Re: Tracking adresses of write and read

[Fabrice Gautier <gautier@email.enst.fr>]

On Wed, 16 May 2001 18:50:53 +0200
Erik Mouw <J.A.K.Mouw@ITS.TUDelft.NL> wrote:

> On Wed, May 16, 2001 at 05:34:06PM +0200, Fabrice Gautier wrote:
> > On Wed, 16 May 2001 15:24:05 +0200
> > Erik Mouw <J.A.K.Mouw@ITS.TUDelft.NL> wrote:
> 
> No. The userland program shouldn't even *care* about the physical
> address. As long as the kernel does the physical to virtual mapping
> correct the userland program should work.

That's what i'm trying to see: Where does the kernel physically write
when my program write in the mmaped area. I don't want the user space program
to handle physical adresses. It just care about it for debugging
purposes, I just want the physical adres as information.

> 
> > I think both virt_to_phys and printk are kernel functions, not syscall
> > or libc function.
> 
> What did you expect on a mailing list with "kernel" in its name? :)

Don't know, maybe there's a kernel syscall, some way to intercept the
memory acesses so that maybe i can use virt_to_phys, some sort of kernel
hack, some way to use a kernel debugger to get this information, some
way for the user space program to access kernel funtions or structures,
anything I could use to get the information, but i don't know...

Since you point out that a user program doesn't normally have to care
about physical adresses, I guess the possible solution have something to
do with kernel...

Thanks, and regards,

-- 
Fabrice Gautier <gautier@email.enstfr>

-
Kernelnewbies: Help each other learn about the Linux kernel.
Archive:       http://mail.nl.linux.org/
IRC Channel:   irc.openprojects.net / #kernelnewbies
Web Page:      http://www.kernelnewbies.org/