[PATCH v1 0/2] don't use vmemmap_populate() to initialize shadow

[PATCH v1 0/2] don't use vmemmap_populate() to initialize shadow

[Pavel Tatashin]

Andrey Ryabinin asked to replace the three patches in my series:

x86-kasan-add-and-use-kasan_map_populate.patch
arm64-kasan-add-and-use-kasan_map_populate.patch
arm64-kasan-avoid-using-vmemmap_populate-to-initialise-shadow.patch

With two patches in this thread:

x86/mm/kasan: don't use vmemmap_populate() to initialize shadow
arm64/mm/kasan: don't use vmemmap_populate() to initialize shadow

Pavel Tatashin (2):
  x86/mm/kasan: don't use vmemmap_populate() to initialize shadow
  arm64/mm/kasan: don't use vmemmap_populate() to initialize shadow

 arch/arm64/Kconfig          |   2 +-
 arch/arm64/mm/kasan_init.c  | 130 ++++++++++++++++++++++++----------------
 arch/x86/Kconfig            |   2 +-
 arch/x86/mm/kasan_init_64.c | 143 +++++++++++++++++++++++++++++++++++++++++---
 4 files changed, 218 insertions(+), 59 deletions(-)

-- 
2.15.0

[PATCH v1 1/2] x86/mm/kasan: don't use vmemmap_populate() to initialize shadow

[Pavel Tatashin]

The kasan shadow is currently mapped using vmemmap_populate() since that
provides a semi-convenient way to map pages into init_top_pgt. However,
since that no longer zeroes the mapped pages, it is not suitable for kasan,
which requires zeroed shadow memory.

Add kasan_populate_shadow() interface and use it instead of
vmemmap_populate(). Besides, this allows us to take advantage of gigantic
pages and use them to populate the shadow, which should save us some memory
wasted on page tables and reduce TLB pressure.

Signed-off-by: Andrey Ryabinin <aryabinin@virtuozzo.com>
Signed-off-by: Pavel Tatashin <pasha.tatashin@oracle.com>
---
 arch/x86/Kconfig            |   2 +-
 arch/x86/mm/kasan_init_64.c | 143 +++++++++++++++++++++++++++++++++++++++++---
 2 files changed, 137 insertions(+), 8 deletions(-)

diff --git a/arch/x86/Kconfig b/arch/x86/Kconfig
index 2fdb23313dd5..2896c14da4c1 100644
--- a/arch/x86/Kconfig
+++ b/arch/x86/Kconfig
@@ -108,7 +108,7 @@ config X86
 	select HAVE_ARCH_AUDITSYSCALL
 	select HAVE_ARCH_HUGE_VMAP		if X86_64 || X86_PAE
 	select HAVE_ARCH_JUMP_LABEL
-	select HAVE_ARCH_KASAN			if X86_64 && SPARSEMEM_VMEMMAP
+	select HAVE_ARCH_KASAN			if X86_64
 	select HAVE_ARCH_KGDB
 	select HAVE_ARCH_KMEMCHECK
 	select HAVE_ARCH_MMAP_RND_BITS		if MMU
diff --git a/arch/x86/mm/kasan_init_64.c b/arch/x86/mm/kasan_init_64.c
index 8f5be3eb40dd..31c779b2c3a0 100644
--- a/arch/x86/mm/kasan_init_64.c
+++ b/arch/x86/mm/kasan_init_64.c
@@ -4,19 +4,148 @@
 #include <linux/bootmem.h>
 #include <linux/kasan.h>
 #include <linux/kdebug.h>
+#include <linux/memblock.h>
 #include <linux/mm.h>
 #include <linux/sched.h>
 #include <linux/sched/task.h>
 #include <linux/vmalloc.h>
 
 #include <asm/e820/types.h>
+#include <asm/pgalloc.h>
 #include <asm/tlbflush.h>
 #include <asm/sections.h>
 #include <asm/pgtable.h>
 
 extern struct range pfn_mapped[E820_MAX_ENTRIES];
 
-static int __init map_range(struct range *range)
+static __init void *early_alloc(size_t size, int nid)
+{
+	return memblock_virt_alloc_try_nid_nopanic(size, size,
+		__pa(MAX_DMA_ADDRESS), BOOTMEM_ALLOC_ACCESSIBLE, nid);
+}
+
+static void __init kasan_populate_pmd(pmd_t *pmd, unsigned long addr,
+				      unsigned long end, int nid)
+{
+	pte_t *pte;
+
+	if (pmd_none(*pmd)) {
+		void *p;
+
+		if (boot_cpu_has(X86_FEATURE_PSE) &&
+		    ((end - addr) == PMD_SIZE) &&
+		    IS_ALIGNED(addr, PMD_SIZE)) {
+			p = early_alloc(PMD_SIZE, nid);
+			if (p && pmd_set_huge(pmd, __pa(p), PAGE_KERNEL))
+				return;
+			else if (p)
+				memblock_free(__pa(p), PMD_SIZE);
+		}
+
+		p = early_alloc(PAGE_SIZE, nid);
+		pmd_populate_kernel(&init_mm, pmd, p);
+	}
+
+	pte = pte_offset_kernel(pmd, addr);
+	do {
+		pte_t entry;
+		void *p;
+
+		if (!pte_none(*pte))
+			continue;
+
+		p = early_alloc(PAGE_SIZE, nid);
+		entry = pfn_pte(PFN_DOWN(__pa(p)), PAGE_KERNEL);
+		set_pte_at(&init_mm, addr, pte, entry);
+	} while (pte++, addr += PAGE_SIZE, addr != end);
+}
+
+static void __init kasan_populate_pud(pud_t *pud, unsigned long addr,
+				      unsigned long end, int nid)
+{
+	pmd_t *pmd;
+	unsigned long next;
+
+	if (pud_none(*pud)) {
+		void *p;
+
+		if (boot_cpu_has(X86_FEATURE_GBPAGES) &&
+		    ((end - addr) == PUD_SIZE) &&
+		    IS_ALIGNED(addr, PUD_SIZE)) {
+			p = early_alloc(PUD_SIZE, nid);
+			if (p && pud_set_huge(pud, __pa(p), PAGE_KERNEL))
+				return;
+			else if (p)
+				memblock_free(__pa(p), PUD_SIZE);
+		}
+
+		p = early_alloc(PAGE_SIZE, nid);
+		pud_populate(&init_mm, pud, p);
+	}
+
+	pmd = pmd_offset(pud, addr);
+	do {
+		next = pmd_addr_end(addr, end);
+		if (!pmd_large(*pmd))
+			kasan_populate_pmd(pmd, addr, next, nid);
+	} while (pmd++, addr = next, addr != end);
+}
+
+static void __init kasan_populate_p4d(p4d_t *p4d, unsigned long addr,
+				      unsigned long end, int nid)
+{
+	pud_t *pud;
+	unsigned long next;
+
+	if (p4d_none(*p4d)) {
+		void *p = early_alloc(PAGE_SIZE, nid);
+
+		p4d_populate(&init_mm, p4d, p);
+	}
+
+	pud = pud_offset(p4d, addr);
+	do {
+		next = pud_addr_end(addr, end);
+		if (!pud_large(*pud))
+			kasan_populate_pud(pud, addr, next, nid);
+	} while (pud++, addr = next, addr != end);
+}
+
+static void __init kasan_populate_pgd(pgd_t *pgd, unsigned long addr,
+				      unsigned long end, int nid)
+{
+	void *p;
+	p4d_t *p4d;
+	unsigned long next;
+
+	if (pgd_none(*pgd)) {
+		p = early_alloc(PAGE_SIZE, nid);
+		pgd_populate(&init_mm, pgd, p);
+	}
+
+	p4d = p4d_offset(pgd, addr);
+	do {
+		next = p4d_addr_end(addr, end);
+		kasan_populate_p4d(p4d, addr, next, nid);
+	} while (p4d++, addr = next, addr != end);
+}
+
+static void __init kasan_populate_shadow(unsigned long addr, unsigned long end,
+					 int nid)
+{
+	pgd_t *pgd;
+	unsigned long next;
+
+	addr = addr & PAGE_MASK;
+	end = round_up(end, PAGE_SIZE);
+	pgd = pgd_offset_k(addr);
+	do {
+		next = pgd_addr_end(addr, end);
+		kasan_populate_pgd(pgd, addr, next, nid);
+	} while (pgd++, addr = next, addr != end);
+}
+
+static void __init map_range(struct range *range)
 {
 	unsigned long start;
 	unsigned long end;
@@ -24,7 +153,7 @@ static int __init map_range(struct range *range)
 	start = (unsigned long)kasan_mem_to_shadow(pfn_to_kaddr(range->start));
 	end = (unsigned long)kasan_mem_to_shadow(pfn_to_kaddr(range->end));
 
-	return vmemmap_populate(start, end, NUMA_NO_NODE);
+	kasan_populate_shadow(start, end, early_pfn_to_nid(range->start));
 }
 
 static void __init clear_pgds(unsigned long start,
@@ -130,16 +259,16 @@ void __init kasan_init(void)
 		if (pfn_mapped[i].end == 0)
 			break;
 
-		if (map_range(&pfn_mapped[i]))
-			panic("kasan: unable to allocate shadow!");
+		map_range(&pfn_mapped[i]);
 	}
+
 	kasan_populate_zero_shadow(
 		kasan_mem_to_shadow((void *)PAGE_OFFSET + MAXMEM),
 		kasan_mem_to_shadow((void *)__START_KERNEL_map));
 
-	vmemmap_populate((unsigned long)kasan_mem_to_shadow(_stext),
-			(unsigned long)kasan_mem_to_shadow(_end),
-			NUMA_NO_NODE);
+	kasan_populate_shadow((unsigned long)kasan_mem_to_shadow(_stext),
+			      (unsigned long)kasan_mem_to_shadow(_end),
+			      early_pfn_to_nid(__pa(_stext)));
 
 	kasan_populate_zero_shadow(kasan_mem_to_shadow((void *)MODULES_END),
 			(void *)KASAN_SHADOW_END);
-- 
2.15.0

[PATCH v1 2/2] arm64/mm/kasan: don't use vmemmap_populate() to initialize shadow

[Pavel Tatashin]

The kasan shadow is currently mapped using vmemmap_populate() since that
provides a semi-convenient way to map pages into init_top_pgt. However,
since that no longer zeroes the mapped pages, it is not suitable for kasan,
which requires zeroed shadow memory.

Add kasan_populate_shadow() interface and use it instead of
vmemmap_populate(). Besides, this allows us to take advantage of gigantic
pages and use them to populate the shadow, which should save us some memory
wasted on page tables and reduce TLB pressure.

Signed-off-by: Will Deacon <will.deacon@arm.com>
Signed-off-by: Pavel Tatashin <pasha.tatashin@oracle.com>
---
 arch/arm64/Kconfig         |   2 +-
 arch/arm64/mm/kasan_init.c | 130 ++++++++++++++++++++++++++++-----------------
 2 files changed, 81 insertions(+), 51 deletions(-)

diff --git a/arch/arm64/Kconfig b/arch/arm64/Kconfig
index 0df64a6a56d4..888580b9036e 100644
--- a/arch/arm64/Kconfig
+++ b/arch/arm64/Kconfig
@@ -68,7 +68,7 @@ config ARM64
 	select HAVE_ARCH_BITREVERSE
 	select HAVE_ARCH_HUGE_VMAP
 	select HAVE_ARCH_JUMP_LABEL
-	select HAVE_ARCH_KASAN if SPARSEMEM_VMEMMAP && !(ARM64_16K_PAGES && ARM64_VA_BITS_48)
+	select HAVE_ARCH_KASAN if !(ARM64_16K_PAGES && ARM64_VA_BITS_48)
 	select HAVE_ARCH_KGDB
 	select HAVE_ARCH_MMAP_RND_BITS
 	select HAVE_ARCH_MMAP_RND_COMPAT_BITS if COMPAT
diff --git a/arch/arm64/mm/kasan_init.c b/arch/arm64/mm/kasan_init.c
index 81f03959a4ab..acba49fb5aac 100644
--- a/arch/arm64/mm/kasan_init.c
+++ b/arch/arm64/mm/kasan_init.c
@@ -11,6 +11,7 @@
  */
 
 #define pr_fmt(fmt) "kasan: " fmt
+#include <linux/bootmem.h>
 #include <linux/kasan.h>
 #include <linux/kernel.h>
 #include <linux/sched/task.h>
@@ -35,77 +36,117 @@ static pgd_t tmp_pg_dir[PTRS_PER_PGD] __initdata __aligned(PGD_SIZE);
  * with the physical address from __pa_symbol.
  */
 
-static void __init kasan_early_pte_populate(pmd_t *pmd, unsigned long addr,
-					unsigned long end)
+static phys_addr_t __init kasan_alloc_zeroed_page(int node)
 {
-	pte_t *pte;
-	unsigned long next;
+	void *p = memblock_virt_alloc_try_nid(PAGE_SIZE, PAGE_SIZE,
+					      __pa(MAX_DMA_ADDRESS),
+					      MEMBLOCK_ALLOC_ACCESSIBLE, node);
+	return __pa(p);
+}
+
+static pte_t *__init kasan_pte_offset(pmd_t *pmd, unsigned long addr, int node,
+				      bool early)
+{
+	if (pmd_none(*pmd)) {
+		phys_addr_t pte_phys = early ? __pa_symbol(kasan_zero_pte)
+					     : kasan_alloc_zeroed_page(node);
+		__pmd_populate(pmd, pte_phys, PMD_TYPE_TABLE);
+	}
+
+	return early ? pte_offset_kimg(pmd, addr)
+		     : pte_offset_kernel(pmd, addr);
+}
 
-	if (pmd_none(*pmd))
-		__pmd_populate(pmd, __pa_symbol(kasan_zero_pte), PMD_TYPE_TABLE);
+static pmd_t *__init kasan_pmd_offset(pud_t *pud, unsigned long addr, int node,
+				      bool early)
+{
+	if (pud_none(*pud)) {
+		phys_addr_t pmd_phys = early ? __pa_symbol(kasan_zero_pmd)
+					     : kasan_alloc_zeroed_page(node);
+		__pud_populate(pud, pmd_phys, PMD_TYPE_TABLE);
+	}
+
+	return early ? pmd_offset_kimg(pud, addr) : pmd_offset(pud, addr);
+}
+
+static pud_t *__init kasan_pud_offset(pgd_t *pgd, unsigned long addr, int node,
+				      bool early)
+{
+	if (pgd_none(*pgd)) {
+		phys_addr_t pud_phys = early ? __pa_symbol(kasan_zero_pud)
+					     : kasan_alloc_zeroed_page(node);
+		__pgd_populate(pgd, pud_phys, PMD_TYPE_TABLE);
+	}
+
+	return early ? pud_offset_kimg(pgd, addr) : pud_offset(pgd, addr);
+}
+
+static void __init kasan_pte_populate(pmd_t *pmd, unsigned long addr,
+				      unsigned long end, int node, bool early)
+{
+	unsigned long next;
+	pte_t *pte = kasan_pte_offset(pmd, addr, node, early);
 
-	pte = pte_offset_kimg(pmd, addr);
 	do {
+		phys_addr_t page_phys = early ? __pa_symbol(kasan_zero_page)
+					      : kasan_alloc_zeroed_page(node);
 		next = addr + PAGE_SIZE;
-		set_pte(pte, pfn_pte(sym_to_pfn(kasan_zero_page),
-					PAGE_KERNEL));
+		set_pte(pte, pfn_pte(__phys_to_pfn(page_phys), PAGE_KERNEL));
 	} while (pte++, addr = next, addr != end && pte_none(*pte));
 }
 
-static void __init kasan_early_pmd_populate(pud_t *pud,
-					unsigned long addr,
-					unsigned long end)
+static void __init kasan_pmd_populate(pud_t *pud, unsigned long addr,
+				      unsigned long end, int node, bool early)
 {
-	pmd_t *pmd;
 	unsigned long next;
+	pmd_t *pmd = kasan_pmd_offset(pud, addr, node, early);
 
-	if (pud_none(*pud))
-		__pud_populate(pud, __pa_symbol(kasan_zero_pmd), PMD_TYPE_TABLE);
-
-	pmd = pmd_offset_kimg(pud, addr);
 	do {
 		next = pmd_addr_end(addr, end);
-		kasan_early_pte_populate(pmd, addr, next);
+		kasan_pte_populate(pmd, addr, next, node, early);
 	} while (pmd++, addr = next, addr != end && pmd_none(*pmd));
 }
 
-static void __init kasan_early_pud_populate(pgd_t *pgd,
-					unsigned long addr,
-					unsigned long end)
+static void __init kasan_pud_populate(pgd_t *pgd, unsigned long addr,
+				      unsigned long end, int node, bool early)
 {
-	pud_t *pud;
 	unsigned long next;
+	pud_t *pud = kasan_pud_offset(pgd, addr, node, early);
 
-	if (pgd_none(*pgd))
-		__pgd_populate(pgd, __pa_symbol(kasan_zero_pud), PUD_TYPE_TABLE);
-
-	pud = pud_offset_kimg(pgd, addr);
 	do {
 		next = pud_addr_end(addr, end);
-		kasan_early_pmd_populate(pud, addr, next);
+		kasan_pmd_populate(pud, addr, next, node, early);
 	} while (pud++, addr = next, addr != end && pud_none(*pud));
 }
 
-static void __init kasan_map_early_shadow(void)
+static void __init kasan_pgd_populate(unsigned long addr, unsigned long end,
+				      int node, bool early)
 {
-	unsigned long addr = KASAN_SHADOW_START;
-	unsigned long end = KASAN_SHADOW_END;
 	unsigned long next;
 	pgd_t *pgd;
 
 	pgd = pgd_offset_k(addr);
 	do {
 		next = pgd_addr_end(addr, end);
-		kasan_early_pud_populate(pgd, addr, next);
+		kasan_pud_populate(pgd, addr, next, node, early);
 	} while (pgd++, addr = next, addr != end);
 }
 
+/* The early shadow maps everything to a single page of zeroes */
 asmlinkage void __init kasan_early_init(void)
 {
 	BUILD_BUG_ON(KASAN_SHADOW_OFFSET != KASAN_SHADOW_END - (1UL << 61));
 	BUILD_BUG_ON(!IS_ALIGNED(KASAN_SHADOW_START, PGDIR_SIZE));
 	BUILD_BUG_ON(!IS_ALIGNED(KASAN_SHADOW_END, PGDIR_SIZE));
-	kasan_map_early_shadow();
+	kasan_pgd_populate(KASAN_SHADOW_START, KASAN_SHADOW_END, NUMA_NO_NODE,
+			   true);
+}
+
+/* Set up full kasan mappings, ensuring that the mapped pages are zeroed */
+static void __init kasan_map_populate(unsigned long start, unsigned long end,
+				      int node)
+{
+	kasan_pgd_populate(start & PAGE_MASK, PAGE_ALIGN(end), node, false);
 }
 
 /*
@@ -142,8 +183,8 @@ void __init kasan_init(void)
 	struct memblock_region *reg;
 	int i;
 
-	kimg_shadow_start = (u64)kasan_mem_to_shadow(_text);
-	kimg_shadow_end = (u64)kasan_mem_to_shadow(_end);
+	kimg_shadow_start = (u64)kasan_mem_to_shadow(_text) & PAGE_MASK;
+	kimg_shadow_end = PAGE_ALIGN((u64)kasan_mem_to_shadow(_end));
 
 	mod_shadow_start = (u64)kasan_mem_to_shadow((void *)MODULES_VADDR);
 	mod_shadow_end = (u64)kasan_mem_to_shadow((void *)MODULES_END);
@@ -161,19 +202,8 @@ void __init kasan_init(void)
 
 	clear_pgds(KASAN_SHADOW_START, KASAN_SHADOW_END);
 
-	vmemmap_populate(kimg_shadow_start, kimg_shadow_end,
-			 pfn_to_nid(virt_to_pfn(lm_alias(_text))));
-
-	/*
-	 * vmemmap_populate() has populated the shadow region that covers the
-	 * kernel image with SWAPPER_BLOCK_SIZE mappings, so we have to round
-	 * the start and end addresses to SWAPPER_BLOCK_SIZE as well, to prevent
-	 * kasan_populate_zero_shadow() from replacing the page table entries
-	 * (PMD or PTE) at the edges of the shadow region for the kernel
-	 * image.
-	 */
-	kimg_shadow_start = round_down(kimg_shadow_start, SWAPPER_BLOCK_SIZE);
-	kimg_shadow_end = round_up(kimg_shadow_end, SWAPPER_BLOCK_SIZE);
+	kasan_map_populate(kimg_shadow_start, kimg_shadow_end,
+			   pfn_to_nid(virt_to_pfn(lm_alias(_text))));
 
 	kasan_populate_zero_shadow((void *)KASAN_SHADOW_START,
 				   (void *)mod_shadow_start);
@@ -191,9 +221,9 @@ void __init kasan_init(void)
 		if (start >= end)
 			break;
 
-		vmemmap_populate((unsigned long)kasan_mem_to_shadow(start),
-				(unsigned long)kasan_mem_to_shadow(end),
-				pfn_to_nid(virt_to_pfn(start)));
+		kasan_map_populate((unsigned long)kasan_mem_to_shadow(start),
+				   (unsigned long)kasan_mem_to_shadow(end),
+				   pfn_to_nid(virt_to_pfn(start)));
 	}
 
 	/*
-- 
2.15.0

[tip:x86/urgent] x86/mm/kasan: Don't use vmemmap_populate() to initialize shadow

[tip-bot for Andrey Ryabinin]

Commit-ID:  f68d62a56708b0c19dca7a998f408510f2fbc3a8
Gitweb:     https://git.kernel.org/tip/f68d62a56708b0c19dca7a998f408510f2fbc3a8
Author:     Andrey Ryabinin <aryabinin@virtuozzo.com>
AuthorDate: Wed, 15 Nov 2017 17:36:35 -0800
Committer:  Ingo Molnar <mingo@kernel.org>
CommitDate: Wed, 22 Nov 2017 07:18:35 +0100

x86/mm/kasan: Don't use vmemmap_populate() to initialize shadow

[ Note, this commit is a cherry-picked version of:

    d17a1d97dc20: ("x86/mm/kasan: don't use vmemmap_populate() to initialize shadow")

  ... for easier x86 entry code testing and back-porting. ]

The KASAN shadow is currently mapped using vmemmap_populate() since that
provides a semi-convenient way to map pages into init_top_pgt.  However,
since that no longer zeroes the mapped pages, it is not suitable for
KASAN, which requires zeroed shadow memory.

Add kasan_populate_shadow() interface and use it instead of
vmemmap_populate().  Besides, this allows us to take advantage of
gigantic pages and use them to populate the shadow, which should save us
some memory wasted on page tables and reduce TLB pressure.

Link: http://lkml.kernel.org/r/20171103185147.2688-2-pasha.tatashin@oracle.com
Signed-off-by: Andrey Ryabinin <aryabinin@virtuozzo.com>
Signed-off-by: Pavel Tatashin <pasha.tatashin@oracle.com>
Cc: Andy Lutomirski <luto@kernel.org>
Cc: Steven Sistare <steven.sistare@oracle.com>
Cc: Daniel Jordan <daniel.m.jordan@oracle.com>
Cc: Bob Picco <bob.picco@oracle.com>
Cc: Michal Hocko <mhocko@suse.com>
Cc: Alexander Potapenko <glider@google.com>
Cc: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Cc: Catalin Marinas <catalin.marinas@arm.com>
Cc: Christian Borntraeger <borntraeger@de.ibm.com>
Cc: David S. Miller <davem@davemloft.net>
Cc: Dmitry Vyukov <dvyukov@google.com>
Cc: Heiko Carstens <heiko.carstens@de.ibm.com>
Cc: "H. Peter Anvin" <hpa@zytor.com>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: Mark Rutland <mark.rutland@arm.com>
Cc: Matthew Wilcox <willy@infradead.org>
Cc: Mel Gorman <mgorman@techsingularity.net>
Cc: Michal Hocko <mhocko@kernel.org>
Cc: Sam Ravnborg <sam@ravnborg.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Will Deacon <will.deacon@arm.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
---
 arch/x86/Kconfig            |   2 +-
 arch/x86/mm/kasan_init_64.c | 143 +++++++++++++++++++++++++++++++++++++++++---
 2 files changed, 137 insertions(+), 8 deletions(-)

diff --git a/arch/x86/Kconfig b/arch/x86/Kconfig
index a0623f0..09dcc94 100644
--- a/arch/x86/Kconfig
+++ b/arch/x86/Kconfig
@@ -110,7 +110,7 @@ config X86
 	select HAVE_ARCH_AUDITSYSCALL
 	select HAVE_ARCH_HUGE_VMAP		if X86_64 || X86_PAE
 	select HAVE_ARCH_JUMP_LABEL
-	select HAVE_ARCH_KASAN			if X86_64 && SPARSEMEM_VMEMMAP
+	select HAVE_ARCH_KASAN			if X86_64
 	select HAVE_ARCH_KGDB
 	select HAVE_ARCH_KMEMCHECK
 	select HAVE_ARCH_MMAP_RND_BITS		if MMU
diff --git a/arch/x86/mm/kasan_init_64.c b/arch/x86/mm/kasan_init_64.c
index 2b60dc6..99dfed6 100644
--- a/arch/x86/mm/kasan_init_64.c
+++ b/arch/x86/mm/kasan_init_64.c
@@ -4,12 +4,14 @@
 #include <linux/bootmem.h>
 #include <linux/kasan.h>
 #include <linux/kdebug.h>
+#include <linux/memblock.h>
 #include <linux/mm.h>
 #include <linux/sched.h>
 #include <linux/sched/task.h>
 #include <linux/vmalloc.h>
 
 #include <asm/e820/types.h>
+#include <asm/pgalloc.h>
 #include <asm/tlbflush.h>
 #include <asm/sections.h>
 #include <asm/pgtable.h>
@@ -18,7 +20,134 @@ extern struct range pfn_mapped[E820_MAX_ENTRIES];
 
 static p4d_t tmp_p4d_table[PTRS_PER_P4D] __initdata __aligned(PAGE_SIZE);
 
-static int __init map_range(struct range *range)
+static __init void *early_alloc(size_t size, int nid)
+{
+	return memblock_virt_alloc_try_nid_nopanic(size, size,
+		__pa(MAX_DMA_ADDRESS), BOOTMEM_ALLOC_ACCESSIBLE, nid);
+}
+
+static void __init kasan_populate_pmd(pmd_t *pmd, unsigned long addr,
+				      unsigned long end, int nid)
+{
+	pte_t *pte;
+
+	if (pmd_none(*pmd)) {
+		void *p;
+
+		if (boot_cpu_has(X86_FEATURE_PSE) &&
+		    ((end - addr) == PMD_SIZE) &&
+		    IS_ALIGNED(addr, PMD_SIZE)) {
+			p = early_alloc(PMD_SIZE, nid);
+			if (p && pmd_set_huge(pmd, __pa(p), PAGE_KERNEL))
+				return;
+			else if (p)
+				memblock_free(__pa(p), PMD_SIZE);
+		}
+
+		p = early_alloc(PAGE_SIZE, nid);
+		pmd_populate_kernel(&init_mm, pmd, p);
+	}
+
+	pte = pte_offset_kernel(pmd, addr);
+	do {
+		pte_t entry;
+		void *p;
+
+		if (!pte_none(*pte))
+			continue;
+
+		p = early_alloc(PAGE_SIZE, nid);
+		entry = pfn_pte(PFN_DOWN(__pa(p)), PAGE_KERNEL);
+		set_pte_at(&init_mm, addr, pte, entry);
+	} while (pte++, addr += PAGE_SIZE, addr != end);
+}
+
+static void __init kasan_populate_pud(pud_t *pud, unsigned long addr,
+				      unsigned long end, int nid)
+{
+	pmd_t *pmd;
+	unsigned long next;
+
+	if (pud_none(*pud)) {
+		void *p;
+
+		if (boot_cpu_has(X86_FEATURE_GBPAGES) &&
+		    ((end - addr) == PUD_SIZE) &&
+		    IS_ALIGNED(addr, PUD_SIZE)) {
+			p = early_alloc(PUD_SIZE, nid);
+			if (p && pud_set_huge(pud, __pa(p), PAGE_KERNEL))
+				return;
+			else if (p)
+				memblock_free(__pa(p), PUD_SIZE);
+		}
+
+		p = early_alloc(PAGE_SIZE, nid);
+		pud_populate(&init_mm, pud, p);
+	}
+
+	pmd = pmd_offset(pud, addr);
+	do {
+		next = pmd_addr_end(addr, end);
+		if (!pmd_large(*pmd))
+			kasan_populate_pmd(pmd, addr, next, nid);
+	} while (pmd++, addr = next, addr != end);
+}
+
+static void __init kasan_populate_p4d(p4d_t *p4d, unsigned long addr,
+				      unsigned long end, int nid)
+{
+	pud_t *pud;
+	unsigned long next;
+
+	if (p4d_none(*p4d)) {
+		void *p = early_alloc(PAGE_SIZE, nid);
+
+		p4d_populate(&init_mm, p4d, p);
+	}
+
+	pud = pud_offset(p4d, addr);
+	do {
+		next = pud_addr_end(addr, end);
+		if (!pud_large(*pud))
+			kasan_populate_pud(pud, addr, next, nid);
+	} while (pud++, addr = next, addr != end);
+}
+
+static void __init kasan_populate_pgd(pgd_t *pgd, unsigned long addr,
+				      unsigned long end, int nid)
+{
+	void *p;
+	p4d_t *p4d;
+	unsigned long next;
+
+	if (pgd_none(*pgd)) {
+		p = early_alloc(PAGE_SIZE, nid);
+		pgd_populate(&init_mm, pgd, p);
+	}
+
+	p4d = p4d_offset(pgd, addr);
+	do {
+		next = p4d_addr_end(addr, end);
+		kasan_populate_p4d(p4d, addr, next, nid);
+	} while (p4d++, addr = next, addr != end);
+}
+
+static void __init kasan_populate_shadow(unsigned long addr, unsigned long end,
+					 int nid)
+{
+	pgd_t *pgd;
+	unsigned long next;
+
+	addr = addr & PAGE_MASK;
+	end = round_up(end, PAGE_SIZE);
+	pgd = pgd_offset_k(addr);
+	do {
+		next = pgd_addr_end(addr, end);
+		kasan_populate_pgd(pgd, addr, next, nid);
+	} while (pgd++, addr = next, addr != end);
+}
+
+static void __init map_range(struct range *range)
 {
 	unsigned long start;
 	unsigned long end;
@@ -26,7 +155,7 @@ static int __init map_range(struct range *range)
 	start = (unsigned long)kasan_mem_to_shadow(pfn_to_kaddr(range->start));
 	end = (unsigned long)kasan_mem_to_shadow(pfn_to_kaddr(range->end));
 
-	return vmemmap_populate(start, end, NUMA_NO_NODE);
+	kasan_populate_shadow(start, end, early_pfn_to_nid(range->start));
 }
 
 static void __init clear_pgds(unsigned long start,
@@ -189,16 +318,16 @@ void __init kasan_init(void)
 		if (pfn_mapped[i].end == 0)
 			break;
 
-		if (map_range(&pfn_mapped[i]))
-			panic("kasan: unable to allocate shadow!");
+		map_range(&pfn_mapped[i]);
 	}
+
 	kasan_populate_zero_shadow(
 		kasan_mem_to_shadow((void *)PAGE_OFFSET + MAXMEM),
 		kasan_mem_to_shadow((void *)__START_KERNEL_map));
 
-	vmemmap_populate((unsigned long)kasan_mem_to_shadow(_stext),
-			(unsigned long)kasan_mem_to_shadow(_end),
-			NUMA_NO_NODE);
+	kasan_populate_shadow((unsigned long)kasan_mem_to_shadow(_stext),
+			      (unsigned long)kasan_mem_to_shadow(_end),
+			      early_pfn_to_nid(__pa(_stext)));
 
 	kasan_populate_zero_shadow(kasan_mem_to_shadow((void *)MODULES_END),
 			(void *)KASAN_SHADOW_END);

Re: [tip:x86/urgent] x86/mm/kasan: Don't use vmemmap_populate() to initialize shadow

[Andy Lutomirski]

On Tue, Nov 21, 2017 at 11:46 PM, tip-bot for Andrey Ryabinin
<tipbot@zytor.com> wrote:
> Commit-ID:  f68d62a56708b0c19dca7a998f408510f2fbc3a8
> Gitweb:     https://git.kernel.org/tip/f68d62a56708b0c19dca7a998f408510f2fbc3a8
> Author:     Andrey Ryabinin <aryabinin@virtuozzo.com>
> AuthorDate: Wed, 15 Nov 2017 17:36:35 -0800
> Committer:  Ingo Molnar <mingo@kernel.org>
> CommitDate: Wed, 22 Nov 2017 07:18:35 +0100
>
> x86/mm/kasan: Don't use vmemmap_populate() to initialize shadow
>
> [ Note, this commit is a cherry-picked version of:
>
>     d17a1d97dc20: ("x86/mm/kasan: don't use vmemmap_populate() to initialize shadow")
>
>   ... for easier x86 entry code testing and back-porting. ]

I'm confused.  Isn't this patch already in Linus' tree, merged in (I
think) 7c225c69f86c934e3be9be63ecde754e286838d7?