]> granicus.if.org Git - postgresql/commit
Be more tense about not creating tuplestores with randomAccess = true unless
authorTom Lane <tgl@sss.pgh.pa.us>
Wed, 29 Oct 2008 00:00:39 +0000 (00:00 +0000)
committerTom Lane <tgl@sss.pgh.pa.us>
Wed, 29 Oct 2008 00:00:39 +0000 (00:00 +0000)
commit05bba3d176e0adc1a032d5c8c6ea2a7622c7dd0d
tree59140f72d76bd4fb7fc7e8bbad6ccf57084f042a
parente3e3d2a789e34ff6572bdf693beb1516a228c5ff
Be more tense about not creating tuplestores with randomAccess = true unless
backwards scan could actually happen.  In particular, pass a flag to
materialize-mode SRFs that tells them whether they need to require random
access.  In passing, also suppress unneeded backward-scan overhead for a
Portal's holdStore tuplestore.  Per my proposal about reducing I/O costs for
tuplestores.
contrib/tablefunc/tablefunc.c
contrib/xml2/xpath.c
src/backend/commands/prepare.c
src/backend/executor/execQual.c
src/backend/executor/nodeFunctionscan.c
src/backend/utils/fmgr/README
src/backend/utils/mmgr/portalmem.c
src/include/executor/executor.h
src/include/nodes/execnodes.h
src/pl/plperl/plperl.c
src/pl/plpgsql/src/pl_exec.c