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

Re: dpt_i2o and i2o_block on amd64 etch

Lennart Sorensen wrote:

On Fri, May 11, 2007 at 02:01:21PM -0500, Neil Gunton wrote:
I installed a server back in 2005 which has an Adaptec SmartRaid 2015S card. I had problems installing AMD64, which I eventually worked around and documented here: 

http://lists.debian.org/debian-amd64/2005/09/msg00201.html
Now I am considering rebuilding the server from the ground up with Etch/AMD64 netinst (repartitioning, replacing disks). I would like to install, if at all possible, without going through the song and dance that was required back in 2005 (bootstrapping off CentOS etc). 

So, I am wondering if anybody can tell me whether:

a) dpt_i2o is now included with the stock netinst amd64 kernel, or
b) i2o_block is now including and working with stock netinst kernel


Well the 2.6.18 kernel on Etch amd64 has:
CONFIG_I2O=m
CONFIG_I2O_LCT_NOTIFY_ON_CHANGES=y
CONFIG_I2O_EXT_ADAPTEC=y
CONFIG_I2O_EXT_ADAPTEC_DMA64=y
CONFIG_I2O_CONFIG=m
CONFIG_I2O_CONFIG_OLD_IOCTL=y
CONFIG_I2O_BUS=m
CONFIG_I2O_BLOCK=m
CONFIG_I2O_SCSI=m
CONFIG_I2O_PROC=m

... so I think i2o_block should be covered.  I seem to recall dpt_i2o
isn't 64bit safe and I don't think anyone has tried to fix it.
I don't really mind which driver it wants to use, as long as I can get things running again, ideally with a quick, straightforward install. The server is colocated and so this will be a road trip up to Chicago. I'd like to know roughly what to expect. 

Looks like it should work.  The kernel on sarge never had the i2o stuff
enabled so it was a pain for sure.


Hi Lennart,
Thanks very much, that's reassuring. I still have a tiny doubt, however, since I remember CentOS x86_64 did not recognize the disks at first, but then gave you an option during the install to manually select a driver, which fortunately included i2o_block. Debian, as I recall, never had such an option.
So, obviously, it was possible for the kernel to have the module be available, but not automatically recognized as "the" module to use.
Can we be fairly certain that Etch will actually recognize these disks automatically, or is it possible that the driver would be in the same state as the old CentOS - present, but somehow not automatic? And if that happened, is there a way now for me to tell Debian to use it during the boot?
Sorry if this is a dumb question, it's just I am so used to weird things that take me by surprise... don't want this road trip to be another epic experience. 

Thanks again!

/Neil
Reply to: