STAMPINF and INF2CAT

I’m having a problem with STAMPINF and INF2CAT.

I create an INF file containing the $ARCH$ symbol for substitution by STAMPINF.

STAMPINF replaces symbol $ARCH$ with either x86 or x64 depending on os.
I’m specifying: os:6_3_X64,Server6_3_X64,8_X64,Server8_X64
so I get x64 substituted for $ARCH$.

INF2CAT then declares that the INF does not have NTAMD64 decorated model sections.

Any ideas?

xxxxx@compuserve.com wrote:

I’m having a problem with STAMPINF and INF2CAT.

I create an INF file containing the $ARCH$ symbol for substitution by STAMPINF.

STAMPINF replaces symbol $ARCH$ with either x86 or x64 depending on os.
I’m specifying: os:6_3_X64,Server6_3_X64,8_X64,Server8_X64
so I get x64 substituted for $ARCH$.

No, you don’t. You want AMD64. If you are running STAMPINF in a batch
file in of a WDK build environment, you can leave off the “-a” parameter
altogether. It will read the _BUILDARCH environment variable, which is
either x86 or AMD64. If you are running it standalone outside of WDK,
then you need to specify either “-a x86” or “-a AMD64”.

The AMD64 vs x64 confusion is a historical anomaly. In the early days,
Intel bet its money on the Itanium, which was “ia64”. AMD came up with
their 64-bit architecture, which they called “amd64”. Those were the
terms used in the DDKs of the time.

As the years went by, Itanium died and Intel adopted the amd64
architecture as em64t. Eventually, Intel complained about the use of
the term “amd64” for what was now a more generic architecture, so some
of the tools started switching to x64. The INF format, however, is
steeped in the mists of antiquity. That’s why it is still valid to say
Signature=$Chicago$
even though Chicago, which was the code name for the original release of
Windows 95, ceased to be relevant more than 20 years ago.

INF2CAT then declares that the INF does not have NTAMD64 decorated model sections.

Right, because it doesn’t.


Tim Roberts, xxxxx@probo.com
Providenza & Boekelheide, Inc.

Tim Roberts wrote:

xxxxx@compuserve.com wrote:

It occurred to me just as a hit “send” that it was a bit ironic for me
to be poking fun at antiquated computer concepts in an email addressed
to someone at “compuserve.com”.


Tim Roberts, xxxxx@probo.com
Providenza & Boekelheide, Inc.

In fact you can put any old rubbish with the STAMPINF -a parameter.
-a anyoldrubbish
produces:
[Manufacturer]
%Tape_Devices%=Tape_Devices,NTanyoldrubbish

The STAMPINF documentation doesn’t mention anyoldrubbish - just any old rubbish as follows:

-a [architecture]
The values for the architecture string are x86, 64 (for Itanium-based platforms), and x64 (for amd64 platforms).

Glad someone remembers Compuserve.

Richard Jenkins wrote:

Glad someone remembers Compuserve.

CB simulator!!

xxxxx@compuserve.com wrote:

In fact you can put any old rubbish with the STAMPINF -a parameter.
-a anyoldrubbish
produces:
[Manufacturer]
%Tape_Devices%=Tape_Devices,NTanyoldrubbish

The STAMPINF documentation doesn’t mention anyoldrubbish - just any old rubbish as follows:

-a [architecture]
The values for the architecture string are x86, 64 (for Itanium-based platforms), and x64 (for amd64 platforms).

Does it really say that? Where? If so, that’s a rather severe bug.
The architecture string must be “amd64” for AMD64 platforms. That’s
dictated by the INF syntax.

Glad someone remembers Compuserve.

I was 75310,2475. For all I know, I can still be reached there.


Tim Roberts, xxxxx@probo.com
Providenza & Boekelheide, Inc.

Tim Roberts wrote:

xxxxx@compuserve.com wrote:
> In fact you can put any old rubbish with the STAMPINF -a parameter.
> -a anyoldrubbish
> produces:
> [Manufacturer]
> %Tape_Devices%=Tape_Devices,NTanyoldrubbish
>
> The STAMPINF documentation doesn’t mention anyoldrubbish - just any old rubbish as follows:
>
> -a [architecture]
> The values for the architecture string are x86, 64 (for Itanium-based platforms), and x64 (for amd64 platforms).
Does it really say that? Where? If so, that’s a rather severe bug.
The architecture string must be “amd64” for AMD64 platforms. That’s
dictated by the INF syntax.

Further, it must be “ia64” for Itanium platforms, not “64”.


Tim Roberts, xxxxx@probo.com
Providenza & Boekelheide, Inc.

This is where I read the STAMPINF documentation:

http://msdn.microsoft.com/en-us/library/windows/hardware/ff552789(v=vs.85).aspx

71477,2703 … either that WAS my CIS number or it’s my credit card number and it doesn’t have a comma. But I’m thinking CIS number, based on Tim’s pattern.

Peter
OSR
@OSRDrivers

xxxxx@compuserve.com wrote:

This is where I read the STAMPINF documentation:

http://msdn.microsoft.com/en-us/library/windows/hardware/ff552789(v=vs.85).aspx

I’ve submitted feedback to get the page corrected.


Tim Roberts, xxxxx@probo.com
Providenza & Boekelheide, Inc.

72617,2636

That rolls off my fingertips automatically just like my best friend’s phone
number from childhood.

Dave Cattley