From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from na01-bn1-obe.outbound.protection.outlook.com (mail-bn1on0145.outbound.protection.outlook.com [157.56.110.145]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-SHA384 (256/256 bits)) (No client certificate requested) by lists.ozlabs.org (Postfix) with ESMTPS id 900821A0B4A for ; Wed, 18 Feb 2015 20:33:08 +1100 (AEDT) From: Bogdan Purcareata To: , , , , Subject: [PATCH 2/2] powerpc/kvm: Limit MAX_VCPUS for guests running on RT Linux Date: Wed, 18 Feb 2015 09:32:35 +0000 Message-ID: <1424251955-308-3-git-send-email-bogdan.purcareata@freescale.com> In-Reply-To: <1424251955-308-1-git-send-email-bogdan.purcareata@freescale.com> References: <1424251955-308-1-git-send-email-bogdan.purcareata@freescale.com> MIME-Version: 1.0 Content-Type: text/plain Cc: scottwood@freescale.com, mihai.caraman@freescale.com, linux-kernel@vger.kernel.org, Bogdan Purcareata List-Id: Linux on PowerPC Developers Mail List List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Due to the introduction of the raw_spinlock for the KVM openpic, guests with a high number of VCPUs may induce great latencies on the underlying RT Linux system (e.g. cyclictest reports latencies of ~15ms for guests with 24 VCPUs). This can be further aggravated by sending a lot of external interrupts to the guest. A malicious app can abuse this scenario, causing a DoS of the host Linux. Until the KVM openpic code is refactored to use finer lock granularity, impose a limitation on the number of VCPUs a guest can have when running on a PREEMPT_RT_FULL system with KVM_MPIC emulation. Signed-off-by: Mihai Caraman Signed-off-by: Bogdan Purcareata Reviewed-by: Scott Wood --- arch/powerpc/include/asm/kvm_host.h | 6 ++++++ 1 file changed, 6 insertions(+) diff --git a/arch/powerpc/include/asm/kvm_host.h b/arch/powerpc/include/asm/kvm_host.h index 8ef0512..6f6b928 100644 --- a/arch/powerpc/include/asm/kvm_host.h +++ b/arch/powerpc/include/asm/kvm_host.h @@ -36,8 +36,14 @@ #include #include +#if defined(CONFIG_PREEMPT_RT_FULL) && defined(CONFIG_KVM_MPIC) +/* Limit the number of vcpus due to in-kernel mpic concurrency */ +#define KVM_MAX_VCPUS 4 +#define KVM_MAX_VCORES 4 +#else #define KVM_MAX_VCPUS NR_CPUS #define KVM_MAX_VCORES NR_CPUS +#endif #define KVM_USER_MEM_SLOTS 32 #define KVM_MEM_SLOTS_NUM KVM_USER_MEM_SLOTS -- 2.1.4