changelog

Linux下的中文输入法

   ____  _   _
  |  _ \| |_| |__
  | |_) | __| '_ \             ``Just because everything is different
  |  __/| |_| | | |              doesn't mean anything has changed.''
  |_|    \__|_| |_|                             -- Irene Peter

  GNU Pth - The GNU Portable Threads

  ChangeLog
  =========

  This is the list of all(!) changes to this Pth source tree. For a list
  of just the user-visible and/or major changes please have a look at
  the NEWS file.

     ____    ___  
    |___ \  / _ \ 
      __) || | | |
     / __/ | |_| |
  __|_____(_)___/_________________________________________________________
                  
  Changes between 2.0.2 and 2.0.3 (12-Sep-2004 to 03-Dec-2004)

   *) Resurrect the old "make striptease" command by fixing the
      configure run-time in the stripped down source tree and fixing the
      list of source files.
      [Ralf S. Engelschall]

   *) The pth_uctx_save() and pth_uctx_restore() API functions
      unfortunately were broken by design because they are C
      _functions_. This leads to one more deadly nesting on the
      run-time stack which effectively caused the pth_mctx_restore()
      in pth_uctx_restore() to return to the end of pth_uctx_save()
      but then the control flow unfortunately returns to the
      pth_uctx_restore() caller instead of the pth_uctx_save() caller
      because the call to pth_uctx_restore() had already overwritten the
      run-time stack position where the original return address for the
      pth_uctx_save() call was stored.
      
      The only workaround would be to #define pth_uctx_save() and
      pth_uctx_restore() as C _macros_, but this then would require that
      lots of the GNU Pth internals from pth_mctx.c would have to be
      exported in the GNU Pth API (which in turn is not acceptable). So,
      the only consequence is to remove the two functions again from the
      GNU Pth API.
      [Ralf S. Engelschall, Stefan Brantschen <[e-mail]>]

   *) Enhance internal pth_mctx_save() if getcontext(3) is used for the
      machine context saving by better emulating the setjmp(3) style
      return code semantics.
      [Ralf S. Engelschall]

   *) Do not break in pth_msgport_find() if a message port
      was created with a NULL name.
      [Ralf S. Engelschall]

   *) Better error checking in pth_key_xxx() functions.
      [Ralf S. Engelschall]

   *) Added PTH_CTRL_FAVOURNEW control which allows the user
      to disable the favouring of new threads on scheduling
      to get more strict priority based scheduling behavior.
      [Ralf S. Engelschall, Vinu V <[e-mail]>]

   *) Upgraded build environment to GNU libtool 1.5.10
      [Ralf S. Engelschall]

  Changes between 2.0.1 and 2.0.2 (13-Jul-2004 to 12-Sep-2004)

   *) Added hard syscall wrappers for send(2) and recv(2).
      [NetBSD pkgsrc patches]

   *) Internally handle errors returned from pth_event() in order
      to pass them upstream to the caller in pth_high.c functions.
      [Ralf S. Engelschall, NetBSD pkgsrc patches]

   *) Fix syscall wrapper for sendto(2).
      [NetBSD pkgsrc patches]

   *) Use "(char *)NULL" instead of plain "NULL" in last argument to
      execl(2) to avoid GCC 3.5's "warning: missing sentinel in function call".
      [Ralf S. Engelschall]

   *) Workaround Autoconf/m4 problems in pth.m4 by using "m4_define"
      instead of just "define" and no using m4_undefine at all.
      [Dr. Andreas Mueller <[e-mail]>]

   *) Avoid side-effects by defining _XOPEN_SOURCE only temporarily
      for inclusion of <poll.h> instead of permanently.
      [Graham Nash <[e-mail]>]

   *) Workaround GCC 2.x and 3.5 (3.4 worked fine) macro parsing
      behaviour by using a pre-declared function pointer type "typedef
      int (*pth_event_func_t)(void *);" instead of using an inlined 
      "int (*)(void *)".
      [Ralf S. Engelschall, Robert Anderson <[e-mail]>]

   *) Fixed prototype for pthread_attr_setschedparam(3) from
      to use a "const struct sched_param *" argument.
      [Rafael 羦ila de Esp韓dola <[e-mail]>]

   *) Upgraded build environment to GNU libtool 1.5.8
      and GNU shtool 2.0.1.
      [Ralf S. Engelschall]

  Changes between 2.0.0 and 2.0.1 (17-Feb-2003 to 13-Jul-2004)

   *) Fix handling/casting of "int (*)(void *)" callback functions.
      [Ralf S. Engelschall]

   *) Resolve symbol conflict in pth_string.c between pow10/round
      and GCC builtins.
      [Ralf S. Engelschall]

   *) Use GCC 3.3 option "-fno-strict-aliasing" (if available) under
      Autoconf option "--enable-debug" because mainly pth_mctx.c
      contains important and correct pointer casting constructs which
      are not acceptable in "strict aliasing" for GCC.
      [Ralf S. Engelschall]

   *) Upgraded build environment to GNU autoconf 2.59, GNU libtool 1.5.6
      and GNU shtool 2.0.0.
      [Ralf S. Engelschall]

   *) Fixed implementation so that --enable-pthread now correctly
      implies --enable-syscall-soft.
      [Steve Alstrin <[e-mail]>, Ralf S. Engelschall]

   *) Replace some 0 by more politically correct NULL in pth_string.c
      [Ralf S. Engelschall]

   *) Adjusted all copyright messages for new year 2004.
      [Ralf S. Engelschall]

  Changes between 2.0b2 and 2.0.0 (04-Dec-2002 to 17-Feb-2003)

   *) Final source tree polishing before release.
      [Ralf S. Engelschall]

   *) Wrote summary for Pth 2.0 for the ANNOUNCE document.
      [Ralf S. Engelschall]

   *) Adjusted all copyright messages for new year 2003.
      [Ralf S. Engelschall]

  Changes between 2.0b1 and 2.0b2 (08-Nov-2002 to 04-Dec-2002)

   *) Upgraded built environment to GNU Autoconf 2.57.
      [Ralf S. Engelschall]

   *) Fixed compiler warnings under Linux: require inclusion of <time.h>
      [Joseph Wayne Norton <[e-mail]>]

   *) Fixed error handling in pth_init(3): it now correctly
      returns an error instead of abort(3)'ing.
      [Ralf S. Engelschall]

   *) Added soft syscall mapping for nanosleep(3) and
      usleep(3) functions also to the Pthread API. 
      [Ralf S. Engelschall]

  Changes between 2.0b0 and 2.0b1 (07-Nov-2002 to 08-Nov-2002)

   *) Make pth_poll(3) more compliant to POSIX.1-2001/SUSv3 poll(2).
      [Ralf S. Engelschall, Nick Hudson <[e-mail]>]

   *) Make pth_select(3) more compliant to POSIX.1-2001/SUSv3 select(2).
      This especially fixes the polling-only situation (timeout = (0,0)).
      [Ralf S. Engelschall]

   *) Make sure that in the even manager a polling-only select(2) call
      uses a correctly initialized timeout parameter.
      [Ralf S. Engelschall]

   *) Added soft and hard syscall mapping for nanosleep(3) and
      usleep(3) functions. 
      [Ralf S. Engelschall]

   *) Fix poll(2) semantics: remove POLLRDNORM from polling
      result if POLLHUP is detected.
      [Paolo Bonzini <[e-mail]>]

   *) Updated PORTING file to include community feedback from
      the last months.
      [Ralf S. Engelschall]

  Changes between 1.4.1 and 2.0b0 (27-Jan-2002 to 07-Nov-2002)

   *) Fix dropping of scheduler thread pools and this way fix memory leak.
      [Ralf S. Engelschall]

   *) Check for PTH_FDMODE_ERROR error conditions internally, too.
      [Ralf S. Engelschall]

   *) Use the new PTH_STATUS_FAILED event status in the scheduler's
      event-manager for filedescriptor events if the internal select(2)
      call returned with an error. Additionally this PTH_STATUS_FAILED
      is recognized by the high-level API functions (pth_select, etc)
      and produce the necessary POSIX conforming return codes (usually
      -1 and errno == EBADF).
      [Ralf S. Engelschall, Thanh Luu <[e-mail]>]

   *) The function "int pth_event_occurred(pth_event_t)" was replaced
      with "pth_status_t pth_event_status(pth_event_t)" where pth_status_t 
      can have values of PTH_STATUS_PENDING (replacing the old FALSE return 
      value of pth_event_occurred), PTH_STATUS_OCCURRED (replacing
      the old TRUE return value of pth_event_occurred), and
      PTH_STATUS_FAILED (a new return value indicating an error in
      processing the event). This was scheduler/event-manager errors can
      be indicated which happended while processing the event. For
      backward compatibility reasons, a macro pth_event_occurred() was
      added. This will be removed soon.
      [Ralf S. Engelschall]

   *) Upgraded to latest version of the OSSP Autoconf macro
      AC_CHECK_EXTLIB in order to fix problems with the current
      Autoconf 2.54 environment.
      [Ralf S. Engelschall]

   *) Add Autoconf option --with-fdsetsize=NUM which allows to
      force a particular (usually higher than the default) FD_SETSIZE
      value for building Pth. Additionally Pth makes sure that the
      application using Pth does not use an even larger size.
      [Nick Hudson <[e-mail]>, Ralf S. Engelschall]

   *) Added thread attribute PTH_ATTR_DISPATCHES which (in bounded
      attribute objects) is incremented every time the context
      is switched to the associated thread. This can be used for
      statistical information.
      [Ralf S. Engelschall]

   *) Added a stand-alone sub-API for manual user-space context
      switching. It is somewhat modeled after the POSIX ucontext(3)
      facility and consists of an opaque data type pth_uctx_t and
      the management functions pth_uctx_create(), pth_uctx_make(),
      pth_uctx_save(), pth_uctx_restore(), pth_uctx_switch() and
      pth_uctx_destroy(). These functions are based on the same
      underlying machine context switching facility (pth_mctx)
      the threads in GNU Pth are using. This facility can be used
      to implement co-routines without a full real multithreading
      environment or even to implement an own multithreading
      environment.
      [Ralf S. Engelschall]

   *) Add a Pth variant of the new POSIX pselect(2) function, including
      soft and hard syscall mapping support for it.
      [Ralf S. Engelschall]

   *) More POSIX compliance for pth_select() in case of invalid
      timeout values and invalid filedescriptors.
      [Ralf S. Engelschall]

   *) Internally switch from "errno_shield {...}" to "pth_shield {...}"
      and from "return_errno(..)" to "return pth_error(...)" in order to
      make the internal error handling a little bit more consistent.
      [Ralf S. Engelschall]

   *) Cleaned and speeded up the pth_exit() processing.
      [Ralf S. Engelschall]

   *) Added POSIX-compliant sanity checks for bad filedescriptors
      to mostly all filedescriptor-based I/O functions in pth_high.c
      [Ralf S. Engelschall]

   *) Added pth_nanosleep() function.
      [Nick Hudson <[e-mail]>, Ralf S. Engelschall]

   *) Allow a NULL name for pth_msgport_create() in order to
      support locally scoped message ports which are not searched via
      pth_msgport_find().
      [Ralf S. Engelschall, Matthew Mondor <[e-mail]>]

   *) Added "hard syscall mapping" support for using 
      SYS___sigprocmask14 instead of SYS_sigprocmask on NetBSD.
      [Nick Hudson <[e-mail]>]

   *) Changes three direct sigprocmask(2) calls with the corresponding
      "hard syscall mapping" macro calls.
      [Nick Hudson <[e-mail]>]

   *) Add Autoconf support for GNU Hurd.
      [B. Douglas Hilton <[e-mail]>]

   *) Completely rewrote the "hard syscall mapping". 

      Previously the internal syscall exit points were based on syscall(2)
      only. This is problematic because it by-passes the C library glue
      code which sometimes performs necessary assembly fiddling in order
      to call the underlying system calls correctly. Additionally,
      syscall(2) does not exists everywhere.

      Now the internal exit points are based on a by-syscall dynamically
      selected combination of RTLD_NEXT+dlsym(2), dlopen(2)+dlsym(2) and
      the known syscall(2) (in this fallback order). This way the "hard
      syscall mapping" is a lot more portable and flexible.
      [Ralf S. Engelschall, Jonathan Schilling <[e-mail]>]

   *) Added support to pth_poll(3) for the poll(2) XPG.4 flags
      POLLD{RD,WR}{NORM,BAND}.
      [Jason Evans <[e-mail]>, Ralf S. Engelschall]

   *) Fixed a long-standing termination bug in pth_exit(3):
      The event handler of pth_exit(3) didn't let pth_exit(3) finish if
      there were any threads on the "dead queue" (where non-detached
      terminated threads are put). Instead it re-entered the scheduler
      which in turn aborted with "**Pth** SCHEDULER INTERNAL ERROR: no
      more thread(s) available to schedule!?!?". This is now fixed by
      not counting the "dead queue" for the determination whether the
      process as a whole should terminate or not.
      [Jonathan Schilling <[e-mail]>]

   *) The manual page stated that system(3) is supported in the
      "Soft System Call Mapping", but in fact it was forgotten in the
      implementation.
      [Jonathan Schilling <[e-mail]>]

   *) Add #define _PTHREAD_T to pthread.h to guard under SCO UnixWare 7
      and OpenUNIX 8.
      [Jonathan Schilling <[e-mail]>]

   *) Upgraded to Autoconf 2.54 environment and cleaned up internal
      Autoconf macro usage.
      [Ralf S. Engelschall]

   *) Fixed lots of English errors in the manual page.
      [Felix Berger <[e-mail]>]

   *) Add optional support for OSSP ex based exception handling. GNU
      Pth (still) does not throw exceptions by itself, but handles the