Cripples bus on VAXstation 2000

[majenkotech]

I am attempting to get SCSI working on the VAXstation 2000. This SCSI is strange, though. It was originally only intended to operate with a TK50 tape drive, but there are VMS patches and ROM patches available to make it work with other SCSI devices too.

Connecting the BlueSCSI (v2, desktop, nightly universal build) cripples the SCSI bus though (no devices get found, and all IDs just return a hard error).

I have two theories:

1. The SCSI is just so old and broken it can't work with the BlueSCSI, or (more likely)
2. The host initiator is fixed on SCSI ID 0 rather than the "normal" ID of 7.

There are ROM patches that allow changing the initiator ID to something else, but I don't have those burned to chip yet (I am fighting with my EPROM programmer, it's about 30 years old....).

Could there be an issue with the BlueSCSI whereby it doesn't like the initiator being on ID 0?

(BTW, for testing purposes I connected the BlueSCSI to a SUN Ultra60 and it worked fine, so my hardware is not at fault).

[erichelgeson]

BlueSCSI doesn't care what the initiator ID is, it can be 0, or 7 or anything in between. Just cant overlap with an existing device.

Please use the current release instead of the nightly so we have a stable base to debug on. Please completely go through the "What to try first" in the troubleshooting guide (even if you dont think it's applicable, again for a stable base to test on)

Then please provide a debug log so we can see what is going on.

Also note there are some notes in the wiki about VAX but not your specific model https://github.com/BlueSCSI/BlueSCSI-v2/wiki/Compatibility#dec--compaq

[majenkotech]

Ok, I have had a quick probe of the SCSI bus and it is just completely dead. All I see is a brief blip on some of the control lines, then a reset, and that's all.

All the other signals remain high at all times. Just ATN, BSY, ACK and SEL pulse low briefly, followed by a longer period of RST and SEL being low. And then it fails. I'm going to try probing the bus with the CDROM attached to see what differs in that initial section if I can arrange some form of connection to it (atm I'm connected direct to the bluescsi)

---

Ok. This paints a very very different picture. For a start /RST never gets asserted.

/ACK and /SEL both remain low after /ATN and /BSY both rise, which I guess means "There's someone here" (though I am not conversant in SCSI).

The /RST being asserted might be a way of the initiator trying to get everything into a known good state, but then it doesn't try anything again, possibly because /SEL remains low for a moment after /RST rises? Not knowing the protocol I am just stabbing in the dark here.

[majenkotech]

Ok. More insights. I have had the RST being asserted with the CDROM arrangement. In that one it is only the RST that is asserted after the ATN/BSY/ACK/SEL phase, not both RST and SEL. And the VS2K aborts immediately it gets that, before it even tries probing the bus for what is there. So my guess is it's the SEL being asserted with RST that is the crippling problem here.

[majenkotech]

There's the good boot first bit in question.

[majenkotech]

Any clue as to what might be the underlying cause here? I have tried looking at the code but I am completely lost with this one.

[erichelgeson]

Hey @majenkotech - first thanks for all the detailed info! We discussed it briefly and nothing stood out to us right away - but we'll dig a bit deeper - free time to look at complex issues is hard to come by these days but we'll for sure dig into this deeper with you.