int ltt_statedump_thread(void *data)
{
struct semaphore work_sema4;
int cpu;
printk(KERN_DEBUG "ltt_statedump_thread\n");
ltt_enumerate_process_states();
ltt_enumerate_file_descriptors();
ltt_enumerate_modules();
ltt_enumerate_vm_maps();
ltt_enumerate_interrupts();
ltt_enumerate_network_ip_interface();
/* Fire off a work queue on each CPU. Their sole purpose in life
* is to guarantee that each CPU has been in a state where is was in
* syscall mode (i.e. not in a trap, an IRQ or a soft IRQ) */
sema_init(&work_sema4, 1 - num_online_cpus());
lock_cpu_hotplug();
for_each_online_cpu(cpu)
{
INIT_WORK(&cpu_work[cpu], ltt_statedump_work_func, &work_sema4);
schedule_delayed_work_on(cpu,&cpu_work[cpu],0);
}
unlock_cpu_hotplug();
/* Wait for all work queues to have completed */
down(&work_sema4);
/* Our work is done */
printk(KERN_DEBUG "trace_statedump_statedump_end\n");
trace_statedump_statedump_end();
do_exit(0);
return 0;
}
static int do_ltt_statedump(void)
{
int cpu;
printk(KERN_DEBUG "do_ltt_statedump\n");
ltt_enumerate_process_states();
ltt_enumerate_file_descriptors();
list_modules();
ltt_enumerate_vm_maps();
list_interrupts();
ltt_enumerate_network_ip_interface();
/* Fire off a work queue on each CPU. Their sole purpose in life
* is to guarantee that each CPU has been in a state where is was in
* syscall mode (i.e. not in a trap, an IRQ or a soft IRQ) */
lock_cpu_hotplug();
atomic_set(&kernel_threads_to_run, num_online_cpus());
__set_current_state(TASK_UNINTERRUPTIBLE);
for_each_online_cpu(cpu) {
INIT_WORK(&cpu_work[cpu], ltt_statedump_work_func, current);
schedule_delayed_work_on(cpu,&cpu_work[cpu],0);
}
unlock_cpu_hotplug();
/* Wait for all threads to run */
schedule();
BUG_ON(atomic_read(&kernel_threads_to_run) != 0);
/* Our work is done */
printk(KERN_DEBUG "trace_statedump_statedump_end\n");
trace_statedump_statedump_end();
return 0;
}
/*
* Update the present map for a cpu node which is going away, and set
* the hard id in the paca(s) to -1 to be consistent with boot time
* convention for non-present cpus.
*/
static void pSeries_remove_processor(struct device_node *np)
{
unsigned int cpu;
int len, nthreads, i;
u32 *intserv;
intserv = (u32 *)get_property(np, "ibm,ppc-interrupt-server#s", &len);
if (!intserv)
return;
nthreads = len / sizeof(u32);
lock_cpu_hotplug();
for (i = 0; i < nthreads; i++) {
for_each_present_cpu(cpu) {
if (get_hard_smp_processor_id(cpu) != intserv[i])
continue;
BUG_ON(cpu_online(cpu));
cpu_clear(cpu, cpu_present_map);
set_hard_smp_processor_id(cpu, -1);
break;
}
if (cpu == NR_CPUS)
printk(KERN_WARNING "Could not find cpu to remove "
"with physical id 0x%x\n", intserv[i]);
}
unlock_cpu_hotplug();
}
static int
__init cpufreq_stats_init(void)
{
int ret;
unsigned int cpu;
spin_lock_init(&cpufreq_stats_lock);
if ((ret = cpufreq_register_notifier(¬ifier_policy_block,
CPUFREQ_POLICY_NOTIFIER)))
return ret;
if ((ret = cpufreq_register_notifier(¬ifier_trans_block,
CPUFREQ_TRANSITION_NOTIFIER))) {
cpufreq_unregister_notifier(¬ifier_policy_block,
CPUFREQ_POLICY_NOTIFIER);
return ret;
}
register_cpu_notifier(&cpufreq_stat_cpu_notifier);
lock_cpu_hotplug();
for_each_online_cpu(cpu) {
cpufreq_stat_cpu_callback(&cpufreq_stat_cpu_notifier, CPU_ONLINE,
(void *)(long)cpu);
}
unlock_cpu_hotplug();
return 0;
}
static int do_ltt_statedump(struct ltt_probe_private_data *call_data)
{
int cpu;
printk(KERN_DEBUG "LTT state dump thread start\n");
ltt_enumerate_process_states(call_data);
ltt_enumerate_file_descriptors(call_data);
list_modules(call_data);
ltt_enumerate_vm_maps(call_data);
list_interrupts(call_data);
ltt_enumerate_network_ip_interface(call_data);
/*
* Fire off a work queue on each CPU. Their sole purpose in life
* is to guarantee that each CPU has been in a state where is was in
* syscall mode (i.e. not in a trap, an IRQ or a soft IRQ).
*/
lock_cpu_hotplug();
atomic_set(&kernel_threads_to_run, num_online_cpus());
work_wake_task = current;
__set_current_state(TASK_UNINTERRUPTIBLE);
for_each_online_cpu(cpu) {
INIT_DELAYED_WORK(&cpu_work[cpu], ltt_statedump_work_func);
schedule_delayed_work_on(cpu, &cpu_work[cpu], 0);
}
unlock_cpu_hotplug();
/* Wait for all threads to run */
schedule();
BUG_ON(atomic_read(&kernel_threads_to_run) != 0);
/* Our work is done */
printk(KERN_DEBUG "LTT state dump end\n");
__trace_mark(0, list_statedump_end, call_data, MARK_NOARGS);
return 0;
}
int mtrr_del_page(int reg, unsigned long base, unsigned long size)
{
int i, max;
mtrr_type ltype;
unsigned long lbase;
unsigned int lsize;
int error = -EINVAL;
if (!mtrr_if)
return -ENXIO;
max = num_var_ranges;
/* No CPU hotplug when we change MTRR entries */
lock_cpu_hotplug();
mutex_lock(&mtrr_mutex);
if (reg < 0) {
/* Search for existing MTRR */
for (i = 0; i < max; ++i) {
mtrr_if->get(i, &lbase, &lsize, <ype);
if (lbase == base && lsize == size) {
reg = i;
break;
}
}
if (reg < 0) {
printk(KERN_DEBUG "mtrr: no MTRR for %lx000,%lx000 found\n", base,
size);
goto out;
}
}
if (reg >= max) {
printk(KERN_WARNING "mtrr: register: %d too big\n", reg);
goto out;
}
if (is_cpu(CYRIX) && !use_intel()) {
if ((reg == 3) && arr3_protected) {
printk(KERN_WARNING "mtrr: ARR3 cannot be changed\n");
goto out;
}
}
mtrr_if->get(reg, &lbase, &lsize, <ype);
if (lsize < 1) {
printk(KERN_WARNING "mtrr: MTRR %d not used\n", reg);
goto out;
}
if (usage_table[reg] < 1) {
printk(KERN_WARNING "mtrr: reg: %d has count=0\n", reg);
goto out;
}
if (--usage_table[reg] < 1)
set_mtrr(reg, 0, 0, 0);
error = reg;
out:
mutex_unlock(&mtrr_mutex);
unlock_cpu_hotplug();
return error;
}
static void do_event_scan_all_cpus(long delay)
{
int cpu;
lock_cpu_hotplug();
cpu = first_cpu(cpu_online_map);
for (;;) {
set_cpus_allowed(current, cpumask_of_cpu(cpu));
do_event_scan(rtas_token("event-scan"));
set_cpus_allowed(current, CPU_MASK_ALL);
/* Drop hotplug lock, and sleep for the specified delay */
unlock_cpu_hotplug();
msleep_interruptible(delay);
lock_cpu_hotplug();
cpu = next_cpu(cpu, cpu_online_map);
if (cpu == NR_CPUS)
break;
}
unlock_cpu_hotplug();
}
/*
* Update cpu_present_map and paca(s) for a new cpu node. The wrinkle
* here is that a cpu device node may represent up to two logical cpus
* in the SMT case. We must honor the assumption in other code that
* the logical ids for sibling SMT threads x and y are adjacent, such
* that x^1 == y and y^1 == x.
*/
static int pSeries_add_processor(struct device_node *np)
{
unsigned int cpu;
cpumask_t candidate_map, tmp = CPU_MASK_NONE;
int err = -ENOSPC, len, nthreads, i;
u32 *intserv;
intserv = (u32 *)get_property(np, "ibm,ppc-interrupt-server#s", &len);
if (!intserv)
return 0;
nthreads = len / sizeof(u32);
for (i = 0; i < nthreads; i++)
cpu_set(i, tmp);
lock_cpu_hotplug();
BUG_ON(!cpus_subset(cpu_present_map, cpu_possible_map));
/* Get a bitmap of unoccupied slots. */
cpus_xor(candidate_map, cpu_possible_map, cpu_present_map);
if (cpus_empty(candidate_map)) {
/* If we get here, it most likely means that NR_CPUS is
* less than the partition's max processors setting.
*/
printk(KERN_ERR "Cannot add cpu %s; this system configuration"
" supports %d logical cpus.\n", np->full_name,
cpus_weight(cpu_possible_map));
goto out_unlock;
}
while (!cpus_empty(tmp))
if (cpus_subset(tmp, candidate_map))
/* Found a range where we can insert the new cpu(s) */
break;
else
cpus_shift_left(tmp, tmp, nthreads);
if (cpus_empty(tmp)) {
printk(KERN_ERR "Unable to find space in cpu_present_map for"
" processor %s with %d thread(s)\n", np->name,
nthreads);
goto out_unlock;
}
for_each_cpu_mask(cpu, tmp) {
BUG_ON(cpu_isset(cpu, cpu_present_map));
cpu_set(cpu, cpu_present_map);
set_hard_smp_processor_id(cpu, *intserv++);
}
static void
__exit cpufreq_stats_exit(void)
{
unsigned int cpu;
cpufreq_unregister_notifier(¬ifier_policy_block,
CPUFREQ_POLICY_NOTIFIER);
cpufreq_unregister_notifier(¬ifier_trans_block,
CPUFREQ_TRANSITION_NOTIFIER);
unregister_cpu_notifier(&cpufreq_stat_cpu_notifier);
lock_cpu_hotplug();
for_each_online_cpu(cpu) {
cpufreq_stat_cpu_callback(&cpufreq_stat_cpu_notifier, CPU_DEAD,
(void *)(long)cpu);
}
unlock_cpu_hotplug();
}
void flow_cache_flush(void)
{
struct flow_flush_info info;
static DECLARE_MUTEX(flow_flush_sem);
/* Don't want cpus going down or up during this. */
lock_cpu_hotplug();
down(&flow_flush_sem);
atomic_set(&info.cpuleft, num_online_cpus());
init_completion(&info.completion);
local_bh_disable();
smp_call_function(flow_cache_flush_per_cpu, &info, 1, 0);
flow_cache_flush_tasklet((unsigned long)&info);
local_bh_enable();
wait_for_completion(&info.completion);
up(&flow_flush_sem);
unlock_cpu_hotplug();
}
int mtrr_add_page(unsigned long base, unsigned long size,
unsigned int type, char increment)
{
int i;
mtrr_type ltype;
unsigned long lbase;
unsigned int lsize;
int error;
if (!mtrr_if)
return -ENXIO;
if ((error = mtrr_if->validate_add_page(base,size,type)))
return error;
if (type >= MTRR_NUM_TYPES) {
printk(KERN_WARNING "mtrr: type: %u invalid\n", type);
return -EINVAL;
}
/* If the type is WC, check that this processor supports it */
if ((type == MTRR_TYPE_WRCOMB) && !have_wrcomb()) {
printk(KERN_WARNING
"mtrr: your processor doesn't support write-combining\n");
return -ENOSYS;
}
if (base & size_or_mask || size & size_or_mask) {
printk(KERN_WARNING "mtrr: base or size exceeds the MTRR width\n");
return -EINVAL;
}
error = -EINVAL;
/* No CPU hotplug when we change MTRR entries */
lock_cpu_hotplug();
/* Search for existing MTRR */
mutex_lock(&mtrr_mutex);
for (i = 0; i < num_var_ranges; ++i) {
mtrr_if->get(i, &lbase, &lsize, <ype);
if (base >= lbase + lsize)
continue;
if ((base < lbase) && (base + size <= lbase))
continue;
/* At this point we know there is some kind of overlap/enclosure */
if ((base < lbase) || (base + size > lbase + lsize)) {
printk(KERN_WARNING
"mtrr: 0x%lx000,0x%lx000 overlaps existing"
" 0x%lx000,0x%x000\n", base, size, lbase,
lsize);
goto out;
}
/* New region is enclosed by an existing region */
if (ltype != type) {
if (type == MTRR_TYPE_UNCACHABLE)
continue;
printk (KERN_WARNING "mtrr: type mismatch for %lx000,%lx000 old: %s new: %s\n",
base, size, mtrr_attrib_to_str(ltype),
mtrr_attrib_to_str(type));
goto out;
}
if (increment)
++usage_table[i];
error = i;
goto out;
}
/* Search for an empty MTRR */
i = mtrr_if->get_free_region(base, size);
if (i >= 0) {
set_mtrr(i, base, size, type);
usage_table[i] = 1;
} else
printk(KERN_INFO "mtrr: no more MTRRs available\n");
error = i;
out:
mutex_unlock(&mtrr_mutex);
unlock_cpu_hotplug();
return error;
}
static int rtasd(void *unused)
{
unsigned int err_type;
int cpu = 0;
int event_scan = rtas_token("event-scan");
cpumask_t all = CPU_MASK_ALL;
int rc;
daemonize("rtasd");
if (event_scan == RTAS_UNKNOWN_SERVICE || get_eventscan_parms() == -1)
goto error;
rtas_log_buf = vmalloc(rtas_error_log_buffer_max*LOG_NUMBER);
if (!rtas_log_buf) {
printk(KERN_ERR "rtasd: no memory\n");
goto error;
}
/* We can use rtas_log_buf now */
no_more_logging = 0;
printk(KERN_ERR "RTAS daemon started\n");
DEBUG("will sleep for %d jiffies\n", (HZ*60/rtas_event_scan_rate) / 2);
/* See if we have any error stored in NVRAM */
memset(logdata, 0, rtas_error_log_max);
rc = nvram_read_error_log(logdata, rtas_error_log_max, &err_type);
if (!rc) {
if (err_type != ERR_FLAG_ALREADY_LOGGED) {
pSeries_log_error(logdata, err_type | ERR_FLAG_BOOT, 0);
}
}
/* First pass. */
lock_cpu_hotplug();
for_each_online_cpu(cpu) {
DEBUG("scheduling on %d\n", cpu);
set_cpus_allowed(current, cpumask_of_cpu(cpu));
DEBUG("watchdog scheduled on cpu %d\n", smp_processor_id());
do_event_scan(event_scan);
set_current_state(TASK_INTERRUPTIBLE);
schedule_timeout(HZ);
}
unlock_cpu_hotplug();
if (surveillance_timeout != -1) {
DEBUG("enabling surveillance\n");
enable_surveillance(surveillance_timeout);
DEBUG("surveillance enabled\n");
}
lock_cpu_hotplug();
cpu = first_cpu_const(mk_cpumask_const(cpu_online_map));
for (;;) {
set_cpus_allowed(current, cpumask_of_cpu(cpu));
do_event_scan(event_scan);
set_cpus_allowed(current, all);
/* Drop hotplug lock, and sleep for a bit (at least
* one second since some machines have problems if we
* call event-scan too quickly). */
unlock_cpu_hotplug();
set_current_state(TASK_INTERRUPTIBLE);
schedule_timeout((HZ*60/rtas_event_scan_rate) / 2);
lock_cpu_hotplug();
cpu = next_cpu_const(cpu, mk_cpumask_const(cpu_online_map));
if (cpu == NR_CPUS)
cpu = first_cpu_const(mk_cpumask_const(cpu_online_map));
}
error:
/* Should delete proc entries */
return -EINVAL;
}