Peter De Schrijver wrote at Monday, October 10, 2011 2:01 AM:

On Fri, Oct 07, 2011 at 08:04:34PM +0200, Stephen Warren wrote:

quoted

Peter De Schrijver wrote at Monday, October 03, 2011 7:19 AM:
...

quoted

diff --git a/arch/arm/mach-tegra/board-dt.c b/arch/arm/mach-tegra/board-dt.c

...

quoted

-	if (of_machine_is_compatible("nvidia,harmony"))
-		harmony_pinmux_init();
-	else if (of_machine_is_compatible("nvidia,seaboard"))
-		seaboard_pinmux_init();
+	for (i = 0; i < ARRAY_SIZE(pinmux_configs); i++) {
+		if (of_machine_is_compatible(pinmux_configs[i].machine)) {
+			pinmux_configs[i].init();
+			break;
+		}
+	}
+
+	if (i == ARRAY_SIZE(pinmux_configs))
+		printk(KERN_WARNING "Unknown platform! Pinmuxing not initialized\n");

Should that be WARN() instead of printk?

pr_warning() might be better. Or do we want a backtrace and a tainted kernel
in this case?

Well, tainting seems like a good idea. I'm less thrilled about the backtrace,
but perhaps it is a good idea to bring attention to this condition.

-- 
nvpublic