mspsim-developers — Discussions regarding mspsim for both developers and users

[Mspsim-developers] Source of error "void: *** IOUnit read from non-existent IO at $0040"

[Marco G. <mrk...@gm...>]

Hi all,
is there some way to understand what $0040 refers to?
"void: *** IOUnit read from non-existent IO at $0040"

It is an error from MSPsim, but I don't know where to find the source of
the error since 0040 doesn't give me any hint.

Thanks in advance.
MG

Re: [Mspsim-developers] void: *** IOUnit read from non-existent IO

[Ben R. <ran...@cs...>]

Hi Marco,

I don't recommend using the "ransford/mspsim" fork of MSPsim for Contiki
development, nor will MSPsim's developers be able to troubleshoot problems
with that fork.  Can you reproduce the problem against mainline MSPsim from
Sourceforge?

-ben


On Fri, Sep 13, 2013 at 3:17 AM, Marco Giordano <mrk...@gm...> wrote:

> I'm trying to investigate the source of the problem.
>
> It seems that the watchdog timer for some reason is not reset.
> From what I understand, the function "scheduleTimer" in Watchdog.java sets
> the Watchdog timer at some time in the future, and it is periodically
> called by the function "write" placed in the same file.
>
> However, I cannot understand where that function "write" is called. When
> the WDT expires, the first line of my error "WDT trigger - will set
> interrupt flag (no reset)" is printed and all the other errors are printed
> consequently.
>
> I would like to know why the timer is missed to be reset, even if I think
> I cannot do much to solve the problem.
>
> thanks,
> MG
>
>
> 2013/9/10 Marco Giordano <mrk...@gm...>
>
>> Hi all,
>> I'm a new subscriber to the mailing list which has a problem with MSPSim.
>>
>> I'm a Contiki user from almost an year for my thesis job.
>> To my purposes, I have added a file to the Contiki core/dev folder, which
>> serves me to simulate the behavior of a temperature and light sensor.
>> I don't know if this was the right way to simulate those kind of sensors,
>> but it worked well till a couple of months ago.
>>
>> In the last period some warning messages begin to appear. I don't know if
>> it is due to the presence of a ctimer I added a couple of month ago in
>> those files.
>>
>> In particular, in most of the simulations, sometimes appears this message:
>>
>> void: *** *IOUnit read from non-existent IO* at ...
>>
>> where after "at" sometimes there's **$0040, others $0042 or $0044.
>>
>> I don't know what's the reason, even because I access to my dev files
>> once every minute but these warning appears once every few hours, and not
>> in all my simulations. Some simulations run clear till the end, without
>> these messages.
>>
>> I have found the line printing this error in this file:
>>
>>
>> https://github.com/ransford/mspsim/blob/master/se/sics/mspsim/core/MSP430Core.java
>>
>> I'm sure that this error is due to my file added to the core/dev
>> directory, what I can not explain is why this error sometimes appears very
>> often, other times never.
>>
>> Moreover, sometimes appear messages like:
>>
>> WDT trigger - will set interrupt flag (no reset)
>> Stack Trace: number of calls: 14 PC: $b834
>>   rtimer_arch_now (../../../cpu/msp430/f1xxx/rtimer-arch.c) called from
>> PC: $a9ba (elapsed: 0)
>>   off (local in ../../../core/dev/cc2420.c) called from PC: $aa2e
>> (elapsed: 348252)
>>   RELEASE_LOCK (local in ../../../core/dev/cc2420.c) called from PC:
>> $ab10 (elapsed: 348275)
>>   cc2420_off (../../../core/dev/cc2420.c) called from PC: $8086 (elapsed:
>> 348366)
>>   off (local in ../../../core/net/mac/contikimac.c) called from PC: $814a
>> (elapsed: 348393)
>>   powercycle_turn_radio_off (local in ../../../core/net/mac/contikimac.c)
>> called from PC: $81d2 (elapsed: 348416)
>>   powercycle (local in ../../../core/net/mac/contikimac.c) called from
>> PC: $95c0 (elapsed: 349969)
>>   rtimer_run_next (../../../core/sys/rtimer.c) called from PC: $b866
>> (elapsed: 349988)
>>  *** Interrupt 6 from PC: $5622
>>   __unpack_f (../../.././gcc/fp-bit.c) called from PC: $5078 (elapsed:
>> 350109)
>>   __gtsf2 (../../.././gcc/fp-bit.c) called from PC: $a0fa (elapsed:
>> 350247)
>>   env_data (../../../core/dev/my-sensors.c) called from PC: $6130
>> (elapsed: 3114618)
>>   process_thread_collect_with_harvester_process (local in
>> .//collect-with-harvester.c) called from PC: $8af8 (elapsed: 3114759)
>>   call_process (local in ../../../core/sys/process.c) called from PC:
>> $8c7a (elapsed: 3114795)
>>   process_run (../../../core/sys/process.c) called from PC: $4250
>> (elapsed: 3904164)
>> Watchdog: illegal write to WDTCTL (0) from $b7d4 - reset!!!!
>>
>> which make one of my node rebooting. The problem is that these messages
>> are different between simulations, so it is very difficult for me to
>> understand the reason, even because I'm not expert in Contiki nor MSPSim.
>>
>> Hope you can give me some explanation of my errors.
>>
>> Thanks in advance.
>> MG
>>
>
>
>
> ------------------------------------------------------------------------------
> How ServiceNow helps IT people transform IT departments:
> 1. Consolidate legacy IT systems to a single system of record for IT
> 2. Standardize and globalize service processes across IT
> 3. Implement zero-touch automation to replace manual, redundant tasks
> http://pubads.g.doubleclick.net/gampad/clk?id=51271111&iu=/4140/ostg.clktrk
> _______________________________________________
> Mspsim-developers mailing list
> Msp...@li...
> https://lists.sourceforge.net/lists/listinfo/mspsim-developers
>
>

Re: [Mspsim-developers] void: *** IOUnit read from non-existent IO

[Marco G. <mrk...@gm...>]

I'm trying to investigate the source of the problem.

It seems that the watchdog timer for some reason is not reset.
>From what I understand, the function "scheduleTimer" in Watchdog.java sets
the Watchdog timer at some time in the future, and it is periodically
called by the function "write" placed in the same file.

However, I cannot understand where that function "write" is called. When
the WDT expires, the first line of my error "WDT trigger - will set
interrupt flag (no reset)" is printed and all the other errors are printed
consequently.

I would like to know why the timer is missed to be reset, even if I think I
cannot do much to solve the problem.

thanks,
MG


2013/9/10 Marco Giordano <mrk...@gm...>

> Hi all,
> I'm a new subscriber to the mailing list which has a problem with MSPSim.
>
> I'm a Contiki user from almost an year for my thesis job.
> To my purposes, I have added a file to the Contiki core/dev folder, which
> serves me to simulate the behavior of a temperature and light sensor.
> I don't know if this was the right way to simulate those kind of sensors,
> but it worked well till a couple of months ago.
>
> In the last period some warning messages begin to appear. I don't know if
> it is due to the presence of a ctimer I added a couple of month ago in
> those files.
>
> In particular, in most of the simulations, sometimes appears this message:
>
> void: *** *IOUnit read from non-existent IO* at ...
>
> where after "at" sometimes there's **$0040, others $0042 or $0044.
>
> I don't know what's the reason, even because I access to my dev files once
> every minute but these warning appears once every few hours, and not in all
> my simulations. Some simulations run clear till the end, without these
> messages.
>
> I have found the line printing this error in this file:
>
>
> https://github.com/ransford/mspsim/blob/master/se/sics/mspsim/core/MSP430Core.java
>
> I'm sure that this error is due to my file added to the core/dev
> directory, what I can not explain is why this error sometimes appears very
> often, other times never.
>
> Moreover, sometimes appear messages like:
>
> WDT trigger - will set interrupt flag (no reset)
> Stack Trace: number of calls: 14 PC: $b834
>   rtimer_arch_now (../../../cpu/msp430/f1xxx/rtimer-arch.c) called from
> PC: $a9ba (elapsed: 0)
>   off (local in ../../../core/dev/cc2420.c) called from PC: $aa2e
> (elapsed: 348252)
>   RELEASE_LOCK (local in ../../../core/dev/cc2420.c) called from PC: $ab10
> (elapsed: 348275)
>   cc2420_off (../../../core/dev/cc2420.c) called from PC: $8086 (elapsed:
> 348366)
>   off (local in ../../../core/net/mac/contikimac.c) called from PC: $814a
> (elapsed: 348393)
>   powercycle_turn_radio_off (local in ../../../core/net/mac/contikimac.c)
> called from PC: $81d2 (elapsed: 348416)
>   powercycle (local in ../../../core/net/mac/contikimac.c) called from PC:
> $95c0 (elapsed: 349969)
>   rtimer_run_next (../../../core/sys/rtimer.c) called from PC: $b866
> (elapsed: 349988)
>  *** Interrupt 6 from PC: $5622
>   __unpack_f (../../.././gcc/fp-bit.c) called from PC: $5078 (elapsed:
> 350109)
>   __gtsf2 (../../.././gcc/fp-bit.c) called from PC: $a0fa (elapsed: 350247)
>   env_data (../../../core/dev/my-sensors.c) called from PC: $6130
> (elapsed: 3114618)
>   process_thread_collect_with_harvester_process (local in
> .//collect-with-harvester.c) called from PC: $8af8 (elapsed: 3114759)
>   call_process (local in ../../../core/sys/process.c) called from PC:
> $8c7a (elapsed: 3114795)
>   process_run (../../../core/sys/process.c) called from PC: $4250
> (elapsed: 3904164)
> Watchdog: illegal write to WDTCTL (0) from $b7d4 - reset!!!!
>
> which make one of my node rebooting. The problem is that these messages
> are different between simulations, so it is very difficult for me to
> understand the reason, even because I'm not expert in Contiki nor MSPSim.
>
> Hope you can give me some explanation of my errors.
>
> Thanks in advance.
> MG
>

[Mspsim-developers] void: *** IOUnit read from non-existent IO

[Marco G. <mrk...@gm...>]

Hi all,
I'm a new subscriber to the mailing list which has a problem with MSPSim.

I'm a Contiki user from almost an year for my thesis job.
To my purposes, I have added a file to the Contiki core/dev folder, which
serves me to simulate the behavior of a temperature and light sensor.
I don't know if this was the right way to simulate those kind of sensors,
but it worked well till a couple of months ago.

In the last period some warning messages begin to appear. I don't know if
it is due to the presence of a ctimer I added a couple of month ago in
those files.

In particular, in most of the simulations, sometimes appears this message:

void: *** *IOUnit read from non-existent IO* at ...

where after "at" sometimes there's **$0040, others $0042 or $0044.

I don't know what's the reason, even because I access to my dev files once
every minute but these warning appears once every few hours, and not in all
my simulations. Some simulations run clear till the end, without these
messages.

I have found the line printing this error in this file:

https://github.com/ransford/mspsim/blob/master/se/sics/mspsim/core/MSP430Core.java

I'm sure that this error is due to my file added to the core/dev directory,
what I can not explain is why this error sometimes appears very often,
other times never.

Moreover, sometimes appear messages like:

WDT trigger - will set interrupt flag (no reset)
Stack Trace: number of calls: 14 PC: $b834
  rtimer_arch_now (../../../cpu/msp430/f1xxx/rtimer-arch.c) called from PC:
$a9ba (elapsed: 0)
  off (local in ../../../core/dev/cc2420.c) called from PC: $aa2e (elapsed:
348252)
  RELEASE_LOCK (local in ../../../core/dev/cc2420.c) called from PC: $ab10
(elapsed: 348275)
  cc2420_off (../../../core/dev/cc2420.c) called from PC: $8086 (elapsed:
348366)
  off (local in ../../../core/net/mac/contikimac.c) called from PC: $814a
(elapsed: 348393)
  powercycle_turn_radio_off (local in ../../../core/net/mac/contikimac.c)
called from PC: $81d2 (elapsed: 348416)
  powercycle (local in ../../../core/net/mac/contikimac.c) called from PC:
$95c0 (elapsed: 349969)
  rtimer_run_next (../../../core/sys/rtimer.c) called from PC: $b866
(elapsed: 349988)
 *** Interrupt 6 from PC: $5622
  __unpack_f (../../.././gcc/fp-bit.c) called from PC: $5078 (elapsed:
350109)
  __gtsf2 (../../.././gcc/fp-bit.c) called from PC: $a0fa (elapsed: 350247)
  env_data (../../../core/dev/my-sensors.c) called from PC: $6130 (elapsed:
3114618)
  process_thread_collect_with_harvester_process (local in
.//collect-with-harvester.c) called from PC: $8af8 (elapsed: 3114759)
  call_process (local in ../../../core/sys/process.c) called from PC: $8c7a
(elapsed: 3114795)
  process_run (../../../core/sys/process.c) called from PC: $4250 (elapsed:
3904164)
Watchdog: illegal write to WDTCTL (0) from $b7d4 - reset!!!!

which make one of my node rebooting. The problem is that these messages are
different between simulations, so it is very difficult for me to understand
the reason, even because I'm not expert in Contiki nor MSPSim.

Hope you can give me some explanation of my errors.

Thanks in advance.
MG

[Mspsim-developers] [PATCH] Fix for CALLA_IND

[Daniele A. <dan...@gm...>]

Dear MSPsim developers,

I found a bug in the current git version of MSPsim: the PC is
incremented too much when emulating CALLA instructions with indirect
addressing mode. The fix is very simple: just remove the line
increasing the PC in the CALLA_IND case. The patch is attached.

Thank you for your hard work on MSPsim and merry Christmas :)

Regards.

-Daniele

Re: [Mspsim-developers] Move to Github?

[Joakim E. <jo...@si...>]

Hello!

Not for the moment - but I guess if it is a good idea
to move it will move sooner or later!

Are you voting on moving?

Best regards,
-- Joakim

Moritz Struebe skrev 2012-10-31 16:59:
> Hi there,
>
> as Cooja moved to GitHub, are there plans to move MSPsim there too?
>
> Cheers
> Morty
>
>
>
> ------------------------------------------------------------------------------
> Everyone hates slow websites. So do we.
> Make your web apps faster with AppDynamics
> Download AppDynamics Lite for free today:
> http://p.sf.net/sfu/appdyn_sfd2d_oct
>
>
>
> _______________________________________________
> Mspsim-developers mailing list
> Msp...@li...
> https://lists.sourceforge.net/lists/listinfo/mspsim-developers
>

[Mspsim-developers] Move to Github?

[Moritz S. <Mor...@in...>]

Hi there,

as Cooja moved to GitHub, are there plans to move MSPsim there too?

Cheers
Morty

-- 
Dipl.-Ing. Moritz 'Morty' Struebe (Wissenschaftlicher Mitarbeiter)
Lehrstuhl für Informatik 4 (Verteilte Systeme und Betriebssysteme)
Friedrich-Alexander-Universität Erlangen-Nürnberg
Martensstr. 1
91058 Erlangen

Tel   : +49 9131 85-25419
Fax   : +49 9131 85-28732
eMail : st...@in...
WWW   : http://www4.informatik.uni-erlangen.de/~morty

[Mspsim-developers] Patch: fixed bug causing incorrect SR update with RRxx instructions

[Daniele A. <dan...@gm...>]

Hi MSPSim developers,

I find out that, when executing MSP430X bitshift instructions
(RRCM/RRAM/RLAM/RRUM), MSPSim does not correctly update the Status
Register. The attached patch fixes the bug.

Regards,

-Daniele

Re: [Mspsim-developers] MSP430X and memory space > 64 kB

[Moritz S. <Mor...@in...>]

On 2012-09-14 16:42, Daniele Alessandrelli wrote:
> void: *** IOUnit read from non-existent IO at $0000
> void: *** IOUnit write to non-existent IO at $0000

This is an access to  a null-pointer.

>
> PS: Perhaps it's my CALLA implementations which are wrong, I can post
> them if you want.

This is always a good idea! ;) I also suggest to write a _very small_
test program. Somthing in the line of
volatile char t [65*1024]
int main(){
for(int i = 1; i < 65*1024; i++) t[i] = 1;
}
to check for errors in mspsim. Don't forget to check the disassambler to
check whether the compiler did it's job right and run that in mspsim.
After all it's not unlikely there's a bug in the compiler.

Cheers
Morty

P.s: I have no idea of the MSP430X platform, this is just some advice
based on my experience with similar problems.

-- 
Dipl.-Ing. Moritz 'Morty' Struebe (Wissenschaftlicher Mitarbeiter)
Lehrstuhl für Informatik 4 (Verteilte Systeme und Betriebssysteme)
Friedrich-Alexander-Universität Erlangen-Nürnberg
Martensstr. 1
91058 Erlangen

Tel   : +49 9131 85-25419
Fax   : +49 9131 85-28732
eMail : st...@in...
WWW   : http://www4.informatik.uni-erlangen.de/~morty

[Mspsim-developers] MSP430X and memory space > 64 kB

[Daniele A. <dan...@gm...>]

Hi everybody,

First of all, did any of you successfully run on MSPsim a firmware
having an address space larger than 64 kB?

I'm asking because I compiled some code (specifically the
"rpl-border-router" Contiki example for Wismote) exceeding the 64 kB
boundary and I cannot simulate it.  I get an exception saying "CALLA:
mode not implemented 134c => 1340".

I implemented the needed addressing modes for CALLA (specifically the
CALLA using register mode and indexed mode), but I get the following
errors which I'm not smart enough to debug ;) Any hint?

void: *** IOUnit read from non-existent IO at $0000
void: *** IOUnit write to non-existent IO at $0000
java.lang.ArrayIndexOutOfBoundsException: -24
	at se.sics.mspsim.core.MSP430Core$1.write(MSP430Core.java:161)
	at se.sics.mspsim.core.MSP430Core.emulateOP(MSP430Core.java:1121)
	at se.sics.mspsim.core.MSP430.run(MSP430.java:109)
	at se.sics.mspsim.core.MSP430.cpuloop(MSP430.java:95)
	at se.sics.mspsim.platform.GenericNode.run(GenericNode.java:263)
	at java.lang.Thread.run(Thread.java:679)

Thanks!

Daniele

PS: Perhaps it's my CALLA implementations which are wrong, I can post
them if you want.

Re: [Mspsim-developers] [PATCH] Add WarnStream to Loggable.

[Moritz S. <Mor...@in...>]

Patches are now well tested and pushed to my git:
http://i4git.cs.fau.de/mspsim.git/tree/patches

Cheers
Morty

On 2012-07-17 18:58, Moritz Struebe wrote:
> Damn, I should go home! Here's the patch!
>
> Cheers
> Morty
>
> On 2012-07-17 18:58, Moritz Struebe wrote:
>> > Ok,
>> >
>> > so here is the fixed version. It allows to add a WarnStream. This stream
>> > will get the warnings, that were previously printed to System.err.
>> >
>> > As a side note: Is there a good reason for the "DEBUG" variable? I'd
>> > expect the JIT to optimitze that call. At least I don't expect that it
>> > has enough impact to justify duplicating the logic whether to debug or not.
>> >
>> > Cheers
>> > Morty


-- 
Dipl.-Ing. Moritz 'Morty' Struebe (Wissenschaftlicher Mitarbeiter)
Lehrstuhl für Informatik 4 (Verteilte Systeme und Betriebssysteme)
Friedrich-Alexander-Universität Erlangen-Nürnberg
Martensstr. 1
91058 Erlangen

Tel   : +49 9131 85-25419
Fax   : +49 9131 85-28732
eMail : st...@in...
WWW   : http://www4.informatik.uni-erlangen.de/~morty

Re: [Mspsim-developers] [PATCH] Add WarnStream to Loggable.

[Moritz S. <Mor...@in...>]

Damn, I should go home! Here's the patch!

Cheers
Morty

On 2012-07-17 18:58, Moritz Struebe wrote:
> Ok,
>
> so here is the fixed version. It allows to add a WarnStream. This stream
> will get the warnings, that were previously printed to System.err.
>
> As a side note: Is there a good reason for the "DEBUG" variable? I'd
> expect the JIT to optimitze that call. At least I don't expect that it
> has enough impact to justify duplicating the logic whether to debug or not.
>
> Cheers
> Morty


-- 
Dipl.-Ing. Moritz 'Morty' Struebe (Wissenschaftlicher Mitarbeiter)
Lehrstuhl für Informatik 4 (Verteilte Systeme und Betriebssysteme)
Friedrich-Alexander-Universität Erlangen-Nürnberg
Martensstr. 1
91058 Erlangen

Tel   : +49 9131 85-25419
Fax   : +49 9131 85-28732
eMail : st...@in...
WWW   : http://www4.informatik.uni-erlangen.de/~morty

[Mspsim-developers] [PATCH] Add WarnStream to Loggable.

[Moritz S. <Mor...@in...>]

Ok,

so here is the fixed version. It allows to add a WarnStream. This stream
will get the warnings, that were previously printed to System.err.

As a side note: Is there a good reason for the "DEBUG" variable? I'd
expect the JIT to optimitze that call. At least I don't expect that it
has enough impact to justify duplicating the logic whether to debug or not.

Cheers
Morty

On 2012-07-17 16:49, Moritz Struebe wrote:
> Here's the follow up patch, that avoids writing to System.err if a
> LogStream is registered for an IOUnit.
>
> Cheers
> Morty


-- 
Dipl.-Ing. Moritz 'Morty' Struebe (Wissenschaftlicher Mitarbeiter)
Lehrstuhl für Informatik 4 (Verteilte Systeme und Betriebssysteme)
Friedrich-Alexander-Universität Erlangen-Nürnberg
Martensstr. 1
91058 Erlangen

Tel   : +49 9131 85-25419
Fax   : +49 9131 85-28732
eMail : st...@in...
WWW   : http://www4.informatik.uni-erlangen.de/~morty

Re: [Mspsim-developers] [PATCH] Only write to System.err, if no LogStream is registered.

[Moritz S. <Mor...@in...>]

Ok, forget this one - a bigger patch is needed to fix this....

Cheers
Morty

On 2012-07-17 16:49, Moritz Struebe wrote:
> Here's the follow up patch, that avoids writing to System.err if a
> LogStream is registered for an IOUnit.
>
> Cheers
> Morty


-- 
Dipl.-Ing. Moritz 'Morty' Struebe (Wissenschaftlicher Mitarbeiter)
Lehrstuhl für Informatik 4 (Verteilte Systeme und Betriebssysteme)
Friedrich-Alexander-Universität Erlangen-Nürnberg
Martensstr. 1
91058 Erlangen

Tel   : +49 9131 85-25419
Fax   : +49 9131 85-28732
eMail : st...@in...
WWW   : http://www4.informatik.uni-erlangen.de/~morty

[Mspsim-developers] [PATCH] Only write to System.err, if no LogStream is registered.

[Moritz S. <Mor...@in...>]

Here's the follow up patch, that avoids writing to System.err if a
LogStream is registered for an IOUnit.

Cheers
Morty

On 2012-07-17 16:38, Moritz Struebe wrote:
> Hi,
>
> this tiny patch adds voidIO to the available IO-Units. While one
> normally does not want to access that directly, one might want to set
> the LogStream.
> The Cooja-Patch, that adds the mote-information to the " *** IOUnit
> write to non-existent IO at ..." warning is already working and will be
> submitted as soon as patches are accepted, again.
>
> Cheers
> Morty


-- 
Dipl.-Ing. Moritz 'Morty' Struebe (Wissenschaftlicher Mitarbeiter)
Lehrstuhl für Informatik 4 (Verteilte Systeme und Betriebssysteme)
Friedrich-Alexander-Universität Erlangen-Nürnberg
Martensstr. 1
91058 Erlangen

Tel   : +49 9131 85-25419
Fax   : +49 9131 85-28732
eMail : st...@in...
WWW   : http://www4.informatik.uni-erlangen.de/~morty

[Mspsim-developers] [PATCH] Access to "voidIO"

[Moritz S. <Mor...@in...>]

Hi,

this tiny patch adds voidIO to the available IO-Units. While one
normally does not want to access that directly, one might want to set
the LogStream.
The Cooja-Patch, that adds the mote-information to the " *** IOUnit
write to non-existent IO at ..." warning is already working and will be
submitted as soon as patches are accepted, again.

Cheers
Morty


-- 
Dipl.-Ing. Moritz 'Morty' Struebe (Wissenschaftlicher Mitarbeiter)
Lehrstuhl für Informatik 4 (Verteilte Systeme und Betriebssysteme)
Friedrich-Alexander-Universität Erlangen-Nürnberg
Martensstr. 1
91058 Erlangen

Tel   : +49 9131 85-25419
Fax   : +49 9131 85-28732
eMail : st...@in...
WWW   : http://www4.informatik.uni-erlangen.de/~morty

Re: [Mspsim-developers] gdbstubs crashes gdb

[Moritz S. <Mor...@in...>]

I rewrote that stub for COOJA [1] and fixed some stuff. But it's more
likely that gdb is buggy, as it shouldn't crash, even if the stub
returns bogus data. I use the gcc 4.4.3 toochain in combination with
7.0.1 gdb for debugging, which works fine. (Still using 3.2.3 for
compilation.)

Cheers
Morty




[1]
http://contikiprojects.svn.sourceforge.net/viewvc/contikiprojects/uni-erlangen.de/COOJA_GDBstub/

On 2012-06-18 17:04, Alexander Bernauer wrote:
> using the current head (commit df95ab4) of MSPSim I get the following
> segmentation fault for msp430-gdb (Ubuntu 7.2~mspgcc-7.2-20110612-1ubuntu1):
>
> ---8<--- Console 1
> $ java -jar mspsim.jar /home/alex/scm/contiki/examples/hello-world/hello-world.sky


-- 
Dipl.-Ing. Moritz 'Morty' Struebe (Wissenschaftlicher Mitarbeiter)
Lehrstuhl für Informatik 4 (Verteilte Systeme und Betriebssysteme)
Friedrich-Alexander-Universität Erlangen-Nürnberg
Martensstr. 1
91058 Erlangen

Tel   : +49 9131 85-25419
Fax   : +49 9131 85-28732
eMail : st...@in...
WWW   : http://www4.informatik.uni-erlangen.de/~morty

Re: [Mspsim-developers] Reading automatic variables

[Moritz S. <Mor...@in...>]

On 2012-06-18 15:49, Alexander Bernauer wrote:
> As far as I can see, DWARF's Call Frame Information table contains all
> necessary information to find the address of an automatic variable at
> runtime.
>
> If necessary, I am willing to contribute the necessary work to add this
> feature to MSPSim. I really need it for my project [1].

Yep, that's what you will have to do. ;)

Cheers
Morty


-- 
Dipl.-Ing. Moritz 'Morty' Struebe (Wissenschaftlicher Mitarbeiter)
Lehrstuhl für Informatik 4 (Verteilte Systeme und Betriebssysteme)
Friedrich-Alexander-Universität Erlangen-Nürnberg
Martensstr. 1
91058 Erlangen

Tel   : +49 9131 85-25419
Fax   : +49 9131 85-28732
eMail : st...@in...
WWW   : http://www4.informatik.uni-erlangen.de/~morty

[Mspsim-developers] gdbstubs crashes gdb

[Alexander B. <ber...@in...>]

Hi

using the current head (commit df95ab4) of MSPSim I get the following
segmentation fault for msp430-gdb (Ubuntu 7.2~mspgcc-7.2-20110612-1ubuntu1):

---8<--- Console 1
$ java -jar mspsim.jar /home/alex/scm/contiki/examples/hello-world/hello-world.sky
[...]
MSPSim>gdbstubs 8877
GDBStubs open server socket port: 8877
[proceed with Console 2]
[debug output - see below]
--->8---

---8<--- Console 2
$ msp430-gdb
[...]
(gdb) file /home/alex/scm/contiki/examples/hello-world/hello-world.sky
Reading symbols from /home/alex/scm/contiki/examples/hello-world/hello-world.sky...done.
(gdb) target remote localhost:8877
Remote debugging using localhost:8877
zsh: segmentation fault  msp430-gdb
--->8---

Is this working for anybody? Which versions of the tools to I have to
use?

The binary has been compiled with msp430-gcc (Ubuntu 4.5.3~mspgcc-4.5.2-20110612-1)
and the example binary is from Contiki's current head (commit 51b7312).

Greetings

Alex

---8<--- debug output
MSPSim>GDBStubs: Read  43 => +
GDBStubs: Read  36 => $
GDBStubs: Read  113 => q
GDBStubs: Read  83 => S
GDBStubs: Read  117 => u
GDBStubs: Read  112 => p
GDBStubs: Read  112 => p
GDBStubs: Read  111 => o
GDBStubs: Read  114 => r
GDBStubs: Read  116 => t
GDBStubs: Read  101 => e
GDBStubs: Read  100 => d
GDBStubs: Read  58 => :
GDBStubs: Read  113 => q
GDBStubs: Read  82 => R
GDBStubs: Read  101 => e
GDBStubs: Read  108 => l
GDBStubs: Read  111 => o
GDBStubs: Read  99 => c
GDBStubs: Read  73 => I
GDBStubs: Read  110 => n
GDBStubs: Read  115 => s
GDBStubs: Read  110 => n
GDBStubs: Read  43 => +
GDBStubs: Read  35 => #
cmd: qSupported:qRelocInsn+
Command unknown
#00
GDBStubs: Read  57 => 9
GDBStubs: Read  97 => a
GDBStubs: Read  43 => +
GDBStubs: Read  36 => $
GDBStubs: Read  72 => H
GDBStubs: Read  103 => g
GDBStubs: Read  48 => 0
GDBStubs: Read  35 => #
cmd: Hg0
OK#9a
GDBStubs: Read  100 => d
GDBStubs: Read  102 => f
GDBStubs: Read  43 => +
GDBStubs: Read  36 => $
GDBStubs: Read  63 => ?
GDBStubs: Read  35 => #
cmd: ?
S01#b4
GDBStubs: Read  51 => 3
GDBStubs: Read  102 => f
GDBStubs: Read  43 => +
GDBStubs: Read  36 => $
GDBStubs: Read  72 => H
GDBStubs: Read  99 => c
GDBStubs: Read  45 => -
GDBStubs: Read  49 => 1
GDBStubs: Read  35 => #
cmd: Hc-1
OK#9a
GDBStubs: Read  48 => 0
GDBStubs: Read  57 => 9
GDBStubs: Read  43 => +
GDBStubs: Read  36 => $
GDBStubs: Read  113 => q
GDBStubs: Read  67 => C
GDBStubs: Read  35 => #
cmd: qC
QC1#c5
GDBStubs: Read  98 => b
GDBStubs: Read  52 => 4
GDBStubs: Read  43 => +
GDBStubs: Read  36 => $
GDBStubs: Read  113 => q
GDBStubs: Read  65 => A
GDBStubs: Read  116 => t
GDBStubs: Read  116 => t
GDBStubs: Read  97 => a
GDBStubs: Read  99 => c
GDBStubs: Read  104 => h
GDBStubs: Read  101 => e
GDBStubs: Read  100 => d
GDBStubs: Read  35 => #
cmd: qAttached
Command unknown
#00
GDBStubs: Read  56 => 8
GDBStubs: Read  102 => f
GDBStubs: Read  43 => +
GDBStubs: Read  36 => $
GDBStubs: Read  113 => q
GDBStubs: Read  79 => O
GDBStubs: Read  102 => f
GDBStubs: Read  102 => f
GDBStubs: Read  115 => s
GDBStubs: Read  101 => e
GDBStubs: Read  116 => t
GDBStubs: Read  115 => s
GDBStubs: Read  35 => #
cmd: qOffsets
Text=0;Data=0;Bss=0#04
GDBStubs: Read  52 => 4
GDBStubs: Read  98 => b
GDBStubs: Read  43 => +
GDBStubs: Read  36 => $
GDBStubs: Read  103 => g
GDBStubs: Read  35 => #
cmd: g
0000000000000000000000000000000000000000000000000000000000000000#00
GDBStubs: Read  54 => 6
GDBStubs: Read  55 => 7
GDBStubs: Read  43 => +
--->8---

[Mspsim-developers] Reading automatic variables

[Alexander B. <ber...@in...>]

Hi

as far as I can see, the current implementation of MSPSim has no support
to read the value of automatic (i.e. function local) variables.

The MapTable returned by ELF.getMap() seems to contain only global and
function static variables. Also, MSPSim's "symbol" command only lists
those.

Does anybody know how I can read the value of automatic variables from
mspsim?

As far as I can see, DWARF's Call Frame Information table contains all
necessary information to find the address of an automatic variable at
runtime.

If necessary, I am willing to contribute the necessary work to add this
feature to MSPSim. I really need it for my project [1].

Greetings

Alex

[1] http://github.com/copton/ocram

Re: [Mspsim-developers] Using the 32-bit multiplier

[Niclas F. <nf...@si...>]

Hi,

Thanks for this patch! It is a great start towards complete emulation of 
the hardware multiplier32. I have pushed it to the MSPSim git master 
repository.

Best regards,
-- Niclas Finne


On 2012-05-04 16:10, Niels Möller wrote:
> ni...@ly... (Niels Möller) writes:
>
>> 2. Looking at the file core/Multiplier32.java, it seems that the
>>     implementation is not complete.
>
> The attached patch makes the implementation a little less incomplete.
> I've tested a 32x32 unsigned multiply, and it seems to work, using the
> tyndall.TyndallNode platform (test program also attached).
>
> Most obvious omissions: I didn't touch the code handling writes to OP2,
> so 32 x 16 won't work. And I don't set any of the SUMEXT and MPYC
> outputs.
>
> This is an afternoon hack, and I'm not at all used to java programming.
> Fully implementing the various modes of operation (saturation,
> fractional, etc) seems like it would be several days work, and then I
> have no idea about the DMA possibilities and the like.
>
> I'm also not entirely sure how it is suppose to operate. The
> specification I have ("MSP430x5xx/MSP430x6xx Family User's Guide",
> SLAU208J) says that writing OP2L "starts" the multiplication, and
> writing OP2H "continues" the multiplication. Are any partial results
> from this visible to the programmer? In particular, if you write OP2L,
> will the low 16 bits of the product be available in RES0, before you
> write to OP2H?
>
> Regards,
> /Niels
>
>
>
>
>
>
>
> ------------------------------------------------------------------------------
> Live Security Virtual Conference
> Exclusive live event will cover all the ways today's security and
> threat landscape has changed and how IT managers can respond. Discussions
> will include endpoint security, mobile security and the latest in malware
> threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/
>
>
>
> _______________________________________________
> Mspsim-developers mailing list
> Msp...@li...
> https://lists.sourceforge.net/lists/listinfo/mspsim-developers

Re: [Mspsim-developers] Using the 32-bit multiplier

[<ni...@ly...>]

ni...@ly... (Niels Möller) writes:

> 2. Looking at the file core/Multiplier32.java, it seems that the
>    implementation is not complete.

The attached patch makes the implementation a little less incomplete.
I've tested a 32x32 unsigned multiply, and it seems to work, using the
tyndall.TyndallNode platform (test program also attached).

Most obvious omissions: I didn't touch the code handling writes to OP2,
so 32 x 16 won't work. And I don't set any of the SUMEXT and MPYC
outputs.

This is an afternoon hack, and I'm not at all used to java programming.
Fully implementing the various modes of operation (saturation,
fractional, etc) seems like it would be several days work, and then I
have no idea about the DMA possibilities and the like.

I'm also not entirely sure how it is suppose to operate. The
specification I have ("MSP430x5xx/MSP430x6xx Family User's Guide",
SLAU208J) says that writing OP2L "starts" the multiplication, and
writing OP2H "continues" the multiplication. Are any partial results
from this visible to the programmer? In particular, if you write OP2L,
will the low 16 bits of the product be available in RES0, before you
write to OP2H?

Regards,
/Niels

Re: [Mspsim-developers] Using the 32-bit multiplier

[<ni...@ly...>]

Hello,

I'm trying to develop and test code for MSP430F5510, using msp430-gcc
and mspsim. I'm doing some crypto code which needs bignum operations, so
using the hardware multiplier is important.

I have some working code using the 16x16 bit multiplier. I then compile
for msp430x1611, and I can run the resulting formware in the simulator
using se.sics.mspsim.platform.sky.SkyNode. Seems to work fine, and I
really appreciate the builtin debugging and profiling support of mspsim
(I have also tried remote debugging with gdb, which would be even
better, but I couldn't get that to work reliably).

Now, I'd like to be able to use the 32-bit multiplier. But there are
some obstacles:

1. I'm not sure which processor to build for and which platform to use.
   The closest I find in the simulator is MSP430F5437 (I'm new to this
   architecture, and I have no good idea on how it differs from the
   ...5510). In mspsim, the platform tyndal.TyndalNode seem to use this
   processor. Is this a good choice?

2. Looking at the file core/Multiplier32.java, it seems that the
   implementation is not complete.

3. The serial hardware seem to be different (different hardware
   registers, maybe more fundamental differences too?) than on
   msp430x1611. To get data out on the serial port, I use the file
   tests/msp430setup.c which among other things implements putchar. But
   this code doesn't work if I compile for msp430f5437 or msp430f5510, I
   get compilation errors when accessing various hardware registers.
   
Any help or suggestions on how to make progress is appreciated. My
current impression is that completing Multiplier32.java should not be a
big deal. I have a lot less clue on how to access the serial port, for
putchar.

Regards,
/Niels Möller

PS. I sent this message a week ago, without first subscribing to the
list. I got a "Your message to Mspsim-developers awaits moderator
approval" back and no other response. If noone wants to do that
moderation (which I fully understand), it would be nice to change the
autoreply to say that messages from non-member are rejected, not
moderated.

Now I'm subscribed to the list, so I hope things should work better.

-- 
Niels Möller. PGP-encrypted email is preferred. Keyid C0B98E26.
Internet email is subject to wholesale government surveillance.

[Mspsim-developers] TimerA Trying to write to non-existent CCTL register: 3

[BILAL M. <Mus...@im...>]

Hello,

  I am new to cooja and I am trying to run the example: 
http://www.sics.se/contiki/wiki/index.php/Setting_up_a_Low-Power_IPv6 
<http://www.sics.se/contiki/wiki/index.php/Setting_up_a_Low-Power_IPv6/RPL_Network> 
and I am getting the following:

  [java] *** Setting up f1611 IO!
      [java]  INFO [Thread-1] (MspMote.java:224) - Loading firmware 
from: /home/bilalmus/contiki/examples/ipv6/rpl-collect/udp-sink.sky
      [java] Set up MSP430 Core with 65536 bytes memory
      [java] *** Setting up f1611 IO!
      [java]  INFO [Thread-1] (MspMote.java:224) - Loading firmware 
from: /home/bilalmus/contiki/examples/ipv6/rpl-collect/udp-sender.sky
      [java] Set up MSP430 Core with 65536 bytes memory
      [java] *** Setting up f1611 IO!
      [java]  INFO [Thread-1] (MspMote.java:224) - Loading firmware 
from: /home/bilalmus/contiki/examples/ipv6/rpl-collect/udp-sender.sky
      [java] Set up MSP430 Core with 65536 bytes memory
      [java] *** Setting up f1611 IO!
      [java]  INFO [Thread-1] (MspMote.java:224) - Loading firmware 
from: /home/bilalmus/contiki/examples/ipv6/rpl-collect/udp-sender.sky
      [java] Set up MSP430 Core with 65536 bytes memory
      [java] *** Setting up f1611 IO!
      [java]  INFO [Thread-1] (MspMote.java:224) - Loading firmware 
from: /home/bilalmus/contiki/examples/ipv6/rpl-collect/udp-sender.sky
      [java] Set up MSP430 Core with 65536 bytes memory
      [java] *** Setting up f1611 IO!
      [java]  INFO [Thread-1] (MspMote.java:224) - Loading firmware 
from: /home/bilalmus/contiki/examples/ipv6/rpl-collect/udp-sender.sky
      [java] Set up MSP430 Core with 65536 bytes memory
      [java] *** Setting up f1611 IO!
      [java]  INFO [Thread-1] (MspMote.java:224) - Loading firmware 
from: /home/bilalmus/contiki/examples/ipv6/rpl-collect/udp-sender.sky
      [java] Set up MSP430 Core with 65536 bytes memory
      [java] *** Setting up f1611 IO!
      [java]  INFO [Thread-1] (MspMote.java:224) - Loading firmware 
from: /home/bilalmus/contiki/examples/ipv6/rpl-collect/udp-sender.sky
      [java] Set up MSP430 Core with 65536 bytes memory
      [java] *** Setting up f1611 IO!
      [java]  INFO [Thread-1] (MspMote.java:224) - Loading firmware 
from: /home/bilalmus/contiki/examples/ipv6/rpl-collect/udp-sender.sky
      [java] Set up MSP430 Core with 65536 bytes memory
      [java] *** Setting up f1611 IO!
      [java]  INFO [Thread-1] (MspMote.java:224) - Loading firmware 
from: /home/bilalmus/contiki/examples/ipv6/rpl-collect/udp-sender.sky
      [java] Set up MSP430 Core with 65536 bytes memory
      [java] *** Setting up f1611 IO!
      [java]  INFO [Thread-1] (MspMote.java:224) - Loading firmware 
from: /home/bilalmus/contiki/examples/ipv6/rpl-collect/udp-sender.sky
      [java] Set up MSP430 Core with 65536 bytes memory
      [java] *** Setting up f1611 IO!
      [java]  INFO [Thread-1] (MspMote.java:224) - Loading firmware 
from: /home/bilalmus/contiki/examples/ipv6/rpl-collect/udp-sender.sky
      [java] Set up MSP430 Core with 65536 bytes memory
      [java] *** Setting up f1611 IO!
      [java]  INFO [Thread-1] (MspMote.java:224) - Loading firmware 
from: /home/bilalmus/contiki/examples/ipv6/rpl-collect/udp-sender.sky
      [java] Set up MSP430 Core with 65536 bytes memory
      [java] *** Setting up f1611 IO!
      [java]  INFO [Thread-1] (MspMote.java:224) - Loading firmware 
from: /home/bilalmus/contiki/examples/ipv6/rpl-collect/udp-sender.sky
      [java] Set up MSP430 Core with 65536 bytes memory
      [java] *** Setting up f1611 IO!
      [java]  INFO [Thread-1] (MspMote.java:224) - Loading firmware 
from: /home/bilalmus/contiki/examples/ipv6/rpl-collect/udp-sender.sky
      [java] Set up MSP430 Core with 65536 bytes memory
      [java] *** Setting up f1611 IO!
      [java]  INFO [Thread-1] (MspMote.java:224) - Loading firmware 
from: /home/bilalmus/contiki/examples/ipv6/rpl-collect/udp-sender.sky
      [java] Set up MSP430 Core with 65536 bytes memory
      [java] *** Setting up f1611 IO!
      [java]  INFO [Thread-1] (MspMote.java:224) - Loading firmware 
from: /home/bilalmus/contiki/examples/ipv6/rpl-collect/udp-sender.sky
      [java] Set up MSP430 Core with 65536 bytes memory
      [java] *** Setting up f1611 IO!
      [java]  INFO [Thread-1] (MspMote.java:224) - Loading firmware 
from: /home/bilalmus/contiki/examples/ipv6/rpl-collect/udp-sender.sky
      [java] Set up MSP430 Core with 65536 bytes memory
      [java] *** Setting up f1611 IO!
      [java]  INFO [Thread-1] (MspMote.java:224) - Loading firmware 
from: /home/bilalmus/contiki/examples/ipv6/rpl-collect/udp-sender.sky
      [java] Set up MSP430 Core with 65536 bytes memory
      [java] *** Setting up f1611 IO!
      [java]  INFO [Thread-1] (MspMote.java:224) - Loading firmware 
from: /home/bilalmus/contiki/examples/ipv6/rpl-collect/udp-sender.sky
      [java] Set up MSP430 Core with 65536 bytes memory
      [java] *** Setting up f1611 IO!
      [java]  INFO [Thread-1] (MspMote.java:224) - Loading firmware 
from: /home/bilalmus/contiki/examples/ipv6/rpl-collect/udp-sender.sky
      [java] Set up MSP430 Core with 65536 bytes memory
      [java] *** Setting up f1611 IO!
      [java]  INFO [Thread-1] (MspMote.java:224) - Loading firmware 
from: /home/bilalmus/contiki/examples/ipv6/rpl-collect/udp-sender.sky
      [java] Set up MSP430 Core with 65536 bytes memory
      [java] *** Setting up f1611 IO!
      [java]  INFO [Thread-1] (MspMote.java:224) - Loading firmware 
from: /home/bilalmus/contiki/examples/ipv6/rpl-collect/udp-sender.sky
      [java] Set up MSP430 Core with 65536 bytes memory
      [java] *** Setting up f1611 IO!
      [java]  INFO [Thread-1] (MspMote.java:224) - Loading firmware 
from: /home/bilalmus/contiki/examples/ipv6/rpl-collect/udp-sender.sky
      [java] Set up MSP430 Core with 65536 bytes memory
      [java] *** Setting up f1611 IO!
      [java]  INFO [Thread-1] (MspMote.java:224) - Loading firmware 
from: /home/bilalmus/contiki/examples/ipv6/rpl-collect/udp-sender.sky
      [java]  INFO [Thread-2] (Simulation.java:248) - Simulation main 
loop started, system time: 1334847602651
      [java] FATAL [Thread-2] (Simulation.java:287) - Simulation stopped 
due to error: null
      [java] se.sics.cooja.ContikiError
      [java]     at se.sics.cooja.mspmote.MspMote.execute(MspMote.java:317)
      [java]     at se.sics.cooja.mspmote.MspMote.execute(MspMote.java:284)
      [java]     at 
se.sics.cooja.motes.AbstractWakeupMote$1.execute(AbstractWakeupMote.java:48)
      [java]     at se.sics.cooja.Simulation.run(Simulation.java:275)
      [java]     at java.lang.Thread.run(Thread.java:679)
      [java] Caused by: se.sics.mspsim.core.EmulationException: TimerA 
Trying to write to non-existent CCTL register: 3
      [java]     at se.sics.mspsim.core.Timer.updateTCCTL(Timer.java:562)
      [java]     at se.sics.mspsim.core.Timer.read(Timer.java:528)
      [java]     at se.sics.mspsim.core.MSP430Core.read(MSP430Core.java:760)
      [java]     at 
se.sics.mspsim.core.MSP430Core.emulateOP(MSP430Core.java:1011)
      [java]     at se.sics.mspsim.core.MSP430.stepMicros(MSP430.java:256)
      [java]     at se.sics.cooja.mspmote.MspMote.execute(MspMote.java:311)
      [java]     ... 4 more
      [java] MSP430: *** IOUnit read from non-existent IO at $0090
      [java] MSP430: *** IOUnit read from non-existent IO at $0092
      [java] MSP430: *** IOUnit read from non-existent IO at $0094
      [java] MSP430: *** IOUnit read from non-existent IO at $0096
      [java] MSP430: *** IOUnit read from non-existent IO at $0098
      [java] MSP430: *** IOUnit read from non-existent IO at $009a
      [java] MSP430: *** IOUnit read from non-existent IO at $009c
      [java] MSP430: *** IOUnit read from non-existent IO at $009e
      [java] MSP430: *** IOUnit read from non-existent IO at $00a0
      [java] MSP430: *** IOUnit read from non-existent IO at $00a2
      [java] MSP430: *** IOUnit read from non-existent IO at $00a4
      [java] MSP430: *** IOUnit read from non-existent IO at $00a6
      [java] MSP430: *** IOUnit read from non-existent IO at $00a8
      [java] MSP430: *** IOUnit read from non-existent IO at $00aa
      [java] MSP430: *** IOUnit read from non-existent IO at $00ac
      [java] MSP430: *** IOUnit read from non-existent IO at $00ae
      [java] MSP430: *** IOUnit read from non-existent IO at $00b0
      [java] MSP430: *** IOUnit read from non-existent IO at $00b2
      [java] MSP430: *** IOUnit read from non-existent IO at $00b4
      [java] MSP430: *** IOUnit read from non-existent IO at $00b6
      [java] MSP430: *** IOUnit read from non-existent IO at $00b8
      [java] MSP430: *** IOUnit read from non-existent IO at $00ba
      [java] MSP430: *** IOUnit read from non-existent IO at $00bc
      [java] MSP430: *** IOUnit read from non-existent IO at $00be
      [java] MSP430: *** IOUnit read from non-existent IO at $00c0
      [java] MSP430: *** IOUnit read from non-existent IO at $00c2
      [java] MSP430: *** IOUnit read from non-existent IO at $00c4
      [java] MSP430: *** IOUnit read from non-existent IO at $00c6
      [java] MSP430: *** IOUnit read from non-existent IO at $00c8
      [java] MSP430: *** IOUnit read from non-existent IO at $00ca
      [java] MSP430: *** IOUnit read from non-existent IO at $00cc
      [java] MSP430: *** IOUnit read from non-existent IO at $00ce
      [java] MSP430: *** IOUnit read from non-existent IO at $00d0
      [java] MSP430: *** IOUnit read from non-existent IO at $00d2
      [java] MSP430: *** IOUnit read from non-existent IO at $00d4
      [java] MSP430: *** IOUnit read from non-existent IO at $00d6
      [java] MSP430: *** IOUnit read from non-existent IO at $00d8
      [java] MSP430: *** IOUnit read from non-existent IO at $00da
      [java] MSP430: *** IOUnit read from non-existent IO at $00dc
      [java] MSP430: *** IOUnit read from non-existent IO at $00de
      [java] MSP430: *** IOUnit read from non-existent IO at $00e0
      [java] MSP430: *** IOUnit read from non-existent IO at $00e2
      [java] MSP430: *** IOUnit read from non-existent IO at $00e4
      [java] MSP430: *** IOUnit read from non-existent IO at $00e6
      [java] MSP430: *** IOUnit read from non-existent IO at $00e8
      [java] MSP430: *** IOUnit read from non-existent IO at $00ea
      [java] MSP430: *** IOUnit read from non-existent IO at $00ec
      [java] MSP430: *** IOUnit read from non-existent IO at $00ee
      [java] MSP430: *** IOUnit read from non-existent IO at $00f0
      [java] MSP430: *** IOUnit read from non-existent IO at $00f2
      [java] MSP430: *** IOUnit read from non-existent IO at $00f4
      [java] MSP430: *** IOUnit read from non-existent IO at $00f6
      [java] MSP430: *** IOUnit read from non-existent IO at $00f8
      [java] MSP430: *** IOUnit read from non-existent IO at $00fa
      [java] MSP430: *** IOUnit read from non-existent IO at $00fc
      [java] MSP430: *** IOUnit read from non-existent IO at $00fe
      [java] MSP430: *** IOUnit read from non-existent IO at $0100
      [java] MSP430: *** IOUnit read from non-existent IO at $0102
      [java] MSP430: *** IOUnit read from non-existent IO at $0104
      [java] MSP430: *** IOUnit read from non-existent IO at $0106
      [java] MSP430: *** IOUnit read from non-existent IO at $0108
      [java] MSP430: *** IOUnit read from non-existent IO at $010a
      [java] MSP430: *** IOUnit read from non-existent IO at $010c
      [java] MSP430: *** IOUnit read from non-existent IO at $010e
      [java] MSP430: *** IOUnit read from non-existent IO at $0110
      [java] MSP430: *** IOUnit read from non-existent IO at $0112
      [java] MSP430: *** IOUnit read from non-existent IO at $0114
      [java] MSP430: *** IOUnit read from non-existent IO at $0116
      [java] MSP430: *** IOUnit read from non-existent IO at $0118
      [java] MSP430: *** IOUnit read from non-existent IO at $011a
      [java] MSP430: *** IOUnit read from non-existent IO at $011c
      [java] MSP430: *** IOUnit read from non-existent IO at $0122
      [java] MSP430: *** IOUnit read from non-existent IO at $0126

I was having the latest version of msp, I removed it to install instead 
an older version(shown below), however I am getting always the same error:

ii  binutils-msp430                      
2.17-2                                          Binutils for MSP430 targets
ii  gcc-msp430                           
3.2.3-2                                         GCC for MSP430 targets
ii  gdb-msp430                           
3.2.3-1                                         GDB for MSP430 targets
ii  msp430-jtag                          
0.0.20061017                                    MSP430 JTAG programmer
ii  msp430-libc                          
20071026                                        C library for MSP430 
targets

Please advise.

Yours sincerely,
Mustapha

[Mspsim-developers] Using the 32-bit multiplier

[<ni...@ly...>]

Hello,

I'm trying to develop and test code for MSP430F5510, using msp430-gcc
and mspsim. I'm doing some crypto code which needs bignum operations, so
using the hardware multiplier is important.

I have some working code using the 16x16 bit multiplier. I then compile
for msp430x1611, and I can run the resulting formware in the simulator
using se.sics.mspsim.platform.sky.SkyNode. Seems to work fine, and I
really appreciate the builtin debugging and profiling support of mspsim
(I have also tried remote debugging with gdb, which would be even
better, but I couldn't get that to work reliably).

Now, I'd like to be able to use the 32-bit multiplier. But there are
some obstacles:

1. I'm not sure which processor to build for and which platform to use.
   The closest I find in the simulator is MSP430F5437 (I'm new to this
   architecture, and I have no good idea on how it differs from the
   ...5510). In mspsim, the platform tyndal.TyndalNode seem to use this
   processor. Is this a good choice?

2. Looking at the file core/Multiplier32.java, it seems that the
   implementation is not complete.

3. The serial hardware seem to be different (different hardware
   registers, maybe more fundamental differences too?) than on
   msp430x1611. To get data out on the serial port, I use the file
   tests/msp430setup.c which among other things implements putchar. But
   this code doesn't work if I compile for msp430f5437 or msp430f5510, I
   get compilation errors when accessing various hardware registers.
   
Any help or suggestions on how to make progress is appreciated. My
current impression is that completing Multiplier32.java should not be a
big deal. I have a lot less clue on how to access the serial port, for
putchar.

Regards,
/Niels Möller

-- 
Niels Möller. PGP-encrypted email is preferred. Keyid C0B98E26.
Internet email is subject to wholesale government surveillance.