Allow XHR timeout attribute to be overridden after send(), per spec

Source/core/loader/DocumentThreadableLoader.cpp

Index: Source/core/loader/DocumentThreadableLoader.cpp
diff --git a/Source/core/loader/DocumentThreadableLoader.cpp b/Source/core/loader/DocumentThreadableLoader.cpp
index 2840dce77f4d9e26c4eae1e3c59f20864f715df6..dbcf5005fb8038c77e6a5a1e9f683bf86cc66b53 100644
--- a/Source/core/loader/DocumentThreadableLoader.cpp
+++ b/Source/core/loader/DocumentThreadableLoader.cpp
@@ -76,6 +76,7 @@ DocumentThreadableLoader::DocumentThreadableLoader(Document& document, Threadabl
     , m_simpleRequest(true)
     , m_async(blockingBehavior == LoadAsynchronously)
     , m_timeoutTimer(this, &DocumentThreadableLoader::didTimeout)
+    , m_requestStarted(0.0)
 {
     ASSERT(client);
     // Setting an outgoing referer is only supported in the async code path.
@@ -138,6 +139,24 @@ DocumentThreadableLoader::~DocumentThreadableLoader()
 {
 }
 
+void DocumentThreadableLoader::overrideTimeout(unsigned long timeoutMilliseconds)
+{
+    if (m_async) {

[tyoshino (SeeGerritForStatus), 2014/05/09 08:04:36]

if (!m_async)
    return;

...

+        m_timeoutTimer.stop();
+        // At the time of this method's implementation, it is only ever called by
+        // XMLHttpRequest, when the timeout attribute is set after sending the
+        // request.
+        //
+        // The XHR request says to resolve the time relative to when the request
+        // was initially sent, however other uses of this method may need to
+        // behave differently, in which case this should be re-arranged somehow.
+        if (timeoutMilliseconds) {
+            double resolvedTime = m_requestStarted + timeoutMilliseconds / 1000.0;
+            m_timeoutTimer.startOneShotExact(resolvedTime, FROM_HERE);

[tyoshino (SeeGerritForStatus), 2014/05/09 08:04:36]

any reason you chose to add this new method to Timer than calculating the time
from now until the new timeout?

[caitp (gmail), 2014/05/09 08:39:49]

For IEEE floating point math, you lose accuracy if you mix small additions with
large ones in a weird order, so adding the extra subtract is potentially an
issue.

But you're right, it's better to have less code, and it's probably not going to
introduce major flakiness, so I'll remove the Timer methods

+        }
+    }
+}
+
 void DocumentThreadableLoader::cancel()
 {
     cancelWithError(ResourceError());
@@ -387,6 +406,8 @@ void DocumentThreadableLoader::loadRequest(const ResourceRequest& request)
     ASSERT(m_sameOriginRequest || requestURL.user().isEmpty());
     ASSERT(m_sameOriginRequest || requestURL.pass().isEmpty());
 
+    m_requestStarted = monotonicallyIncreasingTime();

[tyoshino (SeeGerritForStatus), 2014/05/09 08:04:36]

move this to between L419 and L420 (outside of if-clause)?

+
     ThreadableLoaderOptions options = m_options;
     if (m_async) {
         if (m_actualRequest) {