From be44d2aabce2d62f72d5751d1871b6212bf7a1c7 Mon Sep 17 00:00:00 2001 From: Stas Sergeev Date: Thu, 7 Dec 2006 02:14:01 +0100 Subject: [PATCH] i386: espfix cleanup Clean up the espfix code: - Introduced PER_CPU() macro to be used from asm - Introduced GET_DESC_BASE() macro to be used from asm - Rewrote the fixup code in asm, as calling a C code with the altered %ss appeared to be unsafe - No longer altering the stack from a .fixup section - 16bit per-cpu stack is no longer used, instead the stack segment base is patched the way so that the high word of the kernel and user %esp are the same. - Added the limit-patching for the espfix segment. (Chuck Ebbert) [jeremy@goop.org: use the x86 scaling addressing mode rather than shifting] Signed-off-by: Stas Sergeev Signed-off-by: Andi Kleen Acked-by: Zachary Amsden Acked-by: Chuck Ebbert <76306.1226@compuserve.com> Acked-by: Jan Beulich Cc: Andi Kleen Signed-off-by: Jeremy Fitzhardinge Signed-off-by: Andrew Morton --- arch/i386/kernel/head.S | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'arch/i386/kernel/head.S') diff --git a/arch/i386/kernel/head.S b/arch/i386/kernel/head.S index ca31f18d277c..b1f1df11fcc6 100644 --- a/arch/i386/kernel/head.S +++ b/arch/i386/kernel/head.S @@ -584,7 +584,7 @@ ENTRY(cpu_gdt_table) .quad 0x00009a000000ffff /* 0xc0 APM CS 16 code (16 bit) */ .quad 0x004092000000ffff /* 0xc8 APM DS data */ - .quad 0x0000920000000000 /* 0xd0 - ESPFIX 16-bit SS */ + .quad 0x00c0920000000000 /* 0xd0 - ESPFIX SS */ .quad 0x0000000000000000 /* 0xd8 - unused */ .quad 0x0000000000000000 /* 0xe0 - unused */ .quad 0x0000000000000000 /* 0xe8 - unused */ -- cgit v1.2.3 From 9ca36101a8d74704d78f10910f89d62de96f9dc8 Mon Sep 17 00:00:00 2001 From: Jeremy Fitzhardinge Date: Thu, 7 Dec 2006 02:14:02 +0100 Subject: [PATCH] i386: Basic definitions for i386-pda This patch has the basic definitions of struct i386_pda, and the segment selector in the GDT. asm-i386/pda.h is more or less a direct copy of asm-x86_64/pda.h. The most interesting difference is the use of _proxy_pda, which is used to give gcc a model for the actual memory operations on the real pda structure. No actual reference is ever made to _proxy_pda, so it is never defined. Signed-off-by: Jeremy Fitzhardinge Signed-off-by: Andi Kleen Cc: Chuck Ebbert <76306.1226@compuserve.com> Cc: Zachary Amsden Cc: Jan Beulich Cc: Andi Kleen Signed-off-by: Andrew Morton --- arch/i386/kernel/head.S | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'arch/i386/kernel/head.S') diff --git a/arch/i386/kernel/head.S b/arch/i386/kernel/head.S index b1f1df11fcc6..4a83384c5a61 100644 --- a/arch/i386/kernel/head.S +++ b/arch/i386/kernel/head.S @@ -585,7 +585,7 @@ ENTRY(cpu_gdt_table) .quad 0x004092000000ffff /* 0xc8 APM DS data */ .quad 0x00c0920000000000 /* 0xd0 - ESPFIX SS */ - .quad 0x0000000000000000 /* 0xd8 - unused */ + .quad 0x0000000000000000 /* 0xd8 - PDA */ .quad 0x0000000000000000 /* 0xe0 - unused */ .quad 0x0000000000000000 /* 0xe8 - unused */ .quad 0x0000000000000000 /* 0xf0 - unused */ -- cgit v1.2.3 From f95d47caae5302a63d92be9a0292abc90e2a14e1 Mon Sep 17 00:00:00 2001 From: Jeremy Fitzhardinge Date: Thu, 7 Dec 2006 02:14:02 +0100 Subject: [PATCH] i386: Use %gs as the PDA base-segment in the kernel This patch is the meat of the PDA change. This patch makes several related changes: 1: Most significantly, %gs is now used in the kernel. This means that on entry, the old value of %gs is saved away, and it is reloaded with __KERNEL_PDA. 2: entry.S constructs the stack in the shape of struct pt_regs, and this is passed around the kernel so that the process's saved register state can be accessed. Unfortunately struct pt_regs doesn't currently have space for %gs (or %fs). This patch extends pt_regs to add space for gs (no space is allocated for %fs, since it won't be used, and it would just complicate the code in entry.S to work around the space). 3: Because %gs is now saved on the stack like %ds, %es and the integer registers, there are a number of places where it no longer needs to be handled specially; namely context switch, and saving/restoring the register state in a signal context. 4: And since kernel threads run in kernel space and call normal kernel code, they need to be created with their %gs == __KERNEL_PDA. Signed-off-by: Jeremy Fitzhardinge Signed-off-by: Andi Kleen Cc: Chuck Ebbert <76306.1226@compuserve.com> Cc: Zachary Amsden Cc: Jan Beulich Cc: Andi Kleen Signed-off-by: Andrew Morton --- arch/i386/kernel/head.S | 31 +++++++++++++++++++++++++++---- 1 file changed, 27 insertions(+), 4 deletions(-) (limited to 'arch/i386/kernel/head.S') diff --git a/arch/i386/kernel/head.S b/arch/i386/kernel/head.S index 4a83384c5a61..5b14e95ac8b9 100644 --- a/arch/i386/kernel/head.S +++ b/arch/i386/kernel/head.S @@ -302,6 +302,7 @@ is386: movl $2,%ecx # set MP movl %eax,%cr0 call check_x87 + call setup_pda lgdt cpu_gdt_descr lidt idt_descr ljmp $(__KERNEL_CS),$1f @@ -312,10 +313,13 @@ is386: movl $2,%ecx # set MP movl %eax,%ds movl %eax,%es - xorl %eax,%eax # Clear FS/GS and LDT + xorl %eax,%eax # Clear FS and LDT movl %eax,%fs - movl %eax,%gs lldt %ax + + movl $(__KERNEL_PDA),%eax + mov %eax,%gs + cld # gcc2 wants the direction flag cleared at all times pushl $0 # fake return address for unwinder #ifdef CONFIG_SMP @@ -345,6 +349,23 @@ check_x87: .byte 0xDB,0xE4 /* fsetpm for 287, ignored by 387 */ ret +/* + * Point the GDT at this CPU's PDA. On boot this will be + * cpu_gdt_table and boot_pda; for secondary CPUs, these will be + * that CPU's GDT and PDA. + */ +setup_pda: + /* get the PDA pointer */ + movl start_pda, %eax + + /* slot the PDA address into the GDT */ + mov cpu_gdt_descr+2, %ecx + mov %ax, (__KERNEL_PDA+0+2)(%ecx) /* base & 0x0000ffff */ + shr $16, %eax + mov %al, (__KERNEL_PDA+4+0)(%ecx) /* base & 0x00ff0000 */ + mov %ah, (__KERNEL_PDA+4+3)(%ecx) /* base & 0xff000000 */ + ret + /* * setup_idt * @@ -484,6 +505,8 @@ ENTRY(empty_zero_page) * This starts the data section. */ .data +ENTRY(start_pda) + .long boot_pda ENTRY(stack_start) .long init_thread_union+THREAD_SIZE @@ -525,7 +548,7 @@ idt_descr: # boot GDT descriptor (later on used by CPU#0): .word 0 # 32 bit align gdt_desc.address -cpu_gdt_descr: +ENTRY(cpu_gdt_descr) .word GDT_ENTRIES*8-1 .long cpu_gdt_table @@ -585,7 +608,7 @@ ENTRY(cpu_gdt_table) .quad 0x004092000000ffff /* 0xc8 APM DS data */ .quad 0x00c0920000000000 /* 0xd0 - ESPFIX SS */ - .quad 0x0000000000000000 /* 0xd8 - PDA */ + .quad 0x00cf92000000ffff /* 0xd8 - PDA */ .quad 0x0000000000000000 /* 0xe0 - unused */ .quad 0x0000000000000000 /* 0xe8 - unused */ .quad 0x0000000000000000 /* 0xf0 - unused */ -- cgit v1.2.3 From c9ccf30d77f04064fe5436027ab9d2230c7cdd94 Mon Sep 17 00:00:00 2001 From: Rusty Russell Date: Thu, 7 Dec 2006 02:14:08 +0100 Subject: [PATCH] paravirt: Add startup infrastructure for paravirtualization 1) Each hypervisor writes a probe function to detect whether we are running under that hypervisor. paravirt_probe() registers this function. 2) If vmlinux is booted with ring != 0, we call all the probe functions (with registers except %esp intact) in link order: the winner will not return. Signed-off-by: Rusty Russell Signed-off-by: Chris Wright Signed-off-by: Andi Kleen Cc: Jeremy Fitzhardinge Cc: Zachary Amsden Signed-off-by: Andrew Morton --- arch/i386/kernel/head.S | 33 +++++++++++++++++++++++++++++++++ 1 file changed, 33 insertions(+) (limited to 'arch/i386/kernel/head.S') diff --git a/arch/i386/kernel/head.S b/arch/i386/kernel/head.S index 5b14e95ac8b9..edef5084ce17 100644 --- a/arch/i386/kernel/head.S +++ b/arch/i386/kernel/head.S @@ -55,6 +55,12 @@ */ ENTRY(startup_32) +#ifdef CONFIG_PARAVIRT + movl %cs, %eax + testl $0x3, %eax + jnz startup_paravirt +#endif + /* * Set segments to known values. */ @@ -486,6 +492,33 @@ ignore_int: #endif iret +#ifdef CONFIG_PARAVIRT +startup_paravirt: + cld + movl $(init_thread_union+THREAD_SIZE),%esp + + /* We take pains to preserve all the regs. */ + pushl %edx + pushl %ecx + pushl %eax + + /* paravirt.o is last in link, and that probe fn never returns */ + pushl $__start_paravirtprobe +1: + movl 0(%esp), %eax + pushl (%eax) + movl 8(%esp), %eax + call *(%esp) + popl %eax + + movl 4(%esp), %eax + movl 8(%esp), %ecx + movl 12(%esp), %edx + + addl $4, (%esp) + jmp 1b +#endif + /* * Real beginning of normal "text" segment */ -- cgit v1.2.3