Pi-based Co-Pro on the cheap - 100MHz 6502 for £10? (now 274MHz)

for bbc micro/electron hardware, peripherals & programming issues (NOT emulators!)
User avatar
jgharston
Posts: 3246
Joined: Thu Sep 24, 2009 11:22 am
Location: Whitby/Sheffield
Contact:

Re: Pi-based Co-Pro on the cheap - 100MHz 6502 for £10? (now 274MHz)

Post by jgharston » Thu Feb 09, 2017 12:12 am

jms2 wrote:Some other puzzling things I've noticed with the ARM1176 processor are:
a) *GO isn't recognised, and just hangs the machine. By contrast other * commands which need to be passed across the Tube are recognised, eg *DIN works. If *GO isn't implemented shouldn't it just report "Bad command"?
b) Any other unrecognised * commands also seem to cause a hang, rather than "bad command". I discovered this by mistyping something.
I noticed something odd in the ARM Client *command processing, but assumed I was just misreading it. It looks like there is an actual error there. It looks like host errors aren't being collected correctly.
jms2 wrote:c) The banners at the top of the screen report the OS message, then "Master MMFS" then "BASIC". That can't be right though - 6502 basic won't work and sure enough all I get is a * prompt. I don't understand where this banner comes from.
Which banner? The line that says 'BASIC'? That's not the banner, that's the title of the selected language. What is called the banner is the very first line, the line either displayed by the MOS or by the Tube Client on initialisation (and modified by the 'display banner' OSBYTE) , and says something along the lines of "BBC Computer" or "6502 Tube", etc.

The line that says 'BASIC' says 'BASIC' because BASIC is the default language on your system, so on reset, the MOS is selecting BASIC, printing it's title, then passing it over the the Tube Client for the Tube Client to do whatever it wishes to do with it. If the Tube Client is not a 6502 then the Tube Client will generate an error saying something along the lines of 'This is not XXXX code'.

The MOS has absolutely no knowledge or ability to know what CPU is on the other side of the Tube, it's the responsibility of you, the user, to select an appropriate foreground application to run on the other side of the Tube.

Code: Select all

$ bbcbasic
PDP11 BBC BASIC IV Version 0.25
(C) Copyright J.G.Harston 1989,2005-2015
>_

User avatar
hoglet
Posts: 7600
Joined: Sat Oct 13, 2012 6:21 pm
Location: Bristol
Contact:

Re: Pi-based Co-Pro on the cheap - 100MHz 6502 for £10? (now 274MHz)

Post by hoglet » Thu Feb 09, 2017 9:32 am

I should say the native ARM Co Processor is very untested, and will be full of bugs because it's all new code, and it's been my first exposure to a RISC OS like set of APIs.

I'm very pleased people are starting to use it. I'll slowly work through the comments, and create issues on github for each bug, so things don't fall between the cracks. Please do continue to kick the tyres and report anything that doesn't work.

Dave

User avatar
jms2
Posts: 2051
Joined: Mon Jan 08, 2007 6:38 am
Location: Derby, UK
Contact:

Re: Pi-based Co-Pro on the cheap - 100MHz 6502 for £10? (now 274MHz)

Post by jms2 » Thu Feb 09, 2017 11:34 pm

jgharston wrote:Which banner? The line that says 'BASIC'? That's not the banner, that's the title of the selected language. What is called the banner is the very first line, the line either displayed by the MOS or by the Tube Client on initialisation
Just my misuse of terminology, I wasn't sure what term to use to refer to the language title string.
The line that says 'BASIC' says 'BASIC' because BASIC is the default language on your system, so on reset, the MOS is selecting BASIC, printing it's title, then passing it over the the Tube Client for the Tube Client to do whatever it wishes to do with it. If the Tube Client is not a 6502 then the Tube Client will generate an error saying something along the lines of 'This is not XXXX code'.

The MOS has absolutely no knowledge or ability to know what CPU is on the other side of the Tube, it's the responsibility of you, the user, to select an appropriate foreground application to run on the other side of the Tube.
I see, that all makes sense. I now realise (having checked) that the other non 6502 processors do it as well - I'd just never noticed before. I had assumed that the BASIC message meant that Basic had been copied across and was running, which is logical for a 6502 (which I use quite a bit) but not possible for the other types (which I basically don't use at all except out of curiosity). I had thought they just dropped straight to a * prompt without displaying the host language string.

Just for my understanding, in the second processor environment, does the user gain much benefit from knowing what language is selected on the host? After all, said language is not actually running on the host and not necessarily running on the parasite either.

User avatar
fordp
Posts: 965
Joined: Sun Feb 12, 2012 9:08 pm
Location: Kent, England
Contact:

Re: Pi-based Co-Pro on the cheap - 100MHz 6502 for £10? (now 274MHz)

Post by fordp » Fri Feb 10, 2017 7:48 am

jms2 wrote:I had assumed that the BASIC message meant that Basic had been copied across and was running, which is logical for a 6502 (which I use quite a bit) but not possible for the other types (which I basically don't use at all except out of curiosity). I had thought they just dropped straight to a * prompt without displaying the host language string.
It had occurred to me that on the PiCopro we could load Basic at boot if we wanted using tweaks to the client ROM and having basic on the SD Card of the PI?
FordP (Simon Ellwood)
Time is an illusion. Lunchtime, doubly so!

dp11
Posts: 841
Joined: Sun Aug 12, 2012 8:47 pm
Contact:

Re: Pi-based Co-Pro on the cheap - 100MHz 6502 for £10? (now 274MHz)

Post by dp11 » Sat Feb 11, 2017 6:24 pm

A very quick lashup using just 100R series resistors and Diodes to 3.3v to form a very simple level converter all from the old bits box.
20170211_182001[1].jpg

User avatar
hoglet
Posts: 7600
Joined: Sat Oct 13, 2012 6:21 pm
Location: Bristol
Contact:

Re: Pi-based Co-Pro on the cheap - 100MHz 6502 for £10? (now 274MHz)

Post by hoglet » Sat Feb 11, 2017 6:29 pm

dp11 wrote:A very quick lashup using just 100R series resistors and Diodes to 3.3v to form a very simple level converter all from the old bits box.
Nice... Does it work then?

And if so, does that mean you old level shifter was indeed broken?

Dave

dp11
Posts: 841
Joined: Sun Aug 12, 2012 8:47 pm
Contact:

Re: Pi-based Co-Pro on the cheap - 100MHz 6502 for £10? (now 274MHz)

Post by dp11 » Sat Feb 11, 2017 6:50 pm

Yes it works.

I think the pizero has failed in some way. I'm sure I tested with pi2 and got the same results but must be mistaken . Pi2 and Pi3 work with this new rig. PiZero gives the same failed result.

User avatar
jgharston
Posts: 3246
Joined: Thu Sep 24, 2009 11:22 am
Location: Whitby/Sheffield
Contact:

Re: Pi-based Co-Pro on the cheap - 100MHz 6502 for £10? (now 274MHz)

Post by jgharston » Sat Feb 11, 2017 8:27 pm

fordp wrote:It had occurred to me that on the PiCopro we could load Basic at boot if we wanted using tweaks to the client ROM and having basic on the SD Card of the PI?
On the Sprow ARM CoPro, if the code being entered is not ARM code, then if it is 6502 BASIC (ROM type %01xxxxxx) then on-board ARM BASIC is entered instead. My module handler extension sends a module service call out to see if anything wants to take the non-ARM code and do something with it, for example, start up Z80Tube to run Z80 code, or start up 65Tube to run 6502 code.

Code: Select all

$ bbcbasic
PDP11 BBC BASIC IV Version 0.25
(C) Copyright J.G.Harston 1989,2005-2015
>_

User avatar
simoni
Posts: 438
Joined: Wed May 25, 2016 6:18 pm
Contact:

Re: Pi-based Co-Pro on the cheap - 100MHz 6502 for £10? (now 274MHz)

Post by simoni » Sun Feb 12, 2017 7:19 pm

I'm curious about running BASIC 1.35 in the ARM1176 emulation... I've seen a couple of posts referring to it but I can't find any instructions for how it's done; perhaps my Google-Fu is lacking, but a pointer would be much appreciated :)

User avatar
hoglet
Posts: 7600
Joined: Sat Oct 13, 2012 6:21 pm
Location: Bristol
Contact:

Re: Pi-based Co-Pro on the cheap - 100MHz 6502 for £10? (now 274MHz)

Post by hoglet » Sun Feb 12, 2017 7:43 pm

simoni wrote:I'm curious about running BASIC 1.35 in the ARM1176 emulation... I've seen a couple of posts referring to it but I can't find any instructions for how it's done; perhaps my Google-Fu is lacking, but a pointer would be much appreciated :)
It's not really documented.

You can get the original BAS135 executable from here:
http://mdfs.net/User/JGH/Develop/ARMCoPro/

You need to add it into a .SSD disk image with a load address of &F000 and an execution address of &F104.

Also, I think it fails to run on a Pi 3 for some reason.

Dave

User avatar
simoni
Posts: 438
Joined: Wed May 25, 2016 6:18 pm
Contact:

Re: Pi-based Co-Pro on the cheap - 100MHz 6502 for £10? (now 274MHz)

Post by simoni » Sun Feb 12, 2017 7:51 pm

Also, I think it fails to run on a Pi 3 for some reason.
Yes, I just tried it... displays 'BASIC' and then hangs :( Is there any advantage/disadvantage to using a different Pi (I have 1, 2 and 3 but no zeros)?

Thanks for the prompt help btw :)

dp11
Posts: 841
Joined: Sun Aug 12, 2012 8:47 pm
Contact:

Re: Pi-based Co-Pro on the cheap - 100MHz 6502 for £10? (now 274MHz)

Post by dp11 » Sun Feb 12, 2017 8:35 pm

For the 6502 fast core ( with default Pi clock settings)

PiZero ( 1000MHz ARMv6 core ) : 274MHz
Pi3 (1200MHz ARMv8 core Dual issue ) : 260MHz
Pi2 ( 900MHz ARMv7 core some Dual issue options ): 187MHz

Cores in C might do better with the ARMv8 dual issue core.
I'm looking at if there are any improvements to be had with ARMv7 and ARMv8 pipelining, but as he code sequence for the emulated instructions in the 6502 core are so short there is little scope for more pipeline optimisation/Dual issue.
Last edited by dp11 on Sun Feb 12, 2017 9:02 pm, edited 1 time in total.

User avatar
jgharston
Posts: 3246
Joined: Thu Sep 24, 2009 11:22 am
Location: Whitby/Sheffield
Contact:

Re: Pi-based Co-Pro on the cheap - 100MHz 6502 for £10? (now 274MHz)

Post by jgharston » Sun Feb 12, 2017 8:53 pm

hoglet wrote:You can get the original BAS135 executable from here:
http://mdfs.net/User/JGH/Develop/ARMCoPro/
You need to add it into a .SSD disk image with a load address of &F000 and an execution address of &F104.
Try F12C. The first word in an ARM module is the offset to the entry point, and the first word of BASIC135 is 0000012C. An ARM module is run by doing JMP start+!start, not by doing JMP start*. I've just tested on BeebEm:
A*> LOAD BASIC35 20000
A*> GO 2012C
ARM BBC BASIC V
>
It then starts falling over as the ARMEval doesn't have enough API calls to support it.

*Strictly speaking, by doing IF start?3=0 THEN JMP start+!start ELSE JMP start

Edit: in my 6809 emulator as lots of 6809 BIOS entries are called with JSR (addr) I added extra syntax to the *GO command to allow *GO addr and *GO (addr). I think I'll update my disk-based *GO command to add *GO [addr] to do an ARM-style indirect offset jump.

Code: Select all

$ bbcbasic
PDP11 BBC BASIC IV Version 0.25
(C) Copyright J.G.Harston 1989,2005-2015
>_

User avatar
jms2
Posts: 2051
Joined: Mon Jan 08, 2007 6:38 am
Location: Derby, UK
Contact:

Re: Pi-based Co-Pro on the cheap - 100MHz 6502 for £10? (now 274MHz)

Post by jms2 » Sun Feb 12, 2017 9:07 pm

I'm pretty sure that F12C was the execution address of the file when I first transferred it to an ssd image (ie, the exec address in the . INF file). However, I'll check when I'm back with my Beebs next week.

Edit:

Actually I think I have figured it out. Unrecognised commands don't work, so *BAS135 will hang for that reason. Maybe *RUN BAS135 will work.

User avatar
simoni
Posts: 438
Joined: Wed May 25, 2016 6:18 pm
Contact:

Re: Pi-based Co-Pro on the cheap - 100MHz 6502 for £10? (now 274MHz)

Post by simoni » Sun Feb 12, 2017 10:03 pm

I set the execution address to F12C; in ARM1176 mode *BASIC135 didn't work, *LOAD BASIC135 20000 and *GO 2012C didn't work, also *RUN BASIC135 didn't work.

Interestingly, in the other ARM mode 12 (ARM2?) *BASIC135 caused BASIC V to start, but it crashed as soon as I typed LIST.

User avatar
jgharston
Posts: 3246
Joined: Thu Sep 24, 2009 11:22 am
Location: Whitby/Sheffield
Contact:

Re: Pi-based Co-Pro on the cheap - 100MHz 6502 for £10? (now 274MHz)

Post by jgharston » Mon Feb 13, 2017 12:42 am

I've updated the *GO command on the Tube Utilities Disk to all indirect calls. You can now do */GO addr to jump to the address (as a 6502 JMP addr), */GO (addr) to jump to the destination pointed to by the address (as a 6502 JMP (addr)), and */GO [addr] to jump to address+!address as with a ARM module entry point.

So, you should be able to do:
ARM>*LOAD BASIC137 200000
ARM>*/GO [200000]

Code: Select all

$ bbcbasic
PDP11 BBC BASIC IV Version 0.25
(C) Copyright J.G.Harston 1989,2005-2015
>_

User avatar
jms2
Posts: 2051
Joined: Mon Jan 08, 2007 6:38 am
Location: Derby, UK
Contact:

Re: Pi-based Co-Pro on the cheap - 100MHz 6502 for £10? (now 274MHz)

Post by jms2 » Mon Feb 13, 2017 6:31 am

That sounds ideal, thanks!

Would be good if someone could check it out, as I'm not in a position to do so for a while.

The thing I don't understand is why *Run doesn't work, if the exec address is set correctly.

User avatar
hoglet
Posts: 7600
Joined: Sat Oct 13, 2012 6:21 pm
Location: Bristol
Contact:

Re: Pi-based Co-Pro on the cheap - 100MHz 6502 for £10? (now 274MHz)

Post by hoglet » Mon Feb 13, 2017 8:33 am

I'll try to take a look at this today.

I think part of the confusion is there are two executables in the field, with different size module headers, leading to different execution addresses (F104 vs F12C)

The version I built from sources includes it's own error messages:
https://github.com/hoglet67/PiTubeClien ... a671ebd388

This is so the environment doesn't need to contain them.

Dave

User avatar
fordp
Posts: 965
Joined: Sun Feb 12, 2012 9:08 pm
Location: Kent, England
Contact:

Re: Pi-based Co-Pro on the cheap - 100MHz 6502 for £10? (now 274MHz)

Post by fordp » Mon Feb 13, 2017 9:05 am

hoglet wrote:Hi Thomas,

I see you have just normalized all the line endings across the whole repository:
https://github.com/stardot/b-em/commit/ ... 4a1beb5cf9

I think this will make it impossible now to merge all the different forks back together, as all you will get is conflicts (unless git's merge is far smarter than I give it credit for).

I wonder if it's worth doing a test to confirm this (or not) before more work is done.

I also have some concerns about relying on core.autocrlf (if this is indeed your plan). Is there a way to set this at the repository level?

Dave
Hi Dave,
It is funny that you just posted this elsewhere as I am having great trouble with line endings on my PiDirect fork. Git is using windows line endings and it is breaking all the bash scripts. I am fighting to fix it. I tried to enforce linux line endings both in the repository and on my local machine but it is not working yet :(

I am just trying to merge the latest changes in to my fork to keep it current.

I will keep trying later in the week.
FordP (Simon Ellwood)
Time is an illusion. Lunchtime, doubly so!

User avatar
hoglet
Posts: 7600
Joined: Sat Oct 13, 2012 6:21 pm
Location: Bristol
Contact:

Re: Pi-based Co-Pro on the cheap - 100MHz 6502 for £10? (now 274MHz)

Post by hoglet » Mon Feb 13, 2017 9:11 am

fordp wrote: It is funny that you just posted this elsewhere as I am having great trouble with line endings on my PiDirect fork. Git is using windows line endings and it is breaking all the bash scripts. I am fighting to fix it. I tried to enforce linux line endings both in the repository and on my local machine but it is not working yet :(
You need to make sure core.autocrlf=false

If it's set to true, then the windows git client will change all line endings to windows line endings on check out.

Try a "git config --list" and see what it's current set to.

Dave

User avatar
hoglet
Posts: 7600
Joined: Sat Oct 13, 2012 6:21 pm
Location: Bristol
Contact:

Re: Pi-based Co-Pro on the cheap - 100MHz 6502 for £10? (now 274MHz)

Post by hoglet » Mon Feb 13, 2017 9:58 am

I'm attaching a couple of disk images for testing.

The first contains the original BAS135 (same at the one on Jonathan's site, OWNERRORS=0):
basic135_orig.zip
(37.41 KiB) Downloaded 27 times
The second contains the my recompiled version of BAS135 with local error messages (OWNERRORS=1)
basic135_new.zip
(67.8 KiB) Downloaded 38 times
Both of these seem to work for me on a PI Zero, using MMFS, at least some of the time!

The original version:

Code: Select all

*FX 151,230,15
<ctrl break>
ARM1176 Co Processor 1000MHz

Model B MMFS

BASIC

arm>*DIN15
arm>*INFO BAS135
$.BAS135      00F000 00F12C 00EF48 002
arm>*/BAS135
ARM BBC BASIC V

Starting with 2060540 bytes free.

>_
The rebuilt version:

Code: Select all

*FX 151,230,15
<ctrl break>
ARM1176 Co Processor 1000MHz

Model B MMFS

BASIC

arm>*DIN14
arm>*INFO BAS135
$.BAS135      00F000 00F104 00EFC0 002
arm>*/BAS135
ARM BBC BASIC V version 1.35 (C) Acorn 1989

Starting with 2060540 bytes free.

>_
However, a couple of times I did get the executable just hanging, so there might be an issue here. It's not at all repeatable though.

What filesystem are you using?

Dave

User avatar
hoglet
Posts: 7600
Joined: Sat Oct 13, 2012 6:21 pm
Location: Bristol
Contact:

Re: Pi-based Co-Pro on the cheap - 100MHz 6502 for £10? (now 274MHz)

Post by hoglet » Mon Feb 13, 2017 10:21 am

Just a couple more thoughts.

I suspect the non-repeatable behaviour I'm seeing is a cache consistency issue. Both BBC Basic versions work fine from a cold start, but then trying to switch to the other version usually causes a crash. After loading code, it's important to ensure that the data cache is flushed, and the instruction cache is purged. I suspect I'm not doing that correctly here:
https://github.com/hoglet67/PiTubeDirec ... asm.S#L150

I remember trying to get this and Panos working exposed a lot of MMFS bugs, so those may be in other file systems. e.g. large (>64KB) file support.
https://github.com/hoglet67/MMFS/commit ... b22fdfe79d

Dave

User avatar
hoglet
Posts: 7600
Joined: Sat Oct 13, 2012 6:21 pm
Location: Bristol
Contact:

Re: Pi-based Co-Pro on the cheap - 100MHz 6502 for £10? (now 274MHz)

Post by hoglet » Mon Feb 13, 2017 10:48 am

jms2 wrote: a) *GO isn't recognised, and just hangs the machine. By contrast other * commands which need to be passed across the Tube are recognised, eg *DIN works. If *GO isn't implemented shouldn't it just report "Bad command"?
GO is one of the built-in commands.

So,
*LOAD BAS135 F000
*GO F12C
should work for the original version.

The built-in commands can be seen with *HELP ARM
HELP
TEST
GO
MEM
DIS
FILL
CRC

Their implementation is here.
https://github.com/hoglet67/PiTubeDirec ... commands.c

I'm not doing much error checking on parameters, so I suspect *GO without an address might jump into the weeds.
jms2 wrote: b) Any other unrecognised * commands also seem to cause a hang, rather than "bad command". I discovered this by mistyping something.
I don't see this issue at all.

Can you provide an exact log of what you did so I can try to reproduce it?

Dave

User avatar
hoglet
Posts: 7600
Joined: Sat Oct 13, 2012 6:21 pm
Location: Bristol
Contact:

Re: Pi-based Co-Pro on the cheap - 100MHz 6502 for £10?

Post by hoglet » Mon Feb 13, 2017 12:09 pm

jms2 wrote:I've just tried out the latest build (using the GPU to deal with Tube transfers) on my Master Turbo. This is using an RPi2.
Are you still using the RPI2?

That's almost guaranteed to be your problem with the Native ARM Co Pro, as the cache management is very different. It's definitely failing to load BAS135 on my Pi 3.

It currently only works on the single core Pis.

I've creates an issue for this:
https://github.com/hoglet67/PiTubeDirect/issues/25

I think this is a recent regression, so I'll try some older versions....

Dave

User avatar
hoglet
Posts: 7600
Joined: Sat Oct 13, 2012 6:21 pm
Location: Bristol
Contact:

Re: Pi-based Co-Pro on the cheap - 100MHz 6502 for £10? (now 274MHz)

Post by hoglet » Mon Feb 13, 2017 3:19 pm

I've fixed three separate issues here with the Native ARM Co Pro:
https://github.com/hoglet67/PiTubeDirect/issues/25
https://github.com/hoglet67/PiTubeDirect/issues/26
https://github.com/hoglet67/PiTubeDirect/issues/27

These fixed are in the boa-dev branch.

With these fixes, both builds of BAS135 start reliably on the Pi Models I've been able to test (0 and 3).

John, are you able to build this, or do you need me to post a snapshot build?

Dave

User avatar
jms2
Posts: 2051
Joined: Mon Jan 08, 2007 6:38 am
Location: Derby, UK
Contact:

Re: Pi-based Co-Pro on the cheap - 100MHz 6502 for £10? (now 274MHz)

Post by jms2 » Mon Feb 13, 2017 5:09 pm

Yes Dave, I'm using a Pi2. The process I followed was exactly the same as in your code snippet above, where you describe the use of *GO. I didn't try *HELP ARM, because just doing a *HELP with no parameters didn't reveal any special ARM commands.

It would probably help if you could post up a precomputed version for me, but there's no hurry, I won't be able to do anything with it until the middle of next week anyway.

Thanks for looking into it! :D

User avatar
hoglet
Posts: 7600
Joined: Sat Oct 13, 2012 6:21 pm
Location: Bristol
Contact:

Re: Pi-based Co-Pro on the cheap - 100MHz 6502 for £10? (now 274MHz)

Post by hoglet » Mon Feb 13, 2017 5:16 pm

jms2 wrote: It would probably help if you could post up a precomputed version for me, but there's no hurry, I won't be able to do anything with it until the middle of next week anyway.
Here you go, the latest and greatest from the boa-dev branch:
PiTubeDirect_20170213_1713_dmb.zip
(2.08 MiB) Downloaded 19 times
Dave

User avatar
jgharston
Posts: 3246
Joined: Thu Sep 24, 2009 11:22 am
Location: Whitby/Sheffield
Contact:

Re: Pi-based Co-Pro on the cheap - 100MHz 6502 for £10? (now 274MHz)

Post by jgharston » Mon Feb 13, 2017 11:49 pm

hoglet wrote:
jms2 wrote: b) Any other unrecognised * commands also seem to cause a hang, rather than "bad command". I discovered this by mistyping something.
I don't see this issue at all.
Can you provide an exact log of what you did so I can try to reproduce it?
Also, does error reporting also fail when in BASIC?
ie, test from the supervisor:
ARM> *fred
and test from BASIC:
>*fred
(assuming 'fred' should give a bad command)

Code: Select all

$ bbcbasic
PDP11 BBC BASIC IV Version 0.25
(C) Copyright J.G.Harston 1989,2005-2015
>_

User avatar
jms2
Posts: 2051
Joined: Mon Jan 08, 2007 6:38 am
Location: Derby, UK
Contact:

Re: Pi-based Co-Pro on the cheap - 100MHz 6502 for £10? (now 274MHz)

Post by jms2 » Tue Feb 14, 2017 6:30 am

Jonathan, if I understand the recent posts correctly, all of my issues were caused by an incompatibility with multi core pis, which Dave has now resolved. So hopefully any potential further aspects of the bug I reported will have been cured as well. I'll certainly check though.

User avatar
hoglet
Posts: 7600
Joined: Sat Oct 13, 2012 6:21 pm
Location: Bristol
Contact:

Re: Pi-based Co-Pro on the cheap - 100MHz 6502 for £10? (now 274MHz)

Post by hoglet » Tue Feb 14, 2017 7:25 am

jms2 wrote:Jonathan, if I understand the recent posts correctly, all of my issues were caused by an incompatibility with multi core pis, which Dave has now resolved. So hopefully any potential further aspects of the bug I reported will have been cured as well. I'll certainly check though.
The only issue I could not reproduce was the error handing failure, i.e. on my Pi 3 *JIBBERISH would give Bad Command. So it would be interesting to see if you still have that issue.

Dave

Post Reply