On 2020-04-29 22:21, Will Deacon wrote:
Hi Andrew,
On Wed, Apr 22, 2020 at 05:13:46PM +0100, Andrew Scull wrote:
Errata 1165522, 1319367 and 1530923 each allow TLB entries to be
allocated as a result of a speculative AT instruction. In order to
avoid mandating VHE on certain affected CPUs, apply the workaround to
both the nVHE and the VHE case for all affected CPUs.
Signed-off-by: Andrew Scull <ascull@xxxxxxxxxx>
CC: Marc Zyngier <maz@xxxxxxxxxx>
CC: James Morse <james.morse@xxxxxxx>
CC: Suzuki K Poulose <suzuki.poulose@xxxxxxx>
CC: Will Deacon <will@xxxxxxxxxx>
CC: Steven Price <steven.price@xxxxxxx>
---
* From v1 <20200327143941.195626-1-ascull@xxxxxxxxxx>:
- Restored registers in VHE path
This largely looks good to me, but I don't understand these bits:
diff --git a/arch/arm64/kvm/hyp/switch.c b/arch/arm64/kvm/hyp/switch.c
index 8a1e81a400e0..651820f537fb 100644
--- a/arch/arm64/kvm/hyp/switch.c
+++ b/arch/arm64/kvm/hyp/switch.c
@@ -138,7 +138,7 @@ static void __hyp_text
__activate_traps_nvhe(struct kvm_vcpu *vcpu)
write_sysreg(val, cptr_el2);
- if (cpus_have_final_cap(ARM64_WORKAROUND_SPECULATIVE_AT_NVHE)) {
+ if (cpus_have_const_cap(ARM64_WORKAROUND_SPECULATIVE_AT)) {
It seems like you consistently replace cpus_have_final_cap() with
cpus_have_const_cap(), but I can't figure out why that's required.
Seems like a bad conflict resolution. We definitely want to keep the
final_cap checks for anything that will run at EL2, and probably
everywhere else (if capabilities are not final by the time we hit KVM,
we have bigger problems to solve).
Thanks,
M.
--
Jazz is not dead. It just smells funny...
_______________________________________________
kvmarm mailing list
kvmarm@xxxxxxxxxxxxxxxxxxxxx
https://lists.cs.columbia.edu/mailman/listinfo/kvmarm
