srowen commented on a change in pull request #25186: [SPARK-28433][SQL]Special
scala test case for aarch64
URL: https://github.com/apache/spark/pull/25186#discussion_r304986924
##########
File path:
sql/core/src/test/scala/org/apache/spark/sql/DataFrameAggregateSuite.scala
##########
@@ -728,9 +728,15 @@ class DataFrameAggregateSuite extends QueryTest with
SharedSQLContext {
import java.lang.Float.floatToRawIntBits
import java.lang.Double.doubleToRawLongBits
- // 0.0/0.0 and NaN are different values.
- assert(floatToRawIntBits(0.0f/0.0f) != floatToRawIntBits(Float.NaN))
Review comment:
It sounds like this is the case, but, if several NaN representations that
are _different_ on x86 are the _same_ on ARM then it should still work OK. Let
me run tests to figure it out.
So, I guess the question is, do we test whatever would break if Spark
doesn't handle this? if so then we don't necessarily have to make this specific
assertion is true, because it isn't true in all cases according to the JDK.
----------------------------------------------------------------
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.
For queries about this service, please contact Infrastructure at:
[email protected]
With regards,
Apache Git Services
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
