Renderer.LineLayoutMs UMA histogram to measure line layout time.

Renderer.LineLayoutMs UMA histogram to measure line layout time.

Committed: https://src.chromium.org/viewvc/blink?view=rev&revision=189578

[benjhayden, 2015-02-02 23:46:02]

benjhayden@chromium.org changed reviewers:
+ eae@chromium.org

[benjhayden, 2015-02-02 23:46:02]

Is this what will be impacted by your work to deduplicate line layout? If not,
can you point me to whatever that is?
It looks like this can recurse through RenderBlockFlow::insertFloatingObject()
<- LineBreaker::skipLeadingWhitespace(), so it can potentially include time for
anything called via RenderObject::layout(). It would be good to find a more
specific bit of code to time that didn't recurse to layout().

[eae, 2015-02-02 23:55:02]

On 2015/02/02 23:46:02, benjhayden_chromium wrote:
> Is this what will be impacted by your work to deduplicate line layout? If not,
> can you point me to whatever that is?
> It looks like this can recurse through RenderBlockFlow::insertFloatingObject()
> <- LineBreaker::skipLeadingWhitespace(), so it can potentially include time
for
> anything called via RenderObject::layout(). It would be good to find a more
> specific bit of code to time that didn't recurse to layout().

Yes. Basically it is about reusing work from computePreferedWidths during
line-layout and breaking.

LGTM

[benjhayden, 2015-02-03 00:17:00]

On 2015/02/02 at 23:55:02, eae wrote:
> On 2015/02/02 23:46:02, benjhayden_chromium wrote:
> > Is this what will be impacted by your work to deduplicate line layout? If
not,
> > can you point me to whatever that is?
> > It looks like this can recurse through
RenderBlockFlow::insertFloatingObject()
> > <- LineBreaker::skipLeadingWhitespace(), so it can potentially include time
for
> > anything called via RenderObject::layout(). It would be good to find a more
> > specific bit of code to time that didn't recurse to layout().
> 
> Yes. Basically it is about reusing work from computePreferedWidths during
line-layout and breaking.
> 
> LGTM

Should I not worry about the recursion, or is there a more specific bit of code
that I could time instead?

[eae, 2015-02-03 00:19:57]

On 2015/02/03 00:17:00, benjhayden_chromium wrote:
> On 2015/02/02 at 23:55:02, eae wrote:
> > On 2015/02/02 23:46:02, benjhayden_chromium wrote:
> > > Is this what will be impacted by your work to deduplicate line layout? If
> not,
> > > can you point me to whatever that is?
> > > It looks like this can recurse through
> RenderBlockFlow::insertFloatingObject()
> > > <- LineBreaker::skipLeadingWhitespace(), so it can potentially include
time
> for
> > > anything called via RenderObject::layout(). It would be good to find a
more
> > > specific bit of code to time that didn't recurse to layout().
> > 
> > Yes. Basically it is about reusing work from computePreferedWidths during
> line-layout and breaking.
> > 
> > LGTM
> 
> Should I not worry about the recursion, or is there a more specific bit of
code
> that I could time instead?

I think this is a good start. We might want to add a separate metric for
prefswifth calculation and line breaking but lets revisit that later.

[benjhayden, 2015-02-03 19:55:32]

The CQ bit was checked by benjhayden@chromium.org

[commit-bot: I haz the power, 2015-02-03 19:55:54]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/877293003/20001

[commit-bot: I haz the power, 2015-02-03 22:18:15]

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

[commit-bot: I haz the power, 2015-02-03 22:18:15]

Try jobs failed on following builders:
  mac_blink_rel on tryserver.blink
(http://build.chromium.org/p/tryserver.blink/builders/mac_blink_rel/builds/41086)

[benjhayden, 2015-02-03 22:21:22]

The CQ bit was checked by benjhayden@chromium.org

[commit-bot: I haz the power, 2015-02-03 22:22:11]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/877293003/20001

[commit-bot: I haz the power, 2015-02-04 00:53:48]

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

[commit-bot: I haz the power, 2015-02-04 00:53:49]

Try jobs failed on following builders:
  mac_blink_rel on tryserver.blink
(http://build.chromium.org/p/tryserver.blink/builders/mac_blink_rel/builds/41127)

[benjhayden, 2015-02-05 16:59:23]

The CQ bit was checked by benjhayden@chromium.org

[commit-bot: I haz the power, 2015-02-05 17:00:03]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/877293003/20001

[commit-bot: I haz the power, 2015-02-05 18:16:51]

Committed patchset #2 (id:20001) as
https://src.chromium.org/viewvc/blink?view=rev&revision=189578