[Top] [All Lists]

RE: __initfunc(), kernel coding logic?

Subject: RE: __initfunc(), kernel coding logic?
From: Richard van den Berg <>
Date: Fri, 3 Jul 1998 20:34:38 +0200 (MET DST)
In-reply-to: <>
Hi Harald,

On Thu, 2 Jul 1998, Harald Koerfgen wrote:

> Like Ralf already said: __initfunc(...) just puts the following code into a
> special text section which gets freed after initialization. This has nothing 
> to
> do with the circumstances under which your code is beeing called.

__initfunc() is used in sunlance.c and tc.c and much more drivers so
I thought that was the right place to start. But being called, the code
must be there and it seems that the code isn't in the kernel... Like I
wrote Ralf I put 3 other printk()'s which show at boot, the text of these
printk()'s can be found if I look at the image with Midnight Commander,
but the sentence I wrote is gone. The code I posted was all the code
I experimented with, first I like to discover how to add code to the
kernel. For example I have been messing around with the Makefile's
to change order without any result, in other words, the kernels boot
mechanisms aren't very clear to me.

> If you want to do some experiments with a net driver you may want to edit
> drivers/net/Space.c accordingly.  

Already did that, without any result.

> Hope this helps.

Thanks! Think I can use it...



<Prev in Thread] Current Thread [Next in Thread>