#else
/* Number of 4K pages in DTLB/ITLB combined(in pages): */
int x86_tlbsize;
+ #endif
__u8 x86_virt_bits;
__u8 x86_phys_bits;
- #endif
/* CPUID returned core id bits: */
__u8 x86_coreid_bits;
/* Max extended CPUID function supported: */
unsigned long loops_per_jiffy;
#ifdef CONFIG_SMP
/* cpus sharing the last level cache: */
- cpumask_t llc_shared_map;
+ cpumask_var_t llc_shared_map;
#endif
/* cpuid returned max cores value: */
u16 x86_max_cores;
DECLARE_INIT_PER_CPU(irq_stack_union);
DECLARE_PER_CPU(char *, irq_stack_ptr);
+ DECLARE_PER_CPU(unsigned int, irq_count);
+ extern unsigned long kernel_eflags;
+ extern asmlinkage void ignore_sysret(void);
#else /* X86_64 */
#ifdef CONFIG_CC_STACKPROTECTOR
DECLARE_PER_CPU(unsigned long, stack_canary);
extern void mwait_idle_with_hints(unsigned long eax, unsigned long ecx);
extern void select_idle_routine(const struct cpuinfo_x86 *c);
+extern void init_c1e_mask(void);
extern unsigned long boot_option_idle_override;
extern unsigned long idle_halt;
- #include <linux/init.h>
- #include <linux/kernel.h>
- #include <linux/sched.h>
- #include <linux/string.h>
#include <linux/bootmem.h>
+ #include <linux/linkage.h>
#include <linux/bitops.h>
+ #include <linux/kernel.h>
#include <linux/module.h>
- #include <linux/kgdb.h>
- #include <linux/topology.h>
+ #include <linux/percpu.h>
+ #include <linux/string.h>
#include <linux/delay.h>
+ #include <linux/sched.h>
+ #include <linux/init.h>
+ #include <linux/kgdb.h>
#include <linux/smp.h>
- #include <linux/percpu.h>
- #include <asm/i387.h>
- #include <asm/msr.h>
- #include <asm/io.h>
- #include <asm/linkage.h>
+ #include <linux/io.h>
+
+ #include <asm/stackprotector.h>
#include <asm/mmu_context.h>
+ #include <asm/hypervisor.h>
+ #include <asm/processor.h>
+ #include <asm/sections.h>
+ #include <asm/topology.h>
+ #include <asm/cpumask.h>
+ #include <asm/pgtable.h>
+ #include <asm/atomic.h>
+ #include <asm/proto.h>
+ #include <asm/setup.h>
+ #include <asm/apic.h>
+ #include <asm/desc.h>
+ #include <asm/i387.h>
#include <asm/mtrr.h>
+ #include <asm/numa.h>
+ #include <asm/asm.h>
+ #include <asm/cpu.h>
#include <asm/mce.h>
+ #include <asm/msr.h>
#include <asm/pat.h>
- #include <asm/asm.h>
- #include <asm/numa.h>
#include <asm/smp.h>
- #include <asm/cpu.h>
- #include <asm/cpumask.h>
- #include <asm/apic.h>
#ifdef CONFIG_X86_LOCAL_APIC
#include <asm/uv/uv.h>
#endif
- #include <asm/pgtable.h>
- #include <asm/processor.h>
- #include <asm/desc.h>
- #include <asm/atomic.h>
- #include <asm/proto.h>
- #include <asm/sections.h>
- #include <asm/setup.h>
- #include <asm/hypervisor.h>
- #include <asm/stackprotector.h>
-
#include "cpu.h"
-#ifdef CONFIG_X86_64
-
/* all of these masks are initialized in setup_cpu_local_masks() */
- cpumask_var_t cpu_callin_mask;
- cpumask_var_t cpu_callout_mask;
cpumask_var_t cpu_initialized_mask;
+ cpumask_var_t cpu_callout_mask;
+ cpumask_var_t cpu_callin_mask;
/* representing cpus for which sibling maps can be computed */
cpumask_var_t cpu_sibling_setup_mask;
alloc_bootmem_cpumask_var(&cpu_sibling_setup_mask);
}
- static struct cpu_dev *this_cpu __cpuinitdata;
-#else /* CONFIG_X86_32 */
-
-cpumask_t cpu_sibling_setup_map;
-cpumask_t cpu_callout_map;
-cpumask_t cpu_initialized;
-cpumask_t cpu_callin_map;
-
-#endif /* CONFIG_X86_32 */
-
-
+ static const struct cpu_dev *this_cpu __cpuinitdata;
DEFINE_PER_CPU_PAGE_ALIGNED(struct gdt_page, gdt_page) = { .gdt = {
#ifdef CONFIG_X86_64
* IRET will check the segment types kkeil 2000/10/28
* Also sysret mandates a special GDT layout
*
- * The TLS descriptors are currently at a different place compared to i386.
+ * TLS descriptors are currently at a different place compared to i386.
* Hopefully nobody expects them at a fixed place (Wine?)
*/
- [GDT_ENTRY_KERNEL32_CS] = { { { 0x0000ffff, 0x00cf9b00 } } },
- [GDT_ENTRY_KERNEL_CS] = { { { 0x0000ffff, 0x00af9b00 } } },
- [GDT_ENTRY_KERNEL_DS] = { { { 0x0000ffff, 0x00cf9300 } } },
- [GDT_ENTRY_DEFAULT_USER32_CS] = { { { 0x0000ffff, 0x00cffb00 } } },
- [GDT_ENTRY_DEFAULT_USER_DS] = { { { 0x0000ffff, 0x00cff300 } } },
- [GDT_ENTRY_DEFAULT_USER_CS] = { { { 0x0000ffff, 0x00affb00 } } },
+ [GDT_ENTRY_KERNEL32_CS] = { { { 0x0000ffff, 0x00cf9b00 } } },
+ [GDT_ENTRY_KERNEL_CS] = { { { 0x0000ffff, 0x00af9b00 } } },
+ [GDT_ENTRY_KERNEL_DS] = { { { 0x0000ffff, 0x00cf9300 } } },
+ [GDT_ENTRY_DEFAULT_USER32_CS] = { { { 0x0000ffff, 0x00cffb00 } } },
+ [GDT_ENTRY_DEFAULT_USER_DS] = { { { 0x0000ffff, 0x00cff300 } } },
+ [GDT_ENTRY_DEFAULT_USER_CS] = { { { 0x0000ffff, 0x00affb00 } } },
#else
- [GDT_ENTRY_KERNEL_CS] = { { { 0x0000ffff, 0x00cf9a00 } } },
- [GDT_ENTRY_KERNEL_DS] = { { { 0x0000ffff, 0x00cf9200 } } },
- [GDT_ENTRY_DEFAULT_USER_CS] = { { { 0x0000ffff, 0x00cffa00 } } },
- [GDT_ENTRY_DEFAULT_USER_DS] = { { { 0x0000ffff, 0x00cff200 } } },
+ [GDT_ENTRY_KERNEL_CS] = { { { 0x0000ffff, 0x00cf9a00 } } },
+ [GDT_ENTRY_KERNEL_DS] = { { { 0x0000ffff, 0x00cf9200 } } },
+ [GDT_ENTRY_DEFAULT_USER_CS] = { { { 0x0000ffff, 0x00cffa00 } } },
+ [GDT_ENTRY_DEFAULT_USER_DS] = { { { 0x0000ffff, 0x00cff200 } } },
/*
* Segments used for calling PnP BIOS have byte granularity.
* They code segments and data segments have fixed 64k limits,
* the transfer segment sizes are set at run time.
*/
/* 32-bit code */
- [GDT_ENTRY_PNPBIOS_CS32] = { { { 0x0000ffff, 0x00409a00 } } },
+ [GDT_ENTRY_PNPBIOS_CS32] = { { { 0x0000ffff, 0x00409a00 } } },
/* 16-bit code */
- [GDT_ENTRY_PNPBIOS_CS16] = { { { 0x0000ffff, 0x00009a00 } } },
+ [GDT_ENTRY_PNPBIOS_CS16] = { { { 0x0000ffff, 0x00009a00 } } },
/* 16-bit data */
- [GDT_ENTRY_PNPBIOS_DS] = { { { 0x0000ffff, 0x00009200 } } },
+ [GDT_ENTRY_PNPBIOS_DS] = { { { 0x0000ffff, 0x00009200 } } },
/* 16-bit data */
- [GDT_ENTRY_PNPBIOS_TS1] = { { { 0x00000000, 0x00009200 } } },
+ [GDT_ENTRY_PNPBIOS_TS1] = { { { 0x00000000, 0x00009200 } } },
/* 16-bit data */
- [GDT_ENTRY_PNPBIOS_TS2] = { { { 0x00000000, 0x00009200 } } },
+ [GDT_ENTRY_PNPBIOS_TS2] = { { { 0x00000000, 0x00009200 } } },
/*
* The APM segments have byte granularity and their bases
* are set at run time. All have 64k limits.
*/
/* 32-bit code */
- [GDT_ENTRY_APMBIOS_BASE] = { { { 0x0000ffff, 0x00409a00 } } },
+ [GDT_ENTRY_APMBIOS_BASE] = { { { 0x0000ffff, 0x00409a00 } } },
/* 16-bit code */
- [GDT_ENTRY_APMBIOS_BASE+1] = { { { 0x0000ffff, 0x00009a00 } } },
+ [GDT_ENTRY_APMBIOS_BASE+1] = { { { 0x0000ffff, 0x00009a00 } } },
/* data */
- [GDT_ENTRY_APMBIOS_BASE+2] = { { { 0x0000ffff, 0x00409200 } } },
+ [GDT_ENTRY_APMBIOS_BASE+2] = { { { 0x0000ffff, 0x00409200 } } },
- [GDT_ENTRY_ESPFIX_SS] = { { { 0x00000000, 0x00c09200 } } },
- [GDT_ENTRY_PERCPU] = { { { 0x0000ffff, 0x00cf9200 } } },
+ [GDT_ENTRY_ESPFIX_SS] = { { { 0x00000000, 0x00c09200 } } },
+ [GDT_ENTRY_PERCPU] = { { { 0x0000ffff, 0x00cf9200 } } },
GDT_STACK_CANARY_INIT
#endif
} };
* the CPUID. Add "volatile" to not allow gcc to
* optimize the subsequent calls to this function.
*/
- asm volatile ("pushfl\n\t"
- "pushfl\n\t"
- "popl %0\n\t"
- "movl %0,%1\n\t"
- "xorl %2,%0\n\t"
- "pushl %0\n\t"
- "popfl\n\t"
- "pushfl\n\t"
- "popl %0\n\t"
- "popfl\n\t"
+ asm volatile ("pushfl \n\t"
+ "pushfl \n\t"
+ "popl %0 \n\t"
+ "movl %0, %1 \n\t"
+ "xorl %2, %0 \n\t"
+ "pushl %0 \n\t"
+ "popfl \n\t"
+ "pushfl \n\t"
+ "popl %0 \n\t"
+ "popfl \n\t"
+
: "=&r" (f1), "=&r" (f2)
: "ir" (flag));
static void __cpuinit squash_the_stupid_serial_number(struct cpuinfo_x86 *c)
{
- if (cpu_has(c, X86_FEATURE_PN) && disable_x86_serial_nr) {
- /* Disable processor serial number */
- unsigned long lo, hi;
- rdmsr(MSR_IA32_BBL_CR_CTL, lo, hi);
- lo |= 0x200000;
- wrmsr(MSR_IA32_BBL_CR_CTL, lo, hi);
- printk(KERN_NOTICE "CPU serial number disabled.\n");
- clear_cpu_cap(c, X86_FEATURE_PN);
-
- /* Disabling the serial number may affect the cpuid level */
- c->cpuid_level = cpuid_eax(0);
- }
+ unsigned long lo, hi;
+
+ if (!cpu_has(c, X86_FEATURE_PN) || !disable_x86_serial_nr)
+ return;
+
+ /* Disable processor serial number: */
+
+ rdmsr(MSR_IA32_BBL_CR_CTL, lo, hi);
+ lo |= 0x200000;
+ wrmsr(MSR_IA32_BBL_CR_CTL, lo, hi);
+
+ printk(KERN_NOTICE "CPU serial number disabled.\n");
+ clear_cpu_cap(c, X86_FEATURE_PN);
+
+ /* Disabling the serial number may affect the cpuid level */
+ c->cpuid_level = cpuid_eax(0);
}
static int __init x86_serial_nr_setup(char *s)
u32 feature;
u32 level;
};
+
static const struct cpuid_dependent_feature __cpuinitconst
cpuid_dependent_features[] = {
{ X86_FEATURE_MWAIT, 0x00000005 },
static void __cpuinit filter_cpuid_features(struct cpuinfo_x86 *c, bool warn)
{
const struct cpuid_dependent_feature *df;
+
for (df = cpuid_dependent_features; df->feature; df++) {
+
+ if (!cpu_has(c, df->feature))
+ continue;
/*
* Note: cpuid_level is set to -1 if unavailable, but
* extended_extended_level is set to 0 if unavailable
* when signed; hence the weird messing around with
* signs here...
*/
- if (cpu_has(c, df->feature) &&
- ((s32)df->level < 0 ?
+ if (!((s32)df->level < 0 ?
(u32)df->level > (u32)c->extended_cpuid_level :
- (s32)df->level > (s32)c->cpuid_level)) {
- clear_cpu_cap(c, df->feature);
- if (warn)
- printk(KERN_WARNING
- "CPU: CPU feature %s disabled "
- "due to lack of CPUID level 0x%x\n",
- x86_cap_flags[df->feature],
- df->level);
- }
+ (s32)df->level > (s32)c->cpuid_level))
+ continue;
+
+ clear_cpu_cap(c, df->feature);
+ if (!warn)
+ continue;
+
+ printk(KERN_WARNING
+ "CPU: CPU feature %s disabled, no CPUID level 0x%x\n",
+ x86_cap_flags[df->feature], df->level);
}
}
/*
* Naming convention should be: <Name> [(<Codename>)]
* This table only is used unless init_<vendor>() below doesn't set it;
- * in particular, if CPUID levels 0x80000002..4 are supported, this isn't used
- *
+ * in particular, if CPUID levels 0x80000002..4 are supported, this
+ * isn't used
*/
/* Look up CPU names by table lookup. */
- static char __cpuinit *table_lookup_model(struct cpuinfo_x86 *c)
+ static const char *__cpuinit table_lookup_model(struct cpuinfo_x86 *c)
{
- struct cpu_model_info *info;
+ const struct cpu_model_info *info;
if (c->x86_model >= 16)
return NULL; /* Range check */
load_stack_canary_segment();
}
- /* Current gdt points %fs at the "master" per-cpu area: after this,
- * it's on the real one. */
+ /*
+ * Current gdt points %fs at the "master" per-cpu area: after this,
+ * it's on the real one.
+ */
void switch_to_new_gdt(int cpu)
{
struct desc_ptr gdt_descr;
load_percpu_segment(cpu);
}
- static struct cpu_dev *cpu_devs[X86_VENDOR_NUM] = {};
+ static const struct cpu_dev *__cpuinitdata cpu_devs[X86_VENDOR_NUM] = {};
static void __cpuinit default_init(struct cpuinfo_x86 *c)
{
#endif
}
- static struct cpu_dev __cpuinitdata default_cpu = {
+ static const struct cpu_dev __cpuinitconst default_cpu = {
.c_init = default_init,
.c_vendor = "Unknown",
.c_x86_vendor = X86_VENDOR_UNKNOWN,
if (c->extended_cpuid_level < 0x80000004)
return;
- v = (unsigned int *) c->x86_model_id;
+ v = (unsigned int *)c->x86_model_id;
cpuid(0x80000002, &v[0], &v[1], &v[2], &v[3]);
cpuid(0x80000003, &v[4], &v[5], &v[6], &v[7]);
cpuid(0x80000004, &v[8], &v[9], &v[10], &v[11]);
c->x86_model_id[48] = 0;
- /* Intel chips right-justify this string for some dumb reason;
- undo that brain damage */
+ /*
+ * Intel chips right-justify this string for some dumb reason;
+ * undo that brain damage:
+ */
p = q = &c->x86_model_id[0];
while (*p == ' ')
- p++;
+ p++;
if (p != q) {
- while (*p)
- *q++ = *p++;
- while (q <= &c->x86_model_id[48])
- *q++ = '\0'; /* Zero-pad the rest */
+ while (*p)
+ *q++ = *p++;
+ while (q <= &c->x86_model_id[48])
+ *q++ = '\0'; /* Zero-pad the rest */
}
}
if (smp_num_siblings == 1) {
printk(KERN_INFO "CPU: Hyper-Threading is disabled\n");
- } else if (smp_num_siblings > 1) {
+ goto out;
+ }
- if (smp_num_siblings > nr_cpu_ids) {
- printk(KERN_WARNING "CPU: Unsupported number of siblings %d",
- smp_num_siblings);
- smp_num_siblings = 1;
- return;
- }
+ if (smp_num_siblings <= 1)
+ goto out;
- index_msb = get_count_order(smp_num_siblings);
- c->phys_proc_id = apic->phys_pkg_id(c->initial_apicid, index_msb);
+ if (smp_num_siblings > nr_cpu_ids) {
+ pr_warning("CPU: Unsupported number of siblings %d",
+ smp_num_siblings);
+ smp_num_siblings = 1;
+ return;
+ }
- smp_num_siblings = smp_num_siblings / c->x86_max_cores;
+ index_msb = get_count_order(smp_num_siblings);
+ c->phys_proc_id = apic->phys_pkg_id(c->initial_apicid, index_msb);
- index_msb = get_count_order(smp_num_siblings);
+ smp_num_siblings = smp_num_siblings / c->x86_max_cores;
- core_bits = get_count_order(c->x86_max_cores);
+ index_msb = get_count_order(smp_num_siblings);
- c->cpu_core_id = apic->phys_pkg_id(c->initial_apicid, index_msb) &
- ((1 << core_bits) - 1);
- }
+ core_bits = get_count_order(c->x86_max_cores);
+
+ c->cpu_core_id = apic->phys_pkg_id(c->initial_apicid, index_msb) &
+ ((1 << core_bits) - 1);
out:
if ((c->x86_max_cores * smp_num_siblings) > 1) {
static void __cpuinit get_cpu_vendor(struct cpuinfo_x86 *c)
{
char *v = c->x86_vendor_id;
- int i;
static int printed;
+ int i;
for (i = 0; i < X86_VENDOR_NUM; i++) {
if (!cpu_devs[i])
if (!strcmp(v, cpu_devs[i]->c_ident[0]) ||
(cpu_devs[i]->c_ident[1] &&
!strcmp(v, cpu_devs[i]->c_ident[1]))) {
+
this_cpu = cpu_devs[i];
c->x86_vendor = this_cpu->c_x86_vendor;
return;
if (!printed) {
printed++;
- printk(KERN_ERR "CPU: vendor_id '%s' unknown, using generic init.\n", v);
+ printk(KERN_ERR
+ "CPU: vendor_id '%s' unknown, using generic init.\n", v);
+
printk(KERN_ERR "CPU: Your system may be unstable.\n");
}
/* Intel-defined flags: level 0x00000001 */
if (c->cpuid_level >= 0x00000001) {
u32 junk, tfms, cap0, misc;
+
cpuid(0x00000001, &tfms, &misc, &junk, &cap0);
c->x86 = (tfms >> 8) & 0xf;
c->x86_model = (tfms >> 4) & 0xf;
c->x86_mask = tfms & 0xf;
+
if (c->x86 == 0xf)
c->x86 += (tfms >> 20) & 0xff;
if (c->x86 >= 0x6)
c->x86_model += ((tfms >> 16) & 0xf) << 4;
+
if (cap0 & (1<<19)) {
c->x86_clflush_size = ((misc >> 8) & 0xff) * 8;
c->x86_cache_alignment = c->x86_clflush_size;
/* Intel-defined flags: level 0x00000001 */
if (c->cpuid_level >= 0x00000001) {
u32 capability, excap;
+
cpuid(0x00000001, &tfms, &ebx, &excap, &capability);
c->x86_capability[0] = capability;
c->x86_capability[4] = excap;
/* AMD-defined flags: level 0x80000001 */
xlvl = cpuid_eax(0x80000000);
c->extended_cpuid_level = xlvl;
+
if ((xlvl & 0xffff0000) == 0x80000000) {
if (xlvl >= 0x80000001) {
c->x86_capability[1] = cpuid_edx(0x80000001);
}
}
- #ifdef CONFIG_X86_64
if (c->extended_cpuid_level >= 0x80000008) {
u32 eax = cpuid_eax(0x80000008);
c->x86_virt_bits = (eax >> 8) & 0xff;
c->x86_phys_bits = eax & 0xff;
}
+ #ifdef CONFIG_X86_32
+ else if (cpu_has(c, X86_FEATURE_PAE) || cpu_has(c, X86_FEATURE_PSE36))
+ c->x86_phys_bits = 36;
#endif
if (c->extended_cpuid_level >= 0x80000007)
{
#ifdef CONFIG_X86_64
c->x86_clflush_size = 64;
+ c->x86_phys_bits = 36;
+ c->x86_virt_bits = 48;
#else
c->x86_clflush_size = 32;
+ c->x86_phys_bits = 32;
+ c->x86_virt_bits = 32;
#endif
c->x86_cache_alignment = c->x86_clflush_size;
void __init early_cpu_init(void)
{
- struct cpu_dev **cdev;
+ const struct cpu_dev *const *cdev;
int count = 0;
- printk("KERNEL supported cpus:\n");
+ printk(KERN_INFO "KERNEL supported cpus:\n");
for (cdev = __x86_cpu_dev_start; cdev < __x86_cpu_dev_end; cdev++) {
- struct cpu_dev *cpudev = *cdev;
+ const struct cpu_dev *cpudev = *cdev;
unsigned int j;
if (count >= X86_VENDOR_NUM)
for (j = 0; j < 2; j++) {
if (!cpudev->c_ident[j])
continue;
- printk(" %s %s\n", cpudev->c_vendor,
+ printk(KERN_INFO " %s %s\n", cpudev->c_vendor,
cpudev->c_ident[j]);
}
}
c->x86_coreid_bits = 0;
#ifdef CONFIG_X86_64
c->x86_clflush_size = 64;
+ c->x86_phys_bits = 36;
+ c->x86_virt_bits = 48;
#else
c->cpuid_level = -1; /* CPUID not detected */
c->x86_clflush_size = 32;
+ c->x86_phys_bits = 32;
+ c->x86_virt_bits = 32;
#endif
c->x86_cache_alignment = c->x86_clflush_size;
memset(&c->x86_capability, 0, sizeof c->x86_capability);
squash_the_stupid_serial_number(c);
/*
- * The vendor-specific functions might have changed features. Now
- * we do "generic changes."
+ * The vendor-specific functions might have changed features.
+ * Now we do "generic changes."
*/
/* Filter out anything that depends on CPUID levels we don't have */
/* If the model name is still unset, do table lookup. */
if (!c->x86_model_id[0]) {
- char *p;
+ const char *p;
p = table_lookup_model(c);
if (p)
strcpy(c->x86_model_id, p);
void __init identify_boot_cpu(void)
{
identify_cpu(&boot_cpu_data);
+ init_c1e_mask();
#ifdef CONFIG_X86_32
sysenter_setup();
enable_sep_cpu();
}
struct msr_range {
- unsigned min;
- unsigned max;
+ unsigned min;
+ unsigned max;
};
- static struct msr_range msr_range_array[] __cpuinitdata = {
+ static const struct msr_range msr_range_array[] __cpuinitconst = {
{ 0x00000000, 0x00000418},
{ 0xc0000000, 0xc000040b},
{ 0xc0010000, 0xc0010142},
static void __cpuinit print_cpu_msr(void)
{
+ unsigned index_min, index_max;
unsigned index;
u64 val;
int i;
- unsigned index_min, index_max;
for (i = 0; i < ARRAY_SIZE(msr_range_array); i++) {
index_min = msr_range_array[i].min;
index_max = msr_range_array[i].max;
+
for (index = index_min; index < index_max; index++) {
if (rdmsrl_amd_safe(index, &val))
continue;
}
static int show_msr __cpuinitdata;
+
static __init int setup_show_msr(char *arg)
{
int num;
void __cpuinit print_cpu_info(struct cpuinfo_x86 *c)
{
- char *vendor = NULL;
+ const char *vendor = NULL;
- if (c->x86_vendor < X86_VENDOR_NUM)
+ if (c->x86_vendor < X86_VENDOR_NUM) {
vendor = this_cpu->c_vendor;
- else if (c->cpuid_level >= 0)
- vendor = c->x86_vendor_id;
+ } else {
+ if (c->cpuid_level >= 0)
+ vendor = c->x86_vendor_id;
+ }
if (vendor && !strstr(c->x86_model_id, vendor))
printk(KERN_CONT "%s ", vendor);
static __init int setup_disablecpuid(char *arg)
{
int bit;
+
if (get_option(&arg, &bit) && bit < NCAPINTS*32)
setup_clear_cpu_cap(bit);
else
return 0;
+
return 1;
}
__setup("clearcpuid=", setup_disablecpuid);
DEFINE_PER_CPU_FIRST(union irq_stack_union,
irq_stack_union) __aligned(PAGE_SIZE);
+
DEFINE_PER_CPU(char *, irq_stack_ptr) =
init_per_cpu_var(irq_stack_union.irq_stack) + IRQ_STACK_SIZE - 64;
DEFINE_PER_CPU(unsigned int, irq_count) = -1;
+ /*
+ * Special IST stacks which the CPU switches to when it calls
+ * an IST-marked descriptor entry. Up to 7 stacks (hardware
+ * limit), all of them are 4K, except the debug stack which
+ * is 8K.
+ */
+ static const unsigned int exception_stack_sizes[N_EXCEPTION_STACKS] = {
+ [0 ... N_EXCEPTION_STACKS - 1] = EXCEPTION_STKSZ,
+ [DEBUG_STACK - 1] = DEBUG_STKSZ
+ };
+
static DEFINE_PER_CPU_PAGE_ALIGNED(char, exception_stacks
[(N_EXCEPTION_STACKS - 1) * EXCEPTION_STKSZ + DEBUG_STKSZ])
__aligned(PAGE_SIZE);
- extern asmlinkage void ignore_sysret(void);
-
/* May not be marked __init: used by software suspend */
void syscall_init(void)
{
*/
DEFINE_PER_CPU(struct orig_ist, orig_ist);
- #else /* x86_64 */
+ #else /* CONFIG_X86_64 */
#ifdef CONFIG_CC_STACKPROTECTOR
DEFINE_PER_CPU(unsigned long, stack_canary);
memset(regs, 0, sizeof(struct pt_regs));
regs->fs = __KERNEL_PERCPU;
regs->gs = __KERNEL_STACK_CANARY;
+
return regs;
}
- #endif /* x86_64 */
+ #endif /* CONFIG_X86_64 */
+
+ /*
+ * Clear all 6 debug registers:
+ */
+ static void clear_all_debug_regs(void)
+ {
+ int i;
+
+ for (i = 0; i < 8; i++) {
+ /* Ignore db4, db5 */
+ if ((i == 4) || (i == 5))
+ continue;
+
+ set_debugreg(0, i);
+ }
+ }
/*
* cpu_init() initializes state that is per-CPU. Some data is already
* A lot of state is already set up in PDA init for 64 bit
*/
#ifdef CONFIG_X86_64
+
void __cpuinit cpu_init(void)
{
- int cpu = stack_smp_processor_id();
- struct tss_struct *t = &per_cpu(init_tss, cpu);
- struct orig_ist *orig_ist = &per_cpu(orig_ist, cpu);
- unsigned long v;
+ struct orig_ist *orig_ist;
struct task_struct *me;
+ struct tss_struct *t;
+ unsigned long v;
+ int cpu;
int i;
+ cpu = stack_smp_processor_id();
+ t = &per_cpu(init_tss, cpu);
+ orig_ist = &per_cpu(orig_ist, cpu);
+
#ifdef CONFIG_NUMA
if (cpu != 0 && percpu_read(node_number) == 0 &&
cpu_to_node(cpu) != NUMA_NO_NODE)
* set up and load the per-CPU TSS
*/
if (!orig_ist->ist[0]) {
- static const unsigned int sizes[N_EXCEPTION_STACKS] = {
- [0 ... N_EXCEPTION_STACKS - 1] = EXCEPTION_STKSZ,
- [DEBUG_STACK - 1] = DEBUG_STKSZ
- };
char *estacks = per_cpu(exception_stacks, cpu);
+
for (v = 0; v < N_EXCEPTION_STACKS; v++) {
- estacks += sizes[v];
+ estacks += exception_stack_sizes[v];
orig_ist->ist[v] = t->x86_tss.ist[v] =
(unsigned long)estacks;
}
}
t->x86_tss.io_bitmap_base = offsetof(struct tss_struct, io_bitmap);
+
/*
* <= is required because the CPU will access up to
* 8 bits beyond the end of the IO permission bitmap.
atomic_inc(&init_mm.mm_count);
me->active_mm = &init_mm;
- if (me->mm)
- BUG();
+ BUG_ON(me->mm);
enter_lazy_tlb(&init_mm, me);
load_sp0(t, ¤t->thread);
arch_kgdb_ops.correct_hw_break();
else
#endif
- {
- /*
- * Clear all 6 debug registers:
- */
- set_debugreg(0UL, 0);
- set_debugreg(0UL, 1);
- set_debugreg(0UL, 2);
- set_debugreg(0UL, 3);
- set_debugreg(0UL, 6);
- set_debugreg(0UL, 7);
- }
+ clear_all_debug_regs();
fpu_init();
if (cpumask_test_and_set_cpu(cpu, cpu_initialized_mask)) {
printk(KERN_WARNING "CPU#%d already initialized!\n", cpu);
- for (;;) local_irq_enable();
+ for (;;)
+ local_irq_enable();
}
printk(KERN_INFO "Initializing CPU#%d\n", cpu);
*/
atomic_inc(&init_mm.mm_count);
curr->active_mm = &init_mm;
- if (curr->mm)
- BUG();
+ BUG_ON(curr->mm);
enter_lazy_tlb(&init_mm, curr);
load_sp0(t, thread);
__set_tss_desc(cpu, GDT_ENTRY_DOUBLEFAULT_TSS, &doublefault_tss);
#endif
- /* Clear all 6 debug registers: */
- set_debugreg(0, 0);
- set_debugreg(0, 1);
- set_debugreg(0, 2);
- set_debugreg(0, 3);
- set_debugreg(0, 6);
- set_debugreg(0, 7);
+ clear_all_debug_regs();
/*
* Force FPU initialization:
xsave_init();
}
-
-
#endif
#include <linux/cpufreq.h>
#include <linux/slab.h>
#include <linux/cpumask.h>
+ #include <linux/timex.h>
#include <asm/processor.h>
#include <asm/msr.h>
- #include <asm/timex.h>
+ #include <asm/timer.h>
#include "speedstep-lib.h"
#define PFX "p4-clockmod: "
- #define dprintk(msg...) cpufreq_debug_printk(CPUFREQ_DEBUG_DRIVER, "p4-clockmod", msg)
+ #define dprintk(msg...) cpufreq_debug_printk(CPUFREQ_DEBUG_DRIVER, \
+ "p4-clockmod", msg)
/*
* Duty Cycle (3bits), note DC_DISABLE is not specified in
{
u32 l, h;
- if (!cpu_online(cpu) || (newstate > DC_DISABLE) || (newstate == DC_RESV))
+ if (!cpu_online(cpu) ||
+ (newstate > DC_DISABLE) || (newstate == DC_RESV))
return -EINVAL;
rdmsr_on_cpu(cpu, MSR_IA32_THERM_STATUS, &l, &h);
if (l & 0x01)
dprintk("CPU#%d currently thermal throttled\n", cpu);
- if (has_N44_O17_errata[cpu] && (newstate == DC_25PT || newstate == DC_DFLT))
+ if (has_N44_O17_errata[cpu] &&
+ (newstate == DC_25PT || newstate == DC_DFLT))
newstate = DC_38PT;
rdmsr_on_cpu(cpu, MSR_IA32_THERM_CONTROL, &l, &h);
struct cpufreq_freqs freqs;
int i;
- if (cpufreq_frequency_table_target(policy, &p4clockmod_table[0], target_freq, relation, &newstate))
+ if (cpufreq_frequency_table_target(policy, &p4clockmod_table[0],
+ target_freq, relation, &newstate))
return -EINVAL;
freqs.old = cpufreq_p4_get(policy->cpu);
cpufreq_notify_transition(&freqs, CPUFREQ_PRECHANGE);
}
- /* run on each logical CPU, see section 13.15.3 of IA32 Intel Architecture Software
+ /* run on each logical CPU,
+ * see section 13.15.3 of IA32 Intel Architecture Software
* Developer's Manual, Volume 3
*/
for_each_cpu(i, policy->cpus)
{
if (c->x86 == 0x06) {
if (cpu_has(c, X86_FEATURE_EST))
- printk(KERN_WARNING PFX "Warning: EST-capable CPU detected. "
- "The acpi-cpufreq module offers voltage scaling"
- " in addition of frequency scaling. You should use "
- "that instead of p4-clockmod, if possible.\n");
+ printk(KERN_WARNING PFX "Warning: EST-capable CPU "
+ "detected. The acpi-cpufreq module offers "
+ "voltage scaling in addition of frequency "
+ "scaling. You should use that instead of "
+ "p4-clockmod, if possible.\n");
switch (c->x86_model) {
case 0x0E: /* Core */
case 0x0F: /* Core Duo */
case 0x16: /* Celeron Core */
p4clockmod_driver.flags |= CPUFREQ_CONST_LOOPS;
- return speedstep_get_processor_frequency(SPEEDSTEP_PROCESSOR_PCORE);
+ return speedstep_get_frequency(SPEEDSTEP_CPU_PCORE);
case 0x0D: /* Pentium M (Dothan) */
p4clockmod_driver.flags |= CPUFREQ_CONST_LOOPS;
/* fall through */
case 0x09: /* Pentium M (Banias) */
- return speedstep_get_processor_frequency(SPEEDSTEP_PROCESSOR_PM);
+ return speedstep_get_frequency(SPEEDSTEP_CPU_PM);
}
}
if (c->x86 != 0xF) {
if (!cpu_has(c, X86_FEATURE_EST))
- printk(KERN_WARNING PFX "Unknown p4-clockmod-capable CPU. "
- "Please send an e-mail to <cpufreq@vger.kernel.org>\n");
+ printk(KERN_WARNING PFX "Unknown CPU. "
+ "Please send an e-mail to "
+ "<cpufreq@vger.kernel.org>\n");
return 0;
}
* throttling is active or not. */
p4clockmod_driver.flags |= CPUFREQ_CONST_LOOPS;
- if (speedstep_detect_processor() == SPEEDSTEP_PROCESSOR_P4M) {
+ if (speedstep_detect_processor() == SPEEDSTEP_CPU_P4M) {
printk(KERN_WARNING PFX "Warning: Pentium 4-M detected. "
"The speedstep-ich or acpi cpufreq modules offer "
"voltage scaling in addition of frequency scaling. "
"You should use either one instead of p4-clockmod, "
"if possible.\n");
- return speedstep_get_processor_frequency(SPEEDSTEP_PROCESSOR_P4M);
+ return speedstep_get_frequency(SPEEDSTEP_CPU_P4M);
}
- return speedstep_get_processor_frequency(SPEEDSTEP_PROCESSOR_P4D);
+ return speedstep_get_frequency(SPEEDSTEP_CPU_P4D);
}
unsigned int i;
#ifdef CONFIG_SMP
- cpumask_copy(policy->cpus, &per_cpu(cpu_sibling_map, policy->cpu));
+ cpumask_copy(policy->cpus, cpu_sibling_mask(policy->cpu));
#endif
/* Errata workaround */
dprintk("has errata -- disabling low frequencies\n");
}
+ if (speedstep_detect_processor() == SPEEDSTEP_CPU_P4D &&
+ c->x86_model < 2) {
+ /* switch to maximum frequency and measure result */
+ cpufreq_p4_setdc(policy->cpu, DC_DISABLE);
+ recalibrate_cpu_khz();
+ }
/* get max frequency */
stock_freq = cpufreq_p4_get_frequency(c);
if (!stock_freq)
return -EINVAL;
/* table init */
- for (i=1; (p4clockmod_table[i].frequency != CPUFREQ_TABLE_END); i++) {
- if ((i<2) && (has_N44_O17_errata[policy->cpu]))
+ for (i = 1; (p4clockmod_table[i].frequency != CPUFREQ_TABLE_END); i++) {
+ if ((i < 2) && (has_N44_O17_errata[policy->cpu]))
p4clockmod_table[i].frequency = CPUFREQ_ENTRY_INVALID;
else
p4clockmod_table[i].frequency = (stock_freq * i)/8;
cpufreq_frequency_table_get_attr(p4clockmod_table, policy->cpu);
/* cpuinfo and default policy values */
- policy->cpuinfo.transition_latency = 1000000; /* assumed */
+
+ /* the transition latency is set to be 1 higher than the maximum
+ * transition latency of the ondemand governor */
+ policy->cpuinfo.transition_latency = 10000001;
policy->cur = stock_freq;
return cpufreq_frequency_table_cpuinfo(policy, &p4clockmod_table[0]);
l = DC_DISABLE;
if (l != DC_DISABLE)
- return (stock_freq * l / 8);
+ return stock_freq * l / 8;
return stock_freq;
}
- static struct freq_attr* p4clockmod_attr[] = {
+ static struct freq_attr *p4clockmod_attr[] = {
&cpufreq_freq_attr_scaling_available_freqs,
NULL,
};
ret = cpufreq_register_driver(&p4clockmod_driver);
if (!ret)
- printk(KERN_INFO PFX "P4/Xeon(TM) CPU On-Demand Clock Modulation available\n");
+ printk(KERN_INFO PFX "P4/Xeon(TM) CPU On-Demand Clock "
+ "Modulation available\n");
- return (ret);
+ return ret;
}
}
- MODULE_AUTHOR ("Zwane Mwaikambo <zwane@commfireservices.com>");
- MODULE_DESCRIPTION ("cpufreq driver for Pentium(TM) 4/Xeon(TM)");
- MODULE_LICENSE ("GPL");
+ MODULE_AUTHOR("Zwane Mwaikambo <zwane@commfireservices.com>");
+ MODULE_DESCRIPTION("cpufreq driver for Pentium(TM) 4/Xeon(TM)");
+ MODULE_LICENSE("GPL");
late_initcall(cpufreq_p4_init);
module_exit(cpufreq_p4_exit);
#include <linux/string.h>
#include <linux/cpumask.h>
#include <linux/sched.h> /* for current / set_cpus_allowed() */
+ #include <linux/io.h>
+ #include <linux/delay.h>
#include <asm/msr.h>
- #include <asm/io.h>
- #include <asm/delay.h>
- #ifdef CONFIG_X86_POWERNOW_K8_ACPI
#include <linux/acpi.h>
#include <linux/mutex.h>
#include <acpi/processor.h>
- #endif
#define PFX "powernow-k8: "
#define VERSION "version 2.20.00"
static int cpu_family = CPU_OPTERON;
#ifndef CONFIG_SMP
-DEFINE_PER_CPU(cpumask_t, cpu_core_map);
+static inline const struct cpumask *cpu_core_mask(int cpu)
+{
+ return cpumask_of(0);
+}
#endif
/* Return a frequency in MHz, given an input fid */
return 1000 * find_freq_from_fid(fid);
}
- static u32 find_khz_freq_from_pstate(struct cpufreq_frequency_table *data, u32 pstate)
+ static u32 find_khz_freq_from_pstate(struct cpufreq_frequency_table *data,
+ u32 pstate)
{
return data[pstate].frequency;
}
return 1;
}
- lo = fid | (data->currvid << MSR_C_LO_VID_SHIFT) | MSR_C_LO_INIT_FID_VID;
+ lo = fid;
+ lo |= (data->currvid << MSR_C_LO_VID_SHIFT);
+ lo |= MSR_C_LO_INIT_FID_VID;
dprintk("writing fid 0x%x, lo 0x%x, hi 0x%x\n",
fid, lo, data->plllock * PLL_LOCK_CONVERSION);
do {
wrmsr(MSR_FIDVID_CTL, lo, data->plllock * PLL_LOCK_CONVERSION);
if (i++ > 100) {
- printk(KERN_ERR PFX "Hardware error - pending bit very stuck - no further pstate changes possible\n");
+ printk(KERN_ERR PFX
+ "Hardware error - pending bit very stuck - "
+ "no further pstate changes possible\n");
return 1;
}
} while (query_current_values_with_pending_wait(data));
count_off_irt(data);
if (savevid != data->currvid) {
- printk(KERN_ERR PFX "vid change on fid trans, old 0x%x, new 0x%x\n",
- savevid, data->currvid);
+ printk(KERN_ERR PFX
+ "vid change on fid trans, old 0x%x, new 0x%x\n",
+ savevid, data->currvid);
return 1;
}
if (fid != data->currfid) {
- printk(KERN_ERR PFX "fid trans failed, fid 0x%x, curr 0x%x\n", fid,
- data->currfid);
+ printk(KERN_ERR PFX
+ "fid trans failed, fid 0x%x, curr 0x%x\n", fid,
+ data->currfid);
return 1;
}
return 1;
}
- lo = data->currfid | (vid << MSR_C_LO_VID_SHIFT) | MSR_C_LO_INIT_FID_VID;
+ lo = data->currfid;
+ lo |= (vid << MSR_C_LO_VID_SHIFT);
+ lo |= MSR_C_LO_INIT_FID_VID;
dprintk("writing vid 0x%x, lo 0x%x, hi 0x%x\n",
vid, lo, STOP_GRANT_5NS);
do {
wrmsr(MSR_FIDVID_CTL, lo, STOP_GRANT_5NS);
if (i++ > 100) {
- printk(KERN_ERR PFX "internal error - pending bit very stuck - no further pstate changes possible\n");
+ printk(KERN_ERR PFX "internal error - pending bit "
+ "very stuck - no further pstate "
+ "changes possible\n");
return 1;
}
} while (query_current_values_with_pending_wait(data));
if (savefid != data->currfid) {
- printk(KERN_ERR PFX "fid changed on vid trans, old 0x%x new 0x%x\n",
+ printk(KERN_ERR PFX "fid changed on vid trans, old "
+ "0x%x new 0x%x\n",
savefid, data->currfid);
return 1;
}
if (vid != data->currvid) {
- printk(KERN_ERR PFX "vid trans failed, vid 0x%x, curr 0x%x\n", vid,
- data->currvid);
+ printk(KERN_ERR PFX "vid trans failed, vid 0x%x, "
+ "curr 0x%x\n",
+ vid, data->currvid);
return 1;
}
* Decreasing vid codes represent increasing voltages:
* vid of 0 is 1.550V, vid of 0x1e is 0.800V, vid of VID_OFF is off.
*/
- static int decrease_vid_code_by_step(struct powernow_k8_data *data, u32 reqvid, u32 step)
+ static int decrease_vid_code_by_step(struct powernow_k8_data *data,
+ u32 reqvid, u32 step)
{
if ((data->currvid - reqvid) > step)
reqvid = data->currvid - step;
}
/* Change Opteron/Athlon64 fid and vid, by the 3 phases. */
- static int transition_fid_vid(struct powernow_k8_data *data, u32 reqfid, u32 reqvid)
+ static int transition_fid_vid(struct powernow_k8_data *data,
+ u32 reqfid, u32 reqvid)
{
if (core_voltage_pre_transition(data, reqvid))
return 1;
return 1;
if ((reqfid != data->currfid) || (reqvid != data->currvid)) {
- printk(KERN_ERR PFX "failed (cpu%d): req 0x%x 0x%x, curr 0x%x 0x%x\n",
+ printk(KERN_ERR PFX "failed (cpu%d): req 0x%x 0x%x, "
+ "curr 0x%x 0x%x\n",
smp_processor_id(),
reqfid, reqvid, data->currfid, data->currvid);
return 1;
}
/* Phase 1 - core voltage transition ... setup voltage */
- static int core_voltage_pre_transition(struct powernow_k8_data *data, u32 reqvid)
+ static int core_voltage_pre_transition(struct powernow_k8_data *data,
+ u32 reqvid)
{
u32 rvosteps = data->rvo;
u32 savefid = data->currfid;
u32 maxvid, lo;
- dprintk("ph1 (cpu%d): start, currfid 0x%x, currvid 0x%x, reqvid 0x%x, rvo 0x%x\n",
+ dprintk("ph1 (cpu%d): start, currfid 0x%x, currvid 0x%x, "
+ "reqvid 0x%x, rvo 0x%x\n",
smp_processor_id(),
data->currfid, data->currvid, reqvid, data->rvo);
} else {
dprintk("ph1: changing vid for rvo, req 0x%x\n",
data->currvid - 1);
- if (decrease_vid_code_by_step(data, data->currvid - 1, 1))
+ if (decrease_vid_code_by_step(data, data->currvid-1, 1))
return 1;
rvosteps--;
}
return 1;
if (savefid != data->currfid) {
- printk(KERN_ERR PFX "ph1 err, currfid changed 0x%x\n", data->currfid);
+ printk(KERN_ERR PFX "ph1 err, currfid changed 0x%x\n",
+ data->currfid);
return 1;
}
/* Phase 2 - core frequency transition */
static int core_frequency_transition(struct powernow_k8_data *data, u32 reqfid)
{
- u32 vcoreqfid, vcocurrfid, vcofiddiff, fid_interval, savevid = data->currvid;
+ u32 vcoreqfid, vcocurrfid, vcofiddiff;
+ u32 fid_interval, savevid = data->currvid;
- if ((reqfid < HI_FID_TABLE_BOTTOM) && (data->currfid < HI_FID_TABLE_BOTTOM)) {
- printk(KERN_ERR PFX "ph2: illegal lo-lo transition 0x%x 0x%x\n",
- reqfid, data->currfid);
+ if ((reqfid < HI_FID_TABLE_BOTTOM) &&
+ (data->currfid < HI_FID_TABLE_BOTTOM)) {
+ printk(KERN_ERR PFX "ph2: illegal lo-lo transition "
+ "0x%x 0x%x\n", reqfid, data->currfid);
return 1;
}
if (data->currfid == reqfid) {
- printk(KERN_ERR PFX "ph2 null fid transition 0x%x\n", data->currfid);
+ printk(KERN_ERR PFX "ph2 null fid transition 0x%x\n",
+ data->currfid);
return 0;
}
- dprintk("ph2 (cpu%d): starting, currfid 0x%x, currvid 0x%x, reqfid 0x%x\n",
+ dprintk("ph2 (cpu%d): starting, currfid 0x%x, currvid 0x%x, "
+ "reqfid 0x%x\n",
smp_processor_id(),
data->currfid, data->currvid, reqfid);
if (reqfid > data->currfid) {
if (data->currfid > LO_FID_TABLE_TOP) {
- if (write_new_fid(data, data->currfid + fid_interval)) {
+ if (write_new_fid(data,
+ data->currfid + fid_interval))
return 1;
- }
} else {
if (write_new_fid
- (data, 2 + convert_fid_to_vco_fid(data->currfid))) {
+ (data,
+ 2 + convert_fid_to_vco_fid(data->currfid)))
return 1;
- }
}
} else {
if (write_new_fid(data, data->currfid - fid_interval))
if (data->currfid != reqfid) {
printk(KERN_ERR PFX
- "ph2: mismatch, failed fid transition, curr 0x%x, req 0x%x\n",
+ "ph2: mismatch, failed fid transition, "
+ "curr 0x%x, req 0x%x\n",
data->currfid, reqfid);
return 1;
}
}
/* Phase 3 - core voltage transition flow ... jump to the final vid. */
- static int core_voltage_post_transition(struct powernow_k8_data *data, u32 reqvid)
+ static int core_voltage_post_transition(struct powernow_k8_data *data,
+ u32 reqvid)
{
u32 savefid = data->currfid;
u32 savereqvid = reqvid;
if (data->currvid != reqvid) {
printk(KERN_ERR PFX
- "ph3: failed vid transition\n, req 0x%x, curr 0x%x",
+ "ph3: failed vid transition\n, "
+ "req 0x%x, curr 0x%x",
reqvid, data->currvid);
return 1;
}
if ((eax & CPUID_XFAM) == CPUID_XFAM_K8) {
if (((eax & CPUID_USE_XFAM_XMOD) != CPUID_USE_XFAM_XMOD) ||
((eax & CPUID_XMOD) > CPUID_XMOD_REV_MASK)) {
- printk(KERN_INFO PFX "Processor cpuid %x not supported\n", eax);
+ printk(KERN_INFO PFX
+ "Processor cpuid %x not supported\n", eax);
goto out;
}
}
cpuid(CPUID_FREQ_VOLT_CAPABILITIES, &eax, &ebx, &ecx, &edx);
- if ((edx & P_STATE_TRANSITION_CAPABLE) != P_STATE_TRANSITION_CAPABLE) {
- printk(KERN_INFO PFX "Power state transitions not supported\n");
+ if ((edx & P_STATE_TRANSITION_CAPABLE)
+ != P_STATE_TRANSITION_CAPABLE) {
+ printk(KERN_INFO PFX
+ "Power state transitions not supported\n");
goto out;
}
} else { /* must be a HW Pstate capable processor */
return rc;
}
- static int check_pst_table(struct powernow_k8_data *data, struct pst_s *pst, u8 maxvid)
+ static int check_pst_table(struct powernow_k8_data *data, struct pst_s *pst,
+ u8 maxvid)
{
unsigned int j;
u8 lastfid = 0xff;
j, pst[j].vid);
return -EINVAL;
}
- if (pst[j].vid < data->rvo) { /* vid + rvo >= 0 */
+ if (pst[j].vid < data->rvo) {
+ /* vid + rvo >= 0 */
printk(KERN_ERR FW_BUG PFX "0 vid exceeded with pstate"
" %d\n", j);
return -ENODEV;
}
- if (pst[j].vid < maxvid + data->rvo) { /* vid + rvo >= maxvid */
+ if (pst[j].vid < maxvid + data->rvo) {
+ /* vid + rvo >= maxvid */
printk(KERN_ERR FW_BUG PFX "maxvid exceeded with pstate"
" %d\n", j);
return -ENODEV;
return -EINVAL;
}
if (lastfid > LO_FID_TABLE_TOP)
- printk(KERN_INFO FW_BUG PFX "first fid not from lo freq table\n");
+ printk(KERN_INFO FW_BUG PFX
+ "first fid not from lo freq table\n");
return 0;
}
+ static void invalidate_entry(struct powernow_k8_data *data, unsigned int entry)
+ {
+ data->powernow_table[entry].frequency = CPUFREQ_ENTRY_INVALID;
+ }
+
static void print_basics(struct powernow_k8_data *data)
{
int j;
for (j = 0; j < data->numps; j++) {
- if (data->powernow_table[j].frequency != CPUFREQ_ENTRY_INVALID) {
+ if (data->powernow_table[j].frequency !=
+ CPUFREQ_ENTRY_INVALID) {
if (cpu_family == CPU_HW_PSTATE) {
- printk(KERN_INFO PFX " %d : pstate %d (%d MHz)\n",
- j,
+ printk(KERN_INFO PFX
+ " %d : pstate %d (%d MHz)\n", j,
data->powernow_table[j].index,
data->powernow_table[j].frequency/1000);
} else {
- printk(KERN_INFO PFX " %d : fid 0x%x (%d MHz), vid 0x%x\n",
+ printk(KERN_INFO PFX
+ " %d : fid 0x%x (%d MHz), vid 0x%x\n",
j,
data->powernow_table[j].index & 0xff,
data->powernow_table[j].frequency/1000,
}
}
if (data->batps)
- printk(KERN_INFO PFX "Only %d pstates on battery\n", data->batps);
+ printk(KERN_INFO PFX "Only %d pstates on battery\n",
+ data->batps);
}
- static int fill_powernow_table(struct powernow_k8_data *data, struct pst_s *pst, u8 maxvid)
+ static int fill_powernow_table(struct powernow_k8_data *data,
+ struct pst_s *pst, u8 maxvid)
{
struct cpufreq_frequency_table *powernow_table;
unsigned int j;
- if (data->batps) { /* use ACPI support to get full speed on mains power */
- printk(KERN_WARNING PFX "Only %d pstates usable (use ACPI driver for full range\n", data->batps);
+ if (data->batps) {
+ /* use ACPI support to get full speed on mains power */
+ printk(KERN_WARNING PFX
+ "Only %d pstates usable (use ACPI driver for full "
+ "range\n", data->batps);
data->numps = data->batps;
}
- for ( j=1; j<data->numps; j++ ) {
+ for (j = 1; j < data->numps; j++) {
if (pst[j-1].fid >= pst[j].fid) {
printk(KERN_ERR PFX "PST out of sequence\n");
return -EINVAL;
}
for (j = 0; j < data->numps; j++) {
+ int freq;
powernow_table[j].index = pst[j].fid; /* lower 8 bits */
powernow_table[j].index |= (pst[j].vid << 8); /* upper 8 bits */
- powernow_table[j].frequency = find_khz_freq_from_fid(pst[j].fid);
+ freq = find_khz_freq_from_fid(pst[j].fid);
+ powernow_table[j].frequency = freq;
}
powernow_table[data->numps].frequency = CPUFREQ_TABLE_END;
powernow_table[data->numps].index = 0;
dprintk("cfid 0x%x, cvid 0x%x\n", data->currfid, data->currvid);
data->powernow_table = powernow_table;
- if (first_cpu(per_cpu(cpu_core_map, data->cpu)) == data->cpu)
+ if (cpumask_first(cpu_core_mask(data->cpu)) == data->cpu)
print_basics(data);
for (j = 0; j < data->numps; j++)
- if ((pst[j].fid==data->currfid) && (pst[j].vid==data->currvid))
+ if ((pst[j].fid == data->currfid) &&
+ (pst[j].vid == data->currvid))
return 0;
dprintk("currfid/vid do not match PST, ignoring\n");
}
data->vstable = psb->vstable;
- dprintk("voltage stabilization time: %d(*20us)\n", data->vstable);
+ dprintk("voltage stabilization time: %d(*20us)\n",
+ data->vstable);
dprintk("flags2: 0x%x\n", psb->flags2);
data->rvo = psb->flags2 & 3;
dprintk("numpst: 0x%x\n", psb->num_tables);
cpst = psb->num_tables;
- if ((psb->cpuid == 0x00000fc0) || (psb->cpuid == 0x00000fe0) ){
+ if ((psb->cpuid == 0x00000fc0) ||
+ (psb->cpuid == 0x00000fe0)) {
thiscpuid = cpuid_eax(CPUID_PROCESSOR_SIGNATURE);
- if ((thiscpuid == 0x00000fc0) || (thiscpuid == 0x00000fe0) ) {
+ if ((thiscpuid == 0x00000fc0) ||
+ (thiscpuid == 0x00000fe0))
cpst = 1;
- }
}
if (cpst != 1) {
printk(KERN_ERR FW_BUG PFX "numpst must be 1\n");
data->numps = psb->numps;
dprintk("numpstates: 0x%x\n", data->numps);
- return fill_powernow_table(data, (struct pst_s *)(psb+1), maxvid);
+ return fill_powernow_table(data,
+ (struct pst_s *)(psb+1), maxvid);
}
/*
* If you see this message, complain to BIOS manufacturer. If
* BIOS and Kernel Developer's Guide, which is available on
* www.amd.com
*/
- printk(KERN_ERR PFX "BIOS error - no PSB or ACPI _PSS objects\n");
+ printk(KERN_ERR FW_BUG PFX "No PSB or ACPI _PSS objects\n");
return -ENODEV;
}
- #ifdef CONFIG_X86_POWERNOW_K8_ACPI
- static void powernow_k8_acpi_pst_values(struct powernow_k8_data *data, unsigned int index)
+ static void powernow_k8_acpi_pst_values(struct powernow_k8_data *data,
+ unsigned int index)
{
+ acpi_integer control;
+
if (!data->acpi_data.state_count || (cpu_family == CPU_HW_PSTATE))
return;
- data->irt = (data->acpi_data.states[index].control >> IRT_SHIFT) & IRT_MASK;
- data->rvo = (data->acpi_data.states[index].control >> RVO_SHIFT) & RVO_MASK;
- data->exttype = (data->acpi_data.states[index].control >> EXT_TYPE_SHIFT) & EXT_TYPE_MASK;
- data->plllock = (data->acpi_data.states[index].control >> PLL_L_SHIFT) & PLL_L_MASK;
- data->vidmvs = 1 << ((data->acpi_data.states[index].control >> MVS_SHIFT) & MVS_MASK);
- data->vstable = (data->acpi_data.states[index].control >> VST_SHIFT) & VST_MASK;
- }
+ control = data->acpi_data.states[index].control; data->irt = (control
+ >> IRT_SHIFT) & IRT_MASK; data->rvo = (control >>
+ RVO_SHIFT) & RVO_MASK; data->exttype = (control
+ >> EXT_TYPE_SHIFT) & EXT_TYPE_MASK;
+ data->plllock = (control >> PLL_L_SHIFT) & PLL_L_MASK; data->vidmvs = 1
+ << ((control >> MVS_SHIFT) & MVS_MASK); data->vstable =
+ (control >> VST_SHIFT) & VST_MASK; }
static int powernow_k8_cpu_init_acpi(struct powernow_k8_data *data)
{
struct cpufreq_frequency_table *powernow_table;
int ret_val = -ENODEV;
+ acpi_integer space_id;
if (acpi_processor_register_performance(&data->acpi_data, data->cpu)) {
dprintk("register performance failed: bad ACPI data\n");
goto err_out;
}
- if ((data->acpi_data.control_register.space_id != ACPI_ADR_SPACE_FIXED_HARDWARE) ||
- (data->acpi_data.status_register.space_id != ACPI_ADR_SPACE_FIXED_HARDWARE)) {
+ space_id = data->acpi_data.control_register.space_id;
+ if ((space_id != ACPI_ADR_SPACE_FIXED_HARDWARE) ||
+ (space_id != ACPI_ADR_SPACE_FIXED_HARDWARE)) {
dprintk("Invalid control/status registers (%x - %x)\n",
data->acpi_data.control_register.space_id,
- data->acpi_data.status_register.space_id);
+ space_id);
goto err_out;
}
if (ret_val)
goto err_out_mem;
- powernow_table[data->acpi_data.state_count].frequency = CPUFREQ_TABLE_END;
+ powernow_table[data->acpi_data.state_count].frequency =
+ CPUFREQ_TABLE_END;
powernow_table[data->acpi_data.state_count].index = 0;
data->powernow_table = powernow_table;
/* fill in data */
data->numps = data->acpi_data.state_count;
- if (first_cpu(per_cpu(cpu_core_map, data->cpu)) == data->cpu)
+ if (cpumask_first(cpu_core_mask(data->cpu)) == data->cpu)
print_basics(data);
powernow_k8_acpi_pst_values(data, 0);
err_out:
acpi_processor_unregister_performance(&data->acpi_data, data->cpu);
- /* data->acpi_data.state_count informs us at ->exit() whether ACPI was used */
+ /* data->acpi_data.state_count informs us at ->exit()
+ * whether ACPI was used */
data->acpi_data.state_count = 0;
return ret_val;
}
- static int fill_powernow_table_pstate(struct powernow_k8_data *data, struct cpufreq_frequency_table *powernow_table)
+ static int fill_powernow_table_pstate(struct powernow_k8_data *data,
+ struct cpufreq_frequency_table *powernow_table)
{
int i;
u32 hi = 0, lo = 0;
index = data->acpi_data.states[i].control & HW_PSTATE_MASK;
if (index > data->max_hw_pstate) {
- printk(KERN_ERR PFX "invalid pstate %d - bad value %d.\n", i, index);
- printk(KERN_ERR PFX "Please report to BIOS manufacturer\n");
- powernow_table[i].frequency = CPUFREQ_ENTRY_INVALID;
+ printk(KERN_ERR PFX "invalid pstate %d - "
+ "bad value %d.\n", i, index);
+ printk(KERN_ERR PFX "Please report to BIOS "
+ "manufacturer\n");
+ invalidate_entry(data, i);
continue;
}
rdmsr(MSR_PSTATE_DEF_BASE + index, lo, hi);
if (!(hi & HW_PSTATE_VALID_MASK)) {
dprintk("invalid pstate %d, ignoring\n", index);
- powernow_table[i].frequency = CPUFREQ_ENTRY_INVALID;
+ invalidate_entry(data, i);
continue;
}
powernow_table[i].index = index;
- powernow_table[i].frequency = data->acpi_data.states[i].core_frequency * 1000;
+ powernow_table[i].frequency =
+ data->acpi_data.states[i].core_frequency * 1000;
}
return 0;
}
- static int fill_powernow_table_fidvid(struct powernow_k8_data *data, struct cpufreq_frequency_table *powernow_table)
+ static int fill_powernow_table_fidvid(struct powernow_k8_data *data,
+ struct cpufreq_frequency_table *powernow_table)
{
int i;
int cntlofreq = 0;
+
for (i = 0; i < data->acpi_data.state_count; i++) {
u32 fid;
u32 vid;
+ u32 freq, index;
+ acpi_integer status, control;
if (data->exttype) {
- fid = data->acpi_data.states[i].status & EXT_FID_MASK;
- vid = (data->acpi_data.states[i].status >> VID_SHIFT) & EXT_VID_MASK;
+ status = data->acpi_data.states[i].status;
+ fid = status & EXT_FID_MASK;
+ vid = (status >> VID_SHIFT) & EXT_VID_MASK;
} else {
- fid = data->acpi_data.states[i].control & FID_MASK;
- vid = (data->acpi_data.states[i].control >> VID_SHIFT) & VID_MASK;
+ control = data->acpi_data.states[i].control;
+ fid = control & FID_MASK;
+ vid = (control >> VID_SHIFT) & VID_MASK;
}
dprintk(" %d : fid 0x%x, vid 0x%x\n", i, fid, vid);
- powernow_table[i].index = fid; /* lower 8 bits */
- powernow_table[i].index |= (vid << 8); /* upper 8 bits */
- powernow_table[i].frequency = find_khz_freq_from_fid(fid);
+ index = fid | (vid<<8);
+ powernow_table[i].index = index;
+
+ freq = find_khz_freq_from_fid(fid);
+ powernow_table[i].frequency = freq;
/* verify frequency is OK */
- if ((powernow_table[i].frequency > (MAX_FREQ * 1000)) ||
- (powernow_table[i].frequency < (MIN_FREQ * 1000))) {
- dprintk("invalid freq %u kHz, ignoring\n", powernow_table[i].frequency);
- powernow_table[i].frequency = CPUFREQ_ENTRY_INVALID;
+ if ((freq > (MAX_FREQ * 1000)) || (freq < (MIN_FREQ * 1000))) {
+ dprintk("invalid freq %u kHz, ignoring\n", freq);
+ invalidate_entry(data, i);
continue;
}
- /* verify voltage is OK - BIOSs are using "off" to indicate invalid */
+ /* verify voltage is OK -
+ * BIOSs are using "off" to indicate invalid */
if (vid == VID_OFF) {
dprintk("invalid vid %u, ignoring\n", vid);
- powernow_table[i].frequency = CPUFREQ_ENTRY_INVALID;
+ invalidate_entry(data, i);
continue;
}
/* verify only 1 entry from the lo frequency table */
if (fid < HI_FID_TABLE_BOTTOM) {
if (cntlofreq) {
- /* if both entries are the same, ignore this one ... */
- if ((powernow_table[i].frequency != powernow_table[cntlofreq].frequency) ||
- (powernow_table[i].index != powernow_table[cntlofreq].index)) {
- printk(KERN_ERR PFX "Too many lo freq table entries\n");
+ /* if both entries are the same,
+ * ignore this one ... */
+ if ((freq != powernow_table[cntlofreq].frequency) ||
+ (index != powernow_table[cntlofreq].index)) {
+ printk(KERN_ERR PFX
+ "Too many lo freq table "
+ "entries\n");
return 1;
}
- dprintk("double low frequency table entry, ignoring it.\n");
- powernow_table[i].frequency = CPUFREQ_ENTRY_INVALID;
+ dprintk("double low frequency table entry, "
+ "ignoring it.\n");
+ invalidate_entry(data, i);
continue;
} else
cntlofreq = i;
}
- if (powernow_table[i].frequency != (data->acpi_data.states[i].core_frequency * 1000)) {
- printk(KERN_INFO PFX "invalid freq entries %u kHz vs. %u kHz\n",
- powernow_table[i].frequency,
- (unsigned int) (data->acpi_data.states[i].core_frequency * 1000));
- powernow_table[i].frequency = CPUFREQ_ENTRY_INVALID;
+ if (freq != (data->acpi_data.states[i].core_frequency * 1000)) {
+ printk(KERN_INFO PFX "invalid freq entries "
+ "%u kHz vs. %u kHz\n", freq,
+ (unsigned int)
+ (data->acpi_data.states[i].core_frequency
+ * 1000));
+ invalidate_entry(data, i);
continue;
}
}
static void powernow_k8_cpu_exit_acpi(struct powernow_k8_data *data)
{
if (data->acpi_data.state_count)
- acpi_processor_unregister_performance(&data->acpi_data, data->cpu);
+ acpi_processor_unregister_performance(&data->acpi_data,
+ data->cpu);
free_cpumask_var(data->acpi_data.shared_cpu_map);
}
return 1000 * max_latency;
}
- #else
- static int powernow_k8_cpu_init_acpi(struct powernow_k8_data *data) { return -ENODEV; }
- static void powernow_k8_cpu_exit_acpi(struct powernow_k8_data *data) { return; }
- static void powernow_k8_acpi_pst_values(struct powernow_k8_data *data, unsigned int index) { return; }
- static int get_transition_latency(struct powernow_k8_data *data) { return 0; }
- #endif /* CONFIG_X86_POWERNOW_K8_ACPI */
-
/* Take a frequency, and issue the fid/vid transition command */
- static int transition_frequency_fidvid(struct powernow_k8_data *data, unsigned int index)
+ static int transition_frequency_fidvid(struct powernow_k8_data *data,
+ unsigned int index)
{
u32 fid = 0;
u32 vid = 0;
return 0;
}
- if ((fid < HI_FID_TABLE_BOTTOM) && (data->currfid < HI_FID_TABLE_BOTTOM)) {
+ if ((fid < HI_FID_TABLE_BOTTOM) &&
+ (data->currfid < HI_FID_TABLE_BOTTOM)) {
printk(KERN_ERR PFX
"ignoring illegal change in lo freq table-%x to 0x%x\n",
data->currfid, fid);
}
/* Take a frequency, and issue the hardware pstate transition command */
- static int transition_frequency_pstate(struct powernow_k8_data *data, unsigned int index)
+ static int transition_frequency_pstate(struct powernow_k8_data *data,
+ unsigned int index)
{
u32 pstate = 0;
int res, i;
pstate = index & HW_PSTATE_MASK;
if (pstate > data->max_hw_pstate)
return 0;
- freqs.old = find_khz_freq_from_pstate(data->powernow_table, data->currpstate);
+ freqs.old = find_khz_freq_from_pstate(data->powernow_table,
+ data->currpstate);
freqs.new = find_khz_freq_from_pstate(data->powernow_table, pstate);
for_each_cpu_mask_nr(i, *(data->available_cores)) {
}
/* Driver entry point to switch to the target frequency */
- static int powernowk8_target(struct cpufreq_policy *pol, unsigned targfreq, unsigned relation)
+ static int powernowk8_target(struct cpufreq_policy *pol,
+ unsigned targfreq, unsigned relation)
{
cpumask_t oldmask;
struct powernow_k8_data *data = per_cpu(powernow_data, pol->cpu);
dprintk("targ: curr fid 0x%x, vid 0x%x\n",
data->currfid, data->currvid);
- if ((checkvid != data->currvid) || (checkfid != data->currfid)) {
+ if ((checkvid != data->currvid) ||
+ (checkfid != data->currfid)) {
printk(KERN_INFO PFX
- "error - out of sync, fix 0x%x 0x%x, vid 0x%x 0x%x\n",
- checkfid, data->currfid, checkvid, data->currvid);
+ "error - out of sync, fix 0x%x 0x%x, "
+ "vid 0x%x 0x%x\n",
+ checkfid, data->currfid,
+ checkvid, data->currvid);
}
}
- if (cpufreq_frequency_table_target(pol, data->powernow_table, targfreq, relation, &newstate))
+ if (cpufreq_frequency_table_target(pol, data->powernow_table,
+ targfreq, relation, &newstate))
goto err_out;
mutex_lock(&fidvid_mutex);
mutex_unlock(&fidvid_mutex);
if (cpu_family == CPU_HW_PSTATE)
- pol->cur = find_khz_freq_from_pstate(data->powernow_table, newstate);
+ pol->cur = find_khz_freq_from_pstate(data->powernow_table,
+ newstate);
else
pol->cur = find_khz_freq_from_fid(data->currfid);
ret = 0;
struct powernow_k8_data *data;
cpumask_t oldmask;
int rc;
+ static int print_once;
if (!cpu_online(pol->cpu))
return -ENODEV;
* an UP version, and is deprecated by AMD.
*/
if (num_online_cpus() != 1) {
- #ifndef CONFIG_ACPI_PROCESSOR
- printk(KERN_ERR PFX "ACPI Processor support is required "
- "for SMP systems but is absent. Please load the "
- "ACPI Processor module before starting this "
- "driver.\n");
- #else
- printk(KERN_ERR FW_BUG PFX "Your BIOS does not provide"
- " ACPI _PSS objects in a way that Linux "
- "understands. Please report this to the Linux "
- "ACPI maintainers and complain to your BIOS "
- "vendor.\n");
- #endif
- kfree(data);
- return -ENODEV;
+ /*
+ * Replace this one with print_once as soon as such a
+ * thing gets introduced
+ */
+ if (!print_once) {
+ WARN_ONCE(1, KERN_ERR FW_BUG PFX "Your BIOS "
+ "does not provide ACPI _PSS objects "
+ "in a way that Linux understands. "
+ "Please report this to the Linux ACPI"
+ " maintainers and complain to your "
+ "BIOS vendor.\n");
+ print_once++;
+ }
+ goto err_out;
}
if (pol->cpu != 0) {
printk(KERN_ERR FW_BUG PFX "No ACPI _PSS objects for "
"CPU other than CPU0. Complain to your BIOS "
"vendor.\n");
- kfree(data);
- return -ENODEV;
+ goto err_out;
}
rc = find_psb_table(data);
- if (rc) {
- kfree(data);
- return -ENODEV;
- }
+ if (rc)
+ goto err_out;
+
/* Take a crude guess here.
* That guess was in microseconds, so multiply with 1000 */
pol->cpuinfo.transition_latency = (
if (smp_processor_id() != pol->cpu) {
printk(KERN_ERR PFX "limiting to cpu %u failed\n", pol->cpu);
- goto err_out;
+ goto err_out_unmask;
}
if (pending_bit_stuck()) {
printk(KERN_ERR PFX "failing init, change pending bit set\n");
- goto err_out;
+ goto err_out_unmask;
}
if (query_current_values_with_pending_wait(data))
- goto err_out;
+ goto err_out_unmask;
if (cpu_family == CPU_OPTERON)
fidvid_msr_init();
if (cpu_family == CPU_HW_PSTATE)
cpumask_copy(pol->cpus, cpumask_of(pol->cpu));
else
- cpumask_copy(pol->cpus, &per_cpu(cpu_core_map, pol->cpu));
+ cpumask_copy(pol->cpus, cpu_core_mask(pol->cpu));
data->available_cores = pol->cpus;
if (cpu_family == CPU_HW_PSTATE)
- pol->cur = find_khz_freq_from_pstate(data->powernow_table, data->currpstate);
+ pol->cur = find_khz_freq_from_pstate(data->powernow_table,
+ data->currpstate);
else
pol->cur = find_khz_freq_from_fid(data->currfid);
dprintk("policy current frequency %d kHz\n", pol->cur);
cpufreq_frequency_table_get_attr(data->powernow_table, pol->cpu);
if (cpu_family == CPU_HW_PSTATE)
- dprintk("cpu_init done, current pstate 0x%x\n", data->currpstate);
+ dprintk("cpu_init done, current pstate 0x%x\n",
+ data->currpstate);
else
dprintk("cpu_init done, current fid 0x%x, vid 0x%x\n",
data->currfid, data->currvid);
return 0;
- err_out:
+ err_out_unmask:
set_cpus_allowed_ptr(current, &oldmask);
powernow_k8_cpu_exit_acpi(data);
+ err_out:
kfree(data);
return -ENODEV;
}
- static int __devexit powernowk8_cpu_exit (struct cpufreq_policy *pol)
+ static int __devexit powernowk8_cpu_exit(struct cpufreq_policy *pol)
{
struct powernow_k8_data *data = per_cpu(powernow_data, pol->cpu);
return 0;
}
- static unsigned int powernowk8_get (unsigned int cpu)
+ static unsigned int powernowk8_get(unsigned int cpu)
{
struct powernow_k8_data *data;
cpumask_t oldmask = current->cpus_allowed;
unsigned int khz = 0;
unsigned int first;
- first = first_cpu(per_cpu(cpu_core_map, cpu));
+ first = cpumask_first(cpu_core_mask(cpu));
data = per_cpu(powernow_data, first);
if (!data)
return khz;
}
- static struct freq_attr* powernow_k8_attr[] = {
+ static struct freq_attr *powernow_k8_attr[] = {
&cpufreq_freq_attr_scaling_available_freqs,
NULL,
};
cpufreq_unregister_driver(&cpufreq_amd64_driver);
}
- MODULE_AUTHOR("Paul Devriendt <paul.devriendt@amd.com> and Mark Langsdorf <mark.langsdorf@amd.com>");
+ MODULE_AUTHOR("Paul Devriendt <paul.devriendt@amd.com> and "
+ "Mark Langsdorf <mark.langsdorf@amd.com>");
MODULE_DESCRIPTION("AMD Athlon 64 and Opteron processor frequency driver.");
MODULE_LICENSE("GPL");
/* speedstep_processor
*/
- static unsigned int speedstep_processor = 0;
+ static unsigned int speedstep_processor;
static u32 pmbase;
};
- #define dprintk(msg...) cpufreq_debug_printk(CPUFREQ_DEBUG_DRIVER, "speedstep-ich", msg)
+ #define dprintk(msg...) cpufreq_debug_printk(CPUFREQ_DEBUG_DRIVER, \
+ "speedstep-ich", msg)
/**
*
* Returns: -ENODEV if no register could be found
*/
- static int speedstep_find_register (void)
+ static int speedstep_find_register(void)
{
if (!speedstep_chipset_dev)
return -ENODEV;
*
* Tries to change the SpeedStep state.
*/
- static void speedstep_set_state (unsigned int state)
+ static void speedstep_set_state(unsigned int state)
{
u8 pm2_blk;
u8 value;
dprintk("read at pmbase 0x%x + 0x50 returned 0x%x\n", pmbase, value);
- if (state == (value & 0x1)) {
- dprintk("change to %u MHz succeeded\n", (speedstep_get_processor_frequency(speedstep_processor) / 1000));
- } else {
- printk (KERN_ERR "cpufreq: change failed - I/O error\n");
- }
+ if (state == (value & 0x1))
+ dprintk("change to %u MHz succeeded\n",
+ speedstep_get_frequency(speedstep_processor) / 1000);
+ else
+ printk(KERN_ERR "cpufreq: change failed - I/O error\n");
return;
}
* Tries to activate the SpeedStep status and control registers.
* Returns -EINVAL on an unsupported chipset, and zero on success.
*/
- static int speedstep_activate (void)
+ static int speedstep_activate(void)
{
u16 value = 0;
* functions. Returns the SPEEDSTEP_CHIPSET_-number for the detected
* chipset, or zero on failure.
*/
- static unsigned int speedstep_detect_chipset (void)
+ static unsigned int speedstep_detect_chipset(void)
{
speedstep_chipset_dev = pci_get_subsys(PCI_VENDOR_ID_INTEL,
PCI_DEVICE_ID_INTEL_82801DB_12,
- PCI_ANY_ID,
- PCI_ANY_ID,
+ PCI_ANY_ID, PCI_ANY_ID,
NULL);
if (speedstep_chipset_dev)
return 4; /* 4-M */
speedstep_chipset_dev = pci_get_subsys(PCI_VENDOR_ID_INTEL,
PCI_DEVICE_ID_INTEL_82801CA_12,
- PCI_ANY_ID,
- PCI_ANY_ID,
+ PCI_ANY_ID, PCI_ANY_ID,
NULL);
if (speedstep_chipset_dev)
return 3; /* 3-M */
speedstep_chipset_dev = pci_get_subsys(PCI_VENDOR_ID_INTEL,
PCI_DEVICE_ID_INTEL_82801BA_10,
- PCI_ANY_ID,
- PCI_ANY_ID,
+ PCI_ANY_ID, PCI_ANY_ID,
NULL);
if (speedstep_chipset_dev) {
/* speedstep.c causes lockups on Dell Inspirons 8000 and
hostbridge = pci_get_subsys(PCI_VENDOR_ID_INTEL,
PCI_DEVICE_ID_INTEL_82815_MC,
- PCI_ANY_ID,
- PCI_ANY_ID,
+ PCI_ANY_ID, PCI_ANY_ID,
NULL);
if (!hostbridge)
cpus_allowed = current->cpus_allowed;
set_cpus_allowed_ptr(current, cpus);
- speed = speedstep_get_processor_frequency(speedstep_processor);
+ speed = speedstep_get_frequency(speedstep_processor);
set_cpus_allowed_ptr(current, &cpus_allowed);
dprintk("detected %u kHz as current frequency\n", speed);
return speed;
* speedstep_target - set a new CPUFreq policy
* @policy: new policy
* @target_freq: the target frequency
- * @relation: how that frequency relates to achieved frequency (CPUFREQ_RELATION_L or CPUFREQ_RELATION_H)
+ * @relation: how that frequency relates to achieved frequency
+ * (CPUFREQ_RELATION_L or CPUFREQ_RELATION_H)
*
* Sets a new CPUFreq policy.
*/
- static int speedstep_target (struct cpufreq_policy *policy,
+ static int speedstep_target(struct cpufreq_policy *policy,
unsigned int target_freq,
unsigned int relation)
{
cpumask_t cpus_allowed;
int i;
- if (cpufreq_frequency_table_target(policy, &speedstep_freqs[0], target_freq, relation, &newstate))
+ if (cpufreq_frequency_table_target(policy, &speedstep_freqs[0],
+ target_freq, relation, &newstate))
return -EINVAL;
freqs.old = _speedstep_get(policy->cpus);
* Limit must be within speedstep_low_freq and speedstep_high_freq, with
* at least one border included.
*/
- static int speedstep_verify (struct cpufreq_policy *policy)
+ static int speedstep_verify(struct cpufreq_policy *policy)
{
return cpufreq_frequency_table_verify(policy, &speedstep_freqs[0]);
}
/* only run on CPU to be set, or on its sibling */
#ifdef CONFIG_SMP
- cpumask_copy(policy->cpus, &per_cpu(cpu_sibling_map, policy->cpu));
+ cpumask_copy(policy->cpus, cpu_sibling_mask(policy->cpu));
#endif
cpus_allowed = current->cpus_allowed;
return -EIO;
dprintk("currently at %s speed setting - %i MHz\n",
- (speed == speedstep_freqs[SPEEDSTEP_LOW].frequency) ? "low" : "high",
+ (speed == speedstep_freqs[SPEEDSTEP_LOW].frequency)
+ ? "low" : "high",
(speed / 1000));
/* cpuinfo and default policy values */
result = cpufreq_frequency_table_cpuinfo(policy, speedstep_freqs);
if (result)
- return (result);
+ return result;
- cpufreq_frequency_table_get_attr(speedstep_freqs, policy->cpu);
+ cpufreq_frequency_table_get_attr(speedstep_freqs, policy->cpu);
return 0;
}
return 0;
}
- static struct freq_attr* speedstep_attr[] = {
+ static struct freq_attr *speedstep_attr[] = {
&cpufreq_freq_attr_scaling_available_freqs,
NULL,
};
/* detect processor */
speedstep_processor = speedstep_detect_processor();
if (!speedstep_processor) {
- dprintk("Intel(R) SpeedStep(TM) capable processor not found\n");
+ dprintk("Intel(R) SpeedStep(TM) capable processor "
+ "not found\n");
return -ENODEV;
}
/* detect chipset */
if (!speedstep_detect_chipset()) {
- dprintk("Intel(R) SpeedStep(TM) for this chipset not (yet) available.\n");
+ dprintk("Intel(R) SpeedStep(TM) for this chipset not "
+ "(yet) available.\n");
return -ENODEV;
}
}
- MODULE_AUTHOR ("Dave Jones <davej@redhat.com>, Dominik Brodowski <linux@brodo.de>");
- MODULE_DESCRIPTION ("Speedstep driver for Intel mobile processors on chipsets with ICH-M southbridges.");
- MODULE_LICENSE ("GPL");
+ MODULE_AUTHOR("Dave Jones <davej@redhat.com>, "
+ "Dominik Brodowski <linux@brodo.de>");
+ MODULE_DESCRIPTION("Speedstep driver for Intel mobile processors on chipsets "
+ "with ICH-M southbridges.");
+ MODULE_LICENSE("GPL");
module_init(speedstep_init);
module_exit(speedstep_exit);
};
/* all the cache descriptor types we care about (no TLB or trace cache entries) */
- static struct _cache_table cache_table[] __cpuinitdata =
+ static const struct _cache_table __cpuinitconst cache_table[] =
{
{ 0x06, LVL_1_INST, 8 }, /* 4-way set assoc, 32 byte line size */
{ 0x08, LVL_1_INST, 16 }, /* 4-way set assoc, 32 byte line size */
unsigned long can_disable;
};
-#ifdef CONFIG_PCI
+#if defined(CONFIG_PCI) && defined(CONFIG_SYSFS)
static struct pci_device_id k8_nb_id[] = {
{ PCI_DEVICE(PCI_VENDOR_ID_AMD, 0x1103) },
{ PCI_DEVICE(PCI_VENDOR_ID_AMD, 0x1203) },
unsigned val;
};
- static unsigned short assocs[] __cpuinitdata = {
+ static const unsigned short __cpuinitconst assocs[] = {
[1] = 1, [2] = 2, [4] = 4, [6] = 8,
[8] = 16, [0xa] = 32, [0xb] = 48,
[0xc] = 64,
[0xf] = 0xffff // ??
};
- static unsigned char levels[] __cpuinitdata = { 1, 1, 2, 3 };
- static unsigned char types[] __cpuinitdata = { 1, 2, 3, 3 };
+ static const unsigned char __cpuinitconst levels[] = { 1, 1, 2, 3 };
+ static const unsigned char __cpuinitconst types[] = { 1, 2, 3, 3 };
static void __cpuinit
amd_cpuid4(int leaf, union _cpuid4_leaf_eax *eax,
return 0;
}
-static int
-__cpuinit cpuid4_cache_lookup(int index, struct _cpuid4_info *this_leaf)
-{
- struct _cpuid4_info_regs *leaf_regs =
- (struct _cpuid4_info_regs *)this_leaf;
-
- return cpuid4_cache_lookup_regs(index, leaf_regs);
-}
-
static int __cpuinit find_num_cache_leaves(void)
{
unsigned int eax, ebx, ecx, edx;
return l2;
}
+#ifdef CONFIG_SYSFS
+
/* pointer to _cpuid4_info array (for each cache leaf) */
static DEFINE_PER_CPU(struct _cpuid4_info *, cpuid4_info);
#define CPUID4_INFO_IDX(x, y) (&((per_cpu(cpuid4_info, x))[y]))
per_cpu(cpuid4_info, cpu) = NULL;
}
+static int
+__cpuinit cpuid4_cache_lookup(int index, struct _cpuid4_info *this_leaf)
+{
+ struct _cpuid4_info_regs *leaf_regs =
+ (struct _cpuid4_info_regs *)this_leaf;
+
+ return cpuid4_cache_lookup_regs(index, leaf_regs);
+}
+
static void __cpuinit get_cpu_leaves(void *_retval)
{
int j, *retval = _retval, cpu = smp_processor_id();
return retval;
}
-#ifdef CONFIG_SYSFS
-
#include <linux/kobject.h>
#include <linux/sysfs.h>
};
/* must be called with correct cpu affinity */
- static long threshold_restart_bank(void *_tr)
+ /* Called via smp_call_function_single() */
+ static void threshold_restart_bank(void *_tr)
{
struct thresh_restart *tr = _tr;
u32 mci_misc_hi, mci_misc_lo;
mci_misc_hi |= MASK_COUNT_EN_HI;
wrmsr(tr->b->address, mci_misc_lo, mci_misc_hi);
- return 0;
}
/* cpu init entry point, called from mce.c with preempt off */
tr.b = b;
tr.reset = 0;
tr.old_limit = 0;
- work_on_cpu(b->cpu, threshold_restart_bank, &tr);
+ smp_call_function_single(b->cpu, threshold_restart_bank, &tr, 1);
return end - buf;
}
tr.b = b;
tr.reset = 0;
- work_on_cpu(b->cpu, threshold_restart_bank, &tr);
+ smp_call_function_single(b->cpu, threshold_restart_bank, &tr, 1);
return end - buf;
}
- static long local_error_count(void *_b)
+ struct threshold_block_cross_cpu {
+ struct threshold_block *tb;
+ long retval;
+ };
+
+ static void local_error_count_handler(void *_tbcc)
{
- struct threshold_block *b = _b;
+ struct threshold_block_cross_cpu *tbcc = _tbcc;
+ struct threshold_block *b = tbcc->tb;
u32 low, high;
rdmsr(b->address, low, high);
- return (high & 0xFFF) - (THRESHOLD_MAX - b->threshold_limit);
+ tbcc->retval = (high & 0xFFF) - (THRESHOLD_MAX - b->threshold_limit);
}
static ssize_t show_error_count(struct threshold_block *b, char *buf)
{
- return sprintf(buf, "%lx\n", work_on_cpu(b->cpu, local_error_count, b));
+ struct threshold_block_cross_cpu tbcc = { .tb = b, };
+
+ smp_call_function_single(b->cpu, local_error_count_handler, &tbcc, 1);
+ return sprintf(buf, "%lx\n", tbcc.retval);
}
static ssize_t store_error_count(struct threshold_block *b,
{
struct thresh_restart tr = { .b = b, .reset = 1, .old_limit = 0 };
- work_on_cpu(b->cpu, threshold_restart_bank, &tr);
+ smp_call_function_single(b->cpu, threshold_restart_bank, &tr, 1);
return 1;
}
if ((bank >= NR_BANKS) || (block >= NR_BLOCKS))
return 0;
- if (rdmsr_safe(address, &low, &high))
+ if (rdmsr_safe_on_cpu(cpu, address, &low, &high))
return 0;
if (!(high & MASK_VALID_HI)) {
return err;
}
- static __cpuinit long local_allocate_threshold_blocks(void *_bank)
+ static __cpuinit long
+ local_allocate_threshold_blocks(int cpu, unsigned int bank)
{
- unsigned int *bank = _bank;
-
- return allocate_threshold_blocks(smp_processor_id(), *bank, 0,
- MSR_IA32_MC0_MISC + *bank * 4);
+ return allocate_threshold_blocks(cpu, bank, 0,
+ MSR_IA32_MC0_MISC + bank * 4);
}
/* symlinks sibling shared banks to first core. first core owns dir/files. */
#ifdef CONFIG_SMP
if (cpu_data(cpu).cpu_core_id && shared_bank[bank]) { /* symlink */
- i = cpumask_first(&per_cpu(cpu_core_map, cpu));
+ i = cpumask_first(cpu_core_mask(cpu));
/* first core not up yet */
if (cpu_data(i).cpu_core_id)
if (err)
goto out;
- cpumask_copy(b->cpus, &per_cpu(cpu_core_map, cpu));
+ cpumask_copy(b->cpus, cpu_core_mask(cpu));
per_cpu(threshold_banks, cpu)[bank] = b;
goto out;
}
#ifndef CONFIG_SMP
cpumask_setall(b->cpus);
#else
- cpumask_copy(b->cpus, &per_cpu(cpu_core_map, cpu));
+ cpumask_copy(b->cpus, cpu_core_mask(cpu));
#endif
per_cpu(threshold_banks, cpu)[bank] = b;
- err = work_on_cpu(cpu, local_allocate_threshold_blocks, &bank);
+ err = local_allocate_threshold_blocks(cpu, bank);
if (err)
goto out_free;
for_each_online_cpu (cpu) {
if (cpu == dying)
continue;
- if (set_cpus_allowed_ptr(current, &cpumask_of_cpu(cpu)))
+ if (set_cpus_allowed_ptr(current, cpumask_of(cpu)))
continue;
/* Recheck banks in case CPUs don't all have the same */
if (cmci_supported(&banks))
cmci_discover(banks, 0);
}
- static __cpuinit void intel_init_cmci(void)
+ static void intel_init_cmci(void)
{
int banks;
{
struct task_struct *me = current;
struct thread_struct *t = &me->thread;
+ unsigned long *bp = t->io_bitmap_ptr;
- if (me->thread.io_bitmap_ptr) {
+ if (bp) {
struct tss_struct *tss = &per_cpu(init_tss, get_cpu());
- kfree(t->io_bitmap_ptr);
t->io_bitmap_ptr = NULL;
clear_thread_flag(TIF_IO_BITMAP);
/*
memset(tss->io_bitmap, 0xff, t->io_bitmap_max);
t->io_bitmap_max = 0;
put_cpu();
+ kfree(bp);
}
ds_exit_thread(current);
/*
* Remove this CPU:
*/
- cpu_clear(smp_processor_id(), cpu_online_map);
+ set_cpu_online(smp_processor_id(), false);
disable_local_APIC();
for (;;) {
return 1;
}
-static cpumask_t c1e_mask = CPU_MASK_NONE;
+static cpumask_var_t c1e_mask;
static int c1e_detected;
void c1e_remove_cpu(int cpu)
{
- cpu_clear(cpu, c1e_mask);
+ if (c1e_mask != NULL)
+ cpumask_clear_cpu(cpu, c1e_mask);
}
/*
if (c1e_detected) {
int cpu = smp_processor_id();
- if (!cpu_isset(cpu, c1e_mask)) {
- cpu_set(cpu, c1e_mask);
+ if (!cpumask_test_cpu(cpu, c1e_mask)) {
+ cpumask_set_cpu(cpu, c1e_mask);
/*
* Force broadcast so ACPI can not interfere. Needs
* to run with interrupts enabled as it uses
pm_idle = default_idle;
}
+void __init init_c1e_mask(void)
+{
+ /* If we're using c1e_idle, we need to allocate c1e_mask. */
+ if (pm_idle == c1e_idle) {
+ alloc_cpumask_var(&c1e_mask, GFP_KERNEL);
+ cpumask_clear(c1e_mask);
+ }
+}
+
static int __init idle_setup(char *str)
{
if (!str)
return NULL;
}
+static DEFINE_PER_CPU(cpumask_var_t, uv_flush_tlb_mask);
+
/**
* uv_flush_tlb_others - globally purge translation cache of a virtual
* address or all TLB's
struct mm_struct *mm,
unsigned long va, unsigned int cpu)
{
- static DEFINE_PER_CPU(cpumask_t, flush_tlb_mask);
- struct cpumask *flush_mask = &__get_cpu_var(flush_tlb_mask);
+ struct cpumask *flush_mask = __get_cpu_var(uv_flush_tlb_mask);
int i;
int bit;
int blade;
int node;
int nblades;
int last_blade;
- int cur_cpu = 0;
+ int cur_cpu;
if (!is_uv_system())
return 0;
+ for_each_possible_cpu(cur_cpu)
+ alloc_cpumask_var_node(&per_cpu(uv_flush_tlb_mask, cur_cpu),
+ GFP_KERNEL, cpu_to_node(cur_cpu));
+
uv_bau_retry_limit = 1;
uv_nshift = uv_hub_info->n_val;
uv_mmask = (1UL << uv_hub_info->n_val) - 1;
nblades = 0;
last_blade = -1;
+ cur_cpu = 0;
for_each_online_node(node) {
blade = uv_node_to_blade_id(node);
if (blade == last_blade)
*/
static DEFINE_SPINLOCK(task_group_lock);
+ #ifdef CONFIG_SMP
+ static int root_task_group_empty(void)
+ {
+ return list_empty(&root_task_group.children);
+ }
+ #endif
+
#ifdef CONFIG_FAIR_GROUP_SCHED
#ifdef CONFIG_USER_SCHED
# define INIT_TASK_GROUP_LOAD (2*NICE_0_LOAD)
#else
+ #ifdef CONFIG_SMP
+ static int root_task_group_empty(void)
+ {
+ return 1;
+ }
+ #endif
+
static inline void set_task_rq(struct task_struct *p, unsigned int cpu) { }
static inline struct task_group *task_group(struct task_struct *p)
{
struct rt_prio_array active;
unsigned long rt_nr_running;
#if defined CONFIG_SMP || defined CONFIG_RT_GROUP_SCHED
- int highest_prio; /* highest queued rt task prio */
+ struct {
+ int curr; /* highest queued rt task prio */
+ #ifdef CONFIG_SMP
+ int next; /* next highest */
+ #endif
+ } highest_prio;
#endif
#ifdef CONFIG_SMP
unsigned long rt_nr_migratory;
int overloaded;
+ struct plist_head pushable_tasks;
#endif
int rt_throttled;
u64 rt_time;
unsigned long nr_running;
#define CPU_LOAD_IDX_MAX 5
unsigned long cpu_load[CPU_LOAD_IDX_MAX];
- unsigned char idle_at_tick;
#ifdef CONFIG_NO_HZ
unsigned long last_tick_seen;
unsigned char in_nohz_recently;
struct root_domain *rd;
struct sched_domain *sd;
+ unsigned char idle_at_tick;
/* For active balancing */
int active_balance;
int push_cpu;
/* could above be rq->cfs_rq.exec_clock + rq->rt_rq.rt_runtime ? */
/* sys_sched_yield() stats */
- unsigned int yld_exp_empty;
- unsigned int yld_act_empty;
- unsigned int yld_both_empty;
unsigned int yld_count;
/* schedule() stats */
assert_spin_locked(&task_rq(p)->lock);
- if (unlikely(test_tsk_thread_flag(p, TIF_NEED_RESCHED)))
+ if (test_tsk_need_resched(p))
return;
- set_tsk_thread_flag(p, TIF_NEED_RESCHED);
+ set_tsk_need_resched(p);
cpu = task_cpu(p);
if (cpu == smp_processor_id())
* lockless. The worst case is that the other CPU runs the
* idle task through an additional NOOP schedule()
*/
- set_tsk_thread_flag(rq->idle, TIF_NEED_RESCHED);
+ set_tsk_need_resched(rq->idle);
/* NEED_RESCHED must be visible before we test polling */
smp_mb();
#endif
+ #ifdef CONFIG_PREEMPT
+
/*
- * double_lock_balance - lock the busiest runqueue, this_rq is locked already.
+ * fair double_lock_balance: Safely acquires both rq->locks in a fair
+ * way at the expense of forcing extra atomic operations in all
+ * invocations. This assures that the double_lock is acquired using the
+ * same underlying policy as the spinlock_t on this architecture, which
+ * reduces latency compared to the unfair variant below. However, it
+ * also adds more overhead and therefore may reduce throughput.
*/
- static int double_lock_balance(struct rq *this_rq, struct rq *busiest)
+ static inline int _double_lock_balance(struct rq *this_rq, struct rq *busiest)
+ __releases(this_rq->lock)
+ __acquires(busiest->lock)
+ __acquires(this_rq->lock)
+ {
+ spin_unlock(&this_rq->lock);
+ double_rq_lock(this_rq, busiest);
+
+ return 1;
+ }
+
+ #else
+ /*
+ * Unfair double_lock_balance: Optimizes throughput at the expense of
+ * latency by eliminating extra atomic operations when the locks are
+ * already in proper order on entry. This favors lower cpu-ids and will
+ * grant the double lock to lower cpus over higher ids under contention,
+ * regardless of entry order into the function.
+ */
+ static int _double_lock_balance(struct rq *this_rq, struct rq *busiest)
__releases(this_rq->lock)
__acquires(busiest->lock)
__acquires(this_rq->lock)
{
int ret = 0;
- if (unlikely(!irqs_disabled())) {
- /* printk() doesn't work good under rq->lock */
- spin_unlock(&this_rq->lock);
- BUG_ON(1);
- }
if (unlikely(!spin_trylock(&busiest->lock))) {
if (busiest < this_rq) {
spin_unlock(&this_rq->lock);
return ret;
}
+ #endif /* CONFIG_PREEMPT */
+
+ /*
+ * double_lock_balance - lock the busiest runqueue, this_rq is locked already.
+ */
+ static int double_lock_balance(struct rq *this_rq, struct rq *busiest)
+ {
+ if (unlikely(!irqs_disabled())) {
+ /* printk() doesn't work good under rq->lock */
+ spin_unlock(&this_rq->lock);
+ BUG_ON(1);
+ }
+
+ return _double_lock_balance(this_rq, busiest);
+ }
+
static inline void double_unlock_balance(struct rq *this_rq, struct rq *busiest)
__releases(busiest->lock)
{
static void enqueue_task(struct rq *rq, struct task_struct *p, int wakeup)
{
+ if (wakeup)
+ p->se.start_runtime = p->se.sum_exec_runtime;
+
sched_info_queued(p);
p->sched_class->enqueue_task(rq, p, wakeup);
p->se.on_rq = 1;
static void dequeue_task(struct rq *rq, struct task_struct *p, int sleep)
{
- if (sleep && p->se.last_wakeup) {
- update_avg(&p->se.avg_overlap,
- p->se.sum_exec_runtime - p->se.last_wakeup);
- p->se.last_wakeup = 0;
+ if (sleep) {
+ if (p->se.last_wakeup) {
+ update_avg(&p->se.avg_overlap,
+ p->se.sum_exec_runtime - p->se.last_wakeup);
+ p->se.last_wakeup = 0;
+ } else {
+ update_avg(&p->se.avg_wakeup,
+ sysctl_sched_wakeup_granularity);
+ }
}
sched_info_dequeued(p);
* it must be off the runqueue _entirely_, and not
* preempted!
*
- * So if it wa still runnable (but just not actively
+ * So if it was still runnable (but just not actively
* running right now), it's preempted, and we should
* yield - it could be a while.
*/
sync = 0;
#ifdef CONFIG_SMP
- if (sched_feat(LB_WAKEUP_UPDATE)) {
+ if (sched_feat(LB_WAKEUP_UPDATE) && !root_task_group_empty()) {
struct sched_domain *sd;
this_cpu = raw_smp_processor_id();
activate_task(rq, p, 1);
success = 1;
+ /*
+ * Only attribute actual wakeups done by this task.
+ */
+ if (!in_interrupt()) {
+ struct sched_entity *se = ¤t->se;
+ u64 sample = se->sum_exec_runtime;
+
+ if (se->last_wakeup)
+ sample -= se->last_wakeup;
+ else
+ sample -= se->start_runtime;
+ update_avg(&se->avg_wakeup, sample);
+
+ se->last_wakeup = se->sum_exec_runtime;
+ }
+
out_running:
trace_sched_wakeup(rq, p, success);
check_preempt_curr(rq, p, sync);
p->sched_class->task_wake_up(rq, p);
#endif
out:
- current->se.last_wakeup = current->se.sum_exec_runtime;
-
task_rq_unlock(rq, &flags);
return success;
p->se.prev_sum_exec_runtime = 0;
p->se.last_wakeup = 0;
p->se.avg_overlap = 0;
+ p->se.start_runtime = 0;
+ p->se.avg_wakeup = sysctl_sched_wakeup_granularity;
#ifdef CONFIG_SCHEDSTATS
p->se.wait_start = 0;
/* Want to start with kernel preemption disabled. */
task_thread_info(p)->preempt_count = 1;
#endif
+ plist_node_init(&p->pushable_tasks, MAX_PRIO);
+
put_cpu();
}
#ifdef CONFIG_PREEMPT_NOTIFIERS
/**
- * preempt_notifier_register - tell me when current is being being preempted & rescheduled
+ * preempt_notifier_register - tell me when current is being preempted & rescheduled
* @notifier: notifier struct to register
*/
void preempt_notifier_register(struct preempt_notifier *notifier)
{
struct mm_struct *mm = rq->prev_mm;
long prev_state;
+ #ifdef CONFIG_SMP
+ int post_schedule = 0;
+
+ if (current->sched_class->needs_post_schedule)
+ post_schedule = current->sched_class->needs_post_schedule(rq);
+ #endif
rq->prev_mm = NULL;
finish_arch_switch(prev);
finish_lock_switch(rq, prev);
#ifdef CONFIG_SMP
- if (current->sched_class->post_schedule)
+ if (post_schedule)
current->sched_class->post_schedule(rq);
#endif
struct sched_domain *sd, enum cpu_idle_type idle,
int *all_pinned)
{
+ int tsk_cache_hot = 0;
/*
* We do not migrate tasks that are:
* 1) running (obviously), or
* 2) too many balance attempts have failed.
*/
- if (!task_hot(p, rq->clock, sd) ||
- sd->nr_balance_failed > sd->cache_nice_tries) {
+ tsk_cache_hot = task_hot(p, rq->clock, sd);
+ if (!tsk_cache_hot ||
+ sd->nr_balance_failed > sd->cache_nice_tries) {
#ifdef CONFIG_SCHEDSTATS
- if (task_hot(p, rq->clock, sd)) {
+ if (tsk_cache_hot) {
schedstat_inc(sd, lb_hot_gained[idle]);
schedstat_inc(p, se.nr_forced_migrations);
}
return 1;
}
- if (task_hot(p, rq->clock, sd)) {
+ if (tsk_cache_hot) {
schedstat_inc(p, se.nr_failed_migrations_hot);
return 0;
}
pulled++;
rem_load_move -= p->se.load.weight;
+ #ifdef CONFIG_PREEMPT
+ /*
+ * NEWIDLE balancing is a source of latency, so preemptible kernels
+ * will stop after the first task is pulled to minimize the critical
+ * section.
+ */
+ if (idle == CPU_NEWLY_IDLE)
+ goto out;
+ #endif
+
/*
* We only want to steal up to the prescribed amount of weighted load.
*/
sd, idle, all_pinned, &this_best_prio);
class = class->next;
+ #ifdef CONFIG_PREEMPT
+ /*
+ * NEWIDLE balancing is a source of latency, so preemptible
+ * kernels will stop after the first task is pulled to minimize
+ * the critical section.
+ */
if (idle == CPU_NEWLY_IDLE && this_rq->nr_running)
break;
-
+ #endif
} while (class && max_load_move > total_load_moved);
return total_load_moved > 0;
return 0;
}
-
+ /********** Helpers for find_busiest_group ************************/
/*
- * find_busiest_group finds and returns the busiest CPU group within the
- * domain. It calculates and returns the amount of weighted load which
- * should be moved to restore balance via the imbalance parameter.
+ * sd_lb_stats - Structure to store the statistics of a sched_domain
+ * during load balancing.
*/
- static struct sched_group *
- find_busiest_group(struct sched_domain *sd, int this_cpu,
- unsigned long *imbalance, enum cpu_idle_type idle,
- int *sd_idle, const struct cpumask *cpus, int *balance)
- {
- struct sched_group *busiest = NULL, *this = NULL, *group = sd->groups;
- unsigned long max_load, avg_load, total_load, this_load, total_pwr;
- unsigned long max_pull;
- unsigned long busiest_load_per_task, busiest_nr_running;
- unsigned long this_load_per_task, this_nr_running;
- int load_idx, group_imb = 0;
+ struct sd_lb_stats {
+ struct sched_group *busiest; /* Busiest group in this sd */
+ struct sched_group *this; /* Local group in this sd */
+ unsigned long total_load; /* Total load of all groups in sd */
+ unsigned long total_pwr; /* Total power of all groups in sd */
+ unsigned long avg_load; /* Average load across all groups in sd */
+
+ /** Statistics of this group */
+ unsigned long this_load;
+ unsigned long this_load_per_task;
+ unsigned long this_nr_running;
+
+ /* Statistics of the busiest group */
+ unsigned long max_load;
+ unsigned long busiest_load_per_task;
+ unsigned long busiest_nr_running;
+
+ int group_imb; /* Is there imbalance in this sd */
#if defined(CONFIG_SCHED_MC) || defined(CONFIG_SCHED_SMT)
- int power_savings_balance = 1;
- unsigned long leader_nr_running = 0, min_load_per_task = 0;
- unsigned long min_nr_running = ULONG_MAX;
- struct sched_group *group_min = NULL, *group_leader = NULL;
+ int power_savings_balance; /* Is powersave balance needed for this sd */
+ struct sched_group *group_min; /* Least loaded group in sd */
+ struct sched_group *group_leader; /* Group which relieves group_min */
+ unsigned long min_load_per_task; /* load_per_task in group_min */
+ unsigned long leader_nr_running; /* Nr running of group_leader */
+ unsigned long min_nr_running; /* Nr running of group_min */
#endif
+ };
- max_load = this_load = total_load = total_pwr = 0;
- busiest_load_per_task = busiest_nr_running = 0;
- this_load_per_task = this_nr_running = 0;
+ /*
+ * sg_lb_stats - stats of a sched_group required for load_balancing
+ */
+ struct sg_lb_stats {
+ unsigned long avg_load; /*Avg load across the CPUs of the group */
+ unsigned long group_load; /* Total load over the CPUs of the group */
+ unsigned long sum_nr_running; /* Nr tasks running in the group */
+ unsigned long sum_weighted_load; /* Weighted load of group's tasks */
+ unsigned long group_capacity;
+ int group_imb; /* Is there an imbalance in the group ? */
+ };
- if (idle == CPU_NOT_IDLE)
+ /**
+ * group_first_cpu - Returns the first cpu in the cpumask of a sched_group.
+ * @group: The group whose first cpu is to be returned.
+ */
+ static inline unsigned int group_first_cpu(struct sched_group *group)
+ {
+ return cpumask_first(sched_group_cpus(group));
+ }
+
+ /**
+ * get_sd_load_idx - Obtain the load index for a given sched domain.
+ * @sd: The sched_domain whose load_idx is to be obtained.
+ * @idle: The Idle status of the CPU for whose sd load_icx is obtained.
+ */
+ static inline int get_sd_load_idx(struct sched_domain *sd,
+ enum cpu_idle_type idle)
+ {
+ int load_idx;
+
+ switch (idle) {
+ case CPU_NOT_IDLE:
load_idx = sd->busy_idx;
- else if (idle == CPU_NEWLY_IDLE)
+ break;
+
+ case CPU_NEWLY_IDLE:
load_idx = sd->newidle_idx;
- else
+ break;
+ default:
load_idx = sd->idle_idx;
+ break;
+ }
- do {
- unsigned long load, group_capacity, max_cpu_load, min_cpu_load;
- int local_group;
- int i;
- int __group_imb = 0;
- unsigned int balance_cpu = -1, first_idle_cpu = 0;
- unsigned long sum_nr_running, sum_weighted_load;
- unsigned long sum_avg_load_per_task;
- unsigned long avg_load_per_task;
+ return load_idx;
+ }
- local_group = cpumask_test_cpu(this_cpu,
- sched_group_cpus(group));
- if (local_group)
- balance_cpu = cpumask_first(sched_group_cpus(group));
+ #if defined(CONFIG_SCHED_MC) || defined(CONFIG_SCHED_SMT)
+ /**
+ * init_sd_power_savings_stats - Initialize power savings statistics for
+ * the given sched_domain, during load balancing.
+ *
+ * @sd: Sched domain whose power-savings statistics are to be initialized.
+ * @sds: Variable containing the statistics for sd.
+ * @idle: Idle status of the CPU at which we're performing load-balancing.
+ */
+ static inline void init_sd_power_savings_stats(struct sched_domain *sd,
+ struct sd_lb_stats *sds, enum cpu_idle_type idle)
+ {
+ /*
+ * Busy processors will not participate in power savings
+ * balance.
+ */
+ if (idle == CPU_NOT_IDLE || !(sd->flags & SD_POWERSAVINGS_BALANCE))
+ sds->power_savings_balance = 0;
+ else {
+ sds->power_savings_balance = 1;
+ sds->min_nr_running = ULONG_MAX;
+ sds->leader_nr_running = 0;
+ }
+ }
- /* Tally up the load of all CPUs in the group */
- sum_weighted_load = sum_nr_running = avg_load = 0;
- sum_avg_load_per_task = avg_load_per_task = 0;
+ /**
+ * update_sd_power_savings_stats - Update the power saving stats for a
+ * sched_domain while performing load balancing.
+ *
+ * @group: sched_group belonging to the sched_domain under consideration.
+ * @sds: Variable containing the statistics of the sched_domain
+ * @local_group: Does group contain the CPU for which we're performing
+ * load balancing ?
+ * @sgs: Variable containing the statistics of the group.
+ */
+ static inline void update_sd_power_savings_stats(struct sched_group *group,
+ struct sd_lb_stats *sds, int local_group, struct sg_lb_stats *sgs)
+ {
- max_cpu_load = 0;
- min_cpu_load = ~0UL;
+ if (!sds->power_savings_balance)
+ return;
- for_each_cpu_and(i, sched_group_cpus(group), cpus) {
- struct rq *rq = cpu_rq(i);
+ /*
+ * If the local group is idle or completely loaded
+ * no need to do power savings balance at this domain
+ */
+ if (local_group && (sds->this_nr_running >= sgs->group_capacity ||
+ !sds->this_nr_running))
+ sds->power_savings_balance = 0;
- if (*sd_idle && rq->nr_running)
- *sd_idle = 0;
+ /*
+ * If a group is already running at full capacity or idle,
+ * don't include that group in power savings calculations
+ */
+ if (!sds->power_savings_balance ||
+ sgs->sum_nr_running >= sgs->group_capacity ||
+ !sgs->sum_nr_running)
+ return;
- /* Bias balancing toward cpus of our domain */
- if (local_group) {
- if (idle_cpu(i) && !first_idle_cpu) {
- first_idle_cpu = 1;
- balance_cpu = i;
- }
+ /*
+ * Calculate the group which has the least non-idle load.
+ * This is the group from where we need to pick up the load
+ * for saving power
+ */
+ if ((sgs->sum_nr_running < sds->min_nr_running) ||
+ (sgs->sum_nr_running == sds->min_nr_running &&
+ group_first_cpu(group) > group_first_cpu(sds->group_min))) {
+ sds->group_min = group;
+ sds->min_nr_running = sgs->sum_nr_running;
+ sds->min_load_per_task = sgs->sum_weighted_load /
+ sgs->sum_nr_running;
+ }
- load = target_load(i, load_idx);
- } else {
- load = source_load(i, load_idx);
- if (load > max_cpu_load)
- max_cpu_load = load;
- if (min_cpu_load > load)
- min_cpu_load = load;
- }
+ /*
+ * Calculate the group which is almost near its
+ * capacity but still has some space to pick up some load
+ * from other group and save more power
+ */
+ if (sgs->sum_nr_running > sgs->group_capacity - 1)
+ return;
- avg_load += load;
- sum_nr_running += rq->nr_running;
- sum_weighted_load += weighted_cpuload(i);
+ if (sgs->sum_nr_running > sds->leader_nr_running ||
+ (sgs->sum_nr_running == sds->leader_nr_running &&
+ group_first_cpu(group) < group_first_cpu(sds->group_leader))) {
+ sds->group_leader = group;
+ sds->leader_nr_running = sgs->sum_nr_running;
+ }
+ }
- sum_avg_load_per_task += cpu_avg_load_per_task(i);
- }
+ /**
+ * check_power_save_busiest_group - see if there is potential for some power-savings balance
+ * @sds: Variable containing the statistics of the sched_domain
+ * under consideration.
+ * @this_cpu: Cpu at which we're currently performing load-balancing.
+ * @imbalance: Variable to store the imbalance.
+ *
+ * Description:
+ * Check if we have potential to perform some power-savings balance.
+ * If yes, set the busiest group to be the least loaded group in the
+ * sched_domain, so that it's CPUs can be put to idle.
+ *
+ * Returns 1 if there is potential to perform power-savings balance.
+ * Else returns 0.
+ */
+ static inline int check_power_save_busiest_group(struct sd_lb_stats *sds,
+ int this_cpu, unsigned long *imbalance)
+ {
+ if (!sds->power_savings_balance)
+ return 0;
- /*
- * First idle cpu or the first cpu(busiest) in this sched group
- * is eligible for doing load balancing at this and above
- * domains. In the newly idle case, we will allow all the cpu's
- * to do the newly idle load balance.
- */
- if (idle != CPU_NEWLY_IDLE && local_group &&
- balance_cpu != this_cpu && balance) {
- *balance = 0;
- goto ret;
- }
+ if (sds->this != sds->group_leader ||
+ sds->group_leader == sds->group_min)
+ return 0;
- total_load += avg_load;
- total_pwr += group->__cpu_power;
+ *imbalance = sds->min_load_per_task;
+ sds->busiest = sds->group_min;
- /* Adjust by relative CPU power of the group */
- avg_load = sg_div_cpu_power(group,
- avg_load * SCHED_LOAD_SCALE);
+ if (sched_mc_power_savings >= POWERSAVINGS_BALANCE_WAKEUP) {
+ cpu_rq(this_cpu)->rd->sched_mc_preferred_wakeup_cpu =
+ group_first_cpu(sds->group_leader);
+ }
+ return 1;
- /*
- * Consider the group unbalanced when the imbalance is larger
- * than the average weight of two tasks.
- *
- * APZ: with cgroup the avg task weight can vary wildly and
- * might not be a suitable number - should we keep a
- * normalized nr_running number somewhere that negates
- * the hierarchy?
- */
- avg_load_per_task = sg_div_cpu_power(group,
- sum_avg_load_per_task * SCHED_LOAD_SCALE);
+ }
+ #else /* CONFIG_SCHED_MC || CONFIG_SCHED_SMT */
+ static inline void init_sd_power_savings_stats(struct sched_domain *sd,
+ struct sd_lb_stats *sds, enum cpu_idle_type idle)
+ {
+ return;
+ }
- if ((max_cpu_load - min_cpu_load) > 2*avg_load_per_task)
- __group_imb = 1;
+ static inline void update_sd_power_savings_stats(struct sched_group *group,
+ struct sd_lb_stats *sds, int local_group, struct sg_lb_stats *sgs)
+ {
+ return;
+ }
+
+ static inline int check_power_save_busiest_group(struct sd_lb_stats *sds,
+ int this_cpu, unsigned long *imbalance)
+ {
+ return 0;
+ }
+ #endif /* CONFIG_SCHED_MC || CONFIG_SCHED_SMT */
- group_capacity = group->__cpu_power / SCHED_LOAD_SCALE;
+ /**
+ * update_sg_lb_stats - Update sched_group's statistics for load balancing.
+ * @group: sched_group whose statistics are to be updated.
+ * @this_cpu: Cpu for which load balance is currently performed.
+ * @idle: Idle status of this_cpu
+ * @load_idx: Load index of sched_domain of this_cpu for load calc.
+ * @sd_idle: Idle status of the sched_domain containing group.
+ * @local_group: Does group contain this_cpu.
+ * @cpus: Set of cpus considered for load balancing.
+ * @balance: Should we balance.
+ * @sgs: variable to hold the statistics for this group.
+ */
+ static inline void update_sg_lb_stats(struct sched_group *group, int this_cpu,
+ enum cpu_idle_type idle, int load_idx, int *sd_idle,
+ int local_group, const struct cpumask *cpus,
+ int *balance, struct sg_lb_stats *sgs)
+ {
+ unsigned long load, max_cpu_load, min_cpu_load;
+ int i;
+ unsigned int balance_cpu = -1, first_idle_cpu = 0;
+ unsigned long sum_avg_load_per_task;
+ unsigned long avg_load_per_task;
+
+ if (local_group)
+ balance_cpu = group_first_cpu(group);
+
+ /* Tally up the load of all CPUs in the group */
+ sum_avg_load_per_task = avg_load_per_task = 0;
+ max_cpu_load = 0;
+ min_cpu_load = ~0UL;
+
+ for_each_cpu_and(i, sched_group_cpus(group), cpus) {
+ struct rq *rq = cpu_rq(i);
+
+ if (*sd_idle && rq->nr_running)
+ *sd_idle = 0;
+
+ /* Bias balancing toward cpus of our domain */
if (local_group) {
- this_load = avg_load;
- this = group;
- this_nr_running = sum_nr_running;
- this_load_per_task = sum_weighted_load;
- } else if (avg_load > max_load &&
- (sum_nr_running > group_capacity || __group_imb)) {
- max_load = avg_load;
- busiest = group;
- busiest_nr_running = sum_nr_running;
- busiest_load_per_task = sum_weighted_load;
- group_imb = __group_imb;
+ if (idle_cpu(i) && !first_idle_cpu) {
+ first_idle_cpu = 1;
+ balance_cpu = i;
+ }
+
+ load = target_load(i, load_idx);
+ } else {
+ load = source_load(i, load_idx);
+ if (load > max_cpu_load)
+ max_cpu_load = load;
+ if (min_cpu_load > load)
+ min_cpu_load = load;
}
- #if defined(CONFIG_SCHED_MC) || defined(CONFIG_SCHED_SMT)
- /*
- * Busy processors will not participate in power savings
- * balance.
- */
- if (idle == CPU_NOT_IDLE ||
- !(sd->flags & SD_POWERSAVINGS_BALANCE))
- goto group_next;
+ sgs->group_load += load;
+ sgs->sum_nr_running += rq->nr_running;
+ sgs->sum_weighted_load += weighted_cpuload(i);
- /*
- * If the local group is idle or completely loaded
- * no need to do power savings balance at this domain
- */
- if (local_group && (this_nr_running >= group_capacity ||
- !this_nr_running))
- power_savings_balance = 0;
+ sum_avg_load_per_task += cpu_avg_load_per_task(i);
+ }
- /*
- * If a group is already running at full capacity or idle,
- * don't include that group in power savings calculations
- */
- if (!power_savings_balance || sum_nr_running >= group_capacity
- || !sum_nr_running)
- goto group_next;
+ /*
+ * First idle cpu or the first cpu(busiest) in this sched group
+ * is eligible for doing load balancing at this and above
+ * domains. In the newly idle case, we will allow all the cpu's
+ * to do the newly idle load balance.
+ */
+ if (idle != CPU_NEWLY_IDLE && local_group &&
+ balance_cpu != this_cpu && balance) {
+ *balance = 0;
+ return;
+ }
- /*
- * Calculate the group which has the least non-idle load.
- * This is the group from where we need to pick up the load
- * for saving power
- */
- if ((sum_nr_running < min_nr_running) ||
- (sum_nr_running == min_nr_running &&
- cpumask_first(sched_group_cpus(group)) >
- cpumask_first(sched_group_cpus(group_min)))) {
- group_min = group;
- min_nr_running = sum_nr_running;
- min_load_per_task = sum_weighted_load /
- sum_nr_running;
- }
+ /* Adjust by relative CPU power of the group */
+ sgs->avg_load = sg_div_cpu_power(group,
+ sgs->group_load * SCHED_LOAD_SCALE);
- /*
- * Calculate the group which is almost near its
- * capacity but still has some space to pick up some load
- * from other group and save more power
- */
- if (sum_nr_running <= group_capacity - 1) {
- if (sum_nr_running > leader_nr_running ||
- (sum_nr_running == leader_nr_running &&
- cpumask_first(sched_group_cpus(group)) <
- cpumask_first(sched_group_cpus(group_leader)))) {
- group_leader = group;
- leader_nr_running = sum_nr_running;
- }
+
+ /*
+ * Consider the group unbalanced when the imbalance is larger
+ * than the average weight of two tasks.
+ *
+ * APZ: with cgroup the avg task weight can vary wildly and
+ * might not be a suitable number - should we keep a
+ * normalized nr_running number somewhere that negates
+ * the hierarchy?
+ */
+ avg_load_per_task = sg_div_cpu_power(group,
+ sum_avg_load_per_task * SCHED_LOAD_SCALE);
+
+ if ((max_cpu_load - min_cpu_load) > 2*avg_load_per_task)
+ sgs->group_imb = 1;
+
+ sgs->group_capacity = group->__cpu_power / SCHED_LOAD_SCALE;
+
+ }
+
+ /**
+ * update_sd_lb_stats - Update sched_group's statistics for load balancing.
+ * @sd: sched_domain whose statistics are to be updated.
+ * @this_cpu: Cpu for which load balance is currently performed.
+ * @idle: Idle status of this_cpu
+ * @sd_idle: Idle status of the sched_domain containing group.
+ * @cpus: Set of cpus considered for load balancing.
+ * @balance: Should we balance.
+ * @sds: variable to hold the statistics for this sched_domain.
+ */
+ static inline void update_sd_lb_stats(struct sched_domain *sd, int this_cpu,
+ enum cpu_idle_type idle, int *sd_idle,
+ const struct cpumask *cpus, int *balance,
+ struct sd_lb_stats *sds)
+ {
+ struct sched_group *group = sd->groups;
+ struct sg_lb_stats sgs;
+ int load_idx;
+
+ init_sd_power_savings_stats(sd, sds, idle);
+ load_idx = get_sd_load_idx(sd, idle);
+
+ do {
+ int local_group;
+
+ local_group = cpumask_test_cpu(this_cpu,
+ sched_group_cpus(group));
+ memset(&sgs, 0, sizeof(sgs));
+ update_sg_lb_stats(group, this_cpu, idle, load_idx, sd_idle,
+ local_group, cpus, balance, &sgs);
+
+ if (local_group && balance && !(*balance))
+ return;
+
+ sds->total_load += sgs.group_load;
+ sds->total_pwr += group->__cpu_power;
+
+ if (local_group) {
+ sds->this_load = sgs.avg_load;
+ sds->this = group;
+ sds->this_nr_running = sgs.sum_nr_running;
+ sds->this_load_per_task = sgs.sum_weighted_load;
+ } else if (sgs.avg_load > sds->max_load &&
+ (sgs.sum_nr_running > sgs.group_capacity ||
+ sgs.group_imb)) {
+ sds->max_load = sgs.avg_load;
+ sds->busiest = group;
+ sds->busiest_nr_running = sgs.sum_nr_running;
+ sds->busiest_load_per_task = sgs.sum_weighted_load;
+ sds->group_imb = sgs.group_imb;
}
- group_next:
- #endif
+
+ update_sd_power_savings_stats(group, sds, local_group, &sgs);
group = group->next;
} while (group != sd->groups);
- if (!busiest || this_load >= max_load || busiest_nr_running == 0)
- goto out_balanced;
-
- avg_load = (SCHED_LOAD_SCALE * total_load) / total_pwr;
+ }
- if (this_load >= avg_load ||
- 100*max_load <= sd->imbalance_pct*this_load)
- goto out_balanced;
+ /**
+ * fix_small_imbalance - Calculate the minor imbalance that exists
+ * amongst the groups of a sched_domain, during
+ * load balancing.
+ * @sds: Statistics of the sched_domain whose imbalance is to be calculated.
+ * @this_cpu: The cpu at whose sched_domain we're performing load-balance.
+ * @imbalance: Variable to store the imbalance.
+ */
+ static inline void fix_small_imbalance(struct sd_lb_stats *sds,
+ int this_cpu, unsigned long *imbalance)
+ {
+ unsigned long tmp, pwr_now = 0, pwr_move = 0;
+ unsigned int imbn = 2;
+
+ if (sds->this_nr_running) {
+ sds->this_load_per_task /= sds->this_nr_running;
+ if (sds->busiest_load_per_task >
+ sds->this_load_per_task)
+ imbn = 1;
+ } else
+ sds->this_load_per_task =
+ cpu_avg_load_per_task(this_cpu);
- busiest_load_per_task /= busiest_nr_running;
- if (group_imb)
- busiest_load_per_task = min(busiest_load_per_task, avg_load);
+ if (sds->max_load - sds->this_load + sds->busiest_load_per_task >=
+ sds->busiest_load_per_task * imbn) {
+ *imbalance = sds->busiest_load_per_task;
+ return;
+ }
/*
- * We're trying to get all the cpus to the average_load, so we don't
- * want to push ourselves above the average load, nor do we wish to
- * reduce the max loaded cpu below the average load, as either of these
- * actions would just result in more rebalancing later, and ping-pong
- * tasks around. Thus we look for the minimum possible imbalance.
- * Negative imbalances (*we* are more loaded than anyone else) will
- * be counted as no imbalance for these purposes -- we can't fix that
- * by pulling tasks to us. Be careful of negative numbers as they'll
- * appear as very large values with unsigned longs.
+ * OK, we don't have enough imbalance to justify moving tasks,
+ * however we may be able to increase total CPU power used by
+ * moving them.
*/
- if (max_load <= busiest_load_per_task)
- goto out_balanced;
+ pwr_now += sds->busiest->__cpu_power *
+ min(sds->busiest_load_per_task, sds->max_load);
+ pwr_now += sds->this->__cpu_power *
+ min(sds->this_load_per_task, sds->this_load);
+ pwr_now /= SCHED_LOAD_SCALE;
+
+ /* Amount of load we'd subtract */
+ tmp = sg_div_cpu_power(sds->busiest,
+ sds->busiest_load_per_task * SCHED_LOAD_SCALE);
+ if (sds->max_load > tmp)
+ pwr_move += sds->busiest->__cpu_power *
+ min(sds->busiest_load_per_task, sds->max_load - tmp);
+
+ /* Amount of load we'd add */
+ if (sds->max_load * sds->busiest->__cpu_power <
+ sds->busiest_load_per_task * SCHED_LOAD_SCALE)
+ tmp = sg_div_cpu_power(sds->this,
+ sds->max_load * sds->busiest->__cpu_power);
+ else
+ tmp = sg_div_cpu_power(sds->this,
+ sds->busiest_load_per_task * SCHED_LOAD_SCALE);
+ pwr_move += sds->this->__cpu_power *
+ min(sds->this_load_per_task, sds->this_load + tmp);
+ pwr_move /= SCHED_LOAD_SCALE;
+
+ /* Move if we gain throughput */
+ if (pwr_move > pwr_now)
+ *imbalance = sds->busiest_load_per_task;
+ }
+
+ /**
+ * calculate_imbalance - Calculate the amount of imbalance present within the
+ * groups of a given sched_domain during load balance.
+ * @sds: statistics of the sched_domain whose imbalance is to be calculated.
+ * @this_cpu: Cpu for which currently load balance is being performed.
+ * @imbalance: The variable to store the imbalance.
+ */
+ static inline void calculate_imbalance(struct sd_lb_stats *sds, int this_cpu,
+ unsigned long *imbalance)
+ {
+ unsigned long max_pull;
/*
* In the presence of smp nice balancing, certain scenarios can have
* max load less than avg load(as we skip the groups at or below
* its cpu_power, while calculating max_load..)
*/
- if (max_load < avg_load) {
+ if (sds->max_load < sds->avg_load) {
*imbalance = 0;
- goto small_imbalance;
+ return fix_small_imbalance(sds, this_cpu, imbalance);
}
/* Don't want to pull so many tasks that a group would go idle */
- max_pull = min(max_load - avg_load, max_load - busiest_load_per_task);
+ max_pull = min(sds->max_load - sds->avg_load,
+ sds->max_load - sds->busiest_load_per_task);
/* How much load to actually move to equalise the imbalance */
- *imbalance = min(max_pull * busiest->__cpu_power,
- (avg_load - this_load) * this->__cpu_power)
+ *imbalance = min(max_pull * sds->busiest->__cpu_power,
+ (sds->avg_load - sds->this_load) * sds->this->__cpu_power)
/ SCHED_LOAD_SCALE;
/*
* a think about bumping its value to force at least one task to be
* moved
*/
- if (*imbalance < busiest_load_per_task) {
- unsigned long tmp, pwr_now, pwr_move;
- unsigned int imbn;
-
- small_imbalance:
- pwr_move = pwr_now = 0;
- imbn = 2;
- if (this_nr_running) {
- this_load_per_task /= this_nr_running;
- if (busiest_load_per_task > this_load_per_task)
- imbn = 1;
- } else
- this_load_per_task = cpu_avg_load_per_task(this_cpu);
+ if (*imbalance < sds->busiest_load_per_task)
+ return fix_small_imbalance(sds, this_cpu, imbalance);
- if (max_load - this_load + busiest_load_per_task >=
- busiest_load_per_task * imbn) {
- *imbalance = busiest_load_per_task;
- return busiest;
- }
+ }
+ /******* find_busiest_group() helpers end here *********************/
- /*
- * OK, we don't have enough imbalance to justify moving tasks,
- * however we may be able to increase total CPU power used by
- * moving them.
- */
+ /**
+ * find_busiest_group - Returns the busiest group within the sched_domain
+ * if there is an imbalance. If there isn't an imbalance, and
+ * the user has opted for power-savings, it returns a group whose
+ * CPUs can be put to idle by rebalancing those tasks elsewhere, if
+ * such a group exists.
+ *
+ * Also calculates the amount of weighted load which should be moved
+ * to restore balance.
+ *
+ * @sd: The sched_domain whose busiest group is to be returned.
+ * @this_cpu: The cpu for which load balancing is currently being performed.
+ * @imbalance: Variable which stores amount of weighted load which should
+ * be moved to restore balance/put a group to idle.
+ * @idle: The idle status of this_cpu.
+ * @sd_idle: The idleness of sd
+ * @cpus: The set of CPUs under consideration for load-balancing.
+ * @balance: Pointer to a variable indicating if this_cpu
+ * is the appropriate cpu to perform load balancing at this_level.
+ *
+ * Returns: - the busiest group if imbalance exists.
+ * - If no imbalance and user has opted for power-savings balance,
+ * return the least loaded group whose CPUs can be
+ * put to idle by rebalancing its tasks onto our group.
+ */
+ static struct sched_group *
+ find_busiest_group(struct sched_domain *sd, int this_cpu,
+ unsigned long *imbalance, enum cpu_idle_type idle,
+ int *sd_idle, const struct cpumask *cpus, int *balance)
+ {
+ struct sd_lb_stats sds;
- pwr_now += busiest->__cpu_power *
- min(busiest_load_per_task, max_load);
- pwr_now += this->__cpu_power *
- min(this_load_per_task, this_load);
- pwr_now /= SCHED_LOAD_SCALE;
-
- /* Amount of load we'd subtract */
- tmp = sg_div_cpu_power(busiest,
- busiest_load_per_task * SCHED_LOAD_SCALE);
- if (max_load > tmp)
- pwr_move += busiest->__cpu_power *
- min(busiest_load_per_task, max_load - tmp);
-
- /* Amount of load we'd add */
- if (max_load * busiest->__cpu_power <
- busiest_load_per_task * SCHED_LOAD_SCALE)
- tmp = sg_div_cpu_power(this,
- max_load * busiest->__cpu_power);
- else
- tmp = sg_div_cpu_power(this,
- busiest_load_per_task * SCHED_LOAD_SCALE);
- pwr_move += this->__cpu_power *
- min(this_load_per_task, this_load + tmp);
- pwr_move /= SCHED_LOAD_SCALE;
+ memset(&sds, 0, sizeof(sds));
- /* Move if we gain throughput */
- if (pwr_move > pwr_now)
- *imbalance = busiest_load_per_task;
- }
+ /*
+ * Compute the various statistics relavent for load balancing at
+ * this level.
+ */
+ update_sd_lb_stats(sd, this_cpu, idle, sd_idle, cpus,
+ balance, &sds);
+
+ /* Cases where imbalance does not exist from POV of this_cpu */
+ /* 1) this_cpu is not the appropriate cpu to perform load balancing
+ * at this level.
+ * 2) There is no busy sibling group to pull from.
+ * 3) This group is the busiest group.
+ * 4) This group is more busy than the avg busieness at this
+ * sched_domain.
+ * 5) The imbalance is within the specified limit.
+ * 6) Any rebalance would lead to ping-pong
+ */
+ if (balance && !(*balance))
+ goto ret;
- return busiest;
+ if (!sds.busiest || sds.busiest_nr_running == 0)
+ goto out_balanced;
- out_balanced:
- #if defined(CONFIG_SCHED_MC) || defined(CONFIG_SCHED_SMT)
- if (idle == CPU_NOT_IDLE || !(sd->flags & SD_POWERSAVINGS_BALANCE))
- goto ret;
+ if (sds.this_load >= sds.max_load)
+ goto out_balanced;
- if (this == group_leader && group_leader != group_min) {
- *imbalance = min_load_per_task;
- if (sched_mc_power_savings >= POWERSAVINGS_BALANCE_WAKEUP) {
- cpu_rq(this_cpu)->rd->sched_mc_preferred_wakeup_cpu =
- cpumask_first(sched_group_cpus(group_leader));
- }
- return group_min;
- }
- #endif
+ sds.avg_load = (SCHED_LOAD_SCALE * sds.total_load) / sds.total_pwr;
+
+ if (sds.this_load >= sds.avg_load)
+ goto out_balanced;
+
+ if (100 * sds.max_load <= sd->imbalance_pct * sds.this_load)
+ goto out_balanced;
+
+ sds.busiest_load_per_task /= sds.busiest_nr_running;
+ if (sds.group_imb)
+ sds.busiest_load_per_task =
+ min(sds.busiest_load_per_task, sds.avg_load);
+
+ /*
+ * We're trying to get all the cpus to the average_load, so we don't
+ * want to push ourselves above the average load, nor do we wish to
+ * reduce the max loaded cpu below the average load, as either of these
+ * actions would just result in more rebalancing later, and ping-pong
+ * tasks around. Thus we look for the minimum possible imbalance.
+ * Negative imbalances (*we* are more loaded than anyone else) will
+ * be counted as no imbalance for these purposes -- we can't fix that
+ * by pulling tasks to us. Be careful of negative numbers as they'll
+ * appear as very large values with unsigned longs.
+ */
+ if (sds.max_load <= sds.busiest_load_per_task)
+ goto out_balanced;
+
+ /* Looks like there is an imbalance. Compute it */
+ calculate_imbalance(&sds, this_cpu, imbalance);
+ return sds.busiest;
+
+ out_balanced:
+ /*
+ * There is no obvious imbalance. But check if we can do some balancing
+ * to save power.
+ */
+ if (check_power_save_busiest_group(&sds, this_cpu, imbalance))
+ return sds.busiest;
ret:
*imbalance = 0;
return NULL;
*/
#define MAX_PINNED_INTERVAL 512
+/* Working cpumask for load_balance and load_balance_newidle. */
+static DEFINE_PER_CPU(cpumask_var_t, load_balance_tmpmask);
+
/*
* Check this_cpu to ensure it is balanced within domain. Attempt to move
* tasks if there is an imbalance.
*/
static int load_balance(int this_cpu, struct rq *this_rq,
struct sched_domain *sd, enum cpu_idle_type idle,
- int *balance, struct cpumask *cpus)
+ int *balance)
{
int ld_moved, all_pinned = 0, active_balance = 0, sd_idle = 0;
struct sched_group *group;
unsigned long imbalance;
struct rq *busiest;
unsigned long flags;
+ struct cpumask *cpus = __get_cpu_var(load_balance_tmpmask);
cpumask_setall(cpus);
* this_rq is locked.
*/
static int
-load_balance_newidle(int this_cpu, struct rq *this_rq, struct sched_domain *sd,
- struct cpumask *cpus)
+load_balance_newidle(int this_cpu, struct rq *this_rq, struct sched_domain *sd)
{
struct sched_group *group;
struct rq *busiest = NULL;
int ld_moved = 0;
int sd_idle = 0;
int all_pinned = 0;
+ struct cpumask *cpus = __get_cpu_var(load_balance_tmpmask);
cpumask_setall(cpus);
struct sched_domain *sd;
int pulled_task = 0;
unsigned long next_balance = jiffies + HZ;
- cpumask_var_t tmpmask;
-
- if (!alloc_cpumask_var(&tmpmask, GFP_ATOMIC))
- return;
for_each_domain(this_cpu, sd) {
unsigned long interval;
if (sd->flags & SD_BALANCE_NEWIDLE)
/* If we've pulled tasks over stop searching: */
pulled_task = load_balance_newidle(this_cpu, this_rq,
- sd, tmpmask);
+ sd);
interval = msecs_to_jiffies(sd->balance_interval);
if (time_after(next_balance, sd->last_balance + interval))
*/
this_rq->next_balance = next_balance;
}
- free_cpumask_var(tmpmask);
}
/*
unsigned long next_balance = jiffies + 60*HZ;
int update_next_balance = 0;
int need_serialize;
- cpumask_var_t tmp;
-
- /* Fails alloc? Rebalancing probably not a priority right now. */
- if (!alloc_cpumask_var(&tmp, GFP_ATOMIC))
- return;
for_each_domain(cpu, sd) {
if (!(sd->flags & SD_LOAD_BALANCE))
}
if (time_after_eq(jiffies, sd->last_balance + interval)) {
- if (load_balance(cpu, rq, sd, idle, &balance, tmp)) {
+ if (load_balance(cpu, rq, sd, idle, &balance)) {
/*
* We've pulled tasks over so either we're no
* longer idle, or one of our SMT siblings is
*/
if (likely(update_next_balance))
rq->next_balance = next_balance;
-
- free_cpumask_var(tmp);
}
/*
#endif
}
+ static inline int on_null_domain(int cpu)
+ {
+ return !rcu_dereference(cpu_rq(cpu)->sd);
+ }
+
/*
* Trigger the SCHED_SOFTIRQ if it is time to do periodic load balancing.
*
cpumask_test_cpu(cpu, nohz.cpu_mask))
return;
#endif
- if (time_after_eq(jiffies, rq->next_balance))
+ /* Don't need to rebalance while attached to NULL domain */
+ if (time_after_eq(jiffies, rq->next_balance) &&
+ likely(!on_null_domain(cpu)))
raise_softirq(SCHED_SOFTIRQ);
}
#endif
}
+ static void put_prev_task(struct rq *rq, struct task_struct *prev)
+ {
+ if (prev->state == TASK_RUNNING) {
+ u64 runtime = prev->se.sum_exec_runtime;
+
+ runtime -= prev->se.prev_sum_exec_runtime;
+ runtime = min_t(u64, runtime, 2*sysctl_sched_migration_cost);
+
+ /*
+ * In order to avoid avg_overlap growing stale when we are
+ * indeed overlapping and hence not getting put to sleep, grow
+ * the avg_overlap on preemption.
+ *
+ * We use the average preemption runtime because that
+ * correlates to the amount of cache footprint a task can
+ * build up.
+ */
+ update_avg(&prev->se.avg_overlap, runtime);
+ }
+ prev->sched_class->put_prev_task(rq, prev);
+ }
+
/*
* Pick up the highest-prio task:
*/
static inline struct task_struct *
- pick_next_task(struct rq *rq, struct task_struct *prev)
+ pick_next_task(struct rq *rq)
{
const struct sched_class *class;
struct task_struct *p;
if (unlikely(!rq->nr_running))
idle_balance(cpu, rq);
- prev->sched_class->put_prev_task(rq, prev);
- next = pick_next_task(rq, prev);
+ put_prev_task(rq, prev);
+ next = pick_next_task(rq);
if (likely(prev != next)) {
sched_info_switch(prev, next);
* between schedule and now.
*/
barrier();
- } while (unlikely(test_thread_flag(TIF_NEED_RESCHED)));
+ } while (need_resched());
}
EXPORT_SYMBOL(preempt_schedule);
* between schedule and now.
*/
barrier();
- } while (unlikely(test_thread_flag(TIF_NEED_RESCHED)));
+ } while (need_resched());
}
#endif /* CONFIG_PREEMPT */
if (increment > 40)
increment = 40;
- nice = PRIO_TO_NICE(current->static_prio) + increment;
+ nice = TASK_NICE(current) + increment;
if (nice < -20)
nice = -20;
if (nice > 19)
if (!rq->nr_running)
break;
update_rq_clock(rq);
- next = pick_next_task(rq, rq->curr);
+ next = pick_next_task(rq);
if (!next)
break;
next->sched_class->put_prev_task(rq, next);
{
int group;
- cpumask_and(mask, &per_cpu(cpu_sibling_map, cpu), cpu_map);
+ cpumask_and(mask, topology_thread_cpumask(cpu), cpu_map);
group = cpumask_first(mask);
if (sg)
*sg = &per_cpu(sched_group_core, group).sg;
cpumask_and(mask, cpu_coregroup_mask(cpu), cpu_map);
group = cpumask_first(mask);
#elif defined(CONFIG_SCHED_SMT)
- cpumask_and(mask, &per_cpu(cpu_sibling_map, cpu), cpu_map);
+ cpumask_and(mask, topology_thread_cpumask(cpu), cpu_map);
group = cpumask_first(mask);
#else
group = cpu;
SD_INIT(sd, SIBLING);
set_domain_attribute(sd, attr);
cpumask_and(sched_domain_span(sd),
- &per_cpu(cpu_sibling_map, i), cpu_map);
+ topology_thread_cpumask(i), cpu_map);
sd->parent = p;
p->child = sd;
cpu_to_cpu_group(i, cpu_map, &sd->groups, tmpmask);
/* Set up CPU (sibling) groups */
for_each_cpu(i, cpu_map) {
cpumask_and(this_sibling_map,
- &per_cpu(cpu_sibling_map, i), cpu_map);
+ topology_thread_cpumask(i), cpu_map);
if (i != cpumask_first(this_sibling_map))
continue;
__set_bit(MAX_RT_PRIO, array->bitmap);
#if defined CONFIG_SMP || defined CONFIG_RT_GROUP_SCHED
- rt_rq->highest_prio = MAX_RT_PRIO;
+ rt_rq->highest_prio.curr = MAX_RT_PRIO;
+ #ifdef CONFIG_SMP
+ rt_rq->highest_prio.next = MAX_RT_PRIO;
+ #endif
#endif
#ifdef CONFIG_SMP
rt_rq->rt_nr_migratory = 0;
rt_rq->overloaded = 0;
+ plist_head_init(&rq->rt.pushable_tasks, &rq->lock);
#endif
rt_rq->rt_time = 0;
#endif
#ifdef CONFIG_USER_SCHED
alloc_size *= 2;
+#endif
+#ifdef CONFIG_CPUMASK_OFFSTACK
+ alloc_size += num_possible_cpus() * cpumask_size();
#endif
/*
* As sched_init() is called before page_alloc is setup,
ptr += nr_cpu_ids * sizeof(void **);
#endif /* CONFIG_USER_SCHED */
#endif /* CONFIG_RT_GROUP_SCHED */
+#ifdef CONFIG_CPUMASK_OFFSTACK
+ for_each_possible_cpu(i) {
+ per_cpu(load_balance_tmpmask, i) = (void *)ptr;
+ ptr += cpumask_size();
+ }
+#endif /* CONFIG_CPUMASK_OFFSTACK */
}
#ifdef CONFIG_SMP
struct cpuacct *ca;
int cpu;
- if (!cpuacct_subsys.active)
+ if (unlikely(!cpuacct_subsys.active))
return;
cpu = task_cpu(tsk);
* Move the pages to the [file or anon] inactive list.
*/
pagevec_init(&pvec, 1);
- pgmoved = 0;
lru = LRU_BASE + file * LRU_FILE;
spin_lock_irq(&zone->lru_lock);
*/
reclaim_stat->recent_rotated[!!file] += pgmoved;
+ pgmoved = 0;
while (!list_empty(&l_inactive)) {
page = lru_to_page(&l_inactive);
prefetchw_prev_lru_page(page, &l_inactive, flags);
struct reclaim_state reclaim_state = {
.reclaimed_slab = 0,
};
- node_to_cpumask_ptr(cpumask, pgdat->node_id);
+ const struct cpumask *cpumask = cpumask_of_node(pgdat->node_id);
if (!cpumask_empty(cpumask))
set_cpus_allowed_ptr(tsk, cpumask);
if (action == CPU_ONLINE || action == CPU_ONLINE_FROZEN) {
for_each_node_state(nid, N_HIGH_MEMORY) {
pg_data_t *pgdat = NODE_DATA(nid);
- node_to_cpumask_ptr(mask, pgdat->node_id);
+ const struct cpumask *mask;
+
+ mask = cpumask_of_node(pgdat->node_id);
if (cpumask_any_and(cpu_online_mask, mask) < nr_cpu_ids)
/* One of our CPUs online: restore mask */