run-tests: don't mask errors when a server fails to start

There are sporadic instances of this on Windows.  They seem to happen more
frequently after the test machine is rebooted, although the only way to hit it
on my laptop is to loop certain tests with -j9 for hours.  The problem with
masking out the specific failure is that there's no way to know if it's the same
line in the test that's failing, or if it is random.

The justification for adding this masking in 52e9e63f1495 was that the failures
occur regularly, but that's not the case anymore.  The port number is still
printed, in case that turns out to be useful.

tests/run-tests.py

@@ -1763,9 +1763,9 b' class TestResult(unittest._TextTestResul'
                 servefail, lines = getdiff(expected, got,
                                            test.refpath, test.errpath)
                 if servefail:
-                    raise test.failureException(
+                    self.stream.write(
                         'server failed to start (HGPORT=%s)' % test._startport)
-                else:
+
                     self.stream.write('\n')
                     for line in lines:
                         line = highlightdiff(line, self.color)

tests/test-run-tests.t

@@ -541,10 +541,16 b' Verify that when a process fails to star'
   >   $ echo 'abort: child process failed to start blah'
   > EOF
   $ rt test-serve-fail.t
+  server failed to start (HGPORT=*) (glob)
+  --- $TESTTMP/test-serve-fail.t
+  +++ $TESTTMP/test-serve-fail.t.err
+  @@ -1* +1,2 @@ (glob)
+     $ echo 'abort: child process failed to start blah'
+  +  abort: child process failed to start blah
   
   ERROR: test-serve-fail.t output changed
   !
-  Failed test-serve-fail.t: server failed to start (HGPORT=*) (glob)
+  Failed test-serve-fail.t: output changed
   # Ran 1 tests, 0 skipped, 1 failed.
   python hash seed: * (glob)
   [1]