Patch A devp-pccard and O2 Micro Controller Problem

QNX Newsgroups (Archives) qdn.public.qnxrtp.os
#1

Hello,

I posted a message two days ago regarding a problem with the O2 Micro Carbus
controller.

I have uninstalled the old RTP and reinstalled with patch A which I
downloaded today. I ran ‘uname -a’ and confirmed the patch was fully
installed as per the web recommendation.

The machine does not now lock if a 5V card is plugged in but the driver does
seem to do something very wrong with the voltage switching; I say this
because the resettable fuses supplying the sockets in my notebook trip when
the 5V card is activated. I can tell this because the card is not recognised
(pin returns empty sockets), and insertion of ANY other card (3.3V or 5V)
thereafter doesn’t work. The insertion and removal events are recognised by
devp-pccard (with -vv in invocation), but the card details are always blank.

The only way to clear this problem is to turn off the notebook to reset the
fuses. A simple restart won’t work.

If a power-down is not performed and I reboot to Windows 98 or Win 2K they
both fail to see any PC Cards when plugged in (due to loss of power to the
sockets). However, after a power-down and power-up cycle the Windows systems
correctly detect all cards again, and RTP correctly detects 3.3V cards.
However, if you dare to insert a 5V card again when running RTP, bang go the
fuses - another power-down job.

I will reiterate from the previous post that the Cardbus controller is an O2
Micro OZ6832. I notice that the release notes for patch A mention full
support for these now, although I thought the previous version claimed
claimed that anyway.

This strongly points towards a problem with devp-pccard in my view.

Can anyone help with this problem!

Thanks.

Len Barber.

#2

This is really weird! There is only one bit to turn on 5V and one bit
to turn on 3.3V! devp-pccard has been tested on the Acer notebooks that
have an O2 Micro 6832 CardBus controller and they work fine. What type
of notebook do you have?

Previously, Len Barber wrote in qdn.public.qnxrtp.os:

Hello,

I posted a message two days ago regarding a problem with the O2 Micro Carbus
controller.

I have uninstalled the old RTP and reinstalled with patch A which I
downloaded today. I ran ‘uname -a’ and confirmed the patch was fully
installed as per the web recommendation.

The machine does not now lock if a 5V card is plugged in but the driver does
seem to do something very wrong with the voltage switching; I say this
because the resettable fuses supplying the sockets in my notebook trip when
the 5V card is activated. I can tell this because the card is not recognised
(pin returns empty sockets), and insertion of ANY other card (3.3V or 5V)
thereafter doesn’t work. The insertion and removal events are recognised by
devp-pccard (with -vv in invocation), but the card details are always blank.

The only way to clear this problem is to turn off the notebook to reset the
fuses. A simple restart won’t work.

If a power-down is not performed and I reboot to Windows 98 or Win 2K they
both fail to see any PC Cards when plugged in (due to loss of power to the
sockets). However, after a power-down and power-up cycle the Windows systems
correctly detect all cards again, and RTP correctly detects 3.3V cards.
However, if you dare to insert a 5V card again when running RTP, bang go the
fuses - another power-down job.

I will reiterate from the previous post that the Cardbus controller is an O2
Micro OZ6832. I notice that the release notes for patch A mention full
support for these now, although I thought the previous version claimed
claimed that anyway.

This strongly points towards a problem with devp-pccard in my view.

Can anyone help with this problem!

Thanks.

Len Barber.
\

–

Hugh Brown (613) 591-0931 ext. 209 (voice)
QNX Software Systems Ltd. (613) 591-3579 (fax)
175 Terence Matthews Cres. email: hsbrown@qnx.com
Kanata, Ontario, Canada.
K2M 1W8

#3

The notebook is a UMAX ActionBook 850T.

This problem exists for all 5V cards I can get my hands on, not just a
select few. I am perfectly prepared to believe that there is something
slightly ‘different’ about the notebook’s power set-up around the CardBus
controller, but as I said before this problem does not exist in either
Windows 98 or Windows 2000, hence my conclusion that something devp-pccard
is doing in the card activation sequence is tripping the fuses.

Is there any chance that during the switching the 5V and 3.3V supplies are
connected together briefly (i.e. both bits on at once)? Does the driver turn
off both 3.3V and 5V before activating the one it beleives is appropriate? I
ask this just in case the BIOS in my notebook has left 3.3V active on the
sockets when nothing is inserted (I know it shouldn’t, but just guessing).

Regards,

Len Barber.

#4

Previously, Len Barber wrote in qdn.public.qnxrtp.os:

The notebook is a UMAX ActionBook 850T.

This problem exists for all 5V cards I can get my hands on, not just a
select few. I am perfectly prepared to believe that there is something
slightly ‘different’ about the notebook’s power set-up around the CardBus
controller, but as I said before this problem does not exist in either
Windows 98 or Windows 2000, hence my conclusion that something devp-pccard
is doing in the card activation sequence is tripping the fuses.

Is there any chance that during the switching the 5V and 3.3V supplies are
connected together briefly (i.e. both bits on at once)? Does the driver turn
off both 3.3V and 5V before activating the one it beleives is appropriate? I
ask this just in case the BIOS in my notebook has left 3.3V active on the
sockets when nothing is inserted (I know it shouldn’t, but just guessing).

Regards,

Len Barber.

Responded via e-mail.

–

Hugh Brown (613) 591-0931 ext. 209 (voice)
QNX Software Systems Ltd. (613) 591-3579 (fax)
175 Terence Matthews Cres. email: hsbrown@qnx.com
Kanata, Ontario, Canada.
K2M 1W8