Re: [PATCH v3 1/3] x86/ioreq server(patch for 4.7): Rename p2m_mmio_write_dm to p2m_ioreq_server.

From: "Yu, Zhang" <yu.c.zhang@linux.intel.com>
To: George Dunlap <george.dunlap@citrix.com>,
	Jan Beulich <JBeulich@suse.com>,
	Paul Durrant <Paul.Durrant@citrix.com>
Cc: Kevin Tian <kevin.tian@intel.com>, Wei Liu <wei.liu2@citrix.com>,
	Andrew Cooper <Andrew.Cooper3@citrix.com>,
	"Tim(Xen.org)" <tim@xen.org>,
	"xen-devel@lists.xen.org" <xen-devel@lists.xen.org>,
	Zhiyuan Lv <zhiyuan.lv@intel.com>,
	Jun Nakajima <jun.nakajima@intel.com>,
	"Keir (Xen.org)" <keir@xen.org>
Subject: Re: [PATCH v3 1/3] x86/ioreq server(patch for 4.7): Rename p2m_mmio_write_dm to p2m_ioreq_server.
Date: Tue, 26 Apr 2016 00:20:22 +0800	[thread overview]
Message-ID: <fce49ac6-9d43-cae7-30f7-36cb85dca4a3@linux.intel.com> (raw)
In-Reply-To: <571E42A7.9080503@citrix.com>

On 4/26/2016 12:15 AM, George Dunlap wrote:
> On 25/04/16 16:53, Yu, Zhang wrote:
>>
>>
>> On 4/25/2016 11:38 PM, Jan Beulich wrote:
>>>>>> On 25.04.16 at 17:29, <Paul.Durrant@citrix.com> wrote:
>>>>>  -----Original Message-----
>>>>> From: Yu, Zhang [mailto:yu.c.zhang@linux.intel.com]
>>>>> Sent: 25 April 2016 16:22
>>>>> To: Paul Durrant; George Dunlap
>>>>> Cc: xen-devel@lists.xen.org; Kevin Tian; Keir (Xen.org); Jun Nakajima;
>>>>> Andrew Cooper; Tim (Xen.org); Lv, Zhiyuan; Jan Beulich; Wei Liu
>>>>> Subject: Re: [Xen-devel] [PATCH v3 1/3] x86/ioreq server(patch for
>>>>> 4.7):
>>>>> Rename p2m_mmio_write_dm to p2m_ioreq_server.
>>>>>
>>>>>
>>>>>
>>>>> On 4/25/2016 10:01 PM, Paul Durrant wrote:
>>>>>>> -----Original Message-----
>>>>>>> From: dunlapg@gmail.com [mailto:dunlapg@gmail.com] On Behalf Of
>>>>>>> George Dunlap
>>>>>>> Sent: 25 April 2016 14:39
>>>>>>> To: Yu Zhang
>>>>>>> Cc: xen-devel@lists.xen.org; Kevin Tian; Keir (Xen.org); Jun
>>>>>>> Nakajima;
>>>>>>> Andrew Cooper; Tim (Xen.org); Paul Durrant; Lv, Zhiyuan; Jan
>>>>>>> Beulich; Wei
>>>>> Liu
>>>>>>> Subject: Re: [Xen-devel] [PATCH v3 1/3] x86/ioreq server(patch for
>>>>>>> 4.7):
>>>>>>> Rename p2m_mmio_write_dm to p2m_ioreq_server.
>>>>>>>
>>>>>>> On Mon, Apr 25, 2016 at 11:35 AM, Yu Zhang
>>>>> <yu.c.zhang@linux.intel.com>
>>>>>>> wrote:
>>>>>>>> Previously p2m type p2m_mmio_write_dm was introduced for write-
>>>>>>>> protected memory pages whose write operations are supposed to be
>>>>>>>> forwarded to and emulated by an ioreq server. Yet limitations of
>>>>>>>> rangeset restrict the number of guest pages to be write-protected.
>>>>>>>>
>>>>>>>> This patch replaces the p2m type p2m_mmio_write_dm with a new
>>>>> name:
>>>>>>>> p2m_ioreq_server, which means this p2m type can be claimed by one
>>>>>>>> ioreq server, instead of being tracked inside the rangeset of ioreq
>>>>>>>> server. Patches following up will add the related hvmop handling
>>>>>>>> code which map/unmap type p2m_ioreq_server to/from an ioreq
>>>>> server.
>>>>>>>>
>>>>>>>> changes in v3:
>>>>>>>>   - According to Jan & George's comments, keep
>>>>>>> HVMMEM_mmio_write_dm
>>>>>>>>     for old xen interface versions, and replace it with
>>>>>>>> HVMMEM_unused
>>>>>>>>     for xen interfaces newer than 4.7.0; For p2m_ioreq_server, a new
>>>>>>>>     enum - HVMMEM_ioreq_server is introduced for the get/set mem
>>>>> type
>>>>>>>>     interfaces;
>>>>>>>>   - Add George's Reviewed-by and Acked-by from Tim & Andrew.
>>>>>>>
>>>>>>> Unfortunately these rather contradict each other -- I consider
>>>>>>> Reviewed-by to only stick if the code I've specified hasn't changed
>>>>>>> (or has only changed trivially).
>>>>>>>
>>>>>>> Also...
>>>>>>>
>>>>>>>>
>>>>>>>> changes in v2:
>>>>>>>>   - According to George Dunlap's comments, only rename the p2m type,
>>>>>>>>     with no behavior changes.
>>>>>>>>
>>>>>>>> Signed-off-by: Paul Durrant <paul.durrant@citrix.com>
>>>>>>>> Signed-off-by: Yu Zhang <yu.c.zhang@linux.intel.com>
>>>>>>>> Acked-by: Tim Deegan <tim@xen.org>
>>>>>>>> Acked-by: Andrew Cooper <andrew.cooper3@citrix.com>
>>>>>>>> Reviewed-by: George Dunlap <george.dunlap@citrix.com>
>>>>>>>> Cc: Keir Fraser <keir@xen.org>
>>>>>>>> Cc: Jan Beulich <jbeulich@suse.com>
>>>>>>>> Cc: Andrew Cooper <andrew.cooper3@citrix.com>
>>>>>>>> Cc: Jun Nakajima <jun.nakajima@intel.com>
>>>>>>>> Cc: Kevin Tian <kevin.tian@intel.com>
>>>>>>>> Cc: George Dunlap <george.dunlap@eu.citrix.com>
>>>>>>>> Cc: Tim Deegan <tim@xen.org>
>>>>>>>> ---
>>>>>>>>  xen/arch/x86/hvm/hvm.c          | 14 ++++++++------
>>>>>>>>  xen/arch/x86/mm/p2m-ept.c       |  2 +-
>>>>>>>>  xen/arch/x86/mm/p2m-pt.c        |  2 +-
>>>>>>>>  xen/arch/x86/mm/shadow/multi.c  |  2 +-
>>>>>>>>  xen/include/asm-x86/p2m.h       |  4 ++--
>>>>>>>>  xen/include/public/hvm/hvm_op.h |  8 +++++++-
>>>>>>>>  6 files changed, 20 insertions(+), 12 deletions(-)
>>>>>>>>
>>>>>>>> diff --git a/xen/arch/x86/hvm/hvm.c b/xen/arch/x86/hvm/hvm.c
>>>>>>>> index f24126d..874cb0f 100644
>>>>>>>> --- a/xen/arch/x86/hvm/hvm.c
>>>>>>>> +++ b/xen/arch/x86/hvm/hvm.c
>>>>>>>> @@ -1857,7 +1857,7 @@ int hvm_hap_nested_page_fault(paddr_t gpa,
>>>>>>> unsigned long gla,
>>>>>>>>       */
>>>>>>>>      if ( (p2mt == p2m_mmio_dm) ||
>>>>>>>>           (npfec.write_access &&
>>>>>>>> -          (p2m_is_discard_write(p2mt) || (p2mt ==
>>>>> p2m_mmio_write_dm))) )
>>>>>>>> +          (p2m_is_discard_write(p2mt) || (p2mt ==
>>>>>>>> p2m_ioreq_server))) )
>>>>>>>>      {
>>>>>>>>          __put_gfn(p2m, gfn);
>>>>>>>>          if ( ap2m_active )
>>>>>>>> @@ -5499,8 +5499,8 @@ long do_hvm_op(unsigned long op,
>>>>>>> XEN_GUEST_HANDLE_PARAM(void) arg)
>>>>>>>>              get_gfn_query_unlocked(d, a.pfn, &t);
>>>>>>>>              if ( p2m_is_mmio(t) )
>>>>>>>>                  a.mem_type =  HVMMEM_mmio_dm;
>>>>>>>> -            else if ( t == p2m_mmio_write_dm )
>>>>>>>> -                a.mem_type = HVMMEM_mmio_write_dm;
>>>>>>>> +            else if ( t == p2m_ioreq_server )
>>>>>>>> +                a.mem_type = HVMMEM_ioreq_server;
>>>>>>>>              else if ( p2m_is_readonly(t) )
>>>>>>>>                  a.mem_type =  HVMMEM_ram_ro;
>>>>>>>>              else if ( p2m_is_ram(t) )
>>>>>>>> @@ -5531,7 +5531,8 @@ long do_hvm_op(unsigned long op,
>>>>>>> XEN_GUEST_HANDLE_PARAM(void) arg)
>>>>>>>>              [HVMMEM_ram_rw]  = p2m_ram_rw,
>>>>>>>>              [HVMMEM_ram_ro]  = p2m_ram_ro,
>>>>>>>>              [HVMMEM_mmio_dm] = p2m_mmio_dm,
>>>>>>>> -            [HVMMEM_mmio_write_dm] = p2m_mmio_write_dm
>>>>>>>> +            [HVMMEM_unused] = p2m_invalid,
>>>>>>>> +            [HVMMEM_ioreq_server] = p2m_ioreq_server
>>>>>>>>          };
>>>>>>>>
>>>>>>>>          if ( copy_from_guest(&a, arg, 1) )
>>>>>>>> @@ -5555,7 +5556,8 @@ long do_hvm_op(unsigned long op,
>>>>>>> XEN_GUEST_HANDLE_PARAM(void) arg)
>>>>>>>>               ((a.first_pfn + a.nr - 1) >
>>>>>>>> domain_get_maximum_gpfn(d)) )
>>>>>>>>              goto setmemtype_fail;
>>>>>>>>
>>>>>>>> -        if ( a.hvmmem_type >= ARRAY_SIZE(memtype) )
>>>>>>>> +        if ( a.hvmmem_type >= ARRAY_SIZE(memtype) ||
>>>>>>>> +             unlikely(a.hvmmem_type == HVMMEM_unused) )
>>>>>>>>              goto setmemtype_fail;
>>>>>>>>
>>>>>>>>          while ( a.nr > start_iter )
>>>>>>>> @@ -5579,7 +5581,7 @@ long do_hvm_op(unsigned long op,
>>>>>>> XEN_GUEST_HANDLE_PARAM(void) arg)
>>>>>>>>              }
>>>>>>>>              if ( !p2m_is_ram(t) &&
>>>>>>>>                   (!p2m_is_hole(t) || a.hvmmem_type !=
>>>>>>>> HVMMEM_mmio_dm)
>>>>>>> &&
>>>>>>>> -                 (t != p2m_mmio_write_dm || a.hvmmem_type !=
>>>>>>> HVMMEM_ram_rw) )
>>>>>>>> +                 (t != p2m_ioreq_server || a.hvmmem_type !=
>>>>>>> HVMMEM_ram_rw) )
>>>>>>>>              {
>>>>>>>>                  put_gfn(d, pfn);
>>>>>>>>                  goto setmemtype_fail;
>>>>>>>> diff --git a/xen/arch/x86/mm/p2m-ept.c b/xen/arch/x86/mm/p2m-
>>>>> ept.c
>>>>>>>> index 3cb6868..380ec25 100644
>>>>>>>> --- a/xen/arch/x86/mm/p2m-ept.c
>>>>>>>> +++ b/xen/arch/x86/mm/p2m-ept.c
>>>>>>>> @@ -171,7 +171,7 @@ static void ept_p2m_type_to_flags(struct
>>>>>>> p2m_domain *p2m, ept_entry_t *entry,
>>>>>>>>              entry->a = entry->d = !!cpu_has_vmx_ept_ad;
>>>>>>>>              break;
>>>>>>>>          case p2m_grant_map_ro:
>>>>>>>> -        case p2m_mmio_write_dm:
>>>>>>>> +        case p2m_ioreq_server:
>>>>>>>>              entry->r = 1;
>>>>>>>>              entry->w = entry->x = 0;
>>>>>>>>              entry->a = !!cpu_has_vmx_ept_ad;
>>>>>>>> diff --git a/xen/arch/x86/mm/p2m-pt.c b/xen/arch/x86/mm/p2m-pt.c
>>>>>>>> index 3d80612..eabd2e3 100644
>>>>>>>> --- a/xen/arch/x86/mm/p2m-pt.c
>>>>>>>> +++ b/xen/arch/x86/mm/p2m-pt.c
>>>>>>>> @@ -94,7 +94,7 @@ static unsigned long
>>>>> p2m_type_to_flags(p2m_type_t
>>>>>>> t, mfn_t mfn,
>>>>>>>>      default:
>>>>>>>>          return flags | _PAGE_NX_BIT;
>>>>>>>>      case p2m_grant_map_ro:
>>>>>>>> -    case p2m_mmio_write_dm:
>>>>>>>> +    case p2m_ioreq_server:
>>>>>>>>          return flags | P2M_BASE_FLAGS | _PAGE_NX_BIT;
>>>>>>>>      case p2m_ram_ro:
>>>>>>>>      case p2m_ram_logdirty:
>>>>>>>> diff --git a/xen/arch/x86/mm/shadow/multi.c
>>>>>>> b/xen/arch/x86/mm/shadow/multi.c
>>>>>>>> index e5c8499..c81302a 100644
>>>>>>>> --- a/xen/arch/x86/mm/shadow/multi.c
>>>>>>>> +++ b/xen/arch/x86/mm/shadow/multi.c
>>>>>>>> @@ -3225,7 +3225,7 @@ static int sh_page_fault(struct vcpu *v,
>>>>>>>>
>>>>>>>>      /* Need to hand off device-model MMIO to the device model */
>>>>>>>>      if ( p2mt == p2m_mmio_dm
>>>>>>>> -         || (p2mt == p2m_mmio_write_dm && ft == ft_demand_write) )
>>>>>>>> +         || (p2mt == p2m_ioreq_server && ft == ft_demand_write) )
>>>>>>>>      {
>>>>>>>>          gpa = guest_walk_to_gpa(&gw);
>>>>>>>>          goto mmio;
>>>>>>>> diff --git a/xen/include/asm-x86/p2m.h b/xen/include/asm-x86/p2m.h
>>>>>>>> index 5392eb0..ee2ea9c 100644
>>>>>>>> --- a/xen/include/asm-x86/p2m.h
>>>>>>>> +++ b/xen/include/asm-x86/p2m.h
>>>>>>>> @@ -71,7 +71,7 @@ typedef enum {
>>>>>>>>      p2m_ram_shared = 12,          /* Shared or sharable memory */
>>>>>>>>      p2m_ram_broken = 13,          /* Broken page, access cause
>>>>>>>> domain
>>>>> crash
>>>>>>> */
>>>>>>>>      p2m_map_foreign  = 14,        /* ram pages from foreign
>>>>>>>> domain */
>>>>>>>> -    p2m_mmio_write_dm = 15,       /* Read-only; writes go to the
>>>>>>>> device
>>>>>>> model */
>>>>>>>> +    p2m_ioreq_server = 15,
>>>>>>>>  } p2m_type_t;
>>>>>>>>
>>>>>>>>  /* Modifiers to the query */
>>>>>>>> @@ -112,7 +112,7 @@ typedef unsigned int p2m_query_t;
>>>>>>>>                        | p2m_to_mask(p2m_ram_ro)         \
>>>>>>>>                        | p2m_to_mask(p2m_grant_map_ro)   \
>>>>>>>>                        | p2m_to_mask(p2m_ram_shared)     \
>>>>>>>> -                      | p2m_to_mask(p2m_mmio_write_dm))
>>>>>>>> +                      | p2m_to_mask(p2m_ioreq_server))
>>>>>>>>
>>>>>>>>  /* Write-discard types, which should discard the write
>>>>>>>> operations */
>>>>>>>>  #define P2M_DISCARD_WRITE_TYPES (p2m_to_mask(p2m_ram_ro)     \
>>>>>>>> diff --git a/xen/include/public/hvm/hvm_op.h
>>>>>>> b/xen/include/public/hvm/hvm_op.h
>>>>>>>> index 1606185..b3e45cf 100644
>>>>>>>> --- a/xen/include/public/hvm/hvm_op.h
>>>>>>>> +++ b/xen/include/public/hvm/hvm_op.h
>>>>>>>> @@ -83,7 +83,13 @@ typedef enum {
>>>>>>>>      HVMMEM_ram_rw,             /* Normal read/write guest RAM */
>>>>>>>>      HVMMEM_ram_ro,             /* Read-only; writes are
>>>>>>>> discarded */
>>>>>>>>      HVMMEM_mmio_dm,            /* Reads and write go to the device
>>>>> model
>>>>>>> */
>>>>>>>> -    HVMMEM_mmio_write_dm       /* Read-only; writes go to the
>>>>>>>> device
>>>>>>> model */
>>>>>>>> +#if __XEN_INTERFACE_VERSION__ < 0x00040700
>>>>>>>> +    HVMMEM_mmio_write_dm,      /* Read-only; writes go to the
>>>>>>>> device
>>>>>>> model */
>>>>>>>> +#else
>>>>>>>> +    HVMMEM_unused,             /* Placeholder; setting memory to
>>>>>>>> this
>>>>> type
>>>>>>>> +                                  will fail for code after 4.7.0 */
>>>>>>>> +#endif
>>>>>>>> +    HVMMEM_ioreq_server
>>>>>>>
>>>>>>> Also, I don't think we've had a convincing argument for why this
>>>>>>> patch
>>>>>>> needs to be in 4.7.  The p2m name changes are internal only, and so
>>>>>>> don't need to be made at all; and the old functionality will work as
>>>>>>> well as it ever did.  Furthermore, the whole reason we're in this
>>>>>>> situation is that we checked in a publicly-visible change to the
>>>>>>> interface before it was properly ready; I think we should avoid
>>>>>>> making
>>>>>>> the same mistake this time.
>>>>>>>
>>>>>>> So personally I'd just leave this patch entirely for 4.8; but if Paul
>>>>>>> and/or Jan have strong opinions, then I would say check in only a
>>>>>>> patch to do the #if/#else/#endif, and leave both the p2m type changes
>>>>>>> and the new HVMMEM_ioreq_server enum for when the 4.8
>>>>> development
>>>>>>> window opens.
>>>>>>>
>>>>>>
>>>>>> If the whole series is going in then I think this patch is ok. If
>>>>>> this the
>>>> only
>>>>> patch that is going in for 4.7 then I thing we need the patch to
>>>>> hvm_op.h to
>>>>> deprecate the old type and that's it. We definitely should not
>>>>> introduce an
>>>>> implementation of the type HVMMEM_ioreq_server that has the same
>>>>> hardcoded semantics as the old type and then change it.
>>>>>> The p2m type changes are also wrong. That type needs to be left alone,
>>>>> presumably, so that anything using HVMMEM_mmio_write_dm and
>>>>> compiled to the old interface version continues to function. I think
>>>>> HVMMEM_ioreq_server needs to map to a new p2m type which should be
>>>>> introduced in patch #3.
>>>>>>
>>>>>
>>>>> Sorry, I'm also confused now. :(
>>>>>
>>>>> Do we really want to introduce a new p2m type? Why?
>>>>> My understanding of the previous agreement is that:
>>>>> 1> We need the interface to work on old hypervisor for
>>>>> HVMMEM_mmio_write_dm;
>>>>> 2> We need the interface to return -EINVAL for new hypervisor
>>>>> for HVMMEM_mmio_write_dm;
>>>>> 3> We need the new type HVMMEM_ioreq_server to work on new
>>>>> hypervisor;
>>>>>
>>>>> Did I miss something? Or I totally misunderstood?
>>>>>
>>>>
>>>> I don't know. I'm confused too. What we definitely don't want to do
>>>> is add a
>>>> new HVMMEM type and have it map to the old behaviour, otherwise we're no
>>>> better off.
>>>>
>>>> The question I'm not clear on the answer to is what happens to old code:
>>>>
>>>> Should it continue to compile? If so, should it continue to run.
>>>
>>> We only need to be concerned about the "get type" functionality,
>>> as that's the only thing an arbitrary guest can use. If the
>>> hypercall simply never returns the old type, then old code will
>>> still work (it'll just have some dead code on new Xen), and hence
>>> it compiling against the older interface is fine (and, from general
>>> considerations, a requirement).
>>>
>>
>> Thanks Jan. And I think the answer is yes. The new hypervisor will
>> only return HVMMEM_ioreq_server, which is a different value now.
>
> Right -- but we can't check in this entire series for 4.7; however, Paul
> would like to make it clear that HVMMEM_mmio_write_dm is deprecated; so
> we need a patch which does the #ifdef/#else/#endif clause, and then
> everything else necessary to make sure that things compile properly
> either way, but no p2m changes or additional functionality.
>

Thanks, George.
So what is it good for, with only the #ifdef/#els/#endif clause? It
seems more reasonable to me that we hold all 3 patches till next
version, and at that time the __XEN_INTERFACE_VERSION__ should be
compared against 0x000408000 in the #ifdef I guess?

Yu

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