summaryrefslogtreecommitdiff
path: root/arch/x86/events/intel
diff options
context:
space:
mode:
Diffstat (limited to 'arch/x86/events/intel')
-rw-r--r--arch/x86/events/intel/bts.c151
-rw-r--r--arch/x86/events/intel/core.c414
-rw-r--r--arch/x86/events/intel/cstate.c3
-rw-r--r--arch/x86/events/intel/ds.c68
-rw-r--r--arch/x86/events/intel/knc.c24
-rw-r--r--arch/x86/events/intel/lbr.c46
-rw-r--r--arch/x86/events/intel/p4.c28
-rw-r--r--arch/x86/events/intel/p6.c13
-rw-r--r--arch/x86/events/intel/pt.c37
-rw-r--r--arch/x86/events/intel/uncore.c15
-rw-r--r--arch/x86/events/intel/uncore_discovery.c11
-rw-r--r--arch/x86/events/intel/uncore_nhmex.c71
-rw-r--r--arch/x86/events/intel/uncore_snb.c43
-rw-r--r--arch/x86/events/intel/uncore_snbep.c51
14 files changed, 641 insertions, 334 deletions
diff --git a/arch/x86/events/intel/bts.c b/arch/x86/events/intel/bts.c
index a95e6c91c4d7..61da6b8a3d51 100644
--- a/arch/x86/events/intel/bts.c
+++ b/arch/x86/events/intel/bts.c
@@ -17,6 +17,7 @@
#include <linux/sizes.h>
#include <asm/perf_event.h>
+#include <asm/msr.h>
#include "../perf_event.h"
@@ -80,54 +81,54 @@ static void *
bts_buffer_setup_aux(struct perf_event *event, void **pages,
int nr_pages, bool overwrite)
{
- struct bts_buffer *buf;
+ struct bts_buffer *bb;
struct page *page;
int cpu = event->cpu;
int node = (cpu == -1) ? cpu : cpu_to_node(cpu);
unsigned long offset;
size_t size = nr_pages << PAGE_SHIFT;
- int pg, nbuf, pad;
+ int pg, nr_buf, pad;
/* count all the high order buffers */
- for (pg = 0, nbuf = 0; pg < nr_pages;) {
+ for (pg = 0, nr_buf = 0; pg < nr_pages;) {
page = virt_to_page(pages[pg]);
pg += buf_nr_pages(page);
- nbuf++;
+ nr_buf++;
}
/*
* to avoid interrupts in overwrite mode, only allow one physical
*/
- if (overwrite && nbuf > 1)
+ if (overwrite && nr_buf > 1)
return NULL;
- buf = kzalloc_node(offsetof(struct bts_buffer, buf[nbuf]), GFP_KERNEL, node);
- if (!buf)
+ bb = kzalloc_node(struct_size(bb, buf, nr_buf), GFP_KERNEL, node);
+ if (!bb)
return NULL;
- buf->nr_pages = nr_pages;
- buf->nr_bufs = nbuf;
- buf->snapshot = overwrite;
- buf->data_pages = pages;
- buf->real_size = size - size % BTS_RECORD_SIZE;
+ bb->nr_pages = nr_pages;
+ bb->nr_bufs = nr_buf;
+ bb->snapshot = overwrite;
+ bb->data_pages = pages;
+ bb->real_size = size - size % BTS_RECORD_SIZE;
- for (pg = 0, nbuf = 0, offset = 0, pad = 0; nbuf < buf->nr_bufs; nbuf++) {
+ for (pg = 0, nr_buf = 0, offset = 0, pad = 0; nr_buf < bb->nr_bufs; nr_buf++) {
unsigned int __nr_pages;
page = virt_to_page(pages[pg]);
__nr_pages = buf_nr_pages(page);
- buf->buf[nbuf].page = page;
- buf->buf[nbuf].offset = offset;
- buf->buf[nbuf].displacement = (pad ? BTS_RECORD_SIZE - pad : 0);
- buf->buf[nbuf].size = buf_size(page) - buf->buf[nbuf].displacement;
- pad = buf->buf[nbuf].size % BTS_RECORD_SIZE;
- buf->buf[nbuf].size -= pad;
+ bb->buf[nr_buf].page = page;
+ bb->buf[nr_buf].offset = offset;
+ bb->buf[nr_buf].displacement = (pad ? BTS_RECORD_SIZE - pad : 0);
+ bb->buf[nr_buf].size = buf_size(page) - bb->buf[nr_buf].displacement;
+ pad = bb->buf[nr_buf].size % BTS_RECORD_SIZE;
+ bb->buf[nr_buf].size -= pad;
pg += __nr_pages;
offset += __nr_pages << PAGE_SHIFT;
}
- return buf;
+ return bb;
}
static void bts_buffer_free_aux(void *data)
@@ -135,25 +136,25 @@ static void bts_buffer_free_aux(void *data)
kfree(data);
}
-static unsigned long bts_buffer_offset(struct bts_buffer *buf, unsigned int idx)
+static unsigned long bts_buffer_offset(struct bts_buffer *bb, unsigned int idx)
{
- return buf->buf[idx].offset + buf->buf[idx].displacement;
+ return bb->buf[idx].offset + bb->buf[idx].displacement;
}
static void
-bts_config_buffer(struct bts_buffer *buf)
+bts_config_buffer(struct bts_buffer *bb)
{
int cpu = raw_smp_processor_id();
struct debug_store *ds = per_cpu(cpu_hw_events, cpu).ds;
- struct bts_phys *phys = &buf->buf[buf->cur_buf];
+ struct bts_phys *phys = &bb->buf[bb->cur_buf];
unsigned long index, thresh = 0, end = phys->size;
struct page *page = phys->page;
- index = local_read(&buf->head);
+ index = local_read(&bb->head);
- if (!buf->snapshot) {
- if (buf->end < phys->offset + buf_size(page))
- end = buf->end - phys->offset - phys->displacement;
+ if (!bb->snapshot) {
+ if (bb->end < phys->offset + buf_size(page))
+ end = bb->end - phys->offset - phys->displacement;
index -= phys->offset + phys->displacement;
@@ -168,7 +169,7 @@ bts_config_buffer(struct bts_buffer *buf)
ds->bts_buffer_base = (u64)(long)page_address(page) + phys->displacement;
ds->bts_index = ds->bts_buffer_base + index;
ds->bts_absolute_maximum = ds->bts_buffer_base + end;
- ds->bts_interrupt_threshold = !buf->snapshot
+ ds->bts_interrupt_threshold = !bb->snapshot
? ds->bts_buffer_base + thresh
: ds->bts_absolute_maximum + BTS_RECORD_SIZE;
}
@@ -184,16 +185,16 @@ static void bts_update(struct bts_ctx *bts)
{
int cpu = raw_smp_processor_id();
struct debug_store *ds = per_cpu(cpu_hw_events, cpu).ds;
- struct bts_buffer *buf = perf_get_aux(&bts->handle);
+ struct bts_buffer *bb = perf_get_aux(&bts->handle);
unsigned long index = ds->bts_index - ds->bts_buffer_base, old, head;
- if (!buf)
+ if (!bb)
return;
- head = index + bts_buffer_offset(buf, buf->cur_buf);
- old = local_xchg(&buf->head, head);
+ head = index + bts_buffer_offset(bb, bb->cur_buf);
+ old = local_xchg(&bb->head, head);
- if (!buf->snapshot) {
+ if (!bb->snapshot) {
if (old == head)
return;
@@ -205,9 +206,9 @@ static void bts_update(struct bts_ctx *bts)
* old and head are always in the same physical buffer, so we
* can subtract them to get the data size.
*/
- local_add(head - old, &buf->data_size);
+ local_add(head - old, &bb->data_size);
} else {
- local_set(&buf->data_size, head);
+ local_set(&bb->data_size, head);
}
/*
@@ -218,7 +219,7 @@ static void bts_update(struct bts_ctx *bts)
}
static int
-bts_buffer_reset(struct bts_buffer *buf, struct perf_output_handle *handle);
+bts_buffer_reset(struct bts_buffer *bb, struct perf_output_handle *handle);
/*
* Ordering PMU callbacks wrt themselves and the PMI is done by means
@@ -232,17 +233,17 @@ bts_buffer_reset(struct bts_buffer *buf, struct perf_output_handle *handle);
static void __bts_event_start(struct perf_event *event)
{
struct bts_ctx *bts = this_cpu_ptr(bts_ctx);
- struct bts_buffer *buf = perf_get_aux(&bts->handle);
+ struct bts_buffer *bb = perf_get_aux(&bts->handle);
u64 config = 0;
- if (!buf->snapshot)
+ if (!bb->snapshot)
config |= ARCH_PERFMON_EVENTSEL_INT;
if (!event->attr.exclude_kernel)
config |= ARCH_PERFMON_EVENTSEL_OS;
if (!event->attr.exclude_user)
config |= ARCH_PERFMON_EVENTSEL_USR;
- bts_config_buffer(buf);
+ bts_config_buffer(bb);
/*
* local barrier to make sure that ds configuration made it
@@ -261,13 +262,13 @@ static void bts_event_start(struct perf_event *event, int flags)
{
struct cpu_hw_events *cpuc = this_cpu_ptr(&cpu_hw_events);
struct bts_ctx *bts = this_cpu_ptr(bts_ctx);
- struct bts_buffer *buf;
+ struct bts_buffer *bb;
- buf = perf_aux_output_begin(&bts->handle, event);
- if (!buf)
+ bb = perf_aux_output_begin(&bts->handle, event);
+ if (!bb)
goto fail_stop;
- if (bts_buffer_reset(buf, &bts->handle))
+ if (bts_buffer_reset(bb, &bts->handle))
goto fail_end_stop;
bts->ds_back.bts_buffer_base = cpuc->ds->bts_buffer_base;
@@ -306,27 +307,27 @@ static void bts_event_stop(struct perf_event *event, int flags)
{
struct cpu_hw_events *cpuc = this_cpu_ptr(&cpu_hw_events);
struct bts_ctx *bts = this_cpu_ptr(bts_ctx);
- struct bts_buffer *buf = NULL;
+ struct bts_buffer *bb = NULL;
int state = READ_ONCE(bts->state);
if (state == BTS_STATE_ACTIVE)
__bts_event_stop(event, BTS_STATE_STOPPED);
if (state != BTS_STATE_STOPPED)
- buf = perf_get_aux(&bts->handle);
+ bb = perf_get_aux(&bts->handle);
event->hw.state |= PERF_HES_STOPPED;
if (flags & PERF_EF_UPDATE) {
bts_update(bts);
- if (buf) {
- if (buf->snapshot)
+ if (bb) {
+ if (bb->snapshot)
bts->handle.head =
- local_xchg(&buf->data_size,
- buf->nr_pages << PAGE_SHIFT);
+ local_xchg(&bb->data_size,
+ bb->nr_pages << PAGE_SHIFT);
perf_aux_output_end(&bts->handle,
- local_xchg(&buf->data_size, 0));
+ local_xchg(&bb->data_size, 0));
}
cpuc->ds->bts_index = bts->ds_back.bts_buffer_base;
@@ -382,19 +383,19 @@ void intel_bts_disable_local(void)
}
static int
-bts_buffer_reset(struct bts_buffer *buf, struct perf_output_handle *handle)
+bts_buffer_reset(struct bts_buffer *bb, struct perf_output_handle *handle)
{
unsigned long head, space, next_space, pad, gap, skip, wakeup;
unsigned int next_buf;
struct bts_phys *phys, *next_phys;
int ret;
- if (buf->snapshot)
+ if (bb->snapshot)
return 0;
- head = handle->head & ((buf->nr_pages << PAGE_SHIFT) - 1);
+ head = handle->head & ((bb->nr_pages << PAGE_SHIFT) - 1);
- phys = &buf->buf[buf->cur_buf];
+ phys = &bb->buf[bb->cur_buf];
space = phys->offset + phys->displacement + phys->size - head;
pad = space;
if (space > handle->size) {
@@ -403,10 +404,10 @@ bts_buffer_reset(struct bts_buffer *buf, struct perf_output_handle *handle)
}
if (space <= BTS_SAFETY_MARGIN) {
/* See if next phys buffer has more space */
- next_buf = buf->cur_buf + 1;
- if (next_buf >= buf->nr_bufs)
+ next_buf = bb->cur_buf + 1;
+ if (next_buf >= bb->nr_bufs)
next_buf = 0;
- next_phys = &buf->buf[next_buf];
+ next_phys = &bb->buf[next_buf];
gap = buf_size(phys->page) - phys->displacement - phys->size +
next_phys->displacement;
skip = pad + gap;
@@ -431,8 +432,8 @@ bts_buffer_reset(struct bts_buffer *buf, struct perf_output_handle *handle)
* anymore, so we must not be racing with
* bts_update().
*/
- buf->cur_buf = next_buf;
- local_set(&buf->head, head);
+ bb->cur_buf = next_buf;
+ local_set(&bb->head, head);
}
}
}
@@ -445,7 +446,7 @@ bts_buffer_reset(struct bts_buffer *buf, struct perf_output_handle *handle)
space -= space % BTS_RECORD_SIZE;
}
- buf->end = head + space;
+ bb->end = head + space;
/*
* If we have no space, the lost notification would have been sent when
@@ -462,7 +463,7 @@ int intel_bts_interrupt(void)
struct debug_store *ds = this_cpu_ptr(&cpu_hw_events)->ds;
struct bts_ctx *bts;
struct perf_event *event;
- struct bts_buffer *buf;
+ struct bts_buffer *bb;
s64 old_head;
int err = -ENOSPC, handled = 0;
@@ -485,8 +486,8 @@ int intel_bts_interrupt(void)
if (READ_ONCE(bts->state) == BTS_STATE_STOPPED)
return handled;
- buf = perf_get_aux(&bts->handle);
- if (!buf)
+ bb = perf_get_aux(&bts->handle);
+ if (!bb)
return handled;
/*
@@ -494,26 +495,26 @@ int intel_bts_interrupt(void)
* there's no other way of telling, because the pointer will
* keep moving
*/
- if (buf->snapshot)
+ if (bb->snapshot)
return 0;
- old_head = local_read(&buf->head);
+ old_head = local_read(&bb->head);
bts_update(bts);
/* no new data */
- if (old_head == local_read(&buf->head))
+ if (old_head == local_read(&bb->head))
return handled;
- perf_aux_output_end(&bts->handle, local_xchg(&buf->data_size, 0));
+ perf_aux_output_end(&bts->handle, local_xchg(&bb->data_size, 0));
- buf = perf_aux_output_begin(&bts->handle, event);
- if (buf)
- err = bts_buffer_reset(buf, &bts->handle);
+ bb = perf_aux_output_begin(&bts->handle, event);
+ if (bb)
+ err = bts_buffer_reset(bb, &bts->handle);
if (err) {
WRITE_ONCE(bts->state, BTS_STATE_STOPPED);
- if (buf) {
+ if (bb) {
/*
* BTS_STATE_STOPPED should be visible before
* cleared handle::event
@@ -599,7 +600,11 @@ static void bts_event_read(struct perf_event *event)
static __init int bts_init(void)
{
- if (!boot_cpu_has(X86_FEATURE_DTES64) || !x86_pmu.bts)
+ if (!boot_cpu_has(X86_FEATURE_DTES64))
+ return -ENODEV;
+
+ x86_pmu.bts = boot_cpu_has(X86_FEATURE_BTS);
+ if (!x86_pmu.bts)
return -ENODEV;
if (boot_cpu_has(X86_FEATURE_PTI)) {
diff --git a/arch/x86/events/intel/core.c b/arch/x86/events/intel/core.c
index c5f385413392..466283326630 100644
--- a/arch/x86/events/intel/core.c
+++ b/arch/x86/events/intel/core.c
@@ -23,6 +23,7 @@
#include <asm/intel_pt.h>
#include <asm/apic.h>
#include <asm/cpu_device_id.h>
+#include <asm/msr.h>
#include "../perf_event.h"
@@ -2224,6 +2225,18 @@ static struct extra_reg intel_cmt_extra_regs[] __read_mostly = {
EVENT_EXTRA_END
};
+EVENT_ATTR_STR(topdown-fe-bound, td_fe_bound_skt, "event=0x9c,umask=0x01");
+EVENT_ATTR_STR(topdown-retiring, td_retiring_skt, "event=0xc2,umask=0x02");
+EVENT_ATTR_STR(topdown-be-bound, td_be_bound_skt, "event=0xa4,umask=0x02");
+
+static struct attribute *skt_events_attrs[] = {
+ EVENT_PTR(td_fe_bound_skt),
+ EVENT_PTR(td_retiring_skt),
+ EVENT_PTR(td_bad_spec_cmt),
+ EVENT_PTR(td_be_bound_skt),
+ NULL,
+};
+
#define KNL_OT_L2_HITE BIT_ULL(19) /* Other Tile L2 Hit */
#define KNL_OT_L2_HITF BIT_ULL(20) /* Other Tile L2 Hit */
#define KNL_MCDRAM_LOCAL BIT_ULL(21)
@@ -2285,7 +2298,7 @@ static __always_inline void __intel_pmu_disable_all(bool bts)
{
struct cpu_hw_events *cpuc = this_cpu_ptr(&cpu_hw_events);
- wrmsrl(MSR_CORE_PERF_GLOBAL_CTRL, 0);
+ wrmsrq(MSR_CORE_PERF_GLOBAL_CTRL, 0);
if (bts && test_bit(INTEL_PMC_IDX_FIXED_BTS, cpuc->active_mask))
intel_pmu_disable_bts();
@@ -2294,7 +2307,7 @@ static __always_inline void __intel_pmu_disable_all(bool bts)
static __always_inline void intel_pmu_disable_all(void)
{
__intel_pmu_disable_all(true);
- intel_pmu_pebs_disable_all();
+ static_call_cond(x86_pmu_pebs_disable_all)();
intel_pmu_lbr_disable_all();
}
@@ -2306,11 +2319,11 @@ static void __intel_pmu_enable_all(int added, bool pmi)
intel_pmu_lbr_enable_all(pmi);
if (cpuc->fixed_ctrl_val != cpuc->active_fixed_ctrl_val) {
- wrmsrl(MSR_ARCH_PERFMON_FIXED_CTR_CTRL, cpuc->fixed_ctrl_val);
+ wrmsrq(MSR_ARCH_PERFMON_FIXED_CTR_CTRL, cpuc->fixed_ctrl_val);
cpuc->active_fixed_ctrl_val = cpuc->fixed_ctrl_val;
}
- wrmsrl(MSR_CORE_PERF_GLOBAL_CTRL,
+ wrmsrq(MSR_CORE_PERF_GLOBAL_CTRL,
intel_ctrl & ~cpuc->intel_ctrl_guest_mask);
if (test_bit(INTEL_PMC_IDX_FIXED_BTS, cpuc->active_mask)) {
@@ -2326,7 +2339,7 @@ static void __intel_pmu_enable_all(int added, bool pmi)
static void intel_pmu_enable_all(int added)
{
- intel_pmu_pebs_enable_all();
+ static_call_cond(x86_pmu_pebs_enable_all)();
__intel_pmu_enable_all(added, false);
}
@@ -2426,12 +2439,12 @@ static void intel_pmu_nhm_workaround(void)
}
for (i = 0; i < 4; i++) {
- wrmsrl(MSR_ARCH_PERFMON_EVENTSEL0 + i, nhm_magic[i]);
- wrmsrl(MSR_ARCH_PERFMON_PERFCTR0 + i, 0x0);
+ wrmsrq(MSR_ARCH_PERFMON_EVENTSEL0 + i, nhm_magic[i]);
+ wrmsrq(MSR_ARCH_PERFMON_PERFCTR0 + i, 0x0);
}
- wrmsrl(MSR_CORE_PERF_GLOBAL_CTRL, 0xf);
- wrmsrl(MSR_CORE_PERF_GLOBAL_CTRL, 0x0);
+ wrmsrq(MSR_CORE_PERF_GLOBAL_CTRL, 0xf);
+ wrmsrq(MSR_CORE_PERF_GLOBAL_CTRL, 0x0);
for (i = 0; i < 4; i++) {
event = cpuc->events[i];
@@ -2441,7 +2454,7 @@ static void intel_pmu_nhm_workaround(void)
__x86_pmu_enable_event(&event->hw,
ARCH_PERFMON_EVENTSEL_ENABLE);
} else
- wrmsrl(MSR_ARCH_PERFMON_EVENTSEL0 + i, 0x0);
+ wrmsrq(MSR_ARCH_PERFMON_EVENTSEL0 + i, 0x0);
}
}
@@ -2458,7 +2471,7 @@ static void intel_set_tfa(struct cpu_hw_events *cpuc, bool on)
if (cpuc->tfa_shadow != val) {
cpuc->tfa_shadow = val;
- wrmsrl(MSR_TSX_FORCE_ABORT, val);
+ wrmsrq(MSR_TSX_FORCE_ABORT, val);
}
}
@@ -2489,14 +2502,14 @@ static inline u64 intel_pmu_get_status(void)
{
u64 status;
- rdmsrl(MSR_CORE_PERF_GLOBAL_STATUS, status);
+ rdmsrq(MSR_CORE_PERF_GLOBAL_STATUS, status);
return status;
}
static inline void intel_pmu_ack_status(u64 ack)
{
- wrmsrl(MSR_CORE_PERF_GLOBAL_OVF_CTRL, ack);
+ wrmsrq(MSR_CORE_PERF_GLOBAL_OVF_CTRL, ack);
}
static inline bool event_is_checkpointed(struct perf_event *event)
@@ -2583,7 +2596,7 @@ static void intel_pmu_disable_event(struct perf_event *event)
* so we don't trigger the event without PEBS bit set.
*/
if (unlikely(event->attr.precise_ip))
- intel_pmu_pebs_disable(event);
+ static_call(x86_pmu_pebs_disable)(event);
}
static void intel_pmu_assign_event(struct perf_event *event, int idx)
@@ -2603,6 +2616,9 @@ static void intel_pmu_del_event(struct perf_event *event)
intel_pmu_lbr_del(event);
if (event->attr.precise_ip)
intel_pmu_pebs_del(event);
+ if (is_pebs_counter_event_group(event) ||
+ is_acr_event_group(event))
+ this_cpu_ptr(&cpu_hw_events)->n_late_setup--;
}
static int icl_set_topdown_event_period(struct perf_event *event)
@@ -2619,15 +2635,15 @@ static int icl_set_topdown_event_period(struct perf_event *event)
* Don't need to clear them again.
*/
if (left == x86_pmu.max_period) {
- wrmsrl(MSR_CORE_PERF_FIXED_CTR3, 0);
- wrmsrl(MSR_PERF_METRICS, 0);
+ wrmsrq(MSR_CORE_PERF_FIXED_CTR3, 0);
+ wrmsrq(MSR_PERF_METRICS, 0);
hwc->saved_slots = 0;
hwc->saved_metric = 0;
}
if ((hwc->saved_slots) && is_slots_event(event)) {
- wrmsrl(MSR_CORE_PERF_FIXED_CTR3, hwc->saved_slots);
- wrmsrl(MSR_PERF_METRICS, hwc->saved_metric);
+ wrmsrq(MSR_CORE_PERF_FIXED_CTR3, hwc->saved_slots);
+ wrmsrq(MSR_PERF_METRICS, hwc->saved_metric);
}
perf_event_update_userpage(event);
@@ -2724,12 +2740,12 @@ static u64 intel_update_topdown_event(struct perf_event *event, int metric_end,
if (!val) {
/* read Fixed counter 3 */
- rdpmcl((3 | INTEL_PMC_FIXED_RDPMC_BASE), slots);
+ slots = rdpmc(3 | INTEL_PMC_FIXED_RDPMC_BASE);
if (!slots)
return 0;
/* read PERF_METRICS */
- rdpmcl(INTEL_PMC_FIXED_RDPMC_METRICS, metrics);
+ metrics = rdpmc(INTEL_PMC_FIXED_RDPMC_METRICS);
} else {
slots = val[0];
metrics = val[1];
@@ -2773,8 +2789,8 @@ static u64 intel_update_topdown_event(struct perf_event *event, int metric_end,
if (reset) {
/* The fixed counter 3 has to be written before the PERF_METRICS. */
- wrmsrl(MSR_CORE_PERF_FIXED_CTR3, 0);
- wrmsrl(MSR_PERF_METRICS, 0);
+ wrmsrq(MSR_CORE_PERF_FIXED_CTR3, 0);
+ wrmsrq(MSR_PERF_METRICS, 0);
if (event)
update_saved_topdown_regs(event, 0, 0, metric_end);
}
@@ -2880,6 +2896,52 @@ static void intel_pmu_enable_fixed(struct perf_event *event)
cpuc->fixed_ctrl_val |= bits;
}
+static void intel_pmu_config_acr(int idx, u64 mask, u32 reload)
+{
+ struct cpu_hw_events *cpuc = this_cpu_ptr(&cpu_hw_events);
+ int msr_b, msr_c;
+
+ if (!mask && !cpuc->acr_cfg_b[idx])
+ return;
+
+ if (idx < INTEL_PMC_IDX_FIXED) {
+ msr_b = MSR_IA32_PMC_V6_GP0_CFG_B;
+ msr_c = MSR_IA32_PMC_V6_GP0_CFG_C;
+ } else {
+ msr_b = MSR_IA32_PMC_V6_FX0_CFG_B;
+ msr_c = MSR_IA32_PMC_V6_FX0_CFG_C;
+ idx -= INTEL_PMC_IDX_FIXED;
+ }
+
+ if (cpuc->acr_cfg_b[idx] != mask) {
+ wrmsrl(msr_b + x86_pmu.addr_offset(idx, false), mask);
+ cpuc->acr_cfg_b[idx] = mask;
+ }
+ /* Only need to update the reload value when there is a valid config value. */
+ if (mask && cpuc->acr_cfg_c[idx] != reload) {
+ wrmsrl(msr_c + x86_pmu.addr_offset(idx, false), reload);
+ cpuc->acr_cfg_c[idx] = reload;
+ }
+}
+
+static void intel_pmu_enable_acr(struct perf_event *event)
+{
+ struct hw_perf_event *hwc = &event->hw;
+
+ if (!is_acr_event_group(event) || !event->attr.config2) {
+ /*
+ * The disable doesn't clear the ACR CFG register.
+ * Check and clear the ACR CFG register.
+ */
+ intel_pmu_config_acr(hwc->idx, 0, 0);
+ return;
+ }
+
+ intel_pmu_config_acr(hwc->idx, hwc->config1, -hwc->sample_period);
+}
+
+DEFINE_STATIC_CALL_NULL(intel_pmu_enable_acr_event, intel_pmu_enable_acr);
+
static void intel_pmu_enable_event(struct perf_event *event)
{
u64 enable_mask = ARCH_PERFMON_EVENTSEL_ENABLE;
@@ -2887,16 +2949,19 @@ static void intel_pmu_enable_event(struct perf_event *event)
int idx = hwc->idx;
if (unlikely(event->attr.precise_ip))
- intel_pmu_pebs_enable(event);
+ static_call(x86_pmu_pebs_enable)(event);
switch (idx) {
case 0 ... INTEL_PMC_IDX_FIXED - 1:
if (branch_sample_counters(event))
enable_mask |= ARCH_PERFMON_EVENTSEL_BR_CNTR;
intel_set_masks(event, idx);
+ static_call_cond(intel_pmu_enable_acr_event)(event);
__x86_pmu_enable_event(hwc, enable_mask);
break;
case INTEL_PMC_IDX_FIXED ... INTEL_PMC_IDX_FIXED_BTS - 1:
+ static_call_cond(intel_pmu_enable_acr_event)(event);
+ fallthrough;
case INTEL_PMC_IDX_METRIC_BASE ... INTEL_PMC_IDX_METRIC_END:
intel_pmu_enable_fixed(event);
break;
@@ -2914,12 +2979,51 @@ static void intel_pmu_enable_event(struct perf_event *event)
}
}
+static void intel_pmu_acr_late_setup(struct cpu_hw_events *cpuc)
+{
+ struct perf_event *event, *leader;
+ int i, j, idx;
+
+ for (i = 0; i < cpuc->n_events; i++) {
+ leader = cpuc->event_list[i];
+ if (!is_acr_event_group(leader))
+ continue;
+
+ /* The ACR events must be contiguous. */
+ for (j = i; j < cpuc->n_events; j++) {
+ event = cpuc->event_list[j];
+ if (event->group_leader != leader->group_leader)
+ break;
+ for_each_set_bit(idx, (unsigned long *)&event->attr.config2, X86_PMC_IDX_MAX) {
+ if (WARN_ON_ONCE(i + idx > cpuc->n_events))
+ return;
+ __set_bit(cpuc->assign[i + idx], (unsigned long *)&event->hw.config1);
+ }
+ }
+ i = j - 1;
+ }
+}
+
+void intel_pmu_late_setup(void)
+{
+ struct cpu_hw_events *cpuc = this_cpu_ptr(&cpu_hw_events);
+
+ if (!cpuc->n_late_setup)
+ return;
+
+ intel_pmu_pebs_late_setup(cpuc);
+ intel_pmu_acr_late_setup(cpuc);
+}
+
static void intel_pmu_add_event(struct perf_event *event)
{
if (event->attr.precise_ip)
intel_pmu_pebs_add(event);
if (intel_pmu_needs_branch_stack(event))
intel_pmu_lbr_add(event);
+ if (is_pebs_counter_event_group(event) ||
+ is_acr_event_group(event))
+ this_cpu_ptr(&cpu_hw_events)->n_late_setup++;
}
/*
@@ -2937,7 +3041,7 @@ int intel_pmu_save_and_restart(struct perf_event *event)
*/
if (unlikely(event_is_checkpointed(event))) {
/* No race with NMIs because the counter should not be armed */
- wrmsrl(event->hw.event_base, 0);
+ wrmsrq(event->hw.event_base, 0);
local64_set(&event->hw.prev_count, 0);
}
return static_call(x86_pmu_set_period)(event);
@@ -2976,13 +3080,13 @@ static void intel_pmu_reset(void)
pr_info("clearing PMU state on CPU#%d\n", smp_processor_id());
for_each_set_bit(idx, cntr_mask, INTEL_PMC_MAX_GENERIC) {
- wrmsrl_safe(x86_pmu_config_addr(idx), 0ull);
- wrmsrl_safe(x86_pmu_event_addr(idx), 0ull);
+ wrmsrq_safe(x86_pmu_config_addr(idx), 0ull);
+ wrmsrq_safe(x86_pmu_event_addr(idx), 0ull);
}
for_each_set_bit(idx, fixed_cntr_mask, INTEL_PMC_MAX_FIXED) {
if (fixed_counter_disabled(idx, cpuc->pmu))
continue;
- wrmsrl_safe(x86_pmu_fixed_ctr_addr(idx), 0ull);
+ wrmsrq_safe(x86_pmu_fixed_ctr_addr(idx), 0ull);
}
if (ds)
@@ -2991,7 +3095,7 @@ static void intel_pmu_reset(void)
/* Ack all overflows and disable fixed counters */
if (x86_pmu.version >= 2) {
intel_pmu_ack_status(intel_pmu_get_status());
- wrmsrl(MSR_CORE_PERF_GLOBAL_CTRL, 0);
+ wrmsrq(MSR_CORE_PERF_GLOBAL_CTRL, 0);
}
/* Reset LBRs and LBR freezing */
@@ -3035,8 +3139,7 @@ static void x86_pmu_handle_guest_pebs(struct pt_regs *regs,
continue;
perf_sample_data_init(data, 0, event->hw.last_period);
- if (perf_event_overflow(event, data, regs))
- x86_pmu_stop(event, 0);
+ perf_event_overflow(event, data, regs);
/* Inject one fake event is enough. */
break;
@@ -3101,7 +3204,7 @@ static int handle_pmi_common(struct pt_regs *regs, u64 status)
* Update the MSR if pebs_enabled is changed.
*/
if (pebs_enabled != cpuc->pebs_enabled)
- wrmsrl(MSR_IA32_PEBS_ENABLE, cpuc->pebs_enabled);
+ wrmsrq(MSR_IA32_PEBS_ENABLE, cpuc->pebs_enabled);
/*
* Above PEBS handler (PEBS counters snapshotting) has updated fixed
@@ -3141,6 +3244,7 @@ static int handle_pmi_common(struct pt_regs *regs, u64 status)
for_each_set_bit(bit, (unsigned long *)&status, X86_PMC_IDX_MAX) {
struct perf_event *event = cpuc->events[bit];
+ u64 last_period;
handled++;
@@ -3168,16 +3272,17 @@ static int handle_pmi_common(struct pt_regs *regs, u64 status)
if (is_pebs_counter_event_group(event))
x86_pmu.drain_pebs(regs, &data);
+ last_period = event->hw.last_period;
+
if (!intel_pmu_save_and_restart(event))
continue;
- perf_sample_data_init(&data, 0, event->hw.last_period);
+ perf_sample_data_init(&data, 0, last_period);
if (has_branch_stack(event))
intel_pmu_lbr_save_brstack(&data, cpuc, event);
- if (perf_event_overflow(event, &data, regs))
- x86_pmu_stop(event, 0);
+ perf_event_overflow(event, &data, regs);
}
return handled;
@@ -3739,10 +3844,9 @@ intel_get_event_constraints(struct cpu_hw_events *cpuc, int idx,
if (cpuc->excl_cntrs)
return intel_get_excl_constraints(cpuc, event, idx, c2);
- /* Not all counters support the branch counter feature. */
- if (branch_sample_counters(event)) {
+ if (event->hw.dyn_constraint != ~0ULL) {
c2 = dyn_constraint(cpuc, c2, idx);
- c2->idxmsk64 &= x86_pmu.lbr_counters;
+ c2->idxmsk64 &= event->hw.dyn_constraint;
c2->weight = hweight64(c2->idxmsk64);
}
@@ -4083,6 +4187,39 @@ end:
return start;
}
+static inline bool intel_pmu_has_acr(struct pmu *pmu)
+{
+ return !!hybrid(pmu, acr_cause_mask64);
+}
+
+static bool intel_pmu_is_acr_group(struct perf_event *event)
+{
+ /* The group leader has the ACR flag set */
+ if (is_acr_event_group(event))
+ return true;
+
+ /* The acr_mask is set */
+ if (event->attr.config2)
+ return true;
+
+ return false;
+}
+
+static inline void intel_pmu_set_acr_cntr_constr(struct perf_event *event,
+ u64 *cause_mask, int *num)
+{
+ event->hw.dyn_constraint &= hybrid(event->pmu, acr_cntr_mask64);
+ *cause_mask |= event->attr.config2;
+ *num += 1;
+}
+
+static inline void intel_pmu_set_acr_caused_constr(struct perf_event *event,
+ int idx, u64 cause_mask)
+{
+ if (test_bit(idx, (unsigned long *)&cause_mask))
+ event->hw.dyn_constraint &= hybrid(event->pmu, acr_cause_mask64);
+}
+
static int intel_pmu_hw_config(struct perf_event *event)
{
int ret = x86_pmu_hw_config(event);
@@ -4144,15 +4281,19 @@ static int intel_pmu_hw_config(struct perf_event *event)
leader = event->group_leader;
if (branch_sample_call_stack(leader))
return -EINVAL;
- if (branch_sample_counters(leader))
+ if (branch_sample_counters(leader)) {
num++;
+ leader->hw.dyn_constraint &= x86_pmu.lbr_counters;
+ }
leader->hw.flags |= PERF_X86_EVENT_BRANCH_COUNTERS;
for_each_sibling_event(sibling, leader) {
if (branch_sample_call_stack(sibling))
return -EINVAL;
- if (branch_sample_counters(sibling))
+ if (branch_sample_counters(sibling)) {
num++;
+ sibling->hw.dyn_constraint &= x86_pmu.lbr_counters;
+ }
}
if (num > fls(x86_pmu.lbr_counters))
@@ -4207,6 +4348,94 @@ static int intel_pmu_hw_config(struct perf_event *event)
event->attr.precise_ip)
event->group_leader->hw.flags |= PERF_X86_EVENT_PEBS_CNTR;
+ if (intel_pmu_has_acr(event->pmu) && intel_pmu_is_acr_group(event)) {
+ struct perf_event *sibling, *leader = event->group_leader;
+ struct pmu *pmu = event->pmu;
+ bool has_sw_event = false;
+ int num = 0, idx = 0;
+ u64 cause_mask = 0;
+
+ /* Not support perf metrics */
+ if (is_metric_event(event))
+ return -EINVAL;
+
+ /* Not support freq mode */
+ if (event->attr.freq)
+ return -EINVAL;
+
+ /* PDist is not supported */
+ if (event->attr.config2 && event->attr.precise_ip > 2)
+ return -EINVAL;
+
+ /* The reload value cannot exceeds the max period */
+ if (event->attr.sample_period > x86_pmu.max_period)
+ return -EINVAL;
+ /*
+ * The counter-constraints of each event cannot be finalized
+ * unless the whole group is scanned. However, it's hard
+ * to know whether the event is the last one of the group.
+ * Recalculate the counter-constraints for each event when
+ * adding a new event.
+ *
+ * The group is traversed twice, which may be optimized later.
+ * In the first round,
+ * - Find all events which do reload when other events
+ * overflow and set the corresponding counter-constraints
+ * - Add all events, which can cause other events reload,
+ * in the cause_mask
+ * - Error out if the number of events exceeds the HW limit
+ * - The ACR events must be contiguous.
+ * Error out if there are non-X86 events between ACR events.
+ * This is not a HW limit, but a SW limit.
+ * With the assumption, the intel_pmu_acr_late_setup() can
+ * easily convert the event idx to counter idx without
+ * traversing the whole event list.
+ */
+ if (!is_x86_event(leader))
+ return -EINVAL;
+
+ if (leader->attr.config2)
+ intel_pmu_set_acr_cntr_constr(leader, &cause_mask, &num);
+
+ if (leader->nr_siblings) {
+ for_each_sibling_event(sibling, leader) {
+ if (!is_x86_event(sibling)) {
+ has_sw_event = true;
+ continue;
+ }
+ if (!sibling->attr.config2)
+ continue;
+ if (has_sw_event)
+ return -EINVAL;
+ intel_pmu_set_acr_cntr_constr(sibling, &cause_mask, &num);
+ }
+ }
+ if (leader != event && event->attr.config2) {
+ if (has_sw_event)
+ return -EINVAL;
+ intel_pmu_set_acr_cntr_constr(event, &cause_mask, &num);
+ }
+
+ if (hweight64(cause_mask) > hweight64(hybrid(pmu, acr_cause_mask64)) ||
+ num > hweight64(hybrid(event->pmu, acr_cntr_mask64)))
+ return -EINVAL;
+ /*
+ * In the second round, apply the counter-constraints for
+ * the events which can cause other events reload.
+ */
+ intel_pmu_set_acr_caused_constr(leader, idx++, cause_mask);
+
+ if (leader->nr_siblings) {
+ for_each_sibling_event(sibling, leader)
+ intel_pmu_set_acr_caused_constr(sibling, idx++, cause_mask);
+ }
+
+ if (leader != event)
+ intel_pmu_set_acr_caused_constr(event, idx, cause_mask);
+
+ leader->hw.flags |= PERF_X86_EVENT_ACR;
+ }
+
if ((event->attr.type == PERF_TYPE_HARDWARE) ||
(event->attr.type == PERF_TYPE_HW_CACHE))
return 0;
@@ -4354,7 +4583,7 @@ static struct perf_guest_switch_msr *intel_guest_get_msrs(int *nr, void *data)
.guest = intel_ctrl & ~cpuc->intel_ctrl_host_mask & ~pebs_mask,
};
- if (!x86_pmu.pebs)
+ if (!x86_pmu.ds_pebs)
return arr;
/*
@@ -4952,7 +5181,7 @@ int intel_cpuc_prepare(struct cpu_hw_events *cpuc, int cpu)
goto err;
}
- if (x86_pmu.flags & (PMU_FL_EXCL_CNTRS | PMU_FL_TFA | PMU_FL_BR_CNTR)) {
+ if (x86_pmu.flags & (PMU_FL_EXCL_CNTRS | PMU_FL_TFA | PMU_FL_DYN_CONSTRAINT)) {
size_t sz = X86_PMC_IDX_MAX * sizeof(struct event_constraint);
cpuc->constraint_list = kzalloc_node(sz, GFP_KERNEL, cpu_to_node(cpu));
@@ -5041,7 +5270,7 @@ static inline bool intel_pmu_broken_perf_cap(void)
return false;
}
-static void update_pmu_cap(struct x86_hybrid_pmu *pmu)
+static void update_pmu_cap(struct pmu *pmu)
{
unsigned int cntr, fixed_cntr, ecx, edx;
union cpuid35_eax eax;
@@ -5050,20 +5279,30 @@ static void update_pmu_cap(struct x86_hybrid_pmu *pmu)
cpuid(ARCH_PERFMON_EXT_LEAF, &eax.full, &ebx.full, &ecx, &edx);
if (ebx.split.umask2)
- pmu->config_mask |= ARCH_PERFMON_EVENTSEL_UMASK2;
+ hybrid(pmu, config_mask) |= ARCH_PERFMON_EVENTSEL_UMASK2;
if (ebx.split.eq)
- pmu->config_mask |= ARCH_PERFMON_EVENTSEL_EQ;
+ hybrid(pmu, config_mask) |= ARCH_PERFMON_EVENTSEL_EQ;
if (eax.split.cntr_subleaf) {
cpuid_count(ARCH_PERFMON_EXT_LEAF, ARCH_PERFMON_NUM_COUNTER_LEAF,
&cntr, &fixed_cntr, &ecx, &edx);
- pmu->cntr_mask64 = cntr;
- pmu->fixed_cntr_mask64 = fixed_cntr;
+ hybrid(pmu, cntr_mask64) = cntr;
+ hybrid(pmu, fixed_cntr_mask64) = fixed_cntr;
+ }
+
+ if (eax.split.acr_subleaf) {
+ cpuid_count(ARCH_PERFMON_EXT_LEAF, ARCH_PERFMON_ACR_LEAF,
+ &cntr, &fixed_cntr, &ecx, &edx);
+ /* The mask of the counters which can be reloaded */
+ hybrid(pmu, acr_cntr_mask64) = cntr | ((u64)fixed_cntr << INTEL_PMC_IDX_FIXED);
+
+ /* The mask of the counters which can cause a reload of reloadable counters */
+ hybrid(pmu, acr_cause_mask64) = ecx | ((u64)edx << INTEL_PMC_IDX_FIXED);
}
if (!intel_pmu_broken_perf_cap()) {
/* Perf Metric (Bit 15) and PEBS via PT (Bit 16) are hybrid enumeration */
- rdmsrl(MSR_IA32_PERF_CAPABILITIES, pmu->intel_cap.capabilities);
+ rdmsrq(MSR_IA32_PERF_CAPABILITIES, hybrid(pmu, intel_cap).capabilities);
}
}
@@ -5150,7 +5389,7 @@ static bool init_hybrid_pmu(int cpu)
goto end;
if (this_cpu_has(X86_FEATURE_ARCH_PERFMON_EXT))
- update_pmu_cap(pmu);
+ update_pmu_cap(&pmu->pmu);
intel_pmu_check_hybrid_pmus(pmu);
@@ -5211,7 +5450,7 @@ static void intel_pmu_cpu_starting(int cpu)
if (!is_hybrid() && x86_pmu.intel_cap.perf_metrics) {
union perf_capabilities perf_cap;
- rdmsrl(MSR_IA32_PERF_CAPABILITIES, perf_cap.capabilities);
+ rdmsrq(MSR_IA32_PERF_CAPABILITIES, perf_cap.capabilities);
if (!perf_cap.perf_metrics) {
x86_pmu.intel_cap.perf_metrics = 0;
x86_pmu.intel_ctrl &= ~(1ULL << GLOBAL_CTRL_EN_PERF_METRICS);
@@ -5524,7 +5763,7 @@ static __init void intel_clovertown_quirk(void)
* these chips.
*/
pr_warn("PEBS disabled due to CPU errata\n");
- x86_pmu.pebs = 0;
+ x86_pmu.ds_pebs = 0;
x86_pmu.pebs_constraints = NULL;
}
@@ -5619,24 +5858,24 @@ static bool check_msr(unsigned long msr, u64 mask)
* matches, this is needed to detect certain hardware emulators
* (qemu/kvm) that don't trap on the MSR access and always return 0s.
*/
- if (rdmsrl_safe(msr, &val_old))
+ if (rdmsrq_safe(msr, &val_old))
return false;
/*
- * Only change the bits which can be updated by wrmsrl.
+ * Only change the bits which can be updated by wrmsrq.
*/
val_tmp = val_old ^ mask;
if (is_lbr_from(msr))
val_tmp = lbr_from_signext_quirk_wr(val_tmp);
- if (wrmsrl_safe(msr, val_tmp) ||
- rdmsrl_safe(msr, &val_new))
+ if (wrmsrq_safe(msr, val_tmp) ||
+ rdmsrq_safe(msr, &val_new))
return false;
/*
- * Quirk only affects validation in wrmsr(), so wrmsrl()'s value
- * should equal rdmsrl()'s even with the quirk.
+ * Quirk only affects validation in wrmsr(), so wrmsrq()'s value
+ * should equal rdmsrq()'s even with the quirk.
*/
if (val_new != val_tmp)
return false;
@@ -5647,7 +5886,7 @@ static bool check_msr(unsigned long msr, u64 mask)
/* Here it's sure that the MSR can be safely accessed.
* Restore the old value and return.
*/
- wrmsrl(msr, val_old);
+ wrmsrq(msr, val_old);
return true;
}
@@ -6012,7 +6251,7 @@ tsx_is_visible(struct kobject *kobj, struct attribute *attr, int i)
static umode_t
pebs_is_visible(struct kobject *kobj, struct attribute *attr, int i)
{
- return x86_pmu.pebs ? attr->mode : 0;
+ return x86_pmu.ds_pebs ? attr->mode : 0;
}
static umode_t
@@ -6043,6 +6282,21 @@ td_is_visible(struct kobject *kobj, struct attribute *attr, int i)
return attr->mode;
}
+PMU_FORMAT_ATTR(acr_mask, "config2:0-63");
+
+static struct attribute *format_acr_attrs[] = {
+ &format_attr_acr_mask.attr,
+ NULL
+};
+
+static umode_t
+acr_is_visible(struct kobject *kobj, struct attribute *attr, int i)
+{
+ struct device *dev = kobj_to_dev(kobj);
+
+ return intel_pmu_has_acr(dev_get_drvdata(dev)) ? attr->mode : 0;
+}
+
static struct attribute_group group_events_td = {
.name = "events",
.is_visible = td_is_visible,
@@ -6085,6 +6339,12 @@ static struct attribute_group group_format_evtsel_ext = {
.is_visible = evtsel_ext_is_visible,
};
+static struct attribute_group group_format_acr = {
+ .name = "format",
+ .attrs = format_acr_attrs,
+ .is_visible = acr_is_visible,
+};
+
static struct attribute_group group_default = {
.attrs = intel_pmu_attrs,
.is_visible = default_is_visible,
@@ -6099,6 +6359,7 @@ static const struct attribute_group *attr_update[] = {
&group_format_extra,
&group_format_extra_skl,
&group_format_evtsel_ext,
+ &group_format_acr,
&group_default,
NULL,
};
@@ -6383,6 +6644,7 @@ static const struct attribute_group *hybrid_attr_update[] = {
&group_caps_lbr,
&hybrid_group_format_extra,
&group_format_evtsel_ext,
+ &group_format_acr,
&group_default,
&hybrid_group_cpus,
NULL,
@@ -6575,6 +6837,7 @@ static __always_inline void intel_pmu_init_skt(struct pmu *pmu)
intel_pmu_init_grt(pmu);
hybrid(pmu, event_constraints) = intel_skt_event_constraints;
hybrid(pmu, extra_regs) = intel_cmt_extra_regs;
+ static_call_update(intel_pmu_enable_acr_event, intel_pmu_enable_acr);
}
__init int intel_pmu_init(void)
@@ -6635,6 +6898,7 @@ __init int intel_pmu_init(void)
x86_pmu.pebs_events_mask = intel_pmu_pebs_mask(x86_pmu.cntr_mask64);
x86_pmu.pebs_capable = PEBS_COUNTER_MASK;
+ x86_pmu.config_mask = X86_RAW_EVENT_MASK;
/*
* Quirk: v2 perfmon does not report fixed-purpose events, so
@@ -6651,7 +6915,7 @@ __init int intel_pmu_init(void)
if (boot_cpu_has(X86_FEATURE_PDCM)) {
u64 capabilities;
- rdmsrl(MSR_IA32_PERF_CAPABILITIES, capabilities);
+ rdmsrq(MSR_IA32_PERF_CAPABILITIES, capabilities);
x86_pmu.intel_cap.capabilities = capabilities;
}
@@ -6663,7 +6927,7 @@ __init int intel_pmu_init(void)
if (boot_cpu_has(X86_FEATURE_ARCH_LBR))
intel_pmu_arch_lbr_init();
- intel_ds_init();
+ intel_pebs_init();
x86_add_quirk(intel_arch_events_quirk); /* Install first, so it runs last */
@@ -6674,6 +6938,12 @@ __init int intel_pmu_init(void)
}
/*
+ * Many features on and after V6 require dynamic constraint,
+ * e.g., Arch PEBS, ACR.
+ */
+ if (version >= 6)
+ x86_pmu.flags |= PMU_FL_DYN_CONSTRAINT;
+ /*
* Install the hw-cache-events table:
*/
switch (boot_cpu_data.x86_vfm) {
@@ -6884,6 +7154,18 @@ __init int intel_pmu_init(void)
name = "crestmont";
break;
+ case INTEL_ATOM_DARKMONT_X:
+ intel_pmu_init_skt(NULL);
+ intel_pmu_pebs_data_source_cmt();
+ x86_pmu.pebs_latency_data = cmt_latency_data;
+ x86_pmu.get_event_constraints = cmt_get_event_constraints;
+ td_attr = skt_events_attrs;
+ mem_attr = grt_mem_attrs;
+ extra_attr = cmt_format_attr;
+ pr_cont("Darkmont events, ");
+ name = "darkmont";
+ break;
+
case INTEL_WESTMERE:
case INTEL_WESTMERE_EP:
case INTEL_WESTMERE_EX:
@@ -7433,6 +7715,18 @@ __init int intel_pmu_init(void)
x86_pmu.attr_update = hybrid_attr_update;
}
+ /*
+ * The archPerfmonExt (0x23) includes an enhanced enumeration of
+ * PMU architectural features with a per-core view. For non-hybrid,
+ * each core has the same PMU capabilities. It's good enough to
+ * update the x86_pmu from the booting CPU. For hybrid, the x86_pmu
+ * is used to keep the common capabilities. Still keep the values
+ * from the leaf 0xa. The core specific update will be done later
+ * when a new type is online.
+ */
+ if (!is_hybrid() && boot_cpu_has(X86_FEATURE_ARCH_PERFMON_EXT))
+ update_pmu_cap(NULL);
+
intel_pmu_check_counters_mask(&x86_pmu.cntr_mask64,
&x86_pmu.fixed_cntr_mask64,
&x86_pmu.intel_ctrl);
diff --git a/arch/x86/events/intel/cstate.c b/arch/x86/events/intel/cstate.c
index ae4ec16156bb..ec753e39b007 100644
--- a/arch/x86/events/intel/cstate.c
+++ b/arch/x86/events/intel/cstate.c
@@ -111,6 +111,7 @@
#include <linux/nospec.h>
#include <asm/cpu_device_id.h>
#include <asm/intel-family.h>
+#include <asm/msr.h>
#include "../perf_event.h"
#include "../probe.h"
@@ -320,7 +321,7 @@ static inline u64 cstate_pmu_read_counter(struct perf_event *event)
{
u64 val;
- rdmsrl(event->hw.event_base, val);
+ rdmsrq(event->hw.event_base, val);
return val;
}
diff --git a/arch/x86/events/intel/ds.c b/arch/x86/events/intel/ds.c
index 8d86e91bd5e5..c0b7ac1c7594 100644
--- a/arch/x86/events/intel/ds.c
+++ b/arch/x86/events/intel/ds.c
@@ -10,6 +10,7 @@
#include <asm/tlbflush.h>
#include <asm/insn.h>
#include <asm/io.h>
+#include <asm/msr.h>
#include <asm/timer.h>
#include "../perf_event.h"
@@ -624,7 +625,7 @@ static int alloc_pebs_buffer(int cpu)
int max, node = cpu_to_node(cpu);
void *buffer, *insn_buff, *cea;
- if (!x86_pmu.pebs)
+ if (!x86_pmu.ds_pebs)
return 0;
buffer = dsalloc_pages(bsiz, GFP_KERNEL, cpu);
@@ -659,7 +660,7 @@ static void release_pebs_buffer(int cpu)
struct cpu_hw_events *hwev = per_cpu_ptr(&cpu_hw_events, cpu);
void *cea;
- if (!x86_pmu.pebs)
+ if (!x86_pmu.ds_pebs)
return;
kfree(per_cpu(insn_buffer, cpu));
@@ -734,7 +735,7 @@ void release_ds_buffers(void)
{
int cpu;
- if (!x86_pmu.bts && !x86_pmu.pebs)
+ if (!x86_pmu.bts && !x86_pmu.ds_pebs)
return;
for_each_possible_cpu(cpu)
@@ -750,7 +751,8 @@ void release_ds_buffers(void)
}
for_each_possible_cpu(cpu) {
- release_pebs_buffer(cpu);
+ if (x86_pmu.ds_pebs)
+ release_pebs_buffer(cpu);
release_bts_buffer(cpu);
}
}
@@ -761,15 +763,17 @@ void reserve_ds_buffers(void)
int cpu;
x86_pmu.bts_active = 0;
- x86_pmu.pebs_active = 0;
- if (!x86_pmu.bts && !x86_pmu.pebs)
+ if (x86_pmu.ds_pebs)
+ x86_pmu.pebs_active = 0;
+
+ if (!x86_pmu.bts && !x86_pmu.ds_pebs)
return;
if (!x86_pmu.bts)
bts_err = 1;
- if (!x86_pmu.pebs)
+ if (!x86_pmu.ds_pebs)
pebs_err = 1;
for_each_possible_cpu(cpu) {
@@ -781,7 +785,8 @@ void reserve_ds_buffers(void)
if (!bts_err && alloc_bts_buffer(cpu))
bts_err = 1;
- if (!pebs_err && alloc_pebs_buffer(cpu))
+ if (x86_pmu.ds_pebs && !pebs_err &&
+ alloc_pebs_buffer(cpu))
pebs_err = 1;
if (bts_err && pebs_err)
@@ -793,7 +798,7 @@ void reserve_ds_buffers(void)
release_bts_buffer(cpu);
}
- if (pebs_err) {
+ if (x86_pmu.ds_pebs && pebs_err) {
for_each_possible_cpu(cpu)
release_pebs_buffer(cpu);
}
@@ -805,7 +810,7 @@ void reserve_ds_buffers(void)
if (x86_pmu.bts && !bts_err)
x86_pmu.bts_active = 1;
- if (x86_pmu.pebs && !pebs_err)
+ if (x86_pmu.ds_pebs && !pebs_err)
x86_pmu.pebs_active = 1;
for_each_possible_cpu(cpu) {
@@ -1355,9 +1360,8 @@ static void __intel_pmu_pebs_update_cfg(struct perf_event *event,
}
-static void intel_pmu_late_setup(void)
+void intel_pmu_pebs_late_setup(struct cpu_hw_events *cpuc)
{
- struct cpu_hw_events *cpuc = this_cpu_ptr(&cpu_hw_events);
struct perf_event *event;
u64 pebs_data_cfg = 0;
int i;
@@ -1517,7 +1521,7 @@ static void intel_pmu_pebs_via_pt_enable(struct perf_event *event)
else
value = ds->pebs_event_reset[MAX_PEBS_EVENTS + idx];
}
- wrmsrl(base + idx, value);
+ wrmsrq(base + idx, value);
}
static inline void intel_pmu_drain_large_pebs(struct cpu_hw_events *cpuc)
@@ -1554,7 +1558,7 @@ void intel_pmu_pebs_enable(struct perf_event *event)
*/
intel_pmu_drain_pebs_buffer();
adaptive_pebs_record_size_update();
- wrmsrl(MSR_PEBS_DATA_CFG, pebs_data_cfg);
+ wrmsrq(MSR_PEBS_DATA_CFG, pebs_data_cfg);
cpuc->active_pebs_data_cfg = pebs_data_cfg;
}
}
@@ -1617,7 +1621,7 @@ void intel_pmu_pebs_disable(struct perf_event *event)
intel_pmu_pebs_via_pt_disable(event);
if (cpuc->enabled)
- wrmsrl(MSR_IA32_PEBS_ENABLE, cpuc->pebs_enabled);
+ wrmsrq(MSR_IA32_PEBS_ENABLE, cpuc->pebs_enabled);
hwc->config |= ARCH_PERFMON_EVENTSEL_INT;
}
@@ -1627,7 +1631,7 @@ void intel_pmu_pebs_enable_all(void)
struct cpu_hw_events *cpuc = this_cpu_ptr(&cpu_hw_events);
if (cpuc->pebs_enabled)
- wrmsrl(MSR_IA32_PEBS_ENABLE, cpuc->pebs_enabled);
+ wrmsrq(MSR_IA32_PEBS_ENABLE, cpuc->pebs_enabled);
}
void intel_pmu_pebs_disable_all(void)
@@ -1828,8 +1832,6 @@ static void setup_pebs_fixed_sample_data(struct perf_event *event,
perf_sample_data_init(data, 0, event->hw.last_period);
- data->period = event->hw.last_period;
-
/*
* Use latency for weight (only avail with PEBS-LL)
*/
@@ -2082,7 +2084,6 @@ static void setup_pebs_adaptive_sample_data(struct perf_event *event,
sample_type = event->attr.sample_type;
format_group = basic->format_group;
perf_sample_data_init(data, 0, event->hw.last_period);
- data->period = event->hw.last_period;
setup_pebs_time(event, data, basic->tsc);
@@ -2276,7 +2277,7 @@ intel_pmu_save_and_restart_reload(struct perf_event *event, int count)
WARN_ON(this_cpu_read(cpu_hw_events.enabled));
prev_raw_count = local64_read(&hwc->prev_count);
- rdpmcl(hwc->event_base_rdpmc, new_raw_count);
+ new_raw_count = rdpmc(hwc->event_base_rdpmc);
local64_set(&hwc->prev_count, new_raw_count);
/*
@@ -2359,8 +2360,7 @@ __intel_pmu_pebs_last_event(struct perf_event *event,
* All but the last records are processed.
* The last one is left to be able to call the overflow handler.
*/
- if (perf_event_overflow(event, data, regs))
- x86_pmu_stop(event, 0);
+ perf_event_overflow(event, data, regs);
}
if (hwc->flags & PERF_X86_EVENT_AUTO_RELOAD) {
@@ -2589,8 +2589,8 @@ static void intel_pmu_drain_pebs_nhm(struct pt_regs *iregs, struct perf_sample_d
if (error[bit]) {
perf_log_lost_samples(event, error[bit]);
- if (iregs && perf_event_account_interrupt(event))
- x86_pmu_stop(event, 0);
+ if (iregs)
+ perf_event_account_interrupt(event);
}
if (counts[bit]) {
@@ -2670,10 +2670,10 @@ static void intel_pmu_drain_pebs_icl(struct pt_regs *iregs, struct perf_sample_d
}
/*
- * BTS, PEBS probe and setup
+ * PEBS probe and setup
*/
-void __init intel_ds_init(void)
+void __init intel_pebs_init(void)
{
/*
* No support for 32bit formats
@@ -2681,13 +2681,12 @@ void __init intel_ds_init(void)
if (!boot_cpu_has(X86_FEATURE_DTES64))
return;
- x86_pmu.bts = boot_cpu_has(X86_FEATURE_BTS);
- x86_pmu.pebs = boot_cpu_has(X86_FEATURE_PEBS);
+ x86_pmu.ds_pebs = boot_cpu_has(X86_FEATURE_PEBS);
x86_pmu.pebs_buffer_size = PEBS_BUFFER_SIZE;
if (x86_pmu.version <= 4)
x86_pmu.pebs_no_isolation = 1;
- if (x86_pmu.pebs) {
+ if (x86_pmu.ds_pebs) {
char pebs_type = x86_pmu.intel_cap.pebs_trap ? '+' : '-';
char *pebs_qual = "";
int format = x86_pmu.intel_cap.pebs_format;
@@ -2695,6 +2694,11 @@ void __init intel_ds_init(void)
if (format < 4)
x86_pmu.intel_cap.pebs_baseline = 0;
+ x86_pmu.pebs_enable = intel_pmu_pebs_enable;
+ x86_pmu.pebs_disable = intel_pmu_pebs_disable;
+ x86_pmu.pebs_enable_all = intel_pmu_pebs_enable_all;
+ x86_pmu.pebs_disable_all = intel_pmu_pebs_disable_all;
+
switch (format) {
case 0:
pr_cont("PEBS fmt0%c, ", pebs_type);
@@ -2779,7 +2783,7 @@ void __init intel_ds_init(void)
default:
pr_cont("no PEBS fmt%d%c, ", format, pebs_type);
- x86_pmu.pebs = 0;
+ x86_pmu.ds_pebs = 0;
}
}
}
@@ -2788,8 +2792,8 @@ void perf_restore_debug_store(void)
{
struct debug_store *ds = __this_cpu_read(cpu_hw_events.ds);
- if (!x86_pmu.bts && !x86_pmu.pebs)
+ if (!x86_pmu.bts && !x86_pmu.ds_pebs)
return;
- wrmsrl(MSR_IA32_DS_AREA, (unsigned long)ds);
+ wrmsrq(MSR_IA32_DS_AREA, (unsigned long)ds);
}
diff --git a/arch/x86/events/intel/knc.c b/arch/x86/events/intel/knc.c
index 034a1f6a457c..e614baf42926 100644
--- a/arch/x86/events/intel/knc.c
+++ b/arch/x86/events/intel/knc.c
@@ -5,6 +5,7 @@
#include <linux/types.h>
#include <asm/hardirq.h>
+#include <asm/msr.h>
#include "../perf_event.h"
@@ -159,18 +160,18 @@ static void knc_pmu_disable_all(void)
{
u64 val;
- rdmsrl(MSR_KNC_IA32_PERF_GLOBAL_CTRL, val);
+ rdmsrq(MSR_KNC_IA32_PERF_GLOBAL_CTRL, val);
val &= ~(KNC_ENABLE_COUNTER0|KNC_ENABLE_COUNTER1);
- wrmsrl(MSR_KNC_IA32_PERF_GLOBAL_CTRL, val);
+ wrmsrq(MSR_KNC_IA32_PERF_GLOBAL_CTRL, val);
}
static void knc_pmu_enable_all(int added)
{
u64 val;
- rdmsrl(MSR_KNC_IA32_PERF_GLOBAL_CTRL, val);
+ rdmsrq(MSR_KNC_IA32_PERF_GLOBAL_CTRL, val);
val |= (KNC_ENABLE_COUNTER0|KNC_ENABLE_COUNTER1);
- wrmsrl(MSR_KNC_IA32_PERF_GLOBAL_CTRL, val);
+ wrmsrq(MSR_KNC_IA32_PERF_GLOBAL_CTRL, val);
}
static inline void
@@ -182,7 +183,7 @@ knc_pmu_disable_event(struct perf_event *event)
val = hwc->config;
val &= ~ARCH_PERFMON_EVENTSEL_ENABLE;
- (void)wrmsrl_safe(hwc->config_base + hwc->idx, val);
+ (void)wrmsrq_safe(hwc->config_base + hwc->idx, val);
}
static void knc_pmu_enable_event(struct perf_event *event)
@@ -193,21 +194,21 @@ static void knc_pmu_enable_event(struct perf_event *event)
val = hwc->config;
val |= ARCH_PERFMON_EVENTSEL_ENABLE;
- (void)wrmsrl_safe(hwc->config_base + hwc->idx, val);
+ (void)wrmsrq_safe(hwc->config_base + hwc->idx, val);
}
static inline u64 knc_pmu_get_status(void)
{
u64 status;
- rdmsrl(MSR_KNC_IA32_PERF_GLOBAL_STATUS, status);
+ rdmsrq(MSR_KNC_IA32_PERF_GLOBAL_STATUS, status);
return status;
}
static inline void knc_pmu_ack_status(u64 ack)
{
- wrmsrl(MSR_KNC_IA32_PERF_GLOBAL_OVF_CONTROL, ack);
+ wrmsrq(MSR_KNC_IA32_PERF_GLOBAL_OVF_CONTROL, ack);
}
static int knc_pmu_handle_irq(struct pt_regs *regs)
@@ -241,19 +242,20 @@ again:
for_each_set_bit(bit, (unsigned long *)&status, X86_PMC_IDX_MAX) {
struct perf_event *event = cpuc->events[bit];
+ u64 last_period;
handled++;
if (!test_bit(bit, cpuc->active_mask))
continue;
+ last_period = event->hw.last_period;
if (!intel_pmu_save_and_restart(event))
continue;
- perf_sample_data_init(&data, 0, event->hw.last_period);
+ perf_sample_data_init(&data, 0, last_period);
- if (perf_event_overflow(event, &data, regs))
- x86_pmu_stop(event, 0);
+ perf_event_overflow(event, &data, regs);
}
/*
diff --git a/arch/x86/events/intel/lbr.c b/arch/x86/events/intel/lbr.c
index f44c3d866f24..7aa59966e7c3 100644
--- a/arch/x86/events/intel/lbr.c
+++ b/arch/x86/events/intel/lbr.c
@@ -137,9 +137,9 @@ static void __intel_pmu_lbr_enable(bool pmi)
if (cpuc->lbr_sel)
lbr_select = cpuc->lbr_sel->config & x86_pmu.lbr_sel_mask;
if (!static_cpu_has(X86_FEATURE_ARCH_LBR) && !pmi && cpuc->lbr_sel)
- wrmsrl(MSR_LBR_SELECT, lbr_select);
+ wrmsrq(MSR_LBR_SELECT, lbr_select);
- rdmsrl(MSR_IA32_DEBUGCTLMSR, debugctl);
+ rdmsrq(MSR_IA32_DEBUGCTLMSR, debugctl);
orig_debugctl = debugctl;
if (!static_cpu_has(X86_FEATURE_ARCH_LBR))
@@ -155,10 +155,10 @@ static void __intel_pmu_lbr_enable(bool pmi)
debugctl |= DEBUGCTLMSR_FREEZE_LBRS_ON_PMI;
if (orig_debugctl != debugctl)
- wrmsrl(MSR_IA32_DEBUGCTLMSR, debugctl);
+ wrmsrq(MSR_IA32_DEBUGCTLMSR, debugctl);
if (static_cpu_has(X86_FEATURE_ARCH_LBR))
- wrmsrl(MSR_ARCH_LBR_CTL, lbr_select | ARCH_LBR_CTL_LBREN);
+ wrmsrq(MSR_ARCH_LBR_CTL, lbr_select | ARCH_LBR_CTL_LBREN);
}
void intel_pmu_lbr_reset_32(void)
@@ -166,7 +166,7 @@ void intel_pmu_lbr_reset_32(void)
int i;
for (i = 0; i < x86_pmu.lbr_nr; i++)
- wrmsrl(x86_pmu.lbr_from + i, 0);
+ wrmsrq(x86_pmu.lbr_from + i, 0);
}
void intel_pmu_lbr_reset_64(void)
@@ -174,17 +174,17 @@ void intel_pmu_lbr_reset_64(void)
int i;
for (i = 0; i < x86_pmu.lbr_nr; i++) {
- wrmsrl(x86_pmu.lbr_from + i, 0);
- wrmsrl(x86_pmu.lbr_to + i, 0);
+ wrmsrq(x86_pmu.lbr_from + i, 0);
+ wrmsrq(x86_pmu.lbr_to + i, 0);
if (x86_pmu.lbr_has_info)
- wrmsrl(x86_pmu.lbr_info + i, 0);
+ wrmsrq(x86_pmu.lbr_info + i, 0);
}
}
static void intel_pmu_arch_lbr_reset(void)
{
/* Write to ARCH_LBR_DEPTH MSR, all LBR entries are reset to 0 */
- wrmsrl(MSR_ARCH_LBR_DEPTH, x86_pmu.lbr_nr);
+ wrmsrq(MSR_ARCH_LBR_DEPTH, x86_pmu.lbr_nr);
}
void intel_pmu_lbr_reset(void)
@@ -199,7 +199,7 @@ void intel_pmu_lbr_reset(void)
cpuc->last_task_ctx = NULL;
cpuc->last_log_id = 0;
if (!static_cpu_has(X86_FEATURE_ARCH_LBR) && cpuc->lbr_select)
- wrmsrl(MSR_LBR_SELECT, 0);
+ wrmsrq(MSR_LBR_SELECT, 0);
}
/*
@@ -209,7 +209,7 @@ static inline u64 intel_pmu_lbr_tos(void)
{
u64 tos;
- rdmsrl(x86_pmu.lbr_tos, tos);
+ rdmsrq(x86_pmu.lbr_tos, tos);
return tos;
}
@@ -282,17 +282,17 @@ static u64 lbr_from_signext_quirk_rd(u64 val)
static __always_inline void wrlbr_from(unsigned int idx, u64 val)
{
val = lbr_from_signext_quirk_wr(val);
- wrmsrl(x86_pmu.lbr_from + idx, val);
+ wrmsrq(x86_pmu.lbr_from + idx, val);
}
static __always_inline void wrlbr_to(unsigned int idx, u64 val)
{
- wrmsrl(x86_pmu.lbr_to + idx, val);
+ wrmsrq(x86_pmu.lbr_to + idx, val);
}
static __always_inline void wrlbr_info(unsigned int idx, u64 val)
{
- wrmsrl(x86_pmu.lbr_info + idx, val);
+ wrmsrq(x86_pmu.lbr_info + idx, val);
}
static __always_inline u64 rdlbr_from(unsigned int idx, struct lbr_entry *lbr)
@@ -302,7 +302,7 @@ static __always_inline u64 rdlbr_from(unsigned int idx, struct lbr_entry *lbr)
if (lbr)
return lbr->from;
- rdmsrl(x86_pmu.lbr_from + idx, val);
+ rdmsrq(x86_pmu.lbr_from + idx, val);
return lbr_from_signext_quirk_rd(val);
}
@@ -314,7 +314,7 @@ static __always_inline u64 rdlbr_to(unsigned int idx, struct lbr_entry *lbr)
if (lbr)
return lbr->to;
- rdmsrl(x86_pmu.lbr_to + idx, val);
+ rdmsrq(x86_pmu.lbr_to + idx, val);
return val;
}
@@ -326,7 +326,7 @@ static __always_inline u64 rdlbr_info(unsigned int idx, struct lbr_entry *lbr)
if (lbr)
return lbr->info;
- rdmsrl(x86_pmu.lbr_info + idx, val);
+ rdmsrq(x86_pmu.lbr_info + idx, val);
return val;
}
@@ -380,10 +380,10 @@ void intel_pmu_lbr_restore(void *ctx)
wrlbr_info(lbr_idx, 0);
}
- wrmsrl(x86_pmu.lbr_tos, tos);
+ wrmsrq(x86_pmu.lbr_tos, tos);
if (cpuc->lbr_select)
- wrmsrl(MSR_LBR_SELECT, task_ctx->lbr_sel);
+ wrmsrq(MSR_LBR_SELECT, task_ctx->lbr_sel);
}
static void intel_pmu_arch_lbr_restore(void *ctx)
@@ -475,7 +475,7 @@ void intel_pmu_lbr_save(void *ctx)
task_ctx->tos = tos;
if (cpuc->lbr_select)
- rdmsrl(MSR_LBR_SELECT, task_ctx->lbr_sel);
+ rdmsrq(MSR_LBR_SELECT, task_ctx->lbr_sel);
}
static void intel_pmu_arch_lbr_save(void *ctx)
@@ -752,7 +752,7 @@ void intel_pmu_lbr_read_32(struct cpu_hw_events *cpuc)
u64 lbr;
} msr_lastbranch;
- rdmsrl(x86_pmu.lbr_from + lbr_idx, msr_lastbranch.lbr);
+ rdmsrq(x86_pmu.lbr_from + lbr_idx, msr_lastbranch.lbr);
perf_clear_branch_entry_bitfields(br);
@@ -1602,7 +1602,7 @@ void __init intel_pmu_arch_lbr_init(void)
goto clear_arch_lbr;
/* Apply the max depth of Arch LBR */
- if (wrmsrl_safe(MSR_ARCH_LBR_DEPTH, lbr_nr))
+ if (wrmsrq_safe(MSR_ARCH_LBR_DEPTH, lbr_nr))
goto clear_arch_lbr;
x86_pmu.lbr_depth_mask = eax.split.lbr_depth_mask;
@@ -1618,7 +1618,7 @@ void __init intel_pmu_arch_lbr_init(void)
x86_pmu.lbr_nr = lbr_nr;
if (!!x86_pmu.lbr_counters)
- x86_pmu.flags |= PMU_FL_BR_CNTR;
+ x86_pmu.flags |= PMU_FL_BR_CNTR | PMU_FL_DYN_CONSTRAINT;
if (x86_pmu.lbr_mispred)
static_branch_enable(&x86_lbr_mispred);
diff --git a/arch/x86/events/intel/p4.c b/arch/x86/events/intel/p4.c
index c85a9fc44355..e5fd7367e45d 100644
--- a/arch/x86/events/intel/p4.c
+++ b/arch/x86/events/intel/p4.c
@@ -13,6 +13,7 @@
#include <asm/cpu_device_id.h>
#include <asm/hardirq.h>
#include <asm/apic.h>
+#include <asm/msr.h>
#include "../perf_event.h"
@@ -859,9 +860,9 @@ static inline int p4_pmu_clear_cccr_ovf(struct hw_perf_event *hwc)
u64 v;
/* an official way for overflow indication */
- rdmsrl(hwc->config_base, v);
+ rdmsrq(hwc->config_base, v);
if (v & P4_CCCR_OVF) {
- wrmsrl(hwc->config_base, v & ~P4_CCCR_OVF);
+ wrmsrq(hwc->config_base, v & ~P4_CCCR_OVF);
return 1;
}
@@ -872,7 +873,7 @@ static inline int p4_pmu_clear_cccr_ovf(struct hw_perf_event *hwc)
* the counter has reached zero value and continued counting before
* real NMI signal was received:
*/
- rdmsrl(hwc->event_base, v);
+ rdmsrq(hwc->event_base, v);
if (!(v & ARCH_P4_UNFLAGGED_BIT))
return 1;
@@ -897,8 +898,8 @@ static void p4_pmu_disable_pebs(void)
* So at moment let leave metrics turned on forever -- it's
* ok for now but need to be revisited!
*
- * (void)wrmsrl_safe(MSR_IA32_PEBS_ENABLE, 0);
- * (void)wrmsrl_safe(MSR_P4_PEBS_MATRIX_VERT, 0);
+ * (void)wrmsrq_safe(MSR_IA32_PEBS_ENABLE, 0);
+ * (void)wrmsrq_safe(MSR_P4_PEBS_MATRIX_VERT, 0);
*/
}
@@ -911,7 +912,7 @@ static inline void p4_pmu_disable_event(struct perf_event *event)
* state we need to clear P4_CCCR_OVF, otherwise interrupt get
* asserted again and again
*/
- (void)wrmsrl_safe(hwc->config_base,
+ (void)wrmsrq_safe(hwc->config_base,
p4_config_unpack_cccr(hwc->config) & ~P4_CCCR_ENABLE & ~P4_CCCR_OVF & ~P4_CCCR_RESERVED);
}
@@ -944,8 +945,8 @@ static void p4_pmu_enable_pebs(u64 config)
bind = &p4_pebs_bind_map[idx];
- (void)wrmsrl_safe(MSR_IA32_PEBS_ENABLE, (u64)bind->metric_pebs);
- (void)wrmsrl_safe(MSR_P4_PEBS_MATRIX_VERT, (u64)bind->metric_vert);
+ (void)wrmsrq_safe(MSR_IA32_PEBS_ENABLE, (u64)bind->metric_pebs);
+ (void)wrmsrq_safe(MSR_P4_PEBS_MATRIX_VERT, (u64)bind->metric_vert);
}
static void __p4_pmu_enable_event(struct perf_event *event)
@@ -979,8 +980,8 @@ static void __p4_pmu_enable_event(struct perf_event *event)
*/
p4_pmu_enable_pebs(hwc->config);
- (void)wrmsrl_safe(escr_addr, escr_conf);
- (void)wrmsrl_safe(hwc->config_base,
+ (void)wrmsrq_safe(escr_addr, escr_conf);
+ (void)wrmsrq_safe(hwc->config_base,
(cccr & ~P4_CCCR_RESERVED) | P4_CCCR_ENABLE);
}
@@ -1024,7 +1025,7 @@ static int p4_pmu_set_period(struct perf_event *event)
*
* the former idea is taken from OProfile code
*/
- wrmsrl(hwc->event_base, (u64)(-left) & x86_pmu.cntval_mask);
+ wrmsrq(hwc->event_base, (u64)(-left) & x86_pmu.cntval_mask);
}
return ret;
@@ -1072,8 +1073,7 @@ static int p4_pmu_handle_irq(struct pt_regs *regs)
continue;
- if (perf_event_overflow(event, &data, regs))
- x86_pmu_stop(event, 0);
+ perf_event_overflow(event, &data, regs);
}
if (handled)
@@ -1398,7 +1398,7 @@ __init int p4_pmu_init(void)
*/
for_each_set_bit(i, x86_pmu.cntr_mask, X86_PMC_IDX_MAX) {
reg = x86_pmu_config_addr(i);
- wrmsrl_safe(reg, 0ULL);
+ wrmsrq_safe(reg, 0ULL);
}
return 0;
diff --git a/arch/x86/events/intel/p6.c b/arch/x86/events/intel/p6.c
index 65b45e9d7016..6e41de355bd8 100644
--- a/arch/x86/events/intel/p6.c
+++ b/arch/x86/events/intel/p6.c
@@ -3,6 +3,7 @@
#include <linux/types.h>
#include <asm/cpu_device_id.h>
+#include <asm/msr.h>
#include "../perf_event.h"
@@ -142,9 +143,9 @@ static void p6_pmu_disable_all(void)
u64 val;
/* p6 only has one enable register */
- rdmsrl(MSR_P6_EVNTSEL0, val);
+ rdmsrq(MSR_P6_EVNTSEL0, val);
val &= ~ARCH_PERFMON_EVENTSEL_ENABLE;
- wrmsrl(MSR_P6_EVNTSEL0, val);
+ wrmsrq(MSR_P6_EVNTSEL0, val);
}
static void p6_pmu_enable_all(int added)
@@ -152,9 +153,9 @@ static void p6_pmu_enable_all(int added)
unsigned long val;
/* p6 only has one enable register */
- rdmsrl(MSR_P6_EVNTSEL0, val);
+ rdmsrq(MSR_P6_EVNTSEL0, val);
val |= ARCH_PERFMON_EVENTSEL_ENABLE;
- wrmsrl(MSR_P6_EVNTSEL0, val);
+ wrmsrq(MSR_P6_EVNTSEL0, val);
}
static inline void
@@ -163,7 +164,7 @@ p6_pmu_disable_event(struct perf_event *event)
struct hw_perf_event *hwc = &event->hw;
u64 val = P6_NOP_EVENT;
- (void)wrmsrl_safe(hwc->config_base, val);
+ (void)wrmsrq_safe(hwc->config_base, val);
}
static void p6_pmu_enable_event(struct perf_event *event)
@@ -180,7 +181,7 @@ static void p6_pmu_enable_event(struct perf_event *event)
* to actually enable the events.
*/
- (void)wrmsrl_safe(hwc->config_base, val);
+ (void)wrmsrq_safe(hwc->config_base, val);
}
PMU_FORMAT_ATTR(event, "config:0-7" );
diff --git a/arch/x86/events/intel/pt.c b/arch/x86/events/intel/pt.c
index fa37565f6418..e8cf29d2b10c 100644
--- a/arch/x86/events/intel/pt.c
+++ b/arch/x86/events/intel/pt.c
@@ -18,12 +18,13 @@
#include <linux/slab.h>
#include <linux/device.h>
-#include <asm/cpuid.h>
+#include <asm/cpuid/api.h>
#include <asm/perf_event.h>
#include <asm/insn.h>
#include <asm/io.h>
#include <asm/intel_pt.h>
#include <asm/cpu_device_id.h>
+#include <asm/msr.h>
#include "../perf_event.h"
#include "pt.h"
@@ -194,7 +195,7 @@ static int __init pt_pmu_hw_init(void)
int ret;
long i;
- rdmsrl(MSR_PLATFORM_INFO, reg);
+ rdmsrq(MSR_PLATFORM_INFO, reg);
pt_pmu.max_nonturbo_ratio = (reg & 0xff00) >> 8;
/*
@@ -230,7 +231,7 @@ static int __init pt_pmu_hw_init(void)
* "IA32_VMX_MISC[bit 14]" being 1 means PT can trace
* post-VMXON.
*/
- rdmsrl(MSR_IA32_VMX_MISC, reg);
+ rdmsrq(MSR_IA32_VMX_MISC, reg);
if (reg & BIT(14))
pt_pmu.vmx = true;
}
@@ -426,7 +427,7 @@ static void pt_config_start(struct perf_event *event)
if (READ_ONCE(pt->vmx_on))
perf_aux_output_flag(&pt->handle, PERF_AUX_FLAG_PARTIAL);
else
- wrmsrl(MSR_IA32_RTIT_CTL, ctl);
+ wrmsrq(MSR_IA32_RTIT_CTL, ctl);
WRITE_ONCE(event->hw.aux_config, ctl);
}
@@ -485,12 +486,12 @@ static u64 pt_config_filters(struct perf_event *event)
/* avoid redundant msr writes */
if (pt->filters.filter[range].msr_a != filter->msr_a) {
- wrmsrl(pt_address_ranges[range].msr_a, filter->msr_a);
+ wrmsrq(pt_address_ranges[range].msr_a, filter->msr_a);
pt->filters.filter[range].msr_a = filter->msr_a;
}
if (pt->filters.filter[range].msr_b != filter->msr_b) {
- wrmsrl(pt_address_ranges[range].msr_b, filter->msr_b);
+ wrmsrq(pt_address_ranges[range].msr_b, filter->msr_b);
pt->filters.filter[range].msr_b = filter->msr_b;
}
@@ -509,7 +510,7 @@ static void pt_config(struct perf_event *event)
/* First round: clear STATUS, in particular the PSB byte counter. */
if (!event->hw.aux_config) {
perf_event_itrace_started(event);
- wrmsrl(MSR_IA32_RTIT_STATUS, 0);
+ wrmsrq(MSR_IA32_RTIT_STATUS, 0);
}
reg = pt_config_filters(event);
@@ -569,7 +570,7 @@ static void pt_config_stop(struct perf_event *event)
ctl &= ~RTIT_CTL_TRACEEN;
if (!READ_ONCE(pt->vmx_on))
- wrmsrl(MSR_IA32_RTIT_CTL, ctl);
+ wrmsrq(MSR_IA32_RTIT_CTL, ctl);
WRITE_ONCE(event->hw.aux_config, ctl);
@@ -658,13 +659,13 @@ static void pt_config_buffer(struct pt_buffer *buf)
reg = virt_to_phys(base);
if (pt->output_base != reg) {
pt->output_base = reg;
- wrmsrl(MSR_IA32_RTIT_OUTPUT_BASE, reg);
+ wrmsrq(MSR_IA32_RTIT_OUTPUT_BASE, reg);
}
reg = 0x7f | (mask << 7) | ((u64)buf->output_off << 32);
if (pt->output_mask != reg) {
pt->output_mask = reg;
- wrmsrl(MSR_IA32_RTIT_OUTPUT_MASK, reg);
+ wrmsrq(MSR_IA32_RTIT_OUTPUT_MASK, reg);
}
}
@@ -926,7 +927,7 @@ static void pt_handle_status(struct pt *pt)
int advance = 0;
u64 status;
- rdmsrl(MSR_IA32_RTIT_STATUS, status);
+ rdmsrq(MSR_IA32_RTIT_STATUS, status);
if (status & RTIT_STATUS_ERROR) {
pr_err_ratelimited("ToPA ERROR encountered, trying to recover\n");
@@ -970,7 +971,7 @@ static void pt_handle_status(struct pt *pt)
if (advance)
pt_buffer_advance(buf);
- wrmsrl(MSR_IA32_RTIT_STATUS, status);
+ wrmsrq(MSR_IA32_RTIT_STATUS, status);
}
/**
@@ -985,12 +986,12 @@ static void pt_read_offset(struct pt_buffer *buf)
struct topa_page *tp;
if (!buf->single) {
- rdmsrl(MSR_IA32_RTIT_OUTPUT_BASE, pt->output_base);
+ rdmsrq(MSR_IA32_RTIT_OUTPUT_BASE, pt->output_base);
tp = phys_to_virt(pt->output_base);
buf->cur = &tp->topa;
}
- rdmsrl(MSR_IA32_RTIT_OUTPUT_MASK, pt->output_mask);
+ rdmsrq(MSR_IA32_RTIT_OUTPUT_MASK, pt->output_mask);
/* offset within current output region */
buf->output_off = pt->output_mask >> 32;
/* index of current output region within this table */
@@ -1585,7 +1586,7 @@ void intel_pt_handle_vmx(int on)
/* Turn PTs back on */
if (!on && event)
- wrmsrl(MSR_IA32_RTIT_CTL, event->hw.aux_config);
+ wrmsrq(MSR_IA32_RTIT_CTL, event->hw.aux_config);
local_irq_restore(flags);
}
@@ -1611,7 +1612,7 @@ static void pt_event_start(struct perf_event *event, int mode)
* PMI might have just cleared these, so resume_allowed
* must be checked again also.
*/
- rdmsrl(MSR_IA32_RTIT_STATUS, status);
+ rdmsrq(MSR_IA32_RTIT_STATUS, status);
if (!(status & (RTIT_STATUS_TRIGGEREN |
RTIT_STATUS_ERROR |
RTIT_STATUS_STOPPED)) &&
@@ -1839,7 +1840,7 @@ static __init int pt_init(void)
for_each_online_cpu(cpu) {
u64 ctl;
- ret = rdmsrl_safe_on_cpu(cpu, MSR_IA32_RTIT_CTL, &ctl);
+ ret = rdmsrq_safe_on_cpu(cpu, MSR_IA32_RTIT_CTL, &ctl);
if (!ret && (ctl & RTIT_CTL_TRACEEN))
prior_warn++;
}
@@ -1863,6 +1864,8 @@ static __init int pt_init(void)
if (!intel_pt_validate_hw_cap(PT_CAP_topa_multiple_entries))
pt_pmu.pmu.capabilities = PERF_PMU_CAP_AUX_NO_SG;
+ else
+ pt_pmu.pmu.capabilities = PERF_PMU_CAP_AUX_PREFER_LARGE;
pt_pmu.pmu.capabilities |= PERF_PMU_CAP_EXCLUSIVE |
PERF_PMU_CAP_ITRACE |
diff --git a/arch/x86/events/intel/uncore.c b/arch/x86/events/intel/uncore.c
index a34e50fc4a8f..e0815a12db90 100644
--- a/arch/x86/events/intel/uncore.c
+++ b/arch/x86/events/intel/uncore.c
@@ -3,6 +3,7 @@
#include <asm/cpu_device_id.h>
#include <asm/intel-family.h>
+#include <asm/msr.h>
#include "uncore.h"
#include "uncore_discovery.h"
@@ -150,7 +151,7 @@ u64 uncore_msr_read_counter(struct intel_uncore_box *box, struct perf_event *eve
{
u64 count;
- rdmsrl(event->hw.event_base, count);
+ rdmsrq(event->hw.event_base, count);
return count;
}
@@ -305,17 +306,11 @@ static enum hrtimer_restart uncore_pmu_hrtimer(struct hrtimer *hrtimer)
{
struct intel_uncore_box *box;
struct perf_event *event;
- unsigned long flags;
int bit;
box = container_of(hrtimer, struct intel_uncore_box, hrtimer);
if (!box->n_active || box->cpu != smp_processor_id())
return HRTIMER_NORESTART;
- /*
- * disable local interrupt to prevent uncore_pmu_event_start/stop
- * to interrupt the update process
- */
- local_irq_save(flags);
/*
* handle boxes with an active event list as opposed to active
@@ -328,8 +323,6 @@ static enum hrtimer_restart uncore_pmu_hrtimer(struct hrtimer *hrtimer)
for_each_set_bit(bit, box->active_mask, UNCORE_PMC_IDX_MAX)
uncore_perf_event_update(box, box->events[bit]);
- local_irq_restore(flags);
-
hrtimer_forward_now(hrtimer, ns_to_ktime(box->hrtimer_duration));
return HRTIMER_RESTART;
}
@@ -337,7 +330,7 @@ static enum hrtimer_restart uncore_pmu_hrtimer(struct hrtimer *hrtimer)
void uncore_pmu_start_hrtimer(struct intel_uncore_box *box)
{
hrtimer_start(&box->hrtimer, ns_to_ktime(box->hrtimer_duration),
- HRTIMER_MODE_REL_PINNED);
+ HRTIMER_MODE_REL_PINNED_HARD);
}
void uncore_pmu_cancel_hrtimer(struct intel_uncore_box *box)
@@ -347,7 +340,7 @@ void uncore_pmu_cancel_hrtimer(struct intel_uncore_box *box)
static void uncore_pmu_init_hrtimer(struct intel_uncore_box *box)
{
- hrtimer_setup(&box->hrtimer, uncore_pmu_hrtimer, CLOCK_MONOTONIC, HRTIMER_MODE_REL);
+ hrtimer_setup(&box->hrtimer, uncore_pmu_hrtimer, CLOCK_MONOTONIC, HRTIMER_MODE_REL_HARD);
}
static struct intel_uncore_box *uncore_alloc_box(struct intel_uncore_type *type,
diff --git a/arch/x86/events/intel/uncore_discovery.c b/arch/x86/events/intel/uncore_discovery.c
index 571e44b49691..18a3022f26a0 100644
--- a/arch/x86/events/intel/uncore_discovery.c
+++ b/arch/x86/events/intel/uncore_discovery.c
@@ -5,6 +5,7 @@
*/
#define pr_fmt(fmt) KBUILD_MODNAME ": " fmt
+#include <asm/msr.h>
#include "uncore.h"
#include "uncore_discovery.h"
@@ -441,17 +442,17 @@ static u64 intel_generic_uncore_box_ctl(struct intel_uncore_box *box)
void intel_generic_uncore_msr_init_box(struct intel_uncore_box *box)
{
- wrmsrl(intel_generic_uncore_box_ctl(box), GENERIC_PMON_BOX_CTL_INT);
+ wrmsrq(intel_generic_uncore_box_ctl(box), GENERIC_PMON_BOX_CTL_INT);
}
void intel_generic_uncore_msr_disable_box(struct intel_uncore_box *box)
{
- wrmsrl(intel_generic_uncore_box_ctl(box), GENERIC_PMON_BOX_CTL_FRZ);
+ wrmsrq(intel_generic_uncore_box_ctl(box), GENERIC_PMON_BOX_CTL_FRZ);
}
void intel_generic_uncore_msr_enable_box(struct intel_uncore_box *box)
{
- wrmsrl(intel_generic_uncore_box_ctl(box), 0);
+ wrmsrq(intel_generic_uncore_box_ctl(box), 0);
}
static void intel_generic_uncore_msr_enable_event(struct intel_uncore_box *box,
@@ -459,7 +460,7 @@ static void intel_generic_uncore_msr_enable_event(struct intel_uncore_box *box,
{
struct hw_perf_event *hwc = &event->hw;
- wrmsrl(hwc->config_base, hwc->config);
+ wrmsrq(hwc->config_base, hwc->config);
}
static void intel_generic_uncore_msr_disable_event(struct intel_uncore_box *box,
@@ -467,7 +468,7 @@ static void intel_generic_uncore_msr_disable_event(struct intel_uncore_box *box,
{
struct hw_perf_event *hwc = &event->hw;
- wrmsrl(hwc->config_base, 0);
+ wrmsrq(hwc->config_base, 0);
}
static struct intel_uncore_ops generic_uncore_msr_ops = {
diff --git a/arch/x86/events/intel/uncore_nhmex.c b/arch/x86/events/intel/uncore_nhmex.c
index 466833478e81..8962e7cb21e3 100644
--- a/arch/x86/events/intel/uncore_nhmex.c
+++ b/arch/x86/events/intel/uncore_nhmex.c
@@ -1,6 +1,7 @@
// SPDX-License-Identifier: GPL-2.0
/* Nehalem-EX/Westmere-EX uncore support */
#include <asm/cpu_device_id.h>
+#include <asm/msr.h>
#include "uncore.h"
/* NHM-EX event control */
@@ -200,12 +201,12 @@ DEFINE_UNCORE_FORMAT_ATTR(mask, mask, "config2:0-63");
static void nhmex_uncore_msr_init_box(struct intel_uncore_box *box)
{
- wrmsrl(NHMEX_U_MSR_PMON_GLOBAL_CTL, NHMEX_U_PMON_GLOBAL_EN_ALL);
+ wrmsrq(NHMEX_U_MSR_PMON_GLOBAL_CTL, NHMEX_U_PMON_GLOBAL_EN_ALL);
}
static void nhmex_uncore_msr_exit_box(struct intel_uncore_box *box)
{
- wrmsrl(NHMEX_U_MSR_PMON_GLOBAL_CTL, 0);
+ wrmsrq(NHMEX_U_MSR_PMON_GLOBAL_CTL, 0);
}
static void nhmex_uncore_msr_disable_box(struct intel_uncore_box *box)
@@ -214,12 +215,12 @@ static void nhmex_uncore_msr_disable_box(struct intel_uncore_box *box)
u64 config;
if (msr) {
- rdmsrl(msr, config);
+ rdmsrq(msr, config);
config &= ~((1ULL << uncore_num_counters(box)) - 1);
/* WBox has a fixed counter */
if (uncore_msr_fixed_ctl(box))
config &= ~NHMEX_W_PMON_GLOBAL_FIXED_EN;
- wrmsrl(msr, config);
+ wrmsrq(msr, config);
}
}
@@ -229,18 +230,18 @@ static void nhmex_uncore_msr_enable_box(struct intel_uncore_box *box)
u64 config;
if (msr) {
- rdmsrl(msr, config);
+ rdmsrq(msr, config);
config |= (1ULL << uncore_num_counters(box)) - 1;
/* WBox has a fixed counter */
if (uncore_msr_fixed_ctl(box))
config |= NHMEX_W_PMON_GLOBAL_FIXED_EN;
- wrmsrl(msr, config);
+ wrmsrq(msr, config);
}
}
static void nhmex_uncore_msr_disable_event(struct intel_uncore_box *box, struct perf_event *event)
{
- wrmsrl(event->hw.config_base, 0);
+ wrmsrq(event->hw.config_base, 0);
}
static void nhmex_uncore_msr_enable_event(struct intel_uncore_box *box, struct perf_event *event)
@@ -248,11 +249,11 @@ static void nhmex_uncore_msr_enable_event(struct intel_uncore_box *box, struct p
struct hw_perf_event *hwc = &event->hw;
if (hwc->idx == UNCORE_PMC_IDX_FIXED)
- wrmsrl(hwc->config_base, NHMEX_PMON_CTL_EN_BIT0);
+ wrmsrq(hwc->config_base, NHMEX_PMON_CTL_EN_BIT0);
else if (box->pmu->type->event_mask & NHMEX_PMON_CTL_EN_BIT0)
- wrmsrl(hwc->config_base, hwc->config | NHMEX_PMON_CTL_EN_BIT22);
+ wrmsrq(hwc->config_base, hwc->config | NHMEX_PMON_CTL_EN_BIT22);
else
- wrmsrl(hwc->config_base, hwc->config | NHMEX_PMON_CTL_EN_BIT0);
+ wrmsrq(hwc->config_base, hwc->config | NHMEX_PMON_CTL_EN_BIT0);
}
#define NHMEX_UNCORE_OPS_COMMON_INIT() \
@@ -382,10 +383,10 @@ static void nhmex_bbox_msr_enable_event(struct intel_uncore_box *box, struct per
struct hw_perf_event_extra *reg2 = &hwc->branch_reg;
if (reg1->idx != EXTRA_REG_NONE) {
- wrmsrl(reg1->reg, reg1->config);
- wrmsrl(reg1->reg + 1, reg2->config);
+ wrmsrq(reg1->reg, reg1->config);
+ wrmsrq(reg1->reg + 1, reg2->config);
}
- wrmsrl(hwc->config_base, NHMEX_PMON_CTL_EN_BIT0 |
+ wrmsrq(hwc->config_base, NHMEX_PMON_CTL_EN_BIT0 |
(hwc->config & NHMEX_B_PMON_CTL_EV_SEL_MASK));
}
@@ -467,12 +468,12 @@ static void nhmex_sbox_msr_enable_event(struct intel_uncore_box *box, struct per
struct hw_perf_event_extra *reg2 = &hwc->branch_reg;
if (reg1->idx != EXTRA_REG_NONE) {
- wrmsrl(reg1->reg, 0);
- wrmsrl(reg1->reg + 1, reg1->config);
- wrmsrl(reg1->reg + 2, reg2->config);
- wrmsrl(reg1->reg, NHMEX_S_PMON_MM_CFG_EN);
+ wrmsrq(reg1->reg, 0);
+ wrmsrq(reg1->reg + 1, reg1->config);
+ wrmsrq(reg1->reg + 2, reg2->config);
+ wrmsrq(reg1->reg, NHMEX_S_PMON_MM_CFG_EN);
}
- wrmsrl(hwc->config_base, hwc->config | NHMEX_PMON_CTL_EN_BIT22);
+ wrmsrq(hwc->config_base, hwc->config | NHMEX_PMON_CTL_EN_BIT22);
}
static struct attribute *nhmex_uncore_sbox_formats_attr[] = {
@@ -842,25 +843,25 @@ static void nhmex_mbox_msr_enable_event(struct intel_uncore_box *box, struct per
idx = __BITS_VALUE(reg1->idx, 0, 8);
if (idx != 0xff)
- wrmsrl(__BITS_VALUE(reg1->reg, 0, 16),
+ wrmsrq(__BITS_VALUE(reg1->reg, 0, 16),
nhmex_mbox_shared_reg_config(box, idx));
idx = __BITS_VALUE(reg1->idx, 1, 8);
if (idx != 0xff)
- wrmsrl(__BITS_VALUE(reg1->reg, 1, 16),
+ wrmsrq(__BITS_VALUE(reg1->reg, 1, 16),
nhmex_mbox_shared_reg_config(box, idx));
if (reg2->idx != EXTRA_REG_NONE) {
- wrmsrl(reg2->reg, 0);
+ wrmsrq(reg2->reg, 0);
if (reg2->config != ~0ULL) {
- wrmsrl(reg2->reg + 1,
+ wrmsrq(reg2->reg + 1,
reg2->config & NHMEX_M_PMON_ADDR_MATCH_MASK);
- wrmsrl(reg2->reg + 2, NHMEX_M_PMON_ADDR_MASK_MASK &
+ wrmsrq(reg2->reg + 2, NHMEX_M_PMON_ADDR_MASK_MASK &
(reg2->config >> NHMEX_M_PMON_ADDR_MASK_SHIFT));
- wrmsrl(reg2->reg, NHMEX_M_PMON_MM_CFG_EN);
+ wrmsrq(reg2->reg, NHMEX_M_PMON_MM_CFG_EN);
}
}
- wrmsrl(hwc->config_base, hwc->config | NHMEX_PMON_CTL_EN_BIT0);
+ wrmsrq(hwc->config_base, hwc->config | NHMEX_PMON_CTL_EN_BIT0);
}
DEFINE_UNCORE_FORMAT_ATTR(count_mode, count_mode, "config:2-3");
@@ -1121,31 +1122,31 @@ static void nhmex_rbox_msr_enable_event(struct intel_uncore_box *box, struct per
switch (idx % 6) {
case 0:
- wrmsrl(NHMEX_R_MSR_PORTN_IPERF_CFG0(port), reg1->config);
+ wrmsrq(NHMEX_R_MSR_PORTN_IPERF_CFG0(port), reg1->config);
break;
case 1:
- wrmsrl(NHMEX_R_MSR_PORTN_IPERF_CFG1(port), reg1->config);
+ wrmsrq(NHMEX_R_MSR_PORTN_IPERF_CFG1(port), reg1->config);
break;
case 2:
case 3:
- wrmsrl(NHMEX_R_MSR_PORTN_QLX_CFG(port),
+ wrmsrq(NHMEX_R_MSR_PORTN_QLX_CFG(port),
uncore_shared_reg_config(box, 2 + (idx / 6) * 5));
break;
case 4:
- wrmsrl(NHMEX_R_MSR_PORTN_XBR_SET1_MM_CFG(port),
+ wrmsrq(NHMEX_R_MSR_PORTN_XBR_SET1_MM_CFG(port),
hwc->config >> 32);
- wrmsrl(NHMEX_R_MSR_PORTN_XBR_SET1_MATCH(port), reg1->config);
- wrmsrl(NHMEX_R_MSR_PORTN_XBR_SET1_MASK(port), reg2->config);
+ wrmsrq(NHMEX_R_MSR_PORTN_XBR_SET1_MATCH(port), reg1->config);
+ wrmsrq(NHMEX_R_MSR_PORTN_XBR_SET1_MASK(port), reg2->config);
break;
case 5:
- wrmsrl(NHMEX_R_MSR_PORTN_XBR_SET2_MM_CFG(port),
+ wrmsrq(NHMEX_R_MSR_PORTN_XBR_SET2_MM_CFG(port),
hwc->config >> 32);
- wrmsrl(NHMEX_R_MSR_PORTN_XBR_SET2_MATCH(port), reg1->config);
- wrmsrl(NHMEX_R_MSR_PORTN_XBR_SET2_MASK(port), reg2->config);
+ wrmsrq(NHMEX_R_MSR_PORTN_XBR_SET2_MATCH(port), reg1->config);
+ wrmsrq(NHMEX_R_MSR_PORTN_XBR_SET2_MASK(port), reg2->config);
break;
}
- wrmsrl(hwc->config_base, NHMEX_PMON_CTL_EN_BIT0 |
+ wrmsrq(hwc->config_base, NHMEX_PMON_CTL_EN_BIT0 |
(hwc->config & NHMEX_R_PMON_CTL_EV_SEL_MASK));
}
diff --git a/arch/x86/events/intel/uncore_snb.c b/arch/x86/events/intel/uncore_snb.c
index edb7fd50efe0..a1a96833e30e 100644
--- a/arch/x86/events/intel/uncore_snb.c
+++ b/arch/x86/events/intel/uncore_snb.c
@@ -1,5 +1,6 @@
// SPDX-License-Identifier: GPL-2.0
/* Nehalem/SandBridge/Haswell/Broadwell/Skylake uncore support */
+#include <asm/msr.h>
#include "uncore.h"
#include "uncore_discovery.h"
@@ -260,34 +261,34 @@ static void snb_uncore_msr_enable_event(struct intel_uncore_box *box, struct per
struct hw_perf_event *hwc = &event->hw;
if (hwc->idx < UNCORE_PMC_IDX_FIXED)
- wrmsrl(hwc->config_base, hwc->config | SNB_UNC_CTL_EN);
+ wrmsrq(hwc->config_base, hwc->config | SNB_UNC_CTL_EN);
else
- wrmsrl(hwc->config_base, SNB_UNC_CTL_EN);
+ wrmsrq(hwc->config_base, SNB_UNC_CTL_EN);
}
static void snb_uncore_msr_disable_event(struct intel_uncore_box *box, struct perf_event *event)
{
- wrmsrl(event->hw.config_base, 0);
+ wrmsrq(event->hw.config_base, 0);
}
static void snb_uncore_msr_init_box(struct intel_uncore_box *box)
{
if (box->pmu->pmu_idx == 0) {
- wrmsrl(SNB_UNC_PERF_GLOBAL_CTL,
+ wrmsrq(SNB_UNC_PERF_GLOBAL_CTL,
SNB_UNC_GLOBAL_CTL_EN | SNB_UNC_GLOBAL_CTL_CORE_ALL);
}
}
static void snb_uncore_msr_enable_box(struct intel_uncore_box *box)
{
- wrmsrl(SNB_UNC_PERF_GLOBAL_CTL,
+ wrmsrq(SNB_UNC_PERF_GLOBAL_CTL,
SNB_UNC_GLOBAL_CTL_EN | SNB_UNC_GLOBAL_CTL_CORE_ALL);
}
static void snb_uncore_msr_exit_box(struct intel_uncore_box *box)
{
if (box->pmu->pmu_idx == 0)
- wrmsrl(SNB_UNC_PERF_GLOBAL_CTL, 0);
+ wrmsrq(SNB_UNC_PERF_GLOBAL_CTL, 0);
}
static struct uncore_event_desc snb_uncore_events[] = {
@@ -372,7 +373,7 @@ void snb_uncore_cpu_init(void)
static void skl_uncore_msr_init_box(struct intel_uncore_box *box)
{
if (box->pmu->pmu_idx == 0) {
- wrmsrl(SKL_UNC_PERF_GLOBAL_CTL,
+ wrmsrq(SKL_UNC_PERF_GLOBAL_CTL,
SNB_UNC_GLOBAL_CTL_EN | SKL_UNC_GLOBAL_CTL_CORE_ALL);
}
@@ -383,14 +384,14 @@ static void skl_uncore_msr_init_box(struct intel_uncore_box *box)
static void skl_uncore_msr_enable_box(struct intel_uncore_box *box)
{
- wrmsrl(SKL_UNC_PERF_GLOBAL_CTL,
+ wrmsrq(SKL_UNC_PERF_GLOBAL_CTL,
SNB_UNC_GLOBAL_CTL_EN | SKL_UNC_GLOBAL_CTL_CORE_ALL);
}
static void skl_uncore_msr_exit_box(struct intel_uncore_box *box)
{
if (box->pmu->pmu_idx == 0)
- wrmsrl(SKL_UNC_PERF_GLOBAL_CTL, 0);
+ wrmsrq(SKL_UNC_PERF_GLOBAL_CTL, 0);
}
static struct intel_uncore_ops skl_uncore_msr_ops = {
@@ -504,7 +505,7 @@ static int icl_get_cbox_num(void)
{
u64 num_boxes;
- rdmsrl(ICL_UNC_CBO_CONFIG, num_boxes);
+ rdmsrq(ICL_UNC_CBO_CONFIG, num_boxes);
return num_boxes & ICL_UNC_NUM_CBO_MASK;
}
@@ -525,7 +526,7 @@ static struct intel_uncore_type *tgl_msr_uncores[] = {
static void rkl_uncore_msr_init_box(struct intel_uncore_box *box)
{
if (box->pmu->pmu_idx == 0)
- wrmsrl(SKL_UNC_PERF_GLOBAL_CTL, SNB_UNC_GLOBAL_CTL_EN);
+ wrmsrq(SKL_UNC_PERF_GLOBAL_CTL, SNB_UNC_GLOBAL_CTL_EN);
}
void tgl_uncore_cpu_init(void)
@@ -541,24 +542,24 @@ void tgl_uncore_cpu_init(void)
static void adl_uncore_msr_init_box(struct intel_uncore_box *box)
{
if (box->pmu->pmu_idx == 0)
- wrmsrl(ADL_UNC_PERF_GLOBAL_CTL, SNB_UNC_GLOBAL_CTL_EN);
+ wrmsrq(ADL_UNC_PERF_GLOBAL_CTL, SNB_UNC_GLOBAL_CTL_EN);
}
static void adl_uncore_msr_enable_box(struct intel_uncore_box *box)
{
- wrmsrl(ADL_UNC_PERF_GLOBAL_CTL, SNB_UNC_GLOBAL_CTL_EN);
+ wrmsrq(ADL_UNC_PERF_GLOBAL_CTL, SNB_UNC_GLOBAL_CTL_EN);
}
static void adl_uncore_msr_disable_box(struct intel_uncore_box *box)
{
if (box->pmu->pmu_idx == 0)
- wrmsrl(ADL_UNC_PERF_GLOBAL_CTL, 0);
+ wrmsrq(ADL_UNC_PERF_GLOBAL_CTL, 0);
}
static void adl_uncore_msr_exit_box(struct intel_uncore_box *box)
{
if (box->pmu->pmu_idx == 0)
- wrmsrl(ADL_UNC_PERF_GLOBAL_CTL, 0);
+ wrmsrq(ADL_UNC_PERF_GLOBAL_CTL, 0);
}
static struct intel_uncore_ops adl_uncore_msr_ops = {
@@ -691,7 +692,7 @@ static struct intel_uncore_type mtl_uncore_hac_cbox = {
static void mtl_uncore_msr_init_box(struct intel_uncore_box *box)
{
- wrmsrl(uncore_msr_box_ctl(box), SNB_UNC_GLOBAL_CTL_EN);
+ wrmsrq(uncore_msr_box_ctl(box), SNB_UNC_GLOBAL_CTL_EN);
}
static struct intel_uncore_ops mtl_uncore_msr_ops = {
@@ -758,7 +759,7 @@ static struct intel_uncore_type *lnl_msr_uncores[] = {
static void lnl_uncore_msr_init_box(struct intel_uncore_box *box)
{
if (box->pmu->pmu_idx == 0)
- wrmsrl(LNL_UNC_MSR_GLOBAL_CTL, SNB_UNC_GLOBAL_CTL_EN);
+ wrmsrq(LNL_UNC_MSR_GLOBAL_CTL, SNB_UNC_GLOBAL_CTL_EN);
}
static struct intel_uncore_ops lnl_uncore_msr_ops = {
@@ -1306,12 +1307,12 @@ int skl_uncore_pci_init(void)
/* Nehalem uncore support */
static void nhm_uncore_msr_disable_box(struct intel_uncore_box *box)
{
- wrmsrl(NHM_UNC_PERF_GLOBAL_CTL, 0);
+ wrmsrq(NHM_UNC_PERF_GLOBAL_CTL, 0);
}
static void nhm_uncore_msr_enable_box(struct intel_uncore_box *box)
{
- wrmsrl(NHM_UNC_PERF_GLOBAL_CTL, NHM_UNC_GLOBAL_CTL_EN_PC_ALL | NHM_UNC_GLOBAL_CTL_EN_FC);
+ wrmsrq(NHM_UNC_PERF_GLOBAL_CTL, NHM_UNC_GLOBAL_CTL_EN_PC_ALL | NHM_UNC_GLOBAL_CTL_EN_FC);
}
static void nhm_uncore_msr_enable_event(struct intel_uncore_box *box, struct perf_event *event)
@@ -1319,9 +1320,9 @@ static void nhm_uncore_msr_enable_event(struct intel_uncore_box *box, struct per
struct hw_perf_event *hwc = &event->hw;
if (hwc->idx < UNCORE_PMC_IDX_FIXED)
- wrmsrl(hwc->config_base, hwc->config | SNB_UNC_CTL_EN);
+ wrmsrq(hwc->config_base, hwc->config | SNB_UNC_CTL_EN);
else
- wrmsrl(hwc->config_base, NHM_UNC_FIXED_CTR_CTL_EN);
+ wrmsrq(hwc->config_base, NHM_UNC_FIXED_CTR_CTL_EN);
}
static struct attribute *nhm_uncore_formats_attr[] = {
diff --git a/arch/x86/events/intel/uncore_snbep.c b/arch/x86/events/intel/uncore_snbep.c
index 76d96df1475a..2824dc9950be 100644
--- a/arch/x86/events/intel/uncore_snbep.c
+++ b/arch/x86/events/intel/uncore_snbep.c
@@ -1,6 +1,7 @@
// SPDX-License-Identifier: GPL-2.0
/* SandyBridge-EP/IvyTown uncore support */
#include <asm/cpu_device_id.h>
+#include <asm/msr.h>
#include "uncore.h"
#include "uncore_discovery.h"
@@ -618,9 +619,9 @@ static void snbep_uncore_msr_disable_box(struct intel_uncore_box *box)
msr = uncore_msr_box_ctl(box);
if (msr) {
- rdmsrl(msr, config);
+ rdmsrq(msr, config);
config |= SNBEP_PMON_BOX_CTL_FRZ;
- wrmsrl(msr, config);
+ wrmsrq(msr, config);
}
}
@@ -631,9 +632,9 @@ static void snbep_uncore_msr_enable_box(struct intel_uncore_box *box)
msr = uncore_msr_box_ctl(box);
if (msr) {
- rdmsrl(msr, config);
+ rdmsrq(msr, config);
config &= ~SNBEP_PMON_BOX_CTL_FRZ;
- wrmsrl(msr, config);
+ wrmsrq(msr, config);
}
}
@@ -643,9 +644,9 @@ static void snbep_uncore_msr_enable_event(struct intel_uncore_box *box, struct p
struct hw_perf_event_extra *reg1 = &hwc->extra_reg;
if (reg1->idx != EXTRA_REG_NONE)
- wrmsrl(reg1->reg, uncore_shared_reg_config(box, 0));
+ wrmsrq(reg1->reg, uncore_shared_reg_config(box, 0));
- wrmsrl(hwc->config_base, hwc->config | SNBEP_PMON_CTL_EN);
+ wrmsrq(hwc->config_base, hwc->config | SNBEP_PMON_CTL_EN);
}
static void snbep_uncore_msr_disable_event(struct intel_uncore_box *box,
@@ -653,7 +654,7 @@ static void snbep_uncore_msr_disable_event(struct intel_uncore_box *box,
{
struct hw_perf_event *hwc = &event->hw;
- wrmsrl(hwc->config_base, hwc->config);
+ wrmsrq(hwc->config_base, hwc->config);
}
static void snbep_uncore_msr_init_box(struct intel_uncore_box *box)
@@ -661,7 +662,7 @@ static void snbep_uncore_msr_init_box(struct intel_uncore_box *box)
unsigned msr = uncore_msr_box_ctl(box);
if (msr)
- wrmsrl(msr, SNBEP_PMON_BOX_CTL_INT);
+ wrmsrq(msr, SNBEP_PMON_BOX_CTL_INT);
}
static struct attribute *snbep_uncore_formats_attr[] = {
@@ -1532,7 +1533,7 @@ static void ivbep_uncore_msr_init_box(struct intel_uncore_box *box)
{
unsigned msr = uncore_msr_box_ctl(box);
if (msr)
- wrmsrl(msr, IVBEP_PMON_BOX_CTL_INT);
+ wrmsrq(msr, IVBEP_PMON_BOX_CTL_INT);
}
static void ivbep_uncore_pci_init_box(struct intel_uncore_box *box)
@@ -1783,11 +1784,11 @@ static void ivbep_cbox_enable_event(struct intel_uncore_box *box, struct perf_ev
if (reg1->idx != EXTRA_REG_NONE) {
u64 filter = uncore_shared_reg_config(box, 0);
- wrmsrl(reg1->reg, filter & 0xffffffff);
- wrmsrl(reg1->reg + 6, filter >> 32);
+ wrmsrq(reg1->reg, filter & 0xffffffff);
+ wrmsrq(reg1->reg + 6, filter >> 32);
}
- wrmsrl(hwc->config_base, hwc->config | SNBEP_PMON_CTL_EN);
+ wrmsrq(hwc->config_base, hwc->config | SNBEP_PMON_CTL_EN);
}
static struct intel_uncore_ops ivbep_uncore_cbox_ops = {
@@ -2767,11 +2768,11 @@ static void hswep_cbox_enable_event(struct intel_uncore_box *box,
if (reg1->idx != EXTRA_REG_NONE) {
u64 filter = uncore_shared_reg_config(box, 0);
- wrmsrl(reg1->reg, filter & 0xffffffff);
- wrmsrl(reg1->reg + 1, filter >> 32);
+ wrmsrq(reg1->reg, filter & 0xffffffff);
+ wrmsrq(reg1->reg + 1, filter >> 32);
}
- wrmsrl(hwc->config_base, hwc->config | SNBEP_PMON_CTL_EN);
+ wrmsrq(hwc->config_base, hwc->config | SNBEP_PMON_CTL_EN);
}
static struct intel_uncore_ops hswep_uncore_cbox_ops = {
@@ -2816,7 +2817,7 @@ static void hswep_uncore_sbox_msr_init_box(struct intel_uncore_box *box)
for_each_set_bit(i, (unsigned long *)&init, 64) {
flags |= (1ULL << i);
- wrmsrl(msr, flags);
+ wrmsrq(msr, flags);
}
}
}
@@ -3708,7 +3709,7 @@ static void skx_iio_enable_event(struct intel_uncore_box *box,
{
struct hw_perf_event *hwc = &event->hw;
- wrmsrl(hwc->config_base, hwc->config | SNBEP_PMON_CTL_EN);
+ wrmsrq(hwc->config_base, hwc->config | SNBEP_PMON_CTL_EN);
}
static struct intel_uncore_ops skx_uncore_iio_ops = {
@@ -3765,7 +3766,7 @@ static int skx_msr_cpu_bus_read(int cpu, u64 *topology)
{
u64 msr_value;
- if (rdmsrl_on_cpu(cpu, SKX_MSR_CPU_BUS_NUMBER, &msr_value) ||
+ if (rdmsrq_on_cpu(cpu, SKX_MSR_CPU_BUS_NUMBER, &msr_value) ||
!(msr_value & SKX_MSR_CPU_BUS_VALID_BIT))
return -ENXIO;
@@ -4655,9 +4656,9 @@ static void snr_cha_enable_event(struct intel_uncore_box *box,
struct hw_perf_event_extra *reg1 = &hwc->extra_reg;
if (reg1->idx != EXTRA_REG_NONE)
- wrmsrl(reg1->reg, reg1->config);
+ wrmsrq(reg1->reg, reg1->config);
- wrmsrl(hwc->config_base, hwc->config | SNBEP_PMON_CTL_EN);
+ wrmsrq(hwc->config_base, hwc->config | SNBEP_PMON_CTL_EN);
}
static struct intel_uncore_ops snr_uncore_chabox_ops = {
@@ -5882,9 +5883,9 @@ static void spr_uncore_msr_enable_event(struct intel_uncore_box *box,
struct hw_perf_event_extra *reg1 = &hwc->extra_reg;
if (reg1->idx != EXTRA_REG_NONE)
- wrmsrl(reg1->reg, reg1->config);
+ wrmsrq(reg1->reg, reg1->config);
- wrmsrl(hwc->config_base, hwc->config);
+ wrmsrq(hwc->config_base, hwc->config);
}
static void spr_uncore_msr_disable_event(struct intel_uncore_box *box,
@@ -5894,9 +5895,9 @@ static void spr_uncore_msr_disable_event(struct intel_uncore_box *box,
struct hw_perf_event_extra *reg1 = &hwc->extra_reg;
if (reg1->idx != EXTRA_REG_NONE)
- wrmsrl(reg1->reg, 0);
+ wrmsrq(reg1->reg, 0);
- wrmsrl(hwc->config_base, 0);
+ wrmsrq(hwc->config_base, 0);
}
static int spr_cha_hw_config(struct intel_uncore_box *box, struct perf_event *event)
@@ -6485,7 +6486,7 @@ void spr_uncore_cpu_init(void)
* of UNCORE_SPR_CHA) is incorrect on some SPR variants because of a
* firmware bug. Using the value from SPR_MSR_UNC_CBO_CONFIG to replace it.
*/
- rdmsrl(SPR_MSR_UNC_CBO_CONFIG, num_cbo);
+ rdmsrq(SPR_MSR_UNC_CBO_CONFIG, num_cbo);
/*
* The MSR doesn't work on the EMR XCC, but the firmware bug doesn't impact
* the EMR XCC. Don't let the value from the MSR replace the existing value.
generated by cgit (git 2.34.1) at 2025-05-30 17:43:56 +0000