Hello!
I've recently acquired a used hard-drive (HDD IDE Seagate 40Gb ST340016A - bought it using local forums - i.e. not in the shop) and I have 2 days (48 hours) to test the device for problems and return it back if any problems found.

Currently accessing the device on the test machine (i'll install HDD on another machine that isn't going to be assembled in 48 hours) gives 100% CPU load immediately. Turning DMA on didn't help.

The main questions is - which program should I use to test device (not the filesystem integrity) for problems? There is "seatools" program, but it looks like it doesn't work with IDE drives, only with SATA. And what's this with CPU load? I do not have similar problems with another HDD (WDC WD400BB-00DEA0)

Thanks.

Open a terminal window and start the top utility to see what is causing the CPU load.

You can use the mt utility to test the drive. Here are some examples to retention a tape and to erase a tape header. Here the tape drive is mapped as /dev/mt0.
More information about mt can be found at the following URL.
http://www.linuxcommand.org/man_pages/mt1.html

Your Linux installation might map your IDE tape drive in one of several different ways. It might be mapped as /dev/mt0 or /dev/sr0 or /dev/st0. That depends on things like whether your boot command includes the directive ide=scsi. If it does then you might want to remove it. You really don't need it, but if you have it then you can do more operations on your tape drive that are only available with scsi tape drives.

Sorry, I was talking about a HDD, not the tape device. Did I put a misprint somewhere? Or maybe I do not understand something?

Well, cpu load was caused by wrong drive parameters.
It looks like it was used in 16-bit mode with disabled dma and umask_irq.
I've turned those parameters on (hdparm -c1 -u1 -d1 -Xudma5 /dev/hdd), now cpu load is gone, and drive works with a very good speed.

Problem solved.