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