Re: [PATCH RFC v1 11/13] xen/libxl: allow creating HVM guests without a device model

From: "Jan Beulich" <JBeulich@suse.com>
To: Roger Pau Monne <roger.pau@citrix.com>
Cc: Elena Ufimtseva <elena.ufimtseva@oracle.com>,
	Wei Liu <wei.liu2@citrix.com>,
	Ian Campbell <ian.campbell@citrix.com>,
	Stefano Stabellini <stefano.stabellini@eu.citrix.com>,
	Andrew Cooper <andrew.cooper3@citrix.com>,
	Ian Jackson <ian.jackson@eu.citrix.com>,
	xen-devel@lists.xenproject.org,
	Boris Ostrovsky <boris.ostrovsky@oracle.com>
Subject: Re: [PATCH RFC v1 11/13] xen/libxl: allow creating HVM guests without a device model
Date: Tue, 23 Jun 2015 10:41:07 +0100	[thread overview]
Message-ID: <558945D302000078000880A3@mail.emea.novell.com> (raw)
In-Reply-To: <1434989487-74940-12-git-send-email-roger.pau@citrix.com>

>>> On 22.06.15 at 18:11, <roger.pau@citrix.com> wrote:
> --- a/xen/arch/x86/hvm/hvm.c
> +++ b/xen/arch/x86/hvm/hvm.c
> @@ -343,7 +343,7 @@ u64 hvm_get_guest_tsc_adjust(struct vcpu *v)
>  void hvm_migrate_timers(struct vcpu *v)
>  {
>      /* PVH doesn't use rtc and emulated timers, it uses pvclock mechanism. */
> -    if ( is_pvh_vcpu(v) )
> +    if ( is_pvh_vcpu(v) || v->domain->arch.hvm_domain.no_emu )

Why would you need to keep the is-PVH check when you have ...

> @@ -1485,9 +1485,10 @@ int hvm_domain_initialise(struct domain *d)
>      else
>          d->arch.hvm_domain.io_bitmap = hvm_io_bitmap;
>  
> -    if ( is_pvh_domain(d) )
> +    if ( is_pvh_domain(d) || domcr_flags & DOMCRF_noemu )
>      {
>          register_portio_handler(d, 0, 0x10003, handle_pvh_io);
> +        d->arch.hvm_domain.no_emu = TRUE;

... this (which of course shouldn't use TRUE).

> @@ -2327,6 +2328,9 @@ int hvm_vcpu_initialise(struct vcpu *v)
>          return 0;
>      }
>  
> +    if ( d->arch.hvm_domain.no_emu )
> +        return 0;
> +
>      rc = setup_compat_arg_xlat(v); /* teardown: free_compat_arg_xlat() */

How are you going to get away without an argument translation
area?

Jan