[PATCH 01/04]Create x86 directory to hold x86-specific files.

All of lore.kernel.org
 help / color / mirror / Atom feed

* [PATCH 01/04]Create x86 directory to hold x86-specific files.
@ 2008-07-11  5:43 ` Zhang, Xiantao
  0 siblings, 0 replies; 24+ messages in thread
From: Zhang, Xiantao @ 2008-07-11  5:43 UTC (permalink / raw)
  To: kvm-ia64

[-- Attachment #1: Type: text/plain, Size: 3418 bytes --]

From 03ac444d1ab4446c587e8180ceaba60b9e75b28d Mon Sep 17 00:00:00 2001
From: Xiantao Zhang <xiantao.zhang@intel.com>
Date: Fri, 11 Jul 2008 10:13:08 +0800
Subject: [PATCH] KVM: external module: Moving x86-speicif files to x86
directory.

Create x86 directory to hold x86-specific files.
Signed-off-by: Xiantao Zhang <xiantao.zhang@intel.com>
---
 kernel/Kbuild                             |   10 +++++-----
 kernel/Makefile                           |    2 +-
 kernel/{ => x86}/anon_inodes.c            |    0 
 kernel/{ => x86}/debug.h                  |    0 
 kernel/{ => x86}/external-module-compat.c |    0 
 kernel/{ => x86}/external-module-compat.h |    0 
 kernel/{ => x86}/preempt.c                |    0 
 kernel/{ => x86}/vmx-debug.c              |    0 
 8 files changed, 6 insertions(+), 6 deletions(-)
 rename kernel/{ => x86}/anon_inodes.c (100%)
 rename kernel/{ => x86}/debug.h (100%)
 rename kernel/{ => x86}/external-module-compat.c (100%)
 rename kernel/{ => x86}/external-module-compat.h (100%)
 rename kernel/{ => x86}/preempt.c (100%)
 rename kernel/{ => x86}/vmx-debug.c (100%)

diff --git a/kernel/Kbuild b/kernel/Kbuild
index 676f28d..de56aec 100644
--- a/kernel/Kbuild
+++ b/kernel/Kbuild
@@ -1,11 +1,11 @@
 obj-m := kvm.o kvm-intel.o kvm-amd.o
-kvm-objs := kvm_main.o x86.o mmu.o x86_emulate.o anon_inodes.o irq.o
i8259.o \
-	 lapic.o ioapic.o preempt.o i8254.o coalesced_mmio.o \
-	 external-module-compat.o
+kvm-objs := kvm_main.o x86.o mmu.o x86_emulate.o x86/anon_inodes.o
irq.o i8259.o \
+	 lapic.o ioapic.o x86/preempt.o i8254.o coalesced_mmio.o \
+	 x86/external-module-compat.o
 ifeq ($(CONFIG_KVM_TRACE),y)
 kvm-objs += kvm_trace.o
 endif
-kvm-intel-objs := vmx.o vmx-debug.o external-module-compat.o
-kvm-amd-objs := svm.o external-module-compat.o
+kvm-intel-objs := vmx.o x86/vmx-debug.o x86/external-module-compat.o
+kvm-amd-objs := svm.o x86/external-module-compat.o
 
 CFLAGS_kvm_main.o = -DKVM_MAIN
diff --git a/kernel/Makefile b/kernel/Makefile
index 69b4981..c2c666e 100644
--- a/kernel/Makefile
+++ b/kernel/Makefile
@@ -35,7 +35,7 @@ all::
 	$(MAKE) -C $(KERNELDIR) M=`pwd` \
 		LINUXINCLUDE="-I`pwd`/include -Iinclude
-I`pwd`/include-compat \
 		-include include/linux/autoconf.h \
-		-include `pwd`/external-module-compat.h"
+		-include `pwd`/$(ARCH_DIR)/external-module-compat.h"
 		"$$@"
 
 sync: header-sync source-sync
diff --git a/kernel/anon_inodes.c b/kernel/x86/anon_inodes.c
similarity index 100%
rename from kernel/anon_inodes.c
rename to kernel/x86/anon_inodes.c
diff --git a/kernel/debug.h b/kernel/x86/debug.h
similarity index 100%
rename from kernel/debug.h
rename to kernel/x86/debug.h
diff --git a/kernel/external-module-compat.c
b/kernel/x86/external-module-compat.c
similarity index 100%
rename from kernel/external-module-compat.c
rename to kernel/x86/external-module-compat.c
diff --git a/kernel/external-module-compat.h
b/kernel/x86/external-module-compat.h
similarity index 100%
rename from kernel/external-module-compat.h
rename to kernel/x86/external-module-compat.h
diff --git a/kernel/preempt.c b/kernel/x86/preempt.c
similarity index 100%
rename from kernel/preempt.c
rename to kernel/x86/preempt.c
diff --git a/kernel/vmx-debug.c b/kernel/x86/vmx-debug.c
similarity index 100%
rename from kernel/vmx-debug.c
rename to kernel/x86/vmx-debug.c
-- 
1.5.2

[-- Attachment #2: 0001-KVM-external-module-Moving-x86-speicif-files-to-x8.patch --]
[-- Type: application/octet-stream, Size: 3329 bytes --]

From 03ac444d1ab4446c587e8180ceaba60b9e75b28d Mon Sep 17 00:00:00 2001
From: Xiantao Zhang <xiantao.zhang@intel.com>
Date: Fri, 11 Jul 2008 10:13:08 +0800
Subject: [PATCH] KVM: external module: Moving x86-speicif files to x86 directory.

Create x86 directory to hold x86-specific files.
Signed-off-by: Xiantao Zhang <xiantao.zhang@intel.com>
---
 kernel/Kbuild                             |   10 +++++-----
 kernel/Makefile                           |    2 +-
 kernel/{ => x86}/anon_inodes.c            |    0 
 kernel/{ => x86}/debug.h                  |    0 
 kernel/{ => x86}/external-module-compat.c |    0 
 kernel/{ => x86}/external-module-compat.h |    0 
 kernel/{ => x86}/preempt.c                |    0 
 kernel/{ => x86}/vmx-debug.c              |    0 
 8 files changed, 6 insertions(+), 6 deletions(-)
 rename kernel/{ => x86}/anon_inodes.c (100%)
 rename kernel/{ => x86}/debug.h (100%)
 rename kernel/{ => x86}/external-module-compat.c (100%)
 rename kernel/{ => x86}/external-module-compat.h (100%)
 rename kernel/{ => x86}/preempt.c (100%)
 rename kernel/{ => x86}/vmx-debug.c (100%)

diff --git a/kernel/Kbuild b/kernel/Kbuild
index 676f28d..de56aec 100644
--- a/kernel/Kbuild
+++ b/kernel/Kbuild
@@ -1,11 +1,11 @@
 obj-m := kvm.o kvm-intel.o kvm-amd.o
-kvm-objs := kvm_main.o x86.o mmu.o x86_emulate.o anon_inodes.o irq.o i8259.o \
-	 lapic.o ioapic.o preempt.o i8254.o coalesced_mmio.o \
-	 external-module-compat.o
+kvm-objs := kvm_main.o x86.o mmu.o x86_emulate.o x86/anon_inodes.o irq.o i8259.o \
+	 lapic.o ioapic.o x86/preempt.o i8254.o coalesced_mmio.o \
+	 x86/external-module-compat.o
 ifeq ($(CONFIG_KVM_TRACE),y)
 kvm-objs += kvm_trace.o
 endif
-kvm-intel-objs := vmx.o vmx-debug.o external-module-compat.o
-kvm-amd-objs := svm.o external-module-compat.o
+kvm-intel-objs := vmx.o x86/vmx-debug.o x86/external-module-compat.o
+kvm-amd-objs := svm.o x86/external-module-compat.o
 
 CFLAGS_kvm_main.o = -DKVM_MAIN
diff --git a/kernel/Makefile b/kernel/Makefile
index 69b4981..c2c666e 100644
--- a/kernel/Makefile
+++ b/kernel/Makefile
@@ -35,7 +35,7 @@ all::
 	$(MAKE) -C $(KERNELDIR) M=`pwd` \
 		LINUXINCLUDE="-I`pwd`/include -Iinclude -I`pwd`/include-compat \
 		-include include/linux/autoconf.h \
-		-include `pwd`/external-module-compat.h"
+		-include `pwd`/$(ARCH_DIR)/external-module-compat.h"
 		"$$@"
 
 sync: header-sync source-sync
diff --git a/kernel/anon_inodes.c b/kernel/x86/anon_inodes.c
similarity index 100%
rename from kernel/anon_inodes.c
rename to kernel/x86/anon_inodes.c
diff --git a/kernel/debug.h b/kernel/x86/debug.h
similarity index 100%
rename from kernel/debug.h
rename to kernel/x86/debug.h
diff --git a/kernel/external-module-compat.c b/kernel/x86/external-module-compat.c
similarity index 100%
rename from kernel/external-module-compat.c
rename to kernel/x86/external-module-compat.c
diff --git a/kernel/external-module-compat.h b/kernel/x86/external-module-compat.h
similarity index 100%
rename from kernel/external-module-compat.h
rename to kernel/x86/external-module-compat.h
diff --git a/kernel/preempt.c b/kernel/x86/preempt.c
similarity index 100%
rename from kernel/preempt.c
rename to kernel/x86/preempt.c
diff --git a/kernel/vmx-debug.c b/kernel/x86/vmx-debug.c
similarity index 100%
rename from kernel/vmx-debug.c
rename to kernel/x86/vmx-debug.c
-- 
1.5.2


^ permalink raw reply related	[flat|nested] 24+ messages in thread

* [PATCH 01/04]Create x86 directory to hold x86-specific files.
@ 2008-07-11  5:43 ` Zhang, Xiantao
  0 siblings, 0 replies; 24+ messages in thread
From: Zhang, Xiantao @ 2008-07-11  5:43 UTC (permalink / raw)
  To: kvm-ia64, kvm; +Cc: Avi Kivity

[-- Attachment #1: Type: text/plain, Size: 3419 bytes --]

>From 03ac444d1ab4446c587e8180ceaba60b9e75b28d Mon Sep 17 00:00:00 2001
From: Xiantao Zhang <xiantao.zhang@intel.com>
Date: Fri, 11 Jul 2008 10:13:08 +0800
Subject: [PATCH] KVM: external module: Moving x86-speicif files to x86
directory.

Create x86 directory to hold x86-specific files.
Signed-off-by: Xiantao Zhang <xiantao.zhang@intel.com>
---
 kernel/Kbuild                             |   10 +++++-----
 kernel/Makefile                           |    2 +-
 kernel/{ => x86}/anon_inodes.c            |    0 
 kernel/{ => x86}/debug.h                  |    0 
 kernel/{ => x86}/external-module-compat.c |    0 
 kernel/{ => x86}/external-module-compat.h |    0 
 kernel/{ => x86}/preempt.c                |    0 
 kernel/{ => x86}/vmx-debug.c              |    0 
 8 files changed, 6 insertions(+), 6 deletions(-)
 rename kernel/{ => x86}/anon_inodes.c (100%)
 rename kernel/{ => x86}/debug.h (100%)
 rename kernel/{ => x86}/external-module-compat.c (100%)
 rename kernel/{ => x86}/external-module-compat.h (100%)
 rename kernel/{ => x86}/preempt.c (100%)
 rename kernel/{ => x86}/vmx-debug.c (100%)

diff --git a/kernel/Kbuild b/kernel/Kbuild
index 676f28d..de56aec 100644
--- a/kernel/Kbuild
+++ b/kernel/Kbuild
@@ -1,11 +1,11 @@
 obj-m := kvm.o kvm-intel.o kvm-amd.o
-kvm-objs := kvm_main.o x86.o mmu.o x86_emulate.o anon_inodes.o irq.o
i8259.o \
-	 lapic.o ioapic.o preempt.o i8254.o coalesced_mmio.o \
-	 external-module-compat.o
+kvm-objs := kvm_main.o x86.o mmu.o x86_emulate.o x86/anon_inodes.o
irq.o i8259.o \
+	 lapic.o ioapic.o x86/preempt.o i8254.o coalesced_mmio.o \
+	 x86/external-module-compat.o
 ifeq ($(CONFIG_KVM_TRACE),y)
 kvm-objs += kvm_trace.o
 endif
-kvm-intel-objs := vmx.o vmx-debug.o external-module-compat.o
-kvm-amd-objs := svm.o external-module-compat.o
+kvm-intel-objs := vmx.o x86/vmx-debug.o x86/external-module-compat.o
+kvm-amd-objs := svm.o x86/external-module-compat.o
 
 CFLAGS_kvm_main.o = -DKVM_MAIN
diff --git a/kernel/Makefile b/kernel/Makefile
index 69b4981..c2c666e 100644
--- a/kernel/Makefile
+++ b/kernel/Makefile
@@ -35,7 +35,7 @@ all::
 	$(MAKE) -C $(KERNELDIR) M=`pwd` \
 		LINUXINCLUDE="-I`pwd`/include -Iinclude
-I`pwd`/include-compat \
 		-include include/linux/autoconf.h \
-		-include `pwd`/external-module-compat.h"
+		-include `pwd`/$(ARCH_DIR)/external-module-compat.h"
 		"$$@"
 
 sync: header-sync source-sync
diff --git a/kernel/anon_inodes.c b/kernel/x86/anon_inodes.c
similarity index 100%
rename from kernel/anon_inodes.c
rename to kernel/x86/anon_inodes.c
diff --git a/kernel/debug.h b/kernel/x86/debug.h
similarity index 100%
rename from kernel/debug.h
rename to kernel/x86/debug.h
diff --git a/kernel/external-module-compat.c
b/kernel/x86/external-module-compat.c
similarity index 100%
rename from kernel/external-module-compat.c
rename to kernel/x86/external-module-compat.c
diff --git a/kernel/external-module-compat.h
b/kernel/x86/external-module-compat.h
similarity index 100%
rename from kernel/external-module-compat.h
rename to kernel/x86/external-module-compat.h
diff --git a/kernel/preempt.c b/kernel/x86/preempt.c
similarity index 100%
rename from kernel/preempt.c
rename to kernel/x86/preempt.c
diff --git a/kernel/vmx-debug.c b/kernel/x86/vmx-debug.c
similarity index 100%
rename from kernel/vmx-debug.c
rename to kernel/x86/vmx-debug.c
-- 
1.5.2

[-- Attachment #2: 0001-KVM-external-module-Moving-x86-speicif-files-to-x8.patch --]
[-- Type: application/octet-stream, Size: 3329 bytes --]

From 03ac444d1ab4446c587e8180ceaba60b9e75b28d Mon Sep 17 00:00:00 2001
From: Xiantao Zhang <xiantao.zhang@intel.com>
Date: Fri, 11 Jul 2008 10:13:08 +0800
Subject: [PATCH] KVM: external module: Moving x86-speicif files to x86 directory.

Create x86 directory to hold x86-specific files.
Signed-off-by: Xiantao Zhang <xiantao.zhang@intel.com>
---
 kernel/Kbuild                             |   10 +++++-----
 kernel/Makefile                           |    2 +-
 kernel/{ => x86}/anon_inodes.c            |    0 
 kernel/{ => x86}/debug.h                  |    0 
 kernel/{ => x86}/external-module-compat.c |    0 
 kernel/{ => x86}/external-module-compat.h |    0 
 kernel/{ => x86}/preempt.c                |    0 
 kernel/{ => x86}/vmx-debug.c              |    0 
 8 files changed, 6 insertions(+), 6 deletions(-)
 rename kernel/{ => x86}/anon_inodes.c (100%)
 rename kernel/{ => x86}/debug.h (100%)
 rename kernel/{ => x86}/external-module-compat.c (100%)
 rename kernel/{ => x86}/external-module-compat.h (100%)
 rename kernel/{ => x86}/preempt.c (100%)
 rename kernel/{ => x86}/vmx-debug.c (100%)

diff --git a/kernel/Kbuild b/kernel/Kbuild
index 676f28d..de56aec 100644
--- a/kernel/Kbuild
+++ b/kernel/Kbuild
@@ -1,11 +1,11 @@
 obj-m := kvm.o kvm-intel.o kvm-amd.o
-kvm-objs := kvm_main.o x86.o mmu.o x86_emulate.o anon_inodes.o irq.o i8259.o \
-	 lapic.o ioapic.o preempt.o i8254.o coalesced_mmio.o \
-	 external-module-compat.o
+kvm-objs := kvm_main.o x86.o mmu.o x86_emulate.o x86/anon_inodes.o irq.o i8259.o \
+	 lapic.o ioapic.o x86/preempt.o i8254.o coalesced_mmio.o \
+	 x86/external-module-compat.o
 ifeq ($(CONFIG_KVM_TRACE),y)
 kvm-objs += kvm_trace.o
 endif
-kvm-intel-objs := vmx.o vmx-debug.o external-module-compat.o
-kvm-amd-objs := svm.o external-module-compat.o
+kvm-intel-objs := vmx.o x86/vmx-debug.o x86/external-module-compat.o
+kvm-amd-objs := svm.o x86/external-module-compat.o
 
 CFLAGS_kvm_main.o = -DKVM_MAIN
diff --git a/kernel/Makefile b/kernel/Makefile
index 69b4981..c2c666e 100644
--- a/kernel/Makefile
+++ b/kernel/Makefile
@@ -35,7 +35,7 @@ all::
 	$(MAKE) -C $(KERNELDIR) M=`pwd` \
 		LINUXINCLUDE="-I`pwd`/include -Iinclude -I`pwd`/include-compat \
 		-include include/linux/autoconf.h \
-		-include `pwd`/external-module-compat.h"
+		-include `pwd`/$(ARCH_DIR)/external-module-compat.h"
 		"$$@"
 
 sync: header-sync source-sync
diff --git a/kernel/anon_inodes.c b/kernel/x86/anon_inodes.c
similarity index 100%
rename from kernel/anon_inodes.c
rename to kernel/x86/anon_inodes.c
diff --git a/kernel/debug.h b/kernel/x86/debug.h
similarity index 100%
rename from kernel/debug.h
rename to kernel/x86/debug.h
diff --git a/kernel/external-module-compat.c b/kernel/x86/external-module-compat.c
similarity index 100%
rename from kernel/external-module-compat.c
rename to kernel/x86/external-module-compat.c
diff --git a/kernel/external-module-compat.h b/kernel/x86/external-module-compat.h
similarity index 100%
rename from kernel/external-module-compat.h
rename to kernel/x86/external-module-compat.h
diff --git a/kernel/preempt.c b/kernel/x86/preempt.c
similarity index 100%
rename from kernel/preempt.c
rename to kernel/x86/preempt.c
diff --git a/kernel/vmx-debug.c b/kernel/x86/vmx-debug.c
similarity index 100%
rename from kernel/vmx-debug.c
rename to kernel/x86/vmx-debug.c
-- 
1.5.2


^ permalink raw reply related	[flat|nested] 24+ messages in thread

* Re: [PATCH 01/04]Create x86 directory to hold x86-specific files.
  2008-07-11  5:43 ` Zhang, Xiantao
@ 2008-07-13  8:53   ` Avi Kivity
  -1 siblings, 0 replies; 24+ messages in thread
From: Avi Kivity @ 2008-07-13  8:53 UTC (permalink / raw)
  To: kvm-ia64

Zhang, Xiantao wrote:
> From 03ac444d1ab4446c587e8180ceaba60b9e75b28d Mon Sep 17 00:00:00 2001
> From: Xiantao Zhang <xiantao.zhang@intel.com>
> Date: Fri, 11 Jul 2008 10:13:08 +0800
> Subject: [PATCH] KVM: external module: Moving x86-speicif files to x86
> directory.
>
> Create x86 directory to hold x86-specific files.
> Signed-off-by: Xiantao Zhang <xiantao.zhang@intel.com>
> ---
>  kernel/{ => x86}/anon_inodes.c            |    0 
>   

This isn't really x86 specific.  It's just kernel version dependent.  
The problem is that it is built unconditionally, even if the kernel has 
anon_inodes support.

Please send a patch that wraps the entire file in #ifdef so that we use 
the host kernel's anon_inodes if it is available.

>  kernel/{ => x86}/external-module-compat.c |    0 
>  kernel/{ => x86}/external-module-compat.h |    0 
>   

Parts of this are generic, for example the mutex code.  Please move only 
the x86 specifc parts (even if ia64 doesn't need everything in the 
generic code).

-- 
error compiling committee.c: too many arguments to function

^ permalink raw reply	[flat|nested] 24+ messages in thread

* Re: [PATCH 01/04]Create x86 directory to hold x86-specific files.
@ 2008-07-13  8:53   ` Avi Kivity
  0 siblings, 0 replies; 24+ messages in thread
From: Avi Kivity @ 2008-07-13  8:53 UTC (permalink / raw)
  To: Zhang, Xiantao; +Cc: kvm-ia64, kvm

Zhang, Xiantao wrote:
> From 03ac444d1ab4446c587e8180ceaba60b9e75b28d Mon Sep 17 00:00:00 2001
> From: Xiantao Zhang <xiantao.zhang@intel.com>
> Date: Fri, 11 Jul 2008 10:13:08 +0800
> Subject: [PATCH] KVM: external module: Moving x86-speicif files to x86
> directory.
>
> Create x86 directory to hold x86-specific files.
> Signed-off-by: Xiantao Zhang <xiantao.zhang@intel.com>
> ---
>  kernel/{ => x86}/anon_inodes.c            |    0 
>   

This isn't really x86 specific.  It's just kernel version dependent.  
The problem is that it is built unconditionally, even if the kernel has 
anon_inodes support.

Please send a patch that wraps the entire file in #ifdef so that we use 
the host kernel's anon_inodes if it is available.

>  kernel/{ => x86}/external-module-compat.c |    0 
>  kernel/{ => x86}/external-module-compat.h |    0 
>   

Parts of this are generic, for example the mutex code.  Please move only 
the x86 specifc parts (even if ia64 doesn't need everything in the 
generic code).

-- 
error compiling committee.c: too many arguments to function

^ permalink raw reply	[flat|nested] 24+ messages in thread

* RE: [PATCH 01/04]Create x86 directory to hold x86-specific files.
  2008-07-13  8:53   ` Avi Kivity
@ 2008-07-16  9:11   ` Zhang, Xiantao
  -1 siblings, 0 replies; 24+ messages in thread
From: Zhang, Xiantao @ 2008-07-16  9:11 UTC (permalink / raw)
  To: kvm-ia64

Avi Kivity wrote:
> Zhang, Xiantao wrote:
>> From 03ac444d1ab4446c587e8180ceaba60b9e75b28d Mon Sep 17 00:00:00
>> 2001 From: Xiantao Zhang <xiantao.zhang@intel.com>
>> Date: Fri, 11 Jul 2008 10:13:08 +0800
>> Subject: [PATCH] KVM: external module: Moving x86-speicif files to
>> x86 directory. 
>> 
>> Create x86 directory to hold x86-specific files.
>> Signed-off-by: Xiantao Zhang <xiantao.zhang@intel.com> ---
>>  kernel/{ => x86}/anon_inodes.c            |    0
>> 
> 
> This isn't really x86 specific.  It's just kernel version dependent.
> The problem is that it is built unconditionally, even if the kernel
> has anon_inodes support.
> Please send a patch that wraps the entire file in #ifdef so that we
> use the host kernel's anon_inodes if it is available.
> 
Sure, I will update the patch. 
>>  kernel/{ => x86}/external-module-compat.c |    0
>>  kernel/{ => x86}/external-module-compat.h |    0
>> 
> 
> Parts of this are generic, for example the mutex code.  Please move
> only the x86 specifc parts (even if ia64 doesn't need everything in
> the generic code).'
OK.

^ permalink raw reply	[flat|nested] 24+ messages in thread

* RE: [PATCH 01/04]Create x86 directory to hold x86-specific files.
@ 2008-07-16  9:11   ` Zhang, Xiantao
  0 siblings, 0 replies; 24+ messages in thread
From: Zhang, Xiantao @ 2008-07-16  9:11 UTC (permalink / raw)
  To: Avi Kivity; +Cc: kvm-ia64, kvm

Avi Kivity wrote:
> Zhang, Xiantao wrote:
>> From 03ac444d1ab4446c587e8180ceaba60b9e75b28d Mon Sep 17 00:00:00
>> 2001 From: Xiantao Zhang <xiantao.zhang@intel.com>
>> Date: Fri, 11 Jul 2008 10:13:08 +0800
>> Subject: [PATCH] KVM: external module: Moving x86-speicif files to
>> x86 directory. 
>> 
>> Create x86 directory to hold x86-specific files.
>> Signed-off-by: Xiantao Zhang <xiantao.zhang@intel.com> ---
>>  kernel/{ => x86}/anon_inodes.c            |    0
>> 
> 
> This isn't really x86 specific.  It's just kernel version dependent.
> The problem is that it is built unconditionally, even if the kernel
> has anon_inodes support.
> Please send a patch that wraps the entire file in #ifdef so that we
> use the host kernel's anon_inodes if it is available.
> 
Sure, I will update the patch. 
>>  kernel/{ => x86}/external-module-compat.c |    0
>>  kernel/{ => x86}/external-module-compat.h |    0
>> 
> 
> Parts of this are generic, for example the mutex code.  Please move
> only the x86 specifc parts (even if ia64 doesn't need everything in
> the generic code).'
OK.

^ permalink raw reply	[flat|nested] 24+ messages in thread

* [PATCH 01/04]Create x86 directory to hold x86-specific files.
@ 2008-07-25  9:26   ` Zhang, Xiantao
  0 siblings, 0 replies; 24+ messages in thread
From: Zhang, Xiantao @ 2008-07-25  9:26 UTC (permalink / raw)
  To: kvm-ia64

From c8c8c5a6b3ff943cb645f9ac4ef169490f14cd08 Mon Sep 17 00:00:00 2001
From: Zhang Xiantao <xiantao.zhang@intel.com>
Date: Fri, 25 Jul 2008 16:50:30 +0800
Subject: [PATCH] KVM: external module: Moving x86-speicif files to x86
directory.

Create x86 directory to hold x86-specific files.
Signed-off-by: Zhang Xiantao <xiantao.zhang@intel.com>
---
 kernel/Kbuild                             |    8 +-
 kernel/Makefile                           |    2 +-
 kernel/external-module-compat-comm.h      |  475 ++++++++++++++++++
 kernel/external-module-compat.h           |  759
-----------------------------
 kernel/{ => x86}/debug.h                  |    0 
 kernel/{ => x86}/external-module-compat.c |    0 
 kernel/x86/external-module-compat.h       |  291 +++++++++++
 kernel/{ => x86}/preempt.c                |    0 
 kernel/{ => x86}/vmx-debug.c              |    0 
 9 files changed, 771 insertions(+), 764 deletions(-)
 create mode 100644 kernel/external-module-compat-comm.h
 delete mode 100644 kernel/external-module-compat.h
 rename kernel/{ => x86}/debug.h (100%)
 rename kernel/{ => x86}/external-module-compat.c (100%)
 create mode 100644 kernel/x86/external-module-compat.h
 rename kernel/{ => x86}/preempt.c (100%)
 rename kernel/{ => x86}/vmx-debug.c (100%)

diff --git a/kernel/Kbuild b/kernel/Kbuild
index 676f28d..cb63582 100644
--- a/kernel/Kbuild
+++ b/kernel/Kbuild
@@ -1,11 +1,11 @@
 obj-m := kvm.o kvm-intel.o kvm-amd.o
 kvm-objs := kvm_main.o x86.o mmu.o x86_emulate.o anon_inodes.o irq.o
i8259.o \
-	 lapic.o ioapic.o preempt.o i8254.o coalesced_mmio.o \
-	 external-module-compat.o
+	 lapic.o ioapic.o x86/preempt.o i8254.o coalesced_mmio.o \
+	 x86/external-module-compat.o
 ifeq ($(CONFIG_KVM_TRACE),y)
 kvm-objs += kvm_trace.o
 endif
-kvm-intel-objs := vmx.o vmx-debug.o external-module-compat.o
-kvm-amd-objs := svm.o external-module-compat.o
+kvm-intel-objs := vmx.o x86/vmx-debug.o x86/external-module-compat.o
+kvm-amd-objs := svm.o x86/external-module-compat.o
 
 CFLAGS_kvm_main.o = -DKVM_MAIN
diff --git a/kernel/Makefile b/kernel/Makefile
index 69b4981..c2c666e 100644
--- a/kernel/Makefile
+++ b/kernel/Makefile
@@ -35,7 +35,7 @@ all::
 	$(MAKE) -C $(KERNELDIR) M=`pwd` \
 		LINUXINCLUDE="-I`pwd`/include -Iinclude
-I`pwd`/include-compat \
 		-include include/linux/autoconf.h \
-		-include `pwd`/external-module-compat.h"
+		-include `pwd`/$(ARCH_DIR)/external-module-compat.h"
 		"$$@"
 
 sync: header-sync source-sync
diff --git a/kernel/external-module-compat-comm.h
b/kernel/external-module-compat-comm.h
new file mode 100644
index 0000000..d4bf395
--- /dev/null
+++ b/kernel/external-module-compat-comm.h
@@ -0,0 +1,475 @@
+
+/*
+ * Compatibility header for building as an external module.
+ */
+
+/*
+ * Avoid picking up the kernel's kvm.h in case we have a newer one.
+ */
+
+#include <linux/compiler.h>
+#include <linux/version.h>
+#include <linux/string.h>
+#include <linux/kvm.h>
+#include <linux/kvm_para.h>
+#include <linux/cpu.h>
+#include <linux/time.h>
+#include <asm/processor.h>
+#include <linux/hrtimer.h>
+#include <asm/bitops.h>
+#include <asm/msr.h>
+
+/*
+ * 2.6.16 does not have GFP_NOWAIT
+ */
+
+#include <linux/gfp.h>
+
+#ifndef GFP_NOWAIT
+#define GFP_NOWAIT (GFP_ATOMIC & ~__GFP_HIGH)
+#endif
+
+
+/*
+ * kvm profiling support needs 2.6.20
+ */
+#include <linux/profile.h>
+
+#ifndef KVM_PROFILING
+#define KVM_PROFILING 1234
+#define prof_on       4321
+#endif
+
+/*
+ * smp_call_function_single() is not exported below 2.6.20, and has
different
+ * semantics below 2.6.23.  The 'nonatomic' argument was removed in
2.6.27.
+ */
+#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,27)
+
+int kvm_smp_call_function_single(int cpu, void (*func)(void *info),
+				 void *info, int wait);
+
+#define smp_call_function_single kvm_smp_call_function_single
+
+#endif
+
+/* on_each_cpu() lost an argument in 2.6.27. */
+#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,27)
+
+#define kvm_on_each_cpu(func, info, wait) on_each_cpu(func, info, 0,
wait)
+
+#else
+
+#define kvm_on_each_cpu(func, info, wait) on_each_cpu(func, info, wait)
+
+#endif
+
+/*
+ * The cpu hotplug stubs are broken if !CONFIG_CPU_HOTPLUG
+ */
+
+#if LINUX_VERSION_CODE <= KERNEL_VERSION(2,6,15)
+#define DEFINE_MUTEX(a) DECLARE_MUTEX(a)
+#define mutex_lock_interruptible(a) down_interruptible(a)
+#define mutex_unlock(a) up(a)
+#define mutex_lock(a) down(a)
+#define mutex_init(a) init_MUTEX(a)
+#define mutex_trylock(a) down_trylock(a)
+#define mutex semaphore
+#endif
+
+#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,14)
+#ifndef kzalloc
+#define kzalloc(size,flags)			\
+({						\
+	void *__ret = kmalloc(size, flags);	\
+	if (__ret)				\
+		memset(__ret, 0, size);		\
+	__ret;					\
+})
+#endif
+#endif
+
+#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,17)
+#ifndef kmem_cache_zalloc
+#define kmem_cache_zalloc(cache,flags)			  \
+({							  \
+	void *__ret = kmem_cache_alloc(cache, flags);	  \
+	if (__ret)                                        \
+		memset(__ret, 0, kmem_cache_size(cache)); \
+	__ret;                                            \
+})
+#endif
+#endif
+
+#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,21)
+
+#ifndef CONFIG_HOTPLUG_CPU
+#define register_cpu_notifier(nb) (0)
+#endif
+
+#endif
+
+#include <linux/miscdevice.h>
+#ifndef KVM_MINOR
+#define KVM_MINOR 232
+#endif
+
+#include <linux/notifier.h>
+#ifndef CPU_TASKS_FROZEN
+
+#define CPU_TASKS_FROZEN       0x0010
+#define CPU_ONLINE_FROZEN      (CPU_ONLINE | CPU_TASKS_FROZEN)
+#define CPU_UP_PREPARE_FROZEN  (CPU_UP_PREPARE | CPU_TASKS_FROZEN)
+#define CPU_UP_CANCELED_FROZEN (CPU_UP_CANCELED | CPU_TASKS_FROZEN)
+#define CPU_DOWN_PREPARE_FROZEN        (CPU_DOWN_PREPARE |
CPU_TASKS_FROZEN)
+#define CPU_DOWN_FAILED_FROZEN (CPU_DOWN_FAILED | CPU_TASKS_FROZEN)
+#define CPU_DEAD_FROZEN                (CPU_DEAD | CPU_TASKS_FROZEN)
+
+#endif
+
+#ifndef CPU_DYING
+#define CPU_DYING 0x000A
+#define CPU_DYING_FROZEN (CPU_DYING | CPU_TASKS_FROZEN)
+#endif
+
+#include <asm/system.h>
+
+struct inode;
+#include <linux/anon_inodes.h>
+#define anon_inode_getfd kvm_anon_inode_getfd
+int kvm_init_anon_inodes(void);
+void kvm_exit_anon_inodes(void);
+int anon_inode_getfd(const char *name,
+		     const struct file_operations *fops,
+		     void *priv);
+
+/*
+ * 2.6.23 removed the cache destructor
+ */
+#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,23)
+#  define kmem_cache_create(name, size, align, flags, ctor) \
+	kmem_cache_create(name, size, align, flags, ctor, NULL)
+#endif
+
+/* HRTIMER_MODE_ABS started life with a different name */
+#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,21)
+#define HRTIMER_MODE_ABS HRTIMER_ABS
+#endif
+
+/* div64_u64 is fairly new */
+#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,26)
+
+#define div64_u64 kvm_div64_u64
+
+#ifdef CONFIG_64BIT
+
+static inline uint64_t div64_u64(uint64_t dividend, uint64_t divisor)
+{
+	return dividend / divisor;
+}
+
+#else
+
+uint64_t div64_u64(uint64_t dividend, uint64_t divisor);
+
+#endif
+
+#endif
+
+#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,19)
+
+#ifdef RHEL_RELEASE_CODE
+#if RHEL_RELEASE_CODE >= RHEL_RELEASE_VERSION(5,2)
+#define RHEL_BOOL 1
+#endif
+#endif
+
+#ifndef RHEL_BOOL
+
+typedef _Bool bool;
+
+#endif
+
+#endif
+
+/*
+ * PF_VCPU is a Linux 2.6.24 addition
+ */
+
+#include <linux/sched.h>
+
+#ifndef PF_VCPU
+#define PF_VCPU 0
+#endif
+
+/*
+ * smp_call_function_mask() is not defined/exported below 2.6.24
+ */
+
+#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,24)
+
+int kvm_smp_call_function_mask(cpumask_t mask, void (*func) (void
*info),
+			       void *info, int wait);
+
+#define smp_call_function_mask kvm_smp_call_function_mask
+
+#endif
+
+/* empty_zero_page isn't exported in all kernels */
+#include <asm/pgtable.h>
+
+#define empty_zero_page kvm_empty_zero_page
+
+static char empty_zero_page[PAGE_SIZE];
+
+static inline void blahblah(void)
+{
+	(void)empty_zero_page[0];
+}
+
+/* __mmdrop() is not exported before 2.6.25 */
+#include <linux/sched.h>
+
+#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,25)
+
+#define mmdrop(x) do { (void)(x); } while (0)
+#define mmget(x) do { (void)(x); } while (0)
+
+#else
+
+#define mmget(x) do { atomic_inc(x); } while (0)
+
+#endif
+
+/* pagefault_enable(), page_fault_disable() - 2.6.20 */
+#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,20)
+
+#define pagefault_enable()  do {} while (0)
+#define pagefault_disable() do {} while (0)
+
+#endif
+
+/* vm ops ->fault() was introduced in 2.6.23. */
+#include <linux/mm.h>
+
+#ifdef KVM_MAIN
+#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,23)
+
+struct vm_fault {
+	unsigned int flags;
+	pgoff_t pgoff;
+	void __user *virtual_address;
+	struct page *page;
+};
+
+static int kvm_vcpu_fault(struct vm_area_struct *vma, struct vm_fault
*vmf);
+static int kvm_vm_fault(struct vm_area_struct *vma, struct vm_fault
*vmf);
+
+static inline struct page *kvm_nopage_to_fault(
+	int (*fault)(struct vm_area_struct *vma, struct vm_fault *vmf),
+	struct vm_area_struct *vma,
+	unsigned long address,
+	int *type)
+{
+	struct vm_fault vmf;
+	int ret;
+
+	vmf.pgoff = ((address - vma->vm_start) >> PAGE_SHIFT) +
vma->vm_pgoff;
+	vmf.virtual_address = (void __user *)address;
+	ret = fault(vma, &vmf);
+	if (ret)
+		return NOPAGE_SIGBUS;
+	*type = VM_FAULT_MINOR;
+	return vmf.page;
+}
+
+static inline struct page *__kvm_vcpu_fault(struct vm_area_struct *vma,
+					    unsigned long address,
+					    int *type)
+{
+	return kvm_nopage_to_fault(kvm_vcpu_fault, vma, address, type);
+}
+
+static inline struct page *__kvm_vm_fault(struct vm_area_struct *vma,
+					  unsigned long address,
+					  int *type)
+{
+	return kvm_nopage_to_fault(kvm_vm_fault, vma, address, type);
+}
+
+#define VMA_OPS_FAULT(x) nopage
+#define VMA_OPS_FAULT_FUNC(x) __##x
+
+#else
+
+#define VMA_OPS_FAULT(x) x
+#define VMA_OPS_FAULT_FUNC(x) x
+
+#endif
+#endif
+
+/* simple vfs attribute getter signature has changed to add a return
code */
+
+#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,25)
+
+#define MAKE_SIMPLE_ATTRIBUTE_GETTER(x)       \
+	static u64 x(void *v)                 \
+	{				      \
+		u64 ret = 0;		      \
+					      \
+		__##x(v, &ret);		      \
+		return ret;		      \
+	}
+
+#else
+
+#define MAKE_SIMPLE_ATTRIBUTE_GETTER(x)       \
+	static int x(void *v, u64 *val)	      \
+	{				      \
+		return __##x(v, val);	      \
+	}
+
+#endif
+
+/* set_kset_name() is gone in 2.6.25 */
+
+#if LINUX_VERSION_CODE >= KERNEL_VERSION(2,6,25)
+
+#define set_kset_name(x) .name = x
+
+#endif
+
+#if LINUX_VERSION_CODE >= KERNEL_VERSION(2,6,25)
+#ifndef FASTCALL
+#define FASTCALL(x)	x
+#define fastcall
+#endif
+#endif
+
+#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,23)
+
+static unsigned  __attribute__((__used__)) kvm_tsc_khz = 2000000;
+
+#else
+
+#define kvm_tsc_khz tsc_khz
+
+#endif
+
+#if LINUX_VERSION_CODE <= KERNEL_VERSION(2,6,21)
+
+#include <linux/ktime.h>
+#include <linux/hrtimer.h>
+
+#define ktime_get kvm_ktime_get
+
+static inline ktime_t ktime_get(void)
+{
+	struct timespec now;
+
+	ktime_get_ts(&now);
+
+	return timespec_to_ktime(now);
+}
+
+#endif
+
+/* __aligned arrived in 2.6.21 */
+#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,21)
+#define __aligned(x) __attribute__((__aligned__(x)))
+#endif
+
+#include <linux/mm.h>
+
+/* The shrinker API changed in 2.6.23 */
+#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,23)
+
+struct kvm_shrinker {
+	int (*shrink)(int nr_to_scan, gfp_t gfp_mask);
+	int seeks;
+	struct shrinker *kshrinker;
+};
+
+static inline void register_shrinker(struct kvm_shrinker *shrinker)
+{
+	shrinker->kshrinker = set_shrinker(shrinker->seeks,
shrinker->shrink);
+}
+
+static inline void unregister_shrinker(struct kvm_shrinker *shrinker)
+{
+	if (shrinker->kshrinker)
+		remove_shrinker(shrinker->kshrinker);
+}
+
+#define shrinker kvm_shrinker
+
+#endif
+
+/* clocksource */
+#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,18)
+static inline u32 clocksource_khz2mult(u32 khz, u32 shift_constant)
+{
+	/*  khz = cyc/(Million ns)
+	 *  mult/2^shift  = ns/cyc
+	 *  mult = ns/cyc * 2^shift
+	 *  mult = 1Million/khz * 2^shift
+	 *  mult = 1000000 * 2^shift / khz
+	 *  mult = (1000000<<shift) / khz
+	 */
+	u64 tmp = ((u64)1000000) << shift_constant;
+
+	tmp += khz/2; /* round for do_div */
+	do_div(tmp, khz);
+
+	return (u32)tmp;
+}
+#else
+#include <linux/clocksource.h>
+#endif
+
+/* manually export hrtimer_init/start/cancel */
+#include <linux/kallsyms.h>
+extern void (*hrtimer_init_p)(struct hrtimer *timer, clockid_t
which_clock,
+			      enum hrtimer_mode mode);
+extern int (*hrtimer_start_p)(struct hrtimer *timer, ktime_t tim,
+			      const enum hrtimer_mode mode);
+extern int (*hrtimer_cancel_p)(struct hrtimer *timer);
+
+#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,17) &&
defined(CONFIG_KALLSYMS)
+static inline void hrtimer_kallsyms_resolve(void)
+{
+	hrtimer_init_p = (void *) kallsyms_lookup_name("hrtimer_init");
+	BUG_ON(!hrtimer_init_p);
+	hrtimer_start_p = (void *)
kallsyms_lookup_name("hrtimer_start");
+	BUG_ON(!hrtimer_start_p);
+	hrtimer_cancel_p = (void *)
kallsyms_lookup_name("hrtimer_cancel");
+	BUG_ON(!hrtimer_cancel_p);
+}
+#else
+static inline void hrtimer_kallsyms_resolve(void)
+{
+	hrtimer_init_p = hrtimer_init;
+	hrtimer_start_p = hrtimer_start;
+	hrtimer_cancel_p = hrtimer_cancel;
+}
+#endif
+
+/* handle old hrtimer API with data pointer */
+#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,17)
+static inline void hrtimer_data_pointer(struct hrtimer *timer)
+{
+	timer->data = (void *)timer;
+}
+#else
+static inline void hrtimer_data_pointer(struct hrtimer *timer) {}
+#endif
+
+#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,22)
+
+#define ns_to_timespec kvm_ns_to_timespec
+
+struct timespec kvm_ns_to_timespec(const s64 nsec);
+
+#endif
+
diff --git a/kernel/external-module-compat.h
b/kernel/external-module-compat.h
deleted file mode 100644
index 49751fc..0000000
--- a/kernel/external-module-compat.h
+++ /dev/null
@@ -1,759 +0,0 @@
-
-/*
- * Compatibility header for building as an external module.
- */
-
-/*
- * Avoid picking up the kernel's kvm.h in case we have a newer one.
- */
-
-#include <linux/compiler.h>
-#include <linux/version.h>
-#include <linux/string.h>
-#include <linux/kvm.h>
-#include <linux/kvm_para.h>
-#include <linux/cpu.h>
-#include <linux/time.h>
-#include <asm/processor.h>
-#include <linux/hrtimer.h>
-#include <asm/bitops.h>
-#include <asm/msr.h>
-
-/*
- * 2.6.16 does not have GFP_NOWAIT
- */
-
-#include <linux/gfp.h>
-
-#ifndef GFP_NOWAIT
-#define GFP_NOWAIT (GFP_ATOMIC & ~__GFP_HIGH)
-#endif
-
-
-/*
- * kvm profiling support needs 2.6.20
- */
-#include <linux/profile.h>
-
-#ifndef KVM_PROFILING
-#define KVM_PROFILING 1234
-#define prof_on       4321
-#endif
-
-/*
- * smp_call_function_single() is not exported below 2.6.20, and has
different
- * semantics below 2.6.23.  The 'nonatomic' argument was removed in
2.6.27.
- */
-#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,27)
-
-int kvm_smp_call_function_single(int cpu, void (*func)(void *info),
-				 void *info, int wait);
-
-#define smp_call_function_single kvm_smp_call_function_single
-
-#endif
-
-/* on_each_cpu() lost an argument in 2.6.27. */
-#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,27)
-
-#define kvm_on_each_cpu(func, info, wait) on_each_cpu(func, info, 0,
wait)
-
-#else
-
-#define kvm_on_each_cpu(func, info, wait) on_each_cpu(func, info, wait)
-
-#endif
-
-/*
- * The cpu hotplug stubs are broken if !CONFIG_CPU_HOTPLUG
- */
-
-#if LINUX_VERSION_CODE <= KERNEL_VERSION(2,6,15)
-#define DEFINE_MUTEX(a) DECLARE_MUTEX(a)
-#define mutex_lock_interruptible(a) down_interruptible(a)
-#define mutex_unlock(a) up(a)
-#define mutex_lock(a) down(a)
-#define mutex_init(a) init_MUTEX(a)
-#define mutex_trylock(a) down_trylock(a)
-#define mutex semaphore
-#endif
-
-#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,14)
-#ifndef kzalloc
-#define kzalloc(size,flags)			\
-({						\
-	void *__ret = kmalloc(size, flags);	\
-	if (__ret)				\
-		memset(__ret, 0, size);		\
-	__ret;					\
-})
-#endif
-#endif
-
-#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,17)
-#ifndef kmem_cache_zalloc
-#define kmem_cache_zalloc(cache,flags)			  \
-({							  \
-	void *__ret = kmem_cache_alloc(cache, flags);	  \
-	if (__ret)                                        \
-		memset(__ret, 0, kmem_cache_size(cache)); \
-	__ret;                                            \
-})
-#endif
-#endif
-
-#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,21)
-
-#ifndef CONFIG_HOTPLUG_CPU
-#define register_cpu_notifier(nb) (0)
-#endif
-
-#endif
-
-#include <linux/miscdevice.h>
-#ifndef KVM_MINOR
-#define KVM_MINOR 232
-#endif
-
-#include <linux/notifier.h>
-#ifndef CPU_TASKS_FROZEN
-
-#define CPU_TASKS_FROZEN       0x0010
-#define CPU_ONLINE_FROZEN      (CPU_ONLINE | CPU_TASKS_FROZEN)
-#define CPU_UP_PREPARE_FROZEN  (CPU_UP_PREPARE | CPU_TASKS_FROZEN)
-#define CPU_UP_CANCELED_FROZEN (CPU_UP_CANCELED | CPU_TASKS_FROZEN)
-#define CPU_DOWN_PREPARE_FROZEN        (CPU_DOWN_PREPARE |
CPU_TASKS_FROZEN)
-#define CPU_DOWN_FAILED_FROZEN (CPU_DOWN_FAILED | CPU_TASKS_FROZEN)
-#define CPU_DEAD_FROZEN                (CPU_DEAD | CPU_TASKS_FROZEN)
-
-#endif
-
-#ifndef CPU_DYING
-#define CPU_DYING 0x000A
-#define CPU_DYING_FROZEN (CPU_DYING | CPU_TASKS_FROZEN)
-#endif
-
-#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,22)
-
-#ifndef _EFER_SCE
-#define _EFER_SCE		0  /* SYSCALL/SYSRET */
-#endif
-
-#ifndef EFER_SCE
-#define EFER_SCE		(1<<_EFER_SCE)
-#endif
-
-#endif
-
-/*
- * For set_64bit(), which is in a new file asm/cmpxchg.h in newer
kernels.
- */
-#include <asm/system.h>
-
-struct inode;
-#include <linux/anon_inodes.h>
-#define anon_inode_getfd kvm_anon_inode_getfd
-int kvm_init_anon_inodes(void);
-void kvm_exit_anon_inodes(void);
-int anon_inode_getfd(const char *name,
-		     const struct file_operations *fops,
-		     void *priv);
-
-#include <linux/smp.h>
-
-#ifndef X86_CR0_PE
-#define X86_CR0_PE 0x00000001
-#endif
-
-#ifndef X86_CR0_MP
-#define X86_CR0_MP 0x00000002
-#endif
-
-#ifndef X86_CR0_EM
-#define X86_CR0_EM 0x00000004
-#endif
-
-#ifndef X86_CR0_TS
-#define X86_CR0_TS 0x00000008
-#endif
-
-#ifndef X86_CR0_ET
-#define X86_CR0_ET 0x00000010
-#endif
-
-#ifndef X86_CR0_NE
-#define X86_CR0_NE 0x00000020
-#endif
-
-#ifndef X86_CR0_WP
-#define X86_CR0_WP 0x00010000
-#endif
-
-#ifndef X86_CR0_AM
-#define X86_CR0_AM 0x00040000
-#endif
-
-#ifndef X86_CR0_NW
-#define X86_CR0_NW 0x20000000
-#endif
-
-#ifndef X86_CR0_CD
-#define X86_CR0_CD 0x40000000
-#endif
-
-#ifndef X86_CR0_PG
-#define X86_CR0_PG 0x80000000
-#endif
-
-#ifndef X86_CR3_PWT
-#define X86_CR3_PWT 0x00000008
-#endif
-
-#ifndef X86_CR3_PCD
-#define X86_CR3_PCD 0x00000010
-#endif
-
-#ifndef X86_CR4_VMXE
-#define X86_CR4_VMXE 0x00002000
-#endif
-
-#undef X86_CR8_TPR
-#define X86_CR8_TPR 0x0f
-
-/*
- * 2.6.23 removed the cache destructor
- */
-#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,23)
-#  define kmem_cache_create(name, size, align, flags, ctor) \
-	kmem_cache_create(name, size, align, flags, ctor, NULL)
-#endif
-
-/*
- * 2.6.22 does not define set_64bit() under nonpae
- */
-#ifdef CONFIG_X86_32
-
-#include <asm/cmpxchg.h>
-
-static inline void __kvm_set_64bit(u64 *ptr, u64 val)
-{
-	unsigned int low = val;
-	unsigned int high = val >> 32;
-
-	__asm__ __volatile__ (
-		"\n1:\t"
-		"movl (%0), %%eax\n\t"
-		"movl 4(%0), %%edx\n\t"
-		"lock cmpxchg8b (%0)\n\t"
-		"jnz 1b"
-		: /* no outputs */
-		:	"D"(ptr),
-			"b"(low),
-			"c"(high)
-		:	"ax","dx","memory");
-}
-
-#undef  set_64bit
-#define set_64bit __kvm_set_64bit
-
-static inline unsigned long long __kvm_cmpxchg64(volatile void *ptr,
-						 unsigned long long old,
-						 unsigned long long new)
-{
-	unsigned long long prev;
-	__asm__ __volatile__("lock cmpxchg8b %3"
-			     : "=A"(prev)
-			     : "b"((unsigned long)new),
-			       "c"((unsigned long)(new >> 32)),
-			       "m"(*__xg(ptr)),
-			       "0"(old)
-			     : "memory");
-	return prev;
-}
-
-#define kvm_cmpxchg64(ptr,o,n)\
-	((__typeof__(*(ptr)))__kvm_cmpxchg64((ptr),(unsigned long
long)(o),\
-					(unsigned long long)(n)))
-
-#undef cmpxchg64
-#define cmpxchg64(ptr, o, n) kvm_cmpxchg64(ptr, o, n)
-
-#endif
-
-#ifndef CONFIG_PREEMPT_NOTIFIERS
-/*
- * Include sched|preempt.h before defining CONFIG_PREEMPT_NOTIFIERS to
avoid
- * a miscompile.
- */
-#include <linux/sched.h>
-#include <linux/preempt.h>
-#define CONFIG_PREEMPT_NOTIFIERS
-#define CONFIG_PREEMPT_NOTIFIERS_COMPAT
-
-struct preempt_notifier;
-
-struct preempt_ops {
-	void (*sched_in)(struct preempt_notifier *notifier, int cpu);
-	void (*sched_out)(struct preempt_notifier *notifier,
-			  struct task_struct *next);
-};
-
-struct preempt_notifier {
-	struct list_head link;
-	struct task_struct *tsk;
-	struct preempt_ops *ops;
-};
-
-void preempt_notifier_register(struct preempt_notifier *notifier);
-void preempt_notifier_unregister(struct preempt_notifier *notifier);
-
-static inline void preempt_notifier_init(struct preempt_notifier
*notifier,
-				     struct preempt_ops *ops)
-{
-	notifier->ops = ops;
-}
-
-void start_special_insn(void);
-void end_special_insn(void);
-void in_special_section(void);
-void special_reload_dr7(void);
-
-void preempt_notifier_sys_init(void);
-void preempt_notifier_sys_exit(void);
-
-#else
-
-static inline void start_special_insn(void) {}
-static inline void end_special_insn(void) {}
-static inline void in_special_section(void) {}
-static inline void special_reload_dr7(void) {}
-
-static inline void preempt_notifier_sys_init(void) {}
-static inline void preempt_notifier_sys_exit(void) {}
-
-#endif
-
-/* HRTIMER_MODE_ABS started life with a different name */
-#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,21)
-#define HRTIMER_MODE_ABS HRTIMER_ABS
-#endif
-
-/* div64_u64 is fairly new */
-#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,26)
-
-#define div64_u64 kvm_div64_u64
-
-#ifdef CONFIG_64BIT
-
-static inline uint64_t div64_u64(uint64_t dividend, uint64_t divisor)
-{
-	return dividend / divisor;
-}
-
-#else
-
-uint64_t div64_u64(uint64_t dividend, uint64_t divisor);
-
-#endif
-
-#endif
-
-#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,19)
-
-#ifdef RHEL_RELEASE_CODE
-#if RHEL_RELEASE_CODE >= RHEL_RELEASE_VERSION(5,2)
-#define RHEL_BOOL 1
-#endif
-#endif
-
-#ifndef RHEL_BOOL
-
-typedef _Bool bool;
-
-#endif
-
-#endif
-
-/*
- * PF_VCPU is a Linux 2.6.24 addition
- */
-
-#include <linux/sched.h>
-
-#ifndef PF_VCPU
-#define PF_VCPU 0
-#endif
-
-/*
- * smp_call_function_mask() is not defined/exported below 2.6.24
- */
-
-#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,24)
-
-int kvm_smp_call_function_mask(cpumask_t mask, void (*func) (void
*info),
-			       void *info, int wait);
-
-#define smp_call_function_mask kvm_smp_call_function_mask
-
-#endif
-
-/* CONFIG_HAS_IOMEM is apparently fairly new too (2.6.21 for x86_64).
*/
-#ifndef CONFIG_HAS_IOMEM
-#define CONFIG_HAS_IOMEM 1
-#endif
-
-/* empty_zero_page isn't exported in all kernels */
-#include <asm/pgtable.h>
-
-#define empty_zero_page kvm_empty_zero_page
-
-static char empty_zero_page[PAGE_SIZE];
-
-static inline void blahblah(void)
-{
-	(void)empty_zero_page[0];
-}
-
-/* __mmdrop() is not exported before 2.6.25 */
-#include <linux/sched.h>
-
-#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,25)
-
-#define mmdrop(x) do { (void)(x); } while (0)
-#define mmget(x) do { (void)(x); } while (0)
-
-#else
-
-#define mmget(x) do { atomic_inc(x); } while (0)
-
-#endif
-
-/* X86_FEATURE_NX is missing in some x86_64 kernels */
-
-#include <asm/cpufeature.h>
-
-#ifndef X86_FEATURE_NX
-#define X86_FEATURE_NX (1*32+20)
-#endif
-
-#undef true
-#define true 1
-#undef false
-#define false 0
-
-/* EFER_LMA and EFER_LME are missing in pre 2.6.24 i386 kernels */
-#ifndef EFER_LME
-#define _EFER_LME           8  /* Long mode enable */
-#define _EFER_LMA           10 /* Long mode active (read-only) */
-#define EFER_LME            (1<<_EFER_LME)
-#define EFER_LMA            (1<<_EFER_LMA)
-#endif
-
-/* pagefault_enable(), page_fault_disable() - 2.6.20 */
-#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,20)
-
-#define pagefault_enable()  do {} while (0)
-#define pagefault_disable() do {} while (0)
-
-#endif
-
-/* vm ops ->fault() was introduced in 2.6.23. */
-#include <linux/mm.h>
-
-#ifdef KVM_MAIN
-#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,23)
-
-struct vm_fault {
-	unsigned int flags;
-	pgoff_t pgoff;
-	void __user *virtual_address;
-	struct page *page;
-};
-
-static int kvm_vcpu_fault(struct vm_area_struct *vma, struct vm_fault
*vmf);
-static int kvm_vm_fault(struct vm_area_struct *vma, struct vm_fault
*vmf);
-
-static inline struct page *kvm_nopage_to_fault(
-	int (*fault)(struct vm_area_struct *vma, struct vm_fault *vmf),
-	struct vm_area_struct *vma,
-	unsigned long address,
-	int *type)
-{
-	struct vm_fault vmf;
-	int ret;
-
-	vmf.pgoff = ((address - vma->vm_start) >> PAGE_SHIFT) +
vma->vm_pgoff;
-	vmf.virtual_address = (void __user *)address;
-	ret = fault(vma, &vmf);
-	if (ret)
-		return NOPAGE_SIGBUS;
-	*type = VM_FAULT_MINOR;
-	return vmf.page;
-}
-
-static inline struct page *__kvm_vcpu_fault(struct vm_area_struct *vma,
-					    unsigned long address,
-					    int *type)
-{
-	return kvm_nopage_to_fault(kvm_vcpu_fault, vma, address, type);
-}
-
-static inline struct page *__kvm_vm_fault(struct vm_area_struct *vma,
-					  unsigned long address,
-					  int *type)
-{
-	return kvm_nopage_to_fault(kvm_vm_fault, vma, address, type);
-}
-
-#define VMA_OPS_FAULT(x) nopage
-#define VMA_OPS_FAULT_FUNC(x) __##x
-
-#else
-
-#define VMA_OPS_FAULT(x) x
-#define VMA_OPS_FAULT_FUNC(x) x
-
-#endif
-#endif
-
-/* simple vfs attribute getter signature has changed to add a return
code */
-
-#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,25)
-
-#define MAKE_SIMPLE_ATTRIBUTE_GETTER(x)       \
-	static u64 x(void *v)                 \
-	{				      \
-		u64 ret = 0;		      \
-					      \
-		__##x(v, &ret);		      \
-		return ret;		      \
-	}
-
-#else
-
-#define MAKE_SIMPLE_ATTRIBUTE_GETTER(x)       \
-	static int x(void *v, u64 *val)	      \
-	{				      \
-		return __##x(v, val);	      \
-	}
-
-#endif
-
-/* set_kset_name() is gone in 2.6.25 */
-
-#if LINUX_VERSION_CODE >= KERNEL_VERSION(2,6,25)
-
-#define set_kset_name(x) .name = x
-
-#endif
-
-#if LINUX_VERSION_CODE >= KERNEL_VERSION(2,6,25)
-#ifndef FASTCALL
-#define FASTCALL(x)	x
-#define fastcall
-#endif
-#endif
-
-#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,23)
-
-static unsigned  __attribute__((__used__)) kvm_tsc_khz = 2000000;
-
-#else
-
-#define kvm_tsc_khz tsc_khz
-
-#endif
-
-struct kvm_desc_struct {
-	union {
-		struct { unsigned int a, b; };
-		struct {
-			u16 limit0;
-			u16 base0;
-			unsigned base1: 8, type: 4, s: 1, dpl: 2, p: 1;
-			unsigned limit: 4, avl: 1, l: 1, d: 1, g: 1,
base2: 8;
-		};
-
-	};
-} __attribute__((packed));
-
-struct kvm_ldttss_desc64 {
-	u16 limit0;
-	u16 base0;
-	unsigned base1 : 8, type : 5, dpl : 2, p : 1;
-	unsigned limit1 : 4, zero0 : 3, g : 1, base2 : 8;
-	u32 base3;
-	u32 zero1;
-} __attribute__((packed));
-
-struct kvm_desc_ptr {
-	unsigned short size;
-	unsigned long address;
-} __attribute__((packed));
-
-#if LINUX_VERSION_CODE <= KERNEL_VERSION(2,6,21)
-
-#include <linux/ktime.h>
-#include <linux/hrtimer.h>
-
-#define ktime_get kvm_ktime_get
-
-static inline ktime_t ktime_get(void)
-{
-	struct timespec now;
-
-	ktime_get_ts(&now);
-
-	return timespec_to_ktime(now);
-}
-
-#endif
-
-/* __aligned arrived in 2.6.21 */
-#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,21)
-#define __aligned(x) __attribute__((__aligned__(x)))
-#endif
-
-#include <asm/msr.h>
-#ifndef MSR_FS_BASE
-#define MSR_FS_BASE 0xc0000100
-#endif
-#ifndef MSR_GS_BASE
-#define MSR_GS_BASE 0xc0000101
-#endif
-
-#include <linux/mm.h>
-
-/* The shrinker API changed in 2.6.23 */
-#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,23)
-
-struct kvm_shrinker {
-	int (*shrink)(int nr_to_scan, gfp_t gfp_mask);
-	int seeks;
-	struct shrinker *kshrinker;
-};
-
-static inline void register_shrinker(struct kvm_shrinker *shrinker)
-{
-	shrinker->kshrinker = set_shrinker(shrinker->seeks,
shrinker->shrink);
-}
-
-static inline void unregister_shrinker(struct kvm_shrinker *shrinker)
-{
-	if (shrinker->kshrinker)
-		remove_shrinker(shrinker->kshrinker);
-}
-
-#define shrinker kvm_shrinker
-
-#endif
-
-/* undefine lapic */
-#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,18)
-
-#undef lapic
-
-#endif
-
-/* clocksource */
-#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,18)
-static inline u32 clocksource_khz2mult(u32 khz, u32 shift_constant)
-{
-	/*  khz = cyc/(Million ns)
-	 *  mult/2^shift  = ns/cyc
-	 *  mult = ns/cyc * 2^shift
-	 *  mult = 1Million/khz * 2^shift
-	 *  mult = 1000000 * 2^shift / khz
-	 *  mult = (1000000<<shift) / khz
-	 */
-	u64 tmp = ((u64)1000000) << shift_constant;
-
-	tmp += khz/2; /* round for do_div */
-	do_div(tmp, khz);
-
-	return (u32)tmp;
-}
-#else
-#include <linux/clocksource.h>
-#endif
-
-/* manually export hrtimer_init/start/cancel */
-#include <linux/kallsyms.h>
-extern void (*hrtimer_init_p)(struct hrtimer *timer, clockid_t
which_clock,
-			      enum hrtimer_mode mode);
-extern int (*hrtimer_start_p)(struct hrtimer *timer, ktime_t tim,
-			      const enum hrtimer_mode mode);
-extern int (*hrtimer_cancel_p)(struct hrtimer *timer);
-
-#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,17) &&
defined(CONFIG_KALLSYMS)
-static inline void hrtimer_kallsyms_resolve(void)
-{
-	hrtimer_init_p = (void *) kallsyms_lookup_name("hrtimer_init");
-	BUG_ON(!hrtimer_init_p);
-	hrtimer_start_p = (void *)
kallsyms_lookup_name("hrtimer_start");
-	BUG_ON(!hrtimer_start_p);
-	hrtimer_cancel_p = (void *)
kallsyms_lookup_name("hrtimer_cancel");
-	BUG_ON(!hrtimer_cancel_p);
-}
-#else
-static inline void hrtimer_kallsyms_resolve(void)
-{
-	hrtimer_init_p = hrtimer_init;
-	hrtimer_start_p = hrtimer_start;
-	hrtimer_cancel_p = hrtimer_cancel;
-}
-#endif
-
-/* handle old hrtimer API with data pointer */
-#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,17)
-static inline void hrtimer_data_pointer(struct hrtimer *timer)
-{
-	timer->data = (void *)timer;
-}
-#else
-static inline void hrtimer_data_pointer(struct hrtimer *timer) {}
-#endif
-
-#include <asm/hw_irq.h>
-#ifndef NMI_VECTOR
-#define NMI_VECTOR 2
-#endif
-
-#ifndef MSR_MTRRcap
-#define MSR_MTRRcap            0x0fe
-#define MSR_MTRRfix64K_00000   0x250
-#define MSR_MTRRfix16K_80000   0x258
-#define MSR_MTRRfix16K_A0000   0x259
-#define MSR_MTRRfix4K_C0000    0x268
-#define MSR_MTRRfix4K_C8000    0x269
-#define MSR_MTRRfix4K_D0000    0x26a
-#define MSR_MTRRfix4K_D8000    0x26b
-#define MSR_MTRRfix4K_E0000    0x26c
-#define MSR_MTRRfix4K_E8000    0x26d
-#define MSR_MTRRfix4K_F0000    0x26e
-#define MSR_MTRRfix4K_F8000    0x26f
-#define MSR_MTRRdefType        0x2ff
-#endif
-
-#ifndef MSR_IA32_CR_PAT
-#define MSR_IA32_CR_PAT        0x00000277
-#endif
-
-#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,22)
-
-#define ns_to_timespec kvm_ns_to_timespec
-
-struct timespec kvm_ns_to_timespec(const s64 nsec);
-
-#endif
-
-/* Define DEBUGCTLMSR bits */
-#ifndef DEBUGCTLMSR_LBR
-
-#define _DEBUGCTLMSR_LBR	0 /* last branch recording */
-#define _DEBUGCTLMSR_BTF	1 /* single-step on branches */
-
-#define DEBUGCTLMSR_LBR		(1UL << _DEBUGCTLMSR_LBR)
-#define DEBUGCTLMSR_BTF		(1UL << _DEBUGCTLMSR_BTF)
-
-#endif
diff --git a/kernel/debug.h b/kernel/x86/debug.h
similarity index 100%
rename from kernel/debug.h
rename to kernel/x86/debug.h
diff --git a/kernel/external-module-compat.c
b/kernel/x86/external-module-compat.c
similarity index 100%
rename from kernel/external-module-compat.c
rename to kernel/x86/external-module-compat.c
diff --git a/kernel/x86/external-module-compat.h
b/kernel/x86/external-module-compat.h
new file mode 100644
index 0000000..ce93551
--- /dev/null
+++ b/kernel/x86/external-module-compat.h
@@ -0,0 +1,291 @@
+
+/*
+ * Compatibility header for building as an external module.
+ */
+
+#include <linux/compiler.h>
+#include <linux/version.h>
+
+#include "../external-module-compat-comm.h"
+
+#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,22)
+
+#ifndef _EFER_SCE
+#define _EFER_SCE		0  /* SYSCALL/SYSRET */
+#endif
+
+#ifndef EFER_SCE
+#define EFER_SCE		(1<<_EFER_SCE)
+#endif
+
+#endif
+
+#include <linux/smp.h>
+
+#ifndef X86_CR0_PE
+#define X86_CR0_PE 0x00000001
+#endif
+
+#ifndef X86_CR0_MP
+#define X86_CR0_MP 0x00000002
+#endif
+
+#ifndef X86_CR0_EM
+#define X86_CR0_EM 0x00000004
+#endif
+
+#ifndef X86_CR0_TS
+#define X86_CR0_TS 0x00000008
+#endif
+
+#ifndef X86_CR0_ET
+#define X86_CR0_ET 0x00000010
+#endif
+
+#ifndef X86_CR0_NE
+#define X86_CR0_NE 0x00000020
+#endif
+
+#ifndef X86_CR0_WP
+#define X86_CR0_WP 0x00010000
+#endif
+
+#ifndef X86_CR0_AM
+#define X86_CR0_AM 0x00040000
+#endif
+
+#ifndef X86_CR0_NW
+#define X86_CR0_NW 0x20000000
+#endif
+
+#ifndef X86_CR0_CD
+#define X86_CR0_CD 0x40000000
+#endif
+
+#ifndef X86_CR0_PG
+#define X86_CR0_PG 0x80000000
+#endif
+
+#ifndef X86_CR3_PWT
+#define X86_CR3_PWT 0x00000008
+#endif
+
+#ifndef X86_CR3_PCD
+#define X86_CR3_PCD 0x00000010
+#endif
+
+#ifndef X86_CR4_VMXE
+#define X86_CR4_VMXE 0x00002000
+#endif
+
+#undef X86_CR8_TPR
+#define X86_CR8_TPR 0x0f
+
+/*
+ * 2.6.22 does not define set_64bit() under nonpae
+ */
+#ifdef CONFIG_X86_32
+
+#include <asm/cmpxchg.h>
+
+static inline void __kvm_set_64bit(u64 *ptr, u64 val)
+{
+	unsigned int low = val;
+	unsigned int high = val >> 32;
+
+	__asm__ __volatile__ (
+		"\n1:\t"
+		"movl (%0), %%eax\n\t"
+		"movl 4(%0), %%edx\n\t"
+		"lock cmpxchg8b (%0)\n\t"
+		"jnz 1b"
+		: /* no outputs */
+		:	"D"(ptr),
+			"b"(low),
+			"c"(high)
+		:	"ax","dx","memory");
+}
+
+#undef  set_64bit
+#define set_64bit __kvm_set_64bit
+
+static inline unsigned long long __kvm_cmpxchg64(volatile void *ptr,
+						 unsigned long long old,
+						 unsigned long long new)
+{
+	unsigned long long prev;
+	__asm__ __volatile__("lock cmpxchg8b %3"
+			     : "=A"(prev)
+			     : "b"((unsigned long)new),
+			       "c"((unsigned long)(new >> 32)),
+			       "m"(*__xg(ptr)),
+			       "0"(old)
+			     : "memory");
+	return prev;
+}
+
+#define kvm_cmpxchg64(ptr,o,n)\
+	((__typeof__(*(ptr)))__kvm_cmpxchg64((ptr),(unsigned long
long)(o),\
+					(unsigned long long)(n)))
+
+#undef cmpxchg64
+#define cmpxchg64(ptr, o, n) kvm_cmpxchg64(ptr, o, n)
+
+#endif
+
+#ifndef CONFIG_PREEMPT_NOTIFIERS
+/*
+ * Include sched|preempt.h before defining CONFIG_PREEMPT_NOTIFIERS to
avoid
+ * a miscompile.
+ */
+#include <linux/sched.h>
+#include <linux/preempt.h>
+#define CONFIG_PREEMPT_NOTIFIERS
+#define CONFIG_PREEMPT_NOTIFIERS_COMPAT
+
+struct preempt_notifier;
+
+struct preempt_ops {
+	void (*sched_in)(struct preempt_notifier *notifier, int cpu);
+	void (*sched_out)(struct preempt_notifier *notifier,
+			  struct task_struct *next);
+};
+
+struct preempt_notifier {
+	struct list_head link;
+	struct task_struct *tsk;
+	struct preempt_ops *ops;
+};
+
+void preempt_notifier_register(struct preempt_notifier *notifier);
+void preempt_notifier_unregister(struct preempt_notifier *notifier);
+
+static inline void preempt_notifier_init(struct preempt_notifier
*notifier,
+				     struct preempt_ops *ops)
+{
+	notifier->ops = ops;
+}
+
+void start_special_insn(void);
+void end_special_insn(void);
+void in_special_section(void);
+void special_reload_dr7(void);
+
+void preempt_notifier_sys_init(void);
+void preempt_notifier_sys_exit(void);
+
+#else
+
+static inline void start_special_insn(void) {}
+static inline void end_special_insn(void) {}
+static inline void in_special_section(void) {}
+static inline void special_reload_dr7(void) {}
+
+static inline void preempt_notifier_sys_init(void) {}
+static inline void preempt_notifier_sys_exit(void) {}
+
+#endif
+
+/* CONFIG_HAS_IOMEM is apparently fairly new too (2.6.21 for x86_64).
*/
+#ifndef CONFIG_HAS_IOMEM
+#define CONFIG_HAS_IOMEM 1
+#endif
+
+/* X86_FEATURE_NX is missing in some x86_64 kernels */
+
+#include <asm/cpufeature.h>
+
+#ifndef X86_FEATURE_NX
+#define X86_FEATURE_NX (1*32+20)
+#endif
+
+#undef true
+#define true 1
+#undef false
+#define false 0
+
+/* EFER_LMA and EFER_LME are missing in pre 2.6.24 i386 kernels */
+#ifndef EFER_LME
+#define _EFER_LME           8  /* Long mode enable */
+#define _EFER_LMA           10 /* Long mode active (read-only) */
+#define EFER_LME            (1<<_EFER_LME)
+#define EFER_LMA            (1<<_EFER_LMA)
+#endif
+
+struct kvm_desc_struct {
+	union {
+		struct { unsigned int a, b; };
+		struct {
+			u16 limit0;
+			u16 base0;
+			unsigned base1: 8, type: 4, s: 1, dpl: 2, p: 1;
+			unsigned limit: 4, avl: 1, l: 1, d: 1, g: 1,
base2: 8;
+		};
+
+	};
+} __attribute__((packed));
+
+struct kvm_ldttss_desc64 {
+	u16 limit0;
+	u16 base0;
+	unsigned base1 : 8, type : 5, dpl : 2, p : 1;
+	unsigned limit1 : 4, zero0 : 3, g : 1, base2 : 8;
+	u32 base3;
+	u32 zero1;
+} __attribute__((packed));
+
+struct kvm_desc_ptr {
+	unsigned short size;
+	unsigned long address;
+} __attribute__((packed));
+
+#include <asm/msr.h>
+#ifndef MSR_FS_BASE
+#define MSR_FS_BASE 0xc0000100
+#endif
+#ifndef MSR_GS_BASE
+#define MSR_GS_BASE 0xc0000101
+#endif
+
+/* undefine lapic */
+#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,18)
+
+#undef lapic
+
+#endif
+
+#include <asm/hw_irq.h>
+#ifndef NMI_VECTOR
+#define NMI_VECTOR 2
+#endif
+
+#ifndef MSR_MTRRcap
+#define MSR_MTRRcap            0x0fe
+#define MSR_MTRRfix64K_00000   0x250
+#define MSR_MTRRfix16K_80000   0x258
+#define MSR_MTRRfix16K_A0000   0x259
+#define MSR_MTRRfix4K_C0000    0x268
+#define MSR_MTRRfix4K_C8000    0x269
+#define MSR_MTRRfix4K_D0000    0x26a
+#define MSR_MTRRfix4K_D8000    0x26b
+#define MSR_MTRRfix4K_E0000    0x26c
+#define MSR_MTRRfix4K_E8000    0x26d
+#define MSR_MTRRfix4K_F0000    0x26e
+#define MSR_MTRRfix4K_F8000    0x26f
+#define MSR_MTRRdefType        0x2ff
+#endif
+
+#ifndef MSR_IA32_CR_PAT
+#define MSR_IA32_CR_PAT        0x00000277
+#endif
+
+/* Define DEBUGCTLMSR bits */
+#ifndef DEBUGCTLMSR_LBR
+
+#define _DEBUGCTLMSR_LBR	0 /* last branch recording */
+#define _DEBUGCTLMSR_BTF	1 /* single-step on branches */
+
+#define DEBUGCTLMSR_LBR		(1UL << _DEBUGCTLMSR_LBR)
+#define DEBUGCTLMSR_BTF		(1UL << _DEBUGCTLMSR_BTF)
+
+#endif
diff --git a/kernel/preempt.c b/kernel/x86/preempt.c
similarity index 100%
rename from kernel/preempt.c
rename to kernel/x86/preempt.c
diff --git a/kernel/vmx-debug.c b/kernel/x86/vmx-debug.c
similarity index 100%
rename from kernel/vmx-debug.c
rename to kernel/x86/vmx-debug.c
-- 
1.5.6.2

^ permalink raw reply related	[flat|nested] 24+ messages in thread

* [PATCH 01/04]Create x86 directory to hold x86-specific files.
@ 2008-07-25  9:26   ` Zhang, Xiantao
  0 siblings, 0 replies; 24+ messages in thread
From: Zhang, Xiantao @ 2008-07-25  9:26 UTC (permalink / raw)
  To: kvm-ia64, kvm; +Cc: avi

>From c8c8c5a6b3ff943cb645f9ac4ef169490f14cd08 Mon Sep 17 00:00:00 2001
From: Zhang Xiantao <xiantao.zhang@intel.com>
Date: Fri, 25 Jul 2008 16:50:30 +0800
Subject: [PATCH] KVM: external module: Moving x86-speicif files to x86
directory.

Create x86 directory to hold x86-specific files.
Signed-off-by: Zhang Xiantao <xiantao.zhang@intel.com>
---
 kernel/Kbuild                             |    8 +-
 kernel/Makefile                           |    2 +-
 kernel/external-module-compat-comm.h      |  475 ++++++++++++++++++
 kernel/external-module-compat.h           |  759
-----------------------------
 kernel/{ => x86}/debug.h                  |    0 
 kernel/{ => x86}/external-module-compat.c |    0 
 kernel/x86/external-module-compat.h       |  291 +++++++++++
 kernel/{ => x86}/preempt.c                |    0 
 kernel/{ => x86}/vmx-debug.c              |    0 
 9 files changed, 771 insertions(+), 764 deletions(-)
 create mode 100644 kernel/external-module-compat-comm.h
 delete mode 100644 kernel/external-module-compat.h
 rename kernel/{ => x86}/debug.h (100%)
 rename kernel/{ => x86}/external-module-compat.c (100%)
 create mode 100644 kernel/x86/external-module-compat.h
 rename kernel/{ => x86}/preempt.c (100%)
 rename kernel/{ => x86}/vmx-debug.c (100%)

diff --git a/kernel/Kbuild b/kernel/Kbuild
index 676f28d..cb63582 100644
--- a/kernel/Kbuild
+++ b/kernel/Kbuild
@@ -1,11 +1,11 @@
 obj-m := kvm.o kvm-intel.o kvm-amd.o
 kvm-objs := kvm_main.o x86.o mmu.o x86_emulate.o anon_inodes.o irq.o
i8259.o \
-	 lapic.o ioapic.o preempt.o i8254.o coalesced_mmio.o \
-	 external-module-compat.o
+	 lapic.o ioapic.o x86/preempt.o i8254.o coalesced_mmio.o \
+	 x86/external-module-compat.o
 ifeq ($(CONFIG_KVM_TRACE),y)
 kvm-objs += kvm_trace.o
 endif
-kvm-intel-objs := vmx.o vmx-debug.o external-module-compat.o
-kvm-amd-objs := svm.o external-module-compat.o
+kvm-intel-objs := vmx.o x86/vmx-debug.o x86/external-module-compat.o
+kvm-amd-objs := svm.o x86/external-module-compat.o
 
 CFLAGS_kvm_main.o = -DKVM_MAIN
diff --git a/kernel/Makefile b/kernel/Makefile
index 69b4981..c2c666e 100644
--- a/kernel/Makefile
+++ b/kernel/Makefile
@@ -35,7 +35,7 @@ all::
 	$(MAKE) -C $(KERNELDIR) M=`pwd` \
 		LINUXINCLUDE="-I`pwd`/include -Iinclude
-I`pwd`/include-compat \
 		-include include/linux/autoconf.h \
-		-include `pwd`/external-module-compat.h"
+		-include `pwd`/$(ARCH_DIR)/external-module-compat.h"
 		"$$@"
 
 sync: header-sync source-sync
diff --git a/kernel/external-module-compat-comm.h
b/kernel/external-module-compat-comm.h
new file mode 100644
index 0000000..d4bf395
--- /dev/null
+++ b/kernel/external-module-compat-comm.h
@@ -0,0 +1,475 @@
+
+/*
+ * Compatibility header for building as an external module.
+ */
+
+/*
+ * Avoid picking up the kernel's kvm.h in case we have a newer one.
+ */
+
+#include <linux/compiler.h>
+#include <linux/version.h>
+#include <linux/string.h>
+#include <linux/kvm.h>
+#include <linux/kvm_para.h>
+#include <linux/cpu.h>
+#include <linux/time.h>
+#include <asm/processor.h>
+#include <linux/hrtimer.h>
+#include <asm/bitops.h>
+#include <asm/msr.h>
+
+/*
+ * 2.6.16 does not have GFP_NOWAIT
+ */
+
+#include <linux/gfp.h>
+
+#ifndef GFP_NOWAIT
+#define GFP_NOWAIT (GFP_ATOMIC & ~__GFP_HIGH)
+#endif
+
+
+/*
+ * kvm profiling support needs 2.6.20
+ */
+#include <linux/profile.h>
+
+#ifndef KVM_PROFILING
+#define KVM_PROFILING 1234
+#define prof_on       4321
+#endif
+
+/*
+ * smp_call_function_single() is not exported below 2.6.20, and has
different
+ * semantics below 2.6.23.  The 'nonatomic' argument was removed in
2.6.27.
+ */
+#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,27)
+
+int kvm_smp_call_function_single(int cpu, void (*func)(void *info),
+				 void *info, int wait);
+
+#define smp_call_function_single kvm_smp_call_function_single
+
+#endif
+
+/* on_each_cpu() lost an argument in 2.6.27. */
+#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,27)
+
+#define kvm_on_each_cpu(func, info, wait) on_each_cpu(func, info, 0,
wait)
+
+#else
+
+#define kvm_on_each_cpu(func, info, wait) on_each_cpu(func, info, wait)
+
+#endif
+
+/*
+ * The cpu hotplug stubs are broken if !CONFIG_CPU_HOTPLUG
+ */
+
+#if LINUX_VERSION_CODE <= KERNEL_VERSION(2,6,15)
+#define DEFINE_MUTEX(a) DECLARE_MUTEX(a)
+#define mutex_lock_interruptible(a) down_interruptible(a)
+#define mutex_unlock(a) up(a)
+#define mutex_lock(a) down(a)
+#define mutex_init(a) init_MUTEX(a)
+#define mutex_trylock(a) down_trylock(a)
+#define mutex semaphore
+#endif
+
+#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,14)
+#ifndef kzalloc
+#define kzalloc(size,flags)			\
+({						\
+	void *__ret = kmalloc(size, flags);	\
+	if (__ret)				\
+		memset(__ret, 0, size);		\
+	__ret;					\
+})
+#endif
+#endif
+
+#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,17)
+#ifndef kmem_cache_zalloc
+#define kmem_cache_zalloc(cache,flags)			  \
+({							  \
+	void *__ret = kmem_cache_alloc(cache, flags);	  \
+	if (__ret)                                        \
+		memset(__ret, 0, kmem_cache_size(cache)); \
+	__ret;                                            \
+})
+#endif
+#endif
+
+#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,21)
+
+#ifndef CONFIG_HOTPLUG_CPU
+#define register_cpu_notifier(nb) (0)
+#endif
+
+#endif
+
+#include <linux/miscdevice.h>
+#ifndef KVM_MINOR
+#define KVM_MINOR 232
+#endif
+
+#include <linux/notifier.h>
+#ifndef CPU_TASKS_FROZEN
+
+#define CPU_TASKS_FROZEN       0x0010
+#define CPU_ONLINE_FROZEN      (CPU_ONLINE | CPU_TASKS_FROZEN)
+#define CPU_UP_PREPARE_FROZEN  (CPU_UP_PREPARE | CPU_TASKS_FROZEN)
+#define CPU_UP_CANCELED_FROZEN (CPU_UP_CANCELED | CPU_TASKS_FROZEN)
+#define CPU_DOWN_PREPARE_FROZEN        (CPU_DOWN_PREPARE |
CPU_TASKS_FROZEN)
+#define CPU_DOWN_FAILED_FROZEN (CPU_DOWN_FAILED | CPU_TASKS_FROZEN)
+#define CPU_DEAD_FROZEN                (CPU_DEAD | CPU_TASKS_FROZEN)
+
+#endif
+
+#ifndef CPU_DYING
+#define CPU_DYING 0x000A
+#define CPU_DYING_FROZEN (CPU_DYING | CPU_TASKS_FROZEN)
+#endif
+
+#include <asm/system.h>
+
+struct inode;
+#include <linux/anon_inodes.h>
+#define anon_inode_getfd kvm_anon_inode_getfd
+int kvm_init_anon_inodes(void);
+void kvm_exit_anon_inodes(void);
+int anon_inode_getfd(const char *name,
+		     const struct file_operations *fops,
+		     void *priv);
+
+/*
+ * 2.6.23 removed the cache destructor
+ */
+#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,23)
+#  define kmem_cache_create(name, size, align, flags, ctor) \
+	kmem_cache_create(name, size, align, flags, ctor, NULL)
+#endif
+
+/* HRTIMER_MODE_ABS started life with a different name */
+#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,21)
+#define HRTIMER_MODE_ABS HRTIMER_ABS
+#endif
+
+/* div64_u64 is fairly new */
+#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,26)
+
+#define div64_u64 kvm_div64_u64
+
+#ifdef CONFIG_64BIT
+
+static inline uint64_t div64_u64(uint64_t dividend, uint64_t divisor)
+{
+	return dividend / divisor;
+}
+
+#else
+
+uint64_t div64_u64(uint64_t dividend, uint64_t divisor);
+
+#endif
+
+#endif
+
+#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,19)
+
+#ifdef RHEL_RELEASE_CODE
+#if RHEL_RELEASE_CODE >= RHEL_RELEASE_VERSION(5,2)
+#define RHEL_BOOL 1
+#endif
+#endif
+
+#ifndef RHEL_BOOL
+
+typedef _Bool bool;
+
+#endif
+
+#endif
+
+/*
+ * PF_VCPU is a Linux 2.6.24 addition
+ */
+
+#include <linux/sched.h>
+
+#ifndef PF_VCPU
+#define PF_VCPU 0
+#endif
+
+/*
+ * smp_call_function_mask() is not defined/exported below 2.6.24
+ */
+
+#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,24)
+
+int kvm_smp_call_function_mask(cpumask_t mask, void (*func) (void
*info),
+			       void *info, int wait);
+
+#define smp_call_function_mask kvm_smp_call_function_mask
+
+#endif
+
+/* empty_zero_page isn't exported in all kernels */
+#include <asm/pgtable.h>
+
+#define empty_zero_page kvm_empty_zero_page
+
+static char empty_zero_page[PAGE_SIZE];
+
+static inline void blahblah(void)
+{
+	(void)empty_zero_page[0];
+}
+
+/* __mmdrop() is not exported before 2.6.25 */
+#include <linux/sched.h>
+
+#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,25)
+
+#define mmdrop(x) do { (void)(x); } while (0)
+#define mmget(x) do { (void)(x); } while (0)
+
+#else
+
+#define mmget(x) do { atomic_inc(x); } while (0)
+
+#endif
+
+/* pagefault_enable(), page_fault_disable() - 2.6.20 */
+#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,20)
+
+#define pagefault_enable()  do {} while (0)
+#define pagefault_disable() do {} while (0)
+
+#endif
+
+/* vm ops ->fault() was introduced in 2.6.23. */
+#include <linux/mm.h>
+
+#ifdef KVM_MAIN
+#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,23)
+
+struct vm_fault {
+	unsigned int flags;
+	pgoff_t pgoff;
+	void __user *virtual_address;
+	struct page *page;
+};
+
+static int kvm_vcpu_fault(struct vm_area_struct *vma, struct vm_fault
*vmf);
+static int kvm_vm_fault(struct vm_area_struct *vma, struct vm_fault
*vmf);
+
+static inline struct page *kvm_nopage_to_fault(
+	int (*fault)(struct vm_area_struct *vma, struct vm_fault *vmf),
+	struct vm_area_struct *vma,
+	unsigned long address,
+	int *type)
+{
+	struct vm_fault vmf;
+	int ret;
+
+	vmf.pgoff = ((address - vma->vm_start) >> PAGE_SHIFT) +
vma->vm_pgoff;
+	vmf.virtual_address = (void __user *)address;
+	ret = fault(vma, &vmf);
+	if (ret)
+		return NOPAGE_SIGBUS;
+	*type = VM_FAULT_MINOR;
+	return vmf.page;
+}
+
+static inline struct page *__kvm_vcpu_fault(struct vm_area_struct *vma,
+					    unsigned long address,
+					    int *type)
+{
+	return kvm_nopage_to_fault(kvm_vcpu_fault, vma, address, type);
+}
+
+static inline struct page *__kvm_vm_fault(struct vm_area_struct *vma,
+					  unsigned long address,
+					  int *type)
+{
+	return kvm_nopage_to_fault(kvm_vm_fault, vma, address, type);
+}
+
+#define VMA_OPS_FAULT(x) nopage
+#define VMA_OPS_FAULT_FUNC(x) __##x
+
+#else
+
+#define VMA_OPS_FAULT(x) x
+#define VMA_OPS_FAULT_FUNC(x) x
+
+#endif
+#endif
+
+/* simple vfs attribute getter signature has changed to add a return
code */
+
+#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,25)
+
+#define MAKE_SIMPLE_ATTRIBUTE_GETTER(x)       \
+	static u64 x(void *v)                 \
+	{				      \
+		u64 ret = 0;		      \
+					      \
+		__##x(v, &ret);		      \
+		return ret;		      \
+	}
+
+#else
+
+#define MAKE_SIMPLE_ATTRIBUTE_GETTER(x)       \
+	static int x(void *v, u64 *val)	      \
+	{				      \
+		return __##x(v, val);	      \
+	}
+
+#endif
+
+/* set_kset_name() is gone in 2.6.25 */
+
+#if LINUX_VERSION_CODE >= KERNEL_VERSION(2,6,25)
+
+#define set_kset_name(x) .name = x
+
+#endif
+
+#if LINUX_VERSION_CODE >= KERNEL_VERSION(2,6,25)
+#ifndef FASTCALL
+#define FASTCALL(x)	x
+#define fastcall
+#endif
+#endif
+
+#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,23)
+
+static unsigned  __attribute__((__used__)) kvm_tsc_khz = 2000000;
+
+#else
+
+#define kvm_tsc_khz tsc_khz
+
+#endif
+
+#if LINUX_VERSION_CODE <= KERNEL_VERSION(2,6,21)
+
+#include <linux/ktime.h>
+#include <linux/hrtimer.h>
+
+#define ktime_get kvm_ktime_get
+
+static inline ktime_t ktime_get(void)
+{
+	struct timespec now;
+
+	ktime_get_ts(&now);
+
+	return timespec_to_ktime(now);
+}
+
+#endif
+
+/* __aligned arrived in 2.6.21 */
+#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,21)
+#define __aligned(x) __attribute__((__aligned__(x)))
+#endif
+
+#include <linux/mm.h>
+
+/* The shrinker API changed in 2.6.23 */
+#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,23)
+
+struct kvm_shrinker {
+	int (*shrink)(int nr_to_scan, gfp_t gfp_mask);
+	int seeks;
+	struct shrinker *kshrinker;
+};
+
+static inline void register_shrinker(struct kvm_shrinker *shrinker)
+{
+	shrinker->kshrinker = set_shrinker(shrinker->seeks,
shrinker->shrink);
+}
+
+static inline void unregister_shrinker(struct kvm_shrinker *shrinker)
+{
+	if (shrinker->kshrinker)
+		remove_shrinker(shrinker->kshrinker);
+}
+
+#define shrinker kvm_shrinker
+
+#endif
+
+/* clocksource */
+#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,18)
+static inline u32 clocksource_khz2mult(u32 khz, u32 shift_constant)
+{
+	/*  khz = cyc/(Million ns)
+	 *  mult/2^shift  = ns/cyc
+	 *  mult = ns/cyc * 2^shift
+	 *  mult = 1Million/khz * 2^shift
+	 *  mult = 1000000 * 2^shift / khz
+	 *  mult = (1000000<<shift) / khz
+	 */
+	u64 tmp = ((u64)1000000) << shift_constant;
+
+	tmp += khz/2; /* round for do_div */
+	do_div(tmp, khz);
+
+	return (u32)tmp;
+}
+#else
+#include <linux/clocksource.h>
+#endif
+
+/* manually export hrtimer_init/start/cancel */
+#include <linux/kallsyms.h>
+extern void (*hrtimer_init_p)(struct hrtimer *timer, clockid_t
which_clock,
+			      enum hrtimer_mode mode);
+extern int (*hrtimer_start_p)(struct hrtimer *timer, ktime_t tim,
+			      const enum hrtimer_mode mode);
+extern int (*hrtimer_cancel_p)(struct hrtimer *timer);
+
+#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,17) &&
defined(CONFIG_KALLSYMS)
+static inline void hrtimer_kallsyms_resolve(void)
+{
+	hrtimer_init_p = (void *) kallsyms_lookup_name("hrtimer_init");
+	BUG_ON(!hrtimer_init_p);
+	hrtimer_start_p = (void *)
kallsyms_lookup_name("hrtimer_start");
+	BUG_ON(!hrtimer_start_p);
+	hrtimer_cancel_p = (void *)
kallsyms_lookup_name("hrtimer_cancel");
+	BUG_ON(!hrtimer_cancel_p);
+}
+#else
+static inline void hrtimer_kallsyms_resolve(void)
+{
+	hrtimer_init_p = hrtimer_init;
+	hrtimer_start_p = hrtimer_start;
+	hrtimer_cancel_p = hrtimer_cancel;
+}
+#endif
+
+/* handle old hrtimer API with data pointer */
+#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,17)
+static inline void hrtimer_data_pointer(struct hrtimer *timer)
+{
+	timer->data = (void *)timer;
+}
+#else
+static inline void hrtimer_data_pointer(struct hrtimer *timer) {}
+#endif
+
+#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,22)
+
+#define ns_to_timespec kvm_ns_to_timespec
+
+struct timespec kvm_ns_to_timespec(const s64 nsec);
+
+#endif
+
diff --git a/kernel/external-module-compat.h
b/kernel/external-module-compat.h
deleted file mode 100644
index 49751fc..0000000
--- a/kernel/external-module-compat.h
+++ /dev/null
@@ -1,759 +0,0 @@
-
-/*
- * Compatibility header for building as an external module.
- */
-
-/*
- * Avoid picking up the kernel's kvm.h in case we have a newer one.
- */
-
-#include <linux/compiler.h>
-#include <linux/version.h>
-#include <linux/string.h>
-#include <linux/kvm.h>
-#include <linux/kvm_para.h>
-#include <linux/cpu.h>
-#include <linux/time.h>
-#include <asm/processor.h>
-#include <linux/hrtimer.h>
-#include <asm/bitops.h>
-#include <asm/msr.h>
-
-/*
- * 2.6.16 does not have GFP_NOWAIT
- */
-
-#include <linux/gfp.h>
-
-#ifndef GFP_NOWAIT
-#define GFP_NOWAIT (GFP_ATOMIC & ~__GFP_HIGH)
-#endif
-
-
-/*
- * kvm profiling support needs 2.6.20
- */
-#include <linux/profile.h>
-
-#ifndef KVM_PROFILING
-#define KVM_PROFILING 1234
-#define prof_on       4321
-#endif
-
-/*
- * smp_call_function_single() is not exported below 2.6.20, and has
different
- * semantics below 2.6.23.  The 'nonatomic' argument was removed in
2.6.27.
- */
-#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,27)
-
-int kvm_smp_call_function_single(int cpu, void (*func)(void *info),
-				 void *info, int wait);
-
-#define smp_call_function_single kvm_smp_call_function_single
-
-#endif
-
-/* on_each_cpu() lost an argument in 2.6.27. */
-#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,27)
-
-#define kvm_on_each_cpu(func, info, wait) on_each_cpu(func, info, 0,
wait)
-
-#else
-
-#define kvm_on_each_cpu(func, info, wait) on_each_cpu(func, info, wait)
-
-#endif
-
-/*
- * The cpu hotplug stubs are broken if !CONFIG_CPU_HOTPLUG
- */
-
-#if LINUX_VERSION_CODE <= KERNEL_VERSION(2,6,15)
-#define DEFINE_MUTEX(a) DECLARE_MUTEX(a)
-#define mutex_lock_interruptible(a) down_interruptible(a)
-#define mutex_unlock(a) up(a)
-#define mutex_lock(a) down(a)
-#define mutex_init(a) init_MUTEX(a)
-#define mutex_trylock(a) down_trylock(a)
-#define mutex semaphore
-#endif
-
-#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,14)
-#ifndef kzalloc
-#define kzalloc(size,flags)			\
-({						\
-	void *__ret = kmalloc(size, flags);	\
-	if (__ret)				\
-		memset(__ret, 0, size);		\
-	__ret;					\
-})
-#endif
-#endif
-
-#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,17)
-#ifndef kmem_cache_zalloc
-#define kmem_cache_zalloc(cache,flags)			  \
-({							  \
-	void *__ret = kmem_cache_alloc(cache, flags);	  \
-	if (__ret)                                        \
-		memset(__ret, 0, kmem_cache_size(cache)); \
-	__ret;                                            \
-})
-#endif
-#endif
-
-#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,21)
-
-#ifndef CONFIG_HOTPLUG_CPU
-#define register_cpu_notifier(nb) (0)
-#endif
-
-#endif
-
-#include <linux/miscdevice.h>
-#ifndef KVM_MINOR
-#define KVM_MINOR 232
-#endif
-
-#include <linux/notifier.h>
-#ifndef CPU_TASKS_FROZEN
-
-#define CPU_TASKS_FROZEN       0x0010
-#define CPU_ONLINE_FROZEN      (CPU_ONLINE | CPU_TASKS_FROZEN)
-#define CPU_UP_PREPARE_FROZEN  (CPU_UP_PREPARE | CPU_TASKS_FROZEN)
-#define CPU_UP_CANCELED_FROZEN (CPU_UP_CANCELED | CPU_TASKS_FROZEN)
-#define CPU_DOWN_PREPARE_FROZEN        (CPU_DOWN_PREPARE |
CPU_TASKS_FROZEN)
-#define CPU_DOWN_FAILED_FROZEN (CPU_DOWN_FAILED | CPU_TASKS_FROZEN)
-#define CPU_DEAD_FROZEN                (CPU_DEAD | CPU_TASKS_FROZEN)
-
-#endif
-
-#ifndef CPU_DYING
-#define CPU_DYING 0x000A
-#define CPU_DYING_FROZEN (CPU_DYING | CPU_TASKS_FROZEN)
-#endif
-
-#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,22)
-
-#ifndef _EFER_SCE
-#define _EFER_SCE		0  /* SYSCALL/SYSRET */
-#endif
-
-#ifndef EFER_SCE
-#define EFER_SCE		(1<<_EFER_SCE)
-#endif
-
-#endif
-
-/*
- * For set_64bit(), which is in a new file asm/cmpxchg.h in newer
kernels.
- */
-#include <asm/system.h>
-
-struct inode;
-#include <linux/anon_inodes.h>
-#define anon_inode_getfd kvm_anon_inode_getfd
-int kvm_init_anon_inodes(void);
-void kvm_exit_anon_inodes(void);
-int anon_inode_getfd(const char *name,
-		     const struct file_operations *fops,
-		     void *priv);
-
-#include <linux/smp.h>
-
-#ifndef X86_CR0_PE
-#define X86_CR0_PE 0x00000001
-#endif
-
-#ifndef X86_CR0_MP
-#define X86_CR0_MP 0x00000002
-#endif
-
-#ifndef X86_CR0_EM
-#define X86_CR0_EM 0x00000004
-#endif
-
-#ifndef X86_CR0_TS
-#define X86_CR0_TS 0x00000008
-#endif
-
-#ifndef X86_CR0_ET
-#define X86_CR0_ET 0x00000010
-#endif
-
-#ifndef X86_CR0_NE
-#define X86_CR0_NE 0x00000020
-#endif
-
-#ifndef X86_CR0_WP
-#define X86_CR0_WP 0x00010000
-#endif
-
-#ifndef X86_CR0_AM
-#define X86_CR0_AM 0x00040000
-#endif
-
-#ifndef X86_CR0_NW
-#define X86_CR0_NW 0x20000000
-#endif
-
-#ifndef X86_CR0_CD
-#define X86_CR0_CD 0x40000000
-#endif
-
-#ifndef X86_CR0_PG
-#define X86_CR0_PG 0x80000000
-#endif
-
-#ifndef X86_CR3_PWT
-#define X86_CR3_PWT 0x00000008
-#endif
-
-#ifndef X86_CR3_PCD
-#define X86_CR3_PCD 0x00000010
-#endif
-
-#ifndef X86_CR4_VMXE
-#define X86_CR4_VMXE 0x00002000
-#endif
-
-#undef X86_CR8_TPR
-#define X86_CR8_TPR 0x0f
-
-/*
- * 2.6.23 removed the cache destructor
- */
-#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,23)
-#  define kmem_cache_create(name, size, align, flags, ctor) \
-	kmem_cache_create(name, size, align, flags, ctor, NULL)
-#endif
-
-/*
- * 2.6.22 does not define set_64bit() under nonpae
- */
-#ifdef CONFIG_X86_32
-
-#include <asm/cmpxchg.h>
-
-static inline void __kvm_set_64bit(u64 *ptr, u64 val)
-{
-	unsigned int low = val;
-	unsigned int high = val >> 32;
-
-	__asm__ __volatile__ (
-		"\n1:\t"
-		"movl (%0), %%eax\n\t"
-		"movl 4(%0), %%edx\n\t"
-		"lock cmpxchg8b (%0)\n\t"
-		"jnz 1b"
-		: /* no outputs */
-		:	"D"(ptr),
-			"b"(low),
-			"c"(high)
-		:	"ax","dx","memory");
-}
-
-#undef  set_64bit
-#define set_64bit __kvm_set_64bit
-
-static inline unsigned long long __kvm_cmpxchg64(volatile void *ptr,
-						 unsigned long long old,
-						 unsigned long long new)
-{
-	unsigned long long prev;
-	__asm__ __volatile__("lock cmpxchg8b %3"
-			     : "=A"(prev)
-			     : "b"((unsigned long)new),
-			       "c"((unsigned long)(new >> 32)),
-			       "m"(*__xg(ptr)),
-			       "0"(old)
-			     : "memory");
-	return prev;
-}
-
-#define kvm_cmpxchg64(ptr,o,n)\
-	((__typeof__(*(ptr)))__kvm_cmpxchg64((ptr),(unsigned long
long)(o),\
-					(unsigned long long)(n)))
-
-#undef cmpxchg64
-#define cmpxchg64(ptr, o, n) kvm_cmpxchg64(ptr, o, n)
-
-#endif
-
-#ifndef CONFIG_PREEMPT_NOTIFIERS
-/*
- * Include sched|preempt.h before defining CONFIG_PREEMPT_NOTIFIERS to
avoid
- * a miscompile.
- */
-#include <linux/sched.h>
-#include <linux/preempt.h>
-#define CONFIG_PREEMPT_NOTIFIERS
-#define CONFIG_PREEMPT_NOTIFIERS_COMPAT
-
-struct preempt_notifier;
-
-struct preempt_ops {
-	void (*sched_in)(struct preempt_notifier *notifier, int cpu);
-	void (*sched_out)(struct preempt_notifier *notifier,
-			  struct task_struct *next);
-};
-
-struct preempt_notifier {
-	struct list_head link;
-	struct task_struct *tsk;
-	struct preempt_ops *ops;
-};
-
-void preempt_notifier_register(struct preempt_notifier *notifier);
-void preempt_notifier_unregister(struct preempt_notifier *notifier);
-
-static inline void preempt_notifier_init(struct preempt_notifier
*notifier,
-				     struct preempt_ops *ops)
-{
-	notifier->ops = ops;
-}
-
-void start_special_insn(void);
-void end_special_insn(void);
-void in_special_section(void);
-void special_reload_dr7(void);
-
-void preempt_notifier_sys_init(void);
-void preempt_notifier_sys_exit(void);
-
-#else
-
-static inline void start_special_insn(void) {}
-static inline void end_special_insn(void) {}
-static inline void in_special_section(void) {}
-static inline void special_reload_dr7(void) {}
-
-static inline void preempt_notifier_sys_init(void) {}
-static inline void preempt_notifier_sys_exit(void) {}
-
-#endif
-
-/* HRTIMER_MODE_ABS started life with a different name */
-#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,21)
-#define HRTIMER_MODE_ABS HRTIMER_ABS
-#endif
-
-/* div64_u64 is fairly new */
-#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,26)
-
-#define div64_u64 kvm_div64_u64
-
-#ifdef CONFIG_64BIT
-
-static inline uint64_t div64_u64(uint64_t dividend, uint64_t divisor)
-{
-	return dividend / divisor;
-}
-
-#else
-
-uint64_t div64_u64(uint64_t dividend, uint64_t divisor);
-
-#endif
-
-#endif
-
-#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,19)
-
-#ifdef RHEL_RELEASE_CODE
-#if RHEL_RELEASE_CODE >= RHEL_RELEASE_VERSION(5,2)
-#define RHEL_BOOL 1
-#endif
-#endif
-
-#ifndef RHEL_BOOL
-
-typedef _Bool bool;
-
-#endif
-
-#endif
-
-/*
- * PF_VCPU is a Linux 2.6.24 addition
- */
-
-#include <linux/sched.h>
-
-#ifndef PF_VCPU
-#define PF_VCPU 0
-#endif
-
-/*
- * smp_call_function_mask() is not defined/exported below 2.6.24
- */
-
-#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,24)
-
-int kvm_smp_call_function_mask(cpumask_t mask, void (*func) (void
*info),
-			       void *info, int wait);
-
-#define smp_call_function_mask kvm_smp_call_function_mask
-
-#endif
-
-/* CONFIG_HAS_IOMEM is apparently fairly new too (2.6.21 for x86_64).
*/
-#ifndef CONFIG_HAS_IOMEM
-#define CONFIG_HAS_IOMEM 1
-#endif
-
-/* empty_zero_page isn't exported in all kernels */
-#include <asm/pgtable.h>
-
-#define empty_zero_page kvm_empty_zero_page
-
-static char empty_zero_page[PAGE_SIZE];
-
-static inline void blahblah(void)
-{
-	(void)empty_zero_page[0];
-}
-
-/* __mmdrop() is not exported before 2.6.25 */
-#include <linux/sched.h>
-
-#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,25)
-
-#define mmdrop(x) do { (void)(x); } while (0)
-#define mmget(x) do { (void)(x); } while (0)
-
-#else
-
-#define mmget(x) do { atomic_inc(x); } while (0)
-
-#endif
-
-/* X86_FEATURE_NX is missing in some x86_64 kernels */
-
-#include <asm/cpufeature.h>
-
-#ifndef X86_FEATURE_NX
-#define X86_FEATURE_NX (1*32+20)
-#endif
-
-#undef true
-#define true 1
-#undef false
-#define false 0
-
-/* EFER_LMA and EFER_LME are missing in pre 2.6.24 i386 kernels */
-#ifndef EFER_LME
-#define _EFER_LME           8  /* Long mode enable */
-#define _EFER_LMA           10 /* Long mode active (read-only) */
-#define EFER_LME            (1<<_EFER_LME)
-#define EFER_LMA            (1<<_EFER_LMA)
-#endif
-
-/* pagefault_enable(), page_fault_disable() - 2.6.20 */
-#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,20)
-
-#define pagefault_enable()  do {} while (0)
-#define pagefault_disable() do {} while (0)
-
-#endif
-
-/* vm ops ->fault() was introduced in 2.6.23. */
-#include <linux/mm.h>
-
-#ifdef KVM_MAIN
-#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,23)
-
-struct vm_fault {
-	unsigned int flags;
-	pgoff_t pgoff;
-	void __user *virtual_address;
-	struct page *page;
-};
-
-static int kvm_vcpu_fault(struct vm_area_struct *vma, struct vm_fault
*vmf);
-static int kvm_vm_fault(struct vm_area_struct *vma, struct vm_fault
*vmf);
-
-static inline struct page *kvm_nopage_to_fault(
-	int (*fault)(struct vm_area_struct *vma, struct vm_fault *vmf),
-	struct vm_area_struct *vma,
-	unsigned long address,
-	int *type)
-{
-	struct vm_fault vmf;
-	int ret;
-
-	vmf.pgoff = ((address - vma->vm_start) >> PAGE_SHIFT) +
vma->vm_pgoff;
-	vmf.virtual_address = (void __user *)address;
-	ret = fault(vma, &vmf);
-	if (ret)
-		return NOPAGE_SIGBUS;
-	*type = VM_FAULT_MINOR;
-	return vmf.page;
-}
-
-static inline struct page *__kvm_vcpu_fault(struct vm_area_struct *vma,
-					    unsigned long address,
-					    int *type)
-{
-	return kvm_nopage_to_fault(kvm_vcpu_fault, vma, address, type);
-}
-
-static inline struct page *__kvm_vm_fault(struct vm_area_struct *vma,
-					  unsigned long address,
-					  int *type)
-{
-	return kvm_nopage_to_fault(kvm_vm_fault, vma, address, type);
-}
-
-#define VMA_OPS_FAULT(x) nopage
-#define VMA_OPS_FAULT_FUNC(x) __##x
-
-#else
-
-#define VMA_OPS_FAULT(x) x
-#define VMA_OPS_FAULT_FUNC(x) x
-
-#endif
-#endif
-
-/* simple vfs attribute getter signature has changed to add a return
code */
-
-#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,25)
-
-#define MAKE_SIMPLE_ATTRIBUTE_GETTER(x)       \
-	static u64 x(void *v)                 \
-	{				      \
-		u64 ret = 0;		      \
-					      \
-		__##x(v, &ret);		      \
-		return ret;		      \
-	}
-
-#else
-
-#define MAKE_SIMPLE_ATTRIBUTE_GETTER(x)       \
-	static int x(void *v, u64 *val)	      \
-	{				      \
-		return __##x(v, val);	      \
-	}
-
-#endif
-
-/* set_kset_name() is gone in 2.6.25 */
-
-#if LINUX_VERSION_CODE >= KERNEL_VERSION(2,6,25)
-
-#define set_kset_name(x) .name = x
-
-#endif
-
-#if LINUX_VERSION_CODE >= KERNEL_VERSION(2,6,25)
-#ifndef FASTCALL
-#define FASTCALL(x)	x
-#define fastcall
-#endif
-#endif
-
-#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,23)
-
-static unsigned  __attribute__((__used__)) kvm_tsc_khz = 2000000;
-
-#else
-
-#define kvm_tsc_khz tsc_khz
-
-#endif
-
-struct kvm_desc_struct {
-	union {
-		struct { unsigned int a, b; };
-		struct {
-			u16 limit0;
-			u16 base0;
-			unsigned base1: 8, type: 4, s: 1, dpl: 2, p: 1;
-			unsigned limit: 4, avl: 1, l: 1, d: 1, g: 1,
base2: 8;
-		};
-
-	};
-} __attribute__((packed));
-
-struct kvm_ldttss_desc64 {
-	u16 limit0;
-	u16 base0;
-	unsigned base1 : 8, type : 5, dpl : 2, p : 1;
-	unsigned limit1 : 4, zero0 : 3, g : 1, base2 : 8;
-	u32 base3;
-	u32 zero1;
-} __attribute__((packed));
-
-struct kvm_desc_ptr {
-	unsigned short size;
-	unsigned long address;
-} __attribute__((packed));
-
-#if LINUX_VERSION_CODE <= KERNEL_VERSION(2,6,21)
-
-#include <linux/ktime.h>
-#include <linux/hrtimer.h>
-
-#define ktime_get kvm_ktime_get
-
-static inline ktime_t ktime_get(void)
-{
-	struct timespec now;
-
-	ktime_get_ts(&now);
-
-	return timespec_to_ktime(now);
-}
-
-#endif
-
-/* __aligned arrived in 2.6.21 */
-#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,21)
-#define __aligned(x) __attribute__((__aligned__(x)))
-#endif
-
-#include <asm/msr.h>
-#ifndef MSR_FS_BASE
-#define MSR_FS_BASE 0xc0000100
-#endif
-#ifndef MSR_GS_BASE
-#define MSR_GS_BASE 0xc0000101
-#endif
-
-#include <linux/mm.h>
-
-/* The shrinker API changed in 2.6.23 */
-#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,23)
-
-struct kvm_shrinker {
-	int (*shrink)(int nr_to_scan, gfp_t gfp_mask);
-	int seeks;
-	struct shrinker *kshrinker;
-};
-
-static inline void register_shrinker(struct kvm_shrinker *shrinker)
-{
-	shrinker->kshrinker = set_shrinker(shrinker->seeks,
shrinker->shrink);
-}
-
-static inline void unregister_shrinker(struct kvm_shrinker *shrinker)
-{
-	if (shrinker->kshrinker)
-		remove_shrinker(shrinker->kshrinker);
-}
-
-#define shrinker kvm_shrinker
-
-#endif
-
-/* undefine lapic */
-#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,18)
-
-#undef lapic
-
-#endif
-
-/* clocksource */
-#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,18)
-static inline u32 clocksource_khz2mult(u32 khz, u32 shift_constant)
-{
-	/*  khz = cyc/(Million ns)
-	 *  mult/2^shift  = ns/cyc
-	 *  mult = ns/cyc * 2^shift
-	 *  mult = 1Million/khz * 2^shift
-	 *  mult = 1000000 * 2^shift / khz
-	 *  mult = (1000000<<shift) / khz
-	 */
-	u64 tmp = ((u64)1000000) << shift_constant;
-
-	tmp += khz/2; /* round for do_div */
-	do_div(tmp, khz);
-
-	return (u32)tmp;
-}
-#else
-#include <linux/clocksource.h>
-#endif
-
-/* manually export hrtimer_init/start/cancel */
-#include <linux/kallsyms.h>
-extern void (*hrtimer_init_p)(struct hrtimer *timer, clockid_t
which_clock,
-			      enum hrtimer_mode mode);
-extern int (*hrtimer_start_p)(struct hrtimer *timer, ktime_t tim,
-			      const enum hrtimer_mode mode);
-extern int (*hrtimer_cancel_p)(struct hrtimer *timer);
-
-#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,17) &&
defined(CONFIG_KALLSYMS)
-static inline void hrtimer_kallsyms_resolve(void)
-{
-	hrtimer_init_p = (void *) kallsyms_lookup_name("hrtimer_init");
-	BUG_ON(!hrtimer_init_p);
-	hrtimer_start_p = (void *)
kallsyms_lookup_name("hrtimer_start");
-	BUG_ON(!hrtimer_start_p);
-	hrtimer_cancel_p = (void *)
kallsyms_lookup_name("hrtimer_cancel");
-	BUG_ON(!hrtimer_cancel_p);
-}
-#else
-static inline void hrtimer_kallsyms_resolve(void)
-{
-	hrtimer_init_p = hrtimer_init;
-	hrtimer_start_p = hrtimer_start;
-	hrtimer_cancel_p = hrtimer_cancel;
-}
-#endif
-
-/* handle old hrtimer API with data pointer */
-#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,17)
-static inline void hrtimer_data_pointer(struct hrtimer *timer)
-{
-	timer->data = (void *)timer;
-}
-#else
-static inline void hrtimer_data_pointer(struct hrtimer *timer) {}
-#endif
-
-#include <asm/hw_irq.h>
-#ifndef NMI_VECTOR
-#define NMI_VECTOR 2
-#endif
-
-#ifndef MSR_MTRRcap
-#define MSR_MTRRcap            0x0fe
-#define MSR_MTRRfix64K_00000   0x250
-#define MSR_MTRRfix16K_80000   0x258
-#define MSR_MTRRfix16K_A0000   0x259
-#define MSR_MTRRfix4K_C0000    0x268
-#define MSR_MTRRfix4K_C8000    0x269
-#define MSR_MTRRfix4K_D0000    0x26a
-#define MSR_MTRRfix4K_D8000    0x26b
-#define MSR_MTRRfix4K_E0000    0x26c
-#define MSR_MTRRfix4K_E8000    0x26d
-#define MSR_MTRRfix4K_F0000    0x26e
-#define MSR_MTRRfix4K_F8000    0x26f
-#define MSR_MTRRdefType        0x2ff
-#endif
-
-#ifndef MSR_IA32_CR_PAT
-#define MSR_IA32_CR_PAT        0x00000277
-#endif
-
-#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,22)
-
-#define ns_to_timespec kvm_ns_to_timespec
-
-struct timespec kvm_ns_to_timespec(const s64 nsec);
-
-#endif
-
-/* Define DEBUGCTLMSR bits */
-#ifndef DEBUGCTLMSR_LBR
-
-#define _DEBUGCTLMSR_LBR	0 /* last branch recording */
-#define _DEBUGCTLMSR_BTF	1 /* single-step on branches */
-
-#define DEBUGCTLMSR_LBR		(1UL << _DEBUGCTLMSR_LBR)
-#define DEBUGCTLMSR_BTF		(1UL << _DEBUGCTLMSR_BTF)
-
-#endif
diff --git a/kernel/debug.h b/kernel/x86/debug.h
similarity index 100%
rename from kernel/debug.h
rename to kernel/x86/debug.h
diff --git a/kernel/external-module-compat.c
b/kernel/x86/external-module-compat.c
similarity index 100%
rename from kernel/external-module-compat.c
rename to kernel/x86/external-module-compat.c
diff --git a/kernel/x86/external-module-compat.h
b/kernel/x86/external-module-compat.h
new file mode 100644
index 0000000..ce93551
--- /dev/null
+++ b/kernel/x86/external-module-compat.h
@@ -0,0 +1,291 @@
+
+/*
+ * Compatibility header for building as an external module.
+ */
+
+#include <linux/compiler.h>
+#include <linux/version.h>
+
+#include "../external-module-compat-comm.h"
+
+#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,22)
+
+#ifndef _EFER_SCE
+#define _EFER_SCE		0  /* SYSCALL/SYSRET */
+#endif
+
+#ifndef EFER_SCE
+#define EFER_SCE		(1<<_EFER_SCE)
+#endif
+
+#endif
+
+#include <linux/smp.h>
+
+#ifndef X86_CR0_PE
+#define X86_CR0_PE 0x00000001
+#endif
+
+#ifndef X86_CR0_MP
+#define X86_CR0_MP 0x00000002
+#endif
+
+#ifndef X86_CR0_EM
+#define X86_CR0_EM 0x00000004
+#endif
+
+#ifndef X86_CR0_TS
+#define X86_CR0_TS 0x00000008
+#endif
+
+#ifndef X86_CR0_ET
+#define X86_CR0_ET 0x00000010
+#endif
+
+#ifndef X86_CR0_NE
+#define X86_CR0_NE 0x00000020
+#endif
+
+#ifndef X86_CR0_WP
+#define X86_CR0_WP 0x00010000
+#endif
+
+#ifndef X86_CR0_AM
+#define X86_CR0_AM 0x00040000
+#endif
+
+#ifndef X86_CR0_NW
+#define X86_CR0_NW 0x20000000
+#endif
+
+#ifndef X86_CR0_CD
+#define X86_CR0_CD 0x40000000
+#endif
+
+#ifndef X86_CR0_PG
+#define X86_CR0_PG 0x80000000
+#endif
+
+#ifndef X86_CR3_PWT
+#define X86_CR3_PWT 0x00000008
+#endif
+
+#ifndef X86_CR3_PCD
+#define X86_CR3_PCD 0x00000010
+#endif
+
+#ifndef X86_CR4_VMXE
+#define X86_CR4_VMXE 0x00002000
+#endif
+
+#undef X86_CR8_TPR
+#define X86_CR8_TPR 0x0f
+
+/*
+ * 2.6.22 does not define set_64bit() under nonpae
+ */
+#ifdef CONFIG_X86_32
+
+#include <asm/cmpxchg.h>
+
+static inline void __kvm_set_64bit(u64 *ptr, u64 val)
+{
+	unsigned int low = val;
+	unsigned int high = val >> 32;
+
+	__asm__ __volatile__ (
+		"\n1:\t"
+		"movl (%0), %%eax\n\t"
+		"movl 4(%0), %%edx\n\t"
+		"lock cmpxchg8b (%0)\n\t"
+		"jnz 1b"
+		: /* no outputs */
+		:	"D"(ptr),
+			"b"(low),
+			"c"(high)
+		:	"ax","dx","memory");
+}
+
+#undef  set_64bit
+#define set_64bit __kvm_set_64bit
+
+static inline unsigned long long __kvm_cmpxchg64(volatile void *ptr,
+						 unsigned long long old,
+						 unsigned long long new)
+{
+	unsigned long long prev;
+	__asm__ __volatile__("lock cmpxchg8b %3"
+			     : "=A"(prev)
+			     : "b"((unsigned long)new),
+			       "c"((unsigned long)(new >> 32)),
+			       "m"(*__xg(ptr)),
+			       "0"(old)
+			     : "memory");
+	return prev;
+}
+
+#define kvm_cmpxchg64(ptr,o,n)\
+	((__typeof__(*(ptr)))__kvm_cmpxchg64((ptr),(unsigned long
long)(o),\
+					(unsigned long long)(n)))
+
+#undef cmpxchg64
+#define cmpxchg64(ptr, o, n) kvm_cmpxchg64(ptr, o, n)
+
+#endif
+
+#ifndef CONFIG_PREEMPT_NOTIFIERS
+/*
+ * Include sched|preempt.h before defining CONFIG_PREEMPT_NOTIFIERS to
avoid
+ * a miscompile.
+ */
+#include <linux/sched.h>
+#include <linux/preempt.h>
+#define CONFIG_PREEMPT_NOTIFIERS
+#define CONFIG_PREEMPT_NOTIFIERS_COMPAT
+
+struct preempt_notifier;
+
+struct preempt_ops {
+	void (*sched_in)(struct preempt_notifier *notifier, int cpu);
+	void (*sched_out)(struct preempt_notifier *notifier,
+			  struct task_struct *next);
+};
+
+struct preempt_notifier {
+	struct list_head link;
+	struct task_struct *tsk;
+	struct preempt_ops *ops;
+};
+
+void preempt_notifier_register(struct preempt_notifier *notifier);
+void preempt_notifier_unregister(struct preempt_notifier *notifier);
+
+static inline void preempt_notifier_init(struct preempt_notifier
*notifier,
+				     struct preempt_ops *ops)
+{
+	notifier->ops = ops;
+}
+
+void start_special_insn(void);
+void end_special_insn(void);
+void in_special_section(void);
+void special_reload_dr7(void);
+
+void preempt_notifier_sys_init(void);
+void preempt_notifier_sys_exit(void);
+
+#else
+
+static inline void start_special_insn(void) {}
+static inline void end_special_insn(void) {}
+static inline void in_special_section(void) {}
+static inline void special_reload_dr7(void) {}
+
+static inline void preempt_notifier_sys_init(void) {}
+static inline void preempt_notifier_sys_exit(void) {}
+
+#endif
+
+/* CONFIG_HAS_IOMEM is apparently fairly new too (2.6.21 for x86_64).
*/
+#ifndef CONFIG_HAS_IOMEM
+#define CONFIG_HAS_IOMEM 1
+#endif
+
+/* X86_FEATURE_NX is missing in some x86_64 kernels */
+
+#include <asm/cpufeature.h>
+
+#ifndef X86_FEATURE_NX
+#define X86_FEATURE_NX (1*32+20)
+#endif
+
+#undef true
+#define true 1
+#undef false
+#define false 0
+
+/* EFER_LMA and EFER_LME are missing in pre 2.6.24 i386 kernels */
+#ifndef EFER_LME
+#define _EFER_LME           8  /* Long mode enable */
+#define _EFER_LMA           10 /* Long mode active (read-only) */
+#define EFER_LME            (1<<_EFER_LME)
+#define EFER_LMA            (1<<_EFER_LMA)
+#endif
+
+struct kvm_desc_struct {
+	union {
+		struct { unsigned int a, b; };
+		struct {
+			u16 limit0;
+			u16 base0;
+			unsigned base1: 8, type: 4, s: 1, dpl: 2, p: 1;
+			unsigned limit: 4, avl: 1, l: 1, d: 1, g: 1,
base2: 8;
+		};
+
+	};
+} __attribute__((packed));
+
+struct kvm_ldttss_desc64 {
+	u16 limit0;
+	u16 base0;
+	unsigned base1 : 8, type : 5, dpl : 2, p : 1;
+	unsigned limit1 : 4, zero0 : 3, g : 1, base2 : 8;
+	u32 base3;
+	u32 zero1;
+} __attribute__((packed));
+
+struct kvm_desc_ptr {
+	unsigned short size;
+	unsigned long address;
+} __attribute__((packed));
+
+#include <asm/msr.h>
+#ifndef MSR_FS_BASE
+#define MSR_FS_BASE 0xc0000100
+#endif
+#ifndef MSR_GS_BASE
+#define MSR_GS_BASE 0xc0000101
+#endif
+
+/* undefine lapic */
+#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,18)
+
+#undef lapic
+
+#endif
+
+#include <asm/hw_irq.h>
+#ifndef NMI_VECTOR
+#define NMI_VECTOR 2
+#endif
+
+#ifndef MSR_MTRRcap
+#define MSR_MTRRcap            0x0fe
+#define MSR_MTRRfix64K_00000   0x250
+#define MSR_MTRRfix16K_80000   0x258
+#define MSR_MTRRfix16K_A0000   0x259
+#define MSR_MTRRfix4K_C0000    0x268
+#define MSR_MTRRfix4K_C8000    0x269
+#define MSR_MTRRfix4K_D0000    0x26a
+#define MSR_MTRRfix4K_D8000    0x26b
+#define MSR_MTRRfix4K_E0000    0x26c
+#define MSR_MTRRfix4K_E8000    0x26d
+#define MSR_MTRRfix4K_F0000    0x26e
+#define MSR_MTRRfix4K_F8000    0x26f
+#define MSR_MTRRdefType        0x2ff
+#endif
+
+#ifndef MSR_IA32_CR_PAT
+#define MSR_IA32_CR_PAT        0x00000277
+#endif
+
+/* Define DEBUGCTLMSR bits */
+#ifndef DEBUGCTLMSR_LBR
+
+#define _DEBUGCTLMSR_LBR	0 /* last branch recording */
+#define _DEBUGCTLMSR_BTF	1 /* single-step on branches */
+
+#define DEBUGCTLMSR_LBR		(1UL << _DEBUGCTLMSR_LBR)
+#define DEBUGCTLMSR_BTF		(1UL << _DEBUGCTLMSR_BTF)
+
+#endif
diff --git a/kernel/preempt.c b/kernel/x86/preempt.c
similarity index 100%
rename from kernel/preempt.c
rename to kernel/x86/preempt.c
diff --git a/kernel/vmx-debug.c b/kernel/x86/vmx-debug.c
similarity index 100%
rename from kernel/vmx-debug.c
rename to kernel/x86/vmx-debug.c
-- 
1.5.6.2

^ permalink raw reply related	[flat|nested] 24+ messages in thread

* Re: [PATCH 01/04]Create x86 directory to hold x86-specific files.
  2008-07-25  9:26   ` Zhang, Xiantao
@ 2008-07-27  7:45   ` Avi Kivity
  -1 siblings, 0 replies; 24+ messages in thread
From: Avi Kivity @ 2008-07-27  7:45 UTC (permalink / raw)
  To: kvm-ia64

Zhang, Xiantao wrote:
> From c8c8c5a6b3ff943cb645f9ac4ef169490f14cd08 Mon Sep 17 00:00:00 2001
> From: Zhang Xiantao <xiantao.zhang@intel.com>
> Date: Fri, 25 Jul 2008 16:50:30 +0800
> Subject: [PATCH] KVM: external module: Moving x86-speicif files to x86
> directory.
>
> Create x86 directory to hold x86-specific files.
> Signed-off-by: Zhang Xiantao <xiantao.zhang@intel.com>
>   

You forgot to attach a patch; and your mailer corrupts patches, so I 
can't apply this.

-- 
error compiling committee.c: too many arguments to function


^ permalink raw reply	[flat|nested] 24+ messages in thread

* Re: [PATCH 01/04]Create x86 directory to hold x86-specific files.
@ 2008-07-27  7:45   ` Avi Kivity
  0 siblings, 0 replies; 24+ messages in thread
From: Avi Kivity @ 2008-07-27  7:45 UTC (permalink / raw)
  To: Zhang, Xiantao; +Cc: kvm-ia64, kvm

Zhang, Xiantao wrote:
> From c8c8c5a6b3ff943cb645f9ac4ef169490f14cd08 Mon Sep 17 00:00:00 2001
> From: Zhang Xiantao <xiantao.zhang@intel.com>
> Date: Fri, 25 Jul 2008 16:50:30 +0800
> Subject: [PATCH] KVM: external module: Moving x86-speicif files to x86
> directory.
>
> Create x86 directory to hold x86-specific files.
> Signed-off-by: Zhang Xiantao <xiantao.zhang@intel.com>
>   

You forgot to attach a patch; and your mailer corrupts patches, so I 
can't apply this.

-- 
error compiling committee.c: too many arguments to function


^ permalink raw reply	[flat|nested] 24+ messages in thread

* RE: [PATCH 01/04]Create x86 directory to hold x86-specific files.
  2008-07-27  7:45   ` Avi Kivity
@ 2008-07-27  9:46   ` Zhang, Xiantao
  -1 siblings, 0 replies; 24+ messages in thread
From: Zhang, Xiantao @ 2008-07-27  9:46 UTC (permalink / raw)
  To: kvm-ia64

[-- Attachment #1: Type: text/plain, Size: 615 bytes --]

Avi Kivity wrote:
> Zhang, Xiantao wrote:
>> From c8c8c5a6b3ff943cb645f9ac4ef169490f14cd08 Mon Sep 17 00:00:00
>> 2001 From: Zhang Xiantao <xiantao.zhang@intel.com>
>> Date: Fri, 25 Jul 2008 16:50:30 +0800
>> Subject: [PATCH] KVM: external module: Moving x86-speicif files to
>> x86 directory. 
>> 
>> Create x86 directory to hold x86-specific files.
>> Signed-off-by: Zhang Xiantao <xiantao.zhang@intel.com>
>> 
> 
> You forgot to attach a patch; and your mailer corrupts patches, so I
> can't apply this.

Sorry. Attached! :)
> --
> error compiling committee.c: too many arguments to function


[-- Attachment #2: 0001-KVM-external-module-Moving-x86-speicif-files-to-x8.patch --]
[-- Type: application/octet-stream, Size: 38217 bytes --]

From c8c8c5a6b3ff943cb645f9ac4ef169490f14cd08 Mon Sep 17 00:00:00 2001
From: Zhang Xiantao <xiantao.zhang@intel.com>
Date: Fri, 25 Jul 2008 16:50:30 +0800
Subject: [PATCH] KVM: external module: Moving x86-speicif files to x86 directory.

Create x86 directory to hold x86-specific files.
Signed-off-by: Zhang Xiantao <xiantao.zhang@intel.com>
---
 kernel/Kbuild                             |    8 +-
 kernel/Makefile                           |    2 +-
 kernel/external-module-compat-comm.h      |  475 ++++++++++++++++++
 kernel/external-module-compat.h           |  759 -----------------------------
 kernel/{ => x86}/debug.h                  |    0 
 kernel/{ => x86}/external-module-compat.c |    0 
 kernel/x86/external-module-compat.h       |  291 +++++++++++
 kernel/{ => x86}/preempt.c                |    0 
 kernel/{ => x86}/vmx-debug.c              |    0 
 9 files changed, 771 insertions(+), 764 deletions(-)
 create mode 100644 kernel/external-module-compat-comm.h
 delete mode 100644 kernel/external-module-compat.h
 rename kernel/{ => x86}/debug.h (100%)
 rename kernel/{ => x86}/external-module-compat.c (100%)
 create mode 100644 kernel/x86/external-module-compat.h
 rename kernel/{ => x86}/preempt.c (100%)
 rename kernel/{ => x86}/vmx-debug.c (100%)

diff --git a/kernel/Kbuild b/kernel/Kbuild
index 676f28d..cb63582 100644
--- a/kernel/Kbuild
+++ b/kernel/Kbuild
@@ -1,11 +1,11 @@
 obj-m := kvm.o kvm-intel.o kvm-amd.o
 kvm-objs := kvm_main.o x86.o mmu.o x86_emulate.o anon_inodes.o irq.o i8259.o \
-	 lapic.o ioapic.o preempt.o i8254.o coalesced_mmio.o \
-	 external-module-compat.o
+	 lapic.o ioapic.o x86/preempt.o i8254.o coalesced_mmio.o \
+	 x86/external-module-compat.o
 ifeq ($(CONFIG_KVM_TRACE),y)
 kvm-objs += kvm_trace.o
 endif
-kvm-intel-objs := vmx.o vmx-debug.o external-module-compat.o
-kvm-amd-objs := svm.o external-module-compat.o
+kvm-intel-objs := vmx.o x86/vmx-debug.o x86/external-module-compat.o
+kvm-amd-objs := svm.o x86/external-module-compat.o
 
 CFLAGS_kvm_main.o = -DKVM_MAIN
diff --git a/kernel/Makefile b/kernel/Makefile
index 69b4981..c2c666e 100644
--- a/kernel/Makefile
+++ b/kernel/Makefile
@@ -35,7 +35,7 @@ all::
 	$(MAKE) -C $(KERNELDIR) M=`pwd` \
 		LINUXINCLUDE="-I`pwd`/include -Iinclude -I`pwd`/include-compat \
 		-include include/linux/autoconf.h \
-		-include `pwd`/external-module-compat.h"
+		-include `pwd`/$(ARCH_DIR)/external-module-compat.h"
 		"$$@"
 
 sync: header-sync source-sync
diff --git a/kernel/external-module-compat-comm.h b/kernel/external-module-compat-comm.h
new file mode 100644
index 0000000..d4bf395
--- /dev/null
+++ b/kernel/external-module-compat-comm.h
@@ -0,0 +1,475 @@
+
+/*
+ * Compatibility header for building as an external module.
+ */
+
+/*
+ * Avoid picking up the kernel's kvm.h in case we have a newer one.
+ */
+
+#include <linux/compiler.h>
+#include <linux/version.h>
+#include <linux/string.h>
+#include <linux/kvm.h>
+#include <linux/kvm_para.h>
+#include <linux/cpu.h>
+#include <linux/time.h>
+#include <asm/processor.h>
+#include <linux/hrtimer.h>
+#include <asm/bitops.h>
+#include <asm/msr.h>
+
+/*
+ * 2.6.16 does not have GFP_NOWAIT
+ */
+
+#include <linux/gfp.h>
+
+#ifndef GFP_NOWAIT
+#define GFP_NOWAIT (GFP_ATOMIC & ~__GFP_HIGH)
+#endif
+
+
+/*
+ * kvm profiling support needs 2.6.20
+ */
+#include <linux/profile.h>
+
+#ifndef KVM_PROFILING
+#define KVM_PROFILING 1234
+#define prof_on       4321
+#endif
+
+/*
+ * smp_call_function_single() is not exported below 2.6.20, and has different
+ * semantics below 2.6.23.  The 'nonatomic' argument was removed in 2.6.27.
+ */
+#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,27)
+
+int kvm_smp_call_function_single(int cpu, void (*func)(void *info),
+				 void *info, int wait);
+
+#define smp_call_function_single kvm_smp_call_function_single
+
+#endif
+
+/* on_each_cpu() lost an argument in 2.6.27. */
+#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,27)
+
+#define kvm_on_each_cpu(func, info, wait) on_each_cpu(func, info, 0, wait)
+
+#else
+
+#define kvm_on_each_cpu(func, info, wait) on_each_cpu(func, info, wait)
+
+#endif
+
+/*
+ * The cpu hotplug stubs are broken if !CONFIG_CPU_HOTPLUG
+ */
+
+#if LINUX_VERSION_CODE <= KERNEL_VERSION(2,6,15)
+#define DEFINE_MUTEX(a) DECLARE_MUTEX(a)
+#define mutex_lock_interruptible(a) down_interruptible(a)
+#define mutex_unlock(a) up(a)
+#define mutex_lock(a) down(a)
+#define mutex_init(a) init_MUTEX(a)
+#define mutex_trylock(a) down_trylock(a)
+#define mutex semaphore
+#endif
+
+#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,14)
+#ifndef kzalloc
+#define kzalloc(size,flags)			\
+({						\
+	void *__ret = kmalloc(size, flags);	\
+	if (__ret)				\
+		memset(__ret, 0, size);		\
+	__ret;					\
+})
+#endif
+#endif
+
+#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,17)
+#ifndef kmem_cache_zalloc
+#define kmem_cache_zalloc(cache,flags)			  \
+({							  \
+	void *__ret = kmem_cache_alloc(cache, flags);	  \
+	if (__ret)                                        \
+		memset(__ret, 0, kmem_cache_size(cache)); \
+	__ret;                                            \
+})
+#endif
+#endif
+
+#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,21)
+
+#ifndef CONFIG_HOTPLUG_CPU
+#define register_cpu_notifier(nb) (0)
+#endif
+
+#endif
+
+#include <linux/miscdevice.h>
+#ifndef KVM_MINOR
+#define KVM_MINOR 232
+#endif
+
+#include <linux/notifier.h>
+#ifndef CPU_TASKS_FROZEN
+
+#define CPU_TASKS_FROZEN       0x0010
+#define CPU_ONLINE_FROZEN      (CPU_ONLINE | CPU_TASKS_FROZEN)
+#define CPU_UP_PREPARE_FROZEN  (CPU_UP_PREPARE | CPU_TASKS_FROZEN)
+#define CPU_UP_CANCELED_FROZEN (CPU_UP_CANCELED | CPU_TASKS_FROZEN)
+#define CPU_DOWN_PREPARE_FROZEN        (CPU_DOWN_PREPARE | CPU_TASKS_FROZEN)
+#define CPU_DOWN_FAILED_FROZEN (CPU_DOWN_FAILED | CPU_TASKS_FROZEN)
+#define CPU_DEAD_FROZEN                (CPU_DEAD | CPU_TASKS_FROZEN)
+
+#endif
+
+#ifndef CPU_DYING
+#define CPU_DYING 0x000A
+#define CPU_DYING_FROZEN (CPU_DYING | CPU_TASKS_FROZEN)
+#endif
+
+#include <asm/system.h>
+
+struct inode;
+#include <linux/anon_inodes.h>
+#define anon_inode_getfd kvm_anon_inode_getfd
+int kvm_init_anon_inodes(void);
+void kvm_exit_anon_inodes(void);
+int anon_inode_getfd(const char *name,
+		     const struct file_operations *fops,
+		     void *priv);
+
+/*
+ * 2.6.23 removed the cache destructor
+ */
+#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,23)
+#  define kmem_cache_create(name, size, align, flags, ctor) \
+	kmem_cache_create(name, size, align, flags, ctor, NULL)
+#endif
+
+/* HRTIMER_MODE_ABS started life with a different name */
+#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,21)
+#define HRTIMER_MODE_ABS HRTIMER_ABS
+#endif
+
+/* div64_u64 is fairly new */
+#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,26)
+
+#define div64_u64 kvm_div64_u64
+
+#ifdef CONFIG_64BIT
+
+static inline uint64_t div64_u64(uint64_t dividend, uint64_t divisor)
+{
+	return dividend / divisor;
+}
+
+#else
+
+uint64_t div64_u64(uint64_t dividend, uint64_t divisor);
+
+#endif
+
+#endif
+
+#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,19)
+
+#ifdef RHEL_RELEASE_CODE
+#if RHEL_RELEASE_CODE >= RHEL_RELEASE_VERSION(5,2)
+#define RHEL_BOOL 1
+#endif
+#endif
+
+#ifndef RHEL_BOOL
+
+typedef _Bool bool;
+
+#endif
+
+#endif
+
+/*
+ * PF_VCPU is a Linux 2.6.24 addition
+ */
+
+#include <linux/sched.h>
+
+#ifndef PF_VCPU
+#define PF_VCPU 0
+#endif
+
+/*
+ * smp_call_function_mask() is not defined/exported below 2.6.24
+ */
+
+#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,24)
+
+int kvm_smp_call_function_mask(cpumask_t mask, void (*func) (void *info),
+			       void *info, int wait);
+
+#define smp_call_function_mask kvm_smp_call_function_mask
+
+#endif
+
+/* empty_zero_page isn't exported in all kernels */
+#include <asm/pgtable.h>
+
+#define empty_zero_page kvm_empty_zero_page
+
+static char empty_zero_page[PAGE_SIZE];
+
+static inline void blahblah(void)
+{
+	(void)empty_zero_page[0];
+}
+
+/* __mmdrop() is not exported before 2.6.25 */
+#include <linux/sched.h>
+
+#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,25)
+
+#define mmdrop(x) do { (void)(x); } while (0)
+#define mmget(x) do { (void)(x); } while (0)
+
+#else
+
+#define mmget(x) do { atomic_inc(x); } while (0)
+
+#endif
+
+/* pagefault_enable(), page_fault_disable() - 2.6.20 */
+#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,20)
+
+#define pagefault_enable()  do {} while (0)
+#define pagefault_disable() do {} while (0)
+
+#endif
+
+/* vm ops ->fault() was introduced in 2.6.23. */
+#include <linux/mm.h>
+
+#ifdef KVM_MAIN
+#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,23)
+
+struct vm_fault {
+	unsigned int flags;
+	pgoff_t pgoff;
+	void __user *virtual_address;
+	struct page *page;
+};
+
+static int kvm_vcpu_fault(struct vm_area_struct *vma, struct vm_fault *vmf);
+static int kvm_vm_fault(struct vm_area_struct *vma, struct vm_fault *vmf);
+
+static inline struct page *kvm_nopage_to_fault(
+	int (*fault)(struct vm_area_struct *vma, struct vm_fault *vmf),
+	struct vm_area_struct *vma,
+	unsigned long address,
+	int *type)
+{
+	struct vm_fault vmf;
+	int ret;
+
+	vmf.pgoff = ((address - vma->vm_start) >> PAGE_SHIFT) + vma->vm_pgoff;
+	vmf.virtual_address = (void __user *)address;
+	ret = fault(vma, &vmf);
+	if (ret)
+		return NOPAGE_SIGBUS;
+	*type = VM_FAULT_MINOR;
+	return vmf.page;
+}
+
+static inline struct page *__kvm_vcpu_fault(struct vm_area_struct *vma,
+					    unsigned long address,
+					    int *type)
+{
+	return kvm_nopage_to_fault(kvm_vcpu_fault, vma, address, type);
+}
+
+static inline struct page *__kvm_vm_fault(struct vm_area_struct *vma,
+					  unsigned long address,
+					  int *type)
+{
+	return kvm_nopage_to_fault(kvm_vm_fault, vma, address, type);
+}
+
+#define VMA_OPS_FAULT(x) nopage
+#define VMA_OPS_FAULT_FUNC(x) __##x
+
+#else
+
+#define VMA_OPS_FAULT(x) x
+#define VMA_OPS_FAULT_FUNC(x) x
+
+#endif
+#endif
+
+/* simple vfs attribute getter signature has changed to add a return code */
+
+#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,25)
+
+#define MAKE_SIMPLE_ATTRIBUTE_GETTER(x)       \
+	static u64 x(void *v)                 \
+	{				      \
+		u64 ret = 0;		      \
+					      \
+		__##x(v, &ret);		      \
+		return ret;		      \
+	}
+
+#else
+
+#define MAKE_SIMPLE_ATTRIBUTE_GETTER(x)       \
+	static int x(void *v, u64 *val)	      \
+	{				      \
+		return __##x(v, val);	      \
+	}
+
+#endif
+
+/* set_kset_name() is gone in 2.6.25 */
+
+#if LINUX_VERSION_CODE >= KERNEL_VERSION(2,6,25)
+
+#define set_kset_name(x) .name = x
+
+#endif
+
+#if LINUX_VERSION_CODE >= KERNEL_VERSION(2,6,25)
+#ifndef FASTCALL
+#define FASTCALL(x)	x
+#define fastcall
+#endif
+#endif
+
+#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,23)
+
+static unsigned  __attribute__((__used__)) kvm_tsc_khz = 2000000;
+
+#else
+
+#define kvm_tsc_khz tsc_khz
+
+#endif
+
+#if LINUX_VERSION_CODE <= KERNEL_VERSION(2,6,21)
+
+#include <linux/ktime.h>
+#include <linux/hrtimer.h>
+
+#define ktime_get kvm_ktime_get
+
+static inline ktime_t ktime_get(void)
+{
+	struct timespec now;
+
+	ktime_get_ts(&now);
+
+	return timespec_to_ktime(now);
+}
+
+#endif
+
+/* __aligned arrived in 2.6.21 */
+#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,21)
+#define __aligned(x) __attribute__((__aligned__(x)))
+#endif
+
+#include <linux/mm.h>
+
+/* The shrinker API changed in 2.6.23 */
+#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,23)
+
+struct kvm_shrinker {
+	int (*shrink)(int nr_to_scan, gfp_t gfp_mask);
+	int seeks;
+	struct shrinker *kshrinker;
+};
+
+static inline void register_shrinker(struct kvm_shrinker *shrinker)
+{
+	shrinker->kshrinker = set_shrinker(shrinker->seeks, shrinker->shrink);
+}
+
+static inline void unregister_shrinker(struct kvm_shrinker *shrinker)
+{
+	if (shrinker->kshrinker)
+		remove_shrinker(shrinker->kshrinker);
+}
+
+#define shrinker kvm_shrinker
+
+#endif
+
+/* clocksource */
+#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,18)
+static inline u32 clocksource_khz2mult(u32 khz, u32 shift_constant)
+{
+	/*  khz = cyc/(Million ns)
+	 *  mult/2^shift  = ns/cyc
+	 *  mult = ns/cyc * 2^shift
+	 *  mult = 1Million/khz * 2^shift
+	 *  mult = 1000000 * 2^shift / khz
+	 *  mult = (1000000<<shift) / khz
+	 */
+	u64 tmp = ((u64)1000000) << shift_constant;
+
+	tmp += khz/2; /* round for do_div */
+	do_div(tmp, khz);
+
+	return (u32)tmp;
+}
+#else
+#include <linux/clocksource.h>
+#endif
+
+/* manually export hrtimer_init/start/cancel */
+#include <linux/kallsyms.h>
+extern void (*hrtimer_init_p)(struct hrtimer *timer, clockid_t which_clock,
+			      enum hrtimer_mode mode);
+extern int (*hrtimer_start_p)(struct hrtimer *timer, ktime_t tim,
+			      const enum hrtimer_mode mode);
+extern int (*hrtimer_cancel_p)(struct hrtimer *timer);
+
+#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,17) && defined(CONFIG_KALLSYMS)
+static inline void hrtimer_kallsyms_resolve(void)
+{
+	hrtimer_init_p = (void *) kallsyms_lookup_name("hrtimer_init");
+	BUG_ON(!hrtimer_init_p);
+	hrtimer_start_p = (void *) kallsyms_lookup_name("hrtimer_start");
+	BUG_ON(!hrtimer_start_p);
+	hrtimer_cancel_p = (void *) kallsyms_lookup_name("hrtimer_cancel");
+	BUG_ON(!hrtimer_cancel_p);
+}
+#else
+static inline void hrtimer_kallsyms_resolve(void)
+{
+	hrtimer_init_p = hrtimer_init;
+	hrtimer_start_p = hrtimer_start;
+	hrtimer_cancel_p = hrtimer_cancel;
+}
+#endif
+
+/* handle old hrtimer API with data pointer */
+#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,17)
+static inline void hrtimer_data_pointer(struct hrtimer *timer)
+{
+	timer->data = (void *)timer;
+}
+#else
+static inline void hrtimer_data_pointer(struct hrtimer *timer) {}
+#endif
+
+#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,22)
+
+#define ns_to_timespec kvm_ns_to_timespec
+
+struct timespec kvm_ns_to_timespec(const s64 nsec);
+
+#endif
+
diff --git a/kernel/external-module-compat.h b/kernel/external-module-compat.h
deleted file mode 100644
index 49751fc..0000000
--- a/kernel/external-module-compat.h
+++ /dev/null
@@ -1,759 +0,0 @@
-
-/*
- * Compatibility header for building as an external module.
- */
-
-/*
- * Avoid picking up the kernel's kvm.h in case we have a newer one.
- */
-
-#include <linux/compiler.h>
-#include <linux/version.h>
-#include <linux/string.h>
-#include <linux/kvm.h>
-#include <linux/kvm_para.h>
-#include <linux/cpu.h>
-#include <linux/time.h>
-#include <asm/processor.h>
-#include <linux/hrtimer.h>
-#include <asm/bitops.h>
-#include <asm/msr.h>
-
-/*
- * 2.6.16 does not have GFP_NOWAIT
- */
-
-#include <linux/gfp.h>
-
-#ifndef GFP_NOWAIT
-#define GFP_NOWAIT (GFP_ATOMIC & ~__GFP_HIGH)
-#endif
-
-
-/*
- * kvm profiling support needs 2.6.20
- */
-#include <linux/profile.h>
-
-#ifndef KVM_PROFILING
-#define KVM_PROFILING 1234
-#define prof_on       4321
-#endif
-
-/*
- * smp_call_function_single() is not exported below 2.6.20, and has different
- * semantics below 2.6.23.  The 'nonatomic' argument was removed in 2.6.27.
- */
-#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,27)
-
-int kvm_smp_call_function_single(int cpu, void (*func)(void *info),
-				 void *info, int wait);
-
-#define smp_call_function_single kvm_smp_call_function_single
-
-#endif
-
-/* on_each_cpu() lost an argument in 2.6.27. */
-#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,27)
-
-#define kvm_on_each_cpu(func, info, wait) on_each_cpu(func, info, 0, wait)
-
-#else
-
-#define kvm_on_each_cpu(func, info, wait) on_each_cpu(func, info, wait)
-
-#endif
-
-/*
- * The cpu hotplug stubs are broken if !CONFIG_CPU_HOTPLUG
- */
-
-#if LINUX_VERSION_CODE <= KERNEL_VERSION(2,6,15)
-#define DEFINE_MUTEX(a) DECLARE_MUTEX(a)
-#define mutex_lock_interruptible(a) down_interruptible(a)
-#define mutex_unlock(a) up(a)
-#define mutex_lock(a) down(a)
-#define mutex_init(a) init_MUTEX(a)
-#define mutex_trylock(a) down_trylock(a)
-#define mutex semaphore
-#endif
-
-#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,14)
-#ifndef kzalloc
-#define kzalloc(size,flags)			\
-({						\
-	void *__ret = kmalloc(size, flags);	\
-	if (__ret)				\
-		memset(__ret, 0, size);		\
-	__ret;					\
-})
-#endif
-#endif
-
-#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,17)
-#ifndef kmem_cache_zalloc
-#define kmem_cache_zalloc(cache,flags)			  \
-({							  \
-	void *__ret = kmem_cache_alloc(cache, flags);	  \
-	if (__ret)                                        \
-		memset(__ret, 0, kmem_cache_size(cache)); \
-	__ret;                                            \
-})
-#endif
-#endif
-
-#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,21)
-
-#ifndef CONFIG_HOTPLUG_CPU
-#define register_cpu_notifier(nb) (0)
-#endif
-
-#endif
-
-#include <linux/miscdevice.h>
-#ifndef KVM_MINOR
-#define KVM_MINOR 232
-#endif
-
-#include <linux/notifier.h>
-#ifndef CPU_TASKS_FROZEN
-
-#define CPU_TASKS_FROZEN       0x0010
-#define CPU_ONLINE_FROZEN      (CPU_ONLINE | CPU_TASKS_FROZEN)
-#define CPU_UP_PREPARE_FROZEN  (CPU_UP_PREPARE | CPU_TASKS_FROZEN)
-#define CPU_UP_CANCELED_FROZEN (CPU_UP_CANCELED | CPU_TASKS_FROZEN)
-#define CPU_DOWN_PREPARE_FROZEN        (CPU_DOWN_PREPARE | CPU_TASKS_FROZEN)
-#define CPU_DOWN_FAILED_FROZEN (CPU_DOWN_FAILED | CPU_TASKS_FROZEN)
-#define CPU_DEAD_FROZEN                (CPU_DEAD | CPU_TASKS_FROZEN)
-
-#endif
-
-#ifndef CPU_DYING
-#define CPU_DYING 0x000A
-#define CPU_DYING_FROZEN (CPU_DYING | CPU_TASKS_FROZEN)
-#endif
-
-#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,22)
-
-#ifndef _EFER_SCE
-#define _EFER_SCE		0  /* SYSCALL/SYSRET */
-#endif
-
-#ifndef EFER_SCE
-#define EFER_SCE		(1<<_EFER_SCE)
-#endif
-
-#endif
-
-/*
- * For set_64bit(), which is in a new file asm/cmpxchg.h in newer kernels.
- */
-#include <asm/system.h>
-
-struct inode;
-#include <linux/anon_inodes.h>
-#define anon_inode_getfd kvm_anon_inode_getfd
-int kvm_init_anon_inodes(void);
-void kvm_exit_anon_inodes(void);
-int anon_inode_getfd(const char *name,
-		     const struct file_operations *fops,
-		     void *priv);
-
-#include <linux/smp.h>
-
-#ifndef X86_CR0_PE
-#define X86_CR0_PE 0x00000001
-#endif
-
-#ifndef X86_CR0_MP
-#define X86_CR0_MP 0x00000002
-#endif
-
-#ifndef X86_CR0_EM
-#define X86_CR0_EM 0x00000004
-#endif
-
-#ifndef X86_CR0_TS
-#define X86_CR0_TS 0x00000008
-#endif
-
-#ifndef X86_CR0_ET
-#define X86_CR0_ET 0x00000010
-#endif
-
-#ifndef X86_CR0_NE
-#define X86_CR0_NE 0x00000020
-#endif
-
-#ifndef X86_CR0_WP
-#define X86_CR0_WP 0x00010000
-#endif
-
-#ifndef X86_CR0_AM
-#define X86_CR0_AM 0x00040000
-#endif
-
-#ifndef X86_CR0_NW
-#define X86_CR0_NW 0x20000000
-#endif
-
-#ifndef X86_CR0_CD
-#define X86_CR0_CD 0x40000000
-#endif
-
-#ifndef X86_CR0_PG
-#define X86_CR0_PG 0x80000000
-#endif
-
-#ifndef X86_CR3_PWT
-#define X86_CR3_PWT 0x00000008
-#endif
-
-#ifndef X86_CR3_PCD
-#define X86_CR3_PCD 0x00000010
-#endif
-
-#ifndef X86_CR4_VMXE
-#define X86_CR4_VMXE 0x00002000
-#endif
-
-#undef X86_CR8_TPR
-#define X86_CR8_TPR 0x0f
-
-/*
- * 2.6.23 removed the cache destructor
- */
-#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,23)
-#  define kmem_cache_create(name, size, align, flags, ctor) \
-	kmem_cache_create(name, size, align, flags, ctor, NULL)
-#endif
-
-/*
- * 2.6.22 does not define set_64bit() under nonpae
- */
-#ifdef CONFIG_X86_32
-
-#include <asm/cmpxchg.h>
-
-static inline void __kvm_set_64bit(u64 *ptr, u64 val)
-{
-	unsigned int low = val;
-	unsigned int high = val >> 32;
-
-	__asm__ __volatile__ (
-		"\n1:\t"
-		"movl (%0), %%eax\n\t"
-		"movl 4(%0), %%edx\n\t"
-		"lock cmpxchg8b (%0)\n\t"
-		"jnz 1b"
-		: /* no outputs */
-		:	"D"(ptr),
-			"b"(low),
-			"c"(high)
-		:	"ax","dx","memory");
-}
-
-#undef  set_64bit
-#define set_64bit __kvm_set_64bit
-
-static inline unsigned long long __kvm_cmpxchg64(volatile void *ptr,
-						 unsigned long long old,
-						 unsigned long long new)
-{
-	unsigned long long prev;
-	__asm__ __volatile__("lock cmpxchg8b %3"
-			     : "=A"(prev)
-			     : "b"((unsigned long)new),
-			       "c"((unsigned long)(new >> 32)),
-			       "m"(*__xg(ptr)),
-			       "0"(old)
-			     : "memory");
-	return prev;
-}
-
-#define kvm_cmpxchg64(ptr,o,n)\
-	((__typeof__(*(ptr)))__kvm_cmpxchg64((ptr),(unsigned long long)(o),\
-					(unsigned long long)(n)))
-
-#undef cmpxchg64
-#define cmpxchg64(ptr, o, n) kvm_cmpxchg64(ptr, o, n)
-
-#endif
-
-#ifndef CONFIG_PREEMPT_NOTIFIERS
-/*
- * Include sched|preempt.h before defining CONFIG_PREEMPT_NOTIFIERS to avoid
- * a miscompile.
- */
-#include <linux/sched.h>
-#include <linux/preempt.h>
-#define CONFIG_PREEMPT_NOTIFIERS
-#define CONFIG_PREEMPT_NOTIFIERS_COMPAT
-
-struct preempt_notifier;
-
-struct preempt_ops {
-	void (*sched_in)(struct preempt_notifier *notifier, int cpu);
-	void (*sched_out)(struct preempt_notifier *notifier,
-			  struct task_struct *next);
-};
-
-struct preempt_notifier {
-	struct list_head link;
-	struct task_struct *tsk;
-	struct preempt_ops *ops;
-};
-
-void preempt_notifier_register(struct preempt_notifier *notifier);
-void preempt_notifier_unregister(struct preempt_notifier *notifier);
-
-static inline void preempt_notifier_init(struct preempt_notifier *notifier,
-				     struct preempt_ops *ops)
-{
-	notifier->ops = ops;
-}
-
-void start_special_insn(void);
-void end_special_insn(void);
-void in_special_section(void);
-void special_reload_dr7(void);
-
-void preempt_notifier_sys_init(void);
-void preempt_notifier_sys_exit(void);
-
-#else
-
-static inline void start_special_insn(void) {}
-static inline void end_special_insn(void) {}
-static inline void in_special_section(void) {}
-static inline void special_reload_dr7(void) {}
-
-static inline void preempt_notifier_sys_init(void) {}
-static inline void preempt_notifier_sys_exit(void) {}
-
-#endif
-
-/* HRTIMER_MODE_ABS started life with a different name */
-#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,21)
-#define HRTIMER_MODE_ABS HRTIMER_ABS
-#endif
-
-/* div64_u64 is fairly new */
-#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,26)
-
-#define div64_u64 kvm_div64_u64
-
-#ifdef CONFIG_64BIT
-
-static inline uint64_t div64_u64(uint64_t dividend, uint64_t divisor)
-{
-	return dividend / divisor;
-}
-
-#else
-
-uint64_t div64_u64(uint64_t dividend, uint64_t divisor);
-
-#endif
-
-#endif
-
-#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,19)
-
-#ifdef RHEL_RELEASE_CODE
-#if RHEL_RELEASE_CODE >= RHEL_RELEASE_VERSION(5,2)
-#define RHEL_BOOL 1
-#endif
-#endif
-
-#ifndef RHEL_BOOL
-
-typedef _Bool bool;
-
-#endif
-
-#endif
-
-/*
- * PF_VCPU is a Linux 2.6.24 addition
- */
-
-#include <linux/sched.h>
-
-#ifndef PF_VCPU
-#define PF_VCPU 0
-#endif
-
-/*
- * smp_call_function_mask() is not defined/exported below 2.6.24
- */
-
-#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,24)
-
-int kvm_smp_call_function_mask(cpumask_t mask, void (*func) (void *info),
-			       void *info, int wait);
-
-#define smp_call_function_mask kvm_smp_call_function_mask
-
-#endif
-
-/* CONFIG_HAS_IOMEM is apparently fairly new too (2.6.21 for x86_64). */
-#ifndef CONFIG_HAS_IOMEM
-#define CONFIG_HAS_IOMEM 1
-#endif
-
-/* empty_zero_page isn't exported in all kernels */
-#include <asm/pgtable.h>
-
-#define empty_zero_page kvm_empty_zero_page
-
-static char empty_zero_page[PAGE_SIZE];
-
-static inline void blahblah(void)
-{
-	(void)empty_zero_page[0];
-}
-
-/* __mmdrop() is not exported before 2.6.25 */
-#include <linux/sched.h>
-
-#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,25)
-
-#define mmdrop(x) do { (void)(x); } while (0)
-#define mmget(x) do { (void)(x); } while (0)
-
-#else
-
-#define mmget(x) do { atomic_inc(x); } while (0)
-
-#endif
-
-/* X86_FEATURE_NX is missing in some x86_64 kernels */
-
-#include <asm/cpufeature.h>
-
-#ifndef X86_FEATURE_NX
-#define X86_FEATURE_NX (1*32+20)
-#endif
-
-#undef true
-#define true 1
-#undef false
-#define false 0
-
-/* EFER_LMA and EFER_LME are missing in pre 2.6.24 i386 kernels */
-#ifndef EFER_LME
-#define _EFER_LME           8  /* Long mode enable */
-#define _EFER_LMA           10 /* Long mode active (read-only) */
-#define EFER_LME            (1<<_EFER_LME)
-#define EFER_LMA            (1<<_EFER_LMA)
-#endif
-
-/* pagefault_enable(), page_fault_disable() - 2.6.20 */
-#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,20)
-
-#define pagefault_enable()  do {} while (0)
-#define pagefault_disable() do {} while (0)
-
-#endif
-
-/* vm ops ->fault() was introduced in 2.6.23. */
-#include <linux/mm.h>
-
-#ifdef KVM_MAIN
-#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,23)
-
-struct vm_fault {
-	unsigned int flags;
-	pgoff_t pgoff;
-	void __user *virtual_address;
-	struct page *page;
-};
-
-static int kvm_vcpu_fault(struct vm_area_struct *vma, struct vm_fault *vmf);
-static int kvm_vm_fault(struct vm_area_struct *vma, struct vm_fault *vmf);
-
-static inline struct page *kvm_nopage_to_fault(
-	int (*fault)(struct vm_area_struct *vma, struct vm_fault *vmf),
-	struct vm_area_struct *vma,
-	unsigned long address,
-	int *type)
-{
-	struct vm_fault vmf;
-	int ret;
-
-	vmf.pgoff = ((address - vma->vm_start) >> PAGE_SHIFT) + vma->vm_pgoff;
-	vmf.virtual_address = (void __user *)address;
-	ret = fault(vma, &vmf);
-	if (ret)
-		return NOPAGE_SIGBUS;
-	*type = VM_FAULT_MINOR;
-	return vmf.page;
-}
-
-static inline struct page *__kvm_vcpu_fault(struct vm_area_struct *vma,
-					    unsigned long address,
-					    int *type)
-{
-	return kvm_nopage_to_fault(kvm_vcpu_fault, vma, address, type);
-}
-
-static inline struct page *__kvm_vm_fault(struct vm_area_struct *vma,
-					  unsigned long address,
-					  int *type)
-{
-	return kvm_nopage_to_fault(kvm_vm_fault, vma, address, type);
-}
-
-#define VMA_OPS_FAULT(x) nopage
-#define VMA_OPS_FAULT_FUNC(x) __##x
-
-#else
-
-#define VMA_OPS_FAULT(x) x
-#define VMA_OPS_FAULT_FUNC(x) x
-
-#endif
-#endif
-
-/* simple vfs attribute getter signature has changed to add a return code */
-
-#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,25)
-
-#define MAKE_SIMPLE_ATTRIBUTE_GETTER(x)       \
-	static u64 x(void *v)                 \
-	{				      \
-		u64 ret = 0;		      \
-					      \
-		__##x(v, &ret);		      \
-		return ret;		      \
-	}
-
-#else
-
-#define MAKE_SIMPLE_ATTRIBUTE_GETTER(x)       \
-	static int x(void *v, u64 *val)	      \
-	{				      \
-		return __##x(v, val);	      \
-	}
-
-#endif
-
-/* set_kset_name() is gone in 2.6.25 */
-
-#if LINUX_VERSION_CODE >= KERNEL_VERSION(2,6,25)
-
-#define set_kset_name(x) .name = x
-
-#endif
-
-#if LINUX_VERSION_CODE >= KERNEL_VERSION(2,6,25)
-#ifndef FASTCALL
-#define FASTCALL(x)	x
-#define fastcall
-#endif
-#endif
-
-#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,23)
-
-static unsigned  __attribute__((__used__)) kvm_tsc_khz = 2000000;
-
-#else
-
-#define kvm_tsc_khz tsc_khz
-
-#endif
-
-struct kvm_desc_struct {
-	union {
-		struct { unsigned int a, b; };
-		struct {
-			u16 limit0;
-			u16 base0;
-			unsigned base1: 8, type: 4, s: 1, dpl: 2, p: 1;
-			unsigned limit: 4, avl: 1, l: 1, d: 1, g: 1, base2: 8;
-		};
-
-	};
-} __attribute__((packed));
-
-struct kvm_ldttss_desc64 {
-	u16 limit0;
-	u16 base0;
-	unsigned base1 : 8, type : 5, dpl : 2, p : 1;
-	unsigned limit1 : 4, zero0 : 3, g : 1, base2 : 8;
-	u32 base3;
-	u32 zero1;
-} __attribute__((packed));
-
-struct kvm_desc_ptr {
-	unsigned short size;
-	unsigned long address;
-} __attribute__((packed));
-
-#if LINUX_VERSION_CODE <= KERNEL_VERSION(2,6,21)
-
-#include <linux/ktime.h>
-#include <linux/hrtimer.h>
-
-#define ktime_get kvm_ktime_get
-
-static inline ktime_t ktime_get(void)
-{
-	struct timespec now;
-
-	ktime_get_ts(&now);
-
-	return timespec_to_ktime(now);
-}
-
-#endif
-
-/* __aligned arrived in 2.6.21 */
-#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,21)
-#define __aligned(x) __attribute__((__aligned__(x)))
-#endif
-
-#include <asm/msr.h>
-#ifndef MSR_FS_BASE
-#define MSR_FS_BASE 0xc0000100
-#endif
-#ifndef MSR_GS_BASE
-#define MSR_GS_BASE 0xc0000101
-#endif
-
-#include <linux/mm.h>
-
-/* The shrinker API changed in 2.6.23 */
-#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,23)
-
-struct kvm_shrinker {
-	int (*shrink)(int nr_to_scan, gfp_t gfp_mask);
-	int seeks;
-	struct shrinker *kshrinker;
-};
-
-static inline void register_shrinker(struct kvm_shrinker *shrinker)
-{
-	shrinker->kshrinker = set_shrinker(shrinker->seeks, shrinker->shrink);
-}
-
-static inline void unregister_shrinker(struct kvm_shrinker *shrinker)
-{
-	if (shrinker->kshrinker)
-		remove_shrinker(shrinker->kshrinker);
-}
-
-#define shrinker kvm_shrinker
-
-#endif
-
-/* undefine lapic */
-#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,18)
-
-#undef lapic
-
-#endif
-
-/* clocksource */
-#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,18)
-static inline u32 clocksource_khz2mult(u32 khz, u32 shift_constant)
-{
-	/*  khz = cyc/(Million ns)
-	 *  mult/2^shift  = ns/cyc
-	 *  mult = ns/cyc * 2^shift
-	 *  mult = 1Million/khz * 2^shift
-	 *  mult = 1000000 * 2^shift / khz
-	 *  mult = (1000000<<shift) / khz
-	 */
-	u64 tmp = ((u64)1000000) << shift_constant;
-
-	tmp += khz/2; /* round for do_div */
-	do_div(tmp, khz);
-
-	return (u32)tmp;
-}
-#else
-#include <linux/clocksource.h>
-#endif
-
-/* manually export hrtimer_init/start/cancel */
-#include <linux/kallsyms.h>
-extern void (*hrtimer_init_p)(struct hrtimer *timer, clockid_t which_clock,
-			      enum hrtimer_mode mode);
-extern int (*hrtimer_start_p)(struct hrtimer *timer, ktime_t tim,
-			      const enum hrtimer_mode mode);
-extern int (*hrtimer_cancel_p)(struct hrtimer *timer);
-
-#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,17) && defined(CONFIG_KALLSYMS)
-static inline void hrtimer_kallsyms_resolve(void)
-{
-	hrtimer_init_p = (void *) kallsyms_lookup_name("hrtimer_init");
-	BUG_ON(!hrtimer_init_p);
-	hrtimer_start_p = (void *) kallsyms_lookup_name("hrtimer_start");
-	BUG_ON(!hrtimer_start_p);
-	hrtimer_cancel_p = (void *) kallsyms_lookup_name("hrtimer_cancel");
-	BUG_ON(!hrtimer_cancel_p);
-}
-#else
-static inline void hrtimer_kallsyms_resolve(void)
-{
-	hrtimer_init_p = hrtimer_init;
-	hrtimer_start_p = hrtimer_start;
-	hrtimer_cancel_p = hrtimer_cancel;
-}
-#endif
-
-/* handle old hrtimer API with data pointer */
-#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,17)
-static inline void hrtimer_data_pointer(struct hrtimer *timer)
-{
-	timer->data = (void *)timer;
-}
-#else
-static inline void hrtimer_data_pointer(struct hrtimer *timer) {}
-#endif
-
-#include <asm/hw_irq.h>
-#ifndef NMI_VECTOR
-#define NMI_VECTOR 2
-#endif
-
-#ifndef MSR_MTRRcap
-#define MSR_MTRRcap            0x0fe
-#define MSR_MTRRfix64K_00000   0x250
-#define MSR_MTRRfix16K_80000   0x258
-#define MSR_MTRRfix16K_A0000   0x259
-#define MSR_MTRRfix4K_C0000    0x268
-#define MSR_MTRRfix4K_C8000    0x269
-#define MSR_MTRRfix4K_D0000    0x26a
-#define MSR_MTRRfix4K_D8000    0x26b
-#define MSR_MTRRfix4K_E0000    0x26c
-#define MSR_MTRRfix4K_E8000    0x26d
-#define MSR_MTRRfix4K_F0000    0x26e
-#define MSR_MTRRfix4K_F8000    0x26f
-#define MSR_MTRRdefType        0x2ff
-#endif
-
-#ifndef MSR_IA32_CR_PAT
-#define MSR_IA32_CR_PAT        0x00000277
-#endif
-
-#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,22)
-
-#define ns_to_timespec kvm_ns_to_timespec
-
-struct timespec kvm_ns_to_timespec(const s64 nsec);
-
-#endif
-
-/* Define DEBUGCTLMSR bits */
-#ifndef DEBUGCTLMSR_LBR
-
-#define _DEBUGCTLMSR_LBR	0 /* last branch recording */
-#define _DEBUGCTLMSR_BTF	1 /* single-step on branches */
-
-#define DEBUGCTLMSR_LBR		(1UL << _DEBUGCTLMSR_LBR)
-#define DEBUGCTLMSR_BTF		(1UL << _DEBUGCTLMSR_BTF)
-
-#endif
diff --git a/kernel/debug.h b/kernel/x86/debug.h
similarity index 100%
rename from kernel/debug.h
rename to kernel/x86/debug.h
diff --git a/kernel/external-module-compat.c b/kernel/x86/external-module-compat.c
similarity index 100%
rename from kernel/external-module-compat.c
rename to kernel/x86/external-module-compat.c
diff --git a/kernel/x86/external-module-compat.h b/kernel/x86/external-module-compat.h
new file mode 100644
index 0000000..ce93551
--- /dev/null
+++ b/kernel/x86/external-module-compat.h
@@ -0,0 +1,291 @@
+
+/*
+ * Compatibility header for building as an external module.
+ */
+
+#include <linux/compiler.h>
+#include <linux/version.h>
+
+#include "../external-module-compat-comm.h"
+
+#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,22)
+
+#ifndef _EFER_SCE
+#define _EFER_SCE		0  /* SYSCALL/SYSRET */
+#endif
+
+#ifndef EFER_SCE
+#define EFER_SCE		(1<<_EFER_SCE)
+#endif
+
+#endif
+
+#include <linux/smp.h>
+
+#ifndef X86_CR0_PE
+#define X86_CR0_PE 0x00000001
+#endif
+
+#ifndef X86_CR0_MP
+#define X86_CR0_MP 0x00000002
+#endif
+
+#ifndef X86_CR0_EM
+#define X86_CR0_EM 0x00000004
+#endif
+
+#ifndef X86_CR0_TS
+#define X86_CR0_TS 0x00000008
+#endif
+
+#ifndef X86_CR0_ET
+#define X86_CR0_ET 0x00000010
+#endif
+
+#ifndef X86_CR0_NE
+#define X86_CR0_NE 0x00000020
+#endif
+
+#ifndef X86_CR0_WP
+#define X86_CR0_WP 0x00010000
+#endif
+
+#ifndef X86_CR0_AM
+#define X86_CR0_AM 0x00040000
+#endif
+
+#ifndef X86_CR0_NW
+#define X86_CR0_NW 0x20000000
+#endif
+
+#ifndef X86_CR0_CD
+#define X86_CR0_CD 0x40000000
+#endif
+
+#ifndef X86_CR0_PG
+#define X86_CR0_PG 0x80000000
+#endif
+
+#ifndef X86_CR3_PWT
+#define X86_CR3_PWT 0x00000008
+#endif
+
+#ifndef X86_CR3_PCD
+#define X86_CR3_PCD 0x00000010
+#endif
+
+#ifndef X86_CR4_VMXE
+#define X86_CR4_VMXE 0x00002000
+#endif
+
+#undef X86_CR8_TPR
+#define X86_CR8_TPR 0x0f
+
+/*
+ * 2.6.22 does not define set_64bit() under nonpae
+ */
+#ifdef CONFIG_X86_32
+
+#include <asm/cmpxchg.h>
+
+static inline void __kvm_set_64bit(u64 *ptr, u64 val)
+{
+	unsigned int low = val;
+	unsigned int high = val >> 32;
+
+	__asm__ __volatile__ (
+		"\n1:\t"
+		"movl (%0), %%eax\n\t"
+		"movl 4(%0), %%edx\n\t"
+		"lock cmpxchg8b (%0)\n\t"
+		"jnz 1b"
+		: /* no outputs */
+		:	"D"(ptr),
+			"b"(low),
+			"c"(high)
+		:	"ax","dx","memory");
+}
+
+#undef  set_64bit
+#define set_64bit __kvm_set_64bit
+
+static inline unsigned long long __kvm_cmpxchg64(volatile void *ptr,
+						 unsigned long long old,
+						 unsigned long long new)
+{
+	unsigned long long prev;
+	__asm__ __volatile__("lock cmpxchg8b %3"
+			     : "=A"(prev)
+			     : "b"((unsigned long)new),
+			       "c"((unsigned long)(new >> 32)),
+			       "m"(*__xg(ptr)),
+			       "0"(old)
+			     : "memory");
+	return prev;
+}
+
+#define kvm_cmpxchg64(ptr,o,n)\
+	((__typeof__(*(ptr)))__kvm_cmpxchg64((ptr),(unsigned long long)(o),\
+					(unsigned long long)(n)))
+
+#undef cmpxchg64
+#define cmpxchg64(ptr, o, n) kvm_cmpxchg64(ptr, o, n)
+
+#endif
+
+#ifndef CONFIG_PREEMPT_NOTIFIERS
+/*
+ * Include sched|preempt.h before defining CONFIG_PREEMPT_NOTIFIERS to avoid
+ * a miscompile.
+ */
+#include <linux/sched.h>
+#include <linux/preempt.h>
+#define CONFIG_PREEMPT_NOTIFIERS
+#define CONFIG_PREEMPT_NOTIFIERS_COMPAT
+
+struct preempt_notifier;
+
+struct preempt_ops {
+	void (*sched_in)(struct preempt_notifier *notifier, int cpu);
+	void (*sched_out)(struct preempt_notifier *notifier,
+			  struct task_struct *next);
+};
+
+struct preempt_notifier {
+	struct list_head link;
+	struct task_struct *tsk;
+	struct preempt_ops *ops;
+};
+
+void preempt_notifier_register(struct preempt_notifier *notifier);
+void preempt_notifier_unregister(struct preempt_notifier *notifier);
+
+static inline void preempt_notifier_init(struct preempt_notifier *notifier,
+				     struct preempt_ops *ops)
+{
+	notifier->ops = ops;
+}
+
+void start_special_insn(void);
+void end_special_insn(void);
+void in_special_section(void);
+void special_reload_dr7(void);
+
+void preempt_notifier_sys_init(void);
+void preempt_notifier_sys_exit(void);
+
+#else
+
+static inline void start_special_insn(void) {}
+static inline void end_special_insn(void) {}
+static inline void in_special_section(void) {}
+static inline void special_reload_dr7(void) {}
+
+static inline void preempt_notifier_sys_init(void) {}
+static inline void preempt_notifier_sys_exit(void) {}
+
+#endif
+
+/* CONFIG_HAS_IOMEM is apparently fairly new too (2.6.21 for x86_64). */
+#ifndef CONFIG_HAS_IOMEM
+#define CONFIG_HAS_IOMEM 1
+#endif
+
+/* X86_FEATURE_NX is missing in some x86_64 kernels */
+
+#include <asm/cpufeature.h>
+
+#ifndef X86_FEATURE_NX
+#define X86_FEATURE_NX (1*32+20)
+#endif
+
+#undef true
+#define true 1
+#undef false
+#define false 0
+
+/* EFER_LMA and EFER_LME are missing in pre 2.6.24 i386 kernels */
+#ifndef EFER_LME
+#define _EFER_LME           8  /* Long mode enable */
+#define _EFER_LMA           10 /* Long mode active (read-only) */
+#define EFER_LME            (1<<_EFER_LME)
+#define EFER_LMA            (1<<_EFER_LMA)
+#endif
+
+struct kvm_desc_struct {
+	union {
+		struct { unsigned int a, b; };
+		struct {
+			u16 limit0;
+			u16 base0;
+			unsigned base1: 8, type: 4, s: 1, dpl: 2, p: 1;
+			unsigned limit: 4, avl: 1, l: 1, d: 1, g: 1, base2: 8;
+		};
+
+	};
+} __attribute__((packed));
+
+struct kvm_ldttss_desc64 {
+	u16 limit0;
+	u16 base0;
+	unsigned base1 : 8, type : 5, dpl : 2, p : 1;
+	unsigned limit1 : 4, zero0 : 3, g : 1, base2 : 8;
+	u32 base3;
+	u32 zero1;
+} __attribute__((packed));
+
+struct kvm_desc_ptr {
+	unsigned short size;
+	unsigned long address;
+} __attribute__((packed));
+
+#include <asm/msr.h>
+#ifndef MSR_FS_BASE
+#define MSR_FS_BASE 0xc0000100
+#endif
+#ifndef MSR_GS_BASE
+#define MSR_GS_BASE 0xc0000101
+#endif
+
+/* undefine lapic */
+#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,18)
+
+#undef lapic
+
+#endif
+
+#include <asm/hw_irq.h>
+#ifndef NMI_VECTOR
+#define NMI_VECTOR 2
+#endif
+
+#ifndef MSR_MTRRcap
+#define MSR_MTRRcap            0x0fe
+#define MSR_MTRRfix64K_00000   0x250
+#define MSR_MTRRfix16K_80000   0x258
+#define MSR_MTRRfix16K_A0000   0x259
+#define MSR_MTRRfix4K_C0000    0x268
+#define MSR_MTRRfix4K_C8000    0x269
+#define MSR_MTRRfix4K_D0000    0x26a
+#define MSR_MTRRfix4K_D8000    0x26b
+#define MSR_MTRRfix4K_E0000    0x26c
+#define MSR_MTRRfix4K_E8000    0x26d
+#define MSR_MTRRfix4K_F0000    0x26e
+#define MSR_MTRRfix4K_F8000    0x26f
+#define MSR_MTRRdefType        0x2ff
+#endif
+
+#ifndef MSR_IA32_CR_PAT
+#define MSR_IA32_CR_PAT        0x00000277
+#endif
+
+/* Define DEBUGCTLMSR bits */
+#ifndef DEBUGCTLMSR_LBR
+
+#define _DEBUGCTLMSR_LBR	0 /* last branch recording */
+#define _DEBUGCTLMSR_BTF	1 /* single-step on branches */
+
+#define DEBUGCTLMSR_LBR		(1UL << _DEBUGCTLMSR_LBR)
+#define DEBUGCTLMSR_BTF		(1UL << _DEBUGCTLMSR_BTF)
+
+#endif
diff --git a/kernel/preempt.c b/kernel/x86/preempt.c
similarity index 100%
rename from kernel/preempt.c
rename to kernel/x86/preempt.c
diff --git a/kernel/vmx-debug.c b/kernel/x86/vmx-debug.c
similarity index 100%
rename from kernel/vmx-debug.c
rename to kernel/x86/vmx-debug.c
-- 
1.5.6.2


^ permalink raw reply related	[flat|nested] 24+ messages in thread

* RE: [PATCH 01/04]Create x86 directory to hold x86-specific files.
@ 2008-07-27  9:46   ` Zhang, Xiantao
  0 siblings, 0 replies; 24+ messages in thread
From: Zhang, Xiantao @ 2008-07-27  9:46 UTC (permalink / raw)
  To: Avi Kivity; +Cc: kvm-ia64, kvm

[-- Attachment #1: Type: text/plain, Size: 615 bytes --]

Avi Kivity wrote:
> Zhang, Xiantao wrote:
>> From c8c8c5a6b3ff943cb645f9ac4ef169490f14cd08 Mon Sep 17 00:00:00
>> 2001 From: Zhang Xiantao <xiantao.zhang@intel.com>
>> Date: Fri, 25 Jul 2008 16:50:30 +0800
>> Subject: [PATCH] KVM: external module: Moving x86-speicif files to
>> x86 directory. 
>> 
>> Create x86 directory to hold x86-specific files.
>> Signed-off-by: Zhang Xiantao <xiantao.zhang@intel.com>
>> 
> 
> You forgot to attach a patch; and your mailer corrupts patches, so I
> can't apply this.

Sorry. Attached! :)
> --
> error compiling committee.c: too many arguments to function


[-- Attachment #2: 0001-KVM-external-module-Moving-x86-speicif-files-to-x8.patch --]
[-- Type: application/octet-stream, Size: 38217 bytes --]

From c8c8c5a6b3ff943cb645f9ac4ef169490f14cd08 Mon Sep 17 00:00:00 2001
From: Zhang Xiantao <xiantao.zhang@intel.com>
Date: Fri, 25 Jul 2008 16:50:30 +0800
Subject: [PATCH] KVM: external module: Moving x86-speicif files to x86 directory.

Create x86 directory to hold x86-specific files.
Signed-off-by: Zhang Xiantao <xiantao.zhang@intel.com>
---
 kernel/Kbuild                             |    8 +-
 kernel/Makefile                           |    2 +-
 kernel/external-module-compat-comm.h      |  475 ++++++++++++++++++
 kernel/external-module-compat.h           |  759 -----------------------------
 kernel/{ => x86}/debug.h                  |    0 
 kernel/{ => x86}/external-module-compat.c |    0 
 kernel/x86/external-module-compat.h       |  291 +++++++++++
 kernel/{ => x86}/preempt.c                |    0 
 kernel/{ => x86}/vmx-debug.c              |    0 
 9 files changed, 771 insertions(+), 764 deletions(-)
 create mode 100644 kernel/external-module-compat-comm.h
 delete mode 100644 kernel/external-module-compat.h
 rename kernel/{ => x86}/debug.h (100%)
 rename kernel/{ => x86}/external-module-compat.c (100%)
 create mode 100644 kernel/x86/external-module-compat.h
 rename kernel/{ => x86}/preempt.c (100%)
 rename kernel/{ => x86}/vmx-debug.c (100%)

diff --git a/kernel/Kbuild b/kernel/Kbuild
index 676f28d..cb63582 100644
--- a/kernel/Kbuild
+++ b/kernel/Kbuild
@@ -1,11 +1,11 @@
 obj-m := kvm.o kvm-intel.o kvm-amd.o
 kvm-objs := kvm_main.o x86.o mmu.o x86_emulate.o anon_inodes.o irq.o i8259.o \
-	 lapic.o ioapic.o preempt.o i8254.o coalesced_mmio.o \
-	 external-module-compat.o
+	 lapic.o ioapic.o x86/preempt.o i8254.o coalesced_mmio.o \
+	 x86/external-module-compat.o
 ifeq ($(CONFIG_KVM_TRACE),y)
 kvm-objs += kvm_trace.o
 endif
-kvm-intel-objs := vmx.o vmx-debug.o external-module-compat.o
-kvm-amd-objs := svm.o external-module-compat.o
+kvm-intel-objs := vmx.o x86/vmx-debug.o x86/external-module-compat.o
+kvm-amd-objs := svm.o x86/external-module-compat.o
 
 CFLAGS_kvm_main.o = -DKVM_MAIN
diff --git a/kernel/Makefile b/kernel/Makefile
index 69b4981..c2c666e 100644
--- a/kernel/Makefile
+++ b/kernel/Makefile
@@ -35,7 +35,7 @@ all::
 	$(MAKE) -C $(KERNELDIR) M=`pwd` \
 		LINUXINCLUDE="-I`pwd`/include -Iinclude -I`pwd`/include-compat \
 		-include include/linux/autoconf.h \
-		-include `pwd`/external-module-compat.h"
+		-include `pwd`/$(ARCH_DIR)/external-module-compat.h"
 		"$$@"
 
 sync: header-sync source-sync
diff --git a/kernel/external-module-compat-comm.h b/kernel/external-module-compat-comm.h
new file mode 100644
index 0000000..d4bf395
--- /dev/null
+++ b/kernel/external-module-compat-comm.h
@@ -0,0 +1,475 @@
+
+/*
+ * Compatibility header for building as an external module.
+ */
+
+/*
+ * Avoid picking up the kernel's kvm.h in case we have a newer one.
+ */
+
+#include <linux/compiler.h>
+#include <linux/version.h>
+#include <linux/string.h>
+#include <linux/kvm.h>
+#include <linux/kvm_para.h>
+#include <linux/cpu.h>
+#include <linux/time.h>
+#include <asm/processor.h>
+#include <linux/hrtimer.h>
+#include <asm/bitops.h>
+#include <asm/msr.h>
+
+/*
+ * 2.6.16 does not have GFP_NOWAIT
+ */
+
+#include <linux/gfp.h>
+
+#ifndef GFP_NOWAIT
+#define GFP_NOWAIT (GFP_ATOMIC & ~__GFP_HIGH)
+#endif
+
+
+/*
+ * kvm profiling support needs 2.6.20
+ */
+#include <linux/profile.h>
+
+#ifndef KVM_PROFILING
+#define KVM_PROFILING 1234
+#define prof_on       4321
+#endif
+
+/*
+ * smp_call_function_single() is not exported below 2.6.20, and has different
+ * semantics below 2.6.23.  The 'nonatomic' argument was removed in 2.6.27.
+ */
+#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,27)
+
+int kvm_smp_call_function_single(int cpu, void (*func)(void *info),
+				 void *info, int wait);
+
+#define smp_call_function_single kvm_smp_call_function_single
+
+#endif
+
+/* on_each_cpu() lost an argument in 2.6.27. */
+#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,27)
+
+#define kvm_on_each_cpu(func, info, wait) on_each_cpu(func, info, 0, wait)
+
+#else
+
+#define kvm_on_each_cpu(func, info, wait) on_each_cpu(func, info, wait)
+
+#endif
+
+/*
+ * The cpu hotplug stubs are broken if !CONFIG_CPU_HOTPLUG
+ */
+
+#if LINUX_VERSION_CODE <= KERNEL_VERSION(2,6,15)
+#define DEFINE_MUTEX(a) DECLARE_MUTEX(a)
+#define mutex_lock_interruptible(a) down_interruptible(a)
+#define mutex_unlock(a) up(a)
+#define mutex_lock(a) down(a)
+#define mutex_init(a) init_MUTEX(a)
+#define mutex_trylock(a) down_trylock(a)
+#define mutex semaphore
+#endif
+
+#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,14)
+#ifndef kzalloc
+#define kzalloc(size,flags)			\
+({						\
+	void *__ret = kmalloc(size, flags);	\
+	if (__ret)				\
+		memset(__ret, 0, size);		\
+	__ret;					\
+})
+#endif
+#endif
+
+#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,17)
+#ifndef kmem_cache_zalloc
+#define kmem_cache_zalloc(cache,flags)			  \
+({							  \
+	void *__ret = kmem_cache_alloc(cache, flags);	  \
+	if (__ret)                                        \
+		memset(__ret, 0, kmem_cache_size(cache)); \
+	__ret;                                            \
+})
+#endif
+#endif
+
+#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,21)
+
+#ifndef CONFIG_HOTPLUG_CPU
+#define register_cpu_notifier(nb) (0)
+#endif
+
+#endif
+
+#include <linux/miscdevice.h>
+#ifndef KVM_MINOR
+#define KVM_MINOR 232
+#endif
+
+#include <linux/notifier.h>
+#ifndef CPU_TASKS_FROZEN
+
+#define CPU_TASKS_FROZEN       0x0010
+#define CPU_ONLINE_FROZEN      (CPU_ONLINE | CPU_TASKS_FROZEN)
+#define CPU_UP_PREPARE_FROZEN  (CPU_UP_PREPARE | CPU_TASKS_FROZEN)
+#define CPU_UP_CANCELED_FROZEN (CPU_UP_CANCELED | CPU_TASKS_FROZEN)
+#define CPU_DOWN_PREPARE_FROZEN        (CPU_DOWN_PREPARE | CPU_TASKS_FROZEN)
+#define CPU_DOWN_FAILED_FROZEN (CPU_DOWN_FAILED | CPU_TASKS_FROZEN)
+#define CPU_DEAD_FROZEN                (CPU_DEAD | CPU_TASKS_FROZEN)
+
+#endif
+
+#ifndef CPU_DYING
+#define CPU_DYING 0x000A
+#define CPU_DYING_FROZEN (CPU_DYING | CPU_TASKS_FROZEN)
+#endif
+
+#include <asm/system.h>
+
+struct inode;
+#include <linux/anon_inodes.h>
+#define anon_inode_getfd kvm_anon_inode_getfd
+int kvm_init_anon_inodes(void);
+void kvm_exit_anon_inodes(void);
+int anon_inode_getfd(const char *name,
+		     const struct file_operations *fops,
+		     void *priv);
+
+/*
+ * 2.6.23 removed the cache destructor
+ */
+#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,23)
+#  define kmem_cache_create(name, size, align, flags, ctor) \
+	kmem_cache_create(name, size, align, flags, ctor, NULL)
+#endif
+
+/* HRTIMER_MODE_ABS started life with a different name */
+#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,21)
+#define HRTIMER_MODE_ABS HRTIMER_ABS
+#endif
+
+/* div64_u64 is fairly new */
+#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,26)
+
+#define div64_u64 kvm_div64_u64
+
+#ifdef CONFIG_64BIT
+
+static inline uint64_t div64_u64(uint64_t dividend, uint64_t divisor)
+{
+	return dividend / divisor;
+}
+
+#else
+
+uint64_t div64_u64(uint64_t dividend, uint64_t divisor);
+
+#endif
+
+#endif
+
+#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,19)
+
+#ifdef RHEL_RELEASE_CODE
+#if RHEL_RELEASE_CODE >= RHEL_RELEASE_VERSION(5,2)
+#define RHEL_BOOL 1
+#endif
+#endif
+
+#ifndef RHEL_BOOL
+
+typedef _Bool bool;
+
+#endif
+
+#endif
+
+/*
+ * PF_VCPU is a Linux 2.6.24 addition
+ */
+
+#include <linux/sched.h>
+
+#ifndef PF_VCPU
+#define PF_VCPU 0
+#endif
+
+/*
+ * smp_call_function_mask() is not defined/exported below 2.6.24
+ */
+
+#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,24)
+
+int kvm_smp_call_function_mask(cpumask_t mask, void (*func) (void *info),
+			       void *info, int wait);
+
+#define smp_call_function_mask kvm_smp_call_function_mask
+
+#endif
+
+/* empty_zero_page isn't exported in all kernels */
+#include <asm/pgtable.h>
+
+#define empty_zero_page kvm_empty_zero_page
+
+static char empty_zero_page[PAGE_SIZE];
+
+static inline void blahblah(void)
+{
+	(void)empty_zero_page[0];
+}
+
+/* __mmdrop() is not exported before 2.6.25 */
+#include <linux/sched.h>
+
+#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,25)
+
+#define mmdrop(x) do { (void)(x); } while (0)
+#define mmget(x) do { (void)(x); } while (0)
+
+#else
+
+#define mmget(x) do { atomic_inc(x); } while (0)
+
+#endif
+
+/* pagefault_enable(), page_fault_disable() - 2.6.20 */
+#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,20)
+
+#define pagefault_enable()  do {} while (0)
+#define pagefault_disable() do {} while (0)
+
+#endif
+
+/* vm ops ->fault() was introduced in 2.6.23. */
+#include <linux/mm.h>
+
+#ifdef KVM_MAIN
+#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,23)
+
+struct vm_fault {
+	unsigned int flags;
+	pgoff_t pgoff;
+	void __user *virtual_address;
+	struct page *page;
+};
+
+static int kvm_vcpu_fault(struct vm_area_struct *vma, struct vm_fault *vmf);
+static int kvm_vm_fault(struct vm_area_struct *vma, struct vm_fault *vmf);
+
+static inline struct page *kvm_nopage_to_fault(
+	int (*fault)(struct vm_area_struct *vma, struct vm_fault *vmf),
+	struct vm_area_struct *vma,
+	unsigned long address,
+	int *type)
+{
+	struct vm_fault vmf;
+	int ret;
+
+	vmf.pgoff = ((address - vma->vm_start) >> PAGE_SHIFT) + vma->vm_pgoff;
+	vmf.virtual_address = (void __user *)address;
+	ret = fault(vma, &vmf);
+	if (ret)
+		return NOPAGE_SIGBUS;
+	*type = VM_FAULT_MINOR;
+	return vmf.page;
+}
+
+static inline struct page *__kvm_vcpu_fault(struct vm_area_struct *vma,
+					    unsigned long address,
+					    int *type)
+{
+	return kvm_nopage_to_fault(kvm_vcpu_fault, vma, address, type);
+}
+
+static inline struct page *__kvm_vm_fault(struct vm_area_struct *vma,
+					  unsigned long address,
+					  int *type)
+{
+	return kvm_nopage_to_fault(kvm_vm_fault, vma, address, type);
+}
+
+#define VMA_OPS_FAULT(x) nopage
+#define VMA_OPS_FAULT_FUNC(x) __##x
+
+#else
+
+#define VMA_OPS_FAULT(x) x
+#define VMA_OPS_FAULT_FUNC(x) x
+
+#endif
+#endif
+
+/* simple vfs attribute getter signature has changed to add a return code */
+
+#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,25)
+
+#define MAKE_SIMPLE_ATTRIBUTE_GETTER(x)       \
+	static u64 x(void *v)                 \
+	{				      \
+		u64 ret = 0;		      \
+					      \
+		__##x(v, &ret);		      \
+		return ret;		      \
+	}
+
+#else
+
+#define MAKE_SIMPLE_ATTRIBUTE_GETTER(x)       \
+	static int x(void *v, u64 *val)	      \
+	{				      \
+		return __##x(v, val);	      \
+	}
+
+#endif
+
+/* set_kset_name() is gone in 2.6.25 */
+
+#if LINUX_VERSION_CODE >= KERNEL_VERSION(2,6,25)
+
+#define set_kset_name(x) .name = x
+
+#endif
+
+#if LINUX_VERSION_CODE >= KERNEL_VERSION(2,6,25)
+#ifndef FASTCALL
+#define FASTCALL(x)	x
+#define fastcall
+#endif
+#endif
+
+#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,23)
+
+static unsigned  __attribute__((__used__)) kvm_tsc_khz = 2000000;
+
+#else
+
+#define kvm_tsc_khz tsc_khz
+
+#endif
+
+#if LINUX_VERSION_CODE <= KERNEL_VERSION(2,6,21)
+
+#include <linux/ktime.h>
+#include <linux/hrtimer.h>
+
+#define ktime_get kvm_ktime_get
+
+static inline ktime_t ktime_get(void)
+{
+	struct timespec now;
+
+	ktime_get_ts(&now);
+
+	return timespec_to_ktime(now);
+}
+
+#endif
+
+/* __aligned arrived in 2.6.21 */
+#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,21)
+#define __aligned(x) __attribute__((__aligned__(x)))
+#endif
+
+#include <linux/mm.h>
+
+/* The shrinker API changed in 2.6.23 */
+#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,23)
+
+struct kvm_shrinker {
+	int (*shrink)(int nr_to_scan, gfp_t gfp_mask);
+	int seeks;
+	struct shrinker *kshrinker;
+};
+
+static inline void register_shrinker(struct kvm_shrinker *shrinker)
+{
+	shrinker->kshrinker = set_shrinker(shrinker->seeks, shrinker->shrink);
+}
+
+static inline void unregister_shrinker(struct kvm_shrinker *shrinker)
+{
+	if (shrinker->kshrinker)
+		remove_shrinker(shrinker->kshrinker);
+}
+
+#define shrinker kvm_shrinker
+
+#endif
+
+/* clocksource */
+#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,18)
+static inline u32 clocksource_khz2mult(u32 khz, u32 shift_constant)
+{
+	/*  khz = cyc/(Million ns)
+	 *  mult/2^shift  = ns/cyc
+	 *  mult = ns/cyc * 2^shift
+	 *  mult = 1Million/khz * 2^shift
+	 *  mult = 1000000 * 2^shift / khz
+	 *  mult = (1000000<<shift) / khz
+	 */
+	u64 tmp = ((u64)1000000) << shift_constant;
+
+	tmp += khz/2; /* round for do_div */
+	do_div(tmp, khz);
+
+	return (u32)tmp;
+}
+#else
+#include <linux/clocksource.h>
+#endif
+
+/* manually export hrtimer_init/start/cancel */
+#include <linux/kallsyms.h>
+extern void (*hrtimer_init_p)(struct hrtimer *timer, clockid_t which_clock,
+			      enum hrtimer_mode mode);
+extern int (*hrtimer_start_p)(struct hrtimer *timer, ktime_t tim,
+			      const enum hrtimer_mode mode);
+extern int (*hrtimer_cancel_p)(struct hrtimer *timer);
+
+#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,17) && defined(CONFIG_KALLSYMS)
+static inline void hrtimer_kallsyms_resolve(void)
+{
+	hrtimer_init_p = (void *) kallsyms_lookup_name("hrtimer_init");
+	BUG_ON(!hrtimer_init_p);
+	hrtimer_start_p = (void *) kallsyms_lookup_name("hrtimer_start");
+	BUG_ON(!hrtimer_start_p);
+	hrtimer_cancel_p = (void *) kallsyms_lookup_name("hrtimer_cancel");
+	BUG_ON(!hrtimer_cancel_p);
+}
+#else
+static inline void hrtimer_kallsyms_resolve(void)
+{
+	hrtimer_init_p = hrtimer_init;
+	hrtimer_start_p = hrtimer_start;
+	hrtimer_cancel_p = hrtimer_cancel;
+}
+#endif
+
+/* handle old hrtimer API with data pointer */
+#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,17)
+static inline void hrtimer_data_pointer(struct hrtimer *timer)
+{
+	timer->data = (void *)timer;
+}
+#else
+static inline void hrtimer_data_pointer(struct hrtimer *timer) {}
+#endif
+
+#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,22)
+
+#define ns_to_timespec kvm_ns_to_timespec
+
+struct timespec kvm_ns_to_timespec(const s64 nsec);
+
+#endif
+
diff --git a/kernel/external-module-compat.h b/kernel/external-module-compat.h
deleted file mode 100644
index 49751fc..0000000
--- a/kernel/external-module-compat.h
+++ /dev/null
@@ -1,759 +0,0 @@
-
-/*
- * Compatibility header for building as an external module.
- */
-
-/*
- * Avoid picking up the kernel's kvm.h in case we have a newer one.
- */
-
-#include <linux/compiler.h>
-#include <linux/version.h>
-#include <linux/string.h>
-#include <linux/kvm.h>
-#include <linux/kvm_para.h>
-#include <linux/cpu.h>
-#include <linux/time.h>
-#include <asm/processor.h>
-#include <linux/hrtimer.h>
-#include <asm/bitops.h>
-#include <asm/msr.h>
-
-/*
- * 2.6.16 does not have GFP_NOWAIT
- */
-
-#include <linux/gfp.h>
-
-#ifndef GFP_NOWAIT
-#define GFP_NOWAIT (GFP_ATOMIC & ~__GFP_HIGH)
-#endif
-
-
-/*
- * kvm profiling support needs 2.6.20
- */
-#include <linux/profile.h>
-
-#ifndef KVM_PROFILING
-#define KVM_PROFILING 1234
-#define prof_on       4321
-#endif
-
-/*
- * smp_call_function_single() is not exported below 2.6.20, and has different
- * semantics below 2.6.23.  The 'nonatomic' argument was removed in 2.6.27.
- */
-#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,27)
-
-int kvm_smp_call_function_single(int cpu, void (*func)(void *info),
-				 void *info, int wait);
-
-#define smp_call_function_single kvm_smp_call_function_single
-
-#endif
-
-/* on_each_cpu() lost an argument in 2.6.27. */
-#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,27)
-
-#define kvm_on_each_cpu(func, info, wait) on_each_cpu(func, info, 0, wait)
-
-#else
-
-#define kvm_on_each_cpu(func, info, wait) on_each_cpu(func, info, wait)
-
-#endif
-
-/*
- * The cpu hotplug stubs are broken if !CONFIG_CPU_HOTPLUG
- */
-
-#if LINUX_VERSION_CODE <= KERNEL_VERSION(2,6,15)
-#define DEFINE_MUTEX(a) DECLARE_MUTEX(a)
-#define mutex_lock_interruptible(a) down_interruptible(a)
-#define mutex_unlock(a) up(a)
-#define mutex_lock(a) down(a)
-#define mutex_init(a) init_MUTEX(a)
-#define mutex_trylock(a) down_trylock(a)
-#define mutex semaphore
-#endif
-
-#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,14)
-#ifndef kzalloc
-#define kzalloc(size,flags)			\
-({						\
-	void *__ret = kmalloc(size, flags);	\
-	if (__ret)				\
-		memset(__ret, 0, size);		\
-	__ret;					\
-})
-#endif
-#endif
-
-#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,17)
-#ifndef kmem_cache_zalloc
-#define kmem_cache_zalloc(cache,flags)			  \
-({							  \
-	void *__ret = kmem_cache_alloc(cache, flags);	  \
-	if (__ret)                                        \
-		memset(__ret, 0, kmem_cache_size(cache)); \
-	__ret;                                            \
-})
-#endif
-#endif
-
-#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,21)
-
-#ifndef CONFIG_HOTPLUG_CPU
-#define register_cpu_notifier(nb) (0)
-#endif
-
-#endif
-
-#include <linux/miscdevice.h>
-#ifndef KVM_MINOR
-#define KVM_MINOR 232
-#endif
-
-#include <linux/notifier.h>
-#ifndef CPU_TASKS_FROZEN
-
-#define CPU_TASKS_FROZEN       0x0010
-#define CPU_ONLINE_FROZEN      (CPU_ONLINE | CPU_TASKS_FROZEN)
-#define CPU_UP_PREPARE_FROZEN  (CPU_UP_PREPARE | CPU_TASKS_FROZEN)
-#define CPU_UP_CANCELED_FROZEN (CPU_UP_CANCELED | CPU_TASKS_FROZEN)
-#define CPU_DOWN_PREPARE_FROZEN        (CPU_DOWN_PREPARE | CPU_TASKS_FROZEN)
-#define CPU_DOWN_FAILED_FROZEN (CPU_DOWN_FAILED | CPU_TASKS_FROZEN)
-#define CPU_DEAD_FROZEN                (CPU_DEAD | CPU_TASKS_FROZEN)
-
-#endif
-
-#ifndef CPU_DYING
-#define CPU_DYING 0x000A
-#define CPU_DYING_FROZEN (CPU_DYING | CPU_TASKS_FROZEN)
-#endif
-
-#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,22)
-
-#ifndef _EFER_SCE
-#define _EFER_SCE		0  /* SYSCALL/SYSRET */
-#endif
-
-#ifndef EFER_SCE
-#define EFER_SCE		(1<<_EFER_SCE)
-#endif
-
-#endif
-
-/*
- * For set_64bit(), which is in a new file asm/cmpxchg.h in newer kernels.
- */
-#include <asm/system.h>
-
-struct inode;
-#include <linux/anon_inodes.h>
-#define anon_inode_getfd kvm_anon_inode_getfd
-int kvm_init_anon_inodes(void);
-void kvm_exit_anon_inodes(void);
-int anon_inode_getfd(const char *name,
-		     const struct file_operations *fops,
-		     void *priv);
-
-#include <linux/smp.h>
-
-#ifndef X86_CR0_PE
-#define X86_CR0_PE 0x00000001
-#endif
-
-#ifndef X86_CR0_MP
-#define X86_CR0_MP 0x00000002
-#endif
-
-#ifndef X86_CR0_EM
-#define X86_CR0_EM 0x00000004
-#endif
-
-#ifndef X86_CR0_TS
-#define X86_CR0_TS 0x00000008
-#endif
-
-#ifndef X86_CR0_ET
-#define X86_CR0_ET 0x00000010
-#endif
-
-#ifndef X86_CR0_NE
-#define X86_CR0_NE 0x00000020
-#endif
-
-#ifndef X86_CR0_WP
-#define X86_CR0_WP 0x00010000
-#endif
-
-#ifndef X86_CR0_AM
-#define X86_CR0_AM 0x00040000
-#endif
-
-#ifndef X86_CR0_NW
-#define X86_CR0_NW 0x20000000
-#endif
-
-#ifndef X86_CR0_CD
-#define X86_CR0_CD 0x40000000
-#endif
-
-#ifndef X86_CR0_PG
-#define X86_CR0_PG 0x80000000
-#endif
-
-#ifndef X86_CR3_PWT
-#define X86_CR3_PWT 0x00000008
-#endif
-
-#ifndef X86_CR3_PCD
-#define X86_CR3_PCD 0x00000010
-#endif
-
-#ifndef X86_CR4_VMXE
-#define X86_CR4_VMXE 0x00002000
-#endif
-
-#undef X86_CR8_TPR
-#define X86_CR8_TPR 0x0f
-
-/*
- * 2.6.23 removed the cache destructor
- */
-#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,23)
-#  define kmem_cache_create(name, size, align, flags, ctor) \
-	kmem_cache_create(name, size, align, flags, ctor, NULL)
-#endif
-
-/*
- * 2.6.22 does not define set_64bit() under nonpae
- */
-#ifdef CONFIG_X86_32
-
-#include <asm/cmpxchg.h>
-
-static inline void __kvm_set_64bit(u64 *ptr, u64 val)
-{
-	unsigned int low = val;
-	unsigned int high = val >> 32;
-
-	__asm__ __volatile__ (
-		"\n1:\t"
-		"movl (%0), %%eax\n\t"
-		"movl 4(%0), %%edx\n\t"
-		"lock cmpxchg8b (%0)\n\t"
-		"jnz 1b"
-		: /* no outputs */
-		:	"D"(ptr),
-			"b"(low),
-			"c"(high)
-		:	"ax","dx","memory");
-}
-
-#undef  set_64bit
-#define set_64bit __kvm_set_64bit
-
-static inline unsigned long long __kvm_cmpxchg64(volatile void *ptr,
-						 unsigned long long old,
-						 unsigned long long new)
-{
-	unsigned long long prev;
-	__asm__ __volatile__("lock cmpxchg8b %3"
-			     : "=A"(prev)
-			     : "b"((unsigned long)new),
-			       "c"((unsigned long)(new >> 32)),
-			       "m"(*__xg(ptr)),
-			       "0"(old)
-			     : "memory");
-	return prev;
-}
-
-#define kvm_cmpxchg64(ptr,o,n)\
-	((__typeof__(*(ptr)))__kvm_cmpxchg64((ptr),(unsigned long long)(o),\
-					(unsigned long long)(n)))
-
-#undef cmpxchg64
-#define cmpxchg64(ptr, o, n) kvm_cmpxchg64(ptr, o, n)
-
-#endif
-
-#ifndef CONFIG_PREEMPT_NOTIFIERS
-/*
- * Include sched|preempt.h before defining CONFIG_PREEMPT_NOTIFIERS to avoid
- * a miscompile.
- */
-#include <linux/sched.h>
-#include <linux/preempt.h>
-#define CONFIG_PREEMPT_NOTIFIERS
-#define CONFIG_PREEMPT_NOTIFIERS_COMPAT
-
-struct preempt_notifier;
-
-struct preempt_ops {
-	void (*sched_in)(struct preempt_notifier *notifier, int cpu);
-	void (*sched_out)(struct preempt_notifier *notifier,
-			  struct task_struct *next);
-};
-
-struct preempt_notifier {
-	struct list_head link;
-	struct task_struct *tsk;
-	struct preempt_ops *ops;
-};
-
-void preempt_notifier_register(struct preempt_notifier *notifier);
-void preempt_notifier_unregister(struct preempt_notifier *notifier);
-
-static inline void preempt_notifier_init(struct preempt_notifier *notifier,
-				     struct preempt_ops *ops)
-{
-	notifier->ops = ops;
-}
-
-void start_special_insn(void);
-void end_special_insn(void);
-void in_special_section(void);
-void special_reload_dr7(void);
-
-void preempt_notifier_sys_init(void);
-void preempt_notifier_sys_exit(void);
-
-#else
-
-static inline void start_special_insn(void) {}
-static inline void end_special_insn(void) {}
-static inline void in_special_section(void) {}
-static inline void special_reload_dr7(void) {}
-
-static inline void preempt_notifier_sys_init(void) {}
-static inline void preempt_notifier_sys_exit(void) {}
-
-#endif
-
-/* HRTIMER_MODE_ABS started life with a different name */
-#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,21)
-#define HRTIMER_MODE_ABS HRTIMER_ABS
-#endif
-
-/* div64_u64 is fairly new */
-#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,26)
-
-#define div64_u64 kvm_div64_u64
-
-#ifdef CONFIG_64BIT
-
-static inline uint64_t div64_u64(uint64_t dividend, uint64_t divisor)
-{
-	return dividend / divisor;
-}
-
-#else
-
-uint64_t div64_u64(uint64_t dividend, uint64_t divisor);
-
-#endif
-
-#endif
-
-#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,19)
-
-#ifdef RHEL_RELEASE_CODE
-#if RHEL_RELEASE_CODE >= RHEL_RELEASE_VERSION(5,2)
-#define RHEL_BOOL 1
-#endif
-#endif
-
-#ifndef RHEL_BOOL
-
-typedef _Bool bool;
-
-#endif
-
-#endif
-
-/*
- * PF_VCPU is a Linux 2.6.24 addition
- */
-
-#include <linux/sched.h>
-
-#ifndef PF_VCPU
-#define PF_VCPU 0
-#endif
-
-/*
- * smp_call_function_mask() is not defined/exported below 2.6.24
- */
-
-#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,24)
-
-int kvm_smp_call_function_mask(cpumask_t mask, void (*func) (void *info),
-			       void *info, int wait);
-
-#define smp_call_function_mask kvm_smp_call_function_mask
-
-#endif
-
-/* CONFIG_HAS_IOMEM is apparently fairly new too (2.6.21 for x86_64). */
-#ifndef CONFIG_HAS_IOMEM
-#define CONFIG_HAS_IOMEM 1
-#endif
-
-/* empty_zero_page isn't exported in all kernels */
-#include <asm/pgtable.h>
-
-#define empty_zero_page kvm_empty_zero_page
-
-static char empty_zero_page[PAGE_SIZE];
-
-static inline void blahblah(void)
-{
-	(void)empty_zero_page[0];
-}
-
-/* __mmdrop() is not exported before 2.6.25 */
-#include <linux/sched.h>
-
-#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,25)
-
-#define mmdrop(x) do { (void)(x); } while (0)
-#define mmget(x) do { (void)(x); } while (0)
-
-#else
-
-#define mmget(x) do { atomic_inc(x); } while (0)
-
-#endif
-
-/* X86_FEATURE_NX is missing in some x86_64 kernels */
-
-#include <asm/cpufeature.h>
-
-#ifndef X86_FEATURE_NX
-#define X86_FEATURE_NX (1*32+20)
-#endif
-
-#undef true
-#define true 1
-#undef false
-#define false 0
-
-/* EFER_LMA and EFER_LME are missing in pre 2.6.24 i386 kernels */
-#ifndef EFER_LME
-#define _EFER_LME           8  /* Long mode enable */
-#define _EFER_LMA           10 /* Long mode active (read-only) */
-#define EFER_LME            (1<<_EFER_LME)
-#define EFER_LMA            (1<<_EFER_LMA)
-#endif
-
-/* pagefault_enable(), page_fault_disable() - 2.6.20 */
-#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,20)
-
-#define pagefault_enable()  do {} while (0)
-#define pagefault_disable() do {} while (0)
-
-#endif
-
-/* vm ops ->fault() was introduced in 2.6.23. */
-#include <linux/mm.h>
-
-#ifdef KVM_MAIN
-#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,23)
-
-struct vm_fault {
-	unsigned int flags;
-	pgoff_t pgoff;
-	void __user *virtual_address;
-	struct page *page;
-};
-
-static int kvm_vcpu_fault(struct vm_area_struct *vma, struct vm_fault *vmf);
-static int kvm_vm_fault(struct vm_area_struct *vma, struct vm_fault *vmf);
-
-static inline struct page *kvm_nopage_to_fault(
-	int (*fault)(struct vm_area_struct *vma, struct vm_fault *vmf),
-	struct vm_area_struct *vma,
-	unsigned long address,
-	int *type)
-{
-	struct vm_fault vmf;
-	int ret;
-
-	vmf.pgoff = ((address - vma->vm_start) >> PAGE_SHIFT) + vma->vm_pgoff;
-	vmf.virtual_address = (void __user *)address;
-	ret = fault(vma, &vmf);
-	if (ret)
-		return NOPAGE_SIGBUS;
-	*type = VM_FAULT_MINOR;
-	return vmf.page;
-}
-
-static inline struct page *__kvm_vcpu_fault(struct vm_area_struct *vma,
-					    unsigned long address,
-					    int *type)
-{
-	return kvm_nopage_to_fault(kvm_vcpu_fault, vma, address, type);
-}
-
-static inline struct page *__kvm_vm_fault(struct vm_area_struct *vma,
-					  unsigned long address,
-					  int *type)
-{
-	return kvm_nopage_to_fault(kvm_vm_fault, vma, address, type);
-}
-
-#define VMA_OPS_FAULT(x) nopage
-#define VMA_OPS_FAULT_FUNC(x) __##x
-
-#else
-
-#define VMA_OPS_FAULT(x) x
-#define VMA_OPS_FAULT_FUNC(x) x
-
-#endif
-#endif
-
-/* simple vfs attribute getter signature has changed to add a return code */
-
-#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,25)
-
-#define MAKE_SIMPLE_ATTRIBUTE_GETTER(x)       \
-	static u64 x(void *v)                 \
-	{				      \
-		u64 ret = 0;		      \
-					      \
-		__##x(v, &ret);		      \
-		return ret;		      \
-	}
-
-#else
-
-#define MAKE_SIMPLE_ATTRIBUTE_GETTER(x)       \
-	static int x(void *v, u64 *val)	      \
-	{				      \
-		return __##x(v, val);	      \
-	}
-
-#endif
-
-/* set_kset_name() is gone in 2.6.25 */
-
-#if LINUX_VERSION_CODE >= KERNEL_VERSION(2,6,25)
-
-#define set_kset_name(x) .name = x
-
-#endif
-
-#if LINUX_VERSION_CODE >= KERNEL_VERSION(2,6,25)
-#ifndef FASTCALL
-#define FASTCALL(x)	x
-#define fastcall
-#endif
-#endif
-
-#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,23)
-
-static unsigned  __attribute__((__used__)) kvm_tsc_khz = 2000000;
-
-#else
-
-#define kvm_tsc_khz tsc_khz
-
-#endif
-
-struct kvm_desc_struct {
-	union {
-		struct { unsigned int a, b; };
-		struct {
-			u16 limit0;
-			u16 base0;
-			unsigned base1: 8, type: 4, s: 1, dpl: 2, p: 1;
-			unsigned limit: 4, avl: 1, l: 1, d: 1, g: 1, base2: 8;
-		};
-
-	};
-} __attribute__((packed));
-
-struct kvm_ldttss_desc64 {
-	u16 limit0;
-	u16 base0;
-	unsigned base1 : 8, type : 5, dpl : 2, p : 1;
-	unsigned limit1 : 4, zero0 : 3, g : 1, base2 : 8;
-	u32 base3;
-	u32 zero1;
-} __attribute__((packed));
-
-struct kvm_desc_ptr {
-	unsigned short size;
-	unsigned long address;
-} __attribute__((packed));
-
-#if LINUX_VERSION_CODE <= KERNEL_VERSION(2,6,21)
-
-#include <linux/ktime.h>
-#include <linux/hrtimer.h>
-
-#define ktime_get kvm_ktime_get
-
-static inline ktime_t ktime_get(void)
-{
-	struct timespec now;
-
-	ktime_get_ts(&now);
-
-	return timespec_to_ktime(now);
-}
-
-#endif
-
-/* __aligned arrived in 2.6.21 */
-#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,21)
-#define __aligned(x) __attribute__((__aligned__(x)))
-#endif
-
-#include <asm/msr.h>
-#ifndef MSR_FS_BASE
-#define MSR_FS_BASE 0xc0000100
-#endif
-#ifndef MSR_GS_BASE
-#define MSR_GS_BASE 0xc0000101
-#endif
-
-#include <linux/mm.h>
-
-/* The shrinker API changed in 2.6.23 */
-#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,23)
-
-struct kvm_shrinker {
-	int (*shrink)(int nr_to_scan, gfp_t gfp_mask);
-	int seeks;
-	struct shrinker *kshrinker;
-};
-
-static inline void register_shrinker(struct kvm_shrinker *shrinker)
-{
-	shrinker->kshrinker = set_shrinker(shrinker->seeks, shrinker->shrink);
-}
-
-static inline void unregister_shrinker(struct kvm_shrinker *shrinker)
-{
-	if (shrinker->kshrinker)
-		remove_shrinker(shrinker->kshrinker);
-}
-
-#define shrinker kvm_shrinker
-
-#endif
-
-/* undefine lapic */
-#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,18)
-
-#undef lapic
-
-#endif
-
-/* clocksource */
-#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,18)
-static inline u32 clocksource_khz2mult(u32 khz, u32 shift_constant)
-{
-	/*  khz = cyc/(Million ns)
-	 *  mult/2^shift  = ns/cyc
-	 *  mult = ns/cyc * 2^shift
-	 *  mult = 1Million/khz * 2^shift
-	 *  mult = 1000000 * 2^shift / khz
-	 *  mult = (1000000<<shift) / khz
-	 */
-	u64 tmp = ((u64)1000000) << shift_constant;
-
-	tmp += khz/2; /* round for do_div */
-	do_div(tmp, khz);
-
-	return (u32)tmp;
-}
-#else
-#include <linux/clocksource.h>
-#endif
-
-/* manually export hrtimer_init/start/cancel */
-#include <linux/kallsyms.h>
-extern void (*hrtimer_init_p)(struct hrtimer *timer, clockid_t which_clock,
-			      enum hrtimer_mode mode);
-extern int (*hrtimer_start_p)(struct hrtimer *timer, ktime_t tim,
-			      const enum hrtimer_mode mode);
-extern int (*hrtimer_cancel_p)(struct hrtimer *timer);
-
-#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,17) && defined(CONFIG_KALLSYMS)
-static inline void hrtimer_kallsyms_resolve(void)
-{
-	hrtimer_init_p = (void *) kallsyms_lookup_name("hrtimer_init");
-	BUG_ON(!hrtimer_init_p);
-	hrtimer_start_p = (void *) kallsyms_lookup_name("hrtimer_start");
-	BUG_ON(!hrtimer_start_p);
-	hrtimer_cancel_p = (void *) kallsyms_lookup_name("hrtimer_cancel");
-	BUG_ON(!hrtimer_cancel_p);
-}
-#else
-static inline void hrtimer_kallsyms_resolve(void)
-{
-	hrtimer_init_p = hrtimer_init;
-	hrtimer_start_p = hrtimer_start;
-	hrtimer_cancel_p = hrtimer_cancel;
-}
-#endif
-
-/* handle old hrtimer API with data pointer */
-#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,17)
-static inline void hrtimer_data_pointer(struct hrtimer *timer)
-{
-	timer->data = (void *)timer;
-}
-#else
-static inline void hrtimer_data_pointer(struct hrtimer *timer) {}
-#endif
-
-#include <asm/hw_irq.h>
-#ifndef NMI_VECTOR
-#define NMI_VECTOR 2
-#endif
-
-#ifndef MSR_MTRRcap
-#define MSR_MTRRcap            0x0fe
-#define MSR_MTRRfix64K_00000   0x250
-#define MSR_MTRRfix16K_80000   0x258
-#define MSR_MTRRfix16K_A0000   0x259
-#define MSR_MTRRfix4K_C0000    0x268
-#define MSR_MTRRfix4K_C8000    0x269
-#define MSR_MTRRfix4K_D0000    0x26a
-#define MSR_MTRRfix4K_D8000    0x26b
-#define MSR_MTRRfix4K_E0000    0x26c
-#define MSR_MTRRfix4K_E8000    0x26d
-#define MSR_MTRRfix4K_F0000    0x26e
-#define MSR_MTRRfix4K_F8000    0x26f
-#define MSR_MTRRdefType        0x2ff
-#endif
-
-#ifndef MSR_IA32_CR_PAT
-#define MSR_IA32_CR_PAT        0x00000277
-#endif
-
-#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,22)
-
-#define ns_to_timespec kvm_ns_to_timespec
-
-struct timespec kvm_ns_to_timespec(const s64 nsec);
-
-#endif
-
-/* Define DEBUGCTLMSR bits */
-#ifndef DEBUGCTLMSR_LBR
-
-#define _DEBUGCTLMSR_LBR	0 /* last branch recording */
-#define _DEBUGCTLMSR_BTF	1 /* single-step on branches */
-
-#define DEBUGCTLMSR_LBR		(1UL << _DEBUGCTLMSR_LBR)
-#define DEBUGCTLMSR_BTF		(1UL << _DEBUGCTLMSR_BTF)
-
-#endif
diff --git a/kernel/debug.h b/kernel/x86/debug.h
similarity index 100%
rename from kernel/debug.h
rename to kernel/x86/debug.h
diff --git a/kernel/external-module-compat.c b/kernel/x86/external-module-compat.c
similarity index 100%
rename from kernel/external-module-compat.c
rename to kernel/x86/external-module-compat.c
diff --git a/kernel/x86/external-module-compat.h b/kernel/x86/external-module-compat.h
new file mode 100644
index 0000000..ce93551
--- /dev/null
+++ b/kernel/x86/external-module-compat.h
@@ -0,0 +1,291 @@
+
+/*
+ * Compatibility header for building as an external module.
+ */
+
+#include <linux/compiler.h>
+#include <linux/version.h>
+
+#include "../external-module-compat-comm.h"
+
+#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,22)
+
+#ifndef _EFER_SCE
+#define _EFER_SCE		0  /* SYSCALL/SYSRET */
+#endif
+
+#ifndef EFER_SCE
+#define EFER_SCE		(1<<_EFER_SCE)
+#endif
+
+#endif
+
+#include <linux/smp.h>
+
+#ifndef X86_CR0_PE
+#define X86_CR0_PE 0x00000001
+#endif
+
+#ifndef X86_CR0_MP
+#define X86_CR0_MP 0x00000002
+#endif
+
+#ifndef X86_CR0_EM
+#define X86_CR0_EM 0x00000004
+#endif
+
+#ifndef X86_CR0_TS
+#define X86_CR0_TS 0x00000008
+#endif
+
+#ifndef X86_CR0_ET
+#define X86_CR0_ET 0x00000010
+#endif
+
+#ifndef X86_CR0_NE
+#define X86_CR0_NE 0x00000020
+#endif
+
+#ifndef X86_CR0_WP
+#define X86_CR0_WP 0x00010000
+#endif
+
+#ifndef X86_CR0_AM
+#define X86_CR0_AM 0x00040000
+#endif
+
+#ifndef X86_CR0_NW
+#define X86_CR0_NW 0x20000000
+#endif
+
+#ifndef X86_CR0_CD
+#define X86_CR0_CD 0x40000000
+#endif
+
+#ifndef X86_CR0_PG
+#define X86_CR0_PG 0x80000000
+#endif
+
+#ifndef X86_CR3_PWT
+#define X86_CR3_PWT 0x00000008
+#endif
+
+#ifndef X86_CR3_PCD
+#define X86_CR3_PCD 0x00000010
+#endif
+
+#ifndef X86_CR4_VMXE
+#define X86_CR4_VMXE 0x00002000
+#endif
+
+#undef X86_CR8_TPR
+#define X86_CR8_TPR 0x0f
+
+/*
+ * 2.6.22 does not define set_64bit() under nonpae
+ */
+#ifdef CONFIG_X86_32
+
+#include <asm/cmpxchg.h>
+
+static inline void __kvm_set_64bit(u64 *ptr, u64 val)
+{
+	unsigned int low = val;
+	unsigned int high = val >> 32;
+
+	__asm__ __volatile__ (
+		"\n1:\t"
+		"movl (%0), %%eax\n\t"
+		"movl 4(%0), %%edx\n\t"
+		"lock cmpxchg8b (%0)\n\t"
+		"jnz 1b"
+		: /* no outputs */
+		:	"D"(ptr),
+			"b"(low),
+			"c"(high)
+		:	"ax","dx","memory");
+}
+
+#undef  set_64bit
+#define set_64bit __kvm_set_64bit
+
+static inline unsigned long long __kvm_cmpxchg64(volatile void *ptr,
+						 unsigned long long old,
+						 unsigned long long new)
+{
+	unsigned long long prev;
+	__asm__ __volatile__("lock cmpxchg8b %3"
+			     : "=A"(prev)
+			     : "b"((unsigned long)new),
+			       "c"((unsigned long)(new >> 32)),
+			       "m"(*__xg(ptr)),
+			       "0"(old)
+			     : "memory");
+	return prev;
+}
+
+#define kvm_cmpxchg64(ptr,o,n)\
+	((__typeof__(*(ptr)))__kvm_cmpxchg64((ptr),(unsigned long long)(o),\
+					(unsigned long long)(n)))
+
+#undef cmpxchg64
+#define cmpxchg64(ptr, o, n) kvm_cmpxchg64(ptr, o, n)
+
+#endif
+
+#ifndef CONFIG_PREEMPT_NOTIFIERS
+/*
+ * Include sched|preempt.h before defining CONFIG_PREEMPT_NOTIFIERS to avoid
+ * a miscompile.
+ */
+#include <linux/sched.h>
+#include <linux/preempt.h>
+#define CONFIG_PREEMPT_NOTIFIERS
+#define CONFIG_PREEMPT_NOTIFIERS_COMPAT
+
+struct preempt_notifier;
+
+struct preempt_ops {
+	void (*sched_in)(struct preempt_notifier *notifier, int cpu);
+	void (*sched_out)(struct preempt_notifier *notifier,
+			  struct task_struct *next);
+};
+
+struct preempt_notifier {
+	struct list_head link;
+	struct task_struct *tsk;
+	struct preempt_ops *ops;
+};
+
+void preempt_notifier_register(struct preempt_notifier *notifier);
+void preempt_notifier_unregister(struct preempt_notifier *notifier);
+
+static inline void preempt_notifier_init(struct preempt_notifier *notifier,
+				     struct preempt_ops *ops)
+{
+	notifier->ops = ops;
+}
+
+void start_special_insn(void);
+void end_special_insn(void);
+void in_special_section(void);
+void special_reload_dr7(void);
+
+void preempt_notifier_sys_init(void);
+void preempt_notifier_sys_exit(void);
+
+#else
+
+static inline void start_special_insn(void) {}
+static inline void end_special_insn(void) {}
+static inline void in_special_section(void) {}
+static inline void special_reload_dr7(void) {}
+
+static inline void preempt_notifier_sys_init(void) {}
+static inline void preempt_notifier_sys_exit(void) {}
+
+#endif
+
+/* CONFIG_HAS_IOMEM is apparently fairly new too (2.6.21 for x86_64). */
+#ifndef CONFIG_HAS_IOMEM
+#define CONFIG_HAS_IOMEM 1
+#endif
+
+/* X86_FEATURE_NX is missing in some x86_64 kernels */
+
+#include <asm/cpufeature.h>
+
+#ifndef X86_FEATURE_NX
+#define X86_FEATURE_NX (1*32+20)
+#endif
+
+#undef true
+#define true 1
+#undef false
+#define false 0
+
+/* EFER_LMA and EFER_LME are missing in pre 2.6.24 i386 kernels */
+#ifndef EFER_LME
+#define _EFER_LME           8  /* Long mode enable */
+#define _EFER_LMA           10 /* Long mode active (read-only) */
+#define EFER_LME            (1<<_EFER_LME)
+#define EFER_LMA            (1<<_EFER_LMA)
+#endif
+
+struct kvm_desc_struct {
+	union {
+		struct { unsigned int a, b; };
+		struct {
+			u16 limit0;
+			u16 base0;
+			unsigned base1: 8, type: 4, s: 1, dpl: 2, p: 1;
+			unsigned limit: 4, avl: 1, l: 1, d: 1, g: 1, base2: 8;
+		};
+
+	};
+} __attribute__((packed));
+
+struct kvm_ldttss_desc64 {
+	u16 limit0;
+	u16 base0;
+	unsigned base1 : 8, type : 5, dpl : 2, p : 1;
+	unsigned limit1 : 4, zero0 : 3, g : 1, base2 : 8;
+	u32 base3;
+	u32 zero1;
+} __attribute__((packed));
+
+struct kvm_desc_ptr {
+	unsigned short size;
+	unsigned long address;
+} __attribute__((packed));
+
+#include <asm/msr.h>
+#ifndef MSR_FS_BASE
+#define MSR_FS_BASE 0xc0000100
+#endif
+#ifndef MSR_GS_BASE
+#define MSR_GS_BASE 0xc0000101
+#endif
+
+/* undefine lapic */
+#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,18)
+
+#undef lapic
+
+#endif
+
+#include <asm/hw_irq.h>
+#ifndef NMI_VECTOR
+#define NMI_VECTOR 2
+#endif
+
+#ifndef MSR_MTRRcap
+#define MSR_MTRRcap            0x0fe
+#define MSR_MTRRfix64K_00000   0x250
+#define MSR_MTRRfix16K_80000   0x258
+#define MSR_MTRRfix16K_A0000   0x259
+#define MSR_MTRRfix4K_C0000    0x268
+#define MSR_MTRRfix4K_C8000    0x269
+#define MSR_MTRRfix4K_D0000    0x26a
+#define MSR_MTRRfix4K_D8000    0x26b
+#define MSR_MTRRfix4K_E0000    0x26c
+#define MSR_MTRRfix4K_E8000    0x26d
+#define MSR_MTRRfix4K_F0000    0x26e
+#define MSR_MTRRfix4K_F8000    0x26f
+#define MSR_MTRRdefType        0x2ff
+#endif
+
+#ifndef MSR_IA32_CR_PAT
+#define MSR_IA32_CR_PAT        0x00000277
+#endif
+
+/* Define DEBUGCTLMSR bits */
+#ifndef DEBUGCTLMSR_LBR
+
+#define _DEBUGCTLMSR_LBR	0 /* last branch recording */
+#define _DEBUGCTLMSR_BTF	1 /* single-step on branches */
+
+#define DEBUGCTLMSR_LBR		(1UL << _DEBUGCTLMSR_LBR)
+#define DEBUGCTLMSR_BTF		(1UL << _DEBUGCTLMSR_BTF)
+
+#endif
diff --git a/kernel/preempt.c b/kernel/x86/preempt.c
similarity index 100%
rename from kernel/preempt.c
rename to kernel/x86/preempt.c
diff --git a/kernel/vmx-debug.c b/kernel/x86/vmx-debug.c
similarity index 100%
rename from kernel/vmx-debug.c
rename to kernel/x86/vmx-debug.c
-- 
1.5.6.2


^ permalink raw reply related	[flat|nested] 24+ messages in thread

* Re: [PATCH 01/04]Create x86 directory to hold x86-specific files.
  2008-07-27  9:46   ` Zhang, Xiantao
@ 2008-07-29 13:14   ` Avi Kivity
  -1 siblings, 0 replies; 24+ messages in thread
From: Avi Kivity @ 2008-07-29 13:14 UTC (permalink / raw)
  To: kvm-ia64

Zhang, Xiantao wrote:
> Avi Kivity wrote:
>   
>> Zhang, Xiantao wrote:
>>     
>>> From c8c8c5a6b3ff943cb645f9ac4ef169490f14cd08 Mon Sep 17 00:00:00
>>> 2001 From: Zhang Xiantao <xiantao.zhang@intel.com>
>>> Date: Fri, 25 Jul 2008 16:50:30 +0800
>>> Subject: [PATCH] KVM: external module: Moving x86-speicif files to
>>> x86 directory. 
>>>
>>> Create x86 directory to hold x86-specific files.
>>> Signed-off-by: Zhang Xiantao <xiantao.zhang@intel.com>
>>>
>>>       
>> You forgot to attach a patch; and your mailer corrupts patches, so I
>> can't apply this.
>>     
>
> Sorry. Attached! :)
>   

Sorry 2, external-module-compat.h has changed!  Please update and I'll 
apply it quicker this time.

-- 
error compiling committee.c: too many arguments to function


^ permalink raw reply	[flat|nested] 24+ messages in thread

* Re: [PATCH 01/04]Create x86 directory to hold x86-specific files.
@ 2008-07-29 13:14   ` Avi Kivity
  0 siblings, 0 replies; 24+ messages in thread
From: Avi Kivity @ 2008-07-29 13:14 UTC (permalink / raw)
  To: Zhang, Xiantao; +Cc: kvm-ia64, kvm

Zhang, Xiantao wrote:
> Avi Kivity wrote:
>   
>> Zhang, Xiantao wrote:
>>     
>>> From c8c8c5a6b3ff943cb645f9ac4ef169490f14cd08 Mon Sep 17 00:00:00
>>> 2001 From: Zhang Xiantao <xiantao.zhang@intel.com>
>>> Date: Fri, 25 Jul 2008 16:50:30 +0800
>>> Subject: [PATCH] KVM: external module: Moving x86-speicif files to
>>> x86 directory. 
>>>
>>> Create x86 directory to hold x86-specific files.
>>> Signed-off-by: Zhang Xiantao <xiantao.zhang@intel.com>
>>>
>>>       
>> You forgot to attach a patch; and your mailer corrupts patches, so I
>> can't apply this.
>>     
>
> Sorry. Attached! :)
>   

Sorry 2, external-module-compat.h has changed!  Please update and I'll 
apply it quicker this time.

-- 
error compiling committee.c: too many arguments to function


^ permalink raw reply	[flat|nested] 24+ messages in thread

* RE: [PATCH 01/04]Create x86 directory to hold x86-specific files.
  2008-07-29 13:14   ` Avi Kivity
@ 2008-07-30  1:42   ` Zhang, Xiantao
  -1 siblings, 0 replies; 24+ messages in thread
From: Zhang, Xiantao @ 2008-07-30  1:42 UTC (permalink / raw)
  To: kvm-ia64

[-- Attachment #1: Type: text/plain, Size: 793 bytes --]

Avi Kivity wrote:
> Zhang, Xiantao wrote:
>> Avi Kivity wrote:
>> 
>>> Zhang, Xiantao wrote:
>>> 
>>>> From c8c8c5a6b3ff943cb645f9ac4ef169490f14cd08 Mon Sep 17 00:00:00
>>>> 2001 From: Zhang Xiantao <xiantao.zhang@intel.com>
>>>> Date: Fri, 25 Jul 2008 16:50:30 +0800
>>>> Subject: [PATCH] KVM: external module: Moving x86-speicif files to
>>>> x86 directory. 
>>>> 
>>>> Create x86 directory to hold x86-specific files.
>>>> Signed-off-by: Zhang Xiantao <xiantao.zhang@intel.com>
>>>> 
>>>> 
>>> You forgot to attach a patch; and your mailer corrupts patches, so
>>> I can't apply this. 
>>> 
>> 
>> Sorry. Attached! :)
>> 
> 
> Sorry 2, external-module-compat.h has changed!  Please update and I'll
> apply it quicker this time.

Updated and attached it again :-)

[-- Attachment #2: 0001-KVM-external-module-Moving-x86-speicif-files-to-x8.patch --]
[-- Type: application/octet-stream, Size: 38940 bytes --]

From dae628bf166e6b2c2a8c86340c0f50dbc1746af5 Mon Sep 17 00:00:00 2001
From: Zhang Xiantao <xiantao.zhang@intel.com>
Date: Fri, 25 Jul 2008 16:50:30 +0800
Subject: [PATCH] KVM: external module: Moving x86-speicif files to x86 directory.

Create x86 directory to hold x86-specific files.
Signed-off-by: Zhang Xiantao <xiantao.zhang@intel.com>
---
 kernel/Kbuild                             |    8 +-
 kernel/Makefile                           |    2 +-
 kernel/external-module-compat-comm.h      |  497 ++++++++++++++++++
 kernel/external-module-compat.h           |  781 -----------------------------
 kernel/{ => x86}/debug.h                  |    0 
 kernel/{ => x86}/external-module-compat.c |    0 
 kernel/x86/external-module-compat.h       |  291 +++++++++++
 kernel/{ => x86}/preempt.c                |    0 
 kernel/{ => x86}/vmx-debug.c              |    0 
 9 files changed, 793 insertions(+), 786 deletions(-)
 create mode 100644 kernel/external-module-compat-comm.h
 delete mode 100644 kernel/external-module-compat.h
 rename kernel/{ => x86}/debug.h (100%)
 rename kernel/{ => x86}/external-module-compat.c (100%)
 create mode 100644 kernel/x86/external-module-compat.h
 rename kernel/{ => x86}/preempt.c (100%)
 rename kernel/{ => x86}/vmx-debug.c (100%)

diff --git a/kernel/Kbuild b/kernel/Kbuild
index 676f28d..cb63582 100644
--- a/kernel/Kbuild
+++ b/kernel/Kbuild
@@ -1,11 +1,11 @@
 obj-m := kvm.o kvm-intel.o kvm-amd.o
 kvm-objs := kvm_main.o x86.o mmu.o x86_emulate.o anon_inodes.o irq.o i8259.o \
-	 lapic.o ioapic.o preempt.o i8254.o coalesced_mmio.o \
-	 external-module-compat.o
+	 lapic.o ioapic.o x86/preempt.o i8254.o coalesced_mmio.o \
+	 x86/external-module-compat.o
 ifeq ($(CONFIG_KVM_TRACE),y)
 kvm-objs += kvm_trace.o
 endif
-kvm-intel-objs := vmx.o vmx-debug.o external-module-compat.o
-kvm-amd-objs := svm.o external-module-compat.o
+kvm-intel-objs := vmx.o x86/vmx-debug.o x86/external-module-compat.o
+kvm-amd-objs := svm.o x86/external-module-compat.o
 
 CFLAGS_kvm_main.o = -DKVM_MAIN
diff --git a/kernel/Makefile b/kernel/Makefile
index 69b4981..c2c666e 100644
--- a/kernel/Makefile
+++ b/kernel/Makefile
@@ -35,7 +35,7 @@ all::
 	$(MAKE) -C $(KERNELDIR) M=`pwd` \
 		LINUXINCLUDE="-I`pwd`/include -Iinclude -I`pwd`/include-compat \
 		-include include/linux/autoconf.h \
-		-include `pwd`/external-module-compat.h"
+		-include `pwd`/$(ARCH_DIR)/external-module-compat.h"
 		"$$@"
 
 sync: header-sync source-sync
diff --git a/kernel/external-module-compat-comm.h b/kernel/external-module-compat-comm.h
new file mode 100644
index 0000000..f5169b8
--- /dev/null
+++ b/kernel/external-module-compat-comm.h
@@ -0,0 +1,497 @@
+
+/*
+ * Compatibility header for building as an external module.
+ */
+
+/*
+ * Avoid picking up the kernel's kvm.h in case we have a newer one.
+ */
+
+#include <linux/compiler.h>
+#include <linux/version.h>
+#include <linux/string.h>
+#include <linux/kvm.h>
+#include <linux/kvm_para.h>
+#include <linux/cpu.h>
+#include <linux/time.h>
+#include <asm/processor.h>
+#include <linux/hrtimer.h>
+#include <asm/bitops.h>
+#include <asm/msr.h>
+
+/*
+ * 2.6.16 does not have GFP_NOWAIT
+ */
+
+#include <linux/gfp.h>
+
+#ifndef GFP_NOWAIT
+#define GFP_NOWAIT (GFP_ATOMIC & ~__GFP_HIGH)
+#endif
+
+
+/*
+ * kvm profiling support needs 2.6.20
+ */
+#include <linux/profile.h>
+
+#ifndef KVM_PROFILING
+#define KVM_PROFILING 1234
+#define prof_on       4321
+#endif
+
+/*
+ * smp_call_function_single() is not exported below 2.6.20, and has different
+ * semantics below 2.6.23.  The 'nonatomic' argument was removed in 2.6.27.
+ */
+#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,27)
+
+int kvm_smp_call_function_single(int cpu, void (*func)(void *info),
+				 void *info, int wait);
+
+#define smp_call_function_single kvm_smp_call_function_single
+
+#endif
+
+/* on_each_cpu() lost an argument in 2.6.27. */
+#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,27)
+
+#define kvm_on_each_cpu(func, info, wait) on_each_cpu(func, info, 0, wait)
+
+#else
+
+#define kvm_on_each_cpu(func, info, wait) on_each_cpu(func, info, wait)
+
+#endif
+
+/*
+ * The cpu hotplug stubs are broken if !CONFIG_CPU_HOTPLUG
+ */
+
+#if LINUX_VERSION_CODE <= KERNEL_VERSION(2,6,15)
+#define DEFINE_MUTEX(a) DECLARE_MUTEX(a)
+#define mutex_lock_interruptible(a) down_interruptible(a)
+#define mutex_unlock(a) up(a)
+#define mutex_lock(a) down(a)
+#define mutex_init(a) init_MUTEX(a)
+#define mutex_trylock(a) down_trylock(a)
+#define mutex semaphore
+#endif
+
+#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,14)
+#ifndef kzalloc
+#define kzalloc(size,flags)			\
+({						\
+	void *__ret = kmalloc(size, flags);	\
+	if (__ret)				\
+		memset(__ret, 0, size);		\
+	__ret;					\
+})
+#endif
+#endif
+
+#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,17)
+#ifndef kmem_cache_zalloc
+#define kmem_cache_zalloc(cache,flags)			  \
+({							  \
+	void *__ret = kmem_cache_alloc(cache, flags);	  \
+	if (__ret)                                        \
+		memset(__ret, 0, kmem_cache_size(cache)); \
+	__ret;                                            \
+})
+#endif
+#endif
+
+#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,21)
+
+#ifndef CONFIG_HOTPLUG_CPU
+#define register_cpu_notifier(nb) (0)
+#endif
+
+#endif
+
+#include <linux/miscdevice.h>
+#ifndef KVM_MINOR
+#define KVM_MINOR 232
+#endif
+
+#include <linux/notifier.h>
+#ifndef CPU_TASKS_FROZEN
+
+#define CPU_TASKS_FROZEN       0x0010
+#define CPU_ONLINE_FROZEN      (CPU_ONLINE | CPU_TASKS_FROZEN)
+#define CPU_UP_PREPARE_FROZEN  (CPU_UP_PREPARE | CPU_TASKS_FROZEN)
+#define CPU_UP_CANCELED_FROZEN (CPU_UP_CANCELED | CPU_TASKS_FROZEN)
+#define CPU_DOWN_PREPARE_FROZEN        (CPU_DOWN_PREPARE | CPU_TASKS_FROZEN)
+#define CPU_DOWN_FAILED_FROZEN (CPU_DOWN_FAILED | CPU_TASKS_FROZEN)
+#define CPU_DEAD_FROZEN                (CPU_DEAD | CPU_TASKS_FROZEN)
+
+#endif
+
+#ifndef CPU_DYING
+#define CPU_DYING 0x000A
+#define CPU_DYING_FROZEN (CPU_DYING | CPU_TASKS_FROZEN)
+#endif
+
+#include <asm/system.h>
+
+struct inode;
+#include <linux/anon_inodes.h>
+#define anon_inode_getfd kvm_anon_inode_getfd
+int kvm_init_anon_inodes(void);
+void kvm_exit_anon_inodes(void);
+int anon_inode_getfd(const char *name,
+		     const struct file_operations *fops,
+		     void *priv , int flags);
+
+/*
+ * 2.6.23 removed the cache destructor
+ */
+#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,23)
+#  define kmem_cache_create(name, size, align, flags, ctor) \
+	kmem_cache_create(name, size, align, flags, ctor, NULL)
+#endif
+
+/* HRTIMER_MODE_ABS started life with a different name */
+#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,21)
+#define HRTIMER_MODE_ABS HRTIMER_ABS
+#endif
+
+/* div64_u64 is fairly new */
+#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,26)
+
+#define div64_u64 kvm_div64_u64
+
+#ifdef CONFIG_64BIT
+
+static inline uint64_t div64_u64(uint64_t dividend, uint64_t divisor)
+{
+	return dividend / divisor;
+}
+
+#else
+
+uint64_t div64_u64(uint64_t dividend, uint64_t divisor);
+
+#endif
+
+#endif
+
+#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,19)
+
+#ifdef RHEL_RELEASE_CODE
+#if RHEL_RELEASE_CODE >= RHEL_RELEASE_VERSION(5,2)
+#define RHEL_BOOL 1
+#endif
+#endif
+
+#ifndef RHEL_BOOL
+
+typedef _Bool bool;
+
+#endif
+
+#endif
+
+/*
+ * PF_VCPU is a Linux 2.6.24 addition
+ */
+
+#include <linux/sched.h>
+
+#ifndef PF_VCPU
+#define PF_VCPU 0
+#endif
+
+/*
+ * smp_call_function_mask() is not defined/exported below 2.6.24
+ */
+
+#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,24)
+
+int kvm_smp_call_function_mask(cpumask_t mask, void (*func) (void *info),
+			       void *info, int wait);
+
+#define smp_call_function_mask kvm_smp_call_function_mask
+
+#endif
+
+/* empty_zero_page isn't exported in all kernels */
+#include <asm/pgtable.h>
+
+#define empty_zero_page kvm_empty_zero_page
+
+static char empty_zero_page[PAGE_SIZE];
+
+static inline void blahblah(void)
+{
+	(void)empty_zero_page[0];
+}
+
+/* __mmdrop() is not exported before 2.6.25 */
+#include <linux/sched.h>
+
+#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,25)
+
+#define mmdrop(x) do { (void)(x); } while (0)
+#define mmget(x) do { (void)(x); } while (0)
+
+#else
+
+#define mmget(x) do { atomic_inc(x); } while (0)
+
+#endif
+
+/* pagefault_enable(), page_fault_disable() - 2.6.20 */
+#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,20)
+
+static inline void pagefault_disable(void)
+{
+	inc_preempt_count();
+	/*
+	 * make sure to have issued the store before a pagefault
+	 * can hit.
+	 */
+	barrier();
+}
+
+static inline void pagefault_enable(void)
+{
+	/*
+	 * make sure to issue those last loads/stores before enabling
+	 * the pagefault handler again.
+	 */
+	barrier();
+	dec_preempt_count();
+	/*
+	 * make sure we do..
+	 */
+	barrier();
+	preempt_check_resched();
+}
+
+#endif
+
+/* vm ops ->fault() was introduced in 2.6.23. */
+#include <linux/mm.h>
+
+#ifdef KVM_MAIN
+#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,23)
+
+struct vm_fault {
+	unsigned int flags;
+	pgoff_t pgoff;
+	void __user *virtual_address;
+	struct page *page;
+};
+
+static int kvm_vcpu_fault(struct vm_area_struct *vma, struct vm_fault *vmf);
+static int kvm_vm_fault(struct vm_area_struct *vma, struct vm_fault *vmf);
+
+static inline struct page *kvm_nopage_to_fault(
+	int (*fault)(struct vm_area_struct *vma, struct vm_fault *vmf),
+	struct vm_area_struct *vma,
+	unsigned long address,
+	int *type)
+{
+	struct vm_fault vmf;
+	int ret;
+
+	vmf.pgoff = ((address - vma->vm_start) >> PAGE_SHIFT) + vma->vm_pgoff;
+	vmf.virtual_address = (void __user *)address;
+	ret = fault(vma, &vmf);
+	if (ret)
+		return NOPAGE_SIGBUS;
+	*type = VM_FAULT_MINOR;
+	return vmf.page;
+}
+
+static inline struct page *__kvm_vcpu_fault(struct vm_area_struct *vma,
+					    unsigned long address,
+					    int *type)
+{
+	return kvm_nopage_to_fault(kvm_vcpu_fault, vma, address, type);
+}
+
+static inline struct page *__kvm_vm_fault(struct vm_area_struct *vma,
+					  unsigned long address,
+					  int *type)
+{
+	return kvm_nopage_to_fault(kvm_vm_fault, vma, address, type);
+}
+
+#define VMA_OPS_FAULT(x) nopage
+#define VMA_OPS_FAULT_FUNC(x) __##x
+
+#else
+
+#define VMA_OPS_FAULT(x) x
+#define VMA_OPS_FAULT_FUNC(x) x
+
+#endif
+#endif
+
+/* simple vfs attribute getter signature has changed to add a return code */
+
+#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,25)
+
+#define MAKE_SIMPLE_ATTRIBUTE_GETTER(x)       \
+	static u64 x(void *v)                 \
+	{				      \
+		u64 ret = 0;		      \
+					      \
+		__##x(v, &ret);		      \
+		return ret;		      \
+	}
+
+#else
+
+#define MAKE_SIMPLE_ATTRIBUTE_GETTER(x)       \
+	static int x(void *v, u64 *val)	      \
+	{				      \
+		return __##x(v, val);	      \
+	}
+
+#endif
+
+/* set_kset_name() is gone in 2.6.25 */
+
+#if LINUX_VERSION_CODE >= KERNEL_VERSION(2,6,25)
+
+#define set_kset_name(x) .name = x
+
+#endif
+
+#if LINUX_VERSION_CODE >= KERNEL_VERSION(2,6,25)
+#ifndef FASTCALL
+#define FASTCALL(x)	x
+#define fastcall
+#endif
+#endif
+
+#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,23)
+
+static unsigned  __attribute__((__used__)) kvm_tsc_khz = 2000000;
+
+#else
+
+#define kvm_tsc_khz tsc_khz
+
+#endif
+
+#if LINUX_VERSION_CODE <= KERNEL_VERSION(2,6,21)
+
+#include <linux/ktime.h>
+#include <linux/hrtimer.h>
+
+#define ktime_get kvm_ktime_get
+
+static inline ktime_t ktime_get(void)
+{
+	struct timespec now;
+
+	ktime_get_ts(&now);
+
+	return timespec_to_ktime(now);
+}
+
+#endif
+
+/* __aligned arrived in 2.6.21 */
+#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,21)
+#define __aligned(x) __attribute__((__aligned__(x)))
+#endif
+
+#include <linux/mm.h>
+
+/* The shrinker API changed in 2.6.23 */
+#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,23)
+
+struct kvm_shrinker {
+	int (*shrink)(int nr_to_scan, gfp_t gfp_mask);
+	int seeks;
+	struct shrinker *kshrinker;
+};
+
+static inline void register_shrinker(struct kvm_shrinker *shrinker)
+{
+	shrinker->kshrinker = set_shrinker(shrinker->seeks, shrinker->shrink);
+}
+
+static inline void unregister_shrinker(struct kvm_shrinker *shrinker)
+{
+	if (shrinker->kshrinker)
+		remove_shrinker(shrinker->kshrinker);
+}
+
+#define shrinker kvm_shrinker
+
+#endif
+
+/* clocksource */
+#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,18)
+static inline u32 clocksource_khz2mult(u32 khz, u32 shift_constant)
+{
+	/*  khz = cyc/(Million ns)
+	 *  mult/2^shift  = ns/cyc
+	 *  mult = ns/cyc * 2^shift
+	 *  mult = 1Million/khz * 2^shift
+	 *  mult = 1000000 * 2^shift / khz
+	 *  mult = (1000000<<shift) / khz
+	 */
+	u64 tmp = ((u64)1000000) << shift_constant;
+
+	tmp += khz/2; /* round for do_div */
+	do_div(tmp, khz);
+
+	return (u32)tmp;
+}
+#else
+#include <linux/clocksource.h>
+#endif
+
+/* manually export hrtimer_init/start/cancel */
+#include <linux/kallsyms.h>
+extern void (*hrtimer_init_p)(struct hrtimer *timer, clockid_t which_clock,
+			      enum hrtimer_mode mode);
+extern int (*hrtimer_start_p)(struct hrtimer *timer, ktime_t tim,
+			      const enum hrtimer_mode mode);
+extern int (*hrtimer_cancel_p)(struct hrtimer *timer);
+
+#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,17) && defined(CONFIG_KALLSYMS)
+static inline void hrtimer_kallsyms_resolve(void)
+{
+	hrtimer_init_p = (void *) kallsyms_lookup_name("hrtimer_init");
+	BUG_ON(!hrtimer_init_p);
+	hrtimer_start_p = (void *) kallsyms_lookup_name("hrtimer_start");
+	BUG_ON(!hrtimer_start_p);
+	hrtimer_cancel_p = (void *) kallsyms_lookup_name("hrtimer_cancel");
+	BUG_ON(!hrtimer_cancel_p);
+}
+#else
+static inline void hrtimer_kallsyms_resolve(void)
+{
+	hrtimer_init_p = hrtimer_init;
+	hrtimer_start_p = hrtimer_start;
+	hrtimer_cancel_p = hrtimer_cancel;
+}
+#endif
+
+/* handle old hrtimer API with data pointer */
+#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,17)
+static inline void hrtimer_data_pointer(struct hrtimer *timer)
+{
+	timer->data = (void *)timer;
+}
+#else
+static inline void hrtimer_data_pointer(struct hrtimer *timer) {}
+#endif
+
+#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,22)
+
+#define ns_to_timespec kvm_ns_to_timespec
+
+struct timespec kvm_ns_to_timespec(const s64 nsec);
+
+#endif
+
diff --git a/kernel/external-module-compat.h b/kernel/external-module-compat.h
deleted file mode 100644
index a27103c..0000000
--- a/kernel/external-module-compat.h
+++ /dev/null
@@ -1,781 +0,0 @@
-
-/*
- * Compatibility header for building as an external module.
- */
-
-/*
- * Avoid picking up the kernel's kvm.h in case we have a newer one.
- */
-
-#include <linux/compiler.h>
-#include <linux/version.h>
-#include <linux/string.h>
-#include <linux/kvm.h>
-#include <linux/kvm_para.h>
-#include <linux/cpu.h>
-#include <linux/time.h>
-#include <asm/processor.h>
-#include <linux/hrtimer.h>
-#include <asm/bitops.h>
-#include <asm/msr.h>
-
-/*
- * 2.6.16 does not have GFP_NOWAIT
- */
-
-#include <linux/gfp.h>
-
-#ifndef GFP_NOWAIT
-#define GFP_NOWAIT (GFP_ATOMIC & ~__GFP_HIGH)
-#endif
-
-
-/*
- * kvm profiling support needs 2.6.20
- */
-#include <linux/profile.h>
-
-#ifndef KVM_PROFILING
-#define KVM_PROFILING 1234
-#define prof_on       4321
-#endif
-
-/*
- * smp_call_function_single() is not exported below 2.6.20, and has different
- * semantics below 2.6.23.  The 'nonatomic' argument was removed in 2.6.27.
- */
-#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,27)
-
-int kvm_smp_call_function_single(int cpu, void (*func)(void *info),
-				 void *info, int wait);
-
-#define smp_call_function_single kvm_smp_call_function_single
-
-#endif
-
-/* on_each_cpu() lost an argument in 2.6.27. */
-#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,27)
-
-#define kvm_on_each_cpu(func, info, wait) on_each_cpu(func, info, 0, wait)
-
-#else
-
-#define kvm_on_each_cpu(func, info, wait) on_each_cpu(func, info, wait)
-
-#endif
-
-/*
- * The cpu hotplug stubs are broken if !CONFIG_CPU_HOTPLUG
- */
-
-#if LINUX_VERSION_CODE <= KERNEL_VERSION(2,6,15)
-#define DEFINE_MUTEX(a) DECLARE_MUTEX(a)
-#define mutex_lock_interruptible(a) down_interruptible(a)
-#define mutex_unlock(a) up(a)
-#define mutex_lock(a) down(a)
-#define mutex_init(a) init_MUTEX(a)
-#define mutex_trylock(a) down_trylock(a)
-#define mutex semaphore
-#endif
-
-#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,14)
-#ifndef kzalloc
-#define kzalloc(size,flags)			\
-({						\
-	void *__ret = kmalloc(size, flags);	\
-	if (__ret)				\
-		memset(__ret, 0, size);		\
-	__ret;					\
-})
-#endif
-#endif
-
-#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,17)
-#ifndef kmem_cache_zalloc
-#define kmem_cache_zalloc(cache,flags)			  \
-({							  \
-	void *__ret = kmem_cache_alloc(cache, flags);	  \
-	if (__ret)                                        \
-		memset(__ret, 0, kmem_cache_size(cache)); \
-	__ret;                                            \
-})
-#endif
-#endif
-
-#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,21)
-
-#ifndef CONFIG_HOTPLUG_CPU
-#define register_cpu_notifier(nb) (0)
-#endif
-
-#endif
-
-#include <linux/miscdevice.h>
-#ifndef KVM_MINOR
-#define KVM_MINOR 232
-#endif
-
-#include <linux/notifier.h>
-#ifndef CPU_TASKS_FROZEN
-
-#define CPU_TASKS_FROZEN       0x0010
-#define CPU_ONLINE_FROZEN      (CPU_ONLINE | CPU_TASKS_FROZEN)
-#define CPU_UP_PREPARE_FROZEN  (CPU_UP_PREPARE | CPU_TASKS_FROZEN)
-#define CPU_UP_CANCELED_FROZEN (CPU_UP_CANCELED | CPU_TASKS_FROZEN)
-#define CPU_DOWN_PREPARE_FROZEN        (CPU_DOWN_PREPARE | CPU_TASKS_FROZEN)
-#define CPU_DOWN_FAILED_FROZEN (CPU_DOWN_FAILED | CPU_TASKS_FROZEN)
-#define CPU_DEAD_FROZEN                (CPU_DEAD | CPU_TASKS_FROZEN)
-
-#endif
-
-#ifndef CPU_DYING
-#define CPU_DYING 0x000A
-#define CPU_DYING_FROZEN (CPU_DYING | CPU_TASKS_FROZEN)
-#endif
-
-#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,22)
-
-#ifndef _EFER_SCE
-#define _EFER_SCE		0  /* SYSCALL/SYSRET */
-#endif
-
-#ifndef EFER_SCE
-#define EFER_SCE		(1<<_EFER_SCE)
-#endif
-
-#endif
-
-/*
- * For set_64bit(), which is in a new file asm/cmpxchg.h in newer kernels.
- */
-#include <asm/system.h>
-
-struct inode;
-#include <linux/anon_inodes.h>
-#define anon_inode_getfd kvm_anon_inode_getfd
-int kvm_init_anon_inodes(void);
-void kvm_exit_anon_inodes(void);
-int anon_inode_getfd(const char *name,
-		     const struct file_operations *fops,
-		     void *priv, int flags);
-
-#include <linux/smp.h>
-
-#ifndef X86_CR0_PE
-#define X86_CR0_PE 0x00000001
-#endif
-
-#ifndef X86_CR0_MP
-#define X86_CR0_MP 0x00000002
-#endif
-
-#ifndef X86_CR0_EM
-#define X86_CR0_EM 0x00000004
-#endif
-
-#ifndef X86_CR0_TS
-#define X86_CR0_TS 0x00000008
-#endif
-
-#ifndef X86_CR0_ET
-#define X86_CR0_ET 0x00000010
-#endif
-
-#ifndef X86_CR0_NE
-#define X86_CR0_NE 0x00000020
-#endif
-
-#ifndef X86_CR0_WP
-#define X86_CR0_WP 0x00010000
-#endif
-
-#ifndef X86_CR0_AM
-#define X86_CR0_AM 0x00040000
-#endif
-
-#ifndef X86_CR0_NW
-#define X86_CR0_NW 0x20000000
-#endif
-
-#ifndef X86_CR0_CD
-#define X86_CR0_CD 0x40000000
-#endif
-
-#ifndef X86_CR0_PG
-#define X86_CR0_PG 0x80000000
-#endif
-
-#ifndef X86_CR3_PWT
-#define X86_CR3_PWT 0x00000008
-#endif
-
-#ifndef X86_CR3_PCD
-#define X86_CR3_PCD 0x00000010
-#endif
-
-#ifndef X86_CR4_VMXE
-#define X86_CR4_VMXE 0x00002000
-#endif
-
-#undef X86_CR8_TPR
-#define X86_CR8_TPR 0x0f
-
-/*
- * 2.6.23 removed the cache destructor
- */
-#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,23)
-#  define kmem_cache_create(name, size, align, flags, ctor) \
-	kmem_cache_create(name, size, align, flags, ctor, NULL)
-#endif
-
-/*
- * 2.6.22 does not define set_64bit() under nonpae
- */
-#ifdef CONFIG_X86_32
-
-#include <asm/cmpxchg.h>
-
-static inline void __kvm_set_64bit(u64 *ptr, u64 val)
-{
-	unsigned int low = val;
-	unsigned int high = val >> 32;
-
-	__asm__ __volatile__ (
-		"\n1:\t"
-		"movl (%0), %%eax\n\t"
-		"movl 4(%0), %%edx\n\t"
-		"lock cmpxchg8b (%0)\n\t"
-		"jnz 1b"
-		: /* no outputs */
-		:	"D"(ptr),
-			"b"(low),
-			"c"(high)
-		:	"ax","dx","memory");
-}
-
-#undef  set_64bit
-#define set_64bit __kvm_set_64bit
-
-static inline unsigned long long __kvm_cmpxchg64(volatile void *ptr,
-						 unsigned long long old,
-						 unsigned long long new)
-{
-	unsigned long long prev;
-	__asm__ __volatile__("lock cmpxchg8b %3"
-			     : "=A"(prev)
-			     : "b"((unsigned long)new),
-			       "c"((unsigned long)(new >> 32)),
-			       "m"(*__xg(ptr)),
-			       "0"(old)
-			     : "memory");
-	return prev;
-}
-
-#define kvm_cmpxchg64(ptr,o,n)\
-	((__typeof__(*(ptr)))__kvm_cmpxchg64((ptr),(unsigned long long)(o),\
-					(unsigned long long)(n)))
-
-#undef cmpxchg64
-#define cmpxchg64(ptr, o, n) kvm_cmpxchg64(ptr, o, n)
-
-#endif
-
-#ifndef CONFIG_PREEMPT_NOTIFIERS
-/*
- * Include sched|preempt.h before defining CONFIG_PREEMPT_NOTIFIERS to avoid
- * a miscompile.
- */
-#include <linux/sched.h>
-#include <linux/preempt.h>
-#define CONFIG_PREEMPT_NOTIFIERS
-#define CONFIG_PREEMPT_NOTIFIERS_COMPAT
-
-struct preempt_notifier;
-
-struct preempt_ops {
-	void (*sched_in)(struct preempt_notifier *notifier, int cpu);
-	void (*sched_out)(struct preempt_notifier *notifier,
-			  struct task_struct *next);
-};
-
-struct preempt_notifier {
-	struct list_head link;
-	struct task_struct *tsk;
-	struct preempt_ops *ops;
-};
-
-void preempt_notifier_register(struct preempt_notifier *notifier);
-void preempt_notifier_unregister(struct preempt_notifier *notifier);
-
-static inline void preempt_notifier_init(struct preempt_notifier *notifier,
-				     struct preempt_ops *ops)
-{
-	notifier->ops = ops;
-}
-
-void start_special_insn(void);
-void end_special_insn(void);
-void in_special_section(void);
-void special_reload_dr7(void);
-
-void preempt_notifier_sys_init(void);
-void preempt_notifier_sys_exit(void);
-
-#else
-
-static inline void start_special_insn(void) {}
-static inline void end_special_insn(void) {}
-static inline void in_special_section(void) {}
-static inline void special_reload_dr7(void) {}
-
-static inline void preempt_notifier_sys_init(void) {}
-static inline void preempt_notifier_sys_exit(void) {}
-
-#endif
-
-/* HRTIMER_MODE_ABS started life with a different name */
-#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,21)
-#define HRTIMER_MODE_ABS HRTIMER_ABS
-#endif
-
-/* div64_u64 is fairly new */
-#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,26)
-
-#define div64_u64 kvm_div64_u64
-
-#ifdef CONFIG_64BIT
-
-static inline uint64_t div64_u64(uint64_t dividend, uint64_t divisor)
-{
-	return dividend / divisor;
-}
-
-#else
-
-uint64_t div64_u64(uint64_t dividend, uint64_t divisor);
-
-#endif
-
-#endif
-
-#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,19)
-
-#ifdef RHEL_RELEASE_CODE
-#if RHEL_RELEASE_CODE >= RHEL_RELEASE_VERSION(5,2)
-#define RHEL_BOOL 1
-#endif
-#endif
-
-#ifndef RHEL_BOOL
-
-typedef _Bool bool;
-
-#endif
-
-#endif
-
-/*
- * PF_VCPU is a Linux 2.6.24 addition
- */
-
-#include <linux/sched.h>
-
-#ifndef PF_VCPU
-#define PF_VCPU 0
-#endif
-
-/*
- * smp_call_function_mask() is not defined/exported below 2.6.24
- */
-
-#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,24)
-
-int kvm_smp_call_function_mask(cpumask_t mask, void (*func) (void *info),
-			       void *info, int wait);
-
-#define smp_call_function_mask kvm_smp_call_function_mask
-
-#endif
-
-/* CONFIG_HAS_IOMEM is apparently fairly new too (2.6.21 for x86_64). */
-#ifndef CONFIG_HAS_IOMEM
-#define CONFIG_HAS_IOMEM 1
-#endif
-
-/* empty_zero_page isn't exported in all kernels */
-#include <asm/pgtable.h>
-
-#define empty_zero_page kvm_empty_zero_page
-
-static char empty_zero_page[PAGE_SIZE];
-
-static inline void blahblah(void)
-{
-	(void)empty_zero_page[0];
-}
-
-/* __mmdrop() is not exported before 2.6.25 */
-#include <linux/sched.h>
-
-#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,25)
-
-#define mmdrop(x) do { (void)(x); } while (0)
-#define mmget(x) do { (void)(x); } while (0)
-
-#else
-
-#define mmget(x) do { atomic_inc(x); } while (0)
-
-#endif
-
-/* X86_FEATURE_NX is missing in some x86_64 kernels */
-
-#include <asm/cpufeature.h>
-
-#ifndef X86_FEATURE_NX
-#define X86_FEATURE_NX (1*32+20)
-#endif
-
-#undef true
-#define true 1
-#undef false
-#define false 0
-
-/* EFER_LMA and EFER_LME are missing in pre 2.6.24 i386 kernels */
-#ifndef EFER_LME
-#define _EFER_LME           8  /* Long mode enable */
-#define _EFER_LMA           10 /* Long mode active (read-only) */
-#define EFER_LME            (1<<_EFER_LME)
-#define EFER_LMA            (1<<_EFER_LMA)
-#endif
-
-/* pagefault_enable(), page_fault_disable() - 2.6.20 */
-#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,20)
-
-static inline void pagefault_disable(void)
-{
-	inc_preempt_count();
-	/*
-	 * make sure to have issued the store before a pagefault
-	 * can hit.
-	 */
-	barrier();
-}
-
-static inline void pagefault_enable(void)
-{
-	/*
-	 * make sure to issue those last loads/stores before enabling
-	 * the pagefault handler again.
-	 */
-	barrier();
-	dec_preempt_count();
-	/*
-	 * make sure we do..
-	 */
-	barrier();
-	preempt_check_resched();
-}
-
-#endif
-
-/* vm ops ->fault() was introduced in 2.6.23. */
-#include <linux/mm.h>
-
-#ifdef KVM_MAIN
-#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,23)
-
-struct vm_fault {
-	unsigned int flags;
-	pgoff_t pgoff;
-	void __user *virtual_address;
-	struct page *page;
-};
-
-static int kvm_vcpu_fault(struct vm_area_struct *vma, struct vm_fault *vmf);
-static int kvm_vm_fault(struct vm_area_struct *vma, struct vm_fault *vmf);
-
-static inline struct page *kvm_nopage_to_fault(
-	int (*fault)(struct vm_area_struct *vma, struct vm_fault *vmf),
-	struct vm_area_struct *vma,
-	unsigned long address,
-	int *type)
-{
-	struct vm_fault vmf;
-	int ret;
-
-	vmf.pgoff = ((address - vma->vm_start) >> PAGE_SHIFT) + vma->vm_pgoff;
-	vmf.virtual_address = (void __user *)address;
-	ret = fault(vma, &vmf);
-	if (ret)
-		return NOPAGE_SIGBUS;
-	*type = VM_FAULT_MINOR;
-	return vmf.page;
-}
-
-static inline struct page *__kvm_vcpu_fault(struct vm_area_struct *vma,
-					    unsigned long address,
-					    int *type)
-{
-	return kvm_nopage_to_fault(kvm_vcpu_fault, vma, address, type);
-}
-
-static inline struct page *__kvm_vm_fault(struct vm_area_struct *vma,
-					  unsigned long address,
-					  int *type)
-{
-	return kvm_nopage_to_fault(kvm_vm_fault, vma, address, type);
-}
-
-#define VMA_OPS_FAULT(x) nopage
-#define VMA_OPS_FAULT_FUNC(x) __##x
-
-#else
-
-#define VMA_OPS_FAULT(x) x
-#define VMA_OPS_FAULT_FUNC(x) x
-
-#endif
-#endif
-
-/* simple vfs attribute getter signature has changed to add a return code */
-
-#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,25)
-
-#define MAKE_SIMPLE_ATTRIBUTE_GETTER(x)       \
-	static u64 x(void *v)                 \
-	{				      \
-		u64 ret = 0;		      \
-					      \
-		__##x(v, &ret);		      \
-		return ret;		      \
-	}
-
-#else
-
-#define MAKE_SIMPLE_ATTRIBUTE_GETTER(x)       \
-	static int x(void *v, u64 *val)	      \
-	{				      \
-		return __##x(v, val);	      \
-	}
-
-#endif
-
-/* set_kset_name() is gone in 2.6.25 */
-
-#if LINUX_VERSION_CODE >= KERNEL_VERSION(2,6,25)
-
-#define set_kset_name(x) .name = x
-
-#endif
-
-#if LINUX_VERSION_CODE >= KERNEL_VERSION(2,6,25)
-#ifndef FASTCALL
-#define FASTCALL(x)	x
-#define fastcall
-#endif
-#endif
-
-#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,23)
-
-static unsigned  __attribute__((__used__)) kvm_tsc_khz = 2000000;
-
-#else
-
-#define kvm_tsc_khz tsc_khz
-
-#endif
-
-struct kvm_desc_struct {
-	union {
-		struct { unsigned int a, b; };
-		struct {
-			u16 limit0;
-			u16 base0;
-			unsigned base1: 8, type: 4, s: 1, dpl: 2, p: 1;
-			unsigned limit: 4, avl: 1, l: 1, d: 1, g: 1, base2: 8;
-		};
-
-	};
-} __attribute__((packed));
-
-struct kvm_ldttss_desc64 {
-	u16 limit0;
-	u16 base0;
-	unsigned base1 : 8, type : 5, dpl : 2, p : 1;
-	unsigned limit1 : 4, zero0 : 3, g : 1, base2 : 8;
-	u32 base3;
-	u32 zero1;
-} __attribute__((packed));
-
-struct kvm_desc_ptr {
-	unsigned short size;
-	unsigned long address;
-} __attribute__((packed));
-
-#if LINUX_VERSION_CODE <= KERNEL_VERSION(2,6,21)
-
-#include <linux/ktime.h>
-#include <linux/hrtimer.h>
-
-#define ktime_get kvm_ktime_get
-
-static inline ktime_t ktime_get(void)
-{
-	struct timespec now;
-
-	ktime_get_ts(&now);
-
-	return timespec_to_ktime(now);
-}
-
-#endif
-
-/* __aligned arrived in 2.6.21 */
-#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,21)
-#define __aligned(x) __attribute__((__aligned__(x)))
-#endif
-
-#include <asm/msr.h>
-#ifndef MSR_FS_BASE
-#define MSR_FS_BASE 0xc0000100
-#endif
-#ifndef MSR_GS_BASE
-#define MSR_GS_BASE 0xc0000101
-#endif
-
-#include <linux/mm.h>
-
-/* The shrinker API changed in 2.6.23 */
-#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,23)
-
-struct kvm_shrinker {
-	int (*shrink)(int nr_to_scan, gfp_t gfp_mask);
-	int seeks;
-	struct shrinker *kshrinker;
-};
-
-static inline void register_shrinker(struct kvm_shrinker *shrinker)
-{
-	shrinker->kshrinker = set_shrinker(shrinker->seeks, shrinker->shrink);
-}
-
-static inline void unregister_shrinker(struct kvm_shrinker *shrinker)
-{
-	if (shrinker->kshrinker)
-		remove_shrinker(shrinker->kshrinker);
-}
-
-#define shrinker kvm_shrinker
-
-#endif
-
-/* undefine lapic */
-#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,18)
-
-#undef lapic
-
-#endif
-
-/* clocksource */
-#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,18)
-static inline u32 clocksource_khz2mult(u32 khz, u32 shift_constant)
-{
-	/*  khz = cyc/(Million ns)
-	 *  mult/2^shift  = ns/cyc
-	 *  mult = ns/cyc * 2^shift
-	 *  mult = 1Million/khz * 2^shift
-	 *  mult = 1000000 * 2^shift / khz
-	 *  mult = (1000000<<shift) / khz
-	 */
-	u64 tmp = ((u64)1000000) << shift_constant;
-
-	tmp += khz/2; /* round for do_div */
-	do_div(tmp, khz);
-
-	return (u32)tmp;
-}
-#else
-#include <linux/clocksource.h>
-#endif
-
-/* manually export hrtimer_init/start/cancel */
-#include <linux/kallsyms.h>
-extern void (*hrtimer_init_p)(struct hrtimer *timer, clockid_t which_clock,
-			      enum hrtimer_mode mode);
-extern int (*hrtimer_start_p)(struct hrtimer *timer, ktime_t tim,
-			      const enum hrtimer_mode mode);
-extern int (*hrtimer_cancel_p)(struct hrtimer *timer);
-
-#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,17) && defined(CONFIG_KALLSYMS)
-static inline void hrtimer_kallsyms_resolve(void)
-{
-	hrtimer_init_p = (void *) kallsyms_lookup_name("hrtimer_init");
-	BUG_ON(!hrtimer_init_p);
-	hrtimer_start_p = (void *) kallsyms_lookup_name("hrtimer_start");
-	BUG_ON(!hrtimer_start_p);
-	hrtimer_cancel_p = (void *) kallsyms_lookup_name("hrtimer_cancel");
-	BUG_ON(!hrtimer_cancel_p);
-}
-#else
-static inline void hrtimer_kallsyms_resolve(void)
-{
-	hrtimer_init_p = hrtimer_init;
-	hrtimer_start_p = hrtimer_start;
-	hrtimer_cancel_p = hrtimer_cancel;
-}
-#endif
-
-/* handle old hrtimer API with data pointer */
-#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,17)
-static inline void hrtimer_data_pointer(struct hrtimer *timer)
-{
-	timer->data = (void *)timer;
-}
-#else
-static inline void hrtimer_data_pointer(struct hrtimer *timer) {}
-#endif
-
-#include <asm/hw_irq.h>
-#ifndef NMI_VECTOR
-#define NMI_VECTOR 2
-#endif
-
-#ifndef MSR_MTRRcap
-#define MSR_MTRRcap            0x0fe
-#define MSR_MTRRfix64K_00000   0x250
-#define MSR_MTRRfix16K_80000   0x258
-#define MSR_MTRRfix16K_A0000   0x259
-#define MSR_MTRRfix4K_C0000    0x268
-#define MSR_MTRRfix4K_C8000    0x269
-#define MSR_MTRRfix4K_D0000    0x26a
-#define MSR_MTRRfix4K_D8000    0x26b
-#define MSR_MTRRfix4K_E0000    0x26c
-#define MSR_MTRRfix4K_E8000    0x26d
-#define MSR_MTRRfix4K_F0000    0x26e
-#define MSR_MTRRfix4K_F8000    0x26f
-#define MSR_MTRRdefType        0x2ff
-#endif
-
-#ifndef MSR_IA32_CR_PAT
-#define MSR_IA32_CR_PAT        0x00000277
-#endif
-
-#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,22)
-
-#define ns_to_timespec kvm_ns_to_timespec
-
-struct timespec kvm_ns_to_timespec(const s64 nsec);
-
-#endif
-
-/* Define DEBUGCTLMSR bits */
-#ifndef DEBUGCTLMSR_LBR
-
-#define _DEBUGCTLMSR_LBR	0 /* last branch recording */
-#define _DEBUGCTLMSR_BTF	1 /* single-step on branches */
-
-#define DEBUGCTLMSR_LBR		(1UL << _DEBUGCTLMSR_LBR)
-#define DEBUGCTLMSR_BTF		(1UL << _DEBUGCTLMSR_BTF)
-
-#endif
diff --git a/kernel/debug.h b/kernel/x86/debug.h
similarity index 100%
rename from kernel/debug.h
rename to kernel/x86/debug.h
diff --git a/kernel/external-module-compat.c b/kernel/x86/external-module-compat.c
similarity index 100%
rename from kernel/external-module-compat.c
rename to kernel/x86/external-module-compat.c
diff --git a/kernel/x86/external-module-compat.h b/kernel/x86/external-module-compat.h
new file mode 100644
index 0000000..ce93551
--- /dev/null
+++ b/kernel/x86/external-module-compat.h
@@ -0,0 +1,291 @@
+
+/*
+ * Compatibility header for building as an external module.
+ */
+
+#include <linux/compiler.h>
+#include <linux/version.h>
+
+#include "../external-module-compat-comm.h"
+
+#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,22)
+
+#ifndef _EFER_SCE
+#define _EFER_SCE		0  /* SYSCALL/SYSRET */
+#endif
+
+#ifndef EFER_SCE
+#define EFER_SCE		(1<<_EFER_SCE)
+#endif
+
+#endif
+
+#include <linux/smp.h>
+
+#ifndef X86_CR0_PE
+#define X86_CR0_PE 0x00000001
+#endif
+
+#ifndef X86_CR0_MP
+#define X86_CR0_MP 0x00000002
+#endif
+
+#ifndef X86_CR0_EM
+#define X86_CR0_EM 0x00000004
+#endif
+
+#ifndef X86_CR0_TS
+#define X86_CR0_TS 0x00000008
+#endif
+
+#ifndef X86_CR0_ET
+#define X86_CR0_ET 0x00000010
+#endif
+
+#ifndef X86_CR0_NE
+#define X86_CR0_NE 0x00000020
+#endif
+
+#ifndef X86_CR0_WP
+#define X86_CR0_WP 0x00010000
+#endif
+
+#ifndef X86_CR0_AM
+#define X86_CR0_AM 0x00040000
+#endif
+
+#ifndef X86_CR0_NW
+#define X86_CR0_NW 0x20000000
+#endif
+
+#ifndef X86_CR0_CD
+#define X86_CR0_CD 0x40000000
+#endif
+
+#ifndef X86_CR0_PG
+#define X86_CR0_PG 0x80000000
+#endif
+
+#ifndef X86_CR3_PWT
+#define X86_CR3_PWT 0x00000008
+#endif
+
+#ifndef X86_CR3_PCD
+#define X86_CR3_PCD 0x00000010
+#endif
+
+#ifndef X86_CR4_VMXE
+#define X86_CR4_VMXE 0x00002000
+#endif
+
+#undef X86_CR8_TPR
+#define X86_CR8_TPR 0x0f
+
+/*
+ * 2.6.22 does not define set_64bit() under nonpae
+ */
+#ifdef CONFIG_X86_32
+
+#include <asm/cmpxchg.h>
+
+static inline void __kvm_set_64bit(u64 *ptr, u64 val)
+{
+	unsigned int low = val;
+	unsigned int high = val >> 32;
+
+	__asm__ __volatile__ (
+		"\n1:\t"
+		"movl (%0), %%eax\n\t"
+		"movl 4(%0), %%edx\n\t"
+		"lock cmpxchg8b (%0)\n\t"
+		"jnz 1b"
+		: /* no outputs */
+		:	"D"(ptr),
+			"b"(low),
+			"c"(high)
+		:	"ax","dx","memory");
+}
+
+#undef  set_64bit
+#define set_64bit __kvm_set_64bit
+
+static inline unsigned long long __kvm_cmpxchg64(volatile void *ptr,
+						 unsigned long long old,
+						 unsigned long long new)
+{
+	unsigned long long prev;
+	__asm__ __volatile__("lock cmpxchg8b %3"
+			     : "=A"(prev)
+			     : "b"((unsigned long)new),
+			       "c"((unsigned long)(new >> 32)),
+			       "m"(*__xg(ptr)),
+			       "0"(old)
+			     : "memory");
+	return prev;
+}
+
+#define kvm_cmpxchg64(ptr,o,n)\
+	((__typeof__(*(ptr)))__kvm_cmpxchg64((ptr),(unsigned long long)(o),\
+					(unsigned long long)(n)))
+
+#undef cmpxchg64
+#define cmpxchg64(ptr, o, n) kvm_cmpxchg64(ptr, o, n)
+
+#endif
+
+#ifndef CONFIG_PREEMPT_NOTIFIERS
+/*
+ * Include sched|preempt.h before defining CONFIG_PREEMPT_NOTIFIERS to avoid
+ * a miscompile.
+ */
+#include <linux/sched.h>
+#include <linux/preempt.h>
+#define CONFIG_PREEMPT_NOTIFIERS
+#define CONFIG_PREEMPT_NOTIFIERS_COMPAT
+
+struct preempt_notifier;
+
+struct preempt_ops {
+	void (*sched_in)(struct preempt_notifier *notifier, int cpu);
+	void (*sched_out)(struct preempt_notifier *notifier,
+			  struct task_struct *next);
+};
+
+struct preempt_notifier {
+	struct list_head link;
+	struct task_struct *tsk;
+	struct preempt_ops *ops;
+};
+
+void preempt_notifier_register(struct preempt_notifier *notifier);
+void preempt_notifier_unregister(struct preempt_notifier *notifier);
+
+static inline void preempt_notifier_init(struct preempt_notifier *notifier,
+				     struct preempt_ops *ops)
+{
+	notifier->ops = ops;
+}
+
+void start_special_insn(void);
+void end_special_insn(void);
+void in_special_section(void);
+void special_reload_dr7(void);
+
+void preempt_notifier_sys_init(void);
+void preempt_notifier_sys_exit(void);
+
+#else
+
+static inline void start_special_insn(void) {}
+static inline void end_special_insn(void) {}
+static inline void in_special_section(void) {}
+static inline void special_reload_dr7(void) {}
+
+static inline void preempt_notifier_sys_init(void) {}
+static inline void preempt_notifier_sys_exit(void) {}
+
+#endif
+
+/* CONFIG_HAS_IOMEM is apparently fairly new too (2.6.21 for x86_64). */
+#ifndef CONFIG_HAS_IOMEM
+#define CONFIG_HAS_IOMEM 1
+#endif
+
+/* X86_FEATURE_NX is missing in some x86_64 kernels */
+
+#include <asm/cpufeature.h>
+
+#ifndef X86_FEATURE_NX
+#define X86_FEATURE_NX (1*32+20)
+#endif
+
+#undef true
+#define true 1
+#undef false
+#define false 0
+
+/* EFER_LMA and EFER_LME are missing in pre 2.6.24 i386 kernels */
+#ifndef EFER_LME
+#define _EFER_LME           8  /* Long mode enable */
+#define _EFER_LMA           10 /* Long mode active (read-only) */
+#define EFER_LME            (1<<_EFER_LME)
+#define EFER_LMA            (1<<_EFER_LMA)
+#endif
+
+struct kvm_desc_struct {
+	union {
+		struct { unsigned int a, b; };
+		struct {
+			u16 limit0;
+			u16 base0;
+			unsigned base1: 8, type: 4, s: 1, dpl: 2, p: 1;
+			unsigned limit: 4, avl: 1, l: 1, d: 1, g: 1, base2: 8;
+		};
+
+	};
+} __attribute__((packed));
+
+struct kvm_ldttss_desc64 {
+	u16 limit0;
+	u16 base0;
+	unsigned base1 : 8, type : 5, dpl : 2, p : 1;
+	unsigned limit1 : 4, zero0 : 3, g : 1, base2 : 8;
+	u32 base3;
+	u32 zero1;
+} __attribute__((packed));
+
+struct kvm_desc_ptr {
+	unsigned short size;
+	unsigned long address;
+} __attribute__((packed));
+
+#include <asm/msr.h>
+#ifndef MSR_FS_BASE
+#define MSR_FS_BASE 0xc0000100
+#endif
+#ifndef MSR_GS_BASE
+#define MSR_GS_BASE 0xc0000101
+#endif
+
+/* undefine lapic */
+#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,18)
+
+#undef lapic
+
+#endif
+
+#include <asm/hw_irq.h>
+#ifndef NMI_VECTOR
+#define NMI_VECTOR 2
+#endif
+
+#ifndef MSR_MTRRcap
+#define MSR_MTRRcap            0x0fe
+#define MSR_MTRRfix64K_00000   0x250
+#define MSR_MTRRfix16K_80000   0x258
+#define MSR_MTRRfix16K_A0000   0x259
+#define MSR_MTRRfix4K_C0000    0x268
+#define MSR_MTRRfix4K_C8000    0x269
+#define MSR_MTRRfix4K_D0000    0x26a
+#define MSR_MTRRfix4K_D8000    0x26b
+#define MSR_MTRRfix4K_E0000    0x26c
+#define MSR_MTRRfix4K_E8000    0x26d
+#define MSR_MTRRfix4K_F0000    0x26e
+#define MSR_MTRRfix4K_F8000    0x26f
+#define MSR_MTRRdefType        0x2ff
+#endif
+
+#ifndef MSR_IA32_CR_PAT
+#define MSR_IA32_CR_PAT        0x00000277
+#endif
+
+/* Define DEBUGCTLMSR bits */
+#ifndef DEBUGCTLMSR_LBR
+
+#define _DEBUGCTLMSR_LBR	0 /* last branch recording */
+#define _DEBUGCTLMSR_BTF	1 /* single-step on branches */
+
+#define DEBUGCTLMSR_LBR		(1UL << _DEBUGCTLMSR_LBR)
+#define DEBUGCTLMSR_BTF		(1UL << _DEBUGCTLMSR_BTF)
+
+#endif
diff --git a/kernel/preempt.c b/kernel/x86/preempt.c
similarity index 100%
rename from kernel/preempt.c
rename to kernel/x86/preempt.c
diff --git a/kernel/vmx-debug.c b/kernel/x86/vmx-debug.c
similarity index 100%
rename from kernel/vmx-debug.c
rename to kernel/x86/vmx-debug.c
-- 
1.5.6.2


^ permalink raw reply related	[flat|nested] 24+ messages in thread

* RE: [PATCH 01/04]Create x86 directory to hold x86-specific files.
@ 2008-07-30  1:42   ` Zhang, Xiantao
  0 siblings, 0 replies; 24+ messages in thread
From: Zhang, Xiantao @ 2008-07-30  1:42 UTC (permalink / raw)
  To: Avi Kivity; +Cc: kvm-ia64, kvm

[-- Attachment #1: Type: text/plain, Size: 793 bytes --]

Avi Kivity wrote:
> Zhang, Xiantao wrote:
>> Avi Kivity wrote:
>> 
>>> Zhang, Xiantao wrote:
>>> 
>>>> From c8c8c5a6b3ff943cb645f9ac4ef169490f14cd08 Mon Sep 17 00:00:00
>>>> 2001 From: Zhang Xiantao <xiantao.zhang@intel.com>
>>>> Date: Fri, 25 Jul 2008 16:50:30 +0800
>>>> Subject: [PATCH] KVM: external module: Moving x86-speicif files to
>>>> x86 directory. 
>>>> 
>>>> Create x86 directory to hold x86-specific files.
>>>> Signed-off-by: Zhang Xiantao <xiantao.zhang@intel.com>
>>>> 
>>>> 
>>> You forgot to attach a patch; and your mailer corrupts patches, so
>>> I can't apply this. 
>>> 
>> 
>> Sorry. Attached! :)
>> 
> 
> Sorry 2, external-module-compat.h has changed!  Please update and I'll
> apply it quicker this time.

Updated and attached it again :-)

[-- Attachment #2: 0001-KVM-external-module-Moving-x86-speicif-files-to-x8.patch --]
[-- Type: application/octet-stream, Size: 38940 bytes --]

From dae628bf166e6b2c2a8c86340c0f50dbc1746af5 Mon Sep 17 00:00:00 2001
From: Zhang Xiantao <xiantao.zhang@intel.com>
Date: Fri, 25 Jul 2008 16:50:30 +0800
Subject: [PATCH] KVM: external module: Moving x86-speicif files to x86 directory.

Create x86 directory to hold x86-specific files.
Signed-off-by: Zhang Xiantao <xiantao.zhang@intel.com>
---
 kernel/Kbuild                             |    8 +-
 kernel/Makefile                           |    2 +-
 kernel/external-module-compat-comm.h      |  497 ++++++++++++++++++
 kernel/external-module-compat.h           |  781 -----------------------------
 kernel/{ => x86}/debug.h                  |    0 
 kernel/{ => x86}/external-module-compat.c |    0 
 kernel/x86/external-module-compat.h       |  291 +++++++++++
 kernel/{ => x86}/preempt.c                |    0 
 kernel/{ => x86}/vmx-debug.c              |    0 
 9 files changed, 793 insertions(+), 786 deletions(-)
 create mode 100644 kernel/external-module-compat-comm.h
 delete mode 100644 kernel/external-module-compat.h
 rename kernel/{ => x86}/debug.h (100%)
 rename kernel/{ => x86}/external-module-compat.c (100%)
 create mode 100644 kernel/x86/external-module-compat.h
 rename kernel/{ => x86}/preempt.c (100%)
 rename kernel/{ => x86}/vmx-debug.c (100%)

diff --git a/kernel/Kbuild b/kernel/Kbuild
index 676f28d..cb63582 100644
--- a/kernel/Kbuild
+++ b/kernel/Kbuild
@@ -1,11 +1,11 @@
 obj-m := kvm.o kvm-intel.o kvm-amd.o
 kvm-objs := kvm_main.o x86.o mmu.o x86_emulate.o anon_inodes.o irq.o i8259.o \
-	 lapic.o ioapic.o preempt.o i8254.o coalesced_mmio.o \
-	 external-module-compat.o
+	 lapic.o ioapic.o x86/preempt.o i8254.o coalesced_mmio.o \
+	 x86/external-module-compat.o
 ifeq ($(CONFIG_KVM_TRACE),y)
 kvm-objs += kvm_trace.o
 endif
-kvm-intel-objs := vmx.o vmx-debug.o external-module-compat.o
-kvm-amd-objs := svm.o external-module-compat.o
+kvm-intel-objs := vmx.o x86/vmx-debug.o x86/external-module-compat.o
+kvm-amd-objs := svm.o x86/external-module-compat.o
 
 CFLAGS_kvm_main.o = -DKVM_MAIN
diff --git a/kernel/Makefile b/kernel/Makefile
index 69b4981..c2c666e 100644
--- a/kernel/Makefile
+++ b/kernel/Makefile
@@ -35,7 +35,7 @@ all::
 	$(MAKE) -C $(KERNELDIR) M=`pwd` \
 		LINUXINCLUDE="-I`pwd`/include -Iinclude -I`pwd`/include-compat \
 		-include include/linux/autoconf.h \
-		-include `pwd`/external-module-compat.h"
+		-include `pwd`/$(ARCH_DIR)/external-module-compat.h"
 		"$$@"
 
 sync: header-sync source-sync
diff --git a/kernel/external-module-compat-comm.h b/kernel/external-module-compat-comm.h
new file mode 100644
index 0000000..f5169b8
--- /dev/null
+++ b/kernel/external-module-compat-comm.h
@@ -0,0 +1,497 @@
+
+/*
+ * Compatibility header for building as an external module.
+ */
+
+/*
+ * Avoid picking up the kernel's kvm.h in case we have a newer one.
+ */
+
+#include <linux/compiler.h>
+#include <linux/version.h>
+#include <linux/string.h>
+#include <linux/kvm.h>
+#include <linux/kvm_para.h>
+#include <linux/cpu.h>
+#include <linux/time.h>
+#include <asm/processor.h>
+#include <linux/hrtimer.h>
+#include <asm/bitops.h>
+#include <asm/msr.h>
+
+/*
+ * 2.6.16 does not have GFP_NOWAIT
+ */
+
+#include <linux/gfp.h>
+
+#ifndef GFP_NOWAIT
+#define GFP_NOWAIT (GFP_ATOMIC & ~__GFP_HIGH)
+#endif
+
+
+/*
+ * kvm profiling support needs 2.6.20
+ */
+#include <linux/profile.h>
+
+#ifndef KVM_PROFILING
+#define KVM_PROFILING 1234
+#define prof_on       4321
+#endif
+
+/*
+ * smp_call_function_single() is not exported below 2.6.20, and has different
+ * semantics below 2.6.23.  The 'nonatomic' argument was removed in 2.6.27.
+ */
+#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,27)
+
+int kvm_smp_call_function_single(int cpu, void (*func)(void *info),
+				 void *info, int wait);
+
+#define smp_call_function_single kvm_smp_call_function_single
+
+#endif
+
+/* on_each_cpu() lost an argument in 2.6.27. */
+#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,27)
+
+#define kvm_on_each_cpu(func, info, wait) on_each_cpu(func, info, 0, wait)
+
+#else
+
+#define kvm_on_each_cpu(func, info, wait) on_each_cpu(func, info, wait)
+
+#endif
+
+/*
+ * The cpu hotplug stubs are broken if !CONFIG_CPU_HOTPLUG
+ */
+
+#if LINUX_VERSION_CODE <= KERNEL_VERSION(2,6,15)
+#define DEFINE_MUTEX(a) DECLARE_MUTEX(a)
+#define mutex_lock_interruptible(a) down_interruptible(a)
+#define mutex_unlock(a) up(a)
+#define mutex_lock(a) down(a)
+#define mutex_init(a) init_MUTEX(a)
+#define mutex_trylock(a) down_trylock(a)
+#define mutex semaphore
+#endif
+
+#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,14)
+#ifndef kzalloc
+#define kzalloc(size,flags)			\
+({						\
+	void *__ret = kmalloc(size, flags);	\
+	if (__ret)				\
+		memset(__ret, 0, size);		\
+	__ret;					\
+})
+#endif
+#endif
+
+#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,17)
+#ifndef kmem_cache_zalloc
+#define kmem_cache_zalloc(cache,flags)			  \
+({							  \
+	void *__ret = kmem_cache_alloc(cache, flags);	  \
+	if (__ret)                                        \
+		memset(__ret, 0, kmem_cache_size(cache)); \
+	__ret;                                            \
+})
+#endif
+#endif
+
+#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,21)
+
+#ifndef CONFIG_HOTPLUG_CPU
+#define register_cpu_notifier(nb) (0)
+#endif
+
+#endif
+
+#include <linux/miscdevice.h>
+#ifndef KVM_MINOR
+#define KVM_MINOR 232
+#endif
+
+#include <linux/notifier.h>
+#ifndef CPU_TASKS_FROZEN
+
+#define CPU_TASKS_FROZEN       0x0010
+#define CPU_ONLINE_FROZEN      (CPU_ONLINE | CPU_TASKS_FROZEN)
+#define CPU_UP_PREPARE_FROZEN  (CPU_UP_PREPARE | CPU_TASKS_FROZEN)
+#define CPU_UP_CANCELED_FROZEN (CPU_UP_CANCELED | CPU_TASKS_FROZEN)
+#define CPU_DOWN_PREPARE_FROZEN        (CPU_DOWN_PREPARE | CPU_TASKS_FROZEN)
+#define CPU_DOWN_FAILED_FROZEN (CPU_DOWN_FAILED | CPU_TASKS_FROZEN)
+#define CPU_DEAD_FROZEN                (CPU_DEAD | CPU_TASKS_FROZEN)
+
+#endif
+
+#ifndef CPU_DYING
+#define CPU_DYING 0x000A
+#define CPU_DYING_FROZEN (CPU_DYING | CPU_TASKS_FROZEN)
+#endif
+
+#include <asm/system.h>
+
+struct inode;
+#include <linux/anon_inodes.h>
+#define anon_inode_getfd kvm_anon_inode_getfd
+int kvm_init_anon_inodes(void);
+void kvm_exit_anon_inodes(void);
+int anon_inode_getfd(const char *name,
+		     const struct file_operations *fops,
+		     void *priv , int flags);
+
+/*
+ * 2.6.23 removed the cache destructor
+ */
+#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,23)
+#  define kmem_cache_create(name, size, align, flags, ctor) \
+	kmem_cache_create(name, size, align, flags, ctor, NULL)
+#endif
+
+/* HRTIMER_MODE_ABS started life with a different name */
+#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,21)
+#define HRTIMER_MODE_ABS HRTIMER_ABS
+#endif
+
+/* div64_u64 is fairly new */
+#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,26)
+
+#define div64_u64 kvm_div64_u64
+
+#ifdef CONFIG_64BIT
+
+static inline uint64_t div64_u64(uint64_t dividend, uint64_t divisor)
+{
+	return dividend / divisor;
+}
+
+#else
+
+uint64_t div64_u64(uint64_t dividend, uint64_t divisor);
+
+#endif
+
+#endif
+
+#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,19)
+
+#ifdef RHEL_RELEASE_CODE
+#if RHEL_RELEASE_CODE >= RHEL_RELEASE_VERSION(5,2)
+#define RHEL_BOOL 1
+#endif
+#endif
+
+#ifndef RHEL_BOOL
+
+typedef _Bool bool;
+
+#endif
+
+#endif
+
+/*
+ * PF_VCPU is a Linux 2.6.24 addition
+ */
+
+#include <linux/sched.h>
+
+#ifndef PF_VCPU
+#define PF_VCPU 0
+#endif
+
+/*
+ * smp_call_function_mask() is not defined/exported below 2.6.24
+ */
+
+#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,24)
+
+int kvm_smp_call_function_mask(cpumask_t mask, void (*func) (void *info),
+			       void *info, int wait);
+
+#define smp_call_function_mask kvm_smp_call_function_mask
+
+#endif
+
+/* empty_zero_page isn't exported in all kernels */
+#include <asm/pgtable.h>
+
+#define empty_zero_page kvm_empty_zero_page
+
+static char empty_zero_page[PAGE_SIZE];
+
+static inline void blahblah(void)
+{
+	(void)empty_zero_page[0];
+}
+
+/* __mmdrop() is not exported before 2.6.25 */
+#include <linux/sched.h>
+
+#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,25)
+
+#define mmdrop(x) do { (void)(x); } while (0)
+#define mmget(x) do { (void)(x); } while (0)
+
+#else
+
+#define mmget(x) do { atomic_inc(x); } while (0)
+
+#endif
+
+/* pagefault_enable(), page_fault_disable() - 2.6.20 */
+#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,20)
+
+static inline void pagefault_disable(void)
+{
+	inc_preempt_count();
+	/*
+	 * make sure to have issued the store before a pagefault
+	 * can hit.
+	 */
+	barrier();
+}
+
+static inline void pagefault_enable(void)
+{
+	/*
+	 * make sure to issue those last loads/stores before enabling
+	 * the pagefault handler again.
+	 */
+	barrier();
+	dec_preempt_count();
+	/*
+	 * make sure we do..
+	 */
+	barrier();
+	preempt_check_resched();
+}
+
+#endif
+
+/* vm ops ->fault() was introduced in 2.6.23. */
+#include <linux/mm.h>
+
+#ifdef KVM_MAIN
+#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,23)
+
+struct vm_fault {
+	unsigned int flags;
+	pgoff_t pgoff;
+	void __user *virtual_address;
+	struct page *page;
+};
+
+static int kvm_vcpu_fault(struct vm_area_struct *vma, struct vm_fault *vmf);
+static int kvm_vm_fault(struct vm_area_struct *vma, struct vm_fault *vmf);
+
+static inline struct page *kvm_nopage_to_fault(
+	int (*fault)(struct vm_area_struct *vma, struct vm_fault *vmf),
+	struct vm_area_struct *vma,
+	unsigned long address,
+	int *type)
+{
+	struct vm_fault vmf;
+	int ret;
+
+	vmf.pgoff = ((address - vma->vm_start) >> PAGE_SHIFT) + vma->vm_pgoff;
+	vmf.virtual_address = (void __user *)address;
+	ret = fault(vma, &vmf);
+	if (ret)
+		return NOPAGE_SIGBUS;
+	*type = VM_FAULT_MINOR;
+	return vmf.page;
+}
+
+static inline struct page *__kvm_vcpu_fault(struct vm_area_struct *vma,
+					    unsigned long address,
+					    int *type)
+{
+	return kvm_nopage_to_fault(kvm_vcpu_fault, vma, address, type);
+}
+
+static inline struct page *__kvm_vm_fault(struct vm_area_struct *vma,
+					  unsigned long address,
+					  int *type)
+{
+	return kvm_nopage_to_fault(kvm_vm_fault, vma, address, type);
+}
+
+#define VMA_OPS_FAULT(x) nopage
+#define VMA_OPS_FAULT_FUNC(x) __##x
+
+#else
+
+#define VMA_OPS_FAULT(x) x
+#define VMA_OPS_FAULT_FUNC(x) x
+
+#endif
+#endif
+
+/* simple vfs attribute getter signature has changed to add a return code */
+
+#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,25)
+
+#define MAKE_SIMPLE_ATTRIBUTE_GETTER(x)       \
+	static u64 x(void *v)                 \
+	{				      \
+		u64 ret = 0;		      \
+					      \
+		__##x(v, &ret);		      \
+		return ret;		      \
+	}
+
+#else
+
+#define MAKE_SIMPLE_ATTRIBUTE_GETTER(x)       \
+	static int x(void *v, u64 *val)	      \
+	{				      \
+		return __##x(v, val);	      \
+	}
+
+#endif
+
+/* set_kset_name() is gone in 2.6.25 */
+
+#if LINUX_VERSION_CODE >= KERNEL_VERSION(2,6,25)
+
+#define set_kset_name(x) .name = x
+
+#endif
+
+#if LINUX_VERSION_CODE >= KERNEL_VERSION(2,6,25)
+#ifndef FASTCALL
+#define FASTCALL(x)	x
+#define fastcall
+#endif
+#endif
+
+#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,23)
+
+static unsigned  __attribute__((__used__)) kvm_tsc_khz = 2000000;
+
+#else
+
+#define kvm_tsc_khz tsc_khz
+
+#endif
+
+#if LINUX_VERSION_CODE <= KERNEL_VERSION(2,6,21)
+
+#include <linux/ktime.h>
+#include <linux/hrtimer.h>
+
+#define ktime_get kvm_ktime_get
+
+static inline ktime_t ktime_get(void)
+{
+	struct timespec now;
+
+	ktime_get_ts(&now);
+
+	return timespec_to_ktime(now);
+}
+
+#endif
+
+/* __aligned arrived in 2.6.21 */
+#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,21)
+#define __aligned(x) __attribute__((__aligned__(x)))
+#endif
+
+#include <linux/mm.h>
+
+/* The shrinker API changed in 2.6.23 */
+#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,23)
+
+struct kvm_shrinker {
+	int (*shrink)(int nr_to_scan, gfp_t gfp_mask);
+	int seeks;
+	struct shrinker *kshrinker;
+};
+
+static inline void register_shrinker(struct kvm_shrinker *shrinker)
+{
+	shrinker->kshrinker = set_shrinker(shrinker->seeks, shrinker->shrink);
+}
+
+static inline void unregister_shrinker(struct kvm_shrinker *shrinker)
+{
+	if (shrinker->kshrinker)
+		remove_shrinker(shrinker->kshrinker);
+}
+
+#define shrinker kvm_shrinker
+
+#endif
+
+/* clocksource */
+#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,18)
+static inline u32 clocksource_khz2mult(u32 khz, u32 shift_constant)
+{
+	/*  khz = cyc/(Million ns)
+	 *  mult/2^shift  = ns/cyc
+	 *  mult = ns/cyc * 2^shift
+	 *  mult = 1Million/khz * 2^shift
+	 *  mult = 1000000 * 2^shift / khz
+	 *  mult = (1000000<<shift) / khz
+	 */
+	u64 tmp = ((u64)1000000) << shift_constant;
+
+	tmp += khz/2; /* round for do_div */
+	do_div(tmp, khz);
+
+	return (u32)tmp;
+}
+#else
+#include <linux/clocksource.h>
+#endif
+
+/* manually export hrtimer_init/start/cancel */
+#include <linux/kallsyms.h>
+extern void (*hrtimer_init_p)(struct hrtimer *timer, clockid_t which_clock,
+			      enum hrtimer_mode mode);
+extern int (*hrtimer_start_p)(struct hrtimer *timer, ktime_t tim,
+			      const enum hrtimer_mode mode);
+extern int (*hrtimer_cancel_p)(struct hrtimer *timer);
+
+#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,17) && defined(CONFIG_KALLSYMS)
+static inline void hrtimer_kallsyms_resolve(void)
+{
+	hrtimer_init_p = (void *) kallsyms_lookup_name("hrtimer_init");
+	BUG_ON(!hrtimer_init_p);
+	hrtimer_start_p = (void *) kallsyms_lookup_name("hrtimer_start");
+	BUG_ON(!hrtimer_start_p);
+	hrtimer_cancel_p = (void *) kallsyms_lookup_name("hrtimer_cancel");
+	BUG_ON(!hrtimer_cancel_p);
+}
+#else
+static inline void hrtimer_kallsyms_resolve(void)
+{
+	hrtimer_init_p = hrtimer_init;
+	hrtimer_start_p = hrtimer_start;
+	hrtimer_cancel_p = hrtimer_cancel;
+}
+#endif
+
+/* handle old hrtimer API with data pointer */
+#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,17)
+static inline void hrtimer_data_pointer(struct hrtimer *timer)
+{
+	timer->data = (void *)timer;
+}
+#else
+static inline void hrtimer_data_pointer(struct hrtimer *timer) {}
+#endif
+
+#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,22)
+
+#define ns_to_timespec kvm_ns_to_timespec
+
+struct timespec kvm_ns_to_timespec(const s64 nsec);
+
+#endif
+
diff --git a/kernel/external-module-compat.h b/kernel/external-module-compat.h
deleted file mode 100644
index a27103c..0000000
--- a/kernel/external-module-compat.h
+++ /dev/null
@@ -1,781 +0,0 @@
-
-/*
- * Compatibility header for building as an external module.
- */
-
-/*
- * Avoid picking up the kernel's kvm.h in case we have a newer one.
- */
-
-#include <linux/compiler.h>
-#include <linux/version.h>
-#include <linux/string.h>
-#include <linux/kvm.h>
-#include <linux/kvm_para.h>
-#include <linux/cpu.h>
-#include <linux/time.h>
-#include <asm/processor.h>
-#include <linux/hrtimer.h>
-#include <asm/bitops.h>
-#include <asm/msr.h>
-
-/*
- * 2.6.16 does not have GFP_NOWAIT
- */
-
-#include <linux/gfp.h>
-
-#ifndef GFP_NOWAIT
-#define GFP_NOWAIT (GFP_ATOMIC & ~__GFP_HIGH)
-#endif
-
-
-/*
- * kvm profiling support needs 2.6.20
- */
-#include <linux/profile.h>
-
-#ifndef KVM_PROFILING
-#define KVM_PROFILING 1234
-#define prof_on       4321
-#endif
-
-/*
- * smp_call_function_single() is not exported below 2.6.20, and has different
- * semantics below 2.6.23.  The 'nonatomic' argument was removed in 2.6.27.
- */
-#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,27)
-
-int kvm_smp_call_function_single(int cpu, void (*func)(void *info),
-				 void *info, int wait);
-
-#define smp_call_function_single kvm_smp_call_function_single
-
-#endif
-
-/* on_each_cpu() lost an argument in 2.6.27. */
-#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,27)
-
-#define kvm_on_each_cpu(func, info, wait) on_each_cpu(func, info, 0, wait)
-
-#else
-
-#define kvm_on_each_cpu(func, info, wait) on_each_cpu(func, info, wait)
-
-#endif
-
-/*
- * The cpu hotplug stubs are broken if !CONFIG_CPU_HOTPLUG
- */
-
-#if LINUX_VERSION_CODE <= KERNEL_VERSION(2,6,15)
-#define DEFINE_MUTEX(a) DECLARE_MUTEX(a)
-#define mutex_lock_interruptible(a) down_interruptible(a)
-#define mutex_unlock(a) up(a)
-#define mutex_lock(a) down(a)
-#define mutex_init(a) init_MUTEX(a)
-#define mutex_trylock(a) down_trylock(a)
-#define mutex semaphore
-#endif
-
-#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,14)
-#ifndef kzalloc
-#define kzalloc(size,flags)			\
-({						\
-	void *__ret = kmalloc(size, flags);	\
-	if (__ret)				\
-		memset(__ret, 0, size);		\
-	__ret;					\
-})
-#endif
-#endif
-
-#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,17)
-#ifndef kmem_cache_zalloc
-#define kmem_cache_zalloc(cache,flags)			  \
-({							  \
-	void *__ret = kmem_cache_alloc(cache, flags);	  \
-	if (__ret)                                        \
-		memset(__ret, 0, kmem_cache_size(cache)); \
-	__ret;                                            \
-})
-#endif
-#endif
-
-#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,21)
-
-#ifndef CONFIG_HOTPLUG_CPU
-#define register_cpu_notifier(nb) (0)
-#endif
-
-#endif
-
-#include <linux/miscdevice.h>
-#ifndef KVM_MINOR
-#define KVM_MINOR 232
-#endif
-
-#include <linux/notifier.h>
-#ifndef CPU_TASKS_FROZEN
-
-#define CPU_TASKS_FROZEN       0x0010
-#define CPU_ONLINE_FROZEN      (CPU_ONLINE | CPU_TASKS_FROZEN)
-#define CPU_UP_PREPARE_FROZEN  (CPU_UP_PREPARE | CPU_TASKS_FROZEN)
-#define CPU_UP_CANCELED_FROZEN (CPU_UP_CANCELED | CPU_TASKS_FROZEN)
-#define CPU_DOWN_PREPARE_FROZEN        (CPU_DOWN_PREPARE | CPU_TASKS_FROZEN)
-#define CPU_DOWN_FAILED_FROZEN (CPU_DOWN_FAILED | CPU_TASKS_FROZEN)
-#define CPU_DEAD_FROZEN                (CPU_DEAD | CPU_TASKS_FROZEN)
-
-#endif
-
-#ifndef CPU_DYING
-#define CPU_DYING 0x000A
-#define CPU_DYING_FROZEN (CPU_DYING | CPU_TASKS_FROZEN)
-#endif
-
-#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,22)
-
-#ifndef _EFER_SCE
-#define _EFER_SCE		0  /* SYSCALL/SYSRET */
-#endif
-
-#ifndef EFER_SCE
-#define EFER_SCE		(1<<_EFER_SCE)
-#endif
-
-#endif
-
-/*
- * For set_64bit(), which is in a new file asm/cmpxchg.h in newer kernels.
- */
-#include <asm/system.h>
-
-struct inode;
-#include <linux/anon_inodes.h>
-#define anon_inode_getfd kvm_anon_inode_getfd
-int kvm_init_anon_inodes(void);
-void kvm_exit_anon_inodes(void);
-int anon_inode_getfd(const char *name,
-		     const struct file_operations *fops,
-		     void *priv, int flags);
-
-#include <linux/smp.h>
-
-#ifndef X86_CR0_PE
-#define X86_CR0_PE 0x00000001
-#endif
-
-#ifndef X86_CR0_MP
-#define X86_CR0_MP 0x00000002
-#endif
-
-#ifndef X86_CR0_EM
-#define X86_CR0_EM 0x00000004
-#endif
-
-#ifndef X86_CR0_TS
-#define X86_CR0_TS 0x00000008
-#endif
-
-#ifndef X86_CR0_ET
-#define X86_CR0_ET 0x00000010
-#endif
-
-#ifndef X86_CR0_NE
-#define X86_CR0_NE 0x00000020
-#endif
-
-#ifndef X86_CR0_WP
-#define X86_CR0_WP 0x00010000
-#endif
-
-#ifndef X86_CR0_AM
-#define X86_CR0_AM 0x00040000
-#endif
-
-#ifndef X86_CR0_NW
-#define X86_CR0_NW 0x20000000
-#endif
-
-#ifndef X86_CR0_CD
-#define X86_CR0_CD 0x40000000
-#endif
-
-#ifndef X86_CR0_PG
-#define X86_CR0_PG 0x80000000
-#endif
-
-#ifndef X86_CR3_PWT
-#define X86_CR3_PWT 0x00000008
-#endif
-
-#ifndef X86_CR3_PCD
-#define X86_CR3_PCD 0x00000010
-#endif
-
-#ifndef X86_CR4_VMXE
-#define X86_CR4_VMXE 0x00002000
-#endif
-
-#undef X86_CR8_TPR
-#define X86_CR8_TPR 0x0f
-
-/*
- * 2.6.23 removed the cache destructor
- */
-#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,23)
-#  define kmem_cache_create(name, size, align, flags, ctor) \
-	kmem_cache_create(name, size, align, flags, ctor, NULL)
-#endif
-
-/*
- * 2.6.22 does not define set_64bit() under nonpae
- */
-#ifdef CONFIG_X86_32
-
-#include <asm/cmpxchg.h>
-
-static inline void __kvm_set_64bit(u64 *ptr, u64 val)
-{
-	unsigned int low = val;
-	unsigned int high = val >> 32;
-
-	__asm__ __volatile__ (
-		"\n1:\t"
-		"movl (%0), %%eax\n\t"
-		"movl 4(%0), %%edx\n\t"
-		"lock cmpxchg8b (%0)\n\t"
-		"jnz 1b"
-		: /* no outputs */
-		:	"D"(ptr),
-			"b"(low),
-			"c"(high)
-		:	"ax","dx","memory");
-}
-
-#undef  set_64bit
-#define set_64bit __kvm_set_64bit
-
-static inline unsigned long long __kvm_cmpxchg64(volatile void *ptr,
-						 unsigned long long old,
-						 unsigned long long new)
-{
-	unsigned long long prev;
-	__asm__ __volatile__("lock cmpxchg8b %3"
-			     : "=A"(prev)
-			     : "b"((unsigned long)new),
-			       "c"((unsigned long)(new >> 32)),
-			       "m"(*__xg(ptr)),
-			       "0"(old)
-			     : "memory");
-	return prev;
-}
-
-#define kvm_cmpxchg64(ptr,o,n)\
-	((__typeof__(*(ptr)))__kvm_cmpxchg64((ptr),(unsigned long long)(o),\
-					(unsigned long long)(n)))
-
-#undef cmpxchg64
-#define cmpxchg64(ptr, o, n) kvm_cmpxchg64(ptr, o, n)
-
-#endif
-
-#ifndef CONFIG_PREEMPT_NOTIFIERS
-/*
- * Include sched|preempt.h before defining CONFIG_PREEMPT_NOTIFIERS to avoid
- * a miscompile.
- */
-#include <linux/sched.h>
-#include <linux/preempt.h>
-#define CONFIG_PREEMPT_NOTIFIERS
-#define CONFIG_PREEMPT_NOTIFIERS_COMPAT
-
-struct preempt_notifier;
-
-struct preempt_ops {
-	void (*sched_in)(struct preempt_notifier *notifier, int cpu);
-	void (*sched_out)(struct preempt_notifier *notifier,
-			  struct task_struct *next);
-};
-
-struct preempt_notifier {
-	struct list_head link;
-	struct task_struct *tsk;
-	struct preempt_ops *ops;
-};
-
-void preempt_notifier_register(struct preempt_notifier *notifier);
-void preempt_notifier_unregister(struct preempt_notifier *notifier);
-
-static inline void preempt_notifier_init(struct preempt_notifier *notifier,
-				     struct preempt_ops *ops)
-{
-	notifier->ops = ops;
-}
-
-void start_special_insn(void);
-void end_special_insn(void);
-void in_special_section(void);
-void special_reload_dr7(void);
-
-void preempt_notifier_sys_init(void);
-void preempt_notifier_sys_exit(void);
-
-#else
-
-static inline void start_special_insn(void) {}
-static inline void end_special_insn(void) {}
-static inline void in_special_section(void) {}
-static inline void special_reload_dr7(void) {}
-
-static inline void preempt_notifier_sys_init(void) {}
-static inline void preempt_notifier_sys_exit(void) {}
-
-#endif
-
-/* HRTIMER_MODE_ABS started life with a different name */
-#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,21)
-#define HRTIMER_MODE_ABS HRTIMER_ABS
-#endif
-
-/* div64_u64 is fairly new */
-#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,26)
-
-#define div64_u64 kvm_div64_u64
-
-#ifdef CONFIG_64BIT
-
-static inline uint64_t div64_u64(uint64_t dividend, uint64_t divisor)
-{
-	return dividend / divisor;
-}
-
-#else
-
-uint64_t div64_u64(uint64_t dividend, uint64_t divisor);
-
-#endif
-
-#endif
-
-#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,19)
-
-#ifdef RHEL_RELEASE_CODE
-#if RHEL_RELEASE_CODE >= RHEL_RELEASE_VERSION(5,2)
-#define RHEL_BOOL 1
-#endif
-#endif
-
-#ifndef RHEL_BOOL
-
-typedef _Bool bool;
-
-#endif
-
-#endif
-
-/*
- * PF_VCPU is a Linux 2.6.24 addition
- */
-
-#include <linux/sched.h>
-
-#ifndef PF_VCPU
-#define PF_VCPU 0
-#endif
-
-/*
- * smp_call_function_mask() is not defined/exported below 2.6.24
- */
-
-#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,24)
-
-int kvm_smp_call_function_mask(cpumask_t mask, void (*func) (void *info),
-			       void *info, int wait);
-
-#define smp_call_function_mask kvm_smp_call_function_mask
-
-#endif
-
-/* CONFIG_HAS_IOMEM is apparently fairly new too (2.6.21 for x86_64). */
-#ifndef CONFIG_HAS_IOMEM
-#define CONFIG_HAS_IOMEM 1
-#endif
-
-/* empty_zero_page isn't exported in all kernels */
-#include <asm/pgtable.h>
-
-#define empty_zero_page kvm_empty_zero_page
-
-static char empty_zero_page[PAGE_SIZE];
-
-static inline void blahblah(void)
-{
-	(void)empty_zero_page[0];
-}
-
-/* __mmdrop() is not exported before 2.6.25 */
-#include <linux/sched.h>
-
-#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,25)
-
-#define mmdrop(x) do { (void)(x); } while (0)
-#define mmget(x) do { (void)(x); } while (0)
-
-#else
-
-#define mmget(x) do { atomic_inc(x); } while (0)
-
-#endif
-
-/* X86_FEATURE_NX is missing in some x86_64 kernels */
-
-#include <asm/cpufeature.h>
-
-#ifndef X86_FEATURE_NX
-#define X86_FEATURE_NX (1*32+20)
-#endif
-
-#undef true
-#define true 1
-#undef false
-#define false 0
-
-/* EFER_LMA and EFER_LME are missing in pre 2.6.24 i386 kernels */
-#ifndef EFER_LME
-#define _EFER_LME           8  /* Long mode enable */
-#define _EFER_LMA           10 /* Long mode active (read-only) */
-#define EFER_LME            (1<<_EFER_LME)
-#define EFER_LMA            (1<<_EFER_LMA)
-#endif
-
-/* pagefault_enable(), page_fault_disable() - 2.6.20 */
-#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,20)
-
-static inline void pagefault_disable(void)
-{
-	inc_preempt_count();
-	/*
-	 * make sure to have issued the store before a pagefault
-	 * can hit.
-	 */
-	barrier();
-}
-
-static inline void pagefault_enable(void)
-{
-	/*
-	 * make sure to issue those last loads/stores before enabling
-	 * the pagefault handler again.
-	 */
-	barrier();
-	dec_preempt_count();
-	/*
-	 * make sure we do..
-	 */
-	barrier();
-	preempt_check_resched();
-}
-
-#endif
-
-/* vm ops ->fault() was introduced in 2.6.23. */
-#include <linux/mm.h>
-
-#ifdef KVM_MAIN
-#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,23)
-
-struct vm_fault {
-	unsigned int flags;
-	pgoff_t pgoff;
-	void __user *virtual_address;
-	struct page *page;
-};
-
-static int kvm_vcpu_fault(struct vm_area_struct *vma, struct vm_fault *vmf);
-static int kvm_vm_fault(struct vm_area_struct *vma, struct vm_fault *vmf);
-
-static inline struct page *kvm_nopage_to_fault(
-	int (*fault)(struct vm_area_struct *vma, struct vm_fault *vmf),
-	struct vm_area_struct *vma,
-	unsigned long address,
-	int *type)
-{
-	struct vm_fault vmf;
-	int ret;
-
-	vmf.pgoff = ((address - vma->vm_start) >> PAGE_SHIFT) + vma->vm_pgoff;
-	vmf.virtual_address = (void __user *)address;
-	ret = fault(vma, &vmf);
-	if (ret)
-		return NOPAGE_SIGBUS;
-	*type = VM_FAULT_MINOR;
-	return vmf.page;
-}
-
-static inline struct page *__kvm_vcpu_fault(struct vm_area_struct *vma,
-					    unsigned long address,
-					    int *type)
-{
-	return kvm_nopage_to_fault(kvm_vcpu_fault, vma, address, type);
-}
-
-static inline struct page *__kvm_vm_fault(struct vm_area_struct *vma,
-					  unsigned long address,
-					  int *type)
-{
-	return kvm_nopage_to_fault(kvm_vm_fault, vma, address, type);
-}
-
-#define VMA_OPS_FAULT(x) nopage
-#define VMA_OPS_FAULT_FUNC(x) __##x
-
-#else
-
-#define VMA_OPS_FAULT(x) x
-#define VMA_OPS_FAULT_FUNC(x) x
-
-#endif
-#endif
-
-/* simple vfs attribute getter signature has changed to add a return code */
-
-#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,25)
-
-#define MAKE_SIMPLE_ATTRIBUTE_GETTER(x)       \
-	static u64 x(void *v)                 \
-	{				      \
-		u64 ret = 0;		      \
-					      \
-		__##x(v, &ret);		      \
-		return ret;		      \
-	}
-
-#else
-
-#define MAKE_SIMPLE_ATTRIBUTE_GETTER(x)       \
-	static int x(void *v, u64 *val)	      \
-	{				      \
-		return __##x(v, val);	      \
-	}
-
-#endif
-
-/* set_kset_name() is gone in 2.6.25 */
-
-#if LINUX_VERSION_CODE >= KERNEL_VERSION(2,6,25)
-
-#define set_kset_name(x) .name = x
-
-#endif
-
-#if LINUX_VERSION_CODE >= KERNEL_VERSION(2,6,25)
-#ifndef FASTCALL
-#define FASTCALL(x)	x
-#define fastcall
-#endif
-#endif
-
-#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,23)
-
-static unsigned  __attribute__((__used__)) kvm_tsc_khz = 2000000;
-
-#else
-
-#define kvm_tsc_khz tsc_khz
-
-#endif
-
-struct kvm_desc_struct {
-	union {
-		struct { unsigned int a, b; };
-		struct {
-			u16 limit0;
-			u16 base0;
-			unsigned base1: 8, type: 4, s: 1, dpl: 2, p: 1;
-			unsigned limit: 4, avl: 1, l: 1, d: 1, g: 1, base2: 8;
-		};
-
-	};
-} __attribute__((packed));
-
-struct kvm_ldttss_desc64 {
-	u16 limit0;
-	u16 base0;
-	unsigned base1 : 8, type : 5, dpl : 2, p : 1;
-	unsigned limit1 : 4, zero0 : 3, g : 1, base2 : 8;
-	u32 base3;
-	u32 zero1;
-} __attribute__((packed));
-
-struct kvm_desc_ptr {
-	unsigned short size;
-	unsigned long address;
-} __attribute__((packed));
-
-#if LINUX_VERSION_CODE <= KERNEL_VERSION(2,6,21)
-
-#include <linux/ktime.h>
-#include <linux/hrtimer.h>
-
-#define ktime_get kvm_ktime_get
-
-static inline ktime_t ktime_get(void)
-{
-	struct timespec now;
-
-	ktime_get_ts(&now);
-
-	return timespec_to_ktime(now);
-}
-
-#endif
-
-/* __aligned arrived in 2.6.21 */
-#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,21)
-#define __aligned(x) __attribute__((__aligned__(x)))
-#endif
-
-#include <asm/msr.h>
-#ifndef MSR_FS_BASE
-#define MSR_FS_BASE 0xc0000100
-#endif
-#ifndef MSR_GS_BASE
-#define MSR_GS_BASE 0xc0000101
-#endif
-
-#include <linux/mm.h>
-
-/* The shrinker API changed in 2.6.23 */
-#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,23)
-
-struct kvm_shrinker {
-	int (*shrink)(int nr_to_scan, gfp_t gfp_mask);
-	int seeks;
-	struct shrinker *kshrinker;
-};
-
-static inline void register_shrinker(struct kvm_shrinker *shrinker)
-{
-	shrinker->kshrinker = set_shrinker(shrinker->seeks, shrinker->shrink);
-}
-
-static inline void unregister_shrinker(struct kvm_shrinker *shrinker)
-{
-	if (shrinker->kshrinker)
-		remove_shrinker(shrinker->kshrinker);
-}
-
-#define shrinker kvm_shrinker
-
-#endif
-
-/* undefine lapic */
-#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,18)
-
-#undef lapic
-
-#endif
-
-/* clocksource */
-#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,18)
-static inline u32 clocksource_khz2mult(u32 khz, u32 shift_constant)
-{
-	/*  khz = cyc/(Million ns)
-	 *  mult/2^shift  = ns/cyc
-	 *  mult = ns/cyc * 2^shift
-	 *  mult = 1Million/khz * 2^shift
-	 *  mult = 1000000 * 2^shift / khz
-	 *  mult = (1000000<<shift) / khz
-	 */
-	u64 tmp = ((u64)1000000) << shift_constant;
-
-	tmp += khz/2; /* round for do_div */
-	do_div(tmp, khz);
-
-	return (u32)tmp;
-}
-#else
-#include <linux/clocksource.h>
-#endif
-
-/* manually export hrtimer_init/start/cancel */
-#include <linux/kallsyms.h>
-extern void (*hrtimer_init_p)(struct hrtimer *timer, clockid_t which_clock,
-			      enum hrtimer_mode mode);
-extern int (*hrtimer_start_p)(struct hrtimer *timer, ktime_t tim,
-			      const enum hrtimer_mode mode);
-extern int (*hrtimer_cancel_p)(struct hrtimer *timer);
-
-#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,17) && defined(CONFIG_KALLSYMS)
-static inline void hrtimer_kallsyms_resolve(void)
-{
-	hrtimer_init_p = (void *) kallsyms_lookup_name("hrtimer_init");
-	BUG_ON(!hrtimer_init_p);
-	hrtimer_start_p = (void *) kallsyms_lookup_name("hrtimer_start");
-	BUG_ON(!hrtimer_start_p);
-	hrtimer_cancel_p = (void *) kallsyms_lookup_name("hrtimer_cancel");
-	BUG_ON(!hrtimer_cancel_p);
-}
-#else
-static inline void hrtimer_kallsyms_resolve(void)
-{
-	hrtimer_init_p = hrtimer_init;
-	hrtimer_start_p = hrtimer_start;
-	hrtimer_cancel_p = hrtimer_cancel;
-}
-#endif
-
-/* handle old hrtimer API with data pointer */
-#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,17)
-static inline void hrtimer_data_pointer(struct hrtimer *timer)
-{
-	timer->data = (void *)timer;
-}
-#else
-static inline void hrtimer_data_pointer(struct hrtimer *timer) {}
-#endif
-
-#include <asm/hw_irq.h>
-#ifndef NMI_VECTOR
-#define NMI_VECTOR 2
-#endif
-
-#ifndef MSR_MTRRcap
-#define MSR_MTRRcap            0x0fe
-#define MSR_MTRRfix64K_00000   0x250
-#define MSR_MTRRfix16K_80000   0x258
-#define MSR_MTRRfix16K_A0000   0x259
-#define MSR_MTRRfix4K_C0000    0x268
-#define MSR_MTRRfix4K_C8000    0x269
-#define MSR_MTRRfix4K_D0000    0x26a
-#define MSR_MTRRfix4K_D8000    0x26b
-#define MSR_MTRRfix4K_E0000    0x26c
-#define MSR_MTRRfix4K_E8000    0x26d
-#define MSR_MTRRfix4K_F0000    0x26e
-#define MSR_MTRRfix4K_F8000    0x26f
-#define MSR_MTRRdefType        0x2ff
-#endif
-
-#ifndef MSR_IA32_CR_PAT
-#define MSR_IA32_CR_PAT        0x00000277
-#endif
-
-#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,22)
-
-#define ns_to_timespec kvm_ns_to_timespec
-
-struct timespec kvm_ns_to_timespec(const s64 nsec);
-
-#endif
-
-/* Define DEBUGCTLMSR bits */
-#ifndef DEBUGCTLMSR_LBR
-
-#define _DEBUGCTLMSR_LBR	0 /* last branch recording */
-#define _DEBUGCTLMSR_BTF	1 /* single-step on branches */
-
-#define DEBUGCTLMSR_LBR		(1UL << _DEBUGCTLMSR_LBR)
-#define DEBUGCTLMSR_BTF		(1UL << _DEBUGCTLMSR_BTF)
-
-#endif
diff --git a/kernel/debug.h b/kernel/x86/debug.h
similarity index 100%
rename from kernel/debug.h
rename to kernel/x86/debug.h
diff --git a/kernel/external-module-compat.c b/kernel/x86/external-module-compat.c
similarity index 100%
rename from kernel/external-module-compat.c
rename to kernel/x86/external-module-compat.c
diff --git a/kernel/x86/external-module-compat.h b/kernel/x86/external-module-compat.h
new file mode 100644
index 0000000..ce93551
--- /dev/null
+++ b/kernel/x86/external-module-compat.h
@@ -0,0 +1,291 @@
+
+/*
+ * Compatibility header for building as an external module.
+ */
+
+#include <linux/compiler.h>
+#include <linux/version.h>
+
+#include "../external-module-compat-comm.h"
+
+#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,22)
+
+#ifndef _EFER_SCE
+#define _EFER_SCE		0  /* SYSCALL/SYSRET */
+#endif
+
+#ifndef EFER_SCE
+#define EFER_SCE		(1<<_EFER_SCE)
+#endif
+
+#endif
+
+#include <linux/smp.h>
+
+#ifndef X86_CR0_PE
+#define X86_CR0_PE 0x00000001
+#endif
+
+#ifndef X86_CR0_MP
+#define X86_CR0_MP 0x00000002
+#endif
+
+#ifndef X86_CR0_EM
+#define X86_CR0_EM 0x00000004
+#endif
+
+#ifndef X86_CR0_TS
+#define X86_CR0_TS 0x00000008
+#endif
+
+#ifndef X86_CR0_ET
+#define X86_CR0_ET 0x00000010
+#endif
+
+#ifndef X86_CR0_NE
+#define X86_CR0_NE 0x00000020
+#endif
+
+#ifndef X86_CR0_WP
+#define X86_CR0_WP 0x00010000
+#endif
+
+#ifndef X86_CR0_AM
+#define X86_CR0_AM 0x00040000
+#endif
+
+#ifndef X86_CR0_NW
+#define X86_CR0_NW 0x20000000
+#endif
+
+#ifndef X86_CR0_CD
+#define X86_CR0_CD 0x40000000
+#endif
+
+#ifndef X86_CR0_PG
+#define X86_CR0_PG 0x80000000
+#endif
+
+#ifndef X86_CR3_PWT
+#define X86_CR3_PWT 0x00000008
+#endif
+
+#ifndef X86_CR3_PCD
+#define X86_CR3_PCD 0x00000010
+#endif
+
+#ifndef X86_CR4_VMXE
+#define X86_CR4_VMXE 0x00002000
+#endif
+
+#undef X86_CR8_TPR
+#define X86_CR8_TPR 0x0f
+
+/*
+ * 2.6.22 does not define set_64bit() under nonpae
+ */
+#ifdef CONFIG_X86_32
+
+#include <asm/cmpxchg.h>
+
+static inline void __kvm_set_64bit(u64 *ptr, u64 val)
+{
+	unsigned int low = val;
+	unsigned int high = val >> 32;
+
+	__asm__ __volatile__ (
+		"\n1:\t"
+		"movl (%0), %%eax\n\t"
+		"movl 4(%0), %%edx\n\t"
+		"lock cmpxchg8b (%0)\n\t"
+		"jnz 1b"
+		: /* no outputs */
+		:	"D"(ptr),
+			"b"(low),
+			"c"(high)
+		:	"ax","dx","memory");
+}
+
+#undef  set_64bit
+#define set_64bit __kvm_set_64bit
+
+static inline unsigned long long __kvm_cmpxchg64(volatile void *ptr,
+						 unsigned long long old,
+						 unsigned long long new)
+{
+	unsigned long long prev;
+	__asm__ __volatile__("lock cmpxchg8b %3"
+			     : "=A"(prev)
+			     : "b"((unsigned long)new),
+			       "c"((unsigned long)(new >> 32)),
+			       "m"(*__xg(ptr)),
+			       "0"(old)
+			     : "memory");
+	return prev;
+}
+
+#define kvm_cmpxchg64(ptr,o,n)\
+	((__typeof__(*(ptr)))__kvm_cmpxchg64((ptr),(unsigned long long)(o),\
+					(unsigned long long)(n)))
+
+#undef cmpxchg64
+#define cmpxchg64(ptr, o, n) kvm_cmpxchg64(ptr, o, n)
+
+#endif
+
+#ifndef CONFIG_PREEMPT_NOTIFIERS
+/*
+ * Include sched|preempt.h before defining CONFIG_PREEMPT_NOTIFIERS to avoid
+ * a miscompile.
+ */
+#include <linux/sched.h>
+#include <linux/preempt.h>
+#define CONFIG_PREEMPT_NOTIFIERS
+#define CONFIG_PREEMPT_NOTIFIERS_COMPAT
+
+struct preempt_notifier;
+
+struct preempt_ops {
+	void (*sched_in)(struct preempt_notifier *notifier, int cpu);
+	void (*sched_out)(struct preempt_notifier *notifier,
+			  struct task_struct *next);
+};
+
+struct preempt_notifier {
+	struct list_head link;
+	struct task_struct *tsk;
+	struct preempt_ops *ops;
+};
+
+void preempt_notifier_register(struct preempt_notifier *notifier);
+void preempt_notifier_unregister(struct preempt_notifier *notifier);
+
+static inline void preempt_notifier_init(struct preempt_notifier *notifier,
+				     struct preempt_ops *ops)
+{
+	notifier->ops = ops;
+}
+
+void start_special_insn(void);
+void end_special_insn(void);
+void in_special_section(void);
+void special_reload_dr7(void);
+
+void preempt_notifier_sys_init(void);
+void preempt_notifier_sys_exit(void);
+
+#else
+
+static inline void start_special_insn(void) {}
+static inline void end_special_insn(void) {}
+static inline void in_special_section(void) {}
+static inline void special_reload_dr7(void) {}
+
+static inline void preempt_notifier_sys_init(void) {}
+static inline void preempt_notifier_sys_exit(void) {}
+
+#endif
+
+/* CONFIG_HAS_IOMEM is apparently fairly new too (2.6.21 for x86_64). */
+#ifndef CONFIG_HAS_IOMEM
+#define CONFIG_HAS_IOMEM 1
+#endif
+
+/* X86_FEATURE_NX is missing in some x86_64 kernels */
+
+#include <asm/cpufeature.h>
+
+#ifndef X86_FEATURE_NX
+#define X86_FEATURE_NX (1*32+20)
+#endif
+
+#undef true
+#define true 1
+#undef false
+#define false 0
+
+/* EFER_LMA and EFER_LME are missing in pre 2.6.24 i386 kernels */
+#ifndef EFER_LME
+#define _EFER_LME           8  /* Long mode enable */
+#define _EFER_LMA           10 /* Long mode active (read-only) */
+#define EFER_LME            (1<<_EFER_LME)
+#define EFER_LMA            (1<<_EFER_LMA)
+#endif
+
+struct kvm_desc_struct {
+	union {
+		struct { unsigned int a, b; };
+		struct {
+			u16 limit0;
+			u16 base0;
+			unsigned base1: 8, type: 4, s: 1, dpl: 2, p: 1;
+			unsigned limit: 4, avl: 1, l: 1, d: 1, g: 1, base2: 8;
+		};
+
+	};
+} __attribute__((packed));
+
+struct kvm_ldttss_desc64 {
+	u16 limit0;
+	u16 base0;
+	unsigned base1 : 8, type : 5, dpl : 2, p : 1;
+	unsigned limit1 : 4, zero0 : 3, g : 1, base2 : 8;
+	u32 base3;
+	u32 zero1;
+} __attribute__((packed));
+
+struct kvm_desc_ptr {
+	unsigned short size;
+	unsigned long address;
+} __attribute__((packed));
+
+#include <asm/msr.h>
+#ifndef MSR_FS_BASE
+#define MSR_FS_BASE 0xc0000100
+#endif
+#ifndef MSR_GS_BASE
+#define MSR_GS_BASE 0xc0000101
+#endif
+
+/* undefine lapic */
+#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,18)
+
+#undef lapic
+
+#endif
+
+#include <asm/hw_irq.h>
+#ifndef NMI_VECTOR
+#define NMI_VECTOR 2
+#endif
+
+#ifndef MSR_MTRRcap
+#define MSR_MTRRcap            0x0fe
+#define MSR_MTRRfix64K_00000   0x250
+#define MSR_MTRRfix16K_80000   0x258
+#define MSR_MTRRfix16K_A0000   0x259
+#define MSR_MTRRfix4K_C0000    0x268
+#define MSR_MTRRfix4K_C8000    0x269
+#define MSR_MTRRfix4K_D0000    0x26a
+#define MSR_MTRRfix4K_D8000    0x26b
+#define MSR_MTRRfix4K_E0000    0x26c
+#define MSR_MTRRfix4K_E8000    0x26d
+#define MSR_MTRRfix4K_F0000    0x26e
+#define MSR_MTRRfix4K_F8000    0x26f
+#define MSR_MTRRdefType        0x2ff
+#endif
+
+#ifndef MSR_IA32_CR_PAT
+#define MSR_IA32_CR_PAT        0x00000277
+#endif
+
+/* Define DEBUGCTLMSR bits */
+#ifndef DEBUGCTLMSR_LBR
+
+#define _DEBUGCTLMSR_LBR	0 /* last branch recording */
+#define _DEBUGCTLMSR_BTF	1 /* single-step on branches */
+
+#define DEBUGCTLMSR_LBR		(1UL << _DEBUGCTLMSR_LBR)
+#define DEBUGCTLMSR_BTF		(1UL << _DEBUGCTLMSR_BTF)
+
+#endif
diff --git a/kernel/preempt.c b/kernel/x86/preempt.c
similarity index 100%
rename from kernel/preempt.c
rename to kernel/x86/preempt.c
diff --git a/kernel/vmx-debug.c b/kernel/x86/vmx-debug.c
similarity index 100%
rename from kernel/vmx-debug.c
rename to kernel/x86/vmx-debug.c
-- 
1.5.6.2


^ permalink raw reply related	[flat|nested] 24+ messages in thread

* Re: [PATCH 01/04]Create x86 directory to hold x86-specific files.
  2008-07-27  9:46   ` Zhang, Xiantao
@ 2008-07-30 14:04   ` Avi Kivity
  -1 siblings, 0 replies; 24+ messages in thread
From: Avi Kivity @ 2008-07-30 14:04 UTC (permalink / raw)
  To: kvm-ia64

Zhang, Xiantao wrote:
> Avi Kivity wrote:
>   
>> Zhang, Xiantao wrote:
>>     
>>> From c8c8c5a6b3ff943cb645f9ac4ef169490f14cd08 Mon Sep 17 00:00:00
>>> 2001 From: Zhang Xiantao <xiantao.zhang@intel.com>
>>> Date: Fri, 25 Jul 2008 16:50:30 +0800
>>> Subject: [PATCH] KVM: external module: Moving x86-speicif files to
>>> x86 directory. 
>>>
>>> Create x86 directory to hold x86-specific files.
>>> Signed-off-by: Zhang Xiantao <xiantao.zhang@intel.com>
>>>
>>>       
>> You forgot to attach a patch; and your mailer corrupts patches, so I
>> can't apply this.
>>     
>
> Sorry. Attached! :)
>   

Applied all, but I had to revert the change to the install command in 
the last patch.  My install doesn't like the -C option.  Please check if 
further changes are necessary.

-- 
error compiling committee.c: too many arguments to function


^ permalink raw reply	[flat|nested] 24+ messages in thread

* Re: [PATCH 01/04]Create x86 directory to hold x86-specific files.
@ 2008-07-30 14:04   ` Avi Kivity
  0 siblings, 0 replies; 24+ messages in thread
From: Avi Kivity @ 2008-07-30 14:04 UTC (permalink / raw)
  To: Zhang, Xiantao; +Cc: kvm-ia64, kvm

Zhang, Xiantao wrote:
> Avi Kivity wrote:
>   
>> Zhang, Xiantao wrote:
>>     
>>> From c8c8c5a6b3ff943cb645f9ac4ef169490f14cd08 Mon Sep 17 00:00:00
>>> 2001 From: Zhang Xiantao <xiantao.zhang@intel.com>
>>> Date: Fri, 25 Jul 2008 16:50:30 +0800
>>> Subject: [PATCH] KVM: external module: Moving x86-speicif files to
>>> x86 directory. 
>>>
>>> Create x86 directory to hold x86-specific files.
>>> Signed-off-by: Zhang Xiantao <xiantao.zhang@intel.com>
>>>
>>>       
>> You forgot to attach a patch; and your mailer corrupts patches, so I
>> can't apply this.
>>     
>
> Sorry. Attached! :)
>   

Applied all, but I had to revert the change to the install command in 
the last patch.  My install doesn't like the -C option.  Please check if 
further changes are necessary.

-- 
error compiling committee.c: too many arguments to function


^ permalink raw reply	[flat|nested] 24+ messages in thread

* Re: [PATCH 01/04]Create x86 directory to hold x86-specific files.
  2008-07-27  9:46   ` Zhang, Xiantao
@ 2008-07-31  3:25   ` Yang, Sheng
  -1 siblings, 0 replies; 24+ messages in thread
From: Yang, Sheng @ 2008-07-31  3:25 UTC (permalink / raw)
  To: kvm-ia64

On Sunday 27 July 2008 17:46:26 Zhang, Xiantao wrote:
> Avi Kivity wrote:
> > Zhang, Xiantao wrote:
> >> From c8c8c5a6b3ff943cb645f9ac4ef169490f14cd08 Mon Sep 17
> >> 00:00:00 2001 From: Zhang Xiantao <xiantao.zhang@intel.com>
> >> Date: Fri, 25 Jul 2008 16:50:30 +0800
> >> Subject: [PATCH] KVM: external module: Moving x86-speicif files
> >> to x86 directory.
> >>
> >> Create x86 directory to hold x86-specific files.
> >> Signed-off-by: Zhang Xiantao <xiantao.zhang@intel.com>
> >
> > You forgot to attach a patch; and your mailer corrupts patches,
> > so I can't apply this.
>
> Sorry. Attached! :)
>

This broke kvm script in kvm-userspace. We need a update of it...

-- 
regards
Yang, Sheng


^ permalink raw reply	[flat|nested] 24+ messages in thread

* Re: [PATCH 01/04]Create x86 directory to hold x86-specific files.
@ 2008-07-31  3:25   ` Yang, Sheng
  0 siblings, 0 replies; 24+ messages in thread
From: Yang, Sheng @ 2008-07-31  3:25 UTC (permalink / raw)
  To: kvm; +Cc: Zhang, Xiantao, Avi Kivity, kvm-ia64

On Sunday 27 July 2008 17:46:26 Zhang, Xiantao wrote:
> Avi Kivity wrote:
> > Zhang, Xiantao wrote:
> >> From c8c8c5a6b3ff943cb645f9ac4ef169490f14cd08 Mon Sep 17
> >> 00:00:00 2001 From: Zhang Xiantao <xiantao.zhang@intel.com>
> >> Date: Fri, 25 Jul 2008 16:50:30 +0800
> >> Subject: [PATCH] KVM: external module: Moving x86-speicif files
> >> to x86 directory.
> >>
> >> Create x86 directory to hold x86-specific files.
> >> Signed-off-by: Zhang Xiantao <xiantao.zhang@intel.com>
> >
> > You forgot to attach a patch; and your mailer corrupts patches,
> > so I can't apply this.
>
> Sorry. Attached! :)
>

This broke kvm script in kvm-userspace. We need a update of it...

-- 
regards
Yang, Sheng


^ permalink raw reply	[flat|nested] 24+ messages in thread

* RE: [PATCH 01/04]Create x86 directory to hold x86-specific files.
  2008-07-31  3:25   ` Yang, Sheng
@ 2008-07-31  3:53   ` Zhang, Xiantao
  -1 siblings, 0 replies; 24+ messages in thread
From: Zhang, Xiantao @ 2008-07-31  3:53 UTC (permalink / raw)
  To: kvm-ia64

Yang, Sheng wrote:
> On Sunday 27 July 2008 17:46:26 Zhang, Xiantao wrote:
>> Avi Kivity wrote:
>>> Zhang, Xiantao wrote:
>>>> From c8c8c5a6b3ff943cb645f9ac4ef169490f14cd08 Mon Sep 17
>>>> 00:00:00 2001 From: Zhang Xiantao <xiantao.zhang@intel.com>
>>>> Date: Fri, 25 Jul 2008 16:50:30 +0800
>>>> Subject: [PATCH] KVM: external module: Moving x86-speicif files to
>>>> x86 directory. 
>>>> 
>>>> Create x86 directory to hold x86-specific files.
>>>> Signed-off-by: Zhang Xiantao <xiantao.zhang@intel.com>
>>> 
>>> You forgot to attach a patch; and your mailer corrupts patches,
>>> so I can't apply this.
>> 
>> Sorry. Attached! :)
>> 
> 
> This broke kvm script in kvm-userspace. We need a update of it...

Why break kvm scripts ?
xiantao

^ permalink raw reply	[flat|nested] 24+ messages in thread

* RE: [PATCH 01/04]Create x86 directory to hold x86-specific files.
@ 2008-07-31  3:53   ` Zhang, Xiantao
  0 siblings, 0 replies; 24+ messages in thread
From: Zhang, Xiantao @ 2008-07-31  3:53 UTC (permalink / raw)
  To: Yang, Sheng, kvm; +Cc: Avi Kivity, kvm-ia64

Yang, Sheng wrote:
> On Sunday 27 July 2008 17:46:26 Zhang, Xiantao wrote:
>> Avi Kivity wrote:
>>> Zhang, Xiantao wrote:
>>>> From c8c8c5a6b3ff943cb645f9ac4ef169490f14cd08 Mon Sep 17
>>>> 00:00:00 2001 From: Zhang Xiantao <xiantao.zhang@intel.com>
>>>> Date: Fri, 25 Jul 2008 16:50:30 +0800
>>>> Subject: [PATCH] KVM: external module: Moving x86-speicif files to
>>>> x86 directory. 
>>>> 
>>>> Create x86 directory to hold x86-specific files.
>>>> Signed-off-by: Zhang Xiantao <xiantao.zhang@intel.com>
>>> 
>>> You forgot to attach a patch; and your mailer corrupts patches,
>>> so I can't apply this.
>> 
>> Sorry. Attached! :)
>> 
> 
> This broke kvm script in kvm-userspace. We need a update of it...

Why break kvm scripts ?
xiantao

^ permalink raw reply	[flat|nested] 24+ messages in thread

* RE: [PATCH 01/04]Create x86 directory to hold x86-specific files.
  2008-07-30 14:04   ` Avi Kivity
@ 2008-08-01  2:48   ` Zhang, Xiantao
  -1 siblings, 0 replies; 24+ messages in thread
From: Zhang, Xiantao @ 2008-08-01  2:48 UTC (permalink / raw)
  To: kvm-ia64

Avi Kivity wrote:
> Zhang, Xiantao wrote:
>> Avi Kivity wrote:
>> 
>>> Zhang, Xiantao wrote:
>>> 
>>>> From c8c8c5a6b3ff943cb645f9ac4ef169490f14cd08 Mon Sep 17 00:00:00
>>>> 2001 From: Zhang Xiantao <xiantao.zhang@intel.com>
>>>> Date: Fri, 25 Jul 2008 16:50:30 +0800
>>>> Subject: [PATCH] KVM: external module: Moving x86-speicif files to
>>>> x86 directory. 
>>>> 
>>>> Create x86 directory to hold x86-specific files.
>>>> Signed-off-by: Zhang Xiantao <xiantao.zhang@intel.com>
>>>> 
>>>> 
>>> You forgot to attach a patch; and your mailer corrupts patches, so
>>> I can't apply this. 
>>> 
>> 
>> Sorry. Attached! :)
>> 
> 
> Applied all, but I had to revert the change to the install command in
> the last patch.  My install doesn't like the -C option.  Please check
> if further changes are necessary.
Hi, Avi
    -t should be a new argument for installer, and doesn't exist in many
older Linux systems, for example, my RHEL4.3 can't recognize it at all.
How to make it work well in these systems ? 
Xiantao

^ permalink raw reply	[flat|nested] 24+ messages in thread

* RE: [PATCH 01/04]Create x86 directory to hold x86-specific files.
@ 2008-08-01  2:48   ` Zhang, Xiantao
  0 siblings, 0 replies; 24+ messages in thread
From: Zhang, Xiantao @ 2008-08-01  2:48 UTC (permalink / raw)
  To: Avi Kivity; +Cc: kvm-ia64, kvm

Avi Kivity wrote:
> Zhang, Xiantao wrote:
>> Avi Kivity wrote:
>> 
>>> Zhang, Xiantao wrote:
>>> 
>>>> From c8c8c5a6b3ff943cb645f9ac4ef169490f14cd08 Mon Sep 17 00:00:00
>>>> 2001 From: Zhang Xiantao <xiantao.zhang@intel.com>
>>>> Date: Fri, 25 Jul 2008 16:50:30 +0800
>>>> Subject: [PATCH] KVM: external module: Moving x86-speicif files to
>>>> x86 directory. 
>>>> 
>>>> Create x86 directory to hold x86-specific files.
>>>> Signed-off-by: Zhang Xiantao <xiantao.zhang@intel.com>
>>>> 
>>>> 
>>> You forgot to attach a patch; and your mailer corrupts patches, so
>>> I can't apply this. 
>>> 
>> 
>> Sorry. Attached! :)
>> 
> 
> Applied all, but I had to revert the change to the install command in
> the last patch.  My install doesn't like the -C option.  Please check
> if further changes are necessary.
Hi, Avi
    -t should be a new argument for installer, and doesn't exist in many
older Linux systems, for example, my RHEL4.3 can't recognize it at all.
How to make it work well in these systems ? 
Xiantao

^ permalink raw reply	[flat|nested] 24+ messages in thread

end of thread, other threads:[~2008-08-01  2:48 UTC | newest]

Thread overview: 24+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2008-07-11  5:43 [PATCH 01/04]Create x86 directory to hold x86-specific files Zhang, Xiantao
2008-07-11  5:43 ` Zhang, Xiantao
2008-07-13  8:53 ` Avi Kivity
2008-07-13  8:53   ` Avi Kivity
2008-07-16  9:11 ` Zhang, Xiantao
2008-07-16  9:11   ` Zhang, Xiantao
2008-07-25  9:26 ` Zhang, Xiantao
2008-07-25  9:26   ` Zhang, Xiantao
2008-07-27  7:45 ` Avi Kivity
2008-07-27  7:45   ` Avi Kivity
2008-07-27  9:46 ` Zhang, Xiantao
2008-07-27  9:46   ` Zhang, Xiantao
2008-07-29 13:14 ` Avi Kivity
2008-07-29 13:14   ` Avi Kivity
2008-07-30  1:42 ` Zhang, Xiantao
2008-07-30  1:42   ` Zhang, Xiantao
2008-07-30 14:04 ` Avi Kivity
2008-07-30 14:04   ` Avi Kivity
2008-07-31  3:25 ` Yang, Sheng
2008-07-31  3:25   ` Yang, Sheng
2008-07-31  3:53 ` Zhang, Xiantao
2008-07-31  3:53   ` Zhang, Xiantao
2008-08-01  2:48 ` Zhang, Xiantao
2008-08-01  2:48   ` Zhang, Xiantao

This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.