[PATCH v6 0/3] Add support for Fujitsu A64FX processor

Shuuichirou Ishii posted 3 patches 2 years, 7 months ago
Test checkpatch passed
Patches applied successfully (tree, apply log)
git fetch https://github.com/patchew-project/qemu tags/patchew/20210831082940.2811719-1-ishii.shuuichir@fujitsu.com
Maintainers: Paolo Bonzini <pbonzini@redhat.com>, Laurent Vivier <lvivier@redhat.com>, Thomas Huth <thuth@redhat.com>, Peter Maydell <peter.maydell@linaro.org>
docs/system/arm/virt.rst       |  1 +
hw/arm/virt.c                  |  1 +
target/arm/cpu64.c             | 48 ++++++++++++++++++++++++++++++++++
tests/qtest/arm-cpu-features.c | 13 +++++++++
4 files changed, 63 insertions(+)
[PATCH v6 0/3] Add support for Fujitsu A64FX processor
Posted by Shuuichirou Ishii 2 years, 7 months ago
This is the v6 patch series.

v6:
For patch 1[1/3], added the commit messages that the Identification registers
value are defined based on FX700, and has been tested and confirmed.

v5:
A64FX supports only 128, 256, and 512bit SVE vector lengths,
but the QEMU implementation prior to v4 did not have an API
to specify a specific vector length.
Andrew has implemented an API (cpu->sve_vq_supported, commit:5401b1e08d
etc) to solve this, so we have changed the implementation to use that API.

v4:
The following changes have been made to match the SVE specification of
the A64FX processor.
- Implemented internally only the vector lengths of 128, 256, and 512 bit
  supported by the A64FX processor.
- Removed sve and sve-max-vq properties due to the above changes, and
  fixed them so that no explicit options can be specified.

v3:
When we created the v2 patch series, we based it on the v1 patch series
that had not been merged into the upstream, so we created the v3 patch
series as a patch series that can be applied independently.

v2:
No features have been added or removed from the v1 patch series. Removal
of unused definitions that were added in excess, and consolidation of
patches for the purpose of functional consistency.

For patch 1, Implemented Identification registers for A64FX processor.
HPC extension registers will be implemented in the future.
For SVE, the A64FX processor supports only 128,256 and 512bit vector lengths.
The Identification registers value are defined based on the FX700,
and have been tested and confirmed.

For patch 2, A64FX processor can now be used by specifying the -cpu
a64fx option when the -macine virt option is specified.

For patch 3, added A64FX processor related tests.

Shuuichirou Ishii (3):
  target-arm: Add support for Fujitsu A64FX
  hw/arm/virt: target-arm: Add A64FX processor support to virt machine
  tests/arm-cpu-features: Add A64FX processor related

 docs/system/arm/virt.rst       |  1 +
 hw/arm/virt.c                  |  1 +
 target/arm/cpu64.c             | 48 ++++++++++++++++++++++++++++++++++
 tests/qtest/arm-cpu-features.c | 13 +++++++++
 4 files changed, 63 insertions(+)

-- 
2.27.0


Re: [PATCH v6 0/3] Add support for Fujitsu A64FX processor
Posted by Peter Maydell 2 years, 7 months ago
On Tue, 31 Aug 2021 at 09:29, Shuuichirou Ishii
<ishii.shuuichir@fujitsu.com> wrote:
>
> This is the v6 patch series.
>
> v6:
> For patch 1[1/3], added the commit messages that the Identification registers
> value are defined based on FX700, and has been tested and confirmed.


Applied to target-arm.next. Thanks!

-- PMM