From mboxrd@z Thu Jan 1 00:00:00 1970 From: robherring2@gmail.com (Rob Herring) Date: Tue, 07 Feb 2012 16:52:14 -0600 Subject: [PATCH 2/3] ARM: make disable_fiq macro optional In-Reply-To: References: <1328562200-8810-1-git-send-email-robherring2@gmail.com> <1328562200-8810-3-git-send-email-robherring2@gmail.com> <4F30B225.60605@gmail.com> Message-ID: <4F31AB1E.4010800@gmail.com> To: linux-arm-kernel@lists.infradead.org List-Id: linux-arm-kernel.lists.infradead.org On 02/06/2012 11:24 PM, Nicolas Pitre wrote: > On Mon, 6 Feb 2012, Rob Herring wrote: > >> Nicolas, >> >> On 02/06/2012 04:34 PM, Nicolas Pitre wrote: >>> On Mon, 6 Feb 2012, Rob Herring wrote: >>> >>>> From: Rob Herring >>>> >>>> Only rpc uses disable_fiq, so make its use conditional on ARCH_RPC. >>>> >>>> With this, entry-macro.S is no longer needed for platforms that >>>> select MULTI_IRQ_HANDLER and the include of it can be conditional. >>> >>> Using set_fiq_handler() in the rpc boot code we could get rid of >>> disable_fiq entirely. >>> >> >> That's already being done here: >> arch/arm/mach-rpc/dma.c:290: set_fiq_handler(fiqhandler_start, >> fiqhandler_length); >> >> Can the macro just be deleted? Or a default handler needs to be setup >> earlier? > > That's the handler installed by the RPC DMA code. Another one is needed > by default for when that code is not active. Wouldn't the fiq be masked then? rpc_init_irq masks out the interrupts in the same register as disable_fiq macro: iomd_writeb(0, IOMD_FIQMASK); Rob