Bug when join 2 table with [BUG]

[nic9lif3]

When I process this sql:

SELECT 4 TIME_ID
	,STOCK_ID
	,MAX(CASE WHEN TIME_ID BETWEEN 4 - 1597 + 1 AND 4 - 1 + 1 THEN VAR12 ELSE null END) MAX_VAR12_1M_1597M
FROM (
	SELECT *
	FROM VAR1_TMP_BOOK_INFO
	WHERE TIME_ID BETWEEN 4 - 17711 + 1
			AND 4
	) VAR1_TMP_BOOK_INFO
JOIN (
	SELECT stock_id stock_id1
		,time_id time_id1
		,seconds_in_bucket seconds_in_bucket1
		,LAG(VAR1, 1) OVER (
			PARTITION BY stock_id
			,time_id ORDER BY seconds_in_bucket
			) VAR1_LAG
	FROM VAR1_TMP_BOOK_INFO
	WHERE TIME_ID BETWEEN 4 - 1
			AND 4
	) VAR1_TMP_BOOK_INFO_1 ON VAR1_TMP_BOOK_INFO.stock_id = VAR1_TMP_BOOK_INFO_1.stock_id1
	AND VAR1_TMP_BOOK_INFO.time_id = VAR1_TMP_BOOK_INFO_1.time_id1
	AND VAR1_TMP_BOOK_INFO.seconds_in_bucket = VAR1_TMP_BOOK_INFO_1.seconds_in_bucket1
GROUP BY STOCK_ID

then it raises error

RunExecuteGraphError: [RunExecuteGraph Error] Ral failure at: /opt/conda/envs/rapids/conda-bld/blazingsql_1623363481746/work/engine/src/Interpreter/interpreter_cpp.cu:330: Operations between literals is not supported

If I remove condition in agg, it works:

SELECT 4 TIME_ID
	,STOCK_ID
	,MAX(CASE WHEN 1=1 THEN VAR12 ELSE null END) MAX_VAR12_1M_1597M
FROM (
	SELECT *
	FROM VAR1_TMP_BOOK_INFO
	WHERE TIME_ID BETWEEN 4 - 17711 + 1
			AND 4
	) VAR1_TMP_BOOK_INFO
JOIN (
	SELECT stock_id stock_id1
		,time_id time_id1
		,seconds_in_bucket seconds_in_bucket1
		,LAG(VAR1, 1) OVER (
			PARTITION BY stock_id
			,time_id ORDER BY seconds_in_bucket
			) VAR1_LAG
	FROM VAR1_TMP_BOOK_INFO
	WHERE TIME_ID BETWEEN 4 - 1
			AND 4
	) VAR1_TMP_BOOK_INFO_1 ON VAR1_TMP_BOOK_INFO.stock_id = VAR1_TMP_BOOK_INFO_1.stock_id1
	AND VAR1_TMP_BOOK_INFO.time_id = VAR1_TMP_BOOK_INFO_1.time_id1
	AND VAR1_TMP_BOOK_INFO.seconds_in_bucket = VAR1_TMP_BOOK_INFO_1.seconds_in_bucket1
GROUP BY STOCK_ID

or if I remove join table it also works:

SELECT 4 TIME_ID
	,STOCK_ID
	,MAX(CASE WHEN TIME_ID BETWEEN 4 - 1597 + 1 AND 4 - 1 + 1 THEN VAR12 ELSE null END) MAX_VAR12_1M_1597M
FROM (
	SELECT *
	FROM VAR1_TMP_BOOK_INFO
	WHERE TIME_ID BETWEEN 4 - 17711 + 1
			AND 4
	) VAR1_TMP_BOOK_INFO
GROUP BY STOCK_ID

The source table likes that:

	TIME_ID	STOCK_ID	SECONDS_IN_BUCKET	VAR1	VAR12
0	4	0	1	0.000306751	-0.590466
1	4	0	5	0.00060095	-10.1625
2	4	0	0	0.000306751	-0.539158