On Sat, 12 Feb 2005 01:47:27 +0200
Ion-Mihai Tetcu <[EMAIL PROTECTED]> wrote:
> On Tue, 8 Feb 2005 08:39:15 -0800
> Kris Kennaway <[EMAIL PROTECTED]> wrote:
>
> > On Tue, Feb 08, 2005 at 03:58:22PM +0200, Ion-Mihai Tetcu wrote:
> > > On Tue, 8 Feb 2005 14:45:17 +0200 (EET)
> > > "Viktor Ivanov" <
On Tue, 8 Feb 2005 08:39:15 -0800
Kris Kennaway <[EMAIL PROTECTED]> wrote:
> On Tue, Feb 08, 2005 at 03:58:22PM +0200, Ion-Mihai Tetcu wrote:
> > On Tue, 8 Feb 2005 14:45:17 +0200 (EET)
> > "Viktor Ivanov" <[EMAIL PROTECTED]> wrote:
> >
[ ... ]
> > > On the other hand, using SCHED_ULE improves
On Feb 10, 2005, at 18:32, Doug Poland wrote:
On Thu, February 10, 2005 10:39 am, Julio Capote said:
I upgraded to 5-STABLE + ULE + PREEMPT last night, been using my
desktop system heavily for about 6 hours, and no crashing, AWESOME
performance under heavy load. Im curious to see some 1(O) vs ULE
b
On Feb 10, 2005, at 18:22, Mipam wrote:
I noticed this:
You're running RELENG_5 on a dual Xeon 3.06GHz box
You enabled DEVICE_POLLING and ULE as well.
But normally compiling the kernel with SMP and polling won't work.
Don't you use a SMP kernel or?
I saw several pages stating that DEVICE_POLLING ac
On Thu, February 10, 2005 11:43 am, Kevin Oberman said:
>> Date: Thu, 10 Feb 2005 11:32:34 -0600 (CST) From: "Doug Poland"
>> Precisely how does one upgrade 5-STABLE to include ULE + PREMPT?
>> Obviously I'll cvsup RELENG_5, but don't know how to set ULE +
>> PREMPT.
>>
> Put them in your configura
Cvsup to RELENG_5, and change your kernel config ot include ULE and
PREMPTION support; I suggest reading the "keeping up to date" and
"kernel compiling" chapters on the handbook, if you haven't done so
already.
-Julio
On Thu, 2005-02-10 at 11:32 -0600, Doug Poland wrote:
> On Thu, February 10,
On Thu, February 10, 2005 10:39 am, Julio Capote said:
> I upgraded to 5-STABLE + ULE + PREEMPT last night, been using my
> desktop system heavily for about 6 hours, and no crashing, AWESOME
> performance under heavy load. Im curious to see some 1(O) vs ULE
> benches..
>
Precisely how does one upgr
Hi,
I noticed this:
You're running RELENG_5 on a dual Xeon 3.06GHz box
You enabled DEVICE_POLLING and ULE as well.
But normally compiling the kernel with SMP and polling won't work.
Don't you use a SMP kernel or?
Bye,
Mipam.
___
freebsd-stable@freebsd
I upgraded to 5-STABLE + ULE + PREEMPT last night, been using my desktop
system heavily for about 6 hours, and no crashing, AWESOME performance
under heavy load. Im curious to see some 1(O) vs ULE benches..
On Thu, 2005-02-10 at 17:00 +0100, Frode Nordahl wrote:
> On Feb 10, 2005, at 16:30, Rand
On Feb 10, 2005, at 16:30, Randy Bush wrote:
how about stability? last time i tried ule, the system died in
a few days.
Don't know yet.
I will keep trashing the server and post here if it crashes.
Regards,
Frode Nordahl
randy
___
freebsd-stable@freebsd.o
how about stability? last time i tried ule, the system died in
a few days.
randy
___
freebsd-stable@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-stable
To unsubscribe, send any mail to "[EMAIL PROTECTED]"
[ Cc'ing Bill Paul, the author of the xl(4) driver ]
> Also, as far as I understand, some extra boost comes
> from using polling the devices; however, not all
> devices do support this yet.
>
> I use SMP (dual intel CPUs) with if_xl lan card,
> but I can't use polling with this device.
>
> ULE +
Simon Barner wrote:
> Frode Nordahl wrote:
>
>>I was surprised by the actual difference in system
>> performance and
>>usability between 4BSD and ULE under such loads!
>>
>>If you haven't tried it on your heavy trafic server
>>yet, go and do it
>>right now! :-)
>
>
> Just a short "me too". In
Frode Nordahl wrote:
> I was surprised by the actual difference in system performance and
> usability between 4BSD and ULE under such loads!
>
> If you haven't tried it on your heavy trafic server yet, go and do it
> right now! :-)
Just a short "me too". In my case it isn't a heavy traffic serv
Hello,
I have been reading curiously about the ULE scheduler on the lists for
quite some time without ever getting any clear good or bad feeling
about it, so I thought it was about time to give it some real
experience.
I briefly tried it on my desktop about a year a go, and felt that
everythin
On Tue, 8 Feb 2005 16:13:22 +0200 (EET)
"Viktor Ivanov" <[EMAIL PROTECTED]> wrote:
> On Tue, Ôåâðóàðè 8, 2005 15:58, Ion-Mihai Tetcu êàçà:
> >> I've been using only SCHED_ULE on my UP WS, even when there was #error
> >> def. It never broke, not even once :) Though I think there's trouble
> >> with
I compiled my kernel with ULE this morning on my AMD64 workstation to help
test. All seems good so far. Anything in particular to keep an eye on?
-Mike
On Tue, 8 Feb 2005, Michael Nottebrock wrote:
On Tuesday, 8. February 2005 14:02, Mipam wrote:
Okay clear, but the fact that it's in 5-stable sug
On Tue, 8 Feb 2005 08:39:15 -0800
Kris Kennaway <[EMAIL PROTECTED]> wrote:
> On Tue, Feb 08, 2005 at 03:58:22PM +0200, Ion-Mihai Tetcu wrote:
> > On Tue, 8 Feb 2005 14:45:17 +0200 (EET)
> > "Viktor Ivanov" <[EMAIL PROTECTED]> wrote:
> >
> > > On Tue, 8, 2005 14:33, Michael Nottebrock ???
Mipam <[EMAIL PROTECTED]> wrote:
> Thanks for all comments on this topic.
> A good point was made upon the fact that there were always options
> available that weren't stable in X-stable.
> The docs contained appropriate warnings about it you mentioned.
I guess you're referring to my comment.
On Tue, Feb 08, 2005 at 03:58:22PM +0200, Ion-Mihai Tetcu wrote:
> On Tue, 8 Feb 2005 14:45:17 +0200 (EET)
> "Viktor Ivanov" <[EMAIL PROTECTED]> wrote:
>
> > On Tue, 8, 2005 14:33, Michael Nottebrock :
> > > On Tuesday, 8. February 2005 13:07, Mipam wrote:
> > >> I saw several changes
Thanks for all comments on this topic.
A good point was made upon the fact that there were always options
available that weren't stable in X-stable.
The docs contained appropriate warnings about it you mentioned.
Cool, I wish to read the docs on ULE and possible
warnings about using ULE, where ca
On Tue, Февруари 8, 2005 15:58, Ion-Mihai Tetcu каза:
>> I've been using only SCHED_ULE on my UP WS, even when there was #error
>> def. It never broke, not even once :) Though I think there's trouble
>> with SMP and/or HTT. I tried it once on a P4 and it paniced.
>>
>> On the other hand, using SCHE
On Tue, 8 Feb 2005 14:45:17 +0200 (EET)
"Viktor Ivanov" <[EMAIL PROTECTED]> wrote:
> On Tue, Ôåâðóàðè 8, 2005 14:33, Michael Nottebrock êàçà:
> > On Tuesday, 8. February 2005 13:07, Mipam wrote:
> >> I saw several changes to sched_ule.c in the 5 stable branch.
> >> Beneath is one of them:
> >>
> >
Mipam <[EMAIL PROTECTED]> wrote:
> > On Tuesday, 8. February 2005 14:02, Mipam wrote:
> > > Okay clear, but the fact that it's in 5-stable suggests the it's stable
> > > to
> > > use, else why would it be in 5-stable.
> > > Maybe i'm completly wrong in this interpretation?
> [...]
> I thoug
On Tuesday, 8. February 2005 14:22, Ion-Mihai Tetcu wrote:
> On Tue, 8 Feb 2005 14:51:17 +0200 (EET)
>
> "Viktor Ivanov" <[EMAIL PROTECTED]> wrote:
> > On Tue, Ôåâðóàðè 8, 2005 14:43, Ion-Mihai Tetcu êàçà:
> > > Could you tell us again after a week ? There used to be a panic when
> > > using rtprio
On Tue, 8 Feb 2005, Michael Nottebrock wrote:
> On Tuesday, 8. February 2005 14:02, Mipam wrote:
>
> > Okay clear, but the fact that it's in 5-stable suggests the it's stable to
> > use, else why would it be in 5-stable.
>
> The changes that have been merged to stable have been tested for some t
On Tuesday, 8. February 2005 14:02, Mipam wrote:
> Okay clear, but the fact that it's in 5-stable suggests the it's stable to
> use, else why would it be in 5-stable.
The changes that have been merged to stable have been tested for some time in
6-CURRENT, so they're not completely experimental,
On Tue, 8 Feb 2005 14:51:17 +0200 (EET)
"Viktor Ivanov" <[EMAIL PROTECTED]> wrote:
> On Tue, Ôåâðóàðè 8, 2005 14:43, Ion-Mihai Tetcu êàçà:
> > Could you tell us again after a week ? There used to be a panic when
> > using rtprio to raise the priority of a running process, do you know if
> > it's f
On Tue, 8 Feb 2005 14:00:03 +0100
Michael Nottebrock <[EMAIL PROTECTED]> wrote:
> On Tuesday, 8. February 2005 13:43, Ion-Mihai Tetcu wrote:
> > There used to be a panic when
> > using rtprio to raise the priority of a running process, do you know if
> > it's fix ?
>
> I never got any panics with
On Tue, 8 Feb 2005, Michael Nottebrock wrote:
> On Tuesday, 8. February 2005 13:38, Mipam wrote:
>
> > Okay, so then the ULE sched is fairly stable then?
> > But it's still not the default scheduler?
>
> It will never become the default scheduler in 5.x again. 5.x went into
> -STABLE
> mode wi
On Tuesday, 8. February 2005 13:43, Ion-Mihai Tetcu wrote:
> There used to be a panic when
> using rtprio to raise the priority of a running process, do you know if
> it's fix ?
I never got any panics with ULE back when it was available, so I can't tell.
If you care about ULE, turn it on and see
On Tuesday, 8. February 2005 13:38, Mipam wrote:
> Okay, so then the ULE sched is fairly stable then?
> But it's still not the default scheduler?
It will never become the default scheduler in 5.x again. 5.x went into -STABLE
mode with 4BSD, and that's why the default will remain 4BSD.
> Is it s
On Tue, Февруари 8, 2005 14:43, Ion-Mihai Tetcu каза:
> Could you tell us again after a week ? There used to be a panic when
> using rtprio to raise the priority of a running process, do you know if
> it's fix ?
>
I never had those, and I usually run mplayer with rtprio 30... Though
mplayer never
On Tue, Февруари 8, 2005 14:33, Michael Nottebrock каза:
> On Tuesday, 8. February 2005 13:07, Mipam wrote:
>> I saw several changes to sched_ule.c in the 5 stable branch.
>> Beneath is one of them:
>>
>> http://lists.freebsd.org/pipermail/cvs-src/2005-February/039863.html
>>
>> Is the ULE schedule
On Tue, 8 Feb 2005 13:33:04 +0100
Michael Nottebrock <[EMAIL PROTECTED]> wrote:
> On Tuesday, 8. February 2005 13:07, Mipam wrote:
> > Hi,
> >
> > I saw several changes to sched_ule.c in the 5 stable branch.
> > Beneath is one of them:
> >
> > http://lists.freebsd.org/pipermail/cvs-src/2005-Februa
On Tue, 8 Feb 2005, Michael Nottebrock wrote:
> On Tuesday, 8. February 2005 13:07, Mipam wrote:
> > Hi,
> >
> > I saw several changes to sched_ule.c in the 5 stable branch.
> > Beneath is one of them:
> >
> > http://lists.freebsd.org/pipermail/cvs-src/2005-February/039863.html
> >
> > Is the ULE
On Tuesday, 8. February 2005 13:07, Mipam wrote:
> Hi,
>
> I saw several changes to sched_ule.c in the 5 stable branch.
> Beneath is one of them:
>
> http://lists.freebsd.org/pipermail/cvs-src/2005-February/039863.html
>
> Is the ULE scheduler still far from stable in RELENG_5 or not?
You can now
Hi,
I saw several changes to sched_ule.c in the 5 stable branch.
Beneath is one of them:
http://lists.freebsd.org/pipermail/cvs-src/2005-February/039863.html
Is the ULE scheduler still far from stable in RELENG_5 or not?
Bye,
Mipam.
___
freebsd-stabl
38 matches
Mail list logo