ix86_expand_setcc to clear + set strict_low_part

Jeffrey A Law law@cygnus.com
Tue Jan 11 05:29:00 GMT 2000


  In message < ord7ra6zfl.fsf@garnize.lsd.ic.unicamp.br >you write:
  > --=-=-=
  > 
  > I was investigating why ix86_expand_setcc said:
  > 
  >   type = 0;
  >   /* %%% reload problems with in-out.  Revisit.  */
  >   type = 1;
  > 
  > In fact, removing the assignment `type=1' caused stage1 to fail to
  > build stage2, with various ICEs within verify_local_live_at_start(),
  > at the end of peephole2_optimize().
  > 
  > I managed to create a minimal testcase:

Please install the testcase into the testsuite.

I believe rth already commented on the patch itself.

jeff



More information about the Gcc-patches mailing list