Linux-mm Archive on lore.kernel.org
 help / color / Atom feed
* [PATCH 0/2 v2] mm, slab: consolidate KMALLOC_MAX_SIZE
@ 2016-12-20 13:06 Michal Hocko
  2016-12-20 13:06 ` [PATCH 1/2] mm, slab: make sure that KMALLOC_MAX_SIZE will fit into MAX_ORDER Michal Hocko
  2016-12-20 13:06 ` [PATCH 2/2] bpf: do not use KMALLOC_SHIFT_MAX Michal Hocko
  0 siblings, 2 replies; 3+ messages in thread
From: Michal Hocko @ 2016-12-20 13:06 UTC (permalink / raw)
  To: Andrew Morton
  Cc: Cristopher Lameter, Alexei Starovoitov, Andrey Konovalov, netdev,
	linux-mm, LKML

Hi,
this is the second version of the patchset previously posted here [1].
Alexei has insisted on the patches reordering which I've done in this
series. I've also updated the changelog of the second patch to mention
why KMALLOC_SHIFT_MAX has been used.

Andrey has revealed a discrepancy between KMALLOC_MAX_SIZE and the
maximum supported page allocator size [2]. The underlying problem
should be fixed in the ep_write_iter code of course, but I do not feel
qualified to do that. The discrepancy which it reveals (see patch 2)
is worth fixing anyway, though.

While I was looking into the code, I've noticed that the only code which
uses KMALLOC_SHIFT_MAX outside of the slab code is bpf so I've updated
it to use KMALLOC_MAX_SIZE instead. There shouldn't be any real reason
to use KMALLOC_SHIFT_MAX which is a slab internal constant same as
KMALLOC_SHIFT_{LOW,HIGH}

[1] http://lkml.kernel.org/r/20161215164722.21586-1-mhocko@kernel.org
[2] http://lkml.kernel.org/r/CAAeHK+ztusS68DejO8AH3nn-EfiYQpD5FmBwmqKG8BWvoqPNqQ@mail.gm

--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@kvack.org.  For more info on Linux MM,
see: http://www.linux-mm.org/ .
Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>

^ permalink raw reply	[flat|nested] 3+ messages in thread

* [PATCH 1/2] mm, slab: make sure that KMALLOC_MAX_SIZE will fit into MAX_ORDER
  2016-12-20 13:06 [PATCH 0/2 v2] mm, slab: consolidate KMALLOC_MAX_SIZE Michal Hocko
@ 2016-12-20 13:06 ` Michal Hocko
  2016-12-20 13:06 ` [PATCH 2/2] bpf: do not use KMALLOC_SHIFT_MAX Michal Hocko
  1 sibling, 0 replies; 3+ messages in thread
From: Michal Hocko @ 2016-12-20 13:06 UTC (permalink / raw)
  To: Andrew Morton
  Cc: Cristopher Lameter, Alexei Starovoitov, Andrey Konovalov, netdev,
	linux-mm, LKML, Michal Hocko

From: Michal Hocko <mhocko@suse.com>

Andrey Konovalov has reported the following warning triggered by
the syzkaller fuzzer.

WARNING: CPU: 1 PID: 9935 at mm/page_alloc.c:3511
__alloc_pages_nodemask+0x159c/0x1e20
Kernel panic - not syncing: panic_on_warn set ...

CPU: 1 PID: 9935 Comm: syz-executor0 Not tainted 4.9.0-rc7+ #34
Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS Bochs 01/01/2011
 ffff88006949f2c8 ffffffff81f96b8a ffffffff00000200 1ffff1000d293dec
 ffffed000d293de4 0000000000000a06 0000000041b58ab3 ffffffff8598b510
 ffffffff81f968f8 0000000041b58ab3 ffffffff85942a58 ffffffff81432860
Call Trace:
 [<     inline     >] __dump_stack lib/dump_stack.c:15
 [<ffffffff81f96b8a>] dump_stack+0x292/0x398 lib/dump_stack.c:51
 [<ffffffff8168c88e>] panic+0x1cb/0x3a9 kernel/panic.c:179
 [<ffffffff812b80b4>] __warn+0x1c4/0x1e0 kernel/panic.c:542
 [<ffffffff812b831c>] warn_slowpath_null+0x2c/0x40 kernel/panic.c:585
 [<     inline     >] __alloc_pages_slowpath mm/page_alloc.c:3511
 [<ffffffff816c08ac>] __alloc_pages_nodemask+0x159c/0x1e20 mm/page_alloc.c:3781
 [<ffffffff817cde17>] alloc_pages_current+0x1c7/0x6b0 mm/mempolicy.c:2072
 [<     inline     >] alloc_pages include/linux/gfp.h:469
 [<ffffffff8172fd8f>] kmalloc_order+0x1f/0x70 mm/slab_common.c:1015
 [<ffffffff8172fdff>] kmalloc_order_trace+0x1f/0x160 mm/slab_common.c:1026
 [<     inline     >] kmalloc_large include/linux/slab.h:422
 [<ffffffff817e01f0>] __kmalloc+0x210/0x2d0 mm/slub.c:3723
 [<     inline     >] kmalloc include/linux/slab.h:495
 [<ffffffff832262a7>] ep_write_iter+0x167/0xb50 drivers/usb/gadget/legacy/inode.c:664
 [<     inline     >] new_sync_write fs/read_write.c:499
 [<ffffffff817fdcd3>] __vfs_write+0x483/0x760 fs/read_write.c:512
 [<ffffffff817ff720>] vfs_write+0x170/0x4e0 fs/read_write.c:560
 [<     inline     >] SYSC_write fs/read_write.c:607
 [<ffffffff81803b2b>] SyS_write+0xfb/0x230 fs/read_write.c:599
 [<ffffffff84f47ec1>] entry_SYSCALL_64_fastpath+0x1f/0xc2

The issue is caused by a lack of size check for the request size in
ep_write_iter which should be fixed. It, however, points to another
problem, that SLUB defines KMALLOC_MAX_SIZE too large because the its
KMALLOC_SHIFT_MAX is (MAX_ORDER + PAGE_SHIFT) which means that the
resulting page allocator request might be MAX_ORDER which is too large
(see __alloc_pages_slowpath). The same applies to the SLOB allocator
which allows even larger sizes. Make sure that they are capped properly
and never request more than MAX_ORDER order.

Reported-by: Andrey Konovalov <andreyknvl@google.com>
Signed-off-by: Michal Hocko <mhocko@suse.com>
Acked-by: Christoph Lameter <cl@linux.com>
---
 include/linux/slab.h | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/include/linux/slab.h b/include/linux/slab.h
index 084b12bad198..4c5363566815 100644
--- a/include/linux/slab.h
+++ b/include/linux/slab.h
@@ -226,7 +226,7 @@ static inline const char *__check_heap_object(const void *ptr,
  * (PAGE_SIZE*2).  Larger requests are passed to the page allocator.
  */
 #define KMALLOC_SHIFT_HIGH	(PAGE_SHIFT + 1)
-#define KMALLOC_SHIFT_MAX	(MAX_ORDER + PAGE_SHIFT)
+#define KMALLOC_SHIFT_MAX	(MAX_ORDER + PAGE_SHIFT - 1)
 #ifndef KMALLOC_SHIFT_LOW
 #define KMALLOC_SHIFT_LOW	3
 #endif
@@ -239,7 +239,7 @@ static inline const char *__check_heap_object(const void *ptr,
  * be allocated from the same page.
  */
 #define KMALLOC_SHIFT_HIGH	PAGE_SHIFT
-#define KMALLOC_SHIFT_MAX	30
+#define KMALLOC_SHIFT_MAX	(MAX_ORDER + PAGE_SHIFT - 1)
 #ifndef KMALLOC_SHIFT_LOW
 #define KMALLOC_SHIFT_LOW	3
 #endif
-- 
2.10.2

--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@kvack.org.  For more info on Linux MM,
see: http://www.linux-mm.org/ .
Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>

^ permalink raw reply	[flat|nested] 3+ messages in thread

* [PATCH 2/2] bpf: do not use KMALLOC_SHIFT_MAX
  2016-12-20 13:06 [PATCH 0/2 v2] mm, slab: consolidate KMALLOC_MAX_SIZE Michal Hocko
  2016-12-20 13:06 ` [PATCH 1/2] mm, slab: make sure that KMALLOC_MAX_SIZE will fit into MAX_ORDER Michal Hocko
@ 2016-12-20 13:06 ` Michal Hocko
  1 sibling, 0 replies; 3+ messages in thread
From: Michal Hocko @ 2016-12-20 13:06 UTC (permalink / raw)
  To: Andrew Morton
  Cc: Cristopher Lameter, Alexei Starovoitov, Andrey Konovalov, netdev,
	linux-mm, LKML, Michal Hocko

From: Michal Hocko <mhocko@suse.com>

01b3f52157ff ("bpf: fix allocation warnings in bpf maps and integer
overflow") has added checks for the maximum allocateable size. It
(ab)used KMALLOC_SHIFT_MAX for that purpose. While this is not incorrect
it is not very clean because we already have KMALLOC_MAX_SIZE for this
very reason so let's change both checks to use KMALLOC_MAX_SIZE instead.

The original motivation for using KMALLOC_SHIFT_MAX was to work around
an incorrect KMALLOC_MAX_SIZE which could lead to allocation warnings
but it is no longer needed since "slab: make sure that KMALLOC_MAX_SIZE
will fit into MAX_ORDER".

Signed-off-by: Michal Hocko <mhocko@suse.com>
Acked-by: Christoph Lameter <cl@linux.com>
---
 kernel/bpf/arraymap.c | 2 +-
 kernel/bpf/hashtab.c  | 2 +-
 2 files changed, 2 insertions(+), 2 deletions(-)

diff --git a/kernel/bpf/arraymap.c b/kernel/bpf/arraymap.c
index a2ac051c342f..229a5d5df977 100644
--- a/kernel/bpf/arraymap.c
+++ b/kernel/bpf/arraymap.c
@@ -56,7 +56,7 @@ static struct bpf_map *array_map_alloc(union bpf_attr *attr)
 	    attr->value_size == 0 || attr->map_flags)
 		return ERR_PTR(-EINVAL);
 
-	if (attr->value_size >= 1 << (KMALLOC_SHIFT_MAX - 1))
+	if (attr->value_size > KMALLOC_MAX_SIZE)
 		/* if value_size is bigger, the user space won't be able to
 		 * access the elements.
 		 */
diff --git a/kernel/bpf/hashtab.c b/kernel/bpf/hashtab.c
index ad1bc67aff1b..c5ec7dc71c84 100644
--- a/kernel/bpf/hashtab.c
+++ b/kernel/bpf/hashtab.c
@@ -181,7 +181,7 @@ static struct bpf_map *htab_map_alloc(union bpf_attr *attr)
 		 */
 		goto free_htab;
 
-	if (htab->map.value_size >= (1 << (KMALLOC_SHIFT_MAX - 1)) -
+	if (htab->map.value_size >= KMALLOC_MAX_SIZE -
 	    MAX_BPF_STACK - sizeof(struct htab_elem))
 		/* if value_size is bigger, the user space won't be able to
 		 * access the elements via bpf syscall. This check also makes
-- 
2.10.2

--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@kvack.org.  For more info on Linux MM,
see: http://www.linux-mm.org/ .
Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>

^ permalink raw reply	[flat|nested] 3+ messages in thread

end of thread, back to index

Thread overview: 3+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2016-12-20 13:06 [PATCH 0/2 v2] mm, slab: consolidate KMALLOC_MAX_SIZE Michal Hocko
2016-12-20 13:06 ` [PATCH 1/2] mm, slab: make sure that KMALLOC_MAX_SIZE will fit into MAX_ORDER Michal Hocko
2016-12-20 13:06 ` [PATCH 2/2] bpf: do not use KMALLOC_SHIFT_MAX Michal Hocko

Linux-mm Archive on lore.kernel.org

Archives are clonable:
	git clone --mirror https://lore.kernel.org/linux-mm/0 linux-mm/git/0.git

	# If you have public-inbox 1.1+ installed, you may
	# initialize and index your mirror using the following commands:
	public-inbox-init -V2 linux-mm linux-mm/ https://lore.kernel.org/linux-mm \
		linux-mm@kvack.org
	public-inbox-index linux-mm

Example config snippet for mirrors

Newsgroup available over NNTP:
	nntp://nntp.lore.kernel.org/org.kvack.linux-mm


AGPL code for this site: git clone https://public-inbox.org/public-inbox.git