* [PATCH v3 0/6] powerpc/perf: Export memory hierarchy level @ 2017-03-23 3:26 Madhavan Srinivasan 2017-03-23 3:26 ` [PATCH v3 1/6] powerpc/perf: Define big-endian version of perf_mem_data_src Madhavan Srinivasan ` (5 more replies) 0 siblings, 6 replies; 15+ messages in thread From: Madhavan Srinivasan @ 2017-03-23 3:26 UTC (permalink / raw) To: mpe; +Cc: linux-kernel, linuxppc-dev, Madhavan Srinivasan Power8/Power9 Perforence Monitoring Unit (PMU) supports different sampling modes (SM) such as Random Instruction Sampling (RIS), Random Load/Store Facility Sampling (RLS) and Random Branch Sampling (RBS). Sample mode RLS updates Sampled Instruction Event Register [SIER] bits with memory hierarchy information for a cache reload. Patchset exports the hierarchy information to the user via the perf_mem_data_src object from SIER. Patchset is a rebase of the work posted previously with minor updates to it. https://lkml.org/lkml/2015/6/11/92 Changelog v2: -Updated the commit messages -Fixed isa207_find_source() to consider all the possible sier[ldst] values. Changelog v1: - Fixed author-ship for the first patch and added suka's "Signed-off-by:". Madhavan Srinivasan (5): powerpc/perf: Export memory hierarchy info to user space powerpc/perf: Support to export MMCRA[TEC*] field to userspace powerpc/perf: Support to export SIERs bit in Power8 powerpc/perf: Support to export SIERs bit in Power9 powerpc/perf: Add Power8 mem_access event to sysfs Sukadev Bhattiprolu (1): powerpc/perf: Define big-endian version of perf_mem_data_src arch/powerpc/include/asm/perf_event_server.h | 3 + arch/powerpc/perf/core-book3s.c | 8 +++ arch/powerpc/perf/isa207-common.c | 88 ++++++++++++++++++++++++++++ arch/powerpc/perf/isa207-common.h | 26 +++++++- arch/powerpc/perf/power8-events-list.h | 6 ++ arch/powerpc/perf/power8-pmu.c | 4 ++ arch/powerpc/perf/power9-pmu.c | 2 + include/uapi/linux/perf_event.h | 16 +++++ tools/include/uapi/linux/perf_event.h | 16 +++++ 9 files changed, 168 insertions(+), 1 deletion(-) -- 2.7.4 ^ permalink raw reply [flat|nested] 15+ messages in thread
* [PATCH v3 1/6] powerpc/perf: Define big-endian version of perf_mem_data_src 2017-03-23 3:26 [PATCH v3 0/6] powerpc/perf: Export memory hierarchy level Madhavan Srinivasan @ 2017-03-23 3:26 ` Madhavan Srinivasan 2017-03-23 3:26 ` [PATCH v3 2/6] powerpc/perf: Export memory hierarchy info to user space Madhavan Srinivasan ` (4 subsequent siblings) 5 siblings, 0 replies; 15+ messages in thread From: Madhavan Srinivasan @ 2017-03-23 3:26 UTC (permalink / raw) To: mpe Cc: linux-kernel, linuxppc-dev, Sukadev Bhattiprolu, Peter Zijlstra, Ingo Molnar, Arnaldo Carvalho de Melo, Alexander Shishkin, Wang Nan, Alexei Starovoitov, Stephane Eranian, Madhavan Srinivasan From: Sukadev Bhattiprolu <sukadev@linux.vnet.ibm.com> perf_mem_data_src is an union that is initialized via the ->val field and accessed via the bitmap fields. For this to work on big endian platforms (Which is broken now), we also need a big-endian represenation of perf_mem_data_src. i.e, in a big endian system, if user request PERF_SAMPLE_DATA_SRC (perf report -d), will get the default value from perf_sample_data_init(), which is PERF_MEM_NA. Value for PERF_MEM_NA is constructed using shifts: /* TLB access */ #define PERF_MEM_TLB_NA 0x01 /* not available */ ... #define PERF_MEM_TLB_SHIFT 26 #define PERF_MEM_S(a, s) \ (((__u64)PERF_MEM_##a##_##s) << PERF_MEM_##a##_SHIFT) #define PERF_MEM_NA (PERF_MEM_S(OP, NA) |\ PERF_MEM_S(LVL, NA) |\ PERF_MEM_S(SNOOP, NA) |\ PERF_MEM_S(LOCK, NA) |\ PERF_MEM_S(TLB, NA)) Which works out as: ((0x01 << 0) | (0x01 << 5) | (0x01 << 19) | (0x01 << 24) | (0x01 << 26)) Which means the PERF_MEM_NA value comes out of the kernel as 0x5080021 in CPU endian. But then in the perf tool, the code uses the bitfields to inspect the value, and currently the bitfields are defined using little endian ordering. So eg. in perf_mem__tlb_scnprintf() we see: data_src->val = 0x5080021 op = 0x0 lvl = 0x0 snoop = 0x0 lock = 0x0 dtlb = 0x0 rsvd = 0x5080021 Patch does a minimal fix of adding big endian definition of the bitfields to match the values that are already exported by the kernel on big endian. And it makes no change on little endian. Cc: Peter Zijlstra <peterz@infradead.org> Cc: Ingo Molnar <mingo@redhat.com> Cc: Arnaldo Carvalho de Melo <acme@kernel.org> Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com> Cc: Wang Nan <wangnan0@huawei.com> Cc: Alexei Starovoitov <ast@kernel.org> Cc: Stephane Eranian <eranian@google.com> Signed-off-by: Sukadev Bhattiprolu <sukadev@linux.vnet.ibm.com> Signed-off-by: Madhavan Srinivasan <maddy@linux.vnet.ibm.com> --- Changelog v2: -Added Michael Ellerman's explanation to comiit message. Changelog v1: -Fixed author-ship and added suka's "Signed-off-by:". include/uapi/linux/perf_event.h | 16 ++++++++++++++++ tools/include/uapi/linux/perf_event.h | 16 ++++++++++++++++ 2 files changed, 32 insertions(+) diff --git a/include/uapi/linux/perf_event.h b/include/uapi/linux/perf_event.h index c66a485a24ac..c4af1159a200 100644 --- a/include/uapi/linux/perf_event.h +++ b/include/uapi/linux/perf_event.h @@ -891,6 +891,7 @@ enum perf_callchain_context { #define PERF_FLAG_PID_CGROUP (1UL << 2) /* pid=cgroup id, per-cpu mode only */ #define PERF_FLAG_FD_CLOEXEC (1UL << 3) /* O_CLOEXEC */ +#if defined(__LITTLE_ENDIAN_BITFIELD) union perf_mem_data_src { __u64 val; struct { @@ -902,6 +903,21 @@ union perf_mem_data_src { mem_rsvd:31; }; }; +#elif defined(__BIG_ENDIAN_BITFIELD) +union perf_mem_data_src { + __u64 val; + struct { + __u64 mem_rsvd:31, + mem_dtlb:7, /* tlb access */ + mem_lock:2, /* lock instr */ + mem_snoop:5, /* snoop mode */ + mem_lvl:14, /* memory hierarchy level */ + mem_op:5; /* type of opcode */ + }; +}; +#else +#error "Unknown endianness" +#endif /* type of opcode (load/store/prefetch,code) */ #define PERF_MEM_OP_NA 0x01 /* not available */ diff --git a/tools/include/uapi/linux/perf_event.h b/tools/include/uapi/linux/perf_event.h index c66a485a24ac..c4af1159a200 100644 --- a/tools/include/uapi/linux/perf_event.h +++ b/tools/include/uapi/linux/perf_event.h @@ -891,6 +891,7 @@ enum perf_callchain_context { #define PERF_FLAG_PID_CGROUP (1UL << 2) /* pid=cgroup id, per-cpu mode only */ #define PERF_FLAG_FD_CLOEXEC (1UL << 3) /* O_CLOEXEC */ +#if defined(__LITTLE_ENDIAN_BITFIELD) union perf_mem_data_src { __u64 val; struct { @@ -902,6 +903,21 @@ union perf_mem_data_src { mem_rsvd:31; }; }; +#elif defined(__BIG_ENDIAN_BITFIELD) +union perf_mem_data_src { + __u64 val; + struct { + __u64 mem_rsvd:31, + mem_dtlb:7, /* tlb access */ + mem_lock:2, /* lock instr */ + mem_snoop:5, /* snoop mode */ + mem_lvl:14, /* memory hierarchy level */ + mem_op:5; /* type of opcode */ + }; +}; +#else +#error "Unknown endianness" +#endif /* type of opcode (load/store/prefetch,code) */ #define PERF_MEM_OP_NA 0x01 /* not available */ -- 2.7.4 ^ permalink raw reply related [flat|nested] 15+ messages in thread
* [PATCH v3 2/6] powerpc/perf: Export memory hierarchy info to user space 2017-03-23 3:26 [PATCH v3 0/6] powerpc/perf: Export memory hierarchy level Madhavan Srinivasan 2017-03-23 3:26 ` [PATCH v3 1/6] powerpc/perf: Define big-endian version of perf_mem_data_src Madhavan Srinivasan @ 2017-03-23 3:26 ` Madhavan Srinivasan 2017-03-23 3:26 ` [PATCH v3 3/6] powerpc/perf: Support to export MMCRA[TEC*] field to userspace Madhavan Srinivasan ` (3 subsequent siblings) 5 siblings, 0 replies; 15+ messages in thread From: Madhavan Srinivasan @ 2017-03-23 3:26 UTC (permalink / raw) To: mpe Cc: linux-kernel, linuxppc-dev, Madhavan Srinivasan, Benjamin Herrenschmidt, Paul Mackerras, Thomas Gleixner, Sebastian Andrzej Siewior, Anna-Maria Gleixner, Daniel Axtens, Peter Zijlstra, Ingo Molnar, Arnaldo Carvalho de Melo, Alexander Shishkin, Wang Nan, Alexei Starovoitov, Stephane Eranian, Sukadev Bhattiprolu The LDST field and DATA_SRC in SIER identifies the memory hierarchy level (eg: L1, L2 etc), from which a data-cache miss for a marked instruction was satisfied. Use the 'perf_mem_data_src' object to export this hierarchy level to user space. Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org> Cc: Paul Mackerras <paulus@samba.org> Cc: Thomas Gleixner <tglx@linutronix.de> Cc: Sebastian Andrzej Siewior <bigeasy@linutronix.de> Cc: Anna-Maria Gleixner <anna-maria@linutronix.de> Cc: Daniel Axtens <dja@axtens.net> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Ingo Molnar <mingo@redhat.com> Cc: Arnaldo Carvalho de Melo <acme@kernel.org> Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com> Cc: Wang Nan <wangnan0@huawei.com> Cc: Alexei Starovoitov <ast@kernel.org> Cc: Stephane Eranian <eranian@google.com> Signed-off-by: Sukadev Bhattiprolu <sukadev@linux.vnet.ibm.com> Signed-off-by: Madhavan Srinivasan <maddy@linux.vnet.ibm.com> --- Changelog v2: - Fixed isa207_find_source() to consider all the possible sier[ldst] values. arch/powerpc/include/asm/perf_event_server.h | 2 + arch/powerpc/perf/core-book3s.c | 4 ++ arch/powerpc/perf/isa207-common.c | 80 ++++++++++++++++++++++++++++ arch/powerpc/perf/isa207-common.h | 16 +++++- 4 files changed, 101 insertions(+), 1 deletion(-) diff --git a/arch/powerpc/include/asm/perf_event_server.h b/arch/powerpc/include/asm/perf_event_server.h index ae0a23091a9b..446cdcd9b7f5 100644 --- a/arch/powerpc/include/asm/perf_event_server.h +++ b/arch/powerpc/include/asm/perf_event_server.h @@ -38,6 +38,8 @@ struct power_pmu { unsigned long *valp); int (*get_alternatives)(u64 event_id, unsigned int flags, u64 alt[]); + void (*get_mem_data_src)(union perf_mem_data_src *dsrc, + u32 flags, struct pt_regs *regs); u64 (*bhrb_filter_map)(u64 branch_sample_type); void (*config_bhrb)(u64 pmu_bhrb_filter); void (*disable_pmc)(unsigned int pmc, unsigned long mmcr[]); diff --git a/arch/powerpc/perf/core-book3s.c b/arch/powerpc/perf/core-book3s.c index 2ff13249f87a..e241ebebab6f 100644 --- a/arch/powerpc/perf/core-book3s.c +++ b/arch/powerpc/perf/core-book3s.c @@ -2049,6 +2049,10 @@ static void record_and_restart(struct perf_event *event, unsigned long val, data.br_stack = &cpuhw->bhrb_stack; } + if (event->attr.sample_type & PERF_SAMPLE_DATA_SRC && + ppmu->get_mem_data_src) + ppmu->get_mem_data_src(&data.data_src, ppmu->flags, regs); + if (perf_event_overflow(event, &data, regs)) power_pmu_stop(event, 0); } diff --git a/arch/powerpc/perf/isa207-common.c b/arch/powerpc/perf/isa207-common.c index cd951fd231c4..41cc053ee692 100644 --- a/arch/powerpc/perf/isa207-common.c +++ b/arch/powerpc/perf/isa207-common.c @@ -148,6 +148,86 @@ static bool is_thresh_cmp_valid(u64 event) return true; } +static inline u64 isa207_find_source(u64 idx, u32 sub_idx) +{ + u64 ret = PERF_MEM_NA; + + switch(idx) { + case 0: + /* Nothing to do */ + break; + case 1: + ret = PLH(LVL, L1); + break; + case 2: + ret = PLH(LVL, L2); + break; + case 3: + ret = PLH(LVL, L3); + break; + case 4: + if (sub_idx <= 1) + ret = PLH(LVL, LOC_RAM); + else if (sub_idx > 1 && sub_idx <= 2) + ret = PLH(LVL, REM_RAM1); + else + ret = PLH(LVL, REM_RAM2); + ret |= P(SNOOP, HIT); + break; + case 5: + ret = PLH(LVL, REM_CCE1); + if ((sub_idx == 0) || (sub_idx == 2) || (sub_idx == 4)) + ret |= P(SNOOP, HIT); + else if ((sub_idx == 1) || (sub_idx == 3) || (sub_idx == 5)) + ret |= P(SNOOP, HITM); + break; + case 6: + ret = PLH(LVL, REM_CCE2); + if ((sub_idx == 0) || (sub_idx == 2)) + ret |= P(SNOOP, HIT); + else if ((sub_idx == 1) || (sub_idx == 3)) + ret |= P(SNOOP, HITM); + break; + case 7: + ret = PSM(LVL, L1); + break; + } + + return ret; +} + +static inline bool is_load_store_inst(u64 sier) +{ + u64 val; + val = (sier & ISA207_SIER_TYPE_MASK) >> ISA207_SIER_TYPE_SHIFT; + + /* 1 = load, 2 = store */ + return val == 1 || val == 2; +} + +void isa207_get_mem_data_src(union perf_mem_data_src *dsrc, u32 flags, + struct pt_regs *regs) +{ + u64 idx; + u32 sub_idx; + u64 sier; + + /* Skip if no SIER support */ + if (!(flags & PPMU_HAS_SIER)) { + dsrc->val = 0; + return; + } + + sier = mfspr(SPRN_SIER); + if (is_load_store_inst(sier)) { + idx = (sier & ISA207_SIER_LDST_MASK) >> ISA207_SIER_LDST_SHIFT; + sub_idx = (sier & ISA207_SIER_DATA_SRC_MASK) >> ISA207_SIER_DATA_SRC_SHIFT; + + dsrc->val = isa207_find_source(idx, sub_idx); + } +} + + int isa207_get_constraint(u64 event, unsigned long *maskp, unsigned long *valp) { unsigned int unit, pmc, cache, ebb; diff --git a/arch/powerpc/perf/isa207-common.h b/arch/powerpc/perf/isa207-common.h index 899210f14ee4..592aa0917cf3 100644 --- a/arch/powerpc/perf/isa207-common.h +++ b/arch/powerpc/perf/isa207-common.h @@ -260,6 +260,19 @@ #define MAX_ALT 2 #define MAX_PMU_COUNTERS 6 +#define ISA207_SIER_TYPE_SHIFT 15 +#define ISA207_SIER_TYPE_MASK (0x7ull << ISA207_SIER_TYPE_SHIFT) + +#define ISA207_SIER_LDST_SHIFT 1 +#define ISA207_SIER_LDST_MASK (0x7ull << ISA207_SIER_LDST_SHIFT) + +#define ISA207_SIER_DATA_SRC_SHIFT 53 +#define ISA207_SIER_DATA_SRC_MASK (0x7ull << ISA207_SIER_DATA_SRC_SHIFT) + +#define P(a, b) PERF_MEM_S(a, b) +#define PLH(a, b) (P(OP, LOAD) | P(LVL, HIT) | P(a, b)) +#define PSM(a, b) (P(OP, STORE) | P(LVL, MISS) | P(a, b)) + int isa207_get_constraint(u64 event, unsigned long *maskp, unsigned long *valp); int isa207_compute_mmcr(u64 event[], int n_ev, unsigned int hwc[], unsigned long mmcr[], @@ -267,6 +280,7 @@ int isa207_compute_mmcr(u64 event[], int n_ev, void isa207_disable_pmc(unsigned int pmc, unsigned long mmcr[]); int isa207_get_alternatives(u64 event, u64 alt[], const unsigned int ev_alt[][MAX_ALT], int size); - +void isa207_get_mem_data_src(union perf_mem_data_src *dsrc, u32 flags, + struct pt_regs *regs); #endif -- 2.7.4 ^ permalink raw reply related [flat|nested] 15+ messages in thread
* [PATCH v3 3/6] powerpc/perf: Support to export MMCRA[TEC*] field to userspace 2017-03-23 3:26 [PATCH v3 0/6] powerpc/perf: Export memory hierarchy level Madhavan Srinivasan 2017-03-23 3:26 ` [PATCH v3 1/6] powerpc/perf: Define big-endian version of perf_mem_data_src Madhavan Srinivasan 2017-03-23 3:26 ` [PATCH v3 2/6] powerpc/perf: Export memory hierarchy info to user space Madhavan Srinivasan @ 2017-03-23 3:26 ` Madhavan Srinivasan 2017-03-23 3:26 ` [PATCH v3 4/6] powerpc/perf: Support to export SIERs bit in Power8 Madhavan Srinivasan ` (2 subsequent siblings) 5 siblings, 0 replies; 15+ messages in thread From: Madhavan Srinivasan @ 2017-03-23 3:26 UTC (permalink / raw) To: mpe Cc: linux-kernel, linuxppc-dev, Madhavan Srinivasan, Benjamin Herrenschmidt, Paul Mackerras, Thomas Gleixner, Sebastian Andrzej Siewior, Anna-Maria Gleixner, Daniel Axtens, Sukadev Bhattiprolu, Peter Zijlstra, Ingo Molnar, Arnaldo Carvalho de Melo, Alexander Shishkin, Wang Nan, Alexei Starovoitov, Stephane Eranian Threshold feature when used with MMCRA [Threshold Event Counter Event], MMCRA[Threshold Start event] and MMCRA[Threshold End event] will update MMCRA[Threashold Event Counter Exponent] and MMCRA[Threshold Event Counter Multiplier] with the corresponding threshold event count values. Patch to export MMCRA[TECX/TECM] to userspace in 'weight' field of struct perf_sample_data. Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org> Cc: Paul Mackerras <paulus@samba.org> Cc: Thomas Gleixner <tglx@linutronix.de> Cc: Sebastian Andrzej Siewior <bigeasy@linutronix.de> Cc: Anna-Maria Gleixner <anna-maria@linutronix.de> Cc: Daniel Axtens <dja@axtens.net> Cc: Sukadev Bhattiprolu <sukadev@linux.vnet.ibm.com> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Ingo Molnar <mingo@redhat.com> Cc: Arnaldo Carvalho de Melo <acme@kernel.org> Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com> Cc: Wang Nan <wangnan0@huawei.com> Cc: Alexei Starovoitov <ast@kernel.org> Cc: Stephane Eranian <eranian@google.com> Signed-off-by: Madhavan Srinivasan <maddy@linux.vnet.ibm.com> --- arch/powerpc/include/asm/perf_event_server.h | 1 + arch/powerpc/perf/core-book3s.c | 4 ++++ arch/powerpc/perf/isa207-common.c | 8 ++++++++ arch/powerpc/perf/isa207-common.h | 10 ++++++++++ 4 files changed, 23 insertions(+) diff --git a/arch/powerpc/include/asm/perf_event_server.h b/arch/powerpc/include/asm/perf_event_server.h index 446cdcd9b7f5..723bf48e7494 100644 --- a/arch/powerpc/include/asm/perf_event_server.h +++ b/arch/powerpc/include/asm/perf_event_server.h @@ -40,6 +40,7 @@ struct power_pmu { u64 alt[]); void (*get_mem_data_src)(union perf_mem_data_src *dsrc, u32 flags, struct pt_regs *regs); + void (*get_mem_weight)(u64 *weight); u64 (*bhrb_filter_map)(u64 branch_sample_type); void (*config_bhrb)(u64 pmu_bhrb_filter); void (*disable_pmc)(unsigned int pmc, unsigned long mmcr[]); diff --git a/arch/powerpc/perf/core-book3s.c b/arch/powerpc/perf/core-book3s.c index e241ebebab6f..6c2d4168daec 100644 --- a/arch/powerpc/perf/core-book3s.c +++ b/arch/powerpc/perf/core-book3s.c @@ -2053,6 +2053,10 @@ static void record_and_restart(struct perf_event *event, unsigned long val, ppmu->get_mem_data_src) ppmu->get_mem_data_src(&data.data_src, ppmu->flags, regs); + if (event->attr.sample_type & PERF_SAMPLE_WEIGHT && + ppmu->get_mem_weight) + ppmu->get_mem_weight(&data.weight); + if (perf_event_overflow(event, &data, regs)) power_pmu_stop(event, 0); } diff --git a/arch/powerpc/perf/isa207-common.c b/arch/powerpc/perf/isa207-common.c index 41cc053ee692..292f6a242bb4 100644 --- a/arch/powerpc/perf/isa207-common.c +++ b/arch/powerpc/perf/isa207-common.c @@ -227,6 +227,14 @@ void isa207_get_mem_data_src(union perf_mem_data_src *dsrc, u32 flags, } } +void isa207_get_mem_weight(u64 *weight) +{ + u64 mmcra = mfspr(SPRN_MMCRA); + u64 exp = MMCRA_THR_CTR_EXP(mmcra); + u64 mantissa = MMCRA_THR_CTR_MANT(mmcra); + + *weight = mantissa << (2 * exp); +} int isa207_get_constraint(u64 event, unsigned long *maskp, unsigned long *valp) { diff --git a/arch/powerpc/perf/isa207-common.h b/arch/powerpc/perf/isa207-common.h index 592aa0917cf3..23e0516df4a4 100644 --- a/arch/powerpc/perf/isa207-common.h +++ b/arch/powerpc/perf/isa207-common.h @@ -248,6 +248,15 @@ #define MMCRA_SDAR_MODE_TLB (1ull << MMCRA_SDAR_MODE_SHIFT) #define MMCRA_SDAR_MODE_NO_UPDATES ~(0x3ull << MMCRA_SDAR_MODE_SHIFT) #define MMCRA_IFM_SHIFT 30 +#define MMCRA_THR_CTR_MANT_SHIFT 19 +#define MMCRA_THR_CTR_MANT_MASK 0x7Ful +#define MMCRA_THR_CTR_MANT(v) (((v) >> MMCRA_THR_CTR_MANT_SHIFT) &\ + MMCRA_THR_CTR_MANT_MASK) + +#define MMCRA_THR_CTR_EXP_SHIFT 27 +#define MMCRA_THR_CTR_EXP_MASK 0x7ul +#define MMCRA_THR_CTR_EXP(v) (((v) >> MMCRA_THR_CTR_EXP_SHIFT) &\ + MMCRA_THR_CTR_EXP_MASK) /* MMCR1 Threshold Compare bit constant for power9 */ #define p9_MMCRA_THR_CMP_SHIFT 45 @@ -282,5 +291,6 @@ int isa207_get_alternatives(u64 event, u64 alt[], const unsigned int ev_alt[][MAX_ALT], int size); void isa207_get_mem_data_src(union perf_mem_data_src *dsrc, u32 flags, struct pt_regs *regs); +void isa207_get_mem_weight(u64 *weight); #endif -- 2.7.4 ^ permalink raw reply related [flat|nested] 15+ messages in thread
* [PATCH v3 4/6] powerpc/perf: Support to export SIERs bit in Power8 2017-03-23 3:26 [PATCH v3 0/6] powerpc/perf: Export memory hierarchy level Madhavan Srinivasan ` (2 preceding siblings ...) 2017-03-23 3:26 ` [PATCH v3 3/6] powerpc/perf: Support to export MMCRA[TEC*] field to userspace Madhavan Srinivasan @ 2017-03-23 3:26 ` Madhavan Srinivasan 2017-03-23 3:26 ` [PATCH v3 5/6] powerpc/perf: Support to export SIERs bit in Power9 Madhavan Srinivasan 2017-03-23 3:26 ` [PATCH v3 6/6] powerpc/perf: Add Power8 mem_access event to sysfs Madhavan Srinivasan 5 siblings, 0 replies; 15+ messages in thread From: Madhavan Srinivasan @ 2017-03-23 3:26 UTC (permalink / raw) To: mpe Cc: linux-kernel, linuxppc-dev, Madhavan Srinivasan, Benjamin Herrenschmidt, Paul Mackerras, Daniel Axtens, Andrew Donnellan, Peter Zijlstra, Ingo Molnar, Arnaldo Carvalho de Melo, Alexander Shishkin, Wang Nan, Alexei Starovoitov, Stephane Eranian, Sukadev Bhattiprolu Patch to export SIER bits to userspace via perf_mem_data_src and perf_sample_data struct. Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org> Cc: Paul Mackerras <paulus@samba.org> Cc: Daniel Axtens <dja@axtens.net> Cc: Andrew Donnellan <andrew.donnellan@au1.ibm.com> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Ingo Molnar <mingo@redhat.com> Cc: Arnaldo Carvalho de Melo <acme@kernel.org> Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com> Cc: Wang Nan <wangnan0@huawei.com> Cc: Alexei Starovoitov <ast@kernel.org> Cc: Stephane Eranian <eranian@google.com> Signed-off-by: Sukadev Bhattiprolu <sukadev@linux.vnet.ibm.com> Signed-off-by: Madhavan Srinivasan <maddy@linux.vnet.ibm.com> --- arch/powerpc/perf/power8-pmu.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/arch/powerpc/perf/power8-pmu.c b/arch/powerpc/perf/power8-pmu.c index ce15b19a7962..932d7536f0eb 100644 --- a/arch/powerpc/perf/power8-pmu.c +++ b/arch/powerpc/perf/power8-pmu.c @@ -325,6 +325,8 @@ static struct power_pmu power8_pmu = { .bhrb_filter_map = power8_bhrb_filter_map, .get_constraint = isa207_get_constraint, .get_alternatives = power8_get_alternatives, + .get_mem_data_src = isa207_get_mem_data_src, + .get_mem_weight = isa207_get_mem_weight, .disable_pmc = isa207_disable_pmc, .flags = PPMU_HAS_SIER | PPMU_ARCH_207S, .n_generic = ARRAY_SIZE(power8_generic_events), -- 2.7.4 ^ permalink raw reply related [flat|nested] 15+ messages in thread
* [PATCH v3 5/6] powerpc/perf: Support to export SIERs bit in Power9 2017-03-23 3:26 [PATCH v3 0/6] powerpc/perf: Export memory hierarchy level Madhavan Srinivasan ` (3 preceding siblings ...) 2017-03-23 3:26 ` [PATCH v3 4/6] powerpc/perf: Support to export SIERs bit in Power8 Madhavan Srinivasan @ 2017-03-23 3:26 ` Madhavan Srinivasan 2017-03-23 3:26 ` [PATCH v3 6/6] powerpc/perf: Add Power8 mem_access event to sysfs Madhavan Srinivasan 5 siblings, 0 replies; 15+ messages in thread From: Madhavan Srinivasan @ 2017-03-23 3:26 UTC (permalink / raw) To: mpe Cc: linux-kernel, linuxppc-dev, Madhavan Srinivasan, Benjamin Herrenschmidt, Paul Mackerras, Sukadev Bhattiprolu, Daniel Axtens, Andrew Donnellan, Peter Zijlstra, Ingo Molnar, Arnaldo Carvalho de Melo, Alexander Shishkin, Wang Nan, Alexei Starovoitov, Stephane Eranian Patch to export SIER bits to userspace via perf_mem_data_src and perf_sample_data struct. Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org> Cc: Paul Mackerras <paulus@samba.org> Cc: Sukadev Bhattiprolu <sukadev@linux.vnet.ibm.com> Cc: Daniel Axtens <dja@axtens.net> Cc: Andrew Donnellan <andrew.donnellan@au1.ibm.com> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Ingo Molnar <mingo@redhat.com> Cc: Arnaldo Carvalho de Melo <acme@kernel.org> Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com> Cc: Wang Nan <wangnan0@huawei.com> Cc: Alexei Starovoitov <ast@kernel.org> Cc: Stephane Eranian <eranian@google.com> Signed-off-by: Madhavan Srinivasan <maddy@linux.vnet.ibm.com> --- arch/powerpc/perf/power9-pmu.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/arch/powerpc/perf/power9-pmu.c b/arch/powerpc/perf/power9-pmu.c index 7f6582708e06..018f8e90ac35 100644 --- a/arch/powerpc/perf/power9-pmu.c +++ b/arch/powerpc/perf/power9-pmu.c @@ -427,6 +427,8 @@ static struct power_pmu power9_pmu = { .bhrb_filter_map = power9_bhrb_filter_map, .get_constraint = isa207_get_constraint, .get_alternatives = power9_get_alternatives, + .get_mem_data_src = isa207_get_mem_data_src, + .get_mem_weight = isa207_get_mem_weight, .disable_pmc = isa207_disable_pmc, .flags = PPMU_HAS_SIER | PPMU_ARCH_207S, .n_generic = ARRAY_SIZE(power9_generic_events), -- 2.7.4 ^ permalink raw reply related [flat|nested] 15+ messages in thread
* [PATCH v3 6/6] powerpc/perf: Add Power8 mem_access event to sysfs 2017-03-23 3:26 [PATCH v3 0/6] powerpc/perf: Export memory hierarchy level Madhavan Srinivasan ` (4 preceding siblings ...) 2017-03-23 3:26 ` [PATCH v3 5/6] powerpc/perf: Support to export SIERs bit in Power9 Madhavan Srinivasan @ 2017-03-23 3:26 ` Madhavan Srinivasan 5 siblings, 0 replies; 15+ messages in thread From: Madhavan Srinivasan @ 2017-03-23 3:26 UTC (permalink / raw) To: mpe Cc: linux-kernel, linuxppc-dev, Madhavan Srinivasan, Benjamin Herrenschmidt, Paul Mackerras, Sukadev Bhattiprolu, Daniel Axtens, Andrew Donnellan, Peter Zijlstra, Ingo Molnar, Arnaldo Carvalho de Melo, Alexander Shishkin, Wang Nan, Alexei Starovoitov, Stephane Eranian Patch add "mem_access" event to sysfs. This as-is not a raw event supported by Power8 pmu. Instead, it is formed based on raw event encoding specificed in isa207-common.h. Primary PMU event used here is PM_MRK_INST_CMPL. This event tracks only the completed marked instructions. Random sampling mode (MMCRA[SM]) with Random Instruction Sampling (RIS) is enabled to mark type of instructions. With Random sampling in RLS mode with PM_MRK_INST_CMPL event, the LDST /DATA_SRC fields in SIER identifies the memory hierarchy level (eg: L1, L2 etc) statisfied a data-cache miss for a marked instruction. Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org> Cc: Paul Mackerras <paulus@samba.org> Cc: Sukadev Bhattiprolu <sukadev@linux.vnet.ibm.com> Cc: Daniel Axtens <dja@axtens.net> Cc: Andrew Donnellan <andrew.donnellan@au1.ibm.com> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Ingo Molnar <mingo@redhat.com> Cc: Arnaldo Carvalho de Melo <acme@kernel.org> Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com> Cc: Wang Nan <wangnan0@huawei.com> Cc: Alexei Starovoitov <ast@kernel.org> Cc: Stephane Eranian <eranian@google.com> Signed-off-by: Madhavan Srinivasan <maddy@linux.vnet.ibm.com> --- arch/powerpc/perf/power8-events-list.h | 6 ++++++ arch/powerpc/perf/power8-pmu.c | 2 ++ 2 files changed, 8 insertions(+) diff --git a/arch/powerpc/perf/power8-events-list.h b/arch/powerpc/perf/power8-events-list.h index 3a2e6e8ebb92..0f1d184627cc 100644 --- a/arch/powerpc/perf/power8-events-list.h +++ b/arch/powerpc/perf/power8-events-list.h @@ -89,3 +89,9 @@ EVENT(PM_MRK_FILT_MATCH, 0x2013c) EVENT(PM_MRK_FILT_MATCH_ALT, 0x3012e) /* Alternate event code for PM_LD_MISS_L1 */ EVENT(PM_LD_MISS_L1_ALT, 0x400f0) +/* + * Memory Access Event -- mem_access + * Primary PMU event used here is PM_MRK_INST_CMPL, along with + * Random Load/Store Facility Sampling (RIS) in Random sampling mode (MMCRA[SM]). + */ +EVENT(MEM_ACCESS, 0x10401e0) diff --git a/arch/powerpc/perf/power8-pmu.c b/arch/powerpc/perf/power8-pmu.c index 932d7536f0eb..5463516e369b 100644 --- a/arch/powerpc/perf/power8-pmu.c +++ b/arch/powerpc/perf/power8-pmu.c @@ -90,6 +90,7 @@ GENERIC_EVENT_ATTR(branch-instructions, PM_BRU_FIN); GENERIC_EVENT_ATTR(branch-misses, PM_BR_MPRED_CMPL); GENERIC_EVENT_ATTR(cache-references, PM_LD_REF_L1); GENERIC_EVENT_ATTR(cache-misses, PM_LD_MISS_L1); +GENERIC_EVENT_ATTR(mem_access, MEM_ACCESS); CACHE_EVENT_ATTR(L1-dcache-load-misses, PM_LD_MISS_L1); CACHE_EVENT_ATTR(L1-dcache-loads, PM_LD_REF_L1); @@ -120,6 +121,7 @@ static struct attribute *power8_events_attr[] = { GENERIC_EVENT_PTR(PM_BR_MPRED_CMPL), GENERIC_EVENT_PTR(PM_LD_REF_L1), GENERIC_EVENT_PTR(PM_LD_MISS_L1), + GENERIC_EVENT_PTR(MEM_ACCESS), CACHE_EVENT_PTR(PM_LD_MISS_L1), CACHE_EVENT_PTR(PM_LD_REF_L1), -- 2.7.4 ^ permalink raw reply related [flat|nested] 15+ messages in thread
* [PATCH v3 0/6] powerpc/perf: Export memory hierarchy level @ 2017-04-11 1:51 Madhavan Srinivasan 2017-04-11 1:51 ` [PATCH v3 1/6] powerpc/perf: Define big-endian version of perf_mem_data_src Madhavan Srinivasan 0 siblings, 1 reply; 15+ messages in thread From: Madhavan Srinivasan @ 2017-04-11 1:51 UTC (permalink / raw) To: mpe Cc: linux-kernel, linuxppc-dev, benh, paulus, sukadev, andrew.donnellan, peterz, mingo, acme, alexander.shishkin, wangnan0, ast, eranian, Madhavan Srinivasan Power8/Power9 Perforence Monitoring Unit (PMU) supports different sampling modes (SM) such as Random Instruction Sampling (RIS), Random Load/Store Facility Sampling (RLS) and Random Branch Sampling (RBS). Sample mode RLS updates Sampled Instruction Event Register [SIER] bits with memory hierarchy information for a cache reload. Patchset exports the hierarchy information to the user via the perf_mem_data_src object from SIER. Patchset is a rebase of the work posted previously with minor updates to it. https://lkml.org/lkml/2015/6/11/92 Changelog v3: -Removed is_load_store() and merged the same to get_memdata_src callback -Added a check to update OP_LOAD or OP_STORE in data_src->val Changelog v2: -Updated the commit messages -Fixed isa207_find_source() to consider all the possible sier[ldst] values. Changelog v1: - Fixed author-ship for the first patch and added suka's "Signed-off-by:". Madhavan Srinivasan (5): powerpc/perf: Export memory hierarchy info to user space powerpc/perf: Support to export MMCRA[TEC*] field to userspace powerpc/perf: Support to export SIERs bit in Power8 powerpc/perf: Support to export SIERs bit in Power9 powerpc/perf: Add Power8 mem_access event to sysfs Sukadev Bhattiprolu (1): powerpc/perf: Define big-endian version of perf_mem_data_src arch/powerpc/include/asm/perf_event_server.h | 3 + arch/powerpc/perf/core-book3s.c | 8 +++ arch/powerpc/perf/isa207-common.c | 82 ++++++++++++++++++++++++++++ arch/powerpc/perf/isa207-common.h | 26 ++++++++- arch/powerpc/perf/power8-events-list.h | 6 ++ arch/powerpc/perf/power8-pmu.c | 4 ++ arch/powerpc/perf/power9-pmu.c | 2 + include/uapi/linux/perf_event.h | 16 ++++++ tools/include/uapi/linux/perf_event.h | 16 ++++++ 9 files changed, 162 insertions(+), 1 deletion(-) -- 2.7.4 ^ permalink raw reply [flat|nested] 15+ messages in thread
* [PATCH v3 1/6] powerpc/perf: Define big-endian version of perf_mem_data_src 2017-04-11 1:51 [PATCH v3 0/6] powerpc/perf: Export memory hierarchy level Madhavan Srinivasan @ 2017-04-11 1:51 ` Madhavan Srinivasan 2017-04-13 12:38 ` Peter Zijlstra 0 siblings, 1 reply; 15+ messages in thread From: Madhavan Srinivasan @ 2017-04-11 1:51 UTC (permalink / raw) To: mpe Cc: linux-kernel, linuxppc-dev, benh, paulus, sukadev, andrew.donnellan, peterz, mingo, acme, alexander.shishkin, wangnan0, ast, eranian, Madhavan Srinivasan From: Sukadev Bhattiprolu <sukadev@linux.vnet.ibm.com> perf_mem_data_src is an union that is initialized via the ->val field and accessed via the bitmap fields. For this to work on big endian platforms (Which is broken now), we also need a big-endian represenation of perf_mem_data_src. i.e, in a big endian system, if user request PERF_SAMPLE_DATA_SRC (perf report -d), will get the default value from perf_sample_data_init(), which is PERF_MEM_NA. Value for PERF_MEM_NA is constructed using shifts: /* TLB access */ #define PERF_MEM_TLB_NA 0x01 /* not available */ ... #define PERF_MEM_TLB_SHIFT 26 #define PERF_MEM_S(a, s) \ (((__u64)PERF_MEM_##a##_##s) << PERF_MEM_##a##_SHIFT) #define PERF_MEM_NA (PERF_MEM_S(OP, NA) |\ PERF_MEM_S(LVL, NA) |\ PERF_MEM_S(SNOOP, NA) |\ PERF_MEM_S(LOCK, NA) |\ PERF_MEM_S(TLB, NA)) Which works out as: ((0x01 << 0) | (0x01 << 5) | (0x01 << 19) | (0x01 << 24) | (0x01 << 26)) Which means the PERF_MEM_NA value comes out of the kernel as 0x5080021 in CPU endian. But then in the perf tool, the code uses the bitfields to inspect the value, and currently the bitfields are defined using little endian ordering. So eg. in perf_mem__tlb_scnprintf() we see: data_src->val = 0x5080021 op = 0x0 lvl = 0x0 snoop = 0x0 lock = 0x0 dtlb = 0x0 rsvd = 0x5080021 Patch does a minimal fix of adding big endian definition of the bitfields to match the values that are already exported by the kernel on big endian. And it makes no change on little endian. Signed-off-by: Sukadev Bhattiprolu <sukadev@linux.vnet.ibm.com> Signed-off-by: Madhavan Srinivasan <maddy@linux.vnet.ibm.com> --- include/uapi/linux/perf_event.h | 16 ++++++++++++++++ tools/include/uapi/linux/perf_event.h | 16 ++++++++++++++++ 2 files changed, 32 insertions(+) diff --git a/include/uapi/linux/perf_event.h b/include/uapi/linux/perf_event.h index c66a485a24ac..c4af1159a200 100644 --- a/include/uapi/linux/perf_event.h +++ b/include/uapi/linux/perf_event.h @@ -891,6 +891,7 @@ enum perf_callchain_context { #define PERF_FLAG_PID_CGROUP (1UL << 2) /* pid=cgroup id, per-cpu mode only */ #define PERF_FLAG_FD_CLOEXEC (1UL << 3) /* O_CLOEXEC */ +#if defined(__LITTLE_ENDIAN_BITFIELD) union perf_mem_data_src { __u64 val; struct { @@ -902,6 +903,21 @@ union perf_mem_data_src { mem_rsvd:31; }; }; +#elif defined(__BIG_ENDIAN_BITFIELD) +union perf_mem_data_src { + __u64 val; + struct { + __u64 mem_rsvd:31, + mem_dtlb:7, /* tlb access */ + mem_lock:2, /* lock instr */ + mem_snoop:5, /* snoop mode */ + mem_lvl:14, /* memory hierarchy level */ + mem_op:5; /* type of opcode */ + }; +}; +#else +#error "Unknown endianness" +#endif /* type of opcode (load/store/prefetch,code) */ #define PERF_MEM_OP_NA 0x01 /* not available */ diff --git a/tools/include/uapi/linux/perf_event.h b/tools/include/uapi/linux/perf_event.h index c66a485a24ac..c4af1159a200 100644 --- a/tools/include/uapi/linux/perf_event.h +++ b/tools/include/uapi/linux/perf_event.h @@ -891,6 +891,7 @@ enum perf_callchain_context { #define PERF_FLAG_PID_CGROUP (1UL << 2) /* pid=cgroup id, per-cpu mode only */ #define PERF_FLAG_FD_CLOEXEC (1UL << 3) /* O_CLOEXEC */ +#if defined(__LITTLE_ENDIAN_BITFIELD) union perf_mem_data_src { __u64 val; struct { @@ -902,6 +903,21 @@ union perf_mem_data_src { mem_rsvd:31; }; }; +#elif defined(__BIG_ENDIAN_BITFIELD) +union perf_mem_data_src { + __u64 val; + struct { + __u64 mem_rsvd:31, + mem_dtlb:7, /* tlb access */ + mem_lock:2, /* lock instr */ + mem_snoop:5, /* snoop mode */ + mem_lvl:14, /* memory hierarchy level */ + mem_op:5; /* type of opcode */ + }; +}; +#else +#error "Unknown endianness" +#endif /* type of opcode (load/store/prefetch,code) */ #define PERF_MEM_OP_NA 0x01 /* not available */ -- 2.7.4 ^ permalink raw reply related [flat|nested] 15+ messages in thread
* Re: [PATCH v3 1/6] powerpc/perf: Define big-endian version of perf_mem_data_src 2017-04-11 1:51 ` [PATCH v3 1/6] powerpc/perf: Define big-endian version of perf_mem_data_src Madhavan Srinivasan @ 2017-04-13 12:38 ` Peter Zijlstra 2017-04-13 13:23 ` Michael Ellerman ` (2 more replies) 0 siblings, 3 replies; 15+ messages in thread From: Peter Zijlstra @ 2017-04-13 12:38 UTC (permalink / raw) To: Madhavan Srinivasan Cc: mpe, linux-kernel, linuxppc-dev, benh, paulus, sukadev, andrew.donnellan, mingo, acme, alexander.shishkin, wangnan0, ast, eranian On Tue, Apr 11, 2017 at 07:21:05AM +0530, Madhavan Srinivasan wrote: > From: Sukadev Bhattiprolu <sukadev@linux.vnet.ibm.com> > > perf_mem_data_src is an union that is initialized via the ->val field > and accessed via the bitmap fields. For this to work on big endian > platforms (Which is broken now), we also need a big-endian represenation > of perf_mem_data_src. i.e, in a big endian system, if user request > PERF_SAMPLE_DATA_SRC (perf report -d), will get the default value from > perf_sample_data_init(), which is PERF_MEM_NA. Value for PERF_MEM_NA > is constructed using shifts: > > /* TLB access */ > #define PERF_MEM_TLB_NA 0x01 /* not available */ > ... > #define PERF_MEM_TLB_SHIFT 26 > > #define PERF_MEM_S(a, s) \ > (((__u64)PERF_MEM_##a##_##s) << PERF_MEM_##a##_SHIFT) > > #define PERF_MEM_NA (PERF_MEM_S(OP, NA) |\ > PERF_MEM_S(LVL, NA) |\ > PERF_MEM_S(SNOOP, NA) |\ > PERF_MEM_S(LOCK, NA) |\ > PERF_MEM_S(TLB, NA)) > > Which works out as: > > ((0x01 << 0) | (0x01 << 5) | (0x01 << 19) | (0x01 << 24) | (0x01 << 26)) > > Which means the PERF_MEM_NA value comes out of the kernel as 0x5080021 > in CPU endian. > > But then in the perf tool, the code uses the bitfields to inspect the > value, and currently the bitfields are defined using little endian > ordering. > > So eg. in perf_mem__tlb_scnprintf() we see: > data_src->val = 0x5080021 > op = 0x0 > lvl = 0x0 > snoop = 0x0 > lock = 0x0 > dtlb = 0x0 > rsvd = 0x5080021 > > Patch does a minimal fix of adding big endian definition of the bitfields > to match the values that are already exported by the kernel on big endian. > And it makes no change on little endian. I think it is important to note that there are no current big-endian users. So 'fixing' this will not break anybody and will ensure future users (next patch) will work correctly. Aside from that amendment, Acked-by: Peter Zijlstra (Intel) <peterz@infradead.org> ^ permalink raw reply [flat|nested] 15+ messages in thread
* Re: [PATCH v3 1/6] powerpc/perf: Define big-endian version of perf_mem_data_src 2017-04-13 12:38 ` Peter Zijlstra @ 2017-04-13 13:23 ` Michael Ellerman 2017-04-17 3:46 ` Madhavan Srinivasan 2017-04-17 3:46 ` Madhavan Srinivasan 2017-04-19 4:50 ` Michael Ellerman 2 siblings, 1 reply; 15+ messages in thread From: Michael Ellerman @ 2017-04-13 13:23 UTC (permalink / raw) To: Peter Zijlstra, Madhavan Srinivasan Cc: linux-kernel, linuxppc-dev, benh, paulus, sukadev, andrew.donnellan, mingo, acme, alexander.shishkin, wangnan0, ast, eranian Peter Zijlstra <peterz@infradead.org> writes: > On Tue, Apr 11, 2017 at 07:21:05AM +0530, Madhavan Srinivasan wrote: >> From: Sukadev Bhattiprolu <sukadev@linux.vnet.ibm.com> >> >> perf_mem_data_src is an union that is initialized via the ->val field >> and accessed via the bitmap fields. For this to work on big endian >> platforms (Which is broken now), we also need a big-endian represenation >> of perf_mem_data_src. i.e, in a big endian system, if user request >> PERF_SAMPLE_DATA_SRC (perf report -d), will get the default value from >> perf_sample_data_init(), which is PERF_MEM_NA. Value for PERF_MEM_NA >> is constructed using shifts: >> >> /* TLB access */ >> #define PERF_MEM_TLB_NA 0x01 /* not available */ >> ... >> #define PERF_MEM_TLB_SHIFT 26 >> >> #define PERF_MEM_S(a, s) \ >> (((__u64)PERF_MEM_##a##_##s) << PERF_MEM_##a##_SHIFT) >> >> #define PERF_MEM_NA (PERF_MEM_S(OP, NA) |\ >> PERF_MEM_S(LVL, NA) |\ >> PERF_MEM_S(SNOOP, NA) |\ >> PERF_MEM_S(LOCK, NA) |\ >> PERF_MEM_S(TLB, NA)) >> >> Which works out as: >> >> ((0x01 << 0) | (0x01 << 5) | (0x01 << 19) | (0x01 << 24) | (0x01 << 26)) >> >> Which means the PERF_MEM_NA value comes out of the kernel as 0x5080021 >> in CPU endian. >> >> But then in the perf tool, the code uses the bitfields to inspect the >> value, and currently the bitfields are defined using little endian >> ordering. >> >> So eg. in perf_mem__tlb_scnprintf() we see: >> data_src->val = 0x5080021 >> op = 0x0 >> lvl = 0x0 >> snoop = 0x0 >> lock = 0x0 >> dtlb = 0x0 >> rsvd = 0x5080021 >> >> Patch does a minimal fix of adding big endian definition of the bitfields >> to match the values that are already exported by the kernel on big endian. >> And it makes no change on little endian. > > I think it is important to note that there are no current big-endian > users. So 'fixing' this will not break anybody and will ensure future > users (next patch) will work correctly. Sure I'll fold in something along those lines. > Aside from that amendment, > > Acked-by: Peter Zijlstra (Intel) <peterz@infradead.org> Thanks. cheers ^ permalink raw reply [flat|nested] 15+ messages in thread
* Re: [PATCH v3 1/6] powerpc/perf: Define big-endian version of perf_mem_data_src 2017-04-13 13:23 ` Michael Ellerman @ 2017-04-17 3:46 ` Madhavan Srinivasan 0 siblings, 0 replies; 15+ messages in thread From: Madhavan Srinivasan @ 2017-04-17 3:46 UTC (permalink / raw) To: Michael Ellerman, Peter Zijlstra Cc: linux-kernel, linuxppc-dev, benh, paulus, sukadev, andrew.donnellan, mingo, acme, alexander.shishkin, wangnan0, ast, eranian On Thursday 13 April 2017 06:53 PM, Michael Ellerman wrote: > Peter Zijlstra <peterz@infradead.org> writes: > >> On Tue, Apr 11, 2017 at 07:21:05AM +0530, Madhavan Srinivasan wrote: >>> From: Sukadev Bhattiprolu <sukadev@linux.vnet.ibm.com> >>> >>> perf_mem_data_src is an union that is initialized via the ->val field >>> and accessed via the bitmap fields. For this to work on big endian >>> platforms (Which is broken now), we also need a big-endian represenation >>> of perf_mem_data_src. i.e, in a big endian system, if user request >>> PERF_SAMPLE_DATA_SRC (perf report -d), will get the default value from >>> perf_sample_data_init(), which is PERF_MEM_NA. Value for PERF_MEM_NA >>> is constructed using shifts: >>> >>> /* TLB access */ >>> #define PERF_MEM_TLB_NA 0x01 /* not available */ >>> ... >>> #define PERF_MEM_TLB_SHIFT 26 >>> >>> #define PERF_MEM_S(a, s) \ >>> (((__u64)PERF_MEM_##a##_##s) << PERF_MEM_##a##_SHIFT) >>> >>> #define PERF_MEM_NA (PERF_MEM_S(OP, NA) |\ >>> PERF_MEM_S(LVL, NA) |\ >>> PERF_MEM_S(SNOOP, NA) |\ >>> PERF_MEM_S(LOCK, NA) |\ >>> PERF_MEM_S(TLB, NA)) >>> >>> Which works out as: >>> >>> ((0x01 << 0) | (0x01 << 5) | (0x01 << 19) | (0x01 << 24) | (0x01 << 26)) >>> >>> Which means the PERF_MEM_NA value comes out of the kernel as 0x5080021 >>> in CPU endian. >>> >>> But then in the perf tool, the code uses the bitfields to inspect the >>> value, and currently the bitfields are defined using little endian >>> ordering. >>> >>> So eg. in perf_mem__tlb_scnprintf() we see: >>> data_src->val = 0x5080021 >>> op = 0x0 >>> lvl = 0x0 >>> snoop = 0x0 >>> lock = 0x0 >>> dtlb = 0x0 >>> rsvd = 0x5080021 >>> >>> Patch does a minimal fix of adding big endian definition of the bitfields >>> to match the values that are already exported by the kernel on big endian. >>> And it makes no change on little endian. >> I think it is important to note that there are no current big-endian >> users. So 'fixing' this will not break anybody and will ensure future >> users (next patch) will work correctly. > Sure I'll fold in something along those lines. Thanks mpe. Maddy > >> Aside from that amendment, >> >> Acked-by: Peter Zijlstra (Intel) <peterz@infradead.org> > Thanks. > > cheers > ^ permalink raw reply [flat|nested] 15+ messages in thread
* Re: [PATCH v3 1/6] powerpc/perf: Define big-endian version of perf_mem_data_src 2017-04-13 12:38 ` Peter Zijlstra 2017-04-13 13:23 ` Michael Ellerman @ 2017-04-17 3:46 ` Madhavan Srinivasan 2017-04-19 4:50 ` Michael Ellerman 2 siblings, 0 replies; 15+ messages in thread From: Madhavan Srinivasan @ 2017-04-17 3:46 UTC (permalink / raw) To: Peter Zijlstra Cc: mpe, linux-kernel, linuxppc-dev, benh, paulus, sukadev, andrew.donnellan, mingo, acme, alexander.shishkin, wangnan0, ast, eranian On Thursday 13 April 2017 06:08 PM, Peter Zijlstra wrote: > On Tue, Apr 11, 2017 at 07:21:05AM +0530, Madhavan Srinivasan wrote: >> From: Sukadev Bhattiprolu <sukadev@linux.vnet.ibm.com> >> >> perf_mem_data_src is an union that is initialized via the ->val field >> and accessed via the bitmap fields. For this to work on big endian >> platforms (Which is broken now), we also need a big-endian represenation >> of perf_mem_data_src. i.e, in a big endian system, if user request >> PERF_SAMPLE_DATA_SRC (perf report -d), will get the default value from >> perf_sample_data_init(), which is PERF_MEM_NA. Value for PERF_MEM_NA >> is constructed using shifts: >> >> /* TLB access */ >> #define PERF_MEM_TLB_NA 0x01 /* not available */ >> ... >> #define PERF_MEM_TLB_SHIFT 26 >> >> #define PERF_MEM_S(a, s) \ >> (((__u64)PERF_MEM_##a##_##s) << PERF_MEM_##a##_SHIFT) >> >> #define PERF_MEM_NA (PERF_MEM_S(OP, NA) |\ >> PERF_MEM_S(LVL, NA) |\ >> PERF_MEM_S(SNOOP, NA) |\ >> PERF_MEM_S(LOCK, NA) |\ >> PERF_MEM_S(TLB, NA)) >> >> Which works out as: >> >> ((0x01 << 0) | (0x01 << 5) | (0x01 << 19) | (0x01 << 24) | (0x01 << 26)) >> >> Which means the PERF_MEM_NA value comes out of the kernel as 0x5080021 >> in CPU endian. >> >> But then in the perf tool, the code uses the bitfields to inspect the >> value, and currently the bitfields are defined using little endian >> ordering. >> >> So eg. in perf_mem__tlb_scnprintf() we see: >> data_src->val = 0x5080021 >> op = 0x0 >> lvl = 0x0 >> snoop = 0x0 >> lock = 0x0 >> dtlb = 0x0 >> rsvd = 0x5080021 >> >> Patch does a minimal fix of adding big endian definition of the bitfields >> to match the values that are already exported by the kernel on big endian. >> And it makes no change on little endian. > I think it is important to note that there are no current big-endian > users. So 'fixing' this will not break anybody and will ensure future > users (next patch) will work correctly. > > Aside from that amendment, > > Acked-by: Peter Zijlstra (Intel) <peterz@infradead.org> Thanks Maddy ^ permalink raw reply [flat|nested] 15+ messages in thread
* Re: [PATCH v3 1/6] powerpc/perf: Define big-endian version of perf_mem_data_src 2017-04-13 12:38 ` Peter Zijlstra 2017-04-13 13:23 ` Michael Ellerman 2017-04-17 3:46 ` Madhavan Srinivasan @ 2017-04-19 4:50 ` Michael Ellerman 2017-04-19 14:32 ` Madhavan Srinivasan 2 siblings, 1 reply; 15+ messages in thread From: Michael Ellerman @ 2017-04-19 4:50 UTC (permalink / raw) To: Peter Zijlstra, Madhavan Srinivasan Cc: linux-kernel, linuxppc-dev, benh, paulus, sukadev, andrew.donnellan, mingo, acme, alexander.shishkin, wangnan0, ast, eranian Peter Zijlstra <peterz@infradead.org> writes: > On Tue, Apr 11, 2017 at 07:21:05AM +0530, Madhavan Srinivasan wrote: >> From: Sukadev Bhattiprolu <sukadev@linux.vnet.ibm.com> >> >> perf_mem_data_src is an union that is initialized via the ->val field >> and accessed via the bitmap fields. For this to work on big endian >> platforms (Which is broken now), we also need a big-endian represenation >> of perf_mem_data_src. i.e, in a big endian system, if user request >> PERF_SAMPLE_DATA_SRC (perf report -d), will get the default value from >> perf_sample_data_init(), which is PERF_MEM_NA. Value for PERF_MEM_NA >> is constructed using shifts: >> >> /* TLB access */ >> #define PERF_MEM_TLB_NA 0x01 /* not available */ >> ... >> #define PERF_MEM_TLB_SHIFT 26 >> >> #define PERF_MEM_S(a, s) \ >> (((__u64)PERF_MEM_##a##_##s) << PERF_MEM_##a##_SHIFT) >> >> #define PERF_MEM_NA (PERF_MEM_S(OP, NA) |\ >> PERF_MEM_S(LVL, NA) |\ >> PERF_MEM_S(SNOOP, NA) |\ >> PERF_MEM_S(LOCK, NA) |\ >> PERF_MEM_S(TLB, NA)) >> >> Which works out as: >> >> ((0x01 << 0) | (0x01 << 5) | (0x01 << 19) | (0x01 << 24) | (0x01 << 26)) >> >> Which means the PERF_MEM_NA value comes out of the kernel as 0x5080021 >> in CPU endian. >> >> But then in the perf tool, the code uses the bitfields to inspect the >> value, and currently the bitfields are defined using little endian >> ordering. >> >> So eg. in perf_mem__tlb_scnprintf() we see: >> data_src->val = 0x5080021 >> op = 0x0 >> lvl = 0x0 >> snoop = 0x0 >> lock = 0x0 >> dtlb = 0x0 >> rsvd = 0x5080021 >> >> Patch does a minimal fix of adding big endian definition of the bitfields >> to match the values that are already exported by the kernel on big endian. >> And it makes no change on little endian. > > I think it is important to note that there are no current big-endian > users. So 'fixing' this will not break anybody and will ensure future > users (next patch) will work correctly. Actually that's only partly true. As I describe above the PERF_MEM_NA value is currently exported on BE platforms when a user requests it. So I added this text after the output from perf_mem__tlb_scnprintf(): Because of the way the perf tool code is written this is still displayed to the user as "N/A", so there is no bug visible at the UI level. Currently there are no big endian architectures which export a meaningful value (ie. other than PERF_MEM_NA), so the extent of the bug on big endian platforms is that the PERF_MEM_NA value is exported incorrectly as described above. Subsequent patches will add support on big endian powerpc for populating the data source value. Hope that is clear. It also occurred to me that we don't actually have to redefine the whole union, it's only the bitfields that matter, so we could reduce the diff to: diff --git a/include/uapi/linux/perf_event.h b/include/uapi/linux/perf_event.h index c66a485a24ac..97152c79df6b 100644 --- a/include/uapi/linux/perf_event.h +++ b/include/uapi/linux/perf_event.h @@ -894,12 +894,23 @@ enum perf_callchain_context { union perf_mem_data_src { __u64 val; struct { +#if defined(__LITTLE_ENDIAN_BITFIELD) __u64 mem_op:5, /* type of opcode */ mem_lvl:14, /* memory hierarchy level */ mem_snoop:5, /* snoop mode */ mem_lock:2, /* lock instr */ mem_dtlb:7, /* tlb access */ mem_rsvd:31; +#elif defined(__BIG_ENDIAN_BITFIELD) + __u64 mem_rsvd:31, + mem_dtlb:7, /* tlb access */ + mem_lock:2, /* lock instr */ + mem_snoop:5, /* snoop mode */ + mem_lvl:14, /* memory hierarchy level */ + mem_op:5; /* type of opcode */ +#else +#error "Unknown endianness" +#endif }; }; That looks better to me, thoughts? cheers ^ permalink raw reply related [flat|nested] 15+ messages in thread
* Re: [PATCH v3 1/6] powerpc/perf: Define big-endian version of perf_mem_data_src 2017-04-19 4:50 ` Michael Ellerman @ 2017-04-19 14:32 ` Madhavan Srinivasan 2017-04-19 22:16 ` Michael Ellerman 0 siblings, 1 reply; 15+ messages in thread From: Madhavan Srinivasan @ 2017-04-19 14:32 UTC (permalink / raw) To: Michael Ellerman, Peter Zijlstra Cc: linux-kernel, linuxppc-dev, benh, paulus, sukadev, andrew.donnellan, mingo, acme, alexander.shishkin, wangnan0, ast, eranian On Wednesday 19 April 2017 10:20 AM, Michael Ellerman wrote: > Peter Zijlstra <peterz@infradead.org> writes: > >> On Tue, Apr 11, 2017 at 07:21:05AM +0530, Madhavan Srinivasan wrote: >>> From: Sukadev Bhattiprolu <sukadev@linux.vnet.ibm.com> >>> >>> perf_mem_data_src is an union that is initialized via the ->val field >>> and accessed via the bitmap fields. For this to work on big endian >>> platforms (Which is broken now), we also need a big-endian represenation >>> of perf_mem_data_src. i.e, in a big endian system, if user request >>> PERF_SAMPLE_DATA_SRC (perf report -d), will get the default value from >>> perf_sample_data_init(), which is PERF_MEM_NA. Value for PERF_MEM_NA >>> is constructed using shifts: >>> >>> /* TLB access */ >>> #define PERF_MEM_TLB_NA 0x01 /* not available */ >>> ... >>> #define PERF_MEM_TLB_SHIFT 26 >>> >>> #define PERF_MEM_S(a, s) \ >>> (((__u64)PERF_MEM_##a##_##s) << PERF_MEM_##a##_SHIFT) >>> >>> #define PERF_MEM_NA (PERF_MEM_S(OP, NA) |\ >>> PERF_MEM_S(LVL, NA) |\ >>> PERF_MEM_S(SNOOP, NA) |\ >>> PERF_MEM_S(LOCK, NA) |\ >>> PERF_MEM_S(TLB, NA)) >>> >>> Which works out as: >>> >>> ((0x01 << 0) | (0x01 << 5) | (0x01 << 19) | (0x01 << 24) | (0x01 << 26)) >>> >>> Which means the PERF_MEM_NA value comes out of the kernel as 0x5080021 >>> in CPU endian. >>> >>> But then in the perf tool, the code uses the bitfields to inspect the >>> value, and currently the bitfields are defined using little endian >>> ordering. >>> >>> So eg. in perf_mem__tlb_scnprintf() we see: >>> data_src->val = 0x5080021 >>> op = 0x0 >>> lvl = 0x0 >>> snoop = 0x0 >>> lock = 0x0 >>> dtlb = 0x0 >>> rsvd = 0x5080021 >>> >>> Patch does a minimal fix of adding big endian definition of the bitfields >>> to match the values that are already exported by the kernel on big endian. >>> And it makes no change on little endian. >> I think it is important to note that there are no current big-endian >> users. So 'fixing' this will not break anybody and will ensure future >> users (next patch) will work correctly. > Actually that's only partly true. As I describe above the PERF_MEM_NA > value is currently exported on BE platforms when a user requests it. > > So I added this text after the output from perf_mem__tlb_scnprintf(): > > Because of the way the perf tool code is written this is still displayed to the > user as "N/A", so there is no bug visible at the UI level. > > Currently there are no big endian architectures which export a meaningful > value (ie. other than PERF_MEM_NA), so the extent of the bug on big endian > platforms is that the PERF_MEM_NA value is exported incorrectly as described > above. Subsequent patches will add support on big endian powerpc for populating > the data source value. > > > Hope that is clear. > > It also occurred to me that we don't actually have to redefine the whole > union, it's only the bitfields that matter, so we could reduce the diff > to: > > diff --git a/include/uapi/linux/perf_event.h b/include/uapi/linux/perf_event.h > index c66a485a24ac..97152c79df6b 100644 > --- a/include/uapi/linux/perf_event.h > +++ b/include/uapi/linux/perf_event.h > @@ -894,12 +894,23 @@ enum perf_callchain_context { > union perf_mem_data_src { > __u64 val; > struct { > +#if defined(__LITTLE_ENDIAN_BITFIELD) > __u64 mem_op:5, /* type of opcode */ > mem_lvl:14, /* memory hierarchy level */ > mem_snoop:5, /* snoop mode */ > mem_lock:2, /* lock instr */ > mem_dtlb:7, /* tlb access */ > mem_rsvd:31; > +#elif defined(__BIG_ENDIAN_BITFIELD) > + __u64 mem_rsvd:31, > + mem_dtlb:7, /* tlb access */ > + mem_lock:2, /* lock instr */ > + mem_snoop:5, /* snoop mode */ > + mem_lvl:14, /* memory hierarchy level */ > + mem_op:5; /* type of opcode */ > +#else > +#error "Unknown endianness" > +#endif > }; > }; > > > That looks better to me, thoughts? Yep. Looks fine to me and also tested the same. Maddy > > cheers > ^ permalink raw reply [flat|nested] 15+ messages in thread
* Re: [PATCH v3 1/6] powerpc/perf: Define big-endian version of perf_mem_data_src 2017-04-19 14:32 ` Madhavan Srinivasan @ 2017-04-19 22:16 ` Michael Ellerman 0 siblings, 0 replies; 15+ messages in thread From: Michael Ellerman @ 2017-04-19 22:16 UTC (permalink / raw) To: Madhavan Srinivasan, Peter Zijlstra Cc: linux-kernel, linuxppc-dev, benh, paulus, sukadev, andrew.donnellan, mingo, acme, alexander.shishkin, wangnan0, ast, eranian Madhavan Srinivasan <maddy@linux.vnet.ibm.com> writes: > On Wednesday 19 April 2017 10:20 AM, Michael Ellerman wrote: >> It also occurred to me that we don't actually have to redefine the whole >> union, it's only the bitfields that matter, so we could reduce the diff >> to: >> >> diff --git a/include/uapi/linux/perf_event.h b/include/uapi/linux/perf_event.h >> index c66a485a24ac..97152c79df6b 100644 >> --- a/include/uapi/linux/perf_event.h >> +++ b/include/uapi/linux/perf_event.h >> @@ -894,12 +894,23 @@ enum perf_callchain_context { >> union perf_mem_data_src { >> __u64 val; >> struct { >> +#if defined(__LITTLE_ENDIAN_BITFIELD) >> __u64 mem_op:5, /* type of opcode */ >> mem_lvl:14, /* memory hierarchy level */ >> mem_snoop:5, /* snoop mode */ >> mem_lock:2, /* lock instr */ >> mem_dtlb:7, /* tlb access */ >> mem_rsvd:31; >> +#elif defined(__BIG_ENDIAN_BITFIELD) >> + __u64 mem_rsvd:31, >> + mem_dtlb:7, /* tlb access */ >> + mem_lock:2, /* lock instr */ >> + mem_snoop:5, /* snoop mode */ >> + mem_lvl:14, /* memory hierarchy level */ >> + mem_op:5; /* type of opcode */ >> +#else >> +#error "Unknown endianness" >> +#endif >> }; >> }; >> >> >> That looks better to me, thoughts? > > Yep. Looks fine to me and also tested the same. I merged the original version, as that's what Peterz acked and I didn't want to block the series any longer. I'll send an incremental patch to do the cleanup. cheers ^ permalink raw reply [flat|nested] 15+ messages in thread
end of thread, other threads:[~2017-04-19 22:16 UTC | newest] Thread overview: 15+ messages (download: mbox.gz / follow: Atom feed) -- links below jump to the message on this page -- 2017-03-23 3:26 [PATCH v3 0/6] powerpc/perf: Export memory hierarchy level Madhavan Srinivasan 2017-03-23 3:26 ` [PATCH v3 1/6] powerpc/perf: Define big-endian version of perf_mem_data_src Madhavan Srinivasan 2017-03-23 3:26 ` [PATCH v3 2/6] powerpc/perf: Export memory hierarchy info to user space Madhavan Srinivasan 2017-03-23 3:26 ` [PATCH v3 3/6] powerpc/perf: Support to export MMCRA[TEC*] field to userspace Madhavan Srinivasan 2017-03-23 3:26 ` [PATCH v3 4/6] powerpc/perf: Support to export SIERs bit in Power8 Madhavan Srinivasan 2017-03-23 3:26 ` [PATCH v3 5/6] powerpc/perf: Support to export SIERs bit in Power9 Madhavan Srinivasan 2017-03-23 3:26 ` [PATCH v3 6/6] powerpc/perf: Add Power8 mem_access event to sysfs Madhavan Srinivasan 2017-04-11 1:51 [PATCH v3 0/6] powerpc/perf: Export memory hierarchy level Madhavan Srinivasan 2017-04-11 1:51 ` [PATCH v3 1/6] powerpc/perf: Define big-endian version of perf_mem_data_src Madhavan Srinivasan 2017-04-13 12:38 ` Peter Zijlstra 2017-04-13 13:23 ` Michael Ellerman 2017-04-17 3:46 ` Madhavan Srinivasan 2017-04-17 3:46 ` Madhavan Srinivasan 2017-04-19 4:50 ` Michael Ellerman 2017-04-19 14:32 ` Madhavan Srinivasan 2017-04-19 22:16 ` Michael Ellerman
This is a public inbox, see mirroring instructions for how to clone and mirror all data and code used for this inbox; as well as URLs for NNTP newsgroup(s).