From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from eggs.gnu.org ([209.51.188.92]:54109) by lists.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1glU07-0006j6-2F for qemu-devel@nongnu.org; Mon, 21 Jan 2019 02:28:07 -0500 Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71) (envelope-from ) id 1glU06-0000bk-6U for qemu-devel@nongnu.org; Mon, 21 Jan 2019 02:28:07 -0500 Received: from mx1.redhat.com ([209.132.183.28]:52502) by eggs.gnu.org with esmtps (TLS1.0:DHE_RSA_AES_256_CBC_SHA1:32) (Exim 4.71) (envelope-from ) id 1glU05-0000ZG-UQ for qemu-devel@nongnu.org; Mon, 21 Jan 2019 02:28:06 -0500 Date: Mon, 21 Jan 2019 02:21:39 -0500 (EST) From: Pankaj Gupta Message-ID: <790023026.65652877.1548055299697.JavaMail.zimbra@redhat.com> In-Reply-To: <20190121061143.GB21562@tiger-server> References: <20190116103615-mutt-send-email-mst@kernel.org> <20190121061143.GB21562@tiger-server> MIME-Version: 1.0 Content-Type: text/plain; charset=utf-8 Content-Transfer-Encoding: quoted-printable Subject: Re: [Qemu-devel] [PATCH V9 6/6] docs: Added MAP_SYNC documentation List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , To: Yi Zhang Cc: "Michael S. Tsirkin" , xiaoguangrong eric , stefanha@redhat.com, pbonzini@redhat.com, yu c zhang , ehabkost@redhat.com, qemu-devel@nongnu.org, imammedo@redhat.com, dan j williams > > > --- > > > docs/nvdimm.txt | 21 ++++++++++++++++++++- > > > qemu-options.hx | 4 ++++ > > > 2 files changed, 24 insertions(+), 1 deletion(-) > > >=20 > > > diff --git a/docs/nvdimm.txt b/docs/nvdimm.txt > > > index 5f158a6..565ba73 100644 > > > --- a/docs/nvdimm.txt > > > +++ b/docs/nvdimm.txt > > > @@ -142,11 +142,30 @@ backend of vNVDIMM: > > > Guest Data Persistence > > > ---------------------- > > > =20 > > > +vNVDIMM is designed and implemented to guarantee the guest data > > > +persistence on the backends even on the host crash and power > >=20 > > in case of a host crash or a power failure > >=20 > > (be consistent) > >=20 > > > +failures. However, there are still some requirements and limitations > > > +as explained below. > > > + > > > Though QEMU supports multiple types of vNVDIMM backends on Linux, > > > -currently the only one that can guarantee the guest write persistenc= e > > > +if MAP_SYNC is not supported by the host kernel and the backends, > > > +the only backend that can guarantee the guest write persistence > > > is the device DAX on the real NVDIMM device (e.g., /dev/dax0.0), to > > > which all guest access do not involve any host-side kernel cache. > > > =20 > > > +mmap(2) flag MAP_SYNC is added since Linux kernel 4.15. On such > > > +systems, QEMU can mmap(2) the backend with MAP_SYNC, which can ensur= e > >=20 > > which ensures > >=20 > > > +filesystem metadata consistent > >=20 > > consistency > >=20 > > > even after a system crash or power > >=20 > > a power > >=20 > > > +failure. Besides the host kernel support, enabling MAP_SYNC in QEMU > > > +also requires: > > > + > > > + - the backend is a file supporting DAX, e.g., a file on an ext4 or > > > + xfs file system mounted with '-o dax', > > > + > > > + - 'share' option of memory-backend-file is 'on'. > >=20 > > why does share need to be on? > According to mmap(2) man page, > MAP_SYNC is available only with the MAP_SHARED_VALIDATE map=E2=80=90ping = type. > I will add this to documentation, Thanks. MAP_SHARE_VALIDATE should be used to validate MAP_SYNC with MAP_SHARED. #define MAP_SHARED_VALIDATE 0x03 In contrast, MAP_SHARED is: #define MAP_SHARED 0x01 /* Share changes */ I think we should use MAP_SHARED_VALIDATE with MAP_SYNC here? Thanks, Pankaj=20 > >=20 > > > + > > > + - 'pmem' option of memory-backend-file is 'on' > > > + > > > When using other types of backends, it's suggested to set 'unarmed' > > > option of '-device nvdimm' to 'on', which sets the unarmed flag of t= he > > > guest NVDIMM region mapping structure. This unarmed flag indicates > > > diff --git a/qemu-options.hx b/qemu-options.hx > > > index 08f8516..545cb8a 100644 > > > --- a/qemu-options.hx > > > +++ b/qemu-options.hx > > > @@ -4002,6 +4002,10 @@ using the SNIA NVM programming model (e.g. Int= el > > > NVDIMM). > > > If @option{pmem} is set to 'on', QEMU will take necessary operations= to > > > guarantee the persistence of its own writes to @option{mem-path} > > > (e.g. in vNVDIMM label emulation and live migration). > > > +Also, we will map the backend-file with MAP_SYNC flag, which can ens= ure > > > +the file metadata is in sync to @option{mem-path} even on the host c= rash > > > +and power failures. MAP_SYNC requires supports from both the host ke= rnel > >=20 > > requires support > >=20 > > > +(since Linux kernel 4.15) and @option{mem-path} (only files supporti= ng > > > DAX). > > > =20 > > > @item -object > > > memory-backend-ram,id=3D@var{id},merge=3D@var{on|off},dump=3D@var{on= |off},share=3D@var{on|off},prealloc=3D@var{on|off},size=3D@var{size},host-n= odes=3D@var{host-nodes},policy=3D@var{default|preferred|bind|interleave} > > > =20 > > > -- > > > 2.7.4 >=20