./ChangeLog:

2000-12-28  Ross Johnson  <rpj@special.ise.canberra.edu.au>

        * private.c (ptw32_threadStart): Unhandled exceptions are
        now passed through to the system to deal with. This is consistent
        with normal Windows behaviour. C++ applications may use
        set_terminate() to override the default behaviour which is
        to call ptw32_terminate(). Ptw32_terminate() cleans up some
        POSIX thread stuff before calling the system default function
        which calls abort(). The users termination function should conform
        to standard C++ semantics which is to not return. It should
        exit the thread (call pthread_exit()) or exit the application.
        * private.c (ptw32_terminate): Added as the default set_terminate()
        function. It calls the system default function after cleaning up
        some POSIX thread stuff.

        * implement.h (ptw32_try_enter_critical_section): Move
        declaration.
        * global.c (ptw32_try_enter_critical_section): Moved
        from dll.c.
        * dll.c: Move process and thread attach/detach code into
        functions in nonportable.c.
        * nonportable.c (pthread_win32_process_attach_np): Process
        attach code from dll.c is now available to static linked
        applications.
        * nonportable.c (pthread_win32_process_detach_np): Likewise.
        * nonportable.c (pthread_win32_thread_attach_np): Likewise.
        * nonportable.c (pthread_win32_thread_detach_np): Likewise.

        * pthread.h: Add new non-portable prototypes for static
        linked applications.

        * GNUmakefile (OPT): Increase optimisation flag and remove
        debug info flag.

        * pthread.def: Add new non-portable exports for static
        linked applications.

2000-12-11  Ross Johnson  <rpj@special.ise.canberra.edu.au>

        * FAQ: Update Answer 6 re getting a fully working
        Mingw32 built library.

2000-09-09  Ross Johnson  <rpj@special.ise.canberra.edu.au>

        * pthread.h (ctime_r): Fix arg.

./tests/ChangeLog:
2000-12-28  Ross Johnson  <rpj@special.ise.canberra.edu.au>

        * eyal1.c: Increase thread work loads.
        * exception2.c: New test.
        * exception3.c: New test.
        * Makefile: Add new tests exception2.c and exception3.c.
        * GNUmakefile: Likewise.

2000-12-11  Ross Johnson  <rpj@special.ise.canberra.edu.au>

        * cleanup3.c: Remove unused variable.
        * cleanup2.c: Likewise.
        * exception1.c: Throw an exception rather than use
        a deliberate zero divide so that catch(...) will
        handle it under Mingw32. Mingw32 now builds the
        library correctly to pass all tests - see Thomas
        Pfaff's detailed instructions re needed changes
        to Mingw32 in the Pthreads-Win32 FAQ.

1 files changed, 34 insertions, 22 deletions

diff --git a/README b/README
index 25792ca..93bf8cd 100644
--- a/README
+++ b/README
@@ -71,17 +71,6 @@ Known bugs in this snapshot
 

 1. Asynchronous cancelation only works on Intel X86 machines.

 

-2. Running the test "join1.c" with the library built with Mingw32

-and the GNUmakefile included, the test fails with a segmentation (invalid

-page access) exception. The fault appears to be in the assembler code

-emmitted by the compiler [to handle exception contexts] at the

-end of the try block in ptw32_threadStart().

-

-3. There are problems using the libpthreadw32.a stub archive derived

-from either of pthreadVSE.dll or pthreadVCE.dll. The cleanup1.c test

-fails. This is now an expected result of having different EH and cleanup

-handler schemes in the library and application.

-

 

 Caveats

 -------

@@ -144,6 +133,34 @@ pthread_getw32threadhandle_np(pthread_t thread);
 	win32 specific attributes of the thread.

 

 int

+pthread_delay_np (const struct timespec *interval);

+

+        This routine causes a thread to delay execution for a specific period of time.

+        This period ends at the current time plus the specified interval. The routine

+        will not return before the end of the period is reached, but may return an

+        arbitrary amount of time after the period has gone by. This can be due to

+        system load, thread priorities, and system timer granularity.

+

+        Specifying an interval of zero (0) seconds and zero (0) nanoseconds is

+        allowed and can be used to force the thread to give up the processor or to

+        deliver a pending cancelation request.

+

+        This routine is a cancelation point.

+

+        The timespec structure contains the following two fields:

+

+                tv_sec is an integer number of seconds.

+                tv_nsec is an integer number of nanoseconds. 

+

+        Return Values

+

+        If an error condition occurs, this routine returns an integer value

+        indicating the type of error. Possible return values are as follows:

+

+        0          Successful completion. 

+        [EINVAL]   The value specified by interval is invalid. 

+

+int

 pthreadCancelableWait (HANDLE waitHandle);

 

 int

@@ -156,7 +173,7 @@ pthreadCancelableTimedWait (HANDLE waitHandle, DWORD timeout);
 	pthread_cancel has been called. It is implemented using

 	WaitForMultipleObjects on 'waitHandle' and a manually

 	reset w32 event used to implement pthread_cancel.

- 

+

 

 Building under VC++ using either C++ EH or Structured EH

 --------------------------------------------------------

@@ -182,20 +199,15 @@ nmake clean VSE
 Building under Mingw32

 ----------------------

 

-The dll can be built with Mingw32 gcc-2.95.2 or later.

+The dll can be built with Mingw32 gcc-2.95.2-1 after you've

+made the changes to Mingw32 desribed in Question 6 of the FAQ.

+Hopefully versions after gcc-2.95.2-1 won't require the fix.

+

 Run "make" in the source directory (uses GNUmakefile). This builds

 pthreadGCE.dll and libpthreadw32.a.

 

-To generate the libpthreadw32.a file from pthreadVCE.dll rather than

-building the library with G++, run "make fake.a". (You must have

-pthreadVCE.dll already built - see above.)

-Please note that VC++ and GNU exception handling is implemented

-completely differently and so automatic cleanup of objects and

-the propagation of exceptions themselves between the two schemes is not

-going to happen.

-

 You can run the testsuite by changing to the "tests" directory and

-running "make clean" and then "make". See the "Known bugs" section above.

+running "make clean" and then "make".

 

 

 Building the library under Cygwin