Re: [B.A.T.M.A.N.] [PATCH 3/3] alfred: Mount debugfs before reducing capabilities

[Andrew Lunn <andrew@lunn.ch>]

On Thu, Mar 10, 2016 at 04:09:18PM +0100, Simon Wunderlich wrote:
> Hi Andrew,
> 
> On Tuesday 02 February 2016 03:25:12 Andrew Lunn wrote:
> > On Fri, Jan 29, 2016 at 01:14:53PM +0100, Simon Wunderlich wrote:
> > > On Thursday 28 January 2016 04:53:15 Andrew Lunn wrote:
> > [...]
> > > > @@ -160,6 +161,9 @@ static struct globals *alfred_init(int argc, char
> > > > *argv[]) {NULL,			0,			NULL,	0},
> > > > 
> > > >  	};
> > > > 
> > > > +	/* We need full capabilities to mount debugfs, so do that now */
> > > > +	debugfs_mount(NULL);
> > > > +
> > > > 
> > > >  	ret = reduce_capabilities();
> > > >  	if (ret < 0)
> > > >  	
> > > >  		return NULL;
> > > 
> > > Can't we remove the other calls to debugfs_mount() ? I see 3 more calls in
> > > alfred ...
> > 
> > Yes, the other calls within this binary are probably failing, due to
> > reduced capabilities. I can remove them in a v3 patch.
> > 
> > 	Andrew
> 
> Are you still planning on sending a fixed patchset? :)

Hi Simon

I have a fixed up version. But at the moment, it is not clear what is
happening with the kernel patches. Maybe we need to change the debugfs
API in order to make it easier to build with older kernels?

    Andrew