[tip:perf/core] perf/x86/intel/pt: Add support for PTWRITE and power event tracing

From: tip-bot for Alexander Shishkin
Date: Mon Sep 19 2016 - 19:26:31 EST


Commit-ID: 8ee83b2ab3d1987cbd80c9f2c6f2b12fed87b51e
Gitweb: http://git.kernel.org/tip/8ee83b2ab3d1987cbd80c9f2c6f2b12fed87b51e
Author: Alexander Shishkin <alexander.shishkin@xxxxxxxxxxxxxxx>
AuthorDate: Fri, 16 Sep 2016 16:48:19 +0300
Committer: Thomas Gleixner <tglx@xxxxxxxxxxxxx>
CommitDate: Tue, 20 Sep 2016 01:18:28 +0200

perf/x86/intel/pt: Add support for PTWRITE and power event tracing

The Intel PT facility grew some new functionality:

* PTWRITE packet carries the payload of the new PTWRITE instruction
that can be used to instrument Intel PT traces with user-supplied
data. Packets of this type are only generated if 'ptwrite' capability
is set and PTWEn bit is set in the event attribute's config. Flow
update packets (FUP) can be generated on PTWRITE packets if FUPonPTW
config bit is set. Setting these bits is not allowed if 'ptwrite'
capability is not set.

* PWRE, PWRX, MWAIT, EXSTOP packets communicate core power management
events. These depend on 'power_event_tracing' capability and are
enabled by setting PwrEvtEn bit in the event attribute.

Extend the driver capabilities and provide the proper sanity checks in the
event validation function.

[ tglx: Massaged changelog ]

Signed-off-by: Alexander Shishkin <alexander.shishkin@xxxxxxxxxxxxxxx>
Cc: Arnaldo Carvalho de Melo <acme@xxxxxxxxxxxxx>
Cc: Peter Zijlstra <a.p.zijlstra@xxxxxxxxx>
Cc: vince@xxxxxxxxxx
Cc: eranian@xxxxxxxxxx
Cc: Adrian Hunter <adrian.hunter@xxxxxxxxx>
Link: http://lkml.kernel.org/r/20160916134819.1978-1-alexander.shishkin@xxxxxxxxxxxxxxx
Signed-off-by: Thomas Gleixner <tglx@xxxxxxxxxxxxx>

---
arch/x86/events/intel/pt.c | 24 +++++++++++++++++++++++-
arch/x86/events/intel/pt.h | 5 +++++
2 files changed, 28 insertions(+), 1 deletion(-)

diff --git a/arch/x86/events/intel/pt.c b/arch/x86/events/intel/pt.c
index 04bb5fb..18d18fd 100644
--- a/arch/x86/events/intel/pt.c
+++ b/arch/x86/events/intel/pt.c
@@ -69,6 +69,8 @@ static struct pt_cap_desc {
PT_CAP(psb_cyc, 0, CR_EBX, BIT(1)),
PT_CAP(ip_filtering, 0, CR_EBX, BIT(2)),
PT_CAP(mtc, 0, CR_EBX, BIT(3)),
+ PT_CAP(ptwrite, 0, CR_EBX, BIT(4)),
+ PT_CAP(power_event_trace, 0, CR_EBX, BIT(5)),
PT_CAP(topa_output, 0, CR_ECX, BIT(0)),
PT_CAP(topa_multiple_entries, 0, CR_ECX, BIT(1)),
PT_CAP(single_range_output, 0, CR_ECX, BIT(2)),
@@ -259,10 +261,16 @@ fail:
#define RTIT_CTL_MTC (RTIT_CTL_MTC_EN | \
RTIT_CTL_MTC_RANGE)

+#define RTIT_CTL_PTW (RTIT_CTL_PTW_EN | \
+ RTIT_CTL_FUP_ON_PTW)
+
#define PT_CONFIG_MASK (RTIT_CTL_TSC_EN | \
RTIT_CTL_DISRETC | \
RTIT_CTL_CYC_PSB | \
- RTIT_CTL_MTC)
+ RTIT_CTL_MTC | \
+ RTIT_CTL_PWR_EVT_EN | \
+ RTIT_CTL_FUP_ON_PTW | \
+ RTIT_CTL_PTW_EN)

static bool pt_event_valid(struct perf_event *event)
{
@@ -311,6 +319,20 @@ static bool pt_event_valid(struct perf_event *event)
return false;
}

+ if (config & RTIT_CTL_PWR_EVT_EN &&
+ !pt_cap_get(PT_CAP_power_event_trace))
+ return false;
+
+ if (config & RTIT_CTL_PTW) {
+ if (!pt_cap_get(PT_CAP_ptwrite))
+ return false;
+
+ /* FUPonPTW without PTW doesn't make sense */
+ if ((config & RTIT_CTL_FUP_ON_PTW) &&
+ !(config & RTIT_CTL_PTW_EN))
+ return false;
+ }
+
return true;
}

diff --git a/arch/x86/events/intel/pt.h b/arch/x86/events/intel/pt.h
index efffa4a..53473c2 100644
--- a/arch/x86/events/intel/pt.h
+++ b/arch/x86/events/intel/pt.h
@@ -26,11 +26,14 @@
#define RTIT_CTL_CYCLEACC BIT(1)
#define RTIT_CTL_OS BIT(2)
#define RTIT_CTL_USR BIT(3)
+#define RTIT_CTL_PWR_EVT_EN BIT(4)
+#define RTIT_CTL_FUP_ON_PTW BIT(5)
#define RTIT_CTL_CR3EN BIT(7)
#define RTIT_CTL_TOPA BIT(8)
#define RTIT_CTL_MTC_EN BIT(9)
#define RTIT_CTL_TSC_EN BIT(10)
#define RTIT_CTL_DISRETC BIT(11)
+#define RTIT_CTL_PTW_EN BIT(12)
#define RTIT_CTL_BRANCH_EN BIT(13)
#define RTIT_CTL_MTC_RANGE_OFFSET 14
#define RTIT_CTL_MTC_RANGE (0x0full << RTIT_CTL_MTC_RANGE_OFFSET)
@@ -91,6 +94,8 @@ enum pt_capabilities {
PT_CAP_psb_cyc,
PT_CAP_ip_filtering,
PT_CAP_mtc,
+ PT_CAP_ptwrite,
+ PT_CAP_power_event_trace,
PT_CAP_topa_output,
PT_CAP_topa_multiple_entries,
PT_CAP_single_range_output,