[PATCH] Handle spurious page fault for hugetlb region

The hugetlb pages are currently pre-faulted.  At the time of mmap of
hugepages, we populate the new PTEs.  It is possible that HW has already
cached some of the unused PTEs internally.  These stale entries never
get a chance to be purged in existing control flow.

This patch extends the check in page fault code for hugepages.  Check if
a faulted address falls with in size for the hugetlb file backing it.
We return VM_FAULT_MINOR for these cases (assuming that the arch
specific page-faulting code purges the stale entry for the archs that
need it).

Signed-off-by: Rohit Seth <rohit.seth@intel.com>

[ This is apparently arguably an ia64 port bug. But the code won't
  hurt, and for now it fixes a real problem on some ia64 machines ]

Signed-off-by: Linus Torvalds <torvalds@osdl.org>

include/linux/hugetlb.h

@@ -155,11 +155,24 @@ static inline void set_file_hugepages(struct file *file)
 {
 	file->f_op = &hugetlbfs_file_operations;
 }
+
+static inline int valid_hugetlb_file_off(struct vm_area_struct *vma, 
+					  unsigned long address) 
+{
+	struct inode *inode = vma->vm_file->f_dentry->d_inode;
+	loff_t file_off = address - vma->vm_start;
+
+	file_off += (vma->vm_pgoff << PAGE_SHIFT);
+
+	return (file_off < inode->i_size);
+}
+
 #else /* !CONFIG_HUGETLBFS */
 
 #define is_file_hugepages(file)		0
 #define set_file_hugepages(file)	BUG()
 #define hugetlb_zero_setup(size)	ERR_PTR(-ENOSYS)
+#define valid_hugetlb_file_off(vma, address) 	0
 
 #endif /* !CONFIG_HUGETLBFS */
 

mm/memory.c

@@ -2045,8 +2045,18 @@ int __handle_mm_fault(struct mm_struct *mm, struct vm_area_struct * vma,
 
 	inc_page_state(pgfault);
 
-	if (is_vm_hugetlb_page(vma))
-		return VM_FAULT_SIGBUS;	/* mapping truncation does this. */
+	if (unlikely(is_vm_hugetlb_page(vma))) {
+		if (valid_hugetlb_file_off(vma, address))
+			/* We get here only if there was a stale(zero) TLB entry 
+			 * (because of  HW prefetching). 
+			 * Low-level arch code (if needed) should have already
+			 * purged the stale entry as part of this fault handling.  
+			 * Here we just return.
+			 */
+			return VM_FAULT_MINOR; 
+		else
+			return VM_FAULT_SIGBUS;	/* mapping truncation does this. */
+	}
 
 	/*
 	 * We need the page table lock to synchronize with kswapd