From: andrey.konovalov@linux.dev
To: Marco Elver <elver@google.com>,
Alexander Potapenko <glider@google.com>,
Vincenzo Frascino <vincenzo.frascino@arm.com>,
Catalin Marinas <catalin.marinas@arm.com>,
Peter Collingbourne <pcc@google.com>
Cc: Andrey Konovalov <andreyknvl@gmail.com>,
Dmitry Vyukov <dvyukov@google.com>,
Andrey Ryabinin <aryabinin@virtuozzo.com>,
kasan-dev@googlegroups.com,
Andrew Morton <akpm@linux-foundation.org>,
linux-mm@kvack.org, Will Deacon <will@kernel.org>,
linux-arm-kernel@lists.infradead.org,
Evgenii Stepanov <eugenis@google.com>,
linux-kernel@vger.kernel.org,
Andrey Konovalov <andreyknvl@google.com>
Subject: [PATCH 26/31] kasan, page_alloc: allow skipping unpoisoning for HW_TAGS
Date: Tue, 30 Nov 2021 23:08:08 +0100 [thread overview]
Message-ID: <e60cbad6f4f8ee08137671c008c83ab26255e9bf.1638308023.git.andreyknvl@google.com> (raw)
In-Reply-To: <cover.1638308023.git.andreyknvl@google.com>
From: Andrey Konovalov <andreyknvl@google.com>
This patch add a new GFP flag __GFP_SKIP_KASAN_UNPOISON that allows
skipping KASAN poisoning for page_alloc allocations. The flag is only
effective with HW_TAGS KASAN.
This flag will be used by vmalloc code for page_alloc allocations
backing vmalloc() mappings in the following patch. The reason to skip
KASAN poisoning for these pages in page_alloc is because vmalloc code
will be poisoning them instead.
Signed-off-by: Andrey Konovalov <andreyknvl@google.com>
---
include/linux/gfp.h | 13 +++++++++----
mm/page_alloc.c | 24 +++++++++++++++++-------
2 files changed, 26 insertions(+), 11 deletions(-)
diff --git a/include/linux/gfp.h b/include/linux/gfp.h
index dddd7597689f..a4c8ff3fbed1 100644
--- a/include/linux/gfp.h
+++ b/include/linux/gfp.h
@@ -54,9 +54,10 @@ struct vm_area_struct;
#define ___GFP_THISNODE 0x200000u
#define ___GFP_ACCOUNT 0x400000u
#define ___GFP_ZEROTAGS 0x800000u
-#define ___GFP_SKIP_KASAN_POISON 0x1000000u
+#define ___GFP_SKIP_KASAN_UNPOISON 0x1000000u
+#define ___GFP_SKIP_KASAN_POISON 0x2000000u
#ifdef CONFIG_LOCKDEP
-#define ___GFP_NOLOCKDEP 0x2000000u
+#define ___GFP_NOLOCKDEP 0x4000000u
#else
#define ___GFP_NOLOCKDEP 0
#endif
@@ -235,6 +236,9 @@ struct vm_area_struct;
* %__GFP_ZEROTAGS zeroes memory tags at allocation time if the memory itself
* is being zeroed (either via __GFP_ZERO or via init_on_alloc).
*
+ * %__GFP_SKIP_KASAN_UNPOISON skips KASAN unpoisoning on page allocation.
+ * Currently only has an effect in HW tags mode.
+ *
* %__GFP_SKIP_KASAN_POISON returns a page which does not need to be poisoned
* on deallocation. Typically used for userspace pages. Currently only has an
* effect in HW tags mode.
@@ -243,13 +247,14 @@ struct vm_area_struct;
#define __GFP_COMP ((__force gfp_t)___GFP_COMP)
#define __GFP_ZERO ((__force gfp_t)___GFP_ZERO)
#define __GFP_ZEROTAGS ((__force gfp_t)___GFP_ZEROTAGS)
-#define __GFP_SKIP_KASAN_POISON ((__force gfp_t)___GFP_SKIP_KASAN_POISON)
+#define __GFP_SKIP_KASAN_UNPOISON ((__force gfp_t)___GFP_SKIP_KASAN_UNPOISON)
+#define __GFP_SKIP_KASAN_POISON ((__force gfp_t)___GFP_SKIP_KASAN_POISON)
/* Disable lockdep for GFP context tracking */
#define __GFP_NOLOCKDEP ((__force gfp_t)___GFP_NOLOCKDEP)
/* Room for N __GFP_FOO bits */
-#define __GFP_BITS_SHIFT (25 + IS_ENABLED(CONFIG_LOCKDEP))
+#define __GFP_BITS_SHIFT (26 + IS_ENABLED(CONFIG_LOCKDEP))
#define __GFP_BITS_MASK ((__force gfp_t)((1 << __GFP_BITS_SHIFT) - 1))
/**
diff --git a/mm/page_alloc.c b/mm/page_alloc.c
index 4eb341351124..3afebc037fcd 100644
--- a/mm/page_alloc.c
+++ b/mm/page_alloc.c
@@ -2381,6 +2381,21 @@ static bool check_new_pages(struct page *page, unsigned int order)
return false;
}
+static inline bool should_skip_kasan_unpoison(gfp_t flags, bool init_tags)
+{
+ /* Don't skip if a software KASAN mode is enabled. */
+ if (!IS_ENABLED(CONFIG_KASAN_HW_TAGS))
+ return false;
+
+ /*
+ * For hardware tag-based KASAN, skip if either:
+ *
+ * 1. Memory tags have already been cleared via tag_clear_highpage().
+ * 2. Skipping has been requested via __GFP_SKIP_KASAN_UNPOISON.
+ */
+ return init_tags || (flags & __GFP_SKIP_KASAN_UNPOISON);
+}
+
inline void post_alloc_hook(struct page *page, unsigned int order,
gfp_t gfp_flags)
{
@@ -2420,13 +2435,8 @@ inline void post_alloc_hook(struct page *page, unsigned int order,
/* Note that memory is already initialized by the loop above. */
init = false;
}
- /*
- * If either a software KASAN mode is enabled, or,
- * in the case of hardware tag-based KASAN,
- * if memory tags have not been cleared via tag_clear_highpage().
- */
- if (!IS_ENABLED(CONFIG_KASAN_HW_TAGS) || !init_tags) {
- /* Mark shadow memory or set memory tags. */
+ if (!should_skip_kasan_unpoison(gfp_flags, init_tags)) {
+ /* Unpoison shadow memory or set memory tags. */
kasan_unpoison_pages(page, order, init);
/* Note that memory is already initialized by KASAN. */
--
2.25.1
next prev parent reply other threads:[~2021-11-30 22:13 UTC|newest]
Thread overview: 56+ messages / expand[flat|nested] mbox.gz Atom feed top
2021-11-30 21:39 [PATCH 00/31] kasan, vmalloc, arm64: add vmalloc tagging support for SW/HW_TAGS andrey.konovalov
2021-11-30 21:39 ` [PATCH 01/31] kasan, page_alloc: deduplicate should_skip_kasan_poison andrey.konovalov
2021-11-30 21:39 ` [PATCH 02/31] kasan, page_alloc: move tag_clear_highpage out of kernel_init_free_pages andrey.konovalov
2021-12-02 15:24 ` Alexander Potapenko
2021-11-30 21:39 ` [PATCH 03/31] kasan, page_alloc: merge kasan_free_pages into free_pages_prepare andrey.konovalov
2021-12-02 15:32 ` Alexander Potapenko
2021-11-30 21:39 ` [PATCH 04/31] kasan, page_alloc: simplify kasan_poison_pages call site andrey.konovalov
2021-12-01 14:09 ` Marco Elver
2021-12-06 21:07 ` Andrey Konovalov
2021-11-30 21:39 ` [PATCH 05/31] kasan, page_alloc: init memory of skipped pages on free andrey.konovalov
2021-11-30 21:41 ` [PATCH 06/31] mm: clarify __GFP_ZEROTAGS comment andrey.konovalov
2021-11-30 21:41 ` [PATCH 07/31] kasan: only apply __GFP_ZEROTAGS when memory is zeroed andrey.konovalov
2021-12-02 15:40 ` Alexander Potapenko
2021-11-30 21:41 ` [PATCH 08/31] kasan, page_alloc: refactor init checks in post_alloc_hook andrey.konovalov
2021-12-02 16:13 ` Alexander Potapenko
2021-12-06 21:09 ` Andrey Konovalov
2021-12-16 10:59 ` Alexander Potapenko
2021-11-30 21:42 ` [PATCH 09/31] kasan, page_alloc: merge kasan_alloc_pages into post_alloc_hook andrey.konovalov
2021-11-30 21:52 ` [PATCH 10/31] kasan, page_alloc: combine tag_clear_highpage calls in post_alloc_hook andrey.konovalov
2021-11-30 22:05 ` [PATCH 11/31] kasan, page_alloc: move SetPageSkipKASanPoison " andrey.konovalov
2021-11-30 22:05 ` [PATCH 12/31] kasan, page_alloc: move kernel_init_free_pages " andrey.konovalov
2021-11-30 22:05 ` [PATCH 13/31] kasan, page_alloc: simplify kasan_unpoison_pages call site andrey.konovalov
2021-11-30 22:06 ` [PATCH 14/31] kasan: clean up metadata byte definitions andrey.konovalov
2021-11-30 22:06 ` [PATCH 15/31] kasan: define KASAN_VMALLOC_INVALID for SW_TAGS andrey.konovalov
2021-11-30 22:06 ` [PATCH 16/31] kasan, x86, arm64, s390: rename functions for modules shadow andrey.konovalov
2021-11-30 22:06 ` [PATCH 17/31] kasan, vmalloc: drop outdated VM_KASAN comment andrey.konovalov
2021-11-30 22:07 ` [PATCH 18/31] kasan: reorder vmalloc hooks andrey.konovalov
2021-11-30 22:07 ` [PATCH 19/31] kasan: add wrappers for " andrey.konovalov
2021-11-30 22:07 ` [PATCH 20/31] kasan, vmalloc: reset tags in vmalloc functions andrey.konovalov
2021-12-02 14:17 ` Marco Elver
2021-12-06 21:08 ` Andrey Konovalov
2021-11-30 22:07 ` [PATCH 21/31] kasan, fork: don't tag stacks allocated with vmalloc andrey.konovalov
2021-12-02 14:27 ` Marco Elver
2021-12-06 21:08 ` Andrey Konovalov
2021-11-30 22:07 ` [PATCH 22/31] kasan, vmalloc: add vmalloc support to SW_TAGS andrey.konovalov
2021-11-30 22:07 ` [PATCH 23/31] kasan, arm64: allow KASAN_VMALLOC with SW_TAGS andrey.konovalov
2021-12-03 12:37 ` Marco Elver
2021-12-06 21:10 ` Andrey Konovalov
2021-11-30 22:07 ` [PATCH 24/31] kasan, vmalloc, arm64: mark vmalloc mappings as pgprot_tagged andrey.konovalov
2021-12-03 12:42 ` Marco Elver
2021-12-06 21:12 ` Andrey Konovalov
2021-11-30 22:08 ` [PATCH 25/31] kasan, vmalloc: don't unpoison VM_ALLOC pages before mapping andrey.konovalov
2021-11-30 22:08 ` andrey.konovalov [this message]
2021-11-30 22:08 ` [PATCH 27/31] kasan, vmalloc: add vmalloc support to HW_TAGS andrey.konovalov
2021-12-03 12:41 ` Marco Elver
2021-12-06 21:12 ` Andrey Konovalov
2021-11-30 22:08 ` [PATCH 28/31] kasan: add kasan.vmalloc command line flag andrey.konovalov
2021-12-03 12:09 ` Marco Elver
2021-12-06 21:09 ` Andrey Konovalov
2021-11-30 22:08 ` [PATCH 29/31] kasan, arm64: allow KASAN_VMALLOC with HW_TAGS andrey.konovalov
2021-12-01 11:35 ` Marco Elver
2021-12-06 21:10 ` Andrey Konovalov
2021-12-03 12:40 ` Marco Elver
2021-12-06 21:10 ` Andrey Konovalov
2021-11-30 22:08 ` [PATCH 30/31] kasan: documentation updates andrey.konovalov
2021-11-30 22:08 ` [PATCH 31/31] kasan: improve vmalloc tests andrey.konovalov
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=e60cbad6f4f8ee08137671c008c83ab26255e9bf.1638308023.git.andreyknvl@google.com \
--to=andrey.konovalov@linux.dev \
--cc=akpm@linux-foundation.org \
--cc=andreyknvl@gmail.com \
--cc=andreyknvl@google.com \
--cc=aryabinin@virtuozzo.com \
--cc=catalin.marinas@arm.com \
--cc=dvyukov@google.com \
--cc=elver@google.com \
--cc=eugenis@google.com \
--cc=glider@google.com \
--cc=kasan-dev@googlegroups.com \
--cc=linux-arm-kernel@lists.infradead.org \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-mm@kvack.org \
--cc=pcc@google.com \
--cc=vincenzo.frascino@arm.com \
--cc=will@kernel.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox