| ??? 07/08/09 19:31 Read: times Msg Score: +1 +1 Informative |
#166881 - internals of SDCC Responding to: ???'s previous message |
Sorry if I wasn't soon enough, but here goes...
This requirement of needing to know about the ISR's when compiling the file with main() is a shortcut from earlier days to circumvent the use of .org statements. When SDCC encounters the implementation of main() it also inserts startup code into the generated asm. The reset vector and interrupt vectors are also a part of it. So SDCC needs to know all ISR's at that time. The .org statement is only allowed in absolute code segments and this prevents relocating the code. When time allows I will one day remove this requirement and let SDCC issue the vector when it encounters the ISR implementation. Maarten |
| Topic | Author | Date |
| Pseudo timers make programming delays easy. | 01/01/70 00:00 | |
| volatile + racing condition | 01/01/70 00:00 | |
| slow processors | 01/01/70 00:00 | |
| You beat me to it... | 01/01/70 00:00 | |
| Timers_0.1 available. | 01/01/70 00:00 | |
| SDCC | 01/01/70 00:00 | |
| ISR defining with SDCC | 01/01/70 00:00 | |
| oh, I just read it in the manual | 01/01/70 00:00 | |
| only conditionally, as #ifdef SDCC | 01/01/70 00:00 | |
| SDCC and ISRs | 01/01/70 00:00 | |
| Prototyping ISRs | 01/01/70 00:00 | |
| you can see it as if.... | 01/01/70 00:00 | |
| SDCC Quirk? | 01/01/70 00:00 | |
internals of SDCC | 01/01/70 00:00 | |
| duh | 01/01/70 00:00 | |
| Too quick | 01/01/70 00:00 | |
| I see something else... | 01/01/70 00:00 | |
| That helped. | 01/01/70 00:00 | |
| Oops! Timers_0.2 available. | 01/01/70 00:00 | |
| you persist | 01/01/70 00:00 | |
| Good idea! | 01/01/70 00:00 | |
| atomicity | 01/01/70 00:00 | |
| No | 01/01/70 00:00 | |
| I gladly, click on a link .... | 01/01/70 00:00 | |
| Direct link | 01/01/70 00:00 | |
| that was clearly possible, I wonder why ... | 01/01/70 00:00 | |
| one more thing, now we are digging deep | 01/01/70 00:00 |