Initialize GIN metapage correctly when replaying metapage-update WAL record.

I broke this with my WAL format refactoring patch. Before that, the metapage
was read from disk, and modified in-place regardless of the LSN. That was
always a bit silly, as there's no need to read the old page version from
disk disk when we're overwriting it anyway. So that was changed in 9.5, but
I failed to add a GinInitPage call to initialize the page-headers correctly.
Usually you wouldn't notice, because the metapage is already in the page
cache and is not zeroed.

One way to reproduce this is to perform a VACUUM on an already vacuumed
table (so that the vacuum has no real work to do), immediately after a
checkpoint, and then perform an immediate shutdown. After recovery, the
page headers of the metapage will be incorrectly all-zeroes.

Reported by Jeff Janes

Author: hlinnaka

src/backend/access/gin/ginxlog.c

@@ -512,6 +512,7 @@ ginRedoUpdateMetapage(XLogReaderState *record)
 	Assert(BufferGetBlockNumber(metabuffer) == GIN_METAPAGE_BLKNO);
 	metapage = BufferGetPage(metabuffer);
 
+	GinInitPage(metapage, GIN_META, BufferGetPageSize(metabuffer));
 	memcpy(GinPageGetMeta(metapage), &data->metadata, sizeof(GinMetaPageData));
 	PageSetLSN(metapage, lsn);
 	MarkBufferDirty(metabuffer);