perf_event.h 15.5 KB
Newer Older
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
/*
 * Performance events x86 architecture header
 *
 *  Copyright (C) 2008 Thomas Gleixner <tglx@linutronix.de>
 *  Copyright (C) 2008-2009 Red Hat, Inc., Ingo Molnar
 *  Copyright (C) 2009 Jaswinder Singh Rajput
 *  Copyright (C) 2009 Advanced Micro Devices, Inc., Robert Richter
 *  Copyright (C) 2008-2009 Red Hat, Inc., Peter Zijlstra <pzijlstr@redhat.com>
 *  Copyright (C) 2009 Intel Corporation, <markus.t.metzger@intel.com>
 *  Copyright (C) 2009 Google, Inc., Stephane Eranian
 *
 *  For licencing details see kernel-base/COPYING
 */

#include <linux/perf_event.h>

17 18 19 20 21 22 23 24 25 26 27 28
#if 0
#undef wrmsrl
#define wrmsrl(msr, val) 						\
do {									\
	unsigned int _msr = (msr);					\
	u64 _val = (val);						\
	trace_printk("wrmsrl(%x, %Lx)\n", (unsigned int)(_msr),		\
			(unsigned long long)(_val));			\
	native_write_msr((_msr), (u32)(_val), (u32)(_val >> 32));	\
} while (0)
#endif

29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47
/*
 *          |   NHM/WSM    |      SNB     |
 * register -------------------------------
 *          |  HT  | no HT |  HT  | no HT |
 *-----------------------------------------
 * offcore  | core | core  | cpu  | core  |
 * lbr_sel  | core | core  | cpu  | core  |
 * ld_lat   | cpu  | core  | cpu  | core  |
 *-----------------------------------------
 *
 * Given that there is a small number of shared regs,
 * we can pre-allocate their slot in the per-cpu
 * per-core reg tables.
 */
enum extra_reg_type {
	EXTRA_REG_NONE  = -1,	/* not used */

	EXTRA_REG_RSP_0 = 0,	/* offcore_response_0 */
	EXTRA_REG_RSP_1 = 1,	/* offcore_response_1 */
48
	EXTRA_REG_LBR   = 2,	/* lbr_select */
49 50 51 52 53 54 55 56 57 58 59 60

	EXTRA_REG_MAX		/* number of entries needed */
};

struct event_constraint {
	union {
		unsigned long	idxmsk[BITS_TO_LONGS(X86_PMC_IDX_MAX)];
		u64		idxmsk64;
	};
	u64	code;
	u64	cmask;
	int	weight;
61
	int	overlap;
62 63 64 65 66 67 68 69 70 71
};

struct amd_nb {
	int nb_id;  /* NorthBridge id */
	int refcnt; /* reference count */
	struct perf_event *owners[X86_PMC_IDX_MAX];
	struct event_constraint event_constraints[X86_PMC_IDX_MAX];
};

/* The maximal number of PEBS events: */
72
#define MAX_PEBS_EVENTS		8
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

/*
 * A debug store configuration.
 *
 * We only support architectures that use 64bit fields.
 */
struct debug_store {
	u64	bts_buffer_base;
	u64	bts_index;
	u64	bts_absolute_maximum;
	u64	bts_interrupt_threshold;
	u64	pebs_buffer_base;
	u64	pebs_index;
	u64	pebs_absolute_maximum;
	u64	pebs_interrupt_threshold;
	u64	pebs_event_reset[MAX_PEBS_EVENTS];
};

/*
 * Per register state.
 */
struct er_account {
	raw_spinlock_t		lock;	/* per-core: protect structure */
	u64                 config;	/* extra MSR config */
	u64                 reg;	/* extra MSR number */
	atomic_t            ref;	/* reference count */
};

/*
 * Per core/cpu state
 *
 * Used to coordinate shared registers between HT threads or
 * among events on a single PMU.
 */
struct intel_shared_regs {
	struct er_account       regs[EXTRA_REG_MAX];
	int                     refcnt;		/* per-core: #HT threads */
	unsigned                core_id;	/* per-core: core id */
};

#define MAX_LBR_ENTRIES		16

struct cpu_hw_events {
	/*
	 * Generic x86 PMC bits
	 */
	struct perf_event	*events[X86_PMC_IDX_MAX]; /* in counter order */
	unsigned long		active_mask[BITS_TO_LONGS(X86_PMC_IDX_MAX)];
	unsigned long		running[BITS_TO_LONGS(X86_PMC_IDX_MAX)];
	int			enabled;

	int			n_events;
	int			n_added;
	int			n_txn;
	int			assign[X86_PMC_IDX_MAX]; /* event to counter assignment */
	u64			tags[X86_PMC_IDX_MAX];
	struct perf_event	*event_list[X86_PMC_IDX_MAX]; /* in enabled order */

	unsigned int		group_flag;
132
	int			is_fake;
133 134 135 136 137 138 139 140 141 142 143 144 145 146

	/*
	 * Intel DebugStore bits
	 */
	struct debug_store	*ds;
	u64			pebs_enabled;

	/*
	 * Intel LBR bits
	 */
	int				lbr_users;
	void				*lbr_context;
	struct perf_branch_stack	lbr_stack;
	struct perf_branch_entry	lbr_entries[MAX_LBR_ENTRIES];
147
	struct er_account		*lbr_sel;
148
	u64				br_sel;
149

150 151 152 153 154 155 156
	/*
	 * Intel host/guest exclude bits
	 */
	u64				intel_ctrl_guest_mask;
	u64				intel_ctrl_host_mask;
	struct perf_guest_switch_msr	guest_switch_msrs[X86_PMC_IDX_MAX];

157 158 159 160 161 162 163 164 165
	/*
	 * manage shared (per-core, per-cpu) registers
	 * used on Intel NHM/WSM/SNB
	 */
	struct intel_shared_regs	*shared_regs;

	/*
	 * AMD specific bits
	 */
166 167 168
	struct amd_nb			*amd_nb;
	/* Inverted mask of bits to clear in the perf_ctr ctrl registers */
	u64				perf_ctr_virt_mask;
169 170 171 172

	void				*kfree_on_online;
};

173
#define __EVENT_CONSTRAINT(c, n, m, w, o) {\
174 175 176 177
	{ .idxmsk64 = (n) },		\
	.code = (c),			\
	.cmask = (m),			\
	.weight = (w),			\
178
	.overlap = (o),			\
179 180 181
}

#define EVENT_CONSTRAINT(c, n, m)	\
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
	__EVENT_CONSTRAINT(c, n, m, HWEIGHT(n), 0)

/*
 * The overlap flag marks event constraints with overlapping counter
 * masks. This is the case if the counter mask of such an event is not
 * a subset of any other counter mask of a constraint with an equal or
 * higher weight, e.g.:
 *
 *  c_overlaps = EVENT_CONSTRAINT_OVERLAP(0, 0x09, 0);
 *  c_another1 = EVENT_CONSTRAINT(0, 0x07, 0);
 *  c_another2 = EVENT_CONSTRAINT(0, 0x38, 0);
 *
 * The event scheduler may not select the correct counter in the first
 * cycle because it needs to know which subsequent events will be
 * scheduled. It may fail to schedule the events then. So we set the
 * overlap flag for such constraints to give the scheduler a hint which
 * events to select for counter rescheduling.
 *
 * Care must be taken as the rescheduling algorithm is O(n!) which
 * will increase scheduling cycles for an over-commited system
 * dramatically.  The number of such EVENT_CONSTRAINT_OVERLAP() macros
 * and its counter masks must be kept at a minimum.
 */
#define EVENT_CONSTRAINT_OVERLAP(c, n, m)	\
	__EVENT_CONSTRAINT(c, n, m, HWEIGHT(n), 1)
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

/*
 * Constraint on the Event code.
 */
#define INTEL_EVENT_CONSTRAINT(c, n)	\
	EVENT_CONSTRAINT(c, n, ARCH_PERFMON_EVENTSEL_EVENT)

/*
 * Constraint on the Event code + UMask + fixed-mask
 *
 * filter mask to validate fixed counter events.
 * the following filters disqualify for fixed counters:
 *  - inv
 *  - edge
 *  - cnt-mask
 *  The other filters are supported by fixed counters.
 *  The any-thread option is supported starting with v3.
 */
#define FIXED_EVENT_CONSTRAINT(c, n)	\
	EVENT_CONSTRAINT(c, (1ULL << (32+n)), X86_RAW_EVENT_MASK)

/*
 * Constraint on the Event code + UMask
 */
#define INTEL_UEVENT_CONSTRAINT(c, n)	\
	EVENT_CONSTRAINT(c, n, INTEL_ARCH_EVENT_MASK)

#define EVENT_CONSTRAINT_END		\
	EVENT_CONSTRAINT(0, 0, 0)

#define for_each_event_constraint(e, c)	\
	for ((e) = (c); (e)->weight; (e)++)

/*
 * Extra registers for specific events.
 *
 * Some events need large masks and require external MSRs.
 * Those extra MSRs end up being shared for all events on
 * a PMU and sometimes between PMU of sibling HT threads.
 * In either case, the kernel needs to handle conflicting
 * accesses to those extra, shared, regs. The data structure
 * to manage those registers is stored in cpu_hw_event.
 */
struct extra_reg {
	unsigned int		event;
	unsigned int		msr;
	u64			config_mask;
	u64			valid_mask;
	int			idx;  /* per_xxx->regs[] reg index */
};

#define EVENT_EXTRA_REG(e, ms, m, vm, i) {	\
	.event = (e),		\
	.msr = (ms),		\
	.config_mask = (m),	\
	.valid_mask = (vm),	\
	.idx = EXTRA_REG_##i	\
	}

#define INTEL_EVENT_EXTRA_REG(event, msr, vm, idx)	\
	EVENT_EXTRA_REG(event, msr, ARCH_PERFMON_EVENTSEL_EVENT, vm, idx)

#define EVENT_EXTRA_END EVENT_EXTRA_REG(0, 0, 0, 0, RSP_0)

union perf_capabilities {
	struct {
		u64	lbr_format:6;
		u64	pebs_trap:1;
		u64	pebs_arch_reg:1;
		u64	pebs_format:4;
		u64	smm_freeze:1;
	};
	u64	capabilities;
};

282 283 284 285 286
struct x86_pmu_quirk {
	struct x86_pmu_quirk *next;
	void (*func)(void);
};

287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309
union x86_pmu_config {
	struct {
		u64 event:8,
		    umask:8,
		    usr:1,
		    os:1,
		    edge:1,
		    pc:1,
		    interrupt:1,
		    __reserved1:1,
		    en:1,
		    inv:1,
		    cmask:8,
		    event2:4,
		    __reserved2:4,
		    go:1,
		    ho:1;
	} bits;
	u64 value;
};

#define X86_CONFIG(args...) ((union x86_pmu_config){.bits = {args}}).value

310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333
/*
 * struct x86_pmu - generic x86 pmu
 */
struct x86_pmu {
	/*
	 * Generic x86 PMC bits
	 */
	const char	*name;
	int		version;
	int		(*handle_irq)(struct pt_regs *);
	void		(*disable_all)(void);
	void		(*enable_all)(int added);
	void		(*enable)(struct perf_event *);
	void		(*disable)(struct perf_event *);
	int		(*hw_config)(struct perf_event *event);
	int		(*schedule_events)(struct cpu_hw_events *cpuc, int n, int *assign);
	unsigned	eventsel;
	unsigned	perfctr;
	u64		(*event_map)(int);
	int		max_events;
	int		num_counters;
	int		num_counters_fixed;
	int		cntval_bits;
	u64		cntval_mask;
334 335 336 337 338
	union {
			unsigned long events_maskl;
			unsigned long events_mask[BITS_TO_LONGS(ARCH_PERFMON_EVENTS_COUNT)];
	};
	int		events_mask_len;
339 340 341 342 343 344 345 346 347
	int		apic;
	u64		max_period;
	struct event_constraint *
			(*get_event_constraints)(struct cpu_hw_events *cpuc,
						 struct perf_event *event);

	void		(*put_event_constraints)(struct cpu_hw_events *cpuc,
						 struct perf_event *event);
	struct event_constraint *event_constraints;
348
	struct x86_pmu_quirk *quirks;
349 350
	int		perfctr_second_write;

351 352 353 354
	/*
	 * sysfs attrs
	 */
	int		attr_rdpmc;
355
	struct attribute **format_attrs;
356 357 358 359

	/*
	 * CPU Hotplug hooks
	 */
360 361 362 363
	int		(*cpu_prepare)(int cpu);
	void		(*cpu_starting)(int cpu);
	void		(*cpu_dying)(int cpu);
	void		(*cpu_dead)(int cpu);
364 365

	void		(*check_microcode)(void);
366
	void		(*flush_branch_stack)(void);
367 368 369 370 371 372 373 374 375 376

	/*
	 * Intel Arch Perfmon v2+
	 */
	u64			intel_ctrl;
	union perf_capabilities intel_cap;

	/*
	 * Intel DebugStore bits
	 */
377
	unsigned int	bts		:1,
378 379 380 381
			bts_active	:1,
			pebs		:1,
			pebs_active	:1,
			pebs_broken	:1;
382 383 384
	int		pebs_record_size;
	void		(*drain_pebs)(struct pt_regs *regs);
	struct event_constraint *pebs_constraints;
385
	void		(*pebs_aliases)(struct perf_event *event);
386
	int 		max_pebs_events;
387 388 389 390 391 392

	/*
	 * Intel LBR
	 */
	unsigned long	lbr_tos, lbr_from, lbr_to; /* MSR base regs       */
	int		lbr_nr;			   /* hardware stack size */
393 394
	u64		lbr_sel_mask;		   /* LBR_SELECT valid bits */
	const int	*lbr_sel_map;		   /* lbr_select mappings */
395 396 397 398 399 400

	/*
	 * Extra registers for events
	 */
	struct extra_reg *extra_regs;
	unsigned int er_flags;
401 402 403 404 405

	/*
	 * Intel host/guest support (KVM)
	 */
	struct perf_guest_switch_msr *(*guest_get_msrs)(int *nr);
406 407
};

408 409 410 411 412 413 414 415 416
#define x86_add_quirk(func_)						\
do {									\
	static struct x86_pmu_quirk __quirk __initdata = {		\
		.func = func_,						\
	};								\
	__quirk.next = x86_pmu.quirks;					\
	x86_pmu.quirks = &__quirk;					\
} while (0)

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
#define ERF_NO_HT_SHARING	1
#define ERF_HAS_RSP_1		2

extern struct x86_pmu x86_pmu __read_mostly;

DECLARE_PER_CPU(struct cpu_hw_events, cpu_hw_events);

int x86_perf_event_set_period(struct perf_event *event);

/*
 * Generalized hw caching related hw_event table, filled
 * in on a per model basis. A value of 0 means
 * 'not supported', -1 means 'hw_event makes no sense on
 * this CPU', any other value means the raw hw_event
 * ID.
 */

#define C(x) PERF_COUNT_HW_CACHE_##x

extern u64 __read_mostly hw_cache_event_ids
				[PERF_COUNT_HW_CACHE_MAX]
				[PERF_COUNT_HW_CACHE_OP_MAX]
				[PERF_COUNT_HW_CACHE_RESULT_MAX];
extern u64 __read_mostly hw_cache_extra_regs
				[PERF_COUNT_HW_CACHE_MAX]
				[PERF_COUNT_HW_CACHE_OP_MAX]
				[PERF_COUNT_HW_CACHE_RESULT_MAX];

u64 x86_perf_event_update(struct perf_event *event);

static inline int x86_pmu_addr_offset(int index)
{
	int offset;

	/* offset = X86_FEATURE_PERFCTR_CORE ? index << 1 : index */
	alternative_io(ASM_NOP2,
		       "shll $1, %%eax",
		       X86_FEATURE_PERFCTR_CORE,
		       "=a" (offset),
		       "a"  (index));

	return offset;
}

static inline unsigned int x86_pmu_config_addr(int index)
{
	return x86_pmu.eventsel + x86_pmu_addr_offset(index);
}

static inline unsigned int x86_pmu_event_addr(int index)
{
	return x86_pmu.perfctr + x86_pmu_addr_offset(index);
}

int x86_setup_perfctr(struct perf_event *event);

int x86_pmu_hw_config(struct perf_event *event);

void x86_pmu_disable_all(void);

static inline void __x86_pmu_enable_event(struct hw_perf_event *hwc,
					  u64 enable_mask)
{
480 481
	u64 disable_mask = __this_cpu_read(cpu_hw_events.perf_ctr_virt_mask);

482 483
	if (hwc->extra_reg.reg)
		wrmsrl(hwc->extra_reg.reg, hwc->extra_reg.config);
484
	wrmsrl(hwc->config_base, (hwc->config | enable_mask) & ~disable_mask);
485 486 487 488
}

void x86_pmu_enable_all(int added);

489 490
int perf_assign_events(struct event_constraint **constraints, int n,
			int wmin, int wmax, int *assign);
491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509
int x86_schedule_events(struct cpu_hw_events *cpuc, int n, int *assign);

void x86_pmu_stop(struct perf_event *event, int flags);

static inline void x86_pmu_disable_event(struct perf_event *event)
{
	struct hw_perf_event *hwc = &event->hw;

	wrmsrl(hwc->config_base, hwc->config);
}

void x86_pmu_enable_event(struct perf_event *event);

int x86_pmu_handle_irq(struct pt_regs *regs);

extern struct event_constraint emptyconstraint;

extern struct event_constraint unconstrained;

510 511 512 513 514 515 516 517 518
static inline bool kernel_ip(unsigned long ip)
{
#ifdef CONFIG_X86_32
	return ip > PAGE_OFFSET;
#else
	return (long)ip < 0;
#endif
}

519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538
/*
 * Not all PMUs provide the right context information to place the reported IP
 * into full context. Specifically segment registers are typically not
 * supplied.
 *
 * Assuming the address is a linear address (it is for IBS), we fake the CS and
 * vm86 mode using the known zero-based code segment and 'fix up' the registers
 * to reflect this.
 *
 * Intel PEBS/LBR appear to typically provide the effective address, nothing
 * much we can do about that but pray and treat it like a linear address.
 */
static inline void set_linear_ip(struct pt_regs *regs, unsigned long ip)
{
	regs->cs = kernel_ip(ip) ? __KERNEL_CS : __USER_CS;
	if (regs->flags & X86_VM_MASK)
		regs->flags ^= (PERF_EFLAGS_VM | X86_VM_MASK);
	regs->ip = ip;
}

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
#ifdef CONFIG_CPU_SUP_AMD

int amd_pmu_init(void);

#else /* CONFIG_CPU_SUP_AMD */

static inline int amd_pmu_init(void)
{
	return 0;
}

#endif /* CONFIG_CPU_SUP_AMD */

#ifdef CONFIG_CPU_SUP_INTEL

int intel_pmu_save_and_restart(struct perf_event *event);

struct event_constraint *
x86_get_event_constraints(struct cpu_hw_events *cpuc, struct perf_event *event);

struct intel_shared_regs *allocate_shared_regs(int cpu);

int intel_pmu_init(void);

void init_debug_store_on_cpu(int cpu);

void fini_debug_store_on_cpu(int cpu);

void release_ds_buffers(void);

void reserve_ds_buffers(void);

extern struct event_constraint bts_constraint;

void intel_pmu_enable_bts(u64 config);

void intel_pmu_disable_bts(void);

int intel_pmu_drain_bts_buffer(void);

extern struct event_constraint intel_core2_pebs_event_constraints[];

extern struct event_constraint intel_atom_pebs_event_constraints[];

extern struct event_constraint intel_nehalem_pebs_event_constraints[];

extern struct event_constraint intel_westmere_pebs_event_constraints[];

extern struct event_constraint intel_snb_pebs_event_constraints[];

struct event_constraint *intel_pebs_constraints(struct perf_event *event);

void intel_pmu_pebs_enable(struct perf_event *event);

void intel_pmu_pebs_disable(struct perf_event *event);

void intel_pmu_pebs_enable_all(void);

void intel_pmu_pebs_disable_all(void);

void intel_ds_init(void);

void intel_pmu_lbr_reset(void);

void intel_pmu_lbr_enable(struct perf_event *event);

void intel_pmu_lbr_disable(struct perf_event *event);

void intel_pmu_lbr_enable_all(void);

void intel_pmu_lbr_disable_all(void);

void intel_pmu_lbr_read(void);

void intel_pmu_lbr_init_core(void);

void intel_pmu_lbr_init_nhm(void);

void intel_pmu_lbr_init_atom(void);

619 620
void intel_pmu_lbr_init_snb(void);

621 622
int intel_pmu_setup_lbr_filter(struct perf_event *event);

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
int p4_pmu_init(void);

int p6_pmu_init(void);

#else /* CONFIG_CPU_SUP_INTEL */

static inline void reserve_ds_buffers(void)
{
}

static inline void release_ds_buffers(void)
{
}

static inline int intel_pmu_init(void)
{
	return 0;
}

static inline struct intel_shared_regs *allocate_shared_regs(int cpu)
{
	return NULL;
}

#endif /* CONFIG_CPU_SUP_INTEL */