[PATCH] kexec: x86 kexec core

This is the i386 implementation of kexec. Signed-off-by: Eric Biederman <ebiederm@xmission.com> Signed-off-by: Andrew Morton <akpm@osdl.org> Signed-off-by: Linus Torvalds <torvalds@osdl.org>

[PATCH] kexec: x86 kexec core
This is the i386 implementation of kexec. Signed-off-by: Eric Biederman <ebiederm@xmission.com> Signed-off-by: Andrew Morton <akpm@osdl.org> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
Eric W. Biederman · Linus Torvalds
1 parent dd2a13054f
Showing 7 changed files with 429 additions and 1 deletions Side-by-side Diff
arch/i386/Kconfig
arch/i386/kernel/Makefile
arch/i386/kernel/crash.c
arch/i386/kernel/machine_kexec.c
arch/i386/kernel/relocate_kernel.S
arch/i386/kernel/syscall_table.S
include/asm-i386/kexec.h
@@ -953,6 +953,23 @@
  
 	  Don't change this unless you know what you are doing.
  
+config KEXEC
+	bool "kexec system call (EXPERIMENTAL)"
+	depends on EXPERIMENTAL
+	help
+	  kexec is a system call that implements the ability to shutdown your
+	  current kernel, and to start another kernel.  It is like a reboot
+	  but it is indepedent of the system firmware.   And like a reboot
+	  you can start any kernel with it, not just Linux.
+
+	  The name comes from the similiarity to the exec system call.
+
+	  It is an ongoing process to be certain the hardware in a machine
+	  is properly shutdown, so do not be surprised if this code does not
+	  initially work for you.  It may help to enable device hotplugging
+	  support.  As of this writing the exact hardware interface is
+	  strongly in flux, so no good recommendation can be made.
+
 endmenu
  
  
@@ -24,6 +24,7 @@
 obj-$(CONFIG_X86_LOCAL_APIC)	+= apic.o nmi.o
 obj-$(CONFIG_X86_IO_APIC)	+= io_apic.o
 obj-$(CONFIG_X86_REBOOTFIXUPS)	+= reboot_fixups.o
+obj-$(CONFIG_KEXEC)		+= machine_kexec.o relocate_kernel.o crash.o
 obj-$(CONFIG_X86_NUMAQ)		+= numaq.o
 obj-$(CONFIG_X86_SUMMIT_NUMA)	+= summit.o
 obj-$(CONFIG_KPROBES)		+= kprobes.o
+/*
+ * Architecture specific (i386) functions for kexec based crash dumps.
+ *
+ * Created by: Hariprasad Nellitheertha (hari@in.ibm.com)
+ *
+ * Copyright (C) IBM Corporation, 2004. All rights reserved.
+ *
+ */
+
+#include <linux/init.h>
+#include <linux/types.h>
+#include <linux/kernel.h>
+#include <linux/smp.h>
+#include <linux/irq.h>
+#include <linux/reboot.h>
+#include <linux/kexec.h>
+#include <linux/irq.h>
+#include <linux/delay.h>
+#include <linux/elf.h>
+#include <linux/elfcore.h>
+
+#include <asm/processor.h>
+#include <asm/hardirq.h>
+#include <asm/nmi.h>
+#include <asm/hw_irq.h>
+
+#define MAX_NOTE_BYTES 1024
+typedef u32 note_buf_t[MAX_NOTE_BYTES/4];
+
+note_buf_t crash_notes[NR_CPUS];
+
+void machine_crash_shutdown(void)
+{
+	/* This function is only called after the system
+	 * has paniced or is otherwise in a critical state.
+	 * The minimum amount of code to allow a kexec'd kernel
+	 * to run successfully needs to happen here.
+	 *
+	 * In practice this means shooting down the other cpus in
+	 * an SMP system.
+	 */
+}
+/*
+ * machine_kexec.c - handle transition of Linux booting another kernel
+ * Copyright (C) 2002-2005 Eric Biederman  <ebiederm@xmission.com>
+ *
+ * This source code is licensed under the GNU General Public License,
+ * Version 2.  See the file COPYING for more details.
+ */
+
+#include <linux/mm.h>
+#include <linux/kexec.h>
+#include <linux/delay.h>
+#include <asm/pgtable.h>
+#include <asm/pgalloc.h>
+#include <asm/tlbflush.h>
+#include <asm/mmu_context.h>
+#include <asm/io.h>
+#include <asm/apic.h>
+#include <asm/cpufeature.h>
+
+static inline unsigned long read_cr3(void)
+{
+	unsigned long cr3;
+	asm volatile("movl %%cr3,%0": "=r"(cr3));
+	return cr3;
+}
+
+#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);
+}
+#endif
+
+
+static void set_idt(void *newidt, __u16 limit)
+{
+	unsigned char curidt[6];
+
+	/* ia32 supports unaliged loads & stores */
+	(*(__u16 *)(curidt)) = limit;
+	(*(__u32 *)(curidt +2)) = (unsigned long)(newidt);
+
+	__asm__ __volatile__ (
+		"lidt %0\n"
+		: "=m" (curidt)
+		);
+};
+
+
+static void set_gdt(void *newgdt, __u16 limit)
+{
+	unsigned char curgdt[6];
+
+	/* ia32 supports unaligned loads & stores */
+	(*(__u16 *)(curgdt)) = limit;
+	(*(__u32 *)(curgdt +2)) = (unsigned long)(newgdt);
+
+	__asm__ __volatile__ (
+		"lgdt %0\n"
+		: "=m" (curgdt)
+		);
+};
+
+static void load_segments(void)
+{
+#define __STR(X) #X
+#define STR(X) __STR(X)
+
+	__asm__ __volatile__ (
+		"\tljmp $"STR(__KERNEL_CS)",$1f\n"
+		"\t1:\n"
+		"\tmovl $"STR(__KERNEL_DS)",%eax\n"
+		"\tmovl %eax,%ds\n"
+		"\tmovl %eax,%es\n"
+		"\tmovl %eax,%fs\n"
+		"\tmovl %eax,%gs\n"
+		"\tmovl %eax,%ss\n"
+		);
+#undef STR
+#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;
+
+const extern unsigned char relocate_new_kernel[];
+extern void relocate_new_kernel_end(void);
+const extern unsigned int relocate_new_kernel_size;
+
+/*
+ * A architecture hook called to validate the
+ * proposed image and prepare the control pages
+ * as needed.  The pages for KEXEC_CONTROL_CODE_SIZE
+ * have been allocated, but the segments have yet
+ * been copied into the kernel.
+ *
+ * Do what every setup is needed on image and the
+ * reboot code buffer to allow us to avoid allocations
+ * later.
+ *
+ * Currently nothing.
+ */
+int machine_kexec_prepare(struct kimage *image)
+{
+	return 0;
+}
+
+/*
+ * Undo anything leftover by machine_kexec_prepare
+ * when an image is freed.
+ */
+void machine_kexec_cleanup(struct kimage *image)
+{
+}
+
+/*
+ * Do not allocate memory (or fail in any way) in machine_kexec().
+ * We are past the point of no return, committed to rebooting now.
+ */
+NORET_TYPE void machine_kexec(struct kimage *image)
+{
+	unsigned long page_list;
+	unsigned long reboot_code_buffer;
+	relocate_new_kernel_t rnk;
+
+	/* 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);
+
+	/* The segment registers are funny things, they are
+	 * automatically loaded from a table, in memory wherever you
+	 * set them to a specific selector, but this table is never
+	 * accessed again you set the segment to a different selector.
+	 *
+	 * The more common model is are caches where the behide
+	 * the scenes work is done, but is also dropped at arbitrary
+	 * times.
+	 *
+	 * I take advantage of this here by force loading the
+	 * segments, before I zap the gdt with an invalid value.
+	 */
+	load_segments();
+	/* The gdt & idt are now invalid.
+	 * If you want to load them you must set up your own idt & gdt.
+	 */
+	set_gdt(phys_to_virt(0),0);
+	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.S - put the kernel image in place to boot
+ * Copyright (C) 2002-2004 Eric Biederman  <ebiederm@xmission.com>
+ *
+ * This source code is licensed under the GNU General Public License,
+ * Version 2.  See the file COPYING for more details.
+ */
+
+#include <linux/linkage.h>
+
+	/*
+	 * 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  12(%esp), %edx /* start address */
+	movl  16(%esp), %ecx /* cpu_has_pae */
+
+	/* zero out flags, and disable interrupts */
+	pushl $0
+	popfl
+
+	/* set a new stack at the bottom of our page... */
+	lea   4096(%ebp), %esp
+
+	/* store the parameters back on the stack */
+	pushl   %edx /* store the start address */
+
+	/* Set cr0 to a known state:
+	 * 31 0 == Paging disabled
+	 * 18 0 == Alignment check disabled
+	 * 16 0 == Write protect disabled
+	 * 3  0 == No task switch
+	 * 2  0 == Don't do FP software emulation.
+	 * 0  1 == Proctected mode enabled
+	 */
+	movl	%cr0, %eax
+	andl	$~((1<<31)|(1<<18)|(1<<16)|(1<<3)|(1<<2)), %eax
+	orl	$(1<<0), %eax
+	movl	%eax, %cr0
+
+	/* clear cr4 if applicable */
+	testl	%ecx, %ecx
+	jz	1f
+	/* Set cr4 to a known state:
+	 * Setting everything to zero seems safe.
+	 */
+	movl	%cr4, %eax
+	andl	$0, %eax
+	movl	%eax, %cr4
+
+	jmp 1f
+1:
+
+	/* Flush the TLB (needed?) */
+	xorl	%eax, %eax
+	movl	%eax, %cr3
+
+	/* Do the copies */
+	movl	%ebx, %ecx
+	jmp	1f
+
+0:	/* top, read another word from the indirection page */
+	movl	(%ebx), %ecx
+	addl	$4, %ebx
+1:
+	testl	$0x1,   %ecx  /* is it a destination page */
+	jz	2f
+	movl	%ecx,	%edi
+	andl	$0xfffff000, %edi
+	jmp     0b
+2:
+	testl	$0x2,	%ecx  /* is it an indirection page */
+	jz	2f
+	movl	%ecx,	%ebx
+	andl	$0xfffff000, %ebx
+	jmp     0b
+2:
+	testl   $0x4,   %ecx /* is it the done indicator */
+	jz      2f
+	jmp     3f
+2:
+	testl   $0x8,   %ecx /* is it the source indicator */
+	jz      0b	     /* Ignore it otherwise */
+	movl    %ecx,   %esi /* For every source page do a copy */
+	andl    $0xfffff000, %esi
+
+	movl    $1024, %ecx
+	rep ; movsl
+	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, it's handy, and not processor dependent.
+	 */
+	xorl	%eax, %eax
+	movl	%eax, %cr3
+
+	/* set all of the registers to known values */
+	/* leave %esp alone */
+
+	xorl	%eax, %eax
+	xorl	%ebx, %ebx
+	xorl    %ecx, %ecx
+	xorl    %edx, %edx
+	xorl    %esi, %esi
+	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
@@ -283,7 +283,7 @@
 	.long sys_mq_timedreceive	/* 280 */
 	.long sys_mq_notify
 	.long sys_mq_getsetattr
-	.long sys_ni_syscall		/* reserved for kexec */
+	.long sys_kexec_load
 	.long sys_waitid
 	.long sys_ni_syscall		/* 285 */ /* available */
 	.long sys_add_key
+#ifndef _I386_KEXEC_H
+#define _I386_KEXEC_H
+
+#include <asm/fixmap.h>
+
+/*
+ * KEXEC_SOURCE_MEMORY_LIMIT maximum page get_free_page can return.
+ * I.e. Maximum page that is mapped directly into kernel memory,
+ * and kmap is not required.
+ *
+ * Someone correct me if FIXADDR_START - PAGEOFFSET is not the correct
+ * calculation for the amount of memory directly mappable into the
+ * kernel memory space.
+ */
+
+/* Maximum physical address we can use pages from */
+#define KEXEC_SOURCE_MEMORY_LIMIT (-1UL)
+/* Maximum address we can reach in physical address mode */
+#define KEXEC_DESTINATION_MEMORY_LIMIT (-1UL)
+/* Maximum address we can use for the control code buffer */
+#define KEXEC_CONTROL_MEMORY_LIMIT TASK_SIZE
+
+#define KEXEC_CONTROL_CODE_SIZE	4096
+
+/* The native architecture */
+#define KEXEC_ARCH KEXEC_ARCH_386
+
+#endif /* _I386_KEXEC_H */
...	...	@@ -953,6 +953,23 @@
953	953
954	954	Don't change this unless you know what you are doing.
955	955
	956	+config KEXEC
	957	+ bool "kexec system call (EXPERIMENTAL)"
	958	+ depends on EXPERIMENTAL
	959	+ help
	960	+ kexec is a system call that implements the ability to shutdown your
	961	+ current kernel, and to start another kernel. It is like a reboot
	962	+ but it is indepedent of the system firmware. And like a reboot
	963	+ you can start any kernel with it, not just Linux.
	964	+
	965	+ The name comes from the similiarity to the exec system call.
	966	+
	967	+ It is an ongoing process to be certain the hardware in a machine
	968	+ is properly shutdown, so do not be surprised if this code does not
	969	+ initially work for you. It may help to enable device hotplugging
	970	+ support. As of this writing the exact hardware interface is
	971	+ strongly in flux, so no good recommendation can be made.
	972	+
956	973	endmenu
957	974
958	975
...	...	@@ -24,6 +24,7 @@
24	24	obj-$(CONFIG_X86_LOCAL_APIC) += apic.o nmi.o
25	25	obj-$(CONFIG_X86_IO_APIC) += io_apic.o
26	26	obj-$(CONFIG_X86_REBOOTFIXUPS) += reboot_fixups.o
	27	+obj-$(CONFIG_KEXEC) += machine_kexec.o relocate_kernel.o crash.o
27	28	obj-$(CONFIG_X86_NUMAQ) += numaq.o
28	29	obj-$(CONFIG_X86_SUMMIT_NUMA) += summit.o
29	30	obj-$(CONFIG_KPROBES) += kprobes.o
	1	+/*
	2	+ * Architecture specific (i386) functions for kexec based crash dumps.
	3	+ *
	4	+ * Created by: Hariprasad Nellitheertha (hari@in.ibm.com)
	5	+ *
	6	+ * Copyright (C) IBM Corporation, 2004. All rights reserved.
	7	+ *
	8	+ */
	9	+
	10	+#include <linux/init.h>
	11	+#include <linux/types.h>
	12	+#include <linux/kernel.h>
	13	+#include <linux/smp.h>
	14	+#include <linux/irq.h>
	15	+#include <linux/reboot.h>
	16	+#include <linux/kexec.h>
	17	+#include <linux/irq.h>
	18	+#include <linux/delay.h>
	19	+#include <linux/elf.h>
	20	+#include <linux/elfcore.h>
	21	+
	22	+#include <asm/processor.h>
	23	+#include <asm/hardirq.h>
	24	+#include <asm/nmi.h>
	25	+#include <asm/hw_irq.h>
	26	+
	27	+#define MAX_NOTE_BYTES 1024
	28	+typedef u32 note_buf_t[MAX_NOTE_BYTES/4];
	29	+
	30	+note_buf_t crash_notes[NR_CPUS];
	31	+
	32	+void machine_crash_shutdown(void)
	33	+{
	34	+ /* This function is only called after the system
	35	+ * has paniced or is otherwise in a critical state.
	36	+ * The minimum amount of code to allow a kexec'd kernel
	37	+ * to run successfully needs to happen here.
	38	+ *
	39	+ * In practice this means shooting down the other cpus in
	40	+ * an SMP system.
	41	+ */
	42	+}
	1	+/*
	2	+ * machine_kexec.c - handle transition of Linux booting another kernel
	3	+ * Copyright (C) 2002-2005 Eric Biederman <ebiederm@xmission.com>
	4	+ *
	5	+ * This source code is licensed under the GNU General Public License,
	6	+ * Version 2. See the file COPYING for more details.
	7	+ */
	8	+
	9	+#include <linux/mm.h>
	10	+#include <linux/kexec.h>
	11	+#include <linux/delay.h>
	12	+#include <asm/pgtable.h>
	13	+#include <asm/pgalloc.h>
	14	+#include <asm/tlbflush.h>
	15	+#include <asm/mmu_context.h>
	16	+#include <asm/io.h>
	17	+#include <asm/apic.h>
	18	+#include <asm/cpufeature.h>
	19	+
	20	+static inline unsigned long read_cr3(void)
	21	+{
	22	+ unsigned long cr3;
	23	+ asm volatile("movl %%cr3,%0": "=r"(cr3));
	24	+ return cr3;
	25	+}
	26	+
	27	+#define PAGE_ALIGNED __attribute__ ((__aligned__(PAGE_SIZE)))
	28	+
	29	+#define L0_ATTR (_PAGE_PRESENT \| _PAGE_RW \| _PAGE_ACCESSED \| _PAGE_DIRTY)
	30	+#define L1_ATTR (_PAGE_PRESENT \| _PAGE_RW \| _PAGE_ACCESSED \| _PAGE_DIRTY)
	31	+#define L2_ATTR (_PAGE_PRESENT)
	32	+
	33	+#define LEVEL0_SIZE (1UL << 12UL)
	34	+
	35	+#ifndef CONFIG_X86_PAE
	36	+#define LEVEL1_SIZE (1UL << 22UL)
	37	+static u32 pgtable_level1[1024] PAGE_ALIGNED;
	38	+
	39	+static void identity_map_page(unsigned long address)
	40	+{
	41	+ unsigned long level1_index, level2_index;
	42	+ u32 *pgtable_level2;
	43	+
	44	+ /* Find the current page table */
	45	+ pgtable_level2 = __va(read_cr3());
	46	+
	47	+ /* Find the indexes of the physical address to identity map */
	48	+ level1_index = (address % LEVEL1_SIZE)/LEVEL0_SIZE;
	49	+ level2_index = address / LEVEL1_SIZE;
	50	+
	51	+ /* Identity map the page table entry */
	52	+ pgtable_level1[level1_index] = address \| L0_ATTR;
	53	+ pgtable_level2[level2_index] = __pa(pgtable_level1) \| L1_ATTR;
	54	+
	55	+ /* Flush the tlb so the new mapping takes effect.
	56	+ * Global tlb entries are not flushed but that is not an issue.
	57	+ */
	58	+ load_cr3(pgtable_level2);
	59	+}
	60	+
	61	+#else
	62	+#define LEVEL1_SIZE (1UL << 21UL)
	63	+#define LEVEL2_SIZE (1UL << 30UL)
	64	+static u64 pgtable_level1[512] PAGE_ALIGNED;
	65	+static u64 pgtable_level2[512] PAGE_ALIGNED;
	66	+
	67	+static void identity_map_page(unsigned long address)
	68	+{
	69	+ unsigned long level1_index, level2_index, level3_index;
	70	+ u64 *pgtable_level3;
	71	+
	72	+ /* Find the current page table */
	73	+ pgtable_level3 = __va(read_cr3());
	74	+
	75	+ /* Find the indexes of the physical address to identity map */
	76	+ level1_index = (address % LEVEL1_SIZE)/LEVEL0_SIZE;
	77	+ level2_index = (address % LEVEL2_SIZE)/LEVEL1_SIZE;
	78	+ level3_index = address / LEVEL2_SIZE;
	79	+
	80	+ /* Identity map the page table entry */
	81	+ pgtable_level1[level1_index] = address \| L0_ATTR;
	82	+ pgtable_level2[level2_index] = __pa(pgtable_level1) \| L1_ATTR;
	83	+ set_64bit(&pgtable_level3[level3_index], __pa(pgtable_level2) \| L2_ATTR);
	84	+
	85	+ /* Flush the tlb so the new mapping takes effect.
	86	+ * Global tlb entries are not flushed but that is not an issue.
	87	+ */
	88	+ load_cr3(pgtable_level3);
	89	+}
	90	+#endif
	91	+
	92	+
	93	+static void set_idt(void *newidt, __u16 limit)
	94	+{
	95	+ unsigned char curidt[6];
	96	+
	97	+ /* ia32 supports unaliged loads & stores */
	98	+ ((__u16 )(curidt)) = limit;
	99	+ ((__u32 )(curidt +2)) = (unsigned long)(newidt);
	100	+
	101	+ __asm__ __volatile__ (
	102	+ "lidt %0\n"
	103	+ : "=m" (curidt)
	104	+ );
	105	+};
	106	+
	107	+
	108	+static void set_gdt(void *newgdt, __u16 limit)
	109	+{
	110	+ unsigned char curgdt[6];
	111	+
	112	+ /* ia32 supports unaligned loads & stores */
	113	+ ((__u16 )(curgdt)) = limit;
	114	+ ((__u32 )(curgdt +2)) = (unsigned long)(newgdt);
	115	+
	116	+ __asm__ __volatile__ (
	117	+ "lgdt %0\n"
	118	+ : "=m" (curgdt)
	119	+ );
	120	+};
	121	+
	122	+static void load_segments(void)
	123	+{
	124	+#define __STR(X) #X
	125	+#define STR(X) __STR(X)
	126	+
	127	+ __asm__ __volatile__ (
	128	+ "\tljmp $"STR(__KERNEL_CS)",$1f\n"
	129	+ "\t1:\n"
	130	+ "\tmovl $"STR(__KERNEL_DS)",%eax\n"
	131	+ "\tmovl %eax,%ds\n"
	132	+ "\tmovl %eax,%es\n"
	133	+ "\tmovl %eax,%fs\n"
	134	+ "\tmovl %eax,%gs\n"
	135	+ "\tmovl %eax,%ss\n"
	136	+ );
	137	+#undef STR
	138	+#undef __STR
	139	+}
	140	+
	141	+typedef asmlinkage NORET_TYPE void (*relocate_new_kernel_t)(
	142	+ unsigned long indirection_page, unsigned long reboot_code_buffer,
	143	+ unsigned long start_address, unsigned int has_pae) ATTRIB_NORET;
	144	+
	145	+const extern unsigned char relocate_new_kernel[];
	146	+extern void relocate_new_kernel_end(void);
	147	+const extern unsigned int relocate_new_kernel_size;
	148	+
	149	+/*
	150	+ * A architecture hook called to validate the
	151	+ * proposed image and prepare the control pages
	152	+ * as needed. The pages for KEXEC_CONTROL_CODE_SIZE
	153	+ * have been allocated, but the segments have yet
	154	+ * been copied into the kernel.
	155	+ *
	156	+ * Do what every setup is needed on image and the
	157	+ * reboot code buffer to allow us to avoid allocations
	158	+ * later.
	159	+ *
	160	+ * Currently nothing.
	161	+ */
	162	+int machine_kexec_prepare(struct kimage *image)
	163	+{
	164	+ return 0;
	165	+}
	166	+
	167	+/*
	168	+ * Undo anything leftover by machine_kexec_prepare
	169	+ * when an image is freed.
	170	+ */
	171	+void machine_kexec_cleanup(struct kimage *image)
	172	+{
	173	+}
	174	+
	175	+/*
	176	+ * Do not allocate memory (or fail in any way) in machine_kexec().
	177	+ * We are past the point of no return, committed to rebooting now.
	178	+ */
	179	+NORET_TYPE void machine_kexec(struct kimage *image)
	180	+{
	181	+ unsigned long page_list;
	182	+ unsigned long reboot_code_buffer;
	183	+ relocate_new_kernel_t rnk;
	184	+
	185	+ /* Interrupts aren't acceptable while we reboot */
	186	+ local_irq_disable();
	187	+
	188	+ /* Compute some offsets */
	189	+ reboot_code_buffer = page_to_pfn(image->control_code_page) << PAGE_SHIFT;
	190	+ page_list = image->head;
	191	+
	192	+ /* Set up an identity mapping for the reboot_code_buffer */
	193	+ identity_map_page(reboot_code_buffer);
	194	+
	195	+ /* copy it out */
	196	+ memcpy((void *)reboot_code_buffer, relocate_new_kernel, relocate_new_kernel_size);
	197	+
	198	+ /* The segment registers are funny things, they are
	199	+ * automatically loaded from a table, in memory wherever you
	200	+ * set them to a specific selector, but this table is never
	201	+ * accessed again you set the segment to a different selector.
	202	+ *
	203	+ * The more common model is are caches where the behide
	204	+ * the scenes work is done, but is also dropped at arbitrary
	205	+ * times.
	206	+ *
	207	+ * I take advantage of this here by force loading the
	208	+ * segments, before I zap the gdt with an invalid value.
	209	+ */
	210	+ load_segments();
	211	+ /* The gdt & idt are now invalid.
	212	+ * If you want to load them you must set up your own idt & gdt.
	213	+ */
	214	+ set_gdt(phys_to_virt(0),0);
	215	+ set_idt(phys_to_virt(0),0);
	216	+
	217	+ /* now call it */
	218	+ rnk = (relocate_new_kernel_t) reboot_code_buffer;
	219	+ (*rnk)(page_list, reboot_code_buffer, image->start, cpu_has_pae);
	220	+}
	1	+/*
	2	+ * relocate_kernel.S - put the kernel image in place to boot
	3	+ * Copyright (C) 2002-2004 Eric Biederman <ebiederm@xmission.com>
	4	+ *
	5	+ * This source code is licensed under the GNU General Public License,
	6	+ * Version 2. See the file COPYING for more details.
	7	+ */
	8	+
	9	+#include <linux/linkage.h>
	10	+
	11	+ /*
	12	+ * Must be relocatable PIC code callable as a C function, that once
	13	+ * it starts can not use the previous processes stack.
	14	+ */
	15	+ .globl relocate_new_kernel
	16	+relocate_new_kernel:
	17	+ /* read the arguments and say goodbye to the stack */
	18	+ movl 4(%esp), %ebx /* page_list */
	19	+ movl 8(%esp), %ebp /* reboot_code_buffer */
	20	+ movl 12(%esp), %edx /* start address */
	21	+ movl 16(%esp), %ecx /* cpu_has_pae */
	22	+
	23	+ /* zero out flags, and disable interrupts */
	24	+ pushl $0
	25	+ popfl
	26	+
	27	+ /* set a new stack at the bottom of our page... */
	28	+ lea 4096(%ebp), %esp
	29	+
	30	+ /* store the parameters back on the stack */
	31	+ pushl %edx /* store the start address */
	32	+
	33	+ /* Set cr0 to a known state:
	34	+ * 31 0 == Paging disabled
	35	+ * 18 0 == Alignment check disabled
	36	+ * 16 0 == Write protect disabled
	37	+ * 3 0 == No task switch
	38	+ * 2 0 == Don't do FP software emulation.
	39	+ * 0 1 == Proctected mode enabled
	40	+ */
	41	+ movl %cr0, %eax
	42	+ andl $~((1<<31)\|(1<<18)\|(1<<16)\|(1<<3)\|(1<<2)), %eax
	43	+ orl $(1<<0), %eax
	44	+ movl %eax, %cr0
	45	+
	46	+ /* clear cr4 if applicable */
	47	+ testl %ecx, %ecx
	48	+ jz 1f
	49	+ /* Set cr4 to a known state:
	50	+ * Setting everything to zero seems safe.
	51	+ */
	52	+ movl %cr4, %eax
	53	+ andl $0, %eax
	54	+ movl %eax, %cr4
	55	+
	56	+ jmp 1f
	57	+1:
	58	+
	59	+ /* Flush the TLB (needed?) */
	60	+ xorl %eax, %eax
	61	+ movl %eax, %cr3
	62	+
	63	+ /* Do the copies */
	64	+ movl %ebx, %ecx
	65	+ jmp 1f
	66	+
	67	+0: /* top, read another word from the indirection page */
	68	+ movl (%ebx), %ecx
	69	+ addl $4, %ebx
	70	+1:
	71	+ testl $0x1, %ecx /* is it a destination page */
	72	+ jz 2f
	73	+ movl %ecx, %edi
	74	+ andl $0xfffff000, %edi
	75	+ jmp 0b
	76	+2:
	77	+ testl $0x2, %ecx /* is it an indirection page */
	78	+ jz 2f
	79	+ movl %ecx, %ebx
	80	+ andl $0xfffff000, %ebx
	81	+ jmp 0b
	82	+2:
	83	+ testl $0x4, %ecx /* is it the done indicator */
	84	+ jz 2f
	85	+ jmp 3f
	86	+2:
	87	+ testl $0x8, %ecx /* is it the source indicator */
	88	+ jz 0b /* Ignore it otherwise */
	89	+ movl %ecx, %esi /* For every source page do a copy */
	90	+ andl $0xfffff000, %esi
	91	+
	92	+ movl $1024, %ecx
	93	+ rep ; movsl
	94	+ jmp 0b
	95	+
	96	+3:
	97	+
	98	+ /* To be certain of avoiding problems with self-modifying code
	99	+ * I need to execute a serializing instruction here.
	100	+ * So I flush the TLB, it's handy, and not processor dependent.
	101	+ */
	102	+ xorl %eax, %eax
	103	+ movl %eax, %cr3
	104	+
	105	+ /* set all of the registers to known values */
	106	+ /* leave %esp alone */
	107	+
	108	+ xorl %eax, %eax
	109	+ xorl %ebx, %ebx
	110	+ xorl %ecx, %ecx
	111	+ xorl %edx, %edx
	112	+ xorl %esi, %esi
	113	+ xorl %edi, %edi
	114	+ xorl %ebp, %ebp
	115	+ ret
	116	+relocate_new_kernel_end:
	117	+
	118	+ .globl relocate_new_kernel_size
	119	+relocate_new_kernel_size:
	120	+ .long relocate_new_kernel_end - relocate_new_kernel
...	...	@@ -283,7 +283,7 @@
283	283	.long sys_mq_timedreceive /* 280 */
284	284	.long sys_mq_notify
285	285	.long sys_mq_getsetattr
286		- .long sys_ni_syscall /* reserved for kexec */
	286	+ .long sys_kexec_load
287	287	.long sys_waitid
288	288	.long sys_ni_syscall /* 285 / / available */
289	289	.long sys_add_key
	1	+#ifndef _I386_KEXEC_H
	2	+#define _I386_KEXEC_H
	3	+
	4	+#include <asm/fixmap.h>
	5	+
	6	+/*
	7	+ * KEXEC_SOURCE_MEMORY_LIMIT maximum page get_free_page can return.
	8	+ * I.e. Maximum page that is mapped directly into kernel memory,
	9	+ * and kmap is not required.
	10	+ *
	11	+ * Someone correct me if FIXADDR_START - PAGEOFFSET is not the correct
	12	+ * calculation for the amount of memory directly mappable into the
	13	+ * kernel memory space.
	14	+ */
	15	+
	16	+/* Maximum physical address we can use pages from */
	17	+#define KEXEC_SOURCE_MEMORY_LIMIT (-1UL)
	18	+/* Maximum address we can reach in physical address mode */
	19	+#define KEXEC_DESTINATION_MEMORY_LIMIT (-1UL)
	20	+/* Maximum address we can use for the control code buffer */
	21	+#define KEXEC_CONTROL_MEMORY_LIMIT TASK_SIZE
	22	+
	23	+#define KEXEC_CONTROL_CODE_SIZE 4096
	24	+
	25	+/* The native architecture */
	26	+#define KEXEC_ARCH KEXEC_ARCH_386
	27	+
	28	+#endif /* _I386_KEXEC_H */