blob: ba0aacde94fba2472d33ba9da5764a2b8cc5f63e [file] [log] [blame]
Ashok Rajc8094062006-01-08 01:03:17 -08001 CPU hotplug Support in Linux(tm) Kernel
2
3 Maintainers:
4 CPU Hotplug Core:
5 Rusty Russell <rusty@rustycorp.com.au>
6 Srivatsa Vaddagiri <vatsa@in.ibm.com>
7 i386:
8 Zwane Mwaikambo <zwane@arm.linux.org.uk>
9 ppc64:
10 Nathan Lynch <nathanl@austin.ibm.com>
11 Joel Schopp <jschopp@austin.ibm.com>
12 ia64/x86_64:
13 Ashok Raj <ashok.raj@intel.com>
Heiko Carstens255acee2006-02-17 13:52:46 -080014 s390:
15 Heiko Carstens <heiko.carstens@de.ibm.com>
Ashok Rajc8094062006-01-08 01:03:17 -080016
17Authors: Ashok Raj <ashok.raj@intel.com>
18Lots of feedback: Nathan Lynch <nathanl@austin.ibm.com>,
19 Joel Schopp <jschopp@austin.ibm.com>
20
21Introduction
22
23Modern advances in system architectures have introduced advanced error
24reporting and correction capabilities in processors. CPU architectures permit
25partitioning support, where compute resources of a single CPU could be made
26available to virtual machine environments. There are couple OEMS that
27support NUMA hardware which are hot pluggable as well, where physical
28node insertion and removal require support for CPU hotplug.
29
30Such advances require CPUs available to a kernel to be removed either for
31provisioning reasons, or for RAS purposes to keep an offending CPU off
32system execution path. Hence the need for CPU hotplug support in the
33Linux kernel.
34
35A more novel use of CPU-hotplug support is its use today in suspend
36resume support for SMP. Dual-core and HT support makes even
37a laptop run SMP kernels which didn't support these methods. SMP support
38for suspend/resume is a work in progress.
39
40General Stuff about CPU Hotplug
41--------------------------------
42
43Command Line Switches
44---------------------
45maxcpus=n Restrict boot time cpus to n. Say if you have 4 cpus, using
46 maxcpus=2 will only boot 2. You can choose to bring the
47 other cpus later online, read FAQ's for more info.
48
Satoru Takeuchica926e82006-10-19 23:29:06 -070049additional_cpus=n (*) Use this to limit hotpluggable cpus. This option sets
Heiko Carstens255acee2006-02-17 13:52:46 -080050 cpu_possible_map = cpu_present_map + additional_cpus
Ashok Raj8f8b11382006-02-16 14:01:48 -080051
Heiko Carstens6303dbf2006-02-20 18:27:58 -080052(*) Option valid only for following architectures
Heiko Carstens48483b32008-01-26 14:11:05 +010053- x86_64, ia64
Heiko Carstens6303dbf2006-02-20 18:27:58 -080054
Ashok Raj8f8b11382006-02-16 14:01:48 -080055ia64 and x86_64 use the number of disabled local apics in ACPI tables MADT
56to determine the number of potentially hot-pluggable cpus. The implementation
Matt LaPlante5d3f0832006-11-30 05:21:10 +010057should only rely on this to count the # of cpus, but *MUST* not rely on the
58apicid values in those tables for disabled apics. In the event BIOS doesn't
Ashok Raj8f8b11382006-02-16 14:01:48 -080059mark such hot-pluggable cpus as disabled entries, one could use this
60parameter "additional_cpus=x" to represent those cpus in the cpu_possible_map.
61
Heiko Carstens6303dbf2006-02-20 18:27:58 -080062s390 uses the number of cpus it detects at IPL time to also the number of bits
63in cpu_possible_map. If it is desired to add additional cpus at a later time
64the number should be specified using this option or the possible_cpus option.
Ashok Raj8f8b11382006-02-16 14:01:48 -080065
Heiko Carstens37a33022006-02-17 13:52:47 -080066possible_cpus=n [s390 only] use this to set hotpluggable cpus.
67 This option sets possible_cpus bits in
68 cpu_possible_map. Thus keeping the numbers of bits set
69 constant even if the machine gets rebooted.
70 This option overrides additional_cpus.
71
Ashok Rajc8094062006-01-08 01:03:17 -080072CPU maps and such
73-----------------
74[More on cpumaps and primitive to manipulate, please check
75include/linux/cpumask.h that has more descriptive text.]
76
77cpu_possible_map: Bitmap of possible CPUs that can ever be available in the
78system. This is used to allocate some boot time memory for per_cpu variables
79that aren't designed to grow/shrink as CPUs are made available or removed.
80Once set during boot time discovery phase, the map is static, i.e no bits
81are added or removed anytime. Trimming it accurately for your system needs
82upfront can save some boot time memory. See below for how we use heuristics
83in x86_64 case to keep this under check.
84
85cpu_online_map: Bitmap of all CPUs currently online. Its set in __cpu_up()
86after a cpu is available for kernel scheduling and ready to receive
87interrupts from devices. Its cleared when a cpu is brought down using
88__cpu_disable(), before which all OS services including interrupts are
89migrated to another target CPU.
90
91cpu_present_map: Bitmap of CPUs currently present in the system. Not all
92of them may be online. When physical hotplug is processed by the relevant
93subsystem (e.g ACPI) can change and new bit either be added or removed
94from the map depending on the event is hot-add/hot-remove. There are currently
95no locking rules as of now. Typical usage is to init topology during boot,
96at which time hotplug is disabled.
97
98You really dont need to manipulate any of the system cpu maps. They should
99be read-only for most use. When setting up per-cpu resources almost always use
KAMEZAWA Hiroyuki3c30a752006-03-28 01:56:39 -0800100cpu_possible_map/for_each_possible_cpu() to iterate.
Ashok Rajc8094062006-01-08 01:03:17 -0800101
102Never use anything other than cpumask_t to represent bitmap of CPUs.
103
Satoru Takeuchica926e82006-10-19 23:29:06 -0700104 #include <linux/cpumask.h>
Ashok Rajc8094062006-01-08 01:03:17 -0800105
Satoru Takeuchica926e82006-10-19 23:29:06 -0700106 for_each_possible_cpu - Iterate over cpu_possible_map
107 for_each_online_cpu - Iterate over cpu_online_map
108 for_each_present_cpu - Iterate over cpu_present_map
109 for_each_cpu_mask(x,mask) - Iterate over some random collection of cpu mask.
Ashok Rajc8094062006-01-08 01:03:17 -0800110
Satoru Takeuchica926e82006-10-19 23:29:06 -0700111 #include <linux/cpu.h>
Gautham R Shenoy86ef5c92008-01-25 21:08:02 +0100112 get_online_cpus() and put_online_cpus():
Ashok Rajc8094062006-01-08 01:03:17 -0800113
Gautham R Shenoy86ef5c92008-01-25 21:08:02 +0100114The above calls are used to inhibit cpu hotplug operations. While the
115cpu_hotplug.refcount is non zero, the cpu_online_map will not change.
116If you merely need to avoid cpus going away, you could also use
117preempt_disable() and preempt_enable() for those sections.
118Just remember the critical section cannot call any
Ashok Rajc8094062006-01-08 01:03:17 -0800119function that can sleep or schedule this process away. The preempt_disable()
120will work as long as stop_machine_run() is used to take a cpu down.
121
122CPU Hotplug - Frequently Asked Questions.
123
Satoru Takeuchica926e82006-10-19 23:29:06 -0700124Q: How to enable my kernel to support CPU hotplug?
Ashok Rajc8094062006-01-08 01:03:17 -0800125A: When doing make defconfig, Enable CPU hotplug support
126
127 "Processor type and Features" -> Support for Hotpluggable CPUs
128
129Make sure that you have CONFIG_HOTPLUG, and CONFIG_SMP turned on as well.
130
131You would need to enable CONFIG_HOTPLUG_CPU for SMP suspend/resume support
132as well.
133
134Q: What architectures support CPU hotplug?
135A: As of 2.6.14, the following architectures support CPU hotplug.
136
137i386 (Intel), ppc, ppc64, parisc, s390, ia64 and x86_64
138
139Q: How to test if hotplug is supported on the newly built kernel?
140A: You should now notice an entry in sysfs.
141
142Check if sysfs is mounted, using the "mount" command. You should notice
143an entry as shown below in the output.
144
Satoru Takeuchica926e82006-10-19 23:29:06 -0700145 ....
146 none on /sys type sysfs (rw)
147 ....
Ashok Rajc8094062006-01-08 01:03:17 -0800148
Satoru Takeuchica926e82006-10-19 23:29:06 -0700149If this is not mounted, do the following.
Ashok Rajc8094062006-01-08 01:03:17 -0800150
Satoru Takeuchica926e82006-10-19 23:29:06 -0700151 #mkdir /sysfs
152 #mount -t sysfs sys /sys
Ashok Rajc8094062006-01-08 01:03:17 -0800153
Satoru Takeuchica926e82006-10-19 23:29:06 -0700154Now you should see entries for all present cpu, the following is an example
Ashok Rajc8094062006-01-08 01:03:17 -0800155in a 8-way system.
156
Satoru Takeuchica926e82006-10-19 23:29:06 -0700157 #pwd
158 #/sys/devices/system/cpu
159 #ls -l
160 total 0
161 drwxr-xr-x 10 root root 0 Sep 19 07:44 .
162 drwxr-xr-x 13 root root 0 Sep 19 07:45 ..
163 drwxr-xr-x 3 root root 0 Sep 19 07:44 cpu0
164 drwxr-xr-x 3 root root 0 Sep 19 07:44 cpu1
165 drwxr-xr-x 3 root root 0 Sep 19 07:44 cpu2
166 drwxr-xr-x 3 root root 0 Sep 19 07:44 cpu3
167 drwxr-xr-x 3 root root 0 Sep 19 07:44 cpu4
168 drwxr-xr-x 3 root root 0 Sep 19 07:44 cpu5
169 drwxr-xr-x 3 root root 0 Sep 19 07:44 cpu6
170 drwxr-xr-x 3 root root 0 Sep 19 07:48 cpu7
Ashok Rajc8094062006-01-08 01:03:17 -0800171
172Under each directory you would find an "online" file which is the control
173file to logically online/offline a processor.
174
175Q: Does hot-add/hot-remove refer to physical add/remove of cpus?
176A: The usage of hot-add/remove may not be very consistently used in the code.
Satoru Takeuchica926e82006-10-19 23:29:06 -0700177CONFIG_HOTPLUG_CPU enables logical online/offline capability in the kernel.
Ashok Rajc8094062006-01-08 01:03:17 -0800178To support physical addition/removal, one would need some BIOS hooks and
179the platform should have something like an attention button in PCI hotplug.
180CONFIG_ACPI_HOTPLUG_CPU enables ACPI support for physical add/remove of CPUs.
181
182Q: How do i logically offline a CPU?
183A: Do the following.
184
Satoru Takeuchica926e82006-10-19 23:29:06 -0700185 #echo 0 > /sys/devices/system/cpu/cpuX/online
Ashok Rajc8094062006-01-08 01:03:17 -0800186
Satoru Takeuchica926e82006-10-19 23:29:06 -0700187Once the logical offline is successful, check
Ashok Rajc8094062006-01-08 01:03:17 -0800188
Satoru Takeuchica926e82006-10-19 23:29:06 -0700189 #cat /proc/interrupts
Ashok Rajc8094062006-01-08 01:03:17 -0800190
Satoru Takeuchica926e82006-10-19 23:29:06 -0700191You should now not see the CPU that you removed. Also online file will report
Ashok Rajc8094062006-01-08 01:03:17 -0800192the state as 0 when a cpu if offline and 1 when its online.
193
Satoru Takeuchica926e82006-10-19 23:29:06 -0700194 #To display the current cpu state.
195 #cat /sys/devices/system/cpu/cpuX/online
Ashok Rajc8094062006-01-08 01:03:17 -0800196
197Q: Why cant i remove CPU0 on some systems?
198A: Some architectures may have some special dependency on a certain CPU.
199
200For e.g in IA64 platforms we have ability to sent platform interrupts to the
201OS. a.k.a Corrected Platform Error Interrupts (CPEI). In current ACPI
202specifications, we didn't have a way to change the target CPU. Hence if the
203current ACPI version doesn't support such re-direction, we disable that CPU
204by making it not-removable.
205
206In such cases you will also notice that the online file is missing under cpu0.
207
208Q: How do i find out if a particular CPU is not removable?
209A: Depending on the implementation, some architectures may show this by the
210absence of the "online" file. This is done if it can be determined ahead of
211time that this CPU cannot be removed.
212
213In some situations, this can be a run time check, i.e if you try to remove the
214last CPU, this will not be permitted. You can find such failures by
215investigating the return value of the "echo" command.
216
217Q: What happens when a CPU is being logically offlined?
218A: The following happen, listed in no particular order :-)
219
220- A notification is sent to in-kernel registered modules by sending an event
Rafael J. Wysocki8bb78442007-05-09 02:35:10 -0700221 CPU_DOWN_PREPARE or CPU_DOWN_PREPARE_FROZEN, depending on whether or not the
222 CPU is being offlined while tasks are frozen due to a suspend operation in
223 progress
Cliff Wickman470fd642007-10-18 23:40:46 -0700224- All processes are migrated away from this outgoing CPU to new CPUs.
225 The new CPU is chosen from each process' current cpuset, which may be
226 a subset of all online CPUs.
Ashok Rajc8094062006-01-08 01:03:17 -0800227- All interrupts targeted to this CPU is migrated to a new CPU
228- timers/bottom half/task lets are also migrated to a new CPU
229- Once all services are migrated, kernel calls an arch specific routine
230 __cpu_disable() to perform arch specific cleanup.
231- Once this is successful, an event for successful cleanup is sent by an event
Rafael J. Wysocki8bb78442007-05-09 02:35:10 -0700232 CPU_DEAD (or CPU_DEAD_FROZEN if tasks are frozen due to a suspend while the
233 CPU is being offlined).
Ashok Rajc8094062006-01-08 01:03:17 -0800234
235 "It is expected that each service cleans up when the CPU_DOWN_PREPARE
236 notifier is called, when CPU_DEAD is called its expected there is nothing
237 running on behalf of this CPU that was offlined"
238
239Q: If i have some kernel code that needs to be aware of CPU arrival and
240 departure, how to i arrange for proper notification?
241A: This is what you would need in your kernel code to receive notifications.
242
Satoru Takeuchica926e82006-10-19 23:29:06 -0700243 #include <linux/cpu.h>
244 static int __cpuinit foobar_cpu_callback(struct notifier_block *nfb,
Ashok Rajc8094062006-01-08 01:03:17 -0800245 unsigned long action, void *hcpu)
246 {
247 unsigned int cpu = (unsigned long)hcpu;
248
249 switch (action) {
250 case CPU_ONLINE:
Rafael J. Wysocki8bb78442007-05-09 02:35:10 -0700251 case CPU_ONLINE_FROZEN:
Ashok Rajc8094062006-01-08 01:03:17 -0800252 foobar_online_action(cpu);
253 break;
254 case CPU_DEAD:
Rafael J. Wysocki8bb78442007-05-09 02:35:10 -0700255 case CPU_DEAD_FROZEN:
Ashok Rajc8094062006-01-08 01:03:17 -0800256 foobar_dead_action(cpu);
257 break;
258 }
259 return NOTIFY_OK;
260 }
261
Chandra Seetharaman7c7165c2006-07-30 03:03:36 -0700262 static struct notifier_block __cpuinitdata foobar_cpu_notifer =
Ashok Rajc8094062006-01-08 01:03:17 -0800263 {
264 .notifier_call = foobar_cpu_callback,
265 };
266
Chandra Seetharaman7c7165c2006-07-30 03:03:36 -0700267You need to call register_cpu_notifier() from your init function.
268Init functions could be of two types:
2691. early init (init function called when only the boot processor is online).
2702. late init (init function called _after_ all the CPUs are online).
Ashok Rajc8094062006-01-08 01:03:17 -0800271
Chandra Seetharaman7c7165c2006-07-30 03:03:36 -0700272For the first case, you should add the following to your init function
Ashok Rajc8094062006-01-08 01:03:17 -0800273
274 register_cpu_notifier(&foobar_cpu_notifier);
275
Chandra Seetharaman7c7165c2006-07-30 03:03:36 -0700276For the second case, you should add the following to your init function
277
278 register_hotcpu_notifier(&foobar_cpu_notifier);
279
Ashok Rajc8094062006-01-08 01:03:17 -0800280You can fail PREPARE notifiers if something doesn't work to prepare resources.
281This will stop the activity and send a following CANCELED event back.
282
283CPU_DEAD should not be failed, its just a goodness indication, but bad
284things will happen if a notifier in path sent a BAD notify code.
285
286Q: I don't see my action being called for all CPUs already up and running?
287A: Yes, CPU notifiers are called only when new CPUs are on-lined or offlined.
288 If you need to perform some action for each cpu already in the system, then
289
Satoru Takeuchica926e82006-10-19 23:29:06 -0700290 for_each_online_cpu(i) {
Ashok Rajc8094062006-01-08 01:03:17 -0800291 foobar_cpu_callback(&foobar_cpu_notifier, CPU_UP_PREPARE, i);
Satoru Takeuchica926e82006-10-19 23:29:06 -0700292 foobar_cpu_callback(&foobar_cpu_notifier, CPU_ONLINE, i);
293 }
Ashok Rajc8094062006-01-08 01:03:17 -0800294
295Q: If i would like to develop cpu hotplug support for a new architecture,
296 what do i need at a minimum?
297A: The following are what is required for CPU hotplug infrastructure to work
298 correctly.
299
300 - Make sure you have an entry in Kconfig to enable CONFIG_HOTPLUG_CPU
301 - __cpu_up() - Arch interface to bring up a CPU
302 - __cpu_disable() - Arch interface to shutdown a CPU, no more interrupts
303 can be handled by the kernel after the routine
304 returns. Including local APIC timers etc are
305 shutdown.
306 - __cpu_die() - This actually supposed to ensure death of the CPU.
307 Actually look at some example code in other arch
308 that implement CPU hotplug. The processor is taken
309 down from the idle() loop for that specific
310 architecture. __cpu_die() typically waits for some
311 per_cpu state to be set, to ensure the processor
312 dead routine is called to be sure positively.
313
314Q: I need to ensure that a particular cpu is not removed when there is some
315 work specific to this cpu is in progress.
316A: First switch the current thread context to preferred cpu
317
Satoru Takeuchica926e82006-10-19 23:29:06 -0700318 int my_func_on_cpu(int cpu)
319 {
320 cpumask_t saved_mask, new_mask = CPU_MASK_NONE;
321 int curr_cpu, err = 0;
Ashok Rajc8094062006-01-08 01:03:17 -0800322
Satoru Takeuchica926e82006-10-19 23:29:06 -0700323 saved_mask = current->cpus_allowed;
324 cpu_set(cpu, new_mask);
325 err = set_cpus_allowed(current, new_mask);
Ashok Rajc8094062006-01-08 01:03:17 -0800326
Satoru Takeuchica926e82006-10-19 23:29:06 -0700327 if (err)
328 return err;
Ashok Rajc8094062006-01-08 01:03:17 -0800329
Satoru Takeuchica926e82006-10-19 23:29:06 -0700330 /*
331 * If we got scheduled out just after the return from
332 * set_cpus_allowed() before running the work, this ensures
333 * we stay locked.
334 */
335 curr_cpu = get_cpu();
Ashok Rajc8094062006-01-08 01:03:17 -0800336
Satoru Takeuchica926e82006-10-19 23:29:06 -0700337 if (curr_cpu != cpu) {
338 err = -EAGAIN;
339 goto ret;
340 } else {
341 /*
342 * Do work : But cant sleep, since get_cpu() disables preempt
343 */
344 }
345 ret:
346 put_cpu();
347 set_cpus_allowed(current, saved_mask);
348 return err;
349 }
Ashok Rajc8094062006-01-08 01:03:17 -0800350
351
352Q: How do we determine how many CPUs are available for hotplug.
353A: There is no clear spec defined way from ACPI that can give us that
354 information today. Based on some input from Natalie of Unisys,
355 that the ACPI MADT (Multiple APIC Description Tables) marks those possible
356 CPUs in a system with disabled status.
357
358 Andi implemented some simple heuristics that count the number of disabled
359 CPUs in MADT as hotpluggable CPUS. In the case there are no disabled CPUS
360 we assume 1/2 the number of CPUs currently present can be hotplugged.
361
362 Caveat: Today's ACPI MADT can only provide 256 entries since the apicid field
363 in MADT is only 8 bits.
364
365User Space Notification
366
367Hotplug support for devices is common in Linux today. Its being used today to
368support automatic configuration of network, usb and pci devices. A hotplug
369event can be used to invoke an agent script to perform the configuration task.
370
371You can add /etc/hotplug/cpu.agent to handle hotplug notification user space
372scripts.
373
374 #!/bin/bash
375 # $Id: cpu.agent
376 # Kernel hotplug params include:
377 #ACTION=%s [online or offline]
378 #DEVPATH=%s
379 #
380 cd /etc/hotplug
381 . ./hotplug.functions
382
383 case $ACTION in
384 online)
385 echo `date` ":cpu.agent" add cpu >> /tmp/hotplug.txt
386 ;;
387 offline)
388 echo `date` ":cpu.agent" remove cpu >>/tmp/hotplug.txt
389 ;;
390 *)
391 debug_mesg CPU $ACTION event not supported
392 exit 1
393 ;;
394 esac