[ale] Re: [ale] boot parms ide=???
jb at sourceillustrated.com
Wed Feb 19 10:13:17 EST 2003
Sounds like Dow is a prime candidate for writing the How-to "Tuning Drive
I/O with hdparm".
Dow, can you have it finished by this afternoon, so I can play tonight? :-)
Dow Hurst said:
> If you set the kernel ide parameter to 66 then I wonder if you limit
> throughput to 66 max. You could check this with hdparm by booting with
> different ide parm values and just testing transfer rates. I know that
> tuning my wife's SuSE 7.3 install with hdparm I went from 6Mb to ~25Mb
> per sec just by setting dma. On my SuSE 8.1 machine I found dma already
> set and a transfer rate of ~40Mb per sec was pretty consistent. This
> is limited by the bus so you could back calculate where the limiting
> factor would be. A video card data transfer would be another way to
> test throughput for a PCI based card. Does AGP go thru the bus or
> bypass it on another faster route? I don't know enough about how this
> stuff works now. I used to have the old 86, 286, and 386 stuff down
> pretty good as to how it works. ;-)
> One thought is if you have a faster IDE controller are you loading the
> right kernel driver to take advantage of the speed available? Is there
> a driver specific for your controller? I know the RAID cards have
> faster controllers and specific drivers since I dealt with the HPT372
> driver on the motherboard I'm using now. My 40Mb specs are not from the
> 133 RAID HPT372 controller but the 100 main board controller.
> Geoffrey wrote:
>> I've been trying to research the various boot parameters available.
>> Particularily regarding ide parms. I find references to a doc
>> README.ide in the kernel source, but I've got two sets of source,
>> neither has the file. I finally tracked one down, but no reference to
>> the ide parms I'm looking for.
>> In particular, I had a box that had an ata66 controller and passing
>> ide=66 was recognized.
>> Following that process I have another box with an ata100 controller,
>> but when attempting ide=100, I receive an error that this is invalid.
>> Valid values are 20-66.
>> I've seen references to ide=100 in searches, so I assumed this was
>> These are 2.4.19 kernels, pretty current.
>> Anyone care to shed some light on this??
> Dow Hurst Office: 770-499-3428
> Systems Support Specialist Fax: 770-423-6744
> 1000 Chastain Rd., Bldg. 12
> Chemistry Department SC428 Email:dhurst at kennesaw.edu
> Kennesaw State University Dow.Hurst at mindspring.com
> Kennesaw, GA 30144
> *Computational Chemistry is fun!*
> Ale mailing list
> Ale at ale.org
Ale mailing list
Ale at ale.org
More information about the Ale