[ECOS] Re: ecos-discuss Digest 17 Apr 2003 16:10:44 -0000 Issue 1239

[ECOS] Re: ecos-discuss Digest 17 Apr 2003 16:10:44 -0000 Issue 1239

[Pierre Habraken]

> ecos-discuss-digest-help@sources.redhat.com wrote:
> 
> Subject: Re:[ECOS] Re: arm evaluator e7t hangs when executing example progs/tests
> Date: Thu, 17 Apr 2003 00:07:55 +0200
> From: "Matt Kober" <mattkober@web.de>
> To: "JonathanLarmour" <jifl@eCosCentric.com>
> CC: ecos-discuss@sources.redhat.com
> 
> Chris Garry provided me with a redboot image and the twothreads.out
> file that worked on his
> e7t. i had no luck with it. i even had a chance to try the files on 2
> other e7t's but they showed
> the same behavior.
> i did some debugging of a simple interrupt prog written in c (without
> eCos) and it seems like
> the external interrupt that i used doesn't get processed correctly.
> the interrupt pending bit is
> set but the cpu doesn't vector to the isr. i have to check that out
> more carefully tomorrow.
> 
> has anyone had a similar problem before ? i mean it seems to be a
> general problem with
> all e7t's i have tried.

I had this kind of behavior when running the e7t with Angel because the
latter starts application programs in user mode.
Using Redboot instead of Angel solves the problem.

Pierre
-- 
________________________________________________________________________
Pierre HABRAKEN - mailto:Pierre.Habraken@imag.fr
Tél: 04 76 82 72 83 - Fax: 04 76 82 72 87
IMAG-LSR BP72 38402 SAINT MARTIN D'HERES Cedex
________________________________________________________________________

-- 
Before posting, please read the FAQ: http://sources.redhat.com/fom/ecos
and search the list archive: http://sources.redhat.com/ml/ecos-discuss

Re: [ECOS] Re: ecos-discuss Digest 17 Apr 2003 16:10:44 -0000 Issue 1239

[Jonathan Larmour]

Pierre Habraken wrote:
>>Subject: Re:[ECOS] Re: arm evaluator e7t hangs when executing example progs/tests
>>Date: Thu, 17 Apr 2003 00:07:55 +0200
>>From: "Matt Kober" <mattkober@web.de>
>>To: "JonathanLarmour" <jifl@eCosCentric.com>
>>CC: ecos-discuss@sources.redhat.com
>>
>>Chris Garry provided me with a redboot image and the twothreads.out
>>file that worked on his
>>e7t. i had no luck with it. i even had a chance to try the files on 2
>>other e7t's but they showed
>>the same behavior.
>>i did some debugging of a simple interrupt prog written in c (without
>>eCos) and it seems like
>>the external interrupt that i used doesn't get processed correctly.
>>the interrupt pending bit is
>>set but the cpu doesn't vector to the isr. i have to check that out
>>more carefully tomorrow.
> 
> I had this kind of behavior when running the e7t with Angel because the
> latter starts application programs in user mode.
> Using Redboot instead of Angel solves the problem.

That could explain it, except that Matt said he tried RedBoot unless 
perhaps Matt put RedBoot on the board but still used Angel to load the 
applications?

Jifl
-- 
eCosCentric    http://www.eCosCentric.com/    The eCos and RedBoot experts
--[ "You can complain because roses have thorns, or you ]--
--[  can rejoice because thorns have roses." -Lincoln   ]-- Opinions==mine


-- 
Before posting, please read the FAQ: http://sources.redhat.com/fom/ecos
and search the list archive: http://sources.redhat.com/ml/ecos-discuss

Re: [ECOS] Re: ecos-discuss Digest 17 Apr 2003 16:10:44 -0000 Issue 1239

[Alex Schuilenburg]

Matt Kober wrote:
> that was actually a great hint, but i wasn't using angel at all. i had angel residing in flash though and i noticed that it always runs for a moment before redboot gets a chance to run. so angel probably messed up the memory map and mode settings. is that what could have happened?

Yes.  Angel is a nasty piece of work as it always assumes it has full 
control over the CPU (or did at least in my day).  When using Angel to 
bootstrap another RTOS from RAM, I always had to add a while delay loop 
of around 1 sec to allow Angel to "settle down" before I could reliably 
turn off interrupts, reconfigure the MMU, IRQ vectors etc.

HTH
-- Alex


> but shouldn't redboot initialize all the registers again when it's run? i would think so.
> 
> i solved the problem by rebuilding redboot with asserts and tracing enabled, but i totally don't understand why this should make a difference. fact is that my own test prog and all the kernel tests run just fine now.
> 
> Matt
> 
> 
> 
> 
>>Pierre Habraken wrote:
>>
>>>>Subject: Re:[ECOS] Re: arm evaluator e7t hangs when executing example progs/tests
> 
> 
>>>I had this kind of behavior when running the e7t with Angel because the
>>>latter starts application programs in user mode.
>>>Using Redboot instead of Angel solves the problem.
>>
>>That could explain it, except that Matt said he tried RedBoot unless 
>>perhaps Matt put RedBoot on the board but still used Angel to load the 
>>applications?
>>
>>Jifl
>>-- 
>>eCosCentric    http://www.eCosCentric.com/    The eCos and RedBoot experts
>>--[ "You can complain because roses have thorns, or you ]--
>>--[  can rejoice because thorns have roses." -Lincoln   ]-- Opinions==mine
>>
> 
> 
> 
> ______________________________________________________________________________
> UNICEF bittet um Spenden fur die Kinder im Irak! Hier online an
> UNICEF spenden: https://spenden.web.de/unicef/special/?mc=021101
> 
> 



-- 
Before posting, please read the FAQ: http://sources.redhat.com/fom/ecos
and search the list archive: http://sources.redhat.com/ml/ecos-discuss

Re: [ECOS] Re: ecos-discuss Digest 17 Apr 2003 16:10:44 -0000 Issue 1239

[Jonathan Larmour]

Matt Kober wrote:
> that was actually a great hint, but i wasn't using angel at all. i had
> angel residing in flash though and i noticed that it always runs for a
> moment before redboot gets a chance to run.

That's normal.

> so angel probably messed up
> the memory map and mode settings. is that what could have happened? but
> shouldn't redboot initialize all the registers again when it's run? i
> would think so.

Except that it works fine for others.... although I suppose it _could_ be 
some difference between Angel versions in use. I just can't tell from 
here, sorry.

> i solved the problem by rebuilding redboot with asserts and tracing
> enabled, but i totally don't understand why this should make a
> difference. fact is that my own test prog and all the kernel tests run
> just fine now.

Erk! That's more than a little strange. That sounds like it might be 
something like uninitialized data, or an alignment issue. Hmm. 
Unfortunately again, nothing I can do to tell which from here!

Jifl
-- 
eCosCentric    http://www.eCosCentric.com/    The eCos and RedBoot experts
--[ "You can complain because roses have thorns, or you ]--
--[  can rejoice because thorns have roses." -Lincoln   ]-- Opinions==mine


-- 
Before posting, please read the FAQ: http://sources.redhat.com/fom/ecos
and search the list archive: http://sources.redhat.com/ml/ecos-discuss

[ECOS] Re: ecos-discuss Digest 17 Apr 2003 16:10:44 -0000 Issue 1239

[Matt Kober]

that was actually a great hint, but i wasn't using angel at all. i had angel residing in flash though and i noticed that it always runs for a moment before redboot gets a chance to run. so angel probably messed up the memory map and mode settings. is that what could have happened?
but shouldn't redboot initialize all the registers again when it's run? i would think so.

i solved the problem by rebuilding redboot with asserts and tracing enabled, but i totally don't understand why this should make a difference. fact is that my own test prog and all the kernel tests run just fine now.

Matt



> Pierre Habraken wrote:
> >>Subject: Re:[ECOS] Re: arm evaluator e7t hangs when executing example progs/tests

> > 
> > I had this kind of behavior when running the e7t with Angel because the
> > latter starts application programs in user mode.
> > Using Redboot instead of Angel solves the problem.
> 
> That could explain it, except that Matt said he tried RedBoot unless 
> perhaps Matt put RedBoot on the board but still used Angel to load the 
> applications?
> 
> Jifl
> -- 
> eCosCentric    http://www.eCosCentric.com/    The eCos and RedBoot experts
> --[ "You can complain because roses have thorns, or you ]--
> --[  can rejoice because thorns have roses." -Lincoln   ]-- Opinions==mine
> 


______________________________________________________________________________
UNICEF bittet um Spenden fur die Kinder im Irak! Hier online an
UNICEF spenden: https://spenden.web.de/unicef/special/?mc=021101


-- 
Before posting, please read the FAQ: http://sources.redhat.com/fom/ecos
and search the list archive: http://sources.redhat.com/ml/ecos-discuss