The patch titled
Subject: kasan, mm: optimize krealloc poisoning
has been added to the -mm tree. Its filename is
kasan-mm-optimize-krealloc-poisoning.patch
This patch should soon appear at
https://ozlabs.org/~akpm/mmots/broken-out/kasan-mm-optimize-krealloc-poisoning.patch
and later at
https://ozlabs.org/~akpm/mmotm/broken-out/kasan-mm-optimize-krealloc-poisoning.patch
Before you just go and hit "reply", please:
a) Consider who else should be cc'ed
b) Prefer to cc a suitable mailing list as well
c) Ideally: find the original patch on the mailing list and do a
reply-to-all to that, adding suitable additional cc's
*** Remember to use Documentation/process/submit-checklist.rst when testing your code ***
The -mm tree is included into linux-next and is updated
there every 3-4 working days
------------------------------------------------------
From: Andrey Konovalov <andreyknvl@xxxxxxxxxx>
Subject: kasan, mm: optimize krealloc poisoning
Currently, krealloc() always calls ksize(), which unpoisons the whole
object including the redzone. This is inefficient, as kasan_krealloc()
repoisons the redzone for objects that fit into the same buffer.
This patch changes krealloc() instrumentation to use uninstrumented
__ksize() that doesn't unpoison the memory. Instead, kasan_kreallos() is
changed to unpoison the memory excluding the redzone.
For objects that don't fit into the old allocation, this patch disables
KASAN accessibility checks when copying memory into a new object instead
of unpoisoning it.
Link: https://lkml.kernel.org/r/9bef90327c9cb109d736c40115684fd32f49e6b0.1612546384.git.andreyknvl@xxxxxxxxxx
Signed-off-by: Andrey Konovalov <andreyknvl@xxxxxxxxxx>
Cc: Alexander Potapenko <glider@xxxxxxxxxx>
Cc: Andrey Ryabinin <aryabinin@xxxxxxxxxxxxx>
Cc: Branislav Rankov <Branislav.Rankov@xxxxxxx>
Cc: Catalin Marinas <catalin.marinas@xxxxxxx>
Cc: Dmitry Vyukov <dvyukov@xxxxxxxxxx>
Cc: Evgenii Stepanov <eugenis@xxxxxxxxxx>
Cc: Kevin Brodsky <kevin.brodsky@xxxxxxx>
Cc: Marco Elver <elver@xxxxxxxxxx>
Cc: Peter Collingbourne <pcc@xxxxxxxxxx>
Cc: Vincenzo Frascino <vincenzo.frascino@xxxxxxx>
Cc: Will Deacon <will.deacon@xxxxxxx>
Signed-off-by: Andrew Morton <akpm@xxxxxxxxxxxxxxxxxxxx>
---
mm/kasan/common.c | 12 ++++++++++--
mm/slab_common.c | 20 ++++++++++++++------
2 files changed, 24 insertions(+), 8 deletions(-)
--- a/mm/kasan/common.c~kasan-mm-optimize-krealloc-poisoning
+++ a/mm/kasan/common.c
@@ -476,7 +476,7 @@ static void *____kasan_kmalloc(struct km
/*
* The object has already been unpoisoned by kasan_slab_alloc() for
- * kmalloc() or by ksize() for krealloc().
+ * kmalloc() or by kasan_krealloc() for krealloc().
*/
/*
@@ -526,7 +526,7 @@ void * __must_check __kasan_kmalloc_larg
/*
* The object has already been unpoisoned by kasan_alloc_pages() for
- * alloc_pages() or by ksize() for krealloc().
+ * alloc_pages() or by kasan_krealloc() for krealloc().
*/
/*
@@ -554,8 +554,16 @@ void * __must_check __kasan_krealloc(con
if (unlikely(object == ZERO_SIZE_PTR))
return (void *)object;
+ /*
+ * Unpoison the object's data.
+ * Part of it might already have been unpoisoned, but it's unknown
+ * how big that part is.
+ */
+ kasan_unpoison(object, size);
+
page = virt_to_head_page(object);
+ /* Piggy-back on kmalloc() instrumentation to poison the redzone. */
if (unlikely(!PageSlab(page)))
return __kasan_kmalloc_large(object, size, flags);
else
--- a/mm/slab_common.c~kasan-mm-optimize-krealloc-poisoning
+++ a/mm/slab_common.c
@@ -1061,19 +1061,27 @@ static __always_inline void *__do_kreall
void *ret;
size_t ks;
- if (likely(!ZERO_OR_NULL_PTR(p)) && !kasan_check_byte(p))
- return NULL;
-
- ks = ksize(p);
+ /* Don't use instrumented ksize to allow precise KASAN poisoning. */
+ if (likely(!ZERO_OR_NULL_PTR(p))) {
+ if (!kasan_check_byte(p))
+ return NULL;
+ ks = kfence_ksize(p) ?: __ksize(p);
+ } else
+ ks = 0;
+ /* If the object still fits, repoison it precisely. */
if (ks >= new_size) {
p = kasan_krealloc((void *)p, new_size, flags);
return (void *)p;
}
ret = kmalloc_track_caller(new_size, flags);
- if (ret && p)
- memcpy(ret, p, ks);
+ if (ret && p) {
+ /* Disable KASAN checks as the object's redzone is accessed. */
+ kasan_disable_current();
+ memcpy(ret, kasan_reset_tag(p), ks);
+ kasan_enable_current();
+ }
return ret;
}
_
Patches currently in -mm which might be from andreyknvl@xxxxxxxxxx are
kasan-prefix-global-functions-with-kasan_.patch
kasan-clarify-hw_tags-impact-on-tbi.patch
kasan-clean-up-comments-in-tests.patch
kasan-add-macros-to-simplify-checking-test-constraints.patch
kasan-add-match-all-tag-tests.patch
kasan-arm64-allow-using-kunit-tests-with-hw_tags-mode.patch
kasan-rename-config_test_kasan_module.patch
kasan-add-compiler-barriers-to-kunit_expect_kasan_fail.patch
kasan-adapt-kmalloc_uaf2-test-to-hw_tags-mode.patch
kasan-fix-memory-corruption-in-kasan_bitops_tags-test.patch
kasan-move-_ret_ip_-to-inline-wrappers.patch
kasan-fix-bug-detection-via-ksize-for-hw_tags-mode.patch
kasan-add-proper-page-allocator-tests.patch
kasan-add-a-test-for-kmem_cache_alloc-free_bulk.patch
kasan-dont-run-tests-when-kasan-is-not-enabled.patch
kfence-kasan-make-kfence-compatible-with-kasan-fix-2.patch
kasan-mm-dont-save-alloc-stacks-twice.patch
kasan-mm-optimize-kmalloc-poisoning.patch
kasan-optimize-large-kmalloc-poisoning.patch
kasan-clean-up-setting-free-info-in-kasan_slab_free.patch
kasan-unify-large-kfree-checks.patch
kasan-rework-krealloc-tests.patch
kasan-mm-fail-krealloc-on-freed-objects.patch
kasan-mm-optimize-krealloc-poisoning.patch
kasan-ensure-poisoning-size-alignment.patch
arm64-kasan-simplify-and-inline-mte-functions.patch
kasan-inline-hw_tags-helper-functions.patch
arm64-kasan-export-mte-symbols-for-kasan-tests.patch
kasan-clarify-that-only-first-bug-is-reported-in-hw_tags.patch
