1. New flag - BM_JUST_DIRTIED - added for BufferDesc;
2. All data "dirtiers" (WriteBuffer and WriteNoReleaseBuffer)
   set this flag (and BM_DIRTY too);
3. All data "flushers" (FlushBuffer, BufferSync and BufferReplace)
   turn this flag off just before calling smgr[blind]write/smgrflush
   and check this flag after flushing buffer: if it turned ON then
   BM_DIRTY will stay ON.

losing data changes.

   smgrcreate ().
2. Checking rdesc->rd_tmpunlinked in heap_destroy () & heap_destroyr ()
   before calling smgrunlink ().

Must find alternate solution to libpq shared library build problem.

don't accidentally use non-GNU make.

Remove RELEASE.patchlevel...it hasn't been modified since 1.02 *sigh*

Pointed out by Andrew

pointed out by Andrew...

sure exactly what it is it fixes...

Makefile.custom and pointers at the customize script...

by caller.

SetBufferWriteMode () added;
FlushBuffer () fixed: now directly calls smgrflush () and
	releases buffer only if required by caller.

after writing next free XID.

release buffer if called from WriteNoReleaseBuffer ())

POinted out by: Martin S. Utesch <utesch@aut.tu-freiberg.de>

Some compilers recognize this error.

all the other port related dependencies are ifeq, not ifneq..

If I'm wrong, someone tell me or change it back...