Fmod perf.cc update #16
Conversation
This reverts commit 942c117.
arturbac
left a comment
arturbac
left a comment
There was a problem hiding this comment.
thre is already fmod function object and function also
perf_test_suite/perf.cc
Outdated
| } | ||
| }; | ||
|
|
||
| static fixed_t fmod(fixed_t& a, fixed_t& b) |
There was a problem hiding this comment.
what it is for ?
there are already fobj and function for this also operator %
and all other just directly use it liek div above ..
struct div_test
{
template<typename value_type>
auto operator()(value_type const & tp)
{
return tp / tp;
}
};There was a problem hiding this comment.
@UF4007 FYI:
"OK, I rebased the fmod branch to the master branch and reverted some changes in the fmod"
Procedure for PR usually should look like:
- fork
- create in Your fork a branch for development
- commit changes to this branch and submit PR from this branch
- DO NOT submit changes from master branch, it complicates things for me to amend it and for You too in the future sync with master
After it get squashed forget about that branch. I do not and many other devs do any merges, usually we squash PR into single commit as it gives better git history about changes in code.
There was a problem hiding this comment.
Thus, before every time I decided to PR something, I must leave the old fork and reconstruct a new fork?
There was a problem hiding this comment.
what it is for ?
there are already fobj and function for this also operator %
and all other just directly use it liek div above ..struct div_test { template<typename value_type> auto operator()(value_type const & tp) { return tp / tp; } };
It is for comparison with the calculate speed of floating numbers. We cannot use operator% on floating numbers like 3.14%2.72. I don't want to make things complicated so I make the temporary "fixed_t" fmod function for benchmark use.
|
Updated. |
Update perf.cc for mod operation benchmark.
Somehow, the Files changed appear a lot of files you have been merged to branch master.
————————
OK, I rebased the fmod branch to the master branch and reverted some changes in the fmod branch which is after your last merge in the fmod branch. Now it seems to be back on the right track.
I know it may not be the best way, but I have poor experience in such an open source project, I am very sorry. I usually do teamwork on a private git server which have no PR mechanism.