which probably isn't needed, since we are going to take off -Werror
anyway, right?

(but I didn't test FUNC OP PARAM and PARAM OP FUNC - no time)

2. IndexScanableOperand now uses match_indexkey_operand
instead of equal_indexkey_var (if we have some index on attribute X
then we shouldn't use it for 'where some_func(X) OP CONST').

ExecInitIndexScan now works with operands of Param type and
(!!!) postquel_execute() now substitutes param values
before calling postquel_start().

/usr/include/limits.h (which quiets the costsize.c warnings)...under
FreeBSD, /usr/include/limits.h *includes* machine/limits.h, while under
Solaris, there is no such things as /usr/include/machine...

Problem with Solaris pointed out by Mark Wahl

Pointed out by Mark Wahl

Invalidate vacuum relation cache to use new row counts from vacuum.

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.