Reply by "Fri...@fl.priv.at [msp430]"●March 2, 20152015-03-02
Reginald Beardsley p...@yahoo.com [msp430] wrote on 2015-03-02
20:24 MET: > Thanks. Let hope it isn't borked. Playing w/
CCS 6.0 broke a lot of
> stuff. It will talk to the F5529LP, but nothing else can do that
> now. So I went backwards by quite a bit. Which is really annoying
> because I just got a beta of Mecrisp for the F5529LP from Matthias
> but now don't have a functioning way to load an image. The CCS 6.0
> load UI won't accept offsets for text and data. I really am amazed
> that a commercial product would be so screwed up.
Just run the flasher executable without any options and it should detect
any bricked debuggers and recover them. If that doesn't work for you,
I'm afraid to say, use the windows version of MSP430 Flasher in a VM.
Reply by "Reg...@yahoo.com [msp430]"●March 2, 20152015-03-02
Thanks. Let hope it isn't borked. Playing w/ CCS 6.0 broke a lot of
stuff. It will talk to the F5529LP, but nothing else can do that now. So I
went backwards by quite a bit. Which is really annoying because I just got a
beta of Mecrisp for the F5529LP from Matthias but now don't have a
functioning way to load an image. The CCS 6.0 load UI won't accept offsets
for text and data. I really am amazed that a commercial product would be so
screwed up.
Posted by: Reginald Beardsley
Reply by "Fri...@fl.priv.at [msp430]"●March 2, 20152015-03-02
Hello!
Reginald Beardsley p...@yahoo.com [msp430] wrote on 2015-02-28
22:20 MET: > I'll try CCS again when 6.1 comes out.
BTW I just discovered that CCS v6.1 it out, apparently released when I
was at Embedded World and when TI released the new CC26xx SOC's there.
Reply by "Reg...@yahoo.com [msp430]"●March 2, 20152015-03-02
The behavior is inconsistent. On one trial vim took input from the device, but
the command line did not. That seemed odd, but I didn't pursue the matter.
On a 2nd trial some time later using a different terminal window, both vim and
the command line took input from both buttons. Neither case crashed Linux.
However, that means nothing as I've had it work properly several times and
then crash before.
mspdebug 0.23 is having problems with the ezfet firmware which got updated at
least once by CCS. TI's USB firmware for the F5529LP got munged. Prior to
running CCS I had loaded the blinky demo using Energia with the board clocked at
both 16 & 25 MHz. That no longer works. It reports "no unused FET found".
I'll have to figure out how to load a plain hex image using CCS or the
serial bootloader.
Linux crashing is a separate matter, which I do not intend to pursue. I
don't use Linux for anything important. I use it to run software that
can't be compiled on Solaris or is only available as a binary blob for
Linux.
Posted by: Reginald Beardsley
Reply by "Fri...@fl.priv.at [msp430]"●March 2, 20152015-03-02
Reginald Beardsley p...@yahoo.com [msp430] wrote on 2015-03-02
16:26 MET: > I already answered that question, though not
intentionally.
Nope. You said you've changed focus to the console, away from the
editor.
> > .... On one
>> > occasion the source code editor was the active focus and "hello
>> > world" was inserted in the source code which I then removed, changed
>> > focus and then pressed buttons 1 & 2 in the console tab without
>> > problems.
And I should have said ALWAYS keep the focus inside an editor window
when pressing one of the buttons. Even better run an external editor
program, like gedit, kate, kwrite or whatever.
--
MfG / Regards
Friedrich Lobenstock
Posted by: Friedrich Lobenstock
Reply by "Reg...@yahoo.com [msp430]"●March 2, 20152015-03-02
I already answered that question, though not intentionally.
> .... On one > occasion the source code editor was the active focus
and "hello
> world" was inserted in the source code which I then removed, changed
> focus and then pressed buttons 1 & 2 in the console tab without
> problems.
I should note that the board does not show up as an HID on OpenIndiana, the
Solaris derivative. I've not investigated further. There are a number of
potential causes. I suspect that the root cause lies with the TI driverlib not
correctly configuring the USB device. Linux sort of configures it when it
connects, but doesn't get it quite correct. OI cfgadm(1m) shows the device
as an unconfigured USB hub, but does not show either the mass storage device or
the HID. I've not queried the device with CentOS yet but will do that
later today.
Linux would be completely unusable if keyboard input when the focus was not in
an editor or terminal window caused it to crash. So I think the answer is
it's not quite an HID device. It's close enough to trick Linux
whereas OI more correctly refuses to accept its claims.
Have Fun!
Reg
--------
On Mon, 3/2/15, Friedrich Lobenstock f...@fl.priv.at [msp430] wrote:
Subject: [msp430] Re: Need help with msp430 software
To: m...
Date: Monday, March 2, 2015, 8:19 AM
Hi!
Reginald Beardsley p...@yahoo.com [msp430] wrote on
2015-02-28
22:20 MET:
> In the process of doing this I built and ran the
entire demo
> successfully several times. I got both button 1 &
2 to display text
> in the console window. But on one occasion, pressing
button 1 caused
> Linux to crash and on the other, pressing button 2
caused it to
> crash. In both cases the console was the active focus.
On one
> occasion the source code editor was the active focus
and "hello
> world" was inserted in the source code which I
then removed, changed
> focus and then pressed buttons 1 & 2 in the console
tab without
> problems. The last time it crashed I had just pressed
button 1
> twice, gotten "hello world" twice. It then
crashed when I pressed
> button 2. I had touched neither mouse nor keyboard
between pressing
> the 2 buttons.
Just a quick check, does your system work as normal when an
editor
program (gedit, kate, kwrite, LibreOffice Write, or the
like) has the
active focus INSTEAD of the console when you press the
buttons?
--
MfG / Regards
Friedrich Lobenstock
Posted by: Reginald Beardsley
Reply by "Fri...@fl.priv.at [msp430]"●March 2, 20152015-03-02
Hi!
Reginald Beardsley p...@yahoo.com [msp430] wrote on 2015-02-28
22:20 MET: > In the process of doing this I built and ran the
entire demo
> successfully several times. I got both button 1 & 2 to display text
> in the console window. But on one occasion, pressing button 1 caused
> Linux to crash and on the other, pressing button 2 caused it to
> crash. In both cases the console was the active focus. On one
> occasion the source code editor was the active focus and "hello
> world" was inserted in the source code which I then removed, changed
> focus and then pressed buttons 1 & 2 in the console tab without
> problems. The last time it crashed I had just pressed button 1
> twice, gotten "hello world" twice. It then crashed when I pressed
> button 2. I had touched neither mouse nor keyboard between pressing
> the 2 buttons.
Just a quick check, does your system work as normal when an editor
program (gedit, kate, kwrite, LibreOffice Write, or the like) has the
active focus INSTEAD of the console when you press the buttons?
--
MfG / Regards
Friedrich Lobenstock
Posted by: Friedrich Lobenstock
Reply by "Reg...@yahoo.com [msp430]"●February 28, 20152015-02-28
I imported, built and ran the F5529LP OutofBox code limited demo from the
"Development Tools" section of MSP430Ware following the 4 steps given in the
Resource Explorer dialog. I also did the build and run using the CCS UI after
using the Resource Explorer dialog just to create the project. In between I
deleted the projects from disk, not just the project list.
When I started debug on the code it stopped at the watchdog line. I then
resumed execution. CentOS then automatically mounted the mass storage device
and brought up a file manager box showing the files on the device. I did this
5-6 times after it crashed the first time trying to sort out what the 4 steps
were doing and trying to find out where the Resource Explorer was getting its
instructions so I could see what they were.
In the process of doing this I built and ran the entire demo successfully
several times. I got both button 1 & 2 to display text in the console window.
But on one occasion, pressing button 1 caused Linux to crash and on the other,
pressing button 2 caused it to crash. In both cases the console was the active
focus. On one occasion the source code editor was the active focus and "hello
world" was inserted in the source code which I then removed, changed focus and
then pressed buttons 1 & 2 in the console tab without problems. The last time
it crashed I had just pressed button 1 twice, gotten "hello world" twice. It
then crashed when I pressed button 2. I had touched neither mouse nor keyboard
between pressing the 2 buttons.
/var/crash was empty and /var/log/messages showed nothing from the time of the
button 1 crash. I didn't bother to look after the last crash. While I am
aware that other people seem to accept such behavior, I do not.
There are a very large number of Linux distros. I tried several with
unsatisfactory results until I tried CentOS 6.5. That seemed to work with all
the packages (e.g. Energia, OpenOCD, CCS, etc) I was interested in trying out.
After fooling around with the other packages and several ARM boards, I decided
it was time to play with the F5529LP and try CCS. It may well work with
OpenSUSE, but I'm tired of installing and configuring Linux distros just to
discover that B now works, but A is broken. I'll try CCS again when 6.1
comes out.
If you've got CCS 6 working properly on Linux be happy and don't
change your distro. If you decide to try to replicate what I encountered be
careful!
Have Fun!
Reg
FWIW I've been on the calling end of an immediate transfer support call
with Sun Microsystems when I had systems hanging at work. Crashing the host is
a BIG deal with me. Yes, you will crash a Sun server if you unplug the keyboard
(pre-USB) and then plug it back in. But you're not supposed to do that and
will only do it once deliberately. Attempting to format a 3 TB disk on Solaris
10_u8 will kernel panic the system, but you get a crash dump and it's
logged to /var/adm/messages. No one made 3 TB disks when u8 came out. Other
than that, I've never seen a Sun system crash except for hardware faults.
That level of reliability was expected of AIX, IRIX, Ultrix, HP-UX and all the
other *nix systems I've used except for some supercomputers from Intel and
Alliant. Thankfully I managed to avoid intimate contact with those two. So
yeah, I'm an old guy with an attitude problem. I expect things to work
correctly and for things to be fixed promptly when they
don't.
Posted by: Reginald Beardsley
Reply by "Fri...@fl.priv.at [msp430]"●February 28, 20152015-02-28
Hello!
Reginald Beardsley p...@yahoo.com [msp430] wrote on 2015-02-28
15:02 MET: > I got the "black screen of death" on CentOS Linux 3
times playing w/
> the F5529LP. Sometimes CCS worked properly (both button 1 & 2) and
> sometimes it crashed (both button 1 & 2). Lots of UI things (e.g.
> one click project creation, change background color) did not work as
> all. I have no tolerance for software that crashes. I *really*
> don't like it when the OS crashes and does not produce a crash dump
> or even a panic message in the system log file. That's the main
> reason I've done so little w/ MCUs. I fix bugs as soon as I know
> how to reproduce them, but other people seem to prefer adding new
> features instead. If I can find something that works reliably I'll
> continue. Otherwise I'll find a different amusement.
You are not giving any details on which exact steps you're following to
get to this "black screen of death" as you call it.
Have you got a project where this is reproducible? If yes, can you
possibly export it and put it somewhere for download?
--
MfG / Regards
Friedrich Lobenstock
Posted by: Friedrich Lobenstock
Reply by "Reg...@yahoo.com [msp430]"●February 28, 20152015-02-28
Mike,
Sorry wrong guy to ask. I avoid Windows and thus have no experience that could
guide you. I got the "black screen of death" on CentOS Linux 3 times playing w/
the F5529LP. Sometimes CCS worked properly (both button 1 & 2) and sometimes it
crashed (both button 1 & 2). Lots of UI things (e.g. one click project
creation, change background color) did not work as all. I have no tolerance for
software that crashes. I *really* don't like it when the OS crashes and
does not produce a crash dump or even a panic message in the system log file.
That's the main reason I've done so little w/ MCUs. I fix bugs as
soon as I know how to reproduce them, but other people seem to prefer adding new
features instead. If I can find something that works reliably I'll
continue. Otherwise I'll find a different amusement.
I'm an old big systems guy playing w/ toys. There are some real embedded
system pros here. Onestone probably can help. He has high standards, but I
don't think unreasonable.
If you want to use assembly language I suggest you look at Alfred Arnold's
AS:
I've only used it to build Mecrisp forth, but it's got a fabulous
manual and looks absolutely first rate.
Good luck,
Reg
--------
On Sat, 2/28/15, 'Mike Carpenters' m...@sbcglobal.net [msp430]
wrote:
Subject: [msp430] Need help with msp430 software
To: "Reg Beardsley"
Date: Saturday, February 28, 2015, 5:41 AM
Reg,
I have just joined
the msp430
group and am having trouble finding correct info on
programming one of the chips
in the set. 24F169 64 pin device and have not had much luck
programming the
BLINKER demo (partially because the pin outs in the example
are for the wrong
port). Do you know of a good reference book on using the
CCS compiler
under Windows XP or a suitable replacement such as IAR? I am
not concerned
wheather it is supported or not as long as it is workable in
assembly code. Any
help or descriptions of problem areas would be highly
appreciated. Thanks Mike
Carpenter (214)375 9848 internet address
m...@sbcglobal.net
#yiv7020511084 div#yiv7020511084ygrp-mlmsg
#yiv7020511084ygrp-msg p a span.yiv7020511084yshortcuts {
font-family:Verdana;font-size:10px;font-weight:normal;}