CINXE.COM

<?xml version="1.0" encoding="UTF-8" standalone="yes"?> <!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd"><html xmlns="http://www.w3.org/1999/xhtml"><head><meta http-equiv="Content-Type" content="text/html; charset=UTF-8" /><title>LKML: Matthias Andree: Re: RFC: Starting a stable kernel series off the 2.6 kernel</title><link href="/css/message.css" rel="stylesheet" type="text/css" /><link href="/css/wrap.css" rel="alternate stylesheet" type="text/css" title="wrap" /><link href="/css/nowrap.css" rel="stylesheet" type="text/css" title="nowrap" /><link href="/favicon.ico" rel="shortcut icon" /><script src="/js/simple-calendar.js" type="text/javascript"></script><script src="/js/styleswitcher.js" type="text/javascript"></script><link rel="alternate" type="application/rss+xml" title="lkml.org : last 100 messages" href="/rss.php" /><link rel="alternate" type="application/rss+xml" title="lkml.org : last messages by Matthias Andree" href="/groupie.php?aid=29" /><script> var _paq = window._paq = window._paq || []; /* tracker methods like "setCustomDimension" should be called before "trackPageView" */ _paq.push(["setDoNotTrack", true]); _paq.push(["disableCookies"]); _paq.push(['trackPageView']); _paq.push(['enableLinkTracking']); (function() { var u="//m.lkml.org/"; _paq.push(['setTrackerUrl', u+'matomo.php']); _paq.push(['setSiteId', '1']); var d=document, g=d.createElement('script'), s=d.getElementsByTagName('script')[0]; g.async=true; g.src=u+'matomo.js'; s.parentNode.insertBefore(g,s); })(); </script></head><body onload="es.jasper.simpleCalendar.init();" itemscope="itemscope" itemtype="http://schema.org/BlogPosting"><table border="0" cellpadding="0" cellspacing="0"><tr><td width="180" align="center"><a href="/"><img style="border:0;width:135px;height:32px" src="/images/toprowlk.gif" alt="lkml.org" /></a></td><td width="32">聽</td><td class="nb"><div><a class="nb" href="/lkml"> [lkml]</a> 聽 <a class="nb" href="/lkml/2005"> [2005]</a> 聽 <a class="nb" href="/lkml/2005/12"> [Dec]</a> 聽 <a class="nb" href="/lkml/2005/12/3"> [3]</a> 聽 <a class="nb" href="/lkml/last100"> [last100]</a> 聽 <a href="/rss.php"><img src="/images/rss-or.gif" border="0" alt="RSS Feed" /></a></div><div>Views: <a href="#" class="nowrap" onclick="setActiveStyleSheet('wrap');return false;">[wrap]</a><a href="#" class="wrap" onclick="setActiveStyleSheet('nowrap');return false;">[no wrap]</a> 聽 <a class="nb" href="/lkml/mheaders/2005/12/3/80" onclick="this.href='/lkml/headers'+'/2005/12/3/80';">[headers]</a>聽 <a href="/lkml/bounce/2005/12/3/80">[forward]</a>聽 </div></td><td width="32">聽</td></tr><tr><td valign="top"><div class="es-jasper-simpleCalendar" baseurl="/lkml/"></div><div class="threadlist">Messages in this thread</div><ul class="threadlist"><li class="root"><a href="/lkml/2005/12/3/55">First message in thread</a></li><li><a href="/lkml/2005/12/3/59">Arjan van de Ven</a><ul><li><a href="/lkml/2005/12/3/63">Adrian Bunk</a><ul><li><a href="/lkml/2005/12/3/65">Arjan van de Ven</a><ul><li><a href="/lkml/2005/12/4/37">Denis Vlasenko</a></li><li><a href="/lkml/2005/12/5/84">Michael Frank</a><ul><li><a href="/lkml/2005/12/5/368">Horst von Brand</a></li></ul></li></ul></li><li class="origin"><a href="/lkml/2005/12/3/81">Matthias Andree</a><ul><li><a href="/lkml/2005/12/3/81">Otavio Salvador</a></li><li><a href="/lkml/2005/12/3/84">David Ranson</a><ul><li><a href="/lkml/2005/12/3/88">Steven Rostedt</a></li></ul></li><li><a href="/lkml/2005/12/3/91">Arjan van de Ven</a><ul><li><a href="/lkml/2005/12/3/94">"M."</a></li><li><a href="/lkml/2005/12/3/167">Matthias Andree</a></li></ul></li></ul></li><li><a href="/lkml/2005/12/3/111">Jeff Garzik</a><ul><li><a href="/lkml/2005/12/3/127">Greg KH</a><ul><li><a href="/lkml/2005/12/3/197">Dmitry Torokhov</a></li></ul></li><li><a href="/lkml/2005/12/3/157">Matthias Andree</a></li><li><a href="/lkml/2005/12/4/71">Michael Frank</a></li></ul></li><li><a href="/lkml/2005/12/5/108">Rob Landley</a></li><li><a href="/lkml/2005/12/10/95">Ryan Anderson</a></li></ul></li></ul></li></ul></td><td width="32" rowspan="2" class="c" valign="top"><img src="/images/icornerl.gif" width="32" height="32" alt="/" /></td><td class="c" rowspan="2" valign="top" style="padding-top: 1em"><table><tr><td><table><tr><td class="lp">Date</td><td class="rp" itemprop="datePublished">Sat, 3 Dec 2005 17:27:55 +0100</td></tr><tr><td class="lp">From</td><td class="rp" itemprop="author">Matthias Andree <></td></tr><tr><td class="lp">Subject</td><td class="rp" itemprop="name">Re: RFC: Starting a stable kernel series off the 2.6 kernel</td></tr></table></td><td></td></tr></table><pre itemprop="articleBody">On Sat, 03 Dec 2005, Adrian Bunk wrote:<br /><br />> Things will change as time passes by, but then there's the possibility <br />> to open the next branch and bring the older branch into a security-fixes <br />> only mode.<br />> <br />> > If the current model doesn't work as you claim it doesn't, then maybe<br />> > the model needs finetuning. Right now the biggest pain is the userland<br />> > ABI changes that need new packages; sometimes (often) for no real hard<br />> > reason. Maybe we should just stop doing those bits, they're not in any<br />> > fundamental way blocking general progress (sure there's some code bloat<br />> > due to it, but I guess we'll just have to live with that).<br />> <br />> IOW, we should e.g. ensure that today's udev will still work flawlessly <br />> with kernel 2.6.30 (sic)?<br /><br />Exactly that, and kernel interfaces going away just to annoy binary<br />module providers also hurts third-party OSS modules, such as<br />Fujitsu-Siemens's ServerView agents.<br /><br />I found myself chasing some genuine bugs in that code to please GCC 4<br />(which doesn't tolerate extern blah declarations before static blah<br />definitions), and some idiotic breakage when inter_module_get was<br />dropped somewhen between 2.6.8 (where the modules compiled just fine)<br />and 2.6.13 (where they didn't as functions had been removed). And<br />there's no point arguing about deprecation warnings being around,<br />interfaces can be removed between 2.6 and 2.8 but not between<br />2.6.foo.bar and 2.6.foo.baz. Why not use<br /><br />#if LINUX_VERSION_CODE >= KERNEL_VERSION(2,8,0)<br /><br />or similar to automatically disable such interfaces at the given<br />version?<br /><br />> This could work, but it should be officially announced that e.g. a <br />> userspace running kernel 2.6.15 must work flawlessly with _any_ future <br />> 2.6 kernel.<br /><br />Right. This effectively means to call the beast 2.8.0 if you feel you<br />must break the applications.<br /><br />> For how many years do you think we will be able to ensure that this will <br />> stay true?<br /><br />Well, the current model, since it has been in effect, is just "let's<br />break the interfaces at will, but let's not hint anyone, so let's always<br />just bump the patchlevel no matter how intrusive the change is", and the<br />bandaid 2.6.M.N releases cannot cover the fact that the M -> M+1<br />transition causes breakage.<br /><br />Effectively, 2.6.X behaves like a bleeding edge development ("everything<br />changes") kernel such as 2.3 or 2.5; it isn't stable because some code<br />monkeys claim it is. You cannot declare the can of pea soup in front of<br />you "open" either.<br /><br />Linux is in fact light years away from being "stable", and while it was<br />relatively easy to move forward between 2.4.X releases and even 2.4.X<br />and some early 2.6.X release, moving between 2.6.X and 2.6.Y means<br />switching user-space, too, and chances are the new user-space doesn't<br />work with the old kernel.<br /><br />--<br />Matthias Andree<br />-<br />To unsubscribe from this list: send the line "unsubscribe linux-kernel" in<br />the body of a message to majordomo@vger.kernel.org<br />More majordomo info at <a href="http://vger.kernel.org/majordomo-info.html">http://vger.kernel.org/majordomo-info.html</a><br />Please read the FAQ at <a href="http://www.tux.org/lkml/">http://www.tux.org/lkml/</a><br /><br /></pre></td><td width="32" rowspan="2" class="c" valign="top"><img src="/images/icornerr.gif" width="32" height="32" alt="\" /></td></tr><tr><td align="right" valign="bottom"> 聽 </td></tr><tr><td align="right" valign="bottom">聽</td><td class="c" valign="bottom" style="padding-bottom: 0px"><img src="/images/bcornerl.gif" width="32" height="32" alt="\" /></td><td class="c">聽</td><td class="c" valign="bottom" style="padding-bottom: 0px"><img src="/images/bcornerr.gif" width="32" height="32" alt="/" /></td></tr><tr><td align="right" valign="top" colspan="2"> 聽 </td><td class="lm">Last update: 2005-12-03 17:30 聽聽 [from the cache]<br />漏2003-2020 <a href="http://blog.jasper.es/"><span itemprop="editor">Jasper Spaans</span></a>|hosted at <a href="https://www.digitalocean.com/?refcode=9a8e99d24cf9">Digital Ocean</a> and my Meterkast|<a href="http://blog.jasper.es/categories.html#lkml-ref">Read the blog</a></td><td>聽</td></tr></table><script language="javascript" src="/js/styleswitcher.js" type="text/javascript"></script></body></html>