Conclusion

Since this was a customer's actual crash dump, we had to contact the customer with our findings. Depending on the customer and his understanding of what we were faced with, the conversation could have been uncomfortable at best because we really didn't have the problem figured out.

The tail end of the telephone conversation went something like this:

"So, let me see if I get this right. You're telling me that either we have a hardware problem or we have a software problem?" the customer wisely summed up the news.

"Yes, that's what we believe."

"And, if I understand you correctly, you're telling me that my system will crash again?" he further asked.

"Um, yeah, that's our feeling."

The customer laughed!

He understood exactly what we had found and appreciated what we were faced with. And, thanks to our analysis, he was now armed with enough information to go back to his device-driver writers and ask for their assistance.

He was also now up to date with Solaris 2 issues and fixes to discovered bugs , especially in the semsys module.

It's been over a year since we've heard from this customer about this system. His hardware is fine, and we suspect his new device drivers are coming along nicely .