headers
Bsderss | 1 Sep 2005 02:51
Picon
Favicon

Re: cheap RTEMS/PowerPC platform [was: Re: Open Desktop Workstation]


--- Karel Gardas <kgardas@...> wrote:

> On Wed, 24 Aug 2005, Joel Sherrill
> <joel@...> wrote:
> 
> >>> Hmm i have no idea what the mentioned hardware
> cost, so i can't answer
> >>> that question. But apple's mac mini is about
> 500USD which isn't that
> >>> expensive, when you want a general PowerPC.
> >> 
> >> 
> >> The problem is that IIRC it does not even provide
> serial port which is 
> >> needed for serial console. Pegasos seems to
> provide one RS-232.
> >> 
> >> BTW: Anobody here get RTEMS working on MOAB or
> CerfCube?
> >> http://www.tamsinc.com/3011/index.htm
> >>
>
http://www.intrinsyc.com/products/mob_ref_sys/cerfcube_405_ep/index.asp
> >
> > The Avnet evaluation board for the Coldfire that
> had a BSP submitted
> > recently is inexpensive.  Do they have any PowerPC
> boards like that?
> >
(Continue reading)

Permalink | Reply | 0 comments |
headers
Fabrício de Novaes Kucinskis | 1 Sep 2005 20:58
Picon
Favicon

Strange behavior when running RTEMS on an ERC32 chipset Tharsys board

I'm working with a Tharsys board with a ERC32 chipset (TSC691E + TSC69È +
TSC69É), and I noticed a very strange behavior when the RTEMS is being
initialized.

When trying to execute the example program "rtems-task" in the board, the
RTEMS freezes for some seconds, and then the watchdog timer resets the
board.

After many hours of debugging, I discovered that the problem occurs when the
RTEMS tries to read _ CLOCK_SPEED, stored in the address 0x020007e0.  This
address is in the trap table area of the ERC32.

At the moment where the RTEMS tries to read this value, what is written in
this position of memory is not the clock value, but the instruction "ta 0".

What I noticed was that, in the first instructions of the RTEMS, the Trap
Base Register (TBR) is configured. At this moment, some addresses of the
trap table have its routines modified, including the address of
_CLOCK_SPEED.

What I don't understand is that the overlapping of trap table occurs in the
execution of the instruction that modifies the TBR, what for me doesn't make
sense. I verified the system registers, and none trap occurred at this
moment.

It is important to notice that this problem doesn't happen in the Sparc
Instruction Simulator.

I solved the problem with a small patch, rewriting the address _ CLOCK_SPEED
after the TBR configuration, but I would like to know what is triggering the
(Continue reading)

Permalink | Reply | 4 comments |
headers
Joel Sherrill | 1 Sep 2005 23:38

Re: Strange behavior when running RTEMS on an ERC32 chipset Tharsys board

What version of RTEMS is this with?  I was seeing an issue today with
the CVS head on SIS.  It turned out that the start.s code is copying
initialized data from ROM to RAM but I didn't see anywhere in the 
linkcmds where the initialized data was actually placed there. 
Commenting out "copy_data" loop fixed it.

Can anybody comment on what sections need to be added to the
linkcmds to make this work?

-joel

Fabrício de Novaes Kucinskis wrote:
> I'm working with a Tharsys board with a ERC32 chipset (TSC691E + TSC69È +
> TSC69É), and I noticed a very strange behavior when the RTEMS is being
> initialized.
> 
> When trying to execute the example program "rtems-task" in the board, the
> RTEMS freezes for some seconds, and then the watchdog timer resets the
> board.
> 
> After many hours of debugging, I discovered that the problem occurs when the
> RTEMS tries to read _ CLOCK_SPEED, stored in the address 0x020007e0.  This
> address is in the trap table area of the ERC32.
> 
> At the moment where the RTEMS tries to read this value, what is written in
> this position of memory is not the clock value, but the instruction "ta 0".
> 
> What I noticed was that, in the first instructions of the RTEMS, the Trap
> Base Register (TBR) is configured. At this moment, some addresses of the
> trap table have its routines modified, including the address of
(Continue reading)

Permalink | Reply | 3 comments |
headers
Joel Sherrill | 1 Sep 2005 23:41

Mailing list archives


I don't know what went wrong and am not even sure what I did but
the mailing list archives for August and September are now there.
The same script has been running for years now as a cron job.

Hopefully the cron job won't mess them up again.

--

-- 
Joel Sherrill, Ph.D.             Director of Research & Development
joel@...                 On-Line Applications Research
Ask me about RTEMS: a free RTOS  Huntsville AL 35805
    Support Available             (256) 722-9985
Permalink | Reply | 0 comments |
headers
Joel Sherrill | 2 Sep 2005 00:14

RTEMS 4.6.4 Available

Hi,

It's that time again.  The list of fixes committed to the
4.6 branch reached a level that makes a new release a good
idea. When I reviewed the PRs that had been closed since 4.6.2,
it was clear that another release was needed.  38 PRs were
closed.  See http://www.rtems.com/rtems-4.6/index.html#464
for details.

It can be downloaded from

ftp://ftp.rtems.com/pub/rtems/4.6.4

OR

http://http.rtems.com/ftp/pub/rtems/4.6.4

I do not consider the 4.6 branch dead even with this release.
If there are sufficient problems reported, then there is nothing
preventing a 4.6.5.

FYI I plan to cut 4.6.99.2 from the CVS head tomorrow so we
have a baseline for reporting problems and -- hopefully soon --
forking a 4.7 branch.

--

-- 
Joel Sherrill, Ph.D.             Director of Research & Development
joel@...                 On-Line Applications Research
Ask me about RTEMS: a free RTOS  Huntsville AL 35805
    Support Available             (256) 722-9985
(Continue reading)

Permalink | Reply | 0 comments |
headers
Jiri Gaisler | 2 Sep 2005 00:25

[Fwd: Re: RES: Developent Environments for RTEMS]


For those who like graphical IDEs for software development,
we have now released a new version of the eclipse/cdt plugin
for leon2/3. The plugin requires eclipse-3.1 and the new
cdt-3.0.0. It is installed using the eclipse software update
mechnism, see :

http://www.gaisler.com/eclipse/cdt/

The plugin can display the current plug&play information
of leon3 systems, and has also a leon2/AT697 register
display. Here is a screen dump on what it can look like:

http://www.gaisler.com/eclipse/cdt/eclipse.gif

To debug on leon hardware, you need grmon-1.1.1 or later.

There is no support for ERC32 (yet), but the plugin is
provided in full source code under GPL so feel free to
add it ...

Fabrício de Novaes Kucinskis wrote:
> Wow - how many options!
> 
> 
> First of all, I have to apologize for the wrong information in my previous
> message, about the Gaisler's Eclipse Plugin. I understood that the plugin
> was paid, and delivered WITH the simulator.
> 
> Thank you for the info, Jiri, and forgive me. Next time, I'll read the
(Continue reading)

Permalink | Reply | 0 comments |
headers
Joel Sherrill | 2 Sep 2005 19:23

RTEMS Snapshot 4.6.99.2 and new RPMs


Hi,

I have cut the 4.6.99.2 snapshot and placed it on the ftp
site.  All problems fixed in the 4.6 release series should be
fixed in this source as well.  In addition, this includes
new BSPs, improvements, and additional features. See
http://www.rtems.org/rtems-4.7/index.html for an overview
of the changes from the 4.6 series.

I do not expect that this snapshot is perfect.  I
do believe it is in pretty good shape though.

For the snapshots to turn into a 4.7 release branch,
we need to discover and squash bugs.  Please test on
your hardware and update the appropriate BSP specific
page on the Wiki to indicate your results.  If there
are problems, report them.

It is available at:

ftp://www.rtems.com/pub/rtems/4.6.99.2/

or

http://www.rtems.com/ftp/pub/rtems/4.6.99.2/

The following new tool RPMs are also available and correspond
to what was used to test this snapshot.  The apt repository
information has been updated to reflect the availability of
(Continue reading)

Permalink | Reply | 2 comments |
headers
Joel Sherrill | 7 Sep 2005 16:31

Please Report RTEMS Results


Hi,

It should come as no surprise that no single person has all the
boards in the RTEMS source tree.  In that light, I am begging users
to report on their success or failure with particular boards and
versions.

I would like each BSP page on the Wiki to include some Test Results.
At the moment, I am happy with a "it runs OK" or "I had to apply the
patch with PR XXX to get it to run".  If you can say it runs hello and
ticker, that is actually very important information.  If you can comment
on networking, that is important.  For the pc386 BSP, knowing which NIC
drivers are getting use is important.

Here is an example of one I did for psim.  This is the actual Wiki text
from http://www.rtems.com/wiki/index.php/Psim:

==Test Reports==

4.6.4: [[User:JoelSherrill]] reports that it runs fine on gdb 5.3.

4.6.99.2: [[User::JoelSherrill]] reports that it runs fine on gdb 6.3.

If you are using a 4.5.x series or anything earlier in the 4.6 series,
please feel free to add that information.

If you are testing against CVS, use the version number "CVS-HEAD DATE"
where DATE is when it was last updated.
(Continue reading)

Permalink | Reply | 3 comments |
headers
Ivan Galkin | 7 Sep 2005 21:31
Favicon

Call for SPARCengine 1e BSP

Inspired by Joel's invitation, here's a call for a BSP port to 
SPARCengine 1e board.

SPARCengine 1e is a SPARC V7 board in VME, circa 1991.
This CPU was used as a prototype for the ERC32 chipset.
The existing RTEMS ERC32 BSP does not work on SPARCengine 1e (see below).

Motivation for BSP port:
An ERC32 evaluation board can be found at $15-18K from Atmel or Tharsys, 
whereas a used Force CPU-1CE or CPU-2CE boards are a mere $100 on ebay. 
The ERC32 VME boards continue to be a viable choice of 
radiation-hardened embedded systems for space applications.

Known problems with Rtems ERC32 on SPARCEngine 1e:
ERC32 BSP is in ELF format, whereas SPARCengine bootloader supports 
a.out-sunos-big only.
Very little expert help can be found, the board is too old.
Original SUN libraries for this board are hard to find.

Potential problems:
MMU got redesigned in ERC32.
VME support looks board-specific.

Hope to hear from someone interested.
Ivan
Permalink | Reply | 13 comments |
headers
Joel Sherrill | 7 Sep 2005 22:14

Re: Call for SPARCengine 1e BSP

Ivan Galkin wrote:
> Inspired by Joel's invitation, here's a call for a BSP port to 
> SPARCengine 1e board.

Does this match any of the hardware in the list at?

http://www.netbsd.org/Ports/sparc/

If you can boot NetBSD, then you have (OK someone has) all the driver 
information you need.  Not in the best form but there.

I updated your Wiki entry to include information on how to
address the a.out boot loader and som eother thoughts.  See

http://www.rtems.com/wiki/index.php/SPARCengine_1e

> SPARCengine 1e is a SPARC V7 board in VME, circa 1991.
> This CPU was used as a prototype for the ERC32 chipset.
> The existing RTEMS ERC32 BSP does not work on SPARCengine 1e (see below).
> 
> Motivation for BSP port:
> An ERC32 evaluation board can be found at $15-18K from Atmel or Tharsys, 
> whereas a used Force CPU-1CE or CPU-2CE boards are a mere $100 on ebay. 
> The ERC32 VME boards continue to be a viable choice of 
> radiation-hardened embedded systems for space applications.

Does Force have any documentation?  How about the ESA folks who 
originally did the ERC32?  I know it is old but Jiri.... any thoughts?

> Known problems with Rtems ERC32 on SPARCEngine 1e:
(Continue reading)

Permalink | Reply | 12 comments |

Gmane