tracing: ignore any IOErrors when writing to pipe

When the pager forks off the main process, we can end up with the pipe
closed prematurely. Rather than break hg entirely when that happens
and tracing is active, just let lingering events disappear as needed.

Differential Revision: https://phab.mercurial-scm.org/D4445

hgdemandimport/tracing.py

@@ -28,7 +28,17 b' def log(whencefmt, *whenceargs):'
         _session = os.environ.get('HGCATAPULTSESSION', 'none')
     whence = whencefmt % whenceargs
     try:
-        _pipe.write('START %s %s\n' % (_session, whence))
+        # Both writes to the pipe are wrapped in try/except to ignore
+        # errors, as we can see mysterious errors in here if the pager
+        # is active. Presumably other conditions could trigger
+        # problems too.
+        try:
+            _pipe.write('START %s %s\n' % (_session, whence))
+        except IOError:
+            pass
         yield
     finally:
-        _pipe.write('END %s %s\n' % (_session, whence))
+        try:
+            _pipe.write('END %s %s\n' % (_session, whence))
+        except IOError:
+            pass