[OpenJDK 2D-Dev] [9] RFR JDK-8164931 : Verify if writer.abort() works properly for all writers in IIOWriteProgressListener.

Sergey Bylokhov Sergey.Bylokhov at oracle.com
Fri Sep 16 11:06:14 UTC 2016


Looks fine, expect the changes in TIFFImageWriter.java. I am not sure I 
understand why abortRequested was moved from the end of insert();

On 16.09.16 12:04, Jayathirth D V wrote:
> Hi,
>
>
>
> Please review the following fix in JDK9 at your convenience:
>
>
>
> This issue is similar to
> https://bugs.openjdk.java.net/browse/JDK-4924727 where we made changes
> to all ImageReader plugins.
>
>
>
> Bug : https://bugs.openjdk.java.net/browse/JDK-8164931
>
>
>
> Webrev : http://cr.openjdk.java.net/~jdv/8164931/webrev.00/
>
>
>
> Issue : Verify that when we issue ImageWriter.abort() in
> IIOWriteProgressListener callbacks whether writing is aborted properly.
>
>
>
> Root cause : In many writer plugins we are not checking for
> abortRequested() right after IIOWriteProgressListener callbacks. In
> which case writing may continue until we check for abortRequested().
> Also in some writers we are not calling clearAbortRequest() before every
> write() call.
>
>
>
> Solution : Check for abortRequested() after every
> IIOWriteProgressListener callbacks and before every write() call we
> should have clearAbortRequest() called.
>
>
>
> In case of JPEG clearAbortRequest() is overridden in JPEGImageWriter and
> it clears native abort flag also so there is no change in
> JPEGImageWriter. WBMP changes will be done in JDK-8164930 as
> checkSampleModel() is failing for WBMP.
>
> Thanks,
>
> Jay
>
>
>


-- 
Best regards, Sergey.



More information about the 2d-dev mailing list