391028 - drawImage with broken PNG draws random memory

Status: RESOLVED FIXED

(Core :: Graphics: Canvas2D, defect)

Description

[Philip Taylor]

User-Agent:       Opera/9.22 (X11; Linux i686; U; en)
Build Identifier: 

With a broken PNG file (which causes "The image [...] cannot be displayed, because it contains errors." when opened in the browser normally), calling drawImage results in random stuff being drawn onto the canvas each time the page is reloaded.

Reproduced with:
    Mozilla/5.0 (Windows; U; Windows NT 5.2; en-GB; rv:1.8.1.6) Gecko/20070725 Firefox/2.0.0.6
    Mozilla/5.0 (X11; U; Linux i686; en-GB; rv:1.8.1.5) Gecko/20070713 Firefox/2.0.0.5
    Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.9a8pre) Gecko/2007080504 Minefield/3.0a8pre

Couldn't reproduce in
    Mozilla/5.0 (Windows; U; Windows NT 5.2; en-US; rv:1.9a8pre) Gecko/2007080505 Minefield/3.0a8pre
(it only ever gave a black rectangle).

HTML 5 says at http://www.whatwg.org/specs/web-apps/current-work/#complete that img.complete should be false if the image wasn't valid, and http://www.whatwg.org/specs/web-apps/current-work/#drawimage says that should cause INVALID_STATE_ERR in drawImage.


Reproducible: Always

Comment 1

[Philip Taylor]

Attachment: test case

Comment 2

[Vladimir Vukicevic [:vlad] [:vladv] (needinfo me, slow to respond)]

Attachment: add a check for completed load status

This should fix it; something similar can go in on the branch.

Comment 3

[Vladimir Vukicevic [:vlad] [:vladv] (needinfo me, slow to respond)]

Attachment: er. followup.

Wrong line in the previous patch, sigh.

Comment 4

[Daniel Veditz [:dveditz]]

Comment on attachment 282781 [details] [diff] [review]
add a check for completed load status

approved for 1.8.1.10, a=dveditz for release-drivers

Comment 5

[Daniel Veditz [:dveditz]]

Checked into 1.8 branch

Comment 6

[Carsten Book [:Tomcat]]

verified fixed 1.8.1.10 using Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.8.1.10) Gecko/2007111504 Firefox/2.0.0.10 and the testcase

- adding verified keyword

Comment 7

[Arthur]

Has this caused bug 405584?

Comment 8

[Alexander Sack]

(In reply to comment #7)
> Has this caused bug 405584?
> 

yes, the followup attachment 284556 [details] [diff] [review] wasn't checked in. 

Comment 9

[:Gavin Sharp [email: [e-mail]]]

This is still FIXED on the trunk. Bug 405584  tracks the branch regression.

Comment 10

[Justin Wood (:Callek)]

Comment on attachment 284556 [details] [diff] [review]
er. followup.

appears to have landed, though without approval. (Granted, I do feel its a safe and needed landing, but I wonder why it skipped approval)

Comment 11

[Reed Loden [:reed]]

(In reply to comment #10)
> (From update of attachment 284556 [details] [diff] [review])
> appears to have landed, though without approval. (Granted, I do feel its a safe
> and needed landing, but I wonder why it skipped approval)

The patch had a=release-drivers to land. The flag just wasn't marked as such on the patch. I'm sure dveditz or somebody will set the flag when they get to it, but it definitely had approval to land. :)

Comment 12

[Daniel Veditz [:dveditz]]

Sorry for the missing approvals, this one-bugfix release was mostly handled real-time on IRC. Nick had approvals to land on the _RELBRANCH and the mozilla 1.8 branch

Comment 13

[Al Billings [:abillings - ex-MoCo]]

Verified for 1.8.1.12 with Mozilla/5.0 (Macintosh; U; Intel Mac OS X; en-US; rv:1.8.1.12pre) Gecko/2008012208 BonEcho/2.0.0.12pre.

Comment 14

[Alexander Sack]

Attachment: combined for 1.8.0

Comment 15

[Christopher Aillon (sabbatical, not receiving bugmail)]

Comment on attachment 307242 [details] [diff] [review]
combined for 1.8.0

a=caillon for 1.8.0.15

Comment 16

[Christopher Aillon (sabbatical, not receiving bugmail)]

Attachment 307242 [details] [diff] committed to 1.8.0 branch