Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1752623AbdFLLyC (ORCPT ); Mon, 12 Jun 2017 07:54:02 -0400 Received: from mx2.suse.de ([195.135.220.15]:60396 "EHLO mx1.suse.de" rhost-flags-OK-OK-OK-FAIL) by vger.kernel.org with ESMTP id S1752374AbdFLLyA (ORCPT ); Mon, 12 Jun 2017 07:54:00 -0400 From: Juergen Gross To: linux-kernel@vger.kernel.org, xen-devel@lists.xenproject.org Cc: boris.ostrovsky@oracle.com, Juergen Gross Subject: [PATCH] xen: allocate page for shared info page from low memory Date: Mon, 12 Jun 2017 13:53:56 +0200 Message-Id: <20170612115356.8312-1-jgross@suse.com> X-Mailer: git-send-email 2.12.3 Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Length: 3327 Lines: 101 In a HVM guest the kernel allocates the page for mapping the shared info structure via extend_brk() today. This will lead to a drop of performance as the underlying EPT entry will have to be split up into 4kB entries as the single shared info page is located in hypervisor memory. The issue has been detected by using the libmicro munmap test: unmapping 8kB of memory was faster by nearly a factor of two when no pv interfaces were active in the HVM guest. So instead of taking a page from memory which might be mapped via large EPT entries use a page which is already mapped via a 4kB EPT entry: we can take a page from the first 1MB of memory as the video memory at 640kB disallows using larger EPT entries. Signed-off-by: Juergen Gross --- arch/x86/xen/enlighten_hvm.c | 31 ++++++++++++++++++++++++------- arch/x86/xen/enlighten_pv.c | 2 -- 2 files changed, 24 insertions(+), 9 deletions(-) diff --git a/arch/x86/xen/enlighten_hvm.c b/arch/x86/xen/enlighten_hvm.c index a6d014f47e52..c19477b6e43a 100644 --- a/arch/x86/xen/enlighten_hvm.c +++ b/arch/x86/xen/enlighten_hvm.c @@ -1,5 +1,6 @@ #include #include +#include #include #include @@ -10,9 +11,11 @@ #include #include #include +#include #include #include +#include #include "xen-ops.h" #include "mmu.h" @@ -22,20 +25,34 @@ void __ref xen_hvm_init_shared_info(void) { int cpu; struct xen_add_to_physmap xatp; - static struct shared_info *shared_info_page; + u64 pa; + + if (HYPERVISOR_shared_info == &xen_dummy_shared_info) { + /* + * Search for a free page starting at 4kB physical address. + * Low memory is preferred to avoid an EPT large page split up + * by the mapping. + * Starting below X86_RESERVE_LOW (usually 64kB) is fine as + * the BIOS used for HVM guests is well behaved and won't + * clobber memory other than the first 4kB. + */ + for (pa = PAGE_SIZE; + !e820__mapped_all(pa, pa + PAGE_SIZE, E820_TYPE_RAM) || + memblock_is_reserved(pa); + pa += PAGE_SIZE) + ; + + memblock_reserve(pa, PAGE_SIZE); + HYPERVISOR_shared_info = __va(pa); + } - if (!shared_info_page) - shared_info_page = (struct shared_info *) - extend_brk(PAGE_SIZE, PAGE_SIZE); xatp.domid = DOMID_SELF; xatp.idx = 0; xatp.space = XENMAPSPACE_shared_info; - xatp.gpfn = __pa(shared_info_page) >> PAGE_SHIFT; + xatp.gpfn = virt_to_pfn(HYPERVISOR_shared_info); if (HYPERVISOR_memory_op(XENMEM_add_to_physmap, &xatp)) BUG(); - HYPERVISOR_shared_info = (struct shared_info *)shared_info_page; - /* xen_vcpu is a pointer to the vcpu_info struct in the shared_info * page, we use it in the event channel upcall and in some pvclock * related functions. We don't need the vcpu_info placement diff --git a/arch/x86/xen/enlighten_pv.c b/arch/x86/xen/enlighten_pv.c index f33eef4ebd12..a9a67ecf2c07 100644 --- a/arch/x86/xen/enlighten_pv.c +++ b/arch/x86/xen/enlighten_pv.c @@ -89,8 +89,6 @@ void *xen_initial_gdt; -RESERVE_BRK(shared_info_page_brk, PAGE_SIZE); - static int xen_cpu_up_prepare_pv(unsigned int cpu); static int xen_cpu_dead_pv(unsigned int cpu); -- 2.12.3