| /tools/testing/selftests/powerpc/pmu/ |
| A D | count_instructions.c | 30 u64 overhead, bool report) in do_count_loop() argument 45 expected = instructions + overhead; in do_count_loop() 76 u64 current, overhead; in determine_overhead() local 80 overhead = events[0].result.value; in determine_overhead() 85 if (current < overhead) { in determine_overhead() 87 overhead = current; in determine_overhead() 91 return overhead; in determine_overhead() 97 u64 overhead; in test_body() local 112 overhead = determine_overhead(events); in test_body() 116 FAIL_IF(do_count_loop(events, 1000000, overhead, true)); in test_body() [all …]
|
| A D | count_stcx_fail.c | 30 u64 overhead, bool report) in do_count_loop() argument 81 u64 current, overhead; in determine_overhead() local 85 overhead = events[0].result.value; in determine_overhead() 90 if (current < overhead) { in determine_overhead() 92 overhead = current; in determine_overhead() 96 return overhead; in determine_overhead() 105 u64 overhead; in test_body() local 129 overhead = determine_overhead(events); in test_body() 133 FAIL_IF(do_count_loop(events, 1000000, overhead, true)); in test_body() 136 FAIL_IF(do_count_loop(events, 10000000, overhead, true)); in test_body() [all …]
|
| /tools/testing/selftests/powerpc/pmu/ebb/ |
| A D | instruction_count_test.c | 26 uint64_t overhead, bool report) in do_count_loop() argument 46 expected = instructions + overhead; in do_count_loop() 72 uint64_t current, overhead; in determine_overhead() local 76 overhead = event->result.value; in determine_overhead() 81 if (current < overhead) { in determine_overhead() 83 overhead = current; in determine_overhead() 87 return overhead; in determine_overhead() 112 uint64_t overhead; in instruction_count() local 131 overhead = determine_overhead(&event); in instruction_count() 135 FAIL_IF(do_count_loop(&event, 0x100000, overhead, true)); in instruction_count() [all …]
|
| /tools/perf/Documentation/ |
| A D | callchain-overhead-calculation.txt | 3 The CPU overhead can be shown in two columns as 'Children' and 'Self' 5 wall-clock overhead). The 'self' overhead is simply calculated by 8 'self' overhead values should be 100%. 10 The 'children' overhead is calculated by adding all period values of 11 the child functions so that it can show the total overhead of the 16 It might be confusing that the sum of all the 'children' overhead 19 can find which function has the most overhead even if samples are 46 in the usual (self-overhead-only) output of perf report: 98 Likewise, the 'self' overhead of 'bar' (40%) was added to the 99 'children' overhead of 'main' and '\_\_libc_start_main'. [all …]
|
| A D | cpu-and-latency-overheads.txt | 8 Perf can measure overhead for both of these times (shown in 'overhead' and 11 Optimizing CPU overhead is useful to improve 'throughput', while optimizing 12 latency overhead is useful to improve 'latency'. It's important to understand 20 128 seconds of CPU time). The CPU overhead is: 'foo' - 6.6%, 'bar' - 93.4%. 21 While the latency overhead is: 'foo' - 90%, 'bar' - 10%. If we try to optimize 22 running time of the program looking at the (wrong in this case) CPU overhead, 26 By default, perf shows only CPU overhead. To show latency overhead, use 38 To sort by latency overhead, use 'perf report --latency':
|
| A D | perf-report.txt | 47 the overhead and latency columns. See --percentage for more info. 57 the overhead and latency columns. See --percentage for more info. 62 the overhead and latency columns. See --percentage for more info. 117 - overhead: CPU overhead percentage of sample. 118 - latency: latency (wall-clock) overhead percentage of sample. 147 By default, overhead, comm, dso and symbol keys are used. 148 (i.e. --sort overhead,comm,dso,symbol). 303 - percent: display overhead percent (default) 457 Do not show entries which have an overhead under that percent. 541 for other events to reduce the overhead. [all …]
|
| A D | perf-config.txt | 182 'top' means a overhead percentage which is more than 5%. 187 'medium' means a overhead percentage which has more than 0.5%. 190 'normal' means the rest of overhead percentages 435 current overhead (33.33%). 478 (meaning that it assumes 100% total overhead of 'foo'). 480 The 'graph' uses absolute overhead value of 'foo' as total so each of 481 'bar' and 'baz' callchain will have 20.00% of overhead. 519 histogram entries. Entries having an overhead lower than this 521 is '10', only entries which have more than 10% of overhead will be 531 and show (accumulated) total overhead as well as 'Self' overhead. [all …]
|
| A D | perf-top.txt | 134 local_weight, abort, in_tx, transaction, overhead, sample, period. 140 overhead, overhead_sys, overhead_us, overhead_children, sample and period. 155 percentage of the overhead column. See --percentage for more info. 159 percentage of the overhead column. See --percentage for more info. 163 percentage of the overhead column. See --percentage for more info. 199 enabled. See the `overhead calculation' section for more details. 216 Do not show entries which have an overhead under that percent. 220 Determine how to display the overhead percentage of filtered entries. 357 Alternatively one can ask for a group and then two overhead columns 435 include::callchain-overhead-calculation.txt[]
|
| A D | perf-mem.txt | 79 overhead compared to other values. 158 Unlike linkperf:perf-report[1], which calculates overhead from the actual 159 sample period, perf-mem overhead is calculated using sample weight. E.g. 167 $ perf report -F overhead,symbol 171 $ perf mem report -F overhead,symbol
|
| A D | perf-bench.txt | 71 Benchmark overhead of uprobe + BPF. 130 This is useful to check cgroup context switching overhead.
|
| A D | tips.txt | 45 Order by the overhead of source file name and line number: perf report -s srcline 56 To separate samples by time use perf report --sort time,overhead,sym
|
| A D | perf-amd-ibs.txt | 95 overhead compared to other values. 176 # perf mem report -F overhead,cache,snoop,comm
|
| A D | perf-diff.txt | 114 overhead and 1 (default) means sorting by computed value of column 1 120 Determine how to display the overhead percentage of filtered entries.
|
| A D | perf-annotate.txt | 154 Do not show functions which have an overhead under that percent on
|
| A D | perf-arm-spe.txt | 242 SPE related (SAMPLE_* etc) core PMU events can be used without the overhead of
|
| A D | perf-c2c.txt | 64 overhead compared to other values.
|
| A D | perf-record.txt | 407 Do not update the buildid cache. This saves some overhead in situations 598 size thus lowering runtime profiling overhead. 649 overhead. You can still switch them on with:
|
| /tools/testing/selftests/net/forwarding/ |
| A D | tc_taprio.sh | 156 tc qdisc replace dev $if_name root stab overhead 24 taprio num_tc 8 \ 379 tc qdisc replace dev $swp2 root stab overhead 24 taprio num_tc 8 \
|
| /tools/sched_ext/ |
| A D | README.md | 187 ticks, and without having to incur the overhead of making scheduling decisions. 190 benefits from minimizing scheduling overhead and timer ticks. An example of 205 significantly lowered hierarchical scheduling overhead.
|
| /tools/include/uapi/linux/ |
| A D | pkt_sched.h | 93 unsigned short overhead; member 105 int overhead; member
|
| /tools/memory-model/Documentation/ |
| A D | ordering.txt | 116 of magnitude greater overhead than smp_mb(), atomic_xchg(), and so on. 140 This works, but the added smp_mb() adds needless overhead for 151 incurring no unnecessary overhead on x86. There are a number of 285 But smp_mb() often incurs much higher overhead than does 379 for RCU read-side critical sections incur very low overhead because 553 Used properly, unmarked C-language accesses can reduce overhead on
|
| A D | recipes.txt | 412 Release-acquire chains are a low-overhead, flexible, and easy-to-use 471 Despite this limitation, release-acquire chains are low-overhead as
|
| /tools/virtio/virtio-trace/ |
| A D | README | 5 overhead. Trace agent has the following functions:
|
| /tools/perf/ui/ |
| A D | hist.c | 638 HPP_PERCENT_FNS(overhead, period, PERF_HPP_FMT_TYPE__PERCENT) in HPP_PERCENT_FNS() argument 749 HPP__COLOR_PRINT_FNS("Overhead", overhead, OVERHEAD),
|
| /tools/perf/ui/gtk/ |
| A D | hists.c | 71 __HPP_COLOR_PERCENT_FN(overhead, period) in __HPP_COLOR_PERCENT_FN() argument
|