src/hotspot/cpu/ppc/templateInterpreterGenerator_ppc.cpp

rev 53302 : 8216060: [PPC64] Vector CRC implementation should be used by interpreter and be faster for short arrays
Reviewed-by: gromero, goetz

*** 1,8 ****
  /*
!  * Copyright (c) 2014, 2018, Oracle and/or its affiliates. All rights reserved.
!  * Copyright (c) 2015, 2018, SAP SE. All rights reserved.
   * DO NOT ALTER OR REMOVE COPYRIGHT NOTICES OR THIS FILE HEADER.
   *
   * This code is free software; you can redistribute it and/or modify it
   * under the terms of the GNU General Public License version 2 only, as
   * published by the Free Software Foundation.
--- 1,8 ----
  /*
!  * Copyright (c) 2014, 2019, Oracle and/or its affiliates. All rights reserved.
!  * Copyright (c) 2015, 2019, SAP SE. All rights reserved.
   * DO NOT ALTER OR REMOVE COPYRIGHT NOTICES OR THIS FILE HEADER.
   *
   * This code is free software; you can redistribute it and/or modify it
   * under the terms of the GNU General Public License version 2 only, as
   * published by the Free Software Foundation.
*** 1916,1929 ****
        __ addi(data, data, arrayOopDesc::base_offset_in_bytes(T_BYTE));
      }
  
      StubRoutines::ppc64::generate_load_crc_table_addr(_masm, table);
  
!     // Performance measurements show the 1word and 2word variants to be almost equivalent,
!     // with very light advantages for the 1word variant. We chose the 1word variant for
!     // code compactness.
      __ kernel_crc32_1word(crc, data, dataLen, table, t0, t1, t2, t3, tc0, tc1, tc2, tc3, true);
  
      // Restore caller sp for c2i case (from compiled) and for resized sender frame (from interpreted).
      __ resize_frame_absolute(R21_sender_SP, R11_scratch1, R0);
      __ blr();
  
--- 1916,1931 ----
        __ addi(data, data, arrayOopDesc::base_offset_in_bytes(T_BYTE));
      }
  
      StubRoutines::ppc64::generate_load_crc_table_addr(_masm, table);
  
!     if (!VM_Version::has_vpmsumb()) {
        __ kernel_crc32_1word(crc, data, dataLen, table, t0, t1, t2, t3, tc0, tc1, tc2, tc3, true);
+     } else {
+       StubRoutines::ppc64::generate_load_crc_constants_addr(_masm, t0);
+       __ kernel_crc32_vpmsum(crc, data, dataLen, table, t0, t1, t2, t3, tc0, tc1, tc2, true);
+     }
  
      // Restore caller sp for c2i case (from compiled) and for resized sender frame (from interpreted).
      __ resize_frame_absolute(R21_sender_SP, R11_scratch1, R0);
      __ blr();
  
*** 2004,2017 ****
        __ addi(data, data, arrayOopDesc::base_offset_in_bytes(T_BYTE));
      }
  
      StubRoutines::ppc64::generate_load_crc32c_table_addr(_masm, table);
  
!     // Performance measurements show the 1word and 2word variants to be almost equivalent,
!     // with very light advantages for the 1word variant. We chose the 1word variant for
!     // code compactness.
      __ kernel_crc32_1word(crc, data, dataLen, table, t0, t1, t2, t3, tc0, tc1, tc2, tc3, false);
  
      // Restore caller sp for c2i case (from compiled) and for resized sender frame (from interpreted).
      __ resize_frame_absolute(R21_sender_SP, R11_scratch1, R0);
      __ blr();
  
--- 2006,2021 ----
        __ addi(data, data, arrayOopDesc::base_offset_in_bytes(T_BYTE));
      }
  
      StubRoutines::ppc64::generate_load_crc32c_table_addr(_masm, table);
  
!     if (!VM_Version::has_vpmsumb()) {
        __ kernel_crc32_1word(crc, data, dataLen, table, t0, t1, t2, t3, tc0, tc1, tc2, tc3, false);
+     } else {
+       StubRoutines::ppc64::generate_load_crc32c_constants_addr(_masm, t0);
+       __ kernel_crc32_vpmsum(crc, data, dataLen, table, t0, t1, t2, t3, tc0, tc1, tc2, false);
+     }
  
      // Restore caller sp for c2i case (from compiled) and for resized sender frame (from interpreted).
      __ resize_frame_absolute(R21_sender_SP, R11_scratch1, R0);
      __ blr();