[PATCH] i386: Avoid overwriting the current pgd (V4, i386)

kexec: Avoid overwriting the current pgd (V4, i386)

This patch upgrades the i386-specific kexec code to avoid overwriting the
current pgd. Overwriting the current pgd is bad when CONFIG_CRASH_DUMP is used
to start a secondary kernel that dumps the memory of the previous kernel.

The code introduces a new set of page tables. These tables are used to provide
an executable identity mapping without overwriting the current pgd.

Signed-off-by: Magnus Damm <magnus@valinux.co.jp>
Signed-off-by: Andi Kleen <ak@suse.de>

3 files changed, 201 insertions, 105 deletions

diff --git a/arch/i386/kernel/machine_kexec.c b/arch/i386/kernel/machine_kexec.c
index 66c3dc99a65..91966bafb3d 100644
--- a/arch/i386/kernel/machine_kexec.c
+++ b/arch/i386/kernel/machine_kexec.c
@@ -21,70 +21,13 @@
 #include <asm/system.h>
 
 #define PAGE_ALIGNED __attribute__ ((__aligned__(PAGE_SIZE)))
-
-#define L0_ATTR (_PAGE_PRESENT | _PAGE_RW | _PAGE_ACCESSED | _PAGE_DIRTY)
-#define L1_ATTR (_PAGE_PRESENT | _PAGE_RW | _PAGE_ACCESSED | _PAGE_DIRTY)
-#define L2_ATTR (_PAGE_PRESENT)
-
-#define LEVEL0_SIZE (1UL << 12UL)
-
-#ifndef CONFIG_X86_PAE
-#define LEVEL1_SIZE (1UL << 22UL)
-static u32 pgtable_level1[1024] PAGE_ALIGNED;
-
-static void identity_map_page(unsigned long address)
-{
-	unsigned long level1_index, level2_index;
-	u32 *pgtable_level2;
-
-	/* Find the current page table */
-	pgtable_level2 = __va(read_cr3());
-
-	/* Find the indexes of the physical address to identity map */
-	level1_index = (address % LEVEL1_SIZE)/LEVEL0_SIZE;
-	level2_index = address / LEVEL1_SIZE;
-
-	/* Identity map the page table entry */
-	pgtable_level1[level1_index] = address | L0_ATTR;
-	pgtable_level2[level2_index] = __pa(pgtable_level1) | L1_ATTR;
-
-	/* Flush the tlb so the new mapping takes effect.
-	 * Global tlb entries are not flushed but that is not an issue.
-	 */
-	load_cr3(pgtable_level2);
-}
-
-#else
-#define LEVEL1_SIZE (1UL << 21UL)
-#define LEVEL2_SIZE (1UL << 30UL)
-static u64 pgtable_level1[512] PAGE_ALIGNED;
-static u64 pgtable_level2[512] PAGE_ALIGNED;
-
-static void identity_map_page(unsigned long address)
-{
-	unsigned long level1_index, level2_index, level3_index;
-	u64 *pgtable_level3;
-
-	/* Find the current page table */
-	pgtable_level3 = __va(read_cr3());
-
-	/* Find the indexes of the physical address to identity map */
-	level1_index = (address % LEVEL1_SIZE)/LEVEL0_SIZE;
-	level2_index = (address % LEVEL2_SIZE)/LEVEL1_SIZE;
-	level3_index = address / LEVEL2_SIZE;
-
-	/* Identity map the page table entry */
-	pgtable_level1[level1_index] = address | L0_ATTR;
-	pgtable_level2[level2_index] = __pa(pgtable_level1) | L1_ATTR;
-	set_64bit(&pgtable_level3[level3_index],
-					       __pa(pgtable_level2) | L2_ATTR);
-
-	/* Flush the tlb so the new mapping takes effect.
-	 * Global tlb entries are not flushed but that is not an issue.
-	 */
-	load_cr3(pgtable_level3);
-}
+static u32 kexec_pgd[1024] PAGE_ALIGNED;
+#ifdef CONFIG_X86_PAE
+static u32 kexec_pmd0[1024] PAGE_ALIGNED;
+static u32 kexec_pmd1[1024] PAGE_ALIGNED;
 #endif
+static u32 kexec_pte0[1024] PAGE_ALIGNED;
+static u32 kexec_pte1[1024] PAGE_ALIGNED;
 
 static void set_idt(void *newidt, __u16 limit)
 {
@@ -128,16 +71,6 @@ static void load_segments(void)
 #undef __STR
 }
 
-typedef asmlinkage NORET_TYPE void (*relocate_new_kernel_t)(
-					unsigned long indirection_page,
-					unsigned long reboot_code_buffer,
-					unsigned long start_address,
-					unsigned int has_pae) ATTRIB_NORET;
-
-extern const unsigned char relocate_new_kernel[];
-extern void relocate_new_kernel_end(void);
-extern const unsigned int relocate_new_kernel_size;
-
 /*
  * A architecture hook called to validate the
  * proposed image and prepare the control pages
@@ -170,25 +103,29 @@ void machine_kexec_cleanup(struct kimage *image)
  */
 NORET_TYPE void machine_kexec(struct kimage *image)
 {
-	unsigned long page_list;
-	unsigned long reboot_code_buffer;
-
-	relocate_new_kernel_t rnk;
+	unsigned long page_list[PAGES_NR];
+	void *control_page;
 
 	/* Interrupts aren't acceptable while we reboot */
 	local_irq_disable();
 
-	/* Compute some offsets */
-	reboot_code_buffer = page_to_pfn(image->control_code_page)
-								<< PAGE_SHIFT;
-	page_list = image->head;
-
-	/* Set up an identity mapping for the reboot_code_buffer */
-	identity_map_page(reboot_code_buffer);
-
-	/* copy it out */
-	memcpy((void *)reboot_code_buffer, relocate_new_kernel,
-						relocate_new_kernel_size);
+	control_page = page_address(image->control_code_page);
+	memcpy(control_page, relocate_kernel, PAGE_SIZE);
+
+	page_list[PA_CONTROL_PAGE] = __pa(control_page);
+	page_list[VA_CONTROL_PAGE] = (unsigned long)relocate_kernel;
+	page_list[PA_PGD] = __pa(kexec_pgd);
+	page_list[VA_PGD] = (unsigned long)kexec_pgd;
+#ifdef CONFIG_X86_PAE
+	page_list[PA_PMD_0] = __pa(kexec_pmd0);
+	page_list[VA_PMD_0] = (unsigned long)kexec_pmd0;
+	page_list[PA_PMD_1] = __pa(kexec_pmd1);
+	page_list[VA_PMD_1] = (unsigned long)kexec_pmd1;
+#endif
+	page_list[PA_PTE_0] = __pa(kexec_pte0);
+	page_list[VA_PTE_0] = (unsigned long)kexec_pte0;
+	page_list[PA_PTE_1] = __pa(kexec_pte1);
+	page_list[VA_PTE_1] = (unsigned long)kexec_pte1;
 
 	/* The segment registers are funny things, they have both a
 	 * visible and an invisible part.  Whenever the visible part is
@@ -207,8 +144,8 @@ NORET_TYPE void machine_kexec(struct kimage *image)
 	set_idt(phys_to_virt(0),0);
 
 	/* now call it */
-	rnk = (relocate_new_kernel_t) reboot_code_buffer;
-	(*rnk)(page_list, reboot_code_buffer, image->start, cpu_has_pae);
+	relocate_kernel((unsigned long)image->head, (unsigned long)page_list,
+			image->start, cpu_has_pae);
 }
 
 /* crashkernel=size@addr specifies the location to reserve for
diff --git a/arch/i386/kernel/relocate_kernel.S b/arch/i386/kernel/relocate_kernel.S
index d312616effa..f151d6fae46 100644
--- a/arch/i386/kernel/relocate_kernel.S
+++ b/arch/i386/kernel/relocate_kernel.S
@@ -7,16 +7,138 @@
  */
 
 #include <linux/linkage.h>
+#include <asm/page.h>
+#include <asm/kexec.h>
+
+/*
+ * Must be relocatable PIC code callable as a C function
+ */
+
+#define PTR(x) (x << 2)
+#define PAGE_ALIGNED (1 << PAGE_SHIFT)
+#define PAGE_ATTR 0x63 /* _PAGE_PRESENT|_PAGE_RW|_PAGE_ACCESSED|_PAGE_DIRTY */
+#define PAE_PGD_ATTR 0x01 /* _PAGE_PRESENT */
+
+	.text
+	.align PAGE_ALIGNED
+	.globl relocate_kernel
+relocate_kernel:
+	movl	8(%esp), %ebp /* list of pages */
+
+#ifdef CONFIG_X86_PAE
+	/* map the control page at its virtual address */
+
+	movl	PTR(VA_PGD)(%ebp), %edi
+	movl	PTR(VA_CONTROL_PAGE)(%ebp), %eax
+	andl	$0xc0000000, %eax
+	shrl	$27, %eax
+	addl	%edi, %eax
+
+	movl	PTR(PA_PMD_0)(%ebp), %edx
+	orl	$PAE_PGD_ATTR, %edx
+	movl	%edx, (%eax)
+
+	movl	PTR(VA_PMD_0)(%ebp), %edi
+	movl	PTR(VA_CONTROL_PAGE)(%ebp), %eax
+	andl	$0x3fe00000, %eax
+	shrl	$18, %eax
+	addl	%edi, %eax
+
+	movl	PTR(PA_PTE_0)(%ebp), %edx
+	orl	$PAGE_ATTR, %edx
+	movl	%edx, (%eax)
+
+	movl	PTR(VA_PTE_0)(%ebp), %edi
+	movl	PTR(VA_CONTROL_PAGE)(%ebp), %eax
+	andl	$0x001ff000, %eax
+	shrl	$9, %eax
+	addl	%edi, %eax
+
+	movl	PTR(PA_CONTROL_PAGE)(%ebp), %edx
+	orl	$PAGE_ATTR, %edx
+	movl	%edx, (%eax)
+
+	/* identity map the control page at its physical address */
+
+	movl	PTR(VA_PGD)(%ebp), %edi
+	movl	PTR(PA_CONTROL_PAGE)(%ebp), %eax
+	andl	$0xc0000000, %eax
+	shrl	$27, %eax
+	addl	%edi, %eax
+
+	movl	PTR(PA_PMD_1)(%ebp), %edx
+	orl	$PAE_PGD_ATTR, %edx
+	movl	%edx, (%eax)
+
+	movl	PTR(VA_PMD_1)(%ebp), %edi
+	movl	PTR(PA_CONTROL_PAGE)(%ebp), %eax
+	andl	$0x3fe00000, %eax
+	shrl	$18, %eax
+	addl	%edi, %eax
+
+	movl	PTR(PA_PTE_1)(%ebp), %edx
+	orl	$PAGE_ATTR, %edx
+	movl	%edx, (%eax)
+
+	movl	PTR(VA_PTE_1)(%ebp), %edi
+	movl	PTR(PA_CONTROL_PAGE)(%ebp), %eax
+	andl	$0x001ff000, %eax
+	shrl	$9, %eax
+	addl	%edi, %eax
+
+	movl	PTR(PA_CONTROL_PAGE)(%ebp), %edx
+	orl	$PAGE_ATTR, %edx
+	movl	%edx, (%eax)
+#else
+	/* map the control page at its virtual address */
+
+	movl	PTR(VA_PGD)(%ebp), %edi
+	movl	PTR(VA_CONTROL_PAGE)(%ebp), %eax
+	andl	$0xffc00000, %eax
+	shrl	$20, %eax
+	addl	%edi, %eax
+
+	movl	PTR(PA_PTE_0)(%ebp), %edx
+	orl	$PAGE_ATTR, %edx
+	movl	%edx, (%eax)
+
+	movl	PTR(VA_PTE_0)(%ebp), %edi
+	movl	PTR(VA_CONTROL_PAGE)(%ebp), %eax
+	andl	$0x003ff000, %eax
+	shrl	$10, %eax
+	addl	%edi, %eax
+
+	movl	PTR(PA_CONTROL_PAGE)(%ebp), %edx
+	orl	$PAGE_ATTR, %edx
+	movl	%edx, (%eax)
+
+	/* identity map the control page at its physical address */
+
+	movl	PTR(VA_PGD)(%ebp), %edi
+	movl	PTR(PA_CONTROL_PAGE)(%ebp), %eax
+	andl	$0xffc00000, %eax
+	shrl	$20, %eax
+	addl	%edi, %eax
+
+	movl	PTR(PA_PTE_1)(%ebp), %edx
+	orl	$PAGE_ATTR, %edx
+	movl	%edx, (%eax)
+
+	movl	PTR(VA_PTE_1)(%ebp), %edi
+	movl	PTR(PA_CONTROL_PAGE)(%ebp), %eax
+	andl	$0x003ff000, %eax
+	shrl	$10, %eax
+	addl	%edi, %eax
+
+	movl	PTR(PA_CONTROL_PAGE)(%ebp), %edx
+	orl	$PAGE_ATTR, %edx
+	movl	%edx, (%eax)
+#endif
 
-	/*
-	 * Must be relocatable PIC code callable as a C function, that once
-	 * it starts can not use the previous processes stack.
-	 */
-	.globl relocate_new_kernel
 relocate_new_kernel:
 	/* read the arguments and say goodbye to the stack */
 	movl  4(%esp), %ebx /* page_list */
-	movl  8(%esp), %ebp /* reboot_code_buffer */
+	movl  8(%esp), %ebp /* list of pages */
 	movl  12(%esp), %edx /* start address */
 	movl  16(%esp), %ecx /* cpu_has_pae */
 
@@ -24,11 +146,26 @@ relocate_new_kernel:
 	pushl $0
 	popfl
 
-	/* set a new stack at the bottom of our page... */
-	lea   4096(%ebp), %esp
+	/* get physical address of control page now */
+	/* this is impossible after page table switch */
+	movl	PTR(PA_CONTROL_PAGE)(%ebp), %edi
 
-	/* store the parameters back on the stack */
-	pushl   %edx /* store the start address */
+	/* switch to new set of page tables */
+	movl	PTR(PA_PGD)(%ebp), %eax
+	movl	%eax, %cr3
+
+	/* setup a new stack at the end of the physical control page */
+	lea	4096(%edi), %esp
+
+	/* jump to identity mapped page */
+	movl    %edi, %eax
+	addl    $(identity_mapped - relocate_kernel), %eax
+	pushl   %eax
+	ret
+
+identity_mapped:
+	/* store the start address on the stack */
+	pushl   %edx
 
 	/* Set cr0 to a known state:
 	 * 31 0 == Paging disabled
@@ -113,8 +250,3 @@ relocate_new_kernel:
 	xorl    %edi, %edi
 	xorl    %ebp, %ebp
 	ret
-relocate_new_kernel_end:
-
-	.globl relocate_new_kernel_size
-relocate_new_kernel_size:
-	.long relocate_new_kernel_end - relocate_new_kernel
diff --git a/include/asm-i386/kexec.h b/include/asm-i386/kexec.h
index 53f0e06672d..4dfc9f5ed03 100644
--- a/include/asm-i386/kexec.h
+++ b/include/asm-i386/kexec.h
@@ -1,6 +1,26 @@
 #ifndef _I386_KEXEC_H
 #define _I386_KEXEC_H
 
+#define PA_CONTROL_PAGE  0
+#define VA_CONTROL_PAGE  1
+#define PA_PGD           2
+#define VA_PGD           3
+#define PA_PTE_0         4
+#define VA_PTE_0         5
+#define PA_PTE_1         6
+#define VA_PTE_1         7
+#ifdef CONFIG_X86_PAE
+#define PA_PMD_0         8
+#define VA_PMD_0         9
+#define PA_PMD_1         10
+#define VA_PMD_1         11
+#define PAGES_NR         12
+#else
+#define PAGES_NR         8
+#endif
+
+#ifndef __ASSEMBLY__
+
 #include <asm/fixmap.h>
 #include <asm/ptrace.h>
 #include <asm/string.h>
@@ -72,5 +92,12 @@ static inline void crash_setup_regs(struct pt_regs *newregs,
                newregs->eip = (unsigned long)current_text_addr();
        }
 }
+asmlinkage NORET_TYPE void
+relocate_kernel(unsigned long indirection_page,
+		unsigned long control_page,
+		unsigned long start_address,
+		unsigned int has_pae) ATTRIB_NORET;
+
+#endif /* __ASSEMBLY__ */
 
 #endif /* _I386_KEXEC_H */