linux/arch/x86/kernel/head_32.S

716 lines
17 KiB
ArmAsm
Raw Normal View History

/*
*
* Copyright (C) 1991, 1992 Linus Torvalds
*
* Enhanced CPU detection and feature setting code by Mike Jagdis
* and Martin Mares, November 1997.
*/
.text
#include <linux/threads.h>
#include <linux/init.h>
#include <linux/linkage.h>
#include <asm/segment.h>
#include <asm/page.h>
#include <asm/pgtable.h>
#include <asm/desc.h>
#include <asm/cache.h>
#include <asm/thread_info.h>
#include <asm/asm-offsets.h>
#include <asm/setup.h>
#include <asm/processor-flags.h>
/* Physical address */
#define pa(X) ((X) - __PAGE_OFFSET)
/*
* References to members of the new_cpu_data structure.
*/
#define X86 new_cpu_data+CPUINFO_x86
#define X86_VENDOR new_cpu_data+CPUINFO_x86_vendor
#define X86_MODEL new_cpu_data+CPUINFO_x86_model
#define X86_MASK new_cpu_data+CPUINFO_x86_mask
#define X86_HARD_MATH new_cpu_data+CPUINFO_hard_math
#define X86_CPUID new_cpu_data+CPUINFO_cpuid_level
#define X86_CAPABILITY new_cpu_data+CPUINFO_x86_capability
#define X86_VENDOR_ID new_cpu_data+CPUINFO_x86_vendor_id
/*
* This is how much memory *in addition to the memory covered up to
* and including _end* we need mapped initially.
* We need:
* - one bit for each possible page, but only in low memory, which means
* 2^32/4096/8 = 128K worst case (4G/4G split.)
* - enough space to map all low memory, which means
* (2^32/4096) / 1024 pages (worst case, non PAE)
* (2^32/4096) / 512 + 4 pages (worst case for PAE)
* - a few pages for allocator use before the kernel pagetable has
* been set up
*
* Modulo rounding, each megabyte assigned here requires a kilobyte of
* memory, which is currently unreclaimed.
*
* This should be a multiple of a page.
*/
LOW_PAGES = 1<<(32-PAGE_SHIFT_asm)
/*
* To preserve the DMA pool in PAGEALLOC kernels, we'll allocate
* pagetables from above the 16MB DMA limit, so we'll have to set
* up pagetables 16MB more (worst-case):
*/
#ifdef CONFIG_DEBUG_PAGEALLOC
LOW_PAGES = LOW_PAGES + 0x1000000
#endif
#if PTRS_PER_PMD > 1
PAGE_TABLE_SIZE = (LOW_PAGES / PTRS_PER_PMD) + PTRS_PER_PGD
#else
PAGE_TABLE_SIZE = (LOW_PAGES / PTRS_PER_PGD)
#endif
BOOTBITMAP_SIZE = LOW_PAGES / 8
ALLOCATOR_SLOP = 4
INIT_MAP_BEYOND_END = BOOTBITMAP_SIZE + (PAGE_TABLE_SIZE + ALLOCATOR_SLOP)*PAGE_SIZE_asm
/*
* 32-bit kernel entrypoint; only used by the boot CPU. On entry,
* %esi points to the real-mode code as a 32-bit pointer.
* CS and DS must be 4 GB flat segments, but we don't depend on
* any particular GDT layout, because we load our own as soon as we
* can.
*/
.section .text.head,"ax",@progbits
ENTRY(startup_32)
/* test KEEP_SEGMENTS flag to see if the bootloader is asking
us to not reload segments */
testb $(1<<6), BP_loadflags(%esi)
jnz 2f
/*
* Set segments to known values.
*/
lgdt pa(boot_gdt_descr)
movl $(__BOOT_DS),%eax
movl %eax,%ds
movl %eax,%es
movl %eax,%fs
movl %eax,%gs
2:
/*
* Clear BSS first so that there are no surprises...
*/
cld
xorl %eax,%eax
movl $pa(__bss_start),%edi
movl $pa(__bss_stop),%ecx
subl %edi,%ecx
shrl $2,%ecx
rep ; stosl
/*
* Copy bootup parameters out of the way.
* Note: %esi still has the pointer to the real-mode data.
* With the kexec as boot loader, parameter segment might be loaded beyond
* kernel image and might not even be addressable by early boot page tables.
* (kexec on panic case). Hence copy out the parameters before initializing
* page tables.
*/
movl $pa(boot_params),%edi
movl $(PARAM_SIZE/4),%ecx
cld
rep
movsl
movl pa(boot_params) + NEW_CL_POINTER,%esi
andl %esi,%esi
jz 1f # No comand line
movl $pa(boot_command_line),%edi
movl $(COMMAND_LINE_SIZE/4),%ecx
rep
movsl
1:
#ifdef CONFIG_PARAVIRT
/* This is can only trip for a broken bootloader... */
cmpw $0x207, pa(boot_params + BP_version)
jb default_entry
/* Paravirt-compatible boot parameters. Look to see what architecture
we're booting under. */
movl pa(boot_params + BP_hardware_subarch), %eax
cmpl $num_subarch_entries, %eax
jae bad_subarch
movl pa(subarch_entries)(,%eax,4), %eax
subl $__PAGE_OFFSET, %eax
jmp *%eax
bad_subarch:
WEAK(lguest_entry)
WEAK(xen_entry)
/* Unknown implementation; there's really
nothing we can do at this point. */
ud2a
__INITDATA
subarch_entries:
.long default_entry /* normal x86/PC */
.long lguest_entry /* lguest hypervisor */
.long xen_entry /* Xen hypervisor */
num_subarch_entries = (. - subarch_entries) / 4
.previous
#endif /* CONFIG_PARAVIRT */
/*
* Initialize page tables. This creates a PDE and a set of page
* tables, which are located immediately beyond _end. The variable
* init_pg_tables_end is set up to point to the first "safe" location.
* Mappings are created both at virtual address 0 (identity mapping)
* and PAGE_OFFSET for up to _end+sizeof(page tables)+INIT_MAP_BEYOND_END.
*
* Note that the stack is not yet set up!
*/
default_entry:
#ifdef CONFIG_X86_PAE
/*
* In PAE mode swapper_pg_dir is statically defined to contain enough
* entries to cover the VMSPLIT option (that is the top 1, 2 or 3
* entries). The identity mapping is handled by pointing two PGD
* entries to the first kernel PMD.
*
* Note the upper half of each PMD or PTE are always zero at
* this stage.
*/
x86: fix asm warning in head_32.S On Mon, May 19, 2008 at 04:10:02PM -0700, Linus Torvalds wrote: > It also causes these warnings on 32-bit PAE: > > AS arch/x86/kernel/head_32.o > arch/x86/kernel/head_32.S: Assembler messages: > arch/x86/kernel/head_32.S:225: Warning: left operand is a bignum; integer 0 assumed > arch/x86/kernel/head_32.S:609: Warning: left operand is a bignum; integer 0 assumed > > and I do not see why (the end result seems to be identical). Fix head_32.S gcc bignum warnings when CONFIG_PAE=y. arch/x86/kernel/head_32.S: Assembler messages: arch/x86/kernel/head_32.S:225: Warning: left operand is a bignum; integer 0 assumed arch/x86/kernel/head_32.S:609: Warning: left operand is a bignum; integer 0 assumed The assembler was stumbling over the 64-bit constant 0x100000000 in the KPMDS #define. Testing: a cmp(1) on head_32.o before and after shows the binary is unchanged. Signed-off-by: Joe Korty <joe.korty@ccur.com Cc: Hugh Dickins <hugh@veritas.com> Cc: Theodore Tso <tytso@mit.edu> Cc: Gabriel C <nix.or.die@googlemail.com> Cc: Keith Packard <keithp@keithp.com> Cc: "Pallipadi Venkatesh" <venkatesh.pallipadi@intel.com> Cc: Eric Anholt <eric@anholt.net> Cc: "Siddha Suresh B" <suresh.b.siddha@intel.com> Cc: bugme-daemon@bugzilla.kernel.org Cc: airlied@linux.ie Cc: "Barnes Jesse" <jesse.barnes@intel.com> Cc: Jeremy Fitzhardinge <jeremy@goop.org> Cc: Andrew Morton <akpm@linux-foundation.org> Cc: "Rafael J. Wysocki" <rjw@sisk.pl> Cc: Linus Torvalds <torvalds@linux-foundation.org> Signed-off-by: Thomas Gleixner <tglx@linutronix.de> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2008-06-02 21:21:06 +00:00
#define KPMDS (((-__PAGE_OFFSET) >> 30) & 3) /* Number of kernel PMDs */
xorl %ebx,%ebx /* %ebx is kept at zero */
movl $pa(pg0), %edi
movl %edi, pa(init_pg_tables_start)
movl $pa(swapper_pg_pmd), %edx
movl $PTE_IDENT_ATTR, %eax
10:
leal PDE_IDENT_ATTR(%edi),%ecx /* Create PMD entry */
movl %ecx,(%edx) /* Store PMD entry */
/* Upper half already zero */
addl $8,%edx
movl $512,%ecx
11:
stosl
xchgl %eax,%ebx
stosl
xchgl %eax,%ebx
addl $0x1000,%eax
loop 11b
/*
* End condition: we must map up to and including INIT_MAP_BEYOND_END
* bytes beyond the end of our own page tables.
*/
leal (INIT_MAP_BEYOND_END+PTE_IDENT_ATTR)(%edi),%ebp
cmpl %ebp,%eax
jb 10b
1:
movl %edi,pa(init_pg_tables_end)
shrl $12, %eax
movl %eax, pa(max_pfn_mapped)
/* Do early initialization of the fixmap area */
movl $pa(swapper_pg_fixmap)+PDE_IDENT_ATTR,%eax
movl %eax,pa(swapper_pg_pmd+0x1000*KPMDS-8)
#else /* Not PAE */
page_pde_offset = (__PAGE_OFFSET >> 20);
movl $pa(pg0), %edi
movl %edi, pa(init_pg_tables_start)
movl $pa(swapper_pg_dir), %edx
movl $PTE_IDENT_ATTR, %eax
10:
leal PDE_IDENT_ATTR(%edi),%ecx /* Create PDE entry */
movl %ecx,(%edx) /* Store identity PDE entry */
movl %ecx,page_pde_offset(%edx) /* Store kernel PDE entry */
addl $4,%edx
movl $1024, %ecx
11:
stosl
addl $0x1000,%eax
loop 11b
/*
* End condition: we must map up to and including INIT_MAP_BEYOND_END
* bytes beyond the end of our own page tables; the +0x007 is
* the attribute bits
*/
leal (INIT_MAP_BEYOND_END+PTE_IDENT_ATTR)(%edi),%ebp
cmpl %ebp,%eax
jb 10b
movl %edi,pa(init_pg_tables_end)
shrl $12, %eax
movl %eax, pa(max_pfn_mapped)
/* Do early initialization of the fixmap area */
movl $pa(swapper_pg_fixmap)+PDE_IDENT_ATTR,%eax
movl %eax,pa(swapper_pg_dir+0xffc)
#endif
jmp 3f
/*
* Non-boot CPU entry point; entered from trampoline.S
* We can't lgdt here, because lgdt itself uses a data segment, but
* we know the trampoline has already loaded the boot_gdt for us.
*
* If cpu hotplug is not supported then this code can go in init section
* which will be freed later
*/
#ifndef CONFIG_HOTPLUG_CPU
.section .init.text,"ax",@progbits
#endif
#ifdef CONFIG_SMP
ENTRY(startup_32_smp)
cld
movl $(__BOOT_DS),%eax
movl %eax,%ds
movl %eax,%es
movl %eax,%fs
movl %eax,%gs
#endif /* CONFIG_SMP */
3:
/*
* New page tables may be in 4Mbyte page mode and may
* be using the global pages.
*
* NOTE! If we are on a 486 we may have no cr4 at all!
* So we do not try to touch it unless we really have
* some bits in it to set. This won't work if the BSP
* implements cr4 but this AP does not -- very unlikely
* but be warned! The same applies to the pse feature
* if not equally supported. --macro
*
* NOTE! We have to correct for the fact that we're
* not yet offset PAGE_OFFSET..
*/
#define cr4_bits pa(mmu_cr4_features)
movl cr4_bits,%edx
andl %edx,%edx
jz 6f
movl %cr4,%eax # Turn on paging options (PSE,PAE,..)
orl %edx,%eax
movl %eax,%cr4
btl $5, %eax # check if PAE is enabled
jnc 6f
/* Check if extended functions are implemented */
movl $0x80000000, %eax
cpuid
cmpl $0x80000000, %eax
jbe 6f
mov $0x80000001, %eax
cpuid
/* Execute Disable bit supported? */
btl $20, %edx
jnc 6f
/* Setup EFER (Extended Feature Enable Register) */
movl $0xc0000080, %ecx
rdmsr
btsl $11, %eax
/* Make changes effective */
wrmsr
6:
/*
* Enable paging
*/
movl $pa(swapper_pg_dir),%eax
movl %eax,%cr3 /* set the page table pointer.. */
movl %cr0,%eax
orl $X86_CR0_PG,%eax
movl %eax,%cr0 /* ..and set paging (PG) bit */
ljmp $__BOOT_CS,$1f /* Clear prefetch and normalize %eip */
1:
/* Set up the stack pointer */
lss stack_start,%esp
/*
* Initialize eflags. Some BIOS's leave bits like NT set. This would
* confuse the debugger if this code is traced.
* XXX - best to initialize before switching to protected mode.
*/
pushl $0
popfl
#ifdef CONFIG_SMP
cmpb $0, ready
jz 1f /* Initial CPU cleans BSS */
jmp checkCPUtype
1:
#endif /* CONFIG_SMP */
/*
* start system 32-bit setup. We need to re-do some of the things done
* in 16-bit mode for the "real" operations.
*/
call setup_idt
checkCPUtype:
movl $-1,X86_CPUID # -1 for no CPUID initially
/* check if it is 486 or 386. */
/*
* XXX - this does a lot of unnecessary setup. Alignment checks don't
* apply at our cpl of 0 and the stack ought to be aligned already, and
* we don't need to preserve eflags.
*/
movb $3,X86 # at least 386
pushfl # push EFLAGS
popl %eax # get EFLAGS
movl %eax,%ecx # save original EFLAGS
xorl $0x240000,%eax # flip AC and ID bits in EFLAGS
pushl %eax # copy to EFLAGS
popfl # set EFLAGS
pushfl # get new EFLAGS
popl %eax # put it in eax
xorl %ecx,%eax # change in flags
pushl %ecx # restore original EFLAGS
popfl
testl $0x40000,%eax # check if AC bit changed
je is386
movb $4,X86 # at least 486
testl $0x200000,%eax # check if ID bit changed
je is486
/* get vendor info */
xorl %eax,%eax # call CPUID with 0 -> return vendor ID
cpuid
movl %eax,X86_CPUID # save CPUID level
movl %ebx,X86_VENDOR_ID # lo 4 chars
movl %edx,X86_VENDOR_ID+4 # next 4 chars
movl %ecx,X86_VENDOR_ID+8 # last 4 chars
orl %eax,%eax # do we have processor info as well?
je is486
movl $1,%eax # Use the CPUID instruction to get CPU type
cpuid
movb %al,%cl # save reg for future use
andb $0x0f,%ah # mask processor family
movb %ah,X86
andb $0xf0,%al # mask model
shrb $4,%al
movb %al,X86_MODEL
andb $0x0f,%cl # mask mask revision
movb %cl,X86_MASK
movl %edx,X86_CAPABILITY
is486: movl $0x50022,%ecx # set AM, WP, NE and MP
jmp 2f
is386: movl $2,%ecx # set MP
2: movl %cr0,%eax
andl $0x80000011,%eax # Save PG,PE,ET
orl %ecx,%eax
movl %eax,%cr0
call check_x87
lgdt early_gdt_descr
lidt idt_descr
ljmp $(__KERNEL_CS),$1f
1: movl $(__KERNEL_DS),%eax # reload all the segment registers
movl %eax,%ss # after changing gdt.
movl %eax,%fs # gets reset once there's real percpu
movl $(__USER_DS),%eax # DS/ES contains default USER segment
movl %eax,%ds
movl %eax,%es
xorl %eax,%eax # Clear GS and LDT
movl %eax,%gs
lldt %ax
cld # gcc2 wants the direction flag cleared at all times
pushl $0 # fake return address for unwinder
#ifdef CONFIG_SMP
movb ready, %cl
movb $1, ready
cmpb $0,%cl # the first CPU calls start_kernel
je 1f
movl $(__KERNEL_PERCPU), %eax
movl %eax,%fs # set this cpu's percpu
movl (stack_start), %esp
1:
#endif /* CONFIG_SMP */
jmp *(initial_code)
/*
* We depend on ET to be correct. This checks for 287/387.
*/
check_x87:
movb $0,X86_HARD_MATH
clts
fninit
fstsw %ax
cmpb $0,%al
je 1f
movl %cr0,%eax /* no coprocessor: have to set bits */
xorl $4,%eax /* set EM */
movl %eax,%cr0
ret
ALIGN
1: movb $1,X86_HARD_MATH
.byte 0xDB,0xE4 /* fsetpm for 287, ignored by 387 */
ret
/*
* setup_idt
*
* sets up a idt with 256 entries pointing to
* ignore_int, interrupt gates. It doesn't actually load
* idt - that can be done only after paging has been enabled
* and the kernel moved to PAGE_OFFSET. Interrupts
* are enabled elsewhere, when we can be relatively
* sure everything is ok.
*
* Warning: %esi is live across this function.
*/
setup_idt:
lea ignore_int,%edx
movl $(__KERNEL_CS << 16),%eax
movw %dx,%ax /* selector = 0x0010 = cs */
movw $0x8E00,%dx /* interrupt gate - dpl=0, present */
lea idt_table,%edi
mov $256,%ecx
rp_sidt:
movl %eax,(%edi)
movl %edx,4(%edi)
addl $8,%edi
dec %ecx
jne rp_sidt
.macro set_early_handler handler,trapno
lea \handler,%edx
movl $(__KERNEL_CS << 16),%eax
movw %dx,%ax
movw $0x8E00,%dx /* interrupt gate - dpl=0, present */
lea idt_table,%edi
movl %eax,8*\trapno(%edi)
movl %edx,8*\trapno+4(%edi)
.endm
set_early_handler handler=early_divide_err,trapno=0
set_early_handler handler=early_illegal_opcode,trapno=6
set_early_handler handler=early_protection_fault,trapno=13
set_early_handler handler=early_page_fault,trapno=14
ret
early_divide_err:
xor %edx,%edx
pushl $0 /* fake errcode */
jmp early_fault
early_illegal_opcode:
movl $6,%edx
pushl $0 /* fake errcode */
jmp early_fault
early_protection_fault:
movl $13,%edx
jmp early_fault
early_page_fault:
movl $14,%edx
jmp early_fault
early_fault:
cld
#ifdef CONFIG_PRINTK
pusha
movl $(__KERNEL_DS),%eax
movl %eax,%ds
movl %eax,%es
cmpl $2,early_recursion_flag
je hlt_loop
incl early_recursion_flag
movl %cr2,%eax
pushl %eax
pushl %edx /* trapno */
pushl $fault_msg
#ifdef CONFIG_EARLY_PRINTK
call early_printk
#else
call printk
#endif
#endif
call dump_stack
hlt_loop:
hlt
jmp hlt_loop
/* This is the default interrupt "handler" :-) */
ALIGN
ignore_int:
cld
#ifdef CONFIG_PRINTK
pushl %eax
pushl %ecx
pushl %edx
pushl %es
pushl %ds
movl $(__KERNEL_DS),%eax
movl %eax,%ds
movl %eax,%es
cmpl $2,early_recursion_flag
je hlt_loop
incl early_recursion_flag
pushl 16(%esp)
pushl 24(%esp)
pushl 32(%esp)
pushl 40(%esp)
pushl $int_msg
#ifdef CONFIG_EARLY_PRINTK
call early_printk
#else
call printk
#endif
addl $(5*4),%esp
popl %ds
popl %es
popl %edx
popl %ecx
popl %eax
#endif
iret
.section .cpuinit.data,"wa"
.align 4
ENTRY(initial_code)
.long i386_start_kernel
.section .text
/*
* Real beginning of normal "text" segment
*/
ENTRY(stext)
ENTRY(_stext)
/*
* BSS section
*/
xen: Core Xen implementation This patch is a rollup of all the core pieces of the Xen implementation, including: - booting and setup - pagetable setup - privileged instructions - segmentation - interrupt flags - upcalls - multicall batching BOOTING AND SETUP The vmlinux image is decorated with ELF notes which tell the Xen domain builder what the kernel's requirements are; the domain builder then constructs the address space accordingly and starts the kernel. Xen has its own entrypoint for the kernel (contained in an ELF note). The ELF notes are set up by xen-head.S, which is included into head.S. In principle it could be linked separately, but it seems to provoke lots of binutils bugs. Because the domain builder starts the kernel in a fairly sane state (32-bit protected mode, paging enabled, flat segments set up), there's not a lot of setup needed before starting the kernel proper. The main steps are: 1. Install the Xen paravirt_ops, which is simply a matter of a structure assignment. 2. Set init_mm to use the Xen-supplied pagetables (analogous to the head.S generated pagetables in a native boot). 3. Reserve address space for Xen, since it takes a chunk at the top of the address space for its own use. 4. Call start_kernel() PAGETABLE SETUP Once we hit the main kernel boot sequence, it will end up calling back via paravirt_ops to set up various pieces of Xen specific state. One of the critical things which requires a bit of extra care is the construction of the initial init_mm pagetable. Because Xen places tight constraints on pagetables (an active pagetable must always be valid, and must always be mapped read-only to the guest domain), we need to be careful when constructing the new pagetable to keep these constraints in mind. It turns out that the easiest way to do this is use the initial Xen-provided pagetable as a template, and then just insert new mappings for memory where a mapping doesn't already exist. This means that during pagetable setup, it uses a special version of xen_set_pte which ignores any attempt to remap a read-only page as read-write (since Xen will map its own initial pagetable as RO), but lets other changes to the ptes happen, so that things like NX are set properly. PRIVILEGED INSTRUCTIONS AND SEGMENTATION When the kernel runs under Xen, it runs in ring 1 rather than ring 0. This means that it is more privileged than user-mode in ring 3, but it still can't run privileged instructions directly. Non-performance critical instructions are dealt with by taking a privilege exception and trapping into the hypervisor and emulating the instruction, but more performance-critical instructions have their own specific paravirt_ops. In many cases we can avoid having to do any hypercalls for these instructions, or the Xen implementation is quite different from the normal native version. The privileged instructions fall into the broad classes of: Segmentation: setting up the GDT and the GDT entries, LDT, TLS and so on. Xen doesn't allow the GDT to be directly modified; all GDT updates are done via hypercalls where the new entries can be validated. This is important because Xen uses segment limits to prevent the guest kernel from damaging the hypervisor itself. Traps and exceptions: Xen uses a special format for trap entrypoints, so when the kernel wants to set an IDT entry, it needs to be converted to the form Xen expects. Xen sets int 0x80 up specially so that the trap goes straight from userspace into the guest kernel without going via the hypervisor. sysenter isn't supported. Kernel stack: The esp0 entry is extracted from the tss and provided to Xen. TLB operations: the various TLB calls are mapped into corresponding Xen hypercalls. Control registers: all the control registers are privileged. The most important is cr3, which points to the base of the current pagetable, and we handle it specially. Another instruction we treat specially is CPUID, even though its not privileged. We want to control what CPU features are visible to the rest of the kernel, and so CPUID ends up going into a paravirt_op. Xen implements this mainly to disable the ACPI and APIC subsystems. INTERRUPT FLAGS Xen maintains its own separate flag for masking events, which is contained within the per-cpu vcpu_info structure. Because the guest kernel runs in ring 1 and not 0, the IF flag in EFLAGS is completely ignored (and must be, because even if a guest domain disables interrupts for itself, it can't disable them overall). (A note on terminology: "events" and interrupts are effectively synonymous. However, rather than using an "enable flag", Xen uses a "mask flag", which blocks event delivery when it is non-zero.) There are paravirt_ops for each of cli/sti/save_fl/restore_fl, which are implemented to manage the Xen event mask state. The only thing worth noting is that when events are unmasked, we need to explicitly see if there's a pending event and call into the hypervisor to make sure it gets delivered. UPCALLS Xen needs a couple of upcall (or callback) functions to be implemented by each guest. One is the event upcalls, which is how events (interrupts, effectively) are delivered to the guests. The other is the failsafe callback, which is used to report errors in either reloading a segment register, or caused by iret. These are implemented in i386/kernel/entry.S so they can jump into the normal iret_exc path when necessary. MULTICALL BATCHING Xen provides a multicall mechanism, which allows multiple hypercalls to be issued at once in order to mitigate the cost of trapping into the hypervisor. This is particularly useful for context switches, since the 4-5 hypercalls they would normally need (reload cr3, update TLS, maybe update LDT) can be reduced to one. This patch implements a generic batching mechanism for hypercalls, which gets used in many places in the Xen code. Signed-off-by: Jeremy Fitzhardinge <jeremy@xensource.com> Signed-off-by: Chris Wright <chrisw@sous-sol.org> Cc: Ian Pratt <ian.pratt@xensource.com> Cc: Christian Limpach <Christian.Limpach@cl.cam.ac.uk> Cc: Adrian Bunk <bunk@stusta.de>
2007-07-18 01:37:04 +00:00
.section ".bss.page_aligned","wa"
.align PAGE_SIZE_asm
#ifdef CONFIG_X86_PAE
swapper_pg_pmd:
.fill 1024*KPMDS,4,0
#else
ENTRY(swapper_pg_dir)
.fill 1024,4,0
#endif
swapper_pg_fixmap:
.fill 1024,4,0
ENTRY(empty_zero_page)
.fill 4096,1,0
/*
* This starts the data section.
*/
#ifdef CONFIG_X86_PAE
.section ".data.page_aligned","wa"
/* Page-aligned for the benefit of paravirt? */
.align PAGE_SIZE_asm
ENTRY(swapper_pg_dir)
.long pa(swapper_pg_pmd+PGD_IDENT_ATTR),0 /* low identity map */
# if KPMDS == 3
.long pa(swapper_pg_pmd+PGD_IDENT_ATTR),0
.long pa(swapper_pg_pmd+PGD_IDENT_ATTR+0x1000),0
.long pa(swapper_pg_pmd+PGD_IDENT_ATTR+0x2000),0
# elif KPMDS == 2
.long 0,0
.long pa(swapper_pg_pmd+PGD_IDENT_ATTR),0
.long pa(swapper_pg_pmd+PGD_IDENT_ATTR+0x1000),0
# elif KPMDS == 1
.long 0,0
.long 0,0
.long pa(swapper_pg_pmd+PGD_IDENT_ATTR),0
# else
# error "Kernel PMDs should be 1, 2 or 3"
# endif
.align PAGE_SIZE_asm /* needs to be page-sized too */
#endif
.data
ENTRY(stack_start)
.long init_thread_union+THREAD_SIZE
.long __BOOT_DS
ready: .byte 0
early_recursion_flag:
.long 0
int_msg:
.asciz "Unknown interrupt or fault at EIP %p %p %p\n"
fault_msg:
/* fault info: */
.ascii "BUG: Int %d: CR2 %p\n"
/* pusha regs: */
.ascii " EDI %p ESI %p EBP %p ESP %p\n"
.ascii " EBX %p EDX %p ECX %p EAX %p\n"
/* fault frame: */
.ascii " err %p EIP %p CS %p flg %p\n"
.ascii "Stack: %p %p %p %p %p %p %p %p\n"
.ascii " %p %p %p %p %p %p %p %p\n"
.asciz " %p %p %p %p %p %p %p %p\n"
#include "../../x86/xen/xen-head.S"
xen: Core Xen implementation This patch is a rollup of all the core pieces of the Xen implementation, including: - booting and setup - pagetable setup - privileged instructions - segmentation - interrupt flags - upcalls - multicall batching BOOTING AND SETUP The vmlinux image is decorated with ELF notes which tell the Xen domain builder what the kernel's requirements are; the domain builder then constructs the address space accordingly and starts the kernel. Xen has its own entrypoint for the kernel (contained in an ELF note). The ELF notes are set up by xen-head.S, which is included into head.S. In principle it could be linked separately, but it seems to provoke lots of binutils bugs. Because the domain builder starts the kernel in a fairly sane state (32-bit protected mode, paging enabled, flat segments set up), there's not a lot of setup needed before starting the kernel proper. The main steps are: 1. Install the Xen paravirt_ops, which is simply a matter of a structure assignment. 2. Set init_mm to use the Xen-supplied pagetables (analogous to the head.S generated pagetables in a native boot). 3. Reserve address space for Xen, since it takes a chunk at the top of the address space for its own use. 4. Call start_kernel() PAGETABLE SETUP Once we hit the main kernel boot sequence, it will end up calling back via paravirt_ops to set up various pieces of Xen specific state. One of the critical things which requires a bit of extra care is the construction of the initial init_mm pagetable. Because Xen places tight constraints on pagetables (an active pagetable must always be valid, and must always be mapped read-only to the guest domain), we need to be careful when constructing the new pagetable to keep these constraints in mind. It turns out that the easiest way to do this is use the initial Xen-provided pagetable as a template, and then just insert new mappings for memory where a mapping doesn't already exist. This means that during pagetable setup, it uses a special version of xen_set_pte which ignores any attempt to remap a read-only page as read-write (since Xen will map its own initial pagetable as RO), but lets other changes to the ptes happen, so that things like NX are set properly. PRIVILEGED INSTRUCTIONS AND SEGMENTATION When the kernel runs under Xen, it runs in ring 1 rather than ring 0. This means that it is more privileged than user-mode in ring 3, but it still can't run privileged instructions directly. Non-performance critical instructions are dealt with by taking a privilege exception and trapping into the hypervisor and emulating the instruction, but more performance-critical instructions have their own specific paravirt_ops. In many cases we can avoid having to do any hypercalls for these instructions, or the Xen implementation is quite different from the normal native version. The privileged instructions fall into the broad classes of: Segmentation: setting up the GDT and the GDT entries, LDT, TLS and so on. Xen doesn't allow the GDT to be directly modified; all GDT updates are done via hypercalls where the new entries can be validated. This is important because Xen uses segment limits to prevent the guest kernel from damaging the hypervisor itself. Traps and exceptions: Xen uses a special format for trap entrypoints, so when the kernel wants to set an IDT entry, it needs to be converted to the form Xen expects. Xen sets int 0x80 up specially so that the trap goes straight from userspace into the guest kernel without going via the hypervisor. sysenter isn't supported. Kernel stack: The esp0 entry is extracted from the tss and provided to Xen. TLB operations: the various TLB calls are mapped into corresponding Xen hypercalls. Control registers: all the control registers are privileged. The most important is cr3, which points to the base of the current pagetable, and we handle it specially. Another instruction we treat specially is CPUID, even though its not privileged. We want to control what CPU features are visible to the rest of the kernel, and so CPUID ends up going into a paravirt_op. Xen implements this mainly to disable the ACPI and APIC subsystems. INTERRUPT FLAGS Xen maintains its own separate flag for masking events, which is contained within the per-cpu vcpu_info structure. Because the guest kernel runs in ring 1 and not 0, the IF flag in EFLAGS is completely ignored (and must be, because even if a guest domain disables interrupts for itself, it can't disable them overall). (A note on terminology: "events" and interrupts are effectively synonymous. However, rather than using an "enable flag", Xen uses a "mask flag", which blocks event delivery when it is non-zero.) There are paravirt_ops for each of cli/sti/save_fl/restore_fl, which are implemented to manage the Xen event mask state. The only thing worth noting is that when events are unmasked, we need to explicitly see if there's a pending event and call into the hypervisor to make sure it gets delivered. UPCALLS Xen needs a couple of upcall (or callback) functions to be implemented by each guest. One is the event upcalls, which is how events (interrupts, effectively) are delivered to the guests. The other is the failsafe callback, which is used to report errors in either reloading a segment register, or caused by iret. These are implemented in i386/kernel/entry.S so they can jump into the normal iret_exc path when necessary. MULTICALL BATCHING Xen provides a multicall mechanism, which allows multiple hypercalls to be issued at once in order to mitigate the cost of trapping into the hypervisor. This is particularly useful for context switches, since the 4-5 hypercalls they would normally need (reload cr3, update TLS, maybe update LDT) can be reduced to one. This patch implements a generic batching mechanism for hypercalls, which gets used in many places in the Xen code. Signed-off-by: Jeremy Fitzhardinge <jeremy@xensource.com> Signed-off-by: Chris Wright <chrisw@sous-sol.org> Cc: Ian Pratt <ian.pratt@xensource.com> Cc: Christian Limpach <Christian.Limpach@cl.cam.ac.uk> Cc: Adrian Bunk <bunk@stusta.de>
2007-07-18 01:37:04 +00:00
/*
* The IDT and GDT 'descriptors' are a strange 48-bit object
* only used by the lidt and lgdt instructions. They are not
* like usual segment descriptors - they consist of a 16-bit
* segment size, and 32-bit linear address value:
*/
.globl boot_gdt_descr
.globl idt_descr
ALIGN
# early boot GDT descriptor (must use 1:1 address mapping)
.word 0 # 32 bit align gdt_desc.address
boot_gdt_descr:
.word __BOOT_DS+7
.long boot_gdt - __PAGE_OFFSET
.word 0 # 32-bit align idt_desc.address
idt_descr:
.word IDT_ENTRIES*8-1 # idt contains 256 entries
.long idt_table
# boot GDT descriptor (later on used by CPU#0):
.word 0 # 32 bit align gdt_desc.address
ENTRY(early_gdt_descr)
.word GDT_ENTRIES*8-1
.long per_cpu__gdt_page /* Overwritten for secondary CPUs */
/*
* The boot_gdt must mirror the equivalent in setup.S and is
* used only for booting.
*/
.align L1_CACHE_BYTES
ENTRY(boot_gdt)
.fill GDT_ENTRY_BOOT_CS,8,0
.quad 0x00cf9a000000ffff /* kernel 4GB code at 0x00000000 */
.quad 0x00cf92000000ffff /* kernel 4GB data at 0x00000000 */