Quoting [EMAIL PROTECTED] ([EMAIL PROTECTED]): > I suppose there are cases where this is valid, but for my original point about > communciations - it just isn't feasible to use IRQ4 (the first comm port) by > more than one program concurrently, is it? > > In my case, I had reprogrammed IRQ4 so when a char came in from the modem, I > could stuff it to a buffer for use later on, and send data to the buffer (and > ultimately use the IRQ to send it back out again when the buffer was full) > when I had my reply ready. I certainly wouldn't want another program taking > chars from my modem while I'm using it.
And that's just what the kernel can enforce through its privilege. > How would you accomplish the same in Linux without using interrupt routines? Take a look at /usr/src/linux/drivers/char/serial.c for how linux does it. It's not that linux doesn't use interrupt routines, it's just that you (as a user in a user process) don't run them directly. You (as a kernel hacker) can do it in kernel-space within a device driver, like serial.c. If you take the kernel source as a model, it's not that difficult to write a simple device driver for a simple device, like, for example, a card I have for driving a motorised specimen stage. You just have to stick to the rules when interacting with the kernel. It's also about the only C programming I've ever bothered with. Cheers, -- Email: [EMAIL PROTECTED] Tel: +44 1908 653 739 Fax: +44 1908 655 151 Snail: David Wright, Earth Science Dept., Milton Keynes, England, MK7 6AA Disclaimer: These addresses are only for reaching me, and do not signify official stationery. Views expressed here are either my own or plagiarised.
