diff options
| author | Robert Haas <rhaas@postgresql.org> | 2015-01-23 11:58:31 -0500 |
|---|---|---|
| committer | Robert Haas <rhaas@postgresql.org> | 2015-01-23 11:58:31 -0500 |
| commit | 5cefbf5a6c4466ac6b1cc2a4316b4eba9108c802 (patch) | |
| tree | fb074b1dfe7beffcadf890a3d7264a9043fe53c6 /src/backend/utils | |
| parent | 6a3c6ba0ba2bf86254d43b40569cdf3012cf12aa (diff) | |
| download | postgresql-5cefbf5a6c4466ac6b1cc2a4316b4eba9108c802.tar.gz postgresql-5cefbf5a6c4466ac6b1cc2a4316b4eba9108c802.zip | |
Don't use abbreviated keys for the final merge pass.
When we write tuples out to disk and read them back in, the abbreviated
keys become non-abbreviated, because the readtup routines don't know
anything about abbreviation. But without this fix, the rest of the
code still thinks the abbreviation-aware compartor should be used,
so chaos ensues.
Report by Andrew Gierth; patch by Peter Geoghegan.
Diffstat (limited to 'src/backend/utils')
| -rw-r--r-- | src/backend/utils/sort/tuplesort.c | 16 |
1 files changed, 16 insertions, 0 deletions
diff --git a/src/backend/utils/sort/tuplesort.c b/src/backend/utils/sort/tuplesort.c index 6d3aa889bc5..b6e302fc9ca 100644 --- a/src/backend/utils/sort/tuplesort.c +++ b/src/backend/utils/sort/tuplesort.c @@ -2172,6 +2172,22 @@ mergeruns(Tuplesortstate *state) return; } + if (state->sortKeys->abbrev_converter) + { + /* + * If there are multiple runs to be merged, when we go to read back + * tuples from disk, abbreviated keys will not have been stored, and we + * don't care to regenerate them. Disable abbreviation from this point + * on. + */ + state->sortKeys->abbrev_converter = NULL; + state->sortKeys->comparator = state->sortKeys->abbrev_full_comparator; + + /* Not strictly necessary, but be tidy */ + state->sortKeys->abbrev_abort = NULL; + state->sortKeys->abbrev_full_comparator = NULL; + } + /* End of step D2: rewind all output tapes to prepare for merging */ for (tapenum = 0; tapenum < state->tapeRange; tapenum++) LogicalTapeRewind(state->tapeset, tapenum, false); |