2011-09-24 04:29:55 +04:00
|
|
|
/*
|
|
|
|
* Linux performance counter support for MIPS.
|
|
|
|
*
|
|
|
|
* Copyright (C) 2010 MIPS Technologies, Inc.
|
2011-09-24 04:29:55 +04:00
|
|
|
* Copyright (C) 2011 Cavium Networks, Inc.
|
2011-09-24 04:29:55 +04:00
|
|
|
* Author: Deng-Cheng Zhu
|
|
|
|
*
|
|
|
|
* This code is based on the implementation for ARM, which is in turn
|
|
|
|
* based on the sparc64 perf event code and the x86 code. Performance
|
|
|
|
* counter access is based on the MIPS Oprofile code. And the callchain
|
|
|
|
* support references the code of MIPS stacktrace.c.
|
|
|
|
*
|
|
|
|
* This program is free software; you can redistribute it and/or modify
|
|
|
|
* it under the terms of the GNU General Public License version 2 as
|
|
|
|
* published by the Free Software Foundation.
|
|
|
|
*/
|
|
|
|
|
|
|
|
#include <linux/cpumask.h>
|
|
|
|
#include <linux/interrupt.h>
|
|
|
|
#include <linux/smp.h>
|
|
|
|
#include <linux/kernel.h>
|
|
|
|
#include <linux/perf_event.h>
|
|
|
|
#include <linux/uaccess.h>
|
|
|
|
|
|
|
|
#include <asm/irq.h>
|
|
|
|
#include <asm/irq_regs.h>
|
|
|
|
#include <asm/stacktrace.h>
|
|
|
|
#include <asm/time.h> /* For perf_irq */
|
|
|
|
|
|
|
|
#define MIPS_MAX_HWEVENTS 4
|
2012-07-14 00:44:54 +04:00
|
|
|
#define MIPS_TCS_PER_COUNTER 2
|
|
|
|
#define MIPS_CPUID_TO_COUNTER_MASK (MIPS_TCS_PER_COUNTER - 1)
|
2011-09-24 04:29:55 +04:00
|
|
|
|
|
|
|
struct cpu_hw_events {
|
|
|
|
/* Array of events on this cpu. */
|
|
|
|
struct perf_event *events[MIPS_MAX_HWEVENTS];
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Set the bit (indexed by the counter number) when the counter
|
|
|
|
* is used for an event.
|
|
|
|
*/
|
|
|
|
unsigned long used_mask[BITS_TO_LONGS(MIPS_MAX_HWEVENTS)];
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Software copy of the control register for each performance counter.
|
|
|
|
* MIPS CPUs vary in performance counters. They use this differently,
|
|
|
|
* and even may not use it.
|
|
|
|
*/
|
|
|
|
unsigned int saved_ctrl[MIPS_MAX_HWEVENTS];
|
|
|
|
};
|
|
|
|
DEFINE_PER_CPU(struct cpu_hw_events, cpu_hw_events) = {
|
|
|
|
.saved_ctrl = {0},
|
|
|
|
};
|
|
|
|
|
|
|
|
/* The description of MIPS performance events. */
|
|
|
|
struct mips_perf_event {
|
|
|
|
unsigned int event_id;
|
|
|
|
/*
|
|
|
|
* MIPS performance counters are indexed starting from 0.
|
|
|
|
* CNTR_EVEN indicates the indexes of the counters to be used are
|
|
|
|
* even numbers.
|
|
|
|
*/
|
|
|
|
unsigned int cntr_mask;
|
|
|
|
#define CNTR_EVEN 0x55555555
|
|
|
|
#define CNTR_ODD 0xaaaaaaaa
|
2011-09-24 04:29:55 +04:00
|
|
|
#define CNTR_ALL 0xffffffff
|
2011-09-24 04:29:55 +04:00
|
|
|
#ifdef CONFIG_MIPS_MT_SMP
|
|
|
|
enum {
|
|
|
|
T = 0,
|
|
|
|
V = 1,
|
|
|
|
P = 2,
|
|
|
|
} range;
|
|
|
|
#else
|
|
|
|
#define T
|
|
|
|
#define V
|
|
|
|
#define P
|
|
|
|
#endif
|
|
|
|
};
|
|
|
|
|
|
|
|
static struct mips_perf_event raw_event;
|
|
|
|
static DEFINE_MUTEX(raw_event_mutex);
|
|
|
|
|
|
|
|
#define C(x) PERF_COUNT_HW_CACHE_##x
|
|
|
|
|
|
|
|
struct mips_pmu {
|
2011-09-24 04:29:55 +04:00
|
|
|
u64 max_period;
|
|
|
|
u64 valid_count;
|
|
|
|
u64 overflow;
|
2011-09-24 04:29:55 +04:00
|
|
|
const char *name;
|
|
|
|
int irq;
|
|
|
|
u64 (*read_counter)(unsigned int idx);
|
|
|
|
void (*write_counter)(unsigned int idx, u64 val);
|
|
|
|
const struct mips_perf_event *(*map_raw_event)(u64 config);
|
|
|
|
const struct mips_perf_event (*general_event_map)[PERF_COUNT_HW_MAX];
|
|
|
|
const struct mips_perf_event (*cache_event_map)
|
|
|
|
[PERF_COUNT_HW_CACHE_MAX]
|
|
|
|
[PERF_COUNT_HW_CACHE_OP_MAX]
|
|
|
|
[PERF_COUNT_HW_CACHE_RESULT_MAX];
|
|
|
|
unsigned int num_counters;
|
|
|
|
};
|
|
|
|
|
2011-09-24 04:29:55 +04:00
|
|
|
static struct mips_pmu mipspmu;
|
|
|
|
|
|
|
|
#define M_CONFIG1_PC (1 << 4)
|
|
|
|
|
2013-01-22 15:59:30 +04:00
|
|
|
#define M_PERFCTL_EXL (1 << 0)
|
|
|
|
#define M_PERFCTL_KERNEL (1 << 1)
|
|
|
|
#define M_PERFCTL_SUPERVISOR (1 << 2)
|
|
|
|
#define M_PERFCTL_USER (1 << 3)
|
|
|
|
#define M_PERFCTL_INTERRUPT_ENABLE (1 << 4)
|
2011-09-24 04:29:55 +04:00
|
|
|
#define M_PERFCTL_EVENT(event) (((event) & 0x3ff) << 5)
|
2013-01-22 15:59:30 +04:00
|
|
|
#define M_PERFCTL_VPEID(vpe) ((vpe) << 16)
|
2012-07-14 00:44:54 +04:00
|
|
|
|
|
|
|
#ifdef CONFIG_CPU_BMIPS5000
|
|
|
|
#define M_PERFCTL_MT_EN(filter) 0
|
|
|
|
#else /* !CONFIG_CPU_BMIPS5000 */
|
2011-09-24 04:29:55 +04:00
|
|
|
#define M_PERFCTL_MT_EN(filter) ((filter) << 20)
|
2012-07-14 00:44:54 +04:00
|
|
|
#endif /* CONFIG_CPU_BMIPS5000 */
|
|
|
|
|
2013-01-22 15:59:30 +04:00
|
|
|
#define M_TC_EN_ALL M_PERFCTL_MT_EN(0)
|
|
|
|
#define M_TC_EN_VPE M_PERFCTL_MT_EN(1)
|
|
|
|
#define M_TC_EN_TC M_PERFCTL_MT_EN(2)
|
|
|
|
#define M_PERFCTL_TCID(tcid) ((tcid) << 22)
|
|
|
|
#define M_PERFCTL_WIDE (1 << 30)
|
|
|
|
#define M_PERFCTL_MORE (1 << 31)
|
|
|
|
#define M_PERFCTL_TC (1 << 30)
|
2011-09-24 04:29:55 +04:00
|
|
|
|
|
|
|
#define M_PERFCTL_COUNT_EVENT_WHENEVER (M_PERFCTL_EXL | \
|
|
|
|
M_PERFCTL_KERNEL | \
|
|
|
|
M_PERFCTL_USER | \
|
|
|
|
M_PERFCTL_SUPERVISOR | \
|
|
|
|
M_PERFCTL_INTERRUPT_ENABLE)
|
|
|
|
|
|
|
|
#ifdef CONFIG_MIPS_MT_SMP
|
|
|
|
#define M_PERFCTL_CONFIG_MASK 0x3fff801f
|
|
|
|
#else
|
|
|
|
#define M_PERFCTL_CONFIG_MASK 0x1f
|
|
|
|
#endif
|
|
|
|
#define M_PERFCTL_EVENT_MASK 0xfe0
|
|
|
|
|
|
|
|
|
2012-07-14 00:44:53 +04:00
|
|
|
#ifdef CONFIG_MIPS_PERF_SHARED_TC_COUNTERS
|
2011-09-24 04:29:55 +04:00
|
|
|
static int cpu_has_mipsmt_pertccounters;
|
|
|
|
|
|
|
|
static DEFINE_RWLOCK(pmuint_rwlock);
|
|
|
|
|
2012-07-14 00:44:54 +04:00
|
|
|
#if defined(CONFIG_CPU_BMIPS5000)
|
|
|
|
#define vpe_id() (cpu_has_mipsmt_pertccounters ? \
|
|
|
|
0 : (smp_processor_id() & MIPS_CPUID_TO_COUNTER_MASK))
|
|
|
|
#else
|
2011-09-24 04:29:55 +04:00
|
|
|
/*
|
|
|
|
* FIXME: For VSMP, vpe_id() is redefined for Perf-events, because
|
|
|
|
* cpu_data[cpuid].vpe_id reports 0 for _both_ CPUs.
|
|
|
|
*/
|
|
|
|
#define vpe_id() (cpu_has_mipsmt_pertccounters ? \
|
2012-07-14 00:44:54 +04:00
|
|
|
0 : smp_processor_id())
|
|
|
|
#endif
|
2011-09-24 04:29:55 +04:00
|
|
|
|
|
|
|
/* Copied from op_model_mipsxx.c */
|
|
|
|
static unsigned int vpe_shift(void)
|
|
|
|
{
|
|
|
|
if (num_possible_cpus() > 1)
|
|
|
|
return 1;
|
|
|
|
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
static unsigned int counters_total_to_per_cpu(unsigned int counters)
|
|
|
|
{
|
|
|
|
return counters >> vpe_shift();
|
|
|
|
}
|
|
|
|
|
2012-07-14 00:44:53 +04:00
|
|
|
#else /* !CONFIG_MIPS_PERF_SHARED_TC_COUNTERS */
|
2011-09-24 04:29:55 +04:00
|
|
|
#define vpe_id() 0
|
|
|
|
|
2012-07-14 00:44:53 +04:00
|
|
|
#endif /* CONFIG_MIPS_PERF_SHARED_TC_COUNTERS */
|
2011-09-24 04:29:55 +04:00
|
|
|
|
|
|
|
static void resume_local_counters(void);
|
|
|
|
static void pause_local_counters(void);
|
|
|
|
static irqreturn_t mipsxx_pmu_handle_irq(int, void *);
|
|
|
|
static int mipsxx_pmu_handle_shared_irq(void);
|
|
|
|
|
|
|
|
static unsigned int mipsxx_pmu_swizzle_perf_idx(unsigned int idx)
|
|
|
|
{
|
|
|
|
if (vpe_id() == 1)
|
|
|
|
idx = (idx + 2) & 3;
|
|
|
|
return idx;
|
|
|
|
}
|
|
|
|
|
|
|
|
static u64 mipsxx_pmu_read_counter(unsigned int idx)
|
|
|
|
{
|
|
|
|
idx = mipsxx_pmu_swizzle_perf_idx(idx);
|
|
|
|
|
|
|
|
switch (idx) {
|
|
|
|
case 0:
|
|
|
|
/*
|
|
|
|
* The counters are unsigned, we must cast to truncate
|
|
|
|
* off the high bits.
|
|
|
|
*/
|
|
|
|
return (u32)read_c0_perfcntr0();
|
|
|
|
case 1:
|
|
|
|
return (u32)read_c0_perfcntr1();
|
|
|
|
case 2:
|
|
|
|
return (u32)read_c0_perfcntr2();
|
|
|
|
case 3:
|
|
|
|
return (u32)read_c0_perfcntr3();
|
|
|
|
default:
|
|
|
|
WARN_ONCE(1, "Invalid performance counter number (%d)\n", idx);
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
static u64 mipsxx_pmu_read_counter_64(unsigned int idx)
|
|
|
|
{
|
|
|
|
idx = mipsxx_pmu_swizzle_perf_idx(idx);
|
|
|
|
|
|
|
|
switch (idx) {
|
|
|
|
case 0:
|
|
|
|
return read_c0_perfcntr0_64();
|
|
|
|
case 1:
|
|
|
|
return read_c0_perfcntr1_64();
|
|
|
|
case 2:
|
|
|
|
return read_c0_perfcntr2_64();
|
|
|
|
case 3:
|
|
|
|
return read_c0_perfcntr3_64();
|
|
|
|
default:
|
|
|
|
WARN_ONCE(1, "Invalid performance counter number (%d)\n", idx);
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
static void mipsxx_pmu_write_counter(unsigned int idx, u64 val)
|
|
|
|
{
|
|
|
|
idx = mipsxx_pmu_swizzle_perf_idx(idx);
|
|
|
|
|
|
|
|
switch (idx) {
|
|
|
|
case 0:
|
|
|
|
write_c0_perfcntr0(val);
|
|
|
|
return;
|
|
|
|
case 1:
|
|
|
|
write_c0_perfcntr1(val);
|
|
|
|
return;
|
|
|
|
case 2:
|
|
|
|
write_c0_perfcntr2(val);
|
|
|
|
return;
|
|
|
|
case 3:
|
|
|
|
write_c0_perfcntr3(val);
|
|
|
|
return;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
static void mipsxx_pmu_write_counter_64(unsigned int idx, u64 val)
|
|
|
|
{
|
|
|
|
idx = mipsxx_pmu_swizzle_perf_idx(idx);
|
|
|
|
|
|
|
|
switch (idx) {
|
|
|
|
case 0:
|
|
|
|
write_c0_perfcntr0_64(val);
|
|
|
|
return;
|
|
|
|
case 1:
|
|
|
|
write_c0_perfcntr1_64(val);
|
|
|
|
return;
|
|
|
|
case 2:
|
|
|
|
write_c0_perfcntr2_64(val);
|
|
|
|
return;
|
|
|
|
case 3:
|
|
|
|
write_c0_perfcntr3_64(val);
|
|
|
|
return;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
static unsigned int mipsxx_pmu_read_control(unsigned int idx)
|
|
|
|
{
|
|
|
|
idx = mipsxx_pmu_swizzle_perf_idx(idx);
|
|
|
|
|
|
|
|
switch (idx) {
|
|
|
|
case 0:
|
|
|
|
return read_c0_perfctrl0();
|
|
|
|
case 1:
|
|
|
|
return read_c0_perfctrl1();
|
|
|
|
case 2:
|
|
|
|
return read_c0_perfctrl2();
|
|
|
|
case 3:
|
|
|
|
return read_c0_perfctrl3();
|
|
|
|
default:
|
|
|
|
WARN_ONCE(1, "Invalid performance counter number (%d)\n", idx);
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
static void mipsxx_pmu_write_control(unsigned int idx, unsigned int val)
|
|
|
|
{
|
|
|
|
idx = mipsxx_pmu_swizzle_perf_idx(idx);
|
|
|
|
|
|
|
|
switch (idx) {
|
|
|
|
case 0:
|
|
|
|
write_c0_perfctrl0(val);
|
|
|
|
return;
|
|
|
|
case 1:
|
|
|
|
write_c0_perfctrl1(val);
|
|
|
|
return;
|
|
|
|
case 2:
|
|
|
|
write_c0_perfctrl2(val);
|
|
|
|
return;
|
|
|
|
case 3:
|
|
|
|
write_c0_perfctrl3(val);
|
|
|
|
return;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
static int mipsxx_pmu_alloc_counter(struct cpu_hw_events *cpuc,
|
|
|
|
struct hw_perf_event *hwc)
|
|
|
|
{
|
|
|
|
int i;
|
|
|
|
|
|
|
|
/*
|
|
|
|
* We only need to care the counter mask. The range has been
|
|
|
|
* checked definitely.
|
|
|
|
*/
|
|
|
|
unsigned long cntr_mask = (hwc->event_base >> 8) & 0xffff;
|
|
|
|
|
|
|
|
for (i = mipspmu.num_counters - 1; i >= 0; i--) {
|
|
|
|
/*
|
|
|
|
* Note that some MIPS perf events can be counted by both
|
|
|
|
* even and odd counters, wheresas many other are only by
|
|
|
|
* even _or_ odd counters. This introduces an issue that
|
|
|
|
* when the former kind of event takes the counter the
|
|
|
|
* latter kind of event wants to use, then the "counter
|
|
|
|
* allocation" for the latter event will fail. In fact if
|
|
|
|
* they can be dynamically swapped, they both feel happy.
|
|
|
|
* But here we leave this issue alone for now.
|
|
|
|
*/
|
|
|
|
if (test_bit(i, &cntr_mask) &&
|
|
|
|
!test_and_set_bit(i, cpuc->used_mask))
|
|
|
|
return i;
|
|
|
|
}
|
|
|
|
|
|
|
|
return -EAGAIN;
|
|
|
|
}
|
|
|
|
|
|
|
|
static void mipsxx_pmu_enable_event(struct hw_perf_event *evt, int idx)
|
|
|
|
{
|
mips: Replace __get_cpu_var uses
__get_cpu_var() is used for multiple purposes in the kernel source. One of
them is address calculation via the form &__get_cpu_var(x). This calculates
the address for the instance of the percpu variable of the current processor
based on an offset.
Other use cases are for storing and retrieving data from the current
processors percpu area. __get_cpu_var() can be used as an lvalue when
writing data or on the right side of an assignment.
__get_cpu_var() is defined as :
#define __get_cpu_var(var) (*this_cpu_ptr(&(var)))
__get_cpu_var() always only does an address determination. However, store
and retrieve operations could use a segment prefix (or global register on
other platforms) to avoid the address calculation.
this_cpu_write() and this_cpu_read() can directly take an offset into a
percpu area and use optimized assembly code to read and write per cpu
variables.
This patch converts __get_cpu_var into either an explicit address
calculation using this_cpu_ptr() or into a use of this_cpu operations that
use the offset. Thereby address calculations are avoided and less registers
are used when code is generated.
At the end of the patch set all uses of __get_cpu_var have been removed so
the macro is removed too.
The patch set includes passes over all arches as well. Once these operations
are used throughout then specialized macros can be defined in non -x86
arches as well in order to optimize per cpu access by f.e. using a global
register that may be set to the per cpu base.
Transformations done to __get_cpu_var()
1. Determine the address of the percpu instance of the current processor.
DEFINE_PER_CPU(int, y);
int *x = &__get_cpu_var(y);
Converts to
int *x = this_cpu_ptr(&y);
2. Same as #1 but this time an array structure is involved.
DEFINE_PER_CPU(int, y[20]);
int *x = __get_cpu_var(y);
Converts to
int *x = this_cpu_ptr(y);
3. Retrieve the content of the current processors instance of a per cpu
variable.
DEFINE_PER_CPU(int, y);
int x = __get_cpu_var(y)
Converts to
int x = __this_cpu_read(y);
4. Retrieve the content of a percpu struct
DEFINE_PER_CPU(struct mystruct, y);
struct mystruct x = __get_cpu_var(y);
Converts to
memcpy(&x, this_cpu_ptr(&y), sizeof(x));
5. Assignment to a per cpu variable
DEFINE_PER_CPU(int, y)
__get_cpu_var(y) = x;
Converts to
__this_cpu_write(y, x);
6. Increment/Decrement etc of a per cpu variable
DEFINE_PER_CPU(int, y);
__get_cpu_var(y)++
Converts to
__this_cpu_inc(y)
Cc: Ralf Baechle <ralf@linux-mips.org>
Signed-off-by: Christoph Lameter <cl@linux.com>
Signed-off-by: Tejun Heo <tj@kernel.org>
2014-08-17 21:30:44 +04:00
|
|
|
struct cpu_hw_events *cpuc = this_cpu_ptr(&cpu_hw_events);
|
2011-09-24 04:29:55 +04:00
|
|
|
|
|
|
|
WARN_ON(idx < 0 || idx >= mipspmu.num_counters);
|
|
|
|
|
|
|
|
cpuc->saved_ctrl[idx] = M_PERFCTL_EVENT(evt->event_base & 0xff) |
|
|
|
|
(evt->config_base & M_PERFCTL_CONFIG_MASK) |
|
|
|
|
/* Make sure interrupt enabled. */
|
|
|
|
M_PERFCTL_INTERRUPT_ENABLE;
|
2012-07-14 00:44:54 +04:00
|
|
|
if (IS_ENABLED(CONFIG_CPU_BMIPS5000))
|
|
|
|
/* enable the counter for the calling thread */
|
|
|
|
cpuc->saved_ctrl[idx] |=
|
|
|
|
(1 << (12 + vpe_id())) | M_PERFCTL_TC;
|
|
|
|
|
2011-09-24 04:29:55 +04:00
|
|
|
/*
|
|
|
|
* We do not actually let the counter run. Leave it until start().
|
|
|
|
*/
|
|
|
|
}
|
|
|
|
|
|
|
|
static void mipsxx_pmu_disable_event(int idx)
|
|
|
|
{
|
mips: Replace __get_cpu_var uses
__get_cpu_var() is used for multiple purposes in the kernel source. One of
them is address calculation via the form &__get_cpu_var(x). This calculates
the address for the instance of the percpu variable of the current processor
based on an offset.
Other use cases are for storing and retrieving data from the current
processors percpu area. __get_cpu_var() can be used as an lvalue when
writing data or on the right side of an assignment.
__get_cpu_var() is defined as :
#define __get_cpu_var(var) (*this_cpu_ptr(&(var)))
__get_cpu_var() always only does an address determination. However, store
and retrieve operations could use a segment prefix (or global register on
other platforms) to avoid the address calculation.
this_cpu_write() and this_cpu_read() can directly take an offset into a
percpu area and use optimized assembly code to read and write per cpu
variables.
This patch converts __get_cpu_var into either an explicit address
calculation using this_cpu_ptr() or into a use of this_cpu operations that
use the offset. Thereby address calculations are avoided and less registers
are used when code is generated.
At the end of the patch set all uses of __get_cpu_var have been removed so
the macro is removed too.
The patch set includes passes over all arches as well. Once these operations
are used throughout then specialized macros can be defined in non -x86
arches as well in order to optimize per cpu access by f.e. using a global
register that may be set to the per cpu base.
Transformations done to __get_cpu_var()
1. Determine the address of the percpu instance of the current processor.
DEFINE_PER_CPU(int, y);
int *x = &__get_cpu_var(y);
Converts to
int *x = this_cpu_ptr(&y);
2. Same as #1 but this time an array structure is involved.
DEFINE_PER_CPU(int, y[20]);
int *x = __get_cpu_var(y);
Converts to
int *x = this_cpu_ptr(y);
3. Retrieve the content of the current processors instance of a per cpu
variable.
DEFINE_PER_CPU(int, y);
int x = __get_cpu_var(y)
Converts to
int x = __this_cpu_read(y);
4. Retrieve the content of a percpu struct
DEFINE_PER_CPU(struct mystruct, y);
struct mystruct x = __get_cpu_var(y);
Converts to
memcpy(&x, this_cpu_ptr(&y), sizeof(x));
5. Assignment to a per cpu variable
DEFINE_PER_CPU(int, y)
__get_cpu_var(y) = x;
Converts to
__this_cpu_write(y, x);
6. Increment/Decrement etc of a per cpu variable
DEFINE_PER_CPU(int, y);
__get_cpu_var(y)++
Converts to
__this_cpu_inc(y)
Cc: Ralf Baechle <ralf@linux-mips.org>
Signed-off-by: Christoph Lameter <cl@linux.com>
Signed-off-by: Tejun Heo <tj@kernel.org>
2014-08-17 21:30:44 +04:00
|
|
|
struct cpu_hw_events *cpuc = this_cpu_ptr(&cpu_hw_events);
|
2011-09-24 04:29:55 +04:00
|
|
|
unsigned long flags;
|
|
|
|
|
|
|
|
WARN_ON(idx < 0 || idx >= mipspmu.num_counters);
|
|
|
|
|
|
|
|
local_irq_save(flags);
|
|
|
|
cpuc->saved_ctrl[idx] = mipsxx_pmu_read_control(idx) &
|
|
|
|
~M_PERFCTL_COUNT_EVENT_WHENEVER;
|
|
|
|
mipsxx_pmu_write_control(idx, cpuc->saved_ctrl[idx]);
|
|
|
|
local_irq_restore(flags);
|
|
|
|
}
|
2011-09-24 04:29:55 +04:00
|
|
|
|
|
|
|
static int mipspmu_event_set_period(struct perf_event *event,
|
|
|
|
struct hw_perf_event *hwc,
|
|
|
|
int idx)
|
|
|
|
{
|
2011-09-24 04:29:55 +04:00
|
|
|
u64 left = local64_read(&hwc->period_left);
|
|
|
|
u64 period = hwc->sample_period;
|
2011-09-24 04:29:55 +04:00
|
|
|
int ret = 0;
|
|
|
|
|
2011-09-24 04:29:55 +04:00
|
|
|
if (unlikely((left + period) & (1ULL << 63))) {
|
|
|
|
/* left underflowed by more than period. */
|
2011-09-24 04:29:55 +04:00
|
|
|
left = period;
|
|
|
|
local64_set(&hwc->period_left, left);
|
|
|
|
hwc->last_period = period;
|
|
|
|
ret = 1;
|
2011-09-24 04:29:55 +04:00
|
|
|
} else if (unlikely((left + period) <= period)) {
|
|
|
|
/* left underflowed by less than period. */
|
2011-09-24 04:29:55 +04:00
|
|
|
left += period;
|
|
|
|
local64_set(&hwc->period_left, left);
|
|
|
|
hwc->last_period = period;
|
|
|
|
ret = 1;
|
|
|
|
}
|
|
|
|
|
2011-09-24 04:29:55 +04:00
|
|
|
if (left > mipspmu.max_period) {
|
|
|
|
left = mipspmu.max_period;
|
|
|
|
local64_set(&hwc->period_left, left);
|
|
|
|
}
|
2011-09-24 04:29:55 +04:00
|
|
|
|
2011-09-24 04:29:55 +04:00
|
|
|
local64_set(&hwc->prev_count, mipspmu.overflow - left);
|
2011-09-24 04:29:55 +04:00
|
|
|
|
2011-09-24 04:29:55 +04:00
|
|
|
mipspmu.write_counter(idx, mipspmu.overflow - left);
|
2011-09-24 04:29:55 +04:00
|
|
|
|
|
|
|
perf_event_update_userpage(event);
|
|
|
|
|
|
|
|
return ret;
|
|
|
|
}
|
|
|
|
|
|
|
|
static void mipspmu_event_update(struct perf_event *event,
|
|
|
|
struct hw_perf_event *hwc,
|
|
|
|
int idx)
|
|
|
|
{
|
2011-09-24 04:29:55 +04:00
|
|
|
u64 prev_raw_count, new_raw_count;
|
2011-09-24 04:29:55 +04:00
|
|
|
u64 delta;
|
|
|
|
|
|
|
|
again:
|
|
|
|
prev_raw_count = local64_read(&hwc->prev_count);
|
2011-09-24 04:29:55 +04:00
|
|
|
new_raw_count = mipspmu.read_counter(idx);
|
2011-09-24 04:29:55 +04:00
|
|
|
|
|
|
|
if (local64_cmpxchg(&hwc->prev_count, prev_raw_count,
|
|
|
|
new_raw_count) != prev_raw_count)
|
|
|
|
goto again;
|
|
|
|
|
2011-09-24 04:29:55 +04:00
|
|
|
delta = new_raw_count - prev_raw_count;
|
2011-09-24 04:29:55 +04:00
|
|
|
|
|
|
|
local64_add(delta, &event->count);
|
|
|
|
local64_sub(delta, &hwc->period_left);
|
|
|
|
}
|
|
|
|
|
|
|
|
static void mipspmu_start(struct perf_event *event, int flags)
|
|
|
|
{
|
|
|
|
struct hw_perf_event *hwc = &event->hw;
|
|
|
|
|
|
|
|
if (flags & PERF_EF_RELOAD)
|
|
|
|
WARN_ON_ONCE(!(hwc->state & PERF_HES_UPTODATE));
|
|
|
|
|
|
|
|
hwc->state = 0;
|
|
|
|
|
|
|
|
/* Set the period for the event. */
|
|
|
|
mipspmu_event_set_period(event, hwc, hwc->idx);
|
|
|
|
|
|
|
|
/* Enable the event. */
|
2011-09-24 04:29:55 +04:00
|
|
|
mipsxx_pmu_enable_event(hwc, hwc->idx);
|
2011-09-24 04:29:55 +04:00
|
|
|
}
|
|
|
|
|
|
|
|
static void mipspmu_stop(struct perf_event *event, int flags)
|
|
|
|
{
|
|
|
|
struct hw_perf_event *hwc = &event->hw;
|
|
|
|
|
|
|
|
if (!(hwc->state & PERF_HES_STOPPED)) {
|
|
|
|
/* We are working on a local event. */
|
2011-09-24 04:29:55 +04:00
|
|
|
mipsxx_pmu_disable_event(hwc->idx);
|
2011-09-24 04:29:55 +04:00
|
|
|
barrier();
|
|
|
|
mipspmu_event_update(event, hwc, hwc->idx);
|
|
|
|
hwc->state |= PERF_HES_STOPPED | PERF_HES_UPTODATE;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
static int mipspmu_add(struct perf_event *event, int flags)
|
|
|
|
{
|
mips: Replace __get_cpu_var uses
__get_cpu_var() is used for multiple purposes in the kernel source. One of
them is address calculation via the form &__get_cpu_var(x). This calculates
the address for the instance of the percpu variable of the current processor
based on an offset.
Other use cases are for storing and retrieving data from the current
processors percpu area. __get_cpu_var() can be used as an lvalue when
writing data or on the right side of an assignment.
__get_cpu_var() is defined as :
#define __get_cpu_var(var) (*this_cpu_ptr(&(var)))
__get_cpu_var() always only does an address determination. However, store
and retrieve operations could use a segment prefix (or global register on
other platforms) to avoid the address calculation.
this_cpu_write() and this_cpu_read() can directly take an offset into a
percpu area and use optimized assembly code to read and write per cpu
variables.
This patch converts __get_cpu_var into either an explicit address
calculation using this_cpu_ptr() or into a use of this_cpu operations that
use the offset. Thereby address calculations are avoided and less registers
are used when code is generated.
At the end of the patch set all uses of __get_cpu_var have been removed so
the macro is removed too.
The patch set includes passes over all arches as well. Once these operations
are used throughout then specialized macros can be defined in non -x86
arches as well in order to optimize per cpu access by f.e. using a global
register that may be set to the per cpu base.
Transformations done to __get_cpu_var()
1. Determine the address of the percpu instance of the current processor.
DEFINE_PER_CPU(int, y);
int *x = &__get_cpu_var(y);
Converts to
int *x = this_cpu_ptr(&y);
2. Same as #1 but this time an array structure is involved.
DEFINE_PER_CPU(int, y[20]);
int *x = __get_cpu_var(y);
Converts to
int *x = this_cpu_ptr(y);
3. Retrieve the content of the current processors instance of a per cpu
variable.
DEFINE_PER_CPU(int, y);
int x = __get_cpu_var(y)
Converts to
int x = __this_cpu_read(y);
4. Retrieve the content of a percpu struct
DEFINE_PER_CPU(struct mystruct, y);
struct mystruct x = __get_cpu_var(y);
Converts to
memcpy(&x, this_cpu_ptr(&y), sizeof(x));
5. Assignment to a per cpu variable
DEFINE_PER_CPU(int, y)
__get_cpu_var(y) = x;
Converts to
__this_cpu_write(y, x);
6. Increment/Decrement etc of a per cpu variable
DEFINE_PER_CPU(int, y);
__get_cpu_var(y)++
Converts to
__this_cpu_inc(y)
Cc: Ralf Baechle <ralf@linux-mips.org>
Signed-off-by: Christoph Lameter <cl@linux.com>
Signed-off-by: Tejun Heo <tj@kernel.org>
2014-08-17 21:30:44 +04:00
|
|
|
struct cpu_hw_events *cpuc = this_cpu_ptr(&cpu_hw_events);
|
2011-09-24 04:29:55 +04:00
|
|
|
struct hw_perf_event *hwc = &event->hw;
|
|
|
|
int idx;
|
|
|
|
int err = 0;
|
|
|
|
|
|
|
|
perf_pmu_disable(event->pmu);
|
|
|
|
|
|
|
|
/* To look for a free counter for this event. */
|
2011-09-24 04:29:55 +04:00
|
|
|
idx = mipsxx_pmu_alloc_counter(cpuc, hwc);
|
2011-09-24 04:29:55 +04:00
|
|
|
if (idx < 0) {
|
|
|
|
err = idx;
|
|
|
|
goto out;
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* If there is an event in the counter we are going to use then
|
|
|
|
* make sure it is disabled.
|
|
|
|
*/
|
|
|
|
event->hw.idx = idx;
|
2011-09-24 04:29:55 +04:00
|
|
|
mipsxx_pmu_disable_event(idx);
|
2011-09-24 04:29:55 +04:00
|
|
|
cpuc->events[idx] = event;
|
|
|
|
|
|
|
|
hwc->state = PERF_HES_STOPPED | PERF_HES_UPTODATE;
|
|
|
|
if (flags & PERF_EF_START)
|
|
|
|
mipspmu_start(event, PERF_EF_RELOAD);
|
|
|
|
|
|
|
|
/* Propagate our changes to the userspace mapping. */
|
|
|
|
perf_event_update_userpage(event);
|
|
|
|
|
|
|
|
out:
|
|
|
|
perf_pmu_enable(event->pmu);
|
|
|
|
return err;
|
|
|
|
}
|
|
|
|
|
|
|
|
static void mipspmu_del(struct perf_event *event, int flags)
|
|
|
|
{
|
mips: Replace __get_cpu_var uses
__get_cpu_var() is used for multiple purposes in the kernel source. One of
them is address calculation via the form &__get_cpu_var(x). This calculates
the address for the instance of the percpu variable of the current processor
based on an offset.
Other use cases are for storing and retrieving data from the current
processors percpu area. __get_cpu_var() can be used as an lvalue when
writing data or on the right side of an assignment.
__get_cpu_var() is defined as :
#define __get_cpu_var(var) (*this_cpu_ptr(&(var)))
__get_cpu_var() always only does an address determination. However, store
and retrieve operations could use a segment prefix (or global register on
other platforms) to avoid the address calculation.
this_cpu_write() and this_cpu_read() can directly take an offset into a
percpu area and use optimized assembly code to read and write per cpu
variables.
This patch converts __get_cpu_var into either an explicit address
calculation using this_cpu_ptr() or into a use of this_cpu operations that
use the offset. Thereby address calculations are avoided and less registers
are used when code is generated.
At the end of the patch set all uses of __get_cpu_var have been removed so
the macro is removed too.
The patch set includes passes over all arches as well. Once these operations
are used throughout then specialized macros can be defined in non -x86
arches as well in order to optimize per cpu access by f.e. using a global
register that may be set to the per cpu base.
Transformations done to __get_cpu_var()
1. Determine the address of the percpu instance of the current processor.
DEFINE_PER_CPU(int, y);
int *x = &__get_cpu_var(y);
Converts to
int *x = this_cpu_ptr(&y);
2. Same as #1 but this time an array structure is involved.
DEFINE_PER_CPU(int, y[20]);
int *x = __get_cpu_var(y);
Converts to
int *x = this_cpu_ptr(y);
3. Retrieve the content of the current processors instance of a per cpu
variable.
DEFINE_PER_CPU(int, y);
int x = __get_cpu_var(y)
Converts to
int x = __this_cpu_read(y);
4. Retrieve the content of a percpu struct
DEFINE_PER_CPU(struct mystruct, y);
struct mystruct x = __get_cpu_var(y);
Converts to
memcpy(&x, this_cpu_ptr(&y), sizeof(x));
5. Assignment to a per cpu variable
DEFINE_PER_CPU(int, y)
__get_cpu_var(y) = x;
Converts to
__this_cpu_write(y, x);
6. Increment/Decrement etc of a per cpu variable
DEFINE_PER_CPU(int, y);
__get_cpu_var(y)++
Converts to
__this_cpu_inc(y)
Cc: Ralf Baechle <ralf@linux-mips.org>
Signed-off-by: Christoph Lameter <cl@linux.com>
Signed-off-by: Tejun Heo <tj@kernel.org>
2014-08-17 21:30:44 +04:00
|
|
|
struct cpu_hw_events *cpuc = this_cpu_ptr(&cpu_hw_events);
|
2011-09-24 04:29:55 +04:00
|
|
|
struct hw_perf_event *hwc = &event->hw;
|
|
|
|
int idx = hwc->idx;
|
|
|
|
|
2011-09-24 04:29:55 +04:00
|
|
|
WARN_ON(idx < 0 || idx >= mipspmu.num_counters);
|
2011-09-24 04:29:55 +04:00
|
|
|
|
|
|
|
mipspmu_stop(event, PERF_EF_UPDATE);
|
|
|
|
cpuc->events[idx] = NULL;
|
|
|
|
clear_bit(idx, cpuc->used_mask);
|
|
|
|
|
|
|
|
perf_event_update_userpage(event);
|
|
|
|
}
|
|
|
|
|
|
|
|
static void mipspmu_read(struct perf_event *event)
|
|
|
|
{
|
|
|
|
struct hw_perf_event *hwc = &event->hw;
|
|
|
|
|
|
|
|
/* Don't read disabled counters! */
|
|
|
|
if (hwc->idx < 0)
|
|
|
|
return;
|
|
|
|
|
|
|
|
mipspmu_event_update(event, hwc, hwc->idx);
|
|
|
|
}
|
|
|
|
|
|
|
|
static void mipspmu_enable(struct pmu *pmu)
|
|
|
|
{
|
2012-07-14 00:44:53 +04:00
|
|
|
#ifdef CONFIG_MIPS_PERF_SHARED_TC_COUNTERS
|
2011-09-24 04:29:55 +04:00
|
|
|
write_unlock(&pmuint_rwlock);
|
|
|
|
#endif
|
|
|
|
resume_local_counters();
|
2011-09-24 04:29:55 +04:00
|
|
|
}
|
|
|
|
|
2011-09-24 04:29:55 +04:00
|
|
|
/*
|
|
|
|
* MIPS performance counters can be per-TC. The control registers can
|
|
|
|
* not be directly accessed accross CPUs. Hence if we want to do global
|
|
|
|
* control, we need cross CPU calls. on_each_cpu() can help us, but we
|
|
|
|
* can not make sure this function is called with interrupts enabled. So
|
|
|
|
* here we pause local counters and then grab a rwlock and leave the
|
|
|
|
* counters on other CPUs alone. If any counter interrupt raises while
|
|
|
|
* we own the write lock, simply pause local counters on that CPU and
|
|
|
|
* spin in the handler. Also we know we won't be switched to another
|
|
|
|
* CPU after pausing local counters and before grabbing the lock.
|
|
|
|
*/
|
2011-09-24 04:29:55 +04:00
|
|
|
static void mipspmu_disable(struct pmu *pmu)
|
|
|
|
{
|
2011-09-24 04:29:55 +04:00
|
|
|
pause_local_counters();
|
2012-07-14 00:44:53 +04:00
|
|
|
#ifdef CONFIG_MIPS_PERF_SHARED_TC_COUNTERS
|
2011-09-24 04:29:55 +04:00
|
|
|
write_lock(&pmuint_rwlock);
|
|
|
|
#endif
|
2011-09-24 04:29:55 +04:00
|
|
|
}
|
|
|
|
|
|
|
|
static atomic_t active_events = ATOMIC_INIT(0);
|
|
|
|
static DEFINE_MUTEX(pmu_reserve_mutex);
|
|
|
|
static int (*save_perf_irq)(void);
|
|
|
|
|
|
|
|
static int mipspmu_get_irq(void)
|
|
|
|
{
|
|
|
|
int err;
|
|
|
|
|
2011-09-24 04:29:55 +04:00
|
|
|
if (mipspmu.irq >= 0) {
|
2011-09-24 04:29:55 +04:00
|
|
|
/* Request my own irq handler. */
|
2011-09-24 04:29:55 +04:00
|
|
|
err = request_irq(mipspmu.irq, mipsxx_pmu_handle_irq,
|
2015-01-28 00:45:53 +03:00
|
|
|
IRQF_PERCPU | IRQF_NOBALANCING |
|
|
|
|
IRQF_NO_THREAD | IRQF_NO_SUSPEND |
|
|
|
|
IRQF_SHARED,
|
|
|
|
"mips_perf_pmu", &mipspmu);
|
2011-09-24 04:29:55 +04:00
|
|
|
if (err) {
|
2014-10-04 20:50:42 +04:00
|
|
|
pr_warn("Unable to request IRQ%d for MIPS performance counters!\n",
|
|
|
|
mipspmu.irq);
|
2011-09-24 04:29:55 +04:00
|
|
|
}
|
|
|
|
} else if (cp0_perfcount_irq < 0) {
|
|
|
|
/*
|
|
|
|
* We are sharing the irq number with the timer interrupt.
|
|
|
|
*/
|
|
|
|
save_perf_irq = perf_irq;
|
2011-09-24 04:29:55 +04:00
|
|
|
perf_irq = mipsxx_pmu_handle_shared_irq;
|
2011-09-24 04:29:55 +04:00
|
|
|
err = 0;
|
|
|
|
} else {
|
2014-10-04 20:50:42 +04:00
|
|
|
pr_warn("The platform hasn't properly defined its interrupt controller\n");
|
2011-09-24 04:29:55 +04:00
|
|
|
err = -ENOENT;
|
|
|
|
}
|
|
|
|
|
|
|
|
return err;
|
|
|
|
}
|
|
|
|
|
|
|
|
static void mipspmu_free_irq(void)
|
|
|
|
{
|
2011-09-24 04:29:55 +04:00
|
|
|
if (mipspmu.irq >= 0)
|
2015-01-28 00:45:53 +03:00
|
|
|
free_irq(mipspmu.irq, &mipspmu);
|
2011-09-24 04:29:55 +04:00
|
|
|
else if (cp0_perfcount_irq < 0)
|
|
|
|
perf_irq = save_perf_irq;
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* mipsxx/rm9000/loongson2 have different performance counters, they have
|
|
|
|
* specific low-level init routines.
|
|
|
|
*/
|
|
|
|
static void reset_counters(void *arg);
|
|
|
|
static int __hw_perf_event_init(struct perf_event *event);
|
|
|
|
|
|
|
|
static void hw_perf_event_destroy(struct perf_event *event)
|
|
|
|
{
|
|
|
|
if (atomic_dec_and_mutex_lock(&active_events,
|
|
|
|
&pmu_reserve_mutex)) {
|
|
|
|
/*
|
|
|
|
* We must not call the destroy function with interrupts
|
|
|
|
* disabled.
|
|
|
|
*/
|
|
|
|
on_each_cpu(reset_counters,
|
2011-09-24 04:29:55 +04:00
|
|
|
(void *)(long)mipspmu.num_counters, 1);
|
2011-09-24 04:29:55 +04:00
|
|
|
mipspmu_free_irq();
|
|
|
|
mutex_unlock(&pmu_reserve_mutex);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
static int mipspmu_event_init(struct perf_event *event)
|
|
|
|
{
|
|
|
|
int err = 0;
|
|
|
|
|
2012-02-10 02:20:59 +04:00
|
|
|
/* does not support taken branch sampling */
|
|
|
|
if (has_branch_stack(event))
|
|
|
|
return -EOPNOTSUPP;
|
|
|
|
|
2011-09-24 04:29:55 +04:00
|
|
|
switch (event->attr.type) {
|
|
|
|
case PERF_TYPE_RAW:
|
|
|
|
case PERF_TYPE_HARDWARE:
|
|
|
|
case PERF_TYPE_HW_CACHE:
|
|
|
|
break;
|
|
|
|
|
|
|
|
default:
|
|
|
|
return -ENOENT;
|
|
|
|
}
|
|
|
|
|
2011-09-24 04:29:55 +04:00
|
|
|
if (event->cpu >= nr_cpumask_bits ||
|
|
|
|
(event->cpu >= 0 && !cpu_online(event->cpu)))
|
2011-09-24 04:29:55 +04:00
|
|
|
return -ENODEV;
|
|
|
|
|
|
|
|
if (!atomic_inc_not_zero(&active_events)) {
|
|
|
|
mutex_lock(&pmu_reserve_mutex);
|
|
|
|
if (atomic_read(&active_events) == 0)
|
|
|
|
err = mipspmu_get_irq();
|
|
|
|
|
|
|
|
if (!err)
|
|
|
|
atomic_inc(&active_events);
|
|
|
|
mutex_unlock(&pmu_reserve_mutex);
|
|
|
|
}
|
|
|
|
|
|
|
|
if (err)
|
|
|
|
return err;
|
|
|
|
|
2011-11-21 23:28:48 +04:00
|
|
|
return __hw_perf_event_init(event);
|
2011-09-24 04:29:55 +04:00
|
|
|
}
|
|
|
|
|
|
|
|
static struct pmu pmu = {
|
|
|
|
.pmu_enable = mipspmu_enable,
|
|
|
|
.pmu_disable = mipspmu_disable,
|
|
|
|
.event_init = mipspmu_event_init,
|
|
|
|
.add = mipspmu_add,
|
|
|
|
.del = mipspmu_del,
|
|
|
|
.start = mipspmu_start,
|
|
|
|
.stop = mipspmu_stop,
|
|
|
|
.read = mipspmu_read,
|
|
|
|
};
|
|
|
|
|
|
|
|
static unsigned int mipspmu_perf_event_encode(const struct mips_perf_event *pev)
|
|
|
|
{
|
|
|
|
/*
|
|
|
|
* Top 8 bits for range, next 16 bits for cntr_mask, lowest 8 bits for
|
|
|
|
* event_id.
|
|
|
|
*/
|
|
|
|
#ifdef CONFIG_MIPS_MT_SMP
|
|
|
|
return ((unsigned int)pev->range << 24) |
|
|
|
|
(pev->cntr_mask & 0xffff00) |
|
|
|
|
(pev->event_id & 0xff);
|
|
|
|
#else
|
|
|
|
return (pev->cntr_mask & 0xffff00) |
|
|
|
|
(pev->event_id & 0xff);
|
|
|
|
#endif
|
|
|
|
}
|
|
|
|
|
|
|
|
static const struct mips_perf_event *mipspmu_map_general_event(int idx)
|
|
|
|
{
|
|
|
|
|
2012-07-14 00:44:50 +04:00
|
|
|
if ((*mipspmu.general_event_map)[idx].cntr_mask == 0)
|
|
|
|
return ERR_PTR(-EOPNOTSUPP);
|
|
|
|
return &(*mipspmu.general_event_map)[idx];
|
2011-09-24 04:29:55 +04:00
|
|
|
}
|
|
|
|
|
|
|
|
static const struct mips_perf_event *mipspmu_map_cache_event(u64 config)
|
|
|
|
{
|
|
|
|
unsigned int cache_type, cache_op, cache_result;
|
|
|
|
const struct mips_perf_event *pev;
|
|
|
|
|
|
|
|
cache_type = (config >> 0) & 0xff;
|
|
|
|
if (cache_type >= PERF_COUNT_HW_CACHE_MAX)
|
|
|
|
return ERR_PTR(-EINVAL);
|
|
|
|
|
|
|
|
cache_op = (config >> 8) & 0xff;
|
|
|
|
if (cache_op >= PERF_COUNT_HW_CACHE_OP_MAX)
|
|
|
|
return ERR_PTR(-EINVAL);
|
|
|
|
|
|
|
|
cache_result = (config >> 16) & 0xff;
|
|
|
|
if (cache_result >= PERF_COUNT_HW_CACHE_RESULT_MAX)
|
|
|
|
return ERR_PTR(-EINVAL);
|
|
|
|
|
2011-09-24 04:29:55 +04:00
|
|
|
pev = &((*mipspmu.cache_event_map)
|
2011-09-24 04:29:55 +04:00
|
|
|
[cache_type]
|
|
|
|
[cache_op]
|
|
|
|
[cache_result]);
|
|
|
|
|
2012-07-14 00:44:50 +04:00
|
|
|
if (pev->cntr_mask == 0)
|
2011-09-24 04:29:55 +04:00
|
|
|
return ERR_PTR(-EOPNOTSUPP);
|
|
|
|
|
|
|
|
return pev;
|
|
|
|
|
|
|
|
}
|
|
|
|
|
|
|
|
static int validate_group(struct perf_event *event)
|
|
|
|
{
|
|
|
|
struct perf_event *sibling, *leader = event->group_leader;
|
|
|
|
struct cpu_hw_events fake_cpuc;
|
|
|
|
|
|
|
|
memset(&fake_cpuc, 0, sizeof(fake_cpuc));
|
|
|
|
|
2011-11-21 23:28:47 +04:00
|
|
|
if (mipsxx_pmu_alloc_counter(&fake_cpuc, &leader->hw) < 0)
|
2011-11-09 20:56:37 +04:00
|
|
|
return -EINVAL;
|
2011-09-24 04:29:55 +04:00
|
|
|
|
|
|
|
list_for_each_entry(sibling, &leader->sibling_list, group_entry) {
|
2011-11-21 23:28:47 +04:00
|
|
|
if (mipsxx_pmu_alloc_counter(&fake_cpuc, &sibling->hw) < 0)
|
2011-11-09 20:56:37 +04:00
|
|
|
return -EINVAL;
|
2011-09-24 04:29:55 +04:00
|
|
|
}
|
|
|
|
|
2011-11-21 23:28:47 +04:00
|
|
|
if (mipsxx_pmu_alloc_counter(&fake_cpuc, &event->hw) < 0)
|
2011-11-09 20:56:37 +04:00
|
|
|
return -EINVAL;
|
2011-09-24 04:29:55 +04:00
|
|
|
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
/* This is needed by specific irq handlers in perf_event_*.c */
|
|
|
|
static void handle_associated_event(struct cpu_hw_events *cpuc,
|
|
|
|
int idx, struct perf_sample_data *data,
|
|
|
|
struct pt_regs *regs)
|
|
|
|
{
|
|
|
|
struct perf_event *event = cpuc->events[idx];
|
|
|
|
struct hw_perf_event *hwc = &event->hw;
|
|
|
|
|
|
|
|
mipspmu_event_update(event, hwc, idx);
|
|
|
|
data->period = event->hw.last_period;
|
|
|
|
if (!mipspmu_event_set_period(event, hwc, idx))
|
|
|
|
return;
|
|
|
|
|
|
|
|
if (perf_event_overflow(event, data, regs))
|
2011-09-24 04:29:55 +04:00
|
|
|
mipsxx_pmu_disable_event(idx);
|
2011-09-24 04:29:55 +04:00
|
|
|
}
|
2010-10-12 15:37:24 +04:00
|
|
|
|
|
|
|
|
2011-09-24 04:29:55 +04:00
|
|
|
static int __n_counters(void)
|
2010-10-12 15:37:24 +04:00
|
|
|
{
|
|
|
|
if (!(read_c0_config1() & M_CONFIG1_PC))
|
|
|
|
return 0;
|
|
|
|
if (!(read_c0_perfctrl0() & M_PERFCTL_MORE))
|
|
|
|
return 1;
|
|
|
|
if (!(read_c0_perfctrl1() & M_PERFCTL_MORE))
|
|
|
|
return 2;
|
|
|
|
if (!(read_c0_perfctrl2() & M_PERFCTL_MORE))
|
|
|
|
return 3;
|
|
|
|
|
|
|
|
return 4;
|
|
|
|
}
|
|
|
|
|
2011-09-24 04:29:55 +04:00
|
|
|
static int n_counters(void)
|
2010-10-12 15:37:24 +04:00
|
|
|
{
|
|
|
|
int counters;
|
|
|
|
|
|
|
|
switch (current_cpu_type()) {
|
|
|
|
case CPU_R10000:
|
|
|
|
counters = 2;
|
|
|
|
break;
|
|
|
|
|
|
|
|
case CPU_R12000:
|
|
|
|
case CPU_R14000:
|
2015-01-21 15:59:45 +03:00
|
|
|
case CPU_R16000:
|
2010-10-12 15:37:24 +04:00
|
|
|
counters = 4;
|
|
|
|
break;
|
|
|
|
|
|
|
|
default:
|
|
|
|
counters = __n_counters();
|
|
|
|
}
|
|
|
|
|
|
|
|
return counters;
|
|
|
|
}
|
|
|
|
|
|
|
|
static void reset_counters(void *arg)
|
|
|
|
{
|
|
|
|
int counters = (int)(long)arg;
|
|
|
|
switch (counters) {
|
|
|
|
case 4:
|
2011-09-24 04:29:55 +04:00
|
|
|
mipsxx_pmu_write_control(3, 0);
|
|
|
|
mipspmu.write_counter(3, 0);
|
2010-10-12 15:37:24 +04:00
|
|
|
case 3:
|
2011-09-24 04:29:55 +04:00
|
|
|
mipsxx_pmu_write_control(2, 0);
|
|
|
|
mipspmu.write_counter(2, 0);
|
2010-10-12 15:37:24 +04:00
|
|
|
case 2:
|
2011-09-24 04:29:55 +04:00
|
|
|
mipsxx_pmu_write_control(1, 0);
|
|
|
|
mipspmu.write_counter(1, 0);
|
2010-10-12 15:37:24 +04:00
|
|
|
case 1:
|
2011-09-24 04:29:55 +04:00
|
|
|
mipsxx_pmu_write_control(0, 0);
|
|
|
|
mipspmu.write_counter(0, 0);
|
2010-10-12 15:37:24 +04:00
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2014-02-10 21:48:54 +04:00
|
|
|
/* 24K/34K/1004K/interAptiv/loongson1 cores share the same event map. */
|
2010-10-12 15:37:24 +04:00
|
|
|
static const struct mips_perf_event mipsxxcore_event_map
|
|
|
|
[PERF_COUNT_HW_MAX] = {
|
|
|
|
[PERF_COUNT_HW_CPU_CYCLES] = { 0x00, CNTR_EVEN | CNTR_ODD, P },
|
|
|
|
[PERF_COUNT_HW_INSTRUCTIONS] = { 0x01, CNTR_EVEN | CNTR_ODD, T },
|
|
|
|
[PERF_COUNT_HW_BRANCH_INSTRUCTIONS] = { 0x02, CNTR_EVEN, T },
|
|
|
|
[PERF_COUNT_HW_BRANCH_MISSES] = { 0x02, CNTR_ODD, T },
|
|
|
|
};
|
|
|
|
|
2014-02-10 21:48:53 +04:00
|
|
|
/* 74K/proAptiv core has different branch event code. */
|
2014-02-10 21:48:52 +04:00
|
|
|
static const struct mips_perf_event mipsxxcore_event_map2
|
2010-10-12 15:37:24 +04:00
|
|
|
[PERF_COUNT_HW_MAX] = {
|
|
|
|
[PERF_COUNT_HW_CPU_CYCLES] = { 0x00, CNTR_EVEN | CNTR_ODD, P },
|
|
|
|
[PERF_COUNT_HW_INSTRUCTIONS] = { 0x01, CNTR_EVEN | CNTR_ODD, T },
|
|
|
|
[PERF_COUNT_HW_BRANCH_INSTRUCTIONS] = { 0x27, CNTR_EVEN, T },
|
|
|
|
[PERF_COUNT_HW_BRANCH_MISSES] = { 0x27, CNTR_ODD, T },
|
|
|
|
};
|
|
|
|
|
2015-03-29 05:54:08 +03:00
|
|
|
static const struct mips_perf_event loongson3_event_map[PERF_COUNT_HW_MAX] = {
|
|
|
|
[PERF_COUNT_HW_CPU_CYCLES] = { 0x00, CNTR_EVEN },
|
|
|
|
[PERF_COUNT_HW_INSTRUCTIONS] = { 0x00, CNTR_ODD },
|
|
|
|
[PERF_COUNT_HW_BRANCH_INSTRUCTIONS] = { 0x01, CNTR_EVEN },
|
|
|
|
[PERF_COUNT_HW_BRANCH_MISSES] = { 0x01, CNTR_ODD },
|
|
|
|
};
|
|
|
|
|
2011-09-24 04:29:55 +04:00
|
|
|
static const struct mips_perf_event octeon_event_map[PERF_COUNT_HW_MAX] = {
|
|
|
|
[PERF_COUNT_HW_CPU_CYCLES] = { 0x01, CNTR_ALL },
|
|
|
|
[PERF_COUNT_HW_INSTRUCTIONS] = { 0x03, CNTR_ALL },
|
|
|
|
[PERF_COUNT_HW_CACHE_REFERENCES] = { 0x2b, CNTR_ALL },
|
2013-01-22 15:59:30 +04:00
|
|
|
[PERF_COUNT_HW_CACHE_MISSES] = { 0x2e, CNTR_ALL },
|
2011-09-24 04:29:55 +04:00
|
|
|
[PERF_COUNT_HW_BRANCH_INSTRUCTIONS] = { 0x08, CNTR_ALL },
|
|
|
|
[PERF_COUNT_HW_BRANCH_MISSES] = { 0x09, CNTR_ALL },
|
|
|
|
[PERF_COUNT_HW_BUS_CYCLES] = { 0x25, CNTR_ALL },
|
|
|
|
};
|
|
|
|
|
2012-07-14 00:44:54 +04:00
|
|
|
static const struct mips_perf_event bmips5000_event_map
|
|
|
|
[PERF_COUNT_HW_MAX] = {
|
|
|
|
[PERF_COUNT_HW_CPU_CYCLES] = { 0x00, CNTR_EVEN | CNTR_ODD, T },
|
|
|
|
[PERF_COUNT_HW_INSTRUCTIONS] = { 0x01, CNTR_EVEN | CNTR_ODD, T },
|
|
|
|
[PERF_COUNT_HW_BRANCH_MISSES] = { 0x02, CNTR_ODD, T },
|
|
|
|
};
|
|
|
|
|
2012-10-31 16:01:28 +04:00
|
|
|
static const struct mips_perf_event xlp_event_map[PERF_COUNT_HW_MAX] = {
|
|
|
|
[PERF_COUNT_HW_CPU_CYCLES] = { 0x01, CNTR_ALL },
|
|
|
|
[PERF_COUNT_HW_INSTRUCTIONS] = { 0x18, CNTR_ALL }, /* PAPI_TOT_INS */
|
|
|
|
[PERF_COUNT_HW_CACHE_REFERENCES] = { 0x04, CNTR_ALL }, /* PAPI_L1_ICA */
|
|
|
|
[PERF_COUNT_HW_CACHE_MISSES] = { 0x07, CNTR_ALL }, /* PAPI_L1_ICM */
|
|
|
|
[PERF_COUNT_HW_BRANCH_INSTRUCTIONS] = { 0x1b, CNTR_ALL }, /* PAPI_BR_CN */
|
|
|
|
[PERF_COUNT_HW_BRANCH_MISSES] = { 0x1c, CNTR_ALL }, /* PAPI_BR_MSP */
|
|
|
|
};
|
|
|
|
|
2014-02-10 21:48:54 +04:00
|
|
|
/* 24K/34K/1004K/interAptiv/loongson1 cores share the same cache event map. */
|
2010-10-12 15:37:24 +04:00
|
|
|
static const struct mips_perf_event mipsxxcore_cache_map
|
|
|
|
[PERF_COUNT_HW_CACHE_MAX]
|
|
|
|
[PERF_COUNT_HW_CACHE_OP_MAX]
|
|
|
|
[PERF_COUNT_HW_CACHE_RESULT_MAX] = {
|
|
|
|
[C(L1D)] = {
|
|
|
|
/*
|
|
|
|
* Like some other architectures (e.g. ARM), the performance
|
|
|
|
* counters don't differentiate between read and write
|
|
|
|
* accesses/misses, so this isn't strictly correct, but it's the
|
|
|
|
* best we can do. Writes and reads get combined.
|
|
|
|
*/
|
|
|
|
[C(OP_READ)] = {
|
|
|
|
[C(RESULT_ACCESS)] = { 0x0a, CNTR_EVEN, T },
|
|
|
|
[C(RESULT_MISS)] = { 0x0b, CNTR_EVEN | CNTR_ODD, T },
|
|
|
|
},
|
|
|
|
[C(OP_WRITE)] = {
|
|
|
|
[C(RESULT_ACCESS)] = { 0x0a, CNTR_EVEN, T },
|
|
|
|
[C(RESULT_MISS)] = { 0x0b, CNTR_EVEN | CNTR_ODD, T },
|
|
|
|
},
|
|
|
|
},
|
|
|
|
[C(L1I)] = {
|
|
|
|
[C(OP_READ)] = {
|
|
|
|
[C(RESULT_ACCESS)] = { 0x09, CNTR_EVEN, T },
|
|
|
|
[C(RESULT_MISS)] = { 0x09, CNTR_ODD, T },
|
|
|
|
},
|
|
|
|
[C(OP_WRITE)] = {
|
|
|
|
[C(RESULT_ACCESS)] = { 0x09, CNTR_EVEN, T },
|
|
|
|
[C(RESULT_MISS)] = { 0x09, CNTR_ODD, T },
|
|
|
|
},
|
|
|
|
[C(OP_PREFETCH)] = {
|
|
|
|
[C(RESULT_ACCESS)] = { 0x14, CNTR_EVEN, T },
|
|
|
|
/*
|
|
|
|
* Note that MIPS has only "hit" events countable for
|
|
|
|
* the prefetch operation.
|
|
|
|
*/
|
|
|
|
},
|
|
|
|
},
|
|
|
|
[C(LL)] = {
|
|
|
|
[C(OP_READ)] = {
|
|
|
|
[C(RESULT_ACCESS)] = { 0x15, CNTR_ODD, P },
|
|
|
|
[C(RESULT_MISS)] = { 0x16, CNTR_EVEN, P },
|
|
|
|
},
|
|
|
|
[C(OP_WRITE)] = {
|
|
|
|
[C(RESULT_ACCESS)] = { 0x15, CNTR_ODD, P },
|
|
|
|
[C(RESULT_MISS)] = { 0x16, CNTR_EVEN, P },
|
|
|
|
},
|
|
|
|
},
|
|
|
|
[C(DTLB)] = {
|
|
|
|
[C(OP_READ)] = {
|
|
|
|
[C(RESULT_ACCESS)] = { 0x06, CNTR_EVEN, T },
|
|
|
|
[C(RESULT_MISS)] = { 0x06, CNTR_ODD, T },
|
|
|
|
},
|
|
|
|
[C(OP_WRITE)] = {
|
|
|
|
[C(RESULT_ACCESS)] = { 0x06, CNTR_EVEN, T },
|
|
|
|
[C(RESULT_MISS)] = { 0x06, CNTR_ODD, T },
|
|
|
|
},
|
|
|
|
},
|
|
|
|
[C(ITLB)] = {
|
|
|
|
[C(OP_READ)] = {
|
|
|
|
[C(RESULT_ACCESS)] = { 0x05, CNTR_EVEN, T },
|
|
|
|
[C(RESULT_MISS)] = { 0x05, CNTR_ODD, T },
|
|
|
|
},
|
|
|
|
[C(OP_WRITE)] = {
|
|
|
|
[C(RESULT_ACCESS)] = { 0x05, CNTR_EVEN, T },
|
|
|
|
[C(RESULT_MISS)] = { 0x05, CNTR_ODD, T },
|
|
|
|
},
|
|
|
|
},
|
|
|
|
[C(BPU)] = {
|
|
|
|
/* Using the same code for *HW_BRANCH* */
|
|
|
|
[C(OP_READ)] = {
|
|
|
|
[C(RESULT_ACCESS)] = { 0x02, CNTR_EVEN, T },
|
|
|
|
[C(RESULT_MISS)] = { 0x02, CNTR_ODD, T },
|
|
|
|
},
|
|
|
|
[C(OP_WRITE)] = {
|
|
|
|
[C(RESULT_ACCESS)] = { 0x02, CNTR_EVEN, T },
|
|
|
|
[C(RESULT_MISS)] = { 0x02, CNTR_ODD, T },
|
|
|
|
},
|
2011-04-23 01:37:06 +04:00
|
|
|
},
|
2010-10-12 15:37:24 +04:00
|
|
|
};
|
|
|
|
|
2014-02-10 21:48:53 +04:00
|
|
|
/* 74K/proAptiv core has completely different cache event map. */
|
2014-02-10 21:48:52 +04:00
|
|
|
static const struct mips_perf_event mipsxxcore_cache_map2
|
2010-10-12 15:37:24 +04:00
|
|
|
[PERF_COUNT_HW_CACHE_MAX]
|
|
|
|
[PERF_COUNT_HW_CACHE_OP_MAX]
|
|
|
|
[PERF_COUNT_HW_CACHE_RESULT_MAX] = {
|
|
|
|
[C(L1D)] = {
|
|
|
|
/*
|
|
|
|
* Like some other architectures (e.g. ARM), the performance
|
|
|
|
* counters don't differentiate between read and write
|
|
|
|
* accesses/misses, so this isn't strictly correct, but it's the
|
|
|
|
* best we can do. Writes and reads get combined.
|
|
|
|
*/
|
|
|
|
[C(OP_READ)] = {
|
|
|
|
[C(RESULT_ACCESS)] = { 0x17, CNTR_ODD, T },
|
|
|
|
[C(RESULT_MISS)] = { 0x18, CNTR_ODD, T },
|
|
|
|
},
|
|
|
|
[C(OP_WRITE)] = {
|
|
|
|
[C(RESULT_ACCESS)] = { 0x17, CNTR_ODD, T },
|
|
|
|
[C(RESULT_MISS)] = { 0x18, CNTR_ODD, T },
|
|
|
|
},
|
|
|
|
},
|
|
|
|
[C(L1I)] = {
|
|
|
|
[C(OP_READ)] = {
|
|
|
|
[C(RESULT_ACCESS)] = { 0x06, CNTR_EVEN, T },
|
|
|
|
[C(RESULT_MISS)] = { 0x06, CNTR_ODD, T },
|
|
|
|
},
|
|
|
|
[C(OP_WRITE)] = {
|
|
|
|
[C(RESULT_ACCESS)] = { 0x06, CNTR_EVEN, T },
|
|
|
|
[C(RESULT_MISS)] = { 0x06, CNTR_ODD, T },
|
|
|
|
},
|
|
|
|
[C(OP_PREFETCH)] = {
|
|
|
|
[C(RESULT_ACCESS)] = { 0x34, CNTR_EVEN, T },
|
|
|
|
/*
|
|
|
|
* Note that MIPS has only "hit" events countable for
|
|
|
|
* the prefetch operation.
|
|
|
|
*/
|
|
|
|
},
|
|
|
|
},
|
|
|
|
[C(LL)] = {
|
|
|
|
[C(OP_READ)] = {
|
|
|
|
[C(RESULT_ACCESS)] = { 0x1c, CNTR_ODD, P },
|
2013-10-08 19:17:48 +04:00
|
|
|
[C(RESULT_MISS)] = { 0x1d, CNTR_EVEN, P },
|
2010-10-12 15:37:24 +04:00
|
|
|
},
|
|
|
|
[C(OP_WRITE)] = {
|
|
|
|
[C(RESULT_ACCESS)] = { 0x1c, CNTR_ODD, P },
|
2013-10-08 19:17:48 +04:00
|
|
|
[C(RESULT_MISS)] = { 0x1d, CNTR_EVEN, P },
|
2010-10-12 15:37:24 +04:00
|
|
|
},
|
|
|
|
},
|
2014-02-10 21:48:53 +04:00
|
|
|
/*
|
|
|
|
* 74K core does not have specific DTLB events. proAptiv core has
|
|
|
|
* "speculative" DTLB events which are numbered 0x63 (even/odd) and
|
|
|
|
* not included here. One can use raw events if really needed.
|
|
|
|
*/
|
2010-10-12 15:37:24 +04:00
|
|
|
[C(ITLB)] = {
|
|
|
|
[C(OP_READ)] = {
|
|
|
|
[C(RESULT_ACCESS)] = { 0x04, CNTR_EVEN, T },
|
|
|
|
[C(RESULT_MISS)] = { 0x04, CNTR_ODD, T },
|
|
|
|
},
|
|
|
|
[C(OP_WRITE)] = {
|
|
|
|
[C(RESULT_ACCESS)] = { 0x04, CNTR_EVEN, T },
|
|
|
|
[C(RESULT_MISS)] = { 0x04, CNTR_ODD, T },
|
|
|
|
},
|
|
|
|
},
|
|
|
|
[C(BPU)] = {
|
|
|
|
/* Using the same code for *HW_BRANCH* */
|
|
|
|
[C(OP_READ)] = {
|
|
|
|
[C(RESULT_ACCESS)] = { 0x27, CNTR_EVEN, T },
|
|
|
|
[C(RESULT_MISS)] = { 0x27, CNTR_ODD, T },
|
|
|
|
},
|
|
|
|
[C(OP_WRITE)] = {
|
|
|
|
[C(RESULT_ACCESS)] = { 0x27, CNTR_EVEN, T },
|
|
|
|
[C(RESULT_MISS)] = { 0x27, CNTR_ODD, T },
|
|
|
|
},
|
2011-04-23 01:37:06 +04:00
|
|
|
},
|
2010-10-12 15:37:24 +04:00
|
|
|
};
|
|
|
|
|
2015-03-29 05:54:08 +03:00
|
|
|
static const struct mips_perf_event loongson3_cache_map
|
|
|
|
[PERF_COUNT_HW_CACHE_MAX]
|
|
|
|
[PERF_COUNT_HW_CACHE_OP_MAX]
|
|
|
|
[PERF_COUNT_HW_CACHE_RESULT_MAX] = {
|
|
|
|
[C(L1D)] = {
|
|
|
|
/*
|
|
|
|
* Like some other architectures (e.g. ARM), the performance
|
|
|
|
* counters don't differentiate between read and write
|
|
|
|
* accesses/misses, so this isn't strictly correct, but it's the
|
|
|
|
* best we can do. Writes and reads get combined.
|
|
|
|
*/
|
|
|
|
[C(OP_READ)] = {
|
|
|
|
[C(RESULT_MISS)] = { 0x04, CNTR_ODD },
|
|
|
|
},
|
|
|
|
[C(OP_WRITE)] = {
|
|
|
|
[C(RESULT_MISS)] = { 0x04, CNTR_ODD },
|
|
|
|
},
|
|
|
|
},
|
|
|
|
[C(L1I)] = {
|
|
|
|
[C(OP_READ)] = {
|
|
|
|
[C(RESULT_MISS)] = { 0x04, CNTR_EVEN },
|
|
|
|
},
|
|
|
|
[C(OP_WRITE)] = {
|
|
|
|
[C(RESULT_MISS)] = { 0x04, CNTR_EVEN },
|
|
|
|
},
|
|
|
|
},
|
|
|
|
[C(DTLB)] = {
|
|
|
|
[C(OP_READ)] = {
|
|
|
|
[C(RESULT_MISS)] = { 0x09, CNTR_ODD },
|
|
|
|
},
|
|
|
|
[C(OP_WRITE)] = {
|
|
|
|
[C(RESULT_MISS)] = { 0x09, CNTR_ODD },
|
|
|
|
},
|
|
|
|
},
|
|
|
|
[C(ITLB)] = {
|
|
|
|
[C(OP_READ)] = {
|
|
|
|
[C(RESULT_MISS)] = { 0x0c, CNTR_ODD },
|
|
|
|
},
|
|
|
|
[C(OP_WRITE)] = {
|
|
|
|
[C(RESULT_MISS)] = { 0x0c, CNTR_ODD },
|
|
|
|
},
|
|
|
|
},
|
|
|
|
[C(BPU)] = {
|
|
|
|
/* Using the same code for *HW_BRANCH* */
|
|
|
|
[C(OP_READ)] = {
|
|
|
|
[C(RESULT_ACCESS)] = { 0x02, CNTR_EVEN },
|
|
|
|
[C(RESULT_MISS)] = { 0x02, CNTR_ODD },
|
|
|
|
},
|
|
|
|
[C(OP_WRITE)] = {
|
|
|
|
[C(RESULT_ACCESS)] = { 0x02, CNTR_EVEN },
|
|
|
|
[C(RESULT_MISS)] = { 0x02, CNTR_ODD },
|
|
|
|
},
|
|
|
|
},
|
|
|
|
};
|
|
|
|
|
2012-07-14 00:44:54 +04:00
|
|
|
/* BMIPS5000 */
|
|
|
|
static const struct mips_perf_event bmips5000_cache_map
|
|
|
|
[PERF_COUNT_HW_CACHE_MAX]
|
|
|
|
[PERF_COUNT_HW_CACHE_OP_MAX]
|
|
|
|
[PERF_COUNT_HW_CACHE_RESULT_MAX] = {
|
|
|
|
[C(L1D)] = {
|
|
|
|
/*
|
|
|
|
* Like some other architectures (e.g. ARM), the performance
|
|
|
|
* counters don't differentiate between read and write
|
|
|
|
* accesses/misses, so this isn't strictly correct, but it's the
|
|
|
|
* best we can do. Writes and reads get combined.
|
|
|
|
*/
|
|
|
|
[C(OP_READ)] = {
|
|
|
|
[C(RESULT_ACCESS)] = { 12, CNTR_EVEN, T },
|
|
|
|
[C(RESULT_MISS)] = { 12, CNTR_ODD, T },
|
|
|
|
},
|
|
|
|
[C(OP_WRITE)] = {
|
|
|
|
[C(RESULT_ACCESS)] = { 12, CNTR_EVEN, T },
|
|
|
|
[C(RESULT_MISS)] = { 12, CNTR_ODD, T },
|
|
|
|
},
|
|
|
|
},
|
|
|
|
[C(L1I)] = {
|
|
|
|
[C(OP_READ)] = {
|
|
|
|
[C(RESULT_ACCESS)] = { 10, CNTR_EVEN, T },
|
|
|
|
[C(RESULT_MISS)] = { 10, CNTR_ODD, T },
|
|
|
|
},
|
|
|
|
[C(OP_WRITE)] = {
|
|
|
|
[C(RESULT_ACCESS)] = { 10, CNTR_EVEN, T },
|
|
|
|
[C(RESULT_MISS)] = { 10, CNTR_ODD, T },
|
|
|
|
},
|
|
|
|
[C(OP_PREFETCH)] = {
|
|
|
|
[C(RESULT_ACCESS)] = { 23, CNTR_EVEN, T },
|
|
|
|
/*
|
|
|
|
* Note that MIPS has only "hit" events countable for
|
|
|
|
* the prefetch operation.
|
|
|
|
*/
|
|
|
|
},
|
|
|
|
},
|
|
|
|
[C(LL)] = {
|
|
|
|
[C(OP_READ)] = {
|
|
|
|
[C(RESULT_ACCESS)] = { 28, CNTR_EVEN, P },
|
|
|
|
[C(RESULT_MISS)] = { 28, CNTR_ODD, P },
|
|
|
|
},
|
|
|
|
[C(OP_WRITE)] = {
|
|
|
|
[C(RESULT_ACCESS)] = { 28, CNTR_EVEN, P },
|
|
|
|
[C(RESULT_MISS)] = { 28, CNTR_ODD, P },
|
|
|
|
},
|
|
|
|
},
|
|
|
|
[C(BPU)] = {
|
|
|
|
/* Using the same code for *HW_BRANCH* */
|
|
|
|
[C(OP_READ)] = {
|
|
|
|
[C(RESULT_MISS)] = { 0x02, CNTR_ODD, T },
|
|
|
|
},
|
|
|
|
[C(OP_WRITE)] = {
|
|
|
|
[C(RESULT_MISS)] = { 0x02, CNTR_ODD, T },
|
|
|
|
},
|
|
|
|
},
|
|
|
|
};
|
|
|
|
|
2011-09-24 04:29:55 +04:00
|
|
|
|
|
|
|
static const struct mips_perf_event octeon_cache_map
|
|
|
|
[PERF_COUNT_HW_CACHE_MAX]
|
|
|
|
[PERF_COUNT_HW_CACHE_OP_MAX]
|
|
|
|
[PERF_COUNT_HW_CACHE_RESULT_MAX] = {
|
|
|
|
[C(L1D)] = {
|
|
|
|
[C(OP_READ)] = {
|
|
|
|
[C(RESULT_ACCESS)] = { 0x2b, CNTR_ALL },
|
|
|
|
[C(RESULT_MISS)] = { 0x2e, CNTR_ALL },
|
|
|
|
},
|
|
|
|
[C(OP_WRITE)] = {
|
|
|
|
[C(RESULT_ACCESS)] = { 0x30, CNTR_ALL },
|
|
|
|
},
|
|
|
|
},
|
|
|
|
[C(L1I)] = {
|
|
|
|
[C(OP_READ)] = {
|
|
|
|
[C(RESULT_ACCESS)] = { 0x18, CNTR_ALL },
|
|
|
|
},
|
|
|
|
[C(OP_PREFETCH)] = {
|
|
|
|
[C(RESULT_ACCESS)] = { 0x19, CNTR_ALL },
|
|
|
|
},
|
|
|
|
},
|
|
|
|
[C(DTLB)] = {
|
|
|
|
/*
|
|
|
|
* Only general DTLB misses are counted use the same event for
|
|
|
|
* read and write.
|
|
|
|
*/
|
|
|
|
[C(OP_READ)] = {
|
|
|
|
[C(RESULT_MISS)] = { 0x35, CNTR_ALL },
|
|
|
|
},
|
|
|
|
[C(OP_WRITE)] = {
|
|
|
|
[C(RESULT_MISS)] = { 0x35, CNTR_ALL },
|
|
|
|
},
|
|
|
|
},
|
|
|
|
[C(ITLB)] = {
|
|
|
|
[C(OP_READ)] = {
|
|
|
|
[C(RESULT_MISS)] = { 0x37, CNTR_ALL },
|
|
|
|
},
|
|
|
|
},
|
|
|
|
};
|
|
|
|
|
2012-10-31 16:01:28 +04:00
|
|
|
static const struct mips_perf_event xlp_cache_map
|
|
|
|
[PERF_COUNT_HW_CACHE_MAX]
|
|
|
|
[PERF_COUNT_HW_CACHE_OP_MAX]
|
|
|
|
[PERF_COUNT_HW_CACHE_RESULT_MAX] = {
|
|
|
|
[C(L1D)] = {
|
|
|
|
[C(OP_READ)] = {
|
|
|
|
[C(RESULT_ACCESS)] = { 0x31, CNTR_ALL }, /* PAPI_L1_DCR */
|
|
|
|
[C(RESULT_MISS)] = { 0x30, CNTR_ALL }, /* PAPI_L1_LDM */
|
|
|
|
},
|
|
|
|
[C(OP_WRITE)] = {
|
|
|
|
[C(RESULT_ACCESS)] = { 0x2f, CNTR_ALL }, /* PAPI_L1_DCW */
|
|
|
|
[C(RESULT_MISS)] = { 0x2e, CNTR_ALL }, /* PAPI_L1_STM */
|
|
|
|
},
|
|
|
|
},
|
|
|
|
[C(L1I)] = {
|
|
|
|
[C(OP_READ)] = {
|
|
|
|
[C(RESULT_ACCESS)] = { 0x04, CNTR_ALL }, /* PAPI_L1_ICA */
|
|
|
|
[C(RESULT_MISS)] = { 0x07, CNTR_ALL }, /* PAPI_L1_ICM */
|
|
|
|
},
|
|
|
|
},
|
|
|
|
[C(LL)] = {
|
|
|
|
[C(OP_READ)] = {
|
|
|
|
[C(RESULT_ACCESS)] = { 0x35, CNTR_ALL }, /* PAPI_L2_DCR */
|
|
|
|
[C(RESULT_MISS)] = { 0x37, CNTR_ALL }, /* PAPI_L2_LDM */
|
|
|
|
},
|
|
|
|
[C(OP_WRITE)] = {
|
|
|
|
[C(RESULT_ACCESS)] = { 0x34, CNTR_ALL }, /* PAPI_L2_DCA */
|
|
|
|
[C(RESULT_MISS)] = { 0x36, CNTR_ALL }, /* PAPI_L2_DCM */
|
|
|
|
},
|
|
|
|
},
|
|
|
|
[C(DTLB)] = {
|
|
|
|
/*
|
|
|
|
* Only general DTLB misses are counted use the same event for
|
|
|
|
* read and write.
|
|
|
|
*/
|
|
|
|
[C(OP_READ)] = {
|
|
|
|
[C(RESULT_MISS)] = { 0x2d, CNTR_ALL }, /* PAPI_TLB_DM */
|
|
|
|
},
|
|
|
|
[C(OP_WRITE)] = {
|
|
|
|
[C(RESULT_MISS)] = { 0x2d, CNTR_ALL }, /* PAPI_TLB_DM */
|
|
|
|
},
|
|
|
|
},
|
|
|
|
[C(ITLB)] = {
|
|
|
|
[C(OP_READ)] = {
|
|
|
|
[C(RESULT_MISS)] = { 0x08, CNTR_ALL }, /* PAPI_TLB_IM */
|
|
|
|
},
|
|
|
|
[C(OP_WRITE)] = {
|
|
|
|
[C(RESULT_MISS)] = { 0x08, CNTR_ALL }, /* PAPI_TLB_IM */
|
|
|
|
},
|
|
|
|
},
|
|
|
|
[C(BPU)] = {
|
|
|
|
[C(OP_READ)] = {
|
|
|
|
[C(RESULT_MISS)] = { 0x25, CNTR_ALL },
|
|
|
|
},
|
|
|
|
},
|
|
|
|
};
|
|
|
|
|
2010-10-12 15:37:24 +04:00
|
|
|
#ifdef CONFIG_MIPS_MT_SMP
|
2011-09-24 04:29:55 +04:00
|
|
|
static void check_and_calc_range(struct perf_event *event,
|
|
|
|
const struct mips_perf_event *pev)
|
2010-10-12 15:37:24 +04:00
|
|
|
{
|
|
|
|
struct hw_perf_event *hwc = &event->hw;
|
|
|
|
|
|
|
|
if (event->cpu >= 0) {
|
|
|
|
if (pev->range > V) {
|
|
|
|
/*
|
|
|
|
* The user selected an event that is processor
|
|
|
|
* wide, while expecting it to be VPE wide.
|
|
|
|
*/
|
|
|
|
hwc->config_base |= M_TC_EN_ALL;
|
|
|
|
} else {
|
|
|
|
/*
|
|
|
|
* FIXME: cpu_data[event->cpu].vpe_id reports 0
|
|
|
|
* for both CPUs.
|
|
|
|
*/
|
|
|
|
hwc->config_base |= M_PERFCTL_VPEID(event->cpu);
|
|
|
|
hwc->config_base |= M_TC_EN_VPE;
|
|
|
|
}
|
|
|
|
} else
|
|
|
|
hwc->config_base |= M_TC_EN_ALL;
|
|
|
|
}
|
|
|
|
#else
|
2011-09-24 04:29:55 +04:00
|
|
|
static void check_and_calc_range(struct perf_event *event,
|
|
|
|
const struct mips_perf_event *pev)
|
2010-10-12 15:37:24 +04:00
|
|
|
{
|
|
|
|
}
|
|
|
|
#endif
|
|
|
|
|
|
|
|
static int __hw_perf_event_init(struct perf_event *event)
|
|
|
|
{
|
|
|
|
struct perf_event_attr *attr = &event->attr;
|
|
|
|
struct hw_perf_event *hwc = &event->hw;
|
|
|
|
const struct mips_perf_event *pev;
|
|
|
|
int err;
|
|
|
|
|
|
|
|
/* Returning MIPS event descriptor for generic perf event. */
|
|
|
|
if (PERF_TYPE_HARDWARE == event->attr.type) {
|
|
|
|
if (event->attr.config >= PERF_COUNT_HW_MAX)
|
|
|
|
return -EINVAL;
|
|
|
|
pev = mipspmu_map_general_event(event->attr.config);
|
|
|
|
} else if (PERF_TYPE_HW_CACHE == event->attr.type) {
|
|
|
|
pev = mipspmu_map_cache_event(event->attr.config);
|
|
|
|
} else if (PERF_TYPE_RAW == event->attr.type) {
|
|
|
|
/* We are working on the global raw event. */
|
|
|
|
mutex_lock(&raw_event_mutex);
|
2011-09-24 04:29:55 +04:00
|
|
|
pev = mipspmu.map_raw_event(event->attr.config);
|
2010-10-12 15:37:24 +04:00
|
|
|
} else {
|
|
|
|
/* The event type is not (yet) supported. */
|
|
|
|
return -EOPNOTSUPP;
|
|
|
|
}
|
|
|
|
|
|
|
|
if (IS_ERR(pev)) {
|
|
|
|
if (PERF_TYPE_RAW == event->attr.type)
|
|
|
|
mutex_unlock(&raw_event_mutex);
|
|
|
|
return PTR_ERR(pev);
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* We allow max flexibility on how each individual counter shared
|
|
|
|
* by the single CPU operates (the mode exclusion and the range).
|
|
|
|
*/
|
|
|
|
hwc->config_base = M_PERFCTL_INTERRUPT_ENABLE;
|
|
|
|
|
|
|
|
/* Calculate range bits and validate it. */
|
|
|
|
if (num_possible_cpus() > 1)
|
|
|
|
check_and_calc_range(event, pev);
|
|
|
|
|
|
|
|
hwc->event_base = mipspmu_perf_event_encode(pev);
|
|
|
|
if (PERF_TYPE_RAW == event->attr.type)
|
|
|
|
mutex_unlock(&raw_event_mutex);
|
|
|
|
|
|
|
|
if (!attr->exclude_user)
|
|
|
|
hwc->config_base |= M_PERFCTL_USER;
|
|
|
|
if (!attr->exclude_kernel) {
|
|
|
|
hwc->config_base |= M_PERFCTL_KERNEL;
|
|
|
|
/* MIPS kernel mode: KSU == 00b || EXL == 1 || ERL == 1 */
|
|
|
|
hwc->config_base |= M_PERFCTL_EXL;
|
|
|
|
}
|
|
|
|
if (!attr->exclude_hv)
|
|
|
|
hwc->config_base |= M_PERFCTL_SUPERVISOR;
|
|
|
|
|
|
|
|
hwc->config_base &= M_PERFCTL_CONFIG_MASK;
|
|
|
|
/*
|
|
|
|
* The event can belong to another cpu. We do not assign a local
|
|
|
|
* counter for it for now.
|
|
|
|
*/
|
|
|
|
hwc->idx = -1;
|
|
|
|
hwc->config = 0;
|
|
|
|
|
|
|
|
if (!hwc->sample_period) {
|
2011-09-24 04:29:55 +04:00
|
|
|
hwc->sample_period = mipspmu.max_period;
|
2010-10-12 15:37:24 +04:00
|
|
|
hwc->last_period = hwc->sample_period;
|
|
|
|
local64_set(&hwc->period_left, hwc->sample_period);
|
|
|
|
}
|
|
|
|
|
|
|
|
err = 0;
|
2011-11-21 23:28:48 +04:00
|
|
|
if (event->group_leader != event)
|
2010-10-12 15:37:24 +04:00
|
|
|
err = validate_group(event);
|
|
|
|
|
|
|
|
event->destroy = hw_perf_event_destroy;
|
2011-11-21 23:28:48 +04:00
|
|
|
|
|
|
|
if (err)
|
|
|
|
event->destroy(event);
|
|
|
|
|
2010-10-12 15:37:24 +04:00
|
|
|
return err;
|
|
|
|
}
|
|
|
|
|
|
|
|
static void pause_local_counters(void)
|
|
|
|
{
|
mips: Replace __get_cpu_var uses
__get_cpu_var() is used for multiple purposes in the kernel source. One of
them is address calculation via the form &__get_cpu_var(x). This calculates
the address for the instance of the percpu variable of the current processor
based on an offset.
Other use cases are for storing and retrieving data from the current
processors percpu area. __get_cpu_var() can be used as an lvalue when
writing data or on the right side of an assignment.
__get_cpu_var() is defined as :
#define __get_cpu_var(var) (*this_cpu_ptr(&(var)))
__get_cpu_var() always only does an address determination. However, store
and retrieve operations could use a segment prefix (or global register on
other platforms) to avoid the address calculation.
this_cpu_write() and this_cpu_read() can directly take an offset into a
percpu area and use optimized assembly code to read and write per cpu
variables.
This patch converts __get_cpu_var into either an explicit address
calculation using this_cpu_ptr() or into a use of this_cpu operations that
use the offset. Thereby address calculations are avoided and less registers
are used when code is generated.
At the end of the patch set all uses of __get_cpu_var have been removed so
the macro is removed too.
The patch set includes passes over all arches as well. Once these operations
are used throughout then specialized macros can be defined in non -x86
arches as well in order to optimize per cpu access by f.e. using a global
register that may be set to the per cpu base.
Transformations done to __get_cpu_var()
1. Determine the address of the percpu instance of the current processor.
DEFINE_PER_CPU(int, y);
int *x = &__get_cpu_var(y);
Converts to
int *x = this_cpu_ptr(&y);
2. Same as #1 but this time an array structure is involved.
DEFINE_PER_CPU(int, y[20]);
int *x = __get_cpu_var(y);
Converts to
int *x = this_cpu_ptr(y);
3. Retrieve the content of the current processors instance of a per cpu
variable.
DEFINE_PER_CPU(int, y);
int x = __get_cpu_var(y)
Converts to
int x = __this_cpu_read(y);
4. Retrieve the content of a percpu struct
DEFINE_PER_CPU(struct mystruct, y);
struct mystruct x = __get_cpu_var(y);
Converts to
memcpy(&x, this_cpu_ptr(&y), sizeof(x));
5. Assignment to a per cpu variable
DEFINE_PER_CPU(int, y)
__get_cpu_var(y) = x;
Converts to
__this_cpu_write(y, x);
6. Increment/Decrement etc of a per cpu variable
DEFINE_PER_CPU(int, y);
__get_cpu_var(y)++
Converts to
__this_cpu_inc(y)
Cc: Ralf Baechle <ralf@linux-mips.org>
Signed-off-by: Christoph Lameter <cl@linux.com>
Signed-off-by: Tejun Heo <tj@kernel.org>
2014-08-17 21:30:44 +04:00
|
|
|
struct cpu_hw_events *cpuc = this_cpu_ptr(&cpu_hw_events);
|
2011-09-24 04:29:55 +04:00
|
|
|
int ctr = mipspmu.num_counters;
|
2010-10-12 15:37:24 +04:00
|
|
|
unsigned long flags;
|
|
|
|
|
|
|
|
local_irq_save(flags);
|
2011-09-24 04:29:55 +04:00
|
|
|
do {
|
|
|
|
ctr--;
|
|
|
|
cpuc->saved_ctrl[ctr] = mipsxx_pmu_read_control(ctr);
|
|
|
|
mipsxx_pmu_write_control(ctr, cpuc->saved_ctrl[ctr] &
|
|
|
|
~M_PERFCTL_COUNT_EVENT_WHENEVER);
|
|
|
|
} while (ctr > 0);
|
2010-10-12 15:37:24 +04:00
|
|
|
local_irq_restore(flags);
|
|
|
|
}
|
|
|
|
|
|
|
|
static void resume_local_counters(void)
|
|
|
|
{
|
mips: Replace __get_cpu_var uses
__get_cpu_var() is used for multiple purposes in the kernel source. One of
them is address calculation via the form &__get_cpu_var(x). This calculates
the address for the instance of the percpu variable of the current processor
based on an offset.
Other use cases are for storing and retrieving data from the current
processors percpu area. __get_cpu_var() can be used as an lvalue when
writing data or on the right side of an assignment.
__get_cpu_var() is defined as :
#define __get_cpu_var(var) (*this_cpu_ptr(&(var)))
__get_cpu_var() always only does an address determination. However, store
and retrieve operations could use a segment prefix (or global register on
other platforms) to avoid the address calculation.
this_cpu_write() and this_cpu_read() can directly take an offset into a
percpu area and use optimized assembly code to read and write per cpu
variables.
This patch converts __get_cpu_var into either an explicit address
calculation using this_cpu_ptr() or into a use of this_cpu operations that
use the offset. Thereby address calculations are avoided and less registers
are used when code is generated.
At the end of the patch set all uses of __get_cpu_var have been removed so
the macro is removed too.
The patch set includes passes over all arches as well. Once these operations
are used throughout then specialized macros can be defined in non -x86
arches as well in order to optimize per cpu access by f.e. using a global
register that may be set to the per cpu base.
Transformations done to __get_cpu_var()
1. Determine the address of the percpu instance of the current processor.
DEFINE_PER_CPU(int, y);
int *x = &__get_cpu_var(y);
Converts to
int *x = this_cpu_ptr(&y);
2. Same as #1 but this time an array structure is involved.
DEFINE_PER_CPU(int, y[20]);
int *x = __get_cpu_var(y);
Converts to
int *x = this_cpu_ptr(y);
3. Retrieve the content of the current processors instance of a per cpu
variable.
DEFINE_PER_CPU(int, y);
int x = __get_cpu_var(y)
Converts to
int x = __this_cpu_read(y);
4. Retrieve the content of a percpu struct
DEFINE_PER_CPU(struct mystruct, y);
struct mystruct x = __get_cpu_var(y);
Converts to
memcpy(&x, this_cpu_ptr(&y), sizeof(x));
5. Assignment to a per cpu variable
DEFINE_PER_CPU(int, y)
__get_cpu_var(y) = x;
Converts to
__this_cpu_write(y, x);
6. Increment/Decrement etc of a per cpu variable
DEFINE_PER_CPU(int, y);
__get_cpu_var(y)++
Converts to
__this_cpu_inc(y)
Cc: Ralf Baechle <ralf@linux-mips.org>
Signed-off-by: Christoph Lameter <cl@linux.com>
Signed-off-by: Tejun Heo <tj@kernel.org>
2014-08-17 21:30:44 +04:00
|
|
|
struct cpu_hw_events *cpuc = this_cpu_ptr(&cpu_hw_events);
|
2011-09-24 04:29:55 +04:00
|
|
|
int ctr = mipspmu.num_counters;
|
2010-10-12 15:37:24 +04:00
|
|
|
|
2011-09-24 04:29:55 +04:00
|
|
|
do {
|
|
|
|
ctr--;
|
|
|
|
mipsxx_pmu_write_control(ctr, cpuc->saved_ctrl[ctr]);
|
|
|
|
} while (ctr > 0);
|
2010-10-12 15:37:24 +04:00
|
|
|
}
|
|
|
|
|
|
|
|
static int mipsxx_pmu_handle_shared_irq(void)
|
|
|
|
{
|
mips: Replace __get_cpu_var uses
__get_cpu_var() is used for multiple purposes in the kernel source. One of
them is address calculation via the form &__get_cpu_var(x). This calculates
the address for the instance of the percpu variable of the current processor
based on an offset.
Other use cases are for storing and retrieving data from the current
processors percpu area. __get_cpu_var() can be used as an lvalue when
writing data or on the right side of an assignment.
__get_cpu_var() is defined as :
#define __get_cpu_var(var) (*this_cpu_ptr(&(var)))
__get_cpu_var() always only does an address determination. However, store
and retrieve operations could use a segment prefix (or global register on
other platforms) to avoid the address calculation.
this_cpu_write() and this_cpu_read() can directly take an offset into a
percpu area and use optimized assembly code to read and write per cpu
variables.
This patch converts __get_cpu_var into either an explicit address
calculation using this_cpu_ptr() or into a use of this_cpu operations that
use the offset. Thereby address calculations are avoided and less registers
are used when code is generated.
At the end of the patch set all uses of __get_cpu_var have been removed so
the macro is removed too.
The patch set includes passes over all arches as well. Once these operations
are used throughout then specialized macros can be defined in non -x86
arches as well in order to optimize per cpu access by f.e. using a global
register that may be set to the per cpu base.
Transformations done to __get_cpu_var()
1. Determine the address of the percpu instance of the current processor.
DEFINE_PER_CPU(int, y);
int *x = &__get_cpu_var(y);
Converts to
int *x = this_cpu_ptr(&y);
2. Same as #1 but this time an array structure is involved.
DEFINE_PER_CPU(int, y[20]);
int *x = __get_cpu_var(y);
Converts to
int *x = this_cpu_ptr(y);
3. Retrieve the content of the current processors instance of a per cpu
variable.
DEFINE_PER_CPU(int, y);
int x = __get_cpu_var(y)
Converts to
int x = __this_cpu_read(y);
4. Retrieve the content of a percpu struct
DEFINE_PER_CPU(struct mystruct, y);
struct mystruct x = __get_cpu_var(y);
Converts to
memcpy(&x, this_cpu_ptr(&y), sizeof(x));
5. Assignment to a per cpu variable
DEFINE_PER_CPU(int, y)
__get_cpu_var(y) = x;
Converts to
__this_cpu_write(y, x);
6. Increment/Decrement etc of a per cpu variable
DEFINE_PER_CPU(int, y);
__get_cpu_var(y)++
Converts to
__this_cpu_inc(y)
Cc: Ralf Baechle <ralf@linux-mips.org>
Signed-off-by: Christoph Lameter <cl@linux.com>
Signed-off-by: Tejun Heo <tj@kernel.org>
2014-08-17 21:30:44 +04:00
|
|
|
struct cpu_hw_events *cpuc = this_cpu_ptr(&cpu_hw_events);
|
2010-10-12 15:37:24 +04:00
|
|
|
struct perf_sample_data data;
|
2011-09-24 04:29:55 +04:00
|
|
|
unsigned int counters = mipspmu.num_counters;
|
|
|
|
u64 counter;
|
2010-10-12 15:37:24 +04:00
|
|
|
int handled = IRQ_NONE;
|
|
|
|
struct pt_regs *regs;
|
|
|
|
|
2012-07-14 00:44:51 +04:00
|
|
|
if (cpu_has_perf_cntr_intr_bit && !(read_c0_cause() & CAUSEF_PCI))
|
2010-10-12 15:37:24 +04:00
|
|
|
return handled;
|
|
|
|
/*
|
|
|
|
* First we pause the local counters, so that when we are locked
|
|
|
|
* here, the counters are all paused. When it gets locked due to
|
|
|
|
* perf_disable(), the timer interrupt handler will be delayed.
|
|
|
|
*
|
|
|
|
* See also mipsxx_pmu_start().
|
|
|
|
*/
|
|
|
|
pause_local_counters();
|
2012-07-14 00:44:53 +04:00
|
|
|
#ifdef CONFIG_MIPS_PERF_SHARED_TC_COUNTERS
|
2010-10-12 15:37:24 +04:00
|
|
|
read_lock(&pmuint_rwlock);
|
|
|
|
#endif
|
|
|
|
|
|
|
|
regs = get_irq_regs();
|
|
|
|
|
2012-04-02 22:19:08 +04:00
|
|
|
perf_sample_data_init(&data, 0, 0);
|
2010-10-12 15:37:24 +04:00
|
|
|
|
|
|
|
switch (counters) {
|
|
|
|
#define HANDLE_COUNTER(n) \
|
|
|
|
case n + 1: \
|
|
|
|
if (test_bit(n, cpuc->used_mask)) { \
|
2011-09-24 04:29:55 +04:00
|
|
|
counter = mipspmu.read_counter(n); \
|
|
|
|
if (counter & mipspmu.overflow) { \
|
|
|
|
handle_associated_event(cpuc, n, &data, regs); \
|
2010-10-12 15:37:24 +04:00
|
|
|
handled = IRQ_HANDLED; \
|
|
|
|
} \
|
|
|
|
}
|
|
|
|
HANDLE_COUNTER(3)
|
|
|
|
HANDLE_COUNTER(2)
|
|
|
|
HANDLE_COUNTER(1)
|
|
|
|
HANDLE_COUNTER(0)
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Do all the work for the pending perf events. We can do this
|
|
|
|
* in here because the performance counter interrupt is a regular
|
|
|
|
* interrupt, not NMI.
|
|
|
|
*/
|
|
|
|
if (handled == IRQ_HANDLED)
|
2011-01-21 11:19:17 +03:00
|
|
|
irq_work_run();
|
2010-10-12 15:37:24 +04:00
|
|
|
|
2012-07-14 00:44:53 +04:00
|
|
|
#ifdef CONFIG_MIPS_PERF_SHARED_TC_COUNTERS
|
2010-10-12 15:37:24 +04:00
|
|
|
read_unlock(&pmuint_rwlock);
|
|
|
|
#endif
|
|
|
|
resume_local_counters();
|
|
|
|
return handled;
|
|
|
|
}
|
|
|
|
|
2011-09-24 04:29:55 +04:00
|
|
|
static irqreturn_t mipsxx_pmu_handle_irq(int irq, void *dev)
|
2010-10-12 15:37:24 +04:00
|
|
|
{
|
|
|
|
return mipsxx_pmu_handle_shared_irq();
|
|
|
|
}
|
|
|
|
|
|
|
|
/* 24K */
|
|
|
|
#define IS_BOTH_COUNTERS_24K_EVENT(b) \
|
|
|
|
((b) == 0 || (b) == 1 || (b) == 11)
|
|
|
|
|
|
|
|
/* 34K */
|
|
|
|
#define IS_BOTH_COUNTERS_34K_EVENT(b) \
|
|
|
|
((b) == 0 || (b) == 1 || (b) == 11)
|
|
|
|
#ifdef CONFIG_MIPS_MT_SMP
|
|
|
|
#define IS_RANGE_P_34K_EVENT(r, b) \
|
|
|
|
((b) == 0 || (r) == 18 || (b) == 21 || (b) == 22 || \
|
|
|
|
(b) == 25 || (b) == 39 || (r) == 44 || (r) == 174 || \
|
|
|
|
(r) == 176 || ((b) >= 50 && (b) <= 55) || \
|
|
|
|
((b) >= 64 && (b) <= 67))
|
2013-01-22 15:59:30 +04:00
|
|
|
#define IS_RANGE_V_34K_EVENT(r) ((r) == 47)
|
2010-10-12 15:37:24 +04:00
|
|
|
#endif
|
|
|
|
|
|
|
|
/* 74K */
|
|
|
|
#define IS_BOTH_COUNTERS_74K_EVENT(b) \
|
|
|
|
((b) == 0 || (b) == 1)
|
|
|
|
|
2014-02-10 21:48:53 +04:00
|
|
|
/* proAptiv */
|
|
|
|
#define IS_BOTH_COUNTERS_PROAPTIV_EVENT(b) \
|
|
|
|
((b) == 0 || (b) == 1)
|
2014-07-04 14:08:57 +04:00
|
|
|
/* P5600 */
|
|
|
|
#define IS_BOTH_COUNTERS_P5600_EVENT(b) \
|
|
|
|
((b) == 0 || (b) == 1)
|
2014-02-10 21:48:53 +04:00
|
|
|
|
2010-10-12 15:37:24 +04:00
|
|
|
/* 1004K */
|
|
|
|
#define IS_BOTH_COUNTERS_1004K_EVENT(b) \
|
|
|
|
((b) == 0 || (b) == 1 || (b) == 11)
|
|
|
|
#ifdef CONFIG_MIPS_MT_SMP
|
|
|
|
#define IS_RANGE_P_1004K_EVENT(r, b) \
|
|
|
|
((b) == 0 || (r) == 18 || (b) == 21 || (b) == 22 || \
|
|
|
|
(b) == 25 || (b) == 36 || (b) == 39 || (r) == 44 || \
|
|
|
|
(r) == 174 || (r) == 176 || ((b) >= 50 && (b) <= 59) || \
|
|
|
|
(r) == 188 || (b) == 61 || (b) == 62 || \
|
|
|
|
((b) >= 64 && (b) <= 67))
|
|
|
|
#define IS_RANGE_V_1004K_EVENT(r) ((r) == 47)
|
|
|
|
#endif
|
|
|
|
|
2014-02-10 21:48:54 +04:00
|
|
|
/* interAptiv */
|
|
|
|
#define IS_BOTH_COUNTERS_INTERAPTIV_EVENT(b) \
|
|
|
|
((b) == 0 || (b) == 1 || (b) == 11)
|
|
|
|
#ifdef CONFIG_MIPS_MT_SMP
|
|
|
|
/* The P/V/T info is not provided for "(b) == 38" in SUM, assume P. */
|
|
|
|
#define IS_RANGE_P_INTERAPTIV_EVENT(r, b) \
|
|
|
|
((b) == 0 || (r) == 18 || (b) == 21 || (b) == 22 || \
|
|
|
|
(b) == 25 || (b) == 36 || (b) == 38 || (b) == 39 || \
|
|
|
|
(r) == 44 || (r) == 174 || (r) == 176 || ((b) >= 50 && \
|
|
|
|
(b) <= 59) || (r) == 188 || (b) == 61 || (b) == 62 || \
|
|
|
|
((b) >= 64 && (b) <= 67))
|
|
|
|
#define IS_RANGE_V_INTERAPTIV_EVENT(r) ((r) == 47 || (r) == 175)
|
|
|
|
#endif
|
|
|
|
|
2012-07-14 00:44:54 +04:00
|
|
|
/* BMIPS5000 */
|
|
|
|
#define IS_BOTH_COUNTERS_BMIPS5000_EVENT(b) \
|
|
|
|
((b) == 0 || (b) == 1)
|
|
|
|
|
|
|
|
|
2010-10-12 15:37:24 +04:00
|
|
|
/*
|
2014-07-04 14:08:56 +04:00
|
|
|
* For most cores the user can use 0-255 raw events, where 0-127 for the events
|
|
|
|
* of even counters, and 128-255 for odd counters. Note that bit 7 is used to
|
|
|
|
* indicate the even/odd bank selector. So, for example, when user wants to take
|
|
|
|
* the Event Num of 15 for odd counters (by referring to the user manual), then
|
|
|
|
* 128 needs to be added to 15 as the input for the event config, i.e., 143 (0x8F)
|
|
|
|
* to be used.
|
|
|
|
*
|
|
|
|
* Some newer cores have even more events, in which case the user can use raw
|
|
|
|
* events 0-511, where 0-255 are for the events of even counters, and 256-511
|
|
|
|
* are for odd counters, so bit 8 is used to indicate the even/odd bank selector.
|
2010-10-12 15:37:24 +04:00
|
|
|
*/
|
2011-09-24 04:29:55 +04:00
|
|
|
static const struct mips_perf_event *mipsxx_pmu_map_raw_event(u64 config)
|
2010-10-12 15:37:24 +04:00
|
|
|
{
|
2014-07-04 14:08:56 +04:00
|
|
|
/* currently most cores have 7-bit event numbers */
|
2010-10-12 15:37:24 +04:00
|
|
|
unsigned int raw_id = config & 0xff;
|
|
|
|
unsigned int base_id = raw_id & 0x7f;
|
|
|
|
|
|
|
|
switch (current_cpu_type()) {
|
|
|
|
case CPU_24K:
|
|
|
|
if (IS_BOTH_COUNTERS_24K_EVENT(base_id))
|
|
|
|
raw_event.cntr_mask = CNTR_EVEN | CNTR_ODD;
|
|
|
|
else
|
|
|
|
raw_event.cntr_mask =
|
|
|
|
raw_id > 127 ? CNTR_ODD : CNTR_EVEN;
|
|
|
|
#ifdef CONFIG_MIPS_MT_SMP
|
|
|
|
/*
|
|
|
|
* This is actually doing nothing. Non-multithreading
|
|
|
|
* CPUs will not check and calculate the range.
|
|
|
|
*/
|
|
|
|
raw_event.range = P;
|
|
|
|
#endif
|
|
|
|
break;
|
|
|
|
case CPU_34K:
|
|
|
|
if (IS_BOTH_COUNTERS_34K_EVENT(base_id))
|
|
|
|
raw_event.cntr_mask = CNTR_EVEN | CNTR_ODD;
|
|
|
|
else
|
|
|
|
raw_event.cntr_mask =
|
|
|
|
raw_id > 127 ? CNTR_ODD : CNTR_EVEN;
|
|
|
|
#ifdef CONFIG_MIPS_MT_SMP
|
|
|
|
if (IS_RANGE_P_34K_EVENT(raw_id, base_id))
|
|
|
|
raw_event.range = P;
|
|
|
|
else if (unlikely(IS_RANGE_V_34K_EVENT(raw_id)))
|
|
|
|
raw_event.range = V;
|
|
|
|
else
|
|
|
|
raw_event.range = T;
|
|
|
|
#endif
|
|
|
|
break;
|
|
|
|
case CPU_74K:
|
2014-01-18 01:03:50 +04:00
|
|
|
case CPU_1074K:
|
2010-10-12 15:37:24 +04:00
|
|
|
if (IS_BOTH_COUNTERS_74K_EVENT(base_id))
|
|
|
|
raw_event.cntr_mask = CNTR_EVEN | CNTR_ODD;
|
|
|
|
else
|
|
|
|
raw_event.cntr_mask =
|
|
|
|
raw_id > 127 ? CNTR_ODD : CNTR_EVEN;
|
|
|
|
#ifdef CONFIG_MIPS_MT_SMP
|
|
|
|
raw_event.range = P;
|
2014-02-10 21:48:53 +04:00
|
|
|
#endif
|
|
|
|
break;
|
|
|
|
case CPU_PROAPTIV:
|
|
|
|
if (IS_BOTH_COUNTERS_PROAPTIV_EVENT(base_id))
|
|
|
|
raw_event.cntr_mask = CNTR_EVEN | CNTR_ODD;
|
|
|
|
else
|
|
|
|
raw_event.cntr_mask =
|
|
|
|
raw_id > 127 ? CNTR_ODD : CNTR_EVEN;
|
|
|
|
#ifdef CONFIG_MIPS_MT_SMP
|
|
|
|
raw_event.range = P;
|
2014-07-04 14:08:57 +04:00
|
|
|
#endif
|
|
|
|
break;
|
|
|
|
case CPU_P5600:
|
2015-07-09 12:40:36 +03:00
|
|
|
case CPU_I6400:
|
2014-07-04 14:08:57 +04:00
|
|
|
/* 8-bit event numbers */
|
|
|
|
raw_id = config & 0x1ff;
|
|
|
|
base_id = raw_id & 0xff;
|
|
|
|
if (IS_BOTH_COUNTERS_P5600_EVENT(base_id))
|
|
|
|
raw_event.cntr_mask = CNTR_EVEN | CNTR_ODD;
|
|
|
|
else
|
|
|
|
raw_event.cntr_mask =
|
|
|
|
raw_id > 255 ? CNTR_ODD : CNTR_EVEN;
|
|
|
|
#ifdef CONFIG_MIPS_MT_SMP
|
|
|
|
raw_event.range = P;
|
2010-10-12 15:37:24 +04:00
|
|
|
#endif
|
|
|
|
break;
|
|
|
|
case CPU_1004K:
|
|
|
|
if (IS_BOTH_COUNTERS_1004K_EVENT(base_id))
|
|
|
|
raw_event.cntr_mask = CNTR_EVEN | CNTR_ODD;
|
|
|
|
else
|
|
|
|
raw_event.cntr_mask =
|
|
|
|
raw_id > 127 ? CNTR_ODD : CNTR_EVEN;
|
|
|
|
#ifdef CONFIG_MIPS_MT_SMP
|
|
|
|
if (IS_RANGE_P_1004K_EVENT(raw_id, base_id))
|
|
|
|
raw_event.range = P;
|
|
|
|
else if (unlikely(IS_RANGE_V_1004K_EVENT(raw_id)))
|
|
|
|
raw_event.range = V;
|
|
|
|
else
|
|
|
|
raw_event.range = T;
|
2014-02-10 21:48:54 +04:00
|
|
|
#endif
|
|
|
|
break;
|
|
|
|
case CPU_INTERAPTIV:
|
|
|
|
if (IS_BOTH_COUNTERS_INTERAPTIV_EVENT(base_id))
|
|
|
|
raw_event.cntr_mask = CNTR_EVEN | CNTR_ODD;
|
|
|
|
else
|
|
|
|
raw_event.cntr_mask =
|
|
|
|
raw_id > 127 ? CNTR_ODD : CNTR_EVEN;
|
|
|
|
#ifdef CONFIG_MIPS_MT_SMP
|
|
|
|
if (IS_RANGE_P_INTERAPTIV_EVENT(raw_id, base_id))
|
|
|
|
raw_event.range = P;
|
|
|
|
else if (unlikely(IS_RANGE_V_INTERAPTIV_EVENT(raw_id)))
|
|
|
|
raw_event.range = V;
|
|
|
|
else
|
|
|
|
raw_event.range = T;
|
2010-10-12 15:37:24 +04:00
|
|
|
#endif
|
|
|
|
break;
|
2012-07-14 00:44:54 +04:00
|
|
|
case CPU_BMIPS5000:
|
|
|
|
if (IS_BOTH_COUNTERS_BMIPS5000_EVENT(base_id))
|
|
|
|
raw_event.cntr_mask = CNTR_EVEN | CNTR_ODD;
|
|
|
|
else
|
|
|
|
raw_event.cntr_mask =
|
|
|
|
raw_id > 127 ? CNTR_ODD : CNTR_EVEN;
|
2015-03-29 05:54:08 +03:00
|
|
|
break;
|
|
|
|
case CPU_LOONGSON3:
|
|
|
|
raw_event.cntr_mask = raw_id > 127 ? CNTR_ODD : CNTR_EVEN;
|
|
|
|
break;
|
2010-10-12 15:37:24 +04:00
|
|
|
}
|
|
|
|
|
2014-07-04 14:08:56 +04:00
|
|
|
raw_event.event_id = base_id;
|
|
|
|
|
2010-10-12 15:37:24 +04:00
|
|
|
return &raw_event;
|
|
|
|
}
|
|
|
|
|
2011-09-24 04:29:55 +04:00
|
|
|
static const struct mips_perf_event *octeon_pmu_map_raw_event(u64 config)
|
|
|
|
{
|
|
|
|
unsigned int raw_id = config & 0xff;
|
|
|
|
unsigned int base_id = raw_id & 0x7f;
|
|
|
|
|
|
|
|
|
|
|
|
raw_event.cntr_mask = CNTR_ALL;
|
|
|
|
raw_event.event_id = base_id;
|
|
|
|
|
|
|
|
if (current_cpu_type() == CPU_CAVIUM_OCTEON2) {
|
|
|
|
if (base_id > 0x42)
|
|
|
|
return ERR_PTR(-EOPNOTSUPP);
|
|
|
|
} else {
|
|
|
|
if (base_id > 0x3a)
|
|
|
|
return ERR_PTR(-EOPNOTSUPP);
|
|
|
|
}
|
|
|
|
|
|
|
|
switch (base_id) {
|
|
|
|
case 0x00:
|
|
|
|
case 0x0f:
|
|
|
|
case 0x1e:
|
|
|
|
case 0x1f:
|
|
|
|
case 0x2f:
|
|
|
|
case 0x34:
|
|
|
|
case 0x3b ... 0x3f:
|
|
|
|
return ERR_PTR(-EOPNOTSUPP);
|
|
|
|
default:
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
|
|
|
|
return &raw_event;
|
|
|
|
}
|
|
|
|
|
2012-10-31 16:01:28 +04:00
|
|
|
static const struct mips_perf_event *xlp_pmu_map_raw_event(u64 config)
|
|
|
|
{
|
|
|
|
unsigned int raw_id = config & 0xff;
|
|
|
|
|
|
|
|
/* Only 1-63 are defined */
|
|
|
|
if ((raw_id < 0x01) || (raw_id > 0x3f))
|
|
|
|
return ERR_PTR(-EOPNOTSUPP);
|
|
|
|
|
|
|
|
raw_event.cntr_mask = CNTR_ALL;
|
|
|
|
raw_event.event_id = raw_id;
|
|
|
|
|
|
|
|
return &raw_event;
|
|
|
|
}
|
|
|
|
|
2010-10-12 15:37:24 +04:00
|
|
|
static int __init
|
|
|
|
init_hw_perf_events(void)
|
|
|
|
{
|
|
|
|
int counters, irq;
|
2011-09-24 04:29:55 +04:00
|
|
|
int counter_bits;
|
2010-10-12 15:37:24 +04:00
|
|
|
|
|
|
|
pr_info("Performance counters: ");
|
|
|
|
|
|
|
|
counters = n_counters();
|
|
|
|
if (counters == 0) {
|
|
|
|
pr_cont("No available PMU.\n");
|
|
|
|
return -ENODEV;
|
|
|
|
}
|
|
|
|
|
2012-07-14 00:44:53 +04:00
|
|
|
#ifdef CONFIG_MIPS_PERF_SHARED_TC_COUNTERS
|
2010-10-12 15:37:24 +04:00
|
|
|
cpu_has_mipsmt_pertccounters = read_c0_config7() & (1<<19);
|
|
|
|
if (!cpu_has_mipsmt_pertccounters)
|
|
|
|
counters = counters_total_to_per_cpu(counters);
|
|
|
|
#endif
|
|
|
|
|
2014-09-19 01:47:12 +04:00
|
|
|
if (get_c0_perfcount_int)
|
|
|
|
irq = get_c0_perfcount_int();
|
2015-01-28 00:45:49 +03:00
|
|
|
else if (cp0_perfcount_irq >= 0)
|
2014-09-19 01:47:12 +04:00
|
|
|
irq = MIPS_CPU_IRQ_BASE + cp0_perfcount_irq;
|
|
|
|
else
|
|
|
|
irq = -1;
|
2010-10-12 15:37:24 +04:00
|
|
|
|
2011-09-24 04:29:55 +04:00
|
|
|
mipspmu.map_raw_event = mipsxx_pmu_map_raw_event;
|
2010-10-12 15:37:24 +04:00
|
|
|
|
|
|
|
switch (current_cpu_type()) {
|
|
|
|
case CPU_24K:
|
2011-09-24 04:29:55 +04:00
|
|
|
mipspmu.name = "mips/24K";
|
|
|
|
mipspmu.general_event_map = &mipsxxcore_event_map;
|
|
|
|
mipspmu.cache_event_map = &mipsxxcore_cache_map;
|
2010-10-12 15:37:24 +04:00
|
|
|
break;
|
|
|
|
case CPU_34K:
|
2011-09-24 04:29:55 +04:00
|
|
|
mipspmu.name = "mips/34K";
|
|
|
|
mipspmu.general_event_map = &mipsxxcore_event_map;
|
|
|
|
mipspmu.cache_event_map = &mipsxxcore_cache_map;
|
2010-10-12 15:37:24 +04:00
|
|
|
break;
|
|
|
|
case CPU_74K:
|
2011-09-24 04:29:55 +04:00
|
|
|
mipspmu.name = "mips/74K";
|
2014-02-10 21:48:52 +04:00
|
|
|
mipspmu.general_event_map = &mipsxxcore_event_map2;
|
|
|
|
mipspmu.cache_event_map = &mipsxxcore_cache_map2;
|
2010-10-12 15:37:24 +04:00
|
|
|
break;
|
2014-02-10 21:48:53 +04:00
|
|
|
case CPU_PROAPTIV:
|
|
|
|
mipspmu.name = "mips/proAptiv";
|
|
|
|
mipspmu.general_event_map = &mipsxxcore_event_map2;
|
|
|
|
mipspmu.cache_event_map = &mipsxxcore_cache_map2;
|
|
|
|
break;
|
2014-07-04 14:08:57 +04:00
|
|
|
case CPU_P5600:
|
|
|
|
mipspmu.name = "mips/P5600";
|
|
|
|
mipspmu.general_event_map = &mipsxxcore_event_map2;
|
|
|
|
mipspmu.cache_event_map = &mipsxxcore_cache_map2;
|
|
|
|
break;
|
2015-07-09 12:40:36 +03:00
|
|
|
case CPU_I6400:
|
|
|
|
mipspmu.name = "mips/I6400";
|
|
|
|
mipspmu.general_event_map = &mipsxxcore_event_map2;
|
|
|
|
mipspmu.cache_event_map = &mipsxxcore_cache_map2;
|
|
|
|
break;
|
2010-10-12 15:37:24 +04:00
|
|
|
case CPU_1004K:
|
2011-09-24 04:29:55 +04:00
|
|
|
mipspmu.name = "mips/1004K";
|
|
|
|
mipspmu.general_event_map = &mipsxxcore_event_map;
|
|
|
|
mipspmu.cache_event_map = &mipsxxcore_cache_map;
|
2010-10-12 15:37:24 +04:00
|
|
|
break;
|
2014-01-18 01:03:50 +04:00
|
|
|
case CPU_1074K:
|
|
|
|
mipspmu.name = "mips/1074K";
|
|
|
|
mipspmu.general_event_map = &mipsxxcore_event_map;
|
|
|
|
mipspmu.cache_event_map = &mipsxxcore_cache_map;
|
|
|
|
break;
|
2014-02-10 21:48:54 +04:00
|
|
|
case CPU_INTERAPTIV:
|
|
|
|
mipspmu.name = "mips/interAptiv";
|
|
|
|
mipspmu.general_event_map = &mipsxxcore_event_map;
|
|
|
|
mipspmu.cache_event_map = &mipsxxcore_cache_map;
|
|
|
|
break;
|
2012-06-20 23:05:32 +04:00
|
|
|
case CPU_LOONGSON1:
|
|
|
|
mipspmu.name = "mips/loongson1";
|
|
|
|
mipspmu.general_event_map = &mipsxxcore_event_map;
|
|
|
|
mipspmu.cache_event_map = &mipsxxcore_cache_map;
|
|
|
|
break;
|
2015-03-29 05:54:08 +03:00
|
|
|
case CPU_LOONGSON3:
|
|
|
|
mipspmu.name = "mips/loongson3";
|
|
|
|
mipspmu.general_event_map = &loongson3_event_map;
|
|
|
|
mipspmu.cache_event_map = &loongson3_cache_map;
|
|
|
|
break;
|
2011-09-24 04:29:55 +04:00
|
|
|
case CPU_CAVIUM_OCTEON:
|
|
|
|
case CPU_CAVIUM_OCTEON_PLUS:
|
|
|
|
case CPU_CAVIUM_OCTEON2:
|
|
|
|
mipspmu.name = "octeon";
|
|
|
|
mipspmu.general_event_map = &octeon_event_map;
|
|
|
|
mipspmu.cache_event_map = &octeon_cache_map;
|
|
|
|
mipspmu.map_raw_event = octeon_pmu_map_raw_event;
|
|
|
|
break;
|
2012-07-14 00:44:54 +04:00
|
|
|
case CPU_BMIPS5000:
|
|
|
|
mipspmu.name = "BMIPS5000";
|
|
|
|
mipspmu.general_event_map = &bmips5000_event_map;
|
|
|
|
mipspmu.cache_event_map = &bmips5000_cache_map;
|
|
|
|
break;
|
2012-10-31 16:01:28 +04:00
|
|
|
case CPU_XLP:
|
|
|
|
mipspmu.name = "xlp";
|
|
|
|
mipspmu.general_event_map = &xlp_event_map;
|
|
|
|
mipspmu.cache_event_map = &xlp_cache_map;
|
|
|
|
mipspmu.map_raw_event = xlp_pmu_map_raw_event;
|
|
|
|
break;
|
2010-10-12 15:37:24 +04:00
|
|
|
default:
|
|
|
|
pr_cont("Either hardware does not support performance "
|
|
|
|
"counters, or not yet implemented.\n");
|
|
|
|
return -ENODEV;
|
|
|
|
}
|
|
|
|
|
2011-09-24 04:29:55 +04:00
|
|
|
mipspmu.num_counters = counters;
|
|
|
|
mipspmu.irq = irq;
|
|
|
|
|
|
|
|
if (read_c0_perfctrl0() & M_PERFCTL_WIDE) {
|
|
|
|
mipspmu.max_period = (1ULL << 63) - 1;
|
|
|
|
mipspmu.valid_count = (1ULL << 63) - 1;
|
|
|
|
mipspmu.overflow = 1ULL << 63;
|
|
|
|
mipspmu.read_counter = mipsxx_pmu_read_counter_64;
|
|
|
|
mipspmu.write_counter = mipsxx_pmu_write_counter_64;
|
|
|
|
counter_bits = 64;
|
|
|
|
} else {
|
|
|
|
mipspmu.max_period = (1ULL << 31) - 1;
|
|
|
|
mipspmu.valid_count = (1ULL << 31) - 1;
|
|
|
|
mipspmu.overflow = 1ULL << 31;
|
|
|
|
mipspmu.read_counter = mipsxx_pmu_read_counter;
|
|
|
|
mipspmu.write_counter = mipsxx_pmu_write_counter;
|
|
|
|
counter_bits = 32;
|
|
|
|
}
|
|
|
|
|
|
|
|
on_each_cpu(reset_counters, (void *)(long)counters, 1);
|
|
|
|
|
|
|
|
pr_cont("%s PMU enabled, %d %d-bit counters available to each "
|
|
|
|
"CPU, irq %d%s\n", mipspmu.name, counters, counter_bits, irq,
|
|
|
|
irq < 0 ? " (share with timer interrupt)" : "");
|
2010-10-12 15:37:24 +04:00
|
|
|
|
MIPS, Perf-events: Work with the new PMU interface
This is the MIPS part of the following commits by Peter Zijlstra:
- a4eaf7f14675cb512d69f0c928055e73d0c6d252
perf: Rework the PMU methods
Replace pmu::{enable,disable,start,stop,unthrottle} with
pmu::{add,del,start,stop}, all of which take a flags argument.
The new interface extends the capability to stop a counter while
keeping it scheduled on the PMU. We replace the throttled state with
the generic stopped state.
This also allows us to efficiently stop/start counters over certain
code paths (like IRQ handlers).
It also allows scheduling a counter without it starting, allowing for
a generic frozen state (useful for rotating stopped counters).
The stopped state is implemented in two different ways, depending on
how the architecture implemented the throttled state:
1) We disable the counter:
a) the pmu has per-counter enable bits, we flip that
b) we program a NOP event, preserving the counter state
2) We store the counter state and ignore all read/overflow events
For MIPSXX, the stopped state is implemented in the way of 1.b as above.
- 33696fc0d141bbbcb12f75b69608ea83282e3117
perf: Per PMU disable
Changes perf_disable() into perf_pmu_disable().
- 24cd7f54a0d47e1d5b3de29e2456bfbd2d8447b7
perf: Reduce perf_disable() usage
Since the current perf_disable() usage is only an optimization,
remove it for now. This eases the removal of the __weak
hw_perf_enable() interface.
- b0a873ebbf87bf38bf70b5e39a7cadc96099fa13
perf: Register PMU implementations
Simple registration interface for struct pmu, this provides the
infrastructure for removing all the weak functions.
- 51b0fe39549a04858001922919ab355dee9bdfcf
perf: Deconstify struct pmu
sed -ie 's/const struct pmu\>/struct pmu/g' `git grep -l "const struct pmu\>"`
Reported-by: Wu Zhangjin <wuzhangjin@gmail.com>
Acked-by: David Daney <ddaney@caviumnetworks.com>
Signed-off-by: Deng-Cheng Zhu <dengcheng.zhu@gmail.com>
To: a.p.zijlstra@chello.nl
To: fweisbec@gmail.com
To: will.deacon@arm.com
Cc: linux-mips@linux-mips.org
Cc: linux-kernel@vger.kernel.org
Cc: wuzhangjin@gmail.com
Cc: paulus@samba.org
Cc: mingo@elte.hu
Cc: acme@redhat.com
Cc: dengcheng.zhu@gmail.com
Cc: matt@console-pimps.org
Cc: sshtylyov@mvista.com
Cc: ddaney@caviumnetworks.com
Patchwork: http://patchwork.linux-mips.org/patch/2012/
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
2011-01-21 11:19:18 +03:00
|
|
|
perf_pmu_register(&pmu, "cpu", PERF_TYPE_RAW);
|
|
|
|
|
2010-10-12 15:37:24 +04:00
|
|
|
return 0;
|
|
|
|
}
|
2010-11-25 20:38:29 +03:00
|
|
|
early_initcall(init_hw_perf_events);
|