Drop needless self protections from ResourceLoader.

Drop needless self protections from ResourceLoader.

This is a refactoring and does not change behavior.

BUG=None

Committed: https://src.chromium.org/viewvc/blink?view=rev&revision=185413

[yhirano, 2014-11-14 06:04:42]

yhirano@chromium.org changed reviewers:
+ kouhei@chromium.org

[kouhei (in TOK), 2014-11-14 07:35:10]

kouhei@chromium.org changed reviewers:
+ japhet@chromium.org

[kouhei (in TOK), 2014-11-14 07:35:11]

Hmm. I'm not sure if this change is really safe.

Would you provide more justification for this change?

[yhirano, 2014-11-14 09:32:35]

On 2014/11/14 07:35:11, kouhei wrote:
> Hmm. I'm not sure if this change is really safe.
> 
> Would you provide more justification for this change?

Removing protection from didSendData should be safe, because there is only one
statement.
In releaseResources,
 - m_loader::cancel does nothing with the blink resource loader[1].
 - m_loader destructor does nothing other than cancel.
 - m_host.clear() may call Resource methods in ResourceFetcher::clearPreloads,
but this change does not change its behavior because we already clear
ResourceLoader reference from Resource at L109-L111.

[yhirano, 2014-11-14 09:32:46]

On 2014/11/14 09:32:35, yhirano wrote:
> On 2014/11/14 07:35:11, kouhei wrote:
> > Hmm. I'm not sure if this change is really safe.
> > 
> > Would you provide more justification for this change?
> 
> Removing protection from didSendData should be safe, because there is only one
> statement.
> In releaseResources,
>  - m_loader::cancel does nothing with the blink resource loader[1].
>  - m_loader destructor does nothing other than cancel.
>  - m_host.clear() may call Resource methods in ResourceFetcher::clearPreloads,
> but this change does not change its behavior because we already clear
> ResourceLoader reference from Resource at L109-L111.

1:
https://code.google.com/p/chromium/codesearch#chromium/src/content/child/web_...

[kouhei (in TOK), 2014-11-14 09:55:54]

Discussed offline. The whole hack seems to be historically inherited from old
implementation for Mac. We inspected callsites but couldn't find WebURLLoader
impl which relys on this hack.

lgtm

[yhirano1, 2014-11-14 09:57:53]

Nate, can you take a look?

[Nate Chapin, 2014-11-14 18:04:17]

A quick look around agrees with your analysis. LGTM

[yhirano, 2014-11-17 02:31:57]

On 2014/11/14 18:04:17, Nate Chapin wrote:
> A quick look around agrees with your analysis. LGTM

Thank you!

[yhirano, 2014-11-17 02:32:04]

The CQ bit was checked by yhirano@chromium.org

[commit-bot: I haz the power, 2014-11-17 02:33:09]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/728453002/20001

[commit-bot: I haz the power, 2014-11-17 03:08:41]

Committed patchset #2 (id:20001) as 185413