summaryrefslogtreecommitdiff
path: root/PLAN
blob: ffc0ba1c6eed7230393f30c6e01675f764071594 (plain) (blame)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
Turn on blinkenlights from the PDP core

There's probably a bug in the UART TX code that's dropping characters when it gets full

Debounce switches

Implement switch features:
    Start
    Load_Add
    Deposit
    Examine
    Continue
    Stop
    Single_Step
    Single_Inst
    The DF* IF* and SR* data switches

Don't ignore 6000 and 6001

Add "interrupts enabled" flag
6000, 6001, 6002 should properly handle it
    6000 - skip next instruction if interrupts are enabled; disable interrupts (we should probably do this immediately)
    6001 - enable interrupts (after next instruction)
    6002 - disable interrupts (we should probably do this immediately)

Jules thinks clock is most likely interrupt source
    pdp8 source tries to map clock interrupts to "real time" - we might want to mangle that to make it deterministic

Sources of interrupts observed:
    PTR flag is set from loader, might be ignorable
    Memory management is poked from inside ISRs, might be ignorable (MMU should be optional)
    TTO ready for output interrupt - this looks like a very likely candidate
    TTI input ready interrupt - this looks like a very likely candidate
        Looks like these happen even when there is no input available - might be caused by the program itself doing something a little funny