[llvm-bugs] [Bug 40632] New: [AArch64] Very bad code generation with shuffles and llvm.aarch64.neon.smull.v4i32

[via llvm-bugs]

https://bugs.llvm.org/show_bug.cgi?id=40632

            Bug ID: 40632
           Summary: [AArch64] Very bad code generation with shuffles and
                    llvm.aarch64.neon.smull.v4i32
           Product: libraries
           Version: trunk
          Hardware: PC
                OS: Windows NT
            Status: NEW
          Severity: enhancement
          Priority: P
         Component: Backend: AArch64
          Assignee: unassignedbugs at nondot.org
          Reporter: efriedma at quicinc.com
                CC: arnaud.degrandmaison at arm.com, florian_hahn at apple.com,
                    llvm-bugs at lists.llvm.org, maratek at gmail.com,
                    peter.smith at linaro.org, resistor at mac.com,
                    Ties.Stuij at arm.com

#include <arm_neon.h>

int32x4_t f(int32x4_t vacc, int16x8_t vb, int16x8_t va) {
    return vmull_lane_s16(vget_low_s16(vb), vget_low_s16(va), 0);
}

Produces:

        dup     v0.8h, v2.h[0]
        ext     v0.16b, v0.16b, v0.16b, #8
        smull   v0.4s, v1.4h, v0.4h
        ret

Something is obviously going terribly wrong here.

I'll have a patch soon, I think.

-- 
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/20190206/e61f888f/attachment.html>