diff options
| author | Heikki Linnakangas <heikki.linnakangas@iki.fi> | 2014-05-13 14:16:28 +0300 |
|---|---|---|
| committer | Heikki Linnakangas <heikki.linnakangas@iki.fi> | 2014-05-13 15:27:14 +0300 |
| commit | d5b912c905efd531a6228d08c98cd92d49cdd94c (patch) | |
| tree | 5e90a778d6ac6f0bdcf1094e3de7a26ab18c6e6e | |
| parent | b65a258a512049eba567283d93bd63b22b77916a (diff) | |
| download | postgresql-d5b912c905efd531a6228d08c98cd92d49cdd94c.tar.gz postgresql-d5b912c905efd531a6228d08c98cd92d49cdd94c.zip | |
Initialize padding bytes in btree_gist varbit support.
The code expands a varbit gist leaf key to a node key by copying the bit
data twice in a varlen datum, as both the lower and upper key. The lower key
was expanded to INTALIGN size, but the padding bytes were not initialized.
That's a problem because when the lower/upper keys are compared, the padding
bytes are used compared too, when the values are otherwise equal. That could
lead to incorrect query results.
REINDEX is advised for any btree_gist indexes on bit or bit varying data
type, to fix any garbage padding bytes on disk.
Per Valgrind, reported by Andres Freund. Backpatch to all supported
versions.
| -rw-r--r-- | contrib/btree_gist/btree_bit.c | 12 |
1 files changed, 8 insertions, 4 deletions
diff --git a/contrib/btree_gist/btree_bit.c b/contrib/btree_gist/btree_bit.c index d94abcb3cf5..94eaf6df3e8 100644 --- a/contrib/btree_gist/btree_bit.c +++ b/contrib/btree_gist/btree_bit.c @@ -83,10 +83,14 @@ static bytea * gbt_bit_xfrm(bytea *leaf) { bytea *out = leaf; - int s = INTALIGN(VARBITBYTES(leaf) + VARHDRSZ); - - out = palloc(s); - SET_VARSIZE(out, s); + int sz = VARBITBYTES(leaf) + VARHDRSZ; + int padded_sz = INTALIGN(sz); + + out = (bytea *) palloc(padded_sz); + /* initialize the padding bytes to zero */ + while (sz < padded_sz) + ((char *) out)[sz++] = 0; + SET_VARSIZE(out, padded_sz); memcpy((void *) VARDATA(out), (void *) VARBITS(leaf), VARBITBYTES(leaf)); return out; } |