pgsql: Teach CLUSTER to use seqscan-and-sort when it's faster than inde

Teach CLUSTER to use seqscan-and-sort when it's faster than indexscan.

... or at least, when the planner's cost estimates say it will be faster.

Leonardo Francalanci, reviewed by Itagaki Takahiro and Tom Lane

Branch
------
master

Details
-------
http://git.postgresql.org/gitweb?p=postgresql.git;a=commitdiff;h=3ba11d3df2115b04171a8eda8e0389e02578d8d0

Modified Files
--------------
doc/src/sgml/ref/cluster.sgml | 67 +++----
src/backend/commands/cluster.c | 161 ++++++++++++----
src/backend/optimizer/path/costsize.c | 45 +++--
src/backend/optimizer/plan/createplan.c | 3 +
src/backend/optimizer/plan/planmain.c | 3 +-
src/backend/optimizer/plan/planner.c | 123 ++++++++++++-
src/backend/optimizer/prep/prepunion.c | 3 +-
src/backend/optimizer/util/pathnode.c | 2 +
src/backend/optimizer/util/plancat.c | 86 +++++++--
src/backend/utils/sort/tuplesort.c | 310 +++++++++++++++++++++++++++++++
src/include/optimizer/cost.h | 1 +
src/include/optimizer/plancat.h | 2 +
src/include/optimizer/planner.h | 2 +
src/include/utils/tuplesort.h | 43 +++--
14 files changed, 713 insertions(+), 138 deletions(-)