[Click] used count after click-uninstall

[Abel Navarro]

Hi Eddie,

I configured Click with the default options (no extra parameters). Also have 
tried with "--enable-dmalloc" and "--enable-kassert" getting the same 
results. We are only using modules from Click. We have not coded any element 
neither modified any of them.

The configuration first loaded works properly all the time but something gets 
wrong with the module usage count. I don't know how to face the problem and 
appreciate any ideas you can provide.

About Endace, I don't know how much can explain, but surely it's based on the 
Click router project, which we found wonderful (it brings fresh air to the 
routing/forwarding scene).

Abel

On Saturday 06 November 2004 12:57, Eddie Kohler wrote:
> Abel Navarro wrote:
> > Hi Guys,
> >
> > When I uninstall the click kernel module, the use count of the module is
> > 2, and that prevents the module to be unloaded. I've been looking at the
> > mailing list archive for some ideas and found a case almost like mine
> > (not solved):
> >
> > https://amsterdam.lcs.mit.edu/pipermail/click/2004-April/002680.html
> >
> > The sympthoms are the same, but I'm using click-1.4.1 and also tried the
> > latest version of the cvs.
> >
> > When I load the test configuration on the click distribution
> > (test.click), the click kernel module reports an usage of 8. More complex
> > configurations report higher module usage (in my case 49). But in all
> > cases the remaining use count after "unloading" is 2.
>
> Hi Abel,
>
> Can you report the options you used to install Click?  And did you include
> any elements of your own?
>
> Endace, huh?  Cool!  :)
> Eddie
>
> > The output from click-uninstall is:
> >
> > # click-uninstall -V
> > click-uninstall: Installing blank configuration in kernel
> > click-uninstall: Removing Click module
> > click: Device or resource busy
> > click-uninstall: cannot remove Click module from kernel
> >
> > Dmesg reports the kernel module loaded, but doesn't reports wether is
> > unloaded.
> >
> > Any ideas are welcome.
> >
> > Thanks
> >
> > Abel
> >
> >
> > _______________________________________________
> > click mailing list
> > click at amsterdam.lcs.mit.edu
> > https://amsterdam.lcs.mit.edu/mailman/listinfo/click