Results 11 to 19 of 19
Although not excluded, I doubt that something is wrong with my hardware.
The disk geometry got screwed when I was resizing partitions with PartitionMagic. So I tried changing the geometry ...
- 09-03-2009 #11
Although not excluded, I doubt that something is wrong with my hardware.
The disk geometry got screwed when I was resizing partitions with PartitionMagic. So I tried changing the geometry using TestDisk geometry feature, and that messed up the CHS values.
I believe that Gparted and PartitionMagic are not agreeing on some partitioning standards!
Still, I believe that the disk will be fine after the low-level format.Fedora 13, GNOME
HP Compaq NC6000, Intel Pentium M Processor 725, ATI Mobility Radeon 9600, Intel PRO 2200BG
HP Compaq DC7100, Intel Pentium 4 Processor 541, Intel 915G Express, D-Link DWL-510 AirPlus G
- 09-03-2009 #12Linux User
- Join Date
- Dec 2007
- Location
- Idaho USA
- Posts
- 351
I have no idea just what that feature does (maybe change values in the bios ?) so you may very will be correct.TestDisk geometry feature
That I can believe. I have had Partition Magic make changes in the MBR/VBR that was not correct.I believe that Gparted and PartitionMagic are not agreeing on some partitioning standards!
- 09-03-2009 #13
Backup of data is always a good start

... after that run grub find command
does grub find the kernel ?Code:find /boot/vmlinuz find /vmlinuz
Ed: I have not used Parted Magic since it trashed a drive for me
. Check for kernel image using grub and live CD ...
Last edited by Jonathan183; 09-03-2009 at 08:45 PM.
- 09-04-2009 #14
I found that the correct kernel is "/boot/vmlinuz-2.6.27.29-0.1-default"
The correct GRUB input is then:
Although Linux found many issues with /dev/hda3 FS it started properly. However, I don't trust it any more so I'm going to reinstall everything.Code:kernel /boot/vmlinuz-2.6.27.29-0.1-default root=/dev/disk/by-id/scsi-SATA_FUJITSU_MHV2040_NT30T5C2AWCT-part3 splash=silent
I'm also going to check my hardware to make sure that nothing else is wrong.
Thanks for all the help!Fedora 13, GNOME
HP Compaq NC6000, Intel Pentium M Processor 725, ATI Mobility Radeon 9600, Intel PRO 2200BG
HP Compaq DC7100, Intel Pentium 4 Processor 541, Intel 915G Express, D-Link DWL-510 AirPlus G
- 09-09-2009 #15
Lostfarmer@ you were right. It is a hardware problem.
After I replaced the HDD, testdisk still shows the same CHS warning.
The real bad news is that this is the notebook computer. There are no IDE cables to replace, and there are no other IDE connectors available.
PS. I apologize for wrongfully blaming Partition Magic to cause this problem.Fedora 13, GNOME
HP Compaq NC6000, Intel Pentium M Processor 725, ATI Mobility Radeon 9600, Intel PRO 2200BG
HP Compaq DC7100, Intel Pentium 4 Processor 541, Intel 915G Express, D-Link DWL-510 AirPlus G
- 09-09-2009 #16Linux User
- Join Date
- Dec 2007
- Location
- Idaho USA
- Posts
- 351
I know not a thing about notebooks, can you reset the BIOS ? If so do it, that could fix error.
- 09-10-2009 #17
After the BIOS reset something is changed, but it is not perfect yet.
Testdisk does not show any of the old warnings:
However a new warning appears:Code:Warning: Incorrect number of heads/cylinder 240 (NTFS) != 16 (HD) Warning: Bad ending/starting head (CHS and LBA don't match)
Code:Warning: The current number of heads per cylinder is 240 but the correct value may be 16.
Fedora 13, GNOME
HP Compaq NC6000, Intel Pentium M Processor 725, ATI Mobility Radeon 9600, Intel PRO 2200BG
HP Compaq DC7100, Intel Pentium 4 Processor 541, Intel 915G Express, D-Link DWL-510 AirPlus G
- 09-10-2009 #18Linux User
- Join Date
- Dec 2007
- Location
- Idaho USA
- Posts
- 351
Reading 'testdisk' web sit, they indicate the program must guess the CHS values and sometimes it is wrong. That is way they have the geometry feature to change the way testdisk will see the disk. They do not say at all that the disk may be bad only they are in error when it guesses.
If the error comes after hdd detection, but dueing looking at the partitions it could be the CHS value in the partition table is wrong. Partition Magic will also check the values. The CHS addressing is normally not use , only the LBA addressing mode is used.
You may not have any problem , only a 'testdisk' problem. If linux/XP can read/write to all partition with no problems then would say all is OK.
Does Partition Magic indicate any errors ?
Does both hdd's show 'testdisk' error ?
Can you connect hdd 'external adapter' to a different comp and run testdisk ?
Can you boot into hdd installed linux ?
- 09-10-2009 #19
So far I didn't experience any read/write errors with the HDD.
I can boot normally. Both OS's, Linux and Windows, operate normally.
Partition Magic does not give any warnings/errors.
Gparted also does not give any warnings/errors.
Powermax does not report any issues (all tests passed).
Testdisk is the only software that reports this issue (on both HDD's).
Moreover, testdisk diagnosis is not consistent. There is some OS dependency observed.
In Linux testdisk reports:
In DOS and Windows testdisk reports:Code:Warning: The current number of heads per cylinder is 240 but the correct value may be 255.
It seams as testdisk is indeed guessing. As you explained:Code:Warning: The current number of heads per cylinder is 240 but the correct value may be 16.
I'm starting to become convinced that, after all, nothing is wrong with my hardware.Reading 'testdisk' web sit, they indicate the program must guess the CHS values and sometimes it is wrong. That is way they have the geometry feature to change the way testdisk will see the disk. They do not say at all that the disk may be bad only they are in error when it guesses.Fedora 13, GNOME
HP Compaq NC6000, Intel Pentium M Processor 725, ATI Mobility Radeon 9600, Intel PRO 2200BG
HP Compaq DC7100, Intel Pentium 4 Processor 541, Intel 915G Express, D-Link DWL-510 AirPlus G




