Blank screen or distorted image because of wrong default AGPMode value
Affects | Status | Importance | Assigned to | Milestone | |
---|---|---|---|---|---|
X.Org X server |
Fix Released
|
Medium
|
|||
xserver-xorg-video-ati (Ubuntu) |
Fix Released
|
High
|
Bryce Harrington | ||
Bug Description
Binary package hint: xserver-
Since the updates a few days ago on gutsy, my screen shows horizontal tearing streaks with only the mouse pointer shown normally (see http://
Usplash is displayed normally, gdm and gnome are scrambled.
This problem was first seen on Gutsy with the upload of the 1:6.6.193-1ubuntu1 ati driver, and worked fine on Feisty and Breezy. Downgrading xserver-
[Workaround]
Disable 3D (including Desktop Effects) by adding this to your Modules section:
Section "Module"
...
Disable "dri"
EndSection
If the above does not fix the issue for you, then you are not seeing this particular bug, and should report it separately.
UPDATE:
Try setting 'Option "AGPMode" "4"' to the Device section.
Wouter Stomp (wouterstomp-deactivatedaccount) wrote : | #1 |
Wouter Stomp (wouterstomp-deactivatedaccount) wrote : | #2 |
Wouter Stomp (wouterstomp-deactivatedaccount) wrote : | #3 |
These might also be the same problem:
http://
http://
Wouter Stomp (wouterstomp-deactivatedaccount) wrote : | #4 |
Btw, Usplash is displayed normally, gdm and gnome are scrambled
KIAaze (zohn-joidberg) wrote : | #5 |
I also have this problem with an ATI Radeon IGP 345M on a Compaq Presario 2500 laptop.
My screen is less scrambled (I can recognize the login screen and even log in), but it still makes the system unusable.
The splash screen works normally for me too and the display is normal on an external monitor.
http://
http://
Changed in xserver-xorg-video-ati: | |
status: | New → Confirmed |
mabovo (mabovo) wrote : | #6 |
Same to me with ATI Radeon 9600 using ati driver. Solved it replacing by fglrx 8.40.4 driver and reconfiguring Xorg.
Wouter Stomp (wouterstomp-deactivatedaccount) wrote : | #7 |
This also happens when running from the live cd (tribe 5).
Will Farrington (wcfarrington) wrote : | #8 |
I can confirm this with the 1.6.7.191 driver on Ubuntu Gutsy with a Radeon 7500.
Lil (vicky-lamburn-deactivatedaccount) wrote : | #9 |
Possibly the same issue on my Mobility Radeon 7500 in my T40. Gutsy boots fine and shows Gnome no problems. The tearing issue happens when I start playing with display-config-gtk such as changing the driver or configuring a second screen and then hitting Test. Completely jams up the system and only powering off with holding down the power button will shut it down.
I can't use the fglrx driver as it does not support the Radeon Mobility 7500.
raggar (mbaart) wrote : | #10 |
Have the same issue on my acer aspire 5600 with a x1300 ati card.
I also have trouble with loading the live cd, after booting I get a blank screen. Later on I get a warning that the x server didn't load the driver. So I can't install from the live cd desktop.
mabovo (mabovo) wrote : | #11 |
Last kernel update 2.6.22.11 using ati driver 6.7.192 and xserver-xorg 12ubuntu4 made a regression on Radeon 9600 cards and continues the status of "Screen displays garbage"
Changed in xserver-xorg-video-ati: | |
importance: | Undecided → High |
Bryce Harrington (bryce) wrote : | #12 |
I just spoke with Alex Deucher about this, and showed the screenshot. He's aware of the problem but doesn't have an idea about what may be causing it or how to fix it. It seems to have been an issue for a while in the -ati driver. It's probably unrelated to the kernel version (if you can prove otherwise, let us know).
Can people experiencing this problem please do the following tests?
1. Try disabling tiling by adding Option "ColorTile" "false" to your driver section.
2. Try disabling DRI by adding Disable "DRI" in your module section
3. Try both the above together
Also, please attach your /etc/X11/xorg.conf and /var/log/Xorg.0.log in each of the above cases. Also include the output of lspci -vvnn.
More screenshots/photos of screens showing the garbage could also be quite valuable, so please also attach those if you have them.
Changed in xserver-xorg-video-ati: | |
status: | Confirmed → Incomplete |
Wouter Stomp (wouterstomp-deactivatedaccount) wrote : | #13 |
Wouter Stomp (wouterstomp-deactivatedaccount) wrote : | #14 |
Wouter Stomp (wouterstomp-deactivatedaccount) wrote : | #15 |
Wouter Stomp (wouterstomp-deactivatedaccount) wrote : | #16 |
Tried those 3 tests, none of them fixed the problem (btw. there is no driver section as far as I know, I assume you meant the device section, which is what I tried?).
Wouter Stomp (wouterstomp-deactivatedaccount) wrote : | #18 |
Wouter Stomp (wouterstomp-deactivatedaccount) wrote : | #19 |
I don't think this is the same as the linked upstream bug freedesktop-bugs #11399. That bug is reported on breezy and feisty, while I had no problem on these. The problem for me seems to have started since the upload of the 1:6.6.193-1ubuntu1 ati driver (or around that time). Early in the Gutsy release cycle I didn't have the problem. Also on the screenshots in the upstream bug report I don't see the cursor anywhere, while as you can see on my screenshots the cursor is appearing crystal clear against a scrambled background.
Wouter Stomp (wouterstomp-deactivatedaccount) wrote : | #20 |
Downgrading xserver-
mabovo (mabovo) wrote : | #21 |
- The only diff noted are the missing lines on Xorg.0.log file attached Edit (49.5 KiB, image/jpeg)
I had to downgrade to 6.6.193-1ubuntu1 too.
mabovo (mabovo) wrote : | #22 |
- Xorg.0.log.missing_lines Edit (32.9 KiB, text/plain)
01:00.0 VGA compatible controller: ATI Technologies Inc RV350 AR [Radeon 9600] (prog-if 00 [VGA])
Subsystem: C.P. Technology Co. Ltd Unknown device 2079
Control: I/O+ Mem+ BusMaster+ SpecCycle- MemWINV- VGASnoop- ParErr- Stepping- SERR+ FastB2B-
Status: Cap+ 66MHz+ UDF- FastB2B+ ParErr- DEVSEL=medium >TAbort- <TAbort- <MAbort- >SERR- <PERR-
Latency: 64 (2000ns min), Cache Line Size: 64 bytes
Interrupt: pin A routed to IRQ 20
Region 0: Memory at d0000000 (32-bit, prefetchable) [size=256M]
Region 1: I/O ports at e000 [size=256]
Region 2: Memory at fbe00000 (32-bit, non-prefetchable) [size=64K]
Expansion ROM at fbd00000 [disabled] [size=128K]
01:00.1 Display controller: ATI Technologies Inc RV350 AR [Radeon 9600] (Secondary)
Subsystem: C.P. Technology Co. Ltd Unknown device 2078
Control: I/O+ Mem+ BusMaster+ SpecCycle- MemWINV- VGASnoop- ParErr- Stepping- SERR- FastB2B-
Status: Cap+ 66MHz+ UDF- FastB2B+ ParErr- DEVSEL=medium >TAbort- <TAbort- <MAbort- >SERR- <PERR-
Latency: 64 (2000ns min), Cache Line Size: 64 bytes
Region 0: Memory at e0000000 (32-bit, prefetchable) [size=256M]
Region 1: Memory at fbf00000 (32-bit, non-prefetchable) [size=64K]
mabovo (mabovo) wrote : | #23 |
Changed in xorg-server: | |
status: | Unknown → Confirmed |
mabovo (mabovo) wrote : | #24 |
Ati driver 1.6.7.192-1ubuntu2, xserver-xorg-core 2:1.3.0.
mabovo (mabovo) wrote : Re: Screen displays garbage solved disabling DRI | #25 |
- Xorg.0.log with DRI disabled and ColorTile false Edit (96.6 KiB, text/plain)
Radeon 9600 could recover from scratched screen with ati driver 1.6.7.192 following Bryce tips.
mabovo (mabovo) wrote : Re: Screen displays garbage | #26 |
mabovo (mabovo) wrote : | #27 |
magi (magnus-hagdorn-marsupium) wrote : | #28 |
magi (magnus-hagdorn-marsupium) wrote : | #29 |
I did some more experimentation and found that if I disable glx and dri support in the xorg.conf I can start the xserver albeit without 3D graphics support. Hope this indicates where the problem might be.
Wouter Stomp (wouterstomp-deactivatedaccount) wrote : | #30 |
How do you disable glx support?
Bryce Harrington (bryce) wrote : | #31 |
You can disable by adding this to your Modules section:
Section "Module"
...
Disable "glx"
Disable "dri"
EndSection
It's not sufficient to simply comment out the Load lines if they're there; because Xorg now auto-loads modules if not explicitly mentioned.
GiuseppeVerde (launchpad-digitasaru) wrote : | #32 |
My symptoms: on my notebook (Dell Inspiron 8600 w/ ATI Mobile Radeon RV350 (M10)), the new driver causes the LCD to be blank, with odd-colored vertical lines appearing somewhat randomly, and the entire screen fading very slowly to white).
Reverting back to xserver-
Disabling ColorTiling, GLX, and DRI did not help anything.
lspci output:
01:00.0 VGA compatible controller: ATI Technologies Inc RV350 [Mobility Radeon 9
600 M10] (prog-if 00 [VGA])
Subsystem: Dell Unknown device 2001
Control: I/O+ Mem+ BusMaster+ SpecCycle- MemWINV- VGASnoop+ ParErr- Step
ping- SERR+ FastB2B-
Status: Cap+ 66MHz+ UDF- FastB2B+ ParErr- DEVSEL=medium >TAbort- <TAbort
- <MAbort- >SERR- <PERR-
Latency: 32 (2000ns min), Cache Line Size: 32 bytes
Interrupt: pin A routed to IRQ 11
Region 0: Memory at d0000000 (32-bit, prefetchable) [size=256M]
Region 1: I/O ports at c000 [size=256]
Region 2: Memory at fcff0000 (32-bit, non-prefetchable) [size=64K]
Expansion ROM at fc000000 [disabled] [size=128K]
bit- FW+ AGP3- Rate=x1,x2,x4
1
-,D3cold-)
GiuseppeVerde (launchpad-digitasaru) wrote : | #33 |
Wouter Stomp (wouterstomp-deactivatedaccount) wrote : | #34 |
Disabling glx in xorg.conf fixes the problem for me.
Wouter Stomp (wouterstomp-deactivatedaccount) wrote : | #35 |
But desktop effects is not working now when I disable glx, while I had desktop effects working with the 6.6.3 ati driver.
Wouter Stomp (wouterstomp-deactivatedaccount) wrote : | #36 |
Btw. no need to disable dri, only glx is enough.
Changed in xserver-xorg-video-ati: | |
status: | Incomplete → Confirmed |
Changed in xorg-server: | |
status: | Confirmed → Fix Released |
Timo Aaltonen (tjaalton) wrote : | #37 |
.194 is now uploaded to be included post-beta.
Changed in xserver-xorg-video-ati: | |
status: | Confirmed → Fix Committed |
Wouter Stomp (wouterstomp-deactivatedaccount) wrote : | #38 |
I tried 6.7.194 from http://
Changed in xserver-xorg-video-ati: | |
status: | Fix Committed → Confirmed |
In freedesktop.org Bugzilla #12571, Timo Aaltonen (tjaalton) wrote : | #39 |
Forwarded from https:/
2007-08-17
Since the updates a few days ago on gutsy, my screen only shows garbage with only the mouse pointer shown normally. This is on a radeon 9250.
(the version which was uploaded mid-August is 6.6.193)
In freedesktop.org Bugzilla #12571, Timo Aaltonen (tjaalton) wrote : | #40 |
disabling glx makes the problem to go away.
In freedesktop.org Bugzilla #12571, Timo Aaltonen (tjaalton) wrote : | #41 |
In freedesktop.org Bugzilla #12571, Timo Aaltonen (tjaalton) wrote : | #42 |
6.7.194 is still broken for him.. logs are coming.
In freedesktop.org Bugzilla #12571, Michel-tungstengraphics (michel-tungstengraphics) wrote : | #43 |
(In reply to comment #1)
> disabling glx makes the problem to go away.
More likely what matters is the DRI, which is also disabled implicitly when GLX is disabled.
AFAICT nobody actually tried with tiling disabled - the option name is "ColorTiling".
In freedesktop.org Bugzilla #12571, Timo Aaltonen (tjaalton) wrote : | #44 |
Ok, I'll ask that.
Here is the log with glx enabled:
In freedesktop.org Bugzilla #12571, Timo Aaltonen (tjaalton) wrote : | #45 |
on trying ColorTiling:
"Tried that, doesn't help. Also tried only disabling DRI as mentioned in the upstream report, but that doesn't do anything either. Only disabling glx works."
In freedesktop.org Bugzilla #12571, Michel-tungstengraphics (michel-tungstengraphics) wrote : | #46 |
(In reply to comment #6)
> on trying ColorTiling: [...]
Saw that... again, the option name was spelt incorrectly in the downstream bug report, and the log file shows it misspelt in the same way and thus ineffective.
Timo Aaltonen (tjaalton) wrote : Re: Screen displays garbage | #47 |
Wouter: please attach the Xorg.0.log as well.
Changed in xserver-xorg-video-ati: | |
status: | Confirmed → Incomplete |
Timo Aaltonen (tjaalton) wrote : | #48 |
reported upstream.
Wouter Stomp (wouterstomp-deactivatedaccount) wrote : | #49 |
Changed in xserver-xorg-video-ati: | |
status: | Incomplete → Confirmed |
Timo Aaltonen (tjaalton) wrote : | #50 |
Thank you. Also, try setting :
Option "ColorTiling" "false"
for the driver.
Wouter Stomp (wouterstomp-deactivatedaccount) wrote : | #51 |
Tried that, doesn't help. Also tried only disabling DRI as mentioned in the upstream report, but that doesn't do anything either. Only disabling glx works.
Changed in xorg-server: | |
status: | Unknown → Incomplete |
Wouter Stomp (wouterstomp-deactivatedaccount) wrote : | #52 |
Tried correct spelling ColorTiling instead of ColorTile as mentioned upstream, doesn't work.
Timo Aaltonen (tjaalton) wrote : | #53 |
Wouter:
please attach logs with different settings:
-just glx disabled
-just dri disabled
-ColorTiling set false
Wouter Stomp (wouterstomp-deactivatedaccount) wrote : | #54 |
Wouter Stomp (wouterstomp-deactivatedaccount) wrote : | #55 |
Wouter Stomp (wouterstomp-deactivatedaccount) wrote : | #56 |
Wouter Stomp (wouterstomp-deactivatedaccount) wrote : | #57 |
- Xorg.0.nodri2.log Edit (47.1 KiB, text/plain)
OK, I just found out the problem is dri. I had Disable "DRI" in my xorg.conf, but it should be Disable "dri". Disabling dri fixes the problem for me. A new Xorg.0.log withouth dri attached.
KIAaze (zohn-joidberg) wrote : | #58 |
- logfiles for all 4 cases: 0=initial xorg.conf, 1=colortile false, 2=disable DRI, 3=1+2 Edit (45.8 KiB, application/x-tar)
Ok, I finally have some time to do this.
So here are all the logs.
Screen remained scrambled in all cases (I did ctrl+alt+backspace after each).
I put all logfiles (/etc/X11/
case 0 = initial xorg.conf
case 1 = colortile false
case 2 = disable DRI
case 3 = colortile false+disable DRI
KIAaze (zohn-joidberg) wrote : | #59 |
- more xorg.conf tests Edit (147.6 KiB, application/x-tar)
Ok, I did a few more tests using "ColorTiling" and "dri" instead of "ColorTile" and "DRI".
I also removed the comments on the DRI section at the end of xorg.conf.
None of the configurations solved the problem.
Screen is still slightly scrambled and has an offset to the left. (i.e. like this "bcdea")
I'll try to take a picture if I can get my hands on a camera.
KIAaze (zohn-joidberg) wrote : | #60 |
It seems DRI (direct rendering?) isn't working for me anymore.
After all those tests, I downgraded to the old driver again, and now all my games run slow.
glxinfo says no direct rendering and whenever I run glxgears, X crashes.
I did a lot of updates today since I haven't been able to do that for a month, so I don't know if it's still related to this ATI driver problem or something else.
Has there been any XGL update lately by any chance?
KIAaze (zohn-joidberg) wrote : | #61 |
And yes, I did remove the "Disable "dri"" in xorg.conf. I even tried out older xorg.conf backups.
In freedesktop.org Bugzilla #12571, Timo Aaltonen (tjaalton) wrote : | #62 |
Disabling dri is the only thing which made it work:
GiuseppeVerde (launchpad-digitasaru) wrote : Re: Screen displays garbage | #63 |
FWIW, the latest ATI driver in Gutsy works for me again.
Changed in xorg-server: | |
status: | Incomplete → Confirmed |
dmf86 (diogomiguelferreira86) wrote : | #64 |
Same issue here, @ Gutsy with all the updates and in a ATI Radeon 9200 Pro. Got X to wrok changing the driver to "Vesa", but without 3D of course...
mabovo (mabovo) wrote : | #65 |
Gutsy Beta fresh install on Radeon 9600 RV350 with Vesa installed as default. Forcing manually xorg.conf to work with latest ati driver and fighting against displayconfig-gtk to fix ati driver without success, finally Gutsy accepted ati driver after some crashes. Desktop effects now is enabled and Compiz plugins work like a charm.
Henrik Nilsen Omma (henrik) wrote : | #66 |
- logs.tar.gz Edit (12.4 KiB, application/x-tar)
I see the same problem on a Radeon Xpress 200 with the latest Gutsy. Various logs attached (sorry for the .tgz but there are about 10 log files).
I tried recovering with various GUI tools like displayconfig-gtk, without luck and had to resort to vesa in the end.
Wouter Stomp (wouterstomp-deactivatedaccount) wrote : | #67 |
Is reverting back to the feisty version of the ATI driver an option for Gutsy? It worked perfectly and works well with compiz too.
Tormod Volden (tormodvolden) wrote : | #68 |
Wouter, I believe the new ati driver fixes more than it breaks. Of course going back to the old driver would be an option for individual users like you. It seems like most of the other reports in this bug are different issues than yours though.
mabovo (mabovo) wrote : | #69 |
Stellarium causes a few symptoms of garbage in a small piece of the screen just when the program begins.
I am using the ati driver 1.6.7.194 TV6 with a Radeon 9600. If I disable desktop effects in Compiz this behavior do not happen.
mabovo (mabovo) wrote : | #70 |
- stellarium.config.ini.txt Edit (6.9 KiB, text/plain)
Stellarium runs perfect in a Radeon 8500 with CRT monitor.
In order to get it running with a Radeon 9600 with LCD monitor the [video] section of config.ini must be set to 24 bpp_mode and screen reolution to 1280x1024.
Quinn Reynolds (quinn-reynolds) wrote : | #71 |
I can confirm this occurs with the latest Gutsy Beta CD on my laptop, a Compaq NC6000 with ATI Mobility Radeon 9600, 1400x1050 lcd.
I can't get the LiveCD to run in either normal or safe mode without getting a screen full of gibberish as soon as X fires up.
Tommy (stateq2) wrote : | #72 |
I'm confirming this bug on a ATI Mobility Radeon 7500. Changing to the vesa driver in my xorg.conf "fixes" it, but obviously things are much slower.
dmf86 (diogomiguelferreira86) wrote : | #73 |
No fix yet? :( There's no way to adapt GG to the old ATI working driver of FF, or vice-versa?
Bryce Harrington (bryce) wrote : | #74 |
Unfortunately this bug did not get sorted out for Gutsy-final. However, since -ati had a lot of new code and is still being stabilized, we anticipate doing at least one hotfix for the -ati driver for Gutsy. Because it will be a hotfix, though, we need to be absolutely certain there is no chance of regressions for other users, so it may take some time to ensure the fix solves this problem and causes absolutely no others.
I suspect that most people commenting on this thread are having unrelated issues. Unfortunately, due to the bug's generic title "Screen displays garbage", I think that's causing all sundry of -ati issues to collect here. It is possible that some of these disparate issues have the same or similar root causes, but in general unless you experience exactly the same symptoms as Wouter, then it's likely you are having a distinct bug that should be reported separately.
mabovo (mabovo) wrote : | #75 |
Today last updates disabled desktop effects and couldn't recovery it working on xorg.conf and ati driver 1.6.7.195. There is not symptoms of garbage in the screen but I think other packages maybe are interfering with ati driver.
dmf86 (diogomiguelferreira86) wrote : | #76 |
We'll be awaintig :) Can't move to GG until there... :( Keep up the good work!
Bryce Harrington (bryce) wrote : | #77 |
dmf86, of course you can install and run the Feisty 6.6.3 -ati driver as a workaround for this issue. However, for Gutsy we wish to focus on attaining a true fix for the issue, without needing to revert to the old driver. -ati 6.7.x contains a wealth of bug fixes for long standing issues we wish to have, and even though it brings a few new issues such as this one, upstream has been quite responsive to bug reports and we anticipate putting in a request for doing an update with a newer -ati, perhaps after 6.8.0 is out.
dmf86 (diogomiguelferreira86) wrote : | #78 |
Take your time, do a good work. :) Sorry for the "rushing"! ;)
Aron van Ammers (aron-multitof) wrote : | #79 |
Just reproduced this bug in Gutsy with a 9200SE. For me disabling GLX solves it.
Unfortunately I have been looking for solutions for hours and I didn't find this thread until 15 minutes ago :(. It would have been helpful if the subject contained "radeon" or "ati".
Anyway, I'm a little happier because at least I can run my normal resolution again. Now all I'm waiting for is for GLX/DRI to work so I can run Compiz again! And as with many Linux issues, it takes ages to solve it but you learn a lot on the way :) (in this case about configuring Xorg et al).
Aron van Ammers (aron-multitof) wrote : | #80 |
- Xorg.0.log Edit (50.5 KiB, text/plain)
Installing the Feisty driver (6.6.3) does not solve the issue for me with a 9200SE. When GLX/DRI are enabled I still get garbled display.
JW (arch0njw) wrote : Re: [rv280] Screen displays garbage, mouse cursor normal (regression) | #81 |
"me too"
me@machine:~$ uname -a
Linux machine 2.6.22-14-generic #1 SMP Sun Oct 14 23:05:12 GMT 2007 i686 GNU/Linux
00:00.0 Host bridge: ATI Technologies Inc RS200/RS200M AGP Bridge [IGP 340M] (rev 02)
00:01.0 PCI bridge: ATI Technologies Inc PCI Bridge [IGP 340M]
01:05.0 VGA compatible controller: ATI Technologies Inc Radeon IGP 330M/340M/350M
I've followed bunch of the above suggestions with no positive results. I tried running the Feisty LiveCD and the ati driver was used successfully.
Is there any other information that would be helpful?
Bryce Harrington (bryce) wrote : Re: [ati radeon rv280] Screen displays garbage, mouse cursor normal (regression) | #82 |
JW, please see http://
In general, "me too" posts are discouraged, and you should strive to add some piece of value to the discussion. At a minimum you must provide proof that you are indeed seeing the exact same issue as the original poster, such as matching screenshots, similar logs, or that the workarounds solved it for you as well.
You've indicated that the workarounds people posted above did not give you positive results, which would tend to suggest that you're not seeing Wouter's bug at all, but something completely different.
description: | updated |
Bryce Harrington (bryce) wrote : Re: [ati radeon rv280] Screen displays horizontal streaking, mouse cursor normal; worked okay with Feisty | #83 |
We've posted a new 1:6.7.196-1 -ati driver to Hardy. Hardy Alpha-1 will be released some time later this week, if someone would like to test if it has a fix. I reviewed the changelog and none of the changes appear relevant to this bug, so my guess is that it won't have any effect, but would still be of interest to know.
Wouter Stomp (wouterstomp-deactivatedaccount) wrote : | #84 |
Just tested it, indeed no effect.
description: | updated |
Tormod Volden (tormodvolden) wrote : | #85 |
Are all the affected cards AGP cards? Please try:
Option "BusType" "PCI"
in the Device section of xorg.conf
JW (arch0njw) wrote : | #86 |
Per request, I've filed another bug here:
https:/
The steps here do not work for me, so the assertion is that I am experiencing a different issue.
Thanks for the pointers, Bryce!
Joachim Noreiko (jnoreiko) wrote : | #87 |
I'm seeing this with the onboard graphics on a Foxconn K8M890M2MB -- though for me the effect doesn't show on a screenshot.
Wouter Stomp (wouterstomp-deactivatedaccount) wrote : | #88 |
Great news! Option "AGPMode" "4" fixes the problem.
Bug 181602 and Bug 180343 might be duplicates of this bug?
In freedesktop.org Bugzilla #12571, Timo Aaltonen (tjaalton) wrote : | #89 |
This bug is actually about the AGPMode mess that's going on :) Setting it to "4" works for the reporter.
description: | updated |
In freedesktop.org Bugzilla #12571, Michel-tungstengraphics (michel-tungstengraphics) wrote : | #90 |
Unfortunately, as far as we can tell it's just impossible to always pick a default value for AGPMode that works, and in general leaving it unchanged from what the BIOS set up seems most reliable. I'm afraid it looks like the cases where that doesn't work need to change it in the BIOS setup or xorg.conf.
In freedesktop.org Bugzilla #12571, Bugzi09-fdo-tormod (bugzi09-fdo-tormod) wrote : | #91 |
If I understand the current code correctly, the default is to leave it unchanged for v3 cards, but to use x1 for v2 cards. This breaks for some v2 cards. I think the old behaviour of leaving v2 cards unchanged also broke on some cards.
In freedesktop.org Bugzilla #12571, Michel-tungstengraphics (michel-tungstengraphics) wrote : | #92 |
Sure, this bug is about V3 though, but the same problem also occurred when we defaulted to 4x for that. Right now it's looking like leaving it unchanged is most reliable overall, and we should probably switch back to that for non-V3 as well.
magi (magnus-hagdorn-marsupium) wrote : | #93 |
yup, I can confirm this as well. Excellent, it works!
Denis_AeC (denis.romand-aliceadsl) wrote : | #94 |
I just tested Hardy alpha 3 cd (and live usb), but it does not work : as soon as session is started, the screen is no more readible (as in first screen shot above), and the mouse is still correct and can be moved. For Gutsy I already filled a bug (that I can't find now ??) that has been linked to this one.
Video card is ATI RV280 [Radeon 9200 SE] on PCI.
So, for the moment, Feisty : OK, Gutsy : KO, Hardy alpha3 : KO
Aron van Ammers (aron-multitof) wrote : | #95 |
I can confirm the AGPMode 4 workaround as well on Radeon 9200 SE. Now happily using Compiz on Gutsy with the normal drivers, finally :).
Denis, does the AGPMode workaround work for you with those live CD's? If so, maybe the component that detects hardware and generates xorg.conf should take it into account.
Denis_AeC (denis.romand-aliceadsl) wrote : | #96 |
- xorg.conf - Valid Edit (1.1 KiB, text/plain)
Operations :
- boot on live usb hardy alpha 3
- messy screen when session starts
- ctl (?) alt F1 => console session
- look at xorg.conf : not a "classic" one ???
- can not see where to put agpmode !
Normally, the LiveCD is supposed to boot correctly by itself ;-) Users should not have to modify anything in configuration to make it work !
Timo Aaltonen (tjaalton) wrote : | #97 |
The official word from upstream is that the change to default AGPMode to "1" was a mistake, and it should be changed back to "4", since it seems to be more common. There is no reliable way to automatically get the right mode for those where it fails, so the only option for those is to either modify xorg.conf or play with BIOS settings.
Tormod Volden (tormodvolden) wrote : | #98 |
Upstream has now changed to just use (as default) the mode that was set up by the firmware/BIOS. Test packages at https:/
Changed in xserver-xorg-video-ati: | |
status: | Confirmed → Fix Committed |
Wouter Stomp (wouterstomp-deactivatedaccount) wrote : | #99 |
I tried with xserver-
Tormod Volden (tormodvolden) wrote : | #100 |
Wouter, so AGPMode 4 fixes the problem, but the new driver does not? Please attach your new Xorg.0.log.
Tormod Volden (tormodvolden) wrote : | #101 |
Wouter, I just looked at your old log and realized that you have an AGP v3 card (8x or 4x). The upstream change is for v2 cards (4x 2x 1x).
Changed in xserver-xorg-video-ati: | |
status: | Fix Committed → Confirmed |
KIAaze (zohn-joidberg) wrote : | #102 |
It's working! :)
I currently have those installed:
xserver-
xserver-xorg 1:7.2-5ubuntu13
libc6 2.7-5ubuntu2
And my screen works correctly. I didn't even have to change my xorg.conf.
Direct rendering works too.
Desktops effects still can't be enabled (beryl used to work) and I still have some graphical glitches in some openGL games and screenlets with black borders. But I'm not sure if those problems are related to the driver.
To upgrade to the new ATI driver, I just replaced gutsy with hardy in the sources.list and upgraded the 3 mentioned packages without doing a full dist-upgrade.
But I'll probably be reinstalling soon anyway and then I'll do a full upgrade to hardy.
KIAaze (zohn-joidberg) wrote : | #103 |
Don't know if this is related to this bug directly anymore, but:
After upgrade to hardy:
-display working
-direct rendering working
-Compiz working!!! (never worked before)
-Even some games in wine working that didn't work before (but this might be due to some wine update or other)
Please keep it that way for the final release. :)
Wouter Stomp (wouterstomp-deactivatedaccount) wrote : | #104 |
So what is the plan for hardy for AGP v3 cards? Could agpmode be set to 4 by default, solving this problem?
Timo Aaltonen (tjaalton) wrote : | #105 |
Wouter: no, unfortunately there is no way to make it work for everybody, this is confirmed by upstream. Changing it to 4 would break it for others like we've seen during Hardy, before 6.8.0-1
Denis_AeC (denis.romand-aliceadsl) wrote : | #106 |
I downloaded alpha5 CD and tested it on my pc : same thing as usual, display doesn't work when booting on cd. On another PC, it's OK. It is frustrating as I had no problem on Feisty. I dropped Gutsy because I didn't succeed in having 3D working, after having difficulties to get a proper video. "My" bug :
https:/
So what about Hardy ? Can I hope to have a CD working on my PC ? Must I install Hardy and try to modify xorg or other things to have correct display ? Will I succeed in having 3D ?
Timo wrote : "no, unfortunately there is no way to make it work for everybody, this is confirmed by upstream. Changing it to 4 would break it for others like we've seen during Hardy, before 6.8.0-1"
??? Does that mean that there is no way to have a correct display for every PC ? How do others linux distribution perform ? It is the first since many years that a boot CD does not work properly on one of my PCs ...
At least, a list of malfunctionning cards and a clear tutorial on how to get video cards properly working should be widely dispatched. (For me, operations that I did for modifying xorg.conf are clearly linked to sorcery ;-) )
Timo Aaltonen (tjaalton) wrote : | #107 |
Denis_AeC: maybe I was just confused.. as Tormod said, upstream reverted the change that broke 6.7.197 for some users, and the current driver should be better. Alpha5 did not have 6.8.0.
Wouter Stomp (wouterstomp-deactivatedaccount) wrote : | #108 |
Timo: then how is it possible it worked fine on feisty and earlier? Or where there similar reports back then?
Wouter Stomp (wouterstomp-deactivatedaccount) wrote : | #109 |
And was the default ever 4 during Hardy? Before 6.8.0-1? It never worked for me, so that seems odd...
Timo Aaltonen (tjaalton) wrote : | #110 |
Wouter: not that I know of, but you should just try a daily livecd to find out if it works :)
Wouter Stomp (wouterstomp-deactivatedaccount) wrote : | #111 |
I know how to try and I regularly do so, but it was never fixed. What was the situation during feisty and earlier? Where there any bug reports bag then for agpv3 cards? And is it known exactly which configurations are affected by this problem?
Timo Aaltonen (tjaalton) wrote : | #112 |
I don't know the history.
Have you tried the latest hardy or not? 6.8.0 was uploaded after alpha5. If you have tried, please attach your Xorg.0.log as was requested.
Denis_AeC (denis.romand-aliceadsl) wrote : | #113 |
- Second xorg.conf, no modif - Valid Edit (1.2 KiB, text/plain)
Hardy : not OK on Beta, but ...
As soon as Hardy Beta appeared, I downloaded it and gave it a try on my PC :
- boot on liveusb : crazy video as "usual" when session starts
- ctl alt F1, stop gdm, xorg.conf modif, start gdm : fail, try to remove lock : no lock, start gdm again : ok and then correct video, directly in 1280*1024. But when looking in xorg.conf, I saw that I put an erroneous QGPMode ...
So, second test :
- boot on liveusb : crazy video as "usual" when session starts
- ctl alt F1, stop gdm, start gdm : fail, start gdm again : ok and then correct video, directly in 1280*1024. No problem to start applications, download packages from web, ...
So, when booting on CD or liveUSB, if it could immediately be correct, it would be GREAT !
My video card is ATI RV280 Radeon 9200 SE
Flat screen is Hyundai L70S+ 1280*1024
"My" bug : <https:/
Denis_AeC (denis.romand-aliceadsl) wrote : | #114 |
- Current xorg.conf : quite minimal ! - Valid Edit (1.2 KiB, text/plain)
Hardy Beta again , but installed on PC :
After having tested Hardy Beta on LiveUSB (see previous post), I installed it on my home PC. This of course required to perform the steps as described in previous post to have a liveUSB session with a correct display. But then, I started installation that ran without problem up to reboot.
- Reboot after installation : correct up to session start, and then bad video as "usual"
- ctl alt F1, stop gdm, start gdm : no change in video !
- ctl alt F1, stop gdm, xorg.conf modification (adding Option "AGPMode" "4") start gdm : OK, video correct. And then, application installation, grub & fstab modifications, reboot, ... and now, compiz activation with cube and effects : no problem !
So, Hardy is OK for me, but installation is not as good as expected and this video problem will probably block some (many ?) people.
Timo Aaltonen (tjaalton) wrote : | #115 |
Denic_AeC: try forcing the AGPMode from BIOS, and commit out the option from xorg.conf. Upstream thinks that the driver cannot be made to work right for everyone, and the only choice is to force it from BIOS.
chkur (chkur) wrote : | #116 |
- xorg.conf - Valid Edit (1.2 KiB, text/plain)
I get this problem on Ubuntu 8.04 and LinuxMint 5 on my ATI IGP 320. When I try to boot normally from LiveCD, I get blank screen. When I boot in "compatibility mode" I get resolution 1280x1024x60Hz instead of 1024x768x85Hz. If I change resolution to 1024x768 I get picture as described here. When I installed Ubuntu on hard drive I get the same. I tried to disable dri and glx in xorg.conf but problem is still here. My videocard is integrated so I can't force AGPMode to make a problem go away. I shall attach my xorg.conf and Xorg.0.log from LiveCD - booting in normal and compatibility mode.
chkur (chkur) wrote : | #117 |
chkur (chkur) wrote : | #118 |
chkur (chkur) wrote : | #119 |
chkur (chkur) wrote : | #120 |
Changed in xserver-xorg-video-ati: | |
status: | Confirmed → Triaged |
Wouter Stomp (wouterstomp-deactivatedaccount) wrote : | #121 |
Apparently quite a few people with a variety of ATI video cards are affected by this. Is there any data on which configurations would working and broken with which setting of the agpmode? And how many people use these?
Bryce Harrington (bryce) wrote : | #122 |
Hi Wouter and everyone,
I talked with upstream a bit about this AGPMode issue. It seems extremely hardware-specific as to what AGP Mode to use. So I coded up a quirk system that will let us build a table in the driver of ATI video cards vs. host bridges, and the AGP Mode they need. The data on the configurations I need for setting this is:
* AGPMode value needed
* Make/Model of laptop or motherboard
* Output of lspci -vvnn
* Is the system all factory hardware, or have any parts been replaced?
* Is there an AGP Mode in the system BIOS?
- If so, is it set to the factory default?
You can test various values of AGPMode by setting the option in your xorg.conf's Device section like this:
Section "Device"
...
Option "AGPMode" "2"
EndSection
Possible values are 1, 2, 4, and 8.
For more details about this issue and the quirk, I've written a section about it in this page: https:/
Wouter Stomp (wouterstomp-deactivatedaccount) wrote : | #123 |
Thanks for again looking into this! My motherboard died a few days ago and I decided to buy a new computer, so I won't be able to provide much more additional information. But the things you asked:
ATI Radeon 9250 AGP 128mb, Model name: 9250AD128BY
Motherboard: ASRock K7VT4A+ (printed on the motherboard itself it says AGP 8X)
AGP Mode needed to function: 4
lspci output is in one of the first comments of this bug
All factory hardware, nothing replaced
There are some agp settings in the bios, never touched them.
Because of the broken motherboard I won't be able to do more testing.
There are several more reports in the forums about this, where setting the agp mode fixed the problem, it might be worthwhile to request feedback there as well.
Bryce Harrington (bryce) wrote : | #124 |
Wouter, ok great thanks for letting me know that data. I'll put in a quirk for it.
If you wouldn't mind posting about this to the forums, that could be helpful. See my comment #109 for the info required.
Bryce Harrington (bryce) wrote : | #125 |
Changed in xserver-xorg-video-ati: | |
assignee: | nobody → bryceharrington |
status: | Triaged → In Progress |
KIAaze (zohn-joidberg) wrote : | #126 |
- lspci-vvnn.log Edit (8.9 KiB, text/plain)
Everything is working fine for me now. No more scrambled screen. (this bug) No more multiple loading bars and unaccessible virtual terminals. (bug #245849 & bug #253243).
And direct rendering is still there.
I haven't tried out different AGP settings yet. AGP mode 4 works well for me.
Here's my system info:
* ubu #133192
AGPMode needed: 4
Model: Compaq Presario 2500
Host Bridge: [0600]: ATI Technologies Inc RS200/RS200M AGP Bridge [IGP 340M] [1002:cbb2] (rev 02)
Graphics Card: [0300]: ATI Technologies Inc Radeon IGP 330M/340M/350M [1002:4337]
HW changes: None (except some added RAM)
BIOS: (Unknown)
KIAaze (zohn-joidberg) wrote : | #127 |
- xorg.conf - Valid Edit (2.6 KiB, text/plain)
Just in case, here's my xorg.conf & other info too.
uname -a:
Linux my-laptop 2.6.27-2-generic #1 SMP Thu Aug 28 17:20:02 UTC 2008 i686 GNU/Linux
apt-cache policy xserver-
xserver-
Installed: 1:6.9.0+
Candidate: 1:6.9.0+
Version table:
*** 1:6.9.0+
500 http://
100 /var/lib/
JW (arch0njw) wrote : | #128 |
- 1133192-lspci.out Edit (8.6 KiB, text/plain)
AGPMode value needed
>>> not using one -- will test modes later and see if that changes anything.
Make/Model of laptop or motherboard
>>> HP Pavilion ZE5500
Output of lspci -vvnn
>>> attached
Is the system all factory hardware, or have any parts been replaced?
>>> Added memory to take it to 1GB (was 512MB)
Is there an AGP Mode in the system BIOS?
>>> No. I have also not upgraded the BIOS as it is a pain in the butt under Linux -- maybe a new version of the BIOS would have this.
JW (arch0njw) wrote : | #129 |
Steve Langasek (vorlon) wrote : | #130 |
freeze exception granted for the agp quirk mode patch, please upload.
Wouter Stomp (wouterstomp-deactivatedaccount) wrote : | #131 |
In 1:6.9.0+
Bryce Harrington (bryce) wrote : | #132 |
Wouter, the uploaded fix probably would only fix your issue (or anyone else with identical hardware to you).
Anyway, I'm going to go ahead and assume it's fixed. You can reopen if it reoccurs.
Changed in xserver-xorg-video-ati: | |
status: | In Progress → Fix Released |
In freedesktop.org Bugzilla #12571, agd5f (agd5f) wrote : | #133 |
if this is still an issue with the driver from git, we can add a quirk for your card/gpu combo.
Michal Suchanek (hramrach) wrote : | #134 |
I have lockups with recent Debian testing on a K8 board with Radeon 9250.
Earlier the card had to be run in 4x mode although the only option in BIOS is to run 8x.
I just removed the card to save myself the headache.
The mainboard BIOS probably sucks.
Tormod Volden (tormodvolden) wrote : | #135 |
Many bugs were marked as duplicates of this bug, but apparently not all the reported issues have been fixed - this is because we need to add a workaround for each card/motherboard combination.
If you still need to specify AGPMode manually on your system, please see https:/
Changed in xorg-server: | |
importance: | Unknown → Medium |
Changed in xorg-server: | |
importance: | Medium → Unknown |
status: | Confirmed → Fix Released |
Changed in xorg-server: | |
importance: | Unknown → Medium |
Many people with ATI graphics cards report the same or similar problems:
http:// ubuntuforums. org/showthread. php?t=526947 blogs.sun. com/richb/ entry/latest_ ubuntu_ gutsy_updates_ blew
http://