x86, kexec: x86_64: add kexec jump support for x86_64

Impact: New major feature

This patch add kexec jump support for x86_64. More information about
kexec jump can be found in corresponding x86_32 support patch.

Signed-off-by: Huang Ying <ying.huang@intel.com>
Signed-off-by: H. Peter Anvin <hpa@zytor.com>

arch/x86/Kconfig

@@ -1431,7 +1431,7 @@ config CRASH_DUMP
 config KEXEC_JUMP
 	bool "kexec jump (EXPERIMENTAL)"
 	depends on EXPERIMENTAL
-	depends on KEXEC && HIBERNATION && X86_32
+	depends on KEXEC && HIBERNATION
 	---help---
 	  Jump between original kernel and kexeced kernel and invoke
 	  code in physical address mode via KEXEC

arch/x86/include/asm/kexec.h

@@ -9,13 +9,13 @@
 # define PAGES_NR		4
 #else
 # define PA_CONTROL_PAGE	0
-# define PA_TABLE_PAGE		1
-# define PAGES_NR		2
+# define VA_CONTROL_PAGE	1
+# define PA_TABLE_PAGE		2
+# define PA_SWAP_PAGE		3
+# define PAGES_NR		4
 #endif
 
-#ifdef CONFIG_X86_32
 # define KEXEC_CONTROL_CODE_MAX_SIZE	2048
-#endif
 
 #ifndef __ASSEMBLY__
 
@@ -136,10 +136,11 @@ relocate_kernel(unsigned long indirection_page,
 		unsigned int has_pae,
 		unsigned int preserve_context);
 #else
-NORET_TYPE void
+unsigned long
 relocate_kernel(unsigned long indirection_page,
 		unsigned long page_list,
-		unsigned long start_address) ATTRIB_NORET;
+		unsigned long start_address,
+		unsigned int preserve_context);
 #endif
 
 #define ARCH_HAS_KIMAGE_ARCH

arch/x86/kernel/machine_kexec_64.c

@@ -13,6 +13,7 @@
 #include <linux/numa.h>
 #include <linux/ftrace.h>
 #include <linux/io.h>
+#include <linux/suspend.h>
 
 #include <asm/pgtable.h>
 #include <asm/tlbflush.h>
@@ -270,19 +271,43 @@ void machine_kexec(struct kimage *image)
 {
 	unsigned long page_list[PAGES_NR];
 	void *control_page;
+	int save_ftrace_enabled;
 
-	tracer_disable();
+#ifdef CONFIG_KEXEC_JUMP
+	if (kexec_image->preserve_context)
+		save_processor_state();
+#endif
+
+	save_ftrace_enabled = __ftrace_enabled_save();
 
 	/* Interrupts aren't acceptable while we reboot */
 	local_irq_disable();
 
+	if (image->preserve_context) {
+#ifdef CONFIG_X86_IO_APIC
+		/*
+		 * We need to put APICs in legacy mode so that we can
+		 * get timer interrupts in second kernel. kexec/kdump
+		 * paths already have calls to disable_IO_APIC() in
+		 * one form or other. kexec jump path also need
+		 * one.
+		 */
+		disable_IO_APIC();
+#endif
+	}
+
 	control_page = page_address(image->control_code_page) + PAGE_SIZE;
-	memcpy(control_page, relocate_kernel, PAGE_SIZE);
+	memcpy(control_page, relocate_kernel, KEXEC_CONTROL_CODE_MAX_SIZE);
 
 	page_list[PA_CONTROL_PAGE] = virt_to_phys(control_page);
+	page_list[VA_CONTROL_PAGE] = (unsigned long)control_page;
 	page_list[PA_TABLE_PAGE] =
 	  (unsigned long)__pa(page_address(image->control_code_page));
 
+	if (image->type == KEXEC_TYPE_DEFAULT)
+		page_list[PA_SWAP_PAGE] = (page_to_pfn(image->swap_page)
+						<< PAGE_SHIFT);
+
 	/*
 	 * The segment registers are funny things, they have both a
 	 * visible and an invisible part.  Whenever the visible part is
@@ -302,8 +327,17 @@ void machine_kexec(struct kimage *image)
 	set_idt(phys_to_virt(0), 0);
 
 	/* now call it */
-	relocate_kernel((unsigned long)image->head, (unsigned long)page_list,
-			image->start);
+	image->start = relocate_kernel((unsigned long)image->head,
+				       (unsigned long)page_list,
+				       image->start,
+				       image->preserve_context);
+
+#ifdef CONFIG_KEXEC_JUMP
+	if (kexec_image->preserve_context)
+		restore_processor_state();
+#endif
+
+	__ftrace_enabled_restore(save_ftrace_enabled);
 }
 
 void arch_crash_save_vmcoreinfo(void)

arch/x86/kernel/relocate_kernel_64.S

@@ -19,6 +19,24 @@
 #define PTR(x) (x << 3)
 #define PAGE_ATTR (_PAGE_PRESENT | _PAGE_RW | _PAGE_ACCESSED | _PAGE_DIRTY)
 
+/*
+ * control_page + KEXEC_CONTROL_CODE_MAX_SIZE
+ * ~ control_page + PAGE_SIZE are used as data storage and stack for
+ * jumping back
+ */
+#define DATA(offset)		(KEXEC_CONTROL_CODE_MAX_SIZE+(offset))
+
+/* Minimal CPU state */
+#define RSP			DATA(0x0)
+#define CR0			DATA(0x8)
+#define CR3			DATA(0x10)
+#define CR4			DATA(0x18)
+
+/* other data */
+#define CP_PA_TABLE_PAGE	DATA(0x20)
+#define CP_PA_SWAP_PAGE		DATA(0x28)
+#define CP_PA_BACKUP_PAGES_MAP	DATA(0x30)
+
 	.text
 	.align PAGE_SIZE
 	.code64
@@ -28,8 +46,27 @@ relocate_kernel:
 	 * %rdi indirection_page
 	 * %rsi page_list
 	 * %rdx start address
+	 * %rcx preserve_context
 	 */
 
+	/* Save the CPU context, used for jumping back */
+	pushq %rbx
+	pushq %rbp
+	pushq %r12
+	pushq %r13
+	pushq %r14
+	pushq %r15
+	pushf
+
+	movq	PTR(VA_CONTROL_PAGE)(%rsi), %r11
+	movq	%rsp, RSP(%r11)
+	movq	%cr0, %rax
+	movq	%rax, CR0(%r11)
+	movq	%cr3, %rax
+	movq	%rax, CR3(%r11)
+	movq	%cr4, %rax
+	movq	%rax, CR4(%r11)
+
 	/* zero out flags, and disable interrupts */
 	pushq $0
 	popfq
@@ -41,10 +78,18 @@ relocate_kernel:
 	movq	PTR(PA_CONTROL_PAGE)(%rsi), %r8
 
 	/* get physical address of page table now too */
-	movq	PTR(PA_TABLE_PAGE)(%rsi), %rcx
+	movq	PTR(PA_TABLE_PAGE)(%rsi), %r9
+
+	/* get physical address of swap page now */
+	movq	PTR(PA_SWAP_PAGE)(%rsi), %r10
+
+	/* save some information for jumping back */
+	movq	%r9, CP_PA_TABLE_PAGE(%r11)
+	movq	%r10, CP_PA_SWAP_PAGE(%r11)
+	movq	%rdi, CP_PA_BACKUP_PAGES_MAP(%r11)
 
 	/* Switch to the identity mapped page tables */
-	movq	%rcx, %cr3
+	movq	%r9, %cr3
 
 	/* setup a new stack at the end of the physical control page */
 	lea	PAGE_SIZE(%r8), %rsp
@@ -83,9 +128,87 @@ identity_mapped:
 1:
 
 	/* Flush the TLB (needed?) */
-	movq	%rcx, %cr3
+	movq	%r9, %cr3
+
+	movq	%rcx, %r11
+	call	swap_pages
+
+	/*
+	 * To be certain of avoiding problems with self-modifying code
+	 * I need to execute a serializing instruction here.
+	 * So I flush the TLB by reloading %cr3 here, it's handy,
+	 * and not processor dependent.
+	 */
+	movq	%cr3, %rax
+	movq	%rax, %cr3
+
+	/*
+	 * set all of the registers to known values
+	 * leave %rsp alone
+	 */
+
+	testq	%r11, %r11
+	jnz 1f
+	xorq	%rax, %rax
+	xorq	%rbx, %rbx
+	xorq    %rcx, %rcx
+	xorq    %rdx, %rdx
+	xorq    %rsi, %rsi
+	xorq    %rdi, %rdi
+	xorq    %rbp, %rbp
+	xorq	%r8,  %r8
+	xorq	%r9,  %r9
+	xorq	%r10, %r9
+	xorq	%r11, %r11
+	xorq	%r12, %r12
+	xorq	%r13, %r13
+	xorq	%r14, %r14
+	xorq	%r15, %r15
+
+	ret
+
+1:
+	popq	%rdx
+	leaq	PAGE_SIZE(%r10), %rsp
+	call	*%rdx
+
+	/* get the re-entry point of the peer system */
+	movq	0(%rsp), %rbp
+	call	1f
+1:
+	popq	%r8
+	subq	$(1b - relocate_kernel), %r8
+	movq	CP_PA_SWAP_PAGE(%r8), %r10
+	movq	CP_PA_BACKUP_PAGES_MAP(%r8), %rdi
+	movq	CP_PA_TABLE_PAGE(%r8), %rax
+	movq	%rax, %cr3
+	lea	PAGE_SIZE(%r8), %rsp
+	call	swap_pages
+	movq	$virtual_mapped, %rax
+	pushq	%rax
+	ret
+
+virtual_mapped:
+	movq	RSP(%r8), %rsp
+	movq	CR4(%r8), %rax
+	movq	%rax, %cr4
+	movq	CR3(%r8), %rax
+	movq	CR0(%r8), %r8
+	movq	%rax, %cr3
+	movq	%r8, %cr0
+	movq	%rbp, %rax
+
+	popf
+	popq	%r15
+	popq	%r14
+	popq	%r13
+	popq	%r12
+	popq	%rbp
+	popq	%rbx
+	ret
 
 	/* Do the copies */
+swap_pages:
 	movq	%rdi, %rcx 	/* Put the page_list in %rcx */
 	xorq	%rdi, %rdi
 	xorq	%rsi, %rsi
@@ -117,39 +240,27 @@ identity_mapped:
 	movq	%rcx,   %rsi  /* For ever source page do a copy */
 	andq	$0xfffffffffffff000, %rsi
 
+	movq	%rdi, %rdx
+	movq	%rsi, %rax
+
+	movq	%r10, %rdi
 	movq	$512,   %rcx
 	rep ; movsq
-	jmp	0b
-3:
 
-	/*
-	 * To be certain of avoiding problems with self-modifying code
-	 * I need to execute a serializing instruction here.
-	 * So I flush the TLB by reloading %cr3 here, it's handy,
-	 * and not processor dependent.
-	 */
-	movq	%cr3, %rax
-	movq	%rax, %cr3
-
-	/*
-	 * set all of the registers to known values
-	 * leave %rsp alone
-	 */
+	movq	%rax, %rdi
+	movq	%rdx, %rsi
+	movq	$512,   %rcx
+	rep ; movsq
 
-	xorq	%rax, %rax
-	xorq	%rbx, %rbx
-	xorq    %rcx, %rcx
-	xorq    %rdx, %rdx
-	xorq    %rsi, %rsi
-	xorq    %rdi, %rdi
-	xorq    %rbp, %rbp
-	xorq	%r8,  %r8
-	xorq	%r9,  %r9
-	xorq	%r10, %r9
-	xorq	%r11, %r11
-	xorq	%r12, %r12
-	xorq	%r13, %r13
-	xorq	%r14, %r14
-	xorq	%r15, %r15
+	movq	%rdx, %rdi
+	movq	%r10, %rsi
+	movq	$512,   %rcx
+	rep ; movsq
 
+	lea	PAGE_SIZE(%rax), %rsi
+	jmp	0b
+3:
 	ret
+
+	.globl kexec_control_code_size
+.set kexec_control_code_size, . - relocate_kernel

arch/x86/kernel/vmlinux_64.lds.S

@@ -275,3 +275,10 @@ ASSERT((_end - _text <= KERNEL_IMAGE_SIZE),
 ASSERT((per_cpu__irq_stack_union == 0),
         "irq_stack_union is not at start of per-cpu area");
 #endif
+
+#ifdef CONFIG_KEXEC
+#include <asm/kexec.h>
+
+ASSERT(kexec_control_code_size <= KEXEC_CONTROL_CODE_MAX_SIZE,
+       "kexec control code size is too big")
+#endif