I'm having problems getting my Dell Inspiron 9300 to see my DVD-RW as a SATA device (/dev/sdb[x]) instead of a IDE device (/dev/hdb[x]). This seriously impedes performance of the drive. Having done a little searching around the Internet (what a tool!) it appears the problem may be that both the ide and ata_piix drivers are included in the kernel as modules and that this results in the ata_piix driver installing before the ide driver and as such the DVD-RW drive is then seen as an IDE device. It's suggested that the ide driver be compiled into the kernel so that is installs first which results in the ata_piix driver then picking up the DVD-RW drive and getting it Right(TM). Can someone interpret what I'm saying (;-]) and then comment on how these two drivers are included in the kernel and whether or not this might affect the outcomes. Alternatively, can someone suggest another fix for this. R.

On Fri, Sep 29, 2006 at 12:28:37PM +0000, michiel wrote: > kernel parameters: > > libata atapi_enabled=1 ide hdc=noprobe Asides from the hdc=noprobe, the rest of the command line does nothing in current kernels. libata.atapi_enabled=1 *used* to do something, but now that's the default (Also you missed the all important '.'). And 'ide' doesn't do anything, and never did afair.

On Mon, Oct 02, 2006 at 11:23:09AM +1000, Rodd Clarkson wrote: > On Fri, 2006-09-29 at 15:20 -0400, Dave Jones wrote: > > On Fri, Sep 29, 2006 at 12:28:37PM +0000, michiel wrote: > > > > > kernel parameters: > > > > > > libata atapi_enabled=1 ide hdc=noprobe > > > > Asides from the hdc=noprobe, the rest of the command line does nothing > > in current kernels. libata.atapi_enabled=1 *used* to do something, > > but now that's the default (Also you missed the all important '.'). > > And 'ide' doesn't do anything, and never did afair. > > I hate to ask this question, but why do I need to manually supply a > command line argument for this to 'Just Work'? It's a bug. You're correct in thinking that it should "just work". Dave

On Mon, Oct 02, 2006 at 02:11:17PM +1000, Rodd Clarkson wrote: > > > > > kernel parameters: > > > > > libata atapi_enabled=1 ide hdc=noprobe > > > > Asides from the hdc=noprobe, the rest of the command line does nothing > > > > in current kernels. libata.atapi_enabled=1 *used* to do something, > > > > but now that's the default (Also you missed the all important '.'). > > > > And 'ide' doesn't do anything, and never did afair. > > > > > > I hate to ask this question, but why do I need to manually supply a > > > command line argument for this to 'Just Work'? > > > > It's a bug. You're correct in thinking that it should "just work". > > Thanks Dave. I take it someone is working on fixing this bug? Is there > a bugzilla entry somewhere I could track? Doesn't ring a bell, and it doesn't show up on the SATA bug tracker list.. (https://bugzilla.redhat.com/bugzilla/showdependencytree.cgi?id=FCMETA_SATA) Though that's fallen a bit by the wayside in recent times, so there may be SATA bugs that aren't listed on there. It won't hurt to file a new one.