[PATCH net-next v5 0/7] net: skb: isolate skb data area allocations into a separate bucket

[PATCH net-next v5 0/7] net: skb: isolate skb data area allocations into a separate bucket

[Kees Cook]

Hi!

This gets the buckets able to handle memcg (GFP_KERNEL_ACCOUNT) with
isolation (since it's common due to AF_UNIX), and GFP_DMA with fall back
(since it's rare). It gave me an excuse to build out bucket kunit tests
too, and that (and LLM review) found a couple other issues that needed
fixing too.

The bulk of this is mm/slab, but the final patch is netdev, which Paolo
acked in v4, so I'm hoping this whole series can go via slab?

Thanks!

-Kees

 v5:
 - 2/7: kmem_buckets_create() and kmem_buckets_create_types() take an
   alignment, where kmem_cache_create() has it; 0 keeps v4's behaviour,
   the alignment of the kmalloc cache being mirrored. All callers pass 0.
   New subject: "mm/slab: Let kmem_buckets_create() take an alignment"
   (Harry)
 - 4/7: test that an explicit alignment is honored.
 - 5/7: export mem_cgroup_kmem_disabled() for KUnit only, fixing a
   modular build of slub_kunit (Harry).
 - 6/7: commit log shows the per-set cost of indexing rows by
   kmalloc_cache_type; test that an explicit alignment reaches the
   accounted row too.
 - 7/7: pass the new alignment argument.
 - Collected tags on 1/7 (Harry, Pedro, Hao) and 7/7 (Paolo).
 - v4..v5 diff: https://git.kernel.org/pub/scm/linux/kernel/git/kees/linux.git/diff/?id=dev/v7.3-rc2/skb-buckets/v5&id2=dev/v7.3-rc2/skb-buckets/v4
 v4: https://lore.kernel.org/all/20260921075811.too.775-kees@kernel.org/
 v3: https://lore.kernel.org/all/20260702170728.168755-1-pfalcato@suse.de/


Kees Cook (6):
  mm/slab: Mark the kmem_buckets_create() context as a Context: section
  mm/slab: Let kmem_buckets_create() take an alignment
  mm/slab: Add kmem_buckets_destroy()
  mm/slab: Add tests for the existing kmem_buckets behaviour
  mm/slab: Provide kmalloc type fallback for bucket allocations
  mm/slab: Let a bucket set handle __GFP_ACCOUNT

Pedro Falcato (1):
  net: skb: isolate skb data area allocations into a separate bucket

 include/linux/slab.h   |  66 ++++++-
 mm/slab.h              |  44 ++++-
 ipc/msgutil.c          |   2 +-
 lib/tests/slub_kunit.c | 397 +++++++++++++++++++++++++++++++++++++++++
 mm/memcontrol.c        |   2 +
 mm/slab_common.c       | 173 +++++++++++++++---
 mm/util.c              |   2 +-
 net/core/skbuff.c      |  11 +-
 8 files changed, 663 insertions(+), 34 deletions(-)

-- 
2.34.1

[PATCH net-next v5 1/7] mm/slab: Mark the kmem_buckets_create() context as a Context: section

[Kees Cook]

kmem_buckets_create() had the same sentence about calling context as
kmem_cache_create(), but lacked the "Context:" prefix, so kernel-doc
rendered in the body instead of as a "context" section.

Give it the missing prefix. Additionally fix the "a interrupt" typo
__kmem_cache_create_args() had.

Fixes: b32801d1255be ("mm/slab: Introduce kmem_buckets_create() and family")
Assisted-by: LLM
Reviewed-by: Harry Yoo (Meta) <harry@kernel.org>
Acked-by: Pedro Falcato <pfalcato@suse.de>
Reviewed-by: Hao Li <hao.li@linux.dev>
Signed-off-by: Kees Cook <kees@kernel.org>
---
 mm/slab_common.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/mm/slab_common.c b/mm/slab_common.c
index b19ba1b31484..270408ce5a9d 100644
--- a/mm/slab_common.c
+++ b/mm/slab_common.c
@@ -311,7 +311,7 @@ __kmem_cache_alias(const char *name, unsigned int size, slab_flags_t flags,
  * &SLAB_TYPESAFE_BY_RCU - Slab page (not individual objects) freeing delayed
  * by a grace period - see the full description before using.
  *
- * Context: Cannot be called within a interrupt, but can be interrupted.
+ * Context: Cannot be called within an interrupt, but can be interrupted.
  *
  * Return: a pointer to the cache on success, NULL on failure.
  */
@@ -422,7 +422,7 @@ static struct kmem_cache *kmem_buckets_cache __ro_after_init;
  *		to/from userspace.
  * @ctor: A constructor for the objects, run when new allocations are made.
  *
- * Cannot be called within an interrupt, but can be interrupted.
+ * Context: Cannot be called within an interrupt, but can be interrupted.
  *
  * Return: a pointer to the cache on success, NULL on failure. When
  * CONFIG_SLAB_BUCKETS is not enabled, ZERO_SIZE_PTR is returned, and
-- 
2.34.1

[PATCH net-next v5 2/7] mm/slab: Let kmem_buckets_create() take an alignment

[Kees Cook]

A bucket set is created with kmem_cache_create_usercopy(..., align = 0),
so calculate_alignment() falls back to arch_slab_minalign(), typically 8
bytes. The general kmalloc caches it stands in for are created through
create_boot_cache(), which starts from ARCH_KMALLOC_MINALIGN and raises
it to the largest power-of-two divisor of the size:

	if (flags & SLAB_KMALLOC)
		align = max(align, 1U << (ffs(size) - 1));

This is only a problem when slab metadata is enabled with
CONFIG_KASAN=y, CONFIG_SLUB_DEBUG_ON=y, or "slab_debug=...", because
metadata changes the stride size off a power of two, for example:

	size 128:  bucket align=8 size=224  | kmalloc align=128  size=384
	size 512:  bucket align=8 size=608  | kmalloc align=512  size=1536
	size 2048: bucket align=8 size=2144 | kmalloc align=2048 size=6144

So bucket allocations will fail the IS_ALIGNED(p, ARCH_DMA_MINALIGN)
check, potentially creating problems for non-coherent DMA situation.

As discussed in review, add an alignment argument to
kmem_buckets_create(), where kmem_cache_create() has it. A caller that
needs a particular alignment passes it. With 0, each cache takes the
alignment of the kmalloc cache being mirrored, which is where the size
and the name suffix already come from, so a caller moving from kmalloc()
keeps the alignment it had. Both existing callers pass 0.

Fixes: b32801d1255be ("mm/slab: Introduce kmem_buckets_create() and family")
Assisted-by: LLM
Signed-off-by: Kees Cook <kees@kernel.org>
---
 include/linux/slab.h | 3 ++-
 ipc/msgutil.c        | 2 +-
 mm/slab_common.c     | 9 +++++++--
 mm/util.c            | 2 +-
 4 files changed, 11 insertions(+), 5 deletions(-)

diff --git a/include/linux/slab.h b/include/linux/slab.h
index cda126def67a..94709e4e14f3 100644
--- a/include/linux/slab.h
+++ b/include/linux/slab.h
@@ -890,7 +890,8 @@ void *kmem_cache_alloc_lru_noprof(struct kmem_cache *s, struct list_lru *lru,
 bool kmem_cache_charge(void *objp, gfp_t gfpflags);
 void kmem_cache_free(struct kmem_cache *s, void *objp);
 
-kmem_buckets *kmem_buckets_create(const char *name, slab_flags_t flags,
+kmem_buckets *kmem_buckets_create(const char *name, unsigned int align,
+				  slab_flags_t flags,
 				  unsigned int useroffset, unsigned int usersize,
 				  void (*ctor)(void *));
 
diff --git a/ipc/msgutil.c b/ipc/msgutil.c
index e28f0cecb2ec..b49b8f65a582 100644
--- a/ipc/msgutil.c
+++ b/ipc/msgutil.c
@@ -43,7 +43,7 @@ static kmem_buckets *msg_buckets __ro_after_init;
 
 static int __init init_msg_buckets(void)
 {
-	msg_buckets = kmem_buckets_create("msg_msg", SLAB_ACCOUNT,
+	msg_buckets = kmem_buckets_create("msg_msg", 0, SLAB_ACCOUNT,
 					  sizeof(struct msg_msg),
 					  DATALEN_MSG, NULL);
 
diff --git a/mm/slab_common.c b/mm/slab_common.c
index 270408ce5a9d..301f3d4f4ca2 100644
--- a/mm/slab_common.c
+++ b/mm/slab_common.c
@@ -415,6 +415,9 @@ static struct kmem_cache *kmem_buckets_cache __ro_after_init;
  *			 allocations via kmem_buckets_alloc()
  * @name: A prefix string which is used in /proc/slabinfo to identify this
  *	  cache. The individual caches with have their sizes as the suffix.
+ * @align: The required alignment for the objects, or 0 to give each cache
+ *	   the alignment of the kmalloc cache of the same size, as a caller
+ *	   moving from kmalloc() may depend on.
  * @flags: SLAB flags (see kmem_cache_create() for details).
  * @useroffset: Starting offset within an allocation that may be copied
  *		to/from userspace.
@@ -429,7 +432,8 @@ static struct kmem_cache *kmem_buckets_cache __ro_after_init;
  * subsequent calls to kmem_buckets_alloc() will fall back to kmalloc().
  * (i.e. callers only need to check for NULL on failure.)
  */
-kmem_buckets *kmem_buckets_create(const char *name, slab_flags_t flags,
+kmem_buckets *kmem_buckets_create(const char *name, unsigned int align,
+				  slab_flags_t flags,
 				  unsigned int useroffset,
 				  unsigned int usersize,
 				  void (*ctor)(void *))
@@ -487,7 +491,8 @@ kmem_buckets *kmem_buckets_create(const char *name, slab_flags_t flags,
 			if (WARN_ON(!cache_name))
 				goto fail;
 			(*b)[aligned_idx] = kmem_cache_create_usercopy(cache_name, size,
-					0, flags, cache_useroffset,
+					align ?: kmalloc_caches[KMALLOC_NORMAL][idx]->align,
+					flags, cache_useroffset,
 					cache_usersize, ctor);
 			kfree(cache_name);
 			if (WARN_ON(!(*b)[aligned_idx]))
diff --git a/mm/util.c b/mm/util.c
index bf0513d1d3d0..3f04d77ef4b8 100644
--- a/mm/util.c
+++ b/mm/util.c
@@ -199,7 +199,7 @@ static kmem_buckets *user_buckets __ro_after_init;
 
 static int __init init_user_buckets(void)
 {
-	user_buckets = kmem_buckets_create("memdup_user", 0, 0, INT_MAX, NULL);
+	user_buckets = kmem_buckets_create("memdup_user", 0, 0, 0, INT_MAX, NULL);
 
 	return 0;
 }
-- 
2.34.1

[PATCH net-next v5 3/7] mm/slab: Add kmem_buckets_destroy()

[Kees Cook]

kmem_buckets_create() intentionally had no "destroy" counterpart. Every
caller has lived in core kernel code and creates its set once at boot,
so nothing has needed to take one down. However, KUnit tests may be
built module, so we need it now to support the coming tests.

Some caches have size aliases, so the same pointer is stored at more
then one index, so we have to save it, clear all matching instances, and
then free the saved cache pointer. (This is what the bitmap was tracking
before in the "allocation failed" error path.)

When CONFIG_SLAB_BUCKETS=n the whole body compiles away, matching the
ZERO_SIZE_PTR that kmem_buckets_create() hands back in that configuration.

Link: https://lore.kernel.org/all/20240809073309.2134488-1-kees@kernel.org/
Assisted-by: LLM
Signed-off-by: Kees Cook <kees@kernel.org>
---
 include/linux/slab.h |  1 +
 mm/slab_common.c     | 49 +++++++++++++++++++++++++++++++++++++-------
 2 files changed, 43 insertions(+), 7 deletions(-)

diff --git a/include/linux/slab.h b/include/linux/slab.h
index 94709e4e14f3..4e6e74b3a990 100644
--- a/include/linux/slab.h
+++ b/include/linux/slab.h
@@ -894,6 +894,7 @@ kmem_buckets *kmem_buckets_create(const char *name, unsigned int align,
 				  slab_flags_t flags,
 				  unsigned int useroffset, unsigned int usersize,
 				  void (*ctor)(void *));
+void kmem_buckets_destroy(kmem_buckets *bucket);
 
 /*
  * Bulk allocation and freeing operations. These are accelerated in an
diff --git a/mm/slab_common.c b/mm/slab_common.c
index 301f3d4f4ca2..d66e5e56a0f1 100644
--- a/mm/slab_common.c
+++ b/mm/slab_common.c
@@ -438,12 +438,9 @@ kmem_buckets *kmem_buckets_create(const char *name, unsigned int align,
 				  unsigned int usersize,
 				  void (*ctor)(void *))
 {
-	unsigned long mask = 0;
 	unsigned int idx;
 	kmem_buckets *b;
 
-	BUILD_BUG_ON(ARRAY_SIZE(kmalloc_caches[KMALLOC_NORMAL]) > BITS_PER_LONG);
-
 	/*
 	 * When the separate buckets API is not built in, just return
 	 * a non-NULL value for the kmem_buckets pointer, which will be
@@ -497,7 +494,6 @@ kmem_buckets *kmem_buckets_create(const char *name, unsigned int align,
 			kfree(cache_name);
 			if (WARN_ON(!(*b)[aligned_idx]))
 				goto fail;
-			set_bit(aligned_idx, &mask);
 		}
 		if (idx != aligned_idx)
 			(*b)[idx] = (*b)[aligned_idx];
@@ -506,14 +502,53 @@ kmem_buckets *kmem_buckets_create(const char *name, unsigned int align,
 	return b;
 
 fail:
-	for_each_set_bit(idx, &mask, ARRAY_SIZE(kmalloc_caches[KMALLOC_NORMAL]))
-		kmem_cache_destroy((*b)[idx]);
-	kmem_cache_free(kmem_buckets_cache, b);
+	kmem_buckets_destroy(b);
 
 	return NULL;
 }
 EXPORT_SYMBOL(kmem_buckets_create);
 
+/**
+ * kmem_buckets_destroy - Destroy a set of caches made by kmem_buckets_create()
+ * @bucket: The set to destroy, which may be NULL.
+ *
+ * Destroys each cache in @bucket and then frees @bucket itself. As for
+ * kmem_cache_destroy(), every object allocated from @bucket must have been
+ * freed beforehand, and @bucket must not be used afterwards.
+ *
+ * Context: Process context. May sleep, as kmem_cache_destroy() takes the
+ *	    slab mutex and can wait on RCU callbacks for each cache.
+ */
+void kmem_buckets_destroy(kmem_buckets *bucket)
+{
+	unsigned int idx, i;
+
+	if (!IS_ENABLED(CONFIG_SLAB_BUCKETS) || ZERO_OR_NULL_PTR(bucket))
+		return;
+
+	for (idx = 0; idx < ARRAY_SIZE(kmalloc_caches[KMALLOC_NORMAL]); idx++) {
+		struct kmem_cache *cache = (*bucket)[idx];
+
+		if (!cache)
+			continue;
+
+		/*
+		 * Sizes below arch_slab_minalign() share one cache, which
+		 * kmem_buckets_create() then stores at each of their indices.
+		 * Drop every reference to it before destroying it, so that no
+		 * later pass reads a pointer to a cache that is already gone.
+		 */
+		for (i = idx; i < ARRAY_SIZE(kmalloc_caches[KMALLOC_NORMAL]); i++)
+			if ((*bucket)[i] == cache)
+				(*bucket)[i] = NULL;
+
+		kmem_cache_destroy(cache);
+	}
+
+	kmem_cache_free(kmem_buckets_cache, bucket);
+}
+EXPORT_SYMBOL(kmem_buckets_destroy);
+
 /*
  * For a given kmem_cache, kmem_cache_destroy() should only be called
  * once or there will be a use-after-free problem. The actual deletion
-- 
2.34.1

[PATCH net-next v5 4/7] mm/slab: Add tests for the existing kmem_buckets behaviour

[Kees Cook]

kmem_buckets has had no test coverage since it was added. Add tests,
including stuff unique to the bucket design:

 - A bucket allocation comes from a cache of the set's own, and that
   cache carries SLAB_NO_MERGE.

 - Each size class is served by the set, including 96 and 192, which are
   not powers of two and are filled in from an aligned index by
   kmem_buckets_create(). Sizes above KMALLOC_MAX_CACHE_SIZE go to the
   page allocator instead, bucket set or not.

 - Each cache is aligned like the kmalloc cache it mirrors, or, when the
   set was created with an alignment, to that one.

 - With CONFIG_SLAB_BUCKETS=n, kmem_buckets_create() still returns a
   non-NULL (zero size alloc pointer), so that callers only have to check
   for failure, and allocations through it come from the general caches.

 - Destroying a set takes its caches down rather than only freeing the
   set, which is what a module creating one on each load depends on.
   A set freed without its caches would leave the names taken, and the
   next load would warn about every one of them.

The tests skip rather than compile out, which is useful for testing
the CONFIG_SLAB_BUCKETS=n behaviors.

Built and tests passing (with expected skips) on ARCH=x86_64 defconfig
with GCC 16.2.0, with CONFIG_SLAB_BUCKETS as y and n.

Assisted-by: LLM
Signed-off-by: Kees Cook <kees@kernel.org>
---
 lib/tests/slub_kunit.c | 252 +++++++++++++++++++++++++++++++++++++++++
 1 file changed, 252 insertions(+)

diff --git a/lib/tests/slub_kunit.c b/lib/tests/slub_kunit.c
index e3b63f0338d5..3c923a3af825 100644
--- a/lib/tests/slub_kunit.c
+++ b/lib/tests/slub_kunit.c
@@ -1,6 +1,7 @@
 // SPDX-License-Identifier: GPL-2.0
 #include <kunit/test.h>
 #include <kunit/test-bug.h>
+#include <kunit/resource.h>
 #include <linux/mm.h>
 #include <linux/slab.h>
 #include <linux/module.h>
@@ -474,6 +475,251 @@ static int test_init(struct kunit *test)
 	return 0;
 }
 
+/* Destroy buckets on test exit so a failed KUNIT_ASSERT_*() doesn't leak. */
+KUNIT_DEFINE_ACTION_WRAPPER(destroy_buckets, kmem_buckets_destroy, kmem_buckets *);
+
+#define KUNIT_ASSERT_BUCKETS_CREATED(test, b)					\
+	do {									\
+		KUNIT_ASSERT_NOT_NULL(test, b);					\
+		KUNIT_ASSERT_EQ(test, 0,					\
+				kunit_add_action_or_reset(test,			\
+							  destroy_buckets, b)); \
+	} while (0)
+
+/*
+ * The cache an allocation came from, or NULL if it came from no cache at
+ * all, e.g. a size too big for any of them is served by the page allocator.
+ */
+static struct kmem_cache *cache_of(void *p)
+{
+	struct slab *slab = virt_to_slab(p);
+
+	return slab ? slab->slab_cache : NULL;
+}
+
+/*
+ * A bucket set exists to keep its allocations out of the caches everything
+ * else uses, so check the two things that make that true: they come from a
+ * cache of the set's own, and that cache is never merged into another.
+ */
+static void test_kmem_buckets_isolation(struct kunit *test)
+{
+	struct kmem_cache *bucket_cache, *general_cache;
+	kmem_buckets *b;
+	void *p, *q;
+
+	if (!IS_ENABLED(CONFIG_SLAB_BUCKETS))
+		kunit_skip(test, "needs CONFIG_SLAB_BUCKETS");
+
+	b = kmem_buckets_create("isolated_buckets", 0, 0, 0, INT_MAX, NULL);
+	KUNIT_ASSERT_BUCKETS_CREATED(test, b);
+
+	/*
+	 * Free each allocation before asserting on the next one: the cache
+	 * outlives its objects, so nothing below needs them, and an assertion
+	 * that leaves one behind would make the deferred teardown report a
+	 * cache that is still in use.
+	 */
+	p = kmem_buckets_alloc(b, 128, GFP_KERNEL);
+	KUNIT_ASSERT_NOT_NULL(test, p);
+	bucket_cache = cache_of(p);
+	kfree(p);
+	KUNIT_ASSERT_NOT_NULL(test, bucket_cache);
+
+	KUNIT_EXPECT_TRUE_MSG(test, strstarts(bucket_cache->name, "isolated_buckets-"),
+			      "expected a bucket cache, got %s", bucket_cache->name);
+
+	/*
+	 * Cache merging is on by default, and a bucket cache merged into a
+	 * same-sized general one would quietly undo the whole separation.
+	 */
+	KUNIT_EXPECT_TRUE(test, bucket_cache->flags & SLAB_NO_MERGE);
+
+	q = kmalloc(128, GFP_KERNEL);
+	KUNIT_ASSERT_NOT_NULL(test, q);
+	general_cache = cache_of(q);
+	kfree(q);
+	KUNIT_ASSERT_NOT_NULL(test, general_cache);
+
+	KUNIT_EXPECT_PTR_NE(test, bucket_cache, general_cache);
+}
+
+/*
+ * Every size class gets its own cache in the set, including the ones that
+ * are not powers of two and are filled in from an aligned index. Sizes past
+ * the largest cache are served by the page allocator, bucket set or not.
+ */
+static void test_kmem_buckets_sizes(struct kunit *test)
+{
+	static const size_t sizes[] = { 8, 96, 192, 1024, 4096 };
+	struct kmem_cache *c;
+	kmem_buckets *b;
+	void *p;
+	int i;
+
+	if (!IS_ENABLED(CONFIG_SLAB_BUCKETS))
+		kunit_skip(test, "needs CONFIG_SLAB_BUCKETS");
+
+	b = kmem_buckets_create("sized_buckets", 0, 0, 0, INT_MAX, NULL);
+	KUNIT_ASSERT_BUCKETS_CREATED(test, b);
+
+	for (i = 0; i < ARRAY_SIZE(sizes); i++) {
+		p = kmem_buckets_alloc(b, sizes[i], GFP_KERNEL);
+		KUNIT_ASSERT_NOT_NULL(test, p);
+		c = cache_of(p);
+		kfree(p);
+		KUNIT_ASSERT_NOT_NULL(test, c);
+
+		KUNIT_EXPECT_TRUE_MSG(test, strstarts(c->name, "sized_buckets-"),
+				      "size %zu: expected a bucket cache, got %s",
+				      sizes[i], c->name);
+		KUNIT_EXPECT_GE(test, c->object_size, sizes[i]);
+	}
+
+	/* Too big for any cache: a folio from the page allocator, not a slab. */
+	p = kmem_buckets_alloc(b, KMALLOC_MAX_CACHE_SIZE + 1, GFP_KERNEL);
+	KUNIT_ASSERT_NOT_NULL(test, p);
+	c = cache_of(p);
+	kfree(p);
+
+	KUNIT_EXPECT_NULL(test, c);
+}
+
+/*
+ * A bucket cache stands in for a kmalloc cache, so it has to be aligned like
+ * one. The DMA layer decides whether a buffer needs bouncing from its size,
+ * on the grounds that a kmalloc cache of that size is already aligned for
+ * the device, so a weaker alignment here is not something a caller can see
+ * coming. Without slab debugging the size implies the alignment and this
+ * holds either way; with it, only the cache's own alignment does.
+ */
+static void test_kmem_buckets_alignment(struct kunit *test)
+{
+	static const size_t sizes[] = { 128, 512, 2048 };
+	struct kmem_cache *bucket_cache, *general_cache;
+	kmem_buckets *b;
+	void *p;
+	int i;
+
+	if (!IS_ENABLED(CONFIG_SLAB_BUCKETS))
+		kunit_skip(test, "needs CONFIG_SLAB_BUCKETS");
+
+	b = kmem_buckets_create("aligned_buckets", 0, 0, 0, INT_MAX, NULL);
+	KUNIT_ASSERT_BUCKETS_CREATED(test, b);
+
+	for (i = 0; i < ARRAY_SIZE(sizes); i++) {
+		p = kmem_buckets_alloc(b, sizes[i], GFP_KERNEL);
+		KUNIT_ASSERT_NOT_NULL(test, p);
+		bucket_cache = cache_of(p);
+		KUNIT_EXPECT_TRUE_MSG(test,
+				      IS_ALIGNED((unsigned long)p, ARCH_DMA_MINALIGN),
+				      "size %zu: object %p is not %d byte aligned",
+				      sizes[i], p, (int)ARCH_DMA_MINALIGN);
+		kfree(p);
+
+		p = kmalloc(sizes[i], GFP_KERNEL);
+		KUNIT_ASSERT_NOT_NULL(test, p);
+		general_cache = cache_of(p);
+		kfree(p);
+
+		KUNIT_ASSERT_NOT_NULL(test, bucket_cache);
+		KUNIT_ASSERT_NOT_NULL(test, general_cache);
+		KUNIT_EXPECT_EQ_MSG(test, bucket_cache->align, general_cache->align,
+				    "size %zu: bucket cache aligned to %u, %s to %u",
+				    sizes[i], bucket_cache->align,
+				    general_cache->name, general_cache->align);
+	}
+}
+
+/*
+ * A set created with an alignment gives every one of its caches that
+ * alignment in place of the kmalloc caches' own. 256 is stronger than
+ * kmalloc's alignment for the 64 and 128 byte caches and weaker than it
+ * for the 512 and 2048 byte ones, so this checks the override both ways.
+ */
+static void test_kmem_buckets_explicit_alignment(struct kunit *test)
+{
+	static const size_t sizes[] = { 64, 128, 512, 2048 };
+	const unsigned int align = 256;
+	struct kmem_cache *c;
+	kmem_buckets *b;
+	void *p;
+	int i;
+
+	if (!IS_ENABLED(CONFIG_SLAB_BUCKETS))
+		kunit_skip(test, "needs CONFIG_SLAB_BUCKETS");
+
+	b = kmem_buckets_create("explicit_buckets", align, 0, 0, INT_MAX, NULL);
+	KUNIT_ASSERT_BUCKETS_CREATED(test, b);
+
+	for (i = 0; i < ARRAY_SIZE(sizes); i++) {
+		p = kmem_buckets_alloc(b, sizes[i], GFP_KERNEL);
+		KUNIT_ASSERT_NOT_NULL(test, p);
+		c = cache_of(p);
+		KUNIT_EXPECT_TRUE_MSG(test, IS_ALIGNED((unsigned long)p, align),
+				      "size %zu: object %p is not %u byte aligned",
+				      sizes[i], p, align);
+		kfree(p);
+
+		KUNIT_ASSERT_NOT_NULL(test, c);
+		KUNIT_EXPECT_EQ_MSG(test, c->align, align,
+				    "size %zu: bucket cache aligned to %u, not %u",
+				    sizes[i], c->align, align);
+	}
+}
+
+/*
+ * With the feature compiled out, kmem_buckets_create() still returns
+ * something non-NULL so that callers only have to check for failure, and
+ * allocations through it work (i.e. come from the general caches).
+ */
+static void test_kmem_buckets_disabled(struct kunit *test)
+{
+	kmem_buckets *b;
+	struct kmem_cache *c;
+	void *p;
+
+	if (IS_ENABLED(CONFIG_SLAB_BUCKETS))
+		kunit_skip(test, "only meaningful without CONFIG_SLAB_BUCKETS");
+
+	b = kmem_buckets_create("disabled_buckets", 0, 0, 0, INT_MAX, NULL);
+	KUNIT_ASSERT_BUCKETS_CREATED(test, b);
+
+	p = kmem_buckets_alloc(b, 128, GFP_KERNEL);
+	KUNIT_ASSERT_NOT_NULL(test, p);
+	c = cache_of(p);
+	kfree(p);
+	KUNIT_ASSERT_NOT_NULL(test, c);
+
+	KUNIT_EXPECT_TRUE_MSG(test, !strstarts(c->name, "disabled_buckets-"),
+			      "expected a general cache, got %s", c->name);
+}
+
+/* Destroying a set has to take its caches down, not just free the set. */
+static void test_kmem_buckets_destroy(struct kunit *test)
+{
+	kmem_buckets *b;
+	void *p;
+
+	if (!IS_ENABLED(CONFIG_SLAB_BUCKETS))
+		kunit_skip(test, "needs CONFIG_SLAB_BUCKETS");
+
+	b = kmem_buckets_create("destroyed_buckets", 0, 0, 0, INT_MAX, NULL);
+	KUNIT_ASSERT_NOT_NULL(test, b);
+
+	/*
+	 * Deliberately leaked, as test_leak_destroy() leaks its own: the
+	 * teardown below has to find it. kmem_cache_destroy() unlists the
+	 * cache either way, so the name is still released.
+	 */
+	p = kmem_buckets_alloc(b, 128, GFP_KERNEL);
+	KUNIT_EXPECT_NOT_NULL(test, p);
+
+	kmem_buckets_destroy(b);
+
+	KUNIT_EXPECT_EQ(test, 2, slab_errors);
+}
+
 static struct kunit_case test_cases[] = {
 	KUNIT_CASE(test_clobber_zone),
 
@@ -495,6 +741,12 @@ static struct kunit_case test_cases[] = {
 #if defined(CONFIG_KPROBES) && defined(CONFIG_SMP)
 	KUNIT_CASE_SLOW(test_kmalloc_nolock_and_friends_kprobe),
 #endif
+	KUNIT_CASE(test_kmem_buckets_isolation),
+	KUNIT_CASE(test_kmem_buckets_sizes),
+	KUNIT_CASE(test_kmem_buckets_alignment),
+	KUNIT_CASE(test_kmem_buckets_explicit_alignment),
+	KUNIT_CASE(test_kmem_buckets_disabled),
+	KUNIT_CASE(test_kmem_buckets_destroy),
 	{}
 };
 
-- 
2.34.1

[PATCH net-next v5 5/7] mm/slab: Provide kmalloc type fallback for bucket allocations

[Kees Cook]

kmem_buckets_create() clones kmalloc_caches[KMALLOC_NORMAL].
kmalloc_slab() figures out the kmalloc type the caller asks for, but
then ignored it whenever a bucket set was in use, returning a normal
cache regardless. This would be a problem if a caller asked for GFP_DMA,
__GFP_ACCOUNT, etc. None of the current users do this, so there is no
problem, but it makes adding new users fragile. For example, skb data[1]
needs to handle GFP_DMA (rarely) and __GFP_ACCOUNT (often).

Send those allocations to the general caches instead so nothing breaks and
regular allocations remain isolated with the bucket. The kmem_bucket_type
enum contains only a single item here, but will be expanded in the next
patch.

The test checks mem_cgroup_kmem_disabled() before expecting an accounted
cache, so export it for KUnit only, which a modular build of the test
needs to link.

Built and tests pass with ARCH=x86_64 defconfig with GCC 16.2.0, with
CONFIG_SLAB_BUCKETS as y and n.

Assisted-by: LLM
Link: https://lore.kernel.org/all/04debe19-bbe8-4b5f-9668-753d1f97832d@redhat.com/ [1]
Signed-off-by: Kees Cook <kees@kernel.org>
---
 include/linux/slab.h   | 13 ++++++++++
 mm/slab.h              | 23 ++++++++++++++++--
 lib/tests/slub_kunit.c | 54 ++++++++++++++++++++++++++++++++++++++++++
 mm/memcontrol.c        |  2 ++
 4 files changed, 90 insertions(+), 2 deletions(-)

diff --git a/include/linux/slab.h b/include/linux/slab.h
index 4e6e74b3a990..bdd00235d4b0 100644
--- a/include/linux/slab.h
+++ b/include/linux/slab.h
@@ -742,6 +742,19 @@ typedef struct kmem_cache * kmem_buckets[KMALLOC_SHIFT_HIGH + 1];
 
 extern kmem_buckets kmalloc_caches[NR_KMALLOC_TYPES];
 
+/*
+ * The kmalloc types a bucket set can hold a copy of. This is deliberately not
+ * enum kmalloc_cache_type: the KMALLOC_PARTITION copies are all "normal" to a
+ * bucket set, which already separates what they were there to separate, so
+ * indexing by those would mean up to KMALLOC_PARTITION_CACHES_NR unusable
+ * rows per set. Allocations of any type not listed here are served by the
+ * general caches.
+ */
+enum kmem_bucket_type {
+	KMEM_BUCKET_NORMAL = 0,
+	NR_KMEM_BUCKET_TYPES
+};
+
 /*
  * Define gfp bits that should not be set for KMALLOC_NORMAL.
  */
diff --git a/mm/slab.h b/mm/slab.h
index 8fd6835e4235..7f1bfee83b92 100644
--- a/mm/slab.h
+++ b/mm/slab.h
@@ -421,6 +421,26 @@ static inline unsigned int size_index_elem(unsigned int bytes)
 	return (bytes - 1) / 8;
 }
 
+/*
+ * Which set of buckets to use for the given kmalloc_cache_type. If not
+ * handled by the kmem_buckets, fall back to general caches.
+ */
+static inline kmem_buckets *
+kmalloc_choose_bucket(kmem_buckets *bucket, enum kmalloc_cache_type type)
+{
+	enum kmem_bucket_type btype;
+
+	if (!bucket)
+		return &kmalloc_caches[type];
+
+	if (type <= KMALLOC_PARTITION_END)
+		btype = KMEM_BUCKET_NORMAL;
+	else
+		return &kmalloc_caches[type];	/* No set holds a row for it. */
+
+	return &bucket[btype];
+}
+
 /*
  * Find the kmem_cache structure that serves a given size of
  * allocation
@@ -438,8 +458,7 @@ kmalloc_slab(size_t size, kmem_buckets *b, gfp_t flags, kmalloc_token_t token,
 	if (alloc_flags & SLAB_ALLOC_NO_OBJ_EXT)
 		type = KMALLOC_NO_OBJ_EXT;
 
-	if (!b)
-		b = &kmalloc_caches[type];
+	b = kmalloc_choose_bucket(b, type);
 	if (size <= 192)
 		index = kmalloc_size_index[size_index_elem(size)];
 	else
diff --git a/lib/tests/slub_kunit.c b/lib/tests/slub_kunit.c
index 3c923a3af825..9768de01e6f9 100644
--- a/lib/tests/slub_kunit.c
+++ b/lib/tests/slub_kunit.c
@@ -720,6 +720,58 @@ static void test_kmem_buckets_destroy(struct kunit *test)
 	KUNIT_EXPECT_EQ(test, 2, slab_errors);
 }
 
+/*
+ * A bucket set holds only the kmalloc types it was created with, so an
+ * allocation that asks for a different one has to come from the general
+ * caches. Check that it does, rather than being served a normal cache that
+ * does not satisfy what the flags asked for.
+ */
+static void test_kmem_buckets_type_fallback(struct kunit *test)
+{
+	struct kmem_cache *c;
+	kmem_buckets *b;
+	void *p;
+
+	if (!IS_ENABLED(CONFIG_SLAB_BUCKETS))
+		kunit_skip(test, "needs CONFIG_SLAB_BUCKETS");
+
+	b = kmem_buckets_create("test_buckets", 0, 0, 0, INT_MAX, NULL);
+	KUNIT_ASSERT_BUCKETS_CREATED(test, b);
+
+	/* A plain allocation stays isolated in the bucket set. */
+	p = kmem_buckets_alloc(b, 128, GFP_KERNEL);
+	KUNIT_ASSERT_NOT_NULL(test, p);
+	c = cache_of(p);
+	kfree(p);
+	KUNIT_ASSERT_NOT_NULL(test, c);
+
+	KUNIT_EXPECT_TRUE_MSG(test, strstarts(c->name, "test_buckets-"),
+			      "expected a bucket cache, got %s", c->name);
+
+	/* One that needs ZONE_DMA cannot, so it falls back. */
+	if (IS_ENABLED(CONFIG_ZONE_DMA)) {
+		p = kmem_buckets_alloc(b, 128, GFP_KERNEL | GFP_DMA);
+		KUNIT_ASSERT_NOT_NULL(test, p);
+		c = cache_of(p);
+		kfree(p);
+		KUNIT_ASSERT_NOT_NULL(test, c);
+
+		KUNIT_EXPECT_TRUE_MSG(test, strstarts(c->name, "dma-kmalloc-"),
+				      "expected a DMA cache, got %s", c->name);
+	}
+
+	/* Nor can one that has to be accounted. */
+	if (IS_ENABLED(CONFIG_MEMCG) && !mem_cgroup_kmem_disabled()) {
+		p = kmem_buckets_alloc(b, 128, GFP_KERNEL | __GFP_ACCOUNT);
+		KUNIT_ASSERT_NOT_NULL(test, p);
+		c = virt_to_slab(p)->slab_cache;
+		kfree(p);
+
+		KUNIT_EXPECT_TRUE_MSG(test, strstarts(c->name, "kmalloc-cg-"),
+				      "expected an accounted cache, got %s", c->name);
+	}
+}
+
 static struct kunit_case test_cases[] = {
 	KUNIT_CASE(test_clobber_zone),
 
@@ -747,6 +799,7 @@ static struct kunit_case test_cases[] = {
 	KUNIT_CASE(test_kmem_buckets_explicit_alignment),
 	KUNIT_CASE(test_kmem_buckets_disabled),
 	KUNIT_CASE(test_kmem_buckets_destroy),
+	KUNIT_CASE(test_kmem_buckets_type_fallback),
 	{}
 };
 
@@ -757,5 +810,6 @@ static struct kunit_suite test_suite = {
 };
 kunit_test_suite(test_suite);
 
+MODULE_IMPORT_NS("EXPORTED_FOR_KUNIT_TESTING");
 MODULE_DESCRIPTION("Kunit tests for slub allocator");
 MODULE_LICENSE("GPL");
diff --git a/mm/memcontrol.c b/mm/memcontrol.c
index 1271d390b617..5ceeb5a0b614 100644
--- a/mm/memcontrol.c
+++ b/mm/memcontrol.c
@@ -62,6 +62,7 @@
 #include <linux/seq_buf.h>
 #include <linux/sched/isolation.h>
 #include <linux/kmemleak.h>
+#include <kunit/visibility.h>
 #include "internal.h"
 #include "swap.h"
 #include "swap_table.h"
@@ -135,6 +136,7 @@ bool mem_cgroup_kmem_disabled(void)
 {
 	return cgroup_memory_nokmem;
 }
+EXPORT_SYMBOL_IF_KUNIT(mem_cgroup_kmem_disabled);
 
 static void memcg_uncharge(struct mem_cgroup *memcg, unsigned int nr_pages);
 
-- 
2.34.1

[PATCH net-next v5 6/7] mm/slab: Let a bucket set handle __GFP_ACCOUNT

[Kees Cook]

A bucket set holds one row of caches, cloned from KMALLOC_NORMAL, and an
allocation of any other kmalloc type falls back to the general caches.
Extend this to handle __GFP_ACCOUNT, so that a single bucket user can
isolate either GFP_KERNEL or GFP_KERNEL_ACCOUNT allocations, as is
needed for skb data, where AF_UNIX uses:

	sk->sk_allocation = GFP_KERNEL_ACCOUNT;

The coverage is selected at bucket creation time:

	b = kmem_buckets_create_types(name, 0, flags, 0, INT_MAX, NULL,
				      BIT(KMEM_BUCKET_NORMAL) |
				      BIT(KMEM_BUCKET_CGROUP));

The prior kmem_buckets_create() function keeps its name and defaults
to only KMEM_BUCKET_NORMAL, leaving existing users as-is.

Only the accounted type is offered. Nothing wants a reclaimable or
no-obj-ext row, and of the twelve places passing GFP_DMA to an skb
allocator, all are rare hardware: b44, b43legacy, prestera, and s390 ctcm.

The choice is made at creation rather than every set getting every type
because the rows, when populated, are not free. Each holds 13 caches, and
a cache is a 1208 byte struct plus an unconditional per-cpu allocation,
a node struct, and an entry in /proc/slabinfo and under /sys/kernel/slab.

The rows are indexed by enum kmem_bucket_type, not enum
kmalloc_cache_type, which counts each KMALLOC_PARTITION copy as a type
of its own. Indexed by kmalloc_cache_type, a set would carry rows it
can never use: on x86_64 with CONFIG_KMALLOC_PARTITION_CACHES=y,
CONFIG_MEMCG=y and CONFIG_ZONE_DMA=y, 20 rows of 112 bytes (2240 bytes)
per set, against the 2 rows (224 bytes) used here.

KMEM_BUCKET_CGROUP collapses to KMEM_BUCKET_NORMAL without CONFIG_MEMCG,
exactly as KMALLOC_CGROUP does, so NR_KMEM_BUCKET_TYPES is 1 there and a
bucket set is the same single row it is today. Where the type is asked for
but the system is not creating caches of it (under "cgroup.memory=nokmem")
the row is aliased to the normal one, as new_kmalloc_cache() does for the
general caches, so those allocations stay isolated rather than falling
back to the general caches.

Built and tests pass (and skip as expected) on ARCH=x86_64 defconfig
with GCC 16.2.0 in all combinations of CONFIG_SLAB_BUCKETS=y/n and
CONFIG_MEMCG=y/n/y+"cgroup.memory=nokmem".

Assisted-by: LLM
Signed-off-by: Kees Cook <kees@kernel.org>
---
 include/linux/slab.h   |  53 +++++++++++++--
 mm/slab.h              |  23 ++++++-
 lib/tests/slub_kunit.c | 101 ++++++++++++++++++++++++++--
 mm/slab_common.c       | 149 ++++++++++++++++++++++++++++++++---------
 4 files changed, 283 insertions(+), 43 deletions(-)

diff --git a/include/linux/slab.h b/include/linux/slab.h
index bdd00235d4b0..ee359dc90333 100644
--- a/include/linux/slab.h
+++ b/include/linux/slab.h
@@ -752,6 +752,11 @@ extern kmem_buckets kmalloc_caches[NR_KMALLOC_TYPES];
  */
 enum kmem_bucket_type {
 	KMEM_BUCKET_NORMAL = 0,
+#ifdef CONFIG_MEMCG
+	KMEM_BUCKET_CGROUP,
+#else
+	KMEM_BUCKET_CGROUP = KMEM_BUCKET_NORMAL,
+#endif
 	NR_KMEM_BUCKET_TYPES
 };
 
@@ -903,10 +908,50 @@ void *kmem_cache_alloc_lru_noprof(struct kmem_cache *s, struct list_lru *lru,
 bool kmem_cache_charge(void *objp, gfp_t gfpflags);
 void kmem_cache_free(struct kmem_cache *s, void *objp);
 
-kmem_buckets *kmem_buckets_create(const char *name, unsigned int align,
-				  slab_flags_t flags,
-				  unsigned int useroffset, unsigned int usersize,
-				  void (*ctor)(void *));
+kmem_buckets *kmem_buckets_create_types(const char *name, unsigned int align,
+					slab_flags_t flags,
+					unsigned int useroffset, unsigned int usersize,
+					void (*ctor)(void *),
+					unsigned int type_mask);
+
+/**
+ * kmem_buckets_create - Create a set of caches that handle dynamic sized
+ *			 allocations via kmem_buckets_alloc()
+ * @name: A prefix string which is used in /proc/slabinfo to identify this
+ *	  cache. The individual caches with have their sizes as the suffix.
+ * @align: The required alignment for the objects, or 0 to give each cache
+ *	   the alignment of the kmalloc cache of the same size, as a caller
+ *	   moving from kmalloc() may depend on.
+ * @flags: SLAB flags (see kmem_cache_create() for details).
+ * @useroffset: Starting offset within an allocation that may be copied
+ *		to/from userspace.
+ * @usersize: How many bytes, starting at @useroffset, may be copied
+ *		to/from userspace.
+ * @ctor: A constructor for the objects, run when new allocations are made.
+ *
+ * Covers KMEM_BUCKET_NORMAL only. Allocations needing another kmalloc type
+ * are served by the general caches, keeping the type they asked for and
+ * losing only the isolation. Use kmem_buckets_create_types() to cover more.
+ *
+ * Context: Cannot be called within an interrupt, but can be interrupted.
+ *
+ * Return: a pointer to the cache on success, NULL on failure. When
+ * CONFIG_SLAB_BUCKETS is not enabled, ZERO_SIZE_PTR is returned, and
+ * subsequent calls to kmem_buckets_alloc() will fall back to kmalloc().
+ * (i.e. callers only need to check for NULL on failure.)
+ */
+static inline kmem_buckets *kmem_buckets_create(const char *name,
+						unsigned int align,
+						slab_flags_t flags,
+						unsigned int useroffset,
+						unsigned int usersize,
+						void (*ctor)(void *))
+{
+	return kmem_buckets_create_types(name, align, flags, useroffset,
+					 usersize, ctor,
+					 BIT(KMEM_BUCKET_NORMAL));
+}
+
 void kmem_buckets_destroy(kmem_buckets *bucket);
 
 /*
diff --git a/mm/slab.h b/mm/slab.h
index 7f1bfee83b92..2af44e09edda 100644
--- a/mm/slab.h
+++ b/mm/slab.h
@@ -435,10 +435,31 @@ kmalloc_choose_bucket(kmem_buckets *bucket, enum kmalloc_cache_type type)
 
 	if (type <= KMALLOC_PARTITION_END)
 		btype = KMEM_BUCKET_NORMAL;
+	else if (IS_ENABLED(CONFIG_MEMCG) && type == KMALLOC_CGROUP)
+		btype = KMEM_BUCKET_CGROUP;
 	else
 		return &kmalloc_caches[type];	/* No set holds a row for it. */
 
-	return &bucket[btype];
+	/*
+	 * Either this row was created, and holds a cache everywhere the
+	 * general caches hold one, or it was never created and holds nothing.
+	 * Test with the KMALLOC_SHIFT_LOW which exists in every configuration.
+	 */
+	if (likely(bucket[btype][KMALLOC_SHIFT_LOW]))
+		return &bucket[btype];
+
+	/*
+	 * A row this set _could_ have held, but was not created with: the type
+	 * mask passed to kmem_buckets_create_types() did not cover what its
+	 * callers actually tried to allocate. Report the mismatch but still
+	 * fall back to the general caches.
+	 *
+	 * At present, only __GFP_ACCOUNT can be missing.
+	 */
+	WARN_ONCE(1,
+		  "kmem_buckets: __GFP_ACCOUNT needs BIT(KMEM_BUCKET_CGROUP) in create mask\n");
+
+	return &kmalloc_caches[type];
 }
 
 /*
diff --git a/lib/tests/slub_kunit.c b/lib/tests/slub_kunit.c
index 9768de01e6f9..bc5e52f20179 100644
--- a/lib/tests/slub_kunit.c
+++ b/lib/tests/slub_kunit.c
@@ -760,15 +760,104 @@ static void test_kmem_buckets_type_fallback(struct kunit *test)
 				      "expected a DMA cache, got %s", c->name);
 	}
 
-	/* Nor can one that has to be accounted. */
+	/*
+	 * An accounted allocation would fall back too, but a bucket set can
+	 * hold that type, so reaching the fallback means the create mask was
+	 * wrong and kmalloc_slab() warns. Not exercised here for that reason;
+	 * test_kmem_buckets_type_covered() checks the type that is asked for.
+	 */
+}
+
+/*
+ * A bucket set created for a kmalloc type keeps those allocations isolated
+ * too, rather than sending them to the general caches. Where nothing creates
+ * accounted caches at all, the row aliases the normal one, so this also
+ * covers tearing down a set whose rows share their caches.
+ */
+static void test_kmem_buckets_type_covered(struct kunit *test)
+{
+	struct kmem_cache *c, *normal_cache;
+	kmem_buckets *b;
+	void *p;
+
+	if (!IS_ENABLED(CONFIG_SLAB_BUCKETS))
+		kunit_skip(test, "needs CONFIG_SLAB_BUCKETS");
+
+	b = kmem_buckets_create_types("covered_buckets", 0, 0, 0, INT_MAX, NULL,
+				      BIT(KMEM_BUCKET_NORMAL) |
+				      BIT(KMEM_BUCKET_CGROUP));
+	KUNIT_ASSERT_BUCKETS_CREATED(test, b);
+
+	p = kmem_buckets_alloc(b, 128, GFP_KERNEL);
+	KUNIT_ASSERT_NOT_NULL(test, p);
+	normal_cache = cache_of(p);
+	kfree(p);
+	KUNIT_ASSERT_NOT_NULL(test, normal_cache);
+
+	KUNIT_EXPECT_TRUE_MSG(test, strstarts(normal_cache->name, "covered_buckets-128"),
+			      "expected the normal bucket cache, got %s",
+			      normal_cache->name);
+
+	/* Accounted, and still in the bucket set rather than kmalloc-cg-*. */
+	p = kmem_buckets_alloc(b, 128, GFP_KERNEL | __GFP_ACCOUNT);
+	KUNIT_ASSERT_NOT_NULL(test, p);
+	c = cache_of(p);
+	kfree(p);
+	KUNIT_ASSERT_NOT_NULL(test, c);
+
 	if (IS_ENABLED(CONFIG_MEMCG) && !mem_cgroup_kmem_disabled()) {
-		p = kmem_buckets_alloc(b, 128, GFP_KERNEL | __GFP_ACCOUNT);
+		KUNIT_EXPECT_TRUE_MSG(test, strstarts(c->name, "covered_buckets-cg-"),
+				      "expected the accounted bucket cache, got %s",
+				      c->name);
+		KUNIT_EXPECT_TRUE(test, c->flags & SLAB_ACCOUNT);
+	} else {
+		/*
+		 * Nothing is creating accounted caches, so the row aliases
+		 * the normal one and the allocation lands there -- isolated
+		 * still, just not separately accounted.
+		 */
+		KUNIT_EXPECT_PTR_EQ(test, c, normal_cache);
+	}
+}
+
+/*
+ * The alignment a set is created with reaches every row it holds, not just
+ * the normal one. 256 is stronger than kmalloc's alignment for 128 byte
+ * objects, so a row built without it would show here.
+ */
+static void test_kmem_buckets_type_covered_alignment(struct kunit *test)
+{
+	static const gfp_t gfps[] = { GFP_KERNEL, GFP_KERNEL | __GFP_ACCOUNT };
+	const unsigned int align = 256;
+	struct kmem_cache *c;
+	kmem_buckets *b;
+	void *p;
+	int i;
+
+	if (!IS_ENABLED(CONFIG_SLAB_BUCKETS))
+		kunit_skip(test, "needs CONFIG_SLAB_BUCKETS");
+
+	b = kmem_buckets_create_types("covered_aligned", align, 0, 0, INT_MAX,
+				      NULL, BIT(KMEM_BUCKET_NORMAL) |
+				      BIT(KMEM_BUCKET_CGROUP));
+	KUNIT_ASSERT_BUCKETS_CREATED(test, b);
+
+	for (i = 0; i < ARRAY_SIZE(gfps); i++) {
+		p = kmem_buckets_alloc(b, 128, gfps[i]);
 		KUNIT_ASSERT_NOT_NULL(test, p);
-		c = virt_to_slab(p)->slab_cache;
+		c = cache_of(p);
+		KUNIT_EXPECT_TRUE_MSG(test, IS_ALIGNED((unsigned long)p, align),
+				      "gfp %pGg: object %p is not %u byte aligned",
+				      &gfps[i], p, align);
 		kfree(p);
 
-		KUNIT_EXPECT_TRUE_MSG(test, strstarts(c->name, "kmalloc-cg-"),
-				      "expected an accounted cache, got %s", c->name);
+		KUNIT_ASSERT_NOT_NULL(test, c);
+		KUNIT_EXPECT_TRUE_MSG(test, strstarts(c->name, "covered_aligned-"),
+				      "gfp %pGg: expected a bucket cache, got %s",
+				      &gfps[i], c->name);
+		KUNIT_EXPECT_EQ_MSG(test, c->align, align,
+				    "gfp %pGg: %s aligned to %u, not %u",
+				    &gfps[i], c->name, c->align, align);
 	}
 }
 
@@ -800,6 +889,8 @@ static struct kunit_case test_cases[] = {
 	KUNIT_CASE(test_kmem_buckets_disabled),
 	KUNIT_CASE(test_kmem_buckets_destroy),
 	KUNIT_CASE(test_kmem_buckets_type_fallback),
+	KUNIT_CASE(test_kmem_buckets_type_covered),
+	KUNIT_CASE(test_kmem_buckets_type_covered_alignment),
 	{}
 };
 
diff --git a/mm/slab_common.c b/mm/slab_common.c
index d66e5e56a0f1..c045e2d41ee5 100644
--- a/mm/slab_common.c
+++ b/mm/slab_common.c
@@ -410,9 +410,15 @@ EXPORT_SYMBOL(__kmem_cache_create_args);
 
 static struct kmem_cache *kmem_buckets_cache __ro_after_init;
 
+static int kmem_buckets_create_row(kmem_buckets *b,
+				   enum kmalloc_cache_type type,
+				   const char *name, unsigned int align,
+				   slab_flags_t flags, unsigned int useroffset,
+				   unsigned int usersize, void (*ctor)(void *));
+
 /**
- * kmem_buckets_create - Create a set of caches that handle dynamic sized
- *			 allocations via kmem_buckets_alloc()
+ * kmem_buckets_create_types - Create a set of caches that handle dynamic sized
+ *			       allocations via kmem_buckets_alloc()
  * @name: A prefix string which is used in /proc/slabinfo to identify this
  *	  cache. The individual caches with have their sizes as the suffix.
  * @align: The required alignment for the objects, or 0 to give each cache
@@ -424,6 +430,11 @@ static struct kmem_cache *kmem_buckets_cache __ro_after_init;
  * @usersize: How many bytes, starting at @useroffset, may be copied
  *		to/from userspace.
  * @ctor: A constructor for the objects, run when new allocations are made.
+ * @type_mask: Which kmalloc types to hold caches for, as a mask of
+ *	       BIT(KMEM_BUCKET_*). KMEM_BUCKET_NORMAL is always included.
+ *	       Allocations of a type that is not covered are served by the
+ *	       general caches instead, so a caller need not know in advance
+ *	       which types its own callers will ask for.
  *
  * Context: Cannot be called within an interrupt, but can be interrupted.
  *
@@ -432,13 +443,14 @@ static struct kmem_cache *kmem_buckets_cache __ro_after_init;
  * subsequent calls to kmem_buckets_alloc() will fall back to kmalloc().
  * (i.e. callers only need to check for NULL on failure.)
  */
-kmem_buckets *kmem_buckets_create(const char *name, unsigned int align,
-				  slab_flags_t flags,
-				  unsigned int useroffset,
-				  unsigned int usersize,
-				  void (*ctor)(void *))
-{
-	unsigned int idx;
+kmem_buckets *kmem_buckets_create_types(const char *name, unsigned int align,
+					slab_flags_t flags,
+					unsigned int useroffset,
+					unsigned int usersize,
+					void (*ctor)(void *),
+					unsigned int type_mask)
+{
+	enum kmem_bucket_type btype;
 	kmem_buckets *b;
 
 	/*
@@ -457,20 +469,87 @@ kmem_buckets *kmem_buckets_create(const char *name, unsigned int align,
 		return NULL;
 
 	flags |= SLAB_NO_MERGE;
+	type_mask |= BIT(KMEM_BUCKET_NORMAL);
+
+	for (btype = 0; btype < NR_KMEM_BUCKET_TYPES; btype++) {
+		enum kmalloc_cache_type src = KMALLOC_NORMAL;
+		slab_flags_t type_flags = 0;
+
+		if (!(type_mask & BIT(btype)))
+			continue;
+
+		/*
+		 * Under CONFIG_MEMCG=n the two types are the same value, so
+		 * the IS_ENABLED() is what keeps the normal row out of here.
+		 */
+		if (IS_ENABLED(CONFIG_MEMCG) && btype == KMEM_BUCKET_CGROUP) {
+			/*
+			 * Aliasing below reads the normal row, so this loop
+			 * must have built it already. That holds only while
+			 * the normal type sorts first.
+			 */
+			BUILD_BUG_ON(KMEM_BUCKET_CGROUP <= KMEM_BUCKET_NORMAL);
+
+			/*
+			 * Nothing anywhere is creating accounted caches, as
+			 * with "cgroup.memory=nokmem". Point this row's
+			 * entries at the normal row's caches, the way
+			 * new_kmalloc_cache() aliases kmalloc_caches[] for
+			 * the same reason. Leaving the row empty instead
+			 * would send every accounted allocation out of the
+			 * set and into the general caches.
+			 */
+			if (mem_cgroup_kmem_disabled()) {
+				memcpy(b[btype], b[KMEM_BUCKET_NORMAL],
+				       sizeof(b[btype]));
+				continue;
+			}
+
+			type_flags = SLAB_ACCOUNT;
+			src = KMALLOC_CGROUP;
+		}
+
+		if (kmem_buckets_create_row(&b[btype], src, name, align,
+					    flags | type_flags, useroffset,
+					    usersize, ctor))
+			goto fail;
+	}
+
+	return b;
+
+fail:
+	kmem_buckets_destroy(b);
+
+	return NULL;
+}
+EXPORT_SYMBOL(kmem_buckets_create_types);
+
+/*
+ * Build one row of @b by mirroring the general caches of @type: a cache per
+ * kmalloc size, each named "@name-" followed by that cache's own suffix, so
+ * a row of KMALLOC_CGROUP ("kmalloc-cg-96") gets "@name-cg-96".
+ */
+static int kmem_buckets_create_row(kmem_buckets *b,
+				   enum kmalloc_cache_type type,
+				   const char *name, unsigned int align,
+				   slab_flags_t flags, unsigned int useroffset,
+				   unsigned int usersize, void (*ctor)(void *))
+{
+	unsigned int idx;
 
 	for (idx = 0; idx < ARRAY_SIZE(kmalloc_caches[KMALLOC_NORMAL]); idx++) {
 		char *short_size, *cache_name;
 		unsigned int cache_useroffset, cache_usersize;
 		unsigned int size, aligned_idx;
 
-		if (!kmalloc_caches[KMALLOC_NORMAL][idx])
+		if (!kmalloc_caches[type][idx])
 			continue;
 
-		size = kmalloc_caches[KMALLOC_NORMAL][idx]->object_size;
+		size = kmalloc_caches[type][idx]->object_size;
 		if (!size)
 			continue;
 
-		short_size = strchr(kmalloc_caches[KMALLOC_NORMAL][idx]->name, '-');
+		short_size = strchr(kmalloc_caches[type][idx]->name, '-');
 		if (WARN_ON(!short_size))
 			goto fail;
 
@@ -488,7 +567,7 @@ kmem_buckets *kmem_buckets_create(const char *name, unsigned int align,
 			if (WARN_ON(!cache_name))
 				goto fail;
 			(*b)[aligned_idx] = kmem_cache_create_usercopy(cache_name, size,
-					align ?: kmalloc_caches[KMALLOC_NORMAL][idx]->align,
+					align ?: kmalloc_caches[type][idx]->align,
 					flags, cache_useroffset,
 					cache_usersize, ctor);
 			kfree(cache_name);
@@ -499,14 +578,11 @@ kmem_buckets *kmem_buckets_create(const char *name, unsigned int align,
 			(*b)[idx] = (*b)[aligned_idx];
 	}
 
-	return b;
+	return 0;
 
 fail:
-	kmem_buckets_destroy(b);
-
-	return NULL;
+	return -ENOMEM;
 }
-EXPORT_SYMBOL(kmem_buckets_create);
 
 /**
  * kmem_buckets_destroy - Destroy a set of caches made by kmem_buckets_create()
@@ -521,28 +597,35 @@ EXPORT_SYMBOL(kmem_buckets_create);
  */
 void kmem_buckets_destroy(kmem_buckets *bucket)
 {
+	enum kmem_bucket_type btype, t;
 	unsigned int idx, i;
 
 	if (!IS_ENABLED(CONFIG_SLAB_BUCKETS) || ZERO_OR_NULL_PTR(bucket))
 		return;
 
-	for (idx = 0; idx < ARRAY_SIZE(kmalloc_caches[KMALLOC_NORMAL]); idx++) {
-		struct kmem_cache *cache = (*bucket)[idx];
+	for (btype = 0; btype < NR_KMEM_BUCKET_TYPES; btype++) {
+		for (idx = 0; idx < ARRAY_SIZE(bucket[btype]); idx++) {
+			struct kmem_cache *cache = bucket[btype][idx];
 
-		if (!cache)
-			continue;
+			if (!cache)
+				continue;
 
-		/*
-		 * Sizes below arch_slab_minalign() share one cache, which
-		 * kmem_buckets_create() then stores at each of their indices.
-		 * Drop every reference to it before destroying it, so that no
-		 * later pass reads a pointer to a cache that is already gone.
-		 */
-		for (i = idx; i < ARRAY_SIZE(kmalloc_caches[KMALLOC_NORMAL]); i++)
-			if ((*bucket)[i] == cache)
-				(*bucket)[i] = NULL;
+			/*
+			 * A cache is reachable from more than one entry: sizes
+			 * below arch_slab_minalign() share one, and a row that
+			 * kmem_buckets_create_types() aliased onto the normal
+			 * one under "cgroup.memory=nokmem" holds all of them a
+			 * second time. Drop every reference before destroying
+			 * it, so that no later pass reads a pointer to a cache
+			 * that is already gone.
+			 */
+			for (t = 0; t < NR_KMEM_BUCKET_TYPES; t++)
+				for (i = 0; i < ARRAY_SIZE(bucket[t]); i++)
+					if (bucket[t][i] == cache)
+						bucket[t][i] = NULL;
 
-		kmem_cache_destroy(cache);
+			kmem_cache_destroy(cache);
+		}
 	}
 
 	kmem_cache_free(kmem_buckets_cache, bucket);
@@ -1092,7 +1175,7 @@ void __init create_kmalloc_caches(void)
 
 	if (IS_ENABLED(CONFIG_SLAB_BUCKETS))
 		kmem_buckets_cache = kmem_cache_create("kmalloc_buckets",
-						       sizeof(kmem_buckets),
+						       sizeof(kmem_buckets) * NR_KMEM_BUCKET_TYPES,
 						       0, SLAB_NO_MERGE, NULL);
 }
 
-- 
2.34.1

[PATCH net-next v5 7/7] net: skb: isolate skb data area allocations into a separate bucket

[Kees Cook]

From: Pedro Falcato <pfalcato@suse.de>

SKB data area allocations (as done from alloc_skb()) use kmalloc().
These allocations can be variably sized and their contents can be more
or less controlled from userspace, which makes them useful for attackers
that want to overwrite a use-after-free'd object from the same kmalloc slab
(which often just requires the sizes to roughly match into the same kmalloc
bucket). [0] is an easy example of an exploit that uses netlink skb
allocation to target another similarly-sized accidentally freed object.

While other mitigations like CONFIG_RANDOM_KMALLOC_CACHES exist, these are
probabilistic. Use the existing kmem buckets API to further isolate these
allocations in a guaranteed fashion, when CONFIG_SLAB_BUCKETS=y.

Ask for the accounted kmalloc type as well as the normal one. AF_UNIX
sets sk_allocation to GFP_KERNEL_ACCOUNT, so without it every AF_UNIX
skb data area would fall back to the general caches, and those are the
ones most worth isolating. GFP_DMA is left to fall back, being passed to
an skb allocator only by rare devices.

Link: https://github.com/google/security-research/blob/master/pocs/linux/kernelctf/CVE-2023-4207_lts_cos_mitigation_2/docs/exploit.md [0]
Reviewed-by: Kees Cook <kees@kernel.org>
Signed-off-by: Pedro Falcato <pfalcato@suse.de>
Acked-by: Paolo Abeni <pabeni@redhat.com>
Signed-off-by: Kees Cook <kees@kernel.org>
---
 net/core/skbuff.c | 11 +++++++++--
 1 file changed, 9 insertions(+), 2 deletions(-)

diff --git a/net/core/skbuff.c b/net/core/skbuff.c
index 966af3beed94..e0660b2dbc19 100644
--- a/net/core/skbuff.c
+++ b/net/core/skbuff.c
@@ -586,6 +586,8 @@ struct sk_buff *napi_build_skb(void *data, unsigned int frag_size)
 }
 EXPORT_SYMBOL(napi_build_skb);
 
+static kmem_buckets *skb_data_buckets __ro_after_init;
+
 static void *kmalloc_pfmemalloc(size_t obj_size, gfp_t flags, int node)
 {
 	if (!gfp_pfmemalloc_allowed(flags))
@@ -593,7 +595,8 @@ static void *kmalloc_pfmemalloc(size_t obj_size, gfp_t flags, int node)
 	if (!obj_size)
 		return kmem_cache_alloc_node(net_hotdata.skb_small_head_cache,
 					     flags, node);
-	return kmalloc_node_track_caller(obj_size, flags, node);
+	return kmem_buckets_alloc_node_track_caller(skb_data_buckets, obj_size,
+						    flags, node);
 }
 
 /*
@@ -634,7 +637,7 @@ static void *kmalloc_reserve(unsigned int *size, gfp_t flags, int node,
 	 * Try a regular allocation, when that fails and we're not entitled
 	 * to the reserves, fail.
 	 */
-	obj = kmalloc_node_track_caller(obj_size,
+	obj = kmem_buckets_alloc_node_track_caller(skb_data_buckets, obj_size,
 					flags | __GFP_NOMEMALLOC | __GFP_NOWARN,
 					node);
 	if (likely(obj))
@@ -5235,6 +5238,10 @@ void __init skb_init(void)
 						0,
 						SKB_SMALL_HEAD_HEADROOM,
 						NULL);
+	skb_data_buckets = kmem_buckets_create_types("skb_data", 0, SLAB_PANIC,
+						     0, INT_MAX, NULL,
+						     BIT(KMEM_BUCKET_NORMAL) |
+						     BIT(KMEM_BUCKET_CGROUP));
 	skb_extensions_init();
 }
 
-- 
2.34.1