@dawkot To put it simply, what Araq says is: in the first case, the macro
operates directly on procA at compile time so it behaves as expected but in the
second case, it actually operates on p argument (which has no implementation as
it's not an actual procedure, therefore its implementation is nil) and then you
provide procA as a value of p argument at runtime. But what was done at compile
time remains the same so the runtime argument passing is irrelevant.
What you need is a template, as templates' arguments are passed at compile time
rather than runtime:
import macros
macro procImplRepr(p: proc): string =
p.symbol.getImpl.repr
proc procA = discard
template procB(p: proc): string = p.procImplRepr
echo procA.procImplRepr
echo procB(procA)
# both print the same value
