From mboxrd@z Thu Jan 1 00:00:00 1970 From: "Enrico Weigelt, metux IT consult" Subject: Re: [PATCH V2 1/2] serial/8250: Add support for NI-Serial PXI/PXIe+485 devices Date: Thu, 4 Jul 2019 20:17:56 +0200 Message-ID: <9f08515a-3498-709e-d72b-538a54b704ad@metux.net> References: <20190703072435.34152-1-je.yen.tam@ni.com> <63697b10-98a7-86b7-2513-3eb793ed03c5@metux.net> <20190703144744.GB24961@kroah.com> <20190703155941.GB19567@kroah.com> Mime-Version: 1.0 Content-Type: text/plain; charset=utf-8 Content-Transfer-Encoding: 7bit Return-path: In-Reply-To: <20190703155941.GB19567@kroah.com> Content-Language: en-US Sender: linux-kernel-owner@vger.kernel.org To: Greg KH Cc: jeyentam , linux-serial@vger.kernel.org, linux-kernel@vger.kernel.org List-Id: linux-serial@vger.kernel.org On 03.07.19 17:59, Greg KH wrote: >> That's sad, because in embedded world we often have to care about >> code size, so making those devices optional would be of great help. > > Really? are you sure? Try it and see what you really end up saving. Just in 8250_pci.c alone, I could (depending on actual target) drop about 80% LoC. Haven't tested how much the difference in image size will really be. But if you really like to know and would consider taking some patches for making those things optional, I'll take some time and do the dirty work. By the way: would you still be opposed when these options are masked behind CONFIG_EXPERT or CONFIG_EMBEDDED ? I can imagine this could be a good compromise. --mtx -- Enrico Weigelt, metux IT consult Free software and Linux embedded engineering info@metux.net -- +49-151-27565287