Hi, >> ... why we don't provide a >> generic interrpt wrapper as part of the C library? > Maybe because it's too slow to do anything serious? Is from your perspective the overhead caused be the wrapper to high or is the called C code too slow? Anyway on the Apple2 the overhead caused by the firmware plus OS is already quite high. Compared to that the overhead of the wrapper is rather low... > At least this is one > reason that caused me not to offer such a function. Someone with enough > knowledge can easily adapt the functions from the debug module. All others > should better not use interrupt handlers in C. Surely, but that code isn't specific to the task at all. So the question is why do it again and again... > That said, I'm not against such a function if there is public demand. I need it once so I'll implement it. I personally don't need it in the C library. I'm offering to try to implement in a way suitable for library integration. That's all. > What I > fear is people calling printf and similar stuff and messing up the runtime > environment. Or doing stuff that causes overruns of the interrupt handler > thread. These problems are quite nasty. It works ok a hundred times and then > fails randomly. That sort of reminds me of me avoiding comments in complicated code sections to keep others from changing my code if they don't really know hat they're doing ;-)) However I've been told that this isn't the right way... > No need to fail. Just use a function vector that may be set any times. But that makes it slower ;-) > See above. What I fear are hordes of people experiencing interrupt handler > problems. Writing interrupt handlers in C is actually *more* difficult than > writing the in assembler. Hm, I think that depends heavily on the task at hand. In my case I've got an existing C code designed from the start to run in an interrupt handler - it doesn't even call memcy(). But it's quite a lot and it contains a quite some if()s, while()s and switch()es and typed array addressing operations and pointer arithmetic. I don't see why I should translate all that into 6502 asm if I have cc65 at hand... Anyway, I'm not into a discussion on this topic. If there's no momentum then I'll just do it locally for my current task. Regards, Oliver ---------------------------------------------------------------------- To unsubscribe from the list send mail to majordomo@musoftware.de with the string "unsubscribe cc65" in the body(!) of the mail.Received on Wed Jan 18 00:02:31 2012
This archive was generated by hypermail 2.1.8 : 2012-01-18 00:02:34 CET