From mboxrd@z Thu Jan 1 00:00:00 1970 From: jjhiblot@gmail.com (Jean-Jacques Hiblot) Date: Mon, 03 Mar 2014 09:55:58 +0100 Subject: [PATCH] irq: Changed the return type of irq_chip.irq_startup() from unsigned int to int In-Reply-To: <20140301160114.GE21483@n2100.arm.linux.org.uk> References: <1393521679-32331-1-git-send-email-jjhiblot@traphandler.com> <7150036.UpHTd1PK1K@radagast> <20140301160114.GE21483@n2100.arm.linux.org.uk> Message-ID: <5314439E.6070908@gmail.com> To: linux-arm-kernel@lists.infradead.org List-Id: linux-arm-kernel.lists.infradead.org Le 01/03/2014 17:01, Russell King - ARM Linux a ?crit : > On Sat, Mar 01, 2014 at 04:31:13PM +0100, Thomas Gleixner wrote: >> On Fri, 28 Feb 2014, James Hogan wrote: >>> On Thursday 27 February 2014 18:21:19 Jean-Jacques Hiblot wrote: >>>> drivers/irqchip/irq-metag.c | 4 ++-- >>> >>> I think you've missed out drivers/irqchip/irq-metag-ext.c >> >> There is a good reason why I asked to do this with coccinelle. > > Go easy, not everyone understands that complex tool - I certainly don't. > I ended up deleting it from my system after several failed attempts, and > I came to the conclusion that to use it properly, you also needed a Julia > installed along side you to learn how to write its scripts. > > The big problem is it /doesn't/ do what it advertises to - I ended up > writing expression after expression to make it detect all the various > forms of what I wanted it to, which as I understand the tool is precisely > the opposite of what you're supposed to do with it. > > I ended up coming to the conclusion that a well formed sed and/or grep > expression was far more useful and effective than trying to figure out > coccinelle. > Actually I tried to use coccinelle for this, but to paraphrase Thomas: it was a failure on its own. The tool is complex. I'll send shortly a second version of the patch with the missing change(s). Jean-Jacques