I/O errors on hard disk on Linux boot

Question

Here is a screenshot of booting Arch.

I guess the reason is that I force poweroff my Arch linux many times. (I already force poweroff my Arch because my firefox flash plugin use too much memory to stop my system.)

Note: I can boot my Windows 7 system on the same drive disk. So I think it is not a disk problem, mostly is a partion problem.

Update: I check out more information, the partion /dev/sda9 is /home directory. And always error on same sector 798717984. I use DiskGenius software under Windows to check error. Then found an error. and that partion is not formated.

I want to recover my Arch linux. How to solve this ?

If I can not fix this error, then how to get the partion data out ?

Update2: I really hope to save this partion data out. Because I have a lot of important things in this partion. I think the first step is backup this bad partion or whole hard drive into an image file (what image file ?), then let someone who can fix this partion to fix.

More update: After I use DiskGenius software to fix the partion sector error. Then I use e2fsck to check. get error:

fsck.ext4: Bad magic number in super-block while trying to open /dev/sda9.
/dev/sda9: The superblock could not be read or does not describe a correct ext2 filesystem.
VFS: can't find ext4 filesystem.

(my this broken partion /home -> /dev/sda9 is ext4 when I create it before.)

And I execute command # mke2fs /dev/sda9 to get block information:

OS type: Linux
Block size: 4096 (log=2)
Fragment size=4096 (log=2)
Stride = 0 blocks, stripe width = 0 blocks
65536 inodes, 261888 blocks
13094 blocks (5.00%) reserved for super user
First data block = 0
Maximum filesystem bloack = 268435456
8 block groups
32768 blocks per group, 32768 fragments per group
8192 inodes per group
Super block backup stored on blocks:
    32768, 988304, 163840, 229376

Answer 1

Are you able to login and use Arch Linux too, other than the error messages cluttering the console? If so, then most likely your hard drive is failing, just not completely dead yet. The line that says "I/O error" indicates that the kernel encountered an error trying to read data from the harddrive, and the lines beginning with "ata1.00" provide detail about the internals of the read request in the hardware. Windows doesn't show such messages, which is probably why you don't see any problems there, yet.

If you had file system corruption caused by killing the power, then the kernel should still be able to read the data from the drive, but wouldn't be able to interpret what files the data belongs to. That would result in a different set of errors.

Another way to tell would be to reboot and see if you get a similar error but with different details, e.g. a sector number different than 798717984. If so, that means the error is occurring somewhat randomly, which is another sign of hardware failure. Again, this is mostly likely your hard drive, though it is possible another hardware component could be failing.

I'd suggest making a backup and replacing the drive before it's too late.

EDIT AFTER OP UPDATE:

If only one sector is bad, you could use e2fsck -c -y as goldilocks suggested and continue to use the drive if that fixes the error. But modern drives have transparent error-correction built-in, and in my experience, by the time the OS starts to detect errors in the course of normal I/O, the drive is very close to the end of it's life.

Regardless of what course of action you take, absolutely make sure you have a good backup of the entire drive before trying to repair anything!