[Qemu-devel] [qemu PATCH v4 0/4] support NFIT platform capabilities

Ross Zwisler posted 4 patches 5 years, 11 months ago
Patches applied successfully (tree, apply log)
git fetch https://github.com/patchew-project/qemu tags/patchew/20180521163203.26590-1-ross.zwisler@linux.intel.com
Test checkpatch passed
Test docker-mingw@fedora passed
Test docker-quick@centos7 passed
Test s390x passed
docs/nvdimm.txt                       |  27 ++++++++++++++++++++
hw/acpi/nvdimm.c                      |  45 +++++++++++++++++++++++++++++++---
hw/i386/pc.c                          |  31 +++++++++++++++++++++++
hw/mem/nvdimm.c                       |   2 +-
include/hw/i386/pc.h                  |   1 +
include/hw/mem/nvdimm.h               |   7 +++++-
tests/.gitignore                      |   1 +
tests/acpi-test-data/pc/NFIT.dimmpxm  | Bin 224 -> 240 bytes
tests/acpi-test-data/q35/NFIT.dimmpxm | Bin 224 -> 240 bytes
tests/bios-tables-test.c              |   2 +-
10 files changed, 109 insertions(+), 7 deletions(-)
[Qemu-devel] [qemu PATCH v4 0/4] support NFIT platform capabilities
Posted by Ross Zwisler 5 years, 11 months ago
Changes since v3:
 * Updated the text in docs/nvdimm.txt to make it clear that the value
   being passed in on the command line in an integer made up of various
   bit fields. (Rob Elliott)
 
 * Updated the "Highest Valid Capability" byte to be dynamic based on
   the highest valid bit in the user's input. (Rob Elliott)

---

The first 2 patches in this series clean up some things I noticed while
coding.

Patch 3 adds support for the new Platform Capabilities Structure, which
was added to the NFIT in ACPI 6.2 Errata A.  We add a machine command
line option "nvdimm-cap":

    -machine pc,accel=kvm,nvdimm,nvdimm-cap=2

which allows the user to pass in a value for this structure.  When such
a value is passed in we will generate the new NFIT subtable.

Patch 4 adds code to the "make check" self test infrastructure so that
we generate the new Platform Capabilities Structure, and adds it to the
expected NFIT output so that we test for it.

Ross Zwisler (4):
  nvdimm: fix typo in label-size definition
  tests/.gitignore: add entry for generated file
  nvdimm, acpi: support NFIT platform capabilities
  ACPI testing: test NFIT platform capabilities

 docs/nvdimm.txt                       |  27 ++++++++++++++++++++
 hw/acpi/nvdimm.c                      |  45 +++++++++++++++++++++++++++++++---
 hw/i386/pc.c                          |  31 +++++++++++++++++++++++
 hw/mem/nvdimm.c                       |   2 +-
 include/hw/i386/pc.h                  |   1 +
 include/hw/mem/nvdimm.h               |   7 +++++-
 tests/.gitignore                      |   1 +
 tests/acpi-test-data/pc/NFIT.dimmpxm  | Bin 224 -> 240 bytes
 tests/acpi-test-data/q35/NFIT.dimmpxm | Bin 224 -> 240 bytes
 tests/bios-tables-test.c              |   2 +-
 10 files changed, 109 insertions(+), 7 deletions(-)

-- 
2.14.3


Re: [Qemu-devel] [qemu PATCH v4 0/4] support NFIT platform capabilities
Posted by Michael S. Tsirkin 5 years, 11 months ago
On Mon, May 21, 2018 at 10:31:59AM -0600, Ross Zwisler wrote:
> Changes since v3:
>  * Updated the text in docs/nvdimm.txt to make it clear that the value
>    being passed in on the command line in an integer made up of various
>    bit fields. (Rob Elliott)
>  
>  * Updated the "Highest Valid Capability" byte to be dynamic based on
>    the highest valid bit in the user's input. (Rob Elliott)

Igor could you review pls? I think your comments have been addressed.

> ---
> 
> The first 2 patches in this series clean up some things I noticed while
> coding.
> 
> Patch 3 adds support for the new Platform Capabilities Structure, which
> was added to the NFIT in ACPI 6.2 Errata A.  We add a machine command
> line option "nvdimm-cap":
> 
>     -machine pc,accel=kvm,nvdimm,nvdimm-cap=2
> 
> which allows the user to pass in a value for this structure.  When such
> a value is passed in we will generate the new NFIT subtable.
> 
> Patch 4 adds code to the "make check" self test infrastructure so that
> we generate the new Platform Capabilities Structure, and adds it to the
> expected NFIT output so that we test for it.
> 
> Ross Zwisler (4):
>   nvdimm: fix typo in label-size definition
>   tests/.gitignore: add entry for generated file
>   nvdimm, acpi: support NFIT platform capabilities
>   ACPI testing: test NFIT platform capabilities
> 
>  docs/nvdimm.txt                       |  27 ++++++++++++++++++++
>  hw/acpi/nvdimm.c                      |  45 +++++++++++++++++++++++++++++++---
>  hw/i386/pc.c                          |  31 +++++++++++++++++++++++
>  hw/mem/nvdimm.c                       |   2 +-
>  include/hw/i386/pc.h                  |   1 +
>  include/hw/mem/nvdimm.h               |   7 +++++-
>  tests/.gitignore                      |   1 +
>  tests/acpi-test-data/pc/NFIT.dimmpxm  | Bin 224 -> 240 bytes
>  tests/acpi-test-data/q35/NFIT.dimmpxm | Bin 224 -> 240 bytes
>  tests/bios-tables-test.c              |   2 +-
>  10 files changed, 109 insertions(+), 7 deletions(-)
> 
> -- 
> 2.14.3

Re: [Qemu-devel] [qemu PATCH v4 0/4] support NFIT platform capabilities
Posted by Ross Zwisler 5 years, 10 months ago
On Fri, May 25, 2018 at 08:51:22PM +0300, Michael S. Tsirkin wrote:
> On Mon, May 21, 2018 at 10:31:59AM -0600, Ross Zwisler wrote:
> > Changes since v3:
> >  * Updated the text in docs/nvdimm.txt to make it clear that the value
> >    being passed in on the command line in an integer made up of various
> >    bit fields. (Rob Elliott)
> >  
> >  * Updated the "Highest Valid Capability" byte to be dynamic based on
> >    the highest valid bit in the user's input. (Rob Elliott)
> 
> Igor could you review pls? I think your comments have been addressed.

Ping?  Igor, can you pick this up?

Re: [Qemu-devel] [qemu PATCH v4 0/4] support NFIT platform capabilities
Posted by Ross Zwisler 5 years, 11 months ago
Ping on this series.  Rob, I think I've addressed all your feedback.  Can you
please verify?

Thanks,
- Ross

On Mon, May 21, 2018 at 10:31:59AM -0600, Ross Zwisler wrote:
> Changes since v3:
>  * Updated the text in docs/nvdimm.txt to make it clear that the value
>    being passed in on the command line in an integer made up of various
>    bit fields. (Rob Elliott)
>  
>  * Updated the "Highest Valid Capability" byte to be dynamic based on
>    the highest valid bit in the user's input. (Rob Elliott)
> 
> ---
> 
> The first 2 patches in this series clean up some things I noticed while
> coding.
> 
> Patch 3 adds support for the new Platform Capabilities Structure, which
> was added to the NFIT in ACPI 6.2 Errata A.  We add a machine command
> line option "nvdimm-cap":
> 
>     -machine pc,accel=kvm,nvdimm,nvdimm-cap=2
> 
> which allows the user to pass in a value for this structure.  When such
> a value is passed in we will generate the new NFIT subtable.
> 
> Patch 4 adds code to the "make check" self test infrastructure so that
> we generate the new Platform Capabilities Structure, and adds it to the
> expected NFIT output so that we test for it.
> 
> Ross Zwisler (4):
>   nvdimm: fix typo in label-size definition
>   tests/.gitignore: add entry for generated file
>   nvdimm, acpi: support NFIT platform capabilities
>   ACPI testing: test NFIT platform capabilities
> 
>  docs/nvdimm.txt                       |  27 ++++++++++++++++++++
>  hw/acpi/nvdimm.c                      |  45 +++++++++++++++++++++++++++++++---
>  hw/i386/pc.c                          |  31 +++++++++++++++++++++++
>  hw/mem/nvdimm.c                       |   2 +-
>  include/hw/i386/pc.h                  |   1 +
>  include/hw/mem/nvdimm.h               |   7 +++++-
>  tests/.gitignore                      |   1 +
>  tests/acpi-test-data/pc/NFIT.dimmpxm  | Bin 224 -> 240 bytes
>  tests/acpi-test-data/q35/NFIT.dimmpxm | Bin 224 -> 240 bytes
>  tests/bios-tables-test.c              |   2 +-
>  10 files changed, 109 insertions(+), 7 deletions(-)
> 
> -- 
> 2.14.3
> 

Re: [Qemu-devel] [qemu PATCH v4 0/4] support NFIT platform capabilities
Posted by Elliott, Robert (Persistent Memory) 5 years, 10 months ago
> Ping on this series.  Rob, I think I've addressed all your feedback.
> Can you please verify?

I haven't tested it, but it reads OK.  I'm OK with just extending
the valid count for bits set to one for now; we can add a new
argument later if a need arises for extending it to express new bits
set to zero.