Summary: | SDVO-Card + DVI-Port + modesetting branch + 1680px-screen result in bogus resolution | ||||||
---|---|---|---|---|---|---|---|
Product: | xorg | Reporter: | Johannes Truschnigg <johannes.truschnigg> | ||||
Component: | Driver/intel | Assignee: | Eric Anholt <eric> | ||||
Status: | RESOLVED FIXED | QA Contact: | |||||
Severity: | normal | ||||||
Priority: | high | CC: | dcm | ||||
Version: | git | ||||||
Hardware: | x86 (IA32) | ||||||
OS: | Linux (All) | ||||||
Whiteboard: | |||||||
i915 platform: | i915 features: | ||||||
Attachments: |
|
Description
Johannes Truschnigg
2007-01-16 12:17:20 UTC
It's me again - since DRI started to work again with the driver version I checked out today, I tried starting ioquake3 with software rendering forced (since my version of Mesa is too old for hardware-accelerated GLX to make the cut), and after quitting the game, I'm back at a perfect 1680x1050px-desktop and quite happy with it. I could imagine that info might come in handy when tracking that weird issue down. Cheers! This bug occurs for me also. I have a Intel DG965WH motherboard and I'm using a Wintec-branded SDVO card and its DVI output. My monitor is a Dell 2000FP with a native resolution of 1600x1200. When I start X with the mode setting branch, my monitor reports that it is being fed at 1656x1200. The desktop itself is still exactly 1600x1200, but is left-justified and squished slightly into the 1656x1200 mode. I will attach a patch that works around the problem for me. Created attachment 8517 [details] [review] Workaround This patch avoids the problem by eliminating the mysterious requirement "H Blanking size must be < 63*8". Presumably that requirement is in there for a reason, but it is incompatable with a lot of high resolution modelines. If I simply remove the test, it "works for me". Yeah, great stuff - works for me too. My rather strange "workaround" (starting an OpenGL application fullscreen, making it switch resolutions, and get back to my desktop with proper 1680x1050) ceased to work about two days ago after updating the driver once more, alongside breaking DRI, by the way. David Moore's patch fixes the issue for me. - Tristan Committed an even more aggressive version of this patch, which ajax and I had talked about on IRC a while back. Thanks! |
Use of freedesktop.org services, including Bugzilla, is subject to our Code of Conduct. How we collect and use information is described in our Privacy Policy.