From mboxrd@z Thu Jan  1 00:00:00 1970
From: Alessandro Rubini <rubini-list@gnudd.com>
Date: Tue, 26 Oct 2010 23:40:42 +0200
Subject: [U-Boot] [PATCH 3/3 v2] Replace CONFIG_SYS_GBL_DATA_SIZE by
 auto-generated value
In-Reply-To: <20101026213148.89A6F152451@gemini.denx.de>
References: <20101026213148.89A6F152451@gemini.denx.de>
	<20101026211233.BB83B152451@gemini.denx.de>
	<20101026203811.D0EF1152451@gemini.denx.de>
	<1288104730-25651-1-git-send-email-wd@denx.de>
	<1288101601-24871-4-git-send-email-wd@denx.de>
	<20101026195429.GA1349@morgana.i.gnudd.com>
	<20101026205756.GA2139@morgana.i.gnudd.com>
	<20101026211821.GA2528@morgana.i.gnudd.com>
Message-ID: <20101026214042.GA2897@morgana.i.gnudd.com>
List-Id: <u-boot.lists.denx.de>
MIME-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
Content-Transfer-Encoding: 7bit
To: u-boot@lists.denx.de

> I am not convinced, as we don't have an array context here.

But sizeof(struct x) doesn't depend on how struct x is used.
You can declare a pointer and then allocate for an array.

> I don't see that with
> 
> 	struct foo x;
> 	struct foo y[N];
> 
> we have a guarantee that sizeof(x) == sizeof(y[0]). 

Yes, I see this requirement. But maybe we're already bikeshedding:
I won't get offended if you keep the global +15 & ~15.

regards
/alessandro