Oilpan: trace part objects as per the GC plugin's wishes.

Oilpan: trace part objects as per the GC plugin's wishes.

Following https://src.chromium.org/viewvc/chrome?view=rev&revision=287310,
the Blink clang plugin insists that part objects are to be traced
via Visitor's fallback method

  template<typename T>void trace(const T& t);

Fix up some uses that were not.

R=haraken
BUG=340522
NOTRY=true

Committed: https://src.chromium.org/viewvc/blink?view=rev&revision=180152

[sof, 2014-08-13 07:35:57]

Please take a look.

[haraken, 2014-08-13 07:42:35]

LGTM

[sof, 2014-08-13 07:43:28]

The CQ bit was checked by sigbjornf@opera.com

[commit-bot: I haz the power, 2014-08-13 07:44:19]

CQ is trying da patch. Follow status at
 https://chromium-status.appspot.com/cq/sigbjornf@opera.com/471463002/1

[commit-bot: I haz the power, 2014-08-13 07:44:49]

Change committed as 180152

[Mads Ager (chromium), 2014-08-13 07:45:21]

LGTM, thank you!!!

Ian, we should update the plugin to do at least limited checks on the forms of
trace calls that we see in trace methods.

[Mads Ager (chromium), 2014-08-13 07:47:59]

On 2014/08/13 07:45:21, Mads Ager (chromium) wrote:
> LGTM, thank you!!!
> 
> Ian, we should update the plugin to do at least limited checks on the forms of
> trace calls that we see in trace methods.

Disregard that comment Ian, that is exactly what is going on here. :-)

One reason why this is really important (and I thought that was what prompted
Sigbjorn to do this) is that the visitor overload will soon have a crucial null
vtable check: https://codereview.chromium.org/455363002/

[sof, 2014-08-13 07:50:29]

On 2014/08/13 07:47:59, Mads Ager (chromium) wrote:
> On 2014/08/13 07:45:21, Mads Ager (chromium) wrote:
> > LGTM, thank you!!!
> > 
> > Ian, we should update the plugin to do at least limited checks on the forms
of
> > trace calls that we see in trace methods.
> 
> Disregard that comment Ian, that is exactly what is going on here. :-)
> 
> One reason why this is really important (and I thought that was what prompted
> Sigbjorn to do this) is that the visitor overload will soon have a crucial
null
> vtable check: https://codereview.chromium.org/455363002/

Nah, just doing trivial stuff like building with a ToT version of the GC plugin
(the extra constraints on trace() usage in the referred-to CL is most welcome,
btw.)