Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1753422AbbGGDex (ORCPT ); Mon, 6 Jul 2015 23:34:53 -0400 Received: from aserp1040.oracle.com ([141.146.126.69]:34171 "EHLO aserp1040.oracle.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751993AbbGGDeq (ORCPT ); Mon, 6 Jul 2015 23:34:46 -0400 From: Boris Ostrovsky To: david.vrabel@citrix.com, konrad.wilk@oracle.com Cc: roger.pau@citrix.com, elena.ufimtseva@oracle.com, stefano.stabellini@eu.citrix.com, tim@xen.org, jbeulich@suse.com, andrew.cooper3@citrix.com, ian.campbell@citrix.com, wei.liu2@citrix.com, ian.jackson@eu.citrix.com, xen-devel@lists.xenproject.org, linux-kernel@vger.kernel.org, boris.ostrovsky@oracle.com Subject: [PATCH 1/6] xen/x86/pvh: Save %rbx in xen_pvh_early_cpu_init() Date: Mon, 6 Jul 2015 23:34:20 -0400 Message-Id: <1436240065-10813-2-git-send-email-boris.ostrovsky@oracle.com> X-Mailer: git-send-email 1.8.1.4 In-Reply-To: <1436240065-10813-1-git-send-email-boris.ostrovsky@oracle.com> References: <1436240065-10813-1-git-send-email-boris.ostrovsky@oracle.com> X-Source-IP: aserv0021.oracle.com [141.146.126.233] Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Length: 3686 Lines: 108 x86-64 ABI requires that functions preserve %rbx. When xen_pvh_early_cpu_init() is executed on boot cpu it is invoked as a function and 'cpuid' instruction will clobber %rbx. (This is not a concern on secondary processors since there xen_pvh_early_cpu_init() is the entry point and thus does not need to preserve registers.) Since we cannot access stack on secondary processors (PTE's NX bit will cause #GP on AMD --- see commit a2ef5dc2c7cb ("x86/xen: Set EFER.NX and EFER.SCE in PVH guests")) we can't always save %rbx on the stack. We work around this by creating a new entry point for those processors --- xen_pvh_early_cpu_init_secondary(). Note that on secondary CPUs we will load %rbx with garbage, which is OK. As a side effect of these changes we don't need to save %rsi anymore, since we can use %rbx instead of %rsi as a temporary register. (We could store %rbx into another scratch register such as %r8 but since we will need new entry point for 32-bit PVH anyway we go with this approach for symmetry) Signed-off-by: Boris Ostrovsky --- arch/x86/xen/smp.c | 3 ++- arch/x86/xen/smp.h | 4 ++++ arch/x86/xen/xen-head.S | 14 +++++++------- 3 files changed, 13 insertions(+), 8 deletions(-) diff --git a/arch/x86/xen/smp.c b/arch/x86/xen/smp.c index 8648438..ca7ee1f 100644 --- a/arch/x86/xen/smp.c +++ b/arch/x86/xen/smp.c @@ -423,7 +423,8 @@ cpu_initialize_context(unsigned int cpu, struct task_struct *idle) * bit set. This means before DS/SS is touched, NX in * EFER must be set. Hence the following assembly glue code. */ - ctxt->user_regs.eip = (unsigned long)xen_pvh_early_cpu_init; + ctxt->user_regs.eip = + (unsigned long)xen_pvh_early_cpu_init_secondary; ctxt->user_regs.rdi = cpu; ctxt->user_regs.rsi = true; /* entry == true */ } diff --git a/arch/x86/xen/smp.h b/arch/x86/xen/smp.h index 963d62a..bf2b43c 100644 --- a/arch/x86/xen/smp.h +++ b/arch/x86/xen/smp.h @@ -10,10 +10,14 @@ extern void xen_send_IPI_self(int vector); #ifdef CONFIG_XEN_PVH extern void xen_pvh_early_cpu_init(int cpu, bool entry); +extern void xen_pvh_early_cpu_init_secondary(int cpu, bool entry); #else static inline void xen_pvh_early_cpu_init(int cpu, bool entry) { } +static inline void xen_pvh_early_cpu_init_secondary(int cpu, bool entry) +{ +} #endif #endif diff --git a/arch/x86/xen/xen-head.S b/arch/x86/xen/xen-head.S index 8afdfcc..b1508a8 100644 --- a/arch/x86/xen/xen-head.S +++ b/arch/x86/xen/xen-head.S @@ -56,28 +56,28 @@ ENTRY(startup_xen) * @entry: true if this is a secondary vcpu coming up on this entry * point, false if this is the boot CPU being initialized for * the first time (%rsi) - * - * Note: This is called as a function on the boot CPU, and is the entry point - * on the secondary CPU. */ ENTRY(xen_pvh_early_cpu_init) - mov %rsi, %r11 + mov %rbx, -8(%rsp) +/* Entry point for secondary CPUs */ +ENTRY(xen_pvh_early_cpu_init_secondary) /* Gather features to see if NX implemented. */ mov $0x80000001, %eax cpuid - mov %edx, %esi + mov %edx, %ebx mov $MSR_EFER, %ecx rdmsr bts $_EFER_SCE, %eax - bt $20, %esi + bt $20, %ebx jnc 1f /* No NX, skip setting it */ bts $_EFER_NX, %eax 1: wrmsr + mov -8(%rsp), %rbx #ifdef CONFIG_SMP - cmp $0, %r11b + cmp $0, %esi jne cpu_bringup_and_idle #endif ret -- 1.8.1.4 -- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/