User-Agent: Mozilla/5.0 (Windows NT 5.0; rv:2.0b9pre) Gecko/20101218 Firefox/4.0b9pre Build Identifier: Mozilla/5.0 (Windows NT 5.0; rv:2.0b9pre) Gecko/20101218 Firefox/4.0b9pre If an html5 video is scaled, either by CTRL - or in the web page with the html markup, in Windows 2000 black vertical lines appear. This doesn't happen in Ubuntu 10.10 (it's another computer, I don't know if it is caused by hardware or software). Graphics card: Radeon IGP 320m Reproducible: Always Steps to Reproduce: 1.Open html5 video 2.Ctrl - Actual Results: Black vertical lines above video content Expected Results: Scaled video
Please post the graphics section from about:support Also, can you attach a screenshot showing the issue.
(In reply to comment #1) > Please post the graphics section from about:support > > Also, can you attach a screenshot showing the issue. Adapter Description: RADEON IGP 320M Vendor: ID1002 Device: ID4336 Adapter RAM: Unknown Adapter Drivers: ati2dvag Driver Version: 18.104.22.16818 Driver Date: 11-7-2002 Direct2D Enabled: false DirectWrite Enabled: false WebGL Renderer:(WebGL unavailable) GPU Accelerated Windows: 0/1 It is definitely a software problem: I have tried the same thing on the same computer using Ubuntu 6.10 and it works. I'll soon attach the screenshot
I'm not sure if this is the same issue as bug 618458 or not, but in both cases it is related to scaling.
(In reply to comment #4) > I'm not sure if this is the same issue as bug 618458 or not, but in both cases > it is related to scaling. I can't help with code, but it could be related to bug 618458 as sometimes green parts appear too. I have encountered the same problem on Win Xp too. To sum up: 2001 computer (radeon...) | 2009 computer (Mobile Intel (r) 4 Series Express) ****************************************************************************** Windows 2000: error | Ubuntu 10.10: works Windows XP: error | Windows 7 : works Ubuntu 6.10: works | Windows XP : works
What CPU model does the computer have where you're seeing this problem?
(In reply to comment #6) > What CPU model does the computer have where you're seeing this problem? Compaq, EvoN1015V x86 Family 6 Model 8 Stepping 1 AuthenticAMD ~1656 MHZ
Thanks, that appears to be an Athlon XP (Thoroughbred B), which has support for MMX and SSE, so it should be using the SSE version of the colour conversion code. Your paste of about:support earlier confirms that you're not using the hardware accelerated path for this. I tested both the SSE and plain C (just in case we're falling back to C on your CPU) versions of the code on a Windows 7 machine (with HW acceleration disabled), but still failed to reproduce this problem.
From your original steps, it sounds like the video is playing when you hit Ctrl and - to scale the video. Is that the case? Does the corruption appear instantly at that point? Starting from the first step, if you pause the video so that the current frame is not corrupt and then scale down with Ctrl and -, do you see corruption in that case? If not, does it appear once you start playing the video?
Oh, this might be a duplicate of bug 616778, which is now fixed. Can you please test the latest nightly build?
(In reply to comment #10) > Oh, this might be a duplicate of bug 616778, which is now fixed. Can you > please test the latest nightly build? You're right, now it works for me. Mark as resolved duplicate?
Thanks for confirming!