Skip to content
Snippets Groups Projects
  1. Sep 13, 2004
  2. Sep 11, 2004
    • rtoy's avatar
      Port of SBCL's float-accuracy compilation policy. Intended to make · 571091c7
      rtoy authored
      double-float-epsilon actually be epsilon on x86.  No effect on other
      ports.
      
      The default precision is now 53-bit (double-float) instead of 64-bit
      (80-bit floats).  However, to preserve C expectations, all calls to C
      have the precision set to 64-bit.  This slows down calls to C, but we
      try to make syscalls and such fast by not changing precision for the
      call.
      
      By default ext:float-accuracy is 3.
      
      Use boot7.lisp to bootstrap.
      571091c7
  3. Sep 10, 2004
  4. Sep 09, 2004
  5. Sep 08, 2004
    • rtoy's avatar
      Make UPGRADED-COMPLEX-PART-TYPE return RATIONAL for rational parts. · 61d4e4e6
      rtoy authored
      (Should we have it do something else for (EQL 0)?)
      61d4e4e6
    • rtoy's avatar
      o Change the def-type-translator for COMPLEX to fix some confusion in · d321a141
      rtoy authored
        the compiler about what COMPLEX and (COMPLEX REAL) are.  They should
        be the same.
        - Make (COMPLEX *) be the same as (COMPLEX REAL), because we were
          creating a plain COMPLEX, which isn't what we want.
        - If the typespec is a subtype of RATIONAL, return (COMPLEX
          RATIONAL).
      o Make (COMPLEX REAL) print out as COMPLEX instead of as a union of
        specialized complex types.
      o Modify CTYPE-OF-NUMBER so that it no longer creates (COMPLEX
        (INTEGER low high)) types; these are now (COMPLEX RATIONAL).  This
        fixes a bug where (type-of #c(0 1/2)) was giving an error.  (This is
        fixes the bug by removing the offending code.  But do we really want
        and need to support (COMPLEX INTEGER) types in the compiler?
        Shouldn't (COMPLEX RATIONAL) be good enough?)
      d321a141
    • rtoy's avatar
      Port lazy sxhash on symbols and one-at-a-time hash to ppc. Use · 96757501
      rtoy authored
      boot6-sxhash to bootstrap this change for ppc.
      96757501
  6. Sep 07, 2004
  7. Sep 03, 2004
    • rtoy's avatar
      The CLHS entry for *PRINT-LENGTH* says · 9836e815
      rtoy authored
          If an object to be recursively printed has components and is at a
          level equal to or greater than the value of *print-level*, then
          the object is printed as ``#''.
      
      Adjust the default structure printer and %print-unreadable-object to
      do this.  This fixes the PRINT-LEVEL.2, PRINT-LEVEL.6, PRINT-LEVEL.8,
      and PRINT-LEVEL.9. test failures.
      9836e815
  8. Sep 02, 2004
    • rtoy's avatar
      o PPRINT-INDENT should accept any real for the indentation. We · 24d3f886
      rtoy authored
        silently truncate that to integer.
      
      o PPRINT-LOGICAL-BLOCK is supposed to signal a type-error if :suffix,
        :prefix, or :per-line-prefix is supplied but is not a string.  We
        check for this, and make sure we don't pass in NIL in other parts of
        the code.
      
      These issues were found in ansi-tests.
      24d3f886
  9. Sep 01, 2004
  10. Aug 31, 2004
  11. Aug 30, 2004
    • rtoy's avatar
      COMPUTE-TAB-SIZE: · 6a5a1c87
      rtoy authored
      o Previous changes were wrong.  This version handles relative and
        section-relative columns correctly,  I think.
      
      PPRINT-TABULAR
      o The computation of colinc was wrong---I don't think we need to add
        one to tabsize.
      o I don't think we should output the space character.  It results in
        one extra space.
      
      With these changes, the format tests in ansi-tests for ~T and friends
      pass.  I think these are right now, but I'm not 100% sure.
      6a5a1c87
    • rtoy's avatar
      Apply patch from Jan Rychter that implements return-from-frame · d0eecd42
      rtoy authored
      (debug-return).  No bootstrap file or cross-compile needed.
      Return-from-frame only works when debug = 3.
      d0eecd42
  12. Aug 28, 2004
  13. Aug 27, 2004
  14. Aug 23, 2004
  15. Aug 22, 2004
  16. Aug 20, 2004
  17. Aug 17, 2004
  18. Aug 13, 2004
  19. Aug 04, 2004
  20. Aug 03, 2004
  21. Aug 02, 2004
  22. Jul 30, 2004
  23. Jul 29, 2004
    • rtoy's avatar
      (defun foo () (zot)) when compiled and run gave a silly message about · 7d79b30c
      rtoy authored
      #C(0d0 0d0) failing a typecase.  Then doing a backtrace gave another
      error about some negative number not being a positive fixnum.
      
      The former error is caused by a bad definition of undefined_tramp.
      The latter error is caused by the debugger not being able to compute
      the offset correctly.
      
      (Fixes ported from sbcl.)
      7d79b30c
Loading