[llvm-bugs] [Bug 41188] New: Unprofitable target specific vectorization

[via llvm-bugs]

https://bugs.llvm.org/show_bug.cgi?id=41188

            Bug ID: 41188
           Summary: Unprofitable target specific vectorization
           Product: libraries
           Version: trunk
          Hardware: PC
                OS: Linux
            Status: NEW
          Severity: enhancement
          Priority: P
         Component: Backend: X86
          Assignee: unassignedbugs at nondot.org
          Reporter: david.bolvansky at gmail.com
                CC: craig.topper at gmail.com, llvm-bugs at lists.llvm.org,
                    llvm-dev at redking.me.uk, spatel+llvm at rotateright.com

I checked this "benchmark": https://godbolt.org/z/IrAj-o

-O3 on haswell X86-64, 4720HQ - cca 1.9s
-O3 -mtune=native -march=native - cca 2.1s


Seems like we have a regression there too. Just looking at the godbolt results,
the slower code contains a vectorized code, standard -O3 avoids vectorization.

Maybe cost model issue for Haswell ?

-- 
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/20190321/9a1e6fe8/attachment.html>