coresight-etm4x.c 32.0 KB
Newer Older
1 2 3
// SPDX-License-Identifier: GPL-2.0
/*
 * Copyright (c) 2014, The Linux Foundation. All rights reserved.
4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21
 */

#include <linux/kernel.h>
#include <linux/moduleparam.h>
#include <linux/init.h>
#include <linux/types.h>
#include <linux/device.h>
#include <linux/io.h>
#include <linux/err.h>
#include <linux/fs.h>
#include <linux/slab.h>
#include <linux/delay.h>
#include <linux/smp.h>
#include <linux/sysfs.h>
#include <linux/stat.h>
#include <linux/clk.h>
#include <linux/cpu.h>
#include <linux/coresight.h>
22
#include <linux/coresight-pmu.h>
23 24 25 26
#include <linux/pm_wakeup.h>
#include <linux/amba/bus.h>
#include <linux/seq_file.h>
#include <linux/uaccess.h>
27
#include <linux/perf_event.h>
28 29
#include <linux/pm_runtime.h>
#include <asm/sections.h>
30
#include <asm/local.h>
31
#include <asm/virt.h>
32 33

#include "coresight-etm4x.h"
34
#include "coresight-etm-perf.h"
35 36 37 38 39 40 41

static int boot_enable;
module_param_named(boot_enable, boot_enable, int, S_IRUGO);

/* The number of ETMv4 currently registered */
static int etm4_count;
static struct etmv4_drvdata *etmdrvdata[NR_CPUS];
42 43 44
static void etm4_set_default_config(struct etmv4_config *config);
static int etm4_set_event_filters(struct etmv4_drvdata *drvdata,
				  struct perf_event *event);
45

46 47
static enum cpuhp_state hp_online;

48
static void etm4_os_unlock(struct etmv4_drvdata *drvdata)
49 50 51
{
	/* Writing any value to ETMOSLAR unlocks the trace registers */
	writel_relaxed(0x0, drvdata->base + TRCOSLAR);
52
	drvdata->os_unlock = true;
53 54 55 56 57
	isb();
}

static bool etm4_arch_supported(u8 arch)
{
58 59
	/* Mask out the minor version number */
	switch (arch & 0xf0) {
60 61 62 63 64 65 66 67
	case ETM_ARCH_V4:
		break;
	default:
		return false;
	}
	return true;
}

68 69 70 71 72 73 74
static int etm4_cpu_id(struct coresight_device *csdev)
{
	struct etmv4_drvdata *drvdata = dev_get_drvdata(csdev->dev.parent);

	return drvdata->cpu;
}

75 76 77 78
static int etm4_trace_id(struct coresight_device *csdev)
{
	struct etmv4_drvdata *drvdata = dev_get_drvdata(csdev->dev.parent);

79
	return drvdata->trcid;
80 81
}

82 83 84 85 86 87
struct etm4_enable_arg {
	struct etmv4_drvdata *drvdata;
	int rc;
};

static int etm4_enable_hw(struct etmv4_drvdata *drvdata)
88
{
89
	int i, rc;
90
	struct etmv4_config *config = &drvdata->config;
91 92 93 94 95

	CS_UNLOCK(drvdata->base);

	etm4_os_unlock(drvdata);

96 97 98 99
	rc = coresight_claim_device_unlocked(drvdata->base);
	if (rc)
		goto done;

100 101 102 103 104 105
	/* Disable the trace unit before programming trace registers */
	writel_relaxed(0, drvdata->base + TRCPRGCTLR);

	/* wait for TRCSTATR.IDLE to go up */
	if (coresight_timeout(drvdata->base, TRCSTATR, TRCSTATR_IDLE_BIT, 1))
		dev_err(drvdata->dev,
106
			"timeout while waiting for Idle Trace Status\n");
107

108 109
	writel_relaxed(config->pe_sel, drvdata->base + TRCPROCSELR);
	writel_relaxed(config->cfg, drvdata->base + TRCCONFIGR);
110 111
	/* nothing specific implemented */
	writel_relaxed(0x0, drvdata->base + TRCAUXCTLR);
112 113 114 115 116 117 118
	writel_relaxed(config->eventctrl0, drvdata->base + TRCEVENTCTL0R);
	writel_relaxed(config->eventctrl1, drvdata->base + TRCEVENTCTL1R);
	writel_relaxed(config->stall_ctrl, drvdata->base + TRCSTALLCTLR);
	writel_relaxed(config->ts_ctrl, drvdata->base + TRCTSCTLR);
	writel_relaxed(config->syncfreq, drvdata->base + TRCSYNCPR);
	writel_relaxed(config->ccctlr, drvdata->base + TRCCCCTLR);
	writel_relaxed(config->bb_ctrl, drvdata->base + TRCBBCTLR);
119
	writel_relaxed(drvdata->trcid, drvdata->base + TRCTRACEIDR);
120 121 122
	writel_relaxed(config->vinst_ctrl, drvdata->base + TRCVICTLR);
	writel_relaxed(config->viiectlr, drvdata->base + TRCVIIECTLR);
	writel_relaxed(config->vissctlr,
123
		       drvdata->base + TRCVISSCTLR);
124
	writel_relaxed(config->vipcssctlr,
125 126
		       drvdata->base + TRCVIPCSSCTLR);
	for (i = 0; i < drvdata->nrseqstate - 1; i++)
127
		writel_relaxed(config->seq_ctrl[i],
128
			       drvdata->base + TRCSEQEVRn(i));
129 130 131
	writel_relaxed(config->seq_rst, drvdata->base + TRCSEQRSTEVR);
	writel_relaxed(config->seq_state, drvdata->base + TRCSEQSTR);
	writel_relaxed(config->ext_inp, drvdata->base + TRCEXTINSELR);
132
	for (i = 0; i < drvdata->nr_cntr; i++) {
133
		writel_relaxed(config->cntrldvr[i],
134
			       drvdata->base + TRCCNTRLDVRn(i));
135
		writel_relaxed(config->cntr_ctrl[i],
136
			       drvdata->base + TRCCNTCTLRn(i));
137
		writel_relaxed(config->cntr_val[i],
138 139
			       drvdata->base + TRCCNTVRn(i));
	}
140

141 142 143 144 145
	/*
	 * Resource selector pair 0 is always implemented and reserved.  As
	 * such start at 2.
	 */
	for (i = 2; i < drvdata->nr_resource * 2; i++)
146
		writel_relaxed(config->res_ctrl[i],
147 148 149
			       drvdata->base + TRCRSCTLRn(i));

	for (i = 0; i < drvdata->nr_ss_cmp; i++) {
150
		writel_relaxed(config->ss_ctrl[i],
151
			       drvdata->base + TRCSSCCRn(i));
152
		writel_relaxed(config->ss_status[i],
153
			       drvdata->base + TRCSSCSRn(i));
154
		writel_relaxed(config->ss_pe_cmp[i],
155 156 157
			       drvdata->base + TRCSSPCICRn(i));
	}
	for (i = 0; i < drvdata->nr_addr_cmp; i++) {
158
		writeq_relaxed(config->addr_val[i],
159
			       drvdata->base + TRCACVRn(i));
160
		writeq_relaxed(config->addr_acc[i],
161 162 163
			       drvdata->base + TRCACATRn(i));
	}
	for (i = 0; i < drvdata->numcidc; i++)
164
		writeq_relaxed(config->ctxid_pid[i],
165
			       drvdata->base + TRCCIDCVRn(i));
166 167
	writel_relaxed(config->ctxid_mask0, drvdata->base + TRCCIDCCTLR0);
	writel_relaxed(config->ctxid_mask1, drvdata->base + TRCCIDCCTLR1);
168 169

	for (i = 0; i < drvdata->numvmidc; i++)
170
		writeq_relaxed(config->vmid_val[i],
171
			       drvdata->base + TRCVMIDCVRn(i));
172 173
	writel_relaxed(config->vmid_mask0, drvdata->base + TRCVMIDCCTLR0);
	writel_relaxed(config->vmid_mask1, drvdata->base + TRCVMIDCCTLR1);
174

175 176 177 178 179 180 181
	/*
	 * Request to keep the trace unit powered and also
	 * emulation of powerdown
	 */
	writel_relaxed(readl_relaxed(drvdata->base + TRCPDCR) | TRCPDCR_PU,
		       drvdata->base + TRCPDCR);

182 183 184 185 186 187
	/* Enable the trace unit */
	writel_relaxed(1, drvdata->base + TRCPRGCTLR);

	/* wait for TRCSTATR.IDLE to go back down to '0' */
	if (coresight_timeout(drvdata->base, TRCSTATR, TRCSTATR_IDLE_BIT, 0))
		dev_err(drvdata->dev,
188
			"timeout while waiting for Idle Trace Status\n");
189

190
done:
191 192
	CS_LOCK(drvdata->base);

193 194 195
	dev_dbg(drvdata->dev, "cpu: %d enable smp call done: %d\n",
		drvdata->cpu, rc);
	return rc;
196 197 198 199 200 201 202 203 204
}

static void etm4_enable_hw_smp_call(void *info)
{
	struct etm4_enable_arg *arg = info;

	if (WARN_ON(!arg))
		return;
	arg->rc = etm4_enable_hw(arg->drvdata);
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
/*
 * The goal of function etm4_config_timestamp_event() is to configure a
 * counter that will tell the tracer to emit a timestamp packet when it
 * reaches zero.  This is done in order to get a more fine grained idea
 * of when instructions are executed so that they can be correlated
 * with execution on other CPUs.
 *
 * To do this the counter itself is configured to self reload and
 * TRCRSCTLR1 (always true) used to get the counter to decrement.  From
 * there a resource selector is configured with the counter and the
 * timestamp control register to use the resource selector to trigger the
 * event that will insert a timestamp packet in the stream.
 */
static int etm4_config_timestamp_event(struct etmv4_drvdata *drvdata)
{
	int ctridx, ret = -EINVAL;
	int counter, rselector;
	u32 val = 0;
	struct etmv4_config *config = &drvdata->config;

	/* No point in trying if we don't have at least one counter */
	if (!drvdata->nr_cntr)
		goto out;

	/* Find a counter that hasn't been initialised */
	for (ctridx = 0; ctridx < drvdata->nr_cntr; ctridx++)
		if (config->cntr_val[ctridx] == 0)
			break;

	/* All the counters have been configured already, bail out */
	if (ctridx == drvdata->nr_cntr) {
		pr_debug("%s: no available counter found\n", __func__);
		ret = -ENOSPC;
		goto out;
	}

	/*
	 * Searching for an available resource selector to use, starting at
	 * '2' since every implementation has at least 2 resource selector.
	 * ETMIDR4 gives the number of resource selector _pairs_,
	 * hence multiply by 2.
	 */
	for (rselector = 2; rselector < drvdata->nr_resource * 2; rselector++)
		if (!config->res_ctrl[rselector])
			break;

	if (rselector == drvdata->nr_resource * 2) {
		pr_debug("%s: no available resource selector found\n",
			 __func__);
		ret = -ENOSPC;
		goto out;
	}

	/* Remember what counter we used */
	counter = 1 << ctridx;

	/*
	 * Initialise original and reload counter value to the smallest
	 * possible value in order to get as much precision as we can.
	 */
	config->cntr_val[ctridx] = 1;
	config->cntrldvr[ctridx] = 1;

	/* Set the trace counter control register */
	val =  0x1 << 16	|  /* Bit 16, reload counter automatically */
	       0x0 << 7		|  /* Select single resource selector */
	       0x1;		   /* Resource selector 1, i.e always true */

	config->cntr_ctrl[ctridx] = val;

	val = 0x2 << 16		| /* Group 0b0010 - Counter and sequencers */
	      counter << 0;	  /* Counter to use */

	config->res_ctrl[rselector] = val;

	val = 0x0 << 7		| /* Select single resource selector */
	      rselector;	  /* Resource selector */

	config->ts_ctrl = val;

	ret = 0;
out:
	return ret;
}

292
static int etm4_parse_event_config(struct etmv4_drvdata *drvdata,
293
				   struct perf_event *event)
294
{
295
	int ret = 0;
296
	struct etmv4_config *config = &drvdata->config;
297
	struct perf_event_attr *attr = &event->attr;
298

299 300 301 302
	if (!attr) {
		ret = -EINVAL;
		goto out;
	}
303 304 305 306 307 308 309 310 311 312 313

	/* Clear configuration from previous run */
	memset(config, 0, sizeof(struct etmv4_config));

	if (attr->exclude_kernel)
		config->mode = ETM_MODE_EXCL_KERN;

	if (attr->exclude_user)
		config->mode = ETM_MODE_EXCL_USER;

	/* Always start from the default config */
314 315 316 317 318 319
	etm4_set_default_config(config);

	/* Configure filters specified on the perf cmd line, if any. */
	ret = etm4_set_event_filters(drvdata, event);
	if (ret)
		goto out;
320 321

	/* Go from generic option to ETMv4 specifics */
322 323 324 325 326
	if (attr->config & BIT(ETM_OPT_CYCACC)) {
		config->cfg |= BIT(4);
		/* TRM: Must program this for cycacc to work */
		config->ccctlr = ETM_CYC_THRESHOLD_DEFAULT;
	}
327 328 329 330 331 332 333 334 335 336 337 338 339 340 341
	if (attr->config & BIT(ETM_OPT_TS)) {
		/*
		 * Configure timestamps to be emitted at regular intervals in
		 * order to correlate instructions executed on different CPUs
		 * (CPU-wide trace scenarios).
		 */
		ret = etm4_config_timestamp_event(drvdata);

		/*
		 * No need to go further if timestamp intervals can't
		 * be configured.
		 */
		if (ret)
			goto out;

342 343
		/* bit[11], Global timestamp tracing bit */
		config->cfg |= BIT(11);
344
	}
345 346 347 348 349

	if (attr->config & BIT(ETM_OPT_CTXTID))
		/* bit[6], Context ID tracing bit */
		config->cfg |= BIT(ETM4_CFG_BIT_CTXTID);

350 351 352 353
	/* return stack - enable if selected and supported */
	if ((attr->config & BIT(ETM_OPT_RETSTK)) && drvdata->retstack)
		/* bit[12], Return stack enable bit */
		config->cfg |= BIT(12);
354

355 356
out:
	return ret;
357 358 359
}

static int etm4_enable_perf(struct coresight_device *csdev,
360
			    struct perf_event *event)
361
{
362
	int ret = 0;
363 364
	struct etmv4_drvdata *drvdata = dev_get_drvdata(csdev->dev.parent);

365 366 367 368
	if (WARN_ON_ONCE(drvdata->cpu != smp_processor_id())) {
		ret = -EINVAL;
		goto out;
	}
369 370

	/* Configure the tracer based on the session's specifics */
371 372 373
	ret = etm4_parse_event_config(drvdata, event);
	if (ret)
		goto out;
374
	/* And enable it */
375
	ret = etm4_enable_hw(drvdata);
376

377 378
out:
	return ret;
379 380
}

381
static int etm4_enable_sysfs(struct coresight_device *csdev)
382 383
{
	struct etmv4_drvdata *drvdata = dev_get_drvdata(csdev->dev.parent);
384
	struct etm4_enable_arg arg = { 0 };
385 386 387 388 389 390 391 392
	int ret;

	spin_lock(&drvdata->spinlock);

	/*
	 * Executing etm4_enable_hw on the cpu whose ETM is being enabled
	 * ensures that register writes occur when cpu is powered.
	 */
393
	arg.drvdata = drvdata;
394
	ret = smp_call_function_single(drvdata->cpu,
395 396 397 398 399
				       etm4_enable_hw_smp_call, &arg, 1);
	if (!ret)
		ret = arg.rc;
	if (!ret)
		drvdata->sticky_enable = true;
400 401
	spin_unlock(&drvdata->spinlock);

402 403
	if (!ret)
		dev_dbg(drvdata->dev, "ETM tracing enabled\n");
404 405 406
	return ret;
}

407
static int etm4_enable(struct coresight_device *csdev,
408
		       struct perf_event *event, u32 mode)
409 410 411 412 413 414 415 416 417 418 419 420 421 422 423
{
	int ret;
	u32 val;
	struct etmv4_drvdata *drvdata = dev_get_drvdata(csdev->dev.parent);

	val = local_cmpxchg(&drvdata->mode, CS_MODE_DISABLED, mode);

	/* Someone is already using the tracer */
	if (val)
		return -EBUSY;

	switch (mode) {
	case CS_MODE_SYSFS:
		ret = etm4_enable_sysfs(csdev);
		break;
424
	case CS_MODE_PERF:
425
		ret = etm4_enable_perf(csdev, event);
426
		break;
427 428 429 430 431 432 433 434 435 436 437
	default:
		ret = -EINVAL;
	}

	/* The tracer didn't start */
	if (ret)
		local_set(&drvdata->mode, CS_MODE_DISABLED);

	return ret;
}

438 439 440 441 442 443 444
static void etm4_disable_hw(void *info)
{
	u32 control;
	struct etmv4_drvdata *drvdata = info;

	CS_UNLOCK(drvdata->base);

445 446 447 448 449
	/* power can be removed from the trace unit now */
	control = readl_relaxed(drvdata->base + TRCPDCR);
	control &= ~TRCPDCR_PU;
	writel_relaxed(control, drvdata->base + TRCPDCR);

450 451 452 453 454 455 456 457 458 459
	control = readl_relaxed(drvdata->base + TRCPRGCTLR);

	/* EN, bit[0] Trace unit enable bit */
	control &= ~0x1;

	/* make sure everything completes before disabling */
	mb();
	isb();
	writel_relaxed(control, drvdata->base + TRCPRGCTLR);

460 461
	coresight_disclaim_device_unlocked(drvdata->base);

462 463 464 465 466
	CS_LOCK(drvdata->base);

	dev_dbg(drvdata->dev, "cpu: %d disable smp call done\n", drvdata->cpu);
}

467 468
static int etm4_disable_perf(struct coresight_device *csdev,
			     struct perf_event *event)
469
{
470 471
	u32 control;
	struct etm_filters *filters = event->hw.addr_filters;
472 473 474 475 476 477
	struct etmv4_drvdata *drvdata = dev_get_drvdata(csdev->dev.parent);

	if (WARN_ON_ONCE(drvdata->cpu != smp_processor_id()))
		return -EINVAL;

	etm4_disable_hw(drvdata);
478 479 480 481 482 483 484 485 486 487 488

	/*
	 * Check if the start/stop logic was active when the unit was stopped.
	 * That way we can re-enable the start/stop logic when the process is
	 * scheduled again.  Configuration of the start/stop logic happens in
	 * function etm4_set_event_filters().
	 */
	control = readl_relaxed(drvdata->base + TRCVICTLR);
	/* TRCVICTLR::SSSTATUS, bit[9] */
	filters->ssstatus = (control & BIT(9));

489 490 491
	return 0;
}

492
static void etm4_disable_sysfs(struct coresight_device *csdev)
493 494 495 496 497 498 499 500 501
{
	struct etmv4_drvdata *drvdata = dev_get_drvdata(csdev->dev.parent);

	/*
	 * Taking hotplug lock here protects from clocks getting disabled
	 * with tracing being left on (crash scenario) if user disable occurs
	 * after cpu online mask indicates the cpu is offline but before the
	 * DYING hotplug callback is serviced by the ETM driver.
	 */
502
	cpus_read_lock();
503 504 505 506 507 508 509 510 511
	spin_lock(&drvdata->spinlock);

	/*
	 * Executing etm4_disable_hw on the cpu whose ETM is being disabled
	 * ensures that register writes occur when cpu is powered.
	 */
	smp_call_function_single(drvdata->cpu, etm4_disable_hw, drvdata, 1);

	spin_unlock(&drvdata->spinlock);
512
	cpus_read_unlock();
513

514
	dev_dbg(drvdata->dev, "ETM tracing disabled\n");
515 516
}

517 518
static void etm4_disable(struct coresight_device *csdev,
			 struct perf_event *event)
519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535
{
	u32 mode;
	struct etmv4_drvdata *drvdata = dev_get_drvdata(csdev->dev.parent);

	/*
	 * For as long as the tracer isn't disabled another entity can't
	 * change its status.  As such we can read the status here without
	 * fearing it will change under us.
	 */
	mode = local_read(&drvdata->mode);

	switch (mode) {
	case CS_MODE_DISABLED:
		break;
	case CS_MODE_SYSFS:
		etm4_disable_sysfs(csdev);
		break;
536
	case CS_MODE_PERF:
537
		etm4_disable_perf(csdev, event);
538
		break;
539 540 541 542 543 544
	}

	if (mode)
		local_set(&drvdata->mode, CS_MODE_DISABLED);
}

545
static const struct coresight_ops_source etm4_source_ops = {
546
	.cpu_id		= etm4_cpu_id,
547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565
	.trace_id	= etm4_trace_id,
	.enable		= etm4_enable,
	.disable	= etm4_disable,
};

static const struct coresight_ops etm4_cs_ops = {
	.source_ops	= &etm4_source_ops,
};

static void etm4_init_arch_data(void *info)
{
	u32 etmidr0;
	u32 etmidr1;
	u32 etmidr2;
	u32 etmidr3;
	u32 etmidr4;
	u32 etmidr5;
	struct etmv4_drvdata *drvdata = info;

566 567 568
	/* Make sure all registers are accessible */
	etm4_os_unlock(drvdata);

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
	CS_UNLOCK(drvdata->base);

	/* find all capabilities of the tracing unit */
	etmidr0 = readl_relaxed(drvdata->base + TRCIDR0);

	/* INSTP0, bits[2:1] P0 tracing support field */
	if (BMVAL(etmidr0, 1, 1) && BMVAL(etmidr0, 2, 2))
		drvdata->instrp0 = true;
	else
		drvdata->instrp0 = false;

	/* TRCBB, bit[5] Branch broadcast tracing support bit */
	if (BMVAL(etmidr0, 5, 5))
		drvdata->trcbb = true;
	else
		drvdata->trcbb = false;

	/* TRCCOND, bit[6] Conditional instruction tracing support bit */
	if (BMVAL(etmidr0, 6, 6))
		drvdata->trccond = true;
	else
		drvdata->trccond = false;

	/* TRCCCI, bit[7] Cycle counting instruction bit */
	if (BMVAL(etmidr0, 7, 7))
		drvdata->trccci = true;
	else
		drvdata->trccci = false;

	/* RETSTACK, bit[9] Return stack bit */
	if (BMVAL(etmidr0, 9, 9))
		drvdata->retstack = true;
	else
		drvdata->retstack = false;

	/* NUMEVENT, bits[11:10] Number of events field */
	drvdata->nr_event = BMVAL(etmidr0, 10, 11);
	/* QSUPP, bits[16:15] Q element support field */
	drvdata->q_support = BMVAL(etmidr0, 15, 16);
	/* TSSIZE, bits[28:24] Global timestamp size field */
	drvdata->ts_size = BMVAL(etmidr0, 24, 28);

	/* base architecture of trace unit */
	etmidr1 = readl_relaxed(drvdata->base + TRCIDR1);
	/*
	 * TRCARCHMIN, bits[7:4] architecture the minor version number
	 * TRCARCHMAJ, bits[11:8] architecture major versin number
	 */
	drvdata->arch = BMVAL(etmidr1, 4, 11);

	/* maximum size of resources */
	etmidr2 = readl_relaxed(drvdata->base + TRCIDR2);
	/* CIDSIZE, bits[9:5] Indicates the Context ID size */
	drvdata->ctxid_size = BMVAL(etmidr2, 5, 9);
	/* VMIDSIZE, bits[14:10] Indicates the VMID size */
	drvdata->vmid_size = BMVAL(etmidr2, 10, 14);
	/* CCSIZE, bits[28:25] size of the cycle counter in bits minus 12 */
	drvdata->ccsize = BMVAL(etmidr2, 25, 28);

	etmidr3 = readl_relaxed(drvdata->base + TRCIDR3);
	/* CCITMIN, bits[11:0] minimum threshold value that can be programmed */
	drvdata->ccitmin = BMVAL(etmidr3, 0, 11);
	/* EXLEVEL_S, bits[19:16] Secure state instruction tracing */
	drvdata->s_ex_level = BMVAL(etmidr3, 16, 19);
	/* EXLEVEL_NS, bits[23:20] Non-secure state instruction tracing */
	drvdata->ns_ex_level = BMVAL(etmidr3, 20, 23);

	/*
	 * TRCERR, bit[24] whether a trace unit can trace a
	 * system error exception.
	 */
	if (BMVAL(etmidr3, 24, 24))
		drvdata->trc_error = true;
	else
		drvdata->trc_error = false;

	/* SYNCPR, bit[25] implementation has a fixed synchronization period? */
	if (BMVAL(etmidr3, 25, 25))
		drvdata->syncpr = true;
	else
		drvdata->syncpr = false;

	/* STALLCTL, bit[26] is stall control implemented? */
	if (BMVAL(etmidr3, 26, 26))
		drvdata->stallctl = true;
	else
		drvdata->stallctl = false;

	/* SYSSTALL, bit[27] implementation can support stall control? */
	if (BMVAL(etmidr3, 27, 27))
		drvdata->sysstall = true;
	else
		drvdata->sysstall = false;

	/* NUMPROC, bits[30:28] the number of PEs available for tracing */
	drvdata->nr_pe = BMVAL(etmidr3, 28, 30);

	/* NOOVERFLOW, bit[31] is trace overflow prevention supported */
	if (BMVAL(etmidr3, 31, 31))
		drvdata->nooverflow = true;
	else
		drvdata->nooverflow = false;

	/* number of resources trace unit supports */
	etmidr4 = readl_relaxed(drvdata->base + TRCIDR4);
	/* NUMACPAIRS, bits[0:3] number of addr comparator pairs for tracing */
	drvdata->nr_addr_cmp = BMVAL(etmidr4, 0, 3);
	/* NUMPC, bits[15:12] number of PE comparator inputs for tracing */
	drvdata->nr_pe_cmp = BMVAL(etmidr4, 12, 15);
678 679 680 681 682 683 684
	/*
	 * NUMRSPAIR, bits[19:16]
	 * The number of resource pairs conveyed by the HW starts at 0, i.e a
	 * value of 0x0 indicate 1 resource pair, 0x1 indicate two and so on.
	 * As such add 1 to the value of NUMRSPAIR for a better representation.
	 */
	drvdata->nr_resource = BMVAL(etmidr4, 16, 19) + 1;
685 686 687 688 689 690 691 692 693 694 695 696 697 698 699 700 701 702 703 704 705 706 707 708 709 710 711 712 713 714 715 716 717 718 719
	/*
	 * NUMSSCC, bits[23:20] the number of single-shot
	 * comparator control for tracing
	 */
	drvdata->nr_ss_cmp = BMVAL(etmidr4, 20, 23);
	/* NUMCIDC, bits[27:24] number of Context ID comparators for tracing */
	drvdata->numcidc = BMVAL(etmidr4, 24, 27);
	/* NUMVMIDC, bits[31:28] number of VMID comparators for tracing */
	drvdata->numvmidc = BMVAL(etmidr4, 28, 31);

	etmidr5 = readl_relaxed(drvdata->base + TRCIDR5);
	/* NUMEXTIN, bits[8:0] number of external inputs implemented */
	drvdata->nr_ext_inp = BMVAL(etmidr5, 0, 8);
	/* TRACEIDSIZE, bits[21:16] indicates the trace ID width */
	drvdata->trcid_size = BMVAL(etmidr5, 16, 21);
	/* ATBTRIG, bit[22] implementation can support ATB triggers? */
	if (BMVAL(etmidr5, 22, 22))
		drvdata->atbtrig = true;
	else
		drvdata->atbtrig = false;
	/*
	 * LPOVERRIDE, bit[23] implementation supports
	 * low-power state override
	 */
	if (BMVAL(etmidr5, 23, 23))
		drvdata->lpoverride = true;
	else
		drvdata->lpoverride = false;
	/* NUMSEQSTATE, bits[27:25] number of sequencer states implemented */
	drvdata->nrseqstate = BMVAL(etmidr5, 25, 27);
	/* NUMCNTR, bits[30:28] number of counters available for tracing */
	drvdata->nr_cntr = BMVAL(etmidr5, 28, 30);
	CS_LOCK(drvdata->base);
}

720
static void etm4_set_default_config(struct etmv4_config *config)
721 722
{
	/* disable all events tracing */
723 724
	config->eventctrl0 = 0x0;
	config->eventctrl1 = 0x0;
725 726

	/* disable stalling */
727
	config->stall_ctrl = 0x0;
728

729 730 731
	/* enable trace synchronization every 4096 bytes, if available */
	config->syncfreq = 0xC;

732
	/* disable timestamp event */
733
	config->ts_ctrl = 0x0;
734

735 736
	/* TRCVICTLR::EVENT = 0x01, select the always on logic */
	config->vinst_ctrl |= BIT(0);
737
}
738

739
static u64 etm4_get_ns_access_type(struct etmv4_config *config)
740
{
741
	u64 access_type = 0;
742

743 744 745 746 747 748 749 750
	/*
	 * EXLEVEL_NS, bits[15:12]
	 * The Exception levels are:
	 *   Bit[12] Exception level 0 - Application
	 *   Bit[13] Exception level 1 - OS
	 *   Bit[14] Exception level 2 - Hypervisor
	 *   Bit[15] Never implemented
	 */
751 752 753 754 755 756 757 758
	if (!is_kernel_in_hyp_mode()) {
		/* Stay away from hypervisor mode for non-VHE */
		access_type =  ETM_EXLEVEL_NS_HYP;
		if (config->mode & ETM_MODE_EXCL_KERN)
			access_type |= ETM_EXLEVEL_NS_OS;
	} else if (config->mode & ETM_MODE_EXCL_KERN) {
		access_type = ETM_EXLEVEL_NS_HYP;
	}
759 760 761 762

	if (config->mode & ETM_MODE_EXCL_USER)
		access_type |= ETM_EXLEVEL_NS_APP;

763 764 765 766 767 768 769
	return access_type;
}

static u64 etm4_get_access_type(struct etmv4_config *config)
{
	u64 access_type = etm4_get_ns_access_type(config);

770
	/*
771 772 773 774 775 776 777
	 * EXLEVEL_S, bits[11:8], don't trace anything happening
	 * in secure state.
	 */
	access_type |= (ETM_EXLEVEL_S_APP	|
			ETM_EXLEVEL_S_OS	|
			ETM_EXLEVEL_S_HYP);

778 779 780 781 782 783 784 785
	return access_type;
}

static void etm4_set_comparator_filter(struct etmv4_config *config,
				       u64 start, u64 stop, int comparator)
{
	u64 access_type = etm4_get_access_type(config);

786
	/* First half of default address comparator */
787 788 789
	config->addr_val[comparator] = start;
	config->addr_acc[comparator] = access_type;
	config->addr_type[comparator] = ETM_ADDR_TYPE_RANGE;
790 791

	/* Second half of default address comparator */
792 793 794 795 796 797 798 799 800 801 802 803 804 805 806 807 808 809 810 811 812 813
	config->addr_val[comparator + 1] = stop;
	config->addr_acc[comparator + 1] = access_type;
	config->addr_type[comparator + 1] = ETM_ADDR_TYPE_RANGE;

	/*
	 * Configure the ViewInst function to include this address range
	 * comparator.
	 *
	 * @comparator is divided by two since it is the index in the
	 * etmv4_config::addr_val array but register TRCVIIECTLR deals with
	 * address range comparator _pairs_.
	 *
	 * Therefore:
	 *	index 0 -> compatator pair 0
	 *	index 2 -> comparator pair 1
	 *	index 4 -> comparator pair 2
	 *	...
	 *	index 14 -> comparator pair 7
	 */
	config->viiectlr |= BIT(comparator / 2);
}

814 815 816 817 818 819 820 821 822 823 824 825 826 827 828 829 830 831 832 833 834
static void etm4_set_start_stop_filter(struct etmv4_config *config,
				       u64 address, int comparator,
				       enum etm_addr_type type)
{
	int shift;
	u64 access_type = etm4_get_access_type(config);

	/* Configure the comparator */
	config->addr_val[comparator] = address;
	config->addr_acc[comparator] = access_type;
	config->addr_type[comparator] = type;

	/*
	 * Configure ViewInst Start-Stop control register.
	 * Addresses configured to start tracing go from bit 0 to n-1,
	 * while those configured to stop tracing from 16 to 16 + n-1.
	 */
	shift = (type == ETM_ADDR_TYPE_START ? 0 : 16);
	config->vissctlr |= BIT(shift + comparator);
}

835 836 837
static void etm4_set_default_filter(struct etmv4_config *config)
{
	u64 start, stop;
838 839

	/*
840 841
	 * Configure address range comparator '0' to encompass all
	 * possible addresses.
842
	 */
843 844 845 846 847
	start = 0x0;
	stop = ~0x0;

	etm4_set_comparator_filter(config, start, stop,
				   ETM_DEFAULT_ADDR_COMP);
848

849 850 851 852 853 854 855
	/*
	 * TRCVICTLR::SSSTATUS == 1, the start-stop logic is
	 * in the started state
	 */
	config->vinst_ctrl |= BIT(9);

	/* No start-stop filtering for ViewInst */
856
	config->vissctlr = 0x0;
857 858
}

859 860 861 862 863 864 865 866 867 868 869 870 871 872 873 874 875 876
static void etm4_set_default(struct etmv4_config *config)
{
	if (WARN_ON_ONCE(!config))
		return;

	/*
	 * Make default initialisation trace everything
	 *
	 * Select the "always true" resource selector on the
	 * "Enablign Event" line and configure address range comparator
	 * '0' to trace all the possible address range.  From there
	 * configure the "include/exclude" engine to include address
	 * range comparator '0'.
	 */
	etm4_set_default_config(config);
	etm4_set_default_filter(config);
}

877 878 879 880 881 882 883 884 885 886 887 888 889 890 891 892 893 894 895 896 897 898
static int etm4_get_next_comparator(struct etmv4_drvdata *drvdata, u32 type)
{
	int nr_comparator, index = 0;
	struct etmv4_config *config = &drvdata->config;

	/*
	 * nr_addr_cmp holds the number of comparator _pair_, so time 2
	 * for the total number of comparators.
	 */
	nr_comparator = drvdata->nr_addr_cmp * 2;

	/* Go through the tally of comparators looking for a free one. */
	while (index < nr_comparator) {
		switch (type) {
		case ETM_ADDR_TYPE_RANGE:
			if (config->addr_type[index] == ETM_ADDR_TYPE_NONE &&
			    config->addr_type[index + 1] == ETM_ADDR_TYPE_NONE)
				return index;

			/* Address range comparators go in pairs */
			index += 2;
			break;
899 900 901 902 903 904 905 906
		case ETM_ADDR_TYPE_START:
		case ETM_ADDR_TYPE_STOP:
			if (config->addr_type[index] == ETM_ADDR_TYPE_NONE)
				return index;

			/* Start/stop address can have odd indexes */
			index += 1;
			break;
907 908 909 910 911 912 913 914 915 916 917 918 919
		default:
			return -EINVAL;
		}
	}

	/* If we are here all the comparators have been used. */
	return -ENOSPC;
}

static int etm4_set_event_filters(struct etmv4_drvdata *drvdata,
				  struct perf_event *event)
{
	int i, comparator, ret = 0;
920
	u64 address;
921 922 923 924 925 926 927 928 929 930 931 932 933 934 935 936 937 938 939 940 941 942 943 944 945 946 947 948 949 950 951 952 953 954 955 956 957 958 959 960 961 962
	struct etmv4_config *config = &drvdata->config;
	struct etm_filters *filters = event->hw.addr_filters;

	if (!filters)
		goto default_filter;

	/* Sync events with what Perf got */
	perf_event_addr_filters_sync(event);

	/*
	 * If there are no filters to deal with simply go ahead with
	 * the default filter, i.e the entire address range.
	 */
	if (!filters->nr_filters)
		goto default_filter;

	for (i = 0; i < filters->nr_filters; i++) {
		struct etm_filter *filter = &filters->etm_filter[i];
		enum etm_addr_type type = filter->type;

		/* See if a comparator is free. */
		comparator = etm4_get_next_comparator(drvdata, type);
		if (comparator < 0) {
			ret = comparator;
			goto out;
		}

		switch (type) {
		case ETM_ADDR_TYPE_RANGE:
			etm4_set_comparator_filter(config,
						   filter->start_addr,
						   filter->stop_addr,
						   comparator);
			/*
			 * TRCVICTLR::SSSTATUS == 1, the start-stop logic is
			 * in the started state
			 */
			config->vinst_ctrl |= BIT(9);

			/* No start-stop filtering for ViewInst */
			config->vissctlr = 0x0;
			break;
963 964 965 966 967 968 969 970 971 972 973 974 975 976 977 978 979 980 981 982 983 984 985 986 987 988 989 990
		case ETM_ADDR_TYPE_START:
		case ETM_ADDR_TYPE_STOP:
			/* Get the right start or stop address */
			address = (type == ETM_ADDR_TYPE_START ?
				   filter->start_addr :
				   filter->stop_addr);

			/* Configure comparator */
			etm4_set_start_stop_filter(config, address,
						   comparator, type);

			/*
			 * If filters::ssstatus == 1, trace acquisition was
			 * started but the process was yanked away before the
			 * the stop address was hit.  As such the start/stop
			 * logic needs to be re-started so that tracing can
			 * resume where it left.
			 *
			 * The start/stop logic status when a process is
			 * scheduled out is checked in function
			 * etm4_disable_perf().
			 */
			if (filters->ssstatus)
				config->vinst_ctrl |= BIT(9);

			/* No include/exclude filtering for ViewInst */
			config->viiectlr = 0x0;
			break;
991 992 993 994 995 996 997 998 999 1000 1001 1002 1003 1004 1005 1006
		default:
			ret = -EINVAL;
			goto out;
		}
	}

	goto out;


default_filter:
	etm4_set_default_filter(config);

out:
	return ret;
}

1007 1008 1009 1010 1011 1012 1013 1014 1015 1016 1017 1018 1019 1020 1021 1022
void etm4_config_trace_mode(struct etmv4_config *config)
{
	u32 addr_acc, mode;

	mode = config->mode;
	mode &= (ETM_MODE_EXCL_KERN | ETM_MODE_EXCL_USER);

	/* excluding kernel AND user space doesn't make sense */
	WARN_ON_ONCE(mode == (ETM_MODE_EXCL_KERN | ETM_MODE_EXCL_USER));

	/* nothing to do if neither flags are set */
	if (!(mode & ETM_MODE_EXCL_KERN) && !(mode & ETM_MODE_EXCL_USER))
		return;

	addr_acc = config->addr_acc[ETM_DEFAULT_ADDR_COMP];
	/* clear default config */
1023 1024
	addr_acc &= ~(ETM_EXLEVEL_NS_APP | ETM_EXLEVEL_NS_OS |
		      ETM_EXLEVEL_NS_HYP);
1025

1026
	addr_acc |= etm4_get_ns_access_type(config);
1027 1028 1029 1030 1031

	config->addr_acc[ETM_DEFAULT_ADDR_COMP] = addr_acc;
	config->addr_acc[ETM_DEFAULT_ADDR_COMP + 1] = addr_acc;
}

1032
static int etm4_online_cpu(unsigned int cpu)
1033 1034
{
	if (!etmdrvdata[cpu])
1035
		return 0;
1036

1037 1038 1039 1040
	if (etmdrvdata[cpu]->boot_enable && !etmdrvdata[cpu]->sticky_enable)
		coresight_enable(etmdrvdata[cpu]->csdev);
	return 0;
}
1041

1042 1043 1044 1045 1046 1047 1048 1049 1050
static int etm4_starting_cpu(unsigned int cpu)
{
	if (!etmdrvdata[cpu])
		return 0;

	spin_lock(&etmdrvdata[cpu]->spinlock);
	if (!etmdrvdata[cpu]->os_unlock) {
		etm4_os_unlock(etmdrvdata[cpu]);
		etmdrvdata[cpu]->os_unlock = true;
1051
	}
1052 1053 1054 1055 1056

	if (local_read(&etmdrvdata[cpu]->mode))
		etm4_enable_hw(etmdrvdata[cpu]);
	spin_unlock(&etmdrvdata[cpu]->spinlock);
	return 0;
1057 1058
}

1059 1060 1061 1062 1063 1064 1065 1066 1067 1068 1069
static int etm4_dying_cpu(unsigned int cpu)
{
	if (!etmdrvdata[cpu])
		return 0;

	spin_lock(&etmdrvdata[cpu]->spinlock);
	if (local_read(&etmdrvdata[cpu]->mode))
		etm4_disable_hw(etmdrvdata[cpu]);
	spin_unlock(&etmdrvdata[cpu]->spinlock);
	return 0;
}
1070

1071 1072 1073 1074 1075
static void etm4_init_trace_id(struct etmv4_drvdata *drvdata)
{
	drvdata->trcid = coresight_get_trace_id(drvdata->cpu);
}

1076 1077 1078 1079 1080 1081 1082 1083
static int etm4_probe(struct amba_device *adev, const struct amba_id *id)
{
	int ret;
	void __iomem *base;
	struct device *dev = &adev->dev;
	struct coresight_platform_data *pdata = NULL;
	struct etmv4_drvdata *drvdata;
	struct resource *res = &adev->res;
1084
	struct coresight_desc desc = { 0 };
1085 1086 1087 1088 1089 1090 1091 1092 1093 1094 1095 1096 1097 1098 1099 1100 1101 1102 1103 1104 1105 1106 1107 1108 1109 1110 1111
	struct device_node *np = adev->dev.of_node;

	drvdata = devm_kzalloc(dev, sizeof(*drvdata), GFP_KERNEL);
	if (!drvdata)
		return -ENOMEM;

	if (np) {
		pdata = of_get_coresight_platform_data(dev, np);
		if (IS_ERR(pdata))
			return PTR_ERR(pdata);
		adev->dev.platform_data = pdata;
	}

	drvdata->dev = &adev->dev;
	dev_set_drvdata(dev, drvdata);

	/* Validity for the resource is already checked by the AMBA core */
	base = devm_ioremap_resource(dev, res);
	if (IS_ERR(base))
		return PTR_ERR(base);

	drvdata->base = base;

	spin_lock_init(&drvdata->spinlock);

	drvdata->cpu = pdata ? pdata->cpu : 0;

1112
	cpus_read_lock();
1113 1114 1115 1116 1117 1118
	etmdrvdata[drvdata->cpu] = drvdata;

	if (smp_call_function_single(drvdata->cpu,
				etm4_init_arch_data,  drvdata, 1))
		dev_err(dev, "ETM arch init failed\n");

1119
	if (!etm4_count++) {
1120 1121 1122 1123 1124 1125
		cpuhp_setup_state_nocalls_cpuslocked(CPUHP_AP_ARM_CORESIGHT_STARTING,
						     "arm/coresight4:starting",
						     etm4_starting_cpu, etm4_dying_cpu);
		ret = cpuhp_setup_state_nocalls_cpuslocked(CPUHP_AP_ONLINE_DYN,
							   "arm/coresight4:online",
							   etm4_online_cpu, NULL);
1126 1127 1128 1129
		if (ret < 0)
			goto err_arch_supported;
		hp_online = ret;
	}
1130

1131
	cpus_read_unlock();
1132 1133 1134 1135 1136

	if (etm4_arch_supported(drvdata->arch) == false) {
		ret = -EINVAL;
		goto err_arch_supported;
	}
1137 1138 1139

	etm4_init_trace_id(drvdata);
	etm4_set_default(&drvdata->config);
1140

1141 1142 1143 1144 1145 1146 1147
	desc.type = CORESIGHT_DEV_TYPE_SOURCE;
	desc.subtype.source_subtype = CORESIGHT_DEV_SUBTYPE_SOURCE_PROC;
	desc.ops = &etm4_cs_ops;
	desc.pdata = pdata;
	desc.dev = dev;
	desc.groups = coresight_etmv4_groups;
	drvdata->csdev = coresight_register(&desc);
1148 1149
	if (IS_ERR(drvdata->csdev)) {
		ret = PTR_ERR(drvdata->csdev);
1150
		goto err_arch_supported;
1151 1152
	}

1153 1154 1155 1156 1157 1158 1159
	ret = etm_perf_symlink(drvdata->csdev, true);
	if (ret) {
		coresight_unregister(drvdata->csdev);
		goto err_arch_supported;
	}

	pm_runtime_put(&adev->dev);
1160 1161
	dev_info(dev, "CPU%d: ETM v%d.%d initialized\n",
		 drvdata->cpu, drvdata->arch >> 4, drvdata->arch & 0xf);
1162 1163 1164 1165 1166 1167 1168 1169 1170

	if (boot_enable) {
		coresight_enable(drvdata->csdev);
		drvdata->boot_enable = true;
	}

	return 0;

err_arch_supported:
1171
	if (--etm4_count == 0) {
1172
		cpuhp_remove_state_nocalls(CPUHP_AP_ARM_CORESIGHT_STARTING);
1173 1174 1175
		if (hp_online)
			cpuhp_remove_state_nocalls(hp_online);
	}
1176 1177 1178
	return ret;
}

1179 1180 1181 1182 1183 1184
static struct amba_cs_uci_id uci_id_etm4[] = {
	{
		/*  ETMv4 UCI data */
		.devarch	= 0x47704a13,
		.devarch_mask	= 0xfff0ffff,
		.devtype	= 0x00000013,
1185
	}
1186
};
1187

1188
static const struct amba_id etm4_ids[] = {
1189 1190 1191 1192 1193
	CS_AMBA_ID(0x000bb95d),		/* Cortex-A53 */
	CS_AMBA_ID(0x000bb95e),		/* Cortex-A57 */
	CS_AMBA_ID(0x000bb95a),		/* Cortex-A72 */
	CS_AMBA_ID(0x000bb959),		/* Cortex-A73 */
	CS_AMBA_UCI_ID(0x000bb9da, uci_id_etm4),	/* Cortex-A35 */
1194
	{},
1195 1196 1197 1198 1199
};

static struct amba_driver etm4x_driver = {
	.drv = {
		.name   = "coresight-etm4x",
1200
		.suppress_bind_attrs = true,
1201 1202 1203 1204
	},
	.probe		= etm4_probe,
	.id_table	= etm4_ids,
};
1205
builtin_amba_driver(etm4x_driver);