[llvm-bugs] [Bug 39196] New: Benchmark regression vs GCC: mcf

[via llvm-bugs]

https://bugs.llvm.org/show_bug.cgi?id=39196

            Bug ID: 39196
           Summary: Benchmark regression vs GCC: mcf
           Product: libraries
           Version: trunk
          Hardware: PC
                OS: Linux
            Status: NEW
          Severity: enhancement
          Priority: P
         Component: Scalar Optimizations
          Assignee: unassignedbugs at nondot.org
          Reporter: paulsson at linux.vnet.ibm.com
                CC: llvm-bugs at lists.llvm.org

It seems that GCC manages to hoist a store inside a loop to a single store
after the loop, while clang stores in every iteration.

This seems to relate to concurrency and to the ability of the compiler to prove
that the value is stored at least once. For discussion and a reduced test case,
see: http://lists.llvm.org/pipermail/llvm-dev/2018-September/126281.html

The function is in pbeampp.c

-- 
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/20181006/6fa9f7a4/attachment.html>