New semantics of our interrupt devices

[sy2002]

• We need to document these semantics somewhere (possible interrupt device documentation, best practice programming, ...)
• One new best practice is, to set the ISR to the own RTI before disabling an interrupt.
• The timer will be disabled if any of the two registers prescaler or timer is zero.
• The VGA scanline interrupt will be disabled, if bit 15 of the scanline register is set.
• Test programs (timer test, fancy, ...) need to be adjusted
• Emulator needs to be adjusted
• sysdef.asm (and comments within sysdef.asm)?
• Others?

[sy2002]

We must not forget to adjust the assembler and C test programs to adhere to these new semantics and to our new ISR best practices. I adjusted test_programs/fancy.asm but all the others are still adhering to the old semantics.

[sy2002]

We also must not forget to adjust all assembler and C test programs to activate interrupts at first because interrupts are now disabled by default (see #139 (comment))

[MJoergen]

I suggest the Monitor enables global interrupt.

[sy2002]

@MJoergen Can you take this one and finalize it? The remaining TODOs as far as I oversee them correctly are:

• Hardware: Adjust the VGA to the new semantics
• Hardware: Adjust the timer module to the new semantics and also adjust the comments in the two file's headers as they describe the old semantics
• Make sure the Monitor enables the global interrupt
• Adjust fancy to the new semantics
• Adjust all other interrupt test and demo programs in assembler and in C