IV65273: COMPILER DOES NOT USE VECTOR OPS FOR MEMSET

XL C/C++ for AIX Fix Pack 1 (September 2014 PTF) for 13.1
XL C/C++ for AIX Fix Pack 2 (December 2014 PTF) for 13.1
XL C/C++ for AIX Fix Pack 3 (April 2015 PTF) for 13.1
XL C/C++ for AIX Fix Pack 4 (July 2015 PTF) for 13.1
XL C for AIX Fix Pack 5 (November 2015 Update) for 13.1
XL C/C++ for AIX Fix Pack 5 (November 2015 PTF) for 13.1
XL C/C++ for AIX Fix Pack 6 (February 2016 PTF) for 13.1
XL C for AIX Fix Pack 6 (February 2016 Update) for 13.1
XL C for AIX Fix Pack 7 (May 2016 Update) for 13.1
XL C/C++ for AIX Fix Pack 7 (May 2016 PTF) for 13.1
XL C for AIX Fix Pack 8 (August 2016 Update) for 13.1
XL C/C++ for AIX Fix Pack 8 (August 2016 PTF) for 13.1
XL C for AIX Fix Pack 9 (June 2017 Update) for 13.1
XL C/C++ for AIX Fix Pack 9 (June 2017 PTF) for 13.1
XL C for AIX Fix Pack 10 (February 2018 Update) for 13.1
XL C/C++ for AIX Fix Pack 10 (February 2018 PTF) for 13.1
XL C for AIX Fix Pack 1 (September 2014 PTF) for 13.1
XL C for AIX Fix Pack 2 (December 2014 PTF) for 13.1
XL C for AIX Fix Pack 3 (April 2015 PTF) for 13.1
XL C for AIX Fix Pack 4 (July 2015 PTF) for 13.1
XL C/C++ for AIX Fix Pack 11 (November 2018 PTF) for 13.1
XL C for AIX Fix Pack 11 (November 2018) for 13.1

APAR status

• Closed as program error.

Error description

• xlC does not take the opportunity to use vector splat immediate
  for memset.
  For input values -16..+15 for second parameter a single vspltisb
  instruction can be used to construct a vector.
  
  ===== COMPILE COMMAND:
  xlC -q64 -O2 -qlist test.cpp -qarch=pwr7 -qaltivec
  
  ===== TESTCASE:
  $ cat test.cpp
  #include <builtins.h>
  #include <string.h>
  
  void memset16(void *out)
  {
     __alignx(4,out);
     memset(out, 0, 16);
  }
  $
  
  
  ===== ACTUAL OUTPUT:
    VR's set/used:   ---- ---- ---- ----  ---- ---- ---- ----
       | 000000                           PDEF     memset16(void
  *)
       | 000000                           AKA       memset16__FPv
      4|                                  PROC      out,gr3
    398| 000000 addi     38000000   1     LI        gr0=0
    398| 000004 std      F8030000   1     ST8
  (char)(gr3,0)=gr0
    398| 000008 std      F8030008   1     ST8
  (char)(gr3,8)=gr0
      8| 00000C bclr     4E800020   1     BA        lr
       |               Tag Table
       | 000010        00000000 00092000 00000000 00000010
       |               Instruction count            4
  
  
  
  ===== EXPECTED OUTPUT:
  The output could also look like:
  xxlxor 0,0,0
  stvd2x 0,0,<out>
  

Local fix

• N/A
  

Problem summary

• USERS AFFECTED:
  Users who use memset are affected by this issue.
  
  PROBLEM DESCRIPTION:
  The compiler was generating inefficient code for memset code
  when -qarch=pwr7 was specified.
  

Problem conclusion

• The compiler has been fixed so that when changing memset into
  vector instructions, the compiler uses two other instructions
  for edge cases to generate 0x00 and 0xff to have one cycle saved
  on power 7.
  

Temporary fix

Comments

APAR Information

• APAR is sysrouted FROM one or more of the following:

  IV63941

• APAR is sysrouted TO one or more of the following:

  LI78244

Fix information

• Fixed component name

  XL C FOR AIX

• Fixed component ID

  5725C7100

Applicable component levels

• RD10 PSY

     UP