diff options
| author | Ernie Rael <errael@raelity.com> | 2023-03-16 21:43:15 +0000 |
|---|---|---|
| committer | Bram Moolenaar <Bram@vim.org> | 2023-03-16 21:43:15 +0000 |
| commit | 076de79ad832558267b3ff903c048df2f4c1a5d6 (patch) | |
| tree | e0412a99ca8e44db8702b98a6a1cec93a0c561b9 /runtime/doc/repeat.txt | |
| parent | 16110ccf11de7a41cb0db0ae1ecb829e5a1b98fc (diff) | |
patch 9.0.1411: accuracy of profiling is not optimalv9.0.1411
Problem: Accuracy of profiling is not optimal.
Solution: Use CLOCK_MONOTONIC if possible. (Ernie Rael, closes #12129)
Diffstat (limited to 'runtime/doc/repeat.txt')
| -rw-r--r-- | runtime/doc/repeat.txt | 7 |
1 files changed, 4 insertions, 3 deletions
diff --git a/runtime/doc/repeat.txt b/runtime/doc/repeat.txt index c8a0f5d3d1..054b8705c5 100644 --- a/runtime/doc/repeat.txt +++ b/runtime/doc/repeat.txt @@ -1148,9 +1148,10 @@ long you take to respond to the input() prompt is irrelevant. Profiling should give a good indication of where time is spent, but keep in mind there are various things that may clobber the results: -- The accuracy of the time measured depends on the gettimeofday() system - function. It may only be as accurate as 1/100 second, even though the times - are displayed in micro seconds. +- The accuracy of the time measured depends on the gettimeofday(), or + clock_gettime if available, system function. The accuracy ranges from 1/100 + second to nano seconds. With clock_gettime the times are displayed in nano + seconds, otherwise micro seconds. You can use `has("prof_nsec")`. - Real elapsed time is measured, if other processes are busy they may cause delays at unpredictable moments. You may want to run the profiling several |