From mboxrd@z Thu Jan 1 00:00:00 1970 From: David Brownell Subject: Re: [PATCH 11/9] move twl4030-gpio to drivers/gpio Date: Sat, 27 Sep 2008 22:16:14 -0700 Message-ID: <200809272216.14672.david-b@pacbell.net> References: <1222427996-7018-1-git-send-email-felipe.balbi@nokia.com> <200809271645.20782.david-b@pacbell.net> <20080928031411.GF8468@frodo> Mime-Version: 1.0 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: 7bit Return-path: Received: from smtp126.sbc.mail.sp1.yahoo.com ([69.147.65.185]:40629 "HELO smtp126.sbc.mail.sp1.yahoo.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with SMTP id S1750818AbYI1FQQ (ORCPT ); Sun, 28 Sep 2008 01:16:16 -0400 In-Reply-To: <20080928031411.GF8468@frodo> Content-Disposition: inline Sender: linux-omap-owner@vger.kernel.org List-Id: linux-omap@vger.kernel.org To: me@felipebalbi.com Cc: Felipe Balbi , linux-omap@vger.kernel.org, Tony Lindgren , Jean Delvare On Saturday 27 September 2008, Felipe Balbi wrote: > On Sat, Sep 27, 2008 at 04:45:20PM -0700, David Brownell wrote: > > On Saturday 27 September 2008, Felipe Balbi wrote: > > > We need something like the patch below (I can't find a substitute > > > to twl4030_set_gpio_edge_ctrl and twl4030_set_gpio_debounce): > > > > Something like that, yes. I have other patches queued up though. > > > > And meanwhile, your build seems goofed ... fix that first. > > Yeah, it only happens when both mmc and twl4030-gpio are built as > modules. Still sounds weird since we have EXPORT_SYMBOL() for those > twl-specific crap. Don't yet build twl4030 as a module. The vmlinux object can't link against symbols in other objects. ;) As I noted, the modular option is for build testing; when this goes to mainline, this code should be able to build on most any platform. (Not just OMAP; not just ARM.) I don't think any of the OMAP boards which actually use this code will be able to use that as a module, though. When hsmmc.c behaves again, I've got patches to get rid of its use of those twl-specific symbols. Then that linkage issue will go away. > > For debouncing, see this LKML RFC: > > > > http://marc.info/?l=linux-kernel&m=122228602206108&w=2 > > the debouncing mechanism looks really good :-) Kind of brainless ... except for the observation that we pretty much need one of those. ;) > That's true. Didn't got that at first look. We should use something > like: > > return gpio_get_value_cansleep(irq_to_gpio(irq)); > > right ?? I want to avoid the irq_to_gpio() bit. Eventually the whole MMC/SD initialization needs reworking ... I doubt Overo will be the only board using one SD for storage, and an SDIO for WLAN. > it's good that you've been playing with twl4030-gpio. You've written > gpiolib anyways, so you might better than I do where stuff should go. Or in some cases: where it's about time to add mechanisms to gpiolib. ;) - dave