From mboxrd@z Thu Jan 1 00:00:00 1970 From: thomas.petazzoni@free-electrons.com (Thomas Petazzoni) Date: Wed, 4 Jun 2014 11:40:06 +0200 Subject: [PATCH] ARM: orion5x: fix mvebu_mbus_dt_init call In-Reply-To: <4280897.V3UkQjotOL@wuerfel> References: <4280897.V3UkQjotOL@wuerfel> Message-ID: <20140604114006.476cef06@free-electrons.com> To: linux-arm-kernel@lists.infradead.org List-Id: linux-arm-kernel.lists.infradead.org Dear Arnd Bergmann, On Wed, 04 Jun 2014 11:06:40 +0200, Arnd Bergmann wrote: > The prototype for mvebu_mbus_dt_init() changed around the same time > as a new caller was added to orion5x. This adds the missing argument > to make orion5x behave correctly to avoid this build error: > > mach-orion5x/board-dt.c: In function 'orion5x_dt_init': > mach-orion5x/board-dt.c:48:2: error: too few arguments to function 'mvebu_mbus_dt_init' > BUG_ON(mvebu_mbus_dt_init()); > ^ > In file included from mach-orion5x/board-dt.c:18:0: > include/linux/mbus.h:76:5: note: declared here > int mvebu_mbus_dt_init(bool is_coherent); > ^ > > Signed-off-by: Arnd Bergmann Acked-by: Thomas Petazzoni > Thomas, I assume you were aware of this problem since you > wrote the two patches that clashed, but for some reason we > screwed it up when merging them into arm-soc. Should we apply > this directly as a bug fix to arm-soc or did you already have > another solution planned? Yes, I am indeed aware of this change. I think I did a patch to fix it at some point, but maybe it got lost in the merging process, or simply I don't remember correctly what happened. It's indeed caused by an interaction between changes in the mvebu-mbus driver and the conversion of mach-orion5x to DT. Your patch is the correct solution. Thanks a lot! Thomas -- Thomas Petazzoni, CTO, Free Electrons Embedded Linux, Kernel and Android engineering http://free-electrons.com