Re: [RFC v6 10/13] target/s390x: use kvm_enabled() to wrap call to kvm_s390_get_hpage_1m

[Al Cho]

Hi Cornelia,

Sorry for missing the reply.

I think it may not be worth it, as you said it seem to be the only call site for kvm_s390_get_hpage_1m().

So I think we could keep it.

Thanks,

            AL

---

From: Cornelia Huck <cohuck@redhat.com>
Sent: Wednesday, June 30, 2021 11:21 PM
To: Al Cho <ACho@suse.com>; qemu-devel@nongnu.org <qemu-devel@nongnu.org>; qemu-s390x@nongnu.org <qemu-s390x@nongnu.org>
Cc: Claudio Fontana <Claudio.Fontana@suse.com>; Al Cho <ACho@suse.com>; José Ricardo Ziviani <jose.ziviani@suse.com>; Claudio Fontana <cfontana@suse.de>
Subject: Re: [RFC v6 10/13] target/s390x: use kvm_enabled() to wrap call to kvm_s390_get_hpage_1m

 

On Tue, Jun 29 2021, "Cho, Yu-Chen" <acho@suse.com> wrote:

> this will allow to remove the kvm stubs.
>
> Signed-off-by: Claudio Fontana <cfontana@suse.de>
> Signed-off-by: Cho, Yu-Chen <acho@suse.com>
> ---
>  target/s390x/diag.c | 3 ++-
>  1 file changed, 2 insertions(+), 1 deletion(-)
>
> diff --git a/target/s390x/diag.c b/target/s390x/diag.c
> index c17a2498a7..8405f69df0 100644
> --- a/target/s390x/diag.c
> +++ b/target/s390x/diag.c
> @@ -20,6 +20,7 @@
>  #include "hw/s390x/ipl.h"
>  #include "hw/s390x/s390-virtio-ccw.h"
>  #include "hw/s390x/pv.h"
> +#include "sysemu/kvm.h"
>  #include "kvm_s390x.h"
> 
>  int handle_diag_288(CPUS390XState *env, uint64_t r1, uint64_t r3)
> @@ -168,7 +169,7 @@ out:
>              return;
>          }
> 
> -        if (kvm_s390_get_hpage_1m()) {
> +        if (kvm_enabled() && kvm_s390_get_hpage_1m()) {

I think I asked before whether we should introduce a
s390_huge_page_backing() wrapper (which might be overkill)... any
opinions on that? I'm not really opposed to this patch here, either.

>              error_report("Protected VMs can currently not be backed with "
>                           "huge pages");
>              env->regs[r1 + 1] = DIAG_308_RC_INVAL_FOR_PV;