RTP on Insyde XPressRom????

Has anyone out there recently used RTP 6.x on a MediaGX/Geode
platform running the InsydeSW XpressRom product?

When trying to boot this we get several warning messages about
PnP BIOS not being detected by QNX and then it starts trying to
detect the IDE device and locks up during the process.

Also of note, this is on a Geode SC1200 – the first one of which
I have tried to run RTP on.

I am aware that the current version of RTP has some serious
issues on x86 boxes without a proper PCI BIOS implementation
(which I feel is inexcusable for an embedded RTOS – QNX4
worked nicely in this situation and if RTP can’t, QNX will be
losing a very substantial market IMHO) but, at least from what
I can tell, this doesn’t seem to be a PCI BIOS issue (at least I’m
not getting the PCI Bios error message that I am used to getting).

Has anyone else out there had any luck with similar configurations?

\

Michael Burkey (mailto:Michael.Burkey@Nexwarecorp.com)
Nexware Corp. (http://www.nexwarecorp.com)
Software Engineer
865.546.9998 x201

I am aware that the current version of RTP has some serious
issues on x86 boxes without a proper PCI BIOS implementation
(which I feel is inexcusable for an embedded RTOS –

What does QNX RtP (a development system) requiring a PCI BIOS, have to
do with QNX Neutrino’s suitability (or unsuitability) for being embedded
on a BIOSless target ?

Rennie

Hi Michael,

Assuming the MediaGX board is not your development system.

Does the MediaGX board contain a PCI bus? If not then you
shouldn’t be running pci-bios in your OS image. Could you
give a bit more detail on your OS image and what the system
contains. We have several MediaGX boards here that run fine
with QNX RTOS 6.1.

Thanks.

Erick.


Michael D. Burkey <michael.burkey@nexwarecorp.com> wrote:

Has anyone out there recently used RTP 6.x on a MediaGX/Geode
platform running the InsydeSW XpressRom product?

When trying to boot this we get several warning messages about
PnP BIOS not being detected by QNX and then it starts trying to
detect the IDE device and locks up during the process.

Also of note, this is on a Geode SC1200 – the first one of which
I have tried to run RTP on.

I am aware that the current version of RTP has some serious
issues on x86 boxes without a proper PCI BIOS implementation
(which I feel is inexcusable for an embedded RTOS – QNX4
worked nicely in this situation and if RTP can’t, QNX will be
losing a very substantial market IMHO) but, at least from what
I can tell, this doesn’t seem to be a PCI BIOS issue (at least I’m
not getting the PCI Bios error message that I am used to getting).

Has anyone else out there had any luck with similar configurations?


Michael Burkey (mailto:> Michael.Burkey@Nexwarecorp.com> )
Nexware Corp. (> http://www.nexwarecorp.com> )
Software Engineer
865.546.9998 x201

When I say “RTP” I mean QNX 6.0/NTO/Augusta/et al.

The issue is that although you can boot a VERY stripped
down NTO system without having “pci-bios” running you
can’t do a heck of a lot as all the network drivers require it
in order to work.

Under QNX4, the older _CA_PCI… calls could be used
to access the PCI BIOS directly (if you had one) or hard
coded port addresses could used for the network drivers.

The problem is that a lot of systems out there may have a
PCI bus but NOT a PCI Bios (i.e. an embedded BIOSless
target) and still need a network driver that can work.

For non-x86 platforms, QNX provides drivers such as
“pci-raven” that, AFAIK, configure the PCI bus and
devices via a bare metal interface straight to the hardware.
They need to provide the same for biosless x86 platforms!
Either that or they need to provide versions of the device
drivers that just allow you to specify what io/memory
locations the devices are hard coded to be at and not
bother trying to probe for them via pci-bios.

“Rennie Allen” <RAllen@csical.com> wrote in message news:64F00D816A85D51198390050046F80C9A31F@exchangecal.hq.csical.com

I am aware that the current version of RTP has some serious
issues on x86 boxes without a proper PCI BIOS implementation
(which I feel is inexcusable for an embedded RTOS –

What does QNX RtP (a development system) requiring a PCI BIOS, have to
do with QNX Neutrino’s suitability (or unsuitability) for being embedded
on a BIOSless target ?

Rennie

Well, we have tracked this down a bit farther on our own.

As I stated, I didn’t really think it was a pci-bios issue.
However enum IS complaining that no “PnP” BIOS
can be found (i.e. the ISA PNP BIOS) but it is saying that
that is not an error – just a warning.

I am TOO familiar with running RTP/NTO on the
MediaGX/Geode platform and have numerous boards
here that work quite nicely. However, this one doesn’t
seem to work real well at the moment.

What I am really asking is if anyone there has actually
tested with the newest XpressRom from InsydeSW on
an SC1200 Geode platform which is quite a bit different
than the older Geode GXm/GX1/GXLV cores.

We have now been able to boot the unit by creating a
VERY stripped down boot image and are trying to localize
it to find out whats crashing.

So far, it looks like devc-ser8250 is the culprit. Which
doesn’t suprise me because, as I understand it, the SC1200
pulls the UARTs into the core and they are basically
virtualized using the VSAII routines just like they did for
their SB Audio and VGA implementations. Which is to
say a “virtual” 8250 UART being handled through an
SMI interrupt routine.

However, even with devc-8250 pulled, we still haven’t
gotten Photon to run on this core yet. Which also doesn’t
really surprise me as they have added quite a few new
video features to the SC1200 (Video Input Port, etc).

Has anyone there actually tested the “SC” series at all yet???

\

Michael Burkey (mailto:Michael.Burkey@Nexwarecorp.com)
Nexware Corp. (http://www.nexwarecorp.com)
Software Engineer
865.546.9998 x201



“Hardware Support Account” <hw@qnx.com> wrote in message news:9mj3ok$ijb$2@nntp.qnx.com

Hi Michael,

Assuming the MediaGX board is not your development system.

Does the MediaGX board contain a PCI bus? If not then you
shouldn’t be running pci-bios in your OS image. Could you
give a bit more detail on your OS image and what the system
contains. We have several MediaGX boards here that run fine
with QNX RTOS 6.1.

Thanks.

Erick.


Michael D. Burkey <> michael.burkey@nexwarecorp.com> > wrote:
Has anyone out there recently used RTP 6.x on a MediaGX/Geode
platform running the InsydeSW XpressRom product?

When trying to boot this we get several warning messages about
PnP BIOS not being detected by QNX and then it starts trying to
detect the IDE device and locks up during the process.

Also of note, this is on a Geode SC1200 – the first one of which
I have tried to run RTP on.

I am aware that the current version of RTP has some serious
issues on x86 boxes without a proper PCI BIOS implementation
(which I feel is inexcusable for an embedded RTOS – QNX4
worked nicely in this situation and if RTP can’t, QNX will be
losing a very substantial market IMHO) but, at least from what
I can tell, this doesn’t seem to be a PCI BIOS issue (at least I’m
not getting the PCI Bios error message that I am used to getting).

Has anyone else out there had any luck with similar configurations?

\

Michael Burkey (mailto:> Michael.Burkey@Nexwarecorp.com> )
Nexware Corp. (> http://www.nexwarecorp.com> )
Software Engineer
865.546.9998 x201
\

The plot thickens…

Well, here is what we have learned in probing the Geode SC1200.

  1. devc-8250 does not work properly and never returns.
    (as the 8250’s are now “virtual” this doesn’t suprise me)
  2. if we start USB printer support devu-ohci dies.
    I would guess this is because their may still be some hooks
    from back into the “real” parallel printer drives somewhere
    in the code and – guess what – the printer port on the SC1200
    is now virtualized just like the UARTs are…
  3. we got Photon working but it looks a bit “strange”
    In 16 bit color mode it looks almost palletized (grainy) but
    seems to function ok. However, the manner in which the
    overlay regions is handled in the SC1200 appears to have
    changed somewhat. At the very least, the colorspace mapping
    appears to be different from what it used to be. Devg-MediaGX
    is returning that the requested colorspace is available and it
    does work, however, the mapping is reversed for U and it looks
    like the V component is completely missing. I’ll check the data
    sheets for the SC1200 shortly and see what looks different.

So, is there any hope of seeing new drivers to fix these
problems in the near future? Or workarounds?


Michael Burkey (mailto:Michael.Burkey@Nexwarecorp.com)
Nexware Corp. (http://www.nexwarecorp.com)
Software Engineer
865.546.9998 x201




“Hardware Support Account” <hw@qnx.com> wrote in message news:9mj3ok$ijb$2@nntp.qnx.com

Hi Michael,

Assuming the MediaGX board is not your development system.

Does the MediaGX board contain a PCI bus? If not then you
shouldn’t be running pci-bios in your OS image. Could you
give a bit more detail on your OS image and what the system
contains. We have several MediaGX boards here that run fine
with QNX RTOS 6.1.

Thanks.

Erick.


Michael D. Burkey <> michael.burkey@nexwarecorp.com> > wrote:
Has anyone out there recently used RTP 6.x on a MediaGX/Geode
platform running the InsydeSW XpressRom product?

When trying to boot this we get several warning messages about
PnP BIOS not being detected by QNX and then it starts trying to
detect the IDE device and locks up during the process.

Also of note, this is on a Geode SC1200 – the first one of which
I have tried to run RTP on.

I am aware that the current version of RTP has some serious
issues on x86 boxes without a proper PCI BIOS implementation
(which I feel is inexcusable for an embedded RTOS – QNX4
worked nicely in this situation and if RTP can’t, QNX will be
losing a very substantial market IMHO) but, at least from what
I can tell, this doesn’t seem to be a PCI BIOS issue (at least I’m
not getting the PCI Bios error message that I am used to getting).

Has anyone else out there had any luck with similar configurations?

\

Michael Burkey (mailto:> Michael.Burkey@Nexwarecorp.com> )
Nexware Corp. (> http://www.nexwarecorp.com> )
Software Engineer
865.546.9998 x201
\

I have now compared the datasheets for the video
controller config registers video input format between
the SC1200 and the CS5530 and they do NOT
match each other in 16bit format except for one mode.

Any hope of getting a new Devg-MediaGx in the future???

Or the source in the DDK so I can fix it myself?
(which I know is doubtful due to potential NDA issues w/NatSemi)


Michael Burkey (mailto:Michael.Burkey@Nexwarecorp.com)
Nexware Corp. (http://www.nexwarecorp.com)
Software Engineer
865.546.9998 x201


“Michael D. Burkey” <michael.burkey@nexwarecorp.com> wrote in message news:9mji44$9pt$1@inn.qnx.com

The plot thickens…

Well, here is what we have learned in probing the Geode SC1200.

  1. devc-8250 does not work properly and never returns.
    (as the 8250’s are now “virtual” this doesn’t suprise me)
  2. if we start USB printer support devu-ohci dies.
    I would guess this is because their may still be some hooks
    from back into the “real” parallel printer drives somewhere
    in the code and – guess what – the printer port on the SC1200
    is now virtualized just like the UARTs are…
  3. we got Photon working but it looks a bit “strange”
    In 16 bit color mode it looks almost palletized (grainy) but
    seems to function ok. However, the manner in which the
    overlay regions is handled in the SC1200 appears to have
    changed somewhat. At the very least, the colorspace mapping
    appears to be different from what it used to be. Devg-MediaGX
    is returning that the requested colorspace is available and it
    does work, however, the mapping is reversed for U and it looks
    like the V component is completely missing. I’ll check the data
    sheets for the SC1200 shortly and see what looks different.

So, is there any hope of seeing new drivers to fix these
problems in the near future? Or workarounds?


Michael Burkey (mailto:> Michael.Burkey@Nexwarecorp.com> )
Nexware Corp. (> http://www.nexwarecorp.com> )
Software Engineer
865.546.9998 x201




“Hardware Support Account” <> hw@qnx.com> > wrote in message news:9mj3ok$ijb$> 2@nntp.qnx.com> …
Hi Michael,

Assuming the MediaGX board is not your development system.

Does the MediaGX board contain a PCI bus? If not then you
shouldn’t be running pci-bios in your OS image. Could you
give a bit more detail on your OS image and what the system
contains. We have several MediaGX boards here that run fine
with QNX RTOS 6.1.

Thanks.

Erick.


Michael D. Burkey <> michael.burkey@nexwarecorp.com> > wrote:
Has anyone out there recently used RTP 6.x on a MediaGX/Geode
platform running the InsydeSW XpressRom product?

When trying to boot this we get several warning messages about
PnP BIOS not being detected by QNX and then it starts trying to
detect the IDE device and locks up during the process.

Also of note, this is on a Geode SC1200 – the first one of which
I have tried to run RTP on.

I am aware that the current version of RTP has some serious
issues on x86 boxes without a proper PCI BIOS implementation
(which I feel is inexcusable for an embedded RTOS – QNX4
worked nicely in this situation and if RTP can’t, QNX will be
losing a very substantial market IMHO) but, at least from what
I can tell, this doesn’t seem to be a PCI BIOS issue (at least I’m
not getting the PCI Bios error message that I am used to getting).

Has anyone else out there had any luck with similar configurations?

\

Michael Burkey (mailto:> Michael.Burkey@Nexwarecorp.com> )
Nexware Corp. (> http://www.nexwarecorp.com> )
Software Engineer
865.546.9998 x201


\

Ok, I understand what your point is now, but AFAIK the drivers shouldn’t
require a PCI BIOS as long as all the pertinent data is provided to them
on the command line (which should’nt be a problem in an embedded
system).

-----Original Message-----
From: Michael D. Burkey [mailto:michael.burkey@nexwarecorp.com]
Posted At: Wednesday, August 29, 2001 11:57 AM
Posted To: installation
Conversation: RTP on Insyde XPressRom???
Subject: Re: RTP on Insyde XPressRom???


When I say “RTP” I mean QNX 6.0/NTO/Augusta/et al.

The issue is that although you can boot a VERY stripped
down NTO system without having “pci-bios” running you
can’t do a heck of a lot as all the network drivers require it
in order to work.

Under QNX4, the older _CA_PCI… calls could be used
to access the PCI BIOS directly (if you had one) or hard
coded port addresses could used for the network drivers.

The problem is that a lot of systems out there may have a
PCI bus but NOT a PCI Bios (i.e. an embedded BIOSless
target) and still need a network driver that can work.

For non-x86 platforms, QNX provides drivers such as
“pci-raven” that, AFAIK, configure the PCI bus and
devices via a bare metal interface straight to the hardware.
They need to provide the same for biosless x86 platforms!
Either that or they need to provide versions of the device
drivers that just allow you to specify what io/memory
locations the devices are hard coded to be at and not
bother trying to probe for them via pci-bios.

“Rennie Allen” <RAllen@csical.com> wrote in message
news:64F00D816A85D51198390050046F80C9A31F@exchangecal.hq.csical.com

I am aware that the current version of RTP has some serious
issues on x86 boxes without a proper PCI BIOS implementation
(which I feel is inexcusable for an embedded RTOS –

What does QNX RtP (a development system) requiring a PCI BIOS, have to
do with QNX Neutrino’s suitability (or unsuitability) for being
embedded
on a BIOSless target ?

Rennie

I agree…however the current NTO network drivers (AFAIK)
no longer accept the port addresses/etc via command line
options – i.e. if the don’t see “pci-bios” they simply refuse
to run at all.

Has anyone out there had any luck getting a true NTO
(6.0C/6.1) system up and running on a machine without a PCI Bios
without having to resort to writing your own device drivers for
the NIC’s, et al?

“Rennie Allen” <RAllen@csical.com> wrote in message news:64F00D816A85D51198390050046F80C9A495@exchangecal.hq.csical.com

Ok, I understand what your point is now, but AFAIK the drivers shouldn’t
require a PCI BIOS as long as all the pertinent data is provided to them
on the command line (which should’nt be a problem in an embedded
system).

-----Original Message-----
From: Michael D. Burkey [mailto:> michael.burkey@nexwarecorp.com> ]
Posted At: Wednesday, August 29, 2001 11:57 AM
Posted To: installation
Conversation: RTP on Insyde XPressRom???
Subject: Re: RTP on Insyde XPressRom???


When I say “RTP” I mean QNX 6.0/NTO/Augusta/et al.

The issue is that although you can boot a VERY stripped
down NTO system without having “pci-bios” running you
can’t do a heck of a lot as all the network drivers require it
in order to work.

Under QNX4, the older _CA_PCI… calls could be used
to access the PCI BIOS directly (if you had one) or hard
coded port addresses could used for the network drivers.

The problem is that a lot of systems out there may have a
PCI bus but NOT a PCI Bios (i.e. an embedded BIOSless
target) and still need a network driver that can work.

For non-x86 platforms, QNX provides drivers such as
“pci-raven” that, AFAIK, configure the PCI bus and
devices via a bare metal interface straight to the hardware.
They need to provide the same for biosless x86 platforms!
Either that or they need to provide versions of the device
drivers that just allow you to specify what io/memory
locations the devices are hard coded to be at and not
bother trying to probe for them via pci-bios.

“Rennie Allen” <> RAllen@csical.com> > wrote in message
news:> 64F00D816A85D51198390050046F80C9A31F@exchangecal.hq.csical.com> …
I am aware that the current version of RTP has some serious
issues on x86 boxes without a proper PCI BIOS implementation
(which I feel is inexcusable for an embedded RTOS –

What does QNX RtP (a development system) requiring a PCI BIOS, have to
do with QNX Neutrino’s suitability (or unsuitability) for being
embedded
on a BIOSless target ?

Rennie

Hi Mike,

We are currently looking into getting a board with the
SC1200, so we may come up with a solution. We do
have an NDA with NatSemi, so cannot put the source in
the DDK.

Thanks,

Joe

Michael D. Burkey <michael.burkey@nexwarecorp.com> wrote:

I have now compared the datasheets for the video
controller config registers video input format between
the SC1200 and the CS5530 and they do NOT
match each other in 16bit format except for one mode.

Any hope of getting a new Devg-MediaGx in the future???

Or the source in the DDK so I can fix it myself?
(which I know is doubtful due to potential NDA issues w/NatSemi)


Michael Burkey (mailto:> Michael.Burkey@Nexwarecorp.com> )
Nexware Corp. (> http://www.nexwarecorp.com> )
Software Engineer
865.546.9998 x201



“Michael D. Burkey” <> michael.burkey@nexwarecorp.com> > wrote in message news:9mji44$9pt$> 1@inn.qnx.com> …
The plot thickens…

Well, here is what we have learned in probing the Geode SC1200.

  1. devc-8250 does not work properly and never returns.
    (as the 8250’s are now “virtual” this doesn’t suprise me)
  2. if we start USB printer support devu-ohci dies.
    I would guess this is because their may still be some hooks
    from back into the “real” parallel printer drives somewhere
    in the code and – guess what – the printer port on the SC1200
    is now virtualized just like the UARTs are…
  3. we got Photon working but it looks a bit “strange”
    In 16 bit color mode it looks almost palletized (grainy) but
    seems to function ok. However, the manner in which the
    overlay regions is handled in the SC1200 appears to have
    changed somewhat. At the very least, the colorspace mapping
    appears to be different from what it used to be. Devg-MediaGX
    is returning that the requested colorspace is available and it
    does work, however, the mapping is reversed for U and it looks
    like the V component is completely missing. I’ll check the data
    sheets for the SC1200 shortly and see what looks different.

So, is there any hope of seeing new drivers to fix these
problems in the near future? Or workarounds?


Michael Burkey (mailto:> Michael.Burkey@Nexwarecorp.com> )
Nexware Corp. (> http://www.nexwarecorp.com> )
Software Engineer
865.546.9998 x201




“Hardware Support Account” <> hw@qnx.com> > wrote in message news:9mj3ok$ijb$> 2@nntp.qnx.com> …
Hi Michael,

Assuming the MediaGX board is not your development system.

Does the MediaGX board contain a PCI bus? If not then you
shouldn’t be running pci-bios in your OS image. Could you
give a bit more detail on your OS image and what the system
contains. We have several MediaGX boards here that run fine
with QNX RTOS 6.1.

Thanks.

Erick.


Michael D. Burkey <> michael.burkey@nexwarecorp.com> > wrote:
Has anyone out there recently used RTP 6.x on a MediaGX/Geode
platform running the InsydeSW XpressRom product?

When trying to boot this we get several warning messages about
PnP BIOS not being detected by QNX and then it starts trying to
detect the IDE device and locks up during the process.

Also of note, this is on a Geode SC1200 – the first one of which
I have tried to run RTP on.

I am aware that the current version of RTP has some serious
issues on x86 boxes without a proper PCI BIOS implementation
(which I feel is inexcusable for an embedded RTOS – QNX4
worked nicely in this situation and if RTP can’t, QNX will be
losing a very substantial market IMHO) but, at least from what
I can tell, this doesn’t seem to be a PCI BIOS issue (at least I’m
not getting the PCI Bios error message that I am used to getting).

Has anyone else out there had any luck with similar configurations?

\

Michael Burkey (mailto:> Michael.Burkey@Nexwarecorp.com> )
Nexware Corp. (> http://www.nexwarecorp.com> )
Software Engineer
865.546.9998 x201


\

Hi,

There is a new devg-cyrix.so posted on the developers corner under New Stuff.
http://deverlopers.qnx.com

Regards,

Joe

Hardware Support Account <hw@qnx.com> wrote:

Hi Mike,

We are currently looking into getting a board with the
SC1200, so we may come up with a solution. We do
have an NDA with NatSemi, so cannot put the source in
the DDK.

Thanks,

Joe

Michael D. Burkey <> michael.burkey@nexwarecorp.com> > wrote:
I have now compared the datasheets for the video
controller config registers video input format between
the SC1200 and the CS5530 and they do NOT
match each other in 16bit format except for one mode.

Any hope of getting a new Devg-MediaGx in the future???

Or the source in the DDK so I can fix it myself?
(which I know is doubtful due to potential NDA issues w/NatSemi)


Michael Burkey (mailto:> Michael.Burkey@Nexwarecorp.com> )
Nexware Corp. (> http://www.nexwarecorp.com> )
Software Engineer
865.546.9998 x201



“Michael D. Burkey” <> michael.burkey@nexwarecorp.com> > wrote in message news:9mji44$9pt$> 1@inn.qnx.com> …
The plot thickens…

Well, here is what we have learned in probing the Geode SC1200.

  1. devc-8250 does not work properly and never returns.
    (as the 8250’s are now “virtual” this doesn’t suprise me)
  2. if we start USB printer support devu-ohci dies.
    I would guess this is because their may still be some hooks
    from back into the “real” parallel printer drives somewhere
    in the code and – guess what – the printer port on the SC1200
    is now virtualized just like the UARTs are…
  3. we got Photon working but it looks a bit “strange”
    In 16 bit color mode it looks almost palletized (grainy) but
    seems to function ok. However, the manner in which the
    overlay regions is handled in the SC1200 appears to have
    changed somewhat. At the very least, the colorspace mapping
    appears to be different from what it used to be. Devg-MediaGX
    is returning that the requested colorspace is available and it
    does work, however, the mapping is reversed for U and it looks
    like the V component is completely missing. I’ll check the data
    sheets for the SC1200 shortly and see what looks different.

So, is there any hope of seeing new drivers to fix these
problems in the near future? Or workarounds?


Michael Burkey (mailto:> Michael.Burkey@Nexwarecorp.com> )
Nexware Corp. (> http://www.nexwarecorp.com> )
Software Engineer
865.546.9998 x201




“Hardware Support Account” <> hw@qnx.com> > wrote in message news:9mj3ok$ijb$> 2@nntp.qnx.com> …
Hi Michael,

Assuming the MediaGX board is not your development system.

Does the MediaGX board contain a PCI bus? If not then you
shouldn’t be running pci-bios in your OS image. Could you
give a bit more detail on your OS image and what the system
contains. We have several MediaGX boards here that run fine
with QNX RTOS 6.1.

Thanks.

Erick.


Michael D. Burkey <> michael.burkey@nexwarecorp.com> > wrote:
Has anyone out there recently used RTP 6.x on a MediaGX/Geode
platform running the InsydeSW XpressRom product?

When trying to boot this we get several warning messages about
PnP BIOS not being detected by QNX and then it starts trying to
detect the IDE device and locks up during the process.

Also of note, this is on a Geode SC1200 – the first one of which
I have tried to run RTP on.

I am aware that the current version of RTP has some serious
issues on x86 boxes without a proper PCI BIOS implementation
(which I feel is inexcusable for an embedded RTOS – QNX4
worked nicely in this situation and if RTP can’t, QNX will be
losing a very substantial market IMHO) but, at least from what
I can tell, this doesn’t seem to be a PCI BIOS issue (at least I’m
not getting the PCI Bios error message that I am used to getting).

Has anyone else out there had any luck with similar configurations?

\

Michael Burkey (mailto:> Michael.Burkey@Nexwarecorp.com> )
Nexware Corp. (> http://www.nexwarecorp.com> )
Software Engineer
865.546.9998 x201


\

Thanks.

Once we can get our software migrated from 6.0 to 6.1
I will be sure to test it out. Which, is quite frankly, proving
to be a bit of a bear due to the numerous architecture changes
involved.

Has there been any work done on testing a new version of
devc-ser8250 yet?


\

Michael Burkey (mailto:Michael.Burkey@Nexwarecorp.com)
Nexware Corp. (http://www.nexwarecorp.com)
Software Engineer
865.546.9998 x201



“Hardware Support Account” <hw@qnx.com> wrote in message news:9niq3d$au7$1@nntp.qnx.com

Hi,

There is a new devg-cyrix.so posted on the developers corner under New Stuff.
http://deverlopers.qnx.com

Regards,

Joe

Hardware Support Account <> hw@qnx.com> > wrote:
Hi Mike,

We are currently looking into getting a board with the
SC1200, so we may come up with a solution. We do
have an NDA with NatSemi, so cannot put the source in
the DDK.

Thanks,

Joe

Michael D. Burkey <> michael.burkey@nexwarecorp.com> > wrote:
I have now compared the datasheets for the video
controller config registers video input format between
the SC1200 and the CS5530 and they do NOT
match each other in 16bit format except for one mode.

Any hope of getting a new Devg-MediaGx in the future???

Or the source in the DDK so I can fix it myself?
(which I know is doubtful due to potential NDA issues w/NatSemi)


Michael Burkey (mailto:> Michael.Burkey@Nexwarecorp.com> )
Nexware Corp. (> http://www.nexwarecorp.com> )
Software Engineer
865.546.9998 x201


“Michael D. Burkey” <> michael.burkey@nexwarecorp.com> > wrote in message news:9mji44$9pt$> 1@inn.qnx.com> …
The plot thickens…

Well, here is what we have learned in probing the Geode SC1200.

  1. devc-8250 does not work properly and never returns.
    (as the 8250’s are now “virtual” this doesn’t suprise me)
  2. if we start USB printer support devu-ohci dies.
    I would guess this is because their may still be some hooks
    from back into the “real” parallel printer drives somewhere
    in the code and – guess what – the printer port on the SC1200
    is now virtualized just like the UARTs are…
  3. we got Photon working but it looks a bit “strange”
    In 16 bit color mode it looks almost palletized (grainy) but
    seems to function ok. However, the manner in which the
    overlay regions is handled in the SC1200 appears to have
    changed somewhat. At the very least, the colorspace mapping
    appears to be different from what it used to be. Devg-MediaGX
    is returning that the requested colorspace is available and it
    does work, however, the mapping is reversed for U and it looks
    like the V component is completely missing. I’ll check the data
    sheets for the SC1200 shortly and see what looks different.

So, is there any hope of seeing new drivers to fix these
problems in the near future? Or workarounds?


Michael Burkey (mailto:> Michael.Burkey@Nexwarecorp.com> )
Nexware Corp. (> http://www.nexwarecorp.com> )
Software Engineer
865.546.9998 x201




“Hardware Support Account” <> hw@qnx.com> > wrote in message news:9mj3ok$ijb$> 2@nntp.qnx.com> …
Hi Michael,

Assuming the MediaGX board is not your development system.

Does the MediaGX board contain a PCI bus? If not then you
shouldn’t be running pci-bios in your OS image. Could you
give a bit more detail on your OS image and what the system
contains. We have several MediaGX boards here that run fine
with QNX RTOS 6.1.

Thanks.

Erick.


Michael D. Burkey <> michael.burkey@nexwarecorp.com> > wrote:
Has anyone out there recently used RTP 6.x on a MediaGX/Geode
platform running the InsydeSW XpressRom product?

When trying to boot this we get several warning messages about
PnP BIOS not being detected by QNX and then it starts trying to
detect the IDE device and locks up during the process.

Also of note, this is on a Geode SC1200 – the first one of which
I have tried to run RTP on.

I am aware that the current version of RTP has some serious
issues on x86 boxes without a proper PCI BIOS implementation
(which I feel is inexcusable for an embedded RTOS – QNX4
worked nicely in this situation and if RTP can’t, QNX will be
losing a very substantial market IMHO) but, at least from what
I can tell, this doesn’t seem to be a PCI BIOS issue (at least I’m
not getting the PCI Bios error message that I am used to getting).

Has anyone else out there had any luck with similar configurations?

\

Michael Burkey (mailto:> Michael.Burkey@Nexwarecorp.com> )
Nexware Corp. (> http://www.nexwarecorp.com> )
Software Engineer
865.546.9998 x201




\