> -----Original Message-----
> From: Michael S. Tsirkin [mailto:mst@redhat.com]
> Sent: 29 March 2020 11:46
> To: Shameerali Kolothum Thodi <shameerali.kolothum.thodi@huawei.com>
> Cc: qemu-devel@nongnu.org; qemu-arm@nongnu.org;
> eric.auger@redhat.com; imammedo@redhat.com; peter.maydell@linaro.org;
> shannon.zhaosl@gmail.com; xiaoguangrong.eric@gmail.com;
> david@redhat.com; xuwei (O) <xuwei5@huawei.com>; lersek@redhat.com;
> Linuxarm <linuxarm@huawei.com>
> Subject: Re: [PATCH v3 00/10] ARM virt: Add NVDIMM support
> 
> On Wed, Mar 11, 2020 at 05:20:04PM +0000, Shameer Kolothum wrote:
> > This series adds NVDIMM support to arm/virt platform.
> 
> 
> So I'm still confused about whether there's a bugfix here
> that we need for 5.0. If yes pls post just that part
> with acks included and for-5.0 in the subject.

Ok. I can send the first 4 patches in this series as general fixes,
but as I mentioned earlier they only matter if we add nvdimm arm/virt
support. The only case I am not sure that may break x86 is the RSDP
table update and the resulting size inconsistency during migration 
discussed in patch #2.
 
Anyways I think it is better to send those separately.

Thanks,
Shameer 

> > The series reuses some of the patches posted by Eric
> > in his earlier attempt here[1].
> >
> > This also include few fixes to qemu in general which were
> > discovered while adding nvdimm support to arm/virt.
> >
> > Patch #2 addresses the issue[2] that, during migration, the
> > source and destination might end up with an inconsistency
> > in acpi table memory region sizes.
> >
> > Patch #3 is to fix the qemu_ram_resize() callback issue[2].
> >
> > Patch #4 is another fix to the nvdimm aml issue discussed
> > here[3].
> >
> > I have done a basic sanity testing of NVDIMM devices
> > with Guest booting with ACPI. Further testing is always
> > welcome.
> >
> > Please let me know your feedback.
> >
> > Thanks,
> > Shameer
> >
> > [1] https://patchwork.kernel.org/cover/10830777/
> > [2] https://patchwork.kernel.org/patch/11339591/
> > [3] https://patchwork.kernel.org/cover/11174959/
> >
> > v2 --> v3
> >  - Added patch #1 and # 2 to fix the inconsistency in acpi
> >    table memory region sizes during migration. Thanks to
> >    David H.
> >  - The fix for qemu_ram_resize() callback was modified to
> >    the one in patch #3. Again thanks to David H.
> >  - Addressed comments from MST and Eric on tests added.
> >  - Addressed comments from Igor/MST on Integer size in patch #4
> >  - Added Eric's R-by to patch #7.
> >
> > v1 --> v2
> >  -Reworked patch #1 and now fix is inside qemu_ram_resize().
> >  -Added patch #2 to fix the nvdim aml issue.
> >  -Dropped support to DT cold plug.
> >  -Updated test_acpi_virt_tcg_memhp() with pc-dimm and nvdimms(patch
> #7)
> >
> > David Hildenbrand (1):
> >   exec: Fix for qemu_ram_resize() callback
> >
> > Kwangwoo Lee (2):
> >   nvdimm: Use configurable ACPI IO base and size
> >   hw/arm/virt: Add nvdimm hot-plug infrastructure
> >
> > Shameer Kolothum (7):
> >   acpi: Use macro for table-loader file name
> >   fw_cfg: Migrate ACPI table mr sizes separately
> >   hw/acpi/nvdimm: Fix for NVDIMM incorrect DSM output buffer length
> >   hw/arm/virt: Add nvdimm hotplug support
> >   tests: Update ACPI tables list for upcoming arm/virt test changes
> >   tests/bios-tables-test: Update arm/virt memhp test
> >   tests/acpi: add expected tables for bios-tables-test
> >
> >  docs/specs/acpi_hw_reduced_hotplug.rst |   1 +
> >  exec.c                                 |  14 +++-
> >  hw/acpi/generic_event_device.c         |  15 ++++-
> >  hw/acpi/nvdimm.c                       |  72
> +++++++++++++++++----
> >  hw/arm/Kconfig                         |   1 +
> >  hw/arm/virt-acpi-build.c               |   8 ++-
> >  hw/arm/virt.c                          |  35 ++++++++--
> >  hw/core/machine.c                      |   1 +
> >  hw/i386/acpi-build.c                   |   8 ++-
> >  hw/i386/acpi-build.h                   |   3 +
> >  hw/i386/pc_piix.c                      |   2 +
> >  hw/i386/pc_q35.c                       |   2 +
> >  hw/mem/Kconfig                         |   2 +-
> >  hw/nvram/fw_cfg.c                      |  86
> ++++++++++++++++++++++++-
> >  include/hw/acpi/aml-build.h            |   1 +
> >  include/hw/acpi/generic_event_device.h |   1 +
> >  include/hw/arm/virt.h                  |   1 +
> >  include/hw/mem/nvdimm.h                |   3 +
> >  include/hw/nvram/fw_cfg.h              |   6 ++
> >  tests/data/acpi/pc/SSDT.dimmpxm        | Bin 685 -> 734 bytes
> >  tests/data/acpi/q35/SSDT.dimmpxm       | Bin 685 -> 734 bytes
> >  tests/data/acpi/virt/DSDT.memhp        | Bin 6644 -> 6668 bytes
> >  tests/data/acpi/virt/NFIT.memhp        | Bin 0 -> 224 bytes
> >  tests/data/acpi/virt/SSDT.memhp        | Bin 0 -> 736 bytes
> >  tests/qtest/bios-tables-test.c         |   9 ++-
> >  25 files changed, 244 insertions(+), 27 deletions(-)
> >  create mode 100644 tests/data/acpi/virt/NFIT.memhp
> >  create mode 100644 tests/data/acpi/virt/SSDT.memhp
> >
> > --
> > 2.17.1
> >

On 30.03.20 10:44, Shameerali Kolothum Thodi wrote:
> 
> 
>> -----Original Message-----
>> From: Michael S. Tsirkin [mailto:mst@redhat.com]
>> Sent: 29 March 2020 11:46
>> To: Shameerali Kolothum Thodi <shameerali.kolothum.thodi@huawei.com>
>> Cc: qemu-devel@nongnu.org; qemu-arm@nongnu.org;
>> eric.auger@redhat.com; imammedo@redhat.com; peter.maydell@linaro.org;
>> shannon.zhaosl@gmail.com; xiaoguangrong.eric@gmail.com;
>> david@redhat.com; xuwei (O) <xuwei5@huawei.com>; lersek@redhat.com;
>> Linuxarm <linuxarm@huawei.com>
>> Subject: Re: [PATCH v3 00/10] ARM virt: Add NVDIMM support
>>
>> On Wed, Mar 11, 2020 at 05:20:04PM +0000, Shameer Kolothum wrote:
>>> This series adds NVDIMM support to arm/virt platform.
>>
>>
>> So I'm still confused about whether there's a bugfix here
>> that we need for 5.0. If yes pls post just that part
>> with acks included and for-5.0 in the subject.
> 
> Ok. I can send the first 4 patches in this series as general fixes,
> but as I mentioned earlier they only matter if we add nvdimm arm/virt
> support. The only case I am not sure that may break x86 is the RSDP
> table update and the resulting size inconsistency during migration 
> discussed in patch #2.

I think at least the first three patches fix real issues that are
possible on x86.

-- 
Thanks,

David / dhildenb