From mboxrd@z Thu Jan 1 00:00:00 1970 From: Heikki Krogerus Subject: Re: [PATCH v3 01/21 resend] serial: rewrite pxa2xx-uart to use 8250_core Date: Thu, 19 Dec 2013 13:05:02 +0200 Message-ID: <20131219110502.GC20686@xps8300> References: <1386901645-28895-1-git-send-email-ynvich@gmail.com> <1387309071-22382-1-git-send-email-ynvich@gmail.com> <1387309071-22382-2-git-send-email-ynvich@gmail.com> <20131219085145.GB20686@xps8300> <1387445713.31516.30.camel@host5.omatika.ru> <1387447296.31516.37.camel@host5.omatika.ru> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Return-path: Received: from mga01.intel.com ([192.55.52.88]:63661 "EHLO mga01.intel.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752165Ab3LSLFK (ORCPT ); Thu, 19 Dec 2013 06:05:10 -0500 Content-Disposition: inline In-Reply-To: <1387447296.31516.37.camel@host5.omatika.ru> Sender: linux-serial-owner@vger.kernel.org List-Id: linux-serial@vger.kernel.org To: Sergei Ianovich Cc: linux-kernel@vger.kernel.org, linux-arm-kernel@lists.infradead.org, Greg Kroah-Hartman , Russell King , Jiri Slaby , Grant Likely , Rob Herring , James Cameron , Zhou Zhu , Haojian Zhuang , Arnd Bergmann , "open list:SERIAL DRIVERS" , "open list:OPEN FIRMWARE AND..." On Thu, Dec 19, 2013 at 02:01:36PM +0400, Sergei Ianovich wrote: > On Thu, 2013-12-19 at 13:35 +0400, Sergei Ianovich wrote: > > On Thu, 2013-12-19 at 10:51 +0200, Heikki Krogerus wrote: > > > I noticed one more thing. I'm sorry about commenting this late. > > > > > > Since you set the type PORT_XSCALE, don's you want to use > > > UPF_FIXED_TYPE flag instead of the UPF_SKIP_TEST here? Otherwise the > > > type will just get overridden and autoconfig() will be executed. > > > > UPF_FIXED_PORT was my first guess. However, it didn't work. > > > > I've also tried to do ioremap in serial_pxa_probe instead of using > > UPF_IOREMAP. This didn't work as well. > > I didn't clarify properly. > > UPF_SKIP_TEST skips only the loopback test. This test could work on PXA, > but requires PXA-specific pre-initialization. There are more test after > the loopback one, which will fail if we are not on PXA. So I decided > that skipping it is easier than tricky rewriting of autoconfig(). OK. So there is no problem here. Sorry about the noise. Thanks, -- heikki