[PATCH v2] x86/mm/vmfault: Make vmalloc_fault() handle large pages

[Toshi Kani]

The following oops was observed when a read syscall was made to
a pmem device after a huge amount (>512GB) of vmalloc ranges was
allocated by ioremap() on a x86_64 system.

 BUG: unable to handle kernel paging request at ffff880840000ff8
 IP: [<ffffffff810664ae>] vmalloc_fault+0x1be/0x300
 PGD c7f03a067 PUD 0
 Oops: 0000 [#1] SM
   :
 Call Trace:
 [<ffffffff81067335>] __do_page_fault+0x285/0x3e0
 [<ffffffff810674bf>] do_page_fault+0x2f/0x80
 [<ffffffff810d6d85>] ? put_prev_entity+0x35/0x7a0
 [<ffffffff817a6888>] page_fault+0x28/0x30
 [<ffffffff813bb976>] ? memcpy_erms+0x6/0x10
 [<ffffffff817a0845>] ? schedule+0x35/0x80
 [<ffffffffa006350a>] ? pmem_rw_bytes+0x6a/0x190 [nd_pmem]
 [<ffffffff817a3713>] ? schedule_timeout+0x183/0x240
 [<ffffffffa028d2b3>] btt_log_read+0x63/0x140 [nd_btt]
   :
 [<ffffffff811201d0>] ? __symbol_put+0x60/0x60
 [<ffffffff8122dc60>] ? kernel_read+0x50/0x80
 [<ffffffff81124489>] SyS_finit_module+0xb9/0xf0
 [<ffffffff817a4632>] entry_SYSCALL_64_fastpath+0x1a/0xa4

Since 4.1, ioremap() supports large page (pud/pmd) mappings in
x86_64 and PAE.  vmalloc_fault() however assumes that the vmalloc
range is limited to pte mappings.

vmalloc faults do not normally happen in ioremap'd ranges since
ioremap() sets up the kernel page tables, which are shared by
user processes.  pgd_ctor() sets the kernel's pgd entries to
user's during fork().  When allocation of the vmalloc ranges
crosses a 512GB boundary, ioremap() allocates a new pud table
and updates the kernel pgd entry to point it.  If user process's
pgd entry does not have this update yet, a read/write syscall
to the range will cause a vmalloc fault, which hits the Oops
above as it does not handle a large page properly.

Following changes are made to vmalloc_fault().

64-bit:
- No change for the pgd sync operation as it handles large
  pages already.
- Add pud_huge() and pmd_huge() to the validation code to
  handle large pages.
- Change pud_page_vaddr() to pud_pfn() since an ioremap range
  is not directly mapped (while the if-statement still works
  with a bogus addr).
- Change pmd_page() to pmd_pfn() since an ioremap range is not
  backed by struct page (while the if-statement still works
  with a bogus addr).

32-bit:
- No change for the sync operation since the index3 pgd entry
  covers the entire vmalloc range, which is always valid.
  (A separate change to sync pgd entry is necessary if this
   memory layout is changed regardless of the page size.)
- Add pmd_huge() to the validation code to handle large pages.
  This is for completeness since vmalloc_fault() won't happen
  in ioremap'd ranges as its pgd entry is always valid.

Reported-by: Henning Schild <henning.schild@xxxxxxxxxxx>
Signed-off-by: Toshi Kani <toshi.kani@xxxxxxx>
Cc: Thomas Gleixner <tglx@xxxxxxxxxxxxx>
Cc: Ingo Molnar <mingo@xxxxxxxxxx>
Cc: "H. Peter Anvin" <hpa@xxxxxxxxx>
Cc: Borislav Petkov <bp@xxxxxxxxx>
---
When this patch is accepted, please copy to stable up to 4.1.
v2: Add more descriptions about the issue in the change log.
---
 arch/x86/mm/fault.c |   15 +++++++++++----
 1 file changed, 11 insertions(+), 4 deletions(-)

diff --git a/arch/x86/mm/fault.c b/arch/x86/mm/fault.c
index eef44d9..e830c71 100644
--- a/arch/x86/mm/fault.c
+++ b/arch/x86/mm/fault.c
@@ -287,6 +287,9 @@ static noinline int vmalloc_fault(unsigned long address)
 	if (!pmd_k)
 		return -1;
 
+	if (pmd_huge(*pmd_k))
+		return 0;
+
 	pte_k = pte_offset_kernel(pmd_k, address);
 	if (!pte_present(*pte_k))
 		return -1;
@@ -360,8 +363,6 @@ void vmalloc_sync_all(void)
  * 64-bit:
  *
  *   Handle a fault on the vmalloc area
- *
- * This assumes no large pages in there.
  */
 static noinline int vmalloc_fault(unsigned long address)
 {
@@ -403,17 +404,23 @@ static noinline int vmalloc_fault(unsigned long address)
 	if (pud_none(*pud_ref))
 		return -1;
 
-	if (pud_none(*pud) || pud_page_vaddr(*pud) != pud_page_vaddr(*pud_ref))
+	if (pud_none(*pud) || pud_pfn(*pud) != pud_pfn(*pud_ref))
 		BUG();
 
+	if (pud_huge(*pud))
+		return 0;
+
 	pmd = pmd_offset(pud, address);
 	pmd_ref = pmd_offset(pud_ref, address);
 	if (pmd_none(*pmd_ref))
 		return -1;
 
-	if (pmd_none(*pmd) || pmd_page(*pmd) != pmd_page(*pmd_ref))
+	if (pmd_none(*pmd) || pmd_pfn(*pmd) != pmd_pfn(*pmd_ref))
 		BUG();
 
+	if (pmd_huge(*pmd))
+		return 0;
+
 	pte_ref = pte_offset_kernel(pmd_ref, address);
 	if (!pte_present(*pte_ref))
 		return -1;