[PATCH 1/3] perf, x86: Add new cache events table for Haswell

[PATCH 1/3] perf, x86: Add new cache events table for Haswell

[Andi Kleen]

From: Andi Kleen <ak@linux.intel.com>

Haswell offcore events are quite different from Sandy Bridge.
Add a new table to handle Haswell properly.

Note that the offcore bits listed in the SDM are not quite correct
(this is currently being fixed). An uptodate list of bits is
in the patch.

The basic setup is similar to Sandy Bridge. The prefetch columns
have been removed, as prefetch counting is not very reliable
on Haswell. One L1 event that is not in the event list anymore
has been also removed.

- data reads do not include code reads (comparable to earlier Sandy
Bridge tables)
- data counts include speculative execution (except L1 write, dtlb, bpu)
- remote node access includes both remote memory, remote cache, remote mmio.
- prefetches are not included in the counts for consistency
(different from Sandy Bridge, which includes prefetches in the remote node)

The events with additional caveats have references to the specification update.

v2: Change name of variable.
v3: Based on older Broadwell patch, but now just for Haswell.
    Various fixes with events being more similar to the older
    Sandy Bridge tables.
v4: Add aliases. Drop two PF bits.
v5: Don't use SUPPLIER_NONE. Remote node is DRAM only. More aliases.
Signed-off-by: Andi Kleen <ak@linux.intel.com>
---
 arch/x86/kernel/cpu/perf_event_intel.c | 195 ++++++++++++++++++++++++++++++++-
 1 file changed, 193 insertions(+), 2 deletions(-)

diff --git a/arch/x86/kernel/cpu/perf_event_intel.c b/arch/x86/kernel/cpu/perf_event_intel.c
index 498b6d9..5cd7e2b 100644
--- a/arch/x86/kernel/cpu/perf_event_intel.c
+++ b/arch/x86/kernel/cpu/perf_event_intel.c
@@ -415,6 +415,197 @@ static __initconst const u64 snb_hw_cache_event_ids
 
 };
 
+/*
+ * Notes on the events:
+ * - data reads do not include code reads (comparable to earlier tables)
+ * - data counts include speculative execution (except L1 write, dtlb, bpu)
+ * - remote node access includes remote memory, remote cache, remote mmio.
+ * - prefetches are not included in the counts because they are not
+ *   reliably counted.
+ * The events with additional caveats have references to the specification update.
+ */
+
+#define HSW_DEMAND_DATA_RD		BIT_ULL(0)
+#define HSW_DEMAND_RFO			BIT_ULL(1)
+#define HSW_ANY_RESPONSE		BIT_ULL(16)
+#define HSW_SUPPLIER_NONE		BIT_ULL(17)
+#define HSW_L3_MISS_LOCAL_DRAM		BIT_ULL(22)
+#define HSW_L3_MISS_REMOTE_HOP0		BIT_ULL(27)
+#define HSW_L3_MISS_REMOTE_HOP1		BIT_ULL(28)
+#define HSW_L3_MISS_REMOTE_HOP2P	BIT_ULL(29)
+#define HSW_L3_MISS			(HSW_L3_MISS_LOCAL_DRAM| \
+					 HSW_L3_MISS_REMOTE_HOP0|HSW_L3_MISS_REMOTE_HOP1| \
+					 HSW_L3_MISS_REMOTE_HOP2P)
+#define HSW_SNOOP_NONE			BIT_ULL(31)
+#define HSW_SNOOP_NOT_NEEDED		BIT_ULL(32)
+#define HSW_SNOOP_MISS			BIT_ULL(33)
+#define HSW_SNOOP_HIT_NO_FWD		BIT_ULL(34)
+#define HSW_SNOOP_HIT_WITH_FWD		BIT_ULL(35)
+#define HSW_SNOOP_HITM			BIT_ULL(36)
+#define HSW_SNOOP_NON_DRAM		BIT_ULL(37)
+#define HSW_ANY_SNOOP			(HSW_SNOOP_NONE| \
+					 HSW_SNOOP_NOT_NEEDED|HSW_SNOOP_MISS| \
+					 HSW_SNOOP_HIT_NO_FWD|HSW_SNOOP_HIT_WITH_FWD| \
+					 HSW_SNOOP_HITM|HSW_SNOOP_NON_DRAM)
+#define HSW_SNOOP_DRAM			(HSW_ANY_SNOOP & ~HSW_SNOOP_NON_DRAM)
+#define HSW_DEMAND_READ			HSW_DEMAND_DATA_RD
+#define HSW_DEMAND_WRITE		HSW_DEMAND_RFO
+#define HSW_L3_MISS_REMOTE		(HSW_L3_MISS_REMOTE_HOP0|\
+					 HSW_L3_MISS_REMOTE_HOP1|HSW_L3_MISS_REMOTE_HOP2P)
+#define HSW_LLC_ACCESS			HSW_ANY_RESPONSE
+
+static __initconst const u64 hsw_hw_cache_event_ids
+				[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) ] = 0x81d0,	/* MEM_UOPS_RETIRED.ALL_LOADS, HSM30 */
+		[ C(RESULT_MISS)   ] = 0x151,	/* L1D.REPLACEMENT */
+	},
+	[ C(OP_WRITE) ] = {
+		[ C(RESULT_ACCESS) ] = 0x82d0,	/* MEM_UOPS_RETIRED.ALL_STORES, HSM30 */
+		[ C(RESULT_MISS)   ] = 0x0,
+	},
+	[ C(OP_PREFETCH) ] = {
+		[ C(RESULT_ACCESS) ] = 0x0,
+		[ C(RESULT_MISS)   ] = 0x0,
+	},
+ },
+ [ C(L1I ) ] = {
+	[ C(OP_READ) ] = {
+		[ C(RESULT_ACCESS) ] = 0x0,
+		[ C(RESULT_MISS)   ] = 0x280,	/* ICACHE.MISSES */
+	},
+	[ C(OP_WRITE) ] = {
+		[ C(RESULT_ACCESS) ] = -1,
+		[ C(RESULT_MISS)   ] = -1,
+	},
+	[ C(OP_PREFETCH) ] = {
+		[ C(RESULT_ACCESS) ] = 0x0,
+		[ C(RESULT_MISS)   ] = 0x0,
+	},
+ },
+ [ C(LL  ) ] = {
+	[ C(OP_READ) ] = {
+		[ C(RESULT_ACCESS) ] = 0x1b7,	/* OFFCORE_RESPONSE */
+		[ C(RESULT_MISS)   ] = 0x1b7,	/* OFFCORE_RESPONSE */
+	},
+	[ C(OP_WRITE) ] = {
+		[ C(RESULT_ACCESS) ] = 0x1b7,	/* OFFCORE_RESPONSE */
+		[ C(RESULT_MISS)   ] = 0x1b7,	/* OFFCORE_RESPONSE */
+	},
+	[ C(OP_PREFETCH) ] = {
+		[ C(RESULT_ACCESS) ] = 0x0,
+		[ C(RESULT_MISS)   ] = 0x0,
+	},
+ },
+ [ C(DTLB) ] = {
+	[ C(OP_READ) ] = {
+		[ C(RESULT_ACCESS) ] = 0x81d0,	/* MEM_UOPS_RETIRED.ALL_LOADS, HSM30 */
+		[ C(RESULT_MISS)   ] = 0x108,	/* DTLB_LOAD_MISSES.MISS_CAUSES_A_WALK */
+	},
+	[ C(OP_WRITE) ] = {
+		[ C(RESULT_ACCESS) ] = 0x82d0,	/* MEM_UOPS_RETIRED.ALL_STORES, HSM30 */
+		[ C(RESULT_MISS)   ] = 0x149,	/* DTLB_STORE_MISSES.MISS_CAUSES_A_WALK */
+	},
+	[ C(OP_PREFETCH) ] = {
+		[ C(RESULT_ACCESS) ] = 0x0,
+		[ C(RESULT_MISS)   ] = 0x0,
+	},
+ },
+ [ C(ITLB) ] = {
+	[ C(OP_READ) ] = {
+		[ C(RESULT_ACCESS) ] = 0x6085,	/* ITLB_MISSES.STLB_HIT */
+		[ C(RESULT_MISS)   ] = 0x185,	/* ITLB_MISSES.MISS_CAUSES_A_WALK */
+	},
+	[ C(OP_WRITE) ] = {
+		[ C(RESULT_ACCESS) ] = -1,
+		[ C(RESULT_MISS)   ] = -1,
+	},
+	[ C(OP_PREFETCH) ] = {
+		[ C(RESULT_ACCESS) ] = -1,
+		[ C(RESULT_MISS)   ] = -1,
+	},
+ },
+ [ C(BPU ) ] = {
+	[ C(OP_READ) ] = {
+		[ C(RESULT_ACCESS) ] = 0xc4,	/* BR_INST_RETIRED.ALL_BRANCHES */
+		[ C(RESULT_MISS)   ] = 0xc5,	/* BR_MISP_RETIRED.ALL_BRANCHES */
+	},
+	[ C(OP_WRITE) ] = {
+		[ C(RESULT_ACCESS) ] = -1,
+		[ C(RESULT_MISS)   ] = -1,
+	},
+	[ C(OP_PREFETCH) ] = {
+		[ C(RESULT_ACCESS) ] = -1,
+		[ C(RESULT_MISS)   ] = -1,
+	},
+ },
+ [ C(NODE) ] = {
+	[ C(OP_READ) ] = {
+		[ C(RESULT_ACCESS) ] = 0x1b7,	/* OFFCORE_RESPONSE */
+		[ C(RESULT_MISS)   ] = 0x1b7,	/* OFFCORE_RESPONSE */
+	},
+	[ C(OP_WRITE) ] = {
+		[ C(RESULT_ACCESS) ] = 0x1b7,	/* OFFCORE_RESPONSE */
+		[ C(RESULT_MISS)   ] = 0x1b7,	/* OFFCORE_RESPONSE */
+	},
+	[ C(OP_PREFETCH) ] = {
+		[ C(RESULT_ACCESS) ] = 0x0,
+		[ C(RESULT_MISS)   ] = 0x0,
+	},
+ },
+};
+
+static __initconst const u64 hsw_hw_cache_extra_regs
+				[PERF_COUNT_HW_CACHE_MAX]
+				[PERF_COUNT_HW_CACHE_OP_MAX]
+				[PERF_COUNT_HW_CACHE_RESULT_MAX] =
+{
+ [ C(LL  ) ] = {
+	[ C(OP_READ) ] = {
+		[ C(RESULT_ACCESS) ] = HSW_DEMAND_READ|
+				       HSW_LLC_ACCESS,
+		[ C(RESULT_MISS)   ] = HSW_DEMAND_READ|
+				       HSW_L3_MISS|HSW_ANY_SNOOP,
+	},
+	[ C(OP_WRITE) ] = {
+		[ C(RESULT_ACCESS) ] = HSW_DEMAND_WRITE|
+				       HSW_LLC_ACCESS,
+		[ C(RESULT_MISS)   ] = HSW_DEMAND_WRITE|
+				       HSW_L3_MISS|HSW_ANY_SNOOP,
+	},
+	[ C(OP_PREFETCH) ] = {
+		[ C(RESULT_ACCESS) ] = 0x0,
+		[ C(RESULT_MISS)   ] = 0x0,
+	},
+ },
+ [ C(NODE) ] = {
+	[ C(OP_READ) ] = {
+		[ C(RESULT_ACCESS) ] = HSW_DEMAND_READ|
+				       HSW_L3_MISS_LOCAL_DRAM|
+				       HSW_SNOOP_DRAM,
+		[ C(RESULT_MISS)   ] = HSW_DEMAND_READ|
+				       HSW_L3_MISS_REMOTE|
+				       HSW_SNOOP_DRAM,
+	},
+	[ C(OP_WRITE) ] = {
+		[ C(RESULT_ACCESS) ] = HSW_DEMAND_WRITE|
+				       HSW_L3_MISS_LOCAL_DRAM|
+				       HSW_SNOOP_DRAM,
+		[ C(RESULT_MISS)   ] = HSW_DEMAND_WRITE|
+				       HSW_L3_MISS_REMOTE|
+				       HSW_SNOOP_DRAM,
+	},
+	[ C(OP_PREFETCH) ] = {
+		[ C(RESULT_ACCESS) ] = 0x0,
+		[ C(RESULT_MISS)   ] = 0x0,
+	},
+ },
+};
+
 static __initconst const u64 westmere_hw_cache_event_ids
 				[PERF_COUNT_HW_CACHE_MAX]
 				[PERF_COUNT_HW_CACHE_OP_MAX]
@@ -2546,8 +2737,8 @@ __init int intel_pmu_init(void)
 	case 69: /* 22nm Haswell ULT */
 	case 70: /* 22nm Haswell + GT3e (Intel Iris Pro graphics) */
 		x86_pmu.late_ack = true;
-		memcpy(hw_cache_event_ids, snb_hw_cache_event_ids, sizeof(hw_cache_event_ids));
-		memcpy(hw_cache_extra_regs, snb_hw_cache_extra_regs, sizeof(hw_cache_extra_regs));
+		memcpy(hw_cache_event_ids, hsw_hw_cache_event_ids, sizeof(hw_cache_event_ids));
+		memcpy(hw_cache_extra_regs, hsw_hw_cache_extra_regs, sizeof(hw_cache_extra_regs));
 
 		intel_pmu_lbr_init_snb();
 
-- 
1.9.3

[PATCH 2/3] perf, x86: Add Broadwell core support

[Andi Kleen]

From: Andi Kleen <ak@linux.intel.com>

Add Broadwell support for Broadwell to perf.

The basic support is very similar to Haswell. We use the new cache
event list added for Haswell earlier. The only differences
are a few bits related to remote nodes. To avoid an extra,
mostly identical, table these are patched up in the initialization code.

The constraint list has one new event that needs to be handled over Haswell.

Includes code and testing from Kan Liang.

v2: Remove unnamed model numbers.
v3: Rename cache event list to hsw_*. Change names.
v4: Use symbolic names for cache events. Improve comments and description.
    Fix sparse warnings (Fengguang Wu)
    Add Xeon D model number.
    Remove cache event table (in separate patch)
    Patch up remote node differences (Kan Liang)
v5: Update offcore bits
Signed-off-by: Andi Kleen <ak@linux.intel.com>
---
 arch/x86/kernel/cpu/perf_event_intel.c | 47 ++++++++++++++++++++++++++++++++++
 1 file changed, 47 insertions(+)

diff --git a/arch/x86/kernel/cpu/perf_event_intel.c b/arch/x86/kernel/cpu/perf_event_intel.c
index 5cd7e2b..259e06b 100644
--- a/arch/x86/kernel/cpu/perf_event_intel.c
+++ b/arch/x86/kernel/cpu/perf_event_intel.c
@@ -220,6 +220,15 @@ static struct event_constraint intel_hsw_event_constraints[] = {
 	EVENT_CONSTRAINT_END
 };
 
+struct event_constraint intel_bdw_event_constraints[] = {
+	FIXED_EVENT_CONSTRAINT(0x00c0, 0),	/* INST_RETIRED.ANY */
+	FIXED_EVENT_CONSTRAINT(0x003c, 1),	/* CPU_CLK_UNHALTED.CORE */
+	FIXED_EVENT_CONSTRAINT(0x0300, 2),	/* CPU_CLK_UNHALTED.REF */
+	INTEL_UEVENT_CONSTRAINT(0x148, 0x4),	/* L1D_PEND_MISS.PENDING */
+	INTEL_EVENT_CONSTRAINT(0xa3, 0x4),	/* CYCLE_ACTIVITY.* */
+	EVENT_CONSTRAINT_END
+};
+
 static u64 intel_pmu_event_map(int hw_event)
 {
 	return intel_perfmon_event_map[hw_event];
@@ -454,6 +463,12 @@ static __initconst const u64 snb_hw_cache_event_ids
 					 HSW_L3_MISS_REMOTE_HOP1|HSW_L3_MISS_REMOTE_HOP2P)
 #define HSW_LLC_ACCESS			HSW_ANY_RESPONSE
 
+#define BDW_L3_MISS_LOCAL		BIT(26)
+#define BDW_L3_MISS			(BDW_L3_MISS_LOCAL| \
+					 HSW_L3_MISS_REMOTE_HOP0|HSW_L3_MISS_REMOTE_HOP1| \
+					 HSW_L3_MISS_REMOTE_HOP2P)
+
+
 static __initconst const u64 hsw_hw_cache_event_ids
 				[PERF_COUNT_HW_CACHE_MAX]
 				[PERF_COUNT_HW_CACHE_OP_MAX]
@@ -2757,6 +2772,38 @@ __init int intel_pmu_init(void)
 		pr_cont("Haswell events, ");
 		break;
 
+	case 61: /* 14nm Broadwell Core-M */
+	case 86: /* 14nm Broadwell Xeon D */
+		x86_pmu.late_ack = true;
+		memcpy(hw_cache_event_ids, hsw_hw_cache_event_ids, sizeof(hw_cache_event_ids));
+		memcpy(hw_cache_extra_regs, hsw_hw_cache_extra_regs, sizeof(hw_cache_extra_regs));
+
+		/* L3_MISS_LOCAL_DRAM is BIT(26) in Broadwell */
+		hw_cache_extra_regs[C(LL)][C(OP_READ)][C(RESULT_MISS)] = HSW_DEMAND_READ |
+									 BDW_L3_MISS|HSW_SNOOP_DRAM;
+		hw_cache_extra_regs[C(LL)][C(OP_WRITE)][C(RESULT_MISS)] = HSW_DEMAND_WRITE|BDW_L3_MISS|
+									  HSW_SNOOP_DRAM;
+		hw_cache_extra_regs[C(NODE)][C(OP_READ)][C(RESULT_ACCESS)] = HSW_DEMAND_READ|
+									     BDW_L3_MISS_LOCAL|HSW_SNOOP_DRAM;
+		hw_cache_extra_regs[C(NODE)][C(OP_WRITE)][C(RESULT_ACCESS)] = HSW_DEMAND_WRITE|
+									      BDW_L3_MISS_LOCAL|HSW_SNOOP_DRAM;
+
+		intel_pmu_lbr_init_snb();
+
+		x86_pmu.event_constraints = intel_bdw_event_constraints;
+		x86_pmu.pebs_constraints = intel_hsw_pebs_event_constraints;
+		x86_pmu.extra_regs = intel_snbep_extra_regs;
+		x86_pmu.pebs_aliases = intel_pebs_aliases_snb;
+		/* all extra regs are per-cpu when HT is on */
+		x86_pmu.er_flags |= ERF_HAS_RSP_1;
+		x86_pmu.er_flags |= ERF_NO_HT_SHARING;
+
+		x86_pmu.hw_config = hsw_hw_config;
+		x86_pmu.get_event_constraints = hsw_get_event_constraints;
+		x86_pmu.cpu_events = hsw_events_attrs;
+		pr_cont("Broadwell events, ");
+		break;
+
 	default:
 		switch (x86_pmu.version) {
 		case 1:
-- 
1.9.3

[tip:perf/core] perf/x86/intel: Add Broadwell core support

[tip-bot for Andi Kleen]

Commit-ID:  91f1b70582c62576f429cf78d53751c66677553d
Gitweb:     http://git.kernel.org/tip/91f1b70582c62576f429cf78d53751c66677553d
Author:     Andi Kleen <ak@linux.intel.com>
AuthorDate: Tue, 17 Feb 2015 18:18:05 -0800
Committer:  Ingo Molnar <mingo@kernel.org>
CommitDate: Fri, 27 Mar 2015 09:14:02 +0100

perf/x86/intel: Add Broadwell core support

Add Broadwell support for Broadwell to perf.

The basic support is very similar to Haswell. We use the new cache
event list added for Haswell earlier. The only differences
are a few bits related to remote nodes. To avoid an extra,
mostly identical, table these are patched up in the initialization code.

The constraint list has one new event that needs to be handled over Haswell.

Includes code and testing from Kan Liang.

Signed-off-by: Andi Kleen <ak@linux.intel.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Link: http://lkml.kernel.org/r/1424225886-18652-2-git-send-email-andi@firstfloor.org
Signed-off-by: Ingo Molnar <mingo@kernel.org>
---
 arch/x86/kernel/cpu/perf_event_intel.c | 47 ++++++++++++++++++++++++++++++++++
 1 file changed, 47 insertions(+)

diff --git a/arch/x86/kernel/cpu/perf_event_intel.c b/arch/x86/kernel/cpu/perf_event_intel.c
index 5ef64bf..2883853 100644
--- a/arch/x86/kernel/cpu/perf_event_intel.c
+++ b/arch/x86/kernel/cpu/perf_event_intel.c
@@ -220,6 +220,15 @@ static struct event_constraint intel_hsw_event_constraints[] = {
 	EVENT_CONSTRAINT_END
 };
 
+struct event_constraint intel_bdw_event_constraints[] = {
+	FIXED_EVENT_CONSTRAINT(0x00c0, 0),	/* INST_RETIRED.ANY */
+	FIXED_EVENT_CONSTRAINT(0x003c, 1),	/* CPU_CLK_UNHALTED.CORE */
+	FIXED_EVENT_CONSTRAINT(0x0300, 2),	/* CPU_CLK_UNHALTED.REF */
+	INTEL_UEVENT_CONSTRAINT(0x148, 0x4),	/* L1D_PEND_MISS.PENDING */
+	INTEL_EVENT_CONSTRAINT(0xa3, 0x4),	/* CYCLE_ACTIVITY.* */
+	EVENT_CONSTRAINT_END
+};
+
 static u64 intel_pmu_event_map(int hw_event)
 {
 	return intel_perfmon_event_map[hw_event];
@@ -453,6 +462,12 @@ static __initconst const u64 snb_hw_cache_event_ids
 					 HSW_L3_MISS_REMOTE_HOP1|HSW_L3_MISS_REMOTE_HOP2P)
 #define HSW_LLC_ACCESS			HSW_ANY_RESPONSE
 
+#define BDW_L3_MISS_LOCAL		BIT(26)
+#define BDW_L3_MISS			(BDW_L3_MISS_LOCAL| \
+					 HSW_L3_MISS_REMOTE_HOP0|HSW_L3_MISS_REMOTE_HOP1| \
+					 HSW_L3_MISS_REMOTE_HOP2P)
+
+
 static __initconst const u64 hsw_hw_cache_event_ids
 				[PERF_COUNT_HW_CACHE_MAX]
 				[PERF_COUNT_HW_CACHE_OP_MAX]
@@ -2730,6 +2745,38 @@ __init int intel_pmu_init(void)
 		pr_cont("Haswell events, ");
 		break;
 
+	case 61: /* 14nm Broadwell Core-M */
+	case 86: /* 14nm Broadwell Xeon D */
+		x86_pmu.late_ack = true;
+		memcpy(hw_cache_event_ids, hsw_hw_cache_event_ids, sizeof(hw_cache_event_ids));
+		memcpy(hw_cache_extra_regs, hsw_hw_cache_extra_regs, sizeof(hw_cache_extra_regs));
+
+		/* L3_MISS_LOCAL_DRAM is BIT(26) in Broadwell */
+		hw_cache_extra_regs[C(LL)][C(OP_READ)][C(RESULT_MISS)] = HSW_DEMAND_READ |
+									 BDW_L3_MISS|HSW_SNOOP_DRAM;
+		hw_cache_extra_regs[C(LL)][C(OP_WRITE)][C(RESULT_MISS)] = HSW_DEMAND_WRITE|BDW_L3_MISS|
+									  HSW_SNOOP_DRAM;
+		hw_cache_extra_regs[C(NODE)][C(OP_READ)][C(RESULT_ACCESS)] = HSW_DEMAND_READ|
+									     BDW_L3_MISS_LOCAL|HSW_SNOOP_DRAM;
+		hw_cache_extra_regs[C(NODE)][C(OP_WRITE)][C(RESULT_ACCESS)] = HSW_DEMAND_WRITE|
+									      BDW_L3_MISS_LOCAL|HSW_SNOOP_DRAM;
+
+		intel_pmu_lbr_init_snb();
+
+		x86_pmu.event_constraints = intel_bdw_event_constraints;
+		x86_pmu.pebs_constraints = intel_hsw_pebs_event_constraints;
+		x86_pmu.extra_regs = intel_snbep_extra_regs;
+		x86_pmu.pebs_aliases = intel_pebs_aliases_snb;
+		/* all extra regs are per-cpu when HT is on */
+		x86_pmu.er_flags |= ERF_HAS_RSP_1;
+		x86_pmu.er_flags |= ERF_NO_HT_SHARING;
+
+		x86_pmu.hw_config = hsw_hw_config;
+		x86_pmu.get_event_constraints = hsw_get_event_constraints;
+		x86_pmu.cpu_events = hsw_events_attrs;
+		pr_cont("Broadwell events, ");
+		break;
+
 	default:
 		switch (x86_pmu.version) {
 		case 1:

[PATCH 3/3] perf, x86: Add INST_RETIRED.ALL workarounds

[Andi Kleen]

From: Andi Kleen <ak@linux.intel.com>

On Broadwell INST_RETIRED.ALL cannot be used with any period
that doesn't have the lowest 6 bits cleared. And the period
should not be smaller than 128.

Add a new callback to enforce this, and set it for Broadwell.

This is erratum BDM57 and BDM11.

How does this handle the case when an app requests a specific
period with some of the bottom bits set

The apps thinks it is sampling at X occurences per sample, when it is
in fact at X - 63 (worst case).

Short answer:

Any useful instruction sampling period needs to be 4-6 orders
of magnitude larger than 128, as an PMI every 128 instructions
would instantly overwhelm the system and be throttled.
So the +-64 error from this is really small compared to the
period, much smaller than normal system jitter.

Long answer:

<write up by Peter:>

IFF we guarantee perf_event_attr::sample_period >= 128.

Suppose we start out with sample_period=192; then we'll set period_left
to 192, we'll end up with left = 128 (we truncate the lower bits). We
get an interrupt, find that period_left = 64 (>0 so we return 0 and
don't get an overflow handler), up that to 128. Then we trigger again,
at n=256. Then we find period_left = -64 (<=0 so we return 1 and do get
an overflow). We increment with sample_period so we get left = 128. We
fire again, at n=384, period_left = 0 (<=0 so we return 1 and get an
overflow). And on and on.

So while the individual interrupts are 'wrong' we get then with
interval=256,128 in exactly the right ratio to average out at 192. And
this works for everything >=128.

So the num_samples*fixed_period thing is still entirely correct +- 127,
which is good enough I'd say, as you already have that error anyhow.

So no need to 'fix' the tools, al we need to do is refuse to create
INST_RETIRED:ALL events with sample_period < 128.

v2: Use correct event name in description. Use EVENT() macro.
v3: Use INTEL_ARCH_EVENT_MASK. Error out for events with too small period.
v4: Expand description.
Signed-off-by: Andi Kleen <ak@linux.intel.com>
---
 arch/x86/kernel/cpu/perf_event.c       |  9 +++++++++
 arch/x86/kernel/cpu/perf_event.h       |  1 +
 arch/x86/kernel/cpu/perf_event_intel.c | 19 +++++++++++++++++++
 3 files changed, 29 insertions(+)

diff --git a/arch/x86/kernel/cpu/perf_event.c b/arch/x86/kernel/cpu/perf_event.c
index 143e5f5..66451a6 100644
--- a/arch/x86/kernel/cpu/perf_event.c
+++ b/arch/x86/kernel/cpu/perf_event.c
@@ -445,6 +445,12 @@ int x86_pmu_hw_config(struct perf_event *event)
 	if (event->attr.type == PERF_TYPE_RAW)
 		event->hw.config |= event->attr.config & X86_RAW_EVENT_MASK;
 
+	if (event->attr.sample_period && x86_pmu.limit_period) {
+		if (x86_pmu.limit_period(event, event->attr.sample_period) >
+				event->attr.sample_period)
+			return -EINVAL;
+	}
+
 	return x86_setup_perfctr(event);
 }
 
@@ -982,6 +988,9 @@ int x86_perf_event_set_period(struct perf_event *event)
 	if (left > x86_pmu.max_period)
 		left = x86_pmu.max_period;
 
+	if (x86_pmu.limit_period)
+		left = x86_pmu.limit_period(event, left);
+
 	per_cpu(pmc_prev_left[idx], smp_processor_id()) = left;
 
 	/*
diff --git a/arch/x86/kernel/cpu/perf_event.h b/arch/x86/kernel/cpu/perf_event.h
index 4e6cdb0..6ce77bd 100644
--- a/arch/x86/kernel/cpu/perf_event.h
+++ b/arch/x86/kernel/cpu/perf_event.h
@@ -449,6 +449,7 @@ struct x86_pmu {
 	struct x86_pmu_quirk *quirks;
 	int		perfctr_second_write;
 	bool		late_ack;
+	unsigned	(*limit_period)(struct perf_event *event, unsigned l);
 
 	/*
 	 * sysfs attrs
diff --git a/arch/x86/kernel/cpu/perf_event_intel.c b/arch/x86/kernel/cpu/perf_event_intel.c
index 259e06b..222274c 100644
--- a/arch/x86/kernel/cpu/perf_event_intel.c
+++ b/arch/x86/kernel/cpu/perf_event_intel.c
@@ -2111,6 +2111,24 @@ hsw_get_event_constraints(struct cpu_hw_events *cpuc, struct perf_event *event)
 	return c;
 }
 
+/*
+ * Broadwell:
+ * The INST_RETIRED.ALL period always needs to have lowest
+ * 6bits cleared (BDM57). It shall not use a period smaller
+ * than 100 (BDM11). We combine the two to enforce
+ * a min-period of 128.
+ */
+static unsigned bdw_limit_period(struct perf_event *event, unsigned left)
+{
+	if ((event->hw.config & INTEL_ARCH_EVENT_MASK) ==
+			X86_CONFIG(.event=0xc0, .umask=0x01)) {
+		if (left < 128)
+			left = 128;
+		left &= ~0x3fu;
+	}
+	return left;
+}
+
 PMU_FORMAT_ATTR(event,	"config:0-7"	);
 PMU_FORMAT_ATTR(umask,	"config:8-15"	);
 PMU_FORMAT_ATTR(edge,	"config:18"	);
@@ -2801,6 +2819,7 @@ __init int intel_pmu_init(void)
 		x86_pmu.hw_config = hsw_hw_config;
 		x86_pmu.get_event_constraints = hsw_get_event_constraints;
 		x86_pmu.cpu_events = hsw_events_attrs;
+		x86_pmu.limit_period = bdw_limit_period;
 		pr_cont("Broadwell events, ");
 		break;
 
-- 
1.9.3

Re: [PATCH 3/3] perf, x86: Add INST_RETIRED.ALL workarounds

[Ingo Molnar]

* Andi Kleen <andi@firstfloor.org> wrote:

> From: Andi Kleen <ak@linux.intel.com>
> 
> On Broadwell INST_RETIRED.ALL cannot be used with any period
> that doesn't have the lowest 6 bits cleared. And the period
> should not be smaller than 128.

Sloppy changelog: a most basic question is not answered by the 
changelog: what happens in practice when the period is set to a 
smaller value than 128?

> +/*
> + * Broadwell:
> + * The INST_RETIRED.ALL period always needs to have lowest
> + * 6bits cleared (BDM57). It shall not use a period smaller
> + * than 100 (BDM11). We combine the two to enforce
> + * a min-period of 128.
> + */

Sloppy comment: that's not what we do:

> +static unsigned bdw_limit_period(struct perf_event *event, unsigned left)
> +{
> +	if ((event->hw.config & INTEL_ARCH_EVENT_MASK) ==
> +			X86_CONFIG(.event=0xc0, .umask=0x01)) {
> +		if (left < 128)
> +			left = 128;
> +		left &= ~0x3fu;
> +	}
> +	return left;

We enforce a minimum period of 128 and round the requested period to 
64.

I think in this case it would be useful to tooling if we updated the 
syscall attribute with the real period value that was used, to not 
skew tooling output.

Thanks,

	Ingo

Re: [PATCH 3/3] perf, x86: Add INST_RETIRED.ALL workarounds

[Peter Zijlstra]

On Mon, Mar 23, 2015 at 10:38:54AM +0100, Ingo Molnar wrote:
> 
> * Andi Kleen <andi@firstfloor.org> wrote:
> 
> > From: Andi Kleen <ak@linux.intel.com>
> > 
> > On Broadwell INST_RETIRED.ALL cannot be used with any period
> > that doesn't have the lowest 6 bits cleared. And the period
> > should not be smaller than 128.
> 
> Sloppy changelog: a most basic question is not answered by the 
> changelog: what happens in practice when the period is set to a 
> smaller value than 128?

http://www.intel.com/content/dam/www/public/us/en/documents/specification-updates/5th-gen-core-family-spec-update.pdf

BDM11 and BDM55 (not 57) tell us that the PMU will generate crap output
if you don't do this. Non-fatal but gibberish.

> > +/*
> > + * Broadwell:
> > + * The INST_RETIRED.ALL period always needs to have lowest
> > + * 6bits cleared (BDM57). It shall not use a period smaller
> > + * than 100 (BDM11). We combine the two to enforce
> > + * a min-period of 128.
> > + */
> 
> Sloppy comment: that's not what we do:
> 
> > +static unsigned bdw_limit_period(struct perf_event *event, unsigned left)
> > +{
> > +	if ((event->hw.config & INTEL_ARCH_EVENT_MASK) ==
> > +			X86_CONFIG(.event=0xc0, .umask=0x01)) {
> > +		if (left < 128)
> > +			left = 128;
> > +		left &= ~0x3fu;
> > +	}
> > +	return left;
> 
> We enforce a minimum period of 128 and round the requested period to 
> 64.

Not quite, we enforce a min period of 128 but otherwise mask bit0-5, no
rounding up.

Which is pretty much what the comment says.

> I think in this case it would be useful to tooling if we updated the 
> syscall attribute with the real period value that was used, to not 
> skew tooling output.

Seeing how we already have a fuzz of up to sample_period events; we
don't know how far into the last period we are when we stop the event,
it might have been 1 event away from generating a PMI, this patch
doesn't actually add significantly to that.

Also, the effective period is the one specified, if the requested period
< 128 we simply reject the event creation. If its any larger we iterate
around the requested sample period with steps of 64 but such that we
average out on the requested period. There is no 'real' period to copy
back.

Another way to look at this is that we use a form of pulse density
modulation to create the desired period using the larger step size; or
perhaps compare it to Breshenham's line drawing algorithm.

Re: [PATCH 3/3] perf, x86: Add INST_RETIRED.ALL workarounds

[Ingo Molnar]

* Peter Zijlstra <peterz@infradead.org> wrote:

> On Mon, Mar 23, 2015 at 10:38:54AM +0100, Ingo Molnar wrote:
> > 
> > * Andi Kleen <andi@firstfloor.org> wrote:
> > 
> > > From: Andi Kleen <ak@linux.intel.com>
> > > 
> > > On Broadwell INST_RETIRED.ALL cannot be used with any period
> > > that doesn't have the lowest 6 bits cleared. And the period
> > > should not be smaller than 128.
> > 
> > Sloppy changelog: a most basic question is not answered by the 
> > changelog: what happens in practice when the period is set to a 
> > smaller value than 128?
> 
> http://www.intel.com/content/dam/www/public/us/en/documents/specification-updates/5th-gen-core-family-spec-update.pdf
> 
> BDM11 and BDM55 (not 57) tell us that the PMU will generate crap output
> if you don't do this. Non-fatal but gibberish.

Should be part of the changelog?

> > > +/*
> > > + * Broadwell:
> > > + * The INST_RETIRED.ALL period always needs to have lowest
> > > + * 6bits cleared (BDM57). It shall not use a period smaller
> > > + * than 100 (BDM11). We combine the two to enforce
> > > + * a min-period of 128.
> > > + */
> > 
> > Sloppy comment: that's not what we do:
> > 
> > > +static unsigned bdw_limit_period(struct perf_event *event, unsigned left)
> > > +{
> > > +	if ((event->hw.config & INTEL_ARCH_EVENT_MASK) ==
> > > +			X86_CONFIG(.event=0xc0, .umask=0x01)) {
> > > +		if (left < 128)
> > > +			left = 128;
> > > +		left &= ~0x3fu;
> > > +	}
> > > +	return left;
> > 
> > We enforce a minimum period of 128 and round the requested period to 
> > 64.
> 
> Not quite, we enforce a min period of 128 but otherwise mask bit0-5, no
> rounding up.

So I did not say rounding up, I meant this sentence:

> > > + *   [...] We combine the two to enforce
> > > + * a min-period of 128.

IMO ambiguously suggests that the result of the combination of the two 
is to enforce a min-period of 128. Would somethin like this:

	          We combine the two to enforce
          a min-period of 128, rounded (down) to multiples of 64.
          The original period is still kept by the core code and is 
          approximated in the long run via these slightly fuzzed 
          hardware-periods.

work with you?

> > I think in this case it would be useful to tooling if we updated 
> > the syscall attribute with the real period value that was used, to 
> > not skew tooling output.
> 
> Seeing how we already have a fuzz of up to sample_period events; we 
> don't know how far into the last period we are when we stop the 
> event, it might have been 1 event away from generating a PMI, this 
> patch doesn't actually add significantly to that.
> 
> Also, the effective period is the one specified, if the requested 
> period < 128 we simply reject the event creation. If its any larger 
> we iterate around the requested sample period with steps of 64 but 
> such that we average out on the requested period. There is no 'real' 
> period to copy back.

Yeah, fair enough.

Thanks,

	Ingo

Re: [PATCH 3/3] perf, x86: Add INST_RETIRED.ALL workarounds

[Peter Zijlstra]

On Mon, Mar 23, 2015 at 11:39:00AM +0100, Ingo Molnar wrote:
> > http://www.intel.com/content/dam/www/public/us/en/documents/specification-updates/5th-gen-core-family-spec-update.pdf
> > 
> > BDM11 and BDM55 (not 57) tell us that the PMU will generate crap output
> > if you don't do this. Non-fatal but gibberish.
> 
> Should be part of the changelog?

Sure, lemme go make that happen.

> So I did not say rounding up, I meant this sentence:
> 
> > > > + *   [...] We combine the two to enforce
> > > > + * a min-period of 128.
> 
> IMO ambiguously suggests that the result of the combination of the two 
> is to enforce a min-period of 128. Would somethin like this:
> 
> 	          We combine the two to enforce
>           a min-period of 128, rounded (down) to multiples of 64.
>           The original period is still kept by the core code and is 
>           approximated in the long run via these slightly fuzzed 
>           hardware-periods.

Like so then?

---
Subject: perf, x86: Add INST_RETIRED.ALL workarounds
From: Andi Kleen <ak@linux.intel.com>
Date: Tue, 17 Feb 2015 18:18:06 -0800

On Broadwell INST_RETIRED.ALL cannot be used with any period
that doesn't have the lowest 6 bits cleared. And the period
should not be smaller than 128.

This is erratum BDM11 and BDM55;
http://www.intel.com/content/dam/www/public/us/en/documents/specification-updates/5th-gen-core-family-spec-update.pdf

BDM11: When using a period < 100; we may get incorrect PEBS/PMI
interrupts and/or an invalid counter state.
BDM55: When bit0-5 of the period are !0 we may get redundant PEBS
records on overflow.

Add a new callback to enforce this, and set it for Broadwell.

How does this handle the case when an app requests a specific
period with some of the bottom bits set?

Short answer:

Any useful instruction sampling period needs to be 4-6 orders
of magnitude larger than 128, as an PMI every 128 instructions
would instantly overwhelm the system and be throttled.
So the +-64 error from this is really small compared to the
period, much smaller than normal system jitter.

Long answer (by Peterz):

IFF we guarantee perf_event_attr::sample_period >= 128.

Suppose we start out with sample_period=192; then we'll set period_left
to 192, we'll end up with left = 128 (we truncate the lower bits). We
get an interrupt, find that period_left = 64 (>0 so we return 0 and
don't get an overflow handler), up that to 128. Then we trigger again,
at n=256. Then we find period_left = -64 (<=0 so we return 1 and do get
an overflow). We increment with sample_period so we get left = 128. We
fire again, at n=384, period_left = 0 (<=0 so we return 1 and get an
overflow). And on and on.

So while the individual interrupts are 'wrong' we get then with
interval=256,128 in exactly the right ratio to average out at 192. And
this works for everything >=128.

So the num_samples*fixed_period thing is still entirely correct +- 127,
which is good enough I'd say, as you already have that error anyhow.

So no need to 'fix' the tools, al we need to do is refuse to create
INST_RETIRED:ALL events with sample_period < 128.

Signed-off-by: Andi Kleen <ak@linux.intel.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
---
 arch/x86/kernel/cpu/perf_event.c       |    9 +++++++++
 arch/x86/kernel/cpu/perf_event.h       |    1 +
 arch/x86/kernel/cpu/perf_event_intel.c |   27 +++++++++++++++++++++++++++
 3 files changed, 37 insertions(+)

--- a/arch/x86/kernel/cpu/perf_event.c
+++ b/arch/x86/kernel/cpu/perf_event.c
@@ -451,6 +451,12 @@ int x86_pmu_hw_config(struct perf_event
 	if (event->attr.type == PERF_TYPE_RAW)
 		event->hw.config |= event->attr.config & X86_RAW_EVENT_MASK;
 
+	if (event->attr.sample_period && x86_pmu.limit_period) {
+		if (x86_pmu.limit_period(event, event->attr.sample_period) >
+				event->attr.sample_period)
+			return -EINVAL;
+	}
+
 	return x86_setup_perfctr(event);
 }
 
@@ -988,6 +994,9 @@ int x86_perf_event_set_period(struct per
 	if (left > x86_pmu.max_period)
 		left = x86_pmu.max_period;
 
+	if (x86_pmu.limit_period)
+		left = x86_pmu.limit_period(event, left);
+
 	per_cpu(pmc_prev_left[idx], smp_processor_id()) = left;
 
 	/*
--- a/arch/x86/kernel/cpu/perf_event.h
+++ b/arch/x86/kernel/cpu/perf_event.h
@@ -451,6 +451,7 @@ struct x86_pmu {
 	struct x86_pmu_quirk *quirks;
 	int		perfctr_second_write;
 	bool		late_ack;
+	unsigned	(*limit_period)(struct perf_event *event, unsigned l);
 
 	/*
 	 * sysfs attrs
--- a/arch/x86/kernel/cpu/perf_event_intel.c
+++ b/arch/x86/kernel/cpu/perf_event_intel.c
@@ -2097,6 +2097,32 @@ hsw_get_event_constraints(struct cpu_hw_
 	return c;
 }
 
+/*
+ * Broadwell:
+ *
+ * The INST_RETIRED.ALL period always needs to have lowest 6 bits cleared
+ * (BDM55) and it must not use a period smaller than 100 (BDM11). We combine
+ * the two to enforce a minimum period of 128 (the smallest value that has bits
+ * 0-5 cleared and >= 100).
+ *
+ * Because of how the code in x86_perf_event_set_period() works, the truncation
+ * of the lower 6 bits is 'harmless' as we'll occasionally add a longer period
+ * to make up for the 'lost' events due to carrying the 'error' in period_left.
+ *
+ * Therefore the effective (average) period matches the requested period,
+ * despite coarser hardware granularity.
+ */
+static unsigned bdw_limit_period(struct perf_event *event, unsigned left)
+{
+	if ((event->hw.config & INTEL_ARCH_EVENT_MASK) ==
+			X86_CONFIG(.event=0xc0, .umask=0x01)) {
+		if (left < 128)
+			left = 128;
+		left &= ~0x3fu;
+	}
+	return left;
+}
+
 PMU_FORMAT_ATTR(event,	"config:0-7"	);
 PMU_FORMAT_ATTR(umask,	"config:8-15"	);
 PMU_FORMAT_ATTR(edge,	"config:18"	);
@@ -2775,6 +2801,7 @@ __init int intel_pmu_init(void)
 		x86_pmu.hw_config = hsw_hw_config;
 		x86_pmu.get_event_constraints = hsw_get_event_constraints;
 		x86_pmu.cpu_events = hsw_events_attrs;
+		x86_pmu.limit_period = bdw_limit_period;
 		pr_cont("Broadwell events, ");
 		break;
 

Re: [PATCH 3/3] perf, x86: Add INST_RETIRED.ALL workarounds

[Ingo Molnar]

* Peter Zijlstra <peterz@infradead.org> wrote:

> On Mon, Mar 23, 2015 at 11:39:00AM +0100, Ingo Molnar wrote:
> > > http://www.intel.com/content/dam/www/public/us/en/documents/specification-updates/5th-gen-core-family-spec-update.pdf
> > > 
> > > BDM11 and BDM55 (not 57) tell us that the PMU will generate crap output
> > > if you don't do this. Non-fatal but gibberish.
> > 
> > Should be part of the changelog?
> 
> Sure, lemme go make that happen.
> 
> > So I did not say rounding up, I meant this sentence:
> > 
> > > > > + *   [...] We combine the two to enforce
> > > > > + * a min-period of 128.
> > 
> > IMO ambiguously suggests that the result of the combination of the two 
> > is to enforce a min-period of 128. Would somethin like this:
> > 
> > 	          We combine the two to enforce
> >           a min-period of 128, rounded (down) to multiples of 64.
> >           The original period is still kept by the core code and is 
> >           approximated in the long run via these slightly fuzzed 
> >           hardware-periods.
> 
> Like so then?

Yeah, looks good to me!

Thanks,

	Ingo

[tip:perf/core] perf/x86/intel: Add INST_RETIRED.ALL workarounds

[tip-bot for Andi Kleen]

Commit-ID:  294fe0f52a44c6f207211de0686c369a961b5533
Gitweb:     http://git.kernel.org/tip/294fe0f52a44c6f207211de0686c369a961b5533
Author:     Andi Kleen <ak@linux.intel.com>
AuthorDate: Tue, 17 Feb 2015 18:18:06 -0800
Committer:  Ingo Molnar <mingo@kernel.org>
CommitDate: Fri, 27 Mar 2015 09:14:03 +0100

perf/x86/intel: Add INST_RETIRED.ALL workarounds

On Broadwell INST_RETIRED.ALL cannot be used with any period
that doesn't have the lowest 6 bits cleared. And the period
should not be smaller than 128.

This is erratum BDM11 and BDM55:

  http://www.intel.com/content/dam/www/public/us/en/documents/specification-updates/5th-gen-core-family-spec-update.pdf

BDM11: When using a period < 100; we may get incorrect PEBS/PMI
interrupts and/or an invalid counter state.
BDM55: When bit0-5 of the period are !0 we may get redundant PEBS
records on overflow.

Add a new callback to enforce this, and set it for Broadwell.

How does this handle the case when an app requests a specific
period with some of the bottom bits set?

Short answer:

Any useful instruction sampling period needs to be 4-6 orders
of magnitude larger than 128, as an PMI every 128 instructions
would instantly overwhelm the system and be throttled.
So the +-64 error from this is really small compared to the
period, much smaller than normal system jitter.

Long answer (by Peterz):

IFF we guarantee perf_event_attr::sample_period >= 128.

Suppose we start out with sample_period=192; then we'll set period_left
to 192, we'll end up with left = 128 (we truncate the lower bits). We
get an interrupt, find that period_left = 64 (>0 so we return 0 and
don't get an overflow handler), up that to 128. Then we trigger again,
at n=256. Then we find period_left = -64 (<=0 so we return 1 and do get
an overflow). We increment with sample_period so we get left = 128. We
fire again, at n=384, period_left = 0 (<=0 so we return 1 and get an
overflow). And on and on.

So while the individual interrupts are 'wrong' we get then with
interval=256,128 in exactly the right ratio to average out at 192. And
this works for everything >=128.

So the num_samples*fixed_period thing is still entirely correct +- 127,
which is good enough I'd say, as you already have that error anyhow.

So no need to 'fix' the tools, al we need to do is refuse to create
INST_RETIRED:ALL events with sample_period < 128.

Signed-off-by: Andi Kleen <ak@linux.intel.com>
[ Updated comments and changelog a bit. ]
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Link: http://lkml.kernel.org/r/1424225886-18652-3-git-send-email-andi@firstfloor.org
Signed-off-by: Ingo Molnar <mingo@kernel.org>
---
 arch/x86/kernel/cpu/perf_event.c       |  9 +++++++++
 arch/x86/kernel/cpu/perf_event.h       |  1 +
 arch/x86/kernel/cpu/perf_event_intel.c | 27 +++++++++++++++++++++++++++
 3 files changed, 37 insertions(+)

diff --git a/arch/x86/kernel/cpu/perf_event.c b/arch/x86/kernel/cpu/perf_event.c
index e0dab5c..ec6e982 100644
--- a/arch/x86/kernel/cpu/perf_event.c
+++ b/arch/x86/kernel/cpu/perf_event.c
@@ -451,6 +451,12 @@ int x86_pmu_hw_config(struct perf_event *event)
 	if (event->attr.type == PERF_TYPE_RAW)
 		event->hw.config |= event->attr.config & X86_RAW_EVENT_MASK;
 
+	if (event->attr.sample_period && x86_pmu.limit_period) {
+		if (x86_pmu.limit_period(event, event->attr.sample_period) >
+				event->attr.sample_period)
+			return -EINVAL;
+	}
+
 	return x86_setup_perfctr(event);
 }
 
@@ -988,6 +994,9 @@ int x86_perf_event_set_period(struct perf_event *event)
 	if (left > x86_pmu.max_period)
 		left = x86_pmu.max_period;
 
+	if (x86_pmu.limit_period)
+		left = x86_pmu.limit_period(event, left);
+
 	per_cpu(pmc_prev_left[idx], smp_processor_id()) = left;
 
 	/*
diff --git a/arch/x86/kernel/cpu/perf_event.h b/arch/x86/kernel/cpu/perf_event.h
index a371d27..87e5081 100644
--- a/arch/x86/kernel/cpu/perf_event.h
+++ b/arch/x86/kernel/cpu/perf_event.h
@@ -451,6 +451,7 @@ struct x86_pmu {
 	struct x86_pmu_quirk *quirks;
 	int		perfctr_second_write;
 	bool		late_ack;
+	unsigned	(*limit_period)(struct perf_event *event, unsigned l);
 
 	/*
 	 * sysfs attrs
diff --git a/arch/x86/kernel/cpu/perf_event_intel.c b/arch/x86/kernel/cpu/perf_event_intel.c
index 2883853..fc6dbc4 100644
--- a/arch/x86/kernel/cpu/perf_event_intel.c
+++ b/arch/x86/kernel/cpu/perf_event_intel.c
@@ -2096,6 +2096,32 @@ hsw_get_event_constraints(struct cpu_hw_events *cpuc, struct perf_event *event)
 	return c;
 }
 
+/*
+ * Broadwell:
+ *
+ * The INST_RETIRED.ALL period always needs to have lowest 6 bits cleared
+ * (BDM55) and it must not use a period smaller than 100 (BDM11). We combine
+ * the two to enforce a minimum period of 128 (the smallest value that has bits
+ * 0-5 cleared and >= 100).
+ *
+ * Because of how the code in x86_perf_event_set_period() works, the truncation
+ * of the lower 6 bits is 'harmless' as we'll occasionally add a longer period
+ * to make up for the 'lost' events due to carrying the 'error' in period_left.
+ *
+ * Therefore the effective (average) period matches the requested period,
+ * despite coarser hardware granularity.
+ */
+static unsigned bdw_limit_period(struct perf_event *event, unsigned left)
+{
+	if ((event->hw.config & INTEL_ARCH_EVENT_MASK) ==
+			X86_CONFIG(.event=0xc0, .umask=0x01)) {
+		if (left < 128)
+			left = 128;
+		left &= ~0x3fu;
+	}
+	return left;
+}
+
 PMU_FORMAT_ATTR(event,	"config:0-7"	);
 PMU_FORMAT_ATTR(umask,	"config:8-15"	);
 PMU_FORMAT_ATTR(edge,	"config:18"	);
@@ -2774,6 +2800,7 @@ __init int intel_pmu_init(void)
 		x86_pmu.hw_config = hsw_hw_config;
 		x86_pmu.get_event_constraints = hsw_get_event_constraints;
 		x86_pmu.cpu_events = hsw_events_attrs;
+		x86_pmu.limit_period = bdw_limit_period;
 		pr_cont("Broadwell events, ");
 		break;
 

Re: [PATCH 1/3] perf, x86: Add new cache events table for Haswell

[Ingo Molnar]

* Andi Kleen <andi@firstfloor.org> wrote:

> From: Andi Kleen <ak@linux.intel.com>
> 
> Haswell offcore events are quite different from Sandy Bridge.
> Add a new table to handle Haswell properly.
> 
> Note that the offcore bits listed in the SDM are not quite correct
> (this is currently being fixed). An uptodate list of bits is
> in the patch.
> 
> The basic setup is similar to Sandy Bridge. The prefetch columns
> have been removed, as prefetch counting is not very reliable
> on Haswell. One L1 event that is not in the event list anymore
> has been also removed.
> 
> - data reads do not include code reads (comparable to earlier Sandy
> Bridge tables)
> - data counts include speculative execution (except L1 write, dtlb, bpu)
> - remote node access includes both remote memory, remote cache, remote mmio.
> - prefetches are not included in the counts for consistency
> (different from Sandy Bridge, which includes prefetches in the remote node)
> 
> The events with additional caveats have references to the specification update.

> +		[ C(RESULT_ACCESS) ] = 0x81d0,	/* MEM_UOPS_RETIRED.ALL_LOADS, HSM30 */
> +		[ C(RESULT_ACCESS) ] = 0x82d0,	/* MEM_UOPS_RETIRED.ALL_STORES, HSM30 */
> +		[ C(RESULT_ACCESS) ] = 0x81d0,	/* MEM_UOPS_RETIRED.ALL_LOADS, HSM30 */
> +		[ C(RESULT_ACCESS) ] = 0x82d0,	/* MEM_UOPS_RETIRED.ALL_STORES, HSM30 */

So that 'HSM30' is code for the specification update?

You'll need to properly describe HSM30 at least once instead of using 
obfuscation.

Thanks,

	Ingo

Re: [PATCH 1/3] perf, x86: Add new cache events table for Haswell

[Peter Zijlstra]

On Mon, Mar 23, 2015 at 10:45:07AM +0100, Ingo Molnar wrote:
> 
> * Andi Kleen <andi@firstfloor.org> wrote:
> 
> > From: Andi Kleen <ak@linux.intel.com>
> > 
> > Haswell offcore events are quite different from Sandy Bridge.
> > Add a new table to handle Haswell properly.
> > 
> > Note that the offcore bits listed in the SDM are not quite correct
> > (this is currently being fixed). An uptodate list of bits is
> > in the patch.
> > 
> > The basic setup is similar to Sandy Bridge. The prefetch columns
> > have been removed, as prefetch counting is not very reliable
> > on Haswell. One L1 event that is not in the event list anymore
> > has been also removed.
> > 
> > - data reads do not include code reads (comparable to earlier Sandy
> > Bridge tables)
> > - data counts include speculative execution (except L1 write, dtlb, bpu)
> > - remote node access includes both remote memory, remote cache, remote mmio.
> > - prefetches are not included in the counts for consistency
> > (different from Sandy Bridge, which includes prefetches in the remote node)
> > 
> > The events with additional caveats have references to the specification update.
> 
> > +		[ C(RESULT_ACCESS) ] = 0x81d0,	/* MEM_UOPS_RETIRED.ALL_LOADS, HSM30 */
> > +		[ C(RESULT_ACCESS) ] = 0x82d0,	/* MEM_UOPS_RETIRED.ALL_STORES, HSM30 */
> > +		[ C(RESULT_ACCESS) ] = 0x81d0,	/* MEM_UOPS_RETIRED.ALL_LOADS, HSM30 */
> > +		[ C(RESULT_ACCESS) ] = 0x82d0,	/* MEM_UOPS_RETIRED.ALL_STORES, HSM30 */
> 
> So that 'HSM30' is code for the specification update?

Yep; found it in:
http://www.intel.com/content/dam/www/public/us/en/documents/specification-updates/4th-gen-core-family-mobile-specification-update.pdf

> You'll need to properly describe HSM30 at least once instead of using 
> obfuscation.


HSM30.
Problem:
Performance Monitor Counters May Produce Incorrect Results
When operating with SMT enabled, a memory at-retirement performance monitoring
event (from the list below) may be dropped or may increment an enabled event on the
corresponding counter with the same number on the physical core’s other thread rather
than the thread experiencing the event. Processors with SMT disabled in BIOS are not
affected by this erratum.
The list of affected memory at-retirement events is as follows:
MEM_UOP_RETIRED.LOADS
MEM_UOP_RETIRED.STORES
MEM_UOP_RETIRED.LOCK
MEM_UOP_RETIRED.SPLIT
MEM_UOP_RETIRED.STLB_MISS
MEM_LOAD_UOPS_RETIRED.HIT_LFB
MEM_LOAD_UOPS_RETIRED.L1_HIT
MEM_LOAD_UOPS_RETIRED.L2_HIT
MEM_LOAD_UOPS_RETIRED.L3_HIT
MEM_LOAD_UOPS_L3_HIT_RETIRED.XSNP_HIT
MEM_LOAD_UOPS_L3_HIT_RETIRED.XSNP_HITM
MEM_LOAD_UOPS_L3_HIT_RETIRED.XSNP_MISS
MEM_LOAD_UOPS_L3_HIT_RETIRED.XSNP_NONE
MEM_LOAD_UOPS_RETIRED.L3_MISS
MEM_LOAD_UOPS_L3_MISS_RETIRED.LOCAL_DRAM
MEM_LOAD_UOPS_L3_MISS_RETIRED.REMOTE_DRAM
MEM_LOAD_UOPS_RETIRED.L2_MISS
Implication:
Due to this erratum, certain performance monitoring event will produce unreliable
results during hyper-threaded operation.
Workaround:
None identified.
Status:
For the steppings affected, see the Summary Table of Changes.

Stephane is working on patches to address this. It affects multiple
generations.

Re: [PATCH 1/3] perf, x86: Add new cache events table for Haswell

[Ingo Molnar]

* Peter Zijlstra <peterz@infradead.org> wrote:

> On Mon, Mar 23, 2015 at 10:45:07AM +0100, Ingo Molnar wrote:
> > 
> > * Andi Kleen <andi@firstfloor.org> wrote:
> > 
> > > From: Andi Kleen <ak@linux.intel.com>
> > > 
> > > Haswell offcore events are quite different from Sandy Bridge.
> > > Add a new table to handle Haswell properly.
> > > 
> > > Note that the offcore bits listed in the SDM are not quite correct
> > > (this is currently being fixed). An uptodate list of bits is
> > > in the patch.
> > > 
> > > The basic setup is similar to Sandy Bridge. The prefetch columns
> > > have been removed, as prefetch counting is not very reliable
> > > on Haswell. One L1 event that is not in the event list anymore
> > > has been also removed.
> > > 
> > > - data reads do not include code reads (comparable to earlier Sandy
> > > Bridge tables)
> > > - data counts include speculative execution (except L1 write, dtlb, bpu)
> > > - remote node access includes both remote memory, remote cache, remote mmio.
> > > - prefetches are not included in the counts for consistency
> > > (different from Sandy Bridge, which includes prefetches in the remote node)
> > > 
> > > The events with additional caveats have references to the specification update.
> > 
> > > +		[ C(RESULT_ACCESS) ] = 0x81d0,	/* MEM_UOPS_RETIRED.ALL_LOADS, HSM30 */
> > > +		[ C(RESULT_ACCESS) ] = 0x82d0,	/* MEM_UOPS_RETIRED.ALL_STORES, HSM30 */
> > > +		[ C(RESULT_ACCESS) ] = 0x81d0,	/* MEM_UOPS_RETIRED.ALL_LOADS, HSM30 */
> > > +		[ C(RESULT_ACCESS) ] = 0x82d0,	/* MEM_UOPS_RETIRED.ALL_STORES, HSM30 */
> > 
> > So that 'HSM30' is code for the specification update?
> 
> Yep; found it in:
> http://www.intel.com/content/dam/www/public/us/en/documents/specification-updates/4th-gen-core-family-mobile-specification-update.pdf
> 
> > You'll need to properly describe HSM30 at least once instead of using 
> > obfuscation.
> 
> 
> HSM30.
> Problem:
> Performance Monitor Counters May Produce Incorrect Results
> When operating with SMT enabled, a memory at-retirement performance monitoring
> event (from the list below) may be dropped or may increment an enabled event on the
> corresponding counter with the same number on the physical core’s other thread rather
> than the thread experiencing the event. Processors with SMT disabled in BIOS are not
> affected by this erratum.
> The list of affected memory at-retirement events is as follows:
> MEM_UOP_RETIRED.LOADS
> MEM_UOP_RETIRED.STORES
> MEM_UOP_RETIRED.LOCK
> MEM_UOP_RETIRED.SPLIT
> MEM_UOP_RETIRED.STLB_MISS
> MEM_LOAD_UOPS_RETIRED.HIT_LFB
> MEM_LOAD_UOPS_RETIRED.L1_HIT
> MEM_LOAD_UOPS_RETIRED.L2_HIT
> MEM_LOAD_UOPS_RETIRED.L3_HIT
> MEM_LOAD_UOPS_L3_HIT_RETIRED.XSNP_HIT
> MEM_LOAD_UOPS_L3_HIT_RETIRED.XSNP_HITM
> MEM_LOAD_UOPS_L3_HIT_RETIRED.XSNP_MISS
> MEM_LOAD_UOPS_L3_HIT_RETIRED.XSNP_NONE
> MEM_LOAD_UOPS_RETIRED.L3_MISS
> MEM_LOAD_UOPS_L3_MISS_RETIRED.LOCAL_DRAM
> MEM_LOAD_UOPS_L3_MISS_RETIRED.REMOTE_DRAM
> MEM_LOAD_UOPS_RETIRED.L2_MISS
> Implication:
> Due to this erratum, certain performance monitoring event will produce unreliable
> results during hyper-threaded operation.
> Workaround:
> None identified.
> Status:
> For the steppings affected, see the Summary Table of Changes.
> 
> Stephane is working on patches to address this. It affects multiple
> generations.

Ok - then at minimum a minimal summary of 'HSM30' should be added to 
the code, so that people know what's happening, without having to dig 
out Intel documents (which might be gone years down the line).

Thanks,

	Ingo

[tip:perf/core] perf/x86/intel: Add new cache events table for Haswell

[tip-bot for Andi Kleen]

Commit-ID:  0f1b5ca240c65ed9533f193720f337bf24fb2f2f
Gitweb:     http://git.kernel.org/tip/0f1b5ca240c65ed9533f193720f337bf24fb2f2f
Author:     Andi Kleen <ak@linux.intel.com>
AuthorDate: Tue, 17 Feb 2015 18:18:04 -0800
Committer:  Ingo Molnar <mingo@kernel.org>
CommitDate: Fri, 27 Mar 2015 09:14:01 +0100

perf/x86/intel: Add new cache events table for Haswell

Haswell offcore events are quite different from Sandy Bridge.
Add a new table to handle Haswell properly.

Note that the offcore bits listed in the SDM are not quite correct
(this is currently being fixed). An uptodate list of bits is
in the patch.

The basic setup is similar to Sandy Bridge. The prefetch columns
have been removed, as prefetch counting is not very reliable
on Haswell. One L1 event that is not in the event list anymore
has been also removed.

- data reads do not include code reads (comparable to earlier Sandy Bridge tables)
- data counts include speculative execution (except L1 write, dtlb, bpu)
- remote node access includes both remote memory, remote cache, remote mmio.
- prefetches are not included in the counts for consistency
  (different from Sandy Bridge, which includes prefetches in the remote node)

Signed-off-by: Andi Kleen <ak@linux.intel.com>
[ Removed the HSM30 comments; we don't have them for SNB/IVB either. ]
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Link: http://lkml.kernel.org/r/1424225886-18652-1-git-send-email-andi@firstfloor.org
Signed-off-by: Ingo Molnar <mingo@kernel.org>
---
 arch/x86/kernel/cpu/perf_event_intel.c | 194 ++++++++++++++++++++++++++++++++-
 1 file changed, 192 insertions(+), 2 deletions(-)

diff --git a/arch/x86/kernel/cpu/perf_event_intel.c b/arch/x86/kernel/cpu/perf_event_intel.c
index 9f1dd18..5ef64bf 100644
--- a/arch/x86/kernel/cpu/perf_event_intel.c
+++ b/arch/x86/kernel/cpu/perf_event_intel.c
@@ -415,6 +415,196 @@ static __initconst const u64 snb_hw_cache_event_ids
 
 };
 
+/*
+ * Notes on the events:
+ * - data reads do not include code reads (comparable to earlier tables)
+ * - data counts include speculative execution (except L1 write, dtlb, bpu)
+ * - remote node access includes remote memory, remote cache, remote mmio.
+ * - prefetches are not included in the counts because they are not
+ *   reliably counted.
+ */
+
+#define HSW_DEMAND_DATA_RD		BIT_ULL(0)
+#define HSW_DEMAND_RFO			BIT_ULL(1)
+#define HSW_ANY_RESPONSE		BIT_ULL(16)
+#define HSW_SUPPLIER_NONE		BIT_ULL(17)
+#define HSW_L3_MISS_LOCAL_DRAM		BIT_ULL(22)
+#define HSW_L3_MISS_REMOTE_HOP0		BIT_ULL(27)
+#define HSW_L3_MISS_REMOTE_HOP1		BIT_ULL(28)
+#define HSW_L3_MISS_REMOTE_HOP2P	BIT_ULL(29)
+#define HSW_L3_MISS			(HSW_L3_MISS_LOCAL_DRAM| \
+					 HSW_L3_MISS_REMOTE_HOP0|HSW_L3_MISS_REMOTE_HOP1| \
+					 HSW_L3_MISS_REMOTE_HOP2P)
+#define HSW_SNOOP_NONE			BIT_ULL(31)
+#define HSW_SNOOP_NOT_NEEDED		BIT_ULL(32)
+#define HSW_SNOOP_MISS			BIT_ULL(33)
+#define HSW_SNOOP_HIT_NO_FWD		BIT_ULL(34)
+#define HSW_SNOOP_HIT_WITH_FWD		BIT_ULL(35)
+#define HSW_SNOOP_HITM			BIT_ULL(36)
+#define HSW_SNOOP_NON_DRAM		BIT_ULL(37)
+#define HSW_ANY_SNOOP			(HSW_SNOOP_NONE| \
+					 HSW_SNOOP_NOT_NEEDED|HSW_SNOOP_MISS| \
+					 HSW_SNOOP_HIT_NO_FWD|HSW_SNOOP_HIT_WITH_FWD| \
+					 HSW_SNOOP_HITM|HSW_SNOOP_NON_DRAM)
+#define HSW_SNOOP_DRAM			(HSW_ANY_SNOOP & ~HSW_SNOOP_NON_DRAM)
+#define HSW_DEMAND_READ			HSW_DEMAND_DATA_RD
+#define HSW_DEMAND_WRITE		HSW_DEMAND_RFO
+#define HSW_L3_MISS_REMOTE		(HSW_L3_MISS_REMOTE_HOP0|\
+					 HSW_L3_MISS_REMOTE_HOP1|HSW_L3_MISS_REMOTE_HOP2P)
+#define HSW_LLC_ACCESS			HSW_ANY_RESPONSE
+
+static __initconst const u64 hsw_hw_cache_event_ids
+				[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) ] = 0x81d0,	/* MEM_UOPS_RETIRED.ALL_LOADS */
+		[ C(RESULT_MISS)   ] = 0x151,	/* L1D.REPLACEMENT */
+	},
+	[ C(OP_WRITE) ] = {
+		[ C(RESULT_ACCESS) ] = 0x82d0,	/* MEM_UOPS_RETIRED.ALL_STORES */
+		[ C(RESULT_MISS)   ] = 0x0,
+	},
+	[ C(OP_PREFETCH) ] = {
+		[ C(RESULT_ACCESS) ] = 0x0,
+		[ C(RESULT_MISS)   ] = 0x0,
+	},
+ },
+ [ C(L1I ) ] = {
+	[ C(OP_READ) ] = {
+		[ C(RESULT_ACCESS) ] = 0x0,
+		[ C(RESULT_MISS)   ] = 0x280,	/* ICACHE.MISSES */
+	},
+	[ C(OP_WRITE) ] = {
+		[ C(RESULT_ACCESS) ] = -1,
+		[ C(RESULT_MISS)   ] = -1,
+	},
+	[ C(OP_PREFETCH) ] = {
+		[ C(RESULT_ACCESS) ] = 0x0,
+		[ C(RESULT_MISS)   ] = 0x0,
+	},
+ },
+ [ C(LL  ) ] = {
+	[ C(OP_READ) ] = {
+		[ C(RESULT_ACCESS) ] = 0x1b7,	/* OFFCORE_RESPONSE */
+		[ C(RESULT_MISS)   ] = 0x1b7,	/* OFFCORE_RESPONSE */
+	},
+	[ C(OP_WRITE) ] = {
+		[ C(RESULT_ACCESS) ] = 0x1b7,	/* OFFCORE_RESPONSE */
+		[ C(RESULT_MISS)   ] = 0x1b7,	/* OFFCORE_RESPONSE */
+	},
+	[ C(OP_PREFETCH) ] = {
+		[ C(RESULT_ACCESS) ] = 0x0,
+		[ C(RESULT_MISS)   ] = 0x0,
+	},
+ },
+ [ C(DTLB) ] = {
+	[ C(OP_READ) ] = {
+		[ C(RESULT_ACCESS) ] = 0x81d0,	/* MEM_UOPS_RETIRED.ALL_LOADS */
+		[ C(RESULT_MISS)   ] = 0x108,	/* DTLB_LOAD_MISSES.MISS_CAUSES_A_WALK */
+	},
+	[ C(OP_WRITE) ] = {
+		[ C(RESULT_ACCESS) ] = 0x82d0,	/* MEM_UOPS_RETIRED.ALL_STORES */
+		[ C(RESULT_MISS)   ] = 0x149,	/* DTLB_STORE_MISSES.MISS_CAUSES_A_WALK */
+	},
+	[ C(OP_PREFETCH) ] = {
+		[ C(RESULT_ACCESS) ] = 0x0,
+		[ C(RESULT_MISS)   ] = 0x0,
+	},
+ },
+ [ C(ITLB) ] = {
+	[ C(OP_READ) ] = {
+		[ C(RESULT_ACCESS) ] = 0x6085,	/* ITLB_MISSES.STLB_HIT */
+		[ C(RESULT_MISS)   ] = 0x185,	/* ITLB_MISSES.MISS_CAUSES_A_WALK */
+	},
+	[ C(OP_WRITE) ] = {
+		[ C(RESULT_ACCESS) ] = -1,
+		[ C(RESULT_MISS)   ] = -1,
+	},
+	[ C(OP_PREFETCH) ] = {
+		[ C(RESULT_ACCESS) ] = -1,
+		[ C(RESULT_MISS)   ] = -1,
+	},
+ },
+ [ C(BPU ) ] = {
+	[ C(OP_READ) ] = {
+		[ C(RESULT_ACCESS) ] = 0xc4,	/* BR_INST_RETIRED.ALL_BRANCHES */
+		[ C(RESULT_MISS)   ] = 0xc5,	/* BR_MISP_RETIRED.ALL_BRANCHES */
+	},
+	[ C(OP_WRITE) ] = {
+		[ C(RESULT_ACCESS) ] = -1,
+		[ C(RESULT_MISS)   ] = -1,
+	},
+	[ C(OP_PREFETCH) ] = {
+		[ C(RESULT_ACCESS) ] = -1,
+		[ C(RESULT_MISS)   ] = -1,
+	},
+ },
+ [ C(NODE) ] = {
+	[ C(OP_READ) ] = {
+		[ C(RESULT_ACCESS) ] = 0x1b7,	/* OFFCORE_RESPONSE */
+		[ C(RESULT_MISS)   ] = 0x1b7,	/* OFFCORE_RESPONSE */
+	},
+	[ C(OP_WRITE) ] = {
+		[ C(RESULT_ACCESS) ] = 0x1b7,	/* OFFCORE_RESPONSE */
+		[ C(RESULT_MISS)   ] = 0x1b7,	/* OFFCORE_RESPONSE */
+	},
+	[ C(OP_PREFETCH) ] = {
+		[ C(RESULT_ACCESS) ] = 0x0,
+		[ C(RESULT_MISS)   ] = 0x0,
+	},
+ },
+};
+
+static __initconst const u64 hsw_hw_cache_extra_regs
+				[PERF_COUNT_HW_CACHE_MAX]
+				[PERF_COUNT_HW_CACHE_OP_MAX]
+				[PERF_COUNT_HW_CACHE_RESULT_MAX] =
+{
+ [ C(LL  ) ] = {
+	[ C(OP_READ) ] = {
+		[ C(RESULT_ACCESS) ] = HSW_DEMAND_READ|
+				       HSW_LLC_ACCESS,
+		[ C(RESULT_MISS)   ] = HSW_DEMAND_READ|
+				       HSW_L3_MISS|HSW_ANY_SNOOP,
+	},
+	[ C(OP_WRITE) ] = {
+		[ C(RESULT_ACCESS) ] = HSW_DEMAND_WRITE|
+				       HSW_LLC_ACCESS,
+		[ C(RESULT_MISS)   ] = HSW_DEMAND_WRITE|
+				       HSW_L3_MISS|HSW_ANY_SNOOP,
+	},
+	[ C(OP_PREFETCH) ] = {
+		[ C(RESULT_ACCESS) ] = 0x0,
+		[ C(RESULT_MISS)   ] = 0x0,
+	},
+ },
+ [ C(NODE) ] = {
+	[ C(OP_READ) ] = {
+		[ C(RESULT_ACCESS) ] = HSW_DEMAND_READ|
+				       HSW_L3_MISS_LOCAL_DRAM|
+				       HSW_SNOOP_DRAM,
+		[ C(RESULT_MISS)   ] = HSW_DEMAND_READ|
+				       HSW_L3_MISS_REMOTE|
+				       HSW_SNOOP_DRAM,
+	},
+	[ C(OP_WRITE) ] = {
+		[ C(RESULT_ACCESS) ] = HSW_DEMAND_WRITE|
+				       HSW_L3_MISS_LOCAL_DRAM|
+				       HSW_SNOOP_DRAM,
+		[ C(RESULT_MISS)   ] = HSW_DEMAND_WRITE|
+				       HSW_L3_MISS_REMOTE|
+				       HSW_SNOOP_DRAM,
+	},
+	[ C(OP_PREFETCH) ] = {
+		[ C(RESULT_ACCESS) ] = 0x0,
+		[ C(RESULT_MISS)   ] = 0x0,
+	},
+ },
+};
+
 static __initconst const u64 westmere_hw_cache_event_ids
 				[PERF_COUNT_HW_CACHE_MAX]
 				[PERF_COUNT_HW_CACHE_OP_MAX]
@@ -2520,8 +2710,8 @@ __init int intel_pmu_init(void)
 	case 69: /* 22nm Haswell ULT */
 	case 70: /* 22nm Haswell + GT3e (Intel Iris Pro graphics) */
 		x86_pmu.late_ack = true;
-		memcpy(hw_cache_event_ids, snb_hw_cache_event_ids, sizeof(hw_cache_event_ids));
-		memcpy(hw_cache_extra_regs, snb_hw_cache_extra_regs, sizeof(hw_cache_extra_regs));
+		memcpy(hw_cache_event_ids, hsw_hw_cache_event_ids, sizeof(hw_cache_event_ids));
+		memcpy(hw_cache_extra_regs, hsw_hw_cache_extra_regs, sizeof(hw_cache_extra_regs));
 
 		intel_pmu_lbr_init_hsw();
 

[PATCH 3/5] perf, x86: Add Broadwell core support

[Andi Kleen]

From: Andi Kleen <ak@linux.intel.com>

Add Broadwell support for Broadwell Client to perf.
This is very similar to Haswell.  It uses a new cache event table,
because there were various changes there.

The constraint list has one new event that needs to be handled over Haswell.

The PEBS event list is the same, so we reuse Haswell's.

v2: Remove unnamed model numbers.
v3: Rename cache event list to hsw_*. Change names.
Signed-off-by: Andi Kleen <ak@linux.intel.com>
---
 arch/x86/kernel/cpu/perf_event_intel.c | 150 +++++++++++++++++++++++++++++++++
 1 file changed, 150 insertions(+)

diff --git a/arch/x86/kernel/cpu/perf_event_intel.c b/arch/x86/kernel/cpu/perf_event_intel.c
index 7c9f78e..61b5404 100644
--- a/arch/x86/kernel/cpu/perf_event_intel.c
+++ b/arch/x86/kernel/cpu/perf_event_intel.c
@@ -220,6 +220,15 @@ static struct event_constraint intel_hsw_event_constraints[] = {
 	EVENT_CONSTRAINT_END
 };
 
+struct event_constraint intel_bdw_event_constraints[] = {
+	FIXED_EVENT_CONSTRAINT(0x00c0, 0),	/* INST_RETIRED.ANY */
+	FIXED_EVENT_CONSTRAINT(0x003c, 1),	/* CPU_CLK_UNHALTED.CORE */
+	FIXED_EVENT_CONSTRAINT(0x0300, 2),	/* CPU_CLK_UNHALTED.REF */
+	INTEL_UEVENT_CONSTRAINT(0x148, 0x4),	/* L1D_PEND_MISS.PENDING */
+	INTEL_EVENT_CONSTRAINT(0xa3, 0x4),	/* CYCLE_ACTIVITY.* */
+	EVENT_CONSTRAINT_END
+};
+
 static u64 intel_pmu_event_map(int hw_event)
 {
 	return intel_perfmon_event_map[hw_event];
@@ -415,6 +424,126 @@ static __initconst const u64 snb_hw_cache_event_ids
 
 };
 
+static __initconst const u64 hsw_hw_cache_event_ids
+				[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) ] = 0x81d0, 	/* MEM_UOPS_RETIRED.ALL_LOADS */
+		[ C(RESULT_MISS)   ] = 0x151, 	/* L1D.REPLACEMENT */
+	},
+	[ C(OP_WRITE) ] = {
+		[ C(RESULT_ACCESS) ] = 0x82d0, 	/* MEM_UOPS_RETIRED.ALL_STORES */
+		[ C(RESULT_MISS)   ] = 0x0,
+	},
+	[ C(OP_PREFETCH) ] = {
+		[ C(RESULT_ACCESS) ] = 0x0,
+		[ C(RESULT_MISS)   ] = 0x0,
+	},
+ },
+ [ C(L1I ) ] = {
+	[ C(OP_READ) ] = {
+		[ C(RESULT_ACCESS) ] = 0x0,
+		[ C(RESULT_MISS)   ] = 0x280, 	/* ICACHE.MISSES */
+	},
+	[ C(OP_WRITE) ] = {
+		[ C(RESULT_ACCESS) ] = -1,
+		[ C(RESULT_MISS)   ] = -1,
+	},
+	[ C(OP_PREFETCH) ] = {
+		[ C(RESULT_ACCESS) ] = 0x0,
+		[ C(RESULT_MISS)   ] = 0x0,
+	},
+ },
+ [ C(LL  ) ] = {
+	[ C(OP_READ) ] = {
+		/* OFFCORE_RESPONSE:ALL_DATA_RD|ALL_CODE_RD */
+		[ C(RESULT_ACCESS) ] = 0x1b7,
+		/* OFFCORE_RESPONSE:ALL_DATA_RD|ALL_CODE_RD|SUPPLIER_NONE|
+                   L3_MISS|ANY_SNOOP */
+		[ C(RESULT_MISS)   ] = 0x1b7,
+	},
+	[ C(OP_WRITE) ] = {
+		[ C(RESULT_ACCESS) ] = 0x1b7, 	/* OFFCORE_RESPONSE:ALL_RFO */
+		/* OFFCORE_RESPONSE:ALL_RFO|SUPPLIER_NONE|L3_MISS|ANY_SNOOP */
+		[ C(RESULT_MISS)   ] = 0x1b7,
+	},
+	[ C(OP_PREFETCH) ] = {
+		[ C(RESULT_ACCESS) ] = 0x0,
+		[ C(RESULT_MISS)   ] = 0x0,
+	},
+ },
+ [ C(DTLB) ] = {
+	[ C(OP_READ) ] = {
+		[ C(RESULT_ACCESS) ] = 0x81d0, 	/* MEM_UOPS_RETIRED.ALL_LOADS */
+		[ C(RESULT_MISS)   ] = 0x108, 	/* DTLB_LOAD_MISSES.MISS_CAUSES_A_WALK */
+	},
+	[ C(OP_WRITE) ] = {
+		[ C(RESULT_ACCESS) ] = 0x82d0, 	/* MEM_UOPS_RETIRED.ALL_STORES */
+		[ C(RESULT_MISS)   ] = 0x149, 	/* DTLB_STORE_MISSES.MISS_CAUSES_A_WALK */
+	},
+	[ C(OP_PREFETCH) ] = {
+		[ C(RESULT_ACCESS) ] = 0x0,
+		[ C(RESULT_MISS)   ] = 0x0,
+	},
+ },
+ [ C(ITLB) ] = {
+	[ C(OP_READ) ] = {
+		[ C(RESULT_ACCESS) ] = 0x6085, 	/* ITLB_MISSES.STLB_HIT */
+		[ C(RESULT_MISS)   ] = 0x185, 	/* ITLB_MISSES.MISS_CAUSES_A_WALK */
+	},
+	[ C(OP_WRITE) ] = {
+		[ C(RESULT_ACCESS) ] = -1,
+		[ C(RESULT_MISS)   ] = -1,
+	},
+	[ C(OP_PREFETCH) ] = {
+		[ C(RESULT_ACCESS) ] = -1,
+		[ C(RESULT_MISS)   ] = -1,
+	},
+ },
+ [ C(BPU ) ] = {
+	[ C(OP_READ) ] = {
+		[ C(RESULT_ACCESS) ] = 0xc4, 	/* BR_INST_RETIRED.ALL_BRANCHES */
+		[ C(RESULT_MISS)   ] = 0xc5, 	/* BR_MISP_RETIRED.ALL_BRANCHES */
+	},
+	[ C(OP_WRITE) ] = {
+		[ C(RESULT_ACCESS) ] = -1,
+		[ C(RESULT_MISS)   ] = -1,
+	},
+	[ C(OP_PREFETCH) ] = {
+		[ C(RESULT_ACCESS) ] = -1,
+		[ C(RESULT_MISS)   ] = -1,
+	},
+ },
+};
+
+static __initconst const u64 hsw_hw_cache_extra_regs
+				[PERF_COUNT_HW_CACHE_MAX]
+				[PERF_COUNT_HW_CACHE_OP_MAX]
+				[PERF_COUNT_HW_CACHE_RESULT_MAX] =
+{
+ [ C(LL  ) ] = {
+	[ C(OP_READ) ] = {
+		/* OFFCORE_RESPONSE:ALL_DATA_RD|ALL_CODE_RD */
+		[ C(RESULT_ACCESS) ] = 0x2d5,
+		/* OFFCORE_RESPONSE:ALL_DATA_RD|ALL_CODE_RD|SUPPLIER_NONE|
+                   L3_MISS|ANY_SNOOP */
+		[ C(RESULT_MISS)   ] = 0x3fbc0202d5,
+	},
+	[ C(OP_WRITE) ] = {
+		[ C(RESULT_ACCESS) ] = 0x122, 	/* OFFCORE_RESPONSE:ALL_RFO */
+		/* OFFCORE_RESPONSE:ALL_RFO|SUPPLIER_NONE|L3_MISS|ANY_SNOOP */
+		[ C(RESULT_MISS)   ] = 0x3fbc020122,
+	},
+	[ C(OP_PREFETCH) ] = {
+		[ C(RESULT_ACCESS) ] = 0x0,
+		[ C(RESULT_MISS)   ] = 0x0,
+	},
+ },
+};
+
 static __initconst const u64 westmere_hw_cache_event_ids
 				[PERF_COUNT_HW_CACHE_MAX]
 				[PERF_COUNT_HW_CACHE_OP_MAX]
@@ -2565,6 +2694,27 @@ __init int intel_pmu_init(void)
 		pr_cont("Haswell events, ");
 		break;
 
+	case 61: /* 14nm Broadwell Core-M */
+		x86_pmu.late_ack = true;
+		memcpy(hw_cache_event_ids, hsw_hw_cache_event_ids, sizeof(hw_cache_event_ids));
+		memcpy(hw_cache_extra_regs, hsw_hw_cache_extra_regs, sizeof(hw_cache_extra_regs));
+
+		intel_pmu_lbr_init_snb();
+
+		x86_pmu.event_constraints = intel_bdw_event_constraints;
+		x86_pmu.pebs_constraints = intel_hsw_pebs_event_constraints;
+		x86_pmu.extra_regs = intel_snbep_extra_regs;
+		x86_pmu.pebs_aliases = intel_pebs_aliases_snb;
+		/* all extra regs are per-cpu when HT is on */
+		x86_pmu.er_flags |= ERF_HAS_RSP_1;
+		x86_pmu.er_flags |= ERF_NO_HT_SHARING;
+
+		x86_pmu.hw_config = hsw_hw_config;
+		x86_pmu.get_event_constraints = hsw_get_event_constraints;
+		x86_pmu.cpu_events = hsw_events_attrs;
+		pr_cont("Broadwell events, ");
+		break;
+
 	default:
 		switch (x86_pmu.version) {
 		case 1:
-- 
1.9.3

[tip:perf/core] perf/x86/intel: Add Broadwell core support

[tip-bot for Andi Kleen]

Commit-ID:  86a349a28b249bf4c3b396af3cc550760dedb456
Gitweb:     http://git.kernel.org/tip/86a349a28b249bf4c3b396af3cc550760dedb456
Author:     Andi Kleen <ak@linux.intel.com>
AuthorDate: Tue, 2 Sep 2014 11:44:13 -0700
Committer:  Ingo Molnar <mingo@kernel.org>
CommitDate: Wed, 24 Sep 2014 14:48:18 +0200

perf/x86/intel: Add Broadwell core support

Add Broadwell support for Broadwell Client to perf.  This is very
similar to Haswell.  It uses a new cache event table, because there
were various changes there.

The constraint list has one new event that needs to be handled over
Haswell.

The PEBS event list is the same, so we reuse Haswell's.

[fengguang.wu: make intel_bdw_event_constraints[] static]
Signed-off-by: Andi Kleen <ak@linux.intel.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Cc: eranian@google.com
Link: http://lkml.kernel.org/r/1409683455-29168-3-git-send-email-andi@firstfloor.org
Signed-off-by: Ingo Molnar <mingo@kernel.org>
---
 arch/x86/kernel/cpu/perf_event_intel.c | 150 +++++++++++++++++++++++++++++++++
 1 file changed, 150 insertions(+)

diff --git a/arch/x86/kernel/cpu/perf_event_intel.c b/arch/x86/kernel/cpu/perf_event_intel.c
index 7c9f78e..5671566 100644
--- a/arch/x86/kernel/cpu/perf_event_intel.c
+++ b/arch/x86/kernel/cpu/perf_event_intel.c
@@ -220,6 +220,15 @@ static struct event_constraint intel_hsw_event_constraints[] = {
 	EVENT_CONSTRAINT_END
 };
 
+static struct event_constraint intel_bdw_event_constraints[] = {
+	FIXED_EVENT_CONSTRAINT(0x00c0, 0),	/* INST_RETIRED.ANY */
+	FIXED_EVENT_CONSTRAINT(0x003c, 1),	/* CPU_CLK_UNHALTED.CORE */
+	FIXED_EVENT_CONSTRAINT(0x0300, 2),	/* CPU_CLK_UNHALTED.REF */
+	INTEL_UEVENT_CONSTRAINT(0x148, 0x4),	/* L1D_PEND_MISS.PENDING */
+	INTEL_EVENT_CONSTRAINT(0xa3, 0x4),	/* CYCLE_ACTIVITY.* */
+	EVENT_CONSTRAINT_END
+};
+
 static u64 intel_pmu_event_map(int hw_event)
 {
 	return intel_perfmon_event_map[hw_event];
@@ -415,6 +424,126 @@ static __initconst const u64 snb_hw_cache_event_ids
 
 };
 
+static __initconst const u64 hsw_hw_cache_event_ids
+				[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) ] = 0x81d0, 	/* MEM_UOPS_RETIRED.ALL_LOADS */
+		[ C(RESULT_MISS)   ] = 0x151, 	/* L1D.REPLACEMENT */
+	},
+	[ C(OP_WRITE) ] = {
+		[ C(RESULT_ACCESS) ] = 0x82d0, 	/* MEM_UOPS_RETIRED.ALL_STORES */
+		[ C(RESULT_MISS)   ] = 0x0,
+	},
+	[ C(OP_PREFETCH) ] = {
+		[ C(RESULT_ACCESS) ] = 0x0,
+		[ C(RESULT_MISS)   ] = 0x0,
+	},
+ },
+ [ C(L1I ) ] = {
+	[ C(OP_READ) ] = {
+		[ C(RESULT_ACCESS) ] = 0x0,
+		[ C(RESULT_MISS)   ] = 0x280, 	/* ICACHE.MISSES */
+	},
+	[ C(OP_WRITE) ] = {
+		[ C(RESULT_ACCESS) ] = -1,
+		[ C(RESULT_MISS)   ] = -1,
+	},
+	[ C(OP_PREFETCH) ] = {
+		[ C(RESULT_ACCESS) ] = 0x0,
+		[ C(RESULT_MISS)   ] = 0x0,
+	},
+ },
+ [ C(LL  ) ] = {
+	[ C(OP_READ) ] = {
+		/* OFFCORE_RESPONSE:ALL_DATA_RD|ALL_CODE_RD */
+		[ C(RESULT_ACCESS) ] = 0x1b7,
+		/* OFFCORE_RESPONSE:ALL_DATA_RD|ALL_CODE_RD|SUPPLIER_NONE|
+                   L3_MISS|ANY_SNOOP */
+		[ C(RESULT_MISS)   ] = 0x1b7,
+	},
+	[ C(OP_WRITE) ] = {
+		[ C(RESULT_ACCESS) ] = 0x1b7, 	/* OFFCORE_RESPONSE:ALL_RFO */
+		/* OFFCORE_RESPONSE:ALL_RFO|SUPPLIER_NONE|L3_MISS|ANY_SNOOP */
+		[ C(RESULT_MISS)   ] = 0x1b7,
+	},
+	[ C(OP_PREFETCH) ] = {
+		[ C(RESULT_ACCESS) ] = 0x0,
+		[ C(RESULT_MISS)   ] = 0x0,
+	},
+ },
+ [ C(DTLB) ] = {
+	[ C(OP_READ) ] = {
+		[ C(RESULT_ACCESS) ] = 0x81d0, 	/* MEM_UOPS_RETIRED.ALL_LOADS */
+		[ C(RESULT_MISS)   ] = 0x108, 	/* DTLB_LOAD_MISSES.MISS_CAUSES_A_WALK */
+	},
+	[ C(OP_WRITE) ] = {
+		[ C(RESULT_ACCESS) ] = 0x82d0, 	/* MEM_UOPS_RETIRED.ALL_STORES */
+		[ C(RESULT_MISS)   ] = 0x149, 	/* DTLB_STORE_MISSES.MISS_CAUSES_A_WALK */
+	},
+	[ C(OP_PREFETCH) ] = {
+		[ C(RESULT_ACCESS) ] = 0x0,
+		[ C(RESULT_MISS)   ] = 0x0,
+	},
+ },
+ [ C(ITLB) ] = {
+	[ C(OP_READ) ] = {
+		[ C(RESULT_ACCESS) ] = 0x6085, 	/* ITLB_MISSES.STLB_HIT */
+		[ C(RESULT_MISS)   ] = 0x185, 	/* ITLB_MISSES.MISS_CAUSES_A_WALK */
+	},
+	[ C(OP_WRITE) ] = {
+		[ C(RESULT_ACCESS) ] = -1,
+		[ C(RESULT_MISS)   ] = -1,
+	},
+	[ C(OP_PREFETCH) ] = {
+		[ C(RESULT_ACCESS) ] = -1,
+		[ C(RESULT_MISS)   ] = -1,
+	},
+ },
+ [ C(BPU ) ] = {
+	[ C(OP_READ) ] = {
+		[ C(RESULT_ACCESS) ] = 0xc4, 	/* BR_INST_RETIRED.ALL_BRANCHES */
+		[ C(RESULT_MISS)   ] = 0xc5, 	/* BR_MISP_RETIRED.ALL_BRANCHES */
+	},
+	[ C(OP_WRITE) ] = {
+		[ C(RESULT_ACCESS) ] = -1,
+		[ C(RESULT_MISS)   ] = -1,
+	},
+	[ C(OP_PREFETCH) ] = {
+		[ C(RESULT_ACCESS) ] = -1,
+		[ C(RESULT_MISS)   ] = -1,
+	},
+ },
+};
+
+static __initconst const u64 hsw_hw_cache_extra_regs
+				[PERF_COUNT_HW_CACHE_MAX]
+				[PERF_COUNT_HW_CACHE_OP_MAX]
+				[PERF_COUNT_HW_CACHE_RESULT_MAX] =
+{
+ [ C(LL  ) ] = {
+	[ C(OP_READ) ] = {
+		/* OFFCORE_RESPONSE:ALL_DATA_RD|ALL_CODE_RD */
+		[ C(RESULT_ACCESS) ] = 0x2d5,
+		/* OFFCORE_RESPONSE:ALL_DATA_RD|ALL_CODE_RD|SUPPLIER_NONE|
+                   L3_MISS|ANY_SNOOP */
+		[ C(RESULT_MISS)   ] = 0x3fbc0202d5ull,
+	},
+	[ C(OP_WRITE) ] = {
+		[ C(RESULT_ACCESS) ] = 0x122, 	/* OFFCORE_RESPONSE:ALL_RFO */
+		/* OFFCORE_RESPONSE:ALL_RFO|SUPPLIER_NONE|L3_MISS|ANY_SNOOP */
+		[ C(RESULT_MISS)   ] = 0x3fbc020122ull,
+	},
+	[ C(OP_PREFETCH) ] = {
+		[ C(RESULT_ACCESS) ] = 0x0,
+		[ C(RESULT_MISS)   ] = 0x0,
+	},
+ },
+};
+
 static __initconst const u64 westmere_hw_cache_event_ids
 				[PERF_COUNT_HW_CACHE_MAX]
 				[PERF_COUNT_HW_CACHE_OP_MAX]
@@ -2565,6 +2694,27 @@ __init int intel_pmu_init(void)
 		pr_cont("Haswell events, ");
 		break;
 
+	case 61: /* 14nm Broadwell Core-M */
+		x86_pmu.late_ack = true;
+		memcpy(hw_cache_event_ids, hsw_hw_cache_event_ids, sizeof(hw_cache_event_ids));
+		memcpy(hw_cache_extra_regs, hsw_hw_cache_extra_regs, sizeof(hw_cache_extra_regs));
+
+		intel_pmu_lbr_init_snb();
+
+		x86_pmu.event_constraints = intel_bdw_event_constraints;
+		x86_pmu.pebs_constraints = intel_hsw_pebs_event_constraints;
+		x86_pmu.extra_regs = intel_snbep_extra_regs;
+		x86_pmu.pebs_aliases = intel_pebs_aliases_snb;
+		/* all extra regs are per-cpu when HT is on */
+		x86_pmu.er_flags |= ERF_HAS_RSP_1;
+		x86_pmu.er_flags |= ERF_NO_HT_SHARING;
+
+		x86_pmu.hw_config = hsw_hw_config;
+		x86_pmu.get_event_constraints = hsw_get_event_constraints;
+		x86_pmu.cpu_events = hsw_events_attrs;
+		pr_cont("Broadwell events, ");
+		break;
+
 	default:
 		switch (x86_pmu.version) {
 		case 1: