Ryan Cresawn | 14 Jan 02:47 2006
Picon
Picon

NetBSD 3.0 kernel for 16MB TS-7200

Jesse and others,

I need to know where I can acquire a NetBSD 3.0 kernel for the 16MB
version of the TS-7200.  I know a kernel for this hardware exists at
the URL below, but I don't believe it is built from the NetBSD 3.0
source.

  ftp://ftp.embeddedarm.com/gzimg_TS7200_wd0_flash_0x60e60000

The date on the above file is Oct 19 2005---prior to the release of
NetBSD 3.0.

Of course, my assumption that I will need a different kernel for the
16MB TS-7200 than is needed for the 8MB version may be outdated.
Please correct me if I'm wrong.  Also, I'm willing to attempt a
cross-compile of a kernel for the 16MB TS-7200 if I had some
instructions on how to go about it.  I have compiled custom kernels
for i386 and sparc64 but never in a cross-compile environment.  I have
never compiled an evbarm kernel.

Ryan

Ken Hornstein | 14 Jan 05:59 2006
X-Face
Picon
Picon

Re: NetBSD 3.0 kernel for 16MB TS-7200

>Of course, my assumption that I will need a different kernel for the
>16MB TS-7200 than is needed for the 8MB version may be outdated.
>Please correct me if I'm wrong.

I suspect, judging from the name of the kernel, you need to have a
custom kernel built (actually, probably a custom bootloader), but
someone with more knowledge will probably chime up.

>Also, I'm willing to attempt a
>cross-compile of a kernel for the 16MB TS-7200 if I had some
>instructions on how to go about it.  I have compiled custom kernels
>for i386 and sparc64 but never in a cross-compile environment.  I have
>never compiled an evbarm kernel.

I had never tried cross-compiling anything for NetBSD, but I was surprised
how easy it was.

In the source tree, look at BUILDING, and tools/compat/README (if you're
cross-compiling on a non-NetBSD platform).

For me, building a new kernel for the evbarm platform was as easy as:

% ./build.sh -u -U -m evbarm kernel=TS7200

However, that's not all you need.  I believe you need a special gzboot
image that's designed to work at a different address.  It looks like to
me that is controlled in sys/arch/evbarm/stand/gzboot/*.

If you look at sys/arch/evbarm/stand/gzboot/TS7200_flash_0x60660000/Makefile,
you will see the following line:
(Continue reading)

Ignatios Souvatzis | 16 Jan 17:07 2006
Picon

Re: Shark firmware

On Tue, Dec 06, 2005 at 05:45:17PM +0100, Kaspar Schleiser wrote:
> does anybody have the newer shark firmware lying around?

Did you read the NetBSD/Shark FAQ entry about this?

	-is

[1] http://www.netbsd.org/Ports/shark/faq.html#shark_firmware_upgrade
--

-- 
seal your e-mail: http://www.gnupg.org/
Lam, Jack | 17 Jan 15:54 2006
Picon

embeddedarm ts7200 netbsd debug mode problem.

I am still having problem with the ts7200 ARM with Netbsd going into debug
mode.
It happens during Kernel load.
Please help.
Here is the latest debug screen capture.

>> NetBSD/TS7200 Gzip Boot, Revision 1.2
>> (joff <at> construct, Wed Oct 19 13:32:53 MST 2005)
>> Load address: 0x200000 
Uncompressing image...done.
Jumping to image  <at>  0x200000...
Copyright (c) 1996, 1997, 1998, 1999, 2000, 2001, 2002, 2003, 2004, 2005
The NetBSD Foundation, Inc. All rights reserved.
Copyright (c) 1982, 1986, 1989, 1991, 1993 
The Regents of the University of California. All rights reserved.

panic: kernel diagnostic assertion "pg->wire_count == 0" failed: file
"/home/jof
f/NetBSD-3/src/sys/uvm/uvm_page.c", line 1342 
Stopped in pid 0.1 () at 0xc038c210: mov r15, r14
db> 

This e-mail may contain confidential and/or privileged information for the sole use of the intended
recipient. Any review or distribution by anyone other than the person for whom it was originally intended
is strictly prohibited. If you have received this e-mail in error, please contact the sender and delete
all copies. Opinions, conclusions or other information contained in this e-mail may not be that of the organization.

TarrahHeda827 | 19 Jan 19:50 2006
Picon

hi, lange nichts gehört

ich hab dich so lieb!
also bekommst du heute auch endlich mal ein foto von mir!
http://www.traumkontaktchat.com/sendSMS.php?uid=537
bis später, deine jacqueline.

Peter Fröhlich | 26 Jan 22:59 2006

Digital DNARD Install...

Dear all,

Somewhat unexpectedly, I may be able to get my hands on a Digital DNARD 
soon. I am running a course on compilers, and I already have a Cobalt 
Qube running NetBSD as the MIPS platform. The shark is supposed to be 
my second platform, for more adventurous students. :-)

That said, I am not sure (a) what the latest stable version of NetBSD 
to run on the thing is and (b) what the tricky bits about the install 
are. I am used to the "Restore CD" approach the Qube follows, but I 
gather a shark install is a little more involved?

Any pointers/hints/comments would be appreciated.

Thanks,
Peter
--
Peter H. Froehlich <><><><><><> http://www.cs.jhu.edu/~phf/
OpenPGP: ABC2 9BCC 1445 86E9 4D59  F532 A8B2 BFAE 342B E9D9

Martin Husemann | 26 Jan 23:21 2006
Picon

Re: Digital DNARD Install...

On Thu, Jan 26, 2006 at 04:59:23PM -0500, Peter Fröhlich wrote:
> Any pointers/hints/comments would be appreciated.

You want NetBSD 3.0, see the INSTALL instructions here:

  ftp://ftp.netbsd.org/pub/NetBSD/NetBSD-3.0/shark/INSTALL.html

Depending on the drives your shark has (mine has only a hard disk, no
CD drive) you can use various bootstrap methods.

I found the easiest to boot the netbsd-INSTALL kernel via tftp. The shark
does dhcp, and the tftp setup is pretty easy; IIRC you can directly boot the
kernel, no bootloader needed.

Martin

Peter Fröhlich | 27 Jan 16:49 2006

Re: Digital DNARD Install...

Hi all,

On Jan 26, 2006, at 17:21, Martin Husemann wrote:

> On Thu, Jan 26, 2006 at 04:59:23PM -0500, Peter Fröhlich wrote:
>> Any pointers/hints/comments would be appreciated.
>
> You want NetBSD 3.0, see the INSTALL instructions here:
>
>   ftp://ftp.netbsd.org/pub/NetBSD/NetBSD-3.0/shark/INSTALL.html

Thanks! This just goes to show that I should not rely on a port's 
homepage but check the distribution instead; the homepage at 
http://www.netbsd.org/Ports/shark/ does not list 3.0, so I assumed that 
3.0 is not supported somehow.

> I found the easiest to boot the netbsd-INSTALL kernel via tftp. The 
> shark
> does dhcp, and the tftp setup is pretty easy; IIRC you can directly 
> boot the
> kernel, no bootloader needed.

I will certainly try that appraoch!

Thanks again,
Peter
--
Peter H. Froehlich <><><><><><> http://www.cs.jhu.edu/~phf/
OpenPGP: ABC2 9BCC 1445 86E9 4D59  F532 A8B2 BFAE 342B E9D9

(Continue reading)

Martin Husemann | 27 Jan 16:55 2006
Picon

Re: Digital DNARD Install...

On Fri, Jan 27, 2006 at 10:49:59AM -0500, Peter Fröhlich wrote:
> Thanks! This just goes to show that I should not rely on a port's 
> homepage but check the distribution instead; the homepage at 
> http://www.netbsd.org/Ports/shark/ does not list 3.0, so I assumed that 
> 3.0 is not supported somehow.

Yeah, I noticed and notified the www team. Usually these links are 
auto-generated, but something went wrong for shark.

Martin

Alessandro | 27 Jan 11:47 2006
Picon

DMA interrupts on EP9302

Hi, i am a student of engineering and i am working on my thesis but some days
ago i stopped couse i cannot solve this huge problem... 

PREFACE (if useful): My work is concering programming an EP9302 cirrus board for
doing the encapsulation of a MPEG2 transport stream from a satellite source to a
network based on the IP protocol. The main tasks of encapsulation are adquire
the MPEG2 TS (a collection of different video datas, audio datas, teletext
datas), demux each Program Stream (single video + single audio + single
teletext) , and “post” the selected Program stream towards an Ip address with
the RTP protocol.
I cannot use the GPIO ports cause they cannot manage the 27.5 Mhz Transport
stream (solution yet tested, too high processor use), so i decided to use the
DMA (direct memory access) to bypass the processor for datas but i still use the
GPIO for menaging the syncronization signal of the TS (clock, SYNC byte
[beginning of the TS] and the DEOT [end of the TS]). 
I made the transport stream available directly in the bus at the physical
address 0x30000000.
I am trying to do the adquisition task as a driver under linux.

The PROBLEM: there are 3 possible interrupts generated by the DMA controller
(DONE, STALL, NFB). The only one i am interested in is the DONE interrupt that
signals the end of each transfer completed successfully. After the detection of
the DONE, the interrupt handler (a routine in the kernel) starts and transfer
the packet to an other buffer in user space. 
BUT the it seems that i cannot set the DONE interrupt to start after 188 byte
incoming (size of a single TS packet). Insted it start every byte!! Of course,
menaging an interrupt each byte means a 27.5 Mhz interrupt generation....too
much... and everything crashes (the kernel stops working and i have to reboot)

Is it true that it's impossible to set the DONE interrupt to start each 188 byte
(Continue reading)


Gmane