Details | Last modification | View Log | RSS feed
| Rev | Author | Line No. | Line |
|---|---|---|---|
| 2 | pj | 1 | 1) Check save_fpu & c... It works, but the code is not very clean... |
| 2 | Clean it. |
||
| 3 | VERY LOW PRIORITY |
||
| 4 | |||
| 5 | 2) Test the interrupt mechanism... DONE |
||
| 6 | Clean the act_int interface... (remove context? Like wake_up...) |
||
| 7 | At least, look at it... |
||
| 8 | MEDIUM PRIORITY |
||
| 9 | |||
| 10 | 3) Time: clean up the time-exact code... Implement the Pentium timestamp |
||
| 11 | code... |
||
| 12 | MEDIUM PRIORITY |
||
| 13 | |||
| 14 | 4) Interface: |
||
| 15 | init(...) look at the parameters |
||
| 16 | event_post(time, handler, parameters) |
||
| 17 | It returns a pointer to an event structure, allocated from |
||
| 18 | a static array... Why not passing the pointer to an event |
||
| 19 | structure (allocated by the user) as a parameter? |
||
| 20 | irq_bind(irq, handler, flags) OK. |
||
| 21 | context_create(task, stack, arg, killer, ctrl) |
||
| 22 | As for event_post... Why not passing a pointer to the |
||
| 23 | context as a parameter? |
||
| 24 | context_delete(c) OK. |
||
| 25 | get_time() Define something like TIME_HIGHRES, for the pentium |
||
| 26 | timestamp... |
||
| 27 | context_save() OK. |
||
| 28 | context_load(c) OK. |
||
| 29 | as_create() |
||
| 30 | OK? |
||
| 31 | as_bind(as, ph_addr, l_addr, size) |
||
| 32 | Interface is OK. It currently provides only half of the |
||
| 33 | functionalities (no paging)... |
||
| 34 | HIGH PRIORITY |
||
| 35 | |||
| 36 | 5) More tests on ASs... |
||
| 37 | HIGH PRIORITY |
||
| 38 | |||
| 39 | 6) X: more MultiBoot compliance... |
||
| 40 | MEDIUM PRIORITY |
||
| 41 | |||
| 42 | 7) Provide a mechanism for system calls (invoked through interrupts?) |
||
| 43 |