Hanging in the Delay loop

Hanging in the Delay loop

[clark]

        Hello All,

        Does anybody know why a machine would hang in the Calibration Delay
Loop just before it spits out the BogoMIPs rating. Our board does however,
when we hook it up to the ICE it and reset it with the ICE it does not.

        Many thanks for any help you can give me.

        Conn Clark


** Sent via the linuxppc-embedded mail list. See http://lists.linuxppc.org/

Re: Hanging in the Delay loop

[Dan Malek]

clark@esteem.com wrote:

>         Does anybody know why a machine would hang in the Calibration Delay
> Loop just before it spits out the BogoMIPs rating.

This happens if the decrementer isn't running.  Is this an 8xx or 8260?
They both have the ability to disable the timebase, which includes the
decrementer (other processors don't do this).  An ICE often does more
than you know, which could be an unexpected register setting that causes
this to happen.


	-- Dan

** Sent via the linuxppc-embedded mail list. See http://lists.linuxppc.org/

Re: Hanging in the Delay loop

[Gabriel Paubert]

On Tue, 10 Oct 2000, Dan Malek wrote:

>
> clark@esteem.com wrote:
>
> >         Does anybody know why a machine would hang in the Calibration Delay
> > Loop just before it spits out the BogoMIPs rating.
>
> This happens if the decrementer isn't running.  Is this an 8xx or 8260?
> They both have the ability to disable the timebase, which includes the
> decrementer (other processors don't do this).  An ICE often does more
> than you know, which could be an unexpected register setting that causes
> this to happen.

Hmmm..., actually all processors have a TBEN pin (except the 601) and many
chipsets have a register which is designed to control this. This would be
very useful to synchronize timebases on SMP systems, the problem is to
write support code for all platforms (is there a way to control this pin
on SMP G4).

	Regards,
	Gabriel.


** Sent via the linuxppc-embedded mail list. See http://lists.linuxppc.org/

Re: Hanging in the Delay loop

[Dan Malek]

Gabriel Paubert wrote:

> Hmmm..., actually all processors have a TBEN pin (except the 601) and many
> chipsets have a register which is designed to control this.

Cool.  I didn't know most systems have external hardware to control
this.  The embedded processors have software control in some
internal registers.


	-- Dan

** Sent via the linuxppc-embedded mail list. See http://lists.linuxppc.org/

Re: Hanging in the Delay loop

[Gabriel Paubert]

On Tue, 10 Oct 2000, Dan Malek wrote:

> Gabriel Paubert wrote:
>
> > Hmmm..., actually all processors have a TBEN pin (except the 601) and many
> > chipsets have a register which is designed to control this.
>
> Cool.  I didn't know most systems have external hardware to control
> this.  The embedded processors have software control in some
> internal registers.

I'd wish they would have done this for the desktop processors too, to
avoid chipset dependent, or even board dependent, code. The main use is
SMP timebase synchronization but it is important.

	Gabriel.


** Sent via the linuxppc-embedded mail list. See http://lists.linuxppc.org/

Re: Hanging in the Delay loop

[Dan Malek]

Gabriel Paubert wrote:

> .... The main use is
> SMP timebase synchronization but it is important.

I'll just make one little suggestion, based upon lots of years
of multiprocessor development.  Find a software implementation that
doesn't require synchronized clocks local to each processor, and
then be grateful for the performance enhancement when you can have
synchronized clocks.

	-- Dan

** Sent via the linuxppc-embedded mail list. See http://lists.linuxppc.org/

Re: Hanging in the Delay loop

[Gabriel Paubert]

On Tue, 10 Oct 2000, Dan Malek wrote:

> Gabriel Paubert wrote:
>
> > .... The main use is
> > SMP timebase synchronization but it is important.
>
> I'll just make one little suggestion, based upon lots of years
> of multiprocessor development.  Find a software implementation that
> doesn't require synchronized clocks local to each processor, and
> then be grateful for the performance enhancement when you can have
> synchronized clocks.

I can't figure how an implementation which works without synchronized
clocks will not allow to trivially synchronize the timebases.
Of course I mean an implementation with good precision.

	Regards,
	Gabriel.


** Sent via the linuxppc-embedded mail list. See http://lists.linuxppc.org/

exchanges file in minicom

[Zhaobin Zhu]

Hi,

I boot linux over IDE disk.
The only connection between host and target is RS232 which is used
to download and as console(minicom).

After kernel is running on target, how can I send/receive files between
host(minicom/linux) and target ?

For ascii file, I can use "cat  < /dev/ttyS0" on target, and
"ascii-xfr .." on host. How about Binary files ?

Thanks,
--
 Zhaobin Zhu
 zzhu@emc.com


** Sent via the linuxppc-embedded mail list. See http://lists.linuxppc.org/

Re: exchanges file in minicom

[Eurico Inocencio]

Hello

>After kernel is running on target, how can I send/receive files between
>host(minicom/linux) and target ?
>For ascii file, I can use "cat  < /dev/ttyS0" on target, and
>"ascii-xfr .." on host. How about Binary files ?

Try ZMODEM, minicom can send and receive files using this
protocol, on the target you need to compile the ´rz' and
or 'sz' programs on your target, or like we do here, integrate
the ZMODEM protocol into the target application.

Like many others we have been using ZMODEM for a number
of years, with both minicom and the win98 "hyper terminal",
both are great for binary file transfers to the target.

Best Regards
Eurico
-----------------------------------------------------------------------
  Eurico Mendes Inocencio          CDFtel,
Lda
  Ph. No:  +351.22.3389843         Rua Goncalo Cristovao, 347 - Sala 205
  FAX No:  +351.22.3389845         4000-270 PORTO
  E-mail:  mailto:emi@cdftel.pt    Portugal
-----------------------------------------------------------------------


** Sent via the linuxppc-embedded mail list. See http://lists.linuxppc.org/

Re: exchanges file in minicom

[Jerry Van Baren]

You can use uuencode to convert binary to an ASCII encoding, transfer
the (now ASCII) file via cat (or use the minicom capture feature), and
then use uudecode to convert the file back into binary.  You can use
tar to bundle files together and gzip to make them smaller before you
uuencode them (making them bigger again :-).

This is technically inferior to the zmodem approach, but all the tools
should already be on your systems.

gvb


At 05:59 PM 10/10/00 -0400, Zhaobin Zhu wrote:

>Hi,
>
>I boot linux over IDE disk.
>The only connection between host and target is RS232 which is used
>to download and as console(minicom).
>
>After kernel is running on target, how can I send/receive files between
>host(minicom/linux) and target ?
>
>For ascii file, I can use "cat  < /dev/ttyS0" on target, and
>"ascii-xfr .." on host. How about Binary files ?
>
>Thanks,
>--
>  Zhaobin Zhu
>  zzhu@emc.com
>
>


** Sent via the linuxppc-embedded mail list. See http://lists.linuxppc.org/

Safe shutdown on IDE boot

[Zhaobin Zhu]

Hi,

I boot Hardhat linux on IDE disk on sp7400.

But I can boot it only once.  The next time I boot,
I see
VFS: Mounted root (ext2 filesystem) readonly.
Freeing unused kernel memory: 284k init

and hangs there.

I put the disk back to Host and found some problem with fs on this disk
and I can fix it by "e2fsck".

I guess i had problem with safely shutdown.
Anybody knows how to do the "shutdown" safely ?

--
 Zhaobin Zhu
 zzhu@emc.com


** Sent via the linuxppc-embedded mail list. See http://lists.linuxppc.org/

Re: Safe shutdown on IDE boot

[Alex Shnitman]

Hi, Zhaobin!

On Tue, Oct 10, 2000 at 02:01:53PM -0400, you wrote the following:

> I boot Hardhat linux on IDE disk on sp7400.
>
> But I can boot it only once.  The next time I boot,
> I see
> VFS: Mounted root (ext2 filesystem) readonly.
> Freeing unused kernel memory: 284k init
>
> and hangs there.
>
> I put the disk back to Host and found some problem with fs on this disk
> and I can fix it by "e2fsck".
>
> I guess i had problem with safely shutdown.
> Anybody knows how to do the "shutdown" safely ?

I have the exact same problem when I boot without an HD but with NFS
root over the network. So I don't think it has to do with shutting
down properly. All eyes to our MAG and Wizard. ;-)


--
Alex Shnitman                            | http://www.debian.org
alexsh@hectic.net, alexsh@linux.org.il   +-----------------------
http://alexsh.hectic.net    UIN 188956    PGP key on web page
       E1 F2 7B 6C A0 31 80 28  63 B8 02 BA 65 C7 8B BA

I drive way too fast to worry about cholesterol.

** Sent via the linuxppc-embedded mail list. See http://lists.linuxppc.org/