Messages in this thread | | | Date | Sat, 6 Sep 2008 23:07:50 +0200 (CEST) | From | Thomas Gleixner <> | Subject | Re: [RFC patch 0/4] TSC calibration improvements |
| |
On Sat, 6 Sep 2008, Linus Torvalds wrote: > On Sat, 6 Sep 2008, Thomas Gleixner wrote: > > > > If Alok has the second check in place and is actually worried about > > that 288us impact, then we can add the following (untested), which > > does not impact the speed of the check. > > Guys, please. > > Show some _taste_.
Tell the hardware dudes who made that crap so difficult
> Dammit, stop adding random crap to "native_calibrate_tsc()" and make it > look like total and utter SHIT. > > If you want to do that > > tsc1 = tsc_read_refs(&ref1, hpet); > .. > tsc2 = tsc_read_refs(&ref1, hpet); > > around calibration and comparing it, then do it *once*. Do it over the > whole thing. Do it in a function of its own, instead of making this > horrible and unreadable mess.
Over which _whole_ thing ? You want to have the very very fast thing, which is not reliable under all circumstances as Alok pointed out and I merily added a sanity check around that for testing.
> This patch may be fine as a "let's check if it works" thing, but please > don't send out total SH*T to public lists.
Why not ? We want to figure out if it solves the problem and sending it to public lists is the fastest way to get it tested.
> Some _tasted_ in programming, please!
What we apply finally is a totally different thing.
Thanks,
tglx
| |