Skip to content
Snippets Groups Projects
  1. May 02, 1997
  2. Apr 25, 1997
  3. Apr 20, 1997
  4. Apr 13, 1997
  5. Apr 09, 1997
    • dtc's avatar
      New value for the :pure option of defstruct, :substructure. Purify · 3d400aaa
      dtc authored
      will place the instance in the static space but still move the
      substructure lists and arrays to the RO space. This is needed for
      compact-info environments which cache the last name in the instance
      which needs to be scavenged, but the hash arrays can still be placed
      in the RO space (the cache name was not being scavenge which could
      cause a false hit if very unlucky). Shall uses this new option in
      compiler/globaldb.lisp after people have had a chance to bootstrap in
      this new support.
      3d400aaa
  6. Apr 02, 1997
  7. Apr 01, 1997
  8. Mar 26, 1997
  9. Mar 25, 1997
  10. Mar 20, 1997
  11. Mar 15, 1997
  12. Mar 13, 1997
  13. Mar 05, 1997
  14. Feb 27, 1997
  15. Feb 25, 1997
  16. Feb 23, 1997
  17. Feb 22, 1997
    • dtc's avatar
      Use a separate stack for alien value allocation on the x86. It had · 1975d13a
      dtc authored
      been sharing the control stack with the compiler which could cause
      trouble (pushes and pops out of order, holes in arguments built up on
      the stack, exposed stack that can be written over on interrupt, etc)
      
      Assigned a static symbol, x86::*alien-stack*, and by default set it to
      the old C stack. A program many rebind this if it needs more room or
      wants the allocation elsewhere.
      
      The *alien-stack* is saved with the dynamic environment so is restored
      with a non-local exist. It's not save on function entry/exit so you
      need to take care to match an allocation with a deallocation.
      
      Makes use of new VOPs which are already in place.
      
      Backout an old alien stack deallocation hack.
      1975d13a
    • pw's avatar
      6d8260e5
  18. Feb 21, 1997
    • dtc's avatar
      Fix multiplication bug. Multiply-bignum-and-fixnum was using the · f3718883
      dtc authored
      following to convert a fixnum to an (unsigned-byte 32) representation:
      (%fixnum-to-digit (if pos val (- val)))
      This broken when val was (- #x20000000) as #x20000000 is not a fixnum as
      expected by %fixnum-to-digit.
      Similar problem in multiply-fixnums.
      It may be best to just leave it up to the compiler to move the positive
      fixnum to an unsinged register here.
      f3718883
  19. Feb 20, 1997
  20. Feb 19, 1997
  21. Feb 15, 1997
    • dtc's avatar
      Change the index used in the SAP functions to a fixnum. This is handy · 4053ef66
      dtc authored
      on the x86 as the stack grows down and offsets are often negative. It
      also fixes an inconsistency with sap- which accepts SAPs but could
      fail in some cases due to compiler transforms to other sap operations
      which only accept a positive index.  This also requires slight backend
      VOPS changes (currently only supported on the x86 port). The other
      ports should pick this one up.
      4053ef66
  22. Feb 13, 1997
  23. Feb 12, 1997
  24. Feb 11, 1997
  25. Feb 10, 1997
  26. Feb 09, 1997
    • pw's avatar
      In maybe-diddle-case, when working on a PATTERN object, the original · 9ad69e85
      pw authored
      object was returned wrapped up as the first piece of a new PATTERN.
      This was traced to a typo where 'thing' (the original pattern) was
      passed to 'function' instead of the lambda arg 'piece'.
      Now, (directory "HOST:DIR;NAME*") does something reasonable.
      9ad69e85
  27. Feb 08, 1997
Loading