[OpenJDK 2D-Dev] [10] RFR JDK-8190512: PngReader incorrectly throws IllegalArgumentException for malformed images with negative dimensions
Jayathirth D V
jayathirth.d.v at oracle.com
Fri Nov 10 08:31:34 UTC 2017
Hi Brian,
Submitter shared PNG image with negative IHDR height also.
And he also clarified from his side that there are no copyright issues.
I have updated the test case to check for negative width and negative height input.
Please find updated webrev for review:
http://cr.openjdk.java.net/~jdv/8190512/webrev.01/
Thanks,
Jay
From: Jayathirth D V
Sent: Thursday, November 09, 2017 12:38 PM
To: Brian Burkhalter
Cc: 2d-dev
Subject: Re: [OpenJDK 2D-Dev] [10] RFR JDK-8190512: PngReader incorrectly throws IllegalArgumentException for malformed images with negative dimensions
Hi Brian,
Thanks for your review.
The encoded string present in test case has only the critical chunks, I have checked and verified the same with Phil also offline and it looks like it doesn't have any text chunks like iTXt, tEXt, and zTXt which may contain copyright information. I have sent mail to submitter also to ask for copyright issues, if any with the provided images. Waiting for the response from submitter. But looking into encoded stream it looks to be safe from copyright issues.
Regarding negative height in IHDR. I think we can't create PNG images with negative IHDR height using ImageIO so I have asked the submitter for another sample image with negative height and non-negative width. I can get such an image I will use encoded stream and update the test case.
Thanks,
Jay
From: Brian Burkhalter
Sent: Thursday, November 09, 2017 12:52 AM
To: Jayathirth D V
Cc: Prahalad Kumar Narayanan; 2d-dev
Subject: Re: [OpenJDK 2D-Dev] [10] RFR JDK-8190512: PngReader incorrectly throws IllegalArgumentException for malformed images with negative dimensions
Hi Jay,
I don't see anything wrong with using an encoded string if it is not derived from a copyrighted image: it's better than checking an image binary into the SCM. It would be nice however to test for negative height as well in order to get full coverage.
Thanks,
Brian
On Nov 8, 2017, at 2:35 AM, Jayathirth D V <HYPERLINK "mailto:jayathirth.d.v at oracle.com"jayathirth.d.v at oracle.com> wrote:
I would like to get inputs from others also so that we can decide to use or not to use the encoded string for image information from submitter in the JDK.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.openjdk.java.net/pipermail/2d-dev/attachments/20171110/71b313ce/attachment.html>
More information about the 2d-dev
mailing list