Fix scale on axis may be negative.

Fix scale on axis may be negative.

This patch modifies judgment condition whether is equal to zero for double type, avoid unnecessary sqrt calls in more case.
If "a" and "b" are both equal to zero, and "c" is less then zero, currently ScaleOnAxis will return "c" which is negative. So, we should ensure that the result is greater than zero.

BUG=none.

Committed: https://src.chromium.org/viewvc/chrome?view=rev&revision=267078

[Yupei Wang, 2014-04-29 02:13:09]

Hi Aelias, I'm very sorry, yesterday I sent a codereview using an invalid
account. Now I send it again.

[Yupei Wang, 2014-04-29 03:02:33]

On 2014/04/29 02:13:09, Yupei Wang wrote:
> Hi Aelias, I'm very sorry, yesterday I sent a codereview using an invalid
> account. Now I send it again.

tomhudson	
16 hours, 39 minutes ago #37
On 2014/04/28 09:33:27, Yupei Wang wrote:
> > The failures appear unrelated to my change.

These messages are relevant:

** Presubmit Messages **
If this change has an associated bug, add BUG=[bug number].
---Reply---
This change has not an associated bug I was observed, it just something 
I noticed on code inspection. So, BUG=none.

** Presubmit Warnings **
perryuwang@gmail.com is not in AUTHORS file. If you are a new contributor,
please visit
http://www.chromium.org/developers/contributing-code and read the "Legal"
section
If you are a chromite, verify the contributor signed the CLA.

(1) I'm not convinced that ComputeTransform2dScaleComponents() should be
returning positive values if the actual scale is negative; the name isn't
sufficient documentation for us to deduce that. Why should we consider negative
values an error? Having an entry at http://crbug.com would give us a place to
discuss that outside the context of this CL.
---Reply---
(1) ComputeTransform2dScaleComponents() is used to calculate page scale on
layer_tree_host_common.cc,
page scale factor is greater than zero.
(2) Before optimization(Review URL: https://codereview.chromium.org/47403002),
ScaleOnAxis is return 
std::sqrt(a * a + b * b + c * c), it is always greater than zero or equal to
zero.
After optimization, it may be negative. It makes ScaleOnAxis return value
difference.
So, we should keep return value consistent in the optimization process.


(2) You're in AUTHORS from a @tencent account; you'll get this warning if you
don't commit from that account.
---Reply---
I has changed my account from @gmail.com to @tencent.com.

[aelias_OOO_until_Jul13, 2014-04-29 03:03:11]

lgtm

[Yupei Wang, 2014-04-29 03:22:33]

The CQ bit was checked by perryuwang@tencent.com

[commit-bot: I haz the power, 2014-04-29 03:23:23]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/perryuwang@tencent.com/253083002/1

[commit-bot: I haz the power, 2014-04-29 04:00:40]

The CQ bit was unchecked by commit-bot@chromium.org

[commit-bot: I haz the power, 2014-04-29 04:00:40]

Try jobs failed on following builders:
  tryserver.chromium on chromium_presubmit

[James Wei, 2014-04-30 05:00:53]

The CQ bit was checked by wistoch@chromium.org

[commit-bot: I haz the power, 2014-04-30 05:02:10]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/perryuwang@tencent.com/253083002/1

[commit-bot: I haz the power, 2014-04-30 05:09:13]

Change committed as 267078