From owner-freebsd-arm@freebsd.org Fri Jul 13 16:01:35 2018 Return-Path: Delivered-To: freebsd-arm@mailman.ysv.freebsd.org Received: from mx1.freebsd.org (mx1.freebsd.org [IPv6:2610:1c1:1:606c::19:1]) by mailman.ysv.freebsd.org (Postfix) with ESMTP id 85B041041EEC for ; Fri, 13 Jul 2018 16:01:35 +0000 (UTC) (envelope-from wlosh@bsdimp.com) Received: from mail-it0-x22d.google.com (mail-it0-x22d.google.com [IPv6:2607:f8b0:4001:c0b::22d]) (using TLSv1.2 with cipher ECDHE-RSA-AES128-GCM-SHA256 (128/128 bits)) (Client CN "smtp.gmail.com", Issuer "Google Internet Authority G2" (verified OK)) by mx1.freebsd.org (Postfix) with ESMTPS id 16D758C73C for ; Fri, 13 Jul 2018 16:01:35 +0000 (UTC) (envelope-from wlosh@bsdimp.com) Received: by mail-it0-x22d.google.com with SMTP id p17-v6so12196942itc.2 for ; Fri, 13 Jul 2018 09:01:35 -0700 (PDT) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=bsdimp-com.20150623.gappssmtp.com; s=20150623; h=mime-version:sender:in-reply-to:references:from:date:message-id :subject:to:cc; bh=M3kYmAF87uiUduimP87hXgkHMCHsYJIdxvOonnlVfcY=; b=vwRFllBrE29g4NYoV+XUEJIdWD55Gc7pdX5VoH/I7nlGyV8U4MkQ5kHjzz96JUj/YB zDo6vD6MS7Bxu1nCdmrynoIE1TrTHBFNCXS+cUBZg8xDui3I5X6W4VXTKPkSHsH3bD+h gUDDuGW9afjV1VSg8YdJUZPiCaC5wgEdCRPvbsIGLx8zXT0OSiB8L2sAiMjsUlwIZNzH 43BqyWGGXjG5YZ2YRKA2aGk41eOGAkMnUmKNkT3D97y3Rqy5PUrNiEkcfrfg7KTELnps rWGYqhGEeTITGexvTfA1Sw9PoTDKDEPraKGatk0/QOhf0gGUJ8T4ETSUuQLT9KZqnkGC NZxw== X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20161025; h=x-gm-message-state:mime-version:sender:in-reply-to:references:from :date:message-id:subject:to:cc; bh=M3kYmAF87uiUduimP87hXgkHMCHsYJIdxvOonnlVfcY=; b=JP2CLzNax617G8hrSu1+VEhNe2RpIrCpMvTORGTPlz0Q2jrkM/jLV4QjDzERmWiT2m DOn3THxIZomAvN0ZgBzt6L05pnskLgImV5+p+Y61bqI1boU2FmmuxHsRb+X20DpNbB/1 q14ZIUGMAlTJW2PWqxQijk/pd7Wvvl5AjMCUzZEHG/A6rPxtAwP/Xj6uyHvRmk3ZsYKP /ui0EDQv1epyq9/jruORP7oFDF/fj9cA6YBvsnfjzbDp0WowpWwoXOvoR73O7m0Al04h UZ7PwJQCecGLgmGBXtZzlU+coKPRM6R038YrV5S5udKBKpBo1ubNhZzW6qs2wXKXjPCY kUew== X-Gm-Message-State: AOUpUlHKWspWEQ4rLij+IlUB/2fbXhhEm3HhQLC48IBV+HR/GJ0u9u+9 1Br1A9aLctS1FurWFSw+ALj3LUhi1VTPxme3ro/qiw== X-Google-Smtp-Source: AAOMgpcB7UgUEAd7Fk1aVP9V6RcIht+u2UOHC6UZ9dqqCcPNMJkMN6DUcEusDQ1uiEzbfYMdFyQ6Wkpv+IWEzu4RwyY= X-Received: by 2002:a24:d50a:: with SMTP id a10-v6mr5285060itg.73.1531497694075; Fri, 13 Jul 2018 09:01:34 -0700 (PDT) MIME-Version: 1.0 Sender: wlosh@bsdimp.com Received: by 2002:a4f:1183:0:0:0:0:0 with HTTP; Fri, 13 Jul 2018 09:01:33 -0700 (PDT) X-Originating-IP: [2603:300b:6:5100:1052:acc7:f9de:2b6d] In-Reply-To: <1531496771.66719.50.camel@freebsd.org> References: <201807122111.w6CLBjP2080376@pdx.rh.CN85.dnsmgr.net> <20180713095616.GA60331@ci0.org> <1531496771.66719.50.camel@freebsd.org> From: Warner Losh Date: Fri, 13 Jul 2018 10:01:33 -0600 X-Google-Sender-Auth: xAgTm_lYPT3AXXDXW2-si8YQRzQ Message-ID: Subject: Re: Retiring two more: Atmel and Xscale To: Ian Lepore Cc: Olivier Houchard , "freebsd-arm@freebsd.org" , "Rodney W. Grimes" Content-Type: text/plain; charset="UTF-8" X-Content-Filtered-By: Mailman/MimeDel 2.1.27 X-BeenThere: freebsd-arm@freebsd.org X-Mailman-Version: 2.1.27 Precedence: list List-Id: "Porting FreeBSD to ARM processors." List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Fri, 13 Jul 2018 16:01:35 -0000 On Fri, Jul 13, 2018 at 9:46 AM, Ian Lepore wrote: > On Fri, 2018-07-13 at 11:56 +0200, Olivier Houchard wrote: > > Hi, > > > > On Thu, Jul 12, 2018 at 03:54:09PM -0600, Warner Losh wrote: > > > > > > On Thu, Jul 12, 2018 at 3:11 PM, Rodney W. Grimes < > > > freebsd-rwg@pdx.rh.cn85.dnsmgr.net> wrote: > > > > > > > > > > > > > > > > > I'd like to retire Atmel and Xscale support prior to 12. The > > > > > atmel stuff > > > > > has been dodgy for a long time and generally lacks the memory > > > > > to run > > > > > FreeBSD well. We have no current users of it. > > > > > > > > > > The Xscale stuff is old enough to not be relevant. There's no > > > > > known > > > > users. > > > > > > > > > > Part of it is already set to go with the big endian stuff, but > > > > > is there a > > > > > reason to keep the rest? > > > > I thought a depreciation policy was in process to help deal with > > > > all of these. > > > > > > > It will, but these two are no-brainers. The last known users of > > > both of > > > these sub-ports are done with them (Atmel since 8.2, Xscale maybe > > > longer). > > > The one Atmel product I'm aware of will never upgrade past 8.2. The > > > Xscale > > > stuff has not been used even longer in any serious way (since the 7 > > > timeframe, though there were one or two possible single users). > > > > > > If there's no objections, I'll just remove. If there are > > > objections, it > > > will be easy enough to do under that umbrella. > > > > > I think I'm the last known Xscale "user", and I would be more than > > glad > > to see it gone :) > > > > Regards, > > > > Olivier > > > > And I (well, $work) was the last known major user of atmel stuff. We > were never able to get any version of freebsd after 8.2 to be stable > enough on at91 hardware to ship products. That was after investing a > couple hundred hours in trying to track down and fix causes of panics > and disk data corruption in freebsd 9, 10, and 11 on our hardware. I > never tried 12, because $work hasn't been interested in investing any > more resources in such old hardware. > > I'm not convinced freebsd > 8 is really robust any armv4 or v5 systems, > but we do have plenty of reports of "it works for me", so we've been > doing what we can to keep it limping along for those folks. Afaik, that > community is now down to users of Marvell Kirkwood chips (Sheeva and > other *Plug systems). > I know mine have been robust enough to do only toy-level things. My Atmel 11-current board was my dnsmasq server for my development network for a while. USB didn't work, however, and I had to hack the ate driver to fix some issues. I could run it off the SD card or netboot it. The issues that Ian and I had talked about wrt unaligned I/O were present in USB but not mmc. I echo Ian's experience in that I saw all kinds of panics with this between 9-current and 10-current, however. I ran this way only a few weeks, and it wasn't as robust as just dropping in a Banana Pi A20 because the board I had sometimes wouldn't boot right (u-boot issues). I've run the Banana Pi since then w/o thinking about it other than making DNS changes on it. Warner