aboutsummaryrefslogtreecommitdiffstats
path: root/arch/x86
diff options
context:
space:
mode:
authorIngo Molnar <mingo@elte.hu>2008-11-10 03:16:27 -0500
committerIngo Molnar <mingo@elte.hu>2008-11-10 03:16:27 -0500
commit4ecd33d930591d41fe356160593a9076467b961c (patch)
treeb9051a334540bbce38db1b2b03cebb4cf1d51f73 /arch/x86
parent7d5a78cd98c3a5eb83bd6a061c5ea6ef1e9b8fcb (diff)
parentf7160c7573615ec82c691e294cf80d920b5d588d (diff)
Merge commit 'v2.6.28-rc4' into x86/apic
Diffstat (limited to 'arch/x86')
-rw-r--r--arch/x86/Kconfig12
-rw-r--r--arch/x86/Kconfig.cpu24
-rw-r--r--arch/x86/boot/compressed/.gitignore2
-rw-r--r--arch/x86/include/asm/cpufeature.h2
-rw-r--r--arch/x86/include/asm/dma-mapping.h4
-rw-r--r--arch/x86/include/asm/es7000/wakecpu.h9
-rw-r--r--arch/x86/include/asm/fixmap.h4
-rw-r--r--arch/x86/include/asm/fixmap_32.h4
-rw-r--r--arch/x86/include/asm/ftrace.h4
-rw-r--r--arch/x86/include/asm/highmem.h5
-rw-r--r--arch/x86/include/asm/io.h6
-rw-r--r--arch/x86/include/asm/iommu.h1
-rw-r--r--arch/x86/include/asm/irq_vectors.h20
-rw-r--r--arch/x86/include/asm/kvm_host.h3
-rw-r--r--arch/x86/include/asm/mach-default/mach_wakecpu.h9
-rw-r--r--arch/x86/include/asm/msr.h2
-rw-r--r--arch/x86/include/asm/pgtable-3level.h4
-rw-r--r--arch/x86/include/asm/smp.h6
-rw-r--r--arch/x86/include/asm/topology.h7
-rw-r--r--arch/x86/include/asm/tsc.h8
-rw-r--r--arch/x86/include/asm/uv/uv_hub.h1
-rw-r--r--arch/x86/include/asm/voyager.h1
-rw-r--r--arch/x86/kernel/Makefile3
-rw-r--r--arch/x86/kernel/amd_iommu.c9
-rw-r--r--arch/x86/kernel/cpu/addon_cpuid_features.c2
-rw-r--r--arch/x86/kernel/cpu/common.c7
-rw-r--r--arch/x86/kernel/e820.c8
-rw-r--r--arch/x86/kernel/entry_32.S4
-rw-r--r--arch/x86/kernel/entry_64.S4
-rw-r--r--arch/x86/kernel/ftrace.c50
-rw-r--r--arch/x86/kernel/genx2apic_uv_x.c7
-rw-r--r--arch/x86/kernel/i386_ksyms_32.c2
-rw-r--r--arch/x86/kernel/io_apic.c2
-rw-r--r--arch/x86/kernel/k8.c1
-rw-r--r--arch/x86/kernel/machine_kexec_32.c5
-rw-r--r--arch/x86/kernel/microcode_amd.c2
-rw-r--r--arch/x86/kernel/microcode_core.c4
-rw-r--r--arch/x86/kernel/pci-dma.c16
-rw-r--r--arch/x86/kernel/pci-gart_64.c2
-rw-r--r--arch/x86/kernel/pci-swiotlb_64.c14
-rw-r--r--arch/x86/kernel/reboot.c6
-rw-r--r--arch/x86/kernel/tlb_32.c6
-rw-r--r--arch/x86/kernel/tlb_64.c5
-rw-r--r--arch/x86/kernel/tsc.c12
-rw-r--r--arch/x86/kernel/vsmp_64.c2
-rw-r--r--arch/x86/kernel/x8664_ksyms_64.c2
-rw-r--r--arch/x86/kvm/i8254.c11
-rw-r--r--arch/x86/kvm/i8254.h1
-rw-r--r--arch/x86/kvm/mmu.c1
-rw-r--r--arch/x86/kvm/x86.c6
-rw-r--r--arch/x86/lguest/boot.c32
-rw-r--r--arch/x86/mach-voyager/setup.c2
-rw-r--r--arch/x86/mach-voyager/voyager_smp.c14
-rw-r--r--arch/x86/mm/Makefile2
-rw-r--r--arch/x86/mm/gup.c2
-rw-r--r--arch/x86/mm/init_32.c3
-rw-r--r--arch/x86/mm/init_64.c75
-rw-r--r--arch/x86/mm/iomap_32.c59
-rw-r--r--arch/x86/mm/ioremap.c22
-rw-r--r--arch/x86/mm/pageattr.c8
-rw-r--r--arch/x86/mm/pat.c4
-rw-r--r--arch/x86/oprofile/op_model_ppro.c9
-rw-r--r--arch/x86/xen/Makefile2
-rw-r--r--arch/x86/xen/enlighten.c5
-rw-r--r--arch/x86/xen/mmu.c31
65 files changed, 395 insertions, 207 deletions
diff --git a/arch/x86/Kconfig b/arch/x86/Kconfig
index 350bee1d54d..4cf0ab13d18 100644
--- a/arch/x86/Kconfig
+++ b/arch/x86/Kconfig
@@ -28,7 +28,7 @@ config X86
28 select HAVE_KRETPROBES 28 select HAVE_KRETPROBES
29 select HAVE_FTRACE_MCOUNT_RECORD 29 select HAVE_FTRACE_MCOUNT_RECORD
30 select HAVE_DYNAMIC_FTRACE 30 select HAVE_DYNAMIC_FTRACE
31 select HAVE_FTRACE 31 select HAVE_FUNCTION_TRACER
32 select HAVE_KVM if ((X86_32 && !X86_VOYAGER && !X86_VISWS && !X86_NUMAQ) || X86_64) 32 select HAVE_KVM if ((X86_32 && !X86_VOYAGER && !X86_VISWS && !X86_NUMAQ) || X86_64)
33 select HAVE_ARCH_KGDB if !X86_VOYAGER 33 select HAVE_ARCH_KGDB if !X86_VOYAGER
34 select HAVE_ARCH_TRACEHOOK 34 select HAVE_ARCH_TRACEHOOK
@@ -231,6 +231,10 @@ config SMP
231 231
232 If you don't know what to do here, say N. 232 If you don't know what to do here, say N.
233 233
234config X86_HAS_BOOT_CPU_ID
235 def_bool y
236 depends on X86_VOYAGER
237
234config X86_FIND_SMP_CONFIG 238config X86_FIND_SMP_CONFIG
235 def_bool y 239 def_bool y
236 depends on X86_MPPARSE || X86_VOYAGER 240 depends on X86_MPPARSE || X86_VOYAGER
@@ -1490,7 +1494,7 @@ config HAVE_ARCH_EARLY_PFN_TO_NID
1490 def_bool X86_64 1494 def_bool X86_64
1491 depends on NUMA 1495 depends on NUMA
1492 1496
1493menu "Power management options" 1497menu "Power management and ACPI options"
1494 depends on !X86_VOYAGER 1498 depends on !X86_VOYAGER
1495 1499
1496config ARCH_HIBERNATION_HEADER 1500config ARCH_HIBERNATION_HEADER
@@ -1890,6 +1894,10 @@ config SYSVIPC_COMPAT
1890endmenu 1894endmenu
1891 1895
1892 1896
1897config HAVE_ATOMIC_IOMAP
1898 def_bool y
1899 depends on X86_32
1900
1893source "net/Kconfig" 1901source "net/Kconfig"
1894 1902
1895source "drivers/Kconfig" 1903source "drivers/Kconfig"
diff --git a/arch/x86/Kconfig.cpu b/arch/x86/Kconfig.cpu
index 0b7c4a3f065..b815664fe37 100644
--- a/arch/x86/Kconfig.cpu
+++ b/arch/x86/Kconfig.cpu
@@ -513,19 +513,19 @@ config CPU_SUP_UMC_32
513 If unsure, say N. 513 If unsure, say N.
514 514
515config X86_DS 515config X86_DS
516 bool "Debug Store support" 516 def_bool X86_PTRACE_BTS
517 default y 517 depends on X86_DEBUGCTLMSR
518 help
519 Add support for Debug Store.
520 This allows the kernel to provide a memory buffer to the hardware
521 to store various profiling and tracing events.
522 518
523config X86_PTRACE_BTS 519config X86_PTRACE_BTS
524 bool "ptrace interface to Branch Trace Store" 520 bool "Branch Trace Store"
525 default y 521 default y
526 depends on (X86_DS && X86_DEBUGCTLMSR) 522 depends on X86_DEBUGCTLMSR
527 help 523 help
528 Add a ptrace interface to allow collecting an execution trace 524 This adds a ptrace interface to the hardware's branch trace store.
529 of the traced task. 525
530 This collects control flow changes in a (cyclic) buffer and allows 526 Debuggers may use it to collect an execution trace of the debugged
531 debuggers to fill in the gaps and show an execution trace of the debuggee. 527 application in order to answer the question 'how did I get here?'.
528 Debuggers may trace user mode as well as kernel mode.
529
530 Say Y unless there is no application development on this machine
531 and you want to save a small amount of code size.
diff --git a/arch/x86/boot/compressed/.gitignore b/arch/x86/boot/compressed/.gitignore
index be0ed065249..63eff3b04d0 100644
--- a/arch/x86/boot/compressed/.gitignore
+++ b/arch/x86/boot/compressed/.gitignore
@@ -1 +1,3 @@
1relocs 1relocs
2vmlinux.bin.all
3vmlinux.relocs
diff --git a/arch/x86/include/asm/cpufeature.h b/arch/x86/include/asm/cpufeature.h
index f73e95d75b4..cfdf8c2c5c3 100644
--- a/arch/x86/include/asm/cpufeature.h
+++ b/arch/x86/include/asm/cpufeature.h
@@ -91,7 +91,7 @@
91#define X86_FEATURE_11AP (3*32+19) /* "" Bad local APIC aka 11AP */ 91#define X86_FEATURE_11AP (3*32+19) /* "" Bad local APIC aka 11AP */
92#define X86_FEATURE_NOPL (3*32+20) /* The NOPL (0F 1F) instructions */ 92#define X86_FEATURE_NOPL (3*32+20) /* The NOPL (0F 1F) instructions */
93#define X86_FEATURE_AMDC1E (3*32+21) /* AMD C1E detected */ 93#define X86_FEATURE_AMDC1E (3*32+21) /* AMD C1E detected */
94#define X86_FEATURE_XTOPOLOGY (3*32+21) /* cpu topology enum extensions */ 94#define X86_FEATURE_XTOPOLOGY (3*32+22) /* cpu topology enum extensions */
95 95
96/* Intel-defined CPU features, CPUID level 0x00000001 (ecx), word 4 */ 96/* Intel-defined CPU features, CPUID level 0x00000001 (ecx), word 4 */
97#define X86_FEATURE_XMM3 (4*32+ 0) /* "pni" SSE-3 */ 97#define X86_FEATURE_XMM3 (4*32+ 0) /* "pni" SSE-3 */
diff --git a/arch/x86/include/asm/dma-mapping.h b/arch/x86/include/asm/dma-mapping.h
index 4a5397bfce2..7f225a4b2a2 100644
--- a/arch/x86/include/asm/dma-mapping.h
+++ b/arch/x86/include/asm/dma-mapping.h
@@ -255,9 +255,11 @@ static inline unsigned long dma_alloc_coherent_mask(struct device *dev,
255 255
256static inline gfp_t dma_alloc_coherent_gfp_flags(struct device *dev, gfp_t gfp) 256static inline gfp_t dma_alloc_coherent_gfp_flags(struct device *dev, gfp_t gfp)
257{ 257{
258#ifdef CONFIG_X86_64
259 unsigned long dma_mask = dma_alloc_coherent_mask(dev, gfp); 258 unsigned long dma_mask = dma_alloc_coherent_mask(dev, gfp);
260 259
260 if (dma_mask <= DMA_24BIT_MASK)
261 gfp |= GFP_DMA;
262#ifdef CONFIG_X86_64
261 if (dma_mask <= DMA_32BIT_MASK && !(gfp & GFP_DMA)) 263 if (dma_mask <= DMA_32BIT_MASK && !(gfp & GFP_DMA))
262 gfp |= GFP_DMA32; 264 gfp |= GFP_DMA32;
263#endif 265#endif
diff --git a/arch/x86/include/asm/es7000/wakecpu.h b/arch/x86/include/asm/es7000/wakecpu.h
index 3ffc5a7bf66..39849346191 100644
--- a/arch/x86/include/asm/es7000/wakecpu.h
+++ b/arch/x86/include/asm/es7000/wakecpu.h
@@ -50,10 +50,9 @@ static inline void restore_NMI_vector(unsigned short *high, unsigned short *low)
50{ 50{
51} 51}
52 52
53#if APIC_DEBUG 53#define inquire_remote_apic(apicid) do { \
54 #define inquire_remote_apic(apicid) __inquire_remote_apic(apicid) 54 if (apic_verbosity >= APIC_DEBUG) \
55#else 55 __inquire_remote_apic(apicid); \
56 #define inquire_remote_apic(apicid) {} 56 } while (0)
57#endif
58 57
59#endif /* __ASM_MACH_WAKECPU_H */ 58#endif /* __ASM_MACH_WAKECPU_H */
diff --git a/arch/x86/include/asm/fixmap.h b/arch/x86/include/asm/fixmap.h
index 8668a94f850..23696d44a0a 100644
--- a/arch/x86/include/asm/fixmap.h
+++ b/arch/x86/include/asm/fixmap.h
@@ -9,6 +9,10 @@
9 9
10extern int fixmaps_set; 10extern int fixmaps_set;
11 11
12extern pte_t *kmap_pte;
13extern pgprot_t kmap_prot;
14extern pte_t *pkmap_page_table;
15
12void __native_set_fixmap(enum fixed_addresses idx, pte_t pte); 16void __native_set_fixmap(enum fixed_addresses idx, pte_t pte);
13void native_set_fixmap(enum fixed_addresses idx, 17void native_set_fixmap(enum fixed_addresses idx,
14 unsigned long phys, pgprot_t flags); 18 unsigned long phys, pgprot_t flags);
diff --git a/arch/x86/include/asm/fixmap_32.h b/arch/x86/include/asm/fixmap_32.h
index 09f29ab5c13..c7115c1d721 100644
--- a/arch/x86/include/asm/fixmap_32.h
+++ b/arch/x86/include/asm/fixmap_32.h
@@ -28,10 +28,8 @@ extern unsigned long __FIXADDR_TOP;
28#include <asm/acpi.h> 28#include <asm/acpi.h>
29#include <asm/apicdef.h> 29#include <asm/apicdef.h>
30#include <asm/page.h> 30#include <asm/page.h>
31#ifdef CONFIG_HIGHMEM
32#include <linux/threads.h> 31#include <linux/threads.h>
33#include <asm/kmap_types.h> 32#include <asm/kmap_types.h>
34#endif
35 33
36/* 34/*
37 * Here we define all the compile-time 'special' virtual 35 * Here we define all the compile-time 'special' virtual
@@ -75,10 +73,8 @@ enum fixed_addresses {
75#ifdef CONFIG_X86_CYCLONE_TIMER 73#ifdef CONFIG_X86_CYCLONE_TIMER
76 FIX_CYCLONE_TIMER, /*cyclone timer register*/ 74 FIX_CYCLONE_TIMER, /*cyclone timer register*/
77#endif 75#endif
78#ifdef CONFIG_HIGHMEM
79 FIX_KMAP_BEGIN, /* reserved pte's for temporary kernel mappings */ 76 FIX_KMAP_BEGIN, /* reserved pte's for temporary kernel mappings */
80 FIX_KMAP_END = FIX_KMAP_BEGIN+(KM_TYPE_NR*NR_CPUS)-1, 77 FIX_KMAP_END = FIX_KMAP_BEGIN+(KM_TYPE_NR*NR_CPUS)-1,
81#endif
82#ifdef CONFIG_PCI_MMCONFIG 78#ifdef CONFIG_PCI_MMCONFIG
83 FIX_PCIE_MCFG, 79 FIX_PCIE_MCFG,
84#endif 80#endif
diff --git a/arch/x86/include/asm/ftrace.h b/arch/x86/include/asm/ftrace.h
index 47f7e65e6c1..9e8bc29b8b1 100644
--- a/arch/x86/include/asm/ftrace.h
+++ b/arch/x86/include/asm/ftrace.h
@@ -1,7 +1,7 @@
1#ifndef _ASM_X86_FTRACE_H 1#ifndef _ASM_X86_FTRACE_H
2#define _ASM_X86_FTRACE_H 2#define _ASM_X86_FTRACE_H
3 3
4#ifdef CONFIG_FTRACE 4#ifdef CONFIG_FUNCTION_TRACER
5#define MCOUNT_ADDR ((long)(mcount)) 5#define MCOUNT_ADDR ((long)(mcount))
6#define MCOUNT_INSN_SIZE 5 /* sizeof mcount call */ 6#define MCOUNT_INSN_SIZE 5 /* sizeof mcount call */
7 7
@@ -19,6 +19,6 @@ static inline unsigned long ftrace_call_adjust(unsigned long addr)
19} 19}
20#endif 20#endif
21 21
22#endif /* CONFIG_FTRACE */ 22#endif /* CONFIG_FUNCTION_TRACER */
23 23
24#endif /* _ASM_X86_FTRACE_H */ 24#endif /* _ASM_X86_FTRACE_H */
diff --git a/arch/x86/include/asm/highmem.h b/arch/x86/include/asm/highmem.h
index a3b3b7c3027..bf9276bea66 100644
--- a/arch/x86/include/asm/highmem.h
+++ b/arch/x86/include/asm/highmem.h
@@ -25,14 +25,11 @@
25#include <asm/kmap_types.h> 25#include <asm/kmap_types.h>
26#include <asm/tlbflush.h> 26#include <asm/tlbflush.h>
27#include <asm/paravirt.h> 27#include <asm/paravirt.h>
28#include <asm/fixmap.h>
28 29
29/* declarations for highmem.c */ 30/* declarations for highmem.c */
30extern unsigned long highstart_pfn, highend_pfn; 31extern unsigned long highstart_pfn, highend_pfn;
31 32
32extern pte_t *kmap_pte;
33extern pgprot_t kmap_prot;
34extern pte_t *pkmap_page_table;
35
36/* 33/*
37 * Right now we initialize only a single pte table. It can be extended 34 * Right now we initialize only a single pte table. It can be extended
38 * easily, subsequent pte tables have to be allocated in one physical 35 * easily, subsequent pte tables have to be allocated in one physical
diff --git a/arch/x86/include/asm/io.h b/arch/x86/include/asm/io.h
index 5618a103f39..ac2abc88cd9 100644
--- a/arch/x86/include/asm/io.h
+++ b/arch/x86/include/asm/io.h
@@ -82,9 +82,9 @@ extern void __iomem *ioremap_wc(unsigned long offset, unsigned long size);
82extern void early_ioremap_init(void); 82extern void early_ioremap_init(void);
83extern void early_ioremap_clear(void); 83extern void early_ioremap_clear(void);
84extern void early_ioremap_reset(void); 84extern void early_ioremap_reset(void);
85extern void *early_ioremap(unsigned long offset, unsigned long size); 85extern void __iomem *early_ioremap(unsigned long offset, unsigned long size);
86extern void *early_memremap(unsigned long offset, unsigned long size); 86extern void __iomem *early_memremap(unsigned long offset, unsigned long size);
87extern void early_iounmap(void *addr, unsigned long size); 87extern void early_iounmap(void __iomem *addr, unsigned long size);
88extern void __iomem *fix_ioremap(unsigned idx, unsigned long phys); 88extern void __iomem *fix_ioremap(unsigned idx, unsigned long phys);
89 89
90 90
diff --git a/arch/x86/include/asm/iommu.h b/arch/x86/include/asm/iommu.h
index 98e28ea8cd1..e4a552d4446 100644
--- a/arch/x86/include/asm/iommu.h
+++ b/arch/x86/include/asm/iommu.h
@@ -7,7 +7,6 @@ extern struct dma_mapping_ops nommu_dma_ops;
7extern int force_iommu, no_iommu; 7extern int force_iommu, no_iommu;
8extern int iommu_detected; 8extern int iommu_detected;
9extern int dmar_disabled; 9extern int dmar_disabled;
10extern int forbid_dac;
11 10
12extern unsigned long iommu_nr_pages(unsigned long addr, unsigned long len); 11extern unsigned long iommu_nr_pages(unsigned long addr, unsigned long len);
13 12
diff --git a/arch/x86/include/asm/irq_vectors.h b/arch/x86/include/asm/irq_vectors.h
index d843ed0e9b2..0005adb0f94 100644
--- a/arch/x86/include/asm/irq_vectors.h
+++ b/arch/x86/include/asm/irq_vectors.h
@@ -101,30 +101,22 @@
101#define LAST_VM86_IRQ 15 101#define LAST_VM86_IRQ 15
102#define invalid_vm86_irq(irq) ((irq) < 3 || (irq) > 15) 102#define invalid_vm86_irq(irq) ((irq) < 3 || (irq) > 15)
103 103
104#ifdef CONFIG_X86_64 104#if defined(CONFIG_X86_IO_APIC) && !defined(CONFIG_X86_VOYAGER)
105# if NR_CPUS < MAX_IO_APICS 105# if NR_CPUS < MAX_IO_APICS
106# define NR_IRQS (NR_VECTORS + (32 * NR_CPUS)) 106# define NR_IRQS (NR_VECTORS + (32 * NR_CPUS))
107# else 107# else
108# define NR_IRQS (NR_VECTORS + (32 * MAX_IO_APICS)) 108# define NR_IRQS (NR_VECTORS + (32 * MAX_IO_APICS))
109# endif 109# endif
110 110
111#elif !defined(CONFIG_X86_VOYAGER) 111#elif defined(CONFIG_X86_VOYAGER)
112 112
113# if defined(CONFIG_X86_IO_APIC) || defined(CONFIG_PARAVIRT) || defined(CONFIG_X86_VISWS) 113# define NR_IRQS 224
114
115# define NR_IRQS 224
116
117# else /* IO_APIC || PARAVIRT */
118
119# define NR_IRQS 16
120
121# endif
122 114
123#else /* !VISWS && !VOYAGER */ 115#else /* IO_APIC || VOYAGER */
124 116
125# define NR_IRQS 224 117# define NR_IRQS 16
126 118
127#endif /* VISWS */ 119#endif
128 120
129/* Voyager specific defines */ 121/* Voyager specific defines */
130/* These define the CPIs we use in linux */ 122/* These define the CPIs we use in linux */
diff --git a/arch/x86/include/asm/kvm_host.h b/arch/x86/include/asm/kvm_host.h
index 65679d00633..8346be87cfa 100644
--- a/arch/x86/include/asm/kvm_host.h
+++ b/arch/x86/include/asm/kvm_host.h
@@ -364,6 +364,9 @@ struct kvm_arch{
364 364
365 struct page *ept_identity_pagetable; 365 struct page *ept_identity_pagetable;
366 bool ept_identity_pagetable_done; 366 bool ept_identity_pagetable_done;
367
368 unsigned long irq_sources_bitmap;
369 unsigned long irq_states[KVM_IOAPIC_NUM_PINS];
367}; 370};
368 371
369struct kvm_vm_stat { 372struct kvm_vm_stat {
diff --git a/arch/x86/include/asm/mach-default/mach_wakecpu.h b/arch/x86/include/asm/mach-default/mach_wakecpu.h
index d5c0b826a4f..9d80db91e99 100644
--- a/arch/x86/include/asm/mach-default/mach_wakecpu.h
+++ b/arch/x86/include/asm/mach-default/mach_wakecpu.h
@@ -33,10 +33,9 @@ static inline void restore_NMI_vector(unsigned short *high, unsigned short *low)
33{ 33{
34} 34}
35 35
36#if APIC_DEBUG 36#define inquire_remote_apic(apicid) do { \
37 #define inquire_remote_apic(apicid) __inquire_remote_apic(apicid) 37 if (apic_verbosity >= APIC_DEBUG) \
38#else 38 __inquire_remote_apic(apicid); \
39 #define inquire_remote_apic(apicid) {} 39 } while (0)
40#endif
41 40
42#endif /* _ASM_X86_MACH_DEFAULT_MACH_WAKECPU_H */ 41#endif /* _ASM_X86_MACH_DEFAULT_MACH_WAKECPU_H */
diff --git a/arch/x86/include/asm/msr.h b/arch/x86/include/asm/msr.h
index 46be2fa7ac2..c2a812ebde8 100644
--- a/arch/x86/include/asm/msr.h
+++ b/arch/x86/include/asm/msr.h
@@ -108,9 +108,7 @@ static __always_inline unsigned long long __native_read_tsc(void)
108{ 108{
109 DECLARE_ARGS(val, low, high); 109 DECLARE_ARGS(val, low, high);
110 110
111 rdtsc_barrier();
112 asm volatile("rdtsc" : EAX_EDX_RET(val, low, high)); 111 asm volatile("rdtsc" : EAX_EDX_RET(val, low, high));
113 rdtsc_barrier();
114 112
115 return EAX_EDX_VAL(val, low, high); 113 return EAX_EDX_VAL(val, low, high);
116} 114}
diff --git a/arch/x86/include/asm/pgtable-3level.h b/arch/x86/include/asm/pgtable-3level.h
index fb16cec702e..52597aeadff 100644
--- a/arch/x86/include/asm/pgtable-3level.h
+++ b/arch/x86/include/asm/pgtable-3level.h
@@ -120,13 +120,13 @@ static inline void pud_clear(pud_t *pudp)
120 write_cr3(pgd); 120 write_cr3(pgd);
121} 121}
122 122
123#define pud_page(pud) ((struct page *) __va(pud_val(pud) & PTE_PFN_MASK)) 123#define pud_page(pud) pfn_to_page(pud_val(pud) >> PAGE_SHIFT)
124 124
125#define pud_page_vaddr(pud) ((unsigned long) __va(pud_val(pud) & PTE_PFN_MASK)) 125#define pud_page_vaddr(pud) ((unsigned long) __va(pud_val(pud) & PTE_PFN_MASK))
126 126
127 127
128/* Find an entry in the second-level page table.. */ 128/* Find an entry in the second-level page table.. */
129#define pmd_offset(pud, address) ((pmd_t *)pud_page(*(pud)) + \ 129#define pmd_offset(pud, address) ((pmd_t *)pud_page_vaddr(*(pud)) + \
130 pmd_index(address)) 130 pmd_index(address))
131 131
132#ifdef CONFIG_SMP 132#ifdef CONFIG_SMP
diff --git a/arch/x86/include/asm/smp.h b/arch/x86/include/asm/smp.h
index 2766021aef8..d12811ce51d 100644
--- a/arch/x86/include/asm/smp.h
+++ b/arch/x86/include/asm/smp.h
@@ -225,5 +225,11 @@ static inline int hard_smp_processor_id(void)
225 225
226#endif /* CONFIG_X86_LOCAL_APIC */ 226#endif /* CONFIG_X86_LOCAL_APIC */
227 227
228#ifdef CONFIG_X86_HAS_BOOT_CPU_ID
229extern unsigned char boot_cpu_id;
230#else
231#define boot_cpu_id 0
232#endif
233
228#endif /* __ASSEMBLY__ */ 234#endif /* __ASSEMBLY__ */
229#endif /* _ASM_X86_SMP_H */ 235#endif /* _ASM_X86_SMP_H */
diff --git a/arch/x86/include/asm/topology.h b/arch/x86/include/asm/topology.h
index 90ac7718469..4850e4b02b6 100644
--- a/arch/x86/include/asm/topology.h
+++ b/arch/x86/include/asm/topology.h
@@ -154,7 +154,7 @@ extern unsigned long node_remap_size[];
154 154
155#endif 155#endif
156 156
157/* sched_domains SD_NODE_INIT for NUMAQ machines */ 157/* sched_domains SD_NODE_INIT for NUMA machines */
158#define SD_NODE_INIT (struct sched_domain) { \ 158#define SD_NODE_INIT (struct sched_domain) { \
159 .min_interval = 8, \ 159 .min_interval = 8, \
160 .max_interval = 32, \ 160 .max_interval = 32, \
@@ -169,8 +169,9 @@ extern unsigned long node_remap_size[];
169 .flags = SD_LOAD_BALANCE \ 169 .flags = SD_LOAD_BALANCE \
170 | SD_BALANCE_EXEC \ 170 | SD_BALANCE_EXEC \
171 | SD_BALANCE_FORK \ 171 | SD_BALANCE_FORK \
172 | SD_SERIALIZE \ 172 | SD_WAKE_AFFINE \
173 | SD_WAKE_BALANCE, \ 173 | SD_WAKE_BALANCE \
174 | SD_SERIALIZE, \
174 .last_balance = jiffies, \ 175 .last_balance = jiffies, \
175 .balance_interval = 1, \ 176 .balance_interval = 1, \
176} 177}
diff --git a/arch/x86/include/asm/tsc.h b/arch/x86/include/asm/tsc.h
index 38ae163cc91..9cd83a8e40d 100644
--- a/arch/x86/include/asm/tsc.h
+++ b/arch/x86/include/asm/tsc.h
@@ -34,6 +34,8 @@ static inline cycles_t get_cycles(void)
34 34
35static __always_inline cycles_t vget_cycles(void) 35static __always_inline cycles_t vget_cycles(void)
36{ 36{
37 cycles_t cycles;
38
37 /* 39 /*
38 * We only do VDSOs on TSC capable CPUs, so this shouldnt 40 * We only do VDSOs on TSC capable CPUs, so this shouldnt
39 * access boot_cpu_data (which is not VDSO-safe): 41 * access boot_cpu_data (which is not VDSO-safe):
@@ -42,7 +44,11 @@ static __always_inline cycles_t vget_cycles(void)
42 if (!cpu_has_tsc) 44 if (!cpu_has_tsc)
43 return 0; 45 return 0;
44#endif 46#endif
45 return (cycles_t)__native_read_tsc(); 47 rdtsc_barrier();
48 cycles = (cycles_t)__native_read_tsc();
49 rdtsc_barrier();
50
51 return cycles;
46} 52}
47 53
48extern void tsc_init(void); 54extern void tsc_init(void);
diff --git a/arch/x86/include/asm/uv/uv_hub.h b/arch/x86/include/asm/uv/uv_hub.h
index c6ad93e315c..7a5782610b2 100644
--- a/arch/x86/include/asm/uv/uv_hub.h
+++ b/arch/x86/include/asm/uv/uv_hub.h
@@ -13,6 +13,7 @@
13 13
14#include <linux/numa.h> 14#include <linux/numa.h>
15#include <linux/percpu.h> 15#include <linux/percpu.h>
16#include <linux/timer.h>
16#include <asm/types.h> 17#include <asm/types.h>
17#include <asm/percpu.h> 18#include <asm/percpu.h>
18 19
diff --git a/arch/x86/include/asm/voyager.h b/arch/x86/include/asm/voyager.h
index 9c811d2e6f9..b3e64730762 100644
--- a/arch/x86/include/asm/voyager.h
+++ b/arch/x86/include/asm/voyager.h
@@ -520,6 +520,7 @@ extern void voyager_restart(void);
520extern void voyager_cat_power_off(void); 520extern void voyager_cat_power_off(void);
521extern void voyager_cat_do_common_interrupt(void); 521extern void voyager_cat_do_common_interrupt(void);
522extern void voyager_handle_nmi(void); 522extern void voyager_handle_nmi(void);
523extern void voyager_smp_intr_init(void);
523/* Commands for the following are */ 524/* Commands for the following are */
524#define VOYAGER_PSI_READ 0 525#define VOYAGER_PSI_READ 0
525#define VOYAGER_PSI_WRITE 1 526#define VOYAGER_PSI_WRITE 1
diff --git a/arch/x86/kernel/Makefile b/arch/x86/kernel/Makefile
index d7e5a58ee22..e489ff9cb3e 100644
--- a/arch/x86/kernel/Makefile
+++ b/arch/x86/kernel/Makefile
@@ -6,11 +6,12 @@ extra-y := head_$(BITS).o head$(BITS).o head.o init_task.o vmlinu
6 6
7CPPFLAGS_vmlinux.lds += -U$(UTS_MACHINE) 7CPPFLAGS_vmlinux.lds += -U$(UTS_MACHINE)
8 8
9ifdef CONFIG_FTRACE 9ifdef CONFIG_FUNCTION_TRACER
10# Do not profile debug and lowlevel utilities 10# Do not profile debug and lowlevel utilities
11CFLAGS_REMOVE_tsc.o = -pg 11CFLAGS_REMOVE_tsc.o = -pg
12CFLAGS_REMOVE_rtc.o = -pg 12CFLAGS_REMOVE_rtc.o = -pg
13CFLAGS_REMOVE_paravirt-spinlocks.o = -pg 13CFLAGS_REMOVE_paravirt-spinlocks.o = -pg
14CFLAGS_REMOVE_ftrace.o = -pg
14endif 15endif
15 16
16# 17#
diff --git a/arch/x86/kernel/amd_iommu.c b/arch/x86/kernel/amd_iommu.c
index a8fd9ebdc8e..331b318304e 100644
--- a/arch/x86/kernel/amd_iommu.c
+++ b/arch/x86/kernel/amd_iommu.c
@@ -50,7 +50,7 @@ static int dma_ops_unity_map(struct dma_ops_domain *dma_dom,
50/* returns !0 if the IOMMU is caching non-present entries in its TLB */ 50/* returns !0 if the IOMMU is caching non-present entries in its TLB */
51static int iommu_has_npcache(struct amd_iommu *iommu) 51static int iommu_has_npcache(struct amd_iommu *iommu)
52{ 52{
53 return iommu->cap & IOMMU_CAP_NPCACHE; 53 return iommu->cap & (1UL << IOMMU_CAP_NPCACHE);
54} 54}
55 55
56/**************************************************************************** 56/****************************************************************************
@@ -536,6 +536,9 @@ static void dma_ops_free_addresses(struct dma_ops_domain *dom,
536{ 536{
537 address >>= PAGE_SHIFT; 537 address >>= PAGE_SHIFT;
538 iommu_area_free(dom->bitmap, address, pages); 538 iommu_area_free(dom->bitmap, address, pages);
539
540 if (address + pages >= dom->next_bit)
541 dom->need_flush = true;
539} 542}
540 543
541/**************************************************************************** 544/****************************************************************************
@@ -992,8 +995,10 @@ static void __unmap_single(struct amd_iommu *iommu,
992 995
993 dma_ops_free_addresses(dma_dom, dma_addr, pages); 996 dma_ops_free_addresses(dma_dom, dma_addr, pages);
994 997
995 if (amd_iommu_unmap_flush) 998 if (amd_iommu_unmap_flush || dma_dom->need_flush) {
996 iommu_flush_pages(iommu, dma_dom->domain.id, dma_addr, size); 999 iommu_flush_pages(iommu, dma_dom->domain.id, dma_addr, size);
1000 dma_dom->need_flush = false;
1001 }
997} 1002}
998 1003
999/* 1004/*
diff --git a/arch/x86/kernel/cpu/addon_cpuid_features.c b/arch/x86/kernel/cpu/addon_cpuid_features.c
index 0d9c993aa93..ef8f831af82 100644
--- a/arch/x86/kernel/cpu/addon_cpuid_features.c
+++ b/arch/x86/kernel/cpu/addon_cpuid_features.c
@@ -69,7 +69,7 @@ void __cpuinit init_scattered_cpuid_features(struct cpuinfo_x86 *c)
69 */ 69 */
70void __cpuinit detect_extended_topology(struct cpuinfo_x86 *c) 70void __cpuinit detect_extended_topology(struct cpuinfo_x86 *c)
71{ 71{
72#ifdef CONFIG_SMP 72#ifdef CONFIG_X86_SMP
73 unsigned int eax, ebx, ecx, edx, sub_index; 73 unsigned int eax, ebx, ecx, edx, sub_index;
74 unsigned int ht_mask_width, core_plus_mask_width; 74 unsigned int ht_mask_width, core_plus_mask_width;
75 unsigned int core_select_mask, core_level_siblings; 75 unsigned int core_select_mask, core_level_siblings;
diff --git a/arch/x86/kernel/cpu/common.c b/arch/x86/kernel/cpu/common.c
index 25581dcb280..b9c9ea0217a 100644
--- a/arch/x86/kernel/cpu/common.c
+++ b/arch/x86/kernel/cpu/common.c
@@ -20,6 +20,7 @@
20#include <asm/pat.h> 20#include <asm/pat.h>
21#include <asm/asm.h> 21#include <asm/asm.h>
22#include <asm/numa.h> 22#include <asm/numa.h>
23#include <asm/smp.h>
23#ifdef CONFIG_X86_LOCAL_APIC 24#ifdef CONFIG_X86_LOCAL_APIC
24#include <asm/mpspec.h> 25#include <asm/mpspec.h>
25#include <asm/apic.h> 26#include <asm/apic.h>
@@ -549,6 +550,10 @@ static void __init early_identify_cpu(struct cpuinfo_x86 *c)
549 this_cpu->c_early_init(c); 550 this_cpu->c_early_init(c);
550 551
551 validate_pat_support(c); 552 validate_pat_support(c);
553
554#ifdef CONFIG_SMP
555 c->cpu_index = boot_cpu_id;
556#endif
552} 557}
553 558
554void __init early_cpu_init(void) 559void __init early_cpu_init(void)
@@ -1134,7 +1139,7 @@ void __cpuinit cpu_init(void)
1134 /* 1139 /*
1135 * Boot processor to setup the FP and extended state context info. 1140 * Boot processor to setup the FP and extended state context info.
1136 */ 1141 */
1137 if (!smp_processor_id()) 1142 if (smp_processor_id() == boot_cpu_id)
1138 init_thread_xstate(); 1143 init_thread_xstate();
1139 1144
1140 xsave_init(); 1145 xsave_init();
diff --git a/arch/x86/kernel/e820.c b/arch/x86/kernel/e820.c
index ce97bf3bed1..7aafeb5263e 100644
--- a/arch/x86/kernel/e820.c
+++ b/arch/x86/kernel/e820.c
@@ -1290,15 +1290,17 @@ void __init e820_reserve_resources(void)
1290 res->start = e820.map[i].addr; 1290 res->start = e820.map[i].addr;
1291 res->end = end; 1291 res->end = end;
1292 1292
1293 res->flags = IORESOURCE_MEM | IORESOURCE_BUSY; 1293 res->flags = IORESOURCE_MEM;
1294 1294
1295 /* 1295 /*
1296 * don't register the region that could be conflicted with 1296 * don't register the region that could be conflicted with
1297 * pci device BAR resource and insert them later in 1297 * pci device BAR resource and insert them later in
1298 * pcibios_resource_survey() 1298 * pcibios_resource_survey()
1299 */ 1299 */
1300 if (e820.map[i].type != E820_RESERVED || res->start < (1ULL<<20)) 1300 if (e820.map[i].type != E820_RESERVED || res->start < (1ULL<<20)) {
1301 res->flags |= IORESOURCE_BUSY;
1301 insert_resource(&iomem_resource, res); 1302 insert_resource(&iomem_resource, res);
1303 }
1302 res++; 1304 res++;
1303 } 1305 }
1304 1306
@@ -1318,7 +1320,7 @@ void __init e820_reserve_resources_late(void)
1318 res = e820_res; 1320 res = e820_res;
1319 for (i = 0; i < e820.nr_map; i++) { 1321 for (i = 0; i < e820.nr_map; i++) {
1320 if (!res->parent && res->end) 1322 if (!res->parent && res->end)
1321 reserve_region_with_split(&iomem_resource, res->start, res->end, res->name); 1323 insert_resource_expand_to_fit(&iomem_resource, res);
1322 res++; 1324 res++;
1323 } 1325 }
1324} 1326}
diff --git a/arch/x86/kernel/entry_32.S b/arch/x86/kernel/entry_32.S
index dd65143941a..28b597ef9ca 100644
--- a/arch/x86/kernel/entry_32.S
+++ b/arch/x86/kernel/entry_32.S
@@ -1149,7 +1149,7 @@ ENDPROC(xen_failsafe_callback)
1149 1149
1150#endif /* CONFIG_XEN */ 1150#endif /* CONFIG_XEN */
1151 1151
1152#ifdef CONFIG_FTRACE 1152#ifdef CONFIG_FUNCTION_TRACER
1153#ifdef CONFIG_DYNAMIC_FTRACE 1153#ifdef CONFIG_DYNAMIC_FTRACE
1154 1154
1155ENTRY(mcount) 1155ENTRY(mcount)
@@ -1204,7 +1204,7 @@ trace:
1204 jmp ftrace_stub 1204 jmp ftrace_stub
1205END(mcount) 1205END(mcount)
1206#endif /* CONFIG_DYNAMIC_FTRACE */ 1206#endif /* CONFIG_DYNAMIC_FTRACE */
1207#endif /* CONFIG_FTRACE */ 1207#endif /* CONFIG_FUNCTION_TRACER */
1208 1208
1209.section .rodata,"a" 1209.section .rodata,"a"
1210#include "syscall_table_32.S" 1210#include "syscall_table_32.S"
diff --git a/arch/x86/kernel/entry_64.S b/arch/x86/kernel/entry_64.S
index 09e7145484c..b86f332c96a 100644
--- a/arch/x86/kernel/entry_64.S
+++ b/arch/x86/kernel/entry_64.S
@@ -61,7 +61,7 @@
61 61
62 .code64 62 .code64
63 63
64#ifdef CONFIG_FTRACE 64#ifdef CONFIG_FUNCTION_TRACER
65#ifdef CONFIG_DYNAMIC_FTRACE 65#ifdef CONFIG_DYNAMIC_FTRACE
66ENTRY(mcount) 66ENTRY(mcount)
67 retq 67 retq
@@ -138,7 +138,7 @@ trace:
138 jmp ftrace_stub 138 jmp ftrace_stub
139END(mcount) 139END(mcount)
140#endif /* CONFIG_DYNAMIC_FTRACE */ 140#endif /* CONFIG_DYNAMIC_FTRACE */
141#endif /* CONFIG_FTRACE */ 141#endif /* CONFIG_FUNCTION_TRACER */
142 142
143#ifndef CONFIG_PREEMPT 143#ifndef CONFIG_PREEMPT
144#define retint_kernel retint_restore_args 144#define retint_kernel retint_restore_args
diff --git a/arch/x86/kernel/ftrace.c b/arch/x86/kernel/ftrace.c
index d073d981a73..50ea0ac8c9b 100644
--- a/arch/x86/kernel/ftrace.c
+++ b/arch/x86/kernel/ftrace.c
@@ -21,8 +21,7 @@
21#include <asm/nops.h> 21#include <asm/nops.h>
22 22
23 23
24/* Long is fine, even if it is only 4 bytes ;-) */ 24static unsigned char ftrace_nop[MCOUNT_INSN_SIZE];
25static unsigned long *ftrace_nop;
26 25
27union ftrace_code_union { 26union ftrace_code_union {
28 char code[MCOUNT_INSN_SIZE]; 27 char code[MCOUNT_INSN_SIZE];
@@ -33,17 +32,17 @@ union ftrace_code_union {
33}; 32};
34 33
35 34
36static int notrace ftrace_calc_offset(long ip, long addr) 35static int ftrace_calc_offset(long ip, long addr)
37{ 36{
38 return (int)(addr - ip); 37 return (int)(addr - ip);
39} 38}
40 39
41notrace unsigned char *ftrace_nop_replace(void) 40unsigned char *ftrace_nop_replace(void)
42{ 41{
43 return (char *)ftrace_nop; 42 return ftrace_nop;
44} 43}
45 44
46notrace unsigned char *ftrace_call_replace(unsigned long ip, unsigned long addr) 45unsigned char *ftrace_call_replace(unsigned long ip, unsigned long addr)
47{ 46{
48 static union ftrace_code_union calc; 47 static union ftrace_code_union calc;
49 48
@@ -57,7 +56,7 @@ notrace unsigned char *ftrace_call_replace(unsigned long ip, unsigned long addr)
57 return calc.code; 56 return calc.code;
58} 57}
59 58
60notrace int 59int
61ftrace_modify_code(unsigned long ip, unsigned char *old_code, 60ftrace_modify_code(unsigned long ip, unsigned char *old_code,
62 unsigned char *new_code) 61 unsigned char *new_code)
63{ 62{
@@ -66,26 +65,31 @@ ftrace_modify_code(unsigned long ip, unsigned char *old_code,
66 /* 65 /*
67 * Note: Due to modules and __init, code can 66 * Note: Due to modules and __init, code can
68 * disappear and change, we need to protect against faulting 67 * disappear and change, we need to protect against faulting
69 * as well as code changing. 68 * as well as code changing. We do this by using the
69 * probe_kernel_* functions.
70 * 70 *
71 * No real locking needed, this code is run through 71 * No real locking needed, this code is run through
72 * kstop_machine, or before SMP starts. 72 * kstop_machine, or before SMP starts.
73 */ 73 */
74 if (__copy_from_user_inatomic(replaced, (char __user *)ip, MCOUNT_INSN_SIZE))
75 return 1;
76 74
75 /* read the text we want to modify */
76 if (probe_kernel_read(replaced, (void *)ip, MCOUNT_INSN_SIZE))
77 return -EFAULT;
78
79 /* Make sure it is what we expect it to be */
77 if (memcmp(replaced, old_code, MCOUNT_INSN_SIZE) != 0) 80 if (memcmp(replaced, old_code, MCOUNT_INSN_SIZE) != 0)
78 return 2; 81 return -EINVAL;
79 82
80 WARN_ON_ONCE(__copy_to_user_inatomic((char __user *)ip, new_code, 83 /* replace the text with the new text */
81 MCOUNT_INSN_SIZE)); 84 if (probe_kernel_write((void *)ip, new_code, MCOUNT_INSN_SIZE))
85 return -EPERM;
82 86
83 sync_core(); 87 sync_core();
84 88
85 return 0; 89 return 0;
86} 90}
87 91
88notrace int ftrace_update_ftrace_func(ftrace_func_t func) 92int ftrace_update_ftrace_func(ftrace_func_t func)
89{ 93{
90 unsigned long ip = (unsigned long)(&ftrace_call); 94 unsigned long ip = (unsigned long)(&ftrace_call);
91 unsigned char old[MCOUNT_INSN_SIZE], *new; 95 unsigned char old[MCOUNT_INSN_SIZE], *new;
@@ -98,13 +102,6 @@ notrace int ftrace_update_ftrace_func(ftrace_func_t func)
98 return ret; 102 return ret;
99} 103}
100 104
101notrace int ftrace_mcount_set(unsigned long *data)
102{
103 /* mcount is initialized as a nop */
104 *data = 0;
105 return 0;
106}
107
108int __init ftrace_dyn_arch_init(void *data) 105int __init ftrace_dyn_arch_init(void *data)
109{ 106{
110 extern const unsigned char ftrace_test_p6nop[]; 107 extern const unsigned char ftrace_test_p6nop[];
@@ -127,9 +124,6 @@ int __init ftrace_dyn_arch_init(void *data)
127 * TODO: check the cpuid to determine the best nop. 124 * TODO: check the cpuid to determine the best nop.
128 */ 125 */
129 asm volatile ( 126 asm volatile (
130 "jmp ftrace_test_jmp\n"
131 /* This code needs to stay around */
132 ".section .text, \"ax\"\n"
133 "ftrace_test_jmp:" 127 "ftrace_test_jmp:"
134 "jmp ftrace_test_p6nop\n" 128 "jmp ftrace_test_p6nop\n"
135 "nop\n" 129 "nop\n"
@@ -140,8 +134,6 @@ int __init ftrace_dyn_arch_init(void *data)
140 "jmp 1f\n" 134 "jmp 1f\n"
141 "ftrace_test_nop5:" 135 "ftrace_test_nop5:"
142 ".byte 0x66,0x66,0x66,0x66,0x90\n" 136 ".byte 0x66,0x66,0x66,0x66,0x90\n"
143 "jmp 1f\n"
144 ".previous\n"
145 "1:" 137 "1:"
146 ".section .fixup, \"ax\"\n" 138 ".section .fixup, \"ax\"\n"
147 "2: movl $1, %0\n" 139 "2: movl $1, %0\n"
@@ -156,15 +148,15 @@ int __init ftrace_dyn_arch_init(void *data)
156 switch (faulted) { 148 switch (faulted) {
157 case 0: 149 case 0:
158 pr_info("ftrace: converting mcount calls to 0f 1f 44 00 00\n"); 150 pr_info("ftrace: converting mcount calls to 0f 1f 44 00 00\n");
159 ftrace_nop = (unsigned long *)ftrace_test_p6nop; 151 memcpy(ftrace_nop, ftrace_test_p6nop, MCOUNT_INSN_SIZE);
160 break; 152 break;
161 case 1: 153 case 1:
162 pr_info("ftrace: converting mcount calls to 66 66 66 66 90\n"); 154 pr_info("ftrace: converting mcount calls to 66 66 66 66 90\n");
163 ftrace_nop = (unsigned long *)ftrace_test_nop5; 155 memcpy(ftrace_nop, ftrace_test_nop5, MCOUNT_INSN_SIZE);
164 break; 156 break;
165 case 2: 157 case 2:
166 pr_info("ftrace: converting mcount calls to jmp . + 5\n"); 158 pr_info("ftrace: converting mcount calls to jmp . + 5\n");
167 ftrace_nop = (unsigned long *)ftrace_test_jmp; 159 memcpy(ftrace_nop, ftrace_test_jmp, MCOUNT_INSN_SIZE);
168 break; 160 break;
169 } 161 }
170 162
diff --git a/arch/x86/kernel/genx2apic_uv_x.c b/arch/x86/kernel/genx2apic_uv_x.c
index 680a06557c5..2c7dbdb9827 100644
--- a/arch/x86/kernel/genx2apic_uv_x.c
+++ b/arch/x86/kernel/genx2apic_uv_x.c
@@ -15,7 +15,6 @@
15#include <linux/ctype.h> 15#include <linux/ctype.h>
16#include <linux/init.h> 16#include <linux/init.h>
17#include <linux/sched.h> 17#include <linux/sched.h>
18#include <linux/bootmem.h>
19#include <linux/module.h> 18#include <linux/module.h>
20#include <linux/hardirq.h> 19#include <linux/hardirq.h>
21#include <asm/smp.h> 20#include <asm/smp.h>
@@ -398,16 +397,16 @@ void __init uv_system_init(void)
398 printk(KERN_DEBUG "UV: Found %d blades\n", uv_num_possible_blades()); 397 printk(KERN_DEBUG "UV: Found %d blades\n", uv_num_possible_blades());
399 398
400 bytes = sizeof(struct uv_blade_info) * uv_num_possible_blades(); 399 bytes = sizeof(struct uv_blade_info) * uv_num_possible_blades();
401 uv_blade_info = alloc_bootmem_pages(bytes); 400 uv_blade_info = kmalloc(bytes, GFP_KERNEL);
402 401
403 get_lowmem_redirect(&lowmem_redir_base, &lowmem_redir_size); 402 get_lowmem_redirect(&lowmem_redir_base, &lowmem_redir_size);
404 403
405 bytes = sizeof(uv_node_to_blade[0]) * num_possible_nodes(); 404 bytes = sizeof(uv_node_to_blade[0]) * num_possible_nodes();
406 uv_node_to_blade = alloc_bootmem_pages(bytes); 405 uv_node_to_blade = kmalloc(bytes, GFP_KERNEL);
407 memset(uv_node_to_blade, 255, bytes); 406 memset(uv_node_to_blade, 255, bytes);
408 407
409 bytes = sizeof(uv_cpu_to_blade[0]) * num_possible_cpus(); 408 bytes = sizeof(uv_cpu_to_blade[0]) * num_possible_cpus();
410 uv_cpu_to_blade = alloc_bootmem_pages(bytes); 409 uv_cpu_to_blade = kmalloc(bytes, GFP_KERNEL);
411 memset(uv_cpu_to_blade, 255, bytes); 410 memset(uv_cpu_to_blade, 255, bytes);
412 411
413 blade = 0; 412 blade = 0;
diff --git a/arch/x86/kernel/i386_ksyms_32.c b/arch/x86/kernel/i386_ksyms_32.c
index dd7ebee446a..43cec6bdda6 100644
--- a/arch/x86/kernel/i386_ksyms_32.c
+++ b/arch/x86/kernel/i386_ksyms_32.c
@@ -5,7 +5,7 @@
5#include <asm/desc.h> 5#include <asm/desc.h>
6#include <asm/ftrace.h> 6#include <asm/ftrace.h>
7 7
8#ifdef CONFIG_FTRACE 8#ifdef CONFIG_FUNCTION_TRACER
9/* mcount is defined in assembly */ 9/* mcount is defined in assembly */
10EXPORT_SYMBOL(mcount); 10EXPORT_SYMBOL(mcount);
11#endif 11#endif
diff --git a/arch/x86/kernel/io_apic.c b/arch/x86/kernel/io_apic.c
index b764d7429c6..7a3f2028e2e 100644
--- a/arch/x86/kernel/io_apic.c
+++ b/arch/x86/kernel/io_apic.c
@@ -3611,6 +3611,8 @@ int __init probe_nr_irqs(void)
3611 /* something wrong ? */ 3611 /* something wrong ? */
3612 if (nr < nr_min) 3612 if (nr < nr_min)
3613 nr = nr_min; 3613 nr = nr_min;
3614 if (WARN_ON(nr > NR_IRQS))
3615 nr = NR_IRQS;
3614 3616
3615 return nr; 3617 return nr;
3616} 3618}
diff --git a/arch/x86/kernel/k8.c b/arch/x86/kernel/k8.c
index 304d8bad655..cbc4332a77b 100644
--- a/arch/x86/kernel/k8.c
+++ b/arch/x86/kernel/k8.c
@@ -18,7 +18,6 @@ static u32 *flush_words;
18struct pci_device_id k8_nb_ids[] = { 18struct pci_device_id k8_nb_ids[] = {
19 { PCI_DEVICE(PCI_VENDOR_ID_AMD, PCI_DEVICE_ID_AMD_K8_NB_MISC) }, 19 { PCI_DEVICE(PCI_VENDOR_ID_AMD, PCI_DEVICE_ID_AMD_K8_NB_MISC) },
20 { PCI_DEVICE(PCI_VENDOR_ID_AMD, PCI_DEVICE_ID_AMD_10H_NB_MISC) }, 20 { PCI_DEVICE(PCI_VENDOR_ID_AMD, PCI_DEVICE_ID_AMD_10H_NB_MISC) },
21 { PCI_DEVICE(PCI_VENDOR_ID_AMD, PCI_DEVICE_ID_AMD_11H_NB_MISC) },
22 {} 21 {}
23}; 22};
24EXPORT_SYMBOL(k8_nb_ids); 23EXPORT_SYMBOL(k8_nb_ids);
diff --git a/arch/x86/kernel/machine_kexec_32.c b/arch/x86/kernel/machine_kexec_32.c
index 0732adba05c..7a385746509 100644
--- a/arch/x86/kernel/machine_kexec_32.c
+++ b/arch/x86/kernel/machine_kexec_32.c
@@ -162,7 +162,10 @@ void machine_kexec(struct kimage *image)
162 page_list[VA_PTE_0] = (unsigned long)kexec_pte0; 162 page_list[VA_PTE_0] = (unsigned long)kexec_pte0;
163 page_list[PA_PTE_1] = __pa(kexec_pte1); 163 page_list[PA_PTE_1] = __pa(kexec_pte1);
164 page_list[VA_PTE_1] = (unsigned long)kexec_pte1; 164 page_list[VA_PTE_1] = (unsigned long)kexec_pte1;
165 page_list[PA_SWAP_PAGE] = (page_to_pfn(image->swap_page) << PAGE_SHIFT); 165
166 if (image->type == KEXEC_TYPE_DEFAULT)
167 page_list[PA_SWAP_PAGE] = (page_to_pfn(image->swap_page)
168 << PAGE_SHIFT);
166 169
167 /* The segment registers are funny things, they have both a 170 /* The segment registers are funny things, they have both a
168 * visible and an invisible part. Whenever the visible part is 171 * visible and an invisible part. Whenever the visible part is
diff --git a/arch/x86/kernel/microcode_amd.c b/arch/x86/kernel/microcode_amd.c
index 7a1f8eeac2c..5f8e5d75a25 100644
--- a/arch/x86/kernel/microcode_amd.c
+++ b/arch/x86/kernel/microcode_amd.c
@@ -39,7 +39,7 @@
39#include <asm/microcode.h> 39#include <asm/microcode.h>
40 40
41MODULE_DESCRIPTION("AMD Microcode Update Driver"); 41MODULE_DESCRIPTION("AMD Microcode Update Driver");
42MODULE_AUTHOR("Peter Oruba <peter.oruba@amd.com>"); 42MODULE_AUTHOR("Peter Oruba");
43MODULE_LICENSE("GPL v2"); 43MODULE_LICENSE("GPL v2");
44 44
45#define UCODE_MAGIC 0x00414d44 45#define UCODE_MAGIC 0x00414d44
diff --git a/arch/x86/kernel/microcode_core.c b/arch/x86/kernel/microcode_core.c
index 936d8d55f23..82fb2809ce3 100644
--- a/arch/x86/kernel/microcode_core.c
+++ b/arch/x86/kernel/microcode_core.c
@@ -480,8 +480,8 @@ static int __init microcode_init(void)
480 480
481 printk(KERN_INFO 481 printk(KERN_INFO
482 "Microcode Update Driver: v" MICROCODE_VERSION 482 "Microcode Update Driver: v" MICROCODE_VERSION
483 " <tigran@aivazian.fsnet.co.uk>" 483 " <tigran@aivazian.fsnet.co.uk>,"
484 " <peter.oruba@amd.com>\n"); 484 " Peter Oruba\n");
485 485
486 return 0; 486 return 0;
487} 487}
diff --git a/arch/x86/kernel/pci-dma.c b/arch/x86/kernel/pci-dma.c
index 1972266e8ba..19262482021 100644
--- a/arch/x86/kernel/pci-dma.c
+++ b/arch/x86/kernel/pci-dma.c
@@ -9,6 +9,8 @@
9#include <asm/calgary.h> 9#include <asm/calgary.h>
10#include <asm/amd_iommu.h> 10#include <asm/amd_iommu.h>
11 11
12static int forbid_dac __read_mostly;
13
12struct dma_mapping_ops *dma_ops; 14struct dma_mapping_ops *dma_ops;
13EXPORT_SYMBOL(dma_ops); 15EXPORT_SYMBOL(dma_ops);
14 16
@@ -291,3 +293,17 @@ void pci_iommu_shutdown(void)
291} 293}
292/* Must execute after PCI subsystem */ 294/* Must execute after PCI subsystem */
293fs_initcall(pci_iommu_init); 295fs_initcall(pci_iommu_init);
296
297#ifdef CONFIG_PCI
298/* Many VIA bridges seem to corrupt data for DAC. Disable it here */
299
300static __devinit void via_no_dac(struct pci_dev *dev)
301{
302 if ((dev->class >> 8) == PCI_CLASS_BRIDGE_PCI && forbid_dac == 0) {
303 printk(KERN_INFO "PCI: VIA PCI bridge detected."
304 "Disabling DAC.\n");
305 forbid_dac = 1;
306 }
307}
308DECLARE_PCI_FIXUP_FINAL(PCI_VENDOR_ID_VIA, PCI_ANY_ID, via_no_dac);
309#endif
diff --git a/arch/x86/kernel/pci-gart_64.c b/arch/x86/kernel/pci-gart_64.c
index e3f75bbcede..a42b02b4df6 100644
--- a/arch/x86/kernel/pci-gart_64.c
+++ b/arch/x86/kernel/pci-gart_64.c
@@ -744,7 +744,7 @@ void __init gart_iommu_init(void)
744 long i; 744 long i;
745 745
746 if (cache_k8_northbridges() < 0 || num_k8_northbridges == 0) { 746 if (cache_k8_northbridges() < 0 || num_k8_northbridges == 0) {
747 printk(KERN_INFO "PCI-GART: No AMD northbridge found.\n"); 747 printk(KERN_INFO "PCI-GART: No AMD GART found.\n");
748 return; 748 return;
749 } 749 }
750 750
diff --git a/arch/x86/kernel/pci-swiotlb_64.c b/arch/x86/kernel/pci-swiotlb_64.c
index c4ce0332759..3c539d111ab 100644
--- a/arch/x86/kernel/pci-swiotlb_64.c
+++ b/arch/x86/kernel/pci-swiotlb_64.c
@@ -18,9 +18,21 @@ swiotlb_map_single_phys(struct device *hwdev, phys_addr_t paddr, size_t size,
18 return swiotlb_map_single(hwdev, phys_to_virt(paddr), size, direction); 18 return swiotlb_map_single(hwdev, phys_to_virt(paddr), size, direction);
19} 19}
20 20
21static void *x86_swiotlb_alloc_coherent(struct device *hwdev, size_t size,
22 dma_addr_t *dma_handle, gfp_t flags)
23{
24 void *vaddr;
25
26 vaddr = dma_generic_alloc_coherent(hwdev, size, dma_handle, flags);
27 if (vaddr)
28 return vaddr;
29
30 return swiotlb_alloc_coherent(hwdev, size, dma_handle, flags);
31}
32
21struct dma_mapping_ops swiotlb_dma_ops = { 33struct dma_mapping_ops swiotlb_dma_ops = {
22 .mapping_error = swiotlb_dma_mapping_error, 34 .mapping_error = swiotlb_dma_mapping_error,
23 .alloc_coherent = swiotlb_alloc_coherent, 35 .alloc_coherent = x86_swiotlb_alloc_coherent,
24 .free_coherent = swiotlb_free_coherent, 36 .free_coherent = swiotlb_free_coherent,
25 .map_single = swiotlb_map_single_phys, 37 .map_single = swiotlb_map_single_phys,
26 .unmap_single = swiotlb_unmap_single, 38 .unmap_single = swiotlb_unmap_single,
diff --git a/arch/x86/kernel/reboot.c b/arch/x86/kernel/reboot.c
index f4c93f1cfc1..724adfc63cb 100644
--- a/arch/x86/kernel/reboot.c
+++ b/arch/x86/kernel/reboot.c
@@ -29,11 +29,7 @@ EXPORT_SYMBOL(pm_power_off);
29 29
30static const struct desc_ptr no_idt = {}; 30static const struct desc_ptr no_idt = {};
31static int reboot_mode; 31static int reboot_mode;
32/* 32enum reboot_type reboot_type = BOOT_KBD;
33 * Keyboard reset and triple fault may result in INIT, not RESET, which
34 * doesn't work when we're in vmx root mode. Try ACPI first.
35 */
36enum reboot_type reboot_type = BOOT_ACPI;
37int reboot_force; 33int reboot_force;
38 34
39#if defined(CONFIG_X86_32) && defined(CONFIG_SMP) 35#if defined(CONFIG_X86_32) && defined(CONFIG_SMP)
diff --git a/arch/x86/kernel/tlb_32.c b/arch/x86/kernel/tlb_32.c
index e00534b3353..f4049f3513b 100644
--- a/arch/x86/kernel/tlb_32.c
+++ b/arch/x86/kernel/tlb_32.c
@@ -154,6 +154,12 @@ void native_flush_tlb_others(const cpumask_t *cpumaskp, struct mm_struct *mm,
154 flush_mm = mm; 154 flush_mm = mm;
155 flush_va = va; 155 flush_va = va;
156 cpus_or(flush_cpumask, cpumask, flush_cpumask); 156 cpus_or(flush_cpumask, cpumask, flush_cpumask);
157
158 /*
159 * Make the above memory operations globally visible before
160 * sending the IPI.
161 */
162 smp_mb();
157 /* 163 /*
158 * We have to send the IPI only to 164 * We have to send the IPI only to
159 * CPUs affected. 165 * CPUs affected.
diff --git a/arch/x86/kernel/tlb_64.c b/arch/x86/kernel/tlb_64.c
index dcbf7a1159e..8f919ca6949 100644
--- a/arch/x86/kernel/tlb_64.c
+++ b/arch/x86/kernel/tlb_64.c
@@ -183,6 +183,11 @@ void native_flush_tlb_others(const cpumask_t *cpumaskp, struct mm_struct *mm,
183 cpus_or(f->flush_cpumask, cpumask, f->flush_cpumask); 183 cpus_or(f->flush_cpumask, cpumask, f->flush_cpumask);
184 184
185 /* 185 /*
186 * Make the above memory operations globally visible before
187 * sending the IPI.
188 */
189 smp_mb();
190 /*
186 * We have to send the IPI only to 191 * We have to send the IPI only to
187 * CPUs affected. 192 * CPUs affected.
188 */ 193 */
diff --git a/arch/x86/kernel/tsc.c b/arch/x86/kernel/tsc.c
index 161bb850fc4..424093b157d 100644
--- a/arch/x86/kernel/tsc.c
+++ b/arch/x86/kernel/tsc.c
@@ -55,7 +55,7 @@ u64 native_sched_clock(void)
55 rdtscll(this_offset); 55 rdtscll(this_offset);
56 56
57 /* return the value in ns */ 57 /* return the value in ns */
58 return cycles_2_ns(this_offset); 58 return __cycles_2_ns(this_offset);
59} 59}
60 60
61/* We need to define a real function for sched_clock, to override the 61/* We need to define a real function for sched_clock, to override the
@@ -759,7 +759,7 @@ __cpuinit int unsynchronized_tsc(void)
759 if (!cpu_has_tsc || tsc_unstable) 759 if (!cpu_has_tsc || tsc_unstable)
760 return 1; 760 return 1;
761 761
762#ifdef CONFIG_SMP 762#ifdef CONFIG_X86_SMP
763 if (apic_is_clustered_box()) 763 if (apic_is_clustered_box())
764 return 1; 764 return 1;
765#endif 765#endif
@@ -813,10 +813,6 @@ void __init tsc_init(void)
813 cpu_khz = calibrate_cpu(); 813 cpu_khz = calibrate_cpu();
814#endif 814#endif
815 815
816 lpj = ((u64)tsc_khz * 1000);
817 do_div(lpj, HZ);
818 lpj_fine = lpj;
819
820 printk("Detected %lu.%03lu MHz processor.\n", 816 printk("Detected %lu.%03lu MHz processor.\n",
821 (unsigned long)cpu_khz / 1000, 817 (unsigned long)cpu_khz / 1000,
822 (unsigned long)cpu_khz % 1000); 818 (unsigned long)cpu_khz % 1000);
@@ -836,6 +832,10 @@ void __init tsc_init(void)
836 /* now allow native_sched_clock() to use rdtsc */ 832 /* now allow native_sched_clock() to use rdtsc */
837 tsc_disabled = 0; 833 tsc_disabled = 0;
838 834
835 lpj = ((u64)tsc_khz * 1000);
836 do_div(lpj, HZ);
837 lpj_fine = lpj;
838
839 use_tsc_delay(); 839 use_tsc_delay();
840 /* Check and install the TSC clocksource */ 840 /* Check and install the TSC clocksource */
841 dmi_check_system(bad_tsc_dmi_table); 841 dmi_check_system(bad_tsc_dmi_table);
diff --git a/arch/x86/kernel/vsmp_64.c b/arch/x86/kernel/vsmp_64.c
index 7766d36983f..a688f3bfaec 100644
--- a/arch/x86/kernel/vsmp_64.c
+++ b/arch/x86/kernel/vsmp_64.c
@@ -78,7 +78,7 @@ static unsigned __init_or_module vsmp_patch(u8 type, u16 clobbers, void *ibuf,
78 78
79static void __init set_vsmp_pv_ops(void) 79static void __init set_vsmp_pv_ops(void)
80{ 80{
81 void *address; 81 void __iomem *address;
82 unsigned int cap, ctl, cfg; 82 unsigned int cap, ctl, cfg;
83 83
84 /* set vSMP magic bits to indicate vSMP capable kernel */ 84 /* set vSMP magic bits to indicate vSMP capable kernel */
diff --git a/arch/x86/kernel/x8664_ksyms_64.c b/arch/x86/kernel/x8664_ksyms_64.c
index b545f371b5f..695e426aa35 100644
--- a/arch/x86/kernel/x8664_ksyms_64.c
+++ b/arch/x86/kernel/x8664_ksyms_64.c
@@ -12,7 +12,7 @@
12#include <asm/desc.h> 12#include <asm/desc.h>
13#include <asm/ftrace.h> 13#include <asm/ftrace.h>
14 14
15#ifdef CONFIG_FTRACE 15#ifdef CONFIG_FUNCTION_TRACER
16/* mcount is defined in assembly */ 16/* mcount is defined in assembly */
17EXPORT_SYMBOL(mcount); 17EXPORT_SYMBOL(mcount);
18#endif 18#endif
diff --git a/arch/x86/kvm/i8254.c b/arch/x86/kvm/i8254.c
index 11c6725fb79..8772dc94682 100644
--- a/arch/x86/kvm/i8254.c
+++ b/arch/x86/kvm/i8254.c
@@ -545,6 +545,12 @@ struct kvm_pit *kvm_create_pit(struct kvm *kvm)
545 if (!pit) 545 if (!pit)
546 return NULL; 546 return NULL;
547 547
548 mutex_lock(&kvm->lock);
549 pit->irq_source_id = kvm_request_irq_source_id(kvm);
550 mutex_unlock(&kvm->lock);
551 if (pit->irq_source_id < 0)
552 return NULL;
553
548 mutex_init(&pit->pit_state.lock); 554 mutex_init(&pit->pit_state.lock);
549 mutex_lock(&pit->pit_state.lock); 555 mutex_lock(&pit->pit_state.lock);
550 spin_lock_init(&pit->pit_state.inject_lock); 556 spin_lock_init(&pit->pit_state.inject_lock);
@@ -587,6 +593,7 @@ void kvm_free_pit(struct kvm *kvm)
587 mutex_lock(&kvm->arch.vpit->pit_state.lock); 593 mutex_lock(&kvm->arch.vpit->pit_state.lock);
588 timer = &kvm->arch.vpit->pit_state.pit_timer.timer; 594 timer = &kvm->arch.vpit->pit_state.pit_timer.timer;
589 hrtimer_cancel(timer); 595 hrtimer_cancel(timer);
596 kvm_free_irq_source_id(kvm, kvm->arch.vpit->irq_source_id);
590 mutex_unlock(&kvm->arch.vpit->pit_state.lock); 597 mutex_unlock(&kvm->arch.vpit->pit_state.lock);
591 kfree(kvm->arch.vpit); 598 kfree(kvm->arch.vpit);
592 } 599 }
@@ -595,8 +602,8 @@ void kvm_free_pit(struct kvm *kvm)
595static void __inject_pit_timer_intr(struct kvm *kvm) 602static void __inject_pit_timer_intr(struct kvm *kvm)
596{ 603{
597 mutex_lock(&kvm->lock); 604 mutex_lock(&kvm->lock);
598 kvm_set_irq(kvm, 0, 1); 605 kvm_set_irq(kvm, kvm->arch.vpit->irq_source_id, 0, 1);
599 kvm_set_irq(kvm, 0, 0); 606 kvm_set_irq(kvm, kvm->arch.vpit->irq_source_id, 0, 0);
600 mutex_unlock(&kvm->lock); 607 mutex_unlock(&kvm->lock);
601} 608}
602 609
diff --git a/arch/x86/kvm/i8254.h b/arch/x86/kvm/i8254.h
index e436d4983aa..4178022b97a 100644
--- a/arch/x86/kvm/i8254.h
+++ b/arch/x86/kvm/i8254.h
@@ -44,6 +44,7 @@ struct kvm_pit {
44 struct kvm_io_device speaker_dev; 44 struct kvm_io_device speaker_dev;
45 struct kvm *kvm; 45 struct kvm *kvm;
46 struct kvm_kpit_state pit_state; 46 struct kvm_kpit_state pit_state;
47 int irq_source_id;
47}; 48};
48 49
49#define KVM_PIT_BASE_ADDRESS 0x40 50#define KVM_PIT_BASE_ADDRESS 0x40
diff --git a/arch/x86/kvm/mmu.c b/arch/x86/kvm/mmu.c
index 99c239c5c0a..2a5e64881d9 100644
--- a/arch/x86/kvm/mmu.c
+++ b/arch/x86/kvm/mmu.c
@@ -2634,6 +2634,7 @@ static int kvm_pv_mmu_write(struct kvm_vcpu *vcpu,
2634static int kvm_pv_mmu_flush_tlb(struct kvm_vcpu *vcpu) 2634static int kvm_pv_mmu_flush_tlb(struct kvm_vcpu *vcpu)
2635{ 2635{
2636 kvm_x86_ops->tlb_flush(vcpu); 2636 kvm_x86_ops->tlb_flush(vcpu);
2637 set_bit(KVM_REQ_MMU_SYNC, &vcpu->requests);
2637 return 1; 2638 return 1;
2638} 2639}
2639 2640
diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c
index 4f0677d1eae..f1f8ff2f1fa 100644
--- a/arch/x86/kvm/x86.c
+++ b/arch/x86/kvm/x86.c
@@ -1742,7 +1742,8 @@ long kvm_arch_vm_ioctl(struct file *filp,
1742 goto out; 1742 goto out;
1743 if (irqchip_in_kernel(kvm)) { 1743 if (irqchip_in_kernel(kvm)) {
1744 mutex_lock(&kvm->lock); 1744 mutex_lock(&kvm->lock);
1745 kvm_set_irq(kvm, irq_event.irq, irq_event.level); 1745 kvm_set_irq(kvm, KVM_USERSPACE_IRQ_SOURCE_ID,
1746 irq_event.irq, irq_event.level);
1746 mutex_unlock(&kvm->lock); 1747 mutex_unlock(&kvm->lock);
1747 r = 0; 1748 r = 0;
1748 } 1749 }
@@ -4013,6 +4014,9 @@ struct kvm *kvm_arch_create_vm(void)
4013 INIT_LIST_HEAD(&kvm->arch.active_mmu_pages); 4014 INIT_LIST_HEAD(&kvm->arch.active_mmu_pages);
4014 INIT_LIST_HEAD(&kvm->arch.assigned_dev_head); 4015 INIT_LIST_HEAD(&kvm->arch.assigned_dev_head);
4015 4016
4017 /* Reserve bit 0 of irq_sources_bitmap for userspace irq source */
4018 set_bit(KVM_USERSPACE_IRQ_SOURCE_ID, &kvm->arch.irq_sources_bitmap);
4019
4016 return kvm; 4020 return kvm;
4017} 4021}
4018 4022
diff --git a/arch/x86/lguest/boot.c b/arch/x86/lguest/boot.c
index 48ee4f9435f..a5d8e1ace1c 100644
--- a/arch/x86/lguest/boot.c
+++ b/arch/x86/lguest/boot.c
@@ -367,10 +367,9 @@ static void lguest_cpuid(unsigned int *ax, unsigned int *bx,
367 * lazily after a task switch, and Linux uses that gratefully, but wouldn't a 367 * lazily after a task switch, and Linux uses that gratefully, but wouldn't a
368 * name like "FPUTRAP bit" be a little less cryptic? 368 * name like "FPUTRAP bit" be a little less cryptic?
369 * 369 *
370 * We store cr0 (and cr3) locally, because the Host never changes it. The 370 * We store cr0 locally because the Host never changes it. The Guest sometimes
371 * Guest sometimes wants to read it and we'd prefer not to bother the Host 371 * wants to read it and we'd prefer not to bother the Host unnecessarily. */
372 * unnecessarily. */ 372static unsigned long current_cr0;
373static unsigned long current_cr0, current_cr3;
374static void lguest_write_cr0(unsigned long val) 373static void lguest_write_cr0(unsigned long val)
375{ 374{
376 lazy_hcall(LHCALL_TS, val & X86_CR0_TS, 0, 0); 375 lazy_hcall(LHCALL_TS, val & X86_CR0_TS, 0, 0);
@@ -399,17 +398,23 @@ static unsigned long lguest_read_cr2(void)
399 return lguest_data.cr2; 398 return lguest_data.cr2;
400} 399}
401 400
401/* See lguest_set_pte() below. */
402static bool cr3_changed = false;
403
402/* cr3 is the current toplevel pagetable page: the principle is the same as 404/* cr3 is the current toplevel pagetable page: the principle is the same as
403 * cr0. Keep a local copy, and tell the Host when it changes. */ 405 * cr0. Keep a local copy, and tell the Host when it changes. The only
406 * difference is that our local copy is in lguest_data because the Host needs
407 * to set it upon our initial hypercall. */
404static void lguest_write_cr3(unsigned long cr3) 408static void lguest_write_cr3(unsigned long cr3)
405{ 409{
410 lguest_data.pgdir = cr3;
406 lazy_hcall(LHCALL_NEW_PGTABLE, cr3, 0, 0); 411 lazy_hcall(LHCALL_NEW_PGTABLE, cr3, 0, 0);
407 current_cr3 = cr3; 412 cr3_changed = true;
408} 413}
409 414
410static unsigned long lguest_read_cr3(void) 415static unsigned long lguest_read_cr3(void)
411{ 416{
412 return current_cr3; 417 return lguest_data.pgdir;
413} 418}
414 419
415/* cr4 is used to enable and disable PGE, but we don't care. */ 420/* cr4 is used to enable and disable PGE, but we don't care. */
@@ -498,13 +503,13 @@ static void lguest_set_pmd(pmd_t *pmdp, pmd_t pmdval)
498 * to forget all of them. Fortunately, this is very rare. 503 * to forget all of them. Fortunately, this is very rare.
499 * 504 *
500 * ... except in early boot when the kernel sets up the initial pagetables, 505 * ... except in early boot when the kernel sets up the initial pagetables,
501 * which makes booting astonishingly slow. So we don't even tell the Host 506 * which makes booting astonishingly slow: 1.83 seconds! So we don't even tell
502 * anything changed until we've done the first page table switch. */ 507 * the Host anything changed until we've done the first page table switch,
508 * which brings boot back to 0.25 seconds. */
503static void lguest_set_pte(pte_t *ptep, pte_t pteval) 509static void lguest_set_pte(pte_t *ptep, pte_t pteval)
504{ 510{
505 *ptep = pteval; 511 *ptep = pteval;
506 /* Don't bother with hypercall before initial setup. */ 512 if (cr3_changed)
507 if (current_cr3)
508 lazy_hcall(LHCALL_FLUSH_TLB, 1, 0, 0); 513 lazy_hcall(LHCALL_FLUSH_TLB, 1, 0, 0);
509} 514}
510 515
@@ -521,7 +526,7 @@ static void lguest_set_pte(pte_t *ptep, pte_t pteval)
521static void lguest_flush_tlb_single(unsigned long addr) 526static void lguest_flush_tlb_single(unsigned long addr)
522{ 527{
523 /* Simply set it to zero: if it was not, it will fault back in. */ 528 /* Simply set it to zero: if it was not, it will fault back in. */
524 lazy_hcall(LHCALL_SET_PTE, current_cr3, addr, 0); 529 lazy_hcall(LHCALL_SET_PTE, lguest_data.pgdir, addr, 0);
525} 530}
526 531
527/* This is what happens after the Guest has removed a large number of entries. 532/* This is what happens after the Guest has removed a large number of entries.
@@ -581,6 +586,9 @@ static void __init lguest_init_IRQ(void)
581 586
582 for (i = 0; i < LGUEST_IRQS; i++) { 587 for (i = 0; i < LGUEST_IRQS; i++) {
583 int vector = FIRST_EXTERNAL_VECTOR + i; 588 int vector = FIRST_EXTERNAL_VECTOR + i;
589 /* Some systems map "vectors" to interrupts weirdly. Lguest has
590 * a straightforward 1 to 1 mapping, so force that here. */
591 __get_cpu_var(vector_irq)[vector] = i;
584 if (vector != SYSCALL_VECTOR) { 592 if (vector != SYSCALL_VECTOR) {
585 set_intr_gate(vector, interrupt[vector]); 593 set_intr_gate(vector, interrupt[vector]);
586 set_irq_chip_and_handler_name(i, &lguest_irq_controller, 594 set_irq_chip_and_handler_name(i, &lguest_irq_controller,
diff --git a/arch/x86/mach-voyager/setup.c b/arch/x86/mach-voyager/setup.c
index 6bbdd633864..a580b9562e7 100644
--- a/arch/x86/mach-voyager/setup.c
+++ b/arch/x86/mach-voyager/setup.c
@@ -27,7 +27,7 @@ static struct irqaction irq2 = {
27void __init intr_init_hook(void) 27void __init intr_init_hook(void)
28{ 28{
29#ifdef CONFIG_SMP 29#ifdef CONFIG_SMP
30 smp_intr_init(); 30 voyager_smp_intr_init();
31#endif 31#endif
32 32
33 setup_irq(2, &irq2); 33 setup_irq(2, &irq2);
diff --git a/arch/x86/mach-voyager/voyager_smp.c b/arch/x86/mach-voyager/voyager_smp.c
index 0f6e8a6523a..0e331652681 100644
--- a/arch/x86/mach-voyager/voyager_smp.c
+++ b/arch/x86/mach-voyager/voyager_smp.c
@@ -90,6 +90,7 @@ static void ack_vic_irq(unsigned int irq);
90static void vic_enable_cpi(void); 90static void vic_enable_cpi(void);
91static void do_boot_cpu(__u8 cpuid); 91static void do_boot_cpu(__u8 cpuid);
92static void do_quad_bootstrap(void); 92static void do_quad_bootstrap(void);
93static void initialize_secondary(void);
93 94
94int hard_smp_processor_id(void); 95int hard_smp_processor_id(void);
95int safe_smp_processor_id(void); 96int safe_smp_processor_id(void);
@@ -344,6 +345,12 @@ static void do_quad_bootstrap(void)
344 } 345 }
345} 346}
346 347
348void prefill_possible_map(void)
349{
350 /* This is empty on voyager because we need a much
351 * earlier detection which is done in find_smp_config */
352}
353
347/* Set up all the basic stuff: read the SMP config and make all the 354/* Set up all the basic stuff: read the SMP config and make all the
348 * SMP information reflect only the boot cpu. All others will be 355 * SMP information reflect only the boot cpu. All others will be
349 * brought on-line later. */ 356 * brought on-line later. */
@@ -413,6 +420,7 @@ void __init smp_store_cpu_info(int id)
413 struct cpuinfo_x86 *c = &cpu_data(id); 420 struct cpuinfo_x86 *c = &cpu_data(id);
414 421
415 *c = boot_cpu_data; 422 *c = boot_cpu_data;
423 c->cpu_index = id;
416 424
417 identify_secondary_cpu(c); 425 identify_secondary_cpu(c);
418} 426}
@@ -650,6 +658,8 @@ void __init smp_boot_cpus(void)
650 smp_tune_scheduling(); 658 smp_tune_scheduling();
651 */ 659 */
652 smp_store_cpu_info(boot_cpu_id); 660 smp_store_cpu_info(boot_cpu_id);
661 /* setup the jump vector */
662 initial_code = (unsigned long)initialize_secondary;
653 printk("CPU%d: ", boot_cpu_id); 663 printk("CPU%d: ", boot_cpu_id);
654 print_cpu_info(&cpu_data(boot_cpu_id)); 664 print_cpu_info(&cpu_data(boot_cpu_id));
655 665
@@ -702,7 +712,7 @@ void __init smp_boot_cpus(void)
702 712
703/* Reload the secondary CPUs task structure (this function does not 713/* Reload the secondary CPUs task structure (this function does not
704 * return ) */ 714 * return ) */
705void __init initialize_secondary(void) 715static void __init initialize_secondary(void)
706{ 716{
707#if 0 717#if 0
708 // AC kernels only 718 // AC kernels only
@@ -1248,7 +1258,7 @@ static void handle_vic_irq(unsigned int irq, struct irq_desc *desc)
1248#define QIC_SET_GATE(cpi, vector) \ 1258#define QIC_SET_GATE(cpi, vector) \
1249 set_intr_gate((cpi) + QIC_DEFAULT_CPI_BASE, (vector)) 1259 set_intr_gate((cpi) + QIC_DEFAULT_CPI_BASE, (vector))
1250 1260
1251void __init smp_intr_init(void) 1261void __init voyager_smp_intr_init(void)
1252{ 1262{
1253 int i; 1263 int i;
1254 1264
diff --git a/arch/x86/mm/Makefile b/arch/x86/mm/Makefile
index 59f89b434b4..fea4565ff57 100644
--- a/arch/x86/mm/Makefile
+++ b/arch/x86/mm/Makefile
@@ -1,7 +1,7 @@
1obj-y := init_$(BITS).o fault.o ioremap.o extable.o pageattr.o mmap.o \ 1obj-y := init_$(BITS).o fault.o ioremap.o extable.o pageattr.o mmap.o \
2 pat.o pgtable.o gup.o 2 pat.o pgtable.o gup.o
3 3
4obj-$(CONFIG_X86_32) += pgtable_32.o 4obj-$(CONFIG_X86_32) += pgtable_32.o iomap_32.o
5 5
6obj-$(CONFIG_HUGETLB_PAGE) += hugetlbpage.o 6obj-$(CONFIG_HUGETLB_PAGE) += hugetlbpage.o
7obj-$(CONFIG_X86_PTDUMP) += dump_pagetables.o 7obj-$(CONFIG_X86_PTDUMP) += dump_pagetables.o
diff --git a/arch/x86/mm/gup.c b/arch/x86/mm/gup.c
index 4ba373c5b8c..be54176e9eb 100644
--- a/arch/x86/mm/gup.c
+++ b/arch/x86/mm/gup.c
@@ -233,7 +233,7 @@ int get_user_pages_fast(unsigned long start, int nr_pages, int write,
233 len = (unsigned long) nr_pages << PAGE_SHIFT; 233 len = (unsigned long) nr_pages << PAGE_SHIFT;
234 end = start + len; 234 end = start + len;
235 if (unlikely(!access_ok(write ? VERIFY_WRITE : VERIFY_READ, 235 if (unlikely(!access_ok(write ? VERIFY_WRITE : VERIFY_READ,
236 start, len))) 236 (void __user *)start, len)))
237 goto slow_irqon; 237 goto slow_irqon;
238 238
239 /* 239 /*
diff --git a/arch/x86/mm/init_32.c b/arch/x86/mm/init_32.c
index 8396868e82c..c483f424207 100644
--- a/arch/x86/mm/init_32.c
+++ b/arch/x86/mm/init_32.c
@@ -334,7 +334,6 @@ int devmem_is_allowed(unsigned long pagenr)
334 return 0; 334 return 0;
335} 335}
336 336
337#ifdef CONFIG_HIGHMEM
338pte_t *kmap_pte; 337pte_t *kmap_pte;
339pgprot_t kmap_prot; 338pgprot_t kmap_prot;
340 339
@@ -357,6 +356,7 @@ static void __init kmap_init(void)
357 kmap_prot = PAGE_KERNEL; 356 kmap_prot = PAGE_KERNEL;
358} 357}
359 358
359#ifdef CONFIG_HIGHMEM
360static void __init permanent_kmaps_init(pgd_t *pgd_base) 360static void __init permanent_kmaps_init(pgd_t *pgd_base)
361{ 361{
362 unsigned long vaddr; 362 unsigned long vaddr;
@@ -436,7 +436,6 @@ static void __init set_highmem_pages_init(void)
436#endif /* !CONFIG_NUMA */ 436#endif /* !CONFIG_NUMA */
437 437
438#else 438#else
439# define kmap_init() do { } while (0)
440# define permanent_kmaps_init(pgd_base) do { } while (0) 439# define permanent_kmaps_init(pgd_base) do { } while (0)
441# define set_highmem_pages_init() do { } while (0) 440# define set_highmem_pages_init() do { } while (0)
442#endif /* CONFIG_HIGHMEM */ 441#endif /* CONFIG_HIGHMEM */
diff --git a/arch/x86/mm/init_64.c b/arch/x86/mm/init_64.c
index b8e461d4941..9db01db6e3c 100644
--- a/arch/x86/mm/init_64.c
+++ b/arch/x86/mm/init_64.c
@@ -350,8 +350,10 @@ phys_pte_init(pte_t *pte_page, unsigned long addr, unsigned long end,
350 * pagetable pages as RO. So assume someone who pre-setup 350 * pagetable pages as RO. So assume someone who pre-setup
351 * these mappings are more intelligent. 351 * these mappings are more intelligent.
352 */ 352 */
353 if (pte_val(*pte)) 353 if (pte_val(*pte)) {
354 pages++;
354 continue; 355 continue;
356 }
355 357
356 if (0) 358 if (0)
357 printk(" pte=%p addr=%lx pte=%016lx\n", 359 printk(" pte=%p addr=%lx pte=%016lx\n",
@@ -418,8 +420,10 @@ phys_pmd_init(pmd_t *pmd_page, unsigned long address, unsigned long end,
418 * not differ with respect to page frame and 420 * not differ with respect to page frame and
419 * attributes. 421 * attributes.
420 */ 422 */
421 if (page_size_mask & (1 << PG_LEVEL_2M)) 423 if (page_size_mask & (1 << PG_LEVEL_2M)) {
424 pages++;
422 continue; 425 continue;
426 }
423 new_prot = pte_pgprot(pte_clrhuge(*(pte_t *)pmd)); 427 new_prot = pte_pgprot(pte_clrhuge(*(pte_t *)pmd));
424 } 428 }
425 429
@@ -499,8 +503,10 @@ phys_pud_init(pud_t *pud_page, unsigned long addr, unsigned long end,
499 * not differ with respect to page frame and 503 * not differ with respect to page frame and
500 * attributes. 504 * attributes.
501 */ 505 */
502 if (page_size_mask & (1 << PG_LEVEL_1G)) 506 if (page_size_mask & (1 << PG_LEVEL_1G)) {
507 pages++;
503 continue; 508 continue;
509 }
504 prot = pte_pgprot(pte_clrhuge(*(pte_t *)pud)); 510 prot = pte_pgprot(pte_clrhuge(*(pte_t *)pud));
505 } 511 }
506 512
@@ -665,12 +671,13 @@ unsigned long __init_refok init_memory_mapping(unsigned long start,
665 unsigned long last_map_addr = 0; 671 unsigned long last_map_addr = 0;
666 unsigned long page_size_mask = 0; 672 unsigned long page_size_mask = 0;
667 unsigned long start_pfn, end_pfn; 673 unsigned long start_pfn, end_pfn;
674 unsigned long pos;
668 675
669 struct map_range mr[NR_RANGE_MR]; 676 struct map_range mr[NR_RANGE_MR];
670 int nr_range, i; 677 int nr_range, i;
671 int use_pse, use_gbpages; 678 int use_pse, use_gbpages;
672 679
673 printk(KERN_INFO "init_memory_mapping\n"); 680 printk(KERN_INFO "init_memory_mapping: %016lx-%016lx\n", start, end);
674 681
675 /* 682 /*
676 * Find space for the kernel direct mapping tables. 683 * Find space for the kernel direct mapping tables.
@@ -704,35 +711,50 @@ unsigned long __init_refok init_memory_mapping(unsigned long start,
704 711
705 /* head if not big page alignment ?*/ 712 /* head if not big page alignment ?*/
706 start_pfn = start >> PAGE_SHIFT; 713 start_pfn = start >> PAGE_SHIFT;
707 end_pfn = ((start + (PMD_SIZE - 1)) >> PMD_SHIFT) 714 pos = start_pfn << PAGE_SHIFT;
715 end_pfn = ((pos + (PMD_SIZE - 1)) >> PMD_SHIFT)
708 << (PMD_SHIFT - PAGE_SHIFT); 716 << (PMD_SHIFT - PAGE_SHIFT);
709 nr_range = save_mr(mr, nr_range, start_pfn, end_pfn, 0); 717 if (start_pfn < end_pfn) {
718 nr_range = save_mr(mr, nr_range, start_pfn, end_pfn, 0);
719 pos = end_pfn << PAGE_SHIFT;
720 }
710 721
711 /* big page (2M) range*/ 722 /* big page (2M) range*/
712 start_pfn = ((start + (PMD_SIZE - 1))>>PMD_SHIFT) 723 start_pfn = ((pos + (PMD_SIZE - 1))>>PMD_SHIFT)
713 << (PMD_SHIFT - PAGE_SHIFT); 724 << (PMD_SHIFT - PAGE_SHIFT);
714 end_pfn = ((start + (PUD_SIZE - 1))>>PUD_SHIFT) 725 end_pfn = ((pos + (PUD_SIZE - 1))>>PUD_SHIFT)
715 << (PUD_SHIFT - PAGE_SHIFT); 726 << (PUD_SHIFT - PAGE_SHIFT);
716 if (end_pfn > ((end>>PUD_SHIFT)<<(PUD_SHIFT - PAGE_SHIFT))) 727 if (end_pfn > ((end>>PMD_SHIFT)<<(PMD_SHIFT - PAGE_SHIFT)))
717 end_pfn = ((end>>PUD_SHIFT)<<(PUD_SHIFT - PAGE_SHIFT)); 728 end_pfn = ((end>>PMD_SHIFT)<<(PMD_SHIFT - PAGE_SHIFT));
718 nr_range = save_mr(mr, nr_range, start_pfn, end_pfn, 729 if (start_pfn < end_pfn) {
719 page_size_mask & (1<<PG_LEVEL_2M)); 730 nr_range = save_mr(mr, nr_range, start_pfn, end_pfn,
731 page_size_mask & (1<<PG_LEVEL_2M));
732 pos = end_pfn << PAGE_SHIFT;
733 }
720 734
721 /* big page (1G) range */ 735 /* big page (1G) range */
722 start_pfn = end_pfn; 736 start_pfn = ((pos + (PUD_SIZE - 1))>>PUD_SHIFT)
723 end_pfn = (end>>PUD_SHIFT) << (PUD_SHIFT - PAGE_SHIFT); 737 << (PUD_SHIFT - PAGE_SHIFT);
724 nr_range = save_mr(mr, nr_range, start_pfn, end_pfn, 738 end_pfn = (end >> PUD_SHIFT) << (PUD_SHIFT - PAGE_SHIFT);
739 if (start_pfn < end_pfn) {
740 nr_range = save_mr(mr, nr_range, start_pfn, end_pfn,
725 page_size_mask & 741 page_size_mask &
726 ((1<<PG_LEVEL_2M)|(1<<PG_LEVEL_1G))); 742 ((1<<PG_LEVEL_2M)|(1<<PG_LEVEL_1G)));
743 pos = end_pfn << PAGE_SHIFT;
744 }
727 745
728 /* tail is not big page (1G) alignment */ 746 /* tail is not big page (1G) alignment */
729 start_pfn = end_pfn; 747 start_pfn = ((pos + (PMD_SIZE - 1))>>PMD_SHIFT)
730 end_pfn = (end>>PMD_SHIFT) << (PMD_SHIFT - PAGE_SHIFT); 748 << (PMD_SHIFT - PAGE_SHIFT);
731 nr_range = save_mr(mr, nr_range, start_pfn, end_pfn, 749 end_pfn = (end >> PMD_SHIFT) << (PMD_SHIFT - PAGE_SHIFT);
732 page_size_mask & (1<<PG_LEVEL_2M)); 750 if (start_pfn < end_pfn) {
751 nr_range = save_mr(mr, nr_range, start_pfn, end_pfn,
752 page_size_mask & (1<<PG_LEVEL_2M));
753 pos = end_pfn << PAGE_SHIFT;
754 }
733 755
734 /* tail is not big page (2M) alignment */ 756 /* tail is not big page (2M) alignment */
735 start_pfn = end_pfn; 757 start_pfn = pos>>PAGE_SHIFT;
736 end_pfn = end>>PAGE_SHIFT; 758 end_pfn = end>>PAGE_SHIFT;
737 nr_range = save_mr(mr, nr_range, start_pfn, end_pfn, 0); 759 nr_range = save_mr(mr, nr_range, start_pfn, end_pfn, 0);
738 760
@@ -831,12 +853,12 @@ int arch_add_memory(int nid, u64 start, u64 size)
831 unsigned long nr_pages = size >> PAGE_SHIFT; 853 unsigned long nr_pages = size >> PAGE_SHIFT;
832 int ret; 854 int ret;
833 855
834 last_mapped_pfn = init_memory_mapping(start, start + size-1); 856 last_mapped_pfn = init_memory_mapping(start, start + size);
835 if (last_mapped_pfn > max_pfn_mapped) 857 if (last_mapped_pfn > max_pfn_mapped)
836 max_pfn_mapped = last_mapped_pfn; 858 max_pfn_mapped = last_mapped_pfn;
837 859
838 ret = __add_pages(zone, start_pfn, nr_pages); 860 ret = __add_pages(zone, start_pfn, nr_pages);
839 WARN_ON(1); 861 WARN_ON_ONCE(ret);
840 862
841 return ret; 863 return ret;
842} 864}
@@ -878,6 +900,7 @@ static struct kcore_list kcore_mem, kcore_vmalloc, kcore_kernel,
878void __init mem_init(void) 900void __init mem_init(void)
879{ 901{
880 long codesize, reservedpages, datasize, initsize; 902 long codesize, reservedpages, datasize, initsize;
903 unsigned long absent_pages;
881 904
882 start_periodic_check_for_corruption(); 905 start_periodic_check_for_corruption();
883 906
@@ -893,8 +916,9 @@ void __init mem_init(void)
893#else 916#else
894 totalram_pages = free_all_bootmem(); 917 totalram_pages = free_all_bootmem();
895#endif 918#endif
896 reservedpages = max_pfn - totalram_pages - 919
897 absent_pages_in_range(0, max_pfn); 920 absent_pages = absent_pages_in_range(0, max_pfn);
921 reservedpages = max_pfn - totalram_pages - absent_pages;
898 after_bootmem = 1; 922 after_bootmem = 1;
899 923
900 codesize = (unsigned long) &_etext - (unsigned long) &_text; 924 codesize = (unsigned long) &_etext - (unsigned long) &_text;
@@ -911,10 +935,11 @@ void __init mem_init(void)
911 VSYSCALL_END - VSYSCALL_START); 935 VSYSCALL_END - VSYSCALL_START);
912 936
913 printk(KERN_INFO "Memory: %luk/%luk available (%ldk kernel code, " 937 printk(KERN_INFO "Memory: %luk/%luk available (%ldk kernel code, "
914 "%ldk reserved, %ldk data, %ldk init)\n", 938 "%ldk absent, %ldk reserved, %ldk data, %ldk init)\n",
915 (unsigned long) nr_free_pages() << (PAGE_SHIFT-10), 939 (unsigned long) nr_free_pages() << (PAGE_SHIFT-10),
916 max_pfn << (PAGE_SHIFT-10), 940 max_pfn << (PAGE_SHIFT-10),
917 codesize >> 10, 941 codesize >> 10,
942 absent_pages << (PAGE_SHIFT-10),
918 reservedpages << (PAGE_SHIFT-10), 943 reservedpages << (PAGE_SHIFT-10),
919 datasize >> 10, 944 datasize >> 10,
920 initsize >> 10); 945 initsize >> 10);
diff --git a/arch/x86/mm/iomap_32.c b/arch/x86/mm/iomap_32.c
new file mode 100644
index 00000000000..d0151d8ce45
--- /dev/null
+++ b/arch/x86/mm/iomap_32.c
@@ -0,0 +1,59 @@
1/*
2 * Copyright © 2008 Ingo Molnar
3 *
4 * This program is free software; you can redistribute it and/or modify
5 * it under the terms of the GNU General Public License as published by
6 * the Free Software Foundation; either version 2 of the License, or
7 * (at your option) any later version.
8 *
9 * This program is distributed in the hope that it will be useful, but
10 * WITHOUT ANY WARRANTY; without even the implied warranty of
11 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
12 * General Public License for more details.
13 *
14 * You should have received a copy of the GNU General Public License along
15 * with this program; if not, write to the Free Software Foundation, Inc.,
16 * 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA.
17 */
18
19#include <asm/iomap.h>
20#include <linux/module.h>
21
22/* Map 'pfn' using fixed map 'type' and protections 'prot'
23 */
24void *
25iomap_atomic_prot_pfn(unsigned long pfn, enum km_type type, pgprot_t prot)
26{
27 enum fixed_addresses idx;
28 unsigned long vaddr;
29
30 pagefault_disable();
31
32 idx = type + KM_TYPE_NR*smp_processor_id();
33 vaddr = __fix_to_virt(FIX_KMAP_BEGIN + idx);
34 set_pte(kmap_pte-idx, pfn_pte(pfn, prot));
35 arch_flush_lazy_mmu_mode();
36
37 return (void*) vaddr;
38}
39EXPORT_SYMBOL_GPL(iomap_atomic_prot_pfn);
40
41void
42iounmap_atomic(void *kvaddr, enum km_type type)
43{
44 unsigned long vaddr = (unsigned long) kvaddr & PAGE_MASK;
45 enum fixed_addresses idx = type + KM_TYPE_NR*smp_processor_id();
46
47 /*
48 * Force other mappings to Oops if they'll try to access this pte
49 * without first remap it. Keeping stale mappings around is a bad idea
50 * also, in case the page changes cacheability attributes or becomes
51 * a protected page in a hypervisor.
52 */
53 if (vaddr == __fix_to_virt(FIX_KMAP_BEGIN+idx))
54 kpte_clear_flush(kmap_pte-idx, vaddr);
55
56 arch_flush_lazy_mmu_mode();
57 pagefault_enable();
58}
59EXPORT_SYMBOL_GPL(iounmap_atomic);
diff --git a/arch/x86/mm/ioremap.c b/arch/x86/mm/ioremap.c
index ae71e11eb3e..d4c4307ff3e 100644
--- a/arch/x86/mm/ioremap.c
+++ b/arch/x86/mm/ioremap.c
@@ -387,7 +387,7 @@ static void __iomem *ioremap_default(resource_size_t phys_addr,
387 unsigned long size) 387 unsigned long size)
388{ 388{
389 unsigned long flags; 389 unsigned long flags;
390 void *ret; 390 void __iomem *ret;
391 int err; 391 int err;
392 392
393 /* 393 /*
@@ -399,11 +399,11 @@ static void __iomem *ioremap_default(resource_size_t phys_addr,
399 if (err < 0) 399 if (err < 0)
400 return NULL; 400 return NULL;
401 401
402 ret = (void *) __ioremap_caller(phys_addr, size, flags, 402 ret = __ioremap_caller(phys_addr, size, flags,
403 __builtin_return_address(0)); 403 __builtin_return_address(0));
404 404
405 free_memtype(phys_addr, phys_addr + size); 405 free_memtype(phys_addr, phys_addr + size);
406 return (void __iomem *)ret; 406 return ret;
407} 407}
408 408
409void __iomem *ioremap_prot(resource_size_t phys_addr, unsigned long size, 409void __iomem *ioremap_prot(resource_size_t phys_addr, unsigned long size,
@@ -622,7 +622,7 @@ static inline void __init early_clear_fixmap(enum fixed_addresses idx)
622 __early_set_fixmap(idx, 0, __pgprot(0)); 622 __early_set_fixmap(idx, 0, __pgprot(0));
623} 623}
624 624
625static void *prev_map[FIX_BTMAPS_SLOTS] __initdata; 625static void __iomem *prev_map[FIX_BTMAPS_SLOTS] __initdata;
626static unsigned long prev_size[FIX_BTMAPS_SLOTS] __initdata; 626static unsigned long prev_size[FIX_BTMAPS_SLOTS] __initdata;
627static int __init check_early_ioremap_leak(void) 627static int __init check_early_ioremap_leak(void)
628{ 628{
@@ -645,7 +645,7 @@ static int __init check_early_ioremap_leak(void)
645} 645}
646late_initcall(check_early_ioremap_leak); 646late_initcall(check_early_ioremap_leak);
647 647
648static void __init *__early_ioremap(unsigned long phys_addr, unsigned long size, pgprot_t prot) 648static void __init __iomem *__early_ioremap(unsigned long phys_addr, unsigned long size, pgprot_t prot)
649{ 649{
650 unsigned long offset, last_addr; 650 unsigned long offset, last_addr;
651 unsigned int nrpages; 651 unsigned int nrpages;
@@ -713,23 +713,23 @@ static void __init *__early_ioremap(unsigned long phys_addr, unsigned long size,
713 if (early_ioremap_debug) 713 if (early_ioremap_debug)
714 printk(KERN_CONT "%08lx + %08lx\n", offset, fix_to_virt(idx0)); 714 printk(KERN_CONT "%08lx + %08lx\n", offset, fix_to_virt(idx0));
715 715
716 prev_map[slot] = (void *) (offset + fix_to_virt(idx0)); 716 prev_map[slot] = (void __iomem *)(offset + fix_to_virt(idx0));
717 return prev_map[slot]; 717 return prev_map[slot];
718} 718}
719 719
720/* Remap an IO device */ 720/* Remap an IO device */
721void __init *early_ioremap(unsigned long phys_addr, unsigned long size) 721void __init __iomem *early_ioremap(unsigned long phys_addr, unsigned long size)
722{ 722{
723 return __early_ioremap(phys_addr, size, PAGE_KERNEL_IO); 723 return __early_ioremap(phys_addr, size, PAGE_KERNEL_IO);
724} 724}
725 725
726/* Remap memory */ 726/* Remap memory */
727void __init *early_memremap(unsigned long phys_addr, unsigned long size) 727void __init __iomem *early_memremap(unsigned long phys_addr, unsigned long size)
728{ 728{
729 return __early_ioremap(phys_addr, size, PAGE_KERNEL); 729 return __early_ioremap(phys_addr, size, PAGE_KERNEL);
730} 730}
731 731
732void __init early_iounmap(void *addr, unsigned long size) 732void __init early_iounmap(void __iomem *addr, unsigned long size)
733{ 733{
734 unsigned long virt_addr; 734 unsigned long virt_addr;
735 unsigned long offset; 735 unsigned long offset;
@@ -779,7 +779,7 @@ void __init early_iounmap(void *addr, unsigned long size)
779 --idx; 779 --idx;
780 --nrpages; 780 --nrpages;
781 } 781 }
782 prev_map[slot] = 0; 782 prev_map[slot] = NULL;
783} 783}
784 784
785void __this_fixmap_does_not_exist(void) 785void __this_fixmap_does_not_exist(void)
diff --git a/arch/x86/mm/pageattr.c b/arch/x86/mm/pageattr.c
index f1dc1b75d16..e89d24815f2 100644
--- a/arch/x86/mm/pageattr.c
+++ b/arch/x86/mm/pageattr.c
@@ -67,18 +67,18 @@ static void split_page_count(int level)
67 67
68void arch_report_meminfo(struct seq_file *m) 68void arch_report_meminfo(struct seq_file *m)
69{ 69{
70 seq_printf(m, "DirectMap4k: %8lu kB\n", 70 seq_printf(m, "DirectMap4k: %8lu kB\n",
71 direct_pages_count[PG_LEVEL_4K] << 2); 71 direct_pages_count[PG_LEVEL_4K] << 2);
72#if defined(CONFIG_X86_64) || defined(CONFIG_X86_PAE) 72#if defined(CONFIG_X86_64) || defined(CONFIG_X86_PAE)
73 seq_printf(m, "DirectMap2M: %8lu kB\n", 73 seq_printf(m, "DirectMap2M: %8lu kB\n",
74 direct_pages_count[PG_LEVEL_2M] << 11); 74 direct_pages_count[PG_LEVEL_2M] << 11);
75#else 75#else
76 seq_printf(m, "DirectMap4M: %8lu kB\n", 76 seq_printf(m, "DirectMap4M: %8lu kB\n",
77 direct_pages_count[PG_LEVEL_2M] << 12); 77 direct_pages_count[PG_LEVEL_2M] << 12);
78#endif 78#endif
79#ifdef CONFIG_X86_64 79#ifdef CONFIG_X86_64
80 if (direct_gbpages) 80 if (direct_gbpages)
81 seq_printf(m, "DirectMap1G: %8lu kB\n", 81 seq_printf(m, "DirectMap1G: %8lu kB\n",
82 direct_pages_count[PG_LEVEL_1G] << 20); 82 direct_pages_count[PG_LEVEL_1G] << 20);
83#endif 83#endif
84} 84}
diff --git a/arch/x86/mm/pat.c b/arch/x86/mm/pat.c
index 738fd0f2495..eb1bf000d12 100644
--- a/arch/x86/mm/pat.c
+++ b/arch/x86/mm/pat.c
@@ -481,12 +481,16 @@ static inline int range_is_allowed(unsigned long pfn, unsigned long size)
481 return 1; 481 return 1;
482} 482}
483#else 483#else
484/* This check is needed to avoid cache aliasing when PAT is enabled */
484static inline int range_is_allowed(unsigned long pfn, unsigned long size) 485static inline int range_is_allowed(unsigned long pfn, unsigned long size)
485{ 486{
486 u64 from = ((u64)pfn) << PAGE_SHIFT; 487 u64 from = ((u64)pfn) << PAGE_SHIFT;
487 u64 to = from + size; 488 u64 to = from + size;
488 u64 cursor = from; 489 u64 cursor = from;
489 490
491 if (!pat_enabled)
492 return 1;
493
490 while (cursor < to) { 494 while (cursor < to) {
491 if (!devmem_is_allowed(pfn)) { 495 if (!devmem_is_allowed(pfn)) {
492 printk(KERN_INFO 496 printk(KERN_INFO
diff --git a/arch/x86/oprofile/op_model_ppro.c b/arch/x86/oprofile/op_model_ppro.c
index 0620d6d45f7..3f1b81a83e2 100644
--- a/arch/x86/oprofile/op_model_ppro.c
+++ b/arch/x86/oprofile/op_model_ppro.c
@@ -27,8 +27,7 @@ static int num_counters = 2;
27static int counter_width = 32; 27static int counter_width = 32;
28 28
29#define CTR_IS_RESERVED(msrs, c) (msrs->counters[(c)].addr ? 1 : 0) 29#define CTR_IS_RESERVED(msrs, c) (msrs->counters[(c)].addr ? 1 : 0)
30#define CTR_READ(l, h, msrs, c) do {rdmsr(msrs->counters[(c)].addr, (l), (h)); } while (0) 30#define CTR_OVERFLOWED(n) (!((n) & (1ULL<<(counter_width-1))))
31#define CTR_OVERFLOWED(n) (!((n) & (1U<<(counter_width-1))))
32 31
33#define CTRL_IS_RESERVED(msrs, c) (msrs->controls[(c)].addr ? 1 : 0) 32#define CTRL_IS_RESERVED(msrs, c) (msrs->controls[(c)].addr ? 1 : 0)
34#define CTRL_READ(l, h, msrs, c) do {rdmsr((msrs->controls[(c)].addr), (l), (h)); } while (0) 33#define CTRL_READ(l, h, msrs, c) do {rdmsr((msrs->controls[(c)].addr), (l), (h)); } while (0)
@@ -124,14 +123,14 @@ static void ppro_setup_ctrs(struct op_msrs const * const msrs)
124static int ppro_check_ctrs(struct pt_regs * const regs, 123static int ppro_check_ctrs(struct pt_regs * const regs,
125 struct op_msrs const * const msrs) 124 struct op_msrs const * const msrs)
126{ 125{
127 unsigned int low, high; 126 u64 val;
128 int i; 127 int i;
129 128
130 for (i = 0 ; i < num_counters; ++i) { 129 for (i = 0 ; i < num_counters; ++i) {
131 if (!reset_value[i]) 130 if (!reset_value[i])
132 continue; 131 continue;
133 CTR_READ(low, high, msrs, i); 132 rdmsrl(msrs->counters[i].addr, val);
134 if (CTR_OVERFLOWED(low)) { 133 if (CTR_OVERFLOWED(val)) {
135 oprofile_add_sample(regs, i); 134 oprofile_add_sample(regs, i);
136 wrmsrl(msrs->counters[i].addr, -reset_value[i]); 135 wrmsrl(msrs->counters[i].addr, -reset_value[i]);
137 } 136 }
diff --git a/arch/x86/xen/Makefile b/arch/x86/xen/Makefile
index 313947940a1..6dcefba7836 100644
--- a/arch/x86/xen/Makefile
+++ b/arch/x86/xen/Makefile
@@ -1,4 +1,4 @@
1ifdef CONFIG_FTRACE 1ifdef CONFIG_FUNCTION_TRACER
2# Do not profile debug and lowlevel utilities 2# Do not profile debug and lowlevel utilities
3CFLAGS_REMOVE_spinlock.o = -pg 3CFLAGS_REMOVE_spinlock.o = -pg
4CFLAGS_REMOVE_time.o = -pg 4CFLAGS_REMOVE_time.o = -pg
diff --git a/arch/x86/xen/enlighten.c b/arch/x86/xen/enlighten.c
index b61534c7a4c..5e4686d70f6 100644
--- a/arch/x86/xen/enlighten.c
+++ b/arch/x86/xen/enlighten.c
@@ -863,15 +863,16 @@ static void xen_alloc_ptpage(struct mm_struct *mm, unsigned long pfn, unsigned l
863 if (PagePinned(virt_to_page(mm->pgd))) { 863 if (PagePinned(virt_to_page(mm->pgd))) {
864 SetPagePinned(page); 864 SetPagePinned(page);
865 865
866 vm_unmap_aliases();
866 if (!PageHighMem(page)) { 867 if (!PageHighMem(page)) {
867 make_lowmem_page_readonly(__va(PFN_PHYS((unsigned long)pfn))); 868 make_lowmem_page_readonly(__va(PFN_PHYS((unsigned long)pfn)));
868 if (level == PT_PTE && USE_SPLIT_PTLOCKS) 869 if (level == PT_PTE && USE_SPLIT_PTLOCKS)
869 pin_pagetable_pfn(MMUEXT_PIN_L1_TABLE, pfn); 870 pin_pagetable_pfn(MMUEXT_PIN_L1_TABLE, pfn);
870 } else 871 } else {
871 /* make sure there are no stray mappings of 872 /* make sure there are no stray mappings of
872 this page */ 873 this page */
873 kmap_flush_unused(); 874 kmap_flush_unused();
874 vm_unmap_aliases(); 875 }
875 } 876 }
876} 877}
877 878
diff --git a/arch/x86/xen/mmu.c b/arch/x86/xen/mmu.c
index d4d52f5a1cf..688936044dc 100644
--- a/arch/x86/xen/mmu.c
+++ b/arch/x86/xen/mmu.c
@@ -246,11 +246,21 @@ xmaddr_t arbitrary_virt_to_machine(void *vaddr)
246{ 246{
247 unsigned long address = (unsigned long)vaddr; 247 unsigned long address = (unsigned long)vaddr;
248 unsigned int level; 248 unsigned int level;
249 pte_t *pte = lookup_address(address, &level); 249 pte_t *pte;
250 unsigned offset = address & ~PAGE_MASK; 250 unsigned offset;
251 251
252 BUG_ON(pte == NULL); 252 /*
253 * if the PFN is in the linear mapped vaddr range, we can just use
254 * the (quick) virt_to_machine() p2m lookup
255 */
256 if (virt_addr_valid(vaddr))
257 return virt_to_machine(vaddr);
253 258
259 /* otherwise we have to do a (slower) full page-table walk */
260
261 pte = lookup_address(address, &level);
262 BUG_ON(pte == NULL);
263 offset = address & ~PAGE_MASK;
254 return XMADDR(((phys_addr_t)pte_mfn(*pte) << PAGE_SHIFT) + offset); 264 return XMADDR(((phys_addr_t)pte_mfn(*pte) << PAGE_SHIFT) + offset);
255} 265}
256 266
@@ -410,7 +420,7 @@ void xen_ptep_modify_prot_commit(struct mm_struct *mm, unsigned long addr,
410 420
411 xen_mc_batch(); 421 xen_mc_batch();
412 422
413 u.ptr = virt_to_machine(ptep).maddr | MMU_PT_UPDATE_PRESERVE_AD; 423 u.ptr = arbitrary_virt_to_machine(ptep).maddr | MMU_PT_UPDATE_PRESERVE_AD;
414 u.val = pte_val_ma(pte); 424 u.val = pte_val_ma(pte);
415 xen_extend_mmu_update(&u); 425 xen_extend_mmu_update(&u);
416 426
@@ -840,13 +850,16 @@ static int xen_pin_page(struct mm_struct *mm, struct page *page,
840 read-only, and can be pinned. */ 850 read-only, and can be pinned. */
841static void __xen_pgd_pin(struct mm_struct *mm, pgd_t *pgd) 851static void __xen_pgd_pin(struct mm_struct *mm, pgd_t *pgd)
842{ 852{
853 vm_unmap_aliases();
854
843 xen_mc_batch(); 855 xen_mc_batch();
844 856
845 if (xen_pgd_walk(mm, xen_pin_page, USER_LIMIT)) { 857 if (xen_pgd_walk(mm, xen_pin_page, USER_LIMIT)) {
846 /* re-enable interrupts for kmap_flush_unused */ 858 /* re-enable interrupts for flushing */
847 xen_mc_issue(0); 859 xen_mc_issue(0);
860
848 kmap_flush_unused(); 861 kmap_flush_unused();
849 vm_unmap_aliases(); 862
850 xen_mc_batch(); 863 xen_mc_batch();
851 } 864 }
852 865
@@ -864,7 +877,7 @@ static void __xen_pgd_pin(struct mm_struct *mm, pgd_t *pgd)
864#else /* CONFIG_X86_32 */ 877#else /* CONFIG_X86_32 */
865#ifdef CONFIG_X86_PAE 878#ifdef CONFIG_X86_PAE
866 /* Need to make sure unshared kernel PMD is pinnable */ 879 /* Need to make sure unshared kernel PMD is pinnable */
867 xen_pin_page(mm, virt_to_page(pgd_page(pgd[pgd_index(TASK_SIZE)])), 880 xen_pin_page(mm, pgd_page(pgd[pgd_index(TASK_SIZE)]),
868 PT_PMD); 881 PT_PMD);
869#endif 882#endif
870 xen_do_pin(MMUEXT_PIN_L3_TABLE, PFN_DOWN(__pa(pgd))); 883 xen_do_pin(MMUEXT_PIN_L3_TABLE, PFN_DOWN(__pa(pgd)));
@@ -981,7 +994,7 @@ static void __xen_pgd_unpin(struct mm_struct *mm, pgd_t *pgd)
981 994
982#ifdef CONFIG_X86_PAE 995#ifdef CONFIG_X86_PAE
983 /* Need to make sure unshared kernel PMD is unpinned */ 996 /* Need to make sure unshared kernel PMD is unpinned */
984 xen_unpin_page(mm, virt_to_page(pgd_page(pgd[pgd_index(TASK_SIZE)])), 997 xen_unpin_page(mm, pgd_page(pgd[pgd_index(TASK_SIZE)]),
985 PT_PMD); 998 PT_PMD);
986#endif 999#endif
987 1000
generated by cgit v1.2.2 (git 2.25.0) at 2026-01-08 18:37:41 -0500