Messages in this thread |  | | Date | Mon, 24 Feb 2014 11:50:02 -0800 | From | John Stultz <> | Subject | Re: [PATCH] [RFC] time: Improve negative offset handling in timekeeping_inject_offset |
| |
On 02/22/2014 02:01 AM, Thomas Gleixner wrote: > On Fri, 21 Feb 2014, John Stultz wrote: >> But yes, we can just leave it as is, and it is a bit academic. But > To be honest, I was mostly arguing due to the academic nature. :) > > The important point is that we restrict it to -1e9 < tv_nsec < > 1e9. The signed/unsigned combos are not that interesting as long as > our internal helpers cope nicely with that.
And for this patch it does check abs(tv_nsec) is sane.
I still think the patch is a good idea, so I'll submit it again after further testing. But if folks object, then I'll drop it.
thanks -john
|  |