Re: [PATCH] x86/hvm: Drop more remains of the PVHv1 implementation

[Andrew Cooper <andrew.cooper3@citrix.com>]

On 19/07/17 15:12, Roger Pau Monné wrote:
> On Wed, Jul 19, 2017 at 02:27:31PM +0100, Andrew Cooper wrote:
>> These functions don't need is_hvm_{vcpu,domain}() predicates.
>>
>> hvmop_set_evtchn_upcall_vector() does need the predicate to prevent a PV
>> caller accessing the hvm union, but swap the copy_from_guest() and
>> is_hvm_domain() predicate to avoid reading the hypercall parameter if we not
>> going to use it.
> IC, certain HVMOPs are available to PV guests (ie: the control domain).

At the very least, the control domain needs to use HVMOP_getparam for
construction and migration purposes.  As a result, PV guests have always
had blanket reign on HVMOPs.

>
>> Signed-off-by: Andrew Cooper <andrew.cooper3@citrix.com>
> Reviewed-by: Roger Pau Monné <roger.pau@citrix.com>
>
> Thanks. Just one style nit.
>
>> CC: George Dunlap <George.Dunlap@eu.citrix.com>
>> CC: Jan Beulich <JBeulich@suse.com>
>> CC: Wei Liu <wei.liu2@citrix.com>
>> CC: Paul Durrant <paul.durrant@citrix.com>
>> CC: Roger Pau Monné <roger.pau@citrix.com>
>> ---
>>  xen/arch/x86/hvm/hvm.c | 15 ++++++---------
>>  1 file changed, 6 insertions(+), 9 deletions(-)
>>
>> diff --git a/xen/arch/x86/hvm/hvm.c b/xen/arch/x86/hvm/hvm.c
>> index 8145385..4fef616 100644
>> --- a/xen/arch/x86/hvm/hvm.c
>> +++ b/xen/arch/x86/hvm/hvm.c
>> @@ -506,8 +506,7 @@ void hvm_do_resume(struct vcpu *v)
>>  {
>>      check_wakeup_from_wait();
>>  
>> -    if ( is_hvm_domain(v->domain) )
>> -        pt_restore_timer(v);
>> +    pt_restore_timer(v);
>>  
>>      if ( !handle_hvm_io_completion(v) )
>>          return;
>> @@ -1544,8 +1543,7 @@ void hvm_vcpu_destroy(struct vcpu *v)
>>      tasklet_kill(&v->arch.hvm_vcpu.assert_evtchn_irq_tasklet);
>>      hvm_funcs.vcpu_destroy(v);
>>  
>> -    if ( is_hvm_vcpu(v) )
>> -        vlapic_destroy(v);
>> +    vlapic_destroy(v);
>>  
>>      hvm_vcpu_cacheattr_destroy(v);
>>  }
>> @@ -1711,7 +1709,6 @@ int hvm_hap_nested_page_fault(paddr_t gpa, unsigned long gla,
>>       * - newer Windows (like Server 2012) for HPET accesses.
>>       */
>>      if ( !nestedhvm_vcpu_in_guestmode(curr)
>> -         && is_hvm_domain(currd)
>>           && hvm_mmio_internal(gpa) )
> Can this be moved to the previous line?

Will fix on commit.

~Andrew

_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xen.org
https://lists.xen.org/xen-devel