It was taking a reference. This would cause the data to be corrupted by
future reads.

fixes #1909

#1911

#1915

#1847

Check if CanConvert before calling reflect.Value.Convert

Otherwise, it might be possible to panic when closing the pipeline if it
tries to read a connection that should be closed but still has a fatal
error on the wire.

#1920

The underlying type of json.RawMessage is a []byte so to avoid it being
considered binary data we need to handle it specifically. This is done
by registerDefaultPgTypeVariants. In addition, handle json.RawMessage in
the JSONCodec PlanEncode to avoid it being mutated by json.Marshal.

#1763

#1924

The PostgreSQL server will reject messages greater than ~1 GB anyway.
However, worse than that is that a message that is larger than 4 GB
could wrap the 32-bit integer message size and be interpreted by the
server as multiple messages. This could allow a malicious client to
inject arbitrary protocol messages.

GHSA-mrww-27vc-gghv

pgx v5 was not vulnerable to CVE-2024-27289 do to how the sanitizer was
being called. But the sanitizer itself still had the underlying issue.
This commit ports the fix from pgx v4 to v5 to ensure that the issue
does not emerge if pgx uses the sanitizer differently in the future.