<AWT Dev> <AWT dev>[9] Review request for 8156121: "Fail forward" fails for GTK3 if no GTK2 available

Phil Race philip.race at oracle.com
Wed Jun 1 18:37:45 UTC 2016


+1

-phil.

On 05/30/2016 02:00 AM, Semyon Sadetsky wrote:
> Please review the updated webrev:
>
> http://cr.openjdk.java.net/~ssadetsky/8156121/webrev.01/
>
> --Semyon
>
>
> On 5/27/2016 7:30 PM, Phil Race wrote:
>>  67 n_libs = sizeof(gtk_libs) / sizeof(GtkLib);
>>   68         load_order = malloc(1 + n_libs * sizeof(GtkLib *));
>>   69         load_order[n_libs] = 0;
>>
>> I think you meant
>>
>>  malloc((1 + n_libs) * sizeof(GtkLib *));
>>
>> since otherwise you are only allocating one byte for the pointer
>> and will have a buffer overrun.
>>
>> But it would be better to use calloc :-
>>
>>
>> calloc(n_libs+1,sizeof(GtkLib *));
>>
>> and then you can dispense with line 69 since calloc zero initialises 
>> memory.
>>
>> -phil.
>>
>>
>> On 05/26/2016 05:33 AM, Semyon Sadetsky wrote:
>>> Hello,
>>>
>>> Please review fix for JDK9:
>>>
>>> bug: https://bugs.openjdk.java.net/browse/JDK-8156121
>>>
>>> webrev: http://cr.openjdk.java.net/~ssadetsky/8156121/webrev.00/
>>>
>>> The issue is caused by changes in -Djdk.gtk.version=<gtk version> 
>>> specification. Initially it should be working as a constraint:  only 
>>> allow the specified version of GTK lib (don't use GTK if this 
>>> version is not available). Then after discussions the specification 
>>> was amended to use this property as a advise: the GTK version to 
>>> start loading with (if it fails the next available GTK version will 
>>> be attempted). So that, the code need be harmonized with the 
>>> specification.
>>>
>>> --Semyon
>>>
>>
>



More information about the awt-dev mailing list