Subject: kexec: Move asm segment handling code to the assembly file (x86_64)
From: http://xenbits.xensource.com/xen-unstable.hg (tip 13816)
Patch-mainline: n/a

This patch moves the idt, gdt, and segment handling code from machine_kexec.c
to relocate_kernel.S.  The main reason behind this move is to avoid code 
duplication in the Xen hypervisor. With this patch all code required to kexec
is put on the control page.

On top of that this patch also counts as a cleanup - I think it is much
nicer to write assembly directly in assembly files than wrap inline assembly
in C functions for no apparent reason.

Signed-off-by: Magnus Damm <magnus@xxxxxxxxxxxxx>
Acked-by: jbeulich@xxxxxxxxxx

 Applies to 2.6.19-rc1.
 jb: fixed up register usage for 2.6.30 (bnc#545206)

--- head-2011-08-09.orig/arch/x86/kernel/machine_kexec_64.c	2011-08-09 10:21:04.000000000 +0200
+++ head-2011-08-09/arch/x86/kernel/machine_kexec_64.c	2010-04-15 09:38:56.000000000 +0200
@@ -203,47 +203,6 @@ static int init_pgtable(struct kimage *i
 	return init_transition_pgtable(image, level4p);
 }
 
-static void set_idt(void *newidt, u16 limit)
-{
-	struct desc_ptr curidt;
-
-	/* x86-64 supports unaliged loads & stores */
-	curidt.size    = limit;
-	curidt.address = (unsigned long)newidt;
-
-	__asm__ __volatile__ (
-		"lidtq %0\n"
-		: : "m" (curidt)
-		);
-};
-
-
-static void set_gdt(void *newgdt, u16 limit)
-{
-	struct desc_ptr curgdt;
-
-	/* x86-64 supports unaligned loads & stores */
-	curgdt.size    = limit;
-	curgdt.address = (unsigned long)newgdt;
-
-	__asm__ __volatile__ (
-		"lgdtq %0\n"
-		: : "m" (curgdt)
-		);
-};
-
-static void load_segments(void)
-{
-	__asm__ __volatile__ (
-		"\tmovl %0,%%ds\n"
-		"\tmovl %0,%%es\n"
-		"\tmovl %0,%%ss\n"
-		"\tmovl %0,%%fs\n"
-		"\tmovl %0,%%gs\n"
-		: : "a" (__KERNEL_DS) : "memory"
-		);
-}
-
 int machine_kexec_prepare(struct kimage *image)
 {
 	unsigned long start_pgtable;
@@ -311,24 +270,6 @@ void machine_kexec(struct kimage *image)
 		page_list[PA_SWAP_PAGE] = (page_to_pfn(image->swap_page)
 						<< PAGE_SHIFT);
 
-	/*
-	 * The segment registers are funny things, they have both a
-	 * visible and an invisible part.  Whenever the visible part is
-	 * set to a specific selector, the invisible part is loaded
-	 * with from a table in memory.  At no other time is the
-	 * descriptor table in memory accessed.
-	 *
-	 * 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 */
 	image->start = relocate_kernel((unsigned long)image->head,
 				       (unsigned long)page_list,
--- head-2011-08-09.orig/arch/x86/kernel/relocate_kernel_64.S	2011-08-09 10:19:07.000000000 +0200
+++ head-2011-08-09/arch/x86/kernel/relocate_kernel_64.S	2011-08-09 10:21:24.000000000 +0200
@@ -91,13 +91,30 @@ relocate_kernel:
 	/* Switch to the identity mapped page tables */
 	movq	%r9, %cr3
 
+	/* setup idt */
+	lidtq	idt_80 - relocate_kernel(%r8)
+
+	/* setup gdt */
+	leaq	gdt - relocate_kernel(%r8), %rax
+	movq	%rax, (gdt_80 - relocate_kernel) + 2(%r8)
+	lgdtq	gdt_80 - relocate_kernel(%r8)
+
+	/* setup data segment registers */
+	xorl	%eax, %eax
+	movl	%eax, %ds
+	movl	%eax, %es
+	movl	%eax, %fs
+	movl	%eax, %gs
+	movl	%eax, %ss
+
 	/* setup a new stack at the end of the physical control page */
 	lea	PAGE_SIZE(%r8), %rsp
 
-	/* jump to identity mapped page */
+	/* load new code segment and jump to identity mapped page */
 	addq	$(identity_mapped - relocate_kernel), %r8
+	pushq	$(gdt_cs - gdt)
 	pushq	%r8
-	ret
+	lretq
 
 identity_mapped:
 	/* set return address to 0 if not preserving context */
@@ -264,5 +281,20 @@ swap_pages:
 3:
 	ret
 
+	.align  16
+gdt:
+	.quad	0x0000000000000000	/* NULL descriptor */
+gdt_cs:
+	.quad   0x00af9a000000ffff
+gdt_end:
+
+gdt_80:
+	.word	gdt_end - gdt - 1	/* limit */
+	.quad	0			/* base - filled in by code above */
+
+idt_80:
+	.word	0			/* limit */
+	.quad	0			/* base */
+
 	.globl kexec_control_code_size
 .set kexec_control_code_size, . - relocate_kernel