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: Humberto Massa: Re: non-free firmware in kernel modules, aggregation and unclear copyright notice.</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 Humberto Massa" href="/groupie.php?aid=22864" /><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/2005/4"> [Apr]</a> 聽 <a class="nb" href="/lkml/2005/4/7"> [7]</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/2005/4/7/101" onclick="this.href='/lkml/headers'+'/2005/4/7/101';">[headers]</a>聽 <a href="/lkml/bounce/2005/4/7/101">[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/2005/4/7/101">First message in thread</a></li><li class="origin"><a href="/lkml/2005/4/7/312">Humberto Massa</a><ul><li><a href="/lkml/2005/4/7/312">Henning Makholm</a><ul><li><a href="/lkml/2005/4/7/319">"David Schwartz"</a><ul><li><a href="/lkml/2005/4/7/333">Henning Makholm</a><ul><li><a href="/lkml/2005/4/8/62">Sven Luther</a></li><li><a href="/lkml/2005/4/8/259">"David Schwartz"</a></li></ul></li><li><a href="/lkml/2005/4/8/3">Raul Miller</a></li></ul></li><li><a href="/lkml/2005/4/8/58">Sven Luther</a></li></ul></li><li><a href="/lkml/2005/4/8/63">Sven Luther</a></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">Thu, 07 Apr 2005 09:15:07 -0300</td></tr><tr><td class="lp">From</td><td class="rp" itemprop="author">Humberto Massa <></td></tr><tr><td class="lp">Subject</td><td class="rp" itemprop="name">Re: non-free firmware in kernel modules, aggregation and unclear copyright notice.</td></tr></table></td><td></td></tr></table><pre itemprop="articleBody">David Schwartz wrote:<br /><br /> >>Well whoever wrote that seems to have taken the stand that<br /> >>the openfirmware package was were the firmware came from.<br /> >>The person obviously made a lot of statements without<br /> >>bothering checking out the real source. Well it didn't come<br /> >>from there, I got it from Alteon under a written agreement<br /> >>stating I could distribute the image under the GPL. Since<br /> >>the firmware is simply data to Linux, hence keeping it under<br /> >>the GPL should be just fine.<br /> ><br /> ><br /> >You cannot distribute anything under the GPL if you cannot<br /> >also distribute the source code (the preferred form of the<br /> >software for the purpose of making modifications to it). How<br /> >Linux sees it is irrelevant. For any piece of software, one<br /> >can imagine some processor that can only see it as data. The<br /> >GPL doesn't distinguish between processors.<br /><br />I think this is undisputed.<br /><br /> >Alteon's written agreement notwithstanding, you cannot<br /> >distribute the firmware under the GPL if you cannot provide<br /> >the preferred form of the firmware for the purpose of making<br /> >modifications to it. The firmware does not run on Linux, so<br /> >saying "linux sees it as data" is as absurd as saying I can<br /> >distribute the x86 Linux kernel without the source because my<br /> >calculator can only see it as data.<br /> ><br /> >You cannot distribute the firmware binary under the GPL.<br /> >Period.<br /><br />This is where you are wrong IMMHO. All that is needed for you<br />to distribute the hexdump blob under the GPL is a declaration<br />from the copyright holder saying "this, to me, is the<br />preferred form for modification of the firmware and hence the<br />source code under the GPL."<br /><br /> >Now, if you were trying to say that you could aggregate the<br /> >firmware with another work and distribute the result under<br /> >the GPL, the test would be whether the final result is "mere<br /> >aggregation" or not. This is a fantastically tricky question<br /> >and I don't think anyone on this list could give you<br /> >particularly useful guidance.<br /><br />After a *lot* of discussion, it was deliberated on d-l that<br />this is not that tricky at all, and that the "mere<br />aggregation" clause applies to the combination, for various<br />reasons, with a great degree of safety. (Safer than that,<br />only after court) :-)<br /><br /> >My own opinion is that it's a threshold issue based upon<br /> >several factors. For example -- has the firmware been<br /> >specifically designed to work with the Linux driver or is it<br /> >"generic" firmware? If you can't take the thing you're<br /> >distributing (the combined binary) and extract two works from<br /> >it (the firmware and the work whose source you are offering),<br /> >I cannot see how you can claim it's mere aggregation.<br /><br />Now, if the firmware was specifically designed to work with<br />the linux driver, than it *is* a derivative work on the kernel<br />as a whole and the source code should be provided upon<br />redistribution as per GPL section 3 etc.<br /><br />*BUT* this does not preclude Broadcom from stating: "our<br />engineers generated by hand the hex codes that make our<br />hardware work."<br /><br /> >If you believe the linker "merely aggregates" the object code<br /> >for the driver with the data for the firmware, I can't see<br /> >how you can argue that any linking is anything but mere<br /> >aggregation. In neither case can you separate the linked work<br /> >into the two separate works and in both cases the linker<br /> >provides one work direct access to the other.<br /><br />No-one is saying that the linker "merely aggregates" object<br />code for the driver; what *is* being said is: in the case of<br />firmware, especially if the firmware is neither a derivative<br />work on the kernel (see above) nor the firmware includes part<br />of the kernel (duh), it is *fairly* *safe* to consider the<br />intermixing of firmware bytes with kernel binary image bytes<br />in an ELF object file as mere aggregation.<br /><br /> >If you only distribute the source to the driver and don't put<br /> >a GPL notice in the files that contain the firmware data, I<br /> >think you're okay. I think you're asking for trouble if you<br /> >distribute a combined compiled/linked driver.<br /><br />Disagreed.<br /><br /> >DS<br /><br />HTH,<br /><br />Massa<br /><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: 2005-04-07 14:19 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>