From: Peter Hurley Subject: Re: [PATCH v3 -next 11/11] serial: 8250_early: Remove setup_early_serial8250_console() Date: Fri, 03 Apr 2015 21:16:04 -0400 Message-ID: <551F3B54.3070707@hurleysoftware.com> References: <1425932842-21812-1-git-send-email-peter@hurleysoftware.com> <551CB5DF.1010108@hurleysoftware.com> <551D6EEA.6050704@hurleysoftware.com> <551E6D61.7000501@hurleysoftware.com> <551ED011.1030004@hurleysoftware.com> <551EDB92.9070008@hurleysoftware.com> Mime-Version: 1.0 Content-Type: text/plain; charset=utf-8 Content-Transfer-Encoding: 7bit Return-path: In-Reply-To: Sender: linux-kernel-owner@vger.kernel.org To: Yinghai Lu Cc: Andrew Morton , Greg Kroah-Hartman , Linux Kernel Mailing List , Jiri Slaby , Rob Herring , "linux-serial@vger.kernel.org" List-Id: linux-serial@vger.kernel.org On 04/03/2015 08:52 PM, Yinghai Lu wrote: > On Fri, Apr 3, 2015 at 11:27 AM, Peter Hurley wrote: >> On 04/03/2015 01:44 PM, Yinghai Lu wrote: >>> On Fri, Apr 3, 2015 at 10:38 AM, Peter Hurley wrote: >>>> On 04/03/2015 12:57 PM, Yinghai Lu wrote: >>>> >>>> Wait -- you have earlycon in a product?? >>> >>> What do you mean? >> >> I mean, what will happen if I put in a big debug banner like Steven >> did for ftrace? >> >> Kernel developers need earlycon for debugging arch code; often the earlycon >> is just hacked together especially when it requires fixmap support. >> >> Putting it in a product and _relying on undocumented behavior_ is a bad idea. > > let me repeat again: > when you have "console=uart8250,io,0x3f8", you will have earlycon and > then console. > > That is just for kernel developer for debugging. > > When you have x86 with bunch of dimms and cpus, user will have to wait couple > of minutes to output from console=ttyS0....and if it hang early, no > one would know what happen. > > Also it is only boot time only.... The fix is sitting in your inbox. Pointless to discuss it now.