CINXE.COM
LKML: Jiri Benc: Re: Broadcom 43xx first results
<?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: Jiri Benc: Re: Broadcom 43xx first results</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 Jiri Benc" href="/groupie.php?aid=28844" /><!--Matomo--><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><!--End Matomo Code--></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/2005/12/8"> [8]</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/2005/12/8/66" onclick="this.href='/lkml/headers'+'/2005/12/8/66';">[headers]</a>聽 <a href="/lkml/bounce/2005/12/8/66">[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/2005/12/4/90">First message in thread</a></li><li><a href="/lkml/2005/12/5/171">Jiri Benc</a><ul><li><a href="/lkml/2005/12/5/180">Jeff Garzik</a><ul><li><a href="/lkml/2005/12/5/186">Jiri Benc</a></li><li><a href="/lkml/2005/12/5/206">Dave Jones</a><ul><li><a href="/lkml/2005/12/5/208">Jeff Garzik</a></li><li><a href="/lkml/2005/12/6/152">Harald Welte</a><ul><li><a href="/lkml/2005/12/6/272">Jeff Garzik</a></li><li><a href="/lkml/2005/12/6/397">"David S. Miller"</a></li></ul></li></ul></li><li class="origin"><a href="/lkml/2005/12/8/67">Jiri Benc</a><ul><li><a href="/lkml/2005/12/8/67">Arjan van de Ven</a><ul><li><a href="/lkml/2005/12/8/73">Jiri Benc</a></li></ul></li></ul></li></ul></li><li><a href="/lkml/2005/12/5/181">Christoph Hellwig</a><ul><li><a href="/lkml/2005/12/5/190">Jiri Benc</a><ul><li><a href="/lkml/2005/12/5/192">Christoph Hellwig</a><ul><li><a href="/lkml/2005/12/5/209">Jiri Benc</a></li></ul></li><li><a href="/lkml/2005/12/6/163">Pavel Machek</a></li><li><a href="/lkml/2005/12/7/393">Michael Wu</a><ul><li><a href="/lkml/2005/12/7/411">Jeff Garzik</a></li></ul></li></ul></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">Thu, 8 Dec 2005 13:07:51 +0100</td></tr><tr><td class="lp">From</td><td class="rp" itemprop="author">Jiri Benc <></td></tr><tr><td class="lp">Subject</td><td class="rp" itemprop="name">Re: Broadcom 43xx first results</td></tr></table></td><td></td></tr></table><pre itemprop="articleBody">On Mon, 05 Dec 2005 14:08:28 -0500, Jeff Garzik wrote:<br />> > Unfortunately, the only long-term solution is to rewrite completely the<br />> > current in-kernel ieee80211 code (I would not call it a "stack") or<br />> > replace it with something another. The current code was written for<br />> > Intel devices and it doesn't support anything else - so every developer<br />> <br />> Patently false.<br /><br />Maybe some explanation why current in-kernel ieee80211 code needs to be<br />rewritten will be useful.<br /><br />1. To support WDS and devices capable to associate with multiple<br />networks, ieee80211_device needs to be separated to two (or even more,<br />see below) structures - one hardware dependent (channel and so) and one<br />link dependent (BSSID etc.).<br /><br />2. To support AP mode, you need to keep a list of associated stations.<br />No such list exists now. Furthermore, that list (or that structure) can<br />be reused also by a client to store information about AP it is<br />associated to. And - possibly - for a list of APs it can associate to,<br />i. e. list of found networks. Currently, informations about AP are<br />hardwired into ieee80211_device structure.<br /><br />3. Most of WE calls can be handled by ieee80211 itself. The rest should<br />be propagated to a driver in some easier way than requiring driver to<br />deal with the whole WE stuff itself. Also, exporting callbacks from<br />ieee80211 that driver has to set as particular WE handlers seems to be<br />unnecessary complicated.<br /><br />4. Callbacks like handle_auth() that were added some time ago are not<br />needed (for explanation, see corresponding thread on netdev).<br /><br />5. Some less important things, e. g. current very inefficient code which<br />deals with found networks.<br /><br /><br />-- <br />Jiri Benc<br />SUSE Labs<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: 2005-12-08 13:10 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>