[RV280 9250] No signal to monitor in 8.04 unless NoDRI used

[Simone Navari]

Binary package hint: xserver-xorg-video-ati

I've a problem using Ati Radeon 9250 on Ubuntu 8.04 with open source driver: very often, after splash screen monitor shows me "no signal" message and it goes to standby mode. I checked Xorg.0.log when xserver starts correctly and when it doesn't and I've seen two differences:
1) Some memory area (Ring, GART, Vertex buffers,etc...) are mapped to different addresses
2) When "no signal" message is shown log ends with

(II) RADEON(0): no multimedia table present, disabling Rage Theatre.

while when X starts without problems the log continues.

I attach Xorg.0.log in both cases: working and bugged, and my xorg.conf.

Best Regards,
Simone Navari.
[lspci]
00:00.0 Host bridge [0600]: VIA Technologies, Inc. K8T800Pro Host Bridge [1106:0282]
     Subsystem: ASUSTeK Computer Inc. A8V Deluxe [1043:80a3]
01:00.0 VGA compatible controller [0300]: ATI Technologies Inc RV280 [Radeon 9200 PRO] [1002:5960] (rev 01) (prog-if 00 [VGA controller])
     Subsystem: Giga-byte Technology Unknown device [1458:4054]

[Simone Navari (simnav)]

Please excuse me, I forgot some details: after "no signal" message is shown, monitor stay in standby mode even if I try to switch to any virtual console (Ctrl+Alt+F1,F2,...). Pressing caps key on keyboard led toggles and if I press Ctrl+Alt+Del system reboots so pc seems to be not hanged. Trying to reset X pressing Ctrl+Alt+Breakspace has no visible reaction.

Simone Navari.

[Bryce Harrington (bryce)]

Please attach the output of `lspci -vvnn`. Also, for no signal issues, please try the steps outlined in the following troubleshooting guide:
https://wiki.ubuntu.com/X/Troubleshooting/Resolution

[Simone Navari (simnav)]

As reqeuested I'll attach lspci output. Meanwhile I found a way to use my system simply putting option NoDRI in xorg.conf: no signal message doesn't appear anymore but system is slower and desktop effects don't work.

Simone Navari.

[Bryce Harrington (bryce)]

radeontool in Jaunty alpha-3* can be used to assist in debugging this issue. After installing it, you run it like this:
    radeontool regmatch '*' > regdump_good.txt
    radeontool regmatch '*' > regdump_broke.txt

Run it two times. Once when you have a working screen (for any driver), and once in the broken case (either from the tty console or logged into the sick box remotely). Attach both of those to this bug report, and we can then forward this issue upstream. Thanks ahead of time!

[* Note: radeontool is available in earlier versions of Ubuntu but is too old; if you need to run it on an older version of Ubuntu, you can obtain and build it from the upstream git tree at http://cgit.freedesktop.org/~airlied/radeontool/. Or maybe just try copying the executable from the jaunty system to the older one; I've not tried it but it might work.]

[Bungeur (sanfton-0801)]

I have this one too. Asus K8V SE M/B; Asus graphics card Radeon XPRESS 9250 128M. Kubuntu 8.10. Exact same crashing and cut-off Xorg log behaviour.

In Bug #141551 I read about AGP bus modes causing such problems with other Radeons, so I experimented with that.

System BIOS only lets me choose AGP8X or 4X. I set it to 4X, and in xorg.conf added "AGPMode" "2". This still fails, but examination of the truncated Xorg.0.log shows why: when it tries to set 2X, it reports ILLEGAL MODE - forcing AGP 4X

next I tried:
"BusType" "PCI", and sure enough it works. the log reports:
(--) RADEON(0): Chipset: "ATI Radeon 9250 5960 (AGP)" (ChipID = 0x5960)
(--) RADEON(0): Linear framebuffer at 0x00000000e8000000
(II) RADEON(0): AGP card detected
(**) RADEON(0): Forced into PCI mode

Sadly, like other users in Bug #141551, this slows everything down very badly.

The VESA driver works perfectly, but of course I am looking for the extra features of radeon driver!

[Bungeur (sanfton-0801)]

more observations:
1. I think this is same as bug 141551. This was closed with Quirks to detect various radeon types and force AGPmodes.
https://bugs.launchpad.net/ubuntu/+source/xserver-xorg-video-ati/+bug/141551/comments/55

But, the when chipID=5960 (RV280, XPRESS 9250) is detected, we have another problem: radeon driver does not try to do AGPmode change - see attached log, showing default AGP8X is used.

2. This chip does not accept AGP2X setting, only AGP4X or 8X

3. Xserver fails just after setting up textured video (see log). When it is working OK - with "BusType" "PCI" setting - textured video is followed by

init memap

in the Xorg log.

could this mean it is crashing while setting up the GART? (presume no GART in PCI mode)

4. fault happens from LiveCD. can't boot kubuntu unless failsafe graphics used.

5. hardware works fine, many hours uptime on VESA driver and windows-ME. Nice quiet (fanless) card.

[Bryce Harrington (bryce)]

[Problem]
Blank screen during X boot on Ati Radeon 9250. Goes away if option NoDRI used.

[Original Report]
I've a problem using Ati Radeon 9250 on Ubuntu 8.04 with open source driver: very often, after splash screen monitor shows me "no signal" message and it goes to standby mode. I checked Xorg.0.log when xserver starts correctly and when it doesn't and I've seen two differences:
1) Some memory area (Ring, GART, Vertex buffers,etc...) are mapped to different addresses
2) When "no signal" message is shown log ends with

(II) RADEON(0): no multimedia table present, disabling Rage Theatre.

while when X starts without problems the log continues.

I attach Xorg.0.log in both cases: working and bugged, and my xorg.conf.

I found a way to use my system simply putting option NoDRI in xorg.conf: no signal message doesn't appear anymore but system is slower and desktop effects don't work.

Best Regards,
Simone Navari.
[lspci]
00:00.0 Host bridge [0600]: VIA Technologies, Inc. K8T800Pro Host Bridge [1106:0282]
     Subsystem: ASUSTeK Computer Inc. A8V Deluxe [1043:80a3]
01:00.0 VGA compatible controller [0300]: ATI Technologies Inc RV280 [Radeon 9200 PRO] [1002:5960] (rev 01) (prog-if 00 [VGA controller])
     Subsystem: Giga-byte Technology Unknown device [1458:4054]

[Bryce Harrington (bryce)]

Created an attachment (id=22395)
radeontool output (failure)

[Bryce Harrington (bryce)]

Created an attachment (id=22396)
radeontool output (success)

[Bryce Harrington (bryce)]

Created an attachment (id=22397)
xorg.conf from failure case

[Bryce Harrington (bryce)]

Created an attachment (id=22398)
xorg.conf from working case

[Bryce Harrington (bryce)]

This is forwarded from the following Ubuntu report, btw:
https://bugs.launchpad.net/bugs/293698

[Bryce Harrington (bryce)]

Thanks, I've forwarded this bug upstream to https://bugs.freedesktop.org/show_bug.cgi?id=19842 - Please subscribe to this bug report in case upstream needs further information or wishes you to test something. Thanks ahead of time!

Bungeur, can you also post your lspci -vvnn info, so we can verify that you do indeed have exactly the same card. Your output should match this VGA portion exactly:

01:00.0 VGA compatible controller [0300]: ATI Technologies Inc RV280 [Radeon 9200 PRO] [1002:5960] (rev 01) (prog-if 00 [VGA controller])
     Subsystem: Giga-byte Technology Unknown device [1458:4054]

If you have a different number than 1458:4054, then we'll need you to also provide register dumps as Simone did, in case there are differences with your card.

[Michel Dänzer (michel-daenzer)]

Please also attach the full Xorg.0.log for both cases (or at the very least a diff between them) and set the MIME type to text/plain.

Also, whenever the log file appears truncated, the X server's stderr output needs to be examined for possible more information. With a display manager it should be captured in the display manager's log file.

[agd5f (agd5f)]

Does setting the AGPMode option help? Add the following to the device section of your config:

Option "AGPMode" "x"
where x = 1 or 2 or 4 or 8.

Try and find what AGP mode works. If none of the AGP modes help, try:
Option "BusType" "PCI"

[Bungeur (sanfton-0801)]

(In reply to comment #7)
> Does setting the AGPMode option help? Add the following to the device section
> of your config:
>
> Option "AGPMode" "x"
> where x = 1 or 2 or 4 or 8.
>
> Try and find what AGP mode works. If none of the AGP modes help, try:
> Option "BusType" "PCI"
>

I have almost exact same bug. the interesting thing about AGPmode is that the driver sets this to AGP8x by default. I think earlier drivers set AGP1x for safety.

If I set AGP2x, the driver responds (in xorg.0.log) something like: "illegal mode. AGP8x and 4x only allowed". Using AGP4x crashes the same as 8x.

setting Option "BusType" "PCI" works fine, including the KDE4.1 compositing effects. But, in this mode the PC seems a lot slower, notably the "bouncing K" of the KDE hourglass, window repaints etc. But hours of uptime show no crashes.

I will run the radeontool when I am back at my kubuntu PC later and post logs.

[Bungeur (sanfton-0801)]

(In reply to comment #6)
> Please also attach the full Xorg.0.log for both cases (or at the very least a
> diff between them) and set the MIME type to text/plain.
>
> Also, whenever the log file appears truncated, the X server's stderr output
> needs to be examined for possible more information. With a display manager it
> should be captured in the display manager's log file.
>

(In reply to comment #6)
> Please also attach the full Xorg.0.log for both cases (or at the very least a
> diff between them) and set the MIME type to text/plain.
>
> Also, whenever the log file appears truncated, the X server's stderr output
> needs to be examined for possible more information. With a display manager it
> should be captured in the display manager's log file.
>

[Bungeur (sanfton-0801)]

Created an attachment (id=22468)
Xorg.og for failure case (AGP)

[Bungeur (sanfton-0801)]

Created an attachment (id=22469)
Xorg.log for Working case (BusType PCI)

[Bungeur (sanfton-0801)]

Created an attachment (id=22470)
xorg.conf for working case (uncomment AGPMode line for Failure case)

[Bungeur (sanfton-0801)]

Created an attachment (id=22471)
lspci -vnn output from PC under test

struggling with jaunty radeontool in Kubuntu Intrepid. binary seems not to work, no success with source yet (only 4 weeks into Linux experience!) tipps for compiling from git appreciated.

[Michel Dänzer (michel-daenzer)]

I don't see anything but two copies of comment #6 in comment #9, did you mean to add anything there?

The log file from the failure case does look truncated, please check the X server stderr output.

[Bungeur (sanfton-0801)]

Created an attachment (id=22476)
stdout, stderr from startx (recovery root prompt)

output from staring recovery mode, root prompt, startx > /data/all 2>&1

I wanted to say in the previous posts, that my Xorg.0.log always stops at the same place (just after textured video is set up). It does not matter which AGPMode is selected, or if AGPFastWrite is yes or no.

with BusType PCI, no problems except slightly slow PC, and occasional scrambled windows when first opened (previews or K menus)

[Michel Dänzer (michel-daenzer)]

(In reply to comment #15)
> output from staring recovery mode, root prompt, startx > /data/all 2>&1
>
> I wanted to say in the previous posts, that my Xorg.0.log always stops at the
> same place (just after textured video is set up). It does not matter which
> AGPMode is selected, or if AGPFastWrite is yes or no.

There's no additional information in there... when the problem occurs, can you still ping the machine? Log in remotely?

[Bungeur (sanfton-0801)]

(In reply to comment #16)
> (In reply to comment #15)
> > output from staring recovery mode, root prompt, startx > /data/all 2>&1
> >
> > I wanted to say in the previous posts, that my Xorg.0.log always stops at the
> > same place (just after textured video is set up). It does not matter which
> > AGPMode is selected, or if AGPFastWrite is yes or no.
>
> There's no additional information in there... when the problem occurs, can you
> still ping the machine? Log in remotely?
>

The power button turns OFF the PC, but nothing else (key presses or CTRL-SH-BACKSP) seems to cause any action at all from the HDD-light. in fact, when X starts the screen gives 'no signal' and no more HDD light after that.

I have another PC with Kubuntu Intrepid here. will that do for remote login? I have not done remote access before though.

[Bungeur (sanfton-0801)]

Created an attachment (id=22491)
X errors from var/log/kdm.log See end of the file.

Found something in this one. I imagine this is the KDE log - and it reports some X errors right at the end of the file. This file was copied during a root prompt session, right after starting X in the faulty condition, and entering the login password blindly. Then, I gave it CTRL-Sh-BACKSP...

[josephjoecastle (josephjoecastle)]

hola... estuve mirando y leyendo la ayuda de la persona mencionada (http://albertomilone.com/index.html), y tambien la tuya, instale envyng... segui los pasos que das, instalo perfecto, o sea sin errores en el terminal, pero cuando lo reinicie se quedó bloqueado en la ventana de arranque de ubuntu.. en la de la rayita que carga...y la pantalla se dividio en 4, en la superior izquierda y superior derecha quedo la imagen congelada de ubuntu y la rayitade carga... en las inferiores derecha e izuqierda aparecen rayas de colores... como puedo recuperarel sistema, o como instalo bien mi ati... es una 9250...

Hi i was looking and reading the help of (http://albertomilone.com/index.html), and i follow your help, I Installed enyng.. and follow the steps given, it installed perfect, i mean without mistakes in the terminal, buet when i reboot i was frozen in the launch window of ubuntu, where it have a line loading.. anfd then rhis window was divided in four widnows, rigth upper and left upper had an image of ubuntu line, and the other tow.. had lines of many colors.. can i repair my ubuntu?, how can i install correctly with out mistakes my ATI... it is ATI RADEON 9250, this machine is PIV, IBM, processor.. 2.4gh cache 533, 1 giga ram, disk 300 gigas,

Please help me, thanks.

[agd5f (agd5f)]

Is this still an issue with xf86-video-ati from git master?

[Bungeur (sanfton-0801)]

(In reply to comment #19)
> Is this still an issue with xf86-video-ati from git master?
>

With Kubuntu Karmic, this one appears to be fixed. My hardware is exactly unchanged, as described below.

The Live CD works without having to choose VESA graphics mode, and the system works reliably with a fresh install.

Actually I upgraded from Intrepid, which worked fine for a couple of days, them blew up spectacularly!

The occasional scrambled video output (eg when a menu first pops up) which was present in Intrepid, is also fixed.

Only thing is, the compositing refuses to engage. I think I am happier with 9.10 overall though!

Unless the compositing is a related issue, I would call this one resolved.

[Bryce Harrington (bryce)]

[According to discussion on the upstream bug, this is no longer an issue in Ubuntu]

[vito (ovitorio)]

I'd rather post this on the upstream bug but Bryce's link to it looks broken.

Well, I am afraid it is still an issue in Ubuntu. More of this, considering the date (2009-12-19) Bryce made his comment of a fix released, it looks like the this fix actually brought this problem to my PC.

I have Asus P4PE motherboard, ATI Radeon 9550 (RV350 AS) graphic card and Ubuntu 8.04 installed.

I've never experienced this problem until recently. And I am running PC with the mentioned above configuration from the date Hardy was released. These are almost 2 years for now.

I do not remember exactly but it showed up on my PC at the very first time about 1-2 months ago. It happened from that time only 3 times always in the middle of a session, all 3 times while PC was just idle. The monitor goes black, "no signal" on monitor the keyboard was locked 2 times of 3 completely, and I had to shutdown by pushing a power button on a tower. The last time keyboard responded to Ctrl-Delete and restarted PC. Then I could continue to work as normal.

My first shot was actually a hardware problem. But after some troubleshooting, further investigation and research I came here. And here I think my problem is.

That's a pity, since I would prefer Open Source driver over proprietary one.

[vito (ovitorio)]

Looks like my previous report was largely premature. After looking deeper into issue I've found that my initial "ati" driver install and configuration somehow get screwed up along with some other components like libgl1-mesa-glx and libgl1-mesa-dri.

All at a sudden I found "direct rendering" reset to NO and glxgears showing extremely low frame rate.

After re-installing everything, then running dpkg-reconfigure xserver-xorg I have things running way faster, and I get back my "direct rendering". glxgears shows frame rate about 8 (!) times faster then before.

I do not know yet, if the "no signal" problem has gone. I hope so very much. I'll watch it.

It seems that I misunderstood the "fix released" message as well. I thought I get it on my Hardy with one of the latest updates. Then I realized that it affected Karmic only.

Again, sorry, for the premature report.