Mailing List Archive

# HG changeset patch
# User Alex Williamson <alex.williamson@hp.com>
# Date 1185545750 21600
# Node ID b0bf9ba32bfe341af07da97d57572659c920fd30
# Parent e6974eba1baec726ca1a97d076f0bf8428096bc2
# Parent 88a17da7f3362126182423100a9d7d4c0d854139
merge with linux-2.6.18-xen.hg
---
arch/i386/kernel/swiotlb.c | 745 -----------------------------
arch/i386/Kconfig | 4
arch/i386/Makefile | 1
arch/i386/kernel/Makefile | 26 -
arch/i386/kernel/acpi/Makefile | 4
arch/i386/kernel/acpi/sleep-xen.c | 129 +++++
arch/i386/kernel/cpu/Makefile | 4
arch/i386/kernel/cpu/mtrr/Makefile | 8
arch/i386/kernel/time-xen.c | 4
arch/i386/mm/Makefile | 9
arch/i386/mm/fault-xen.c | 33 -
arch/i386/mm/ioremap-xen.c | 2
arch/i386/pci/Makefile | 5
arch/i386/power/cpu.c | 3
arch/powerpc/platforms/xen/hcall.c | 146 +++++
arch/powerpc/platforms/xen/setup.c | 13
arch/x86_64/Kconfig | 4
arch/x86_64/Makefile | 1
arch/x86_64/ia32/Makefile | 13
arch/x86_64/kernel/Makefile | 18
arch/x86_64/kernel/acpi/Makefile | 4
arch/x86_64/kernel/acpi/sleep-xen.c | 162 ++++++
arch/x86_64/kernel/head-xen.S | 15
arch/x86_64/kernel/suspend.c | 2
arch/x86_64/mm/Makefile | 13
arch/x86_64/pci/Makefile | 7
buildconfigs/linux-defconfig_xen_x86_32 | 7
buildconfigs/linux-defconfig_xen_x86_64 | 7
drivers/acpi/Kconfig | 6
drivers/acpi/hardware/hwsleep.c | 5
drivers/acpi/sleep/main.c | 9
drivers/acpi/sleep/poweroff.c | 2
drivers/xen/balloon/balloon.c | 12
drivers/xen/blkfront/blkfront.c | 2
drivers/xen/blktap/blktap.c | 12
drivers/xen/char/mem.c | 31 -
drivers/xen/core/gnttab.c | 8
drivers/xen/core/machine_reboot.c | 1
drivers/xen/fbfront/xenfb.c | 1
drivers/xen/fbfront/xenkbd.c | 1
drivers/xen/netfront/netfront.c | 58 +-
drivers/xen/netfront/netfront.h | 4
drivers/xen/pciback/conf_space_header.c | 18
drivers/xen/pcifront/xenbus.c | 1
drivers/xen/privcmd/privcmd.c | 1
drivers/xen/xenbus/xenbus_probe.c | 59 +-
drivers/xen/xenbus/xenbus_probe_backend.c | 1
drivers/xen/xenbus/xenbus_xs.c | 6
include/asm-i386/acpi.h | 4
include/asm-i386/mach-xen/asm/gnttab_dma.h | 2
include/asm-i386/mach-xen/asm/io.h | 3
include/asm-powerpc/io.h | 3
include/asm-x86_64/acpi.h | 4
include/asm-x86_64/mach-xen/asm/io.h | 3
include/xen/interface/platform.h | 12
lib/Makefile | 3
lib/swiotlb-xen.c | 745 +++++++++++++++++++++++++++++
scripts/Makefile.build | 12
scripts/Makefile.lib | 6
scripts/Makefile.xen.awk | 34 +
60 files changed, 1528 insertions(+), 930 deletions(-)

diff -r e6974eba1bae -r b0bf9ba32bfe arch/i386/Kconfig
--- a/arch/i386/Kconfig Thu Jul 26 14:33:50 2007 -0600
+++ b/arch/i386/Kconfig Fri Jul 27 08:15:50 2007 -0600
@@ -832,9 +832,7 @@ menu "Power management options (ACPI, AP
menu "Power management options (ACPI, APM)"
depends on !(X86_VOYAGER || XEN_UNPRIVILEGED_GUEST)

-if !X86_XEN
-source kernel/power/Kconfig
-endif
+source "kernel/power/Kconfig"

source "drivers/acpi/Kconfig"

diff -r e6974eba1bae -r b0bf9ba32bfe arch/i386/Makefile
--- a/arch/i386/Makefile Thu Jul 26 14:33:50 2007 -0600
+++ b/arch/i386/Makefile Fri Jul 27 08:15:50 2007 -0600
@@ -116,7 +116,6 @@ PHONY += zImage bzImage compressed zlilo

ifdef CONFIG_XEN
CPPFLAGS := -Iinclude$(if $(KBUILD_SRC),2)/asm/mach-xen $(CPPFLAGS)
-head-y := arch/i386/kernel/head-xen.o arch/i386/kernel/init_task-xen.o
boot := arch/i386/boot-xen
.PHONY: vmlinuz
all: vmlinuz
diff -r e6974eba1bae -r b0bf9ba32bfe arch/i386/kernel/Makefile
--- a/arch/i386/kernel/Makefile Thu Jul 26 14:33:50 2007 -0600
+++ b/arch/i386/kernel/Makefile Fri Jul 27 08:15:50 2007 -0600
@@ -43,12 +43,7 @@ EXTRA_AFLAGS := -traditional
EXTRA_AFLAGS := -traditional

obj-$(CONFIG_SCx200) += scx200.o
-
-ifdef CONFIG_XEN
-vsyscall_note := vsyscall-note-xen.o
-else
-vsyscall_note := vsyscall-note.o
-endif
+obj-$(CONFIG_XEN) += fixup.o

# vsyscall.o contains the vsyscall DSO images as __initdata.
# We must build both images before we can assemble it.
@@ -71,7 +66,7 @@ SYSCFLAGS_vsyscall-int80.so = $(vsyscall

$(obj)/vsyscall-int80.so $(obj)/vsyscall-sysenter.so: \
$(obj)/vsyscall-%.so: $(src)/vsyscall.lds \
- $(obj)/vsyscall-%.o $(obj)/$(vsyscall_note) FORCE
+ $(obj)/vsyscall-%.o $(obj)/vsyscall-note.o FORCE
$(call if_changed,syscall)

# We also create a special relocatable object that should mirror the symbol
@@ -83,20 +78,11 @@ extra-y += vsyscall-syms.o

SYSCFLAGS_vsyscall-syms.o = -r
$(obj)/vsyscall-syms.o: $(src)/vsyscall.lds \
- $(obj)/vsyscall-sysenter.o $(obj)/$(vsyscall_note) FORCE
+ $(obj)/vsyscall-sysenter.o $(obj)/vsyscall-note.o FORCE
$(call if_changed,syscall)

+early_printk-y += ../../x86_64/kernel/early_printk.o
k8-y += ../../x86_64/kernel/k8.o

-ifdef CONFIG_XEN
-include $(srctree)/scripts/Makefile.xen
-
-obj-y += fixup.o
-microcode-$(subst m,y,$(CONFIG_MICROCODE)) := microcode-xen.o
-n-obj-xen := i8259.o timers/ reboot.o smpboot.o trampoline.o
-
-obj-y := $(call filterxen, $(obj-y), $(n-obj-xen))
-obj-y := $(call cherrypickxen, $(obj-y))
-extra-y := $(call cherrypickxen, $(extra-y))
-%/head-xen.o %/head-xen.s: EXTRA_AFLAGS :=
-endif
+disabled-obj-$(CONFIG_XEN) := i8259.o reboot.o smpboot.o trampoline.o
+%/head.o %/head.s: $(if $(CONFIG_XEN),EXTRA_AFLAGS,dummy) :=
diff -r e6974eba1bae -r b0bf9ba32bfe arch/i386/kernel/acpi/Makefile
--- a/arch/i386/kernel/acpi/Makefile Thu Jul 26 14:33:50 2007 -0600
+++ b/arch/i386/kernel/acpi/Makefile Fri Jul 27 08:15:50 2007 -0600
@@ -6,7 +6,3 @@ obj-y += cstate.o processor.o
obj-y += cstate.o processor.o
endif

-ifdef CONFIG_XEN
-include $(srctree)/scripts/Makefile.xen
-obj-y := $(call cherrypickxen, $(obj-y), $(src))
-endif
diff -r e6974eba1bae -r b0bf9ba32bfe arch/i386/kernel/acpi/sleep-xen.c
--- /dev/null Thu Jan 01 00:00:00 1970 +0000
+++ b/arch/i386/kernel/acpi/sleep-xen.c Fri Jul 27 08:15:50 2007 -0600
@@ -0,0 +1,129 @@
+/*
+ * sleep.c - x86-specific ACPI sleep support.
+ *
+ * Copyright (C) 2001-2003 Patrick Mochel
+ * Copyright (C) 2001-2003 Pavel Machek <pavel@suse.cz>
+ */
+
+#include <linux/acpi.h>
+#include <linux/bootmem.h>
+#include <linux/dmi.h>
+#include <linux/cpumask.h>
+
+#include <asm/smp.h>
+
+/* address in low memory of the wakeup routine. */
+unsigned long acpi_wakeup_address = 0;
+unsigned long acpi_video_flags;
+extern char wakeup_start, wakeup_end;
+
+extern unsigned long FASTCALL(acpi_copy_wakeup_routine(unsigned long));
+
+/**
+ * acpi_save_state_mem - save kernel state
+ *
+ * Create an identity mapped page table and copy the wakeup routine to
+ * low memory.
+ */
+int acpi_save_state_mem(void)
+{
+#ifndef CONFIG_ACPI_PV_SLEEP
+ if (!acpi_wakeup_address)
+ return 1;
+ memcpy((void *)acpi_wakeup_address, &wakeup_start,
+ &wakeup_end - &wakeup_start);
+ acpi_copy_wakeup_routine(acpi_wakeup_address);
+#endif
+ return 0;
+}
+
+/*
+ * acpi_restore_state - undo effects of acpi_save_state_mem
+ */
+void acpi_restore_state_mem(void)
+{
+}
+
+/**
+ * acpi_reserve_bootmem - do _very_ early ACPI initialisation
+ *
+ * We allocate a page from the first 1MB of memory for the wakeup
+ * routine for when we come back from a sleep state. The
+ * runtime allocator allows specification of <16MB pages, but not
+ * <1MB pages.
+ */
+void __init acpi_reserve_bootmem(void)
+{
+ if ((&wakeup_end - &wakeup_start) > PAGE_SIZE) {
+ printk(KERN_ERR
+ "ACPI: Wakeup code way too big, S3 disabled.\n");
+ return;
+ }
+
+ acpi_wakeup_address = (unsigned long)alloc_bootmem_low(PAGE_SIZE);
+ if (!acpi_wakeup_address)
+ printk(KERN_ERR "ACPI: Cannot allocate lowmem, S3 disabled.\n");
+}
+
+static int __init acpi_sleep_setup(char *str)
+{
+ while ((str != NULL) && (*str != '\0')) {
+ if (strncmp(str, "s3_bios", 7) == 0)
+ acpi_video_flags = 1;
+ if (strncmp(str, "s3_mode", 7) == 0)
+ acpi_video_flags |= 2;
+ str = strchr(str, ',');
+ if (str != NULL)
+ str += strspn(str, ", \t");
+ }
+ return 1;
+}
+
+__setup("acpi_sleep=", acpi_sleep_setup);
+
+static __init int reset_videomode_after_s3(struct dmi_system_id *d)
+{
+ acpi_video_flags |= 2;
+ return 0;
+}
+
+static __initdata struct dmi_system_id acpisleep_dmi_table[] = {
+ { /* Reset video mode after returning from ACPI S3 sleep */
+ .callback = reset_videomode_after_s3,
+ .ident = "Toshiba Satellite 4030cdt",
+ .matches = {
+ DMI_MATCH(DMI_PRODUCT_NAME, "S4030CDT/4.3"),
+ },
+ },
+ {}
+};
+
+static int __init acpisleep_dmi_init(void)
+{
+ dmi_check_system(acpisleep_dmi_table);
+ return 0;
+}
+
+core_initcall(acpisleep_dmi_init);
+
+#ifdef CONFIG_ACPI_PV_SLEEP
+#include <asm/hypervisor.h>
+#include <xen/interface/platform.h>
+int acpi_notify_hypervisor_state(u8 sleep_state,
+ u32 pm1a_cnt, u32 pm1b_cnt)
+{
+ struct xen_platform_op op = {
+ .cmd = XENPF_enter_acpi_sleep,
+ .interface_version = XENPF_INTERFACE_VERSION,
+ .u = {
+ .enter_acpi_sleep = {
+ .pm1a_cnt_val = (u16)pm1a_cnt,
+ .pm1b_cnt_val = (u16)pm1b_cnt,
+ .sleep_state = sleep_state,
+ },
+ },
+ };
+
+ return HYPERVISOR_platform_op(&op);
+}
+#endif /* CONFIG_ACPI_PV_SLEEP */
diff -r e6974eba1bae -r b0bf9ba32bfe arch/i386/kernel/cpu/Makefile
--- a/arch/i386/kernel/cpu/Makefile Thu Jul 26 14:33:50 2007 -0600
+++ b/arch/i386/kernel/cpu/Makefile Fri Jul 27 08:15:50 2007 -0600
@@ -18,7 +18,3 @@ obj-$(CONFIG_MTRR) += mtrr/
obj-$(CONFIG_MTRR) += mtrr/
obj-$(CONFIG_CPU_FREQ) += cpufreq/

-ifdef CONFIG_XEN
-include $(srctree)/scripts/Makefile.xen
-obj-y := $(call cherrypickxen, $(obj-y), $(src))
-endif
diff -r e6974eba1bae -r b0bf9ba32bfe arch/i386/kernel/cpu/mtrr/Makefile
--- a/arch/i386/kernel/cpu/mtrr/Makefile Thu Jul 26 14:33:50 2007 -0600
+++ b/arch/i386/kernel/cpu/mtrr/Makefile Fri Jul 27 08:15:50 2007 -0600
@@ -3,10 +3,4 @@ obj-y += cyrix.o
obj-y += cyrix.o
obj-y += centaur.o

-ifdef CONFIG_XEN
-include $(srctree)/scripts/Makefile.xen
-n-obj-xen := generic.o state.o amd.o cyrix.o centaur.o
-
-obj-y := $(call filterxen, $(obj-y), $(n-obj-xen))
-obj-y := $(call cherrypickxen, $(obj-y))
-endif
+obj-$(CONFIG_XEN) := main.o if.o
diff -r e6974eba1bae -r b0bf9ba32bfe arch/i386/kernel/swiotlb.c
--- a/arch/i386/kernel/swiotlb.c Thu Jul 26 14:33:50 2007 -0600
+++ /dev/null Thu Jan 01 00:00:00 1970 +0000
@@ -1,745 +0,0 @@
-/*
- * Dynamic DMA mapping support.
- *
- * This implementation is a fallback for platforms that do not support
- * I/O TLBs (aka DMA address translation hardware).
- * Copyright (C) 2000 Asit Mallick <Asit.K.Mallick@intel.com>
- * Copyright (C) 2000 Goutham Rao <goutham.rao@intel.com>
- * Copyright (C) 2000, 2003 Hewlett-Packard Co
- * David Mosberger-Tang <davidm@hpl.hp.com>
- * Copyright (C) 2005 Keir Fraser <keir@xensource.com>
- */
-
-#include <linux/cache.h>
-#include <linux/mm.h>
-#include <linux/module.h>
-#include <linux/pci.h>
-#include <linux/spinlock.h>
-#include <linux/string.h>
-#include <linux/types.h>
-#include <linux/ctype.h>
-#include <linux/init.h>
-#include <linux/bootmem.h>
-#include <linux/highmem.h>
-#include <asm/io.h>
-#include <asm/pci.h>
-#include <asm/dma.h>
-#include <asm/uaccess.h>
-#include <xen/gnttab.h>
-#include <xen/interface/memory.h>
-#include <asm-i386/mach-xen/asm/gnttab_dma.h>
-
-int swiotlb;
-EXPORT_SYMBOL(swiotlb);
-
-#define OFFSET(val,align) ((unsigned long)((val) & ( (align) - 1)))
-
-/*
- * Maximum allowable number of contiguous slabs to map,
- * must be a power of 2. What is the appropriate value ?
- * The complexity of {map,unmap}_single is linearly dependent on this value.
- */
-#define IO_TLB_SEGSIZE 128
-
-/*
- * log of the size of each IO TLB slab. The number of slabs is command line
- * controllable.
- */
-#define IO_TLB_SHIFT 11
-
-int swiotlb_force;
-
-static char *iotlb_virt_start;
-static unsigned long iotlb_nslabs;
-
-/*
- * Used to do a quick range check in swiotlb_unmap_single and
- * swiotlb_sync_single_*, to see if the memory was in fact allocated by this
- * API.
- */
-static unsigned long iotlb_pfn_start, iotlb_pfn_end;
-
-/* Does the given dma address reside within the swiotlb aperture? */
-static inline int in_swiotlb_aperture(dma_addr_t dev_addr)
-{
- unsigned long pfn = mfn_to_local_pfn(dev_addr >> PAGE_SHIFT);
- return (pfn_valid(pfn)
- && (pfn >= iotlb_pfn_start)
- && (pfn < iotlb_pfn_end));
-}
-
-/*
- * When the IOMMU overflows we return a fallback buffer. This sets the size.
- */
-static unsigned long io_tlb_overflow = 32*1024;
-
-void *io_tlb_overflow_buffer;
-
-/*
- * This is a free list describing the number of free entries available from
- * each index
- */
-static unsigned int *io_tlb_list;
-static unsigned int io_tlb_index;
-
-/*
- * We need to save away the original address corresponding to a mapped entry
- * for the sync operations.
- */
-static struct phys_addr {
- struct page *page;
- unsigned int offset;
-} *io_tlb_orig_addr;
-
-/*
- * Protect the above data structures in the map and unmap calls
- */
-static DEFINE_SPINLOCK(io_tlb_lock);
-
-static unsigned int dma_bits;
-static unsigned int __initdata max_dma_bits = 32;
-static int __init
-setup_dma_bits(char *str)
-{
- max_dma_bits = simple_strtoul(str, NULL, 0);
- return 0;
-}
-__setup("dma_bits=", setup_dma_bits);
-
-static int __init
-setup_io_tlb_npages(char *str)
-{
- /* Unlike ia64, the size is aperture in megabytes, not 'slabs'! */
- if (isdigit(*str)) {
- iotlb_nslabs = simple_strtoul(str, &str, 0) <<
- (20 - IO_TLB_SHIFT);
- iotlb_nslabs = ALIGN(iotlb_nslabs, IO_TLB_SEGSIZE);
- /* Round up to power of two (xen_create_contiguous_region). */
- while (iotlb_nslabs & (iotlb_nslabs-1))
- iotlb_nslabs += iotlb_nslabs & ~(iotlb_nslabs-1);
- }
- if (*str == ',')
- ++str;
- /*
- * NB. 'force' enables the swiotlb, but doesn't force its use for
- * every DMA like it does on native Linux. 'off' forcibly disables
- * use of the swiotlb.
- */
- if (!strcmp(str, "force"))
- swiotlb_force = 1;
- else if (!strcmp(str, "off"))
- swiotlb_force = -1;
- return 1;
-}
-__setup("swiotlb=", setup_io_tlb_npages);
-/* make io_tlb_overflow tunable too? */
-
-/*
- * Statically reserve bounce buffer space and initialize bounce buffer data
- * structures for the software IO TLB used to implement the PCI DMA API.
- */
-void
-swiotlb_init_with_default_size (size_t default_size)
-{
- unsigned long i, bytes;
- int rc;
-
- if (!iotlb_nslabs) {
- iotlb_nslabs = (default_size >> IO_TLB_SHIFT);
- iotlb_nslabs = ALIGN(iotlb_nslabs, IO_TLB_SEGSIZE);
- /* Round up to power of two (xen_create_contiguous_region). */
- while (iotlb_nslabs & (iotlb_nslabs-1))
- iotlb_nslabs += iotlb_nslabs & ~(iotlb_nslabs-1);
- }
-
- bytes = iotlb_nslabs * (1UL << IO_TLB_SHIFT);
-
- /*
- * Get IO TLB memory from the low pages
- */
- iotlb_virt_start = alloc_bootmem_low_pages(bytes);
- if (!iotlb_virt_start)
- panic("Cannot allocate SWIOTLB buffer!\n");
-
- dma_bits = get_order(IO_TLB_SEGSIZE << IO_TLB_SHIFT) + PAGE_SHIFT;
- for (i = 0; i < iotlb_nslabs; i += IO_TLB_SEGSIZE) {
- do {
- rc = xen_create_contiguous_region(
- (unsigned long)iotlb_virt_start + (i << IO_TLB_SHIFT),
- get_order(IO_TLB_SEGSIZE << IO_TLB_SHIFT),
- dma_bits);
- } while (rc && dma_bits++ < max_dma_bits);
- if (rc) {
- if (i == 0)
- panic("No suitable physical memory available for SWIOTLB buffer!\n"
- "Use dom0_mem Xen boot parameter to reserve\n"
- "some DMA memory (e.g., dom0_mem=-128M).\n");
- iotlb_nslabs = i;
- i <<= IO_TLB_SHIFT;
- free_bootmem(__pa(iotlb_virt_start + i), bytes - i);
- bytes = i;
- for (dma_bits = 0; i > 0; i -= IO_TLB_SEGSIZE << IO_TLB_SHIFT) {
- unsigned int bits = fls64(virt_to_bus(iotlb_virt_start + i - 1));
-
- if (bits > dma_bits)
- dma_bits = bits;
- }
- break;
- }
- }
-
- /*
- * Allocate and initialize the free list array. This array is used
- * to find contiguous free memory regions of size up to IO_TLB_SEGSIZE.
- */
- io_tlb_list = alloc_bootmem(iotlb_nslabs * sizeof(int));
- for (i = 0; i < iotlb_nslabs; i++)
- io_tlb_list[i] = IO_TLB_SEGSIZE - OFFSET(i, IO_TLB_SEGSIZE);
- io_tlb_index = 0;
- io_tlb_orig_addr = alloc_bootmem(
- iotlb_nslabs * sizeof(*io_tlb_orig_addr));
-
- /*
- * Get the overflow emergency buffer
- */
- io_tlb_overflow_buffer = alloc_bootmem_low(io_tlb_overflow);
- if (!io_tlb_overflow_buffer)
- panic("Cannot allocate SWIOTLB overflow buffer!\n");
-
- do {
- rc = xen_create_contiguous_region(
- (unsigned long)io_tlb_overflow_buffer,
- get_order(io_tlb_overflow),
- dma_bits);
- } while (rc && dma_bits++ < max_dma_bits);
- if (rc)
- panic("No suitable physical memory available for SWIOTLB overflow buffer!\n");
-
- iotlb_pfn_start = __pa(iotlb_virt_start) >> PAGE_SHIFT;
- iotlb_pfn_end = iotlb_pfn_start + (bytes >> PAGE_SHIFT);
-
- printk(KERN_INFO "Software IO TLB enabled: \n"
- " Aperture: %lu megabytes\n"
- " Kernel range: %p - %p\n"
- " Address size: %u bits\n",
- bytes >> 20,
- iotlb_virt_start, iotlb_virt_start + bytes,
- dma_bits);
-}
-
-void
-swiotlb_init(void)
-{
- long ram_end;
- size_t defsz = 64 * (1 << 20); /* 64MB default size */
-
- if (swiotlb_force == 1) {
- swiotlb = 1;
- } else if ((swiotlb_force != -1) &&
- is_running_on_xen() &&
- is_initial_xendomain()) {
- /* Domain 0 always has a swiotlb. */
- ram_end = HYPERVISOR_memory_op(XENMEM_maximum_ram_page, NULL);
- if (ram_end <= 0x7ffff)
- defsz = 2 * (1 << 20); /* 2MB on <2GB on systems. */
- swiotlb = 1;
- }
-
- if (swiotlb)
- swiotlb_init_with_default_size(defsz);
- else
- printk(KERN_INFO "Software IO TLB disabled\n");
-}
-
-/*
- * We use __copy_to_user_inatomic to transfer to the host buffer because the
- * buffer may be mapped read-only (e.g, in blkback driver) but lower-level
- * drivers map the buffer for DMA_BIDIRECTIONAL access. This causes an
- * unnecessary copy from the aperture to the host buffer, and a page fault.
- */
-static void
-__sync_single(struct phys_addr buffer, char *dma_addr, size_t size, int dir)
-{
- if (PageHighMem(buffer.page)) {
- size_t len, bytes;
- char *dev, *host, *kmp;
- len = size;
- while (len != 0) {
- unsigned long flags;
-
- if (((bytes = len) + buffer.offset) > PAGE_SIZE)
- bytes = PAGE_SIZE - buffer.offset;
- local_irq_save(flags); /* protects KM_BOUNCE_READ */
- kmp = kmap_atomic(buffer.page, KM_BOUNCE_READ);
- dev = dma_addr + size - len;
- host = kmp + buffer.offset;
- if (dir == DMA_FROM_DEVICE) {
- if (__copy_to_user_inatomic(host, dev, bytes))
- /* inaccessible */;
- } else
- memcpy(dev, host, bytes);
- kunmap_atomic(kmp, KM_BOUNCE_READ);
- local_irq_restore(flags);
- len -= bytes;
- buffer.page++;
- buffer.offset = 0;
- }
- } else {
- char *host = (char *)phys_to_virt(
- page_to_pseudophys(buffer.page)) + buffer.offset;
- if (dir == DMA_FROM_DEVICE) {
- if (__copy_to_user_inatomic(host, dma_addr, size))
- /* inaccessible */;
- } else if (dir == DMA_TO_DEVICE)
- memcpy(dma_addr, host, size);
- }
-}
-
-/*
- * Allocates bounce buffer and returns its kernel virtual address.
- */
-static void *
-map_single(struct device *hwdev, struct phys_addr buffer, size_t size, int dir)
-{
- unsigned long flags;
- char *dma_addr;
- unsigned int nslots, stride, index, wrap;
- struct phys_addr slot_buf;
- int i;
-
- /*
- * For mappings greater than a page, we limit the stride (and
- * hence alignment) to a page size.
- */
- nslots = ALIGN(size, 1 << IO_TLB_SHIFT) >> IO_TLB_SHIFT;
- if (size > PAGE_SIZE)
- stride = (1 << (PAGE_SHIFT - IO_TLB_SHIFT));
- else
- stride = 1;
-
- BUG_ON(!nslots);
-
- /*
- * Find suitable number of IO TLB entries size that will fit this
- * request and allocate a buffer from that IO TLB pool.
- */
- spin_lock_irqsave(&io_tlb_lock, flags);
- {
- wrap = index = ALIGN(io_tlb_index, stride);
-
- if (index >= iotlb_nslabs)
- wrap = index = 0;
-
- do {
- /*
- * If we find a slot that indicates we have 'nslots'
- * number of contiguous buffers, we allocate the
- * buffers from that slot and mark the entries as '0'
- * indicating unavailable.
- */
- if (io_tlb_list[index] >= nslots) {
- int count = 0;
-
- for (i = index; i < (int)(index + nslots); i++)
- io_tlb_list[i] = 0;
- for (i = index - 1;
- (OFFSET(i, IO_TLB_SEGSIZE) !=
- IO_TLB_SEGSIZE -1) && io_tlb_list[i];
- i--)
- io_tlb_list[i] = ++count;
- dma_addr = iotlb_virt_start +
- (index << IO_TLB_SHIFT);
-
- /*
- * Update the indices to avoid searching in
- * the next round.
- */
- io_tlb_index =
- ((index + nslots) < iotlb_nslabs
- ? (index + nslots) : 0);
-
- goto found;
- }
- index += stride;
- if (index >= iotlb_nslabs)
- index = 0;
- } while (index != wrap);
-
- spin_unlock_irqrestore(&io_tlb_lock, flags);
- return NULL;
- }
- found:
- spin_unlock_irqrestore(&io_tlb_lock, flags);
-
- /*
- * Save away the mapping from the original address to the DMA address.
- * This is needed when we sync the memory. Then we sync the buffer if
- * needed.
- */
- slot_buf = buffer;
- for (i = 0; i < nslots; i++) {
- slot_buf.page += slot_buf.offset >> PAGE_SHIFT;
- slot_buf.offset &= PAGE_SIZE - 1;
- io_tlb_orig_addr[index+i] = slot_buf;
- slot_buf.offset += 1 << IO_TLB_SHIFT;
- }
- if ((dir == DMA_TO_DEVICE) || (dir == DMA_BIDIRECTIONAL))
- __sync_single(buffer, dma_addr, size, DMA_TO_DEVICE);
-
- return dma_addr;
-}
-
-struct phys_addr dma_addr_to_phys_addr(char *dma_addr)
-{
- int index = (dma_addr - iotlb_virt_start) >> IO_TLB_SHIFT;
- struct phys_addr buffer = io_tlb_orig_addr[index];
- buffer.offset += (long)dma_addr & ((1 << IO_TLB_SHIFT) - 1);
- buffer.page += buffer.offset >> PAGE_SHIFT;
- buffer.offset &= PAGE_SIZE - 1;
- return buffer;
-}
-
-/*
- * dma_addr is the kernel virtual address of the bounce buffer to unmap.
- */
-static void
-unmap_single(struct device *hwdev, char *dma_addr, size_t size, int dir)
-{
- unsigned long flags;
- int i, count, nslots = ALIGN(size, 1 << IO_TLB_SHIFT) >> IO_TLB_SHIFT;
- int index = (dma_addr - iotlb_virt_start) >> IO_TLB_SHIFT;
- struct phys_addr buffer = dma_addr_to_phys_addr(dma_addr);
-
- /*
- * First, sync the memory before unmapping the entry
- */
- if ((dir == DMA_FROM_DEVICE) || (dir == DMA_BIDIRECTIONAL))
- __sync_single(buffer, dma_addr, size, DMA_FROM_DEVICE);
-
- /*
- * Return the buffer to the free list by setting the corresponding
- * entries to indicate the number of contigous entries available.
- * While returning the entries to the free list, we merge the entries
- * with slots below and above the pool being returned.
- */
- spin_lock_irqsave(&io_tlb_lock, flags);
- {
- count = ((index + nslots) < ALIGN(index + 1, IO_TLB_SEGSIZE) ?
- io_tlb_list[index + nslots] : 0);
- /*
- * Step 1: return the slots to the free list, merging the
- * slots with superceeding slots
- */
- for (i = index + nslots - 1; i >= index; i--)
- io_tlb_list[i] = ++count;
- /*
- * Step 2: merge the returned slots with the preceding slots,
- * if available (non zero)
- */
- for (i = index - 1;
- (OFFSET(i, IO_TLB_SEGSIZE) !=
- IO_TLB_SEGSIZE -1) && io_tlb_list[i];
- i--)
- io_tlb_list[i] = ++count;
- }
- spin_unlock_irqrestore(&io_tlb_lock, flags);
-}
-
-static void
-sync_single(struct device *hwdev, char *dma_addr, size_t size, int dir)
-{
- struct phys_addr buffer = dma_addr_to_phys_addr(dma_addr);
- BUG_ON((dir != DMA_FROM_DEVICE) && (dir != DMA_TO_DEVICE));
- __sync_single(buffer, dma_addr, size, dir);
-}
-
-static void
-swiotlb_full(struct device *dev, size_t size, int dir, int do_panic)
-{
- /*
- * Ran out of IOMMU space for this operation. This is very bad.
- * Unfortunately the drivers cannot handle this operation properly.
- * unless they check for pci_dma_mapping_error (most don't)
- * When the mapping is small enough return a static buffer to limit
- * the damage, or panic when the transfer is too big.
- */
- printk(KERN_ERR "PCI-DMA: Out of SW-IOMMU space for %lu bytes at "
- "device %s\n", (unsigned long)size, dev ? dev->bus_id : "?");
-
- if (size > io_tlb_overflow && do_panic) {
- if (dir == PCI_DMA_FROMDEVICE || dir == PCI_DMA_BIDIRECTIONAL)
- panic("PCI-DMA: Memory would be corrupted\n");
- if (dir == PCI_DMA_TODEVICE || dir == PCI_DMA_BIDIRECTIONAL)
- panic("PCI-DMA: Random memory would be DMAed\n");
- }
-}
-
-/*
- * Map a single buffer of the indicated size for DMA in streaming mode. The
- * PCI address to use is returned.
- *
- * Once the device is given the dma address, the device owns this memory until
- * either swiotlb_unmap_single or swiotlb_dma_sync_single is performed.
- */
-dma_addr_t
-swiotlb_map_single(struct device *hwdev, void *ptr, size_t size, int dir)
-{
- dma_addr_t dev_addr = gnttab_dma_map_page(virt_to_page(ptr)) +
- offset_in_page(ptr);
- void *map;
- struct phys_addr buffer;
-
- BUG_ON(dir == DMA_NONE);
-
- /*
- * If the pointer passed in happens to be in the device's DMA window,
- * we can safely return the device addr and not worry about bounce
- * buffering it.
- */
- if (!range_straddles_page_boundary(__pa(ptr), size) &&
- !address_needs_mapping(hwdev, dev_addr))
- return dev_addr;
-
- /*
- * Oh well, have to allocate and map a bounce buffer.
- */
- gnttab_dma_unmap_page(dev_addr);
- buffer.page = virt_to_page(ptr);
- buffer.offset = (unsigned long)ptr & ~PAGE_MASK;
- map = map_single(hwdev, buffer, size, dir);
- if (!map) {
- swiotlb_full(hwdev, size, dir, 1);
- map = io_tlb_overflow_buffer;
- }
-
- dev_addr = virt_to_bus(map);
- return dev_addr;
-}
-
-/*
- * Unmap a single streaming mode DMA translation. The dma_addr and size must
- * match what was provided for in a previous swiotlb_map_single call. All
- * other usages are undefined.
- *
- * After this call, reads by the cpu to the buffer are guaranteed to see
- * whatever the device wrote there.
- */
-void
-swiotlb_unmap_single(struct device *hwdev, dma_addr_t dev_addr, size_t size,
- int dir)
-{
- BUG_ON(dir == DMA_NONE);
- if (in_swiotlb_aperture(dev_addr))
- unmap_single(hwdev, bus_to_virt(dev_addr), size, dir);
- else
- gnttab_dma_unmap_page(dev_addr);
-}
-
-/*
- * Make physical memory consistent for a single streaming mode DMA translation
- * after a transfer.
- *
- * If you perform a swiotlb_map_single() but wish to interrogate the buffer
- * using the cpu, yet do not wish to teardown the PCI dma mapping, you must
- * call this function before doing so. At the next point you give the PCI dma
- * address back to the card, you must first perform a
- * swiotlb_dma_sync_for_device, and then the device again owns the buffer
- */
-void
-swiotlb_sync_single_for_cpu(struct device *hwdev, dma_addr_t dev_addr,
- size_t size, int dir)
-{
- BUG_ON(dir == DMA_NONE);
- if (in_swiotlb_aperture(dev_addr))
- sync_single(hwdev, bus_to_virt(dev_addr), size, dir);
-}
-
-void
-swiotlb_sync_single_for_device(struct device *hwdev, dma_addr_t dev_addr,
- size_t size, int dir)
-{
- BUG_ON(dir == DMA_NONE);
- if (in_swiotlb_aperture(dev_addr))
- sync_single(hwdev, bus_to_virt(dev_addr), size, dir);
-}
-
-/*
- * Map a set of buffers described by scatterlist in streaming mode for DMA.
- * This is the scatter-gather version of the above swiotlb_map_single
- * interface. Here the scatter gather list elements are each tagged with the
- * appropriate dma address and length. They are obtained via
- * sg_dma_{address,length}(SG).
- *
- * NOTE: An implementation may be able to use a smaller number of
- * DMA address/length pairs than there are SG table elements.
- * (for example via virtual mapping capabilities)
- * The routine returns the number of addr/length pairs actually
- * used, at most nents.
- *
- * Device ownership issues as mentioned above for swiotlb_map_single are the
- * same here.
- */
-int
-swiotlb_map_sg(struct device *hwdev, struct scatterlist *sg, int nelems,
- int dir)
-{
- struct phys_addr buffer;
- dma_addr_t dev_addr;
- char *map;
- int i;
-
- BUG_ON(dir == DMA_NONE);
-
- for (i = 0; i < nelems; i++, sg++) {
- dev_addr = gnttab_dma_map_page(sg->page) + sg->offset;
-
- if (range_straddles_page_boundary(page_to_pseudophys(sg->page)
- + sg->offset, sg->length)
- || address_needs_mapping(hwdev, dev_addr)) {
- gnttab_dma_unmap_page(dev_addr);
- buffer.page = sg->page;
- buffer.offset = sg->offset;
- map = map_single(hwdev, buffer, sg->length, dir);
- if (!map) {
- /* Don't panic here, we expect map_sg users
- to do proper error handling. */
- swiotlb_full(hwdev, sg->length, dir, 0);
- swiotlb_unmap_sg(hwdev, sg - i, i, dir);
- sg[0].dma_length = 0;
- return 0;
- }
- sg->dma_address = (dma_addr_t)virt_to_bus(map);
- } else
- sg->dma_address = dev_addr;
- sg->dma_length = sg->length;
- }
- return nelems;
-}
-
-/*
- * Unmap a set of streaming mode DMA translations. Again, cpu read rules
- * concerning calls here are the same as for swiotlb_unmap_single() above.
- */
-void
-swiotlb_unmap_sg(struct device *hwdev, struct scatterlist *sg, int nelems,
- int dir)
-{
- int i;
-
- BUG_ON(dir == DMA_NONE);
-
- for (i = 0; i < nelems; i++, sg++)
- if (in_swiotlb_aperture(sg->dma_address))
- unmap_single(hwdev,
- (void *)bus_to_virt(sg->dma_address),
- sg->dma_length, dir);
- else
- gnttab_dma_unmap_page(sg->dma_address);
-}
-
-/*
- * Make physical memory consistent for a set of streaming mode DMA translations
- * after a transfer.
- *
- * The same as swiotlb_sync_single_* but for a scatter-gather list, same rules
- * and usage.
- */
-void
-swiotlb_sync_sg_for_cpu(struct device *hwdev, struct scatterlist *sg,
- int nelems, int dir)
-{
- int i;
-
- BUG_ON(dir == DMA_NONE);
-
- for (i = 0; i < nelems; i++, sg++)
- if (in_swiotlb_aperture(sg->dma_address))
- sync_single(hwdev,
- (void *)bus_to_virt(sg->dma_address),
- sg->dma_length, dir);
-}
-
-void
-swiotlb_sync_sg_for_device(struct device *hwdev, struct scatterlist *sg,
- int nelems, int dir)
-{
- int i;
-
- BUG_ON(dir == DMA_NONE);
-
- for (i = 0; i < nelems; i++, sg++)
- if (in_swiotlb_aperture(sg->dma_address))
- sync_single(hwdev,
- (void *)bus_to_virt(sg->dma_address),
- sg->dma_length, dir);
-}
-
-#ifdef CONFIG_HIGHMEM
-
-dma_addr_t
-swiotlb_map_page(struct device *hwdev, struct page *page,
- unsigned long offset, size_t size,
- enum dma_data_direction direction)
-{
- struct phys_addr buffer;
- dma_addr_t dev_addr;
- char *map;
-
- dev_addr = gnttab_dma_map_page(page) + offset;
- if (address_needs_mapping(hwdev, dev_addr)) {
- gnttab_dma_unmap_page(dev_addr);
- buffer.page = page;
- buffer.offset = offset;
- map = map_single(hwdev, buffer, size, direction);
- if (!map) {
- swiotlb_full(hwdev, size, direction, 1);
- map = io_tlb_overflow_buffer;
- }
- dev_addr = (dma_addr_t)virt_to_bus(map);
- }
-
- return dev_addr;
-}
-
-void
-swiotlb_unmap_page(struct device *hwdev, dma_addr_t dma_address,
- size_t size, enum dma_data_direction direction)
-{
- BUG_ON(direction == DMA_NONE);
- if (in_swiotlb_aperture(dma_address))
- unmap_single(hwdev, bus_to_virt(dma_address), size, direction);
- else
- gnttab_dma_unmap_page(dma_address);
-}
-
-#endif
-
-int
-swiotlb_dma_mapping_error(dma_addr_t dma_addr)
-{
- return (dma_addr == virt_to_bus(io_tlb_overflow_buffer));
-}
-
-/*
- * Return whether the given PCI device DMA address mask can be supported
- * properly. For example, if your device can only drive the low 24-bits
- * during PCI bus mastering, then you would pass 0x00ffffff as the mask to
- * this function.
- */
-int
-swiotlb_dma_supported (struct device *hwdev, u64 mask)
-{
- return (mask >= ((1UL << dma_bits) - 1));
-}
-
-EXPORT_SYMBOL(swiotlb_init);
-EXPORT_SYMBOL(swiotlb_map_single);
-EXPORT_SYMBOL(swiotlb_unmap_single);
-EXPORT_SYMBOL(swiotlb_map_sg);
-EXPORT_SYMBOL(swiotlb_unmap_sg);
-EXPORT_SYMBOL(swiotlb_sync_single_for_cpu);
-EXPORT_SYMBOL(swiotlb_sync_single_for_device);
-EXPORT_SYMBOL(swiotlb_sync_sg_for_cpu);
-EXPORT_SYMBOL(swiotlb_sync_sg_for_device);
-EXPORT_SYMBOL(swiotlb_dma_mapping_error);
-EXPORT_SYMBOL(swiotlb_dma_supported);
diff -r e6974eba1bae -r b0bf9ba32bfe arch/i386/kernel/time-xen.c
--- a/arch/i386/kernel/time-xen.c Thu Jul 26 14:33:50 2007 -0600
+++ b/arch/i386/kernel/time-xen.c Fri Jul 27 08:15:50 2007 -0600
@@ -867,9 +867,9 @@ static int timer_resume(struct sys_devic
return 0;
}

+void time_resume(void);
static struct sysdev_class timer_sysclass = {
- .resume = timer_resume,
- .suspend = timer_suspend,
+ .resume = time_resume,
set_kset_name("timer"),
};

diff -r e6974eba1bae -r b0bf9ba32bfe arch/i386/mm/Makefile
--- a/arch/i386/mm/Makefile Thu Jul 26 14:33:50 2007 -0600
+++ b/arch/i386/mm/Makefile Fri Jul 27 08:15:50 2007 -0600
@@ -8,11 +8,4 @@ obj-$(CONFIG_HUGETLB_PAGE) += hugetlbpag
obj-$(CONFIG_HUGETLB_PAGE) += hugetlbpage.o
obj-$(CONFIG_HIGHMEM) += highmem.o
obj-$(CONFIG_BOOT_IOREMAP) += boot_ioremap.o
-
-ifdef CONFIG_XEN
-include $(srctree)/scripts/Makefile.xen
-
-obj-y += hypervisor.o
-
-obj-y := $(call cherrypickxen, $(obj-y))
-endif
+obj-$(CONFIG_XEN) += hypervisor.o
diff -r e6974eba1bae -r b0bf9ba32bfe arch/i386/mm/fault-xen.c
--- a/arch/i386/mm/fault-xen.c Thu Jul 26 14:33:50 2007 -0600
+++ b/arch/i386/mm/fault-xen.c Fri Jul 27 08:15:50 2007 -0600
@@ -739,18 +739,31 @@ void vmalloc_sync_all(void)
* problematic: insync can only get set bits added, and updates to
* start are only improving performance (without affecting correctness
* if undone).
- */
- static DECLARE_BITMAP(insync, PTRS_PER_PGD);
+ * XEN: To work on PAE, we need to iterate over PMDs rather than PGDs.
+ * This change works just fine with 2-level paging too.
+ */
+#define sync_index(a) ((a) >> PMD_SHIFT)
+ static DECLARE_BITMAP(insync, PTRS_PER_PGD*PTRS_PER_PMD);
static unsigned long start = TASK_SIZE;
unsigned long address;

BUILD_BUG_ON(TASK_SIZE & ~PGDIR_MASK);
- for (address = start; address >= TASK_SIZE; address += PGDIR_SIZE) {
- if (!test_bit(pgd_index(address), insync)) {
+ for (address = start;
+ address >= TASK_SIZE && address < hypervisor_virt_start;
+ address += 1UL << PMD_SHIFT) {
+ if (!test_bit(sync_index(address), insync)) {
unsigned long flags;
struct page *page;

spin_lock_irqsave(&pgd_lock, flags);
+ /*
+ * XEN: vmalloc_sync_one() failure path logic assumes
+ * pgd_list is non-empty.
+ */
+ if (unlikely(!pgd_list)) {
+ spin_unlock_irqrestore(&pgd_lock, flags);
+ return;
+ }
for (page = pgd_list; page; page =
(struct page *)page->index)
if (!vmalloc_sync_one(page_address(page),
@@ -760,10 +773,10 @@ void vmalloc_sync_all(void)
}
spin_unlock_irqrestore(&pgd_lock, flags);
if (!page)
- set_bit(pgd_index(address), insync);
+ set_bit(sync_index(address), insync);
}
- if (address == start && test_bit(pgd_index(address), insync))
- start = address + PGDIR_SIZE;
- }
-}
-#endif
+ if (address == start && test_bit(sync_index(address), insync))
+ start = address + (1UL << PMD_SHIFT);
+ }
+}
+#endif
diff -r e6974eba1bae -r b0bf9ba32bfe arch/i386/mm/ioremap-xen.c
--- a/arch/i386/mm/ioremap-xen.c Thu Jul 26 14:33:50 2007 -0600
+++ b/arch/i386/mm/ioremap-xen.c Fri Jul 27 08:15:50 2007 -0600
@@ -74,7 +74,7 @@ static int __direct_remap_pfn_range(stru

/*
* Fill in the machine address: PTE ptr is done later by
- * __direct_remap_area_pages().
+ * apply_to_page_range().
*/
v->val = __pte_val(pfn_pte_ma(mfn, prot));

diff -r e6974eba1bae -r b0bf9ba32bfe arch/i386/pci/Makefile
--- a/arch/i386/pci/Makefile Thu Jul 26 14:33:50 2007 -0600
+++ b/arch/i386/pci/Makefile Fri Jul 27 08:15:50 2007 -0600
@@ -16,8 +16,3 @@ pci-$(CONFIG_X86_NUMAQ) := numa.o irq.o
pci-$(CONFIG_X86_NUMAQ) := numa.o irq.o

obj-y += $(pci-y) common.o
-
-ifdef CONFIG_XEN
-include $(srctree)/scripts/Makefile.xen
-obj-y := $(call cherrypickxen, $(obj-y))
-endif
diff -r e6974eba1bae -r b0bf9ba32bfe arch/i386/power/cpu.c
--- a/arch/i386/power/cpu.c Thu Jul 26 14:33:50 2007 -0600
+++ b/arch/i386/power/cpu.c Fri Jul 27 08:15:50 2007 -0600
@@ -62,11 +62,12 @@ static void do_fpu_end(void)

static void fix_processor_context(void)
{
+#ifndef CONFIG_X86_NO_TSS
int cpu = smp_processor_id();
struct tss_struct * t = &per_cpu(init_tss, cpu);

set_tss_desc(cpu,t); /* This just modifies memory; should not be necessary. But... This is necessary, because 386 hardware has concept of busy TSS or some similar stupidity. */
-
+#endif
load_TR_desc(); /* This does ltr */
load_LDT(&current->active_mm->context); /* This does lldt */

diff -r e6974eba1bae -r b0bf9ba32bfe arch/powerpc/platforms/xen/hcall.c
--- a/arch/powerpc/platforms/xen/hcall.c Thu Jul 26 14:33:50 2007 -0600
+++ b/arch/powerpc/platforms/xen/hcall.c Fri Jul 27 08:15:50 2007 -0600
@@ -34,6 +34,7 @@
#include <xen/interface/event_channel.h>
#include <xen/interface/physdev.h>
#include <xen/interface/vcpu.h>
+#include <xen/interface/acm_ops.h>
#include <xen/interface/kexec.h>
#include <xen/public/privcmd.h>
#include <asm/hypercall.h>
@@ -680,6 +681,145 @@ static int xenppc_privcmd_event_channel_
return ret;
}

+static int xenppc_acmcmd_op(privcmd_hypercall_t *hypercall)
+{
+ xen_acmctl_t kern_op;
+ xen_acmctl_t __user *user_op = (xen_acmctl_t __user *)hypercall->arg[0];
+ void *op_desc;
+ void *desc = NULL, *desc2 = NULL, *desc3 = NULL, *desc4 = NULL;
+ int ret = 0;
+
+ if (copy_from_user(&kern_op, user_op, sizeof(xen_acmctl_t)))
+ return -EFAULT;
+
+ if (kern_op.interface_version != ACM_INTERFACE_VERSION) {
+ printk(KERN_WARNING "%s: %s %x != %x\n", __func__, current->comm,
+ kern_op.interface_version, ACM_INTERFACE_VERSION);
+ return -EACCES;
+ }
+
+ op_desc = xencomm_map(&kern_op, sizeof(xen_acmctl_t));
+ if (op_desc == NULL)
+ return -ENOMEM;
+
+ switch (kern_op.cmd) {
+ case ACMOP_setpolicy:
+ desc = xencomm_map(
+ xen_guest_handle(kern_op.u.setpolicy.pushcache),
+ kern_op.u.setpolicy.pushcache_size);
+
+ if (desc == NULL)
+ ret = -ENOMEM;
+
+ set_xen_guest_handle(kern_op.u.setpolicy.pushcache,
+ desc);
+ break;
+ case ACMOP_getpolicy:
+ desc = xencomm_map(
+ xen_guest_handle(kern_op.u.getpolicy.pullcache),
+ kern_op.u.getpolicy.pullcache_size);
+
+ if (desc == NULL)
+ ret = -ENOMEM;
+
+ set_xen_guest_handle(kern_op.u.getpolicy.pullcache,
+ desc);
+ break;
+ case ACMOP_dumpstats:
+ desc = xencomm_map(
+ xen_guest_handle(kern_op.u.dumpstats.pullcache),
+ kern_op.u.dumpstats.pullcache_size);
+
+ if (desc == NULL)
+ ret = -ENOMEM;
+
+ set_xen_guest_handle(kern_op.u.dumpstats.pullcache,
+ desc);
+ break;
+ case ACMOP_getssid:
+ desc = xencomm_map(
+ xen_guest_handle(kern_op.u.getssid.ssidbuf),
+ kern_op.u.getssid.ssidbuf_size);
+
+ if (desc == NULL)
+ ret = -ENOMEM;
+
+ set_xen_guest_handle(kern_op.u.getssid.ssidbuf,
+ desc);
+ break;
+ case ACMOP_getdecision:
+ break;
+ case ACMOP_chgpolicy:
+ desc = xencomm_map(
+ xen_guest_handle(kern_op.u.change_policy.policy_pushcache),
+ kern_op.u.change_policy.policy_pushcache_size);
+ desc2 = xencomm_map(
+ xen_guest_handle(kern_op.u.change_policy.del_array),
+ kern_op.u.change_policy.delarray_size);
+ desc3 = xencomm_map(
+ xen_guest_handle(kern_op.u.change_policy.chg_array),
+ kern_op.u.change_policy.chgarray_size);
+ desc4 = xencomm_map(
+ xen_guest_handle(kern_op.u.change_policy.err_array),
+ kern_op.u.change_policy.errarray_size);
+
+ if (desc == NULL || desc2 == NULL ||
+ desc3 == NULL || desc4 == NULL) {
+ ret = -ENOMEM;
+ goto out;
+ }
+
+ set_xen_guest_handle(kern_op.u.change_policy.policy_pushcache,
+ desc);
+ set_xen_guest_handle(kern_op.u.change_policy.del_array,
+ desc2);
+ set_xen_guest_handle(kern_op.u.change_policy.chg_array,
+ desc3);
+ set_xen_guest_handle(kern_op.u.change_policy.err_array,
+ desc4);
+ break;
+ case ACMOP_relabeldoms:
+ desc = xencomm_map(
+ xen_guest_handle(kern_op.u.relabel_doms.relabel_map),
+ kern_op.u.relabel_doms.relabel_map_size);
+ desc2 = xencomm_map(
+ xen_guest_handle(kern_op.u.relabel_doms.err_array),
+ kern_op.u.relabel_doms.errarray_size);
+
+ if (desc == NULL || desc2 == NULL) {
+ ret = -ENOMEM;
+ goto out;
+ }
+
+ set_xen_guest_handle(kern_op.u.relabel_doms.relabel_map,
+ desc);
+ set_xen_guest_handle(kern_op.u.relabel_doms.err_array,
+ desc2);
+ break;
+ default:
+ printk(KERN_ERR "%s: unknown/unsupported acmctl cmd %d\n",
+ __func__, kern_op.cmd);
+ return -ENOSYS;
+ }
+
+ if (ret)
+ goto out; /* error mapping the nested pointer */
+
+ ret = plpar_hcall_norets(XEN_MARK(hypercall->op),op_desc);
+
+ if (copy_to_user(user_op, &kern_op, sizeof(xen_acmctl_t)))
+ ret = -EFAULT;
+
+out:
+ xencomm_free(desc);
+ xencomm_free(desc2);
+ xencomm_free(desc3);
+ xencomm_free(desc4);
+ xencomm_free(op_desc);
+ return ret;
+}
+
+
/* The PowerPC hypervisor runs in a separate address space from Linux
* kernel/userspace, i.e. real mode. We must therefore translate userspace
* pointers to something the hypervisor can make sense of. */
@@ -698,11 +838,11 @@ int privcmd_hypercall(privcmd_hypercall_
return xenppc_privcmd_version(hypercall);
case __HYPERVISOR_event_channel_op:
return xenppc_privcmd_event_channel_op(hypercall);
+ case __HYPERVISOR_acm_op:
+ return xenppc_acmcmd_op(hypercall);
default:
printk(KERN_ERR "%s: unknown hcall (%ld)\n", __func__, hypercall->op);
- /* fallthru */
- /* below are the hcalls we know will fail and its ok */
- case __HYPERVISOR_acm_op:
+ /* maybe we'll get lucky and the hcall needs no translation. */
return plpar_hcall_norets(XEN_MARK(hypercall->op),
hypercall->arg[0],
hypercall->arg[1],
diff -r e6974eba1bae -r b0bf9ba32bfe arch/powerpc/platforms/xen/setup.c
--- a/arch/powerpc/platforms/xen/setup.c Thu Jul 26 14:33:50 2007 -0600
+++ b/arch/powerpc/platforms/xen/setup.c Fri Jul 27 08:15:50 2007 -0600
@@ -39,6 +39,7 @@
#include <asm/machdep.h>
#include <asm/hypervisor.h>
#include <asm/time.h>
+#include <asm/pmc.h>
#include "setup.h"

#ifdef DEBUG
@@ -287,6 +288,17 @@ unsigned int xenppc_get_irq(struct pt_re
* is nothing for do_IRQ() to do.
* XXX This means we aren't using IRQ stacks. */
return NO_IRQ;
+}
+
+static void xenppc_enable_pmcs(void)
+{
+ unsigned long set, reset;
+
+ power4_enable_pmcs();
+
+ set = 1UL << 63;
+ reset = 0;
+ plpar_hcall_norets(H_PERFMON, set, reset);
}

#ifdef CONFIG_KEXEC
@@ -316,6 +328,7 @@ define_machine(xen) {
.calibrate_decr = generic_calibrate_decr,
.progress = xenppc_progress,
.power_save = xenppc_power_save,
+ .enable_pmcs = xenppc_enable_pmcs,
#ifdef CONFIG_KEXEC
.machine_kexec = xen_machine_kexec,
.machine_kexec_prepare = xen_machine_kexec_prepare,
diff -r e6974eba1bae -r b0bf9ba32bfe arch/x86_64/Kconfig
--- a/arch/x86_64/Kconfig Thu Jul 26 14:33:50 2007 -0600
+++ b/arch/x86_64/Kconfig Fri Jul 27 08:15:50 2007 -0600
@@ -594,9 +594,7 @@ menu "Power management options"
menu "Power management options"
depends on !XEN_UNPRIVILEGED_GUEST

-if !X86_64_XEN
-source kernel/power/Kconfig
-endif
+source "kernel/power/Kconfig"

source "drivers/acpi/Kconfig"

diff -r e6974eba1bae -r b0bf9ba32bfe arch/x86_64/Makefile
--- a/arch/x86_64/Makefile Thu Jul 26 14:33:50 2007 -0600
+++ b/arch/x86_64/Makefile Fri Jul 27 08:15:50 2007 -0600
@@ -80,7 +80,6 @@ PHONY += bzImage bzlilo install archmrpr

ifdef CONFIG_XEN
CPPFLAGS := -Iinclude$(if $(KBUILD_SRC),2)/asm/mach-xen $(CPPFLAGS)
-head-y := arch/x86_64/kernel/head-xen.o arch/x86_64/kernel/head64-xen.o arch/x86_64/kernel/init_task.o
LDFLAGS_vmlinux := -e startup_64
boot := arch/i386/boot-xen
.PHONY: vmlinuz
diff -r e6974eba1bae -r b0bf9ba32bfe arch/x86_64/ia32/Makefile
--- a/arch/x86_64/ia32/Makefile Thu Jul 26 14:33:50 2007 -0600
+++ b/arch/x86_64/ia32/Makefile Fri Jul 27 08:15:50 2007 -0600
@@ -37,15 +37,8 @@ AFLAGS_vsyscall-syscall.o = -m32 -Wa,-32

ifdef CONFIG_XEN
AFLAGS_vsyscall-int80.o = -m32 -Wa,-32 -Iarch/i386/kernel
-CFLAGS_syscall32-xen.o += -DUSE_INT80
-AFLAGS_syscall32_syscall-xen.o += -DUSE_INT80
+CFLAGS_syscall32.o += -DUSE_INT80
+AFLAGS_syscall32_syscall.o += -DUSE_INT80

-$(obj)/syscall32_syscall-xen.o: \
- $(foreach F,int80 sysenter syscall,$(obj)/vsyscall-$F.so)
-
-targets := $(foreach F,int80 sysenter syscall,vsyscall-$F.o vsyscall-$F.so)
-
-include $(srctree)/scripts/Makefile.xen
-
-obj-y := $(call cherrypickxen, $(obj-y))
+$(obj)/syscall32_syscall.o: $(obj)/vsyscall-int80.so
endif
diff -r e6974eba1bae -r b0bf9ba32bfe arch/x86_64/kernel/Makefile
--- a/arch/x86_64/kernel/Makefile Thu Jul 26 14:33:50 2007 -0600
+++ b/arch/x86_64/kernel/Makefile Fri Jul 27 08:15:50 2007 -0600
@@ -57,18 +57,8 @@ msr-$(subst m,y,$(CONFIG_X86_MSR)) += .
msr-$(subst m,y,$(CONFIG_X86_MSR)) += ../../i386/kernel/msr.o
alternative-y += ../../i386/kernel/alternative.o

-ifdef CONFIG_XEN
-time-y += ../../i386/kernel/time-xen.o
-pci-dma-y += ../../i386/kernel/pci-dma-xen.o
-microcode-$(subst m,y,$(CONFIG_MICROCODE)) := ../../i386/kernel/microcode-xen.o
-quirks-y := ../../i386/kernel/quirks-xen.o
+time-$(CONFIG_XEN) += ../../i386/kernel/time.o
+pci-dma-$(CONFIG_XEN) += ../../i386/kernel/pci-dma.o

-n-obj-xen := i8259.o reboot.o i8237.o smpboot.o trampoline.o
-
-include $(srctree)/scripts/Makefile.xen
-
-obj-y := $(call filterxen, $(obj-y), $(n-obj-xen))
-obj-y := $(call cherrypickxen, $(obj-y))
-extra-y := $(call cherrypickxen, $(extra-y))
-%/head-xen.o %/head-xen.s: EXTRA_AFLAGS :=
-endif
+disabled-obj-$(CONFIG_XEN) := i8259.o reboot.o smpboot.o trampoline.o
+%/head.o %/head.s: $(if $(CONFIG_XEN),EXTRA_AFLAGS,dummy) :=
diff -r e6974eba1bae -r b0bf9ba32bfe arch/x86_64/kernel/acpi/Makefile
--- a/arch/x86_64/kernel/acpi/Makefile Thu Jul 26 14:33:50 2007 -0600
+++ b/arch/x86_64/kernel/acpi/Makefile Fri Jul 27 08:15:50 2007 -0600
@@ -8,3 +8,7 @@ endif
endif

boot-$(CONFIG_XEN) := ../../../i386/kernel/acpi/boot-xen.o
+ifdef CONFIG_XEN
+include $(srctree)/scripts/Makefile.xen
+obj-y := $(call cherrypickxen, $(obj-y))
+endif
diff -r e6974eba1bae -r b0bf9ba32bfe arch/x86_64/kernel/acpi/sleep-xen.c
--- /dev/null Thu Jan 01 00:00:00 1970 +0000
+++ b/arch/x86_64/kernel/acpi/sleep-xen.c Fri Jul 27 08:15:50 2007 -0600
@@ -0,0 +1,162 @@
+/*
+ * acpi.c - Architecture-Specific Low-Level ACPI Support
+ *
+ * Copyright (C) 2001, 2002 Paul Diefenbaugh <paul.s.diefenbaugh@intel.com>
+ * Copyright (C) 2001 Jun Nakajima <jun.nakajima@intel.com>
+ * Copyright (C) 2001 Patrick Mochel <mochel@osdl.org>
+ * Copyright (C) 2002 Andi Kleen, SuSE Labs (x86-64 port)
+ * Copyright (C) 2003 Pavel Machek, SuSE Labs
+ *
+ * ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
+ *
+ * ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+ */
+
+#include <linux/kernel.h>
+#include <linux/init.h>
+#include <linux/types.h>
+#include <linux/stddef.h>
+#include <linux/slab.h>
+#include <linux/pci.h>
+#include <linux/bootmem.h>
+#include <linux/acpi.h>
+#include <linux/cpumask.h>
+
+#include <asm/mpspec.h>
+#include <asm/io.h>
+#include <asm/apic.h>
+#include <asm/apicdef.h>
+#include <asm/page.h>
+#include <asm/pgtable.h>
+#include <asm/pgalloc.h>
+#include <asm/io_apic.h>
+#include <asm/proto.h>
+#include <asm/tlbflush.h>
+
+/* --------------------------------------------------------------------------
+ Low-Level Sleep Support
+ -------------------------------------------------------------------------- */
+
+#ifdef CONFIG_ACPI_SLEEP
+
+/* address in low memory of the wakeup routine. */
+unsigned long acpi_wakeup_address = 0;
+unsigned long acpi_video_flags;
+extern char wakeup_start, wakeup_end;
+
+extern unsigned long FASTCALL(acpi_copy_wakeup_routine(unsigned long));
+
+static pgd_t low_ptr;
+
+static void init_low_mapping(void)
+{
+ pgd_t *slot0 = pgd_offset(current->mm, 0UL);
+ low_ptr = *slot0;
+ set_pgd(slot0, *pgd_offset(current->mm, PAGE_OFFSET));
+ WARN_ON(num_online_cpus() != 1);
+ local_flush_tlb();
+}
+
+/**
+ * acpi_save_state_mem - save kernel state
+ *
+ * Create an identity mapped page table and copy the wakeup routine to
+ * low memory.
+ */
+int acpi_save_state_mem(void)
+{
+#ifndef CONFIG_ACPI_PV_SLEEP
+ init_low_mapping();
+
+ memcpy((void *)acpi_wakeup_address, &wakeup_start,
+ &wakeup_end - &wakeup_start);
+ acpi_copy_wakeup_routine(acpi_wakeup_address);
+#endif
+ return 0;
+}
+
+/*
+ * acpi_restore_state
+ */
+void acpi_restore_state_mem(void)
+{
+#ifndef CONFIG_ACPI_PV_SLEEP
+ set_pgd(pgd_offset(current->mm, 0UL), low_ptr);
+ local_flush_tlb();
+#endif
+}
+
+/**
+ * acpi_reserve_bootmem - do _very_ early ACPI initialisation
+ *
+ * We allocate a page in low memory for the wakeup
+ * routine for when we come back from a sleep state. The
+ * runtime allocator allows specification of <16M pages, but not
+ * <1M pages.
+ */
+void __init acpi_reserve_bootmem(void)
+{
+ acpi_wakeup_address = (unsigned long)alloc_bootmem_low(PAGE_SIZE);
+ if ((&wakeup_end - &wakeup_start) > PAGE_SIZE)
+ printk(KERN_CRIT
+ "ACPI: Wakeup code way too big, will crash on attempt to suspend\n");
+}
+
+static int __init acpi_sleep_setup(char *str)
+{
+ while ((str != NULL) && (*str != '\0')) {
+ if (strncmp(str, "s3_bios", 7) == 0)
+ acpi_video_flags = 1;
+ if (strncmp(str, "s3_mode", 7) == 0)
+ acpi_video_flags |= 2;
+ str = strchr(str, ',');
+ if (str != NULL)
+ str += strspn(str, ", \t");
+ }
+
+ return 1;
+}
+
+__setup("acpi_sleep=", acpi_sleep_setup);
+
+#ifdef CONFIG_ACPI_PV_SLEEP
+#include <asm/hypervisor.h>
+#include <xen/interface/platform.h>
+int acpi_notify_hypervisor_state(u8 sleep_state,
+ u32 pm1a_cnt, u32 pm1b_cnt)
+{
+ struct xen_platform_op op = {
+ .cmd = XENPF_enter_acpi_sleep,
+ .interface_version = XENPF_INTERFACE_VERSION,
+ .u = {
+ .enter_acpi_sleep = {
+ .pm1a_cnt_val = (u16)pm1a_cnt,
+ .pm1b_cnt_val = (u16)pm1b_cnt,
+ .sleep_state = sleep_state,
+ },
+ },
+ };
+
+ return HYPERVISOR_platform_op(&op);
+}
+#endif /* CONFIG_ACPI_PV_SLEEP */
+
+#endif /*CONFIG_ACPI_SLEEP */
+
+void acpi_pci_link_exit(void)
+{
+}
diff -r e6974eba1bae -r b0bf9ba32bfe arch/x86_64/kernel/head-xen.S
--- a/arch/x86_64/kernel/head-xen.S Thu Jul 26 14:33:50 2007 -0600
+++ b/arch/x86_64/kernel/head-xen.S Fri Jul 27 08:15:50 2007 -0600
@@ -37,6 +37,13 @@ startup_64:
pushq $0 # fake return address
jmp x86_64_start_kernel

+#ifdef CONFIG_ACPI_SLEEP
+.org 0xf00
+ .globl pGDT32
+pGDT32:
+ .word gdt_end-cpu_gdt_table-1
+ .long cpu_gdt_table-__START_KERNEL_map
+#endif
ENTRY(stext)
ENTRY(_stext)

@@ -95,6 +102,14 @@ NEXT_PAGE(hypercall_page)
CFI_ENDPROC

#undef NEXT_PAGE
+
+ .data
+/* Just dummy symbol to allow compilation. Not used in sleep path */
+#ifdef CONFIG_ACPI_SLEEP
+ .align PAGE_SIZE
+ENTRY(wakeup_level4_pgt)
+ .fill 512,8,0
+#endif

.data

diff -r e6974eba1bae -r b0bf9ba32bfe arch/x86_64/kernel/suspend.c
--- a/arch/x86_64/kernel/suspend.c Thu Jul 26 14:33:50 2007 -0600
+++ b/arch/x86_64/kernel/suspend.c Fri Jul 27 08:15:50 2007 -0600
@@ -114,12 +114,14 @@ void restore_processor_state(void)

void fix_processor_context(void)
{
+#ifndef CONFIG_X86_NO_TSS
int cpu = smp_processor_id();
struct tss_struct *t = &per_cpu(init_tss, cpu);

set_tss_desc(cpu,t); /* This just modifies memory; should not be neccessary. But... This is neccessary, because 386 hardware has concept of busy TSS or some similar stupidity. */

cpu_gdt(cpu)[GDT_ENTRY_TSS].type = 9;
+#endif

syscall_init(); /* This sets MSR_*STAR and related */
load_TR_desc(); /* This does ltr */
diff -r e6974eba1bae -r b0bf9ba32bfe arch/x86_64/mm/Makefile
--- a/arch/x86_64/mm/Makefile Thu Jul 26 14:33:50 2007 -0600
+++ b/arch/x86_64/mm/Makefile Fri Jul 27 08:15:50 2007 -0600
@@ -7,15 +7,8 @@ obj-$(CONFIG_NUMA) += numa.o
obj-$(CONFIG_NUMA) += numa.o
obj-$(CONFIG_K8_NUMA) += k8topology.o
obj-$(CONFIG_ACPI_NUMA) += srat.o
+obj-$(CONFIG_XEN) += hypervisor.o

hugetlbpage-y = ../../i386/mm/hugetlbpage.o
-
-ifdef CONFIG_XEN
-include $(srctree)/scripts/Makefile.xen
-
-ioremap-y += ../../i386/mm/ioremap-xen.o
-hypervisor-y += ../../i386/mm/hypervisor.o
-obj-y += hypervisor.o
-
-obj-y := $(call cherrypickxen, $(obj-y))
-endif
+ioremap-$(CONFIG_XEN) := ../../i386/mm/ioremap.o
+hypervisor-y := ../../i386/mm/hypervisor.o
diff -r e6974eba1bae -r b0bf9ba32bfe arch/x86_64/pci/Makefile
--- a/arch/x86_64/pci/Makefile Thu Jul 26 14:33:50 2007 -0600
+++ b/arch/x86_64/pci/Makefile Fri Jul 27 08:15:50 2007 -0600
@@ -28,10 +28,3 @@ fixup-y += ../../i386/pci/fixup.o
fixup-y += ../../i386/pci/fixup.o
i386-y += ../../i386/pci/i386.o
init-y += ../../i386/pci/init.o
-
-ifdef CONFIG_XEN
-irq-y := ../../i386/pci/irq-xen.o
-include $(srctree)/scripts/Makefile.xen
-
-obj-y := $(call cherrypickxen, $(obj-y))
-endif
diff -r e6974eba1bae -r b0bf9ba32bfe buildconfigs/linux-defconfig_xen_x86_32
--- a/buildconfigs/linux-defconfig_xen_x86_32 Thu Jul 26 14:33:50 2007 -0600
+++ b/buildconfigs/linux-defconfig_xen_x86_32 Fri Jul 27 08:15:50 2007 -0600
@@ -202,11 +202,18 @@ CONFIG_ARCH_ENABLE_MEMORY_HOTPLUG=y
# Power management options (ACPI, APM)
#
CONFIG_PM=y
+# CONFIG_PM_LEGACY is not set
+CONFIG_PM_DEBUG=y
+# CONFIG_SOFTWARE_SUSPEND is not set
+CONFIG_SUSPEND_SMP=y

#
# ACPI (Advanced Configuration and Power Interface) Support
#
CONFIG_ACPI=y
+CONFIG_ACPI_SLEEP=y
+CONFIG_ACPI_SLEEP_PROC_FS=y
+# CONFIG_ACPI_SLEEP_PROC_SLEEP is not set
CONFIG_ACPI_AC=m
CONFIG_ACPI_BATTERY=m
CONFIG_ACPI_BUTTON=m
diff -r e6974eba1bae -r b0bf9ba32bfe buildconfigs/linux-defconfig_xen_x86_64
--- a/buildconfigs/linux-defconfig_xen_x86_64 Thu Jul 26 14:33:50 2007 -0600
+++ b/buildconfigs/linux-defconfig_xen_x86_64 Fri Jul 27 08:15:50 2007 -0600
@@ -161,11 +161,18 @@ CONFIG_GENERIC_PENDING_IRQ=y
# Power management options
#
CONFIG_PM=y
+# CONFIG_PM_LEGACY is not set
+CONFIG_PM_DEBUG=y
+# CONFIG_SOFTWARE_SUSPEND is not set
+CONFIG_SUSPEND_SMP=y

#
# ACPI (Advanced Configuration and Power Interface) Support
#
CONFIG_ACPI=y
+CONFIG_ACPI_SLEEP=y
+CONFIG_ACPI_SLEEP_PROC_FS=y
+# CONFIG_ACPI_SLEEP_PROC_SLEEP is not set
CONFIG_ACPI_AC=m
CONFIG_ACPI_BATTERY=m
CONFIG_ACPI_BUTTON=m
diff -r e6974eba1bae -r b0bf9ba32bfe drivers/acpi/Kconfig
--- a/drivers/acpi/Kconfig Thu Jul 26 14:33:50 2007 -0600
+++ b/drivers/acpi/Kconfig Fri Jul 27 08:15:50 2007 -0600
@@ -45,7 +45,7 @@ if ACPI

config ACPI_SLEEP
bool "Sleep States"
- depends on X86 && (!SMP || SUSPEND_SMP) && !XEN
+ depends on X86 && (!SMP || SUSPEND_SMP)
depends on PM
default y
---help---
@@ -363,6 +363,10 @@ config ACPI_SBS
A "Smart Battery" is quite old and quite rare compared
to today's ACPI "Control Method" battery.

+config ACPI_PV_SLEEP
+ bool
+ depends on X86 && XEN
+ default y
endif # ACPI

endmenu
diff -r e6974eba1bae -r b0bf9ba32bfe drivers/acpi/hardware/hwsleep.c
--- a/drivers/acpi/hardware/hwsleep.c Thu Jul 26 14:33:50 2007 -0600
+++ b/drivers/acpi/hardware/hwsleep.c Fri Jul 27 08:15:50 2007 -0600
@@ -327,6 +327,7 @@ acpi_status asmlinkage acpi_enter_sleep_

ACPI_FLUSH_CPU_CACHE();

+#ifndef CONFIG_ACPI_PV_SLEEP
status = acpi_hw_register_write(ACPI_MTX_DO_NOT_LOCK,
ACPI_REGISTER_PM1A_CONTROL,
PM1Acontrol);
@@ -337,6 +338,10 @@ acpi_status asmlinkage acpi_enter_sleep_
status = acpi_hw_register_write(ACPI_MTX_DO_NOT_LOCK,
ACPI_REGISTER_PM1B_CONTROL,
PM1Bcontrol);
+#else
+ status = acpi_notify_hypervisor_state(sleep_state,
+ PM1Acontrol, PM1Bcontrol);
+#endif
if (ACPI_FAILURE(status)) {
return_ACPI_STATUS(status);
}
diff -r e6974eba1bae -r b0bf9ba32bfe drivers/acpi/sleep/main.c
--- a/drivers/acpi/sleep/main.c Thu Jul 26 14:33:50 2007 -0600
+++ b/drivers/acpi/sleep/main.c Fri Jul 27 08:15:50 2007 -0600
@@ -91,7 +91,14 @@ static int acpi_pm_enter(suspend_state_t
break;

case PM_SUSPEND_MEM:
+#ifdef CONFIG_ACPI_PV_SLEEP
+ /* Hyperviosr will save and restore CPU context
+ * and then we can skip low level housekeeping here.
+ */
+ acpi_enter_sleep_state(acpi_state);
+#else
do_suspend_lowlevel();
+#endif
break;

case PM_SUSPEND_DISK:
@@ -145,10 +152,12 @@ static int acpi_pm_finish(suspend_state_
/* reset firmware waking vector */
acpi_set_firmware_waking_vector((acpi_physical_address) 0);

+#ifndef CONFIG_ACPI_PV_SLEEP
if (init_8259A_after_S1) {
printk("Broken toshiba laptop -> kicking interrupts\n");
init_8259A(0);
}
+#endif
return 0;
}

diff -r e6974eba1bae -r b0bf9ba32bfe drivers/acpi/sleep/poweroff.c
--- a/drivers/acpi/sleep/poweroff.c Thu Jul 26 14:33:50 2007 -0600
+++ b/drivers/acpi/sleep/poweroff.c Fri Jul 27 08:15:50 2007 -0600
@@ -20,6 +20,7 @@ int acpi_sleep_prepare(u32 acpi_state)
int acpi_sleep_prepare(u32 acpi_state)
{
#ifdef CONFIG_ACPI_SLEEP
+#ifndef CONFIG_ACPI_PV_SLEEP
/* do we have a wakeup address for S2 and S3? */
if (acpi_state == ACPI_STATE_S3) {
if (!acpi_wakeup_address) {
@@ -30,6 +31,7 @@ int acpi_sleep_prepare(u32 acpi_state)
acpi_wakeup_address));

}
+#endif
ACPI_FLUSH_CPU_CACHE();
acpi_enable_wakeup_device_prep(acpi_state);
#endif
diff -r e6974eba1bae -r b0bf9ba32bfe drivers/xen/balloon/balloon.c
--- a/drivers/xen/balloon/balloon.c Thu Jul 26 14:33:50 2007 -0600
+++ b/drivers/xen/balloon/balloon.c Fri Jul 27 08:15:50 2007 -0600
@@ -83,7 +83,15 @@ static unsigned long frame_list[.PAGE_SIZ

/* VM /proc information for memory */
extern unsigned long totalram_pages;
+
+#ifndef MODULE
extern unsigned long totalhigh_pages;
+#define inc_totalhigh_pages() (totalhigh_pages++)
+#define dec_totalhigh_pages() (totalhigh_pages--)
+#else
+#define inc_totalhigh_pages() ((void)0)
+#define dec_totalhigh_pages() ((void)0)
+#endif

/* List of ballooned pages, threaded through the mem_map array. */
static LIST_HEAD(ballooned_pages);
@@ -119,7 +127,7 @@ static void balloon_append(struct page *
if (PageHighMem(page)) {
list_add_tail(PAGE_TO_LIST(page), &ballooned_pages);
bs.balloon_high++;
- totalhigh_pages--;
+ dec_totalhigh_pages();
} else {
list_add(PAGE_TO_LIST(page), &ballooned_pages);
bs.balloon_low++;
@@ -139,7 +147,7 @@ static struct page *balloon_retrieve(voi

if (PageHighMem(page)) {
bs.balloon_high--;
- totalhigh_pages++;
+ inc_totalhigh_pages();
}
else
bs.balloon_low--;
diff -r e6974eba1bae -r b0bf9ba32bfe drivers/xen/blkfront/blkfront.c
--- a/drivers/xen/blkfront/blkfront.c Thu Jul 26 14:33:50 2007 -0600
+++ b/drivers/xen/blkfront/blkfront.c Fri Jul 27 08:15:50 2007 -0600
@@ -879,7 +879,7 @@ static struct xenbus_device_id blkfront_
{ "vbd" },
{ "" }
};
-
+MODULE_ALIAS("xen:vbd");

static struct xenbus_driver blkfront = {
.name = "vbd",
diff -r e6974eba1bae -r b0bf9ba32bfe drivers/xen/blktap/blktap.c
--- a/drivers/xen/blktap/blktap.c Thu Jul 26 14:33:50 2007 -0600
+++ b/drivers/xen/blktap/blktap.c Fri Jul 27 08:15:50 2007 -0600
@@ -653,6 +653,7 @@ static int blktap_mmap(struct file *filp
struct page **map;
int i;
tap_blkif_t *info = filp->private_data;
+ int ret;

if (info == NULL) {
WPRINTK("blktap: mmap, retrieving idx failed\n");
@@ -674,9 +675,14 @@ static int blktap_mmap(struct file *filp
info->user_vstart = info->rings_vstart + (RING_PAGES << PAGE_SHIFT);

/* Map the ring pages to the start of the region and reserve it. */
- if (remap_pfn_range(vma, vma->vm_start,
- __pa(info->ufe_ring.sring) >> PAGE_SHIFT,
- PAGE_SIZE, vma->vm_page_prot)) {
+ if (xen_feature(XENFEAT_auto_translated_physmap))
+ ret = vm_insert_page(vma, vma->vm_start,
+ virt_to_page(info->ufe_ring.sring));
+ else
+ ret = remap_pfn_range(vma, vma->vm_start,
+ __pa(info->ufe_ring.sring) >> PAGE_SHIFT,
+ PAGE_SIZE, vma->vm_page_prot);
+ if (ret) {
WPRINTK("Mapping user ring failed!\n");
goto fail;
}
diff -r e6974eba1bae -r b0bf9ba32bfe drivers/xen/char/mem.c
--- a/drivers/xen/char/mem.c Thu Jul 26 14:33:50 2007 -0600
+++ b/drivers/xen/char/mem.c Fri Jul 27 08:15:50 2007 -0600
@@ -26,12 +26,13 @@
#include <asm/io.h>
#include <asm/hypervisor.h>

-#ifndef ARCH_HAS_VALID_PHYS_ADDR_RANGE
-static inline int valid_phys_addr_range(unsigned long addr, size_t count)
+static inline int uncached_access(struct file *file)
{
- return 1;
+ if (file->f_flags & O_SYNC)
+ return 1;
+ /* Xen sets correct MTRR type on non-RAM for us. */
+ return 0;
}
-#endif

/*
* This funcion reads the *physical* memory. The f_pos points directly to the
@@ -44,9 +45,6 @@ static ssize_t read_mem(struct file * fi
ssize_t read = 0, sz;
void __iomem *v;

- if (!valid_phys_addr_range(p, count))
- return -EFAULT;
-
while (count > 0) {
/*
* Handle first page in case it's not aligned
@@ -58,7 +56,7 @@ static ssize_t read_mem(struct file * fi

sz = min_t(unsigned long, sz, count);

- v = xen_xlate_dev_mem_ptr(p, sz);
+ v = ioremap(p, sz);
if (IS_ERR(v) || v == NULL) {
/*
* Some programs (e.g., dmidecode) groove off into
@@ -75,7 +73,7 @@ static ssize_t read_mem(struct file * fi
}

ignored = copy_to_user(buf, v, sz);
- xen_xlate_dev_mem_ptr_unmap(v);
+ iounmap(v);
if (ignored)
return -EFAULT;
buf += sz;
@@ -95,9 +93,6 @@ static ssize_t write_mem(struct file * f
ssize_t written = 0, sz;
void __iomem *v;

- if (!valid_phys_addr_range(p, count))
- return -EFAULT;
-
while (count > 0) {
/*
* Handle first page in case it's not aligned
@@ -109,7 +104,7 @@ static ssize_t write_mem(struct file * f

sz = min_t(unsigned long, sz, count);

- v = xen_xlate_dev_mem_ptr(p, sz);
+ v = ioremap(p, sz);
if (v == NULL)
break;
if (IS_ERR(v)) {
@@ -119,7 +114,7 @@ static ssize_t write_mem(struct file * f
}

ignored = copy_from_user(v, buf, sz);
- xen_xlate_dev_mem_ptr_unmap(v);
+ iounmap(v);
if (ignored) {
written += sz - ignored;
if (written)
@@ -137,14 +132,6 @@ static ssize_t write_mem(struct file * f
}

#ifndef ARCH_HAS_DEV_MEM_MMAP_MEM
-static inline int uncached_access(struct file *file)
-{
- if (file->f_flags & O_SYNC)
- return 1;
- /* Xen sets correct MTRR type on non-RAM for us. */
- return 0;
-}
-
static int xen_mmap_mem(struct file * file, struct vm_area_struct * vma)
{
size_t size = vma->vm_end - vma->vm_start;
diff -r e6974eba1bae -r b0bf9ba32bfe drivers/xen/core/gnttab.c
--- a/drivers/xen/core/gnttab.c Thu Jul 26 14:33:50 2007 -0600
+++ b/drivers/xen/core/gnttab.c Fri Jul 27 08:15:50 2007 -0600
@@ -65,8 +65,6 @@ static struct grant_entry *shared;

static struct gnttab_free_callback *gnttab_free_callback_list;

-static DEFINE_SEQLOCK(gnttab_dma_lock);
-
static int gnttab_expand(unsigned int req_entries);

#define RPP (PAGE_SIZE / sizeof(grant_ref_t))
@@ -186,7 +184,7 @@ int gnttab_end_foreign_access_ref(grant_
nflags = shared[ref].flags;
do {
if ((flags = nflags) & (GTF_reading|GTF_writing)) {
- printk(KERN_ALERT "WARNING: g.e. still in use!\n");
+ printk(KERN_DEBUG "WARNING: g.e. still in use!\n");
return 0;
}
} while ((nflags = synch_cmpxchg_subword(&shared[ref].flags, flags, 0)) !=
@@ -206,7 +204,7 @@ void gnttab_end_foreign_access(grant_ref
} else {
/* XXX This needs to be fixed so that the ref and page are
placed on a list to be freed up later. */
- printk(KERN_WARNING
+ printk(KERN_DEBUG
"WARNING: leaking g.e. and page still in use!\n");
}
}
@@ -430,6 +428,8 @@ static inline unsigned int max_nr_grant_

#ifdef CONFIG_XEN

+static DEFINE_SEQLOCK(gnttab_dma_lock);
+
#ifdef CONFIG_X86
static int map_pte_fn(pte_t *pte, struct page *pmd_page,
unsigned long addr, void *data)
diff -r e6974eba1bae -r b0bf9ba32bfe drivers/xen/core/machine_reboot.c
--- a/drivers/xen/core/machine_reboot.c Thu Jul 26 14:33:50 2007 -0600
+++ b/drivers/xen/core/machine_reboot.c Fri Jul 27 08:15:50 2007 -0600
@@ -13,7 +13,6 @@
#include <asm/hypervisor.h>
#include <xen/xenbus.h>
#include <linux/cpu.h>
-#include <linux/kthread.h>
#include <xen/gnttab.h>
#include <xen/xencons.h>
#include <xen/cpu_hotplug.h>
diff -r e6974eba1bae -r b0bf9ba32bfe drivers/xen/fbfront/xenfb.c
--- a/drivers/xen/fbfront/xenfb.c Thu Jul 26 14:33:50 2007 -0600
+++ b/drivers/xen/fbfront/xenfb.c Fri Jul 27 08:15:50 2007 -0600
@@ -718,6 +718,7 @@ static struct xenbus_device_id xenfb_ids
{ "vfb" },
{ "" }
};
+MODULE_ALIAS("xen:vfb");

static struct xenbus_driver xenfb = {
.name = "vfb",
diff -r e6974eba1bae -r b0bf9ba32bfe drivers/xen/fbfront/xenkbd.c
--- a/drivers/xen/fbfront/xenkbd.c Thu Jul 26 14:33:50 2007 -0600
+++ b/drivers/xen/fbfront/xenkbd.c Fri Jul 27 08:15:50 2007 -0600
@@ -299,6 +299,7 @@ static struct xenbus_device_id xenkbd_id
{ "vkbd" },
{ "" }
};
+MODULE_ALIAS("xen:vkbd");

static struct xenbus_driver xenkbd = {
.name = "vkbd",
diff -r e6974eba1bae -r b0bf9ba32bfe drivers/xen/netfront/netfront.c
--- a/drivers/xen/netfront/netfront.c Thu Jul 26 14:33:50 2007 -0600
+++ b/drivers/xen/netfront/netfront.c Fri Jul 27 08:15:50 2007 -0600
@@ -65,10 +65,6 @@
#include <xen/gnttab.h>
#include <xen/hypercall.h>

-#ifdef HAVE_XEN_PLATFORM_COMPAT_H
-#include <xen/platform-compat.h>
-#endif
-
struct netfront_cb {
struct page *page;
unsigned offset;
@@ -1540,7 +1536,7 @@ static void netif_release_tx_bufs(struct
}
}

-static void netif_release_rx_bufs(struct netfront_info *np)
+static void netif_release_rx_bufs_flip(struct netfront_info *np)
{
struct mmu_update *mmu = np->rx_mmu;
struct multicall_entry *mcl = np->rx_mcl;
@@ -1549,11 +1545,6 @@ static void netif_release_rx_bufs(struct
unsigned long mfn;
int xfer = 0, noxfer = 0, unused = 0;
int id, ref, rc;
-
- if (np->copying_receiver) {
- WPRINTK("%s: fix me for copying receiver.\n", __FUNCTION__);
- return;
- }

skb_queue_head_init(&free_list);

@@ -1601,7 +1592,7 @@ static void netif_release_rx_bufs(struct
xfer++;
}

- IPRINTK("%s: %d xfer, %d noxfer, %d unused\n",
+ DPRINTK("%s: %d xfer, %d noxfer, %d unused\n",
__FUNCTION__, xfer, noxfer, unused);

if (xfer) {
@@ -1628,6 +1619,45 @@ static void netif_release_rx_bufs(struct
spin_unlock_bh(&np->rx_lock);
}

+static void netif_release_rx_bufs_copy(struct netfront_info *np)
+{
+ struct sk_buff *skb;
+ int i, ref;
+ int busy = 0, inuse = 0;
+
+ spin_lock_bh(&np->rx_lock);
+
+ for (i = 0; i < NET_RX_RING_SIZE; i++) {
+ ref = np->grant_rx_ref[i];
+
+ if (ref == GRANT_INVALID_REF)
+ continue;
+
+ inuse++;
+
+ skb = np->rx_skbs[i];
+
+ if (!gnttab_end_foreign_access_ref(ref, 0))
+ {
+ busy++;
+ continue;
+ }
+
+ gnttab_release_grant_reference(&np->gref_rx_head, ref);
+ np->grant_rx_ref[i] = GRANT_INVALID_REF;
+ add_id_to_freelist(np->rx_skbs, i);
+
+ skb_shinfo(skb)->nr_frags = 0;
+ dev_kfree_skb(skb);
+ }
+
+ if (busy)
+ DPRINTK("%s: Unable to release %d of %d inuse grant references out of %ld total.\n",
+ __FUNCTION__, busy, inuse, NET_RX_RING_SIZE);
+
+ spin_unlock_bh(&np->rx_lock);
+}
+
static int network_close(struct net_device *dev)
{
struct netfront_info *np = netdev_priv(dev);
@@ -1821,7 +1851,10 @@ static void netif_uninit(struct net_devi
{
struct netfront_info *np = netdev_priv(dev);
netif_release_tx_bufs(np);
- netif_release_rx_bufs(np);
+ if (np->copying_receiver)
+ netif_release_rx_bufs_copy(np);
+ else
+ netif_release_rx_bufs_flip(np);
gnttab_free_grant_references(np->gref_tx_head);
gnttab_free_grant_references(np->gref_rx_head);
}
@@ -2118,6 +2151,7 @@ static struct xenbus_device_id netfront_
{ "vif" },
{ "" }
};
+MODULE_ALIAS("xen:vif");


static struct xenbus_driver netfront = {
diff -r e6974eba1bae -r b0bf9ba32bfe drivers/xen/netfront/netfront.h
--- a/drivers/xen/netfront/netfront.h Thu Jul 26 14:33:50 2007 -0600
+++ b/drivers/xen/netfront/netfront.h Fri Jul 27 08:15:50 2007 -0600
@@ -44,6 +44,10 @@

#include <xen/xenbus.h>

+#ifdef HAVE_XEN_PLATFORM_COMPAT_H
+#include <xen/platform-compat.h>
+#endif
+
/*
* Function pointer table for hooks into a network acceleration
* plugin. These are called at appropriate points from the netfront
diff -r e6974eba1bae -r b0bf9ba32bfe drivers/xen/pciback/conf_space_header.c
--- a/drivers/xen/pciback/conf_space_header.c Thu Jul 26 14:33:50 2007 -0600
+++ b/drivers/xen/pciback/conf_space_header.c Fri Jul 27 08:15:50 2007 -0600
@@ -75,8 +75,15 @@ static int rom_write(struct pci_dev *dev
*/
if (value == ~PCI_ROM_ADDRESS_ENABLE)
bar->which = 1;
- else
+ else {
+ u32 tmpval;
+ pci_read_config_dword(dev, offset, &tmpval);
+ if (tmpval != bar->val && value == bar->val) {
+ /* Allow restoration of bar value. */
+ pci_write_config_dword(dev, offset, bar->val);
+ }
bar->which = 0;
+ }

/* Do we need to support enabling/disabling the rom address here? */

@@ -102,8 +109,15 @@ static int bar_write(struct pci_dev *dev
*/
if (value == ~0)
bar->which = 1;
- else
+ else {
+ u32 tmpval;
+ pci_read_config_dword(dev, offset, &tmpval);
+ if (tmpval != bar->val && value == bar->val) {
+ /* Allow restoration of bar value. */
+ pci_write_config_dword(dev, offset, bar->val);
+ }
bar->which = 0;
+ }

return 0;
}
diff -r e6974eba1bae -r b0bf9ba32bfe drivers/xen/pcifront/xenbus.c
--- a/drivers/xen/pcifront/xenbus.c Thu Jul 26 14:33:50 2007 -0600
+++ b/drivers/xen/pcifront/xenbus.c Fri Jul 27 08:15:50 2007 -0600
@@ -273,6 +273,7 @@ static struct xenbus_device_id xenpci_id
{"pci"},
{{0}},
};
+MODULE_ALIAS("xen:pci");

static struct xenbus_driver xenbus_pcifront_driver = {
.name = "pcifront",
diff -r e6974eba1bae -r b0bf9ba32bfe drivers/xen/privcmd/privcmd.c
--- a/drivers/xen/privcmd/privcmd.c Thu Jul 26 14:33:50 2007 -0600
+++ b/drivers/xen/privcmd/privcmd.c Fri Jul 27 08:15:50 2007 -0600
@@ -18,7 +18,6 @@
#include <linux/highmem.h>
#include <linux/pagemap.h>
#include <linux/seq_file.h>
-#include <linux/kthread.h>
#include <asm/hypervisor.h>

#include <asm/pgalloc.h>
diff -r e6974eba1bae -r b0bf9ba32bfe drivers/xen/xenbus/xenbus_probe.c
--- a/drivers/xen/xenbus/xenbus_probe.c Thu Jul 26 14:33:50 2007 -0600
+++ b/drivers/xen/xenbus/xenbus_probe.c Fri Jul 27 08:15:50 2007 -0600
@@ -42,7 +42,6 @@
#include <linux/fcntl.h>
#include <linux/mm.h>
#include <linux/notifier.h>
-#include <linux/kthread.h>
#include <linux/mutex.h>

#include <asm/io.h>
@@ -166,6 +165,30 @@ static int read_backend_details(struct x
return read_otherend_details(xendev, "backend-id", "backend");
}

+#if LINUX_VERSION_CODE >= KERNEL_VERSION(2,6,16)
+static int xenbus_uevent_frontend(struct device *dev, char **envp,
+ int num_envp, char *buffer, int buffer_size)
+{
+ struct xenbus_device *xdev;
+ int length = 0, i = 0;
+
+ if (dev == NULL)
+ return -ENODEV;
+ xdev = to_xenbus_device(dev);
+ if (xdev == NULL)
+ return -ENODEV;
+
+ /* stuff we want to pass to /sbin/hotplug */
+ add_uevent_var(envp, num_envp, &i, buffer, buffer_size, &length,
+ "XENBUS_TYPE=%s", xdev->devicetype);
+ add_uevent_var(envp, num_envp, &i, buffer, buffer_size, &length,
+ "XENBUS_PATH=%s", xdev->nodename);
+ add_uevent_var(envp, num_envp, &i, buffer, buffer_size, &length,
+ "MODALIAS=xen:%s", xdev->devicetype);
+
+ return 0;
+}
+#endif

/* Bus type for frontend drivers. */
static struct xen_bus_type xenbus_frontend = {
@@ -181,6 +204,7 @@ static struct xen_bus_type xenbus_fronte
.probe = xenbus_dev_probe,
.remove = xenbus_dev_remove,
.shutdown = xenbus_dev_shutdown,
+ .uevent = xenbus_uevent_frontend,
#endif
},
.dev = {
@@ -210,6 +234,7 @@ static void otherend_changed(struct xenb
DPRINTK("state is %d (%s), %s, %s", state, xenbus_strstate(state),
dev->otherend_watch.node, vec[XS_WATCH_PATH]);

+#if LINUX_VERSION_CODE >= KERNEL_VERSION(2,6,16)
/*
* Ignore xenbus transitions during shutdown. This prevents us doing
* work that can fail e.g., when the rootfs is gone.
@@ -223,6 +248,7 @@ static void otherend_changed(struct xenb
xenbus_frontend_closed(dev);
return;
}
+#endif

if (drv->otherend_changed)
drv->otherend_changed(dev, state);
@@ -1008,7 +1034,7 @@ static int is_disconnected_device(struct
return 0;

xendrv = to_xenbus_driver(dev->driver);
- return (xendev->state != XenbusStateConnected ||
+ return (xendev->state < XenbusStateConnected ||
(xendrv->is_ready && !xendrv->is_ready(xendev)));
}

@@ -1033,10 +1059,13 @@ static int print_device_status(struct de
/* Information only: is this too noisy? */
printk(KERN_INFO "XENBUS: Device with no driver: %s\n",
xendev->nodename);
- } else if (xendev->state != XenbusStateConnected) {
+ } else if (xendev->state < XenbusStateConnected) {
+ enum xenbus_state rstate = XenbusStateUnknown;
+ if (xendev->otherend)
+ rstate = xenbus_read_driver_state(xendev->otherend);
printk(KERN_WARNING "XENBUS: Timeout connecting "
- "to device: %s (state %d)\n",
- xendev->nodename, xendev->state);
+ "to device: %s (local state %d, remote state %d)\n",
+ xendev->nodename, xendev->state, rstate);
}

return 0;
@@ -1046,7 +1075,7 @@ static int ready_to_wait_for_devices;
static int ready_to_wait_for_devices;

/*
- * On a 10 second timeout, wait for all devices currently configured. We need
+ * On a 30-second timeout, wait for all devices currently configured. We need
* to do this to guarantee that the filesystems and / or network devices
* needed for boot are available, before we can allow the boot to proceed.
*
@@ -1061,17 +1090,29 @@ static int ready_to_wait_for_devices;
*/
static void wait_for_devices(struct xenbus_driver *xendrv)
{
- unsigned long timeout = jiffies + 10*HZ;
+ unsigned long start = jiffies;
struct device_driver *drv = xendrv ? &xendrv->driver : NULL;
+ unsigned int seconds_waited = 0;

if (!ready_to_wait_for_devices || !is_running_on_xen())
return;

while (exists_disconnected_device(drv)) {
- if (time_after(jiffies, timeout))
- break;
+ if (time_after(jiffies, start + (seconds_waited+5)*HZ)) {
+ if (!seconds_waited)
+ printk(KERN_WARNING "XENBUS: Waiting for "
+ "devices to initialise: ");
+ seconds_waited += 5;
+ printk("%us...", 30 - seconds_waited);
+ if (seconds_waited == 30)
+ break;
+ }
+
schedule_timeout_interruptible(HZ/10);
}
+
+ if (seconds_waited)
+ printk("\n");

bus_for_each_dev(&xenbus_frontend.bus, NULL, drv,
print_device_status);
diff -r e6974eba1bae -r b0bf9ba32bfe drivers/xen/xenbus/xenbus_probe_backend.c
--- a/drivers/xen/xenbus/xenbus_probe_backend.c Thu Jul 26 14:33:50 2007 -0600
+++ b/drivers/xen/xenbus/xenbus_probe_backend.c Fri Jul 27 08:15:50 2007 -0600
@@ -42,7 +42,6 @@
#include <linux/fcntl.h>
#include <linux/mm.h>
#include <linux/notifier.h>
-#include <linux/kthread.h>

#include <asm/io.h>
#include <asm/page.h>
diff -r e6974eba1bae -r b0bf9ba32bfe drivers/xen/xenbus/xenbus_xs.c
--- a/drivers/xen/xenbus/xenbus_xs.c Thu Jul 26 14:33:50 2007 -0600
+++ b/drivers/xen/xenbus/xenbus_xs.c Fri Jul 27 08:15:50 2007 -0600
@@ -51,6 +51,10 @@
#include <xen/platform-compat.h>
#endif

+#ifndef PF_NOFREEZE /* Old kernel (pre-2.6.6). */
+#define PF_NOFREEZE 0
+#endif
+
struct xs_stored_msg {
struct list_head list;

@@ -802,8 +806,8 @@ static int process_msg(void)
msg->u.watch.vec = split(body, msg->hdr.len,
&msg->u.watch.vec_size);
if (IS_ERR(msg->u.watch.vec)) {
+ err = PTR_ERR(msg->u.watch.vec);
kfree(msg);
- err = PTR_ERR(msg->u.watch.vec);
goto out;
}

diff -r e6974eba1bae -r b0bf9ba32bfe include/asm-i386/acpi.h
--- a/include/asm-i386/acpi.h Thu Jul 26 14:33:50 2007 -0600
+++ b/include/asm-i386/acpi.h Fri Jul 27 08:15:50 2007 -0600
@@ -177,6 +177,10 @@ extern unsigned long acpi_wakeup_address
/* early initialization routine */
extern void acpi_reserve_bootmem(void);

+#ifdef CONFIG_ACPI_PV_SLEEP
+extern int acpi_notify_hypervisor_state(u8 sleep_state,
+ u32 pm1a_cnt, u32 pm1b_cnt);
+#endif /* CONFIG_ACPI_PV_SLEEP */
#endif /*CONFIG_ACPI_SLEEP*/

extern u8 x86_acpiid_to_apicid[];
diff -r e6974eba1bae -r b0bf9ba32bfe include/asm-i386/mach-xen/asm/gnttab_dma.h
--- a/include/asm-i386/mach-xen/asm/gnttab_dma.h Thu Jul 26 14:33:50 2007 -0600
+++ b/include/asm-i386/mach-xen/asm/gnttab_dma.h Fri Jul 27 08:15:50 2007 -0600
@@ -30,7 +30,7 @@ static inline maddr_t gnttab_dma_map_pag
static inline maddr_t gnttab_dma_map_page(struct page *page)
{
__gnttab_dma_map_page(page);
- return page_to_bus(page);
+ return ((maddr_t)pfn_to_mfn(page_to_pfn(page)) << PAGE_SHIFT);
}

static inline void gnttab_dma_unmap_page(maddr_t maddr)
diff -r e6974eba1bae -r b0bf9ba32bfe include/asm-i386/mach-xen/asm/io.h
--- a/include/asm-i386/mach-xen/asm/io.h Thu Jul 26 14:33:50 2007 -0600
+++ b/include/asm-i386/mach-xen/asm/io.h Fri Jul 27 08:15:50 2007 -0600
@@ -53,8 +53,7 @@
* Convert a physical pointer to a virtual kernel pointer for /dev/mem
* access
*/
-#define xen_xlate_dev_mem_ptr(p, sz) ioremap(p, sz)
-#define xen_xlate_dev_mem_ptr_unmap(p) iounmap(p)
+#define xlate_dev_mem_ptr(p) __va(p)

/*
* Convert a virtual cached pointer to an uncached pointer
diff -r e6974eba1bae -r b0bf9ba32bfe include/asm-powerpc/io.h
--- a/include/asm-powerpc/io.h Thu Jul 26 14:33:50 2007 -0600
+++ b/include/asm-powerpc/io.h Fri Jul 27 08:15:50 2007 -0600
@@ -457,9 +457,6 @@ out:
*/
#define xlate_dev_mem_ptr(p) __va(p)

-#define xen_xlate_dev_mem_ptr(p,sz) xlate_dev_mem_ptr(p)
-#define xen_xlate_dev_mem_ptr_unmap(p)
-
/*
* Convert a virtual cached pointer to an uncached pointer
*/
diff -r e6974eba1bae -r b0bf9ba32bfe include/asm-x86_64/acpi.h
--- a/include/asm-x86_64/acpi.h Thu Jul 26 14:33:50 2007 -0600
+++ b/include/asm-x86_64/acpi.h Fri Jul 27 08:15:50 2007 -0600
@@ -153,6 +153,10 @@ extern unsigned long acpi_wakeup_address
/* early initialization routine */
extern void acpi_reserve_bootmem(void);

+#ifdef CONFIG_ACPI_PV_SLEEP
+extern int acpi_notify_hypervisor_state(u8 sleep_state,
+ u32 pm1a_cnt, u32 pm1b_cnt);
+#endif /* CONFIG_ACPI_PV_SLEEP */
#endif /*CONFIG_ACPI_SLEEP*/

#define boot_cpu_physical_apicid boot_cpu_id
diff -r e6974eba1bae -r b0bf9ba32bfe include/asm-x86_64/mach-xen/asm/io.h
--- a/include/asm-x86_64/mach-xen/asm/io.h Thu Jul 26 14:33:50 2007 -0600
+++ b/include/asm-x86_64/mach-xen/asm/io.h Fri Jul 27 08:15:50 2007 -0600
@@ -315,8 +315,7 @@ extern int iommu_bio_merge;
* Convert a physical pointer to a virtual kernel pointer for /dev/mem
* access
*/
-#define xen_xlate_dev_mem_ptr(p, sz) ioremap(p, sz)
-#define xen_xlate_dev_mem_ptr_unmap(p) iounmap(p)
+#define xlate_dev_mem_ptr(p) __va(p)

/*
* Convert a virtual cached pointer to an uncached pointer
diff -r e6974eba1bae -r b0bf9ba32bfe include/xen/interface/platform.h
--- a/include/xen/interface/platform.h Thu Jul 26 14:33:50 2007 -0600
+++ b/include/xen/interface/platform.h Fri Jul 27 08:15:50 2007 -0600
@@ -153,6 +153,17 @@ typedef struct xenpf_firmware_info xenpf
typedef struct xenpf_firmware_info xenpf_firmware_info_t;
DEFINE_XEN_GUEST_HANDLE(xenpf_firmware_info_t);

+#define XENPF_enter_acpi_sleep 51
+struct xenpf_enter_acpi_sleep {
+ /* IN variables */
+ uint16_t pm1a_cnt_val; /* PM1a control value. */
+ uint16_t pm1b_cnt_val; /* PM1b control value. */
+ uint32_t sleep_state; /* Which state to enter (Sn). */
+ uint32_t flags; /* Must be zero. */
+};
+typedef struct xenpf_enter_acpi_sleep xenpf_enter_acpi_sleep_t;
+DEFINE_XEN_GUEST_HANDLE(xenpf_enter_acpi_sleep_t);
+
struct xen_platform_op {
uint32_t cmd;
uint32_t interface_version; /* XENPF_INTERFACE_VERSION */
@@ -164,6 +175,7 @@ struct xen_platform_op {
struct xenpf_microcode_update microcode;
struct xenpf_platform_quirk platform_quirk;
struct xenpf_firmware_info firmware_info;
+ struct xenpf_enter_acpi_sleep enter_acpi_sleep;
uint8_t pad[128];
} u;
};
diff -r e6974eba1bae -r b0bf9ba32bfe lib/Makefile
--- a/lib/Makefile Thu Jul 26 14:33:50 2007 -0600
+++ b/lib/Makefile Fri Jul 27 08:15:50 2007 -0600
@@ -52,9 +52,6 @@ obj-$(CONFIG_AUDIT_GENERIC) += audit.o
obj-$(CONFIG_AUDIT_GENERIC) += audit.o

obj-$(CONFIG_SWIOTLB) += swiotlb.o
-ifeq ($(CONFIG_X86),y)
-swiotlb-$(CONFIG_XEN) := ../arch/i386/kernel/swiotlb.o
-endif
ifeq ($(CONFIG_IA64),y)
swiotlb-$(CONFIG_XEN) := ../arch/ia64/xen/swiotlb.o
endif
diff -r e6974eba1bae -r b0bf9ba32bfe lib/swiotlb-xen.c
--- /dev/null Thu Jan 01 00:00:00 1970 +0000
+++ b/lib/swiotlb-xen.c Fri Jul 27 08:15:50 2007 -0600
@@ -0,0 +1,745 @@
+/*
+ * Dynamic DMA mapping support.
+ *
+ * This implementation is a fallback for platforms that do not support
+ * I/O TLBs (aka DMA address translation hardware).
+ * Copyright (C) 2000 Asit Mallick <Asit.K.Mallick@intel.com>
+ * Copyright (C) 2000 Goutham Rao <goutham.rao@intel.com>
+ * Copyright (C) 2000, 2003 Hewlett-Packard Co
+ * David Mosberger-Tang <davidm@hpl.hp.com>
+ * Copyright (C) 2005 Keir Fraser <keir@xensource.com>
+ */
+
+#include <linux/cache.h>
+#include <linux/mm.h>
+#include <linux/module.h>
+#include <linux/pci.h>
+#include <linux/spinlock.h>
+#include <linux/string.h>
+#include <linux/types.h>
+#include <linux/ctype.h>
+#include <linux/init.h>
+#include <linux/bootmem.h>
+#include <linux/highmem.h>
+#include <asm/io.h>
+#include <asm/pci.h>
+#include <asm/dma.h>
+#include <asm/uaccess.h>
+#include <xen/gnttab.h>
+#include <xen/interface/memory.h>
+#include <asm-i386/mach-xen/asm/gnttab_dma.h>
+
+int swiotlb;
+EXPORT_SYMBOL(swiotlb);
+
+#define OFFSET(val,align) ((unsigned long)((val) & ( (align) - 1)))
+
+/*
+ * Maximum allowable number of contiguous slabs to map,
+ * must be a power of 2. What is the appropriate value ?
+ * The complexity of {map,unmap}_single is linearly dependent on this value.
+ */
+#define IO_TLB_SEGSIZE 128
+
+/*
+ * log of the size of each IO TLB slab. The number of slabs is command line
+ * controllable.
+ */
+#define IO_TLB_SHIFT 11
+
+int swiotlb_force;
+
+static char *iotlb_virt_start;
+static unsigned long iotlb_nslabs;
+
+/*
+ * Used to do a quick range check in swiotlb_unmap_single and
+ * swiotlb_sync_single_*, to see if the memory was in fact allocated by this
+ * API.
+ */
+static unsigned long iotlb_pfn_start, iotlb_pfn_end;
+
+/* Does the given dma address reside within the swiotlb aperture? */
+static inline int in_swiotlb_aperture(dma_addr_t dev_addr)
+{
+ unsigned long pfn = mfn_to_local_pfn(dev_addr >> PAGE_SHIFT);
+ return (pfn_valid(pfn)
+ && (pfn >= iotlb_pfn_start)
+ && (pfn < iotlb_pfn_end));
+}
+
+/*
+ * When the IOMMU overflows we return a fallback buffer. This sets the size.
+ */
+static unsigned long io_tlb_overflow = 32*1024;
+
+void *io_tlb_overflow_buffer;
+
+/*
+ * This is a free list describing the number of free entries available from
+ * each index
+ */
+static unsigned int *io_tlb_list;
+static unsigned int io_tlb_index;
+
+/*
+ * We need to save away the original address corresponding to a mapped entry
+ * for the sync operations.
+ */
+static struct phys_addr {
+ struct page *page;
+ unsigned int offset;
+} *io_tlb_orig_addr;
+
+/*
+ * Protect the above data structures in the map and unmap calls
+ */
+static DEFINE_SPINLOCK(io_tlb_lock);
+
+static unsigned int dma_bits;
+static unsigned int __initdata max_dma_bits = 32;
+static int __init
+setup_dma_bits(char *str)
+{
+ max_dma_bits = simple_strtoul(str, NULL, 0);
+ return 0;
+}
+__setup("dma_bits=", setup_dma_bits);
+
+static int __init
+setup_io_tlb_npages(char *str)
+{
+ /* Unlike ia64, the size is aperture in megabytes, not 'slabs'! */
+ if (isdigit(*str)) {
+ iotlb_nslabs = simple_strtoul(str, &str, 0) <<
+ (20 - IO_TLB_SHIFT);
+ iotlb_nslabs = ALIGN(iotlb_nslabs, IO_TLB_SEGSIZE);
+ /* Round up to power of two (xen_create_contiguous_region). */
+ while (iotlb_nslabs & (iotlb_nslabs-1))
+ iotlb_nslabs += iotlb_nslabs & ~(iotlb_nslabs-1);
+ }
+ if (*str == ',')
+ ++str;
+ /*
+ * NB. 'force' enables the swiotlb, but doesn't force its use for
+ * every DMA like it does on native Linux. 'off' forcibly disables
+ * use of the swiotlb.
+ */
+ if (!strcmp(str, "force"))
+ swiotlb_force = 1;
+ else if (!strcmp(str, "off"))
+ swiotlb_force = -1;
+ return 1;
+}
+__setup("swiotlb=", setup_io_tlb_npages);
+/* make io_tlb_overflow tunable too? */
+
+/*
+ * Statically reserve bounce buffer space and initialize bounce buffer data
+ * structures for the software IO TLB used to implement the PCI DMA API.
+ */
+void
+swiotlb_init_with_default_size (size_t default_size)
+{
+ unsigned long i, bytes;
+ int rc;
+
+ if (!iotlb_nslabs) {
+ iotlb_nslabs = (default_size >> IO_TLB_SHIFT);
+ iotlb_nslabs = ALIGN(iotlb_nslabs, IO_TLB_SEGSIZE);
+ /* Round up to power of two (xen_create_contiguous_region). */
+ while (iotlb_nslabs & (iotlb_nslabs-1))
+ iotlb_nslabs += iotlb_nslabs & ~(iotlb_nslabs-1);
+ }
+
+ bytes = iotlb_nslabs * (1UL << IO_TLB_SHIFT);
+
+ /*
+ * Get IO TLB memory from the low pages
+ */
+ iotlb_virt_start = alloc_bootmem_low_pages(bytes);
+ if (!iotlb_virt_start)
+ panic("Cannot allocate SWIOTLB buffer!\n");
+
+ dma_bits = get_order(IO_TLB_SEGSIZE << IO_TLB_SHIFT) + PAGE_SHIFT;
+ for (i = 0; i < iotlb_nslabs; i += IO_TLB_SEGSIZE) {
+ do {
+ rc = xen_create_contiguous_region(
+ (unsigned long)iotlb_virt_start + (i << IO_TLB_SHIFT),
+ get_order(IO_TLB_SEGSIZE << IO_TLB_SHIFT),
+ dma_bits);
+ } while (rc && dma_bits++ < max_dma_bits);
+ if (rc) {
+ if (i == 0)
+ panic("No suitable physical memory available for SWIOTLB buffer!\n"
+ "Use dom0_mem Xen boot parameter to reserve\n"
+ "some DMA memory (e.g., dom0_mem=-128M).\n");
+ iotlb_nslabs = i;
+ i <<= IO_TLB_SHIFT;
+ free_bootmem(__pa(iotlb_virt_start + i), bytes - i);
+ bytes = i;
+ for (dma_bits = 0; i > 0; i -= IO_TLB_SEGSIZE << IO_TLB_SHIFT) {
+ unsigned int bits = fls64(virt_to_bus(iotlb_virt_start + i - 1));
+
+ if (bits > dma_bits)
+ dma_bits = bits;
+ }
+ break;
+ }
+ }
+
+ /*
+ * Allocate and initialize the free list array. This array is used
+ * to find contiguous free memory regions of size up to IO_TLB_SEGSIZE.
+ */
+ io_tlb_list = alloc_bootmem(iotlb_nslabs * sizeof(int));
+ for (i = 0; i < iotlb_nslabs; i++)
+ io_tlb_list[i] = IO_TLB_SEGSIZE - OFFSET(i, IO_TLB_SEGSIZE);
+ io_tlb_index = 0;
+ io_tlb_orig_addr = alloc_bootmem(
+ iotlb_nslabs * sizeof(*io_tlb_orig_addr));
+
+ /*
+ * Get the overflow emergency buffer
+ */
+ io_tlb_overflow_buffer = alloc_bootmem_low(io_tlb_overflow);
+ if (!io_tlb_overflow_buffer)
+ panic("Cannot allocate SWIOTLB overflow buffer!\n");
+
+ do {
+ rc = xen_create_contiguous_region(
+ (unsigned long)io_tlb_overflow_buffer,
+ get_order(io_tlb_overflow),
+ dma_bits);
+ } while (rc && dma_bits++ < max_dma_bits);
+ if (rc)
+ panic("No suitable physical memory available for SWIOTLB overflow buffer!\n");
+
+ iotlb_pfn_start = __pa(iotlb_virt_start) >> PAGE_SHIFT;
+ iotlb_pfn_end = iotlb_pfn_start + (bytes >> PAGE_SHIFT);
+
+ printk(KERN_INFO "Software IO TLB enabled: \n"
+ " Aperture: %lu megabytes\n"
+ " Kernel range: %p - %p\n"
+ " Address size: %u bits\n",
+ bytes >> 20,
+ iotlb_virt_start, iotlb_virt_start + bytes,
+ dma_bits);
+}
+
+void
+swiotlb_init(void)
+{
+ long ram_end;
+ size_t defsz = 64 * (1 << 20); /* 64MB default size */
+
+ if (swiotlb_force == 1) {
+ swiotlb = 1;
+ } else if ((swiotlb_force != -1) &&
+ is_running_on_xen() &&
+ is_initial_xendomain()) {
+ /* Domain 0 always has a swiotlb. */
+ ram_end = HYPERVISOR_memory_op(XENMEM_maximum_ram_page, NULL);
+ if (ram_end <= 0x7ffff)
+ defsz = 2 * (1 << 20); /* 2MB on <2GB on systems. */
+ swiotlb = 1;
+ }
+
+ if (swiotlb)
+ swiotlb_init_with_default_size(defsz);
+ else
+ printk(KERN_INFO "Software IO TLB disabled\n");
+}
+
+/*
+ * We use __copy_to_user_inatomic to transfer to the host buffer because the
+ * buffer may be mapped read-only (e.g, in blkback driver) but lower-level
+ * drivers map the buffer for DMA_BIDIRECTIONAL access. This causes an
+ * unnecessary copy from the aperture to the host buffer, and a page fault.
+ */
+static void
+__sync_single(struct phys_addr buffer, char *dma_addr, size_t size, int dir)
+{
+ if (PageHighMem(buffer.page)) {
+ size_t len, bytes;
+ char *dev, *host, *kmp;
+ len = size;
+ while (len != 0) {
+ unsigned long flags;
+
+ if (((bytes = len) + buffer.offset) > PAGE_SIZE)
+ bytes = PAGE_SIZE - buffer.offset;
+ local_irq_save(flags); /* protects KM_BOUNCE_READ */
+ kmp = kmap_atomic(buffer.page, KM_BOUNCE_READ);
+ dev = dma_addr + size - len;
+ host = kmp + buffer.offset;
+ if (dir == DMA_FROM_DEVICE) {
+ if (__copy_to_user_inatomic(host, dev, bytes))
+ /* inaccessible */;
+ } else
+ memcpy(dev, host, bytes);
+ kunmap_atomic(kmp, KM_BOUNCE_READ);
+ local_irq_restore(flags);
+ len -= bytes;
+ buffer.page++;
+ buffer.offset = 0;
+ }
+ } else {
+ char *host = (char *)phys_to_virt(
+ page_to_pseudophys(buffer.page)) + buffer.offset;
+ if (dir == DMA_FROM_DEVICE) {
+ if (__copy_to_user_inatomic(host, dma_addr, size))
+ /* inaccessible */;
+ } else if (dir == DMA_TO_DEVICE)
+ memcpy(dma_addr, host, size);
+ }
+}
+
+/*
+ * Allocates bounce buffer and returns its kernel virtual address.
+ */
+static void *
+map_single(struct device *hwdev, struct phys_addr buffer, size_t size, int dir)
+{
+ unsigned long flags;
+ char *dma_addr;
+ unsigned int nslots, stride, index, wrap;
+ struct phys_addr slot_buf;
+ int i;
+
+ /*
+ * For mappings greater than a page, we limit the stride (and
+ * hence alignment) to a page size.
+ */
+ nslots = ALIGN(size, 1 << IO_TLB_SHIFT) >> IO_TLB_SHIFT;
+ if (size > PAGE_SIZE)
+ stride = (1 << (PAGE_SHIFT - IO_TLB_SHIFT));
+ else
+ stride = 1;
+
+ BUG_ON(!nslots);
+
+ /*
+ * Find suitable number of IO TLB entries size that will fit this
+ * request and allocate a buffer from that IO TLB pool.
+ */
+ spin_lock_irqsave(&io_tlb_lock, flags);
+ {
+ wrap = index = ALIGN(io_tlb_index, stride);
+
+ if (index >= iotlb_nslabs)
+ wrap = index = 0;
+
+ do {
+ /*
+ * If we find a slot that indicates we have 'nslots'
+ * number of contiguous buffers, we allocate the
+ * buffers from that slot and mark the entries as '0'
+ * indicating unavailable.
+ */
+ if (io_tlb_list[index] >= nslots) {
+ int count = 0;
+
+ for (i = index; i < (int)(index + nslots); i++)
+ io_tlb_list[i] = 0;
+ for (i = index - 1;
+ (OFFSET(i, IO_TLB_SEGSIZE) !=
+ IO_TLB_SEGSIZE -1) && io_tlb_list[i];
+ i--)
+ io_tlb_list[i] = ++count;
+ dma_addr = iotlb_virt_start +
+ (index << IO_TLB_SHIFT);
+
+ /*
+ * Update the indices to avoid searching in
+ * the next round.
+ */
+ io_tlb_index =
+ ((index + nslots) < iotlb_nslabs
+ ? (index + nslots) : 0);
+
+ goto found;
+ }
+ index += stride;
+ if (index >= iotlb_nslabs)
+ index = 0;
+ } while (index != wrap);
+
+ spin_unlock_irqrestore(&io_tlb_lock, flags);
+ return NULL;
+ }
+ found:
+ spin_unlock_irqrestore(&io_tlb_lock, flags);
+
+ /*
+ * Save away the mapping from the original address to the DMA address.
+ * This is needed when we sync the memory. Then we sync the buffer if
+ * needed.
+ */
+ slot_buf = buffer;
+ for (i = 0; i < nslots; i++) {
+ slot_buf.page += slot_buf.offset >> PAGE_SHIFT;
+ slot_buf.offset &= PAGE_SIZE - 1;
+ io_tlb_orig_addr[index+i] = slot_buf;
+ slot_buf.offset += 1 << IO_TLB_SHIFT;
+ }
+ if ((dir == DMA_TO_DEVICE) || (dir == DMA_BIDIRECTIONAL))
+ __sync_single(buffer, dma_addr, size, DMA_TO_DEVICE);
+
+ return dma_addr;
+}
+
+static struct phys_addr dma_addr_to_phys_addr(char *dma_addr)
+{
+ int index = (dma_addr - iotlb_virt_start) >> IO_TLB_SHIFT;
+ struct phys_addr buffer = io_tlb_orig_addr[index];
+ buffer.offset += (long)dma_addr & ((1 << IO_TLB_SHIFT) - 1);
+ buffer.page += buffer.offset >> PAGE_SHIFT;
+ buffer.offset &= PAGE_SIZE - 1;
+ return buffer;
+}
+
+/*
+ * dma_addr is the kernel virtual address of the bounce buffer to unmap.
+ */
+static void
+unmap_single(struct device *hwdev, char *dma_addr, size_t size, int dir)
+{
+ unsigned long flags;
+ int i, count, nslots = ALIGN(size, 1 << IO_TLB_SHIFT) >> IO_TLB_SHIFT;
+ int index = (dma_addr - iotlb_virt_start) >> IO_TLB_SHIFT;
+ struct phys_addr buffer = dma_addr_to_phys_addr(dma_addr);
+
+ /*
+ * First, sync the memory before unmapping the entry
+ */
+ if ((dir == DMA_FROM_DEVICE) || (dir == DMA_BIDIRECTIONAL))
+ __sync_single(buffer, dma_addr, size, DMA_FROM_DEVICE);
+
+ /*
+ * Return the buffer to the free list by setting the corresponding
+ * entries to indicate the number of contigous entries available.
+ * While returning the entries to the free list, we merge the entries
+ * with slots below and above the pool being returned.
+ */
+ spin_lock_irqsave(&io_tlb_lock, flags);
+ {
+ count = ((index + nslots) < ALIGN(index + 1, IO_TLB_SEGSIZE) ?
+ io_tlb_list[index + nslots] : 0);
+ /*
+ * Step 1: return the slots to the free list, merging the
+ * slots with superceeding slots
+ */
+ for (i = index + nslots - 1; i >= index; i--)
+ io_tlb_list[i] = ++count;
+ /*
+ * Step 2: merge the returned slots with the preceding slots,
+ * if available (non zero)
+ */
+ for (i = index - 1;
+ (OFFSET(i, IO_TLB_SEGSIZE) !=
+ IO_TLB_SEGSIZE -1) && io_tlb_list[i];
+ i--)
+ io_tlb_list[i] = ++count;
+ }
+ spin_unlock_irqrestore(&io_tlb_lock, flags);
+}
+
+static void
+sync_single(struct device *hwdev, char *dma_addr, size_t size, int dir)
+{
+ struct phys_addr buffer = dma_addr_to_phys_addr(dma_addr);
+ BUG_ON((dir != DMA_FROM_DEVICE) && (dir != DMA_TO_DEVICE));
+ __sync_single(buffer, dma_addr, size, dir);
+}
+
+static void
+swiotlb_full(struct device *dev, size_t size, int dir, int do_panic)
+{
+ /*
+ * Ran out of IOMMU space for this operation. This is very bad.
+ * Unfortunately the drivers cannot handle this operation properly.
+ * unless they check for pci_dma_mapping_error (most don't)
+ * When the mapping is small enough return a static buffer to limit
+ * the damage, or panic when the transfer is too big.
+ */
+ printk(KERN_ERR "PCI-DMA: Out of SW-IOMMU space for %lu bytes at "
+ "device %s\n", (unsigned long)size, dev ? dev->bus_id : "?");
+
+ if (size > io_tlb_overflow && do_panic) {
+ if (dir == PCI_DMA_FROMDEVICE || dir == PCI_DMA_BIDIRECTIONAL)
+ panic("PCI-DMA: Memory would be corrupted\n");
+ if (dir == PCI_DMA_TODEVICE || dir == PCI_DMA_BIDIRECTIONAL)
+ panic("PCI-DMA: Random memory would be DMAed\n");
+ }
+}
+
+/*
+ * Map a single buffer of the indicated size for DMA in streaming mode. The
+ * PCI address to use is returned.
+ *
+ * Once the device is given the dma address, the device owns this memory until
+ * either swiotlb_unmap_single or swiotlb_dma_sync_single is performed.
+ */
+dma_addr_t
+swiotlb_map_single(struct device *hwdev, void *ptr, size_t size, int dir)
+{
+ dma_addr_t dev_addr = gnttab_dma_map_page(virt_to_page(ptr)) +
+ offset_in_page(ptr);
+ void *map;
+ struct phys_addr buffer;
+
+ BUG_ON(dir == DMA_NONE);
+
+ /*
+ * If the pointer passed in happens to be in the device's DMA window,
+ * we can safely return the device addr and not worry about bounce
+ * buffering it.
+ */
+ if (!range_straddles_page_boundary(__pa(ptr), size) &&
+ !address_needs_mapping(hwdev, dev_addr))
+ return dev_addr;
+
+ /*
+ * Oh well, have to allocate and map a bounce buffer.
+ */
+ gnttab_dma_unmap_page(dev_addr);
+ buffer.page = virt_to_page(ptr);
+ buffer.offset = (unsigned long)ptr & ~PAGE_MASK;
+ map = map_single(hwdev, buffer, size, dir);
+ if (!map) {
+ swiotlb_full(hwdev, size, dir, 1);
+ map = io_tlb_overflow_buffer;
+ }
+
+ dev_addr = virt_to_bus(map);
+ return dev_addr;
+}
+
+/*
+ * Unmap a single streaming mode DMA translation. The dma_addr and size must
+ * match what was provided for in a previous swiotlb_map_single call. All
+ * other usages are undefined.
+ *
+ * After this call, reads by the cpu to the buffer are guaranteed to see
+ * whatever the device wrote there.
+ */
+void
+swiotlb_unmap_single(struct device *hwdev, dma_addr_t dev_addr, size_t size,
+ int dir)
+{
+ BUG_ON(dir == DMA_NONE);
+ if (in_swiotlb_aperture(dev_addr))
+ unmap_single(hwdev, bus_to_virt(dev_addr), size, dir);
+ else
+ gnttab_dma_unmap_page(dev_addr);
+}
+
+/*
+ * Make physical memory consistent for a single streaming mode DMA translation
+ * after a transfer.
+ *
+ * If you perform a swiotlb_map_single() but wish to interrogate the buffer
+ * using the cpu, yet do not wish to teardown the PCI dma mapping, you must
+ * call this function before doing so. At the next point you give the PCI dma
+ * address back to the card, you must first perform a
+ * swiotlb_dma_sync_for_device, and then the device again owns the buffer
+ */
+void
+swiotlb_sync_single_for_cpu(struct device *hwdev, dma_addr_t dev_addr,
+ size_t size, int dir)
+{
+ BUG_ON(dir == DMA_NONE);
+ if (in_swiotlb_aperture(dev_addr))
+ sync_single(hwdev, bus_to_virt(dev_addr), size, dir);
+}
+
+void
+swiotlb_sync_single_for_device(struct device *hwdev, dma_addr_t dev_addr,
+ size_t size, int dir)
+{
+ BUG_ON(dir == DMA_NONE);
+ if (in_swiotlb_aperture(dev_addr))
+ sync_single(hwdev, bus_to_virt(dev_addr), size, dir);
+}
+
+/*
+ * Map a set of buffers described by scatterlist in streaming mode for DMA.
+ * This is the scatter-gather version of the above swiotlb_map_single
+ * interface. Here the scatter gather list elements are each tagged with the
+ * appropriate dma address and length. They are obtained via
+ * sg_dma_{address,length}(SG).
+ *
+ * NOTE: An implementation may be able to use a smaller number of
+ * DMA address/length pairs than there are SG table elements.
+ * (for example via virtual mapping capabilities)
+ * The routine returns the number of addr/length pairs actually
+ * used, at most nents.
+ *
+ * Device ownership issues as mentioned above for swiotlb_map_single are the
+ * same here.
+ */
+int
+swiotlb_map_sg(struct device *hwdev, struct scatterlist *sg, int nelems,
+ int dir)
+{
+ struct phys_addr buffer;
+ dma_addr_t dev_addr;
+ char *map;
+ int i;
+
+ BUG_ON(dir == DMA_NONE);
+
+ for (i = 0; i < nelems; i++, sg++) {
+ dev_addr = gnttab_dma_map_page(sg->page) + sg->offset;
+
+ if (range_straddles_page_boundary(page_to_pseudophys(sg->page)
+ + sg->offset, sg->length)
+ || address_needs_mapping(hwdev, dev_addr)) {
+ gnttab_dma_unmap_page(dev_addr);
+ buffer.page = sg->page;
+ buffer.offset = sg->offset;
+ map = map_single(hwdev, buffer, sg->length, dir);
+ if (!map) {
+ /* Don't panic here, we expect map_sg users
+ to do proper error handling. */
+ swiotlb_full(hwdev, sg->length, dir, 0);
+ swiotlb_unmap_sg(hwdev, sg - i, i, dir);
+ sg[0].dma_length = 0;
+ return 0;
+ }
+ sg->dma_address = (dma_addr_t)virt_to_bus(map);
+ } else
+ sg->dma_address = dev_addr;
+ sg->dma_length = sg->length;
+ }
+ return nelems;
+}
+
+/*
+ * Unmap a set of streaming mode DMA translations. Again, cpu read rules
+ * concerning calls here are the same as for swiotlb_unmap_single() above.
+ */
+void
+swiotlb_unmap_sg(struct device *hwdev, struct scatterlist *sg, int nelems,
+ int dir)
+{
+ int i;
+
+ BUG_ON(dir == DMA_NONE);
+
+ for (i = 0; i < nelems; i++, sg++)
+ if (in_swiotlb_aperture(sg->dma_address))
+ unmap_single(hwdev,
+ (void *)bus_to_virt(sg->dma_address),
+ sg->dma_length, dir);
+ else
+ gnttab_dma_unmap_page(sg->dma_address);
+}
+
+/*
+ * Make physical memory consistent for a set of streaming mode DMA translations
+ * after a transfer.
+ *
+ * The same as swiotlb_sync_single_* but for a scatter-gather list, same rules
+ * and usage.
+ */
+void
+swiotlb_sync_sg_for_cpu(struct device *hwdev, struct scatterlist *sg,
+ int nelems, int dir)
+{
+ int i;
+
+ BUG_ON(dir == DMA_NONE);
+
+ for (i = 0; i < nelems; i++, sg++)
+ if (in_swiotlb_aperture(sg->dma_address))
+ sync_single(hwdev,
+ (void *)bus_to_virt(sg->dma_address),
+ sg->dma_length, dir);
+}
+
+void
+swiotlb_sync_sg_for_device(struct device *hwdev, struct scatterlist *sg,
+ int nelems, int dir)
+{
+ int i;
+
+ BUG_ON(dir == DMA_NONE);
+
+ for (i = 0; i < nelems; i++, sg++)
+ if (in_swiotlb_aperture(sg->dma_address))
+ sync_single(hwdev,
+ (void *)bus_to_virt(sg->dma_address),
+ sg->dma_length, dir);
+}
+
+#ifdef CONFIG_HIGHMEM
+
+dma_addr_t
+swiotlb_map_page(struct device *hwdev, struct page *page,
+ unsigned long offset, size_t size,
+ enum dma_data_direction direction)
+{
+ struct phys_addr buffer;
+ dma_addr_t dev_addr;
+ char *map;
+
+ dev_addr = gnttab_dma_map_page(page) + offset;
+ if (address_needs_mapping(hwdev, dev_addr)) {
+ gnttab_dma_unmap_page(dev_addr);
+ buffer.page = page;
+ buffer.offset = offset;
+ map = map_single(hwdev, buffer, size, direction);
+ if (!map) {
+ swiotlb_full(hwdev, size, direction, 1);
+ map = io_tlb_overflow_buffer;
+ }
+ dev_addr = (dma_addr_t)virt_to_bus(map);
+ }
+
+ return dev_addr;
+}
+
+void
+swiotlb_unmap_page(struct device *hwdev, dma_addr_t dma_address,
+ size_t size, enum dma_data_direction direction)
+{
+ BUG_ON(direction == DMA_NONE);
+ if (in_swiotlb_aperture(dma_address))
+ unmap_single(hwdev, bus_to_virt(dma_address), size, direction);
+ else
+ gnttab_dma_unmap_page(dma_address);
+}
+
+#endif
+
+int
+swiotlb_dma_mapping_error(dma_addr_t dma_addr)
+{
+ return (dma_addr == virt_to_bus(io_tlb_overflow_buffer));
+}
+
+/*
+ * Return whether the given PCI device DMA address mask can be supported
+ * properly. For example, if your device can only drive the low 24-bits
+ * during PCI bus mastering, then you would pass 0x00ffffff as the mask to
+ * this function.
+ */
+int
+swiotlb_dma_supported (struct device *hwdev, u64 mask)
+{
+ return (mask >= ((1UL << dma_bits) - 1));
+}
+
+EXPORT_SYMBOL(swiotlb_init);
+EXPORT_SYMBOL(swiotlb_map_single);
+EXPORT_SYMBOL(swiotlb_unmap_single);
+EXPORT_SYMBOL(swiotlb_map_sg);
+EXPORT_SYMBOL(swiotlb_unmap_sg);
+EXPORT_SYMBOL(swiotlb_sync_single_for_cpu);
+EXPORT_SYMBOL(swiotlb_sync_single_for_device);
+EXPORT_SYMBOL(swiotlb_sync_sg_for_cpu);
+EXPORT_SYMBOL(swiotlb_sync_sg_for_device);
+EXPORT_SYMBOL(swiotlb_dma_mapping_error);
+EXPORT_SYMBOL(swiotlb_dma_supported);
diff -r e6974eba1bae -r b0bf9ba32bfe scripts/Makefile.build
--- a/scripts/Makefile.build Thu Jul 26 14:33:50 2007 -0600
+++ b/scripts/Makefile.build Fri Jul 27 08:15:50 2007 -0600
@@ -66,6 +66,18 @@ endif

ifndef obj
$(warning kbuild: Makefile.build is included improperly)
+endif
+
+ifeq ($(CONFIG_XEN),y)
+$(objtree)/scripts/Makefile.xen: $(srctree)/scripts/Makefile.xen.awk $(srctree)/scripts/Makefile.build
+ @echo ' Updating $@'
+ @$(AWK) -f $< $(filter-out $<,$^) >$@
+
+xen-src-single-used-m := $(patsubst $(srctree)/%,%,$(wildcard $(addprefix $(srctree)/,$(single-used-m:.o=-xen.c))))
+xen-single-used-m := $(xen-src-single-used-m:-xen.c=.o)
+single-used-m := $(filter-out $(xen-single-used-m),$(single-used-m))
+
+-include $(objtree)/scripts/Makefile.xen
endif

# ===========================================================================
diff -r e6974eba1bae -r b0bf9ba32bfe scripts/Makefile.lib
--- a/scripts/Makefile.lib Thu Jul 26 14:33:50 2007 -0600
+++ b/scripts/Makefile.lib Fri Jul 27 08:15:50 2007 -0600
@@ -12,6 +12,12 @@ obj-m := $(filter-out $(obj-y),$(obj-m))
# Filter out objects already built-in

lib-y := $(filter-out $(obj-y), $(sort $(lib-y) $(lib-m)))
+
+# Remove objects forcibly disabled
+
+obj-y := $(filter-out $(disabled-obj-y),$(obj-y))
+obj-m := $(filter-out $(disabled-obj-y),$(obj-m))
+lib-y := $(filter-out $(disabled-obj-y),$(lib-y))


# Handle objects in subdirs
diff -r e6974eba1bae -r b0bf9ba32bfe scripts/Makefile.xen.awk
--- /dev/null Thu Jan 01 00:00:00 1970 +0000
+++ b/scripts/Makefile.xen.awk Fri Jul 27 08:15:50 2007 -0600
@@ -0,0 +1,34 @@
+BEGIN {
+ is_rule = 0
+}
+
+/^[[:space:]]*#/ {
+ next
+}
+
+/^[[:space:]]*$/ {
+ if (is_rule)
+ print("")
+ is_rule = 0
+ next
+}
+
+/:[[:space:]]*%\.[cS][[:space:]]/ {
+ line = gensub(/%.([cS])/, "%-xen.\\1", "g", $0)
+ line = gensub(/(single-used-m)/, "xen-\\1", "g", line)
+ print line
+ is_rule = 1
+ next
+}
+
+/^[^\t]$/ {
+ if (is_rule)
+ print("")
+ is_rule = 0
+ next
+}
+
+is_rule {
+ print $0
+ next
+}

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