Hi Wei,

On 03/09/2018 05:29 PM, Wei Liu wrote:
On Mon, Mar 05, 2018 at 07:38:36AM -0700, Jan Beulich wrote:
On 05.03.18 at 15:11, <julien.gr...@arm.com> wrote:
On 05/03/18 14:00, Jan Beulich wrote:
On 05.03.18 at 14:43, <julien.gr...@arm.com> wrote:
Anyway, I don't have much knowledge on the x86 to make the modification
that you suggested. So I am going to revert to _mfn(0) for x86.

I'd prefer if you didn't, but well, it'll be one of us to clean it up
then.
I can keep as INVALID_MFN. But then either you or Andrew (or anyone x86
folks) would have to provide the patch to skip incrementing invalid MFN
(if I understood correctly your request).

Sigh - this should go together imo. While wrongly incrementing from
zero was bad, wrongly wrapping from INVALID_MFN makes things
worse.


Try this patch?

I am happy to carry this patch at the beginning of my series if you want.


---8<---
 From 8f0024c690c736d17adde0fa765cbbf6fa2846dc Mon Sep 17 00:00:00 2001
From: Wei Liu <wei.l...@citrix.com>
Date: Fri, 9 Mar 2018 17:20:14 +0000
Subject: [PATCH] x86/mm: skip incrementing mfn if it is not a valid mfn

The function is called to fill in page table entries in
populate_pt_range. Skip incrementing mfn if it is invalid.

Signed-off-by: Wei Liu <wei.l...@citrix.com>
---
  xen/arch/x86/mm.c | 15 ++++++++++-----
  1 file changed, 10 insertions(+), 5 deletions(-)

diff --git a/xen/arch/x86/mm.c b/xen/arch/x86/mm.c
index 9b559448a7..5f5577c7c2 100644
--- a/xen/arch/x86/mm.c
+++ b/xen/arch/x86/mm.c
@@ -4731,7 +4731,8 @@ int map_pages_to_xen(
              }
virt += 1UL << L3_PAGETABLE_SHIFT;
-            mfn     += 1UL << (L3_PAGETABLE_SHIFT - PAGE_SHIFT);
+            if ( !mfn_eq(_mfn(mfn), INVALID_MFN) )
+                mfn += 1UL << (L3_PAGETABLE_SHIFT - PAGE_SHIFT);
              nr_mfns -= 1UL << (L3_PAGETABLE_SHIFT - PAGE_SHIFT);
              continue;
          }
@@ -4756,7 +4757,8 @@ int map_pages_to_xen(
                  if ( i > nr_mfns )
                      i = nr_mfns;
                  virt    += i << PAGE_SHIFT;
-                mfn     += i;
+                if ( !mfn_eq(_mfn(mfn), INVALID_MFN) )
+                    mfn += i;
                  nr_mfns -= i;
                  continue;
              }
@@ -4824,7 +4826,8 @@ int map_pages_to_xen(
              }
virt += 1UL << L2_PAGETABLE_SHIFT;
-            mfn     += 1UL << PAGETABLE_ORDER;
+            if ( !mfn_eq(_mfn(mfn), INVALID_MFN) )
+                mfn += 1UL << PAGETABLE_ORDER;
              nr_mfns -= 1UL << PAGETABLE_ORDER;
          }
          else
@@ -4853,7 +4856,8 @@ int map_pages_to_xen(
                      if ( i > nr_mfns )
                          i = nr_mfns;
                      virt    += i << L1_PAGETABLE_SHIFT;
-                    mfn     += i;
+                    if ( !mfn_eq(_mfn(mfn), INVALID_MFN) )
+                        mfn += i;
                      nr_mfns -= i;
                      goto check_l3;
                  }
@@ -4898,7 +4902,8 @@ int map_pages_to_xen(
              }
virt += 1UL << L1_PAGETABLE_SHIFT;
-            mfn     += 1UL;
+            if ( !mfn_eq(_mfn(mfn), INVALID_MFN) )
+                mfn += 1UL;
              nr_mfns -= 1UL;
if ( (flags == PAGE_HYPERVISOR) &&


Cheers,

--
Julien Grall

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

Reply via email to