Reply by Joe Radomski October 26, 20122012-10-26
FYI with a windows host, the free versions of vmware products are all that most people need.  In fact I have a friend that works for them and he told me they purposedly do it this way as they test new features on the free versions and then move them to the premium version once its stable...
as to virtualbox I'll have to give it another look as well... Overall it seemed to work well with the exception of the USB passthrough..
 

>________________________________
>From: Peter Johansson
>To: m...
>Sent: Wednesday, October 24, 2012 12:59 PM
>Subject: Re: [msp430] Windows 7 32-bit and IAR MSP430 workbench IDE
>

>On Tue, Oct 23, 2012 at 4:42 PM, Joe Radomski wrote:
>
>> I am not a mac owner so I havent tried parallels, overall virtual box is a good product but found it doesn't do well with providing real access to the USB ports.
>
>This was definitely the case with older versions of VirtualBox, but I
>hear things have gotten much better with the latest versions.
>
>VMWare was the first to provide solid USB passthrough capabilities so
>that is what I used when I was first making the migration to OSX a
>number of years back. I have been happy with VMware so I have stuck
>with it.
>
>I generally recommend VirtualBox to people just getting started with
>virtualization. If it gets the job done there is generally no need to
>pop for a commercial implementation.
>
>-p.



Beginning Microcontrollers with the MSP430

Reply by Peter Johansson October 24, 20122012-10-24
On Tue, Oct 23, 2012 at 4:42 PM, Joe Radomski wrote:

> I am not a mac owner so I havent tried parallels, overall virtual box is a good product but found it doesn't do well with providing real access to the USB ports.

This was definitely the case with older versions of VirtualBox, but I
hear things have gotten much better with the latest versions.

VMWare was the first to provide solid USB passthrough capabilities so
that is what I used when I was first making the migration to OSX a
number of years back. I have been happy with VMware so I have stuck
with it.

I generally recommend VirtualBox to people just getting started with
virtualization. If it gets the job done there is generally no need to
pop for a commercial implementation.

-p.
Reply by David Brown October 24, 20122012-10-24
On 24/10/2012 03:21, Alex Belov wrote:
> "I have found USB with VirtualBox running on a Windows host to be a bit
> flaky"
>
> IAR works fine with USB connection into VirtualBox for Windows host.
> Tested with SEGGER clone and TI ARM launchpad. Both OK.
>

My point is just that problems with USB to VB on a Windows host are not
unknown - either to me or to others - but I have never had any trouble
when using Linux as the host.

My experience with VB on a Windows host is that if you use a USB device
that has no drivers or programs associated with it on the host, then it
usually works perfectly. But if there are drivers installed in the host
system, then the Windows host can get confused when passing it over to
VB as it has to switch from the host drivers to the VB passthrough
drivers. So if you are planning on using devices within VB machines
(and this probably applies to VMWare too), you minimise the chances of
problems if you avoid installing drivers in the host system.

But if you are building a new machine as a host for virtual machines,
and you have the choice, then I strongly recommend Linux as the host
machine.
> On 10/23/12, David Brown > > wrote:
> > On 23/10/12 22:42, Joe Radomski wrote:
> >> I am not a mac owner so I havent tried parallels, overall virtual box is
> >> a good product but found it doesn't do well with providing real access
> >> to the USB ports. I have some hardrive recovery hardware that the only
> >> way I can get it to work is to use VMware on xp SP3 (or a real xp box)
> >>
> >> vmware (at least on a windows box) allows you to take over a port..
> >>
> >
> > I have found USB with VirtualBox running on a Windows host to be a bit
> > flaky - but I haven't tried it for a while, and there have been many big
> > improvements in VB since I last used it on Windows.
> >
> > Running on a Linux host, I have had no problems with USB in VirtualBox.
> > I use it daily, with both Linux and Windows guests.
> >
> > It is a long time since I've tried VMWare, so I can't judge it - but I
> > started using VB with about version 2, and haven't looked back.
> >
> > (I haven't used IAR tools, and my msp430 tools are all quite old, so I
> > have no experience at this particular setup.)
> >
>

Reply by Alex Belov October 23, 20122012-10-23
"I have found USB with VirtualBox running on a Windows host to be a bit
flaky"

IAR works fine with USB connection into VirtualBox for Windows host.
Tested with SEGGER clone and TI ARM launchpad. Both OK.

On 10/23/12, David Brown wrote:
> On 23/10/12 22:42, Joe Radomski wrote:
>> I am not a mac owner so I havent tried parallels, overall virtual box is
>> a good product but found it doesn't do well with providing real access
>> to the USB ports. I have some hardrive recovery hardware that the only
>> way I can get it to work is to use VMware on xp SP3 (or a real xp box)
>>
>> vmware (at least on a windows box) allows you to take over a port..
>> I have found USB with VirtualBox running on a Windows host to be a bit
> flaky - but I haven't tried it for a while, and there have been many big
> improvements in VB since I last used it on Windows.
>
> Running on a Linux host, I have had no problems with USB in VirtualBox.
> I use it daily, with both Linux and Windows guests.
>
> It is a long time since I've tried VMWare, so I can't judge it - but I
> started using VB with about version 2, and haven't looked back.
>
> (I haven't used IAR tools, and my msp430 tools are all quite old, so I
> have no experience at this particular setup.)
Reply by Mike Raines October 23, 20122012-10-23
Peter,

Thanks for the reply. My last job was on linux but that was 12 years ago and since I'm now 62, I doubt if I will ever go back.
Now I code embedded devices in c and c++ and am forced to develop configuration and testing programs in a Windows environment.
I must admit that I still miss Emacs sometimes.....

Thanks again,
Mike Raines

________________________________
From: m... [mailto:m...] On Behalf Of Peter Johansson
Sent: Tuesday, October 23, 2012 3:00 PM
To: m...
Subject: Re: [msp430] Windows 7 32-bit and IAR MSP430 workbench IDE

On Tue, Oct 23, 2012 at 8:59 AM, Joe Radomski > wrote:

> I have found the best way to run a virtual machine with real hardware is to use VMware and run a virtual XP session... it works great with all sorts of USB drivers that seem to have trouble on win 7 and other virtual machines.. Be warned that VMware must take over exclusive use of the USB ports while running..

I am a Unix (primarily OSX) person and prefer to do things natively in
a Unix environment. But when Windows is the only option, I have had
the best results running Windows XP in a VM. I use VMware as well,
but I hear that both Parallels and VirtualBox are viable options.

I use a pre-patched and stripped installer made with nLite to produce
a lightweight base install. I then fork this base image and install
only the tools required for a particular toolset of project. Every
toolset/project gets its own isolated VM.

I have found that this approach leads to a *drastic* reduction in the
number of headaches.

-p.



Reply by David Brown October 23, 20122012-10-23
On 23/10/12 22:42, Joe Radomski wrote:
> I am not a mac owner so I havent tried parallels, overall virtual box is
> a good product but found it doesn't do well with providing real access
> to the USB ports. I have some hardrive recovery hardware that the only
> way I can get it to work is to use VMware on xp SP3 (or a real xp box)
>
> vmware (at least on a windows box) allows you to take over a port..
>

I have found USB with VirtualBox running on a Windows host to be a bit
flaky - but I haven't tried it for a while, and there have been many big
improvements in VB since I last used it on Windows.

Running on a Linux host, I have had no problems with USB in VirtualBox.
I use it daily, with both Linux and Windows guests.

It is a long time since I've tried VMWare, so I can't judge it - but I
started using VB with about version 2, and haven't looked back.

(I haven't used IAR tools, and my msp430 tools are all quite old, so I
have no experience at this particular setup.)

Reply by Mike Raines October 23, 20122012-10-23
Hey Folks,

Thanks to all who replied. Since I wrote, I have received much better response from IAR. Perhaps it was the urgency of my appeal, but mostly I think it got better when I found a phone number to the East Coast Tech support. Previously, I had been calling and emailing the main IAR tech support that seems to be under a Dave Bailey in California. I was lucky enough to get a tech support guy in Massachusetts named Colin who stuck with us until we got it working. First we tried to make it work right on the old Win XP machine. This was successful, except whenever a debug session is closed by clicking on the red "Stop Debugging" X icon, I get the message that one or more stacks have overflowed. He said that since whenever we used the "Red Hand" icon to break execution and observe the stack, there were no overflows, or even close, I should ignore the overflow warnings that come up when using the red X to stop debugging.

I then switched to the new Windows 7 32-bit machine, removed IAR under XP virtualization, and under Win7, and installed a clean version from the IAR site. Colin called back and, using GoToMeeting, checked out my problems. I did forget to ask him about the two new icons IAR added to my start menu: IAR License Activation, and IAR License Manager?

We were able to get things working properly, although I sometimes still get the stack overflow warning... Mostly my problems resulted from settings that got changed in the machine change. After several changes, the clues did not tip me off to the problems. I also have some ESD issues with my development layout. Whenever I come back with a cup of coffee (actually even without coffee, or tea) and sit down at the desk and take off my shoes, my MSP430F5438 device resets.

Many thanks to Colin and the East Coast IAR Tech Support team. I hate to badmouth IAR, but it has been my experience that sometimes I have to bump the complaints up from tech support to sales or customer care to get real results. Maybe things will be different now that I have the East Coast tech support number.

Again, Thanks to you guys,
Mike Raines

________________________________
From: m... [mailto:m...] On Behalf Of Jon Kirwan
Sent: Monday, October 22, 2012 5:38 PM
To: MSP430 List
Subject: Re: [msp430] Windows 7 32-bit and IAR MSP430 workbench IDE

On Mon, 22 Oct 2012 14:25:42 -0000, you wrote:

>
>Hey Folks,
> I'm about at my wits end with trying to get IAR to work
> with my new Windows 7 32-bit computer. Has anyone else
> experienced these problems, and can anyone please tell me
> how they overcame the issues?
>
> At first, I installed the latest version listed in my
> IAR "my pages" section: 5.50.1. I had many issues getting
> past the licensing and dongle issues. The installation
> placed two new icons on my start menu: IAR license manager
> and IAR license activator, or some such. I finally got past
> these issues by applying patches suggested in two 3-page
> documents from IAR. Now I can view, edit, compile, and link
> my software, but cannot download or debug it. That is not
> entirely true. Maybe it will download and execute 20% of
> the tries, but the debugger will not work at all.
> Typically, the debugger does not stop at main, breakpoints
> do not work, and the dissassembly window shows address 0.
> Also, whenever I end a debug session, I get error messages,
> "one or more stacks have overflowed" or similar.
>
> I then tried the compatibility resolution features
> included with Windows 7 for running older software. No
> improvement. Then I downloaded and installed all the files
> needed to enable Windows 7 to run virtual machine in Windows
> XP, which was what I was using before. I also had to enable
> hardware virtualization in the BIOS. I then removed IAR and
> installed a clean version under virtual Windows XP. After
> another fight involving dongle recognition, I gave up on
> this and , as a sanity check, decided to go back to my old
> Win XP machine.
>
> After some initial problems, it now works (on the old XP
> machine) except I still get the stack overflow errors upon
> ending debug sessions. I have tried two different TI FET
> 430 UIF tools, same results.
>
> Is the IAR MPS430 workbench supposed to work under
> Windows 7? If not, is there a version that does? If it is
> supposed to work why do I have to go through 6 pages of
> fixes to make it work? We paid IAR a lot of bucks for this
> software and we pay $800.00 per year for support. We
> deserve software that works. Their tech support has tried
> twice to fix this and apparently have given up as they no
> longer answer the phone or answer my emails. Anders
> Lindgren, are you out there?

>At My Wits End,
>Mike Raines

I think I wrote about this problem (and others, as well) a
while back in this group. In my case, I was thinking it might
be related to my 64-bit Win7, as opposed to a 32-bit Win7 I
am not running right not. But the problems are the same with
regards to downloading and executing only a small percentage
of the tries. I don't use the IAR kickstart, or feel
comfortable recommending it, much anymore because of that.
Paul Curtis, I believe, suggested that this may be more a
problem with TI-supplied drivers than a matter of IAR's
fault. But Paul didn't say he was having the same troubles
with his own products. So it's not a strong argument without
that additional information. Just a suggestion.

The rest doesn't surprise me, anymore.

I _AM_ surprised that your working through tech suppport has
not produced results, though. Or that Anders hasn't had
something definite to say about it, here. The facts of that
do bother me.

I hope we get to hear the fuller story as it evolves.

Thanks,
Jon



Reply by Joe Radomski October 23, 20122012-10-23
I am not a mac owner so I havent tried parallels, overall virtual box is a good product but found it doesn't do well with providing real access to the USB ports.  I have some hardrive recovery hardware that the only way I can get it to work is to use VMware on xp SP3 (or a real xp box)
vmware (at least on a windows box) allows you to take over a port..

>________________________________
> From: Peter Johansson
>To: m...
>Sent: Tuesday, October 23, 2012 2:59 PM
>Subject: Re: [msp430] Windows 7 32-bit and IAR MSP430 workbench IDE
>
>

>On Tue, Oct 23, 2012 at 8:59 AM, Joe Radomski wrote:
>
>> I have found the best way to run a virtual machine with real hardware is to use VMware and run a virtual XP session... it works great with all sorts of USB drivers that seem to have trouble on win 7 and other virtual machines.. Be warned that VMware must take over exclusive use of the USB ports while running..
>
>I am a Unix (primarily OSX) person and prefer to do things natively in
>a Unix environment. But when Windows is the only option, I have had
>the best results running Windows XP in a VM. I use VMware as well,
>but I hear that both Parallels and VirtualBox are viable options.
>
>I use a pre-patched and stripped installer made with nLite to produce
>a lightweight base install. I then fork this base image and install
>only the tools required for a particular toolset of project. Every
>toolset/project gets its own isolated VM.
>
>I have found that this approach leads to a *drastic* reduction in the
>number of headaches.
>
>-p.
>
>



Reply by Peter Johansson October 23, 20122012-10-23
On Tue, Oct 23, 2012 at 8:59 AM, Joe Radomski wrote:

> I have found the best way to run a virtual machine with real hardware is to use VMware and run a virtual XP session... it works great with all sorts of USB drivers that seem to have trouble on win 7 and other virtual machines.. Be warned that VMware must take over exclusive use of the USB ports while running..

I am a Unix (primarily OSX) person and prefer to do things natively in
a Unix environment. But when Windows is the only option, I have had
the best results running Windows XP in a VM. I use VMware as well,
but I hear that both Parallels and VirtualBox are viable options.

I use a pre-patched and stripped installer made with nLite to produce
a lightweight base install. I then fork this base image and install
only the tools required for a particular toolset of project. Every
toolset/project gets its own isolated VM.

I have found that this approach leads to a *drastic* reduction in the
number of headaches.

-p.
Reply by Joe Radomski October 23, 20122012-10-23
I have found the best way to run a virtual machine with real hardware is to use VMware and run a virtual XP session... it works great with all sorts of USB drivers that seem to have trouble on win 7 and other virtual machines.. Be warned that VMware must take over exclusive use of the USB ports while running..
 
 
>________________________________
>From: Alex Belov
>To: m...
>Sent: Monday, October 22, 2012 11:03 PM
>Subject: Re: [msp430] Windows 7 32-bit and IAR MSP430 workbench IDE
>

>One of the case use VM and run IAR under winxp there.
>
>On 10/22/12, Jon Kirwan wrote:
>> On Mon, 22 Oct 2012 14:25:42 -0000, you wrote:
>>
>>>
>>>Hey Folks,
>>> I'm about at my wits end with trying to get IAR to work
>>> with my new Windows 7 32-bit computer. Has anyone else
>>> experienced these problems, and can anyone please tell me
>>> how they overcame the issues?
>>>
>>> At first, I installed the latest version listed in my
>>> IAR "my pages" section: 5.50.1. I had many issues getting
>>> past the licensing and dongle issues. The installation
>>> placed two new icons on my start menu: IAR license manager
>>> and IAR license activator, or some such. I finally got past
>>> these issues by applying patches suggested in two 3-page
>>> documents from IAR. Now I can view, edit, compile, and link
>>> my software, but cannot download or debug it. That is not
>>> entirely true. Maybe it will download and execute 20% of
>>> the tries, but the debugger will not work at all.
>>> Typically, the debugger does not stop at main, breakpoints
>>> do not work, and the dissassembly window shows address 0.
>>> Also, whenever I end a debug session, I get error messages,
>>> "one or more stacks have overflowed" or similar.
>>>
>>> I then tried the compatibility resolution features
>>> included with Windows 7 for running older software. No
>>> improvement. Then I downloaded and installed all the files
>>> needed to enable Windows 7 to run virtual machine in Windows
>>> XP, which was what I was using before. I also had to enable
>>> hardware virtualization in the BIOS. I then removed IAR and
>>> installed a clean version under virtual Windows XP. After
>>> another fight involving dongle recognition, I gave up on
>>> this and , as a sanity check, decided to go back to my old
>>> Win XP machine.
>>>
>>> After some initial problems, it now works (on the old XP
>>> machine) except I still get the stack overflow errors upon
>>> ending debug sessions. I have tried two different TI FET
>>> 430 UIF tools, same results.
>>>
>>> Is the IAR MPS430 workbench supposed to work under
>>> Windows 7? If not, is there a version that does? If it is
>>> supposed to work why do I have to go through 6 pages of
>>> fixes to make it work? We paid IAR a lot of bucks for this
>>> software and we pay $800.00 per year for support. We
>>> deserve software that works. Their tech support has tried
>>> twice to fix this and apparently have given up as they no
>>> longer answer the phone or answer my emails. Anders
>>> Lindgren, are you out there?
>>
>>>At My Wits End,
>>>Mike Raines
>>
>> I think I wrote about this problem (and others, as well) a
>> while back in this group. In my case, I was thinking it might
>> be related to my 64-bit Win7, as opposed to a 32-bit Win7 I
>> am not running right not. But the problems are the same with
>> regards to downloading and executing only a small percentage
>> of the tries. I don't use the IAR kickstart, or feel
>> comfortable recommending it, much anymore because of that.
>> Paul Curtis, I believe, suggested that this may be more a
>> problem with TI-supplied drivers than a matter of IAR's
>> fault. But Paul didn't say he was having the same troubles
>> with his own products. So it's not a strong argument without
>> that additional information. Just a suggestion.
>>
>> The rest doesn't surprise me, anymore.
>>
>> I _AM_ surprised that your working through tech suppport has
>> not produced results, though. Or that Anders hasn't had
>> something definite to say about it, here. The facts of that
>> do bother me.
>>
>> I hope we get to hear the fuller story as it evolves.
>>
>> Thanks,
>> Jon
>>