dax: re-enable dax pmd mappings

Now that the get_user_pages() path knows how to handle dax-pmd mappings,
remove the protections that disabled dax-pmd support.

Tests available from github.com/pmem/ndctl:

    make TESTS="lib/test-dax.sh lib/test-mmap.sh" check
Signed-off-by: Dan Williams <dan.j.williams@intel.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

fs/Kconfig

@@ -50,7 +50,8 @@ config FS_DAX_PMD
 	bool
 	default FS_DAX
 	depends on FS_DAX
-	depends on BROKEN
+	depends on ZONE_DEVICE
+	depends on TRANSPARENT_HUGEPAGE
 
 endif # BLOCK
 

fs/dax.c

@@ -592,7 +592,7 @@ int __dax_pmd_fault(struct vm_area_struct *vma, unsigned long address,
 	sector_t block;
 	int result = 0;
 
-	/* dax pmd mappings are broken wrt gup and fork */
+	/* dax pmd mappings require pfn_t_devmap() */
 	if (!IS_ENABLED(CONFIG_FS_DAX_PMD))
 		return VM_FAULT_FALLBACK;
 
@@ -717,11 +717,7 @@ int __dax_pmd_fault(struct vm_area_struct *vma, unsigned long address,
 			goto fallback;
 		}
 
-		/*
-		 * TODO: teach vmf_insert_pfn_pmd() to support
-		 * 'pte_special' for pmds
-		 */
-		if (pfn_t_has_page(dax.pfn)) {
+		if (!pfn_t_devmap(dax.pfn)) {
 			dax_unmap_atomic(bdev, &dax);
 			dax_pmd_dbg(&bh, address, "pfn not in memmap");
 			goto fallback;