Bug 115022

You can access a USB stick and mount it manually.

Please send the output of "hwinfo" and the YaST log files from /var/log/YaST 

Created attachment 48716 [details]
Compressed tarball of requested logs

I attached a printer to a USB port on the laptop and this appears to enable the
installer to find the hard disc. I have sent log files of the successful
installaton(update from beta 3). To check I unplugged the printer and attempted
installation again but the installer couldn't find the hard disk again.

Could you please also attach log files from an unsuccessful installation attempt.
If disks are not found without an USB device attached this seems to be more a
problem of hardware detetion than yast2-storage.

I need the log from 'hwinfo --storage --log=xxx' taken during installation. 

and it's not a blocker  

I've added logfile as requested for successful installation. Can't do one yet
for unsuccessful install as I don't have a memory stick. Will get one and
forward info to you.

Created attachment 48993 [details]
hwinfo --storage output

Nick, I really need the logs from a failed attempt to get an idea why it 
failed. (And why you're at it, 'hwinfo --disk', too for good measure.) 
 
You can take the log and then fix the system up by loading the modules 
manually or whatever else is necessary to get it going and then store 
the file on your hard disk. 
 
Maybe you can then just tell me what went wrong. :-) 

I've bought myself a USB stick and produced the log files you need from an
unsuccessful installation.Attacment follows.

Created attachment 49466 [details]
Output from hwinfo --storage + --disk

Kernel messages show that the sis5513 driver has problems with hda 
(but not in comment 8): 
  
  <6>SIS5513: chipset revision 0  
  <6>SIS5513: not 100% native mode: will probe irqs later  
  <6>SIS5513: SiS 962/963 MuTIOL IDE UDMA133 controller  
  <6>    ide0: BM-DMA at 0x2800-0x2807, BIOS settings: hda:DMA, hdb:pio  
  <6>    ide1: BM-DMA at 0x2808-0x280f, BIOS settings: hdc:DMA, hdd:pio  
  <7>Probing IDE interface ide0...  
  <4>hda: MO2OAD4AIC25N040ATMR04-0, ATA DISK drive  
  <4>ide0 at 0x1f0-0x1f7,0x3f6 on irq 14  
  <6>hda: max request size: 128KiB  
  <6>hda: 0 sectors (0 MB) w/8875KiB Cache, CHS=0/0/55  
  <4>hda: set_multmode: status=0x51 { DriveReady SeekComplete Error }  
  <4>hda: set_multmode: error=0x04 { DriveStatusError }  
  <4>ide: failed opcode was: 0xef  
  <6>hda: cache flushes supported  
  <3>hda: INVALID GEOMETRY: 0 PHYSICAL HEADS?  
  

I can confirm that I also have this problem, and I have the same chipset the SiS962. I get the same error as comment #12, and even if I stick a printer into one of the USB ports this does not solve my problem.

I am using a Pundit AB-P 2600 Barebone, with the SiS962 chipset. I am using a maxtor 80GB HDD.

To confirm, I can install SuSE 9.0 on my machine with no errors, but I get the same "No hard disks were found for this installation, please check your hardware!" error message.

Is there any thing I can do to overcome this problem. I have tried insmod=ide-generic, and turning off acpi, and tyring to install using the safe setting. 

nothing works for me, I am stuck without being able to install.

Just to update, I am trying with Suse 10.0 OSS GM, not Beta 4.

I tried various command line options but nothing works, the driver for this chip set is broken in beta 4 onwards. I've got around this problem by installing a version prior to beta 4 and then running update with 10.0.x rather than a fresh install.

Nick, did you try 10.1 Alpha, or is it likely to still be broken in there, also where can I download 10.0 Beta 3, and did you still need the printer in for Beta 3 in order to get it to work.

Steffen, there are no kernel changes in this area between beta3 and beta4. Did something change in the installation system/tools?

No. No changes there.

Matthew, I tried 10.1 alpha and the same problem there too. The mirrors I use no longer carry the old betas, you could try installing 9.x and running update on that. I started with 9.2 (9.3 has the same problem with finding my disc) and updated that with each of the betas in turn and didn't run into any problems.

I may try the upgrade but I don't fancy going from 9.2 upto 10.0, will it not leave me with a fairly unstable system, or should it be fairly straightforward. Having said that I notice that the bug has now been assigned, so perhaps we can expect a work around at some point.

I tried to upgrade from a Suse 9.2 Distribution. Incidentally my chipset is supported perfectly in Suse 9.2, however after rebooting with the Suse 10.0 OSS GM installation disk 1 in the CD drive, my drive was still incorrectly detected, (even with a printer in the usb port), and so I was not given the choice between clean install and upgrade.

Therefore I can not even upgrade to 10.0 from another distribution, this bug is blocking me from using 10.0. Is there any special trick you used to upgrade?

After you have installed 9.2 go to Yast>Installation Source and delete whatever is there.Then put your 10.0 disc in the cdrom drive and select 'Add'. Yast should find the 10.0 disc and insert that into the source media.Make sure it's enabled. Then go to 'System Update' and select the options and packages you want.

Ok so I have done the system upgrade, but YOU online update still thinks I have 9.2 installed. /etc/SuSE-release states I have 10.0. So if I were to do an update I would effectively download all my packages... this is not pain free, any ideas?

After reboot the new version should appear in YOU.

Ok it doesn't because I did the update while I was still running 9.2, and apparently this means that my base hasn't been upgraded. Everyone suggests doing a boot from CD upgrade, but for us that won't work.

So at present I can't upgrade my system, any ideas?

You have to be running the old system to do the update, my YOU settings correctly updated to 10.0 so I don't understand why yours didn't. 

I'm guessing it was because you updated from 10.0 Beta3 to 10.0 Final, and therefore you still is in 10.0 mode. Just my hunch, any idea where I can get the isos for 10.0 RC3, maybe the kind people from SuSE could send me a URL?

Solution for me (if it doesn't work for you don't moan):

Get hold of the 10.0 Beta 3 isos (you will need to hunt the mirrors), install 10.0 beta 3 (to your own preference). Then get hold of the 10.0 OSS GM isos, go into Yast remove the current CD installation source. Add the new 10.0 OSS GM CD to the installation source. Click on the System Upgrade button, and agree to update all installed packages. Let it run, and hopefully when it finishes and after a reboot, you will be running Suse 10.0 OSS GM. 

How can the upgrade solve this issue ? After the upgrade, your kernel is at exactly the same version as on the GM iso. So you'll run into the same problem with the next reboot.

Well, sorry to fault your reasoning, but I have rebooted several times since the kernel update, and my machine boots just fine. I have no idea why this has solved the issue for me, but it also seemed to solve it in comment #3. I have no knowledge of what is wrong with the kernel included on the installation disc, all I know is that this work-around does work. 

Be grateful if you could work out why, so that I won't have to do this when going to 10.1 next year.

We'll try to work out why but need your help. ;-)

On your running system, please run "uname -a" and paste the result here.
Then boot the 10.0 GM and after yast starts up (i.e. when you see the language selection screen), switch to console 2 (Press Ctrl-Alt-F2) and run "uname -a" there too.

Ok so both unames produce the same output which is,

Linux linux 2.6.13-15-default #1 Tue Sep 13 14:56:15 UTC 2005 i686 i686 i386 GNU/Linux

Linux linux 2.6.13-15-default #1 Tue Sep 13 14:56:15 UTC 2005 i686 i686 i386 GNU/Linux

Interesting. So its not the driver alone, its something in the 'environment'.

Can you please also check "/proc/cmdline" and "lsmod" in both environments ?

Created attachment 55223 [details]
cmdline output (when run at boot from install CD)

Created attachment 55224 [details]
hwinfo --all output (when run at boot from install CD)

Created attachment 55225 [details]
lsmod output (when run at boot from install CD)

Created attachment 55226 [details]
cmdline output (when run at current run time)

Created attachment 55227 [details]
hwinfo --all output (when run at current run time)

Created attachment 55228 [details]
lsmod output (when run at current run time)

The required information provided in the above 6 attachments.

lsmod 
cmdline
hwinfo --all 

outputs run at both boot time when running from the install CD for Suse 10.0 OSS GM, and when run during my current environment (10.0 Beta 3 upgraded via YaST system upgrade to 10.0 OSS GM using install CDs).

Matthew, the attachment from comment #38 looks incomplete. The one from comment #35 contains much more data, esp. /var/log/messages.
Can you please re-run "hwinfo --all" on your running system (as root !) ?

And what is the BIOS setting for the master disk on the primary IDE bus ? It looks like its not set to LBA mode.

Created attachment 55309 [details]
hwinfo --all output (when run at current run time as root)

The BIOS setting for the master disk on the primary IDE is set to Auto in the BIOS, it is presumably picking up LBA mode, because I am able to use the full capacity of the disk(80GB), and not limited to CHS mode(8GB).

I have also re-run hwinfo, this time as root.

Created attachment 55310 [details]
/var/log/messages run as root

Hmm, attachment from comment #43 is still incomplete. Steffen, any ideas ?

Matthew, what happens if you force LBA mode in your BIOS ?
Please also attach /var/log/boot.msg

Created attachment 55312 [details]
/var/log/boot.msg

I have tried forcing LBA in the bios (I am sure LBA is selected by auto anyhow), but the outputs do not differ, so I have left machine in the default BIOS configuration. I have attached /var/log/boot.msg as per your last request.

I'm moving this to kernel now. Its quite puzzling ...

Has this bug progressed at all. I am beginning to panic that at some point in February when I get SuSE 10.1 I will need to go through all the pain again!

mass reopening all SuSE Linux bugs that are set to REMIND+LATER to change the resolution to WONTFIX (adapting to new policy)

mass reopening all SuSE Linux bugs that are set to REMIND+LATER to change the resolution to WONTFIX (adapting to new policy)

mass reopening all SuSE Linux bugs that are set to REMIND+LATER to change the resolution to WONTFIX (adapting to new policy)

Closing old LATER+REMIND bugs as WONTFIX - if you still plan to work on it, feel free to reopen and set to ASSIGNED.

In case the report saw repeated reopen comments, it's due to bugzilla timing out on the huge request ;(