[llvm-bugs] [Bug 26244] New: LLVM ERROR: Cannot select (@llvm.maxnum.v2f32)

via llvm-bugs llvm-bugs at lists.llvm.org
Thu Jan 21 11:18:54 PST 2016


https://llvm.org/bugs/show_bug.cgi?id=26244

            Bug ID: 26244
           Summary: LLVM ERROR: Cannot select (@llvm.maxnum.v2f32)
           Product: new-bugs
           Version: trunk
          Hardware: PC
                OS: Linux
            Status: NEW
          Severity: normal
          Priority: P
         Component: new bugs
          Assignee: unassignedbugs at nondot.org
          Reporter: rahulchaudhry at chromium.org
                CC: llvm-bugs at lists.llvm.org
    Classification: Unclassified

Created attachment 15684
  --> https://llvm.org/bugs/attachment.cgi?id=15684&action=edit
Reproducer

See the attached reduced.ll file.

$ llc reduced.ll 
LLVM ERROR: Cannot select: t4: v2f32 = fmaxnum t20, undef:v2f32
  t20: v2f32 = bitcast t19
    t19: v2i32 = ARMISD::VMOVIMM TargetConstant:i32<0>
      t18: i32 = TargetConstant<0>
  t1: v2f32 = undef
In function: foo

Using llvm/clang from tip-of-trunk (as of 01/20/2016).


This issue was found while trying to build Chrome browser for Chrome OS. The
(reduced) set of flags used to compile the original c++ file were:

clang++ \
        -cc1 \
        -triple thumbv7-cros-linux-gnueabi \
        -target-cpu cortex-a8 \
        -target-feature +fp-armv8 \
        -vectorize-slp \
        -std=gnu++11 \
        -emit-obj \
        -O2 \
        test.cpp

Removing the "-vectorize-slp" flag lets the compile go through.

-- 
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/20160121/1e90ad28/attachment.html>


More information about the llvm-bugs mailing list