Correct Resolution for Mobility Radeon
Hi there,
Glad to have my taskbar back (previous posts) thank you :) Now I am working on the resolution. I should be getting (and have with previous releases) 1680X1050, but am getting 1400x1050, which skews the aspect ratio. I found this troubleshooting guide: https://wiki.ubuntu.com/X/Config/Resolution but haven't had luck yet with my ATI Mobility Radeon HD 2600. Does anyone know of any other troubleshooting guides I might try? In the Xorg log, it does appear to be loading the radeon driver. Could I even downgrade the driver? I'm not looking for acceleration, just the correct aspect ratio.
Thank you,
JakeL
p.s. I tried modprobe radeon, but that locks up my computer
Update: I crawled the log a bit further, and it appears to be loading and unloading the radeon driver a couple of times, and then loading the vesa driver, which would explain the resolution. I noticed that it was trying to load the fglrx (acceleration) module and failing -- perhaps this is the reason? Is there an option I could put in my Xorg.conf to prevent it trying to load the acceleration drivers? Thanks again in advance
fglrx is a proprietary driver! Have you tried installing it?! The driver named "radeon" should be loaded. It is free software. Do you have a file /etc/X11/xorg.conf? Can you show us the logs? What is the output of the 'lspci' command?
Apparently fglrx is the first choice of x... I see that as well in my logs and certainly have never had it installed.
I believe the solution to the resolution issue is to make a custom xorg.conf. Maybe start just by explicitly specifying to use the radeon driver.
Once you have it working, read the radeon man page, there are all kinds of interesting options.
radeon will most likely not work under Trisquel. AFAIK it still relies on firmware files / blobs to provide 3D acceleration and even for pure 2D loading the driver will fail without blob.
Using fglrx / radeon with blob means installing a non-free Ubuntu-based kernel.
Fortunately this is not true. 2D works with free software.
This is not true ? hmmmm.....
http://en.gentoo-wiki.com/wiki/Radeon#Device_Firmware
...
The radeon driver requires the device firmware for proper operation. This firmware needs to be compiled into the kernel if the drivers are built-in, or simply installed and available if they are built as modules.
...
Sorry..but this was the first search result I found on firmware files. I had an HD4670 and still have an AMD6310. I am pretty shure that the radeon kernel module needs the corresponding firmware blob to be loaded to the kernel.
I am a member!
I know nothing about ATI, Radeon etc., but from reading this discussion I got the impression there is a bit of confusion here between "firmwareness" and freedom -- a firmware can be FaiF, as long as it is licensed under an appropriated license. (This is just a logical consideration, not an empiric one.)
> I know nothing about ATI, Radeon etc., but from reading this
> discussion I got the impression there is a bit of confusion here
> between "firmwareness" and freedom -- a firmware can be FaiF, as long
> as it is licensed under an appropriated license. (This is just a
> logical consideration, not an empiric one.)
How can be firmware free while there is no source available and no
documentation of the instruction set used? Skilled hackers outside AMD
do not know what this firmware does due to this. (It's not important on
what CPU does the program run.)
Older Radeon firmware is under the MIT license, I consider it a good
example of free licenses being insuffiecient for software freedom.
(It's easy to find also e.g. GPL-licensed sourceless firmware.)
I don't know a solution for the original issue discussed here; X loading
various unavailable drivers is harmless, the X driver radeon usually
works on Trisquel if a new enough kernel is used.
That wiki might include 3D in their definition of working properly.
And GustavoCM could also be right.
Anyways, I'm using ATI Radeon XPRESS 200M 5955 (PCIE) on an unadulterated Trisquel 5.5 at native resolution.
Hi there,
Sorry for the lack of updates. I found that the radeon driver was being blacklisted, so I commented that out, however, when I rebooted, my display to became a bright, flickering static screen that was most alarming! I'll need to boot into a rescue environment to undo the change, but at least it's just one character (the comment symbol) (I guess I should have known better, the same thing happened when I tried 'modprobe radeon' ;) )
I thought I'd try booting onto a 5.5 or 4.0 disc and see if I couldn't figure out from the logs what driver it was using before and perhaps some more information on a modeline. Any other suggestions are welcome -- I'll keep you posted.
Thanks again,
JakeL
Re-Hi,
it could help if you boot up the recent version of Trisquel in Live Mode and simply post your Xorg.log over at pastebin.
I am curious what driver is loaded by Xorg.
Regards,
Holger
Will do - I was able to capture the Xorg log from a 5.5 live disc, and it does look like it was using the radeon driver (will post tonight if I get a chance). If the current (quantal-based) module has been blacklisted from 6.0 (does anyone know why?), would there be a way to downgrade to the packages from 5.5 (while keeping the rest of the release)? It would at least provide a stop-gap measure. If I understand correctly, 6.0 is now the only supported version of Trisquel?
I also noticed that I was using a slightly older install ISO for 6.0 and downloaded the current one. The video didn't come up after 30-45 sec. (no HDD activity, but was that too short a time to wait?) booting from USB (with no persistence layer), so I tried the text install instead (didn't realize that was the net install). Unfortunately, it ultimately derailed over a dependency conflict with firefox-globalmenu (I think, and I don't think that was the only issue). I also noticed that at one point the text install gave the option of installing some newer kernels than the one that was ultimately installed (3.5 vs 3.2?). I might go back and start over with a burned disc. Thanks again.
Hi there,
Sorry it took me so long to get back to this. Still would appreciate the help, though. Here is my Xorg log from Trisquel 5.5 (32-bit) (works):
Here is my Xorg log from Trisquel 5.5 (64-bit) (also works):
Here is my Xorg log from Trisquel 6.0 (32-bit) (works, but is skewed - am booted into the live disc right now):
(I really hope I didn't inadvertently post any sensitive info in the above logs >_<)
I was unable to get an Xorg log from Trisquel 6.0 (64-bit), as after the splash screen, the screen goes blank and all activity stops. I can't get to a terminal screen or anything. This was regardless of whether I booted from CD, USB, or HDD (yes, I want this that much :)
Thanks in advance -- I would so appreciate any suggestions; perhaps at least the above information could get into a bug report. I'll be checking back for updates, both to this thread and to the ISO images, but I think for now I'll try Parabola on my HDD and run Trisquel 5.5 from a USB.
"radeon will most likely not work under Trisquel. AFAIK it still relies on firmware files / blobs to provide 3D acceleration and even for pure 2D loading the driver will fail without blob."
--- Darksoul71
I guess that must be true...
After a little more experimentation with Trisquel 6.0 - on a laptop with an ATI graphics card - I started experiencing some annoying graphical errors and noticed a very slow scrolling speed.
And, after reading some posts in this thread and noticing that, unlike in Trisquel 5.5, I don't have a "plymouth" splash boot any more, I suppose that my laptop was also using the generic "vesa" driver, instead of the "radeon" one.
Indeed, the behaviour of the "radeon" driver was the reason why I ended up not adopting Trisquel 5.5. (https://trisquel.info/en/forum/almost) And, I thought that the problem with my ATI graphics card had been solved. But, I guess it wasn't... Since, from what I read here, and from what I was experiencing, I suppose that my laptop was using the generic "vesa" driver instead - in order to "solve" the problem of, apparently, not being able to load the "radeon" one...
Because I have an urgency in getting this laptop to work properly, I ended up installing another distro - at least, for the meantime. So, I can't make any tests now. (But, as I said... From what I read, I doubt that I will ever be able to solve the problem with my ATI card...)
Anyway... I just wanted to ask:
If anyone ever gets to make Trisquel 6.0 run properly on a amd64 laptop with an ATI RS482 [Radeon Xpress 200M] graphics card (ATI Radeon Xpress 1100, on the sticker), please let us know how to do it.
No, that's not true. I have Radeon XPRESS 200M 5955 (PCIE) and 2D works OK now.
It's not super great but it works. I'm still on 5.5 for what it's worth though. I've noticed that I need to use the nomodeset kernel parameter in order to be able to suspend and hibernate. And the rendering performance was originally pretty bad, especially scrolling. So I tried the older and supposedly worse acceleration architecture XAA and it gave me much better performance.
My (entire) xorg.conf is
Section "Device"
Identifier "Configured Video Device"
Driver "radeon"
Option "ClockGating" "on"
Option "DynamicPM" "on"
Option "AccelMethod" "XAA"
Option "FBTexPercent" "0"
EndSection
If you haven't read the radeon manual page, now would be a good time. Of course, your card might not work even if mine does.
Hi there, lembas.
Thanks for the tip. I will take a look at the "radeon" manual, once I have the time to.
Yes. On 5.5 I could operate my laptop pretty well. (And I have the impression that I also had to add the "nomodeset" parameter - like I did in 6.0, to make the computer be able to boot into the graphical login.) And the only problem I had in 5.5 was that it took a whole minute to shut down, because of error messages related to the "radeon" driver. But, anyway, 5.5 will soon not be supported any more, so what I wish for is that it would be possible to make my graphics card work well on 6.0.
I've informed myself about my graphics card, and it looks like we really might have the same one (or one pretty similar). I was not even aware that mine was also a PCI-E one. So, it looks that I might have something to learn from your experience.
If you can, when you adopt 6.0, please tell us, here, what you did to make it load the "radeon" driver and operate it well.