[llvm-bugs] [Bug 46631] New: Request: ability to run normal inliner at -O0

[via llvm-bugs]

https://bugs.llvm.org/show_bug.cgi?id=46631

            Bug ID: 46631
           Summary: Request: ability to run normal inliner at -O0
           Product: clang
           Version: unspecified
          Hardware: All
                OS: All
            Status: NEW
          Severity: enhancement
          Priority: P
         Component: -New Bugs
          Assignee: unassignedclangbugs at nondot.org
          Reporter: dave at znu.io
                CC: htmldeveloper at gmail.com, llvm-bugs at lists.llvm.org,
                    neeilans at live.com, richard-llvm at metafoo.co.uk

For projects that lean heavily on "zero-cost" abstractions, lack of inlining at
-O0 is really expensive.

Anecdotally speaking, force inlining LLVM's casting logic and a few key tiny
functions at -O0 in Swift allow the debug Swift compiler to build the standard
library module about 20% faster (saving a couple minutes).

The only "downside" of this approach is that breakpoints in debuggers sometimes
map to more than one location.

It would be really great if clang had a way to force the normal inliner to run
at -O0. Thanks for considering this.

-- 
You are receiving this mail because:
You are on the CC list for the bug.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.llvm.org/pipermail/llvm-bugs/attachments/20200708/62d14b56/attachment-0001.html>