There's also a couple of options --fix-fixable and --rebuild-tree mentioned on the manpage. I'd suggest you give it a read. It's probably wise to copy off what data you can first though if you can get at it.

Thanks. Speaking of saving first, a problem I am facing is I plugged in a USB WD My_Book external storage, which used to be /dev/sdv1 or /dev/sdc1(depending on which usb device was inserted first), but it now there is no such device.
I tried a number of other reasonable candidates (all /dev/s*) for example
but it just does not seem to find it.

OOps, I did a check and then a --rebuild-tree on one older slack partition. It looked like it went ok up to the point where it said it fixed it, but I had to reboot(without giving me a prompt) and I
did a hard reboot(unplugging the power, as it did not seem to respond otherwise). When I did again a --check I get
Bad root block 0 (--rebuild-tree did not complete)

Now I try again it looks like it gets to a point:
EP:[<c0146fc6>] find_get_page +0x26/0x50 55:ESP....


Any hints?

This is at the point where I'm out of obvious ideas, and would have to inventory what's wrong with the computer to know what to try next. That would have to be in person. Sorry.

Ok, just want to confirm this: Since apparently although the external disk is powered and apparently working, the Slack system running from the install cd cannot see it, so as to do a backup, how about doing a fresh slack install to an unused partition, mount that and copy from the broken partition(s) to the fresh one, then just get rid of the old partition by reinstalling new linux distros there? Any risks I am missing?

At worst I will have problems with the hw-this can be checked as far as the disk goes if I do a slow format, checking for bad blocks, no?

Couple more things:
1) I tried memtest86. 3,5 hrs later it had reported some 5000000 errors(none ECC)

Not sure what this means. Or what this means. Does this mean replace the RAM? or the HD?

2) I also tried HD check with the latest Ubuntu CD and after some thing it just hung there for more than half and hour, with the progress bar showing no progress.


I do get a prompt, but reiser complains "comparing bitmaps ...ppf-10630: The on disk and the correct bitmaps differ.... Bad vals were found, ...
and says run the rebuild tree.

Does this say anything or should I just start taking components out one by one, as suggested? I obviously cannot take out the RAM, right?

While bad memory can cause those errors, a loose connector can, too. Open the computer up, clean out all the dust and spider webs. While carefully "grounded" to the chassis at all times, remove the RAM (usually by flipping the latches on each side down). Clean the contacts on the RAM sticks and inside the RAM sockets. Re-insert the RAM firmly. While inside, remove and re-attach other connectors. Make sure nothing is loose from your work. Steps like this tend to cover a lot of apparent hardware problems.

BTW, for fellow old timers here, I've done this on IBM mainframe bus+tag I/O cables and it has helped there, too.

ok, will do. But is this really consistent with the symptoms? (i.e. system hanging)?