1perf-diff(1) 2============ 3 4NAME 5---- 6perf-diff - Read perf.data files and display the differential profile 7 8SYNOPSIS 9-------- 10[verse] 11'perf diff' [baseline file] [data file1] [[data file2] ... ] 12 13DESCRIPTION 14----------- 15This command displays the performance difference amongst two or more perf.data 16files captured via perf record. 17 18If no parameters are passed it will assume perf.data.old and perf.data. 19 20The differential profile is displayed only for events matching both 21specified perf.data files. 22 23If no parameters are passed the samples will be sorted by dso and symbol. 24As the perf.data files could come from different binaries, the symbols addresses 25could vary. So perf diff is based on the comparison of the files and 26symbols name. 27 28OPTIONS 29------- 30-D:: 31--dump-raw-trace:: 32 Dump raw trace in ASCII. 33 34--kallsyms=<file>:: 35 kallsyms pathname 36 37-m:: 38--modules:: 39 Load module symbols. WARNING: use only with -k and LIVE kernel 40 41-d:: 42--dsos=:: 43 Only consider symbols in these dsos. CSV that understands 44 file://filename entries. This option will affect the percentage 45 of the Baseline/Delta column. See --percentage for more info. 46 47-C:: 48--comms=:: 49 Only consider symbols in these comms. CSV that understands 50 file://filename entries. This option will affect the percentage 51 of the Baseline/Delta column. See --percentage for more info. 52 53-S:: 54--symbols=:: 55 Only consider these symbols. CSV that understands 56 file://filename entries. This option will affect the percentage 57 of the Baseline/Delta column. See --percentage for more info. 58 59-s:: 60--sort=:: 61 Sort by key(s): pid, comm, dso, symbol, cpu, parent, srcline. 62 Please see description of --sort in the perf-report man page. 63 64-t:: 65--field-separator=:: 66 67 Use a special separator character and don't pad with spaces, replacing 68 all occurrences of this separator in symbol names (and other output) 69 with a '.' character, that thus it's the only non valid separator. 70 71-v:: 72--verbose:: 73 Be verbose, for instance, show the raw counts in addition to the 74 diff. 75 76-f:: 77--force:: 78 Don't do ownership validation. 79 80--symfs=<directory>:: 81 Look for files with symbols relative to this directory. 82 83-b:: 84--baseline-only:: 85 Show only items with match in baseline. 86 87-c:: 88--compute:: 89 Differential computation selection - delta, ratio, wdiff, delta-abs 90 (default is delta-abs). Default can be changed using diff.compute 91 config option. See COMPARISON METHODS section for more info. 92 93-p:: 94--period:: 95 Show period values for both compared hist entries. 96 97-F:: 98--formula:: 99 Show formula for given computation. 100 101-o:: 102--order:: 103 Specify compute sorting column number. 0 means sorting by baseline 104 overhead and 1 (default) means sorting by computed value of column 1 105 (data from the first file other base baseline). Values more than 1 106 can be used only if enough data files are provided. 107 The default value can be set using the diff.order config option. 108 109--percentage:: 110 Determine how to display the overhead percentage of filtered entries. 111 Filters can be applied by --comms, --dsos and/or --symbols options. 112 113 "relative" means it's relative to filtered entries only so that the 114 sum of shown entries will be always 100%. "absolute" means it retains 115 the original value before and after the filter is applied. 116 117COMPARISON 118---------- 119The comparison is governed by the baseline file. The baseline perf.data 120file is iterated for samples. All other perf.data files specified on 121the command line are searched for the baseline sample pair. If the pair 122is found, specified computation is made and result is displayed. 123 124All samples from non-baseline perf.data files, that do not match any 125baseline entry, are displayed with empty space within baseline column 126and possible computation results (delta) in their related column. 127 128Example files samples: 129- file A with samples f1, f2, f3, f4, f6 130- file B with samples f2, f4, f5 131- file C with samples f1, f2, f5 132 133Example output: 134 x - computation takes place for pair 135 b - baseline sample percentage 136 137- perf diff A B C 138 139 baseline/A compute/B compute/C samples 140 --------------------------------------- 141 b x f1 142 b x x f2 143 b f3 144 b x f4 145 b f6 146 x x f5 147 148- perf diff B A C 149 150 baseline/B compute/A compute/C samples 151 --------------------------------------- 152 b x x f2 153 b x f4 154 b x f5 155 x x f1 156 x f3 157 x f6 158 159- perf diff C B A 160 161 baseline/C compute/B compute/A samples 162 --------------------------------------- 163 b x f1 164 b x x f2 165 b x f5 166 x f3 167 x x f4 168 x f6 169 170COMPARISON METHODS 171------------------ 172delta 173~~~~~ 174If specified the 'Delta' column is displayed with value 'd' computed as: 175 176 d = A->period_percent - B->period_percent 177 178with: 179 - A/B being matching hist entry from data/baseline file specified 180 (or perf.data/perf.data.old) respectively. 181 182 - period_percent being the % of the hist entry period value within 183 single data file 184 185 - with filtering by -C, -d and/or -S, period_percent might be changed 186 relative to how entries are filtered. Use --percentage=absolute to 187 prevent such fluctuation. 188 189delta-abs 190~~~~~~~~~ 191Same as 'delta` method, but sort the result with the absolute values. 192 193ratio 194~~~~~ 195If specified the 'Ratio' column is displayed with value 'r' computed as: 196 197 r = A->period / B->period 198 199with: 200 - A/B being matching hist entry from data/baseline file specified 201 (or perf.data/perf.data.old) respectively. 202 203 - period being the hist entry period value 204 205wdiff:WEIGHT-B,WEIGHT-A 206~~~~~~~~~~~~~~~~~~~~~~~ 207If specified the 'Weighted diff' column is displayed with value 'd' computed as: 208 209 d = B->period * WEIGHT-A - A->period * WEIGHT-B 210 211 - A/B being matching hist entry from data/baseline file specified 212 (or perf.data/perf.data.old) respectively. 213 214 - period being the hist entry period value 215 216 - WEIGHT-A/WEIGHT-B being user supplied weights in the the '-c' option 217 behind ':' separator like '-c wdiff:1,2'. 218 - WEIGHT-A being the weight of the data file 219 - WEIGHT-B being the weight of the baseline data file 220 221SEE ALSO 222-------- 223linkperf:perf-record[1], linkperf:perf-report[1] 224