Skip to content
Snippets Groups Projects
  1. Feb 08, 1998
  2. Feb 06, 1998
    • dtc's avatar
      Re-write record-fields-match which is used in the testing of alien · a8b760a4
      dtc authored
      type equivalence.  Cyclic record and pointer types, which CMUCL
      supports, could cause the algorithm to infinitely recurse and overflow
      the stack. The fixed limits the depth of the search by assuming
      pointer types are equivalent beyond a certain depth.
      a8b760a4
  3. Feb 05, 1998
  4. Feb 03, 1998
  5. Feb 02, 1998
    • dtc's avatar
      Patch from Raymond Toy for the :negative-zero-is-not-zero feature: · b6137f53
      dtc authored
      have make-numeric-type canonicalize the bounds so that a lower bound
      of (-0.0) becomes 0.0 and an upper bound of (0.0) becomes -0.0.  Also
      fixes the problem of (subtypep '(float (-0.0)) (float 0.0)), returning
      nil when it should be T.  This is because (float (-0.0)) is
      canonicalized to the identical (float 0.0).
      b6137f53
  6. Jan 30, 1998
  7. Jan 29, 1998
  8. Jan 26, 1998
    • dtc's avatar
      Take more care to inhibit scheduling when modifying process state and · 70464fca
      dtc authored
      the *all-processes* list, function affected: make-process,
      destroy-process, restart-process, disable-process, and enable-process.
      
      Save and restore the process wait return value when a process is
      interrupted in case a process is interrupted before it reads the
      return value.
      70464fca
  9. Jan 25, 1998
  10. Jan 21, 1998
  11. Jan 20, 1998
  12. Jan 19, 1998
  13. Jan 17, 1998
  14. Jan 16, 1998
  15. Jan 15, 1998
  16. Jan 13, 1998
    • dtc's avatar
      The alien object finalizer was being placed on the alien info, but · a04bd1a5
      dtc authored
      this info may be shared among many alien objects making cancellation
      of the finalisation upon alien object deallocation
      problematic. Probably for this reason this cancellation had been
      disabled, however this often resulted in objects being freed twice and
      possibly live objects being freed. To fix this, the finalizer is now
      placed on the alien object rather than the alien info, and this
      finalizer is cancelled before the deallocation of the alien object to
      prevent repeated freeing. Problem tracked down thanks to a bug report
      and some experimentation by Andrei V. Elkin.
      a04bd1a5
    • dtc's avatar
      Have the lisp connection listener search forward for an available · 2a747380
      dtc authored
      port, and add at least some simple password protection.
      2a747380
    • dtc's avatar
      Re-seed the *random-state* as an after save initialisation. · d2b81e7a
      dtc authored
      Move the loading of the random number generator until after the
      loading of save.lisp so that *after-save-initializations* is
      initialised.
      d2b81e7a
  17. Jan 12, 1998
  18. Jan 11, 1998
  19. Jan 08, 1998
  20. Jan 05, 1998
    • dtc's avatar
      Change to the handling of floating point zeros within numeric type · 3fe8030e
      dtc authored
      specifiers by Raymond Toy. With these changes -0.0 and 0.0 and
      considered separate on the float type intervals which is more natural
      for many arithmetic function derive type optimizers.  These changes
      are dependent upon the :negative-zero-is-not-zero feature.
      3fe8030e
    • dtc's avatar
      Patch from Timothy Miller: · bf230599
      dtc authored
      In the debugger, breakpoint <loc> :function <fn> wouldn't actually put
      the breakpoint in the right function if <fn> was different from the
      current function.
      bf230599
  21. Jan 04, 1998
  22. Jan 03, 1998
  23. Jan 02, 1998
  24. Jan 01, 1998
    • dtc's avatar
      Fix a bug in the scrubbing of the current stack-group's eval-stack. · 30a881ca
      dtc authored
      Update the eval-stack for the current stack-group when scrubbing to
      ensuring that the stack-top is valid and maybe flush an old stored
      vector.
      
      Completely scrub the stored binding stack for the current stack group.
      
      Implement the lock waiting in a separate function rather than
      inlining. This ensures that the wait predicate is fast native code so
      that locking in byte-compiled and interpreted code doesn't slow the
      scheduler.
      30a881ca
  25. Dec 31, 1997
Loading