9 files changed, 131 insertions, 22 deletions
diff --git a/arch/x86/crypto/ghash-clmulni-intel_glue.c b/arch/x86/crypto/ghash-clmulni-intel_glue.c
index cbcc8d8ea93a..7a6e68e4f748 100644
--- a/arch/x86/crypto/ghash-clmulni-intel_glue.c
+++ b/arch/x86/crypto/ghash-clmulni-intel_glue.c
@@ -10,6 +10,7 @@
 * by the Free Software Foundation.
 */
+#include <linux/err.h>
 #include <linux/module.h>
 #include <linux/init.h>
 #include <linux/kernel.h>
diff --git a/arch/x86/include/asm/kvm_host.h b/arch/x86/include/asm/kvm_host.h
index 9e6fe391094e..f702f82aa1eb 100644
--- a/arch/x86/include/asm/kvm_host.h
+++ b/arch/x86/include/asm/kvm_host.h
@@ -79,7 +79,7 @@
 #define KVM_NUM_MMU_PAGES (1 << KVM_MMU_HASH_SHIFT)
 #define KVM_MIN_FREE_MMU_PAGES 5
 #define KVM_REFILL_PAGES 25
-#define KVM_MAX_CPUID_ENTRIES 40
+#define KVM_MAX_CPUID_ENTRIES 80
 #define KVM_NR_FIXED_MTRR_REGION 88
 #define KVM_NR_VAR_MTRR 8
diff --git a/arch/x86/kernel/head_32.S b/arch/x86/kernel/head_32.S
index bcece91dd311..f0bea76f6ea5 100644
--- a/arch/x86/kernel/head_32.S
+++ b/arch/x86/kernel/head_32.S
@@ -620,13 +620,13 @@ ENTRY(initial_code)
 __PAGE_ALIGNED_BSS
        .align PAGE_SIZE_asm
 #ifdef CONFIG_X86_PAE
-initial_pg_pmd:
+ENTRY(initial_pg_pmd)
        .fill 1024*KPMDS,4,0
 #else
 ENTRY(initial_page_table)
        .fill 1024,4,0
 #endif
-initial_pg_fixmap:
+ENTRY(initial_pg_fixmap)
        .fill 1024,4,0
 ENTRY(empty_zero_page)
        .fill 4096,1,0
diff --git a/arch/x86/kvm/svm.c b/arch/x86/kvm/svm.c
index 1ca12298ffc7..b81a9b7c2ca4 100644
--- a/arch/x86/kvm/svm.c
+++ b/arch/x86/kvm/svm.c
@@ -3494,6 +3494,10 @@ static void svm_cpuid_update(struct kvm_vcpu *vcpu)
 static void svm_set_supported_cpuid(u32 func, struct kvm_cpuid_entry2 *entry)
 {
        switch (func) {
+        case 0x00000001:
+                /* Mask out xsave bit as long as it is not supported by SVM */
+                entry->ecx &= ~(bit(X86_FEATURE_XSAVE));
+                break;
        case 0x80000001:
                if (nested)
                        entry->ecx |= (1 << 2); /* Set SVM bit */
diff --git a/arch/x86/kvm/vmx.c b/arch/x86/kvm/vmx.c
index ff21fdda0c53..81fcbe9515c5 100644
--- a/arch/x86/kvm/vmx.c
+++ b/arch/x86/kvm/vmx.c
@@ -4227,11 +4227,6 @@ static int vmx_get_lpage_level(void)
                return PT_PDPE_LEVEL;
 }
-static inline u32 bit(int bitno)
-{
-        return 1 << (bitno & 31);
-}
 static void vmx_cpuid_update(struct kvm_vcpu *vcpu)
 {
        struct kvm_cpuid_entry2 *best;
diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c
index cdac9e592aa5..b989e1f1e5d3 100644
--- a/arch/x86/kvm/x86.c
+++ b/arch/x86/kvm/x86.c
@@ -155,11 +155,6 @@ struct kvm_stats_debugfs_item debugfs_entries[] = {
 u64 __read_mostly host_xcr0;
-static inline u32 bit(int bitno)
-{
-        return 1 << (bitno & 31);
-}
 static void kvm_on_user_return(struct user_return_notifier *urn)
 {
        unsigned slot;
@@ -4569,9 +4564,11 @@ static void kvm_timer_init(void)
 #ifdef CONFIG_CPU_FREQ
                struct cpufreq_policy policy;
                memset(&policy, 0, sizeof(policy));
-                cpufreq_get_policy(&policy, get_cpu());
+                cpu = get_cpu();
+                cpufreq_get_policy(&policy, cpu);
                if (policy.cpuinfo.max_freq)
                        max_tsc_khz = policy.cpuinfo.max_freq;
+                put_cpu();
 #endif
                cpufreq_register_notifier(&kvmclock_cpufreq_notifier_block,
                                          CPUFREQ_TRANSITION_NOTIFIER);
@@ -5522,6 +5519,8 @@ int kvm_arch_vcpu_ioctl_set_sregs(struct kvm_vcpu *vcpu,
        mmu_reset_needed |= kvm_read_cr4(vcpu) != sregs->cr4;
        kvm_x86_ops->set_cr4(vcpu, sregs->cr4);
+        if (sregs->cr4 & X86_CR4_OSXSAVE)
+                update_cpuid(vcpu);
        if (!is_long_mode(vcpu) && is_pae(vcpu)) {
                load_pdptrs(vcpu, vcpu->arch.walk_mmu, vcpu->arch.cr3);
                mmu_reset_needed = 1;
diff --git a/arch/x86/kvm/x86.h b/arch/x86/kvm/x86.h
index 2cea414489f3..c600da830ce0 100644
--- a/arch/x86/kvm/x86.h
+++ b/arch/x86/kvm/x86.h
@@ -70,6 +70,11 @@ static inline int is_paging(struct kvm_vcpu *vcpu)
        return kvm_read_cr0_bits(vcpu, X86_CR0_PG);
 }
+static inline u32 bit(int bitno)
+{
+        return 1 << (bitno & 31);
+}
 void kvm_before_handle_nmi(struct kvm_vcpu *vcpu);
 void kvm_after_handle_nmi(struct kvm_vcpu *vcpu);
 int kvm_inject_realmode_interrupt(struct kvm_vcpu *vcpu, int irq);
diff --git a/arch/x86/lguest/boot.c b/arch/x86/lguest/boot.c
index 73b1e1a1f489..4996cf5f73a0 100644
--- a/arch/x86/lguest/boot.c
+++ b/arch/x86/lguest/boot.c
@@ -531,7 +531,10 @@ static void lguest_write_cr3(unsigned long cr3)
 {
        lguest_data.pgdir = cr3;
        lazy_hcall1(LHCALL_NEW_PGTABLE, cr3);
-        cr3_changed = true;
+        /* These two page tables are simple, linear, and used during boot */
+        if (cr3 != __pa(swapper_pg_dir) && cr3 != __pa(initial_page_table))
+                cr3_changed = true;
 }
 static unsigned long lguest_read_cr3(void)
@@ -703,9 +706,9 @@ static void lguest_set_pmd(pmd_t *pmdp, pmd_t pmdval)
 * to forget all of them.  Fortunately, this is very rare.
 *
 * ... except in early boot when the kernel sets up the initial pagetables,
- * which makes booting astonishingly slow: 1.83 seconds!  So we don't even tell
+ * which makes booting astonishingly slow: 48 seconds!  So we don't even tell
- * the Host anything changed until we've done the first page table switch,
+ * the Host anything changed until we've done the first real page table switch,
- * which brings boot back to 0.25 seconds.
+ * which brings boot back to 4.3 seconds.
 */
 static void lguest_set_pte(pte_t *ptep, pte_t pteval)
 {
@@ -1002,7 +1005,7 @@ static void lguest_time_init(void)
        clockevents_register_device(&lguest_clockevent);
        /* Finally, we unblock the timer interrupt. */
-        enable_lguest_irq(0);
+        clear_bit(0, lguest_data.blocked_interrupts);
 }
 /*
@@ -1349,9 +1352,6 @@ __init void lguest_init(void)
         */
        switch_to_new_gdt(0);
-        /* We actually boot with all memory mapped, but let's say 128MB. */
-        max_pfn_mapped = (128*1024*1024) >> PAGE_SHIFT;
        /*
         * The Host<->Guest Switcher lives at the top of our address space, and
         * the Host told us how big it is when we made LGUEST_INIT hypercall:
diff --git a/arch/x86/lguest/i386_head.S b/arch/x86/lguest/i386_head.S
index 4f420c2f2d55..e7d5382ef263 100644
--- a/arch/x86/lguest/i386_head.S
+++ b/arch/x86/lguest/i386_head.S
@@ -4,6 +4,7 @@
 #include <asm/asm-offsets.h>
 #include <asm/thread_info.h>
 #include <asm/processor-flags.h>
+#include <asm/pgtable.h>
 /*G:020
 * Our story starts with the kernel booting into startup_32 in
@@ -37,9 +38,113 @@ ENTRY(lguest_entry)
        /* Set up the initial stack so we can run C code. */
        movl $(init_thread_union+THREAD_SIZE),%esp
+        call init_pagetables
        /* Jumps are relative: we're running __PAGE_OFFSET too low. */
        jmp lguest_init+__PAGE_OFFSET
+/*
+ * Initialize page tables.  This creates a PDE and a set of page
+ * tables, which are located immediately beyond __brk_base.  The variable
+ * _brk_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.
+ *
+ * FIXME: This code is taken verbatim from arch/x86/kernel/head_32.S: they
+ * don't have a stack at this point, so we can't just use call and ret.
+ */
+init_pagetables:
+#if PTRS_PER_PMD > 1
+#define PAGE_TABLE_SIZE(pages) (((pages) / PTRS_PER_PMD) + PTRS_PER_PGD)
+#else
+#define PAGE_TABLE_SIZE(pages) ((pages) / PTRS_PER_PGD)
+#endif
+#define pa(X) ((X) - __PAGE_OFFSET)
+/* Enough space to fit pagetables for the low memory linear map */
+MAPPING_BEYOND_END = \
+        PAGE_TABLE_SIZE(((1<<32) - __PAGE_OFFSET) >> PAGE_SHIFT) << PAGE_SHIFT
+#ifdef CONFIG_X86_PAE
+        /*
+         * In PAE mode initial_page_table 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.
+         */
+#define KPMDS (((-__PAGE_OFFSET) >> 30) & 3) /* Number of kernel PMDs */
+        xorl %ebx,%ebx                          /* %ebx is kept at zero */
+        movl $pa(__brk_base), %edi
+        movl $pa(initial_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 the end + MAPPING_BEYOND_END.
+         */
+        movl $pa(_end) + MAPPING_BEYOND_END + PTE_IDENT_ATTR, %ebp
+        cmpl %ebp,%eax
+        jb 10b
+1:
+        addl $__PAGE_OFFSET, %edi
+        movl %edi, pa(_brk_end)
+        shrl $12, %eax
+        movl %eax, pa(max_pfn_mapped)
+        /* Do early initialization of the fixmap area */
+        movl $pa(initial_pg_fixmap)+PDE_IDENT_ATTR,%eax
+        movl %eax,pa(initial_pg_pmd+0x1000*KPMDS-8)
+#else   /* Not PAE */
+page_pde_offset = (__PAGE_OFFSET >> 20);
+        movl $pa(__brk_base), %edi
+        movl $pa(initial_page_table), %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 the end + MAPPING_BEYOND_END.
+         */
+        movl $pa(_end) + MAPPING_BEYOND_END + PTE_IDENT_ATTR, %ebp
+        cmpl %ebp,%eax
+        jb 10b
+        addl $__PAGE_OFFSET, %edi
+        movl %edi, pa(_brk_end)
+        shrl $12, %eax
+        movl %eax, pa(max_pfn_mapped)
+        /* Do early initialization of the fixmap area */
+        movl $pa(initial_pg_fixmap)+PDE_IDENT_ATTR,%eax
+        movl %eax,pa(initial_page_table+0xffc)
+#endif
+        ret
 /*G:055
 * We create a macro which puts the assembler code between lgstart_ and lgend_
 * markers.  These templates are put in the .text section: they can't be

diff --git a/arch/x86/crypto/ghash-clmulni-intel_glue.c b/arch/x86/crypto/ghash-clmulni-intel_glue.c index cbcc8d8ea93a..7a6e68e4f748 100644 --- a/arch/x86/crypto/ghash-clmulni-intel_glue.c +++ b/arch/x86/crypto/ghash-clmulni-intel_glue.c
@@ -10,6 +10,7 @@
10	* by the Free Software Foundation.	10	* by the Free Software Foundation.
11	*/	11	*/
12		12
		13	#include <linux/err.h>
13	#include <linux/module.h>	14	#include <linux/module.h>
14	#include <linux/init.h>	15	#include <linux/init.h>
15	#include <linux/kernel.h>	16	#include <linux/kernel.h>


diff --git a/arch/x86/include/asm/kvm_host.h b/arch/x86/include/asm/kvm_host.h index 9e6fe391094e..f702f82aa1eb 100644 --- a/arch/x86/include/asm/kvm_host.h +++ b/arch/x86/include/asm/kvm_host.h
@@ -79,7 +79,7 @@
79	#define KVM_NUM_MMU_PAGES (1 << KVM_MMU_HASH_SHIFT)	79	#define KVM_NUM_MMU_PAGES (1 << KVM_MMU_HASH_SHIFT)
80	#define KVM_MIN_FREE_MMU_PAGES 5	80	#define KVM_MIN_FREE_MMU_PAGES 5
81	#define KVM_REFILL_PAGES 25	81	#define KVM_REFILL_PAGES 25
82	#define KVM_MAX_CPUID_ENTRIES 40	82	#define KVM_MAX_CPUID_ENTRIES 80
83	#define KVM_NR_FIXED_MTRR_REGION 88	83	#define KVM_NR_FIXED_MTRR_REGION 88
84	#define KVM_NR_VAR_MTRR 8	84	#define KVM_NR_VAR_MTRR 8
85		85


diff --git a/arch/x86/kernel/head_32.S b/arch/x86/kernel/head_32.S index bcece91dd311..f0bea76f6ea5 100644 --- a/arch/x86/kernel/head_32.S +++ b/arch/x86/kernel/head_32.S
@@ -620,13 +620,13 @@ ENTRY(initial_code)
620	__PAGE_ALIGNED_BSS	620	__PAGE_ALIGNED_BSS
621	.align PAGE_SIZE_asm	621	.align PAGE_SIZE_asm
622	#ifdef CONFIG_X86_PAE	622	#ifdef CONFIG_X86_PAE
623	initial_pg_pmd:	623	ENTRY(initial_pg_pmd)
624	.fill 1024*KPMDS,4,0	624	.fill 1024*KPMDS,4,0
625	#else	625	#else
626	ENTRY(initial_page_table)	626	ENTRY(initial_page_table)
627	.fill 1024,4,0	627	.fill 1024,4,0
628	#endif	628	#endif
629	initial_pg_fixmap:	629	ENTRY(initial_pg_fixmap)
630	.fill 1024,4,0	630	.fill 1024,4,0
631	ENTRY(empty_zero_page)	631	ENTRY(empty_zero_page)
632	.fill 4096,1,0	632	.fill 4096,1,0


diff --git a/arch/x86/kvm/svm.c b/arch/x86/kvm/svm.c index 1ca12298ffc7..b81a9b7c2ca4 100644 --- a/arch/x86/kvm/svm.c +++ b/arch/x86/kvm/svm.c
@@ -3494,6 +3494,10 @@ static void svm_cpuid_update(struct kvm_vcpu *vcpu)
3494	static void svm_set_supported_cpuid(u32 func, struct kvm_cpuid_entry2 *entry)	3494	static void svm_set_supported_cpuid(u32 func, struct kvm_cpuid_entry2 *entry)
3495	{	3495	{
3496	switch (func) {	3496	switch (func) {
		3497	case 0x00000001:
		3498	/* Mask out xsave bit as long as it is not supported by SVM */
		3499	entry->ecx &= ~(bit(X86_FEATURE_XSAVE));
		3500	break;
3497	case 0x80000001:	3501	case 0x80000001:
3498	if (nested)	3502	if (nested)
3499	entry->ecx \|= (1 << 2); /* Set SVM bit */	3503	entry->ecx \|= (1 << 2); /* Set SVM bit */


diff --git a/arch/x86/kvm/vmx.c b/arch/x86/kvm/vmx.c index ff21fdda0c53..81fcbe9515c5 100644 --- a/arch/x86/kvm/vmx.c +++ b/arch/x86/kvm/vmx.c
@@ -4227,11 +4227,6 @@ static int vmx_get_lpage_level(void)
4227	return PT_PDPE_LEVEL;	4227	return PT_PDPE_LEVEL;
4228	}	4228	}
4229		4229
4230	static inline u32 bit(int bitno)
4231	{
4232	return 1 << (bitno & 31);
4233	}
4234
4235	static void vmx_cpuid_update(struct kvm_vcpu *vcpu)	4230	static void vmx_cpuid_update(struct kvm_vcpu *vcpu)
4236	{	4231	{
4237	struct kvm_cpuid_entry2 *best;	4232	struct kvm_cpuid_entry2 *best;


diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c index cdac9e592aa5..b989e1f1e5d3 100644 --- a/arch/x86/kvm/x86.c +++ b/arch/x86/kvm/x86.c
@@ -155,11 +155,6 @@ struct kvm_stats_debugfs_item debugfs_entries[] = {
155		155
156	u64 __read_mostly host_xcr0;	156	u64 __read_mostly host_xcr0;
157		157
158	static inline u32 bit(int bitno)
159	{
160	return 1 << (bitno & 31);
161	}
162
163	static void kvm_on_user_return(struct user_return_notifier *urn)	158	static void kvm_on_user_return(struct user_return_notifier *urn)
164	{	159	{
165	unsigned slot;	160	unsigned slot;
@@ -4569,9 +4564,11 @@ static void kvm_timer_init(void)
4569	#ifdef CONFIG_CPU_FREQ	4564	#ifdef CONFIG_CPU_FREQ
4570	struct cpufreq_policy policy;	4565	struct cpufreq_policy policy;
4571	memset(&policy, 0, sizeof(policy));	4566	memset(&policy, 0, sizeof(policy));
4572	cpufreq_get_policy(&policy, get_cpu());	4567	cpu = get_cpu();
		4568	cpufreq_get_policy(&policy, cpu);
4573	if (policy.cpuinfo.max_freq)	4569	if (policy.cpuinfo.max_freq)
4574	max_tsc_khz = policy.cpuinfo.max_freq;	4570	max_tsc_khz = policy.cpuinfo.max_freq;
		4571	put_cpu();
4575	#endif	4572	#endif
4576	cpufreq_register_notifier(&kvmclock_cpufreq_notifier_block,	4573	cpufreq_register_notifier(&kvmclock_cpufreq_notifier_block,
4577	CPUFREQ_TRANSITION_NOTIFIER);	4574	CPUFREQ_TRANSITION_NOTIFIER);
@@ -5522,6 +5519,8 @@ int kvm_arch_vcpu_ioctl_set_sregs(struct kvm_vcpu *vcpu,
5522		5519
5523	mmu_reset_needed \|= kvm_read_cr4(vcpu) != sregs->cr4;	5520	mmu_reset_needed \|= kvm_read_cr4(vcpu) != sregs->cr4;
5524	kvm_x86_ops->set_cr4(vcpu, sregs->cr4);	5521	kvm_x86_ops->set_cr4(vcpu, sregs->cr4);
		5522	if (sregs->cr4 & X86_CR4_OSXSAVE)
		5523	update_cpuid(vcpu);
5525	if (!is_long_mode(vcpu) && is_pae(vcpu)) {	5524	if (!is_long_mode(vcpu) && is_pae(vcpu)) {
5526	load_pdptrs(vcpu, vcpu->arch.walk_mmu, vcpu->arch.cr3);	5525	load_pdptrs(vcpu, vcpu->arch.walk_mmu, vcpu->arch.cr3);
5527	mmu_reset_needed = 1;	5526	mmu_reset_needed = 1;


diff --git a/arch/x86/kvm/x86.h b/arch/x86/kvm/x86.h index 2cea414489f3..c600da830ce0 100644 --- a/arch/x86/kvm/x86.h +++ b/arch/x86/kvm/x86.h
@@ -70,6 +70,11 @@ static inline int is_paging(struct kvm_vcpu *vcpu)
70	return kvm_read_cr0_bits(vcpu, X86_CR0_PG);	70	return kvm_read_cr0_bits(vcpu, X86_CR0_PG);
71	}	71	}
72		72
		73	static inline u32 bit(int bitno)
		74	{
		75	return 1 << (bitno & 31);
		76	}
		77
73	void kvm_before_handle_nmi(struct kvm_vcpu *vcpu);	78	void kvm_before_handle_nmi(struct kvm_vcpu *vcpu);
74	void kvm_after_handle_nmi(struct kvm_vcpu *vcpu);	79	void kvm_after_handle_nmi(struct kvm_vcpu *vcpu);
75	int kvm_inject_realmode_interrupt(struct kvm_vcpu *vcpu, int irq);	80	int kvm_inject_realmode_interrupt(struct kvm_vcpu *vcpu, int irq);


diff --git a/arch/x86/lguest/boot.c b/arch/x86/lguest/boot.c index 73b1e1a1f489..4996cf5f73a0 100644 --- a/arch/x86/lguest/boot.c +++ b/arch/x86/lguest/boot.c
@@ -531,7 +531,10 @@ static void lguest_write_cr3(unsigned long cr3)
531	{	531	{
532	lguest_data.pgdir = cr3;	532	lguest_data.pgdir = cr3;
533	lazy_hcall1(LHCALL_NEW_PGTABLE, cr3);	533	lazy_hcall1(LHCALL_NEW_PGTABLE, cr3);
534	cr3_changed = true;	534
		535	/* These two page tables are simple, linear, and used during boot */
		536	if (cr3 != __pa(swapper_pg_dir) && cr3 != __pa(initial_page_table))
		537	cr3_changed = true;
535	}	538	}
536		539
537	static unsigned long lguest_read_cr3(void)	540	static unsigned long lguest_read_cr3(void)
@@ -703,9 +706,9 @@ static void lguest_set_pmd(pmd_t *pmdp, pmd_t pmdval)
703	* to forget all of them. Fortunately, this is very rare.	706	* to forget all of them. Fortunately, this is very rare.
704	*	707	*
705	* ... except in early boot when the kernel sets up the initial pagetables,	708	* ... except in early boot when the kernel sets up the initial pagetables,
706	* which makes booting astonishingly slow: 1.83 seconds! So we don't even tell	709	* which makes booting astonishingly slow: 48 seconds! So we don't even tell
707	* the Host anything changed until we've done the first page table switch,	710	* the Host anything changed until we've done the first real page table switch,
708	* which brings boot back to 0.25 seconds.	711	* which brings boot back to 4.3 seconds.
709	*/	712	*/
710	static void lguest_set_pte(pte_t *ptep, pte_t pteval)	713	static void lguest_set_pte(pte_t *ptep, pte_t pteval)
711	{	714	{
@@ -1002,7 +1005,7 @@ static void lguest_time_init(void)
1002	clockevents_register_device(&lguest_clockevent);	1005	clockevents_register_device(&lguest_clockevent);
1003		1006
1004	/* Finally, we unblock the timer interrupt. */	1007	/* Finally, we unblock the timer interrupt. */
1005	enable_lguest_irq(0);	1008	clear_bit(0, lguest_data.blocked_interrupts);
1006	}	1009	}
1007		1010
1008	/*	1011	/*
@@ -1349,9 +1352,6 @@ __init void lguest_init(void)
1349	*/	1352	*/
1350	switch_to_new_gdt(0);	1353	switch_to_new_gdt(0);
1351		1354
1352	/* We actually boot with all memory mapped, but let's say 128MB. */
1353	max_pfn_mapped = (12810241024) >> PAGE_SHIFT;
1354
1355	/*	1355	/*
1356	* The Host<->Guest Switcher lives at the top of our address space, and	1356	* The Host<->Guest Switcher lives at the top of our address space, and
1357	* the Host told us how big it is when we made LGUEST_INIT hypercall:	1357	* the Host told us how big it is when we made LGUEST_INIT hypercall:


diff --git a/arch/x86/lguest/i386_head.S b/arch/x86/lguest/i386_head.S index 4f420c2f2d55..e7d5382ef263 100644 --- a/arch/x86/lguest/i386_head.S +++ b/arch/x86/lguest/i386_head.S
@@ -4,6 +4,7 @@
4	#include <asm/asm-offsets.h>	4	#include <asm/asm-offsets.h>
5	#include <asm/thread_info.h>	5	#include <asm/thread_info.h>
6	#include <asm/processor-flags.h>	6	#include <asm/processor-flags.h>
		7	#include <asm/pgtable.h>
7		8
8	/*G:020	9	/*G:020
9	* Our story starts with the kernel booting into startup_32 in	10	* Our story starts with the kernel booting into startup_32 in
@@ -37,9 +38,113 @@ ENTRY(lguest_entry)
37	/* Set up the initial stack so we can run C code. */	38	/* Set up the initial stack so we can run C code. */
38	movl $(init_thread_union+THREAD_SIZE),%esp	39	movl $(init_thread_union+THREAD_SIZE),%esp
39		40
		41	call init_pagetables
		42
40	/* Jumps are relative: we're running __PAGE_OFFSET too low. */	43	/* Jumps are relative: we're running __PAGE_OFFSET too low. */
41	jmp lguest_init+__PAGE_OFFSET	44	jmp lguest_init+__PAGE_OFFSET
42		45
		46	/*
		47	* Initialize page tables. This creates a PDE and a set of page
		48	* tables, which are located immediately beyond __brk_base. The variable
		49	* _brk_end is set up to point to the first "safe" location.
		50	* Mappings are created both at virtual address 0 (identity mapping)
		51	* and PAGE_OFFSET for up to _end.
		52	*
		53	* FIXME: This code is taken verbatim from arch/x86/kernel/head_32.S: they
		54	* don't have a stack at this point, so we can't just use call and ret.
		55	*/
		56	init_pagetables:
		57	#if PTRS_PER_PMD > 1
		58	#define PAGE_TABLE_SIZE(pages) (((pages) / PTRS_PER_PMD) + PTRS_PER_PGD)
		59	#else
		60	#define PAGE_TABLE_SIZE(pages) ((pages) / PTRS_PER_PGD)
		61	#endif
		62	#define pa(X) ((X) - __PAGE_OFFSET)
		63
		64	/* Enough space to fit pagetables for the low memory linear map */
		65	MAPPING_BEYOND_END = \
		66	PAGE_TABLE_SIZE(((1<<32) - __PAGE_OFFSET) >> PAGE_SHIFT) << PAGE_SHIFT
		67	#ifdef CONFIG_X86_PAE
		68
		69	/*
		70	* In PAE mode initial_page_table is statically defined to contain
		71	* enough entries to cover the VMSPLIT option (that is the top 1, 2 or 3
		72	* entries). The identity mapping is handled by pointing two PGD entries
		73	* to the first kernel PMD.
		74	*
		75	* Note the upper half of each PMD or PTE are always zero at this stage.
		76	*/
		77
		78	#define KPMDS (((-__PAGE_OFFSET) >> 30) & 3) /* Number of kernel PMDs */
		79
		80	xorl %ebx,%ebx /* %ebx is kept at zero */
		81
		82	movl $pa(__brk_base), %edi
		83	movl $pa(initial_pg_pmd), %edx
		84	movl $PTE_IDENT_ATTR, %eax
		85	10:
		86	leal PDE_IDENT_ATTR(%edi),%ecx /* Create PMD entry */
		87	movl %ecx,(%edx) /* Store PMD entry */
		88	/* Upper half already zero */
		89	addl $8,%edx
		90	movl $512,%ecx
		91	11:
		92	stosl
		93	xchgl %eax,%ebx
		94	stosl
		95	xchgl %eax,%ebx
		96	addl $0x1000,%eax
		97	loop 11b
		98
		99	/*
		100	* End condition: we must map up to the end + MAPPING_BEYOND_END.
		101	*/
		102	movl $pa(_end) + MAPPING_BEYOND_END + PTE_IDENT_ATTR, %ebp
		103	cmpl %ebp,%eax
		104	jb 10b
		105	1:
		106	addl $__PAGE_OFFSET, %edi
		107	movl %edi, pa(_brk_end)
		108	shrl $12, %eax
		109	movl %eax, pa(max_pfn_mapped)
		110
		111	/* Do early initialization of the fixmap area */
		112	movl $pa(initial_pg_fixmap)+PDE_IDENT_ATTR,%eax
		113	movl %eax,pa(initial_pg_pmd+0x1000*KPMDS-8)
		114	#else /* Not PAE */
		115
		116	page_pde_offset = (__PAGE_OFFSET >> 20);
		117
		118	movl $pa(__brk_base), %edi
		119	movl $pa(initial_page_table), %edx
		120	movl $PTE_IDENT_ATTR, %eax
		121	10:
		122	leal PDE_IDENT_ATTR(%edi),%ecx /* Create PDE entry */
		123	movl %ecx,(%edx) /* Store identity PDE entry */
		124	movl %ecx,page_pde_offset(%edx) /* Store kernel PDE entry */
		125	addl $4,%edx
		126	movl $1024, %ecx
		127	11:
		128	stosl
		129	addl $0x1000,%eax
		130	loop 11b
		131	/*
		132	* End condition: we must map up to the end + MAPPING_BEYOND_END.
		133	*/
		134	movl $pa(_end) + MAPPING_BEYOND_END + PTE_IDENT_ATTR, %ebp
		135	cmpl %ebp,%eax
		136	jb 10b
		137	addl $__PAGE_OFFSET, %edi
		138	movl %edi, pa(_brk_end)
		139	shrl $12, %eax
		140	movl %eax, pa(max_pfn_mapped)
		141
		142	/* Do early initialization of the fixmap area */
		143	movl $pa(initial_pg_fixmap)+PDE_IDENT_ATTR,%eax
		144	movl %eax,pa(initial_page_table+0xffc)
		145	#endif
		146	ret
		147
43	/*G:055	148	/*G:055
44	* We create a macro which puts the assembler code between lgstart_ and lgend_	149	* We create a macro which puts the assembler code between lgstart_ and lgend_
45	* markers. These templates are put in the .text section: they can't be	150	* markers. These templates are put in the .text section: they can't be