qemu/system-cpus-Fix-pause_all_vcpus-under-concurrent-env.patch
Jiabo Feng c025e9059a QEMU update to version 8.2.0-6:
- coro: support live patch for libcare
- tests/acpi: Update expected ACPI tables for vcpu hotplug(update BinDir)
- arm/virt: Require mc->has_hotpluggable_cpus for cold-plugged vcpu
- arm/virt: Consider has_ged when set mc->has_hotpluggable_cpus
- arm/virt-acpi: Require possible_cpu_arch_ids for build_cpus_aml()
- acpi/ged: Remove cpuhp field of ged
- acpi/ged: Init cpu hotplug only when machine support it
- intc/gicv3: Fixes for vcpu hotplug
- arm/kvm: Set psci smccc filter only with vcpu hotplug
- accel/kvm: Use correct id for parked vcpu
- arm/virt: Fix adjudgement of core_id for vcpu hotplugged
- arm/virt.c: Convey local_err when set psci-conduit
- system/cpus: Fix resume_all_vcpus() under vCPU hotplug condition
- system/cpus: Fix pause_all_vcpus() under concurrent environment
- acpi/cpu: Fix cpu_hotplug_hw_init()
- arm/cpu: Some fixes for arm_cpu_unrealizefn()
- system/physmem: Fix possible double free when destroy cpu as
- hw/arm/virt: Expose cold-booted CPUs as MADT GICC Enabled
- tcg/mttcg: enable threads to unregister in tcg_ctxs[]
- hw/arm: Support hotplug capability check using _OSC method
- target/arm/kvm,tcg: Register/Handle SMCCC hypercall exits to VMM/Qemu
- target/arm/kvm: Write CPU state back to KVM on reset
- target/arm: Add support of *unrealize* ARMCPU during vCPU Hot-unplug
- physmem,gdbstub: Common helping funcs/changes to *unrealize* vCPU
- hw/arm: Changes required for reset and to support next boot
- arm/virt: Update the guest(via GED) about CPU hot-(un)plug events
- hw/intc/arm-gicv3*: Changes required to (re)init the vCPU register info
- hw/arm,gicv3: Changes to update GIC with vCPU hot-plug notification
- arm/virt: Changes to (un)wire GICC<->vCPU IRQs during hot-(un)plug
- arm/virt: Add/update basic hot-(un)plug framework
- hw/acpi: Update ACPI GED framework to support vCPU Hotplug
- arm/virt: Release objects for *disabled* possible vCPUs after init
- hw/acpi: Make _MAT method optional
- hw/arm: MADT Tbl change to size the guest with possible vCPUs
- hw/acpi: Update GED _EVT method AML with cpu scan
- hw/acpi: ACPI/AML Changes to reflect the correct _STA.{PRES,ENA} Bits to Guest
- arm/virt: Make ARM vCPU *present* status ACPI *persistent*
- arm/virt/acpi: Build CPUs AML with CPU Hotplug support
- tests/acpi/bios-tables-test: Allow changes to virt/DSDT file
- acpi/cpu: Add cpu_cppc building support
- arm/virt/acpi: Factor out CPPC building from DSDT CPU aml
- hw/acpi: Update CPUs AML with cpu-(ctrl)dev change
- arm/virt: Create GED dev before *disabled* CPU Objs are destroyed
- arm/virt: Add cpu hotplug events to GED during creation
- hw/acpi: Init GED framework with cpu hotplug events
- hw/acpi: Use qemu_present_cpu() API in ACPI CPU hotplug init
- hw/acpi: Add ACPI CPU hotplug init stub
- arm/acpi: Enable ACPI support for vcpu hotplug
- hw/acpi: Move CPU ctrl-dev MMIO region len macro to common header file
- arm/virt: Init PMU at host for all possible vcpus
- arm/virt,gicv3: Changes to pre-size GIC with possible vcpus @machine init
- arm/virt,kvm: Pre-create disabled possible vCPUs @machine init
- accel/kvm: Extract common KVM vCPU {creation,parking} code
- arm/virt,target/arm: Machine init time change common to vCPU {cold|hot}-plug
- hw/arm/virt: Move setting of common CPU properties in a function
- cpus-common: Add common CPU utility for possible vCPUs
- arm/virt,target/arm: Add new ARMCPU {socket,cluster,core,thread}-id property

Signed-off-by: Jiabo Feng <fengjiabo1@huawei.com>
2024-04-10 21:01:15 +08:00

92 lines
2.6 KiB
Diff

From 401e145800134d0310d613f48c4962a108b8ddda Mon Sep 17 00:00:00 2001
From: Keqian Zhu <zhukeqian1@huawei.com>
Date: Sun, 17 Mar 2024 16:37:03 +0800
Subject: [PATCH] system/cpus: Fix pause_all_vcpus() under concurrent
environment
Both main loop thread and vCPU thread are allowed to call
pause_all_vcpus(), and in general resume_all_vcpus() is called
after it. Two issues live in pause_all_vcpus():
1. There is possibility that during thread T1 waits on
qemu_pause_cond with bql unlocked, other thread has called
pause_all_vcpus() and resume_all_vcpus(), then thread T1 will
stuck, because the condition all_vcpus_paused() is always false.
2. After all_vcpus_paused() has been checked as true, we will
unlock bql to relock replay_mutex. During the bql was unlocked,
the vcpu's state may has been changed by other thread, so we
must retry.
Signed-off-by: Keqian Zhu <zhukeqian1@huawei.com>
---
system/cpus.c | 29 ++++++++++++++++++++++++-----
1 file changed, 24 insertions(+), 5 deletions(-)
diff --git a/system/cpus.c b/system/cpus.c
index a444a747f0..7c5369fa9c 100644
--- a/system/cpus.c
+++ b/system/cpus.c
@@ -551,12 +551,14 @@ static bool all_vcpus_paused(void)
return true;
}
-void pause_all_vcpus(void)
+static void request_pause_all_vcpus(void)
{
CPUState *cpu;
- qemu_clock_enable(QEMU_CLOCK_VIRTUAL, false);
CPU_FOREACH(cpu) {
+ if (cpu->stopped) {
+ continue;
+ }
if (qemu_cpu_is_self(cpu)) {
qemu_cpu_stop(cpu, true);
} else {
@@ -564,6 +566,14 @@ void pause_all_vcpus(void)
qemu_cpu_kick(cpu);
}
}
+}
+
+void pause_all_vcpus(void)
+{
+ qemu_clock_enable(QEMU_CLOCK_VIRTUAL, false);
+
+retry:
+ request_pause_all_vcpus();
/* We need to drop the replay_lock so any vCPU threads woken up
* can finish their replay tasks
@@ -572,14 +582,23 @@ void pause_all_vcpus(void)
while (!all_vcpus_paused()) {
qemu_cond_wait(&qemu_pause_cond, &qemu_global_mutex);
- CPU_FOREACH(cpu) {
- qemu_cpu_kick(cpu);
- }
+ /* During we waited on qemu_pause_cond the bql was unlocked,
+ * the vcpu's state may has been changed by other thread, so
+ * we must request the pause state on all vcpus again.
+ */
+ request_pause_all_vcpus();
}
qemu_mutex_unlock_iothread();
replay_mutex_lock();
qemu_mutex_lock_iothread();
+
+ /* During the bql was unlocked, the vcpu's state may has been
+ * changed by other thread, so we must retry.
+ */
+ if (!all_vcpus_paused()) {
+ goto retry;
+ }
}
void cpu_resume(CPUState *cpu)
--
2.27.0