Re: [Emc-developers] To-Do list

[Chris Morley]

I think we should add a couple more people for push access too John m and 
Philip c come to mind...if they are interested


 Original message 
From: andy pugh 
Date: 2019-04-28 10:07 a.m. (GMT-07:00)
To: EMC developers 
Subject: Re: [Emc-developers] To-Do list

On Sun, 28 Apr 2019 at 14:07, Peter C. Wallace  wrote:

>
> There are a number of Rob Ellenbergs bugfixes for the TP
> that I dont think have been merged
> (someone on the forum just ran into one of the fixed bugs)

I am not sure why he doesn't have direct push access. It isn't like
anyone else understands the new TP better than he does.

--
atp
"A motorcycle is a bicycle with a pandemonium attachment and is
designed for the especial use of mechanical geniuses, daredevils and
lunatics."
— George Fitch, Atlanta Constitution Newspaper, 1916


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

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

Re: [Emc-developers] iocontrol.cc never runs EMC_TOOL_INIT message

[Chris Morley]

Thanks for looking Andy..im working long hours so have got far.

I have done some code work on this and can get tool init to be called but not 
io_init and tool_ init  I still dont quite understand how messages are queued.

Iocontrol v2 handles error conditions on the hardware side.allows linuxcnc to 
be aware of them.
Its work from Michael H so was not integrated completely.

We can make it standard by changing all sample configs. I would wait till after 
2.8 is made.

My 2 cents.


 Original message 
From: andy pugh 
Date: 2019-04-28 10:46 a.m. (GMT-07:00)
To: EMC developers 
Subject: Re: [Emc-developers] iocontrol.cc never runs EMC_TOOL_INIT message

On Thu, 18 Apr 2019 at 04:34, Chris Morley  wrote:
>
> While I was toying with a tool startup problem from the forum, I discovered 
> that iocontrol.cc never runs
> the EMC_TOOL_INIT NML message even though taskclass.cc sends it.

This has been sat in the bottom of my inbox waiting for me to got
round to looking at it.
(And I still haven't, in any detail)

> So when iocontrol is first run the serial number is 0 and the echo is 
> immediately advanced to 1 when updating the HAL pins.
> When the EMC_TOOL_INIT message comes it has a serial number of 1 and echo 
> already has a number of 1.
> So the EMC_TOOL_INIT is never run but the serial/echo number synchs and 
> everything works from there.

I see the same code in iocontrolV2. (What is "better" about
iocontrolv2 and why isn't it the default?)

--
atp
"A motorcycle is a bicycle with a pandemonium attachment and is
designed for the especial use of mechanical geniuses, daredevils and
lunatics."
— George Fitch, Atlanta Constitution Newspaper, 1916


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

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

Re: [Emc-developers] iocontrol.cc never runs EMC_TOOL_INIT message

[andy pugh]

On Thu, 18 Apr 2019 at 04:34, Chris Morley  wrote:
>
> While I was toying with a tool startup problem from the forum, I discovered 
> that iocontrol.cc never runs
> the EMC_TOOL_INIT NML message even though taskclass.cc sends it.

This has been sat in the bottom of my inbox waiting for me to got
round to looking at it.
(And I still haven't, in any detail)

> So when iocontrol is first run the serial number is 0 and the echo is 
> immediately advanced to 1 when updating the HAL pins.
> When the EMC_TOOL_INIT message comes it has a serial number of 1 and echo 
> already has a number of 1.
> So the EMC_TOOL_INIT is never run but the serial/echo number synchs and 
> everything works from there.

I see the same code in iocontrolV2. (What is "better" about
iocontrolv2 and why isn't it the default?)

-- 
atp
"A motorcycle is a bicycle with a pandemonium attachment and is
designed for the especial use of mechanical geniuses, daredevils and
lunatics."
— George Fitch, Atlanta Constitution Newspaper, 1916


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

Re: [Emc-developers] To-Do list

[andy pugh]

On Sun, 28 Apr 2019 at 14:07, Peter C. Wallace  wrote:

>
> There are a number of Rob Ellenbergs bugfixes for the TP
> that I dont think have been merged
> (someone on the forum just ran into one of the fixed bugs)

I am not sure why he doesn't have direct push access. It isn't like
anyone else understands the new TP better than he does.

-- 
atp
"A motorcycle is a bicycle with a pandemonium attachment and is
designed for the especial use of mechanical geniuses, daredevils and
lunatics."
— George Fitch, Atlanta Constitution Newspaper, 1916


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

Re: [Emc-developers] To-Do list

[Peter C. Wallace]

On Sun, 28 Apr 2019, andy pugh wrote:


Date: Sun, 28 Apr 2019 00:12:49 +0100
From: andy pugh 
Reply-To: EMC developers 
To: EMC developers 
Subject: [Emc-developers] To-Do list

Which bugs on the tracker mean that 2.8 can't be released?



I think that the unwanted motion caused by ON_ABORT_COMMAND needs to
be fixed before release.
https://github.com/LinuxCNC/linuxcnc/issues/579

Any others? We have 21 pull requests and 105 bugs..

What else needs to be done?

Update ISO downloads to Stretch. (easy for preempt-rt, already done,
harder or RTAI)



There are a number of Rob Ellenbergs bugfixes for the TP
that I dont think have been merged
(someone on the forum just ran into one of the fixed bugs)

Peter Wallace
Mesa Electronics



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

Re: [Emc-developers] buildbot for arm64

[Gene Heskett]

On Saturday 27 April 2019 23:49:39 Alan Condit wrote:

> Gene,
>
> I built the kernel on the uSD.
>
> I have a 32gb SSD that doesn’t seem to want to work.

Likely used up, some have a pretty poor write life.

> So I ordered a 
> 240gb SSD and a SATA to USB3.0 case from Walmart. I started to order a
> 120gb SSD but the shipping was almost as much as the difference in
> price between the smaller one and the larger one and I got free
> shipping with the larger one and case.
>
> I have an 80gb 2.5” HD (usb2.0) that I tried flashing like a uSD. It
> wouldn’t boot but I was able to mount it and access the files. I guess
> I’ll just have to wait till my SSD arrives.

I've had pretty poor luck with the 2.5" spinning rust drives. The rock64 
killed a wallmart 1Tb 2.5" usb3 drive with the first kernel build. 
Something is wrong with the rock64's usb3 socket according to the kernel 
experts, but it still works ok with usb2 stuff, and thats what I was 
useing when I built a kernel in an hour.

> I bought the 16gb EMMC also, but I also bought the usb to EMMC
> programmer. I just flashed the EMMC with Etcher. I will reflash it,
> when I am satisfied that I have the OS and linuxcnc working. The usb
> programmer has the same kind of socket on it that the board has, so
> you just mount the EMMC on the programmer and flash it like a uSD,
> then take it off the programmer and mount it on the board. When it is
> mounted on the board the board will boot from it instead of the uSD.
>
I didn't see that, or I would have bought it. Inferred from the 
propaganda was that the rock64 could do that by itself.  But if a dd 
like program could do it, it seems to me dmesg ought to ident the eMMC 
so dd would have a of=/dev/target.  But it doesn't seem to recognize it.  
Does yours before its been programmed?
>
> > Date: Sat, 27 Apr 2019 11:54:19 -0400
> > From: Gene Heskett 
> > To: emc-developers@lists.sourceforge.net
> > Subject: Re: [Emc-developers] buildbot for arm64
> > Message-ID: <201904271154.19577.ghesk...@shentel.net>
> > Content-Type: Text/Plain;  charset="iso-8859-1"
> >
> > On Saturday 27 April 2019 11:08:20 Alan Condit wrote:
> >> I have just completed building 4.19.31 LTS preempt_rt for my
> >> RockPro64 (on my RockPro64). I would love to have access to
> >> linuxcnc builds from a buildbot for arm. 

The Odroid's that they were using to build the armhf for me, were a huge 
crashing headache for John K. so he took them back out of the buildbot 
farm in Sept 2018.  So if I want a new LCNC for my pi, I'm going to have 
to git clone master and build it myself. I don't relish the thought, 
with no more memory than the pi has, it will be a make -j2 at the most 
and several hours, and I expect before I do that, I'd better invest in a 
128 Gb u-sd, clone that 32 its running on now to it with dd, and let 
resize2fs adjust it to use the whole card.

Theres supposed to be a one time fuse in the pi which will make it boot 
from a usb2 drive, but I have run that command to blow that fuse 10 or 
more times w/o its working, so its still stuck booting from the u-sd.
I think my pi's are too early.  And I'm still trying to decide if the pi4 
was an April 1 release, and if its real, is it back compatible?  It will 
take more research, probably with the bull shit filter turned way up.

> >> It took about 8 hours to 
> >> compile the kernel, so I am not looking forward to building
> >> linuxcnc on my RockPro64. I could load machinekit but they seem to
> >> never have a stable release and almost no documentation.

So I've noted...

Take care Alan

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 



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