Re: [PATCH v8 3.2.0-rc5 1/9] uprobes: Install and remove breakpoints.

[Steven Rostedt <rostedt@goodmis.org>]

On Wed, 2012-01-04 at 17:51 +0100, Peter Zijlstra wrote:

> > +               if (is_register)
> > +                       ret = install_breakpoint(mm, uprobe, vma, vi->vaddr);
> > +               else
> > +                       remove_breakpoint(mm, uprobe, vi->vaddr);
> > +
> > +               up_read(&mm->mmap_sem);
> > +               mmput(mm);
> > +               if (is_register) {
> > +                       if (ret && ret == -EEXIST)
> > +                               ret = 0;
> > +                       if (ret)
> > +                               break;
> > +               }
> 
> Since you init ret := 0 and remove_breakpoint doesn't change it, this
> conditional on is_register is superfluous.

True, but I would argue that this is easier to understand. That is, we
only break on a failed install_breakpoint (is_register is set). If I
looked at this code and saw:

	if (is_register)
		ret = install_breakpoint()
	else	
		remove_breakpoint()

	[...]

	if (ret && ret == -EEXIST)
		ret = 0;
	if (ret)
		break;

I would first think that there might be a bug. That is, we should have a
ret = remove_breakpoint().

Thus, I would say, either leave this as is and hope gcc is smart enough
to optimize out the if (is_register), or add the comment:

	/* ret will always be zero on remove_breakpoint */
	if (ret && ret == -EEXIST)
		ret = 0;
	if (ret)
		break;

-- Steve

> 
> > +       }
> > +       list_for_each_entry_safe(vi, tmpvi, &try_list, probe_list) {
> > +               list_del(&vi->probe_list);
> > +               kfree(vi);
> > +       }
> > +       return ret;
> > +} 


--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@kvack.org.  For more info on Linux MM,
see: http://www.linux-mm.org/ .
Fight unfair telecom internet charges in Canada: sign http://stopthemeter.ca/
Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>