I am using Spark3.1.1/Hadoop3.2.2 on Solaris 11. When using Spark to write dataframe to Hadoop in parquete format, JVM crashed with SIGBUS error (Java Frame: sun.misc.UnSafe.GetInt). Error happens only when running in Solaris with SPARC 64 bit, not on other platforms such as Centos or Ubuntu. I don't know if there are problems with Spark 3.x on SPARC platform or not? Anybody has any idea? JVM crashed on Spark writing parquet