Post #304,458
2/21/09 12:29:37 PM
|
CD/RW not working in Ubuntu
My laptop isn't seeing any media in the CD/RW drive. The eject button doesn't work, and when I select eject from nautilus I get a message saying no media found. The only way to open the tray is the paperclip in the hole routine. When I close it the light comes on, so it's definitely recognizing that it closed, and I can boot from a bootable CD so it's not a hardware issue.
I've Googled for it, and seen a bunch of people having the same issue with laptops and CD drives. Several people said it's a GRUB problem, and that switching to LILO fixed it. I don't want to go that route, but I'm about out of patience with it.
Any pointers for what I should research or, barring that, a good HOWTO for switching to LILO greatly appreciated.
--
Drew
|
Post #304,470
2/21/09 5:29:13 PM
|
How in the HELL could it be a GRUB v LILO issue?
What?
Did it work at all before?
Okay... pointers to the location of said fixes please.
Sorry, I have to see this to believe it.
How the HELL can a bootloader cause this issue? Ain't now way. Its got to be coincidental that switching to LILO fixes it.
I believe it triggers a rebuild of the device map that a recent update on HAL or UDEV or something similar caused the current device map to become bad or at least parsed wrongly.
|
Post #304,481
2/21/09 8:37:59 PM
|
I'm just telling you what I read
But what I found when looking for those threads was a tip to change something in the BIOS. Guess I'll try that.
--
Drew
|
Post #304,485
2/21/09 10:26:39 PM
|
Linky please so I can go and correct the...
Crap outta them!
|
Post #304,473
2/21/09 5:41:58 PM
|
Is LILO still alive?
I haven't used it for years and years.
|
Post #304,504
2/22/09 6:43:16 AM
|
Wrong device mapped.
Whatever device your kernel thinks is the CD R/W drive is wrong. udev persistent rules again? I don't know.
I can see how it might be the bootloader, if it's passing something to the kernel about which device is the CD R/W (but why?)
Wade.
"Ah -- I take it the doorbell doesn't work?"
|
Post #304,509
2/22/09 8:31:22 AM
|
Which leads us back to...
Installing LILO fixes the device mapping issue and thus is not the root cause but relieves the issue.
I'd imagine that re-installing grub would work after the LILO steps. But is it really worth the pain LILO presents?
I'd bet that: apt-get install --reinstall grub grub-common
Would work just fine... or something similar that causes the machine to re-probe and write out the devices.
|
Post #304,511
2/22/09 10:48:45 AM
|
I'll try that
I completely agree with you that Grub/LILO doesn't sound like the actual problem, but if it's a set of steps that works I'll do it. Any potential gotchas doing the Grub reinstall?
--
Drew
|
Post #304,519
2/22/09 5:31:48 PM
|
None that I know of...
Even if there were, its more recoverable.
|
Post #304,831
2/27/09 9:05:44 PM
|
Nope, didn't work
Any pointers to research this?
--
Drew
|
Post #304,835
2/27/09 10:27:40 PM
|
Re: Nope, didn't work
Dang, where's that rickroll link...
Regards, -scott Welcome to Rivendell, Mr. Anderson.
|
Post #304,837
2/27/09 11:37:04 PM
|
Possible related issue, from the other side of the coin
About 6 months ago, I started running into problems with new HP laptops. The issue usually manifests itself as extremely slow SATA harddrive performance. Intel's ICH8+ chipsets allow for a AHCI/IDE combination mode which confuses the heck out of the kernel.
This can be selected in the BIOS. Locking in AHCI would be the preferred way, but it completely locks out real IDE devices if they are present because the kernel will not go looking for them (in HP's case: the CD drives...)
The other option is to set the BIOS to something other than AHCI, and add the ide0=noprobe boot parameter.
Is there any way you can compare the boot parameters between GRUB and LILO?
(Warning: if this is a dual boot setup with Windows, do not change the BIOS option. Windows will no longer boot.)
|
Post #304,840
2/28/09 12:02:45 AM
|
No dual boot
Going to try playing with BIOS settings.
--
Drew
|
Post #305,016
3/3/09 5:05:46 PM
|
Funnily enough...
I just experienced this and haven't had the time to diagnose it.
In fact the "/dev/hda" device isn't even being created by HAL/UDEV etc.
I made one manually, but that is a pita.
More later, gotta screening of "Watchmen" to go see.
|
Post #306,451
3/30/09 9:44:20 PM
|
Fixed, sort of
Turns out I had a bad drive after all. But even after replacing it, it still didn't work. I found the two places that were recommending Lilo to fix it:
http://ubuntuforums....7&t=282913&page=2
http://ubuntuforums....php/t-353550.html
I also found recommendations to add "noapic" and "acpi=off" to the end of my "kernel" line in menu.lst. The acpi switch seems to have fixed it. It should matter, but everything I care about seems to be working.
Seems this laptop is really light on reliability. The combo drive seems to fail, and the video cable is flaky. That seems to explain the ghosting I mentioned in another thread. But hey, what did I expect for < $400 new?
--
Drew
|
Post #306,454
3/30/09 10:06:28 PM
|
Forgot to mention...
All I had to do was re-run the initrd build scripts... trigger it to think it had never run... forcing grub to redo its thinking and drive map.
Kernel has been funky lately.
All my stuff works now too.
|
Post #306,455
3/30/09 10:08:24 PM
|
How do you do that?
Re-running the build scripts, that is.
--
Drew
|
Post #306,473
3/31/09 10:58:19 AM
|
man mkinitrd, good example there
|
Post #306,474
3/31/09 11:09:16 AM
|
No manual entry for mkinitrd
--
Drew
|
Post #306,481
3/31/09 12:32:02 PM
|
mknod -m 400 /dev/initrd b 1 250
chown root:disk /dev/initrd
from the ubuntu man initrd page,
I would be very careful before running that,
|
Post #306,491
3/31/09 3:00:06 PM
|
If that comes with a warning ...
I'll stick with the solution I've got in place. apci not working doesn't really affect me.
--
Drew
|
Post #306,507
3/31/09 8:26:25 PM
|
Given its Debian based, try update-initramfs
|
Post #306,510
3/31/09 8:40:23 PM
|
Re: Given its Debian based, try update-initramfs
DING DING...
and follow the directions and advice on the developers pages.
|
Post #306,525
3/31/09 9:29:12 PM
|
Holy cow, I gotta lotta kernels on here
Time to go clean up. I think the last two should be plenty.
--
Drew
|