summaryrefslogtreecommitdiff
path: root/arch/powerpc/platforms/powernv/smp.c
blob: 4aff754b6f2c8da6e94ae7dd5eedced19f95245e (plain)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
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
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
/*
 * SMP support for PowerNV machines.
 *
 * Copyright 2011 IBM Corp.
 *
 * This program is free software; you can redistribute it and/or
 * modify it under the terms of the GNU General Public License
 * as published by the Free Software Foundation; either version
 * 2 of the License, or (at your option) any later version.
 */

#include <linux/kernel.h>
#include <linux/module.h>
#include <linux/sched.h>
#include <linux/sched/hotplug.h>
#include <linux/smp.h>
#include <linux/interrupt.h>
#include <linux/delay.h>
#include <linux/init.h>
#include <linux/spinlock.h>
#include <linux/cpu.h>

#include <asm/irq.h>
#include <asm/smp.h>
#include <asm/paca.h>
#include <asm/machdep.h>
#include <asm/cputable.h>
#include <asm/firmware.h>
#include <asm/vdso_datapage.h>
#include <asm/cputhreads.h>
#include <asm/xics.h>
#include <asm/xive.h>
#include <asm/opal.h>
#include <asm/runlatch.h>
#include <asm/code-patching.h>
#include <asm/dbell.h>
#include <asm/kvm_ppc.h>
#include <asm/ppc-opcode.h>
#include <asm/cpuidle.h>

#include "powernv.h"

#ifdef DEBUG
#include <asm/udbg.h>
#define DBG(fmt...) udbg_printf(fmt)
#else
#define DBG(fmt...)
#endif

static void pnv_smp_setup_cpu(int cpu)
{
	if (xive_enabled())
		xive_smp_setup_cpu();
	else if (cpu != boot_cpuid)
		xics_setup_cpu();
}

static int pnv_smp_kick_cpu(int nr)
{
	unsigned int pcpu = get_hard_smp_processor_id(nr);
	unsigned long start_here =
			__pa(ppc_function_entry(generic_secondary_smp_init));
	long rc;
	uint8_t status;

	BUG_ON(nr < 0 || nr >= NR_CPUS);

	/*
	 * If we already started or OPAL is not supported, we just
	 * kick the CPU via the PACA
	 */
	if (paca[nr].cpu_start || !firmware_has_feature(FW_FEATURE_OPAL))
		goto kick;

	/*
	 * At this point, the CPU can either be spinning on the way in
	 * from kexec or be inside OPAL waiting to be started for the
	 * first time. OPAL v3 allows us to query OPAL to know if it
	 * has the CPUs, so we do that
	 */
	rc = opal_query_cpu_status(pcpu, &status);
	if (rc != OPAL_SUCCESS) {
		pr_warn("OPAL Error %ld querying CPU %d state\n", rc, nr);
		return -ENODEV;
	}

	/*
	 * Already started, just kick it, probably coming from
	 * kexec and spinning
	 */
	if (status == OPAL_THREAD_STARTED)
		goto kick;

	/*
	 * Available/inactive, let's kick it
	 */
	if (status == OPAL_THREAD_INACTIVE) {
		pr_devel("OPAL: Starting CPU %d (HW 0x%x)...\n", nr, pcpu);
		rc = opal_start_cpu(pcpu, start_here);
		if (rc != OPAL_SUCCESS) {
			pr_warn("OPAL Error %ld starting CPU %d\n", rc, nr);
			return -ENODEV;
		}
	} else {
		/*
		 * An unavailable CPU (or any other unknown status)
		 * shouldn't be started. It should also
		 * not be in the possible map but currently it can
		 * happen
		 */
		pr_devel("OPAL: CPU %d (HW 0x%x) is unavailable"
			 " (status %d)...\n", nr, pcpu, status);
		return -ENODEV;
	}

kick:
	return smp_generic_kick_cpu(nr);
}

#ifdef CONFIG_HOTPLUG_CPU

static int pnv_smp_cpu_disable(void)
{
	int cpu = smp_processor_id();

	/* This is identical to pSeries... might consolidate by
	 * moving migrate_irqs_away to a ppc_md with default to
	 * the generic fixup_irqs. --BenH.
	 */
	set_cpu_online(cpu, false);
	vdso_data->processorCount--;
	if (cpu == boot_cpuid)
		boot_cpuid = cpumask_any(cpu_online_mask);
	if (xive_enabled())
		xive_smp_disable_cpu();
	else
		xics_migrate_irqs_away();
	return 0;
}

static void pnv_smp_cpu_kill_self(void)
{
	unsigned int cpu;
	unsigned long srr1, wmask;

	/* Standard hot unplug procedure */
	local_irq_disable();
	idle_task_exit();
	current->active_mm = NULL; /* for sanity */
	cpu = smp_processor_id();
	DBG("CPU%d offline\n", cpu);
	generic_set_cpu_dead(cpu);
	smp_wmb();

	wmask = SRR1_WAKEMASK;
	if (cpu_has_feature(CPU_FTR_ARCH_207S))
		wmask = SRR1_WAKEMASK_P8;

	/* We don't want to take decrementer interrupts while we are offline,
	 * so clear LPCR:PECE1. We keep PECE2 (and LPCR_PECE_HVEE on P9)
	 * enabled as to let IPIs in.
	 */
	mtspr(SPRN_LPCR, mfspr(SPRN_LPCR) & ~(u64)LPCR_PECE1);

	/*
	 * Hard-disable interrupts, and then clear irq_happened flags
	 * that we can safely ignore while off-line, since they
	 * are for things for which we do no processing when off-line
	 * (or in the case of HMI, all the processing we need to do
	 * is done in lower-level real-mode code).
	 */
	hard_irq_disable();
	local_paca->irq_happened &= ~(PACA_IRQ_DEC | PACA_IRQ_HMI);

	while (!generic_check_cpu_restart(cpu)) {
		/*
		 * Clear IPI flag, since we don't handle IPIs while
		 * offline, except for those when changing micro-threading
		 * mode, which are handled explicitly below, and those
		 * for coming online, which are handled via
		 * generic_check_cpu_restart() calls.
		 */
		kvmppc_set_host_ipi(cpu, 0);

		ppc64_runlatch_off();
		srr1 = pnv_cpu_offline(cpu);
		ppc64_runlatch_on();

		/*
		 * If the SRR1 value indicates that we woke up due to
		 * an external interrupt, then clear the interrupt.
		 * We clear the interrupt before checking for the
		 * reason, so as to avoid a race where we wake up for
		 * some other reason, find nothing and clear the interrupt
		 * just as some other cpu is sending us an interrupt.
		 * If we returned from power7_nap as a result of
		 * having finished executing in a KVM guest, then srr1
		 * contains 0.
		 */
		if (((srr1 & wmask) == SRR1_WAKEEE) ||
		    ((srr1 & wmask) == SRR1_WAKEHVI) ||
		    (local_paca->irq_happened & PACA_IRQ_EE)) {
			if (cpu_has_feature(CPU_FTR_ARCH_300)) {
				if (xive_enabled())
					xive_flush_interrupt();
				else
					icp_opal_flush_interrupt();
			} else
				icp_native_flush_interrupt();
		} else if ((srr1 & wmask) == SRR1_WAKEHDBELL) {
			unsigned long msg = PPC_DBELL_TYPE(PPC_DBELL_SERVER);
			asm volatile(PPC_MSGCLR(%0) : : "r" (msg));
		}
		local_paca->irq_happened &= ~(PACA_IRQ_EE | PACA_IRQ_DBELL);
		smp_mb();

		if (cpu_core_split_required())
			continue;

		if (srr1 && !generic_check_cpu_restart(cpu))
			DBG("CPU%d Unexpected exit while offline !\n", cpu);
	}

	/* Re-enable decrementer interrupts */
	mtspr(SPRN_LPCR, mfspr(SPRN_LPCR) | LPCR_PECE1);
	DBG("CPU%d coming online...\n", cpu);
}

#endif /* CONFIG_HOTPLUG_CPU */

static int pnv_cpu_bootable(unsigned int nr)
{
	/*
	 * Starting with POWER8, the subcore logic relies on all threads of a
	 * core being booted so that they can participate in split mode
	 * switches. So on those machines we ignore the smt_enabled_at_boot
	 * setting (smt-enabled on the kernel command line).
	 */
	if (cpu_has_feature(CPU_FTR_ARCH_207S))
		return 1;

	return smp_generic_cpu_bootable(nr);
}

static int pnv_smp_prepare_cpu(int cpu)
{
	if (xive_enabled())
		return xive_smp_prepare_cpu(cpu);
	return 0;
}

/* Cause IPI as setup by the interrupt controller (xics or xive) */
static void (*ic_cause_ipi)(int cpu);

static void pnv_cause_ipi(int cpu)
{
	if (doorbell_try_core_ipi(cpu))
		return;

	ic_cause_ipi(cpu);
}

static void pnv_p9_dd1_cause_ipi(int cpu)
{
	int this_cpu = get_cpu();

	/*
	 * POWER9 DD1 has a global addressed msgsnd, but for now we restrict
	 * IPIs to same core, because it requires additional synchronization
	 * for inter-core doorbells which we do not implement.
	 */
	if (cpumask_test_cpu(cpu, cpu_sibling_mask(this_cpu)))
		doorbell_global_ipi(cpu);
	else
		ic_cause_ipi(cpu);

	put_cpu();
}

static void __init pnv_smp_probe(void)
{
	if (xive_enabled())
		xive_smp_probe();
	else
		xics_smp_probe();

	if (cpu_has_feature(CPU_FTR_DBELL)) {
		ic_cause_ipi = smp_ops->cause_ipi;
		WARN_ON(!ic_cause_ipi);

		if (cpu_has_feature(CPU_FTR_ARCH_300)) {
			if (cpu_has_feature(CPU_FTR_POWER9_DD1))
				smp_ops->cause_ipi = pnv_p9_dd1_cause_ipi;
			else
				smp_ops->cause_ipi = doorbell_global_ipi;
		} else {
			smp_ops->cause_ipi = pnv_cause_ipi;
		}
	}
}

static struct smp_ops_t pnv_smp_ops = {
	.message_pass	= NULL, /* Use smp_muxed_ipi_message_pass */
	.cause_ipi	= NULL,	/* Filled at runtime by pnv_smp_probe() */
	.cause_nmi_ipi	= NULL,
	.probe		= pnv_smp_probe,
	.prepare_cpu	= pnv_smp_prepare_cpu,
	.kick_cpu	= pnv_smp_kick_cpu,
	.setup_cpu	= pnv_smp_setup_cpu,
	.cpu_bootable	= pnv_cpu_bootable,
#ifdef CONFIG_HOTPLUG_CPU
	.cpu_disable	= pnv_smp_cpu_disable,
	.cpu_die	= generic_cpu_die,
#endif /* CONFIG_HOTPLUG_CPU */
};

/* This is called very early during platform setup_arch */
void __init pnv_smp_init(void)
{
	smp_ops = &pnv_smp_ops;

#ifdef CONFIG_HOTPLUG_CPU
	ppc_md.cpu_die	= pnv_smp_cpu_kill_self;
#endif
}