[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: forcedeth?



On Thursday 02 May 2019 14:18:53 Pascal Hambourg wrote:

> Le 02/05/2019 à 13:25, Gene Heskett a écrit :
> > Ha anything been done to forcedeth since wheezy?
>
> Why are you asking ?
>
> > I have installed the LCNC version of stretch
>
> What is LCNC and how does if differ from vanilla stretch ?

Probably the biggest diff is the substitution of a fully preempt-rt 
patched kernel, real time enough to run heavy machinery which is what 
LinuxCNC, AKA LCNC linux does.  Not fully RTAI patched, but that guy got 
tired of the fights trying to get his patches into mainstream and walked 
away several years ago. With his patches, timeing of irq response was 
way less than a 10 microsecond timing wibble so it was possible to drive 
stepper motors thru a power amp directly from the parport.  With a 
preempt-rt kernel, the wibbles are in the 30 to 50 microsecond range and 
several venders have cards the offload the cpu's job to their fpga based 
card which can tolerate that well enough that if one of those cards is 
used instead of the parport, a machine that could move 12 to 15 inches a 
minute, can be moved at 200 inches a minute.  But the card costs around 
$80-$260 depending on what it can do, raising the bar for converting an 
old machine to CNC by that much more.
>
> > The network was easy to make work after the install (the installer
> > doesn't keep the settings its given, and which work during the
> > install, but there is no way to get the networking going after the
> > installs reboot on this machine.
>
> The installer uses the same ethernet drivers as the installed system.
> If the NIC driver worked during the installation, it should work
> equally in the installed system and your problem is a configuration
> issue.

Which is exactly why my original post quoted all those config things.

> > To get the Dell to work, all I had to do was
> > edit /etc/network/interfaces.d/setup
>
> (...)
>
> > And restart the networking.
>
> Do not do this. It does not work. ifupdown does not keep state of the
> original configuration, so when you stop the interface using the new
> configuration which does not match th old one, you will get errors.
> Stop the interface(s), edit the config files(s) and start the
> interface(s) instead.

I'll do that, but what diff does it make when its never worked except 
during the install?

> > And theres no way to even trace its failure because strace is not
> > installed. The error message is very explicit, RTNETWORK, whatever
> > the heck that is, says "file exists", followed by "failed."
>
> Configuration issue. You don't need strace.
> 1) Check the contents of all /etc/network/interfaces{,d/*}

Just one file there on both machines, /etc/network/interfaces.d/setup and 
except for the machines ipv4 address's, are identical.  There is no ipv6 
stuff outside of my local network for at least 100 miles, this is West 
by God Virginia. ANY flat land is made by a bulldozer with a nearby 
tanker of #2 diesel.

> 2) Check the current interface status with ip addr
> 2) Run ifdown -v --force eth0 ; ifup -v eth0
>
> > Do I file a bug against forcedeth?
>
> What seriously makes you thing forcedeth is involved ?

That, and the hardware its running, are the remaining major diffs in the 
networking setup between the 2 machines and the two installs.

This machine is running wheezy perfectly and is using a forcedeth that is 
from an lsmod # forcedeth   65295  0 <-- no links?  WTH is it running?
                            ^^^^^
from this choice of kernels:
gene@coyote:~$ ls -l `locate forcedeth.ko`

This is the running kernel
2017 /lib/modules/3.16.0-0.bpo.4-amd64/kernel/drivers/net/ethernet/nvidia/forcedeth.ko
-rw-r--r-- 1 root root  75384 Dec 10                                                    

None of which match the one running, WTH?

Your turn, now I am really confused....

Cheers, Gene Heskett
-- 
"There are four boxes to be used in defense of liberty:
 soap, ballot, jury, and ammo. Please use in that order."
-Ed Howdershelt (Author)
Genes Web page <http://geneslinuxbox.net:6309/gene>


Reply to: