[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Bug#314707: ide0=ata66 -- OBSOLETE OPTION, WILL BE REMOVED SOON!



hello erik,

thanks for your verbose feedback,

On Sat, 18 Jun 2005, Erik Schanze wrote:

> maximilian attems <debian@sternwelten.at>:
> > On Fri, 17 Jun 2005, Erik Schanze wrote:
<snipp>
> It is an ACER Travelmate 291 LCi.
> I attached dmesg with ide0 option, without it and 'lspci -v' output.
> 
> I played a little bit with:
> - newest Debian-Kernel 2.6.8 (testing, where I come from to unstable), 
> - Kernel 2.6.11 (build from kernel-source-2.6.11, debian-logo-patch applied 
>   and HZ set to 100 in <kernelsource>/include/asm-i386/param.h) and 
> - hdparm.
> 
> On Kernel 2.6.8 the option "ide0=ata66" is required to be able to put HD in 
> UltraDMA mode3/4/5 (it is done on boot automagicly). 
> Without it refuse to set these levels with hdparm (-X67 till -X69).
> Error message "ide0: Speed warnings UDMA 3/4/5 is no functional."
> 
> On kernel 2.6.11 the option is not required, because UltraDMA levels till 
> mode5 are possible without this option. Hdaprm does not complain of setting 
> it (-X69), so the option seems needless.(-> obsolete?)
> 
> I have no idea, if it is only a cosmetic issue in boot messages that there is 
> shown "(UDMA33)" and the chipset runs on UDMA100 though or a really drawback.
> My notebook HD seems too slow to determine it: 
> '$ hdparm -t /dev/hda' gives ~23,9 MB/sec in "UltraDMA mode2" (hdparm -X66) 
> and all above levels till "UltraDMA mode5" -X69, so I think my HD is on the 
> limit.
> 
> Desktop-PCs with similar chipset, that also needed this option maybe give 
> other results, because of its faster HDs.

could you please try linux-image-2.6.12 they aren't yet in unstable
due to the new packaging scheme, but you can find them here:
http://charm.itp.tuwien.ac.at/~mattems/

please report dmesg of the 2.6.12 kernel.
if it still does not find the UDMA100, please send in /proc/ioports
of working 2.6.8 and of non working 2.6.12.

--
maks




Reply to: