4bfaaef01a
kexec: Avoid overwriting the current pgd (V4, x86_64) This patch upgrades the x86_64-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>
277 lines
5.6 KiB
ArmAsm
277 lines
5.6 KiB
ArmAsm
/*
|
|
* relocate_kernel.S - put the kernel image in place to boot
|
|
* 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/linkage.h>
|
|
#include <asm/page.h>
|
|
#include <asm/kexec.h>
|
|
|
|
/*
|
|
* Must be relocatable PIC code callable as a C function
|
|
*/
|
|
|
|
#define PTR(x) (x << 3)
|
|
#define PAGE_ALIGNED (1 << PAGE_SHIFT)
|
|
#define PAGE_ATTR 0x63 /* _PAGE_PRESENT|_PAGE_RW|_PAGE_ACCESSED|_PAGE_DIRTY */
|
|
|
|
.text
|
|
.align PAGE_ALIGNED
|
|
.code64
|
|
.globl relocate_kernel
|
|
relocate_kernel:
|
|
/* %rdi indirection_page
|
|
* %rsi page_list
|
|
* %rdx start address
|
|
*/
|
|
|
|
/* map the control page at its virtual address */
|
|
|
|
movq $0x0000ff8000000000, %r10 /* mask */
|
|
mov $(39 - 3), %cl /* bits to shift */
|
|
movq PTR(VA_CONTROL_PAGE)(%rsi), %r11 /* address to map */
|
|
|
|
movq %r11, %r9
|
|
andq %r10, %r9
|
|
shrq %cl, %r9
|
|
|
|
movq PTR(VA_PGD)(%rsi), %r8
|
|
addq %r8, %r9
|
|
movq PTR(PA_PUD_0)(%rsi), %r8
|
|
orq $PAGE_ATTR, %r8
|
|
movq %r8, (%r9)
|
|
|
|
shrq $9, %r10
|
|
sub $9, %cl
|
|
|
|
movq %r11, %r9
|
|
andq %r10, %r9
|
|
shrq %cl, %r9
|
|
|
|
movq PTR(VA_PUD_0)(%rsi), %r8
|
|
addq %r8, %r9
|
|
movq PTR(PA_PMD_0)(%rsi), %r8
|
|
orq $PAGE_ATTR, %r8
|
|
movq %r8, (%r9)
|
|
|
|
shrq $9, %r10
|
|
sub $9, %cl
|
|
|
|
movq %r11, %r9
|
|
andq %r10, %r9
|
|
shrq %cl, %r9
|
|
|
|
movq PTR(VA_PMD_0)(%rsi), %r8
|
|
addq %r8, %r9
|
|
movq PTR(PA_PTE_0)(%rsi), %r8
|
|
orq $PAGE_ATTR, %r8
|
|
movq %r8, (%r9)
|
|
|
|
shrq $9, %r10
|
|
sub $9, %cl
|
|
|
|
movq %r11, %r9
|
|
andq %r10, %r9
|
|
shrq %cl, %r9
|
|
|
|
movq PTR(VA_PTE_0)(%rsi), %r8
|
|
addq %r8, %r9
|
|
movq PTR(PA_CONTROL_PAGE)(%rsi), %r8
|
|
orq $PAGE_ATTR, %r8
|
|
movq %r8, (%r9)
|
|
|
|
/* identity map the control page at its physical address */
|
|
|
|
movq $0x0000ff8000000000, %r10 /* mask */
|
|
mov $(39 - 3), %cl /* bits to shift */
|
|
movq PTR(PA_CONTROL_PAGE)(%rsi), %r11 /* address to map */
|
|
|
|
movq %r11, %r9
|
|
andq %r10, %r9
|
|
shrq %cl, %r9
|
|
|
|
movq PTR(VA_PGD)(%rsi), %r8
|
|
addq %r8, %r9
|
|
movq PTR(PA_PUD_1)(%rsi), %r8
|
|
orq $PAGE_ATTR, %r8
|
|
movq %r8, (%r9)
|
|
|
|
shrq $9, %r10
|
|
sub $9, %cl
|
|
|
|
movq %r11, %r9
|
|
andq %r10, %r9
|
|
shrq %cl, %r9
|
|
|
|
movq PTR(VA_PUD_1)(%rsi), %r8
|
|
addq %r8, %r9
|
|
movq PTR(PA_PMD_1)(%rsi), %r8
|
|
orq $PAGE_ATTR, %r8
|
|
movq %r8, (%r9)
|
|
|
|
shrq $9, %r10
|
|
sub $9, %cl
|
|
|
|
movq %r11, %r9
|
|
andq %r10, %r9
|
|
shrq %cl, %r9
|
|
|
|
movq PTR(VA_PMD_1)(%rsi), %r8
|
|
addq %r8, %r9
|
|
movq PTR(PA_PTE_1)(%rsi), %r8
|
|
orq $PAGE_ATTR, %r8
|
|
movq %r8, (%r9)
|
|
|
|
shrq $9, %r10
|
|
sub $9, %cl
|
|
|
|
movq %r11, %r9
|
|
andq %r10, %r9
|
|
shrq %cl, %r9
|
|
|
|
movq PTR(VA_PTE_1)(%rsi), %r8
|
|
addq %r8, %r9
|
|
movq PTR(PA_CONTROL_PAGE)(%rsi), %r8
|
|
orq $PAGE_ATTR, %r8
|
|
movq %r8, (%r9)
|
|
|
|
relocate_new_kernel:
|
|
/* %rdi indirection_page
|
|
* %rsi page_list
|
|
* %rdx start address
|
|
*/
|
|
|
|
/* zero out flags, and disable interrupts */
|
|
pushq $0
|
|
popfq
|
|
|
|
/* get physical address of control page now */
|
|
/* this is impossible after page table switch */
|
|
movq PTR(PA_CONTROL_PAGE)(%rsi), %r8
|
|
|
|
/* get physical address of page table now too */
|
|
movq PTR(PA_TABLE_PAGE)(%rsi), %rcx
|
|
|
|
/* switch to new set of page tables */
|
|
movq PTR(PA_PGD)(%rsi), %r9
|
|
movq %r9, %cr3
|
|
|
|
/* setup a new stack at the end of the physical control page */
|
|
lea 4096(%r8), %rsp
|
|
|
|
/* jump to identity mapped page */
|
|
addq $(identity_mapped - relocate_kernel), %r8
|
|
pushq %r8
|
|
ret
|
|
|
|
identity_mapped:
|
|
/* store the start address on the stack */
|
|
pushq %rdx
|
|
|
|
/* 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
|