Print stack traces in child processes when browser tests failed.

base/process/launch_win.cc

Index: base/process/launch_win.cc
diff --git a/base/process/launch_win.cc b/base/process/launch_win.cc
index fa59f1ae90744469e397d3bd778cb8b1e98150c4..b91320364d6b8e248d3e8f03a242d65ccef0b77b 100644
--- a/base/process/launch_win.cc
+++ b/base/process/launch_win.cc
@@ -64,8 +64,22 @@ void RouteStdioToConsole() {
   // stdout/stderr on startup (before the handle IDs can be reused).
   // _fileno(stdout) will return -2 (_NO_CONSOLE_FILENO) if stdout was
   // invalid.
-  if (_fileno(stdout) >= 0 || _fileno(stderr) >= 0)
-    return;
+  if (_fileno(stdout) >= 0 || _fileno(stderr) >= 0) {
+    // _fileno was broken for SUBSYSTEM:WINDOWS from VS2010 to VS2012/2013.
+    // http://crbug.com/358267. Confirm that the underlying HANDLE is valid
+    // before aborting.
+
+    // This causes NaCl tests to hang on XP for reasons unclear, perhaps due

[scottmg, 2015/08/12 02:28:14]

It would be better if you can remove this early return, otherwise won't get
stacks on XP. But I'm not sure if there will be problems with some NaCl modules
then.

[jam, 2015/08/12 03:01:32]

I didn't want to deal with debugging nacl tests on xp so I left it as it was in
your cl. we have one xp bot vs many on waterfall and on trybots, so i figured
the most bang for the buck is doing those first. if we find the one xp bot has
renderer crashes then we can spend time on this.

+    // to not being able to inherit handles. Since it's only for debugging,
+    // and redirecting still works, punt for now.
+    if (base::win::GetVersion() < base::win::VERSION_VISTA)
+      return;
+
+    intptr_t stdout_handle = _get_osfhandle(_fileno(stdout));
+    intptr_t stderr_handle = _get_osfhandle(_fileno(stderr));
+    if (stdout_handle >= 0 || stderr_handle >= 0)
+      return;
+  }
 
   if (!AttachConsole(ATTACH_PARENT_PROCESS)) {
     unsigned int result = GetLastError();