SGI: Hardware

kshuff wrote: Try it with a set of digital speakers...you heard mine

I can't quite hear them here - what kind do you have?

recondas wrote:

kshuff wrote: Try it with a set of digital speakers...you heard mine

I can't quite hear them here - what kind do you have?

I meant Dieter, he seen mine in person. I've got a set of Boston Acoustics BA790's, sweet.

kshuff wrote:

recondas wrote:

kshuff wrote: Try it with a set of digital speakers...you heard mine

I can't quite hear them here - what kind do you have?

I meant Dieter, he seen mine in person. I've got a set of Boston Acoustics BA790's, sweet.

Well, when my crazy grandmother finally kicks the bucket, I'll have to stop by your place.

(she lives in Allentown, I live in Orlando)

MooglyGuy wrote: Well, when my crazy grandmother finally kicks the bucket, I'll have to stop by your place.

(she lives in Allentown, I live in Orlando)

Door is always open

I had a stab at getting an Octane PVO working with an MXE set in an O2k tonight - no dice. Although the XIO 'connects' and the PVO shows up in a lot of places, one place it doesn't show up is in the kernel drivers.

I followed the trail down into /usr/cpu/sysgen and discovered that the Octane PVO drivers installed into IP30boot instead of IP27boot , and if I craftily copied the driver over from the IP30 directory to the IP27 directory, I get linker errors referring to the gioio and giobr modules when I try to relink the kernel. Copying gioio.o and giobr.o from another machine doesn't do a thing, and as soon as I remove evo.a from the IP27boot directory, the kernel links right up.

I also softlinked the direct node in the /hw tree to /dev/evo , and this causes videod to attempt to load the driver, but it craps out without doing so and only acknowledges it tried, likely due to the missing kernel driver. No, the videod doesn't run without a valid video device in the system.

Not looking really good - anyone have any comments or suggestions? I can't really imagine that an O2k and an Octane - although architecturally similar - are close enough to let you jambalaya together a working kernel setup.

Dr. Dave wrote: Not looking really good - anyone have any comments or suggestions? I can't really imagine that an O2k and an Octane - although architecturally similar - are close enough to let you jambalaya together a working kernel setup.

I don't know, I have'nt futzed with it in a while. I ran across the same situation, I had to force an IP30 install of the driver, but I didn't try copying over the relevent stuff to the IP27 dir. Did you also install the dmedia and impactdm stuff? I just ran across this with an Octane and forgot to install those as well.

kshuff wrote:

Dr. Dave wrote: Not looking really good - anyone have any comments or suggestions? I can't really imagine that an O2k and an Octane - although architecturally similar - are close enough to let you jambalaya together a working kernel setup.

I don't know, I have'nt futzed with it in a while. I ran across the same situation, I had to force an IP30 install of the driver, but I didn't try copying over the relevent stuff to the IP27 dir. Did you also install the dmedia and impactdm stuff? I just ran across this with an Octane and forgot to install those as well.

Yup (as far as I can tell). Still no dice without the kernel driver.

So, did everyone give up on the IO6 mod and instead fall back to buying an IO6G or IOC3?

I'm working on modding an O2k that I just bought, and would like to just solder on the PS/2 connectors. (I have neither an IO6G nor an IOC3, but I do have a shoebox, so I could hold either. If anyone has a spare of either, and might want to trade one for O2k parts or XIO options, let me know.) I see from the discussions in the past that the keyboard and mouse appear to work in the PROM monitor, so that indicates there's chip-level support, anyway.

I also see that some people have a lot of problems even getting IO6G or IOC3 working at first, so maybe it's just a matter of figuring out which drivers need to point where.

Anyone have any comments on this?

Thanks,
Chris

Has this issue died out and been buried?

My best guess is that there is some sort of firmware PROM on the IO6, and even though there may be chip level support, at the OS level the IO6 doesn't 'know' about the ports.

IO6-G here finally, though the CadDuo-in-a-shoehorn works fine too with appropriate config changes.