* Ingo Molnar <[EMAIL PROTECTED]> wrote:
> thanks. It shows the anomaly in action:
> So all the time references we have show that (no surprise here) 1
> second passed between the two samples. But sched_clock() shows a
> _large_ jump:
>
> .clock : 125652924079659272
> .cloc
* Matt Mackall <[EMAIL PROTECTED]> wrote:
> On Tue, Jun 05, 2007 at 09:50:15PM +0200, Ingo Molnar wrote:
> > Matt, could you run this for 1-2 minutes and send us the sched_debug.txt
> > output?
>
> http://selenic.com/sched_debug.txt.gz
thanks. It shows the anomaly in action:
now at 36529421
On Tue, Jun 05, 2007 at 09:50:15PM +0200, Ingo Molnar wrote:
> Matt, could you run this for 1-2 minutes and send us the sched_debug.txt
> output?
http://selenic.com/sched_debug.txt.gz
--
Mathematics is the supreme nostalgia of our time.
-
To unsubscribe from this list: send the line "unsubscrib
On Tue, Jun 05, 2007 at 09:08:06PM +0200, Ingo Molnar wrote:
>
> * Ingo Molnar <[EMAIL PROTECTED]> wrote:
>
> > Matt, could you send me your .config too? Maybe i can reproduce it
> > with your config.
>
> ok, got it (off-list) and you seem to have the default HZ=250. Could you
> change it to H
* Dmitry Adamushko <[EMAIL PROTECTED]> wrote:
> On 05/06/07, Dmitry Adamushko <[EMAIL PROTECTED]> wrote:
> >> now at 257428593818894 nsecs
> >>
> >> cpu: 0
> >> .nr_running: 3
> >> .raw_weighted_load : 2063
> >> .nr_switches : 242830075
> >> .nr_load_updates
* Matt Mackall <[EMAIL PROTECTED]> wrote:
> On Tue, Jun 05, 2007 at 02:37:38PM +1000, Rusty Russell wrote:
> > On Mon, 2007-06-04 at 23:18 -0500, Matt Mackall wrote:
> > > On Tue, Jun 05, 2007 at 12:31:15PM +1000, Rusty Russell wrote:
> > > > Does this solve it for you?
> > >
> > > Nope. Doesn't
* Ingo Molnar <[EMAIL PROTECTED]> wrote:
> Matt, could you send me your .config too? Maybe i can reproduce it
> with your config.
ok, got it (off-list) and you seem to have the default HZ=250. Could you
change it to HZ=1000 - does the delay get alot smaller?
if yes then this could be roughly
On 05/06/07, Dmitry Adamushko <[EMAIL PROTECTED]> wrote:
> now at 257428593818894 nsecs
>
> cpu: 0
> .nr_running: 3
> .raw_weighted_load : 2063
> .nr_switches : 242830075
> .nr_load_updates : 30172063
> .nr_uninterruptible: 0
> .jiffies
On 05/06/07, Matt Mackall <[EMAIL PROTECTED]> wrote:
> [...]
> Click into the lguest window and trigger the delay.
I did:
while true; do sleep 1; cat /proc/sched_debug > sched_debug.txt; done
and got this, hopefully inside the window:
Sched Debug Version: v0.02
now at 257428593818894 nsecs
c
On Tue, Jun 05, 2007 at 11:23:44AM +0200, Ingo Molnar wrote:
>
> * Matt Mackall <[EMAIL PROTECTED]> wrote:
>
> > Ok, my transcoding just finished as I was writing this. So I've
> > reproduced the problem with this Python script that I had handy:
> >
> > memload.py:
> > #!/usr/bin/python
> > a =
On Tue, Jun 05, 2007 at 09:19:04AM +0200, Ingo Molnar wrote:
>
> * Matt Mackall <[EMAIL PROTECTED]> wrote:
>
> > sleep_max: 57476665627
> > block_max: 18014060106626075
>
> hm, this block_max looks a bit suspect, it's 003fffb1359e341b. Does your
> box m
On Tue, Jun 05, 2007 at 02:37:38PM +1000, Rusty Russell wrote:
> On Mon, 2007-06-04 at 23:18 -0500, Matt Mackall wrote:
> > On Tue, Jun 05, 2007 at 12:31:15PM +1000, Rusty Russell wrote:
> > > Does this solve it for you?
> >
> > Nope. Doesn't accept input and hogs the CPU with lots of system time.
Matt, could you send me your .config too? Maybe i can reproduce it
with your config.
Ingo
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to [EMAIL PROTECTED]
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read
* Matt Mackall <[EMAIL PROTECTED]> wrote:
> Ok, my transcoding just finished as I was writing this. So I've
> reproduced the problem with this Python script that I had handy:
>
> memload.py:
> #!/usr/bin/python
> a = "a" * 16 * 1024 * 1024
> while 1:
> b = a[1:] + "b"
> a = b[1:] + "c"
* Matt Mackall <[EMAIL PROTECTED]> wrote:
> sleep_max: 57476665627
> block_max: 18014060106626075
hm, this block_max looks a bit suspect, it's 003fffb1359e341b. Does your
box make any use of cpufreq? (what CPU is it?)
> strace -f doesn't kill the laten
* Rusty Russell <[EMAIL PROTECTED]> wrote:
> This sounds like the waker process (nice 19) not getting a chance to
> run. You can hack around it for the moment by changing "nice(19)" in
> Documentation/lguest/lguest.c to something less aggressive.
even if the waker runs at nice+19, under CFS (
On Mon, 2007-06-04 at 23:18 -0500, Matt Mackall wrote:
> On Tue, Jun 05, 2007 at 12:31:15PM +1000, Rusty Russell wrote:
> > Does this solve it for you?
>
> Nope. Doesn't accept input and hogs the CPU with lots of system time.
Dumb question: did you replace both the module and the launcher?
Rusty
On Tue, Jun 05, 2007 at 12:31:15PM +1000, Rusty Russell wrote:
> Does this solve it for you?
Nope. Doesn't accept input and hogs the CPU with lots of system time.
--
Mathematics is the supreme nostalgia of our time.
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the
On Tue, Jun 05, 2007 at 10:18:54AM +1000, Rusty Russell wrote:
> On Mon, 2007-06-04 at 12:37 -0500, Matt Mackall wrote:
> > With 2.6.22-rc3-mm1, I've got a long-running video transcoding going
> > on. In other windows, I'm compiling, reading email, etc. with no
> > noticeable problems.
> >
> > If
On Tue, 2007-06-05 at 10:19 +1000, Rusty Russell wrote:
> On Mon, 2007-06-04 at 12:37 -0500, Matt Mackall wrote:
> > With 2.6.22-rc3-mm1, I've got a long-running video transcoding going
> > on. In other windows, I'm compiling, reading email, etc. with no
> > noticeable problems.
> >
> > If I fire
On Tue, Jun 05, 2007 at 11:02:00AM +1000, Rusty Russell wrote:
> On Mon, 2007-06-04 at 17:58 -0500, Olof Johansson wrote:
> >
> > Is this on hvc console? The hvc driver was recently modified to back off
> > it's timing interval when there's no input, the max value is 2 seconds.
> >
> > See b79107
On Tue, 05 Jun 2007 11:02:00 +1000 Rusty Russell <[EMAIL PROTECTED]> wrote:
>
> > See b791072ba3c3b29bab1890963bde52eb944a8451 for the changes in question.
>
> Hmm, I thought that was only supposed to happen for "interruptless"
> consoles?
The comment certainly says it only applies to polled mode
On Mon, 2007-06-04 at 17:58 -0500, Olof Johansson wrote:
> On Mon, Jun 04, 2007 at 12:37:10PM -0500, Matt Mackall wrote:
> > With 2.6.22-rc3-mm1, I've got a long-running video transcoding going
> > on. In other windows, I'm compiling, reading email, etc. with no
> > noticeable problems.
> >
> > If
On Mon, 2007-06-04 at 12:37 -0500, Matt Mackall wrote:
> With 2.6.22-rc3-mm1, I've got a long-running video transcoding going
> on. In other windows, I'm compiling, reading email, etc. with no
> noticeable problems.
>
> If I fire up lguest and leave it sitting at a shell prompt for a
> couple mome
On Mon, Jun 04, 2007 at 12:37:10PM -0500, Matt Mackall wrote:
> With 2.6.22-rc3-mm1, I've got a long-running video transcoding going
> on. In other windows, I'm compiling, reading email, etc. with no
> noticeable problems.
>
> If I fire up lguest and leave it sitting at a shell prompt for a
> coup
* James Morris <[EMAIL PROTECTED]> wrote:
> On Mon, 4 Jun 2007, Ingo Molnar wrote:
>
> > perhaps lguest does something unusual when it idles? Does it fiddle
> > sched_clock() in any way?
>
> One of the lguest processes calls nice(2), which seems like a long
> shot, but also not impossibly rel
* Ingo Molnar <[EMAIL PROTECTED]> wrote:
> > I suggest you try this yourself [...]
>
> yep, i know how to use lguest and i'm already running it :) But i
> cannot trigger the problem, that's why i was curious about your
> host-side stats.
i can run a hundred CPU hogs on the host and the lguest
On Mon, 4 Jun 2007, Ingo Molnar wrote:
> perhaps lguest does something unusual when it idles? Does it fiddle
> sched_clock() in any way?
One of the lguest processes calls nice(2), which seems like a long shot,
but also not impossibly related. (Note that newer patches in the lguest
queue remov
* Matt Mackall <[EMAIL PROTECTED]> wrote:
> So I sent you everything but the stats for the CPU hog.
ok, thx.
> I suggest you try this yourself [...]
yep, i know how to use lguest and i'm already running it :) But i cannot
trigger the problem, that's why i was curious about your host-side
sta
On Mon, Jun 04, 2007 at 09:40:38PM +0200, Ingo Molnar wrote:
>
> * Matt Mackall <[EMAIL PROTECTED]> wrote:
>
> > > ah, so both the shell and the 'competing' CPU hog was running within
> > > the same lguest instance?
> >
> > No.
> >
> > The CPU hog is running in the host. A single instance of l
* Matt Mackall <[EMAIL PROTECTED]> wrote:
> > ah, so both the shell and the 'competing' CPU hog was running within
> > the same lguest instance?
>
> No.
>
> The CPU hog is running in the host. A single instance of lguest using
> busybox is started in another shell. When it reaches a shell pro
On Mon, Jun 04, 2007 at 09:12:29PM +0200, Ingo Molnar wrote:
>
> * Matt Mackall <[EMAIL PROTECTED]> wrote:
>
> > > btw., does this only happen with lguest, or with other idle shells
> > > too?
> >
> > Only noticed with lguest.
>
> ah, so both the shell and the 'competing' CPU hog was running w
* Matt Mackall <[EMAIL PROTECTED]> wrote:
> > btw., does this only happen with lguest, or with other idle shells
> > too?
>
> Only noticed with lguest.
ah, so both the shell and the 'competing' CPU hog was running within the
same lguest instance?
Ingo
-
To unsubscribe from this list:
On Mon, Jun 04, 2007 at 08:11:51PM +0200, Ingo Molnar wrote:
>
> * Matt Mackall <[EMAIL PROTECTED]> wrote:
>
> > With 2.6.22-rc3-mm1, I've got a long-running video transcoding going
> > on. In other windows, I'm compiling, reading email, etc. with no
> > noticeable problems.
> >
> > If I fire
On Mon, Jun 04, 2007 at 07:54:36PM +0200, Ingo Molnar wrote:
>
> * Matt Mackall <[EMAIL PROTECTED]> wrote:
>
> > With 2.6.22-rc3-mm1, I've got a long-running video transcoding going
> > on. In other windows, I'm compiling, reading email, etc. with no
> > noticeable problems.
> >
> > If I fire
* Matt Mackall <[EMAIL PROTECTED]> wrote:
> With 2.6.22-rc3-mm1, I've got a long-running video transcoding going
> on. In other windows, I'm compiling, reading email, etc. with no
> noticeable problems.
>
> If I fire up lguest and leave it sitting at a shell prompt for a
> couple moments, whe
* Matt Mackall <[EMAIL PROTECTED]> wrote:
> With 2.6.22-rc3-mm1, I've got a long-running video transcoding going
> on. In other windows, I'm compiling, reading email, etc. with no
> noticeable problems.
>
> If I fire up lguest and leave it sitting at a shell prompt for a
> couple moments, whe
With 2.6.22-rc3-mm1, I've got a long-running video transcoding going
on. In other windows, I'm compiling, reading email, etc. with no
noticeable problems.
If I fire up lguest and leave it sitting at a shell prompt for a
couple moments, when I return to type something at the prompt, it can
take 2-3
38 matches
Mail list logo