[llvm-bugs] [Bug 40476] New: [X86][SSE] Recognise out of range shift behaviour

via llvm-bugs llvm-bugs at lists.llvm.org
Sat Jan 26 03:14:56 PST 2019


https://bugs.llvm.org/show_bug.cgi?id=40476

            Bug ID: 40476
           Summary: [X86][SSE] Recognise out of range shift behaviour
           Product: libraries
           Version: trunk
          Hardware: PC
                OS: Windows NT
            Status: NEW
          Severity: enhancement
          Priority: P
         Component: Backend: X86
          Assignee: unassignedbugs at nondot.org
          Reporter: llvm-dev at redking.me.uk
                CC: craig.topper at gmail.com, llvm-bugs at lists.llvm.org,
                    llvm-dev at redking.me.uk, spatel+llvm at rotateright.com

As discussed on D46946, we should recognise when we're trying to handle/clamp
out of range vector shift amounts and make use of the SSE/AVX2 behaviours
instead.

For generic shifts this means we'd need to explicitly replace them with the x86
shift intrinsics.

https://godbolt.org/z/RSDKOh

The "uniform" SSE2 shifts are trickier but would benefit from the same
optimization.

-- 
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/20190126/49fb34c7/attachment-0001.html>


More information about the llvm-bugs mailing list