path: root/fr_FR.ISO8859-1/books/handbook/hw/chapter.sgml

<!--
     The FreeBSD Documentation Project
     The FreeBSD French Documentation Project

     $FreeBSD$
     Original revision: n.nn
-->

    <chapter id="hw">
      <title>** Compatibilit&eacute; mat&eacute;rielle</title>
      &trans.a.haby;

      
      <para>Les questions de compatibilit&eacute; mat&eacute;rielle sont aujourd'hui
        les plus probl&eacute;matiques de l'industrie informatique et FreeBSD
        n'en est nullement &agrave; l'abri. De ce point de vue, l'avantage qu'a
        FreeBSD de pouvoir &ecirc;tre utilis&eacute; sur du mat&eacute;riel PC courant et
        peu co&ucirc;teux est aussi une difficult&eacute; lorsqu'il faut supporter
        l'incroyable vari&eacute;t&eacute; de composants disponibles.
        Il est impossible de donner une liste exhaustive des mat&eacute;riels
        compatibles avec FreeBSD, mais ce chapitre est un catalogue des
        pilotes de p&eacute;riph&eacute;riques inclus dans FreeBSD et des mat&eacute;riels que
        chaque pilote supporte. Si possible et appropri&eacute;, des notes ont
        ajout&eacute;es sur les mat&eacute;riels eux-m&ecirc;mes. Vous pouvez aussi vous 
        r&eacute;f&eacute;rer au chapitre <link linkend="kernelconfig-config">Configurer
        le noyau de FreeBSD</link> de ce manuel pour avoir
        la liste des mat&eacute;riels support&eacute;s.</para>
      
      <para>FreeBSD est un projet b&eacute;n&eacute;vole qui n'a pas les moyens de financer
        un service de tests, nous reposons sur vous, les utilisateurs, pour une
        grande part des informations que fournit ce catalogue. Si vous avez
        l'exp&eacute;rience personnelle d'un mat&eacute;riel qui fonctionne ou ne fonctionne
        pas avec FreeBSD, faites-le nous savoir par courrier &eacute;lectronique
        &agrave; &a.doc;. Les questions concernant les mat&eacute;riels compatibles doivent
        &ecirc;tre adress&eacute;es &agrave; &a.questions; (voyez la section 
        <link linkend="eresources-mail">Listes de diffusion</link>
        pour plus d'informations). Quand vous nous faites
        parvenir de l'information ou posez une question, n'oubliez pas s'il vous
        pla&icirc;t de pr&eacute;ciser exactement quelle version de FreeBSD vous utilisez et
        de donner le maximum de d&eacute;tails sur votre configuration 
        mat&eacute;rielle.</para>
      
      <sect1>
        <title>Ressources Internet</title>
        
        <para>Les liens donn&eacute;s ci-dessous se sont av&eacute;r&eacute;s utiles pour guider
          dans les choix de mat&eacute;riels. Bien que les renseignements qu'ils vous
          donnent ne soient pas n&eacute;cessairement sp&eacute;cifiques (ou m&ecirc;me 
          applicables) &agrave; FreeBSD, ils ne d&eacute;pendent pas, pour la plupart 
          du syst&egrave;me d'exploitation. V&eacute;rifiez s'il vous pla&icirc;t dans le guide
          du mat&eacute;riel pour FreeBSD que la configuration que vous avez choisie
          soit compatible avec FreeBSD avant d'acheter quoi que ce soit.</para>
        
        <para>
          <itemizedlist>
            
            <listitem>
              <para><ulink url="http://www.tomshardware.com/">The Pentium
                  Systems Hardware Performance Guide</ulink>&nbsp;-&nbsp;le
                  guide des performances des syst&egrave;mes Pentium.</para>
            </listitem>
            
          </itemizedlist>
        </para>
        
      </sect1>
      
      <sect1 id="hw-configs">
        <title>Exemples de configurations</title>
        
        <para>La liste de configurations ci-dessous ne constitue en aucun
          cas une publicit&eacute; pour un constructeur ou un produit de la 
          part du <emphasis>Projet FreeBSD</emphasis>. Ces informations ne
          sont donn&eacute;es que pour &ecirc;tre utiles et rassemblent simplement les
          exp&eacute;riences de diff&eacute;rentes personnes sur des configurations vari&eacute;es.
          Tarifs indicatifs. Chauss&eacute;e glissante. Attention au chien.</para>
        
        <sect2 id="hw-jordans-picks">
          <title>La s&eacute;lection de Jordan</title>
          
          <para>J'ai obtenu de bons r&eacute;sultats en mettant sur pied des stations 
            de travail et des serveurs avec les composants ci-dessous. Je ne
            peut vous garantir que vous en aurez aussi, ni qu'aucune des marques
            cit&eacute;es restera &ldquo;le meilleur choix&rdquo;. J'essaierai, si 
            possible, de tenir cette liste &agrave; jour, mais ne peux bien &eacute;videmment
            vous assurer qu'elle le soit &agrave; un moment donn&eacute;.</para>
          
          <sect3 id="hw-mb">
            <title>Cartes m&egrave;res</title>
            
            <para>Pour les syst&egrave;mes Pentium Pro (P6), j'aime assez la carte m&egrave;re
               bi-processseurs 
               <ulink url="http://www.tyan.com/html/products.html">Tyan</ulink>
              S1668. Elle fait un sympathique syst&egrave;me &agrave; un ou deux processeurs
              (ce que supporte FreeBSD 3.0) et le prix du Pentium Pro 180/256K 
              a maintenant baiss&eacute; &agrave; un niveau vraiment abordable. Le Pentium Pro
              reste mon processeur favori pour les serveurs (les m&eacute;gahertzs ne
              font pas tout).</para>
            
            <para>Pour les Pentium II, j'ai un s&eacute;rieux pr&eacute;jug&eacute; en faveur de la 
              carte m&egrave;re <ulink url="http://www.asus.com.tw">ASUS</ulink> 
              <ulink url="http://www.asus.com.tw/Products/Motherboard/Pentiumpro/P2l97-s/index.html">P2l97-S</ulink> 
              avec contr&ocirc;leur WIDE SCSI int&eacute;gr&eacute;.</para>
            
            <para>Pour les machines Pentium, la carte m&egrave;re ASUS 
               <ulink url="http://www.asus.com.tw/Products/Motherboard/Pentium/P55tp4/index.html">P55T2P4</ulink> 
               para&icirc;t un bon choix pour un serveur ou une station de travail
               de taille moyenne &agrave; importante. Vous pouvez aussi 
               regarder du c&ocirc;t&eacute; de la carte
               <ulink url="http://asustek.asus.com.tw/FTP/ASUS/Info/Spec/pvi-486sp3.txt">486SP3G</ulink>, 
               si vous cherchez une carte m&egrave;re 486.</para>
            
            <note>
              <para>(Il semble qu'il soit devenu difficile de se procurer ces
                derni&egrave;res, qu'ASUS ne fabrique apparemment plus.) </para>
            </note>
            
            <para>Ceux qui veulent utiliser des syst&egrave;mes plus tol&eacute;rants aux
              erreurs doivent veiller &agrave; employer de la m&eacute;moire avec contr&ocirc;le
              de parit&eacute;, ou ECC, pour des applications non-stop.</para>

            <note>
              <para>La m&eacute;moire ECC entra&icirc;ne une petite perte de performances
                (que vous remarquerez ou non selon votre application) mais vous
                apporte des gains significatifs en termes de tol&eacute;rance 
                d'erreur.</para>
            </note>
          </sect3>
          
          <sect3>
            <title>Contr&ocirc;leurs de disque</title>
            
            <para>C'est un point plus d&eacute;licat. J'utilisais 
              inconditionnellement des contr&ocirc;leurs
              <ulink url="http://www.buslogic.com">Buslogic</ulink>
              pour tout, de l'ISA au PCI, j'incline maintenant plut&ocirc;t vers
              le contr&ocirc;leur <ulink
                url="http://www.adaptec.com">Adaptec</ulink> 1542CF pour l'ISA,
              le contr&ocirc;leur Buslogic Bt747c pour l'EISA et le contr&ocirc;leur
              Adaptec 2940UW pour le PCI.</para>
            
            <para>J'ai aussi eu de bons r&eacute;sultats avec les cartes
              PCI NCR/Symbios, bien qu'il faille s'assurer que 
              votre carte m&egrave;re supporte le mod&egrave;le sans BIOS (s'il n'y
              a rien sur votre carte qui ressemble vaguement &agrave; une puce
              ROM, c'est probablement un mod&egrave;le qui s'attend &agrave; ce que son
              BIOS soit sur la carte m&egrave;re).</para>
            
            <para>Si vous pensez qu'il vous faut plus d'un contr&ocirc;leur SCSI,
              vous pouvez songer &agrave; &eacute;conomiser vos maigres ressources en 
              emplacements PCI en achetant une carte Adaptec 3940, qui 
              int&egrave;gre deux contr&ocirc;leurs PCI sur un seul connecteur.</para>
            
          </sect3>
          
          <sect3 id="hw-disks">
            <title>Disques durs</title>
            
            <para>Pour cette version particuli&egrave;re de la roulette russe, je 
              donnerais peu de conseils pr&eacute;cis sinon pour recommander
              &ldquo;du SCSI plut&ocirc;t que de l'IDE d&egrave;s que vous pouvez vous
              l'offrir&rdquo;. M&ecirc;me sur de petites machines de bureau, le SCSI
              est souvent un meilleur choix parce qu'il vous permet de
              migrer vos disques du serveur vers la machine de bureau lorsque
              les prix en chute des disques en font une solution &eacute;conomiquement
              viable. Si vous avez plus d'une machine &agrave; administrer, ne pensez
              pas seulement en terme de stockage, voyez plut&ocirc;t cela comme
              une cha&icirc;ne alimentaire!</para>
            
            <para>Je ne trouve pas que les disques WIDE SCSI repr&eacute;sentent
              un investissement n&eacute;cessaire, &agrave; moins que vous ne mettiez en place
              un serveur NFS ou des forums de discussion
              qui devront supporter beaucoup d'acc&egrave;s disque pour de nombreux
              utilisateurs.</para>
            
          </sect3>
          
          <sect3 id="hw-jordans-picks-cdrom">
            <title>Lecteur de CD-ROMs</title>
            
            <para>Ma pr&eacute;f&eacute;rence pour le SCSI s'applique aussi aux lecteurs de
              CD-ROMs SCSI, et bien que j'ai toujours eu de bons r&eacute;sultats
              avec le mod&egrave;le <ulink url="http://www.toshiba.com">Toshiba</ulink>
              XM-3501B (qui existe aussi en version tiroir sous la r&eacute;f&eacute;rence
              XM-5401B), je suis maintenant tr&egrave;s partisan du lecteur
              <ulink url="http://www.plextor.com">Plextor</ulink> PX-12CS.
              C'est un lecteur 12x dont les performances et la fiabilit&eacute; sont
              excellentes.</para>
            
            <para>D'une fa&ccedil;on g&eacute;n&eacute;rale, la plupart des lecteurs de CD-ROMs SCSI
              que j'ai vus, sont de fabrication robuste et vous ne vous 
              tromperez pas non plus si vous prenez un mod&egrave;le HP ou NEC. Le prix
              des lecteurs de CD-ROMs SCSI semble avoir aussi consid&eacute;rablement 
              baiss&eacute; ces derniers mois et devient comp&eacute;titif avec celui des 
              lecteurs IDE, alors qu'ils restent techniquement sup&eacute;rieurs. A 
              choisir entre les deux, je ne vois pas de raison de se d&eacute;cider 
              pour un lecteur IDE.</para>
            
          </sect3>
          
          <sect3 id="hw-worm">
            <title>Graveurs de CD-ROMs non r&eacute;inscriptibles</title>
            
            <para>Au moment o&ugrave; j'&eacute;cris ceci, FreeBSD supporte trois types de
              graveurs de CD-ROMs (bien que je pense qu'ils viennent en fait
              tous de chez Phillips): le Phillips CDD 522 (se comporte comme
              le Plasmon), le Plasmon RF4100 et le HP 6020i. J'utilise 
              personnellement le HP 6020i pour graver mes CD-ROMs (avec la
              version 2.2-current de FreeBSD&nbsp;-&nbsp;il ne fonctionne pas 
              avec la version 2.1.5 et les versions ant&eacute;rieures du pilote SCSI)
              qui me donne toute satisfaction. Regardez dans le fichier
              <ulink url="file:/usr/share/examples/worm">/usr/share/examples/worm</ulink>
              sur votre syst&egrave;me 2.2 pour avoir des exemples de proc&eacute;dures pour
              cr&eacute;er des images au format ISO9660 (avec les extensions RockRidge)
              de vos syst&egrave;mes de fichiers et graver ensuite des CD-ROMs avec un
              HP6020i.</para>
            
          </sect3>
          
          <sect3 id="hw-tape">
            <title>Lecteurs de bandes</title>
            
            <para>J'ai obtenu de bons r&eacute;sultats avec les lecteurs
                <ulink url="http://www.Exabyte.COM:80/Products/8mm/8505XL/Rfeatures.html">8mm</ulink> 
               de chez 
               <ulink url="http://www.exabyte.com">Exabyte</ulink> 
               et 
               <ulink url="http://www-dmo.external.hp.com:80/tape/_cpb0001.htm">4mm (DAT)</ulink>
               de chez <ulink url="http://www.hp.com">HP</ulink>.</para>
            
            <para>Pour les sauvegardes, je recommande les Exabytes pour la 
               robustesse (et la plus grande capacit&eacute;) des bandes 8mm.</para>
            
          </sect3>
          
          <sect3 id="hw-video">
            <title>Cartes graphiques</title>
            
            <para>Si vous pouvez aussi vous offrir un serveur X commercial
              pour 99&#36; US de chez 
              <ulink url="http://www.xig.com/">Xi Graphics, Inc. (autrefois, X Inside, Inc)</ulink>
              alors je vous recommande vivement la carte
              <ulink url="http://www.matrox.com/">Matrox</ulink>
              <ulink url="http://www.matrox.com/mgaweb/brochure.htm">Millenium</ulink>. 
              Cette carte est aussi tr&egrave;s bien support&eacute;e par le serveur 
              <ulink url="http://www.xfree86.org/">XFree86</ulink>, 
              qui en est maintenant &agrave; sa version 3.3.2.</para>
            
            <para>Les cartes
               <ulink url="http://www.nine.com/">Number 9</ulink> sont aussi
               un excellent choix&nbsp;-&nbsp;leurs cartes Vision 868 et 968 
               (la s&eacute;rie 9FX) bas&eacute;es sur le circuit S3 sont aussi tr&egrave;s rapides 
               et bien g&eacute;r&eacute;es par le pilote S3 du serveur XFree86.</para>
            
          </sect3>
          
          <sect3 id="hw-monitors">
            <title>Moniteurs</title>
            
            <para>J'ai eu d'excellents r&eacute;sultats avec les moniteurs
               <ulink url="http://cons3.sel.sony.com/SEL/ccpg/display/ms17se2.html">Sony Multiscan 17seII</ulink>, 
              et avec le Viewsonic qui utilise le m&ecirc;me tube (Trinitron). Pour
              des mod&egrave;les au-del&agrave; de 17", tout ce que je peux aujourd'hui 
              conseiller est de ne pas d&eacute;penser moins de 2.500 &#36; pour
              un moniteur 21" ou 1.700 &#36; pour un 20", si vous en avez
              vraiment besoin. Il y de bons &eacute;crans dans 
              la gamme des 20" et plus,
              et il y en a aussi de bon march&eacute;. Malheureusement, il y en a tr&egrave;s
              peu qui soient &agrave; la fois de bonne qualit&eacute; et bon march&eacute;!</para>
            
          </sect3>
          
          <sect3 id="hw-networking">
            <title>R&eacute;seau</title>
            
            <para>Je peux recommander le contr&ocirc;leur <ulink
                url="http://www.smc.com/">SMC</ulink> Ultra 16 pour les
              applications ISA et les cartes SMC EtherPower ou Compex ENET32
              pour les r&eacute;seaux importants bas&eacute;s sur du PCI. Ces deux cartes
              PCI sont construites autour de la puce contr&ocirc;leur Ethernet
              DEC DC21041 et les autres cartes qui employent cette puce, telles
              que la Zynx ZX432 et la DEC DE435, fonctionneront aussi. Pour
              les r&eacute;seaux 100Mbit, les cartes SMC SMC9332DST 10/100MB ou Intel
              Intel EtherExpress Pro/100B font du bon travail, ma pr&eacute;f&eacute;rence
              allant &agrave; la carte Intel EtherExpress.</para>
            
            <para>Si d'un autre c&ocirc;t&eacute; vous cherchez la solution la moins ch&egrave;re
              possible, mais qui fonctionne malgr&eacute; tout raisonnablement, alors
              pratiquement n'importe quel clone NE2000 est un bon choix.</para>
            
          </sect3>
          
          <sect3 id="hw-serial">
            <title>S&eacute;rie</title>
            
            <para>Si vous cherchez des solutions pour un r&eacute;seau s&eacute;rie &agrave; grande
              vitesse, alors <ulink url="http://www.dgii.com/">Digi
                International</ulink> fabrique la s&eacute;rie <ulink
                url="http://www.dgii.com/prodprofiles/profiles-prices/digiprofiles/digispecs/sync570.html">SYNC/570</ulink>, 
                pour laquelle FreeBSD-current a maintenant des pilotes. 
                <ulink url="http://www.etinc.com">Emerging Technologies</ulink>
                fabrique aussi une carte avec des fonctionnalit&eacute;s T1/E1, 
                qui utilise du logiciel qu'il fournit. 
                Je n'ai cependant pas l'exp&eacute;rience personnelle de ces deux 
                produits.</para>
            
            <para>Les possibilit&eacute;s de cartes multi-ports sont quelque peu plus 
                nombreuses, bien que le support par FreeBSD des produits
                <ulink url="http://www.cyclades.com/">Cyclades</ulink> soit
                r&eacute;put&eacute; le plus complet, essentiellement en raison de 
                l'engagement pris par cette compagnie de nous fournir du
                mat&eacute;riel pour &eacute;valuation et des sp&eacute;cifications techniques. J'ai
                entendu dire que la Cyclom-16Ye offrait le meilleur rapport
                prix/performances, mais je n'ai pas consult&eacute; les tarifs r&eacute;cents.
                D'autres cartes multi-ports dont j'ai entendu dire du bien
                sont les BOCA et les AST, et <ulink
                url="http://www.stallion.com/">Stallion
                Technologies</ulink> propose apparemment <ulink
                url="ftp://ftp.stallion.com/drivers/unsupported/freebsd/stalbsd-0.0.4.tar.gz">ici</ulink>
                un pilote non officiel pour ses cartes.</para>
            
          </sect3>
          
          <sect3 id="hw-audio">
            <title>Audio</title>
            
            <para>J'utilise actuellement une AWE32 de <ulink
                url="http://www.creaf.com/">Creative Labs</ulink>, bien qu'&agrave; peu
                pr&egrave;s tout ce qui vient de chez Creative Labs marcherait 
                aujourd'hui. Ce qui ne veut pas dire que d'autres cartes son
                ne marchent pas, simplemement je n'en ai qu'une exp&eacute;rience
                limit&eacute;e (j'aimais bien autrefois les cartes GUS, mais la
                situation des cartes Gravis est d&eacute;licate depuis quelque
                temps).</para>
            
          </sect3>
          
          <sect3 id="hw-vgrabbers">
            <title>Vid&eacute;o</title>
            
            <para>Pour la capture vid&eacute;o, il y a deux bons 
               choix&nbsp;-&nbsp;n'importe
              quelle carte &agrave; base de puce Brooktree BT848, comme les Hauppauge
              ou les WinTV, marchera &agrave; merveille avec FreeBSD. Une autre carte
              que j'utilise est la
              <ulink
                url="http://www.matrox.com/">Matrox</ulink> <ulink
                url="http://www.matrox.com/imgweb/meteor.htm">Meteor</ulink>. 
              FreeBSD supporte aussi la carte d'incrustation vid&eacute;o plus ancienne
              de chez Creative Labs, mais elles deviennent difficiles &agrave; trouver.
              Notez que la carte Meteor <emphasis>ne fonctionnera pas</emphasis>
              avec les cartes m&egrave;res qui ont un contr&ocirc;leur 440FX! Consultez
              la section
              <link linkend="hw-mb">Cartes m&egrave;res</link> pour plus de
              d&eacute;tails. Dans ce cas, il vaut mieux prendre une carte 
              BT848.</para>
            
          </sect3>
        </sect2>
      </sect1>
      
      <sect1 id="hw-core">
        <title>Composants de base/Processeurs</title>
        
        <sect2>
          <title>Cartes m&egrave;res, bus et contr&ocirc;leurs de bus</title>
          
          <sect3>
            <title>* ISA</title>
            <para></para>
          </sect3>
          
          <sect3>
            <title>* EISA</title>
            <para></para>
          </sect3>
          
          <sect3>
            <title>* VLB</title>
            <para></para>
          </sect3>
          
          <sect3 id="hw-mb-pci">
            <title>PCI</title>
            
            <para><emphasis>Contribution de &a.rgrimes;.<!-- <br> -->25 Avril
                1995.</emphasis></para>
            
            <para><emphasis>Mises &agrave; jour de &a.jkh;.</emphasis><!-- <br>
              -->Derni&egrave;re mise &agrave; jour le <emphasis>26 Ao&ucirc;t
              1996.</emphasis></para>
            
            <para>Parmi les contr&ocirc;leurs INTEL PCI, la liste suivante d&eacute;crit
              diff&eacute;rents types de probl&egrave;me connus, et leur gravit&eacute;, du pire
              au meilleur.</para>
            
              <variablelist>
                <varlistentry><term>Mercury:</term>
                  <listitem>
                    <para>Probl&egrave;mes de coh&eacute;rence du cache, en particulier s'il
                     y a des contr&ocirc;leurs de bus ISA en plus du pont ISA/PCI.
                     C'est un probl&egrave;me mat&eacute;riel, la seule solution consiste
                     &agrave; d&eacute;sactiver le cache.</para>
                  </listitem>
                </varlistentry>
                
                <varlistentry><term>Saturn-I <emphasis>(i.e., 82424ZX en i
                   r&eacute;vision 0, 1 ou 2)</emphasis>:</term>
                  
                  <listitem>
                    <para>Probl&egrave;me de coh&eacute;rence lors de la r&eacute;&eacute;criture dans le
                      cache. C'est un probl&egrave;me mat&eacute;riel. La seule parade 
                      consiste &agrave; configurer le cache externe en mode 
                      transparent. Ou &agrave; passer &agrave; la version Saturn-II.</para>
                  </listitem>
                </varlistentry>
                
                <varlistentry><term>Saturn-II <emphasis>(i.e., 82424ZX en 
                   r&eacute;vision 3 ou 4)</emphasis>:</term>
                  
                  <listitem>
                    <para>Fonctionne bien, mais de nombreux fabriquants de
                      carte m&egrave;re ne se pr&eacute;occupent pas du bit SRAM n&eacute;cessaire
                      aux op&eacute;rations de r&eacute;ecriture. On peut y pallier en
                      utilisant le mode transparent ou en g&eacute;rant le bit SRAM.
                      (J'ai fait cela avec une ASUS PCI/I-486SP3G r&eacute;vision 1.6 
                      et des cartes plus r&eacute;centes).</para>
                  </listitem>
                </varlistentry>
                
                <varlistentry><term>Neptune:</term>
                  
                  <listitem>
                    <para>Ne peut g&eacute;rer plus de deux contr&ocirc;leurs de bus. C'est
                      une erreur de conception reconnue par Intel. Parmi les
                      solutions: ne pas utiliser plus de deux contr&ocirc;leurs, 
                      mat&eacute;riel sp&eacute;cialement con&ccedil;u pour remplacer l'arbitre de
                      bus PCI (apparu avec l'Intel Altair et d'autres cartes
                      m&egrave;res pour serveur Intel), et bien s&ucirc;r la r&eacute;ponse 
                      officielle d'Intel, le remplacer par un Triton, nous
                      &ldquo;l'y avons mis&rdquo;.</para>
                  </listitem>
                </varlistentry>
                
                <varlistentry><term>Triton <emphasis>(ie,
                      430FX)</emphasis>:</term>
                  
                  <listitem>
                    <para>Pas de probl&egrave;me de coh&eacute;rence du cache ou de contr&ocirc;le 
                      du bus connu. Mais cette puce n'impl&eacute;mente tout simplement
                      pas le contr&ocirc;le de parit&eacute;. Contournez le probl&egrave;me de 
                      parit&eacute;. Utilisez des cartes Triton-II si vous avez
                      le choix.</para>
                  </listitem>
                </varlistentry>
                
                <varlistentry><term>Triton-II <emphasis>(ie,
                      430HX)</emphasis>:</term>
                  
                  <listitem>
                    <para>Tous les &eacute;chos sur les cartes m&egrave;res avec cette puce
                      sont jusqu'ici favorables. Pas de probl&egrave;me connu.</para>
                  </listitem>
                </varlistentry>
                
                <varlistentry><term>Orion:</term>
                  
                  <listitem>
                    <para>Les premi&egrave;res versions de cette puce souffraient d'un
                      retard en &eacute;criture PCI qui entra&icirc;nait des d&eacute;gradations
                      sensibles de performance des applications gourmandes en
                      trafic sur le bus PCI. Les versions B0 et ult&eacute;rieures de
                      cette puce ont r&eacute;gl&eacute; ce probl&egrave;me.</para>
                  </listitem>
                </varlistentry>
                
                <varlistentry><term><ulink
                      url="http://developer.intel.com/design/pcisets/desktop.htm#440FX">440FX</ulink>:</term>
                  
                  <listitem>
                    <para>Cette puce pour <ulink
                        url="http://www.intel.com/procs/ppro/index.htm">Pentium Pro</ulink>
                    semble fonctionner correctement et ne souffre pas des 
                    probl&egrave;mes qu'ont connus 
                    les premi&egrave;res puces Orion. Il accepte
                    aussi une plus grande vari&eacute;t&eacute; de types de m&eacute;moire, y compris
                    l'ECC et le contr&ocirc;le de parit&eacute;. Le seul probl&egrave;me connu est 
                    que la carte d'acquisition vid&eacute;o Matrox Meteor ne fonctionne
                    pas avec.</para>
                    
                  </listitem>
                </varlistentry>
              </variablelist>
            
          </sect3>
        </sect2>
        
        <sect2>
          <title>Processeurs/Coprocesseurs</title>
          
          <para><emphasis>Contribution de &a.asami;.<!-- <br> -->26 D&eacute;cembre
              1997.</emphasis></para>
          
          <sect3>
            <title>P6 (Pentium Pro/Pentium II)</title>
            
            <para>Le Pentium Pro et le Pentium II fonctionnent parfaitement
               avec FreeBSD.
               De fait, notre site ftp de base <ulink
                url="ftp://ftp.freebsd.org/">ftp.freebsd.org</ulink> (aussi
              appel&eacute; "<filename>ftp.cdrom.com</filename>", le site ftp le plus
              important au monde) utilise FreeBSD sur un Pentium Pro. Des <ulink
                url="ftp://ftp.cdrom.com/archive-info/wcarchive.txt">D&eacute;tails de la configuration</ulink> sont disponibles si vous &ecirc;tes int&eacute;ress&eacute;s.</para>
            
          </sect3>
          
          <sect3>
            <title>Pentium</title>
            
            <para>Les Pentium Intel (P54C), Pentium MMX (P55C), AMD K6 et
              Cyrix/IBM 6x86MX fonctionnent tous avec FreeBSD. Je n'entrerai
              pas dans le d&eacute;tail de savoir lequel est plus rapide que l'autre,
              il y a des zillions de sites Web sur l'Internet pour vous
              l'expliquer &agrave; l'endroit et &agrave; l'envers.
              <!-- smiley --><emphasis>:)</emphasis></para>

            <note>
              <para>Les diff&eacute;rents processeurs ont besoin d'une alimentation
                et d'une ventilation diff&eacute;rentes. Assurez-vous que votre carte
                m&egrave;re fournit la tension exacte requise par votre processeur. Par
                exemple, de nombreuses puces MMX ont besoin d'une alimentation
                d&eacute;doubl&eacute;e (e.g., 2.9V pour l'unit&eacute; centrale, 3.3V pour les 
                entr&eacute;es/sorties). Certaines puces AMD et Cyrix/IBM chauffent
                plus que les puces Intel. Dans ce cas, v&eacute;rifiez que vous avez
                bien les bons radiateurs et ventilateurs (vous pouvez trouver la
                liste des composants certifi&eacute;s sur leurs pages Web).</para>
            </note>
            
            <sect4>
              <title>Vitesses d'horloge</title>
              
              <para><emphasis>Contribution de &a.rgrimes;.<!-- <br> -->1
                  Octobre 1996.</emphasis></para>
              
              <para><emphasis>Mise &agrave; jour de &a.asami;.<!-- <br> -->27 D&eacute;cembre
                  1997.</emphasis></para>
              
              <para>Les machines de la cat&eacute;gorie Pentium utilisent des vitesses
                d'horloge diff&eacute;rentes pour leurs diff&eacute;rents composants. Il y a
                la fr&eacute;quence du processeur, celle du bus m&eacute;moire externe et 
                celle du bus PCI. Il n'est pas toujours exact qu'un processeur
                &ldquo;plus rapide&rdquo; compose un syst&egrave;me plus rapide
                qu'un &ldquo;plus lent&rdquo;, du fait de ces diff&eacute;rentes
                vitesses d'horloge. Voici une table qui donne la liste des
                possibilit&eacute;s:</para>

              <informaltable frame="none">
                <tgroup cols="4">
                  <thead>
                    <row>
                      <entry>Fr&eacute;quence du processeur (MHz)</entry>
                      <entry>Horloge externe et fr&eacute;quence du bus m&eacute;moire (mHz) 
                        [a]</entry>
                      <entry>Coefficient multiplicateur horloge 
                        interne/externe</entry>
                      <entry>Fr&eacute;quence du bus PCI (MHz)</entry>
                    </row>
                  </thead>
                  <tbody>
                    <row>
                      <entry>60</entry>
                      <entry>60</entry>
                      <entry>1.0</entry>
                      <entry>30</entry>
                    </row>

                    <row>
                      <entry>66</entry>
                      <entry>66</entry>
                      <entry>1.0</entry>
                      <entry>33</entry>
                    </row>

                    <row>
                      <entry>75</entry>
                      <entry>50</entry>
                      <entry>1.5</entry>
                      <entry>25</entry>
                    </row>

                    <row>
                      <entry>90</entry>
                      <entry>60</entry>
                      <entry>1.5</entry>
                      <entry>30</entry>
                    </row>

                    <row>
                      <entry>100</entry>
                      <entry>50 [b]
                      </entry>
                      <entry>2</entry>
                      <entry>25</entry>
                    </row>

                    <row>
                      <entry>100</entry>
                      <entry>66</entry>
                      <entry>1.5</entry>
                      <entry>33</entry>
                    </row>

                    <row>
                      <entry>120</entry>
                      <entry>60</entry>
                      <entry>2</entry>
                      <entry>30</entry>
                    </row>

                    <row>
                      <entry>133</entry>
                      <entry>66</entry>
                      <entry>2</entry>
                      <entry>33</entry>
                    </row>

                    <row>
                      <entry>150</entry>
                      <entry>60</entry>
                      <entry>2.5</entry>
                      <entry>30 (Intel, AMD)</entry>
                    </row>

                    <row>
                      <entry>150</entry>
                      <entry>75</entry>
                      <entry>2</entry>
                      <entry>37.5 (Cyrix/IBM 6x86MX)</entry>
                    </row>

                    <row>
                      <entry>166</entry>
                      <entry>66</entry>
                      <entry>2.5</entry>
                      <entry>33</entry>
                    </row>

                    <row>
                      <entry>180</entry>
                      <entry>60</entry>
                      <entry>3</entry>
                      <entry>30</entry>
                    </row>

                    <row>
                      <entry>200</entry>
                      <entry>66</entry>
                      <entry>3</entry>
                      <entry>33</entry>
                    </row>

                    <row>
                      <entry>233</entry>
                      <entry>66</entry>
                      <entry>3.5</entry>
                      <entry>33</entry>
                    </row>
                  </tbody>
                </tgroup>
              </informaltable>
              <para>Remarques:</para>
              <itemizedlist>
                <listitem>
                  <para>[a] 66MHz peut &ecirc;tre en fait 66.667MHz, mais ne pas le 
                    pr&eacute;sumer.</para>
                </listitem>
                <listitem>
                  <para>[b] Le Pentium 100 peut utiliser une horloge externe &agrave;
                    50MHz avec un coefficient multiplicateur de 2 ou &agrave; 66MHz 
                    avec un coefficient multiplicateur de 1.5.</para>
                </listitem>
              </itemizedlist>
              <para>L'id&eacute;al est donc d'avoir un processeur &agrave; 100,
                133, 166, 200 ou 233, sinon qu'avec un coefficient 
                multiplicateur de 3 et plus, le processeur attend apr&egrave;s
                la m&eacute;moire.</para>
              
            </sect4>
            
            <sect4>
              <title>Le bogue de l'AMD K6</title>
              
              <para>En 1997, on a rapport&eacute; des probl&egrave;mes d'erreurs d'acc&egrave;s
                &agrave; la m&eacute;moire lors de compilations intensives avec l'AMD K6.
                Le probl&egrave;me a &eacute;t&eacute; r&eacute;gl&eacute; au troisi&egrave;me trimestre 97. D'apr&egrave;s
                les rapports, les puces K6 dont la date de fabrication est
                &ldquo;9733&rdquo; ou plus (i.e., produites &agrave; partir de la 
                33&egrave;me semaine de 97) n'ont plus ce probl&egrave;me.</para>
              
            </sect4>
          </sect3>
          
          <sect3>
            <title>* 486</title>
            <para></para>
          </sect3>
          
          <sect3>
            <title>* 386</title>
            <para></para>
          </sect3>
          
          <sect3>
            <title>286</title>
            
            <para>D&eacute;sol&eacute;, FreeBSD ne tourne pas sur des machines 80286. Il est
              quasiment impossible de faire tourner les UNIXs cons&eacute;quents et
              dot&eacute;s de fonctionnalit&eacute;s compl&egrave;tes d'aujourd'hui sur de telles
              machines.</para>
            
          </sect3>
        </sect2>
        
        <sect2>
          <title>M&eacute;moire</title>
          
          <para>Il vous faudra au moins 5 MB de m&eacute;moire pour pouvoir installer 
            FreeBSD. Une fois votre syst&egrave;me en &eacute;tat de marche, vous pouvez
            <link linkend="kernelconfig-building">recompiler un noyau</link>
            qui utilisera moins de m&eacute;moire. Avec <filename>boot4.flp</filename>
            vous pouvez vous en sortir avec seulement 4 MB.</para>
          
        </sect2>
        
        <sect2>
          <title>* BIOS</title>

          <para></para>
        </sect2>
      </sect1>
      
      <sect1 id="hw-io">
        <title>*** P&eacute;riph&eacute;riques d'Entr&eacute;e/Sortie</title>

<!--
        &trans.a.dntt;
-->
        
        <sect2>
          <title>* Cartes graphiques</title>

          <para></para>
        </sect2>
        
        <sect2>
          <title>* Cartes son</title>

          <para></para>
        </sect2>
        
        <sect2>
          <title>*** Ports s&eacute;rie et cartes multi-ports</title>

          <sect3 id="uart">
             <title>*** L'UART : Ce que c'est et comment il fonctionne</title>
             &sgml.todo;

<!--
             <para><emphasis>Copyright &copy; 1996 &a.uhclem;, tous droits
             r&eacute;serv&eacute;s.  13 janvier 1996.</emphasis></para>
        
             <para>Le contr&ocirc;leur du r&eacute;cepteur/&eacute;metteur 
             asynchrone universel <emphasis>Universal 
             Asynchronous Receiver/Transmitter (UART)</emphasis>
             est le composant clef du sous-syst&egrave;me de communication
             d'un ordinateur. L'UART prend des octets de donn&eacute;es et
             transmet des bits individuels d'un mani&egrave;re s&eacute;quentielle.
             Au point de destination, un second UART re-assemble les
             bits en octets complets.</para>
            
             <para>Les transmissions s&eacute;ries sont habituellement
             utilis&eacute;es avec les modems et pour les communications
             non-r&eacute;seaux entre les ordinateurs, terminaux et autres
             p&eacute;riph&eacute;riques.</para>
            
             <para>Il y a deux formes de transmission s&eacute;rie : synchrone
             et asynchrone. D&eacute;pendant du mode que votre mat&eacute;riel
             supporte, le nom d'un sous-syst&egrave;me de communication indiquera
             <literal>A</literal> s'il supporte les communications
             asynchrone et <literal>S</literal> s'il supporte les
             communications synchrones. Les deux formes sont d&eacute;crites
             ci-dessous :</para>
            
             <para>Les acronymes les plus courants sont :
          
             <blockquote>
               <para>UART <emphasis>Universal Asynchronous 
               Receiver/Transmitter></emphasis> : R&eacute;cepteur/Emetteur
               universel asynchrone</para>
             </blockquote>
          
             <blockquote>
               <para>USART <emphasis>Universal SYnchronous-Asynchronous 
               Receiver/Transmitter></emphasis> : R&eacute;cepteur/Emetteur
               universel synchrone-asynchrone</para>
             </blockquote></para>

             <sect4>
               <title>Transmission s&eacute;rie synchrones</title>
          
               <para>Les transmissions s&eacute;ries synchrones n&eacute;cessite le
               fait que l'&eacute;metteur et le r&eacute;cepteur partage une horloge
               commune, ou que l'envoyeur fournisse un signal
               d'&eacute;chantillonage 
               ou n'importe quel autre signal temporel afin que le
               r&eacute;cepteur sache quand est-ce qu'il doit
               &ldquo;lire&rdquo; le prochain bit de donn&eacute;es. Dans la
               plupart des formes de communication synchrone, s'il
               n'y a pas de donn&eacute;es disponibles au moment o&ugrave; il faut
               envoyer, un caract&egrave;re de remplissage doit &ecirc;tre envoy&eacute; &agrave;
               la place de telle sorte qu'il y ait toujours des
               donn&eacute;es en cours de transmission.
               La communication synchrone est souvent plus efficace
               parce seulement des bits de donn&eacute;es sont transmis entre
               l'&eacute;metteur et le r&eacute;cepteur, et la communication
               synchrone peut co&ucirc;ter plus cher si du cablage et
               circuit suppl&eacute;mentaires sont n&eacute;cessaires pour partager
               un signal d'horloge entre l'&eacute;metteur et le
               r&eacute;cepteur.</para> 
              
              <para>Une forme de transmission synchrone est celle 
              utilis&eacute;e par les imprimantes et les p&eacute;riph&eacute;riques disques
              non amovibles dans lesquels la donn&eacute;e est envoy&eacute; &agrave; un
              ensemble de c&acirc;bles pendant que l'horloge ou le signal
              d'&eacute;chantillonage est envoy&eacute; par un c&acirc;ble diff&eacute;rent.
              Les imprimantes et les p&eacute;riph&eacute;riques de disques fixes ne
              sont normalement pas des p&eacute;riph&eacute;riques s&eacute;rie parce que la
              plupart des interfaces standards de disques fixes envoient
              un mot entier de donn&eacute;e par signal d'horloge ou
              d'&eacute;chantillonage en utilisant un c&acirc;ble s&eacute;par&eacute; pour chaque
              bit du mot. Dans l'industrie du PC, cela est appel&eacute;
              p&eacute;riph&eacute;rique parall&egrave;le.</para>
                  
              <para>Le mat&eacute;riel de communication s&eacute;rie standard dans un PC
              ne supporte pas les op&eacute;rations synchrones. Ce mode est crit
              ici dans un seul but de comparaison.</para>
           </sect4>
        
           <sect4>
             <title>Transmission parall&egrave;le asynchrone</title>
          
             <para>La transmission asynchrone parall&egrave;le permet
             au donn&eacute;es d'&ecirc;tre transmises sans que l'&eacute;metteur n'ait &agrave;
             envoyer une horloge au r&eacute;cepteur.
             A la place, l'&eacute;metteur et le r&eacute;cepteur doivent s'entendre
             &agrave; l'avance sur les param&ecirc;tres de temps et des bits sp&eacute;ciaux
             sont ajout&eacute;s &agrave; chaque mot utilis&eacute; pour synchroniser les
             unit&eacute;s d'envoi et de r&eacute;ception.</para>
                 
             <para>Lorsqu'un mot est donn&eacute; &agrave; l'UART pour des
             transmissions asynchrones, un bit nomm&eacute; "bit de d&eacute;part"
             est ajout&eacute; au d&eacute;bute de chaque mot transmis. Le bit de
             d&eacute;part est utilis&eacute; pour avertir le r&eacute;cepteur  qu'un mot de
             donn&eacute;e est sur le point d'&ecirc;tre envoy&eacute;, et pour forcer
             l'horloge du r&eacute;cepteur &agrave; se synchroniser avec l'horloge de
             l'&eacute;metteur. Ces deux horloges doivent &ecirc;tre assez pr&eacute;cises
             afin de ne pas avoir un d&eacute;calage de fr&eacute;quence sup&eacute;rieur &agrave;
             10% durant la transmission du reste des bits du mot. (cette
             n&eacute;cessit&eacute; a &eacute;t&eacute; introduit aux jours des t&eacute;l&eacute;printers
             m&eacute;caniques et est fr&eacute;quement rencontr&eacute; par les &eacute;quipements
             &eacute;lectroniques modernes.</para>
                 
             <para>Apr&egrave;s le bit de d&eacute;part, les bits individuels du mot de
             donn&eacute;es sont envoy&eacute;s, avec le bit de poids faible 
             <emphasis> - the Least Significant Bit (LSB) - </emphasis>
             devant &ecirc;tre envoy&eacute; en premier.
             Chaque bit dans la transmission est envoy&eacute; pour exactement
             le m&ecirc;me temps que tous les autres bits, et le r&eacute;cepteur 
             &ldquo;regarde&rdquo; dans le c&acirc;ble &agrave;
	     approximativement la moiti&eacute; de la p&eacute;riode  assign&eacute;e &agrave;
	     chaque bit pour d&eacute;terminer si le bit est un 
             <literal>1</literal> ou un <literal>0</literal>.  
	     Par exemple, si cela prend deux secondes pour
	     envoyer chaque bit, le r&eacute;cepteur examinera le
	     signal pour d&eacute;terminer si c'est un <literal>1</literal> 
	     ou un <literal>0</literal> apr&egrave;s qu'une seconde
	     se soit pass&eacute;e, puis il attendra deux secondes,
	     et examinera la valeur du bit suivant, et ainsi
	     de suite.</para>
              
             <para>L'&eacute;metteur ne sait pas si le r&eacute;cepteur a
	     &ldquo;regard&eacute;&rdquo; la valeur du bit.
	     L'&eacute;metteur a seulement connaissance de
	     l'horloge lui disant de commencer &agrave; transmettre
	     le prochain bit du mot.</para>
              
             <para>Lorsqu'un mot de donn&eacute;e complet a &eacute;t&eacute;
	     envoy&eacute;, l'&eacute;metteur peut ajouter un bit de
	     parit&eacute; que l'&eacute;metteur  g&eacute;n&egrave;re. Le bit de
	     parit&eacute; peut &ecirc;tre utilis&eacute; par le r&eacute;cepteur pour
	     effectuer une v&eacute;rification d'erreur simple. 
	     Puis enfin, un bit d'arr&ecirc;t est envoy&eacute; &agrave;
	     l'emmetteur.</para>
              
             <para>Lorsque le r&eacute;cepteur recoit tous les bits
	     du mot de donn&eacute;es, il peut v&eacute;rifier les bits de
	     parit&eacute; (l'&eacute;metteur et le r&eacute;cepteur doivent
	     s'&ecirc;tre mis d'accord sur le bit de parit&eacute;
	     utilis&eacute;), puis le r&eacute;cepteur attend un bit
	     d'arr&ecirc;t. Si le bit d'arr&ecirc;t n'apparait pas au
	     moment o&ugrave; il est suppos&eacute; le faire, l'UART
	     consid&egrave;re que le mot complet est erron&eacute; et
	     reportera une erreur de fen&ecirc;tre au processeur
	     h&ocirc;te lorsque le mot de donn&eacute;e est lu. La cause
	     usuelle d'une erreur de fen&ecirc;tre est lorsque
	     l'horloge de l'&eacute;metteur et du r&eacute;cepteur ne
	     tournent pas &agrave; la m&ecirc;me vitesse, et que le
	     signal a &eacute;t&eacute; interrompu.</para>
              
             <para>Ind&eacute;pendamment de si les donn&eacute;es ont &eacute;t&eacute;
	     re&ccedil;u correctement ou non, l'UART &eacute;carte
	     automatiquement les bits de d&eacute;part, d'arr&ecirc;t et de
	     parit&eacute;.
	     Si l'&eacute;metteur et le r&eacute;cepteur sont configur&eacute;s
	     identiquement, ces bits ne sont pas pass&eacute; &agrave;
	     l'h&ocirc;te.</para>
              
             <para>Si un autre mot est pr&ecirc;t pour la
	     transmission, le bit de d&eacute;part pour le nouveau mot
	     peut &ecirc;tre envoy&eacute; aussit&ocirc;t que le bit d'arr&ecirc;t pour
	     le mot pr&eacute;c&eacute;dent a &eacute;t&eacute; envoy&eacute;.</para>
              
             <para>Parce que les donn&eacute;es asynchrones sont 
	     &ldquo;auto-descriptibles&rdquo;, s'il n'y a pas
	     de donn&eacute;es &agrave; transmettre, la ligne de transmission
	     peut &ecirc;tre inactive.</para>
          </sect4>
        
          <sect4>
          <title>Autres fonctions UART</title>
          
          <para>En plus de son travail qui est de convertir
	  des donn&eacute;es de parall&egrave;le &agrave; s&eacute;rie en r&eacute;ception, un
	  UART fournit usuellement des circuits addtionnel
	  pour les signaux qui peuvent &ecirc;tre utilis&eacute;s pour
	  indiquer l'&eacute;tat de le m&eacute;dia de transmission, et
	  pour r&eacute;guler le flux de donn&eacute;es dans l'&eacute;ventualit&eacute;
	  o&ugrave; le p&eacute;riph&eacute;rique distant n'est pas pr&eacute;par&eacute; &agrave;
	  accepter plus de donn&eacute;es. Par exemple, quand le
	  p&eacute;riph&eacute;rique connect&eacute; &agrave; l'UART est un modem, le
	  modem peut reporter la pr&eacute;sence d'une portance sur
	  la ligne t&eacute;l&eacute;phonique alors que l'ordinateur peut
	  ordonner au modem de se r&eacute;initialiser ou de ne pas
	  prendre d'appel en prenant en compte ou non ces
	  signaux suppl&eacute;pmentaires. La fonction de chacun
	  de ces signaux suppl&eacute;mentaires est d&eacute;fini dans le
	  standard EIA RS232-C.</para>
        </sect4>
        
        <sect4>
          <title>Les standards RS232-C et V.24</title>
          
          <para>Dans la plupart des syst&egrave;mes informatiques,
	  l'UART est connect&eacute; au circuit g&eacute;n&eacute;rant le signal 
	  correspondant aux sp&eacute;cifications EIA RS232-C.
	  Il y a aussi un standard CCITT nomm&eacute; V.24
	  qui reprend les sp&eacute;cifications incluses dans
	  RS232-C.</para>
              
          <sect5>
            <title>Assignation de bits RS232-C (marques et
	    espaces)</title>
            
            <para>Dans la RS232-C, une valeur de <literal>1</literal> 
	    est appel&eacute;e une <literal>marque</literal> et une
	    valeur de <literal>0</literal> est appel&eacute; un
	    <literal>espace</literal>.  
	    Lorsqu'une ligne de communication est inactive,
	    la ligne est dite &ldquo;marquant&rdquo; ou
	    transmettant continuellement la valeur
	    <literal>1</literal>.</para>
                
            <para>Le bit de d&eacute;part a toujours une valeur de 
	    <literal>0</literal> (un espace). Le bit d'arr&ecirc;t
	    a toujours une valeur de <literal>1</literal>
	    (une marque). Cela signifie qu'il y aura
	    toujours une transition d'une marque (1) &agrave; un
	    espace (0) sur la ligne &agrave; chaque mot, m&ecirc;me si
	    plusieurs mots sont transmis &agrave; la suite.
	    Cela garanti que l'&eacute;metteur et le r&eacute;cepteur
	    peuvent resynchroniser leurs horloges,
	    ind&eacute;pendamment du contenu des bits de donn&eacute;es en
	    train d'&ecirc;tre transmis.</para>
                
            <para>Le temps d'inactivit&eacute; entre les bits de
	    d&eacute;part et d'arr&ecirc;t n'ont pas &agrave; &ecirc;tre exactement
	    multiple (en incluant z&eacute;ro) de la vitesse de
	    transmission de bits sur le lien de
	    communication, mais la plupart des UART sont
	    con&ccedil;us de cette mani&egrave;re pour que cela soit plus
	    simple.</para>
                
            <para>Dans la RS232-C, le signal "marquant" 
	    (un <literal>1</literal>) est repr&eacute;sent&eacute; par 
	    un voltage entre  -2 VDC (tension en courant
	    continu) et -12 VDC, et un
	    signal "d'espace" (un <literal>0</literal>)
	    est repr&eacute;sent&eacute; par un voltage entre 0 et  +12 VDC.
	    L'&eacute;metteur est suppos&eacute; envoyer du  +12 VDC ou
	    -12 VDC, et le r&eacute;cepteur, et le r&eacute;cepteur est
	    suppos&eacute; supporter une perte de voltage sur les
	    longs c&acirc;bles. 
	    Certains &eacute;metteurs avec des p&eacute;riph&eacute;riques de
	    basse &eacute;nergie (comme les ordinateurs portables)
	    utilisent souvent seulement +5 VDC et -5 VDC,
	    mais ces valeurs sont toujours acceptables
	    par un r&eacute;cepteur RS232-C, du moment que la
	    longueur du c&acirc;ble soit courte.</para>
          </sect5>
              
          <sect5>
            <title>Signal de rupture RS232-C (Break Signal)</title>
            
            <para>La RS232-C sp&eacute;cifie aussi un signal 
	    appel&eacute; <literal>rupture</literal> (break),
	    qui est caus&eacute; en envoyant en continu des valeurs
	    d'espacement (ni de d&eacute;part, ni d'arr&ecirc;t).
	    Quand iul n'y a pas d'&eacute;lectricit&eacute; pr&eacute;sent sur le
	    circuit de donn&eacute;es, la ligne est consid&eacute;r&eacute;e en
	    train d'envoyer une <literal>Rupture</literal>.</para>
                
            <para>Le signal <literal>Rupture</literal> peut
	    &ecirc;tre plus long que le temps mis pour envoyer un
	    octet complet plus les bits de d&eacute;part, d'arr&ecirc;t
	    et de parit&eacute;. La plupart des UART peuvent
	    distinguer une erreur de fen&ecirc;tre et une
	    rupture, mais si l'UART ne peut pas le
	    faire, la d&eacute;tection de fen&ecirc;tre peut &ecirc;tre
	    utilis&eacute;e pour identifier les
	    ruptures.</para>
                
            <para>Aux jours du teleprinter, lorsque
	    plusieurs imprimantes &agrave; travers le monde &eacute;taient
	    c&acirc;bl&eacute;es en s&eacute;rie (comme les services de news),
	    toute unit&eacute; pouvait causer une
	    <literal>Rupture</literal> en ouvrant
	    temporairement le circuit complet.
	    Cela &eacute;tait utilis&eacute; pour pouvoir autoriser un emplacement 
	    avec des nouvelles importante d'interrompre un autre
	    emplacement en train de transmettre des informations.</para>
                
            <para>Dans les syst&egrave;mes modernes, il y a deux types de
	    signal de rupture.  Si la rupture est plus longue
	    que 1.6 secondes, on le consid&egrave;re comme une
	    "rupture de modem", et certains modems
	    peuvent &ecirc;tre programm&eacute;s pour terminer la conversation
	    et pour raccrocher ou entrer dans le mode de commande du
	    modem quand le modem d&eacute;tecte ce signal.
	    Si l'rupture est plus courte que 1.6 secondes, cela
	    signifie une rupture de donn&eacute;es et il appartient &agrave;
	    l'ordinateur distant de r&eacute;pondre &agrave; ce signal.
	    Parfois cette forme de rupture est utilis&eacute;e comme
	    signal d'attention ou d'interruption et est parfois 
	    accept&eacute;e comme remplacement pour le caract&egrave;re 
	    ASCII CONTROL-C.</para>
                
            <para>Les marques et espaces sont &eacute;quivalents aux
	    &ldquo;trous&rdquo; et &ldquo;non trou&rdquo; dans les
	    syst&egrave;mes &agrave; cartes perfor&eacute;es.</para>

            <note>
              <para>Les ruptures ne peuvent &ecirc;tre g&eacute;n&eacute;r&eacute;es  depuis des
	      cartes perfor&eacute;es ou n'importe auelle autre valeur
	      d'octet, puisque les octets sont toujours envoy&eacute;s avec
	      un bit de d&eacute;part et un bit d'arr&ecirc;t.
	      L'UART est habituellement capable de  produire du signal
	      d'espacement en r&eacute;ponse &agrave; une commande sp&eacute;ciale 
	      du processeur h&ocirc;te.</para>
            </note>
          </sect5>
          
          <sect5>
            <title>P&eacute;riph&eacute;riques RS232-C DTE et DCE</title>
            
            <para>Les sp&eacute;cifications de RS232-C d&eacute;finit deux types de
	    mat&eacute;riel : le terminal informatique (DTE) et le p&eacute;riph&eacute;rique
	    de transport de donn&eacute;es (DCE). Habituellement, le
	    p&eacute;riph&eacute;rique DTE est un terminal (ou un ordinateur) et le
	    DCE est un modem. A l'autre extr&eacute;mit&eacute; de la conversation &agrave;
	    travers la ligne t&eacute;l&eacute;phonique, le modem de r&eacute;ception est
	    aussi un p&eacute;riph&eacute;rique DCE et l'ordinateur qui est reli&eacute; &agrave;
	    ce modem est un p&eacute;riph&eacute;rique DTE.
	    Le p&eacute;riph&eacute;rique DCE re&ccedil;oit des signaux sur les broches que
	    le p&eacute;riph&eacute;rique de DTE transmet, et vice versa.</para>

            <para>Lorsque deux p&eacute;riph&eacute;riques qui sont tous deux DTE
	    ou tous deux DCE doivent &ecirc;tre connect&eacute;s ensemble sans un
	    modem ou un interm&eacute;diaire similaire entre eux, un NULL
	    modem doit &ecirc;tre utilis&eacute;. Le NULL modem r&eacute;organize
	    &eacute;lectriquement le c&acirc;blage de telle sorte que la sortie
	    de l'&eacute;metteur soit reli&eacute;e au r&eacute;cepteur sur l'autre
	    p&eacute;riph&eacute;rique et vice versa.
	    Des traductions semblables sont ex&eacute;cut&eacute;es sur tous les
	    signaux de contr&ocirc;le de sorte que chaque p&eacute;riph&eacute;rique
	    voit ce qu'il pense &ecirc;tre du signal DCE (ou DTE) depuis
	    l'autre p&eacute;riph&eacute;rique.</para>
                
            <para>Le nombre de signaux produits par les p&eacute;riph&eacute;riques
	    DTE et DCE ne sont pas sym&eacute;triques.
	    Le p&eacute;riph&eacute;rique DTE produit moins de signaux pour le
	    p&eacute;riph&eacute;rique DCE que le p&eacute;riph&eacute;rique DTE re&ccedil;oit depuis le
	    DCE.</para>
          </sect5>
          
          <sect5>
            <title>Assignation des broches RS232-C</title>
            
            <para>Les sp&eacute;cifications EIA RS232-C (et l'&eacute;quivalent ITU,
	    V.24) d&eacute;signe un connecteur 25 broches (habituellement un
	    DB25) et d&eacute;finit le but de la plupart des broches dans ce
	    connecteur.</para>
                
            <para>Dans l'IBM PC et les syst&egrave;mes
	    semblables, un sous-ensemble de signaux de
	    RS232-C est fourni par l'interm&eacute;diaire de neuf
	    connecteurs &agrave; broches (DB9). Les signaux qui ne
	    sont pas inclus sur le connecteur PC, travaille
	    principalement avec l'ex&eacute;cution synchrone, et ce
	    mode de transmission n'est pas support&eacute;s par 
	    l'UART qu'IBM a choisi pour l'usage dans l'IBM 
	    PC.</para>
                
            <para>Selon le constructeur, un DB25, un
	    DB9, ou les deux types de connecteur peuvent
	    &ecirc;tre utilis&eacute;s pour des transmissions de RS232-C. 
	    (l'IBM PC utilise &eacute;galement un connecteur
	    DB25 pour l'interface parall&egrave;le pour 
	    imprimante ce qui pr&ecirc;te &agrave; certaines 
	    confusions).</para>
                
            <para>Ci-dessous, une table des affectations de signaux
	    des connecteurs DB25 et DB9 dans RS232-C</para>

            <informaltable frame="none">
              <tgroup cols="7">
                <thead>
                  <row>
                    <entry>Broche DB25 RS232-C</entry>
                    <entry>Broche DB9 IBM PC</entry>
                    <entry>EIA Circuit Symbol</entry>
                    <entry>CCITT Circuit Symbol</entry>
                    <entry>Common Name</entry>
                    <entry>Signal Source</entry>
                    <entry>Description</entry>
                  </row>
                </thead>
                
                <tbody>
                  <row>
                    <entry>1</entry>
                    <entry>-</entry>
                    <entry>AA</entry>
                    <entry>101</entry>
                    <entry>PG/FG</entry>
                    <entry>-</entry>
                    <entry>Frame/Protective Ground</entry>
                  </row>
                  
                  <row>
                    <entry>2</entry>
                    <entry>3</entry>
                    <entry>BA</entry>
                    <entry>103</entry>
                    <entry>TD</entry>
                    <entry>DTE</entry>
                    <entry>Transmission donn&eacute;es</entry>
                  </row>
                  
                  <row>
                    <entry>3</entry>
                    <entry>2</entry>
                    <entry>BB</entry>
                    <entry>104</entry>
                    <entry>RD</entry>
                    <entry>DCE</entry>
                    <entry>Reception donn&eacute;es</entry>
                  </row>
                  
                  <row>
                    <entry>4</entry>
                    <entry>7</entry>
                    <entry>CA</entry>
                    <entry>105</entry>
                    <entry>RTS</entry>
                    <entry>DTE</entry>
                    <entry>Request to Send</entry>
                  </row>
                  
                  <row>
                    <entry>5</entry>
                    <entry>8</entry>
                    <entry>CB</entry>
                    <entry>106</entry>
                    <entry>CTS</entry>
                    <entry>DCE</entry>
                    <entry>Clear to Send</entry>
                  </row>
                  
                  <row>
                    <entry>6</entry>
                    <entry>6</entry>
                    <entry>CC</entry>
                    <entry>107</entry>
                    <entry>DSR</entry>
                    <entry>DCE</entry>
                    <entry>Data Set Ready</entry>
                  </row>
                  
                  <row>
                    <entry>7</entry>
                    <entry>5</entry>
                    <entry>AV</entry>
                    <entry>102</entry>
                    <entry>SG/GND</entry>
                    <entry>-</entry>
                    <entry>Signal Ground</entry>
                  </row>
                  
                  <row>
                    <entry>8</entry>
                    <entry>1</entry>
                    <entry>CF</entry>
                    <entry>109</entry>
                    <entry>DCD/CD</entry>
                    <entry>DCE</entry>
                    <entry>Data Carrier Detect</entry>
                  </row>
                  
                  <row>
                    <entry>9</entry>
                    <entry>-</entry>
                    <entry>-</entry>
                    <entry>-</entry>
                    <entry>-</entry>
                    <entry>-</entry>
                    <entry>Reserved for Test</entry>
                  </row>
                  
                  <row>
                    <entry>10</entry>
                    <entry>-</entry>
                    <entry>-</entry>
                    <entry>-</entry>
                    <entry>-</entry>
                    <entry>-</entry>
                    <entry>Reserved for Test</entry>
                  </row>
                  
                  <row>
                    <entry>11</entry>
                    <entry>-</entry>
                    <entry>-</entry>
                    <entry>-</entry>
                    <entry>-</entry>
                    <entry>-</entry>
                    <entry>Reserved for Test</entry>
                  </row>
                  
                  <row>
                    <entry>12</entry>
                    <entry>-</entry>
                    <entry>CI</entry>
                    <entry>122</entry>
                    <entry>SRLSD</entry>
                    <entry>DCE</entry>
                    <entry>Sec. Recv. Line Signal Detector</entry>
                  </row>
                  
                  <row>
                    <entry>13</entry>
                    <entry>-</entry>
                    <entry>SCB</entry>
                    <entry>121</entry>
                    <entry>SCTS</entry>
                    <entry>DCE</entry>
                    <entry>Secondary Clear to Send</entry>
                  </row>
                  
                  <row>
                    <entry>14</entry>
                    <entry>-</entry>
                    <entry>SBA</entry>
                    <entry>118</entry>
                    <entry>STD</entry>
                    <entry>DTE</entry>
                    <entry>Secondary Transmit Data</entry>
                  </row>
                  
                  <row>
                    <entry>15</entry>
                    <entry>-</entry>
                    <entry>DB</entry>
                    <entry>114</entry>
                    <entry>TSET</entry>
                    <entry>DCE</entry>
                    <entry>Trans. Sig. Element Timing</entry>
                  </row>
                  
                  <row>
                    <entry>16</entry>
                    <entry>-</entry>
                    <entry>SBB</entry>
                    <entry>119</entry>
                    <entry>SRD</entry>
                    <entry>DCE</entry>
                    <entry>Secondary Received Data</entry>
                  </row>
                  
                  <row>
                    <entry>17</entry>
                    <entry>-</entry>
                    <entry>DD</entry>
                    <entry>115</entry>
                    <entry>RSET</entry>
                    <entry>DCE</entry>
                    <entry>Receiver Signal Element Timing</entry>
                  </row>
                  
                  <row>
                    <entry>18</entry>
                    <entry>-</entry>
                    <entry>-</entry>
                    <entry>141</entry>
                    <entry>LOOP</entry>
                    <entry>DTE</entry>
                    <entry>Local Loopback</entry>
                  </row>
                  
                  <row>
                    <entry>19</entry>
                    <entry>-</entry>
                    <entry>SCA</entry>
                    <entry>120</entry>
                    <entry>SRS</entry>
                    <entry>DTE</entry>
                    <entry>Secondary Request to Send</entry>
                  </row>
                  
                  <row>
                    <entry>20</entry>
                    <entry>4</entry>
                    <entry>CD</entry>
                    <entry>108.2</entry>
                    <entry>DTR</entry>
                    <entry>DTE</entry>
                    <entry>Data Terminal Ready</entry>
                  </row>
                  
                  <row>
                    <entry>21</entry>
                    <entry>-</entry>
                    <entry>-</entry>
                    <entry>-</entry>
                    <entry>RDL</entry>
                    <entry>DTE</entry>
                    <entry>Remote Digital Loopback</entry>
                  </row>
                  
                  <row>
                    <entry>22</entry>
                    <entry>9</entry>
                    <entry>CE</entry>
                    <entry>125</entry>
                    <entry>RI</entry>
                    <entry>DCE</entry>
                    <entry>Ring Indicator</entry>
                  </row>
                  
                  <row>
                    <entry>23</entry>
                    <entry>-</entry>
                    <entry>CH</entry>
                    <entry>111</entry>
                    <entry>DSRS</entry>
                    <entry>DTE</entry>
                    <entry>Data Signal Rate Selector</entry>
                  </row>
                  
                  <row>
                    <entry>24</entry>
                    <entry>-</entry>
                    <entry>DA</entry>
                    <entry>113</entry>
                    <entry>TSET</entry>
                    <entry>DTE</entry>
                    <entry>Trans. Sig. Element Timing</entry>
                  </row>
                  
                  <row>
                    <entry>25</entry>
                    <entry>-</entry>
                    <entry>-</entry>
                    <entry>142</entry>
                    <entry>-</entry>
                    <entry>DCE</entry>
                    <entry>Test Mode</entry>
                  </row>
                </tbody>
              </tgroup>
            </informaltable>
          </sect5>
        </sect4>
        
        <sect4>
          <title>Bits, Baud et symboles</title>
          
          <para>Le baud est une mesure de vitesse de transmission dans la
	  liaison asynchrone. En raison de l'avance en technologie 
	  de transmission par modem, ce terme est fr&eacute;quemment 
	  employ&eacute; abusivement pour d&eacute;signer le d&eacute;bit dans les nouveaux
	  p&eacute;riph&eacute;riques.</para>
              
          <para>Traditionnellement, un d&eacute;bit en baud repr&eacute;sente le
	  nombre de bits qui sont envoy&eacute;s r&eacute;ellement
	  par le m&eacute;dia, pas la quantit&eacute; des donn&eacute;es qui sont
	  d&eacute;plac&eacute;es r&eacute;ellement d'un dispositif de
	  DTE &agrave; l'autre. Le comptage en baud inclut les bits
	  suppl&eacute;mentaires de d&eacute;part, d'arr&ecirc;t et de parit&eacute; qui sont
	  produits par l'UART d' envoi et retir&eacute;s par l'UART de
	  r&eacute;ception. Ceci signifie que les mots de sept bits
	  de donn&eacute;es prennent r&eacute;ellement 10 bits &agrave; transmettre. 
	  Par cons&eacute;quent, un modem
	  capable de d&eacute;placer 300 bits par seconde d'un endroit &agrave;
	  l'autre peut normalement seulement
	  d&eacute;placer 30 mots de 7 bit si la parit&eacute; est utilis&eacute;e et un bit
	  d'arr&ecirc;t et de de d&eacute;but sont pr&eacute;sent.</para>

          <para>Si des mots contenant des donn&eacute;es de 8 bits sont
	  utilis&eacute;s et si des bits de parit&eacute; sont &eacute;galement
	  utilis&eacute;s, le d&eacute;bit tombe &agrave; 27.27 mots par seconde, parce que
	  cela prend maintenant 11 bits pour envoyer les mots 
	  de 8 bits, et le modem envoie toujours seulement 
	  300 bits par seconde.</para>
              
          <para>La formule pour convertir des octets par seconde
	  en une vitesse baud et vice versa &eacute;tait 
	  simple jusqu'&agrave; ce que les modems correcteurs d'erreurs
	  soient arriv&eacute;s. Ces modems re&ccedil;oivent
	  une s&eacute;rie de bits de l'UART dans l'ordinateur
	  h&ocirc;te (m&ecirc;me lorsque des modems internes
	  sont utilis&eacute;s les donn&eacute;es sont encore 
	  fr&eacute;quemment arrang&eacute;s en s&eacute;rie) et convertissent les bits de
	  nouveau en octets. Ces octets sont alors combin&eacute;s dans
	  des paquets et envoy&eacute;s par
	  de la ligne t&eacute;l&eacute;phonique en utilisant une m&eacute;thode de
	  transmission synchrone. Ceci signifie que les
	  bits de parit&eacute;, d'arr&ecirc;t, de d&eacute;but, ajout&eacute;s par l'UART dans
	  le DTE (l'ordinateur) ont &eacute;t&eacute; retir&eacute;s par le
	  modem avant transmission par le modem d' envoi. Quand ces
	  octets sont re&ccedil;us par le modem distant, le modem 
	  distant ajoute des bits de parit&eacute;, de d&eacute;but, d'arr&ecirc;t 
	  aux mots, les convertit en format s&eacute;quentiel et puis 
	  les envoie &agrave; l'UART de r&eacute;ception
	  dans l'ordinateur distant, qui &eacute;limine
	  alors les bits de d&eacute;but, d'arr&ecirc;t et 
	  de parit&eacute;.</para>
              
          <para>La raison pour laquelle toutes ces conversions 
	  suppl&eacute;mentaires sont effectu&eacute;es est de faire de la sorte
	  que les deux modems
	  puissent ex&eacute;cuter les corrections d'erreurs, qui signifie que
	  le modem de r&eacute;ception peut demander au
	  modem d'envoi de renvoyer un bloc de donn&eacute;es qui n'aurait
	  pas &eacute;t&eacute; re&ccedil;u avec un total de
	  contr&ocirc;le correct. Ce contr&ocirc;le est manipul&eacute; par les
	  modems, et les p&eacute;riph&eacute;riques DTE sont ignorant du processus
	  qui s'est produit.</para>
              
          <para>En sautant les bits de parit&eacute;, de d&eacute;but et d'arr&ecirc;t,
	  les bits suppl&eacute;mentaires des donn&eacute;es - que les deux
	  modems doivent partager entre eux pour effectuer des
	  corrections d'erreurs - sont le plus souvent cach&eacute;s lors du calcul
	  du d&eacute;bit de transmission effectivement vu par le mat&eacute;riel
	  DTE d'envoi et de r&eacute;ception.
	  Par exemple, si un modem envoie dix mots de 7 bit &agrave; un
	  autre modem sans compter les bits de
	  parit&eacute;, de d&eacute;but et d'arr&ecirc;t, le modem d' envoi pourra
	  ajouter 30 bits d'information que le
	  modem de r&eacute;ception pourra employer pour faire de correction
	  d'erreurs, ceci sans influencer la vitesse de
	  transmission des vraies donn&eacute;es.</para>
        
          <para>L'utilisation du terme baud est rendu 
	  encore confus
	  par les modems ex&eacute;cutant une compression.
	  Un mot simple de 8 bits &agrave; travers une ligne
	  t&eacute;l&eacute;phonique pourrait en r&eacute;alit&eacute; &ecirc;tre une douzaine de mots
	  qui ont &eacute;t&eacute; transmis au modem d'envoi. Le modem de r&eacute;ception
	  se chargera ensuite de le d&eacute;compresser et lui rendre son
	  contenu inital qu'il passera au DTE de r&eacute;ception.</para>

	  </para>
	  Modern modems also include buffers that allow the rate that
            bits move across the phone line (DCE to DCE) to be a different
            speed than the speed that the bits move between the DTE and DCE on
            both ends of the conversation.  Normally the speed between the DTE
            and DCE is higher than the DCE to DCE speed because of the use of
            compression by the modems.</para>
              
          <para>Because the number of bits needed to describe a byte varied
            during the trip between the two machines plus the differing
            bits-per-seconds speeds that are used present on the DTE-DCE and
            DCE-DCE links, the usage of the term Baud to describe the overall
            communication speed causes problems and can misrepresent the true
            transmission speed.  So Bits Per Second (bps) is the correct term
            to use to describe the transmission rate seen at the DCE to DCE
            interface and Baud or Bits Per Second are acceptable terms to use
            when a connection is made between two systems with a wired
            connection, or if a modem is in use that is not performing
            error-correction or compression.</para>
              
          <para>Modern high speed modems (2400, 9600, 14,400, and 19,200bps)
            in reality still operate at or below 2400 baud, or more
            accurately, 2400 Symbols per second.  High speed modem are able to
            encode more bits of data into each Symbol using a technique called
            Constellation Stuffing, which is why the effective bits per second
            rate of the modem is higher, but the modem continues to operate
            within the limited audio bandwidth that the telephone system
            provides.  Modems operating at 28,800 and higher speeds have
            variable Symbol rates, but the technique is the same.</para>
        </sect4>
 

        <sect4>
          <title>The IBM Personal Computer UART</title>
          
          <para>Starting with the original IBM Personal Computer, IBM selected
            the National Semiconductor INS8250 UART for use in the IBM PC
            Parallel/Serial Adapter.  Subsequent generations of compatible
            computers from IBM and other vendors continued to use the INS8250
            or improved versions of the National Semiconductor UART
            family.</para>
              
          <sect5>
            <title>National Semiconductor UART Family Tree</title>
            
            <para>There have been several versions and subsequent generations
              of the INS8250 UART.  Each major version is described
              below.</para>

            <programlisting>
INS8250  -&gt; INS8250B
  \
   \
    \-&gt; INS8250A -&gt; INS82C50A
             \
              \
               \-&gt; NS16450 -&gt; NS16C450
                        \
                         \
                          \-&gt; NS16550 -&gt; NS16550A -&gt; PC16550D</programlisting>
                
            <variablelist>
              <varlistentry>
                <term>INS8250</term>

                <listitem>
                  <para>This part was used in the original IBM PC and IBM
                    PC/XT.  The original name for this part was the INS8250
                    ACE (Asynchronous Communications Element) and it is made
                    from NMOS technology.</para>
                        
                  <para>The 8250 uses eight I/O ports and has a one-byte send
                    and a one-byte receive buffer.  This original UART has
                    several race conditions and other flaws. The original IBM
                    BIOS includes code to work around these flaws, but this
                    made the BIOS dependent on the flaws being present, so
                    subsequent parts like the 8250A, 16450  or 16550 could not
                    be used in the original IBM PC or IBM PC/XT.</para>
                </listitem>
              </varlistentry>
              
              <varlistentry>
                <term>INS8250-B</term>
                
                <listitem>
                  <para>This is the slower speed of the INS8250 made from NMOS
                    technology.  It contains the same problems as the original
                    INS8250.</para>
                </listitem>
              </varlistentry>
              
              <varlistentry>
                <term>INS8250A</term>
                
                <listitem>
                  <para>An improved version of the INS8250 using XMOS
                    technology with various functional flaws corrected. The
                    INS8250A was used initially in PC clone computers by
                    vendors who used &ldquo;clean&rdquo; BIOS designs. Because
                    of the corrections in the chip, this part could not be
                    used with a BIOS compatible with the INS8250 or
                    INS8250B.</para>
                </listitem>
              </varlistentry>
              
              <varlistentry>
                <term>INS82C50A</term>
                
                <listitem>
                  <para>This is a CMOS version (low power consumption) of the
                    INS8250A and has similar functional
                    characteristics.</para>
                </listitem>
              </varlistentry>
              
              <varlistentry>
                <term>NS16450</term>
                
                <listitem>
                  <para>Same as NS8250A with improvements so it can be used
                    with faster CPU bus designs.  IBM used this part in the
                    IBM AT and updated the IBM BIOS to no longer rely on the
                    bugs in the INS8250.</para>
                </listitem>
              </varlistentry>
              
              <varlistentry>
                <term>NS16C450</term>
                
                <listitem>
                  <para>This is a CMOS version (low power consumption) of the
                    NS16450.</para>
                </listitem>
              </varlistentry>
              
              <varlistentry>
                <term>NS16550</term>
                
                <listitem>
                  <para>Same as NS16450 with a 16-byte send and receive buffer
                    but the buffer design was flawed and could not be reliably
                    be used.</para>
                </listitem>
              </varlistentry>
              
              <varlistentry>
                <term>NS16550A</term>
                
                <listitem>
                  <para>Same as NS16550 with the buffer flaws corrected. The
                    16550A and its successors have become the most popular
                    UART design in the PC industry, mainly due it its ability
                    to reliably handle higher data rates on operating systems
                    with sluggish interrupt response times.</para>
                </listitem>
              </varlistentry>
              
              <varlistentry>
                <term>NS16C552</term>
                
                <listitem>
                  <para>This component consists of two NS16C550A CMOS UARTs in
                    a single package.</para>
                </listitem>
              </varlistentry>
              
              <varlistentry>
                <term>PC16550D</term>
                
                <listitem>
                  <para>Same as NS16550A with subtle flaws corrected. This is
                    revision D of the 16550 family and is the latest design
                    available from National Semiconductor.</para>
                </listitem>
              </varlistentry>
            </variablelist>
          </sect5>
          
          <sect5>
            <title>The NS16550AF and the PC16550D are the same thing</title>
                
            <para>National reorganized their part numbering system a few years
              ago, and the NS16550AFN no longer exists by that name. (If you
              have a NS16550AFN, look at the date code on the part, which is a
              four digit number that usually starts with a nine.  The first
              two digits of the number are the year, and the last two digits
              are the week in that year when the part was packaged.  If you
              have a NS16550AFN, it is probably a few years old.)</para>
                
            <para>The new numbers are like PC16550DV, with minor differences
              in the suffix letters depending on the package material and its
              shape.  (A description of the numbering system can be found
              below.)</para>
                
            <para>It is important to understand that in some stores, you may
              pay &#36;15(US) for a NS16550AFN made in 1990 and in the next
              bin are the new PC16550DN parts with minor fixes that National
              has made since the AFN part was in production, the PC16550DN was
              probably made in the past six months and it costs half (as low
              as &#36;5(US) in volume) as much as the NS16550AFN because they
              are readily available.</para>
                
            <para>As the supply of NS16550AFN chips continues to shrink, the
              price will probably continue to increase until more people
              discover and accept that the PC16550DN really has the same
              function as the old part number.</para>
          </sect5>
          
          <sect5>
            <title>National Semiconductor Part Numbering System</title>
            
            <para>The older  NS<replaceable>nnnnnrqp</replaceable>  part
              numbers are now of the format
              PC<replaceable>nnnnnrgp</replaceable>.</para>
                
            <para>The <replaceable>r</replaceable> is the revision field.  The
              current revision of the 16550 from National Semiconductor is
              <literal>D</literal>.</para>
                
            <para>The <replaceable>p</replaceable> is the package-type field.
              The types are:</para>
            
            <informaltable frame="none">
              <tgroup cols="3">
                <tbody>
                  <row>
                    <entry>"F"</entry>
                    <entry>QFP</entry>
                    <entry>(quad flat pack) L lead type</entry>
                  </row>
                  
                  <row>
                    <entry>"N"</entry>
                    <entry>DIP</entry>
                    <entry>(dual inline package) through hole straight lead
                      type</entry>
                  </row>

                  <row>
                    <entry>"V"</entry>
                    <entry>LPCC</entry>
                    <entry>(lead plastic chip carrier) J lead type</entry>
                  </row>
                </tbody>
              </tgroup>
            </informaltable>
            
            <para>The <replaceable>g</replaceable> is the product grade field.
              If an <literal>I</literal> precedes the package-type letter, it
              indicates an &ldquo;industrial&rdquo; grade part, which has
              higher specs than a standard part but not as high as Military
              Specification (Milspec) component.  This is an optional
              field.</para>
                
            <para>So what we used to call a NS16550AFN (DIP Package) is now
              called a PC16550DN or PC16550DIN.</para>
          </sect5>
        </sect4>
        
        <sect4>
          <title>Other Vendors and Similar UARTs</title>
          
          <para>Over the years, the 8250, 8250A, 16450 and 16550 have been
            licensed or copied by other chip vendors.  In the case of the
            8250, 8250A and 16450, the exact circuit (the
            &ldquo;megacell&rdquo;) was licensed to many vendors, including
            Western Digital and Intel. Other vendors reverse-engineered the
            part or produced emulations that had similar behavior.</para>
              
          <para>In internal modems, the modem designer will frequently emulate
            the 8250A/16450 with the modem microprocessor, and the emulated
            UART will frequently have a hidden buffer consisting of several
            hundred bytes.  Because of the size of the buffer, these
            emulations can be as reliable as a 16550A in their ability to
            handle high speed data.  However, most operating systems will
            still report that the UART is only a 8250A or 16450, and may not
            make effective use of the extra buffering present in the emulated
            UART unless special drivers are used.</para>
              
          <para>Some modem makers are driven by market forces to abandon a
            design that has hundreds of bytes of buffer and instead use a
            16550A UART so that the product will compare favorably in market
            comparisons even though the effective performance may be lowered
            by this action.</para>
              
          <para>A common misconception is that all parts with
            &ldquo;16550A&rdquo; written on them are identical in performance.
            There are differences, and in some cases, outright flaws in most
            of these 16550A clones.</para>
              
          <para>When the NS16550 was developed, the National Semiconductor
            obtained several patents on the design and they also limited
            licensing, making it harder for other vendors to provide a chip
            with similar features.  Because of the patents, reverse-engineered
            designs and emulations had to avoid infringing the claims covered
            by the patents.  Subsequently, these copies almost never perform
            exactly the same as the NS16550A or PC16550D, which are the parts
            most computer and modem makers want to buy but are sometimes
            unwilling to pay the price required to get the genuine
            part.</para>
              
          <para>Some of the differences in the clone 16550A parts are
            unimportant, while others can prevent the device from being used
            at all with a given operating system or driver.  These differences
            may show up when using other drivers, or when particular
            combinations of events occur that were not well tested or
            considered in the Windows driver.  This is because most modem
            vendors and 16550-clone makers use the Microsoft drivers from
            Windows for Workgroups 3.11 and the Microsoft MSD utility as the
            primary tests for compatibility with the NS16550A.  This
            over-simplistic criteria means that if a different operating
            system is used, problems could appear due to subtle differences
            between the clones and genuine components.</para>
              
          <para>National Semiconductor has made available a program named
            <application>COMTEST</application> that performs compatibility
            tests independent of any OS drivers.  It should be remembered that
            the purpose of this type of program is to demonstrate the flaws in
            the products of the competition, so the program will report major
            as well as extremely subtle differences in behavior in the part
            being tested.</para>
              
          <para>In a series of tests performed by the author of this document
            in 1994, components made by National Semiconductor, TI, StarTech,
            and CMD as well as megacells and emulations embedded in internal
            modems were tested with COMTEST.  A difference count for some of
            these components is listed below. Because these tests were
            performed in 1994, they may not reflect the current performance of
            the given product from a vendor.</para>
              
          <para>It should be noted that COMTEST normally aborts when an
            excessive number or certain types of problems have been detected.
            As part of this testing, COMTEST was modified so that it would not
            abort no matter how many differences were encountered.</para>

          <informaltable frame="none">
            <tgroup cols="3">
              <thead>
                <row>
                  <entry>Vendor</entry>
                  <entry>Part Number</entry>
                  <entry>Errors (aka "differences" reported)</entry>
                </row>
              </thead>
              
              <tbody>
                <row>
                  <entry>National</entry>
                  <entry>(PC16550DV)</entry>
                  <entry>0</entry>
                </row>
                
                <row>
                  <entry>National</entry>
                  <entry>(NS16550AFN)</entry>
                  <entry>0</entry>
                </row>
                
                <row>
                  <entry>National</entry>
                  <entry>(NS16C552V)</entry>
                  <entry>0</entry>
                </row>
                
                <row>
                  <entry>TI</entry>
                  <entry>(TL16550AFN)</entry>
                  <entry>3</entry>
                </row>
                
                <row>
                  <entry>CMD</entry>
                  <entry>(16C550PE)</entry>
                  <entry>19</entry>
                </row>
                
                <row>
                  <entry>StarTech</entry>
                  <entry>(ST16C550J)</entry>
                  <entry>23</entry>
                </row>
                
                <row>
                  <entry>Rockwell</entry>
                  <entry>Reference modem with internal 16550 or an
                    emulation (RC144DPi/C3000-25)</entry>
                  <entry>117</entry>
                </row>
                
                <row>
                  <entry>Sierra</entry>
                  <entry>Modem with an internal 16550
                    (SC11951/SC11351)</entry>
                  <entry>91</entry>
                </row>
              </tbody>
            </tgroup>
          </informaltable>
          
          <note>
            <para>To date, the author of this document has not found any
              non-National parts that report zero differences using the
              COMTEST program.  It should also be noted that National has had
              five versions of the 16550 over the years and the newest parts
              behave a bit differently than the classic NS16550AFN that is
              considered the benchmark for functionality.  COMTEST appears to
              turn a blind eye to the differences within the National product
              line and reports no errors on the National parts (except for the
              original 16550) even when there are official erratas that
              describe bugs in the A, B and C revisions of the parts, so this
              bias in COMTEST must be taken into account.</para>
          </note>
          
          <para>It is important to understand that a simple count of
            differences from COMTEST does not reveal a lot about what
            differences are important and which are not.  For example, about
            half of the differences reported in the two modems listed above
            that have internal UARTs were caused by the clone UARTs not
            supporting five- and six-bit character modes.  The real 16550,
            16450, and 8250 UARTs all support these modes and COMTEST checks
            the functionality of these modes so over fifty differences are
            reported.  However, almost no modern modem supports five- or
            six-bit characters, particularly those with error-correction and
            compression capabilities.  This means that the differences related
            to five- and six-bit character modes can be discounted.</para>
              
          <para>Many of the differences COMTEST reports have to do with
            timing.  In many of the clone designs, when the host reads from
            one port, the status bits in some other port may not update in the
            same amount of time (some faster, some slower) as a
            <emphasis>real</emphasis> NS16550AFN and COMTEST looks for these
            differences.  This means that the number of differences can be
            misleading in that one device may only have one or two differences
            but they are extremely serious, and some other device that updates
            the status registers faster or slower than the reference part
            (that would probably never affect the operation of a properly
            written driver) could have dozens of differences reported.</para>
              
          <para>COMTEST can be used as a screening tool to alert the
            administrator to the presence of potentially incompatible
            components that might cause problems or have to be handled as a
            special case.</para>
              
          <para>If you run COMTEST on a 16550 that is in a modem or a modem is
            attached to the serial port, you need to first issue a ATE0&amp;W
            command to the modem so that the modem will not echo any of the
            test characters.  If you forget to do this, COMTEST will report at
            least this one difference:</para>

          <screen>Error (6)...Timeout interrupt failed: IIR = c1  LSR = 61</screen>
        </sect4>
        
        <sect4>
          <title>8250/16450/16550 Registers</title>
          
          <para>The 8250/16450/16550 UART occupies eight contiguous I/O port
            addresses.  In the IBM PC, there are two defined locations for
            these eight ports and they are known collectively as COM1 and
            COM2.  The makers of PC-clones and add-on cards have created two
            additional areas known as COM3 and COM4, but these extra COM ports
            conflict with other hardware on some systems.  The most common
            conflict is with video adapters that provide IBM 8514
            emulation.</para>
              
          <para>COM1 is located from 0x3f8 to 0x3ff and normally uses IRQ 4
            COM2 is located from 0x2f8 to 0x2ff and normally uses IRQ 3 COM3
            is located from 0x3e8 to 0x3ef and has no standardized IRQ COM4 is
            located from 0x2e8 to 0x2ef and has no standardized IRQ.</para>
              
          <para>A description of the I/O ports of the 8250/16450/16550 UART is
            provided below.</para>

          <informaltable frame="none">
            <tgroup cols="3">
              <thead>
                <row>
                  <entry>I/O Port</entry>
                  <entry>Access Allowed</entry>
                  <entry>Description</entry>
                </row>
              </thead>
              
              <tbody>
                <row>
                  <entry>+0x00</entry>
                  <entry>write (DLAB==0)</entry>
                  <entry><para>Transmit Holding Register
                      (THR).</para><para>Information written to this port are
                      treated as data words and will be transmitted by the
                      UART.</para></entry>
                </row>
                
                <row>
                  <entry>+0x00</entry>
                  <entry>read (DLAB==0)</entry>
                  <entry><para>Receive Buffer Register (RBR).</para><para>Any
                      data words received by the UART form the serial link are
                      accessed by the host by reading this
                      port.</para></entry>
                </row>
                
                <row>
                  <entry>+0x00</entry>
                  <entry>write/read (DLAB==1)</entry>
                  <entry><para>Divisor Latch LSB (DLL)</para><para>This value
                      will be divided from the master input clock (in the IBM
                      PC, the master clock is 1.8432MHz) and the resulting
                      clock will determine the baud rate of the UART.  This
                      register holds bits 0 thru 7 of the
                      divisor.</para></entry>
                </row>
                
                <row>
                  <entry>+0x01</entry>
                  <entry>write/read (DLAB==1)</entry>
                  <entry><para>Divisor Latch MSB (DLH)</para><para>This value
                      will be divided from the master input clock (in the IBM
                      PC, the master clock is 1.8432MHz) and the resulting
                      clock will determine the baud rate of the UART.  This
                      register holds bits 8 thru 15 of the
                      divisor.</para></entry>
                </row>
                
                <row>
                  <entry>+0x01</entry>
                  <entry>write/read (DLAB==0)</entry>
                  <entrytbl cols="2">
                    <colspec colnum="1" colname="col1">
                    <colspec colnum="2" colname="col2">
                    <spanspec namest="col1" nameend="col2" spanname="1to2">

                    <tbody>
                      <row>
                        <entry spanname="1to2"><para>Interrupt Enable Register
                            (IER)</para><para>The 8250/16450/16550 UART
                            classifies events into one of four categories.
                            Each category can be configured to generate an
                            interrupt when any of the events occurs.  The
                            8250/16450/16550 UART generates a single external
                            interrupt signal regardless of how many events in
                            the enabled categories have occurred.  It is up to
                            the host processor to respond to the interrupt and
                            then poll the enabled interrupt categories
                            (usually all categories have interrupts enabled)
                            to determine the true cause(s) of the
                            interrupt.</para></entry>
                      </row>
                      
                      <row>
                        <entry>Bit 7</entry>
                        <entry>Reserved, always 0.</entry>
                      </row>
                      
                      <row>
                        <entry>Bit 6</entry>
                        <entry>Reserved, always 0.</entry>
                      </row>
                      
                      <row>
                        <entry>Bit 5</entry>
                        <entry>Reserved, always 0.</entry>
                      </row>
                      
                      <row>
                        <entry>Bit 4</entry>
                        <entry>Reserved, always 0.</entry>
                      </row>
                      
                      <row>
                        <entry>Bit 3</entry>
                        <entry>Enable Modem Status Interrupt (EDSSI). Setting
                          this bit to "1" allows the UART to generate an
                          interrupt when a change occurs on one or more of the
                          status lines.</entry>
                      </row>
                      
                      <row>
                        <entry>Bit 2</entry>
                        <entry>Enable Receiver Line Status Interrupt (ELSI)
                          Setting this bit to "1" causes the UART to generate
                          an interrupt when the an error (or a BREAK signal)
                          has been detected in the incoming data.</entry>
                      </row>
                      
                      <row>
                        <entry>Bit 1</entry>
                        <entry>Enable Transmitter Holding Register Empty
                          Interrupt (ETBEI) Setting this bit to "1" causes the
                          UART to generate an interrupt when the UART has room
                          for one or more additional characters that are to be
                          transmitted.</entry>
                      </row>
                      
                      <row>
                        <entry>Bit 0</entry>
                        <entry>Enable Received Data Available Interrupt
                          (ERBFI) Setting this bit to "1" causes the UART to
                          generate an interrupt when the UART has received
                          enough characters to exceed the trigger level of the
                          FIFO, or the FIFO timer has expired (stale data), or
                          a single character has been received when the FIFO
                          is disabled.</entry>
                      </row>
                    </tbody>
                  </entrytbl>
                </row>
                
                <row>
                  <entry>+0x02</entry>
                  <entry>write</entry>
                  <entrytbl cols="4">
                    <colspec colnum="1" colname="col1">
                    <colspec colnum="2" colname="col2">
                    <colspec colnum="3" colname="col3">
                    <colspec colnum="4" colname="col4">
                    <spanspec namest="col1" nameend="col4" spanname="1to4">
                    <spanspec namest="col2" nameend="col4" spanname="2to4">
                    
                    <tbody>
                      <row>
                        <entry spanname="1to4">FIFO Control Register (FCR)
                          (This port does not exist on the 8250 and 16450
                          UART.)</entry>
                      </row>

                      <row>
                        <entry>Bit 7</entry>
                        <entry spanname="2to4">Receiver Trigger Bit #1</entry>
                      </row>
                      
                      <row>
                        <entry>Bit 6</entry>
                        <entry spanname="2to4"><para>Receiver Trigger Bit
                            #0</para><para>These two bits control at what
                            point the receiver is to generate an interrupt
                            when the FIFO is active.</para></entry>
                      </row>
                      
                      <row>
                        <entry colname="col2">7</entry>
                        <entry colname="col3">6</entry>
                        <entry colname="col4">How many words are received
                          before an interrupt is generated</entry>
                      </row>
                          
                      <row>
                        <entry colname="col2">0</entry>
                        <entry colname="col3">0</entry>
                        <entry colname="col4">1</entry>
                      </row>
                      
                      <row>
                        <entry colname="col2">0</entry>
                        <entry colname="col3">1</entry>
                        <entry colname="col4">4</entry>
                      </row>
                      
                      <row>
                        <entry colname="col2">1</entry>
                        <entry colname="col3">0</entry>
                        <entry colname="col4">8</entry>
                      </row>
                      
                      <row>
                        <entry colname="col2">1</entry>
                        <entry colname="col3">1</entry>
                        <entry colname="col4">14</entry>
                      </row>
                      
                      <row>
                        <entry>Bit 5</entry>
                        <entry spanname="2to4">Reserved, always 0.</entry>
                      </row>
                      
                      <row>
                        <entry>Bit 4</entry>
                        <entry spanname="2to4">Reserved, always 0.</entry>
                      </row>
                      
                      <row>
                        <entry>Bit 3</entry>
                        <entry spanname="2to4">DMA Mode Select.  If Bit 0 is
                          set to "1" (FIFOs enabled), setting this bit changes
                          the operation of the -RXRDY and -TXRDY signals from
                          Mode 0 to Mode 1.</entry>
                      </row>
                      
                      <row>
                        <entry>Bit 2</entry>
                        <entry spanname="2to4">Transmit FIFO Reset.  When a
                          "1" is written to this bit, the contents of the FIFO
                          are discarded.  Any word currently being transmitted
                          will be sent intact.  This function is useful in
                          aborting transfers.</entry>
                      </row>
                      
                      <row>
                        <entry>Bit 1</entry>
                        <entry spanname="2to4">Receiver FIFO Reset.  When a
                          "1" is written to this bit, the contents of the FIFO
                          are discarded.  Any word currently being assembled
                          in the shift register will be received
                          intact.</entry>
                      </row>

                      <row>
                        <entry>Bit 0</entry>
                        <entry spanname="2to4">16550 FIFO Enable.  When set,
                          both the transmit and receive FIFOs are enabled.
                          Any contents in the holding register, shift
                          registers or FIFOs are lost when FIFOs are enabled
                          or disabled.</entry>
                      </row>
                    </tbody>
                  </entrytbl>
                </row>
                
                <row>
                  <entry>+0x02</entry>
                  <entry>read</entry>
                  <entrytbl cols="6">
                    <colspec colnum="1" colname="col1">
                    <colspec colnum="2" colname="col2">
                    <colspec colnum="3" colname="col3">
                    <colspec colnum="4" colname="col4">
                    <colspec colnum="5" colname="col5">
                    <colspec colnum="6" colname="col6">
                    <spanspec namest="col1" nameend="col6" spanname="1to6">
                    <spanspec namest="col2" nameend="col6" spanname="2to6">

                    <tbody>
                      <row>
                        <entry spanname="1to6">Interrupt Identification
                          Register</entry>
                      </row>

                      <row>
                        <entry>Bit 7</entry>
                        <entry spanname="2to6">FIFOs enabled.  On the
                          8250/16450 UART, this bit is zero.</entry>
                      </row>
                      
                      <row>
                        <entry>Bit 6</entry>
                        <entry spanname="2to6">FIFOs enabled.  On the
                          8250/16450 UART, this bit is zero.</entry>
                      </row>
                      
                      <row>
                        <entry>Bit 5</entry>
                        <entry spanname="2to6">Reserved, always 0.</entry>
                      </row>
                      
                      <row>
                        <entry>Bit 4</entry>
                        <entry spanname="2to6">Reserved, always 0.</entry>
                      </row>
                      
                      <row>
                        <entry>Bit 3</entry>
                        <entry spanname="2to6">Interrupt ID Bit #2.  On the
                          8250/16450 UART, this bit is zero.</entry>
                      </row>
                      
                      <row>
                        <entry>Bit 2</entry>
                        <entry spanname="2to6">Interrupt ID Bit #1</entry>
                      </row>
                      
                      <row>
                        <entry>Bit 1</entry>
                        <entry spanname="2to6">Interrupt ID Bit #0.These three
                          bits combine to report the category of event that
                          caused the interrupt that is in progress.  These
                          categories have priorities, so if  multiple
                          categories of events occur at the same time, the
                          UART will report the more important events first and
                          the host must resolve the events in the order they
                          are reported.  All events that caused the current
                          interrupt must be resolved before any new interrupts
                          will be generated.  (This is a limitation of the PC
                          architecture.)</entry>
                      </row>
                      
                      <row>
                        <entry colname="col2">2</entry>
                        <entry colname="col3">1</entry>
                        <entry colname="col4">0</entry>
                        <entry colname="col5">Priority</entry>
                        <entry colname="col6">Description</entry>
                      </row>
                      
                      <row>
                        <entry colname="col2">0</entry>
                        <entry colname="col3">1</entry>
                        <entry colname="col4">1</entry>
                        <entry colname="col5">First</entry>
                        <entry colname="col6">Received Error (OE, PE, BI, or
                          FE)</entry>
                      </row>
                      
                      <row>
                        <entry colname="col2">0</entry>
                        <entry colname="col3">1</entry>
                        <entry colname="col4">0</entry>
                        <entry colname="col5">Second</entry>
                        <entry colname="col6">Received Data Available</entry>
                      </row>
                      
                      <row>
                        <entry colname="col2">1</entry>
                        <entry colname="col3">1</entry>
                        <entry colname="col4">0</entry>
                        <entry colname="col5">Second</entry>
                        <entry colname="col6">Trigger level identification
                          (Stale data in receive buffer)</entry>
                      </row>
                      
                      <row>
                        <entry colname="col2">0</entry>
                        <entry colname="col3">0</entry>
                        <entry colname="col4">1</entry>
                        <entry colname="col5">Third</entry>
                        <entry colname="col6">Transmitter has room for more
                          words (THRE)</entry>
                      </row>
                      
                      <row>
                        <entry colname="col2">0</entry>
                        <entry colname="col3">0</entry>
                        <entry colname="col4">0</entry>
                        <entry colname="col5">Fourth</entry>
                        <entry colname="col6">Modem Status Change (-CTS, -DSR,
                          -RI, or -DCD)</entry>
                      </row>
                      
                      <row>
                        <entry>Bit 0</entry>
                        <entry spanname="2to6">Interrupt Pending Bit.  If this
                          bit is set to "0", then at least one interrupt is
                          pending.</entry>
                      </row>
                    </tbody>
                  </entrytbl>
                </row>
                
                <row>
                  <entry>+0x03</entry>
                  <entry>write/read</entry>
                  <entrytbl cols="5">
                    <colspec colnum="1" colname="col1">
                    <colspec colnum="2" colname="col2">
                    <colspec colnum="3" colname="col3">
                    <colspec colnum="4" colname="col4">
                    <colspec colnum="5" colname="col5">
                    <spanspec namest="col1" nameend="col5" spanname="1to5">
                    <spanspec namest="col2" nameend="col5" spanname="2to5">
                    <spanspec namest="col4" nameend="col5" spanname="4to5">

                    <tbody>
                      <row>
                        <entry spanname="1to5">Line Control Register
                          (LCR)</entry>
                      </row>
                      
                      <row>
                        <entry>Bit 7</entry>
                        <entry spanname="2to5">Divisor Latch Access Bit
                          (DLAB).  When set, access to the data
                          transmit/receive register (THR/RBR) and the
                          Interrupt Enable Register (IER) is disabled.  Any
                          access to these ports is now redirected to the
                          Divisor Latch Registers.  Setting this bit, loading
                          the Divisor Registers, and clearing DLAB should be
                          done with interrupts disabled.</entry>
                      </row>
                      
                      <row>
                        <entry>Bit 6</entry>
                        <entry spanname="2to5">Set Break.  When set to "1",
                          the transmitter begins to transmit continuous
                          Spacing until this bit is set to "0".  This
                          overrides any bits of characters that are being
                          transmitted.</entry>
                      </row>
                      
                      <row>
                        <entry>Bit 5</entry>
                        <entry spanname="2to5">Stick Parity.  When parity is
                          enabled, setting this bit causes parity to always be
                          "1" or "0", based on the value of Bit 4.</entry>
                      </row>
                      
                      <row>
                        <entry>Bit 4</entry>
                        <entry spanname="2to5">Even Parity Select (EPS). When
                          parity is enabled and Bit 5 is "0", setting this bit
                          causes even parity to be transmitted and expected.
                          Otherwise, odd parity is used.</entry>
                      </row>
                      
                      <row>
                        <entry>Bit 3</entry>
                        <entry spanname="2to5">Parity Enable (PEN).  When set
                          to "1", a parity bit is inserted between the last
                          bit of the data and the Stop Bit.  The UART will
                          also expect parity to be present in the received
                          data.</entry>
                      </row>
                      
                      <row>
                        <entry>Bit 2</entry>
                        <entry spanname="2to5">Number of Stop Bits (STB). If
                          set to "1" and using 5-bit data words, 1.5 Stop Bits
                          are transmitted and expected in each data word.  For
                          6, 7 and 8-bit data words, 2 Stop Bits are
                          transmitted and expected.  When this bit is set to
                          "0", one Stop Bit is used on each data word.</entry>
                      </row>
                      
                      <row>
                        <entry>Bit 1</entry>
                        <entry spanname="2to5">Word Length Select Bit #1
                          (WLSB1)</entry>
                      </row>
                      
                      <row>
                        <entry>Bit 0</entry>
                        <entry spanname="2to5">Word Length Select Bit #0
                          (WLSB0)</entry>
                      </row>
                      
                      <row>
                        <entry colname="col2" spanname="2to5">Together these
                          bits specify the number of bits in each data
                          word.</entry>
                      </row>
                      
                      <row>
                        <entry colname="col2">1</entry>
                        <entry colname="col3">0</entry>
                        <entry colname="col4" spanname="4to5">Word
                          Length</entry>
                      </row>
                      
                      <row>
                        <entry colname="col2">0</entry>
                        <entry colname="col3">0</entry>
                        <entry colname="col4" spanname="4to5">5 Data
                          Bits</entry>
                      </row>
                      
                      <row>
                        <entry colname="col2">0</entry>
                        <entry colname="col3">1</entry>
                        <entry colname="col4" spanname="4to5">6 Data
                          Bits</entry>
                      </row>
                      
                      <row>
                        <entry colname="col2">1</entry>
                        <entry colname="col3">0</entry>
                        <entry colname="col4" spanname="4to5">7 Data
                          Bits</entry>
                      </row>
                      
                      <row>
                        <entry colname="col2">1</entry>
                        <entry colname="col3">1</entry>
                        <entry colname="col4" spanname="4to5">8 Data
                          Bits</entry>
                      </row>
                    </tbody>
                  </entrytbl>
                </row>
                
                <row>
                  <entry>+0x04</entry>
                  <entry>write/read</entry>
                  <entrytbl cols="2">
                    <colspec colnum="1" colname="col1">
                    <colspec colnum="2" colname="col2">
                    <spanspec namest="col1" nameend="col2" spanname="1to2">

                    <tbody>
                      <row>
                        <entry spanname="1to2">Modem Control Register
                          (MCR)</entry>
                      </row>
                      
                      <row>
                        <entry>Bit 7</entry>
                        <entry>Reserved, always 0.</entry>
                      </row>
                      
                      <row>
                        <entry>Bit 6</entry>
                        <entry>Reserved, always 0.</entry>
                      </row>
                      
                      <row>
                        <entry>Bit 5</entry>
                        <entry>Reserved, always 0.</entry>
                          </row>
                      
                      <row>
                        <entry>Bit 4</entry>
                        <entry>Loop-Back Enable.  When set to "1", the UART
                          transmitter and receiver are internally connected
                          together to allow diagnostic operations.  In
                          addition, the UART modem control outputs are
                          connected to the UART modem  control inputs.  CTS is
                          connected to RTS, DTR is connected to DSR, OUT1 is
                          connected to RI, and OUT 2 is connected to
                          DCD.</entry>
                      </row>
                      
                      <row>
                        <entry>Bit 3</entry>
                        <entry>OUT 2.  An auxiliary output that the host
                          processor may set high or low.  In the IBM PC serial
                          adapter (and most clones), OUT 2 is used to
                          tri-state (disable) the interrupt signal from the
                          8250/16450/16550 UART.</entry>
                      </row>
                      
                      <row>
                        <entry>Bit 2</entry>
                        <entry>OUT 1.  An auxiliary output that the host
                          processor may set high or low.  This output is not
                          used on the IBM PC serial adapter.</entry>
                      </row>
                      
                      <row>
                        <entry>Bit 1</entry>
                        <entry>Request to Send (RTS).  When set to "1", the
                          output of the UART -RTS line is Low
                          (Active).</entry>
                      </row>
                      
                      <row>
                        <entry>Bit 0</entry>
                        <entry>Data Terminal Ready (DTR).  When set to "1",
                          the output of the UART -DTR line is Low
                          (Active).</entry>
                      </row>
                    </tbody>
                  </entrytbl>
                </row>
                
                <row>
                  <entry>+0x05</entry>
                  <entry>write/read</entry>
                  <entrytbl cols="2">
                    <colspec colnum="1" colname="col1">
                    <colspec colnum="2" colname="col2">
                    <spanspec namest="col1" nameend="col2" spanname="1to2">

                    <tbody>
                      <row>
                        <entry spanname="1to2">Line Status Register
                          (LSR)</entry>
                      </row>
                      
                      <row>
                        <entry>Bit 7</entry>
                        <entry>Error in Receiver FIFO.  On the 8250/16450
                          UART, this bit is zero.  This bit is set to "1" when
                          any of the bytes in the FIFO have one or more of the
                          following error conditions: PE, FE, or BI.</entry>
                      </row>
                      
                      <row>
                        <entry>Bit 6</entry>
                        <entry>Transmitter Empty (TEMT).  When set to "1",
                          there are no words  remaining in the transmit FIFO
                          or the transmit shift register.  The transmitter is
                          completely idle.</entry>
                      </row>
                      
                      <row>
                        <entry>Bit 5</entry>
                        <entry>Transmitter Holding Register Empty (THRE).
                          When set to "1", the FIFO (or holding register) now
                          has room for at least one additional word to
                          transmit.  The transmitter may still be transmitting
                          when this bit is set to "1".</entry>
                      </row>
                      
                      <row>
                        <entry>Bit 4</entry>
                        <entry>Break Interrupt (BI).  The receiver has
                          detected a Break signal.</entry>
                      </row>
                      
                      <row>
                        <entry>Bit 3</entry>
                        <entry>Framing Error (FE).  A Start Bit was detected
                          but the Stop Bit did not appear at the expected
                          time.  The received word is probably
                          garbled.</entry>
                      </row>
                      
                      <row>
                        <entry>Bit 2</entry>
                        <entry>Parity Error (PE).  The parity bit was
                          incorrect for the word received.</entry>
                      </row>
                      
                      <row>
                        <entry>Bit 1</entry>
                        <entry>Overrun Error (OE).  A new word was received
                          and there was no room in the receive buffer.  The
                          newly-arrived word in the shift register is
                          discarded.  On 8250/16450 UARTs, the word in the
                          holding register is discarded and the newly- arrived
                          word is put in the holding register.</entry>
                      </row>
                      
                      <row>
                        <entry>Bit 0</entry>
                        <entry>Data Ready (DR) One or more words are in the
                          receive FIFO that the host may read.  A word must be
                          completely received and moved from the shift
                          register into the FIFO (or holding register for
                          8250/16450 designs) before this bit is set.</entry>
                      </row>
                    </tbody>
                  </entrytbl>
                </row>
                
                <row>
                  <entry>+0x06</entry>
                  <entry>write/read</entry>
                  <entrytbl cols="2">
                    <colspec colnum="1" colname="col1">
                    <colspec colnum="2" colname="col2">
                    <spanspec namest="col1" nameend="col2" spanname="1to2">

                    <tbody>
                      <row>
                        <entry spanname="1to2">Modem Status Register
                          (MSR)</entry>
                      </row>
                      
                      <row>
                        <entry>Bit 7</entry>
                        <entry>Data Carrier Detect (DCD).  Reflects the state
                          of the DCD line on the UART.</entry>
                      </row>
                      
                      <row>
                        <entry>Bit 6</entry>
                        <entry>Ring Indicator (RI).  Reflects the state of the
                          RI line on the UART.</entry>
                      </row>
                      
                      <row>
                        <entry>Bit 5</entry>
                        <entry>Data Set Ready (DSR).  Reflects the state of
                          the DSR line on the UART.</entry>
                      </row>
                      
                      <row>
                        <entry>Bit 4</entry>
                        <entry>Clear To Send (CTS).  Reflects the state of the
                          CTS line on the UART.</entry>
                      </row>
                      
                      <row>
                        <entry>Bit 3</entry>
                        <entry>Delta Data Carrier Detect (DDCD).  Set to "1"
                          if the -DCD line has changed state one more more
                          times since the last time the MSR was read by the
                          host.</entry>
                      </row>
                      
                      <row>
                        <entry>Bit 2</entry>
                        <entry>Trailing Edge Ring Indicator (TERI).  Set to
                          "1" if the -RI line has had a low to high transition
                          since the last time the MSR was read by the
                          host.</entry>
                      </row>
                      
                      <row>
                        <entry>Bit 1</entry>
                        <entry>Delta Data Set Ready (DDSR).  Set to "1" if the
                          -DSR line has changed state one more more times
                          since the last time the MSR was read by the
                          host.</entry>
                      </row>
                      
                      <row>
                        <entry>Bit 0</entry>
                        <entry>Delta Clear To Send (DCTS).  Set to "1" if the
                          -CTS line has changed state one more more times
                          since the last time the MSR was read by the
                          host.</entry>
                      </row>
                    </tbody>
                  </entrytbl>
                </row>
                
                <row>
                  <entry>+0x07</entry>
                  <entry>write/read</entry>
                  <entry>Scratch Register (SCR).  This register performs no
                    function in the UART.  Any value can be written by the
                    host to this location and read by the host later
                    on.</entry>
                </row>
              </tbody>
            </tgroup>
          </informaltable>
        </sect4>
        
        <sect4>
          <title>Beyond the 16550A UART</title>
          
          <para>Although National Semiconductor has not offered any components
            compatible with the 16550 that provide additional features,
            various other vendors have.  Some of these components are
            described below.  It should be understood that to effectively
            utilize these improvements, drivers may have to be provided by the
            chip vendor since most of the popular operating systems do not
            support features beyond those provided by the 16550.</para>
              
          <variablelist>
            <varlistentry>
              <term>ST16650</term>

              <listitem>
                <para>By default this part is similar to the NS16550A, but an
                  extended 32-byte send and receive buffer can be optionally
                  enabled.  Made by Startech.</para>
              </listitem>
            </varlistentry>
            
            <varlistentry>
              <term>TIL16660</term>
              
              <listitem>
                <para>By default this part behaves similar to the NS16550A,
                  but an extended 64-byte send and receive buffer can be
                  optionally enabled.  Made by Texas Instruments.</para>
              </listitem>
            </varlistentry>
            
            <varlistentry>
              <term>Hayes ESP</term>
              
              <listitem>
                <para>This proprietary plug-in card contains a 2048-byte send
                  and receive buffer, and supports data rates to
                  230.4Kbit/sec.  Made by Hayes.</para>
              </listitem>
            </varlistentry>
          </variablelist>
          
          <para>In addition to these &ldquo;dumb&rdquo; UARTs, many vendors
            produce intelligent serial communication boards.  This type of
            design usually provides a microprocessor that interfaces with
            several UARTs, processes and buffers the data, and then alerts the
            main PC processor when necessary.  Because the UARTs are not
            directly accessed by the PC processor in this type of
            communication system, it is not necessary for the vendor to use
            UARTs that are compatible with the 8250, 16450, or the 16550 UART.
            This leaves the designer free to components that may have better
            performance characteristics.</para>
        </sect4>
-->

          </sect3>

          <sect3 id="sio">
             <title>*** Configurer le pilote de p&eacute;riph&eacute;rique
               <devicename>sio</devicename></title>
             &sgml.todo;

<!--
        <para>The <devicename>sio</devicename> driver provides support for
          NS8250-, NS16450-, NS16550 and NS16550A-based EIA RS-232C (CCITT
          V.24) communications interfaces.  Several multiport cards are
          supported as well.  See the &man.sio.4;
          manual page for detailed technical documentation.</para>
            
        <sect4>
          <title>Digi International (DigiBoard) PC/8</title>
          
          <para><emphasis>Contributed by &a.awebster;.  26 August
              1995.</emphasis></para>
              
          <para>Here is a config snippet from a machine with a Digi
            International PC/8 with 16550.  It has 8 modems connected to these
            8 lines, and they work just great.  Do not forget to add
            <literal>options COM_MULTIPORT</literal> or it will not work very
            well!</para>
              
          <programlisting>
device          sio4    at isa? port 0x100 tty flags 0xb05
device          sio5    at isa? port 0x108 tty flags 0xb05
device          sio6    at isa? port 0x110 tty flags 0xb05
device          sio7    at isa? port 0x118 tty flags 0xb05
device          sio8    at isa? port 0x120 tty flags 0xb05
device          sio9    at isa? port 0x128 tty flags 0xb05
device          sio10   at isa? port 0x130 tty flags 0xb05
device          sio11   at isa? port 0x138 tty flags 0xb05 irq 9 vector siointr</programlisting>
              
          <para>The trick in setting this up is that the MSB of the flags
            represent the last SIO port, in this case 11 so flags are
            0xb05.</para>
        </sect4>
        
        <sect4>
          <title>Boca 16</title>
          
          <para><emphasis>Contributed by &a.whiteside;.  26 August
              1995.</emphasis></para>
          
          <para>The procedures to make a Boca 16 port board with FreeBSD are
            pretty straightforward, but you will need a couple things to make
            it work:</para>
          
          <orderedlist>
            <listitem>
              <para>You either need the kernel sources installed so you can
                recompile the necessary options or you will need someone else
                to compile it for you.  The 2.0.5 default kernel does
                <emphasis>not</emphasis> come with multiport support enabled
                and you will need to add a device entry for each port
                anyways.</para>
            </listitem>
            
            <listitem>
              <para>Two, you will need to know the interrupt and IO setting
                for your Boca Board so you can set these options properly in
                the kernel.</para>
            </listitem>
          </orderedlist>
          
          <para>One important note &mdash; the actual UART chips for the Boca
            16 are in the connector box, not on the internal board itself.  So
            if you have it unplugged, probes of those ports will fail.  I have
            never tested booting with the box unplugged and plugging it back
            in, and I suggest you do not either.</para>
              
          <para>If you do not already have a custom kernel configuration file
            set up, refer to <link linkend="kernelconfig">Kernel
              Configuration</link> for general procedures.  The following are
            the specifics for the Boca 16 board and assume you are using the
            kernel name MYKERNEL and editing with vi.</para>
              
          <procedure>
            <step>
              <para>Add the line 
                
                <programlisting>
options COM_MULTIPORT</programlisting>

                to the config file.</para>
            </step>
            
            <step>
              <para>Where the current <literal>device
                  sio<replaceable>n</replaceable></literal> lines are, you
                will need to add 16 more devices.  Only the last device
                includes the interrupt vector for the board.  (See the
                  &man.sio.4; manual page for detail as
                to why.)  The following example is for a Boca Board with an
                interrupt of 3, and a base IO address 100h.  The IO address
                for Each port is +8 hexadecimal from the previous port, thus
                the 100h, 108h, 110h...  addresses.</para>

              <programlisting>
device sio1 at isa? port 0x100 tty flags 0x1005
device sio2 at isa? port 0x108 tty flags 0x1005
device sio3 at isa? port 0x110 tty flags 0x1005
device sio4 at isa? port 0x118 tty flags 0x1005
&hellip;
device sio15 at isa? port 0x170 tty flags 0x1005
device sio16 at isa? port 0x178 tty flags 0x1005 irq 3 vector siointr</programlisting>

              <para>The flags entry <emphasis>must</emphasis> be changed from
                this example unless you are using the exact same sio
                assignments. Flags are set according to
                0x<replaceable>M</replaceable><replaceable>YY</replaceable>
                where <replaceable>M</replaceable> indicates the minor number
                of the master port (the last port on a Boca 16) and
                <replaceable>YY</replaceable> indicates if FIFO is enabled or
                disabled(enabled), IRQ sharing is used(yes) and if there is an
                AST/4 compatible IRQ control register(no).  In this example,
                <programlisting> flags 0x1005</programlisting> indicates that
                the master port is sio16.  If I added another board and
                assigned sio17 through sio28, the flags for all 16 ports on
                <emphasis>that</emphasis> board would be 0x1C05, where 1C
                indicates the minor number of the master port.  Do not change
                the 05 setting.</para>
                </step>
                  
            <step>
              <para>Save and complete the kernel configuration, recompile,
                install and reboot.  Presuming you have successfully installed
                the recompiled kernel and have it set to the correct address
                and IRQ, your boot message should indicate the successful
                probe of the Boca ports as follows: (obviously the sio
                numbers, IO and IRQ could be different)</para>
                                        
              <screen>sio1 at 0x100-0x107 flags 0x1005 on isa
sio1: type 16550A (multiport)
sio2 at 0x108-0x10f flags 0x1005 on isa
sio2: type 16550A (multiport)
sio3 at 0x110-0x117 flags 0x1005 on isa
sio3: type 16550A (multiport)
sio4 at 0x118-0x11f flags 0x1005 on isa
sio4: type 16550A (multiport)
sio5 at 0x120-0x127 flags 0x1005 on isa
sio5: type 16550A (multiport)
sio6 at 0x128-0x12f flags 0x1005 on isa
sio6: type 16550A (multiport)
sio7 at 0x130-0x137 flags 0x1005 on isa
sio7: type 16550A (multiport)
sio8 at 0x138-0x13f flags 0x1005 on isa
sio8: type 16550A (multiport)
sio9 at 0x140-0x147 flags 0x1005 on isa
sio9: type 16550A (multiport)
sio10 at 0x148-0x14f flags 0x1005 on isa
sio10: type 16550A (multiport)
sio11 at 0x150-0x157 flags 0x1005 on isa
sio11: type 16550A (multiport)
sio12 at 0x158-0x15f flags 0x1005 on isa
sio12: type 16550A (multiport)
sio13 at 0x160-0x167 flags 0x1005 on isa
sio13: type 16550A (multiport)
sio14 at 0x168-0x16f flags 0x1005 on isa
sio14: type 16550A (multiport)
sio15 at 0x170-0x177 flags 0x1005 on isa
sio15: type 16550A (multiport)
sio16 at 0x178-0x17f irq 3 flags 0x1005 on isa
sio16: type 16550A (multiport master)</screen>

              <para>If the messages go by too fast to see,

                <screen>&prompt.root; <userinput>dmesg | more</userinput></screen>
                will show you the boot messages.</para>
            </step>
            
            <step>
              <para>Next, appropriate entries in <filename>/dev</filename> for
                the devices must be made using the
                <filename>/dev/MAKEDEV</filename> script. After becoming
                root:</para>
              
              <screen>&prompt.root; <userinput>cd /dev</userinput>
&prompt.root; <userinput>./MAKEDEV tty1</userinput>
&prompt.root; <userinput>./MAKEDEV cua1</userinput>
<emphasis>(everything in between)</emphasis>
&prompt.root; <userinput>./MAKEDEV ttyg</userinput>
&prompt.root; <userinput>./MAKEDEV cuag</userinput></screen>
                      
              <para>If you do not want or need callout devices for some
                reason, you can dispense with making the
                <filename>cua*</filename> devices.</para>
            </step>
            
            <step>
              <para>If you want a quick and sloppy way to make sure the
                devices are working, you can simply plug a modem into each
                port and (as root)
                    
                <screen>&prompt.root; <userinput>echo at &gt; ttyd*</userinput></screen>
                for each device you have made.  You
                <emphasis>should</emphasis> see the RX lights flash for each
                working port.</para>
            </step>
          </procedure>
        </sect4>
-->

          </sect3>
          <sect3 id="cy">
             <title>*** Configurer le pilote de p&eacute;riph&eacute;rique
               <devicename>cy</devicename></title>
             &sgml.todo;
            
<!--
        <para><emphasis>Contributed by &a.alex;.  6 June
            1996.</emphasis></para>
            
        <para>The Cyclades multiport cards are based on the
          <devicename>cy</devicename> driver instead of the usual
          <devicename>sio</devicename> driver used by other multiport cards.
          Configuration is a simple matter of:</para>
        
        <procedure>
          <step>
            <para>Add the <devicename>cy</devicename> device to your <link
                linkend="kernelconfig-config">kernel configuration</link>
              (note that your irq and iomem settings may differ).</para>
                    
            <programlisting>
device cy0 at isa? tty irq 10 iomem 0xd4000 iosiz 0x2000 vector cyintr</programlisting>
          </step>
                
          <step>
            <para><link linkend="kernelconfig-building">Rebuild and
                install</link>  the new kernel.</para>
          </step>
          
          <step>
            <para>Make the <link linkend="kernelconfig-nodes">device
                nodes</link> by typing (the following example assumes an
              8-port board):</para>
                  
            <screen>&prompt.root; <userinput>cd /dev</userinput>
&prompt.root; <userinput>for i in 0 1 2 3 4 5 6 7;do ./MAKEDEV cuac$i ttyc$i;done</userinput></screen>
          </step>
          
          <step>
            <para>If appropriate, add <link linkend="dialup">dialup</link>
              entries to <link linkend="dialup-ttys">/etc/ttys</link> by
              duplicating serial device (<literal>ttyd</literal>) entries and
              using <literal>ttyc</literal> in place of
              <literal>ttyd</literal>.  For example:</para>
            
            <programlisting>
ttyc0   "/usr/libexec/getty std.38400"  unknown on insecure
ttyc1   "/usr/libexec/getty std.38400"  unknown on insecure
ttyc2   "/usr/libexec/getty std.38400"  unknown on insecure
&hellip;
ttyc7   "/usr/libexec/getty std.38400"  unknown on insecure</programlisting>
          </step>
                
          <step>
            <para>Reboot with the new kernel.</para>
          </step>
        </procedure>

-->
          </sect3>
        </sect2>
        
        <sect2>
          <title>* Ports parall&egrave;les</title>
          <para></para>
        </sect2>
        
        <sect2>
          <title>* Modems</title>
          <para></para>
        </sect2>
        
        <sect2>
          <title>* Cartes r&eacute;seau</title>
          <para></para>
        </sect2>
        
        <sect2>
          <title>* Claviers</title>
          <para></para>
        </sect2>
        
        <sect2>
          <title>* Souris</title>
          <para></para>
        </sect2>
        
        <sect2>
          <title>* Autres</title>
          <para></para>
        </sect2>
      </sect1>
      
  <sect1 id="hw-storage">
    <title>** M&eacute;moires de masse</title>
    &trans.a.haby;
        
    <sect2 id="esdi">
      <title>Utiliser des disques durs ESDI</title>
      
      <para><emphasis>Copyright &copy; 1995, &a.wilko;. 24 Septembre 
        1995.</emphasis></para>
	  
      <para>ESDI est l'abr&eacute;viation de <foreignphrase>Enhanced Small 
        Device Interface</foreignphrase>&nbsp;-&nbsp;Interface 
        Am&eacute;lior&eacute;e pour les P&eacute;riph&eacute;riques 
        L&eacute;gers. Elle se base plus ou moins sur la bonne vieille interface
        ST506/412, initialement con&ccedil;ue par Seagate Technology, le 
        fabricant du premier disque Winchester 5.25" bon march&eacute;.</para>
	  
      <para>L'abr&eacute;viation pr&eacute;cise &agrave; juste titre 
        &rdquo;&eacute;tendue&ldquo;. Pour commencer, l'interface est plus 
        rapide, 10 ou 15 Mbits/seconde au lieu des 5 Mbits/seconde des disques 
        &agrave; interface ST412s. Il y a de plus de nouvelles commandes de plus
        haut niveau, qui font que l'interface ESDI est en quelque sorte plus 
        &ldquo;intelligente&rdquo; que les pilotes de 
        p&eacute;riph&eacute;riques du syst&egrave;me d'exploitation. Elle n'est
        cependant pas comparable aux interfaces SCSI. L'ESDI est un standard 
        ANSI.</para>
	  
      <para>La capacit&eacute; de disques est accrue parce qu'il y a plus de
        secteurs par piste. Il y a g&eacute;n&eacute;ralement 35 secteurs par
        pistes, mais j'ai vu des disques de grande capacit&eacute; avec 54
	secteurs par piste.</para>
      
      <para>Bien que l'IDE et le SCSI ait rendu l'ESDI largement 
        obsol&egrave;te, la possibilit&eacute; de se procurer gratuitement ou
        &agrave; peu de frais des disques d'occasion les rend 
        int&eacute;ressants pour les syst&egrave;mes &agrave; budget 
        r&eacute;duit (ou nul).</para>
      
      <sect3>
	<title>Concepts ESDI</title>
	
	<sect4>
	  <title>Connexions</title>
	  
	  <para>L'interface ESDI utilise deux c&acirc;bles par disque. Le 
            premier est une nappe &agrave; 54 broches qui v&eacute;hicule les
            signaux de commandes et d'&eacute;tat entre le contr&ocirc;leur et 
            le disque. Les disques sont cha&icirc;n&eacute;s en s&eacute;rie sur
            ce c&acirc;ble. C'est donc un bus auquel tous les disques sont 
            reli&eacute;s.</para>
	      
	  <para>Le second c&acirc;ble est une nappe &agrave; 20 broches qui
            v&eacute;hicule les donn&eacute;es de et vers le disque. Ce 
            c&acirc;blage est en &eacute;toile, chaque disque est donc
            directement reli&eacute; au contr&ocirc;leur.</para>
	      
	  <para>Autant que je sache, on ne peut mettre que deux disques par
            contr&ocirc;leur ESDI sur un PC. Cela pour des raisons de
            compatibilit&eacute; (?) avec le standard WD1003 qui n'utilise
            qu'un seul bit pour l'adresse des 
            p&eacute;riph&eacute;riques.</para>
	</sect4>
	
	<sect4>
	  <title>Adresses des p&eacute;riph&eacute;riques</title>
	  
	  <para>Sur chaque c&acirc;ble de commande, il peut y avoir au plus 7
            p&eacute;riph&eacute;riques et 1 contr&ocirc;leur. Pour que le
            contr&ocirc;leur puisse identifier l'adresse de chaque disque, il y
            a sur chaque p&eacute;riph&eacute;rique ESDI des cavaliers ou des
            interrupteurs pour d&eacute;finir l'adresse du 
            p&eacute;riph&eacute;rique.</para>
	      
	  <para>Sur les contr&ocirc;leurs de PC, le premier disque a  l'adresse
            0, et le second l'adresse 1. <emphasis>V&eacute;rifiez 
            toujours</emphasis> que l'adresse de chaque disque est 
            diff&eacute;rente&nbsp;! Sur les PC, o&ugrave; il y a au plus deux
            disques par contr&ocirc;leur, le premier disque est le disque 0 et
            le second le disque 1.</para>
	</sect4>
	
	<sect4>
	  <title>Terminaison</title>
	  
	  <para>Le c&acirc;ble s&eacute;rie de commande (rapellez-vous, celui 
            &agrave; 34 broches) doit &ecirc;tre termin&eacute; sur le dernier
            disque de la cha&icirc;ne. Il y a donc sur les disques ESDI une
            r&eacute;sistance de terminaison qui peut &ecirc;tre enlev&eacute;e
            ou d&eacute;sactiv&eacute;e par un cavalier si elle ne doit pas 
            servir.</para>
	  
	  <para>Il n'y a donc qu'un <emphasis>seul</emphasis> disque, celui 
            en fin du c&acirc;ble de commande, dont le terminateur doit
            &ecirc;tre install&eacute; ou activ&eacute;. Le contr&ocirc;leur
            termine automatiquement l'autre extr&eacute;mit&eacute; du 
            c&acirc;ble. Notez bien, s'il vous pla&icirc;t, que cela implique
            que le contr&ocirc;leur soit &agrave; une extr&eacute;mit&eacute; du
            c&acirc;ble, et <emphasis>non</emphasis> au milieu.</para>
	</sect4>
      </sect3>

      <sect3>
         <title>Utiliser les disques ESDI avec FreeBSD</title>

        <para>Pourquoi est-il si difficile d'arriver &agrave; utiliser des
          disques ESDI&nbsp;?</para>
        
        <para>On dit que ceux qui ont essay&eacute; d'utiliser des disques ESDI 
          avec FreeBSD ont d&eacute;velopp&eacute; un sentiment de profonde
          frustration. Divers facteurs oeuvrent contre vous et produisent des
          r&eacute;sultats difficiles &agrave; comprendre si vous n'y avez 
          jamais &eacute;t&eacute; confront&eacute;.</para>
            
        <para>D'o&ugrave; la l&eacute;gende populaire qui veut que l'ESDI et 
          FreeBSD soient d&eacute;finitivement incompatibles. Ce qui suit tente
          de recenser les difficult&eacute;s et leurs solutions.</para>
            
        <sect4>
          <title>Les diff&eacute;rences de vitesse de l'ESDI</title>
          
          <para>Comme on y a d&eacute;j&agrave; fait allusion, il y a deux 
            versions &agrave; vitesse diff&eacute;rente  de l'ESDI. Les disques
            et les contr&ocirc;leurs plus anciens transf&egrave;rent les
            donn&eacute;es &agrave; 10 Mbits/seconde. Les plus r&eacute;cents le
            font &agrave; 15 Mbits/seconde.</para>
              
          <para>Il est facile d'imaginer qu'utiliser des disques &agrave; 
            15 Mbits/seconde pose des probl&egrave;mes avec des
            contr&ocirc;leurs &agrave; 10 Mbits/seconde. Consultez toujours,
            comme d'habitude, la documentation de votre contr&ocirc;leur
            <emphasis>et</emphasis> celle de votre disque pour v&eacute;rifier
            qu'ils sont compatibles.</para>
        </sect4>
        
        <sect4>
          <title>Restez en piste</title>
          
          <para>Les disques ESDI standards ont de 34 &agrave; 36 secteurs par
            piste. La plupart des (anciens) contr&ocirc;leurs n'acceptent pas
            plus de secteurs que cela. Les disques plus r&eacute;cents, de plus
            grande capacit&eacute;, ont un plus grand nombre de secteurs par
            piste. Je poss&egrave;de par exemple un disque de 670 Mo qui a 54
            secteurs par piste.</para>
              
          <para>Dans mon cas, le contr&ocirc;leur ne peut pas g&eacute;rer
            autant de secteurs. Cela fonctionne en utilisant que 35 secteurs par
            piste. D'o&ugrave; un perte important d'espace disque.</para>
              
          <para>Consultez encore une fois les documentations de votre 
            mat&eacute;riel pour plus d'informations. Ne pas respecter les
            sp&eacute;cifications, comme dans mon cas, marchera ou ne marchera
            pas. Essayez ou procurez-vous un contr&ocirc;leur qui r&egrave;gle 
            le probl&egrave;me.</para>
        </sect4>
        
        <sect4>
          <title>Secteurs mat&eacute;riels ou logiciels</title>
          
          <para>La plupart des disques ESDI permettent de choisir avec un 
            cavalier entre des secteurs mat&eacute;riels ou logiciels. Si les 
            secteurs sont mat&eacute;riels, le disque &eacute;mettera une 
            impulsion de d&eacute;but de secteur &agrave; chaque nouveau 
            secteur. Le contr&ocirc;leur utilisera cette impulsion pour savoir
            quand commencer &agrave; lire ou &agrave; &eacute;crire.</para>
          
          <para>Il est possible de choisir la taille des secteurs 
            mat&eacute;riels (habituellement 256, 512 ou 1024 octets par
            secteur format&eacute;). FreeBSD utilise des secteurs de 512
            octets. Le nombre de secteurs par piste varie aussi, bien qu'on
            utilise toujours le m&ecirc;me nombre d'octets par secteur
            format&eacute;. Le nombre d'octets <emphasis>non 
            format&eacute;s</emphasis> par secteur varie, selon que votre
            contr&ocirc;leur a besoin de plus ou moins d'octets 
            suppl&eacute;mentaires pour fonctionner correctement. Avec plus de
            secteurs par piste, vous aurez bien s&ucirc;r plus d'espace
            disponible, mais vous pouvez avoir des probl&egrave;mes si votre
            contr&ocirc;leur a besoin de plus d'octets que le disque ne peut
            lui en laisser &agrave; disposition.</para>
              
          <para>Avec des secteurs logiciels, le contr&ocirc;leur 
            d&eacute;termine lui-m&ecirc;me quand commencer et cesser de lire ou
            &eacute;crire. Pour les disques ESDI, les secteurs sont 
            mat&eacute;riels par d&eacute;faut (au moins pour tous ceux que je
            connais). Je n'ai jamais eu besoin d'essayer d'utiliser des secteurs
            logiciels.</para>
              
          <para>Exp&eacute;rimentez avec les secteurs avant d'installer FreeBSD,
            parce que vous devrez refaire un formatage de bas niveau &agrave;
            chaque fois.</para>
        </sect4>
        
        <sect4>
          <title>Formatage de bas niveau</title>
          
          <para>Il faut faire un formatage de bas niveau des disques ESDI avant
            de pouvoir les utiliser. Il faut les reformater &agrave; chaque
            fois que vous modifier la position des cavaliers qui 
            d&eacute;terminent le nombre de secteurs par piste ou l'orientation
            (horizontale, verticale) du disque. R&eacute;fl&eacute;chissez donc
            d'abord, puis formatez. Ne sous-estimez pas le temps 
            n&eacute;cessaire&nbsp;; pour de gros disques, cela peut prendre des
            heures.</para>
              
          <para>Evitez les utilitaires de formatage de bas niveau qui marquent
            une piste inutilisable d&egrave;s qu'ils trouvent un secteur 
            endommag&eacute;. Non seulement cela gaspille de l'espace disque, 
            mais cela vous posera peut-&ecirc;tre aussi des probl&egrave;mes 
            avec <literal>bad144</literal> (voyez plus bas la section sur le 
            sujet).</para>
        </sect4>
        
        <sect4>
          <title>Correspondances</title>
          
          <para>Les correspondances, bien que ce ne soit pas un probl&egrave;me
            exclusivement r&eacute;serv&eacute; &agrave; l'ESDI, peuvent vous
            poser de vraies difficult&eacute;s. Il y a diff&eacute;rentes sortes
            de correspondances. Elles ont en commun d'essayer de contourner les
            limites impos&eacute;es &agrave; la g&eacute;om&eacute;trie des 
            disques par l'architecture d'origine de l'IBM PC/AT (merci 
            IBM&nbsp;!).</para>

          <para>Il y a tout d'abord la limite bien connue du 1024&egrave;me
            cylindre pour le d&eacute;marrage. Pour qu'un syst&egrave;me (quel
            qu'il soit) d&eacute;marre, le code de d&eacute;marrage doit se 
            trouver quelque part sur les 1024 premiers cylindres. Il n'y a que
            10 bits disponibles pour coder le num&eacute;ro de cylindre. Le 
            nombre de secteurs est limit&eacute; &agrave; 64 (0-63). Quand vous
            y ajoutez la limite de 16 t&ecirc;tes (aussi li&eacute;e &agrave;
            l'architecture), cela vous donne des disques de taille relativement
            faible.</para>
         
          <para>Pour contourner ce probl&egrave;me, les fabricants de 
            contr&ocirc;leurs ESDI pour PC ont ajout&eacute; une extension au
            BIOS en PROM. Cette extension g&egrave;re les entr&eacute;es/sorties
            disque au d&eacute;marrage. (et pour certains syst&egrave;mes
            d'exploitation, toutes les entr&eacute;es/sorties) en utilisant des
            correspondances. Par exemple, un gros disque pourra &ecirc;tre
            d&eacute;crit au syst&egrave;me comme ayant 32 t&ecirc;tes et 64
            secteurs par piste. De la sorte, le nombre de cylindres sera 
            inf&eacute;rieur &agrave; 1024, ce qui pourra &ecirc;tre 
            exploit&eacute; sans probl&egrave;me. Il faut noter que FreeBSD
            n'utilise le BIOS qu'apr&egrave;s que le noyau ait pris le
            contr&ocirc;le. Nous en dirons plus &agrave; ce sujet plus 
            loin.</para>

          <para>Il faut aussi &eacute;tablir des correspondances avec la plupart
            des BIOS anciens qui ne peuvent g&eacute;rer que des disques avec
            17 secteurs par piste (le vieux standard ST412). Les BIOS plus
            r&eacute;cents premettent de d&eacute;finir le type de disque (c'est
            habituellement le type de disque 47).</para>

          <note>
            <para>Quoique vous fassiez des correspondances apr&egrave;s avoir lu
              ce document, n'oubliez pas que si vous avez plusieurs 
              syst&egrave;mes d'exploitation sur le m&ecirc;me disque, ils
              doivent tous utiliser les m&ecirc;mes correspondances.</para> 
          </note>
          
          <para>Pendant que nous en sommes aux correspondances, j'ai vu un 
            mod&egrave;le de contr&ocirc;leur (mais il y en a probablement
            d'autres) qui permet de diviser un disque en plusieurs partitions
            &agrave; l'aide d'une option du BIOS. J'avais choisi 1 disque = 1
            partition, parce que ce contr&ocirc;leur &eacute;crivait cette
            information sur le disque. A la mise sous tension, il la relit et
            transmet au syst&egrave;me les informations bas&eacute;es sur ce
            qu'il y a sur le disque.</para>
        </sect4>
            
        <sect4>
          <title>Secteurs en r&eacute;serve</title>
          
          <para>La plupart des contr&ocirc;leurs ESDI offrent la 
            possibilit&eacute; de r&eacute;affecter les secteurs 
            d&eacute;fectueux. Pendant ou apr&egrave;s le formatage de bas 
            niveau du disque, les secteurs d&eacute;fectueux sont marqu&eacute;s
            comme tels, et des secteurs de remplacement prennent (logiquement
            bien s&ucirc;r) leur place.</para>
              
          <para>Dans la plupart des cas, c'est fait en utilisant N-1 secteurs de
            chaque piste pour les donn&eacute;es et le secteur N lui-m&ecirc;me
            comme secteur de secours. N est le nombre de secteurs physiquement 
            disponibles sur la piste. L'id&eacute;e est que le syst&egrave;me
            d'exploitation voie un disque &rdquo;parfait&ldquo; sans secteur 
            d&eacute;fectueux. Ce n'est pas exploitable dans le cas de 
            FreeBSD.</para>
              
          <para>Le probl&egrave;me est que la correspondance entre les
            <emphasis>mauvaix</emphasis> et les <emphasis>bons</emphasis> 
            secteurs est effectu&eacute;e par le BIOS du contr&ocirc;leur ESDI.
            FreeBSD, qui est un vrai syst&egrave;me d'exploitation 32 bits,
            n'utilise pas le BIOS avant d'avoir d&eacute;marr&eacute;. Il 
            dispose &agrave; la place de pilotes de p&eacute;riph&eacute;rique
            qui dialoguent directement avec le mat&eacute;riel.</para>
              
          <para>Donc, n'utilisez pas les secteurs de r&eacute;serve,
            la r&eacute;affectation des secteurs d&eacute;fectueux, quel que 
            soit le nom que lui donne le fabricant de votre contr&ocirc;leur, si
            vous voulez vous servir du disque avec FreeBSD.</para>
        </sect4>
        
        <sect4>
          <title>Gestion des blocs d&eacute;fectueux</title>
          
          <para>La section pr&eacute;c&eacute;dente nous a laiss&eacute; sur un
            probl&egrave;me. La gestion des blocs d&eacute;fectueux par le
            contr&ocirc;leur n'est pas exploitable, et FreeBSD suppose 
            malgr&eacute; tout que les supports sont sans d&eacute;faut. Pour
            r&eacute;soudre ce probl&egrave;me, FreeBSD utilise l'outil
            <command>bad144</command>. <command>bad144</command> (dont le nom
            vient du standard de gestion des blocs d&eacute;fectueux de Digital
            Equipment) examine une 
            tranche&nbsp;-&nbsp;<foreignphrase>slice</foreignphrase>&nbsp;-&nbsp;FreeBSD
            pour d&eacute;tecter les blocs d&eacute;fectueux. Quand il les a
            trouv&eacute;s, il remplit une table avec les num&eacute;ros de ces
            blocs &agrave; la fin de la tranche.</para>
              
          <para>Quand le disque est en service, les num&eacute;ros des blocs 
            acc&eacute;d&eacute;s sont compar&eacute;s &agrave; ceux 
            stock&eacute;s dans la table lue sur le disque. Quand un bloc 
            demand&eacute; est dans la liste de <command>bad144</command>, on 
            utilise un bloc de remplacement (aussi en fin de tranche). De cette 
            fa&ccedil;on, c'est un support &rdquo;parfait&ldquo; qui est vu par
            les syst&egrave;mes de fichiers de FreeBSD.</para>
              
          <para>L'utilisation de <command>bad144</command> pr&eacute;sente un
            certain nombre d'inconv&eacute;nients. En premier lieu, la tranche
            ne peut comporter plus de 126 secteurs d&eacute;fectueux. Si votre
            disque pr&eacute;sente un gand nombre de secteurs d&eacute;fectueux,
            vous devrez peut-&ecirc;tre le diviser en plusieurs tranches dont
            chacune aura moins de 126 secteurs d&eacute;fectueux. Evitez les
            programmes de formatage de bas niveau qui marquent d&eacute;fectueux
            tous les secteurs d'une piste, d&egrave;s qu'il y a un 
            probl&egrave;me avec la piste. Vous comprennez bien que la limite
            de 126 secteurs est rapidement atteinte avec de tels 
            programmes.</para>
              
          <para>En second lieu, si la tranche contient le syst&egrave;me de 
            fichiers racine, il faut qu'elle soit &agrave; l'int&eacute;rieur 
            des 1024 premiers cylindres. La liste <command>bad144</command> est
            lue au d&eacute;marrage,  en utilisant le BIOS, et cela ne peut
            se faire que si la liste est avant le 
            1025&egrave;me cylindre.</para>

          <note>
            <para>Ce n'est pas seulement le <emphasis>syst&egrave;me</emphasis> 
              de fichiers racine qui doit se trouver dans les 1024 premiers 
              cylindres, mais toute la <emphasis>tranche</emphasis> qui le 
              contient.</para>
          </note>
        </sect4>
        
        <sect4>
          <title>Configuration du noyau</title>
          
          <para>Les disques ESDI sont g&eacute;r&eacute;s par le m&ecirc;me
            pilote <literal>wd</literal> que les disques IDE et ST412 MFM. Le
            pilote <literal>wd</literal> devrait fonctionner avec toutes les
            interfaces compatibles WD1003.</para>
              
          <para>La plupart des mat&eacute;riels ont des cavaliers pour 
            d&eacute;finir les plages d'adresses d'entr&eacute;es/sorties et les
            lignes IRQ. Cela vous permet de mettre deux contr&ocirc;leurs de 
            type <literal>wd</literal> sur un m&ecirc;me syst&egrave;me.</para>
          
          <para>si votre mat&eacute;riel permet des red&eacute;finition
            non-standard, vous pouvez les utiliser avec FreeBSD, d&egrave;s lors
            que vous donnez les informations correctes dans le fichier de
            configuration du noyau. Voici une extrait de fichier de 
            configuration du noyau ( au  fait, ils sont dans 
            <filename>/sys/i386/conf</filename>)&nbsp;:</para>
              
          <programlisting>
# Premier contr&ocirc;leur compatible WD 
controller      wdc0    at isa? port "IO_WD1" bio irq 14 vector wdintr
disk            wd0     at wdc0 drive 0
disk            wd1     at wdc0 drive 1
# Second contr&ocirc;leur compatible WD 
controller      wdc1    at isa? port "IO_WD2" bio irq 15 vector wdintr
disk            wd2     at wdc1 drive 0
disk            wd3     at wdc1 drive 1
          </programlisting>
        </sect4>
      </sect3>

      <sect3>
        <title>Sp&eacute;cificit&eacute;s de certains mat&eacute;riels ESDI</title>
        
        <sect4>
          <title>Contr&ocirc;leurs Adaptec 2320</title>
          
          <para>J'ai r&eacute;ussi &agrave; installer FreeBSD sur un disque ESDI
            avec un contr&ocirc;leur ACB-2320. Il n'y avait pas d'autre 
            syst&egrave;me d'exploitation sur le disque.</para>
              
          <para>Pour cela, j'ai effectu&eacute; un formatage de bas niveau du
            disque avec <command>NEFMT.EXE</command> 
            (t&eacute;l&eacute;chargeable par <command>ftp</command> depuis
            <hostid role="fqdn">www.adaptec.com</hostid>) et r&eacute;pondu
            <literal>NO</literal> &agrave; la question qui me demandait si le
            disque devait &ecirc;tre format&eacute; en laissant un secteur de
            secours par piste. Le BIOS de l'ACD-2320 &eacute;tait 
            d&eacute;sactiv&eacute; et j'ai utilis&eacute; l'option de 
            configuration
            libre&nbsp;-&nbsp;<literal>free configurable</literal>&nbsp;-&nbsp;du
            BIOS du syst&egrave;me pour permettre au BIOS de d&eacute;marrer
            avec.</para>
              
          <para>Avant de me servir de <command>NEFMT.EXE</command>, j'avais
            essay&eacute; de formater le disque avec l'utilitaire inclus dans
            le BIOS de l'ACB-2320. Cela s'est av&eacute;r&eacute; inutilisable,
            parce qu'il ne m'a pas propos&eacute; de d&eacute;sactiver la
            r&eacute;servation du secteur de secours. Avec ces derniers,
            l'installation de FreeBSD &eacute;choue &agrave; l'ex&eacute;cution
            de <command>bad144</command>.</para>
              
          <para>V&eacute;rifiez avec soin de quelle variante
            ACB-232<replaceable>xy</replaceable> vous disposez. Le
            <replaceable>x</replaceable> vaut  <literal>0</literal> ou
            <literal>2</literal>, selon que le contr&ocirc;leur ne dispose pas
            ou inclut un contr&ocirc;leur de lecteur de disquettes.</para>
              
          <para>Le <literal>y</literal> est plus int&eacute;ressant. C'est un
            blanc, un <literal>A-8</literal> ou un <literal>D</literal>. Le
            blanc indique un contr&ocirc;leur &agrave; 10 Mo/seconde. Le
            <literal>A-8</literal> indique un contr&ocirc;leur &agrave; 15 
            Mo/seconde capable de g&eacute;rer 52 secteurs par piste. Le
            <literal>D</literal> est un contr&ocirc;leur &agrave; 15 Mo/seconde
            qui peut aussi g&eacute;rer des disques avec plus de 36 secteurs
            par piste (52 aussii&nbsp;?).</para>
              
          <para>Toutes ces variantes peuvent g&eacute;rer l'entrelacement 1:1.
            Employez-le, FreeBSD est assez rapide pour s'en accommoder.</para>
        </sect4>
        
        <sect4>
          <title>Contr&ocirc;leurs Western Digital WD1007</title>
          
          <para>J'ai r&eacute;ussi &agrave; installer FreeBSD sur un disque ESDI
            avec un contr&ocirc;leur WD1007. Pour &ecirc;tre pr&eacute;cis,
            c'&eacute;tait un contr&ocirc;leur WD1007-WA2. Il y en a d'autres
            variantes.</para>
              
          <para>Pour qu'il fonctionne, j'ai d&eacute;sactiv&eacute; la 
            correspondance entre secteurs et le BIOS du WD1007. Ce qui signifie             que je n'ai pas pu me servir de l'utilitaire de formatage de bas 
            niveau de ce BIOS.  J'ai r&eacute;cup&eacute;r&eacute; 
            <command>WDFMT.EXE</command> sur 
            <hostid role="fqdn">www.wdc.com</hostid>. Il m'a permis de formater
            le disque sans probl&egrave;me.</para>
        </sect4>
        
        <sect4>
          <title>Contr&ocirc;leurs Ultrastor U14F</title>
          
          <para>Selon de nombreux retours sur le r&eacute;seau, les cartes
            Ultrastor ESDI fonctionnent avec FreeBSD. Je n'ai pas plus 
            d'informations sur leur configuration.</para>
        </sect4>
      </sect3>

      <sect3 id="esdi-further-reading">
        <title>Lectures compl&eacute;mentaires</title>

        <para>Si vous avez l'intention d'utiliser s&eacute;rieusement l'ESDI,
          vous devriez avoir la norme officielle &agrave; port&eacute;e de 
          main&nbsp;:</para>
            
        <para>Le document le plus r&eacute;cent du comit&eacute; ANSI X3T10
          est&nbsp;: &rdquo;Enhanced Small Device Interface (ESDI) 
          [X3.170-1990/X3.170a-1991] [X3T10/792D Rev 11]&ldquo;.</para>
                
        <para>Le forum Usenet 
          <ulink URL="news:comp.periphs">comp.periphs</ulink> est un bon endroit
          ou avoir plus d'informations.</para>
            
        <para>Le <foreignphrase>World Wide Web</foreignphrase> (WWW) est aussi
          une excellente source d'informations&nbsp;: Pour les contr&ocirc;leurs
          Adaptec, consultez 
          <ulink URL="http://www.adaptec.com/">http://www.adaptec.com/</ulink>.
          Pour les contr&ocirc;leurs Western Digital, voyez
          <ulink URL="http://www.wdc.com/">http://www.wdc.com/</ulink>.</para>

      </sect3>

      <sect3>
        <title>Remerciements &agrave; ...</title>
        
        <para>Andrew Gordon pour m'avoir envoy&eacute; un contr&ocirc;leur
          Adaptec 2320 et un disque ESDI pour faire des tests.</para>
      </sect3>
    </sect2>
        
    <sect2 id="scsi">
      <title>Qu'est-ce que le SCSI&nbsp;?</title>
      &trans.a.brive;

      <para><emphasis>Copyright &copy; 1995, &a.wilko;.  July 6,
          1996.</emphasis></para>
          
      <para>SCSI est un acronyme pour Small Computer Systems Interface. C'est
	un standard ANSI qui est devenu l'un des premiers bus d'E/S de
	l'industrie informatique. Les bases du standard SCSI proviennent
	de Shugart Associates (les m&ecirc;mes personnes qui ont donn&eacute; au monde les
	premiers mini-disques floppy) quand ils ont introduit le bus SASI
	(Shugart Associates Standard Interface).</para>

      <para>Un effort industriel a d&eacute;marr&eacute; quelque temps plus tard pour
        arriver &agrave; un standard plus strict, permettant &agrave; des p&eacute;riph&eacute;riques de
        diff&eacute;rents vendeurs de travailler ensemble. Cet effort fut reconnu
        par l'ANSI avec le standard SCSI-1. Ce standard (approx. 1985) devient
        rapidement obsol&egrave;te. Le standard courant est SCSI-2 (cf <link
        linkend="scsi-further-reading">Lecture compl&eacute;mentaire</link>),
	avec SCSI-3 en cours de conception.</para>

      <para>En plus d'un standard pour l'interconnexion physique, SCSI d&eacute;finit
	un standard logique (jeu de commandes) auxquels les disques doivent
	adh&eacute;rer. Ce standard "commun" est appell&eacute; le Common Command Set (CCS)
	et a &eacute;t&eacute; d&eacute;velopp&eacute; plus ou moins en parall&egrave;le avec le SCSI-1 ANSI.
	SCSI-2 int&egrave;gre le CCS (r&eacute;vis&eacute;) dans son standard. Les commandes
	d&eacute;pendent du type de p&eacute;riph&eacute;rique&nbsp;; il ne serait pas logique bien 
        s&ucirc;r de d&eacute;finir une commande "Ecriture" pour un scanner.</para>

      <para>Le bus SCSI est un bus parall&egrave;le, qui supporte plusieurs
	variantes. La plus ancienne et plus utilis&eacute;e est un bus de 8 bits
	de large, avec des signaux en collecteur ouvert (single-ended),
	transport&eacute;s sur 50 fils. (Si vous ne savez pas ce que veut dire
	"collecteur ouvert", ne vous en faites pas; c'est justement le sujet
	de ce document). Les architectures modernes utilisent aussi les bus
	de 16 bits avec des signaux diff&eacute;rentiels. Cela permet d'obtenir
	des taux de transferts de 20Mo/s, sur des c&acirc;bles atteignant 25 m&egrave;tres.
	SCSI-2 permet une largeur maximum du bus de 32 bits en utilisant un
	c&acirc;ble suppl&eacute;mentaire. Rapidement, l'Ultra SCSI (appel&eacute; aussi Fast-20)
	et l'Ultra2 (appel&eacute; aussi Fast-40) arrivent. Fast-20 correspond &agrave;
	20 millions de transferts par seconde (20Mo/s sur un bus de 8 bits)
	et Fast-40 correspond &agrave; 40 millions de transferts par seconde
	(40Mo/s sur un bus de 8 bits). La majorit&eacute; des disques vendus 
	aujourd'hui sont des Ultra SCSI (8 ou 16 bits) en collecteur 
        ouvert.</para>

      <para>Bien s&ucirc;r, le bus SCSI n'a pas que des fils de donn&eacute;es, mais
	aussi un certain nombre de signaux de contr&ocirc;le. Un protocole tr&egrave;s
	&eacute;labor&eacute; fait partie du standard pour permettre &agrave; plusieurs
	p&eacute;riph&eacute;riques de se partager le bus de mani&egrave;re efficace.
	En SCSI-2, les donn&eacute;es sont toujours v&eacute;rifi&eacute;es avec un fil s&eacute;par&eacute;
	pour la parit&eacute;. Dans l'architecture pr&eacute;-SCSI-2, la parit&eacute; &eacute;tait
	optionnelle.</para>

      <para>En SCSI-3, des types de bus encore plus rapides sont introduits,
        dont les bus SCSI s&eacute;rie qui r&eacute;duisent l'overhead du cabl&acirc;ge
	(consommation? d&eacute;lai de propagation?) 
        et permettent une longueur de bus maximale plus importante. 
        Vous pourriez voir des noms comme SSA et FiberChannel dans ce contexte.
	Aucun de ces bus s&eacute;rie n'est aujourd'hui d'usage courant (et
	sp&eacute;cialement pas dans l'environnement typique de FreeBSD).
	Pour cette raison, les types de bus s&eacute;rie ne seront plus abord&eacute;s.</para>

      <para>Comme vous auriez pu le deviner de la description pr&eacute;c&eacute;dente, les
	p&eacute;riph&eacute;riques SCSI sont intelligents. Ils doivent l'&ecirc;tre pour adh&eacute;rer
	au standard SCSI (qui est &eacute;pais de plus de 5 cm). Ainsi, pour un
	disque dur par exemple, vous ne sp&eacute;cifiez pas un t&ecirc;te/cylindre/secteur
	pour adresser un bloc particulier, mais simplement le num&eacute;ro du
	bloc que vous voulez.
	Des sch&eacute;mas &eacute;labor&eacute;s de cache, des remplacements automatiques de blocs
	d&eacute;fecteux, etc, sont tous rendus possibles par cette approche de
	&ldquo;p&eacute;riph&eacute;rique intelligent&rdquo;.</para>

      <para>Sur un bus SCSI, chaque paire possible d'abonn&eacute;s peut communiquer.
	Que leur fonction le leur permette est une autre chose, mais le
	standard ne le restreint pas. Pour &eacute;viter le conflit de signaux,
	les deux abonn&eacute;s doivent passer par une phase d'arbitrage de bus 
	avant de l'utiliser.</para>

      <para>La philosophie du SCSI est d'avoir un standard qui permette
	&agrave; des p&eacute;riph&eacute;riques ancien-standard de travailler avec des 
        nouveaux-standard. Ainsi, un vieux p&eacute;riph&eacute;rique SCSI-1 devrait 
        normalement fonctionner sur un bus SCSI-2. Je dis normalement, car il 
        n'est pas absolument s&ucirc;r que l'impl&eacute;mentation d'un ancien p&eacute;riph&eacute;rique 
        suive le (vieux) standard de mani&egrave;re assez proche pour &ecirc;tre acceptable 
        sur un nouveau bus. Les p&eacute;riph&eacute;riques modernes se comportent bien
	g&eacute;n&eacute;ralement, car la standardisation est devenue plus stricte et
	est mieux respect&eacute;e par les fabriquants de p&eacute;riph&eacute;riques.</para>
	
      <para>D'une mani&egrave;re g&eacute;n&eacute;rale, les chances de faire fonctionner
	correctement un ensemble de p&eacute;riph&eacute;riques sur un seul bus, sont
	meilleures quand tous les abonn&eacute;s sont SCSI-2 ou plus r&eacute;cents.
	Cela implique que vous n'avez pas besoin de supprimer tous vos vieux
	mat&eacute;riels quand vous venez d'avoir ce magnifique disque de 2Go&nbsp;:
	je poss&egrave;de un syst&egrave;me sur lequel un disque pr&eacute;-SCSI-1, un
	lecteur de cartouche QIC en SCSI-2, un lecteur de cartouches
	h&eacute;lico&iuml;dal SCSI-1 et 2 disques SCSI-1 fonctionnent assez
	bien ensemble. D'un point de vue des performances, vous
	pourriez toutefois vouloir s&eacute;parer vos plus vieux p&eacute;riph&eacute;riques
	des plus nouveaux (=plus rapides).</para>

      
      <sect3>
        <title>Composants SCSI</title>

	<para>Comme nous l'avons dit pr&eacute;c&eacute;demment, les p&eacute;riph&eacute;riques SCSI sont
	  intelligents. L'id&eacute;e est de mettre les connaissances sur les d&eacute;tails
	  intimes du mat&eacute;riel dans le p&eacute;riph&eacute;rique SCSI lui-m&ecirc;me. De cette
	  fa&ccedil;on, le syst&egrave;me h&ocirc;te n'a pas besoin de se pr&eacute;occuper de savoir,
	  par exemple, combien de t&ecirc;tes poss&egrave;de le disque, ou combien de pistes
	  poss&egrave;de tel d&eacute;rouleur de bandes. Si vous &ecirc;tes curieux, le standard
	  sp&eacute;cifie des commandes avec lesquelles vous pouvez interroger les
	  p&eacute;riph&eacute;riques sur leurs sp&eacute;cificit&eacute;s mat&eacute;rielles. FreeBSD utilise
	  cette possibilit&eacute; pendant le d&eacute;marrage pour d&eacute;terminer quels sont
	  les p&eacute;riph&eacute;riques connect&eacute;s et s'ils ont besoin d'un traitement
	  sp&eacute;cial.</para>

	<para>L'avantage d'avoir des p&eacute;riph&eacute;riques intelligents est
	  &eacute;vident&nbsp;: le pilote de p&eacute;riph&eacute;rique dans l'h&ocirc;te peut &ecirc;tre
	  con&ccedil;u de mani&egrave;re beaucoup plus g&eacute;n&eacute;rique, il n'y a plus besoin de
	  modifier (et valider&nbsp;!) les pilotes pour chaque nouveau
	  p&eacute;riph&eacute;rique bizarre qui est introduit.</para>

	<para>Pour les c&acirc;bles et les connecteurs, il y a une r&egrave;gle d'or&nbsp;:
	  prenez de la qualit&eacute;. Avec des vitesses de bus augmentant tout
	  le temps, vous vous &eacute;pargnerez beaucoup de peine en utilisant du
	  bon mat&eacute;riel.</para>

	<para>Aussi, utilisez des connecteurs plaqu&eacute;s or, des c&acirc;bles blind&eacute;s
	  et des connecteurs robustes et bien v&eacute;rrouill&eacute;s, etc.
	  Deuxi&egrave;me r&egrave;gle d'or&nbsp;: n'utilisez pas des c&acirc;bles plus longs que
	  n&eacute;cessaires. J'ai une fois perdu 3 jours &agrave; pourchasser un probl&egrave;me
	  sur une machine instable, juste pour d&eacute;couvrir que raccourcir
	  le bus SCSI d'un m&egrave;tre r&eacute;solvait le probl&egrave;me. Et la longueur
	  originale du bus respectait bien les sp&eacute;cifications SCSI.</para>

      </sect3>

      <sect3>
        <title>Types de bus SCSI</title>

	<para>D'un point de vue &eacute;lectrique, il existe deux types de bus
	  incompatibles&nbsp;: collecteur ouvert (<foreignphrase>single-ended
	  </foreignphrase>) et diff&eacute;rentiel. Cela signifie qu'il existe deux
	  principaux groupes de p&eacute;riph&eacute;riques et contr&ocirc;leurs SCSI qui ne peuvent
	  &ecirc;tre m&eacute;lang&eacute;s sur le m&ecirc;me bus. Il est toutefois possible d'utiliser
	  un convertisseur mat&eacute;riel sp&eacute;cial pour transformer un bus collecteur
	  ouvert en diff&eacute;rentiel (et vice versa). Les diff&eacute;rences entre les
	  types de bus sont expliqu&eacute;es dans les sections suivantes.</para>

        <para>Dans beaucoup de documentation &agrave; propos du SCSI, il existe une
	  sorte de jargon en usage pour abr&eacute;ger les diff&eacute;rents types de bus.
	  Une petite liste&nbsp;:</para>
            
        <itemizedlist>
          <listitem>
            <para>FWD&nbsp;: Fast Wide Differential (diff&eacute;rentiel large rapide)</para>
          </listitem>
          
          <listitem>
            <para>FND&nbsp;: Fast Narrow Differential (diff&eacute;rentiel &eacute;troit rapide)</para>
          </listitem>
          
          <listitem>
            <para>SE&nbsp;: Single Ended (collecteur ouvert)</para>
          </listitem>
          
          <listitem>
            <para>FN&nbsp;: Fast Narrow (rapide &eacute;troit)</para>
          </listitem>
          
          <listitem>
            <para>etc.</para>
          </listitem>
        </itemizedlist>
        
        
        <para>Avec un minimun d'imagination, on peut bien imaginer ce que
	  cela veut dire.</para>
            
        <para>Large est un peu ambigu, il peut indiquer des bus de 16 ou
	  32 bits. A ma connaissance, la variante en 32 bits n'est pas (encore)
	  utilis&eacute;e, donc normalement large veut dire 16 bits.</para>
            
        <para>Rapide signifie que la cadence sur le bus est un peu diff&eacute;rente,
	  pour qu'un bus &eacute;troit (8 bits) supporte 10 Mo/s au lieu de 5 Mo/s
	  pour un SCSI 'lent'. Comme indiqu&eacute; pr&eacute;c&eacute;demment, des vitesses de
	  bus de 20 et 40 millions de transferts/seconde &eacute;mergent aussi
	  (Fast-20 = Ultra SCSI et Fast40 = Ultra2 SCSI).</para>
	  
        <note>
          <para>Les lignes de donn&eacute;es &gt; 8 ne sont utilis&eacute;es que pour les
	    transferts de donn&eacute;es et l'adressage des p&eacute;riph&eacute;riques. Les
	    transferts des commandes, messages d'&eacute;tat, etc. n'ont lieu que sur
	    les 8 bits de poids faibles. Le standard permet aux p&eacute;riph&eacute;riques
	    &eacute;troits de fonctionner sur un bus large. La largeur de bus
	    utilisable est n&eacute;goci&eacute;e entre les abonn&eacute;s. Vous devez regarder
	    pr&eacute;cis&eacute;ment l'adressage des abonn&eacute;s lorsque vous m&eacute;langez larges
	    et &eacute;troits.</para>
        </note>
        
<!--
        <sect4>
          <title>Single ended buses</title>
	   &sgml.todo;
          
          <para>A single-ended SCSI bus uses signals that are either 5 Volts
            or 0 Volts (indeed, TTL levels) and are relative to a COMMON
            ground reference.  A singled ended 8 bit SCSI bus has
            approximately 25 ground lines, who are all tied to a single `rail'
            on all devices.  A standard single ended bus has a maximum length
            of 6 meters.  If the same bus is used with fast-SCSI devices, the
            maximum length allowed drops to 3 meters.  Fast-SCSI means that
            instead of 5Mbytes/sec the bus allows 10Mbytes/sec
            transfers.</para>
              
          <para>Fast-20 (Ultra SCSI) and Fast-40 allow for 20 and 40 million
            transfers/second respectively.  So, F20 is 20 Mbytes/second on a 8
            bit bus, 40 Mbytes/second on a 16 bit bus etc.  For F20 the max
            bus length is 1.5 meters, for F40 it becomes 0.75 meters.  Be
            aware that F20 is pushing  the limits quite a bit, so you will
            quickly find out if your SCSI bus is electrically sound.</para>

          <note>
            <para>If some devices on your bus use 'fast' to communicate your
              bus must adhere to the length restrictions for fast
              buses!</para>
          </note>
          
          <para>It is obvious that with the newer fast-SCSI devices the bus
            length can become a real bottleneck.  This is why the differential
            SCSI bus was introduced in the SCSI-2 standard.</para>
              
          <para>For connector pinning and connector types please refer to the
            SCSI-2 standard (see <link linkend="scsi-further-reading">Further
              reading</link>) itself, connectors etc are listed there in
            painstaking detail.</para>
          
          <para>Beware of devices using non-standard cabling.  For instance
            Apple uses a 25pin D-type connecter (like the one on serial ports
            and parallel printers).  Considering that the official SCSI bus
            needs 50 pins you can imagine the use of this connector needs some
            'creative cabling'.  The reduction of the number of ground wires
            they used is a bad idea, you better stick to 50 pins cabling  in
            accordance with the SCSI standard.  For Fast-20 and 40 do not even
            think about buses like this.</para>
        </sect4>
        
        <sect4>
          <title>Differential buses</title>
          
          <para>A differential SCSI bus has a maximum length of 25 meters.
            Quite a difference from the 3 meters for a single-ended fast-SCSI
            bus.  The idea behind differential signals is that each bus signal
            has its own return wire.  So, each signal is carried on a
            (preferably twisted) pair of wires.  The voltage difference
            between these two wires determines whether the signal is asserted
            or de-asserted.  To a certain extent the voltage difference
            between ground and the signal wire pair is not relevant (do not
            try 10 kVolts though).</para>
              
          <para>It is beyond the scope of this document to explain why this
            differential idea is so much better.  Just accept that
            electrically seen the use of differential signals gives a much
            better noise margin.  You will normally find differential buses in
            use for inter-cabinet connections.  Because of the lower cost
            single ended is mostly used for shorter buses like inside
            cabinets.</para>
              
          <para>There is nothing that stops you from using differential stuff
            with FreeBSD, as long as you use a controller that has device
            driver support in FreeBSD.  As an example, Adaptec marketed the
            AHA1740 as a single ended board, whereas the AHA1744 was
            differential.  The software interface to the host is identical for
            both.</para>
        </sect4>
        
        <sect4>
          <title>Terminators</title>
          
          <para>Terminators in SCSI terminology are resistor networks that are
            used to get a correct impedance matching.  Impedance matching is
            important to get clean signals on the bus, without reflections or
            ringing.  If you once made a long distance telephone call on a bad
            line you probably know what reflections are.  With 20Mbytes/sec
            traveling over your SCSI bus, you do not want signals echoing
            back.</para>
              
          <para>Terminators come in various incarnations, with more or less
            sophisticated designs.  Of course, there are internal and external
            variants.  Many SCSI devices come with a number of sockets in
            which a number of resistor networks can (must be!) installed.  If
            you remove terminators from a device, carefully store them.  You
            will need them when you ever decide to reconfigure your SCSI bus.
            There is enough variation in even these simple tiny things to make
            finding the exact replacement a frustrating business.  There are
            also SCSI devices that have a single jumper to enable or disable a
            built-in terminator. There are special terminators you can stick
            onto a flat cable bus.  Others look like external connectors, or a
            connector hood without a cable.  So, lots of choice as you can
            see.</para>
              
          <para>There is much debate going on if and when you should switch
            from simple resistor (passive) terminators to active terminators.
            Active terminators contain slightly more elaborate circuit to give
            cleaner bus signals.  The general consensus seems to be that the
            usefulness of active termination increases when you have long
            buses and/or fast devices.  If you ever have problems with your
            SCSI buses you might consider trying an active terminator.  Try to
            borrow one first, they reputedly are quite expensive.</para>
              
          <para>Please keep in mind that terminators for differential and
            single-ended buses are not identical.  You should <emphasis>not
              mix</emphasis> the two variants.</para>
              
          <para>OK, and now where should you install your terminators? This is
            by far the most misunderstood part of SCSI.  And it is by far the
            simplest.  The rule is: <emphasis>every single line on the SCSI
              bus has 2 (two) terminators, one at each end of the
              bus.</emphasis> So, two and not one or three or whatever.  Do
            yourself a favor and stick to this rule.  It will save you endless
            grief, because wrong termination has the potential to introduce
            highly mysterious bugs.  (Note the &ldquo;potential&rdquo; here;
            the nastiest part is that it may or may not work.)</para>
              
          <para>A common pitfall is to have an internal (flat) cable in a
            machine and also an external cable attached to the controller. It
            seems almost everybody forgets to remove the terminators from the
            controller.  The terminator must now be on the last external
            device, and not on the controller! In general, every
            reconfiguration of a SCSI bus must pay attention to this.</para>

          <note>
            <para>Termination is to be done on a per-line basis.  This means
              if you have both narrow and wide buses connected to the same
              host adapter, you need to enable termination on the higher 8
              bits of the bus on the adapter (as well as the last devices on
              each bus, of course).</para>
              </note>
              
          <para>What I did myself is remove all terminators from my SCSI
            devices and controllers.  I own a couple of external terminators,
            for both the Centronics-type external cabling and for the internal
            flat cable connectors.  This makes reconfiguration much
            easier.</para>
              
          <para>On modern devices, sometimes integrated terminators are used.
            These things are special purpose integrated circuits that can be
            dis/en-abled with a control pin.  It is not necessary to
            physically remove them from a device.  You may find them on newer
            host adapters, sometimes they are software configurable, using
            some sort of setup tool.  Some will even auto-detect the cables
            attached to the connectors and automatically set up the
            termination as necessary.  At any rate, consult your
            documentation!</para>
        </sect4>
        
        <sect4>
          <title>Terminator power</title>
          
          <para>The terminators discussed in the previous chapter need power
            to operate properly.  On the SCSI bus, a line is dedicated to this
            purpose.  So, simple huh?</para>
              
          <para>Not so.  Each device can provide its own terminator power to
            the terminator sockets it has on-device.  But if you have external
            terminators, or when the device supplying the terminator power to
            the SCSI bus line is switched off you are in trouble.</para>
              
          <para>The idea is that initiators (these are devices that initiate
            actions on the bus, a discussion follows) must supply terminator
            power.  All SCSI devices are allowed (but not required) to supply
            terminator power.</para>
              
          <para>To allow for un-powered devices on a bus, the terminator power
            must be supplied to the bus via a diode.  This prevents the
            backflow of current to un-powered devices.</para>
              
          <para>To prevent all kinds of nastiness, the terminator power is
            usually fused.  As you can imagine, fuses might blow.  This can,
            but does not have to, lead to a non functional bus.  If multiple
            devices supply terminator power, a single blown fuse will not put
            you out of business.  A single supplier with a blown fuse
            certainly will.  Clever external terminators sometimes have a  LED
            indication that shows whether terminator power is present.</para>
              
          <para>In newer designs auto-restoring fuses that 'reset' themselves
            after some time are sometimes used.</para>
        </sect4>
        
        <sect4>
          <title>Device addressing</title>
          
          <para>Because the SCSI bus is, ehh, a bus there must be a way to
            distinguish or address the different devices connected to
            it.</para>
              
          <para>This is done by means of the SCSI or target ID.  Each device
            has a unique target ID.  You can select the ID to which a device
            must respond using a set of jumpers, or a dip switch, or something
            similar.  Some SCSI host adapters let you change the target ID
            from the boot menu.  (Yet some others will not let you change the
            ID from 7.)  Consult the documentation of your device for more
            information.</para>
          
          <para>Beware of multiple devices configured to use the same ID.
            Chaos normally reigns in this case.  A pitfall is that one of the
            devices sharing the same ID sometimes even manages to answer to
            I/O requests!</para>
              
          <para>For an 8 bit bus, a maximum of 8 targets is possible.  The
            maximum is 8 because the selection is done bitwise using the 8
            data lines on the bus.  For wide buses this increases to the
            number of data lines (usually 16).</para>

          <note>
            <para>A narrow SCSI device can not communicate with a SCSI device
              with a target ID larger than 7.  This means it is generally not
              a good idea to move your SCSI host adapter's target ID to
              something higher than 7 (or your CD-ROM will stop
              working).</para>
          </note>
          
          <para>The higher the SCSI target ID, the higher the priority the
            devices has.  When it comes to arbitration between devices that
            want to use the bus at the same time, the device that has the
            highest SCSI ID will win.  This also means that the SCSI host
            adapter usually uses target ID 7.  Note however that the lower 8
            IDs have higher priorities than the higher 8 IDs on a wide-SCSI
            bus.  Thus, the order of target IDs is: [7 6 .. 1 0 15 14 .. 9 8]
            on a wide-SCSI system.  (If you you are wondering why the lower 8
            have higher priority, read the previous paragraph for a
            hint.)</para>
              
          <para>For a further subdivision, the standard allows for Logical
            Units or LUNs for short.  A single target ID may have multiple
            LUNs.  For example, a tape device including a tape changer may
            have LUN 0 for the tape device itself, and LUN 1 for the tape
            changer.  In this way, the host system can address each of the
            functional units of the tape changer as desired.</para>
        </sect4>
            
        <sect4>
          <title>Bus layout</title>
          
          <para>SCSI buses are linear.  So, not shaped like Y-junctions, star
            topologies, rings, cobwebs or whatever else people might want to
            invent.  One of the most common mistakes is for people with
            wide-SCSI host adapters to connect devices on all three connecters
            (external connector, internal wide connector, internal narrow
            connector).  Don't do that.  It may appear to work if you are
            really lucky, but I can almost guarantee that your system will
            stop functioning at the most unfortunate moment (this is also
            known as &ldquo;Murphy's law&rdquo;).</para>
              
          <para>You might notice that the terminator issue discussed earlier
            becomes rather hairy if your bus is not linear.  Also, if you have
            more connectors than devices on your internal SCSI cable, make
            sure you attach devices on connectors on both ends instead of
            using the connectors in the middle and let one or both ends
            dangle.  This will screw up the termination of the bus.</para>
              
          <para>The electrical characteristics, its noise margins and
            ultimately the reliability of it all are tightly related to linear
            bus rule.</para>
          
          <para><emphasis>Stick to the linear bus rule!</emphasis></para>
        </sect4>
-->

      </sect3>


      <sect3>
        <title>*** Utiliser le SCSI avec FreeBSD</title>
	   &sgml.todo
    <!--
        <sect4>
          <title>About translations, BIOSes and magic...</title>
          
          <para>As stated before, you should first make sure that you have a
            electrically sound bus.</para>
              
          <para>When you want to use a SCSI disk on your PC as boot disk, you
            must aware of some quirks related to PC BIOSes.  The PC BIOS in
            its first incarnation used a low level physical interface to the
            hard disk.  So, you had to tell the BIOS (using a setup tool or a
            BIOS built-in setup) how your disk physically looked like.  This
            involved stating number of heads, number of cylinders, number of
            sectors per track, obscure things like precompensation and reduced
            write current cylinder etc.</para>
              
          <para>One might be inclined to think that since SCSI disks are smart
            you can forget about this.  Alas, the arcane setup issue is still
            present today.  The system BIOS needs to know how to access your
            SCSI disk with the head/cyl/sector method in order to load the
            FreeBSD kernel during boot.</para>
              
          <para>The SCSI host adapter or SCSI controller you have put in your
            AT/EISA/PCI/whatever bus to connect your disk therefore has its
            own on-board BIOS.  During system startup, the SCSI BIOS takes
            over the hard disk interface routines from the system BIOS.  To
            fool the system BIOS, the system setup is normally set to No hard
            disk present.  Obvious, isn't it?</para>
              
          <para>The SCSI BIOS itself presents to the system a so called
            <emphasis>translated</emphasis> drive.  This means that a fake
            drive table is constructed that allows the PC to boot the drive.
            This translation is often (but not always) done using a pseudo
            drive with 64 heads and 32 sectors per track.  By varying the
            number of cylinders, the SCSI BIOS adapts to the actual drive
            size.  It is useful to note that 32 * 64 / 2 = the size of your
            drive in megabytes.  The division by 2 is to get from disk blocks
            that are normally 512 bytes in size to Kbytes.</para>
              
          <para>Right.  All is well now?! No, it is not.  The system BIOS has
            another quirk you might run into.  The number of cylinders of a
            bootable hard disk cannot be greater than 1024.  Using the
            translation above, this is a show-stopper for disks greater than 1
            GB.  With disk capacities going up all the time this is causing
            problems.</para>
              
          <para>Fortunately, the solution is simple: just use another
            translation, e.g. with 128 heads instead of 32.  In most cases new
            SCSI BIOS versions are available to upgrade older SCSI host
            adapters.  Some newer adapters have an option, in the form of a
            jumper or software setup selection, to switch the translation the
            SCSI BIOS uses.</para>
              
          <para>It is very important that <emphasis>all</emphasis> operating
            systems on the disk use the <emphasis>same translation</emphasis>
            to get the right idea about where to find the relevant partitions.
            So, when installing FreeBSD you must answer any questions about
            heads/cylinders etc using the translated values your host adapter
            uses.</para>
              
          <para>Failing to observe the translation issue might lead to
            un-bootable systems or operating systems overwriting each others
            partitions.  Using fdisk you should be able to see all
            partitions.</para>
              
          <para>You might have heard some talk of &ldquo;lying&rdquo; devices?
            Older FreeBSD kernels used to report the geometry of SCSI disks
            when booting.  An example from one of my systems:</para>
              
          <screen>aha0 targ 0 lun 0: &lt;MICROP 1588-15MB1057404HSP4&gt;
sd0: 636MB (1303250 total sec), 1632 cyl, 15 head, 53 sec, bytes/sec 512</screen>

          <para>Newer kernels usually do not report this information.
            e.g.</para>

          <screen>(bt0:0:0): "SEAGATE ST41651 7574" type 0 fixed SCSI 2
sd0(bt0:0:0): Direct-Access 1350MB (2766300 512 byte sectors)</screen>
                
          <para>Why has this changed?</para>
              
          <para>This info is retrieved from the SCSI disk itself.  Newer disks
            often use a technique called zone bit recording.  The idea is that
            on the outer cylinders of the drive there is more space so more
            sectors per track can be put on them.  This results in disks that
            have more tracks on outer cylinders than on the inner cylinders
            and, last but not least, have more capacity.  You can imagine that
            the value reported by the drive when inquiring about the geometry
            now becomes suspect at best, and nearly always misleading.  When
            asked for a geometry , it is nearly always better to supply the
            geometry used by the BIOS, or <emphasis>if the BIOS is never going
              to know about this disk</emphasis>, (e.g. it is not a booting
            disk) to supply a fictitious geometry that is convenient.</para>
        </sect4>
        
        <sect4>
          <title>SCSI subsystem design</title>
          
          <para>FreeBSD uses a layered SCSI subsystem.  For each different
            controller card a device driver is written.  This driver knows all
            the intimate details about the hardware it controls.  The driver
            has a interface to the upper layers of the SCSI subsystem through
            which it receives its commands and reports back any status.</para>
              
          <para>On top of the card drivers there are a number of more generic
            drivers for a class of devices.  More specific: a driver for tape
            devices (abbreviation: st), magnetic disks (sd), CD-ROMs (cd) etc.
            In case you are wondering where you can find this stuff, it all
            lives in <filename>/sys/scsi</filename>.  See the man pages in
            section 4 for more details.</para>
              
          <para>The multi level design allows a decoupling of low-level bit
            banging and more high level stuff.  Adding support for another
            piece of hardware is a much more manageable problem.</para>
        </sect4>
        
        <sect4>
          <title>Kernel configuration</title>
          
          <para>Dependent on your hardware, the kernel configuration file must
            contain one or more lines describing your host adapter(s).  This
            includes I/O addresses, interrupts etc. Consult the man page for
            your adapter driver to get more info. Apart from that, check out
            <filename>/sys/i386/conf/LINT</filename> for an overview of a
            kernel config file.  <filename>LINT</filename> contains every
            possible option you can dream of.  It does
            <emphasis>not</emphasis> imply <filename>LINT</filename> will
            actually get you to a working kernel at all.</para>
          
          <para>Although it is probably stating the obvious: the kernel config
            file should reflect your actual hardware setup.  So, interrupts,
            I/O addresses etc must match the kernel config file.  During
            system boot messages will be displayed to indicate whether the
            configured hardware was actually found.</para>

          <note>
            <para>Note that most of the EISA/PCI drivers (namely
              <devicename>ahb</devicename>, <devicename>ahc</devicename>,
              <devicename>ncr</devicename> and <devicename>amd</devicename>
              will automatically obtain the correct parameters from the host
              adapters themselves at boot time; thus, you just need to write,
              for instance, <literal>controller ahc0</literal>.</para>
          </note>
          
          <para>An example loosely based on the FreeBSD 2.2.5-Release kernel
            config  file <filename>LINT</filename> with some added comments
            (between []):</para>
              
          <programlisting>
# SCSI host adapters: `aha', `ahb', `aic', `bt', `nca'
#
# aha: Adaptec 154x
# ahb: Adaptec 174x
# ahc: Adaptec 274x/284x/294x
# aic: Adaptec 152x and sound cards using the Adaptec AIC-6360 (slow!)
# amd: AMD 53c974 based SCSI cards (e.g., Tekram DC-390 and 390T)
# bt: Most Buslogic controllers
# nca: ProAudioSpectrum cards using the NCR 5380 or Trantor T130
# ncr: NCR/Symbios 53c810/815/825/875 etc based SCSI cards
# uha: UltraStore 14F and 34F
# sea: Seagate ST01/02 8 bit controller (slow!)
# wds: Western Digital WD7000 controller (no scatter/gather!).
#

[For an Adaptec AHA274x/284x/294x/394x etc controller]
controller        ahc0

[For an NCR/Symbios 53c875 based controller]
controller        ncr0

[For an Ultrastor adapter]
controller        uha0        at isa? port "IO_UHA0" bio irq ? drq 5 vector uhaintr

# Map SCSI buses to specific SCSI adapters
controller        scbus0        at ahc0
controller        scbus2 at ncr0
controller        scbus1  at uha0

# The actual SCSI devices
disk sd0 at scbus0 target 0 unit 0        [SCSI disk 0 is at scbus 0, LUN 0]
disk sd1 at scbus0 target 1             [implicit LUN 0 if omitted]
disk sd2 at scbus1 target 3             [SCSI disk on the uha0]
disk sd3 at scbus2 target 4             [SCSI disk on the ncr0]
tape st1 at scbus0 target 6             [SCSI tape at target 6]
device cd0 at scbus?                    [the first ever CD-ROM found, no wiring]</programlisting>
              
          <para>The example above tells the kernel to look for a ahc (Adaptec
            274x) controller, then for an NCR/Symbios board, and so on.  The
            lines following the controller specifications  tell the kernel to
            configure specific devices but <emphasis>only</emphasis> attach
            them when they match the target ID and LUN specified on the
            corresponding bus.</para>
              
          <para>Wired down devices get &ldquo;first shot&rdquo; at the unit
            numbers so the first non &ldquo;wired down&rdquo; device, is
            allocated the unit number  one greater than the highest
            &ldquo;wired down&rdquo; unit number for that kind of device.  So,
            if you had a SCSI tape at target ID 2 it would be configured as
            st2, as the tape at target ID 6 is wired down to unit number
            1.</para>

          <note>
            <para>Wired down devices need not be found to get their unit
              number.  The unit number for a wired down device is reserved for
              that device, even if it is turned off at boot time.  This allows
              the device to be turned on and brought on-line at a later time,
              without rebooting.  Notice that a device's unit number has
              <emphasis>no</emphasis> relationship with its target ID on  the
              SCSI bus.</para>
          </note>
          
          <para>Below is another example of a kernel config file as used by
            FreeBSD version &lt; 2.0.5.  The difference with the first example
            is that devices are not &ldquo;wired down&rdquo;.  &ldquo;Wired
            down&rdquo; means that you specify which SCSI target belongs to
            which device.</para>
              
          <para>A kernel built to the config file below will attach  the first
            SCSI disk it finds to sd0, the second disk to sd1 etc. If you ever
            removed or added a disk, all other devices of the same type (disk
            in this case) would 'move around'.  This implies you have to
            change <filename>/etc/fstab</filename> each time.</para>
              
          <para>Although the old style still works, you  are
            <emphasis>strongly</emphasis> recommended to use this new feature.
            It will save you a lot of grief whenever you shift your hardware
            around on the SCSI buses.  So, when you re-use your old trusty
            config file after upgrading from a pre-FreeBSD2.0.5.R system check
            this out.</para>
          
          <programlisting>
[driver for Adaptec 174x]
controller      ahb0 at isa? bio irq 11 vector ahbintr

[for Adaptec 154x]
controller      aha0    at isa? port "IO_AHA0" bio irq 11 drq 5 vector ahaintr

[for Seagate ST01/02]
controller      sea0    at isa? bio irq 5 iomem 0xc8000 iosiz 0x2000 vector seaintr

controller      scbus0

device          sd0     [support for 4 SCSI harddisks, sd0 up sd3]
device          st0        [support for 2 SCSI tapes]

[for the CD-ROM]
device          cd0     #Only need one of these, the code dynamically grows</programlisting>
              
          <para>Both examples support SCSI disks.  If during boot more devices
            of a specific type (e.g. sd disks) are found than are configured
            in the booting kernel, the system will simply allocate more
            devices, incrementing the unit number starting at the last number
            &ldquo;wired down&rdquo;.  If there are no &ldquo;wired
            down&rdquo; devices then counting starts at unit 0.</para>
              
          <para>Use <command>man 4 scsi</command> to check for the latest info
            on the SCSI subsystem.  For more detailed info on host adapter
            drivers use eg <command>man 4 ahc</command> for info on the
            Adaptec 294x driver.</para>
        </sect4>
        
        <sect4>
          <title>Tuning your SCSI kernel setup</title>
          
          <para>Experience has shown that some devices are slow to respond to
            INQUIRY commands after a SCSI bus reset (which happens at boot
            time).  An INQUIRY command is sent by the kernel on boot to see
            what kind of device (disk, tape, CD-ROM etc) is connected to a
            specific target ID.  This process is called device probing by the
            way.</para>
              
          <para>To work around the 'slow response' problem, FreeBSD allows a
            tunable delay time before the SCSI devices are probed following a
            SCSI bus reset.  You can set this delay time in your kernel
            configuration file using a line like:</para>
              
          <programlisting>
options         SCSI_DELAY=15         #Be pessimistic about Joe SCSI device</programlisting>

          <para>This line sets the delay time to 15 seconds.  On my own system
            I had to use 3 seconds minimum to get my trusty old CD-ROM drive
            to be recognized.  Start with a high value (say 30 seconds or so)
            when you have problems  with device recognition.  If this helps,
            tune it back until it just stays working.</para>
        </sect4>
        
        <sect4 id="scsi-rogue-devices">
          <title>Rogue SCSI devices</title>
              
          <para>Although the SCSI standard tries to be complete and concise,
            it is a complex standard and implementing things correctly is no
            easy task.  Some vendors do a better job then others.</para>
              
          <para>This is exactly where the &ldquo;rogue&rdquo; devices come
            into view. Rogues are devices that are recognized by the FreeBSD
            kernel as behaving slightly (...) non-standard.  Rogue devices are
            reported by the kernel when booting.  An example for two of my
            cartridge tape units:</para>
          
          <screen>Feb 25 21:03:34 yedi /kernel: ahb0 targ 5 lun 0: &lt;TANDBERG TDC 3600       -06:&gt;
Feb 25 21:03:34 yedi /kernel: st0: Tandberg tdc3600 is a known rogue

Mar 29 21:16:37 yedi /kernel: aha0 targ 5 lun 0: &lt;ARCHIVE VIPER 150  21247-005&gt;
Mar 29 21:16:37 yedi /kernel: st1: Archive  Viper 150 is a known rogue </screen>
              
          <para>For instance, there are devices that respond to all LUNs on a
            certain target ID, even if they are actually only one device.  It
            is easy to see that the kernel might be fooled into believing that
            there are 8 LUNs at that particular target ID. The confusion this
            causes is left as an exercise to the reader.</para>
              
          <para>The SCSI subsystem of FreeBSD recognizes devices with bad
            habits by looking at the INQUIRY response they send when probed.
            Because the INQUIRY response also includes the version number of
            the device  firmware, it is even possible that for different
            firmware versions different workarounds are used. See e.g.
            <filename>/sys/scsi/st.c</filename> and
            <filename>/sys/scsi/scsiconf.c</filename> for more info on how
            this is done.</para>
              
          <para>This scheme works fine, but keep in mind that it of course
            only works for devices that are known to be weird.  If you are the
            first to connect your bogus Mumbletech SCSI CD-ROM you might be
            the one that has to define which workaround is needed.</para>
              
          <para>After you got your Mumbletech working, please send the
            required workaround to the FreeBSD development team for inclusion
            in the next release of FreeBSD.  Other Mumbletech owners will be
            grateful  to you.</para>
        </sect4>
        
        <sect4>
          <title>Multiple LUN devices</title>
          
          <para>In some cases you come across devices that use multiple
            logical units (LUNs) on a single SCSI ID.  In most cases FreeBSD
            only probes devices for LUN 0.  An example are so called bridge
            boards that connect 2 non-SCSI harddisks to a SCSI bus (e.g. an
            Emulex MD21 found in old Sun systems).</para>
              
          <para>This means that any devices with LUNs != 0 are not normally
            found during device probe on system boot.  To work around this
            problem you must add an appropriate entry in /sys/scsi/scsiconf.c
            and rebuild your kernel.</para>
              
          <para>Look for a struct that is initialized like below:</para>

          <programlisting>
{
        T_DIRECT, T_FIXED, "MAXTOR", "XT-4170S", "B5A",
        "mx1", SC_ONE_LU
}</programlisting>
              
          <para>For you Mumbletech BRIDGE2000 that has more than one LUN, acts
            as a SCSI disk and has firmware revision 123 you would add
            something like:</para>
              
          <programlisting>
{
        T_DIRECT, T_FIXED, "MUMBLETECH", "BRIDGE2000", "123",
        "sd", SC_MORE_LUS
}</programlisting>
              
          <para>The kernel on boot scans the inquiry data it receives against
            the table and acts accordingly.  See the source for more
            info.</para>
        </sect4>
        
        <sect4>
          <title>Tagged command queueing</title>
          
          <para>Modern SCSI devices, particularly magnetic disks,
            support what is called tagged command queuing (TCQ).</para>
          
          <para>In a nutshell, TCQ allows the device to have multiple I/O
            requests outstanding at the same time.  Because the device is
            intelligent, it can optimise its operations (like head
            positioning) based on its own request queue.  On  SCSI devices
            like RAID (Redundant Array of Independent Disks) arrays the TCQ
            function is indispensable to take advantage of the device's
            inherent parallelism.</para>
              
          <para>Each I/O request is uniquely identified by a &ldquo;tag&rdquo;
            (hence the name tagged command queuing) and this tag is used by
            FreeBSD to see which I/O in the device drivers queue is reported
            as complete by the device.</para>
              
          <para>It should be noted however that TCQ requires device driver
            support and that some devices implemented it &ldquo;not quite
            right&rdquo; in their firmware.  This problem bit me once, and it
            leads to highly mysterious problems.  In such cases, try to
            disable TCQ.</para>
        </sect4>
        
        <sect4>
          <title>Busmaster host adapters</title>
          
          <para>Most, but not all, SCSI host adapters are bus mastering
            controllers.  This means that they can do I/O on their own without
            putting load onto the host CPU for data movement.</para>
              
          <para>This is of course an advantage for a multitasking operating
            system like FreeBSD.  It must be noted however that there might be
            some rough edges.</para>
              
          <para>For instance an Adaptec 1542 controller can be set to use
            different transfer speeds on the host bus (ISA or AT in this
            case).  The controller is settable to different rates because not
            all motherboards can handle the higher speeds.  Problems like
            hangups, bad data etc might be the result of using a higher data
            transfer rate then your motherboard can stomach.</para>
              
          <para>The solution is of course obvious: switch to a lower data
            transfer rate and try if that works better.</para>
              
          <para>In the case of a Adaptec 1542, there is an option that can be
            put into the kernel config file to allow dynamic determination of
            the right, read: fastest feasible, transfer rate.  This option is
            disabled by default:</para>
              
          <programlisting>
options        "TUNE_1542"             #dynamic tune of bus DMA speed</programlisting>
              
          <para>Check the man pages for the host adapter that you use.  Or
            better still, use the ultimate documentation (read: driver
            source).</para>
        </sect4>
    -->
          </sect3>



          <sect3>
            <title>*** R&eacute;soudre les probl&egrave;mes</title>
	      &sgml.todo

   <!--
        <para>The following list is an attempt to give a guideline for the
          most common SCSI problems and their solutions.  It is by no means
          complete.</para>
            
        <itemizedlist>
          <listitem>
            <para>Check for loose connectors and cables.</para>
          </listitem>
          
          <listitem>
            <para>Check and double check the location and number of your
              terminators.</para>
          </listitem>
          
          <listitem>
            <para>Check if your bus has at least one supplier of terminator
              power (especially with external terminators.</para>
          </listitem>
          
          <listitem>
            <para>Check if no double target IDs are used.</para>
          </listitem>
          
          <listitem>
            <para>Check if all devices to be used are powered up.</para>
          </listitem>
          
          <listitem>
            <para>Make a minimal bus config with as little devices as
              possible.</para>
          </listitem>
          
          <listitem>
            <para>If possible, configure your host adapter to use slow bus
              speeds.</para>
          </listitem>
          
          <listitem>
            <para>Disable tagged command queuing to make things as simple as
              possible (for a NCR hostadapter based system see man
              ncrcontrol)</para>
          </listitem>
          
          <listitem>
            <para>If you can compile a kernel, make one with the
              <literal>SCSIDEBUG</literal> option, and try accessing the
              device with debugging turned on for that device.  If your device
              does not even probe at startup, you may have to define the
              address of the device that is failing, and the desired debug
              level in <filename>/sys/scsi/scsidebug.h</filename>. If it
              probes but just does not work, you can use the
                    &man.scsi.8; command to dynamically set a debug level to
              it in a running kernel (if <literal>SCSIDEBUG</literal> is
              defined).  This will give you <emphasis>copious</emphasis>
              debugging output with which to confuse the gurus. See
              <command>man 4 scsi</command> for more exact information.  Also
              look at <command>man 8 scsi</command>.</para>
          </listitem>
        </itemizedlist>
    -->
          </sect3>


          <sect3 id="scsi-further-reading">
            <title>*** Lectures compl&eacute;mentaires</title>
		  &sgml.todo

  <!--
        <para>If you intend to do some serious SCSI hacking, you might want to
          have the official standard at hand:</para>
            
        <para>Approved American National Standards can be purchased from
          ANSI at
          
          <address>
            <otheraddr>13th Floor</otheraddr>
            <street>11 West 42nd Street</street>
            <city>New York</city>
            <state>NY</state> <postcode>10036</postcode>
            Sales Dept: <phone>(212) 642-4900</phone>
          </address>
        </para>
        
        <para>You can also buy many ANSI
          standards and most committee draft documents from Global
          Engineering Documents,

          <address>
            <street>15 Inverness Way East</street>
            <city>Englewood</city>
            <state>CO</state>, <postcode>80112-5704</postcode>
            Phone: <phone>(800) 854-7179</phone>
            Outside USA and Canada: <phone>(303) 792-2181</phone>
            Fax: <fax>(303) 792- 2192</fax>
          </address>
        </para>
        
        <para>Many X3T10 draft documents are available electronically on the
          SCSI BBS (719-574-0424) and on the <hostid
            role="fqdn">ncrinfo.ncr.com</hostid> anonymous ftp site.</para>
            
        <para>Latest X3T10 committee documents are:</para>
        
        <itemizedlist>
          <listitem>
            <para>AT Attachment (ATA or IDE) [X3.221-1994]
              (<emphasis>Approved</emphasis>)</para>
          </listitem>
          
          <listitem>
            <para>ATA Extensions (ATA-2) [X3T10/948D Rev 2i]</para>
          </listitem>
          
          <listitem>
            <para>Enhanced Small Device Interface (ESDI)
              [X3.170-1990/X3.170a-1991]
              (<emphasis>Approved</emphasis>)</para>
          </listitem>
          
          <listitem>
            <para>Small Computer System Interface &mdash; 2 (SCSI-2)
              [X3.131-1994] (<emphasis>Approved</emphasis>)</para>
          </listitem>
          
          <listitem>
            <para>SCSI-2 Common Access Method Transport and SCSI Interface
              Module (CAM)  [X3T10/792D Rev 11]</para>
          </listitem>
        </itemizedlist>
        
        <para>Other publications that might provide you with additional
          information are:</para>
        
        <itemizedlist>
          <listitem>
            <para>&ldquo;SCSI: Understanding the Small Computer System
              Interface&rdquo;, written by NCR  Corporation.  Available from:
              Prentice Hall, Englewood Cliffs, NJ, 07632 Phone: (201) 767-5937
              ISBN 0-13-796855-8</para>
          </listitem>
          
          <listitem>
            <para>&ldquo;Basics of SCSI&rdquo;, a SCSI tutorial written by
              Ancot Corporation Contact Ancot for availability information at:
              Phone: (415) 322-5322  Fax: (415) 322-0455</para>
          </listitem>
          
          <listitem>
            <para>&ldquo;SCSI Interconnection Guide Book&rdquo;, an AMP
              publication (dated 4/93, Catalog  65237) that lists the various
              SCSI connectors and suggests cabling schemes.  Available from
              AMP at (800) 522-6752 or (717) 564-0100</para>
          </listitem>
          
          <listitem>
            <para>&ldquo;Fast Track to SCSI&rdquo;, A Product Guide written by
              Fujitsu.  Available from: Prentice Hall, Englewood Cliffs, NJ,
              07632 Phone: (201) 767-5937 ISBN 0-13-307000-X</para>
          </listitem>
          
          <listitem>
            <para>&ldquo;The SCSI Bench Reference&rdquo;, &ldquo;The SCSI
              Encyclopedia&rdquo;, and the &ldquo;SCSI Tutor&rdquo;, ENDL
              Publications, 14426 Black Walnut Court, Saratoga CA, 95070
              Phone: (408) 867-6642</para>
          </listitem>
          
          <listitem>
            <para>&ldquo;Zadian SCSI Navigator&rdquo; (quick ref. book) and
              &ldquo;Discover the Power of SCSI&rdquo;  (First book along with
              a one-hour video and tutorial book), Zadian Software, Suite 214,
              1210 S. Bascom Ave., San Jose, CA 92128, (408) 293-0800</para>
          </listitem>
        </itemizedlist>
        
        <para>On Usenet the newsgroups <ulink
            URL="news:comp.periphs.scsi">comp.periphs.scsi</ulink> and <ulink
            URL="news:comp.periphs">comp.periphs</ulink> are noteworthy places
          to look for more info.  You can also find the SCSI-Faq there, which
          is posted periodically.</para>
        
        <para>Most major SCSI device and host adapter suppliers operate ftp
          sites and/or BBS systems.  They may be valuable sources of
          information about the devices you own.</para>
    -->
          </sect3>
        </sect2>
        
        <sect2 id="hw-storage-controllers">
          <title>* Contr&ocirc;leurs de disques/bandes</title>
          <sect3>
            <title>* SCSI</title>
                 <para></para>
          </sect3>
          <sect3>
            <title>* IDE</title>
                 <para></para>
          </sect3>
          <sect3>
            <title>* Disquettes</title>
                 <para></para>
          </sect3>
        </sect2>
        
        <sect2>
          <title>*** Disques durs</title>
          
          <sect3>
            <title>*** Disques durs SCSI</title>
            &sgml.todo;
<!--
        <para><emphasis>Contributed by &a.asami;.  17 February
            1998.</emphasis></para>
            
        <para>As mentioned in the <link linkend="scsi">SCSI</link> section,
          virtually all SCSI hard drives sold today are SCSI-2 compliant and
          thus will work fine as long as you connect them to a supported SCSI
          host adapter.  Most problems people encounter are either due to
          badly designed cabling (cable too long, star topology, etc.),
          insufficient termination, or defective parts. Please refer to the
          <link linkend="scsi">SCSI</link> section first if your SCSI hard
          drive is not working.  However, there are a couple of things you may
          want to take into account before you purchase SCSI hard drives for
          your system.</para>
        
        <sect4>
          <title>Rotational speed</title>
          
          <para>Rotational speeds of SCSI drives sold today range from around
            4,500RPM to 10,000RPM.  Most of them are either 5,400RPM or
            7,200RPM.  Even though the 7,200RPM drives can generally transfer
            data faster, they run considerably hotter than their 5,400RPM
            counterparts.  A large fraction of today's disk drive malfunctions
            are heat-related.  If you do not have very good cooling in your PC
            case, you may want to stick with 5,400RPM or slower drives.</para>
              
          <para>Note that newer drives, with higher areal recording densities,
            can deliver much more bits per rotation than older ones.  Today's
            top-of-line 5,400RPM drives can sustain a throughput comparable to
            7,200RPM drives of one or two model generations ago.  The number
            to find on the spec sheet for bandwidth is &ldquo;internal data
            (or transfer) rate&rdquo;.  It is usually in megabits/sec so
            divide it by 8 and you'll get the rough approximation of how much
            megabytes/sec you can get out of the drive.</para>
              
          <para>(If you are a speed maniac and want a 10,000RPM drive for your
            cute little peecee, be my guest; however, those drives become
            extremely hot.  Don't even think about it if you don't have a fan
            blowing air <emphasis>directly at</emphasis> the drive or a
            properly ventilated disk enclosure.)</para>
              
          <para>Obviously, the latest 10,000RPM drives and 7,200RPM drives can
            deliver more data than the latest 5,400RPM drives, so if absolute
            bandwidth is the necessity for your applications, you have little
            choice but to get the faster drives.  Also, if you need low
            latency, faster drives are better; not only do they usually have
            lower average seek times, but also the rotational delay is one
            place where slow-spinning drives can never beat a faster one.
            (The average rotational latency is half the time it takes to
            rotate the drive once; thus, it's 3 milliseconds for 10,000RPM
            drives, 4.2ms for 7,200RPM drives and 5.6ms for 5,400RPM drives.)
            Latency is seek time plus rotational delay. Make sure you
            understand whether you need low latency or more accesses per
            second, though; in the latter case (e.g., news servers), it may
            not be optimal to purchase one big fast drive.  You can achieve
            similar or even better results by using the ccd (concatenated
            disk) driver to create a striped disk array out of multiple slower
            drives for comparable overall cost.</para>
              
          <para>Make sure you have adequate air flow around the drive,
            especially if you are going to use a fast-spinning drive.  You
            generally need at least 1/2" (1.25cm) of spacing above and below a
            drive.  Understand how the air flows through your PC case.  Most
            cases have the power supply suck the air out of the back.  See
            where the air flows in, and put the drive where it will have the
            largest volume of cool air flowing around it. You may need to seal
            some unwanted holes or add a new fan for effective cooling.</para>
              
          <para>Another consideration is noise.  Many 7,200 or faster drives
            generate a high-pitched whine which is quite unpleasant to most
            people.  That, plus the extra fans often required for cooling, may
            make 7,200 or faster drives unsuitable for some office and home
            environments.</para>
        </sect4>
        
        <sect4>
          <title>Form factor</title>
          
          <para>Most SCSI drives sold today are of 3.5" form factor.  They
            come in two different heights; 1.6" (&ldquo;half-height&rdquo;) or
            1" (&ldquo;low-profile&rdquo;).  The half-height drive is the same
            height as a CD-ROM drive.  However, don't forget the spacing rule
            mentioned in the previous section.  If you have three standard
            3.5" drive bays, you will not be able to put three half-height
            drives in there (without frying them, that is).</para>
        </sect4>
        
        <sect4>
          <title>Interface</title>
          
          <para>The majority of SCSI hard drives sold today are Ultra or
            Ultra-wide SCSI.  The maximum bandwidth of Ultra SCSI is 20MB/sec,
            and Ultra-wide SCSI is 40MB/sec.  There is no difference in max
            cable length between Ultra and Ultra-wide; however, the more
            devices you have on the same bus, the sooner you will start having
            bus integrity problems.  Unless you have a well-designed disk
            enclosure, it is not easy to make more than 5 or 6 Ultra SCSI
            drives work on a single bus.</para>
              
          <para>On the other hand, if you need to connect many drives, going
            for Fast-wide SCSI may not be a bad idea.  That will have the same
            max bandwidth as Ultra (narrow) SCSI, while electronically it's
            much easier to get it &ldquo;right&rdquo;.  My advice would be: if
            you want to connect many disks, get wide SCSI drives; they usually
            cost a little more but it may save you down the road.  (Besides,
            if you can't afford the cost difference, you shouldn't be building
            a disk array.)</para>
              
          <para>There are two variant of wide SCSI drives; 68-pin and 80-pin
            SCA (Single Connector Attach).  The SCA drives don't have a
            separate 4-pin power connector, and also read the SCSI ID settings
            through the 80-pin connector.  If you are really serious about
            building a large storage system, get SCA drives and a good SCA
            enclosure (dual power supply with at least one extra fan).  They
            are more electronically sound than 68-pin counterparts because
            there is no &ldquo;stub&rdquo; of the SCSI bus inside the disk
            canister as in arrays built from 68-pin drives.  They are easier
            to install too (you just need to screw the drive in the canister,
            instead of trying to squeeze in your fingers in a tight place to
            hook up all the little cables (like the SCSI ID and disk activity
            LED lines).</para>
        </sect4>
   -->
          </sect3>
          
          <sect3>
            <title>* Disques durs IDE</title>
            <para></para>
          </sect3>
        </sect2>
        
        <sect2>
          <title>*** Contr&ocirc;leurs de bande</title>
<!--
      <para><emphasis>Contributed by &a.jmb;.  2 July
          1996.</emphasis></para>
-->
          <sect3>
            <title>*** Commandes g&eacute;n&eacute;rales d'acc&egrave;s aux bandes</title>
      
	      &sgml.todo
  <!--
        <para>&man.mt.1; provides generic access to the tape drives.  Some of
          the more common commands are <command>rewind</command>,
          <command>erase</command>, and <command>status</command>.  See the
            &man.mt.1; manual page for a detailed description.</para>
            &sgml.todo;
   -->
          </sect3>
          <sect3>
            <title>*** Interfaces et contr&ocirc;leurs</title>
	  &sgml.todo
   <!--
        <para>There are several different interfaces that support tape drives.
          The interfaces are SCSI, IDE, Floppy and Parallel Port. A wide
          variety of tape drives are available for these interfaces.
          Controllers are discussed in <link
            linkend="hw-storage-controllers">Disk/tape
            controllers</link>.</para>
    -->
          </sect3>
          <sect3>
            <title>*** Lecteurs SCSI</title>
	       &sgml.todo
    <!--
        
        <para>The &man.st.4; driver provides support for 8mm (Exabyte), 4mm
          (DAT: Digital Audio Tape), QIC (Quarter-Inch Cartridge), DLT
          (Digital Linear Tape), QIC Minicartridge and 9-track (remember the
          big reels that you see spinning in Hollywood computer rooms) tape
          drives.  See the &man.st.4; manual page for a detailed
          description.</para>
        
        <para>The drives listed below are currently being used by members of
          the FreeBSD community.  They are not the only drives that will work
          with FreeBSD.  They just happen to be the ones that we use.</para>
        
        <sect4>
          <title>4mm (DAT: Digital Audio Tape)</title>
          
          <para><link linkend="hw-storage-python">Archive Python</link></para>
          
          <para><link linkend="hw-storage-hp1533a">HP C1533A</link></para>
              
          <para><link linkend="hw-storage-hp1534a">HP C1534A</link></para>
              
          <para><link linkend="hw-storage-hp35450a">HP 35450A</link></para>
              
          <para><link linkend="hw-storage-hp35470a">HP 35470A</link></para>
              
          <para><link linkend="hw-storage-hp35480a">HP 35480A</link></para>
              
          <para><link linkend="hw-storage-sdt5000">SDT-5000</link></para>
              
          <para><link linkend="hw-storage-wangtek6200">Wangtek
              6200</link></para>
        </sect4>
        
        <sect4>
          <title>8mm (Exabyte)</title>
          
          <para><link linkend="hw-storage-exb8200">EXB-8200</link></para>
          
          <para><link linkend="hw-storage-exb8500">EXB-8500</link></para>
          
          <para><link linkend="hw-storage-exb8505">EXB-8505</link></para>
        </sect4>
        
        <sect4>
          <title>QIC (Quarter-Inch Cartridge)</title>
          
          <para><link linkend="hw-storage-anaconda">Archive Ananconda
              2750</link></para>
              
          <para><link linkend="hw-storage-viper60">Archive Viper
              60</link></para>
              
          <para><link linkend="hw-storage-viper150">Archive Viper
              150</link></para>
              
          <para><link linkend="hw-storage-viper2525">Archive Viper
              2525</link></para>
              
          <para><link linkend="hw-storage-tandberg3600">Tandberg TDC
              3600</link></para>
              
          <para><link linkend="hw-storage-tandberg3620">Tandberg TDC
              3620</link></para>
              
          <para><link linkend="hw-storage-tandberg4222">Tandberg TDC
              4222</link></para>
              
          <para><link linkend="hw-storage-wangtek5525es">Wangtek
              5525ES</link></para>
        </sect4>
        
        <sect4>
          <title>DLT (Digital Linear Tape)</title>
          
          <para><link linkend="hw-storage-dectz87">Digital TZ87</link></para>
        </sect4>
        
        <sect4>
          <title>Mini-Cartridge</title>
          
          <para><link linkend="hw-storage-ctms3200">Conner CTMS
              3200</link></para>
              
          <para><link linkend="hw-storage-exb2501">Exabyte 2501</link></para>
        </sect4>
        
        <sect4>
          <title>Autoloaders/Changers</title>
          
          <para><link linkend="hw-storage-hp1553a">Hewlett-Packard HP C1553A
              Autoloading DDS2</link></para>
        </sect4>
        
  -->
          </sect3>
          <sect3>
            <title>* Lecteurs IDE</title>
            <para></para>
          </sect3>
          <sect3>
            <title>* Lecteurs sur contr&ocirc;leur de disquette</title>
            &sgml.todo
<!--
 o&ugrave; est pass&eacute; le source ?????
-->
             

        
          </sect3>
          <sect3>
            <title>* Lecteurs sur port parall&egrave;le</title>
            <para></para>
          </sect3>
          <sect3>
            <title>*** Informations d&eacute;taill&eacute;es</title>
	       &sgml.todo
    <!--
        <sect4 id="hw-storage-anaconda">
          <title>Archive Anaconda 2750</title>
          
          <para>The boot message identifier for this drive is <literal>ARCHIVE
              ANCDA 2750 28077 -003 type 1 removable SCSI 2</literal></para>
          
          <para>This is a QIC tape drive.</para>
          
          <para>Native capacity is 1.35GB when using QIC-1350 tapes.  This
            drive will read and write QIC-150 (DC6150), QIC-250 (DC6250), and
            QIC-525 (DC6525) tapes as well.</para>
              
          <para>Data transfer rate is 350kB/s using
                &man.dump.8;.  Rates of 530kB/s have been reported when using
            <link linkend="backups-programs-amanda">Amanda</link></para>

          <para>Production of this drive has been discontinued.</para>
          
          <para>The SCSI bus connector on this tape drive is reversed from
            that on most other SCSI devices.  Make sure that you have enough
            SCSI cable to twist the cable one-half turn before and after the
            Archive Anaconda tape drive, or turn your other SCSI devices
            upside-down.</para>
          
          <para>Two kernel code changes are required to use this drive. This
            drive will not work as delivered.</para>
              
          <para>If you have a SCSI-2 controller, short jumper 6. Otherwise,
            the drive behaves are a SCSI-1 device.  When operating as a SCSI-1
            device, this drive, &ldquo;locks&rdquo; the SCSI bus during some
            tape operations, including: fsf, rewind, and rewoffl.</para>
              
          <para>If you are using the NCR SCSI controllers, patch the file
            <filename>/usr/src/sys/pci/ncr.c</filename> (as shown below).
            Build and install a new kernel.</para>
              
          <programlisting>
*** 4831,4835 ****
                };
        
!               if (np-&gt;latetime&gt;4) {
                        /*
                        **      Although we tried to wake it up,
Remplacer les '*' par des '-'
*** 4831,4836 ****
                };
 
!               if (np-&gt;latetime&gt;1200) {
                        /*
                        **      Although we tried to wake it up,</programlisting>
              
          <para>Reported by: &a.jmb;</para>
        </sect4>
        
        <sect4 id="hw-storage-python">
          <title>Archive Python</title>
          
          <para>The boot message identifier for this drive is <literal>ARCHIVE
              Python 28454-XXX4ASB</literal> <literal>type 1 removable SCSI
              2</literal> <literal>density code 0x8c, 512-byte
              blocks</literal></para>
              
          <para>This is a DDS-1 tape drive.</para>
              
          <para>Native capacity is 2.5GB on 90m tapes.</para>
          
          <para>Data transfer rate is XXX.</para>
          
          <para>This drive was repackaged by Sun Microsystems as model
            411.</para>
              
          <para>Reported by: Bob Bishop <email>rb@gid.co.uk</email></para>
        </sect4>
        
        <sect4 id="hw-storage-viper60">
          <title>Archive Viper 60</title>
          
          <para>The boot message identifier for this drive is <literal>ARCHIVE
              VIPER 60 21116 -007</literal> <literal>type 1 removable SCSI
              1</literal></para>
              
          <para>This is a QIC tape drive.</para>
          
          <para>Native capacity is 60MB.</para>
          
          <para>Data transfer rate is XXX.</para>
          
          <para>Production of this drive has been discontinued.</para>
          
          <para>Reported by: Philippe Regnauld
            <email>regnauld@hsc.fr</email></para>
        </sect4>
        
        <sect4 id="hw-storage-viper150">
          <title>Archive Viper 150</title>
          
          <para>The boot message identifier for this drive is <literal>ARCHIVE
              VIPER 150 21531 -004</literal> <literal>Archive Viper 150 is a
              known rogue</literal> <literal>type 1 removable SCSI
              1</literal>.  A multitude of firmware revisions exist for this
            drive.  Your drive may report different numbers (e.g
            <literal>21247 -005</literal>.</para>
              
          <para>This is a QIC tape drive.</para>
          
          <para>Native capacity is 150/250MB.  Both 150MB (DC6150) and 250MB
            (DC6250) tapes have the recording format.  The 250MB tapes are
            approximately 67% longer than the 150MB tapes.  This drive can
            read 120MB tapes as well.  It can not write 120MB tapes.</para>
              
          <para>Data transfer rate is 100kB/s</para>
          
          <para>This drive reads and writes DC6150 (150MB) and DC6250 (250MB)
            tapes.</para>
              
          <para>This drives quirks are known and pre-compiled into the scsi
            tape device driver (&man.st.4;).</para>
              
          <para>Under FreeBSD 2.2-current, use <command>mt blocksize
              512</command> to set the blocksize.  (The particular drive had
            firmware revision 21247 -005.  Other firmware revisions may behave
            differently) Previous versions of FreeBSD did not have this
            problem.</para>
              
          <para>Production of this drive has been discontinued.</para>
          
          <para>Reported by: Pedro A M Vazquez
            <email>vazquez@IQM.Unicamp.BR</email></para>
          
          <para>Mike Smith
            <email>msmith@atrad.adelaide.edu.au</email></para>
        </sect4>
            
        <sect4 id="hw-storage-viper2525">
          <title>Archive Viper 2525</title>
          
          <para>The boot message identifier for this drive is <literal>ARCHIVE
              VIPER 2525 25462 -011</literal> <literal>type 1 removable SCSI
              1</literal></para>
          
          <para>This is a QIC tape drive.</para>
          
          <para>Native capacity is 525MB.</para>
          
          <para>Data transfer rate is 180kB/s at 90 inches/sec.</para>
          
          <para>The drive reads QIC-525, QIC-150, QIC-120 and QIC-24 tapes.
            Writes QIC-525, QIC-150, and QIC-120.</para>
              
          <para>Firmware revisions prior to <literal>25462 -011</literal> are
            bug ridden and will not function properly.</para>
          
          <para>Production of this drive has been discontinued.</para>
        </sect4>
        
        <sect4 id="hw-storage-conner420r">
          <title>Conner 420R</title>
          
          <para>The boot message identifier for this drive is <literal>Conner
              tape</literal>.</para>
              
          <para>This is a floppy controller, minicartridge tape drive.</para>
              
          <para>Native capacity is XXXX</para>
          
          <para>Data transfer rate is XXX</para>
          
          <para>The drive uses QIC-80 tape cartridges.</para>
          
          <para>Reported by: Mark Hannon
            <email>mark@seeware.DIALix.oz.au</email></para>
        </sect4>
        
        <sect4 id="hw-storage-ctms3200">
          <title>Conner CTMS 3200</title>
              
          <para>The boot message identifier for this drive is <literal>CONNER
              CTMS 3200 7.00</literal> <literal>type 1 removable SCSI
              2</literal>.</para>
              
          <para>This is a minicartridge tape drive.</para>
          
          <para>Native capacity is XXXX</para>
          
          <para>Data transfer rate is XXX</para>
          
          <para>The drive uses QIC-3080 tape cartridges.</para>
          
          <para>Reported by: Thomas S. Traylor
            <email>tst@titan.cs.mci.com</email></para>
        </sect4>
        
        <sect4 id="hw-storage-dectz87">
          <title><ulink
              URL="http://www.digital.com/info/Customer-Update/931206004.txt.html">DEC TZ87</ulink></title>
          
          <para>The boot message identifier for this drive is <literal>DEC
              TZ87 (C) DEC 9206</literal> <literal>type 1 removable SCSI
              2</literal> <literal>density code 0x19</literal></para>
              
          <para>This is a DLT tape drive.</para>
          
          <para>Native capacity is 10GB.</para>
          
          <para>This drive supports hardware data compression.</para>
          
          <para>Data transfer rate is 1.2MB/s.</para>
          
          <para>This drive is identical to the Quantum DLT2000.  The drive
            firmware can be set to emulate several well-known drives,
            including an Exabyte 8mm drive.</para>
              
          <para>Reported by: &a.wilko;</para>
        </sect4>
        
        <sect4 id="hw-storage-exb2501">
          <title><ulink
              URL="http://www.Exabyte.COM:80/Products/Minicartridge/2501/Rfeatures.html">Exabyte EXB-2501</ulink></title>
              
          <para>The boot message identifier for this drive is <literal>EXABYTE
              EXB-2501</literal></para>
              
          <para>This is a mini-cartridge tape drive.</para>
              
          <para>Native capacity is 1GB when using MC3000XL
            minicartridges.</para>
              
          <para>Data transfer rate is XXX</para>
          
          <para>This drive can read and write DC2300 (550MB), DC2750 (750MB),
            MC3000 (750MB), and MC3000XL (1GB) minicartridges.</para>
              
          <para>WARNING: This drive does not meet the SCSI-2 specifications.
            The drive locks up completely in response to a SCSI MODE_SELECT
            command unless there is a formatted tape in the drive.  Before
            using this drive, set the tape blocksize with</para>
              
          <screen>&prompt.root; <userinput>mt -f /dev/st0ctl.0 blocksize 1024</userinput></screen>
              
          <para>Before using a minicartridge for the first time, the
            minicartridge must be formated.  FreeBSD 2.1.0-RELEASE and
            earlier:</para>
              
          <screen>&prompt.root; <userinput>/sbin/scsi -f /dev/rst0.ctl -s 600 -c "4 0 0 0 0 0"</userinput></screen>
              
          <para>(Alternatively, fetch a copy of the
            <command>scsiformat</command> shell script from FreeBSD
            2.1.5/2.2.) FreeBSD 2.1.5 and later:</para>
              
          <screen>&prompt.root; <userinput>/sbin/scsiformat -q -w /dev/rst0.ctl</userinput></screen>
              
          <para>Right now, this drive cannot really be recommended for
            FreeBSD.</para>
              
          <para>Reported by: Bob Beaulieu
            <email>ez@eztravel.com</email></para>
        </sect4>
        
        <sect4 id="hw-storage-exb8200">
          <title>Exabyte EXB-8200</title>
          
          <para>The boot message identifier for this drive is <literal>EXABYTE
              EXB-8200 252X</literal> <literal>type 1 removable SCSI
              1</literal></para>
              
          <para>This is an 8mm tape drive.</para>
          
          <para>Native capacity is 2.3GB.</para>
          
          <para>Data transfer rate is 270kB/s.</para>
          
          <para>This drive is fairly slow in responding to the SCSI bus during
            boot.  A custom kernel may be required (set SCSI_DELAY to 10
            seconds).</para>
          
          <para>There are a large number of firmware configurations for this
            drive, some have been customized to a particular vendor's
            hardware.  The firmware can be changed via EPROM
            replacement.</para>
              
          <para>Production of this drive has been discontinued.</para>
          
          <para>Reported by: Mike Smith
            <email>msmith@atrad.adelaide.edu.au</email></para>
        </sect4>
        
        <sect4 id="hw-storage-exb8500">
          <title>Exabyte EXB-8500</title>
          
          <para>The boot message identifier for this drive is <literal>EXABYTE
              EXB-8500-85Qanx0 0415</literal> <literal>type 1 removable SCSI
              2</literal></para>
              
          <para>This is an 8mm tape drive.</para>
          
          <para>Native capacity is 5GB.</para>
          
          <para>Data transfer rate is 300kB/s.</para>
              
          <para>Reported by: Greg Lehey <email>grog@lemis.de</email></para>
        </sect4>
        
        <sect4 id="hw-storage-exb8505">
          <title><ulink
              URL="http://www.Exabyte.COM:80/Products/8mm/8505XL/Rfeatures.html">Exabyte EXB-8505</ulink></title>
              
              <para>The boot message identifier for this drive is
            <literal>EXABYTE EXB-85058SQANXR1 05B0</literal> <literal>type 1
              removable SCSI 2</literal></para>
              
          <para>This is an 8mm tape drive which supports compression, and is
            upward compatible with the EXB-5200 and EXB-8500.</para>
          
          <para>Native capacity is 5GB.</para>
          
          <para>The drive supports hardware data compression.</para>
          
          <para>Data transfer rate is 300kB/s.</para>
          
          <para>Reported by: Glen Foster
            <email>gfoster@gfoster.com</email></para>
        </sect4>
        
        <sect4 id="hw-storage-hp1533a">
          <title>Hewlett-Packard HP C1533A</title>
          
          <para>The boot message identifier for this drive is <literal>HP
              C1533A 9503</literal> <literal>type 1 removable SCSI
              2</literal>.</para>
              
          <para>This is a DDS-2 tape drive.  DDS-2 means hardware data
            compression and narrower tracks for increased data
            capacity.</para>
          
          <para>Native capacity is 4GB when using 120m tapes.  This drive
            supports hardware data compression.</para>
              
          <para>Data transfer rate is 510kB/s.</para>
          
          <para>This drive is used in Hewlett-Packard's SureStore 6000eU and
            6000i tape drives and C1533A DDS-2 DAT drive.</para>
          
          <para>The drive has a block of 8 dip switches.  The proper settings
            for FreeBSD are: 1 ON; 2 ON; 3 OFF; 4 ON; 5 ON; 6 ON; 7 ON; 8
            ON.</para>
          
          <informaltable frame="none">
            <tgroup cols="3">
              <thead>
                <row>
                  <entry>switch 1</entry>
                  <entry>switch 2</entry>
                  <entry>Result</entry>
                </row>
              </thead>
              
              <tbody>
                <row>
                  <entry>On</entry>
                  <entry>On</entry>
                  <entry>Compression enabled at power-on, with host
                    control</entry>
                </row>
                
                <row>
                  <entry>On</entry>
                  <entry>Off</entry>
                  <entry>Compression enabled at power-on, no host
                    control</entry>
                </row>
                
                <row>
                  <entry>Off</entry>
                  <entry>On</entry>
                  <entry>Compression disabled at power-on, with host
                    control</entry>
                </row>
                
                <row>
                  <entry>Off</entry>
                  <entry>Off</entry>
                  <entry>Compression disabled at power-on, no host
                    control</entry>
                </row>
              </tbody>
            </tgroup>
          </informaltable>
          
          <para>Switch 3 controls MRS (Media Recognition System).  MRS tapes
            have stripes on the transparent leader.  These identify the tape
            as DDS (Digital Data Storage) grade media.  Tapes that do not have
            the stripes will be treated as write-protected.  Switch 3 OFF
            enables MRS.  Switch 3 ON disables MRS.</para>
              
              <para>See <ulink URL="http://www.hp.com/tape/c_intro.html">HP
              SureStore Tape Products</ulink> and <ulink
              URL="http://www.impediment.com/hp/hp_technical.html">Hewlett-Packard
              Disk and Tape Technical Information</ulink> for more information
            on configuring this drive.</para>
              
          <para><emphasis>Warning:</emphasis> Quality control on these drives
            varies greatly.  One FreeBSD core-team member has returned 2 of
            these drives.  Neither lasted more than 5 months.</para>
              
          <para>Reported by: &a.se;</para>
        </sect4>
        
        <sect4 id="hw-storage-hp1534a">
          <title>Hewlett-Packard HP 1534A</title>
          
          <para>The boot message identifier for this drive is <literal>HP
              HP35470A T503</literal> <literal>type 1 removable SCSI
              2</literal> <literal>Sequential-Access density code 0x13,
              variable blocks</literal>.</para>
              
          <para>This is a DDS-1 tape drive.  DDS-1 is the original DAT tape
            format.</para>
              
          <para>Native capacity is 2GB when using 90m tapes.</para>
              
          <para>Data transfer rate is 183kB/s.</para>
          
          <para>The same mechanism is used in Hewlett-Packard's SureStore
            <ulink URL="http://www.dmo.hp.com/tape/sst2000.htm">2000i</ulink>
            tape drive, C35470A DDS format DAT drive, C1534A DDS format DAT
            drive and HP C1536A DDS format DAT drive.</para>
              
          <para>The HP C1534A DDS format DAT drive has two indicator lights,
            one green and one amber.  The green one indicates tape action:
            slow flash during load, steady when loaded, fast flash during
            read/write operations.  The amber one indicates warnings: slow
            flash when cleaning is required or tape is nearing the end of its
            useful life, steady indicates an hard fault.  (factory service
            required?)</para>
              
          <para>Reported by Gary Crutcher
            <email>gcrutchr@nightflight.com</email></para>
        </sect4>
        
        <sect4 id="hw-storage-hp1553a">
          <title>Hewlett-Packard HP C1553A Autoloading DDS2</title>
          
          <para>The boot message identifier for this drive is "".</para>
          
          <para>This is a DDS-2 tape drive with a tape changer.  DDS-2 means
            hardware data compression and narrower tracks for increased data
            capacity.</para>
              
          <para>Native capacity is 24GB when using 120m tapes.  This drive
            supports hardware data compression.</para>
              
          <para>Data transfer rate is 510kB/s (native).</para>
              
          <para>This drive is used in Hewlett-Packard's SureStore <ulink
              URL="http://www.dmo.hp.com/tape/sst12000.htm">12000e</ulink>
            tape drive.</para>
              
          <para>The drive has two selectors on the rear panel.  The selector
            closer to the fan is SCSI id.  The other selector should be set to
            7.</para>
              
          <para>There are four internal switches.  These should be set: 1 ON;
            2 ON; 3 ON; 4 OFF.</para>
              
          <para>At present the kernel drivers do not automatically change
            tapes at the end of a volume.  This shell script can be used to
            change tapes:</para>
              
          <programlisting>
#!/bin/sh
PATH="/sbin:/usr/sbin:/bin:/usr/bin"; export PATH

usage()
{
        echo "Usage: dds_changer [123456ne] raw-device-name
        echo "1..6 = Select cartridge"
        echo "next cartridge"
        echo "eject magazine"
        exit 2
}

if [ $# -ne 2 ] ; then
        usage
fi

cdb3=0
cdb4=0
cdb5=0

case $1 in
        [123456])
                cdb3=$1
                cdb4=1
                ;;
        n)
                ;;
        e)
                cdb5=0x80
                ;;
        ?)
                usage
                ;;
esac

scsi -f $2 -s 100 -c "1b 0 0 $cdb3 $cdb4 $cdb5"</programlisting>
        </sect4>
        
        <sect4 id="hw-storage-hp35450a">
          <title>Hewlett-Packard HP 35450A</title>
          
          <para>The boot message identifier for this drive is <literal>HP
              HP35450A -A C620</literal> <literal>type 1 removable SCSI
              2</literal> <literal>Sequential-Access density code
              0x13</literal></para>
              
          <para>This is a DDS-1 tape drive.  DDS-1 is the original DAT tape
            format.</para>
              
          <para>Native capacity is 1.2GB.</para>
              
          <para>Data transfer rate is 160kB/s.</para>
          
          <para>Reported by: mark thompson
            <email>mark.a.thompson@pobox.com</email></para>
        </sect4>
        
        <sect4 id="hw-storage-hp35470a">
          <title>Hewlett-Packard HP 35470A</title>
          
          <para>The boot message identifier for this drive is <literal>HP
              HP35470A 9 09</literal> <literal>type 1 removable SCSI
              2</literal></para>
              
          <para>This is a DDS-1 tape drive.  DDS-1 is the original DAT tape
            format.</para>
              
          <para>Native capacity is 2GB when using 90m tapes.</para>
              
          <para>Data transfer rate is 183kB/s.</para>
          
          <para>The same mechanism is used in Hewlett-Packard's SureStore
            <ulink URL="http://www.dmo.hp.com/tape/sst2000.htm">2000i</ulink>
            tape drive, C35470A DDS format DAT drive, C1534A DDS format DAT
            drive, and HP C1536A DDS format DAT drive.</para>
              
          <para><emphasis>Warning:</emphasis> Quality control on these drives
            varies greatly.  One FreeBSD core-team member has returned 5 of
            these drives.  None lasted more than 9 months.</para>
              
          <para>Reported by: David Dawes
            <email>dawes@rf900.physics.usyd.edu.au</email> (9 09)</para>
              
        </sect4>
        
        <sect4 id="hw-storage-hp35480a">
          <title>Hewlett-Packard HP 35480A</title>
          
          <para>The boot message identifier for this drive is <literal>HP
              HP35480A 1009</literal> <literal>type 1 removable SCSI
              2</literal> <literal>Sequential-Access density code
              0x13</literal>.</para>
              
          <para>This is a DDS-DC tape drive.  DDS-DC is DDS-1 with hardware
            data compression.  DDS-1 is the original DAT tape format.</para>
              
          <para>Native capacity is 2GB when using 90m tapes.  It cannot handle
            120m tapes.  This drive supports hardware data compression.
            Please refer to the section on <link
              linkend="hw-storage-hp1533a">HP C1533A</link> for the proper
            switch settings.</para>
              
          <para>Data transfer rate is 183kB/s.</para>
          
          <para>This drive is used in Hewlett-Packard's SureStore <ulink
              URL="http://www.dmo.hp.com/tape/sst5000.htm">5000eU</ulink> and
            <ulink URL="http://www.dmo.hp.com/tape/sst5000.htm">5000i</ulink>
            tape drives and C35480A DDS format DAT drive..</para>
              
          <para>This drive will occasionally hang during a tape eject
            operation (<command>mt offline</command>). Pressing the front
            panel button will eject the tape and bring the tape drive back to
            life.</para>
              
          <para>WARNING: HP 35480-03110 only.  On at least two occasions this
            tape drive when used with FreeBSD 2.1.0, an IBM Server 320 and an
            2940W SCSI controller resulted in all SCSI disk partitions being
            lost.  The problem has not be analyzed or resolved at this
            time.</para>
        </sect4>
        
        <sect4 id="hw-storage-sdt5000">
          <title><ulink
              URL="http://www.sel.sony.com/SEL/ccpg/storage/tape/t5000.html">Sony SDT-5000</ulink></title>
              
          <para>There are at least two significantly different models: one is
            a DDS-1 and the other DDS-2.  The DDS-1 version is
            <literal>SDT-5000 3.02</literal>.  The DDS-2 version is
            <literal>SONY SDT-5000 327M</literal>. The DDS-2 version has a 1MB
            cache.  This cache is able to keep the tape streaming in almost
            any circumstances.</para>
              
          <para>The boot message identifier for this drive is <literal>SONY
              SDT-5000 3.02</literal> <literal>type 1 removable SCSI
              2</literal> <literal>Sequential-Access density code
              0x13</literal></para>
              
          <para>Native capacity is 4GB when using 120m tapes.  This drive
            supports hardware data compression.</para>
              
          <para>Data transfer rate is depends upon the model or the drive. The
            rate is 630kB/s for the <literal>SONY SDT-5000 327M</literal>
            while compressing the data.  For the <literal>SONY SDT-5000
              3.02</literal>, the data transfer rate is 225kB/s.</para>
              
          <para>In order to get this drive to stream, set the blocksize to 512
            bytes (<command>mt blocksize 512</command>) reported by Kenneth
            Merry ken@ulc199.residence.gatech.edu</para>
              
          <para><literal>SONY SDT-5000 327M</literal> information reported by
            Charles Henrich henrich@msu.edu</para>
              
          <para>Reported by: &a.jmz;</para>
        </sect4>
        
        <sect4 id="hw-storage-tandberg3600">
          <title>Tandberg TDC 3600</title>
          
          <para>The boot message identifier for this drive is
            <literal>TANDBERG TDC 3600 =08:</literal> <literal>type 1
              removable SCSI 2</literal></para>
              
          <para>This is a QIC tape drive.</para>
          
          <para>Native capacity is 150/250MB.</para>
          
          <para>This drive has quirks which are known and work around code is
            present in the scsi tape device driver (&man.st.4;).
            Upgrading the firmware to XXX version will fix the quirks and
            provide SCSI 2 capabilities.</para>
              
          <para>Data transfer rate is 80kB/s.</para>
          
          <para>IBM and Emerald units will not work.  Replacing the firmware
            EPROM of these units will solve the problem.</para>
              
          <para>Reported by: Michael Smith
            <email>msmith@atrad.adelaide.edu.au</email></para>
        </sect4>
        
        <sect4 id="hw-storage-tandberg3620">
          <title>Tandberg TDC 3620</title>
          
          <para>This is very similar to the <link
              linkend="hw-storage-tandberg3600">Tandberg TDC 3600</link>
            drive.</para>
              
          <para>Reported by: &a.joerg;</para>
        </sect4>
        
        <sect4 id="hw-storage-tandberg4222">
          <title>Tandberg TDC 4222</title>
          
          <para>The boot message identifier for this drive is
            <literal>TANDBERG TDC 4222 =07</literal> <literal>type 1 removable
              SCSI 2</literal></para>
              
          <para>This is a QIC tape drive.</para>
          
          <para>Native capacity is 2.5GB.  The drive will read all cartridges
            from the 60 MB (DC600A) upwards, and write 150 MB (DC6150)
            upwards.  Hardware compression is optionally supported for the 2.5
            GB cartridges.</para>
              
          <para>This drives quirks are known and pre-compiled into the scsi
            tape device driver (&man.st.4;) beginning with FreeBSD
            2.2-current.  For previous versions of FreeBSD, use
            <command>mt</command> to read one block from the tape, rewind the
            tape, and then execute the backup program (<command>mt fsr 1; mt
              rewind; dump ...</command>)</para>
              
          <para>Data transfer rate is 600kB/s (vendor claim with compression),
            350 KB/s can even be reached in start/stop mode. The rate
            decreases for smaller cartridges.</para>
          
          <para>Reported by: &a.joerg;</para>
        </sect4>
        
        <sect4 id="hw-storage-wangtek5525es">
          <title>Wangtek 5525ES</title>
          
          <para>The boot message identifier for this drive is <literal>WANGTEK
              5525ES SCSI REV7 3R1</literal> <literal>type 1 removable SCSI
              1</literal> <literal>density code 0x11, 1024-byte
              blocks</literal></para>
              
          <para>This is a QIC tape drive.</para>
          
          <para>Native capacity is 525MB.</para>
          
          <para>Data transfer rate is 180kB/s.</para>
          
          <para>The drive reads 60, 120, 150, and 525MB tapes.  The drive will
            not write 60MB (DC600 cartridge) tapes.  In order to overwrite 120
            and 150 tapes reliably, first erase (<command>mt erase</command>)
            the tape.  120 and 150 tapes used a wider track (fewer tracks per
            tape) than 525MB tapes. The &ldquo;extra&rdquo; width of the
            previous tracks is not overwritten, as a result the new data lies
            in a band surrounded on both sides by the previous data unless the
            tape have been erased.</para>
              
          <para>This drives quirks are known and pre-compiled into the scsi
            tape device driver (&man.st.4;).</para>
          
          <para>Other firmware revisions that are known to work are:
            M75D</para>
          
          <para>Reported by: Marc van Kempen <email>marc@bowtie.nl</email>
            <literal>REV73R1</literal> Andrew Gordon
            <email>Andrew.Gordon@net-tel.co.uk</email>
            <literal>M75D</literal></para>
        </sect4>
        
        <sect4 id="hw-storage-wangtek6200">
          <title>Wangtek 6200</title>
          
          <para>The boot message identifier for this drive is <literal>WANGTEK
              6200-HS 4B18</literal> <literal>type 1 removable SCSI
              2</literal> <literal>Sequential-Access density code
              0x13</literal></para>
          
          <para>This is a DDS-1 tape drive.</para>
          
          <para>Native capacity is 2GB using 90m tapes.</para>
          
          <para>Data transfer rate is 150kB/s.</para>
          
          <para>Reported by: Tony Kimball <email>alk@Think.COM</email></para>
        </sect4>
  -->
          </sect3>
          <sect3>
            <title>* Lecteurs posant probl&egrave;me</title>
            <para></para>
          </sect3>
        </sect2>
        
        <sect2>
          <title>*** Contr&ocirc;leurs de CD-ROMs</title>

	       &sgml.todo
    <!--
      
      <para><emphasis>Contribution de &a.obrien;.  23 Novembre
              1997.</emphasis></para>
      
      <para>Comme mentionn&eacute; dans 
      <link linkend="hw-jordans-picks-cdrom">Jordan's Picks</link> 
      Generally speaking those in <emphasis>The FreeBSD
      Project</emphasis> prefer SCSI CDROM drives over IDE CDROM drives.
        However not all SCSI CDROM drives are equal.  Some feel the quality of
        some SCSI CDROM drives have been deteriorating to that of IDE CDROM
        drives.  Toshiba used to be the favored stand-by, but many on the SCSI
        mailing list have found displeasure with the 12x speed XM-5701TA as
        its volume (when playing audio CDROMs) is not controllable by the
        various audio player software.</para>
          
      <para>Another area where SCSI CDROM manufacturers are cutting corners is
        adhearance to the <link linkend="scsi-further-reading">SCSI
          specification</link>. Many SCSI CDROMs will respond to <link
          linkend="scsi-rogue-devices">multiple LUNs</link> for its target
        address.  Known violators include the 6x Teac CD-56S 1.0D.</para>
            <para></para>

  -->
        </sect2>
        
        <sect2>
          <title>* Autres</title>
          <para></para>
        </sect2>
        
        <sect2>
          <title>* Ajouter et reconfigurer des disques</title>
          <para></para>
        </sect2>
        
      </sect1>
      
      <sect1 id="hw-other">
        <title>* Autres</title>
        
        <sect2>
          <title>* PCMCIA</title>
          <para></para>
        </sect2>
      </sect1>
    </chapter>
<!--    
     Local Variables:
     mode: sgml
     sgml-declaration: "../chapter.decl"
     sgml-indent-data: t
     sgml-omittag: nil
     sgml-shorttag: nil
     sgml-always-quote-attributes: t 
     sgml-minimize-attributes: max
     sgml-parent-document: ("../handbook.sgml" "part" "chapter")
     End: 
-->