From mboxrd@z Thu Jan  1 00:00:00 1970
From: Jason Gunthorpe <jgunthorpe-ePGOBjL8dl3ta4EC/59zMFaTQe2KTcn/@public.gmane.org>
Subject: Re: [PATCH v2 0/3] TPM2.0: Added eventlog support for
 TPM2.0
Date: Thu, 18 Aug 2016 13:55:21 -0600
Message-ID: <20160818195521.GC3676@obsidianresearch.com>
References: <1470771295-15680-1-git-send-email-nayna@linux.vnet.ibm.com>
	<20160810113243.GF13929@intel.com>
	<20160810171900.GA11543@intel.com>
	<57AC5802.1090109@linux.vnet.ibm.com>
	<20160813025915.GC26929@obsidianresearch.com>
	<57B2D429.8050508@linux.vnet.ibm.com>
Mime-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
Content-Transfer-Encoding: 7bit
Return-path: <tpmdd-devel-bounces-5NWGOfrQmneRv+LV9MX5uipxlwaOVQ5f@public.gmane.org>
Content-Disposition: inline
In-Reply-To: <57B2D429.8050508-23VcF4HTsmIX0ybBhKVfKdBPR1lH4CV8@public.gmane.org>
List-Unsubscribe: <https://lists.sourceforge.net/lists/listinfo/tpmdd-devel>,
	<mailto:tpmdd-devel-request-5NWGOfrQmneRv+LV9MX5uipxlwaOVQ5f@public.gmane.org?subject=unsubscribe>
List-Archive: <http://sourceforge.net/mailarchive/forum.php?forum_name=tpmdd-devel>
List-Post: <mailto:tpmdd-devel-5NWGOfrQmneRv+LV9MX5uipxlwaOVQ5f@public.gmane.org>
List-Help: <mailto:tpmdd-devel-request-5NWGOfrQmneRv+LV9MX5uipxlwaOVQ5f@public.gmane.org?subject=help>
List-Subscribe: <https://lists.sourceforge.net/lists/listinfo/tpmdd-devel>,
	<mailto:tpmdd-devel-request-5NWGOfrQmneRv+LV9MX5uipxlwaOVQ5f@public.gmane.org?subject=subscribe>
Errors-To: tpmdd-devel-bounces-5NWGOfrQmneRv+LV9MX5uipxlwaOVQ5f@public.gmane.org
To: Nayna <nayna-23VcF4HTsmIX0ybBhKVfKdBPR1lH4CV8@public.gmane.org>
Cc: tpmdd-devel-5NWGOfrQmneRv+LV9MX5uipxlwaOVQ5f@public.gmane.org
List-Id: tpmdd-devel@lists.sourceforge.net

On Tue, Aug 16, 2016 at 02:21:53PM +0530, Nayna wrote:

> I tried the suggested approach and since ACPI specific functions won't be
> available for arch using CONFIG_OF, so the compilation fails and vice versa
> for CONFIG_ACPI..

Right, you need to stub out the read_X function with an empty inline
when not available.

> Jason, for understanding.. can you explain the issue with existing design
> for read_log, where arch specific read_log is compiled based on CONFIG
> option.. And then, we can just change in Makefile.. where currently it is
> like
> 
> ifdef CONFIG_ACPI
>        tpm-y += tpm_acpi.o
> else
> ifdef CONFIG_OF
>        tpm-y += tpm_of.o
> endif
> endif

The 'else' is the main issue

 
> this can be changed to
> tpm-$(CONFIG_ACPI) += tpm_acpi.o
> tpm-$(CONFIG_OF) += tpm_of.o

This is fine by me, *however* only if the .o files do not define
overlapping symbols as they do today (eg tpm_acpi contains read_log_acpi)

Use some small #ifdef logic in the header file to create the inline
stub.

Jason

------------------------------------------------------------------------------