diff mbox series

[v2,1/2] xen/pvh: set xen_domain_type to HVM in xen_pvh_init

Message ID 20190423130416.68935-1-roger.pau@citrix.com (mailing list archive)
State Accepted
Commit c9f804d64bb93c8dbf957df1d7e9de11380e522d
Headers show
Series [v2,1/2] xen/pvh: set xen_domain_type to HVM in xen_pvh_init | expand

Commit Message

Roger Pau Monné April 23, 2019, 1:04 p.m. UTC
Or else xen_domain() returns false despite xen_pvh being set.

Signed-off-by: Roger Pau Monné <roger.pau@citrix.com>
---
Cc: Boris Ostrovsky <boris.ostrovsky@oracle.com>
Cc: Juergen Gross <jgross@suse.com>
Cc: xen-devel@lists.xenproject.org
---
 arch/x86/xen/enlighten_pvh.c | 1 +
 1 file changed, 1 insertion(+)

Comments

Boris Ostrovsky April 23, 2019, 1:37 p.m. UTC | #1
On 4/23/19 9:04 AM, Roger Pau Monne wrote:
> Or else xen_domain() returns false despite xen_pvh being set.

Is this new xen_domain() invocation somewhere in EFI initialization that
you add in the second patch?

Asking because I am trying to figure out whether this needs to go to
stable tree.

-boris

>
> Signed-off-by: Roger Pau Monné <roger.pau@citrix.com>
> ---
> Cc: Boris Ostrovsky <boris.ostrovsky@oracle.com>
> Cc: Juergen Gross <jgross@suse.com>
> Cc: xen-devel@lists.xenproject.org
> ---
>  arch/x86/xen/enlighten_pvh.c | 1 +
>  1 file changed, 1 insertion(+)
>
> diff --git a/arch/x86/xen/enlighten_pvh.c b/arch/x86/xen/enlighten_pvh.c
> index 35b7599d2d0b..bbffa409e0e8 100644
> --- a/arch/x86/xen/enlighten_pvh.c
> +++ b/arch/x86/xen/enlighten_pvh.c
> @@ -27,6 +27,7 @@ void __init xen_pvh_init(void)
>  	u64 pfn;
>  
>  	xen_pvh = 1;
> +	xen_domain_type = XEN_HVM_DOMAIN;
>  	xen_start_flags = pvh_start_info.flags;
>  
>  	msr = cpuid_ebx(xen_cpuid_base() + 2);
Roger Pau Monné April 23, 2019, 2:56 p.m. UTC | #2
On Tue, Apr 23, 2019 at 09:37:51AM -0400, Boris Ostrovsky wrote:
> On 4/23/19 9:04 AM, Roger Pau Monne wrote:
> > Or else xen_domain() returns false despite xen_pvh being set.
> 
> Is this new xen_domain() invocation somewhere in EFI initialization that
> you add in the second patch?

Yes, there's a xen_initial_domain() call in xen_efi_probe() which
would return false without this fix.

> Asking because I am trying to figure out whether this needs to go to
> stable tree.

I'm not aware of this being an issue without the EFI code that I add
in patch 2, xen_domain_type would get set slightly later in the boot
process.

I guess both patches should be backported to 4.19 since that's the
only LTS release that has PVH dom0 support IIRC, and it would be good
to get this fixed.

Thanks, Roger.
diff mbox series

Patch

diff --git a/arch/x86/xen/enlighten_pvh.c b/arch/x86/xen/enlighten_pvh.c
index 35b7599d2d0b..bbffa409e0e8 100644
--- a/arch/x86/xen/enlighten_pvh.c
+++ b/arch/x86/xen/enlighten_pvh.c
@@ -27,6 +27,7 @@  void __init xen_pvh_init(void)
 	u64 pfn;
 
 	xen_pvh = 1;
+	xen_domain_type = XEN_HVM_DOMAIN;
 	xen_start_flags = pvh_start_info.flags;
 
 	msr = cpuid_ebx(xen_cpuid_base() + 2);