harlow
asked on
fdisk and big EIDE drives
I am trying to install Linux (RedHat 4.1), OS/2, and maybe even Win 95 on a new 6.4 Gb EIDE drive from Western Digital. The drive has 13,328 cylinders, 15 heads, and 63 sectors. The fdisk programs in Linux, OS/2, and Win 95 all see the disk differently, and none of them seems to be able to partition it correctly. In all cases, the fdisk program thinks it knows what it is doing, but when you try to install an OS on the drive, the partition table is found to be hosed and the partitions inaccessible.
I have heard that IBM has a fixpack for Warp 4 which takes care of the problem, but I am not sure about Linux (and don't really care about Win 95...) Linux fdisk reports a "Begin", a "start", and an "end" for each partition I create. The start and end cylinders (or sectors if you use that option) are reported correctly, but the Begin entry always seems to be a multiple of 1024 cylinders that is somewhere near (above or below) the start value. When I tell fdisk to verify the partition table, sure enough, it says that most of the partitions overlap.
The RedHat installation kernel detects the drive and reports its geometry correctly, and my BIOS is new enough that it seems to understand the drive too. Is there a newer version of fdisk that avoids this problem? Am I doing something wrong?
Justin Harlow
I have heard that IBM has a fixpack for Warp 4 which takes care of the problem, but I am not sure about Linux (and don't really care about Win 95...) Linux fdisk reports a "Begin", a "start", and an "end" for each partition I create. The start and end cylinders (or sectors if you use that option) are reported correctly, but the Begin entry always seems to be a multiple of 1024 cylinders that is somewhere near (above or below) the start value. When I tell fdisk to verify the partition table, sure enough, it says that most of the partitions overlap.
The RedHat installation kernel detects the drive and reports its geometry correctly, and my BIOS is new enough that it seems to understand the drive too. Is there a newer version of fdisk that avoids this problem? Am I doing something wrong?
Justin Harlow
ASKER CERTIFIED SOLUTION
membership
This solution is only available to members.
To access this solution, you must be a member of Experts Exchange.
ASKER