My SuSe freezed, so I had to push reset.
Since then it does not start, it can not mount the / partition and dies with a kernel panic.
I have the following partitions:
hdb5 /boot (ext2)
hdb6 swap
hdb7 / (reiserfs)
I have a SuSe install CD so I could get into rescue mode and run:
reiserfsck --rebuild-tree /dev/hdb7
which finished soon (cca. 2s), and showed:

bread: read 4096 bytes returned -1 (blocks 16, dev3)

It did not seem to help, since I still cannot start SuSe. I get the same error at bootup.

I had a lot of data that I do not want to loose.
Is there any way to recover the reiserfs filesystem on /dev/hdb7?

One of the things that you need to do is to run fsck against hdb5. You might be able to do this using SuSE rescue or perhaps knoppix. If you have another linux system installed on your machine you could use that to run fsck against hdb5.

/dev/hdb5 is mounted as /boot and there is no problem with that filesystem.

I have problem with /dev/hdb7 (reiserfs), which should be mounted as /, but cannot be mounted any longer.

I do not know if fsck is the right tool to correct errors on a reiserfs filesystem. However, I tried it in different ways:
fsck /dev/hdb7
- said there was no ext2 filesystem found
fsck -reiserfs /dev/hdb7 and also fsck -reiser /dev/hdb7
- one of them (I do not remember which) seemed to do something, but finished very fast and did not correct any errors. It said something is wrong with the block count and that I should start it with the -b option and specify a different block count. I, however, have the foggiest idea what is wrong and do not know what block count to specify.
Anyway, is not reiserfsck the right tool to correct errors on a reiserfs filesystem?

Here are the last boot messages:

hdb: dma_intr: status=0x51 {DriveReady SeekComplete Error}
hdb: dma_intr: error 0x40 {UncorrectableError}, LBAsect=964091, sector 128
isofs_read_super: bread failed, dev=03:47, iso_blknum=32, block=64
hdb: dma_intr: status=0x51 {DriveReady SeekComplete Error}
hdb: dma_intr: error 0x40 {UncorrectableError}, LBAsect=964091, sector 128
end_request: I/O error, dev 03:47 (hdb), sector 128
read_superblock: bread failed (dev 03:47, block 64, size 1024)
read_superblock: can't find a reiserfs filesystem on dev 03:47, block 8, size 1024
kernel panic: VFS: Unable to mount root fs on 03:47

Based on the above messages, I might think that there is something wrong with the dma mode. But then why can I mount the other partitions on the same HDD?
Anyway, I disabled dma in the bios, and checked with hdparm that it is not used (at least when I boot up from the rescue CD - is it possible that dma is used when I try to boot up the system from hdd?).

Any hope you see?

By the way, could someone send me a compiled reiserfsck? My e-mail address is ja_szucs@freestart.hu.

My rescue CD has a very ancient version of it, which does not even know the --rebuild-sb parameter that I might need.
I downloaded the source of the new version of reiserfsck, but I came to realize that compilation is out of question under windows or under a rescue system of Linux...
Thanks in advance!