Re: [PATCH 2/6] mm/vmalloc: set area's page_order after allocation succeeds

__vmalloc_area_node() calls set_vm_area_page_order() to set area's
page_order before actually allocating pages to populate the area.

If allocation of large pages in HUGE_VMAP case fails midway, this leaves
the area with elevated page_order throughout the cleanup path.

There is no actual issue with this because the only place that currently
relies on area->page_order on the cleanup path is the loop calculating
the direct map alias range in vm_reset_perms() and it anyway skips
unpopulated pages.

But having set_vm_area_page_order() in the middle of __vmalloc_area_node()
makes things very obscure, hard to reason about and error prone against
future changes of the cleanup path.

Move the call to set_vm_area_page_order() after __vmalloc_area_node()
succeeded where page order is guaranteed.

Signed-off-by: Mike Rapoport (Microsoft) <rppt@kernel.org>
---
 mm/vmalloc.c | 11 +++++++++--
 1 file changed, 9 insertions(+), 2 deletions(-)

diff --git a/mm/vmalloc.c b/mm/vmalloc.c
index 22566e0b6e38..6822f0fe9583 100644
--- a/mm/vmalloc.c
+++ b/mm/vmalloc.c

@@ -3901,8 +3901,7 @@ static void *__vmalloc_area_node(struct vm_struct *area, gfp_t gfp_mask,
 		goto fail;
 	}
 
-	set_vm_area_page_order(area, page_shift - PAGE_SHIFT);
-	page_order = vm_area_page_order(area);
+	page_order = page_shift - PAGE_SHIFT;
 
 	/*
 	 * High-order nofail allocations are really expensive and

@@ -4106,6 +4105,14 @@ void *__vmalloc_node_range_noprof(unsigned long size, unsigned long align,
 	if (!ret)
 		goto fail;
 
+	/*
+	 * Set area->page_order once it's known exactly that the order of the
+	 * pages the area contains.
+	 * Even if we succeeded to partially populate the area with large pages,
+	 * still treat the area as populated with order-0 pages.
+	 */
+	set_vm_area_page_order(area, shift - PAGE_SHIFT);
+
 	/*
 	 * Mark the pages as accessible, now that they are mapped.
 	 * The condition for setting KASAN_VMALLOC_INIT should complement the

-- 
2.53.0