test: fix flaky test-regress-nodejsGH-897

Even after being moved to `sequential` in
1ce05ad, `test-regress-nodejsGH-897` still
was occasionally flaky on Raspberry Pi devices on CI.

The test is especially sensitive to resource constraints. It failed
reliably on my laptop if I moved it to `parallel` and ran 32 competing
node test processes. Even for a flaky test, that's unusually low. I
typically don't see problems, even for flaky tests, until I get up to
around four times that number.

On a Raspberry Pi, of course, that sensitivity to resource constraints
will manifest much sooner.

This change checks the order of timers firing, rather than the duration
before a timer is fired. This eliminates the sensitivity to resource
constraints. The test can now be moved back to `parallel`. I am able to
run many copies of the test simultaneously without seeing test failures.

Fixes: nodejs#10073

test/parallel/test-regress-GH-897.js

@@ -0,0 +1,20 @@
+'use strict';
+
+// Test for bug where a timer duration greater than 0 ms but less than 1 ms
+// resulted in the duration being set for 1000 ms. The expected behavior is
+// that the timeout would be set for 1 ms, and thus fire before timers set
+// with values greater than 1ms.
+//
+// Ref: https://github.com/nodejs/node-v0.x-archive/pull/897
+
+const common = require('../common');
+
+let timer;
+
+setTimeout(function() {
+  clearTimeout(timer);
+}, 0.1); // 0.1 should be treated the same as 1, not 1000...
+
+timer = setTimeout(function() {
+  common.fail('timers fired out of order');
+}, 2); // ...so this timer should fire second.