1 i synthesize instruction 1 i synthesize instructions events
2 y synthesize cycles even 2 y synthesize cycles events
3 b synthesize branches ev 3 b synthesize branches events (branch misses for Arm SPE)
4 c synthesize branches ev 4 c synthesize branches events (calls only)
5 r synthesize branches ev 5 r synthesize branches events (returns only)
6 x synthesize transaction 6 x synthesize transactions events
7 w synthesize ptwrite eve 7 w synthesize ptwrite events
8 p synthesize power event 8 p synthesize power events (incl. PSB events for Intel PT)
9 o synthesize other event 9 o synthesize other events recorded due to the use
10 of aux-output (refer t 10 of aux-output (refer to perf record)
11 I synthesize interrupt o 11 I synthesize interrupt or similar (asynchronous) events
12 (e.g. Intel PT Event T 12 (e.g. Intel PT Event Trace)
13 e synthesize error event 13 e synthesize error events
14 d create a debug log 14 d create a debug log
15 f synthesize first level 15 f synthesize first level cache events
16 m synthesize last level 16 m synthesize last level cache events
17 M synthesize memory even 17 M synthesize memory events
18 t synthesize TLB events 18 t synthesize TLB events
19 a synthesize remote acce 19 a synthesize remote access events
20 g synthesize a call chai 20 g synthesize a call chain (use with i or x)
21 G synthesize a call chai 21 G synthesize a call chain on existing event records
22 l synthesize last branch 22 l synthesize last branch entries (use with i or x)
23 L synthesize last branch 23 L synthesize last branch entries on existing event records
24 s skip initial number of 24 s skip initial number of events
25 q quicker (less detailed 25 q quicker (less detailed) decoding
26 A approximate IPC 26 A approximate IPC
27 Z prefer to ignore times 27 Z prefer to ignore timestamps (so-called "timeless" decoding)
28 T use the timestamp trac 28 T use the timestamp trace as kernel time
29 29
30 The default is all events i.e. the sam 30 The default is all events i.e. the same as --itrace=iybxwpe,
31 except for perf script where it is --i 31 except for perf script where it is --itrace=ce
32 32
33 In addition, the period (default 10000 33 In addition, the period (default 100000, except for perf script where it is 1)
34 for instructions events can be specifi 34 for instructions events can be specified in units of:
35 35
36 i instructions 36 i instructions
37 t ticks 37 t ticks
38 ms milliseconds 38 ms milliseconds
39 us microseconds 39 us microseconds
40 ns nanoseconds (default) 40 ns nanoseconds (default)
41 41
42 Also the call chain size (default 16, 42 Also the call chain size (default 16, max. 1024) for instructions or
43 transactions events can be specified. 43 transactions events can be specified.
44 44
45 Also the number of last branch entries 45 Also the number of last branch entries (default 64, max. 1024) for
46 instructions or transactions events ca 46 instructions or transactions events can be specified.
47 47
48 Similar to options g and l, size may a 48 Similar to options g and l, size may also be specified for options G and L.
49 On x86, note that G and L work poorly 49 On x86, note that G and L work poorly when data has been recorded with
50 large PEBS. Refer linkperf:perf-intel- 50 large PEBS. Refer linkperf:perf-intel-pt[1] man page for details.
51 51
52 It is also possible to skip events gen 52 It is also possible to skip events generated (instructions, branches, transactions,
53 ptwrite, power) at the beginning. This 53 ptwrite, power) at the beginning. This is useful to ignore initialization code.
54 54
55 --itrace=i0nss1000000 55 --itrace=i0nss1000000
56 56
57 skips the first million instructions. 57 skips the first million instructions.
58 58
59 The 'e' option may be followed by flag 59 The 'e' option may be followed by flags which affect what errors will or
60 will not be reported. Each flag must b 60 will not be reported. Each flag must be preceded by either '+' or '-'.
61 The flags are: 61 The flags are:
62 o overflow 62 o overflow
63 l trace data lost 63 l trace data lost
64 64
65 If supported, the 'd' option may be fo 65 If supported, the 'd' option may be followed by flags which affect what
66 debug messages will or will not be log 66 debug messages will or will not be logged. Each flag must be preceded
67 by either '+' or '-'. The flags are: 67 by either '+' or '-'. The flags are:
68 a all perf events 68 a all perf events
69 e output only on errors 69 e output only on errors (size configurable - see linkperf:perf-config[1])
70 o output to stdout 70 o output to stdout
71 71
72 If supported, the 'q' option may be re 72 If supported, the 'q' option may be repeated to increase the effect.
Linux® is a registered trademark of Linus Torvalds in the United States and other countries.
TOMOYO® is a registered trademark of NTT DATA CORPORATION.