Closed ndellingwood closed 3 months ago
Yeah, I'm seeing similar errors in my block sptrsv PR. I'm wondering how that cleanup PR made it past AT testing.
@jgfouca I don't think the intel/19 CI job tests a debug build?
They aren't but I'm seeing the same mishandling of complex data on one of the weaver jobs (but not the other for some reason).
Nightly tests are failing in in the
sparse_sptrsv_kokkos_complex_double_int_int_TestDevice
subtest with intel compilers and nvhpc with output:This began following merge of #2280 , @jgfouca can you investigate?
Reproducer (sems-intel-19.0.5):