Well, I finally found a DVD-ROM to test on, though I'm not 100% sure if it's free of CSS (mplayer doesn't seem to indicate it's using CSS, so that's a good sign). The read speed for the disk ends up topping out at around 3MB/s with the CPU using about 80%-90% CPU time in kernel space. A test of reading a DVD-R 4x shows around 4.5MB/s with the same CPU usage though a lot more stutter in apps.
Volker, are you using ide-scsi or ide-cd? I'm avoid ide-scsi, regardless of burn rate, because of issues which I descripted in the PS on my first post--a side note is that with ide-scsi, dma and non-dma did behave different with dma buring at the full 4x. If you didn't do burning with ide-cd, could you post those results instead or if you are using ide-cd could you reaffirm that? I'm actually beginning to wonder if the issue is partially my ide cable itself. I salvaged it from another relatively modern computer (a gateway computer bought in late '99/early '00), so I assumed that the cable wouldn't be an issue. And my processor is a 1.26Ghz Athlon XP, so I highly doubt that it has anything to do with the CPU itself. It very much sounds like a scheduling issue, really, and I'd like to find some other people with the same issue to figure out the commonality. If there is one, it'll be a lot easier to step forward to the lkml to see if they can reproduce and debug what we believe is the source of the problem. A small note that my burner will only hypothetically burn at 1x or 4x (those are the two speeds k3b lists). Trying to set 1x through k3b always ends up producing an error and an abort message prior to the burn even starting. Trying to run the following: growisofs -speed=1 -dvd-compat -Z /dev/ide/host0/bus1/target0/lun0/ cd=<premastered dvd iso> Produces: :-[ unable to GET EVENT: FFFFFFFF ] Executing 'builtin_dd if=<premastered dvd iso> of=/dev/ide/host0/bus1/target0/ lun0/cd obs=32k seek=0' :-[ Failed to change write speed: 5540->1385 So, I get the impression it's not possible to burn at 1x. Given all this, My question is should I test out another IDE cable given how ide-scsi works but has issues and ide-cd is horribly slow, or does the behavior rule out it being the ide cable's fault? I can always try out another cable if someone thinks it'll help. Kuwanger On Monday 15 December 2003 11:53 am, Joerg Schilling wrote: > From: Andy Polyakov <[EMAIL PROTECTED]> > > >> Looks like you don"t have DMA :-( > > > >Lack of DMA doesn't manifest itself as 0.4x. Requestor's computer (which > >was described in originating post) should perform over 2x without DMA. > >Lack of DMA doesn't manifest itself as "1 to 2 CPU spikes per second," > >but as high load all the time. The comment is hardly relevant. A. > > ^^^^^^^^^^^^^^^^^^^^^^^^^^^^ > Seems to be just another time where you like > to tell that you are not interested in > a cooperation, why? > > > The spikes may _well_ be a result if missing DMA (depending on the > buffer management of the drive). > > It _really_ makes sense to run the readcd speed test. > > > > Jörg