2005-04-16 22:20:36 +00:00
|
|
|
/* CPU control.
|
|
|
|
* (C) 2001, 2002, 2003, 2004 Rusty Russell
|
|
|
|
*
|
|
|
|
* This code is licenced under the GPL.
|
|
|
|
*/
|
|
|
|
#include <linux/proc_fs.h>
|
|
|
|
#include <linux/smp.h>
|
|
|
|
#include <linux/init.h>
|
|
|
|
#include <linux/notifier.h>
|
|
|
|
#include <linux/sched.h>
|
|
|
|
#include <linux/unistd.h>
|
|
|
|
#include <linux/cpu.h>
|
|
|
|
#include <linux/module.h>
|
|
|
|
#include <linux/kthread.h>
|
|
|
|
#include <linux/stop_machine.h>
|
2006-06-26 07:24:32 +00:00
|
|
|
#include <linux/mutex.h>
|
include cleanup: Update gfp.h and slab.h includes to prepare for breaking implicit slab.h inclusion from percpu.h
percpu.h is included by sched.h and module.h and thus ends up being
included when building most .c files. percpu.h includes slab.h which
in turn includes gfp.h making everything defined by the two files
universally available and complicating inclusion dependencies.
percpu.h -> slab.h dependency is about to be removed. Prepare for
this change by updating users of gfp and slab facilities include those
headers directly instead of assuming availability. As this conversion
needs to touch large number of source files, the following script is
used as the basis of conversion.
http://userweb.kernel.org/~tj/misc/slabh-sweep.py
The script does the followings.
* Scan files for gfp and slab usages and update includes such that
only the necessary includes are there. ie. if only gfp is used,
gfp.h, if slab is used, slab.h.
* When the script inserts a new include, it looks at the include
blocks and try to put the new include such that its order conforms
to its surrounding. It's put in the include block which contains
core kernel includes, in the same order that the rest are ordered -
alphabetical, Christmas tree, rev-Xmas-tree or at the end if there
doesn't seem to be any matching order.
* If the script can't find a place to put a new include (mostly
because the file doesn't have fitting include block), it prints out
an error message indicating which .h file needs to be added to the
file.
The conversion was done in the following steps.
1. The initial automatic conversion of all .c files updated slightly
over 4000 files, deleting around 700 includes and adding ~480 gfp.h
and ~3000 slab.h inclusions. The script emitted errors for ~400
files.
2. Each error was manually checked. Some didn't need the inclusion,
some needed manual addition while adding it to implementation .h or
embedding .c file was more appropriate for others. This step added
inclusions to around 150 files.
3. The script was run again and the output was compared to the edits
from #2 to make sure no file was left behind.
4. Several build tests were done and a couple of problems were fixed.
e.g. lib/decompress_*.c used malloc/free() wrappers around slab
APIs requiring slab.h to be added manually.
5. The script was run on all .h files but without automatically
editing them as sprinkling gfp.h and slab.h inclusions around .h
files could easily lead to inclusion dependency hell. Most gfp.h
inclusion directives were ignored as stuff from gfp.h was usually
wildly available and often used in preprocessor macros. Each
slab.h inclusion directive was examined and added manually as
necessary.
6. percpu.h was updated not to include slab.h.
7. Build test were done on the following configurations and failures
were fixed. CONFIG_GCOV_KERNEL was turned off for all tests (as my
distributed build env didn't work with gcov compiles) and a few
more options had to be turned off depending on archs to make things
build (like ipr on powerpc/64 which failed due to missing writeq).
* x86 and x86_64 UP and SMP allmodconfig and a custom test config.
* powerpc and powerpc64 SMP allmodconfig
* sparc and sparc64 SMP allmodconfig
* ia64 SMP allmodconfig
* s390 SMP allmodconfig
* alpha SMP allmodconfig
* um on x86_64 SMP allmodconfig
8. percpu.h modifications were reverted so that it could be applied as
a separate patch and serve as bisection point.
Given the fact that I had only a couple of failures from tests on step
6, I'm fairly confident about the coverage of this conversion patch.
If there is a breakage, it's likely to be something in one of the arch
headers which should be easily discoverable easily on most builds of
the specific arch.
Signed-off-by: Tejun Heo <tj@kernel.org>
Guess-its-ok-by: Christoph Lameter <cl@linux-foundation.org>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: Lee Schermerhorn <Lee.Schermerhorn@hp.com>
2010-03-24 08:04:11 +00:00
|
|
|
#include <linux/gfp.h>
|
2005-04-16 22:20:36 +00:00
|
|
|
|
2008-12-13 10:49:41 +00:00
|
|
|
#ifdef CONFIG_SMP
|
2008-12-29 22:35:14 +00:00
|
|
|
/* Serializes the updates to cpu_online_mask, cpu_present_mask */
|
2006-07-23 19:12:16 +00:00
|
|
|
static DEFINE_MUTEX(cpu_add_remove_lock);
|
2005-04-16 22:20:36 +00:00
|
|
|
|
2010-05-26 21:43:36 +00:00
|
|
|
/*
|
|
|
|
* The following two API's must be used when attempting
|
|
|
|
* to serialize the updates to cpu_online_mask, cpu_present_mask.
|
|
|
|
*/
|
|
|
|
void cpu_maps_update_begin(void)
|
|
|
|
{
|
|
|
|
mutex_lock(&cpu_add_remove_lock);
|
|
|
|
}
|
|
|
|
|
|
|
|
void cpu_maps_update_done(void)
|
|
|
|
{
|
|
|
|
mutex_unlock(&cpu_add_remove_lock);
|
|
|
|
}
|
|
|
|
|
2006-10-17 07:10:35 +00:00
|
|
|
static __cpuinitdata RAW_NOTIFIER_HEAD(cpu_chain);
|
2005-04-16 22:20:36 +00:00
|
|
|
|
2006-09-26 06:32:48 +00:00
|
|
|
/* If set, cpu_up and cpu_down will return -EBUSY and do nothing.
|
|
|
|
* Should always be manipulated under cpu_add_remove_lock
|
|
|
|
*/
|
|
|
|
static int cpu_hotplug_disabled;
|
|
|
|
|
2010-05-26 21:43:36 +00:00
|
|
|
#ifdef CONFIG_HOTPLUG_CPU
|
|
|
|
|
2008-01-25 20:08:01 +00:00
|
|
|
static struct {
|
|
|
|
struct task_struct *active_writer;
|
|
|
|
struct mutex lock; /* Synchronizes accesses to refcount, */
|
|
|
|
/*
|
|
|
|
* Also blocks the new readers during
|
|
|
|
* an ongoing cpu hotplug operation.
|
|
|
|
*/
|
|
|
|
int refcount;
|
2009-06-23 04:18:12 +00:00
|
|
|
} cpu_hotplug = {
|
|
|
|
.active_writer = NULL,
|
|
|
|
.lock = __MUTEX_INITIALIZER(cpu_hotplug.lock),
|
|
|
|
.refcount = 0,
|
|
|
|
};
|
2008-01-25 20:08:01 +00:00
|
|
|
|
2008-01-25 20:08:02 +00:00
|
|
|
void get_online_cpus(void)
|
2005-11-28 21:43:46 +00:00
|
|
|
{
|
2008-01-25 20:08:01 +00:00
|
|
|
might_sleep();
|
|
|
|
if (cpu_hotplug.active_writer == current)
|
2006-07-23 19:12:16 +00:00
|
|
|
return;
|
2008-01-25 20:08:01 +00:00
|
|
|
mutex_lock(&cpu_hotplug.lock);
|
|
|
|
cpu_hotplug.refcount++;
|
|
|
|
mutex_unlock(&cpu_hotplug.lock);
|
|
|
|
|
2005-11-28 21:43:46 +00:00
|
|
|
}
|
2008-01-25 20:08:02 +00:00
|
|
|
EXPORT_SYMBOL_GPL(get_online_cpus);
|
2005-11-09 05:34:24 +00:00
|
|
|
|
2008-01-25 20:08:02 +00:00
|
|
|
void put_online_cpus(void)
|
2005-11-28 21:43:46 +00:00
|
|
|
{
|
2008-01-25 20:08:01 +00:00
|
|
|
if (cpu_hotplug.active_writer == current)
|
2006-07-23 19:12:16 +00:00
|
|
|
return;
|
2008-01-25 20:08:01 +00:00
|
|
|
mutex_lock(&cpu_hotplug.lock);
|
2008-04-29 08:00:29 +00:00
|
|
|
if (!--cpu_hotplug.refcount && unlikely(cpu_hotplug.active_writer))
|
|
|
|
wake_up_process(cpu_hotplug.active_writer);
|
2008-01-25 20:08:01 +00:00
|
|
|
mutex_unlock(&cpu_hotplug.lock);
|
|
|
|
|
2005-11-28 21:43:46 +00:00
|
|
|
}
|
2008-01-25 20:08:02 +00:00
|
|
|
EXPORT_SYMBOL_GPL(put_online_cpus);
|
2005-11-28 21:43:46 +00:00
|
|
|
|
2008-01-25 20:08:01 +00:00
|
|
|
/*
|
|
|
|
* This ensures that the hotplug operation can begin only when the
|
|
|
|
* refcount goes to zero.
|
|
|
|
*
|
|
|
|
* Note that during a cpu-hotplug operation, the new readers, if any,
|
|
|
|
* will be blocked by the cpu_hotplug.lock
|
|
|
|
*
|
2008-04-29 08:00:29 +00:00
|
|
|
* Since cpu_hotplug_begin() is always called after invoking
|
|
|
|
* cpu_maps_update_begin(), we can be sure that only one writer is active.
|
2008-01-25 20:08:01 +00:00
|
|
|
*
|
|
|
|
* Note that theoretically, there is a possibility of a livelock:
|
|
|
|
* - Refcount goes to zero, last reader wakes up the sleeping
|
|
|
|
* writer.
|
|
|
|
* - Last reader unlocks the cpu_hotplug.lock.
|
|
|
|
* - A new reader arrives at this moment, bumps up the refcount.
|
|
|
|
* - The writer acquires the cpu_hotplug.lock finds the refcount
|
|
|
|
* non zero and goes to sleep again.
|
|
|
|
*
|
|
|
|
* However, this is very difficult to achieve in practice since
|
2008-01-25 20:08:02 +00:00
|
|
|
* get_online_cpus() not an api which is called all that often.
|
2008-01-25 20:08:01 +00:00
|
|
|
*
|
|
|
|
*/
|
|
|
|
static void cpu_hotplug_begin(void)
|
|
|
|
{
|
|
|
|
cpu_hotplug.active_writer = current;
|
2008-04-29 08:00:29 +00:00
|
|
|
|
|
|
|
for (;;) {
|
|
|
|
mutex_lock(&cpu_hotplug.lock);
|
|
|
|
if (likely(!cpu_hotplug.refcount))
|
|
|
|
break;
|
|
|
|
__set_current_state(TASK_UNINTERRUPTIBLE);
|
2008-01-25 20:08:01 +00:00
|
|
|
mutex_unlock(&cpu_hotplug.lock);
|
|
|
|
schedule();
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
static void cpu_hotplug_done(void)
|
|
|
|
{
|
|
|
|
cpu_hotplug.active_writer = NULL;
|
|
|
|
mutex_unlock(&cpu_hotplug.lock);
|
|
|
|
}
|
2010-05-26 21:43:36 +00:00
|
|
|
|
|
|
|
#else /* #if CONFIG_HOTPLUG_CPU */
|
|
|
|
static void cpu_hotplug_begin(void) {}
|
|
|
|
static void cpu_hotplug_done(void) {}
|
|
|
|
#endif /* #esle #if CONFIG_HOTPLUG_CPU */
|
|
|
|
|
2005-04-16 22:20:36 +00:00
|
|
|
/* Need to know about CPUs going up/down? */
|
2008-04-29 07:58:51 +00:00
|
|
|
int __ref register_cpu_notifier(struct notifier_block *nb)
|
2005-04-16 22:20:36 +00:00
|
|
|
{
|
2006-10-17 07:10:35 +00:00
|
|
|
int ret;
|
2008-01-25 20:08:01 +00:00
|
|
|
cpu_maps_update_begin();
|
2006-10-17 07:10:35 +00:00
|
|
|
ret = raw_notifier_chain_register(&cpu_chain, nb);
|
2008-01-25 20:08:01 +00:00
|
|
|
cpu_maps_update_done();
|
2006-10-17 07:10:35 +00:00
|
|
|
return ret;
|
2005-04-16 22:20:36 +00:00
|
|
|
}
|
2006-06-27 09:54:08 +00:00
|
|
|
|
2010-05-26 21:43:28 +00:00
|
|
|
static int __cpu_notify(unsigned long val, void *v, int nr_to_call,
|
|
|
|
int *nr_calls)
|
|
|
|
{
|
2010-05-26 21:43:29 +00:00
|
|
|
int ret;
|
|
|
|
|
|
|
|
ret = __raw_notifier_call_chain(&cpu_chain, val, v, nr_to_call,
|
2010-05-26 21:43:28 +00:00
|
|
|
nr_calls);
|
2010-05-26 21:43:29 +00:00
|
|
|
|
|
|
|
return notifier_to_errno(ret);
|
2010-05-26 21:43:28 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
static int cpu_notify(unsigned long val, void *v)
|
|
|
|
{
|
|
|
|
return __cpu_notify(val, v, -1, NULL);
|
|
|
|
}
|
|
|
|
|
2010-05-27 17:32:08 +00:00
|
|
|
#ifdef CONFIG_HOTPLUG_CPU
|
|
|
|
|
2010-05-26 21:43:28 +00:00
|
|
|
static void cpu_notify_nofail(unsigned long val, void *v)
|
|
|
|
{
|
2010-05-27 17:32:08 +00:00
|
|
|
BUG_ON(cpu_notify(val, v));
|
2010-05-26 21:43:28 +00:00
|
|
|
}
|
|
|
|
|
2005-04-16 22:20:36 +00:00
|
|
|
EXPORT_SYMBOL(register_cpu_notifier);
|
|
|
|
|
2008-04-29 07:58:48 +00:00
|
|
|
void __ref unregister_cpu_notifier(struct notifier_block *nb)
|
2005-04-16 22:20:36 +00:00
|
|
|
{
|
2008-01-25 20:08:01 +00:00
|
|
|
cpu_maps_update_begin();
|
2006-10-17 07:10:35 +00:00
|
|
|
raw_notifier_chain_unregister(&cpu_chain, nb);
|
2008-01-25 20:08:01 +00:00
|
|
|
cpu_maps_update_done();
|
2005-04-16 22:20:36 +00:00
|
|
|
}
|
|
|
|
EXPORT_SYMBOL(unregister_cpu_notifier);
|
|
|
|
|
|
|
|
static inline void check_for_tasks(int cpu)
|
|
|
|
{
|
|
|
|
struct task_struct *p;
|
|
|
|
|
|
|
|
write_lock_irq(&tasklist_lock);
|
|
|
|
for_each_process(p) {
|
2010-01-21 15:34:27 +00:00
|
|
|
if (task_cpu(p) == cpu && p->state == TASK_RUNNING &&
|
2005-04-16 22:20:36 +00:00
|
|
|
(!cputime_eq(p->utime, cputime_zero) ||
|
|
|
|
!cputime_eq(p->stime, cputime_zero)))
|
2010-01-25 13:56:34 +00:00
|
|
|
printk(KERN_WARNING "Task %s (pid = %d) is on cpu %d "
|
|
|
|
"(state = %ld, flags = %x)\n",
|
|
|
|
p->comm, task_pid_nr(p), cpu,
|
|
|
|
p->state, p->flags);
|
2005-04-16 22:20:36 +00:00
|
|
|
}
|
|
|
|
write_unlock_irq(&tasklist_lock);
|
|
|
|
}
|
|
|
|
|
2007-05-24 09:23:10 +00:00
|
|
|
struct take_cpu_down_param {
|
2010-03-15 09:10:23 +00:00
|
|
|
struct task_struct *caller;
|
2007-05-24 09:23:10 +00:00
|
|
|
unsigned long mod;
|
|
|
|
void *hcpu;
|
|
|
|
};
|
|
|
|
|
2005-04-16 22:20:36 +00:00
|
|
|
/* Take this CPU down. */
|
2008-04-29 07:58:50 +00:00
|
|
|
static int __ref take_cpu_down(void *_param)
|
2005-04-16 22:20:36 +00:00
|
|
|
{
|
2007-05-24 09:23:10 +00:00
|
|
|
struct take_cpu_down_param *param = _param;
|
2010-03-15 09:10:23 +00:00
|
|
|
unsigned int cpu = (unsigned long)param->hcpu;
|
2005-04-16 22:20:36 +00:00
|
|
|
int err;
|
|
|
|
|
|
|
|
/* Ensure this CPU doesn't handle any more interrupts. */
|
|
|
|
err = __cpu_disable();
|
|
|
|
if (err < 0)
|
2005-06-25 21:54:50 +00:00
|
|
|
return err;
|
2005-04-16 22:20:36 +00:00
|
|
|
|
2010-05-26 21:43:28 +00:00
|
|
|
cpu_notify(CPU_DYING | param->mod, param->hcpu);
|
2008-08-31 17:58:49 +00:00
|
|
|
|
2010-03-15 09:10:23 +00:00
|
|
|
if (task_cpu(param->caller) == cpu)
|
|
|
|
move_task_off_dead_cpu(cpu, param->caller);
|
2005-06-25 21:54:50 +00:00
|
|
|
/* Force idle task to run as soon as we yield: it should
|
|
|
|
immediately notice cpu is offline and die quickly. */
|
|
|
|
sched_idle_next();
|
|
|
|
return 0;
|
2005-04-16 22:20:36 +00:00
|
|
|
}
|
|
|
|
|
2006-09-26 06:32:48 +00:00
|
|
|
/* Requires cpu_add_remove_lock to be held */
|
2008-04-29 07:58:50 +00:00
|
|
|
static int __ref _cpu_down(unsigned int cpu, int tasks_frozen)
|
2005-04-16 22:20:36 +00:00
|
|
|
{
|
2007-05-09 09:34:04 +00:00
|
|
|
int err, nr_calls = 0;
|
|
|
|
void *hcpu = (void *)(long)cpu;
|
2007-05-09 09:35:10 +00:00
|
|
|
unsigned long mod = tasks_frozen ? CPU_TASKS_FROZEN : 0;
|
2007-05-24 09:23:10 +00:00
|
|
|
struct take_cpu_down_param tcd_param = {
|
2010-03-15 09:10:23 +00:00
|
|
|
.caller = current,
|
2007-05-24 09:23:10 +00:00
|
|
|
.mod = mod,
|
|
|
|
.hcpu = hcpu,
|
|
|
|
};
|
2005-04-16 22:20:36 +00:00
|
|
|
|
2006-09-26 06:32:48 +00:00
|
|
|
if (num_online_cpus() == 1)
|
|
|
|
return -EBUSY;
|
2005-04-16 22:20:36 +00:00
|
|
|
|
2006-09-26 06:32:48 +00:00
|
|
|
if (!cpu_online(cpu))
|
|
|
|
return -EINVAL;
|
2005-04-16 22:20:36 +00:00
|
|
|
|
2008-01-25 20:08:01 +00:00
|
|
|
cpu_hotplug_begin();
|
2009-12-16 17:04:32 +00:00
|
|
|
set_cpu_active(cpu, false);
|
2010-05-26 21:43:28 +00:00
|
|
|
err = __cpu_notify(CPU_DOWN_PREPARE | mod, hcpu, -1, &nr_calls);
|
2010-05-26 21:43:29 +00:00
|
|
|
if (err) {
|
2009-11-25 12:31:39 +00:00
|
|
|
set_cpu_active(cpu, true);
|
|
|
|
|
2007-10-18 10:05:12 +00:00
|
|
|
nr_calls--;
|
2010-05-26 21:43:28 +00:00
|
|
|
__cpu_notify(CPU_DOWN_FAILED | mod, hcpu, nr_calls, NULL);
|
2005-04-16 22:20:36 +00:00
|
|
|
printk("%s: attempt to take down CPU %u failed\n",
|
2008-04-30 07:55:08 +00:00
|
|
|
__func__, cpu);
|
2007-05-09 09:34:03 +00:00
|
|
|
goto out_release;
|
2005-04-16 22:20:36 +00:00
|
|
|
}
|
|
|
|
|
2008-12-31 23:42:28 +00:00
|
|
|
err = __stop_machine(take_cpu_down, &tcd_param, cpumask_of(cpu));
|
2008-07-28 17:16:29 +00:00
|
|
|
if (err) {
|
2009-11-25 12:31:39 +00:00
|
|
|
set_cpu_active(cpu, true);
|
2005-04-16 22:20:36 +00:00
|
|
|
/* CPU didn't die: tell everyone. Can't complain. */
|
2010-05-26 21:43:28 +00:00
|
|
|
cpu_notify_nofail(CPU_DOWN_FAILED | mod, hcpu);
|
2005-04-16 22:20:36 +00:00
|
|
|
|
2010-03-15 09:10:23 +00:00
|
|
|
goto out_release;
|
2006-10-28 17:38:57 +00:00
|
|
|
}
|
2008-07-28 17:16:29 +00:00
|
|
|
BUG_ON(cpu_online(cpu));
|
2005-04-16 22:20:36 +00:00
|
|
|
|
|
|
|
/* Wait for it to sleep (leaving idle task). */
|
|
|
|
while (!idle_cpu(cpu))
|
|
|
|
yield();
|
|
|
|
|
|
|
|
/* This actually kills the CPU. */
|
|
|
|
__cpu_die(cpu);
|
|
|
|
|
|
|
|
/* CPU is completely dead: tell everyone. Too late to complain. */
|
2010-05-26 21:43:28 +00:00
|
|
|
cpu_notify_nofail(CPU_DEAD | mod, hcpu);
|
2005-04-16 22:20:36 +00:00
|
|
|
|
|
|
|
check_for_tasks(cpu);
|
|
|
|
|
2007-05-09 09:34:03 +00:00
|
|
|
out_release:
|
2008-01-25 20:08:01 +00:00
|
|
|
cpu_hotplug_done();
|
2010-05-26 21:43:28 +00:00
|
|
|
if (!err)
|
|
|
|
cpu_notify_nofail(CPU_POST_DEAD | mod, hcpu);
|
2006-09-26 06:32:48 +00:00
|
|
|
return err;
|
|
|
|
}
|
|
|
|
|
2008-04-29 07:58:50 +00:00
|
|
|
int __ref cpu_down(unsigned int cpu)
|
2006-09-26 06:32:48 +00:00
|
|
|
{
|
2008-12-22 11:36:30 +00:00
|
|
|
int err;
|
2006-09-26 06:32:48 +00:00
|
|
|
|
2008-01-25 20:08:01 +00:00
|
|
|
cpu_maps_update_begin();
|
2008-07-15 11:43:49 +00:00
|
|
|
|
|
|
|
if (cpu_hotplug_disabled) {
|
2006-09-26 06:32:48 +00:00
|
|
|
err = -EBUSY;
|
2008-07-15 11:43:49 +00:00
|
|
|
goto out;
|
|
|
|
}
|
|
|
|
|
|
|
|
err = _cpu_down(cpu, 0);
|
2006-09-26 06:32:48 +00:00
|
|
|
|
2008-07-15 11:43:49 +00:00
|
|
|
out:
|
2008-01-25 20:08:01 +00:00
|
|
|
cpu_maps_update_done();
|
2005-04-16 22:20:36 +00:00
|
|
|
return err;
|
|
|
|
}
|
2008-04-29 06:35:56 +00:00
|
|
|
EXPORT_SYMBOL(cpu_down);
|
2005-04-16 22:20:36 +00:00
|
|
|
#endif /*CONFIG_HOTPLUG_CPU*/
|
|
|
|
|
2006-09-26 06:32:48 +00:00
|
|
|
/* Requires cpu_add_remove_lock to be held */
|
2007-05-09 09:35:10 +00:00
|
|
|
static int __cpuinit _cpu_up(unsigned int cpu, int tasks_frozen)
|
2005-04-16 22:20:36 +00:00
|
|
|
{
|
2007-05-09 09:34:03 +00:00
|
|
|
int ret, nr_calls = 0;
|
2005-04-16 22:20:36 +00:00
|
|
|
void *hcpu = (void *)(long)cpu;
|
2007-05-09 09:35:10 +00:00
|
|
|
unsigned long mod = tasks_frozen ? CPU_TASKS_FROZEN : 0;
|
2005-04-16 22:20:36 +00:00
|
|
|
|
2006-09-26 06:32:48 +00:00
|
|
|
if (cpu_online(cpu) || !cpu_present(cpu))
|
|
|
|
return -EINVAL;
|
2005-11-09 05:34:24 +00:00
|
|
|
|
2008-01-25 20:08:01 +00:00
|
|
|
cpu_hotplug_begin();
|
2010-05-26 21:43:28 +00:00
|
|
|
ret = __cpu_notify(CPU_UP_PREPARE | mod, hcpu, -1, &nr_calls);
|
2010-05-26 21:43:29 +00:00
|
|
|
if (ret) {
|
2007-10-18 10:05:12 +00:00
|
|
|
nr_calls--;
|
2005-04-16 22:20:36 +00:00
|
|
|
printk("%s: attempt to bring up CPU %u failed\n",
|
2008-04-30 07:55:08 +00:00
|
|
|
__func__, cpu);
|
2005-04-16 22:20:36 +00:00
|
|
|
goto out_notify;
|
|
|
|
}
|
|
|
|
|
|
|
|
/* Arch-specific enabling code. */
|
|
|
|
ret = __cpu_up(cpu);
|
|
|
|
if (ret != 0)
|
|
|
|
goto out_notify;
|
2006-03-24 17:45:21 +00:00
|
|
|
BUG_ON(!cpu_online(cpu));
|
2005-04-16 22:20:36 +00:00
|
|
|
|
2009-03-31 04:05:12 +00:00
|
|
|
set_cpu_active(cpu, true);
|
sched, cpu hotplug: fix set_cpus_allowed() use in hotplug callbacks
Mark Langsdorf reported:
> One of my co-workers noticed that the powernow-k8
> driver no longer restarts when a CPU core is
> hot-disabled and then hot-enabled on AMD quad-core
> systems.
>
> The following comands work fine on 2.6.26 and fail
> on 2.6.27-rc1:
>
> echo 0 > /sys/devices/system/cpu/cpu3/online
> echo 1 > /sys/devices/system/cpu/cpu3/online
> find /sys -name cpufreq
>
> For 2.6.26, the find will return a cpufreq
> directory for each processor. In 2.6.27-rc1,
> the cpu3 directory is missing.
>
> After digging through the code, the following
> logic is failing when the core is hot-enabled
> at runtime. The code works during the boot
> sequence.
>
> cpumask_t = current->cpus_allowed;
> set_cpus_allowed_ptr(current, &cpumask_of_cpu(cpu));
> if (smp_processor_id() != cpu)
> return -ENODEV;
So set the CPU active before calling the CPU_ONLINE notifier chain,
there are a handful of notifiers that use set_cpus_allowed().
This fix also solves the problem with x86-microcode. I've sent
alternative patches for microcode, but as this "rely on
set_cpus_allowed_ptr() being workable in cpu-hotplug(CPU_ONLINE, ...)"
assumption seems to be more broad than what we thought, perhaps this fix
should be applied.
With this patch we define that by the moment CPU_ONLINE is being sent,
a 'cpu' is online and ready for tasks to be migrated onto it.
Signed-off-by: Dmitry Adamushko <dmitry.adamushko@gmail.com>
Reported-by: Mark Langsdorf <mark.langsdorf@amd.com>
Tested-by: Mark Langsdorf <mark.langsdorf@amd.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
2008-07-30 10:34:04 +00:00
|
|
|
|
2005-04-16 22:20:36 +00:00
|
|
|
/* Now call notifier in preparation. */
|
2010-05-26 21:43:28 +00:00
|
|
|
cpu_notify(CPU_ONLINE | mod, hcpu);
|
2005-04-16 22:20:36 +00:00
|
|
|
|
|
|
|
out_notify:
|
|
|
|
if (ret != 0)
|
2010-05-26 21:43:28 +00:00
|
|
|
__cpu_notify(CPU_UP_CANCELED | mod, hcpu, nr_calls, NULL);
|
2008-01-25 20:08:01 +00:00
|
|
|
cpu_hotplug_done();
|
2006-09-26 06:32:48 +00:00
|
|
|
|
|
|
|
return ret;
|
|
|
|
}
|
|
|
|
|
2007-01-11 07:15:34 +00:00
|
|
|
int __cpuinit cpu_up(unsigned int cpu)
|
2006-09-26 06:32:48 +00:00
|
|
|
{
|
|
|
|
int err = 0;
|
2010-05-24 21:32:41 +00:00
|
|
|
|
|
|
|
#ifdef CONFIG_MEMORY_HOTPLUG
|
|
|
|
int nid;
|
|
|
|
pg_data_t *pgdat;
|
|
|
|
#endif
|
|
|
|
|
2008-12-31 23:42:28 +00:00
|
|
|
if (!cpu_possible(cpu)) {
|
2007-10-19 06:40:47 +00:00
|
|
|
printk(KERN_ERR "can't online cpu %d because it is not "
|
|
|
|
"configured as may-hotadd at boot time\n", cpu);
|
2010-03-05 21:42:38 +00:00
|
|
|
#if defined(CONFIG_IA64)
|
2007-10-19 06:40:47 +00:00
|
|
|
printk(KERN_ERR "please check additional_cpus= boot "
|
|
|
|
"parameter\n");
|
|
|
|
#endif
|
|
|
|
return -EINVAL;
|
|
|
|
}
|
2006-09-26 06:32:48 +00:00
|
|
|
|
2010-05-24 21:32:41 +00:00
|
|
|
#ifdef CONFIG_MEMORY_HOTPLUG
|
|
|
|
nid = cpu_to_node(cpu);
|
|
|
|
if (!node_online(nid)) {
|
|
|
|
err = mem_online_node(nid);
|
|
|
|
if (err)
|
|
|
|
return err;
|
|
|
|
}
|
|
|
|
|
|
|
|
pgdat = NODE_DATA(nid);
|
|
|
|
if (!pgdat) {
|
|
|
|
printk(KERN_ERR
|
|
|
|
"Can't online cpu %d due to NULL pgdat\n", cpu);
|
|
|
|
return -ENOMEM;
|
|
|
|
}
|
|
|
|
|
2010-05-24 21:32:52 +00:00
|
|
|
if (pgdat->node_zonelists->_zonerefs->zone == NULL) {
|
|
|
|
mutex_lock(&zonelists_mutex);
|
2010-05-24 21:32:51 +00:00
|
|
|
build_all_zonelists(NULL);
|
2010-05-24 21:32:52 +00:00
|
|
|
mutex_unlock(&zonelists_mutex);
|
|
|
|
}
|
2010-05-24 21:32:41 +00:00
|
|
|
#endif
|
|
|
|
|
2008-01-25 20:08:01 +00:00
|
|
|
cpu_maps_update_begin();
|
2008-07-15 11:43:49 +00:00
|
|
|
|
|
|
|
if (cpu_hotplug_disabled) {
|
2006-09-26 06:32:48 +00:00
|
|
|
err = -EBUSY;
|
2008-07-15 11:43:49 +00:00
|
|
|
goto out;
|
|
|
|
}
|
|
|
|
|
|
|
|
err = _cpu_up(cpu, 0);
|
|
|
|
|
|
|
|
out:
|
2008-01-25 20:08:01 +00:00
|
|
|
cpu_maps_update_done();
|
2006-09-26 06:32:48 +00:00
|
|
|
return err;
|
|
|
|
}
|
|
|
|
|
2007-08-31 06:56:29 +00:00
|
|
|
#ifdef CONFIG_PM_SLEEP_SMP
|
2008-12-31 23:42:28 +00:00
|
|
|
static cpumask_var_t frozen_cpus;
|
2006-09-26 06:32:48 +00:00
|
|
|
|
|
|
|
int disable_nonboot_cpus(void)
|
|
|
|
{
|
2010-05-27 20:16:22 +00:00
|
|
|
int cpu, first_cpu, error = 0;
|
2006-09-26 06:32:48 +00:00
|
|
|
|
2008-01-25 20:08:01 +00:00
|
|
|
cpu_maps_update_begin();
|
2008-12-31 23:42:28 +00:00
|
|
|
first_cpu = cpumask_first(cpu_online_mask);
|
2009-12-16 17:04:32 +00:00
|
|
|
/*
|
|
|
|
* We take down all of the non-boot CPUs in one shot to avoid races
|
2006-09-26 06:32:48 +00:00
|
|
|
* with the userspace trying to use the CPU hotplug at the same time
|
|
|
|
*/
|
2008-12-31 23:42:28 +00:00
|
|
|
cpumask_clear(frozen_cpus);
|
2009-11-25 12:31:39 +00:00
|
|
|
|
2006-09-26 06:32:48 +00:00
|
|
|
printk("Disabling non-boot CPUs ...\n");
|
|
|
|
for_each_online_cpu(cpu) {
|
|
|
|
if (cpu == first_cpu)
|
|
|
|
continue;
|
2007-05-09 09:35:10 +00:00
|
|
|
error = _cpu_down(cpu, 1);
|
2009-11-18 00:22:13 +00:00
|
|
|
if (!error)
|
2008-12-31 23:42:28 +00:00
|
|
|
cpumask_set_cpu(cpu, frozen_cpus);
|
2009-11-18 00:22:13 +00:00
|
|
|
else {
|
2006-09-26 06:32:48 +00:00
|
|
|
printk(KERN_ERR "Error taking CPU%d down: %d\n",
|
|
|
|
cpu, error);
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
}
|
2009-07-01 02:31:07 +00:00
|
|
|
|
2006-09-26 06:32:48 +00:00
|
|
|
if (!error) {
|
|
|
|
BUG_ON(num_online_cpus() > 1);
|
|
|
|
/* Make sure the CPUs won't be enabled by someone else */
|
|
|
|
cpu_hotplug_disabled = 1;
|
|
|
|
} else {
|
2006-12-23 15:55:29 +00:00
|
|
|
printk(KERN_ERR "Non-boot CPUs are not disabled\n");
|
2006-09-26 06:32:48 +00:00
|
|
|
}
|
2008-01-25 20:08:01 +00:00
|
|
|
cpu_maps_update_done();
|
2006-09-26 06:32:48 +00:00
|
|
|
return error;
|
|
|
|
}
|
|
|
|
|
2009-08-20 01:05:36 +00:00
|
|
|
void __weak arch_enable_nonboot_cpus_begin(void)
|
|
|
|
{
|
|
|
|
}
|
|
|
|
|
|
|
|
void __weak arch_enable_nonboot_cpus_end(void)
|
|
|
|
{
|
|
|
|
}
|
|
|
|
|
2008-02-08 12:21:55 +00:00
|
|
|
void __ref enable_nonboot_cpus(void)
|
2006-09-26 06:32:48 +00:00
|
|
|
{
|
|
|
|
int cpu, error;
|
|
|
|
|
|
|
|
/* Allow everyone to use the CPU hotplug again */
|
2008-01-25 20:08:01 +00:00
|
|
|
cpu_maps_update_begin();
|
2006-09-26 06:32:48 +00:00
|
|
|
cpu_hotplug_disabled = 0;
|
2008-12-31 23:42:28 +00:00
|
|
|
if (cpumask_empty(frozen_cpus))
|
2007-04-02 06:49:49 +00:00
|
|
|
goto out;
|
2006-09-26 06:32:48 +00:00
|
|
|
|
|
|
|
printk("Enabling non-boot CPUs ...\n");
|
2009-08-20 01:05:36 +00:00
|
|
|
|
|
|
|
arch_enable_nonboot_cpus_begin();
|
|
|
|
|
2008-12-31 23:42:28 +00:00
|
|
|
for_each_cpu(cpu, frozen_cpus) {
|
2007-05-09 09:35:10 +00:00
|
|
|
error = _cpu_up(cpu, 1);
|
2006-09-26 06:32:48 +00:00
|
|
|
if (!error) {
|
|
|
|
printk("CPU%d is up\n", cpu);
|
|
|
|
continue;
|
|
|
|
}
|
2007-04-02 06:49:49 +00:00
|
|
|
printk(KERN_WARNING "Error taking CPU%d up: %d\n", cpu, error);
|
2006-09-26 06:32:48 +00:00
|
|
|
}
|
2009-08-20 01:05:36 +00:00
|
|
|
|
|
|
|
arch_enable_nonboot_cpus_end();
|
|
|
|
|
2008-12-31 23:42:28 +00:00
|
|
|
cpumask_clear(frozen_cpus);
|
2007-04-02 06:49:49 +00:00
|
|
|
out:
|
2008-01-25 20:08:01 +00:00
|
|
|
cpu_maps_update_done();
|
2005-04-16 22:20:36 +00:00
|
|
|
}
|
2008-12-31 23:42:28 +00:00
|
|
|
|
|
|
|
static int alloc_frozen_cpus(void)
|
|
|
|
{
|
|
|
|
if (!alloc_cpumask_var(&frozen_cpus, GFP_KERNEL|__GFP_ZERO))
|
|
|
|
return -ENOMEM;
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
core_initcall(alloc_frozen_cpus);
|
2007-08-31 06:56:29 +00:00
|
|
|
#endif /* CONFIG_PM_SLEEP_SMP */
|
2008-05-29 18:17:02 +00:00
|
|
|
|
2008-09-07 14:57:22 +00:00
|
|
|
/**
|
|
|
|
* notify_cpu_starting(cpu) - call the CPU_STARTING notifiers
|
|
|
|
* @cpu: cpu that just started
|
|
|
|
*
|
|
|
|
* This function calls the cpu_chain notifiers with CPU_STARTING.
|
|
|
|
* It must be called by the arch code on the new cpu, before the new cpu
|
|
|
|
* enables interrupts and before the "boot" cpu returns from __cpu_up().
|
|
|
|
*/
|
2008-11-22 17:36:44 +00:00
|
|
|
void __cpuinit notify_cpu_starting(unsigned int cpu)
|
2008-09-07 14:57:22 +00:00
|
|
|
{
|
|
|
|
unsigned long val = CPU_STARTING;
|
|
|
|
|
|
|
|
#ifdef CONFIG_PM_SLEEP_SMP
|
2008-12-31 23:42:28 +00:00
|
|
|
if (frozen_cpus != NULL && cpumask_test_cpu(cpu, frozen_cpus))
|
2008-09-07 14:57:22 +00:00
|
|
|
val = CPU_STARTING_FROZEN;
|
|
|
|
#endif /* CONFIG_PM_SLEEP_SMP */
|
2010-05-26 21:43:28 +00:00
|
|
|
cpu_notify(val, (void *)(long)cpu);
|
2008-09-07 14:57:22 +00:00
|
|
|
}
|
|
|
|
|
2008-05-29 18:17:02 +00:00
|
|
|
#endif /* CONFIG_SMP */
|
2008-07-25 01:21:29 +00:00
|
|
|
|
cpu masks: optimize and clean up cpumask_of_cpu()
Clean up and optimize cpumask_of_cpu(), by sharing all the zero words.
Instead of stupidly generating all possible i=0...NR_CPUS 2^i patterns
creating a huge array of constant bitmasks, realize that the zero words
can be shared.
In other words, on a 64-bit architecture, we only ever need 64 of these
arrays - with a different bit set in one single world (with enough zero
words around it so that we can create any bitmask by just offsetting in
that big array). And then we just put enough zeroes around it that we
can point every single cpumask to be one of those things.
So when we have 4k CPU's, instead of having 4k arrays (of 4k bits each,
with one bit set in each array - 2MB memory total), we have exactly 64
arrays instead, each 8k bits in size (64kB total).
And then we just point cpumask(n) to the right position (which we can
calculate dynamically). Once we have the right arrays, getting
"cpumask(n)" ends up being:
static inline const cpumask_t *get_cpu_mask(unsigned int cpu)
{
const unsigned long *p = cpu_bit_bitmap[1 + cpu % BITS_PER_LONG];
p -= cpu / BITS_PER_LONG;
return (const cpumask_t *)p;
}
This brings other advantages and simplifications as well:
- we are not wasting memory that is just filled with a single bit in
various different places
- we don't need all those games to re-create the arrays in some dense
format, because they're already going to be dense enough.
if we compile a kernel for up to 4k CPU's, "wasting" that 64kB of memory
is a non-issue (especially since by doing this "overlapping" trick we
probably get better cache behaviour anyway).
[ mingo@elte.hu:
Converted Linus's mails into a commit. See:
http://lkml.org/lkml/2008/7/27/156
http://lkml.org/lkml/2008/7/28/320
Also applied a family filter - which also has the side-effect of leaving
out the bits where Linus calls me an idio... Oh, never mind ;-)
]
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Cc: Rusty Russell <rusty@rustcorp.com.au>
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: Al Viro <viro@ZenIV.linux.org.uk>
Cc: Mike Travis <travis@sgi.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
2008-07-28 18:32:33 +00:00
|
|
|
/*
|
|
|
|
* cpu_bit_bitmap[] is a special, "compressed" data structure that
|
|
|
|
* represents all NR_CPUS bits binary values of 1<<nr.
|
|
|
|
*
|
2008-12-31 23:42:28 +00:00
|
|
|
* It is used by cpumask_of() to get a constant address to a CPU
|
cpu masks: optimize and clean up cpumask_of_cpu()
Clean up and optimize cpumask_of_cpu(), by sharing all the zero words.
Instead of stupidly generating all possible i=0...NR_CPUS 2^i patterns
creating a huge array of constant bitmasks, realize that the zero words
can be shared.
In other words, on a 64-bit architecture, we only ever need 64 of these
arrays - with a different bit set in one single world (with enough zero
words around it so that we can create any bitmask by just offsetting in
that big array). And then we just put enough zeroes around it that we
can point every single cpumask to be one of those things.
So when we have 4k CPU's, instead of having 4k arrays (of 4k bits each,
with one bit set in each array - 2MB memory total), we have exactly 64
arrays instead, each 8k bits in size (64kB total).
And then we just point cpumask(n) to the right position (which we can
calculate dynamically). Once we have the right arrays, getting
"cpumask(n)" ends up being:
static inline const cpumask_t *get_cpu_mask(unsigned int cpu)
{
const unsigned long *p = cpu_bit_bitmap[1 + cpu % BITS_PER_LONG];
p -= cpu / BITS_PER_LONG;
return (const cpumask_t *)p;
}
This brings other advantages and simplifications as well:
- we are not wasting memory that is just filled with a single bit in
various different places
- we don't need all those games to re-create the arrays in some dense
format, because they're already going to be dense enough.
if we compile a kernel for up to 4k CPU's, "wasting" that 64kB of memory
is a non-issue (especially since by doing this "overlapping" trick we
probably get better cache behaviour anyway).
[ mingo@elte.hu:
Converted Linus's mails into a commit. See:
http://lkml.org/lkml/2008/7/27/156
http://lkml.org/lkml/2008/7/28/320
Also applied a family filter - which also has the side-effect of leaving
out the bits where Linus calls me an idio... Oh, never mind ;-)
]
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Cc: Rusty Russell <rusty@rustcorp.com.au>
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: Al Viro <viro@ZenIV.linux.org.uk>
Cc: Mike Travis <travis@sgi.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
2008-07-28 18:32:33 +00:00
|
|
|
* mask value that has a single bit set only.
|
|
|
|
*/
|
2008-07-25 01:21:29 +00:00
|
|
|
|
cpu masks: optimize and clean up cpumask_of_cpu()
Clean up and optimize cpumask_of_cpu(), by sharing all the zero words.
Instead of stupidly generating all possible i=0...NR_CPUS 2^i patterns
creating a huge array of constant bitmasks, realize that the zero words
can be shared.
In other words, on a 64-bit architecture, we only ever need 64 of these
arrays - with a different bit set in one single world (with enough zero
words around it so that we can create any bitmask by just offsetting in
that big array). And then we just put enough zeroes around it that we
can point every single cpumask to be one of those things.
So when we have 4k CPU's, instead of having 4k arrays (of 4k bits each,
with one bit set in each array - 2MB memory total), we have exactly 64
arrays instead, each 8k bits in size (64kB total).
And then we just point cpumask(n) to the right position (which we can
calculate dynamically). Once we have the right arrays, getting
"cpumask(n)" ends up being:
static inline const cpumask_t *get_cpu_mask(unsigned int cpu)
{
const unsigned long *p = cpu_bit_bitmap[1 + cpu % BITS_PER_LONG];
p -= cpu / BITS_PER_LONG;
return (const cpumask_t *)p;
}
This brings other advantages and simplifications as well:
- we are not wasting memory that is just filled with a single bit in
various different places
- we don't need all those games to re-create the arrays in some dense
format, because they're already going to be dense enough.
if we compile a kernel for up to 4k CPU's, "wasting" that 64kB of memory
is a non-issue (especially since by doing this "overlapping" trick we
probably get better cache behaviour anyway).
[ mingo@elte.hu:
Converted Linus's mails into a commit. See:
http://lkml.org/lkml/2008/7/27/156
http://lkml.org/lkml/2008/7/28/320
Also applied a family filter - which also has the side-effect of leaving
out the bits where Linus calls me an idio... Oh, never mind ;-)
]
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Cc: Rusty Russell <rusty@rustcorp.com.au>
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: Al Viro <viro@ZenIV.linux.org.uk>
Cc: Mike Travis <travis@sgi.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
2008-07-28 18:32:33 +00:00
|
|
|
/* cpu_bit_bitmap[0] is empty - so we can back into it */
|
|
|
|
#define MASK_DECLARE_1(x) [x+1][0] = 1UL << (x)
|
|
|
|
#define MASK_DECLARE_2(x) MASK_DECLARE_1(x), MASK_DECLARE_1(x+1)
|
|
|
|
#define MASK_DECLARE_4(x) MASK_DECLARE_2(x), MASK_DECLARE_2(x+2)
|
|
|
|
#define MASK_DECLARE_8(x) MASK_DECLARE_4(x), MASK_DECLARE_4(x+4)
|
2008-07-25 01:21:29 +00:00
|
|
|
|
cpu masks: optimize and clean up cpumask_of_cpu()
Clean up and optimize cpumask_of_cpu(), by sharing all the zero words.
Instead of stupidly generating all possible i=0...NR_CPUS 2^i patterns
creating a huge array of constant bitmasks, realize that the zero words
can be shared.
In other words, on a 64-bit architecture, we only ever need 64 of these
arrays - with a different bit set in one single world (with enough zero
words around it so that we can create any bitmask by just offsetting in
that big array). And then we just put enough zeroes around it that we
can point every single cpumask to be one of those things.
So when we have 4k CPU's, instead of having 4k arrays (of 4k bits each,
with one bit set in each array - 2MB memory total), we have exactly 64
arrays instead, each 8k bits in size (64kB total).
And then we just point cpumask(n) to the right position (which we can
calculate dynamically). Once we have the right arrays, getting
"cpumask(n)" ends up being:
static inline const cpumask_t *get_cpu_mask(unsigned int cpu)
{
const unsigned long *p = cpu_bit_bitmap[1 + cpu % BITS_PER_LONG];
p -= cpu / BITS_PER_LONG;
return (const cpumask_t *)p;
}
This brings other advantages and simplifications as well:
- we are not wasting memory that is just filled with a single bit in
various different places
- we don't need all those games to re-create the arrays in some dense
format, because they're already going to be dense enough.
if we compile a kernel for up to 4k CPU's, "wasting" that 64kB of memory
is a non-issue (especially since by doing this "overlapping" trick we
probably get better cache behaviour anyway).
[ mingo@elte.hu:
Converted Linus's mails into a commit. See:
http://lkml.org/lkml/2008/7/27/156
http://lkml.org/lkml/2008/7/28/320
Also applied a family filter - which also has the side-effect of leaving
out the bits where Linus calls me an idio... Oh, never mind ;-)
]
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Cc: Rusty Russell <rusty@rustcorp.com.au>
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: Al Viro <viro@ZenIV.linux.org.uk>
Cc: Mike Travis <travis@sgi.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
2008-07-28 18:32:33 +00:00
|
|
|
const unsigned long cpu_bit_bitmap[BITS_PER_LONG+1][BITS_TO_LONGS(NR_CPUS)] = {
|
|
|
|
|
|
|
|
MASK_DECLARE_8(0), MASK_DECLARE_8(8),
|
|
|
|
MASK_DECLARE_8(16), MASK_DECLARE_8(24),
|
|
|
|
#if BITS_PER_LONG > 32
|
|
|
|
MASK_DECLARE_8(32), MASK_DECLARE_8(40),
|
|
|
|
MASK_DECLARE_8(48), MASK_DECLARE_8(56),
|
2008-07-25 01:21:29 +00:00
|
|
|
#endif
|
|
|
|
};
|
cpu masks: optimize and clean up cpumask_of_cpu()
Clean up and optimize cpumask_of_cpu(), by sharing all the zero words.
Instead of stupidly generating all possible i=0...NR_CPUS 2^i patterns
creating a huge array of constant bitmasks, realize that the zero words
can be shared.
In other words, on a 64-bit architecture, we only ever need 64 of these
arrays - with a different bit set in one single world (with enough zero
words around it so that we can create any bitmask by just offsetting in
that big array). And then we just put enough zeroes around it that we
can point every single cpumask to be one of those things.
So when we have 4k CPU's, instead of having 4k arrays (of 4k bits each,
with one bit set in each array - 2MB memory total), we have exactly 64
arrays instead, each 8k bits in size (64kB total).
And then we just point cpumask(n) to the right position (which we can
calculate dynamically). Once we have the right arrays, getting
"cpumask(n)" ends up being:
static inline const cpumask_t *get_cpu_mask(unsigned int cpu)
{
const unsigned long *p = cpu_bit_bitmap[1 + cpu % BITS_PER_LONG];
p -= cpu / BITS_PER_LONG;
return (const cpumask_t *)p;
}
This brings other advantages and simplifications as well:
- we are not wasting memory that is just filled with a single bit in
various different places
- we don't need all those games to re-create the arrays in some dense
format, because they're already going to be dense enough.
if we compile a kernel for up to 4k CPU's, "wasting" that 64kB of memory
is a non-issue (especially since by doing this "overlapping" trick we
probably get better cache behaviour anyway).
[ mingo@elte.hu:
Converted Linus's mails into a commit. See:
http://lkml.org/lkml/2008/7/27/156
http://lkml.org/lkml/2008/7/28/320
Also applied a family filter - which also has the side-effect of leaving
out the bits where Linus calls me an idio... Oh, never mind ;-)
]
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Cc: Rusty Russell <rusty@rustcorp.com.au>
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: Al Viro <viro@ZenIV.linux.org.uk>
Cc: Mike Travis <travis@sgi.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
2008-07-28 18:32:33 +00:00
|
|
|
EXPORT_SYMBOL_GPL(cpu_bit_bitmap);
|
2008-11-05 02:39:10 +00:00
|
|
|
|
|
|
|
const DECLARE_BITMAP(cpu_all_bits, NR_CPUS) = CPU_BITS_ALL;
|
|
|
|
EXPORT_SYMBOL(cpu_all_bits);
|
2008-12-29 22:35:14 +00:00
|
|
|
|
|
|
|
#ifdef CONFIG_INIT_ALL_POSSIBLE
|
|
|
|
static DECLARE_BITMAP(cpu_possible_bits, CONFIG_NR_CPUS) __read_mostly
|
|
|
|
= CPU_BITS_ALL;
|
|
|
|
#else
|
|
|
|
static DECLARE_BITMAP(cpu_possible_bits, CONFIG_NR_CPUS) __read_mostly;
|
|
|
|
#endif
|
|
|
|
const struct cpumask *const cpu_possible_mask = to_cpumask(cpu_possible_bits);
|
|
|
|
EXPORT_SYMBOL(cpu_possible_mask);
|
|
|
|
|
|
|
|
static DECLARE_BITMAP(cpu_online_bits, CONFIG_NR_CPUS) __read_mostly;
|
|
|
|
const struct cpumask *const cpu_online_mask = to_cpumask(cpu_online_bits);
|
|
|
|
EXPORT_SYMBOL(cpu_online_mask);
|
|
|
|
|
|
|
|
static DECLARE_BITMAP(cpu_present_bits, CONFIG_NR_CPUS) __read_mostly;
|
|
|
|
const struct cpumask *const cpu_present_mask = to_cpumask(cpu_present_bits);
|
|
|
|
EXPORT_SYMBOL(cpu_present_mask);
|
|
|
|
|
|
|
|
static DECLARE_BITMAP(cpu_active_bits, CONFIG_NR_CPUS) __read_mostly;
|
|
|
|
const struct cpumask *const cpu_active_mask = to_cpumask(cpu_active_bits);
|
|
|
|
EXPORT_SYMBOL(cpu_active_mask);
|
2008-12-29 22:35:16 +00:00
|
|
|
|
|
|
|
void set_cpu_possible(unsigned int cpu, bool possible)
|
|
|
|
{
|
|
|
|
if (possible)
|
|
|
|
cpumask_set_cpu(cpu, to_cpumask(cpu_possible_bits));
|
|
|
|
else
|
|
|
|
cpumask_clear_cpu(cpu, to_cpumask(cpu_possible_bits));
|
|
|
|
}
|
|
|
|
|
|
|
|
void set_cpu_present(unsigned int cpu, bool present)
|
|
|
|
{
|
|
|
|
if (present)
|
|
|
|
cpumask_set_cpu(cpu, to_cpumask(cpu_present_bits));
|
|
|
|
else
|
|
|
|
cpumask_clear_cpu(cpu, to_cpumask(cpu_present_bits));
|
|
|
|
}
|
|
|
|
|
|
|
|
void set_cpu_online(unsigned int cpu, bool online)
|
|
|
|
{
|
|
|
|
if (online)
|
|
|
|
cpumask_set_cpu(cpu, to_cpumask(cpu_online_bits));
|
|
|
|
else
|
|
|
|
cpumask_clear_cpu(cpu, to_cpumask(cpu_online_bits));
|
|
|
|
}
|
|
|
|
|
|
|
|
void set_cpu_active(unsigned int cpu, bool active)
|
|
|
|
{
|
|
|
|
if (active)
|
|
|
|
cpumask_set_cpu(cpu, to_cpumask(cpu_active_bits));
|
|
|
|
else
|
|
|
|
cpumask_clear_cpu(cpu, to_cpumask(cpu_active_bits));
|
|
|
|
}
|
|
|
|
|
|
|
|
void init_cpu_present(const struct cpumask *src)
|
|
|
|
{
|
|
|
|
cpumask_copy(to_cpumask(cpu_present_bits), src);
|
|
|
|
}
|
|
|
|
|
|
|
|
void init_cpu_possible(const struct cpumask *src)
|
|
|
|
{
|
|
|
|
cpumask_copy(to_cpumask(cpu_possible_bits), src);
|
|
|
|
}
|
|
|
|
|
|
|
|
void init_cpu_online(const struct cpumask *src)
|
|
|
|
{
|
|
|
|
cpumask_copy(to_cpumask(cpu_online_bits), src);
|
|
|
|
}
|