linux-kernel.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
* regression 4.8+ commit 8ea1d2a (mm, frontswap: convert frontswap_enabled to static key) cause memory leak on swapon
@ 2016-10-26 12:10 Christian Borntraeger
  2016-10-26 13:42 ` [PATCH] mm, frontswap: make sure allocated frontswap map is assigned Vlastimil Babka
  0 siblings, 1 reply; 2+ messages in thread
From: Christian Borntraeger @ 2016-10-26 12:10 UTC (permalink / raw)
  To: Vlastimil Babka; +Cc: Andrew Morton, LKML, Linux MM, Konrad Rzeszutek Wilk

Vlastimil,

with commit 8ea1d2a1985a7ae096e ("mm, frontswap: convert frontswap_enabled to static key")
kmemleak complains about a memory leak in swapon

unreferenced object 0x3e09ba56000 (size 32112640):
  comm "swapon", pid 7852, jiffies 4294968787 (age 1490.770s)
  hex dump (first 32 bytes):
    00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00  ................
    00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00  ................
  backtrace:
    [<00000000003a2504>] __vmalloc_node_range+0x194/0x2d8
    [<00000000003a2918>] vzalloc+0x58/0x68
    [<00000000003b0af0>] SyS_swapon+0xd60/0x12f8
    [<0000000000a3dc2e>] system_call+0xd6/0x270
    [<ffffffffffffffff>] 0xffffffffffffffff


Turns out kmemleak is right. We now allocate the frontswap map depending on the kernel config
(and no longer on the enablement)

swapfile.c:
[...]
      if (IS_ENABLED(CONFIG_FRONTSWAP))
                frontswap_map = vzalloc(BITS_TO_LONGS(maxpages) * sizeof(long));

but later on this is passed along
--> enable_swap_info(p, prio, swap_map, cluster_info, frontswap_map);

and ignored if frontswap is disabled
--> frontswap_init(p->type, frontswap_map);
static inline void frontswap_init(unsigned type, unsigned long *map)
{
        if (frontswap_enabled())
                __frontswap_init(type, map);
}

Thing is, that frontswap map is never freed.

Christian

^ permalink raw reply	[flat|nested] 2+ messages in thread

* [PATCH] mm, frontswap: make sure allocated frontswap map is assigned
  2016-10-26 12:10 regression 4.8+ commit 8ea1d2a (mm, frontswap: convert frontswap_enabled to static key) cause memory leak on swapon Christian Borntraeger
@ 2016-10-26 13:42 ` Vlastimil Babka
  0 siblings, 0 replies; 2+ messages in thread
From: Vlastimil Babka @ 2016-10-26 13:42 UTC (permalink / raw)
  To: Andrew Morton, Christian Borntraeger
  Cc: linux-kernel, linux-mm, Vlastimil Babka, stable,
	Konrad Rzeszutek Wilk, Boris Ostrovsky, David Vrabel,
	Juergen Gross, Kirill A. Shutemov

Christian Borntraeger reports:

with commit 8ea1d2a1985a7ae096e ("mm, frontswap: convert frontswap_enabled to
static key") kmemleak complains about a memory leak in swapon

unreferenced object 0x3e09ba56000 (size 32112640):
  comm "swapon", pid 7852, jiffies 4294968787 (age 1490.770s)
  hex dump (first 32 bytes):
    00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00  ................
    00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00  ................
  backtrace:
    [<00000000003a2504>] __vmalloc_node_range+0x194/0x2d8
    [<00000000003a2918>] vzalloc+0x58/0x68
    [<00000000003b0af0>] SyS_swapon+0xd60/0x12f8
    [<0000000000a3dc2e>] system_call+0xd6/0x270
    [<ffffffffffffffff>] 0xffffffffffffffff

Turns out kmemleak is right. We now allocate the frontswap map depending on the
kernel config (and no longer on the enablement)

swapfile.c:
[...]
      if (IS_ENABLED(CONFIG_FRONTSWAP))
                frontswap_map = vzalloc(BITS_TO_LONGS(maxpages) * sizeof(long));

but later on this is passed along
--> enable_swap_info(p, prio, swap_map, cluster_info, frontswap_map);

and ignored if frontswap is disabled
--> frontswap_init(p->type, frontswap_map);
static inline void frontswap_init(unsigned type, unsigned long *map)
{
        if (frontswap_enabled())
                __frontswap_init(type, map);
}

Thing is, that frontswap map is never freed.

===

The leakage is relatively not that bad, because swapon is an infrequent and
privileged operation. However, if the first frontswap backend is registered
after a swap type has been already enabled, it will WARN_ON in
frontswap_register_ops() and frontswap will not be available for the swap type.

Fix this by making sure the map is assigned by frontswap_init() as long as
CONFIG_FRONTSWAP is enabled.

Reported-by: Christian Borntraeger <borntraeger@de.ibm.com>
Fixes: 8ea1d2a1985a ("mm, frontswap: convert frontswap_enabled to static key")
Cc: stable@vger.kernel.org
Cc: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>
Cc: Boris Ostrovsky <boris.ostrovsky@oracle.com>
Cc: David Vrabel <david.vrabel@citrix.com>
Cc: Juergen Gross <jgross@suse.com>
Cc: "Kirill A. Shutemov" <kirill.shutemov@linux.intel.com>
Signed-off-by: Vlastimil Babka <vbabka@suse.cz>
---
 include/linux/frontswap.h | 5 +++--
 1 file changed, 3 insertions(+), 2 deletions(-)

diff --git a/include/linux/frontswap.h b/include/linux/frontswap.h
index c46d2aa16d81..1d18af034554 100644
--- a/include/linux/frontswap.h
+++ b/include/linux/frontswap.h
@@ -106,8 +106,9 @@ static inline void frontswap_invalidate_area(unsigned type)
 
 static inline void frontswap_init(unsigned type, unsigned long *map)
 {
-	if (frontswap_enabled())
-		__frontswap_init(type, map);
+#ifdef CONFIG_FRONTSWAP
+	__frontswap_init(type, map);
+#endif
 }
 
 #endif /* _LINUX_FRONTSWAP_H */
-- 
2.10.1

^ permalink raw reply related	[flat|nested] 2+ messages in thread

end of thread, other threads:[~2016-10-26 13:42 UTC | newest]

Thread overview: 2+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2016-10-26 12:10 regression 4.8+ commit 8ea1d2a (mm, frontswap: convert frontswap_enabled to static key) cause memory leak on swapon Christian Borntraeger
2016-10-26 13:42 ` [PATCH] mm, frontswap: make sure allocated frontswap map is assigned Vlastimil Babka

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).