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
|
/*-
* Copyright (c) 2003-2008 Joseph Koshy
* Copyright (c) 2007 The FreeBSD Foundation
* All rights reserved.
*
* Portions of this software were developed by A. Joseph Koshy under
* sponsorship from the FreeBSD Foundation and Google, Inc.
*
* Redistribution and use in source and binary forms, with or without
* modification, are permitted provided that the following conditions
* are met:
* 1. Redistributions of source code must retain the above copyright
* notice, this list of conditions and the following disclaimer.
* 2. Redistributions in binary form must reproduce the above copyright
* notice, this list of conditions and the following disclaimer in the
* documentation and/or other materials provided with the distribution.
*
* THIS SOFTWARE IS PROVIDED BY THE AUTHORS AND CONTRIBUTORS ``AS IS'' AND
* ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
* IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
* ARE DISCLAIMED. IN NO EVENT SHALL THE AUTHORS OR CONTRIBUTORS BE LIABLE
* FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
* DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
* OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
* HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
* LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
* OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
* SUCH DAMAGE.
*/
#include <sys/cdefs.h>
__FBSDID("$FreeBSD$");
#include "opt_hwpmc_hooks.h"
#include <sys/types.h>
#include <sys/ctype.h>
#include <sys/param.h>
#include <sys/malloc.h>
#include <sys/kernel.h>
#include <sys/lock.h>
#include <sys/mutex.h>
#include <sys/pmc.h>
#include <sys/pmckern.h>
#include <sys/smp.h>
#include <sys/sysctl.h>
#include <sys/systm.h>
#ifdef HWPMC_HOOKS
FEATURE(hwpmc_hooks, "Kernel support for HW PMC");
#define PMC_KERNEL_VERSION PMC_VERSION
#else
#define PMC_KERNEL_VERSION 0
#endif
MALLOC_DECLARE(M_PMCHOOKS);
MALLOC_DEFINE(M_PMCHOOKS, "pmchooks", "Memory space for PMC hooks");
const int pmc_kernel_version = PMC_KERNEL_VERSION;
/* Hook variable. */
int (*pmc_hook)(struct thread *td, int function, void *arg) = NULL;
/* Interrupt handler */
int (*pmc_intr)(int cpu, struct trapframe *tf) = NULL;
/* Bitmask of CPUs requiring servicing at hardclock time */
volatile cpuset_t pmc_cpumask;
/*
* A global count of SS mode PMCs. When non-zero, this means that
* we have processes that are sampling the system as a whole.
*/
volatile int pmc_ss_count;
/*
* Since PMC(4) may not be loaded in the current kernel, the
* convention followed is that a non-NULL value of 'pmc_hook' implies
* the presence of this kernel module.
*
* This requires us to protect 'pmc_hook' with a
* shared (sx) lock -- thus making the process of calling into PMC(4)
* somewhat more expensive than a simple 'if' check and indirect call.
*/
struct sx pmc_sx;
/*
* PMC Soft per cpu trapframe.
*/
struct trapframe pmc_tf[MAXCPU];
/*
* PMC Soft use a global table to store registered events.
*/
SYSCTL_NODE(_kern, OID_AUTO, hwpmc, CTLFLAG_RW, 0, "HWPMC parameters");
static int pmc_softevents = 16;
TUNABLE_INT(PMC_SYSCTL_NAME_PREFIX "softevents", &pmc_softevents);
SYSCTL_INT(_kern_hwpmc, OID_AUTO, softevents, CTLFLAG_TUN|CTLFLAG_RD,
&pmc_softevents, 0, "maximum number of soft events");
struct mtx pmc_softs_mtx;
int pmc_softs_count;
struct pmc_soft **pmc_softs;
MTX_SYSINIT(pmc_soft_mtx, &pmc_softs_mtx, "pmc-softs", MTX_SPIN);
static void
pmc_init_sx(void)
{
sx_init_flags(&pmc_sx, "pmc-sx", SX_NOWITNESS);
}
SYSINIT(pmcsx, SI_SUB_LOCK, SI_ORDER_MIDDLE, pmc_init_sx, NULL);
/*
* Helper functions.
*/
/*
* A note on the CPU numbering scheme used by the hwpmc(4) driver.
*
* CPUs are denoted using numbers in the range 0..[pmc_cpu_max()-1].
* CPUs could be numbered "sparsely" in this range; the predicate
* `pmc_cpu_is_present()' is used to test whether a given CPU is
* physically present.
*
* Further, a CPU that is physically present may be administratively
* disabled or otherwise unavailable for use by hwpmc(4). The
* `pmc_cpu_is_active()' predicate tests for CPU usability. An
* "active" CPU participates in thread scheduling and can field
* interrupts raised by PMC hardware.
*
* On systems with hyperthreaded CPUs, multiple logical CPUs may share
* PMC hardware resources. For such processors one logical CPU is
* denoted as the primary owner of the in-CPU PMC resources. The
* pmc_cpu_is_primary() predicate is used to distinguish this primary
* CPU from the others.
*/
int
pmc_cpu_is_active(int cpu)
{
#ifdef SMP
return (pmc_cpu_is_present(cpu) &&
!CPU_ISSET(cpu, &hlt_cpus_mask));
#else
return (1);
#endif
}
/* Deprecated. */
int
pmc_cpu_is_disabled(int cpu)
{
return (!pmc_cpu_is_active(cpu));
}
int
pmc_cpu_is_present(int cpu)
{
#ifdef SMP
return (!CPU_ABSENT(cpu));
#else
return (1);
#endif
}
int
pmc_cpu_is_primary(int cpu)
{
#ifdef SMP
return (!CPU_ISSET(cpu, &logical_cpus_mask));
#else
return (1);
#endif
}
/*
* Return the maximum CPU number supported by the system. The return
* value is used for scaling internal data structures and for runtime
* checks.
*/
unsigned int
pmc_cpu_max(void)
{
#ifdef SMP
return (mp_maxid+1);
#else
return (1);
#endif
}
#ifdef INVARIANTS
/*
* Return the count of CPUs in the `active' state in the system.
*/
int
pmc_cpu_max_active(void)
{
#ifdef SMP
/*
* When support for CPU hot-plugging is added to the kernel,
* this function would change to return the current number
* of "active" CPUs.
*/
return (mp_ncpus);
#else
return (1);
#endif
}
#endif
/*
* Cleanup event name:
* - remove duplicate '_'
* - all uppercase
*/
static void
pmc_soft_namecleanup(char *name)
{
char *p, *q;
p = q = name;
for ( ; *p == '_' ; p++)
;
for ( ; *p ; p++) {
if (*p == '_' && (*(p + 1) == '_' || *(p + 1) == '\0'))
continue;
else
*q++ = toupper(*p);
}
*q = '\0';
}
void
pmc_soft_ev_register(struct pmc_soft *ps)
{
static int warned = 0;
int n;
ps->ps_running = 0;
ps->ps_ev.pm_ev_code = 0; /* invalid */
pmc_soft_namecleanup(ps->ps_ev.pm_ev_name);
mtx_lock_spin(&pmc_softs_mtx);
if (pmc_softs_count >= pmc_softevents) {
/*
* XXX Reusing events can enter a race condition where
* new allocated event will be used as an old one.
*/
for (n = 0; n < pmc_softevents; n++)
if (pmc_softs[n] == NULL)
break;
if (n == pmc_softevents) {
mtx_unlock_spin(&pmc_softs_mtx);
if (!warned) {
printf("hwpmc: too many soft events, "
"increase kern.hwpmc.softevents tunable\n");
warned = 1;
}
return;
}
ps->ps_ev.pm_ev_code = PMC_EV_SOFT_FIRST + n;
pmc_softs[n] = ps;
} else {
ps->ps_ev.pm_ev_code = PMC_EV_SOFT_FIRST + pmc_softs_count;
pmc_softs[pmc_softs_count++] = ps;
}
mtx_unlock_spin(&pmc_softs_mtx);
}
void
pmc_soft_ev_deregister(struct pmc_soft *ps)
{
KASSERT(ps != NULL, ("pmc_soft_deregister: called with NULL"));
mtx_lock_spin(&pmc_softs_mtx);
if (ps->ps_ev.pm_ev_code != 0 &&
(ps->ps_ev.pm_ev_code - PMC_EV_SOFT_FIRST) < pmc_softevents) {
KASSERT(ps->ps_ev.pm_ev_code >= PMC_EV_SOFT_FIRST &&
ps->ps_ev.pm_ev_code <= PMC_EV_SOFT_LAST,
("pmc_soft_deregister: invalid event value"));
pmc_softs[ps->ps_ev.pm_ev_code - PMC_EV_SOFT_FIRST] = NULL;
}
mtx_unlock_spin(&pmc_softs_mtx);
}
struct pmc_soft *
pmc_soft_ev_acquire(enum pmc_event ev)
{
struct pmc_soft *ps;
if (ev == 0 || (ev - PMC_EV_SOFT_FIRST) >= pmc_softevents)
return NULL;
KASSERT(ev >= PMC_EV_SOFT_FIRST &&
ev <= PMC_EV_SOFT_LAST,
("event out of range"));
mtx_lock_spin(&pmc_softs_mtx);
ps = pmc_softs[ev - PMC_EV_SOFT_FIRST];
if (ps == NULL)
mtx_unlock_spin(&pmc_softs_mtx);
return ps;
}
void
pmc_soft_ev_release(struct pmc_soft *ps)
{
mtx_unlock_spin(&pmc_softs_mtx);
}
/*
* Initialise hwpmc.
*/
static void
init_hwpmc(void *dummy __unused)
{
if (pmc_softevents <= 0 ||
pmc_softevents > PMC_EV_DYN_COUNT) {
(void) printf("hwpmc: tunable \"softevents\"=%d out of "
"range.\n", pmc_softevents);
pmc_softevents = PMC_EV_DYN_COUNT;
}
pmc_softs = malloc(pmc_softevents * sizeof(struct pmc_soft *), M_PMCHOOKS, M_NOWAIT|M_ZERO);
KASSERT(pmc_softs != NULL, ("cannot allocate soft events table"));
}
SYSINIT(hwpmc, SI_SUB_KDTRACE, SI_ORDER_FIRST, init_hwpmc, NULL);
|