linux/arch/s390/kernel/topology.c

481 lines
10 KiB
C
Raw Normal View History

/*
* Copyright IBM Corp. 2007, 2011
* Author(s): Heiko Carstens <heiko.carstens@de.ibm.com>
*/
#define KMSG_COMPONENT "cpu"
#define pr_fmt(fmt) KMSG_COMPONENT ": " fmt
#include <linux/workqueue.h>
#include <linux/bootmem.h>
#include <linux/cpuset.h>
#include <linux/device.h>
#include <linux/export.h>
#include <linux/kernel.h>
#include <linux/sched.h>
#include <linux/init.h>
#include <linux/delay.h>
#include <linux/cpu.h>
#include <linux/smp.h>
#include <linux/mm.h>
#include <asm/sysinfo.h>
[S390] Vertical cpu management. If vertical cpu polarization is active then the hypervisor will dispatch certain cpus for a longer time than other cpus for maximum performance. For example if a guest would have three virtual cpus, each of them with a share of 33 percent, then in case of vertical cpu polarization all of the processing time would be combined to a single cpu which would run all the time, while the other two cpus would get nearly no cpu time. There are three different types of vertical cpus: high, medium and low. Low cpus hardly get any real cpu time, while high cpus get a full real cpu. Medium cpus get something in between. In order to switch between the two possible modes (default is horizontal) a 0 for horizontal polarization or a 1 for vertical polarization must be written to the dispatching sysfs attribute: /sys/devices/system/cpu/dispatching The polarization of each single cpu can be figured out by the polarization sysfs attribute of each cpu: /sys/devices/system/cpu/cpuX/polarization horizontal, vertical:high, vertical:medium, vertical:low or unknown. When switching polarization the polarization attribute may contain the value unknown until the configuration change is done and the kernel has figured out the new polarization of each cpu. Note that running a system with different types of vertical cpus may result in significant performance regressions. If possible only one type of vertical cpus should be used. All other cpus should be offlined. Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com> Signed-off-by: Heiko Carstens <heiko.carstens@de.ibm.com>
2008-04-17 13:46:13 +08:00
#define PTF_HORIZONTAL (0UL)
#define PTF_VERTICAL (1UL)
#define PTF_CHECK (2UL)
struct mask_info {
struct mask_info *next;
unsigned char id;
cpumask_t mask;
};
static void set_topology_timer(void);
static void topology_work_fn(struct work_struct *work);
static struct sysinfo_15_1_x *tl_info;
static int topology_enabled = 1;
static DECLARE_WORK(topology_work, topology_work_fn);
/* topology_lock protects the socket and book linked lists */
static DEFINE_SPINLOCK(topology_lock);
static struct mask_info socket_info;
static struct mask_info book_info;
struct cpu_topology_s390 cpu_topology[NR_CPUS];
EXPORT_SYMBOL_GPL(cpu_topology);
static cpumask_t cpu_group_map(struct mask_info *info, unsigned int cpu)
{
cpumask_t mask;
cpumask_copy(&mask, cpumask_of(cpu));
if (!topology_enabled || !MACHINE_HAS_TOPOLOGY)
return mask;
for (; info; info = info->next) {
if (cpumask_test_cpu(cpu, &info->mask))
return info->mask;
}
return mask;
}
static struct mask_info *add_cpus_to_mask(struct topology_cpu *tl_cpu,
struct mask_info *book,
struct mask_info *socket,
int one_socket_per_cpu)
{
unsigned int cpu;
for_each_set_bit(cpu, &tl_cpu->mask[0], TOPOLOGY_CPU_BITS) {
unsigned int rcpu;
int lcpu;
rcpu = TOPOLOGY_CPU_BITS - 1 - cpu + tl_cpu->origin;
lcpu = smp_find_processor_id(rcpu);
if (lcpu < 0)
continue;
cpumask_set_cpu(lcpu, &book->mask);
cpu_topology[lcpu].book_id = book->id;
cpumask_set_cpu(lcpu, &socket->mask);
cpu_topology[lcpu].core_id = rcpu;
if (one_socket_per_cpu) {
cpu_topology[lcpu].socket_id = rcpu;
socket = socket->next;
} else {
cpu_topology[lcpu].socket_id = socket->id;
}
smp_cpu_set_polarization(lcpu, tl_cpu->pp);
}
return socket;
}
static void clear_masks(void)
{
struct mask_info *info;
info = &socket_info;
while (info) {
cpumask_clear(&info->mask);
info = info->next;
}
info = &book_info;
while (info) {
cpumask_clear(&info->mask);
info = info->next;
}
}
static union topology_entry *next_tle(union topology_entry *tle)
{
if (!tle->nl)
return (union topology_entry *)((struct topology_cpu *)tle + 1);
return (union topology_entry *)((struct topology_container *)tle + 1);
}
static void __tl_to_masks_generic(struct sysinfo_15_1_x *info)
{
struct mask_info *socket = &socket_info;
struct mask_info *book = &book_info;
union topology_entry *tle, *end;
[S390] Vertical cpu management. If vertical cpu polarization is active then the hypervisor will dispatch certain cpus for a longer time than other cpus for maximum performance. For example if a guest would have three virtual cpus, each of them with a share of 33 percent, then in case of vertical cpu polarization all of the processing time would be combined to a single cpu which would run all the time, while the other two cpus would get nearly no cpu time. There are three different types of vertical cpus: high, medium and low. Low cpus hardly get any real cpu time, while high cpus get a full real cpu. Medium cpus get something in between. In order to switch between the two possible modes (default is horizontal) a 0 for horizontal polarization or a 1 for vertical polarization must be written to the dispatching sysfs attribute: /sys/devices/system/cpu/dispatching The polarization of each single cpu can be figured out by the polarization sysfs attribute of each cpu: /sys/devices/system/cpu/cpuX/polarization horizontal, vertical:high, vertical:medium, vertical:low or unknown. When switching polarization the polarization attribute may contain the value unknown until the configuration change is done and the kernel has figured out the new polarization of each cpu. Note that running a system with different types of vertical cpus may result in significant performance regressions. If possible only one type of vertical cpus should be used. All other cpus should be offlined. Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com> Signed-off-by: Heiko Carstens <heiko.carstens@de.ibm.com>
2008-04-17 13:46:13 +08:00
tle = info->tle;
end = (union topology_entry *)((unsigned long)info + info->length);
while (tle < end) {
switch (tle->nl) {
case 2:
book = book->next;
book->id = tle->container.id;
break;
case 1:
socket = socket->next;
socket->id = tle->container.id;
break;
case 0:
add_cpus_to_mask(&tle->cpu, book, socket, 0);
break;
default:
clear_masks();
return;
}
tle = next_tle(tle);
}
}
static void __tl_to_masks_z10(struct sysinfo_15_1_x *info)
{
struct mask_info *socket = &socket_info;
struct mask_info *book = &book_info;
union topology_entry *tle, *end;
tle = info->tle;
end = (union topology_entry *)((unsigned long)info + info->length);
while (tle < end) {
switch (tle->nl) {
case 1:
book = book->next;
book->id = tle->container.id;
break;
case 0:
socket = add_cpus_to_mask(&tle->cpu, book, socket, 1);
break;
default:
clear_masks();
return;
}
tle = next_tle(tle);
}
}
static void tl_to_masks(struct sysinfo_15_1_x *info)
{
struct cpuid cpu_id;
spin_lock_irq(&topology_lock);
get_cpu_id(&cpu_id);
clear_masks();
switch (cpu_id.machine) {
case 0x2097:
case 0x2098:
__tl_to_masks_z10(info);
break;
default:
__tl_to_masks_generic(info);
}
spin_unlock_irq(&topology_lock);
}
[S390] Vertical cpu management. If vertical cpu polarization is active then the hypervisor will dispatch certain cpus for a longer time than other cpus for maximum performance. For example if a guest would have three virtual cpus, each of them with a share of 33 percent, then in case of vertical cpu polarization all of the processing time would be combined to a single cpu which would run all the time, while the other two cpus would get nearly no cpu time. There are three different types of vertical cpus: high, medium and low. Low cpus hardly get any real cpu time, while high cpus get a full real cpu. Medium cpus get something in between. In order to switch between the two possible modes (default is horizontal) a 0 for horizontal polarization or a 1 for vertical polarization must be written to the dispatching sysfs attribute: /sys/devices/system/cpu/dispatching The polarization of each single cpu can be figured out by the polarization sysfs attribute of each cpu: /sys/devices/system/cpu/cpuX/polarization horizontal, vertical:high, vertical:medium, vertical:low or unknown. When switching polarization the polarization attribute may contain the value unknown until the configuration change is done and the kernel has figured out the new polarization of each cpu. Note that running a system with different types of vertical cpus may result in significant performance regressions. If possible only one type of vertical cpus should be used. All other cpus should be offlined. Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com> Signed-off-by: Heiko Carstens <heiko.carstens@de.ibm.com>
2008-04-17 13:46:13 +08:00
static void topology_update_polarization_simple(void)
{
int cpu;
mutex_lock(&smp_cpu_state_mutex);
for_each_possible_cpu(cpu)
smp_cpu_set_polarization(cpu, POLARIZATION_HRZ);
[S390] Vertical cpu management. If vertical cpu polarization is active then the hypervisor will dispatch certain cpus for a longer time than other cpus for maximum performance. For example if a guest would have three virtual cpus, each of them with a share of 33 percent, then in case of vertical cpu polarization all of the processing time would be combined to a single cpu which would run all the time, while the other two cpus would get nearly no cpu time. There are three different types of vertical cpus: high, medium and low. Low cpus hardly get any real cpu time, while high cpus get a full real cpu. Medium cpus get something in between. In order to switch between the two possible modes (default is horizontal) a 0 for horizontal polarization or a 1 for vertical polarization must be written to the dispatching sysfs attribute: /sys/devices/system/cpu/dispatching The polarization of each single cpu can be figured out by the polarization sysfs attribute of each cpu: /sys/devices/system/cpu/cpuX/polarization horizontal, vertical:high, vertical:medium, vertical:low or unknown. When switching polarization the polarization attribute may contain the value unknown until the configuration change is done and the kernel has figured out the new polarization of each cpu. Note that running a system with different types of vertical cpus may result in significant performance regressions. If possible only one type of vertical cpus should be used. All other cpus should be offlined. Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com> Signed-off-by: Heiko Carstens <heiko.carstens@de.ibm.com>
2008-04-17 13:46:13 +08:00
mutex_unlock(&smp_cpu_state_mutex);
}
static int ptf(unsigned long fc)
{
int rc;
asm volatile(
" .insn rre,0xb9a20000,%1,%1\n"
" ipm %0\n"
" srl %0,28\n"
: "=d" (rc)
[S390] Vertical cpu management. If vertical cpu polarization is active then the hypervisor will dispatch certain cpus for a longer time than other cpus for maximum performance. For example if a guest would have three virtual cpus, each of them with a share of 33 percent, then in case of vertical cpu polarization all of the processing time would be combined to a single cpu which would run all the time, while the other two cpus would get nearly no cpu time. There are three different types of vertical cpus: high, medium and low. Low cpus hardly get any real cpu time, while high cpus get a full real cpu. Medium cpus get something in between. In order to switch between the two possible modes (default is horizontal) a 0 for horizontal polarization or a 1 for vertical polarization must be written to the dispatching sysfs attribute: /sys/devices/system/cpu/dispatching The polarization of each single cpu can be figured out by the polarization sysfs attribute of each cpu: /sys/devices/system/cpu/cpuX/polarization horizontal, vertical:high, vertical:medium, vertical:low or unknown. When switching polarization the polarization attribute may contain the value unknown until the configuration change is done and the kernel has figured out the new polarization of each cpu. Note that running a system with different types of vertical cpus may result in significant performance regressions. If possible only one type of vertical cpus should be used. All other cpus should be offlined. Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com> Signed-off-by: Heiko Carstens <heiko.carstens@de.ibm.com>
2008-04-17 13:46:13 +08:00
: "d" (fc) : "cc");
return rc;
}
int topology_set_cpu_management(int fc)
{
int cpu, rc;
[S390] Vertical cpu management. If vertical cpu polarization is active then the hypervisor will dispatch certain cpus for a longer time than other cpus for maximum performance. For example if a guest would have three virtual cpus, each of them with a share of 33 percent, then in case of vertical cpu polarization all of the processing time would be combined to a single cpu which would run all the time, while the other two cpus would get nearly no cpu time. There are three different types of vertical cpus: high, medium and low. Low cpus hardly get any real cpu time, while high cpus get a full real cpu. Medium cpus get something in between. In order to switch between the two possible modes (default is horizontal) a 0 for horizontal polarization or a 1 for vertical polarization must be written to the dispatching sysfs attribute: /sys/devices/system/cpu/dispatching The polarization of each single cpu can be figured out by the polarization sysfs attribute of each cpu: /sys/devices/system/cpu/cpuX/polarization horizontal, vertical:high, vertical:medium, vertical:low or unknown. When switching polarization the polarization attribute may contain the value unknown until the configuration change is done and the kernel has figured out the new polarization of each cpu. Note that running a system with different types of vertical cpus may result in significant performance regressions. If possible only one type of vertical cpus should be used. All other cpus should be offlined. Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com> Signed-off-by: Heiko Carstens <heiko.carstens@de.ibm.com>
2008-04-17 13:46:13 +08:00
if (!MACHINE_HAS_TOPOLOGY)
[S390] Vertical cpu management. If vertical cpu polarization is active then the hypervisor will dispatch certain cpus for a longer time than other cpus for maximum performance. For example if a guest would have three virtual cpus, each of them with a share of 33 percent, then in case of vertical cpu polarization all of the processing time would be combined to a single cpu which would run all the time, while the other two cpus would get nearly no cpu time. There are three different types of vertical cpus: high, medium and low. Low cpus hardly get any real cpu time, while high cpus get a full real cpu. Medium cpus get something in between. In order to switch between the two possible modes (default is horizontal) a 0 for horizontal polarization or a 1 for vertical polarization must be written to the dispatching sysfs attribute: /sys/devices/system/cpu/dispatching The polarization of each single cpu can be figured out by the polarization sysfs attribute of each cpu: /sys/devices/system/cpu/cpuX/polarization horizontal, vertical:high, vertical:medium, vertical:low or unknown. When switching polarization the polarization attribute may contain the value unknown until the configuration change is done and the kernel has figured out the new polarization of each cpu. Note that running a system with different types of vertical cpus may result in significant performance regressions. If possible only one type of vertical cpus should be used. All other cpus should be offlined. Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com> Signed-off-by: Heiko Carstens <heiko.carstens@de.ibm.com>
2008-04-17 13:46:13 +08:00
return -EOPNOTSUPP;
if (fc)
rc = ptf(PTF_VERTICAL);
else
rc = ptf(PTF_HORIZONTAL);
if (rc)
return -EBUSY;
for_each_possible_cpu(cpu)
smp_cpu_set_polarization(cpu, POLARIZATION_UNKNOWN);
return rc;
}
static void update_cpu_masks(void)
{
unsigned long flags;
int cpu;
spin_lock_irqsave(&topology_lock, flags);
for_each_possible_cpu(cpu) {
cpu_topology[cpu].core_mask = cpu_group_map(&socket_info, cpu);
cpu_topology[cpu].book_mask = cpu_group_map(&book_info, cpu);
if (!MACHINE_HAS_TOPOLOGY) {
cpu_topology[cpu].core_id = cpu;
cpu_topology[cpu].socket_id = cpu;
cpu_topology[cpu].book_id = cpu;
}
}
spin_unlock_irqrestore(&topology_lock, flags);
}
void store_topology(struct sysinfo_15_1_x *info)
{
if (topology_max_mnest >= 3)
stsi(info, 15, 1, 3);
else
stsi(info, 15, 1, 2);
}
int arch_update_cpu_topology(void)
{
struct sysinfo_15_1_x *info = tl_info;
cpu: convert 'cpu' and 'machinecheck' sysdev_class to a regular subsystem This moves the 'cpu sysdev_class' over to a regular 'cpu' subsystem and converts the devices to regular devices. The sysdev drivers are implemented as subsystem interfaces now. After all sysdev classes are ported to regular driver core entities, the sysdev implementation will be entirely removed from the kernel. Userspace relies on events and generic sysfs subsystem infrastructure from sysdev devices, which are made available with this conversion. Cc: Haavard Skinnemoen <hskinnemoen@gmail.com> Cc: Hans-Christian Egtvedt <egtvedt@samfundet.no> Cc: Tony Luck <tony.luck@intel.com> Cc: Fenghua Yu <fenghua.yu@intel.com> Cc: Arnd Bergmann <arnd@arndb.de> Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org> Cc: Paul Mackerras <paulus@samba.org> Cc: Martin Schwidefsky <schwidefsky@de.ibm.com> Cc: Heiko Carstens <heiko.carstens@de.ibm.com> Cc: Paul Mundt <lethal@linux-sh.org> Cc: "David S. Miller" <davem@davemloft.net> Cc: Chris Metcalf <cmetcalf@tilera.com> Cc: Thomas Gleixner <tglx@linutronix.de> Cc: Ingo Molnar <mingo@redhat.com> Cc: "H. Peter Anvin" <hpa@zytor.com> Cc: Borislav Petkov <bp@amd64.org> Cc: Tigran Aivazian <tigran@aivazian.fsnet.co.uk> Cc: Len Brown <lenb@kernel.org> Cc: Zhang Rui <rui.zhang@intel.com> Cc: Dave Jones <davej@redhat.com> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Russell King <rmk+kernel@arm.linux.org.uk> Cc: Andrew Morton <akpm@linux-foundation.org> Cc: Arjan van de Ven <arjan@linux.intel.com> Cc: "Rafael J. Wysocki" <rjw@sisk.pl> Cc: "Srivatsa S. Bhat" <srivatsa.bhat@linux.vnet.ibm.com> Signed-off-by: Kay Sievers <kay.sievers@vrfy.org> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
2011-12-22 06:29:42 +08:00
struct device *dev;
int cpu;
if (!MACHINE_HAS_TOPOLOGY) {
update_cpu_masks();
[S390] Vertical cpu management. If vertical cpu polarization is active then the hypervisor will dispatch certain cpus for a longer time than other cpus for maximum performance. For example if a guest would have three virtual cpus, each of them with a share of 33 percent, then in case of vertical cpu polarization all of the processing time would be combined to a single cpu which would run all the time, while the other two cpus would get nearly no cpu time. There are three different types of vertical cpus: high, medium and low. Low cpus hardly get any real cpu time, while high cpus get a full real cpu. Medium cpus get something in between. In order to switch between the two possible modes (default is horizontal) a 0 for horizontal polarization or a 1 for vertical polarization must be written to the dispatching sysfs attribute: /sys/devices/system/cpu/dispatching The polarization of each single cpu can be figured out by the polarization sysfs attribute of each cpu: /sys/devices/system/cpu/cpuX/polarization horizontal, vertical:high, vertical:medium, vertical:low or unknown. When switching polarization the polarization attribute may contain the value unknown until the configuration change is done and the kernel has figured out the new polarization of each cpu. Note that running a system with different types of vertical cpus may result in significant performance regressions. If possible only one type of vertical cpus should be used. All other cpus should be offlined. Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com> Signed-off-by: Heiko Carstens <heiko.carstens@de.ibm.com>
2008-04-17 13:46:13 +08:00
topology_update_polarization_simple();
return 0;
[S390] Vertical cpu management. If vertical cpu polarization is active then the hypervisor will dispatch certain cpus for a longer time than other cpus for maximum performance. For example if a guest would have three virtual cpus, each of them with a share of 33 percent, then in case of vertical cpu polarization all of the processing time would be combined to a single cpu which would run all the time, while the other two cpus would get nearly no cpu time. There are three different types of vertical cpus: high, medium and low. Low cpus hardly get any real cpu time, while high cpus get a full real cpu. Medium cpus get something in between. In order to switch between the two possible modes (default is horizontal) a 0 for horizontal polarization or a 1 for vertical polarization must be written to the dispatching sysfs attribute: /sys/devices/system/cpu/dispatching The polarization of each single cpu can be figured out by the polarization sysfs attribute of each cpu: /sys/devices/system/cpu/cpuX/polarization horizontal, vertical:high, vertical:medium, vertical:low or unknown. When switching polarization the polarization attribute may contain the value unknown until the configuration change is done and the kernel has figured out the new polarization of each cpu. Note that running a system with different types of vertical cpus may result in significant performance regressions. If possible only one type of vertical cpus should be used. All other cpus should be offlined. Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com> Signed-off-by: Heiko Carstens <heiko.carstens@de.ibm.com>
2008-04-17 13:46:13 +08:00
}
store_topology(info);
tl_to_masks(info);
update_cpu_masks();
for_each_online_cpu(cpu) {
cpu: convert 'cpu' and 'machinecheck' sysdev_class to a regular subsystem This moves the 'cpu sysdev_class' over to a regular 'cpu' subsystem and converts the devices to regular devices. The sysdev drivers are implemented as subsystem interfaces now. After all sysdev classes are ported to regular driver core entities, the sysdev implementation will be entirely removed from the kernel. Userspace relies on events and generic sysfs subsystem infrastructure from sysdev devices, which are made available with this conversion. Cc: Haavard Skinnemoen <hskinnemoen@gmail.com> Cc: Hans-Christian Egtvedt <egtvedt@samfundet.no> Cc: Tony Luck <tony.luck@intel.com> Cc: Fenghua Yu <fenghua.yu@intel.com> Cc: Arnd Bergmann <arnd@arndb.de> Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org> Cc: Paul Mackerras <paulus@samba.org> Cc: Martin Schwidefsky <schwidefsky@de.ibm.com> Cc: Heiko Carstens <heiko.carstens@de.ibm.com> Cc: Paul Mundt <lethal@linux-sh.org> Cc: "David S. Miller" <davem@davemloft.net> Cc: Chris Metcalf <cmetcalf@tilera.com> Cc: Thomas Gleixner <tglx@linutronix.de> Cc: Ingo Molnar <mingo@redhat.com> Cc: "H. Peter Anvin" <hpa@zytor.com> Cc: Borislav Petkov <bp@amd64.org> Cc: Tigran Aivazian <tigran@aivazian.fsnet.co.uk> Cc: Len Brown <lenb@kernel.org> Cc: Zhang Rui <rui.zhang@intel.com> Cc: Dave Jones <davej@redhat.com> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Russell King <rmk+kernel@arm.linux.org.uk> Cc: Andrew Morton <akpm@linux-foundation.org> Cc: Arjan van de Ven <arjan@linux.intel.com> Cc: "Rafael J. Wysocki" <rjw@sisk.pl> Cc: "Srivatsa S. Bhat" <srivatsa.bhat@linux.vnet.ibm.com> Signed-off-by: Kay Sievers <kay.sievers@vrfy.org> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
2011-12-22 06:29:42 +08:00
dev = get_cpu_device(cpu);
kobject_uevent(&dev->kobj, KOBJ_CHANGE);
}
return 1;
}
[S390] cpu topology: Fix possible deadlock. When we get a notification that cpu topology changed, we schedule a work struct which just calls arch_reinit_sched_domains. This function in turn calls get_online_cpus() which results int the lockdep warning below. After all it turnded out that it's not legal to call get_online_cpus() from the context of a multi-threaded work queue. It could deadlock this way: process 0 (events/cpu-x): -> run_workqueue -> removes my work_struct from the work queue -> calls work_struct->fn -> get_online_cpus() -> locks on cpu_hotplug.lock since process 1 below is doing cpu hotplug process 1: -> cpu_down (for cpu-x) -> cpu_hotplug_begin (holds cpu_hotplug.lock now) -> cpu-x dead -> notifier_call_chain with CPU_DEAD -> cleanup_workqueue_thread -> flush_cpu_workqueue (succeeds) -> kthread_stop for events/cpu-x -> now kthread_stop waits for my work_struct to complete from within process 0. -> dead. A single threaded workqueue wouldn't have such problems, however there is no such common queue available and it's not worth to create one for the very rare calls to arch_reinit_sched_domains. So we just create a kernel thread from our work struct which calls arch_reinit_sched_domains and are done with it. Thanks to Oleg Nesterov and Peter Zijlstra for helping me figuring out that this isn't a false positive lockdep warning: ======================================================= [ INFO: possible circular locking dependency detected ] 2.6.25-03562-g3dc5063-dirty #12 ------------------------------------------------------- events/3/14 is trying to acquire lock: (&cpu_hotplug.lock){--..}, at: [<0000000000076094>] get_online_cpus+0x50/0x78 but task is already holding lock: (topology_work){--..}, at: [<0000000000059cde>] run_workqueue+0x106/0x278 which lock already depends on the new lock. the existing dependency chain (in reverse order) is: -> #2 (topology_work){--..}: [<000000000006fc74>] __lock_acquire+0x1010/0x111c [<000000000006fe40>] lock_acquire+0xc0/0xf8 [<0000000000059d48>] run_workqueue+0x170/0x278 [<0000000000059edc>] worker_thread+0x8c/0xf0 [<000000000005f5bc>] kthread+0x68/0xa0 [<000000000001a33e>] kernel_thread_starter+0x6/0xc [<000000000001a338>] kernel_thread_starter+0x0/0xc -> #1 (events){--..}: [<000000000006fc74>] __lock_acquire+0x1010/0x111c [<000000000006fe40>] lock_acquire+0xc0/0xf8 [<000000000005a23c>] cleanup_workqueue_thread+0x60/0xa8 [<00000000003b2ab8>] workqueue_cpu_callback+0xbc/0x170 [<00000000003bba80>] notifier_call_chain+0x5c/0xa4 [<00000000000655a2>] __raw_notifier_call_chain+0x26/0x38 [<00000000000655e2>] raw_notifier_call_chain+0x2e/0x40 [<0000000000075e00>] cpu_down+0x228/0x31c [<00000000003b1dd8>] store_online+0x64/0xb8 [<00000000001e7128>] sysdev_store+0x48/0x58 [<0000000000121cd2>] sysfs_write_file+0x126/0x1c0 [<00000000000c1944>] vfs_write+0xb0/0x15c [<00000000000c20e6>] sys_write+0x56/0x88 [<0000000000027a68>] sys32_write+0x34/0x4c [<0000000000023f70>] sysc_noemu+0x10/0x16 [<0000000077f3f186>] 0x77f3f186 -> #0 (&cpu_hotplug.lock){--..}: [<000000000006fa84>] __lock_acquire+0xe20/0x111c [<000000000006fe40>] lock_acquire+0xc0/0xf8 [<00000000003b701c>] mutex_lock_nested+0xd0/0x364 [<0000000000076094>] get_online_cpus+0x50/0x78 [<000000000003a03e>] arch_reinit_sched_domains+0x26/0x58 [<000000000002700e>] topology_work_fn+0x26/0x34 [<0000000000059d4e>] run_workqueue+0x176/0x278 [<0000000000059edc>] worker_thread+0x8c/0xf0 [<000000000005f5bc>] kthread+0x68/0xa0 [<000000000001a33e>] kernel_thread_starter+0x6/0xc [<000000000001a338>] kernel_thread_starter+0x0/0xc other info that might help us debug this: 2 locks held by events/3/14: #0: (events){--..}, at: [<0000000000059cde>] run_workqueue+0x106/0x278 #1: (topology_work){--..}, at: [<0000000000059cde>] run_workqueue+0x106/0x278 stack backtrace: CPU: 3 Not tainted 2.6.25-03562-g3dc5063-dirty #12 Process events/3 (pid: 14, task: 000000002fb04038, ksp: 000000002fb0bd70) 0400000000000000 000000002fb0ba40 0000000000000002 0000000000000000 000000002fb0bae0 000000002fb0ba58 000000002fb0ba58 0000000000016488 0000000000000000 000000002fb0bd70 0000000000000000 0000000000000000 000000002fb0ba40 000000000000000c 000000002fb0ba40 000000002fb0bab0 00000000003c99e0 0000000000016488 000000002fb0ba40 000000002fb0ba90 Call Trace: ([<00000000000163fc>] show_trace+0x138/0x158) [<00000000000164e2>] show_stack+0xc6/0xf8 [<0000000000016624>] dump_stack+0xb0/0xc0 [<000000000006cd36>] print_circular_bug_tail+0xa2/0xb4 [<000000000006fa84>] __lock_acquire+0xe20/0x111c [<000000000006fe40>] lock_acquire+0xc0/0xf8 [<00000000003b701c>] mutex_lock_nested+0xd0/0x364 [<0000000000076094>] get_online_cpus+0x50/0x78 [<000000000003a03e>] arch_reinit_sched_domains+0x26/0x58 [<000000000002700e>] topology_work_fn+0x26/0x34 [<0000000000059d4e>] run_workqueue+0x176/0x278 [<0000000000059edc>] worker_thread+0x8c/0xf0 [<000000000005f5bc>] kthread+0x68/0xa0 [<000000000001a33e>] kernel_thread_starter+0x6/0xc [<000000000001a338>] kernel_thread_starter+0x0/0xc INFO: lockdep is turned off. Cc: Oleg Nesterov <oleg@tv-sign.ru> Cc: Peter Zijlstra <peterz@infradead.org> Signed-off-by: Heiko Carstens <heiko.carstens@de.ibm.com> Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
2008-04-30 19:38:41 +08:00
static void topology_work_fn(struct work_struct *work)
{
rebuild_sched_domains();
}
[S390] Vertical cpu management. If vertical cpu polarization is active then the hypervisor will dispatch certain cpus for a longer time than other cpus for maximum performance. For example if a guest would have three virtual cpus, each of them with a share of 33 percent, then in case of vertical cpu polarization all of the processing time would be combined to a single cpu which would run all the time, while the other two cpus would get nearly no cpu time. There are three different types of vertical cpus: high, medium and low. Low cpus hardly get any real cpu time, while high cpus get a full real cpu. Medium cpus get something in between. In order to switch between the two possible modes (default is horizontal) a 0 for horizontal polarization or a 1 for vertical polarization must be written to the dispatching sysfs attribute: /sys/devices/system/cpu/dispatching The polarization of each single cpu can be figured out by the polarization sysfs attribute of each cpu: /sys/devices/system/cpu/cpuX/polarization horizontal, vertical:high, vertical:medium, vertical:low or unknown. When switching polarization the polarization attribute may contain the value unknown until the configuration change is done and the kernel has figured out the new polarization of each cpu. Note that running a system with different types of vertical cpus may result in significant performance regressions. If possible only one type of vertical cpus should be used. All other cpus should be offlined. Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com> Signed-off-by: Heiko Carstens <heiko.carstens@de.ibm.com>
2008-04-17 13:46:13 +08:00
void topology_schedule_update(void)
{
schedule_work(&topology_work);
}
static void topology_timer_fn(unsigned long ignored)
{
[S390] Vertical cpu management. If vertical cpu polarization is active then the hypervisor will dispatch certain cpus for a longer time than other cpus for maximum performance. For example if a guest would have three virtual cpus, each of them with a share of 33 percent, then in case of vertical cpu polarization all of the processing time would be combined to a single cpu which would run all the time, while the other two cpus would get nearly no cpu time. There are three different types of vertical cpus: high, medium and low. Low cpus hardly get any real cpu time, while high cpus get a full real cpu. Medium cpus get something in between. In order to switch between the two possible modes (default is horizontal) a 0 for horizontal polarization or a 1 for vertical polarization must be written to the dispatching sysfs attribute: /sys/devices/system/cpu/dispatching The polarization of each single cpu can be figured out by the polarization sysfs attribute of each cpu: /sys/devices/system/cpu/cpuX/polarization horizontal, vertical:high, vertical:medium, vertical:low or unknown. When switching polarization the polarization attribute may contain the value unknown until the configuration change is done and the kernel has figured out the new polarization of each cpu. Note that running a system with different types of vertical cpus may result in significant performance regressions. If possible only one type of vertical cpus should be used. All other cpus should be offlined. Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com> Signed-off-by: Heiko Carstens <heiko.carstens@de.ibm.com>
2008-04-17 13:46:13 +08:00
if (ptf(PTF_CHECK))
topology_schedule_update();
set_topology_timer();
}
static struct timer_list topology_timer =
TIMER_DEFERRED_INITIALIZER(topology_timer_fn, 0, 0);
static atomic_t topology_poll = ATOMIC_INIT(0);
static void set_topology_timer(void)
{
if (atomic_add_unless(&topology_poll, -1, 0))
mod_timer(&topology_timer, jiffies + HZ / 10);
else
mod_timer(&topology_timer, jiffies + HZ * 60);
}
void topology_expect_change(void)
{
if (!MACHINE_HAS_TOPOLOGY)
return;
/* This is racy, but it doesn't matter since it is just a heuristic.
* Worst case is that we poll in a higher frequency for a bit longer.
*/
if (atomic_read(&topology_poll) > 60)
return;
atomic_add(60, &topology_poll);
set_topology_timer();
}
static int __init early_parse_topology(char *p)
{
if (strncmp(p, "off", 3))
return 0;
topology_enabled = 0;
return 0;
}
early_param("topology", early_parse_topology);
static void __init alloc_masks(struct sysinfo_15_1_x *info,
struct mask_info *mask, int offset)
{
int i, nr_masks;
nr_masks = info->mag[TOPOLOGY_NR_MAG - offset];
for (i = 0; i < info->mnest - offset; i++)
nr_masks *= info->mag[TOPOLOGY_NR_MAG - offset - 1 - i];
nr_masks = max(nr_masks, 1);
for (i = 0; i < nr_masks; i++) {
mask->next = alloc_bootmem_align(
roundup_pow_of_two(sizeof(struct mask_info)),
roundup_pow_of_two(sizeof(struct mask_info)));
mask = mask->next;
}
}
void __init s390_init_cpu_topology(void)
{
struct sysinfo_15_1_x *info;
int i;
if (!MACHINE_HAS_TOPOLOGY)
return;
tl_info = alloc_bootmem_pages(PAGE_SIZE);
info = tl_info;
store_topology(info);
pr_info("The CPU configuration topology of the machine is:");
for (i = 0; i < TOPOLOGY_NR_MAG; i++)
printk(KERN_CONT " %d", info->mag[i]);
printk(KERN_CONT " / %d\n", info->mnest);
alloc_masks(info, &socket_info, 1);
alloc_masks(info, &book_info, 2);
}
static int cpu_management;
static ssize_t dispatching_show(struct device *dev,
struct device_attribute *attr,
char *buf)
{
ssize_t count;
mutex_lock(&smp_cpu_state_mutex);
count = sprintf(buf, "%d\n", cpu_management);
mutex_unlock(&smp_cpu_state_mutex);
return count;
}
static ssize_t dispatching_store(struct device *dev,
struct device_attribute *attr,
const char *buf,
size_t count)
{
int val, rc;
char delim;
if (sscanf(buf, "%d %c", &val, &delim) != 1)
return -EINVAL;
if (val != 0 && val != 1)
return -EINVAL;
rc = 0;
get_online_cpus();
mutex_lock(&smp_cpu_state_mutex);
if (cpu_management == val)
goto out;
rc = topology_set_cpu_management(val);
if (rc)
goto out;
cpu_management = val;
topology_expect_change();
out:
mutex_unlock(&smp_cpu_state_mutex);
put_online_cpus();
return rc ? rc : count;
}
static DEVICE_ATTR(dispatching, 0644, dispatching_show,
dispatching_store);
static ssize_t cpu_polarization_show(struct device *dev,
struct device_attribute *attr, char *buf)
{
int cpu = dev->id;
ssize_t count;
mutex_lock(&smp_cpu_state_mutex);
switch (smp_cpu_get_polarization(cpu)) {
case POLARIZATION_HRZ:
count = sprintf(buf, "horizontal\n");
break;
case POLARIZATION_VL:
count = sprintf(buf, "vertical:low\n");
break;
case POLARIZATION_VM:
count = sprintf(buf, "vertical:medium\n");
break;
case POLARIZATION_VH:
count = sprintf(buf, "vertical:high\n");
break;
default:
count = sprintf(buf, "unknown\n");
break;
}
mutex_unlock(&smp_cpu_state_mutex);
return count;
}
static DEVICE_ATTR(polarization, 0444, cpu_polarization_show, NULL);
static struct attribute *topology_cpu_attrs[] = {
&dev_attr_polarization.attr,
NULL,
};
static struct attribute_group topology_cpu_attr_group = {
.attrs = topology_cpu_attrs,
};
int topology_cpu_init(struct cpu *cpu)
{
return sysfs_create_group(&cpu->dev.kobj, &topology_cpu_attr_group);
}
const struct cpumask *cpu_coregroup_mask(int cpu)
{
return &cpu_topology[cpu].core_mask;
}
static const struct cpumask *cpu_book_mask(int cpu)
{
return &cpu_topology[cpu].book_mask;
}
static struct sched_domain_topology_level s390_topology[] = {
{ cpu_coregroup_mask, cpu_core_flags, SD_INIT_NAME(MC) },
{ cpu_book_mask, SD_INIT_NAME(BOOK) },
{ cpu_cpu_mask, SD_INIT_NAME(DIE) },
{ NULL, },
};
static int __init topology_init(void)
{
if (MACHINE_HAS_TOPOLOGY)
set_topology_timer();
else
topology_update_polarization_simple();
return device_create_file(cpu_subsys.dev_root, &dev_attr_dispatching);
}
device_initcall(topology_init);
static int __init early_topology_init(void)
{
set_sched_topology(s390_topology);
return 0;
}
early_initcall(early_topology_init);