[llvm-bugs] [Bug 31118] New: AVX512: inefficient code generation

[via llvm-bugs]

https://llvm.org/bugs/show_bug.cgi?id=31118

            Bug ID: 31118
           Summary: AVX512: inefficient code generation
           Product: libraries
           Version: trunk
          Hardware: PC
                OS: Linux
            Status: NEW
          Severity: normal
          Priority: P
         Component: Backend: X86
          Assignee: unassignedbugs at nondot.org
          Reporter: Roland.Schulz at Intel.com
                CC: llvm-bugs at lists.llvm.org
    Classification: Unclassified

Compared to ICC 17u1 on KNL, clang produces inefficient code for intrinsic code

Steps to reproduce:
git clone -b clang https://github.com/rolandschulz/gromacs.git
cd gromacs
mkdir clang
cd clang
CC=clang CXX=clang++ cmake -DGMX_MPI=on -DGMX_SIMD=AVX_512_KNL
-DGMX_BUILD_HELP=OFF -DBUILD_SHARED_LIBS=no -DGMX_CYCLE_SUBCOUNTERS=yes ..
make
cd ..
mkdir icc
cd icc
CC=icc CXX=icpc cmake .. -DGMX_FFT_LIBRARY=mkl -DGMX_SIMD=AVX_512_KNL
-DGMX_MPI=on -DGMX_BUILD_HELP=OFF -DBUILD_SHARED_LIBS=no
-DGMX_CYCLE_SUBCOUNTERS=yes

On KNL (7250) for both binaries (clang/bin and icc/bin)
OMP_WAIT_POLICY=active mpirun -n 34 numactl -m 1 gmx_mpi mdrun -ntomp 4 -nsteps
8000 -resetstep 4000 -noconfout  -nstlist 35 -npme 7 -notunepme -pin on

My results (values in second for function which is the most important compute
kernel and mostly intrinsics):
awk /Nonb/'{print $6}' md.log
ICC: 4.921
Clang: 5.349

-- 
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/20161122/42f8a38a/attachment-0001.html>