Lit. Buffer overruns 1726

So this was a world-wide lapse of hardware thinking. It sure seemed like we had all of them. ;-)

Ours wasn't either. I was just wondering if that had changed. XORing to swap may be "faster" but it sure isn't safer if you can't keep track of sequence when it's interrupted.

Wasn't it put under a feature test switch?

Now I can't remember what changes we had done in the KL-microcode. ISTM there was one or two generated by SMP.

That's not how we did it. We would put the "kernel" and the source on a separate tape with a SPD number and description. Our SMP tape consisted of a directory of the tape, a beware file, CPNSER.MAC, CPNSER.REL.

If MONGEN's question of number of CPUs was greater than 1, CPNSER.REL would get copied into TOPS10.REL and the feature test switch for multi-CPUs in all the other monitor modules would get turned on.

You did backwards from us :-). We had to fight PHBs because they wanted to encode the number of CPUs supported into the file names. This was contrary to JMF's extensibility plans.

BAH

Subtract a hundred and four for e-mail.