From: www.kernel.org
Subject: Linux 2.6.21
Patch-mainline: 2.6.21
Automatically created from "patches.kernel.org/patch-2.6.21" by xen-port-patches.py
Acked-by: jbeulich@novell.com
--- head.orig/arch/x86/Kconfig 2012-04-10 16:14:33.000000000 +0200
+++ head/arch/x86/Kconfig 2012-02-08 11:31:09.000000000 +0100
@@ -101,9 +101,11 @@ config GENERIC_CMOS_UPDATE
config CLOCKSOURCE_WATCHDOG
def_bool y
+ depends on !X86_XEN
config GENERIC_CLOCKEVENTS
def_bool y
+ depends on !X86_XEN
config ARCH_CLOCKSOURCE_DATA
def_bool y
@@ -111,7 +113,7 @@ config ARCH_CLOCKSOURCE_DATA
config GENERIC_CLOCKEVENTS_BROADCAST
def_bool y
- depends on X86_64 || (X86_32 && X86_LOCAL_APIC)
+ depends on X86_64 || (X86_32 && X86_LOCAL_APIC && !X86_XEN)
config LOCKDEP_SUPPORT
def_bool y
--- head.orig/arch/x86/kernel/Makefile 2012-04-10 16:13:59.000000000 +0200
+++ head/arch/x86/kernel/Makefile 2012-04-10 16:14:45.000000000 +0200
@@ -120,7 +120,7 @@ ifeq ($(CONFIG_X86_64),y)
pci-dma_64-$(CONFIG_XEN) += pci-dma_32.o
endif
-disabled-obj-$(CONFIG_XEN) := early-quirks.o i8237.o i8253.o i8259_$(BITS).o \
- reboot.o smpboot_$(BITS).o tsc_$(BITS).o
+disabled-obj-$(CONFIG_XEN) := early-quirks.o hpet.o i8237.o i8253.o i8259_$(BITS).o \
+ reboot.o smpboot_$(BITS).o tsc_$(BITS).o tsc_sync.o
disabled-obj-$(CONFIG_XEN_UNPRIVILEGED_GUEST) += mpparse_64.o
%/head_$(BITS).o %/head_$(BITS).s: $(if $(CONFIG_XEN),EXTRA_AFLAGS,dummy) :=
--- head.orig/arch/x86/kernel/apic/apic-xen.c 2011-01-31 17:29:16.000000000 +0100
+++ head/arch/x86/kernel/apic/apic-xen.c 2011-01-31 17:32:22.000000000 +0100
@@ -25,6 +25,8 @@
#include <linux/kernel_stat.h>
#include <linux/sysdev.h>
#include <linux/cpu.h>
+#include <linux/clockchips.h>
+#include <linux/acpi_pmtmr.h>
#include <linux/module.h>
#include <asm/atomic.h>
@@ -56,83 +58,26 @@ static cpumask_t timer_bcast_ipi;
*/
/*
- * Debug level
+ * Debug level, exported for io_apic.c
*/
int apic_verbosity;
#ifndef CONFIG_XEN
static int modern_apic(void)
{
- unsigned int lvr, version;
/* AMD systems use old APIC versions, so check the CPU */
if (boot_cpu_data.x86_vendor == X86_VENDOR_AMD &&
- boot_cpu_data.x86 >= 0xf)
+ boot_cpu_data.x86 >= 0xf)
return 1;
- lvr = apic_read(APIC_LVR);
- version = GET_APIC_VERSION(lvr);
- return version >= 0x14;
+ return lapic_get_version() >= 0x14;
}
#endif /* !CONFIG_XEN */
-/*
- * 'what should we do if we get a hw irq event on an illegal vector'.
- * each architecture has to answer this themselves.
- */
-void ack_bad_irq(unsigned int irq)
-{
- printk("unexpected IRQ trap at vector %02x\n", irq);
- /*
- * Currently unexpected vectors happen only on SMP and APIC.
- * We _must_ ack these because every local APIC has only N
- * irq slots per priority level, and a 'hanging, unacked' IRQ
- * holds up an irq slot - in excessive cases (when multiple
- * unexpected vectors occur) that might lock up the APIC
- * completely.
- * But only ack when the APIC is enabled -AK
- */
- if (cpu_has_apic)
- ack_APIC_irq();
-}
-
int get_physical_broadcast(void)
{
return 0xff;
}
-#ifndef CONFIG_XEN
-#ifndef CONFIG_SMP
-static void up_apic_timer_interrupt_call(void)
-{
- int cpu = smp_processor_id();
-
- /*
- * the NMI deadlock-detector uses this.
- */
- per_cpu(irq_stat, cpu).apic_timer_irqs++;
-
- smp_local_timer_interrupt();
-}
-#endif
-
-void smp_send_timer_broadcast_ipi(void)
-{
- cpumask_t mask;
-
- cpus_and(mask, cpu_online_map, timer_bcast_ipi);
- if (!cpus_empty(mask)) {
-#ifdef CONFIG_SMP
- send_IPI_mask(mask, LOCAL_TIMER_VECTOR);
-#else
- /*
- * We can directly call the apic timer interrupt handler
- * in UP case. Minus all irq related functions
- */
- up_apic_timer_interrupt_call();
-#endif
- }
-}
-#endif
-
int setup_profiling_timer(unsigned int multiplier)
{
return -EINVAL;
--- head.orig/arch/x86/kernel/cpu/common-xen.c 2011-01-31 17:32:16.000000000 +0100
+++ head/arch/x86/kernel/cpu/common-xen.c 2011-01-31 17:32:22.000000000 +0100
@@ -612,7 +612,7 @@ void __init early_cpu_init(void)
struct pt_regs * __devinit idle_regs(struct pt_regs *regs)
{
memset(regs, 0, sizeof(struct pt_regs));
- regs->xgs = __KERNEL_PDA;
+ regs->xfs = __KERNEL_PDA;
return regs;
}
@@ -669,12 +669,12 @@ struct i386_pda boot_pda = {
.pcurrent = &init_task,
};
-static inline void set_kernel_gs(void)
+static inline void set_kernel_fs(void)
{
- /* Set %gs for this CPU's PDA. Memory clobber is to create a
+ /* Set %fs for this CPU's PDA. Memory clobber is to create a
barrier with respect to any PDA operations, so the compiler
doesn't move any before here. */
- asm volatile ("mov %0, %%gs" : : "r" (__KERNEL_PDA) : "memory");
+ asm volatile ("mov %0, %%fs" : : "r" (__KERNEL_PDA) : "memory");
}
/* Initialize the CPU's GDT and PDA. The boot CPU does this for
@@ -732,7 +732,7 @@ void __cpuinit cpu_set_gdt(int cpu)
}
BUG_ON(HYPERVISOR_set_gdt(frames, (cpu_gdt_descr->size + 1) / 8));
- set_kernel_gs();
+ set_kernel_fs();
}
/* Common CPU init for both boot and secondary CPUs */
@@ -777,8 +777,8 @@ static void __cpuinit _cpu_init(int cpu,
__set_tss_desc(cpu, GDT_ENTRY_DOUBLEFAULT_TSS, &doublefault_tss);
#endif
- /* Clear %fs. */
- asm volatile ("mov %0, %%fs" : : "r" (0));
+ /* Clear %gs. */
+ asm volatile ("mov %0, %%gs" : : "r" (0));
/* Clear all 6 debug registers: */
set_debugreg(0, 0);
--- head.orig/arch/x86/kernel/e820_32-xen.c 2011-01-31 17:32:16.000000000 +0100
+++ head/arch/x86/kernel/e820_32-xen.c 2011-01-31 17:32:22.000000000 +0100
@@ -14,6 +14,7 @@
#include <asm/pgtable.h>
#include <asm/page.h>
#include <asm/e820.h>
+#include <asm/setup.h>
#include <xen/interface/memory.h>
#ifdef CONFIG_EFI
@@ -157,21 +158,22 @@ static struct resource standard_io_resou
.flags = IORESOURCE_BUSY | IORESOURCE_IO
} };
-static int romsignature(const unsigned char *x)
+#define ROMSIGNATURE 0xaa55
+
+static int __init romsignature(const unsigned char *rom)
{
unsigned short sig;
- int ret = 0;
- if (probe_kernel_address((const unsigned short *)x, sig) == 0)
- ret = (sig == 0xaa55);
- return ret;
+
+ return probe_kernel_address((const unsigned short *)rom, sig) == 0 &&
+ sig == ROMSIGNATURE;
}
static int __init romchecksum(unsigned char *rom, unsigned long length)
{
- unsigned char *p, sum = 0;
+ unsigned char sum;
- for (p = rom; p < rom + length; p++)
- sum += *p;
+ for (sum = 0; length; length--)
+ sum += *rom++;
return sum == 0;
}
--- head.orig/arch/x86/kernel/entry_32-xen.S 2011-01-31 17:32:16.000000000 +0100
+++ head/arch/x86/kernel/entry_32-xen.S 2011-04-26 09:08:37.000000000 +0200
@@ -30,7 +30,7 @@
* 18(%esp) - %eax
* 1C(%esp) - %ds
* 20(%esp) - %es
- * 24(%esp) - %gs
+ * 24(%esp) - %fs
* 28(%esp) - orig_eax
* 2C(%esp) - %eip
* 30(%esp) - %cs
@@ -102,9 +102,9 @@ NMI_MASK = 0x80000000
#define SAVE_ALL \
cld; \
- pushl %gs; \
+ pushl %fs; \
CFI_ADJUST_CFA_OFFSET 4;\
- /*CFI_REL_OFFSET gs, 0;*/\
+ /*CFI_REL_OFFSET fs, 0;*/\
pushl %es; \
CFI_ADJUST_CFA_OFFSET 4;\
/*CFI_REL_OFFSET es, 0;*/\
@@ -136,7 +136,7 @@ NMI_MASK = 0x80000000
movl %edx, %ds; \
movl %edx, %es; \
movl $(__KERNEL_PDA), %edx; \
- movl %edx, %gs
+ movl %edx, %fs
#define RESTORE_INT_REGS \
popl %ebx; \
@@ -169,9 +169,9 @@ NMI_MASK = 0x80000000
2: popl %es; \
CFI_ADJUST_CFA_OFFSET -4;\
/*CFI_RESTORE es;*/\
-3: popl %gs; \
+3: popl %fs; \
CFI_ADJUST_CFA_OFFSET -4;\
- /*CFI_RESTORE gs;*/\
+ /*CFI_RESTORE fs;*/\
.pushsection .fixup,"ax"; \
4: movl $0,(%esp); \
jmp 1b; \
@@ -230,6 +230,7 @@ ENTRY(ret_from_fork)
CFI_ADJUST_CFA_OFFSET -4
jmp syscall_exit
CFI_ENDPROC
+END(ret_from_fork)
/*
* Return to user mode is not as complex as all this looks,
@@ -261,6 +262,7 @@ ENTRY(resume_userspace)
# int/exception return?
jne work_pending
jmp restore_all
+END(ret_from_exception)
#ifdef CONFIG_PREEMPT
ENTRY(resume_kernel)
@@ -275,6 +277,7 @@ need_resched:
jz restore_all
call preempt_schedule_irq
jmp need_resched
+END(resume_kernel)
#endif
CFI_ENDPROC
@@ -352,16 +355,17 @@ sysenter_past_esp:
movl PT_OLDESP(%esp), %ecx
xorl %ebp,%ebp
TRACE_IRQS_ON
-1: mov PT_GS(%esp), %gs
+1: mov PT_FS(%esp), %fs
ENABLE_INTERRUPTS_SYSEXIT
CFI_ENDPROC
.pushsection .fixup,"ax"
-2: movl $0,PT_GS(%esp)
+2: movl $0,PT_FS(%esp)
jmp 1b
.section __ex_table,"a"
.align 4
.long 1b,2b
.popsection
+ENDPROC(sysenter_entry)
# pv sysenter call handler stub
ENTRY(sysenter_entry_pv)
@@ -533,6 +537,7 @@ hypervisor_iret:
jmp hypercall_page + (__HYPERVISOR_iret * 32)
#endif
CFI_ENDPROC
+ENDPROC(system_call)
# perform work that needs to be done immediately before resumption
ALIGN
@@ -578,6 +583,7 @@ work_notifysig_v86:
xorl %edx, %edx
call do_notify_resume
jmp resume_userspace_sig
+END(work_pending)
# perform syscall exit tracing
ALIGN
@@ -593,6 +599,7 @@ syscall_trace_entry:
cmpl $(nr_syscalls), %eax
jnae syscall_call
jmp syscall_exit
+END(syscall_trace_entry)
# perform syscall exit tracing
ALIGN
@@ -606,6 +613,7 @@ syscall_exit_work:
movl $1, %edx
call do_syscall_trace
jmp resume_userspace
+END(syscall_exit_work)
CFI_ENDPROC
RING0_INT_FRAME # can't unwind into user space anyway
@@ -616,16 +624,18 @@ syscall_fault:
GET_THREAD_INFO(%ebp)
movl $-EFAULT,PT_EAX(%esp)
jmp resume_userspace
+END(syscall_fault)
syscall_badsys:
movl $-ENOSYS,PT_EAX(%esp)
jmp resume_userspace
+END(syscall_badsys)
CFI_ENDPROC
#ifndef CONFIG_XEN
#define FIXUP_ESPFIX_STACK \
/* since we are on a wrong stack, we cant make it a C code :( */ \
- movl %gs:PDA_cpu, %ebx; \
+ movl %fs:PDA_cpu, %ebx; \
PER_CPU(cpu_gdt_descr, %ebx); \
movl GDS_address(%ebx), %ebx; \
GET_DESC_BASE(GDT_ENTRY_ESPFIX_SS, %ebx, %eax, %ax, %al, %ah); \
@@ -656,9 +666,9 @@ syscall_badsys:
ENTRY(interrupt)
.text
-vector=0
ENTRY(irq_entries_start)
RING0_INT_FRAME
+vector=0
.rept NR_IRQS
ALIGN
.if vector
@@ -667,11 +677,16 @@ ENTRY(irq_entries_start)
1: pushl $~(vector)
CFI_ADJUST_CFA_OFFSET 4
jmp common_interrupt
-.data
+ .previous
.long 1b
-.text
+ .text
vector=vector+1
.endr
+END(irq_entries_start)
+
+.previous
+END(interrupt)
+.previous
/*
* the CPU automatically disables interrupts when executing an IRQ vector,
@@ -684,6 +699,7 @@ common_interrupt:
movl %esp,%eax
call do_IRQ
jmp ret_from_intr
+ENDPROC(common_interrupt)
CFI_ENDPROC
#define BUILD_INTERRUPT(name, nr) \
@@ -696,10 +712,16 @@ ENTRY(name) \
movl %esp,%eax; \
call smp_/**/name; \
jmp ret_from_intr; \
- CFI_ENDPROC
+ CFI_ENDPROC; \
+ENDPROC(name)
/* The include is where all of the SMP etc. interrupts come from */
#include "entry_arch.h"
+
+/* This alternate entry is needed because we hijack the apic LVTT */
+#if defined(CONFIG_VMI) && defined(CONFIG_X86_LOCAL_APIC)
+BUILD_INTERRUPT(apic_vmi_timer_interrupt,LOCAL_TIMER_VECTOR)
+#endif
#else
#define UNWIND_ESPFIX_STACK
#endif
@@ -710,7 +732,7 @@ KPROBE_ENTRY(page_fault)
CFI_ADJUST_CFA_OFFSET 4
ALIGN
error_code:
- /* the function address is in %gs's slot on the stack */
+ /* the function address is in %fs's slot on the stack */
pushl %es
CFI_ADJUST_CFA_OFFSET 4
/*CFI_REL_OFFSET es, 0*/
@@ -739,20 +761,20 @@ error_code:
CFI_ADJUST_CFA_OFFSET 4
CFI_REL_OFFSET ebx, 0
cld
- pushl %gs
+ pushl %fs
CFI_ADJUST_CFA_OFFSET 4
- /*CFI_REL_OFFSET gs, 0*/
+ /*CFI_REL_OFFSET fs, 0*/
movl $(__KERNEL_PDA), %ecx
- movl %ecx, %gs
+ movl %ecx, %fs
UNWIND_ESPFIX_STACK
popl %ecx
CFI_ADJUST_CFA_OFFSET -4
/*CFI_REGISTER es, ecx*/
- movl PT_GS(%esp), %edi # get the function address
+ movl PT_FS(%esp), %edi # get the function address
movl PT_ORIG_EAX(%esp), %edx # get the error code
movl $-1, PT_ORIG_EAX(%esp) # no syscall to restart
- mov %ecx, PT_GS(%esp)
- /*CFI_REL_OFFSET gs, ES*/
+ mov %ecx, PT_FS(%esp)
+ /*CFI_REL_OFFSET fs, ES*/
movl $(__USER_DS), %ecx
movl %ecx, %ds
movl %ecx, %es
@@ -844,7 +866,7 @@ critical_fixup_table:
.byte 6 # pop %eax
.byte 7 # pop %ds
.byte 8 # pop %es
- .byte 9,9 # pop %gs
+ .byte 9,9 # pop %fs
.byte 10,10,10 # add $4,%esp
.byte 11 # iret
.byte -1,-1,-1,-1 # movb $1,1(%esi) = __DISABLE_INTERRUPTS
@@ -909,6 +931,7 @@ ENTRY(coprocessor_error)
CFI_ADJUST_CFA_OFFSET 4
jmp error_code
CFI_ENDPROC
+END(coprocessor_error)
ENTRY(simd_coprocessor_error)
RING0_INT_FRAME
@@ -918,6 +941,7 @@ ENTRY(simd_coprocessor_error)
CFI_ADJUST_CFA_OFFSET 4
jmp error_code
CFI_ENDPROC
+END(simd_coprocessor_error)
ENTRY(device_not_available)
RING0_INT_FRAME
@@ -940,6 +964,7 @@ device_available_emulate:
call math_state_restore
jmp ret_from_exception
CFI_ENDPROC
+END(device_not_available)
#ifndef CONFIG_XEN
/*
@@ -1101,10 +1126,12 @@ ENTRY(native_iret)
.align 4
.long 1b,iret_exc
.previous
+END(native_iret)
ENTRY(native_irq_enable_sysexit)
sti
sysexit
+END(native_irq_enable_sysexit)
#endif
KPROBE_ENTRY(int3)
@@ -1127,6 +1154,7 @@ ENTRY(overflow)
CFI_ADJUST_CFA_OFFSET 4
jmp error_code
CFI_ENDPROC
+END(overflow)
ENTRY(bounds)
RING0_INT_FRAME
@@ -1136,6 +1164,7 @@ ENTRY(bounds)
CFI_ADJUST_CFA_OFFSET 4
jmp error_code
CFI_ENDPROC
+END(bounds)
ENTRY(invalid_op)
RING0_INT_FRAME
@@ -1145,6 +1174,7 @@ ENTRY(invalid_op)
CFI_ADJUST_CFA_OFFSET 4
jmp error_code
CFI_ENDPROC
+END(invalid_op)
ENTRY(coprocessor_segment_overrun)
RING0_INT_FRAME
@@ -1154,6 +1184,7 @@ ENTRY(coprocessor_segment_overrun)
CFI_ADJUST_CFA_OFFSET 4
jmp error_code
CFI_ENDPROC
+END(coprocessor_segment_overrun)
ENTRY(invalid_TSS)
RING0_EC_FRAME
@@ -1161,6 +1192,7 @@ ENTRY(invalid_TSS)
CFI_ADJUST_CFA_OFFSET 4
jmp error_code
CFI_ENDPROC
+END(invalid_TSS)
ENTRY(segment_not_present)
RING0_EC_FRAME
@@ -1168,6 +1200,7 @@ ENTRY(segment_not_present)
CFI_ADJUST_CFA_OFFSET 4
jmp error_code
CFI_ENDPROC
+END(segment_not_present)
ENTRY(stack_segment)
RING0_EC_FRAME
@@ -1175,6 +1208,7 @@ ENTRY(stack_segment)
CFI_ADJUST_CFA_OFFSET 4
jmp error_code
CFI_ENDPROC
+END(stack_segment)
KPROBE_ENTRY(general_protection)
RING0_EC_FRAME
@@ -1190,6 +1224,7 @@ ENTRY(alignment_check)
CFI_ADJUST_CFA_OFFSET 4
jmp error_code
CFI_ENDPROC
+END(alignment_check)
ENTRY(divide_error)
RING0_INT_FRAME
@@ -1199,6 +1234,7 @@ ENTRY(divide_error)
CFI_ADJUST_CFA_OFFSET 4
jmp error_code
CFI_ENDPROC
+END(divide_error)
#ifdef CONFIG_X86_MCE
ENTRY(machine_check)
@@ -1209,6 +1245,7 @@ ENTRY(machine_check)
CFI_ADJUST_CFA_OFFSET 4
jmp error_code
CFI_ENDPROC
+END(machine_check)
#endif
#ifndef CONFIG_XEN
@@ -1220,6 +1257,7 @@ ENTRY(spurious_interrupt_bug)
CFI_ADJUST_CFA_OFFSET 4
jmp error_code
CFI_ENDPROC
+END(spurious_interrupt_bug)
#endif /* !CONFIG_XEN */
ENTRY(fixup_4gb_segment)
@@ -1228,6 +1266,7 @@ ENTRY(fixup_4gb_segment)
CFI_ADJUST_CFA_OFFSET 4
jmp error_code
CFI_ENDPROC
+END(fixup_4gb_segment)
ENTRY(kernel_thread_helper)
pushl $0 # fake return address for unwinder
--- head.orig/arch/x86/kernel/head_32-xen.S 2011-08-09 10:36:13.000000000 +0200
+++ head/arch/x86/kernel/head_32-xen.S 2011-08-09 10:36:25.000000000 +0200
@@ -28,6 +28,7 @@
#define X86_CAPABILITY new_cpu_data+CPUINFO_x86_capability
#define X86_VENDOR_ID new_cpu_data+CPUINFO_x86_vendor_id
+.section .text.head,"ax",@progbits
#define VIRT_ENTRY_OFFSET 0x0
.org VIRT_ENTRY_OFFSET
ENTRY(startup_32)
@@ -61,11 +62,11 @@ ENTRY(startup_32)
movb $1,X86_HARD_MATH
- xorl %eax,%eax # Clear FS
- movl %eax,%fs
+ xorl %eax,%eax # Clear GS
+ movl %eax,%gs
movl $(__KERNEL_PDA),%eax
- mov %eax,%gs
+ mov %eax,%fs
cld # gcc2 wants the direction flag cleared at all times
@@ -76,7 +77,7 @@ ENTRY(startup_32)
* Point the GDT at this CPU's PDA. This will be
* cpu_gdt_table and boot_pda.
*/
-setup_pda:
+ENTRY(setup_pda)
/* get the PDA pointer */
movl $boot_pda, %eax
--- head.orig/arch/x86/kernel/io_apic_32-xen.c 2011-01-31 17:32:16.000000000 +0100
+++ head/arch/x86/kernel/io_apic_32-xen.c 2011-01-31 17:32:22.000000000 +0100
@@ -167,7 +167,7 @@ static inline void io_apic_write(unsigne
*/
static inline void io_apic_modify(unsigned int apic, unsigned int reg, unsigned int value)
{
- volatile struct io_apic *io_apic = io_apic_base(apic);
+ volatile struct io_apic __iomem *io_apic = io_apic_base(apic);
if (sis_apic_bug)
writel(reg, &io_apic->index);
writel(value, &io_apic->data);
@@ -392,7 +392,7 @@ static void set_ioapic_affinity_irq(unsi
break;
entry = irq_2_pin + entry->next;
}
- set_native_irq_info(irq, cpumask);
+ irq_desc[irq].affinity = cpumask;
spin_unlock_irqrestore(&ioapic_lock, flags);
}
@@ -531,8 +531,8 @@ static void do_irq_balance(void)
package_index = CPU_TO_PACKAGEINDEX(i);
for (j = 0; j < NR_IRQS; j++) {
unsigned long value_now, delta;
- /* Is this an active IRQ? */
- if (!irq_desc[j].action)
+ /* Is this an active IRQ or balancing disabled ? */
+ if (!irq_desc[j].action || irq_balancing_disabled(j))
continue;
if ( package_index == i )
IRQ_DELTA(package_index,j) = 0;
@@ -785,7 +785,7 @@ failed:
return 0;
}
-int __init irqbalance_disable(char *str)
+int __devinit irqbalance_disable(char *str)
{
irqbalance_disabled = 1;
return 1;
@@ -1329,11 +1329,9 @@ static void ioapic_register_intr(int irq
trigger == IOAPIC_LEVEL)
set_irq_chip_and_handler_name(irq, &ioapic_chip,
handle_fasteoi_irq, "fasteoi");
- else {
- irq_desc[irq].status |= IRQ_DELAYED_DISABLE;
+ else
set_irq_chip_and_handler_name(irq, &ioapic_chip,
handle_edge_irq, "edge");
- }
set_intr_gate(vector, interrupt[irq]);
}
#else
@@ -1407,7 +1405,6 @@ static void __init setup_IO_APIC_irqs(vo
}
spin_lock_irqsave(&ioapic_lock, flags);
__ioapic_write_entry(apic, pin, entry);
- set_native_irq_info(irq, TARGET_CPUS);
spin_unlock_irqrestore(&ioapic_lock, flags);
}
}
@@ -1638,7 +1635,7 @@ void /*__init*/ print_local_APIC(void *
v = apic_read(APIC_LVR);
printk(KERN_INFO "... APIC VERSION: %08x\n", v);
ver = GET_APIC_VERSION(v);
- maxlvt = get_maxlvt();
+ maxlvt = lapic_get_maxlvt();
v = apic_read(APIC_TASKPRI);
printk(KERN_DEBUG "... APIC TASKPRI: %08x (%02x)\n", v, v & APIC_TPRI_MASK);
@@ -1976,7 +1973,7 @@ static void __init setup_ioapic_ids_from
#endif
#ifndef CONFIG_XEN
-static int no_timer_check __initdata;
+int no_timer_check __initdata;
static int __init notimercheck(char *s)
{
@@ -2369,7 +2366,7 @@ static inline void __init check_timer(vo
disable_8259A_irq(0);
set_irq_chip_and_handler_name(0, &lapic_chip, handle_fasteoi_irq,
- "fasteio");
+ "fasteoi");
apic_write_around(APIC_LVT0, APIC_DM_FIXED | vector); /* Fixed mode */
enable_8259A_irq(0);
@@ -2662,7 +2659,7 @@ static void set_msi_irq_affinity(unsigne
msg.address_lo |= MSI_ADDR_DEST_ID(dest);
write_msi_msg(irq, &msg);
- set_native_irq_info(irq, mask);
+ irq_desc[irq].affinity = mask;
}
#endif /* CONFIG_SMP */
@@ -2681,25 +2678,32 @@ static struct irq_chip msi_chip = {
.retrigger = ioapic_retrigger_irq,
};
-int arch_setup_msi_irq(unsigned int irq, struct pci_dev *dev)
+int arch_setup_msi_irq(struct pci_dev *dev, struct msi_desc *desc)
{
struct msi_msg msg;
- int ret;
+ int irq, ret;
+ irq = create_irq();
+ if (irq < 0)
+ return irq;
+
+ set_irq_msi(irq, desc);
ret = msi_compose_msg(dev, irq, &msg);
- if (ret < 0)
+ if (ret < 0) {
+ destroy_irq(irq);
return ret;
+ }
write_msi_msg(irq, &msg);
set_irq_chip_and_handler_name(irq, &msi_chip, handle_edge_irq,
"edge");
- return 0;
+ return irq;
}
void arch_teardown_msi_irq(unsigned int irq)
{
- return;
+ destroy_irq(irq);
}
#endif /* CONFIG_PCI_MSI */
@@ -2739,7 +2743,7 @@ static void set_ht_irq_affinity(unsigned
dest = cpu_mask_to_apicid(mask);
target_ht_irq(irq, dest);
- set_native_irq_info(irq, mask);
+ irq_desc[irq].affinity = mask;
}
#endif
@@ -2947,7 +2951,6 @@ int io_apic_set_pci_routing (int ioapic,
spin_lock_irqsave(&ioapic_lock, flags);
__ioapic_write_entry(ioapic, pin, entry);
- set_native_irq_info(irq, TARGET_CPUS);
spin_unlock_irqrestore(&ioapic_lock, flags);
return 0;
--- head.orig/arch/x86/kernel/microcode_core-xen.c 2011-12-01 14:59:46.000000000 +0100
+++ head/arch/x86/kernel/microcode_core-xen.c 2011-01-31 17:32:22.000000000 +0100
@@ -108,7 +108,7 @@ static ssize_t microcode_write (struct f
return ret;
}
-static struct file_operations microcode_fops = {
+static const struct file_operations microcode_fops = {
.owner = THIS_MODULE,
.write = microcode_write,
.open = microcode_open,
--- head.orig/arch/x86/kernel/mpparse_32-xen.c 2011-01-31 17:32:16.000000000 +0100
+++ head/arch/x86/kernel/mpparse_32-xen.c 2011-01-31 17:32:22.000000000 +0100
@@ -1079,7 +1079,7 @@ int mp_register_gsi(u32 gsi, int trigger
static int gsi_to_irq[MAX_GSI_NUM];
/* Don't set up the ACPI SCI because it's already set up */
- if (acpi_fadt.sci_int == gsi)
+ if (acpi_gbl_FADT.sci_interrupt == gsi)
return gsi;
ioapic = mp_find_ioapic(gsi);
@@ -1136,7 +1136,7 @@ int mp_register_gsi(u32 gsi, int trigger
/*
* Don't assign IRQ used by ACPI SCI
*/
- if (gsi == acpi_fadt.sci_int)
+ if (gsi == acpi_gbl_FADT.sci_interrupt)
gsi = pci_irq++;
gsi_to_irq[irq] = gsi;
} else {
--- head.orig/arch/x86/kernel/pci-dma-xen.c 2011-01-31 17:32:16.000000000 +0100
+++ head/arch/x86/kernel/pci-dma-xen.c 2011-01-31 17:32:22.000000000 +0100
@@ -308,7 +308,7 @@ int dma_declare_coherent_memory(struct d
return DMA_MEMORY_IO;
free1_out:
- kfree(dev->dma_mem->bitmap);
+ kfree(dev->dma_mem);
out:
if (mem_base)
iounmap(mem_base);
--- head.orig/arch/x86/kernel/pcspeaker.c 2012-06-06 13:23:57.000000000 +0200
+++ head/arch/x86/kernel/pcspeaker.c 2011-01-31 17:32:22.000000000 +0100
@@ -6,6 +6,11 @@ static __init int add_pcspkr(void)
{
struct platform_device *pd;
+#ifdef CONFIG_XEN
+ if (!is_initial_xendomain())
+ return 0;
+#endif
+
pd = platform_device_register_simple("pcspkr", -1, NULL, 0);
return IS_ERR(pd) ? PTR_ERR(pd) : 0;
--- head.orig/arch/x86/kernel/process_32-xen.c 2011-01-31 17:32:16.000000000 +0100
+++ head/arch/x86/kernel/process_32-xen.c 2011-01-31 17:32:22.000000000 +0100
@@ -38,6 +38,7 @@
#include <linux/ptrace.h>
#include <linux/random.h>
#include <linux/personality.h>
+#include <linux/tick.h>
#include <asm/uaccess.h>
#include <asm/pgtable.h>
@@ -160,6 +161,7 @@ void cpu_idle(void)
/* endless idle loop with no priority at all */
while (1) {
+ tick_nohz_stop_sched_tick();
while (!need_resched()) {
void (*idle)(void);
@@ -175,6 +177,7 @@ void cpu_idle(void)
__get_cpu_var(irq_stat).idle_timestamp = jiffies;
idle();
}
+ tick_nohz_restart_sched_tick();
preempt_enable_no_resched();
schedule();
preempt_disable();
@@ -247,8 +250,8 @@ void show_regs(struct pt_regs * regs)
regs->eax,regs->ebx,regs->ecx,regs->edx);
printk("ESI: %08lx EDI: %08lx EBP: %08lx",
regs->esi, regs->edi, regs->ebp);
- printk(" DS: %04x ES: %04x GS: %04x\n",
- 0xffff & regs->xds,0xffff & regs->xes, 0xffff & regs->xgs);
+ printk(" DS: %04x ES: %04x FS: %04x\n",
+ 0xffff & regs->xds,0xffff & regs->xes, 0xffff & regs->xfs);
cr0 = read_cr0();
cr2 = read_cr2();
@@ -279,7 +282,7 @@ int kernel_thread(int (*fn)(void *), voi
regs.xds = __USER_DS;
regs.xes = __USER_DS;
- regs.xgs = __KERNEL_PDA;
+ regs.xfs = __KERNEL_PDA;
regs.orig_eax = -1;
regs.eip = (unsigned long) kernel_thread_helper;
regs.xcs = __KERNEL_CS | get_kernel_rpl();
@@ -356,7 +359,7 @@ int copy_thread(int nr, unsigned long cl
p->thread.eip = (unsigned long) ret_from_fork;
- savesegment(fs,p->thread.fs);
+ savesegment(gs,p->thread.gs);
tsk = current;
if (unlikely(test_tsk_thread_flag(tsk, TIF_IO_BITMAP))) {
@@ -434,8 +437,8 @@ void dump_thread(struct pt_regs * regs,
dump->regs.eax = regs->eax;
dump->regs.ds = regs->xds;
dump->regs.es = regs->xes;
- savesegment(fs,dump->regs.fs);
- dump->regs.gs = regs->xgs;
+ dump->regs.fs = regs->xfs;
+ savesegment(gs,dump->regs.gs);
dump->regs.orig_eax = regs->orig_eax;
dump->regs.eip = regs->eip;
dump->regs.cs = regs->xcs;
@@ -637,16 +640,6 @@ struct task_struct fastcall * __switch_t
prefetch(&next->i387.fxsave);
/*
- * Restore %fs if needed.
- *
- * Glibc normally makes %fs be zero.
- */
- if (unlikely(next->fs))
- loadsegment(fs, next->fs);
-
- write_pda(pcurrent, next_p);
-
- /*
* Now maybe handle debug registers
*/
if (unlikely(task_thread_info(next_p)->flags & _TIF_WORK_CTXSW))
@@ -654,6 +647,15 @@ struct task_struct fastcall * __switch_t
disable_tsc(prev_p, next_p);
+ /*
+ * Leave lazy mode, flushing any hypercalls made here.
+ * This must be done before restoring TLS segments so
+ * the GDT and LDT are properly updated, and must be
+ * done before math_state_restore, so the TS bit is up
+ * to date.
+ */
+ arch_leave_lazy_cpu_mode();
+
/* If the task has used fpu the last 5 timeslices, just do a full
* restore of the math state immediately to avoid the trap; the
* chances of needing FPU soon are obviously high now
@@ -661,6 +663,14 @@ struct task_struct fastcall * __switch_t
if (next_p->fpu_counter > 5)
math_state_restore();
+ /*
+ * Restore %gs if needed (which is common)
+ */
+ if (prev->gs | next->gs)
+ loadsegment(gs, next->gs);
+
+ write_pda(pcurrent, next_p);
+
return prev_p;
}
--- head.orig/arch/x86/kernel/setup_32-xen.c 2012-06-06 13:48:43.000000000 +0200
+++ head/arch/x86/kernel/setup_32-xen.c 2012-06-06 13:49:07.000000000 +0200
@@ -33,7 +33,6 @@
#include <linux/initrd.h>
#include <linux/bootmem.h>
#include <linux/seq_file.h>
-#include <linux/platform_device.h>
#include <linux/console.h>
#include <linux/mca.h>
#include <linux/root_dev.h>
@@ -148,7 +147,7 @@ unsigned long saved_videomode;
#define RAMDISK_PROMPT_FLAG 0x8000
#define RAMDISK_LOAD_FLAG 0x4000
-static char command_line[COMMAND_LINE_SIZE];
+static char __initdata command_line[COMMAND_LINE_SIZE];
unsigned char __initdata boot_params[PARAM_SIZE];
@@ -649,8 +648,8 @@ void __init setup_arch(char **cmdline_p)
if ((i = MAX_GUEST_CMDLINE) > COMMAND_LINE_SIZE)
i = COMMAND_LINE_SIZE;
- memcpy(saved_command_line, xen_start_info->cmd_line, i);
- saved_command_line[i - 1] = '\0';
+ memcpy(boot_command_line, xen_start_info->cmd_line, i);
+ boot_command_line[i - 1] = '\0';
parse_early_param();
if (user_defined_memmap) {
@@ -658,11 +657,19 @@ void __init setup_arch(char **cmdline_p)
print_memory_map("user");
}
- strlcpy(command_line, saved_command_line, COMMAND_LINE_SIZE);
+ strlcpy(command_line, boot_command_line, COMMAND_LINE_SIZE);
*cmdline_p = command_line;
max_low_pfn = setup_memory();
+#ifdef CONFIG_VMI
+ /*
+ * Must be after max_low_pfn is determined, and before kernel
+ * pagetables are setup.
+ */
+ vmi_init();
+#endif
+
/*
* NOTE: before this point _nobody_ is allowed to allocate
* any memory using the bootmem allocator. Although the
@@ -823,7 +830,6 @@ void __init setup_arch(char **cmdline_p)
conswitchp = &dummy_con;
#endif
}
- tsc_init();
}
static int
@@ -833,31 +839,3 @@ xen_panic_event(struct notifier_block *t
/* we're never actually going to get here... */
return NOTIFY_DONE;
}
-
-static __init int add_pcspkr(void)
-{
- struct platform_device *pd;
- int ret;
-
- if (!is_initial_xendomain())
- return 0;
-
- pd = platform_device_alloc("pcspkr", -1);
- if (!pd)
- return -ENOMEM;
-
- ret = platform_device_add(pd);
- if (ret)
- platform_device_put(pd);
-
- return ret;
-}
-device_initcall(add_pcspkr);
-
-/*
- * Local Variables:
- * mode:c
- * c-file-style:"k&r"
- * c-basic-offset:8
- * End:
- */
--- head.orig/arch/x86/kernel/smp_32-xen.c 2011-01-31 17:32:16.000000000 +0100
+++ head/arch/x86/kernel/smp_32-xen.c 2011-01-31 17:32:22.000000000 +0100
@@ -335,8 +335,7 @@ static void flush_tlb_others(cpumask_t c
/*
* i'm not happy about this global shared spinlock in the
* MM hot path, but we'll see how contended it is.
- * Temporarily this turns IRQs off, so that lockups are
- * detected by the NMI watchdog.
+ * AK: x86-64 has a faster method that could be ported.
*/
spin_lock(&tlbstate_lock);
@@ -361,7 +360,7 @@ static void flush_tlb_others(cpumask_t c
while (!cpus_empty(flush_cpumask))
/* nothing. lockup detection does not belong here */
- mb();
+ cpu_relax();
flush_mm = NULL;
flush_va = 0;
--- head.orig/arch/x86/kernel/time-xen.c 2011-07-12 11:10:37.000000000 +0200
+++ head/arch/x86/kernel/time-xen.c 2011-07-12 11:13:30.000000000 +0200
@@ -50,6 +50,7 @@
#include <linux/kernel_stat.h>
#include <linux/posix-timers.h>
#include <linux/cpufreq.h>
+#include <linux/clocksource.h>
#include <asm/io.h>
#include <asm/smp.h>
@@ -74,25 +75,17 @@
#include <xen/evtchn.h>
#include <xen/interface/vcpu.h>
-#if defined (__i386__)
-#include <asm/i8259.h>
+#ifdef CONFIG_X86_32
#include <asm/i8253.h>
DEFINE_SPINLOCK(i8253_lock);
EXPORT_SYMBOL(i8253_lock);
-#endif
-
-#define XEN_SHIFT 22
-
int pit_latch_buggy; /* extern */
-
-#if defined(__x86_64__)
-unsigned long vxtime_hz = PIT_TICK_RATE;
-struct vxtime_data __vxtime __section_vxtime; /* for vsyscalls */
+#else
volatile unsigned long __jiffies __section_jiffies = INITIAL_JIFFIES;
-struct timespec __xtime __section_xtime;
-struct timezone __sys_tz __section_sys_tz;
#endif
+#define XEN_SHIFT 22
+
unsigned int cpu_khz; /* Detected as we calibrate the TSC */
EXPORT_SYMBOL(cpu_khz);
@@ -206,7 +199,7 @@ static inline u64 scale_delta(u64 delta,
return product;
}
-void init_cpu_khz(void)
+static void init_cpu_khz(void)
{
u64 __cpu_khz = 1000000ULL << 32;
struct vcpu_time_info *info = &vcpu_info(0)->time;
@@ -225,16 +218,6 @@ static u64 get_nsec_offset(struct shadow
return scale_delta(delta, shadow->tsc_to_nsec_mul, shadow->tsc_shift);
}
-#ifdef CONFIG_X86_64
-static unsigned long get_usec_offset(struct shadow_time_info *shadow)
-{
- u64 now, delta;
- rdtscll(now);
- delta = now - shadow->tsc_timestamp;
- return scale_delta(delta, shadow->tsc_to_usec_mul, shadow->tsc_shift);
-}
-#endif
-
static void __update_wallclock(time_t sec, long nsec)
{
long wtm_nsec, xtime_nsec;
@@ -258,7 +241,7 @@ static void __update_wallclock(time_t se
set_normalized_timespec(&wall_to_monotonic, wtm_sec, wtm_nsec);
}
-static void update_wallclock(void)
+static void update_wallclock(bool local)
{
shared_info_t *s = HYPERVISOR_shared_info;
@@ -270,7 +253,7 @@ static void update_wallclock(void)
rmb();
} while ((s->wc_version & 1) | (shadow_tv_version ^ s->wc_version));
- if (!independent_wallclock)
+ if (local)
__update_wallclock(shadow_tv.tv_sec, shadow_tv.tv_nsec);
}
@@ -347,134 +330,21 @@ void rtc_cmos_write(unsigned char val, u
}
EXPORT_SYMBOL(rtc_cmos_write);
-#ifdef CONFIG_X86_64
-
-static struct {
- spinlock_t lock;
- struct timeval tv;
- u32 version;
-} monotonic = { .lock = SPIN_LOCK_UNLOCKED };
-
-/*
- * This version of gettimeofday has microsecond resolution
- * and better than microsecond precision on fast x86 machines with TSC.
- */
-void do_gettimeofday(struct timeval *tv)
-{
- unsigned long seq;
- unsigned long usec, sec;
- unsigned long flags;
- s64 nsec;
- unsigned int cpu;
- struct shadow_time_info *shadow;
- u32 local_time_version, monotonic_version;
-
- cpu = get_cpu();
- shadow = &per_cpu(shadow_time, cpu);
-
- do {
- local_time_version = shadow->version;
- seq = read_seqbegin(&xtime_lock);
-
- usec = get_usec_offset(shadow);
-
- sec = xtime.tv_sec;
- usec += (xtime.tv_nsec / NSEC_PER_USEC);
-
- nsec = shadow->system_timestamp - processed_system_time;
- __normalize_time(&sec, &nsec);
- usec += (long)nsec / NSEC_PER_USEC;
-
- monotonic_version = monotonic.version;
-
- if (unlikely(!time_values_up_to_date(cpu))) {
- /*
- * We may have blocked for a long time,
- * rendering our calculations invalid
- * (e.g. the time delta may have
- * overflowed). Detect that and recalculate
- * with fresh values.
- */
- get_time_values_from_xen(cpu);
- continue;
- }
- } while (read_seqretry(&xtime_lock, seq) ||
- (local_time_version != shadow->version));
-
- put_cpu();
-
- while (usec >= USEC_PER_SEC) {
- usec -= USEC_PER_SEC;
- sec++;
- }
-
- spin_lock_irqsave(&monotonic.lock, flags);
- if (unlikely(sec < monotonic.tv.tv_sec) ||
- (sec == monotonic.tv.tv_sec && usec <= monotonic.tv.tv_usec)) {
- sec = monotonic.tv.tv_sec;
- usec = monotonic.tv.tv_usec;
- } else if (likely(monotonic_version == monotonic.version)) {
- monotonic.tv.tv_sec = sec;
- monotonic.tv.tv_usec = usec;
- }
- spin_unlock_irqrestore(&monotonic.lock, flags);
-
- tv->tv_sec = sec;
- tv->tv_usec = usec;
-}
-
-EXPORT_SYMBOL(do_gettimeofday);
-
-/* Reset monotonic gettimeofday() timeval. */
-static inline void monotonic_reset(void)
-{
- spin_lock(&monotonic.lock);
- monotonic.tv.tv_sec = 0;
- monotonic.tv.tv_usec = 0;
- ++monotonic.version;
- spin_unlock(&monotonic.lock);
-}
-
-int do_settimeofday(struct timespec *tv)
-#elif defined(CONFIG_XEN_PRIVILEGED_GUEST)
+#ifdef CONFIG_XEN_PRIVILEGED_GUEST
int xen_update_wallclock(const struct timespec *tv)
-#endif
-#if defined(CONFIG_X86_64) || defined(CONFIG_XEN_PRIVILEGED_GUEST)
{
time_t sec;
s64 nsec;
unsigned int cpu;
struct shadow_time_info *shadow;
struct xen_platform_op op;
-#ifdef CONFIG_X86_64
- int warp = 0;
- if (unlikely(!tv)) {
- monotonic_reset();
- if (!is_initial_xendomain() || independent_wallclock)
- return 0;
- tv = &xtime;
- warp = 1;
- }
-
- if ((unsigned long)tv->tv_nsec >= NSEC_PER_SEC)
- return -EINVAL;
-
- if (!is_initial_xendomain() && !independent_wallclock)
- return -EPERM;
-#else
if (!is_initial_xendomain() || independent_wallclock)
return -EPERM;
-#endif
cpu = get_cpu();
shadow = &per_cpu(shadow_time, cpu);
-#ifdef CONFIG_X86_64
- if (!warp)
- write_seqlock_irq(&xtime_lock);
-#endif
-
/*
* Ensure we don't get blocked for a long time so that our time delta
* overflows. If that were to happen then our shadow time values would
@@ -489,45 +359,19 @@ int xen_update_wallclock(const struct ti
sec = tv->tv_sec;
__normalize_time(&sec, &nsec);
- if (is_initial_xendomain() && !independent_wallclock) {
- op.cmd = XENPF_settime;
- op.u.settime.secs = sec;
- op.u.settime.nsecs = nsec;
- op.u.settime.system_time = shadow->system_timestamp;
- WARN_ON(HYPERVISOR_platform_op(&op));
- update_wallclock();
-#ifdef CONFIG_X86_64
- if (warp) {
- put_cpu();
- return 0;
- }
-#endif
- } else if (independent_wallclock) {
- nsec -= shadow->system_timestamp;
- __normalize_time(&sec, &nsec);
- __update_wallclock(sec, nsec);
- }
-#ifdef CONFIG_X86_64
- ntp_clear();
-
- monotonic_reset();
-
- write_sequnlock_irq(&xtime_lock);
-#endif
+ op.cmd = XENPF_settime;
+ op.u.settime.secs = sec;
+ op.u.settime.nsecs = nsec;
+ op.u.settime.system_time = shadow->system_timestamp;
+ WARN_ON(HYPERVISOR_platform_op(&op));
+ update_wallclock(false);
put_cpu();
-#ifdef CONFIG_X86_64
- clock_was_set();
-#endif
return 0;
}
#endif
-#ifdef CONFIG_X86_64
-EXPORT_SYMBOL(do_settimeofday);
-#endif
-
static void sync_xen_wallclock(unsigned long dummy);
static DEFINE_TIMER(sync_xen_wallclock_timer, sync_xen_wallclock, 0, 0);
static void sync_xen_wallclock(unsigned long dummy)
@@ -551,7 +395,7 @@ static void sync_xen_wallclock(unsigned
op.u.settime.system_time = processed_system_time;
WARN_ON(HYPERVISOR_platform_op(&op));
- update_wallclock();
+ update_wallclock(false);
write_sequnlock_irq(&xtime_lock);
@@ -576,15 +420,7 @@ static int set_rtc_mmss(unsigned long no
return retval;
}
-#ifdef CONFIG_X86_64
-/* monotonic_clock(): returns # of nanoseconds passed since time_init()
- * Note: This function is required to return accurate
- * time even in the absence of multiple timer ticks.
- */
-unsigned long long monotonic_clock(void)
-#else
unsigned long long sched_clock(void)
-#endif
{
unsigned int cpu = get_cpu();
struct shadow_time_info *shadow = &per_cpu(shadow_time, cpu);
@@ -604,21 +440,18 @@ unsigned long long sched_clock(void)
return time;
}
-#ifdef CONFIG_X86_64
-EXPORT_SYMBOL(monotonic_clock);
-
-unsigned long long sched_clock(void)
-{
- return monotonic_clock();
-}
-#endif
unsigned long profile_pc(struct pt_regs *regs)
{
unsigned long pc = instruction_pointer(regs);
#if defined(CONFIG_SMP) || defined(__x86_64__)
- if (!user_mode_vm(regs) && in_lock_functions(pc)) {
+# ifdef __i386__
+ if (!v8086_mode(regs) && SEGMENT_IS_KERNEL_CODE(regs->xcs)
+# else
+ if (!user_mode(regs)
+# endif
+ && in_lock_functions(pc)) {
# ifdef CONFIG_FRAME_POINTER
# ifdef __i386__
return ((unsigned long *)regs->ebp)[1];
@@ -627,14 +460,11 @@ unsigned long profile_pc(struct pt_regs
# endif
# else
# ifdef __i386__
- unsigned long *sp;
- if ((regs->xcs & 2) == 0)
- sp = (unsigned long *)®s->esp;
- else
- sp = (unsigned long *)regs->esp;
+ unsigned long *sp = (unsigned long *)®s->esp;
# else
unsigned long *sp = (unsigned long *)regs->rsp;
# endif
+
/* Return address is either directly at stack pointer
or above a saved eflags. Eflags has bits 22-31 zero,
kernel addresses don't. */
@@ -726,7 +556,8 @@ irqreturn_t timer_interrupt(int irq, voi
}
if (shadow_tv_version != HYPERVISOR_shared_info->wc_version) {
- update_wallclock();
+ update_wallclock(!is_initial_xendomain()
+ && !independent_wallclock);
schedule_clock_was_set_work = 1;
}
@@ -791,19 +622,6 @@ irqreturn_t timer_interrupt(int irq, voi
return IRQ_HANDLED;
}
-#ifndef CONFIG_X86_64
-
-void tsc_init(void)
-{
- init_cpu_khz();
- printk(KERN_INFO "Xen reported: %u.%03u MHz processor.\n",
- cpu_khz / 1000, cpu_khz % 1000);
-
- use_tsc_delay();
-}
-
-#include <linux/clocksource.h>
-
void mark_tsc_unstable(void)
{
#ifndef CONFIG_XEN /* XXX Should tell the hypervisor about this fact. */
@@ -859,21 +677,9 @@ static struct clocksource clocksource_xe
.mask = CLOCKSOURCE_MASK(64),
.mult = 1 << XEN_SHIFT, /* time directly in nanoseconds */
.shift = XEN_SHIFT,
- .is_continuous = 1,
+ .flags = CLOCK_SOURCE_IS_CONTINUOUS,
};
-static int __init init_xen_clocksource(void)
-{
- clocksource_xen.mult = clocksource_khz2mult(cpu_khz,
- clocksource_xen.shift);
-
- return clocksource_register(&clocksource_xen);
-}
-
-module_init(init_xen_clocksource);
-
-#endif
-
static void init_missing_ticks_accounting(unsigned int cpu)
{
struct vcpu_register_runstate_memory_area area;
@@ -894,7 +700,7 @@ static void init_missing_ticks_accountin
}
/* not static: needed by APM */
-unsigned long get_cmos_time(void)
+unsigned long read_persistent_clock(void)
{
unsigned long retval;
unsigned long flags;
@@ -907,11 +713,11 @@ unsigned long get_cmos_time(void)
return retval;
}
-EXPORT_SYMBOL(get_cmos_time);
static void sync_cmos_clock(unsigned long dummy);
static DEFINE_TIMER(sync_cmos_timer, sync_cmos_clock, 0, 0);
+int no_sync_cmos_clock;
static void sync_cmos_clock(unsigned long dummy)
{
@@ -955,7 +761,8 @@ static void sync_cmos_clock(unsigned lon
void notify_arch_cmos_timer(void)
{
- mod_timer(&sync_cmos_timer, jiffies + 1);
+ if (!no_sync_cmos_clock)
+ mod_timer(&sync_cmos_timer, jiffies + 1);
mod_timer(&sync_xen_wallclock_timer, jiffies + 1);
}
@@ -988,29 +795,11 @@ static int time_init_device(void)
device_initcall(time_init_device);
-#ifdef CONFIG_HPET_TIMER
extern void (*late_time_init)(void);
-/* Duplicate of time_init() below, with hpet_enable part added */
-static void __init hpet_time_init(void)
-{
- struct timespec ts;
- ts.tv_sec = get_cmos_time();
- ts.tv_nsec = (INITIAL_JIFFIES % HZ) * (NSEC_PER_SEC / HZ);
-
- do_settimeofday(&ts);
-
- if ((hpet_enable() >= 0) && hpet_use_timer) {
- printk("Using HPET for base-timer\n");
- }
-
- do_time_init();
-}
-#endif
/* Dynamically-mapped IRQ. */
DEFINE_PER_CPU(int, timer_irq);
-extern void (*late_time_init)(void);
static void setup_cpu0_timer_irq(void)
{
per_cpu(timer_irq, 0) =
@@ -1018,7 +807,7 @@ static void setup_cpu0_timer_irq(void)
VIRQ_TIMER,
0,
timer_interrupt,
- IRQF_DISABLED|IRQF_TIMER,
+ IRQF_DISABLED|IRQF_TIMER|IRQF_NOBALANCING,
"timer0",
NULL);
BUG_ON(per_cpu(timer_irq, 0) < 0);
@@ -1030,16 +819,9 @@ static struct vcpu_set_periodic_timer xe
void __init time_init(void)
{
-#ifdef CONFIG_HPET_TIMER
- if (is_hpet_capable()) {
- /*
- * HPET initialization needs to do memory-mapped io. So, let
- * us do a late initialization after mem_init().
- */
- late_time_init = hpet_time_init;
- return;
- }
-#endif
+ init_cpu_khz();
+ printk(KERN_INFO "Xen reported: %u.%03u MHz processor.\n",
+ cpu_khz / 1000, cpu_khz % 1000);
switch (HYPERVISOR_vcpu_op(VCPUOP_set_periodic_timer, 0,
&xen_set_periodic_tick)) {
@@ -1058,18 +840,12 @@ void __init time_init(void)
per_cpu(processed_system_time, 0) = processed_system_time;
init_missing_ticks_accounting(0);
- update_wallclock();
+ clocksource_register(&clocksource_xen);
-#ifdef CONFIG_X86_64
- init_cpu_khz();
- printk(KERN_INFO "Xen reported: %u.%03u MHz processor.\n",
- cpu_khz / 1000, cpu_khz % 1000);
+ update_wallclock(false);
- vxtime.mode = VXTIME_TSC;
- vxtime.quot = (1000000L << 32) / vxtime_hz;
- vxtime.tsc_quot = (1000L << 32) / cpu_khz;
- sync_core();
- rdtscll(vxtime.last_tsc);
+#ifndef CONFIG_X86_64
+ use_tsc_delay();
#endif
/* Cannot request_irq() until kmem is initialised. */
@@ -1204,7 +980,7 @@ void time_resume(void)
processed_system_time = per_cpu(shadow_time, 0).system_timestamp;
- update_wallclock();
+ update_wallclock(false);
}
#ifdef CONFIG_SMP
@@ -1239,7 +1015,7 @@ int __cpuinit local_setup_timer(unsigned
irq = bind_virq_to_irqhandler(VIRQ_TIMER,
cpu,
timer_interrupt,
- IRQF_DISABLED|IRQF_TIMER,
+ IRQF_DISABLED|IRQF_TIMER|IRQF_NOBALANCING,
timer_name[cpu],
NULL);
if (irq < 0)
@@ -1328,7 +1104,7 @@ static ctl_table xen_table[] = {
};
static int __init xen_sysctl_init(void)
{
- (void)register_sysctl_table(xen_table, 0);
+ (void)register_sysctl_table(xen_table);
return 0;
}
__initcall(xen_sysctl_init);
--- head.orig/arch/x86/kernel/traps_32-xen.c 2011-01-31 17:32:16.000000000 +0100
+++ head/arch/x86/kernel/traps_32-xen.c 2011-01-31 17:32:22.000000000 +0100
@@ -100,6 +100,7 @@ asmlinkage void fixup_4gb_segment(void);
asmlinkage void machine_check(void);
int kstack_depth_to_print = 24;
+static unsigned int code_bytes = 64;
ATOMIC_NOTIFIER_HEAD(i386die_chain);
int register_die_notifier(struct notifier_block *nb)
@@ -297,10 +298,11 @@ void show_registers(struct pt_regs *regs
int i;
int in_kernel = 1;
unsigned long esp;
- unsigned short ss;
+ unsigned short ss, gs;
esp = (unsigned long) (®s->esp);
savesegment(ss, ss);
+ savesegment(gs, gs);
if (user_mode_vm(regs)) {
in_kernel = 0;
esp = regs->esp;
@@ -319,8 +321,8 @@ void show_registers(struct pt_regs *regs
regs->eax, regs->ebx, regs->ecx, regs->edx);
printk(KERN_EMERG "esi: %08lx edi: %08lx ebp: %08lx esp: %08lx\n",
regs->esi, regs->edi, regs->ebp, esp);
- printk(KERN_EMERG "ds: %04x es: %04x ss: %04x\n",
- regs->xds & 0xffff, regs->xes & 0xffff, ss);
+ printk(KERN_EMERG "ds: %04x es: %04x fs: %04x gs: %04x ss: %04x\n",
+ regs->xds & 0xffff, regs->xes & 0xffff, regs->xfs & 0xffff, gs, ss);
printk(KERN_EMERG "Process %.*s (pid: %d, ti=%p task=%p task.ti=%p)",
TASK_COMM_LEN, current->comm, current->pid,
current_thread_info(), current, current->thread_info);
@@ -330,7 +332,8 @@ void show_registers(struct pt_regs *regs
*/
if (in_kernel) {
u8 *eip;
- int code_bytes = 64;
+ unsigned int code_prologue = code_bytes * 43 / 64;
+ unsigned int code_len = code_bytes;
unsigned char c;
printk("\n" KERN_EMERG "Stack: ");
@@ -338,14 +341,14 @@ void show_registers(struct pt_regs *regs
printk(KERN_EMERG "Code: ");
- eip = (u8 *)regs->eip - 43;
+ eip = (u8 *)regs->eip - code_prologue;
if (eip < (u8 *)PAGE_OFFSET ||
probe_kernel_address(eip, c)) {
/* try starting at EIP */
eip = (u8 *)regs->eip;
- code_bytes = 32;
+ code_len = code_len - code_prologue + 1;
}
- for (i = 0; i < code_bytes; i++, eip++) {
+ for (i = 0; i < code_len; i++, eip++) {
if (eip < (u8 *)PAGE_OFFSET ||
probe_kernel_address(eip, c)) {
printk(" Bad EIP value.");
@@ -1134,3 +1137,13 @@ static int __init kstack_setup(char *s)
return 1;
}
__setup("kstack=", kstack_setup);
+
+static int __init code_bytes_setup(char *s)
+{
+ code_bytes = simple_strtoul(s, NULL, 0);
+ if (code_bytes > 8192)
+ code_bytes = 8192;
+
+ return 1;
+}
+__setup("code_bytes=", code_bytes_setup);
--- head.orig/arch/x86/mm/fault_32-xen.c 2011-01-31 17:32:16.000000000 +0100
+++ head/arch/x86/mm/fault_32-xen.c 2011-01-31 17:32:22.000000000 +0100
@@ -46,43 +46,17 @@ int unregister_page_fault_notifier(struc
}
EXPORT_SYMBOL_GPL(unregister_page_fault_notifier);
-static inline int notify_page_fault(enum die_val val, const char *str,
- struct pt_regs *regs, long err, int trap, int sig)
+static inline int notify_page_fault(struct pt_regs *regs, long err)
{
struct die_args args = {
.regs = regs,
- .str = str,
+ .str = "page fault",
.err = err,
- .trapnr = trap,
- .signr = sig
+ .trapnr = 14,
+ .signr = SIGSEGV
};
- return atomic_notifier_call_chain(¬ify_page_fault_chain, val, &args);
-}
-
-/*
- * Unlock any spinlocks which will prevent us from getting the
- * message out
- */
-void bust_spinlocks(int yes)
-{
- int loglevel_save = console_loglevel;
-
- if (yes) {
- oops_in_progress = 1;
- return;
- }
-#ifdef CONFIG_VT
- unblank_screen();
-#endif
- oops_in_progress = 0;
- /*
- * OK, the message is on the console. Now we call printk()
- * without oops_in_progress set so that printk will give klogd
- * a poke. Hold onto your hats...
- */
- console_loglevel = 15; /* NMI oopser may have shut the console up */
- printk(" ");
- console_loglevel = loglevel_save;
+ return atomic_notifier_call_chain(¬ify_page_fault_chain,
+ DIE_PAGE_FAULT, &args);
}
/*
@@ -476,8 +450,7 @@ fastcall void __kprobes do_page_fault(st
/* Can take a spurious fault if mapping changes R/O -> R/W. */
if (spurious_fault(regs, address, error_code))
return;
- if (notify_page_fault(DIE_PAGE_FAULT, "page fault", regs, error_code, 14,
- SIGSEGV) == NOTIFY_STOP)
+ if (notify_page_fault(regs, error_code) == NOTIFY_STOP)
return;
/*
* Don't take the mm semaphore here. If we fixup a prefetch
@@ -486,8 +459,7 @@ fastcall void __kprobes do_page_fault(st
goto bad_area_nosemaphore;
}
- if (notify_page_fault(DIE_PAGE_FAULT, "page fault", regs, error_code, 14,
- SIGSEGV) == NOTIFY_STOP)
+ if (notify_page_fault(regs, error_code) == NOTIFY_STOP)
return;
/* It's safe to allow irq's after cr2 has been saved and the vmalloc
--- head.orig/arch/x86/mm/highmem_32-xen.c 2011-01-31 17:32:16.000000000 +0100
+++ head/arch/x86/mm/highmem_32-xen.c 2011-01-31 17:32:22.000000000 +0100
@@ -33,14 +33,16 @@ static void *__kmap_atomic(struct page *
/* even !CONFIG_PREEMPT needs this, for in_atomic in do_page_fault */
pagefault_disable();
+
+ idx = type + KM_TYPE_NR*smp_processor_id();
+ BUG_ON(!pte_none(*(kmap_pte-idx)));
+
if (!PageHighMem(page))
return page_address(page);
- idx = type + KM_TYPE_NR*smp_processor_id();
vaddr = __fix_to_virt(FIX_KMAP_BEGIN + idx);
- if (!pte_none(*(kmap_pte-idx)))
- BUG();
set_pte_at(&init_mm, vaddr, kmap_pte-idx, mk_pte(page, prot));
+ /*arch_flush_lazy_mmu_mode();*/
return (void*) vaddr;
}
@@ -94,6 +96,7 @@ void *kmap_atomic_pfn(unsigned long pfn,
idx = type + KM_TYPE_NR*smp_processor_id();
vaddr = __fix_to_virt(FIX_KMAP_BEGIN + idx);
set_pte(kmap_pte-idx, pfn_pte(pfn, kmap_prot));
+ /*arch_flush_lazy_mmu_mode();*/
return (void*) vaddr;
}
--- head.orig/arch/x86/mm/init_32-xen.c 2011-01-31 17:32:16.000000000 +0100
+++ head/arch/x86/mm/init_32-xen.c 2011-01-31 17:32:22.000000000 +0100
@@ -66,6 +66,7 @@ static pmd_t * __init one_md_table_init(
#ifdef CONFIG_X86_PAE
pmd_table = (pmd_t *) alloc_bootmem_low_pages(PAGE_SIZE);
+ paravirt_alloc_pd(__pa(pmd_table) >> PAGE_SHIFT);
make_lowmem_page_readonly(pmd_table, XENFEAT_writable_page_tables);
set_pgd(pgd, __pgd(__pa(pmd_table) | _PAGE_PRESENT));
pud = pud_offset(pgd, 0);
@@ -87,6 +88,7 @@ static pte_t * __init one_page_table_ini
{
if (pmd_none(*pmd)) {
pte_t *page_table = (pte_t *) alloc_bootmem_low_pages(PAGE_SIZE);
+ paravirt_alloc_pt(__pa(page_table) >> PAGE_SHIFT);
make_lowmem_page_readonly(page_table,
XENFEAT_writable_page_tables);
set_pmd(pmd, __pmd(__pa(page_table) | _PAGE_TABLE));
--- head.orig/arch/x86/mm/pgtable_32-xen.c 2011-01-31 17:32:16.000000000 +0100
+++ head/arch/x86/mm/pgtable_32-xen.c 2011-01-31 17:32:22.000000000 +0100
@@ -149,6 +149,8 @@ void __set_fixmap (enum fixed_addresses
void __init reserve_top_address(unsigned long reserve)
{
BUG_ON(fixmaps > 0);
+ printk(KERN_INFO "Reserving virtual address space above 0x%08x\n",
+ (int)-reserve);
__FIXADDR_TOP = -reserve - PAGE_SIZE;
__VMALLOC_RESERVE += reserve;
}
@@ -259,6 +261,12 @@ void pgd_ctor(void *pgd, struct kmem_cac
swapper_pg_dir + USER_PTRS_PER_PGD,
KERNEL_PGD_PTRS);
memset(pgd, 0, USER_PTRS_PER_PGD*sizeof(pgd_t));
+
+ /* must happen under lock */
+ paravirt_alloc_pd_clone(__pa(pgd) >> PAGE_SHIFT,
+ __pa(swapper_pg_dir) >> PAGE_SHIFT,
+ USER_PTRS_PER_PGD, PTRS_PER_PGD - USER_PTRS_PER_PGD);
+
pgd_list_add(pgd);
spin_unlock_irqrestore(&pgd_lock, flags);
}
@@ -269,6 +277,7 @@ void pgd_dtor(void *pgd, struct kmem_cac
{
unsigned long flags; /* can be called from interrupt context */
+ paravirt_release_pd(__pa(pgd) >> PAGE_SHIFT);
spin_lock_irqsave(&pgd_lock, flags);
pgd_list_del(pgd);
spin_unlock_irqrestore(&pgd_lock, flags);
@@ -299,6 +308,7 @@ pgd_t *pgd_alloc(struct mm_struct *mm)
pmd_t *pmd = kmem_cache_alloc(pmd_cache, GFP_KERNEL);
if (!pmd)
goto out_oom;
+ paravirt_alloc_pd(__pa(pmd) >> PAGE_SHIFT);
set_pgd(&pgd[i], __pgd(1 + __pa(pmd)));
}
return pgd;
@@ -321,6 +331,7 @@ pgd_t *pgd_alloc(struct mm_struct *mm)
pmd[i] = kmem_cache_alloc(pmd_cache, GFP_KERNEL);
if (!pmd[i])
goto out_oom;
+ paravirt_alloc_pd(__pa(pmd) >> PAGE_SHIFT);
}
spin_lock_irqsave(&pgd_lock, flags);
@@ -361,12 +372,17 @@ pgd_t *pgd_alloc(struct mm_struct *mm)
out_oom:
if (HAVE_SHARED_KERNEL_PMD) {
- for (i--; i >= 0; i--)
- kmem_cache_free(pmd_cache,
- (void *)__va(pgd_val(pgd[i])-1));
+ for (i--; i >= 0; i--) {
+ pgd_t pgdent = pgd[i];
+ void* pmd = (void *)__va(pgd_val(pgdent)-1);
+ paravirt_release_pd(__pa(pmd) >> PAGE_SHIFT);
+ kmem_cache_free(pmd_cache, pmd);
+ }
} else {
- for (i--; i >= 0; i--)
+ for (i--; i >= 0; i--) {
+ paravirt_release_pd(__pa(pmd[i]) >> PAGE_SHIFT);
kmem_cache_free(pmd_cache, pmd[i]);
+ }
kfree(pmd);
}
kmem_cache_free(pgd_cache, pgd);
@@ -390,7 +406,9 @@ void pgd_free(pgd_t *pgd)
/* in the PAE case user pgd entries are overwritten before usage */
if (PTRS_PER_PMD > 1) {
for (i = 0; i < USER_PTRS_PER_PGD; ++i) {
- pmd_t *pmd = (void *)__va(pgd_val(pgd[i])-1);
+ pgd_t pgdent = pgd[i];
+ void* pmd = (void *)__va(pgd_val(pgdent)-1);
+ paravirt_release_pd(__pa(pmd) >> PAGE_SHIFT);
kmem_cache_free(pmd_cache, pmd);
}
--- head.orig/arch/x86/ia32/ia32entry-xen.S 2011-01-31 17:29:16.000000000 +0100
+++ head/arch/x86/ia32/ia32entry-xen.S 2011-01-31 17:32:22.000000000 +0100
@@ -465,7 +465,7 @@ ia32_sys_call_table:
.quad sys32_vm86_warning /* vm86old */
.quad compat_sys_wait4
.quad sys_swapoff /* 115 */
- .quad sys32_sysinfo
+ .quad compat_sys_sysinfo
.quad sys32_ipc
.quad sys_fsync
.quad stub32_sigreturn
@@ -510,7 +510,7 @@ ia32_sys_call_table:
.quad sys_sched_yield
.quad sys_sched_get_priority_max
.quad sys_sched_get_priority_min /* 160 */
- .quad sys_sched_rr_get_interval
+ .quad sys32_sched_rr_get_interval
.quad compat_sys_nanosleep
.quad sys_mremap
.quad sys_setresuid16
@@ -668,4 +668,5 @@ ia32_sys_call_table:
.quad compat_sys_vmsplice
.quad compat_sys_move_pages
.quad sys_getcpu
+ .quad sys_epoll_pwait
ia32_syscall_end:
--- head.orig/arch/x86/kernel/e820_64-xen.c 2011-01-31 17:29:16.000000000 +0100
+++ head/arch/x86/kernel/e820_64-xen.c 2011-01-31 17:32:22.000000000 +0100
@@ -90,6 +90,13 @@ static inline int bad_addr(unsigned long
return 1;
}
+#ifdef CONFIG_NUMA
+ /* NUMA memory to node map */
+ if (last >= nodemap_addr && addr < nodemap_addr + nodemap_size) {
+ *addrp = nodemap_addr + nodemap_size;
+ return 1;
+ }
+#endif
/* XXX ramdisk image here? */
#else
if (last < (table_end<<PAGE_SHIFT)) {
@@ -215,6 +222,37 @@ unsigned long __init e820_end_of_ram(voi
}
/*
+ * Find the hole size in the range.
+ */
+unsigned long __init e820_hole_size(unsigned long start, unsigned long end)
+{
+ unsigned long ram = 0;
+ int i;
+
+ for (i = 0; i < e820.nr_map; i++) {
+ struct e820entry *ei = &e820.map[i];
+ unsigned long last, addr;
+
+ if (ei->type != E820_RAM ||
+ ei->addr+ei->size <= start ||
+ ei->addr >= end)
+ continue;
+
+ addr = round_up(ei->addr, PAGE_SIZE);
+ if (addr < start)
+ addr = start;
+
+ last = round_down(ei->addr + ei->size, PAGE_SIZE);
+ if (last >= end)
+ last = end;
+
+ if (last > addr)
+ ram += last - addr;
+ }
+ return ((end - start) - ram);
+}
+
+/*
* Mark e820 reserved areas as busy for the resource manager.
*/
void __init e820_reserve_resources(struct e820entry *e820, int nr_map)
@@ -751,7 +789,7 @@ static int __init parse_memmap_opt(char
}
early_param("memmap", parse_memmap_opt);
-void finish_e820_parsing(void)
+void __init finish_e820_parsing(void)
{
if (userdef) {
printk(KERN_INFO "user-defined physical RAM map:\n");
--- head.orig/arch/x86/kernel/entry_64-xen.S 2011-10-07 11:18:32.000000000 +0200
+++ head/arch/x86/kernel/entry_64-xen.S 2011-10-07 11:20:12.000000000 +0200
@@ -632,6 +632,9 @@ END(invalidate_interrupt\num)
ENTRY(call_function_interrupt)
apicinterrupt CALL_FUNCTION_VECTOR,smp_call_function_interrupt
END(call_function_interrupt)
+ENTRY(irq_move_cleanup_interrupt)
+ apicinterrupt IRQ_MOVE_CLEANUP_VECTOR,smp_irq_move_cleanup_interrupt
+END(irq_move_cleanup_interrupt)
#endif
ENTRY(apic_timer_interrupt)
--- head.orig/arch/x86/kernel/head64-xen.c 2011-01-31 17:32:16.000000000 +0100
+++ head/arch/x86/kernel/head64-xen.c 2011-01-31 17:32:22.000000000 +0100
@@ -45,8 +45,6 @@ static void __init clear_bss(void)
#define OLD_CL_BASE_ADDR 0x90000
#define OLD_CL_OFFSET 0x90022
-extern char saved_command_line[];
-
static void __init copy_bootdata(char *real_mode_data)
{
#ifndef CONFIG_XEN
@@ -62,14 +60,14 @@ static void __init copy_bootdata(char *r
new_data = OLD_CL_BASE_ADDR + * (u16 *) OLD_CL_OFFSET;
}
command_line = (char *) ((u64)(new_data));
- memcpy(saved_command_line, command_line, COMMAND_LINE_SIZE);
+ memcpy(boot_command_line, command_line, COMMAND_LINE_SIZE);
#else
int max_cmdline;
if ((max_cmdline = MAX_GUEST_CMDLINE) > COMMAND_LINE_SIZE)
max_cmdline = COMMAND_LINE_SIZE;
- memcpy(saved_command_line, xen_start_info->cmd_line, max_cmdline);
- saved_command_line[max_cmdline-1] = '\0';
+ memcpy(boot_command_line, xen_start_info->cmd_line, max_cmdline);
+ boot_command_line[max_cmdline-1] = '\0';
#endif
}
--- head.orig/arch/x86/kernel/io_apic_64-xen.c 2011-01-31 17:32:16.000000000 +0100
+++ head/arch/x86/kernel/io_apic_64-xen.c 2011-01-31 17:32:22.000000000 +0100
@@ -36,6 +36,7 @@
#include <acpi/acpi_bus.h>
#endif
+#include <asm/idle.h>
#include <asm/io.h>
#include <asm/smp.h>
#include <asm/desc.h>
@@ -47,7 +48,20 @@
#include <asm/msidef.h>
#include <asm/hypertransport.h>
-static int assign_irq_vector(int irq, cpumask_t mask, cpumask_t *result);
+struct irq_cfg {
+#ifndef CONFIG_XEN
+ cpumask_t domain;
+ cpumask_t old_domain;
+#endif
+ unsigned move_cleanup_count;
+ u8 vector;
+ u8 move_in_progress : 1;
+};
+
+/* irq_cfg is indexed by the sum of all RTEs in all I/O APICs. */
+struct irq_cfg irq_cfg[NR_IRQS] __read_mostly;
+
+static int assign_irq_vector(int irq, cpumask_t mask);
#define __apicdebuginit __init
@@ -89,7 +103,7 @@ int nr_ioapic_registers[MAX_IO_APICS];
* Rough estimation of how many shared IRQs there are, can
* be changed anytime.
*/
-#define MAX_PLUS_SHARED_IRQS NR_IRQ_VECTORS
+#define MAX_PLUS_SHARED_IRQS NR_IRQS
#define PIN_MAP_SIZE (MAX_PLUS_SHARED_IRQS + NR_IRQS)
/*
@@ -262,21 +276,19 @@ static void __target_IO_APIC_irq(unsigne
static void set_ioapic_affinity_irq(unsigned int irq, cpumask_t mask)
{
+ struct irq_cfg *cfg = irq_cfg + irq;
unsigned long flags;
unsigned int dest;
cpumask_t tmp;
- int vector;
cpus_and(tmp, mask, cpu_online_map);
if (cpus_empty(tmp))
- tmp = TARGET_CPUS;
-
- cpus_and(mask, tmp, CPU_MASK_ALL);
+ return;
- vector = assign_irq_vector(irq, mask, &tmp);
- if (vector < 0)
+ if (assign_irq_vector(irq, mask))
return;
+ cpus_and(tmp, cfg->domain, mask);
dest = cpu_mask_to_apicid(tmp);
/*
@@ -285,8 +297,8 @@ static void set_ioapic_affinity_irq(unsi
dest = SET_APIC_LOGICAL_ID(dest);
spin_lock_irqsave(&ioapic_lock, flags);
- __target_IO_APIC_irq(irq, dest, vector);
- set_native_irq_info(irq, mask);
+ __target_IO_APIC_irq(irq, dest, cfg->vector);
+ irq_desc[irq].affinity = mask;
spin_unlock_irqrestore(&ioapic_lock, flags);
}
#endif
@@ -332,11 +344,11 @@ static void add_pin_to_irq(unsigned int
reg = io_apic_read(entry->apic, 0x10 + R + pin*2); \
reg ACTION; \
io_apic_modify(entry->apic, reg); \
+ FINAL; \
if (!entry->next) \
break; \
entry = irq_2_pin + entry->next; \
} \
- FINAL; \
}
#define DO_ACTION(name,R,ACTION, FINAL) \
@@ -669,77 +681,62 @@ static int pin_2_irq(int idx, int apic,
return irq;
}
-static inline int IO_APIC_irq_trigger(int irq)
-{
- int apic, idx, pin;
-
- for (apic = 0; apic < nr_ioapics; apic++) {
- for (pin = 0; pin < nr_ioapic_registers[apic]; pin++) {
- idx = find_irq_entry(apic,pin,mp_INT);
- if ((idx != -1) && (irq == pin_2_irq(idx,apic,pin)))
- return irq_trigger(idx);
- }
- }
- /*
- * nonexistent IRQs are edge default
- */
- return 0;
-}
-
-/* irq_vectors is indexed by the sum of all RTEs in all I/O APICs. */
-static u8 irq_vector[NR_IRQ_VECTORS] __read_mostly;
-
-static int __assign_irq_vector(int irq, cpumask_t mask, cpumask_t *result)
+static int __assign_irq_vector(int irq, cpumask_t mask)
{
- int vector;
struct physdev_irq irq_op;
+ struct irq_cfg *cfg;
- BUG_ON((unsigned)irq >= NR_IRQ_VECTORS);
+ BUG_ON((unsigned)irq >= NR_IRQS);
if (irq < PIRQ_BASE || irq - PIRQ_BASE >= NR_PIRQS)
return -EINVAL;
- cpus_and(*result, mask, cpu_online_map);
+ cfg = &irq_cfg[irq];
+
+ if ((cfg->move_in_progress) || cfg->move_cleanup_count)
+ return -EBUSY;
- if (irq_vector[irq] > 0)
- return irq_vector[irq];
+ if (cfg->vector)
+ return 0;
irq_op.irq = irq;
if (HYPERVISOR_physdev_op(PHYSDEVOP_alloc_irq_vector, &irq_op))
return -ENOSPC;
- vector = irq_op.vector;
- irq_vector[irq] = vector;
+ cfg->vector = irq_op.vector;
- return vector;
+ return 0;
}
-static int assign_irq_vector(int irq, cpumask_t mask, cpumask_t *result)
+static int assign_irq_vector(int irq, cpumask_t mask)
{
- int vector;
+ int err;
unsigned long flags;
spin_lock_irqsave(&vector_lock, flags);
- vector = __assign_irq_vector(irq, mask, result);
+ err = __assign_irq_vector(irq, mask);
spin_unlock_irqrestore(&vector_lock, flags);
- return vector;
+ return err;
}
#ifndef CONFIG_XEN
static void __clear_irq_vector(int irq)
{
+ struct irq_cfg *cfg;
cpumask_t mask;
int cpu, vector;
- BUG_ON(!irq_vector[irq]);
+ BUG_ON((unsigned)irq >= NR_IRQS);
+ cfg = &irq_cfg[irq];
+ BUG_ON(!cfg->vector);
- vector = irq_vector[irq];
- cpus_and(mask, irq_domain[irq], cpu_online_map);
+ vector = cfg->vector;
+ cpus_and(mask, cfg->domain, cpu_online_map);
for_each_cpu_mask(cpu, mask)
per_cpu(vector_irq, cpu)[vector] = -1;
- irq_vector[irq] = 0;
- irq_domain[irq] = CPU_MASK_NONE;
+ cfg->vector = 0;
+ cfg->domain = CPU_MASK_NONE;
}
void __setup_vector_irq(int cpu)
@@ -749,10 +746,10 @@ void __setup_vector_irq(int cpu)
int irq, vector;
/* Mark the inuse vectors */
- for (irq = 0; irq < NR_IRQ_VECTORS; ++irq) {
- if (!cpu_isset(cpu, irq_domain[irq]))
+ for (irq = 0; irq < NR_IRQS; ++irq) {
+ if (!cpu_isset(cpu, irq_cfg[irq].domain))
continue;
- vector = irq_vector[irq];
+ vector = irq_cfg[irq].vector;
per_cpu(vector_irq, cpu)[vector] = irq;
}
/* Mark the free vectors */
@@ -760,41 +757,49 @@ void __setup_vector_irq(int cpu)
irq = per_cpu(vector_irq, cpu)[vector];
if (irq < 0)
continue;
- if (!cpu_isset(cpu, irq_domain[irq]))
+ if (!cpu_isset(cpu, irq_cfg[irq].domain))
per_cpu(vector_irq, cpu)[vector] = -1;
}
}
-extern void (*interrupt[NR_IRQS])(void);
-
static struct irq_chip ioapic_chip;
-#define IOAPIC_AUTO -1
-#define IOAPIC_EDGE 0
-#define IOAPIC_LEVEL 1
-
-static void ioapic_register_intr(int irq, int vector, unsigned long trigger)
+static void ioapic_register_intr(int irq, unsigned long trigger)
{
- if ((trigger == IOAPIC_AUTO && IO_APIC_irq_trigger(irq)) ||
- trigger == IOAPIC_LEVEL)
+ if (trigger)
set_irq_chip_and_handler_name(irq, &ioapic_chip,
handle_fasteoi_irq, "fasteoi");
- else {
- irq_desc[irq].status |= IRQ_DELAYED_DISABLE;
+ else
set_irq_chip_and_handler_name(irq, &ioapic_chip,
handle_edge_irq, "edge");
- }
}
#else
-#define ioapic_register_intr(irq, vector, trigger) evtchn_register_pirq(irq)
+#define ioapic_register_intr(irq, trigger) evtchn_register_pirq(irq)
#endif /* !CONFIG_XEN */
-static void __init setup_IO_APIC_irq(int apic, int pin, int idx, int irq)
+static void setup_IO_APIC_irq(int apic, int pin, unsigned int irq,
+ int trigger, int polarity)
{
+ struct irq_cfg *cfg = irq_cfg + irq;
struct IO_APIC_route_entry entry;
- int vector;
- unsigned long flags;
+ cpumask_t mask;
+
+ if (!IO_APIC_IRQ(irq))
+ return;
+ mask = TARGET_CPUS;
+ if (assign_irq_vector(irq, mask))
+ return;
+
+#ifndef CONFIG_XEN
+ cpus_and(mask, cfg->domain, mask);
+#endif
+
+ apic_printk(APIC_VERBOSE,KERN_DEBUG
+ "IOAPIC[%d]: Set routing entry (%d-%d -> 0x%x -> "
+ "IRQ %d Mode:%i Active:%i)\n",
+ apic, mp_ioapics[apic].mpc_apicid, pin, cfg->vector,
+ irq, trigger, polarity);
/*
* add it to the IO-APIC irq-routing table:
@@ -803,41 +808,23 @@ static void __init setup_IO_APIC_irq(int
entry.delivery_mode = INT_DELIVERY_MODE;
entry.dest_mode = INT_DEST_MODE;
+ entry.dest = cpu_mask_to_apicid(mask);
entry.mask = 0; /* enable IRQ */
- entry.dest.logical.logical_dest = cpu_mask_to_apicid(TARGET_CPUS);
-
- entry.trigger = irq_trigger(idx);
- entry.polarity = irq_polarity(idx);
+ entry.trigger = trigger;
+ entry.polarity = polarity;
+ entry.vector = cfg->vector;
- if (irq_trigger(idx)) {
- entry.trigger = 1;
+ /* Mask level triggered irqs.
+ * Use IRQ_DELAYED_DISABLE for edge triggered irqs.
+ */
+ if (trigger)
entry.mask = 1;
- entry.dest.logical.logical_dest = cpu_mask_to_apicid(TARGET_CPUS);
- }
-
- if (/* !apic && */ !IO_APIC_IRQ(irq))
- return;
- if (IO_APIC_IRQ(irq)) {
- cpumask_t mask;
- vector = assign_irq_vector(irq, TARGET_CPUS, &mask);
- if (vector < 0)
- return;
-
- entry.dest.logical.logical_dest = cpu_mask_to_apicid(mask);
- entry.vector = vector;
-
- ioapic_register_intr(irq, vector, IOAPIC_AUTO);
- if (!apic && (irq < 16))
- disable_8259A_irq(irq);
- }
+ ioapic_register_intr(irq, trigger);
+ if (irq < 16)
+ disable_8259A_irq(irq);
ioapic_write_entry(apic, pin, entry);
-
- spin_lock_irqsave(&ioapic_lock, flags);
- set_native_irq_info(irq, TARGET_CPUS);
- spin_unlock_irqrestore(&ioapic_lock, flags);
-
}
static void __init setup_IO_APIC_irqs(void)
@@ -862,8 +849,8 @@ static void __init setup_IO_APIC_irqs(vo
irq = pin_2_irq(idx, apic, pin);
add_pin_to_irq(irq, apic, pin);
- setup_IO_APIC_irq(apic, pin, idx, irq);
-
+ setup_IO_APIC_irq(apic, pin, irq,
+ irq_trigger(idx), irq_polarity(idx));
}
}
@@ -894,7 +881,7 @@ static void __init setup_ExtINT_IRQ0_pin
*/
entry.dest_mode = INT_DEST_MODE;
entry.mask = 0; /* unmask IRQ now */
- entry.dest.logical.logical_dest = cpu_mask_to_apicid(TARGET_CPUS);
+ entry.dest = cpu_mask_to_apicid(TARGET_CPUS);
entry.delivery_mode = INT_DELIVERY_MODE;
entry.polarity = 0;
entry.trigger = 0;
@@ -994,18 +981,17 @@ void __apicdebuginit print_IO_APIC(void)
printk(KERN_DEBUG ".... IRQ redirection table:\n");
- printk(KERN_DEBUG " NR Log Phy Mask Trig IRR Pol"
- " Stat Dest Deli Vect: \n");
+ printk(KERN_DEBUG " NR Dst Mask Trig IRR Pol"
+ " Stat Dmod Deli Vect: \n");
for (i = 0; i <= reg_01.bits.entries; i++) {
struct IO_APIC_route_entry entry;
entry = ioapic_read_entry(apic, i);
- printk(KERN_DEBUG " %02x %03X %02X ",
+ printk(KERN_DEBUG " %02x %03X ",
i,
- entry.dest.logical.logical_dest,
- entry.dest.physical.physical_dest
+ entry.dest
);
printk("%1d %1d %1d %1d %1d %1d %1d %02X\n",
@@ -1269,8 +1255,7 @@ void disable_IO_APIC(void)
entry.dest_mode = 0; /* Physical */
entry.delivery_mode = dest_ExtINT; /* ExtInt */
entry.vector = 0;
- entry.dest.physical.physical_dest =
- GET_APIC_ID(apic_read(APIC_ID));
+ entry.dest = GET_APIC_ID(apic_read(APIC_ID));
/*
* Add it to the IO-APIC irq-routing table:
@@ -1355,16 +1340,15 @@ static unsigned int startup_ioapic_irq(u
static int ioapic_retrigger_irq(unsigned int irq)
{
+ struct irq_cfg *cfg = &irq_cfg[irq];
cpumask_t mask;
- unsigned vector;
unsigned long flags;
spin_lock_irqsave(&vector_lock, flags);
- vector = irq_vector[irq];
cpus_clear(mask);
- cpu_set(first_cpu(irq_domain[irq]), mask);
+ cpu_set(first_cpu(cfg->domain), mask);
- send_IPI_mask(mask, vector);
+ send_IPI_mask(mask, cfg->vector);
spin_unlock_irqrestore(&vector_lock, flags);
return 1;
@@ -1379,8 +1363,68 @@ static int ioapic_retrigger_irq(unsigned
* races.
*/
+#ifdef CONFIG_SMP
+asmlinkage void smp_irq_move_cleanup_interrupt(void)
+{
+ unsigned vector, me;
+ ack_APIC_irq();
+ exit_idle();
+ irq_enter();
+
+ me = smp_processor_id();
+ for (vector = FIRST_EXTERNAL_VECTOR; vector < NR_VECTORS; vector++) {
+ unsigned int irq;
+ struct irq_desc *desc;
+ struct irq_cfg *cfg;
+ irq = __get_cpu_var(vector_irq)[vector];
+ if (irq >= NR_IRQS)
+ continue;
+
+ desc = irq_desc + irq;
+ cfg = irq_cfg + irq;
+ spin_lock(&desc->lock);
+ if (!cfg->move_cleanup_count)
+ goto unlock;
+
+ if ((vector == cfg->vector) && cpu_isset(me, cfg->domain))
+ goto unlock;
+
+ __get_cpu_var(vector_irq)[vector] = -1;
+ cfg->move_cleanup_count--;
+unlock:
+ spin_unlock(&desc->lock);
+ }
+
+ irq_exit();
+}
+
+static void irq_complete_move(unsigned int irq)
+{
+ struct irq_cfg *cfg = irq_cfg + irq;
+ unsigned vector, me;
+
+ if (likely(!cfg->move_in_progress))
+ return;
+
+ vector = ~get_irq_regs()->orig_rax;
+ me = smp_processor_id();
+ if ((vector == cfg->vector) &&
+ cpu_isset(smp_processor_id(), cfg->domain)) {
+ cpumask_t cleanup_mask;
+
+ cpus_and(cleanup_mask, cfg->old_domain, cpu_online_map);
+ cfg->move_cleanup_count = cpus_weight(cleanup_mask);
+ send_IPI_mask(cleanup_mask, IRQ_MOVE_CLEANUP_VECTOR);
+ cfg->move_in_progress = 0;
+ }
+}
+#else
+static inline void irq_complete_move(unsigned int irq) {}
+#endif
+
static void ack_apic_edge(unsigned int irq)
{
+ irq_complete_move(irq);
move_native_irq(irq);
ack_APIC_irq();
}
@@ -1389,6 +1433,7 @@ static void ack_apic_level(unsigned int
{
int do_unmask_irq = 0;
+ irq_complete_move(irq);
#if defined(CONFIG_GENERIC_PENDING_IRQ) || defined(CONFIG_IRQBALANCE)
/* If we are moving the irq we need to mask it */
if (unlikely(irq_desc[irq].status & IRQ_MOVE_PENDING)) {
@@ -1440,7 +1485,7 @@ static inline void init_IO_APIC_traps(vo
*/
for (irq = 0; irq < NR_IRQS ; irq++) {
int tmp = irq;
- if (IO_APIC_IRQ(tmp) && !irq_vector[tmp]) {
+ if (IO_APIC_IRQ(tmp) && !irq_cfg[tmp].vector) {
/*
* Hmm.. We don't have an entry for this,
* so default to an old-fashioned 8259
@@ -1538,7 +1583,7 @@ static inline void unlock_ExtINT_logic(v
entry1.dest_mode = 0; /* physical delivery */
entry1.mask = 0; /* unmask IRQ now */
- entry1.dest.physical.physical_dest = hard_smp_processor_id();
+ entry1.dest = hard_smp_processor_id();
entry1.delivery_mode = dest_ExtINT;
entry1.polarity = entry0.polarity;
entry1.trigger = 0;
@@ -1582,15 +1627,14 @@ static inline void unlock_ExtINT_logic(v
*/
static inline void check_timer(void)
{
+ struct irq_cfg *cfg = irq_cfg + 0;
int apic1, pin1, apic2, pin2;
- int vector;
- cpumask_t mask;
/*
* get/set the timer IRQ vector:
*/
disable_8259A_irq(0);
- vector = assign_irq_vector(0, TARGET_CPUS, &mask);
+ assign_irq_vector(0, TARGET_CPUS);
/*
* Subtle, code in do_timer_interrupt() expects an AEOI
@@ -1610,7 +1654,7 @@ static inline void check_timer(void)
apic2 = ioapic_i8259.apic;
apic_printk(APIC_VERBOSE,KERN_INFO "..TIMER: vector=0x%02X apic1=%d pin1=%d apic2=%d pin2=%d\n",
- vector, apic1, pin1, apic2, pin2);
+ cfg->vector, apic1, pin1, apic2, pin2);
if (pin1 != -1) {
/*
@@ -1641,7 +1685,7 @@ static inline void check_timer(void)
/*
* legacy devices should be connected to IO APIC #0
*/
- setup_ExtINT_IRQ0_pin(apic2, pin2, vector);
+ setup_ExtINT_IRQ0_pin(apic2, pin2, cfg->vector);
if (timer_irq_works()) {
apic_printk(APIC_VERBOSE," works.\n");
nmi_watchdog_default();
@@ -1666,14 +1710,14 @@ static inline void check_timer(void)
disable_8259A_irq(0);
irq_desc[0].chip = &lapic_irq_type;
- apic_write(APIC_LVT0, APIC_DM_FIXED | vector); /* Fixed mode */
+ apic_write(APIC_LVT0, APIC_DM_FIXED | cfg->vector); /* Fixed mode */
enable_8259A_irq(0);
if (timer_irq_works()) {
apic_printk(APIC_VERBOSE," works.\n");
return;
}
- apic_write(APIC_LVT0, APIC_LVT_MASKED | APIC_DM_FIXED | vector);
+ apic_write(APIC_LVT0, APIC_LVT_MASKED | APIC_DM_FIXED | cfg->vector);
apic_printk(APIC_VERBOSE," failed.\n");
apic_printk(APIC_VERBOSE, KERN_INFO "...trying to set up timer as ExtINT IRQ...");
@@ -1828,19 +1872,16 @@ int create_irq(void)
/* Allocate an unused irq */
int irq;
int new;
- int vector = 0;
unsigned long flags;
- cpumask_t mask;
irq = -ENOSPC;
spin_lock_irqsave(&vector_lock, flags);
for (new = (NR_IRQS - 1); new >= 0; new--) {
if (platform_legacy_irq(new))
continue;
- if (irq_vector[new] != 0)
+ if (irq_cfg[new].vector != 0)
continue;
- vector = __assign_irq_vector(new, TARGET_CPUS, &mask);
- if (likely(vector > 0))
+ if (__assign_irq_vector(new, TARGET_CPUS) == 0)
irq = new;
break;
}
@@ -1871,12 +1912,15 @@ void destroy_irq(unsigned int irq)
#if defined(CONFIG_PCI_MSI) && !defined(CONFIG_XEN)
static int msi_compose_msg(struct pci_dev *pdev, unsigned int irq, struct msi_msg *msg)
{
- int vector;
+ struct irq_cfg *cfg = irq_cfg + irq;
+ int err;
unsigned dest;
cpumask_t tmp;
- vector = assign_irq_vector(irq, TARGET_CPUS, &tmp);
- if (vector >= 0) {
+ tmp = TARGET_CPUS;
+ err = assign_irq_vector(irq, tmp);
+ if (!err) {
+ cpus_and(tmp, cfg->domain, tmp);
dest = cpu_mask_to_apicid(tmp);
msg->address_hi = MSI_ADDR_BASE_HI;
@@ -1896,40 +1940,38 @@ static int msi_compose_msg(struct pci_de
((INT_DELIVERY_MODE != dest_LowestPrio) ?
MSI_DATA_DELIVERY_FIXED:
MSI_DATA_DELIVERY_LOWPRI) |
- MSI_DATA_VECTOR(vector);
+ MSI_DATA_VECTOR(cfg->vector);
}
- return vector;
+ return err;
}
#ifdef CONFIG_SMP
static void set_msi_irq_affinity(unsigned int irq, cpumask_t mask)
{
+ struct irq_cfg *cfg = irq_cfg + irq;
struct msi_msg msg;
unsigned int dest;
cpumask_t tmp;
- int vector;
cpus_and(tmp, mask, cpu_online_map);
if (cpus_empty(tmp))
- tmp = TARGET_CPUS;
-
- cpus_and(mask, tmp, CPU_MASK_ALL);
+ return;
- vector = assign_irq_vector(irq, mask, &tmp);
- if (vector < 0)
+ if (assign_irq_vector(irq, mask))
return;
+ cpus_and(tmp, cfg->domain, mask);
dest = cpu_mask_to_apicid(tmp);
read_msi_msg(irq, &msg);
msg.data &= ~MSI_DATA_VECTOR_MASK;
- msg.data |= MSI_DATA_VECTOR(vector);
+ msg.data |= MSI_DATA_VECTOR(cfg->vector);
msg.address_lo &= ~MSI_ADDR_DEST_ID_MASK;
msg.address_lo |= MSI_ADDR_DEST_ID(dest);
write_msi_msg(irq, &msg);
- set_native_irq_info(irq, mask);
+ irq_desc[irq].affinity = mask;
}
#endif /* CONFIG_SMP */
@@ -1948,24 +1990,31 @@ static struct irq_chip msi_chip = {
.retrigger = ioapic_retrigger_irq,
};
-int arch_setup_msi_irq(unsigned int irq, struct pci_dev *dev)
+int arch_setup_msi_irq(struct pci_dev *dev, struct msi_desc *desc)
{
struct msi_msg msg;
- int ret;
+ int irq, ret;
+ irq = create_irq();
+ if (irq < 0)
+ return irq;
+
+ set_irq_msi(irq, desc);
ret = msi_compose_msg(dev, irq, &msg);
- if (ret < 0)
+ if (ret < 0) {
+ destroy_irq(irq);
return ret;
+ }
write_msi_msg(irq, &msg);
set_irq_chip_and_handler_name(irq, &msi_chip, handle_edge_irq, "edge");
- return 0;
+ return irq;
}
void arch_teardown_msi_irq(unsigned int irq)
{
- return;
+ destroy_irq(irq);
}
#endif /* CONFIG_PCI_MSI */
@@ -1993,24 +2042,22 @@ static void target_ht_irq(unsigned int i
static void set_ht_irq_affinity(unsigned int irq, cpumask_t mask)
{
+ struct irq_cfg *cfg = irq_cfg + irq;
unsigned int dest;
cpumask_t tmp;
- int vector;
cpus_and(tmp, mask, cpu_online_map);
if (cpus_empty(tmp))
- tmp = TARGET_CPUS;
-
- cpus_and(mask, tmp, CPU_MASK_ALL);
+ return;
- vector = assign_irq_vector(irq, mask, &tmp);
- if (vector < 0)
+ if (assign_irq_vector(irq, mask))
return;
+ cpus_and(tmp, cfg->domain, mask);
dest = cpu_mask_to_apicid(tmp);
- target_ht_irq(irq, dest, vector);
- set_native_irq_info(irq, mask);
+ target_ht_irq(irq, dest, cfg->vector);
+ irq_desc[irq].affinity = mask;
}
#endif
@@ -2027,14 +2074,17 @@ static struct irq_chip ht_irq_chip = {
int arch_setup_ht_irq(unsigned int irq, struct pci_dev *dev)
{
- int vector;
+ struct irq_cfg *cfg = irq_cfg + irq;
+ int err;
cpumask_t tmp;
- vector = assign_irq_vector(irq, TARGET_CPUS, &tmp);
- if (vector >= 0) {
+ tmp = TARGET_CPUS;
+ err = assign_irq_vector(irq, tmp);
+ if (!err) {
struct ht_irq_msg msg;
unsigned dest;
+ cpus_and(tmp, cfg->domain, tmp);
dest = cpu_mask_to_apicid(tmp);
msg.address_hi = HT_IRQ_HIGH_DEST_ID(dest);
@@ -2042,7 +2092,7 @@ int arch_setup_ht_irq(unsigned int irq,
msg.address_lo =
HT_IRQ_LOW_BASE |
HT_IRQ_LOW_DEST_ID(dest) |
- HT_IRQ_LOW_VECTOR(vector) |
+ HT_IRQ_LOW_VECTOR(cfg->vector) |
((INT_DEST_MODE == 0) ?
HT_IRQ_LOW_DM_PHYSICAL :
HT_IRQ_LOW_DM_LOGICAL) |
@@ -2057,7 +2107,7 @@ int arch_setup_ht_irq(unsigned int irq,
set_irq_chip_and_handler_name(irq, &ht_irq_chip,
handle_edge_irq, "edge");
}
- return vector;
+ return err;
}
#endif /* CONFIG_HT_IRQ */
@@ -2082,13 +2132,8 @@ int __init io_apic_get_redir_entries (in
}
-int io_apic_set_pci_routing (int ioapic, int pin, int irq, int edge_level, int active_high_low)
+int io_apic_set_pci_routing (int ioapic, int pin, int irq, int triggering, int polarity)
{
- struct IO_APIC_route_entry entry;
- unsigned long flags;
- int vector;
- cpumask_t mask;
-
if (!IO_APIC_IRQ(irq)) {
apic_printk(APIC_QUIET,KERN_ERR "IOAPIC[%d]: Invalid reference to IRQ 0\n",
ioapic);
@@ -2101,42 +2146,7 @@ int io_apic_set_pci_routing (int ioapic,
if (irq >= 16)
add_pin_to_irq(irq, ioapic, pin);
-
- vector = assign_irq_vector(irq, TARGET_CPUS, &mask);
- if (vector < 0)
- return vector;
-
- /*
- * Generate a PCI IRQ routing entry and program the IOAPIC accordingly.
- * Note that we mask (disable) IRQs now -- these get enabled when the
- * corresponding device driver registers for this IRQ.
- */
-
- memset(&entry,0,sizeof(entry));
-
- entry.delivery_mode = INT_DELIVERY_MODE;
- entry.dest_mode = INT_DEST_MODE;
- entry.dest.logical.logical_dest = cpu_mask_to_apicid(mask);
- entry.trigger = edge_level;
- entry.polarity = active_high_low;
- entry.mask = 1; /* Disabled (masked) */
- entry.vector = vector & 0xff;
-
- apic_printk(APIC_VERBOSE,KERN_DEBUG "IOAPIC[%d]: Set PCI routing entry (%d-%d -> 0x%x -> "
- "IRQ %d Mode:%i Active:%i)\n", ioapic,
- mp_ioapics[ioapic].mpc_apicid, pin, entry.vector, irq,
- edge_level, active_high_low);
-
- ioapic_register_intr(irq, entry.vector, edge_level);
-
- if (!ioapic && (irq < 16))
- disable_8259A_irq(irq);
-
- ioapic_write_entry(ioapic, pin, entry);
-
- spin_lock_irqsave(&ioapic_lock, flags);
- set_native_irq_info(irq, TARGET_CPUS);
- spin_unlock_irqrestore(&ioapic_lock, flags);
+ setup_IO_APIC_irq(ioapic, pin, irq, triggering, polarity);
return 0;
}
@@ -2169,8 +2179,10 @@ void __init setup_ioapic_dest(void)
* when you have too many devices, because at that time only boot
* cpu is online.
*/
- if(!irq_vector[irq])
- setup_IO_APIC_irq(ioapic, pin, irq_entry, irq);
+ if (!irq_cfg[irq].vector)
+ setup_IO_APIC_irq(ioapic, pin, irq,
+ irq_trigger(irq_entry),
+ irq_polarity(irq_entry));
else
set_ioapic_affinity_irq(irq, TARGET_CPUS);
}
--- head.orig/arch/x86/kernel/mpparse_64-xen.c 2011-01-31 17:32:16.000000000 +0100
+++ head/arch/x86/kernel/mpparse_64-xen.c 2011-01-31 17:32:22.000000000 +0100
@@ -60,9 +60,9 @@ unsigned long mp_lapic_addr = 0;
/* Processor that is doing the boot up */
unsigned int boot_cpu_id = -1U;
/* Internal processor count */
-unsigned int num_processors __initdata = 0;
+unsigned int num_processors __cpuinitdata = 0;
-unsigned disabled_cpus __initdata;
+unsigned disabled_cpus __cpuinitdata;
/* Bitmask of physically existing CPUs */
physid_mask_t phys_cpu_present_map = PHYSID_MASK_NONE;
@@ -808,7 +808,7 @@ int mp_register_gsi(u32 gsi, int trigger
return gsi;
/* Don't set up the ACPI SCI because it's already set up */
- if (acpi_fadt.sci_int == gsi)
+ if (acpi_gbl_FADT.sci_interrupt == gsi)
return gsi;
ioapic = mp_find_ioapic(gsi);
--- head.orig/arch/x86/kernel/process_64-xen.c 2011-02-02 08:30:59.000000000 +0100
+++ head/arch/x86/kernel/process_64-xen.c 2011-02-02 08:31:05.000000000 +0100
@@ -337,14 +337,17 @@ void load_gs_index(unsigned gs)
void flush_thread(void)
{
struct task_struct *tsk = current;
- struct thread_info *t = current_thread_info();
- if (t->flags & _TIF_ABI_PENDING) {
- t->flags ^= (_TIF_ABI_PENDING | _TIF_IA32);
- if (t->flags & _TIF_IA32)
+ if (test_tsk_thread_flag(tsk, TIF_ABI_PENDING)) {
+ clear_tsk_thread_flag(tsk, TIF_ABI_PENDING);
+ if (test_tsk_thread_flag(tsk, TIF_IA32)) {
+ clear_tsk_thread_flag(tsk, TIF_IA32);
+ } else {
+ set_tsk_thread_flag(tsk, TIF_IA32);
current_thread_info()->status |= TS_COMPAT;
+ }
}
- t->flags &= ~_TIF_DEBUG;
+ clear_tsk_thread_flag(tsk, TIF_DEBUG);
tsk->thread.debugreg0 = 0;
tsk->thread.debugreg1 = 0;
--- head.orig/arch/x86/kernel/setup_64-xen.c 2012-06-06 13:48:44.000000000 +0200
+++ head/arch/x86/kernel/setup_64-xen.c 2012-06-06 13:49:08.000000000 +0200
@@ -141,7 +141,7 @@ EXPORT_SYMBOL_GPL(edid_info);
extern int root_mountflags;
-char command_line[COMMAND_LINE_SIZE];
+char __initdata command_line[COMMAND_LINE_SIZE];
struct resource standard_io_resources[] = {
{ .name = "dma1", .start = 0x00, .end = 0x1f,
@@ -179,134 +179,6 @@ struct resource code_resource = {
.flags = IORESOURCE_RAM,
};
-#define IORESOURCE_ROM (IORESOURCE_BUSY | IORESOURCE_READONLY | IORESOURCE_MEM)
-
-static struct resource system_rom_resource = {
- .name = "System ROM",
- .start = 0xf0000,
- .end = 0xfffff,
- .flags = IORESOURCE_ROM,
-};
-
-static struct resource extension_rom_resource = {
- .name = "Extension ROM",
- .start = 0xe0000,
- .end = 0xeffff,
- .flags = IORESOURCE_ROM,
-};
-
-static struct resource adapter_rom_resources[] = {
- { .name = "Adapter ROM", .start = 0xc8000, .end = 0,
- .flags = IORESOURCE_ROM },
- { .name = "Adapter ROM", .start = 0, .end = 0,
- .flags = IORESOURCE_ROM },
- { .name = "Adapter ROM", .start = 0, .end = 0,
- .flags = IORESOURCE_ROM },
- { .name = "Adapter ROM", .start = 0, .end = 0,
- .flags = IORESOURCE_ROM },
- { .name = "Adapter ROM", .start = 0, .end = 0,
- .flags = IORESOURCE_ROM },
- { .name = "Adapter ROM", .start = 0, .end = 0,
- .flags = IORESOURCE_ROM }
-};
-
-static struct resource video_rom_resource = {
- .name = "Video ROM",
- .start = 0xc0000,
- .end = 0xc7fff,
- .flags = IORESOURCE_ROM,
-};
-
-static struct resource video_ram_resource = {
- .name = "Video RAM area",
- .start = 0xa0000,
- .end = 0xbffff,
- .flags = IORESOURCE_RAM,
-};
-
-#define romsignature(x) (*(unsigned short *)(x) == 0xaa55)
-
-static int __init romchecksum(unsigned char *rom, unsigned long length)
-{
- unsigned char *p, sum = 0;
-
- for (p = rom; p < rom + length; p++)
- sum += *p;
- return sum == 0;
-}
-
-static void __init probe_roms(void)
-{
- unsigned long start, length, upper;
- unsigned char *rom;
- int i;
-
-#ifdef CONFIG_XEN
- /* Nothing to do if not running in dom0. */
- if (!is_initial_xendomain())
- return;
-#endif
-
- /* video rom */
- upper = adapter_rom_resources[0].start;
- for (start = video_rom_resource.start; start < upper; start += 2048) {
- rom = isa_bus_to_virt(start);
- if (!romsignature(rom))
- continue;
-
- video_rom_resource.start = start;
-
- /* 0 < length <= 0x7f * 512, historically */
- length = rom[2] * 512;
-
- /* if checksum okay, trust length byte */
- if (length && romchecksum(rom, length))
- video_rom_resource.end = start + length - 1;
-
- request_resource(&iomem_resource, &video_rom_resource);
- break;
- }
-
- start = (video_rom_resource.end + 1 + 2047) & ~2047UL;
- if (start < upper)
- start = upper;
-
- /* system rom */
- request_resource(&iomem_resource, &system_rom_resource);
- upper = system_rom_resource.start;
-
- /* check for extension rom (ignore length byte!) */
- rom = isa_bus_to_virt(extension_rom_resource.start);
- if (romsignature(rom)) {
- length = extension_rom_resource.end - extension_rom_resource.start + 1;
- if (romchecksum(rom, length)) {
- request_resource(&iomem_resource, &extension_rom_resource);
- upper = extension_rom_resource.start;
- }
- }
-
- /* check for adapter roms on 2k boundaries */
- for (i = 0; i < ARRAY_SIZE(adapter_rom_resources) && start < upper;
- start += 2048) {
- rom = isa_bus_to_virt(start);
- if (!romsignature(rom))
- continue;
-
- /* 0 < length <= 0x7f * 512, historically */
- length = rom[2] * 512;
-
- /* but accept any length that fits if checksum okay */
- if (!length || start + length > upper || !romchecksum(rom, length))
- continue;
-
- adapter_rom_resources[i].start = start;
- adapter_rom_resources[i].end = start + length - 1;
- request_resource(&iomem_resource, &adapter_rom_resources[i]);
-
- start = adapter_rom_resources[i++].end & ~2047UL;
- }
-}
-
#ifdef CONFIG_PROC_VMCORE
/* elfcorehdr= specifies the location of elf core header
* stored by the crashed kernel. This option will be passed
@@ -404,7 +276,7 @@ void __init setup_arch(char **cmdline_p)
#ifdef CONFIG_XEN
extern struct e820map machine_e820;
- printk(KERN_INFO "Command line: %s\n", saved_command_line);
+ printk(KERN_INFO "Command line: %s\n", boot_command_line);
/* Register a call for panic conditions. */
atomic_notifier_chain_register(&panic_notifier_list, &xen_panic_block);
@@ -431,7 +303,7 @@ void __init setup_arch(char **cmdline_p)
ARCH_SETUP
#else
- printk(KERN_INFO "Command line: %s\n", saved_command_line);
+ printk(KERN_INFO "Command line: %s\n", boot_command_line);
ROOT_DEV = old_decode_dev(ORIG_ROOT_DEV);
screen_info = SCREEN_INFO;
@@ -462,7 +334,7 @@ void __init setup_arch(char **cmdline_p)
early_identify_cpu(&boot_cpu_data);
- strlcpy(command_line, saved_command_line, COMMAND_LINE_SIZE);
+ strlcpy(command_line, boot_command_line, COMMAND_LINE_SIZE);
*cmdline_p = command_line;
parse_early_param();
@@ -532,6 +404,11 @@ void __init setup_arch(char **cmdline_p)
/* reserve ebda region */
if (ebda_addr)
reserve_bootmem_generic(ebda_addr, ebda_size);
+#ifdef CONFIG_NUMA
+ /* reserve nodemap region */
+ if (nodemap_addr)
+ reserve_bootmem_generic(nodemap_addr, nodemap_size);
+#endif
#ifdef CONFIG_SMP
/*
@@ -730,10 +607,8 @@ void __init setup_arch(char **cmdline_p)
#endif
/*
- * Request address space for all standard RAM and ROM resources
- * and also for regions reported as reserved by the e820.
+ * We trust e820 completely. No explicit ROM probing in memory.
*/
- probe_roms();
#ifdef CONFIG_XEN
if (is_initial_xendomain())
e820_reserve_resources(machine_e820.map, machine_e820.nr_map);
@@ -742,8 +617,6 @@ void __init setup_arch(char **cmdline_p)
e820_mark_nosave_regions();
#endif
- request_resource(&iomem_resource, &video_ram_resource);
-
{
unsigned i;
/* request I/O space for devices used on all i[345]86 PCs */
@@ -1322,7 +1195,8 @@ static int show_cpuinfo(struct seq_file
NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL,
NULL, NULL, NULL, "syscall", NULL, NULL, NULL, NULL,
NULL, NULL, NULL, NULL, "nx", NULL, "mmxext", NULL,
- NULL, "fxsr_opt", NULL, "rdtscp", NULL, "lm", "3dnowext", "3dnow",
+ NULL, "fxsr_opt", "pdpe1gb", "rdtscp", NULL, "lm",
+ "3dnowext", "3dnow",
/* Transmeta-defined */
"recovery", "longrun", NULL, "lrti", NULL, NULL, NULL, NULL,
@@ -1340,7 +1214,7 @@ static int show_cpuinfo(struct seq_file
/* Intel-defined (#2) */
"pni", NULL, NULL, "monitor", "ds_cpl", "vmx", "smx", "est",
"tm2", "ssse3", "cid", NULL, NULL, "cx16", "xtpr", NULL,
- NULL, NULL, "dca", NULL, NULL, NULL, NULL, NULL,
+ NULL, NULL, "dca", NULL, NULL, NULL, NULL, "popcnt",
NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL,
/* VIA/Cyrix/Centaur-defined */
@@ -1350,8 +1224,10 @@ static int show_cpuinfo(struct seq_file
NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL,
/* AMD-defined (#2) */
- "lahf_lm", "cmp_legacy", "svm", NULL, "cr8_legacy", NULL, NULL, NULL,
- NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL,
+ "lahf_lm", "cmp_legacy", "svm", "extapic", "cr8_legacy",
+ "altmovcr8", "abm", "sse4a",
+ "misalignsse", "3dnowprefetch",
+ "osvw", "ibs", NULL, NULL, NULL, NULL,
NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL,
NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL,
};
@@ -1362,6 +1238,9 @@ static int show_cpuinfo(struct seq_file
"ttp", /* thermal trip */
"tm",
"stc",
+ "100mhzsteps",
+ "hwpstate",
+ NULL, /* tsc invariant mapped to constant_tsc */
NULL,
/* nothing */ /* constant_tsc - moved to flags */
};
@@ -1478,26 +1357,3 @@ struct seq_operations cpuinfo_op = {
.stop = c_stop,
.show = show_cpuinfo,
};
-
-#if defined(CONFIG_INPUT_PCSPKR) || defined(CONFIG_INPUT_PCSPKR_MODULE)
-#include <linux/platform_device.h>
-static __init int add_pcspkr(void)
-{
- struct platform_device *pd;
- int ret;
-
- if (!is_initial_xendomain())
- return 0;
-
- pd = platform_device_alloc("pcspkr", -1);
- if (!pd)
- return -ENOMEM;
-
- ret = platform_device_add(pd);
- if (ret)
- platform_device_put(pd);
-
- return ret;
-}
-device_initcall(add_pcspkr);
-#endif
--- head.orig/arch/x86/kernel/vsyscall_64-xen.c 2011-01-31 17:32:16.000000000 +0100
+++ head/arch/x86/kernel/vsyscall_64-xen.c 2011-01-31 17:32:22.000000000 +0100
@@ -26,6 +26,7 @@
#include <linux/seqlock.h>
#include <linux/jiffies.h>
#include <linux/sysctl.h>
+#include <linux/clocksource.h>
#include <linux/getcpu.h>
#include <linux/cpu.h>
#include <linux/smp.h>
@@ -34,6 +35,7 @@
#include <asm/vsyscall.h>
#include <asm/pgtable.h>
#include <asm/page.h>
+#include <asm/unistd.h>
#include <asm/fixmap.h>
#include <asm/errno.h>
#include <asm/io.h>
@@ -44,56 +46,41 @@
#define __vsyscall(nr) __attribute__ ((unused,__section__(".vsyscall_" #nr)))
#define __syscall_clobber "r11","rcx","memory"
-int __sysctl_vsyscall __section_sysctl_vsyscall = 1;
-seqlock_t __xtime_lock __section_xtime_lock = SEQLOCK_UNLOCKED;
+struct vsyscall_gtod_data_t {
+ seqlock_t lock;
+ int sysctl_enabled;
+ struct timeval wall_time_tv;
+ struct timezone sys_tz;
+ cycle_t offset_base;
+ struct clocksource clock;
+};
int __vgetcpu_mode __section_vgetcpu_mode;
-#include <asm/unistd.h>
-
-static __always_inline void timeval_normalize(struct timeval * tv)
+struct vsyscall_gtod_data_t __vsyscall_gtod_data __section_vsyscall_gtod_data =
{
- time_t __sec;
-
- __sec = tv->tv_usec / 1000000;
- if (__sec) {
- tv->tv_usec %= 1000000;
- tv->tv_sec += __sec;
- }
-}
+ .lock = SEQLOCK_UNLOCKED,
+ .sysctl_enabled = 1,
+};
-static __always_inline void do_vgettimeofday(struct timeval * tv)
+void update_vsyscall(struct timespec *wall_time, struct clocksource *clock)
{
- long sequence, t;
- unsigned long sec, usec;
+ unsigned long flags;
- do {
- sequence = read_seqbegin(&__xtime_lock);
-
- sec = __xtime.tv_sec;
- usec = __xtime.tv_nsec / 1000;
-
- if (__vxtime.mode != VXTIME_HPET) {
- t = get_cycles_sync();
- if (t < __vxtime.last_tsc)
- t = __vxtime.last_tsc;
- usec += ((t - __vxtime.last_tsc) *
- __vxtime.tsc_quot) >> 32;
- /* See comment in x86_64 do_gettimeofday. */
- } else {
- usec += ((readl((void __iomem *)
- fix_to_virt(VSYSCALL_HPET) + 0xf0) -
- __vxtime.last) * __vxtime.quot) >> 32;
- }
- } while (read_seqretry(&__xtime_lock, sequence));
-
- tv->tv_sec = sec + usec / 1000000;
- tv->tv_usec = usec % 1000000;
+ write_seqlock_irqsave(&vsyscall_gtod_data.lock, flags);
+ /* copy vsyscall data */
+ vsyscall_gtod_data.clock = *clock;
+ vsyscall_gtod_data.wall_time_tv.tv_sec = wall_time->tv_sec;
+ vsyscall_gtod_data.wall_time_tv.tv_usec = wall_time->tv_nsec/1000;
+ vsyscall_gtod_data.sys_tz = sys_tz;
+ write_sequnlock_irqrestore(&vsyscall_gtod_data.lock, flags);
}
-/* RED-PEN may want to readd seq locking, but then the variable should be write-once. */
+/* RED-PEN may want to readd seq locking, but then the variable should be
+ * write-once.
+ */
static __always_inline void do_get_tz(struct timezone * tz)
{
- *tz = __sys_tz;
+ *tz = __vsyscall_gtod_data.sys_tz;
}
static __always_inline int gettimeofday(struct timeval *tv, struct timezone *tz)
@@ -101,7 +88,8 @@ static __always_inline int gettimeofday(
int ret;
asm volatile("vsysc2: syscall"
: "=a" (ret)
- : "0" (__NR_gettimeofday),"D" (tv),"S" (tz) : __syscall_clobber );
+ : "0" (__NR_gettimeofday),"D" (tv),"S" (tz)
+ : __syscall_clobber );
return ret;
}
@@ -114,10 +102,44 @@ static __always_inline long time_syscall
return secs;
}
+static __always_inline void do_vgettimeofday(struct timeval * tv)
+{
+ cycle_t now, base, mask, cycle_delta;
+ unsigned long seq, mult, shift, nsec_delta;
+ cycle_t (*vread)(void);
+ do {
+ seq = read_seqbegin(&__vsyscall_gtod_data.lock);
+
+ vread = __vsyscall_gtod_data.clock.vread;
+ if (unlikely(!__vsyscall_gtod_data.sysctl_enabled || !vread)) {
+ gettimeofday(tv,NULL);
+ return;
+ }
+ now = vread();
+ base = __vsyscall_gtod_data.clock.cycle_last;
+ mask = __vsyscall_gtod_data.clock.mask;
+ mult = __vsyscall_gtod_data.clock.mult;
+ shift = __vsyscall_gtod_data.clock.shift;
+
+ *tv = __vsyscall_gtod_data.wall_time_tv;
+
+ } while (read_seqretry(&__vsyscall_gtod_data.lock, seq));
+
+ /* calculate interval: */
+ cycle_delta = (now - base) & mask;
+ /* convert to nsecs: */
+ nsec_delta = (cycle_delta * mult) >> shift;
+
+ /* convert to usecs and add to timespec: */
+ tv->tv_usec += nsec_delta / NSEC_PER_USEC;
+ while (tv->tv_usec > USEC_PER_SEC) {
+ tv->tv_sec += 1;
+ tv->tv_usec -= USEC_PER_SEC;
+ }
+}
+
int __vsyscall(0) vgettimeofday(struct timeval * tv, struct timezone * tz)
{
- if (!__sysctl_vsyscall)
- return gettimeofday(tv,tz);
if (tv)
do_vgettimeofday(tv);
if (tz)
@@ -129,11 +151,11 @@ int __vsyscall(0) vgettimeofday(struct t
* unlikely */
time_t __vsyscall(1) vtime(time_t *t)
{
- if (!__sysctl_vsyscall)
+ if (unlikely(!__vsyscall_gtod_data.sysctl_enabled))
return time_syscall(t);
else if (t)
- *t = __xtime.tv_sec;
- return __xtime.tv_sec;
+ *t = __vsyscall_gtod_data.wall_time_tv.tv_sec;
+ return __vsyscall_gtod_data.wall_time_tv.tv_sec;
}
/* Fast way to get current CPU and node.
@@ -210,7 +232,7 @@ static int vsyscall_sysctl_change(ctl_ta
ret = -ENOMEM;
goto out;
}
- if (!sysctl_vsyscall) {
+ if (!vsyscall_gtod_data.sysctl_enabled) {
writew(SYSCALL, map1);
writew(SYSCALL, map2);
} else {
@@ -232,16 +254,17 @@ static int vsyscall_sysctl_nostrat(ctl_t
static ctl_table kernel_table2[] = {
{ .ctl_name = 99, .procname = "vsyscall64",
- .data = &sysctl_vsyscall, .maxlen = sizeof(int), .mode = 0644,
+ .data = &vsyscall_gtod_data.sysctl_enabled, .maxlen = sizeof(int),
+ .mode = 0644,
.strategy = vsyscall_sysctl_nostrat,
.proc_handler = vsyscall_sysctl_change },
- { 0, }
+ {}
};
static ctl_table kernel_root_table2[] = {
{ .ctl_name = CTL_KERN, .procname = "kernel", .mode = 0555,
.child = kernel_table2 },
- { 0 },
+ {}
};
#endif
@@ -304,14 +327,14 @@ static int __init vsyscall_init(void)
BUG_ON((unsigned long) &vgetcpu != VSYSCALL_ADDR(__NR_vgetcpu));
map_vsyscall();
#ifdef CONFIG_XEN
- sysctl_vsyscall = 0; /* disable vgettimeofay() */
+ vsyscall_gtod_data.sysctl_enabled = 0; /* disable vgettimeofay() */
if (boot_cpu_has(X86_FEATURE_RDTSCP))
vgetcpu_mode = VGETCPU_RDTSCP;
else
vgetcpu_mode = VGETCPU_LSL;
#endif
#ifdef CONFIG_SYSCTL
- register_sysctl_table(kernel_root_table2, 0);
+ register_sysctl_table(kernel_root_table2);
#endif
on_each_cpu(cpu_vsyscall_init, NULL, 0, 1);
hotcpu_notifier(cpu_vsyscall_notifier, 0);
--- head.orig/arch/x86/mm/fault_64-xen.c 2011-07-26 09:27:24.000000000 +0200
+++ head/arch/x86/mm/fault_64-xen.c 2011-07-26 09:27:34.000000000 +0200
@@ -56,38 +56,17 @@ int unregister_page_fault_notifier(struc
}
EXPORT_SYMBOL_GPL(unregister_page_fault_notifier);
-static inline int notify_page_fault(enum die_val val, const char *str,
- struct pt_regs *regs, long err, int trap, int sig)
+static inline int notify_page_fault(struct pt_regs *regs, long err)
{
struct die_args args = {
.regs = regs,
- .str = str,
+ .str = "page fault",
.err = err,
- .trapnr = trap,
- .signr = sig
+ .trapnr = 14,
+ .signr = SIGSEGV
};
- return atomic_notifier_call_chain(¬ify_page_fault_chain, val, &args);
-}
-
-void bust_spinlocks(int yes)
-{
- int loglevel_save = console_loglevel;
- if (yes) {
- oops_in_progress = 1;
- } else {
-#ifdef CONFIG_VT
- unblank_screen();
-#endif
- oops_in_progress = 0;
- /*
- * OK, the message is on the console. Now we call printk()
- * without oops_in_progress set so that printk will give klogd
- * a poke. Hold onto your hats...
- */
- console_loglevel = 15; /* NMI oopser may have shut the console up */
- printk(" ");
- console_loglevel = loglevel_save;
- }
+ return atomic_notifier_call_chain(¬ify_page_fault_chain,
+ DIE_PAGE_FAULT, &args);
}
/* Sometimes the CPU reports invalid exceptions on prefetch.
@@ -438,8 +417,7 @@ asmlinkage void __kprobes do_page_fault(
/* Can take a spurious fault if mapping changes R/O -> R/W. */
if (spurious_fault(regs, address, error_code))
return;
- if (notify_page_fault(DIE_PAGE_FAULT, "page fault", regs, error_code, 14,
- SIGSEGV) == NOTIFY_STOP)
+ if (notify_page_fault(regs, error_code) == NOTIFY_STOP)
return;
/*
* Don't take the mm semaphore here. If we fixup a prefetch
@@ -448,8 +426,7 @@ asmlinkage void __kprobes do_page_fault(
goto bad_area_nosemaphore;
}
- if (notify_page_fault(DIE_PAGE_FAULT, "page fault", regs, error_code, 14,
- SIGSEGV) == NOTIFY_STOP)
+ if (notify_page_fault(regs, error_code) == NOTIFY_STOP)
return;
if (likely(regs->eflags & X86_EFLAGS_IF))
--- head.orig/arch/x86/mm/init_64-xen.c 2011-01-31 17:32:16.000000000 +0100
+++ head/arch/x86/mm/init_64-xen.c 2011-01-31 17:32:22.000000000 +0100
@@ -1145,20 +1145,30 @@ int kern_addr_valid(unsigned long addr)
extern int exception_trace, page_fault_trace;
static ctl_table debug_table2[] = {
- { 99, "exception-trace", &exception_trace, sizeof(int), 0644, NULL,
- proc_dointvec },
- { 0, }
+ {
+ .ctl_name = 99,
+ .procname = "exception-trace",
+ .data = &exception_trace,
+ .maxlen = sizeof(int),
+ .mode = 0644,
+ .proc_handler = proc_dointvec
+ },
+ {}
};
static ctl_table debug_root_table2[] = {
- { .ctl_name = CTL_DEBUG, .procname = "debug", .mode = 0555,
- .child = debug_table2 },
- { 0 },
+ {
+ .ctl_name = CTL_DEBUG,
+ .procname = "debug",
+ .mode = 0555,
+ .child = debug_table2
+ },
+ {}
};
static __init int x8664_sysctl_init(void)
{
- register_sysctl_table(debug_root_table2, 1);
+ register_sysctl_table(debug_root_table2);
return 0;
}
__initcall(x8664_sysctl_init);
--- head.orig/arch/x86/mm/pageattr_64-xen.c 2011-01-31 17:32:16.000000000 +0100
+++ head/arch/x86/mm/pageattr_64-xen.c 2011-01-31 17:32:22.000000000 +0100
@@ -350,8 +350,8 @@ static void flush_kernel_map(void *arg)
void *adr = page_address(pg);
if (cpu_has_clflush)
cache_flush_page(adr);
- __flush_tlb_one(adr);
}
+ __flush_tlb_all();
}
static inline void flush_map(struct list_head *l)
@@ -376,6 +376,7 @@ static void revert_page(unsigned long ad
pud_t *pud;
pmd_t *pmd;
pte_t large_pte;
+ unsigned long pfn;
pgd = pgd_offset_k(address);
BUG_ON(pgd_none(*pgd));
@@ -383,7 +384,8 @@ static void revert_page(unsigned long ad
BUG_ON(pud_none(*pud));
pmd = pmd_offset(pud, address);
BUG_ON(__pmd_val(*pmd) & _PAGE_PSE);
- large_pte = mk_pte_phys(__pa(address) & LARGE_PAGE_MASK, ref_prot);
+ pfn = (__pa(address) & LARGE_PAGE_MASK) >> PAGE_SHIFT;
+ large_pte = pfn_pte(pfn, ref_prot);
large_pte = pte_mkhuge(large_pte);
set_pte((pte_t *)pmd, large_pte);
}
--- head.orig/drivers/acpi/processor_extcntl.c 2011-01-31 14:53:38.000000000 +0100
+++ head/drivers/acpi/processor_extcntl.c 2011-01-31 17:32:22.000000000 +0100
@@ -32,9 +32,8 @@
#define ACPI_PROCESSOR_COMPONENT 0x01000000
#define ACPI_PROCESSOR_CLASS "processor"
-#define ACPI_PROCESSOR_DRIVER_NAME "ACPI Processor Driver"
#define _COMPONENT ACPI_PROCESSOR_COMPONENT
-ACPI_MODULE_NAME("acpi_processor")
+ACPI_MODULE_NAME("processor_extcntl")
static int processor_extcntl_parse_csd(struct acpi_processor *pr);
static int processor_extcntl_get_performance(struct acpi_processor *pr);
@@ -56,24 +55,17 @@ static int processor_notify_smm(void)
return 0;
/* Can't write pstate_cnt to smi_cmd if either value is zero */
- if ((!acpi_fadt.smi_cmd) || (!acpi_fadt.pstate_cnt)) {
+ if (!acpi_gbl_FADT.smi_command || !acpi_gbl_FADT.pstate_control) {
ACPI_DEBUG_PRINT((ACPI_DB_INFO,"No SMI port or pstate_cnt\n"));
return 0;
}
ACPI_DEBUG_PRINT((ACPI_DB_INFO,
"Writing pstate_cnt [0x%x] to smi_cmd [0x%x]\n",
- acpi_fadt.pstate_cnt, acpi_fadt.smi_cmd));
+ acpi_gbl_FADT.pstate_control, acpi_gbl_FADT.smi_command));
- /* FADT v1 doesn't support pstate_cnt, many BIOS vendors use
- * it anyway, so we need to support it... */
- if (acpi_fadt_is_v1) {
- ACPI_DEBUG_PRINT((ACPI_DB_INFO,
- "Using v1.0 FADT reserved value for pstate_cnt\n"));
- }
-
- status = acpi_os_write_port(acpi_fadt.smi_cmd,
- (u32) acpi_fadt.pstate_cnt, 8);
+ status = acpi_os_write_port(acpi_gbl_FADT.smi_command,
+ acpi_gbl_FADT.pstate_control, 8);
if (ACPI_FAILURE(status))
return status;
--- head.orig/drivers/pci/msi-xen.c 2011-10-07 11:18:20.000000000 +0200
+++ head/drivers/pci/msi-xen.c 2011-11-03 12:03:26.000000000 +0100
@@ -41,8 +41,6 @@ struct msi_dev_list {
struct list_head list;
spinlock_t pirq_list_lock;
struct list_head pirq_list_head;
- /* Used for saving/restoring MSI-X tables */
- void __iomem *mask_base;
/* Store default pre-assigned irq */
unsigned int default_irq;
};
@@ -51,14 +49,38 @@ struct msi_pirq_entry {
struct list_head list;
int pirq;
int entry_nr;
-#ifdef CONFIG_PM
- /* PM save area for MSIX address/data */
- u32 address_hi_save;
- u32 address_lo_save;
- u32 data_save;
-#endif
};
+static void msi_set_enable(struct pci_dev *dev, int enable)
+{
+ int pos;
+ u16 control;
+
+ pos = pci_find_capability(dev, PCI_CAP_ID_MSI);
+ if (pos) {
+ pci_read_config_word(dev, pos + PCI_MSI_FLAGS, &control);
+ control &= ~PCI_MSI_FLAGS_ENABLE;
+ if (enable)
+ control |= PCI_MSI_FLAGS_ENABLE;
+ pci_write_config_word(dev, pos + PCI_MSI_FLAGS, control);
+ }
+}
+
+static void msix_set_enable(struct pci_dev *dev, int enable)
+{
+ int pos;
+ u16 control;
+
+ pos = pci_find_capability(dev, PCI_CAP_ID_MSIX);
+ if (pos) {
+ pci_read_config_word(dev, pos + PCI_MSIX_FLAGS, &control);
+ control &= ~PCI_MSIX_FLAGS_ENABLE;
+ if (enable)
+ control |= PCI_MSIX_FLAGS_ENABLE;
+ pci_write_config_word(dev, pos + PCI_MSIX_FLAGS, control);
+ }
+}
+
static struct msi_dev_list *get_msi_dev_pirq_list(struct pci_dev *dev)
{
struct msi_dev_list *msi_dev_list, *ret = NULL;
@@ -168,8 +190,8 @@ static int msi_unmap_pirq(struct pci_dev
int rc;
unmap.domid = msi_get_dev_owner(dev);
- /* See comments in msi_map_pirq_to_vector, input parameter pirq
- * mean irq number only if the device belongs to dom0 itself.
+ /* See comments in msi_map_vector, input parameter pirq means
+ * irq number only if the device belongs to dom0 itself.
*/
unmap.pirq = (unmap.domid != DOMID_SELF)
? pirq : evtchn_get_xen_pirq(pirq);
@@ -205,8 +227,7 @@ static u64 find_table_base(struct pci_de
/*
* Protected by msi_lock
*/
-static int msi_map_pirq_to_vector(struct pci_dev *dev, int pirq,
- int entry_nr, u64 table_base)
+static int msi_map_vector(struct pci_dev *dev, int entry_nr, u64 table_base)
{
struct physdev_map_pirq map_irq;
int rc = -EINVAL;
@@ -217,7 +238,7 @@ static int msi_map_pirq_to_vector(struct
map_irq.domid = domid;
map_irq.type = MAP_PIRQ_TYPE_MSI_SEG;
map_irq.index = -1;
- map_irq.pirq = pirq < 0 ? -1 : evtchn_get_xen_pirq(pirq);
+ map_irq.pirq = -1;
map_irq.bus = dev->bus->number | (pci_domain_nr(dev->bus) << 16);
map_irq.devfn = dev->devfn;
map_irq.entry_nr = entry_nr;
@@ -241,7 +262,7 @@ static int msi_map_pirq_to_vector(struct
if (rc < 0)
return rc;
- /* This happens when MSI support is not enabled in Xen. */
+ /* This happens when MSI support is not enabled in older Xen. */
if (rc == 0 && map_irq.pirq < 0)
return -ENOSYS;
@@ -253,236 +274,52 @@ static int msi_map_pirq_to_vector(struct
* to another domain, and will be 'Linux irq' if it belongs to dom0.
*/
return ((domid != DOMID_SELF) ?
- map_irq.pirq : evtchn_map_pirq(pirq, map_irq.pirq));
-}
-
-static int msi_map_vector(struct pci_dev *dev, int entry_nr, u64 table_base)
-{
- return msi_map_pirq_to_vector(dev, -1, entry_nr, table_base);
+ map_irq.pirq : evtchn_map_pirq(-1, map_irq.pirq));
}
static int msi_init(void)
{
- static int status = 0;
-
- if (pci_msi_quirk) {
- pci_msi_enable = 0;
- printk(KERN_WARNING "PCI: MSI quirk detected. MSI disabled.\n");
- status = -EINVAL;
- }
-
- return status;
-}
-
-void pci_scan_msi_device(struct pci_dev *dev) { }
-
-void disable_msi_mode(struct pci_dev *dev, int pos, int type)
-{
- u16 control;
-
- pci_read_config_word(dev, msi_control_reg(pos), &control);
- if (type == PCI_CAP_ID_MSI) {
- /* Set enabled bits to single MSI & enable MSI_enable bit */
- msi_disable(control);
- pci_write_config_word(dev, msi_control_reg(pos), control);
- dev->msi_enabled = 0;
- } else {
- msix_disable(control);
- pci_write_config_word(dev, msi_control_reg(pos), control);
- dev->msix_enabled = 0;
- }
-
- pci_intx(dev, 1); /* enable intx */
-}
-
-static void enable_msi_mode(struct pci_dev *dev, int pos, int type)
-{
- u16 control;
-
- pci_read_config_word(dev, msi_control_reg(pos), &control);
- if (type == PCI_CAP_ID_MSI) {
- /* Set enabled bits to single MSI & enable MSI_enable bit */
- msi_enable(control, 1);
- pci_write_config_word(dev, msi_control_reg(pos), control);
- dev->msi_enabled = 1;
- } else {
- msix_enable(control);
- pci_write_config_word(dev, msi_control_reg(pos), control);
- dev->msix_enabled = 1;
- }
-
- pci_intx(dev, 0); /* disable intx */
-}
-
-#ifdef CONFIG_PM
-int pci_save_msi_state(struct pci_dev *dev)
-{
- int pos, i = 0;
- u16 control;
- struct pci_cap_saved_state *save_state;
- u32 *cap;
-
- pos = pci_find_capability(dev, PCI_CAP_ID_MSI);
- if (pos <= 0 || dev->no_msi)
- return 0;
-
- pci_read_config_word(dev, msi_control_reg(pos), &control);
- if (!(control & PCI_MSI_FLAGS_ENABLE))
- return 0;
-
- save_state = kzalloc(sizeof(struct pci_cap_saved_state) + sizeof(u32) * 5,
- GFP_KERNEL);
- if (!save_state) {
- printk(KERN_ERR "Out of memory in pci_save_msi_state\n");
- return -ENOMEM;
- }
- cap = &save_state->data[0];
-
- pci_read_config_dword(dev, pos, &cap[i++]);
- control = cap[0] >> 16;
- pci_read_config_dword(dev, pos + PCI_MSI_ADDRESS_LO, &cap[i++]);
- if (control & PCI_MSI_FLAGS_64BIT) {
- pci_read_config_dword(dev, pos + PCI_MSI_ADDRESS_HI, &cap[i++]);
- pci_read_config_dword(dev, pos + PCI_MSI_DATA_64, &cap[i++]);
- } else
- pci_read_config_dword(dev, pos + PCI_MSI_DATA_32, &cap[i++]);
- if (control & PCI_MSI_FLAGS_MASKBIT)
- pci_read_config_dword(dev, pos + PCI_MSI_MASK_BIT, &cap[i++]);
- save_state->cap_nr = PCI_CAP_ID_MSI;
- pci_add_saved_cap(dev, save_state);
return 0;
}
+#ifdef CONFIG_PM
void pci_restore_msi_state(struct pci_dev *dev)
{
- int i = 0, pos;
- u16 control;
- struct pci_cap_saved_state *save_state;
- u32 *cap;
+ int rc = -ENOSYS;
- save_state = pci_find_saved_cap(dev, PCI_CAP_ID_MSI);
- pos = pci_find_capability(dev, PCI_CAP_ID_MSI);
- if (!save_state || pos <= 0)
+ if (!dev->msi_enabled && !dev->msix_enabled)
return;
- cap = &save_state->data[0];
-
- control = cap[i++] >> 16;
- pci_write_config_dword(dev, pos + PCI_MSI_ADDRESS_LO, cap[i++]);
- if (control & PCI_MSI_FLAGS_64BIT) {
- pci_write_config_dword(dev, pos + PCI_MSI_ADDRESS_HI, cap[i++]);
- pci_write_config_dword(dev, pos + PCI_MSI_DATA_64, cap[i++]);
- } else
- pci_write_config_dword(dev, pos + PCI_MSI_DATA_32, cap[i++]);
- if (control & PCI_MSI_FLAGS_MASKBIT)
- pci_write_config_dword(dev, pos + PCI_MSI_MASK_BIT, cap[i++]);
- pci_write_config_word(dev, pos + PCI_MSI_FLAGS, control);
- enable_msi_mode(dev, pos, PCI_CAP_ID_MSI);
- pci_remove_saved_cap(save_state);
- kfree(save_state);
-}
-
-int pci_save_msix_state(struct pci_dev *dev)
-{
- int pos;
- u16 control;
- struct pci_cap_saved_state *save_state;
- unsigned long flags;
- struct msi_dev_list *msi_dev_entry;
- struct msi_pirq_entry *pirq_entry;
- void __iomem *base;
-
- pos = pci_find_capability(dev, PCI_CAP_ID_MSIX);
- if (pos <= 0 || dev->no_msi)
- return 0;
-
- /* save the capability */
- pci_read_config_word(dev, msi_control_reg(pos), &control);
- if (!(control & PCI_MSIX_FLAGS_ENABLE))
- return 0;
- msi_dev_entry = get_msi_dev_pirq_list(dev);
- /* If we failed to map the MSI-X table at pci_enable_msix,
- * We could not support saving them here.
- */
- if (!(base = msi_dev_entry->mask_base))
- return -ENOMEM;
-
- save_state = kzalloc(sizeof(struct pci_cap_saved_state) + sizeof(u16),
- GFP_KERNEL);
- if (!save_state) {
- printk(KERN_ERR "Out of memory in pci_save_msix_state\n");
- return -ENOMEM;
- }
- *((u16 *)&save_state->data[0]) = control;
+ pci_intx(dev, 0); /* disable intx */
+ if (dev->msi_enabled)
+ msi_set_enable(dev, 0);
+ if (dev->msix_enabled)
+ msix_set_enable(dev, 0);
+
+ if (pci_seg_supported) {
+ struct physdev_pci_device restore = {
+ .seg = pci_domain_nr(dev->bus),
+ .bus = dev->bus->number,
+ .devfn = dev->devfn
+ };
- spin_lock_irqsave(&msi_dev_entry->pirq_list_lock, flags);
- list_for_each_entry(pirq_entry, &msi_dev_entry->pirq_list_head, list) {
- int j;
-
- /* save the table */
- j = pirq_entry->entry_nr;
- pirq_entry->address_lo_save =
- readl(base + j * PCI_MSIX_ENTRY_SIZE +
- PCI_MSIX_ENTRY_LOWER_ADDR_OFFSET);
- pirq_entry->address_hi_save =
- readl(base + j * PCI_MSIX_ENTRY_SIZE +
- PCI_MSIX_ENTRY_UPPER_ADDR_OFFSET);
- pirq_entry->data_save =
- readl(base + j * PCI_MSIX_ENTRY_SIZE +
- PCI_MSIX_ENTRY_DATA_OFFSET);
+ rc = HYPERVISOR_physdev_op(PHYSDEVOP_restore_msi_ext,
+ &restore);
}
- spin_unlock_irqrestore(&msi_dev_entry->pirq_list_lock, flags);
-
- save_state->cap_nr = PCI_CAP_ID_MSIX;
- pci_add_saved_cap(dev, save_state);
- return 0;
-}
-
-void pci_restore_msix_state(struct pci_dev *dev)
-{
- u16 save;
- int pos, j;
- void __iomem *base;
- struct pci_cap_saved_state *save_state;
- unsigned long flags;
- struct msi_dev_list *msi_dev_entry;
- struct msi_pirq_entry *pirq_entry;
-
- save_state = pci_find_saved_cap(dev, PCI_CAP_ID_MSIX);
- if (!save_state)
- return;
-
- save = *((u16 *)&save_state->data[0]);
- pci_remove_saved_cap(save_state);
- kfree(save_state);
-
- pos = pci_find_capability(dev, PCI_CAP_ID_MSIX);
- if (pos <= 0)
- return;
-
- msi_dev_entry = get_msi_dev_pirq_list(dev);
- base = msi_dev_entry->mask_base;
+#if CONFIG_XEN_COMPAT < 0x040200
+ if (rc == -ENOSYS && !pci_domain_nr(dev->bus)) {
+ struct physdev_restore_msi restore = {
+ .bus = dev->bus->number,
+ .devfn = dev->devfn
+ };
- spin_lock_irqsave(&msi_dev_entry->pirq_list_lock, flags);
- list_for_each_entry(pirq_entry, &msi_dev_entry->pirq_list_head, list) {
- /* route the table */
- j = pirq_entry->entry_nr;
- writel(pirq_entry->address_lo_save,
- base + j * PCI_MSIX_ENTRY_SIZE +
- PCI_MSIX_ENTRY_LOWER_ADDR_OFFSET);
- writel(pirq_entry->address_hi_save,
- base + j * PCI_MSIX_ENTRY_SIZE +
- PCI_MSIX_ENTRY_UPPER_ADDR_OFFSET);
- writel(pirq_entry->data_save,
- base + j * PCI_MSIX_ENTRY_SIZE +
- PCI_MSIX_ENTRY_DATA_OFFSET);
+ pci_seg_supported = false;
+ rc = HYPERVISOR_physdev_op(PHYSDEVOP_restore_msi, &restore);
}
- spin_unlock_irqrestore(&msi_dev_entry->pirq_list_lock, flags);
-
- pci_write_config_word(dev, msi_control_reg(pos), save);
- enable_msi_mode(dev, pos, PCI_CAP_ID_MSIX);
-}
#endif
+ WARN(rc && rc != -ENOSYS, "restore_msi -> %d\n", rc);
+}
+#endif /* CONFIG_PM */
/**
* msi_capability_init - configure device's MSI capability structure
@@ -498,6 +335,8 @@ static int msi_capability_init(struct pc
int pos, pirq;
u16 control;
+ msi_set_enable(dev, 0); /* Ensure msi is disabled as I set it up */
+
pos = pci_find_capability(dev, PCI_CAP_ID_MSI);
pci_read_config_word(dev, msi_control_reg(pos), &control);
@@ -506,7 +345,8 @@ static int msi_capability_init(struct pc
return -EBUSY;
/* Set MSI enabled bits */
- enable_msi_mode(dev, pos, PCI_CAP_ID_MSI);
+ pci_intx(dev, 0); /* disable intx */
+ msi_set_enable(dev, 1);
dev->msi_enabled = 1;
dev->irq = pirq;
@@ -527,25 +367,20 @@ static int msix_capability_init(struct p
struct msix_entry *entries, int nvec)
{
u64 table_base;
- u16 control;
- int pirq, i, j, mapped, pos, nr_entries;
+ int pirq, i, j, mapped, pos;
struct msi_dev_list *msi_dev_entry = get_msi_dev_pirq_list(dev);
struct msi_pirq_entry *pirq_entry;
if (!msi_dev_entry)
return -ENOMEM;
+ msix_set_enable(dev, 0);/* Ensure msix is disabled as I set it up */
+
pos = pci_find_capability(dev, PCI_CAP_ID_MSIX);
table_base = find_table_base(dev, pos);
if (!table_base)
return -ENODEV;
- pci_read_config_word(dev, msi_control_reg(pos), &control);
- nr_entries = multi_msix_capable(control);
- if (!msi_dev_entry->mask_base)
- msi_dev_entry->mask_base =
- ioremap_nocache(table_base, nr_entries * PCI_MSIX_ENTRY_SIZE);
-
/* MSI-X Table Initialization */
for (i = 0; i < nvec; i++) {
mapped = 0;
@@ -584,7 +419,8 @@ static int msix_capability_init(struct p
return avail;
}
- enable_msi_mode(dev, pos, PCI_CAP_ID_MSIX);
+ pci_intx(dev, 0); /* disable intx */
+ msix_set_enable(dev, 1);
dev->msix_enabled = 1;
return 0;
@@ -670,17 +506,14 @@ int pci_enable_msi(struct pci_dev* dev)
/* Check whether driver already requested for MSI-X irqs */
if (dev->msix_enabled) {
printk(KERN_INFO "PCI: %s: Can't enable MSI. "
- "Device already has MSI-X irq assigned\n",
- pci_name(dev));
- dev->irq = temp;
+ "Device already has MSI-X enabled\n",
+ pci_name(dev));
return -EINVAL;
}
status = msi_capability_init(dev);
if ( !status )
msi_dev_entry->default_irq = temp;
- else
- dev->irq = temp;
return status;
}
@@ -688,7 +521,6 @@ int pci_enable_msi(struct pci_dev* dev)
extern void pci_frontend_disable_msi(struct pci_dev* dev);
void pci_disable_msi(struct pci_dev* dev)
{
- int pos;
int pirq;
struct msi_dev_list *msi_dev_entry = get_msi_dev_pirq_list(dev);
@@ -697,13 +529,11 @@ void pci_disable_msi(struct pci_dev* dev
if (!dev)
return;
+ if (!dev->msi_enabled)
+ return;
+
if (!is_initial_xendomain()) {
#ifdef CONFIG_XEN_PCIDEV_FRONTEND
- if (!(dev->msi_enabled)) {
- printk(KERN_INFO "PCI: %s: Device did not enabled MSI.\n",
- pci_name(dev));
- return;
- }
evtchn_map_pirq(dev->irq, 0);
pci_frontend_disable_msi(dev);
dev->irq = msi_dev_entry->default_irq;
@@ -712,23 +542,15 @@ void pci_disable_msi(struct pci_dev* dev
return;
}
- pos = pci_find_capability(dev, PCI_CAP_ID_MSI);
- if (!pos)
- return;
-
- if (!(dev->msi_enabled)) {
- printk(KERN_INFO "PCI: %s: Device did not enabled MSI.\n",
- pci_name(dev));
- return;
- }
-
pirq = dev->irq;
/* Restore dev->irq to its default pin-assertion vector */
dev->irq = msi_dev_entry->default_irq;
msi_unmap_pirq(dev, pirq);
/* Disable MSI mode */
- disable_msi_mode(dev, pos, PCI_CAP_ID_MSI);
+ msi_set_enable(dev, 0);
+ pci_intx(dev, 1); /* enable intx */
+ dev->msi_enabled = 0;
}
/**
@@ -825,7 +647,6 @@ int pci_enable_msix(struct pci_dev* dev,
printk(KERN_INFO "PCI: %s: Can't enable MSI-X. "
"Device already has an MSI irq assigned\n",
pci_name(dev));
- dev->irq = temp;
return -EINVAL;
}
@@ -833,8 +654,6 @@ int pci_enable_msix(struct pci_dev* dev,
if ( !status )
msi_dev_entry->default_irq = temp;
- else
- dev->irq = temp;
return status;
}
@@ -842,52 +661,29 @@ int pci_enable_msix(struct pci_dev* dev,
extern void pci_frontend_disable_msix(struct pci_dev* dev);
void pci_disable_msix(struct pci_dev* dev)
{
- int pos;
- u16 control;
-
if (!pci_msi_enable)
return;
if (!dev)
return;
- if (!dev->msix_enabled) {
- printk(KERN_INFO "PCI: %s: Device did not enabled MSI-X.\n",
- pci_name(dev));
+
+ if (!dev->msix_enabled)
return;
- }
- if (!is_initial_xendomain()) {
+ if (!is_initial_xendomain())
#ifdef CONFIG_XEN_PCIDEV_FRONTEND
- struct msi_dev_list *msi_dev_entry;
- struct msi_pirq_entry *pirq_entry, *tmp;
-
pci_frontend_disable_msix(dev);
-
- msi_dev_entry = get_msi_dev_pirq_list(dev);
- list_for_each_entry_safe(pirq_entry, tmp,
- &msi_dev_entry->pirq_list_head, list) {
- evtchn_map_pirq(pirq_entry->pirq, 0);
- list_del(&pirq_entry->list);
- kfree(pirq_entry);
- }
-
- dev->irq = msi_dev_entry->default_irq;
- dev->msix_enabled = 0;
-#endif
- return;
- }
-
- pos = pci_find_capability(dev, PCI_CAP_ID_MSIX);
- if (!pos)
- return;
-
- pci_read_config_word(dev, msi_control_reg(pos), &control);
- if (!(control & PCI_MSIX_FLAGS_ENABLE))
+#else
return;
+#endif
msi_remove_pci_irq_vectors(dev);
/* Disable MSI mode */
- disable_msi_mode(dev, pos, PCI_CAP_ID_MSIX);
+ if (is_initial_xendomain()) {
+ msix_set_enable(dev, 0);
+ pci_intx(dev, 1); /* enable intx */
+ }
+ dev->msix_enabled = 0;
}
/**
@@ -922,8 +718,6 @@ void msi_remove_pci_irq_vectors(struct p
kfree(pirq_entry);
}
spin_unlock_irqrestore(&msi_dev_entry->pirq_list_lock, flags);
- iounmap(msi_dev_entry->mask_base);
- msi_dev_entry->mask_base = NULL;
dev->irq = msi_dev_entry->default_irq;
}
--- head.orig/drivers/xen/balloon/sysfs.c 2012-02-29 10:23:59.000000000 +0100
+++ head/drivers/xen/balloon/sysfs.c 2011-01-31 17:32:22.000000000 +0100
@@ -34,6 +34,7 @@
#include <linux/stat.h>
#include <linux/string.h>
#include <linux/sysdev.h>
+#include <linux/module.h>
#include "common.h"
#ifdef HAVE_XEN_PLATFORM_COMPAT_H
--- head.orig/drivers/xen/core/evtchn.c 2012-01-26 13:26:14.000000000 +0100
+++ head/drivers/xen/core/evtchn.c 2011-01-31 17:32:22.000000000 +0100
@@ -144,7 +144,7 @@ static void bind_evtchn_to_cpu(unsigned
BUG_ON(!test_bit(chn, s->evtchn_mask));
if (irq != -1)
- set_native_irq_info(irq, cpumask_of_cpu(cpu));
+ irq_desc[irq].affinity = cpumask_of_cpu(cpu);
clear_bit(chn, (unsigned long *)cpu_evtchn_mask[cpu_evtchn[chn]]);
set_bit(chn, (unsigned long *)cpu_evtchn_mask[cpu]);
@@ -157,7 +157,7 @@ static void init_evtchn_cpu_bindings(voi
/* By default all event channels notify CPU#0. */
for (i = 0; i < NR_IRQS; i++)
- set_native_irq_info(i, cpumask_of_cpu(0));
+ irq_desc[i].affinity = cpumask_of_cpu(0);
memset(cpu_evtchn, 0, sizeof(cpu_evtchn));
for_each_possible_cpu(i)
@@ -745,10 +745,10 @@ static void ack_dynirq(unsigned int irq)
static void end_dynirq(unsigned int irq)
{
- move_masked_irq(irq);
-
- if (!(irq_desc[irq].status & IRQ_DISABLED))
+ if (!(irq_desc[irq].status & IRQ_DISABLED)) {
+ move_masked_irq(irq);
unmask_dynirq(irq);
+ }
}
static struct irq_chip dynirq_chip = {
@@ -892,13 +892,14 @@ static void unmask_pirq(unsigned int irq
static void end_pirq(unsigned int irq)
{
- move_masked_irq(irq);
-
if ((irq_desc[irq].status & (IRQ_DISABLED|IRQ_PENDING)) ==
(IRQ_DISABLED|IRQ_PENDING))
shutdown_pirq(irq);
- else
+ else {
+ if (!(irq_desc[irq].status & IRQ_DISABLED))
+ move_masked_irq(irq);
unmask_pirq(irq);
+ }
}
static struct irq_chip pirq_chip = {
--- head.orig/drivers/xen/core/machine_kexec.c 2011-12-21 10:11:41.000000000 +0100
+++ head/drivers/xen/core/machine_kexec.c 2012-02-17 09:55:35.000000000 +0100
@@ -217,7 +217,7 @@ void xen_machine_kexec_unload(struct kim
* stop all CPUs and kexec. That is it combines machine_shutdown()
* and machine_kexec() in Linux kexec terms.
*/
-NORET_TYPE void machine_kexec(struct kimage *image)
+void __noreturn machine_kexec(struct kimage *image)
{
xen_kexec_exec_t xke;
--- head.orig/drivers/xen/core/smpboot.c 2012-03-22 16:07:26.000000000 +0100
+++ head/drivers/xen/core/smpboot.c 2012-01-20 14:45:36.000000000 +0100
@@ -112,7 +112,7 @@ static int __cpuinit xen_smp_intr_init(u
rc = bind_ipi_to_irqhandler(RESCHEDULE_VECTOR,
cpu,
smp_reschedule_interrupt,
- SA_INTERRUPT,
+ IRQF_DISABLED|IRQF_NOBALANCING,
resched_name[cpu],
NULL);
if (rc < 0)
@@ -123,7 +123,7 @@ static int __cpuinit xen_smp_intr_init(u
rc = bind_ipi_to_irqhandler(CALL_FUNCTION_VECTOR,
cpu,
smp_call_function_interrupt,
- SA_INTERRUPT,
+ IRQF_DISABLED|IRQF_NOBALANCING,
callfunc_name[cpu],
NULL);
if (rc < 0)
@@ -247,7 +247,7 @@ void __init smp_prepare_cpus(unsigned in
{
unsigned int cpu;
struct task_struct *idle;
- int apicid, acpiid;
+ int apicid;
struct vcpu_get_physid cpu_id;
#ifdef __x86_64__
struct desc_ptr *gdt_descr;
@@ -256,14 +256,8 @@ void __init smp_prepare_cpus(unsigned in
#endif
apicid = 0;
- if (HYPERVISOR_vcpu_op(VCPUOP_get_physid, 0, &cpu_id) == 0) {
+ if (HYPERVISOR_vcpu_op(VCPUOP_get_physid, 0, &cpu_id) == 0)
apicid = xen_vcpu_physid_to_x86_apicid(cpu_id.phys_id);
- acpiid = xen_vcpu_physid_to_x86_acpiid(cpu_id.phys_id);
-#ifdef CONFIG_ACPI
- if (acpiid != 0xff)
- x86_acpiid_to_apicid[acpiid] = apicid;
-#endif
- }
boot_cpu_data.apicid = apicid;
cpu_data[0] = boot_cpu_data;
@@ -318,14 +312,8 @@ void __init smp_prepare_cpus(unsigned in
XENFEAT_writable_descriptor_tables);
apicid = cpu;
- if (HYPERVISOR_vcpu_op(VCPUOP_get_physid, cpu, &cpu_id) == 0) {
+ if (HYPERVISOR_vcpu_op(VCPUOP_get_physid, cpu, &cpu_id) == 0)
apicid = xen_vcpu_physid_to_x86_apicid(cpu_id.phys_id);
- acpiid = xen_vcpu_physid_to_x86_acpiid(cpu_id.phys_id);
-#ifdef CONFIG_ACPI
- if (acpiid != 0xff)
- x86_acpiid_to_apicid[acpiid] = apicid;
-#endif
- }
cpu_data[cpu] = boot_cpu_data;
cpu_data[cpu].apicid = apicid;
--- head.orig/drivers/xen/netback/interface.c 2011-11-03 12:01:26.000000000 +0100
+++ head/drivers/xen/netback/interface.c 2011-11-03 12:03:20.000000000 +0100
@@ -165,7 +165,7 @@ static void netbk_get_drvinfo(struct net
struct ethtool_drvinfo *info)
{
strcpy(info->driver, "netbk");
- strcpy(info->bus_info, dev->class_dev.dev->bus_id);
+ strcpy(info->bus_info, dev->dev.parent->bus_id);
}
static const struct netif_stat {
--- head.orig/drivers/xen/netfront/netfront.c 2012-06-06 13:47:56.000000000 +0200
+++ head/drivers/xen/netfront/netfront.c 2012-06-06 13:49:14.000000000 +0200
@@ -1807,7 +1807,7 @@ static void netfront_get_drvinfo(struct
struct ethtool_drvinfo *info)
{
strcpy(info->driver, "netfront");
- strcpy(info->bus_info, dev->class_dev.dev->bus_id);
+ strcpy(info->bus_info, dev->dev.parent->bus_id);
}
static int network_connect(struct net_device *dev)
@@ -1935,20 +1935,19 @@ static struct ethtool_ops network_ethtoo
};
#ifdef CONFIG_SYSFS
-static ssize_t show_rxbuf_min(struct class_device *cd, char *buf)
+static ssize_t show_rxbuf_min(struct device *dev,
+ struct device_attribute *attr, char *buf)
{
- struct net_device *netdev = container_of(cd, struct net_device,
- class_dev);
- struct netfront_info *info = netdev_priv(netdev);
+ struct netfront_info *info = netdev_priv(to_net_dev(dev));
return sprintf(buf, "%u\n", info->rx_min_target);
}
-static ssize_t store_rxbuf_min(struct class_device *cd,
+static ssize_t store_rxbuf_min(struct device *dev,
+ struct device_attribute *attr,
const char *buf, size_t len)
{
- struct net_device *netdev = container_of(cd, struct net_device,
- class_dev);
+ struct net_device *netdev = to_net_dev(dev);
struct netfront_info *np = netdev_priv(netdev);
char *endp;
unsigned long target;
@@ -1978,20 +1977,19 @@ static ssize_t store_rxbuf_min(struct cl
return len;
}
-static ssize_t show_rxbuf_max(struct class_device *cd, char *buf)
+static ssize_t show_rxbuf_max(struct device *dev,
+ struct device_attribute *attr, char *buf)
{
- struct net_device *netdev = container_of(cd, struct net_device,
- class_dev);
- struct netfront_info *info = netdev_priv(netdev);
+ struct netfront_info *info = netdev_priv(to_net_dev(dev));
return sprintf(buf, "%u\n", info->rx_max_target);
}
-static ssize_t store_rxbuf_max(struct class_device *cd,
+static ssize_t store_rxbuf_max(struct device *dev,
+ struct device_attribute *attr,
const char *buf, size_t len)
{
- struct net_device *netdev = container_of(cd, struct net_device,
- class_dev);
+ struct net_device *netdev = to_net_dev(dev);
struct netfront_info *np = netdev_priv(netdev);
char *endp;
unsigned long target;
@@ -2021,16 +2019,15 @@ static ssize_t store_rxbuf_max(struct cl
return len;
}
-static ssize_t show_rxbuf_cur(struct class_device *cd, char *buf)
+static ssize_t show_rxbuf_cur(struct device *dev,
+ struct device_attribute *attr, char *buf)
{
- struct net_device *netdev = container_of(cd, struct net_device,
- class_dev);
- struct netfront_info *info = netdev_priv(netdev);
+ struct netfront_info *info = netdev_priv(to_net_dev(dev));
return sprintf(buf, "%u\n", info->rx_target);
}
-static const struct class_device_attribute xennet_attrs[] = {
+static struct device_attribute xennet_attrs[] = {
__ATTR(rxbuf_min, S_IRUGO|S_IWUSR, show_rxbuf_min, store_rxbuf_min),
__ATTR(rxbuf_max, S_IRUGO|S_IWUSR, show_rxbuf_max, store_rxbuf_max),
__ATTR(rxbuf_cur, S_IRUGO, show_rxbuf_cur, NULL),
@@ -2042,8 +2039,8 @@ static int xennet_sysfs_addif(struct net
int error = 0;
for (i = 0; i < ARRAY_SIZE(xennet_attrs); i++) {
- error = class_device_create_file(&netdev->class_dev,
- &xennet_attrs[i]);
+ error = device_create_file(&netdev->dev,
+ &xennet_attrs[i]);
if (error)
goto fail;
}
@@ -2051,8 +2048,7 @@ static int xennet_sysfs_addif(struct net
fail:
while (--i >= 0)
- class_device_remove_file(&netdev->class_dev,
- &xennet_attrs[i]);
+ device_remove_file(&netdev->dev, &xennet_attrs[i]);
return error;
}
@@ -2060,10 +2056,8 @@ static void xennet_sysfs_delif(struct ne
{
int i;
- for (i = 0; i < ARRAY_SIZE(xennet_attrs); i++) {
- class_device_remove_file(&netdev->class_dev,
- &xennet_attrs[i]);
- }
+ for (i = 0; i < ARRAY_SIZE(xennet_attrs); i++)
+ device_remove_file(&netdev->dev, &xennet_attrs[i]);
}
#endif /* CONFIG_SYSFS */
--- head.orig/drivers/xen/tpmback/common.h 2011-12-21 10:33:05.000000000 +0100
+++ head/drivers/xen/tpmback/common.h 2011-11-28 09:46:47.000000000 +0100
@@ -63,7 +63,7 @@ typedef struct tpmif_st {
void tpmif_disconnect_complete(tpmif_t * tpmif);
tpmif_t *tpmif_find(domid_t domid, struct backend_info *bi);
-void tpmif_interface_init(void);
+int tpmif_interface_init(void);
void tpmif_interface_exit(void);
void tpmif_schedule_work(tpmif_t * tpmif);
void tpmif_deschedule_work(tpmif_t * tpmif);
--- head.orig/drivers/xen/tpmback/interface.c 2011-04-11 14:30:27.000000000 +0200
+++ head/drivers/xen/tpmback/interface.c 2011-04-11 14:31:42.000000000 +0200
@@ -119,13 +119,14 @@ void tpmif_disconnect_complete(tpmif_t *
free_tpmif(tpmif);
}
-void __init tpmif_interface_init(void)
+int __init tpmif_interface_init(void)
{
tpmif_cachep = kmem_cache_create("tpmif_cache", sizeof (tpmif_t),
0, 0, NULL, NULL);
+ return tpmif_cachep ? 0 : -ENOMEM;
}
-void __exit tpmif_interface_exit(void)
+void tpmif_interface_exit(void)
{
kmem_cache_destroy(tpmif_cachep);
}
--- head.orig/drivers/xen/tpmback/tpmback.c 2011-01-31 17:29:16.000000000 +0100
+++ head/drivers/xen/tpmback/tpmback.c 2011-01-31 17:32:22.000000000 +0100
@@ -918,22 +918,30 @@ static int __init tpmback_init(void)
spin_lock_init(&tpm_schedule_list_lock);
INIT_LIST_HEAD(&tpm_schedule_list);
- tpmif_interface_init();
- tpmif_xenbus_init();
+ rc = tpmif_interface_init();
+ if (!rc) {
+ rc = tpmif_xenbus_init();
+ if (rc)
+ tpmif_interface_exit();
+ }
+ if (rc) {
+ misc_deregister(&vtpms_miscdevice);
+ return rc;
+ }
printk(KERN_ALERT "Successfully initialized TPM backend driver.\n");
return 0;
}
-
module_init(tpmback_init);
-void __exit tpmback_exit(void)
+static void __exit tpmback_exit(void)
{
vtpm_release_packets(NULL, 0);
tpmif_xenbus_exit();
tpmif_interface_exit();
misc_deregister(&vtpms_miscdevice);
}
+module_exit(tpmback_exit)
MODULE_LICENSE("Dual BSD/GPL");
--- head.orig/drivers/xen/xenoprof/xenoprofile.c 2012-02-16 12:29:10.000000000 +0100
+++ head/drivers/xen/xenoprof/xenoprofile.c 2012-02-16 12:29:42.000000000 +0100
@@ -239,7 +239,7 @@ static int bind_virq(void)
result = bind_virq_to_irqhandler(VIRQ_XENOPROF,
i,
xenoprof_ovf_interrupt,
- SA_INTERRUPT,
+ IRQF_DISABLED|IRQF_NOBALANCING,
"xenoprof",
NULL);
--- head.orig/arch/x86/include/mach-xen/asm/desc_32.h 2011-01-31 17:32:16.000000000 +0100
+++ head/arch/x86/include/mach-xen/asm/desc_32.h 2011-01-31 17:32:22.000000000 +0100
@@ -21,7 +21,7 @@ struct Xgt_desc_struct {
extern struct Xgt_desc_struct idt_descr;
DECLARE_PER_CPU(struct Xgt_desc_struct, cpu_gdt_descr);
-
+extern struct Xgt_desc_struct early_gdt_descr;
static inline struct desc_struct *get_cpu_gdt_table(unsigned int cpu)
{
--- head.orig/arch/x86/include/mach-xen/asm/hypervisor.h 2011-01-31 17:32:16.000000000 +0100
+++ head/arch/x86/include/mach-xen/asm/hypervisor.h 2011-01-31 17:32:22.000000000 +0100
@@ -160,6 +160,19 @@ static inline void arch_leave_lazy_mmu_m
#define arch_use_lazy_mmu_mode() unlikely(__get_cpu_var(xen_lazy_mmu))
#endif
+#if 0 /* All uses are in places potentially called asynchronously, but
+ * asynchronous code should rather not make use of lazy mode at all.
+ * Therefore, all uses of this function get commented out, proper
+ * detection of asynchronous invocations is added whereever needed,
+ * and this function is disabled to catch any new (improper) uses.
+ */
+static inline void arch_flush_lazy_mmu_mode(void)
+{
+ if (arch_use_lazy_mmu_mode())
+ xen_multicall_flush(false);
+}
+#endif
+
#else /* !CONFIG_XEN || MODULE */
static inline void xen_multicall_flush(bool ignore) {}
@@ -217,7 +230,7 @@ HYPERVISOR_block(
return rc;
}
-static inline void /*__noreturn*/
+static inline void __noreturn
HYPERVISOR_shutdown(
unsigned int reason)
{
--- head.orig/arch/x86/include/mach-xen/asm/irqflags_32.h 2011-01-31 17:32:16.000000000 +0100
+++ head/arch/x86/include/mach-xen/asm/irqflags_32.h 2011-01-31 17:32:22.000000000 +0100
@@ -108,7 +108,7 @@ sysexit_scrit: /**** START OF SYSEXIT CR
sysexit_ecrit: /**** END OF SYSEXIT CRITICAL REGION ****/ ; \
mov $__KERNEL_PDA, %ecx ; \
push %esp ; \
- mov %ecx, %gs ; \
+ mov %ecx, %fs ; \
call evtchn_do_upcall ; \
add $4,%esp ; \
jmp ret_from_intr
--- head.orig/arch/x86/include/mach-xen/asm/mmu_context_32.h 2011-01-31 17:32:16.000000000 +0100
+++ head/arch/x86/include/mach-xen/asm/mmu_context_32.h 2011-01-31 17:32:22.000000000 +0100
@@ -27,13 +27,13 @@ static inline void enter_lazy_tlb(struct
static inline void __prepare_arch_switch(void)
{
/*
- * Save away %fs. No need to save %gs, as it was saved on the
+ * Save away %gs. No need to save %fs, as it was saved on the
* stack on entry. No need to save %es and %ds, as those are
* always kernel segments while inside the kernel.
*/
- asm volatile ( "mov %%fs,%0"
- : "=m" (current->thread.fs));
- asm volatile ( "movl %0,%%fs"
+ asm volatile ( "mov %%gs,%0"
+ : "=m" (current->thread.gs));
+ asm volatile ( "movl %0,%%gs"
: : "r" (0) );
}
@@ -95,7 +95,7 @@ static inline void switch_mm(struct mm_s
}
#define deactivate_mm(tsk, mm) \
- asm("movl %0,%%fs": :"r" (0));
+ asm("movl %0,%%gs": :"r" (0));
static inline void activate_mm(struct mm_struct *prev, struct mm_struct *next)
{
--- head.orig/arch/x86/include/mach-xen/asm/pgalloc_32.h 2008-07-21 11:00:33.000000000 +0200
+++ head/arch/x86/include/mach-xen/asm/pgalloc_32.h 2011-01-31 17:32:22.000000000 +0100
@@ -6,12 +6,23 @@
#include <linux/mm.h> /* for struct page */
#include <asm/io.h> /* for phys_to_virt and page_to_pseudophys */
-#define pmd_populate_kernel(mm, pmd, pte) \
- set_pmd(pmd, __pmd(_PAGE_TABLE + __pa(pte)))
+#define paravirt_alloc_pt(pfn) do { } while (0)
+#define paravirt_alloc_pd(pfn) do { } while (0)
+#define paravirt_alloc_pd(pfn) do { } while (0)
+#define paravirt_alloc_pd_clone(pfn, clonepfn, start, count) do { } while (0)
+#define paravirt_release_pt(pfn) do { } while (0)
+#define paravirt_release_pd(pfn) do { } while (0)
+
+#define pmd_populate_kernel(mm, pmd, pte) \
+do { \
+ paravirt_alloc_pt(__pa(pte) >> PAGE_SHIFT); \
+ set_pmd(pmd, __pmd(_PAGE_TABLE + __pa(pte))); \
+} while (0)
#define pmd_populate(mm, pmd, pte) \
do { \
unsigned long pfn = page_to_pfn(pte); \
+ paravirt_alloc_pt(pfn); \
if (test_bit(PG_pinned, &virt_to_page((mm)->pgd)->flags)) { \
if (!PageHighMem(pte)) \
BUG_ON(HYPERVISOR_update_va_mapping( \
@@ -42,7 +53,11 @@ static inline void pte_free_kernel(pte_t
extern void pte_free(struct page *pte);
-#define __pte_free_tlb(tlb,pte) tlb_remove_page((tlb),(pte))
+#define __pte_free_tlb(tlb,pte) \
+do { \
+ paravirt_release_pt(page_to_pfn(pte)); \
+ tlb_remove_page((tlb),(pte)); \
+} while (0)
#ifdef CONFIG_X86_PAE
/*
--- head.orig/arch/x86/include/mach-xen/asm/pgtable_32.h 2011-01-31 17:32:16.000000000 +0100
+++ head/arch/x86/include/mach-xen/asm/pgtable_32.h 2011-01-31 17:32:22.000000000 +0100
@@ -275,6 +275,7 @@ static inline pte_t pte_mkhuge(pte_t pte
*/
#define pte_update(mm, addr, ptep) do { } while (0)
#define pte_update_defer(mm, addr, ptep) do { } while (0)
+#define paravirt_map_pt_hook(slot, va, pfn) do { } while (0)
/*
* We only update the dirty/accessed state if we set
@@ -490,12 +491,24 @@ extern pte_t *lookup_address(unsigned lo
#endif
#if defined(CONFIG_HIGHPTE)
-#define pte_offset_map(dir, address) \
- ((pte_t *)kmap_atomic_pte(pmd_page(*(dir)),KM_PTE0) + \
- pte_index(address))
-#define pte_offset_map_nested(dir, address) \
- ((pte_t *)kmap_atomic_pte(pmd_page(*(dir)),KM_PTE1) + \
- pte_index(address))
+#define pte_offset_map(dir, address) \
+({ \
+ pte_t *__ptep; \
+ unsigned pfn = pmd_val(*(dir)) >> PAGE_SHIFT; \
+ __ptep = (pte_t *)kmap_atomic_pte(pfn_to_page(pfn),KM_PTE0); \
+ paravirt_map_pt_hook(KM_PTE0,__ptep, pfn); \
+ __ptep = __ptep + pte_index(address); \
+ __ptep; \
+})
+#define pte_offset_map_nested(dir, address) \
+({ \
+ pte_t *__ptep; \
+ unsigned pfn = pmd_val(*(dir)) >> PAGE_SHIFT; \
+ __ptep = (pte_t *)kmap_atomic_pte(pfn_to_page(pfn),KM_PTE1); \
+ paravirt_map_pt_hook(KM_PTE1,__ptep, pfn); \
+ __ptep = __ptep + pte_index(address); \
+ __ptep; \
+})
#define pte_unmap(pte) kunmap_atomic(pte, KM_PTE0)
#define pte_unmap_nested(pte) kunmap_atomic(pte, KM_PTE1)
#else
--- head.orig/arch/x86/include/mach-xen/asm/processor_32.h 2011-01-31 17:32:16.000000000 +0100
+++ head/arch/x86/include/mach-xen/asm/processor_32.h 2011-01-31 17:32:22.000000000 +0100
@@ -431,7 +431,7 @@ struct thread_struct {
.vm86_info = NULL, \
.sysenter_cs = __KERNEL_CS, \
.io_bitmap_ptr = NULL, \
- .gs = __KERNEL_PDA, \
+ .fs = __KERNEL_PDA, \
}
/*
@@ -449,8 +449,8 @@ struct thread_struct {
}
#define start_thread(regs, new_eip, new_esp) do { \
- __asm__("movl %0,%%fs": :"r" (0)); \
- regs->xgs = 0; \
+ __asm__("movl %0,%%gs": :"r" (0)); \
+ regs->xfs = 0; \
set_fs(USER_DS); \
regs->xds = __USER_DS; \
regs->xes = __USER_DS; \
--- head.orig/arch/x86/include/mach-xen/asm/smp_32.h 2011-01-31 17:32:16.000000000 +0100
+++ head/arch/x86/include/mach-xen/asm/smp_32.h 2011-01-31 17:32:22.000000000 +0100
@@ -52,6 +52,11 @@ extern void cpu_exit_clear(void);
extern void cpu_uninit(void);
#endif
+#ifndef CONFIG_PARAVIRT
+#define startup_ipi_hook(phys_apicid, start_eip, start_esp) \
+do { } while (0)
+#endif
+
/*
* This function is needed by all SMP systems. It must _always_ be valid
* from the initial startup. We map APIC_BASE very early in page_setup(),
--- head.orig/arch/x86/include/mach-xen/asm/pgtable_64.h 2011-01-31 17:32:16.000000000 +0100
+++ head/arch/x86/include/mach-xen/asm/pgtable_64.h 2011-01-31 17:32:22.000000000 +0100
@@ -416,15 +416,6 @@ static inline int pmd_large(pmd_t pte) {
#define mk_pte(page, pgprot) pfn_pte(page_to_pfn(page), (pgprot))
#define mk_pte_huge(entry) (__pte_val(entry) |= _PAGE_PRESENT | _PAGE_PSE)
-/* physical address -> PTE */
-static inline pte_t mk_pte_phys(unsigned long physpage, pgprot_t pgprot)
-{
- unsigned long pteval;
- pteval = physpage | pgprot_val(pgprot);
- pteval &= __supported_pte_mask;
- return __pte(pteval);
-}
-
/* Change flags of a PTE */
static inline pte_t pte_modify(pte_t pte, pgprot_t newprot)
{
--- head.orig/arch/x86/include/mach-xen/asm/smp_64.h 2011-01-31 17:32:16.000000000 +0100
+++ head/arch/x86/include/mach-xen/asm/smp_64.h 2011-01-31 17:32:22.000000000 +0100
@@ -7,6 +7,7 @@
#include <linux/threads.h>
#include <linux/cpumask.h>
#include <linux/bitops.h>
+#include <linux/init.h>
extern int disable_apic;
#ifdef CONFIG_X86_LOCAL_APIC
@@ -73,7 +74,7 @@ extern int __cpu_disable(void);
extern void __cpu_die(unsigned int cpu);
extern void prefill_possible_map(void);
extern unsigned num_processors;
-extern unsigned disabled_cpus;
+extern unsigned __cpuinitdata disabled_cpus;
#define NO_PROC_ID 0xFF /* No processor magic marker */
--- head.orig/include/xen/xenbus.h 2011-12-21 10:02:58.000000000 +0100
+++ head/include/xen/xenbus.h 2011-12-21 10:42:02.000000000 +0100
@@ -118,6 +118,7 @@ struct xenbus_driver {
#define DEFINE_XENBUS_DRIVER(var, drvname, methods...) \
struct xenbus_driver var ## _driver = { \
.driver.name = drvname + 0 ?: var ## _ids->devicetype, \
+ .driver.mod_name = KBUILD_MODNAME, \
XENBUS_DRIVER_SET_OWNER(THIS_MODULE) \
.ids = var ## _ids, ## methods \
}
@@ -127,8 +128,8 @@ static inline struct xenbus_driver *to_x
return container_of(drv, struct xenbus_driver, driver);
}
-int xenbus_register_frontend(struct xenbus_driver *drv);
-int xenbus_register_backend(struct xenbus_driver *drv);
+int __must_check xenbus_register_frontend(struct xenbus_driver *drv);
+int __must_check xenbus_register_backend(struct xenbus_driver *drv);
void xenbus_unregister_driver(struct xenbus_driver *drv);
struct xenbus_transaction
--- head.orig/lib/swiotlb-xen.c 2010-09-16 13:31:46.000000000 +0200
+++ head/lib/swiotlb-xen.c 2011-01-31 17:32:22.000000000 +0100
@@ -143,8 +143,8 @@ __setup("swiotlb=", setup_io_tlb_npages)
* 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)
+void __init
+swiotlb_init_with_default_size(size_t default_size)
{
unsigned long i, bytes;
int rc;
@@ -229,7 +229,7 @@ swiotlb_init_with_default_size (size_t d
dma_bits);
}
-void
+void __init
swiotlb_init(void)
{
long ram_end;
@@ -480,8 +480,8 @@ swiotlb_full(struct device *dev, size_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 : "?");
+ printk(KERN_ERR "PCI-DMA: Out of SW-IOMMU space for %zu bytes at "
+ "device %s\n", size, dev ? dev->bus_id : "?");
if (size > io_tlb_overflow && do_panic) {
if (dir == PCI_DMA_FROMDEVICE || dir == PCI_DMA_BIDIRECTIONAL)
@@ -562,7 +562,7 @@ swiotlb_unmap_single(struct device *hwde
* address back to the card, you must first perform a
* swiotlb_dma_sync_for_device, and then the device again owns the buffer
*/
-static inline void
+static void
swiotlb_sync_single(struct device *hwdev, dma_addr_t dev_addr,
size_t size, int dir, int target)
{
@@ -630,7 +630,7 @@ swiotlb_map_sg(struct device *hwdev, str
sg[0].dma_length = 0;
return 0;
}
- sg->dma_address = (dma_addr_t)virt_to_bus(map);
+ sg->dma_address = virt_to_bus(map);
} else
sg->dma_address = dev_addr;
sg->dma_length = sg->length;
@@ -652,8 +652,7 @@ swiotlb_unmap_sg(struct device *hwdev, s
for (i = 0; i < nelems; i++, sg++)
if (in_swiotlb_aperture(sg->dma_address))
- unmap_single(hwdev,
- (void *)bus_to_virt(sg->dma_address),
+ unmap_single(hwdev, bus_to_virt(sg->dma_address),
sg->dma_length, dir);
else
gnttab_dma_unmap_page(sg->dma_address);
@@ -666,7 +665,7 @@ swiotlb_unmap_sg(struct device *hwdev, s
* The same as swiotlb_sync_single_* but for a scatter-gather list, same rules
* and usage.
*/
-static inline void
+static void
swiotlb_sync_sg(struct device *hwdev, struct scatterlist *sg,
int nelems, int dir, int target)
{