Don't think so. Could you provide an example?

For "any encoding other than US-ASCII and iso-8859-1" to matter while decoding, you'd need wide characters, in which case Perl will croak with "Wide character in subroutine entry at .../Encode.pm line..." before returning anything. Any non-wide characters will simply be treated as iso-8859-1, because decode('iso-8859-1',...) is telling Perl they are.

Here's a test that works with single-byte encodings:

>perl -MEncode -wle"for (0..255) { print if chr ne decode($ARGV[0], ch
+r) }" iso-8859-1

>perl -MEncode -wle"for (0x20..0x7E,0x0A..0xFF) { print if chr ne deco
+de($ARGV[0], chr) }" iso-8859-2
161
162
163
165
166
169
...
[download]

Here's a UTF-8 example:

>perl -MEncode -wle"$chr=qq{\342\231\240}; print 'diff' if $chr ne dec
+ode('UTF-8', $chr)"
diff
[download]

Yes, but that's not what the question was. It was asked whether there is any value of $x for which test($x) with the given code (i.e. decode('iso-8859-1',...)) returns 0.  By modifying the code you can of course achieve anything...  But this is kind of like

"Is there any case in which this code will return 0?"

sub test {
    if (1) {
        return 1;
    } else {
        retrun 0;
    }
}
[download]

Yes, there is:

sub test {
    if (0) {
        return 1;
    } else {
        retrun 0;
    }
}
[download]