Skip to contentSkip to conversation

mm/percpu.c: append alignment sanity checkup to avoid memory leakage

PATCH In review zijun_hu 2 participants 2 messages Oct 14, 2016 – Oct 19, 2016

Subsystems (2)
lore
Share & tools
lore

zijun_hu

author
#0 lore raw

PATCH 1/1 mm/percpu.c: append alignment sanity checkup to avoid memory leakage

From: zijun_hu <[email protected]>

the percpu allocator only works well currently when allocates a power of 2 aligned area, but there aren't any hints for alignment requirement, so memory leakage maybe be caused if allocate other alignment areas

the alignment must be a even at least since the LSB of a chunk->map element is used as free/in-use flag of a area; besides, the alignment must be a power of 2 too since ALIGN() doesn't work well for other alignment always but is adopted by pcpu_fit_in_area(). IOW, the current allocator only works well for a power of 2 aligned area allocation.

see below opposite example for why a odd alignment doesn't work lets assume area [16, 36) is free but its previous one is in-use, we want to allocate a @size == 8 and @align == 7 area. the larger area [16, 36) is split to three areas [16, 21), [21, 29), [29, 36) eventually. however, due to the usage for a chunk->map element, the actual offset of the aim area [21, 29) is 21 but is recorded in relevant element as 20; moreover the residual tail free area [29, 36) is mistook as in-use and is lost silently

unlike macro roundup(), ALIGN(x, a) doesn't work if @a isn't a power of 2 for example, roundup(10, 6) == 12 but ALIGN(10, 6) == 10, and the latter result isn't desired obviously.

fix it by appending sanity checkup for alignment requirement

---
 include/linux/kernel.h | 1 +
 mm/percpu.c            | 3 ++-
 2 files changed, 3 insertions(+), 1 deletion(-)
include/linux/kernel.h +1 -0
index bc6ed52a39b9..0dc0b21bd164 100644
@@ -45,6 +45,7 @@
 
 #define REPEAT_BYTE(x)	((~0ul / 0xff) * (x))
 
+/* @a is a power of 2 value */
 #define ALIGN(x, a)		__ALIGN_KERNEL((x), (a))
 #define __ALIGN_MASK(x, mask)	__ALIGN_KERNEL_MASK((x), (mask))
 #define PTR_ALIGN(p, a)		((typeof(p))ALIGN((unsigned long)(p), (a)))
mm/percpu.c +2 -1
index 255714302394..10ba3f9a3826 100644
@@ -886,7 +886,8 @@ static void __percpu *pcpu_alloc(size_t size, size_t align, bool reserved,
 
 	size = ALIGN(size, 2);
 
-	if (unlikely(!size || size > PCPU_MIN_UNIT_SIZE || align > PAGE_SIZE)) {
+	if (unlikely(!size || size > PCPU_MIN_UNIT_SIZE || align > PAGE_SIZE
+				|| !is_power_of_2(align))) {
 		WARN(true, "illegal size (%zu) or align (%zu) for percpu allocation\n",
 		     size, align);
 		return NULL;
Signature · 7 lines
1.9.1

--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to [email protected].  For more info on Linux MM,
see: http://www.linux-mm.org/ .
Don't email: <a href=mailto:"[email protected]"> [email protected] </a>
maintainer
#1 lore raw

Hello,

I updated the patch description and code style a bit and applied to percpu/for-4.10.

Thanks.

------ 8< ------