[GIT PULL] KVM/riscv changes for 7.4

[Anup Patel]

On Thu, Oct 22, 2026 at 6:36 PM Paolo Bonzini <pbonzini at redhat.com> wrote:
>
> On Thu, Oct 22, 2026 at 2:51 PM Anup Patel <anup at brainfault.org> wrote:
> >
> > Hi Paolo,
> >
> > I was on vacation last week; hence, I am sending
> > the KVM RISC-V 7.4 PR a bit late. Apologies for
> > the inconvenience.
>
> The problem is not the PR, it's that *none* of these changes were in
> linux-next until yesterday.  This is not the first time that I notice
> this and remind you of it...
>
> These are 80% bugfixes or otherwise not new features (e.g. 3/4/5/6 in
> the list below would be fine even for -rcN), so I'll pull
> nevertheless. But *please* try to commit stuff as soon as it's ready!

I generally update riscv_kvm_next branch much earlier before
merge window but this time I missed. Going forward, I will ensure
riscv_kvm_next branch is updated sooner.

Thanks,
Anup


>
> Paolo
>
> >
> > We have the following KVM RISC-V changes for 7.4:
> > 1) Lazy V-extension enablement for KVM selftests
> > 2) Implement KVM_PRE_FAULT_MEMORY ioctl() and enable corresponding KVM selftests
> > 3) Show sepc/scause/stval CSRs in KVM selftests vcpu dump
> > 4) Use a raw lock for IMSIC wakeup state
> > 5) Reject vCPU creation after AIA initialization
> > 6) Updated next branch name in KVM/riscv MAINTAINERS entry
> > 7) Allow Zicclsm, Ziccamoa, Ziccif, Za64rs for KVM guest
> >
> > In addition to the above, we also have the following fixes
> > and minor improvements:
> > 1) Avoid redundant interrupt check for pending timers
> > 2) Fix build with vector support disabled
> > 3) Fix IPI delivery for out-of-order vcpu_id
> > 4) Replace imsic switch-case macros with range checks
>