flight 5&6: install hang on ATI x700/x800

I can confirm this bug, it seems to be h/w specific though.

I have a Targa Traveller 826 laptop, see https://wiki.ubuntu.com/LaptopTestingTeam/TargaTraveller826 for details.

I have the same problem: Booting from cd, I can see the ubuntu/kubuntu logo, but the boot cd menu never appears.

I tried all four ubuntu i386 and amd64 install and live cds, as well as their kubuntu counterparts.

I've never seen this problem with any earlier release i installed before, certainly not with any dapper flight release up to and including flight-4.

Regards,
Dominik

I have tried with Kubuntu Dapper flight 5 and I have the same problem.

I add an information : The boot menu appears with Dapper flight 4.

I can confirm this bug as well. I have a Fujitsu-Siemens Amilo M1437G laptop. Specs here: http://www.fujitsu-siemens.com/Resources/19/321597266.pdf

Thanks

Same problem with Kubuntu Flight 5 on HP nw8240. More info in bug 34586.

Laptop testing page:
https://wiki.ubuntu.com/LaptopTestingTeam/HPNW8240/Kubuntu

I consider this critical as it cannot be installed on several systems from different vendors.

I have same problem but the menu items appears...but when splash-boot indicate 'Mounting root file system' it stopped and splash-boot to disappear and the boot stopped at the message 'set - thermal udev'. More info in bug 36827 https://launchpad.net/malone/bugs/36827 .

I'm trying to hang to grub many parameters but don't work.

Regars,
gp

PeppeP: If the menu items appear, you don't have the same problem.

Out of curiosity, if you press Enter, does it boot anyway? That is, I wonder if it's just a display problem.

no, it doesn't. system seems to be completely frozen. even ctrl-alt-del/backspace doesn't work. The laptop's fan starts humming after a wee while of waiting, perhaps the cpu is stuck in an infinite loop....

I can confirm Dominik's observation on my HP nw8240.

I confirm Dominik's observation on my PC.

fireboot: what type of computer do you have? Who is the manufactuer and what is the graphics card etc.

Paul Sladen :
Motherboard : MSI K8N Neo4 Platinum
Processor : Amd Athlon 64 X2 3800+
Graphics card : Sapphire Radeon X800 GTO²
Hard disk : Western Digital WDC2500KS 250GB Sata2

I'm having a similar problem, also on a Fujitsu Siemens motherboard (X1170, I think, using the on-board graphics and an Intel Pentium 3 "coppermine" processor). The install menu comes up and the kernel loads, but it then comes up with a blank screen, with a cusor flashing at the top left (no, it doesn't accept input from the keyboard). I've tried two different processors and think this is motherboard specific. This happens with Ubuntu and Kubuntu Dapper f5 Cds. Breezy installed O.K. Dapper installed fine from the same CD on an old Dell (until the HD controller died) and will run on the Fujitsu using the Dell's hard drive.

I have tested just released Kubuntu Flight 6 install CD and the problem is still reproducable on HP nw8240.

Yep. The same problem with Ubuntu Dapper Flight 6. Only thing I can see is the Ubuntu logo before the hard freeze on Fujitsu Siemens Amilo M1437G laptop. I confirm Dominik's observation.

Yes the same Problem with my sapphire x800 gto quadrat 2 and gigabyte k8nxp-sli mobo with amd64 winchester,leadtek pvr,audigy2zs. No menu. I have tried Ubuntu Dapper Drake flight 5,Kubuntu, Live,32bit,64bit,and Install Cds. All the same Problem!

I have the same problem with Ubuntu/Kubuntu dapper flight 6

Okay, some things can you try:

  Does the caps-lock key+LED work?

  Does deleteing 'splash' and 'quiet' options from the boot commandline work.

  Does switching gfxboot to text-mode first before pressing *boot* work.

Paul Sladen:
Caps-Lock key+LED doesn't work. Same thing with num-lock. I have to shutdown the computer from the power button, because of the hard freeze.

Could you give instructions how to test the last two cases? I never get to the boot commandline. I only see the graphical ubuntu logo and so no menu nor commandline.

Thank you

I can confirm NDGsleeper observations on HP nw8240.

AKAIK, there is no way to get rid of splash at this point as this is done by boot loader and no kernel is running yet. There are just couple of messages before that splash screen appears, but they are so fast that is impossible to read them (I think the first like is ISO LINUX loader or something like that).

It seems that all valid reports (reports from PeppeP and Roy Niblett are for different problem) are on machines with different newer ATI graphic cards.

For summary:
fireboot: Sapphire Radeon X800 GTO² (amd64 desktop)
DominikBodi: ATI Mobility Radeon X700 (Targa Traveller 826)
NDGsleeper: ATI Mobility Radeon X700 (FSAmilo M1437G)
Luka Renko: ATI FireGL V5000 PCI-Express (HP nw8240)
Hagbard Celine: sapphire x800 (amd64 desktop)

Are all this cards PCI-Express?
Can all reporters please post output of "sudo lspci -v" for your card?

Mine is:
0000:01:00.0 VGA compatible controller: ATI Technologies Inc Radeon Mobility X70
0 (PCIE) (prog-if 00 [VGA])
        Subsystem: Hewlett-Packard Company: Unknown device 0940
        Flags: bus master, fast devsel, latency 0, IRQ 10
        Memory at c0000000 (32-bit, prefetchable) [size=128M]
        I/O ports at 2000 [size=256]
        Memory at c8800000 (32-bit, non-prefetchable) [size=64K]
        Expansion ROM at c8820000 [disabled] [size=128K]
        Capabilities: [50] Power Management version 2
        Capabilities: [58] #10 [0001]
        Capabilities: [80] Message Signalled Interrupts: 64bit+ Queue=0/0 Enable-

I have missed another report from duplicate bug 37580.

Hagbard Celine: x850xt PCI-Express (amd64 desktop)

Mine (Using PCLinuxOS for now):
03:00.0 VGA compatible controller: ATI Technologies Inc Radeon Mobility X700 (PCIE) (prog-if 00 [VGA])
        Subsystem: Fujitsu Siemens Computer GmbH Unknown device 107c
        Flags: bus master, fast devsel, latency 0, IRQ 10
        Memory at d0000000 (64-bit, prefetchable) [size=128M]
        Memory at ffdf0000 (64-bit, non-prefetchable) [size=64K]
        I/O ports at d000 [size=256]
        Expansion ROM at ffdc0000 [disabled] [size=128K]
        Capabilities: [50] Power Management version 2
        Capabilities: [58] Express Endpoint IRQ 0
        Capabilities: [80] Message Signalled Interrupts: 64bit+ Queue=0/0 Enable-
        Capabilities: [100] Advanced Error Reporting

Mine (Using vidalinux)

05:00.0 VGA compatible controller: ATI Technologies Inc R430 [Radeon X800 XL] (PCIe) (prog-if 00 [VGA])
        Subsystem: PC Partner Limited Unknown device 1600
        Flags: bus master, fast devsel, latency 0, IRQ 3
        Memory at d0000000 (64-bit, prefetchable) [size=256M]
        Memory at fd7f0000 (64-bit, non-prefetchable) [size=64K]
        I/O ports at 6c00 [size=256]
        [virtual] Expansion ROM at fd700000 [disabled] [size=128K]
        Capabilities: [50] Power Management version 2
        Capabilities: [58] Express Endpoint IRQ 0
        Capabilities: [80] Message Signalled Interrupts: 64bit+ Queue=0/0 Enable-
        Capabilities: [100] Advanced Error Reporting

05:00.1 Display controller: ATI Technologies Inc R430 [Radeon X800 XL] (PCIe) Secondary
        Subsystem: PC Partner Limited Unknown device 1601
        Flags: fast devsel
        Memory at fd7e0000 (64-bit, non-prefetchable) [disabled] [size=64K]
        Capabilities: [50] Power Management version 2
        Capabilities: [58] Express Endpoint IRQ 0

Same bug is also being discussed in this forum thread:
http://www.ubuntuforums.org/showthread.php?t=153989

Looks like all are ATI x700/x800 cards.

I'd like to try to isolate the point at which this bug was introduced. It seems likely to me (although not certain) that it was due to some change in the gfxboot-theme-ubuntu code. Accordingly, I've prepared six reduced CD images with various different versions of this code. I'd like anyone affected by this bug to download each of these, burn them to CD, and tell me exactly which images (if any) successfully display the full install menu. The images will not actually install a working system, so don't bother trying that.

The images are here, and are about 8MB each:

  http://people.ubuntu.com/~cjwatson/gfxboot/34592/

Thanks in advance.

Colin, I have tested all and only last one (gfxboot-theme-ubuntu-0.1.17-i386.iso) exposed the same behaviour as Flight 5/6 install CD.
I will be offline for couple of hours, but available in evening for further tests if you will have something new to test.

OK, thanks. That narrows it down to essentially a single change, which was the addition of parsing for the gfxboot-background and gfxboot-foreground keywords in isolinux.cfg.

I've prepared another test image (http://people.ubuntu.com/~cjwatson/gfxboot/34592/gfxboot-theme-ubuntu-0.1.17.1-i386.iso). This has one bug fixed (which should be irrelevant in this case) and a number of debugging tracepoints added. When you boot into this image, a debugging window should appear at the top left-hand corner of the screen. Its contents are only useful if you have some other materials to hand and a good bit of experience at deciphering them, so don't worry about that for now. I would like you to press Escape, *wait* for the contents of the debugging window to change, and repeat; count exactly the number of times that you press Escape (which correspond roughly to lines in isolinux.cfg). In a working system, you will press Escape 64 times, and then the install menu will appear; in your system, it should freeze before that. Tell me the number of times you pressed Escape before the freeze, and we'll take it from there.

Thanks in advance!

It hangs on 63rd press of Esc.

The screen contains something like this:

Left side:
0: 709a0.4
-----
err 0
ip 45dd 4606.5

Right side:
ffffffffff .5
cec .5
4779.65
45db.25

Note: during the steps mostly (only?) left side of debug messages change.

I made the same tests as Luka Renko on my FS-A1437G with exactly the same kind of observations. I hope you already got all the required information for solving the problem. Looking forward to install the next dapper release!

I have the same results as Luka.

Luka and I went through some further debugging on IRC, and nailed it down to a 'free' call, indicating dodgy memory handling. I looked through the code and believe that this change fixes the problem. I'd appreciate it if folks affected by this could try tomorrow's daily install CD (http://cdimage.ubuntu.com/daily/current/, but wait for about 18 hours from now before downloading it; make sure the timestamp shown on that page is current when you do) and let me know if the hang has been fixed.

gfxboot-theme-ubuntu (0.1.18) dapper; urgency=low

  * Fix parsing of gfxboot-foreground keyword.
  * Add translations for new xforcevesa boot options (closes: Malone
    #27020).
  * Update language names from localechooser 0.27ubuntu13, adding Thai.
  * Fix comments in .po/.pot files: maintainer comments should start with
    "#.", not "# ".
  * Update translations from Rosetta: Breton, German, Greek, Spanish,
    Finnish, French, Galician, Hungarian, Italian, Lithuanian, Norwegian
    Bokmål, Polish, Portuguese, Portuguese (Brazil), Romanian, Russian,
    Slovak, Swedish, Chinese (China), Chinese (Taiwan).
  * Update font to add newly-required characters.
  * Fix memory handling in isolinux.cfg parser: take copies of all labels,
    and free a pointer to the start of the file, not the end (closes: Malone
    #34592).

 -- Colin Watson <email address hidden> Fri, 7 Apr 2006 18:02:33 +0100

Colin, I can confirm this bug is fixed with this daily build: http://cdimage.ubuntu.com/daily/20060408/dapper-install-i386.iso

Menu worked, but I now again get bug 31974, but could install from text mode.

It's work correctly for me too. I'm using dapper, thanks Colin :)

Hooray; one down, er, some to go. Thanks all!

Thanks I am downloading todays 20060410 daily right now and giving it a try. Lets hope I am able to go beyond the boot splash. I have been strugling with this for a while now, and it indeed a relief to see this fixed. Thanks again!

Hardware HP Compaq nw8240 laptop ATI FireGL V5000

Please confirm if issue still persists on Dapper Beta. Thanks.

Jerome

oopps please ignore. (didnt notice the status)

Confirmed as fixed in Ubuntu Beta. ;-)