Junchao Zhang

@liruipeng It will be helpful if we can set up a call. Thanks.

> @lucbv @jczhang07 Hi guys, is there an update on this? I have some overlapping changes that I would like to put in to help support rocsparse in Tpetra, but...

I still met the same error today with updated Perlmutter default settings.

Yes, to impose boundary conditions and to implement PETSc MatZerosRows (https://petsc.org/release/manualpages/Mat/MatZeroRows) with a Hypre CSRMatrix underneath.

@lucbv I forgot this MR hadn't been merged yet until I tested a SpMV benchmark on Frontier recently and saw the terrible performance again (https://github.com/kokkos/kokkos-kernels/pull/1469#issue-1306297831). I rebased the MR off...

@brian-kelley Could you test this PR in CI?

On a Crusher compute, with `export HIP_VISIBLE_DEVICES=`, I could still run petsc GPU tests. I used rocm/5.2.0.

Yes, I did have `oneapi/eng-compiler/2024.07.30.002`. From the AUTOTEST script, I re-tried with `--with-sycl --enable-unified-memory` only, but the errors still existed.

With hypre/master, that error went away. But there were new errors with respect to MPI symbols. I think I know why. `./configure --prefix=/home/jczhang/petsc/arch-kokkos-dbg MAKE=/opt/aurora/24.180.3/spack/unified/0.8.0/install/linux-sles15-x86_64/gcc-12.2.0/gmake-4.4.1-6g37exp/bin/gmake --libdir=/home/jczhang/petsc/arch-kokkos-dbg/lib CC="mpicc" CFLAGS="-fPIC -Wno-sign-conversion -Wno-float-conversion -Wno-implicit-float-conversion...

Attached. [config.log](https://github.com/user-attachments/files/18612288/config.log) [Makefile.config.txt](https://github.com/user-attachments/files/18612295/Makefile.config.txt)