TABLESAMPLE corrupts memory #16258
Description
What happens?
When I do this...
CREATE OR REPLACE TABLE blah as (
SELECT *
FROM range(10_000_000)
);
SELECT * FROM blah TABLESAMPLE 100 ROWS;
SELECT * FROM blah TABLESAMPLE 100 ROWS;
SELECT * FROM blah TABLESAMPLE 100 ROWS;
SELECT * FROM blah TABLESAMPLE 100 ROWS;
SELECT * FROM blah TABLESAMPLE 100 ROWS;
SELECT * FROM blah TABLESAMPLE 100 ROWS;
SELECT * FROM blah TABLESAMPLE 100 ROWS;
SELECT * FROM blah TABLESAMPLE 100 ROWS;
SELECT * FROM blah TABLESAMPLE 100 ROWS;
SELECT * FROM blah TABLESAMPLE 100 ROWS;
One of those selects will throw, usually the second.
duckdb(24789,0x1fed14240) malloc: Incorrect checksum for freed object 0x152085800: probably modified after being freed.
Corrupt value: 0x2f2525
duckdb(24789,0x1fed14240) malloc: *** set a breakpoint in malloc_error_break to debug
To Reproduce
CREATE OR REPLACE TABLE blah as (
SELECT *
FROM range(10_000_000)
);
SELECT * FROM blah TABLESAMPLE 100 ROWS;
SELECT * FROM blah TABLESAMPLE 100 ROWS;
SELECT * FROM blah TABLESAMPLE 100 ROWS;
SELECT * FROM blah TABLESAMPLE 100 ROWS;
SELECT * FROM blah TABLESAMPLE 100 ROWS;
SELECT * FROM blah TABLESAMPLE 100 ROWS;
SELECT * FROM blah TABLESAMPLE 100 ROWS;
SELECT * FROM blah TABLESAMPLE 100 ROWS;
SELECT * FROM blah TABLESAMPLE 100 ROWS;
SELECT * FROM blah TABLESAMPLE 100 ROWS;
OS:
macOS Sequoia 15.2
DuckDB Version:
v1.2.0 5f5512b
DuckDB Client:
cli
Hardware:
2021 Macbook Pro, M1 Max 64GB
Full Name:
Jason Punyon
Affiliation:
Sumble
What is the latest build you tested with? If possible, we recommend testing with the latest nightly build.
Local release build at 28c95be
Did you include all relevant data sets for reproducing the issue?
Yes
Did you include all code required to reproduce the issue?
- Yes, I have
Did you include all relevant configuration (e.g., CPU architecture, Python version, Linux distribution) to reproduce the issue?
- Yes, I have