移去rt

hc

2023-12-08 01573e231f18eb2d99162747186f59511f56b64d

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