Wes Kurdziolek wrote: >I remember seeing a message on lkml that this is fixed in 2.4.20-pre7. > >On Tue, 2002-10-01 at 11:12, Ibukun Olumuyiwa wrote: > > >>No, I don't have a solution to the DMA problem yet. I remember trying >>hdparm, but not even hdparm could enable DMA on the hard drive (try it -- >>it will refuse. weird.) >> >>I haven't spent much time on this though...I was thinking of recompiling >>the kernel and using a generic driver rather than the intel driver which >>it is probably using. >> >>Ibukun >> >>On Tue 01 Oct 2002, Philipp H?felfinger @ Tiscalinet wrote: >> >> >>>Hi... >>> >>>Am Die, 2002-10-01 um 08.53 schrieb Ibukun Olumuyiwa: >>> >>> >>>>I also had IDE problems...especially with the system seizing >>>>intermittently during hard disk writes, and an indication that the >>>>system was unable to turn on DMA. >>>> >>>> >>>How did you solve this DMA problems? Do you have a patch that works... >>>I tried the latest prepatch for kernel 2.4.20 but the problem wasn't >>>solved. >>> >>>cu >>> >>>Philipp >>> Yes I had that problem too, signaled it several times during the debugging of 9.0, but it could not be fixed apparently. The fact that hdparm cannot enable DMA is related to that bug. There are patches published by the main maintener of the IDE drivers for the kernel, Andre Hedrick. I can't find again his personal site, but if you find it (I posted his address on that list previously, it should not be difficult) you'll find a lot of patches.
This bug has been widely discussed on the kernel mailing list : http://lists.insecure.org/linux-kernel/2002/Jun/0091.html http://www.uwsg.iu.edu/hypermail/linux/kernel/0208.1/0103.html http://www.mail-archive.com/linux-hardware@senator-bedfellow.mit.edu/msg02774.html http://hypermail.idiosynkrasia.net/linux-kernel/archived/2002/week31/0028.html http://www.cs.helsinki.fi/linux/linux-kernel/2002-34/0347.html By the way, the IDE stuff is completely rewritten in the new kernel 2.5. Expect enhancements and problem solving (I hope...) Eric
