Xen project Mailing List

On Tue, Jun 23, 2015 at 2:52 PM, Ed White <edmund.h.white@xxxxxxxxx> wrote:

On 06/23/2015 11:15 AM, Lengyel, Tamas wrote:
> On Mon, Jun 22, 2015 at 2:56 PM, Ed White <edmund.h.white@xxxxxxxxx> wrote:
>
>> Add the remaining routines required to support enabling the alternate
>> p2m functionality.
>>
>> Signed-off-by: Ed White <edmund.h.white@xxxxxxxxx>
>> ---
>> xen/arch/x86/hvm/hvm.c | 60 +++++-
>> xen/arch/x86/mm/hap/Makefile | 1 +
>> xen/arch/x86/mm/hap/altp2m_hap.c | 103 +++++++++
>> xen/arch/x86/mm/p2m-ept.c | 3 +
>> xen/arch/x86/mm/p2m.c | 405
>> ++++++++++++++++++++++++++++++++++++
>> xen/include/asm-x86/hvm/altp2mhvm.h | 4 +
>> xen/include/asm-x86/p2m.h | 33 +++
>> 7 files changed, 601 insertions(+), 8 deletions(-)
>> create mode 100644 xen/arch/x86/mm/hap/altp2m_hap.c
>>
>> diff --git a/xen/arch/x86/hvm/hvm.c b/xen/arch/x86/hvm/hvm.c
>> index d75c12d..b758ee1 100644
>> --- a/xen/arch/x86/hvm/hvm.c
>> +++ b/xen/arch/x86/hvm/hvm.c
>> @@ -2786,10 +2786,11 @@ int hvm_hap_nested_page_fault(paddr_t gpa,
>> unsigned long gla,
>> p2m_access_t p2ma;
>> mfn_t mfn;
>> struct vcpu *v = current;
>> - struct p2m_domain *p2m;
>> + struct p2m_domain *p2m, *hostp2m;
>> int rc, fall_through = 0, paged = 0;
>> int sharing_enomem = 0;
>> vm_event_request_t *req_ptr = NULL;
>> + int altp2m_active = 0;
>>
>> /* On Nested Virtualization, walk the guest page table.
>> * If this succeeds, all is fine.
>> @@ -2845,15 +2846,33 @@ int hvm_hap_nested_page_fault(paddr_t gpa,
>> unsigned long gla,
>> {
>> if ( !handle_mmio_with_translation(gla, gpa >> PAGE_SHIFT, npfec)
>> )
>> hvm_inject_hw_exception(TRAP_gp_fault, 0);
>> - rc = 1;
>> - goto out;
>> + return 1;
>> }
>>
>> - p2m = p2m_get_hostp2m(v->domain);
>> - mfn = get_gfn_type_access(p2m, gfn, &p2mt, &p2ma,
>> + altp2m_active = altp2mhvm_active(v->domain);
>> +
>> + /* Take a lock on the host p2m speculatively, to avoid potential
>> + * locking order problems later and to handle unshare etc.
>> + */
>> + hostp2m = p2m_get_hostp2m(v->domain);
>> + mfn = get_gfn_type_access(hostp2m, gfn, &p2mt, &p2ma,
>> P2M_ALLOC | (npfec.write_access ?
>> P2M_UNSHARE : 0),
>> NULL);
>>
>> + if ( altp2m_active )
>> + {
>> + if ( altp2mhvm_hap_nested_page_fault(v, gpa, gla, npfec, &p2m) ==
>> 1 )
>> + {
>> + /* entry was lazily copied from host -- retry */
>> + __put_gfn(hostp2m, gfn);
>> + return 1;
>> + }
>> +
>> + mfn = get_gfn_type_access(p2m, gfn, &p2mt, &p2ma, 0, NULL);
>> + }
>> + else
>> + p2m = hostp2m;
>> +
>> /* Check access permissions first, then handle faults */
>> if ( mfn_x(mfn) != INVALID_MFN )
>> {
>> @@ -2893,6 +2912,20 @@ int hvm_hap_nested_page_fault(paddr_t gpa, unsigned
>> long gla,
>>
>> if ( violation )
>> {
>> + /* Should #VE be emulated for this fault? */
>> + if ( p2m_is_altp2m(p2m) && !cpu_has_vmx_virt_exceptions )
>> + {
>> + unsigned int sve;
>> +
>> + p2m->get_entry_full(p2m, gfn, &p2mt, &p2ma, 0, NULL,
>> &sve);
>> +
>> + if ( !sve && ahvm_vcpu_emulate_ve(v) )
>>
>
> This line generates the following compile-time error: "hvm.c:2923:51:
> error: ‘v’ undeclared (first use in this function)". Did you mean to pass
> curr to ahvm_vcpu_emulate_ve instead of v?
>
> I would recommend doing a compile-test on each patch of the series to catch
> small things like this. Travis-ci has been working really great for me to
> automate that process (https://github.com/tklengyel/xen/compare/travis) ;)
>

I don't know why you are seeing that error, you can clearly see that v is
defined and initialised at the start of the containing function, and has
been used earlier.

I always compile test every patch individually.

Ed

Was applying the series to the latest master branch. Things have apparently changed enough in that frame to cause breakage =) Sorry for the noise.

www.novetta.com

Tamas K Lengyel

Senior Security Researcher

7921 Jones Branch Drive

McLean VA 22102

Email tlengyel@novetta.com

Re: [Xen-devel] [PATCH v2 09/12] x86/altp2m: add remaining support routines.