[llvm-bugs] [Bug 46249] New: [X86] VPSHUFD and VPERMQ with immediates combined into VPERMD with variable mask

[via llvm-bugs]

https://bugs.llvm.org/show_bug.cgi?id=46249

            Bug ID: 46249
           Summary: [X86] VPSHUFD and VPERMQ with immediates combined into
                    VPERMD with variable mask
           Product: libraries
           Version: trunk
          Hardware: PC
                OS: All
            Status: NEW
          Severity: enhancement
          Priority: P
         Component: Backend: X86
          Assignee: unassignedbugs at nondot.org
          Reporter: craig.topper 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

As noted here https://twitter.com/damageboy/status/1270266352406802432

We combined

t45: v16i32 = vector_shuffle<3,2,1,0,7,6,5,4,11,10,9,8,15,14,13,12> t44,
undef:v16i32
t46: v16i32 = vector_shuffle<8,9,10,11,12,13,14,15,0,1,2,3,4,5,6,7> t45,
undef:v16i32

into 

t258: v16i32 = vector_shuffle<11,10,9,8,15,14,13,12,3,2,1,0,7,6,5,4>

Which then required vpermd.


This happened before lowering and lowering doesn't know to recreate the two
shuffles. Simon, would hasFastVariableShuffle() also have created a VPERMD if
lowering had managed to make 2 immediate shuffles?

-- 
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/20200609/178571bf/attachment.html>