From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org X-Spam-Level: X-Spam-Status: No, score=-7.0 required=3.0 tests=HEADER_FROM_DIFFERENT_DOMAINS, INCLUDES_PATCH,MAILING_LIST_MULTI,SIGNED_OFF_BY,SPF_HELO_NONE,SPF_PASS autolearn=ham autolearn_force=no version=3.4.0 Received: from mail.kernel.org (mail.kernel.org [198.145.29.99]) by smtp.lore.kernel.org (Postfix) with ESMTP id D088FC04AAF for ; Thu, 16 May 2019 12:00:00 +0000 (UTC) Received: from vger.kernel.org (vger.kernel.org [209.132.180.67]) by mail.kernel.org (Postfix) with ESMTP id 9AEEB20818 for ; Thu, 16 May 2019 12:00:00 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1727492AbfEPL77 (ORCPT ); Thu, 16 May 2019 07:59:59 -0400 Received: from mx1.redhat.com ([209.132.183.28]:42668 "EHLO mx1.redhat.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1727281AbfEPL77 (ORCPT ); Thu, 16 May 2019 07:59:59 -0400 Received: from smtp.corp.redhat.com (int-mx03.intmail.prod.int.phx2.redhat.com [10.5.11.13]) (using TLSv1.2 with cipher AECDH-AES256-SHA (256/256 bits)) (No client certificate requested) by mx1.redhat.com (Postfix) with ESMTPS id 61F213082135; Thu, 16 May 2019 11:59:58 +0000 (UTC) Received: from thuth.remote.csb (ovpn-116-68.ams2.redhat.com [10.36.116.68]) by smtp.corp.redhat.com (Postfix) with ESMTP id 389AE60619; Thu, 16 May 2019 11:59:51 +0000 (UTC) Subject: Re: [RFC PATCH 2/4] KVM: selftests: Align memory region addresses to 1M on s390x To: David Hildenbrand , Christian Borntraeger , Janosch Frank , kvm@vger.kernel.org Cc: Paolo Bonzini , =?UTF-8?B?UmFkaW0gS3LEjW3DocWZ?= , Shuah Khan , Cornelia Huck , Andrew Jones , linux-kernel@vger.kernel.org, linux-kselftest@vger.kernel.org, linux-s390@vger.kernel.org References: <20190516111253.4494-1-thuth@redhat.com> <20190516111253.4494-3-thuth@redhat.com> From: Thomas Huth Openpgp: preference=signencrypt Autocrypt: addr=thuth@redhat.com; keydata= xsFNBFH7eUwBEACzyOXKU+5Pcs6wNpKzrlJwzRl3VGZt95VCdb+FgoU9g11m7FWcOafrVRwU yYkTm9+7zBUc0sW5AuPGR/dp3pSLX/yFWsA/UB4nJsHqgDvDU7BImSeiTrnpMOTXb7Arw2a2 4CflIyFqjCpfDM4MuTmzTjXq4Uov1giGE9X6viNo1pxyEpd7PanlKNnf4PqEQp06X4IgUacW tSGj6Gcns1bCuHV8OPWLkf4hkRnu8hdL6i60Yxz4E6TqlrpxsfYwLXgEeswPHOA6Mn4Cso9O 0lewVYfFfsmokfAVMKWzOl1Sr0KGI5T9CpmRfAiSHpthhHWnECcJFwl72NTi6kUcUzG4se81 O6n9d/kTj7pzTmBdfwuOZ0YUSqcqs0W+l1NcASSYZQaDoD3/SLk+nqVeCBB4OnYOGhgmIHNW 0CwMRO/GK+20alxzk//V9GmIM2ACElbfF8+Uug3pqiHkVnKqM7W9/S1NH2qmxB6zMiJUHlTH gnVeZX0dgH27mzstcF786uPcdEqS0KJuxh2kk5IvUSL3Qn3ZgmgdxBMyCPciD/1cb7/Ahazr 3ThHQXSHXkH/aDXdfLsKVuwDzHLVSkdSnZdt5HHh75/NFHxwaTlydgfHmFFwodK8y/TjyiGZ zg2Kje38xnz8zKn9iesFBCcONXS7txENTzX0z80WKBhK+XSFJwARAQABzRxUaG9tYXMgSHV0 aCA8dGguaHV0aEBnbXguZGU+wsF7BBMBAgAlAhsDBgsJCAcDAgYVCAIJCgsEFgIDAQIeAQIX gAUCUfuWKwIZAQAKCRAu2dd0/nAttbe/EACb9hafyOb2FmhUqeAiBORSsUifFacQ7laVjcgR I4um8CSHvxijYftpkM2EdAtmXIKgbNDpQoXcWLXB9lu9mLgTO4DVT00TRR65ikn3FCWcyT74 ENTOzRKyKLsDCjhXKPblTPIQbYAUCOWElcyAPm0ERd62fA/rKNxgIiNo/l4UODOMoOJm2/Ox ZoTckW68Eqv7k9L7m7j+Hn3hoDTjAmcCBJt+j7pOhzWvCbqoNOIH8C8qvPaNlrba+R/K6jkO 6jZkTbYQpGIofEQJ/TNn38IsNGpI1ALTHWFtoMxp3j2Imz0REO6dRE2fHRN8sVlHgkoeGhmY NbDsDE1jFQOEObFnu0euk//7BXU7tGOHckVAZ8T1smiRPHfQU7UEH2a/grndxJ+PNeM5w7n2 l+FN3cf2KgPotCK2s9MjSdZA7C5e3rFYO8lqiqTJKvc62vqp3e7B0Kjyy5/QtzSOejBij2QL xkKSFNtxIz4MtuxN8e3IDQNxsKry3nF7R4MDvouXlMo6wP9KuyNWb+vFJt9GtbgfDMIFVamp ZfhEWzWRJH4VgksENA4K/BzjEHCcbTUb1TFsiB1VRnBPJ0SqlvifnfKk6HcpkDk6Pg8Q5FOJ gbNHrdgXsm+m/9GF2zUUr+rOlhVbK23TUqKqPfwnD7uxjpakVcJnsVCFqJpZi1F/ga9IN87B TQRR+3lMARAAtp831HniPHb9AuKq3wj83ujZK8lH5RLrfVsB4X1wi47bwo56BqhXpR/zxPTR eOFT0gnbw9UkphVc7uk/alnXMDEmgvnuxv89PwIQX6k3qLABeV7ykJQG/WT5HQ6+2DdGtVw3 2vjYAPiWQeETsgWRRQMDR0/hwp8s8tL/UodwYCScH6Vxx9pdy353L1fK4Bb9G73a+9FPjp9l x+WwKTsltVqSBuSjyZQ3c3EE8qbTidXZxB38JwARH8yN3TX+t65cbBqLl/zRUUUTapHQpUEd yoAsHIml32e4q+3xdLtTdlLi7FgPBItSazcqZPjEcYW73UAuLcmQmfJlQ5PkDiuqcitn+KzH /1pqsTU7QFZjbmSMJyXY0TDErOFuMOjf20b6arcpEqse1V3IKrb+nqqA2azboRm3pEANLAJw iVTwK3qwGRgK5ut6N/Znv20VEHkFUsRAZoOusrIRfR5HFDxlXguAdEz8M/hxXFYYXqOoaCYy 6pJxTjy0Y/tIfmS/g9Bnp8qg9wsrsnk0+XRnDVPak++G3Uq9tJPwpJbyO0vcqEI3vAXkAB7X VXLzvFwi66RrsPUoDkuzj+aCNumtOePDOCpXQGPpKl+l1aYRMN/+lNSk3+1sVuc2C07WnYyE gV/cbEVklPmKrNwu6DeUyD0qI/bVzKMWZAiB1r56hsGeyYcAEQEAAcLBXwQYAQIACQUCUft5 TAIbDAAKCRAu2dd0/nAttYTwEACLAS/THRqXRKb17PQmKwZHerUvZm2klo+lwQ3wNQBHUJAT p2R9ULexyXrJPqjUpy7+voz+FcKiuQBTKyieiIxO46oMxsbXGZ70o3gxjxdYdgimUD6U8PPd JH8tfAL4BR5FZNjspcnscN2jgbF4OrpDeOLyBaj6HPmElNPtECHWCaf1xbIFsZxSDGMA6cUh 0uX3Q8VI7JN1AR2cfiIRY7NrIlWYucJxyKjO3ivWm69nCtsHiJ0wcF8KlVo7F2eLaufo0K8A ynL8SHMF3VEyxsXOP2f1UR9T2Ur30MXcTBpjUxml1TX3RWY5uH89Js/jlIugBwuAmacJ7JYh lTg6sF/GNc4nPb4kk2yktNWTade+TzsllYlJPaorD2Qe8qX0iFUhFC6y9+O6mP4ZvWoYapp9 ezYNuebMgEr93ob1+4sFg3812wNP01WqsGtWCJHnPv/JoonFdMzD/bIkXGEJMk6ks2kxQQZq g6Ik/s/vxOfao/xCn8nHt7GwvVy41795hzK6tbSl+BuyCRp0vfPRP34OnK7+jR2nvQpJu/pU rCELuGwT9hsYkUPjVd4lfylN3mzEc6iAv/wwjsc0DRTSQCpXT3v2ymTAsRKrVaEZLibTXaf+ WslxWek3xNYRiqwwWAJuL652eAlxUgQ5ZS+fXBRTiQpJ+F26I/2lccScRd9G5w== Organization: Red Hat Message-ID: <395e1b02-09b7-9420-33e1-a3abb36282f0@redhat.com> Date: Thu, 16 May 2019 13:59:50 +0200 User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:60.0) Gecko/20100101 Thunderbird/60.6.1 MIME-Version: 1.0 In-Reply-To: Content-Type: text/plain; charset=utf-8 Content-Language: en-US Content-Transfer-Encoding: 7bit X-Scanned-By: MIMEDefang 2.79 on 10.5.11.13 X-Greylist: Sender IP whitelisted, not delayed by milter-greylist-4.5.16 (mx1.redhat.com [10.5.110.42]); Thu, 16 May 2019 11:59:58 +0000 (UTC) Sender: linux-kernel-owner@vger.kernel.org Precedence: bulk List-ID: X-Mailing-List: linux-kernel@vger.kernel.org On 16/05/2019 13.30, David Hildenbrand wrote: > On 16.05.19 13:12, Thomas Huth wrote: >> On s390x, there is a constraint that memory regions have to be aligned >> to 1M (or running the VM will fail). Introduce a new "alignment" variable >> in the vm_userspace_mem_region_add() function which now can be used for >> both, huge page and s390x alignment requirements. >> >> Signed-off-by: Thomas Huth >> --- >> tools/testing/selftests/kvm/lib/kvm_util.c | 21 +++++++++++++++++----- >> 1 file changed, 16 insertions(+), 5 deletions(-) >> >> diff --git a/tools/testing/selftests/kvm/lib/kvm_util.c b/tools/testing/selftests/kvm/lib/kvm_util.c >> index 8d63ccb93e10..64a0da6efe3d 100644 >> --- a/tools/testing/selftests/kvm/lib/kvm_util.c >> +++ b/tools/testing/selftests/kvm/lib/kvm_util.c >> @@ -559,6 +559,7 @@ void vm_userspace_mem_region_add(struct kvm_vm *vm, >> unsigned long pmem_size = 0; >> struct userspace_mem_region *region; >> size_t huge_page_size = KVM_UTIL_PGS_PER_HUGEPG * vm->page_size; >> + size_t alignment; >> >> TEST_ASSERT((guest_paddr % vm->page_size) == 0, "Guest physical " >> "address not on a page boundary.\n" >> @@ -608,9 +609,20 @@ void vm_userspace_mem_region_add(struct kvm_vm *vm, >> TEST_ASSERT(region != NULL, "Insufficient Memory"); >> region->mmap_size = npages * vm->page_size; >> >> - /* Enough memory to align up to a huge page. */ >> +#ifdef __s390x__ >> + /* On s390x, the host address must be aligned to 1M (due to PGSTEs) */ >> + alignment = 0x100000; > > This corresponds to huge_page_size, maybe you can exploit this fact here. > > Something like > > alignment = 1; > > /* On s390x, the host address must always be aligned to the THP size */ > #ifndef __s390x__ > if (src_type == VM_MEM_SRC_ANONYMOUS_THP) > #endif > alignment = huge_page_size; > > Maybe in a nicer fashion. Not sure. Hmm, but if I've got your explanation on IRC right, it's rather a coincidence that the huge page size matches the alignment requirements for KVM memslots, isn't it? So I think the code would look rather confusing if I'd try to shorten it this way...? Thomas From mboxrd@z Thu Jan 1 00:00:00 1970 From: thuth at redhat.com (Thomas Huth) Date: Thu, 16 May 2019 13:59:50 +0200 Subject: [RFC PATCH 2/4] KVM: selftests: Align memory region addresses to 1M on s390x In-Reply-To: References: <20190516111253.4494-1-thuth@redhat.com> <20190516111253.4494-3-thuth@redhat.com> Message-ID: <395e1b02-09b7-9420-33e1-a3abb36282f0@redhat.com> On 16/05/2019 13.30, David Hildenbrand wrote: > On 16.05.19 13:12, Thomas Huth wrote: >> On s390x, there is a constraint that memory regions have to be aligned >> to 1M (or running the VM will fail). Introduce a new "alignment" variable >> in the vm_userspace_mem_region_add() function which now can be used for >> both, huge page and s390x alignment requirements. >> >> Signed-off-by: Thomas Huth >> --- >> tools/testing/selftests/kvm/lib/kvm_util.c | 21 +++++++++++++++++----- >> 1 file changed, 16 insertions(+), 5 deletions(-) >> >> diff --git a/tools/testing/selftests/kvm/lib/kvm_util.c b/tools/testing/selftests/kvm/lib/kvm_util.c >> index 8d63ccb93e10..64a0da6efe3d 100644 >> --- a/tools/testing/selftests/kvm/lib/kvm_util.c >> +++ b/tools/testing/selftests/kvm/lib/kvm_util.c >> @@ -559,6 +559,7 @@ void vm_userspace_mem_region_add(struct kvm_vm *vm, >> unsigned long pmem_size = 0; >> struct userspace_mem_region *region; >> size_t huge_page_size = KVM_UTIL_PGS_PER_HUGEPG * vm->page_size; >> + size_t alignment; >> >> TEST_ASSERT((guest_paddr % vm->page_size) == 0, "Guest physical " >> "address not on a page boundary.\n" >> @@ -608,9 +609,20 @@ void vm_userspace_mem_region_add(struct kvm_vm *vm, >> TEST_ASSERT(region != NULL, "Insufficient Memory"); >> region->mmap_size = npages * vm->page_size; >> >> - /* Enough memory to align up to a huge page. */ >> +#ifdef __s390x__ >> + /* On s390x, the host address must be aligned to 1M (due to PGSTEs) */ >> + alignment = 0x100000; > > This corresponds to huge_page_size, maybe you can exploit this fact here. > > Something like > > alignment = 1; > > /* On s390x, the host address must always be aligned to the THP size */ > #ifndef __s390x__ > if (src_type == VM_MEM_SRC_ANONYMOUS_THP) > #endif > alignment = huge_page_size; > > Maybe in a nicer fashion. Not sure. Hmm, but if I've got your explanation on IRC right, it's rather a coincidence that the huge page size matches the alignment requirements for KVM memslots, isn't it? So I think the code would look rather confusing if I'd try to shorten it this way...? Thomas From mboxrd@z Thu Jan 1 00:00:00 1970 From: thuth@redhat.com (Thomas Huth) Date: Thu, 16 May 2019 13:59:50 +0200 Subject: [RFC PATCH 2/4] KVM: selftests: Align memory region addresses to 1M on s390x In-Reply-To: References: <20190516111253.4494-1-thuth@redhat.com> <20190516111253.4494-3-thuth@redhat.com> Message-ID: <395e1b02-09b7-9420-33e1-a3abb36282f0@redhat.com> Content-Type: text/plain; charset="UTF-8" Message-ID: <20190516115950.tv-XvdIohYjps3e9wkQt4jPDOlq1oBAk1bHUX_Mh8So@z> On 16/05/2019 13.30, David Hildenbrand wrote: > On 16.05.19 13:12, Thomas Huth wrote: >> On s390x, there is a constraint that memory regions have to be aligned >> to 1M (or running the VM will fail). Introduce a new "alignment" variable >> in the vm_userspace_mem_region_add() function which now can be used for >> both, huge page and s390x alignment requirements. >> >> Signed-off-by: Thomas Huth >> --- >> tools/testing/selftests/kvm/lib/kvm_util.c | 21 +++++++++++++++++----- >> 1 file changed, 16 insertions(+), 5 deletions(-) >> >> diff --git a/tools/testing/selftests/kvm/lib/kvm_util.c b/tools/testing/selftests/kvm/lib/kvm_util.c >> index 8d63ccb93e10..64a0da6efe3d 100644 >> --- a/tools/testing/selftests/kvm/lib/kvm_util.c >> +++ b/tools/testing/selftests/kvm/lib/kvm_util.c >> @@ -559,6 +559,7 @@ void vm_userspace_mem_region_add(struct kvm_vm *vm, >> unsigned long pmem_size = 0; >> struct userspace_mem_region *region; >> size_t huge_page_size = KVM_UTIL_PGS_PER_HUGEPG * vm->page_size; >> + size_t alignment; >> >> TEST_ASSERT((guest_paddr % vm->page_size) == 0, "Guest physical " >> "address not on a page boundary.\n" >> @@ -608,9 +609,20 @@ void vm_userspace_mem_region_add(struct kvm_vm *vm, >> TEST_ASSERT(region != NULL, "Insufficient Memory"); >> region->mmap_size = npages * vm->page_size; >> >> - /* Enough memory to align up to a huge page. */ >> +#ifdef __s390x__ >> + /* On s390x, the host address must be aligned to 1M (due to PGSTEs) */ >> + alignment = 0x100000; > > This corresponds to huge_page_size, maybe you can exploit this fact here. > > Something like > > alignment = 1; > > /* On s390x, the host address must always be aligned to the THP size */ > #ifndef __s390x__ > if (src_type == VM_MEM_SRC_ANONYMOUS_THP) > #endif > alignment = huge_page_size; > > Maybe in a nicer fashion. Not sure. Hmm, but if I've got your explanation on IRC right, it's rather a coincidence that the huge page size matches the alignment requirements for KVM memslots, isn't it? So I think the code would look rather confusing if I'd try to shorten it this way...? Thomas