日期

05 Oct 2022

 参与者

使用 @ 提及姓名以列出参会人员

• Jonathan Cameron

• Lorenzo Pieralisi

• @James Morse

• Salil Mehta

• @ilkka@os.amperecomputing.com

• @shameer Kolothum

 目标

Meeting Minutes:

1. Salil presented some updates on the testing of the James Kernel patches with the QEMU.

   1. https://git.gitlab.arm.com/linux-arm/linux-jm.gitvirtual_cpu_hotplug/rfc/v0

   2. Slides share as attached

2. Forward ported QEMU with some fixes was shared (by Salil)

   1. https://github.com/salil-mehta/qemu.gitvirt-cpuhp-armv8/rfc-v1-port29092022

3. Some discussions on the use of _STA.EnableBit during remove of cpu which was causing crash.

   1. There is an assumption in the patches that _STA.ENA=0 while cpus are being removed.

4. Issue of present == possible with the James patches was also discussed

   1. Could we get around this by conditionally making cpus present in the kernel (by Salil)

      1. "ACPI says present but Linux still says not present" is an inconsistent representation and can lead to future maintenance problems (by James)

      2. What kind of future problems? (needs more debate?)

5. Issue with removing cold-booted cpus was also discussed

   1. Jonathan suggested keeping a variable in the kernel to identify the cpu which was earlier cold-booted or we could even use GICC Enabled/online-capable flag bits from MADT Table.

6. A thought to evolve the ACPI handshake protocol between firmware and kernel was also discussed

   1. Jonathan floated an idea of using the _OSC method ?

   2. James mentioned the limitation that ifGICC.Enabled=1 during boot then none of the _STA fields could change as it effects the functionality of the 'kexec'

7. James would be using the forward ported QEMU repo for further testing and fixing. He might not be available for next few weeks as he would shift his focus on MPAM

8. QEMU still has lots of issue to be resolved, Salil shall continue to work in refining those and help James in further resolving the issues with his approach

9. A variant of James approach with conditionally making CPU present has also been shared with the ARM folks for their humble consideration.

   1. https://github.com/salil-mehta/linux.gitvirt-cpuhp-arm64/rfc-v2/jmorse-variant-with-cond-present-cpu

   2. This has been found working in all the cases. Although, the issue about inconsistency between ACPI and kernel needs a thorough discussion!

Recordings:

• https://linaro-org.zoom.us/rec/share/u5f3QC-lVSeaQZ5Xwmwy0jxDroPA7NPmd7Uf4ICBbSYNE9gR0FujDee31PtwyTNb.8VdwBw_b6UrQALUH
  密码：N0*&pcfb

• Topics

 行动项

添加行动事项，完成未解决的问题或讨论主题:

 决策

输入“/决策”记录您在本次会议中做出的决定: