[U-Boot-Users] Virtex-II Pro and Instruction Prefetching

[U-Boot-Users] Virtex-II Pro and Instruction Prefetching

[Frank Smith]

Hello,

I'm working on getting PPCBoot 2.0.0 and Linux running on a custom
Virtex-II Pro development board.

I'm currently encountering problems with my DDR SDRAM controller --
it can't seem to handle instruction prefetching yet.  So to sidestep the
issue for now, I want to configure the 405 so that instructions
are not prefetched.

I've disabled instruction caching altogether (ICCR = 0x00000000) and I
have a core configuration register (CCR0) value of 0x00700000, which
according to the documentation has "prefetching for non-cachable
regions" disabled.

Even so, ChipScope is telling me that instructions are still 
being prefetched from SDRAM (sometimes 2 instructions, sometimes 4,
and sometimes 8 instructions at a time).  Is there anything else I have
to do to disable prefetching of non-cachable instructions?


Thanks,
Frank.

[U-Boot-Users] Virtex-II Pro and Instruction Prefetching

[Robert Schwebel]

On Sat, Jun 14, 2003 at 09:27:14AM -0300, Frank Smith wrote:
> I'm working on getting PPCBoot 2.0.0 and Linux running on a custom
> Virtex-II Pro development board.

Forget that old version. Use U-Boot and start with the latest release or
with CVS.

Robert 
-- 
 Dipl.-Ing. Robert Schwebel | http://www.pengutronix.de
 Pengutronix - Linux Solutions for Science and Industry
   Braunschweiger Str. 79,  31134 Hildesheim, Germany
   Handelsregister:  Amtsgericht Hildesheim, HRA 2686
    Phone: +49-5121-28619-0 |  Fax: +49-5121-28619-4

[U-Boot-Users] Re: Virtex-II Pro and Instruction Prefetching

[Wolfgang Denk]

In message <03Jun14.092013adt.7305@dragon.appliedmicro.ns.ca> you wrote:
> 
> I'm working on getting PPCBoot 2.0.0 and Linux running on a custom
> Virtex-II Pro development board.

May I ask why you  use  PPCBoot  (a  dead  project)  instead  of  the
successor project U-Boot?

> I'm currently encountering problems with my DDR SDRAM controller --

This is a very common problem. Seems 99% of all people who attempt to
port U-Boot see this soner or later.

> it can't seem to handle instruction prefetching yet.  So to sidestep the
> issue for now, I want to configure the 405 so that instructions
> are not prefetched.

Don't. The problem will bit you later again. Solve problems  as  they
appear.


Wolfgang Denk

-- 
Software Engineering:  Embedded and Realtime Systems,  Embedded Linux
Phone: (+49)-8142-4596-87  Fax: (+49)-8142-4596-88  Email: wd at denx.de
Violence in reality is quite different from theory.
	-- Spock, "The Cloud Minders", stardate 5818.4

[U-Boot-Users] Re: Virtex-II Pro and Instruction Prefetching

[Frank Smith]

On Sat, 14 Jun 2003, Wolfgang Denk wrote:

> In message <03Jun14.092013adt.7305@dragon.appliedmicro.ns.ca> you wrote:
> > 
> > I'm working on getting PPCBoot 2.0.0 and Linux running on a custom
> > Virtex-II Pro development board.
> 
> May I ask why you  use  PPCBoot  (a  dead  project)  instead  of  the
> successor project U-Boot?

Just haven't gotten around to it.  I promise I will soon (although I think
there's a bug in cpu/ppc4xx/start.S I've pointed out previously that
nobody has rendered an opinion on as of yet :-)

> > I'm currently encountering problems with my DDR SDRAM controller --
> 
> This is a very common problem. Seems 99% of all people who attempt to
> port U-Boot see this soner or later.
> 
> > it can't seem to handle instruction prefetching yet.  So to sidestep the
> > issue for now, I want to configure the 405 so that instructions
> > are not prefetched.
> 
> Don't. The problem will bit you later again. Solve problems  as  they
> appear.

Point taken.  I've identified the problem, and someone else is working on
solving it, so I'm just looking for a way to make some progress until it
gets solved.


Thanks,
-Frank.

[U-Boot-Users] Re: Virtex-II Pro and Instruction Prefetching

[Wolfgang Denk]

In message <03Jun14.105725adt.7305@dragon.appliedmicro.ns.ca> you wrote:
> 
> > May I ask why you  use  PPCBoot  (a  dead  project)  instead  of  the
> > successor project U-Boot?
> 
> Just haven't gotten around to it.  I promise I will soon (although I think

Do it now. Don't waste time on old stuff.

> there's a bug in cpu/ppc4xx/start.S I've pointed out previously that
> nobody has rendered an opinion on as of yet :-)

You mean the STD_EXCEPTION vs. CRIT_EXCEPTION ?

I am waiting for Stefan to pick this up :-)


But actually it should not matter at all: your code is never supposed
to run into such an exception.  If  it's  crashing  anyway  there  is
little difference if it's crashing "correctly" or not ;-)

Best regards,

Wolfgang Denk

-- 
Software Engineering:  Embedded and Realtime Systems,  Embedded Linux
Phone: (+49)-8142-4596-87  Fax: (+49)-8142-4596-88  Email: wd at denx.de
The following statement is not true.  The previous statement is true.

[U-Boot-Users] RE: Virtex-II Pro and Instruction Prefetching

[Frank Smith]

> -----Original Message-----
> From: wd at denx.de [mailto:wd at denx.de]

> > there's a bug in cpu/ppc4xx/start.S I've pointed out previously that
> > nobody has rendered an opinion on as of yet :-)
> 
> You mean the STD_EXCEPTION vs. CRIT_EXCEPTION ?
> 
> I am waiting for Stefan to pick this up :-)
> 
> But actually it should not matter at all: your code is never supposed
> to run into such an exception.  If  it's  crashing  anyway  there  is
> little difference if it's crashing "correctly" or not ;-)

The problem with it crashing incorrectly is that the Oops message 
lies to you, as it's reporting the contents of SRR0/SRR1, not SRR2/SRR3 
as it should be for a critical exception. Makes it more difficult to
debug when you think the wrong instruction is causing the exception.

Thanks,
-Frank.
---
Outgoing mail is certified Virus Free.
Checked by AVG anti-virus system (http://www.grisoft.com).
Version: 6.0.488 / Virus Database: 287 - Release Date: 6/5/03

[U-Boot-Users] RE: Virtex-II Pro and Instruction Prefetching

[Stefan Roese]

Hi!

> > > there's a bug in cpu/ppc4xx/start.S I've pointed out previously that
> > > nobody has rendered an opinion on as of yet :-)
> > 
> > You mean the STD_EXCEPTION vs. CRIT_EXCEPTION ?
> > 
> > I am waiting for Stefan to pick this up :-)
> > 
> > But actually it should not matter at all: your code is never supposed
> > to run into such an exception.  If  it's  crashing  anyway  there  is
> > little difference if it's crashing "correctly" or not ;-)
> 
> The problem with it crashing incorrectly is that the Oops message 
> lies to you, as it's reporting the contents of SRR0/SRR1, not SRR2/SRR3 
> as it should be for a critical exception. Makes it more difficult to
> debug when you think the wrong instruction is causing the exception.

OK, this seems to be correct. Let's give it a try!

Best regards,
Stefan.