pmpatrick

In linux, X refers to the graphical windowing system. All GUIs, e.g. KDE, Gnome, ect, in linux run on top of the X windowing system. Also the graphical login screen runs on the X windowing system. If the X system is not properly configured, you either won't get a picture on the screen or you will have some problems with that picture, system hangs, ect. Repeatedly hitting the restart button has the potential to really bugger up your file system. To my knowledge fsck only checks and repairs file system integrity; it does not directly test the hard drive itself. The fact that it is running tells me that hitting the restart button may have caused some problems with the file system, i.e. hitting the restart button resulted in the file system on partition /dev/hda1 not being cleanly unmounted with resulting file system corruption. Linux really hates being shut down this way, more so than windows. However, if you have good reasons to suspect a problem with the hard drive itself, I would definetly recommend downloading the hard drive utility from the manufacturer as I suggested in my previous post. If you want to test the theory that it is the X windowing system causing the problem you can run linux in what is referred to as text mode. In text mode the X system never starts up and you run from a command prompt like DOS.To do so, when the lilo boot screen first comes up immediately press the Escape button. At the command prompt type "linux 3" without quotes and press enter. The system should then boot to a command line style login. Type your user name and press enter then your password and press enter again and you should be at a command prompt. If you get this far, I think you have pretty good evidence that the X windowing system is not configured correctly. If you want to try to start X from here type "startx" without quotes at the command prompt and see what happens.

Also forgot to add, if you can get to a command line by using Ctrl-Alt-F2, you can cleanly reboot or halt the system by typing reboot or halt at the command prompt.

I assume that when the monitor goes black, you are turning the computer off and then back on to reboot. That is why you are getting the fsck message. The fsck command runs automatically on boot if the file system was not unmounted cleanly (i.e. you turn the computer off without going through the shut down procedure). The purpose of fsck is to repair the corrupted file system caused by the improper shutdown. The exit code from fsck is the sum of all the conditions that apply, in your case 1,2, and 8. 1 means errors were found and corrected; 2 means a reboot is suggested; and 8 means that fsck encountered an operational error. The 8 code is the troublesome code here. It could mean the superblock is corrupted. That's the bad news. The good news is that I doubt there is anything wrong with the hard drive other than the file system may be corrupted. If you want some peace of mind on this point and know the make and model of the hard drive, go to the manufacturer's website and you can probably download a utlility which will allow you to run fairly comprehensive tests on the drive and make repairs if necessary. Usually you just download the file, click on the downloaded file and you are prompted to put a floppy in. The diagnostic program is then written to the floppy which you can boot off of in the affected computer and run the diagnostics. Back to the problem at hand, I think your real problem is with your X system since you've indicated problems configuring your monitor. Doing the hard shut down when the screen goes black is only making matters worse. Try booting up again and note any error messages. If the screen goes black again, press Ctrl-Alt-F2. That should bring you to a command line login prompt. Enter your user name and password, then type su and your root password. You can then run the X configuration utility by typing XFdrake. From here you can reset monitor settings, grapics card, etc and test the X system to see how it is working.

If you can network the two computers this link may help re doing a network install: http://www.club-nihil.net/mub/viewtopic.ph...rk+installation Also, you could try doing a hard drive install by transferring the install CDs to a Fat partition on your laptop and following the procedure here: http://www.mandrakeuser.org/docs/install/ihdinst.html

Try going to Mandrake Control Center>Network & Internet>Connection. Under LAN Configuration check to see if under "State" it says up. Also note the driver being used. You can reconfigure the NIC and internet connection by ticking the Wizzard button. I didn't see any Intel NICs listed under supported hardware on Mandrake's site so you may have some problems with this card. If you reconfigure through Wizzard, you will have an opportunity to select the driver. See if there's an option for your card you can select.

I have used my Radeon 7500 with Mandrake 9.0 on two different intel computers, one a P3 the other a P4 and had no problems with either installation. However, I have noticed posts reporting problems with the Radeon 7500. Haven't really noticed whether they were using intel or AMD chipsets but it's possible that the combination of AMD+Radeon 7500 doesn't sit well with Mandrake.

Here's a link discussing several small linux distros: http://www.linuxdot.org/cgi-bin/columns.cgi?id=23 I can't imagine getting X running on 16 megs but I guess anything is possible. I think you will have better luck trying to pare down a mini distro however. Also, on the bootable CD distros, you might want to give Knoppix a try.

You can reset lilo from graphical to text mode by going into Mandrake Control Center>Boot>Boot Config and tick the Configuration button. In the pop up window that comes up there should be an entry for lilo text in the Bootloader to Use section.

Just in case anyone is interested, I installed Redhat 8 and dma is enabled Was getting 38MB/sec compared to 5MB/sec in Mandrake 9.0. I also spent some time on Google researching the issue and this does not appear to be solely a Mandrake problem but rather a kernel problem which only Redhat has patched. I saw several posts from SUSE users with identical complaints re getting dma enabled. Hopefully, Mandrake will take care of this in its next release.

Thanks for the tip. Checked Mandrake Expert and found the post you referred to. The kernel which fixed the problem is 2.4.19.18. The Mandrake 9.0 kerenel is 2.4.19-16mdk. I beleive Redhat 8.0 is 2.4.18 something. Are you sure dma is enabled in the earlier Redhat kernel with the 845E chipset? Please check using #hdparm -v /dev/hdx I'd like to know before I bother with trying to install Redhat. Thanks for your help.

Sorry, typo in the above. The ISA chipset is located on the bus at 00:1f:0.

Did a little further investigation and this is what I found: 1. Checked the hardware profile in Mandrake Control Center. The following device is listed as located on the bus at 0:1f:1 - "8280DB 8456/GL Chipset ISA Bridge". The IDE/ATA controller is located on the bus at 0:1f:1; 2.Checked the data transfer rate of the hard drives using hdparm -t /dev/hdx and got a low rate of around 5MB/sec. With my old P3 motherboard and the same drives, I used to get over 20MB/sec; 3.Checked whether DMA was enabled with hdparm -v /dev/hdx and it wasn't on either drive. DMA was set to zero on both; 4. Now comes the wierd part. Tried to enable DMA with hdparm -d1 /dev/hdx and got the following error message; dev/hdb: setting using_dma to 1 (on) HDIO_SET_DMA failed: Operation not permitted using_dma = 0 (off) Has anyone with the collision of resources error message gotten similar results and does anyone have any suggestions on how to enable DMA?

This is bizarre. I have exactly the same error message after upgrading my motherboard/cpu to an Intel D845EBG2/P4(2.53GHz). It must be something in the 845E chipset. I think the Asus board uses this Intel chipset as do most other P4 motherboard manufacturers. I also had identical symptoms as mad_penguin when I tried to use my Promise Ultra 100 PCI controller card. Mandrake 9.0 installer didn't see the on board Intel IDE controller that my CD drive and my CD burner were on. Wound up pulling the card and putting everything on the on board IDE controller and the install went fine except for the resource collission message. I had no problems with Mandrake 9.0 using the Promise card on my P3 motherboard.

Try checking your partitions in Partition Magic. I've noticed that whenever I've installed either Red Hat or Mandrake on partitions created and formated in Partition Magic, if I run Partition Magic again errors are reported re the partition table, usually two per partition touched by the Linux install. It's a popup window that you will see when you run Partition Magic. It asks if you want to correct the errors; I usually say yes and have no problems afterwards. Also, check your entry in /etc/fstab for the windows partition that your having trouble with. Here's what mine looks like and I've had no problems:/dev/hde1 /mnt/win_c vfat iocharset=iso8859-1,codepage=850,umask=0 0 0

I still had windows 98 on my laptop so copying the CDs to disk wasn't a problem. I'm really not familiar with DOS so I can't help you there. However, I did find something that might shed some further light on the problem if your CD drive is connected via a pcmcia card like mine was. Apparently, if this is the case, you should make a boot floppy using pcmcia.img found in the Images directory of CD1. See : http://www.linux-mandrake.com/en/doc/82/en...l-bootdisk.html Wish I'd seen that when I was doing my install!! If you do connect your CD drive this way, it's worth a shot. Perhaps the drivers in pcmcia.img will eliminate the read errors. Also, the Readme file in the Images directory of CD1 has some limited info on installing over a network.

Try this: 1.Go to Mandrake Control Center>Mount Points>CD Burner.Tick Options then OK and make sure user and and noauto are ticked and supermount is not ticked.Click OK then OK again; 2. Open /etc/lilo.conf in root as follows: # gedit /etc/lilo.conf Check the append line under the linux lablel and see if the following appears there: hde=ide-scsi. If it does delete it, reboot and see if you can can mount your cdrw drive as /dev/hde. This line is what provides the scsi emulation for your cdrw drive. Deleting it may allow you to read from the drive but you will then not be able to write to the drive. For this you need the scsi emualtion but the scsi emualtion can prohibit you from accessing the drive as a normal CD drive. By the way, before you make any of these changes, you may want to see if you can burn a CD on the drive, i.e. see if the drive is accesible to any CD burning program.

If you want to install off your hard drive, here is the link for that: http://www.mandrakeuser.org/docs/install/ihdinst.html. You have to make a boot floppy using the hd image file on CD 1 as described in the link. The image directory of CD1 also has a network image on it so I imagine you could create a boot floppy per the above using this image and install from a network drive. I'm sure it can be done but I can't find any documentation on it. You seem to have thoroughly checked out your CDs. If they are good, the only point of error would seem to be read errors from your CD drive on your laptop. I had a similar problem when I recently tried to install on a laptop that had a CD drive that connected through the pcmcia slot. Wound up doing an install from the hard drive.

If you have your cmos set up to print the boot info on the screen, the hard drive manufaturer will be printed out on the first screen.

If you know the manufacturer of the hard drive, go to its website and download the appropriate hard drive utility for your drive. With this utility you can run diagnostics on the hard drive and do a low level format which zero fills every sector of your hard drive. This will totally clean the drive and leave it in the same condition as a new drive.

According to this link (http://www.scyld.com/network/) SiS 900 is supported in kernels 2.2.13 and above. MD 9.0 uses kernel 2.4.19 so I would think you would be alright.

Make a copy of your boot floppy and make sure it works so you have a backup. I'm pretty sure you should add the option to the append line.The separate commands on this line are separated by a space like so: ..........hdb=ide-scsi linux ide=nodma

The above should take care of defining a mount point and creating an empty directory in /mnt for a mount point, but if you want to actually burn CDs with that new cdrw you will also have to edit the "append" line in lilo.conf which is in /etc by adding: "hdx=ide-scsi" where "x" is the drive letter for your cdrw. Your new append line for your linux boot menu entry in lilo.conf should probably look something like this: append="quiet devfs=mount hdx=ide-scsi" The reason for this edit is that all CD burners in linux have to run under scsi emulation in order to burn CDs.

What you describe is what can happen when your MBR gets messed up, i.e. '9' will be continuosly printed across the screen when you try to boot. I think it means your BIOS can't find something in your MBR. I get the same thing when I use Powerquest's Drive Image to restore but I can always get lilo back by using the rescue utility on CD1 to reinstall lilo. If you made a boot floppy you can try booting off that and checking the boot configuration in Mandrake Control Center. I'm not that familiar with XP, but I'm sure there must be a way to restore the windows MBR. I know in win 98 and earlier, you just boot off the start up disk and enter the following command at the DOS prompt: fdisk /mbr. There should be a space between "fdisk" and "/mbr".

Wow. You've been busy! If your getting the same result with both Mandrake and Red Hat, it would appear to point to some basic hardware incompatability or possibly some problem with your bios set up. Your hard drive is certainly not being read from what I can see in your prior post and it can't be because of a corrupted partition table since you wiped the disk clean. If you can give a listing of your hardware set up, some one might be able to come up with some more ideas. Give the make and model of your computer along with: processor, motherboard chipset, IDE controller, hard drive model, BIOS, memory, CDROM, ect.

Thanks for your suggestions. Still can't get the thing to work. Here's what my research has revealed. Spiedra, you are correct that you should not attempt to mount an audio cd( see: http://www.tldp.org/HOWTO/CDROM-HOWTO/x289...9.html#AEN1008). All programs which attempt to read an audio cd report no audio cd present. When I disable supermount in mcc and reboot supermount loads again despite the fact that I ticked the save changes tab. My conjecturre is that supermount is interfering with audio cd detection by attempting to mount the device when an audio cd is present. I also recall that when I first installed MD 9, it would take Konqueror nearly 30 sec. to list the fies in /mnt which I understand to be part of the supermount problem. At some point that behavior totally disappeared and I could access /mnt without delay; my suspicion is that at the same time I probably lost audio cd capability.Bottom line is I don't really need to play audio CDs as my entire music collection is already ripped to mp3s so I'm not going to mess with this anymore until a fix for supermount comes out. Thanks again for your help.