From Bugzilla Helper: User-Agent: Mozilla/5.0 (Windows; U; Windows NT 5.1; en-US; rv:1.0.0) Gecko/20020530 Description of problem: When I try to use my tape drive (Seagate Travan 10/20GB IDE), I encounter reproducible system freezes at the end of whatever tape operation is attempted. For example, restore -t will start the drive in motion, whirr for a bit (as normal) but then freeze the system solid at the very moment you expect it to complete. This also happens with tar, cpio but /not/ mt. I am accessing the drive using the ide-tape module, on /dev/nht0. This problem did not occur under RedHat 7.1 Version-Release number of selected component (if applicable): How reproducible: Always Steps to Reproduce: 1.put tape in drive 2.restore -t 3.Crash Actual Results: System freeze. Requires reboot. Expected Results: Contents of first file on tape should be listed Additional info: System is an Athlong 1.2GHz, with Abit KT7A motherboard (VIA chipset)
Something enabled DMA again, it seems. I thought we fixed that in initscripts? mt does not transfer data blocks, only commands. As a workaround "hdparm -d 0 /dev/hdc" should work.
hdparm -d 0 /dev/hdd does indeed work, so this probably is the DMA bug I'd thought was fixed. Which init script would you recommend putting the hdparm line in?
I forget if RHL8 had multiply /etc/sysconfig/harddisk*, or only "harddisks". On RHL 9, copy /etc/sysconfig/harddisks into /etc/sysconfig/harddiskhdd and edit USE_DMA statement. If RHL8 does not support that, add the command into /etc/rc.d/rc.local. Editing "harddisks" will kill DMA on disks, so don't do it. There were some arguments that ide-tape can be fixed for DMA, unfortunately it seems there is no way to find which tapes were broken except white/black lists.