/*
* ita_rr.c - example of how to use data range restriction with the Itanium PMU
*
* Copyright (c) 2002-2006 Hewlett-Packard Development Company, L.P.
* Contributed by Stephane Eranian <eranian@hpl.hp.com>
*
* Permission is hereby granted, free of charge, to any person obtaining a copy
* of this software and associated documentation files (the "Software"), to deal
* in the Software without restriction, including without limitation the rights
* to use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies
* of the Software, and to permit persons to whom the Software is furnished to do so,
* subject to the following conditions:
*
* The above copyright notice and this permission notice shall be included in all
* copies or substantial portions of the Software.
*
* THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR IMPLIED,
* INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, FITNESS FOR A
* PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT
* HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF
* CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE
* OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.
*
* This file is part of libpfm, a performance monitoring support library for
* applications on Linux/ia64.
*/
#include <sys/types.h>
#include <stdio.h>
#include <stdlib.h>
#include <stdarg.h>
#include <errno.h>
#include <unistd.h>
#include <string.h>
#include <signal.h>
#include <perfmon/perfmon.h>
#include <perfmon/pfmlib_itanium.h>
#define N_LOOP 100000000U
#if defined(__ECC) && defined(__INTEL_COMPILER)
/* if you do not have this file, your compiler is too old */
#include <ia64intrin.h>
#define clear_psr_ac() __rum(1UL<<3)
#elif defined(__GNUC__)
static inline void
clear_psr_ac(void)
{
__asm__ __volatile__("rum psr.ac;;" ::: "memory" );
}
#else
#error "You need to define clear_psr_ac() for your compiler"
#endif
#define TEST_DATA_COUNT 16
#define NUM_PMCS PFMLIB_MAX_PMCS
#define NUM_PMDS PFMLIB_MAX_PMDS
#define MAX_PMU_NAME_LEN 32
#define MAX_EVT_NAME_LEN 128
typedef struct {
char *event_name;
unsigned long expected_value;
} event_desc_t;
static event_desc_t event_list[]={
{ "misaligned_loads_retired", N_LOOP },
{ "misaligned_stores_retired", N_LOOP },
{ NULL, 0UL}
};
typedef union {
unsigned long l_tab[2];
unsigned int i_tab[4];
unsigned short s_tab[8];
unsigned char c_tab[16];
} test_data_t;
static int
do_test(test_data_t *data)
{
unsigned int *l, v;
l = (unsigned int *)(data->c_tab+1);
if (((unsigned long)l & 0x1) == 0) {
printf("Data is not unaligned, can't run test\n");
return -1;
}
v = *l;
v++;
*l = v;
return 0;
}
static void fatal_error(char *fmt,...) __attribute__((noreturn));
static void
fatal_error(char *fmt, ...)
{
va_list ap;
va_start(ap, fmt);
vfprintf(stderr, fmt, ap);
va_end(ap);
exit(1);
}
int
main(int argc, char **argv)
{
event_desc_t *p;
test_data_t *test_data, *test_data_fake;
unsigned long range_start, range_end;
int ret, type = 0;
pfmlib_input_param_t inp;
pfmlib_output_param_t outp;
pfmlib_ita_input_param_t ita_inp;
pfmlib_ita_output_param_t ita_outp;
pfarg_reg_t pd[NUM_PMDS];
pfarg_reg_t pc[NUM_PMCS];
pfarg_dbreg_t dbrs[8];
pfarg_context_t ctx[1];
pfarg_load_t load_args;
pfmlib_options_t pfmlib_options;
unsigned int i;
int id;
char name[MAX_EVT_NAME_LEN];
/*
* Initialize pfm library (required before we can use it)
*/
if (pfm_initialize() != PFMLIB_SUCCESS) {
fatal_error("Can't initialize library\n");
}
/*
* Let's make sure we run this on the right CPU family
*/
pfm_get_pmu_type(&type);
if (type != PFMLIB_ITANIUM_PMU) {
char model[MAX_PMU_NAME_LEN];
pfm_get_pmu_name(model, MAX_PMU_NAME_LEN);
fatal_error("this program does not work with %s PMU\n", model);
}
/*
* pass options to library (optional)
*/
memset(&pfmlib_options, 0, sizeof(pfmlib_options));
pfmlib_options.pfm_debug = 0; /* set to 1 for debug */
pfmlib_options.pfm_verbose = 0; /* set to 1 for debug */
pfm_set_options(&pfmlib_options);
/*
* now let's allocate the data structure we will be monitoring
*/
test_data = (test_data_t *)malloc(sizeof(test_data_t)*TEST_DATA_COUNT);
if (test_data == NULL) {
fatal_error("cannot allocate test data structure");
}
test_data_fake = (test_data_t *)malloc(sizeof(test_data_t)*TEST_DATA_COUNT);
if (test_data_fake == NULL) {
fatal_error("cannot allocate test data structure");
}
/*
* Compute the range we are interested in
*/
range_start = (unsigned long)test_data;
range_end = range_start + sizeof(test_data_t)*TEST_DATA_COUNT;
memset(pd, 0, sizeof(pd));
memset(pc, 0, sizeof(pc));
memset(ctx, 0, sizeof(ctx));
memset(dbrs,0, sizeof(dbrs));
memset(&load_args, 0, sizeof(load_args));
/*
* prepare parameters to library. we don't use any Itanium
* specific features here. so the pfp_model is NULL.
*/
memset(&inp,0, sizeof(inp));
memset(&outp,0, sizeof(outp));
memset(&ita_inp,0, sizeof(ita_inp));
memset(&ita_outp,0, sizeof(ita_outp));
/*
* find requested event
*/
p = event_list;
for (i=0; p->event_name ; i++, p++) {
if (pfm_find_event(p->event_name, &inp.pfp_events[i].event) != PFMLIB_SUCCESS) {
fatal_error("Cannot find %s event\n", p->event_name);
}
}
/*
* set the privilege mode:
* PFM_PLM3 : user level only
*/
inp.pfp_dfl_plm = PFM_PLM3;
/*
* how many counters we use
*/
inp.pfp_event_count = i;
/*
* We use the library to figure out how to program the debug registers
* to cover the data range we are interested in. The rr_end parameter
* must point to the byte after the last of the range (C-style range).
*
* Because of the masking mechanism and therefore alignment constraints used to implement
* this feature, it may not be possible to exactly cover a given range. It may be that
* the coverage exceeds the desired range. So it is possible to capture noise if
* the surrounding addresses are also heavily used. You can figure out, the actual
* start and end offsets of the generated range by checking the rr_soff and rr_eoff fields
* in the pfmlib_ita_output_param_t structure when coming back from the library call.
*
* Upon return, the pfmlib_ita_output_param_t.pfp_ita_drange.rr_dbr array is programmed and
* the number of entries used to cover the range is in rr_nbr_used.
*/
/*
* We indicate that we are using a Data Range Restriction feature.
* In this particular case this will cause, pfm_dispatch_events() to
* add pmc13 to the list of PMC registers to initialize and the
*/
ita_inp.pfp_ita_drange.rr_used = 1;
ita_inp.pfp_ita_drange.rr_limits[0].rr_start = range_start;
ita_inp.pfp_ita_drange.rr_limits[0].rr_end = range_end;
/*
* use the library to find the monitors to use
*
* upon return, cnt contains the number of entries
* used in pc[].
*/
if ((ret=pfm_dispatch_events(&inp, &ita_inp, &outp, &ita_outp)) != PFMLIB_SUCCESS) {
fatal_error("cannot configure events: %s\n", pfm_strerror(ret));
}
printf("data range : [0x%016lx-0x%016lx): %d pair of debug registers used\n"
"start_offset:-0x%lx end_offset:+0x%lx\n",
range_start,
range_end,
ita_outp.pfp_ita_drange.rr_nbr_used >> 1,
ita_outp.pfp_ita_drange.rr_infos[0].rr_soff,
ita_outp.pfp_ita_drange.rr_infos[0].rr_eoff);
printf("fake data range: [0x%016lx-0x%016lx)\n",
(unsigned long)test_data_fake,
(unsigned long)test_data_fake+sizeof(test_data_t)*TEST_DATA_COUNT);
/*
* now create the context for self monitoring/per-task
*/
if (perfmonctl(0, PFM_CREATE_CONTEXT, ctx, 1) == -1) {
if (errno == ENOSYS) {
fatal_error("Your kernel does not have performance monitoring support!\n");
}
fatal_error("cannot create PFM context %s\n", strerror(errno));
}
/*
* extract our file descriptor
*/
id = ctx[0].ctx_fd;
/*
* Now prepare the argument to initialize the PMDs and PMCS.
* We must pfp_pmc_count to determine the number of PMC to intialize.
* We must use pfp_event_count to determine the number of PMD to initialize.
* Some events cause extra PMCs to be used, so pfp_pmc_count may be >= pfp_event_count.
*
* This step is new compared to libpfm-2.x. It is necessary because the library no
* longer knows about the kernel data structures.
*/
for (i=0; i < outp.pfp_pmc_count; i++) {
pc[i].reg_num = outp.pfp_pmcs[i].reg_num;
pc[i].reg_value = outp.pfp_pmcs[i].reg_value;
}
/*
* the PMC controlling the event ALWAYS come first, that's why this loop
* is safe even when extra PMC are needed to support a particular event.
*/
for (i=0; i < inp.pfp_event_count; i++) {
pd[i].reg_num = pc[i].reg_num;
}
/*
* propagate the setup for the debug registers from the library to the arguments
* to the perfmonctl() syscall. The library does not know the type of the syscall
* anymore.
*/
for (i=0; i < ita_outp.pfp_ita_drange.rr_nbr_used; i++) {
dbrs[i].dbreg_num = ita_outp.pfp_ita_drange.rr_br[i].reg_num;
dbrs[i].dbreg_value = ita_outp.pfp_ita_drange.rr_br[i].reg_value;
}
/*
* Program the data debug registers.
*
* IMPORTANT: programming the debug register MUST always be done before the PMCs
* otherwise the kernel will fail on PFM_WRITE_PMCS. This is for security reasons.
*/
if (perfmonctl(id, PFM_WRITE_DBRS, dbrs, ita_outp.pfp_ita_drange.rr_nbr_used) == -1) {
fatal_error("perfmonctl error PFM_WRITE_DBRS errno %d\n",errno);
}
/*
* Now program the registers
*
* We don't use the save variable to indicate the number of elements passed to
* the kernel because, as we said earlier, pc may contain more elements than
* the number of events we specified, i.e., contains more than coutning monitors.
*/
if (perfmonctl(id, PFM_WRITE_PMCS, pc, outp.pfp_pmc_count) == -1) {
fatal_error("perfmonctl error PFM_WRITE_PMCS errno %d\n",errno);
}
if (perfmonctl(id, PFM_WRITE_PMDS, pd, inp.pfp_event_count) == -1) {
fatal_error("perfmonctl error PFM_WRITE_PMDS errno %d\n",errno);
}
/*
* now we load (i.e., attach) the context to ourself
*/
load_args.load_pid = getpid();
if (perfmonctl(id, PFM_LOAD_CONTEXT, &load_args, 1) == -1) {
fatal_error("perfmonctl error PFM_LOAD_CONTEXT errno %d\n",errno);
}
/*
* Let's make sure that the hardware does the unaligned accesses (do not use the
* kernel software handler otherwise the PMU won't see the unaligned fault).
*/
clear_psr_ac();
/*
* Let's roll now.
*
* The idea behind this test is to have two dynamically allocated data structures
* which are access in a unaligned fashion. But we want to capture only the unaligned
* accesses on one of the two. So the debug registers are programmed to cover the
* first one ONLY. Then we activate monotoring and access the two data structures.
* This is an artificial example just to demonstrate how to use data address range
* restrictions.
*/
pfm_self_start(id);
for (i=0; i < N_LOOP; i++) {
do_test(test_data);
do_test(test_data_fake);
}
pfm_self_stop(id);
/*
* now read the results
*/
if (perfmonctl(id, PFM_READ_PMDS, pd, inp.pfp_event_count) == -1) {
fatal_error( "perfmonctl error READ_PMDS errno %d\n",errno);
}
/*
* print the results
*
* It is important to realize, that the first event we specified may not
* be in PMD4. Not all events can be measured by any monitor. That's why
* we need to use the pc[] array to figure out where event i was allocated.
*
* For this example, we expect to see a value of 1 for both misaligned loads
* and misaligned stores. But it can be two when the test_data and test_data_fake
* are allocate very close from each other and the range created with the debug
* registers is larger then test_data.
*
*/
for (i=0; i < inp.pfp_event_count; i++) {
pfm_get_full_event_name(&inp.pfp_events[i], name, MAX_EVT_NAME_LEN);
printf("PMD%u %20lu %s (expected %lu)\n",
pd[i].reg_num,
pd[i].reg_value,
name, event_list[i].expected_value);
if (pd[i].reg_value != event_list[i].expected_value) {
printf("error: Result should be %lu for %s\n", event_list[i].expected_value, name);
break;
}
}
/*
* let's stop this now
*/
close(id);
free(test_data);
free(test_data_fake);
return 0;
}