[PATCH v4 0/2] kasan: bugfix and cleanup

[PATCH v4 0/2] kasan: bugfix and cleanup

[Peter Collingbourne]

Hi,

This patch series reverts a cleanup patch that turned out to introduce
a bug, and does some cleanup of its own by removing some flags that I
realized were redundant while investigating the bug.

Rebased onto linux-next at akpm's request.

Peter

Peter Collingbourne (2):
  Revert "kasan: drop skip_kasan_poison variable in free_pages_prepare"
  kasan: remove PG_skip_kasan_poison flag

 include/linux/gfp_types.h      | 30 ++++++------
 include/linux/page-flags.h     |  9 ----
 include/trace/events/mmflags.h | 13 +-----
 mm/kasan/hw_tags.c             |  2 +-
 mm/page_alloc.c                | 84 +++++++++++++---------------------
 mm/vmalloc.c                   |  2 +-
 6 files changed, 49 insertions(+), 91 deletions(-)

-- 
2.40.0.rc1.284.g88254d51c5-goog

[PATCH v4 1/2] Revert "kasan: drop skip_kasan_poison variable in free_pages_prepare"

[Peter Collingbourne]

This reverts commit 487a32ec24be819e747af8c2ab0d5c515508086a.

The should_skip_kasan_poison() function reads the PG_skip_kasan_poison
flag from page->flags. However, this line of code in free_pages_prepare():

page->flags &= ~PAGE_FLAGS_CHECK_AT_PREP;

clears most of page->flags, including PG_skip_kasan_poison, before calling
should_skip_kasan_poison(), which meant that it would never return true
as a result of the page flag being set. Therefore, fix the code to call
should_skip_kasan_poison() before clearing the flags, as we were doing
before the reverted patch.

This fixes a measurable performance regression introduced in the
reverted commit, where munmap() takes longer than intended if HW
tags KASAN is supported and enabled at runtime. Without this patch,
we see a single-digit percentage performance regression in a particular
mmap()-heavy benchmark when enabling HW tags KASAN, and with the patch,
there is no statistically significant performance impact when enabling
HW tags KASAN.

Signed-off-by: Peter Collingbourne <pcc@google.com>
Fixes: 487a32ec24be ("kasan: drop skip_kasan_poison variable in free_pages_prepare")
Cc: <stable@vger.kernel.org> # 6.1
Link: https://linux-review.googlesource.com/id/Ic4f13affeebd20548758438bb9ed9ca40e312b79
Reviewed-by: Andrey Konovalov <andreyknvl@gmail.com>
---
 mm/page_alloc.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/mm/page_alloc.c b/mm/page_alloc.c
index 1c54790c2d17..c58ebf21ce63 100644
--- a/mm/page_alloc.c
+++ b/mm/page_alloc.c
@@ -1413,6 +1413,7 @@ static __always_inline bool free_pages_prepare(struct page *page,
 			unsigned int order, fpi_t fpi_flags)
 {
 	int bad = 0;
+	bool skip_kasan_poison = should_skip_kasan_poison(page, fpi_flags);
 	bool init = want_init_on_free();
 
 	VM_BUG_ON_PAGE(PageTail(page), page);
@@ -1489,7 +1490,7 @@ static __always_inline bool free_pages_prepare(struct page *page,
 	 * With hardware tag-based KASAN, memory tags must be set before the
 	 * page becomes unavailable via debug_pagealloc or arch_free_page.
 	 */
-	if (!should_skip_kasan_poison(page, fpi_flags)) {
+	if (!skip_kasan_poison) {
 		kasan_poison_pages(page, order, init);
 
 		/* Memory is already initialized if KASAN did it internally. */
-- 
2.40.0.rc1.284.g88254d51c5-goog

[PATCH v4 2/2] kasan: remove PG_skip_kasan_poison flag

[Peter Collingbourne]

Code inspection reveals that PG_skip_kasan_poison is redundant with
kasantag, because the former is intended to be set iff the latter is
the match-all tag. It can also be observed that it's basically pointless
to poison pages which have kasantag=0, because any pages with this tag
would have been pointed to by pointers with match-all tags, so poisoning
the pages would have little to no effect in terms of bug detection.
Therefore, change the condition in should_skip_kasan_poison() to check
kasantag instead, and remove PG_skip_kasan_poison and associated flags.

Signed-off-by: Peter Collingbourne <pcc@google.com>
Link: https://linux-review.googlesource.com/id/I57f825f2eaeaf7e8389d6cf4597c8a5821359838
Reviewed-by: Andrey Konovalov <andreyknvl@gmail.com>
---
v4:
- rebased to linux-next

v3:
- update comments

v2:
- also remove GFP_SKIP_KASAN_POISON and FPI_SKIP_KASAN_POISON
- rename GFP_SKIP_KASAN_UNPOISON to GFP_SKIP_KASAN
- update comments
- simplify control flow by removing reset_tags

 include/linux/gfp_types.h      | 30 ++++++-------
 include/linux/page-flags.h     |  9 ----
 include/trace/events/mmflags.h | 13 +-----
 mm/kasan/hw_tags.c             |  2 +-
 mm/page_alloc.c                | 81 +++++++++++++---------------------
 mm/vmalloc.c                   |  2 +-
 6 files changed, 47 insertions(+), 90 deletions(-)

diff --git a/include/linux/gfp_types.h b/include/linux/gfp_types.h
index 5088637fe5c2..6583a58670c5 100644
--- a/include/linux/gfp_types.h
+++ b/include/linux/gfp_types.h
@@ -47,16 +47,14 @@ typedef unsigned int __bitwise gfp_t;
 #define ___GFP_ACCOUNT		0x400000u
 #define ___GFP_ZEROTAGS		0x800000u
 #ifdef CONFIG_KASAN_HW_TAGS
-#define ___GFP_SKIP_ZERO		0x1000000u
-#define ___GFP_SKIP_KASAN_UNPOISON	0x2000000u
-#define ___GFP_SKIP_KASAN_POISON	0x4000000u
+#define ___GFP_SKIP_ZERO	0x1000000u
+#define ___GFP_SKIP_KASAN	0x2000000u
 #else
-#define ___GFP_SKIP_ZERO		0
-#define ___GFP_SKIP_KASAN_UNPOISON	0
-#define ___GFP_SKIP_KASAN_POISON	0
+#define ___GFP_SKIP_ZERO	0
+#define ___GFP_SKIP_KASAN	0
 #endif
 #ifdef CONFIG_LOCKDEP
-#define ___GFP_NOLOCKDEP	0x8000000u
+#define ___GFP_NOLOCKDEP	0x4000000u
 #else
 #define ___GFP_NOLOCKDEP	0
 #endif
@@ -234,25 +232,24 @@ typedef unsigned int __bitwise gfp_t;
  * memory tags at the same time as zeroing memory has minimal additional
  * performace impact.
  *
- * %__GFP_SKIP_KASAN_UNPOISON makes KASAN skip unpoisoning on page allocation.
- * Only effective in HW_TAGS mode.
- *
- * %__GFP_SKIP_KASAN_POISON makes KASAN skip poisoning on page deallocation.
- * Typically, used for userspace pages. Only effective in HW_TAGS mode.
+ * %__GFP_SKIP_KASAN makes KASAN skip unpoisoning on page allocation.
+ * Used for userspace and vmalloc pages; the latter are unpoisoned by
+ * kasan_unpoison_vmalloc instead. For userspace pages, results in
+ * poisoning being skipped as well, see should_skip_kasan_poison for
+ * details. Only effective in HW_TAGS mode.
  */
 #define __GFP_NOWARN	((__force gfp_t)___GFP_NOWARN)
 #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_ZERO ((__force gfp_t)___GFP_SKIP_ZERO)
-#define __GFP_SKIP_KASAN_UNPOISON ((__force gfp_t)___GFP_SKIP_KASAN_UNPOISON)
-#define __GFP_SKIP_KASAN_POISON   ((__force gfp_t)___GFP_SKIP_KASAN_POISON)
+#define __GFP_SKIP_KASAN ((__force gfp_t)___GFP_SKIP_KASAN)
 
 /* Disable lockdep for GFP context tracking */
 #define __GFP_NOLOCKDEP ((__force gfp_t)___GFP_NOLOCKDEP)
 
 /* Room for N __GFP_FOO bits */
-#define __GFP_BITS_SHIFT (27 + 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))
 
 /**
@@ -335,8 +332,7 @@ typedef unsigned int __bitwise gfp_t;
 #define GFP_DMA		__GFP_DMA
 #define GFP_DMA32	__GFP_DMA32
 #define GFP_HIGHUSER	(GFP_USER | __GFP_HIGHMEM)
-#define GFP_HIGHUSER_MOVABLE	(GFP_HIGHUSER | __GFP_MOVABLE | \
-			 __GFP_SKIP_KASAN_POISON | __GFP_SKIP_KASAN_UNPOISON)
+#define GFP_HIGHUSER_MOVABLE	(GFP_HIGHUSER | __GFP_MOVABLE | __GFP_SKIP_KASAN)
 #define GFP_TRANSHUGE_LIGHT	((GFP_HIGHUSER_MOVABLE | __GFP_COMP | \
 			 __GFP_NOMEMALLOC | __GFP_NOWARN) & ~__GFP_RECLAIM)
 #define GFP_TRANSHUGE	(GFP_TRANSHUGE_LIGHT | __GFP_DIRECT_RECLAIM)
diff --git a/include/linux/page-flags.h b/include/linux/page-flags.h
index 57287102c5bd..dcda20c47b8f 100644
--- a/include/linux/page-flags.h
+++ b/include/linux/page-flags.h
@@ -135,9 +135,6 @@ enum pageflags {
 #ifdef CONFIG_ARCH_USES_PG_ARCH_X
 	PG_arch_2,
 	PG_arch_3,
-#endif
-#ifdef CONFIG_KASAN_HW_TAGS
-	PG_skip_kasan_poison,
 #endif
 	__NR_PAGEFLAGS,
 
@@ -594,12 +591,6 @@ TESTCLEARFLAG(Young, young, PF_ANY)
 PAGEFLAG(Idle, idle, PF_ANY)
 #endif
 
-#ifdef CONFIG_KASAN_HW_TAGS
-PAGEFLAG(SkipKASanPoison, skip_kasan_poison, PF_HEAD)
-#else
-PAGEFLAG_FALSE(SkipKASanPoison, skip_kasan_poison)
-#endif
-
 /*
  * PageReported() is used to track reported free pages within the Buddy
  * allocator. We can use the non-atomic version of the test and set
diff --git a/include/trace/events/mmflags.h b/include/trace/events/mmflags.h
index b28218b7998e..b63e7c0fbbe5 100644
--- a/include/trace/events/mmflags.h
+++ b/include/trace/events/mmflags.h
@@ -55,8 +55,7 @@
 #ifdef CONFIG_KASAN_HW_TAGS
 #define __def_gfpflag_names_kasan ,			\
 	gfpflag_string(__GFP_SKIP_ZERO),		\
-	gfpflag_string(__GFP_SKIP_KASAN_POISON),	\
-	gfpflag_string(__GFP_SKIP_KASAN_UNPOISON)
+	gfpflag_string(__GFP_SKIP_KASAN)
 #else
 #define __def_gfpflag_names_kasan
 #endif
@@ -96,13 +95,6 @@
 #define IF_HAVE_PG_ARCH_X(_name)
 #endif
 
-#ifdef CONFIG_KASAN_HW_TAGS
-#define IF_HAVE_PG_SKIP_KASAN_POISON(_name) \
-	,{1UL << PG_##_name, __stringify(_name)}
-#else
-#define IF_HAVE_PG_SKIP_KASAN_POISON(_name)
-#endif
-
 #define DEF_PAGEFLAG_NAME(_name) { 1UL <<  PG_##_name, __stringify(_name) }
 
 #define __def_pageflag_names						\
@@ -133,8 +125,7 @@ IF_HAVE_PG_HWPOISON(hwpoison)						\
 IF_HAVE_PG_IDLE(idle)							\
 IF_HAVE_PG_IDLE(young)							\
 IF_HAVE_PG_ARCH_X(arch_2)						\
-IF_HAVE_PG_ARCH_X(arch_3)						\
-IF_HAVE_PG_SKIP_KASAN_POISON(skip_kasan_poison)
+IF_HAVE_PG_ARCH_X(arch_3)
 
 #define show_page_flags(flags)						\
 	(flags) ? __print_flags(flags, "|",				\
diff --git a/mm/kasan/hw_tags.c b/mm/kasan/hw_tags.c
index d1bcb0205327..bb4f56e5bdec 100644
--- a/mm/kasan/hw_tags.c
+++ b/mm/kasan/hw_tags.c
@@ -318,7 +318,7 @@ void *__kasan_unpoison_vmalloc(const void *start, unsigned long size,
 	 * Thus, for VM_ALLOC mappings, hardware tag-based KASAN only tags
 	 * the first virtual mapping, which is created by vmalloc().
 	 * Tagging the page_alloc memory backing that vmalloc() allocation is
-	 * skipped, see ___GFP_SKIP_KASAN_UNPOISON.
+	 * skipped, see ___GFP_SKIP_KASAN.
 	 *
 	 * For non-VM_ALLOC allocations, page_alloc memory is tagged as usual.
 	 */
diff --git a/mm/page_alloc.c b/mm/page_alloc.c
index c58ebf21ce63..680a4d76460e 100644
--- a/mm/page_alloc.c
+++ b/mm/page_alloc.c
@@ -112,17 +112,6 @@ typedef int __bitwise fpi_t;
  */
 #define FPI_TO_TAIL		((__force fpi_t)BIT(1))
 
-/*
- * Don't poison memory with KASAN (only for the tag-based modes).
- * During boot, all non-reserved memblock memory is exposed to page_alloc.
- * Poisoning all that memory lengthens boot time, especially on systems with
- * large amount of RAM. This flag is used to skip that poisoning.
- * This is only done for the tag-based KASAN modes, as those are able to
- * detect memory corruptions with the memory tags assigned by default.
- * All memory allocated normally after boot gets poisoned as usual.
- */
-#define FPI_SKIP_KASAN_POISON	((__force fpi_t)BIT(2))
-
 /* prevent >1 _updater_ of zone percpu pageset ->high and ->batch fields */
 static DEFINE_MUTEX(pcp_batch_high_lock);
 #define MIN_PERCPU_PAGELIST_HIGH_FRACTION (8)
@@ -1370,13 +1359,19 @@ static int free_tail_pages_check(struct page *head_page, struct page *page)
 /*
  * Skip KASAN memory poisoning when either:
  *
- * 1. Deferred memory initialization has not yet completed,
- *    see the explanation below.
- * 2. Skipping poisoning is requested via FPI_SKIP_KASAN_POISON,
- *    see the comment next to it.
- * 3. Skipping poisoning is requested via __GFP_SKIP_KASAN_POISON,
- *    see the comment next to it.
- * 4. The allocation is excluded from being checked due to sampling,
+ * 1. For generic KASAN: deferred memory initialization has not yet completed.
+ *    Tag-based KASAN modes skip pages freed via deferred memory initialization
+ *    using page tags instead (see below).
+ * 2. For tag-based KASAN modes: the page has a match-all KASAN tag, indicating
+ *    that error detection is disabled for accesses via the page address.
+ *
+ * Pages will have match-all tags in the following circumstances:
+ *
+ * 1. Pages are being initialized for the first time, including during deferred
+ *    memory init; see the call to page_kasan_tag_reset in __init_single_page.
+ * 2. The allocation was not unpoisoned due to __GFP_SKIP_KASAN, with the
+ *    exception of pages unpoisoned by kasan_unpoison_vmalloc.
+ * 3. The allocation was excluded from being checked due to sampling,
  *    see the call to kasan_unpoison_pages.
  *
  * Poisoning pages during deferred memory init will greatly lengthen the
@@ -1392,10 +1387,10 @@ static int free_tail_pages_check(struct page *head_page, struct page *page)
  */
 static inline bool should_skip_kasan_poison(struct page *page, fpi_t fpi_flags)
 {
-	return deferred_pages_enabled() ||
-	       (!IS_ENABLED(CONFIG_KASAN_GENERIC) &&
-		(fpi_flags & FPI_SKIP_KASAN_POISON)) ||
-	       PageSkipKASanPoison(page);
+	if (IS_ENABLED(CONFIG_KASAN_GENERIC))
+		return deferred_pages_enabled();
+
+	return page_kasan_tag(page) == 0xff;
 }
 
 static void kernel_init_pages(struct page *page, int numpages)
@@ -1730,7 +1725,7 @@ void __free_pages_core(struct page *page, unsigned int order)
 	 * Bypass PCP and place fresh pages right to the tail, primarily
 	 * relevant for memory onlining.
 	 */
-	__free_pages_ok(page, order, FPI_TO_TAIL | FPI_SKIP_KASAN_POISON);
+	__free_pages_ok(page, order, FPI_TO_TAIL);
 }
 
 #ifdef CONFIG_NUMA
@@ -2396,9 +2391,9 @@ static inline bool should_skip_kasan_unpoison(gfp_t flags)
 
 	/*
 	 * With hardware tag-based KASAN enabled, skip if this has been
-	 * requested via __GFP_SKIP_KASAN_UNPOISON.
+	 * requested via __GFP_SKIP_KASAN.
 	 */
-	return flags & __GFP_SKIP_KASAN_UNPOISON;
+	return flags & __GFP_SKIP_KASAN;
 }
 
 static inline bool should_skip_init(gfp_t flags)
@@ -2417,7 +2412,6 @@ inline void post_alloc_hook(struct page *page, unsigned int order,
 	bool init = !want_init_on_free() && want_init_on_alloc(gfp_flags) &&
 			!should_skip_init(gfp_flags);
 	bool zero_tags = init && (gfp_flags & __GFP_ZEROTAGS);
-	bool reset_tags = true;
 	int i;
 
 	set_page_private(page, 0);
@@ -2451,37 +2445,22 @@ inline void post_alloc_hook(struct page *page, unsigned int order,
 		/* Take note that memory was initialized by the loop above. */
 		init = false;
 	}
-	if (!should_skip_kasan_unpoison(gfp_flags)) {
-		/* Try unpoisoning (or setting tags) and initializing memory. */
-		if (kasan_unpoison_pages(page, order, init)) {
-			/* Take note that memory was initialized by KASAN. */
-			if (kasan_has_integrated_init())
-				init = false;
-			/* Take note that memory tags were set by KASAN. */
-			reset_tags = false;
-		} else {
-			/*
-			 * KASAN decided to exclude this allocation from being
-			 * (un)poisoned due to sampling. Make KASAN skip
-			 * poisoning when the allocation is freed.
-			 */
-			SetPageSkipKASanPoison(page);
-		}
-	}
-	/*
-	 * If memory tags have not been set by KASAN, reset the page tags to
-	 * ensure page_address() dereferencing does not fault.
-	 */
-	if (reset_tags) {
+	if (!should_skip_kasan_unpoison(gfp_flags) &&
+	    kasan_unpoison_pages(page, order, init)) {
+		/* Take note that memory was initialized by KASAN. */
+		if (kasan_has_integrated_init())
+			init = false;
+	} else {
+		/*
+		 * If memory tags have not been set by KASAN, reset the page
+		 * tags to ensure page_address() dereferencing does not fault.
+		 */
 		for (i = 0; i != 1 << order; ++i)
 			page_kasan_tag_reset(page + i);
 	}
 	/* If memory is still not initialized, initialize it now. */
 	if (init)
 		kernel_init_pages(page, 1 << order);
-	/* Propagate __GFP_SKIP_KASAN_POISON to page flags. */
-	if (kasan_hw_tags_enabled() && (gfp_flags & __GFP_SKIP_KASAN_POISON))
-		SetPageSkipKASanPoison(page);
 
 	set_page_owner(page, order, gfp_flags);
 	page_table_check_alloc(page, order);
diff --git a/mm/vmalloc.c b/mm/vmalloc.c
index bef6cf2b4d46..5e60e9792cbf 100644
--- a/mm/vmalloc.c
+++ b/mm/vmalloc.c
@@ -3188,7 +3188,7 @@ void *__vmalloc_node_range(unsigned long size, unsigned long align,
 			 * pages backing VM_ALLOC mapping. Memory is instead
 			 * poisoned and zeroed by kasan_unpoison_vmalloc().
 			 */
-			gfp_mask |= __GFP_SKIP_KASAN_UNPOISON | __GFP_SKIP_ZERO;
+			gfp_mask |= __GFP_SKIP_KASAN | __GFP_SKIP_ZERO;
 		}
 
 		/* Take note that the mapping is PAGE_KERNEL. */
-- 
2.40.0.rc1.284.g88254d51c5-goog