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: Steve French: Re: inode_change_ok</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 Steve French" href="/groupie.php?aid=10026" /><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/2005/11"> [Nov]</a> 聽 <a class="nb" href="/lkml/2005/11/28"> [28]</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/2005/11/28/131" onclick="this.href='/lkml/headers'+'/2005/11/28/131';">[headers]</a>聽 <a href="/lkml/bounce/2005/11/28/131">[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/2005/11/28/78">First message in thread</a></li><li><a href="/lkml/2005/11/28/80">Trond Myklebust</a><ul><li><a href="/lkml/2005/11/28/83">Trond Myklebust</a><ul><li class="origin"><a href="">Steve French</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">Mon, 28 Nov 2005 15:05:15 -0600</td></tr><tr><td class="lp">From</td><td class="rp" itemprop="author">Steve French <></td></tr><tr><td class="lp">Subject</td><td class="rp" itemprop="name">Re: inode_change_ok</td></tr></table></td><td></td></tr></table><pre itemprop="articleBody">Trond Myklebust wrote:<br /><br />>CAP_CHOWN would therefore only be useful in the corner case where the<br />>server is allowing full root privileges to a client, and where the<br />>client is using capabilities to limit the root account's (or a setuid<br />>processes') ability to exercise those full privileges.<br />><br />>Cheers,<br />> Trond<br />><br />><br />> <br />><br />It sounds like I do need to make the change to cifs to add the optional <br />(based on the mount parm the admin can choose which model) call to <br />inode_change_ok to make the permissions check consistent when "noperm" <br />is not enabled on that mount , but I was trying to figure out how <br />important that change to the cifs code was and whether it was a priority <br />to get in. I realize that this corner case is far less common in the <br />wild for nfs. The corner case you describe above did seem to come up <br />today for cifs though (where the admin had in effect, trusted the <br />client, and allowed the client to mount as root with full root priv on <br />the server - I can imagine this being valid in their particular case, <br />with a trusted client os, over a trusted lan, but obviously would not be <br />the most common scenario). Obviously they did not turn on <br />multiusermount in /proc/fs/cifs (ie the ability to send different uids <br />on the wire depending on the uid of the client calling process) so the <br />only perm check that would be useful that was occuring at the client the <br />way they had chosen to configure/mount was happening in <br />generic_permission in the client. In their particular configuration <br />open and other paths in the vfs that call permission behaved fine and <br />the permission checks worked as expected - but chown always worked <br />(which is of course wrong) since the vfs does not call permission but <br />expects the vfs (if at all) to call inode_change_ok - so open failing, <br />but chown working was confusing to them.<br /><br />Of course it does not make it much easier trying to compare with other <br />filesystems - ncpfs does seem to call both (generic_permission <br />inode_change_ok) and but the afs model (openafs) is far more complicated <br />and somewhat confusing (and interestingly afs does not call <br />generic_permission but does call inode_change_ok) so it is not really <br />possible to draw too many conclusions for how afs approached the same <br />two usage models (apparently afs also had a third usage model based on <br />machine address)<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: 2005-11-28 21:10 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>