1. 14 Nov, 2018 1 commit
  2. 22 Oct, 2018 1 commit
  3. 25 Nov, 2017 1 commit
  4. 27 Mar, 2014 1 commit
    • Artem Shinkarov's avatar
      DBUG_UNREACHABLE macro to replace DBUG_ASSERT (<false-expr> cases · 491595f3
      Artem Shinkarov authored
      We introduce DBUG_UNREACHABLE macro to mark the places in the code
      that must not happen.  We produce an error message stating that we
      hit such a place in the code.
      
          * debug.[ch]:    DBUG_UNREACHABLE macro defined for both DBUG_OFF
                           and !DBUG_OFF cases.
      
                           exit_function was moved outside of !DBUG_OFF
                           region.
      
          * ctinfo.[ch]:   Removed CTIterminateCompilationSilient in favour
                           of existing CTIexit and replaced occurences.
      
          * main.c         Adjusted comments to match doxygen style, set
                           exit_function unconditionally to DBUG_OFF.
      491595f3
  5. 17 Apr, 2013 1 commit
    • Artem Shinkarov's avatar
      Removing $Id$ and $Log$ entries from the files. · a3df33b7
      Artem Shinkarov authored
      After migration to git the annotations became useless, as they cannot be
      expanded under git.  Part of the annotations were wrong even under svn.
      As for logs, all the logs can be retrieved per file using
      git log <filename>
      a3df33b7
  6. 21 Aug, 2012 1 commit
    • Robert Bernecky's avatar
      WLF assumes that, if the result of a producerWL is AKS, · 6021ebb1
      Robert Bernecky authored
      then GENARRAY_SHAPE of that WL is AKV.
      
      A recent extension to WLPROP, allowing it to work on CONDFUNS
      (not checked in yet), causes an assert failure in WLF, because the 
      producerWL, when copied into the CONDFUN, has AKS result
      shape, but GENARRAY_SHAPE, now an N_arg to the CONDFUN,
      is no longer AKV, because WLBSC got its fingers in the pie,
      and generates an AKS N_array for that shape.
      
      The change here merely makes the GENARRAY_SHAPE check for AKV
      early enough that WLF does not attempt to perform the fold.
      
      Presumably, on another trip through CYC/SAACYC, things
      will untangle and the WLF will happen.
      
      6021ebb1
  7. 13 Jul, 2012 1 commit
  8. 21 Jun, 2012 2 commits
  9. 18 Jun, 2012 1 commit
  10. 27 Apr, 2012 1 commit
  11. 21 Aug, 2011 2 commits
  12. 18 Aug, 2011 1 commit
  13. 17 Aug, 2011 1 commit
  14. 03 Jul, 2011 1 commit
  15. 19 Nov, 2009 1 commit
  16. 06 Jun, 2009 2 commits
  17. 03 May, 2009 1 commit
  18. 07 Apr, 2009 1 commit
    • Robert Bernecky's avatar
      Mostly boilerplate and setup for extended WLF. · 562292c3
      Robert Bernecky authored
      1. -ssaiv enables code to force WL index variables
         into SSA form. This definitely kills WLF today, 
         manifesting itself as references to undefined IVs.
         If you use it with -nowlf, you may get further.
      
      2. -doswlf extended:
        - Operates on multiple WL partitions in both folder
          and foldee WLs.
      
        - Maps modarray(foldee) WLs into genarray(shape(foldee)),
          to eliminate reference to foldee data.
      
        - Slightly improved analysis of WL bounds. However,
          the whole array shapes and bounds
          thing remains a dog's breakfast. 
      
      Operation to invoke extended WLF requires:
      
       -ssaiv -doswlf -dosaa
      
      and may also require
      
       -nowlf
      
      562292c3
  19. 11 Feb, 2009 1 commit
  20. 27 Jan, 2009 1 commit
    • Robert Bernecky's avatar
      1. Flattening code for phases 10 and 11. There's lot more to do, · e79a856a
      Robert Bernecky authored
         but I want to get some stuff checked in. Theoretically, this is
         all enabled only with sac2c option -ssaiv. The intent is to make
         that option go away, and to make flattened code the standard for
         all code in the above phases.
      
      2. -d treecheck code to detect non-flattened GENERATOR_BOUND1 and
         GENERATOR_BOUND2, which are the only ones I absolutely need to
         be flattened, at present.
      
      
      e79a856a
  21. 06 Nov, 2008 2 commits
  22. 05 Nov, 2008 2 commits
  23. 20 Jun, 2007 1 commit
  24. 05 Feb, 2007 3 commits
  25. 29 Jan, 2007 1 commit
  26. 25 Oct, 2006 1 commit
  27. 04 Jul, 2006 1 commit
  28. 03 Jul, 2006 1 commit
  29. 24 Mar, 2006 1 commit
  30. 22 Nov, 2005 1 commit
  31. 18 Nov, 2005 1 commit
  32. 10 Nov, 2005 1 commit
  33. 10 Oct, 2005 1 commit