KVM: arm64: A510 errata workaround and fixes for fixup_guest_exit()
Status from this message: Applied to fixes, thanks!
Get series with b4
b4 am [email protected]Versions newest first · linked by author and title
Changes in v2 from the cover letter
* Moved the SPSR_EL2 fixup into a helper called earlier
* Use final cap
* Dropped the IS_ENABLED() check
Patches in v2 Applied
| # | Subject and files | Size | Replies | Applied |
|---|---|---|---|---|
| 1/4 | arm64: Add Cortex-A510 CPU part definitionarch/ | +2 | 0 replies | |
| 2/4 | KVM: arm64: Avoid consuming a stale esr value when SError occurarch/ | +2 | 0 replies | |
| 3/4 | KVM: arm64: Stop handle_exit() from handling HVC twice when an SError occursarch/ | +8 | 0 replies | |
| 4/4 | KVM: arm64: Workaround Cortex-A510's single-step and PAC trap errataDocumentation/ arch/ arch/ +2 more | +46 | 0 replies |
Reviews trailers per patch
1 of 4 patches has a Reviewed-by or Acked-by.
| Patch | Suzuki K Poulose | Catalin Marinas |
|---|---|---|
| 1/4 | R-b | A-b |
| 2/4 | ||
| 3/4 | ||
| 4/4 |
Trailers given in replies are filled; those carried in the patches from earlier versions are outlined with dashes.
Cover letter [PATCH v2 0/4] KVM: arm64: A510 errata workaround and fixes for fixup_guest_exit()
Hello!
Changes since v1:
* Moved the SPSR_EL2 fixup into a helper called earlier
* Use final cap
* Dropped the IS_ENABLED() check
---
Early Cortex-A510 parts have a nasty erratum where two ERETs, pointer-auth and software step conspire to corrupt SPSR_EL2. A guest can only trigger this when it is being stepped by EL2, which gives EL2 the opportunity to work around the erratum. Patch 4 does this, the SDEN is available from:
https://developer.arm.com/documentation/SDEN2397239/900
Patches 2 and 3 fix two issues with the adjacent code where a stale esr value could be used to alter the ELR_EL2 when an IRQ synchronises an SError, and when an HVC synchronises an SError, the HVC may be handled twice, (not just execute twice).
There are three series that would add the Cortex-A510 part macros. I've picked Anshuman's patch that does this, on the assumption that makes someone's life easier. I haven't spotted that patch on the arm64/for-next/fixes branch, so I've not included the hash in the prerequisite field of the CC-stable.
Let me know if you want this reposted once that value is known.
This series is based on v5.17-rc1 and can be retrieved from:
https://git.gitlab.arm.com/linux-arm/linux-jm.git a510_errata/kvm_bits/v2
Thanks,
James
Anshuman Khandual (1):
arm64: Add Cortex-A510 CPU part definition
James Morse (3):
KVM: arm64: Avoid consuming a stale esr value when SError occur
KVM: arm64: Stop handle_exit() from handling HVC twice when an SError
occurs
KVM: arm64: Workaround Cortex-A510's single-step and PAC trap errata
Documentation/arm64/silicon-errata.rst | 2 ++ arch/arm64/Kconfig | 16 ++++++++++++++++ arch/arm64/include/asm/cputype.h | 2 ++ arch/arm64/kernel/cpu_errata.c | 8 ++++++++ arch/arm64/kvm/handle_exit.c | 8 ++++++++ arch/arm64/kvm/hyp/include/hyp/switch.h | 23 +++++++++++++++++++++-- arch/arm64/tools/cpucaps | 1 + 7 files changed, 58 insertions(+), 2 deletions(-)
Signature · 6 lines
2.30.2 _______________________________________________ kvmarm mailing list [email protected] https://lists.cs.columbia.edu/mailman/listinfo/kvmarm