Messages in this thread | | | Date | Wed, 2 Jan 2008 21:23:15 +0100 | From | Ingo Molnar <> | Subject | Re: [PATCH] x86: kprobes change kprobe_handler flow |
| |
* Abhishek Sagar <sagar.abhishek@gmail.com> wrote:
> > > + default: > > > + /* impossible cases */ > > > + BUG(); > > > > I think no need to check that here. > > It may not get hit at runtime but is quite informative.
sidenote: please use WARN_ON(1) instead. In the case of the impossible triggering somehow (the Linux kernel is notorious for that ;-), we'd have a nice WARN_ON(1) in the dmesg, instead of a crashed box.
Ingo
| |