Re: [RFC] [PATCH 1/2 v2] x86: introduce int3-based instructionpatching

[Jiri Kosina]

On Thu, 11 Jul 2013, Steven Rostedt wrote:

> > The current code assumes that one of the two code sequences is a NOP,
> > and therefore that jumping over the region is legal.  This does not
> > allow for transitioning one active code sequence to another.
> 
> Correct, and I think we should keep the two changes separate, as the NOP
> case is trivial. No need to complicate the trivial and common updates
> (jump_labels and ftrace). But for things like kprobes, we could do a bit
> more complex code, but it should probably be separate.
> 
> Perhaps call this text_poke_nop_bp()?

Hmm ... I don't think this is exactly precise, at least as long as the 
implementation in the patchset I have submitted is concerned.

Yes, most use cases (jump labels, perhaps ftrace) will simply be skipping 
over the patched region, pretending that NOP has been there; but the 
handler provided to text_poke_bp() is completely free to do any other kind 
of trickery.

The one that jump label provides in PATCH 2/2 really just skips over the 
region, yes. But the interface potentially allows for more.

-- 
Jiri Kosina
SUSE Labs
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/