Attempt to fix floating point test failure

The previous behavior depending on exact float values, it seemed
to work OK on amd64 and i386, but other CI platforms are
experiencing non-deterministic test failures with it. Relax test
slightly to hopefully pass on such platforms.

test/ruby/test_float.rb

@@ -898,7 +898,7 @@ class TestFloat < Test::Unit::TestCase
       a = rand
       b = a+rand*1000
       s = (b - a) / 10
-      b = a + s*10
+      b = a + s*9.999999
       seq = (a...b).step(s)
       assert_equal(10, seq.to_a.length, seq.inspect)
     end