From mboxrd@z Thu Jan  1 00:00:00 1970
From: arnd@arndb.de (Arnd Bergmann)
Date: Tue, 29 Apr 2014 13:18:30 +0200
Subject: [PATCH v4 5/5] ARM: DRA7: Add support for soc_is_dra74x() and
 soc_is_dra72x() varients
In-Reply-To: <1398769513-8736-6-git-send-email-rnayak@ti.com>
References: <1398769513-8736-1-git-send-email-rnayak@ti.com>
 <1398769513-8736-6-git-send-email-rnayak@ti.com>
Message-ID: <4792826.emf6IohHd6@wuerfel>
To: linux-arm-kernel@lists.infradead.org
List-Id: linux-arm-kernel.lists.infradead.org

On Tuesday 29 April 2014 16:35:13 Rajendra Nayak wrote:
> @@ -393,7 +395,12 @@ IS_OMAP_TYPE(3430, 0x3430)
>  
>  #if defined(CONFIG_SOC_DRA7XX)
>  #undef soc_is_dra7xx
> +#undef soc_is_dra74x
> +#undef soc_is_dra72x
>  #define soc_is_dra7xx()        (of_machine_is_compatible("ti,dra7"))
> +#define soc_is_dra74x()        (of_machine_is_compatible("ti,dra74"))
> +#define soc_is_dra72x()        (of_machine_is_compatible("ti,dra72"))
> +

You shouldn't normally have to define these. Why are they needed?

Maybe it's better to wait for a user to show up, and then we can decide
whether we actually want to have them this way, or if there is a better
solution for the particular use case.

Normally, we'd want to make run-time decisions based on properties
of the nodes a driver is working on, not the global machine compatible
string.

	Arnd