[Emc-users] Debian install from live image

2015-10-17 Thread jrmitchellj . 
Since Lucid has reached end of support I decided to back up my config and
install from the 2.7 live CD.
Having done that, installed all of the updates, and adjustments to my
config files,I ran the latency test.
Overall the latency numbers are 30% higher than the Lucid install, but
still in an acceptable range.
Setting "isolcpus=1" in grub has no effect on latency, as it did in the
Lucid install.
Now when I run LinuxCNC, I am getting RTAPI error on startup
intermittently.  When it does start without the error, the error shows up
within 5 to 10 minutes.

Attached are the dmesg, and the -v startup for linuxcnc.

Watching the system monitor while LinuxCNC is running for spikes in the CPU
utilization shows nothing unusual, other than the second cpu is not being
utilized at all.
I have always thought that isolating a cpu from the linux scheduler was to
allow the RTAI kernel to use it without the system interrupts interfering
with it.

Is there some setup that i missed to mask the RTAI kernel to use the
isolated cpu? Or are there any "taskset" statements that need to be changed
to properly utilize the isolated cpu?

I know just enough about real time OSes to be really dangerous!

Any help is always appreciated!

--J. Ray Mitchell Jr.
jrmitche...@gmail.com
(818)324-7573


The dogmas of the quiet past, are inadequate to the stormy present. The
occasion is piled high with difficulty, and we must rise with the occasion.
As our case is new, so we must think anew and act anew. We must disenthrall
ourselves, and then we shall save our country.*Abraham Lincoln
*, *Annual message
to Congress, December 1, 1862*
*16th president of US (1809 - 1865)*
[0.00] Initializing cgroup subsys cpuset
[0.00] Initializing cgroup subsys cpu
[0.00] Linux version 3.4-9-rtai-686-pae (Debian 3.4.55-4linuxcnc) () 
(gcc version 4.6.3 (Debian 4.6.3-14) ) #1 SMP PREEMPT Debian 3.4.55-4linuxcnc
[0.00] BIOS-provided physical RAM map:
[0.00]  BIOS-e820:  - 0008f000 (usable)
[0.00]  BIOS-e820: 0008f000 - 0009 (reserved)
[0.00]  BIOS-e820: 0009 - 0009ec00 (usable)
[0.00]  BIOS-e820: 0009ec00 - 000a (reserved)
[0.00]  BIOS-e820: 000e - 0010 (reserved)
[0.00]  BIOS-e820: 0010 - 7ee98000 (usable)
[0.00]  BIOS-e820: 7ee98000 - 7eebf000 (reserved)
[0.00]  BIOS-e820: 7eebf000 - 7ef48000 (usable)
[0.00]  BIOS-e820: 7ef48000 - 7efbf000 (ACPI NVS)
[0.00]  BIOS-e820: 7efbf000 - 7eff1000 (usable)
[0.00]  BIOS-e820: 7eff1000 - 7efff000 (ACPI data)
[0.00]  BIOS-e820: 7efff000 - 7f00 (usable)
[0.00]  BIOS-e820: 7f00 - 8000 (reserved)
[0.00]  BIOS-e820: f800 - fc00 (reserved)
[0.00]  BIOS-e820: fff0 - 0001 (reserved)
[0.00] Notice: NX (Execute Disable) protection cannot be enabled: 
non-PAE kernel!
[0.00] SMBIOS 2.5 present.
[0.00] DMI:  /D510MO, BIOS 
MOPNV10N.86A.0542.2011.0729.1300 07/29/2011
[0.00] e820 update range:  - 0001 (usable) 
==> (reserved)
[0.00] e820 remove range: 000a - 0010 (usable)
[0.00] last_pfn = 0x7f000 max_arch_pfn = 0x10
[0.00] MTRR default type: uncachable
[0.00] MTRR fixed ranges enabled:
[0.00]   0-9 write-back
[0.00]   A-B uncachable
[0.00]   C-D write-protect
[0.00]   E-F uncachable
[0.00] MTRR variable ranges enabled:
[0.00]   0 base 0 mask F8000 write-back
[0.00]   1 base 07F00 mask FFF00 uncachable
[0.00]   2 base 0 mask F write-protect
[0.00]   3 base 0FFF1 mask F write-protect
[0.00]   4 disabled
[0.00]   5 disabled
[0.00]   6 disabled
[0.00] x86 PAT enabled: cpu 0, old 0x7040600070406, new 0x7010600070106
[0.00] found SMP MP-table at [c00fbe70] fbe70
[0.00] initial memory mapped : 0 - 0180
[0.00] Base memory trampoline at [c009a000] 9a000 size 16384
[0.00] init_memory_mapping: -377fe000
[0.00]  00 - 40 page 4k
[0.00]  40 - 003740 page 2M
[0.00]  003740 - 00377fe000 page 4k
[0.00] kernel direct mapping tables up to 0x377fdfff @ [mem 
0x017f8000-0x017f]
[0.00] RAMDISK: 3652c000 - 3728e000
[0.00] ACPI: RSDP 000f2070 00024 (v02 INTEL )
[0.00] ACPI: XSDT 7effe120 0004C (v01 INTEL  MOPNV10N 021E  
0113)
[0.00] ACPI: FACP 7effd000 000F4 (v03 INTEL  MOPNV10N

Re: [Emc-users] Debian install from live image

2015-10-17 Thread Karlsson & Wang 
> Since Lucid has reached end of support I decided to back up my config and
> install from the 2.7 live CD.
> Having done that, installed all of the updates, and adjustments to my
> config files,I ran the latency test.
> Overall the latency numbers are 30% higher than the Lucid install, but
> still in an acceptable range.
> Setting "isolcpus=1" in grub has no effect on latency, as it did in the
> Lucid install.
> Now when I run LinuxCNC, I am getting RTAPI error on startup
> intermittently.  When it does start without the error, the error shows up
> within 5 to 10 minutes.
> 
> Attached are the dmesg, and the -v startup for linuxcnc.
> 
> Watching the system monitor while LinuxCNC is running for spikes in the CPU
> utilization shows nothing unusual, other than the second cpu is not being
> utilized at all.
> I have always thought that isolating a cpu from the linux scheduler was to
> allow the RTAI kernel to use it without the system interrupts interfering
> with it.
> 
> Is there some setup that i missed to mask the RTAI kernel to use the
> isolated cpu? Or are there any "taskset" statements that need to be changed
> to properly utilize the isolated cpu?
> 
> I know just enough about real time OSes to be really dangerous!
> 
> Any help is always appreciated!
> 
> --J. Ray Mitchell Jr.
> jrmitche...@gmail.com
> (818)324-7573

If there are more than one core one serve the interrupts and the second is used 
for real time scheduling?

--
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Debian install from live image

2015-10-17 Thread jrmitchellj . 
On modern linux kernels, and especially on an RTOS kernel, there are ways
to assign tasks to specific cores of a multicore system.
Affinity is one way that I have used on a project that had data coming in
from DAQ cards at a rapid rate, and I don't want to loose any samples.  So
I assigned the data collection to one core, then the processing & packaging
is done on the other core.

Most of the real time stuff that is currently in kernels is from the
RT-Preempt RTOS project.  A news article I saw recently states that more of
it will be brought into the mainline kernel in the near future.

It really seems weird to me  that if you isolate a core from the standard
kernel scheduler, that you are not scheduling task on it with the RTAI
kernel patch.

Ray M.

--J. Ray Mitchell Jr.
jrmitche...@gmail.com
(818)324-7573


The dogmas of the quiet past, are inadequate to the stormy present. The
occasion is piled high with difficulty, and we must rise with the occasion.
As our case is new, so we must think anew and act anew. We must disenthrall
ourselves, and then we shall save our country.*Abraham Lincoln
*, *Annual message
to Congress, December 1, 1862*
*16th president of US (1809 - 1865)*

On Sat, Oct 17, 2015 at 1:31 PM, Karlsson & Wang <
nicklas.karls...@karlssonwang.se> wrote:

> > Since Lucid has reached end of support I decided to back up my config and
> > install from the 2.7 live CD.
> > Having done that, installed all of the updates, and adjustments to my
> > config files,I ran the latency test.
> > Overall the latency numbers are 30% higher than the Lucid install, but
> > still in an acceptable range.
> > Setting "isolcpus=1" in grub has no effect on latency, as it did in the
> > Lucid install.
> > Now when I run LinuxCNC, I am getting RTAPI error on startup
> > intermittently.  When it does start without the error, the error shows up
> > within 5 to 10 minutes.
> >
> > Attached are the dmesg, and the -v startup for linuxcnc.
> >
> > Watching the system monitor while LinuxCNC is running for spikes in the
> CPU
> > utilization shows nothing unusual, other than the second cpu is not being
> > utilized at all.
> > I have always thought that isolating a cpu from the linux scheduler was
> to
> > allow the RTAI kernel to use it without the system interrupts interfering
> > with it.
> >
> > Is there some setup that i missed to mask the RTAI kernel to use the
> > isolated cpu? Or are there any "taskset" statements that need to be
> changed
> > to properly utilize the isolated cpu?
> >
> > I know just enough about real time OSes to be really dangerous!
> >
> > Any help is always appreciated!
> >
> > --J. Ray Mitchell Jr.
> > jrmitche...@gmail.com
> > (818)324-7573
>
> If there are more than one core one serve the interrupts and the second is
> used for real time scheduling?
>
>
> --
> ___
> Emc-users mailing list
> Emc-users@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/emc-users
>
--
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Debian install from live image

2015-10-17 Thread jrmitchellj . 
By the way, I just checked from a backup of my Lucid configuration, and it
is not using the second core for anything as well.  It just has better
latency to begin with.

Ray

--J. Ray Mitchell Jr.
jrmitche...@gmail.com
(818)324-7573


The dogmas of the quiet past, are inadequate to the stormy present. The
occasion is piled high with difficulty, and we must rise with the occasion.
As our case is new, so we must think anew and act anew. We must disenthrall
ourselves, and then we shall save our country.*Abraham Lincoln
*, *Annual message
to Congress, December 1, 1862*
*16th president of US (1809 - 1865)*

On Sat, Oct 17, 2015 at 2:10 PM, jrmitchellj . 
wrote:

> On modern linux kernels, and especially on an RTOS kernel, there are ways
> to assign tasks to specific cores of a multicore system.
> Affinity is one way that I have used on a project that had data coming in
> from DAQ cards at a rapid rate, and I don't want to loose any samples.  So
> I assigned the data collection to one core, then the processing & packaging
> is done on the other core.
>
> Most of the real time stuff that is currently in kernels is from the
> RT-Preempt RTOS project.  A news article I saw recently states that more of
> it will be brought into the mainline kernel in the near future.
>
> It really seems weird to me  that if you isolate a core from the standard
> kernel scheduler, that you are not scheduling task on it with the RTAI
> kernel patch.
>
> Ray M.
>
> --J. Ray Mitchell Jr.
> jrmitche...@gmail.com
> (818)324-7573
>
>
> The dogmas of the quiet past, are inadequate to the stormy present. The
> occasion is piled high with difficulty, and we must rise with the occasion.
> As our case is new, so we must think anew and act anew. We must disenthrall
> ourselves, and then we shall save our country.*Abraham Lincoln
> *, *Annual message
> to Congress, December 1, 1862*
> *16th president of US (1809 - 1865)*
>
> On Sat, Oct 17, 2015 at 1:31 PM, Karlsson & Wang <
> nicklas.karls...@karlssonwang.se> wrote:
>
>> > Since Lucid has reached end of support I decided to back up my config
>> and
>> > install from the 2.7 live CD.
>> > Having done that, installed all of the updates, and adjustments to my
>> > config files,I ran the latency test.
>> > Overall the latency numbers are 30% higher than the Lucid install, but
>> > still in an acceptable range.
>> > Setting "isolcpus=1" in grub has no effect on latency, as it did in the
>> > Lucid install.
>> > Now when I run LinuxCNC, I am getting RTAPI error on startup
>> > intermittently.  When it does start without the error, the error shows
>> up
>> > within 5 to 10 minutes.
>> >
>> > Attached are the dmesg, and the -v startup for linuxcnc.
>> >
>> > Watching the system monitor while LinuxCNC is running for spikes in the
>> CPU
>> > utilization shows nothing unusual, other than the second cpu is not
>> being
>> > utilized at all.
>> > I have always thought that isolating a cpu from the linux scheduler was
>> to
>> > allow the RTAI kernel to use it without the system interrupts
>> interfering
>> > with it.
>> >
>> > Is there some setup that i missed to mask the RTAI kernel to use the
>> > isolated cpu? Or are there any "taskset" statements that need to be
>> changed
>> > to properly utilize the isolated cpu?
>> >
>> > I know just enough about real time OSes to be really dangerous!
>> >
>> > Any help is always appreciated!
>> >
>> > --J. Ray Mitchell Jr.
>> > jrmitche...@gmail.com
>> > (818)324-7573
>>
>> If there are more than one core one serve the interrupts and the second
>> is used for real time scheduling?
>>
>>
>> --
>> ___
>> Emc-users mailing list
>> Emc-users@lists.sourceforge.net
>> https://lists.sourceforge.net/lists/listinfo/emc-users
>>
>
>
--
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users