Messages in this thread | | | From | roque@di ... | Subject | Re: Wrong use of IRQ-handling in ftape? | Date | Sat, 16 Mar 1996 00:41:26 +0100 |
| |
>>>>> "Hans" == Hans Georg Zezschwitz <redjack@science-products.com> writes:
Hans> Hello, I noticed the change in request_irq and free_irq Hans> together with the introduction of shared interrupts.
Hans> As writing a section of the Kernel Korner in Linux Journal, Hans> I wondered what the new "device id" meant.
Hans> As Shared Interrupts are a concept supported by PCI, which Hans> introduces device ids as well, I saw those to concepts Hans> together and believed that the device id should is only Hans> useful when SA_SHIRQ is set (when calling Hans> request_irq). Howether, when looking at patch-1.3.74.gz, I Hans> noticed that the calls to "request_irq" and "free_irq" don't Hans> support shared irqs, but anyway set the device id. This Hans> should not be harmful, but who understood the new Hans> irq-concept wrong, me or the "patcher"?
Beeing pratical, the new dev_id field adds a new void * to the irq action struture that is very handy to use in device drivers even if you are not sharing interrupts since it allows you to get a pointer to your dev struture without going through an irq_to_dev map (these days device drivers are designed to support multiple instances of a driver).
I've checked the code and it seams to me that the kernel doesn't make any assumptions on the dev_id field so is up to the driver to use it as a pointer to whatever it wants....
I can be wrong of course....
regards, Pedro.
| |