| lists.openwall.net | lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC | |
|
Open Source and information security mailing list archives
| ||
|
Message-ID: <20200116124626.1155-9-yezengruan@huawei.com>
Date: Thu, 16 Jan 2020 20:46:26 +0800
From: Zengruan Ye <yezengruan@...wei.com>
To: <linux-kernel@...r.kernel.org>,
<linux-arm-kernel@...ts.infradead.org>,
<kvmarm@...ts.cs.columbia.edu>, <kvm@...r.kernel.org>,
<linux-doc@...r.kernel.org>,
<virtualization@...ts.linux-foundation.org>
CC: <yezengruan@...wei.com>, <maz@...nel.org>, <james.morse@....com>,
<linux@...linux.org.uk>, <suzuki.poulose@....com>,
<julien.thierry.kdev@...il.com>, <catalin.marinas@....com>,
<mark.rutland@....com>, <will@...nel.org>, <steven.price@....com>,
<daniel.lezcano@...aro.org>, <wanghaibin.wang@...wei.com>,
<peterz@...radead.org>, <longman@...hat.com>
Subject: [PATCH v3 8/8] KVM: arm64: Support the vCPU preemption check
Support the vcpu_is_preempted() functionality under KVM/arm64. This will
enhance lock performance on overcommitted hosts (more runnable vCPUs
than physical CPUs in the system) as doing busy waits for preempted
vCPUs will hurt system performance far worse than early yielding.
unix benchmark result:
host: kernel 5.5.0-rc5, HiSilicon Kunpeng920, 8 CPUs
guest: kernel 5.5.0-rc5, 16 vCPUs
test-case | after-patch | before-patch
----------------------------------------+-------------------+------------------
Dhrystone 2 using register variables | 334600751.0 lps | 335319028.3 lps
Double-Precision Whetstone | 32856.1 MWIPS | 32849.6 MWIPS
Execl Throughput | 3662.1 lps | 2718.0 lps
File Copy 1024 bufsize 2000 maxblocks | 432906.4 KBps | 158011.8 KBps
File Copy 256 bufsize 500 maxblocks | 116023.0 KBps | 37664.0 KBps
File Copy 4096 bufsize 8000 maxblocks | 1432769.8 KBps | 441108.8 KBps
Pipe Throughput | 6405029.6 lps | 6021457.6 lps
Pipe-based Context Switching | 185872.7 lps | 184255.3 lps
Process Creation | 4025.7 lps | 3706.6 lps
Shell Scripts (1 concurrent) | 6745.6 lpm | 6436.1 lpm
Shell Scripts (8 concurrent) | 998.7 lpm | 931.1 lpm
System Call Overhead | 3913363.1 lps | 3883287.8 lps
----------------------------------------+-------------------+------------------
System Benchmarks Index Score | 1835.1 | 1327.6
Signed-off-by: Zengruan Ye <yezengruan@...wei.com>
---
arch/arm64/include/asm/paravirt.h | 3 +
arch/arm64/kernel/paravirt.c | 125 ++++++++++++++++++++++++++++++
arch/arm64/kernel/setup.c | 2 +
include/linux/cpuhotplug.h | 1 +
4 files changed, 131 insertions(+)
diff --git a/arch/arm64/include/asm/paravirt.h b/arch/arm64/include/asm/paravirt.h
index 7b1c81b544bb..ca3a2c7881f3 100644
--- a/arch/arm64/include/asm/paravirt.h
+++ b/arch/arm64/include/asm/paravirt.h
@@ -29,6 +29,8 @@ static inline u64 paravirt_steal_clock(int cpu)
int __init pv_time_init(void);
+int __init pv_lock_init(void);
+
__visible bool __native_vcpu_is_preempted(int cpu);
static inline bool pv_vcpu_is_preempted(int cpu)
@@ -39,6 +41,7 @@ static inline bool pv_vcpu_is_preempted(int cpu)
#else
#define pv_time_init() do {} while (0)
+#define pv_lock_init() do {} while (0)
#endif // CONFIG_PARAVIRT
diff --git a/arch/arm64/kernel/paravirt.c b/arch/arm64/kernel/paravirt.c
index d8f1ba8c22ce..39c6c35e59c6 100644
--- a/arch/arm64/kernel/paravirt.c
+++ b/arch/arm64/kernel/paravirt.c
@@ -22,6 +22,8 @@
#include <asm/paravirt.h>
#include <asm/pvclock-abi.h>
#include <asm/smp_plat.h>
+#include <asm/pvlock-abi.h>
+#include <asm/hypervisor.h>
struct static_key paravirt_steal_enabled;
struct static_key paravirt_steal_rq_enabled;
@@ -35,6 +37,10 @@ struct pv_time_stolen_time_region {
struct pvclock_vcpu_stolen_time *kaddr;
};
+struct pv_lock_state_region {
+ struct pvlock_vcpu_state *kaddr;
+};
+
static DEFINE_PER_CPU(struct pv_time_stolen_time_region, stolen_time_region);
static bool steal_acc = true;
@@ -158,3 +164,122 @@ int __init pv_time_init(void)
return 0;
}
+
+static DEFINE_PER_CPU(struct pv_lock_state_region, lock_state_region);
+
+static bool kvm_vcpu_is_preempted(int cpu)
+{
+ struct pv_lock_state_region *reg;
+ __le64 preempted_le;
+
+ reg = per_cpu_ptr(&lock_state_region, cpu);
+ if (!reg->kaddr) {
+ pr_warn_once("PV lock enabled but not configured for cpu %d\n",
+ cpu);
+ return false;
+ }
+
+ preempted_le = le64_to_cpu(READ_ONCE(reg->kaddr->preempted));
+
+ return !!preempted_le;
+}
+
+static int pvlock_vcpu_state_dying_cpu(unsigned int cpu)
+{
+ struct pv_lock_state_region *reg;
+
+ reg = this_cpu_ptr(&lock_state_region);
+ if (!reg->kaddr)
+ return 0;
+
+ memunmap(reg->kaddr);
+ memset(reg, 0, sizeof(*reg));
+
+ return 0;
+}
+
+static int init_pvlock_vcpu_state(unsigned int cpu)
+{
+ struct pv_lock_state_region *reg;
+ struct arm_smccc_res res;
+
+ reg = this_cpu_ptr(&lock_state_region);
+
+ arm_smccc_1_1_invoke(ARM_SMCCC_VENDOR_HYP_KVM_PV_LOCK_FUNC_ID,
+ KVM_PV_LOCK_PREEMPTED, &res);
+
+ if (res.a0 == SMCCC_RET_NOT_SUPPORTED) {
+ pr_warn("Failed to init PV lock data structure\n");
+ return -EINVAL;
+ }
+
+ reg->kaddr = memremap(res.a0,
+ sizeof(struct pvlock_vcpu_state),
+ MEMREMAP_WB);
+
+ if (!reg->kaddr) {
+ pr_warn("Failed to map PV lock data structure\n");
+ return -ENOMEM;
+ }
+
+ return 0;
+}
+
+static int kvm_arm_init_pvlock(void)
+{
+ int ret;
+
+ ret = cpuhp_setup_state(CPUHP_AP_ARM_KVM_PVLOCK_STARTING,
+ "hypervisor/arm/pvlock:starting",
+ init_pvlock_vcpu_state,
+ pvlock_vcpu_state_dying_cpu);
+ if (ret < 0) {
+ pr_warn("PV lock init failed\n");
+ return ret;
+ }
+
+ return 0;
+}
+
+static bool has_kvm_pvlock(void)
+{
+ struct arm_smccc_res res;
+
+ /* To detect the presence of PV lock support we require SMCCC 1.1+ */
+ if (psci_ops.smccc_version < SMCCC_VERSION_1_1)
+ return false;
+
+ if (!kvm_arm_hyp_service_available(
+ ARM_SMCCC_VENDOR_HYP_KVM_PV_LOCK_FUNC_ID)) {
+ pr_warn("PV lock is not available\n");
+ return false;
+ }
+
+ arm_smccc_1_1_invoke(ARM_SMCCC_VENDOR_HYP_KVM_PV_LOCK_FUNC_ID,
+ KVM_PV_LOCK_FEATURES, &res);
+
+ if (res.a0 != SMCCC_RET_SUCCESS)
+ return false;
+
+ return true;
+}
+
+int __init pv_lock_init(void)
+{
+ int ret;
+
+ if (is_hyp_mode_available())
+ return 0;
+
+ if (!has_kvm_pvlock())
+ return 0;
+
+ ret = kvm_arm_init_pvlock();
+ if (ret)
+ return ret;
+
+ pv_ops.lock.vcpu_is_preempted = kvm_vcpu_is_preempted;
+ pr_info("using PV lock preempted\n");
+
+ return 0;
+}
diff --git a/arch/arm64/kernel/setup.c b/arch/arm64/kernel/setup.c
index ad595c30ded0..b2a6c2a435ba 100644
--- a/arch/arm64/kernel/setup.c
+++ b/arch/arm64/kernel/setup.c
@@ -376,6 +376,8 @@ void __init setup_arch(char **cmdline_p)
smp_init_cpus();
smp_build_mpidr_hash();
+ pv_lock_init();
+
/* Init percpu seeds for random tags after cpus are set up. */
kasan_init_tags();
diff --git a/include/linux/cpuhotplug.h b/include/linux/cpuhotplug.h
index e51ee772b9f5..f72ff95ab63a 100644
--- a/include/linux/cpuhotplug.h
+++ b/include/linux/cpuhotplug.h
@@ -138,6 +138,7 @@ enum cpuhp_state {
CPUHP_AP_DUMMY_TIMER_STARTING,
CPUHP_AP_ARM_XEN_STARTING,
CPUHP_AP_ARM_KVMPV_STARTING,
+ CPUHP_AP_ARM_KVM_PVLOCK_STARTING,
CPUHP_AP_ARM_CORESIGHT_STARTING,
CPUHP_AP_ARM64_ISNDEP_STARTING,
CPUHP_AP_SMPCFD_DYING,
--
2.19.1
Powered by blists - more mailing lists