From: Eric W. Biederman Date: Sat, 25 Jun 2005 21:58:02 +0000 (-0700) Subject: [PATCH] kexec: x86_64 kexec implementation X-Git-Tag: v2.6.13-rc1~68^2~108 X-Git-Url: https://err.no/cgi-bin/gitweb.cgi?a=commitdiff_plain;h=5234f5eb04abbbfa306ccfbc2ccbb6e73f515b15;p=linux-2.6 [PATCH] kexec: x86_64 kexec implementation This is the x86_64 implementation of machine kexec. 32bit compatibility support has been implemented, and machine_kexec has been enhanced to not care about the changing internal kernel paget table structures. From: Alexander Nyberg build fix Signed-off-by: Eric Biederman Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- diff --git a/arch/x86_64/Kconfig b/arch/x86_64/Kconfig index 56c02cf639..d09437b5c4 100644 --- a/arch/x86_64/Kconfig +++ b/arch/x86_64/Kconfig @@ -380,6 +380,23 @@ config PHYSICAL_START 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. + config SECCOMP bool "Enable seccomp to safely compute untrusted bytecode" depends on PROC_FS diff --git a/arch/x86_64/ia32/ia32entry.S b/arch/x86_64/ia32/ia32entry.S index f3ca0db85b..cc935427d5 100644 --- a/arch/x86_64/ia32/ia32entry.S +++ b/arch/x86_64/ia32/ia32entry.S @@ -589,7 +589,7 @@ ia32_sys_call_table: .quad compat_sys_mq_timedreceive /* 280 */ .quad compat_sys_mq_notify .quad compat_sys_mq_getsetattr - .quad quiet_ni_syscall /* reserved for kexec */ + .quad compat_sys_kexec_load /* reserved for kexec */ .quad compat_sys_waitid .quad quiet_ni_syscall /* sys_altroot */ .quad sys_add_key diff --git a/arch/x86_64/kernel/Makefile b/arch/x86_64/kernel/Makefile index 5ca4a4598f..48f9e2c19c 100644 --- a/arch/x86_64/kernel/Makefile +++ b/arch/x86_64/kernel/Makefile @@ -20,6 +20,7 @@ obj-$(CONFIG_SMP) += smp.o smpboot.o trampoline.o obj-$(CONFIG_X86_LOCAL_APIC) += apic.o nmi.o obj-$(CONFIG_X86_IO_APIC) += io_apic.o mpparse.o \ genapic.o genapic_cluster.o genapic_flat.o +obj-$(CONFIG_KEXEC) += machine_kexec.o relocate_kernel.o crash.o obj-$(CONFIG_PM) += suspend.o obj-$(CONFIG_SOFTWARE_SUSPEND) += suspend_asm.o obj-$(CONFIG_CPU_FREQ) += cpufreq/ diff --git a/arch/x86_64/kernel/crash.c b/arch/x86_64/kernel/crash.c new file mode 100644 index 0000000000..7caf8a49d0 --- /dev/null +++ b/arch/x86_64/kernel/crash.c @@ -0,0 +1,38 @@ +/* + * Architecture specific (x86_64) functions for kexec based crash dumps. + * + * Created by: Hariprasad Nellitheertha (hari@in.ibm.com) + * + * Copyright (C) IBM Corporation, 2004. All rights reserved. + * + */ + +#include +#include +#include +#include +#include +#include +#include + +#include +#include +#include +#include + +#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. + */ +} diff --git a/arch/x86_64/kernel/machine_kexec.c b/arch/x86_64/kernel/machine_kexec.c new file mode 100644 index 0000000000..200b5993f8 --- /dev/null +++ b/arch/x86_64/kernel/machine_kexec.c @@ -0,0 +1,245 @@ +/* + * machine_kexec.c - handle transition of Linux booting another kernel + * Copyright (C) 2002-2005 Eric Biederman + * + * This source code is licensed under the GNU General Public License, + * Version 2. See the file COPYING for more details. + */ + +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include + +#define LEVEL0_SIZE (1UL << 12UL) +#define LEVEL1_SIZE (1UL << 21UL) +#define LEVEL2_SIZE (1UL << 30UL) +#define LEVEL3_SIZE (1UL << 39UL) +#define LEVEL4_SIZE (1UL << 48UL) + +#define L0_ATTR (_PAGE_PRESENT | _PAGE_RW | _PAGE_ACCESSED | _PAGE_DIRTY) +#define L1_ATTR (_PAGE_PRESENT | _PAGE_RW | _PAGE_ACCESSED | _PAGE_DIRTY | _PAGE_PSE) +#define L2_ATTR (_PAGE_PRESENT | _PAGE_RW | _PAGE_ACCESSED | _PAGE_DIRTY) +#define L3_ATTR (_PAGE_PRESENT | _PAGE_RW | _PAGE_ACCESSED | _PAGE_DIRTY) + +static void init_level2_page( + u64 *level2p, unsigned long addr) +{ + unsigned long end_addr; + addr &= PAGE_MASK; + end_addr = addr + LEVEL2_SIZE; + while(addr < end_addr) { + *(level2p++) = addr | L1_ATTR; + addr += LEVEL1_SIZE; + } +} + +static int init_level3_page(struct kimage *image, + u64 *level3p, unsigned long addr, unsigned long last_addr) +{ + unsigned long end_addr; + int result; + result = 0; + addr &= PAGE_MASK; + end_addr = addr + LEVEL3_SIZE; + while((addr < last_addr) && (addr < end_addr)) { + struct page *page; + u64 *level2p; + page = kimage_alloc_control_pages(image, 0); + if (!page) { + result = -ENOMEM; + goto out; + } + level2p = (u64 *)page_address(page); + init_level2_page(level2p, addr); + *(level3p++) = __pa(level2p) | L2_ATTR; + addr += LEVEL2_SIZE; + } + /* clear the unused entries */ + while(addr < end_addr) { + *(level3p++) = 0; + addr += LEVEL2_SIZE; + } +out: + return result; +} + + +static int init_level4_page(struct kimage *image, + u64 *level4p, unsigned long addr, unsigned long last_addr) +{ + unsigned long end_addr; + int result; + result = 0; + addr &= PAGE_MASK; + end_addr = addr + LEVEL4_SIZE; + while((addr < last_addr) && (addr < end_addr)) { + struct page *page; + u64 *level3p; + page = kimage_alloc_control_pages(image, 0); + if (!page) { + result = -ENOMEM; + goto out; + } + level3p = (u64 *)page_address(page); + result = init_level3_page(image, level3p, addr, last_addr); + if (result) { + goto out; + } + *(level4p++) = __pa(level3p) | L3_ATTR; + addr += LEVEL3_SIZE; + } + /* clear the unused entries */ + while(addr < end_addr) { + *(level4p++) = 0; + addr += LEVEL3_SIZE; + } + out: + return result; +} + + +static int init_pgtable(struct kimage *image, unsigned long start_pgtable) +{ + u64 *level4p; + level4p = (u64 *)__va(start_pgtable); + return init_level4_page(image, level4p, 0, end_pfn << PAGE_SHIFT); +} + +static void set_idt(void *newidt, u16 limit) +{ + unsigned char curidt[10]; + + /* x86-64 supports unaliged loads & stores */ + (*(u16 *)(curidt)) = limit; + (*(u64 *)(curidt +2)) = (unsigned long)(newidt); + + __asm__ __volatile__ ( + "lidt %0\n" + : "=m" (curidt) + ); +}; + + +static void set_gdt(void *newgdt, u16 limit) +{ + unsigned char curgdt[10]; + + /* x86-64 supports unaligned loads & stores */ + (*(u16 *)(curgdt)) = limit; + (*(u64 *)(curgdt +2)) = (unsigned long)(newgdt); + + __asm__ __volatile__ ( + "lgdt %0\n" + : "=m" (curgdt) + ); +}; + +static void load_segments(void) +{ + __asm__ __volatile__ ( + "\tmovl $"STR(__KERNEL_DS)",%eax\n" + "\tmovl %eax,%ds\n" + "\tmovl %eax,%es\n" + "\tmovl %eax,%ss\n" + "\tmovl %eax,%fs\n" + "\tmovl %eax,%gs\n" + ); +#undef STR +#undef __STR +} + +typedef NORET_TYPE void (*relocate_new_kernel_t)( + unsigned long indirection_page, unsigned long control_code_buffer, + unsigned long start_address, unsigned long pgtable) ATTRIB_NORET; + +const extern unsigned char relocate_new_kernel[]; +const extern unsigned long relocate_new_kernel_size; + +int machine_kexec_prepare(struct kimage *image) +{ + unsigned long start_pgtable, control_code_buffer; + int result; + + /* Calculate the offsets */ + start_pgtable = page_to_pfn(image->control_code_page) << PAGE_SHIFT; + control_code_buffer = start_pgtable + 4096UL; + + /* Setup the identity mapped 64bit page table */ + result = init_pgtable(image, start_pgtable); + if (result) { + return result; + } + + /* Place the code in the reboot code buffer */ + memcpy(__va(control_code_buffer), relocate_new_kernel, relocate_new_kernel_size); + + return 0; +} + +void machine_kexec_cleanup(struct kimage *image) +{ + return; +} + +/* + * 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 control_code_buffer; + unsigned long start_pgtable; + relocate_new_kernel_t rnk; + + /* Interrupts aren't acceptable while we reboot */ + local_irq_disable(); + + /* Calculate the offsets */ + page_list = image->head; + start_pgtable = page_to_pfn(image->control_code_page) << PAGE_SHIFT; + control_code_buffer = start_pgtable + 4096UL; + + /* Set the low half of the page table to my identity mapped + * page table for kexec. Leave the high half pointing at the + * kernel pages. Don't bother to flush the global pages + * as that will happen when I fully switch to my identity mapped + * page table anyway. + */ + memcpy(__va(read_cr3()), __va(start_pgtable), PAGE_SIZE/2); + __flush_tlb(); + + + /* 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 unless you set the segment to a different selector. + * + * The more common model 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) control_code_buffer; + (*rnk)(page_list, control_code_buffer, image->start, start_pgtable); +} diff --git a/arch/x86_64/kernel/relocate_kernel.S b/arch/x86_64/kernel/relocate_kernel.S new file mode 100644 index 0000000000..d24fa9b72a --- /dev/null +++ b/arch/x86_64/kernel/relocate_kernel.S @@ -0,0 +1,143 @@ +/* + * relocate_kernel.S - put the kernel image in place to boot + * Copyright (C) 2002-2005 Eric Biederman + * + * This source code is licensed under the GNU General Public License, + * Version 2. See the file COPYING for more details. + */ + +#include + + /* + * 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 + .code64 +relocate_new_kernel: + /* %rdi page_list + * %rsi reboot_code_buffer + * %rdx start address + * %rcx page_table + * %r8 arg5 + * %r9 arg6 + */ + + /* zero out flags, and disable interrupts */ + pushq $0 + popfq + + /* set a new stack at the bottom of our page... */ + lea 4096(%rsi), %rsp + + /* store the parameters back on the stack */ + pushq %rdx /* store the start address */ + + /* Set cr0 to a known state: + * 31 1 == Paging enabled + * 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 + */ + movq %cr0, %rax + andq $~((1<<18)|(1<<16)|(1<<3)|(1<<2)), %rax + orl $((1<<31)|(1<<0)), %eax + movq %rax, %cr0 + + /* Set cr4 to a known state: + * 10 0 == xmm exceptions disabled + * 9 0 == xmm registers instructions disabled + * 8 0 == performance monitoring counter disabled + * 7 0 == page global disabled + * 6 0 == machine check exceptions disabled + * 5 1 == physical address extension enabled + * 4 0 == page size extensions disabled + * 3 0 == Debug extensions disabled + * 2 0 == Time stamp disable (disabled) + * 1 0 == Protected mode virtual interrupts disabled + * 0 0 == VME disabled + */ + + movq $((1<<5)), %rax + movq %rax, %cr4 + + jmp 1f +1: + + /* Switch to the identity mapped page tables, + * and flush the TLB. + */ + movq %rcx, %cr3 + + /* Do the copies */ + movq %rdi, %rcx /* Put the page_list in %rcx */ + xorq %rdi, %rdi + xorq %rsi, %rsi + jmp 1f + +0: /* top, read another word for the indirection page */ + + movq (%rbx), %rcx + addq $8, %rbx +1: + testq $0x1, %rcx /* is it a destination page? */ + jz 2f + movq %rcx, %rdi + andq $0xfffffffffffff000, %rdi + jmp 0b +2: + testq $0x2, %rcx /* is it an indirection page? */ + jz 2f + movq %rcx, %rbx + andq $0xfffffffffffff000, %rbx + jmp 0b +2: + testq $0x4, %rcx /* is it the done indicator? */ + jz 2f + jmp 3f +2: + testq $0x8, %rcx /* is it the source indicator? */ + jz 0b /* Ignore it otherwise */ + movq %rcx, %rsi /* For ever source page do a copy */ + andq $0xfffffffffffff000, %rsi + + 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 */ + + 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 +relocate_new_kernel_end: + + .globl relocate_new_kernel_size +relocate_new_kernel_size: + .quad relocate_new_kernel_end - relocate_new_kernel diff --git a/include/asm-x86_64/kexec.h b/include/asm-x86_64/kexec.h new file mode 100644 index 0000000000..dc33646dc7 --- /dev/null +++ b/include/asm-x86_64/kexec.h @@ -0,0 +1,28 @@ +#ifndef _X86_64_KEXEC_H +#define _X86_64_KEXEC_H + +#include +#include + +/* + * 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. + * + * So far x86_64 is limited to 40 physical address bits. + */ + +/* Maximum physical address we can use pages from */ +#define KEXEC_SOURCE_MEMORY_LIMIT (0xFFFFFFFFFFUL) +/* Maximum address we can reach in physical address mode */ +#define KEXEC_DESTINATION_MEMORY_LIMIT (0xFFFFFFFFFFUL) +/* Maximum address we can use for the control pages */ +#define KEXEC_CONTROL_MEMORY_LIMIT (0xFFFFFFFFFFUL) + +/* Allocate one page for the pdp and the second for the code */ +#define KEXEC_CONTROL_CODE_SIZE (4096UL + 4096UL) + +/* The native architecture */ +#define KEXEC_ARCH KEXEC_ARCH_X86_64 + +#endif /* _X86_64_KEXEC_H */ diff --git a/include/asm-x86_64/unistd.h b/include/asm-x86_64/unistd.h index 3c9af6fd43..d767adcbf0 100644 --- a/include/asm-x86_64/unistd.h +++ b/include/asm-x86_64/unistd.h @@ -552,7 +552,7 @@ __SYSCALL(__NR_mq_notify, sys_mq_notify) #define __NR_mq_getsetattr 245 __SYSCALL(__NR_mq_getsetattr, sys_mq_getsetattr) #define __NR_kexec_load 246 -__SYSCALL(__NR_kexec_load, sys_ni_syscall) +__SYSCALL(__NR_kexec_load, sys_kexec_load) #define __NR_waitid 247 __SYSCALL(__NR_waitid, sys_waitid) #define __NR_add_key 248