Skip to content
Snippets Groups Projects
  1. Oct 09, 2004
  2. Oct 07, 2004
  3. Oct 06, 2004
  4. Oct 04, 2004
  5. Sep 23, 2004
  6. Sep 21, 2004
  7. Sep 13, 2004
  8. 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
  9. Sep 08, 2004
  10. Sep 03, 2004
  11. Aug 30, 2004
  12. Aug 23, 2004
    • emarsden's avatar
      · 41def4d2
      emarsden authored
      When naming a component in FIND-COMPONENT-NAME, don't require the component's
      context to be printable readably. Fixes a bug triggered by cl-pdf, where the
      context is a CLOS object which is not readably printable.
      41def4d2
  13. Aug 22, 2004
  14. Aug 20, 2004
  15. Aug 16, 2004
  16. Aug 12, 2004
  17. Aug 09, 2004
  18. Aug 08, 2004
  19. Aug 04, 2004
    • rtoy's avatar
      On sparc, undefined_tramp and closure_tramp are foreign data, not · efd44e7b
      rtoy authored
      foreign functions for linkage table stuff.  (This should probably be
      changed.)
      
      In cold-fdefinition-object and cold-fset, we call
      lookup-foreign-symbol and stuff away the address of the entry in the
      linkage-table.  I think we really want the contents of the entry since
      that contains the actual address for these functions.
      
      Thus, change lookup-foreign-symbol to take an link-type arg, and for
      sparc, when the link-type is :data, return the value from the
      linkage-table entry instead of returning the address of the
      linkage-table entry.
      
      I hope this fixes the occasional problem with undefined functions
      giving sigbus instead of an undefined-function error when loading
      kernel.core.
      efd44e7b
  20. Aug 03, 2004
  21. Aug 02, 2004
  22. Jul 30, 2004
  23. Jul 29, 2004
  24. Jul 28, 2004
  25. Jul 27, 2004
  26. Jul 25, 2004
    • pmai's avatar
      Commit of the compiler backend portion of the PPC/Darwin port. · cd45202a
      pmai authored
      This includes various fixes from SBCL and other sources to the PPC backend.
      It also does away with Gary Byers' renumbering of the type tags for
      function and instance pointers, which allowed for a slightly clever and
      presumably faster function calling convention.  These changes have been
      conditionalized out with the PPC-FUN-HACK conditional, so that we might
      revisit the changes in the future.  The current port has only been tested
      without PPC-FUN-HACK, but this might change.
      cd45202a
  27. Jul 21, 2004
  28. Jul 20, 2004
  29. Jul 19, 2004
Loading