On Wed, Aug 05, 2026 at 01:32:26PM +0800, Mark Brown wrote:
> On Wed, Aug 05, 2026 at 05:48:04PM +0800, Yao Yuan wrote:
> > On Mon, Aug 03, 2026 at 11:56:48PM +0800, Mark Brown wrote:
>
> > >
> > > if (ctxt_has_s1poe(&vcpu->arch.ctxt))
> > > write_sysreg_el1(__vcpu_sys_reg(vcpu, POR_EL2),
> > > SYS_POR);
> > > +
> > > + if (ctxt_has_gcs(&vcpu->arch.ctxt)) {
> > > + write_sysreg_el1(__vcpu_sys_reg(vcpu, GCSCR_EL2),
> > > SYS_GCSCR);
> > > + write_sysreg_el1(__vcpu_sys_reg(vcpu, GCSPR_EL2),
> > > SYS_GCSPR);
> > > + }
>
> > Looks the s1pie checking for GCS vEL2 context switch is
> > necessary yet IIUC, KVM does vEL2 context switch for nested
> > L1's EL2, it should follow same rule as non-nested L1's EL1
> > context switch.
>
> Right, it's not really needed in either case but hopefully helps code
> generation a bit. Thanks for spotting that.
>
> Please delete unneeded context from mails when replying. Doing this
> makes it much easier to find your reply in the message, helping ensure
> it won't be missed by people scrolling through the irrelevant quoted
> material.
Sure. let me try delete more next time.
I delete some but decide to leave the whole context switch
diffs to help other guys to see the realted changes when
read my comment.
