change system file

hc

2024-10-09 244b2c5ca8b14627e4a17755e5922221e121c771

 kernel/tools/perf/Documentation/perf-c2c.txt
..
..
@@ -19,8 +19,11 @@
19
19
 The perf c2c tool provides means for Shared Data C2C/HITM analysis. It allows
20
20
 you to track down the cacheline contentions.
21
21
 
22
-The tool is based on x86's load latency and precise store facility events
23
-provided by Intel CPUs. These events provide:
22
+On x86, the tool is based on load latency and precise store facility events
23
+provided by Intel CPUs. On PowerPC, the tool uses random instruction sampling
24
+with thresholding feature.
25
+
26
+These events provide:
24
27
   - memory address of the access
25
28
   - type of the access (load and store details)
26
29
   - latency (in cycles) of the load access
..
..
@@ -37,7 +40,7 @@
37
40
 --------------
38
41
 -e::
39
42
 --event=::
40
-	Select the PMU event. Use 'perf mem record -e list'
43
+	Select the PMU event. Use 'perf c2c record -e list'
41
44
 	to list available events.
42
45
 
43
46
 -v::
..
..
@@ -46,7 +49,7 @@
46
49
 
47
50
 -l::
48
51
 --ldlat::
49
-	Configure mem-loads latency.
52
+	Configure mem-loads latency. (x86 only)
50
53
 
51
54
 -k::
52
55
 --all-kernel::
..
..
@@ -108,6 +111,17 @@
108
111
 --display::
109
112
 	Switch to HITM type (rmt, lcl) to display and sort on. Total HITMs as default.
110
113
 
114
+--stitch-lbr::
115
+	Show callgraph with stitched LBRs, which may have more complete
116
+	callgraph. The perf.data file must have been obtained using
117
+	perf c2c record --call-graph lbr.
118
+	Disabled by default. In common cases with call stack overflows,
119
+	it can recreate better call stacks than the default lbr call stack
120
+	output. But this approach is not full proof. There can be cases
121
+	where it creates incorrect call stacks from incorrect matches.
122
+	The known limitations include exception handing such as
123
+	setjmp/longjmp will have calls/returns not match.
124
+
111
125
 C2C RECORD
112
126
 ----------
113
127
 The perf c2c record command setup options related to HITM cacheline analysis
..
..
@@ -119,10 +133,15 @@
119
133
   -W,-d,--phys-data,--sample-cpu
120
134
 
121
135
 Unless specified otherwise with '-e' option, following events are monitored by
122
-default:
136
+default on x86:
123
137
 
124
138
   cpu/mem-loads,ldlat=30/P
125
139
   cpu/mem-stores/P
140
+
141
+and following on PowerPC:
142
+
143
+  cpu/mem-loads/
144
+  cpu/mem-stores/
126
145
 
127
146
 User can pass any 'perf record' option behind '--' mark, like (to enable
128
147
 callchains and system wide monitoring):
..
..
@@ -155,34 +174,36 @@
155
174
   Cacheline
156
175
   - cacheline address (hex number)
157
176
 
158
-  Total records
159
-  - sum of all cachelines accesses
160
-
161
177
   Rmt/Lcl Hitm
162
178
   - cacheline percentage of all Remote/Local HITM accesses
163
179
 
164
-  LLC Load Hitm - Total, Lcl, Rmt
180
+  LLC Load Hitm - Total, LclHitm, RmtHitm
165
181
   - count of Total/Local/Remote load HITMs
166
182
 
167
-  Store Reference - Total, L1Hit, L1Miss
168
-    Total - all store accesses
169
-    L1Hit - store accesses that hit L1
170
-    L1Hit - store accesses that missed L1
183
+  Total records
184
+  - sum of all cachelines accesses
171
185
 
172
-  Load Dram
173
-  - count of local and remote DRAM accesses
174
-
175
-  LLC Ld Miss
176
-  - count of all accesses that missed LLC
177
-
178
-  Total Loads
186
+  Total loads
179
187
   - sum of all load accesses
188
+
189
+  Total stores
190
+  - sum of all store accesses
191
+
192
+  Store Reference - L1Hit, L1Miss
193
+    L1Hit - store accesses that hit L1
194
+    L1Miss - store accesses that missed L1
180
195
 
181
196
   Core Load Hit - FB, L1, L2
182
197
   - count of load hits in FB (Fill Buffer), L1 and L2 cache
183
198
 
184
-  LLC Load Hit - Llc, Rmt
185
-  - count of LLC and Remote load hits
199
+  LLC Load Hit - LlcHit, LclHitm
200
+  - count of LLC load accesses, includes LLC hits and LLC HITMs
201
+
202
+  RMT Load Hit - RmtHit, RmtHitm
203
+  - count of remote load accesses, includes remote hits and remote HITMs
204
+
205
+  Load Dram - Lcl, Rmt
206
+  - count of local and remote DRAM accesses
186
207
 
187
208
 For each offset in the 2) list we display following data:
188
209