obsolete: fix n^2 marker computation behavior

Previously, if you ran obsolete.createmarkers with a bunch of markers that did
not have successors (like when you do a prune), it encountered a n^2 computation
behavior because the loop would read the changelog (to get ctx.parents()), then
add a marker, in a loop.  Adding a marker invalidated the computehidden cache,
and reading the changelog recomputed it.

This resulted in pruning 150 commits taking 150+ seconds in a large repo.

The fix is to break the reading part of the loop to be separate from the writing
part.

mercurial/obsolete.py

@@ -1225,6 +1225,7 b' def createmarkers(repo, relations, flag='
         metadata['user'] = repo.ui.username()
     tr = repo.transaction('add-obsolescence-marker')
     try:
+        markerargs = []
         for rel in relations:
             prec = rel[0]
             sucs = rel[1]
@@ -1243,6 +1244,15 b' def createmarkers(repo, relations, flag='
                 npare = tuple(p.node() for p in prec.parents())
             if nprec in nsucs:
                 raise error.Abort("changeset %s cannot obsolete itself" % prec)
+
+            # Creating the marker causes the hidden cache to become invalid,
+            # which causes recomputation when we ask for prec.parents() above.
+            # Resulting in n^2 behavior.  So let's prepare all of the args
+            # first, then create the markers.
+            markerargs.append((nprec, nsucs, npare, localmetadata))
+
+        for args in markerargs:
+            nprec, nsucs, npare, localmetadata = args
             repo.obsstore.create(tr, nprec, nsucs, flag, parents=npare,
                                  date=date, metadata=localmetadata)
             repo.filteredrevcache.clear()