xref: /freebsd/lib/libpmc/pmc.haswellxeon.3 (revision eda14cbc264d6969b02f2b1994cef11148e914f1)
1.\"
2.\" Copyright (c) 2013 Hiren Panchasara <hiren.panchasara@gmail.com>
3.\" All rights reserved.
4.\"
5.\" Redistribution and use in source and binary forms, with or without
6.\" modification, are permitted provided that the following conditions
7.\" are met:
8.\" 1. Redistributions of source code must retain the above copyright
9.\"    notice, this list of conditions and the following disclaimer.
10.\" 2. Redistributions in binary form must reproduce the above copyright
11.\"    notice, this list of conditions and the following disclaimer in the
12.\"    documentation and/or other materials provided with the distribution.
13.\"
14.\" THIS SOFTWARE IS PROVIDED BY THE AUTHOR AND CONTRIBUTORS ``AS IS'' AND
15.\" ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
16.\" IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
17.\" ARE DISCLAIMED.  IN NO EVENT SHALL THE AUTHOR OR CONTRIBUTORS BE LIABLE
18.\" FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
19.\" DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
20.\" OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
21.\" HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
22.\" LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
23.\" OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
24.\" SUCH DAMAGE.
25.\"
26.\" $FreeBSD$
27.\"
28.Dd November 21, 2014
29.Dt PMC.HASWELLXEON 3
30.Os
31.Sh NAME
32.Nm pmc.haswellxeon
33.Nd measurement events for
34.Tn Intel
35.Tn Haswell Xeon
36family CPUs
37.Sh LIBRARY
38.Lb libpmc
39.Sh SYNOPSIS
40.In pmc.h
41.Sh DESCRIPTION
42.Tn Intel
43.Tn "Haswell"
44CPUs contain PMCs conforming to version 2 of the
45.Tn Intel
46performance measurement architecture.
47These CPUs may contain up to two classes of PMCs:
48.Bl -tag -width "Li PMC_CLASS_IAP"
49.It Li PMC_CLASS_IAF
50Fixed-function counters that count only one hardware event per counter.
51.It Li PMC_CLASS_IAP
52Programmable counters that may be configured to count one of a defined
53set of hardware events.
54.El
55.Pp
56The number of PMCs available in each class and their widths need to be
57determined at run time by calling
58.Xr pmc_cpuinfo 3 .
59.Pp
60Intel Haswell Xeon PMCs are documented in
61.Rs
62.%B "Intel(R) 64 and IA-32 Architectures Software Developer's Manual"
63.%T "Combined Volumes: 1, 2A, 2B, 2C, 3A, 3B and 3C"
64.%N "Order Number: 325462-052US"
65.%D September 2014
66.%Q "Intel Corporation"
67.Re
68.Ss HASWELL FIXED FUNCTION PMCS
69These PMCs and their supported events are documented in
70.Xr pmc.iaf 3 .
71.Ss HASWELL PROGRAMMABLE PMCS
72The programmable PMCs support the following capabilities:
73.Bl -column "PMC_CAP_INTERRUPT" "Support"
74.It Em Capability Ta Em Support
75.It PMC_CAP_CASCADE Ta \&No
76.It PMC_CAP_EDGE Ta Yes
77.It PMC_CAP_INTERRUPT Ta Yes
78.It PMC_CAP_INVERT Ta Yes
79.It PMC_CAP_READ Ta Yes
80.It PMC_CAP_PRECISE Ta \&No
81.It PMC_CAP_SYSTEM Ta Yes
82.It PMC_CAP_TAGGING Ta \&No
83.It PMC_CAP_THRESHOLD Ta Yes
84.It PMC_CAP_USER Ta Yes
85.It PMC_CAP_WRITE Ta Yes
86.El
87.Ss Event Qualifiers
88Event specifiers for these PMCs support the following common
89qualifiers:
90.Bl -tag -width indent
91.It Li rsp= Ns Ar value
92Configure the Off-core Response bits.
93.Bl -tag -width indent
94.It Li DMND_DATA_RD
95Counts the number of demand and DCU prefetch data reads of full
96and partial cachelines as well as demand data page table entry
97cacheline reads. Does not count L2 data read prefetches or
98instruction fetches.
99.It Li REQ_DMND_RFO
100Counts the number of demand and DCU prefetch reads for ownership (RFO)
101requests generated by a write to data cacheline. Does not count L2 RFO
102prefetches.
103.It Li REQ_DMND_IFETCH
104Counts the number of demand and DCU prefetch instruction cacheline reads.
105Does not count L2 code read prefetches.
106.It Li REQ_WB
107Counts the number of writeback (modified to exclusive) transactions.
108.It Li REQ_PF_DATA_RD
109Counts the number of data cacheline reads generated by L2 prefetchers.
110.It Li REQ_PF_RFO
111Counts the number of RFO requests generated by L2 prefetchers.
112.It Li REQ_PF_IFETCH
113Counts the number of code reads generated by L2 prefetchers.
114.It Li REQ_PF_LLC_DATA_RD
115L2 prefetcher to L3 for loads.
116.It Li REQ_PF_LLC_RFO
117RFO requests generated by L2 prefetcher
118.It Li REQ_PF_LLC_IFETCH
119L2 prefetcher to L3 for instruction fetches.
120.It Li REQ_BUS_LOCKS
121Bus lock and split lock requests.
122.It Li REQ_STRM_ST
123Streaming store requests.
124.It Li REQ_OTHER
125Any other request that crosses IDI, including I/O.
126.It Li RES_ANY
127Catch all value for any response types.
128.It Li RES_SUPPLIER_NO_SUPP
129No Supplier Information available.
130.It Li RES_SUPPLIER_LLC_HITM
131M-state initial lookup stat in L3.
132.It Li RES_SUPPLIER_LLC_HITE
133E-state.
134.It Li RES_SUPPLIER_LLC_HITS
135S-state.
136.It Li RES_SUPPLIER_LLC_HITF
137F-state.
138.It Li RES_SUPPLIER_LOCAL
139Local DRAM Controller.
140.It Li RES_SNOOP_SNP_NONE
141No details on snoop-related information.
142.It Li RES_SNOOP_SNP_NO_NEEDED
143No snoop was needed to satisfy the request.
144.It Li RES_SNOOP_SNP_MISS
145A snoop was needed and it missed all snooped caches:
146-For LLC Hit, ReslHitl was returned by all cores
147-For LLC Miss, Rspl was returned by all sockets and data was returned from
148DRAM.
149.It Li RES_SNOOP_HIT_NO_FWD
150A snoop was needed and it hits in at least one snooped cache. Hit denotes a
151cache-line was valid before snoop effect. This includes:
152-Snoop Hit w/ Invalidation (LLC Hit, RFO)
153-Snoop Hit, Left Shared (LLC Hit/Miss, IFetch/Data_RD)
154-Snoop Hit w/ Invalidation and No Forward (LLC Miss, RFO Hit S)
155In the LLC Miss case, data is returned from DRAM.
156.It Li RES_SNOOP_HIT_FWD
157A snoop was needed and data was forwarded from a remote socket.
158This includes:
159-Snoop Forward Clean, Left Shared (LLC Hit/Miss, IFetch/Data_RD/RFT).
160.It Li RES_SNOOP_HITM
161A snoop was needed and it HitM-ed in local or remote cache. HitM denotes a
162cache-line was in modified state before effect as a results of snoop. This
163includes:
164-Snoop HitM w/ WB (LLC miss, IFetch/Data_RD)
165-Snoop Forward Modified w/ Invalidation (LLC Hit/Miss, RFO)
166-Snoop MtoS (LLC Hit, IFetch/Data_RD).
167.It Li RES_NON_DRAM
168Target was non-DRAM system address. This includes MMIO transactions.
169.El
170.It Li cmask= Ns Ar value
171Configure the PMC to increment only if the number of configured
172events measured in a cycle is greater than or equal to
173.Ar value .
174.It Li edge
175Configure the PMC to count the number of de-asserted to asserted
176transitions of the conditions expressed by the other qualifiers.
177If specified, the counter will increment only once whenever a
178condition becomes true, irrespective of the number of clocks during
179which the condition remains true.
180.It Li inv
181Invert the sense of comparison when the
182.Dq Li cmask
183qualifier is present, making the counter increment when the number of
184events per cycle is less than the value specified by the
185.Dq Li cmask
186qualifier.
187.It Li os
188Configure the PMC to count events happening at processor privilege
189level 0.
190.It Li usr
191Configure the PMC to count events occurring at privilege levels 1, 2
192or 3.
193.El
194.Pp
195If neither of the
196.Dq Li os
197or
198.Dq Li usr
199qualifiers are specified, the default is to enable both.
200.Ss Event Specifiers (Programmable PMCs)
201Haswell programmable PMCs support the following events:
202.Bl -tag -width indent
203.It Li LD_BLOCKS.STORE_FORWARD
204.Pq Event 03H , Umask 02H
205Loads blocked by overlapping with store buffer that
206cannot be forwarded.
207.It Li MISALIGN_MEM_REF.LOADS
208.Pq Event 05H , Umask 01H
209Speculative cache-line split load uops dispatched to
210L1D.
211.It Li MISALIGN_MEM_REF.STORES
212.Pq Event 05H , Umask 02H
213Speculative cache-line split Store-address uops
214dispatched to L1D.
215.It Li LD_BLOCKS_PARTIAL.ADDRESS_ALIAS
216.Pq Event 07H , Umask 01H
217False dependencies in MOB due to partial compare
218on address.
219.It Li DTLB_LOAD_MISSES.MISS_CAUSES_A_WALK
220.Pq Event 08H , Umask 01H
221Misses in all TLB levels that cause a page walk of any
222page size.
223.It Li DTLB_LOAD_MISSES.WALK_COMPLETED_4K
224.Pq Event 08H , Umask 02H
225Completed page walks due to demand load misses
226that caused 4K page walks in any TLB levels.
227.It Li DTLB_LOAD_MISSES.WALK_COMPLETED_2M_4K
228.Pq Event 08H , Umask 02H
229Completed page walks due to demand load misses
230that caused 2M/4M page walks in any TLB levels.
231.It Li DTLB_LOAD_MISSES.WALK_COMPLETED
232.Pq Event 08H , Umask 0EH
233Completed page walks in any TLB of any page size
234due to demand load misses
235.It Li DTLB_LOAD_MISSES.WALK_DURATION
236.Pq Event 08H , Umask 10H
237Cycle PMH is busy with a walk.
238.It Li DTLB_LOAD_MISSES.STLB_HIT_4K
239.Pq Event 08H , Umask 20H
240Load misses that missed DTLB but hit STLB (4K).
241.It Li DTLB_LOAD_MISSES.STLB_HIT_2M
242.Pq Event 08H , Umask 40H
243Load misses that missed DTLB but hit STLB (2M).
244.It Li DTLB_LOAD_MISSES.STLB_HIT
245.Pq Event 08H , Umask 60H
246Number of cache load STLB hits. No page walk.
247.It Li DTLB_LOAD_MISSES.PDE_CACHE_MISS
248.Pq Event 08H , Umask 80H
249DTLB demand load misses with low part of linear-to-
250physical address translation missed
251.It Li INT_MISC.RECOVERY_CYCLES
252.Pq Event 0DH , Umask 03H
253Cycles waiting to recover after Machine Clears
254except JEClear. Set Cmask= 1.
255.It Li UOPS_ISSUED.ANY
256.Pq Event 0EH , Umask 01H
257ncrements each cycle the # of Uops issued by the
258RAT to RS.
259Set Cmask = 1, Inv = 1, Any= 1to count stalled cycles
260of this core.
261.It Li UOPS_ISSUED.FLAGS_MERGE
262.Pq Event 0EH , Umask 10H
263Number of flags-merge uops allocated. Such uops
264adds delay.
265.It Li UOPS_ISSUED.SLOW_LEA
266.Pq Event 0EH , Umask 20H
267Number of slow LEA or similar uops allocated. Such
268uop has 3 sources (e.g. 2 sources + immediate)
269regardless if as a result of LEA instruction or not.
270.It Li UOPS_ISSUED.SiNGLE_MUL
271.Pq Event 0EH , Umask 40H
272Number of multiply packed/scalar single precision
273uops allocated.
274.It Li L2_RQSTS.DEMAND_DATA_RD_MISS
275.Pq Event 24H , Umask 21H
276Demand Data Read requests that missed L2, no
277rejects.
278.It Li L2_RQSTS.DEMAND_DATA_RD_HIT
279.Pq Event 24H , Umask 41H
280Demand Data Read requests that hit L2 cache.
281.It Li L2_RQSTS.ALL_DEMAND_DATA_RD
282.Pq Event 24H , Umask E1H
283Counts any demand and L1 HW prefetch data load
284requests to L2.
285.It Li L2_RQSTS.RFO_HIT
286.Pq Event 24H , Umask 42H
287Counts the number of store RFO requests that hit
288the L2 cache.
289.It Li L2_RQSTS.RFO_MISS
290.Pq Event 24H , Umask 22H
291Counts the number of store RFO requests that miss
292the L2 cache.
293.It Li L2_RQSTS.ALL_RFO
294.Pq Event 24H , Umask E2H
295Counts all L2 store RFO requests.
296.It Li L2_RQSTS.CODE_RD_HIT
297.Pq Event 24H , Umask 44H
298Number of instruction fetches that hit the L2 cache.
299.It Li L2_RQSTS.CODE_RD_MISS
300.Pq Event 24H , Umask 24H
301Number of instruction fetches that missed the L2
302cache.
303.It Li L2_RQSTS.ALL_DEMAND_MISS
304.Pq Event 24H , Umask 27H
305Demand requests that miss L2 cache.
306.It Li L2_RQSTS.ALL_DEMAND_REFERENCES
307.Pq Event 24H , Umask E7H
308Demand requests to L2 cache.
309.It Li L2_RQSTS.ALL_CODE_RD
310.Pq Event 24H , Umask E4H
311Counts all L2 code requests.
312.It Li L2_RQSTS.L2_PF_HIT
313.Pq Event 24H , Umask 50H
314Counts all L2 HW prefetcher requests that hit L2.
315.It Li L2_RQSTS.L2_PF_MISS
316.Pq Event 24H , Umask 30H
317Counts all L2 HW prefetcher requests that missed
318L2.
319.It Li L2_RQSTS.ALL_PF
320.Pq Event 24H , Umask F8H
321Counts all L2 HW prefetcher requests.
322.It Li L2_RQSTS.MISS
323.Pq Event 24H , Umask 3FH
324All requests that missed L2.
325.It Li L2_RQSTS.REFERENCES
326.Pq Event 24H , Umask FFH
327All requests to L2 cache.
328.It Li L2_DEMAND_RQSTS.WB_HIT
329.Pq Event 27H , Umask 50H
330Not rejected writebacks that hit L2 cache
331.It Li LONGEST_LAT_CACHE.REFERENCE
332.Pq Event 2EH , Umask 4FH
333This event counts requests originating from the core
334that reference a cache line in the last level cache.
335.It Li LONGEST_LAT_CACHE.MISS
336.Pq Event 2EH , Umask 41H
337This event counts each cache miss condition for
338references to the last level cache.
339.It Li CPU_CLK_UNHALTED.THREAD_P
340.Pq Event 3CH , Umask 00H
341Counts the number of thread cycles while the thread
342is not in a halt state. The thread enters the halt state
343when it is running the HLT instruction. The core
344frequency may change from time to time due to
345power or thermal throttling.
346.It Li CPU_CLK_THREAD_UNHALTED.REF_XCLK
347.Pq Event 3CH , Umask 01H
348Increments at the frequency of XCLK (100 MHz)
349when not halted.
350.It Li L1D_PEND_MISS.PENDING
351.Pq Event 48H , Umask 01H
352Increments the number of outstanding L1D misses
353every cycle. Set Cmaks = 1 and Edge =1 to count
354occurrences.
355.It Li DTLB_STORE_MISSES.MISS_CAUSES_A_WALK
356.Pq Event 49H , Umask 01H
357Miss in all TLB levels causes an page walk of any
358page size (4K/2M/4M/1G).
359.It Li DTLB_STORE_MISSES.WALK_COMPLETED_4K
360.Pq Event 49H , Umask 02H
361Completed page walks due to store misses in one or
362more TLB levels of 4K page structure.
363.It Li DTLB_STORE_MISSES.WALK_COMPLETED_2M_4M
364.Pq Event 49H , Umask 04H
365Completed page walks due to store misses in one or
366more TLB levels of 2M/4M page structure.
367.It Li DTLB_STORE_MISSES.WALK_COMPLETED
368.Pq Event 49H , Umask 0EH
369Completed page walks due to store miss in any TLB
370levels of any page size (4K/2M/4M/1G).
371.It Li DTLB_STORE_MISSES.WALK_DURATION
372.Pq Event 49H , Umask 10H
373Cycles PMH is busy with this walk.
374.It Li DTLB_STORE_MISSES.STLB_HIT_4K
375.Pq Event 49H , Umask 20H
376Store misses that missed DTLB but hit STLB (4K).
377.It Li DTLB_STORE_MISSES.STLB_HIT_2M
378.Pq Event 49H , Umask 40H
379Store misses that missed DTLB but hit STLB (2M).
380.It Li DTLB_STORE_MISSES.STLB_HIT
381.Pq Event 49H , Umask 60H
382Store operations that miss the first TLB level but hit
383the second and do not cause page walks.
384.It Li DTLB_STORE_MISSES.PDE_CACHE_MISS
385.Pq Event 49H , Umask 80H
386DTLB store misses with low part of linear-to-physical
387address translation missed.
388.It Li LOAD_HIT_PRE.SW_PF
389.Pq Event 4CH , Umask 01H
390Non-SW-prefetch load dispatches that hit fill buffer
391allocated for S/W prefetch.
392.It Li LOAD_HIT_PRE.HW_PF
393.Pq Event 4CH , Umask 02H
394Non-SW-prefetch load dispatches that hit fill buffer
395allocated for H/W prefetch.
396.It Li L1D.REPLACEMENT
397.Pq Event 51H , Umask 01H
398Counts the number of lines brought into the L1 data
399cache.
400.It Li MOVE_ELIMINATION.INT_NOT_ELIMINATED
401.Pq Event 58H , Umask 04H
402Number of integer Move Elimination candidate uops
403that were not eliminated.
404.It Li MOVE_ELIMINATION.SMID_NOT_ELIMINATED
405.Pq Event 58H , Umask 08H
406Number of SIMD Move Elimination candidate uops
407that were not eliminated.
408.It Li MOVE_ELIMINATION.INT_ELIMINATED
409.Pq Event 58H , Umask 01H
410Unhalted core cycles when the thread is in ring 0.
411.It Li MOVE_ELIMINATION.SMID_ELIMINATED
412.Pq Event 58H , Umask 02H
413Number of SIMD Move Elimination candidate uops
414that were eliminated.
415.It Li CPL_CYCLES.RING0
416.Pq Event 5CH , Umask 02H
417Unhalted core cycles when the thread is in ring 0.
418.It Li CPL_CYCLES.RING123
419.Pq Event 5CH , Umask 01H
420Unhalted core cycles when the thread is not in ring 0.
421.It Li RS_EVENTS.EMPTY_CYCLES
422.Pq Event 5EH , Umask 01H
423Cycles the RS is empty for the thread.
424.It Li OFFCORE_REQUESTS_OUTSTANDING.DEMAND_DATA_RD
425.Pq Event 60H , Umask 01H
426Offcore outstanding Demand Data Read transactions
427in SQ to uncore. Set Cmask=1 to count cycles.
428.It Li OFFCORE_REQUESTS_OUTSTANDING.DEMAND_CORE_RD
429.Pq Event 60H , Umask 02H
430Offcore outstanding Demand code Read transactions
431in SQ to uncore. Set Cmask=1 to count cycles.
432.It Li OFFCORE_REQUESTS_OUTSTANDING.DEMAND_RFO
433.Pq Event 60H , Umask 04H
434Offcore outstanding RFO store transactions in SQ to
435uncore. Set Cmask=1 to count cycles.
436.It Li OFFCORE_REQUESTS_OUTSTANDING.ALL_DATA_RD
437.Pq Event 60H , Umask 08H
438Offcore outstanding cacheable data read
439transactions in SQ to uncore. Set Cmask=1 to count
440cycles.
441.It Li LOCK_CYCLES.SPLIT_LOCK_UC_LOCK_DURATION
442.Pq Event 63H , Umask 01H
443Cycles in which the L1D and L2 are locked, due to a
444UC lock or split lock.
445.It Li LOCK_CYCLES.CACHE_LOCK_DURATION
446.Pq Event 63H , Umask 02H
447Cycles in which the L1D is locked.
448.It Li IDQ.EMPTY
449.Pq Event 79H , Umask 02H
450Counts cycles the IDQ is empty.
451.It Li IDQ.MITE_UOPS
452.Pq Event 79H , Umask 04H
453Increment each cycle # of uops delivered to IDQ from
454MITE path.
455Set Cmask = 1 to count cycles.
456.It Li IDQ.DSB_UOPS
457.Pq Event 79H , Umask 08H
458Increment each cycle. # of uops delivered to IDQ
459from DSB path.
460Set Cmask = 1 to count cycles.
461.It Li IDQ.MS_DSB_UOPS
462.Pq Event 79H , Umask 10H
463Increment each cycle # of uops delivered to IDQ
464when MS_busy by DSB. Set Cmask = 1 to count
465cycles. Add Edge=1 to count # of delivery.
466.It Li IDQ.MS_MITE_UOPS
467.Pq Event 79H , Umask 20H
468ncrement each cycle # of uops delivered to IDQ
469when MS_busy by MITE. Set Cmask = 1 to count
470cycles.
471.It Li IDQ.MS_UOPS
472.Pq Event 79H , Umask 30H
473Increment each cycle # of uops delivered to IDQ from
474MS by either DSB or MITE. Set Cmask = 1 to count
475cycles.
476.It Li IDQ.ALL_DSB_CYCLES_ANY_UOPS
477.Pq Event 79H , Umask 18H
478Counts cycles DSB is delivered at least one uops. Set
479Cmask = 1.
480.It Li IDQ.ALL_DSB_CYCLES_4_UOPS
481.Pq Event 79H , Umask 18H
482Counts cycles DSB is delivered four uops. Set Cmask
483=4.
484.It Li IDQ.ALL_MITE_CYCLES_ANY_UOPS
485.Pq Event 79H , Umask 24H
486Counts cycles MITE is delivered at least one uops. Set
487Cmask = 1.
488.It Li IDQ.ALL_MITE_CYCLES_4_UOPS
489.Pq Event 79H , Umask 24H
490Counts cycles MITE is delivered four uops. Set Cmask
491=4.
492.It Li IDQ.MITE_ALL_UOPS
493.Pq Event 79H , Umask 3CH
494# of uops delivered to IDQ from any path.
495.It Li ICACHE.MISSES
496.Pq Event 80H , Umask 02H
497Number of Instruction Cache, Streaming Buffer and
498Victim Cache Misses. Includes UC accesses.
499.It Li ITLB_MISSES.MISS_CAUSES_A_WALK
500.Pq Event 85H , Umask 01H
501Misses in ITLB that causes a page walk of any page
502size.
503.It Li ITLB_MISSES.WALK_COMPLETED_4K
504.Pq Event 85H , Umask 02H
505Completed page walks due to misses in ITLB 4K page
506entries.
507.It Li TLB_MISSES.WALK_COMPLETED_2M_4M
508.Pq Event 85H , Umask 04H
509Completed page walks due to misses in ITLB 2M/4M
510page entries.
511.It Li ITLB_MISSES.WALK_COMPLETED
512.Pq Event 85H , Umask 0EH
513Completed page walks in ITLB of any page size.
514.It Li ITLB_MISSES.WALK_DURATION
515.Pq Event 85H , Umask 10H
516Cycle PMH is busy with a walk.
517.It Li ITLB_MISSES.STLB_HIT_4K
518.Pq Event 85H , Umask 20H
519ITLB misses that hit STLB (4K).
520.It Li ITLB_MISSES.STLB_HIT_2M
521.Pq Event 85H , Umask 40H
522ITLB misses that hit STLB (2K).
523.It Li ITLB_MISSES.STLB_HIT
524.Pq Event 85H , Umask 60H
525TLB misses that hit STLB. No page walk.
526.It Li ILD_STALL.LCP
527.Pq Event 87H , Umask 01H
528Stalls caused by changing prefix length of the
529instruction.
530.It Li ILD_STALL.IQ_FULL
531.Pq Event 87H , Umask 04H
532Stall cycles due to IQ is full.
533.It Li BR_INST_EXEC.NONTAKEN_COND
534.Pq Event 88H , Umask 41H
535Count conditional near branch instructions that were executed (but not
536necessarily retired) and not taken.
537.It Li BR_INST_EXEC.TAKEN_COND
538.Pq Event 88H , Umask 81H
539Count conditional near branch instructions that were executed (but not
540necessarily retired) and taken.
541.It Li BR_INST_EXEC.DIRECT_JMP
542.Pq Event 88H , Umask 82H
543Count all unconditional near branch instructions excluding calls and
544indirect branches.
545.It Li BR_INST_EXEC.INDIRECT_JMP_NON_CALL_RET
546.Pq Event 88H , Umask 84H
547Count executed indirect near branch instructions that are not calls nor
548returns.
549.It Li BR_INST_EXEC.RETURN_NEAR
550.Pq Event 88H , Umask 88H
551Count indirect near branches that have a return mnemonic.
552.It Li BR_INST_EXEC.DIRECT_NEAR_CALL
553.Pq Event 88H , Umask 90H
554Count unconditional near call branch instructions, excluding non call
555branch, executed.
556.It Li BR_INST_EXEC.INDIRECT_NEAR_CALL
557.Pq Event 88H , Umask A0H
558Count indirect near calls, including both register and memory indirect,
559executed.
560.It Li BR_INST_EXEC.ALL_BRANCHES
561.Pq Event 88H , Umask FFH
562Counts all near executed branches (not necessarily retired).
563.It Li BR_MISP_EXEC.NONTAKEN_COND
564.Pq Event 89H , Umask 41H
565Count conditional near branch instructions mispredicted as nontaken.
566.It Li BR_MISP_EXEC.TAKEN_COND
567.Pq Event 89H , Umask 81H
568Count conditional near branch instructions mispredicted as taken.
569.It Li BR_MISP_EXEC.INDIRECT_JMP_NON_CALL_RET
570.Pq Event 89H , Umask 84H
571Count mispredicted indirect near branch instructions that are not calls
572nor returns.
573.It Li BR_MISP_EXEC.RETURN_NEAR
574.Pq Event 89H , Umask 88H
575Count mispredicted indirect near branches that have a return mnemonic.
576.It Li BR_MISP_EXEC.DIRECT_NEAR_CALL
577.Pq Event 89H , Umask 90H
578Count mispredicted unconditional near call branch instructions, excluding
579non call branch, executed.
580.It Li BR_MISP_EXEC.INDIRECT_NEAR_CALL
581.Pq Event 89H , Umask A0H
582Count mispredicted indirect near calls, including both register and memory
583indirect, executed.
584.It Li BR_MISP_EXEC.ALL_BRANCHES
585.Pq Event 89H , Umask FFH
586Counts all mispredicted near executed branches (not necessarily retired).
587.It Li IDQ_UOPS_NOT_DELIVERED.CORE
588.Pq Event 9CH , Umask 01H
589Count number of non-delivered uops to RAT per
590thread.
591.It Li UOPS_EXECUTED_PORT.PORT_0
592.Pq Event A1H , Umask 01H
593Cycles which a Uop is dispatched on port 0 in this
594thread.
595.It Li UOPS_EXECUTED_PORT.PORT_1
596.Pq Event A1H , Umask 02H
597Cycles which a Uop is dispatched on port 1 in this
598thread.
599.It Li UOPS_EXECUTED_PORT.PORT_2
600.Pq Event A1H , Umask 04H
601Cycles which a Uop is dispatched on port 2 in this
602thread.
603.It Li UOPS_EXECUTED_PORT.PORT_3
604.Pq Event A1H , Umask 08H
605Cycles which a Uop is dispatched on port 3 in this
606thread.
607.It Li UOPS_EXECUTED_PORT.PORT_4
608.Pq Event A1H , Umask 10H
609Cycles which a Uop is dispatched on port 4 in this
610thread.
611.It Li UOPS_EXECUTED_PORT.PORT_5
612.Pq Event A1H , Umask 20H
613Cycles which a Uop is dispatched on port 5 in this
614thread.
615.It Li UOPS_EXECUTED_PORT.PORT_6
616.Pq Event A1H , Umask 40H
617Cycles which a Uop is dispatched on port 6 in this
618thread.
619.It Li UOPS_EXECUTED_PORT.PORT_7
620.Pq Event A1H , Umask 80H
621Cycles which a Uop is dispatched on port 7 in this
622thread.
623.It Li RESOURCE_STALLS.ANY
624.Pq Event A2H , Umask 01H
625Cycles Allocation is stalled due to Resource Related
626reason.
627.It Li RESOURCE_STALLS.RS
628.Pq Event A2H , Umask 04H
629Cycles stalled due to no eligible RS entry available.
630.It Li RESOURCE_STALLS.SB
631.Pq Event A2H , Umask 08H
632Cycles stalled due to no store buffers available (not
633including draining form sync).
634.It Li RESOURCE_STALLS.ROB
635.Pq Event A2H , Umask 10H
636Cycles stalled due to re-order buffer full.
637.It Li CYCLE_ACTIVITY.CYCLES_L2_PENDING
638.Pq Event A3H , Umask 01H
639Cycles with pending L2 miss loads. Set Cmask=2 to
640count cycle.
641.It Li CYCLE_ACTIVITY.CYCLES_LDM_PENDING
642.Pq Event A3H , Umask 02H
643Cycles with pending memory loads. Set Cmask=2 to
644count cycle.
645.It Li CYCLE_ACTIVITY.STALLS_L2_PENDING
646.Pq Event A3H , Umask 05H
647Number of loads missed L2.
648.It Li CYCLE_ACTIVITY.CYCLES_L1D_PENDING
649.Pq Event A3H , Umask 08H
650Cycles with pending L1 cache miss loads. Set
651Cmask=8 to count cycle.
652.It Li ITLB.ITLB_FLUSH
653.Pq Event AEH , Umask 01H
654Counts the number of ITLB flushes, includes
6554k/2M/4M pages.
656.It Li OFFCORE_REQUESTS.DEMAND_DATA_RD
657.Pq Event B0H , Umask 01H
658Demand data read requests sent to uncore.
659.It Li OFFCORE_REQUESTS.DEMAND_CODE_RD
660.Pq Event B0H , Umask 02H
661Demand code read requests sent to uncore.
662.It Li OFFCORE_REQUESTS.DEMAND_RFO
663.Pq Event B0H , Umask 04H
664Demand RFO read requests sent to uncore, including
665regular RFOs, locks, ItoM.
666.It Li OFFCORE_REQUESTS.ALL_DATA_RD
667.Pq Event B0H , Umask 08H
668Data read requests sent to uncore (demand and
669prefetch).
670.It Li UOPS_EXECUTED.CORE
671.Pq Event B1H , Umask 02H
672Counts total number of uops to be executed per-core
673each cycle.
674.It Li OFF_CORE_RESPONSE_0
675.Pq Event B7H , Umask 01H
676Requires MSR 01A6H
677.It Li OFF_CORE_RESPONSE_1
678.Pq Event BBH , Umask 01H
679Requires MSR 01A7H
680.It Li PAGE_WALKER_LOADS.DTLB_L1
681.Pq Event BCH , Umask 11H
682Number of DTLB page walker loads that hit in the
683L1+FB.
684.It Li PAGE_WALKER_LOADS.ITLB_L1
685.Pq Event BCH , Umask 21H
686Number of ITLB page walker loads that hit in the
687L1+FB.
688.It Li PAGE_WALKER_LOADS.DTLB_L2
689.Pq Event BCH , Umask 12H
690Number of DTLB page walker loads that hit in the L2.
691.It Li PAGE_WALKER_LOADS.ITLB_L2
692.Pq Event BCH , Umask 22H
693Number of ITLB page walker loads that hit in the L2.
694.It Li PAGE_WALKER_LOADS.DTLB_L3
695.Pq Event BCH , Umask 14H
696Number of DTLB page walker loads that hit in the L3.
697.It Li PAGE_WALKER_LOADS.ITLB_L3
698.Pq Event BCH , Umask 24H
699Number of ITLB page walker loads that hit in the L3.
700.It Li PAGE_WALKER_LOADS.DTLB_MEMORY
701.Pq Event BCH , Umask 18H
702Number of DTLB page walker loads from memory.
703.It Li PAGE_WALKER_LOADS.ITLB_MEMORY
704.Pq Event BCH , Umask 28H
705Number of ITLB page walker loads from memory.
706.It Li TLB_FLUSH.DTLB_THREAD
707.Pq Event BDH , Umask 01H
708DTLB flush attempts of the thread-specific entries.
709.It Li TLB_FLUSH.STLB_ANY
710.Pq Event BDH , Umask 20H
711Count number of STLB flush attempts.
712.It Li INST_RETIRED.ANY_P
713.Pq Event C0H , Umask 00H
714Number of instructions at retirement.
715.It Li INST_RETIRED.ALL
716.Pq Event C0H , Umask 01H
717Precise instruction retired event with HW to reduce
718effect of PEBS shadow in IP distribution.
719.It Li OTHER_ASSISTS.AVX_TO_SSE
720.Pq Event C1H , Umask 08H
721Number of transitions from AVX-256 to legacy SSE
722when penalty applicable.
723.It Li OTHER_ASSISTS.SSE_TO_AVX
724.Pq Event C1H , Umask 10H
725Number of transitions from SSE to AVX-256 when
726penalty applicable.
727.It Li OTHER_ASSISTS.ANY_WB_ASSIST
728.Pq Event C1H , Umask 40H
729Number of microcode assists invoked by HW upon
730uop writeback.
731.It Li UOPS_RETIRED.ALL
732.Pq Event C2H , Umask 01H
733Counts the number of micro-ops retired, Use
734cmask=1 and invert to count active cycles or stalled
735cycles.
736.It Li UOPS_RETIRED.RETIRE_SLOTS
737.Pq Event C2H , Umask 02H
738Counts the number of retirement slots used each
739cycle.
740.It Li MACHINE_CLEARS.MEMORY_ORDERING
741.Pq Event C3H , Umask 02H
742Counts the number of machine clears due to memory
743order conflicts.
744.It Li MACHINE_CLEARS.SMC
745.Pq Event C3H , Umask 04H
746Number of self-modifying-code machine clears
747detected.
748.It Li MACHINE_CLEARS.MASKMOV
749.Pq Event C3H , Umask 20H
750Counts the number of executed AVX masked load
751operations that refer to an illegal address range with
752the mask bits set to 0.
753.It Li BR_INST_RETIRED.ALL_BRANCHES
754.Pq Event C4H , Umask 00H
755Branch instructions at retirement.
756.It Li BR_INST_RETIRED.CONDITIONAL
757.Pq Event C4H , Umask 01H
758Counts the number of conditional branch instructions Supports PEBS
759retired.
760.It Li BR_INST_RETIRED.NEAR_CALL
761.Pq Event C4H , Umask 02H
762Direct and indirect near call instructions retired.
763.It Li BR_INST_RETIRED.ALL_BRANCHES
764.Pq Event C4H , Umask 04H
765Counts the number of branch instructions retired.
766.It Li BR_INST_RETIRED.NEAR_RETURN
767.Pq Event C4H , Umask 08H
768Counts the number of near return instructions
769retired.
770.It Li BR_INST_RETIRED.NOT_TAKEN
771.Pq Event C4H , Umask 10H
772Counts the number of not taken branch instructions
773retired.
774 It Li BR_INST_RETIRED.NEAR_TAKEN
775.Pq Event C4H , Umask 20H
776Number of near taken branches retired.
777.It Li BR_INST_RETIRED.FAR_BRANCH
778.Pq Event C4H , Umask 40H
779Number of far branches retired.
780.It Li BR_MISP_RETIRED.ALL_BRANCHES
781.Pq Event C5H , Umask 00H
782Mispredicted branch instructions at retirement
783.It Li BR_MISP_RETIRED.CONDITIONAL
784.Pq Event C5H , Umask 01H
785Mispredicted conditional branch instructions retired.
786.It Li BR_MISP_RETIRED.CONDITIONAL
787.Pq Event C5H , Umask 04H
788Mispredicted macro branch instructions retired.
789.It Li FP_ASSIST.X87_OUTPUT
790.Pq Event CAH , Umask 02H
791Number of X87 FP assists due to Output values.
792.It Li FP_ASSIST.X87_INPUT
793.Pq Event CAH , Umask 04H
794Number of X87 FP assists due to input values.
795.It Li FP_ASSIST.SIMD_OUTPUT
796.Pq Event CAH , Umask 08H
797Number of SIMD FP assists due to Output values.
798.It Li FP_ASSIST.SIMD_INPUT
799.Pq Event CAH , Umask 10H
800Number of SIMD FP assists due to input values.
801.It Li FP_ASSIST.ANY
802.Pq Event CAH , Umask 1EH
803Cycles with any input/output SSE* or FP assists.
804.It Li ROB_MISC_EVENTS.LBR_INSERTS
805.Pq Event CCH , Umask 20H
806Count cases of saving new LBR records by hardware.
807.It Li MEM_TRANS_RETIRED.LOAD_LATENCY
808.Pq Event CDH , Umask 01H
809Randomly sampled loads whose latency is above a
810user defined threshold. A small fraction of the overall
811loads are sampled due to randomization.
812.It Li MEM_UOPS_RETIRED.STLB_MISS_LOADS
813.Pq Event D0H , Umask 11H
814Count retired load uops that missed the STLB.
815.It Li MEM_UOPS_RETIRED.STLB_MISS_STORES
816.Pq Event D0H , Umask 12H
817Count retired store uops that missed the STLB.
818.It Li MEM_UOPS_RETIRED.SPLIT_LOADS
819.Pq Event D0H , Umask 41H
820Count retired load uops that were split across a cache line.
821.It Li MEM_UOPS_RETIRED.SPLIT_STORES
822.Pq Event D0H , Umask 42H
823Count retired store uops that were split across a cache line.
824.It Li MEM_UOPS_RETIRED.ALL_LOADS
825.Pq Event D0H , Umask 81H
826Count all retired load uops.
827.It Li MEM_UOPS_RETIRED.ALL_STORES
828.Pq Event D0H , Umask 82H
829Count all retired store uops.
830.It Li MEM_LOAD_UOPS_RETIRED.L1_HIT
831.Pq Event D1H , Umask 01H
832Retired load uops with L1 cache hits as data sources.
833.It Li MEM_LOAD_UOPS_RETIRED.L2_HIT
834.Pq Event D1H , Umask 02H
835Retired load uops with L2 cache hits as data sources.
836.It Li MEM_LOAD_UOPS_RETIRED.LLC_HIT
837.Pq Event D1H , Umask 04H
838Retired load uops with LLC cache hits as data
839sources.
840.It Li MEM_LOAD_UOPS_RETIRED.L2_MISS
841.Pq Event D1H , Umask 10H
842Retired load uops missed L2. Unknown data source
843excluded.
844.It Li MEM_LOAD_UOPS_RETIRED.HIT_LFB
845.Pq Event D1H , Umask 40H
846Retired load uops which data sources were load uops
847missed L1 but hit FB due to preceding miss to the
848same cache line with data not ready.
849.It Li MEM_LOAD_UOPS_LLC_HIT_RETIRED.XSNP_MISS
850.Pq Event D2H , Umask 01H
851Retired load uops which data sources were LLC hit
852and cross-core snoop missed in on-pkg core cache.
853.It Li MEM_LOAD_UOPS_LLC_HIT_RETIRED.XSNP_HIT
854.Pq Event D2H , Umask 02H
855Retired load uops which data sources were LLC and
856cross-core snoop hits in on-pkg core cache.
857.It Li MEM_LOAD_UOPS_LLC_HIT_RETIRED.XSNP_HITM
858.Pq Event D2H , Umask 04H
859Retired load uops which data sources were HitM
860responses from shared LLC.
861.It Li MEM_LOAD_UOPS_LLC_HIT_RETIRED.XSNP_NONE
862.Pq Event D2H , Umask 08H
863Retired load uops which data sources were hits in
864LLC without snoops required.
865.It Li MEM_LOAD_UOPS_LLC_MISS_RETIRED.LOCAL_DRAM
866.Pq Event D3H , Umask 01H
867Retired load uops which data sources missed LLC but
868serviced from local dram.
869.It Li BACLEARS.ANY
870.Pq Event E6H , Umask 1FH
871Number of front end re-steers due to BPU
872misprediction.
873.It Li L2_TRANS.DEMAND_DATA_RD
874.Pq Event F0H , Umask 01H
875Demand Data Read requests that access L2 cache.
876.It Li L2_TRANS.RFO
877.Pq Event F0H , Umask 02H
878RFO requests that access L2 cache.
879.It Li L2_TRANS.CODE_RD
880.Pq Event F0H , Umask 04H
881L2 cache accesses when fetching instructions.
882.It Li L2_TRANS.ALL_PF
883.Pq Event F0H , Umask 08H
884Any MLC or LLC HW prefetch accessing L2, including
885rejects.
886.It Li L2_TRANS.L1D_WB
887.Pq Event F0H , Umask 10H
888L1D writebacks that access L2 cache.
889.It Li L2_TRANS.L2_FILL
890.Pq Event F0H , Umask 20H
891L2 fill requests that access L2 cache.
892.It Li L2_TRANS.L2_WB
893.Pq Event F0H , Umask 40H
894L2 writebacks that access L2 cache.
895.It Li L2_TRANS.ALL_REQUESTS
896.Pq Event F0H , Umask 80H
897Transactions accessing L2 pipe.
898.It Li L2_LINES_IN.I
899.Pq Event F1H , Umask 01H
900L2 cache lines in I state filling L2.
901.It Li L2_LINES_IN.S
902.Pq Event F1H , Umask 02H
903L2 cache lines in S state filling L2.
904.It Li L2_LINES_IN.E
905.Pq Event F1H , Umask 04H
906L2 cache lines in E state filling L2.
907.It Li L2_LINES_IN.ALL
908.Pq Event F1H , Umask 07H
909L2 cache lines filling L2.
910.It Li L2_LINES_OUT.DEMAND_CLEAN
911.Pq Event F2H , Umask 05H
912Clean L2 cache lines evicted by demand.
913.It Li L2_LINES_OUT.DEMAND_DIRTY
914.Pq Event F2H , Umask 06H
915Dirty L2 cache lines evicted by demand.
916.El
917.Sh SEE ALSO
918.Xr pmc 3 ,
919.Xr pmc.atom 3 ,
920.Xr pmc.core 3 ,
921.Xr pmc.iaf 3 ,
922.Xr pmc.ucf 3 ,
923.Xr pmc.k7 3 ,
924.Xr pmc.k8 3 ,
925.Xr pmc.p4 3 ,
926.Xr pmc.p5 3 ,
927.Xr pmc.p6 3 ,
928.Xr pmc.corei7 3 ,
929.Xr pmc.corei7uc 3 ,
930.Xr pmc.haswell 3 ,
931.Xr pmc.haswelluc 3 ,
932.Xr pmc.ivybridge 3 ,
933.Xr pmc.ivybridgexeon 3 ,
934.Xr pmc.sandybridge 3 ,
935.Xr pmc.sandybridgeuc 3 ,
936.Xr pmc.sandybridgexeon 3 ,
937.Xr pmc.westmere 3 ,
938.Xr pmc.westmereuc 3 ,
939.Xr pmc.soft 3 ,
940.Xr pmc.tsc 3 ,
941.Xr pmc_cpuinfo 3 ,
942.Xr pmclog 3 ,
943.Xr hwpmc 4
944.Sh HISTORY
945Support for the Haswell Xeon microarchitecture first appeared in
946.Fx 10.2 .
947.Sh AUTHORS
948The
949.Lb libpmc
950library was written by
951.An "Joseph Koshy"
952.Aq jkoshy@FreeBSD.org .
953The support for the Haswell Xeon
954microarchitecture was written by
955.An "Randall Stewart"
956.Aq rrs@FreeBSD.org .
957