Load HRTF database only when the panner is set to HRTF

Load HRTF database only when the panner is set to HRTF

Currently, the creation of any panner node causes the HRTF database to
be loaded.  However, the default panning model is "equalpower" (about
99.8%) so loading the database is a waste of time and memory.
(According to https://codereview.chromium.org/176683003/, it takes
30MB to hold the database.)

Thus, only load the HRTF database when the panning model is set to
HRTF.

Can't test this easily, but did verify with a debugger that the
database is only created when the panning model is set to HRTF and not
before.

BUG=592787
TEST=none

Committed: https://crrev.com/6610df622ffb33ade51e5ccdb92b7d6b0e2b81ce
Cr-Commit-Position: refs/heads/master@{#407307}

[Raymond Toy, 2016-07-21 16:36:55]

rtoy@chromium.org changed reviewers:
+ hongchan@chromium.org

[Raymond Toy, 2016-07-21 16:36:55]

PTAL.

Not quite sure if it's ok style to have a case statement fall through to the
next.  I can rearrange the code if you like.

[hongchan, 2016-07-21 16:48:53]

https://codereview.chromium.org/2171703003/diff/1/third_party/WebKit/Source/m...
File third_party/WebKit/Source/modules/webaudio/PannerNode.cpp (right):

https://codereview.chromium.org/2171703003/diff/1/third_party/WebKit/Source/m...
third_party/WebKit/Source/modules/webaudio/PannerNode.cpp:293: switch (model) {
If this fall through is necessary, why do we even need the switch here?

We can load HRTF database only when the model is HRTF, and we always perform the
rest of the step. No?

[Raymond Toy, 2016-07-21 16:57:34]

https://codereview.chromium.org/2171703003/diff/1/third_party/WebKit/Source/m...
File third_party/WebKit/Source/modules/webaudio/PannerNode.cpp (right):

https://codereview.chromium.org/2171703003/diff/1/third_party/WebKit/Source/m...
third_party/WebKit/Source/modules/webaudio/PannerNode.cpp:293: switch (model) {
On 2016/07/21 16:48:53, hoch wrote:
> If this fall through is necessary, why do we even need the switch here?
> 
> We can load HRTF database only when the model is HRTF, and we always perform
the
> rest of the step. No?

Good point. I guess it's a safety check that we didn't call this with an invalid
value (the default case in the switch).

[Raymond Toy, 2016-07-22 20:43:52]

Anything more I need to do?

[hongchan, 2016-07-22 20:48:45]

lgtm

[Raymond Toy, 2016-07-22 21:15:12]

The CQ bit was checked by rtoy@chromium.org

[commit-bot: I haz the power, 2016-07-22 21:15:56]

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, 2016-07-23 00:01:21]

Committed patchset #2 (id:20001)

[commit-bot: I haz the power, 2016-07-23 00:02:58]

Description was changed from

==========
Load HRTF database only when the panner is set to HRTF

Currently, the creation of any panner node causes the HRTF database to
be loaded.  However, the default panning model is "equalpower" (about
99.8%) so loading the database is a waste of time and memory.
(According to https://codereview.chromium.org/176683003/, it takes
30MB to hold the database.)

Thus, only load the HRTF database when the panning model is set to
HRTF.

Can't test this easily, but did verify with a debugger that the
database is only created when the panning model is set to HRTF and not
before.

BUG=592787
TEST=none
==========

to

==========
Load HRTF database only when the panner is set to HRTF

Currently, the creation of any panner node causes the HRTF database to
be loaded.  However, the default panning model is "equalpower" (about
99.8%) so loading the database is a waste of time and memory.
(According to https://codereview.chromium.org/176683003/, it takes
30MB to hold the database.)

Thus, only load the HRTF database when the panning model is set to
HRTF.

Can't test this easily, but did verify with a debugger that the
database is only created when the panning model is set to HRTF and not
before.

BUG=592787
TEST=none

Committed: https://crrev.com/6610df622ffb33ade51e5ccdb92b7d6b0e2b81ce
Cr-Commit-Position: refs/heads/master@{#407307}
==========

[commit-bot: I haz the power, 2016-07-23 00:02:59]

Patchset 2 (id:??) landed as
https://crrev.com/6610df622ffb33ade51e5ccdb92b7d6b0e2b81ce
Cr-Commit-Position: refs/heads/master@{#407307}