My MLTP EAC Hack

While Tom had been writing his latest prank, I had, independently, developed one of my own.

As I said, the IBM 1800 equivalent of a Black Screen of Death was a MuLTiPle Error Alert Control event. A MLTP EAC event, which we called simply "MLTP EAC", involved the system executing the Error Alert Control program, and during that execution, generating a second error. EAC was not written to be reentrant, so, if it was executed twice before completing the first, would retreat to the MLPT EAC sequence of code.

An MLTP EAC sequence would simply, oh so simply, take an 8 character (4 word) section of memory, and type that 8 characters on the system console. That 8 characters was coded, by TSX, as

MLTP EAC

The MLTP EAC message intended to simply inform the operator of the computer that an unrecoverable system error has occurred, and a hard system boot (Power Off To Reset) was now required.

Now, there was nothing magical about that terse 8 characters. It was coded into the system, when the system was compiled. TSX being very primitive, it took not a lot of work to locate and patch those 8 characters, and I wrote a pair of programs:

1. Customise the "MLTP EAC" character string to one of our own choosing.
   
2. Generate a MLTP EAC sequence (ie, crash the system).
   

>>More