smp.c 9.21 KB
Newer Older
1
2
3
/*
 *	Intel SMP support routines.
 *
4
 *	(c) 1995 Alan Cox, Building #3 <alan@lxorguk.ukuu.org.uk>
Ingo Molnar's avatar
Ingo Molnar committed
5
 *	(c) 1998-99, 2000, 2009 Ingo Molnar <mingo@redhat.com>
6
7
8
9
10
11
12
13
 *      (c) 2002,2003 Andi Kleen, SuSE Labs.
 *
 *	i386 and x86_64 integration by Glauber Costa <gcosta@redhat.com>
 *
 *	This code is released under the GNU General Public License version 2 or
 *	later.
 */

Glauber Costa's avatar
Glauber Costa committed
14
15
16
17
18
#include <linux/init.h>

#include <linux/mm.h>
#include <linux/delay.h>
#include <linux/spinlock.h>
19
#include <linux/export.h>
Glauber Costa's avatar
Glauber Costa committed
20
21
22
23
24
#include <linux/kernel_stat.h>
#include <linux/mc146818rtc.h>
#include <linux/cache.h>
#include <linux/interrupt.h>
#include <linux/cpu.h>
25
#include <linux/gfp.h>
Glauber Costa's avatar
Glauber Costa committed
26
27
28
29
30

#include <asm/mtrr.h>
#include <asm/tlbflush.h>
#include <asm/mmu_context.h>
#include <asm/proto.h>
Ingo Molnar's avatar
Ingo Molnar committed
31
#include <asm/apic.h>
32
#include <asm/nmi.h>
33
#include <asm/mce.h>
34
#include <asm/trace/irq_vectors.h>
35
#include <asm/kexec.h>
36
#include <asm/virtext.h>
37

38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
/*
 *	Some notes on x86 processor bugs affecting SMP operation:
 *
 *	Pentium, Pentium Pro, II, III (and all CPUs) have bugs.
 *	The Linux implications for SMP are handled as follows:
 *
 *	Pentium III / [Xeon]
 *		None of the E1AP-E3AP errata are visible to the user.
 *
 *	E1AP.	see PII A1AP
 *	E2AP.	see PII A2AP
 *	E3AP.	see PII A3AP
 *
 *	Pentium II / [Xeon]
 *		None of the A1AP-A3AP errata are visible to the user.
 *
 *	A1AP.	see PPro 1AP
 *	A2AP.	see PPro 2AP
 *	A3AP.	see PPro 7AP
 *
 *	Pentium Pro
 *		None of 1AP-9AP errata are visible to the normal user,
 *	except occasional delivery of 'spurious interrupt' as trap #15.
 *	This is very rare and a non-problem.
 *
 *	1AP.	Linux maps APIC as non-cacheable
 *	2AP.	worked around in hardware
 *	3AP.	fixed in C0 and above steppings microcode update.
 *		Linux does not use excessive STARTUP_IPIs.
 *	4AP.	worked around in hardware
 *	5AP.	symmetric IO mode (normal Linux operation) not affected.
 *		'noapic' mode has vector 0xf filled out properly.
 *	6AP.	'noapic' mode might be affected - fixed in later steppings
 *	7AP.	We do not assume writes to the LVT deassering IRQs
 *	8AP.	We do not enable low power mode (deep sleep) during MP bootup
 *	9AP.	We do not use mixed mode
 *
 *	Pentium
 *		There is a marginal case where REP MOVS on 100MHz SMP
 *	machines with B stepping processors can fail. XXX should provide
 *	an L1cache=Writethrough or L1cache=off option.
 *
 *		B stepping CPUs may hang. There are hardware work arounds
 *	for this. We warn about it in case your board doesn't have the work
 *	arounds. Basically that's so I can tell anyone with a B stepping
 *	CPU and SMP problems "tough".
 *
 *	Specific items [From Pentium Processor Specification Update]
 *
 *	1AP.	Linux doesn't use remote read
 *	2AP.	Linux doesn't trust APIC errors
 *	3AP.	We work around this
 *	4AP.	Linux never generated 3 interrupts of the same priority
 *		to cause a lost local interrupt.
 *	5AP.	Remote read is never used
 *	6AP.	not affected - worked around in hardware
 *	7AP.	not affected - worked around in hardware
 *	8AP.	worked around in hardware - we get explicit CS errors if not
 *	9AP.	only 'noapic' mode affected. Might generate spurious
 *		interrupts, we log only the first one and count the
 *		rest silently.
 *	10AP.	not affected - worked around in hardware
 *	11AP.	Linux reads the APIC between writes to avoid this, as per
 *		the documentation. Make sure you preserve this as it affects
 *		the C stepping chips too.
 *	12AP.	not affected - worked around in hardware
 *	13AP.	not affected - worked around in hardware
 *	14AP.	we always deassert INIT during bootup
 *	15AP.	not affected - worked around in hardware
 *	16AP.	not affected - worked around in hardware
 *	17AP.	not affected - worked around in hardware
 *	18AP.	not affected - worked around in hardware
 *	19AP.	not affected - worked around in BIOS
 *
 *	If this sounds worrying believe me these bugs are either ___RARE___,
 *	or are signal timing bugs worked around in hardware and there's
 *	about nothing of note with C stepping upwards.
 */
Glauber Costa's avatar
Glauber Costa committed
116

117
static atomic_t stopping_cpu = ATOMIC_INIT(-1);
118
static bool smp_no_nmi_ipi = false;
119

Glauber Costa's avatar
Glauber Costa committed
120
121
122
123
124
125
126
/*
 * this function sends a 'reschedule' IPI to another CPU.
 * it goes straight through and wastes no time serializing
 * anything. Worst case is that we lose a reschedule ...
 */
static void native_smp_send_reschedule(int cpu)
{
127
	if (unlikely(cpu_is_offline(cpu))) {
128
		WARN(1, "sched: Unexpected reschedule of offline CPU#%d!\n", cpu);
129
130
		return;
	}
131
	apic->send_IPI(cpu, RESCHEDULE_VECTOR);
Glauber Costa's avatar
Glauber Costa committed
132
133
}

134
void native_send_call_func_single_ipi(int cpu)
Glauber Costa's avatar
Glauber Costa committed
135
{
136
	apic->send_IPI(cpu, CALL_FUNCTION_SINGLE_VECTOR);
Glauber Costa's avatar
Glauber Costa committed
137
138
}

139
void native_send_call_func_ipi(const struct cpumask *mask)
Glauber Costa's avatar
Glauber Costa committed
140
{
141
	cpumask_var_t allbutself;
Glauber Costa's avatar
Glauber Costa committed
142

143
	if (!alloc_cpumask_var(&allbutself, GFP_ATOMIC)) {
144
		apic->send_IPI_mask(mask, CALL_FUNCTION_VECTOR);
145
146
		return;
	}
Glauber Costa's avatar
Glauber Costa committed
147

148
149
150
151
152
	cpumask_copy(allbutself, cpu_online_mask);
	cpumask_clear_cpu(smp_processor_id(), allbutself);

	if (cpumask_equal(mask, allbutself) &&
	    cpumask_equal(cpu_online_mask, cpu_callout_mask))
153
		apic->send_IPI_allbutself(CALL_FUNCTION_VECTOR);
Glauber Costa's avatar
Glauber Costa committed
154
	else
155
		apic->send_IPI_mask(mask, CALL_FUNCTION_VECTOR);
156
157

	free_cpumask_var(allbutself);
Glauber Costa's avatar
Glauber Costa committed
158
159
}

160
161
162
163
164
165
static int smp_stop_nmi_callback(unsigned int val, struct pt_regs *regs)
{
	/* We are registered on stopping cpu too, avoid spurious NMI */
	if (raw_smp_processor_id() == atomic_read(&stopping_cpu))
		return NMI_HANDLED;

166
	cpu_emergency_vmxoff();
167
168
169
170
171
	stop_this_cpu(NULL);

	return NMI_HANDLED;
}

Glauber Costa's avatar
Glauber Costa committed
172
173
174
175
/*
 * this function calls the 'stop' function on all other CPUs in the system.
 */

176
asmlinkage __visible void smp_reboot_interrupt(void)
177
{
178
	ipi_entering_ack_irq();
179
	cpu_emergency_vmxoff();
180
181
182
183
	stop_this_cpu(NULL);
	irq_exit();
}

184
static void native_stop_other_cpus(int wait)
Glauber Costa's avatar
Glauber Costa committed
185
186
{
	unsigned long flags;
187
	unsigned long timeout;
Glauber Costa's avatar
Glauber Costa committed
188
189
190
191

	if (reboot_force)
		return;

192
193
194
	/*
	 * Use an own vector here because smp_call_function
	 * does lots of things not suitable in a panic situation.
195
196
197
198
199
200
201
202
203
204
	 */

	/*
	 * We start by using the REBOOT_VECTOR irq.
	 * The irq is treated as a sync point to allow critical
	 * regions of code on other cpus to release their spin locks
	 * and re-enable irqs.  Jumping straight to an NMI might
	 * accidentally cause deadlocks with further shutdown/panic
	 * code.  By syncing, we give the cpus up to one second to
	 * finish their work before we force them off with the NMI.
205
206
	 */
	if (num_online_cpus() > 1) {
207
208
209
210
211
212
213
		/* did someone beat us here? */
		if (atomic_cmpxchg(&stopping_cpu, -1, safe_smp_processor_id()) != -1)
			return;

		/* sync above data before sending IRQ */
		wmb();

214
215
		apic->send_IPI_allbutself(REBOOT_VECTOR);

216
217
218
219
220
221
		/*
		 * Don't wait longer than a second if the caller
		 * didn't ask us to wait.
		 */
		timeout = USEC_PER_SEC;
		while (num_online_cpus() > 1 && (wait || timeout--))
222
223
			udelay(1);
	}
224
225
	
	/* if the REBOOT_VECTOR didn't work, try with the NMI */
226
	if ((num_online_cpus() > 1) && (!smp_no_nmi_ipi))  {
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
		if (register_nmi_handler(NMI_LOCAL, smp_stop_nmi_callback,
					 NMI_FLAG_FIRST, "smp_stop"))
			/* Note: we ignore failures here */
			/* Hope the REBOOT_IRQ is good enough */
			goto finish;

		/* sync above data before sending IRQ */
		wmb();

		pr_emerg("Shutting down cpus with NMI\n");

		apic->send_IPI_allbutself(NMI_VECTOR);

		/*
		 * Don't wait longer than a 10 ms if the caller
		 * didn't ask us to wait.
		 */
		timeout = USEC_PER_MSEC * 10;
		while (num_online_cpus() > 1 && (wait || timeout--))
			udelay(1);
	}
248

249
finish:
Glauber Costa's avatar
Glauber Costa committed
250
251
	local_irq_save(flags);
	disable_local_APIC();
252
	mcheck_cpu_clear(this_cpu_ptr(&cpu_info));
Glauber Costa's avatar
Glauber Costa committed
253
254
255
256
	local_irq_restore(flags);
}

/*
257
258
 * Reschedule call back. KVM uses this interrupt to force a cpu out of
 * guest mode
Glauber Costa's avatar
Glauber Costa committed
259
 */
260
__visible void __irq_entry smp_reschedule_interrupt(struct pt_regs *regs)
261
262
{
	ack_APIC_irq();
263
	inc_irq_stat(irq_resched_count);
264
	kvm_set_cpu_l1tf_flush_l1d();
265

266
	if (trace_resched_ipi_enabled()) {
267
268
269
270
271
272
273
274
275
276
277
		/*
		 * scheduler_ipi() might call irq_enter() as well, but
		 * nested calls are fine.
		 */
		irq_enter();
		trace_reschedule_entry(RESCHEDULE_VECTOR);
		scheduler_ipi();
		trace_reschedule_exit(RESCHEDULE_VECTOR);
		irq_exit();
		return;
	}
278
	scheduler_ipi();
279
}
Glauber Costa's avatar
Glauber Costa committed
280

281
__visible void __irq_entry smp_call_function_interrupt(struct pt_regs *regs)
282
{
283
	ipi_entering_ack_irq();
284
	trace_call_function_entry(CALL_FUNCTION_VECTOR);
285
	inc_irq_stat(irq_call_count);
286
287
	generic_smp_call_function_interrupt();
	trace_call_function_exit(CALL_FUNCTION_VECTOR);
288
	exiting_irq();
Glauber Costa's avatar
Glauber Costa committed
289
290
}

291
__visible void __irq_entry smp_call_function_single_interrupt(struct pt_regs *r)
292
{
293
	ipi_entering_ack_irq();
294
	trace_call_function_single_entry(CALL_FUNCTION_SINGLE_VECTOR);
295
296
	inc_irq_stat(irq_call_count);
	generic_smp_call_function_single_interrupt();
297
298
299
300
	trace_call_function_single_exit(CALL_FUNCTION_SINGLE_VECTOR);
	exiting_irq();
}

301
302
static int __init nonmi_ipi_setup(char *str)
{
303
304
	smp_no_nmi_ipi = true;
	return 1;
305
306
307
308
}

__setup("nonmi_ipi", nonmi_ipi_setup);

Glauber Costa's avatar
Glauber Costa committed
309
struct smp_ops smp_ops = {
310
311
312
	.smp_prepare_boot_cpu	= native_smp_prepare_boot_cpu,
	.smp_prepare_cpus	= native_smp_prepare_cpus,
	.smp_cpus_done		= native_smp_cpus_done,
Glauber Costa's avatar
Glauber Costa committed
313

314
	.stop_other_cpus	= native_stop_other_cpus,
315
316
317
#if defined(CONFIG_KEXEC_CORE)
	.crash_stop_other_cpus	= kdump_nmi_shootdown_cpus,
#endif
318
	.smp_send_reschedule	= native_smp_send_reschedule,
319

320
321
322
323
	.cpu_up			= native_cpu_up,
	.cpu_die		= native_cpu_die,
	.cpu_disable		= native_cpu_disable,
	.play_dead		= native_play_dead,
324

325
	.send_call_func_ipi	= native_send_call_func_ipi,
326
	.send_call_func_single_ipi = native_send_call_func_single_ipi,
Glauber Costa's avatar
Glauber Costa committed
327
328
};
EXPORT_SYMBOL_GPL(smp_ops);