Code signing pcmcia drivers

Hello,

We've been having some problems with our PCMCIA driver since the leap to
Mobile 5.0. We now suspect it might be related to the new security features
of Mobile 5.0. It is currently not signed, however it appears to load fine
when the card is inserted and we also are receiving a valid handle when we
call CreateFile and success when we call WriteFile. Nothing is streaming to
or from the port though. We are beginning to think that our driver has
'partial trust' status and is not being able to access some API's. Is this
possible? Is there a way to verify this?

Another quick question regarding code signing: What am I suppose to do
during development when I compile and test my PCMCIA driver 50 times a day
on the target device? It could get very costly and time consuming to sign
code each time you wanted to test something on the target hardware.

Thomas

John>
Tue Nov 01 14:00:06 CST 2005

> We are beginning to think that our driver has 'partial trust' status and
> is not being able to access some API's. Is this possible? Is there a way
> to verify this?

As document
http://msdn.microsoft.com/library/default.asp?url=/library/en-us/guide_ppc/html/ppc_wce51conapplicationsecurityonmobiledevicesozup.asp
explains, it all depends on policies, which are set by OEMs.

On the PPC devices, which I have, all code, which is allowed to run,
runs with full trust in kernel mode. Call CeGetCurrentTrust to verify
the current trust.

> What am I suppose to do during development

Just sign your DLLs with one of "TEST ONLY ..." certificates,
which come with PPC and Smartphone SDK, then install it on the device
(to install the certificate copy SdkCerts.cab to the PPC, then invoke it
on PPC (using Explorer).

John

John>
Tue Nov 01 14:20:00 CST 2005

I correct myself.

The same document sais:
"Applications running on Pocket PC have access to all system resources."

So, if you driver is loaded, it should have full trust.

Thomas
Tue Nov 01 16:10:41 CST 2005

Hi John,

Well according to you, the documentation, and CeGetCurrentTrust() my driver
appears to have full trust. CeGetCurrentTrust() is returning the value:
"OEM_CERTIFY_TRUST (2) The OEM trusts the module to perform any OS function
calls." And the SdkCerts.cab is preventing the Unknown Publisher pop-up
when the CompactFlash card is inserted (which loads my driver).

This appears to rule out security issues with my pcmcia driver. Which I
don't know if this is good or bad since I'm back to square one as to why my
pcmcia stream driver isn't working with Mobile 5.0. Wish my kernel tracker
worked!!

Thanks for your help John.

Thomas





<John> wrote in message news:mamdnSPIgP5tUPreRVn-hA@speakeasy.net...
>I correct myself.
>
> The same document sais:
> "Applications running on Pocket PC have access to all system resources."
>
> So, if you driver is loaded, it should have full trust.
>

John>
Tue Nov 01 17:01:16 CST 2005

> Wish my kernel tracker worked!!

I am just curious, how the tracker would have helped to diagnose this
problem?

Thomas
Tue Nov 01 17:55:53 CST 2005


<John> wrote in message news:OuidnbpCpMQhbvreRVn-qg@speakeasy.net...
>> Wish my kernel tracker worked!!
>
> I am just curious, how the tracker would have helped to diagnose this
> problem?
>

I don't pretend to be an expert on the Kernel Tracker but I would definitely
see a difference (of my drivers interactions) between Mobile 2003 and Mobile
5.0.

Any more understanding can only be a good thing, since there is no DDK,
support, or documentation for Windows Mobile 5.0.

To quote a few features from the Kernel Tracker docs:

a.. Thread interactions
a.. Internal dependencies
a.. System state information
a.. System events, mapped onto the thread that was executing at the time
they occurred
a.. System interrupts
a.. All processes and threads in the system, including when these processes
and threads are:
a.. Created
a.. Run
a.. Stopped
a.. Sleeping

John>
Tue Nov 01 18:21:19 CST 2005

As you described your problem, your attempts to write to the card's
registers
don't produce any externally visible effect.
I don't see how tracking interrupts, events, threads, etc. will help in this
particular case.
I am not an expert on PC cards (debugged them just once or twice).
I wonder whether you can at least read your card's identification
information.

"Thomas Magma" <somewhere@overtherainbow.com> wrote in message
news:O6JtN$z3FHA.3460@TK2MSFTNGP12.phx.gbl...
>
> <John> wrote in message news:OuidnbpCpMQhbvreRVn-qg@speakeasy.net...
>>> Wish my kernel tracker worked!!
>>
>> I am just curious, how the tracker would have helped to diagnose this
>> problem?
>>
>
> I don't pretend to be an expert on the Kernel Tracker but I would
> definitely see a difference (of my drivers interactions) between Mobile
> 2003 and Mobile 5.0.
>
> Any more understanding can only be a good thing, since there is no DDK,
> support, or documentation for Windows Mobile 5.0.
>
> To quote a few features from the Kernel Tracker docs:
>
> a.. Thread interactions
> a.. Internal dependencies
> a.. System state information
> a.. System events, mapped onto the thread that was executing at the time
> they occurred
> a.. System interrupts
> a.. All processes and threads in the system, including when these
> processes and threads are:
> a.. Created
> a.. Run
> a.. Stopped
> a.. Sleeping
>
>

Thomas
Wed Nov 02 11:09:53 CST 2005

grrrrrrrrrr

I just noticed something about my pcmcia driver...my driver used the
cardmapwindow function in pcmcia.dll and CE 5.0 docs now says that it is not
supported and returns null.

How is a developer suppose to know when MS has changed something? Are the
changes to MS support dlls summarized and published anywhere?

Thomas

<John> wrote in message
news:W6SdnZLAU_Idm_XenZ2dnUVZ_sydnZ2d@speakeasy.net...
> As you described your problem, your attempts to write to the card's
> registers
> don't produce any externally visible effect.
> I don't see how tracking interrupts, events, threads, etc. will help in
> this
> particular case.
> I am not an expert on PC cards (debugged them just once or twice).
> I wonder whether you can at least read your card's identification
> information.
>
> "Thomas Magma" <somewhere@overtherainbow.com> wrote in message
> news:O6JtN$z3FHA.3460@TK2MSFTNGP12.phx.gbl...
>>
>> <John> wrote in message news:OuidnbpCpMQhbvreRVn-qg@speakeasy.net...
>>>> Wish my kernel tracker worked!!
>>>
>>> I am just curious, how the tracker would have helped to diagnose this
>>> problem?
>>>
>>
>> I don't pretend to be an expert on the Kernel Tracker but I would
>> definitely see a difference (of my drivers interactions) between Mobile
>> 2003 and Mobile 5.0.
>>
>> Any more understanding can only be a good thing, since there is no DDK,
>> support, or documentation for Windows Mobile 5.0.
>>
>> To quote a few features from the Kernel Tracker docs:
>>
>> a.. Thread interactions
>> a.. Internal dependencies
>> a.. System state information
>> a.. System events, mapped onto the thread that was executing at the time
>> they occurred
>> a.. System interrupts
>> a.. All processes and threads in the system, including when these
>> processes and threads are:
>> a.. Created
>> a.. Run
>> a.. Stopped
>> a.. Sleeping
>>
>>
>
>

Chris
Wed Nov 02 11:31:20 CST 2005

The PB 5.0 documentation covers what changed in the underlying OS.

--
Chris Tacke
Co-founder
OpenNETCF.org
Are you using the SDF? Let's do a case study.
Email us at d c s @ o p e n n e t c f . c o m
http://www.opennetcf.org/donate


"Thomas Magma" <somewhere@overtherainbow.com> wrote in message
news:uOOk8A93FHA.700@TK2MSFTNGP15.phx.gbl...
> grrrrrrrrrr
>
> I just noticed something about my pcmcia driver...my driver used the
> cardmapwindow function in pcmcia.dll and CE 5.0 docs now says that it is
> not supported and returns null.
>
> How is a developer suppose to know when MS has changed something? Are the
> changes to MS support dlls summarized and published anywhere?
>
> Thomas
>
> <John> wrote in message
> news:W6SdnZLAU_Idm_XenZ2dnUVZ_sydnZ2d@speakeasy.net...
>> As you described your problem, your attempts to write to the card's
>> registers
>> don't produce any externally visible effect.
>> I don't see how tracking interrupts, events, threads, etc. will help in
>> this
>> particular case.
>> I am not an expert on PC cards (debugged them just once or twice).
>> I wonder whether you can at least read your card's identification
>> information.
>>
>> "Thomas Magma" <somewhere@overtherainbow.com> wrote in message
>> news:O6JtN$z3FHA.3460@TK2MSFTNGP12.phx.gbl...
>>>
>>> <John> wrote in message news:OuidnbpCpMQhbvreRVn-qg@speakeasy.net...
>>>>> Wish my kernel tracker worked!!
>>>>
>>>> I am just curious, how the tracker would have helped to diagnose this
>>>> problem?
>>>>
>>>
>>> I don't pretend to be an expert on the Kernel Tracker but I would
>>> definitely see a difference (of my drivers interactions) between Mobile
>>> 2003 and Mobile 5.0.
>>>
>>> Any more understanding can only be a good thing, since there is no DDK,
>>> support, or documentation for Windows Mobile 5.0.
>>>
>>> To quote a few features from the Kernel Tracker docs:
>>>
>>> a.. Thread interactions
>>> a.. Internal dependencies
>>> a.. System state information
>>> a.. System events, mapped onto the thread that was executing at the time
>>> they occurred
>>> a.. System interrupts
>>> a.. All processes and threads in the system, including when these
>>> processes and threads are:
>>> a.. Created
>>> a.. Run
>>> a.. Stopped
>>> a.. Sleeping
>>>
>>>
>>
>>
>
>

Thomas
Wed Nov 02 12:27:01 CST 2005

Are you suppose to look for changes function by function? In the case of
CardMapWindow which changed, it doesn't show up in any summary in the PB 5.0
documentation. In fact there are several locations in the documentation that
still point to CardMapWindow as a valid function.

Thomas

"Chris Tacke, eMVP" <ctacke@spamfree-opennetcf.org> wrote in message
news:%236Ck%23M93FHA.3540@TK2MSFTNGP10.phx.gbl...
> The PB 5.0 documentation covers what changed in the underlying OS.
>
> --
> Chris Tacke
> Co-founder
> OpenNETCF.org
> Are you using the SDF? Let's do a case study.
> Email us at d c s @ o p e n n e t c f . c o m
> http://www.opennetcf.org/donate
>
>
> "Thomas Magma" <somewhere@overtherainbow.com> wrote in message
> news:uOOk8A93FHA.700@TK2MSFTNGP15.phx.gbl...
>> grrrrrrrrrr
>>
>> I just noticed something about my pcmcia driver...my driver used the
>> cardmapwindow function in pcmcia.dll and CE 5.0 docs now says that it is
>> not supported and returns null.
>>
>> How is a developer suppose to know when MS has changed something? Are the
>> changes to MS support dlls summarized and published anywhere?
>>
>> Thomas
>>
>> <John> wrote in message
>> news:W6SdnZLAU_Idm_XenZ2dnUVZ_sydnZ2d@speakeasy.net...
>>> As you described your problem, your attempts to write to the card's
>>> registers
>>> don't produce any externally visible effect.
>>> I don't see how tracking interrupts, events, threads, etc. will help in
>>> this
>>> particular case.
>>> I am not an expert on PC cards (debugged them just once or twice).
>>> I wonder whether you can at least read your card's identification
>>> information.
>>>
>>> "Thomas Magma" <somewhere@overtherainbow.com> wrote in message
>>> news:O6JtN$z3FHA.3460@TK2MSFTNGP12.phx.gbl...
>>>>
>>>> <John> wrote in message news:OuidnbpCpMQhbvreRVn-qg@speakeasy.net...
>>>>>> Wish my kernel tracker worked!!
>>>>>
>>>>> I am just curious, how the tracker would have helped to diagnose this
>>>>> problem?
>>>>>
>>>>
>>>> I don't pretend to be an expert on the Kernel Tracker but I would
>>>> definitely see a difference (of my drivers interactions) between Mobile
>>>> 2003 and Mobile 5.0.
>>>>
>>>> Any more understanding can only be a good thing, since there is no DDK,
>>>> support, or documentation for Windows Mobile 5.0.
>>>>
>>>> To quote a few features from the Kernel Tracker docs:
>>>>
>>>> a.. Thread interactions
>>>> a.. Internal dependencies
>>>> a.. System state information
>>>> a.. System events, mapped onto the thread that was executing at the
>>>> time they occurred
>>>> a.. System interrupts
>>>> a.. All processes and threads in the system, including when these
>>>> processes and threads are:
>>>> a.. Created
>>>> a.. Run
>>>> a.. Stopped
>>>> a.. Sleeping
>>>>
>>>>
>>>
>>>
>>
>>
>
>

John>
Wed Nov 02 12:48:02 CST 2005

Just a quick search for Mobile PC Card API on microsoft.com reveals
http://msdn.microsoft.com/library/default.asp?url=/library/en-us/wceddk5/html/wce50orideprecatedpcmciadriverconversiontobusagnosticmodel.asp
and
http://msdn.microsoft.com/library/default.asp?url=/library/en-us/wceddk5/html/wce50lrfPCCardDriverReference.asp
It has the description of the new API and a link
to the legacy API, which says:
CardMapWindow
This function is obsolete and not supported as of Windows CE 5.0.


"Thomas Magma" <somewhere@overtherainbow.com> wrote in message
news:e44hCs93FHA.1184@TK2MSFTNGP12.phx.gbl...
> Are you suppose to look for changes function by function? In the case of
> CardMapWindow which changed, it doesn't show up in any summary in the PB
> 5.0 documentation. In fact there are several locations in the
> documentation that still point to CardMapWindow as a valid function.
>
> Thomas
>
> "Chris Tacke, eMVP" <ctacke@spamfree-opennetcf.org> wrote in message
> news:%236Ck%23M93FHA.3540@TK2MSFTNGP10.phx.gbl...
>> The PB 5.0 documentation covers what changed in the underlying OS.
>>
>> --
>> Chris Tacke
>> Co-founder
>> OpenNETCF.org
>> Are you using the SDF? Let's do a case study.
>> Email us at d c s @ o p e n n e t c f . c o m
>> http://www.opennetcf.org/donate
>>
>>
>> "Thomas Magma" <somewhere@overtherainbow.com> wrote in message
>> news:uOOk8A93FHA.700@TK2MSFTNGP15.phx.gbl...
>>> grrrrrrrrrr
>>>
>>> I just noticed something about my pcmcia driver...my driver used the
>>> cardmapwindow function in pcmcia.dll and CE 5.0 docs now says that it is
>>> not supported and returns null.
>>>
>>> How is a developer suppose to know when MS has changed something? Are
>>> the changes to MS support dlls summarized and published anywhere?
>>>
>>> Thomas
>>>
>>> <John> wrote in message
>>> news:W6SdnZLAU_Idm_XenZ2dnUVZ_sydnZ2d@speakeasy.net...
>>>> As you described your problem, your attempts to write to the card's
>>>> registers
>>>> don't produce any externally visible effect.
>>>> I don't see how tracking interrupts, events, threads, etc. will help in
>>>> this
>>>> particular case.
>>>> I am not an expert on PC cards (debugged them just once or twice).
>>>> I wonder whether you can at least read your card's identification
>>>> information.
>>>>
>>>> "Thomas Magma" <somewhere@overtherainbow.com> wrote in message
>>>> news:O6JtN$z3FHA.3460@TK2MSFTNGP12.phx.gbl...
>>>>>
>>>>> <John> wrote in message news:OuidnbpCpMQhbvreRVn-qg@speakeasy.net...
>>>>>>> Wish my kernel tracker worked!!
>>>>>>
>>>>>> I am just curious, how the tracker would have helped to diagnose this
>>>>>> problem?
>>>>>>
>>>>>
>>>>> I don't pretend to be an expert on the Kernel Tracker but I would
>>>>> definitely see a difference (of my drivers interactions) between
>>>>> Mobile 2003 and Mobile 5.0.
>>>>>
>>>>> Any more understanding can only be a good thing, since there is no
>>>>> DDK, support, or documentation for Windows Mobile 5.0.
>>>>>
>>>>> To quote a few features from the Kernel Tracker docs:
>>>>>
>>>>> a.. Thread interactions
>>>>> a.. Internal dependencies
>>>>> a.. System state information
>>>>> a.. System events, mapped onto the thread that was executing at the
>>>>> time they occurred
>>>>> a.. System interrupts
>>>>> a.. All processes and threads in the system, including when these
>>>>> processes and threads are:
>>>>> a.. Created
>>>>> a.. Run
>>>>> a.. Stopped
>>>>> a.. Sleeping
>>>>>
>>>>>
>>>>
>>>>
>>>
>>>
>>
>>
>
>

Chris
Wed Nov 02 12:46:53 CST 2005

If you open PB 5.0 Help's TOC you'll see something like this:

- Welcome to CE 5.0
- Overview
- CE Architecture
- What's New in CE
+ tutorial
+ Getting Assistance
+ Migrating From an earl;ier version of CE
etc.

Lots of info in there. "What's New" and "Migrating" being most important
for what you're asking.

--
Chris Tacke
Co-founder
OpenNETCF.org
Are you using the SDF? Let's do a case study.
Email us at d c s @ o p e n n e t c f . c o m
http://www.opennetcf.org/donate


"Thomas Magma" <somewhere@overtherainbow.com> wrote in message
news:e44hCs93FHA.1184@TK2MSFTNGP12.phx.gbl...
> Are you suppose to look for changes function by function? In the case of
> CardMapWindow which changed, it doesn't show up in any summary in the PB
> 5.0 documentation. In fact there are several locations in the
> documentation that still point to CardMapWindow as a valid function.
>
> Thomas
>
> "Chris Tacke, eMVP" <ctacke@spamfree-opennetcf.org> wrote in message
> news:%236Ck%23M93FHA.3540@TK2MSFTNGP10.phx.gbl...
>> The PB 5.0 documentation covers what changed in the underlying OS.
>>
>> --
>> Chris Tacke
>> Co-founder
>> OpenNETCF.org
>> Are you using the SDF? Let's do a case study.
>> Email us at d c s @ o p e n n e t c f . c o m
>> http://www.opennetcf.org/donate
>>
>>
>> "Thomas Magma" <somewhere@overtherainbow.com> wrote in message
>> news:uOOk8A93FHA.700@TK2MSFTNGP15.phx.gbl...
>>> grrrrrrrrrr
>>>
>>> I just noticed something about my pcmcia driver...my driver used the
>>> cardmapwindow function in pcmcia.dll and CE 5.0 docs now says that it is
>>> not supported and returns null.
>>>
>>> How is a developer suppose to know when MS has changed something? Are
>>> the changes to MS support dlls summarized and published anywhere?
>>>
>>> Thomas
>>>
>>> <John> wrote in message
>>> news:W6SdnZLAU_Idm_XenZ2dnUVZ_sydnZ2d@speakeasy.net...
>>>> As you described your problem, your attempts to write to the card's
>>>> registers
>>>> don't produce any externally visible effect.
>>>> I don't see how tracking interrupts, events, threads, etc. will help in
>>>> this
>>>> particular case.
>>>> I am not an expert on PC cards (debugged them just once or twice).
>>>> I wonder whether you can at least read your card's identification
>>>> information.
>>>>
>>>> "Thomas Magma" <somewhere@overtherainbow.com> wrote in message
>>>> news:O6JtN$z3FHA.3460@TK2MSFTNGP12.phx.gbl...
>>>>>
>>>>> <John> wrote in message news:OuidnbpCpMQhbvreRVn-qg@speakeasy.net...
>>>>>>> Wish my kernel tracker worked!!
>>>>>>
>>>>>> I am just curious, how the tracker would have helped to diagnose this
>>>>>> problem?
>>>>>>
>>>>>
>>>>> I don't pretend to be an expert on the Kernel Tracker but I would
>>>>> definitely see a difference (of my drivers interactions) between
>>>>> Mobile 2003 and Mobile 5.0.
>>>>>
>>>>> Any more understanding can only be a good thing, since there is no
>>>>> DDK, support, or documentation for Windows Mobile 5.0.
>>>>>
>>>>> To quote a few features from the Kernel Tracker docs:
>>>>>
>>>>> a.. Thread interactions
>>>>> a.. Internal dependencies
>>>>> a.. System state information
>>>>> a.. System events, mapped onto the thread that was executing at the
>>>>> time they occurred
>>>>> a.. System interrupts
>>>>> a.. All processes and threads in the system, including when these
>>>>> processes and threads are:
>>>>> a.. Created
>>>>> a.. Run
>>>>> a.. Stopped
>>>>> a.. Sleeping
>>>>>
>>>>>
>>>>
>>>>
>>>
>>>
>>
>>
>
>