some CppCheck and related usage cleanups
#5979
Conversation
|
This is a continuation of the This change is already buried somewhere in my countless branches but I did the changes from scratch now that I encountered this again. |
| * and if it's possible at all */ | ||
| bool isUnusedFunctionCheckEnabled() const | ||
| { | ||
| return useSingleJob() && checks.isEnabled(Checks::unusedFunction); |
There was a problem hiding this comment.
hmm.. I wonder why it doesn't check useSingleJob() || !buildDir.empty()
There was a problem hiding this comment.
Just keeping the current behavior intact for now.
There was a problem hiding this comment.
Unsurprisingly tis check turned out flawed. This is addressed in #6207.
| if (mSettings.checks.isEnabled(Checks::unusedFunction) && | ||
| mSettings.useSingleJob() && | ||
| mSettings.buildDir.empty()) { | ||
| if (mSettings.isUnusedFunctionCheckEnabled() && mSettings.buildDir.empty()) { |
There was a problem hiding this comment.
hmm.. I don't understand why we required that the buildDir is empty. Don't we want to run this checking here for some reason if the build dir is empty..
There was a problem hiding this comment.
I have no idea. Currently I am keeping the current behavior intact. I will look into this when I look at the build directory related false positives. But that requires adding a lot more general testing of the build directory first.
|
Okay - now I do recall why those changes were buried in the first place because I encountered exactly these testing failures. |
I was passing the wrong suppressions to the reporting function. Should be fine now. |
|
Merging as there has been no feedback for almost a whole week. It is also blocking traffic as usual. |
No description provided.