conv-float-u32 has never taken the absolute value of its input.

Through and including version 2.1.7, there *was* a bug in conv-float-u32
which caused a negative value (e.g., -1) to output as a very large value
(e.g., 4294967295 or 0xFFFFFFFF) when 'clamp' is TRUE.  In 2.2.0 and
newer, negative values will be correctly clamped to 0 in the output.
As far as I can tell, the behavior when 'clamp' is FALSE (the default)
hasn't ever changed.

This change was not specifically called out in the release notes for
version 2.2.0.

If you have specific examples of input and output values (along with the
value of the 'clamp' parameter) that have changed between emc versions
(please note which versions!), I'd be happy to tell you if it's a result
of this bugfix, or a result of some other difference I'm not aware of.

Jeff

-------------------------------------------------------------------------
This SF.net email is sponsored by the 2008 JavaOne(SM) Conference 
Don't miss this year's exciting event. There's still time to save $100. 
Use priority code J8TL2D2. 
http://ad.doubleclick.net/clk;198757673;13503038;p?http://java.sun.com/javaone
_______________________________________________
Emc-users mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/emc-users

Reply via email to