Adding another control test for Chrome Endure that uses WebDriver.

Adding another control test for Chrome Endure that uses WebDriver.

The new control test, testControlAttachDetachDOMTreeWebDriver, behaves the
same as the existing control test testControlAttachDetachDOMTree, except it
uses webdriver to click on buttons to attach/detach a DOM tree, rather than
relying on Javascript setTimeout calls to drive the attaching/detaching.

BUG=chromium-os:27590
TEST=Verified the test runs successfully on my local linux machine.

Committed: https://src.chromium.org/viewvc/chrome?view=rev&revision=126190

[frankf1, 2012-03-10 02:57:17]

What's the advantage of using webdriver here?

http://codereview.chromium.org/9687001/diff/1/chrome/test/data/chrome_endure/...
File chrome/test/data/chrome_endure/endurance_control_webdriver.html (right):

http://codereview.chromium.org/9687001/diff/1/chrome/test/data/chrome_endure/...
chrome/test/data/chrome_endure/endurance_control_webdriver.html:31:
document.body.removeChild(document.body.childNodes[5]);
can you query the node by name instead of the position it happens to be?

http://codereview.chromium.org/9687001/diff/1/chrome/test/functional/perf_end...
File chrome/test/functional/perf_endure.py (right):

http://codereview.chromium.org/9687001/diff/1/chrome/test/functional/perf_end...
chrome/test/functional/perf_endure.py:291: def
testControlAttachDetachDOMTreeWebDriver(self):
There are a lot of duplication between this and previous function. Please factor
this out.

http://codereview.chromium.org/9687001/diff/1/chrome/test/functional/perf_end...
chrome/test/functional/perf_endure.py:326: time.time() - self._test_start_time)
alignment issue

[dennis_jeffrey, 2012-03-12 18:48:06]

Thanks!

The new control test added in this CL was explicitly designed to use WebDriver,
in an effort to verify that the use of WebDriver itself doesn't adversely affect
the memory bloat measurements.  It's essentially a control test for WebDriver.

http://codereview.chromium.org/9687001/diff/1/chrome/test/data/chrome_endure/...
File chrome/test/data/chrome_endure/endurance_control_webdriver.html (right):

http://codereview.chromium.org/9687001/diff/1/chrome/test/data/chrome_endure/...
chrome/test/data/chrome_endure/endurance_control_webdriver.html:31:
document.body.removeChild(document.body.childNodes[5]);
On 2012/03/10 02:57:17, frankf1 wrote:
> can you query the node by name instead of the position it happens to be?

Done - I gave the root node an ID.

http://codereview.chromium.org/9687001/diff/1/chrome/test/functional/perf_end...
File chrome/test/functional/perf_endure.py (right):

http://codereview.chromium.org/9687001/diff/1/chrome/test/functional/perf_end...
chrome/test/functional/perf_endure.py:291: def
testControlAttachDetachDOMTreeWebDriver(self):
On 2012/03/10 02:57:17, frankf1 wrote:
> There are a lot of duplication between this and previous function. Please
factor
> this out.

Good point.  Done.  Also added a TODO to make the same change for the other
Chrome Endure tests.

http://codereview.chromium.org/9687001/diff/1/chrome/test/functional/perf_end...
chrome/test/functional/perf_endure.py:326: time.time() - self._test_start_time)
On 2012/03/10 02:57:17, frankf1 wrote:
> alignment issue

I indented this line 4 spaces underneath the right-hand side of the previous
line, since it's a continuation of that expression.  Which indentation do you
recommend?  Would you prefer that I indent with respect to the start of the
previous line instead?  Or line it up underneath the start of the right-hand
side in the previous line?

[frankf1, 2012-03-12 19:10:09]

lgtm

[dennis_jeffrey, 2012-03-12 19:11:32]

Please take one more quick look; I adjusted the indentation as per our offline
discussion.

[dennis_jeffrey, 2012-03-12 19:21:36]

Thanks for the offline LGTM of the latest patch set.  I'll submit this now.