Re: Is it really correct to check for breakpoint in kernel space against ptracer's address space?

[Ruslan Kabatsayev]

On Tue, May 10, 2016 at 6:23 PM, Andy Lutomirski <luto@xxxxxxxxxxxxxx> wrote:
> Anyway, what kernel are you on?  I think I fixed your specific issue in:
>
> commit 27747f8bc355a2808ca9e490ab6866acd85b4c16
> Author: Andy Lutomirski <luto@xxxxxxxxxx>
> Date:   Thu Jul 30 20:32:42 2015 -0700
>
>     perf/x86/hw_breakpoints: Fix check for kernel-space breakpoints
>
> If that does fix it and it's a problem for you on older kernels, you
> could ask for a backport.

Indeed, I was using a 3.12 kernel. This should indeed fix it, thanks.

BTW, in that commit, why isn't va>=TASK_SIZE_MAX redundant with the
second operand of ||? If va is aligned to a multiple of len (so
va+len-1 doesn't overflow), then in what cases would the first
condition be true with the second being false?