From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1755025AbZIPH4Q (ORCPT ); Wed, 16 Sep 2009 03:56:16 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S1754665AbZIPH4P (ORCPT ); Wed, 16 Sep 2009 03:56:15 -0400 Received: from mx1.redhat.com ([209.132.183.28]:43869 "EHLO mx1.redhat.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1753959AbZIPH4O (ORCPT ); Wed, 16 Sep 2009 03:56:14 -0400 Date: Wed, 16 Sep 2009 03:56:17 -0400 (EDT) From: Miroslav Rezanina To: Jeremy Fitzhardinge Cc: linux-kernel@vger.kernel.org, xen-devel@lists.xensource.com Message-ID: <1250041003.152221253087777235.JavaMail.root@zmail06.collab.prod.int.phx2.redhat.com> In-Reply-To: <1582654680.152201253087647836.JavaMail.root@zmail06.collab.prod.int.phx2.redhat.com> Subject: Re: [PATCH][v2.6.29][XEN] Return unused memory to hypervisor MIME-Version: 1.0 Content-Type: text/plain; charset=utf-8 Content-Transfer-Encoding: 7bit X-Originating-IP: [10.5.5.71] Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org > From: "Jeremy Fitzhardinge" > To: "Miroslav Rezanina" > Cc: linux-kernel@vger.kernel.org, xen-devel@lists.xensource.com > Sent: Friday, September 4, 2009 1:26:24 AM GMT +01:00 Amsterdam / Berlin / Bern / Rome / Stockholm / Vienna > Subject: Re: [PATCH][v2.6.29][XEN] Return unused memory to hypervisor > > On 08/19/09 06:05, Miroslav Rezanina wrote: > > when running linux as XEN guest and use boot parameter mem= to set > memory lower then is assigned to guest, not used memory should be > returned to hypervisor as free. This is working with kernel available > on xen.org pages, but is not working with kernel 2.6.29. Comparing > both kernels I found code for returning unused memory to hypervisor is > missing. Following patch add this functionality to 2.6.29 kernel. > > > > Are you planning on submitting a revised patch along the lines I > suggested? > > Thanks, > J General version of patch. This version checks the e820 map for holes and returns all memory that is not mapped. Patch: =========== diff --git a/arch/x86/xen/enlighten.c b/arch/x86/xen/enlighten.c index b58e963..acc9166 100644 --- a/arch/x86/xen/enlighten.c +++ b/arch/x86/xen/enlighten.c @@ -31,6 +31,7 @@ #include #include #include +#include #include #include #include @@ -122,6 +123,59 @@ static int have_vcpu_info_placement = #endif ; +void __init xen_return_unused_memory(void) +{ + static struct e820map holes = { + .nr_map = 0 + }; + struct xen_memory_reservation reservation = { + .address_bits = 0, + .extent_order = 0, + .domid = DOMID_SELF + }; + unsigned long last_end = 0; + int i; + + for(i=0;i last_end) { + holes.map[holes.nr_map].addr = last_end; + holes.map[holes.nr_map].size = + e820.map[i].addr - last_end; + holes.nr_map++; + } + last_end = e820.map[i].addr + e820.map[i].size; + } + + if (last_end < PFN_PHYS((u64)xen_start_info->nr_pages)) { + holes.map[holes.nr_map].addr=last_end; + holes.map[holes.nr_map].size = + PFN_PHYS((u64)xen_start_info->nr_pages) - last_end; + holes.nr_map++; + } + + if (holes.nr_map == 0) + return; + + for(i=0;i> PAGE_SHIFT; + unsigned long epfn = ((holes.map[i].addr + holes.map[i].size) >> PAGE_SHIFT); + int ret; + + if (spfn % PAGE_SIZE != 0) + spfn++; + + if (spfn >= epfn) + continue; + + set_xen_guest_handle(reservation.extent_start, + ((unsigned long *)xen_start_info->mfn_list) + spfn); + + reservation.nr_extents = epfn - spfn; + ret = HYPERVISOR_memory_op(XENMEM_decrease_reservation, + &reservation); + BUG_ON (ret != epfn - spfn); + } +} static void xen_vcpu_setup(int cpu) { @@ -1057,6 +1111,8 @@ static __init void xen_post_allocator_init(void) SetPagePinned(virt_to_page(level3_user_vsyscall)); #endif xen_mark_init_mm_pinned(); + +/* xen_return_unused_memory(); */ } /* This is called once we have the cpu_possible_map */ -- Miroslav Rezanina Software Engineer - Virtualization Team - XEN kernel