Hi,
bluemoon wrote:Geometrian wrote:It does not make sense to spend time overengineering a collection of industry-quality individualized interrupt handlers when I can't even get a trivial example to work properly!
You miss the point. It is however the opposite, using single entry and followed by a switch case is indeed overcomplicate the logic,
and provide individual interrupt handler for each, and only such IRQ/exception, is considered trival in general sense.
Anyway, all we can do is tell the story from our experience, to give a bit more information on the subject, it's still you to make the design choice.
Yes.
Imagine if, instead of writing an OS, you were trying to travel 50000 Km to the north.
First you've cut & pasted a bunch of code from a tutorial (which was only intended to help explain something and was never intended for an actual OS). This is like catching a fast bus to a town 150 Km to the west.
Then you've got a bug (possibly from not really understanding the code because it was cut & pasted). So far it's been about 3 days of trying to fix it. This is like waking another 150 Km to the west.
Once the bug is fixed, you'll start implementing more code that depends "tutorial interrupt handling". This will be another 200 Km of walking west.
None of this travelling west will get you any closer to your goal in the north.
I can understand that you are learning things on your journey west, and this learning might make travelling west seem useful. However, you'd learn the same things if you were actually travelling north.
Cheers,
Brendan
