[PATCH 00 of 24] More Xen updates

[PATCH 00 of 24] More Xen updates

[Jeremy Fitzhardinge]

Hi Ingo,

More patches.  There are three groups:

- some code movements to make common code available to other
  architectures (namely, ia64)
- Xen device driver updates, including a new xen paravirt framebuffer driver,
- The Xen balloon driver (shrink only, so no dependency on memory hotplug yet)
- some "using smp_processor_id while preemptable" warning fixes,
  mostly as a result of the core kernel's tendency to do tlb flushes
  and update init_mm while preemptable,
- other little Xen cleanups

Thanks,
	J

[PATCH 01 of 24] xen: add missing __HYPERVISOR_arch_[0-7] definisions which ia64 needs

[Jeremy Fitzhardinge]

From: Isaku Yamahata <yamahata@valinux.co.jp>

Add xen hypercall numbers defined for arch specific use.
ia64/xen domU uses __HYPERVISOR_arch_1 to manipulate paravirtualized
IOSAPIC. Although all those constants aren't used yet by IA64 at this
moment, add all arch specific hypercall numbers.

Signed-off-by: Isaku Yamahata <yamahata@valinux.co.jp>
Signed-off-by: Jeremy Fitzhardinge <jeremy.fitzhardinge@citrix.com>
---
 include/xen/interface/xen.h |   11 ++++++++++-
 1 file changed, 10 insertions(+), 1 deletion(-)

diff --git a/include/xen/interface/xen.h b/include/xen/interface/xen.h
--- a/include/xen/interface/xen.h
+++ b/include/xen/interface/xen.h
@@ -57,6 +57,16 @@
 #define __HYPERVISOR_event_channel_op     32
 #define __HYPERVISOR_physdev_op           33
 #define __HYPERVISOR_hvm_op               34
+
+/* Architecture-specific hypercall definitions. */
+#define __HYPERVISOR_arch_0               48
+#define __HYPERVISOR_arch_1               49
+#define __HYPERVISOR_arch_2               50
+#define __HYPERVISOR_arch_3               51
+#define __HYPERVISOR_arch_4               52
+#define __HYPERVISOR_arch_5               53
+#define __HYPERVISOR_arch_6               54
+#define __HYPERVISOR_arch_7               55
 
 /*
  * VIRTUAL INTERRUPTS

[PATCH 02 of 24] xen: add missing VIRQ_ARCH_[0-7] definitions which ia64/xen needs

[Jeremy Fitzhardinge]

From: Isaku Yamahata <yamahata@valinux.co.jp>

Add xen VIRQ numbers defined for arch specific use.
ia64/xen domU uses VIRQ_ARCH_0 for virtual itc timer.
Although all those constants aren't used yet by ia64
at this moment, add all arch specific VIRQ numbers.

Signed-off-by: Isaku Yamahata <yamahata@valinux.co.jp>
Signed-off-by: Jeremy Fitzhardinge <jeremy.fitzhardinge@citrix.com>
---
 include/xen/interface/xen.h |   13 +++++++++++--
 1 file changed, 11 insertions(+), 2 deletions(-)

diff --git a/include/xen/interface/xen.h b/include/xen/interface/xen.h
--- a/include/xen/interface/xen.h
+++ b/include/xen/interface/xen.h
@@ -78,8 +78,18 @@
 #define VIRQ_CONSOLE    2  /* (DOM0) Bytes received on emergency console. */
 #define VIRQ_DOM_EXC    3  /* (DOM0) Exceptional event for some domain.   */
 #define VIRQ_DEBUGGER   6  /* (DOM0) A domain has paused for debugging.   */
-#define NR_VIRQS        8
 
+/* Architecture-specific VIRQ definitions. */
+#define VIRQ_ARCH_0    16
+#define VIRQ_ARCH_1    17
+#define VIRQ_ARCH_2    18
+#define VIRQ_ARCH_3    19
+#define VIRQ_ARCH_4    20
+#define VIRQ_ARCH_5    21
+#define VIRQ_ARCH_6    22
+#define VIRQ_ARCH_7    23
+
+#define NR_VIRQS       24
 /*
  * MMU-UPDATE REQUESTS
  *

[PATCH 03 of 24] xen: add missing definitions for xen grant table which ia64/xen needs

[Jeremy Fitzhardinge]

From: Isaku Yamahata <yamahata@valinux.co.jp>

Add xen handles realted definitions for grant table which ia64/xen
needs.
Pointer argumsnts for ia64/xen hypercall are passed in pseudo physical
address (guest physical address) so that it is required to convert
guest kernel virtual address into pseudo physical address right before
issuing hypercall.
The xen guest handle represents such arguments.
Define necessary handles and helper functions.

Signed-off-by: Isaku Yamahata <yamahata@valinux.co.jp>
Signed-off-by: Jeremy Fitzhardinge <jeremy.fitzhardinge@citrix.com>
---
 drivers/xen/grant-table.c           |    2 +-
 include/asm-x86/xen/interface.h     |   24 ++++++++++++++++++++++++
 include/xen/interface/grant_table.h |   12 ++++++++----
 3 files changed, 33 insertions(+), 5 deletions(-)

diff --git a/drivers/xen/grant-table.c b/drivers/xen/grant-table.c
--- a/drivers/xen/grant-table.c
+++ b/drivers/xen/grant-table.c
@@ -466,7 +466,7 @@
 
 	setup.dom        = DOMID_SELF;
 	setup.nr_frames  = nr_gframes;
-	setup.frame_list = frames;
+	set_xen_guest_handle(setup.frame_list, frames);
 
 	rc = HYPERVISOR_grant_table_op(GNTTABOP_setup_table, &setup, 1);
 	if (rc == -ENOSYS) {
diff --git a/include/asm-x86/xen/interface.h b/include/asm-x86/xen/interface.h
--- a/include/asm-x86/xen/interface.h
+++ b/include/asm-x86/xen/interface.h
@@ -21,6 +21,30 @@
 	__DEFINE_GUEST_HANDLE(name, struct name)
 #define DEFINE_GUEST_HANDLE(name) __DEFINE_GUEST_HANDLE(name, name)
 #define GUEST_HANDLE(name)        __guest_handle_ ## name
+
+#ifdef __XEN__
+#if defined(__i386__)
+#define set_xen_guest_handle(hnd, val)			\
+	do {						\
+		if (sizeof(hnd) == 8)			\
+			*(uint64_t *)&(hnd) = 0;	\
+		(hnd).p = val;				\
+	} while (0)
+#elif defined(__x86_64__)
+#define set_xen_guest_handle(hnd, val)	do { (hnd).p = val; } while (0)
+#endif
+#else
+#if defined(__i386__)
+#define set_xen_guest_handle(hnd, val)			\
+	do {						\
+		if (sizeof(hnd) == 8)			\
+			*(uint64_t *)&(hnd) = 0;	\
+		(hnd) = val;				\
+	} while (0)
+#elif defined(__x86_64__)
+#define set_xen_guest_handle(hnd, val)	do { (hnd) = val; } while (0)
+#endif
+#endif
 
 #ifndef __ASSEMBLY__
 /* Guest handles for primitive C types. */
diff --git a/include/xen/interface/grant_table.h b/include/xen/interface/grant_table.h
--- a/include/xen/interface/grant_table.h
+++ b/include/xen/interface/grant_table.h
@@ -185,6 +185,7 @@
     grant_handle_t handle;
     uint64_t dev_bus_addr;
 };
+DEFINE_GUEST_HANDLE_STRUCT(gnttab_map_grant_ref);
 
 /*
  * GNTTABOP_unmap_grant_ref: Destroy one or more grant-reference mappings
@@ -206,6 +207,7 @@
     /* OUT parameters. */
     int16_t  status;              /* GNTST_* */
 };
+DEFINE_GUEST_HANDLE_STRUCT(gnttab_unmap_grant_ref);
 
 /*
  * GNTTABOP_setup_table: Set up a grant table for <dom> comprising at least
@@ -223,8 +225,9 @@
     uint32_t nr_frames;
     /* OUT parameters. */
     int16_t  status;              /* GNTST_* */
-    ulong *frame_list;
+    GUEST_HANDLE(ulong) frame_list;
 };
+DEFINE_GUEST_HANDLE_STRUCT(gnttab_setup_table);
 
 /*
  * GNTTABOP_dump_table: Dump the contents of the grant table to the
@@ -237,6 +240,7 @@
     /* OUT parameters. */
     int16_t status;               /* GNTST_* */
 };
+DEFINE_GUEST_HANDLE_STRUCT(gnttab_dump_table);
 
 /*
  * GNTTABOP_transfer_grant_ref: Transfer <frame> to a foreign domain. The
@@ -255,7 +259,7 @@
     /* OUT parameters. */
     int16_t       status;
 };
-
+DEFINE_GUEST_HANDLE_STRUCT(gnttab_transfer);
 
 /*
  * GNTTABOP_copy: Hypervisor based copy
@@ -296,6 +300,7 @@
 	/* OUT parameters. */
 	int16_t       status;
 };
+DEFINE_GUEST_HANDLE_STRUCT(gnttab_copy);
 
 /*
  * GNTTABOP_query_size: Query the current and maximum sizes of the shared
@@ -313,7 +318,7 @@
     uint32_t max_nr_frames;
     int16_t  status;              /* GNTST_* */
 };
-
+DEFINE_GUEST_HANDLE_STRUCT(gnttab_query_size);
 
 /*
  * Bitfield values for update_pin_status.flags.

[PATCH 04 of 24] xen: add missing definitions in include/xen/interface/vcpu.h which ia64/xen needs

[Jeremy Fitzhardinge]

From: Isaku Yamahata <yamahata@valinux.co.jp>

Add xen handles realted definitions for xen vcpu which ia64/xen needs.
Pointer argumsnts for ia64/xen hypercall are passed in pseudo physical
address (guest physical address) so that it is required to convert
guest kernel virtual address into pseudo physical address.
The xen guest handle represents such arguments.

Signed-off-by: Isaku Yamahata <yamahata@valinux.co.jp>
Signed-off-by: Jeremy Fitzhardinge <jeremy.fitzhardinge@citrix.com>
---
 include/xen/interface/vcpu.h |    6 +++++-
 1 file changed, 5 insertions(+), 1 deletion(-)

diff --git a/include/xen/interface/vcpu.h b/include/xen/interface/vcpu.h
--- a/include/xen/interface/vcpu.h
+++ b/include/xen/interface/vcpu.h
@@ -85,6 +85,7 @@
 		 */
 		uint64_t time[4];
 };
+DEFINE_GUEST_HANDLE_STRUCT(vcpu_runstate_info);
 
 /* VCPU is currently running on a physical CPU. */
 #define RUNSTATE_running  0
@@ -119,6 +120,7 @@
 #define VCPUOP_register_runstate_memory_area 5
 struct vcpu_register_runstate_memory_area {
 		union {
+				GUEST_HANDLE(vcpu_runstate_info) h;
 				struct vcpu_runstate_info *v;
 				uint64_t p;
 		} addr;
@@ -134,6 +136,7 @@
 struct vcpu_set_periodic_timer {
 		uint64_t period_ns;
 };
+DEFINE_GUEST_HANDLE_STRUCT(vcpu_set_periodic_timer);
 
 /*
  * Set or stop a VCPU's single-shot timer. Every VCPU has one single-shot
@@ -145,6 +148,7 @@
 		uint64_t timeout_abs_ns;
 		uint32_t flags;			   /* VCPU_SSHOTTMR_??? */
 };
+DEFINE_GUEST_HANDLE_STRUCT(vcpu_set_singleshot_timer);
 
 /* Flags to VCPUOP_set_singleshot_timer. */
  /* Require the timeout to be in the future (return -ETIME if it's passed). */
@@ -164,5 +168,6 @@
     uint32_t offset; /* offset within page */
     uint32_t rsvd;   /* unused */
 };
+DEFINE_GUEST_HANDLE_STRUCT(vcpu_register_vcpu_info);
 
 #endif /* __XEN_PUBLIC_VCPU_H__ */

[PATCH 05 of 24] xen: move features.c from arch/x86/xen/features.c to drivers/xen

[Jeremy Fitzhardinge]

From: Isaku Yamahata <yamahata@valinux.co.jp>

ia64/xen also uses it too. Move it into common place so that
ia64/xen can share the code.

Signed-off-by: Isaku Yamahata <yamahata@valinux.co.jp>
Signed-off-by: Jeremy Fitzhardinge <jeremy.fitzhardinge@citrix.com>
---
 arch/x86/xen/Makefile  |    2 +-
 drivers/xen/Makefile   |    2 +-
 drivers/xen/features.c |    1 -
 3 files changed, 2 insertions(+), 3 deletions(-)

diff --git a/arch/x86/xen/Makefile b/arch/x86/xen/Makefile
--- a/arch/x86/xen/Makefile
+++ b/arch/x86/xen/Makefile
@@ -1,4 +1,4 @@
-obj-y		:= enlighten.o setup.o features.o multicalls.o mmu.o \
+obj-y		:= enlighten.o setup.o multicalls.o mmu.o \
 			events.o time.o manage.o xen-asm.o
 
 obj-$(CONFIG_SMP)	+= smp.o
diff --git a/drivers/xen/Makefile b/drivers/xen/Makefile
--- a/drivers/xen/Makefile
+++ b/drivers/xen/Makefile
@@ -1,2 +1,2 @@
-obj-y	+= grant-table.o
+obj-y	+= grant-table.o features.o
 obj-y	+= xenbus/
diff --git a/arch/x86/xen/features.c b/drivers/xen/features.c
rename from arch/x86/xen/features.c
rename to drivers/xen/features.c

[PATCH 06 of 24] xen: Move events.c to drivers/xen for IA64/Xen support

[Jeremy Fitzhardinge]

From: Isaku Yamahata <yamahata@valinux.co.jp>

move arch/x86/xen/events.c undedr drivers/xen to share codes
with x86 and ia64. And minor adjustment to compile.
ia64/xen also uses events.c

Signed-off-by: Yaozu (Eddie) Dong <eddie.dong@intel.com>
Signed-off-by: Isaku Yamahata <yamahata@valinux.co.jp>
Signed-off-by: Jeremy Fitzhardinge <jeremy.fitzhardinge@citrix.com>
---
 arch/x86/xen/Makefile  |    2 +-
 arch/x86/xen/xen-ops.h |    2 +-
 drivers/xen/Makefile   |    2 +-
 drivers/xen/events.c   |    3 +--
 include/xen/xen-ops.h  |    9 ++++++++-
 5 files changed, 12 insertions(+), 6 deletions(-)

diff --git a/arch/x86/xen/Makefile b/arch/x86/xen/Makefile
--- a/arch/x86/xen/Makefile
+++ b/arch/x86/xen/Makefile
@@ -1,4 +1,4 @@
 obj-y		:= enlighten.o setup.o multicalls.o mmu.o \
-			events.o time.o manage.o xen-asm.o
+			time.o manage.o xen-asm.o
 
 obj-$(CONFIG_SMP)	+= smp.o
diff --git a/arch/x86/xen/xen-ops.h b/arch/x86/xen/xen-ops.h
--- a/arch/x86/xen/xen-ops.h
+++ b/arch/x86/xen/xen-ops.h
@@ -3,6 +3,7 @@
 
 #include <linux/init.h>
 #include <linux/irqreturn.h>
+#include <xen/xen-ops.h>
 
 /* These are code, but not functions.  Defined in entry.S */
 extern const char xen_hypervisor_callback[];
@@ -10,7 +11,6 @@
 
 void xen_copy_trap_info(struct trap_info *traps);
 
-DECLARE_PER_CPU(struct vcpu_info *, xen_vcpu);
 DECLARE_PER_CPU(unsigned long, xen_cr3);
 DECLARE_PER_CPU(unsigned long, xen_current_cr3);
 
diff --git a/drivers/xen/Makefile b/drivers/xen/Makefile
--- a/drivers/xen/Makefile
+++ b/drivers/xen/Makefile
@@ -1,2 +1,2 @@
-obj-y	+= grant-table.o features.o
+obj-y	+= grant-table.o features.o events.o
 obj-y	+= xenbus/
diff --git a/arch/x86/xen/events.c b/drivers/xen/events.c
rename from arch/x86/xen/events.c
rename to drivers/xen/events.c
--- a/arch/x86/xen/events.c
+++ b/drivers/xen/events.c
@@ -33,11 +33,10 @@
 #include <asm/xen/hypercall.h>
 #include <asm/xen/hypervisor.h>
 
+#include <xen/xen-ops.h>
 #include <xen/events.h>
 #include <xen/interface/xen.h>
 #include <xen/interface/event_channel.h>
-
-#include "xen-ops.h"
 
 /*
  * This lock protects updates to the following mapping and reference-count
diff --git a/include/xen/xen-ops.h b/include/xen/xen-ops.h
new file mode 100644
--- /dev/null
+++ b/include/xen/xen-ops.h
@@ -0,0 +1,8 @@
+#ifndef INCLUDE_XEN_OPS_H
+#define INCLUDE_XEN_OPS_H
+
+#include <linux/percpu.h>
+
+DECLARE_PER_CPU(struct vcpu_info *, xen_vcpu);
+
+#endif /* INCLUDE_XEN_OPS_H */

[PATCH 07 of 24] Xen: Make events.c portable for ia64/xen support

[Jeremy Fitzhardinge]

From: Isaku Yamahata <yamahata@valinux.co.jp>

Remove x86 dependency in drivers/xen/events.c for ia64/xen support
introducing include/asm/xen/events.h.
Introduce xen_irqs_disabled() to hide regs->flags
Introduce xen_do_IRQ() to hide regs->orig_ax.
make enum ipi_vector definition arch specific. ia64/xen needs four vectors.
Add one rmb() because on ia64 xchg() isn't barrier.

Signed-off-by: Isaku Yamahata <yamahata@valinux.co.jp>
Signed-off-by: Jeremy Fitzhardinge <jeremy.fitzhardinge@citrix.com>
---
 drivers/xen/events.c         |   13 +++++++------
 include/asm-x86/xen/events.h |   22 ++++++++++++++++++++++
 include/xen/events.h         |   10 +---------
 3 files changed, 30 insertions(+), 15 deletions(-)

diff --git a/drivers/xen/events.c b/drivers/xen/events.c
--- a/drivers/xen/events.c
+++ b/drivers/xen/events.c
@@ -469,7 +469,7 @@
 	for_each_online_cpu(i) {
 		struct vcpu_info *v = per_cpu(xen_vcpu, i);
 		printk("%d: masked=%d pending=%d event_sel %08lx\n  ", i,
-			(get_irq_regs() && i == cpu) ? !(get_irq_regs()->flags & X86_EFLAGS_IF) : v->evtchn_upcall_mask,
+			(get_irq_regs() && i == cpu) ? xen_irqs_disabled(get_irq_regs()) : v->evtchn_upcall_mask,
 			v->evtchn_upcall_pending,
 			v->evtchn_pending_sel);
 	}
@@ -527,7 +527,10 @@
 		if (__get_cpu_var(nesting_count)++)
 			goto out;
 
-		/* NB. No need for a barrier here -- XCHG is a barrier on x86. */
+#ifndef CONFIG_X86 /* No need for a barrier -- XCHG is a barrier on x86. */
+		/* Clear master flag /before/ clearing selector flag. */
+		rmb();
+#endif
 		pending_words = xchg(&vcpu_info->evtchn_pending_sel, 0);
 		while (pending_words != 0) {
 			unsigned long pending_bits;
@@ -539,10 +542,8 @@
 				int port = (word_idx * BITS_PER_LONG) + bit_idx;
 				int irq = evtchn_to_irq[port];
 
-				if (irq != -1) {
-					regs->orig_ax = ~irq;
-					do_IRQ(regs);
-				}
+				if (irq != -1)
+					xen_do_IRQ(irq, regs);
 			}
 		}
 
diff --git a/include/asm-x86/xen/events.h b/include/asm-x86/xen/events.h
new file mode 100644
--- /dev/null
+++ b/include/asm-x86/xen/events.h
@@ -0,0 +1,22 @@
+#ifndef __XEN_EVENTS_H
+#define __XEN_EVENTS_H
+
+enum ipi_vector {
+	XEN_RESCHEDULE_VECTOR,
+	XEN_CALL_FUNCTION_VECTOR,
+
+	XEN_NR_IPIS,
+};
+
+static inline int xen_irqs_disabled(struct pt_regs *regs)
+{
+	return raw_irqs_disabled_flags(regs->flags);
+}
+
+static inline void xen_do_IRQ(int irq, struct pt_regs *regs)
+{
+	regs->orig_ax = ~irq;
+	do_IRQ(regs);
+}
+
+#endif /* __XEN_EVENTS_H */
diff --git a/include/xen/events.h b/include/xen/events.h
--- a/include/xen/events.h
+++ b/include/xen/events.h
@@ -5,13 +5,7 @@
 
 #include <xen/interface/event_channel.h>
 #include <asm/xen/hypercall.h>
-
-enum ipi_vector {
-	XEN_RESCHEDULE_VECTOR,
-	XEN_CALL_FUNCTION_VECTOR,
-
-	XEN_NR_IPIS,
-};
+#include <asm/xen/events.h>
 
 int bind_evtchn_to_irq(unsigned int evtchn);
 int bind_evtchn_to_irqhandler(unsigned int evtchn,

[PATCH 08 of 24] Re: [PATCH 08/12] xen: add resend_irq_on_evtchn() definition into events.c

[Jeremy Fitzhardinge]

From: Isaku Yamahata <yamahata@valinux.co.jp>

Define resend_irq_on_evtchn() which ia64/xen uses.
Although it isn't used by current x86/xen code, it's arch generic
so that put it into common code.

Signed-off-by: Isaku Yamahata <yamahata@valinux.co.jp>
Signed-off-by: Jeremy Fitzhardinge <jeremy.fitzhardinge@citrix.com>
---
 drivers/xen/events.c |   16 ++++++++++++++++
 include/xen/events.h |    3 +--
 2 files changed, 17 insertions(+), 2 deletions(-)

diff --git a/drivers/xen/events.c b/drivers/xen/events.c
--- a/drivers/xen/events.c
+++ b/drivers/xen/events.c
@@ -586,6 +586,22 @@
 	rebind_irq_to_cpu(irq, tcpu);
 }
 
+int resend_irq_on_evtchn(unsigned int irq)
+{
+	int masked, evtchn = evtchn_from_irq(irq);
+	struct shared_info *s = HYPERVISOR_shared_info;
+
+	if (!VALID_EVTCHN(evtchn))
+		return 1;
+
+	masked = sync_test_and_set_bit(evtchn, s->evtchn_mask);
+	sync_set_bit(evtchn, s->evtchn_pending);
+	if (!masked)
+		unmask_evtchn(evtchn);
+
+	return 1;
+}
+
 static void enable_dynirq(unsigned int irq)
 {
 	int evtchn = evtchn_from_irq(irq);
diff --git a/include/xen/events.h b/include/xen/events.h
--- a/include/xen/events.h
+++ b/include/xen/events.h
@@ -31,6 +31,7 @@
 void unbind_from_irqhandler(unsigned int irq, void *dev_id);
 
 void xen_send_IPI_one(unsigned int cpu, enum ipi_vector vector);
+int resend_irq_on_evtchn(unsigned int irq);
 
 static inline void notify_remote_via_evtchn(int port)
 {

[PATCH 09 of 24] xen: make include/xen/page.h portable moving those definitions under asm dir

[Jeremy Fitzhardinge]

From: Isaku Yamahata <yamahata@valinux.co.jp>

The definitions in include/asm/xen/page.h are arch specific.
ia64/xen wants to define its own version. So move them to arch specific
directory and keep include/xen/page.h in order not to break compilation.

Signed-off-by: Isaku Yamahata <yamahata@valinux.co.jp>
Signed-off-by: Jeremy Fitzhardinge <jeremy.fitzhardinge@citrix.com>
---
 include/asm-x86/xen/page.h |  168 +++++++++++++++++++++++++++++++++++++++++++
 include/xen/page.h         |  169 --------------------------------------------
 2 files changed, 169 insertions(+), 168 deletions(-)

diff --git a/include/asm-x86/xen/page.h b/include/asm-x86/xen/page.h
new file mode 100644
--- /dev/null
+++ b/include/asm-x86/xen/page.h
@@ -0,0 +1,168 @@
+#ifndef __XEN_PAGE_H
+#define __XEN_PAGE_H
+
+#include <linux/pfn.h>
+
+#include <asm/uaccess.h>
+#include <asm/pgtable.h>
+
+#include <xen/features.h>
+
+/* Xen machine address */
+typedef struct xmaddr {
+	phys_addr_t maddr;
+} xmaddr_t;
+
+/* Xen pseudo-physical address */
+typedef struct xpaddr {
+	phys_addr_t paddr;
+} xpaddr_t;
+
+#define XMADDR(x)	((xmaddr_t) { .maddr = (x) })
+#define XPADDR(x)	((xpaddr_t) { .paddr = (x) })
+
+/**** MACHINE <-> PHYSICAL CONVERSION MACROS ****/
+#define INVALID_P2M_ENTRY	(~0UL)
+#define FOREIGN_FRAME_BIT	(1UL<<31)
+#define FOREIGN_FRAME(m)	((m) | FOREIGN_FRAME_BIT)
+
+extern unsigned long *phys_to_machine_mapping;
+
+static inline unsigned long pfn_to_mfn(unsigned long pfn)
+{
+	if (xen_feature(XENFEAT_auto_translated_physmap))
+		return pfn;
+
+	return phys_to_machine_mapping[(unsigned int)(pfn)] &
+		~FOREIGN_FRAME_BIT;
+}
+
+static inline int phys_to_machine_mapping_valid(unsigned long pfn)
+{
+	if (xen_feature(XENFEAT_auto_translated_physmap))
+		return 1;
+
+	return (phys_to_machine_mapping[pfn] != INVALID_P2M_ENTRY);
+}
+
+static inline unsigned long mfn_to_pfn(unsigned long mfn)
+{
+	unsigned long pfn;
+
+	if (xen_feature(XENFEAT_auto_translated_physmap))
+		return mfn;
+
+#if 0
+	if (unlikely((mfn >> machine_to_phys_order) != 0))
+		return max_mapnr;
+#endif
+
+	pfn = 0;
+	/*
+	 * The array access can fail (e.g., device space beyond end of RAM).
+	 * In such cases it doesn't matter what we return (we return garbage),
+	 * but we must handle the fault without crashing!
+	 */
+	__get_user(pfn, &machine_to_phys_mapping[mfn]);
+
+	return pfn;
+}
+
+static inline xmaddr_t phys_to_machine(xpaddr_t phys)
+{
+	unsigned offset = phys.paddr & ~PAGE_MASK;
+	return XMADDR(PFN_PHYS((u64)pfn_to_mfn(PFN_DOWN(phys.paddr))) | offset);
+}
+
+static inline xpaddr_t machine_to_phys(xmaddr_t machine)
+{
+	unsigned offset = machine.maddr & ~PAGE_MASK;
+	return XPADDR(PFN_PHYS((u64)mfn_to_pfn(PFN_DOWN(machine.maddr))) | offset);
+}
+
+/*
+ * We detect special mappings in one of two ways:
+ *  1. If the MFN is an I/O page then Xen will set the m2p entry
+ *     to be outside our maximum possible pseudophys range.
+ *  2. If the MFN belongs to a different domain then we will certainly
+ *     not have MFN in our p2m table. Conversely, if the page is ours,
+ *     then we'll have p2m(m2p(MFN))==MFN.
+ * If we detect a special mapping then it doesn't have a 'struct page'.
+ * We force !pfn_valid() by returning an out-of-range pointer.
+ *
+ * NB. These checks require that, for any MFN that is not in our reservation,
+ * there is no PFN such that p2m(PFN) == MFN. Otherwise we can get confused if
+ * we are foreign-mapping the MFN, and the other domain as m2p(MFN) == PFN.
+ * Yikes! Various places must poke in INVALID_P2M_ENTRY for safety.
+ *
+ * NB2. When deliberately mapping foreign pages into the p2m table, you *must*
+ *      use FOREIGN_FRAME(). This will cause pte_pfn() to choke on it, as we
+ *      require. In all the cases we care about, the FOREIGN_FRAME bit is
+ *      masked (e.g., pfn_to_mfn()) so behaviour there is correct.
+ */
+static inline unsigned long mfn_to_local_pfn(unsigned long mfn)
+{
+	extern unsigned long max_mapnr;
+	unsigned long pfn = mfn_to_pfn(mfn);
+	if ((pfn < max_mapnr)
+	    && !xen_feature(XENFEAT_auto_translated_physmap)
+	    && (phys_to_machine_mapping[pfn] != mfn))
+		return max_mapnr; /* force !pfn_valid() */
+	return pfn;
+}
+
+static inline void set_phys_to_machine(unsigned long pfn, unsigned long mfn)
+{
+	if (xen_feature(XENFEAT_auto_translated_physmap)) {
+		BUG_ON(pfn != mfn && mfn != INVALID_P2M_ENTRY);
+		return;
+	}
+	phys_to_machine_mapping[pfn] = mfn;
+}
+
+/* VIRT <-> MACHINE conversion */
+#define virt_to_machine(v)	(phys_to_machine(XPADDR(__pa(v))))
+#define virt_to_mfn(v)		(pfn_to_mfn(PFN_DOWN(__pa(v))))
+#define mfn_to_virt(m)		(__va(mfn_to_pfn(m) << PAGE_SHIFT))
+
+static inline unsigned long pte_mfn(pte_t pte)
+{
+	return (pte.pte & ~_PAGE_NX) >> PAGE_SHIFT;
+}
+
+static inline pte_t mfn_pte(unsigned long page_nr, pgprot_t pgprot)
+{
+	pte_t pte;
+
+	pte.pte = ((phys_addr_t)page_nr << PAGE_SHIFT) |
+		(pgprot_val(pgprot) & __supported_pte_mask);
+
+	return pte;
+}
+
+static inline pteval_t pte_val_ma(pte_t pte)
+{
+	return pte.pte;
+}
+
+static inline pte_t __pte_ma(pteval_t x)
+{
+	return (pte_t) { .pte = x };
+}
+
+#ifdef CONFIG_X86_PAE
+#define pmd_val_ma(v) ((v).pmd)
+#define pud_val_ma(v) ((v).pgd.pgd)
+#define __pmd_ma(x)	((pmd_t) { (x) } )
+#else  /* !X86_PAE */
+#define pmd_val_ma(v)	((v).pud.pgd.pgd)
+#endif	/* CONFIG_X86_PAE */
+
+#define pgd_val_ma(x)	((x).pgd)
+
+
+xmaddr_t arbitrary_virt_to_machine(unsigned long address);
+void make_lowmem_page_readonly(void *vaddr);
+void make_lowmem_page_readwrite(void *vaddr);
+
+#endif /* __XEN_PAGE_H */
diff --git a/include/xen/page.h b/include/xen/page.h
--- a/include/xen/page.h
+++ b/include/xen/page.h
@@ -1,168 +1,1 @@
-#ifndef __XEN_PAGE_H
-#define __XEN_PAGE_H
-
-#include <linux/pfn.h>
-
-#include <asm/uaccess.h>
-#include <asm/pgtable.h>
-
-#include <xen/features.h>
-
-/* Xen machine address */
-typedef struct xmaddr {
-	phys_addr_t maddr;
-} xmaddr_t;
-
-/* Xen pseudo-physical address */
-typedef struct xpaddr {
-	phys_addr_t paddr;
-} xpaddr_t;
-
-#define XMADDR(x)	((xmaddr_t) { .maddr = (x) })
-#define XPADDR(x)	((xpaddr_t) { .paddr = (x) })
-
-/**** MACHINE <-> PHYSICAL CONVERSION MACROS ****/
-#define INVALID_P2M_ENTRY	(~0UL)
-#define FOREIGN_FRAME_BIT	(1UL<<31)
-#define FOREIGN_FRAME(m)	((m) | FOREIGN_FRAME_BIT)
-
-extern unsigned long *phys_to_machine_mapping;
-
-static inline unsigned long pfn_to_mfn(unsigned long pfn)
-{
-	if (xen_feature(XENFEAT_auto_translated_physmap))
-		return pfn;
-
-	return phys_to_machine_mapping[(unsigned int)(pfn)] &
-		~FOREIGN_FRAME_BIT;
-}
-
-static inline int phys_to_machine_mapping_valid(unsigned long pfn)
-{
-	if (xen_feature(XENFEAT_auto_translated_physmap))
-		return 1;
-
-	return (phys_to_machine_mapping[pfn] != INVALID_P2M_ENTRY);
-}
-
-static inline unsigned long mfn_to_pfn(unsigned long mfn)
-{
-	unsigned long pfn;
-
-	if (xen_feature(XENFEAT_auto_translated_physmap))
-		return mfn;
-
-#if 0
-	if (unlikely((mfn >> machine_to_phys_order) != 0))
-		return max_mapnr;
-#endif
-
-	pfn = 0;
-	/*
-	 * The array access can fail (e.g., device space beyond end of RAM).
-	 * In such cases it doesn't matter what we return (we return garbage),
-	 * but we must handle the fault without crashing!
-	 */
-	__get_user(pfn, &machine_to_phys_mapping[mfn]);
-
-	return pfn;
-}
-
-static inline xmaddr_t phys_to_machine(xpaddr_t phys)
-{
-	unsigned offset = phys.paddr & ~PAGE_MASK;
-	return XMADDR(PFN_PHYS((u64)pfn_to_mfn(PFN_DOWN(phys.paddr))) | offset);
-}
-
-static inline xpaddr_t machine_to_phys(xmaddr_t machine)
-{
-	unsigned offset = machine.maddr & ~PAGE_MASK;
-	return XPADDR(PFN_PHYS((u64)mfn_to_pfn(PFN_DOWN(machine.maddr))) | offset);
-}
-
-/*
- * We detect special mappings in one of two ways:
- *  1. If the MFN is an I/O page then Xen will set the m2p entry
- *     to be outside our maximum possible pseudophys range.
- *  2. If the MFN belongs to a different domain then we will certainly
- *     not have MFN in our p2m table. Conversely, if the page is ours,
- *     then we'll have p2m(m2p(MFN))==MFN.
- * If we detect a special mapping then it doesn't have a 'struct page'.
- * We force !pfn_valid() by returning an out-of-range pointer.
- *
- * NB. These checks require that, for any MFN that is not in our reservation,
- * there is no PFN such that p2m(PFN) == MFN. Otherwise we can get confused if
- * we are foreign-mapping the MFN, and the other domain as m2p(MFN) == PFN.
- * Yikes! Various places must poke in INVALID_P2M_ENTRY for safety.
- *
- * NB2. When deliberately mapping foreign pages into the p2m table, you *must*
- *      use FOREIGN_FRAME(). This will cause pte_pfn() to choke on it, as we
- *      require. In all the cases we care about, the FOREIGN_FRAME bit is
- *      masked (e.g., pfn_to_mfn()) so behaviour there is correct.
- */
-static inline unsigned long mfn_to_local_pfn(unsigned long mfn)
-{
-	extern unsigned long max_mapnr;
-	unsigned long pfn = mfn_to_pfn(mfn);
-	if ((pfn < max_mapnr)
-	    && !xen_feature(XENFEAT_auto_translated_physmap)
-	    && (phys_to_machine_mapping[pfn] != mfn))
-		return max_mapnr; /* force !pfn_valid() */
-	return pfn;
-}
-
-static inline void set_phys_to_machine(unsigned long pfn, unsigned long mfn)
-{
-	if (xen_feature(XENFEAT_auto_translated_physmap)) {
-		BUG_ON(pfn != mfn && mfn != INVALID_P2M_ENTRY);
-		return;
-	}
-	phys_to_machine_mapping[pfn] = mfn;
-}
-
-/* VIRT <-> MACHINE conversion */
-#define virt_to_machine(v)	(phys_to_machine(XPADDR(__pa(v))))
-#define virt_to_mfn(v)		(pfn_to_mfn(PFN_DOWN(__pa(v))))
-#define mfn_to_virt(m)		(__va(mfn_to_pfn(m) << PAGE_SHIFT))
-
-static inline unsigned long pte_mfn(pte_t pte)
-{
-	return (pte.pte & ~_PAGE_NX) >> PAGE_SHIFT;
-}
-
-static inline pte_t mfn_pte(unsigned long page_nr, pgprot_t pgprot)
-{
-	pte_t pte;
-
-	pte.pte = ((phys_addr_t)page_nr << PAGE_SHIFT) |
-		(pgprot_val(pgprot) & __supported_pte_mask);
-
-	return pte;
-}
-
-static inline pteval_t pte_val_ma(pte_t pte)
-{
-	return pte.pte;
-}
-
-static inline pte_t __pte_ma(pteval_t x)
-{
-	return (pte_t) { .pte = x };
-}
-
-#ifdef CONFIG_X86_PAE
-#define pmd_val_ma(v) ((v).pmd)
-#define pud_val_ma(v) ((v).pgd.pgd)
-#define __pmd_ma(x)	((pmd_t) { (x) } )
-#else  /* !X86_PAE */
-#define pmd_val_ma(v)	((v).pud.pgd.pgd)
-#endif	/* CONFIG_X86_PAE */
-
-#define pgd_val_ma(x)	((x).pgd)
-
-
-xmaddr_t arbitrary_virt_to_machine(unsigned long address);
-void make_lowmem_page_readonly(void *vaddr);
-void make_lowmem_page_readwrite(void *vaddr);
-
-#endif /* __XEN_PAGE_H */
+#include <asm/xen/page.h>

[PATCH 10 of 24] xen: replace callers of alloc_vm_area()/free_vm_area() with xen_ prefixed one

[Jeremy Fitzhardinge]

From: Isaku Yamahata <yamahata@valinux.co.jp>

Don't use alloc_vm_area()/free_vm_area() directly, instead define
xen_alloc_vm_area()/xen_free_vm_area() and use them.

alloc_vm_area()/free_vm_area() are used to allocate/free area which
are for grant table mapping. Xen/x86 grant table is based on virtual
address so that alloc_vm_area()/free_vm_area() are suitable.
On the other hand Xen/ia64 (and Xen/powerpc) grant table is based on
pseudo physical address (guest physical address) so that allocation
should be done differently.
The original version of xenified Linux/IA64 have its own
allocate_vm_area()/free_vm_area() definitions which don't allocate vm area
contradictory to those names.
Now vanilla Linux already has its definitions so that it's impossible
to have IA64 definitions of allocate_vm_area()/free_vm_area().
Instead introduce xen_allocate_vm_area()/xen_free_vm_area() and use them.

Signed-off-by: Isaku Yamahata <yamahata@valinux.co.jp>
Signed-off-by: Jeremy Fitzhardinge <jeremy.fitzhardinge@citrix.com>
---
 drivers/xen/grant-table.c          |    2 +-
 drivers/xen/xenbus/xenbus_client.c |    6 +++---
 include/asm-x86/xen/grant_table.h  |    7 +++++++
 include/xen/grant_table.h          |    3 +--
 4 files changed, 12 insertions(+), 6 deletions(-)

diff --git a/drivers/xen/grant-table.c b/drivers/xen/grant-table.c
--- a/drivers/xen/grant-table.c
+++ b/drivers/xen/grant-table.c
@@ -478,7 +478,7 @@
 
 	if (shared == NULL) {
 		struct vm_struct *area;
-		area = alloc_vm_area(PAGE_SIZE * max_nr_grant_frames());
+		area = xen_alloc_vm_area(PAGE_SIZE * max_nr_grant_frames());
 		BUG_ON(area == NULL);
 		shared = area->addr;
 	}
diff --git a/drivers/xen/xenbus/xenbus_client.c b/drivers/xen/xenbus/xenbus_client.c
--- a/drivers/xen/xenbus/xenbus_client.c
+++ b/drivers/xen/xenbus/xenbus_client.c
@@ -399,7 +399,7 @@
 
 	*vaddr = NULL;
 
-	area = alloc_vm_area(PAGE_SIZE);
+	area = xen_alloc_vm_area(PAGE_SIZE);
 	if (!area)
 		return -ENOMEM;
 
@@ -409,7 +409,7 @@
 		BUG();
 
 	if (op.status != GNTST_okay) {
-		free_vm_area(area);
+		xen_free_vm_area(area);
 		xenbus_dev_fatal(dev, op.status,
 				 "mapping in shared page %d from domain %d",
 				 gnt_ref, dev->otherend_id);
@@ -508,7 +508,7 @@
 		BUG();
 
 	if (op.status == GNTST_okay)
-		free_vm_area(area);
+		xen_free_vm_area(area);
 	else
 		xenbus_dev_error(dev, op.status,
 				 "unmapping page at handle %d error %d",
diff --git a/include/asm-x86/xen/grant_table.h b/include/asm-x86/xen/grant_table.h
new file mode 100644
--- /dev/null
+++ b/include/asm-x86/xen/grant_table.h
@@ -0,0 +1,7 @@
+#ifndef __XEN_GRANT_TABLE_H
+#define __XEN_GRANT_TABLE_H
+
+#define xen_alloc_vm_area(size)	alloc_vm_area(size)
+#define xen_free_vm_area(area)	free_vm_area(area)
+
+#endif /* __XEN_GRANT_TABLE_H */
diff --git a/include/xen/grant_table.h b/include/xen/grant_table.h
--- a/include/xen/grant_table.h
+++ b/include/xen/grant_table.h
@@ -39,6 +39,7 @@
 
 #include <asm/xen/hypervisor.h>
 #include <xen/interface/grant_table.h>
+#include <asm/xen/grant_table.h>
 
 /* NR_GRANT_FRAMES must be less than or equal to that configured in Xen */
 #define NR_GRANT_FRAMES 4

[PATCH 11 of 24] xen: make grant table arch portable

[Jeremy Fitzhardinge]

From: Isaku Yamahata <yamahata@valinux.co.jp>

split out x86 specific part from grant-table.c and
allow ia64/xen specific initialization.
ia64/xen grant table is based on pseudo physical address
(guest physical address) unlike x86/xen. On ia64 init_mm
doesn't map identity straight mapped area.
ia64/xen specific grant table initialization is necessary.

Signed-off-by: Isaku Yamahata <yamahata@valinux.co.jp>
---
 arch/x86/xen/Makefile      |    2
 arch/x86/xen/grant-table.c |   91 ++++++++++++++++++++++++++++++++++++++++++++
 drivers/xen/grant-table.c  |   35 +---------------
 include/xen/grant_table.h  |    7 ++-
 4 files changed, 101 insertions(+), 34 deletions(-)

diff --git a/arch/x86/xen/Makefile b/arch/x86/xen/Makefile
--- a/arch/x86/xen/Makefile
+++ b/arch/x86/xen/Makefile
@@ -1,4 +1,4 @@
 obj-y		:= enlighten.o setup.o multicalls.o mmu.o \
-			time.o manage.o xen-asm.o
+			time.o manage.o xen-asm.o grant-table.o
 
 obj-$(CONFIG_SMP)	+= smp.o
diff --git a/arch/x86/xen/grant-table.c b/arch/x86/xen/grant-table.c
new file mode 100644
--- /dev/null
+++ b/arch/x86/xen/grant-table.c
@@ -0,0 +1,91 @@
+/******************************************************************************
+ * grant_table.c
+ * x86 specific part
+ *
+ * Granting foreign access to our memory reservation.
+ *
+ * Copyright (c) 2005-2006, Christopher Clark
+ * Copyright (c) 2004-2005, K A Fraser
+ * Copyright (c) 2008 Isaku Yamahata <yamahata at valinux co jp>
+ *                    VA Linux Systems Japan. Split out x86 specific part.
+ *
+ * This program is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU General Public License version 2
+ * as published by the Free Software Foundation; or, when distributed
+ * separately from the Linux kernel or incorporated into other
+ * software packages, subject to the following license:
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a copy
+ * of this source file (the "Software"), to deal in the Software without
+ * restriction, including without limitation the rights to use, copy, modify,
+ * merge, publish, distribute, sublicense, and/or sell copies of the Software,
+ * and to permit persons to whom the Software is furnished to do so, subject to
+ * the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
+ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
+ * IN THE SOFTWARE.
+ */
+
+#include <linux/sched.h>
+#include <linux/mm.h>
+#include <linux/vmalloc.h>
+
+#include <xen/interface/xen.h>
+#include <xen/page.h>
+#include <xen/grant_table.h>
+
+#include <asm/pgtable.h>
+
+static int map_pte_fn(pte_t *pte, struct page *pmd_page,
+		      unsigned long addr, void *data)
+{
+	unsigned long **frames = (unsigned long **)data;
+
+	set_pte_at(&init_mm, addr, pte, mfn_pte((*frames)[0], PAGE_KERNEL));
+	(*frames)++;
+	return 0;
+}
+
+static int unmap_pte_fn(pte_t *pte, struct page *pmd_page,
+			unsigned long addr, void *data)
+{
+
+	set_pte_at(&init_mm, addr, pte, __pte(0));
+	return 0;
+}
+
+int arch_gnttab_map_shared(unsigned long *frames, unsigned long nr_gframes,
+			   unsigned long max_nr_gframes,
+			   struct grant_entry **__shared)
+{
+	int rc;
+	struct grant_entry *shared = *__shared;
+
+	if (shared == NULL) {
+		struct vm_struct *area =
+			xen_alloc_vm_area(PAGE_SIZE * max_nr_gframes);
+		BUG_ON(area == NULL);
+		shared = area->addr;
+		*__shared = shared;
+	}
+
+	rc = apply_to_page_range(&init_mm, (unsigned long)shared,
+				 PAGE_SIZE * nr_gframes,
+				 map_pte_fn, &frames);
+	return rc;
+}
+
+void arch_gnttab_unmap_shared(struct grant_entry *shared,
+			      unsigned long nr_gframes)
+{
+	apply_to_page_range(&init_mm, (unsigned long)shared,
+			    PAGE_SIZE * nr_gframes, unmap_pte_fn, NULL);
+}
diff --git a/drivers/xen/grant-table.c b/drivers/xen/grant-table.c
--- a/drivers/xen/grant-table.c
+++ b/drivers/xen/grant-table.c
@@ -435,24 +435,6 @@
 	return xen_max;
 }
 
-static int map_pte_fn(pte_t *pte, struct page *pmd_page,
-		      unsigned long addr, void *data)
-{
-	unsigned long **frames = (unsigned long **)data;
-
-	set_pte_at(&init_mm, addr, pte, mfn_pte((*frames)[0], PAGE_KERNEL));
-	(*frames)++;
-	return 0;
-}
-
-static int unmap_pte_fn(pte_t *pte, struct page *pmd_page,
-			unsigned long addr, void *data)
-{
-
-	set_pte_at(&init_mm, addr, pte, __pte(0));
-	return 0;
-}
-
 static int gnttab_map(unsigned int start_idx, unsigned int end_idx)
 {
 	struct gnttab_setup_table setup;
@@ -476,17 +458,9 @@
 
 	BUG_ON(rc || setup.status);
 
-	if (shared == NULL) {
-		struct vm_struct *area;
-		area = xen_alloc_vm_area(PAGE_SIZE * max_nr_grant_frames());
-		BUG_ON(area == NULL);
-		shared = area->addr;
-	}
-	rc = apply_to_page_range(&init_mm, (unsigned long)shared,
-				 PAGE_SIZE * nr_gframes,
-				 map_pte_fn, &frames);
+	rc = arch_gnttab_map_shared(frames, nr_gframes, max_nr_grant_frames(),
+				    &shared);
 	BUG_ON(rc);
-	frames -= nr_gframes; /* adjust after map_pte_fn() */
 
 	kfree(frames);
 
@@ -502,10 +476,7 @@
 
 static int gnttab_suspend(void)
 {
-	apply_to_page_range(&init_mm, (unsigned long)shared,
-			    PAGE_SIZE * nr_grant_frames,
-			    unmap_pte_fn, NULL);
-
+	arch_gnttab_unmap_shared(shared, nr_grant_frames);
 	return 0;
 }
 
diff --git a/include/xen/grant_table.h b/include/xen/grant_table.h
--- a/include/xen/grant_table.h
+++ b/include/xen/grant_table.h
@@ -103,6 +103,12 @@
 void gnttab_grant_foreign_transfer_ref(grant_ref_t, domid_t domid,
 				       unsigned long pfn);
 
+int arch_gnttab_map_shared(unsigned long *frames, unsigned long nr_gframes,
+			   unsigned long max_nr_gframes,
+			   struct grant_entry **__shared);
+void arch_gnttab_unmap_shared(struct grant_entry *shared,
+			      unsigned long nr_gframes);
+
 #define gnttab_map_vaddr(map) ((void *)(map.host_virt_addr))
 
 #endif /* __ASM_GNTTAB_H__ */

[PATCH 12 of 24] xen: import arch generic part of xencomm

[Jeremy Fitzhardinge]

From: Isaku Yamahata <yamahata@valinux.co.jp>

On xen/ia64 and xen/powerpc hypercall arguments are passed by pseudo
physical address (guest physical address) so that it's necessary to
convert from virtual address into pseudo physical address. The frame
work is called xencomm.
Import arch generic part of xencomm.

Signed-off-by: Isaku Yamahata <yamahata@valinux.co.jp>
Signed-off-by: Jeremy Fitzhardinge <jeremy.fitzhardinge@citrix.com>
---
 drivers/xen/Makefile            |    1
 drivers/xen/xencomm.c           |  232 +++++++++++++++++++++++++++++++++++++++
 include/xen/interface/xencomm.h |   41 ++++++
 include/xen/xencomm.h           |   78 ++++++++++++-
 4 files changed, 351 insertions(+), 1 deletion(-)

diff --git a/drivers/xen/Makefile b/drivers/xen/Makefile
--- a/drivers/xen/Makefile
+++ b/drivers/xen/Makefile
@@ -1,2 +1,3 @@
 obj-y	+= grant-table.o features.o events.o
 obj-y	+= xenbus/
+obj-$(CONFIG_XEN_XENCOMM)	+= xencomm.o
diff --git a/drivers/xen/xencomm.c b/drivers/xen/xencomm.c
new file mode 100644
--- /dev/null
+++ b/drivers/xen/xencomm.c
@@ -0,0 +1,232 @@
+/*
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307 USA
+ *
+ * Copyright (C) IBM Corp. 2006
+ *
+ * Authors: Hollis Blanchard <hollisb@us.ibm.com>
+ */
+
+#include <linux/gfp.h>
+#include <linux/mm.h>
+#include <asm/page.h>
+#include <xen/xencomm.h>
+#include <xen/interface/xen.h>
+#ifdef __ia64__
+#include <asm/xen/xencomm.h>	/* for is_kern_addr() */
+#endif
+
+#ifdef HAVE_XEN_PLATFORM_COMPAT_H
+#include <xen/platform-compat.h>
+#endif
+
+static int xencomm_init(struct xencomm_desc *desc,
+			void *buffer, unsigned long bytes)
+{
+	unsigned long recorded = 0;
+	int i = 0;
+
+	while ((recorded < bytes) && (i < desc->nr_addrs)) {
+		unsigned long vaddr = (unsigned long)buffer + recorded;
+		unsigned long paddr;
+		int offset;
+		int chunksz;
+
+		offset = vaddr % PAGE_SIZE; /* handle partial pages */
+		chunksz = min(PAGE_SIZE - offset, bytes - recorded);
+
+		paddr = xencomm_vtop(vaddr);
+		if (paddr == ~0UL) {
+			printk(KERN_DEBUG "%s: couldn't translate vaddr %lx\n",
+			       __func__, vaddr);
+			return -EINVAL;
+		}
+
+		desc->address[i++] = paddr;
+		recorded += chunksz;
+	}
+
+	if (recorded < bytes) {
+		printk(KERN_DEBUG
+		       "%s: could only translate %ld of %ld bytes\n",
+		       __func__, recorded, bytes);
+		return -ENOSPC;
+	}
+
+	/* mark remaining addresses invalid (just for safety) */
+	while (i < desc->nr_addrs)
+		desc->address[i++] = XENCOMM_INVALID;
+
+	desc->magic = XENCOMM_MAGIC;
+
+	return 0;
+}
+
+static struct xencomm_desc *xencomm_alloc(gfp_t gfp_mask,
+					  void *buffer, unsigned long bytes)
+{
+	struct xencomm_desc *desc;
+	unsigned long buffer_ulong = (unsigned long)buffer;
+	unsigned long start = buffer_ulong & PAGE_MASK;
+	unsigned long end = (buffer_ulong + bytes) | ~PAGE_MASK;
+	unsigned long nr_addrs = (end - start + 1) >> PAGE_SHIFT;
+	unsigned long size = sizeof(*desc) +
+		sizeof(desc->address[0]) * nr_addrs;
+
+	/*
+	 * slab allocator returns at least sizeof(void*) aligned pointer.
+	 * When sizeof(*desc) > sizeof(void*), struct xencomm_desc might
+	 * cross page boundary.
+	 */
+	if (sizeof(*desc) > sizeof(void *)) {
+		unsigned long order = get_order(size);
+		desc = (struct xencomm_desc *)__get_free_pages(gfp_mask,
+							       order);
+		if (desc == NULL)
+			return NULL;
+
+		desc->nr_addrs =
+			((PAGE_SIZE << order) - sizeof(struct xencomm_desc)) /
+			sizeof(*desc->address);
+	} else {
+		desc = kmalloc(size, gfp_mask);
+		if (desc == NULL)
+			return NULL;
+
+		desc->nr_addrs = nr_addrs;
+	}
+	return desc;
+}
+
+void xencomm_free(struct xencomm_handle *desc)
+{
+	if (desc && !((ulong)desc & XENCOMM_INLINE_FLAG)) {
+		struct xencomm_desc *desc__ = (struct xencomm_desc *)desc;
+		if (sizeof(*desc__) > sizeof(void *)) {
+			unsigned long size = sizeof(*desc__) +
+				sizeof(desc__->address[0]) * desc__->nr_addrs;
+			unsigned long order = get_order(size);
+			free_pages((unsigned long)__va(desc), order);
+		} else
+			kfree(__va(desc));
+	}
+}
+
+static int xencomm_create(void *buffer, unsigned long bytes,
+			  struct xencomm_desc **ret, gfp_t gfp_mask)
+{
+	struct xencomm_desc *desc;
+	int rc;
+
+	pr_debug("%s: %p[%ld]\n", __func__, buffer, bytes);
+
+	if (bytes == 0) {
+		/* don't create a descriptor; Xen recognizes NULL. */
+		BUG_ON(buffer != NULL);
+		*ret = NULL;
+		return 0;
+	}
+
+	BUG_ON(buffer == NULL); /* 'bytes' is non-zero */
+
+	desc = xencomm_alloc(gfp_mask, buffer, bytes);
+	if (!desc) {
+		printk(KERN_DEBUG "%s failure\n", "xencomm_alloc");
+		return -ENOMEM;
+	}
+
+	rc = xencomm_init(desc, buffer, bytes);
+	if (rc) {
+		printk(KERN_DEBUG "%s failure: %d\n", "xencomm_init", rc);
+		xencomm_free((struct xencomm_handle *)__pa(desc));
+		return rc;
+	}
+
+	*ret = desc;
+	return 0;
+}
+
+/* check if memory address is within VMALLOC region  */
+static int is_phys_contiguous(unsigned long addr)
+{
+	if (!is_kernel_addr(addr))
+		return 0;
+
+	return (addr < VMALLOC_START) || (addr >= VMALLOC_END);
+}
+
+static struct xencomm_handle *xencomm_create_inline(void *ptr)
+{
+	unsigned long paddr;
+
+	BUG_ON(!is_phys_contiguous((unsigned long)ptr));
+
+	paddr = (unsigned long)xencomm_pa(ptr);
+	BUG_ON(paddr & XENCOMM_INLINE_FLAG);
+	return (struct xencomm_handle *)(paddr | XENCOMM_INLINE_FLAG);
+}
+
+/* "mini" routine, for stack-based communications: */
+static int xencomm_create_mini(void *buffer,
+	unsigned long bytes, struct xencomm_mini *xc_desc,
+	struct xencomm_desc **ret)
+{
+	int rc = 0;
+	struct xencomm_desc *desc;
+	BUG_ON(((unsigned long)xc_desc) % sizeof(*xc_desc) != 0);
+
+	desc = (void *)xc_desc;
+
+	desc->nr_addrs = XENCOMM_MINI_ADDRS;
+
+	rc = xencomm_init(desc, buffer, bytes);
+	if (!rc)
+		*ret = desc;
+
+	return rc;
+}
+
+struct xencomm_handle *xencomm_map(void *ptr, unsigned long bytes)
+{
+	int rc;
+	struct xencomm_desc *desc;
+
+	if (is_phys_contiguous((unsigned long)ptr))
+		return xencomm_create_inline(ptr);
+
+	rc = xencomm_create(ptr, bytes, &desc, GFP_KERNEL);
+
+	if (rc || desc == NULL)
+		return NULL;
+
+	return xencomm_pa(desc);
+}
+
+struct xencomm_handle *__xencomm_map_no_alloc(void *ptr, unsigned long bytes,
+			struct xencomm_mini *xc_desc)
+{
+	int rc;
+	struct xencomm_desc *desc = NULL;
+
+	if (is_phys_contiguous((unsigned long)ptr))
+		return xencomm_create_inline(ptr);
+
+	rc = xencomm_create_mini(ptr, bytes, xc_desc,
+				&desc);
+
+	if (rc)
+		return NULL;
+
+	return xencomm_pa(desc);
+}
diff --git a/include/xen/interface/xencomm.h b/include/xen/interface/xencomm.h
new file mode 100644
--- /dev/null
+++ b/include/xen/interface/xencomm.h
@@ -0,0 +1,41 @@
+/*
+ * Permission is hereby granted, free of charge, to any person obtaining a copy
+ * of this software and associated documentation files (the "Software"), to
+ * deal in the Software without restriction, including without limitation the
+ * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or
+ * sell copies of the Software, and to permit persons to whom the Software is
+ * furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
+ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
+ * DEALINGS IN THE SOFTWARE.
+ *
+ * Copyright (C) IBM Corp. 2006
+ */
+
+#ifndef _XEN_XENCOMM_H_
+#define _XEN_XENCOMM_H_
+
+/* A xencomm descriptor is a scatter/gather list containing physical
+ * addresses corresponding to a virtually contiguous memory area. The
+ * hypervisor translates these physical addresses to machine addresses to copy
+ * to and from the virtually contiguous area.
+ */
+
+#define XENCOMM_MAGIC 0x58434F4D /* 'XCOM' */
+#define XENCOMM_INVALID (~0UL)
+
+struct xencomm_desc {
+    uint32_t magic;
+    uint32_t nr_addrs; /* the number of entries in address[] */
+    uint64_t address[0];
+};
+
+#endif /* _XEN_XENCOMM_H_ */
diff --git a/include/xen/xencomm.h b/include/xen/xencomm.h
new file mode 100644
--- /dev/null
+++ b/include/xen/xencomm.h
@@ -0,0 +1,77 @@
+/*
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307 USA
+ *
+ * Copyright (C) IBM Corp. 2006
+ *
+ * Authors: Hollis Blanchard <hollisb@us.ibm.com>
+ *          Jerone Young <jyoung5@us.ibm.com>
+ */
+
+#ifndef _LINUX_XENCOMM_H_
+#define _LINUX_XENCOMM_H_
+
+#include <xen/interface/xencomm.h>
+
+#define XENCOMM_MINI_ADDRS 3
+struct xencomm_mini {
+	struct xencomm_desc _desc;
+	uint64_t address[XENCOMM_MINI_ADDRS];
+};
+
+/* To avoid additionnal virt to phys conversion, an opaque structure is
+   presented.  */
+struct xencomm_handle;
+
+extern void xencomm_free(struct xencomm_handle *desc);
+extern struct xencomm_handle *xencomm_map(void *ptr, unsigned long bytes);
+extern struct xencomm_handle *__xencomm_map_no_alloc(void *ptr,
+			unsigned long bytes,  struct xencomm_mini *xc_area);
+
+#if 0
+#define XENCOMM_MINI_ALIGNED(xc_desc, n)				\
+	struct xencomm_mini xc_desc ## _base[(n)]			\
+	__attribute__((__aligned__(sizeof(struct xencomm_mini))));	\
+	struct xencomm_mini *xc_desc = &xc_desc ## _base[0];
+#else
+/*
+ * gcc bug workaround:
+ * http://gcc.gnu.org/bugzilla/show_bug.cgi?id=16660
+ * gcc doesn't handle properly stack variable with
+ * __attribute__((__align__(sizeof(struct xencomm_mini))))
+ */
+#define XENCOMM_MINI_ALIGNED(xc_desc, n)				\
+	unsigned char xc_desc ## _base[((n) + 1 ) *			\
+				       sizeof(struct xencomm_mini)];	\
+	struct xencomm_mini *xc_desc = (struct xencomm_mini *)		\
+		((unsigned long)xc_desc ## _base +			\
+		 (sizeof(struct xencomm_mini) -				\
+		  ((unsigned long)xc_desc ## _base) %			\
+		  sizeof(struct xencomm_mini)));
+#endif
+#define xencomm_map_no_alloc(ptr, bytes)			\
+	({ XENCOMM_MINI_ALIGNED(xc_desc, 1);			\
+		__xencomm_map_no_alloc(ptr, bytes, xc_desc); })
+
+/* provided by architecture code: */
+extern unsigned long xencomm_vtop(unsigned long vaddr);
+
+static inline void *xencomm_pa(void *ptr)
+{
+	return (void *)xencomm_vtop((unsigned long)ptr);
+}
+
+#define xen_guest_handle(hnd)  ((hnd).p)
+
+#endif /* _LINUX_XENCOMM_H_ */

[PATCH 13 of 24] [PATCH] xen: Make xen-blkfront write its protocol ABI to xenstore

[Jeremy Fitzhardinge]

From: Markus Armbruster <armbru@redhat.com>

Frontends are expected to write their protocol ABI to xenstore.  Since
the protocol ABI defaults to the backend's native ABI, things work
fine without that as long as the frontend's native ABI is identical to
the backend's native ABI.  This is not the case for xen-blkfront
running 32-on-64, because its ABI differs between 32 and 64 bit, and
thus needs this fix.

Based on http://xenbits.xensource.com/xen-unstable.hg?rev/c545932a18f3
and http://xenbits.xensource.com/xen-unstable.hg?rev/ffe52263b430 by
Gerd Hoffmann <kraxel@suse.de>

Signed-off-by: Markus Armbruster <armbru@redhat.com>
Signed-off-by: Jeremy Fitzhardinge <Jeremy.Fitzhardinge@citrix.com>
---
 drivers/block/xen-blkfront.c         |    7 +++++++
 include/xen/interface/io/protocols.h |   22 +++++++++++++++++++++-
 2 files changed, 28 insertions(+), 1 deletion(-)

diff --git a/drivers/block/xen-blkfront.c b/drivers/block/xen-blkfront.c
--- a/drivers/block/xen-blkfront.c
+++ b/drivers/block/xen-blkfront.c
@@ -47,6 +47,7 @@
 
 #include <xen/interface/grant_table.h>
 #include <xen/interface/io/blkif.h>
+#include <xen/interface/io/protocols.h>
 
 #include <asm/xen/hypervisor.h>
 
@@ -614,6 +615,12 @@
 		message = "writing event-channel";
 		goto abort_transaction;
 	}
+	err = xenbus_printf(xbt, dev->nodename, "protocol", "%s",
+			    XEN_IO_PROTO_ABI_NATIVE);
+	if (err) {
+		message = "writing protocol";
+		goto abort_transaction;
+	}
 
 	err = xenbus_transaction_end(xbt, 0);
 	if (err) {
diff --git a/include/xen/interface/io/protocols.h b/include/xen/interface/io/protocols.h
new file mode 100644
--- /dev/null
+++ b/include/xen/interface/io/protocols.h
@@ -0,0 +1,21 @@
+#ifndef __XEN_PROTOCOLS_H__
+#define __XEN_PROTOCOLS_H__
+
+#define XEN_IO_PROTO_ABI_X86_32     "x86_32-abi"
+#define XEN_IO_PROTO_ABI_X86_64     "x86_64-abi"
+#define XEN_IO_PROTO_ABI_IA64       "ia64-abi"
+#define XEN_IO_PROTO_ABI_POWERPC64  "powerpc64-abi"
+
+#if defined(__i386__)
+# define XEN_IO_PROTO_ABI_NATIVE XEN_IO_PROTO_ABI_X86_32
+#elif defined(__x86_64__)
+# define XEN_IO_PROTO_ABI_NATIVE XEN_IO_PROTO_ABI_X86_64
+#elif defined(__ia64__)
+# define XEN_IO_PROTO_ABI_NATIVE XEN_IO_PROTO_ABI_IA64
+#elif defined(__powerpc64__)
+# define XEN_IO_PROTO_ABI_NATIVE XEN_IO_PROTO_ABI_POWERPC64
+#else
+# error arch fixup needed here
+#endif
+
+#endif

[PATCH 14 of 24] xen/blkfront: use bdget_disk

[Jeremy Fitzhardinge]

info->dev is never initialized to anything, so bdget(info->dev) is
meaningless.  Get rid of info->dev, and use bdget_disk on the gendisk.

Signed-off-by: Jeremy Fitzhardinge <jeremy.fitzhardinge@citrix.com>
Cc: Al Viro <viro@ZenIV.linux.org.uk>
---
 drivers/block/xen-blkfront.c |    3 +--
 1 file changed, 1 insertion(+), 2 deletions(-)

diff --git a/drivers/block/xen-blkfront.c b/drivers/block/xen-blkfront.c
--- a/drivers/block/xen-blkfront.c
+++ b/drivers/block/xen-blkfront.c
@@ -75,7 +75,6 @@
 struct blkfront_info
 {
 	struct xenbus_device *xbdev;
-	dev_t dev;
 	struct gendisk *gd;
 	int vdevice;
 	blkif_vdev_t handle;
@@ -903,7 +902,7 @@
 		break;
 
 	case XenbusStateClosing:
-		bd = bdget(info->dev);
+		bd = bdget_disk(info->gd, 0);
 		if (bd == NULL)
 			xenbus_dev_fatal(dev, -ENODEV, "bdget failed");
 

[PATCH 15 of 24] xen blkfront: Delay wait for block devices until after the disk is added

[Jeremy Fitzhardinge]

From: Christian Limpach <Christian.Limpach@xensource.com>

When the xen block frontend driver is built as a module the module load
is only synchronous up to the point where the frontend and the backend
become connected rather than when the disk is added.

This means that there can be a race on boot between loading the module and
loading the dm-* modules and doing the scan for LVM physical volumes (all
in the initrd). In the failure case the disk is not present until after the
scan for physical volumes is complete.

Taken from:

  http://xenbits.xensource.com/linux-2.6.18-xen.hg?rev/11483a00c017

Signed-off-by: Christian Limpach <Christian.Limpach@xensource.com>
Signed-off-by: Mark McLoughlin <markmc@redhat.com>
Signed-off-by: Jeremy Fitzhardinge <jeremy.fitzhardinge@citrix.com>
---
 drivers/block/xen-blkfront.c      |   11 +++++++++++
 drivers/xen/xenbus/xenbus_probe.c |    5 ++++-
 include/xen/xenbus.h              |    2 +-
 3 files changed, 16 insertions(+), 2 deletions(-)

diff --git a/drivers/block/xen-blkfront.c b/drivers/block/xen-blkfront.c
--- a/drivers/block/xen-blkfront.c
+++ b/drivers/block/xen-blkfront.c
@@ -88,6 +88,7 @@
 	struct blk_shadow shadow[BLK_RING_SIZE];
 	unsigned long shadow_free;
 	int feature_barrier;
+	int is_ready;
 
 	/**
 	 * The number of people holding this device open.  We won't allow a
@@ -839,6 +840,8 @@
 	spin_unlock_irq(&blkif_io_lock);
 
 	add_disk(info->gd);
+
+	info->is_ready = 1;
 }
 
 /**
@@ -931,6 +934,13 @@
 	return 0;
 }
 
+static int blkfront_is_ready(struct xenbus_device *dev)
+{
+	struct blkfront_info *info = dev->dev.driver_data;
+
+	return info->is_ready;
+}
+
 static int blkif_open(struct inode *inode, struct file *filep)
 {
 	struct blkfront_info *info = inode->i_bdev->bd_disk->private_data;
@@ -977,6 +987,7 @@
 	.remove = blkfront_remove,
 	.resume = blkfront_resume,
 	.otherend_changed = backend_changed,
+	.is_ready = blkfront_is_ready,
 };
 
 static int __init xlblk_init(void)
diff --git a/drivers/xen/xenbus/xenbus_probe.c b/drivers/xen/xenbus/xenbus_probe.c
--- a/drivers/xen/xenbus/xenbus_probe.c
+++ b/drivers/xen/xenbus/xenbus_probe.c
@@ -846,6 +846,7 @@
 {
 	struct xenbus_device *xendev = to_xenbus_device(dev);
 	struct device_driver *drv = data;
+	struct xenbus_driver *xendrv;
 
 	/*
 	 * A device with no driver will never connect. We care only about
@@ -858,7 +859,9 @@
 	if (drv && (dev->driver != drv))
 		return 0;
 
-	return (xendev->state != XenbusStateConnected);
+	xendrv = to_xenbus_driver(dev->driver);
+	return (xendev->state != XenbusStateConnected ||
+		(xendrv->is_ready && !xendrv->is_ready(xendev)));
 }
 
 static int exists_disconnected_device(struct device_driver *drv)
diff --git a/include/xen/xenbus.h b/include/xen/xenbus.h
--- a/include/xen/xenbus.h
+++ b/include/xen/xenbus.h
@@ -97,6 +97,7 @@
 	int (*uevent)(struct xenbus_device *, char **, int, char *, int);
 	struct device_driver driver;
 	int (*read_otherend_details)(struct xenbus_device *dev);
+	int (*is_ready)(struct xenbus_device *dev);
 };
 
 static inline struct xenbus_driver *to_xenbus_driver(struct device_driver *drv)

[PATCH 16 of 24] xen: Module autoprobing support for frontend drivers

[Jeremy Fitzhardinge]

From: Mark McLoughlin <markmc@redhat.com>

Add module aliases to support autoprobing modules
for xen frontend devices.

Signed-off-by: Mark McLoughlin <markmc@redhat.com>
---
 drivers/block/xen-blkfront.c      |    1 +
 drivers/net/xen-netfront.c        |    1 +
 drivers/xen/xenbus/xenbus_probe.c |   28 +++++++++++++++++++++++++---
 3 files changed, 27 insertions(+), 3 deletions(-)

diff --git a/drivers/block/xen-blkfront.c b/drivers/block/xen-blkfront.c
--- a/drivers/block/xen-blkfront.c
+++ b/drivers/block/xen-blkfront.c
@@ -1015,3 +1015,4 @@
 MODULE_DESCRIPTION("Xen virtual block device frontend");
 MODULE_LICENSE("GPL");
 MODULE_ALIAS_BLOCKDEV_MAJOR(XENVBD_MAJOR);
+MODULE_ALIAS("xen:vbd");
diff --git a/drivers/net/xen-netfront.c b/drivers/net/xen-netfront.c
--- a/drivers/net/xen-netfront.c
+++ b/drivers/net/xen-netfront.c
@@ -1809,3 +1809,4 @@
 
 MODULE_DESCRIPTION("Xen virtual network device frontend");
 MODULE_LICENSE("GPL");
+MODULE_ALIAS("xen:vif");
diff --git a/drivers/xen/xenbus/xenbus_probe.c b/drivers/xen/xenbus/xenbus_probe.c
--- a/drivers/xen/xenbus/xenbus_probe.c
+++ b/drivers/xen/xenbus/xenbus_probe.c
@@ -88,6 +88,16 @@
 	return match_device(drv->ids, to_xenbus_device(_dev)) != NULL;
 }
 
+static int xenbus_uevent(struct device *_dev, struct kobj_uevent_env *env)
+{
+	struct xenbus_device *dev = to_xenbus_device(_dev);
+
+	if (add_uevent_var(env, "MODALIAS=xen:%s", dev->devicetype))
+		return -ENOMEM;
+
+	return 0;
+}
+
 /* device/<type>/<id> => <type>-<id> */
 static int frontend_bus_id(char bus_id[BUS_ID_SIZE], const char *nodename)
 {
@@ -166,6 +176,7 @@
 	.bus = {
 		.name     = "xen",
 		.match    = xenbus_match,
+		.uevent   = xenbus_uevent,
 		.probe    = xenbus_dev_probe,
 		.remove   = xenbus_dev_remove,
 		.shutdown = xenbus_dev_shutdown,
@@ -438,6 +449,12 @@
 }
 DEVICE_ATTR(devtype, S_IRUSR | S_IRGRP | S_IROTH, xendev_show_devtype, NULL);
 
+static ssize_t xendev_show_modalias(struct device *dev,
+				    struct device_attribute *attr, char *buf)
+{
+	return sprintf(buf, "xen:%s\n", to_xenbus_device(dev)->devicetype);
+}
+DEVICE_ATTR(modalias, S_IRUSR | S_IRGRP | S_IROTH, xendev_show_modalias, NULL);
 
 int xenbus_probe_node(struct xen_bus_type *bus,
 		      const char *type,
@@ -492,10 +509,16 @@
 
 	err = device_create_file(&xendev->dev, &dev_attr_devtype);
 	if (err)
-		goto fail_remove_file;
+		goto fail_remove_nodename;
+
+	err = device_create_file(&xendev->dev, &dev_attr_modalias);
+	if (err)
+		goto fail_remove_devtype;
 
 	return 0;
-fail_remove_file:
+fail_remove_devtype:
+	device_remove_file(&xendev->dev, &dev_attr_devtype);
+fail_remove_nodename:
 	device_remove_file(&xendev->dev, &dev_attr_nodename);
 fail_unregister:
 	device_unregister(&xendev->dev);

[PATCH 17 of 24] xen: Add compatibility aliases for frontend drivers

[Jeremy Fitzhardinge]

From: Mark McLoughlin <markmc@redhat.com>

Before getting merged, xen-blkfront was xenblk and
xen-netfront was xennet.

Temporarily adding compatibility module aliases
eases upgrades from older versions by e.g. allowing
mkinitrd to find the new version of the module.

Signed-off-by: Mark McLoughlin <markmc@redhat.com>
---
 drivers/block/xen-blkfront.c |    1 +
 drivers/net/xen-netfront.c   |    2 +-
 2 files changed, 2 insertions(+), 1 deletion(-)

diff --git a/drivers/block/xen-blkfront.c b/drivers/block/xen-blkfront.c
--- a/drivers/block/xen-blkfront.c
+++ b/drivers/block/xen-blkfront.c
@@ -1016,3 +1016,4 @@
 MODULE_LICENSE("GPL");
 MODULE_ALIAS_BLOCKDEV_MAJOR(XENVBD_MAJOR);
 MODULE_ALIAS("xen:vbd");
+MODULE_ALIAS("xenblk");
diff --git a/drivers/net/xen-netfront.c b/drivers/net/xen-netfront.c
--- a/drivers/net/xen-netfront.c
+++ b/drivers/net/xen-netfront.c
@@ -1810,3 +1810,4 @@
 MODULE_DESCRIPTION("Xen virtual network device frontend");
 MODULE_LICENSE("GPL");
 MODULE_ALIAS("xen:vif");
+MODULE_ALIAS("xennet");

[PATCH 18 of 24] xen pvfb: Para-virtual framebuffer, keyboard and pointer driver

[Jeremy Fitzhardinge]

From: Markus Armbruster <armbru@redhat.com>

This is a pair of Xen para-virtual frontend device drivers:
drivers/video/xen-fbfront.c provides a framebuffer, and
drivers/input/xen-kbdfront provides keyboard and mouse.

The backends run in dom0 user space.

The two drivers are not in two separate patches, because the
intermediate step (one driver, not the other) is somewhat problematic:
the backend in dom0 needs both drivers, and will refuse to complete
device initialization unless they're both present.

Signed-off-by: Markus Armbruster <armbru@redhat.com>

---
 drivers/input/Kconfig            |    9
 drivers/input/Makefile           |    2
 drivers/input/xen-kbdfront.c     |  340 +++++++++++++++++++++++
 drivers/video/Kconfig            |   14
 drivers/video/Makefile           |    1
 drivers/video/xen-fbfront.c      |  550 ++++++++++++++++++++++++++++++++++++++
 include/xen/interface/io/fbif.h  |  124 ++++++++
 include/xen/interface/io/kbdif.h |  114 +++++++
 8 files changed, 1154 insertions(+)

diff --git a/drivers/input/Kconfig b/drivers/input/Kconfig
--- a/drivers/input/Kconfig
+++ b/drivers/input/Kconfig
@@ -149,6 +149,15 @@
 	  To compile this driver as a module, choose M here: the
 	  module will be called apm-power.
 
+config XEN_KBDDEV_FRONTEND
+	tristate "Xen virtual keyboard and mouse support"
+	depends on XEN_FBDEV_FRONTEND
+	default y
+	help
+	  This driver implements the front-end of the Xen virtual
+	  keyboard and mouse device driver.  It communicates with a back-end
+	  in another domain.
+
 comment "Input Device Drivers"
 
 source "drivers/input/keyboard/Kconfig"
diff --git a/drivers/input/Makefile b/drivers/input/Makefile
--- a/drivers/input/Makefile
+++ b/drivers/input/Makefile
@@ -23,3 +23,5 @@
 obj-$(CONFIG_INPUT_MISC)	+= misc/
 
 obj-$(CONFIG_INPUT_APMPOWER)	+= apm-power.o
+
+obj-$(CONFIG_XEN_KBDDEV_FRONTEND)	+= xen-kbdfront.o
diff --git a/drivers/input/xen-kbdfront.c b/drivers/input/xen-kbdfront.c
new file mode 100644
--- /dev/null
+++ b/drivers/input/xen-kbdfront.c
@@ -0,0 +1,340 @@
+/*
+ * Xen para-virtual input device
+ *
+ * Copyright (C) 2005 Anthony Liguori <aliguori@us.ibm.com>
+ * Copyright (C) 2006-2008 Red Hat, Inc., Markus Armbruster <armbru@redhat.com>
+ *
+ *  Based on linux/drivers/input/mouse/sermouse.c
+ *
+ *  This file is subject to the terms and conditions of the GNU General Public
+ *  License. See the file COPYING in the main directory of this archive for
+ *  more details.
+ */
+
+/*
+ * TODO:
+ *
+ * Switch to grant tables together with xen-fbfront.c.
+ */
+
+#include <linux/kernel.h>
+#include <linux/errno.h>
+#include <linux/module.h>
+#include <linux/input.h>
+#include <asm/xen/hypervisor.h>
+#include <xen/events.h>
+#include <xen/page.h>
+#include <xen/interface/io/fbif.h>
+#include <xen/interface/io/kbdif.h>
+#include <xen/xenbus.h>
+
+struct xenkbd_info {
+	struct input_dev *kbd;
+	struct input_dev *ptr;
+	struct xenkbd_page *page;
+	int irq;
+	struct xenbus_device *xbdev;
+	char phys[32];
+};
+
+static int xenkbd_remove(struct xenbus_device *);
+static int xenkbd_connect_backend(struct xenbus_device *, struct xenkbd_info *);
+static void xenkbd_disconnect_backend(struct xenkbd_info *);
+
+/*
+ * Note: if you need to send out events, see xenfb_do_update() for how
+ * to do that.
+ */
+
+static irqreturn_t input_handler(int rq, void *dev_id)
+{
+	struct xenkbd_info *info = dev_id;
+	struct xenkbd_page *page = info->page;
+	__u32 cons, prod;
+
+	prod = page->in_prod;
+	if (prod == page->in_cons)
+		return IRQ_HANDLED;
+	rmb();			/* ensure we see ring contents up to prod */
+	for (cons = page->in_cons; cons != prod; cons++) {
+		union xenkbd_in_event *event;
+		struct input_dev *dev;
+		event = &XENKBD_IN_RING_REF(page, cons);
+
+		dev = info->ptr;
+		switch (event->type) {
+		case XENKBD_TYPE_MOTION:
+			input_report_rel(dev, REL_X, event->motion.rel_x);
+			input_report_rel(dev, REL_Y, event->motion.rel_y);
+			break;
+		case XENKBD_TYPE_KEY:
+			dev = NULL;
+			if (test_bit(event->key.keycode, info->kbd->keybit))
+				dev = info->kbd;
+			if (test_bit(event->key.keycode, info->ptr->keybit))
+				dev = info->ptr;
+			if (dev)
+				input_report_key(dev, event->key.keycode,
+						 event->key.pressed);
+			else
+				printk(KERN_WARNING
+				       "xenkbd: unhandled keycode 0x%x\n",
+				       event->key.keycode);
+			break;
+		case XENKBD_TYPE_POS:
+			input_report_abs(dev, ABS_X, event->pos.abs_x);
+			input_report_abs(dev, ABS_Y, event->pos.abs_y);
+			break;
+		}
+		if (dev)
+			input_sync(dev);
+	}
+	mb();			/* ensure we got ring contents */
+	page->in_cons = cons;
+	notify_remote_via_irq(info->irq);
+
+	return IRQ_HANDLED;
+}
+
+static int __devinit xenkbd_probe(struct xenbus_device *dev,
+				  const struct xenbus_device_id *id)
+{
+	int ret, i;
+	struct xenkbd_info *info;
+	struct input_dev *kbd, *ptr;
+
+	info = kzalloc(sizeof(*info), GFP_KERNEL);
+	if (!info) {
+		xenbus_dev_fatal(dev, -ENOMEM, "allocating info structure");
+		return -ENOMEM;
+	}
+	dev->dev.driver_data = info;
+	info->xbdev = dev;
+	info->irq = -1;
+	snprintf(info->phys, sizeof(info->phys), "xenbus/%s", dev->nodename);
+
+	info->page = (void *)__get_free_page(GFP_KERNEL | __GFP_ZERO);
+	if (!info->page)
+		goto error_nomem;
+
+	/* keyboard */
+	kbd = input_allocate_device();
+	if (!kbd)
+		goto error_nomem;
+	kbd->name = "Xen Virtual Keyboard";
+	kbd->phys = info->phys;
+	kbd->id.bustype = BUS_PCI;
+	kbd->id.vendor = 0x5853;
+	kbd->id.product = 0xffff;
+	kbd->evbit[0] = BIT(EV_KEY);
+	for (i = KEY_ESC; i < KEY_UNKNOWN; i++)
+		set_bit(i, kbd->keybit);
+	for (i = KEY_OK; i < KEY_MAX; i++)
+		set_bit(i, kbd->keybit);
+
+	ret = input_register_device(kbd);
+	if (ret) {
+		input_free_device(kbd);
+		xenbus_dev_fatal(dev, ret, "input_register_device(kbd)");
+		goto error;
+	}
+	info->kbd = kbd;
+
+	/* pointing device */
+	ptr = input_allocate_device();
+	if (!ptr)
+		goto error_nomem;
+	ptr->name = "Xen Virtual Pointer";
+	ptr->phys = info->phys;
+	ptr->id.bustype = BUS_PCI;
+	ptr->id.vendor = 0x5853;
+	ptr->id.product = 0xfffe;
+	ptr->evbit[0] = BIT(EV_KEY) | BIT(EV_REL) | BIT(EV_ABS);
+	for (i = BTN_LEFT; i <= BTN_TASK; i++)
+		set_bit(i, ptr->keybit);
+	ptr->relbit[0] = BIT(REL_X) | BIT(REL_Y);
+	input_set_abs_params(ptr, ABS_X, 0, XENFB_WIDTH, 0, 0);
+	input_set_abs_params(ptr, ABS_Y, 0, XENFB_HEIGHT, 0, 0);
+
+	ret = input_register_device(ptr);
+	if (ret) {
+		input_free_device(ptr);
+		xenbus_dev_fatal(dev, ret, "input_register_device(ptr)");
+		goto error;
+	}
+	info->ptr = ptr;
+
+	ret = xenkbd_connect_backend(dev, info);
+	if (ret < 0)
+		goto error;
+
+	return 0;
+
+ error_nomem:
+	ret = -ENOMEM;
+	xenbus_dev_fatal(dev, ret, "allocating device memory");
+ error:
+	xenkbd_remove(dev);
+	return ret;
+}
+
+static int xenkbd_resume(struct xenbus_device *dev)
+{
+	struct xenkbd_info *info = dev->dev.driver_data;
+
+	xenkbd_disconnect_backend(info);
+	memset(info->page, 0, PAGE_SIZE);
+	return xenkbd_connect_backend(dev, info);
+}
+
+static int xenkbd_remove(struct xenbus_device *dev)
+{
+	struct xenkbd_info *info = dev->dev.driver_data;
+
+	xenkbd_disconnect_backend(info);
+	if (info->kbd)
+		input_unregister_device(info->kbd);
+	if (info->ptr)
+		input_unregister_device(info->ptr);
+	free_page((unsigned long)info->page);
+	kfree(info);
+	return 0;
+}
+
+static int xenkbd_connect_backend(struct xenbus_device *dev,
+				  struct xenkbd_info *info)
+{
+	int ret, evtchn;
+	struct xenbus_transaction xbt;
+
+	ret = xenbus_alloc_evtchn(dev, &evtchn);
+	if (ret)
+		return ret;
+	ret = bind_evtchn_to_irqhandler(evtchn, input_handler,
+					0, dev->devicetype, info);
+	if (ret < 0) {
+		xenbus_free_evtchn(dev, evtchn);
+		xenbus_dev_fatal(dev, ret, "bind_evtchn_to_irqhandler");
+		return ret;
+	}
+	info->irq = ret;
+
+ again:
+	ret = xenbus_transaction_start(&xbt);
+	if (ret) {
+		xenbus_dev_fatal(dev, ret, "starting transaction");
+		return ret;
+	}
+	ret = xenbus_printf(xbt, dev->nodename, "page-ref", "%lu",
+			    virt_to_mfn(info->page));
+	if (ret)
+		goto error_xenbus;
+	ret = xenbus_printf(xbt, dev->nodename, "event-channel", "%u",
+			    evtchn);
+	if (ret)
+		goto error_xenbus;
+	ret = xenbus_transaction_end(xbt, 0);
+	if (ret) {
+		if (ret == -EAGAIN)
+			goto again;
+		xenbus_dev_fatal(dev, ret, "completing transaction");
+		return ret;
+	}
+
+	xenbus_switch_state(dev, XenbusStateInitialised);
+	return 0;
+
+ error_xenbus:
+	xenbus_transaction_end(xbt, 1);
+	xenbus_dev_fatal(dev, ret, "writing xenstore");
+	return ret;
+}
+
+static void xenkbd_disconnect_backend(struct xenkbd_info *info)
+{
+	if (info->irq >= 0)
+		unbind_from_irqhandler(info->irq, info);
+	info->irq = -1;
+}
+
+static void xenkbd_backend_changed(struct xenbus_device *dev,
+				   enum xenbus_state backend_state)
+{
+	struct xenkbd_info *info = dev->dev.driver_data;
+	int ret, val;
+
+	switch (backend_state) {
+	case XenbusStateInitialising:
+	case XenbusStateInitialised:
+	case XenbusStateUnknown:
+	case XenbusStateClosed:
+		break;
+
+	case XenbusStateInitWait:
+InitWait:
+		ret = xenbus_scanf(XBT_NIL, info->xbdev->otherend,
+				   "feature-abs-pointer", "%d", &val);
+		if (ret < 0)
+			val = 0;
+		if (val) {
+			ret = xenbus_printf(XBT_NIL, info->xbdev->nodename,
+					    "request-abs-pointer", "1");
+			if (ret)
+				printk(KERN_WARNING
+				       "xenkbd: can't request abs-pointer");
+		}
+		xenbus_switch_state(dev, XenbusStateConnected);
+		break;
+
+	case XenbusStateConnected:
+		/*
+		 * Work around xenbus race condition: If backend goes
+		 * through InitWait to Connected fast enough, we can
+		 * get Connected twice here.
+		 */
+		if (dev->state != XenbusStateConnected)
+			goto InitWait; /* no InitWait seen yet, fudge it */
+		break;
+
+	case XenbusStateClosing:
+		xenbus_frontend_closed(dev);
+		break;
+	}
+}
+
+static struct xenbus_device_id xenkbd_ids[] = {
+	{ "vkbd" },
+	{ "" }
+};
+
+static struct xenbus_driver xenkbd = {
+	.name = "vkbd",
+	.owner = THIS_MODULE,
+	.ids = xenkbd_ids,
+	.probe = xenkbd_probe,
+	.remove = xenkbd_remove,
+	.resume = xenkbd_resume,
+	.otherend_changed = xenkbd_backend_changed,
+};
+
+static int __init xenkbd_init(void)
+{
+	if (!is_running_on_xen())
+		return -ENODEV;
+
+	/* Nothing to do if running in dom0. */
+	if (is_initial_xendomain())
+		return -ENODEV;
+
+	return xenbus_register_frontend(&xenkbd);
+}
+
+static void __exit xenkbd_cleanup(void)
+{
+	xenbus_unregister_driver(&xenkbd);
+}
+
+module_init(xenkbd_init);
+module_exit(xenkbd_cleanup);
+
+MODULE_LICENSE("GPL");
diff --git a/drivers/video/Kconfig b/drivers/video/Kconfig
--- a/drivers/video/Kconfig
+++ b/drivers/video/Kconfig
@@ -1930,6 +1930,20 @@
 
 	  If unsure, say N.
 
+config XEN_FBDEV_FRONTEND
+	tristate "Xen virtual frame buffer support"
+	depends on FB && XEN
+	select FB_SYS_FILLRECT
+	select FB_SYS_COPYAREA
+	select FB_SYS_IMAGEBLIT
+	select FB_SYS_FOPS
+	select FB_DEFERRED_IO
+	default y
+	help
+	  This driver implements the front-end of the Xen virtual
+	  frame buffer driver.  It communicates with a back-end
+	  in another domain.
+
 source "drivers/video/omap/Kconfig"
 
 source "drivers/video/backlight/Kconfig"
diff --git a/drivers/video/Makefile b/drivers/video/Makefile
--- a/drivers/video/Makefile
+++ b/drivers/video/Makefile
@@ -114,6 +114,7 @@
 obj-$(CONFIG_FB_SM501)            += sm501fb.o
 obj-$(CONFIG_FB_XILINX)           += xilinxfb.o
 obj-$(CONFIG_FB_OMAP)             += omap/
+obj-$(CONFIG_XEN_FBDEV_FRONTEND)  += xen-fbfront.o
 
 # Platform or fallback drivers go here
 obj-$(CONFIG_FB_UVESA)            += uvesafb.o
diff --git a/drivers/video/xen-fbfront.c b/drivers/video/xen-fbfront.c
new file mode 100644
--- /dev/null
+++ b/drivers/video/xen-fbfront.c
@@ -0,0 +1,550 @@
+/*
+ * Xen para-virtual frame buffer device
+ *
+ * Copyright (C) 2005-2006 Anthony Liguori <aliguori@us.ibm.com>
+ * Copyright (C) 2006-2008 Red Hat, Inc., Markus Armbruster <armbru@redhat.com>
+ *
+ *  Based on linux/drivers/video/q40fb.c
+ *
+ *  This file is subject to the terms and conditions of the GNU General Public
+ *  License. See the file COPYING in the main directory of this archive for
+ *  more details.
+ */
+
+/*
+ * TODO:
+ *
+ * Switch to grant tables when they become capable of dealing with the
+ * frame buffer.
+ */
+
+#include <linux/kernel.h>
+#include <linux/errno.h>
+#include <linux/fb.h>
+#include <linux/module.h>
+#include <linux/vmalloc.h>
+#include <linux/mm.h>
+#include <asm/xen/hypervisor.h>
+#include <xen/events.h>
+#include <xen/page.h>
+#include <xen/interface/io/fbif.h>
+#include <xen/interface/io/protocols.h>
+#include <xen/xenbus.h>
+
+struct xenfb_info {
+	unsigned char		*fb;
+	struct fb_info		*fb_info;
+	int			x1, y1, x2, y2;	/* dirty rectangle,
+						   protected by dirty_lock */
+	spinlock_t		dirty_lock;
+	int			nr_pages;
+	int			irq;
+	struct xenfb_page	*page;
+	unsigned long 		*mfns;
+	int			update_wanted; /* XENFB_TYPE_UPDATE wanted */
+
+	struct xenbus_device	*xbdev;
+};
+
+static u32 xenfb_mem_len = XENFB_WIDTH * XENFB_HEIGHT * XENFB_DEPTH / 8;
+
+static int xenfb_remove(struct xenbus_device *);
+static void xenfb_init_shared_page(struct xenfb_info *);
+static int xenfb_connect_backend(struct xenbus_device *, struct xenfb_info *);
+static void xenfb_disconnect_backend(struct xenfb_info *);
+
+static void xenfb_do_update(struct xenfb_info *info,
+			    int x, int y, int w, int h)
+{
+	union xenfb_out_event event;
+	u32 prod;
+
+	event.type = XENFB_TYPE_UPDATE;
+	event.update.x = x;
+	event.update.y = y;
+	event.update.width = w;
+	event.update.height = h;
+
+	prod = info->page->out_prod;
+	/* caller ensures !xenfb_queue_full() */
+	mb();			/* ensure ring space available */
+	XENFB_OUT_RING_REF(info->page, prod) = event;
+	wmb();			/* ensure ring contents visible */
+	info->page->out_prod = prod + 1;
+
+	notify_remote_via_irq(info->irq);
+}
+
+static int xenfb_queue_full(struct xenfb_info *info)
+{
+	u32 cons, prod;
+
+	prod = info->page->out_prod;
+	cons = info->page->out_cons;
+	return prod - cons == XENFB_OUT_RING_LEN;
+}
+
+static void xenfb_refresh(struct xenfb_info *info,
+			  int x1, int y1, int w, int h)
+{
+	unsigned long flags;
+	int y2 = y1 + h - 1;
+	int x2 = x1 + w - 1;
+
+	if (!info->update_wanted)
+		return;
+
+	spin_lock_irqsave(&info->dirty_lock, flags);
+
+	/* Combine with dirty rectangle: */
+	if (info->y1 < y1)
+		y1 = info->y1;
+	if (info->y2 > y2)
+		y2 = info->y2;
+	if (info->x1 < x1)
+		x1 = info->x1;
+	if (info->x2 > x2)
+		x2 = info->x2;
+
+	if (xenfb_queue_full(info)) {
+		/* Can't send right now, stash it in the dirty rectangle */
+		info->x1 = x1;
+		info->x2 = x2;
+		info->y1 = y1;
+		info->y2 = y2;
+		spin_unlock_irqrestore(&info->dirty_lock, flags);
+		return;
+	}
+
+	/* Clear dirty rectangle: */
+	info->x1 = info->y1 = INT_MAX;
+	info->x2 = info->y2 = 0;
+
+	spin_unlock_irqrestore(&info->dirty_lock, flags);
+
+	if (x1 <= x2 && y1 <= y2)
+		xenfb_do_update(info, x1, y1, x2 - x1 + 1, y2 - y1 + 1);
+}
+
+static void xenfb_deferred_io(struct fb_info *fb_info,
+			      struct list_head *pagelist)
+{
+	struct xenfb_info *info = fb_info->par;
+	struct page *page;
+	unsigned long beg, end;
+	int y1, y2, miny, maxy;
+
+	miny = INT_MAX;
+	maxy = 0;
+	list_for_each_entry(page, pagelist, lru) {
+		beg = page->index << PAGE_SHIFT;
+		end = beg + PAGE_SIZE - 1;
+		y1 = beg / fb_info->fix.line_length;
+		y2 = end / fb_info->fix.line_length;
+		if (y2 >= fb_info->var.yres)
+			y2 = fb_info->var.yres - 1;
+		if (miny > y1)
+			miny = y1;
+		if (maxy < y2)
+			maxy = y2;
+	}
+	xenfb_refresh(info, 0, miny, fb_info->var.xres, maxy - miny + 1);
+}
+
+static struct fb_deferred_io xenfb_defio = {
+	.delay		= HZ / 20,
+	.deferred_io	= xenfb_deferred_io,
+};
+
+static int xenfb_setcolreg(unsigned regno, unsigned red, unsigned green,
+			   unsigned blue, unsigned transp,
+			   struct fb_info *info)
+{
+	u32 v;
+
+	if (regno > info->cmap.len)
+		return 1;
+
+#define CNVT_TOHW(val, width) ((((val)<<(width))+0x7FFF-(val))>>16)
+	red = CNVT_TOHW(red, info->var.red.length);
+	green = CNVT_TOHW(green, info->var.green.length);
+	blue = CNVT_TOHW(blue, info->var.blue.length);
+	transp = CNVT_TOHW(transp, info->var.transp.length);
+#undef CNVT_TOHW
+
+	v = (red << info->var.red.offset) |
+	    (green << info->var.green.offset) |
+	    (blue << info->var.blue.offset);
+
+	switch (info->var.bits_per_pixel) {
+	case 16:
+	case 24:
+	case 32:
+		((u32 *)info->pseudo_palette)[regno] = v;
+		break;
+	}
+
+	return 0;
+}
+
+static void xenfb_fillrect(struct fb_info *p, const struct fb_fillrect *rect)
+{
+	struct xenfb_info *info = p->par;
+
+	sys_fillrect(p, rect);
+	xenfb_refresh(info, rect->dx, rect->dy, rect->width, rect->height);
+}
+
+static void xenfb_imageblit(struct fb_info *p, const struct fb_image *image)
+{
+	struct xenfb_info *info = p->par;
+
+	sys_imageblit(p, image);
+	xenfb_refresh(info, image->dx, image->dy, image->width, image->height);
+}
+
+static void xenfb_copyarea(struct fb_info *p, const struct fb_copyarea *area)
+{
+	struct xenfb_info *info = p->par;
+
+	sys_copyarea(p, area);
+	xenfb_refresh(info, area->dx, area->dy, area->width, area->height);
+}
+
+static ssize_t xenfb_write(struct fb_info *p, const char __user *buf,
+			size_t count, loff_t *ppos)
+{
+	struct xenfb_info *info = p->par;
+	ssize_t res;
+
+	res = fb_sys_write(p, buf, count, ppos);
+	xenfb_refresh(info, 0, 0, info->page->width, info->page->height);
+	return res;
+}
+
+static struct fb_ops xenfb_fb_ops = {
+	.owner		= THIS_MODULE,
+	.fb_read	= fb_sys_read,
+	.fb_write	= xenfb_write,
+	.fb_setcolreg	= xenfb_setcolreg,
+	.fb_fillrect	= xenfb_fillrect,
+	.fb_copyarea	= xenfb_copyarea,
+	.fb_imageblit	= xenfb_imageblit,
+};
+
+static irqreturn_t xenfb_event_handler(int rq, void *dev_id)
+{
+	/*
+	 * No in events recognized, simply ignore them all.
+	 * If you need to recognize some, see xen-kbdfront's
+	 * input_handler() for how to do that.
+	 */
+	struct xenfb_info *info = dev_id;
+	struct xenfb_page *page = info->page;
+
+	if (page->in_cons != page->in_prod) {
+		info->page->in_cons = info->page->in_prod;
+		notify_remote_via_irq(info->irq);
+	}
+
+	/* Flush dirty rectangle: */
+	xenfb_refresh(info, INT_MAX, INT_MAX, -INT_MAX, -INT_MAX);
+
+	return IRQ_HANDLED;
+}
+
+static int __devinit xenfb_probe(struct xenbus_device *dev,
+				 const struct xenbus_device_id *id)
+{
+	struct xenfb_info *info;
+	struct fb_info *fb_info;
+	int ret;
+
+	info = kzalloc(sizeof(*info), GFP_KERNEL);
+	if (info == NULL) {
+		xenbus_dev_fatal(dev, -ENOMEM, "allocating info structure");
+		return -ENOMEM;
+	}
+	dev->dev.driver_data = info;
+	info->xbdev = dev;
+	info->irq = -1;
+	info->x1 = info->y1 = INT_MAX;
+	spin_lock_init(&info->dirty_lock);
+
+	info->fb = vmalloc(xenfb_mem_len);
+	if (info->fb == NULL)
+		goto error_nomem;
+	memset(info->fb, 0, xenfb_mem_len);
+
+	info->nr_pages = (xenfb_mem_len + PAGE_SIZE - 1) >> PAGE_SHIFT;
+
+	info->mfns = vmalloc(sizeof(unsigned long) * info->nr_pages);
+	if (!info->mfns)
+		goto error_nomem;
+
+	/* set up shared page */
+	info->page = (void *)__get_free_page(GFP_KERNEL | __GFP_ZERO);
+	if (!info->page)
+		goto error_nomem;
+
+	xenfb_init_shared_page(info);
+
+	/* abusing framebuffer_alloc() to allocate pseudo_palette */
+	fb_info = framebuffer_alloc(sizeof(u32) * 256, NULL);
+	if (fb_info == NULL)
+		goto error_nomem;
+
+	/* complete the abuse: */
+	fb_info->pseudo_palette = fb_info->par;
+	fb_info->par = info;
+
+	fb_info->screen_base = info->fb;
+
+	fb_info->fbops = &xenfb_fb_ops;
+	fb_info->var.xres_virtual = fb_info->var.xres = info->page->width;
+	fb_info->var.yres_virtual = fb_info->var.yres = info->page->height;
+	fb_info->var.bits_per_pixel = info->page->depth;
+
+	fb_info->var.red = (struct fb_bitfield){16, 8, 0};
+	fb_info->var.green = (struct fb_bitfield){8, 8, 0};
+	fb_info->var.blue = (struct fb_bitfield){0, 8, 0};
+
+	fb_info->var.activate = FB_ACTIVATE_NOW;
+	fb_info->var.height = -1;
+	fb_info->var.width = -1;
+	fb_info->var.vmode = FB_VMODE_NONINTERLACED;
+
+	fb_info->fix.visual = FB_VISUAL_TRUECOLOR;
+	fb_info->fix.line_length = info->page->line_length;
+	fb_info->fix.smem_start = 0;
+	fb_info->fix.smem_len = xenfb_mem_len;
+	strcpy(fb_info->fix.id, "xen");
+	fb_info->fix.type = FB_TYPE_PACKED_PIXELS;
+	fb_info->fix.accel = FB_ACCEL_NONE;
+
+	fb_info->flags = FBINFO_FLAG_DEFAULT;
+
+	ret = fb_alloc_cmap(&fb_info->cmap, 256, 0);
+	if (ret < 0) {
+		framebuffer_release(fb_info);
+		xenbus_dev_fatal(dev, ret, "fb_alloc_cmap");
+		goto error;
+	}
+
+	fb_info->fbdefio = &xenfb_defio;
+	fb_deferred_io_init(fb_info);
+
+	ret = register_framebuffer(fb_info);
+	if (ret) {
+		fb_deferred_io_cleanup(fb_info);
+		fb_dealloc_cmap(&fb_info->cmap);
+		framebuffer_release(fb_info);
+		xenbus_dev_fatal(dev, ret, "register_framebuffer");
+		goto error;
+	}
+	info->fb_info = fb_info;
+
+	ret = xenfb_connect_backend(dev, info);
+	if (ret < 0)
+		goto error;
+
+	return 0;
+
+ error_nomem:
+	ret = -ENOMEM;
+	xenbus_dev_fatal(dev, ret, "allocating device memory");
+ error:
+	xenfb_remove(dev);
+	return ret;
+}
+
+static int xenfb_resume(struct xenbus_device *dev)
+{
+	struct xenfb_info *info = dev->dev.driver_data;
+
+	xenfb_disconnect_backend(info);
+	xenfb_init_shared_page(info);
+	return xenfb_connect_backend(dev, info);
+}
+
+static int xenfb_remove(struct xenbus_device *dev)
+{
+	struct xenfb_info *info = dev->dev.driver_data;
+
+	xenfb_disconnect_backend(info);
+	if (info->fb_info) {
+		fb_deferred_io_cleanup(info->fb_info);
+		unregister_framebuffer(info->fb_info);
+		fb_dealloc_cmap(&info->fb_info->cmap);
+		framebuffer_release(info->fb_info);
+	}
+	free_page((unsigned long)info->page);
+	vfree(info->mfns);
+	vfree(info->fb);
+	kfree(info);
+
+	return 0;
+}
+
+static unsigned long vmalloc_to_mfn(void *address)
+{
+	return pfn_to_mfn(vmalloc_to_pfn(address));
+}
+
+static void xenfb_init_shared_page(struct xenfb_info *info)
+{
+	int i;
+
+	for (i = 0; i < info->nr_pages; i++)
+		info->mfns[i] = vmalloc_to_mfn(info->fb + i * PAGE_SIZE);
+
+	info->page->pd[0] = vmalloc_to_mfn(info->mfns);
+	info->page->pd[1] = 0;
+	info->page->width = XENFB_WIDTH;
+	info->page->height = XENFB_HEIGHT;
+	info->page->depth = XENFB_DEPTH;
+	info->page->line_length = (info->page->depth / 8) * info->page->width;
+	info->page->mem_length = xenfb_mem_len;
+	info->page->in_cons = info->page->in_prod = 0;
+	info->page->out_cons = info->page->out_prod = 0;
+}
+
+static int xenfb_connect_backend(struct xenbus_device *dev,
+				 struct xenfb_info *info)
+{
+	int ret, evtchn;
+	struct xenbus_transaction xbt;
+
+	ret = xenbus_alloc_evtchn(dev, &evtchn);
+	if (ret)
+		return ret;
+	ret = bind_evtchn_to_irqhandler(evtchn, xenfb_event_handler,
+					0, dev->devicetype, info);
+	if (ret < 0) {
+		xenbus_free_evtchn(dev, evtchn);
+		xenbus_dev_fatal(dev, ret, "bind_evtchn_to_irqhandler");
+		return ret;
+	}
+	info->irq = ret;
+
+ again:
+	ret = xenbus_transaction_start(&xbt);
+	if (ret) {
+		xenbus_dev_fatal(dev, ret, "starting transaction");
+		return ret;
+	}
+	ret = xenbus_printf(xbt, dev->nodename, "page-ref", "%lu",
+			    virt_to_mfn(info->page));
+	if (ret)
+		goto error_xenbus;
+	ret = xenbus_printf(xbt, dev->nodename, "event-channel", "%u",
+			    evtchn);
+	if (ret)
+		goto error_xenbus;
+	ret = xenbus_printf(xbt, dev->nodename, "protocol", "%s",
+			    XEN_IO_PROTO_ABI_NATIVE);
+	if (ret)
+		goto error_xenbus;
+	ret = xenbus_printf(xbt, dev->nodename, "feature-update", "1");
+	if (ret)
+		goto error_xenbus;
+	ret = xenbus_transaction_end(xbt, 0);
+	if (ret) {
+		if (ret == -EAGAIN)
+			goto again;
+		xenbus_dev_fatal(dev, ret, "completing transaction");
+		return ret;
+	}
+
+	xenbus_switch_state(dev, XenbusStateInitialised);
+	return 0;
+
+ error_xenbus:
+	xenbus_transaction_end(xbt, 1);
+	xenbus_dev_fatal(dev, ret, "writing xenstore");
+	return ret;
+}
+
+static void xenfb_disconnect_backend(struct xenfb_info *info)
+{
+	if (info->irq >= 0)
+		unbind_from_irqhandler(info->irq, info);
+	info->irq = -1;
+}
+
+static void xenfb_backend_changed(struct xenbus_device *dev,
+				  enum xenbus_state backend_state)
+{
+	struct xenfb_info *info = dev->dev.driver_data;
+	int val;
+
+	switch (backend_state) {
+	case XenbusStateInitialising:
+	case XenbusStateInitialised:
+	case XenbusStateUnknown:
+	case XenbusStateClosed:
+		break;
+
+	case XenbusStateInitWait:
+InitWait:
+		xenbus_switch_state(dev, XenbusStateConnected);
+		break;
+
+	case XenbusStateConnected:
+		/*
+		 * Work around xenbus race condition: If backend goes
+		 * through InitWait to Connected fast enough, we can
+		 * get Connected twice here.
+		 */
+		if (dev->state != XenbusStateConnected)
+			goto InitWait; /* no InitWait seen yet, fudge it */
+
+		if (xenbus_scanf(XBT_NIL, info->xbdev->otherend,
+				 "request-update", "%d", &val) < 0)
+			val = 0;
+		if (val)
+			info->update_wanted = 1;
+		break;
+
+	case XenbusStateClosing:
+		xenbus_frontend_closed(dev);
+		break;
+	}
+}
+
+static struct xenbus_device_id xenfb_ids[] = {
+	{ "vfb" },
+	{ "" }
+};
+
+static struct xenbus_driver xenfb = {
+	.name = "vfb",
+	.owner = THIS_MODULE,
+	.ids = xenfb_ids,
+	.probe = xenfb_probe,
+	.remove = xenfb_remove,
+	.resume = xenfb_resume,
+	.otherend_changed = xenfb_backend_changed,
+};
+
+static int __init xenfb_init(void)
+{
+	if (!is_running_on_xen())
+		return -ENODEV;
+
+	/* Nothing to do if running in dom0. */
+	if (is_initial_xendomain())
+		return -ENODEV;
+
+	return xenbus_register_frontend(&xenfb);
+}
+
+static void __exit xenfb_cleanup(void)
+{
+	xenbus_unregister_driver(&xenfb);
+}
+
+module_init(xenfb_init);
+module_exit(xenfb_cleanup);
+
+MODULE_LICENSE("GPL");
diff --git a/include/xen/interface/io/fbif.h b/include/xen/interface/io/fbif.h
new file mode 100644
--- /dev/null
+++ b/include/xen/interface/io/fbif.h
@@ -0,0 +1,124 @@
+/*
+ * fbif.h -- Xen virtual frame buffer device
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a copy
+ * of this software and associated documentation files (the "Software"), to
+ * deal in the Software without restriction, including without limitation the
+ * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or
+ * sell copies of the Software, and to permit persons to whom the Software is
+ * furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
+ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
+ * DEALINGS IN THE SOFTWARE.
+ *
+ * Copyright (C) 2005 Anthony Liguori <aliguori@us.ibm.com>
+ * Copyright (C) 2006 Red Hat, Inc., Markus Armbruster <armbru@redhat.com>
+ */
+
+#ifndef __XEN_PUBLIC_IO_FBIF_H__
+#define __XEN_PUBLIC_IO_FBIF_H__
+
+/* Out events (frontend -> backend) */
+
+/*
+ * Out events may be sent only when requested by backend, and receipt
+ * of an unknown out event is an error.
+ */
+
+/* Event type 1 currently not used */
+/*
+ * Framebuffer update notification event
+ * Capable frontend sets feature-update in xenstore.
+ * Backend requests it by setting request-update in xenstore.
+ */
+#define XENFB_TYPE_UPDATE 2
+
+struct xenfb_update {
+	uint8_t type;		/* XENFB_TYPE_UPDATE */
+	int32_t x;		/* source x */
+	int32_t y;		/* source y */
+	int32_t width;		/* rect width */
+	int32_t height;		/* rect height */
+};
+
+#define XENFB_OUT_EVENT_SIZE 40
+
+union xenfb_out_event {
+	uint8_t type;
+	struct xenfb_update update;
+	char pad[XENFB_OUT_EVENT_SIZE];
+};
+
+/* In events (backend -> frontend) */
+
+/*
+ * Frontends should ignore unknown in events.
+ * No in events currently defined.
+ */
+
+#define XENFB_IN_EVENT_SIZE 40
+
+union xenfb_in_event {
+	uint8_t type;
+	char pad[XENFB_IN_EVENT_SIZE];
+};
+
+/* shared page */
+
+#define XENFB_IN_RING_SIZE 1024
+#define XENFB_IN_RING_LEN (XENFB_IN_RING_SIZE / XENFB_IN_EVENT_SIZE)
+#define XENFB_IN_RING_OFFS 1024
+#define XENFB_IN_RING(page) \
+	((union xenfb_in_event *)((char *)(page) + XENFB_IN_RING_OFFS))
+#define XENFB_IN_RING_REF(page, idx) \
+	(XENFB_IN_RING((page))[(idx) % XENFB_IN_RING_LEN])
+
+#define XENFB_OUT_RING_SIZE 2048
+#define XENFB_OUT_RING_LEN (XENFB_OUT_RING_SIZE / XENFB_OUT_EVENT_SIZE)
+#define XENFB_OUT_RING_OFFS (XENFB_IN_RING_OFFS + XENFB_IN_RING_SIZE)
+#define XENFB_OUT_RING(page) \
+	((union xenfb_out_event *)((char *)(page) + XENFB_OUT_RING_OFFS))
+#define XENFB_OUT_RING_REF(page, idx) \
+	(XENFB_OUT_RING((page))[(idx) % XENFB_OUT_RING_LEN])
+
+struct xenfb_page {
+	uint32_t in_cons, in_prod;
+	uint32_t out_cons, out_prod;
+
+	int32_t width;          /* width of the framebuffer (in pixels) */
+	int32_t height;         /* height of the framebuffer (in pixels) */
+	uint32_t line_length;   /* length of a row of pixels (in bytes) */
+	uint32_t mem_length;    /* length of the framebuffer (in bytes) */
+	uint8_t depth;          /* depth of a pixel (in bits) */
+
+	/*
+	 * Framebuffer page directory
+	 *
+	 * Each directory page holds PAGE_SIZE / sizeof(*pd)
+	 * framebuffer pages, and can thus map up to PAGE_SIZE *
+	 * PAGE_SIZE / sizeof(*pd) bytes.  With PAGE_SIZE == 4096 and
+	 * sizeof(unsigned long) == 4, that's 4 Megs.  Two directory
+	 * pages should be enough for a while.
+	 */
+	unsigned long pd[2];
+};
+
+/*
+ * Wart: xenkbd needs to know resolution.  Put it here until a better
+ * solution is found, but don't leak it to the backend.
+ */
+#ifdef __KERNEL__
+#define XENFB_WIDTH 800
+#define XENFB_HEIGHT 600
+#define XENFB_DEPTH 32
+#endif
+
+#endif
diff --git a/include/xen/interface/io/kbdif.h b/include/xen/interface/io/kbdif.h
new file mode 100644
--- /dev/null
+++ b/include/xen/interface/io/kbdif.h
@@ -0,0 +1,114 @@
+/*
+ * kbdif.h -- Xen virtual keyboard/mouse
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a copy
+ * of this software and associated documentation files (the "Software"), to
+ * deal in the Software without restriction, including without limitation the
+ * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or
+ * sell copies of the Software, and to permit persons to whom the Software is
+ * furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
+ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
+ * DEALINGS IN THE SOFTWARE.
+ *
+ * Copyright (C) 2005 Anthony Liguori <aliguori@us.ibm.com>
+ * Copyright (C) 2006 Red Hat, Inc., Markus Armbruster <armbru@redhat.com>
+ */
+
+#ifndef __XEN_PUBLIC_IO_KBDIF_H__
+#define __XEN_PUBLIC_IO_KBDIF_H__
+
+/* In events (backend -> frontend) */
+
+/*
+ * Frontends should ignore unknown in events.
+ */
+
+/* Pointer movement event */
+#define XENKBD_TYPE_MOTION  1
+/* Event type 2 currently not used */
+/* Key event (includes pointer buttons) */
+#define XENKBD_TYPE_KEY     3
+/*
+ * Pointer position event
+ * Capable backend sets feature-abs-pointer in xenstore.
+ * Frontend requests ot instead of XENKBD_TYPE_MOTION by setting
+ * request-abs-update in xenstore.
+ */
+#define XENKBD_TYPE_POS     4
+
+struct xenkbd_motion {
+	uint8_t type;		/* XENKBD_TYPE_MOTION */
+	int32_t rel_x;		/* relative X motion */
+	int32_t rel_y;		/* relative Y motion */
+};
+
+struct xenkbd_key {
+	uint8_t type;		/* XENKBD_TYPE_KEY */
+	uint8_t pressed;	/* 1 if pressed; 0 otherwise */
+	uint32_t keycode;	/* KEY_* from linux/input.h */
+};
+
+struct xenkbd_position {
+	uint8_t type;		/* XENKBD_TYPE_POS */
+	int32_t abs_x;		/* absolute X position (in FB pixels) */
+	int32_t abs_y;		/* absolute Y position (in FB pixels) */
+};
+
+#define XENKBD_IN_EVENT_SIZE 40
+
+union xenkbd_in_event {
+	uint8_t type;
+	struct xenkbd_motion motion;
+	struct xenkbd_key key;
+	struct xenkbd_position pos;
+	char pad[XENKBD_IN_EVENT_SIZE];
+};
+
+/* Out events (frontend -> backend) */
+
+/*
+ * Out events may be sent only when requested by backend, and receipt
+ * of an unknown out event is an error.
+ * No out events currently defined.
+ */
+
+#define XENKBD_OUT_EVENT_SIZE 40
+
+union xenkbd_out_event {
+	uint8_t type;
+	char pad[XENKBD_OUT_EVENT_SIZE];
+};
+
+/* shared page */
+
+#define XENKBD_IN_RING_SIZE 2048
+#define XENKBD_IN_RING_LEN (XENKBD_IN_RING_SIZE / XENKBD_IN_EVENT_SIZE)
+#define XENKBD_IN_RING_OFFS 1024
+#define XENKBD_IN_RING(page) \
+	((union xenkbd_in_event *)((char *)(page) + XENKBD_IN_RING_OFFS))
+#define XENKBD_IN_RING_REF(page, idx) \
+	(XENKBD_IN_RING((page))[(idx) % XENKBD_IN_RING_LEN])
+
+#define XENKBD_OUT_RING_SIZE 1024
+#define XENKBD_OUT_RING_LEN (XENKBD_OUT_RING_SIZE / XENKBD_OUT_EVENT_SIZE)
+#define XENKBD_OUT_RING_OFFS (XENKBD_IN_RING_OFFS + XENKBD_IN_RING_SIZE)
+#define XENKBD_OUT_RING(page) \
+	((union xenkbd_out_event *)((char *)(page) + XENKBD_OUT_RING_OFFS))
+#define XENKBD_OUT_RING_REF(page, idx) \
+	(XENKBD_OUT_RING((page))[(idx) % XENKBD_OUT_RING_LEN])
+
+struct xenkbd_page {
+	uint32_t in_cons, in_prod;
+	uint32_t out_cons, out_prod;
+};
+
+#endif

[PATCH 19 of 24] x86: fix build problem in pud_populate without CONFIG_PARAVIRT

[Jeremy Fitzhardinge]

[-- Warning: decoded text below may be mangled, UTF-8 assumed --]
[-- Attachment #1: Type: text/plain; charset="us-ascii", Size: 970 bytes --]

asm/paravirt.h ends up including linux/sched.h, which pud_populate needs
for its reference to current.  Specifically include linux/sched.h for it.

In file included from /home/jeremy/hg/xen/paravirt/linux/arch/x86/mm/pgtable.c:3:
include2/asm/pgalloc.h: In function ‘pud_populate’:
include2/asm/pgalloc.h:93: error: dereferencing pointer to incomplete type
make[3]: *** [arch/x86/mm/pgtable.o] Error 1
make[2]: *** [arch/x86/mm/] Error 2
make[1]: *** [sub-make] Error 2
make: *** [all] Error 2

Signed-off-by: Jeremy Fitzhardinge <jeremy.fitzhardinge@citrix.com>
---
 include/asm-x86/pgalloc.h |    1 +
 1 file changed, 1 insertion(+)

diff --git a/include/asm-x86/pgalloc.h b/include/asm-x86/pgalloc.h
--- a/include/asm-x86/pgalloc.h
+++ b/include/asm-x86/pgalloc.h
@@ -4,6 +4,7 @@
 #include <linux/threads.h>
 #include <linux/mm.h>		/* for struct page */
 #include <linux/pagemap.h>
+#include <linux/sched.h>
 
 #ifdef CONFIG_PARAVIRT
 #include <asm/paravirt.h>

[PATCH 20 of 24] xen: disable preemption during tlb flush

[Jeremy Fitzhardinge]

Various places in the kernel flush the tlb even though preemption doens't
guarantee the tlb flush is happening on any particular CPU.  In many cases
this doesn't seem to matter, so don't make a fuss about it.

Signed-off-by: Jeremy Fitzhardinge <jeremy.fitzhardinge@citrix.com>
---
 arch/x86/xen/enlighten.c |   15 +++++++++++++--
 1 file changed, 13 insertions(+), 2 deletions(-)

diff --git a/arch/x86/xen/enlighten.c b/arch/x86/xen/enlighten.c
--- a/arch/x86/xen/enlighten.c
+++ b/arch/x86/xen/enlighten.c
@@ -532,26 +532,37 @@
 static void xen_flush_tlb(void)
 {
 	struct mmuext_op *op;
-	struct multicall_space mcs = xen_mc_entry(sizeof(*op));
+	struct multicall_space mcs;
+
+	preempt_disable();
+
+	mcs = xen_mc_entry(sizeof(*op));
 
 	op = mcs.args;
 	op->cmd = MMUEXT_TLB_FLUSH_LOCAL;
 	MULTI_mmuext_op(mcs.mc, op, 1, NULL, DOMID_SELF);
 
 	xen_mc_issue(PARAVIRT_LAZY_MMU);
+
+	preempt_enable();
 }
 
 static void xen_flush_tlb_single(unsigned long addr)
 {
 	struct mmuext_op *op;
-	struct multicall_space mcs = xen_mc_entry(sizeof(*op));
+	struct multicall_space mcs;
 
+	preempt_disable();
+
+	mcs = xen_mc_entry(sizeof(*op));
 	op = mcs.args;
 	op->cmd = MMUEXT_INVLPG_LOCAL;
 	op->arg1.linear_addr = addr & PAGE_MASK;
 	MULTI_mmuext_op(mcs.mc, op, 1, NULL, DOMID_SELF);
 
 	xen_mc_issue(PARAVIRT_LAZY_MMU);
+
+	preempt_enable();
 }
 
 static void xen_flush_tlb_others(const cpumask_t *cpus, struct mm_struct *mm,

[PATCH 21 of 24] xen: allow set_pte_at on init_mm to be lockless

[Jeremy Fitzhardinge]

The usual pagetable locking protocol doesn't seem to apply to updates
to init_mm, so don't rely on preemption being disabled in xen_set_pte_at
on init_mm.

Signed-off-by: Jeremy Fitzhardinge <jeremy.fitzhardinge@citrix.com>
---
 arch/x86/xen/mmu.c |   12 ++++++++++--
 1 file changed, 10 insertions(+), 2 deletions(-)

diff --git a/arch/x86/xen/mmu.c b/arch/x86/xen/mmu.c
--- a/arch/x86/xen/mmu.c
+++ b/arch/x86/xen/mmu.c
@@ -156,6 +156,10 @@
 void xen_set_pte_at(struct mm_struct *mm, unsigned long addr,
 		    pte_t *ptep, pte_t pteval)
 {
+	/* updates to init_mm may be done without lock */
+	if (mm == &init_mm)
+		preempt_disable();
+
 	if (mm == current->mm || mm == &init_mm) {
 		if (paravirt_get_lazy_mode() == PARAVIRT_LAZY_MMU) {
 			struct multicall_space mcs;
@@ -163,12 +167,16 @@
 
 			MULTI_update_va_mapping(mcs.mc, addr, pteval, 0);
 			xen_mc_issue(PARAVIRT_LAZY_MMU);
-			return;
+			goto out;
 		} else
 			if (HYPERVISOR_update_va_mapping(addr, pteval, 0) == 0)
-				return;
+				goto out;
 	}
 	xen_set_pte(ptep, pteval);
+
+out:
+	if (mm == &init_mm)
+		preempt_enable();
 }
 
 pteval_t xen_pte_val(pte_t pte)

[PATCH 22 of 24] xen: fold xen_sysexit into xen_iret

[Jeremy Fitzhardinge]

xen_sysexit and xen_iret were doing essentially the same thing.  Rather
than having a separate implementation for xen_sysexit, we can just strip
the stack back to an iret frame and jump into xen_iret.  This removes
a lot of code and complexity - specifically, another critical region.

Signed-off-by: Jeremy Fitzhardinge <jeremy.fitzhardinge@citrix.com>
---
 arch/x86/kernel/entry_32.S |    9 -----
 arch/x86/xen/xen-asm.S     |   70 ++++++++------------------------------------
 2 files changed, 15 insertions(+), 64 deletions(-)

diff --git a/arch/x86/kernel/entry_32.S b/arch/x86/kernel/entry_32.S
--- a/arch/x86/kernel/entry_32.S
+++ b/arch/x86/kernel/entry_32.S
@@ -1044,15 +1044,8 @@
 
 	jmp  xen_iret_crit_fixup
 
-1:	cmpl $xen_sysexit_start_crit,%eax
-	jb   2f
-	cmpl $xen_sysexit_end_crit,%eax
-	jae  2f
-
-	jmp xen_sysexit_crit_fixup
-
 ENTRY(xen_do_upcall)
-2:	mov %esp, %eax
+1:	mov %esp, %eax
 	call xen_evtchn_do_upcall
 	jmp  ret_from_intr
 	CFI_ENDPROC
diff --git a/arch/x86/xen/xen-asm.S b/arch/x86/xen/xen-asm.S
--- a/arch/x86/xen/xen-asm.S
+++ b/arch/x86/xen/xen-asm.S
@@ -106,6 +106,20 @@
 	ret
 	ENDPROC(xen_restore_fl_direct)
 	RELOC(xen_restore_fl_direct, 2b+1)
+
+/*
+	We can't use sysexit directly, because we're not running in ring0.
+	But we can easily fake it up using iret.  Assuming xen_sysexit
+	is jumped to with a standard stack frame, we can just strip it
+	back to a standard iret frame and use iret.
+ */
+ENTRY(xen_sysexit)
+	movl PT_EAX(%esp), %eax			/* Shouldn't be necessary? */
+	orl $X86_EFLAGS_IF, PT_EFLAGS(%esp)
+	lea PT_EIP(%esp), %esp
+
+	jmp xen_iret
+ENDPROC(xen_sysexit)
 
 /*
 	This is run where a normal iret would be run, with the same stack setup:
@@ -276,62 +290,6 @@
 2:	jmp xen_do_upcall
 
 
-ENTRY(xen_sysexit)
-	/* Store vcpu_info pointer for easy access.  Do it this
-	   way to avoid having to reload %fs */
-#ifdef CONFIG_SMP
-	GET_THREAD_INFO(%eax)
-	movl TI_cpu(%eax),%eax
-	movl __per_cpu_offset(,%eax,4),%eax
-	mov per_cpu__xen_vcpu(%eax),%eax
-#else
-	movl per_cpu__xen_vcpu, %eax
-#endif
-
-	/* We can't actually use sysexit in a pv guest,
-	   so fake it up with iret */
-	pushl $__USER_DS		/* user stack segment */
-	pushl %ecx			/* user esp */
-	pushl PT_EFLAGS+2*4(%esp)	/* user eflags */
-	pushl $__USER_CS		/* user code segment */
-	pushl %edx			/* user eip */
-
-xen_sysexit_start_crit:
-	/* Unmask events... */
-	movb $0, XEN_vcpu_info_mask(%eax)
-	/* ...and test for pending.
-	   There's a preempt window here, but it doesn't
-	   matter because we're within the critical section. */
-	testb $0xff, XEN_vcpu_info_pending(%eax)
-
-	/* If there's something pending, mask events again so we
-	   can directly inject it back into the kernel. */
-	jnz   1f
-
-	movl PT_EAX+5*4(%esp),%eax
-2:	iret
-1:	movb $1, XEN_vcpu_info_mask(%eax)
-xen_sysexit_end_crit:
-	addl $5*4, %esp		/* remove iret frame */
-	/* no need to re-save regs, but need to restore kernel %fs */
-	mov $__KERNEL_PERCPU, %eax
-	mov %eax, %fs
-	jmp xen_do_upcall
-.section __ex_table,"a"
-	.align 4
-	.long 2b,iret_exc
-.previous
-
-	.globl xen_sysexit_start_crit, xen_sysexit_end_crit
-/*
-	sysexit fixup is easy, since the old frame is still sitting there
-	on the stack.  We just need to remove the new recursive
-	interrupt and return.
- */
-ENTRY(xen_sysexit_crit_fixup)
-	addl $PT_OLDESP+5*4, %esp		/* remove frame+iret */
-	jmp xen_do_upcall
-
 /*
 	Force an event check by making a hypercall,
 	but preserve regs before making the call.

[PATCH 23 of 24] xen: allow compilation with non-flat memory

[Jeremy Fitzhardinge]

There's no real reason we can't support sparsemem/discontigmem, so do so.
This is mostly useful to support hotplug memory.

Signed-off-by: Jeremy Fitzhardinge <jeremy.fitzhardinge@citrix.com>
---
 arch/x86/xen/Kconfig     |    2 +-
 arch/x86/xen/enlighten.c |    2 ++
 2 files changed, 3 insertions(+), 1 deletion(-)

diff --git a/arch/x86/xen/Kconfig b/arch/x86/xen/Kconfig
--- a/arch/x86/xen/Kconfig
+++ b/arch/x86/xen/Kconfig
@@ -6,7 +6,7 @@
 	bool "Xen guest support"
 	select PARAVIRT
 	depends on X86_32
-	depends on X86_CMPXCHG && X86_TSC && !NEED_MULTIPLE_NODES && !(X86_VISWS || X86_VOYAGER)
+	depends on X86_CMPXCHG && X86_TSC && !(X86_VISWS || X86_VOYAGER)
 	help
 	  This is the Linux Xen port.  Enabling this will allow the
 	  kernel to boot in a paravirtualized environment under the
diff --git a/arch/x86/xen/enlighten.c b/arch/x86/xen/enlighten.c
--- a/arch/x86/xen/enlighten.c
+++ b/arch/x86/xen/enlighten.c
@@ -669,7 +669,9 @@
    everything is pinned. */
 static __init void xen_alloc_pte_init(struct mm_struct *mm, u32 pfn)
 {
+#ifdef CONFIG_FLATMEM
 	BUG_ON(mem_map);	/* should only be used early */
+#endif
 	make_lowmem_page_readonly(__va(PFN_PHYS(pfn)));
 }
 

[PATCH 24 of 24] xen: add balloon driver

[Jeremy Fitzhardinge]

The balloon driver allows memory to be dynamically added or removed from the domain,
in order to allow host memory to be balanced between multiple domains.

This patch introduces the Xen balloon driver, though it currently only
allows a domain to be shrunk from its initial size (and re-grown back to
that size).  A later patch will add the ability to grow a domain beyond
its initial size.

Signed-off-by: Jeremy Fitzhardinge <jeremy.fitzhardinge@citrix.com>
---
 drivers/Kconfig                |    2
 drivers/xen/Kconfig            |   19 +
 drivers/xen/Makefile           |    1
 drivers/xen/balloon.c          |  712 ++++++++++++++++++++++++++++++++++++++++
 include/xen/balloon.h          |   61 +++
 include/xen/interface/memory.h |   12
 6 files changed, 799 insertions(+), 8 deletions(-)

diff --git a/drivers/Kconfig b/drivers/Kconfig
--- a/drivers/Kconfig
+++ b/drivers/Kconfig
@@ -97,4 +97,6 @@
 source "drivers/auxdisplay/Kconfig"
 
 source "drivers/uio/Kconfig"
+
+source "drivers/xen/Kconfig"
 endmenu
diff --git a/drivers/xen/Kconfig b/drivers/xen/Kconfig
new file mode 100644
--- /dev/null
+++ b/drivers/xen/Kconfig
@@ -0,0 +1,19 @@
+config XEN_BALLOON
+	bool "Xen memory balloon driver"
+	depends on XEN
+	default y
+	help
+	  The balloon driver allows the Xen domain to request more memory from
+	  the system to expand the domain's memory allocation, or alternatively
+	  return unneeded memory to the system.
+
+config XEN_SCRUB_PAGES
+	bool "Scrub pages before returning them to system"
+	depends on XEN_BALLOON
+	default y
+	help
+	  Scrub pages before returning them to the system for reuse by
+	  other domains.  This makes sure that any confidential data
+	  is not accidentally visible to other domains.  Is it more
+	  secure, but slightly less efficient.
+	  If in doubt, say yes.
diff --git a/drivers/xen/Makefile b/drivers/xen/Makefile
--- a/drivers/xen/Makefile
+++ b/drivers/xen/Makefile
@@ -1,3 +1,4 @@
 obj-y	+= grant-table.o features.o events.o
 obj-y	+= xenbus/
 obj-$(CONFIG_XEN_XENCOMM)	+= xencomm.o
+obj-$(CONFIG_XEN_BALLOON)	+= balloon.o
diff --git a/drivers/xen/balloon.c b/drivers/xen/balloon.c
new file mode 100644
--- /dev/null
+++ b/drivers/xen/balloon.c
@@ -0,0 +1,712 @@
+/******************************************************************************
+ * balloon.c
+ *
+ * Xen balloon driver - enables returning/claiming memory to/from Xen.
+ *
+ * Copyright (c) 2003, B Dragovic
+ * Copyright (c) 2003-2004, M Williamson, K Fraser
+ * Copyright (c) 2005 Dan M. Smith, IBM Corporation
+ *
+ * This program is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU General Public License version 2
+ * as published by the Free Software Foundation; or, when distributed
+ * separately from the Linux kernel or incorporated into other
+ * software packages, subject to the following license:
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a copy
+ * of this source file (the "Software"), to deal in the Software without
+ * restriction, including without limitation the rights to use, copy, modify,
+ * merge, publish, distribute, sublicense, and/or sell copies of the Software,
+ * and to permit persons to whom the Software is furnished to do so, subject to
+ * the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
+ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
+ * IN THE SOFTWARE.
+ */
+
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/sched.h>
+#include <linux/errno.h>
+#include <linux/mm.h>
+#include <linux/bootmem.h>
+#include <linux/pagemap.h>
+#include <linux/highmem.h>
+#include <linux/mutex.h>
+#include <linux/highmem.h>
+#include <linux/list.h>
+#include <linux/sysdev.h>
+
+#include <asm/xen/hypervisor.h>
+#include <asm/page.h>
+#include <asm/pgalloc.h>
+#include <asm/pgtable.h>
+#include <asm/uaccess.h>
+#include <asm/tlb.h>
+
+#include <xen/interface/memory.h>
+#include <xen/balloon.h>
+#include <xen/xenbus.h>
+#include <xen/features.h>
+#include <xen/page.h>
+
+#define PAGES2KB(_p) ((_p)<<(PAGE_SHIFT-10))
+
+#define BALLOON_CLASS_NAME "memory"
+
+struct balloon_stats {
+	/* We aim for 'current allocation' == 'target allocation'. */
+	unsigned long current_pages;
+	unsigned long target_pages;
+	/* We may hit the hard limit in Xen. If we do then we remember it. */
+	unsigned long hard_limit;
+	/*
+	 * Drivers may alter the memory reservation independently, but they
+	 * must inform the balloon driver so we avoid hitting the hard limit.
+	 */
+	unsigned long driver_pages;
+	/* Number of pages in high- and low-memory balloons. */
+	unsigned long balloon_low;
+	unsigned long balloon_high;
+};
+
+static DEFINE_MUTEX(balloon_mutex);
+
+static struct sys_device balloon_sysdev;
+
+static int register_balloon(struct sys_device *sysdev);
+
+/*
+ * Protects atomic reservation decrease/increase against concurrent increases.
+ * Also protects non-atomic updates of current_pages and driver_pages, and
+ * balloon lists.
+ */
+static DEFINE_SPINLOCK(balloon_lock);
+
+static struct balloon_stats balloon_stats;
+
+/* We increase/decrease in batches which fit in a page */
+static unsigned long frame_list[PAGE_SIZE / sizeof(unsigned long)];
+
+/* VM /proc information for memory */
+extern unsigned long totalram_pages;
+
+#ifdef CONFIG_HIGHMEM
+extern unsigned long totalhigh_pages;
+#define inc_totalhigh_pages() (totalhigh_pages++)
+#define dec_totalhigh_pages() (totalhigh_pages--)
+#else
+#define inc_totalhigh_pages() do {} while(0)
+#define dec_totalhigh_pages() do {} while(0)
+#endif
+
+/* List of ballooned pages, threaded through the mem_map array. */
+static LIST_HEAD(ballooned_pages);
+
+/* Main work function, always executed in process context. */
+static void balloon_process(struct work_struct *work);
+static DECLARE_WORK(balloon_worker, balloon_process);
+static struct timer_list balloon_timer;
+
+/* When ballooning out (allocating memory to return to Xen) we don't really
+   want the kernel to try too hard since that can trigger the oom killer. */
+#define GFP_BALLOON \
+	(GFP_HIGHUSER | __GFP_NOWARN | __GFP_NORETRY | __GFP_NOMEMALLOC)
+
+static void scrub_page(struct page *page)
+{
+#ifdef CONFIG_XEN_SCRUB_PAGES
+	if (PageHighMem(page)) {
+		void *v = kmap(page);
+		clear_page(v);
+		kunmap(v);
+	} else {
+		void *v = page_address(page);
+		clear_page(v);
+	}
+#endif
+}
+
+/* balloon_append: add the given page to the balloon. */
+static void balloon_append(struct page *page)
+{
+	/* Lowmem is re-populated first, so highmem pages go at list tail. */
+	if (PageHighMem(page)) {
+		list_add_tail(&page->lru, &ballooned_pages);
+		balloon_stats.balloon_high++;
+		dec_totalhigh_pages();
+	} else {
+		list_add(&page->lru, &ballooned_pages);
+		balloon_stats.balloon_low++;
+	}
+}
+
+/* balloon_retrieve: rescue a page from the balloon, if it is not empty. */
+static struct page *balloon_retrieve(void)
+{
+	struct page *page;
+
+	if (list_empty(&ballooned_pages))
+		return NULL;
+
+	page = list_entry(ballooned_pages.next, struct page, lru);
+	list_del(&page->lru);
+
+	if (PageHighMem(page)) {
+		balloon_stats.balloon_high--;
+		inc_totalhigh_pages();
+	}
+	else
+		balloon_stats.balloon_low--;
+
+	return page;
+}
+
+static struct page *balloon_first_page(void)
+{
+	if (list_empty(&ballooned_pages))
+		return NULL;
+	return list_entry(ballooned_pages.next, struct page, lru);
+}
+
+static struct page *balloon_next_page(struct page *page)
+{
+	struct list_head *next = page->lru.next;
+	if (next == &ballooned_pages)
+		return NULL;
+	return list_entry(next, struct page, lru);
+}
+
+static void balloon_alarm(unsigned long unused)
+{
+	schedule_work(&balloon_worker);
+}
+
+static unsigned long current_target(void)
+{
+	unsigned long target = min(balloon_stats.target_pages, balloon_stats.hard_limit);
+
+	target = min(target,
+		     balloon_stats.current_pages +
+		     balloon_stats.balloon_low +
+		     balloon_stats.balloon_high);
+
+	return target;
+}
+
+static int increase_reservation(unsigned long nr_pages)
+{
+	unsigned long  pfn, i, flags;
+	struct page   *page;
+	long           rc;
+	struct xen_memory_reservation reservation = {
+		.address_bits = 0,
+		.extent_order = 0,
+		.domid        = DOMID_SELF
+	};
+
+	if (nr_pages > ARRAY_SIZE(frame_list))
+		nr_pages = ARRAY_SIZE(frame_list);
+
+	spin_lock_irqsave(&balloon_lock, flags);
+
+	page = balloon_first_page();
+	for (i = 0; i < nr_pages; i++) {
+		BUG_ON(page == NULL);
+		frame_list[i] = page_to_pfn(page);;
+		page = balloon_next_page(page);
+	}
+
+	reservation.extent_start = (unsigned long)frame_list;
+	reservation.nr_extents   = nr_pages;
+	rc = HYPERVISOR_memory_op(
+		XENMEM_populate_physmap, &reservation);
+	if (rc < nr_pages) {
+		if (rc > 0) {
+			int ret;
+
+			/* We hit the Xen hard limit: reprobe. */
+			reservation.nr_extents = rc;
+			ret = HYPERVISOR_memory_op(XENMEM_decrease_reservation,
+					&reservation);
+			BUG_ON(ret != rc);
+		}
+		if (rc >= 0)
+			balloon_stats.hard_limit = (balloon_stats.current_pages + rc -
+						    balloon_stats.driver_pages);
+		goto out;
+	}
+
+	for (i = 0; i < nr_pages; i++) {
+		page = balloon_retrieve();
+		BUG_ON(page == NULL);
+
+		pfn = page_to_pfn(page);
+		BUG_ON(!xen_feature(XENFEAT_auto_translated_physmap) &&
+		       phys_to_machine_mapping_valid(pfn));
+
+		set_phys_to_machine(pfn, frame_list[i]);
+
+		/* Link back into the page tables if not highmem. */
+		if (pfn < max_low_pfn) {
+			int ret;
+			ret = HYPERVISOR_update_va_mapping(
+				(unsigned long)__va(pfn << PAGE_SHIFT),
+				mfn_pte(frame_list[i], PAGE_KERNEL),
+				0);
+			BUG_ON(ret);
+		}
+
+		/* Relinquish the page back to the allocator. */
+		ClearPageReserved(page);
+		init_page_count(page);
+		__free_page(page);
+	}
+
+	balloon_stats.current_pages += nr_pages;
+	totalram_pages = balloon_stats.current_pages;
+
+ out:
+	spin_unlock_irqrestore(&balloon_lock, flags);
+
+	return 0;
+}
+
+static int decrease_reservation(unsigned long nr_pages)
+{
+	unsigned long  pfn, i, flags;
+	struct page   *page;
+	int            need_sleep = 0;
+	int ret;
+	struct xen_memory_reservation reservation = {
+		.address_bits = 0,
+		.extent_order = 0,
+		.domid        = DOMID_SELF
+	};
+
+	if (nr_pages > ARRAY_SIZE(frame_list))
+		nr_pages = ARRAY_SIZE(frame_list);
+
+	for (i = 0; i < nr_pages; i++) {
+		if ((page = alloc_page(GFP_BALLOON)) == NULL) {
+			nr_pages = i;
+			need_sleep = 1;
+			break;
+		}
+
+		pfn = page_to_pfn(page);
+		frame_list[i] = pfn_to_mfn(pfn);
+
+		scrub_page(page);
+	}
+
+	/* Ensure that ballooned highmem pages don't have kmaps. */
+	kmap_flush_unused();
+	flush_tlb_all();
+
+	spin_lock_irqsave(&balloon_lock, flags);
+
+	/* No more mappings: invalidate P2M and add to balloon. */
+	for (i = 0; i < nr_pages; i++) {
+		pfn = mfn_to_pfn(frame_list[i]);
+		set_phys_to_machine(pfn, INVALID_P2M_ENTRY);
+		balloon_append(pfn_to_page(pfn));
+	}
+
+	reservation.extent_start = (unsigned long)frame_list;
+	reservation.nr_extents   = nr_pages;
+	ret = HYPERVISOR_memory_op(XENMEM_decrease_reservation, &reservation);
+	BUG_ON(ret != nr_pages);
+
+	balloon_stats.current_pages -= nr_pages;
+	totalram_pages = balloon_stats.current_pages;
+
+	spin_unlock_irqrestore(&balloon_lock, flags);
+
+	return need_sleep;
+}
+
+/*
+ * We avoid multiple worker processes conflicting via the balloon mutex.
+ * We may of course race updates of the target counts (which are protected
+ * by the balloon lock), or with changes to the Xen hard limit, but we will
+ * recover from these in time.
+ */
+static void balloon_process(struct work_struct *work)
+{
+	int need_sleep = 0;
+	long credit;
+
+	mutex_lock(&balloon_mutex);
+
+	do {
+		credit = current_target() - balloon_stats.current_pages;
+		if (credit > 0)
+			need_sleep = (increase_reservation(credit) != 0);
+		if (credit < 0)
+			need_sleep = (decrease_reservation(-credit) != 0);
+
+#ifndef CONFIG_PREEMPT
+		if (need_resched())
+			schedule();
+#endif
+	} while ((credit != 0) && !need_sleep);
+
+	/* Schedule more work if there is some still to be done. */
+	if (current_target() != balloon_stats.current_pages)
+		mod_timer(&balloon_timer, jiffies + HZ);
+
+	mutex_unlock(&balloon_mutex);
+}
+
+/* Resets the Xen limit, sets new target, and kicks off processing. */
+void balloon_set_new_target(unsigned long target)
+{
+	/* No need for lock. Not read-modify-write updates. */
+	balloon_stats.hard_limit   = ~0UL;
+	balloon_stats.target_pages = target;
+	schedule_work(&balloon_worker);
+}
+
+static struct xenbus_watch target_watch =
+{
+	.node = "memory/target"
+};
+
+/* React to a change in the target key */
+static void watch_target(struct xenbus_watch *watch,
+			 const char **vec, unsigned int len)
+{
+	unsigned long long new_target;
+	int err;
+
+	err = xenbus_scanf(XBT_NIL, "memory", "target", "%llu", &new_target);
+	if (err != 1) {
+		/* This is ok (for domain0 at least) - so just return */
+		return;
+	}
+
+	/* The given memory/target value is in KiB, so it needs converting to
+	 * pages. PAGE_SHIFT converts bytes to pages, hence PAGE_SHIFT - 10.
+	 */
+	balloon_set_new_target(new_target >> (PAGE_SHIFT - 10));
+}
+
+static int balloon_init_watcher(struct notifier_block *notifier,
+				unsigned long event,
+				void *data)
+{
+	int err;
+
+	err = register_xenbus_watch(&target_watch);
+	if (err)
+		printk(KERN_ERR "Failed to set balloon watcher\n");
+
+	return NOTIFY_DONE;
+}
+
+static struct notifier_block xenstore_notifier;
+
+static int __init balloon_init(void)
+{
+	unsigned long pfn;
+	struct page *page;
+
+	if (!is_running_on_xen())
+		return -ENODEV;
+
+	pr_info("xen_balloon: Initialising balloon driver.\n");
+
+	balloon_stats.current_pages = min(xen_start_info->nr_pages, max_pfn);
+	totalram_pages   = balloon_stats.current_pages;
+	balloon_stats.target_pages  = balloon_stats.current_pages;
+	balloon_stats.balloon_low   = 0;
+	balloon_stats.balloon_high  = 0;
+	balloon_stats.driver_pages  = 0UL;
+	balloon_stats.hard_limit    = ~0UL;
+
+	init_timer(&balloon_timer);
+	balloon_timer.data = 0;
+	balloon_timer.function = balloon_alarm;
+
+	register_balloon(&balloon_sysdev);
+
+	/* Initialise the balloon with excess memory space. */
+	for (pfn = xen_start_info->nr_pages; pfn < max_pfn; pfn++) {
+		page = pfn_to_page(pfn);
+		if (!PageReserved(page))
+			balloon_append(page);
+	}
+
+	target_watch.callback = watch_target;
+	xenstore_notifier.notifier_call = balloon_init_watcher;
+
+	register_xenstore_notifier(&xenstore_notifier);
+
+	return 0;
+}
+
+subsys_initcall(balloon_init);
+
+static void balloon_exit(void)
+{
+    /* XXX - release balloon here */
+    return;
+}
+
+module_exit(balloon_exit);
+
+static void balloon_update_driver_allowance(long delta)
+{
+	unsigned long flags;
+
+	spin_lock_irqsave(&balloon_lock, flags);
+	balloon_stats.driver_pages += delta;
+	spin_unlock_irqrestore(&balloon_lock, flags);
+}
+
+static int dealloc_pte_fn(
+	pte_t *pte, struct page *pmd_page, unsigned long addr, void *data)
+{
+	unsigned long mfn = pte_mfn(*pte);
+	int ret;
+	struct xen_memory_reservation reservation = {
+		.nr_extents   = 1,
+		.extent_order = 0,
+		.domid        = DOMID_SELF
+	};
+	reservation.extent_start = (unsigned long)&mfn;
+	set_pte_at(&init_mm, addr, pte, __pte_ma(0ull));
+	set_phys_to_machine(__pa(addr) >> PAGE_SHIFT, INVALID_P2M_ENTRY);
+	ret = HYPERVISOR_memory_op(XENMEM_decrease_reservation, &reservation);
+	BUG_ON(ret != 1);
+	return 0;
+}
+
+static struct page **alloc_empty_pages_and_pagevec(int nr_pages)
+{
+	unsigned long vaddr, flags;
+	struct page *page, **pagevec;
+	int i, ret;
+
+	pagevec = kmalloc(sizeof(page) * nr_pages, GFP_KERNEL);
+	if (pagevec == NULL)
+		return NULL;
+
+	for (i = 0; i < nr_pages; i++) {
+		page = pagevec[i] = alloc_page(GFP_KERNEL);
+		if (page == NULL)
+			goto err;
+
+		vaddr = (unsigned long)page_address(page);
+
+		scrub_page(page);
+
+		spin_lock_irqsave(&balloon_lock, flags);
+
+		if (xen_feature(XENFEAT_auto_translated_physmap)) {
+			unsigned long gmfn = page_to_pfn(page);
+			struct xen_memory_reservation reservation = {
+				.nr_extents   = 1,
+				.extent_order = 0,
+				.domid        = DOMID_SELF
+			};
+			reservation.extent_start = (unsigned long)&gmfn;
+			ret = HYPERVISOR_memory_op(XENMEM_decrease_reservation,
+						   &reservation);
+			if (ret == 1)
+				ret = 0; /* success */
+		} else {
+			ret = apply_to_page_range(&init_mm, vaddr, PAGE_SIZE,
+						  dealloc_pte_fn, NULL);
+		}
+
+		if (ret != 0) {
+			spin_unlock_irqrestore(&balloon_lock, flags);
+			__free_page(page);
+			goto err;
+		}
+
+		totalram_pages = --balloon_stats.current_pages;
+
+		spin_unlock_irqrestore(&balloon_lock, flags);
+	}
+
+ out:
+	schedule_work(&balloon_worker);
+	flush_tlb_all();
+	return pagevec;
+
+ err:
+	spin_lock_irqsave(&balloon_lock, flags);
+	while (--i >= 0)
+		balloon_append(pagevec[i]);
+	spin_unlock_irqrestore(&balloon_lock, flags);
+	kfree(pagevec);
+	pagevec = NULL;
+	goto out;
+}
+
+static void free_empty_pages_and_pagevec(struct page **pagevec, int nr_pages)
+{
+	unsigned long flags;
+	int i;
+
+	if (pagevec == NULL)
+		return;
+
+	spin_lock_irqsave(&balloon_lock, flags);
+	for (i = 0; i < nr_pages; i++) {
+		BUG_ON(page_count(pagevec[i]) != 1);
+		balloon_append(pagevec[i]);
+	}
+	spin_unlock_irqrestore(&balloon_lock, flags);
+
+	kfree(pagevec);
+
+	schedule_work(&balloon_worker);
+}
+
+static void balloon_release_driver_page(struct page *page)
+{
+	unsigned long flags;
+
+	spin_lock_irqsave(&balloon_lock, flags);
+	balloon_append(page);
+	balloon_stats.driver_pages--;
+	spin_unlock_irqrestore(&balloon_lock, flags);
+
+	schedule_work(&balloon_worker);
+}
+
+
+#define BALLOON_SHOW(name, format, args...)			\
+	static ssize_t show_##name(struct sys_device *dev,	\
+				   char *buf)			\
+	{							\
+		return sprintf(buf, format, ##args);		\
+	}							\
+	static SYSDEV_ATTR(name, S_IRUGO, show_##name, NULL)
+
+BALLOON_SHOW(current_kb, "%lu\n", PAGES2KB(balloon_stats.current_pages));
+BALLOON_SHOW(low_kb, "%lu\n", PAGES2KB(balloon_stats.balloon_low));
+BALLOON_SHOW(high_kb, "%lu\n", PAGES2KB(balloon_stats.balloon_high));
+BALLOON_SHOW(hard_limit_kb,
+	     (balloon_stats.hard_limit!=~0UL) ? "%lu\n" : "???\n",
+	     (balloon_stats.hard_limit!=~0UL) ? PAGES2KB(balloon_stats.hard_limit) : 0);
+BALLOON_SHOW(driver_kb, "%lu\n", PAGES2KB(balloon_stats.driver_pages));
+
+static ssize_t show_target_kb(struct sys_device *dev, char *buf)
+{
+	return sprintf(buf, "%lu\n", PAGES2KB(balloon_stats.target_pages));
+}
+
+static ssize_t store_target_kb(struct sys_device *dev,
+			       const char *buf,
+			       size_t count)
+{
+	char memstring[64], *endchar;
+	unsigned long long target_bytes;
+
+	if (!capable(CAP_SYS_ADMIN))
+		return -EPERM;
+
+	if (count <= 1)
+		return -EBADMSG; /* runt */
+	if (count > sizeof(memstring))
+		return -EFBIG;   /* too long */
+	strcpy(memstring, buf);
+
+	target_bytes = memparse(memstring, &endchar);
+	balloon_set_new_target(target_bytes >> PAGE_SHIFT);
+
+	return count;
+}
+
+static SYSDEV_ATTR(target_kb, S_IRUGO | S_IWUSR,
+		   show_target_kb, store_target_kb);
+
+static struct sysdev_attribute *balloon_attrs[] = {
+	&attr_target_kb,
+};
+
+static struct attribute *balloon_info_attrs[] = {
+	&attr_current_kb.attr,
+	&attr_low_kb.attr,
+	&attr_high_kb.attr,
+	&attr_hard_limit_kb.attr,
+	&attr_driver_kb.attr,
+	NULL
+};
+
+static struct attribute_group balloon_info_group = {
+	.name = "info",
+	.attrs = balloon_info_attrs,
+};
+
+static struct sysdev_class balloon_sysdev_class = {
+	.name = BALLOON_CLASS_NAME,
+};
+
+static int register_balloon(struct sys_device *sysdev)
+{
+	int i, error;
+
+	error = sysdev_class_register(&balloon_sysdev_class);
+	if (error)
+		return error;
+
+	sysdev->id = 0;
+	sysdev->cls = &balloon_sysdev_class;
+
+	error = sysdev_register(sysdev);
+	if (error) {
+		sysdev_class_unregister(&balloon_sysdev_class);
+		return error;
+	}
+
+	for (i = 0; i < ARRAY_SIZE(balloon_attrs); i++) {
+		error = sysdev_create_file(sysdev, balloon_attrs[i]);
+		if (error)
+			goto fail;
+	}
+
+	error = sysfs_create_group(&sysdev->kobj, &balloon_info_group);
+	if (error)
+		goto fail;
+
+	return 0;
+
+ fail:
+	while (--i >= 0)
+		sysdev_remove_file(sysdev, balloon_attrs[i]);
+	sysdev_unregister(sysdev);
+	sysdev_class_unregister(&balloon_sysdev_class);
+	return error;
+}
+
+static void unregister_balloon(struct sys_device *sysdev)
+{
+	int i;
+
+	sysfs_remove_group(&sysdev->kobj, &balloon_info_group);
+	for (i = 0; i < ARRAY_SIZE(balloon_attrs); i++)
+		sysdev_remove_file(sysdev, balloon_attrs[i]);
+	sysdev_unregister(sysdev);
+	sysdev_class_unregister(&balloon_sysdev_class);
+}
+
+static void balloon_sysfs_exit(void)
+{
+	unregister_balloon(&balloon_sysdev);
+}
+
+MODULE_LICENSE("GPL");
diff --git a/include/xen/balloon.h b/include/xen/balloon.h
new file mode 100644
--- /dev/null
+++ b/include/xen/balloon.h
@@ -0,0 +1,61 @@
+/******************************************************************************
+ * balloon.h
+ *
+ * Xen balloon driver - enables returning/claiming memory to/from Xen.
+ *
+ * Copyright (c) 2003, B Dragovic
+ * Copyright (c) 2003-2004, M Williamson, K Fraser
+ *
+ * This program is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU General Public License version 2
+ * as published by the Free Software Foundation; or, when distributed
+ * separately from the Linux kernel or incorporated into other
+ * software packages, subject to the following license:
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a copy
+ * of this source file (the "Software"), to deal in the Software without
+ * restriction, including without limitation the rights to use, copy, modify,
+ * merge, publish, distribute, sublicense, and/or sell copies of the Software,
+ * and to permit persons to whom the Software is furnished to do so, subject to
+ * the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
+ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
+ * IN THE SOFTWARE.
+ */
+
+#ifndef __XEN_BALLOON_H__
+#define __XEN_BALLOON_H__
+
+#include <linux/spinlock.h>
+
+#if 0
+/*
+ * Inform the balloon driver that it should allow some slop for device-driver
+ * memory activities.
+ */
+void balloon_update_driver_allowance(long delta);
+
+/* Allocate/free a set of empty pages in low memory (i.e., no RAM mapped). */
+struct page **alloc_empty_pages_and_pagevec(int nr_pages);
+void free_empty_pages_and_pagevec(struct page **pagevec, int nr_pages);
+
+void balloon_release_driver_page(struct page *page);
+
+/*
+ * Prevent the balloon driver from changing the memory reservation during
+ * a driver critical region.
+ */
+extern spinlock_t balloon_lock;
+#define balloon_lock(__flags)   spin_lock_irqsave(&balloon_lock, __flags)
+#define balloon_unlock(__flags) spin_unlock_irqrestore(&balloon_lock, __flags)
+#endif
+
+#endif /* __XEN_BALLOON_H__ */
diff --git a/include/xen/interface/memory.h b/include/xen/interface/memory.h
--- a/include/xen/interface/memory.h
+++ b/include/xen/interface/memory.h
@@ -29,7 +29,7 @@
      *   OUT: GMFN bases of extents that were allocated
      *   (NB. This command also updates the mach_to_phys translation table)
      */
-    GUEST_HANDLE(ulong) extent_start;
+    ulong extent_start;
 
     /* Number of extents, and size/alignment of each (2^extent_order pages). */
     unsigned long  nr_extents;
@@ -50,7 +50,6 @@
     domid_t        domid;
 
 };
-DEFINE_GUEST_HANDLE_STRUCT(xen_memory_reservation);
 
 /*
  * Returns the maximum machine frame number of mapped RAM in this system.
@@ -86,7 +85,7 @@
      * any large discontiguities in the machine address space, 2MB gaps in
      * the machphys table will be represented by an MFN base of zero.
      */
-    GUEST_HANDLE(ulong) extent_start;
+    ulong extent_start;
 
     /*
      * Number of extents written to the above array. This will be smaller
@@ -94,7 +93,6 @@
      */
     unsigned int nr_extents;
 };
-DEFINE_GUEST_HANDLE_STRUCT(xen_machphys_mfn_list);
 
 /*
  * Sets the GPFN at which a particular page appears in the specified guest's
@@ -117,7 +115,6 @@
     /* GPFN where the source mapping page should appear. */
     unsigned long gpfn;
 };
-DEFINE_GUEST_HANDLE_STRUCT(xen_add_to_physmap);
 
 /*
  * Translates a list of domain-specific GPFNs into MFNs. Returns a -ve error
@@ -132,14 +129,13 @@
     unsigned long nr_gpfns;
 
     /* List of GPFNs to translate. */
-    GUEST_HANDLE(ulong) gpfn_list;
+    ulong gpfn_list;
 
     /*
      * Output list to contain MFN translations. May be the same as the input
      * list (in which case each input GPFN is overwritten with the output MFN).
      */
-    GUEST_HANDLE(ulong) mfn_list;
+    ulong mfn_list;
 };
-DEFINE_GUEST_HANDLE_STRUCT(xen_translate_gpfn_list);
 
 #endif /* __XEN_PUBLIC_MEMORY_H__ */

Re: [PATCH 24 of 24] xen: add balloon driver

[Isaku Yamahata]

The following modification to include/xen/interface/memory.h 
removes handles related definitions.
It breaks ia64/xen compilation because ia64/xen uses memory interface
not only for balloon, but also for xen_alloc_vm_area/xen_free_vm_area().

Can it be added again?

thanks,

On Wed, Apr 02, 2008 at 10:54:13AM -0700, Jeremy Fitzhardinge wrote:
> diff --git a/include/xen/interface/memory.h b/include/xen/interface/memory.h
> --- a/include/xen/interface/memory.h
> +++ b/include/xen/interface/memory.h
> @@ -29,7 +29,7 @@
>       *   OUT: GMFN bases of extents that were allocated
>       *   (NB. This command also updates the mach_to_phys translation table)
>       */
> -    GUEST_HANDLE(ulong) extent_start;
> +    ulong extent_start;
>  
>      /* Number of extents, and size/alignment of each (2^extent_order pages). */
>      unsigned long  nr_extents;
> @@ -50,7 +50,6 @@
>      domid_t        domid;
>  
>  };
> -DEFINE_GUEST_HANDLE_STRUCT(xen_memory_reservation);
>  
>  /*
>   * Returns the maximum machine frame number of mapped RAM in this system.
> @@ -86,7 +85,7 @@
>       * any large discontiguities in the machine address space, 2MB gaps in
>       * the machphys table will be represented by an MFN base of zero.
>       */
> -    GUEST_HANDLE(ulong) extent_start;
> +    ulong extent_start;
>  
>      /*
>       * Number of extents written to the above array. This will be smaller
> @@ -94,7 +93,6 @@
>       */
>      unsigned int nr_extents;
>  };
> -DEFINE_GUEST_HANDLE_STRUCT(xen_machphys_mfn_list);
>  
>  /*
>   * Sets the GPFN at which a particular page appears in the specified guest's
> @@ -117,7 +115,6 @@
>      /* GPFN where the source mapping page should appear. */
>      unsigned long gpfn;
>  };
> -DEFINE_GUEST_HANDLE_STRUCT(xen_add_to_physmap);
>  
>  /*
>   * Translates a list of domain-specific GPFNs into MFNs. Returns a -ve error
> @@ -132,14 +129,13 @@
>      unsigned long nr_gpfns;
>  
>      /* List of GPFNs to translate. */
> -    GUEST_HANDLE(ulong) gpfn_list;
> +    ulong gpfn_list;
>  
>      /*
>       * Output list to contain MFN translations. May be the same as the input
>       * list (in which case each input GPFN is overwritten with the output MFN).
>       */
> -    GUEST_HANDLE(ulong) mfn_list;
> +    ulong mfn_list;
>  };
> -DEFINE_GUEST_HANDLE_STRUCT(xen_translate_gpfn_list);
>  
>  #endif /* __XEN_PUBLIC_MEMORY_H__ */
> 
> 
> 

-- 
yamahata

Re: [PATCH 24 of 24] xen: add balloon driver

[Jeremy Fitzhardinge]

Isaku Yamahata wrote:
> The following modification to include/xen/interface/memory.h 
> removes handles related definitions.
> It breaks ia64/xen compilation because ia64/xen uses memory interface
> not only for balloon, but also for xen_alloc_vm_area/xen_free_vm_area().
>
> Can it be added again?
>   

Yes, sure.  We're going to hit that a bit because we needed to remove 
them to make the initial upstreaming smooth.

    J

Re: [PATCH 00 of 24] More Xen updates

[Ingo Molnar]

* Jeremy Fitzhardinge <jeremy@goop.org> wrote:

> Hi Ingo,
> 
> More patches.  There are three groups:
> 
> - some code movements to make common code available to other
>   architectures (namely, ia64)
> - Xen device driver updates, including a new xen paravirt framebuffer driver,
> - The Xen balloon driver (shrink only, so no dependency on memory hotplug yet)
> - some "using smp_processor_id while preemptable" warning fixes,
>   mostly as a result of the core kernel's tendency to do tlb flushes
>   and update init_mm while preemptable,
> - other little Xen cleanups

thanks, applied for testing. The .25 Xen items we've got queued up are:

 Subject: xen: refactor xen_{alloc,release}_{pte,pmd}()
 Subject: xen: do not pin/unpin PMD pages
 Subject: xen: clear PG_pinned in release_{pte,pmd}()

anything missing?

	Ingo

Re: [PATCH 00 of 24] More Xen updates

[Jeremy Fitzhardinge]

[-- Attachment #1: Type: text/plain, Size: 919 bytes --]

Ingo Molnar wrote:
> * Jeremy Fitzhardinge <jeremy@goop.org> wrote:
>
>   
>> Hi Ingo,
>>
>> More patches.  There are three groups:
>>
>> - some code movements to make common code available to other
>>   architectures (namely, ia64)
>> - Xen device driver updates, including a new xen paravirt framebuffer driver,
>> - The Xen balloon driver (shrink only, so no dependency on memory hotplug yet)
>> - some "using smp_processor_id while preemptable" warning fixes,
>>   mostly as a result of the core kernel's tendency to do tlb flushes
>>   and update init_mm while preemptable,
>> - other little Xen cleanups
>>     
>
> thanks, applied for testing. The .25 Xen items we've got queued up are:
>
>  Subject: xen: refactor xen_{alloc,release}_{pte,pmd}()
>  Subject: xen: do not pin/unpin PMD pages
>  Subject: xen: clear PG_pinned in release_{pte,pmd}()
>   

Mark just posted another bugfix patch - attached.

    J



[-- Attachment #2: xen-Fix-grant-table-bug.patch --]
[-- Type: text/plain, Size: 3309 bytes --]

Subject: xen: Fix grant table bug

From: Michael Abd-El-Malek <mabdelmalek@cmu.edu>

A PV OS has two grant table data structures: the grant table itself
and a free list.  The free list is composed of an array of pages,
which grow dynamically as the guest OS requires more grants.  While
the grant table contains 8-byte entries, the free list contains 4-byte
entries.  So we have half as many pages in the free list than in the
grant table.

There was a bug in the free list allocation code. The free list was
indexed as if it was the same size as the grant table.  But it's only
half as large.  So memory got corrupted, and I was seeing crashes in
the slab allocator later on.

Taken from:

  http://xenbits.xensource.com/linux-2.6.18-xen.hg?rev/4018c0da3360

Signed-off-by: Michael Abd-El-Malek <mabdelmalek@cmu.edu>
Signed-off-by: Mark McLoughlin <markmc@redhat.com>
Signed-off-by: Jeremy Fitzhardinge <jeremy.fitzhardinge@citrix.com>
---
 drivers/xen/grant-table.c |   17 ++++++++++-------
 1 file changed, 10 insertions(+), 7 deletions(-)

===================================================================
--- a/drivers/xen/grant-table.c
+++ b/drivers/xen/grant-table.c
@@ -381,11 +381,15 @@ EXPORT_SYMBOL_GPL(gnttab_cancel_free_callback);
 static int grow_gnttab_list(unsigned int more_frames)
 {
 	unsigned int new_nr_grant_frames, extra_entries, i;
+	unsigned int nr_glist_frames, new_nr_glist_frames;
 
 	new_nr_grant_frames = nr_grant_frames + more_frames;
 	extra_entries       = more_frames * GREFS_PER_GRANT_FRAME;
 
-	for (i = nr_grant_frames; i < new_nr_grant_frames; i++) {
+	nr_glist_frames = (nr_grant_frames * GREFS_PER_GRANT_FRAME + RPP - 1) / RPP;
+	new_nr_glist_frames =
+		(new_nr_grant_frames * GREFS_PER_GRANT_FRAME + RPP - 1) / RPP;
+	for (i = nr_glist_frames; i < new_nr_glist_frames; i++) {
 		gnttab_list[i] = (grant_ref_t *)__get_free_page(GFP_ATOMIC);
 		if (!gnttab_list[i])
 			goto grow_nomem;
@@ -407,7 +411,7 @@ static int grow_gnttab_list(unsigned int more_frames)
 	return 0;
 
 grow_nomem:
-	for ( ; i >= nr_grant_frames; i--)
+	for ( ; i >= nr_glist_frames; i--)
 		free_page((unsigned long) gnttab_list[i]);
 	return -ENOMEM;
 }
@@ -530,7 +534,7 @@ static int gnttab_expand(unsigned int req_entries)
 static int __devinit gnttab_init(void)
 {
 	int i;
-	unsigned int max_nr_glist_frames;
+	unsigned int max_nr_glist_frames, nr_glist_frames;
 	unsigned int nr_init_grefs;
 
 	if (!is_running_on_xen())
@@ -543,15 +547,15 @@ static int __devinit gnttab_init(void)
 	 * grant reference free list on the current hypervisor.
 	 */
 	max_nr_glist_frames = (boot_max_nr_grant_frames *
-			       GREFS_PER_GRANT_FRAME /
-			       (PAGE_SIZE / sizeof(grant_ref_t)));
+			       GREFS_PER_GRANT_FRAME / RPP);
 
 	gnttab_list = kmalloc(max_nr_glist_frames * sizeof(grant_ref_t *),
 			      GFP_KERNEL);
 	if (gnttab_list == NULL)
 		return -ENOMEM;
 
-	for (i = 0; i < nr_grant_frames; i++) {
+	nr_glist_frames = (nr_grant_frames * GREFS_PER_GRANT_FRAME + RPP - 1) / RPP;
+	for (i = 0; i < nr_glist_frames; i++) {
 		gnttab_list[i] = (grant_ref_t *)__get_free_page(GFP_KERNEL);
 		if (gnttab_list[i] == NULL)
 			goto ini_nomem;
-- 



_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xensource.com
http://lists.xensource.com/xen-devel