Mark Hounschell schrieb:
Webillo Disperso wrote:
When sharing interrupts as in: 11: 12785 XT-PIC eth0, uhci, bttv, serial (part of 'cat /proc/interrupts' example) the only reason for delay is the kernel, after having called one source irq routine, must ask "any other irq of same number pending?".
Then why does my dual 800Mz intel pIII box show 17, and my 2 dual 1.7Gz intel p4 zeons show 19? I recently read in some PCI spec book that there were 256 available and the REAL limitation was in the vendor bios's. In fact we are considering having a custom bios done for us for these dual 1.7 boxes and they say they can support 256 irq's with it. ???
Mark
Yes, this makes me both look up and also express ?????????????, because from what I remember (True, that's based on old ISA architecture) there is a dedicated line from the card running to the cpu, telling the CPU an IRQ has occured. (hardware). Following this theory, there'd be a need for 256 lines... I'd love to have more than these 16 IRQ's for the same obvious reasons, so I am jumping on the thread to learn more... Juergen -- =========================================== __ _ Juergen Braukmann juergen.braukmann@gmx.de| -o)/ / (_)__ __ ____ __ Tel: 0201-743648 dk4jb@db0qs.#nrw.deu.eu | /\\ /__/ / _ \/ // /\ \/ / ===========================================_\_v __/_/_//_/\_,_/ /_/\_\