[Webkit-unassigned] [Bug 44505] [EFL] Missing plugins support for efl port

[bugzilla-daemon at webkit.org]

https://bugs.webkit.org/show_bug.cgi?id=44505





--- Comment #39 from Mariusz Grzegorczyk <mariusz.g at samsung.com>  2011-08-30 05:14:42 PST ---
(From update of attachment 105367)
View in context: https://bugs.webkit.org/attachment.cgi?id=105367&action=review

>> Source/WebCore/plugins/efl/PluginDataEfl.cpp:12
>> +    version 2 of the License, or (at your option) any later version.
> 
> Other files from this directory uses a different license. Is this intentional?

Code was moved from GTK version, so type of license is kept. Also copyrights were moved with adding new ones.

>>> Source/WebCore/plugins/efl/PluginPackageEfl.cpp:121
>>> +    }
>> 
>> Small nit: you could ditch m_isLoaded and just use m_loadCount instead.
> 
> There can be situation that m_isLoaded is set, but m_loadCount is not incremented when module is opened, but symbols(NP_Initialize, and NP_Shutdown) can't be taken. Same implementation in Gtk.

There can be situation that m_isLoaded is set, but m_loadCount is not incremented when module is opened, but symbols(NP_Initialize, and NP_Shutdown) can't be taken. Same implementation in Gtk.

>>> Source/WebCore/plugins/efl/PluginViewEfl.cpp:106
>>> +            || event->type() == eventNames().mouseoverEvent) {
>> 
>> Minor nit: wouldn't a switch() statement be better here?
> 
> event->type() is type of AtomicString, so can't use switch here

event->type() is type of AtomicString, so can't use switch here

>> Source/WebCore/plugins/efl/PluginViewEfl.cpp:224
>> +        return NPERR_FILE_NOT_FOUND;
> 
> NPERR_FILE_NOT_FOUND should be returned if the file is missing or is invalid; in this case, isn't NPERR_NO_DATA more suitable?

>From PluginDebug.cpp it looks ok: this error means that file is missing, or is INVALID, but no data is for streams.
    "File missing or invalid.", /* NPERR_FILE_NOT_FOUND */
    "Stream contains no data.", /* NPERR_NO_DATA */

>>> Source/WebCore/plugins/efl/PluginViewEfl.cpp:236
>>> +        *result = NPERR_NO_ERROR;
>> 
>> Are value and result guaranteed to be non-null?
> 
> result is guaranteed, but value is not checked in any port. It looks that in this matter webkit trusts plugins. Also type of value is taken from type of NPNVariable which is possibly dangerous(casting from void* to different types of returned value)

result is guaranteed, but value is not checked in any port. It looks that in this matter webkit trusts plugins. Also type of value is taken from type of NPNVariable which is possibly dangerous(casting from void* to different types of returned value)

-- 
Configure bugmail: https://bugs.webkit.org/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are the assignee for the bug.