[PATCH 1/7] cleanup: paravirt unhandled fallthrough

[PATCH 1/7] cleanup: paravirt unhandled fallthrough

[Rusty Russell]

The current code simply calls "start_kernel" directly if we're under a
hypervisor and no paravirt_ops backend wants us, because paravirt.c
registers that as a backend.

This was always a vain hope; start_kernel won't get far without setup.
It's also impossible for paravirt_ops backends which don't sit in the
arch/i386/kernel directory: they can't link before paravirt.o anyway.

Keep it simple: if we pass all the registered paravirt probes, BUG().

Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>

diff -r 4fb7fc327801 arch/i386/kernel/Makefile
--- a/arch/i386/kernel/Makefile	Mon Feb 12 12:37:01 2007 +1100
+++ b/arch/i386/kernel/Makefile	Mon Feb 12 12:55:00 2007 +1100
@@ -41,8 +41,6 @@ obj-$(CONFIG_KWATCH)		+= debugreg.o kwat
 obj-$(CONFIG_KWATCH)		+= debugreg.o kwatch.o
 
 obj-$(CONFIG_VMI)		+= vmi.o vmitime.o
-
-# Make sure this is linked after any other paravirt_ops structs: see head.S
 obj-$(CONFIG_PARAVIRT)		+= paravirt.o
 
 EXTRA_AFLAGS   := -traditional
diff -r 4fb7fc327801 arch/i386/kernel/head.S
--- a/arch/i386/kernel/head.S	Mon Feb 12 12:37:01 2007 +1100
+++ b/arch/i386/kernel/head.S	Mon Feb 12 12:54:19 2007 +1100
@@ -513,10 +513,11 @@ startup_paravirt:
 	pushl	%ecx
 	pushl	%eax
 
-	/* paravirt.o is last in link, and that probe fn never returns */
 	pushl	$__start_paravirtprobe
 1:
 	movl	0(%esp), %eax
+	cmpl	$__stop_paravirtprobe, %eax
+	je	unhandled_paravirt
 	pushl	(%eax)
 	movl	8(%esp), %eax
 	call	*(%esp)
@@ -528,6 +529,10 @@ 1:
 
 	addl	$4, (%esp)
 	jmp	1b
+
+unhandled_paravirt:
+	/* Nothing wanted us: we're screwed. */ 
+	ud2
 #endif
 
 /*
diff -r 4fb7fc327801 arch/i386/kernel/paravirt.c
--- a/arch/i386/kernel/paravirt.c	Mon Feb 12 12:37:01 2007 +1100
+++ b/arch/i386/kernel/paravirt.c	Mon Feb 12 12:54:19 2007 +1100
@@ -481,9 +481,6 @@ static int __init print_banner(void)
 	return 0;
 }
 core_initcall(print_banner);
-
-/* We simply declare start_kernel to be the paravirt probe of last resort. */
-paravirt_probe(start_kernel);
 
 struct paravirt_ops paravirt_ops = {
 	.name = "bare hardware",

[PATCH 2/7] cleanup: Initialize esp0 properly all the time

[Rusty Russell]

Whenever we schedule, __switch_to calls load_esp0 which does:

	tss->esp0 = thread->esp0;

This is never initialized for the initial thread (ie "swapper"), so
when we're scheduling that, we end up setting esp0 to 0.  This is
fine: the swapper never leaves ring 0, so this field is never used.

lguest, however, gets upset that we're trying to used an unmapped page
as our kernel stack.  Rather than work around it there, let's
initialize it.

Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>

===================================================================
--- a/include/asm-i386/processor.h
+++ b/include/asm-i386/processor.h
@@ -421,6 +421,7 @@ struct thread_struct {
 };
 
 #define INIT_THREAD  {							\
+	.esp0 = sizeof(init_stack) + (long)&init_stack,			\
 	.vm86_info = NULL,						\
 	.sysenter_cs = __KERNEL_CS,					\
 	.io_bitmap_ptr = NULL,						\

[PATCH 3/7] cleanup: Make hvc_console.c compile on non-PowerPC

[Rusty Russell]

There's a really nice console helper (esp. for virtual console
drivers) in drivers/char/hvc_console.c.  It has only ever been used
for PowerPC, though, so it uses NO_IRQ which is only defined there.

Let's fix that so it's more widely useful.  By, say, lguest.

Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>

===================================================================
--- a/drivers/char/hvc_console.c
+++ b/drivers/char/hvc_console.c
@@ -48,6 +48,10 @@
 #define HVC_MINOR	0
 
 #define TIMEOUT		(10)
+
+#ifndef NO_IRQ
+#define NO_IRQ 0
+#endif
 
 /*
  * Wait this long per iteration while trying to push buffered data to the

[PATCH 4/7] cleanup: Move mce_disabled to asm/mce.h

[Rusty Russell]

Allows external actors to disable mce.

Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>

===================================================================
--- a/arch/i386/kernel/cpu/mcheck/mce.h
+++ b/arch/i386/kernel/cpu/mcheck/mce.h
@@ -1,4 +1,5 @@
 #include <linux/init.h>
+#include <asm/mce.h>
 
 void amd_mcheck_init(struct cpuinfo_x86 *c);
 void intel_p4_mcheck_init(struct cpuinfo_x86 *c);
@@ -9,6 +10,5 @@ void winchip_mcheck_init(struct cpuinfo_
 /* Call the installed machine check handler for this CPU setup. */
 extern fastcall void (*machine_check_vector)(struct pt_regs *, long error_code);
 
-extern int mce_disabled;
 extern int nr_mce_banks;
 
===================================================================
--- a/include/asm-i386/mce.h
+++ b/include/asm-i386/mce.h
@@ -3,3 +3,5 @@ extern void mcheck_init(struct cpuinfo_x
 #else
 #define mcheck_init(c) do {} while(0)
 #endif
+
+extern int mce_disabled;

[PATCH 5/7] cleanup: Rename cpu_gdt_descr and remove extern declaration from smpboot.c

[Rusty Russell]

When I implemented the DECLARE_PER_CPU(var) macros, I was careful that
people couldn't use "var" in a non-percpu context, by prepending
percpu__.  I never considered that this would allow them to overload
the same name for a per-cpu and a non-percpu variable.

It is only one of many horrors in the i386 boot code, but let's rename
the non-perpcu cpu_gdt_descr to early_gdt_descr (not boot_gdt_descr,
that's something else...)

Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>

===================================================================
--- a/arch/i386/kernel/head.S
+++ b/arch/i386/kernel/head.S
@@ -309,7 +309,7 @@ 2:	movl %cr0,%eax
 
 	call check_x87
 	call setup_pda
-	lgdt cpu_gdt_descr
+	lgdt early_gdt_descr
 	lidt idt_descr
 	ljmp $(__KERNEL_CS),$1f
 1:	movl $(__KERNEL_DS),%eax	# reload all the segment registers
@@ -365,7 +365,7 @@ setup_pda:
 	movl start_pda, %eax
 
 	/* slot the PDA address into the GDT */
-	mov cpu_gdt_descr+2, %ecx
+	mov early_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 */
@@ -588,7 +588,7 @@ idt_descr:
 
 # boot GDT descriptor (later on used by CPU#0):
 	.word 0				# 32 bit align gdt_desc.address
-ENTRY(cpu_gdt_descr)
+ENTRY(early_gdt_descr)
 	.word GDT_ENTRIES*8-1
 	.long cpu_gdt_table
 
===================================================================
--- a/arch/i386/kernel/smpboot.c
+++ b/arch/i386/kernel/smpboot.c
@@ -619,7 +619,6 @@ extern struct {
 	unsigned short ss;
 } stack_start;
 extern struct i386_pda *start_pda;
-extern struct Xgt_desc_struct cpu_gdt_descr;
 
 #ifdef CONFIG_NUMA
 
===================================================================
--- a/include/asm-i386/desc.h
+++ b/include/asm-i386/desc.h
@@ -22,7 +22,7 @@ struct Xgt_desc_struct {
 
 extern struct Xgt_desc_struct idt_descr;
 DECLARE_PER_CPU(struct Xgt_desc_struct, cpu_gdt_descr);
-
+extern struct Xgt_desc_struct early_gdt_descr;
 
 static inline struct desc_struct *get_cpu_gdt_table(unsigned int cpu)
 {

[PATCH 6/7] cleanup: Remove extern declaration from mm/discontig.c, put in header.

[Rusty Russell]

Extern declarations belong in headers.  Times, they are a'changin.

Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>

===================================================================
--- a/arch/i386/mm/discontig.c
+++ b/arch/i386/mm/discontig.c
@@ -101,7 +101,6 @@ extern void add_one_highpage_init(struct
 extern void add_one_highpage_init(struct page *, int, int);
 
 extern struct e820map e820;
-extern unsigned long init_pg_tables_end;
 extern unsigned long highend_pfn, highstart_pfn;
 extern unsigned long max_low_pfn;
 extern unsigned long totalram_pages;
===================================================================
--- a/include/asm-i386/setup.h
+++ b/include/asm-i386/setup.h
@@ -77,6 +77,8 @@ void __init add_memory_region(unsigned l
 void __init add_memory_region(unsigned long long start,
 			      unsigned long long size, int type);
 
+extern unsigned long init_pg_tables_end;
+
 #endif /* __ASSEMBLY__ */
 
 #endif  /*  __KERNEL__  */

[PATCH 7/7] cleanup: make disable_acpi() valid w/o CONFIG_ACPI

[Rusty Russell]

Len Brown <lenb@kernel.org> said:
> Okay, but better to use disable_acpi()
> indeed, since this would be the first code not already inside CONFIG_ACPI
> to invoke disable_acpi(), we could define the inline as empty and you could
> then scratch the #ifdef too.

Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>

===================================================================
--- a/include/asm-i386/acpi.h
+++ b/include/asm-i386/acpi.h
@@ -127,6 +127,7 @@ extern int acpi_irq_balance_set(char *st
 #define acpi_ioapic 0
 static inline void acpi_noirq_set(void) { }
 static inline void acpi_disable_pci(void) { }
+static inline void disable_acpi(void) { }
 
 #endif	/* !CONFIG_ACPI */
 

[PATCH 1/2] lguest preparation: EXPORT_SYMBOL_GPL 5 functions

[Rusty Russell]

lguest does some fairly lowlevel things to support a host, which
normal modules don't need:

math_state_restore:
	When the guest triggers a Device Not Available fault, we need
	to be able to restore the FPU

__put_task_struct:
	We need to hold a reference to another task for inter-guest
	I/O, and put_task_struct() is an inline function which calls
	__put_task_struct.

access_process_vm:
	We need to access another task for inter-guest I/O.

map_vm_area & __get_vm_area:
	We need to map the switcher shim (ie. monitor) at 0xFFC01000.

Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>

===================================================================
--- a/arch/i386/kernel/traps.c
+++ b/arch/i386/kernel/traps.c
@@ -1054,6 +1054,7 @@ asmlinkage void math_state_restore(void)
 	thread->status |= TS_USEDFPU;	/* So we fnsave on switch_to() */
 	tsk->fpu_counter++;
 }
+EXPORT_SYMBOL_GPL(math_state_restore);
 
 #ifndef CONFIG_MATH_EMULATION
 
===================================================================
--- a/kernel/fork.c
+++ b/kernel/fork.c
@@ -126,6 +126,7 @@ void __put_task_struct(struct task_struc
 	if (!profile_handoff_task(tsk))
 		free_task(tsk);
 }
+EXPORT_SYMBOL_GPL(__put_task_struct);
 
 void __init fork_init(unsigned long mempages)
 {
===================================================================
--- a/mm/memory.c
+++ b/mm/memory.c
@@ -2692,3 +2692,4 @@ int access_process_vm(struct task_struct
 
 	return buf - old_buf;
 }
+EXPORT_SYMBOL_GPL(access_process_vm);
===================================================================
--- a/mm/vmalloc.c
+++ b/mm/vmalloc.c
@@ -159,6 +159,7 @@ int map_vm_area(struct vm_struct *area, 
 	flush_cache_vmap((unsigned long) area->addr, end);
 	return err;
 }
+EXPORT_SYMBOL_GPL(map_vm_area);
 
 static struct vm_struct *__get_vm_area_node(unsigned long size, unsigned long flags,
 					    unsigned long start, unsigned long end,
@@ -237,6 +238,7 @@ struct vm_struct *__get_vm_area(unsigned
 {
 	return __get_vm_area_node(size, flags, start, end, -1, GFP_KERNEL);
 }
+EXPORT_SYMBOL_GPL(__get_vm_area);
 
 /**
  *	get_vm_area  -  reserve a contingous kernel virtual area

[PATCH 2/2] lguest preparation: expose futex infrastructure: get_futex_key, get_key_refs and drop_key_refs

[Rusty Russell]

lguest uses the convenient futex infrastructure for inter-domain I/O,
so expose get_futex_key, get_key_refs (renamed get_futex_key_refs) and
drop_key_refs (renamed drop_futex_key_refs).  Also means we need to
expose the union that these use.

No code changes.

Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>

===================================================================
--- a/include/linux/futex.h
+++ b/include/linux/futex.h
@@ -100,6 +100,35 @@ extern int
 extern int
 handle_futex_death(u32 __user *uaddr, struct task_struct *curr, int pi);
 
+/*
+ * Futexes are matched on equal values of this key.
+ * The key type depends on whether it's a shared or private mapping.
+ * Don't rearrange members without looking at hash_futex().
+ *
+ * offset is aligned to a multiple of sizeof(u32) (== 4) by definition.
+ * We set bit 0 to indicate if it's an inode-based key.
+ */
+union futex_key {
+	struct {
+		unsigned long pgoff;
+		struct inode *inode;
+		int offset;
+	} shared;
+	struct {
+		unsigned long address;
+		struct mm_struct *mm;
+		int offset;
+	} private;
+	struct {
+		unsigned long word;
+		void *ptr;
+		int offset;
+	} both;
+};
+int get_futex_key(u32 __user *uaddr, union futex_key *key);
+void get_futex_key_refs(union futex_key *key);
+void drop_futex_key_refs(union futex_key *key);
+
 #ifdef CONFIG_FUTEX
 extern void exit_robust_list(struct task_struct *curr);
 extern void exit_pi_state_list(struct task_struct *curr);
===================================================================
--- a/kernel/futex.c
+++ b/kernel/futex.c
@@ -48,37 +48,12 @@
 #include <linux/pagemap.h>
 #include <linux/syscalls.h>
 #include <linux/signal.h>
+#include <linux/module.h>
 #include <asm/futex.h>
 
 #include "rtmutex_common.h"
 
 #define FUTEX_HASHBITS (CONFIG_BASE_SMALL ? 4 : 8)
-
-/*
- * Futexes are matched on equal values of this key.
- * The key type depends on whether it's a shared or private mapping.
- * Don't rearrange members without looking at hash_futex().
- *
- * offset is aligned to a multiple of sizeof(u32) (== 4) by definition.
- * We set bit 0 to indicate if it's an inode-based key.
- */
-union futex_key {
-	struct {
-		unsigned long pgoff;
-		struct inode *inode;
-		int offset;
-	} shared;
-	struct {
-		unsigned long address;
-		struct mm_struct *mm;
-		int offset;
-	} private;
-	struct {
-		unsigned long word;
-		void *ptr;
-		int offset;
-	} both;
-};
 
 /*
  * Priority Inheritance state:
@@ -175,7 +150,7 @@ static inline int match_futex(union fute
  *
  * Should be called with &current->mm->mmap_sem but NOT any spinlocks.
  */
-static int get_futex_key(u32 __user *uaddr, union futex_key *key)
+int get_futex_key(u32 __user *uaddr, union futex_key *key)
 {
 	unsigned long address = (unsigned long)uaddr;
 	struct mm_struct *mm = current->mm;
@@ -246,6 +221,7 @@ static int get_futex_key(u32 __user *uad
 	}
 	return err;
 }
+EXPORT_SYMBOL_GPL(get_futex_key);
 
 /*
  * Take a reference to the resource addressed by a key.
@@ -254,7 +230,7 @@ static int get_futex_key(u32 __user *uad
  * NOTE: mmap_sem MUST be held between get_futex_key() and calling this
  * function, if it is called at all.  mmap_sem keeps key->shared.inode valid.
  */
-static inline void get_key_refs(union futex_key *key)
+inline void get_futex_key_refs(union futex_key *key)
 {
 	if (key->both.ptr != 0) {
 		if (key->both.offset & 1)
@@ -263,12 +239,13 @@ static inline void get_key_refs(union fu
 			atomic_inc(&key->private.mm->mm_count);
 	}
 }
+EXPORT_SYMBOL_GPL(get_futex_key_refs);
 
 /*
  * Drop a reference to the resource addressed by a key.
  * The hash bucket spinlock must not be held.
  */
-static void drop_key_refs(union futex_key *key)
+void drop_futex_key_refs(union futex_key *key)
 {
 	if (key->both.ptr != 0) {
 		if (key->both.offset & 1)
@@ -277,6 +254,7 @@ static void drop_key_refs(union futex_ke
 			mmdrop(key->private.mm);
 	}
 }
+EXPORT_SYMBOL_GPL(drop_futex_key_refs);
 
 static inline int get_futex_value_locked(u32 *dest, u32 __user *from)
 {
@@ -871,7 +849,7 @@ static int futex_requeue(u32 __user *uad
 				this->lock_ptr = &hb2->lock;
 			}
 			this->key = key2;
-			get_key_refs(&key2);
+			get_futex_key_refs(&key2);
 			drop_count++;
 
 			if (ret - nr_wake >= nr_requeue)
@@ -884,9 +862,9 @@ out_unlock:
 	if (hb1 != hb2)
 		spin_unlock(&hb2->lock);
 
-	/* drop_key_refs() must be called outside the spinlocks. */
+	/* drop_futex_key_refs() must be called outside the spinlocks. */
 	while (--drop_count >= 0)
-		drop_key_refs(&key1);
+		drop_futex_key_refs(&key1);
 
 out:
 	up_read(&current->mm->mmap_sem);
@@ -904,7 +882,7 @@ queue_lock(struct futex_q *q, int fd, st
 
 	init_waitqueue_head(&q->waiters);
 
-	get_key_refs(&q->key);
+	get_futex_key_refs(&q->key);
 	hb = hash_futex(&q->key);
 	q->lock_ptr = &hb->lock;
 
@@ -923,7 +901,7 @@ queue_unlock(struct futex_q *q, struct f
 queue_unlock(struct futex_q *q, struct futex_hash_bucket *hb)
 {
 	spin_unlock(&hb->lock);
-	drop_key_refs(&q->key);
+	drop_futex_key_refs(&q->key);
 }
 
 /*
@@ -978,7 +956,7 @@ static int unqueue_me(struct futex_q *q)
 		ret = 1;
 	}
 
-	drop_key_refs(&q->key);
+	drop_futex_key_refs(&q->key);
 	return ret;
 }
 
@@ -997,7 +975,7 @@ static void unqueue_me_pi(struct futex_q
 
 	spin_unlock(&hb->lock);
 
-	drop_key_refs(&q->key);
+	drop_futex_key_refs(&q->key);
 }
 
 static int futex_wait(u32 __user *uaddr, u32 val, unsigned long time)

[PATCH 1/8] lguest: Kconfig and headers

[Rusty Russell]

Unfortunately, we don't have the build infrastructure for "private"
asm-offsets.h files, so there's a not-so-neat include in
arch/i386/kernel/asm-offsets.c.

The four headers are:
asm/lguest.h:
	Things the guest needs to know (hypercall numbers, etc).
asm/lguest_device.h:
	Things lguest devices need to know (lguest bus registration)
asm/lguest_user.h:
	Things that the lguest userspace utility needs (/dev/lguest
	and some devices)
arch/i386/lguest/lg.h:
	Internal header for the lg module (which consists of 8 files).

Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>

diff -r c2ef4d061458 arch/i386/Kconfig
--- a/arch/i386/Kconfig	Mon Feb 12 12:56:58 2007 +1100
+++ b/arch/i386/Kconfig	Mon Feb 12 12:57:00 2007 +1100
@@ -253,6 +253,27 @@ config ES7000_CLUSTERED_APIC
 	depends on SMP && X86_ES7000 && MPENTIUMIII
 
 source "arch/i386/Kconfig.cpu"
+
+config LGUEST
+	tristate "Linux hypervisor example code"
+	depends on X86 && PARAVIRT && EXPERIMENTAL && !X86_PAE
+	select LGUEST_GUEST
+	select HVC_DRIVER
+	---help---
+	  This is a very simple module which allows you to run
+	  multiple instances of the same Linux kernel, using the
+	  "lguest" command found in the Documentation/lguest directory.
+	  Note that "lguest" is pronounced to rhyme with "fell quest",
+	  not "rustyvisor".  See Documentation/lguest/lguest.txt.
+
+	  If unsure, say N.  If curious, say M.  If masochistic, say Y.
+
+config LGUEST_GUEST
+	bool
+	help
+	  The guest needs code built-in, even if the host has lguest
+	  support as a module.  The drivers are tiny, so we build them
+	  in too.
 
 config HPET_TIMER
 	bool "HPET Timer Support"
diff -r c2ef4d061458 arch/i386/kernel/asm-offsets.c
--- a/arch/i386/kernel/asm-offsets.c	Mon Feb 12 12:56:58 2007 +1100
+++ b/arch/i386/kernel/asm-offsets.c	Mon Feb 12 12:57:00 2007 +1100
@@ -16,6 +16,10 @@
 #include <asm/thread_info.h>
 #include <asm/elf.h>
 #include <asm/pda.h>
+#ifdef CONFIG_LGUEST_GUEST
+#include <asm/lguest.h>
+#include "../lguest/lg.h"
+#endif
 
 #define DEFINE(sym, val) \
         asm volatile("\n->" #sym " %0 " #val : : "i" (val))
@@ -111,4 +115,19 @@ void foo(void)
 	OFFSET(PARAVIRT_iret, paravirt_ops, iret);
 	OFFSET(PARAVIRT_read_cr0, paravirt_ops, read_cr0);
 #endif
+
+#ifdef CONFIG_LGUEST_GUEST
+	BLANK();
+	OFFSET(LGUEST_DATA_irq_enabled, lguest_data, irq_enabled);
+	OFFSET(LGUEST_STATE_host_stackptr, lguest_state, host.stackptr);
+	OFFSET(LGUEST_STATE_host_pgdir, lguest_state, host.pgdir);
+	OFFSET(LGUEST_STATE_host_gdt, lguest_state, host.gdt);
+	OFFSET(LGUEST_STATE_host_idt, lguest_state, host.idt);
+	OFFSET(LGUEST_STATE_regs, lguest_state, regs);
+	OFFSET(LGUEST_STATE_gdt, lguest_state, gdt);
+	OFFSET(LGUEST_STATE_idt, lguest_state, idt);
+	OFFSET(LGUEST_STATE_gdt_table, lguest_state, gdt_table);
+	OFFSET(LGUEST_STATE_trapnum, lguest_state, regs.trapnum);
+	OFFSET(LGUEST_STATE_errcode, lguest_state, regs.errcode);
+#endif
 }
diff -r c2ef4d061458 arch/i386/lguest/lg.h
--- /dev/null	Thu Jan 01 00:00:00 1970 +0000
+++ b/arch/i386/lguest/lg.h	Mon Feb 12 12:59:06 2007 +1100
@@ -0,0 +1,253 @@
+#ifndef _LGUEST_H
+#define _LGUEST_H
+
+#include <asm/desc.h>
+/* 64k ought to be enough for anybody! */
+#define HYPERVISOR_MAP_ORDER 16
+#define HYPERVISOR_PAGES ((1 << HYPERVISOR_MAP_ORDER)/PAGE_SIZE)
+
+#define GDT_ENTRY_LGUEST_CS	10
+#define GDT_ENTRY_LGUEST_DS	11
+#define LGUEST_CS		(GDT_ENTRY_LGUEST_CS * 8)
+#define LGUEST_DS		(GDT_ENTRY_LGUEST_DS * 8)
+
+#ifndef __ASSEMBLY__
+#include <linux/types.h>
+#include <linux/init.h>
+#include <linux/stringify.h>
+#include <linux/binfmts.h>
+#include <linux/futex.h>
+#include <asm/lguest.h>
+#include <asm/lguest_user.h>
+#include <asm/semaphore.h>
+#include "irq_vectors.h"
+
+#define GUEST_DPL 1
+
+struct lguest_regs
+{
+	/* Manually saved part. */
+	u32 cr3;
+	u32 ebx, ecx, edx;
+	u32 esi, edi, ebp;
+	u32 gs;
+	u32 eax;
+	u32 fs, ds, es;
+	u32 trapnum, errcode;
+	/* Trap pushed part */
+	u32 eip;
+	u32 cs;
+	u32 eflags;
+	u32 esp;
+	u32 ss;
+};
+
+__exit void free_pagetables(void);
+__init int init_pagetables(struct page *hype_pages);
+
+/* Full 4G segment descriptors, suitable for CS and DS. */
+#define FULL_EXEC_SEGMENT ((struct desc_struct){0x0000ffff, 0x00cf9b00}) 
+#define FULL_SEGMENT ((struct desc_struct){0x0000ffff, 0x00cf9300}) 
+
+/* Simplified version of IDT. */
+struct host_trap
+{
+	unsigned long addr;
+	int disable_interrupts;
+};
+
+struct lguest_dma_info
+{
+	struct list_head list;
+	union futex_key key;
+	unsigned long dmas;
+	u16 next_dma;
+	u16 num_dmas;
+	u16 guestid;
+	u8 interrupt; 	/* 0 when not registered */
+};
+
+struct pgdir
+{
+	u32 cr3;
+	u32 *pgdir;
+};
+
+/* The private info the thread maintains about the guest. */
+struct lguest
+{
+	struct lguest_state *state;
+	struct lguest_data __user *lguest_data;
+	struct task_struct *tsk;
+	struct mm_struct *mm; 	/* == tsk->mm, but that becomes NULL on exit */
+	u16 guestid;
+	u32 pfn_limit;
+	u32 page_offset;
+	u32 cr2;
+	int timer_on;
+	int halted;
+	int ts;
+	u32 gpf_eip;
+	u32 last_timer;
+	u32 next_hcall;
+	u16 tls_limits[GDT_ENTRY_TLS_ENTRIES];
+
+	/* We keep a small number of these. */
+	u32 pgdidx;
+	struct pgdir pgdirs[4];
+	void *trap_page;
+
+	/* Cached wakeup: we hold a reference to this task. */
+	struct task_struct *wake;
+
+	unsigned long noirq_start, noirq_end;
+	int dma_is_pending;
+	unsigned long pending_dma; /* struct lguest_dma */
+	unsigned long pending_addr; /* address they're sending to */
+
+	unsigned int stack_pages;
+
+	struct lguest_dma_info dma[LGUEST_MAX_DMA];
+
+	/* Dead? */
+	const char *dead;
+
+	/* We intercept page fault (demand shadow paging & cr2 saving)
+	   protection fault (in/out emulation, TLS handling) and
+	   device not available (TS handling). */
+	struct host_trap page_trap, gpf_trap, fpu_trap;
+
+	/* Virtual interrupts */
+	DECLARE_BITMAP(irqs_pending, LGUEST_IRQS);
+	struct host_trap interrupt[LGUEST_IRQS];
+};
+
+extern struct page *hype_pages; /* Contiguous pages. */
+extern struct lguest lguests[];
+extern struct mutex lguest_lock;
+
+/* core.c: */
+/* Entry points in hypervisor */
+const unsigned long *__lguest_default_idt_entries(void);
+struct lguest_state *__lguest_states(void);
+u32 lhread_u32(struct lguest *lg, u32 addr);
+void lhwrite_u32(struct lguest *lg, u32 val, u32 addr);
+void lhread(struct lguest *lg, void *buf, u32 addr, unsigned bytes);
+void lhwrite(struct lguest *lg, u32 addr, const void *buf, unsigned bytes);
+int lguest_address_ok(const struct lguest *lg, unsigned long addr);
+int run_guest(struct lguest *lg, char *__user user);
+int find_free_guest(void);
+
+/* interrupts_and_traps.c: */
+void maybe_do_interrupt(struct lguest *lg);
+int reflect_trap(struct lguest *lg, const struct host_trap *trap, int has_err);
+void check_bug_kill(struct lguest *lg);
+void load_guest_idt_entry(struct lguest *lg, unsigned int i, u32 low, u32 hi);
+
+/* segments.c: */
+void load_guest_gdt(struct lguest *lg, u32 table, u32 num);
+void guest_load_tls(struct lguest *lg,
+		    const struct desc_struct __user *tls_array);
+
+int init_guest_pagetable(struct lguest *lg, u32 pgtable);
+void free_guest_pagetable(struct lguest *lg);
+void guest_new_pagetable(struct lguest *lg, u32 pgtable);
+void guest_set_pud(struct lguest *lg, unsigned long cr3, u32 i);
+void guest_pagetable_clear_all(struct lguest *lg);
+void guest_pagetable_flush_user(struct lguest *lg);
+void guest_set_pte(struct lguest *lg, unsigned long cr3,
+		   unsigned long vaddr, u32 val);
+void map_trap_page(struct lguest *info);
+int demand_page(struct lguest *info, u32 cr2, int write);
+void pin_stack_pages(struct lguest *lg);
+
+int lguest_device_init(void);
+void lguest_device_remove(void);
+void lguest_io_init(void);
+u32 bind_dma(struct lguest *lg,
+	     unsigned long addr, unsigned long udma, u16 numdmas,u8 interrupt);
+int send_dma(struct lguest *info, unsigned long addr,
+	     unsigned long udma);
+void release_all_dma(struct lguest *lg);
+unsigned long get_dma_buffer(struct lguest *lg, unsigned long addr,
+			     unsigned long *interrupt);
+
+void set_wakeup_process(struct lguest *lg, struct task_struct *p);
+int do_async_hcalls(struct lguest *info);
+int hypercall(struct lguest *info, struct lguest_regs *regs);
+
+#define kill_guest(lg, fmt...)					\
+do {								\
+	if (!(lg)->dead) {					\
+		(lg)->dead = kasprintf(GFP_ATOMIC, fmt);	\
+		if (!(lg)->dead)				\
+			(lg)->dead = (void *)1;			\
+	}							\
+} while(0)
+
+static inline unsigned long guest_pa(struct lguest *lg, unsigned long vaddr)
+{
+	return vaddr - lg->page_offset;
+}
+
+/* Hardware-defined TSS structure. */
+struct x86_tss
+{
+	unsigned short	back_link,__blh;
+	unsigned long	esp0;
+	unsigned short	ss0,__ss0pad;
+	unsigned long	esp1;
+	unsigned short	ss1,__ss1pad;
+	unsigned long	esp2;
+	unsigned short	ss2,__ss2pad;
+	unsigned long	cr3;
+	unsigned long	eip;
+	unsigned long	eflags;
+	unsigned long	eax,ecx,edx,ebx;
+	unsigned long	esp; /* We actually use this one to save esp. */
+	unsigned long	ebp;
+	unsigned long	esi;
+	unsigned long	edi;
+	unsigned short	es, __espad;
+	unsigned short	cs, __cspad;
+	unsigned short	ss, __sspad;
+	unsigned short	ds, __dspad;
+	unsigned short	fs, __fspad;
+	unsigned short	gs, __gspad;
+	unsigned short	ldt, __ldtpad;
+	unsigned short	trace, io_bitmap_base;
+};
+
+int fixup_gdt_table(struct desc_struct *gdt, unsigned int num,
+		    struct lguest_regs *regs, struct x86_tss *tss);
+
+struct lguest_host_state
+{
+	struct Xgt_desc_struct	gdt;
+	struct Xgt_desc_struct	idt;
+	unsigned long		pgdir;
+	unsigned long		stackptr;
+};
+
+/* This sits in the high-mapped shim. */
+struct lguest_state
+{
+	/* Task struct. */
+	struct x86_tss tss;
+
+	/* Gate descriptor table. */
+	struct Xgt_desc_struct gdt;
+	struct desc_struct gdt_table[GDT_ENTRIES];
+
+	/* Interrupt descriptor table. */
+	struct Xgt_desc_struct idt;
+	struct desc_struct idt_table[IDT_ENTRIES];
+
+	/* Host state we store while the guest runs. */
+	struct lguest_host_state host;
+
+	/* This is the stack on which we push our regs. */
+	struct lguest_regs regs;
+};
+#endif	/* __ASSEMBLY__ */
+#endif	/* _LGUEST_H */
diff -r c2ef4d061458 include/asm-i386/lguest.h
--- /dev/null	Thu Jan 01 00:00:00 1970 +0000
+++ b/include/asm-i386/lguest.h	Mon Feb 12 12:59:29 2007 +1100
@@ -0,0 +1,84 @@
+/* Things the lguest guest needs to know. */
+#ifndef _ASM_LGUEST_H
+#define _ASM_LGUEST_H
+
+#define LGUEST_MAGIC_EBP 0x4C687970
+#define LGUEST_MAGIC_EDI 0x652D4D65
+#define LGUEST_MAGIC_ESI 0xFFFFFFFF
+
+#define LHCALL_FLUSH_ASYNC	0
+#define LHCALL_LGUEST_INIT	1
+#define LHCALL_CRASH		2
+#define LHCALL_LOAD_GDT		3
+#define LHCALL_NEW_PGTABLE	4
+#define LHCALL_FLUSH_TLB	5
+#define LHCALL_LOAD_IDT_ENTRY	6
+#define LHCALL_SET_STACK	7
+#define LHCALL_TS		8
+#define LHCALL_TIMER_READ	9
+#define LHCALL_TIMER_START	10
+#define LHCALL_HALT		11
+#define LHCALL_GET_WALLCLOCK	12
+#define LHCALL_BIND_DMA		13
+#define LHCALL_SEND_DMA		14
+#define LHCALL_SET_PTE		15
+#define LHCALL_SET_UNKNOWN_PTE	16
+#define LHCALL_SET_PUD		17
+#define LHCALL_LOAD_TLS		18
+
+#define LGUEST_TRAP_ENTRY 0x1F
+
+static inline unsigned long
+hcall(unsigned long call,
+      unsigned long arg1, unsigned long arg2, unsigned long arg3)
+{
+	asm volatile("int $" __stringify(LGUEST_TRAP_ENTRY)
+		     : "=a"(call)
+		     : "a"(call), "d"(arg1), "b"(arg2), "c"(arg3) 
+		     : "memory");
+	return call;
+}
+
+void async_hcall(unsigned long call,
+		 unsigned long arg1, unsigned long arg2, unsigned long arg3);
+
+#define LGUEST_IRQS 32
+
+#define LHCALL_RING_SIZE 64
+struct hcall_ring
+{
+	u32 eax, edx, ebx, ecx;
+};
+
+/* All the good stuff happens here: guest registers it with LGUEST_INIT */
+struct lguest_data
+{
+/* Fields which change during running: */
+	/* 512 == enabled (same as eflags) */
+	unsigned int irq_enabled;
+	/* Blocked interrupts. */
+	DECLARE_BITMAP(interrupts, LGUEST_IRQS); 
+
+	/* Last (userspace) address we got a GPF & reloaded gs. */
+	unsigned int gs_gpf_eip;
+
+	/* Virtual address of page fault. */
+	unsigned long cr2;
+
+	/* Async hypercall ring.  0xFF == done, 0 == pending. */
+	u8 hcall_status[LHCALL_RING_SIZE];
+	struct hcall_ring hcalls[LHCALL_RING_SIZE];
+			
+/* Fields initialized by the hypervisor at boot: */
+	/* Memory not to try to access */
+	unsigned long reserve_mem;
+	/* ID of this guest (used by network driver to set ethernet address) */
+	u16 guestid;
+
+/* Fields initialized by the guest at boot: */
+	/* Instruction range to suppress interrupts even if enabled */
+	unsigned long noirq_start, noirq_end;
+};
+extern struct lguest_data lguest_data;
+extern struct lguest_device_desc *lguest_devices; /* Just past max_pfn */
+#endif	/* _ASM_LGUEST_H */
diff -r c2ef4d061458 include/asm-i386/lguest_device.h
--- /dev/null	Thu Jan 01 00:00:00 1970 +0000
+++ b/include/asm-i386/lguest_device.h	Mon Feb 12 12:57:00 2007 +1100
@@ -0,0 +1,31 @@
+#ifndef _ASM_LGUEST_DEVICE_H
+#define _ASM_LGUEST_DEVICE_H
+/* Everything you need to know about lguest devices. */
+#include <linux/device.h>
+#include <asm/lguest.h>
+#include <asm/lguest_user.h>
+
+struct lguest_device {
+	/* Unique busid, and index into lguest_page->devices[] */
+	/* By convention, each device can use irq index+1 if it wants to. */
+	unsigned int index;
+
+	struct device dev;
+
+	/* Driver can hang data off here. */
+	void *private;
+};
+
+struct lguest_driver {
+	const char *name;
+	struct module *owner;
+	u16 device_type;
+	int (*probe)(struct lguest_device *dev);
+	void (*remove)(struct lguest_device *dev);
+
+	struct device_driver drv;
+};
+
+extern int register_lguest_driver(struct lguest_driver *drv);
+extern void unregister_lguest_driver(struct lguest_driver *drv);
+#endif /* _ASM_LGUEST_DEVICE_H */
diff -r c2ef4d061458 include/asm-i386/lguest_user.h
--- /dev/null	Thu Jan 01 00:00:00 1970 +0000
+++ b/include/asm-i386/lguest_user.h	Mon Feb 12 12:57:00 2007 +1100
@@ -0,0 +1,86 @@
+#ifndef _ASM_LGUEST_USER
+#define _ASM_LGUEST_USER
+/* Everything the "lguest" userspace program needs to know. */
+/* They can register up to 32 arrays of lguest_dma. */
+#define LGUEST_MAX_DMA		32
+/* At most we can dma 16 lguest_dma in one op. */
+#define LGUEST_MAX_DMA_SECTIONS	16
+
+/* How many devices?  Assume each one wants up to two dma arrays per device. */
+#define LGUEST_MAX_DEVICES (LGUEST_MAX_DMA/2)
+
+struct lguest_dma
+{
+	/* 0 if free to be used, filled by hypervisor. */
+ 	u32 used_len;
+	u32 addr[LGUEST_MAX_DMA_SECTIONS];
+	u16 len[LGUEST_MAX_DMA_SECTIONS];
+};
+
+/* This is found at address 0. */
+struct lguest_boot_info
+{
+	u32 max_pfn;
+	u32 initrd_size;
+	char cmdline[256];
+};
+
+struct lguest_block_page
+{
+	/* 0 is a read, 1 is a write. */
+	int type;
+	u32 sector; 	/* Offset in device = sector * 512. */
+	u32 bytes;	/* Length expected to be read/written in bytes */
+	/* 0 = pending, 1 = done, 2 = done, error */
+	int result;
+	u32 num_sectors; /* Disk length = num_sectors * 512 */
+};
+
+/* There is a shared page of these. */
+struct lguest_net
+{
+	union {
+		unsigned char mac[6];
+		struct {
+			u8 promisc;
+			u8 pad;
+			u16 guestid;
+		};
+	};
+};
+
+/* lguest_device_desc->type */
+#define LGUEST_DEVICE_T_CONSOLE	1
+#define LGUEST_DEVICE_T_NET	2
+#define LGUEST_DEVICE_T_BLOCK	3
+
+/* lguest_device_desc->status.  256 and above are device specific. */
+#define LGUEST_DEVICE_S_ACKNOWLEDGE	1 /* We have seen device. */
+#define LGUEST_DEVICE_S_DRIVER		2 /* We have found a driver */
+#define LGUEST_DEVICE_S_DRIVER_OK	4 /* Driver says OK! */
+#define LGUEST_DEVICE_S_REMOVED		8 /* Device has gone away. */
+#define LGUEST_DEVICE_S_REMOVED_ACK	16 /* Driver has been told. */
+#define LGUEST_DEVICE_S_FAILED		128 /* Something actually failed */
+
+#define LGUEST_NET_F_NOCSUM		0x4000 /* Don't bother checksumming */
+#define LGUEST_DEVICE_F_RANDOMNESS	0x8000 /* IRQ is fairly random */
+
+/* We have a page of these descriptors in the lguest_device page. */
+struct lguest_device_desc {
+	u16 type;
+	u16 features;
+	u16 status;
+	u16 num_pages;
+	u32 pfn;
+};
+
+/* Write command first word is a request. */
+enum lguest_req
+{
+	LHREQ_INITIALIZE, /* + pfnlimit, pgdir, start, pageoffset */
+	LHREQ_GETDMA, /* + addr (returns &lguest_dma, irq in ->used_len) */
+	LHREQ_IRQ, /* + irq */
+};
+
+
+#endif /* _ASM_LGUEST_USER */

[PATCH 2/8] lguest: the host code (lg.ko).

[Rusty Russell]

This is the host module (lg.ko) which supports lguest:

arch/i386/lguest/hypervisor.S:
	The actual guest <-> host switching code.  This is compiled into
	a C array, which is mapped to 0xFFC01000 in host and guests.

arch/i386/lguest/core.c:
	The core of the hypervisor, which calls into the assembler
	code which does this actual switch.  Also contains helper
	routines.

arch/i386/lguest/hypercalls.c:
	The entry point for the 19 hypercalls.

arch/i386/lguest/interrupts_and_traps.c:
	Handling of interrupts and traps, except page faults.

arch/i386/lguest/io.c:
	I/O from guest to host, and between guests.

arch/i386/lguest/lguest_user.c:
	/dev/lguest interface for lguest program to launch/control guests.

arch/i386/lguest/page_tables.c:
	Shadow Page table handling: generally we build up the shadow
	page tables by converting from guest page tables when a fault occurs.

arch/i386/lguest/segments.c:
	Segmentation (GDT) handling: we have to ensure they're trimmed
	to avoid guest access to the switching code.

Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
* * *

diff -r a31396449b82 arch/i386/lguest/core.c
--- /dev/null	Thu Jan 01 00:00:00 1970 +0000
+++ b/arch/i386/lguest/core.c	Mon Feb 12 12:59:41 2007 +1100
@@ -0,0 +1,432 @@
+/* World's simplest hypervisor, to test paravirt_ops and show
+ * unbelievers that virtualization is the future.  Plus, it's fun! */
+#include <linux/module.h>
+#include <linux/stringify.h>
+#include <linux/stddef.h>
+#include <linux/io.h>
+#include <linux/mm.h>
+#include <linux/vmalloc.h>
+#include <linux/cpu.h>
+#include <linux/freezer.h>
+#include <asm/lguest.h>
+#include <asm/paravirt.h>
+#include <asm/desc.h>
+#include <asm/pgtable.h>
+#include <asm/uaccess.h>
+#include <asm/poll.h>
+#include <asm/highmem.h>
+#include <asm/asm-offsets.h>
+#include <asm/i387.h>
+#include "lg.h"
+
+/* This is our hypervisor, compiled from hypervisor.S. */
+static char __initdata hypervisor_blob[] = {
+#include "hypervisor-blob.c"
+};
+
+#define MAX_LGUEST_GUESTS						  \
+	(((PAGE_SIZE << HYPERVISOR_PAGE_ORDER) - sizeof(hypervisor_blob)) \
+	 / sizeof(struct lguest_state))
+
+static struct vm_struct *hypervisor_vma;
+static int cpu_had_pge;
+static struct {
+	unsigned long offset;
+	unsigned short segment;
+} lguest_entry __attribute_used__;
+struct page *hype_pages; /* Contiguous pages. */
+struct lguest lguests[MAX_LGUEST_GUESTS];
+DEFINE_MUTEX(lguest_lock);
+
+/* IDT entries are at start of hypervisor. */
+const unsigned long *__lguest_default_idt_entries(void)
+{
+	return (void *)HYPE_ADDR;
+}
+
+/* Next is switch_to_guest */
+static void *__lguest_switch_to_guest(void)
+{
+	return (void *)HYPE_ADDR + HYPE_DATA_SIZE;
+}
+
+/* Then we use everything else to hold guest state. */
+struct lguest_state *__lguest_states(void)
+{
+	return (void *)HYPE_ADDR + sizeof(hypervisor_blob);
+}
+
+static __init int map_hypervisor(void)
+{
+	unsigned int i;
+	int err;
+	struct page *pages[HYPERVISOR_PAGES], **pagep = pages;
+
+	hype_pages = alloc_pages(GFP_KERNEL|__GFP_ZERO, HYPERVISOR_PAGE_ORDER);
+	if (!hype_pages)
+		return -ENOMEM;
+
+	hypervisor_vma = __get_vm_area(PAGE_SIZE << HYPERVISOR_PAGE_ORDER,
+				       VM_ALLOC, HYPE_ADDR, VMALLOC_END);
+	if (!hypervisor_vma) {
+		err = -ENOMEM;
+		printk("lguest: could not map hypervisor pages high\n");
+		goto free_pages;
+	}
+
+	for (i = 0; i < HYPERVISOR_PAGES; i++)
+		pages[i] = hype_pages + i;
+
+	err = map_vm_area(hypervisor_vma, PAGE_KERNEL, &pagep);
+	if (err) {
+		printk("lguest: map_vm_area failed: %i\n", err);
+		goto free_vma;
+	}
+	memcpy(hypervisor_vma->addr, hypervisor_blob, sizeof(hypervisor_blob));
+
+	/* Setup LGUEST segments on all cpus */
+	for_each_possible_cpu(i) {
+		get_cpu_gdt_table(i)[GDT_ENTRY_LGUEST_CS] = FULL_EXEC_SEGMENT;
+		get_cpu_gdt_table(i)[GDT_ENTRY_LGUEST_DS] = FULL_SEGMENT;
+	}
+
+	/* Initialize entry point into hypervisor. */
+	lguest_entry.offset = (long)__lguest_switch_to_guest();
+	lguest_entry.segment = LGUEST_CS;
+
+	printk("lguest: mapped hypervisor at %p\n", hypervisor_vma->addr);
+	return 0;
+
+free_vma:
+	vunmap(hypervisor_vma->addr);
+free_pages:
+	__free_pages(hype_pages, HYPERVISOR_PAGE_ORDER);
+	return err;
+}
+
+static __exit void unmap_hypervisor(void)
+{
+	vunmap(hypervisor_vma->addr);
+	__free_pages(hype_pages, HYPERVISOR_PAGE_ORDER);
+}
+
+/* IN/OUT insns: enough to get us past boot-time probing. */
+static int emulate_insn(struct lguest *lg)
+{
+	u8 insn;
+	unsigned int insnlen = 0, in = 0, shift = 0;
+	unsigned long physaddr = guest_pa(lg, lg->state->regs.eip);
+
+	/* This only works for addresses in linear mapping... */
+	if (lg->state->regs.eip < lg->page_offset)
+		return 0;
+	lhread(lg, &insn, physaddr, 1);
+
+	/* Operand size prefix means it's actually for ax. */
+	if (insn == 0x66) {
+		shift = 16;
+		insnlen = 1;
+		lhread(lg, &insn, physaddr + insnlen, 1);
+	}
+
+	switch (insn & 0xFE) {
+	case 0xE4: /* in     <next byte>,%al */
+		insnlen += 2;
+		in = 1;
+		break;
+	case 0xEC: /* in     (%dx),%al */
+		insnlen += 1;
+		in = 1;
+		break;
+	case 0xE6: /* out    %al,<next byte> */
+		insnlen += 2;
+		break;
+	case 0xEE: /* out    %al,(%dx) */
+		insnlen += 1;
+		break;
+	default:
+		return 0;
+	}
+
+	if (in) {
+		/* Lower bit tells is whether it's a 16 or 32 bit access */
+		if (insn & 0x1)
+			lg->state->regs.eax = 0xFFFFFFFF;
+		else
+			lg->state->regs.eax |= (0xFFFF << shift);
+	}
+	lg->state->regs.eip += insnlen;
+	return 1;
+}
+
+int find_free_guest(void)
+{
+	unsigned int i;
+	for (i = 0; i < MAX_LGUEST_GUESTS; i++)
+		if (!lguests[i].state)
+			return i;
+	return -1;
+}
+
+int lguest_address_ok(const struct lguest *lg, unsigned long addr)
+{
+	return addr / PAGE_SIZE < lg->pfn_limit;
+}
+
+/* Just like get_user, but don't let guest access lguest binary. */
+u32 lhread_u32(struct lguest *lg, u32 addr)
+{
+	u32 val = 0;
+
+	/* Don't let them access lguest_add */
+	if (!lguest_address_ok(lg, addr)
+	    || get_user(val, (u32 __user *)addr) != 0)
+		kill_guest(lg, "bad read address %u", addr);
+	return val;
+}
+
+void lhwrite_u32(struct lguest *lg, u32 addr, u32 val)
+{
+	if (!lguest_address_ok(lg, addr)
+	    || put_user(val, (u32 __user *)addr) != 0)
+		kill_guest(lg, "bad write address %u", addr);
+}
+
+void lhread(struct lguest *lg, void *b, u32 addr, unsigned bytes)
+{
+	if (addr + bytes < addr || !lguest_address_ok(lg, addr+bytes)
+	    || copy_from_user(b, (void __user *)addr, bytes) != 0) {
+		/* copy_from_user should do this, but as we rely on it... */
+		memset(b, 0, bytes);
+		kill_guest(lg, "bad read address %u len %u", addr, bytes);
+	}
+}
+
+void lhwrite(struct lguest *lg, u32 addr, const void *b, unsigned bytes)
+{
+	if (addr + bytes < addr
+	    || !lguest_address_ok(lg, addr+bytes)
+	    || copy_to_user((void __user *)addr, b, bytes) != 0)
+		kill_guest(lg, "bad write address %u len %u", addr, bytes);
+}
+
+/* Saves exporting idt_table from kernel */
+static struct desc_struct *get_idt_table(void)
+{
+	struct Xgt_desc_struct idt;
+
+	asm("sidt %0":"=m" (idt));
+	return (void *)idt.address;
+}
+
+static int usermode(struct lguest_regs *regs)
+{
+	return (regs->cs & SEGMENT_RPL_MASK) == USER_RPL;
+}
+
+/* Trap page resets this when it reloads gs. */
+static int new_gfp_eip(struct lguest *lg, struct lguest_regs *regs)
+{
+	u32 eip;
+	get_user(eip, &lg->lguest_data->gs_gpf_eip);
+	if (eip == regs->eip)
+		return 0;
+	put_user(regs->eip, &lg->lguest_data->gs_gpf_eip);
+	return 1;
+}
+
+static void set_ts(unsigned int guest_ts)
+{
+	u32 cr0;
+	if (guest_ts) {
+		asm("movl %%cr0,%0":"=r" (cr0));
+		if (!(cr0 & 8))
+			asm("movl %0,%%cr0": :"r" (cr0|8));
+	}
+}
+
+static void run_guest_once(struct lguest *lg)
+{
+	unsigned int clobber;
+
+	/* Put eflags on stack, lcall does rest. */
+	asm volatile("pushf; lcall *lguest_entry"
+		     : "=a"(clobber), "=d"(clobber)
+		     : "0"(lg->state), "1"(get_idt_table())
+		     : "memory");
+}
+
+int run_guest(struct lguest *lg, char *__user user)
+{
+	struct lguest_regs *regs = &lg->state->regs;
+
+	while (!lg->dead) {
+		unsigned int cr2 = 0; /* Damn gcc */
+
+		/* Hypercalls first: we might have been out to userspace */
+		if (do_async_hcalls(lg))
+			goto pending_dma;
+
+		if (regs->trapnum == LGUEST_TRAP_ENTRY) {
+			/* Only do hypercall once. */
+			regs->trapnum = 255;
+			if (hypercall(lg, regs))
+				goto pending_dma;
+		}
+
+		if (signal_pending(current))
+			return -EINTR;
+		maybe_do_interrupt(lg);
+
+		try_to_freeze();
+
+		if (lg->dead)
+			break;
+
+		if (lg->halted) {
+			set_current_state(TASK_INTERRUPTIBLE);
+			schedule_timeout(1);
+			continue;
+		}
+
+		/* Restore limits on TLS segments if in user mode. */
+		if (usermode(regs)) {
+			unsigned int i;
+			for (i = 0; i < ARRAY_SIZE(lg->tls_limits); i++)
+				lg->state->gdt_table[GDT_ENTRY_TLS_MIN+i].a
+					|= lg->tls_limits[i];
+		}
+
+		local_irq_disable();
+		map_trap_page(lg);
+
+		/* Host state to be restored after the guest returns. */
+		asm("sidt %0":"=m"(lg->state->host.idt));
+		lg->state->host.gdt = __get_cpu_var(cpu_gdt_descr);
+
+		/* Even if *we* don't want FPU trap, guest might... */
+		set_ts(lg->ts);
+
+		run_guest_once(lg);
+
+		/* Save cr2 now if we page-faulted. */
+		if (regs->trapnum == 14)
+			asm("movl %%cr2,%0" :"=r" (cr2));
+		else if (regs->trapnum == 7)
+			math_state_restore();
+		local_irq_enable();
+
+		switch (regs->trapnum) {
+		case 13: /* We've intercepted a GPF. */
+			if (regs->errcode == 0) {
+				if (emulate_insn(lg))
+					continue;
+
+				/* FIXME: If it's reloading %gs in a loop? */
+				if (usermode(regs) && new_gfp_eip(lg,regs))
+					continue;
+			}
+
+			if (reflect_trap(lg, &lg->gpf_trap, 1))
+				continue;
+			break;
+		case 14: /* We've intercepted a page fault. */
+			if (demand_page(lg, cr2, regs->errcode & 2))
+				continue;
+
+			/* If lguest_data is NULL, this won't hurt. */
+			put_user(cr2, &lg->lguest_data->cr2);
+			if (reflect_trap(lg, &lg->page_trap, 1))
+				continue;
+			kill_guest(lg, "unhandled page fault at %#x"
+				   " (eip=%#x, errcode=%#x)",
+				   cr2, regs->eip, regs->errcode);
+			break;
+		case 7: /* We've intercepted a Device Not Available fault. */
+			/* If they don't want to know, just absorb it. */
+			if (!lg->ts) 
+				continue;
+			if (reflect_trap(lg, &lg->fpu_trap, 0))
+				continue;
+			kill_guest(lg, "unhandled FPU fault at %#x",
+				   regs->eip);
+			break;
+		case 32 ... 255: /* Real interrupt, fall thru */
+			cond_resched();
+		case LGUEST_TRAP_ENTRY: /* Handled at top of loop */
+			continue;
+		case 6: /* Invalid opcode before they installed handler */
+			check_bug_kill(lg);
+		}
+		kill_guest(lg,"unhandled trap %i at %#x (err=%i)",
+			   regs->trapnum, regs->eip, regs->errcode);
+	}
+	return -ENOENT;
+
+pending_dma:
+	put_user(lg->pending_dma, (unsigned long *)user);
+	put_user(lg->pending_addr, (unsigned long *)user+1);
+	return sizeof(unsigned long)*2;
+}
+
+#define STRUCT_LGUEST_ELEM_SIZE(elem) sizeof(((struct lguest_state *)0)->elem)
+
+static void adjust_pge(void *on)
+{
+	if (on)
+		write_cr4(read_cr4() | X86_CR4_PGE);
+	else
+		write_cr4(read_cr4() & ~X86_CR4_PGE);
+}
+ 
+static int __init init(void)
+{
+	int err;
+
+	if (paravirt_enabled())
+		return -EPERM;
+
+	err = map_hypervisor();
+	if (err)
+		return err;
+
+	err = init_pagetables(hype_pages);
+	if (err) {
+		unmap_hypervisor();
+		return err;
+	}
+	lguest_io_init();
+
+	err = lguest_device_init();
+	if (err) {
+		free_pagetables();
+		unmap_hypervisor();
+		return err;
+	}
+	lock_cpu_hotplug();
+	if (cpu_has_pge) { /* We have a broader idea of "global". */
+		cpu_had_pge = 1;
+		on_each_cpu(adjust_pge, 0, 0, 1);
+		clear_bit(X86_FEATURE_PGE, boot_cpu_data.x86_capability);
+	}
+	unlock_cpu_hotplug();
+	return 0;
+}
+
+static void __exit fini(void)
+{
+	lguest_device_remove();
+	free_pagetables();
+	unmap_hypervisor();
+	lock_cpu_hotplug();
+	if (cpu_had_pge) {
+		set_bit(X86_FEATURE_PGE, boot_cpu_data.x86_capability);
+		on_each_cpu(adjust_pge, (void *)1, 0, 1);
+	}
+	unlock_cpu_hotplug();
+}
+
+module_init(init);
+module_exit(fini);
+MODULE_LICENSE("GPL");
+MODULE_AUTHOR("Rusty Russell <rusty@rustcorp.com.au>");
diff -r a31396449b82 arch/i386/lguest/hypercalls.c
--- /dev/null	Thu Jan 01 00:00:00 1970 +0000
+++ b/arch/i386/lguest/hypercalls.c	Mon Feb 12 13:00:04 2007 +1100
@@ -0,0 +1,186 @@
+/*  Actual hypercalls, which allow guests to actually do something.
+    Copyright (C) 2006 Rusty Russell IBM Corporation
+
+    This program is free software; you can redistribute it and/or modify
+    it under the terms of the GNU General Public License as published by
+    the Free Software Foundation; either version 2 of the License, or
+    (at your option) any later version.
+
+    This program is distributed in the hope that it will be useful,
+    but WITHOUT ANY WARRANTY; without even the implied warranty of
+    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+    GNU General Public License for more details.
+
+    You should have received a copy of the GNU General Public License
+    along with this program; if not, write to the Free Software
+    Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA  02110-1301 USA
+*/
+#include <linux/uaccess.h>
+#include <linux/syscalls.h>
+#include <linux/mm.h>
+#include <asm/lguest.h>
+#include <asm/page.h>
+#include <asm/pgtable.h>
+#include <irq_vectors.h>
+#include "lg.h"
+
+static void guest_set_stack(struct lguest *lg,
+			    u32 seg, u32 esp, unsigned int pages)
+{
+	/* You cannot have a stack segment with priv level 0. */
+	if ((seg & 0x3) != GUEST_DPL)
+		kill_guest(lg, "bad stack segment %i", seg);
+	if (pages > 2)
+		kill_guest(lg, "bad stack pages %u", pages);
+	lg->state->tss.ss1 = seg;
+	lg->state->tss.esp1 = esp;
+	lg->stack_pages = pages;
+	pin_stack_pages(lg);
+}
+
+/* Return true if DMA to host userspace now pending. */
+static int do_hcall(struct lguest *lg, struct lguest_regs *regs)
+{
+	switch (regs->eax) {
+	case LHCALL_FLUSH_ASYNC:
+		break;
+	case LHCALL_LGUEST_INIT:
+		kill_guest(lg, "already have lguest_data");
+		break;
+	case LHCALL_CRASH: {
+		char msg[128];
+		lhread(lg, msg, regs->edx, sizeof(msg));
+		msg[sizeof(msg)-1] = '\0';
+		kill_guest(lg, "CRASH: %s", msg);
+		break;
+	}
+	case LHCALL_LOAD_GDT:
+		load_guest_gdt(lg, regs->edx, regs->ebx);
+		break;
+	case LHCALL_NEW_PGTABLE:
+		guest_new_pagetable(lg, regs->edx);
+		break;
+	case LHCALL_FLUSH_TLB:
+		if (regs->edx)
+			guest_pagetable_clear_all(lg);
+		else
+			guest_pagetable_flush_user(lg);
+		break;
+	case LHCALL_LOAD_IDT_ENTRY:
+		load_guest_idt_entry(lg, regs->edx, regs->ebx, regs->ecx);
+		break;
+	case LHCALL_SET_STACK:
+		guest_set_stack(lg, regs->edx, regs->ebx, regs->ecx);
+		break;
+	case LHCALL_TS:
+		lg->ts = regs->edx;
+		break;
+	case LHCALL_TIMER_READ: {
+		u32 now = jiffies;
+		mb();
+		regs->eax = now - lg->last_timer;
+		lg->last_timer = now;
+		break;
+	}
+	case LHCALL_TIMER_START:
+		lg->timer_on = 1;
+		if (regs->edx != HZ)
+			kill_guest(lg, "Bad clock speed %i", regs->edx);
+		lg->last_timer = jiffies;
+		break;
+	case LHCALL_HALT:
+		lg->halted = 1;
+		break;
+	case LHCALL_GET_WALLCLOCK: {
+		struct timeval tv;
+		do_gettimeofday(&tv);
+		regs->eax = tv.tv_sec;
+		break;
+	}
+	case LHCALL_BIND_DMA:
+		regs->eax = bind_dma(lg, regs->edx, regs->ebx,
+				     regs->ecx >> 8, regs->ecx & 0xFF);
+		break;
+	case LHCALL_SEND_DMA:
+		return send_dma(lg, regs->edx, regs->ebx);
+	case LHCALL_SET_PTE:
+		guest_set_pte(lg, regs->edx, regs->ebx, regs->ecx);
+		break;
+	case LHCALL_SET_UNKNOWN_PTE:
+		guest_pagetable_clear_all(lg);
+		break;
+	case LHCALL_SET_PUD:
+		guest_set_pud(lg, regs->edx, regs->ebx);
+		break;
+	case LHCALL_LOAD_TLS:
+		guest_load_tls(lg, (struct desc_struct __user*)regs->edx);
+		break;
+	default:
+		kill_guest(lg, "Bad hypercall %i\n", regs->eax);
+	}
+	return 0;
+}
+
+/* We always do queued calls before actual hypercall. */
+int do_async_hcalls(struct lguest *lg)
+{
+	unsigned int i, pending;
+	u8 st[LHCALL_RING_SIZE];
+
+	if (!lg->lguest_data)
+		return 0;
+
+	copy_from_user(&st, &lg->lguest_data->hcall_status, sizeof(st));
+	for (i = 0; i < ARRAY_SIZE(st); i++) {
+		struct lguest_regs regs;
+		unsigned int n = lg->next_hcall;
+
+		if (st[n] == 0xFF)
+			break;
+
+		if (++lg->next_hcall == LHCALL_RING_SIZE)
+			lg->next_hcall = 0;
+
+		get_user(regs.eax, &lg->lguest_data->hcalls[n].eax);
+		get_user(regs.edx, &lg->lguest_data->hcalls[n].edx);
+		get_user(regs.ecx, &lg->lguest_data->hcalls[n].ecx);
+		get_user(regs.ebx, &lg->lguest_data->hcalls[n].ebx);
+		pending = do_hcall(lg, &regs);
+		put_user(0xFF, &lg->lguest_data->hcall_status[n]);
+		if (pending)
+			return 1;
+	}
+
+	set_wakeup_process(lg, NULL);
+	return 0;
+}
+
+int hypercall(struct lguest *lg, struct lguest_regs *regs)
+{
+	int pending;
+
+	if (!lg->lguest_data) {
+		if (regs->eax != LHCALL_LGUEST_INIT) {
+			kill_guest(lg, "hypercall %i before LGUEST_INIT",
+				   regs->eax);
+			return 0;
+		}
+
+		lg->lguest_data = (struct lguest_data __user *)regs->edx;
+		/* We check here so we can simply copy_to_user/from_user */
+		if (!lguest_address_ok(lg, (long)lg->lguest_data)
+		    || !lguest_address_ok(lg, (long)(lg->lguest_data+1))){
+			kill_guest(lg, "bad guest page %p", lg->lguest_data);
+			return 0;
+		}
+		get_user(lg->noirq_start, &lg->lguest_data->noirq_start);
+		get_user(lg->noirq_end, &lg->lguest_data->noirq_end);
+		/* We reserve the top pgd entry. */
+		put_user(4U*1024*1024, &lg->lguest_data->reserve_mem);
+		put_user(lg->guestid, &lg->lguest_data->guestid);
+		return 0;
+	}
+	pending = do_hcall(lg, regs);
+	set_wakeup_process(lg, NULL);
+	return pending;
+}
diff -r a31396449b82 arch/i386/lguest/hypervisor.S
--- /dev/null	Thu Jan 01 00:00:00 1970 +0000
+++ b/arch/i386/lguest/hypervisor.S	Mon Feb 12 12:59:41 2007 +1100
@@ -0,0 +1,170 @@
+/* This code sits at 0xFFFF1000 to do the low-level guest<->host switch.
+   Layout is: default_idt_entries (1k), then switch_to_guest entry point. */
+#include <linux/linkage.h>
+#include <asm/asm-offsets.h>
+#include "lg.h"
+
+#define SAVE_REGS				\
+	/* Save old guest/host state */		\
+	pushl	%es;				\
+	pushl	%ds;				\
+	pushl	%fs;				\
+	pushl	%eax;				\
+	pushl	%gs;				\
+	pushl	%ebp;				\
+	pushl	%edi;				\
+	pushl	%esi;				\
+	pushl	%edx;				\
+	pushl	%ecx;				\
+	pushl	%ebx;				\
+
+.text
+ENTRY(_start) /* ld complains unless _start is defined. */
+/* %eax contains ptr to target guest state, %edx contains host idt. */
+switch_to_guest:
+	pushl	%ss
+	SAVE_REGS
+	/* Save old stack, switch to guest's stack. */
+	movl	%esp, LGUEST_STATE_host_stackptr(%eax)
+	movl	%eax, %esp
+	/* Guest registers will be at: %esp-$LGUEST_STATE_regs */
+	addl	$LGUEST_STATE_regs, %esp
+	/* Switch to guest's GDT, IDT. */
+	lgdt	LGUEST_STATE_gdt(%eax)
+	lidt	LGUEST_STATE_idt(%eax)
+	/* Save page table top. */
+	movl	%cr3, %ebx
+	movl	%ebx, LGUEST_STATE_host_pgdir(%eax)
+	/* Set host's TSS to available (clear byte 5 bit 2). */
+	movl	(LGUEST_STATE_host_gdt+2)(%eax), %ebx
+	andb	$0xFD, (GDT_ENTRY_TSS*8 + 5)(%ebx)
+	/* Switch to guest page tables */
+	popl	%ebx
+	movl	%ebx, %cr3
+	/* Switch to guest's TSS. */
+	movl	$(GDT_ENTRY_TSS*8), %ebx
+	ltr	%bx
+	/* Restore guest regs */
+	popl	%ebx
+	popl	%ecx
+	popl	%edx
+	popl	%esi
+	popl	%edi
+	popl	%ebp
+	popl	%gs
+	/* Now we've loaded gs, neuter the TLS entries down to 1 byte/page */
+	addl	$(LGUEST_STATE_gdt_table+GDT_ENTRY_TLS_MIN*8), %eax
+	movw	$0,(%eax)
+	movw	$0,8(%eax)
+	movw	$0,16(%eax)
+	popl	%eax
+	popl	%fs
+	popl	%ds
+	popl	%es
+	/* Skip error code and trap number */
+	addl	$8, %esp
+	iret
+
+#define SWITCH_TO_HOST							\
+	SAVE_REGS;							\
+	/* Save old pgdir */						\
+	movl	%cr3, %eax;						\
+	pushl	%eax;							\
+	/* Load lguest ds segment for convenience. */			\
+	movl	$(LGUEST_DS), %eax;					\
+	movl	%eax, %ds;						\
+	/* Now figure out who we are */					\
+	movl	%esp, %eax;						\
+	subl	$LGUEST_STATE_regs, %eax;				\
+	/* Switch to host page tables (GDT, IDT and stack are in host   \
+	   mem, so need this first) */					\
+	movl	LGUEST_STATE_host_pgdir(%eax), %ebx;			\
+	movl	%ebx, %cr3;						\
+	/* Set guest's TSS to available (clear byte 5 bit 2). */	\
+	andb	$0xFD, (LGUEST_STATE_gdt_table+GDT_ENTRY_TSS*8+5)(%eax);\
+	/* Switch to host's GDT & IDT. */				\
+	lgdt	LGUEST_STATE_host_gdt(%eax);				\
+	lidt	LGUEST_STATE_host_idt(%eax);				\
+	/* Switch to host's stack. */					\
+	movl	LGUEST_STATE_host_stackptr(%eax), %esp;			\
+	/* Switch to host's TSS */					\
+	movl	$(GDT_ENTRY_TSS*8), %eax;				\
+	ltr	%ax;							\
+	/* Restore host regs */						\
+	popl	%ebx;							\
+	popl	%ecx;							\
+	popl	%edx;							\
+	popl	%esi;							\
+	popl	%edi;							\
+	popl	%ebp;							\
+	popl	%gs;							\
+	popl	%eax;							\
+	popl	%fs;							\
+	popl	%ds;							\
+	popl	%es;							\
+	popl	%ss
+	
+/* Return to run_guest_once. */
+return_to_host:
+	SWITCH_TO_HOST
+	iret
+
+deliver_to_host:
+	SWITCH_TO_HOST
+decode_idt_and_jmp:
+	/* Decode IDT and jump to hosts' irq handler.  When that does iret, it
+	 * will return to run_guest_once.  This is a feature. */
+	/* We told gcc we'd clobber edx and eax... */
+	movl	LGUEST_STATE_trapnum(%eax), %eax
+	leal	(%edx,%eax,8), %eax
+	movzwl	(%eax),%edx
+	movl	4(%eax), %eax
+	xorw	%ax, %ax
+	orl	%eax, %edx
+	jmp	*%edx
+
+deliver_to_host_with_errcode:
+	SWITCH_TO_HOST
+	pushl	LGUEST_STATE_errcode(%eax)
+	jmp decode_idt_and_jmp
+
+/* Real hardware interrupts are delivered straight to the host.  Others
+   cause us to return to run_guest_once so it can decide what to do.  Note
+   that some of these are overridden by the guest to deliver directly, and
+   never enter here (see load_guest_idt_entry). */
+.macro IRQ_STUB N TARGET
+	.data; .long 1f; .text; 1:
+ /* Make an error number for most traps, which don't have one. */
+ .if (\N <> 2) && (\N <> 8) && (\N < 10 || \N > 14) && (\N <> 17)
+	pushl	$0
+ .endif
+	pushl	$\N
+	jmp	\TARGET
+	ALIGN
+.endm
+
+.macro IRQ_STUBS FIRST LAST TARGET
+ irq=\FIRST
+ .rept \LAST-\FIRST+1
+	IRQ_STUB irq \TARGET
+  irq=irq+1
+ .endr
+.endm
+	
+/* We intercept every interrupt, because we may need to switch back to
+ * host.  Unfortunately we can't tell them apart except by entry
+ * point, so we need 256 entry points.
+ */
+irq_stubs:
+.data
+default_idt_entries:	
+.text
+	IRQ_STUBS 0 1 return_to_host		/* First two traps */
+	IRQ_STUB 2 deliver_to_host_with_errcode	/* NMI */
+	IRQ_STUBS 3 31 return_to_host		/* Rest of traps */
+	IRQ_STUBS 32 127 deliver_to_host	/* Real interrupts */
+	IRQ_STUB 128 return_to_host		/* System call (overridden) */
+	IRQ_STUBS 129 255 deliver_to_host	/* Other real interrupts */
+
+/* Everything after this is used for the lguest_state structs. */
+ALIGN
diff -r a31396449b82 arch/i386/lguest/interrupts_and_traps.c
--- /dev/null	Thu Jan 01 00:00:00 1970 +0000
+++ b/arch/i386/lguest/interrupts_and_traps.c	Mon Feb 12 12:59:41 2007 +1100
@@ -0,0 +1,230 @@
+#include <linux/uaccess.h>
+#include "lg.h"
+
+static void push_guest_stack(struct lguest *lg, u32 __user **gstack, u32 val)
+{
+	lhwrite_u32(lg, (u32)--(*gstack), val);
+}
+
+int reflect_trap(struct lguest *lg, const struct host_trap *trap, int has_err)
+{
+	u32 __user *gstack;
+	u32 eflags, ss, irq_enable;
+	struct lguest_regs *regs = &lg->state->regs;
+
+	if (!trap->addr)
+		return 0;
+
+	/* If they want a ring change, we use new stack and push old ss/esp */
+	if ((regs->ss&0x3) != GUEST_DPL) {
+		gstack = (u32 __user *)guest_pa(lg, lg->state->tss.esp1);
+		ss = lg->state->tss.ss1;
+		push_guest_stack(lg, &gstack, regs->ss);
+		push_guest_stack(lg, &gstack, regs->esp);
+	} else {
+		gstack = (u32 __user *)guest_pa(lg, regs->esp);
+		ss = regs->ss;
+	}
+
+	/* We use IF bit in eflags to indicate whether irqs were disabled
+	   (it's always 0, since irqs are enabled when guest is running). */
+	eflags = regs->eflags;
+	get_user(irq_enable, &lg->lguest_data->irq_enabled);
+	eflags |= (irq_enable & 512);
+
+	push_guest_stack(lg, &gstack, eflags);
+	push_guest_stack(lg, &gstack, regs->cs);
+	push_guest_stack(lg, &gstack, regs->eip);
+
+	if (has_err)
+		push_guest_stack(lg, &gstack, regs->errcode);
+
+	/* Change the real stack so hypervisor returns to trap handler */
+	regs->ss = ss;
+	regs->esp = (u32)gstack + lg->page_offset;
+	regs->cs = (__KERNEL_CS|GUEST_DPL);
+	regs->eip = trap->addr;
+
+	/* GS will be neutered on way back to guest. */
+	put_user(0, &lg->lguest_data->gs_gpf_eip);
+
+	/* Disable interrupts for an interrupt gate. */
+	if (trap->disable_interrupts)
+		put_user(0, &lg->lguest_data->irq_enabled);
+	return 1;
+}
+
+void maybe_do_interrupt(struct lguest *lg)
+{
+	unsigned int irq;
+	DECLARE_BITMAP(irqs, LGUEST_IRQS);
+
+	if (!lg->lguest_data)
+		return;
+
+	/* If timer has changed, set timer interrupt. */
+	if (lg->timer_on && jiffies != lg->last_timer)
+		set_bit(0, lg->irqs_pending);
+
+	/* Mask out any interrupts they have blocked. */
+	copy_from_user(&irqs, lg->lguest_data->interrupts, sizeof(irqs));
+	bitmap_andnot(irqs, lg->irqs_pending, irqs, LGUEST_IRQS);
+
+	irq = find_first_bit(irqs, LGUEST_IRQS);
+	if (irq >= LGUEST_IRQS)
+		return;
+
+	/* If they're halted, we re-enable interrupts. */
+	if (lg->halted) {
+		/* Re-enable interrupts. */
+		put_user(512, &lg->lguest_data->irq_enabled);
+		lg->halted = 0;
+	} else {
+		/* Maybe they have interrupts disabled? */
+		u32 irq_enabled;
+		get_user(irq_enabled, &lg->lguest_data->irq_enabled);
+		if (!irq_enabled)
+			return;
+	}
+
+	if (lg->interrupt[irq].addr != 0) {
+		clear_bit(irq, lg->irqs_pending);
+		reflect_trap(lg, &lg->interrupt[irq], 0);
+	}
+}
+
+void check_bug_kill(struct lguest *lg)
+{
+#ifdef CONFIG_BUG
+	u32 eip = lg->state->regs.eip - PAGE_OFFSET;
+	u16 insn;
+
+	/* This only works for addresses in linear mapping... */
+	if (lg->state->regs.eip < PAGE_OFFSET)
+		return;
+	lhread(lg, &insn, eip, sizeof(insn));
+	if (insn == 0x0b0f) {
+#ifdef CONFIG_DEBUG_BUGVERBOSE
+		u16 l;
+		u32 f;
+		char file[128];
+		lhread(lg, &l, eip+sizeof(insn), sizeof(l));
+		lhread(lg, &f, eip+sizeof(insn)+sizeof(l), sizeof(f));
+		lhread(lg, file, f - PAGE_OFFSET, sizeof(file));
+		file[sizeof(file)-1] = 0;
+		kill_guest(lg, "BUG() at %#x %s:%u", eip, file, l);
+#else
+		kill_guest(lg, "BUG() at %#x", eip);
+#endif	/* CONFIG_DEBUG_BUGVERBOSE */
+	}
+#endif	/* CONFIG_BUG */
+}
+
+static void copy_trap(struct lguest *lg,
+		      struct host_trap *trap,
+		      const struct desc_struct *desc)
+{
+	u8 type = ((desc->b >> 8) & 0xF);
+
+	/* Not present? */
+	if (!(desc->b & 0x8000)) {
+		trap->addr = 0;
+		return;
+	}
+	if (type != 0xE && type != 0xF)
+		kill_guest(lg, "bad IDT type %i", type);
+	trap->disable_interrupts = (type == 0xE);
+	trap->addr = ((desc->a & 0x0000FFFF) | (desc->b & 0xFFFF0000));
+}
+
+/* FIXME: Put this in hypervisor.S and do something clever with relocs? */
+static u8 tramp[] 
+= { 0x0f, 0xa8, 0x0f, 0xa9, /* push %gs; pop %gs */
+    0x36, 0xc7, 0x05, 0x55, 0x55, 0x55, 0x55, 0x00, 0x00, 0x00, 0x00,
+    /* movl 0, %ss:lguest_data.gs_gpf_eip */
+    0xe9, 0x55, 0x55, 0x55, 0x55 /* jmp dstaddr */
+};
+#define TRAMP_MOVL_TARGET_OFF 7
+#define TRAMP_JMP_TARGET_OFF 16
+
+static u32 setup_trampoline(struct lguest *lg, unsigned int i, u32 dstaddr)
+{
+	u32 addr, off;
+
+	off = sizeof(tramp)*i;
+	memcpy(lg->trap_page + off, tramp, sizeof(tramp));
+
+	/* 0 is to be placed in lguest_data.gs_gpf_eip. */
+	addr = (u32)&lg->lguest_data->gs_gpf_eip + lg->page_offset;
+	memcpy(lg->trap_page + off + TRAMP_MOVL_TARGET_OFF, &addr, 4);
+
+	/* Address is relative to where end of jmp will be. */
+	addr = dstaddr - ((-4*1024*1024) + off + sizeof(tramp));
+	memcpy(lg->trap_page + off + TRAMP_JMP_TARGET_OFF, &addr, 4);
+	return (-4*1024*1024) + off;
+}
+
+/* We bounce through the trap page, for two reasons: firstly, we need
+   the interrupt destination always mapped, to avoid double faults,
+   secondly we want to reload %gs to make it innocuous on entering kernel.
+ */
+static void setup_idt(struct lguest *lg,
+		      unsigned int i,
+		      const struct desc_struct *desc)
+{
+	u8 type = ((desc->b >> 8) & 0xF);
+	u32 taddr;
+
+	/* Not present? */
+	if (!(desc->b & 0x8000)) {
+		/* FIXME: When we need this, we'll know... */
+		if (lg->state->idt_table[i].a & 0x8000)
+			kill_guest(lg, "removing interrupts not supported");
+		return;
+	}
+
+	/* We could reflect and disable interrupts, but guest can do itself. */
+	if (type != 0xF)
+		kill_guest(lg, "bad direct IDT %i type %i", i, type);
+
+	taddr = setup_trampoline(lg, i, (desc->a&0xFFFF)|(desc->b&0xFFFF0000));
+
+	lg->state->idt_table[i].a = (((__KERNEL_CS|GUEST_DPL)<<16)
+					| (taddr & 0x0000FFFF));
+	lg->state->idt_table[i].b = (desc->b&0xEF00)|(taddr&0xFFFF0000);
+}
+
+void load_guest_idt_entry(struct lguest *lg, unsigned int i, u32 low, u32 high)
+{
+	struct desc_struct d = { low, high };
+
+	switch (i) {
+	/* Ignore NMI, doublefault, hypercall, spurious interrupt. */
+	case 2:
+	case 8:
+	case 15:
+	case LGUEST_TRAP_ENTRY:
+	/* FIXME: We should handle debug and int3 */
+	case 1:
+	case 3:
+		return;
+	/* We intercept page fault, general protection fault and fpu missing */
+	case 13:
+		copy_trap(lg, &lg->gpf_trap, &d);
+		return;
+	case 14:
+		copy_trap(lg, &lg->page_trap, &d);
+		return;
+	case 7:
+		copy_trap(lg, &lg->fpu_trap, &d);
+		return;
+	}
+
+	/* Other traps go straight to guest. */
+	if (i < FIRST_EXTERNAL_VECTOR || i == SYSCALL_VECTOR)
+		setup_idt(lg, i, &d);
+	/* A virtual interrupt */
+	else if (i < FIRST_EXTERNAL_VECTOR + LGUEST_IRQS)
+		copy_trap(lg, &lg->interrupt[i-FIRST_EXTERNAL_VECTOR], &d);
+}
+
diff -r a31396449b82 arch/i386/lguest/io.c
--- /dev/null	Thu Jan 01 00:00:00 1970 +0000
+++ b/arch/i386/lguest/io.c	Mon Feb 12 12:59:41 2007 +1100
@@ -0,0 +1,413 @@
+/* Simple I/O model for guests, based on shared memory.
+ * Copyright (C) 2006 Rusty Russell IBM Corporation
+ *
+ *  This program is free software; you can redistribute it and/or modify
+ *  it under the terms of the GNU General Public License as published by
+ *  the Free Software Foundation; either version 2 of the License, or
+ *  (at your option) any later version.
+ *
+ *  This program is distributed in the hope that it will be useful,
+ *  but WITHOUT ANY WARRANTY; without even the implied warranty of
+ *  MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ *  GNU General Public License for more details.
+ *
+ *  You should have received a copy of the GNU General Public License
+ *  along with this program; if not, write to the Free Software
+ *  Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA  02110-1301 USA
+ */
+#include <linux/types.h>
+#include <linux/futex.h>
+#include <linux/jhash.h>
+#include <linux/mm.h>
+#include <linux/highmem.h>
+#include <linux/uaccess.h>
+#include "lg.h"
+
+static struct list_head dma_hash[64];
+
+/* FIXME: allow multi-page lengths. */
+static int check_dma_list(struct lguest *lg, const struct lguest_dma *dma)
+{
+	unsigned int i;
+
+	for (i = 0; i < LGUEST_MAX_DMA_SECTIONS; i++) {
+		if (!dma->len[i])
+			return 1;
+		if (!lguest_address_ok(lg, dma->addr[i]))
+			goto kill;
+		if (dma->len[i] > PAGE_SIZE)
+			goto kill;
+		/* We could do over a page, but is it worth it? */
+		if ((dma->addr[i] % PAGE_SIZE) + dma->len[i] > PAGE_SIZE)
+			goto kill;
+	}
+	return 1;
+
+kill:
+	kill_guest(lg, "bad DMA entry: %u@%#x", dma->len[i], dma->addr[i]);
+	return 0;
+}
+
+static unsigned int hash(const union futex_key *key)
+{
+	return jhash2((u32*)&key->both.word,
+		      (sizeof(key->both.word)+sizeof(key->both.ptr))/4,
+		      key->both.offset)
+		% ARRAY_SIZE(dma_hash);
+}
+
+/* Must hold read lock on dmainfo owner's current->mm->mmap_sem */
+static void unlink_dma(struct lguest_dma_info *dmainfo)
+{
+	BUG_ON(!mutex_is_locked(&lguest_lock));
+	dmainfo->interrupt = 0;
+	list_del(&dmainfo->list);
+	drop_futex_key_refs(&dmainfo->key);
+}
+
+static inline int key_eq(const union futex_key *a, const union futex_key *b)
+{
+	return (a->both.word == b->both.word
+		&& a->both.ptr == b->both.ptr
+		&& a->both.offset == b->both.offset);
+}
+
+static u32 unbind_dma(struct lguest *lg,
+		      const union futex_key *key,
+		      unsigned long dmas)
+{
+	int i, ret = 0;
+
+	for (i = 0; i < LGUEST_MAX_DMA; i++) {
+		if (key_eq(key, &lg->dma[i].key) && dmas == lg->dma[i].dmas) {
+			unlink_dma(&lg->dma[i]);
+			ret = 1;
+			break;
+		}
+	}
+	return ret;
+}
+
+u32 bind_dma(struct lguest *lg,
+	     unsigned long addr, unsigned long dmas, u16 numdmas, u8 interrupt)
+{
+	unsigned int i;
+	u32 ret = 0;
+	union futex_key key;
+
+	if (interrupt >= LGUEST_IRQS)
+		return 0;
+
+	mutex_lock(&lguest_lock);
+	down_read(&current->mm->mmap_sem);
+	if (get_futex_key((u32 __user *)addr, &key) != 0) {
+		kill_guest(lg, "bad dma address %#lx", addr);
+		goto unlock;
+	}
+	get_futex_key_refs(&key);
+
+	if (interrupt == 0)
+		ret = unbind_dma(lg, &key, dmas);
+	else {
+		for (i = 0; i < LGUEST_MAX_DMA; i++) {
+			if (lg->dma[i].interrupt == 0) {
+				lg->dma[i].dmas = dmas;
+				lg->dma[i].num_dmas = numdmas;
+				lg->dma[i].next_dma = 0;
+				lg->dma[i].key = key;
+				lg->dma[i].guestid = lg->guestid;
+				lg->dma[i].interrupt = interrupt;
+				list_add(&lg->dma[i].list,
+					 &dma_hash[hash(&key)]);
+				ret = 1;
+				goto unlock;
+			}
+		}
+	}
+	drop_futex_key_refs(&key);
+unlock:
+ 	up_read(&current->mm->mmap_sem);
+	mutex_unlock(&lguest_lock);
+	return ret;
+}
+
+/* lhread from another guest */
+static int lhread_other(struct lguest *lg,
+			void *buf, u32 addr, unsigned bytes)
+{
+	if (addr + bytes < addr
+	    || !lguest_address_ok(lg, addr+bytes)
+	    || access_process_vm(lg->tsk, addr, buf, bytes, 0) != bytes) {
+		memset(buf, 0, bytes);
+		kill_guest(lg, "bad address in registered DMA struct");
+		return 0;
+	}
+	return 1;
+}
+
+/* lhwrite to another guest */
+static int lhwrite_other(struct lguest *lg, u32 addr,
+			 const void *buf, unsigned bytes)
+{
+	if (addr + bytes < addr
+	    || !lguest_address_ok(lg, addr+bytes)
+	    || (access_process_vm(lg->tsk, addr, (void *)buf, bytes, 1)
+		!= bytes)) {
+		kill_guest(lg, "bad address writing to registered DMA");
+		return 0;
+	}
+	return 1;
+}
+
+static u32 copy_data(const struct lguest_dma *src,
+		     const struct lguest_dma *dst,
+		     struct page *pages[])
+{
+	unsigned int totlen, si, di, srcoff, dstoff;
+	void *maddr = NULL;
+
+	totlen = 0;
+	si = di = 0;
+	srcoff = dstoff = 0;
+	while (si < LGUEST_MAX_DMA_SECTIONS && src->len[si]
+	       && di < LGUEST_MAX_DMA_SECTIONS && dst->len[di]) {
+		u32 len = min(src->len[si] - srcoff, dst->len[di] - dstoff);
+
+		if (!maddr)
+			maddr = kmap(pages[di]);
+
+		/* FIXME: This is not completely portable, since
+		   archs do different things for copy_to_user_page. */
+		if (copy_from_user(maddr + (dst->addr[di] + dstoff)%PAGE_SIZE,
+				   (void *__user)src->addr[si], len) != 0) {
+			totlen = 0;
+			break;
+		}
+
+		totlen += len;
+		srcoff += len;
+		dstoff += len;
+		if (srcoff == src->len[si]) {
+			si++;
+			srcoff = 0;
+		}
+		if (dstoff == dst->len[di]) {
+			kunmap(pages[di]);
+			maddr = NULL;
+			di++;
+			dstoff = 0;
+		}
+	}
+
+	if (maddr)
+		kunmap(pages[di]);
+
+	return totlen;
+}
+
+/* Src is us, ie. current. */
+static u32 do_dma(struct lguest *srclg, const struct lguest_dma *src,
+		  struct lguest *dstlg, const struct lguest_dma *dst)
+{
+	int i;
+	u32 ret;
+	struct page *pages[LGUEST_MAX_DMA_SECTIONS];
+
+	if (!check_dma_list(dstlg, dst) || !check_dma_list(srclg, src))
+		return 0;
+
+	/* First get the destination pages */
+	for (i = 0; i < LGUEST_MAX_DMA_SECTIONS; i++) {
+		if (dst->len[i] == 0)
+			break;
+		if (get_user_pages(dstlg->tsk, dstlg->mm,
+				   dst->addr[i], 1, 1, 1, pages+i, NULL)
+		    != 1) {
+			ret = 0;
+			goto drop_pages;
+		}
+	}
+
+	/* Now copy until we run out of src or dst. */
+	ret = copy_data(src, dst, pages);
+
+drop_pages:
+	while (--i >= 0)
+		put_page(pages[i]);
+	return ret;
+}
+
+/* We cache one process to wakeup: helps for batching & wakes outside locks. */
+void set_wakeup_process(struct lguest *lg, struct task_struct *p)
+{
+	if (p == lg->wake)
+		return;
+
+	if (lg->wake) {
+		wake_up_process(lg->wake);
+		put_task_struct(lg->wake);
+	}
+	lg->wake = p;
+	if (lg->wake)
+		get_task_struct(lg->wake);
+}
+
+static int dma_transfer(struct lguest *srclg,
+			unsigned long udma,
+			struct lguest_dma_info *dst)
+{
+	struct lguest_dma dst_dma, src_dma;
+	struct lguest *dstlg;
+	u32 i, dma = 0;
+
+	dstlg = &lguests[dst->guestid];
+	/* Get our dma list. */
+	lhread(srclg, &src_dma, udma, sizeof(src_dma));
+
+	/* We can't deadlock against them dmaing to us, because this
+	 * is all under the lguest_lock. */
+	down_read(&dstlg->mm->mmap_sem);
+
+	for (i = 0; i < dst->num_dmas; i++) {
+		dma = (dst->next_dma + i) % dst->num_dmas;
+		if (!lhread_other(dstlg, &dst_dma,
+				  dst->dmas + dma * sizeof(struct lguest_dma),
+				  sizeof(dst_dma))) {
+			goto fail;
+		}
+		if (!dst_dma.used_len)
+			break;
+	}
+	if (i != dst->num_dmas) {
+		unsigned long used_lenp;
+		unsigned int ret;
+
+		ret = do_dma(srclg, &src_dma, dstlg, &dst_dma);
+		/* Put used length in src. */
+		lhwrite_u32(srclg,
+			    udma+offsetof(struct lguest_dma, used_len), ret);
+		if (ret == 0 && src_dma.len[0] != 0)
+			goto fail;
+
+		/* Make sure destination sees contents before length. */
+		mb();
+		used_lenp = dst->dmas
+			+ dma * sizeof(struct lguest_dma)
+			+ offsetof(struct lguest_dma, used_len);
+		lhwrite_other(dstlg, used_lenp, &ret, sizeof(ret));
+		dst->next_dma++;
+	}
+ 	up_read(&dstlg->mm->mmap_sem);
+
+	/* Do this last so dst doesn't simply sleep on lock. */
+	set_bit(dst->interrupt, dstlg->irqs_pending);
+	set_wakeup_process(srclg, dstlg->tsk);
+	return i == dst->num_dmas;
+
+fail:
+	up_read(&dstlg->mm->mmap_sem);
+	return 0;
+}
+
+int send_dma(struct lguest *lg, unsigned long addr, unsigned long udma)
+{
+	union futex_key key;
+	int pending = 0, empty = 0;
+
+again:
+	mutex_lock(&lguest_lock);
+	down_read(&current->mm->mmap_sem);
+	if (get_futex_key((u32 __user *)addr, &key) != 0) {
+		kill_guest(lg, "bad sending DMA address");
+		goto unlock;
+	}
+	/* Shared mapping?  Look for other guests... */
+	if (key.shared.offset & 1) {
+		struct lguest_dma_info *i, *n;
+		list_for_each_entry_safe(i, n, &dma_hash[hash(&key)], list) {
+			if (i->guestid == lg->guestid)
+				continue;
+			if (!key_eq(&key, &i->key))
+				continue;
+
+			empty += dma_transfer(lg, udma, i);
+			break;
+		}
+		if (empty == 1) {
+			/* Give any recipients one chance to restock. */
+			up_read(&current->mm->mmap_sem);
+			mutex_unlock(&lguest_lock);
+			yield();
+			empty++;
+			goto again;
+		}
+		pending = 0;
+	} else {
+		/* Private mapping: tell our userspace. */
+		lg->dma_is_pending = 1;
+		lg->pending_dma = udma;
+		lg->pending_addr = addr;
+		pending = 1;
+	}
+unlock:
+	up_read(&current->mm->mmap_sem);
+	mutex_unlock(&lguest_lock);
+	return pending;
+}
+
+void release_all_dma(struct lguest *lg)
+{
+	unsigned int i;
+
+	BUG_ON(!mutex_is_locked(&lguest_lock));
+
+	down_read(&lg->mm->mmap_sem);
+	for (i = 0; i < LGUEST_MAX_DMA; i++) {
+		if (lg->dma[i].interrupt)
+			unlink_dma(&lg->dma[i]);
+	}
+	up_read(&lg->mm->mmap_sem);
+}
+
+/* Userspace wants a dma buffer from this guest. */
+unsigned long get_dma_buffer(struct lguest *lg,
+			     unsigned long addr, unsigned long *interrupt)
+{
+	unsigned long ret = 0;
+	union futex_key key;
+	struct lguest_dma_info *i;
+
+	mutex_lock(&lguest_lock);
+	down_read(&current->mm->mmap_sem);
+	if (get_futex_key((u32 __user *)addr, &key) != 0) {
+		kill_guest(lg, "bad registered DMA buffer");
+		goto unlock;
+	}
+	list_for_each_entry(i, &dma_hash[hash(&key)], list) {
+		if (key_eq(&key, &i->key) && i->guestid == lg->guestid) {
+			unsigned int j;
+			for (j = 0; j < i->num_dmas; j++) {
+				struct lguest_dma dma;
+
+				ret = i->dmas + j * sizeof(struct lguest_dma);
+				lhread(lg, &dma, ret, sizeof(dma));
+				if (dma.used_len == 0)
+					break;
+			}
+			*interrupt = i->interrupt;
+			break;
+		}
+	}
+unlock:
+	up_read(&current->mm->mmap_sem);
+	mutex_unlock(&lguest_lock);
+	return ret;
+}
+
+void lguest_io_init(void)
+{
+	unsigned int i;
+
+	for (i = 0; i < ARRAY_SIZE(dma_hash); i++)
+		INIT_LIST_HEAD(&dma_hash[i]);
+}
diff -r a31396449b82 arch/i386/lguest/lg.h
--- a/arch/i386/lguest/lg.h	Mon Feb 12 12:59:39 2007 +1100
+++ b/arch/i386/lguest/lg.h	Mon Feb 12 12:59:41 2007 +1100
@@ -3,8 +3,8 @@
 
 #include <asm/desc.h>
 /* 64k ought to be enough for anybody! */
-#define HYPERVISOR_MAP_ORDER 16
-#define HYPERVISOR_PAGES ((1 << HYPERVISOR_MAP_ORDER)/PAGE_SIZE)
+#define HYPERVISOR_PAGE_ORDER (16 - PAGE_SHIFT)
+#define HYPERVISOR_PAGES (1 << HYPERVISOR_PAGE_ORDER)
 
 #define GDT_ENTRY_LGUEST_CS	10
 #define GDT_ENTRY_LGUEST_DS	11
diff -r a31396449b82 arch/i386/lguest/lguest_user.c
--- /dev/null	Thu Jan 01 00:00:00 1970 +0000
+++ b/arch/i386/lguest/lguest_user.c	Mon Feb 12 12:59:41 2007 +1100
@@ -0,0 +1,242 @@
+/* Userspace control of the guest, via /dev/lguest. */
+#include <linux/uaccess.h>
+#include <linux/miscdevice.h>
+#include <linux/fs.h>
+#include "lg.h"
+
+static struct lguest_state *setup_guest_state(unsigned int num, void *pgdir,
+					      unsigned long start)
+{
+	struct lguest_state *guest = &__lguest_states()[num];
+	unsigned int i;
+	const long *def = __lguest_default_idt_entries();
+	struct lguest_regs *regs;
+
+	guest->gdt_table[GDT_ENTRY_KERNEL_CS] = FULL_EXEC_SEGMENT;
+	guest->gdt_table[GDT_ENTRY_KERNEL_DS] = FULL_SEGMENT;
+	guest->gdt.size = GDT_ENTRIES*8-1;
+	guest->gdt.address = (unsigned long)&guest->gdt_table;
+
+	/* Other guest's IDTs are initialized from default. */
+	guest->idt.size = 8 * IDT_ENTRIES;
+	guest->idt.address = (long)guest->idt_table;
+	for (i = 0; i < IDT_ENTRIES; i++) {
+		u32 flags = 0x8e00;
+
+		/* They can't "int" into any of them except hypercall. */
+		if (i == LGUEST_TRAP_ENTRY)
+			flags |= (GUEST_DPL << 13);
+
+		guest->idt_table[i].a = (LGUEST_CS<<16) | (def[i]&0x0000FFFF);
+		guest->idt_table[i].b = (def[i]&0xFFFF0000) | flags;
+	}
+
+	memset(&guest->tss, 0, sizeof(guest->tss));
+	guest->tss.ss0 = LGUEST_DS;
+	guest->tss.esp0 = (unsigned long)(guest+1);
+	guest->tss.io_bitmap_base = sizeof(guest->tss); /* No I/O for you! */
+
+	/* Write out stack in format lguest expects, so we can switch to it. */
+	regs = &guest->regs;
+	regs->cr3 = __pa(pgdir);
+	regs->eax = regs->ebx = regs->ecx = regs->edx = regs->esp = 0;
+	regs->edi = LGUEST_MAGIC_EDI;
+	regs->ebp = LGUEST_MAGIC_EBP;
+	regs->esi = LGUEST_MAGIC_ESI;
+	regs->gs = regs->fs = 0;
+	regs->ds = regs->es = __KERNEL_DS|GUEST_DPL;
+	regs->trapnum = regs->errcode = 0;
+	regs->eip = start;
+	regs->cs = __KERNEL_CS|GUEST_DPL;
+	regs->eflags = 0x202; 	/* Interrupts enabled. */
+	regs->ss = __KERNEL_DS|GUEST_DPL;
+
+	if (!fixup_gdt_table(guest->gdt_table, ARRAY_SIZE(guest->gdt_table),
+			     &guest->regs, &guest->tss))
+		return NULL;
+
+	return guest;
+}
+
+/* + addr */
+static long user_get_dma(struct lguest *lg, const u32 __user *input)
+{
+	unsigned long addr, udma, irq;
+
+	if (get_user(addr, input) != 0)
+		return -EFAULT;
+	udma = get_dma_buffer(lg, addr, &irq);
+	if (!udma)
+		return -ENOENT;
+
+	/* We put irq number in udma->used_len. */
+	lhwrite_u32(lg, udma + offsetof(struct lguest_dma, used_len), irq);
+	return udma;
+}
+
+/* + irq */
+static int user_send_irq(struct lguest *lg, const u32 __user *input)
+{
+	u32 irq;
+
+	if (get_user(irq, input) != 0)
+		return -EFAULT;
+	if (irq >= LGUEST_IRQS)
+		return -EINVAL;
+	set_bit(irq, lg->irqs_pending);
+	return 0;
+}
+
+static ssize_t read(struct file *file, char __user *user, size_t size,loff_t*o)
+{
+	struct lguest *lg = file->private_data;
+
+	if (!lg)
+		return -EINVAL;
+
+	if (lg->dead) {
+		size_t len;
+
+		if (lg->dead == (void *)-1)
+			return -ENOMEM;
+
+		len = min(size, strlen(lg->dead)+1);
+		if (copy_to_user(user, lg->dead, len) != 0)
+			return -EFAULT;
+		return len;
+	}
+
+	if (lg->dma_is_pending)
+		lg->dma_is_pending = 0;
+
+	return run_guest(lg, user);
+}
+
+/* Take: pfnlimit, pgdir, start, pageoffset. */
+static int initialize(struct file *file, const u32 __user *input)
+{
+	struct lguest *lg;
+	int err, i;
+	u32 args[4];
+
+	if (file->private_data)
+		return -EBUSY;
+
+	if (copy_from_user(args, input, sizeof(args)) != 0)
+		return -EFAULT;
+
+	if (args[1] <= PAGE_SIZE)
+		return -EINVAL;
+
+	mutex_lock(&lguest_lock);
+	i = find_free_guest();
+	if (i < 0) {
+		err = -ENOSPC;
+		goto unlock;
+	}
+	lg = &lguests[i];
+	lg->guestid = i;
+	lg->pfn_limit = args[0];
+	lg->page_offset = args[3];
+
+	lg->trap_page = (u32 *)get_zeroed_page(GFP_KERNEL);
+	if (!lg->trap_page) {
+		err = -ENOMEM;
+		goto release_guest;
+	}
+
+	err = init_guest_pagetable(lg, args[1]);
+	if (err)
+		goto free_trap_page;
+
+	lg->state = setup_guest_state(i, lg->pgdirs[lg->pgdidx].pgdir,args[2]);
+	if (!lg->state) {
+		err = -ENOEXEC;
+		goto release_pgtable;
+	}
+	mutex_unlock(&lguest_lock);
+
+	lg->tsk = current;
+	lg->mm = get_task_mm(current);
+	file->private_data = lg;
+	return sizeof(args);
+
+release_pgtable:
+	free_guest_pagetable(lg);
+free_trap_page:
+	free_page((long)lg->trap_page);
+release_guest:
+	memset(lg, 0, sizeof(*lg));
+unlock:
+	mutex_unlock(&lguest_lock);
+	return err;
+}
+
+static ssize_t write(struct file *file, const char __user *input,
+		     size_t size, loff_t *off)
+{
+	struct lguest *lg = file->private_data;
+	u32 req;
+
+	if (get_user(req, input) != 0)
+		return -EFAULT;
+	input += sizeof(req);
+
+	if (req != LHREQ_INITIALIZE && !lg)
+		return -EINVAL;
+	if (lg && lg->dead)
+		return -ENOENT;
+
+	switch (req) {
+	case LHREQ_INITIALIZE:
+		return initialize(file, (const u32 __user *)input);
+	case LHREQ_GETDMA:
+		return user_get_dma(lg, (const u32 __user *)input);
+	case LHREQ_IRQ:
+		return user_send_irq(lg, (const u32 __user *)input);
+	default:
+		return -EINVAL;
+	}
+}
+
+static int close(struct inode *inode, struct file *file)
+{
+	struct lguest *lg = file->private_data;
+
+	if (!lg)
+		return 0;
+
+	mutex_lock(&lguest_lock);
+	release_all_dma(lg);
+	free_page((long)lg->trap_page);
+	free_guest_pagetable(lg);
+	mmput(lg->mm);
+	if (lg->dead != (void *)1)
+		kfree(lg->dead);
+	memset(lg->state, 0, sizeof(*lg->state));
+	memset(lg, 0, sizeof(*lg));
+	mutex_unlock(&lguest_lock);
+	return 0;
+}
+
+static struct file_operations lguest_fops = {
+	.owner	 = THIS_MODULE,
+	.release = close,
+	.write	 = write,
+	.read	 = read,
+};
+static struct miscdevice lguest_dev = {
+	.minor	= MISC_DYNAMIC_MINOR,
+	.name	= "lguest",
+	.fops	= &lguest_fops,
+};
+
+int __init lguest_device_init(void)
+{
+	return misc_register(&lguest_dev);
+}
+
+void __exit lguest_device_remove(void)
+{
+	misc_deregister(&lguest_dev);
+}
diff -r a31396449b82 arch/i386/lguest/page_tables.c
--- /dev/null	Thu Jan 01 00:00:00 1970 +0000
+++ b/arch/i386/lguest/page_tables.c	Mon Feb 12 12:59:41 2007 +1100
@@ -0,0 +1,374 @@
+/* Shadow page table operations.
+ * Copyright (C) Rusty Russell IBm Corporation 2006.
+ * GPL v2 and any later version */
+#include <linux/mm.h>
+#include <linux/types.h>
+#include <linux/spinlock.h>
+#include <linux/random.h>
+#include <linux/percpu.h>
+#include <asm/tlbflush.h>
+#include "lg.h"
+
+#define PTES_PER_PAGE_SHIFT 10
+#define PTES_PER_PAGE (1 << PTES_PER_PAGE_SHIFT)
+#define HYPERVISOR_PGD_ENTRY (PTES_PER_PAGE - 1)
+
+static DEFINE_PER_CPU(u32 *, hypervisor_pte_pages) = { NULL };
+#define hypervisor_pte_page(cpu) per_cpu(hypervisor_pte_pages, cpu)
+
+static unsigned vaddr_to_pgd(unsigned long vaddr)
+{
+	return vaddr >> (PAGE_SHIFT + PTES_PER_PAGE_SHIFT);
+}
+
+/* These access the real versions. */
+static u32 *toplev(struct lguest *lg, u32 i, unsigned long vaddr)
+{
+	unsigned int index = vaddr_to_pgd(vaddr);
+
+	if (index >= HYPERVISOR_PGD_ENTRY) {
+		kill_guest(lg, "attempt to access hypervisor pages");
+		index = 0;
+	} 
+	return &lg->pgdirs[i].pgdir[index];
+}
+
+static u32 *pteof(struct lguest *lg, u32 top, unsigned long vaddr)
+{
+	u32 *page = __va(top&PAGE_MASK);
+	BUG_ON(!(top & _PAGE_PRESENT));
+	return &page[(vaddr >> PAGE_SHIFT) % PTES_PER_PAGE];
+}
+
+/* These access the guest versions. */
+static u32 gtoplev(struct lguest *lg, unsigned long vaddr)
+{
+	unsigned int index = vaddr >> (PAGE_SHIFT + PTES_PER_PAGE_SHIFT);
+	return lg->pgdirs[lg->pgdidx].cr3 + index * sizeof(u32);
+}
+
+static u32 gpteof(struct lguest *lg, u32 gtop, unsigned long vaddr)
+{
+	u32 gpage = (gtop&PAGE_MASK);
+	BUG_ON(!(gtop & _PAGE_PRESENT));
+	return gpage + ((vaddr >> PAGE_SHIFT) % PTES_PER_PAGE) * sizeof(u32);
+}
+
+static void release_pte(u32 pte)
+{
+	if (pte & _PAGE_PRESENT)
+		put_page(pfn_to_page(pte >> PAGE_SHIFT));
+}
+
+/* Do a virtual -> physical mapping on a user page. */
+static unsigned long get_pfn(unsigned long virtpfn, int write)
+{
+	struct vm_area_struct *vma;
+	struct page *page;
+	unsigned long ret = -1UL;
+
+	down_read(&current->mm->mmap_sem);
+	if (get_user_pages(current, current->mm, virtpfn << PAGE_SHIFT,
+			   1, write, 1, &page, &vma) == 1)
+		ret = page_to_pfn(page);
+	up_read(&current->mm->mmap_sem);
+	return ret;
+}
+
+static u32 check_pgtable_entry(struct lguest *lg, u32 entry)
+{
+	if ((entry & (_PAGE_PWT|_PAGE_PSE))
+	    || (entry >> PAGE_SHIFT) >= lg->pfn_limit)
+		kill_guest(lg, "bad page table entry");
+	return entry & ~_PAGE_GLOBAL;
+}
+
+static u32 get_pte(struct lguest *lg, u32 entry, int write)
+{
+	u32 pfn;
+
+	pfn = get_pfn(entry >> PAGE_SHIFT, write);
+	if (pfn == -1UL) {
+		kill_guest(lg, "failed to get page %u", entry>>PAGE_SHIFT);
+		return 0;
+	}
+	return ((pfn << PAGE_SHIFT) | (entry & (PAGE_SIZE-1)));
+}
+
+/* FIXME: We hold reference to pages, which prevents them from being
+   swapped.  It'd be nice to have a callback when Linux wants to swap out. */
+
+/* We fault pages in, which allows us to update accessed/dirty bits.
+ * Return NULL or the pte page. */
+static int page_in(struct lguest *lg, u32 vaddr, unsigned flags)
+{
+	u32 gtop, gpte;
+	u32 *top, *pte, *ptepage;
+	u32 val;
+
+	gtop = gtoplev(lg, vaddr);
+	val = lhread_u32(lg, gtop);
+	if (!(val & _PAGE_PRESENT))
+		return 0;
+
+	top = toplev(lg, lg->pgdidx, vaddr);
+	if (!(*top & _PAGE_PRESENT)) {
+		/* Get a PTE page for them. */
+		ptepage = (void *)get_zeroed_page(GFP_KERNEL);
+		/* FIXME: Steal from self in this case? */
+		if (!ptepage) {
+			kill_guest(lg, "out of memory allocating pte page");
+			return 0;
+		}
+		val = check_pgtable_entry(lg, val);
+		*top = (__pa(ptepage) | (val & (PAGE_SIZE-1)));
+	} else
+		ptepage = __va(*top & PAGE_MASK);
+
+	gpte = gpteof(lg, val, vaddr);
+	val = lhread_u32(lg, gpte);
+
+	/* No page, or write to readonly page? */
+	if (!(val&_PAGE_PRESENT) || ((flags&_PAGE_DIRTY) && !(val&_PAGE_RW)))
+		return 0;
+
+	pte = pteof(lg, *top, vaddr);
+	val = check_pgtable_entry(lg, val) | flags;
+
+	/* We're done with the old pte. */
+	release_pte(*pte);
+
+	/* We don't make it writable if this isn't a write: later
+	 * write will fault so we can set dirty bit in guest. */
+	if (val & _PAGE_DIRTY)
+		*pte = get_pte(lg, val, 1);
+	else
+		*pte = get_pte(lg, val & ~_PAGE_RW, 0);
+
+	/* Now we update dirty/accessed on guest. */
+	lhwrite_u32(lg, gpte, val);
+	return 1;
+}
+
+int demand_page(struct lguest *lg, u32 vaddr, int write)
+{
+	return page_in(lg, vaddr, (write ? _PAGE_DIRTY : 0)|_PAGE_ACCESSED);
+}
+
+void pin_stack_pages(struct lguest *lg)
+{
+	unsigned int i;
+	u32 stack = lg->state->tss.esp1;
+
+	for (i = 0; i < lg->stack_pages; i++)
+		if (!demand_page(lg, stack - i*PAGE_SIZE, 1))
+			kill_guest(lg, "bad stack page %i@%#x", i, stack);
+}
+
+static unsigned int find_pgdir(struct lguest *lg, u32 pgtable)
+{
+	unsigned int i;
+	for (i = 0; i < ARRAY_SIZE(lg->pgdirs); i++)
+		if (lg->pgdirs[i].cr3 == pgtable)
+			break;
+	return i;
+}
+
+static void release_pgd(struct lguest *lg, u32 *pgd)
+{
+	if (*pgd & _PAGE_PRESENT) {
+		unsigned int i;
+		u32 *ptepage = __va(*pgd & ~(PAGE_SIZE-1));
+		for (i = 0; i < PTES_PER_PAGE; i++)
+			release_pte(ptepage[i]);
+		free_page((long)ptepage);
+		*pgd = 0;
+	}
+}
+
+static void flush_user_mappings(struct lguest *lg, int idx)
+{
+	unsigned int i;
+	for (i = 0; i < vaddr_to_pgd(lg->page_offset); i++)
+		release_pgd(lg, lg->pgdirs[idx].pgdir + i);
+}
+
+void guest_pagetable_flush_user(struct lguest *lg)
+{
+	flush_user_mappings(lg, lg->pgdidx);
+}
+
+static unsigned int new_pgdir(struct lguest *lg, u32 cr3)
+{
+	unsigned int next;
+
+	next = (lg->pgdidx + random32()) % ARRAY_SIZE(lg->pgdirs);
+	if (!lg->pgdirs[next].pgdir) {
+		lg->pgdirs[next].pgdir = (u32 *)get_zeroed_page(GFP_KERNEL);
+		if (!lg->pgdirs[next].pgdir)
+			next = lg->pgdidx;
+	}
+	lg->pgdirs[next].cr3 = cr3;
+	/* Release all the non-kernel mappings. */
+	flush_user_mappings(lg, next);
+
+	return next;
+}
+
+void guest_new_pagetable(struct lguest *lg, u32 pgtable)
+{
+	int newpgdir;
+
+	newpgdir = find_pgdir(lg, pgtable);
+	if (newpgdir == ARRAY_SIZE(lg->pgdirs))
+		newpgdir = new_pgdir(lg, pgtable);
+	lg->pgdidx = newpgdir;
+	lg->state->regs.cr3 = __pa(lg->pgdirs[lg->pgdidx].pgdir);
+	pin_stack_pages(lg);
+}
+
+static void release_all_pagetables(struct lguest *lg)
+{
+	unsigned int i, j;
+
+	for (i = 0; i < ARRAY_SIZE(lg->pgdirs); i++)
+		if (lg->pgdirs[i].pgdir)
+			for (j = 0; j < HYPERVISOR_PGD_ENTRY; j++)
+				release_pgd(lg, lg->pgdirs[i].pgdir + j);
+}
+
+void guest_pagetable_clear_all(struct lguest *lg)
+{
+	release_all_pagetables(lg);
+	pin_stack_pages(lg);
+}
+
+static void do_set_pte(struct lguest *lg, int idx,
+		       unsigned long vaddr, u32 val)
+{
+	u32 *top = toplev(lg, idx, vaddr);
+	if (*top & _PAGE_PRESENT) {
+		u32 *pte = pteof(lg, *top, vaddr);
+		release_pte(*pte);
+		if (val & (_PAGE_DIRTY | _PAGE_ACCESSED)) {
+			val = check_pgtable_entry(lg, val);
+			*pte = get_pte(lg, val, val & _PAGE_DIRTY);
+		} else
+			*pte = 0;
+	}
+}
+
+void guest_set_pte(struct lguest *lg,
+		   unsigned long cr3, unsigned long vaddr, u32 val)
+{
+	/* Kernel mappings must be changed on all top levels. */
+	if (vaddr >= lg->page_offset) {
+		unsigned int i;
+		for (i = 0; i < ARRAY_SIZE(lg->pgdirs); i++)
+			if (lg->pgdirs[i].pgdir)
+				do_set_pte(lg, i, vaddr, val);
+	} else {
+		int pgdir = find_pgdir(lg, cr3);
+		if (pgdir != ARRAY_SIZE(lg->pgdirs))
+			do_set_pte(lg, pgdir, vaddr, val);
+	}
+}
+
+void guest_set_pud(struct lguest *lg, unsigned long cr3, u32 idx)
+{
+	int pgdir;
+
+	if (idx >= HYPERVISOR_PGD_ENTRY)
+		return;
+
+	pgdir = find_pgdir(lg, cr3);
+	if (pgdir < ARRAY_SIZE(lg->pgdirs))
+		release_pgd(lg, lg->pgdirs[pgdir].pgdir + idx);
+}
+
+int init_guest_pagetable(struct lguest *lg, u32 pgtable)
+{
+	/* We assume this in flush_user_mappings, so check now */
+	if (vaddr_to_pgd(lg->page_offset) >= HYPERVISOR_PGD_ENTRY)
+		return -EINVAL;
+	lg->pgdidx = 0;
+	lg->pgdirs[lg->pgdidx].cr3 = pgtable;
+	lg->pgdirs[lg->pgdidx].pgdir = (u32*)get_zeroed_page(GFP_KERNEL);
+	if (!lg->pgdirs[lg->pgdidx].pgdir)
+		return -ENOMEM;
+	return 0;
+}
+
+void free_guest_pagetable(struct lguest *lg)
+{
+	unsigned int i;
+
+	release_all_pagetables(lg);
+	for (i = 0; i < ARRAY_SIZE(lg->pgdirs); i++)
+		free_page((long)lg->pgdirs[i].pgdir);
+}
+
+/* Caller must be preempt-safe */
+void map_trap_page(struct lguest *lg)
+{
+	int cpu = smp_processor_id();
+	
+	hypervisor_pte_page(cpu)[0] = (__pa(lg->trap_page)|_PAGE_PRESENT);
+
+	/* Since hypervisor less that 4MB, we simply mug top pte page. */
+	lg->pgdirs[lg->pgdidx].pgdir[HYPERVISOR_PGD_ENTRY] =
+		(__pa(hypervisor_pte_page(cpu))| _PAGE_KERNEL);
+}
+
+static void free_hypervisor_pte_pages(void)
+{
+	int i;
+	
+	for_each_possible_cpu(i)
+		free_page((long)hypervisor_pte_page(i));
+}
+
+static __init int alloc_hypervisor_pte_pages(void)
+{
+	int i;
+
+	for_each_possible_cpu(i) {
+		hypervisor_pte_page(i) = (u32 *)get_zeroed_page(GFP_KERNEL);
+		if (!hypervisor_pte_page(i)) {
+			free_hypervisor_pte_pages();
+			return -ENOMEM;
+		}
+	}
+	return 0;
+}
+
+static __init void populate_hypervisor_pte_page(int cpu)
+{
+	int i;
+	u32 *pte = hypervisor_pte_page(cpu);
+
+	for (i = 0; i < HYPERVISOR_PAGES; i++) {
+		/* First entry set dynamically in map_trap_page */
+		pte[i+1] = ((page_to_pfn(&hype_pages[i]) << PAGE_SHIFT) 
+			    | _PAGE_KERNEL_EXEC);
+	}
+}
+
+__init int init_pagetables(struct page hype_pages[])
+{
+	int ret;
+	unsigned int i;
+
+	ret = alloc_hypervisor_pte_pages();
+	if (ret)
+		return ret;
+
+	for_each_possible_cpu(i)
+		populate_hypervisor_pte_page(i);
+	return 0;
+}
+
+__exit void free_pagetables(void)
+{
+	free_hypervisor_pte_pages();
+}
diff -r a31396449b82 arch/i386/lguest/segments.c
--- /dev/null	Thu Jan 01 00:00:00 1970 +0000
+++ b/arch/i386/lguest/segments.c	Mon Feb 12 12:59:41 2007 +1100
@@ -0,0 +1,171 @@
+#include "lg.h"
+
+/* Dealing with GDT entries is such a horror, I convert to sanity and back */
+struct decoded_gdt_entry
+{
+	u32 base, limit;
+	union {
+		struct {
+			unsigned type:4;
+			unsigned dtype:1;
+			unsigned dpl:2;
+			unsigned present:1;
+			unsigned unused:4;
+			unsigned avl:1;
+			unsigned mbz:1;
+			unsigned def:1;
+			unsigned page_granularity:1;
+		};
+		u16 raw_attributes;
+	};
+};
+
+static struct decoded_gdt_entry decode_gdt_entry(const struct desc_struct *en)
+{
+	struct decoded_gdt_entry de;
+	de.base = ((en->a >> 16) | ((en->b & 0xff) << 16) 
+		   | (en->b & 0xFF000000));
+	de.limit = ((en->a & 0xFFFF) | (en->b & 0xF0000));
+	de.raw_attributes = (en->b >> 8);
+	return de;
+}
+
+static struct desc_struct encode_gdt_entry(const struct decoded_gdt_entry *de)
+{
+	struct desc_struct en;
+	en.a = ((de->limit & 0xFFFF) | (de->base << 16));
+	en.b = (((de->base >> 16) & 0xFF) 
+		 | ((((u32)de->raw_attributes) & 0xF0FF) << 8)
+		 | (de->limit & 0xF0000)
+		 | (de->base & 0xFF000000));
+	return en;
+}
+
+static int check_desc(const struct decoded_gdt_entry *dec)
+{
+	return (dec->mbz == 0 && dec->dtype == 1 && (dec->type & 4) == 0);
+}
+
+static void check_segment(const struct desc_struct *gdt, u32 *segreg)
+{
+	if (*segreg > 255 || !(gdt[*segreg >> 3].b & 0x8000))
+		*segreg = 0;
+}
+
+/* Ensure our manually-loaded segment regs don't fault in switch_to_guest. */
+static void check_live_segments(const struct desc_struct *gdt,
+				struct lguest_regs *regs)
+{
+	check_segment(gdt, &regs->es);
+	check_segment(gdt, &regs->ds);
+	check_segment(gdt, &regs->fs);
+	check_segment(gdt, &regs->gs);
+}
+
+int fixup_gdt_table(struct desc_struct *gdt, unsigned int num,
+		    struct lguest_regs *regs, struct x86_tss *tss)
+{
+	unsigned int i;
+	struct decoded_gdt_entry dec;
+
+	for (i = 0; i < num; i++) {
+		unsigned long base, length;
+
+		/* We override these ones, so we don't care what they give. */
+		if (i == GDT_ENTRY_TSS
+		    || i == GDT_ENTRY_LGUEST_CS
+		    || i == GDT_ENTRY_LGUEST_DS
+		    || i == GDT_ENTRY_DOUBLEFAULT_TSS)
+			continue;
+
+		dec = decode_gdt_entry(&gdt[i]);
+		if (!dec.present)
+			continue;
+
+		if (!check_desc(&dec))
+			return 0;
+
+		base = dec.base;
+		length = dec.limit + 1;
+		if (dec.page_granularity) {
+			base *= PAGE_SIZE;
+			length *= PAGE_SIZE;
+		}
+
+		/* Unacceptable base? */
+		if (base >= HYPE_ADDR)
+			return 0;
+
+		/* Wrap around or segment overlaps hypervisor mem? */
+		if (!length
+		    || base + length < base
+		    || base + length > HYPE_ADDR) {
+			/* Trim to edge of hypervisor. */
+			length = HYPE_ADDR - base;
+			if (dec.page_granularity)
+				dec.limit = (length / PAGE_SIZE) - 1;
+			else
+				dec.limit = length - 1;
+		}
+		if (dec.dpl == 0)
+			dec.dpl = GUEST_DPL;
+		gdt[i] = encode_gdt_entry(&dec);
+	}
+	check_live_segments(gdt, regs);
+
+	/* Now put in hypervisor data and code segments. */
+	gdt[GDT_ENTRY_LGUEST_CS] = FULL_EXEC_SEGMENT;
+	gdt[GDT_ENTRY_LGUEST_DS] = FULL_SEGMENT;
+
+	/* Finally, TSS entry */
+	dec.base = (unsigned long)tss;
+	dec.limit = sizeof(*tss)-1;
+	dec.type = 0x9;
+	dec.dtype = 0;
+	dec.def = 0;
+	dec.present = 1;
+	dec.mbz = 0;
+	dec.page_granularity = 0;
+	gdt[GDT_ENTRY_TSS] = encode_gdt_entry(&dec);
+
+	return 1;
+}
+
+void load_guest_gdt(struct lguest *lg, u32 table, u32 num)
+{
+	if (num > GDT_ENTRIES)
+		kill_guest(lg, "too many gdt entries %i", num);
+
+	lhread(lg, lg->state->gdt_table, table,
+	       num * sizeof(lg->state->gdt_table[0]));
+	if (!fixup_gdt_table(lg->state->gdt_table, num, 
+			     &lg->state->regs, &lg->state->tss))
+		kill_guest(lg, "bad gdt table");
+}
+
+/* We don't care about limit here, since we only let them use these in
+ * usermode (where lack of USER bit in pagetable protects hypervisor mem).
+ * However, we want to ensure it doesn't fault when loaded, since *we* are
+ * the ones who will load it in switch_to_guest.
+ */
+void guest_load_tls(struct lguest *lg, const struct desc_struct __user *gtls)
+{
+	unsigned int i;
+	struct desc_struct *tls = &lg->state->gdt_table[GDT_ENTRY_TLS_MIN];
+
+	lhread(lg, tls, (u32)gtls, sizeof(*tls)*GDT_ENTRY_TLS_ENTRIES);
+	for (i = 0; i < ARRAY_SIZE(lg->tls_limits); i++) {
+		struct decoded_gdt_entry dec = decode_gdt_entry(&tls[i]);
+
+		if (!dec.present)
+			continue;
+
+		/* We truncate to one byte/page (depending on G bit) to neuter
+		   it, so ensure it's more than 1 page below trap page. */
+		tls[i].a &= 0xFFFF0000;
+		lg->tls_limits[i] = dec.limit;
+		if (!check_desc(&dec) || dec.base > HYPE_ADDR - PAGE_SIZE)
+			kill_guest(lg, "bad TLS descriptor %i", i);
+	}
+	check_live_segments(lg->state->gdt_table, &lg->state->regs);
+}

[PATCH 3/8] lguest: Guest code

[Rusty Russell]

This is the guest code which replaces the parts of paravirt_ops with
hypercalls.  It's fairly trivial.  This patch also includes trivial
bus driver for lguest devices, and an extern declarations for boot_pda
(previously frobbed only from head.S).

Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>

diff -r 3c4f57d11d07 arch/i386/lguest/lguest.c
--- /dev/null	Thu Jan 01 00:00:00 1970 +0000
+++ b/arch/i386/lguest/lguest.c	Mon Feb 12 14:21:57 2007 +1100
@@ -0,0 +1,562 @@
+/*
+ * Lguest specific paravirt-ops implementation
+ *
+ * Copyright (C) 2006, Rusty Russell <rusty@rustcorp.com.au> IBM Corporation.
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful, but
+ * WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE, GOOD TITLE or
+ * NON INFRINGEMENT.  See the GNU General Public License for more
+ * details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
+ */
+#include <linux/kernel.h>
+#include <linux/start_kernel.h>
+#include <linux/string.h>
+#include <linux/console.h>
+#include <linux/screen_info.h>
+#include <linux/irq.h>
+#include <linux/interrupt.h>
+#include <asm/paravirt.h>
+#include <asm/lguest.h>
+#include <asm/lguest_user.h>
+#include <asm/param.h>
+#include <asm/page.h>
+#include <asm/pgtable.h>
+#include <asm/desc.h>
+#include <asm/setup.h>
+#include <asm/e820.h>
+#include <asm/pda.h>
+#include <asm/asm-offsets.h>
+#include <asm/mce.h>
+
+struct lguest_data lguest_data;
+struct lguest_device_desc *lguest_devices;
+static __initdata const struct lguest_boot_info *boot = __va(0);
+
+void async_hcall(unsigned long call,
+		 unsigned long arg1, unsigned long arg2, unsigned long arg3)
+{
+	/* Note: This code assumes we're uniprocessor. */
+	static unsigned int next_call;
+	unsigned long flags;
+
+	local_irq_save(flags);
+	if (lguest_data.hcall_status[next_call] != 0xFF) {
+		/* Table full, so do normal hcall which will flush table. */
+		hcall(call, arg1, arg2, arg3);
+	} else {
+		lguest_data.hcalls[next_call].eax = call;
+		lguest_data.hcalls[next_call].edx = arg1;
+		lguest_data.hcalls[next_call].ebx = arg2;
+		lguest_data.hcalls[next_call].ecx = arg3;
+		wmb();
+		lguest_data.hcall_status[next_call] = 0;
+		if (++next_call == LHCALL_RING_SIZE)
+			next_call = 0;
+	}
+	local_irq_restore(flags);
+}
+
+#ifdef PARAVIRT_LAZY_NONE 	/* Not in 2.6.20. */
+static int lazy_mode;
+static void fastcall lguest_lazy_mode(int mode)
+{
+	lazy_mode = mode;
+	if (mode == PARAVIRT_LAZY_NONE)
+		hcall(LHCALL_FLUSH_ASYNC, 0, 0, 0);
+}
+
+static void lazy_hcall(unsigned long call,
+		       unsigned long arg1,
+		       unsigned long arg2,
+		       unsigned long arg3)
+{
+	if (lazy_mode == PARAVIRT_LAZY_NONE)
+		hcall(call, arg1, arg2, arg3);
+	else
+		async_hcall(call, arg1, arg2, arg3);
+}
+#else
+#define lazy_hcall hcall
+#endif
+
+static unsigned long fastcall save_fl(void)
+{
+	return lguest_data.irq_enabled;
+}
+
+static void fastcall restore_fl(unsigned long flags)
+{
+	/* FIXME: Check if interrupt pending... */
+	lguest_data.irq_enabled = flags;
+}
+
+static void fastcall irq_disable(void)
+{
+	lguest_data.irq_enabled = 0;
+}
+
+static void fastcall irq_enable(void)
+{
+	/* Linux i386 code expects bit 9 set. */
+	/* FIXME: Check if interrupt pending... */
+	lguest_data.irq_enabled = 512;
+}
+
+static void fastcall lguest_load_gdt(const struct Xgt_desc_struct *desc)
+{
+	BUG_ON((desc->size+1)/8 != GDT_ENTRIES);
+	hcall(LHCALL_LOAD_GDT, __pa(desc->address), GDT_ENTRIES, 0);
+}
+
+static void fastcall lguest_load_idt(const struct Xgt_desc_struct *desc)
+{
+	unsigned int i;
+	struct desc_struct *idt = (void *)desc->address;
+
+	for (i = 0; i < (desc->size+1)/8; i++)
+		hcall(LHCALL_LOAD_IDT_ENTRY, i, idt[i].a, idt[i].b);
+}
+
+static int lguest_panic(struct notifier_block *nb, unsigned long l, void *p)
+{
+	hcall(LHCALL_CRASH, __pa(p), 0, 0);
+	return NOTIFY_DONE;
+}
+
+static struct notifier_block paniced = {
+	.notifier_call = lguest_panic
+};
+
+static char *lguest_memory_setup(void)
+{
+	/* We do this here because lockcheck barfs if before start_kernel */
+	atomic_notifier_chain_register(&panic_notifier_list, &paniced);
+
+	e820.nr_map = 0;
+	add_memory_region(0, PFN_PHYS(boot->max_pfn), E820_RAM);
+	return "LGUEST";
+}
+
+static fastcall void lguest_cpuid(unsigned int *eax, unsigned int *ebx,
+				 unsigned int *ecx, unsigned int *edx)
+{
+	int is_feature = (*eax == 1);
+
+	native_cpuid(eax, ebx, ecx, edx);
+	if (is_feature) {
+		unsigned long *excap = (unsigned long *)ecx,
+			*features = (unsigned long *)edx;
+		/* Hypervisor needs to know when we flush kernel pages. */
+		set_bit(X86_FEATURE_PGE, features);
+		/* We don't have any features! */
+		clear_bit(X86_FEATURE_VME, features);
+		clear_bit(X86_FEATURE_DE, features);
+		clear_bit(X86_FEATURE_PSE, features);
+		clear_bit(X86_FEATURE_PAE, features);
+		clear_bit(X86_FEATURE_SEP, features);
+		clear_bit(X86_FEATURE_APIC, features);
+		clear_bit(X86_FEATURE_MTRR, features);
+		/* No MWAIT, either */
+		clear_bit(3, excap);
+	}
+}
+
+static unsigned long current_cr3;
+static void fastcall lguest_write_cr3(unsigned long cr3)
+{
+	hcall(LHCALL_NEW_PGTABLE, cr3, 0, 0);
+	current_cr3 = cr3;
+}
+
+static void fastcall lguest_flush_tlb(void)
+{
+	lazy_hcall(LHCALL_FLUSH_TLB, 0, 0, 0);
+}
+
+static void fastcall lguest_flush_tlb_kernel(void)
+{
+	lazy_hcall(LHCALL_FLUSH_TLB, 1, 0, 0);
+}
+
+static void fastcall lguest_flush_tlb_single(u32 addr)
+{
+	/* Simply set it to zero, and it will fault back in. */
+	lazy_hcall(LHCALL_SET_PTE, current_cr3, addr, 0);
+}
+
+/* FIXME: Eliminate all callers of this. */
+static fastcall void lguest_set_pte(pte_t *ptep, pte_t pteval)
+{
+	*ptep = pteval;
+	/* Don't bother with hypercall before initial setup. */
+	if (current_cr3)
+		hcall(LHCALL_SET_UNKNOWN_PTE, 0, 0, 0);
+}
+
+static fastcall void lguest_set_pte_at(struct mm_struct *mm, u32 addr, pte_t *ptep, pte_t pteval)
+{
+	*ptep = pteval;
+	lazy_hcall(LHCALL_SET_PTE, __pa(mm->pgd), addr, pteval.pte_low);
+}
+
+/* We only support two-level pagetables at the moment. */
+static fastcall void lguest_set_pud(pmd_t *pmdp, pmd_t pmdval)
+{
+	*pmdp = pmdval;
+	lazy_hcall(LHCALL_SET_PUD, __pa(pmdp)&PAGE_MASK,
+		   (__pa(pmdp)&(PAGE_SIZE-1))/4, 0);
+}
+
+#ifdef CONFIG_X86_LOCAL_APIC
+static fastcall void lguest_apic_write(unsigned long reg, unsigned long v)
+{
+}
+
+static fastcall void lguest_apic_write_atomic(unsigned long reg, unsigned long v)
+{
+}
+
+static fastcall unsigned long lguest_apic_read(unsigned long reg)
+{
+	return 0;
+}
+#endif
+
+/* We move eflags word to lguest_data.irq_enabled to restore interrupt
+   state.  For page faults, gpfs and virtual interrupts, the
+   hypervisor has saved eflags manually, otherwise it was delivered
+   directly and so eflags reflects the real machine IF state,
+   ie. interrupts on.  Since the kernel always dies if it takes such a
+   trap with interrupts disabled anyway, turning interrupts back on
+   unconditionally here is OK. */
+asm("lguest_iret:"
+    " pushl	%eax;"
+    " movl	12(%esp), %eax;"
+    "lguest_noirq_start:;"
+    " movl	%eax,%ss:lguest_data+"__stringify(LGUEST_DATA_irq_enabled)";"
+    " popl	%eax;"
+    " iret;"
+    "lguest_noirq_end:");
+extern void fastcall lguest_iret(void);
+extern char lguest_noirq_start[], lguest_noirq_end[];
+
+static void fastcall lguest_load_esp0(struct tss_struct *tss,
+				     struct thread_struct *thread)
+{
+	lazy_hcall(LHCALL_SET_STACK, __KERNEL_DS|0x1, thread->esp0,
+		   THREAD_SIZE/PAGE_SIZE);
+}
+
+static fastcall void lguest_load_tr_desc(void)
+{
+}
+
+static fastcall void lguest_set_ldt(const void *addr, unsigned entries)
+{
+	/* FIXME: Implement. */
+	BUG_ON(entries);
+}
+
+static fastcall void lguest_load_tls(struct thread_struct *t, unsigned int cpu)
+{
+	lazy_hcall(LHCALL_LOAD_TLS, __pa(&t->tls_array), cpu, 0);
+}
+
+static fastcall void lguest_set_debugreg(int regno, unsigned long value)
+{
+	/* FIXME: Implement */
+}
+
+static unsigned int lguest_cr0;
+static fastcall void lguest_clts(void)
+{
+	lazy_hcall(LHCALL_TS, 0, 0, 0);
+	lguest_cr0 &= ~8U;
+}
+
+static fastcall unsigned long lguest_read_cr0(void)
+{
+	return lguest_cr0;
+}
+
+static fastcall void lguest_write_cr0(unsigned long val)
+{
+	hcall(LHCALL_TS, val & 8, 0, 0);
+	lguest_cr0 = val;
+}
+
+static fastcall unsigned long lguest_read_cr2(void)
+{
+	return lguest_data.cr2;
+}
+
+static fastcall unsigned long lguest_read_cr3(void)
+{
+	return current_cr3;
+}
+
+/* Used to enable/disable PGE, but we don't care. */
+static fastcall unsigned long lguest_read_cr4(void)
+{
+	return 0;
+}
+
+static fastcall void lguest_write_cr4(unsigned long val)
+{
+}
+
+static void fastcall lguest_time_irq(unsigned int irq, struct irq_desc *desc)
+{
+	do_timer(hcall(LHCALL_TIMER_READ, 0, 0, 0));
+	update_process_times(user_mode_vm(get_irq_regs()));
+}
+
+static void disable_lguest_irq(unsigned int irq)
+{
+	set_bit(irq, lguest_data.interrupts);
+}
+
+static void enable_lguest_irq(unsigned int irq)
+{
+	clear_bit(irq, lguest_data.interrupts);
+	/* FIXME: If it's pending? */
+}
+
+static struct irq_chip lguest_irq_controller = {
+	.name		= "lguest",
+	.mask		= disable_lguest_irq,
+	.mask_ack	= disable_lguest_irq,
+	.unmask		= enable_lguest_irq,
+};
+
+static void lguest_time_init(void)
+{
+	set_irq_handler(0, lguest_time_irq);
+	hcall(LHCALL_TIMER_START,HZ,0,0);
+}
+
+static void __init lguest_init_IRQ(void)
+{
+	unsigned int i;
+
+	for (i = 0; i < LGUEST_IRQS; i++) {
+		int vector = FIRST_EXTERNAL_VECTOR + i;
+		if (i >= NR_IRQS)
+			break;
+		if (vector != SYSCALL_VECTOR) {
+			set_intr_gate(vector, interrupt[i]);
+			set_irq_chip_and_handler(i, &lguest_irq_controller,
+						 handle_level_irq);
+		}
+	}
+	irq_ctx_init(smp_processor_id());
+}
+
+static inline void native_write_dt_entry(void *dt, int entry, u32 entry_low, u32 entry_high)
+{
+	u32 *lp = (u32 *)((char *)dt + entry*8);
+	lp[0] = entry_low;
+	lp[1] = entry_high;
+}
+
+static fastcall void lguest_write_ldt_entry(void *dt, int entrynum, u32 low, u32 high)
+{
+	/* FIXME: Allow this. */
+	BUG();
+}
+
+static fastcall void lguest_write_gdt_entry(void *dt, int entrynum,
+					   u32 low, u32 high)
+{
+	native_write_dt_entry(dt, entrynum, low, high);
+	hcall(LHCALL_LOAD_GDT, __pa(dt), GDT_ENTRIES, 0);
+}
+
+static fastcall void lguest_write_idt_entry(void *dt, int entrynum,
+					   u32 low, u32 high)
+{
+	native_write_dt_entry(dt, entrynum, low, high);
+	hcall(LHCALL_LOAD_IDT_ENTRY, entrynum, low, high);
+}
+
+#define LGUEST_IRQ "lguest_data+"__stringify(LGUEST_DATA_irq_enabled)
+#define DEF_LGUEST(name, code)				\
+	extern const char start_##name[], end_##name[];		\
+	asm("start_" #name ": " code "; end_" #name ":")
+DEF_LGUEST(cli, "movl $0," LGUEST_IRQ);
+DEF_LGUEST(sti, "movl $512," LGUEST_IRQ);
+DEF_LGUEST(popf, "movl %eax," LGUEST_IRQ);
+DEF_LGUEST(pushf, "movl " LGUEST_IRQ ",%eax");
+DEF_LGUEST(pushf_cli, "movl " LGUEST_IRQ ",%eax; movl $0," LGUEST_IRQ);
+DEF_LGUEST(iret, ".byte 0xE9,0,0,0,0"); /* jmp ... */
+
+static const struct lguest_insns
+{
+	const char *start, *end;
+} lguest_insns[] = {
+	[PARAVIRT_IRQ_DISABLE] = { start_cli, end_cli },
+	[PARAVIRT_IRQ_ENABLE] = { start_sti, end_sti },
+	[PARAVIRT_RESTORE_FLAGS] = { start_popf, end_popf },
+	[PARAVIRT_SAVE_FLAGS] = { start_pushf, end_pushf },
+	[PARAVIRT_SAVE_FLAGS_IRQ_DISABLE] = { start_pushf_cli, end_pushf_cli },
+	[PARAVIRT_INTERRUPT_RETURN] = { start_iret, end_iret },
+};
+static unsigned lguest_patch(u8 type, u16 clobber, void *insns, unsigned len)
+{
+	unsigned int insn_len;
+
+	/* Don't touch it if we don't have a replacement */
+	if (type >= ARRAY_SIZE(lguest_insns) || !lguest_insns[type].start)
+		return len;
+
+	insn_len = lguest_insns[type].end - lguest_insns[type].start;
+
+	/* Similarly if we can't fit replacement. */
+	if (len < insn_len)
+		return len;
+
+	memcpy(insns, lguest_insns[type].start, insn_len);
+	if (type == PARAVIRT_INTERRUPT_RETURN) {
+		/* Jumps are relative. */
+		u32 off = (u32)lguest_iret - ((u32)insns + insn_len);
+		memcpy(insns+1, &off, sizeof(off));
+	}
+	return insn_len;
+}
+
+static void fastcall lguest_safe_halt(void)
+{
+	hcall(LHCALL_HALT, 0, 0, 0);
+}
+
+static unsigned long lguest_get_wallclock(void)
+{
+	return hcall(LHCALL_GET_WALLCLOCK, 0, 0, 0);
+}
+
+static void lguest_power_off(void)
+{
+	hcall(LHCALL_CRASH, __pa("Power down"), 0, 0);
+}
+
+static __attribute_used__ __init void lguest_init(void)
+{
+	paravirt_ops.name = "lguest";
+	paravirt_ops.paravirt_enabled = 1;
+	paravirt_ops.kernel_rpl = 1;
+
+	paravirt_ops.save_fl = save_fl;
+	paravirt_ops.restore_fl = restore_fl;
+	paravirt_ops.irq_disable = irq_disable;
+	paravirt_ops.irq_enable = irq_enable;
+	paravirt_ops.load_gdt = lguest_load_gdt;
+	paravirt_ops.memory_setup = lguest_memory_setup;
+	paravirt_ops.cpuid = lguest_cpuid;
+	paravirt_ops.write_cr3 = lguest_write_cr3;
+	paravirt_ops.flush_tlb_user = lguest_flush_tlb;
+	paravirt_ops.flush_tlb_single = lguest_flush_tlb_single;
+	paravirt_ops.flush_tlb_kernel = lguest_flush_tlb_kernel;
+	paravirt_ops.set_pte = lguest_set_pte;
+	paravirt_ops.set_pte_at = lguest_set_pte_at;
+	paravirt_ops.set_pmd = lguest_set_pud;
+#ifdef CONFIG_X86_LOCAL_APIC
+	paravirt_ops.apic_write = lguest_apic_write;
+	paravirt_ops.apic_write_atomic = lguest_apic_write_atomic;
+	paravirt_ops.apic_read = lguest_apic_read;
+#endif
+	paravirt_ops.load_idt = lguest_load_idt;
+	paravirt_ops.iret = lguest_iret;
+	paravirt_ops.load_esp0 = lguest_load_esp0;
+	paravirt_ops.load_tr_desc = lguest_load_tr_desc;
+	paravirt_ops.set_ldt = lguest_set_ldt;
+	paravirt_ops.load_tls = lguest_load_tls;
+	paravirt_ops.set_debugreg = lguest_set_debugreg;
+	paravirt_ops.clts = lguest_clts;
+	paravirt_ops.read_cr0 = lguest_read_cr0;
+	paravirt_ops.write_cr0 = lguest_write_cr0;
+	paravirt_ops.init_IRQ = lguest_init_IRQ;
+	paravirt_ops.read_cr2 = lguest_read_cr2;
+	paravirt_ops.read_cr3 = lguest_read_cr3;
+	paravirt_ops.read_cr4 = lguest_read_cr4;
+	paravirt_ops.write_cr4 = lguest_write_cr4;
+	paravirt_ops.write_ldt_entry = lguest_write_ldt_entry;
+	paravirt_ops.write_gdt_entry = lguest_write_gdt_entry;
+	paravirt_ops.write_idt_entry = lguest_write_idt_entry;
+	paravirt_ops.patch = lguest_patch;
+	paravirt_ops.safe_halt = lguest_safe_halt;
+	paravirt_ops.get_wallclock = lguest_get_wallclock;
+	paravirt_ops.time_init = lguest_time_init;
+#ifdef PARAVIRT_LAZY_NONE
+	paravirt_ops.set_lazy_mode = lguest_lazy_mode;
+#endif
+
+	memset(lguest_data.hcall_status,0xFF,sizeof(lguest_data.hcall_status));
+	lguest_data.noirq_start = (u32)lguest_noirq_start;
+	lguest_data.noirq_end = (u32)lguest_noirq_end;
+	hcall(LHCALL_LGUEST_INIT, __pa(&lguest_data), 0, 0);
+	strncpy(boot_command_line, boot->cmdline, COMMAND_LINE_SIZE);
+
+	/* We use top of mem for initial pagetables. */
+	init_pg_tables_end = __pa(pg0);
+
+	/* set up PDA descriptor */
+	pack_descriptor((u32 *)&cpu_gdt_table[GDT_ENTRY_PDA].a,
+			(u32 *)&cpu_gdt_table[GDT_ENTRY_PDA].b,
+			(unsigned)&boot_pda, sizeof(boot_pda)-1,
+			0x80 | DESCTYPE_S | 0x02, 0);
+	load_gdt(&early_gdt_descr);
+	asm volatile ("mov %0, %%fs" : : "r" (__KERNEL_PDA) : "memory");
+
+	reserve_top_address(lguest_data.reserve_mem);
+
+	cpu_detect(&new_cpu_data);
+	/* Need this before paging_init. */
+	set_bit(X86_FEATURE_PGE, new_cpu_data.x86_capability);
+	/* Math is always hard! */
+	new_cpu_data.hard_math = 1;
+
+	/* FIXME: Better way? */
+	/* Suppress vgacon startup code */
+	SCREEN_INFO.orig_video_isVGA = VIDEO_TYPE_VLFB;
+
+	add_preferred_console("hvc", 0, NULL);
+
+#ifdef CONFIG_X86_MCE
+	mce_disabled = 1;
+#endif
+
+#ifdef CONFIG_ACPI
+	acpi_disabled = 1;
+	acpi_ht = 0;
+#endif
+	if (boot->initrd_size) {
+		/* We stash this at top of memory. */
+		INITRD_START = boot->max_pfn*PAGE_SIZE - boot->initrd_size;
+		INITRD_SIZE = boot->initrd_size;
+		LOADER_TYPE = 0xFF;
+	}
+
+	pm_power_off = lguest_power_off;
+	start_kernel();
+}
+
+asm("lguest_maybe_init:\n"
+    "	cmpl $"__stringify(LGUEST_MAGIC_EBP)", %ebp\n"
+    "	jne 1f\n"
+    "	cmpl $"__stringify(LGUEST_MAGIC_EDI)", %edi\n"
+    "	jne 1f\n"
+    "	cmpl $"__stringify(LGUEST_MAGIC_ESI)", %esi\n"
+    "	je lguest_init\n"
+    "1: ret");
+extern void asmlinkage lguest_maybe_init(void);
+paravirt_probe(lguest_maybe_init);
diff -r 3c4f57d11d07 arch/i386/lguest/lguest_bus.c
--- /dev/null	Thu Jan 01 00:00:00 1970 +0000
+++ b/arch/i386/lguest/lguest_bus.c	Mon Feb 12 13:59:20 2007 +1100
@@ -0,0 +1,180 @@
+#include <linux/init.h>
+#include <linux/bootmem.h>
+#include <asm/lguest_device.h>
+#include <asm/lguest.h>
+#include <asm/io.h>
+
+static ssize_t type_show(struct device *_dev,
+                         struct device_attribute *attr, char *buf)
+{
+	struct lguest_device *dev = container_of(_dev,struct lguest_device,dev);
+	return sprintf(buf, "%hu", lguest_devices[dev->index].type);
+}
+static ssize_t features_show(struct device *_dev,
+                             struct device_attribute *attr, char *buf)
+{
+	struct lguest_device *dev = container_of(_dev,struct lguest_device,dev);
+	return sprintf(buf, "%hx", lguest_devices[dev->index].features);
+}
+static ssize_t pfn_show(struct device *_dev,
+			 struct device_attribute *attr, char *buf)
+{
+	struct lguest_device *dev = container_of(_dev,struct lguest_device,dev);
+	return sprintf(buf, "%u", lguest_devices[dev->index].pfn);
+}
+static ssize_t status_show(struct device *_dev,
+                           struct device_attribute *attr, char *buf)
+{
+	struct lguest_device *dev = container_of(_dev,struct lguest_device,dev);
+	return sprintf(buf, "%hx", lguest_devices[dev->index].status);
+}
+static ssize_t status_store(struct device *_dev, struct device_attribute *attr,
+                            const char *buf, size_t count)
+{
+	struct lguest_device *dev = container_of(_dev,struct lguest_device,dev);
+	if (sscanf(buf, "%hi", &lguest_devices[dev->index].status) != 1)
+		return -EINVAL;
+	return count;
+}
+static struct device_attribute lguest_dev_attrs[] = {
+	__ATTR_RO(type),
+	__ATTR_RO(features),
+	__ATTR_RO(pfn),
+	__ATTR(status, 0644, status_show, status_store),
+	__ATTR_NULL
+};
+
+static int lguest_dev_match(struct device *_dev, struct device_driver *_drv)
+{
+	struct lguest_device *dev = container_of(_dev,struct lguest_device,dev);
+	struct lguest_driver *drv = container_of(_drv,struct lguest_driver,drv);
+
+	return (drv->device_type == lguest_devices[dev->index].type);
+}
+
+struct lguest_bus {
+	struct bus_type bus;
+	struct device dev;
+};
+
+static struct lguest_bus lguest_bus = {
+	.bus = {
+		.name  = "lguest",
+		.match = lguest_dev_match,
+		.dev_attrs = lguest_dev_attrs,
+	},
+	.dev = {
+		.parent = NULL,
+		.bus_id = "lguest",
+	}
+};
+
+static int lguest_dev_probe(struct device *_dev)
+{
+	int ret;
+	struct lguest_device *dev = container_of(_dev,struct lguest_device,dev);
+	struct lguest_driver *drv = container_of(dev->dev.driver,
+						struct lguest_driver, drv);
+
+	lguest_devices[dev->index].status |= LGUEST_DEVICE_S_DRIVER;
+	ret = drv->probe(dev);
+	if (ret == 0)
+		lguest_devices[dev->index].status |= LGUEST_DEVICE_S_DRIVER_OK;
+	return ret;
+}
+
+static int lguest_dev_remove(struct device *_dev)
+{
+	struct lguest_device *dev = container_of(_dev,struct lguest_device,dev);
+	struct lguest_driver *drv = container_of(dev->dev.driver,
+						struct lguest_driver, drv);
+
+	if (dev->dev.driver && drv->remove)
+		drv->remove(dev);
+	put_device(&dev->dev);
+	return 0;
+}
+
+int register_lguest_driver(struct lguest_driver *drv)
+{
+	if (!lguest_devices)
+		return 0;
+	
+	drv->drv.bus = &lguest_bus.bus;
+	drv->drv.name = drv->name;
+	drv->drv.owner = drv->owner;
+	drv->drv.probe = lguest_dev_probe;
+	drv->drv.remove = lguest_dev_remove;
+
+	return driver_register(&drv->drv);
+}
+EXPORT_SYMBOL_GPL(register_lguest_driver);
+
+void unregister_lguest_driver(struct lguest_driver *drv)
+{
+	if (!lguest_devices)
+		return;
+
+	driver_unregister(&drv->drv);
+}
+EXPORT_SYMBOL_GPL(unregister_lguest_driver);
+
+static void release_lguest_device(struct device *_dev)
+{
+	struct lguest_device *dev = container_of(_dev,struct lguest_device,dev);
+
+	lguest_devices[dev->index].status |= LGUEST_DEVICE_S_REMOVED_ACK;
+	kfree(dev);
+}
+
+static void add_lguest_device(unsigned int index)
+{
+	struct lguest_device *new;
+
+	lguest_devices[index].status |= LGUEST_DEVICE_S_ACKNOWLEDGE;
+	new = kmalloc(sizeof(struct lguest_device), GFP_KERNEL);
+	if (!new) {
+		printk(KERN_EMERG "Cannot allocate lguest device %u\n", index);
+		lguest_devices[index].status |= LGUEST_DEVICE_S_FAILED;
+		return;
+	}
+
+	new->index = index;
+	new->private = NULL;
+	memset(&new->dev, 0, sizeof(new->dev));
+	new->dev.parent = &lguest_bus.dev;
+	new->dev.bus = &lguest_bus.bus;
+	new->dev.release = release_lguest_device;
+	sprintf(new->dev.bus_id, "%u", index);
+	if (device_register(&new->dev) != 0) {
+		printk(KERN_EMERG "Cannot register lguest device %u\n", index);
+		lguest_devices[index].status |= LGUEST_DEVICE_S_FAILED;
+		kfree(new);
+	}
+}
+
+static void scan_devices(void)
+{
+	unsigned int i;
+
+	for (i = 0; i < LGUEST_MAX_DEVICES; i++)
+		if (lguest_devices[i].type)
+			add_lguest_device(i);
+}
+
+static int __init lguest_bus_init(void)
+{
+	if (strcmp(paravirt_ops.name, "lguest") != 0)
+		return 0;
+
+	/* Devices are in page above top of "normal" mem. */
+	lguest_devices = ioremap(max_pfn << PAGE_SHIFT, PAGE_SIZE);
+
+	if (bus_register(&lguest_bus.bus) != 0
+	    || device_register(&lguest_bus.dev) != 0)
+		panic("lguest bus registration failed");
+
+	scan_devices();
+	return 0;
+}
+postcore_initcall(lguest_bus_init);
diff -r 3c4f57d11d07 include/asm-i386/pda.h
--- a/include/asm-i386/pda.h	Mon Feb 12 13:00:10 2007 +1100
+++ b/include/asm-i386/pda.h	Mon Feb 12 13:59:20 2007 +1100
@@ -97,4 +97,5 @@ extern struct i386_pda _proxy_pda;
 #define sub_pda(field,val) pda_to_op("sub",field,val)
 #define or_pda(field,val) pda_to_op("or",field,val)
 
+extern struct i386_pda boot_pda;
 #endif	/* _I386_PDA_H */

[PATCH 4/8] lguest: Makefile

[Rusty Russell]

Finally, we put in the Makefile, so it will build.

Linking the switcher code (hypervisor.S) ready to be copied
into the top of memory is the only non-trivial thing here.

Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>

===================================================================
--- a/arch/i386/Makefile
+++ b/arch/i386/Makefile
@@ -108,6 +108,7 @@ drivers-$(CONFIG_PCI)			+= arch/i386/pci
 # must be linked after kernel/
 drivers-$(CONFIG_OPROFILE)		+= arch/i386/oprofile/
 drivers-$(CONFIG_PM)			+= arch/i386/power/
+drivers-$(CONFIG_LGUEST_GUEST)		+= arch/i386/lguest/
 
 CFLAGS += $(mflags-y)
 AFLAGS += $(mflags-y)
===================================================================
--- /dev/null
+++ b/arch/i386/lguest/Makefile
@@ -0,0 +1,22 @@
+# Guest requires the paravirt_ops replacement and the bus driver.
+obj-$(CONFIG_LGUEST_GUEST) += lguest.o lguest_bus.o
+
+# Host requires the other files, which can be a module.
+obj-$(CONFIG_LGUEST)	+= lg.o
+lg-objs := core.o hypercalls.o page_tables.o interrupts_and_traps.o \
+	segments.o io.o lguest_user.o
+
+# We use top 4MB for guest traps page, then hypervisor. */
+HYPE_ADDR := (0xFFC00000+4096)
+# The data is only 1k (256 interrupt handler pointers)
+HYPE_DATA_SIZE := 1024
+CFLAGS += -DHYPE_ADDR="$(HYPE_ADDR)" -DHYPE_DATA_SIZE="$(HYPE_DATA_SIZE)"
+
+$(obj)/core.o: $(obj)/hypervisor-blob.c
+# This links the hypervisor in the right place and turns it into a C array.
+$(obj)/hypervisor-raw: $(obj)/hypervisor.o
+	@$(LD) -static -Tdata=`printf %#x $$(($(HYPE_ADDR)))` -Ttext=`printf %#x $$(($(HYPE_ADDR)+$(HYPE_DATA_SIZE)))` -o $@ $< && $(OBJCOPY) -O binary $@
+$(obj)/hypervisor-blob.c: $(obj)/hypervisor-raw
+	@od -tx1 -An -v $< | sed -e 's/^ /0x/' -e 's/$$/,/' -e 's/ /,0x/g' > $@
+
+clean-files := hypervisor-blob.c hypervisor-raw

[PATCH 5/8] lguest: trivial guest network driver

[Rusty Russell]

This network driver operates both to the host process, and to other
guests.  It's pretty trivial.

Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>

===================================================================
--- a/drivers/net/Makefile
+++ b/drivers/net/Makefile
@@ -217,3 +217,4 @@ obj-$(CONFIG_FS_ENET) += fs_enet/
 obj-$(CONFIG_FS_ENET) += fs_enet/
 
 obj-$(CONFIG_NETXEN_NIC) += netxen/
+obj-$(CONFIG_LGUEST_GUEST) += lguest_net.o
===================================================================
--- /dev/null
+++ b/drivers/net/lguest_net.c
@@ -0,0 +1,400 @@
+/* A simple network driver for lguest.
+ *
+ * Copyright 2006 Rusty Russell <rusty@rustcorp.com.au> IBM Corporation
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307  USA
+ */
+//#define DEBUG
+#include <linux/netdevice.h>
+#include <linux/etherdevice.h>
+#include <linux/module.h>
+#include <linux/mm_types.h>
+#include <asm/io.h>
+#include <asm/lguest_device.h>
+
+#define SHARED_SIZE		PAGE_SIZE
+#define DATA_SIZE		1500
+#define MAX_LANS		4
+#define NUM_SKBS		8
+/* We overload multicast bit to show promiscuous mode. */
+#define PROMISC_BIT		0x80
+
+struct lguestnet_info
+{
+	/* The shared page. */
+	struct lguest_net *peer;
+	unsigned long peer_phys;
+
+	/* My peerid. */
+	unsigned int me;
+
+	struct net_device_stats stats;
+
+	/* Receive queue. */
+	struct sk_buff *skb[NUM_SKBS];
+	struct lguest_dma dma[NUM_SKBS];
+};
+
+/* How many bytes left in this page. */
+static unsigned int rest_of_page(void *data)
+{
+	return PAGE_SIZE - ((unsigned long)data % PAGE_SIZE);
+}
+
+/* Simple convention: offset 4 * peernum. */
+static unsigned long peer_addr(struct lguestnet_info *info, unsigned peernum)
+{
+	return info->peer_phys + 4 * peernum;
+}
+
+static void skb_to_dma(const struct sk_buff *skb, unsigned int len,
+		       struct lguest_dma *dma)
+{
+	unsigned int i, seg;
+
+	for (i = seg = 0; i < len; seg++, i += rest_of_page(skb->data + i)) {
+		dma->addr[seg] = virt_to_phys(skb->data + i);
+		dma->len[seg] = min((unsigned)(len - i),
+				    rest_of_page(skb->data + i));
+	}
+	for (i = 0; i < skb_shinfo(skb)->nr_frags; i++, seg++) {
+		const skb_frag_t *f = &skb_shinfo(skb)->frags[i];
+		/* Should not happen with MTU less than 64k - 2 * PAGE_SIZE. */
+		if (seg == LGUEST_MAX_DMA_SECTIONS) {
+			printk("Woah dude!  Megapacket!\n");
+			break;
+		}
+		dma->addr[seg] = page_to_phys(f->page) + f->page_offset;
+		dma->len[seg] = f->size;
+	}
+	if (seg < LGUEST_MAX_DMA_SECTIONS)
+		dma->len[seg] = 0;
+}
+
+static void transfer_packet(struct net_device *dev,
+			    struct sk_buff *skb,
+			    unsigned int peernum)
+{
+	struct lguestnet_info *info = dev->priv;
+	struct lguest_dma dma;
+
+	skb_to_dma(skb, skb->len, &dma);
+	pr_debug("xfer length %04x (%u)\n", htons(skb->len), skb->len);
+
+	hcall(LHCALL_SEND_DMA, peer_addr(info,peernum), __pa(&dma),0);
+	if (dma.used_len != skb->len) {
+		info->stats.tx_carrier_errors++;
+		pr_debug("Bad xfer to peer %i: %i of %i (dma %p/%i)\n",
+			 peernum, dma.used_len, skb->len,
+			 (void *)dma.addr[0], dma.len[0]);
+	} else {
+		pr_debug("lguestnet: sent %u bytes\n", skb->len);
+		info->stats.tx_bytes += skb->len;
+		info->stats.tx_packets++;
+	}
+}
+
+static int mac_eq(const unsigned char mac[ETH_ALEN],
+		  struct lguestnet_info *info, unsigned int peer)
+{
+	/* Ignore multicast bit, which peer turns on to mean promisc. */
+	if ((info->peer[peer].promisc & (~PROMISC_BIT)) != mac[0])
+		return 0;
+	return memcmp(mac+1, info->peer[peer].mac+1, ETH_ALEN-1) == 0;
+}
+
+static int unused_peer(const struct lguest_net peer[], unsigned int num)
+{
+	return peer[num].guestid == 0xFFFF;
+}
+
+static int is_broadcast(const unsigned char dest[ETH_ALEN])
+{
+	return dest[0] == 0xFF && dest[1] == 0xFF && dest[2] == 0xFF
+		&& dest[3] == 0xFF && dest[4] == 0xFF && dest[5] == 0xFF;
+}
+
+static int promisc(struct lguestnet_info *info, unsigned int peer)
+{
+	return info->peer[peer].promisc & PROMISC_BIT;
+}
+
+static void lguestnet_set_multicast(struct net_device *dev)
+{
+	struct lguestnet_info *info = dev->priv;
+
+	if (dev->flags & IFF_PROMISC)
+		info->peer[info->me].promisc |= PROMISC_BIT;
+	else
+		info->peer[info->me].promisc &= ~PROMISC_BIT;
+}
+
+static int lguestnet_start_xmit(struct sk_buff *skb, struct net_device *dev)
+{
+	unsigned int i;
+	int broadcast;
+	struct lguestnet_info *info = dev->priv;
+	const unsigned char *dest = ((struct ethhdr *)skb->data)->h_dest;
+
+	broadcast = is_broadcast(dest);
+
+	pr_debug("lguestnet %s: xmit broadcast=%i\n",
+		 dev->name, broadcast);
+	pr_debug("dest: %02x:%02x:%02x:%02x:%02x:%02x\n",
+		 dest[0], dest[1], dest[2], dest[3], dest[4], dest[5]);
+
+	for (i = 0; i < SHARED_SIZE/sizeof(struct lguest_net); i++) {
+		if (i == info->me || unused_peer(info->peer, i))
+			continue;
+
+		if (!broadcast && !promisc(info, i) && !mac_eq(dest, info, i))
+			continue;
+
+		pr_debug("lguestnet %s: sending from %i to %i\n",
+			 dev->name, info->me, i);
+		transfer_packet(dev, skb, i);
+	}
+	dev_kfree_skb(skb);
+	return 0;
+}
+
+static struct sk_buff *lguestnet_alloc_skb(struct net_device *dev, int gfpflags)
+{
+	struct sk_buff *skb;
+
+	skb = alloc_skb(16 + ETH_HLEN + DATA_SIZE, gfpflags);
+	if (!skb)
+		return NULL;
+
+	skb->dev = dev;
+	skb_reserve(skb, 16);
+	return skb;
+}
+
+/* Find a new skb to put in this slot in shared mem. */
+static int fill_slot(struct net_device *dev, unsigned int slot)
+{
+	struct lguestnet_info *info = dev->priv;
+	/* Try to create and register a new one. */
+	info->skb[slot] = lguestnet_alloc_skb(dev, GFP_ATOMIC);
+	if (!info->skb[slot]) {
+		printk("%s: could not fill slot %i\n", dev->name, slot);
+		return -ENOMEM;
+	}
+
+	skb_to_dma(info->skb[slot], ETH_HLEN + DATA_SIZE, &info->dma[slot]);
+	wmb();
+	/* Now we tell hypervisor it can use the slot. */
+	info->dma[slot].used_len = 0;
+
+	pr_debug("lguestnet: %s populating slot %i with %p\n",
+		 dev->name, slot, info->skb[slot]);
+	return 0;
+}
+
+static irqreturn_t lguestnet_rcv(int irq, void *dev_id)
+{
+	struct net_device *dev = dev_id;
+	struct lguestnet_info *info = dev->priv;
+	unsigned int i, done = 0;
+
+	for (i = 0; i < ARRAY_SIZE(info->dma); i++) {
+		unsigned int length;
+		struct sk_buff *skb;
+
+		length = info->dma[i].used_len;
+		if (length == 0)
+			continue;
+
+		done++;
+		skb = info->skb[i];
+		fill_slot(dev, i);
+
+		pr_debug("Received skb %p\n", skb);
+		if (length < 14 || length > 1514) {
+			pr_debug(KERN_WARNING "%s: unbelievable skb len: %i\n",
+				 dev->name, length);
+			dev_kfree_skb(skb);
+			continue;
+		}
+
+		skb_put(skb, length);
+		skb->protocol = eth_type_trans(skb, dev);
+		/* This is a reliable transport. */
+		skb->ip_summed = CHECKSUM_UNNECESSARY;
+		pr_debug("Receiving skb proto 0x%04x len %i type %i\n",
+			 ntohs(skb->protocol), skb->len,skb->pkt_type);
+
+		info->stats.rx_bytes += skb->len;
+		info->stats.rx_packets++;
+		netif_rx(skb);
+	}
+	return done ? IRQ_HANDLED : IRQ_NONE;
+}
+
+static int populate_page(struct net_device *dev)
+{
+	int i;
+	struct lguestnet_info *info = dev->priv;
+
+	pr_debug("lguestnet: peer %i shared page %p me %p\n",
+	         info->me, info->peer, &info->peer[info->me]);
+
+	/* Set up our MAC address */
+	memcpy(info->peer[info->me].mac, dev->dev_addr, ETH_ALEN);
+
+	/* Turn on promisc mode if needed */
+	lguestnet_set_multicast(dev);
+
+	for (i = 0; i < ARRAY_SIZE(info->dma); i++) {
+		if (fill_slot(dev, i) != 0)
+			goto cleanup;
+	}
+	pr_debug("lguestnet: allocated %i dma recv buffers\n", i);
+	if (!hcall(LHCALL_BIND_DMA, peer_addr(info, info->me), __pa(info->dma),
+		   (NUM_SKBS << 8) | dev->irq))
+		goto cleanup;
+	return 0;
+
+cleanup:
+	while (--i >= 0)
+		dev_kfree_skb(info->skb[i]);
+	return -ENOMEM;
+}
+
+static void unpopulate_page(struct lguestnet_info *info)
+{
+	unsigned int i;
+	struct lguest_net *me = &info->peer[info->me];
+
+	/* Clear all trace: others might deliver packets, we'll ignore it. */
+	memset(me, 0, sizeof(*me));
+
+	/* Deregister sg lists. */
+	hcall(LHCALL_BIND_DMA, peer_addr(info, info->me), __pa(info->dma), 0);
+	for (i = 0; i < ARRAY_SIZE(info->dma); i++)
+		dev_kfree_skb(info->skb[i]);
+}
+
+static int lguestnet_open(struct net_device *dev)
+{
+	return populate_page(dev);
+}
+
+static int lguestnet_close(struct net_device *dev)
+{
+	unpopulate_page(dev->priv);
+	return 0;
+}
+
+static struct net_device_stats *lguestnet_get_stats(struct net_device *dev)
+{
+	struct lguestnet_info *info = dev->priv;
+
+	return &info->stats;
+}
+
+static int lguestnet_probe(struct lguest_device *lhdev)
+{
+	int err, irqf = IRQF_SHARED;
+	unsigned long mapsize;
+	struct net_device *dev;
+	struct lguestnet_info *info;
+	struct lguest_device_desc *desc = &lguest_devices[lhdev->index];
+
+	pr_debug("lguest_net: probing for device %i\n", lhdev->index);
+	mapsize = PAGE_SIZE * desc->num_pages;
+
+	dev = alloc_etherdev(sizeof(struct lguestnet_info));
+	if (!dev)
+		return -ENOMEM;
+
+	SET_MODULE_OWNER(dev);
+
+	/* Ethernet defaults with some changes */
+	ether_setup(dev);
+	dev->set_mac_address = NULL;
+	dev->mtu = DATA_SIZE;
+
+	dev->dev_addr[0] = 0x02; /* set local assignment bit (IEEE802) */
+	dev->dev_addr[1] = 0x00;
+	memcpy(&dev->dev_addr[2], &lguest_data.guestid, 2);
+	dev->dev_addr[4] = 0x00;
+	dev->dev_addr[5] = 0x00;
+
+	dev->open = lguestnet_open;
+	dev->stop = lguestnet_close;
+	dev->hard_start_xmit = lguestnet_start_xmit;
+	dev->get_stats = lguestnet_get_stats;
+
+	/* Turning on/off promisc will call dev->set_multicast_list.
+	 * We don't actually support multicast yet */
+	dev->set_multicast_list = lguestnet_set_multicast;
+	dev->mem_start = ((unsigned long)desc->pfn << PAGE_SHIFT);
+	dev->mem_end = dev->mem_start + mapsize;
+	dev->irq = lhdev->index+1;
+	dev->dma = 0;
+	dev->features = NETIF_F_SG;
+	if (desc->features & LGUEST_NET_F_NOCSUM)
+		dev->features |= NETIF_F_NO_CSUM;
+
+	info = dev->priv;
+	info->peer_phys = ((unsigned long)desc->pfn << PAGE_SHIFT);
+	info->peer = (void *)ioremap(info->peer_phys, mapsize);
+	/* This stores our peerid (upper bits reserved for future). */
+	info->me = (desc->features & (mapsize-1));
+
+	/* skbs allocated on open */
+	memset(info->skb, 0, sizeof(info->skb));
+
+	err = register_netdev(dev);
+	if (err) {
+		pr_debug("lguestnet: registering device failed\n");
+		goto free;
+	}
+
+	if (lguest_devices[lhdev->index].features & LGUEST_DEVICE_F_RANDOMNESS)
+		irqf |= IRQF_SAMPLE_RANDOM;
+	if (request_irq(dev->irq, lguestnet_rcv, irqf, "lguestnet", dev) != 0) {
+		pr_debug("lguestnet: could not get net irq %i\n", dev->irq);
+		goto unregister;
+	}
+
+	pr_debug("lguestnet: registered device %s\n", dev->name);
+	lhdev->private = dev;
+	return 0;
+
+unregister:
+	unregister_netdev(dev);
+free:
+	free_netdev(dev);
+	return err;
+}
+
+static struct lguest_driver lguestnet_drv = {
+	.name = "lguestnet",
+	.owner = THIS_MODULE,
+	.device_type = LGUEST_DEVICE_T_NET,
+	.probe = lguestnet_probe,
+};
+
+static __init int lguestnet_init(void)
+{
+	return register_lguest_driver(&lguestnet_drv);
+}
+module_init(lguestnet_init);
+
+MODULE_DESCRIPTION("Lguest network driver");
+MODULE_LICENSE("GPL");

[PATCH 6/8] lguest: trivial guest console driver

[Rusty Russell]

A trivial driver to have a basic lguest console, using the hvc_console
infrastructure.

Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>

diff -r aaa62bd9788a drivers/char/Makefile
--- a/drivers/char/Makefile	Mon Feb 12 13:01:16 2007 +1100
+++ b/drivers/char/Makefile	Mon Feb 12 13:01:45 2007 +1100
@@ -44,6 +44,7 @@ obj-$(CONFIG_SX)		+= sx.o generic_serial
 obj-$(CONFIG_SX)		+= sx.o generic_serial.o
 obj-$(CONFIG_RIO)		+= rio/ generic_serial.o
 obj-$(CONFIG_HVC_CONSOLE)	+= hvc_vio.o hvsi.o
+obj-$(CONFIG_LGUEST_GUEST)	+= hvc_lguest.o
 obj-$(CONFIG_HVC_ISERIES)	+= hvc_iseries.o
 obj-$(CONFIG_HVC_RTAS)		+= hvc_rtas.o
 obj-$(CONFIG_HVC_DRIVER)	+= hvc_console.o
diff -r aaa62bd9788a drivers/char/hvc_lguest.c
--- /dev/null	Thu Jan 01 00:00:00 1970 +0000
+++ b/drivers/char/hvc_lguest.c	Mon Feb 12 13:01:19 2007 +1100
@@ -0,0 +1,99 @@
+/* Simple console for lguest.
+ *
+ * Copyright (C) 2006 Rusty Russell, IBM Corporation
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307  USA
+ */
+#include <linux/err.h>
+#include <linux/init.h>
+#include <asm/lguest_device.h>
+#include "hvc_console.h"
+
+static int cons_irq;
+static int cons_offset;
+static char inbuf[256];
+static struct lguest_dma cons_input = { .used_len = 0,
+					.addr[0] = __pa(inbuf),
+					.len[0] = sizeof(inbuf),
+					.len[1] = 0 };
+
+static int get_chars(u32 vtermno, char *buf, int count)
+{
+	if (!cons_input.used_len)
+		return 0;
+
+	if (cons_input.used_len - cons_offset < count)
+		count = cons_input.used_len - cons_offset;
+
+	memcpy(buf, inbuf + cons_offset, count);
+	cons_offset += count;
+	if (cons_offset == cons_input.used_len) {
+		cons_offset = 0;
+		cons_input.used_len = 0;
+	}
+	return count;
+}
+
+static int put_chars(u32 vtermno, const char *buf, int count)
+{
+	struct lguest_dma dma;
+
+	/* FIXME: what if it's over a page boundary? */
+	dma.len[0] = count;
+	dma.len[1] = 0;
+	dma.addr[0] = __pa(buf);
+
+	hcall(LHCALL_SEND_DMA, 4, __pa(&dma), 0);
+	return count;
+}
+
+struct hv_ops lguest_cons = {
+	.get_chars = get_chars,
+	.put_chars = put_chars,
+};
+
+static int __init cons_init(void)
+{
+	if (strcmp(paravirt_ops.name, "lguest") != 0)
+		return 0;
+
+	return hvc_instantiate(0, 0, &lguest_cons);
+}
+console_initcall(cons_init);
+
+static int lguestcons_probe(struct lguest_device *lhdev)
+{
+	cons_irq = lhdev->index+1;
+	lhdev->private = hvc_alloc(0, cons_irq, &lguest_cons, 256);
+	if (IS_ERR(lhdev->private))
+		return PTR_ERR(lhdev->private);
+
+	if (!hcall(LHCALL_BIND_DMA, 0, __pa(&cons_input), (1<<8)+cons_irq))
+		printk("lguest console: failed to bind buffer.\n");
+	return 0;
+}
+
+static struct lguest_driver lguestcons_drv = {
+	.name = "lguestcons",
+	.owner = THIS_MODULE,
+	.device_type = LGUEST_DEVICE_T_CONSOLE,
+	.probe = lguestcons_probe,
+};
+
+static int __init hvc_lguest_init(void)
+{
+	return register_lguest_driver(&lguestcons_drv);
+}
+module_init(hvc_lguest_init);

[PATCH 7/8] lguest: trivial guest block driver

[Rusty Russell]

A simple block driver for lguest (/dev/lgbX).  Only does one request
at once.

Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>

===================================================================
--- a/drivers/block/Makefile
+++ b/drivers/block/Makefile
@@ -28,4 +28,5 @@ obj-$(CONFIG_VIODASD)		+= viodasd.o
 obj-$(CONFIG_VIODASD)		+= viodasd.o
 obj-$(CONFIG_BLK_DEV_SX8)	+= sx8.o
 obj-$(CONFIG_BLK_DEV_UB)	+= ub.o
+obj-$(CONFIG_LGUEST_GUEST)	+= lguest_blk.o
 
===================================================================
--- /dev/null
+++ b/drivers/block/lguest_blk.c
@@ -0,0 +1,260 @@
+/* A simple block driver for lguest.
+ *
+ * Copyright 2006 Rusty Russell <rusty@rustcorp.com.au> IBM Corporation
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307  USA
+ */
+//#define DEBUG
+#include <linux/init.h>
+#include <linux/types.h>
+#include <linux/blkdev.h>
+#include <linux/interrupt.h>
+#include <asm/lguest_device.h>
+
+static char next_block_index = 'a';
+
+struct blockdev
+{
+	spinlock_t lock;
+
+	/* The disk structure for the kernel. */
+	struct gendisk *disk;
+
+	/* The major number for this disk. */
+	int major;
+	int irq;
+
+	unsigned long phys_addr;
+	/* The ioremap'ed block page. */
+	struct lguest_block_page *lb_page;
+
+	/* We only have a single request outstanding at a time. */
+	struct lguest_dma dma;
+	struct request *req;
+};
+
+static irqreturn_t lgb_irq(int irq, void *_bd)
+{
+	struct blockdev *bd = _bd;
+	unsigned long flags;
+
+	if (!bd->req) {
+		pr_debug("No work!\n");
+		return IRQ_NONE;
+	}
+
+	if (!bd->lb_page->result) {
+		pr_debug("No result!\n");
+		return IRQ_NONE;
+	}
+
+	spin_lock_irqsave(&bd->lock, flags);
+	end_request(bd->req, bd->lb_page->result == 1);
+	bd->req = NULL;
+	bd->dma.used_len = 0;
+	blk_start_queue(bd->disk->queue);
+	spin_unlock_irqrestore(&bd->lock, flags);
+	return IRQ_HANDLED;
+}
+
+static unsigned int req_to_dma(struct request *req, struct lguest_dma *dma)
+{
+	unsigned int i = 0, idx, len = 0;
+	struct bio *bio;
+
+	rq_for_each_bio(bio, req) {
+		struct bio_vec *bvec;
+		bio_for_each_segment(bvec, bio, idx) {
+			BUG_ON(i == LGUEST_MAX_DMA_SECTIONS);
+			BUG_ON(!bvec->bv_len);
+			dma->addr[i] = page_to_phys(bvec->bv_page)
+				+ bvec->bv_offset;
+			dma->len[i] = bvec->bv_len;
+			len += bvec->bv_len;
+			i++;
+		}
+	}
+	if (i < LGUEST_MAX_DMA_SECTIONS)
+		dma->len[i] = 0;
+	return len;
+}
+
+static void empty_dma(struct lguest_dma *dma)
+{
+	dma->len[0] = 0;
+}
+
+static void setup_req(struct blockdev *bd,
+		      int type, struct request *req, struct lguest_dma *dma)
+{
+	bd->lb_page->type = type;
+	bd->lb_page->sector = req->sector;
+	bd->lb_page->result = 0;
+	bd->req = req;
+	bd->lb_page->bytes = req_to_dma(req, dma);
+}
+
+static int do_write(struct blockdev *bd, struct request *req)
+{
+	struct lguest_dma send;
+
+	pr_debug("lgb: WRITE sector %li\n", (long)req->sector);
+	setup_req(bd, 1, req, &send);
+
+	hcall(LHCALL_SEND_DMA, bd->phys_addr, __pa(&send), 0);
+	return 1;
+}
+
+static int do_read(struct blockdev *bd, struct request *req)
+{
+	struct lguest_dma ping;
+
+	pr_debug("lgb: READ sector %li\n", (long)req->sector);
+	setup_req(bd, 0, req, &bd->dma);
+
+	empty_dma(&ping);
+	hcall(LHCALL_SEND_DMA,bd->phys_addr,__pa(&ping),0);
+	return 1;
+}
+
+static void do_lgb_request(request_queue_t *q)
+{
+	struct blockdev *bd;
+	struct request *req;
+	int ok;
+
+again:
+	req = elv_next_request(q);
+	if (!req)
+		return;
+
+	bd = req->rq_disk->private_data;
+	/* Sometimes we get repeated requests after blk_stop_queue. */
+	if (bd->req)
+		return;
+
+	if (!blk_fs_request(req)) {
+		pr_debug("Got non-command 0x%08x\n", req->cmd_type);
+	error:
+		req->errors++;
+		end_request(req, 0);
+		goto again;
+	} else {
+		if (rq_data_dir(req) == WRITE)
+			ok = do_write(req->rq_disk->private_data, req);
+		else
+			ok = do_read(req->rq_disk->private_data, req);
+
+		if (!ok)
+			goto error;
+		/* Wait for interrupt to tell us it's done. */
+		blk_stop_queue(q);
+	}
+}
+
+static struct block_device_operations lguestblk_fops = {
+	.owner = THIS_MODULE,
+};
+
+static int lguestblk_probe(struct lguest_device *lhdev)
+{
+	struct blockdev *bd;
+	int err;
+	int irqflags = IRQF_SHARED;
+
+	bd = kmalloc(sizeof(*bd), GFP_KERNEL);
+	if (!bd)
+		return -ENOMEM;
+
+	spin_lock_init(&bd->lock);
+	bd->phys_addr = (lguest_devices[lhdev->index].pfn << PAGE_SHIFT);
+
+	bd->disk = alloc_disk(1);
+	if (!bd->disk) {
+		err = -ENOMEM;
+		goto out_free_bd;
+	}
+
+	bd->disk->queue = blk_init_queue(do_lgb_request, &bd->lock);
+	if (!bd->disk->queue) {
+		err = -ENOMEM;
+		goto out_put;
+	}
+
+	/* We can only handle a certain number of sg entries */
+	blk_queue_max_hw_segments(bd->disk->queue, LGUEST_MAX_DMA_SECTIONS);
+	/* Buffers must not cross page boundaries */
+	blk_queue_segment_boundary(bd->disk->queue, PAGE_SIZE-1);
+
+	bd->irq = lhdev->index+1;
+	bd->major = register_blkdev(0, "lguestblk");
+	if (bd->major < 0) {
+		err = bd->major;
+		goto out_cleanup_queue;
+	}
+	bd->lb_page = (void *)ioremap(bd->phys_addr, PAGE_SIZE);
+	bd->req = NULL;
+
+	sprintf(bd->disk->disk_name, "lgb%c", next_block_index++);
+	if (lguest_devices[lhdev->index].features & LGUEST_DEVICE_F_RANDOMNESS)
+		irqflags |= IRQF_SAMPLE_RANDOM;
+	err = request_irq(bd->irq, lgb_irq, irqflags, bd->disk->disk_name, bd);
+	if (err)
+		goto out_unmap;
+
+	bd->dma.used_len = 0;
+	bd->dma.len[0] = 0;
+	hcall(LHCALL_BIND_DMA, bd->phys_addr, __pa(&bd->dma), (1<<8)+bd->irq);
+
+	printk(KERN_INFO "%s: device %i at major %d\n",
+	       bd->disk->disk_name, lhdev->index, bd->major);
+
+	bd->disk->major = bd->major;
+	bd->disk->first_minor = 0;
+	bd->disk->private_data = bd;
+	bd->disk->fops = &lguestblk_fops;
+	/* This is initialized to the disk size by the other end. */
+	set_capacity(bd->disk, bd->lb_page->num_sectors);
+	add_disk(bd->disk);
+
+	lhdev->private = bd;
+	return 0;
+
+out_unmap:
+	iounmap(bd->lb_page);
+out_cleanup_queue:
+	blk_cleanup_queue(bd->disk->queue);
+out_put:
+	put_disk(bd->disk);
+out_free_bd:
+	kfree(bd);
+	return err;
+}
+
+static struct lguest_driver lguestblk_drv = {
+	.name = "lguestblk",
+	.owner = THIS_MODULE,
+	.device_type = LGUEST_DEVICE_T_BLOCK,
+	.probe = lguestblk_probe,
+};
+
+static __init int lguestblk_init(void)
+{
+	return register_lguest_driver(&lguestblk_drv);
+}
+module_init(lguestblk_init);
+
+MODULE_DESCRIPTION("Lguest block driver");
+MODULE_LICENSE("GPL");

[PATCH 8/8] lguest: documentatation and example launcher

[Rusty Russell]

Fairly complete documentation for lguest.  I actually want to get rid
of the "coding" part of lguest.txt and roll it into the code itself,
literary-programming-style.

The launcher utility is also here: I don't have delusions of interface
stability, so it makes sense to have it here as an example, and it's
only 1000 lines.

Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>

diff -r 8806a441a0b1 Documentation/dontdiff
--- a/Documentation/dontdiff	Mon Feb 12 13:02:02 2007 +1100
+++ b/Documentation/dontdiff	Mon Feb 12 13:47:43 2007 +1100
@@ -144,3 +144,6 @@ wanxlfw.inc
 wanxlfw.inc
 uImage
 zImage
+hypervisor-blob.c
+lguest.lds
+hypervisor-raw
diff -r 8806a441a0b1 Documentation/lguest/Makefile
--- /dev/null	Thu Jan 01 00:00:00 1970 +0000
+++ b/Documentation/lguest/Makefile	Mon Feb 12 13:48:13 2007 +1100
@@ -0,0 +1,21 @@
+# This creates the demonstration utility "lguest" which runs a Linux guest.
+
+# We rely on CONFIG_PAGE_OFFSET to know where to put lguest binary.
+# Some shells (dash - ubunu) can't handle numbers that big so we cheat.
+include ../../.config
+LGUEST_GUEST_TOP := ($(CONFIG_PAGE_OFFSET) - 0x08000000)
+
+CFLAGS:=-Wall -Wmissing-declarations -Wmissing-prototypes -O3 \
+	-static -DLGUEST_GUEST_TOP="$(LGUEST_GUEST_TOP)" -Wl,-T,lguest.lds
+LDLIBS:=-lz
+
+all: lguest.lds lguest
+
+# The linker script on x86 is so complex the only way of creating one
+# which will link our binary in the right place is to mangle the
+# default one.
+lguest.lds:
+	$(LD) --verbose | awk '/^==========/ { PRINT=1; next; } /SIZEOF_HEADERS/ { gsub(/0x[0-9A-F]*/, "$(LGUEST_GUEST_TOP)") } { if (PRINT) print $$0; }' > $@
+
+clean:
+	rm -f lguest.lds lguest
diff -r 8806a441a0b1 Documentation/lguest/lguest.c
--- /dev/null	Thu Jan 01 00:00:00 1970 +0000
+++ b/Documentation/lguest/lguest.c	Mon Feb 12 13:47:43 2007 +1100
@@ -0,0 +1,989 @@
+/* Simple program to layout "physical" memory for new lguest guest.
+ * Linked high to avoid likely physical memory.  */
+#define _LARGEFILE64_SOURCE
+#define _GNU_SOURCE
+#include <stdio.h>
+#include <string.h>
+#include <unistd.h>
+#include <err.h>
+#include <stdint.h>
+#include <stdlib.h>
+#include <elf.h>
+#include <sys/mman.h>
+#include <sys/types.h>
+#include <sys/stat.h>
+#include <sys/wait.h>
+#include <fcntl.h>
+#include <assert.h>
+#include <stdbool.h>
+#include <errno.h>
+#include <signal.h>
+#include <sys/socket.h>
+#include <sys/ioctl.h>
+#include <sys/time.h>
+#include <time.h>
+#include <netinet/in.h>
+#include <linux/if.h>
+#include <linux/if_tun.h>
+#include <sys/uio.h>
+#include <termios.h>
+#include <zlib.h>
+typedef uint32_t u32;
+typedef uint16_t u16;
+typedef uint8_t u8;
+
+#include "../../include/asm/lguest_user.h"
+
+#define PAGE_PRESENT 0x7 	/* Present, RW, Execute */
+#define NET_PEERNUM 1
+
+static bool verbose;
+#define verbose(args...) \
+	do { if (verbose) printf(args); fflush(stdout); } while(0)
+
+struct devices
+{
+	fd_set infds;
+	int max_infd;
+
+	struct device *dev;
+};
+
+struct device
+{
+	struct device *next;
+	struct lguest_device_desc *desc;
+	void *mem;
+
+	/* Watch this fd if handle_input non-NULL. */
+	int fd;
+	int (*handle_input)(int fd, struct device *me);
+
+	/* Watch DMA to this address if handle_input non-NULL. */
+	unsigned long watch_address;
+	u32 (*handle_output)(int fd, const struct iovec *iov,
+			     unsigned int num, struct device *me);
+
+	/* Device-specific data. */
+	void *priv;
+};
+
+static char buf[1024];
+static struct iovec discard_iov = { .iov_base=buf, .iov_len=sizeof(buf) };
+static int zero_fd;
+
+static u32 memparse(const char *ptr)
+{
+	char *end;
+	unsigned long ret = strtoul(ptr, &end, 0);
+
+	switch (*end) {
+	case 'G':
+	case 'g':
+		ret <<= 10;
+	case 'M':
+	case 'm':
+		ret <<= 10;
+	case 'K':
+	case 'k':
+		ret <<= 10;
+		end++;
+	default:
+		break;
+	}
+	return ret;
+}
+
+static inline unsigned long page_align(unsigned long addr)
+{
+	return ((addr + getpagesize()-1) & ~(getpagesize()-1));
+}
+
+/* initrd gets loaded at top of memory: return length. */
+static unsigned long load_initrd(const char *name, unsigned long end)
+{
+	int ifd;
+	struct stat st;
+	void *iaddr;
+
+	if (!name)
+		return 0;
+
+	ifd = open(name, O_RDONLY, 0);
+	if (ifd < 0)
+		err(1, "Opening initrd '%s'", name);
+		
+	if (fstat(ifd, &st) < 0)
+		err(1, "fstat() on initrd '%s'", name);
+
+	iaddr = mmap((void *)end - st.st_size, st.st_size,
+		     PROT_READ|PROT_EXEC|PROT_WRITE,
+		     MAP_FIXED|MAP_PRIVATE, ifd, 0);
+	if (iaddr != (void *)end - st.st_size)
+		err(1, "Mmaping initrd '%s' returned %p not %p",
+		    name, iaddr, (void *)end - st.st_size);
+	close(ifd);
+	verbose("mapped initrd %s size=%lu @ %p\n", name, st.st_size, iaddr);
+	return st.st_size;
+}
+
+/* First map /dev/zero over entire memory, then insert kernel. */
+static void map_memory(unsigned long mem)
+{
+	if (mmap(0, mem,
+		 PROT_READ|PROT_WRITE|PROT_EXEC,
+		 MAP_FIXED|MAP_PRIVATE, zero_fd, 0) != (void *)0)
+		err(1, "Mmaping /dev/zero for %li bytes", mem);
+}
+
+static u32 finish(unsigned long mem, unsigned long *page_offset,
+		  const char *initrd, unsigned long *ird_size)
+{
+	u32 *pgdir = NULL, *linear = NULL;
+	int i, pte_pages;
+
+	/* This is a top of mem. */
+	*ird_size = load_initrd(initrd, mem);
+
+	/* Below initrd is used as top level of pagetable. */
+	pte_pages = 1 + (mem/getpagesize() + 1023)/1024;
+
+	pgdir = (u32 *)page_align(mem - *ird_size - pte_pages*getpagesize());
+	linear = (void *)pgdir + getpagesize();
+
+	/* Linear map all of memory at page_offset (to top of mem). */
+	if (mem > -*page_offset)
+		mem = -*page_offset;
+
+	for (i = 0; i < mem / getpagesize(); i++)
+		linear[i] = ((i * getpagesize()) | PAGE_PRESENT);
+	verbose("Linear %p-%p (%i-%i) = %#08x-%#08x\n",
+		linear, linear+i-1, 0, i-1, linear[0], linear[i-1]);
+
+	/* Now set up pgd so that this memory is at page_offset */
+	for (i = 0; i < mem / getpagesize(); i += getpagesize()/sizeof(u32)) {
+		pgdir[(i + *page_offset/getpagesize())/1024] 
+			= (((u32)linear + i*sizeof(u32)) | PAGE_PRESENT);
+		verbose("Top level %lu = %#08x\n",
+			(i + *page_offset/getpagesize())/1024,
+			pgdir[(i + *page_offset/getpagesize())/1024]);
+	}
+
+	return (unsigned long)pgdir;
+}
+
+/* Returns the entry point */
+static u32 map_elf(int elf_fd, const Elf32_Ehdr *ehdr, unsigned long mem,
+		   unsigned long *pgdir_addr,
+		   const char *initrd, unsigned long *ird_size,
+		   unsigned long *page_offset)
+{
+	void *addr;
+	Elf32_Phdr phdr[ehdr->e_phnum];
+	unsigned int i;
+
+	/* Sanity checks. */
+	if (ehdr->e_type != ET_EXEC
+	    || ehdr->e_machine != EM_386
+	    || ehdr->e_phentsize != sizeof(Elf32_Phdr)
+	    || ehdr->e_phnum < 1 || ehdr->e_phnum > 65536U/sizeof(Elf32_Phdr))
+		errx(1, "Malformed elf header");
+
+	if (lseek(elf_fd, ehdr->e_phoff, SEEK_SET) < 0)
+		err(1, "Seeking to program headers");
+	if (read(elf_fd, phdr, sizeof(phdr)) != sizeof(phdr))
+		err(1, "Reading program headers");
+
+	map_memory(mem);
+
+	*page_offset = 0;
+	/* We map the loadable segments at virtual addresses corresponding
+	 * to their physical addresses (our virtual == guest physical). */
+	for (i = 0; i < ehdr->e_phnum; i++) {
+		if (phdr[i].p_type != PT_LOAD)
+			continue;
+
+		verbose("Section %i: size %i addr %p\n",
+			i, phdr[i].p_memsz, (void *)phdr[i].p_paddr);
+		/* We map everything private, writable. */
+		if (phdr[i].p_paddr + phdr[i].p_memsz > mem)
+			errx(1, "Segment %i overlaps end of memory", i);
+
+		/* We expect linear address space. */
+		if (!*page_offset)
+			*page_offset = phdr[i].p_vaddr - phdr[i].p_paddr;
+		else if (*page_offset != phdr[i].p_vaddr - phdr[i].p_paddr)
+			errx(1, "Page offset of section %i different", i);
+
+		/* Recent ld versions don't page align any more. */
+		if (phdr[i].p_paddr % getpagesize()) {
+			phdr[i].p_filesz += (phdr[i].p_paddr % getpagesize());
+			phdr[i].p_offset -= (phdr[i].p_paddr % getpagesize());
+			phdr[i].p_paddr -= (phdr[i].p_paddr % getpagesize());
+		}
+		addr = mmap((void *)phdr[i].p_paddr,
+			    phdr[i].p_filesz,
+			    PROT_READ|PROT_WRITE|PROT_EXEC,
+			    MAP_FIXED|MAP_PRIVATE,
+			    elf_fd, phdr[i].p_offset);
+		if (addr != (void *)phdr[i].p_paddr)
+			err(1, "Mmaping vmlinux segment %i returned %p not %p (%p)",
+			    i, addr, (void *)phdr[i].p_paddr, &phdr[i].p_paddr);
+	}
+
+	*pgdir_addr = finish(mem, page_offset, initrd, ird_size);
+	/* Entry is physical address: convert to virtual */
+	return ehdr->e_entry + *page_offset;
+}
+
+static unsigned long intuit_page_offset(unsigned char *img, unsigned long len)
+{
+	unsigned int i, possibilities[256];
+
+	for (i = 0; i + 4 < len; i++) {
+		/* mov 0xXXXXXXXX,%eax */
+		if (img[i] == 0xA1 && ++possibilities[img[i+4]] > 3)
+			return (unsigned long)img[i+4] << 24;
+	}
+	errx(1, "could not determine page offset");
+}
+
+static u32 bzimage(int fd, unsigned long mem, unsigned long *pgdir_addr,
+		   const char *initrd, unsigned long *ird_size,
+		   unsigned long *page_offset)
+{
+	gzFile f;
+	int ret, len = 0;
+	void *img = (void *)0x100000;
+
+	map_memory(mem);
+
+	f = gzdopen(fd, "rb");
+	if (gzdirect(f))
+		errx(1, "did not find correct gzip header");
+	while ((ret = gzread(f, img + len, 65536)) > 0)
+		len += ret;
+	if (ret < 0)
+		err(1, "reading image from bzImage");
+
+	verbose("Unpacked size %i addr %p\n", len, img);
+	*page_offset = intuit_page_offset(img, len);
+	*pgdir_addr = finish(mem, page_offset, initrd, ird_size);
+
+	/* Entry is physical address: convert to virtual */
+	return (u32)img + *page_offset;
+}
+
+static u32 load_bzimage(int bzimage_fd, const Elf32_Ehdr *ehdr, 
+			unsigned long mem, unsigned long *pgdir_addr,
+			const char *initrd, unsigned long *ird_size,
+			unsigned long *page_offset)
+{
+	unsigned char c;
+	int state = 0;
+
+	/* Just brute force it. */
+	while (read(bzimage_fd, &c, 1) == 1) {
+		switch (state) {
+		case 0:
+			if (c == 0x1F)
+				state++;
+			break;
+		case 1:
+			if (c == 0x8B)
+				state++;
+			else
+				state = 0;
+			break;
+		case 2 ... 8:
+			state++;
+			break;
+		case 9:
+			lseek(bzimage_fd, -10, SEEK_CUR);
+			if (c != 0x03) /* Compressed under UNIX. */
+				state = -1;
+			else
+				return bzimage(bzimage_fd, mem, pgdir_addr,
+					       initrd, ird_size, page_offset);
+		}
+	}
+	errx(1, "Could not find kernel in bzImage");
+}
+
+static void *map_pages(unsigned long addr, unsigned int num)
+{
+	if (mmap((void *)addr, getpagesize() * num,
+		 PROT_READ|PROT_WRITE|PROT_EXEC,
+		 MAP_FIXED|MAP_PRIVATE, zero_fd, 0) != (void *)addr)
+		err(1, "Mmaping %u pages of /dev/zero @%p", num, (void *)addr);
+	return (void *)addr;
+}
+
+static struct lguest_device_desc *
+get_dev_entry(struct lguest_device_desc *descs, u16 type, u16 num_pages)
+{
+	static unsigned long top = LGUEST_GUEST_TOP;
+	int i;
+	unsigned long pfn = 0;
+
+	if (num_pages) {
+		top -= num_pages*getpagesize();
+		map_pages(top, num_pages);
+		pfn = top / getpagesize();
+	}
+
+	for (i = 0; i < LGUEST_MAX_DEVICES; i++) {
+		if (!descs[i].type) {
+			descs[i].features = descs[i].status = 0;
+			descs[i].type = type;
+			descs[i].num_pages = num_pages;
+			descs[i].pfn = pfn;
+			return &descs[i];
+		}
+	}
+	errx(1, "too many devices");
+}
+
+static void set_fd(int fd, struct devices *devices)
+{
+	FD_SET(fd, &devices->infds);
+	if (fd > devices->max_infd)
+		devices->max_infd = fd;
+}
+
+static struct device *new_device(struct devices *devices,
+				 struct lguest_device_desc *descs,
+				 u16 type, u16 num_pages,
+				 int fd,
+				 int (*handle_input)(int, struct device *),
+				 unsigned long watch_off,
+				 u32 (*handle_output)(int,
+						      const struct iovec *,
+						      unsigned,
+						      struct device *))
+{
+	struct device *dev = malloc(sizeof(*dev));
+
+	dev->next = devices->dev;
+	devices->dev = dev;
+
+	dev->fd = fd;
+	if (handle_input)
+		set_fd(dev->fd, devices);
+	dev->desc = get_dev_entry(descs, type, num_pages);
+	dev->mem = (void *)(dev->desc->pfn * getpagesize());
+	dev->handle_input = handle_input;
+	dev->watch_address = (unsigned long)dev->mem + watch_off;
+	dev->handle_output = handle_output;
+	return dev;
+}
+
+static int tell_kernel(u32 pagelimit, u32 pgdir, u32 start, u32 page_offset)
+{
+	u32 args[] = { LHREQ_INITIALIZE,
+		       pagelimit, pgdir, start, page_offset };
+	int fd = open("/dev/lguest", O_RDWR);
+
+	if (fd < 0)
+		err(1, "Opening /dev/lguest");
+
+	verbose("Telling kernel limit %u, pgdir %i, e=%#08x page_off=0x%08x\n",
+		pagelimit, pgdir, start, page_offset);
+	if (write(fd, args, sizeof(args)) < 0)
+		err(1, "Writing to /dev/lguest");
+	return fd;
+}
+
+static void concat(char *dst, char *args[])
+{
+	unsigned int i, len = 0;
+
+	for (i = 0; args[i]; i++) {
+		strcpy(dst+len, args[i]);
+		strcat(dst+len, " ");
+		len += strlen(args[i]) + 1;
+	}
+	/* In case it's empty. */
+	dst[len] = '\0';
+}
+
+static void *_check_pointer(unsigned long addr, unsigned int size,
+			    unsigned int line)
+{
+	if (addr >= LGUEST_GUEST_TOP || addr + size >= LGUEST_GUEST_TOP)
+		errx(1, "%s:%i: Invalid address %li", __FILE__, line, addr);
+	return (void *)addr;
+}
+#define check_pointer(addr,size) _check_pointer(addr, size, __LINE__)
+
+/* Returns pointer to dma->used_len */
+static u32 *dma2iov(unsigned long dma, struct iovec iov[], unsigned *num)
+{
+	unsigned int i;
+	struct lguest_dma *udma;
+
+	/* No buffers? */
+	if (dma == 0) {
+		printf("no buffers\n");
+		return NULL;
+	}
+
+	udma = check_pointer(dma, sizeof(*udma));
+	for (i = 0; i < LGUEST_MAX_DMA_SECTIONS; i++) {
+		if (!udma->len[i])
+			break;
+
+		iov[i].iov_base = check_pointer(udma->addr[i], udma->len[i]);
+		iov[i].iov_len = udma->len[i];
+	}
+	*num = i;
+	return &udma->used_len;
+}
+
+static u32 *get_dma_buffer(int fd, void *addr,
+			   struct iovec iov[], unsigned *num, u32 *irq)
+{
+	u32 buf[] = { LHREQ_GETDMA, (u32)addr };
+	unsigned long udma;
+	u32 *res;
+
+	udma = write(fd, buf, sizeof(buf));
+	if (udma == (unsigned long)-1)
+		return NULL;
+
+	/* Kernel stashes irq in ->used_len. */
+	res = dma2iov(udma, iov, num);
+	if (res)
+		*irq = *res;
+	return res;
+}
+
+static void trigger_irq(int fd, u32 irq)
+{
+	u32 buf[] = { LHREQ_IRQ, irq };
+	if (write(fd, buf, sizeof(buf)) != 0)
+		err(1, "Triggering irq %i", irq);
+}
+
+static struct termios orig_term;
+static void restore_term(void)
+{
+	tcsetattr(STDIN_FILENO, TCSANOW, &orig_term);
+}
+
+struct console_abort
+{
+	int count;
+	struct timeval start;
+};
+
+/* We DMA input to buffer bound at start of console page. */
+static int handle_console_input(int fd, struct device *dev)
+{
+	u32 num, irq = 0, *lenp;
+	int len;
+	struct iovec iov[LGUEST_MAX_DMA_SECTIONS];
+	struct console_abort *abort = dev->priv;
+
+	lenp = get_dma_buffer(fd, dev->mem, iov, &num, &irq);
+	if (!lenp) {
+		warn("console: no dma buffer!");
+		iov[0] = discard_iov;
+		num = 1;
+	}
+
+	len = readv(dev->fd, iov, num);
+	if (len <= 0) {
+		warnx("Failed to get console input, ignoring console.");
+		len = 0;
+	}
+
+	if (lenp) {
+		*lenp = len;
+		trigger_irq(fd, irq);
+	}
+
+	/* Three ^C within one second?  Exit. */
+	if (len == 1 && ((char *)iov[0].iov_base)[0] == 3) {
+		if (!abort->count++)
+			gettimeofday(&abort->start, NULL);
+		else if (abort->count == 3) {
+			struct timeval now;
+			gettimeofday(&now, NULL);
+			if (now.tv_sec <= abort->start.tv_sec+1)
+				exit(2);
+			abort->count = 0;
+		}
+	} else
+		abort->count = 0;
+
+	if (!len) {
+		restore_term();
+		return 0;
+	}
+	return 1;
+}
+
+static unsigned long peer_offset(unsigned int peernum)
+{
+	return 4 * peernum;
+}
+
+static u32 handle_tun_output(int fd, const struct iovec *iov,
+			     unsigned num, struct device *dev)
+{
+	/* Now we've seen output, we should warn if we can't get buffers. */
+	*(bool *)dev->priv = true;
+	return writev(dev->fd, iov, num);
+}
+
+static u32 handle_block_output(int fd, const struct iovec *iov,
+			       unsigned num, struct device *dev)
+{
+	struct lguest_block_page *p = dev->mem;
+	u32 irq, reply_num, *lenp;
+	int len;
+	struct iovec reply[LGUEST_MAX_DMA_SECTIONS];
+	off64_t device_len, off = (off64_t)p->sector * 512;
+
+	device_len = *(off64_t *)dev->priv;
+
+	if (off >= device_len)
+		err(1, "Bad offset %llu vs %llu", off, device_len);
+	if (lseek64(dev->fd, off, SEEK_SET) != off)
+		err(1, "Bad seek to sector %i", p->sector);
+
+	verbose("Block: %s at offset %llu\n", p->type ? "WRITE" : "READ", off);
+
+	lenp = get_dma_buffer(fd, dev->mem, reply, &reply_num, &irq);
+	if (!lenp)
+		err(1, "Block request didn't give us a dma buffer");
+
+	if (p->type) {
+		len = writev(dev->fd, iov, num);
+		if (off + len > device_len) {
+			ftruncate(dev->fd, device_len);
+			errx(1, "Write past end %llu+%u", off, len);
+		}
+		*lenp = 0;
+	} else {
+		len = readv(dev->fd, reply, reply_num);
+		*lenp = len;
+	}
+
+	p->result = 1 + (p->bytes != len);
+	trigger_irq(fd, irq);
+	return 0;
+}
+
+#define HIPQUAD(ip)				\
+	((u8)(ip >> 24)),			\
+	((u8)(ip >> 16)),			\
+	((u8)(ip >> 8)),			\
+	((u8)(ip))
+
+static void configure_device(const char *devname, u32 ipaddr,
+			     unsigned char hwaddr[6])
+{
+	struct ifreq ifr;
+	int fd;
+	struct sockaddr_in *sin = (struct sockaddr_in *)&ifr.ifr_addr;
+
+	memset(&ifr, 0, sizeof(ifr));
+	strcpy(ifr.ifr_name, devname);
+	sin->sin_family = AF_INET;
+	sin->sin_addr.s_addr = htonl(ipaddr);
+	fd = socket(PF_INET, SOCK_DGRAM, IPPROTO_IP);
+	if (fd < 0)
+		err(1, "opening IP socket");
+	if (ioctl(fd, SIOCSIFADDR, &ifr) != 0)
+		err(1, "Setting %s interface address", devname);
+	ifr.ifr_flags = IFF_UP;
+	if (ioctl(fd, SIOCSIFFLAGS, &ifr) != 0)
+		err(1, "Bringing interface %s up", devname);
+
+	if (ioctl(fd, SIOCGIFHWADDR, &ifr) != 0)
+		err(1, "getting hw address for %s", devname);
+
+	memcpy(hwaddr, ifr.ifr_hwaddr.sa_data, 6);
+}
+
+/* We send lguest_add signals while input is pending: avoids races. */
+static void wake_parent(int pipefd, struct devices *devices)
+{
+	int parent = getppid();
+	nice(19);
+
+	set_fd(pipefd, devices);
+
+	for (;;) {
+		fd_set rfds = devices->infds;
+
+		select(devices->max_infd+1, &rfds, NULL, NULL, NULL);
+		if (FD_ISSET(pipefd, &rfds)) {
+			int ignorefd;
+			if (read(pipefd, &ignorefd, sizeof(ignorefd)) == 0)
+				exit(0);
+			FD_CLR(ignorefd, &devices->infds);
+		}
+		kill(parent, SIGUSR1);
+	}
+}
+
+/* We don't want signal to kill us, just jerk us out of kernel. */
+static void wakeup(int signo)
+{
+}
+
+static int handle_tun_input(int fd, struct device *dev)
+{
+	u32 irq = 0, num, *lenp;
+	int len;
+	struct iovec iov[LGUEST_MAX_DMA_SECTIONS];
+
+	lenp = get_dma_buffer(fd, dev->mem+peer_offset(NET_PEERNUM), iov, &num,
+			      &irq);
+	if (!lenp) {
+		if (*(bool *)dev->priv)
+			warn("network: no dma buffer!");
+		iov[0] = discard_iov;
+		num = 1;
+	}
+
+	len = readv(dev->fd, iov, num);
+	if (len <= 0)
+		err(1, "reading network");
+	if (lenp) {
+		*lenp = len;
+		trigger_irq(fd, irq);
+	}
+	verbose("tun input packet len %i [%02x %02x] (%s)\n", len,
+		((u8 *)iov[0].iov_base)[0], ((u8 *)iov[0].iov_base)[1],
+		lenp ? "sent" : "discarded");
+	return 1;
+}
+
+/* We use fnctl locks to reserve network slots (autocleanup!) */
+static unsigned int find_slot(int netfd, const char *filename)
+{
+	struct flock fl;
+
+	fl.l_type = F_WRLCK;
+	fl.l_whence = SEEK_SET;
+	fl.l_len = 1;
+	for (fl.l_start = 0;
+	     fl.l_start < getpagesize()/sizeof(struct lguest_net);
+	     fl.l_start++) {
+		if (fcntl(netfd, F_SETLK, &fl) == 0)
+			return fl.l_start;
+	}
+	errx(1, "No free slots in network file %s", filename);
+}
+
+static void setup_net_file(const char *filename,
+			   struct lguest_device_desc *descs,
+			   struct devices *devices)
+{
+	int netfd;
+	struct device *dev;
+
+	netfd = open(filename, O_RDWR, 0);
+	if (netfd < 0) {
+		if (errno == ENOENT) {
+			netfd = open(filename, O_RDWR|O_CREAT, 0600);
+			if (netfd >= 0) {
+				char page[getpagesize()];
+				/* 0xFFFF == NO_GUEST */
+				memset(page, 0xFF, sizeof(page));
+				write(netfd, page, sizeof(page));
+			}
+		}
+		if (netfd < 0)
+			err(1, "cannot open net file '%s'", filename);
+	}
+
+	dev = new_device(devices, descs, LGUEST_DEVICE_T_NET, 1,
+			 -1, NULL, 0, NULL);
+
+	/* This is the slot for the guest to use. */
+	dev->desc->features = find_slot(netfd, filename)|LGUEST_NET_F_NOCSUM;
+	/* We overwrite the /dev/zero mapping with the actual file. */
+	if (mmap(dev->mem, getpagesize(), PROT_READ|PROT_WRITE,
+			 MAP_FIXED|MAP_SHARED, netfd, 0) != dev->mem)
+			err(1, "could not mmap '%s'", filename);
+	verbose("device %p@%p: shared net %s, peer %i\n", dev->desc, 
+		(void *)(dev->desc->pfn * getpagesize()), filename, 
+		dev->desc->features & ~LGUEST_NET_F_NOCSUM);
+}
+
+static u32 str2ip(const char *ipaddr)
+{
+	unsigned int byte[4];
+
+	sscanf(ipaddr, "%u.%u.%u.%u", &byte[0], &byte[1], &byte[2], &byte[3]);
+	return (byte[0] << 24) | (byte[1] << 16) | (byte[2] << 8) | byte[3];
+}
+
+static void setup_tun_net(const char *ipaddr,
+			  struct lguest_device_desc *descs,
+			  struct devices *devices)
+{
+	struct device *dev;
+	struct ifreq ifr;
+	int netfd;
+
+	netfd = open("/dev/net/tun", O_RDWR);
+	if (netfd < 0)
+		err(1, "opening /dev/net/tun");
+
+	memset(&ifr, 0, sizeof(ifr));
+	ifr.ifr_flags = IFF_TAP | IFF_NO_PI;
+	strcpy(ifr.ifr_name, "tap%d");
+	if (ioctl(netfd, TUNSETIFF, &ifr) != 0)
+		err(1, "configuring /dev/net/tun");
+
+	dev = new_device(devices, descs, LGUEST_DEVICE_T_NET, 1,
+			 netfd, handle_tun_input,
+			 peer_offset(0), handle_tun_output);
+	dev->priv = malloc(sizeof(bool));
+	*(bool *)dev->priv = false;
+
+	/* We are peer 0, rest is all NO_GUEST */
+	memset(dev->mem, 0xFF, getpagesize());
+	configure_device(ifr.ifr_name, str2ip(ipaddr), dev->mem);
+
+	/* You will be peer 1: we should create enough jitter to randomize */
+	dev->desc->features = NET_PEERNUM|LGUEST_DEVICE_F_RANDOMNESS;
+	verbose("device %p@%p: tun net %u.%u.%u.%u\n", dev->desc, 
+		(void *)(dev->desc->pfn * getpagesize()),
+		HIPQUAD(str2ip(ipaddr)));
+}
+
+static void setup_block_file(const char *filename,
+			     struct lguest_device_desc *descs,
+			     struct devices *devices)
+{
+	int fd;
+	struct device *dev;
+	off64_t *blocksize;
+	struct lguest_block_page *p;
+
+	fd = open(filename, O_RDWR|O_LARGEFILE|O_DIRECT, 0);
+	if (fd < 0)
+		err(1, "Opening %s", filename);
+
+	dev = new_device(devices, descs, LGUEST_DEVICE_T_BLOCK, 1,
+			 fd, NULL, 0, handle_block_output);
+	dev->desc->features = LGUEST_DEVICE_F_RANDOMNESS;
+	blocksize = dev->priv = malloc(sizeof(*blocksize));
+	*blocksize = lseek64(fd, 0, SEEK_END);
+	p = dev->mem;
+
+	p->num_sectors = *blocksize/512;
+	verbose("device %p@%p: block %i sectors\n", dev->desc, 
+		(void *)(dev->desc->pfn * getpagesize()), p->num_sectors);
+}
+
+static u32 handle_console_output(int fd, const struct iovec *iov,
+				 unsigned num, struct device*dev)
+{
+	return writev(STDOUT_FILENO, iov, num);
+}
+
+static void setup_console(struct lguest_device_desc *descs,
+			  struct devices *devices)
+{
+	struct device *dev;
+
+	if (tcgetattr(STDIN_FILENO, &orig_term) == 0) {
+		struct termios term = orig_term;
+		term.c_lflag &= ~(ISIG|ICANON|ECHO);
+		tcsetattr(STDIN_FILENO, TCSANOW, &term);
+		atexit(restore_term);
+	}
+
+	/* We don't currently require a page for the console. */
+	dev = new_device(devices, descs, LGUEST_DEVICE_T_CONSOLE, 0,
+			 STDIN_FILENO, handle_console_input,
+			 4, handle_console_output);
+	dev->priv = malloc(sizeof(struct console_abort));
+	((struct console_abort *)dev->priv)->count = 0;
+	verbose("device %p@%p: console\n", dev->desc, 
+		(void *)(dev->desc->pfn * getpagesize()));
+}
+
+static const char *get_arg(const char *arg, const char *prefix)
+{
+	if (strncmp(arg, prefix, strlen(prefix)) == 0)
+		return arg + strlen(prefix);
+	return NULL;
+}
+
+static u32 handle_device(int fd, unsigned long dma, unsigned long addr,
+			 struct devices *devices)
+{
+	struct device *i;
+	u32 *lenp;
+	struct iovec iov[LGUEST_MAX_DMA_SECTIONS];
+	unsigned num = 0;
+
+	lenp = dma2iov(dma, iov, &num);
+	if (!lenp)
+		errx(1, "Bad SEND_DMA %li for address %#lx\n", dma, addr);
+
+	for (i = devices->dev; i; i = i->next) {
+		if (i->handle_output && addr == i->watch_address) {
+			*lenp = i->handle_output(fd, iov, num, i);
+			return 0;
+		}
+	}
+	warnx("Pending dma %p, addr %p", (void *)dma, (void *)addr);
+	return 0;
+}
+
+static void handle_input(int fd, int childfd, struct devices *devices)
+{
+	struct timeval poll = { .tv_sec = 0, .tv_usec = 0 };
+
+	for (;;) {
+		struct device *i;
+		fd_set fds = devices->infds;
+
+		if (select(devices->max_infd+1, &fds, NULL, NULL, &poll) == 0)
+			break;
+
+		for (i = devices->dev; i; i = i->next) {
+			if (i->handle_input && FD_ISSET(i->fd, &fds)) {
+				if (!i->handle_input(fd, i)) {
+					FD_CLR(i->fd, &devices->infds);
+					/* Tell child to ignore it too... */
+					write(childfd, &i->fd, sizeof(i->fd));
+				}
+			}
+		}
+	}
+}
+
+int main(int argc, char *argv[])
+{
+	unsigned long mem, pgdir, entry, initrd_size, page_offset;
+	int arg, kern_fd, fd, child, pipefd[2];
+	Elf32_Ehdr hdr;
+	struct sigaction act;
+	sigset_t sigset;
+	struct lguest_device_desc *devdescs;
+	struct devices devices;
+	struct lguest_boot_info *boot = (void *)0;
+	const char *initrd_name = NULL;
+	u32 (*load)(int, const Elf32_Ehdr *ehdr, unsigned long,
+		    unsigned long *, const char *, unsigned long *,
+		    unsigned long *);
+
+	if (argv[1] && strcmp(argv[1], "--verbose") == 0) {
+		verbose = true;
+		argv++;
+		argc--;
+	}
+
+	if (argc < 4)
+		errx(1, "Usage: lguest [--verbose] <mem> vmlinux "
+			"[--sharenet=<filename>|--tunnet=<ipaddr>|--block=<filename>"
+			"|--initrd=<filename>]... [args...]");
+
+	zero_fd = open("/dev/zero", O_RDONLY, 0);
+	if (zero_fd < 0)
+		err(1, "Opening /dev/zero");
+
+	mem = memparse(argv[1]);
+	kern_fd = open(argv[2], O_RDONLY, 0);
+	if (kern_fd < 0)
+		err(1, "Opening %s", argv[2]);
+
+	if (read(kern_fd, &hdr, sizeof(hdr)) != sizeof(hdr))
+		err(1, "Reading %s elf header", argv[2]);
+
+	if (memcmp(hdr.e_ident, ELFMAG, SELFMAG) == 0)
+		load = map_elf;
+	else
+		load = load_bzimage;
+
+	devices.max_infd = -1;
+	devices.dev = NULL;
+	FD_ZERO(&devices.infds);
+
+	devdescs = map_pages(mem, 1);
+	arg = 3;
+	while (argv[arg] && argv[arg][0] == '-') {
+		const char *argval;
+
+		if ((argval = get_arg(argv[arg], "--sharenet=")) != NULL)
+			setup_net_file(argval, devdescs, &devices);
+		else if ((argval = get_arg(argv[arg], "--tunnet=")) != NULL)
+			setup_tun_net(argval, devdescs, &devices);
+		else if ((argval = get_arg(argv[arg], "--block=")) != NULL)
+			setup_block_file(argval, devdescs, &devices);
+		else if ((argval = get_arg(argv[arg], "--initrd=")) != NULL)
+			initrd_name = argval;
+		else
+			errx(1, "unknown arg '%s'", argv[arg]);
+		arg++;
+	}
+
+	entry = load(kern_fd, &hdr, mem, &pgdir, initrd_name, &initrd_size,
+		     &page_offset);
+	setup_console(devdescs, &devices);
+
+	concat(boot->cmdline, argv+arg);
+	boot->max_pfn = mem/getpagesize();
+	boot->initrd_size = initrd_size;
+
+	act.sa_handler = wakeup;
+	sigemptyset(&act.sa_mask);
+	act.sa_flags = 0;
+	sigaction(SIGUSR1, &act, NULL);
+
+	pipe(pipefd);
+	child = fork();
+	if (child == -1)
+		err(1, "forking");
+
+	if (child == 0) {
+		close(pipefd[1]);
+		wake_parent(pipefd[0], &devices);
+	}
+	close(pipefd[0]);
+
+	sigemptyset(&sigset);
+	sigaddset(&sigset, SIGUSR1);
+	sigprocmask(SIG_BLOCK, &sigset, NULL);
+
+	/* LGUEST_GUEST_TOP defined in Makefile, just below us. */
+	fd = tell_kernel(LGUEST_GUEST_TOP/getpagesize(),
+			 pgdir, entry, page_offset);
+
+	for (;;) {
+		unsigned long arr[2];
+		int readval;
+
+		sigprocmask(SIG_UNBLOCK, &sigset, NULL);
+		readval = read(fd, arr, sizeof(arr));
+		sigprocmask(SIG_BLOCK, &sigset, NULL);
+
+		switch (readval) {
+		case sizeof(arr):
+			handle_device(fd, arr[0], arr[1], &devices);
+			break;
+		case -1:
+			if (errno == EINTR)
+				break;
+		default:
+			if (errno == ENOENT) {
+				char reason[1024];
+				if (read(fd, reason, sizeof(reason)) > 0)
+					errx(1, "%s", reason);
+			}
+			err(1, "Running guest failed");
+		}
+		handle_input(fd, pipefd[1], &devices);
+	}
+}
diff -r 8806a441a0b1 Documentation/lguest/lguest.txt
--- /dev/null	Thu Jan 01 00:00:00 1970 +0000
+++ b/Documentation/lguest/lguest.txt	Mon Feb 12 13:47:43 2007 +1100
@@ -0,0 +1,355 @@
+Rusty's Remarkably Unreliable Guide to Lguest
+	- or, A Young Coder's Illustrated Hypervisor
+http://lguest.ozlabs.org
+
+Lguest is designed to be a minimal hypervisor for the Linux kernel, for
+Linux developers and users to experiment with virtualization with the
+minimum of complexity.  Nonetheless, it should have sufficient
+features to make it useful for specific tasks, and, of course, you are
+encouraged to fork and enhance it.
+
+Features:
+
+- Kernel module which runs in a normal kernel.
+- Simple I/O model for communication.
+- Simple program to create new guests.
+- Logo contains cute puppies: http://lguest.ozlabs.org
+
+Developer features:
+
+- Fun to hack on.
+- No ABI: being tied to a specific kernel anyway, you can change anything.
+- Many opportunities for improvement or feature implementation.
+
+Running Lguest:
+
+- You will need to configure your kernel with the following options:
+
+  CONFIG_HIGHMEM64G=n ("High Memory Support" "64GB")[1]
+  CONFIG_TUN=y/m ("Universal TUN/TAP device driver support")
+  CONFIG_EXPERIMENTAL=y ("Prompt for development and/or incomplete code/drivers")
+  CONFIG_PARAVIRT=y ("Paravirtualization support (EXPERIMENTAL)")
+  CONFIG_LGUEST=y/m ("Linux hypervisor example code")
+
+  and I recommend:
+  CONFIG_HZ=100 ("Timer frequency")[2]
+
+  You must have a machine with a TSC: look for "tsc" in /proc/cpuinfo.
+  It's simple to remove this restriction, but everyone has a TSC these
+  days.
+
+- A tool called "lguest" is available in this directory: type "make"
+  to build it.
+
+- Create or find a root disk image.  There are several useful ones
+  around, such as the xm-test tiny root image at 
+	  http://xm-test.xensource.com/ramdisks/initrd-1.1-i386.img
+
+  For more serious work, I usually use a distribution ISO image and
+  install it under qemu, then make multiple copies:
+
+	  dd if=/dev/zero of=rootfile bs=1M count=2048
+	  qemu -cdrom image.iso -hda rootfile -net user -net nic -boot d
+
+- "modprobe lg" if you built it as a module.
+
+- Run an lguest as root:
+
+      Documentation/lguest/lguest 64m vmlinux --tunnet=192.168.19.1 --block=rootfile root=/dev/lgba
+
+   Explanation:
+    64m: the amount of memory to use.
+
+    vmlinux: the kernel image found in the top of your build directory.  You
+       can also use a standard bzImage.
+
+    --tunnet=192.168.19.1: configures a "tap" device for networking with this
+       IP address.
+
+    --block=rootfile: a file or block device which becomes /dev/lgba
+       inside the guest.
+
+    root=/dev/lgba: this (and anything else on the command line) are
+       kernel boot parameters.
+
+- Configuring networking.  I usually have the host masquerade, using
+  "iptables -t nat -o eth0 -j MASQUERADE" and "echo 1 >
+  /proc/sys/net/ipv4/ip_forward".  In this example, I would configure
+  eth0 inside the guest at 192.168.19.2.
+
+- You can also create an inter-guest network using
+  "--sharenet=<filename>": any two guests using the same file are on
+  the same network.  This file is created if it does not exist.
+
+
+Lguest I/O model:
+
+Lguest uses a simplified DMA model plus shared memory for I/O.  Guests
+can communicate with each other if they share underlying memory
+(usually by the lguest program mmaping the same file), but they can
+use any non-shared memory to communicate with the lguest process.
+
+Guests can register DMA buffers at any physical address using the
+LHCALL_BIND_DMA(physaddr, dmabufs, num<<8|irq) hypercall.  "dmabufs"
+is the physical address of an array of "num" "struct lguest_dma": each
+contains a used_len, and an array of physical addresses and lengths.
+When a transfer occurs, the "used_len" field of one of the buffers
+which has used_len 0 will be set to the length transferred and the irq
+will fire.
+
+Using an irq value of 0 unbinds the dma buffers.
+
+To send DMA, the LHCALL_SEND_DMA(physaddr, dma_physaddr) hypercall is
+used, and the bytes used is written to the used_len field.  This can
+be 0 if noone else has bound a DMA buffer to that address or some
+other error.  DMA buffers bound by the same guest are ignored.
+
+
+Hacking on Lguest:
+
+Lguest uses the paravirt_ops infrastructure to override various
+sensitive operations so Linux can run in ring level 1 (rather
+than 0).  These operations make "hypercalls": traps into a tiny shim
+which is mapped at the top of memory which then switches back to the
+host Linux for servicing.  In fact, any real interrupt and many
+traps cause a switch back to the host, which doesn't even notice that
+it was switched out.  This means that the guest process is scheduled
+like any other process, although it spends most of its time in its own
+special address space.
+
+Here are the parts of the hypervisor at the moment:
+
+hypervisor.S:
+	The assembler shim which is mapped at 0xFFC01000 (-4M+1page)
+	in the host and all the guests.  This is built into a .o file
+	and inserted in the source as a C array: it is simply copied
+	into the mapped memory.
+
+	The shim is entered from the host at switch_to_guest with
+	interrupts off: this saves state and switches page tables,
+	GDT, IDT, TSS and stack, then dives into the guest with an
+	iret.
+
+	There are two ways back to the host: a trap or an external
+	interrupt.  A trap, such as a page fault, goes through
+	return_to_host, which simply switches back and irets to the
+	caller (init.c's lcall), which decides what to do.  For an
+	interrupt we call deliver_to_host, which switches to the host
+	then jumps straight to the host interrupt routine: the
+	interrupt routine will do an "iret" at some stage, which, now
+	we've switched stacks, will return to the caller in init.c.
+
+page_tables.c:
+	We cannot let guests control their own pagetables, since they
+	must not access others' memory and their concept of physical
+	addresses is not related to the real physical addresses: the
+	guest "physical" addresses are in fact virtual addresses in
+	the host's lguest thread.  The process of mapping the two
+	can be fairly complicated.
+
+	We keep up to 4 cached page tables.  When a page is referred
+	to by these guest "shadow" pagetables, we keep a reference to
+	it to prevent the Linux kernel from thinking it is unused and
+	paging it out underneath us.
+	FIXME: it would be much better to have a callback in mm_struct.
+
+	The main work is done in page_in.  First we check the
+	top-level guest page table: if that entry is not present, then
+	it's a real guest fault and we reflect it to the guest.
+	Otherwise, we check the real top level, and allocate a new
+	pagetable page if necessary.  Then we check the next level of
+	the guest page table: if that isn't present, or this was a
+	write and the guest entry is read only, we reflect it to the
+	guest.  Otherwise, we check the guest entry, convert the page
+	number to the actual physical page number, then set it in our
+	page table.  At this point we also update the accessed and
+	dirty bits in the guest.
+
+	So a guest's top-level pagetable starts empty, and over time
+	we fault more pages in.  If the guest switches page tables, we
+	see if it's in out 4-entry cache: if not, we clear the
+	non-kernel section of one of them and use that.  (The kernel
+	page table entries will always be the same in all top levels).
+
+	We have to keep the stack pages for the guest kernel mapped at
+	all times, since we point some traps (particularly system
+	calls) directly into the guest.  If the stack were not mapped
+	we would get a double fault, which means we kill the guest.
+
+	Note that there are three page tables for each guest: the
+	Linux host ones which exist for lguest just like any other
+	process, the actual ones used when we switch to running the
+	guest, and the ones inside the guest which it thinks it's
+	using (and we copy to the actual ones after checking).
+
+hypercalls.c:
+	This is where the guest used int 0x1F to ask the hypervisor
+	for something.  The first hypercall is always
+	LHCALL_LGUEST_INIT, which tells us where the "struct
+	lguest_page" is.  We populate the lguest_page with useful
+	information, and it's also used to indicate virtual interrupts
+	and whether the guest expects interrupts to be disabled.
+
+	Most of these calls are fairly self-explanatory, or covered
+	elsewhere.  Note that LHCALL_CRASH allows a guest to get a
+	message out before any devices are enabled, which can be
+	useful for debugging.
+
+	do_async_hypercalls: a ringbuffer in the lguest page allows
+	the guest to queue hypercalls for later execution.  This is
+	useful for hypercall batching during context switch, and for
+	some bulk I/O.  The return value of the hypercall is
+	discarded, so it doesn't make sense to batch some hypercalls.
+	Note that we always do all these "async" calls before any
+	normal hypercall, which means that any hypercall acts as a
+	flush operation.  The only trick is that an async SEND_DMA
+	hypercall may need to be serviced by the host userspace; the
+	run_guest loop is constructed so that we continue servicing
+	hypercalls when we re-enter the loop after host userspace has
+	done the I/O operation.
+
+	setup_trampoline: this populates a stub for direct traps to
+	the guest.  Using a trampoline page (which sits just below the
+	hypervisor at -4M) ensures that the page is always mapped, and
+	also ensures that we reload the %gs register before entering the
+	kernel (see guest_load_tls).
+
+io.c:
+	lguest provides DMA-style transfer, and buffer registration.
+	The guest can dma send to a particular address, or register a
+	set of DMA buffers at a particular address.  This provides
+	inter-guest I/O (for shared addresses, such as a shared mmap)
+	or I/O out to the userspace process (lguest).
+
+	We currently use the futex infrastructure to see if a given
+	address is shared: if it is, we look for another guest which
+	has registered a DMA buffer at this address and copy the data,
+	then interrupt the recipient.  Otherwise, we notify the guest
+	userspace (which has access to all the guest memory) to handle
+	the transfer.
+
+	TODO: We could flip whole pages between guests at this point
+	if we wanted to, however it seems unlikely to be worthwhile.
+	More optimization could be gained by having servers for certain
+	devices within the host kernel itself, avoiding at
+	least two switches into the lguest binary and back.
+
+core.c:
+	This contains the core of lguest, "run_guest", which
+	continuously lcalls into the switch_to_guest routine until
+	something interesting happens.  In particular, we only return
+	to userspace (ie. "lguest") when a signal occurs or the guest
+	does a SEND_DMA destined for host userspace.
+
+	emulate_insn(): we don't paravirtualize io and out
+	instructions, so we trap and emulate them here.  This is only
+	used when the guest is booting and probing for PCI busses,
+	etc.
+
+	lguest_address_ok(): the guest kernel must not be able to
+	access the lguest binary, otherwise it could break out of
+	its virtualization, so all dereferences must use the
+	lhread_u32/lhwrite_u32/lhread/lhwrite routines which check
+	this.
+
+	reflect_trap(): when we decide that the guest should handle a
+	trap (a page fault, a general protection fault, an FPU fault
+	or a virtual interrupt), we manually push a trap frame onto
+	its stack as it expects it to be.  There are two kinds of
+	traps for x86: interrupt gates expect to have interrupts
+	disabled, and trap gates expect interrupts to be left alone.
+	The guest will restore interrupts in lguest_iret.
+
+	Of course, we don't actually let the guest disable interrupts,
+	just prevent us from delivering interupts to that guest (the
+	flag "irq_enabled" in the lguest_page).
+
+	kill_guest: this is used when an error occurs which can only
+	be caused by the guest kernel.  You can continue as normal
+	after this: the guest will exit when it returns to run_thread.
+
+	fixup_gdt_table: we protect the hypervisor shim from being
+	accessed using segments, so we have to trim segments the guest
+	uses to exclude the hypervisor.  The shim itself uses two
+	segments (only accessible to ring 0) which map the entire
+	memory range, and we use our own TSS entry.
+
+	guest_load_tls: glibc implements __thread using
+	thread-local-storage segments.  These segments start at a
+	different offset for each thread, and cover the entire 4GB
+	address space.  glibc then uses huge offsets into this segment
+	to wrap around and access variables below that offset.
+	Unfortunately, we cannot allow this in general, as this would
+	allow access to the hypervisor shim!  Fortunately, x86 page
+	table entries contain a "user" bit, which when cleared makes
+	pages inaccessible to ring level 3.  We clear this bit for the
+	pagetable entries mapping the hypervisor, so we can allow ring
+	3 (ie. userspace) access to 4G segments.  If the guest is in
+	ring 3, we setup the segment limits at the full 4G just before
+	calling into hypervisor.S.  It will reload %gs, then truncate
+	these TLS segments to a single page.  This ensures that any
+	reload of gs gets the truncated segments.  As the guest
+	userspace will also load %gs itself, we ignore the first
+	protection fault that occurs at any given address in userspace
+	(assuming it's caused by use of the truncated segment).  As
+	all traps reload gs explicitly (trampoline page) or implicitly
+	(reflect_trap), they all must reset the pointer to the
+	last-detected faulting instruction, as they will fault again.
+
+device.c:
+	This contains the host userspace interface code	(ie. /dev/lguest).
+
+	The read and write routines are where the userspace program
+	lguest starts and performs I/O to the guest.  The initial
+	write supplies the number of memory pages, the access limit
+	(which is used to ensure the guest doesn't overwrite the
+	lguest binary which sits above this address), the initial
+	guest pagetable top, and the address to jump into the guest
+	image.  Reading from the file causes the guest to run until a
+	signal or I/O is pending.
+
+lguest_bus.c:
+	A simple bus which sits in the lguest_page and indicates what
+	devices are available.  Using the interrupt model it would be
+	easy to make this dynamic.
+
+drivers/net/lguest_net.c:
+	A simple network device, which (invisible to the guest) can be
+	shared between several guests or simply talk to the lguest
+	process.  There is only one unusual element: the sender
+	needs to find the packet destination.
+
+	We manually scan the shared page for mac addresses to decide
+	where to send a packet.  We overload an unusable bit in that
+	mac address to indicate promiscuous mode (so the sender knows
+	to send a copy of all packets to that recipient).
+
+drivers/char/hvc_lguest.c:
+	A simple console.  It could use a shared page as a ringbuffer
+	and merely use the dma mechanism for notifications, but using
+	DMA directly is less code.
+
+	TODO: The console input can be flooded if it doesn't service
+	fast enough, and will lose characters.  If this is a problem,
+	switch to ringbuffer or use multiple DMA buffers and define an
+	ordering.
+
+drivers/block/lguest_blk.c:
+	A simple block device.  It's actually overkill for the current
+	use: talking to the userspace side is synchronous, but this allows
+	it to be served by something else in future.
+
+arch/i386/kernel/lguest.c:
+	The guest paravirt_ops implementation.  The only complexity is
+	in the implementation of lguest_iret: we need to restore the
+	interrupt state and return from the interrupt atomically.  To
+	this end, we tell the hypervisor that it is not to interrupt
+	us in those instructions between the restoration (usually
+	enabling) of interrupts and the actual "iret".
+
+Cheers!
+Rusty Russell rusty@rustcorp.com.au.
+
+[1] These are on various places on the TODO list, waiting for you to
+    get annoyed enough at the limitation to fix it.
+[2] Lguest is not yet tickless when idle.  See [1].

Re: [PATCH 7/8] lguest: trivial guest block driver

[Jens Axboe]

On Mon, Feb 12 2007, Rusty Russell wrote:
> +static irqreturn_t lgb_irq(int irq, void *_bd)
> +{
> +	struct blockdev *bd = _bd;
> +	unsigned long flags;
> +
> +	if (!bd->req) {
> +		pr_debug("No work!\n");
> +		return IRQ_NONE;
> +	}
> +
> +	if (!bd->lb_page->result) {
> +		pr_debug("No result!\n");
> +		return IRQ_NONE;
> +	}
> +
> +	spin_lock_irqsave(&bd->lock, flags);
> +	end_request(bd->req, bd->lb_page->result == 1);
> +	bd->req = NULL;
> +	bd->dma.used_len = 0;
> +	blk_start_queue(bd->disk->queue);
> +	spin_unlock_irqrestore(&bd->lock, flags);
> +	return IRQ_HANDLED;
> +}

You are using the old-style end request handling. So while I generally
discourage use of end_request(), you seem to have a bigger problem here:

> +static unsigned int req_to_dma(struct request *req, struct lguest_dma *dma)
> +{
> +	unsigned int i = 0, idx, len = 0;
> +	struct bio *bio;
> +
> +	rq_for_each_bio(bio, req) {
> +		struct bio_vec *bvec;
> +		bio_for_each_segment(bvec, bio, idx) {
> +			BUG_ON(i == LGUEST_MAX_DMA_SECTIONS);
> +			BUG_ON(!bvec->bv_len);
> +			dma->addr[i] = page_to_phys(bvec->bv_page)
> +				+ bvec->bv_offset;
> +			dma->len[i] = bvec->bv_len;
> +			len += bvec->bv_len;
> +			i++;
> +		}
> +	}
> +	if (i < LGUEST_MAX_DMA_SECTIONS)
> +		dma->len[i] = 0;
> +	return len;
> +}

Here you map the entire request (lets call that segment A..Z), but
end_request() only completes the first chunk of the request. So
elv_next_request() will retrieve the same request again, and you'll then
map B..Z and repeat that transfer. So unless I'm missing some other part
here (just read it over quickly), you are re-doing large parts of a
merged request several times.

So: don't use end_request(). Add some driver helper that does:

static void lgb_end_request(struct blockdev *bd)
{
        int uptodate = bd->lb_page->result == 1;
        struct request *rq = bd->req;

        end_that_request_first(rq, uptodate, req->hard_nr_sectors);
        add_disk_randomness(rq->rq_disk);
        blkdev_dequeue_request(rq);
        end_that_request_last(rq, uptodate);
}

We could probably even make that a block layer helper, I'm sure others
could be cleaned up with that as well. You want to use that helper in
do_lgb_request() as well.

-- 
Jens Axboe

Re: [PATCH 7/8] lguest: trivial guest block driver

[Rusty Russell]

On Mon, 2007-02-12 at 05:43 +0100, Jens Axboe wrote:
> On Mon, Feb 12 2007, Rusty Russell wrote:
> > +	end_request(bd->req, bd->lb_page->result == 1);
>
> You are using the old-style end request handling. So while I generally
> discourage use of end_request(), you seem to have a bigger problem here:

> > +	rq_for_each_bio(bio, req) {
> > +		struct bio_vec *bvec;
> > +		bio_for_each_segment(bvec, bio, idx) {
> > +			BUG_ON(i == LGUEST_MAX_DMA_SECTIONS);
> > +			BUG_ON(!bvec->bv_len);
> > +			dma->addr[i] = page_to_phys(bvec->bv_page)
> > +				+ bvec->bv_offset;
> > +			dma->len[i] = bvec->bv_len;
> > +			len += bvec->bv_len;
> > +			i++;
> > +		}
> > +	}
> > +	if (i < LGUEST_MAX_DMA_SECTIONS)
> > +		dma->len[i] = 0;
> > +	return len;
> > +}
> 
> Here you map the entire request (lets call that segment A..Z), but
> end_request() only completes the first chunk of the request. So
> elv_next_request() will retrieve the same request again, and you'll then
> map B..Z and repeat that transfer. So unless I'm missing some other part
> here (just read it over quickly), you are re-doing large parts of a
> merged request several times.
> 
> So: don't use end_request(). Add some driver helper that does:
> 
> static void lgb_end_request(struct blockdev *bd)
> {
>         int uptodate = bd->lb_page->result == 1;
>         struct request *rq = bd->req;
> 
>         end_that_request_first(rq, uptodate, req->hard_nr_sectors);
>         add_disk_randomness(rq->rq_disk);
>         blkdev_dequeue_request(rq);
>         end_that_request_last(rq, uptodate);
> }
> 
> We could probably even make that a block layer helper, I'm sure others
> could be cleaned up with that as well. You want to use that helper in
> do_lgb_request() as well.

I'm confused.  That code looks like end_request:

void end_request(struct request *req, int uptodate)
{
	if (!end_that_request_first(req, uptodate, req->hard_cur_sectors)) {
		add_disk_randomness(req->rq_disk);
		blkdev_dequeue_request(req);
		end_that_request_last(req, uptodate);
	}
}

Rusty.

Re: [PATCH 7/8] lguest: trivial guest block driver

[Jens Axboe]

On Mon, Feb 12 2007, Rusty Russell wrote:
> On Mon, 2007-02-12 at 05:43 +0100, Jens Axboe wrote:
> > On Mon, Feb 12 2007, Rusty Russell wrote:
> > > +	end_request(bd->req, bd->lb_page->result == 1);
> >
> > You are using the old-style end request handling. So while I generally
> > discourage use of end_request(), you seem to have a bigger problem here:
> 
> > > +	rq_for_each_bio(bio, req) {
> > > +		struct bio_vec *bvec;
> > > +		bio_for_each_segment(bvec, bio, idx) {
> > > +			BUG_ON(i == LGUEST_MAX_DMA_SECTIONS);
> > > +			BUG_ON(!bvec->bv_len);
> > > +			dma->addr[i] = page_to_phys(bvec->bv_page)
> > > +				+ bvec->bv_offset;
> > > +			dma->len[i] = bvec->bv_len;
> > > +			len += bvec->bv_len;
> > > +			i++;
> > > +		}
> > > +	}
> > > +	if (i < LGUEST_MAX_DMA_SECTIONS)
> > > +		dma->len[i] = 0;
> > > +	return len;
> > > +}
> > 
> > Here you map the entire request (lets call that segment A..Z), but
> > end_request() only completes the first chunk of the request. So
> > elv_next_request() will retrieve the same request again, and you'll then
> > map B..Z and repeat that transfer. So unless I'm missing some other part
> > here (just read it over quickly), you are re-doing large parts of a
> > merged request several times.
> > 
> > So: don't use end_request(). Add some driver helper that does:
> > 
> > static void lgb_end_request(struct blockdev *bd)
> > {
> >         int uptodate = bd->lb_page->result == 1;
> >         struct request *rq = bd->req;
> > 
> >         end_that_request_first(rq, uptodate, req->hard_nr_sectors);
> >         add_disk_randomness(rq->rq_disk);
> >         blkdev_dequeue_request(rq);
> >         end_that_request_last(rq, uptodate);
> > }
> > 
> > We could probably even make that a block layer helper, I'm sure others
> > could be cleaned up with that as well. You want to use that helper in
> > do_lgb_request() as well.
> 
> I'm confused.  That code looks like end_request:
> 
> void end_request(struct request *req, int uptodate)
> {
> 	if (!end_that_request_first(req, uptodate, req->hard_cur_sectors)) {
> 		add_disk_randomness(req->rq_disk);
> 		blkdev_dequeue_request(req);
> 		end_that_request_last(req, uptodate);
> 	}
> }

Note hard_cur_sectors vs hard_nr_sectors. The former refers to the first
segment sector count, the latter to the total sector count in the
request. Hence the difference!

-- 
Jens Axboe

Re: [PATCH 7/8] lguest: trivial guest block driver

[Jens Axboe]

On Mon, Feb 12 2007, Jens Axboe wrote:
> On Mon, Feb 12 2007, Rusty Russell wrote:
> > On Mon, 2007-02-12 at 05:43 +0100, Jens Axboe wrote:
> > > On Mon, Feb 12 2007, Rusty Russell wrote:
> > > > +	end_request(bd->req, bd->lb_page->result == 1);
> > >
> > > You are using the old-style end request handling. So while I generally
> > > discourage use of end_request(), you seem to have a bigger problem here:
> > 
> > > > +	rq_for_each_bio(bio, req) {
> > > > +		struct bio_vec *bvec;
> > > > +		bio_for_each_segment(bvec, bio, idx) {
> > > > +			BUG_ON(i == LGUEST_MAX_DMA_SECTIONS);
> > > > +			BUG_ON(!bvec->bv_len);
> > > > +			dma->addr[i] = page_to_phys(bvec->bv_page)
> > > > +				+ bvec->bv_offset;
> > > > +			dma->len[i] = bvec->bv_len;
> > > > +			len += bvec->bv_len;
> > > > +			i++;
> > > > +		}
> > > > +	}
> > > > +	if (i < LGUEST_MAX_DMA_SECTIONS)
> > > > +		dma->len[i] = 0;
> > > > +	return len;
> > > > +}
> > > 
> > > Here you map the entire request (lets call that segment A..Z), but
> > > end_request() only completes the first chunk of the request. So
> > > elv_next_request() will retrieve the same request again, and you'll then
> > > map B..Z and repeat that transfer. So unless I'm missing some other part
> > > here (just read it over quickly), you are re-doing large parts of a
> > > merged request several times.
> > > 
> > > So: don't use end_request(). Add some driver helper that does:
> > > 
> > > static void lgb_end_request(struct blockdev *bd)
> > > {
> > >         int uptodate = bd->lb_page->result == 1;
> > >         struct request *rq = bd->req;
> > > 
> > >         end_that_request_first(rq, uptodate, req->hard_nr_sectors);
> > >         add_disk_randomness(rq->rq_disk);
> > >         blkdev_dequeue_request(rq);
> > >         end_that_request_last(rq, uptodate);
> > > }
> > > 
> > > We could probably even make that a block layer helper, I'm sure others
> > > could be cleaned up with that as well. You want to use that helper in
> > > do_lgb_request() as well.
> > 
> > I'm confused.  That code looks like end_request:
> > 
> > void end_request(struct request *req, int uptodate)
> > {
> > 	if (!end_that_request_first(req, uptodate, req->hard_cur_sectors)) {
> > 		add_disk_randomness(req->rq_disk);
> > 		blkdev_dequeue_request(req);
> > 		end_that_request_last(req, uptodate);
> > 	}
> > }
> 
> Note hard_cur_sectors vs hard_nr_sectors. The former refers to the first
> segment sector count, the latter to the total sector count in the
> request. Hence the difference!

Also, that is why my example code doesn't check the
end_that_request_first() return value, it must be 0 or it would be
buggy. Alternatively,

        if (end_that_request_first(rq, uptodate, rq->hard_nr_sectors))
                BUG()

would make that more clear. Or just a nice little comment. Anyway, if
you fixup lguest I'll add a kernel helper that we can switch it to once
merged.

-- 
Jens Axboe

Re: [PATCH 7/8] lguest: trivial guest block driver

[Rusty Russell]

On Mon, 2007-02-12 at 06:32 +0100, Jens Axboe wrote:
> On Mon, Feb 12 2007, Rusty Russell wrote:
> > On Mon, 2007-02-12 at 05:43 +0100, Jens Axboe wrote:
> > > Here you map the entire request (lets call that segment A..Z), but
> > > end_request() only completes the first chunk of the request. So
> > > elv_next_request() will retrieve the same request again, and you'll then
> > > map B..Z and repeat that transfer. So unless I'm missing some other part
> > > here (just read it over quickly), you are re-doing large parts of a
> > > merged request several times.

virtbench before:
	Time to read from disk (256 kB): 18654562 nsec
After:
	Time to read from disk (256 kB): 8018468 nsec

Thanks Jens!!
Rusty.
PS.  One day I'll buy you a beer and you can explain your nomenclature
theory for the block subsystem 8)

Name: lguest: trivial guest block driver

A simple block driver for lguest (/dev/lgbX).  Only does one request
at once.

Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>

diff -r a155959c419f drivers/block/Makefile
--- a/drivers/block/Makefile	Mon Feb 12 14:26:47 2007 +1100
+++ b/drivers/block/Makefile	Mon Feb 12 14:26:47 2007 +1100
@@ -28,4 +28,5 @@ obj-$(CONFIG_VIODASD)		+= viodasd.o
 obj-$(CONFIG_VIODASD)		+= viodasd.o
 obj-$(CONFIG_BLK_DEV_SX8)	+= sx8.o
 obj-$(CONFIG_BLK_DEV_UB)	+= ub.o
+obj-$(CONFIG_LGUEST_GUEST)	+= lguest_blk.o
 
diff -r a155959c419f drivers/block/lguest_blk.c
--- /dev/null	Thu Jan 01 00:00:00 1970 +0000
+++ b/drivers/block/lguest_blk.c	Mon Feb 12 18:07:05 2007 +1100
@@ -0,0 +1,270 @@
+/* A simple block driver for lguest.
+ *
+ * Copyright 2006 Rusty Russell <rusty@rustcorp.com.au> IBM Corporation
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307  USA
+ */
+//#define DEBUG
+#include <linux/init.h>
+#include <linux/types.h>
+#include <linux/blkdev.h>
+#include <linux/interrupt.h>
+#include <asm/lguest_device.h>
+
+static char next_block_index = 'a';
+
+struct blockdev
+{
+	spinlock_t lock;
+
+	/* The disk structure for the kernel. */
+	struct gendisk *disk;
+
+	/* The major number for this disk. */
+	int major;
+	int irq;
+
+	unsigned long phys_addr;
+	/* The ioremap'ed block page. */
+	struct lguest_block_page *lb_page;
+
+	/* We only have a single request outstanding at a time. */
+	struct lguest_dma dma;
+	struct request *req;
+};
+
+/* Jens gave me this nice helper to end all chunks of a request. */
+static void end_entire_request(struct request *req, int uptodate)
+{
+	if (end_that_request_first(req, uptodate, req->hard_nr_sectors))
+		BUG();
+	add_disk_randomness(req->rq_disk);
+	blkdev_dequeue_request(req);
+	end_that_request_last(req, uptodate);
+}
+
+static irqreturn_t lgb_irq(int irq, void *_bd)
+{
+	struct blockdev *bd = _bd;
+	unsigned long flags;
+
+	if (!bd->req) {
+		pr_debug("No work!\n");
+		return IRQ_NONE;
+	}
+
+	if (!bd->lb_page->result) {
+		pr_debug("No result!\n");
+		return IRQ_NONE;
+	}
+
+	spin_lock_irqsave(&bd->lock, flags);
+	end_entire_request(bd->req, bd->lb_page->result == 1);
+	bd->req = NULL;
+	bd->dma.used_len = 0;
+	blk_start_queue(bd->disk->queue);
+	spin_unlock_irqrestore(&bd->lock, flags);
+	return IRQ_HANDLED;
+}
+
+static unsigned int req_to_dma(struct request *req, struct lguest_dma *dma)
+{
+	unsigned int i = 0, idx, len = 0;
+	struct bio *bio;
+
+	rq_for_each_bio(bio, req) {
+		struct bio_vec *bvec;
+		bio_for_each_segment(bvec, bio, idx) {
+			BUG_ON(i == LGUEST_MAX_DMA_SECTIONS);
+			BUG_ON(!bvec->bv_len);
+			dma->addr[i] = page_to_phys(bvec->bv_page)
+				+ bvec->bv_offset;
+			dma->len[i] = bvec->bv_len;
+			len += bvec->bv_len;
+			i++;
+		}
+	}
+	if (i < LGUEST_MAX_DMA_SECTIONS)
+		dma->len[i] = 0;
+	return len;
+}
+
+static void empty_dma(struct lguest_dma *dma)
+{
+	dma->len[0] = 0;
+}
+
+static void setup_req(struct blockdev *bd,
+		      int type, struct request *req, struct lguest_dma *dma)
+{
+	bd->lb_page->type = type;
+	bd->lb_page->sector = req->sector;
+	bd->lb_page->result = 0;
+	bd->req = req;
+	bd->lb_page->bytes = req_to_dma(req, dma);
+}
+
+static int do_write(struct blockdev *bd, struct request *req)
+{
+	struct lguest_dma send;
+
+	pr_debug("lgb: WRITE sector %li\n", (long)req->sector);
+	setup_req(bd, 1, req, &send);
+
+	hcall(LHCALL_SEND_DMA, bd->phys_addr, __pa(&send), 0);
+	return 1;
+}
+
+static int do_read(struct blockdev *bd, struct request *req)
+{
+	struct lguest_dma ping;
+
+	pr_debug("lgb: READ sector %li\n", (long)req->sector);
+	setup_req(bd, 0, req, &bd->dma);
+
+	empty_dma(&ping);
+	hcall(LHCALL_SEND_DMA,bd->phys_addr,__pa(&ping),0);
+	return 1;
+}
+
+static void do_lgb_request(request_queue_t *q)
+{
+	struct blockdev *bd;
+	struct request *req;
+	int ok;
+
+again:
+	req = elv_next_request(q);
+	if (!req)
+		return;
+
+	bd = req->rq_disk->private_data;
+	/* Sometimes we get repeated requests after blk_stop_queue. */
+	if (bd->req)
+		return;
+
+	if (!blk_fs_request(req)) {
+		pr_debug("Got non-command 0x%08x\n", req->cmd_type);
+	error:
+		req->errors++;
+		end_entire_request(req, 0);
+		goto again;
+	} else {
+		if (rq_data_dir(req) == WRITE)
+			ok = do_write(req->rq_disk->private_data, req);
+		else
+			ok = do_read(req->rq_disk->private_data, req);
+
+		if (!ok)
+			goto error;
+		/* Wait for interrupt to tell us it's done. */
+		blk_stop_queue(q);
+	}
+}
+
+static struct block_device_operations lguestblk_fops = {
+	.owner = THIS_MODULE,
+};
+
+static int lguestblk_probe(struct lguest_device *lhdev)
+{
+	struct blockdev *bd;
+	int err;
+	int irqflags = IRQF_SHARED;
+
+	bd = kmalloc(sizeof(*bd), GFP_KERNEL);
+	if (!bd)
+		return -ENOMEM;
+
+	spin_lock_init(&bd->lock);
+	bd->phys_addr = (lguest_devices[lhdev->index].pfn << PAGE_SHIFT);
+
+	bd->disk = alloc_disk(1);
+	if (!bd->disk) {
+		err = -ENOMEM;
+		goto out_free_bd;
+	}
+
+	bd->disk->queue = blk_init_queue(do_lgb_request, &bd->lock);
+	if (!bd->disk->queue) {
+		err = -ENOMEM;
+		goto out_put;
+	}
+
+	/* We can only handle a certain number of sg entries */
+	blk_queue_max_hw_segments(bd->disk->queue, LGUEST_MAX_DMA_SECTIONS);
+	/* Buffers must not cross page boundaries */
+	blk_queue_segment_boundary(bd->disk->queue, PAGE_SIZE-1);
+
+	bd->irq = lhdev->index+1;
+	bd->major = register_blkdev(0, "lguestblk");
+	if (bd->major < 0) {
+		err = bd->major;
+		goto out_cleanup_queue;
+	}
+	bd->lb_page = (void *)ioremap(bd->phys_addr, PAGE_SIZE);
+	bd->req = NULL;
+
+	sprintf(bd->disk->disk_name, "lgb%c", next_block_index++);
+	if (lguest_devices[lhdev->index].features & LGUEST_DEVICE_F_RANDOMNESS)
+		irqflags |= IRQF_SAMPLE_RANDOM;
+	err = request_irq(bd->irq, lgb_irq, irqflags, bd->disk->disk_name, bd);
+	if (err)
+		goto out_unmap;
+
+	bd->dma.used_len = 0;
+	bd->dma.len[0] = 0;
+	hcall(LHCALL_BIND_DMA, bd->phys_addr, __pa(&bd->dma), (1<<8)+bd->irq);
+
+	printk(KERN_INFO "%s: device %i at major %d\n",
+	       bd->disk->disk_name, lhdev->index, bd->major);
+
+	bd->disk->major = bd->major;
+	bd->disk->first_minor = 0;
+	bd->disk->private_data = bd;
+	bd->disk->fops = &lguestblk_fops;
+	/* This is initialized to the disk size by the other end. */
+	set_capacity(bd->disk, bd->lb_page->num_sectors);
+	add_disk(bd->disk);
+
+	lhdev->private = bd;
+	return 0;
+
+out_unmap:
+	iounmap(bd->lb_page);
+out_cleanup_queue:
+	blk_cleanup_queue(bd->disk->queue);
+out_put:
+	put_disk(bd->disk);
+out_free_bd:
+	kfree(bd);
+	return err;
+}
+
+static struct lguest_driver lguestblk_drv = {
+	.name = "lguestblk",
+	.owner = THIS_MODULE,
+	.device_type = LGUEST_DEVICE_T_BLOCK,
+	.probe = lguestblk_probe,
+};
+
+static __init int lguestblk_init(void)
+{
+	return register_lguest_driver(&lguestblk_drv);
+}
+module_init(lguestblk_init);
+
+MODULE_DESCRIPTION("Lguest block driver");
+MODULE_LICENSE("GPL");

Re: [PATCH 7/8] lguest: trivial guest block driver

[Rusty Russell]

On Mon, 2007-02-12 at 06:32 +0100, Jens Axboe wrote:
> On Mon, Feb 12 2007, Rusty Russell wrote:
> > On Mon, 2007-02-12 at 05:43 +0100, Jens Axboe wrote:
> > > Here you map the entire request (lets call that segment A..Z), but
> > > end_request() only completes the first chunk of the request. So
> > > elv_next_request() will retrieve the same request again, and you'll then
> > > map B..Z and repeat that transfer. So unless I'm missing some other part
> > > here (just read it over quickly), you are re-doing large parts of a
> > > merged request several times.

virtbench before:
	Time to read from disk (256 kB): 18654562 nsec
After:
	Time to read from disk (256 kB): 8018468 nsec

Thanks Jens!!
Rusty.
PS.  One day I'll buy you a beer and you can explain your nomenclature
theory for the block subsystem 8)

Name: lguest: trivial guest block driver

A simple block driver for lguest (/dev/lgbX).  Only does one request
at once.

Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>

diff -r a155959c419f drivers/block/Makefile
--- a/drivers/block/Makefile	Mon Feb 12 14:26:47 2007 +1100
+++ b/drivers/block/Makefile	Mon Feb 12 14:26:47 2007 +1100
@@ -28,4 +28,5 @@ obj-$(CONFIG_VIODASD)		+= viodasd.o
 obj-$(CONFIG_VIODASD)		+= viodasd.o
 obj-$(CONFIG_BLK_DEV_SX8)	+= sx8.o
 obj-$(CONFIG_BLK_DEV_UB)	+= ub.o
+obj-$(CONFIG_LGUEST_GUEST)	+= lguest_blk.o
 
diff -r a155959c419f drivers/block/lguest_blk.c
--- /dev/null	Thu Jan 01 00:00:00 1970 +0000
+++ b/drivers/block/lguest_blk.c	Mon Feb 12 18:07:05 2007 +1100
@@ -0,0 +1,270 @@
+/* A simple block driver for lguest.
+ *
+ * Copyright 2006 Rusty Russell <rusty@rustcorp.com.au> IBM Corporation
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307  USA
+ */
+//#define DEBUG
+#include <linux/init.h>
+#include <linux/types.h>
+#include <linux/blkdev.h>
+#include <linux/interrupt.h>
+#include <asm/lguest_device.h>
+
+static char next_block_index = 'a';
+
+struct blockdev
+{
+	spinlock_t lock;
+
+	/* The disk structure for the kernel. */
+	struct gendisk *disk;
+
+	/* The major number for this disk. */
+	int major;
+	int irq;
+
+	unsigned long phys_addr;
+	/* The ioremap'ed block page. */
+	struct lguest_block_page *lb_page;
+
+	/* We only have a single request outstanding at a time. */
+	struct lguest_dma dma;
+	struct request *req;
+};
+
+/* Jens gave me this nice helper to end all chunks of a request. */
+static void end_entire_request(struct request *req, int uptodate)
+{
+	if (end_that_request_first(req, uptodate, req->hard_nr_sectors))
+		BUG();
+	add_disk_randomness(req->rq_disk);
+	blkdev_dequeue_request(req);
+	end_that_request_last(req, uptodate);
+}
+
+static irqreturn_t lgb_irq(int irq, void *_bd)
+{
+	struct blockdev *bd = _bd;
+	unsigned long flags;
+
+	if (!bd->req) {
+		pr_debug("No work!\n");
+		return IRQ_NONE;
+	}
+
+	if (!bd->lb_page->result) {
+		pr_debug("No result!\n");
+		return IRQ_NONE;
+	}
+
+	spin_lock_irqsave(&bd->lock, flags);
+	end_entire_request(bd->req, bd->lb_page->result == 1);
+	bd->req = NULL;
+	bd->dma.used_len = 0;
+	blk_start_queue(bd->disk->queue);
+	spin_unlock_irqrestore(&bd->lock, flags);
+	return IRQ_HANDLED;
+}
+
+static unsigned int req_to_dma(struct request *req, struct lguest_dma *dma)
+{
+	unsigned int i = 0, idx, len = 0;
+	struct bio *bio;
+
+	rq_for_each_bio(bio, req) {
+		struct bio_vec *bvec;
+		bio_for_each_segment(bvec, bio, idx) {
+			BUG_ON(i == LGUEST_MAX_DMA_SECTIONS);
+			BUG_ON(!bvec->bv_len);
+			dma->addr[i] = page_to_phys(bvec->bv_page)
+				+ bvec->bv_offset;
+			dma->len[i] = bvec->bv_len;
+			len += bvec->bv_len;
+			i++;
+		}
+	}
+	if (i < LGUEST_MAX_DMA_SECTIONS)
+		dma->len[i] = 0;
+	return len;
+}
+
+static void empty_dma(struct lguest_dma *dma)
+{
+	dma->len[0] = 0;
+}
+
+static void setup_req(struct blockdev *bd,
+		      int type, struct request *req, struct lguest_dma *dma)
+{
+	bd->lb_page->type = type;
+	bd->lb_page->sector = req->sector;
+	bd->lb_page->result = 0;
+	bd->req = req;
+	bd->lb_page->bytes = req_to_dma(req, dma);
+}
+
+static int do_write(struct blockdev *bd, struct request *req)
+{
+	struct lguest_dma send;
+
+	pr_debug("lgb: WRITE sector %li\n", (long)req->sector);
+	setup_req(bd, 1, req, &send);
+
+	hcall(LHCALL_SEND_DMA, bd->phys_addr, __pa(&send), 0);
+	return 1;
+}
+
+static int do_read(struct blockdev *bd, struct request *req)
+{
+	struct lguest_dma ping;
+
+	pr_debug("lgb: READ sector %li\n", (long)req->sector);
+	setup_req(bd, 0, req, &bd->dma);
+
+	empty_dma(&ping);
+	hcall(LHCALL_SEND_DMA,bd->phys_addr,__pa(&ping),0);
+	return 1;
+}
+
+static void do_lgb_request(request_queue_t *q)
+{
+	struct blockdev *bd;
+	struct request *req;
+	int ok;
+
+again:
+	req = elv_next_request(q);
+	if (!req)
+		return;
+
+	bd = req->rq_disk->private_data;
+	/* Sometimes we get repeated requests after blk_stop_queue. */
+	if (bd->req)
+		return;
+
+	if (!blk_fs_request(req)) {
+		pr_debug("Got non-command 0x%08x\n", req->cmd_type);
+	error:
+		req->errors++;
+		end_entire_request(req, 0);
+		goto again;
+	} else {
+		if (rq_data_dir(req) == WRITE)
+			ok = do_write(req->rq_disk->private_data, req);
+		else
+			ok = do_read(req->rq_disk->private_data, req);
+
+		if (!ok)
+			goto error;
+		/* Wait for interrupt to tell us it's done. */
+		blk_stop_queue(q);
+	}
+}
+
+static struct block_device_operations lguestblk_fops = {
+	.owner = THIS_MODULE,
+};
+
+static int lguestblk_probe(struct lguest_device *lhdev)
+{
+	struct blockdev *bd;
+	int err;
+	int irqflags = IRQF_SHARED;
+
+	bd = kmalloc(sizeof(*bd), GFP_KERNEL);
+	if (!bd)
+		return -ENOMEM;
+
+	spin_lock_init(&bd->lock);
+	bd->phys_addr = (lguest_devices[lhdev->index].pfn << PAGE_SHIFT);
+
+	bd->disk = alloc_disk(1);
+	if (!bd->disk) {
+		err = -ENOMEM;
+		goto out_free_bd;
+	}
+
+	bd->disk->queue = blk_init_queue(do_lgb_request, &bd->lock);
+	if (!bd->disk->queue) {
+		err = -ENOMEM;
+		goto out_put;
+	}
+
+	/* We can only handle a certain number of sg entries */
+	blk_queue_max_hw_segments(bd->disk->queue, LGUEST_MAX_DMA_SECTIONS);
+	/* Buffers must not cross page boundaries */
+	blk_queue_segment_boundary(bd->disk->queue, PAGE_SIZE-1);
+
+	bd->irq = lhdev->index+1;
+	bd->major = register_blkdev(0, "lguestblk");
+	if (bd->major < 0) {
+		err = bd->major;
+		goto out_cleanup_queue;
+	}
+	bd->lb_page = (void *)ioremap(bd->phys_addr, PAGE_SIZE);
+	bd->req = NULL;
+
+	sprintf(bd->disk->disk_name, "lgb%c", next_block_index++);
+	if (lguest_devices[lhdev->index].features & LGUEST_DEVICE_F_RANDOMNESS)
+		irqflags |= IRQF_SAMPLE_RANDOM;
+	err = request_irq(bd->irq, lgb_irq, irqflags, bd->disk->disk_name, bd);
+	if (err)
+		goto out_unmap;
+
+	bd->dma.used_len = 0;
+	bd->dma.len[0] = 0;
+	hcall(LHCALL_BIND_DMA, bd->phys_addr, __pa(&bd->dma), (1<<8)+bd->irq);
+
+	printk(KERN_INFO "%s: device %i at major %d\n",
+	       bd->disk->disk_name, lhdev->index, bd->major);
+
+	bd->disk->major = bd->major;
+	bd->disk->first_minor = 0;
+	bd->disk->private_data = bd;
+	bd->disk->fops = &lguestblk_fops;
+	/* This is initialized to the disk size by the other end. */
+	set_capacity(bd->disk, bd->lb_page->num_sectors);
+	add_disk(bd->disk);
+
+	lhdev->private = bd;
+	return 0;
+
+out_unmap:
+	iounmap(bd->lb_page);
+out_cleanup_queue:
+	blk_cleanup_queue(bd->disk->queue);
+out_put:
+	put_disk(bd->disk);
+out_free_bd:
+	kfree(bd);
+	return err;
+}
+
+static struct lguest_driver lguestblk_drv = {
+	.name = "lguestblk",
+	.owner = THIS_MODULE,
+	.device_type = LGUEST_DEVICE_T_BLOCK,
+	.probe = lguestblk_probe,
+};
+
+static __init int lguestblk_init(void)
+{
+	return register_lguest_driver(&lguestblk_drv);
+}
+module_init(lguestblk_init);
+
+MODULE_DESCRIPTION("Lguest block driver");
+MODULE_LICENSE("GPL");

Re: [PATCH 7/8] lguest: trivial guest block driver

[Jens Axboe]

On Mon, Feb 12 2007, Rusty Russell wrote:
> On Mon, 2007-02-12 at 06:32 +0100, Jens Axboe wrote:
> > On Mon, Feb 12 2007, Rusty Russell wrote:
> > > On Mon, 2007-02-12 at 05:43 +0100, Jens Axboe wrote:
> > > > Here you map the entire request (lets call that segment A..Z), but
> > > > end_request() only completes the first chunk of the request. So
> > > > elv_next_request() will retrieve the same request again, and you'll then
> > > > map B..Z and repeat that transfer. So unless I'm missing some other part
> > > > here (just read it over quickly), you are re-doing large parts of a
> > > > merged request several times.
> 
> virtbench before:
> 	Time to read from disk (256 kB): 18654562 nsec
> After:
> 	Time to read from disk (256 kB): 8018468 nsec
> 
> Thanks Jens!!

My pleasure, it's not often you get to make that big a performance
improvement with just a little few lines of change :-)

I guess you'll take changes to make this driver queuing as well? It's
pretty important for good guest io performance as well.

-- 
Jens Axboe

Re: [PATCH 7/8] lguest: trivial guest block driver

[Rusty Russell]

On Mon, 2007-02-12 at 16:01 +0100, Jens Axboe wrote:
> On Mon, Feb 12 2007, Rusty Russell wrote:
> > Thanks Jens!!
> 
> My pleasure, it's not often you get to make that big a performance
> improvement with just a little few lines of change :-)

*cough* I deliberately leave these low hanging fruit in lguest to
encourage people to hack on it.  Really.  *cough*

> I guess you'll take changes to make this driver queuing as well? It's
> pretty important for good guest io performance as well.

The question is whether the guest or host should queue.  If you have
multiple guests sharing a disk in the hose, I would think that the host
is better off queuing.  And this would seem to be the common case.

On my todo list is:
1) Implement write barriers, (-> fsync in the host)
2) Make the host userspace program (lguest) async rather than blocking,
3) Allow multiple outstanding requests.

Then it should be useful for other hypervisors.

Cheers!
Rusty.

Re: [PATCH 7/8] lguest: trivial guest block driver

[Rusty Russell]

On Mon, 2007-02-12 at 16:01 +0100, Jens Axboe wrote:
> On Mon, Feb 12 2007, Rusty Russell wrote:
> > Thanks Jens!!
> 
> My pleasure, it's not often you get to make that big a performance
> improvement with just a little few lines of change :-)

*cough* I deliberately leave these low hanging fruit in lguest to
encourage people to hack on it.  Really.  *cough*

> I guess you'll take changes to make this driver queuing as well? It's
> pretty important for good guest io performance as well.

The question is whether the guest or host should queue.  If you have
multiple guests sharing a disk in the hose, I would think that the host
is better off queuing.  And this would seem to be the common case.

On my todo list is:
1) Implement write barriers, (-> fsync in the host)
2) Make the host userspace program (lguest) async rather than blocking,
3) Allow multiple outstanding requests.

Then it should be useful for other hypervisors.

Cheers!
Rusty.

Re: [PATCH 7/8] lguest: trivial guest block driver

[Jens Axboe]

On Tue, Feb 13 2007, Rusty Russell wrote:
> On Mon, 2007-02-12 at 16:01 +0100, Jens Axboe wrote:
> > On Mon, Feb 12 2007, Rusty Russell wrote:
> > > Thanks Jens!!
> > 
> > My pleasure, it's not often you get to make that big a performance
> > improvement with just a little few lines of change :-)
> 
> *cough* I deliberately leave these low hanging fruit in lguest to
> encourage people to hack on it.  Really.  *cough*

:-)

> > I guess you'll take changes to make this driver queuing as well? It's
> > pretty important for good guest io performance as well.
> 
> The question is whether the guest or host should queue.  If you have
> multiple guests sharing a disk in the hose, I would think that the host
> is better off queuing.  And this would seem to be the common case.

You want queuing in the host, since for queuing to make a performance
difference it needs to be propagated all the way down to the device. But
you can't get queuing at the host level if the guest device doesn't
allow it, you'd be serializing anyway. Well for writes you can, but not
synchronous requests.

Basically allow as much to be sent to the host as possible.

> On my todo list is:
> 1) Implement write barriers, (-> fsync in the host)
> 2) Make the host userspace program (lguest) async rather than blocking,
> 3) Allow multiple outstanding requests.
> 
> Then it should be useful for other hypervisors.

I completely agree with your prioritized list.

-- 
Jens Axboe

Re: [PATCH 5/8] lguest: trivial guest network driver

[Herbert Xu]

On Mon, Feb 12, 2007 at 02:52:01PM +1100, Rusty Russell wrote:
>
> +static void skb_to_dma(const struct sk_buff *skb, unsigned int len,
> +		       struct lguest_dma *dma)
> +{
> +	unsigned int i, seg;
> +
> +	for (i = seg = 0; i < len; seg++, i += rest_of_page(skb->data + i)) {
> +		dma->addr[seg] = virt_to_phys(skb->data + i);
> +		dma->len[seg] = min((unsigned)(len - i),
> +				    rest_of_page(skb->data + i));
> +	}
> +	for (i = 0; i < skb_shinfo(skb)->nr_frags; i++, seg++) {
> +		const skb_frag_t *f = &skb_shinfo(skb)->frags[i];
> +		/* Should not happen with MTU less than 64k - 2 * PAGE_SIZE. */
> +		if (seg == LGUEST_MAX_DMA_SECTIONS) {
> +			printk("Woah dude!  Megapacket!\n");
> +			break;
> +		}
> +		dma->addr[seg] = page_to_phys(f->page) + f->page_offset;
> +		dma->len[seg] = f->size;
> +	}

The length field from the skb covers the fragmented parts as well.
So you don't want to use it as the boundary for the loop over the
skb header data.  As it is, if the skb has fragments, the first
loop will try to read them off the header.

Cheers,
-- 
Visit Openswan at http://www.openswan.org/
Email: Herbert Xu ~{PmV>HI~} <herbert@gondor.apana.org.au>
Home Page: http://gondor.apana.org.au/~herbert/
PGP Key: http://gondor.apana.org.au/~herbert/pubkey.txt

Re: [PATCH 5/8] lguest: trivial guest network driver

[Rusty Russell]

On Tue, 2007-02-13 at 02:55 +1100, Herbert Xu wrote:
> On Mon, Feb 12, 2007 at 02:52:01PM +1100, Rusty Russell wrote:
> >
> > +static void skb_to_dma(const struct sk_buff *skb, unsigned int len,
> > +		       struct lguest_dma *dma)
> > +{
> > +	unsigned int i, seg;
> > +
> > +	for (i = seg = 0; i < len; seg++, i += rest_of_page(skb->data + i)) {
> 
> The length field from the skb covers the fragmented parts as well.
> So you don't want to use it as the boundary for the loop over the
> skb header data.  As it is, if the skb has fragments, the first
> loop will try to read them off the header.

Good spotting!  This function needs to be passed skb_headlen(skb),
rather than skb->len.  Patch is below (I renamed the parameter as well,
for clarity).

It's fascinating why this "worked".  Firstly, for inter-guest
communication, I am not calculating checksums, nor checking them.
Secondly, for guest->host, I turn off hw checksumming so the kernel
disables SG and this code is never executed.  Thirdly, for virtbench's
inter-guest sendfile test does not check the data received is actually
correct.  Finally, while we end up producing a packet which is larger
than skb->len of 1514, the DMA receive buffer for the other guest is
only 1514 bytes, so the result of the transfer is 1514 (==skb->len), so
no error reported by the driver.

==
lguest network driver fix: handle scatter-gather packets correctly

Thanks to Herbert Xu for noticing the bug: "len" here is skb_headlen(),
not skb->len.  Renamed the function to clarify, too.

Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>

diff -r ed6484d145a4 drivers/net/lguest_net.c
--- a/drivers/net/lguest_net.c	Tue Feb 13 11:30:39 2007 +1100
+++ b/drivers/net/lguest_net.c	Tue Feb 13 12:07:15 2007 +1100
@@ -59,14 +59,14 @@ static unsigned long peer_addr(struct lg
 	return info->peer_phys + 4 * peernum;
 }
 
-static void skb_to_dma(const struct sk_buff *skb, unsigned int len,
+static void skb_to_dma(const struct sk_buff *skb, unsigned int headlen,
 		       struct lguest_dma *dma)
 {
 	unsigned int i, seg;
 
-	for (i = seg = 0; i < len; seg++, i += rest_of_page(skb->data + i)) {
+	for (i = seg = 0; i < headlen; seg++, i += rest_of_page(skb->data+i)) {
 		dma->addr[seg] = virt_to_phys(skb->data + i);
-		dma->len[seg] = min((unsigned)(len - i),
+		dma->len[seg] = min((unsigned)(headlen - i),
 				    rest_of_page(skb->data + i));
 	}
 	for (i = 0; i < skb_shinfo(skb)->nr_frags; i++, seg++) {
@@ -90,7 +90,7 @@ static void transfer_packet(struct net_d
 	struct lguestnet_info *info = dev->priv;
 	struct lguest_dma dma;
 
-	skb_to_dma(skb, skb->len, &dma);
+	skb_to_dma(skb, skb_headlen(skb), &dma);
 	pr_debug("xfer length %04x (%u)\n", htons(skb->len), skb->len);
 
 	hcall(LHCALL_SEND_DMA, peer_addr(info,peernum), __pa(&dma),0);

Re: [PATCH 5/8] lguest: trivial guest network driver

[Herbert Xu]

On Tue, Feb 13, 2007 at 01:15:18PM +1100, Rusty Russell wrote:
> 
> Good spotting!  This function needs to be passed skb_headlen(skb),
> rather than skb->len.  Patch is below (I renamed the parameter as well,
> for clarity).

How about just dropping that parameter and using skb_headlen(skb)
directly?

Cheers,
-- 
Visit Openswan at http://www.openswan.org/
Email: Herbert Xu ~{PmV>HI~} <herbert@gondor.apana.org.au>
Home Page: http://gondor.apana.org.au/~herbert/
PGP Key: http://gondor.apana.org.au/~herbert/pubkey.txt

Re: [PATCH 5/8] lguest: trivial guest network driver

[Rusty Russell]

On Wed, 2007-02-14 at 01:06 +1100, Herbert Xu wrote:
> On Tue, Feb 13, 2007 at 01:15:18PM +1100, Rusty Russell wrote:
> > 
> > Good spotting!  This function needs to be passed skb_headlen(skb),
> > rather than skb->len.  Patch is below (I renamed the parameter as well,
> > for clarity).
> 
> How about just dropping that parameter and using skb_headlen(skb)
> directly?

It's also used to generate dma structs for outgoing packets.  In that
case, skb_headlen() == 0:

static struct sk_buff *lguestnet_alloc_skb(struct net_device *dev, int gfpflags)
{
	struct sk_buff *skb;

	skb = alloc_skb(16 + ETH_HLEN + DATA_SIZE, gfpflags);
	if (!skb)
		return NULL;

	skb->dev = dev;
	skb_reserve(skb, 16);
	return skb;
}

/* Find a new skb to put in this slot in shared mem. */
static int fill_slot(struct net_device *dev, unsigned int slot)
{
	struct lguestnet_info *info = dev->priv;
	/* Try to create and register a new one. */
	info->skb[slot] = lguestnet_alloc_skb(dev, GFP_ATOMIC);
	if (!info->skb[slot]) {
		printk("%s: could not fill slot %i\n", dev->name, slot);
		return -ENOMEM;
	}

	skb_to_dma(info->skb[slot], ETH_HLEN + DATA_SIZE, &info->dma[slot]);

Cheers,
Rusty.

Re: [PATCH 5/8] lguest: trivial guest network driver

[Herbert Xu]

On Wed, Feb 14, 2007 at 03:47:55PM +1100, Rusty Russell wrote:
> 
> It's also used to generate dma structs for outgoing packets.  In that
> case, skb_headlen() == 0:

I see, in that case you're guaranteed to have no fragments.
Still it feels a bit weird to have a length field that only
applies to the header part.

Cheers,
-- 
Visit Openswan at http://www.openswan.org/
Email: Herbert Xu ~{PmV>HI~} <herbert@gondor.apana.org.au>
Home Page: http://gondor.apana.org.au/~herbert/
PGP Key: http://gondor.apana.org.au/~herbert/pubkey.txt

Re: [PATCH 5/8] lguest: trivial guest network driver

[Rusty Russell]

On Thu, 2007-02-15 at 00:57 +1100, Herbert Xu wrote:
> On Wed, Feb 14, 2007 at 03:47:55PM +1100, Rusty Russell wrote:
> > 
> > It's also used to generate dma structs for outgoing packets.  In that
> > case, skb_headlen() == 0:
> 
> I see, in that case you're guaranteed to have no fragments.
> Still it feels a bit weird to have a length field that only
> applies to the header part.

Indeed.  In future if I want to create fragmented skbs for receive
(large MTU support), I would actually set skb->len and skb->datalen and
then could avoid the extra argument.

Cheers,
Rusty.

Re: [PATCH 1/8] lguest: Kconfig and headers

[James Morris]

FWIW,

I've set up an unofficial git tree with these patches, and will try and 
track changes as they're posted.

git://git.infradead.org/~jmorris/lguest-testing.git


Use the 'current' branch.




- James
-- 
James Morris
<jmorris@namei.org>

Re: [PATCH 7/7] cleanup: make disable_acpi() valid w/o CONFIG_ACPI

[Len Brown]

Applied.

thanks,
-Len

On Sunday 11 February 2007 22:39, Rusty Russell wrote:
> Len Brown <lenb@kernel.org> said:
> > Okay, but better to use disable_acpi()
> > indeed, since this would be the first code not already inside CONFIG_ACPI
> > to invoke disable_acpi(), we could define the inline as empty and you could
> > then scratch the #ifdef too.
> 
> Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
> 
> ===================================================================
> --- a/include/asm-i386/acpi.h
> +++ b/include/asm-i386/acpi.h
> @@ -127,6 +127,7 @@ extern int acpi_irq_balance_set(char *st
>  #define acpi_ioapic 0
>  static inline void acpi_noirq_set(void) { }
>  static inline void acpi_disable_pci(void) { }
> +static inline void disable_acpi(void) { }
>  
>  #endif	/* !CONFIG_ACPI */
>  
> 
> 
> -
> To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html
> Please read the FAQ at  http://www.tux.org/lkml/
> 

Re: [PATCH 5/7] cleanup: Rename cpu_gdt_descr and remove extern declaration from smpboot.c

[Zachary Amsden]

Rusty Russell wrote:
> When I implemented the DECLARE_PER_CPU(var) macros, I was careful that
> people couldn't use "var" in a non-percpu context, by prepending
> percpu__.  I never considered that this would allow them to overload
> the same name for a per-cpu and a non-percpu variable.
>
> It is only one of many horrors in the i386 boot code, but let's rename
> the non-perpcu cpu_gdt_descr to early_gdt_descr (not boot_gdt_descr,
> that's something else...)
>   

Finally, someone not just noticed, but fixed it!  Thank you Rusty.  Was 
on my todo list, but never bubbled up.

Acked-by: Zachary Amsden <zach@vmware.com>