Xtpacz commented on PR #12261:
URL: https://github.com/apache/gluten/pull/12261#issuecomment-4676907218
@zhouyuan Thanks for you review. It is confirmed that this issue only exists
in spark-33, as you mentioned. Here is the fallback log from driver on spark-33:
```
INFO org.apache.spark.sql.execution.GlutenFallbackReporter: Validation
failed for plan: Project[QueryId=4], due to:
- Native validation failed:
|- Validation failed due to exception caught at
file:SubstraitToVeloxPlanValidator.cc line:1450 function:validate, thrown from
file:ExprCompiler.cpp line:311 function:compileCall, reason:Found incompatible
return types for 'divide' (DECIMAL(38, 9) vs. DECIMAL(38, 10)) for input types
(DECIMAL(29, 6), DECIMAL(27, 6)).
```
This confirms that CheckOverflowTransformer use original.child.dataType
(which is left.dataType in spark-33's BinaryArithmetic), so causing the
precision mismatch.
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
