Re: [PATCHSET 00/13] tracing/uprobes: Add support for more fetch methods (v6)

[Oleg Nesterov <oleg@redhat.com>]

On 11/05, Namhyung Kim wrote:
>
> On Mon, 4 Nov 2013 16:01:12 +0100, Oleg Nesterov wrote:
> > Or the syntax should be "name=probe @file/addr" or something like this.
>
> Okay.  Let's call this kind of thing "cross-fetch" (or a better name can
> be suggested).

Yes ;) and I am afraid there was some confusion in our discussion.
I probably confused "probe other binaries" with cross-fetch and vice
versa sometimes.

> This is more complex situation and needs more discussion
> as you said.  So let's skip the discussion for now. :)

Agreed.

> > So far I think that trace_uprobes.c should not play games with vma. At all.
>
> Yes, playing with vma is fragile.  But otherwise how can we get the
> address from the file+offset in random processes?

Yes, this is not as simple as I thought.

Let me repeat, somehow I completely forgot we need to probe other (libc)
binaries (not only the executable itself) and dump the data from that
binary. That is why I wrongly thought that that ->start_data trick can
work.

OK, I see other emails from you. Perhaps we can rely on instruction_pointer(),
(I'll write more emails on this), But if not, then we probably need tu->inode
and vma games in fetch/get_user_vaddr(). I'd still like to avoid this, if
possible, but I am no longer sure.


> >> > Can't we simply implement get_user_vaddr() as
> >> >
> >> > 	static void __user *get_user_vaddr(unsigned long addr, struct trace_uprobe *tu)
> >> > 	{
> >> > 		void __user *vaddr = (void __force __user *)addr;
> >> >
> >> > 		/* A NULL tu means that we already got the vaddr */
> >> > 		if (tu)
> >> > 			vaddr += (current->mm->start_data & PAGE_MASK);
> >> >
> >> > 		return vaddr;
> >> > 	}
> >> >
> >> This can only work with the probes fetching data from the executable,
> >> right?  But as I said it should support any other binaries too.
> >
> > See above, we can't in general read other binaries.
>
> Okay, I need to clarify my words.  I'm not saying about "cross-fetch"
> here, what I wanted to say is adding a probe in some dso and fetch data
> from the dso.
>
> [...snip...]

Yes, sorry for confusion, see above.

Oleg.