cc: PaintTimeCounter display on the HudLayer in continuous painting mode

cc: PaintTimeCounter display on the HudLayer in continuous painting mode

This change adds a display for the PaintTimeCounter to the HudLayer. The display is shown in continuous painting mode. Common functionality with the FPSCounter is refactored into a Graph helper structure and drawing methods.

screenshot: https://docs.google.com/open?id=0B8Y78t3tjy1XOVdvWGFGdXFOMzA
depends on: https://codereview.chromium.org/11827009

BUG=


Committed: https://src.chromium.org/viewvc/chrome?view=rev&revision=177579

[nduca, 2013-01-10 20:06:12]

I like this. Shawnsingh for final review.

[egraether, 2013-01-11 18:34:02]

On 2013/01/10 20:06:12, nduca wrote:
> I like this. Shawnsingh for final review.

@shawnsingh this is ready for review.
PaintTimeCounter got approved: https://codereview.chromium.org/11827009

[shawnsingh, 2013-01-11 18:41:00]

I'll take a look later today =)

In the mean-time, can you please explain the concept of "high" in the Graph? 
I'm not seeing it clearly.   what's the difference between "max" and "high"?

[egraether, 2013-01-11 19:00:42]

On 2013/01/11 18:41:00, shawnsingh wrote:
> I'll take a look later today =)
> 
> In the mean-time, can you please explain the concept of "high" in the Graph? 
> I'm not seeing it clearly.   what's the difference between "max" and "high"?

"max" is the current maximum value of each counter. It is displayed in the top
right of the display as number.

"high" is the current upper bound of the graph visual. It is used to scale the
values in the graph to the correct visual range. "high" has a "defaultHigh".
When this is exceeded by max, "high" will ease towards "max" to accommodate the
whole range of values. Using the easing it is easier for the user to follow the
scaling of the graph.

The "defaultHigh" is chosen so that this scaling won't happen commonly. In the
fpsGraph you can only see it when using the --disable-gpu-vsync flag.

I'm open to do some renamings here.

[shawnsingh, 2013-01-15 22:48:52]

looks good!  I only have very minor nits.

Deferring to nduca@ for the trivial layer_tree_impl.* parts of this patch.

I tried to think of a better name than "high", but couldn't think of one.  Maybe
"range", but that isn't much clearer than high.

https://codereview.chromium.org/11820014/diff/9001/cc/heads_up_display_layer_...
File cc/heads_up_display_layer_impl.h (right):

https://codereview.chromium.org/11820014/diff/9001/cc/heads_up_display_layer_...
cc/heads_up_display_layer_impl.h:51: static double updateHigh(Graph&);
If Graph is supposed to be non-const, let's use a pointer here.  I know some
other places in this file are using non-const references instead of pointers,
but we should also eventually be migrating those from the webkit style to
chromium style eventually, too.  This is one style point I feel we should go
ahead and use chromium style even if the rest of the file is still using webkit
style.

https://codereview.chromium.org/11820014/diff/9001/cc/heads_up_display_layer_...
cc/heads_up_display_layer_impl.h:68: void drawTextLeftAligned(SkCanvas*,
SkPaint&, SkRect bounds, std::string text);
same here and for all the following functions - either pointer or const&

https://codereview.chromium.org/11820014/diff/9001/cc/heads_up_display_layer_...
cc/heads_up_display_layer_impl.h:86: base::TimeTicks m_lastGraphNumbersUpdate;
more nitpickyness -  perhaps this is more clear named as "timeOfLastGraphUpdate"

[egraether, 2013-01-16 00:53:02]

Resolved the nits.

https://codereview.chromium.org/11820014/diff/13001/cc/heads_up_display_layer...
File cc/heads_up_display_layer_impl.h (right):

https://codereview.chromium.org/11820014/diff/13001/cc/heads_up_display_layer...
cc/heads_up_display_layer_impl.h:51: static double updateUpperBound(Graph*);
Renamed "high" to "upperBound".

[nduca, 2013-01-16 00:59:43]

lgtm

[egraether, 2013-01-17 22:07:36]

The PaintTimeCounter was switched to use base::TimeDelta (
https://codereview.chromium.org/11827009 ). I updated this patch to the new API.

@shawnsingh Can you give this a final look please?

[shawnsingh, 2013-01-17 23:30:00]

LGTM too

On 2013/01/17 22:07:36, egraether wrote:
> The PaintTimeCounter was switched to use base::TimeDelta (
> https://codereview.chromium.org/11827009 ). I updated this patch to the new
API.
> 
> @shawnsingh Can you give this a final look please?

[commit-bot: I haz the power, 2013-01-18 00:59:35]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/egraether@chromium.org/11820014/21002

[commit-bot: I haz the power, 2013-01-18 02:49:18]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/egraether@chromium.org/11820014/21002

[commit-bot: I haz the power, 2013-01-18 03:14:26]

Change committed as 177579