); } static void cpu_common_unrealizefn(DeviceState *dev) diff --git a/plugins/core.c b/plugins/core.c index 0726bc7f25..1e5da7853b 100644 --- a/plugins/core.c +++ b/plugins/core.c @@ -65,6 +65,11 @@ static void plugin_cpu_update__locked(gpointer k, gpointer v, gpointer udata) CPUState *cpu = container_of(k, CPUState, cpu_index); run_on_cpu_data mask = RUN_ON_CPU_HOST_ULONG(*plugin.mask); + /* + * There is a race condition between the starting of the vCPU + * thread at the end of cpu_common_realizefn and when realized is + * finally set. + */ if (DEVICE(cpu)->realized) { async_run_on_cpu(cpu, plugin_cpu_update__async, mask); } else { diff --git a/system/cpus.c b/system/cpus.c index d3640c9503..7dd8464c5e 100644 --- a/system/cpus.c +++ b/system/cpus.c @@ -488,11 +488,13 @@ void cpus_kick_thread(CPUState *cpu) void qemu_cpu_kick(CPUState *cpu) { - qemu_cond_broadcast(cpu->halt_cond); - if (cpus_accel->kick_vcpu_thread) { - cpus_accel->kick_vcpu_thread(cpu); - } else { /* default */ - cpus_kick_thread(cpu); + if (cpu->halt_cond) { + qemu_cond_broadcast(cpu->halt_cond); + if (cpus_accel->kick_vcpu_thread) { + cpus_accel->kick_vcpu_thread(cpu); + } else { /* default */ + cpus_kick_thread(cpu); + } } } @@ -674,7 +676,10 @@ void qemu_init_vcpu(CPUState *cpu) cpu->num_ases = 1; cpu_address_space_init(cpu, 0, "cpu-memory", cpu->memory); } +} +void qemu_start_vcpu(CPUState *cpu) +{ /* accelerators all implement the AccelOpsClass */ g_assert(cpus_accel != NULL && cpus_accel->create_vcpu_thread != NULL); cpus_accel->create_vcpu_thread(cpu);