xref: /freebsd/lib/libpmc/pmc.ivybridgexeon.3 (revision 8ddb146abcdf061be9f2c0db7e391697dafad85c)
1.\" Copyright (c) 2013 Hiren Panchasara <hiren.panchasara@gmail.com>
2.\" All rights reserved.
3.\"
4.\" Redistribution and use in source and binary forms, with or without
5.\" modification, are permitted provided that the following conditions
6.\" are met:
7.\" 1. Redistributions of source code must retain the above copyright
8.\"    notice, this list of conditions and the following disclaimer.
9.\" 2. Redistributions in binary form must reproduce the above copyright
10.\"    notice, this list of conditions and the following disclaimer in the
11.\"    documentation and/or other materials provided with the distribution.
12.\"
13.\" THIS SOFTWARE IS PROVIDED BY THE AUTHOR AND CONTRIBUTORS ``AS IS'' AND
14.\" ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
15.\" IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
16.\" ARE DISCLAIMED.  IN NO EVENT SHALL THE AUTHOR OR CONTRIBUTORS BE LIABLE
17.\" FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
18.\" DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
19.\" OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
20.\" HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
21.\" LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
22.\" OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
23.\" SUCH DAMAGE.
24.\"
25.\" $FreeBSD$
26.\"
27.Dd January 25, 2013
28.Dt PMC.IVYBRIDGEXEON 3
29.Os
30.Sh NAME
31.Nm pmc.ivybridgexeon
32.Nd measurement events for
33.Tn Intel
34.Tn Ivy Bridge Xeon
35family CPUs
36.Sh LIBRARY
37.Lb libpmc
38.Sh SYNOPSIS
39.In pmc.h
40.Sh DESCRIPTION
41.Tn Intel
42.Tn "Ivy Bridge Xeon"
43CPUs contain PMCs conforming to version 2 of the
44.Tn Intel
45performance measurement architecture.
46These CPUs may contain up to three classes of PMCs:
47.Bl -tag -width "Li PMC_CLASS_IAP"
48.It Li PMC_CLASS_IAF
49Fixed-function counters that count only one hardware event per counter.
50.It Li PMC_CLASS_IAP
51Programmable counters that may be configured to count one of a defined
52set of hardware events.
53.El
54.Pp
55The number of PMCs available in each class and their widths need to be
56determined at run time by calling
57.Xr pmc_cpuinfo 3 .
58.Pp
59Intel Ivy Bridge Xeon PMCs are documented in
60.Rs
61.%B "Intel(R) 64 and IA-32 Architectures Software Developer's Manual"
62.%N "Order Number: 325462-045US"
63.%D January 2013
64.%Q "Intel Corporation"
65.Re
66.Ss IVYBRIDGE FIXED FUNCTION PMCS
67These PMCs and their supported events are documented in
68.Xr pmc.iaf 3 .
69.Ss IVYBRIDGE PROGRAMMABLE PMCS
70The programmable PMCs support the following capabilities:
71.Bl -column "PMC_CAP_INTERRUPT" "Support"
72.It Em Capability Ta Em Support
73.It PMC_CAP_CASCADE Ta \&No
74.It PMC_CAP_EDGE Ta Yes
75.It PMC_CAP_INTERRUPT Ta Yes
76.It PMC_CAP_INVERT Ta Yes
77.It PMC_CAP_READ Ta Yes
78.It PMC_CAP_PRECISE Ta \&No
79.It PMC_CAP_SYSTEM Ta Yes
80.It PMC_CAP_TAGGING Ta \&No
81.It PMC_CAP_THRESHOLD Ta Yes
82.It PMC_CAP_USER Ta Yes
83.It PMC_CAP_WRITE Ta Yes
84.El
85.Ss Event Qualifiers
86Event specifiers for these PMCs support the following common
87qualifiers:
88.Bl -tag -width indent
89.It Li rsp= Ns Ar value
90Configure the Off-core Response bits.
91.Bl -tag -width indent
92.It Li REQ_DMND_DATA_RD
93Counts the number of demand and DCU prefetch data reads of full and partial
94cachelines as well as demand data page table entry cacheline reads.
95Does not count L2 data read prefetches or instruction fetches.
96.It Li REQ_DMND_RFO
97Counts the number of demand and DCU prefetch reads for ownership (RFO)
98requests generated by a write to data cacheline.
99Does not count L2 RFO prefetches.
100.It Li REQ_DMND_IFETCH
101Counts the number of demand and DCU prefetch instruction cacheline reads.
102Does not count L2 code read prefetches.
103.It Li REQ_WB
104Counts the number of writeback (modified to exclusive) transactions.
105.It Li REQ_PF_DATA_RD
106Counts the number of data cacheline reads generated by L2 prefetchers.
107.It Li REQ_PF_RFO
108Counts the number of RFO requests generated by L2 prefetchers.
109.It Li REQ_PF_IFETCH
110Counts the number of code reads generated by L2 prefetchers.
111.It Li REQ_PF_LLC_DATA_RD
112L2 prefetcher to L3 for loads.
113.It Li REQ_PF_LLC_RFO
114RFO requests generated by L2 prefetcher
115.It Li REQ_PF_LLC_IFETCH
116L2 prefetcher to L3 for instruction fetches.
117.It Li REQ_BUS_LOCKS
118Bus lock and split lock requests.
119.It Li REQ_STRM_ST
120Streaming store requests.
121.It Li REQ_OTHER
122Any other request that crosses IDI, including I/O.
123.It Li RES_ANY
124Catch all value for any response types.
125.It Li RES_SUPPLIER_NO_SUPP
126No Supplier Information available.
127.It Li RES_SUPPLIER_LLC_HITM
128M-state initial lookup stat in L3.
129.It Li RES_SUPPLIER_LLC_HITE
130E-state.
131.It Li RES_SUPPLIER_LLC_HITS
132S-state.
133.It Li RES_SUPPLIER_LLC_HITF
134F-state.
135.It Li RES_SUPPLIER_LOCAL
136Local DRAM Controller.
137.It Li RES_SNOOP_SNP_NONE
138No details on snoop-related information.
139.It Li RES_SNOOP_SNP_NO_NEEDED
140No snoop was needed to satisfy the request.
141.It Li RES_SNOOP_SNP_MISS
142A snoop was needed and it missed all snooped caches:
143-For LLC Hit, ReslHitl was returned by all cores
144-For LLC Miss, Rspl was returned by all sockets and data was returned from
145DRAM.
146.It Li RES_SNOOP_HIT_NO_FWD
147A snoop was needed and it hits in at least one snooped cache.
148Hit denotes a cache-line was valid before snoop effect.
149This includes:
150-Snoop Hit w/ Invalidation (LLC Hit, RFO)
151-Snoop Hit, Left Shared (LLC Hit/Miss, IFetch/Data_RD)
152-Snoop Hit w/ Invalidation and No Forward (LLC Miss, RFO Hit S)
153In the LLC Miss case, data is returned from DRAM.
154.It Li RES_SNOOP_HIT_FWD
155A snoop was needed and data was forwarded from a remote socket.
156This includes:
157-Snoop Forward Clean, Left Shared (LLC Hit/Miss, IFetch/Data_RD/RFT).
158.It Li RES_SNOOP_HITM
159A snoop was needed and it HitM-ed in local or remote cache.
160HitM denotes a cache-line was in modified state before effect as a results of snoop.
161This includes:
162-Snoop HitM w/ WB (LLC miss, IFetch/Data_RD)
163-Snoop Forward Modified w/ Invalidation (LLC Hit/Miss, RFO)
164-Snoop MtoS (LLC Hit, IFetch/Data_RD).
165.It Li RES_NON_DRAM
166Target was non-DRAM system address.
167This includes MMIO transactions.
168.El
169.It Li cmask= Ns Ar value
170Configure the PMC to increment only if the number of configured
171events measured in a cycle is greater than or equal to
172.Ar value .
173.It Li edge
174Configure the PMC to count the number of de-asserted to asserted
175transitions of the conditions expressed by the other qualifiers.
176If specified, the counter will increment only once whenever a
177condition becomes true, irrespective of the number of clocks during
178which the condition remains true.
179.It Li inv
180Invert the sense of comparison when the
181.Dq Li cmask
182qualifier is present, making the counter increment when the number of
183events per cycle is less than the value specified by the
184.Dq Li cmask
185qualifier.
186.It Li os
187Configure the PMC to count events happening at processor privilege
188level 0.
189.It Li usr
190Configure the PMC to count events occurring at privilege levels 1, 2
191or 3.
192.El
193.Pp
194If neither of the
195.Dq Li os
196or
197.Dq Li usr
198qualifiers are specified, the default is to enable both.
199.Ss Event Specifiers (Programmable PMCs)
200Ivy Bridge programmable PMCs support the following events:
201.Bl -tag -width indent
202.It Li LD_BLOCKS.STORE_FORWARD
203.Pq Event 03H , Umask 02H
204loads blocked by overlapping with store buffer that cannot be forwarded .
205.It Li MISALIGN_MEM_REF.LOADS
206.Pq Event 05H , Umask 01H
207Speculative cache-line split load uops dispatched to L1D.
208.It Li MISALIGN_MEM_REF.STORES
209.Pq Event 05H , Umask 02H
210Speculative cache-line split Store- address uops dispatched to L1D.
211.It Li LD_BLOCKS_PARTIAL.ADDRESS_ALIAS
212.Pq Event 07H , Umask 01H
213False dependencies in MOB due to partial compare on address.
214.It Li DTLB_LOAD_MISSES.DEMAND_LD_MISS_CAUSES_A_WALK
215.Pq Event 08H , Umask 81H
216Misses in all TLB levels that cause a page walk of any page size from demand loads.
217.It Li DTLB_LOAD_MISSES.DEMAND_LD_WALK_COMPLETED
218.Pq Event 08H , Umask 82H
219Misses in all TLB levels that caused page walk completed of any size by demand loads.
220.It Li DTLB_LOAD_MISSES.DEMAND_LD_WALK_DURATION
221.Pq Event 08H , Umask 84H
222Cycle PMH is busy with a walk due to demand loads.
223.It Li UOPS_ISSUED.ANY
224.Pq Event 0EH , Umask 01H
225Increments each cycle the # of Uops issued by the RAT to RS.
226Set Cmask = 1, Inv = 1to count stalled cycles.
227Set Cmask = 1, Inv = 1, Any= 1to count stalled cycles of this core.
228.It Li UOPS_ISSUED.FLAGS_MERGE
229.Pq Event 0EH , Umask 10H
230Number of flags-merge uops allocated.
231Such uops adds delay.
232.It Li UOPS_ISSUED.SLOW_LEA
233.Pq Event 0EH , Umask 20H
234Number of slow LEA or similar uops allocated.
235Such uop has 3 sources (e.g. 2 sources + immediate) regardless if as a result of LEA instruction or not.
236.It Li UOPS_ISSUED.SINGLE_MUL
237.Pq Event 0EH , Umask 40H
238Number of multiply packed/scalar single precision uops allocated.
239.It Li ARITH.FPU_DIV_ACTIVE
240.Pq Event 14H , Umask 01H
241Cycles that the divider is active, includes INT and FP.
242Set 'edge =1, cmask=1' to count the number of divides.
243.It Li L2_RQSTS.DEMAND_DATA_RD_HIT
244.Pq Event 24H , Umask 01H
245Demand Data Read requests that hit L2 cache.
246.It Li L2_RQSTS.ALL_DEMAND_DATA_RD
247.Pq Event 24H , Umask 03H
248Counts any demand and L1 HW prefetch data load requests to L2.
249.It Li L2_RQSTS.RFO_HITS
250.Pq Event 24H , Umask 04H
251Counts the number of store RFO requests that hit the L2 cache.
252.It Li L2_RQSTS.RFO_MISS
253.Pq Event 24H , Umask 08H
254Counts the number of store RFO requests that miss the L2 cache.
255.It Li L2_RQSTS.ALL_RFO
256.Pq Event 24H , Umask 0CH
257Counts all L2 store RFO requests.
258.It Li L2_RQSTS.CODE_RD_HIT
259.Pq Event 24H , Umask 10H
260Number of instruction fetches that hit the L2 cache.
261.It Li L2_RQSTS.CODE_RD_MISS
262.Pq Event 24H , Umask 20H
263Number of instruction fetches that missed the L2 cache.
264.It Li L2_RQSTS.ALL_CODE_RD
265.Pq Event 24H , Umask 30H
266Counts all L2 code requests.
267.It Li L2_RQSTS.PF_HIT
268.Pq Event 24H , Umask 40H
269Counts all L2 HW prefetcher requests that hit L2.
270.It Li L2_RQSTS.PF_MISS
271.Pq Event 24H , Umask 80H
272Counts all L2 HW prefetcher requests that missed L2.
273.It Li L2_RQSTS.ALL_PF
274.Pq Event 24H , Umask C0H
275Counts all L2 HW prefetcher requests.
276.It Li L2_STORE_LOCK_RQSTS.MISS
277.Pq Event 27H , Umask 01H
278RFOs that miss cache lines.
279.It Li L2_STORE_LOCK_RQSTS.HIT_M
280.Pq Event 27H , Umask 08H
281RFOs that hit cache lines in M state.
282.It Li L2_STORE_LOCK_RQSTS.ALL
283.Pq Event 27H , Umask 0FH
284RFOs that access cache lines in any state.
285.It Li L2_L1D_WB_RQSTS.MISS
286.Pq Event 28H , Umask 01H
287Not rejected writebacks that missed LLC.
288.It Li L2_L1D_WB_RQSTS.HIT_E
289.Pq Event 28H , Umask 04H
290Not rejected writebacks from L1D to L2 cache lines in E state.
291.It Li L2_L1D_WB_RQSTS.HIT_M
292.Pq Event 28H , Umask 08H
293Not rejected writebacks from L1D to L2 cache lines in M state.
294.It Li L2_L1D_WB_RQSTS.ALL
295.Pq Event 28H , Umask 0FH
296Not rejected writebacks from L1D to L2 cache lines in any state.
297.It Li LONGEST_LAT_CACHE.REFERENCE
298.Pq Event 2EH , Umask 4FH
299This event counts requests originating from the core that reference a cache
300line in the last level cache.
301.It Li LONGEST_LAT_CACHE.MISS
302.Pq Event 2EH , Umask 41H
303This event counts each cache miss condition for references to the last level
304cache.
305.It Li CPU_CLK_UNHALTED.THREAD_P
306.Pq Event 3CH , Umask 00H
307Counts the number of thread cycles while the thread is not in a halt state.
308The thread enters the halt state when it is running the HLT instruction.
309The core frequency may change from time to time due to power or thermal throttling.
310.It Li CPU_CLK_THREAD_UNHALTED.REF_XCLK
311.Pq Event 3CH , Umask 01H
312Increments at the frequency of XCLK (100 MHz) when not halted.
313.It Li L1D_PEND_MISS.PENDING
314.Pq Event 48H , Umask 01H
315Increments the number of outstanding L1D misses every cycle.
316Set Cmaks = 1 and Edge =1 to count occurrences.
317Counter 2 only.
318Set Cmask = 1 to count cycles.
319.It Li DTLB_STORE_MISSES.MISS_CAUSES_A_WALK
320.Pq Event 49H , Umask 01H
321Miss in all TLB levels causes an page walk of any page size (4K/2M/4M/1G).
322.It Li DTLB_STORE_MISSES.WALK_COMPLETED
323.Pq Event 49H , Umask 02H
324Miss in all TLB levels causes a page walk that completes of any page size
325(4K/2M/4M/1G).
326.It Li DTLB_STORE_MISSES.WALK_DURATION
327.Pq Event 49H , Umask 04H
328Cycles PMH is busy with this walk.
329.It Li DTLB_STORE_MISSES.STLB_HIT
330.Pq Event 49H , Umask 10H
331Store operations that miss the first TLB level but hit the second and do not
332cause page walks.
333.It Li LOAD_HIT_PRE.SW_PF
334.Pq Event 4CH , Umask 01H
335Non-SW-prefetch load dispatches that hit fill buffer allocated for S/W prefetch.
336.It Li LOAD_HIT_PRE.HW_PF
337.Pq Event 4CH , Umask 02H
338Non-SW-prefetch load dispatches that hit fill buffer allocated for H/W prefetch.
339.It Li L1D.REPLACEMENT
340.Pq Event 51H , Umask 01H
341Counts the number of lines brought into the L1 data cache.
342.It Li MOVE_ELIMINATION.INT_NOT_ELIMINATED
343.Pq Event 58H , Umask 01H
344Number of integer Move Elimination candidate uops that were not eliminated.
345.It Li MOVE_ELIMINATION.SIMD_NOT_ELIMINATED
346.Pq Event 58H , Umask 02H
347Number of SIMD Move Elimination candidate uops that were not eliminated.
348.It Li MOVE_ELIMINATION.INT_ELIMINATED
349.Pq Event 58H , Umask 04H
350Number of integer Move Elimination candidate uops that were eliminated.
351.It Li MOVE_ELIMINATION.SIMD_ELIMINATED
352.Pq Event 58H , Umask 08H
353Number of SIMD Move Elimination candidate uops that were eliminated.
354.It Li CPL_CYCLES.RING0
355.Pq Event 5CH , Umask 01H
356Unhalted core cycles when the thread is in ring 0.
357Use Edge to count transition.
358.It Li CPL_CYCLES.RING123
359.Pq Event 5CH , Umask 02H
360Unhalted core cycles when the thread is not in ring 0.
361.It Li RS_EVENTS.EMPTY_CYCLES
362.Pq Event 5EH , Umask 01H
363Cycles the RS is empty for the thread.
364.It Li DTLB_LOAD_MISSES.STLB_HIT
365.Pq Event 5FH , Umask 04H
366Counts load operations that missed 1st level DTLB but hit the 2nd level.
367.It Li OFFCORE_REQUESTS_OUTSTANDING.DEMAND_DATA_RD
368.Pq Event 60H , Umask 01H
369Offcore outstanding Demand Data Read transactions in SQ to uncore.
370Set Cmask=1 to count cycles.
371.It Li OFFCORE_REQUESTS_OUTSTANDING.DEMAND_CODE_RD
372.Pq Event 60H , Umask 02H
373Offcore outstanding Demand Code Read transactions in SQ to uncore.
374Set Cmask=1 to count cycles.
375.It Li OFFCORE_REQUESTS_OUTSTANDING.DEMAND_RFO
376.Pq Event 60H , Umask 04H
377Offcore outstanding RFO store transactions in SQ to uncore.
378Set Cmask=1 to count cycles.
379.It Li OFFCORE_REQUESTS_OUTSTANDING.ALL_DATA_RD
380.Pq Event 60H , Umask 08H
381Offcore outstanding cacheable data read transactions in SQ to uncore.
382Set Cmask=1 to count cycles.
383.It Li LOCK_CYCLES.SPLIT_LOCK_UC_LOCK_DURATION
384.Pq Event 63H , Umask 01H
385Cycles in which the L1D and L2 are locked, due to a UC lock or split lock.
386.It Li LOCK_CYCLES.CACHE_LOCK_DURATION
387.Pq Event 63H , Umask 02H
388Cycles in which the L1D is locked.
389.It Li IDQ.EMPTY
390.Pq Event 79H , Umask 02H
391Counts cycles the IDQ is empty.
392.It Li IDQ.MITE_UOPS
393.Pq Event 79H , Umask 04H
394Increment each cycle # of uops delivered to IDQ from MITE path.
395Can combine Umask 04H and 20H.
396Set Cmask = 1 to count cycles.
397.It Li IDQ.DSB_UOPS
398.Pq Event 79H , Umask 08H
399Increment each cycle. # of uops delivered to IDQ from DSB path.
400Can combine Umask 08H and 10H
401Set Cmask = 1 to count cycles.
402.It Li IDQ.MS_DSB_UOPS
403.Pq Event 79H , Umask 10H
404Increment each cycle # of uops delivered to IDQ when MS_busy by DSB.
405Set Cmask = 1 to count cycles.
406Add Edge=1 to count # of delivery.
407Can combine Umask 04H, 08H.
408.It Li IDQ.MS_MITE_UOPS
409.Pq Event 79H , Umask 20H
410Increment each cycle # of uops delivered to IDQ when MS_busy by MITE.
411Set Cmask = 1 to count cycles.
412Can combine Umask 04H, 08H.
413.It Li IDQ.MS_UOPS
414.Pq Event 79H , Umask 30H
415Increment each cycle # of uops delivered to IDQ from MS by either DSB or
416MITE.
417Set Cmask = 1 to count cycles.
418Can combine Umask 04H, 08H.
419.It Li IDQ.ALL_DSB_CYCLES_ANY_UOPS
420.Pq Event 79H , Umask 18H
421Counts cycles DSB is delivered at least one uops.
422Set Cmask = 1.
423.It Li IDQ.ALL_DSB_CYCLES_4_UOPS
424.Pq Event 79H , Umask 18H
425Counts cycles DSB is delivered four uops.
426Set Cmask = 4.
427.It Li IDQ.ALL_MITE_CYCLES_ANY_UOPS
428.Pq Event 79H , Umask 24H
429Counts cycles MITE is delivered at least one uops.
430Set Cmask = 1.
431.It Li IDQ.ALL_MITE_CYCLES_4_UOPS
432.Pq Event 79H , Umask 24H
433Counts cycles MITE is delivered four uops.
434Set Cmask = 4.
435.It Li IDQ.MITE_ALL_UOPS
436.Pq Event 79H , Umask 3CH
437# of uops delivered to IDQ from any path.
438.It Li ICACHE.MISSES
439.Pq Event 80H , Umask 02H
440Number of Instruction Cache, Streaming Buffer and Victim Cache Misses.
441Includes UC accesses.
442.It Li ITLB_MISSES.MISS_CAUSES_A_WALK
443.Pq Event 85H , Umask 01H
444Misses in all ITLB levels that cause page walks.
445.It Li ITLB_MISSES.WALK_COMPLETED
446.Pq Event 85H , Umask 02H
447Misses in all ITLB levels that cause completed page walks.
448.It Li ITLB_MISSES.WALK_DURATION
449.Pq Event 85H , Umask 04H
450Cycle PMH is busy with a walk.
451.It Li ITLB_MISSES.STLB_HIT
452.Pq Event 85H , Umask 10H
453Number of cache load STLB hits.
454No page walk.
455.It Li ILD_STALL.LCP
456.Pq Event 87H , Umask 01H
457Stalls caused by changing prefix length of the instruction.
458.It Li ILD_STALL.IQ_FULL
459.Pq Event 87H , Umask 04H
460Stall cycles due to IQ is full.
461.It Li BR_INST_EXEC.NONTAKEN_COND
462.Pq Event 88H , Umask 41H
463Count conditional near branch instructions that were executed (but not
464necessarily retired) and not taken.
465.It Li BR_INST_EXEC.TAKEN_COND
466.Pq Event 88H , Umask 81H
467Count conditional near branch instructions that were executed (but not
468necessarily retired) and taken.
469.It Li BR_INST_EXEC.DIRECT_JMP
470.Pq Event 88H , Umask 82H
471Count all unconditional near branch instructions excluding calls and
472indirect branches.
473.It Li BR_INST_EXEC.INDIRECT_JMP_NON_CALL_RET
474.Pq Event 88H , Umask 84H
475Count executed indirect near branch instructions that are not calls nor
476returns.
477.It Li BR_INST_EXEC.RETURN_NEAR
478.Pq Event 88H , Umask 88H
479Count indirect near branches that have a return mnemonic.
480.It Li BR_INST_EXEC.DIRECT_NEAR_CALL
481.Pq Event 88H , Umask 90H
482Count unconditional near call branch instructions, excluding non call
483branch, executed.
484.It Li BR_INST_EXEC.INDIRECT_NEAR_CALL
485.Pq Event 88H , Umask A0H
486Count indirect near calls, including both register and memory indirect,
487executed.
488.It Li BR_INST_EXEC.ALL_BRANCHES
489.Pq Event 88H , Umask FFH
490Counts all near executed branches (not necessarily retired).
491.It Li BR_MISP_EXEC.NONTAKEN_COND
492.Pq Event 89H , Umask 41H
493Count conditional near branch instructions mispredicted as nontaken.
494.It Li BR_MISP_EXEC.TAKEN_COND
495.Pq Event 89H , Umask 81H
496Count conditional near branch instructions mispredicted as taken.
497.It Li BR_MISP_EXEC.INDIRECT_JMP_NON_CALL_RET
498.Pq Event 89H , Umask 84H
499Count mispredicted indirect near branch instructions that are not calls
500nor returns.
501.It Li BR_MISP_EXEC.RETURN_NEAR
502.Pq Event 89H , Umask 88H
503Count mispredicted indirect near branches that have a return mnemonic.
504.It Li BR_MISP_EXEC.DIRECT_NEAR_CALL
505.Pq Event 89H , Umask 90H
506Count mispredicted unconditional near call branch instructions, excluding
507non call branch, executed.
508.It Li BR_MISP_EXEC.INDIRECT_NEAR_CALL
509.Pq Event 89H , Umask A0H
510Count mispredicted indirect near calls, including both register and memory
511indirect, executed.
512.It Li BR_MISP_EXEC.ALL_BRANCHES
513.Pq Event 89H , Umask FFH
514Counts all mispredicted near executed branches (not necessarily retired).
515.It Li IDQ_UOPS_NOT_DELIVERED.CORE
516.Pq Event 9CH , Umask 01H
517Count number of non-delivered uops to RAT per thread.
518Use Cmask to qualify uop b/w.
519.It Li UOPS_DISPATCHED_PORT.PORT_0
520.Pq Event A1H , Umask 01H
521Cycles which a Uop is dispatched on port 0.
522.It Li UOPS_DISPATCHED_PORT.PORT_1
523.Pq Event A1H , Umask 02H
524Cycles which a Uop is dispatched on port 1.
525.It Li UOPS_DISPATCHED_PORT.PORT_2_LD
526.Pq Event A1H , Umask 04H
527Cycles which a load uop is dispatched on port 2.
528.It Li UOPS_DISPATCHED_PORT.PORT_2_STA
529.Pq Event A1H , Umask 08H
530Cycles which a store address uop is dispatched on port 2.
531.It Li UOPS_DISPATCHED_PORT.PORT_2
532.Pq Event A1H , Umask 0CH
533Cycles which a Uop is dispatched on port 2.
534.It Li UOPS_DISPATCHED_PORT.PORT_3_LD
535.Pq Event A1H , Umask 10H
536Cycles which a load uop is dispatched on port 3.
537.It Li UOPS_DISPATCHED_PORT.PORT_3_STA
538.Pq Event A1H , Umask 20H
539Cycles which a store address uop is dispatched on port 3.
540.It Li UOPS_DISPATCHED_PORT.PORT_3
541.Pq Event A1H , Umask 30H
542Cycles which a Uop is dispatched on port 3.
543.It Li UOPS_DISPATCHED_PORT.PORT_4
544.Pq Event A1H , Umask 40H
545Cycles which a Uop is dispatched on port 4.
546.It Li UOPS_DISPATCHED_PORT.PORT_5
547.Pq Event A1H , Umask 80H
548Cycles which a Uop is dispatched on port 5.
549.It Li RESOURCE_STALLS.ANY
550.Pq Event A2H , Umask 01H
551Cycles Allocation is stalled due to Resource Related reason.
552.It Li RESOURCE_STALLS.RS
553.Pq Event A2H , Umask 04H
554Cycles stalled due to no eligible RS entry available.
555.It Li RESOURCE_STALLS.SB
556.Pq Event A2H , Umask 08H
557Cycles stalled due to no store buffers available. (not including draining
558form sync).
559.It Li RESOURCE_STALLS.ROB
560.Pq Event A2H , Umask 10H
561Cycles stalled due to re-order buffer full.
562.It Li CYCLE_ACTIVITY.CYCLES_L2_PENDING
563.Pq Event A3H , Umask 01H
564Cycles with pending L2 miss loads.
565Set AnyThread to count per core.
566.It Li CYCLE_ACTIVITY.CYCLES_LDM_PENDING
567.Pq Event A3H , Umask 02H
568Cycles with pending memory loads.
569Set AnyThread to count per core.
570.It Li CYCLE_ACTIVITY.CYCLES_NO_EXECUTE
571.Pq Event A3H , Umask 04H
572Cycles of dispatch stalls.
573Set AnyThread to count per core.
574.It Li CYCLE_ACTIVITY.CYCLES_L1D_PENDING
575.Pq Event A3H , Umask 08H
576Cycles with pending L1 cache miss loads.
577Set AnyThread to count per core.
578.It Li DSB2MITE_SWITCHES.COUNT
579.Pq Event ABH , Umask 01H
580Number of DSB to MITE switches.
581.It Li DSB2MITE_SWITCHES.PENALTY_CYCLES
582.Pq Event ABH , Umask 02H
583Cycles DSB to MITE switches caused delay.
584.It Li DSB_FILL.EXCEED_DSB_LINES
585.Pq Event ACH , Umask 08H
586DSB Fill encountered > 3 DSB lines.
587.It Li ITLB.ITLB_FLUSH
588.Pq Event AEH , Umask 01H
589Counts the number of ITLB flushes, includes 4k/2M/4M pages.
590.It Li OFFCORE_REQUESTS.DEMAND_DATA_RD
591.Pq Event B0H , Umask 01H
592Demand data read requests sent to uncore.
593.It Li OFFCORE_REQUESTS.DEMAND_CODE_RD
594.Pq Event B0H , Umask 02H
595Demand code read requests sent to uncore.
596.It Li OFFCORE_REQUESTS.DEMAND_RFO
597.Pq Event B0H , Umask 04H
598Demand RFO read requests sent to uncore, including regular RFOs, locks,
599ItoM.
600.It Li OFFCORE_REQUESTS.ALL_DATA_RD
601.Pq Event B0H , Umask 08H
602Data read requests sent to uncore (demand and prefetch).
603.It Li UOPS_EXECUTED.THREAD
604.Pq Event B1H , Umask 01H
605Counts total number of uops to be executed per-thread each cycle.
606Set Cmask = 1, INV =1 to count stall cycles.
607.It Li UOPS_EXECUTED.CORE
608.Pq Event B1H , Umask 02H
609Counts total number of uops to be executed per-core each cycle.
610Do not need to set ANY.
611.It Li OFF_CORE_RESPONSE_0
612.Pq Event B7H , Umask 01H
613Off-core Response Performance Monitoring.
614PMC0 only.
615Requires programming MSR 01A6H.
616.It Li OFF_CORE_RESPONSE_1
617.Pq Event BBH , Umask 01H
618Off-core Response Performance Monitoring.
619PMC3 only.
620Requires programming MSR 01A7H.
621.It Li TLB_FLUSH.DTLB_THREAD
622.Pq Event BDH , Umask 01H
623DTLB flush attempts of the thread- specific entries.
624.It Li TLB_FLUSH.STLB_ANY
625.Pq Event BDH , Umask 20H
626Count number of STLB flush attempts.
627.It Li INST_RETIRED.ANY_P
628.Pq Event C0H , Umask 00H
629Number of instructions at retirement.
630.It Li INST_RETIRED.ALL
631.Pq Event C0H , Umask 01H
632Precise instruction retired event with HW to reduce effect of PEBS shadow in
633IP distribution.
634PMC1 only.
635Must quiesce other PMCs.
636.It Li OTHER_ASSISTS.AVX_STORE
637.Pq Event C1H , Umask 08H
638Number of assists associated with 256-bit AVX store operations.
639.It Li OTHER_ASSISTS.AVX_TO_SSE
640.Pq Event C1H , Umask 10H
641Number of transitions from AVX- 256 to legacy SSE when penalty applicable.
642.It Li OTHER_ASSISTS.SSE_TO_AVX
643.Pq Event C1H , Umask 20H
644Number of transitions from SSE to AVX-256 when penalty applicable.
645.It Li UOPS_RETIRED.ALL
646.Pq Event C2H , Umask 01H
647Counts the number of micro-ops retired, Use cmask=1 and invert to count
648active cycles or stalled cycles.
649Supports PEBS, use Any=1 for core granular.
650.It Li UOPS_RETIRED.RETIRE_SLOTS
651.Pq Event C2H , Umask 02H
652Counts the number of retirement slots used each cycle.
653.It Li MACHINE_CLEARS.MEMORY_ORDERING
654.Pq Event C3H , Umask 02H
655Counts the number of machine clears due to memory order conflicts.
656.It Li MACHINE_CLEARS.SMC
657.Pq Event C3H , Umask 04H
658Number of self-modifying-code machine clears detected.
659.It Li MACHINE_CLEARS.MASKMOV
660.Pq Event C3H , Umask 20H
661Counts the number of executed AVX masked load operations that refer to an
662illegal address range with the mask bits set to 0.
663.It Li BR_INST_RETIRED.ALL_BRANCHES
664.Pq Event C4H , Umask 00H
665Branch instructions at retirement.
666.It Li BR_INST_RETIRED.CONDITIONAL
667.Pq Event C4H , Umask 01H
668Counts the number of conditional branch instructions retired.
669Supports PEBS.
670.It Li BR_INST_RETIRED.NEAR_CALL
671.Pq Event C4H , Umask 02H
672Direct and indirect near call instructions retired.
673.It Li BR_INST_RETIRED.ALL_BRANCHES
674.Pq Event C4H , Umask 04H
675Counts the number of branch instructions retired.
676.It Li BR_INST_RETIRED.NEAR_RETURN
677.Pq Event C4H , Umask 08H
678Counts the number of near return instructions retired.
679.It Li BR_INST_RETIRED.NOT_TAKEN
680.Pq Event C4H , Umask 10H
681Counts the number of not taken branch instructions retired.
682.It Li BR_INST_RETIRED.NEAR_TAKEN
683.Pq Event C4H , Umask 20H
684Number of near taken branches retired.
685.It Li BR_INST_RETIRED.FAR_BRANCH
686.Pq Event C4H , Umask 40H
687Number of far branches retired.
688.It Li BR_MISP_RETIRED.ALL_BRANCHES
689.Pq Event C5H , Umask 00H
690Mispredicted branch instructions at retirement.
691.It Li BR_MISP_RETIRED.CONDITIONAL
692.Pq Event C5H , Umask 01H
693Mispredicted conditional branch instructions retired.
694Supports PEBS.
695.It Li BR_MISP_RETIRED.NEAR_CALL
696.Pq Event C5H , Umask 02H
697Direct and indirect mispredicted near call instructions retired.
698.It Li BR_MISP_RETIRED.ALL_BRANCHES
699.Pq Event C5H , Umask 04H
700Mispredicted macro branch instructions retired.
701.It Li BR_MISP_RETIRED.NOT_TAKEN
702.Pq Event C5H , Umask 10H
703Mispredicted not taken branch instructions retired.
704.It Li BR_MISP_RETIRED.TAKEN
705.Pq Event C5H , Umask 20H
706Mispredicted taken branch instructions retired.
707.It Li FP_ASSIST.X87_OUTPUT
708.Pq Event CAH , Umask 02H
709Number of X87 FP assists due to Output values.
710.It Li FP_ASSIST.X87_INPUT
711.Pq Event CAH , Umask 04H
712Number of X87 FP assists due to input values.
713.It Li FP_ASSIST.SIMD_OUTPUT
714.Pq Event CAH , Umask 08H
715Number of SIMD FP assists due to Output values.
716.It Li FP_ASSIST.SIMD_INPUT
717.Pq Event CAH , Umask 10H
718Number of SIMD FP assists due to input values.
719.It Li FP_ASSIST.ANY
720.Pq Event CAH , Umask 1EH
721Cycles with any input/output SSE* or FP assists.
722.It Li ROB_MISC_EVENTS.LBR_INSERTS
723.Pq Event CCH , Umask 20H
724Count cases of saving new LBR records by hardware.
725.It Li MEM_TRANS_RETIRED.LOAD_LATENCY
726.Pq Event CDH , Umask 01H
727Sample loads with specified latency threshold.
728PMC3 only.
729Specify threshold in MSR 0x3F6.
730.It Li MEM_TRANS_RETIRED.PRECISE_STORE
731.Pq Event CDH , Umask 02H
732Sample stores and collect precise store operation via PEBS record.
733PMC3 only.
734.It Li MEM_UOPS_RETIRED.STLB_MISS_LOADS
735.Pq Event D0H , Umask 11H
736Count retired load uops that missed the STLB.
737.It Li MEM_UOPS_RETIRED.STLB_MISS_STORES
738.Pq Event D0H , Umask 12H
739Count retired store uops that missed the STLB.
740.It Li MEM_UOPS_RETIRED.SPLIT_LOADS
741.Pq Event D0H , Umask 41H
742Count retired load uops that were split across a cache line.
743.It Li MEM_UOPS_RETIRED.SPLIT_STORES
744.Pq Event D0H , Umask 42H
745Count retired store uops that were split across a cache line.
746.It Li MEM_UOPS_RETIRED.ALL_LOADS
747.Pq Event D0H , Umask 81H
748Count all retired load uops.
749.It Li MEM_UOPS_RETIRED.ALL_STORES
750.Pq Event D0H , Umask 82H
751Count all retired store uops.
752.It Li MEM_LOAD_UOPS_RETIRED.L1_HIT
753.Pq Event D1H , Umask 01H
754Retired load uops with L1 cache hits as data sources.
755Supports PEBS.
756.It Li MEM_LOAD_UOPS_RETIRED.L2_HIT
757.Pq Event D1H , Umask 02H
758Retired load uops with L2 cache hits as data sources.
759.It Li MEM_LOAD_UOPS_RETIRED.LLC_HIT
760.Pq Event D1H , Umask 04H
761Retired load uops whose data source was LLC hit with no snoop required.
762.It Li MEM_LOAD_UOPS_RETIRED.LLC_MISS
763.Pq Event D1H , Umask 20H
764Retired load uops whose data source is LLC miss.
765.It Li MEM_LOAD_UOPS_RETIRED.HIT_LFB
766.Pq Event D1H , Umask 40H
767Retired load uops which data sources were load uops missed L1 but hit FB due
768to preceding miss to the same cache line with data not ready.
769.It Li MEM_LOAD_UOPS_LLC_HIT_RETIRED.XSNP_MISS
770.Pq Event D2H , Umask 01H
771Retired load uops which data sources were LLC hit and cross-core snoop
772missed in on-pkg core cache.
773Supports PEBS.
774.It Li MEM_LOAD_UOPS_LLC_HIT_RETIRED.XSNP_HIT
775.Pq Event D2H , Umask 02H
776Retired load uops which data sources were LLC and cross-core snoop hits in
777on-pkg core cache.
778Supports PEBS.
779.It Li MEM_LOAD_UOPS_LLC_HIT_RETIRED.XSNP_HITM
780.Pq Event D2H , Umask 04H
781Retired load uops which data sources were HitM responses from shared LLC.
782.It Li MEM_LOAD_UOPS_LLC_HIT_RETIRED.XSNP_NONE
783.Pq Event D2H , Umask 08H
784Retired load uops which data sources were hits in LLC without snoops
785required.
786.It Li MEM_LOAD_UOPS_LLC_MISS_RETIRED.LOCAL_DRAM
787.Pq Event D3H , Umask 01H
788Retired load uops which data sources missed LLC but serviced from local
789dram.
790Supports PEBS.
791.It Li MEM_LOAD_UOPS_LLC_MISS_RETIRED.REMOTE_DRAM
792.Pq Event D3H , Umask 04H
793Retired load uops whose data source was remote DRAM.
794.It Li MEM_LOAD_UOPS_LLC_MISS_RETIRED.REMOTE_HITM
795.Pq Event D3H , Umask 10H
796Retired load uops whose data source was remote HITM.
797.It Li MEM_LOAD_UOPS_LLC_MISS_RETIRED.REMOTE_FWD
798.Pq Event D3H , Umask 20H
799Retired load uops whose data source was forwards from a remote cache.
800.It Li BACLEARS.ANY
801.Pq Event E6H , Umask 1FH
802Number of front end re-steers due to BPU misprediction.
803.It Li L2_TRANS.DEMAND_DATA_RD
804.Pq Event F0H , Umask 01H
805Demand Data Read requests that access L2 cache.
806.It Li L2_TRANS.RFO
807.Pq Event F0H , Umask 02H
808RFO requests that access L2 cache.
809.It Li L2_TRANS.CODE_RD
810.Pq Event F0H , Umask 04H
811L2 cache accesses when fetching instructions.
812.It Li L2_TRANS.ALL_PF
813.Pq Event F0H , Umask 08H
814Any MLC or LLC HW prefetch accessing L2, including rejects.
815.It Li L2_TRANS.L1D_WB
816.Pq Event F0H , Umask 10H
817L1D writebacks that access L2 cache.
818.It Li L2_TRANS.L2_FILL
819.Pq Event F0H , Umask 20H
820L2 fill requests that access L2 cache.
821.It Li L2_TRANS.L2_WB
822.Pq Event F0H , Umask 40H
823L2 writebacks that access L2 cache.
824.It Li L2_TRANS.ALL_REQUESTS
825.Pq Event F0H , Umask 80H
826Transactions accessing L2 pipe.
827.It Li L2_LINES_IN.I
828.Pq Event F1H , Umask 01H
829L2 cache lines in I state filling L2.
830Counting does not cover rejects.
831.It Li L2_LINES_IN.S
832.Pq Event F1H , Umask 02H
833L2 cache lines in S state filling L2.
834Counting does not cover rejects.
835.It Li L2_LINES_IN.E
836.Pq Event F1H , Umask 04H
837L2 cache lines in E state filling L2.
838Counting does not cover rejects.
839.It Li L2_LINES_IN.ALL
840.Pq Event F1H , Umask 07H
841L2 cache lines filling L2.
842Counting does not cover rejects.
843.It Li L2_LINES_OUT.DEMAND_CLEAN
844.Pq Event F2H , Umask 01H
845Clean L2 cache lines evicted by demand.
846.It Li L2_LINES_OUT.DEMAND_DIRTY
847.Pq Event F2H , Umask 02H
848Dirty L2 cache lines evicted by demand.
849.It Li L2_LINES_OUT.PF_CLEAN
850.Pq Event F2H , Umask 04H
851Clean L2 cache lines evicted by the MLC prefetcher.
852.It Li L2_LINES_OUT.PF_DIRTY
853.Pq Event F2H , Umask 08H
854Dirty L2 cache lines evicted by the MLC prefetcher.
855.It Li L2_LINES_OUT.DIRTY_ALL
856.Pq Event F2H , Umask 0AH
857Dirty L2 cache lines filling the L2.
858.El
859.Sh SEE ALSO
860.Xr pmc 3 ,
861.Xr pmc.atom 3 ,
862.Xr pmc.core 3 ,
863.Xr pmc.corei7 3 ,
864.Xr pmc.corei7uc 3 ,
865.Xr pmc.iaf 3 ,
866.Xr pmc.ivybridge 3 ,
867.Xr pmc.k7 3 ,
868.Xr pmc.k8 3 ,
869.Xr pmc.sandybridge 3 ,
870.Xr pmc.sandybridgeuc 3 ,
871.Xr pmc.sandybridgexeon 3 ,
872.Xr pmc.soft 3 ,
873.Xr pmc.tsc 3 ,
874.Xr pmc.ucf 3 ,
875.Xr pmc.westmere 3 ,
876.Xr pmc.westmereuc 3 ,
877.Xr pmc_cpuinfo 3 ,
878.Xr pmclog 3 ,
879.Xr hwpmc 4
880.Sh HISTORY
881The
882.Nm pmc
883library first appeared in
884.Fx 6.0 .
885.Sh AUTHORS
886.An -nosplit
887The
888.Lb libpmc
889library was written by
890.An Joseph Koshy Aq Mt jkoshy@FreeBSD.org .
891The support for the Ivy Bridge Xeon
892microarchitecture was written by
893.An Hiren Panchasara Aq Mt hiren.panchasara@gmail.com .
894