Re: System Call
From: Josef Moellers (josef.moellers_at_fujitsu-siemens.com)
Date: 12/20/04
- Next message: Peter T. Breuer: "Re: System Call"
- Previous message: Kasper Dupont: "Re: System Call"
- In reply to: Kasper Dupont: "Re: System Call"
- Next in thread: Kasper Dupont: "Re: System Call"
- Reply: Kasper Dupont: "Re: System Call"
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]
Date: Mon, 20 Dec 2004 16:17:08 +0100
Kasper Dupont wrote:
> Josef Moellers wrote:
>
>>but the kernel has the penalty of an indirection
>>when calling "Functions exported from a module"!
>
>
> I don't think that is even possible.
>
Then what's the use of a module exporting functions if there's noone who
can call them?
But you're right: the kernel will not be able to call a function merely
because it's exported (the addresses will not automagically find their
way into the kernel _code_), but those functions exported will be
usually be registered somewhere (e.g. by hooking up to the
scsi-mid-layer or the block-device-layer code) and then called by kernel
code.
I.e. access from the kernel to code within a module has the penalty of
an indirection.
-- Josef Möllers (Pinguinpfleger bei FSC) If failure had no penalty success would not be a prize -- T. Pratchett
- Next message: Peter T. Breuer: "Re: System Call"
- Previous message: Kasper Dupont: "Re: System Call"
- In reply to: Kasper Dupont: "Re: System Call"
- Next in thread: Kasper Dupont: "Re: System Call"
- Reply: Kasper Dupont: "Re: System Call"
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]
Relevant Pages
|