I have updated the code. It turned out that much less code has to be moved than I initially thought. :)

Clean up code and only mode definitions.

Having as few code in headers as possible is the general style guide in LLVM, so I'd rather err towards that and put definitions out-of-line as much as possible. Therefore if we can getaway by just moving definitions out-of-line, while keeping rest of the header ordering the same, let's go with that if it's feasible

Thanks for the comments!

Fixed nits. Thanks for the comments!

Delete insertion operator of raw_ostream for char8_t.

An alternative is to cast the char8_t* to char_t* in the << operator of the ostream of to get back the old behavior, e.g. something like

raw_ostream &operator<<(const char8_t *Str) {
    return this->operator<<(reinterpret_cast<const char *>(Str));
}

In D141392#4040037, @ilya-biryukov wrote:

Are there any uses of u8 string literals with formatted_raw_ostream outside tests? What was the effect on the final behavior here?

I just wanted to make sure we are not doing the wrong thing here. If the code used to compile and it now changes behavior, we may run into runtime failures in real code.
If that's the case, we could potentially consider adding overloads for u8 string literals to either support it or with =delete to make sure the code using them will be rewritten.

Thanks!

Additionally check for ODR use and remove inlcude added by accident.

In D139400#3974624, @shafik wrote:

It looks like we have an existing message for the regular function case

Reuse existing error message.

Use isLocalVarDecl to check if we have a local variable.

In D139400#3974654, @shafik wrote:

It looks like the existing diagnostic is issued by CheckDefaultArgumentVisitor. I think you can reuse it here but not 100% on that.

Thanks for the pointer! I will have a look.

Simplify code and update description.

Instead of lokking if there is a pack expansion pass the already computed information whether we have a by ref or a by copy type.

Moved check if we have a pack expansion with reference type to Sema::BuildLambdaExpr.

Fix test.

Remove unnecessary code from test.

Add a diagnostics test for pack expansions to check if args are passed by reference and not by value.

Note: The build failures are in parts of the code that isn't effected by this change.

Thanks for the comments!

Fixed comments and removed unnecessary includes.

Run clang-format on changed files.

Sync to head.

Thanks for the comments!

Add comments.

Add note that the new IgnoreMarcos options isn't guaranteed to be forward compatible.

Add test with global IgnoreMacros=true and readability-function-cognitive-complexity.IgnoreMacros unset.

Sync to head.

Remove unnecessary checks.

Add checks to ensure that notes aren't added for specific options.

In D96281#2596657, @lebedev.ri wrote:

Once again, can we please split this into two patches?
I'm quite happy with the DescribeBasicIncrements stuff, that can be merged right now.

Remove IgnoreMacros option (to be added in a follow up patch) and sync to head.

Thanks for the comments!

Change option name, improve description and minor fixes.

In D96281#2556586, @njames93 wrote:

How does this handle a macro where the argument has complex code.

MACRO(if (...) {});

Add test cases showing that when IgnoreMacros is set to 'true', also macro arguments are ignored.

Thanks for the comments!

Clarified option FlagBasicIncrements and removed base default constructor from class initializer list of FunctionASTVisitor.

Thanks for the comments!

Thanks for the comments!
I double-checked that the renaming went well and hope that I haven't missed anything this time ...

Addressed comments and renamed test to modernize-deprecated-functional

Addressed comments of reviewers.

Alex and Aaron, thanks for the comments!

Addressed comments of reviewer.

Implemented all suggestions from the reviewer.

looks good to me

looks good to me

looks good to me

Very nice!

The Emacs part looks good for me now.

Added two comments. It took some time for me to recall Emacs Lisps ;)

Thanks for the comments!

Updating D22805: [clang-include-fixer] Added Emacs integration for clang-include-fixer.

Thanks for the comments!