| commit | ec8ef12310886f8dae440b616a2f7eb60602d141 | |
| author | Jiří Zárevúcky <zarevucky.jiri@gmail.com> | |
| Wed, 19 Apr 2023 09:31:02 +0000 (19 11:31 +0200) | ||
| committer | Jiří Zárevúcky <zarevucky.jiri@gmail.com> | |
| Wed, 19 Apr 2023 09:31:02 +0000 (19 11:31 +0200) | ||
| tree | c110bc8160f9c090d5cfe479b469c353f1d64031 | treesnapshot (tar.gz zip) |
| parent | 117ad5a23def46f16848a3bd0a7168021c762d3c | commitdiff |
Split find_best_thread() into two functions
try_find_thread() attempts to get a thread from runqueues and returns
NULL when there's none available. find_best_thread() functions as
before and goes to sleep between attempts to find a thread to run.
The purpose of this split is that we can use the non-sleeping version
in the context of a previously running thread to avoid an additional
context switch in case new thread is immediately available.
try_find_thread() attempts to get a thread from runqueues and returns
NULL when there's none available. find_best_thread() functions as
before and goes to sleep between attempts to find a thread to run.
The purpose of this split is that we can use the non-sleeping version
in the context of a previously running thread to avoid an additional
context switch in case new thread is immediately available.
| kernel/generic/src/proc/scheduler.c | diffblobblamehistory |