Re: [Qemu-devel] [PATCH v3 00/32] implement vNVDIMM

[Xiao Guangrong]

On 10/13/2015 12:36 AM, Dan Williams wrote:
> On Sun, Oct 11, 2015 at 9:33 PM, Xiao Guangrong
> <address@hidden> wrote:
> > On 10/11/2015 05:17 AM, Dan Williams wrote:
> > > On Sat, Oct 10, 2015 at 8:52 PM, Xiao Guangrong
> > > <address@hidden> wrote:
> > > [..]
> > > > ====== Test ======
> > > > In host
> > > > 1) create memory backed file, e.g # dd if=zero of=/tmp/nvdimm bs=1G
> > > > count=10
> > > > 2) append "-object memory-backend-file,share,id=mem1,
> > > >      mem-path=/tmp/nvdimm -device nvdimm,memdev=mem1,reserve-label-data,
> > > >      id=nv1" in QEMU command line
> > > >
> > > > In guest, download the latest upsteam kernel (4.2 merge window) and
> > > > enable
> > > > ACPI_NFIT, LIBNVDIMM and BLK_DEV_PMEM.
> > > > 1) insmod drivers/nvdimm/libnvdimm.ko
> > > > 2) insmod drivers/acpi/nfit.ko
> > > > 3) insmod drivers/nvdimm/nd_btt.ko
> > > > 4) insmod drivers/nvdimm/nd_pmem.ko
> > > > You can see the whole nvdimm device used as a single namespace and
> > > > /dev/pmem0
> > > > appears. You can do whatever on /dev/pmem0 including DAX access.
> > > >
> > > > Currently Linux NVDIMM driver does not support namespace operation on
> > > > this
> > > > kind of PMEM, apply below changes to support dynamical namespace:
> > > >
> > > > @@ -798,7 +823,8 @@ static int acpi_nfit_register_dimms(struct
> > > > acpi_nfit_desc *a
> > > >                           continue;
> > > >                   }
> > > >
> > > > -               if (nfit_mem->bdw && nfit_mem->memdev_pmem)
> > > > +               //if (nfit_mem->bdw && nfit_mem->memdev_pmem)
> > > > +               if (nfit_mem->memdev_pmem)
> > > >                           flags |= NDD_ALIASING;
> > >
> > >
> > > This is just for testing purposes, right?  I expect guests can
> >
> >
> > It's used to validate NVDIMM _DSM method and static namespace following
> > NVDIMM specs...
>
> Static namespaces can be emitted without a label.  Linux needs this to
> support existing "label-less" bare metal NVDIMMs.

This is Linux specific? As i did not see it has been documented in the
spec...

> > > sub-divide persistent memory capacity by partitioning the resulting
> > > block device(s).
> >
> >
> > I understand that it's a Linux design... Hmm, can the same expectation
> > apply to PBLK?
>
> BLK-mode is a bit different as those namespaces have both configurable
> sector-size and an optional BTT.  It is possible to expect multiple
> BLK namespaces per a given region with different settings.

Okay, thanks for your nice explanation, Dan!