[v6] hw/arm: GIC 'its=off' ACPI table fixes

[PATCH-for-10.1 v6 0/9] hw/arm: GIC 'its=off' ACPI table fixes

Posted by Gustavo Romero 4 months, 2 weeks ago

Since v2:
- Fixed no_tcg_its inverted logic (rth)

Since v3:
- Fixed remappings in the IORT table when ITS is no present
- Rebased on master and resoled conflics, like no more "no_its"
  flag in VirtMachineClass
- Dropped patch 1/9 because we actually want the instance flags,
  not only the class flags, and the instance flags are the ones
  to be used often when deciding about the presence/absence of a
  machine feature, instead of the negated class flags ("no_*")
- Adapted the other patches that depended on 1/9
- Dropped patch 4/9 in favor of using the instance flag for
  checking if ITS is on or off
- Simplified VM options for the new "its=off" test

Since v4:
- Fixed/restored commit authorship and author tags (philmd)
- Factored out create_its_idmaps to clarify how ID ranges that
  go direct to ITS Group node are created (eauger)
- Addressed all remaining minor review comments (eauger)

Since v5:
- Use of "directed to" in comments and commit messages (eauger)
- Renaming of variables num_id_mappings, smmu_idmaps, and its_idmaps
  to make code clearer (eauger) 

v1: https://lists.gnu.org/archive/html/qemu-devel/2025-03/msg07080.html
v2: https://lists.gnu.org/archive/html/qemu-devel/2025-04/msg00495.html (Patches 6/14 -> 14/14 in the series)
v3: https://lists.gnu.org/archive/html/qemu-devel/2025-04/msg00567.html
v4: https://mail.gnu.org/archive/html/qemu-devel/2025-06/msg02583.html
v5: https://mail.gnu.org/archive/html/qemu-devel/2025-06/msg03793.html

Fix ACPI tables for '-M its=off' CLI option and resolve the issue:

https://gitlab.com/qemu-project/qemu/-/issues/2886


Cheers,
Gustavo

Gustavo Romero (6):
  hw/arm/virt: Simplify logic for setting instance's 'tcg_its' variable
  hw/arm/virt-acpi-build: Improve comment in build_iort
  hw/arm/virt-acpi-build: Factor out create_its_idmaps
  qtest/bios-tables-test: Add blobs for its=off test on aarch64
  hw/arm/virt-acpi-build: Fix ACPI IORT and MADT tables when its=off
  qtest/bios-tables-test: Update blobs for its=off test on aarch64

Philippe Mathieu-Daudé (3):
  hw/intc/gicv3_its: Do not check its_class_name()
  hw/arm/virt: Simplify create_its()
  qtest/bios-tables-test: Add test for when ITS is off on aarch64

 hw/arm/virt-acpi-build.c                  | 190 ++++++++++++++--------
 hw/arm/virt.c                             |  25 +--
 include/hw/intc/arm_gicv3_its_common.h    |   2 +-
 tests/data/acpi/aarch64/virt/APIC.its_off | Bin 0 -> 164 bytes
 tests/data/acpi/aarch64/virt/IORT.its_off | Bin 0 -> 172 bytes
 tests/qtest/bios-tables-test.c            |  21 +++
 6 files changed, 157 insertions(+), 81 deletions(-)
 create mode 100644 tests/data/acpi/aarch64/virt/APIC.its_off
 create mode 100644 tests/data/acpi/aarch64/virt/IORT.its_off

-- 
2.34.1

Re: [PATCH-for-10.1 v6 0/9] hw/arm: GIC 'its=off' ACPI table fixes

Posted by Peter Maydell 4 months, 2 weeks ago

On Sat, 28 Jun 2025 at 20:57, Gustavo Romero <gustavo.romero@linaro.org> wrote:
>
> Since v2:
> - Fixed no_tcg_its inverted logic (rth)
>
> Since v3:
> - Fixed remappings in the IORT table when ITS is no present
> - Rebased on master and resoled conflics, like no more "no_its"
>   flag in VirtMachineClass
> - Dropped patch 1/9 because we actually want the instance flags,
>   not only the class flags, and the instance flags are the ones
>   to be used often when deciding about the presence/absence of a
>   machine feature, instead of the negated class flags ("no_*")
> - Adapted the other patches that depended on 1/9
> - Dropped patch 4/9 in favor of using the instance flag for
>   checking if ITS is on or off
> - Simplified VM options for the new "its=off" test
>
> Since v4:
> - Fixed/restored commit authorship and author tags (philmd)
> - Factored out create_its_idmaps to clarify how ID ranges that
>   go direct to ITS Group node are created (eauger)
> - Addressed all remaining minor review comments (eauger)
>
> Since v5:
> - Use of "directed to" in comments and commit messages (eauger)
> - Renaming of variables num_id_mappings, smmu_idmaps, and its_idmaps
>   to make code clearer (eauger)
>
> v1: https://lists.gnu.org/archive/html/qemu-devel/2025-03/msg07080.html
> v2: https://lists.gnu.org/archive/html/qemu-devel/2025-04/msg00495.html (Patches 6/14 -> 14/14 in the series)
> v3: https://lists.gnu.org/archive/html/qemu-devel/2025-04/msg00567.html
> v4: https://mail.gnu.org/archive/html/qemu-devel/2025-06/msg02583.html
> v5: https://mail.gnu.org/archive/html/qemu-devel/2025-06/msg03793.html
>
> Fix ACPI tables for '-M its=off' CLI option and resolve the issue:
>
> https://gitlab.com/qemu-project/qemu/-/issues/2886


Applied to target-arm.next, thanks. (I fixed a few checkpatch
nits about hardcoded tabs and an overlong line.)

-- PMM