summaryrefslogtreecommitdiff
path: root/arch/mips/kernel/pm-cps.c
blob: 1d2996cd58daec79fc7d05dbeafb0c5c353d0cd4 (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
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
/*
 * Copyright (C) 2014 Imagination Technologies
 * Author: Paul Burton <paul.burton@imgtec.com>
 *
 * 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/cpuhotplug.h>
#include <linux/init.h>
#include <linux/percpu.h>
#include <linux/slab.h>

#include <asm/asm-offsets.h>
#include <asm/cacheflush.h>
#include <asm/cacheops.h>
#include <asm/idle.h>
#include <asm/mips-cps.h>
#include <asm/mipsmtregs.h>
#include <asm/pm.h>
#include <asm/pm-cps.h>
#include <asm/smp-cps.h>
#include <asm/uasm.h>

/*
 * cps_nc_entry_fn - type of a generated non-coherent state entry function
 * @online: the count of online coupled VPEs
 * @nc_ready_count: pointer to a non-coherent mapping of the core ready_count
 *
 * The code entering & exiting non-coherent states is generated at runtime
 * using uasm, in order to ensure that the compiler cannot insert a stray
 * memory access at an unfortunate time and to allow the generation of optimal
 * core-specific code particularly for cache routines. If coupled_coherence
 * is non-zero and this is the entry function for the CPS_PM_NC_WAIT state,
 * returns the number of VPEs that were in the wait state at the point this
 * VPE left it. Returns garbage if coupled_coherence is zero or this is not
 * the entry function for CPS_PM_NC_WAIT.
 */
typedef unsigned (*cps_nc_entry_fn)(unsigned online, u32 *nc_ready_count);

/*
 * The entry point of the generated non-coherent idle state entry/exit
 * functions. Actually per-core rather than per-CPU.
 */
static DEFINE_PER_CPU_READ_MOSTLY(cps_nc_entry_fn[CPS_PM_STATE_COUNT],
				  nc_asm_enter);

/* Bitmap indicating which states are supported by the system */
static DECLARE_BITMAP(state_support, CPS_PM_STATE_COUNT);

/*
 * Indicates the number of coupled VPEs ready to operate in a non-coherent
 * state. Actually per-core rather than per-CPU.
 */
static DEFINE_PER_CPU_ALIGNED(u32*, ready_count);

/* Indicates online CPUs coupled with the current CPU */
static DEFINE_PER_CPU_ALIGNED(cpumask_t, online_coupled);

/*
 * Used to synchronize entry to deep idle states. Actually per-core rather
 * than per-CPU.
 */
static DEFINE_PER_CPU_ALIGNED(atomic_t, pm_barrier);

/* Saved CPU state across the CPS_PM_POWER_GATED state */
DEFINE_PER_CPU_ALIGNED(struct mips_static_suspend_state, cps_cpu_state);

/* A somewhat arbitrary number of labels & relocs for uasm */
static struct uasm_label labels[32];
static struct uasm_reloc relocs[32];

enum mips_reg {
	zero, at, v0, v1, a0, a1, a2, a3,
	t0, t1, t2, t3, t4, t5, t6, t7,
	s0, s1, s2, s3, s4, s5, s6, s7,
	t8, t9, k0, k1, gp, sp, fp, ra,
};

bool cps_pm_support_state(enum cps_pm_state state)
{
	return test_bit(state, state_support);
}

static void coupled_barrier(atomic_t *a, unsigned online)
{
	/*
	 * This function is effectively the same as
	 * cpuidle_coupled_parallel_barrier, which can't be used here since
	 * there's no cpuidle device.
	 */

	if (!coupled_coherence)
		return;

	smp_mb__before_atomic();
	atomic_inc(a);

	while (atomic_read(a) < online)
		cpu_relax();

	if (atomic_inc_return(a) == online * 2) {
		atomic_set(a, 0);
		return;
	}

	while (atomic_read(a) > online)
		cpu_relax();
}

int cps_pm_enter_state(enum cps_pm_state state)
{
	unsigned cpu = smp_processor_id();
	unsigned core = cpu_core(&current_cpu_data);
	unsigned online, left;
	cpumask_t *coupled_mask = this_cpu_ptr(&online_coupled);
	u32 *core_ready_count, *nc_core_ready_count;
	void *nc_addr;
	cps_nc_entry_fn entry;
	struct core_boot_config *core_cfg;
	struct vpe_boot_config *vpe_cfg;

	/* Check that there is an entry function for this state */
	entry = per_cpu(nc_asm_enter, core)[state];
	if (!entry)
		return -EINVAL;

	/* Calculate which coupled CPUs (VPEs) are online */
#if defined(CONFIG_MIPS_MT) || defined(CONFIG_CPU_MIPSR6)
	if (cpu_online(cpu)) {
		cpumask_and(coupled_mask, cpu_online_mask,
			    &cpu_sibling_map[cpu]);
		online = cpumask_weight(coupled_mask);
		cpumask_clear_cpu(cpu, coupled_mask);
	} else
#endif
	{
		cpumask_clear(coupled_mask);
		online = 1;
	}

	/* Setup the VPE to run mips_cps_pm_restore when started again */
	if (IS_ENABLED(CONFIG_CPU_PM) && state == CPS_PM_POWER_GATED) {
		/* Power gating relies upon CPS SMP */
		if (!mips_cps_smp_in_use())
			return -EINVAL;

		core_cfg = &mips_cps_core_bootcfg[core];
		vpe_cfg = &core_cfg->vpe_config[cpu_vpe_id(&current_cpu_data)];
		vpe_cfg->pc = (unsigned long)mips_cps_pm_restore;
		vpe_cfg->gp = (unsigned long)current_thread_info();
		vpe_cfg->sp = 0;
	}

	/* Indicate that this CPU might not be coherent */
	cpumask_clear_cpu(cpu, &cpu_coherent_mask);
	smp_mb__after_atomic();

	/* Create a non-coherent mapping of the core ready_count */
	core_ready_count = per_cpu(ready_count, core);
	nc_addr = kmap_noncoherent(virt_to_page(core_ready_count),
				   (unsigned long)core_ready_count);
	nc_addr += ((unsigned long)core_ready_count & ~PAGE_MASK);
	nc_core_ready_count = nc_addr;

	/* Ensure ready_count is zero-initialised before the assembly runs */
	WRITE_ONCE(*nc_core_ready_count, 0);
	coupled_barrier(&per_cpu(pm_barrier, core), online);

	/* Run the generated entry code */
	left = entry(online, nc_core_ready_count);

	/* Remove the non-coherent mapping of ready_count */
	kunmap_noncoherent();

	/* Indicate that this CPU is definitely coherent */
	cpumask_set_cpu(cpu, &cpu_coherent_mask);

	/*
	 * If this VPE is the first to leave the non-coherent wait state then
	 * it needs to wake up any coupled VPEs still running their wait
	 * instruction so that they return to cpuidle, which can then complete
	 * coordination between the coupled VPEs & provide the governor with
	 * a chance to reflect on the length of time the VPEs were in the
	 * idle state.
	 */
	if (coupled_coherence && (state == CPS_PM_NC_WAIT) && (left == online))
		arch_send_call_function_ipi_mask(coupled_mask);

	return 0;
}

static void cps_gen_cache_routine(u32 **pp, struct uasm_label **pl,
				  struct uasm_reloc **pr,
				  const struct cache_desc *cache,
				  unsigned op, int lbl)
{
	unsigned cache_size = cache->ways << cache->waybit;
	unsigned i;
	const unsigned unroll_lines = 32;

	/* If the cache isn't present this function has it easy */
	if (cache->flags & MIPS_CACHE_NOT_PRESENT)
		return;

	/* Load base address */
	UASM_i_LA(pp, t0, (long)CKSEG0);

	/* Calculate end address */
	if (cache_size < 0x8000)
		uasm_i_addiu(pp, t1, t0, cache_size);
	else
		UASM_i_LA(pp, t1, (long)(CKSEG0 + cache_size));

	/* Start of cache op loop */
	uasm_build_label(pl, *pp, lbl);

	/* Generate the cache ops */
	for (i = 0; i < unroll_lines; i++) {
		if (cpu_has_mips_r6) {
			uasm_i_cache(pp, op, 0, t0);
			uasm_i_addiu(pp, t0, t0, cache->linesz);
		} else {
			uasm_i_cache(pp, op, i * cache->linesz, t0);
		}
	}

	if (!cpu_has_mips_r6)
		/* Update the base address */
		uasm_i_addiu(pp, t0, t0, unroll_lines * cache->linesz);

	/* Loop if we haven't reached the end address yet */
	uasm_il_bne(pp, pr, t0, t1, lbl);
	uasm_i_nop(pp);
}

static int cps_gen_flush_fsb(u32 **pp, struct uasm_label **pl,
			     struct uasm_reloc **pr,
			     const struct cpuinfo_mips *cpu_info,
			     int lbl)
{
	unsigned i, fsb_size = 8;
	unsigned num_loads = (fsb_size * 3) / 2;
	unsigned line_stride = 2;
	unsigned line_size = cpu_info->dcache.linesz;
	unsigned perf_counter, perf_event;
	unsigned revision = cpu_info->processor_id & PRID_REV_MASK;

	/*
	 * Determine whether this CPU requires an FSB flush, and if so which
	 * performance counter/event reflect stalls due to a full FSB.
	 */
	switch (__get_cpu_type(cpu_info->cputype)) {
	case CPU_INTERAPTIV:
		perf_counter = 1;
		perf_event = 51;
		break;

	case CPU_PROAPTIV:
		/* Newer proAptiv cores don't require this workaround */
		if (revision >= PRID_REV_ENCODE_332(1, 1, 0))
			return 0;

		/* On older ones it's unavailable */
		return -1;

	default:
		/* Assume that the CPU does not need this workaround */
		return 0;
	}

	/*
	 * Ensure that the fill/store buffer (FSB) is not holding the results
	 * of a prefetch, since if it is then the CPC sequencer may become
	 * stuck in the D3 (ClrBus) state whilst entering a low power state.
	 */

	/* Preserve perf counter setup */
	uasm_i_mfc0(pp, t2, 25, (perf_counter * 2) + 0); /* PerfCtlN */
	uasm_i_mfc0(pp, t3, 25, (perf_counter * 2) + 1); /* PerfCntN */

	/* Setup perf counter to count FSB full pipeline stalls */
	uasm_i_addiu(pp, t0, zero, (perf_event << 5) | 0xf);
	uasm_i_mtc0(pp, t0, 25, (perf_counter * 2) + 0); /* PerfCtlN */
	uasm_i_ehb(pp);
	uasm_i_mtc0(pp, zero, 25, (perf_counter * 2) + 1); /* PerfCntN */
	uasm_i_ehb(pp);

	/* Base address for loads */
	UASM_i_LA(pp, t0, (long)CKSEG0);

	/* Start of clear loop */
	uasm_build_label(pl, *pp, lbl);

	/* Perform some loads to fill the FSB */
	for (i = 0; i < num_loads; i++)
		uasm_i_lw(pp, zero, i * line_size * line_stride, t0);

	/*
	 * Invalidate the new D-cache entries so that the cache will need
	 * refilling (via the FSB) if the loop is executed again.
	 */
	for (i = 0; i < num_loads; i++) {
		uasm_i_cache(pp, Hit_Invalidate_D,
			     i * line_size * line_stride, t0);
		uasm_i_cache(pp, Hit_Writeback_Inv_SD,
			     i * line_size * line_stride, t0);
	}

	/* Barrier ensuring previous cache invalidates are complete */
	uasm_i_sync(pp, STYPE_SYNC);
	uasm_i_ehb(pp);

	/* Check whether the pipeline stalled due to the FSB being full */
	uasm_i_mfc0(pp, t1, 25, (perf_counter * 2) + 1); /* PerfCntN */

	/* Loop if it didn't */
	uasm_il_beqz(pp, pr, t1, lbl);
	uasm_i_nop(pp);

	/* Restore perf counter 1. The count may well now be wrong... */
	uasm_i_mtc0(pp, t2, 25, (perf_counter * 2) + 0); /* PerfCtlN */
	uasm_i_ehb(pp);
	uasm_i_mtc0(pp, t3, 25, (perf_counter * 2) + 1); /* PerfCntN */
	uasm_i_ehb(pp);

	return 0;
}

static void cps_gen_set_top_bit(u32 **pp, struct uasm_label **pl,
				struct uasm_reloc **pr,
				unsigned r_addr, int lbl)
{
	uasm_i_lui(pp, t0, uasm_rel_hi(0x80000000));
	uasm_build_label(pl, *pp, lbl);
	uasm_i_ll(pp, t1, 0, r_addr);
	uasm_i_or(pp, t1, t1, t0);
	uasm_i_sc(pp, t1, 0, r_addr);
	uasm_il_beqz(pp, pr, t1, lbl);
	uasm_i_nop(pp);
}

static void *cps_gen_entry_code(unsigned cpu, enum cps_pm_state state)
{
	struct uasm_label *l = labels;
	struct uasm_reloc *r = relocs;
	u32 *buf, *p;
	const unsigned r_online = a0;
	const unsigned r_nc_count = a1;
	const unsigned r_pcohctl = t7;
	const unsigned max_instrs = 256;
	unsigned cpc_cmd;
	int err;
	enum {
		lbl_incready = 1,
		lbl_poll_cont,
		lbl_secondary_hang,
		lbl_disable_coherence,
		lbl_flush_fsb,
		lbl_invicache,
		lbl_flushdcache,
		lbl_hang,
		lbl_set_cont,
		lbl_secondary_cont,
		lbl_decready,
	};

	/* Allocate a buffer to hold the generated code */
	p = buf = kcalloc(max_instrs, sizeof(u32), GFP_KERNEL);
	if (!buf)
		return NULL;

	/* Clear labels & relocs ready for (re)use */
	memset(labels, 0, sizeof(labels));
	memset(relocs, 0, sizeof(relocs));

	if (IS_ENABLED(CONFIG_CPU_PM) && state == CPS_PM_POWER_GATED) {
		/* Power gating relies upon CPS SMP */
		if (!mips_cps_smp_in_use())
			goto out_err;

		/*
		 * Save CPU state. Note the non-standard calling convention
		 * with the return address placed in v0 to avoid clobbering
		 * the ra register before it is saved.
		 */
		UASM_i_LA(&p, t0, (long)mips_cps_pm_save);
		uasm_i_jalr(&p, v0, t0);
		uasm_i_nop(&p);
	}

	/*
	 * Load addresses of required CM & CPC registers. This is done early
	 * because they're needed in both the enable & disable coherence steps
	 * but in the coupled case the enable step will only run on one VPE.
	 */
	UASM_i_LA(&p, r_pcohctl, (long)addr_gcr_cl_coherence());

	if (coupled_coherence) {
		/* Increment ready_count */
		uasm_i_sync(&p, STYPE_SYNC_MB);
		uasm_build_label(&l, p, lbl_incready);
		uasm_i_ll(&p, t1, 0, r_nc_count);
		uasm_i_addiu(&p, t2, t1, 1);
		uasm_i_sc(&p, t2, 0, r_nc_count);
		uasm_il_beqz(&p, &r, t2, lbl_incready);
		uasm_i_addiu(&p, t1, t1, 1);

		/* Barrier ensuring all CPUs see the updated r_nc_count value */
		uasm_i_sync(&p, STYPE_SYNC_MB);

		/*
		 * If this is the last VPE to become ready for non-coherence
		 * then it should branch below.
		 */
		uasm_il_beq(&p, &r, t1, r_online, lbl_disable_coherence);
		uasm_i_nop(&p);

		if (state < CPS_PM_POWER_GATED) {
			/*
			 * Otherwise this is not the last VPE to become ready
			 * for non-coherence. It needs to wait until coherence
			 * has been disabled before proceeding, which it will do
			 * by polling for the top bit of ready_count being set.
			 */
			uasm_i_addiu(&p, t1, zero, -1);
			uasm_build_label(&l, p, lbl_poll_cont);
			uasm_i_lw(&p, t0, 0, r_nc_count);
			uasm_il_bltz(&p, &r, t0, lbl_secondary_cont);
			uasm_i_ehb(&p);
			if (cpu_has_mipsmt)
				uasm_i_yield(&p, zero, t1);
			uasm_il_b(&p, &r, lbl_poll_cont);
			uasm_i_nop(&p);
		} else {
			/*
			 * The core will lose power & this VPE will not continue
			 * so it can simply halt here.
			 */
			if (cpu_has_mipsmt) {
				/* Halt the VPE via C0 tchalt register */
				uasm_i_addiu(&p, t0, zero, TCHALT_H);
				uasm_i_mtc0(&p, t0, 2, 4);
			} else if (cpu_has_vp) {
				/* Halt the VP via the CPC VP_STOP register */
				unsigned int vpe_id;

				vpe_id = cpu_vpe_id(&cpu_data[cpu]);
				uasm_i_addiu(&p, t0, zero, 1 << vpe_id);
				UASM_i_LA(&p, t1, (long)addr_cpc_cl_vp_stop());
				uasm_i_sw(&p, t0, 0, t1);
			} else {
				BUG();
			}
			uasm_build_label(&l, p, lbl_secondary_hang);
			uasm_il_b(&p, &r, lbl_secondary_hang);
			uasm_i_nop(&p);
		}
	}

	/*
	 * This is the point of no return - this VPE will now proceed to
	 * disable coherence. At this point we *must* be sure that no other
	 * VPE within the core will interfere with the L1 dcache.
	 */
	uasm_build_label(&l, p, lbl_disable_coherence);

	/* Invalidate the L1 icache */
	cps_gen_cache_routine(&p, &l, &r, &cpu_data[cpu].icache,
			      Index_Invalidate_I, lbl_invicache);

	/* Writeback & invalidate the L1 dcache */
	cps_gen_cache_routine(&p, &l, &r, &cpu_data[cpu].dcache,
			      Index_Writeback_Inv_D, lbl_flushdcache);

	/* Barrier ensuring previous cache invalidates are complete */
	uasm_i_sync(&p, STYPE_SYNC);
	uasm_i_ehb(&p);

	if (mips_cm_revision() < CM_REV_CM3) {
		/*
		* Disable all but self interventions. The load from COHCTL is
		* defined by the interAptiv & proAptiv SUMs as ensuring that the
		*  operation resulting from the preceding store is complete.
		*/
		uasm_i_addiu(&p, t0, zero, 1 << cpu_core(&cpu_data[cpu]));
		uasm_i_sw(&p, t0, 0, r_pcohctl);
		uasm_i_lw(&p, t0, 0, r_pcohctl);

		/* Barrier to ensure write to coherence control is complete */
		uasm_i_sync(&p, STYPE_SYNC);
		uasm_i_ehb(&p);
	}

	/* Disable coherence */
	uasm_i_sw(&p, zero, 0, r_pcohctl);
	uasm_i_lw(&p, t0, 0, r_pcohctl);

	if (state >= CPS_PM_CLOCK_GATED) {
		err = cps_gen_flush_fsb(&p, &l, &r, &cpu_data[cpu],
					lbl_flush_fsb);
		if (err)
			goto out_err;

		/* Determine the CPC command to issue */
		switch (state) {
		case CPS_PM_CLOCK_GATED:
			cpc_cmd = CPC_Cx_CMD_CLOCKOFF;
			break;
		case CPS_PM_POWER_GATED:
			cpc_cmd = CPC_Cx_CMD_PWRDOWN;
			break;
		default:
			BUG();
			goto out_err;
		}

		/* Issue the CPC command */
		UASM_i_LA(&p, t0, (long)addr_cpc_cl_cmd());
		uasm_i_addiu(&p, t1, zero, cpc_cmd);
		uasm_i_sw(&p, t1, 0, t0);

		if (state == CPS_PM_POWER_GATED) {
			/* If anything goes wrong just hang */
			uasm_build_label(&l, p, lbl_hang);
			uasm_il_b(&p, &r, lbl_hang);
			uasm_i_nop(&p);

			/*
			 * There's no point generating more code, the core is
			 * powered down & if powered back up will run from the
			 * reset vector not from here.
			 */
			goto gen_done;
		}

		/* Barrier to ensure write to CPC command is complete */
		uasm_i_sync(&p, STYPE_SYNC);
		uasm_i_ehb(&p);
	}

	if (state == CPS_PM_NC_WAIT) {
		/*
		 * At this point it is safe for all VPEs to proceed with
		 * execution. This VPE will set the top bit of ready_count
		 * to indicate to the other VPEs that they may continue.
		 */
		if (coupled_coherence)
			cps_gen_set_top_bit(&p, &l, &r, r_nc_count,
					    lbl_set_cont);

		/*
		 * VPEs which did not disable coherence will continue
		 * executing, after coherence has been disabled, from this
		 * point.
		 */
		uasm_build_label(&l, p, lbl_secondary_cont);

		/* Now perform our wait */
		uasm_i_wait(&p, 0);
	}

	/*
	 * Re-enable coherence. Note that for CPS_PM_NC_WAIT all coupled VPEs
	 * will run this. The first will actually re-enable coherence & the
	 * rest will just be performing a rather unusual nop.
	 */
	uasm_i_addiu(&p, t0, zero, mips_cm_revision() < CM_REV_CM3
				? CM_GCR_Cx_COHERENCE_COHDOMAINEN
				: CM3_GCR_Cx_COHERENCE_COHEN);

	uasm_i_sw(&p, t0, 0, r_pcohctl);
	uasm_i_lw(&p, t0, 0, r_pcohctl);

	/* Barrier to ensure write to coherence control is complete */
	uasm_i_sync(&p, STYPE_SYNC);
	uasm_i_ehb(&p);

	if (coupled_coherence && (state == CPS_PM_NC_WAIT)) {
		/* Decrement ready_count */
		uasm_build_label(&l, p, lbl_decready);
		uasm_i_sync(&p, STYPE_SYNC_MB);
		uasm_i_ll(&p, t1, 0, r_nc_count);
		uasm_i_addiu(&p, t2, t1, -1);
		uasm_i_sc(&p, t2, 0, r_nc_count);
		uasm_il_beqz(&p, &r, t2, lbl_decready);
		uasm_i_andi(&p, v0, t1, (1 << fls(smp_num_siblings)) - 1);

		/* Barrier ensuring all CPUs see the updated r_nc_count value */
		uasm_i_sync(&p, STYPE_SYNC_MB);
	}

	if (coupled_coherence && (state == CPS_PM_CLOCK_GATED)) {
		/*
		 * At this point it is safe for all VPEs to proceed with
		 * execution. This VPE will set the top bit of ready_count
		 * to indicate to the other VPEs that they may continue.
		 */
		cps_gen_set_top_bit(&p, &l, &r, r_nc_count, lbl_set_cont);

		/*
		 * This core will be reliant upon another core sending a
		 * power-up command to the CPC in order to resume operation.
		 * Thus an arbitrary VPE can't trigger the core leaving the
		 * idle state and the one that disables coherence might as well
		 * be the one to re-enable it. The rest will continue from here
		 * after that has been done.
		 */
		uasm_build_label(&l, p, lbl_secondary_cont);

		/* Barrier ensuring all CPUs see the updated r_nc_count value */
		uasm_i_sync(&p, STYPE_SYNC_MB);
	}

	/* The core is coherent, time to return to C code */
	uasm_i_jr(&p, ra);
	uasm_i_nop(&p);

gen_done:
	/* Ensure the code didn't exceed the resources allocated for it */
	BUG_ON((p - buf) > max_instrs);
	BUG_ON((l - labels) > ARRAY_SIZE(labels));
	BUG_ON((r - relocs) > ARRAY_SIZE(relocs));

	/* Patch branch offsets */
	uasm_resolve_relocs(relocs, labels);

	/* Flush the icache */
	local_flush_icache_range((unsigned long)buf, (unsigned long)p);

	return buf;
out_err:
	kfree(buf);
	return NULL;
}

static int cps_pm_online_cpu(unsigned int cpu)
{
	enum cps_pm_state state;
	unsigned core = cpu_core(&cpu_data[cpu]);
	void *entry_fn, *core_rc;

	for (state = CPS_PM_NC_WAIT; state < CPS_PM_STATE_COUNT; state++) {
		if (per_cpu(nc_asm_enter, core)[state])
			continue;
		if (!test_bit(state, state_support))
			continue;

		entry_fn = cps_gen_entry_code(cpu, state);
		if (!entry_fn) {
			pr_err("Failed to generate core %u state %u entry\n",
			       core, state);
			clear_bit(state, state_support);
		}

		per_cpu(nc_asm_enter, core)[state] = entry_fn;
	}

	if (!per_cpu(ready_count, core)) {
		core_rc = kmalloc(sizeof(u32), GFP_KERNEL);
		if (!core_rc) {
			pr_err("Failed allocate core %u ready_count\n", core);
			return -ENOMEM;
		}
		per_cpu(ready_count, core) = core_rc;
	}

	return 0;
}

static int __init cps_pm_init(void)
{
	/* A CM is required for all non-coherent states */
	if (!mips_cm_present()) {
		pr_warn("pm-cps: no CM, non-coherent states unavailable\n");
		return 0;
	}

	/*
	 * If interrupts were enabled whilst running a wait instruction on a
	 * non-coherent core then the VPE may end up processing interrupts
	 * whilst non-coherent. That would be bad.
	 */
	if (cpu_wait == r4k_wait_irqoff)
		set_bit(CPS_PM_NC_WAIT, state_support);
	else
		pr_warn("pm-cps: non-coherent wait unavailable\n");

	/* Detect whether a CPC is present */
	if (mips_cpc_present()) {
		/* Detect whether clock gating is implemented */
		if (read_cpc_cl_stat_conf() & CPC_Cx_STAT_CONF_CLKGAT_IMPL)
			set_bit(CPS_PM_CLOCK_GATED, state_support);
		else
			pr_warn("pm-cps: CPC does not support clock gating\n");

		/* Power gating is available with CPS SMP & any CPC */
		if (mips_cps_smp_in_use())
			set_bit(CPS_PM_POWER_GATED, state_support);
		else
			pr_warn("pm-cps: CPS SMP not in use, power gating unavailable\n");
	} else {
		pr_warn("pm-cps: no CPC, clock & power gating unavailable\n");
	}

	return cpuhp_setup_state(CPUHP_AP_ONLINE_DYN, "mips/cps_pm:online",
				 cps_pm_online_cpu, NULL);
}
arch_initcall(cps_pm_init);