[PATCH 0/7] xen: Convert __page_to_mfn and __mfn_to_page to use typesafe MFN

[PATCH 0/7] xen: Convert __page_to_mfn and __mfn_to_page to use typesafe MFN

[Julien Grall]

Hi all,

Most of the users of page_to_mfn and mfn_to_page are either overriding
the macros to make them work with mfn_t or use mfn_x/_mfn becaue the rest
of the function use mfn_t.

So I think it is time to make __page_to_mfn and __mfn_to_page using typesafe
MFN.

The first 6 patches will convert of the code to use typesafe MFN, easing
the tree-wide conversion in patch 7.

Cheers,

Cc:  Andrew Cooper <andrew.cooper3@citrix.com>
Cc:  Boris Ostrovsky <boris.ostrovsky@oracle.com>
Cc:  Elena Ufimtseva <elena.ufimtseva@oracle.com>
Cc:  Gang Wei <gang.wei@intel.com>
Cc:  George Dunlap <george.dunlap@eu.citrix.com>
Cc:  George Dunlap <George.Dunlap@eu.citrix.com>
Cc:  Ian Jackson <ian.jackson@eu.citrix.com>
Cc:  Jan Beulich <jbeulich@suse.com>
Cc:  Julien Grall <julien.grall@arm.com>
Cc:  Jun Nakajima <jun.nakajima@intel.com>
Cc:  Kevin Tian <kevin.tian@intel.com>
Cc:  Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>
Cc:  Paul Durrant <paul.durrant@citrix.com>
Cc:  Razvan Cojocaru <rcojocaru@bitdefender.com>
Cc:  Shane Wang <shane.wang@intel.com>
Cc:  Stefano Stabellini <sstabellini@kernel.org>
Cc:  Suravee Suthikulpanit <suravee.suthikulpanit@amd.com>
Cc:  Tamas K Lengyel <tamas@tklengyel.com>
Cc:  Tim Deegan <tim@xen.org>
Cc:  Wei Liu <wei.liu2@citrix.com>

Julien Grall (7):
  xen/arm: domain_build: Clean-up insert_11_bank
  xen/arm32: mm: Rework is_xen_heap_page to avoid nameclash
  xen/x86: Use maddr_to_page and maddr_to_mfn to avoid open-coded >>
    PAGE_SHIFT
  xen/kimage: Remove defined but unused variables
  xen/xenoprof: Convert the file to use typesafe MFN
  xen/tmem: Convert the file common/tmem_xen.c to use typesafe MFN
  xen: Convert __page_to_mfn and __mfn_to_page to use typesafe MFN

 xen/arch/arm/domain_build.c             | 15 ++++++++-------
 xen/arch/arm/kernel.c                   |  2 +-
 xen/arch/arm/mem_access.c               |  2 +-
 xen/arch/arm/mm.c                       |  2 +-
 xen/arch/arm/p2m.c                      |  8 +-------
 xen/arch/x86/cpu/vpmu.c                 |  6 +++---
 xen/arch/x86/debug.c                    |  2 +-
 xen/arch/x86/domain.c                   | 21 +++++++++++----------
 xen/arch/x86/domctl.c                   |  2 +-
 xen/arch/x86/hvm/dm.c                   |  2 +-
 xen/arch/x86/hvm/dom0_build.c           |  6 +++---
 xen/arch/x86/hvm/hvm.c                  | 16 ++++++++--------
 xen/arch/x86/hvm/ioreq.c                |  6 +++---
 xen/arch/x86/hvm/stdvga.c               |  2 +-
 xen/arch/x86/hvm/svm/svm.c              |  4 ++--
 xen/arch/x86/hvm/viridian.c             |  8 ++++----
 xen/arch/x86/hvm/vmx/vmcs.c             |  2 +-
 xen/arch/x86/hvm/vmx/vmx.c              | 10 +++++-----
 xen/arch/x86/hvm/vmx/vvmx.c             |  2 +-
 xen/arch/x86/mm.c                       |  6 ------
 xen/arch/x86/mm/guest_walk.c            |  6 +++---
 xen/arch/x86/mm/hap/guest_walk.c        |  2 +-
 xen/arch/x86/mm/hap/hap.c               |  6 ------
 xen/arch/x86/mm/hap/nested_ept.c        |  2 +-
 xen/arch/x86/mm/mem_sharing.c           |  9 ++-------
 xen/arch/x86/mm/p2m-ept.c               |  4 ++++
 xen/arch/x86/mm/p2m-pod.c               |  6 ------
 xen/arch/x86/mm/p2m.c                   |  6 ------
 xen/arch/x86/mm/paging.c                |  6 ------
 xen/arch/x86/mm/shadow/common.c         |  2 +-
 xen/arch/x86/mm/shadow/multi.c          |  6 +++---
 xen/arch/x86/mm/shadow/private.h        | 16 ++--------------
 xen/arch/x86/numa.c                     |  2 +-
 xen/arch/x86/physdev.c                  |  2 +-
 xen/arch/x86/pv/callback.c              |  6 ------
 xen/arch/x86/pv/descriptor-tables.c     |  6 ------
 xen/arch/x86/pv/dom0_build.c            |  6 ++++++
 xen/arch/x86/pv/domain.c                |  6 ------
 xen/arch/x86/pv/emul-gate-op.c          |  6 ------
 xen/arch/x86/pv/emul-priv-op.c          | 10 ----------
 xen/arch/x86/pv/grant_table.c           |  6 ------
 xen/arch/x86/pv/ro-page-fault.c         |  6 ------
 xen/arch/x86/smpboot.c                  |  6 ------
 xen/arch/x86/tboot.c                    |  4 ++--
 xen/arch/x86/traps.c                    |  2 +-
 xen/arch/x86/x86_64/mm.c                |  6 ++++++
 xen/common/domain.c                     |  4 ++--
 xen/common/event_fifo.c                 |  2 +-
 xen/common/grant_table.c                |  6 ++++++
 xen/common/kimage.c                     | 25 +++++++++++--------------
 xen/common/memory.c                     |  6 ++++++
 xen/common/page_alloc.c                 |  6 ++++++
 xen/common/tmem.c                       |  2 +-
 xen/common/tmem_xen.c                   | 24 ++++++++++++------------
 xen/common/trace.c                      |  6 ++++++
 xen/common/vmap.c                       |  9 +++++----
 xen/common/xenoprof.c                   | 17 +++++++++++------
 xen/drivers/passthrough/amd/iommu_map.c |  6 ++++++
 xen/drivers/passthrough/iommu.c         |  2 +-
 xen/drivers/passthrough/x86/iommu.c     |  2 +-
 xen/include/asm-arm/mm.h                | 22 ++++++++++++----------
 xen/include/asm-arm/p2m.h               |  4 ++--
 xen/include/asm-x86/mm.h                | 12 ++++++------
 xen/include/asm-x86/p2m.h               |  2 +-
 xen/include/asm-x86/page.h              | 32 ++++++++++++++++----------------
 xen/include/xen/domain_page.h           |  4 ++--
 xen/include/xen/tmem_xen.h              |  2 +-
 67 files changed, 206 insertions(+), 258 deletions(-)

-- 
2.11.0


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

[PATCH 1/7] xen/arm: domain_build: Clean-up insert_11_bank

[Julien Grall]

    - Remove spurious ()
    - Add missing spaces
    - Turn 1 << to 1UL <<

Signed-off-by: Julien Grall <julien.grall@linaro.org>

---

Cc: Stefano Stabellini <sstabellini@kernel.org>
---
 xen/arch/arm/domain_build.c | 7 ++++---
 1 file changed, 4 insertions(+), 3 deletions(-)

diff --git a/xen/arch/arm/domain_build.c b/xen/arch/arm/domain_build.c
index 3723dc3f78..093ebf1a8e 100644
--- a/xen/arch/arm/domain_build.c
+++ b/xen/arch/arm/domain_build.c
@@ -109,11 +109,11 @@ static bool insert_11_bank(struct domain *d,
 
     spfn = page_to_mfn(pg);
     start = pfn_to_paddr(spfn);
-    size = pfn_to_paddr((1 << order));
+    size = pfn_to_paddr(1UL << order);
 
     D11PRINT("Allocated %#"PRIpaddr"-%#"PRIpaddr" (%ldMB/%ldMB, order %d)\n",
              start, start + size,
-             1UL << (order+PAGE_SHIFT-20),
+             1UL << (order + PAGE_SHIFT -  20),
              /* Don't want format this as PRIpaddr (16 digit hex) */
              (unsigned long)(kinfo->unassigned_mem >> 20),
              order);
@@ -167,7 +167,8 @@ static bool insert_11_bank(struct domain *d,
          */
         if ( start + size < bank->start && kinfo->mem.nr_banks < NR_MEM_BANKS )
         {
-            memmove(bank + 1, bank, sizeof(*bank)*(kinfo->mem.nr_banks - i));
+            memmove(bank + 1, bank,
+                    sizeof(*bank) * (kinfo->mem.nr_banks - i));
             kinfo->mem.nr_banks++;
             bank->start = start;
             bank->size = size;
-- 
2.11.0


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

Re: [PATCH 1/7] xen/arm: domain_build: Clean-up insert_11_bank

[Andrew Cooper]

On 04/10/2017 19:15, Julien Grall wrote:
> diff --git a/xen/arch/arm/domain_build.c b/xen/arch/arm/domain_build.c
> index 3723dc3f78..093ebf1a8e 100644
> --- a/xen/arch/arm/domain_build.c
> +++ b/xen/arch/arm/domain_build.c
> @@ -109,11 +109,11 @@ static bool insert_11_bank(struct domain *d,
>  
>      spfn = page_to_mfn(pg);
>      start = pfn_to_paddr(spfn);
> -    size = pfn_to_paddr((1 << order));
> +    size = pfn_to_paddr(1UL << order);
>  
>      D11PRINT("Allocated %#"PRIpaddr"-%#"PRIpaddr" (%ldMB/%ldMB, order %d)\n",
>               start, start + size,
> -             1UL << (order+PAGE_SHIFT-20),
> +             1UL << (order + PAGE_SHIFT -  20),

If you are looking to be picky, you've got a double space between the
minus and the 20.  I'm sure this would be trivial to fix on commit.

~Andrew

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

Re: [PATCH 1/7] xen/arm: domain_build: Clean-up insert_11_bank

[Julien Grall]

Hi Andrew,

On 04/10/17 23:39, Andrew Cooper wrote:
> On 04/10/2017 19:15, Julien Grall wrote:
>> diff --git a/xen/arch/arm/domain_build.c b/xen/arch/arm/domain_build.c
>> index 3723dc3f78..093ebf1a8e 100644
>> --- a/xen/arch/arm/domain_build.c
>> +++ b/xen/arch/arm/domain_build.c
>> @@ -109,11 +109,11 @@ static bool insert_11_bank(struct domain *d,
>>   
>>       spfn = page_to_mfn(pg);
>>       start = pfn_to_paddr(spfn);
>> -    size = pfn_to_paddr((1 << order));
>> +    size = pfn_to_paddr(1UL << order);
>>   
>>       D11PRINT("Allocated %#"PRIpaddr"-%#"PRIpaddr" (%ldMB/%ldMB, order %d)\n",
>>                start, start + size,
>> -             1UL << (order+PAGE_SHIFT-20),
>> +             1UL << (order + PAGE_SHIFT -  20),
> 
> If you are looking to be picky, you've got a double space between the
> minus and the 20.  I'm sure this would be trivial to fix on commit.

Argh, on my original patch I had 2 spaces before the minus. Dropped one 
before sending on xen-devel and didn't spot the one after the minus.

I will resend the series with your comments addressed.

Cheers,

-- 
Julien Grall

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

[PATCH 2/7] xen/arm32: mm: Rework is_xen_heap_page to avoid nameclash

[Julien Grall]

The arm32 version of the function is_xen_heap_page currently define a
variable _mfn. This will lead to a compiler when use typesafe MFN in a
folow-up patch:

called object '_mfn' is not a function or function pointer

Fix it by renaming the local variable _mfn to mfn_.

Signed-off-by: Julien Grall <julien.grall@linaro.org>

---

Cc: Stefano Stabellini <sstabellini@kernel.org>
---
 xen/include/asm-arm/mm.h | 6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)

diff --git a/xen/include/asm-arm/mm.h b/xen/include/asm-arm/mm.h
index cd6dfb54b9..737a429409 100644
--- a/xen/include/asm-arm/mm.h
+++ b/xen/include/asm-arm/mm.h
@@ -140,9 +140,9 @@ extern vaddr_t xenheap_virt_start;
 #ifdef CONFIG_ARM_32
 #define is_xen_heap_page(page) is_xen_heap_mfn(page_to_mfn(page))
 #define is_xen_heap_mfn(mfn) ({                                 \
-    unsigned long _mfn = (mfn);                                 \
-    (_mfn >= mfn_x(xenheap_mfn_start) &&                        \
-     _mfn < mfn_x(xenheap_mfn_end));                            \
+    unsigned long mfn_ = (mfn);                                 \
+    (mfn_ >= mfn_x(xenheap_mfn_start) &&                        \
+     mfn_ < mfn_x(xenheap_mfn_end));                            \
 })
 #else
 #define is_xen_heap_page(page) ((page)->count_info & PGC_xen_heap)
-- 
2.11.0


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

[PATCH 3/7] xen/x86: Use maddr_to_page and maddr_to_mfn to avoid open-coded >> PAGE_SHIFT

[Julien Grall]

The constructions _mfn(... > PAGE_SHIFT) and mfn_to_page(... >> PAGE_SHIFT)
could respectively be replaced by maddr_to_mfn(...) and
maddr_to_page(...).

Signed-off-by: Julien Grall <julien.grall@linaro.org>

---

Cc: Elena Ufimtseva <elena.ufimtseva@oracle.com>
Cc: Jan Beulich <jbeulich@suse.com>
Cc: Andrew Cooper <andrew.cooper3@citrix.com>
Cc: Tim Deegan <tim@xen.org>
Cc: George Dunlap <george.dunlap@eu.citrix.com>
---
 xen/arch/x86/debug.c            | 2 +-
 xen/arch/x86/mm/shadow/common.c | 2 +-
 xen/arch/x86/mm/shadow/multi.c  | 6 +++---
 xen/common/kimage.c             | 6 +++---
 4 files changed, 8 insertions(+), 8 deletions(-)

diff --git a/xen/arch/x86/debug.c b/xen/arch/x86/debug.c
index 1c10b84a16..9159f32db4 100644
--- a/xen/arch/x86/debug.c
+++ b/xen/arch/x86/debug.c
@@ -98,7 +98,7 @@ dbg_pv_va2mfn(dbgva_t vaddr, struct domain *dp, uint64_t pgd3val)
     l2_pgentry_t l2e, *l2t;
     l1_pgentry_t l1e, *l1t;
     unsigned long cr3 = (pgd3val ? pgd3val : dp->vcpu[0]->arch.cr3);
-    mfn_t mfn = _mfn(cr3 >> PAGE_SHIFT);
+    mfn_t mfn = maddr_to_mfn(cr3);
 
     DBGP2("vaddr:%lx domid:%d cr3:%lx pgd3:%lx\n", vaddr, dp->domain_id, 
           cr3, pgd3val);
diff --git a/xen/arch/x86/mm/shadow/common.c b/xen/arch/x86/mm/shadow/common.c
index 86186cccdf..f65d2a6523 100644
--- a/xen/arch/x86/mm/shadow/common.c
+++ b/xen/arch/x86/mm/shadow/common.c
@@ -2640,7 +2640,7 @@ static int sh_remove_shadow_via_pointer(struct domain *d, mfn_t smfn)
     ASSERT(sh_type_has_up_pointer(d, sp->u.sh.type));
 
     if (sp->up == 0) return 0;
-    pmfn = _mfn(sp->up >> PAGE_SHIFT);
+    pmfn = maddr_to_mfn(sp->up);
     ASSERT(mfn_valid(pmfn));
     vaddr = map_domain_page(pmfn);
     ASSERT(vaddr);
diff --git a/xen/arch/x86/mm/shadow/multi.c b/xen/arch/x86/mm/shadow/multi.c
index 28030acbf6..1e42e1d8ab 100644
--- a/xen/arch/x86/mm/shadow/multi.c
+++ b/xen/arch/x86/mm/shadow/multi.c
@@ -2425,7 +2425,7 @@ int sh_safe_not_to_sync(struct vcpu *v, mfn_t gl1mfn)
     sp = mfn_to_page(smfn);
     if ( sp->u.sh.count != 1 || !sp->up )
         return 0;
-    smfn = _mfn(sp->up >> PAGE_SHIFT);
+    smfn = maddr_to_mfn(sp->up);
     ASSERT(mfn_valid(smfn));
 
 #if (SHADOW_PAGING_LEVELS == 4)
@@ -2434,7 +2434,7 @@ int sh_safe_not_to_sync(struct vcpu *v, mfn_t gl1mfn)
     ASSERT(sh_type_has_up_pointer(d, SH_type_l2_shadow));
     if ( sp->u.sh.count != 1 || !sp->up )
         return 0;
-    smfn = _mfn(sp->up >> PAGE_SHIFT);
+    smfn = maddr_to_mfn(sp->up);
     ASSERT(mfn_valid(smfn));
 
     /* up to l4 */
@@ -2442,7 +2442,7 @@ int sh_safe_not_to_sync(struct vcpu *v, mfn_t gl1mfn)
     if ( sp->u.sh.count != 1
          || !sh_type_has_up_pointer(d, SH_type_l3_64_shadow) || !sp->up )
         return 0;
-    smfn = _mfn(sp->up >> PAGE_SHIFT);
+    smfn = maddr_to_mfn(sp->up);
     ASSERT(mfn_valid(smfn));
 #endif
 
diff --git a/xen/common/kimage.c b/xen/common/kimage.c
index cf624d10fd..ebc71affd1 100644
--- a/xen/common/kimage.c
+++ b/xen/common/kimage.c
@@ -504,7 +504,7 @@ static void kimage_free_entry(kimage_entry_t entry)
 {
     struct page_info *page;
 
-    page = mfn_to_page(entry >> PAGE_SHIFT);
+    page = maddr_to_page(entry);
     free_domheap_page(page);
 }
 
@@ -636,8 +636,8 @@ static struct page_info *kimage_alloc_page(struct kexec_image *image,
         if ( old )
         {
             /* If so move it. */
-            mfn_t old_mfn = _mfn(*old >> PAGE_SHIFT);
-            mfn_t mfn = _mfn(addr >> PAGE_SHIFT);
+            mfn_t old_mfn = maddr_to_mfn(*old);
+            mfn_t mfn = maddr_to_mfn(addr);
 
             copy_domain_page(mfn, old_mfn);
             clear_domain_page(old_mfn);
-- 
2.11.0


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

Re: [PATCH 3/7] xen/x86: Use maddr_to_page and maddr_to_mfn to avoid open-coded >> PAGE_SHIFT

[Andrew Cooper]

On 04/10/2017 19:15, Julien Grall wrote:
> The constructions _mfn(... > PAGE_SHIFT) and mfn_to_page(... >> PAGE_SHIFT)
> could respectively be replaced by maddr_to_mfn(...) and
> maddr_to_page(...).
>
> Signed-off-by: Julien Grall <julien.grall@linaro.org>

Reviewed-by: Andrew Cooper <andrew.cooper3@citrix.com>

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

[PATCH 4/7] xen/kimage: Remove defined but unused variables

[Julien Grall]

In the function kimage_alloc_normal_control_page, the variables mfn and
emfn are defined but not used. Remove them.

Signed-off-by: Julien Grall <julien.grall@linaro.org>

---

Cc: Andrew Cooper <andrew.cooper3@citrix.com>
---
 xen/common/kimage.c | 3 ---
 1 file changed, 3 deletions(-)

diff --git a/xen/common/kimage.c b/xen/common/kimage.c
index ebc71affd1..07587896a4 100644
--- a/xen/common/kimage.c
+++ b/xen/common/kimage.c
@@ -310,14 +310,11 @@ static struct page_info *kimage_alloc_normal_control_page(
      * destination page.
      */
     do {
-        unsigned long mfn, emfn;
         paddr_t addr, eaddr;
 
         page = kimage_alloc_zeroed_page(memflags);
         if ( !page )
             break;
-        mfn   = page_to_mfn(page);
-        emfn  = mfn + 1;
         addr  = page_to_maddr(page);
         eaddr = addr + PAGE_SIZE;
         if ( kimage_is_destination_range(image, addr, eaddr) )
-- 
2.11.0


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

Re: [PATCH 4/7] xen/kimage: Remove defined but unused variables

[Andrew Cooper]

On 04/10/2017 19:15, Julien Grall wrote:
> In the function kimage_alloc_normal_control_page, the variables mfn and
> emfn are defined but not used. Remove them.
>
> Signed-off-by: Julien Grall <julien.grall@linaro.org>

Oops.

Reviewed-by: Andrew Cooper <andrew.cooper3@citrix.com>

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

[PATCH 5/7] xen/xenoprof: Convert the file to use typesafe MFN

[Julien Grall]

The file common/xenoprof.c is now converted to use typesafe. This is
requiring to override the macros virt_to_mfn and mfn_to_page to make
them work with mfn_t.

Signed-off-by: Julien Grall <julien.grall@linaro.org>

---

Cc: Andrew Cooper <andrew.cooper3@citrix.com>
Cc: George Dunlap <George.Dunlap@eu.citrix.com>
Cc: Ian Jackson <ian.jackson@eu.citrix.com>
Cc: Jan Beulich <jbeulich@suse.com>
Cc: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>
Cc: Stefano Stabellini <sstabellini@kernel.org>
Cc: Tim Deegan <tim@xen.org>
Cc: Wei Liu <wei.liu2@citrix.com>
---
 xen/common/xenoprof.c | 19 +++++++++++++------
 1 file changed, 13 insertions(+), 6 deletions(-)

diff --git a/xen/common/xenoprof.c b/xen/common/xenoprof.c
index a5fe6204a5..98937c9ac6 100644
--- a/xen/common/xenoprof.c
+++ b/xen/common/xenoprof.c
@@ -19,6 +19,12 @@
 #include <xsm/xsm.h>
 #include <xen/hypercall.h>
 
+/* Override macros from asm/page.h to make them work with mfn_t */
+#undef virt_to_mfn
+#define virt_to_mfn(va) _mfn(__virt_to_mfn(va))
+#undef mfn_to_page
+#define mfn_to_page(mfn) __mfn_to_page(mfn_x(mfn))
+
 /* Limit amount of pages used for shared buffer (per domain) */
 #define MAX_OPROF_SHARED_PAGES 32
 
@@ -134,25 +140,26 @@ static void xenoprof_reset_buf(struct domain *d)
 }
 
 static int
-share_xenoprof_page_with_guest(struct domain *d, unsigned long mfn, int npages)
+share_xenoprof_page_with_guest(struct domain *d, mfn_t mfn, int npages)
 {
     int i;
 
     /* Check if previous page owner has released the page. */
     for ( i = 0; i < npages; i++ )
     {
-        struct page_info *page = mfn_to_page(mfn + i);
+        struct page_info *page = mfn_to_page(mfn_add(mfn, i));
         if ( (page->count_info & (PGC_allocated|PGC_count_mask)) != 0 )
         {
             printk(XENLOG_G_INFO "dom%d mfn %#lx page->count_info %#lx\n",
-                   d->domain_id, mfn + i, page->count_info);
+                   d->domain_id, mfn_x(mfn_add(mfn, i)), page->count_info);
             return -EBUSY;
         }
         page_set_owner(page, NULL);
     }
 
     for ( i = 0; i < npages; i++ )
-        share_xen_page_with_guest(mfn_to_page(mfn + i), d, XENSHARE_writable);
+        share_xen_page_with_guest(mfn_to_page(mfn_add(mfn, i)),
+                                  d, XENSHARE_writable);
 
     return 0;
 }
@@ -161,11 +168,11 @@ static void
 unshare_xenoprof_page_with_guest(struct xenoprof *x)
 {
     int i, npages = x->npages;
-    unsigned long mfn = virt_to_mfn(x->rawbuf);
+    mfn_t mfn = virt_to_mfn(x->rawbuf);
 
     for ( i = 0; i < npages; i++ )
     {
-        struct page_info *page = mfn_to_page(mfn + i);
+        struct page_info *page = mfn_to_page(mfn_add(mfn, i));
         BUG_ON(page_get_owner(page) != current->domain);
         if ( test_and_clear_bit(_PGC_allocated, &page->count_info) )
             put_page(page);
-- 
2.11.0


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

Re: [PATCH 5/7] xen/xenoprof: Convert the file to use typesafe MFN

[Andrew Cooper]

On 04/10/2017 19:15, Julien Grall wrote:
> @@ -134,25 +140,26 @@ static void xenoprof_reset_buf(struct domain *d)
>  }
>  
>  static int
> -share_xenoprof_page_with_guest(struct domain *d, unsigned long mfn, int npages)
> +share_xenoprof_page_with_guest(struct domain *d, mfn_t mfn, int npages)
>  {
>      int i;
>  
>      /* Check if previous page owner has released the page. */
>      for ( i = 0; i < npages; i++ )
>      {
> -        struct page_info *page = mfn_to_page(mfn + i);
> +        struct page_info *page = mfn_to_page(mfn_add(mfn, i));

A newline would be nice here...

>          if ( (page->count_info & (PGC_allocated|PGC_count_mask)) != 0 )
>          {
>              printk(XENLOG_G_INFO "dom%d mfn %#lx page->count_info %#lx\n",
> -                   d->domain_id, mfn + i, page->count_info);
> +                   d->domain_id, mfn_x(mfn_add(mfn, i)), page->count_info);
>              return -EBUSY;
>          }
>          page_set_owner(page, NULL);
>      }
>  
>      for ( i = 0; i < npages; i++ )
> -        share_xen_page_with_guest(mfn_to_page(mfn + i), d, XENSHARE_writable);
> +        share_xen_page_with_guest(mfn_to_page(mfn_add(mfn, i)),
> +                                  d, XENSHARE_writable);
>  
>      return 0;
>  }
> @@ -161,11 +168,11 @@ static void
>  unshare_xenoprof_page_with_guest(struct xenoprof *x)
>  {
>      int i, npages = x->npages;
> -    unsigned long mfn = virt_to_mfn(x->rawbuf);
> +    mfn_t mfn = virt_to_mfn(x->rawbuf);
>  
>      for ( i = 0; i < npages; i++ )
>      {
> -        struct page_info *page = mfn_to_page(mfn + i);
> +        struct page_info *page = mfn_to_page(mfn_add(mfn, i));

... and here.  This can easily be fixed on commit, so Reviewed-by:
Andrew Cooper <andrew.cooper3@citrix.com>

>          BUG_ON(page_get_owner(page) != current->domain);
>          if ( test_and_clear_bit(_PGC_allocated, &page->count_info) )
>              put_page(page);


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

[PATCH 6/7] xen/tmem: Convert the file common/tmem_xen.c to use typesafe MFN

[Julien Grall]

The file common/tmem_xen.c is now converted to use typesafe. This is
requiring to override the macro page_to_mfn to make it work with mfn_t.

Note that all variables converted to mfn_t havem there initial value,
when set, switch from 0 to INVALID_MFN. This is fine because the initial
values was always overriden before used.

Signed-off-by: Julien Grall <julien.grall@linaro.org>

---

Cc: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>
---
 xen/common/tmem_xen.c | 28 ++++++++++++++++------------
 1 file changed, 16 insertions(+), 12 deletions(-)

diff --git a/xen/common/tmem_xen.c b/xen/common/tmem_xen.c
index 20f74b268f..8dc031514a 100644
--- a/xen/common/tmem_xen.c
+++ b/xen/common/tmem_xen.c
@@ -14,6 +14,10 @@
 #include <xen/cpu.h>
 #include <xen/init.h>
 
+/* Override macros from asm/page.h to make them work with mfn_t */
+#undef page_to_mfn
+#define page_to_mfn(pg) _mfn(__page_to_mfn(pg))
+
 bool __read_mostly opt_tmem;
 boolean_param("tmem", opt_tmem);
 
@@ -31,7 +35,7 @@ static DEFINE_PER_CPU_READ_MOSTLY(unsigned char *, dstmem);
 static DEFINE_PER_CPU_READ_MOSTLY(void *, scratch_page);
 
 #if defined(CONFIG_ARM)
-static inline void *cli_get_page(xen_pfn_t cmfn, unsigned long *pcli_mfn,
+static inline void *cli_get_page(xen_pfn_t cmfn, mfn_t *pcli_mfn,
                                  struct page_info **pcli_pfp, bool cli_write)
 {
     ASSERT_UNREACHABLE();
@@ -39,14 +43,14 @@ static inline void *cli_get_page(xen_pfn_t cmfn, unsigned long *pcli_mfn,
 }
 
 static inline void cli_put_page(void *cli_va, struct page_info *cli_pfp,
-                                unsigned long cli_mfn, bool mark_dirty)
+                                mfn_t cli_mfn, bool mark_dirty)
 {
     ASSERT_UNREACHABLE();
 }
 #else
 #include <asm/p2m.h>
 
-static inline void *cli_get_page(xen_pfn_t cmfn, unsigned long *pcli_mfn,
+static inline void *cli_get_page(xen_pfn_t cmfn, mfn_t *pcli_mfn,
                                  struct page_info **pcli_pfp, bool cli_write)
 {
     p2m_type_t t;
@@ -68,16 +72,16 @@ static inline void *cli_get_page(xen_pfn_t cmfn, unsigned long *pcli_mfn,
 
     *pcli_mfn = page_to_mfn(page);
     *pcli_pfp = page;
-    return map_domain_page(_mfn(*pcli_mfn));
+    return map_domain_page(*pcli_mfn);
 }
 
 static inline void cli_put_page(void *cli_va, struct page_info *cli_pfp,
-                                unsigned long cli_mfn, bool mark_dirty)
+                                mfn_t cli_mfn, bool mark_dirty)
 {
     if ( mark_dirty )
     {
         put_page_and_type(cli_pfp);
-        paging_mark_dirty(current->domain, _mfn(cli_mfn));
+        paging_mark_dirty(current->domain, cli_mfn);
     }
     else
         put_page(cli_pfp);
@@ -88,14 +92,14 @@ static inline void cli_put_page(void *cli_va, struct page_info *cli_pfp,
 int tmem_copy_from_client(struct page_info *pfp,
     xen_pfn_t cmfn, tmem_cli_va_param_t clibuf)
 {
-    unsigned long tmem_mfn, cli_mfn = 0;
+    mfn_t tmem_mfn, cli_mfn = INVALID_MFN;
     char *tmem_va, *cli_va = NULL;
     struct page_info *cli_pfp = NULL;
     int rc = 1;
 
     ASSERT(pfp != NULL);
     tmem_mfn = page_to_mfn(pfp);
-    tmem_va = map_domain_page(_mfn(tmem_mfn));
+    tmem_va = map_domain_page(tmem_mfn);
     if ( guest_handle_is_null(clibuf) )
     {
         cli_va = cli_get_page(cmfn, &cli_mfn, &cli_pfp, 0);
@@ -125,7 +129,7 @@ int tmem_compress_from_client(xen_pfn_t cmfn,
     unsigned char *wmem = this_cpu(workmem);
     char *scratch = this_cpu(scratch_page);
     struct page_info *cli_pfp = NULL;
-    unsigned long cli_mfn = 0;
+    mfn_t cli_mfn = INVALID_MFN;
     void *cli_va = NULL;
 
     if ( dmem == NULL || wmem == NULL )
@@ -152,7 +156,7 @@ int tmem_compress_from_client(xen_pfn_t cmfn,
 int tmem_copy_to_client(xen_pfn_t cmfn, struct page_info *pfp,
     tmem_cli_va_param_t clibuf)
 {
-    unsigned long tmem_mfn, cli_mfn = 0;
+    mfn_t tmem_mfn, cli_mfn = INVALID_MFN;
     char *tmem_va, *cli_va = NULL;
     struct page_info *cli_pfp = NULL;
     int rc = 1;
@@ -165,7 +169,7 @@ int tmem_copy_to_client(xen_pfn_t cmfn, struct page_info *pfp,
             return -EFAULT;
     }
     tmem_mfn = page_to_mfn(pfp);
-    tmem_va = map_domain_page(_mfn(tmem_mfn));
+    tmem_va = map_domain_page(tmem_mfn);
     if ( cli_va )
     {
         memcpy(cli_va, tmem_va, PAGE_SIZE);
@@ -181,7 +185,7 @@ int tmem_copy_to_client(xen_pfn_t cmfn, struct page_info *pfp,
 int tmem_decompress_to_client(xen_pfn_t cmfn, void *tmem_va,
                                     size_t size, tmem_cli_va_param_t clibuf)
 {
-    unsigned long cli_mfn = 0;
+    mfn_t cli_mfn = INVALID_MFN;
     struct page_info *cli_pfp = NULL;
     void *cli_va = NULL;
     char *scratch = this_cpu(scratch_page);
-- 
2.11.0


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

Re: [PATCH 6/7] xen/tmem: Convert the file common/tmem_xen.c to use typesafe MFN

[Andrew Cooper]

On 04/10/2017 19:15, Julien Grall wrote:
> @@ -68,16 +72,16 @@ static inline void *cli_get_page(xen_pfn_t cmfn, unsigned long *pcli_mfn,
>  
>      *pcli_mfn = page_to_mfn(page);
>      *pcli_pfp = page;

Newline.

> -    return map_domain_page(_mfn(*pcli_mfn));
> +    return map_domain_page(*pcli_mfn);
>  }
>  
>  static inline void cli_put_page(void *cli_va, struct page_info *cli_pfp,
> -                                unsigned long cli_mfn, bool mark_dirty)
> +                                mfn_t cli_mfn, bool mark_dirty)
>  {
>      if ( mark_dirty )
>      {
>          put_page_and_type(cli_pfp);
> -        paging_mark_dirty(current->domain, _mfn(cli_mfn));
> +        paging_mark_dirty(current->domain, cli_mfn);
>      }
>      else
>          put_page(cli_pfp);
> @@ -165,7 +169,7 @@ int tmem_copy_to_client(xen_pfn_t cmfn, struct page_info *pfp,
>              return -EFAULT;
>      }
>      tmem_mfn = page_to_mfn(pfp);
> -    tmem_va = map_domain_page(_mfn(tmem_mfn));
> +    tmem_va = map_domain_page(tmem_mfn);

Newline.

Otherwise, Reviewed-by: Andrew Cooper <andrew.cooper3@citrix.com>

>      if ( cli_va )
>      {
>          memcpy(cli_va, tmem_va, PAGE_SIZE);
>


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

[PATCH 7/7] xen: Convert __page_to_mfn and __mfn_to_page to use typesafe MFN

[Julien Grall]

Most of the users of page_to_mfn and mfn_to_page are either overriding
the macros to make them work with mfn_t or use mfn_x/_mfn because the
rest of the function use mfn_t.

So make __page_to_mfn and __mfn_to_page return mfn_t by default.

Only reasonable clean-ups are done in this patch because it is
already quite big. So some of the files now override page_to_mfn and
mfn_to_page to avoid using mfn_t.

Signed-off-by: Julien Grall <julien.grall@linaro.org>

---

Cc: Stefano Stabellini <sstabellini@kernel.org>
Cc: Julien Grall <julien.grall@arm.com>
Cc: Andrew Cooper <andrew.cooper3@citrix.com>
Cc: George Dunlap <George.Dunlap@eu.citrix.com>
Cc: Ian Jackson <ian.jackson@eu.citrix.com>
Cc: Jan Beulich <jbeulich@suse.com>
Cc: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>
Cc: Tim Deegan <tim@xen.org>
Cc: Wei Liu <wei.liu2@citrix.com>
Cc: Razvan Cojocaru <rcojocaru@bitdefender.com>
Cc: Tamas K Lengyel <tamas@tklengyel.com>
Cc: Paul Durrant <paul.durrant@citrix.com>
Cc: Boris Ostrovsky <boris.ostrovsky@oracle.com>
Cc: Suravee Suthikulpanit <suravee.suthikulpanit@amd.com>
Cc: Jun Nakajima <jun.nakajima@intel.com>
Cc: Kevin Tian <kevin.tian@intel.com>
Cc: George Dunlap <george.dunlap@eu.citrix.com>
Cc: Gang Wei <gang.wei@intel.com>
Cc: Shane Wang <shane.wang@intel.com>
---
 xen/arch/arm/domain_build.c             |  8 ++++----
 xen/arch/arm/kernel.c                   |  2 +-
 xen/arch/arm/mem_access.c               |  2 +-
 xen/arch/arm/mm.c                       |  2 +-
 xen/arch/arm/p2m.c                      |  8 +-------
 xen/arch/x86/cpu/vpmu.c                 |  6 +++---
 xen/arch/x86/domain.c                   | 21 +++++++++++----------
 xen/arch/x86/domctl.c                   |  2 +-
 xen/arch/x86/hvm/dm.c                   |  2 +-
 xen/arch/x86/hvm/dom0_build.c           |  6 +++---
 xen/arch/x86/hvm/hvm.c                  | 16 ++++++++--------
 xen/arch/x86/hvm/ioreq.c                |  6 +++---
 xen/arch/x86/hvm/stdvga.c               |  2 +-
 xen/arch/x86/hvm/svm/svm.c              |  4 ++--
 xen/arch/x86/hvm/viridian.c             |  8 ++++----
 xen/arch/x86/hvm/vmx/vmcs.c             |  2 +-
 xen/arch/x86/hvm/vmx/vmx.c              | 10 +++++-----
 xen/arch/x86/hvm/vmx/vvmx.c             |  2 +-
 xen/arch/x86/mm.c                       |  6 ------
 xen/arch/x86/mm/guest_walk.c            |  6 +++---
 xen/arch/x86/mm/hap/guest_walk.c        |  2 +-
 xen/arch/x86/mm/hap/hap.c               |  6 ------
 xen/arch/x86/mm/hap/nested_ept.c        |  2 +-
 xen/arch/x86/mm/mem_sharing.c           |  9 ++-------
 xen/arch/x86/mm/p2m-ept.c               |  4 ++++
 xen/arch/x86/mm/p2m-pod.c               |  6 ------
 xen/arch/x86/mm/p2m.c                   |  6 ------
 xen/arch/x86/mm/paging.c                |  6 ------
 xen/arch/x86/mm/shadow/private.h        | 16 ++--------------
 xen/arch/x86/numa.c                     |  2 +-
 xen/arch/x86/physdev.c                  |  2 +-
 xen/arch/x86/pv/callback.c              |  6 ------
 xen/arch/x86/pv/descriptor-tables.c     |  6 ------
 xen/arch/x86/pv/dom0_build.c            |  6 ++++++
 xen/arch/x86/pv/domain.c                |  6 ------
 xen/arch/x86/pv/emul-gate-op.c          |  6 ------
 xen/arch/x86/pv/emul-priv-op.c          | 10 ----------
 xen/arch/x86/pv/grant_table.c           |  6 ------
 xen/arch/x86/pv/ro-page-fault.c         |  6 ------
 xen/arch/x86/smpboot.c                  |  6 ------
 xen/arch/x86/tboot.c                    |  4 ++--
 xen/arch/x86/traps.c                    |  2 +-
 xen/arch/x86/x86_64/mm.c                |  6 ++++++
 xen/common/domain.c                     |  4 ++--
 xen/common/event_fifo.c                 |  2 +-
 xen/common/grant_table.c                |  6 ++++++
 xen/common/kimage.c                     | 16 ++++++++--------
 xen/common/memory.c                     |  6 ++++++
 xen/common/page_alloc.c                 |  6 ++++++
 xen/common/tmem.c                       |  2 +-
 xen/common/tmem_xen.c                   |  4 ----
 xen/common/trace.c                      |  6 ++++++
 xen/common/vmap.c                       |  9 +++++----
 xen/common/xenoprof.c                   |  2 --
 xen/drivers/passthrough/amd/iommu_map.c |  6 ++++++
 xen/drivers/passthrough/iommu.c         |  2 +-
 xen/drivers/passthrough/x86/iommu.c     |  2 +-
 xen/include/asm-arm/mm.h                | 16 +++++++++-------
 xen/include/asm-arm/p2m.h               |  4 ++--
 xen/include/asm-x86/mm.h                | 12 ++++++------
 xen/include/asm-x86/p2m.h               |  2 +-
 xen/include/asm-x86/page.h              | 32 ++++++++++++++++----------------
 xen/include/xen/domain_page.h           |  4 ++--
 xen/include/xen/tmem_xen.h              |  2 +-
 64 files changed, 168 insertions(+), 229 deletions(-)

diff --git a/xen/arch/arm/domain_build.c b/xen/arch/arm/domain_build.c
index 093ebf1a8e..0753d03aac 100644
--- a/xen/arch/arm/domain_build.c
+++ b/xen/arch/arm/domain_build.c
@@ -104,11 +104,11 @@ static bool insert_11_bank(struct domain *d,
                            unsigned int order)
 {
     int res, i;
-    paddr_t spfn;
+    mfn_t smfn;
     paddr_t start, size;
 
-    spfn = page_to_mfn(pg);
-    start = pfn_to_paddr(spfn);
+    smfn = page_to_mfn(pg);
+    start = mfn_to_maddr(smfn);
     size = pfn_to_paddr(1UL << order);
 
     D11PRINT("Allocated %#"PRIpaddr"-%#"PRIpaddr" (%ldMB/%ldMB, order %d)\n",
@@ -126,7 +126,7 @@ static bool insert_11_bank(struct domain *d,
         goto fail;
     }
 
-    res = guest_physmap_add_page(d, _gfn(spfn), _mfn(spfn), order);
+    res = guest_physmap_add_page(d, _gfn(mfn_x(smfn)), smfn, order);
     if ( res )
         panic("Failed map pages to DOM0: %d", res);
 
diff --git a/xen/arch/arm/kernel.c b/xen/arch/arm/kernel.c
index 9c183f96da..f391938640 100644
--- a/xen/arch/arm/kernel.c
+++ b/xen/arch/arm/kernel.c
@@ -295,7 +295,7 @@ static __init int kernel_decompress(struct bootmodule *mod)
         iounmap(input);
         return -ENOMEM;
     }
-    mfn = _mfn(page_to_mfn(pages));
+    mfn = page_to_mfn(pages);
     output = __vmap(&mfn, 1 << kernel_order_out, 1, 1, PAGE_HYPERVISOR, VMAP_DEFAULT);
 
     rc = perform_gunzip(output, input, size);
diff --git a/xen/arch/arm/mem_access.c b/xen/arch/arm/mem_access.c
index 0f2cbb81d3..112e291cba 100644
--- a/xen/arch/arm/mem_access.c
+++ b/xen/arch/arm/mem_access.c
@@ -210,7 +210,7 @@ p2m_mem_access_check_and_get_page(vaddr_t gva, unsigned long flag,
     if ( t != p2m_ram_rw )
         goto err;
 
-    page = mfn_to_page(mfn_x(mfn));
+    page = mfn_to_page(mfn);
 
     if ( unlikely(!get_page(page, v->domain)) )
         page = NULL;
diff --git a/xen/arch/arm/mm.c b/xen/arch/arm/mm.c
index 9a37f29ce6..fe7a5da9bb 100644
--- a/xen/arch/arm/mm.c
+++ b/xen/arch/arm/mm.c
@@ -1286,7 +1286,7 @@ int xenmem_add_to_physmap_one(
             return -EINVAL;
         }
 
-        mfn = _mfn(page_to_mfn(page));
+        mfn = page_to_mfn(page);
         t = p2m_map_foreign;
 
         rcu_unlock_domain(od);
diff --git a/xen/arch/arm/p2m.c b/xen/arch/arm/p2m.c
index 0410b1e86b..1e7a0c6c40 100644
--- a/xen/arch/arm/p2m.c
+++ b/xen/arch/arm/p2m.c
@@ -38,12 +38,6 @@ static unsigned int __read_mostly max_vmid = MAX_VMID_8_BIT;
 
 #define P2M_ROOT_PAGES    (1<<P2M_ROOT_ORDER)
 
-/* Override macros from asm/mm.h to make them work with mfn_t */
-#undef mfn_to_page
-#define mfn_to_page(mfn) __mfn_to_page(mfn_x(mfn))
-#undef page_to_mfn
-#define page_to_mfn(pg) _mfn(__page_to_mfn(pg))
-
 unsigned int __read_mostly p2m_ipa_bits;
 
 /* Helpers to lookup the properties of each level */
@@ -98,7 +92,7 @@ void dump_p2m_lookup(struct domain *d, paddr_t addr)
     printk("dom%d IPA 0x%"PRIpaddr"\n", d->domain_id, addr);
 
     printk("P2M @ %p mfn:0x%lx\n",
-           p2m->root, __page_to_mfn(p2m->root));
+           p2m->root, mfn_x(page_to_mfn(p2m->root)));
 
     dump_pt_walk(page_to_maddr(p2m->root), addr,
                  P2M_ROOT_LEVEL, P2M_ROOT_PAGES);
diff --git a/xen/arch/x86/cpu/vpmu.c b/xen/arch/x86/cpu/vpmu.c
index fd2fcacc26..c3acddc53e 100644
--- a/xen/arch/x86/cpu/vpmu.c
+++ b/xen/arch/x86/cpu/vpmu.c
@@ -657,7 +657,7 @@ static void pvpmu_finish(struct domain *d, xen_pmu_params_t *params)
 {
     struct vcpu *v;
     struct vpmu_struct *vpmu;
-    uint64_t mfn;
+    mfn_t mfn;
     void *xenpmu_data;
 
     if ( (params->vcpu >= d->max_vcpus) || (d->vcpu[params->vcpu] == NULL) )
@@ -678,8 +678,8 @@ static void pvpmu_finish(struct domain *d, xen_pmu_params_t *params)
 
     if ( xenpmu_data )
     {
-        mfn = domain_page_map_to_mfn(xenpmu_data);
-        ASSERT(mfn_valid(_mfn(mfn)));
+        mfn = _mfn(domain_page_map_to_mfn(xenpmu_data));
+        ASSERT(mfn_valid(mfn));
         unmap_domain_page_global(xenpmu_data);
         put_page_and_type(mfn_to_page(mfn));
     }
diff --git a/xen/arch/x86/domain.c b/xen/arch/x86/domain.c
index bb1ffa3222..395ef6145a 100644
--- a/xen/arch/x86/domain.c
+++ b/xen/arch/x86/domain.c
@@ -186,7 +186,7 @@ void dump_pageframe_info(struct domain *d)
                 }
             }
             printk("    DomPage %p: caf=%08lx, taf=%" PRtype_info "\n",
-                   _p(page_to_mfn(page)),
+                   _p(mfn_x(page_to_mfn(page))),
                    page->count_info, page->u.inuse.type_info);
         }
         spin_unlock(&d->page_alloc_lock);
@@ -199,7 +199,7 @@ void dump_pageframe_info(struct domain *d)
     page_list_for_each ( page, &d->xenpage_list )
     {
         printk("    XenPage %p: caf=%08lx, taf=%" PRtype_info "\n",
-               _p(page_to_mfn(page)),
+               _p(mfn_x(page_to_mfn(page))),
                page->count_info, page->u.inuse.type_info);
     }
     spin_unlock(&d->page_alloc_lock);
@@ -621,7 +621,8 @@ int arch_domain_soft_reset(struct domain *d)
     struct page_info *page = virt_to_page(d->shared_info), *new_page;
     int ret = 0;
     struct domain *owner;
-    unsigned long mfn, gfn;
+    mfn_t mfn;
+    unsigned long gfn;
     p2m_type_t p2mt;
     unsigned int i;
 
@@ -655,7 +656,7 @@ int arch_domain_soft_reset(struct domain *d)
     ASSERT( owner == d );
 
     mfn = page_to_mfn(page);
-    gfn = mfn_to_gmfn(d, mfn);
+    gfn = mfn_to_gmfn(d, mfn_x(mfn));
 
     /*
      * gfn == INVALID_GFN indicates that the shared_info page was never mapped
@@ -664,7 +665,7 @@ int arch_domain_soft_reset(struct domain *d)
     if ( gfn == gfn_x(INVALID_GFN) )
         goto exit_put_page;
 
-    if ( mfn_x(get_gfn_query(d, gfn, &p2mt)) != mfn )
+    if ( !mfn_eq(get_gfn_query(d, gfn, &p2mt), mfn) )
     {
         printk(XENLOG_G_ERR "Failed to get Dom%d's shared_info GFN (%lx)\n",
                d->domain_id, gfn);
@@ -681,7 +682,7 @@ int arch_domain_soft_reset(struct domain *d)
         goto exit_put_gfn;
     }
 
-    ret = guest_physmap_remove_page(d, _gfn(gfn), _mfn(mfn), PAGE_ORDER_4K);
+    ret = guest_physmap_remove_page(d, _gfn(gfn), mfn, PAGE_ORDER_4K);
     if ( ret )
     {
         printk(XENLOG_G_ERR "Failed to remove Dom%d's shared_info frame %lx\n",
@@ -690,7 +691,7 @@ int arch_domain_soft_reset(struct domain *d)
         goto exit_put_gfn;
     }
 
-    ret = guest_physmap_add_page(d, _gfn(gfn), _mfn(page_to_mfn(new_page)),
+    ret = guest_physmap_add_page(d, _gfn(gfn), page_to_mfn(new_page),
                                  PAGE_ORDER_4K);
     if ( ret )
     {
@@ -988,7 +989,7 @@ int arch_set_info_guest(
                 {
                     if ( (page->u.inuse.type_info & PGT_type_mask) ==
                          PGT_l4_page_table )
-                        done = !fill_ro_mpt(_mfn(page_to_mfn(page)));
+                        done = !fill_ro_mpt(page_to_mfn(page));
 
                     page_unlock(page);
                 }
@@ -1114,7 +1115,7 @@ int arch_set_info_guest(
         l4_pgentry_t *l4tab;
 
         l4tab = map_domain_page(_mfn(pagetable_get_pfn(v->arch.guest_table)));
-        *l4tab = l4e_from_pfn(page_to_mfn(cr3_page),
+        *l4tab = l4e_from_pfn(mfn_x(page_to_mfn(cr3_page)),
             _PAGE_PRESENT|_PAGE_RW|_PAGE_USER|_PAGE_ACCESSED);
         unmap_domain_page(l4tab);
     }
@@ -1941,7 +1942,7 @@ int domain_relinquish_resources(struct domain *d)
         if ( d->arch.pirq_eoi_map != NULL )
         {
             unmap_domain_page_global(d->arch.pirq_eoi_map);
-            put_page_and_type(mfn_to_page(d->arch.pirq_eoi_map_mfn));
+            put_page_and_type(mfn_to_page(_mfn(d->arch.pirq_eoi_map_mfn)));
             d->arch.pirq_eoi_map = NULL;
             d->arch.auto_unmask = 0;
         }
diff --git a/xen/arch/x86/domctl.c b/xen/arch/x86/domctl.c
index 540ba089d7..9292ae5118 100644
--- a/xen/arch/x86/domctl.c
+++ b/xen/arch/x86/domctl.c
@@ -429,7 +429,7 @@ long arch_do_domctl(
         {
             if ( i >= max_pfns )
                 break;
-            mfn = page_to_mfn(page);
+            mfn = mfn_x(page_to_mfn(page));
             if ( copy_to_guest_offset(domctl->u.getmemlist.buffer,
                                       i, &mfn, 1) )
             {
diff --git a/xen/arch/x86/hvm/dm.c b/xen/arch/x86/hvm/dm.c
index 9cf53b551c..1a83f27c0b 100644
--- a/xen/arch/x86/hvm/dm.c
+++ b/xen/arch/x86/hvm/dm.c
@@ -219,7 +219,7 @@ static int modified_memory(struct domain *d,
             page = get_page_from_gfn(d, pfn, NULL, P2M_UNSHARE);
             if ( page )
             {
-                mfn_t gmfn = _mfn(page_to_mfn(page));
+                mfn_t gmfn = page_to_mfn(page);
 
                 paging_mark_dirty(d, gmfn);
                 /*
diff --git a/xen/arch/x86/hvm/dom0_build.c b/xen/arch/x86/hvm/dom0_build.c
index e8f746c70b..7789f6e571 100644
--- a/xen/arch/x86/hvm/dom0_build.c
+++ b/xen/arch/x86/hvm/dom0_build.c
@@ -120,7 +120,7 @@ static int __init pvh_populate_memory_range(struct domain *d,
             continue;
         }
 
-        rc = guest_physmap_add_page(d, _gfn(start), _mfn(page_to_mfn(page)),
+        rc = guest_physmap_add_page(d, _gfn(start), page_to_mfn(page),
                                     order);
         if ( rc != 0 )
         {
@@ -270,7 +270,7 @@ static int __init pvh_setup_vmx_realmode_helpers(struct domain *d)
     }
     write_32bit_pse_identmap(ident_pt);
     unmap_domain_page(ident_pt);
-    put_page(mfn_to_page(mfn_x(mfn)));
+    put_page(mfn_to_page(mfn));
     d->arch.hvm_domain.params[HVM_PARAM_IDENT_PT] = gaddr;
     if ( pvh_add_mem_range(d, gaddr, gaddr + PAGE_SIZE, E820_RESERVED) )
             printk("Unable to set identity page tables as reserved in the memory map\n");
@@ -288,7 +288,7 @@ static void __init pvh_steal_low_ram(struct domain *d, unsigned long start,
 
     for ( mfn = start; mfn < start + nr_pages; mfn++ )
     {
-        struct page_info *pg = mfn_to_page(mfn);
+        struct page_info *pg = mfn_to_page(_mfn(mfn));
         int rc;
 
         rc = unshare_xen_page_with_guest(pg, dom_io);
diff --git a/xen/arch/x86/hvm/hvm.c b/xen/arch/x86/hvm/hvm.c
index 205b4cb685..e8dbc23a51 100644
--- a/xen/arch/x86/hvm/hvm.c
+++ b/xen/arch/x86/hvm/hvm.c
@@ -2211,7 +2211,7 @@ int hvm_set_cr0(unsigned long value, bool_t may_defer)
             v->arch.guest_table = pagetable_from_page(page);
 
             HVM_DBG_LOG(DBG_LEVEL_VMMU, "Update CR3 value = %lx, mfn = %lx",
-                        v->arch.hvm_vcpu.guest_cr[3], page_to_mfn(page));
+                        v->arch.hvm_vcpu.guest_cr[3], mfn_x(page_to_mfn(page)));
         }
     }
     else if ( !(value & X86_CR0_PG) && (old_value & X86_CR0_PG) )
@@ -2546,7 +2546,7 @@ static void *_hvm_map_guest_frame(unsigned long gfn, bool_t permanent,
         if ( unlikely(p2m_is_discard_write(p2mt)) )
             *writable = 0;
         else if ( !permanent )
-            paging_mark_dirty(d, _mfn(page_to_mfn(page)));
+            paging_mark_dirty(d, page_to_mfn(page));
     }
 
     if ( !permanent )
@@ -2588,13 +2588,13 @@ void *hvm_map_guest_frame_ro(unsigned long gfn, bool_t permanent)
 
 void hvm_unmap_guest_frame(void *p, bool_t permanent)
 {
-    unsigned long mfn;
+    mfn_t mfn;
     struct page_info *page;
 
     if ( !p )
         return;
 
-    mfn = domain_page_map_to_mfn(p);
+    mfn = _mfn(domain_page_map_to_mfn(p));
     page = mfn_to_page(mfn);
 
     if ( !permanent )
@@ -2609,7 +2609,7 @@ void hvm_unmap_guest_frame(void *p, bool_t permanent)
         list_for_each_entry(track, &d->arch.hvm_domain.write_map.list, list)
             if ( track->page == page )
             {
-                paging_mark_dirty(d, _mfn(mfn));
+                paging_mark_dirty(d, mfn);
                 list_del(&track->list);
                 xfree(track);
                 break;
@@ -2626,7 +2626,7 @@ void hvm_mapped_guest_frames_mark_dirty(struct domain *d)
 
     spin_lock(&d->arch.hvm_domain.write_map.lock);
     list_for_each_entry(track, &d->arch.hvm_domain.write_map.list, list)
-        paging_mark_dirty(d, _mfn(page_to_mfn(track->page)));
+        paging_mark_dirty(d, page_to_mfn(track->page));
     spin_unlock(&d->arch.hvm_domain.write_map.lock);
 }
 
@@ -3201,7 +3201,7 @@ static enum hvm_translation_result __hvm_copy(
                 if ( xchg(&lastpage, gfn_x(gfn)) != gfn_x(gfn) )
                     dprintk(XENLOG_G_DEBUG,
                             "%pv attempted write to read-only gfn %#lx (mfn=%#lx)\n",
-                            v, gfn_x(gfn), page_to_mfn(page));
+                            v, gfn_x(gfn), mfn_x(page_to_mfn(page)));
             }
             else
             {
@@ -3209,7 +3209,7 @@ static enum hvm_translation_result __hvm_copy(
                     memcpy(p, buf, count);
                 else
                     memset(p, 0, count);
-                paging_mark_dirty(v->domain, _mfn(page_to_mfn(page)));
+                paging_mark_dirty(v->domain, page_to_mfn(page));
             }
         }
         else
diff --git a/xen/arch/x86/hvm/ioreq.c b/xen/arch/x86/hvm/ioreq.c
index f2e0b3f74a..5bd5cd788e 100644
--- a/xen/arch/x86/hvm/ioreq.c
+++ b/xen/arch/x86/hvm/ioreq.c
@@ -268,7 +268,7 @@ static void hvm_remove_ioreq_gfn(
     struct domain *d, struct hvm_ioreq_page *iorp)
 {
     if ( guest_physmap_remove_page(d, _gfn(iorp->gfn),
-                                   _mfn(page_to_mfn(iorp->page)), 0) )
+                                   page_to_mfn(iorp->page), 0) )
         domain_crash(d);
     clear_page(iorp->va);
 }
@@ -281,9 +281,9 @@ static int hvm_add_ioreq_gfn(
     clear_page(iorp->va);
 
     rc = guest_physmap_add_page(d, _gfn(iorp->gfn),
-                                _mfn(page_to_mfn(iorp->page)), 0);
+                                page_to_mfn(iorp->page), 0);
     if ( rc == 0 )
-        paging_mark_dirty(d, _mfn(page_to_mfn(iorp->page)));
+        paging_mark_dirty(d, page_to_mfn(iorp->page));
 
     return rc;
 }
diff --git a/xen/arch/x86/hvm/stdvga.c b/xen/arch/x86/hvm/stdvga.c
index 088fbdf8ce..925bab2438 100644
--- a/xen/arch/x86/hvm/stdvga.c
+++ b/xen/arch/x86/hvm/stdvga.c
@@ -590,7 +590,7 @@ void stdvga_init(struct domain *d)
         if ( pg == NULL )
             break;
         s->vram_page[i] = pg;
-        clear_domain_page(_mfn(page_to_mfn(pg)));
+        clear_domain_page(page_to_mfn(pg));
     }
 
     if ( i == ARRAY_SIZE(s->vram_page) )
diff --git a/xen/arch/x86/hvm/svm/svm.c b/xen/arch/x86/hvm/svm/svm.c
index b9cf423fd9..f50f931598 100644
--- a/xen/arch/x86/hvm/svm/svm.c
+++ b/xen/arch/x86/hvm/svm/svm.c
@@ -1521,7 +1521,7 @@ static int svm_cpu_up_prepare(unsigned int cpu)
         if ( !pg )
             goto err;
 
-        clear_domain_page(_mfn(page_to_mfn(pg)));
+        clear_domain_page(page_to_mfn(pg));
         *this_hsa = page_to_maddr(pg);
     }
 
@@ -1531,7 +1531,7 @@ static int svm_cpu_up_prepare(unsigned int cpu)
         if ( !pg )
             goto err;
 
-        clear_domain_page(_mfn(page_to_mfn(pg)));
+        clear_domain_page(page_to_mfn(pg));
         *this_vmcb = page_to_maddr(pg);
     }
 
diff --git a/xen/arch/x86/hvm/viridian.c b/xen/arch/x86/hvm/viridian.c
index f0fa59d7d5..c5440863b1 100644
--- a/xen/arch/x86/hvm/viridian.c
+++ b/xen/arch/x86/hvm/viridian.c
@@ -354,7 +354,7 @@ static void enable_hypercall_page(struct domain *d)
         if ( page )
             put_page(page);
         gdprintk(XENLOG_WARNING, "Bad GMFN %#"PRI_gfn" (MFN %#"PRI_mfn")\n",
-                 gmfn, page ? page_to_mfn(page) : mfn_x(INVALID_MFN));
+                 gmfn, mfn_x(page ? page_to_mfn(page) : INVALID_MFN));
         return;
     }
 
@@ -414,7 +414,7 @@ static void initialize_vp_assist(struct vcpu *v)
 
  fail:
     gdprintk(XENLOG_WARNING, "Bad GMFN %#"PRI_gfn" (MFN %#"PRI_mfn")\n", gmfn,
-             page ? page_to_mfn(page) : mfn_x(INVALID_MFN));
+             mfn_x(page ? page_to_mfn(page) : INVALID_MFN));
 }
 
 static void teardown_vp_assist(struct vcpu *v)
@@ -427,7 +427,7 @@ static void teardown_vp_assist(struct vcpu *v)
 
     v->arch.hvm_vcpu.viridian.vp_assist.va = NULL;
 
-    page = mfn_to_page(domain_page_map_to_mfn(va));
+    page = mfn_to_page(_mfn(domain_page_map_to_mfn(va)));
 
     unmap_domain_page_global(va);
     put_page_and_type(page);
@@ -494,7 +494,7 @@ static void update_reference_tsc(struct domain *d, bool_t initialize)
         if ( page )
             put_page(page);
         gdprintk(XENLOG_WARNING, "Bad GMFN %#"PRI_gfn" (MFN %#"PRI_mfn")\n",
-                 gmfn, page ? page_to_mfn(page) : mfn_x(INVALID_MFN));
+                 gmfn, mfn_x(page ? page_to_mfn(page) : INVALID_MFN));
         return;
     }
 
diff --git a/xen/arch/x86/hvm/vmx/vmcs.c b/xen/arch/x86/hvm/vmx/vmcs.c
index f62fe7e217..471d224539 100644
--- a/xen/arch/x86/hvm/vmx/vmcs.c
+++ b/xen/arch/x86/hvm/vmx/vmcs.c
@@ -1441,7 +1441,7 @@ int vmx_vcpu_enable_pml(struct vcpu *v)
 
     vmx_vmcs_enter(v);
 
-    __vmwrite(PML_ADDRESS, page_to_mfn(v->arch.hvm_vmx.pml_pg) << PAGE_SHIFT);
+    __vmwrite(PML_ADDRESS, page_to_maddr(v->arch.hvm_vmx.pml_pg));
     __vmwrite(GUEST_PML_INDEX, NR_PML_ENTRIES - 1);
 
     v->arch.hvm_vmx.secondary_exec_control |= SECONDARY_EXEC_ENABLE_PML;
diff --git a/xen/arch/x86/hvm/vmx/vmx.c b/xen/arch/x86/hvm/vmx/vmx.c
index 9cfa9b6965..40b91933bf 100644
--- a/xen/arch/x86/hvm/vmx/vmx.c
+++ b/xen/arch/x86/hvm/vmx/vmx.c
@@ -2951,7 +2951,7 @@ gp_fault:
 static int vmx_alloc_vlapic_mapping(struct domain *d)
 {
     struct page_info *pg;
-    unsigned long mfn;
+    mfn_t mfn;
 
     if ( !cpu_has_vmx_virtualize_apic_accesses )
         return 0;
@@ -2960,10 +2960,10 @@ static int vmx_alloc_vlapic_mapping(struct domain *d)
     if ( !pg )
         return -ENOMEM;
     mfn = page_to_mfn(pg);
-    clear_domain_page(_mfn(mfn));
+    clear_domain_page(mfn);
     share_xen_page_with_guest(pg, d, XENSHARE_writable);
-    d->arch.hvm_domain.vmx.apic_access_mfn = mfn;
-    set_mmio_p2m_entry(d, paddr_to_pfn(APIC_DEFAULT_PHYS_BASE), _mfn(mfn),
+    d->arch.hvm_domain.vmx.apic_access_mfn = mfn_x(mfn);
+    set_mmio_p2m_entry(d, paddr_to_pfn(APIC_DEFAULT_PHYS_BASE), mfn,
                        PAGE_ORDER_4K, p2m_get_hostp2m(d)->default_access);
 
     return 0;
@@ -2974,7 +2974,7 @@ static void vmx_free_vlapic_mapping(struct domain *d)
     unsigned long mfn = d->arch.hvm_domain.vmx.apic_access_mfn;
 
     if ( mfn != 0 )
-        free_shared_domheap_page(mfn_to_page(mfn));
+        free_shared_domheap_page(mfn_to_page(_mfn(mfn)));
 }
 
 static void vmx_install_vlapic_mapping(struct vcpu *v)
diff --git a/xen/arch/x86/hvm/vmx/vvmx.c b/xen/arch/x86/hvm/vmx/vvmx.c
index cd0ee0a307..790a2285e5 100644
--- a/xen/arch/x86/hvm/vmx/vvmx.c
+++ b/xen/arch/x86/hvm/vmx/vvmx.c
@@ -84,7 +84,7 @@ int nvmx_vcpu_initialise(struct vcpu *v)
         }
         v->arch.hvm_vmx.vmread_bitmap = vmread_bitmap;
 
-        clear_domain_page(_mfn(page_to_mfn(vmread_bitmap)));
+        clear_domain_page(page_to_mfn(vmread_bitmap));
 
         vmwrite_bitmap = alloc_domheap_page(NULL, 0);
         if ( !vmwrite_bitmap )
diff --git a/xen/arch/x86/mm.c b/xen/arch/x86/mm.c
index d9df5ca69f..39038723ce 100644
--- a/xen/arch/x86/mm.c
+++ b/xen/arch/x86/mm.c
@@ -129,12 +129,6 @@
 
 #include "pv/mm.h"
 
-/* Override macros from asm/page.h to make them work with mfn_t */
-#undef mfn_to_page
-#define mfn_to_page(mfn) __mfn_to_page(mfn_x(mfn))
-#undef page_to_mfn
-#define page_to_mfn(pg) _mfn(__page_to_mfn(pg))
-
 /* Mapping of the fixmap space needed early. */
 l1_pgentry_t __section(".bss.page_aligned") __aligned(PAGE_SIZE)
     l1_fixmap[L1_PAGETABLE_ENTRIES];
diff --git a/xen/arch/x86/mm/guest_walk.c b/xen/arch/x86/mm/guest_walk.c
index 6055fec1ad..f67aeda3d0 100644
--- a/xen/arch/x86/mm/guest_walk.c
+++ b/xen/arch/x86/mm/guest_walk.c
@@ -469,20 +469,20 @@ guest_walk_tables(struct vcpu *v, struct p2m_domain *p2m,
     if ( l3p )
     {
         unmap_domain_page(l3p);
-        put_page(mfn_to_page(mfn_x(gw->l3mfn)));
+        put_page(mfn_to_page(gw->l3mfn));
     }
 #endif
 #if GUEST_PAGING_LEVELS >= 3
     if ( l2p )
     {
         unmap_domain_page(l2p);
-        put_page(mfn_to_page(mfn_x(gw->l2mfn)));
+        put_page(mfn_to_page(gw->l2mfn));
     }
 #endif
     if ( l1p )
     {
         unmap_domain_page(l1p);
-        put_page(mfn_to_page(mfn_x(gw->l1mfn)));
+        put_page(mfn_to_page(gw->l1mfn));
     }
 
     return walk_ok;
diff --git a/xen/arch/x86/mm/hap/guest_walk.c b/xen/arch/x86/mm/hap/guest_walk.c
index c550017ba4..cb3f9cebe7 100644
--- a/xen/arch/x86/mm/hap/guest_walk.c
+++ b/xen/arch/x86/mm/hap/guest_walk.c
@@ -83,7 +83,7 @@ unsigned long hap_p2m_ga_to_gfn(GUEST_PAGING_LEVELS)(
         *pfec &= ~PFEC_page_present;
         goto out_tweak_pfec;
     }
-    top_mfn = _mfn(page_to_mfn(top_page));
+    top_mfn = page_to_mfn(top_page);
 
     /* Map the top-level table and call the tree-walker */
     ASSERT(mfn_valid(top_mfn));
diff --git a/xen/arch/x86/mm/hap/hap.c b/xen/arch/x86/mm/hap/hap.c
index dc85e828cd..e45c1a1913 100644
--- a/xen/arch/x86/mm/hap/hap.c
+++ b/xen/arch/x86/mm/hap/hap.c
@@ -42,12 +42,6 @@
 
 #include "private.h"
 
-/* Override macros from asm/page.h to make them work with mfn_t */
-#undef mfn_to_page
-#define mfn_to_page(_m) __mfn_to_page(mfn_x(_m))
-#undef page_to_mfn
-#define page_to_mfn(_pg) _mfn(__page_to_mfn(_pg))
-
 /************************************************/
 /*          HAP VRAM TRACKING SUPPORT           */
 /************************************************/
diff --git a/xen/arch/x86/mm/hap/nested_ept.c b/xen/arch/x86/mm/hap/nested_ept.c
index 14b1bb01e9..1738df69f6 100644
--- a/xen/arch/x86/mm/hap/nested_ept.c
+++ b/xen/arch/x86/mm/hap/nested_ept.c
@@ -173,7 +173,7 @@ nept_walk_tables(struct vcpu *v, unsigned long l2ga, ept_walk_t *gw)
             goto map_err;
         gw->lxe[lvl] = lxp[ept_lvl_table_offset(l2ga, lvl)];
         unmap_domain_page(lxp);
-        put_page(mfn_to_page(mfn_x(lxmfn)));
+        put_page(mfn_to_page(lxmfn));
 
         if ( nept_non_present_check(gw->lxe[lvl]) )
             goto non_present;
diff --git a/xen/arch/x86/mm/mem_sharing.c b/xen/arch/x86/mm/mem_sharing.c
index b856028c02..b799bdb77c 100644
--- a/xen/arch/x86/mm/mem_sharing.c
+++ b/xen/arch/x86/mm/mem_sharing.c
@@ -152,11 +152,6 @@ static inline shr_handle_t get_next_handle(void)
 #define mem_sharing_enabled(d) \
     (is_hvm_domain(d) && (d)->arch.hvm_domain.mem_sharing_enabled)
 
-#undef mfn_to_page
-#define mfn_to_page(_m) __mfn_to_page(mfn_x(_m))
-#undef page_to_mfn
-#define page_to_mfn(_pg) _mfn(__page_to_mfn(_pg))
-
 static atomic_t nr_saved_mfns   = ATOMIC_INIT(0); 
 static atomic_t nr_shared_mfns  = ATOMIC_INIT(0);
 
@@ -1185,8 +1180,8 @@ int __mem_sharing_unshare_page(struct domain *d,
         return -ENOMEM;
     }
 
-    s = map_domain_page(_mfn(__page_to_mfn(old_page)));
-    t = map_domain_page(_mfn(__page_to_mfn(page)));
+    s = map_domain_page(page_to_mfn(old_page));
+    t = map_domain_page(page_to_mfn(page));
     memcpy(t, s, PAGE_SIZE);
     unmap_domain_page(s);
     unmap_domain_page(t);
diff --git a/xen/arch/x86/mm/p2m-ept.c b/xen/arch/x86/mm/p2m-ept.c
index 054827aa88..24de202a1b 100644
--- a/xen/arch/x86/mm/p2m-ept.c
+++ b/xen/arch/x86/mm/p2m-ept.c
@@ -33,6 +33,10 @@
 
 #include "mm-locks.h"
 
+/* Override macros from asm/page.h to avoid using typesafe mfn_t. */
+#undef mfn_to_page
+#define mfn_to_page(mfn) __mfn_to_page(_mfn(mfn))
+
 #define atomic_read_ept_entry(__pepte)                              \
     ( (ept_entry_t) { .epte = read_atomic(&(__pepte)->epte) } )
 
diff --git a/xen/arch/x86/mm/p2m-pod.c b/xen/arch/x86/mm/p2m-pod.c
index 0a811ccf28..7a88074c31 100644
--- a/xen/arch/x86/mm/p2m-pod.c
+++ b/xen/arch/x86/mm/p2m-pod.c
@@ -29,12 +29,6 @@
 
 #include "mm-locks.h"
 
-/* Override macros from asm/page.h to make them work with mfn_t */
-#undef mfn_to_page
-#define mfn_to_page(_m) __mfn_to_page(mfn_x(_m))
-#undef page_to_mfn
-#define page_to_mfn(_pg) _mfn(__page_to_mfn(_pg))
-
 #define superpage_aligned(_x)  (((_x)&(SUPERPAGE_PAGES-1))==0)
 
 /* Enforce lock ordering when grabbing the "external" page_alloc lock */
diff --git a/xen/arch/x86/mm/p2m.c b/xen/arch/x86/mm/p2m.c
index 3fbc537da6..2194b35bc7 100644
--- a/xen/arch/x86/mm/p2m.c
+++ b/xen/arch/x86/mm/p2m.c
@@ -47,12 +47,6 @@ bool_t __initdata opt_hap_1gb = 1, __initdata opt_hap_2mb = 1;
 boolean_param("hap_1gb", opt_hap_1gb);
 boolean_param("hap_2mb", opt_hap_2mb);
 
-/* Override macros from asm/page.h to make them work with mfn_t */
-#undef mfn_to_page
-#define mfn_to_page(_m) __mfn_to_page(mfn_x(_m))
-#undef page_to_mfn
-#define page_to_mfn(_pg) _mfn(__page_to_mfn(_pg))
-
 DEFINE_PERCPU_RWLOCK_GLOBAL(p2m_percpu_rwlock);
 
 /* Init the datastructures for later use by the p2m code */
diff --git a/xen/arch/x86/mm/paging.c b/xen/arch/x86/mm/paging.c
index 1e2c9ba4cc..cb97642cbc 100644
--- a/xen/arch/x86/mm/paging.c
+++ b/xen/arch/x86/mm/paging.c
@@ -47,12 +47,6 @@
 /* Per-CPU variable for enforcing the lock ordering */
 DEFINE_PER_CPU(int, mm_lock_level);
 
-/* Override macros from asm/page.h to make them work with mfn_t */
-#undef mfn_to_page
-#define mfn_to_page(_m) __mfn_to_page(mfn_x(_m))
-#undef page_to_mfn
-#define page_to_mfn(_pg) _mfn(__page_to_mfn(_pg))
-
 /************************************************/
 /*              LOG DIRTY SUPPORT               */
 /************************************************/
diff --git a/xen/arch/x86/mm/shadow/private.h b/xen/arch/x86/mm/shadow/private.h
index 6a03370402..b9cc680f4e 100644
--- a/xen/arch/x86/mm/shadow/private.h
+++ b/xen/arch/x86/mm/shadow/private.h
@@ -315,7 +315,7 @@ static inline int page_is_out_of_sync(struct page_info *p)
 
 static inline int mfn_is_out_of_sync(mfn_t gmfn)
 {
-    return page_is_out_of_sync(mfn_to_page(mfn_x(gmfn)));
+    return page_is_out_of_sync(mfn_to_page(gmfn));
 }
 
 static inline int page_oos_may_write(struct page_info *p)
@@ -326,7 +326,7 @@ static inline int page_oos_may_write(struct page_info *p)
 
 static inline int mfn_oos_may_write(mfn_t gmfn)
 {
-    return page_oos_may_write(mfn_to_page(mfn_x(gmfn)));
+    return page_oos_may_write(mfn_to_page(gmfn));
 }
 #endif /* (SHADOW_OPTIMIZATIONS & SHOPT_OUT_OF_SYNC) */
 
@@ -465,18 +465,6 @@ void sh_reset_l3_up_pointers(struct vcpu *v);
  * MFN/page-info handling
  */
 
-/* Override macros from asm/page.h to make them work with mfn_t */
-#undef mfn_to_page
-#define mfn_to_page(_m) __mfn_to_page(mfn_x(_m))
-#undef page_to_mfn
-#define page_to_mfn(_pg) _mfn(__page_to_mfn(_pg))
-
-/* Override pagetable_t <-> struct page_info conversions to work with mfn_t */
-#undef pagetable_get_page
-#define pagetable_get_page(x)   mfn_to_page(pagetable_get_mfn(x))
-#undef pagetable_from_page
-#define pagetable_from_page(pg) pagetable_from_mfn(page_to_mfn(pg))
-
 #define backpointer(sp) _mfn(pdx_to_pfn((unsigned long)(sp)->v.sh.back))
 static inline unsigned long __backpointer(const struct page_info *sp)
 {
diff --git a/xen/arch/x86/numa.c b/xen/arch/x86/numa.c
index 4fc967f893..a87987da6f 100644
--- a/xen/arch/x86/numa.c
+++ b/xen/arch/x86/numa.c
@@ -430,7 +430,7 @@ static void dump_numa(unsigned char key)
         spin_lock(&d->page_alloc_lock);
         page_list_for_each(page, &d->page_list)
         {
-            i = phys_to_nid((paddr_t)page_to_mfn(page) << PAGE_SHIFT);
+            i = phys_to_nid(page_to_maddr(page));
             page_num_node[i]++;
         }
         spin_unlock(&d->page_alloc_lock);
diff --git a/xen/arch/x86/physdev.c b/xen/arch/x86/physdev.c
index 0eb409758f..ba950af4a8 100644
--- a/xen/arch/x86/physdev.c
+++ b/xen/arch/x86/physdev.c
@@ -241,7 +241,7 @@ ret_t do_physdev_op(int cmd, XEN_GUEST_HANDLE_PARAM(void) arg)
         }
 
         if ( cmpxchg(&currd->arch.pirq_eoi_map_mfn,
-                     0, page_to_mfn(page)) != 0 )
+                     0, mfn_x(page_to_mfn(page))) != 0 )
         {
             put_page_and_type(page);
             ret = -EBUSY;
diff --git a/xen/arch/x86/pv/callback.c b/xen/arch/x86/pv/callback.c
index 97d8438600..5957cb5085 100644
--- a/xen/arch/x86/pv/callback.c
+++ b/xen/arch/x86/pv/callback.c
@@ -31,12 +31,6 @@
 
 #include <public/callback.h>
 
-/* Override macros from asm/page.h to make them work with mfn_t */
-#undef mfn_to_page
-#define mfn_to_page(mfn) __mfn_to_page(mfn_x(mfn))
-#undef page_to_mfn
-#define page_to_mfn(pg) _mfn(__page_to_mfn(pg))
-
 static int register_guest_nmi_callback(unsigned long address)
 {
     struct vcpu *curr = current;
diff --git a/xen/arch/x86/pv/descriptor-tables.c b/xen/arch/x86/pv/descriptor-tables.c
index 81973af124..371221a302 100644
--- a/xen/arch/x86/pv/descriptor-tables.c
+++ b/xen/arch/x86/pv/descriptor-tables.c
@@ -25,12 +25,6 @@
 #include <asm/p2m.h>
 #include <asm/pv/mm.h>
 
-/* Override macros from asm/page.h to make them work with mfn_t */
-#undef mfn_to_page
-#define mfn_to_page(mfn) __mfn_to_page(mfn_x(mfn))
-#undef page_to_mfn
-#define page_to_mfn(pg) _mfn(__page_to_mfn(pg))
-
 /*******************
  * Descriptor Tables
  */
diff --git a/xen/arch/x86/pv/dom0_build.c b/xen/arch/x86/pv/dom0_build.c
index dcbee43e8f..e9a893ba47 100644
--- a/xen/arch/x86/pv/dom0_build.c
+++ b/xen/arch/x86/pv/dom0_build.c
@@ -22,6 +22,12 @@
 
 #include "mm.h"
 
+/* Override macros from asm/page.h to avoid using typesafe mfn_t. */
+#undef page_to_mfn
+#define page_to_mfn(pg) mfn_x(__page_to_mfn(pg))
+#undef mfn_to_page
+#define mfn_to_page(mfn) __mfn_to_page(_mfn(mfn))
+
 /* Allow ring-3 access in long mode as guest cannot use ring 1 ... */
 #define BASE_PROT (_PAGE_PRESENT|_PAGE_RW|_PAGE_ACCESSED|_PAGE_USER)
 #define L1_PROT (BASE_PROT|_PAGE_GUEST_KERNEL)
diff --git a/xen/arch/x86/pv/domain.c b/xen/arch/x86/pv/domain.c
index 90d5569be1..4ca3205821 100644
--- a/xen/arch/x86/pv/domain.c
+++ b/xen/arch/x86/pv/domain.c
@@ -16,12 +16,6 @@
 
 #include "mm.h"
 
-/* Override macros from asm/page.h to make them work with mfn_t */
-#undef mfn_to_page
-#define mfn_to_page(mfn) __mfn_to_page(mfn_x(mfn))
-#undef page_to_mfn
-#define page_to_mfn(pg) _mfn(__page_to_mfn(pg))
-
 static void noreturn continue_nonidle_domain(struct vcpu *v)
 {
     check_wakeup_from_wait();
diff --git a/xen/arch/x86/pv/emul-gate-op.c b/xen/arch/x86/pv/emul-gate-op.c
index 0f89c91dff..5cdb54c937 100644
--- a/xen/arch/x86/pv/emul-gate-op.c
+++ b/xen/arch/x86/pv/emul-gate-op.c
@@ -41,12 +41,6 @@
 
 #include "emulate.h"
 
-/* Override macros from asm/page.h to make them work with mfn_t */
-#undef mfn_to_page
-#define mfn_to_page(mfn) __mfn_to_page(mfn_x(mfn))
-#undef page_to_mfn
-#define page_to_mfn(pg) _mfn(__page_to_mfn(pg))
-
 static int read_gate_descriptor(unsigned int gate_sel,
                                 const struct vcpu *v,
                                 unsigned int *sel,
diff --git a/xen/arch/x86/pv/emul-priv-op.c b/xen/arch/x86/pv/emul-priv-op.c
index dd90713acf..9ccbd021ef 100644
--- a/xen/arch/x86/pv/emul-priv-op.c
+++ b/xen/arch/x86/pv/emul-priv-op.c
@@ -43,16 +43,6 @@
 #include "emulate.h"
 #include "mm.h"
 
-/* Override macros from asm/page.h to make them work with mfn_t */
-#undef mfn_to_page
-#define mfn_to_page(mfn) __mfn_to_page(mfn_x(mfn))
-#undef page_to_mfn
-#define page_to_mfn(pg) _mfn(__page_to_mfn(pg))
-
-/***********************
- * I/O emulation support
- */
-
 struct priv_op_ctxt {
     struct x86_emulate_ctxt ctxt;
     struct {
diff --git a/xen/arch/x86/pv/grant_table.c b/xen/arch/x86/pv/grant_table.c
index aaca228c6b..97323367c5 100644
--- a/xen/arch/x86/pv/grant_table.c
+++ b/xen/arch/x86/pv/grant_table.c
@@ -27,12 +27,6 @@
 
 #include "mm.h"
 
-/* Override macros from asm/page.h to make them work with mfn_t */
-#undef mfn_to_page
-#define mfn_to_page(mfn) __mfn_to_page(mfn_x(mfn))
-#undef page_to_mfn
-#define page_to_mfn(pg) _mfn(__page_to_mfn(pg))
-
 static unsigned int grant_to_pte_flags(unsigned int grant_flags,
                                        unsigned int cache_flags)
 {
diff --git a/xen/arch/x86/pv/ro-page-fault.c b/xen/arch/x86/pv/ro-page-fault.c
index 6b2976d3df..a7b7eb5113 100644
--- a/xen/arch/x86/pv/ro-page-fault.c
+++ b/xen/arch/x86/pv/ro-page-fault.c
@@ -33,12 +33,6 @@
 #include "emulate.h"
 #include "mm.h"
 
-/* Override macros from asm/page.h to make them work with mfn_t */
-#undef mfn_to_page
-#define mfn_to_page(mfn) __mfn_to_page(mfn_x(mfn))
-#undef page_to_mfn
-#define page_to_mfn(pg) _mfn(__page_to_mfn(pg))
-
 /*********************
  * Writable Pagetables
  */
diff --git a/xen/arch/x86/smpboot.c b/xen/arch/x86/smpboot.c
index 3ca716c59f..663966bc74 100644
--- a/xen/arch/x86/smpboot.c
+++ b/xen/arch/x86/smpboot.c
@@ -46,12 +46,6 @@
 #include <mach_wakecpu.h>
 #include <smpboot_hooks.h>
 
-/* Override macros from asm/page.h to make them work with mfn_t */
-#undef mfn_to_page
-#define mfn_to_page(mfn) __mfn_to_page(mfn_x(mfn))
-#undef page_to_mfn
-#define page_to_mfn(pg) _mfn(__page_to_mfn(pg))
-
 #define setup_trampoline()    (bootsym_phys(trampoline_realmode_entry))
 
 unsigned long __read_mostly trampoline_phys;
diff --git a/xen/arch/x86/tboot.c b/xen/arch/x86/tboot.c
index 59d7c477f4..e9522f06ec 100644
--- a/xen/arch/x86/tboot.c
+++ b/xen/arch/x86/tboot.c
@@ -184,7 +184,7 @@ static void update_pagetable_mac(vmac_ctx_t *ctx)
 
     for ( mfn = 0; mfn < max_page; mfn++ )
     {
-        struct page_info *page = mfn_to_page(mfn);
+        struct page_info *page = mfn_to_page(_mfn(mfn));
 
         if ( !mfn_valid(_mfn(mfn)) )
             continue;
@@ -276,7 +276,7 @@ static void tboot_gen_xenheap_integrity(const uint8_t key[TB_KEY_SIZE],
     vmac_set_key((uint8_t *)key, &ctx);
     for ( mfn = 0; mfn < max_page; mfn++ )
     {
-        struct page_info *page = __mfn_to_page(mfn);
+        struct page_info *page = mfn_to_page(_mfn(mfn));
 
         if ( !mfn_valid(_mfn(mfn)) )
             continue;
diff --git a/xen/arch/x86/traps.c b/xen/arch/x86/traps.c
index 86506f3747..b85394d1f9 100644
--- a/xen/arch/x86/traps.c
+++ b/xen/arch/x86/traps.c
@@ -811,7 +811,7 @@ int wrmsr_hypervisor_regs(uint32_t idx, uint64_t val)
 
             gdprintk(XENLOG_WARNING,
                      "Bad GMFN %lx (MFN %lx) to MSR %08x\n",
-                     gmfn, page ? page_to_mfn(page) : -1UL, base);
+                     gmfn, page ? mfn_x(page_to_mfn(page)) : -1UL, base);
             return 0;
         }
 
diff --git a/xen/arch/x86/x86_64/mm.c b/xen/arch/x86/x86_64/mm.c
index 11746730b4..971ccfcbbe 100644
--- a/xen/arch/x86/x86_64/mm.c
+++ b/xen/arch/x86/x86_64/mm.c
@@ -40,6 +40,12 @@ asm(".file \"" __FILE__ "\"");
 #include <asm/mem_sharing.h>
 #include <public/memory.h>
 
+/* Override macros from asm/page.h to avoid using typesafe mfn_t. */
+#undef page_to_mfn
+#define page_to_mfn(pg) mfn_x(__page_to_mfn(pg))
+#undef mfn_to_page
+#define mfn_to_page(mfn) __mfn_to_page(_mfn(mfn))
+
 unsigned int __read_mostly m2p_compat_vstart = __HYPERVISOR_COMPAT_VIRT_START;
 
 l2_pgentry_t *compat_idle_pg_table_l2;
diff --git a/xen/common/domain.c b/xen/common/domain.c
index 5aebcf265f..e8302e8e1b 100644
--- a/xen/common/domain.c
+++ b/xen/common/domain.c
@@ -1192,7 +1192,7 @@ int map_vcpu_info(struct vcpu *v, unsigned long gfn, unsigned offset)
     }
 
     v->vcpu_info = new_info;
-    v->vcpu_info_mfn = _mfn(page_to_mfn(page));
+    v->vcpu_info_mfn = page_to_mfn(page);
 
     /* Set new vcpu_info pointer /before/ setting pending flags. */
     smp_wmb();
@@ -1225,7 +1225,7 @@ void unmap_vcpu_info(struct vcpu *v)
 
     vcpu_info_reset(v); /* NB: Clobbers v->vcpu_info_mfn */
 
-    put_page_and_type(mfn_to_page(mfn_x(mfn)));
+    put_page_and_type(mfn_to_page(mfn));
 }
 
 int default_initialise_vcpu(struct vcpu *v, XEN_GUEST_HANDLE_PARAM(void) arg)
diff --git a/xen/common/event_fifo.c b/xen/common/event_fifo.c
index c49f446754..f15adf0eb5 100644
--- a/xen/common/event_fifo.c
+++ b/xen/common/event_fifo.c
@@ -389,7 +389,7 @@ static void unmap_guest_page(void *virt)
         return;
 
     virt = (void *)((unsigned long)virt & PAGE_MASK);
-    page = mfn_to_page(domain_page_map_to_mfn(virt));
+    page = mfn_to_page(_mfn(domain_page_map_to_mfn(virt)));
 
     unmap_domain_page_global(virt);
     put_page_and_type(page);
diff --git a/xen/common/grant_table.c b/xen/common/grant_table.c
index 6d20b17739..2afde596d9 100644
--- a/xen/common/grant_table.c
+++ b/xen/common/grant_table.c
@@ -40,6 +40,12 @@
 #include <xsm/xsm.h>
 #include <asm/flushtlb.h>
 
+/* Override macros from asm/page.h to avoid using typesafe mfn_t. */
+#undef page_to_mfn
+#define page_to_mfn(pg) mfn_x(__page_to_mfn(pg))
+#undef mfn_to_page
+#define mfn_to_page(mfn) __mfn_to_page(_mfn(mfn))
+
 /* Per-domain grant information. */
 struct grant_table {
     /*
diff --git a/xen/common/kimage.c b/xen/common/kimage.c
index 07587896a4..93c7da5c20 100644
--- a/xen/common/kimage.c
+++ b/xen/common/kimage.c
@@ -76,7 +76,7 @@ static struct page_info *kimage_alloc_zeroed_page(unsigned memflags)
     if ( !page )
         return NULL;
 
-    clear_domain_page(_mfn(page_to_mfn(page)));
+    clear_domain_page(page_to_mfn(page));
 
     return page;
 }
@@ -405,7 +405,7 @@ static struct page_info *kimage_alloc_crash_control_page(struct kexec_image *ima
     if ( page )
     {
         image->next_crash_page = hole_end;
-        clear_domain_page(_mfn(page_to_mfn(page)));
+        clear_domain_page(page_to_mfn(page));
     }
 
     return page;
@@ -641,7 +641,7 @@ static struct page_info *kimage_alloc_page(struct kexec_image *image,
             *old = (addr & ~PAGE_MASK) | IND_SOURCE;
             unmap_domain_page(old);
 
-            page = mfn_to_page(mfn_x(old_mfn));
+            page = mfn_to_page(old_mfn);
             break;
         }
         else
@@ -873,22 +873,22 @@ int kimage_build_ind(struct kexec_image *image, unsigned long ind_mfn,
     for ( entry = page; ;  )
     {
         unsigned long ind;
-        unsigned long mfn;
+        mfn_t mfn;
 
         ind = kimage_entry_ind(entry, compat);
-        mfn = kimage_entry_mfn(entry, compat);
+        mfn = _mfn(kimage_entry_mfn(entry, compat));
 
         switch ( ind )
         {
         case IND_DESTINATION:
-            dest = (paddr_t)mfn << PAGE_SHIFT;
+            dest = mfn_to_maddr(mfn);
             ret = kimage_set_destination(image, dest);
             if ( ret < 0 )
                 goto done;
             break;
         case IND_INDIRECTION:
             unmap_domain_page(page);
-            page = map_domain_page(_mfn(mfn));
+            page = map_domain_page(mfn);
             entry = page;
             continue;
         case IND_DONE:
@@ -913,7 +913,7 @@ int kimage_build_ind(struct kexec_image *image, unsigned long ind_mfn,
                 goto done;
             }
 
-            copy_domain_page(_mfn(page_to_mfn(xen_page)), _mfn(mfn));
+            copy_domain_page(page_to_mfn(xen_page), mfn);
             put_page(guest_page);
 
             ret = kimage_add_page(image, page_to_maddr(xen_page));
diff --git a/xen/common/memory.c b/xen/common/memory.c
index ad987e0f29..e467f271c7 100644
--- a/xen/common/memory.c
+++ b/xen/common/memory.c
@@ -29,6 +29,12 @@
 #include <public/memory.h>
 #include <xsm/xsm.h>
 
+/* Override macros from asm/page.h to avoid using typesafe mfn_t. */
+#undef page_to_mfn
+#define page_to_mfn(pg) mfn_x(__page_to_mfn(pg))
+#undef mfn_to_page
+#define mfn_to_page(mfn) __mfn_to_page(_mfn(mfn))
+
 struct memop_args {
     /* INPUT */
     struct domain *domain;     /* Domain to be affected. */
diff --git a/xen/common/page_alloc.c b/xen/common/page_alloc.c
index 472c6fe329..5e7d74e274 100644
--- a/xen/common/page_alloc.c
+++ b/xen/common/page_alloc.c
@@ -150,6 +150,12 @@
 #define p2m_pod_offline_or_broken_replace(pg) BUG_ON(pg != NULL)
 #endif
 
+/* Override macros from asm/page.h to avoid using typesafe mfn_t. */
+#undef page_to_mfn
+#define page_to_mfn(pg) mfn_x(__page_to_mfn(pg))
+#undef mfn_to_page
+#define mfn_to_page(mfn) __mfn_to_page(_mfn(mfn))
+
 /*
  * Comma-separated list of hexadecimal page numbers containing bad bytes.
  * e.g. 'badpage=0x3f45,0x8a321'.
diff --git a/xen/common/tmem.c b/xen/common/tmem.c
index c955cf7167..1adb96f00c 100644
--- a/xen/common/tmem.c
+++ b/xen/common/tmem.c
@@ -243,7 +243,7 @@ static void tmem_persistent_pool_page_put(void *page_va)
     struct page_info *pi;
 
     ASSERT(IS_PAGE_ALIGNED(page_va));
-    pi = mfn_to_page(virt_to_mfn(page_va));
+    pi = mfn_to_page(_mfn(virt_to_mfn(page_va)));
     ASSERT(IS_VALID_PAGE(pi));
     __tmem_free_page_thispool(pi);
 }
diff --git a/xen/common/tmem_xen.c b/xen/common/tmem_xen.c
index 8dc031514a..9131fd9d79 100644
--- a/xen/common/tmem_xen.c
+++ b/xen/common/tmem_xen.c
@@ -14,10 +14,6 @@
 #include <xen/cpu.h>
 #include <xen/init.h>
 
-/* Override macros from asm/page.h to make them work with mfn_t */
-#undef page_to_mfn
-#define page_to_mfn(pg) _mfn(__page_to_mfn(pg))
-
 bool __read_mostly opt_tmem;
 boolean_param("tmem", opt_tmem);
 
diff --git a/xen/common/trace.c b/xen/common/trace.c
index 2e18702317..cf8f8b0997 100644
--- a/xen/common/trace.c
+++ b/xen/common/trace.c
@@ -42,6 +42,12 @@ CHECK_t_buf;
 #define compat_t_rec t_rec
 #endif
 
+/* Override macros from asm/page.h to avoid using typesafe mfn_t. */
+#undef page_to_mfn
+#define page_to_mfn(pg) mfn_x(__page_to_mfn(pg))
+#undef mfn_to_page
+#define mfn_to_page(mfn) __mfn_to_page(_mfn(mfn))
+
 /* opt_tbuf_size: trace buffer size (in pages) for each cpu */
 static unsigned int opt_tbuf_size;
 static unsigned int opt_tevt_mask;
diff --git a/xen/common/vmap.c b/xen/common/vmap.c
index 0b23f8fb97..10f32b29e0 100644
--- a/xen/common/vmap.c
+++ b/xen/common/vmap.c
@@ -36,7 +36,7 @@ void __init vm_init_type(enum vmap_region type, void *start, void *end)
     {
         struct page_info *pg = alloc_domheap_page(NULL, 0);
 
-        map_pages_to_xen(va, page_to_mfn(pg), 1, PAGE_HYPERVISOR);
+        map_pages_to_xen(va, mfn_x(page_to_mfn(pg)), 1, PAGE_HYPERVISOR);
         clear_page((void *)va);
     }
     bitmap_fill(vm_bitmap(type), vm_low[type]);
@@ -107,7 +107,8 @@ static void *vm_alloc(unsigned int nr, unsigned int align,
         {
             unsigned long va = (unsigned long)vm_bitmap(t) + vm_top[t] / 8;
 
-            if ( !map_pages_to_xen(va, page_to_mfn(pg), 1, PAGE_HYPERVISOR) )
+            if ( !map_pages_to_xen(va, mfn_x(page_to_mfn(pg)),
+                                   1, PAGE_HYPERVISOR) )
             {
                 clear_page((void *)va);
                 vm_top[t] += PAGE_SIZE * 8;
@@ -258,7 +259,7 @@ static void *vmalloc_type(size_t size, enum vmap_region type)
         pg = alloc_domheap_page(NULL, 0);
         if ( pg == NULL )
             goto error;
-        mfn[i] = _mfn(page_to_mfn(pg));
+        mfn[i] = page_to_mfn(pg);
     }
 
     va = __vmap(mfn, 1, pages, 1, PAGE_HYPERVISOR, type);
@@ -270,7 +271,7 @@ static void *vmalloc_type(size_t size, enum vmap_region type)
 
  error:
     while ( i-- )
-        free_domheap_page(mfn_to_page(mfn_x(mfn[i])));
+        free_domheap_page(mfn_to_page(mfn[i]));
     xfree(mfn);
     return NULL;
 }
diff --git a/xen/common/xenoprof.c b/xen/common/xenoprof.c
index 98937c9ac6..1f547bca52 100644
--- a/xen/common/xenoprof.c
+++ b/xen/common/xenoprof.c
@@ -22,8 +22,6 @@
 /* Override macros from asm/page.h to make them work with mfn_t */
 #undef virt_to_mfn
 #define virt_to_mfn(va) _mfn(__virt_to_mfn(va))
-#undef mfn_to_page
-#define mfn_to_page(mfn) __mfn_to_page(mfn_x(mfn))
 
 /* Limit amount of pages used for shared buffer (per domain) */
 #define MAX_OPROF_SHARED_PAGES 32
diff --git a/xen/drivers/passthrough/amd/iommu_map.c b/xen/drivers/passthrough/amd/iommu_map.c
index fd2327d3e5..bd62c2ce90 100644
--- a/xen/drivers/passthrough/amd/iommu_map.c
+++ b/xen/drivers/passthrough/amd/iommu_map.c
@@ -25,6 +25,12 @@
 #include "../ats.h"
 #include <xen/pci.h>
 
+/* Override macros from asm/page.h to avoid using typesafe mfn_t. */
+#undef page_to_mfn
+#define page_to_mfn(pg) mfn_x(__page_to_mfn(pg))
+#undef mfn_to_page
+#define mfn_to_page(mfn) __mfn_to_page(_mfn(mfn))
+
 /* Given pfn and page table level, return pde index */
 static unsigned int pfn_to_pde_idx(unsigned long pfn, unsigned int level)
 {
diff --git a/xen/drivers/passthrough/iommu.c b/xen/drivers/passthrough/iommu.c
index 1aecf7cf34..2c44fabf99 100644
--- a/xen/drivers/passthrough/iommu.c
+++ b/xen/drivers/passthrough/iommu.c
@@ -184,7 +184,7 @@ void __hwdom_init iommu_hwdom_init(struct domain *d)
 
         page_list_for_each ( page, &d->page_list )
         {
-            unsigned long mfn = page_to_mfn(page);
+            unsigned long mfn = mfn_x(page_to_mfn(page));
             unsigned long gfn = mfn_to_gmfn(d, mfn);
             unsigned int mapping = IOMMUF_readable;
             int ret;
diff --git a/xen/drivers/passthrough/x86/iommu.c b/xen/drivers/passthrough/x86/iommu.c
index 0253823173..68182afd91 100644
--- a/xen/drivers/passthrough/x86/iommu.c
+++ b/xen/drivers/passthrough/x86/iommu.c
@@ -58,7 +58,7 @@ int arch_iommu_populate_page_table(struct domain *d)
         if ( is_hvm_domain(d) ||
             (page->u.inuse.type_info & PGT_type_mask) == PGT_writable_page )
         {
-            unsigned long mfn = page_to_mfn(page);
+            unsigned long mfn = mfn_x(page_to_mfn(page));
             unsigned long gfn = mfn_to_gmfn(d, mfn);
 
             if ( gfn != gfn_x(INVALID_GFN) )
diff --git a/xen/include/asm-arm/mm.h b/xen/include/asm-arm/mm.h
index 737a429409..3eb4b68761 100644
--- a/xen/include/asm-arm/mm.h
+++ b/xen/include/asm-arm/mm.h
@@ -138,7 +138,7 @@ extern vaddr_t xenheap_virt_start;
 #endif
 
 #ifdef CONFIG_ARM_32
-#define is_xen_heap_page(page) is_xen_heap_mfn(page_to_mfn(page))
+#define is_xen_heap_page(page) is_xen_heap_mfn(mfn_x(__page_to_mfn(page)))
 #define is_xen_heap_mfn(mfn) ({                                 \
     unsigned long mfn_ = (mfn);                                 \
     (mfn_ >= mfn_x(xenheap_mfn_start) &&                        \
@@ -220,12 +220,14 @@ static inline void __iomem *ioremap_wc(paddr_t start, size_t len)
 })
 
 /* Convert between machine frame numbers and page-info structures. */
-#define __mfn_to_page(mfn)  (frame_table + (pfn_to_pdx(mfn) - frametable_base_pdx))
-#define __page_to_mfn(pg)   pdx_to_pfn((unsigned long)((pg) - frame_table) + frametable_base_pdx)
+#define __mfn_to_page(mfn)                                          \
+    (frame_table + (pfn_to_pdx(mfn_x(mfn)) - frametable_base_pdx))
+#define __page_to_mfn(pg)                                           \
+    _mfn(pdx_to_pfn((unsigned long)((pg) - frame_table) + frametable_base_pdx))
 
 /* Convert between machine addresses and page-info structures. */
-#define maddr_to_page(ma) __mfn_to_page((ma) >> PAGE_SHIFT)
-#define page_to_maddr(pg) ((paddr_t)__page_to_mfn(pg) << PAGE_SHIFT)
+#define maddr_to_page(ma) __mfn_to_page(maddr_to_mfn(ma))
+#define page_to_maddr(pg) (mfn_to_maddr(__page_to_mfn(pg)))
 
 /* Convert between frame number and address formats.  */
 #define pfn_to_paddr(pfn) ((paddr_t)(pfn) << PAGE_SHIFT)
@@ -235,7 +237,7 @@ static inline void __iomem *ioremap_wc(paddr_t start, size_t len)
 #define gaddr_to_gfn(ga)    _gfn(paddr_to_pfn(ga))
 #define mfn_to_maddr(mfn)   pfn_to_paddr(mfn_x(mfn))
 #define maddr_to_mfn(ma)    _mfn(paddr_to_pfn(ma))
-#define vmap_to_mfn(va)     paddr_to_pfn(virt_to_maddr((vaddr_t)va))
+#define vmap_to_mfn(va)     maddr_to_mfn(virt_to_maddr((vaddr_t)va))
 #define vmap_to_page(va)    mfn_to_page(vmap_to_mfn(va))
 
 /* Page-align address and convert to frame number format */
@@ -309,7 +311,7 @@ static inline struct page_info *virt_to_page(const void *v)
 
 static inline void *page_to_virt(const struct page_info *pg)
 {
-    return mfn_to_virt(page_to_mfn(pg));
+    return mfn_to_virt(mfn_x(__page_to_mfn(pg)));
 }
 
 struct page_info *get_page_from_gva(struct vcpu *v, vaddr_t va,
diff --git a/xen/include/asm-arm/p2m.h b/xen/include/asm-arm/p2m.h
index faadcfe8fe..87c9994974 100644
--- a/xen/include/asm-arm/p2m.h
+++ b/xen/include/asm-arm/p2m.h
@@ -276,7 +276,7 @@ static inline struct page_info *get_page_from_gfn(
 {
     struct page_info *page;
     p2m_type_t p2mt;
-    unsigned long mfn = mfn_x(p2m_lookup(d, _gfn(gfn), &p2mt));
+    mfn_t mfn = p2m_lookup(d, _gfn(gfn), &p2mt);
 
     if (t)
         *t = p2mt;
@@ -284,7 +284,7 @@ static inline struct page_info *get_page_from_gfn(
     if ( !p2m_is_any_ram(p2mt) )
         return NULL;
 
-    if ( !mfn_valid(_mfn(mfn)) )
+    if ( !mfn_valid(mfn) )
         return NULL;
     page = mfn_to_page(mfn);
 
diff --git a/xen/include/asm-x86/mm.h b/xen/include/asm-x86/mm.h
index f2e0f498c4..984f54c3fa 100644
--- a/xen/include/asm-x86/mm.h
+++ b/xen/include/asm-x86/mm.h
@@ -259,7 +259,7 @@ struct page_info
 
 #define is_xen_heap_page(page) ((page)->count_info & PGC_xen_heap)
 #define is_xen_heap_mfn(mfn) \
-    (__mfn_valid(mfn) && is_xen_heap_page(__mfn_to_page(mfn)))
+    (__mfn_valid(mfn) && is_xen_heap_page(__mfn_to_page(_mfn(mfn))))
 #define is_xen_fixed_mfn(mfn)                     \
     ((((mfn) << PAGE_SHIFT) >= __pa(&_stext)) &&  \
      (((mfn) << PAGE_SHIFT) <= __pa(&__2M_rwdata_end)))
@@ -369,7 +369,7 @@ void put_page_from_l1e(l1_pgentry_t l1e, struct domain *l1e_owner);
 
 static inline bool get_page_from_mfn(mfn_t mfn, struct domain *d)
 {
-    struct page_info *page = __mfn_to_page(mfn_x(mfn));
+    struct page_info *page = __mfn_to_page(mfn);
 
     if ( unlikely(!mfn_valid(mfn)) || unlikely(!get_page(page, d)) )
     {
@@ -463,10 +463,10 @@ extern paddr_t mem_hotplug;
 #define SHARED_M2P(_e)           ((_e) == SHARED_M2P_ENTRY)
 
 #define compat_machine_to_phys_mapping ((unsigned int *)RDWR_COMPAT_MPT_VIRT_START)
-#define _set_gpfn_from_mfn(mfn, pfn) ({                        \
-    struct domain *d = page_get_owner(__mfn_to_page(mfn));     \
-    unsigned long entry = (d && (d == dom_cow)) ?              \
-        SHARED_M2P_ENTRY : (pfn);                              \
+#define _set_gpfn_from_mfn(mfn, pfn) ({                         \
+    struct domain *d = page_get_owner(__mfn_to_page(_mfn(mfn)));    \
+    unsigned long entry = (d && (d == dom_cow)) ?               \
+        SHARED_M2P_ENTRY : (pfn);                               \
     ((void)((mfn) >= (RDWR_COMPAT_MPT_VIRT_END - RDWR_COMPAT_MPT_VIRT_START) / 4 || \
             (compat_machine_to_phys_mapping[(mfn)] = (unsigned int)(entry))), \
      machine_to_phys_mapping[(mfn)] = (entry));                \
diff --git a/xen/include/asm-x86/p2m.h b/xen/include/asm-x86/p2m.h
index 70f00c332f..18eac537c9 100644
--- a/xen/include/asm-x86/p2m.h
+++ b/xen/include/asm-x86/p2m.h
@@ -480,7 +480,7 @@ static inline struct page_info *get_page_from_gfn(
     /* Non-translated guests see 1-1 RAM / MMIO mappings everywhere */
     if ( t )
         *t = likely(d != dom_io) ? p2m_ram_rw : p2m_mmio_direct;
-    page = __mfn_to_page(gfn);
+    page = __mfn_to_page(_mfn(gfn));
     return mfn_valid(_mfn(gfn)) && get_page(page, d) ? page : NULL;
 }
 
diff --git a/xen/include/asm-x86/page.h b/xen/include/asm-x86/page.h
index 45ca742678..8737ef16ff 100644
--- a/xen/include/asm-x86/page.h
+++ b/xen/include/asm-x86/page.h
@@ -88,10 +88,10 @@
     ((paddr_t)(((x).l4 & (PADDR_MASK&PAGE_MASK))))
 
 /* Get pointer to info structure of page mapped by pte (struct page_info *). */
-#define l1e_get_page(x)           (__mfn_to_page(l1e_get_pfn(x)))
-#define l2e_get_page(x)           (__mfn_to_page(l2e_get_pfn(x)))
-#define l3e_get_page(x)           (__mfn_to_page(l3e_get_pfn(x)))
-#define l4e_get_page(x)           (__mfn_to_page(l4e_get_pfn(x)))
+#define l1e_get_page(x)           (__mfn_to_page(l1e_get_mfn(x)))
+#define l2e_get_page(x)           (__mfn_to_page(l2e_get_mfn(x)))
+#define l3e_get_page(x)           (__mfn_to_page(l3e_get_mfn(x)))
+#define l4e_get_page(x)           (__mfn_to_page(l4e_get_mfn(x)))
 
 /* Get pte access flags (unsigned int). */
 #define l1e_get_flags(x)           (get_pte_flags((x).l1))
@@ -157,10 +157,10 @@ static inline l4_pgentry_t l4e_from_paddr(paddr_t pa, unsigned int flags)
 #define l4e_from_intpte(intpte)    ((l4_pgentry_t) { (intpte_t)(intpte) })
 
 /* Construct a pte from a page pointer and access flags. */
-#define l1e_from_page(page, flags) l1e_from_pfn(__page_to_mfn(page), (flags))
-#define l2e_from_page(page, flags) l2e_from_pfn(__page_to_mfn(page), (flags))
-#define l3e_from_page(page, flags) l3e_from_pfn(__page_to_mfn(page), (flags))
-#define l4e_from_page(page, flags) l4e_from_pfn(__page_to_mfn(page), (flags))
+#define l1e_from_page(page, flags) l1e_from_mfn(__page_to_mfn(page), (flags))
+#define l2e_from_page(page, flags) l2e_from_mfn(__page_to_mfn(page), (flags))
+#define l3e_from_page(page, flags) l3e_from_mfn(__page_to_mfn(page), (flags))
+#define l4e_from_page(page, flags) l4e_from_mfn(__page_to_mfn(page), (flags))
 
 /* Add extra flags to an existing pte. */
 #define l1e_add_flags(x, flags)    ((x).l1 |= put_pte_flags(flags))
@@ -215,13 +215,13 @@ static inline l4_pgentry_t l4e_from_paddr(paddr_t pa, unsigned int flags)
 /* Page-table type. */
 typedef struct { u64 pfn; } pagetable_t;
 #define pagetable_get_paddr(x)  ((paddr_t)(x).pfn << PAGE_SHIFT)
-#define pagetable_get_page(x)   __mfn_to_page((x).pfn)
+#define pagetable_get_page(x)   __mfn_to_page(pagetable_get_mfn(x))
 #define pagetable_get_pfn(x)    ((x).pfn)
 #define pagetable_get_mfn(x)    _mfn(((x).pfn))
 #define pagetable_is_null(x)    ((x).pfn == 0)
 #define pagetable_from_pfn(pfn) ((pagetable_t) { (pfn) })
 #define pagetable_from_mfn(mfn) ((pagetable_t) { mfn_x(mfn) })
-#define pagetable_from_page(pg) pagetable_from_pfn(__page_to_mfn(pg))
+#define pagetable_from_page(pg) pagetable_from_mfn(__page_to_mfn(pg))
 #define pagetable_from_paddr(p) pagetable_from_pfn((p)>>PAGE_SHIFT)
 #define pagetable_null()        pagetable_from_pfn(0)
 
@@ -240,12 +240,12 @@ void copy_page_sse2(void *, const void *);
 #define __mfn_to_virt(mfn)  (maddr_to_virt((paddr_t)(mfn) << PAGE_SHIFT))
 
 /* Convert between machine frame numbers and page-info structures. */
-#define __mfn_to_page(mfn)  (frame_table + pfn_to_pdx(mfn))
-#define __page_to_mfn(pg)   pdx_to_pfn((unsigned long)((pg) - frame_table))
+#define __mfn_to_page(mfn)  (frame_table + pfn_to_pdx(mfn_x(mfn)))
+#define __page_to_mfn(pg)   _mfn(pdx_to_pfn((unsigned long)((pg) - frame_table)))
 
 /* Convert between machine addresses and page-info structures. */
-#define __maddr_to_page(ma) __mfn_to_page((ma) >> PAGE_SHIFT)
-#define __page_to_maddr(pg) ((paddr_t)__page_to_mfn(pg) << PAGE_SHIFT)
+#define __maddr_to_page(ma) __mfn_to_page(maddr_to_mfn(ma))
+#define __page_to_maddr(pg) (mfn_to_maddr(__page_to_mfn(pg)))
 
 /* Convert between frame number and address formats.  */
 #define __pfn_to_paddr(pfn) ((paddr_t)(pfn) << PAGE_SHIFT)
@@ -273,8 +273,8 @@ void copy_page_sse2(void *, const void *);
 #define pfn_to_paddr(pfn)   __pfn_to_paddr(pfn)
 #define paddr_to_pfn(pa)    __paddr_to_pfn(pa)
 #define paddr_to_pdx(pa)    pfn_to_pdx(paddr_to_pfn(pa))
-#define vmap_to_mfn(va)     l1e_get_pfn(*virt_to_xen_l1e((unsigned long)(va)))
-#define vmap_to_page(va)    mfn_to_page(vmap_to_mfn(va))
+#define vmap_to_mfn(va)     _mfn(l1e_get_pfn(*virt_to_xen_l1e((unsigned long)(va))))
+#define vmap_to_page(va)    __mfn_to_page(vmap_to_mfn(va))
 
 #endif /* !defined(__ASSEMBLY__) */
 
diff --git a/xen/include/xen/domain_page.h b/xen/include/xen/domain_page.h
index 890bae5b9c..5d1c34528e 100644
--- a/xen/include/xen/domain_page.h
+++ b/xen/include/xen/domain_page.h
@@ -44,11 +44,11 @@ unsigned long domain_page_map_to_mfn(const void *va);
 void *map_domain_page_global(mfn_t mfn);
 void unmap_domain_page_global(const void *va);
 
-#define __map_domain_page(pg)        map_domain_page(_mfn(__page_to_mfn(pg)))
+#define __map_domain_page(pg)        map_domain_page(__page_to_mfn(pg))
 
 static inline void *__map_domain_page_global(const struct page_info *pg)
 {
-    return map_domain_page_global(_mfn(__page_to_mfn(pg)));
+    return map_domain_page_global(page_to_mfn(pg));
 }
 
 #else /* !CONFIG_DOMAIN_PAGE */
diff --git a/xen/include/xen/tmem_xen.h b/xen/include/xen/tmem_xen.h
index 542c0b3f20..8516a0b131 100644
--- a/xen/include/xen/tmem_xen.h
+++ b/xen/include/xen/tmem_xen.h
@@ -25,7 +25,7 @@
 typedef uint32_t pagesize_t;  /* like size_t, must handle largest PAGE_SIZE */
 
 #define IS_PAGE_ALIGNED(addr) IS_ALIGNED((unsigned long)(addr), PAGE_SIZE)
-#define IS_VALID_PAGE(_pi)    mfn_valid(_mfn(page_to_mfn(_pi)))
+#define IS_VALID_PAGE(_pi)    mfn_valid(page_to_mfn(_pi))
 
 extern struct page_list_head tmem_page_list;
 extern spinlock_t tmem_page_list_lock;
-- 
2.11.0


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

Re: [PATCH 7/7] xen: Convert __page_to_mfn and __mfn_to_page to use typesafe MFN

[Razvan Cojocaru]

On 10/04/2017 09:15 PM, Julien Grall wrote:
> Most of the users of page_to_mfn and mfn_to_page are either overriding
> the macros to make them work with mfn_t or use mfn_x/_mfn because the
> rest of the function use mfn_t.
> 
> So make __page_to_mfn and __mfn_to_page return mfn_t by default.
> 
> Only reasonable clean-ups are done in this patch because it is
> already quite big. So some of the files now override page_to_mfn and
> mfn_to_page to avoid using mfn_t.
> 
> Signed-off-by: Julien Grall <julien.grall@linaro.org>

Acked-by: Razvan Cojocaru <rcojocaru@bitdefender.com>


Thanks,
Razvan

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

Re: [PATCH 7/7] xen: Convert __page_to_mfn and __mfn_to_page to use typesafe MFN

[Tamas K Lengyel]

On Wed, Oct 4, 2017 at 12:15 PM, Julien Grall <julien.grall@linaro.org> wrote:
> Most of the users of page_to_mfn and mfn_to_page are either overriding
> the macros to make them work with mfn_t or use mfn_x/_mfn because the
> rest of the function use mfn_t.
>
> So make __page_to_mfn and __mfn_to_page return mfn_t by default.
>
> Only reasonable clean-ups are done in this patch because it is
> already quite big. So some of the files now override page_to_mfn and
> mfn_to_page to avoid using mfn_t.
>
> Signed-off-by: Julien Grall <julien.grall@linaro.org>

Acked-by: Tamas K Lengyel <tamas@tklengyel.com>

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

Re: [PATCH 7/7] xen: Convert __page_to_mfn and __mfn_to_page to use typesafe MFN

[Andrew Cooper]

On 04/10/2017 19:15, Julien Grall wrote:
> diff --git a/xen/arch/arm/domain_build.c b/xen/arch/arm/domain_build.c
> index 093ebf1a8e..0753d03aac 100644
> --- a/xen/arch/arm/domain_build.c
> +++ b/xen/arch/arm/domain_build.c
> @@ -104,11 +104,11 @@ static bool insert_11_bank(struct domain *d,
>                             unsigned int order)
>  {
>      int res, i;
> -    paddr_t spfn;
> +    mfn_t smfn;
>      paddr_t start, size;
>  
> -    spfn = page_to_mfn(pg);
> -    start = pfn_to_paddr(spfn);
> +    smfn = page_to_mfn(pg);
> +    start = mfn_to_maddr(smfn);
>      size = pfn_to_paddr(1UL << order);

Wouldn't it be cleaner to move this renaming into patch 1, along with an
extra set of undef/override, to be taken out here?  (perhaps not given
the rework effort?)

>  
>      D11PRINT("Allocated %#"PRIpaddr"-%#"PRIpaddr" (%ldMB/%ldMB, order %d)\n",
> @@ -678,8 +678,8 @@ static void pvpmu_finish(struct domain *d, xen_pmu_params_t *params)
>  
>      if ( xenpmu_data )
>      {
> -        mfn = domain_page_map_to_mfn(xenpmu_data);
> -        ASSERT(mfn_valid(_mfn(mfn)));
> +        mfn = _mfn(domain_page_map_to_mfn(xenpmu_data));

Seeing as you convert every(?) call to domain_page_map_to_mfn(), it
would be cleaner to change the return type while making the change.

I'd be happy for such a change being folded into this patch, because
doing so would be by far the least disruptive way of making the change.

> +        ASSERT(mfn_valid(mfn));
>          unmap_domain_page_global(xenpmu_data);
>          put_page_and_type(mfn_to_page(mfn));
>      }
> @@ -1185,8 +1180,8 @@ int __mem_sharing_unshare_page(struct domain *d,
>          return -ENOMEM;
>      }
>  
> -    s = map_domain_page(_mfn(__page_to_mfn(old_page)));
> -    t = map_domain_page(_mfn(__page_to_mfn(page)));
> +    s = map_domain_page(page_to_mfn(old_page));
> +    t = map_domain_page(page_to_mfn(page));
>      memcpy(t, s, PAGE_SIZE);
>      unmap_domain_page(s);
>      unmap_domain_page(t);

This whole lot could turn into copy_domain_page()

> diff --git a/xen/arch/x86/pv/descriptor-tables.c b/xen/arch/x86/pv/descriptor-tables.c
> index 81973af124..371221a302 100644
> --- a/xen/arch/x86/pv/descriptor-tables.c
> +++ b/xen/arch/x86/pv/descriptor-tables.c
> @@ -25,12 +25,6 @@
>  #include <asm/p2m.h>
>  #include <asm/pv/mm.h>
>  
> -/* Override macros from asm/page.h to make them work with mfn_t */
> -#undef mfn_to_page
> -#define mfn_to_page(mfn) __mfn_to_page(mfn_x(mfn))
> -#undef page_to_mfn
> -#define page_to_mfn(pg) _mfn(__page_to_mfn(pg))
> -
>  /*******************
>   * Descriptor Tables
>   */

If you're making this change, please take out the Descriptor Tables
comment like you do with I/O below, because the entire file is dedicated
to descriptor table support and it will save me one item on a cleanup
patch :).

> diff --git a/xen/arch/x86/pv/emul-priv-op.c b/xen/arch/x86/pv/emul-priv-op.c
> index dd90713acf..9ccbd021ef 100644
> --- a/xen/arch/x86/pv/emul-priv-op.c
> +++ b/xen/arch/x86/pv/emul-priv-op.c
> @@ -43,16 +43,6 @@
>  #include "emulate.h"
>  #include "mm.h"
>  
> -/* Override macros from asm/page.h to make them work with mfn_t */
> -#undef mfn_to_page
> -#define mfn_to_page(mfn) __mfn_to_page(mfn_x(mfn))
> -#undef page_to_mfn
> -#define page_to_mfn(pg) _mfn(__page_to_mfn(pg))
> -
> -/***********************
> - * I/O emulation support
> - */
> -
>  struct priv_op_ctxt {
>      struct x86_emulate_ctxt ctxt;
>      struct {
> @@ -873,22 +873,22 @@ int kimage_build_ind(struct kexec_image *image, unsigned long ind_mfn,
>      for ( entry = page; ;  )
>      {
>          unsigned long ind;
> -        unsigned long mfn;
> +        mfn_t mfn;
>  
>          ind = kimage_entry_ind(entry, compat);
> -        mfn = kimage_entry_mfn(entry, compat);
> +        mfn = _mfn(kimage_entry_mfn(entry, compat));

Again, modify the return type of kimage_entry_mfn() ?

> diff --git a/xen/include/asm-x86/page.h b/xen/include/asm-x86/page.h
> index 45ca742678..8737ef16ff 100644
> --- a/xen/include/asm-x86/page.h
> +++ b/xen/include/asm-x86/page.h
> @@ -273,8 +273,8 @@ void copy_page_sse2(void *, const void *);
>  #define pfn_to_paddr(pfn)   __pfn_to_paddr(pfn)
>  #define paddr_to_pfn(pa)    __paddr_to_pfn(pa)
>  #define paddr_to_pdx(pa)    pfn_to_pdx(paddr_to_pfn(pa))
> -#define vmap_to_mfn(va)     l1e_get_pfn(*virt_to_xen_l1e((unsigned long)(va)))
> -#define vmap_to_page(va)    mfn_to_page(vmap_to_mfn(va))
> +#define vmap_to_mfn(va)     _mfn(l1e_get_pfn(*virt_to_xen_l1e((unsigned long)(va))))

l1e_get_mfn(*virt_to_xen_l1e((unsigned long)(va)))

> +#define vmap_to_page(va)    __mfn_to_page(vmap_to_mfn(va))
>  
>  #endif /* !defined(__ASSEMBLY__) */
>  
> diff --git a/xen/include/xen/tmem_xen.h b/xen/include/xen/tmem_xen.h
> index 542c0b3f20..8516a0b131 100644
> --- a/xen/include/xen/tmem_xen.h
> +++ b/xen/include/xen/tmem_xen.h
> @@ -25,7 +25,7 @@
>  typedef uint32_t pagesize_t;  /* like size_t, must handle largest PAGE_SIZE */
>  
>  #define IS_PAGE_ALIGNED(addr) IS_ALIGNED((unsigned long)(addr), PAGE_SIZE)
> -#define IS_VALID_PAGE(_pi)    mfn_valid(_mfn(page_to_mfn(_pi)))
> +#define IS_VALID_PAGE(_pi)    mfn_valid(page_to_mfn(_pi))

/sigh  This is tautological.  The definition of a "valid mfn" in this
case is one for which we have frametable entry, and by having a struct
page_info in our hands, this is by definition true (unless you have a
wild pointer, at which point your bug is elsewhere).

IS_VALID_PAGE() is only ever used in assertions and never usefully, so
instead I would remove it entirely rather than trying to fix it up.

As for TMEM itself (Julien: This my no means blocks the patch.  It is
more an observation for Konrad to see about fixing), I see that TMEM is
broken on x86 machines with more than 5TB of RAM, because it is not
legal to call page_to_virt() on a struct page_info allocated from the
domheap (which is why alloc_xenheap_page() returns a void *, and
alloc_domheap_page() specifically doesn't).  The easy fix for this is to
swap the allocation primitives over to using xenheap allocations, which
would remove the need for page_to_virt() and back, or a better fix would
be to not pass everything thing by virtual address (at which point
retaining use of the domheap is fine).

~Andrew

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

Re: [PATCH 7/7] xen: Convert __page_to_mfn and __mfn_to_page to use typesafe MFN

[Jan Beulich]

>>> On 05.10.17 at 01:27, <andrew.cooper3@citrix.com> wrote:
> As for TMEM itself (Julien: This my no means blocks the patch.  It is
> more an observation for Konrad to see about fixing), I see that TMEM is
> broken on x86 machines with more than 5TB of RAM, because it is not
> legal to call page_to_virt() on a struct page_info allocated from the
> domheap (which is why alloc_xenheap_page() returns a void *, and
> alloc_domheap_page() specifically doesn't).  The easy fix for this is to
> swap the allocation primitives over to using xenheap allocations, which
> would remove the need for page_to_virt() and back, or a better fix would
> be to not pass everything thing by virtual address (at which point
> retaining use of the domheap is fine).

For this reason we have

        if ( tmem_enabled() )
        {
           printk(XENLOG_WARNING
                  "TMEM physical RAM limit exceeded, disabling TMEM\n");
           tmem_disable();
        }

in setup.c.

Jan


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

Re: [PATCH 7/7] xen: Convert __page_to_mfn and __mfn_to_page to use typesafe MFN

[Julien Grall]

Hi Andrew,

On 05/10/17 00:27, Andrew Cooper wrote:
> On 04/10/2017 19:15, Julien Grall wrote:
>> diff --git a/xen/arch/arm/domain_build.c b/xen/arch/arm/domain_build.c
>> index 093ebf1a8e..0753d03aac 100644
>> --- a/xen/arch/arm/domain_build.c
>> +++ b/xen/arch/arm/domain_build.c
>> @@ -104,11 +104,11 @@ static bool insert_11_bank(struct domain *d,
>>                              unsigned int order)
>>   {
>>       int res, i;
>> -    paddr_t spfn;
>> +    mfn_t smfn;
>>       paddr_t start, size;
>>   
>> -    spfn = page_to_mfn(pg);
>> -    start = pfn_to_paddr(spfn);
>> +    smfn = page_to_mfn(pg);
>> +    start = mfn_to_maddr(smfn);
>>       size = pfn_to_paddr(1UL << order);
> 
> Wouldn't it be cleaner to move this renaming into patch 1, along with an
> extra set of undef/override, to be taken out here?  (perhaps not given
> the rework effort?)

I moved the clean-up to patch #1 and add a temporary override that will 
be dropped in this patch.

> 
>>   
>>       D11PRINT("Allocated %#"PRIpaddr"-%#"PRIpaddr" (%ldMB/%ldMB, order %d)\n",
>> @@ -678,8 +678,8 @@ static void pvpmu_finish(struct domain *d, xen_pmu_params_t *params)
>>   
>>       if ( xenpmu_data )
>>       {
>> -        mfn = domain_page_map_to_mfn(xenpmu_data);
>> -        ASSERT(mfn_valid(_mfn(mfn)));
>> +        mfn = _mfn(domain_page_map_to_mfn(xenpmu_data));
> 
> Seeing as you convert every(?) call to domain_page_map_to_mfn(), it
> would be cleaner to change the return type while making the change. >
> I'd be happy for such a change being folded into this patch, because
> doing so would be by far the least disruptive way of making the change.

All of them but one are turned to _mfn(domain_page_map_to_mfn()). I have 
folded the conversion in this patch.

> 
>> +        ASSERT(mfn_valid(mfn));
>>           unmap_domain_page_global(xenpmu_data);
>>           put_page_and_type(mfn_to_page(mfn));
>>       }
>> @@ -1185,8 +1180,8 @@ int __mem_sharing_unshare_page(struct domain *d,
>>           return -ENOMEM;
>>       }
>>   
>> -    s = map_domain_page(_mfn(__page_to_mfn(old_page)));
>> -    t = map_domain_page(_mfn(__page_to_mfn(page)));
>> +    s = map_domain_page(page_to_mfn(old_page));
>> +    t = map_domain_page(page_to_mfn(page));
>>       memcpy(t, s, PAGE_SIZE);
>>       unmap_domain_page(s);
>>       unmap_domain_page(t);
> 
> This whole lot could turn into copy_domain_page()

I have added a patch at the beginning of the series to use copy_domain_page.

> 
>> diff --git a/xen/arch/x86/pv/descriptor-tables.c b/xen/arch/x86/pv/descriptor-tables.c
>> index 81973af124..371221a302 100644
>> --- a/xen/arch/x86/pv/descriptor-tables.c
>> +++ b/xen/arch/x86/pv/descriptor-tables.c
>> @@ -25,12 +25,6 @@
>>   #include <asm/p2m.h>
>>   #include <asm/pv/mm.h>
>>   
>> -/* Override macros from asm/page.h to make them work with mfn_t */
>> -#undef mfn_to_page
>> -#define mfn_to_page(mfn) __mfn_to_page(mfn_x(mfn))
>> -#undef page_to_mfn
>> -#define page_to_mfn(pg) _mfn(__page_to_mfn(pg))
>> -
>>   /*******************
>>    * Descriptor Tables
>>    */
> 
> If you're making this change, please take out the Descriptor Tables
> comment like you do with I/O below, because the entire file is dedicated
> to descriptor table support and it will save me one item on a cleanup
> patch :).

It is dropped now.

> 
>> diff --git a/xen/arch/x86/pv/emul-priv-op.c b/xen/arch/x86/pv/emul-priv-op.c
>> index dd90713acf..9ccbd021ef 100644
>> --- a/xen/arch/x86/pv/emul-priv-op.c
>> +++ b/xen/arch/x86/pv/emul-priv-op.c
>> @@ -43,16 +43,6 @@
>>   #include "emulate.h"
>>   #include "mm.h"
>>   
>> -/* Override macros from asm/page.h to make them work with mfn_t */
>> -#undef mfn_to_page
>> -#define mfn_to_page(mfn) __mfn_to_page(mfn_x(mfn))
>> -#undef page_to_mfn
>> -#define page_to_mfn(pg) _mfn(__page_to_mfn(pg))
>> -
>> -/***********************
>> - * I/O emulation support
>> - */
>> -
>>   struct priv_op_ctxt {
>>       struct x86_emulate_ctxt ctxt;
>>       struct {
>> @@ -873,22 +873,22 @@ int kimage_build_ind(struct kexec_image *image, unsigned long ind_mfn,
>>       for ( entry = page; ;  )
>>       {
>>           unsigned long ind;
>> -        unsigned long mfn;
>> +        mfn_t mfn;
>>   
>>           ind = kimage_entry_ind(entry, compat);
>> -        mfn = kimage_entry_mfn(entry, compat);
>> +        mfn = _mfn(kimage_entry_mfn(entry, compat));
> 
> Again, modify the return type of kimage_entry_mfn() ?

I have added a patch at the beginning of the series to switch 
kimage/kexec to mfn_t.

> 
>> diff --git a/xen/include/asm-x86/page.h b/xen/include/asm-x86/page.h
>> index 45ca742678..8737ef16ff 100644
>> --- a/xen/include/asm-x86/page.h
>> +++ b/xen/include/asm-x86/page.h
>> @@ -273,8 +273,8 @@ void copy_page_sse2(void *, const void *);
>>   #define pfn_to_paddr(pfn)   __pfn_to_paddr(pfn)
>>   #define paddr_to_pfn(pa)    __paddr_to_pfn(pa)
>>   #define paddr_to_pdx(pa)    pfn_to_pdx(paddr_to_pfn(pa))
>> -#define vmap_to_mfn(va)     l1e_get_pfn(*virt_to_xen_l1e((unsigned long)(va)))
>> -#define vmap_to_page(va)    mfn_to_page(vmap_to_mfn(va))
>> +#define vmap_to_mfn(va)     _mfn(l1e_get_pfn(*virt_to_xen_l1e((unsigned long)(va))))
> 
> l1e_get_mfn(*virt_to_xen_l1e((unsigned long)(va)))
> 
>> +#define vmap_to_page(va)    __mfn_to_page(vmap_to_mfn(va))
>>   
>>   #endif /* !defined(__ASSEMBLY__) */
>>   
>> diff --git a/xen/include/xen/tmem_xen.h b/xen/include/xen/tmem_xen.h
>> index 542c0b3f20..8516a0b131 100644
>> --- a/xen/include/xen/tmem_xen.h
>> +++ b/xen/include/xen/tmem_xen.h
>> @@ -25,7 +25,7 @@
>>   typedef uint32_t pagesize_t;  /* like size_t, must handle largest PAGE_SIZE */
>>   
>>   #define IS_PAGE_ALIGNED(addr) IS_ALIGNED((unsigned long)(addr), PAGE_SIZE)
>> -#define IS_VALID_PAGE(_pi)    mfn_valid(_mfn(page_to_mfn(_pi)))
>> +#define IS_VALID_PAGE(_pi)    mfn_valid(page_to_mfn(_pi))
> 
> /sigh  This is tautological.  The definition of a "valid mfn" in this
> case is one for which we have frametable entry, and by having a struct
> page_info in our hands, this is by definition true (unless you have a
> wild pointer, at which point your bug is elsewhere).
> 
> IS_VALID_PAGE() is only ever used in assertions and never usefully, so
> instead I would remove it entirely rather than trying to fix it up.

I would be happy to remove IS_VALID_PAGE is a patch at the beginning of 
the series if Konrad is happy with it.

I will probably send a new version today without IS_VALID_PAGE dropped. 
Though I will mention it in the patch to get feedback.

> 
> As for TMEM itself (Julien: This my no means blocks the patch.  It is
> more an observation for Konrad to see about fixing), I see that TMEM is
> broken on x86 machines with more than 5TB of RAM, because it is not
> legal to call page_to_virt() on a struct page_info allocated from the
> domheap (which is why alloc_xenheap_page() returns a void *, and
> alloc_domheap_page() specifically doesn't).  The easy fix for this is to
> swap the allocation primitives over to using xenheap allocations, which
> would remove the need for page_to_virt() and back, or a better fix would
> be to not pass everything thing by virtual address (at which point
> retaining use of the domheap is fine).

Cheers,

-- 
Julien Grall

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

Re: [PATCH 0/7] xen: Convert __page_to_mfn and __mfn_to_page to use typesafe MFN

[Tim Deegan]

At 19:15 +0100 on 04 Oct (1507144519), Julien Grall wrote:
> Hi all,
> 
> Most of the users of page_to_mfn and mfn_to_page are either overriding
> the macros to make them work with mfn_t or use mfn_x/_mfn becaue the rest
> of the function use mfn_t.
> 
> So I think it is time to make __page_to_mfn and __mfn_to_page using typesafe
> MFN.
> 
> The first 6 patches will convert of the code to use typesafe MFN, easing
> the tree-wide conversion in patch 7.

x86 shadow code changes Acked-by: Tim Deegan <tim@xen.org>

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