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: "Artem B. Bityuckiy": Re: [RFC] CryptoAPI & Compression</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 "Artem B. Bityuckiy"" href="/groupie.php?aid=28528" /><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/2005/4/3"> [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/2005/4/3/58" onclick="this.href='/lkml/headers'+'/2005/4/3/58';">[headers]</a>聽 <a href="/lkml/bounce/2005/4/3/58">[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/2005/3/25/104">First message in thread</a></li><li><a href="/lkml/2005/3/28/118">"Artem B. Bityuckiy"</a><ul><li><a href="/lkml/2005/3/29/92">Herbert Xu</a><ul><li><a href="/lkml/2005/3/29/122">"Artem B. Bityuckiy"</a><ul><li><a href="/lkml/2005/3/30/313">Herbert Xu</a><ul><li><a href="/lkml/2005/3/31/88">"Artem B. Bityuckiy"</a></li></ul></li></ul></li></ul></li><li><a href="/lkml/2005/3/31/74">Herbert Xu</a><ul><li class="origin"><a href="/lkml/2005/4/3/60">"Artem B. Bityuckiy"</a><ul><li><a href="/lkml/2005/4/3/60">Herbert Xu</a><ul><li><a href="/lkml/2005/4/3/61">"Artem B. Bityuckiy"</a></li><li><a href="/lkml/2005/4/3/64">"Artem B. Bityuckiy"</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">Sun, 03 Apr 2005 15:41:07 +0400</td></tr><tr><td class="lp">From</td><td class="rp" itemprop="author">"Artem B. Bityuckiy" <></td></tr><tr><td class="lp">Subject</td><td class="rp" itemprop="name">Re: [RFC] CryptoAPI & Compression</td></tr></table></td><td></td></tr></table><pre itemprop="articleBody">Herbert,<br /><br />I also wonder, does it at all correct to use negative windowBits in <br />crypto API? I mean, if windowBits is negative, zlib doesn't produce the <br />proper zstream header, which is incorrect according to RFC-1950. It also <br />doesn't calculate adler32.<br /><br />For example, if we work over an IP network (RFC-2384), the receiving <br />side may be confused by such a "stripped" zstream.<br /><br />Isn't it conceptually right to produce *correct* zstream, with the <br />header and the proper adler32 ?<br /><br />Yes, for JFFS2 we would like to have no adler32, we anyway protect our <br />data by CRC32. But I suppose this should be an additional feature.<br /><br />Comments?<br /><br />Herbert Xu wrote:<br />>>I made the changes to deflate_decompr() because the old version doesn't<br />>>work properly for me. There are 2 changes.<br />>><br />>>1. I've added the following code:<br />>><br />>>------------------------------------------------------------------------<br />>>if (slen > 2 && !(src[1] & PRESET_DICT) /* No preset dictionary */<br />>> && ((src[0] & 0x0f) == Z_DEFLATED) /* Comp. method byte is OK */<br />>> && !(((src[0] << 8) + src[1]) % 31)) { /* CMF*256 + FLG */<br />>> stream->next_in += 2;<br />>> stream->avail_in -= 2;<br />>>}<br />>>------------------------------------------------------------------------<br />> <br />> The reason you need to add this is because the window bits that<br />> was used to produce the compressed data is positive while the window<br />> bits crypto/deflate is using to perform the decompression isn't.<br />> <br />> So what we should do here is turn window bits into a configurable<br />> parameter.<br />> <br />> Once you supply the correct window bits information, the above is<br />> then simply an optimisation.<br />> <br />> Rather than keeping the above optimisation, JFFS should simply do<br />> the compression with a negative window bits value.<br />> <br />> Of course to maintain backwards compatibility you'll need to do this<br />> as a new compression type.<br />> <br /><br /><br />-- <br />Best Regards,<br />Artem B. Bityuckiy,<br />St.-Petersburg, Russia.<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: 2005-04-06 13:31 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>