Summary: | failed screenblanking with newer g-d-s kernels | ||
---|---|---|---|
Product: | Gentoo Linux | Reporter: | Bob <custom_basses> |
Component: | Current packages | Assignee: | Michal Januszewski (RETIRED) <spock> |
Status: | RESOLVED FIXED | ||
Severity: | normal | CC: | imago, kernel, paolo.pedroni |
Priority: | High | ||
Version: | 2004.3 | ||
Hardware: | x86 | ||
OS: | Linux | ||
URL: | http://forums.gentoo.org/viewtopic.php?p=2001891 | ||
Whiteboard: | |||
Package list: | Runtime testing required: | --- | |
Attachments: | fbsplash-blanking-fix.patch |
Description
Bob
2005-01-22 13:10:29 UTC
Screen blanking with development-sources doesn't work either with vesafb and Nvidia GeForce FX5200 GO - but is working with: ATI Rage128 Pro with in kernel compiled aty128fb Matrox G200 with in kernel compiled matroxfb So I'm thinking this is a bug of vesafb in the kernel sources? I believe the problem you've described is caused by fbsplash, not vesafb-tng. I know about the broken blanking code in the recent versions. It'll be fixed ASAP. Reassigning to myself to take the load off from kernel@ and make sure I don't forget about it ;) Same problem here with vesafb-tng and Geforce4 MMX. I have another problem too, the progress bar of silent theme doesn't show up. I'm currently running 2.6.10-gentoo-r6 (gentoo-dev-sources) OK, everyone: is this still a problem with g-d-s-2.6.10-r7? As for the original bugreport - I've noticed one strange thing. According to the ChangeLog there were no vesafb-tng or fbsplash updates to the 2.6.9 kernel since -r2. yes, its still a problem with 2.6.10-r7. as far as the ChangeLog goes, i'm at a loss to explain why 2.6.9-r8 works properly but not the later versions. maybe some information is missing from the changelog? It does not seem to be a problem either with vesafb-tng or with fbsplash, because I have the same problem with a Radeon 9250 using radeofb (as a module) and I have never installed fbsplash. In my case the only thing which is "blanked" is the blinking cursor. this is still a problem on the 2005.0-rc5 Live CD. while installing from the Live CD, the framebuffer blanks only partially -- the outer rim of the screen is blanked, and the text is "blanked" in that the color changes from white to black. but the background image remains on, the cursor remains on, and text flow continues to be rendered, but in black instead of white. seems solved for me with 2.6.11-gentoo-r4 sources (gentoo-dev-sources) still a problem with Matrox Milennium 1 & 2 video boards using vesafb-tng driver under 2.6.11-r4. :( Created attachment 55040 [details, diff]
fbsplash-blanking-fix.patch
I also had the problem with "incomplete blanking"(using radeonfb). Today I made
some bughunting and tried to find the source of the problem. Its indeed
fbsplash broken blanking.
Though I am still not completely understand whats going on, I came up with a
patch which fixes it for me.If yours isnt a different bug(with the same
symptoms) it should work for you too.
(though im not completely sure if it is the right way to fix it)
The patchs looks good -- I included it in the latest version of fbsplash. Thanks, Imago :) It would be nice if the bug reporter and people who have posted comments could test it and let us know whether it solves the issue completely. There have been no comments in almost a month, so I'm assuming the problem is fixed now. For me it's not. I still think it's the kernel's fault, at least in my case. You can make sure that it's the kernel's fault by testing whether you can reproduce the problem on a vanilla kernel.org source tree. If you can, you should probably report the bug on the kernel bugzilla (http://bugme.osdl.org/). still a problem with gentoo-sources 2.6.11-r6, with matrox milennium 1 and 2 series video boards. Could you please check whether you can reproduce the problem with CONFIG_FB_SPLASH=n? Ah, and one more thing -- the patch is not included in gentoo-sources 2.6.11-r6, so in order to test it, you'd have to apply it manually. >> Could you please check whether you can reproduce the problem with
>> CONFIG_FB_SPLASH=n?
Yes, I've checked that for you. Turning off framebuffer support for the bootsplash makes the problem go away. I guess I would have expected that, as the screenblanking problem I had initially described involves failed blanking of the *background* colors in the framebuffer. When the framebuffer is turned off in the kernel, there are no background colors to turn off. ;)
i will try to test the kernel patch later today.
Have you had any luck with it? The problem should be fixed in gentoo-sources 2.6.12. Please reopen if it's not. |