]> granicus.if.org Git - postgresql/commit
Clean up ill-advised attempt to invent a private set of Node tags.
authorTom Lane <tgl@sss.pgh.pa.us>
Sat, 6 Aug 2011 18:53:59 +0000 (14:53 -0400)
committerTom Lane <tgl@sss.pgh.pa.us>
Sat, 6 Aug 2011 18:53:59 +0000 (14:53 -0400)
commitaf0eca1a80e3e3a6fa0ee00df4a7883e4f228d0b
treedce22f3bc875381106a990100d50d72b5a026335
parent34e2ef661b47140594feb4718893b42fc6fc1771
Clean up ill-advised attempt to invent a private set of Node tags.

Somebody thought it'd be cute to invent a set of Node tag numbers that were
defined independently of, and indeed conflicting with, the main tag-number
list.  While this accidentally failed to fail so far, it would certainly
lead to trouble as soon as anyone wanted to, say, apply copyObject to these
node types.  Clang was already complaining about the use of makeNode on
these tags, and I think quite rightly so.  Fix by pushing these node
definitions into the mainstream, including putting replnodes.h where it
belongs.
src/backend/replication/repl_gram.y
src/backend/replication/walsender.c
src/include/nodes/nodes.h
src/include/nodes/replnodes.h [moved from src/include/replication/replnodes.h with 80% similarity]
src/include/replication/basebackup.h