| 1 |
On 9/30/05, Duncan <1i5t5.duncan@×××.net> wrote: |
| 2 |
> Tres Melton posted <1128118297.6208.265.camel@×××××××××.org>, excerpted |
| 3 |
> below, on Fri, 30 Sep 2005 16:11:37 -0600: |
| 4 |
> |
| 5 |
> > You have had so many problems with the other kernels doing this stuff |
| 6 |
> > and now that you have found a good one I really think that you should |
| 7 |
> > compose a message entitled "Latencies in the kernel" and post a summary |
| 8 |
> > of your experiences with the various kernels to the LKML. I think that |
| 9 |
> > it would benefit the developers (and all of us) to know that the kernel |
| 10 |
> > seems to be tuned rather nicely now to tasking these things. Anyway, |
| 11 |
> > I'm glad that it is working for you. |
| 12 |
|
| 13 |
Yes, I've now moved a bit forward to 2.6.14-rc3-rt1. It seems to be |
| 14 |
working slightly better. On the previous 2.6.14-rc2-rt7 I had a few |
| 15 |
times where I got a short burst of xruns. This happened at sort of odd |
| 16 |
times, but it happened. While I've only run the new one since |
| 17 |
yesterday I've seen this happen fewer times, and with fewer xruns. Not |
| 18 |
sure I trust data like that, but it's what I've seen. |
| 19 |
|
| 20 |
As for writing the LKML, I could do that. I have reported that these |
| 21 |
kernels are working pretty well. I think I'd like to build Con |
| 22 |
Kolivas's kernel from source, just to ensure that any patches put |
| 23 |
there by gentoo are not the cause of the problems I'm seeing. Jack is |
| 24 |
possibly being given real time priorities than these other kernels |
| 25 |
I've just built, or so I think. Not at all sure. Don't want to say |
| 26 |
much more about that until I know more. These email threads have a way |
| 27 |
of lasting for years and becoming urban legends. So many people seem |
| 28 |
happy with ck-sources so I think if it doesn't work for me it must be |
| 29 |
me. |
| 30 |
|
| 31 |
> |
| 32 |
> Note that it's not the vanilla kernel, but the vanilla kernel with the |
| 33 |
> real-time patches. |
| 34 |
> |
| 35 |
> Still, yes, posting something to LKML is a good idea, as one of the |
| 36 |
> ongoing debates has to do with the merits of the various real-time patches |
| 37 |
> as opposed to other factors, and this sort of direct user testimony could |
| 38 |
> be influential. Of course, don't be surprised if you immediately get |
| 39 |
> asked to do testing on RT related stuff! |
| 40 |
|
| 41 |
Yes, I have to watch out for this. I recently signed up to look after |
| 42 |
some audio apps under Wine. I think that will keep me quite busy. |
| 43 |
|
| 44 |
> |
| 45 |
> (Noting that I dont' follow LKML directly, but do follow happenings in |
| 46 |
> general thru LWN's weekly coverage and occasionally find myself browsing |
| 47 |
> various LKML threads as well.) |
| 48 |
|
| 49 |
Thanks for writing. |
| 50 |
|
| 51 |
Cheers, |
| 52 |
Mark |
| 53 |
|
| 54 |
-- |
| 55 |
gentoo-amd64@g.o mailing list |
Archives