Re: read_proc issue

[Laurent]

> I've encountered this problem before, too.  What is the "One True Way"
> to do this cleanly?  In other words, if you want to do a calculation
> once every time someone runs "cat /proc/foo", what is the cleanest way
> to do that?  The solution we came up with was to check the file offset
> and only do the calculation if offset == 0, which seems pretty
> hackish.

I've tried it and... it works ! :)
Many many thanks :)

In the meantime, I've also followed Tommy Reynolds' advice to not modify 
global state variables within read_procmem. I've intercepted a syscall which 
does the calculation (I've used the open syscall since it allowed me to 
increase the counter by just running vi on any file ;) ) and put it into a 
buffer which is dumped when the /proc entry is read. Works great that way too.
By the way, as the final module will intercept syscalls like open, creat, 
close, link, unlink, mkdir, etc. , I'm wondering if there'll be a dramatic 
negative impact on file operations performance. Is there any efficient method 
to measure this ?
In any case, thanks for all the help you gave me :)

Regards, 
Laurent Sinitambirivoutin
[e-mail]