From patchwork Wed Sep 6 19:01:39 2023 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Stefan Hajnoczi X-Patchwork-Id: 1830512 Return-Path: X-Original-To: incoming@patchwork.ozlabs.org Delivered-To: patchwork-incoming@legolas.ozlabs.org Authentication-Results: legolas.ozlabs.org; dkim=pass (1024-bit key; unprotected) header.d=redhat.com header.i=@redhat.com header.a=rsa-sha256 header.s=mimecast20190719 header.b=L9rW+8iY; dkim-atps=neutral Authentication-Results: legolas.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=patchwork.ozlabs.org) Received: from lists.gnu.org (lists.gnu.org [209.51.188.17]) (using TLSv1.2 with cipher ECDHE-ECDSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by legolas.ozlabs.org (Postfix) with ESMTPS id 4RgsFY478Nz1yh1 for ; Thu, 7 Sep 2023 05:02:41 +1000 (AEST) Received: from localhost ([::1] helo=lists1p.gnu.org) by lists.gnu.org with esmtp (Exim 4.90_1) (envelope-from ) id 1qdxmm-0006BB-2Q; Wed, 06 Sep 2023 15:01:56 -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 1qdxmh-00069x-Ax for qemu-devel@nongnu.org; Wed, 06 Sep 2023 15:01:51 -0400 Received: from us-smtp-delivery-124.mimecast.com ([170.10.133.124]) by eggs.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256) (Exim 4.90_1) (envelope-from ) id 1qdxme-00087M-JY for qemu-devel@nongnu.org; Wed, 06 Sep 2023 15:01:50 -0400 DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=redhat.com; s=mimecast20190719; t=1694026907; h=from:from:reply-to:subject:subject:date:date:message-id:message-id: to:to:cc:cc:mime-version:mime-version: content-transfer-encoding:content-transfer-encoding: in-reply-to:in-reply-to:references:references; bh=tUCOewx5Uk4Y3QQOUxCFtRBKpEhPDosHOt7NC6SB8DA=; b=L9rW+8iY6jTXxpUmXzeULf4psPz/+BmkvG2bLd3W/py8y/Zd4NKrCCDmRNbK2RZbNMLEx7 R0OvlUzM9vgY4kWKx7GbUG3lltNT487DJTnJAFfsa5mWUAWjlx8j7ydcOibfGZWQiK9iYQ +ZIfShPMa4FnNiEB0zEo6WKecZH88ho= Received: from mimecast-mx02.redhat.com (mimecast-mx02.redhat.com [66.187.233.88]) by relay.mimecast.com with ESMTP with STARTTLS (version=TLSv1.2, cipher=TLS_ECDHE_RSA_WITH_AES_256_GCM_SHA384) id us-mta-512-g3qDoFB3O2G6eete1x3mkg-1; Wed, 06 Sep 2023 15:01:46 -0400 X-MC-Unique: g3qDoFB3O2G6eete1x3mkg-1 Received: from smtp.corp.redhat.com (int-mx09.intmail.prod.int.rdu2.redhat.com [10.11.54.9]) (using TLSv1.2 with cipher AECDH-AES256-SHA (256/256 bits)) (No client certificate requested) by mimecast-mx02.redhat.com (Postfix) with ESMTPS id 8746D88DF61; Wed, 6 Sep 2023 19:01:45 +0000 (UTC) Received: from localhost (unknown [10.39.193.221]) by smtp.corp.redhat.com (Postfix) with ESMTP id 14B88493110; Wed, 6 Sep 2023 19:01:44 +0000 (UTC) From: Stefan Hajnoczi To: qemu-devel@nongnu.org Cc: "Dr. David Alan Gilbert" , Eduardo Habkost , pbonzini@redhat.com, Markus Armbruster , Eric Blake , kwolf@redhat.com, Maxim Levitsky , =?utf-8?q?Daniel_P?= =?utf-8?q?=2E_Berrang=C3=A9?= , Stefan Hajnoczi Subject: [RFC 1/3] hmp: avoid the nested event loop in handle_hmp_command() Date: Wed, 6 Sep 2023 15:01:39 -0400 Message-ID: <20230906190141.1286893-2-stefanha@redhat.com> In-Reply-To: <20230906190141.1286893-1-stefanha@redhat.com> References: <20230906190141.1286893-1-stefanha@redhat.com> MIME-Version: 1.0 X-Scanned-By: MIMEDefang 3.1 on 10.11.54.9 Received-SPF: pass client-ip=170.10.133.124; envelope-from=stefanha@redhat.com; helo=us-smtp-delivery-124.mimecast.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, DKIMWL_WL_HIGH=-0.001, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, DKIM_VALID_EF=-0.1, RCVD_IN_DNSWL_NONE=-0.0001, RCVD_IN_MSPIKE_H4=0.001, RCVD_IN_MSPIKE_WL=0.001, 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: 2.1.29 Precedence: list List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Errors-To: qemu-devel-bounces+incoming=patchwork.ozlabs.org@nongnu.org Sender: qemu-devel-bounces+incoming=patchwork.ozlabs.org@nongnu.org Coroutine HMP commands currently run to completion in a nested event loop with the Big QEMU Lock (BQL) held. The call_rcu thread also uses the BQL and cannot process work while the coroutine monitor command is running. A deadlock occurs when monitor commands attempt to wait for call_rcu work to finish. This patch refactors the HMP monitor to use the existing event loop instead of creating a nested event loop. This will allow the next patches to rely on draining call_rcu work. Signed-off-by: Stefan Hajnoczi --- monitor/hmp.c | 28 +++++++++++++++------------- 1 file changed, 15 insertions(+), 13 deletions(-) diff --git a/monitor/hmp.c b/monitor/hmp.c index 69c1b7e98a..6cff2810aa 100644 --- a/monitor/hmp.c +++ b/monitor/hmp.c @@ -1111,15 +1111,17 @@ typedef struct HandleHmpCommandCo { Monitor *mon; const HMPCommand *cmd; QDict *qdict; - bool done; } HandleHmpCommandCo; -static void handle_hmp_command_co(void *opaque) +static void coroutine_fn handle_hmp_command_co(void *opaque) { HandleHmpCommandCo *data = opaque; + handle_hmp_command_exec(data->mon, data->cmd, data->qdict); monitor_set_cur(qemu_coroutine_self(), NULL); - data->done = true; + qobject_unref(data->qdict); + monitor_resume(data->mon); + g_free(data); } void handle_hmp_command(MonitorHMP *mon, const char *cmdline) @@ -1157,20 +1159,20 @@ void handle_hmp_command(MonitorHMP *mon, const char *cmdline) Monitor *old_mon = monitor_set_cur(qemu_coroutine_self(), &mon->common); handle_hmp_command_exec(&mon->common, cmd, qdict); monitor_set_cur(qemu_coroutine_self(), old_mon); + qobject_unref(qdict); } else { - HandleHmpCommandCo data = { - .mon = &mon->common, - .cmd = cmd, - .qdict = qdict, - .done = false, - }; - Coroutine *co = qemu_coroutine_create(handle_hmp_command_co, &data); + HandleHmpCommandCo *data; /* freed by handle_hmp_command_co() */ + + data = g_new(HandleHmpCommandCo, 1); + data->mon = &mon->common; + data->cmd = cmd; + data->qdict = qdict; /* freed by handle_hmp_command_co() */ + + Coroutine *co = qemu_coroutine_create(handle_hmp_command_co, data); + monitor_suspend(&mon->common); /* resumed by handle_hmp_command_co() */ monitor_set_cur(co, &mon->common); aio_co_enter(qemu_get_aio_context(), co); - AIO_WAIT_WHILE_UNLOCKED(NULL, !data.done); } - - qobject_unref(qdict); } static void cmd_completion(MonitorHMP *mon, const char *name, const char *list) From patchwork Wed Sep 6 19:01:40 2023 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Stefan Hajnoczi X-Patchwork-Id: 1830513 Return-Path: X-Original-To: incoming@patchwork.ozlabs.org Delivered-To: patchwork-incoming@legolas.ozlabs.org Authentication-Results: legolas.ozlabs.org; dkim=pass (1024-bit key; unprotected) header.d=redhat.com header.i=@redhat.com header.a=rsa-sha256 header.s=mimecast20190719 header.b=UR/YYC4M; dkim-atps=neutral Authentication-Results: legolas.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=patchwork.ozlabs.org) Received: from lists.gnu.org (lists.gnu.org [209.51.188.17]) (using TLSv1.2 with cipher ECDHE-ECDSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by legolas.ozlabs.org (Postfix) with ESMTPS id 4RgsG94Dfnz1yh1 for ; Thu, 7 Sep 2023 05:03:13 +1000 (AEST) Received: from localhost ([::1] helo=lists1p.gnu.org) by lists.gnu.org with esmtp (Exim 4.90_1) (envelope-from ) id 1qdxmq-0006Cu-EW; Wed, 06 Sep 2023 15:02:00 -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 1qdxmm-0006BY-PR for qemu-devel@nongnu.org; Wed, 06 Sep 2023 15:01:56 -0400 Received: from us-smtp-delivery-124.mimecast.com ([170.10.133.124]) by eggs.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256) (Exim 4.90_1) (envelope-from ) id 1qdxmj-0008A0-AV for qemu-devel@nongnu.org; Wed, 06 Sep 2023 15:01:55 -0400 DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=redhat.com; s=mimecast20190719; t=1694026912; h=from:from:reply-to:subject:subject:date:date:message-id:message-id: to:to:cc:cc:mime-version:mime-version: content-transfer-encoding:content-transfer-encoding: in-reply-to:in-reply-to:references:references; bh=DCK+lnfLAYtw1mOtOApL/BAriyzaovAFIXtqHAPBx7E=; b=UR/YYC4MefkayGtPdGINbBaadX1I/DGoWbk6x7FoEAdKxdgq7frK+M9RjAIeB1zne3i6D9 ArTUv0v7TGecTvMHyCraDz2KSkeZux3+7reA7+jVxvuY18ECAxtyYFzqDwfGfnVBDV77qn dtZlgE6s5FKj5c1A9dF9mH7BkL3/RKY= Received: from mimecast-mx02.redhat.com (mx-ext.redhat.com [66.187.233.73]) by relay.mimecast.com with ESMTP with STARTTLS (version=TLSv1.2, cipher=TLS_ECDHE_RSA_WITH_AES_256_GCM_SHA384) id us-mta-556-GEOy8nXNO9aTyOXA76zMLA-1; Wed, 06 Sep 2023 15:01:49 -0400 X-MC-Unique: GEOy8nXNO9aTyOXA76zMLA-1 Received: from smtp.corp.redhat.com (int-mx07.intmail.prod.int.rdu2.redhat.com [10.11.54.7]) (using TLSv1.2 with cipher AECDH-AES256-SHA (256/256 bits)) (No client certificate requested) by mimecast-mx02.redhat.com (Postfix) with ESMTPS id 611B61C0754F; Wed, 6 Sep 2023 19:01:47 +0000 (UTC) Received: from localhost (unknown [10.39.193.221]) by smtp.corp.redhat.com (Postfix) with ESMTP id E1DA2140E964; Wed, 6 Sep 2023 19:01:46 +0000 (UTC) From: Stefan Hajnoczi To: qemu-devel@nongnu.org Cc: "Dr. David Alan Gilbert" , Eduardo Habkost , pbonzini@redhat.com, Markus Armbruster , Eric Blake , kwolf@redhat.com, Maxim Levitsky , =?utf-8?q?Daniel_P?= =?utf-8?q?=2E_Berrang=C3=A9?= , Stefan Hajnoczi Subject: [RFC 2/3] rcu: add drain_call_rcu_co() API Date: Wed, 6 Sep 2023 15:01:40 -0400 Message-ID: <20230906190141.1286893-3-stefanha@redhat.com> In-Reply-To: <20230906190141.1286893-1-stefanha@redhat.com> References: <20230906190141.1286893-1-stefanha@redhat.com> MIME-Version: 1.0 X-Scanned-By: MIMEDefang 3.1 on 10.11.54.7 Received-SPF: pass client-ip=170.10.133.124; envelope-from=stefanha@redhat.com; helo=us-smtp-delivery-124.mimecast.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, DKIMWL_WL_HIGH=-0.001, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, DKIM_VALID_EF=-0.1, RCVD_IN_DNSWL_NONE=-0.0001, RCVD_IN_MSPIKE_H4=0.001, RCVD_IN_MSPIKE_WL=0.001, 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: 2.1.29 Precedence: list List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Errors-To: qemu-devel-bounces+incoming=patchwork.ozlabs.org@nongnu.org Sender: qemu-devel-bounces+incoming=patchwork.ozlabs.org@nongnu.org call_drain_rcu() has limitations that make it unsuitable for use in qmp_device_add(). Introduce a new coroutine version of drain_call_rcu() with the same functionality but that does not drop the BQL. The next patch will use it to fix qmp_device_add(). Signed-off-by: Stefan Hajnoczi --- MAINTAINERS | 2 ++ docs/devel/rcu.txt | 21 +++++++++++++++++ include/qemu/rcu.h | 1 + util/rcu-internal.h | 8 +++++++ util/rcu-co.c | 55 +++++++++++++++++++++++++++++++++++++++++++++ util/rcu.c | 3 ++- util/meson.build | 2 +- 7 files changed, 90 insertions(+), 2 deletions(-) create mode 100644 util/rcu-internal.h create mode 100644 util/rcu-co.c diff --git a/MAINTAINERS b/MAINTAINERS index 3b29568ed4..7f98253bda 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -2908,6 +2908,8 @@ F: include/qemu/rcu*.h F: tests/unit/rcutorture.c F: tests/unit/test-rcu-*.c F: util/rcu.c +F: util/rcu-co.c +F: util/rcu-internal.h Human Monitor (HMP) M: Dr. David Alan Gilbert diff --git a/docs/devel/rcu.txt b/docs/devel/rcu.txt index 2e6cc607a1..344764527f 100644 --- a/docs/devel/rcu.txt +++ b/docs/devel/rcu.txt @@ -130,6 +130,27 @@ The core RCU API is small: g_free_rcu(&foo, rcu); + void coroutine_fn drain_call_rcu_co(void); + + drain_call_rcu_co() yields until the reclamation phase is finished. + Reclaimer functions previously submitted with call_rcu1() in this + thread will have finished by the time drain_call_rcu_co() returns. + + void drain_call_rcu(void); + + drain_call_rcu() releases the Big QEMU Lock (BQL), if held, waits until + the reclamation phase is finished, and then re-acquires the BQL, if + previously held. Reclaimer functions previously submitted with + call_rcu1() in this thread will have finished by the time + drain_call_rcu() returns. + + drain_call_rcu() has the following limitations: + 1. It deadlocks when called within an RCU read-side critical section. + 2. All functions on the call stack must be designed to handle dropping + the BQL. + + Prefer drain_call_rcu_co() over drain_call_rcu(). + typeof(*p) qatomic_rcu_read(p); qatomic_rcu_read() is similar to qatomic_load_acquire(), but it makes diff --git a/include/qemu/rcu.h b/include/qemu/rcu.h index fea058aa9f..53055df1dc 100644 --- a/include/qemu/rcu.h +++ b/include/qemu/rcu.h @@ -141,6 +141,7 @@ struct rcu_head { }; void call_rcu1(struct rcu_head *head, RCUCBFunc *func); +void coroutine_fn drain_call_rcu_co(void); void drain_call_rcu(void); /* The operands of the minus operator must have the same type, diff --git a/util/rcu-internal.h b/util/rcu-internal.h new file mode 100644 index 0000000000..7d85366d54 --- /dev/null +++ b/util/rcu-internal.h @@ -0,0 +1,8 @@ +/* SPDX-License-Identifier: LGPL-2.1-or-later */ + +#ifndef RCU_INTERNAL_H +#define RCU_INTERNAL_H + +extern int in_drain_call_rcu; + +#endif /* RCU_INTERNAL_H */ diff --git a/util/rcu-co.c b/util/rcu-co.c new file mode 100644 index 0000000000..920fcacb7a --- /dev/null +++ b/util/rcu-co.c @@ -0,0 +1,55 @@ +/* SPDX-License-Identifier: LGPL-2.1-or-later */ +/* + * RCU APIs for coroutines + * + * The RCU coroutine APIs are kept separate from the main RCU code to avoid + * depending on AioContext APIs in rcu.c. This is necessary because at least + * tests/unit/ptimer-test.c has replacement functions for AioContext APIs that + * conflict with the real functions. + * + * It's also nice to logically separate the core RCU code from the coroutine + * APIs :). + */ +#include "qemu/osdep.h" +#include "block/aio.h" +#include "qemu/atomic.h" +#include "qemu/coroutine.h" +#include "qemu/rcu.h" +#include "rcu-internal.h" + +typedef struct { + struct rcu_head rcu; + Coroutine *co; +} RcuDrainCo; + +static void drain_call_rcu_co_bh(void *opaque) +{ + RcuDrainCo *data = opaque; + + /* Re-enter drain_call_rcu_co() where it yielded */ + aio_co_wake(data->co); +} + +static void drain_call_rcu_co_cb(struct rcu_head *node) +{ + RcuDrainCo *data = container_of(node, RcuDrainCo, rcu); + AioContext *ctx = qemu_coroutine_get_aio_context(data->co); + + /* + * drain_call_rcu_co() might still be running in its thread, so schedule a + * BH in its thread. The BH only runs after the coroutine has yielded. + */ + aio_bh_schedule_oneshot(ctx, drain_call_rcu_co_bh, data); +} + +void coroutine_fn drain_call_rcu_co(void) +{ + RcuDrainCo data = { + .co = qemu_coroutine_self(), + }; + + qatomic_inc(&in_drain_call_rcu); + call_rcu1(&data.rcu, drain_call_rcu_co_cb); + qemu_coroutine_yield(); /* wait for drain_rcu_co_bh() */ + qatomic_dec(&in_drain_call_rcu); +} diff --git a/util/rcu.c b/util/rcu.c index e587bcc483..2519bd7d5c 100644 --- a/util/rcu.c +++ b/util/rcu.c @@ -32,6 +32,7 @@ #include "qemu/thread.h" #include "qemu/main-loop.h" #include "qemu/lockable.h" +#include "rcu-internal.h" #if defined(CONFIG_MALLOC_TRIM) #include #endif @@ -46,7 +47,7 @@ unsigned long rcu_gp_ctr = RCU_GP_LOCKED; QemuEvent rcu_gp_event; -static int in_drain_call_rcu; +int in_drain_call_rcu; static QemuMutex rcu_registry_lock; static QemuMutex rcu_sync_lock; diff --git a/util/meson.build b/util/meson.build index a375160286..849d56f756 100644 --- a/util/meson.build +++ b/util/meson.build @@ -43,7 +43,7 @@ util_ss.add(files('keyval.c')) util_ss.add(files('crc32c.c')) util_ss.add(files('uuid.c')) util_ss.add(files('getauxval.c')) -util_ss.add(files('rcu.c')) +util_ss.add(files('rcu.c', 'rcu-co.c')) if have_membarrier util_ss.add(files('sys_membarrier.c')) endif From patchwork Wed Sep 6 19:01:41 2023 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Stefan Hajnoczi X-Patchwork-Id: 1830511 Return-Path: X-Original-To: incoming@patchwork.ozlabs.org Delivered-To: patchwork-incoming@legolas.ozlabs.org Authentication-Results: legolas.ozlabs.org; dkim=pass (1024-bit key; unprotected) header.d=redhat.com header.i=@redhat.com header.a=rsa-sha256 header.s=mimecast20190719 header.b=ax5EJdDe; dkim-atps=neutral Authentication-Results: legolas.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=patchwork.ozlabs.org) Received: from lists.gnu.org (lists.gnu.org [209.51.188.17]) (using TLSv1.2 with cipher ECDHE-ECDSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by legolas.ozlabs.org (Postfix) with ESMTPS id 4RgsFW51Tmz1yh1 for ; Thu, 7 Sep 2023 05:02:39 +1000 (AEST) Received: from localhost ([::1] helo=lists1p.gnu.org) by lists.gnu.org with esmtp (Exim 4.90_1) (envelope-from ) id 1qdxmo-0006C1-S1; Wed, 06 Sep 2023 15:01:59 -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 1qdxmm-0006BM-5A for qemu-devel@nongnu.org; Wed, 06 Sep 2023 15:01:56 -0400 Received: from us-smtp-delivery-124.mimecast.com ([170.10.133.124]) by eggs.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256) (Exim 4.90_1) (envelope-from ) id 1qdxmj-0008A4-Gf for qemu-devel@nongnu.org; Wed, 06 Sep 2023 15:01:55 -0400 DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=redhat.com; s=mimecast20190719; t=1694026913; h=from:from:reply-to:subject:subject:date:date:message-id:message-id: to:to:cc:cc:mime-version:mime-version: content-transfer-encoding:content-transfer-encoding: in-reply-to:in-reply-to:references:references; bh=/5Q+q6yNqAxBk6g6i9jyTzPE/hWWcqQrxs+KLHIhU8E=; b=ax5EJdDe1MPsmkCD7YpiXYnFm3vVOQIA0nWgFGCGy149f6U2vVNte6LYFd86hSaBX+gSBA gpjwlFOm7dxfRetqiJJt8HxRQ9ZECHk6rkasXmP1FngqmsJm7Q4GDxG3hCdCB0TtrdWNlt 2gcWik6rn13rgK59GPimlwgrBdMcXks= Received: from mimecast-mx02.redhat.com (mx-ext.redhat.com [66.187.233.73]) by relay.mimecast.com with ESMTP with STARTTLS (version=TLSv1.2, cipher=TLS_ECDHE_RSA_WITH_AES_256_GCM_SHA384) id us-mta-594-iwzTqUN_NG2GJF4aKSkC1g-1; Wed, 06 Sep 2023 15:01:49 -0400 X-MC-Unique: iwzTqUN_NG2GJF4aKSkC1g-1 Received: from smtp.corp.redhat.com (int-mx06.intmail.prod.int.rdu2.redhat.com [10.11.54.6]) (using TLSv1.2 with cipher AECDH-AES256-SHA (256/256 bits)) (No client certificate requested) by mimecast-mx02.redhat.com (Postfix) with ESMTPS id 3296A1C07557; Wed, 6 Sep 2023 19:01:49 +0000 (UTC) Received: from localhost (unknown [10.39.193.221]) by smtp.corp.redhat.com (Postfix) with ESMTP id AF352202869C; Wed, 6 Sep 2023 19:01:48 +0000 (UTC) From: Stefan Hajnoczi To: qemu-devel@nongnu.org Cc: "Dr. David Alan Gilbert" , Eduardo Habkost , pbonzini@redhat.com, Markus Armbruster , Eric Blake , kwolf@redhat.com, Maxim Levitsky , =?utf-8?q?Daniel_P?= =?utf-8?q?=2E_Berrang=C3=A9?= , Stefan Hajnoczi Subject: [RFC 3/3] qmp: make qmp_device_add() a coroutine Date: Wed, 6 Sep 2023 15:01:41 -0400 Message-ID: <20230906190141.1286893-4-stefanha@redhat.com> In-Reply-To: <20230906190141.1286893-1-stefanha@redhat.com> References: <20230906190141.1286893-1-stefanha@redhat.com> MIME-Version: 1.0 X-Scanned-By: MIMEDefang 3.1 on 10.11.54.6 Received-SPF: pass client-ip=170.10.133.124; envelope-from=stefanha@redhat.com; helo=us-smtp-delivery-124.mimecast.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, DKIMWL_WL_HIGH=-0.001, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, DKIM_VALID_EF=-0.1, RCVD_IN_DNSWL_NONE=-0.0001, RCVD_IN_MSPIKE_H4=0.001, RCVD_IN_MSPIKE_WL=0.001, 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: 2.1.29 Precedence: list List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Errors-To: qemu-devel-bounces+incoming=patchwork.ozlabs.org@nongnu.org Sender: qemu-devel-bounces+incoming=patchwork.ozlabs.org@nongnu.org It is not safe to call drain_call_rcu() from qmp_device_add() because some call stacks are not prepared for drain_call_rcu() to drop the Big QEMU Lock (BQL). For example, device emulation code is protected by the BQL but when it calls aio_poll() -> ... -> qmp_device_add() -> drain_call_rcu() then the BQL is dropped. See bz#2215192 below for a concrete bug of this type. Another limitation of drain_call_rcu() is that it cannot be invoked within an RCU read-side critical section since the reclamation phase cannot complete until the end of the critical section. Unfortunately, call stacks have been seen where this happens (see bz#2214985 below). Switch to call_drain_rcu_co() to avoid these problems. This requires making qmp_device_add() a coroutine. qdev_device_add() is not designed to be called from coroutines, so it must be invoked from a BH and then switch back to the coroutine. Fixes: 7bed89958bfbf40df9ca681cefbdca63abdde39d ("device_core: use drain_call_rcu in in qmp_device_add") Buglink: https://bugzilla.redhat.com/show_bug.cgi?id=2215192 Buglink: https://bugzilla.redhat.com/show_bug.cgi?id=2214985 Signed-off-by: Stefan Hajnoczi --- qapi/qdev.json | 1 + include/monitor/qdev.h | 3 ++- monitor/qmp-cmds.c | 2 +- softmmu/qdev-monitor.c | 34 ++++++++++++++++++++++++++++++---- hmp-commands.hx | 1 + 5 files changed, 35 insertions(+), 6 deletions(-) diff --git a/qapi/qdev.json b/qapi/qdev.json index 6bc5a733b8..78e9d7f7b8 100644 --- a/qapi/qdev.json +++ b/qapi/qdev.json @@ -79,6 +79,7 @@ ## { 'command': 'device_add', 'data': {'driver': 'str', '*bus': 'str', '*id': 'str'}, + 'coroutine': true, 'gen': false, # so we can get the additional arguments 'features': ['json-cli', 'json-cli-hotplug'] } diff --git a/include/monitor/qdev.h b/include/monitor/qdev.h index 1d57bf6577..1fed9eb9ea 100644 --- a/include/monitor/qdev.h +++ b/include/monitor/qdev.h @@ -5,7 +5,8 @@ void hmp_info_qtree(Monitor *mon, const QDict *qdict); void hmp_info_qdm(Monitor *mon, const QDict *qdict); -void qmp_device_add(QDict *qdict, QObject **ret_data, Error **errp); +void coroutine_fn +qmp_device_add(QDict *qdict, QObject **ret_data, Error **errp); int qdev_device_help(QemuOpts *opts); DeviceState *qdev_device_add(QemuOpts *opts, Error **errp); diff --git a/monitor/qmp-cmds.c b/monitor/qmp-cmds.c index b0f948d337..a7419226fe 100644 --- a/monitor/qmp-cmds.c +++ b/monitor/qmp-cmds.c @@ -202,7 +202,7 @@ static void __attribute__((__constructor__)) monitor_init_qmp_commands(void) qmp_init_marshal(&qmp_commands); qmp_register_command(&qmp_commands, "device_add", - qmp_device_add, 0, 0); + qmp_device_add, QCO_COROUTINE, 0); QTAILQ_INIT(&qmp_cap_negotiation_commands); qmp_register_command(&qmp_cap_negotiation_commands, "qmp_capabilities", diff --git a/softmmu/qdev-monitor.c b/softmmu/qdev-monitor.c index 74f4e41338..85ae62f7cf 100644 --- a/softmmu/qdev-monitor.c +++ b/softmmu/qdev-monitor.c @@ -839,8 +839,28 @@ void hmp_info_qdm(Monitor *mon, const QDict *qdict) qdev_print_devinfos(true); } -void qmp_device_add(QDict *qdict, QObject **ret_data, Error **errp) +typedef struct { + Coroutine *co; + QemuOpts *opts; + Error **errp; + DeviceState *dev; +} QmpDeviceAdd; + +static void qmp_device_add_bh(void *opaque) { + QmpDeviceAdd *data = opaque; + + data->dev = qdev_device_add(data->opts, data->errp); + aio_co_wake(data->co); +} + +void coroutine_fn +qmp_device_add(QDict *qdict, QObject **ret_data, Error **errp) +{ + QmpDeviceAdd data = { + .co = qemu_coroutine_self(), + .errp = errp, + }; QemuOpts *opts; DeviceState *dev; @@ -852,7 +872,13 @@ void qmp_device_add(QDict *qdict, QObject **ret_data, Error **errp) qemu_opts_del(opts); return; } - dev = qdev_device_add(opts, errp); + + /* Perform qdev_device_add() call outside coroutine context */ + data.opts = opts; + aio_bh_schedule_oneshot(qemu_coroutine_get_aio_context(data.co), + qmp_device_add_bh, &data); + qemu_coroutine_yield(); + dev = data.dev; /* * Drain all pending RCU callbacks. This is done because @@ -863,7 +889,7 @@ void qmp_device_add(QDict *qdict, QObject **ret_data, Error **errp) * will finish its job completely once qmp command returns result * to the user */ - drain_call_rcu(); + drain_call_rcu_co(); if (!dev) { qemu_opts_del(opts); @@ -956,7 +982,7 @@ void qmp_device_del(const char *id, Error **errp) } } -void hmp_device_add(Monitor *mon, const QDict *qdict) +void coroutine_fn hmp_device_add(Monitor *mon, const QDict *qdict) { Error *err = NULL; diff --git a/hmp-commands.hx b/hmp-commands.hx index 2cbd0f77a0..c737d1fd64 100644 --- a/hmp-commands.hx +++ b/hmp-commands.hx @@ -695,6 +695,7 @@ ERST .params = "driver[,prop=value][,...]", .help = "add device, like -device on the command line", .cmd = hmp_device_add, + .coroutine = true, .command_completion = device_add_completion, },