Hi guys,
I'm debugging a kernel that has several double variables, but which are displayed by the debugger as single precision.
The debugger correctly identifies the variables' type as double, but when inspecting them the values displayed are shorter than expected.
For example:
I know the variable holds the value 0.27096070445647297, but CodeXL only shows 0.2709607.
This is a problem for me, since I'm using double because I do need the extra precision and when debugging I therefore need to check those smaller digits.
Is there a configuration option or some other way I could inspect the full double value in CodeXL?
Thanks.
