[css-grid] Fix 1% regression on system_health.memory_desktop

[css-grid] Fix 1% regression on system_health.memory_desktop

Apparently crrev.com/407361 added a 1% memory regression. The only thing
that could be increasing memory consumption is the fact that we now require
a vector to store the auto tracks instead of just a single GridTrack. We can
make the Vector just reserve 1 element of capacity by default.

BUG=631057
Committed: https://crrev.com/5d2f646883db3f2abe8a503c9702698d1e3cfab7
Cr-Commit-Position: refs/heads/master@{#415916}

[svillar, 2016-08-30 10:24:09]

svillar@igalia.com changed reviewers:
+ cbiesinger@chromium.org, eae@chromium.org, jfernandez@igalia.com,
rego@igalia.com

[jfernandez, 2016-08-30 11:05:14]

https://codereview.chromium.org/2292033002/diff/1/third_party/WebKit/Source/c...
File third_party/WebKit/Source/core/style/ComputedStyle.cpp (right):

https://codereview.chromium.org/2292033002/diff/1/third_party/WebKit/Source/c...
third_party/WebKit/Source/core/style/ComputedStyle.cpp:2006:
trackSizeList.reserveInitialCapacity(1);
Wouldn't make more sense to pay with Vector's inlineCapacity ? We may get
performance improvements by setting an adjusted inlineCapacity value if we knoe
the vector is going to be small.  Not really sure about the exact value, though.

[svillar, 2016-08-30 13:44:27]

On 2016/08/30 11:05:14, jfernandez wrote:
>
https://codereview.chromium.org/2292033002/diff/1/third_party/WebKit/Source/c...
> File third_party/WebKit/Source/core/style/ComputedStyle.cpp (right):
> 
>
https://codereview.chromium.org/2292033002/diff/1/third_party/WebKit/Source/c...
> third_party/WebKit/Source/core/style/ComputedStyle.cpp:2006:
> trackSizeList.reserveInitialCapacity(1);
> Wouldn't make more sense to pay with Vector's inlineCapacity ? We may get
> performance improvements by setting an adjusted inlineCapacity value if we
knoe
> the vector is going to be small.  Not really sure about the exact value,
though.

That would be nice but by doing so you will be forced to change the signatures
of all the methods in ComputedStyle which tough any grid-auto-row|column
property including setGridAutoRows,setGridAutoColumns....

It's a possibility, but I'd like to try with this simpler approach and
reevaluate later if the regression is still there.

[cbiesinger, 2016-08-30 14:11:57]

lgtm

[eae, 2016-08-30 14:53:18]

LGTM

[svillar, 2016-09-01 07:42:22]

The CQ bit was checked by svillar@igalia.com

[commit-bot: I haz the power, 2016-09-01 07:42:53]

CQ is trying da patch. Follow status at
 
https://chromium-cq-status.appspot.com/v2/patch-status/codereview.chromium.or...

[commit-bot: I haz the power, 2016-09-01 09:00:04]

Committed patchset #1 (id:1)

[commit-bot: I haz the power, 2016-09-01 09:03:24]

Description was changed from

==========
[css-grid] Fix 1% regression on system_health.memory_desktop

Apparently crrev.com/407361 added a 1% memory regression. The only thing
that could be increasing memory consumption is the fact that we now require
a vector to store the auto tracks instead of just a single GridTrack. We can
make the Vector just reserve 1 element of capacity by default.

BUG=631057
==========

to

==========
[css-grid] Fix 1% regression on system_health.memory_desktop

Apparently crrev.com/407361 added a 1% memory regression. The only thing
that could be increasing memory consumption is the fact that we now require
a vector to store the auto tracks instead of just a single GridTrack. We can
make the Vector just reserve 1 element of capacity by default.

BUG=631057

Committed: https://crrev.com/5d2f646883db3f2abe8a503c9702698d1e3cfab7
Cr-Commit-Position: refs/heads/master@{#415916}
==========

[commit-bot: I haz the power, 2016-09-01 09:03:25]

Patchset 1 (id:??) landed as
https://crrev.com/5d2f646883db3f2abe8a503c9702698d1e3cfab7
Cr-Commit-Position: refs/heads/master@{#415916}