Raq2 & 2.6.4 : Strange output fro msomewhere

Raq2 & 2.6.4 : Strange output fro msomewhere

[Massimo Cetra]

Hi!

(i'm new here).
I have managed to compile CVS kernels for 2.6.4 and 2.4.25 trees on my
Raq2... And they work great.

Now...
Under heavy load (generating kernel_headers debian package), i saw the
following:

-cobalt:/proc# uptime
Unknown HZ value! (79) Assume 100.
 06:54:27 up 14 min,  2 users,  load average: 1.77, 1.37, 0.69
cobalt:/proc# sar 2 2
Linux 2.6.4 (cobalt)    04/12/04

06:54:32          CPU     %user     %nice   %system     %idle
06:54:34          all     18.31      0.00     81.69      0.00
06:54:36          all     16.67      0.00     83.33      0.00
Average:          all     17.60      0.00     82.40      0.00
cobalt:/proc# uname -a
Linux cobalt 2.6.4 #2 Mon Apr 12 05:50:49 CEST 2004 mips unknown
cobalt:/proc#


What is : "Unknown HZ value! (79) Assume 100." ????????????????????????

Thanks..

 Massimo

Raq2 & 2.6.4 : Strange output fro msomewhere

[Massimo Cetra]

Hi!

(i'm new here).
I have managed to compile CVS kernels for 2.6.4 and 2.4.25 trees on my
Raq2... And they work great.

Now...
Under heavy load (generating kernel_headers debian package), i saw the
following:

-cobalt:/proc# uptime
Unknown HZ value! (79) Assume 100.
 06:54:27 up 14 min,  2 users,  load average: 1.77, 1.37, 0.69
cobalt:/proc# sar 2 2
Linux 2.6.4 (cobalt)    04/12/04

06:54:32          CPU     %user     %nice   %system     %idle
06:54:34          all     18.31      0.00     81.69      0.00
06:54:36          all     16.67      0.00     83.33      0.00
Average:          all     17.60      0.00     82.40      0.00
cobalt:/proc# uname -a
Linux cobalt 2.6.4 #2 Mon Apr 12 05:50:49 CEST 2004 mips unknown
cobalt:/proc#


What is : "Unknown HZ value! (79) Assume 100." ????????????????????????

Thanks..

 Massimo

Re: Raq2 & 2.6.4 : Strange output fro msomewhere

[Kevin D. Kissell]

I don't have the sources to "uptime" handy, but the concept of a "HZ" value
us something that goes back to AT&T UNIX, representing the number of
clock interrupts per second.  It was originally 60, as it was based on the 
wall current AC frequency in the US, but once 100 became a popular
value once people started having faster CPUs and stopped using their
power supplies to generate the signal.

Historically, it was a constant specified in some kernel/system header
file like "sys/param.h".

So I would tend to conclude that the uptime program does some kind 
of switch-driven computation based on the value of HZ it picks up (at
build time?), that the value for the Raq2 is 79, and that 79 isn't one
of the values recognized by the code.

----- Original Message ----- 
From: "Massimo Cetra" <mcetra@navynet.it>
To: <linux-mips@linux-mips.org>
Sent: Monday, April 12, 2004 7:58
Subject: Raq2 & 2.6.4 : Strange output fro msomewhere


> Hi!
> 
> (i'm new here).
> I have managed to compile CVS kernels for 2.6.4 and 2.4.25 trees on my
> Raq2... And they work great.
> 
> Now...
> Under heavy load (generating kernel_headers debian package), i saw the
> following:
> 
> -cobalt:/proc# uptime
> Unknown HZ value! (79) Assume 100.
>  06:54:27 up 14 min,  2 users,  load average: 1.77, 1.37, 0.69
> cobalt:/proc# sar 2 2
> Linux 2.6.4 (cobalt)    04/12/04
> 
> 06:54:32          CPU     %user     %nice   %system     %idle
> 06:54:34          all     18.31      0.00     81.69      0.00
> 06:54:36          all     16.67      0.00     83.33      0.00
> Average:          all     17.60      0.00     82.40      0.00
> cobalt:/proc# uname -a
> Linux cobalt 2.6.4 #2 Mon Apr 12 05:50:49 CEST 2004 mips unknown
> cobalt:/proc#
> 
> 
> What is : "Unknown HZ value! (79) Assume 100." ????????????????????????
> 
> Thanks..
> 
>  Massimo
> 
> 
> 
> 

Re: Raq2 & 2.6.4 : Strange output fro msomewhere

[Kevin D. Kissell]

I don't have the sources to "uptime" handy, but the concept of a "HZ" value
us something that goes back to AT&T UNIX, representing the number of
clock interrupts per second.  It was originally 60, as it was based on the 
wall current AC frequency in the US, but once 100 became a popular
value once people started having faster CPUs and stopped using their
power supplies to generate the signal.

Historically, it was a constant specified in some kernel/system header
file like "sys/param.h".

So I would tend to conclude that the uptime program does some kind 
of switch-driven computation based on the value of HZ it picks up (at
build time?), that the value for the Raq2 is 79, and that 79 isn't one
of the values recognized by the code.

----- Original Message ----- 
From: "Massimo Cetra" <mcetra@navynet.it>
To: <linux-mips@linux-mips.org>
Sent: Monday, April 12, 2004 7:58
Subject: Raq2 & 2.6.4 : Strange output fro msomewhere


> Hi!
> 
> (i'm new here).
> I have managed to compile CVS kernels for 2.6.4 and 2.4.25 trees on my
> Raq2... And they work great.
> 
> Now...
> Under heavy load (generating kernel_headers debian package), i saw the
> following:
> 
> -cobalt:/proc# uptime
> Unknown HZ value! (79) Assume 100.
>  06:54:27 up 14 min,  2 users,  load average: 1.77, 1.37, 0.69
> cobalt:/proc# sar 2 2
> Linux 2.6.4 (cobalt)    04/12/04
> 
> 06:54:32          CPU     %user     %nice   %system     %idle
> 06:54:34          all     18.31      0.00     81.69      0.00
> 06:54:36          all     16.67      0.00     83.33      0.00
> Average:          all     17.60      0.00     82.40      0.00
> cobalt:/proc# uname -a
> Linux cobalt 2.6.4 #2 Mon Apr 12 05:50:49 CEST 2004 mips unknown
> cobalt:/proc#
> 
> 
> What is : "Unknown HZ value! (79) Assume 100." ????????????????????????
> 
> Thanks..
> 
>  Massimo
> 
> 
> 
> 

Re: Raq2 & 2.6.4 : Strange output fro msomewhere

[Kumba]

Massimo Cetra wrote:

> Hi!
> 
> (i'm new here).
> I have managed to compile CVS kernels for 2.6.4 and 2.4.25 trees on my
> Raq2... And they work great.
> 
> Now...
> Under heavy load (generating kernel_headers debian package), i saw the
> following:
> 
> -cobalt:/proc# uptime
> Unknown HZ value! (79) Assume 100.
>  06:54:27 up 14 min,  2 users,  load average: 1.77, 1.37, 0.69
> cobalt:/proc# sar 2 2
> Linux 2.6.4 (cobalt)    04/12/04
> 
> 06:54:32          CPU     %user     %nice   %system     %idle
> 06:54:34          all     18.31      0.00     81.69      0.00
> 06:54:36          all     16.67      0.00     83.33      0.00
> Average:          all     17.60      0.00     82.40      0.00
> cobalt:/proc# uname -a
> Linux cobalt 2.6.4 #2 Mon Apr 12 05:50:49 CEST 2004 mips unknown
> cobalt:/proc#
> 
> 
> What is : "Unknown HZ value! (79) Assume 100." ????????????????????????
> 
> Thanks..

What does dmesg say your bogomips are?  There's a patch that fixes a 
small calculation error.  Without it, bogomips reports in at ~2,500, 
instead of 250.  Call it a guess, but if you missed that patch, that 
might be your bug.  Otherwise, I'm not too sure.


--Kumba

-- 
"Such is oft the course of deeds that move the wheels of the world: 
small hands do them because they must, while the eyes of the great are 
elsewhere."  --Elrond

Re: Raq2 & 2.6.4 : Strange output fro msomewhere

[Florian Lohoff]

[-- Attachment #1: Type: text/plain, Size: 448 bytes --]

On Mon, Apr 12, 2004 at 06:58:20AM +0100, Massimo Cetra wrote:
> -cobalt:/proc# uptime
> Unknown HZ value! (79) Assume 100.
>  06:54:27 up 14 min,  2 users,  load average: 1.77, 1.37, 0.69

IIRC this is a hint to upgrade your procps :)

2.6 from the Documentation/Changes requires at least 3.2.0

Flo
-- 
Florian Lohoff                  flo@rfc822.org             +49-171-2280134
                        Heisenberg may have been here.

[-- Attachment #2: Type: application/pgp-signature, Size: 189 bytes --]