hdparm -d 1 fail test9-pre8

2000-10-02 Thread Mike Galbraith
Greetings, In order for hdparm -d 1 to work in test9-pre8, I had to reverse this change. (Without being able to enable dma, performance here is muy el-stinko;-) Is enabling dma manually now forbidden? (or am I maybe missing something else?) diff -urN linux-2.4.0-test9-pre7/drivers/ide/ide-pci.

Re: hdparm -d 1 fail test9-pre8

2000-10-02 Thread Jeff Garzik
On Mon, 2 Oct 2000, Mike Galbraith wrote: > In order for hdparm -d 1 to work in test9-pre8, I had to reverse > this change. (Without being able to enable dma, performance here > is muy el-stinko;-) Is enabling dma manually now forbidden? (or > am I maybe missing something else?) If this change

Re: hdparm -d 1 fail test9-pre8

2000-10-02 Thread Ivan Kokshaysky
On Mon, Oct 02, 2000 at 09:24:58AM -0500, Jeff Garzik wrote: > If this change broke your DMA enabling, I think there are other bugs > lurking in the code... > This change also broke CMD646 IDE on alpha lx164. CMD646: IDE controller on PCI bus 00 dev 58 CMD646: chipset revision 1 CMD646: chipset

Re: hdparm -d 1 fail test9-pre8

2000-10-02 Thread Andre Hedrick
On Mon, 2 Oct 2000, Mike Galbraith wrote: > Greetings, > > In order for hdparm -d 1 to work in test9-pre8, I had to reverse > this change. (Without being able to enable dma, performance here > is muy el-stinko;-) Is enabling dma manually now forbidden? (or > am I maybe missing something else?)

Re: hdparm -d 1 fail test9-pre8

2000-10-02 Thread Andre Hedrick
On Mon, 2 Oct 2000, Jeff Garzik wrote: I know that you are still pissed at me but do not do end runs around me. That is not cool, and you do not know where things are going. Heck, I keep changing my mind on stuff as I design it. I would be more appreciative if you at least sent it to me since I