On Fri, 20 Feb 2026 20:02:06 GMT, Roger Riggs <[email protected]> wrote:

> Please describe the semantic impact of the violated assertion. Would it ever 
> be observable?

If concurrent map use red-black-tree correctly(or use it normally as common 
red-balck-tree algorithm), this bug will never be observable.
In the logic, It just like the old code assert 1 is a Object, but I fix it to 1 
is a Integer. More correctly( if 1 is a String it will throw error) , but the 
old code is not a "bug" , will never throw error(if the tree is build and use 
correct). I find it by code review.

This assert just like a ut lack some of branch,a ut not perfect, if the 
red-black-tree operation code is wrong and some case  not as expect happen and 
it will not aware,but if the code run correctly, maybe it not need to fix, it 
decide by your team. 

@RogerRiggs

-------------

PR Comment: https://git.openjdk.org/jdk/pull/24612#issuecomment-3936977232

Reply via email to