Author: hrs
Date: Fri Apr  5 04:08:03 2013
New Revision: 249128
URL: http://svnweb.freebsd.org/changeset/base/249128

Log:
  Trim old imformation and bump version numbers.
  
  Approved by:  re (implicit)

Deleted:
  releng/8.4/release/doc/de_DE.ISO8859-1/
  releng/8.4/release/doc/fr_FR.ISO8859-1/
  releng/8.4/release/doc/ja_JP.eucJP/
  releng/8.4/release/doc/ru_RU.KOI8-R/
  releng/8.4/release/doc/zh_CN.GB2312/
Modified:
  releng/8.4/release/doc/en_US.ISO8859-1/errata/article.xml
  releng/8.4/release/doc/en_US.ISO8859-1/installation/article.xml
  releng/8.4/release/doc/en_US.ISO8859-1/relnotes/article.xml
  releng/8.4/release/doc/share/examples/Makefile.relnotesng
  releng/8.4/release/doc/share/xml/release.ent

Modified: releng/8.4/release/doc/en_US.ISO8859-1/errata/article.xml
==============================================================================
--- releng/8.4/release/doc/en_US.ISO8859-1/errata/article.xml   Fri Apr  5 
02:15:10 2013        (r249127)
+++ releng/8.4/release/doc/en_US.ISO8859-1/errata/article.xml   Fri Apr  5 
04:08:03 2013        (r249128)
@@ -41,7 +41,7 @@
     <pubdate>$FreeBSD$</pubdate>
 
     <copyright>
-      <year>2011</year>
+      <year>2013</year>
       <holder role="mailto:d...@freebsd.org";>The &os; Documentation 
Project</holder>
     </copyright>
 
@@ -83,7 +83,7 @@
 <![ %release.type.release [
       &release.current;
 ]]>
-      will be maintained until the release of &os; &release.next;.</para>
+      will be maintained until the EoL of &os; &release.branch; branch.</para>
   </abstract>
 
   <sect1 id="intro">
@@ -167,13 +167,6 @@
   <sect1 id="late-news">
     <title>Late-Breaking News and Corrections</title>
 
-    <para>A bug in <application>OpenSSL</application> that could cause
-      it to parse past the end of the message was found at the late
-      stage of &release.bugfix; release process.  The &release.bugfix;
-      includes a fix for this issue by importing relevant parts from
-      the OpenSSL CVS.  This could be triggered by an incorrectly
-      formatted ClientHello SSL/TLS handshake messages.  The details
-      can be found at <ulink
-      url="http://www.openssl.org/news/secadv_20110208.txt";></ulink>.</para>
+    <para>No news.</para>
   </sect1>
 </article>

Modified: releng/8.4/release/doc/en_US.ISO8859-1/installation/article.xml
==============================================================================
--- releng/8.4/release/doc/en_US.ISO8859-1/installation/article.xml     Fri Apr 
 5 02:15:10 2013        (r249127)
+++ releng/8.4/release/doc/en_US.ISO8859-1/installation/article.xml     Fri Apr 
 5 04:08:03 2013        (r249128)
@@ -21,7 +21,7 @@
     <pubdate>$FreeBSD$</pubdate>
 
     <copyright>
-      <year>2012</year>
+      <year>2013</year>
 
       <holder role="mailto:d...@freebsd.org";>The &os; Documentation 
Project</holder>
     </copyright>
@@ -62,8 +62,8 @@
        <ulink url="&url.books.handbook;//makeworld.html"></ulink>.</para>
 
       <para>The branch tag to use for updating the source is
-       <literal>RELENG_8_3</literal> for CVS.  For SVN use
-       <filename>releng/8.3</filename>.</para>
+       <literal>RELENG_8_4</literal> for CVS.  For SVN use
+       <filename>releng/8.4</filename>.</para>
     </sect2>
 
     <sect2>
@@ -73,11 +73,11 @@
        upgrades of &arch.i386; and &arch.amd64; systems running
        earlier FreeBSD releases. Systems running
        <literal>7.[01234]-RELEASE</literal>,
-       <literal>8.[012]-RELEASE</literal>,
-       <literal>8.3-BETA1</literal>, or
-       <literal>8.3-RC[12]</literal> can upgrade as follows:</para>
+       <literal>8.[0123]-RELEASE</literal>,
+       <literal>8.4-BETA1</literal>, or
+       <literal>8.4-RC[12]</literal> can upgrade as follows:</para>
 
-      <screen>&prompt.root; freebsd-update upgrade -r 8.3-RELEASE</screen>
+      <screen>&prompt.root; freebsd-update upgrade -r 8.4-RELEASE</screen>
 
       <para>During this process, FreeBSD Update may ask the user to
        help by merging some configuration files or by confirming that

Modified: releng/8.4/release/doc/en_US.ISO8859-1/relnotes/article.xml
==============================================================================
--- releng/8.4/release/doc/en_US.ISO8859-1/relnotes/article.xml Fri Apr  5 
02:15:10 2013        (r249127)
+++ releng/8.4/release/doc/en_US.ISO8859-1/relnotes/article.xml Fri Apr  5 
04:08:03 2013        (r249128)
@@ -16,7 +16,8 @@
   <pubdate>$FreeBSD$</pubdate>
 
   <copyright>
-    <year>2011</year>
+    <year>2013</year>
+
     <holder role="mailto:d...@freebsd.org";>The &os; Documentation 
Project</holder>
   </copyright>
 
@@ -118,6 +119,7 @@
        advisories available from
        <ulink url="http://security.FreeBSD.org/";></ulink>.</para>
 
+<!--
       <informaltable frame="none" pgwide="1">
        <tgroup cols="3">
          <colspec colwidth="1*"/>
@@ -138,1126 +140,88 @@
                <entry>20&nbsp;September&nbsp;2010</entry>
                <entry><para>Integer overflow in bzip2 
decompression</para></entry>
              </row>
-<!-- XXX: not for 8.2
-             <row>
-               <entry><ulink 
url="http://security.freebsd.org/advisories/FreeBSD-SA-10:09.pseudofs.asc";
-                             >SA-10:09.pseudofs</ulink></entry>
-               <entry>10&nbsp;October&nbsp;2010</entry>
-               <entry><para>Spurious mutex unlock</para></entry>
-             </row>
--->
-             <row>
-               <entry><ulink 
url="http://security.freebsd.org/advisories/FreeBSD-SA-10:10.openssl.asc";
-                             >SA-10:10.openssl</ulink></entry>
-               <entry>29&nbsp;November&nbsp;2010</entry>
-               <entry><para>OpenSSL multiple vulnerabilities</para></entry>
-             </row>
            </tbody>
        </tgroup>
       </informaltable>
+-->
     </sect2>
 
     <sect2 id="kernel">
       <title>Kernel Changes</title>
 
-      <para arch="ia64" revision="209326">The maximum number of pages
-       used for DMA bounce buffer pool has been increased from 256 to
-       1024.</para>
-
-      <para arch="powerpc" revision="209765">The default value of
-       <varname>kern.hz</varname> has been increased from 100 to
-       1000.</para>
-
-      <para arch="powerpc" revision="209767">The SMP kernel now works on
-       MPC7400-based Apple desktop machines such as
-       PowerMac3,3.</para>
-
-      <para arch="powerpc" revision="211593">&os;/powerpc now supports
-       DMA bounce buffer which is required on systems with larger RAM
-       than 4GB.</para>
-
-      <para arch="mips" revision="215938">&os;/mips support has been
-       improved.  It now supports SMP on a SWARM with a dual-core
-       Sibyte processor.</para>
-
-      <para arch="mips">&os;/mips now supports Netlogic Microsystems'
-       XLR and XLS multi-core processor families.</para>
-
-      <para arch="sparc64" revision="215598">&os;/sparc64 now supports
-       reservation-based physical memory allocation which provides
-       better performance.</para>
-
-      <para arch="amd64" revision="214620">&os;/amd64 now always sets
-       the KVA space as equal to or larger than physical memory size.
-       The default size was calculated based on one-third of the
-       physical memory size by a code derived from one for i386.  It
-       has been changed because constraints for memory space are not
-       severe on amd64 and this change would help to prevent a
-       <quote>kmem_map too small</quote> panic which often occurs
-       when using ZFS.</para>
-
-      <para arch="amd64,i386" revision="214621">CPU topology detection
-       for Intel CPUs has been improved.</para>
-
-      <para arch="amd64,i386" revision="215513">ACPI suspend/resume
-       functionality support has been improved.</para>
-
-      <para arch="amd64,i386" revision="215513">&os; kernel now
-       supports <function>kern_fpu_enter()</function> and
-       <function>kern_fpu_leave()</function> KPIs which allow the
-       kernel subsystems to use XMM register files used in Intel SSE
-       (Streaming SIMD Extensions).</para>
-
-      <para arch="amd64,i386,ia64" revision="215006">The &man.acpi.4;
-       driver now uses ACPI Reset Register capability by default only
-       when a flag in the FADT which indicates it is available.  This
-       behavior was controlled by a &man.sysctl.8; variable
-       <varname>hw.acpi.handle_reboot</varname> and the default value
-       was always set to <literal>0</literal>.</para>
-
-      <para arch="amd64,i386,ia64" revision="215521">The &man.acpi.4;
-       driver now supports new loader tunables
-       <varname>hw.acpi.install_interface</varname> and
-       <varname>hw.acpi.remove_interface</varname>.  For more
-       details, see &man.acpi.4; manual page.</para>
-
-      <para revision="209788">The &man.alq.9; support has been
-       improved.  The <function>alq_writen()</function> and
-       <function>alq_getn()</function> KPIs have been extended to
-       support variable length messages, which is enabled at ALQ
-       creation time depending on the arguments passed to
-       <function>alq_open()</function>.  Also, the
-       <literal>ALQ_NOACTIVATE</literal> and
-       <literal>ALQ_ORDERED</literal> flags have been added to allow
-       ALQ consumers to have more control over I/O scheduling and
-       resource acquisition respectively.  These extensions are fully
-       backward compatible.</para>
-
-      <para revision="209783">The &man.alq.9; support is now provided
-       as a kernel module <filename>alq.ko</filename>.</para>
-
-      <para revision="209692">The &man.ddb.8; kernel debugger now
-       supports an optional delay in <command>reset</command> and
-       <command>reboot</command> commands.  This allows an
-       administrator to break the system into debugger and trigger
-       automatic textdump when an unattended panic occurs.</para>
-
-      <para revision="212230">The &man.ddb.8; kernel debugger now
-       supports a <command>show cdev</command> command.  This
-       displays the list of all created cdev's, consisting of devfs
-       node name and struct cdev address.</para>
-
-      <para revision="214326">The &os; <filename>GENERIC</filename>
-       kernel is now compiled with <option>KDB</option> and
-       <option>KDB_TRACE</option> options.  From 8.2-RELEASE the
-       kernel supports displaying a stack trace on panic by using
-       &man.stack.9; facility with no debugger backend like
-       &man.ddb.8;.  Note that this does not change the default
-       behaviors of the <filename>GENERIC</filename> kernel on
-       panic.</para>
-
-      <para revision="213554,213556,213560">The following
-       &man.sysctl.8; variables are also now loader tunables:
-       <varname>vm.kmem_size</varname>,
-       <varname>vm.kmem_size_max</varname>, and
-       <varname>vm.kmem_size_min</varname>,
-       <varname>debug.kdb.stop_cpus</varname>,
-       <varname>debug.trace_on_panic</varname>, and
-       <varname>kern.sync_on_panic</varname>.  Also, new
-       &man.sysctl.8; variables <varname>vm.kmem_map_size</varname>
-       for the current kmem map size and
-       <varname>vm.kmem_map_free</varname> for largest contiguous
-       free range in kmem map, <varname>vfs.ncsizefactor</varname>
-       for size factor for namecache, and
-       <varname>vfs.ncnegfactor</varname> for ratio of negative
-       namecache entries have been added.</para>
-
-      <para revision="212427">The &os; &man.memguard.9; framework has
-       been improved to make it able to detect use-after-free of
-       allocated memories over a longer time.  For more details, see
-       &man.memguard.9; manual page.</para>
-
-      <para revision="214756"><literal>PT_LWPINFO</literal> request to
-       obtain information about the kernel thread that caused the
-       traced process to stop in the &man.ptrace.2; process tracing
-       and debugging facility has been improved.  It now reports
-       system call entry and leave events, as well as availability of
-       <varname>siginfo_t</varname> accompanying the reported
-       signal.</para>
-
-      <para revision="214254">The &os; &man.crypto.4; framework
-       (opencrypto) now supports XTS-AES (XEX-TCB-CTS, or XEX-based
-       Tweaked Code Book mode with CipherText Stealing), which is
-       defined in IEEE Std. 1619-2007.</para>
-
-      <para arch="amd64" revision="215788">Xen HVM support in
-       &os;/amd64 kernel has been improved.  For more details, see
-       &man.xen.4; manual page.</para>
-
-      <para arch="amd64,i386" revision="214784">The qpi(4) pseudo bus
-       driver has been added.  This supports extra PCI buses on Intel
-       QPI chipsets where various hardware such as memory controllers
-       for each socket is connected.</para>
+      <para></para>
 
       <sect3 id="boot">
        <title>Boot Loader Changes</title>
 
-       <para revision="213994">&os; now fully supports GPT (GUID
-         Partition Table).  Checksums of primary header and primary
-         partition table are verified properly now.</para>
-
-       <para arch="powerpc" revision="215919,215920">Memory
-         management issues that prevented &os; OpenFirmware loader
-         and netbooting from working have been fixed.</para>
-
-       <para revision="212717">The &man.pxeboot.8; now uses NFS
-         version 3 instead of version 2 by default.</para>
+       <para></para>
       </sect3>
 
       <sect3 id="proc">
        <title>Hardware Support</title>
 
-       <para arch="amd64,i386" revision="215633">The &man.aesni.4;
-         driver has been added.  This supports AES accelerator on
-         Intel CPUs and accelerates AES operations for
-         &man.crypto.4;.</para>
-
-       <para arch="amd64,i386" revision="210476">The &man.aibs.4;
-         driver has been added.  This supports the hardware sensors
-         in ASUS motherboards and replaces the &man.acpi.aiboost.4;
-         driver.</para>
-
-       <para arch="amd64,i386" revision="211914">The &man.coretemp.4;
-         driver now supports Xeon 5500/5600 series.</para>
-
-       <para arch="powerpc" revision="215694">&os;/powerpc now
-         supports the I2C bus in Apple System Management Unit.</para>
-
-       <para arch="powerpc" revision="215695">A device driver that
-         supports CPU temperature sensors on PowerMac 11,2 has been
-         added.</para>
-
-       <para revision="215482">The &man.ehci.4;, &man.ohci.4;, and
-         &man.uhci.4; driver now support LOW speed BULK transfer
-         mode.</para>
-
-       <para arch="amd64,i386" revision="209952">The &man.ichwd.4;
-         driver now supports Intel NM10 Express chipset watchdog
-         timer.</para>
-
-       <para revision="215036">The &man.tpm.4; driver, which supports
-         Trusted Platform Module has been added.</para>
-
-       <para revision="215944">The xhci(4) driver, which supports
-         Extensible Host Controller Interface (xHCI) and USB 3.0, has
-         been added.</para>
+       <para></para>
 
        <sect4 id="mm">
          <title>Multimedia Support</title>
 
-         <para>The &os; Linux emulation subsystem now supports the
-           <application>video4linux</application> API.  This requires
-           native <application>video4linux</application> hardware
-           drivers such as the ones provided by <filename
-           role="ports">multimedia/pwcbsd</filename> and <filename
-           role="ports">multimedia/webcamd</filename>.</para>
-
-         <para revision="215483">MIDI input buffer size in the
-           &man.uaudio.4; driver has been changed.  This fixes a
-           problem where the input appears several seconds
-           late.</para>
-
-         <para revision="215484">An issue in the &man.uaudio.4;
-           driver that prevented some USB audio devices from working
-           has been fixed.</para>
+         <para></para>
        </sect4>
 
        <sect4 id="net-if">
          <title>Network Interface Support</title>
 
-         <para revision="212021">The &man.alc.4; driver now supports
-           Atheros AR8151/AR8152 PCIe Gigabit/Fast Ethernet
-           controllers.</para>
-
-         <para revision="215003">A bug in the &man.alc.4; driver was
-           fixed that could lead to a system freeze when the system
-           was booted without a cable plugged in.  This symptom was
-           found in AR8132 on EEE PC.</para>
-
-         <para revision="212011">The TX interrupt moderation timer in
-           the &man.alc.4; driver has been reduced from 50ms to 1ms.
-           The 50ms timer resulted in a poor UDP performance.</para>
-
-         <para revision="214297,214298,214301">The &man.axe.4; driver
-           has been improved for stability and better performance on
-           the TX packet rate.</para>
-
-         <para revision="214997">The &man.bge.4; driver now supports
-           BCM5718 x2 PCI Express dual-port gigabit Ethernet
-           controller family.  This family is the successor to the
-           BCM5714/BCM5715 family and supports IPv4/IPv6 checksum
-           offloading, TSO, VLAN hardware tagging, jumbo frames,
-           MSI/MSIX, IOV, RSS and TSS.  The current version of the
-           driver supports all hardware features except IOV and
-           RSS/TSS.</para>
-
-         <para revision="211367">A bug in the &man.bge.4; driver which
-           prevented TSO from working in BCM57780 has been
-           fixed.</para>
-
-         <para revision="214700">A bug in the &man.bge.4; driver that
-           could wrongly disable the TX checksum offloading feature
-           as well when one tries to disable only the RX checksum
-           offloading has been fixed.</para>
-
-         <para revision="214714">Some improvements for reliability of
-           the &man.bge.4; driver with BCM5906 controller has been
-           made.</para>
-
-         <para revision="214698">The &man.bge.4; driver now supports
-           hardware MAC statistics in controller's internal memory
-           for BCM5705 or newer Broadcom controllers.  These counters
-           can be accessed via &man.sysctl.8; variable
-           <varname>dev.bge.<replaceable>N</replaceable>.stats.*</varname>
-           and provide useful information to diagnose driver
-           issues.</para>
-
-         <para revision="213109">UDP checksum offloading in the
-           &man.bge.4; driver has been disabled by default.  This is
-           because Broadcom controllers have a bug which can generate
-           UDP datagrams with checksum value <literal>0</literal>
-           when TX UDP checksum offloading is enabled.  The checksum
-           offloading can be enabled by using the following loader
-           tunable:</para>
-
-         
<programlisting>dev.bge.<replaceable>N</replaceable>.forced_udpcsum</programlisting>
-
-         <para revision="213593">A bug in the &man.bge.4; driver that
-           could lead to poor performance on a system with more than
-           4 GB RAM has been fixed.  The cause was that all of
-           Broadcom controllers except the BCM5755 and later have a
-           bug in 4 GB-boundary DMA processing and used the bounce
-           buffer in an inefficient way.</para>
-
-         <para revision="212275">The &man.bwi.4; driver, which supports
-           Broadcom BCM430* and BCM431* family Wireless Ethernet
-           controllers, has been added.  This is not compiled into
-           the <filename>GENERIC</filename> kernel because there are
-           some problems.  The kernel module
-           <filename>if_bwi.ko</filename> is available and can be
-           loaded without recompiling the kernel to enable this
-           driver.</para>
-
-         <para revision="212274">A bug in the &man.bwn.4; driver that
-           prevented WPA authentication from working has been
-           fixed.</para>
-
-         <para revision="215499">A bug in the &man.cdce.4; driver has
-           been fixed.</para>
-
-         <para revision="211848">The &man.cxgb.4; driver now supports
-           the following new &man.sysctl.8; variables:
-           <varname>hw.cxgb.nfilters</varname> sets the maximum
-           number of entries in the hardware filter table,
-           
<varname>dev.cxgbc.<replaceable>N</replaceable>.pkt_timestamp</varname>
-           provides packet timestamp instead of connection hash, and
-           <varname>dev.cxgbc.<replaceable>N</replaceable>.core_clock</varname>
-           provides the core clock frequency in kHz.</para>
-
-         <para>The &man.em.4; driver has been updated to version
-           7.1.9.</para>
-
-         <para>The &man.igb.4; driver has been updated to version
-           2.0.7.</para>
-
-         <para revision="209309">The &man.em.4; and &man.igb.4; drivers
-           now provide statistics counters as &man.sysctl.8; MIB
-           objects.</para>
-
-         <para revision="211241">The &man.em.4; and &man.igb.4; drivers
-           now support the &man.led.4; interface via
-           <filename>/dev/led/em<replaceable>N</replaceable></filename>
-           and
-           <filename>/dev/led/igb<replaceable>N</replaceable></filename>
-           for identification LED control.  The following command
-           line makes the LED blink on <literal>em0</literal>:</para>
-
-         <screen>&prompt.root; echo f2 &gt; /dev/led/em0</screen>
-
-         <para revision="212150">The &man.epair.4; virtual Ethernet
-           interface driver now supports explicit UP/DOWN linkstate.
-           This fixes an issue when it is used with the &man.carp.4;
-           protocol.</para>
-
-         <para revision="214717">The &man.fxp.4; driver now supports
-           TSO over VLAN on i82550 and i82551 controllers.</para>
-
-         <para revision="210673">The &man.iwn.4; driver now supports
-           Intel Wireless WiFi Link 6000 series.  The firmware has
-           been updated to version 9.221.4.1.</para>
-
-         <para revision="212386">The &man.ixgbe.4;
-           driver is now also provided as a kernel module.</para>
-
-         <para revision="209308,215970,217593">The &man.ixgbe.4;
-           driver has been updated to version 2.3.8.  It now supports
-           82599, better interrupt handling, hardware assist to LRO,
-           VM SRIOV interface, and so on.</para>
-
-         <para
-           
revision="211379,215881,215890,2105894,216002,216023,216029,216031,216033">The
-           &man.miibus.4; has been rewritten for the generic IEEE
-           802.3 annex 31B full duplex flow control support.  The
-           &man.alc.4;, &man.bge.4;, &man.bce.4;, &man.cas.4;,
-           &man.fxp.4;, &man.gem.4;, &man.jme.4;, &man.msk.4;,
-           &man.nfe.4;, &man.re.4;, &man.stge.4;, and &man.xl.4;
-           drivers along with atphy(4), bmtphy(4), brgphy(4),
-           e1000phy(4), gentbi(4), inphy(4), ip1000phy(4), jmphy(4),
-           nsgphy(4), nsphyter(4), and &man.rgephy.4; have been
-           updated to support flow control via this facility.</para>
-
-         <para revision="212386">The &man.mwlfw.4;
-           driver is now also provided as a kernel module.</para>
-
-         <para revision="216054">A bug in the &man.mxge.4; driver
-           that prevented TSO from working has been fixed.</para>
-
-         <para revision="215456">The &man.nfe.4; driver now supports
-           WoL (Wake on LAN).</para>
-
-         <para revision="215218">The &man.re.4; driver now supports
-           64-bit DMA addressing for RTL810xE/RTL8168/RTL8111 PCIe
-           controllers.</para>
-
-         <para revision="215405">The &man.re.4; driver now supports
-           hardware interrupt moderation of TX completion interrupts
-           on RTL8169/RTL8168 controllers.</para>
-
-         <para revision="211377">The &man.rl.4; driver now supports WoL
-           (Wake on LAN) on RTL8139B or newer controllers.</para>
-
-         <para revision="215222">The &man.rl.4; driver now supports
-           reading hardware statistics counters by setting a
-           &man.sysctl.8; variable
-           <varname>dev.rl.<replaceable>N</replaceable>.stats</varname>
-           to <literal>1</literal>.</para>
-
-         <para revision="212039">The &man.rl.4; driver now supports a
-           device hint to change a way of register access.  Although
-           some newer RTL8139 controllers support memory-mapped
-           register access, it is difficult to detect the support
-           automatically.  For this reason the driver uses I/O
-           mapping by default and provides the following device hint.
-           If it is set to <literal>0</literal>, the driver uses
-           memory mapping for register access.</para>
-
-         
<programlisting>hint.rl.<replaceable>N</replaceable>.prefer_iomap="0"</programlisting>
-
-         <para>Note that the default value is <literal>1</literal>.</para>
-
-         <para revision="214407">The &man.rl.4; driver has improved
-           interrupt handling.  It now has better TX performance
-           under high RX load.</para>
-
-         <para revision="215216">A bug in the &man.sk.4; driver has
-           been fixed.  It did not program the station address for
-           Yukon controllers and overriding the station address with
-           &man.ifconfig.8; was not possible.</para>
-
-         <para revision="211359">The &man.sk.4; driver now disables TX
-           checksum offloading by default.  This is because some
-           revisions of the Yukon controller generate corrupted frames.
-           The checksum offloading can be enabled manually by using
-           <option>txcsum</option> option in the &man.ifconfig.8;
-           utility.</para>
-
-         <para revision="212468,213612,213628">The &man.sis.4; driver
-           now works on all supported platforms.  Some stability and
-           performance issues have also been fixed.</para>
-
-         <para revision="213637">The &man.sis.4; driver now supports
-           WoL (Wake on LAN) on NS DP8315 controller.</para>
-
-         <para revision="213626">A tunable
-           <varname>dev.sis.<replaceable>N</replaceable>.manual_pad</varname>
-           for the &man.sis.4; driver has been added.  This controls
-           whether padding with 0x00 for short frames is done by CPU,
-           rather than the controller.  The reason why this tunable
-           has been added is that NS DP83815/DP83816 pads them with
-           0xff though RFC 1042 specifies it should be 0x00.  The
-           tunable is disabled by default, which means padding with
-           0xff is used because padding with 0x00 by software needs
-           extra CPU cycles.  Enabling <varname>manual_pad</varname>,
-           by setting this &man.sysctl.8; variable to a non-zero
-           value, forces the use of software padding.</para>
-
-         <para revision="211357">The &man.ste.4; driver now supports
-           a device hint to change the device register access mode.
-           The driver uses memory-mapped register access by default,
-           but this caused stability problems with some old IC Plus
-           Corp (formerly Sundace) controllers.  The following device
-           hint makes the driver use I/O mapping for register
-           access:</para>
-
-         
<programlisting>hint.ste.<replaceable>N</replaceable>.prefer_iomap="1"</programlisting>
-
-         <para revision="213120">The &man.xl.4; driver now supports
-           WoL (Wake on LAN).  Note that not all controllers support
-           this functionality and some need an additional remote
-           wakeup cable.</para>
+         <para></para>
        </sect4>
       </sect3>
 
       <sect3 id="net-proto">
        <title>Network Protocols</title>
 
-       <para revision="213505">An issue in the &man.carp.4; pseudo
-         interface and linkstate changes of the underlying interfaces
-         has been fixed.  This happened when a &man.carp.4; interface
-         was created before the underlying interface and its
-         linkstate became <literal>UP</literal>.</para>
-
-       <para revision="212738,216348">A bug in the &man.ipfw.4;
-         packet filter subsystem has been fixed.  The &man.sysctl.8;
-         variable <varname>net.inet.ip.fw.one_pass</varname> did not
-         work for <literal>netgraph</literal> action and in-kernel
-         NAT.</para>
-
-       <para revision="215342">A new loader tunable
-         <varname>net.link.ifqmaxlen</varname> has been added.  It
-         specifies the default value of send interface queue length.
-         The default value for this parameter is
-         <varname>50</varname>.</para>
-
-       <para revision="212742">The <literal>ngtee</literal> action in
-         the &man.ipfw.4; packet filter subsystem has been changed.
-         It no longer accepts a packet.</para>
-
-       <para revision="214108">A possible panic in the &man.ipfw.4;
-         pseudo interface for logging has been fixed.</para>
-
-       <para revision="209691">IPsec flow distribution has been
-         improved for more parallel processing.</para>
-
-       <para revision="209277">A bug in the &os; IPv4 stack that
-         prevented adding a proxy ARP entry over &man.netgraph.4;
-         interfaces has been fixed.</para>
-
-       <para revision="211435">A bug in the &os; IPv6 stack that
-         prevented an <option>-I</option> in the &man.ping6.8;
-         utility from working with
-         <varname>net.inet6.ip6.use_defaultzone=1</varname> has been
-         fixed.</para>
-
-       <para revision="214111">The &man.lagg.4; interface now
-         supports a &man.sysctl.8; variable
-         <varname>net.link.lagg.failover_rx_all</varname>.  This
-         controls whether to accept input packets on any link in a
-         failover lagg.</para>
-
-       <para revision="215869">The &man.ng.eiface.4; &man.netgraph.4;
-         node now supports VLAN-compatible MTU and an MTU size which
-         is larger than 1500.</para>
-
-       <para revision="212320">The &man.ng.ether.4; &man.netgraph.4;
-         node now supports interface transfer between multiple virtual
-         network stacks by &man.ifconfig.8; <command>vnet</command>
-         command.  A &man.ng.ether.4; node associated with a network
-         interface is now destroyed and recreated when the network
-         interface is moved to another vnet.</para>
-
-       <para revision="209843">A new &man.netgraph.4; node
-         &man.ng.patch.4; has been added.  This performs data
-         modification of packets passing through.  Modifications are
-         restricted to a subset of C language operations on unsigned
-         integers of 8, 16, 32 or 64-bit size.</para>
-
-       <para revision="212905">An ICMP unreachable problem in the
-         &man.pf.4; packet filter subsystem when TSO support is
-         enabled has been fixed.</para>
-
-       <para revision="211538">The TCP bandwidth delay product window
-         limiting algorithm controlled by the &man.sysctl.8; variable
-         <varname>net.inet.tcp.inflight.enable</varname> is now
-         disabled by default.  It has been found that this algorithm
-         is inefficient on a fast network with smaller RTT than 10ms.
-         It had been enabled by default since 5.2-RELEASE, and then
-         had been disabled only if the RTT was lesser than 10ms since
-         7.0-RELEASE.  Pluggable TCP congestion control algorithm
-         modules are planned to be added for the future
-         releases.</para>
-
-       <para revision="211602">A bug in &os; TCP Path MTU discovery
-         which could lead to a wrong calculation for an MTU smaller
-         than 256 octets has been fixed.  Note that this bug did not
-         affect MTUs equal to or larger than 256 octets.</para>
-
-       <para revision="214865,214866">The &os; TCP reassembly
-         implementation has been improved.  A long-standing
-         accounting bug affecting SMP systems has been fixed and the
-         <varname>net.inet.tcp.reass.maxqlen</varname> &man.sysctl.8;
-         variable has been retired in favor of a per-connection
-         dynamic limit based on the receive socket buffer size.  &os;
-         receivers now handle packet loss (particularly losses caused
-         by queue overflows) significantly better than before which
-         improves connection throughput.</para>
-
-       <para revision="211870">The TCP initial window increase in RFC
-         3390 which can be controlled by a &man.sysctl.8; variable
-         <varname>net.inet.tcp.rfc3390</varname> now reduces the
-         congestion window to the restart window if a TCP connection
-         has been idle for one retransmit timeout or more.  For more
-         details, see RFC 5681 Section 4.1.</para>
-
-       <para revision="214859">The &man.siftr.4;, Statistical
-         Information For TCP Research (SIFTR) kernel module has been
-         added.  This is a facility that logs a range of statistics
-         on active TCP connections to a log file.  It provides the
-         ability to make highly granular measurements of TCP
-         connection state, aimed at system administrators, developers
-         and researchers.</para>
-
-       <para revision="212319">&os; virtual network stack (vnet) now
-         supports IPv4 multicast routing.</para>
-
-       <para revision="239616">The IEEE 802.11s element identifiers have
-         been updated to reflect the final version of the amendment.  This
-         update breaks compatibility with older mesh setups but is necessary
-         as the previous IDs are used by another amendment leading to
-         unexpected results when trying to associate with an accesspoint
-         using the affected IDs.</para>
+       <para></para>
       </sect3>
 
       <sect3 id="disks">
        <title>Disks and Storage</title>
 
-       <para revision="211458">The &man.ahci.4; driver now disables NCQ
-         and PMP support on VIA VT8251 because they are unreliable
-         under load.</para>
-
-       <para revision="211718">The &man.ahci.4; driver now uses 15
-         seconds for device reset timeout instead of 10 seconds
-         because some devices need 10 - 12 seconds to spin up.</para>
-
-       <para revision="215825">The &man.arcmsr.4; driver
-         has been updated to version 1.20.00.19.</para>
-
-       <para revision="215173">The &man.ada.4; driver now supports a
-         new &man.sysctl.8; variable
-         <varname>kern.cam.ada.spindown_shutdown</varname> which
-         controls whether or not to spin-down disks when shutting
-         down if the device supports the functionality.  The default
-         value is <literal>1</literal>.</para>
-
-       <para revision="210164">The &man.ata.4; driver
-         now supports limiting initial ATA mode for devices via
-         device hints
-         
<varname>hint.<replaceable>devname</replaceable>.<replaceable>unit</replaceable>.dev<replaceable>N</replaceable
-         >.mode</varname> or
-         
<varname>hint.<replaceable>devname</replaceable>.<replaceable>unit</replaceable>.mode</varname>.
-         The valid values are the same as ones supported in the
-         &man.atacontrol.8; and &man.camcontrol.8; utilities.</para>
-
-       <para revision="210204">The &man.ata.4; driver now disables
-         cable status check on both controller and device side
-         when the loader tunable
-         <varname>hw.ata.ata_dma_check_80pin</varname> is
-         <literal>0</literal>.  The check on controller side was
-         performed regardless of this loader tunable.</para>
-
-       <para revision="215777">The &man.ata.4; driver now reports
-         SATA power management capabilities to the &man.CAM.4; layer when
-         <option>ATA_CAM</option> is enabled.  This allows a device
-         to initiate transitions if controller configured to accept
-         it.  This makes
-         <varname>hint.ata.<replaceable>N</replaceable>.pm_level=1</varname>
-         mode work.</para>
-
-       <para revision="214671">The &man.ata.4; driver has been
-         improved on hotplugging and connection speed reporting
-         support for some Intel SATA controllers including ICH5 and
-         ICH8+ operating in legacy mode.</para>
-
-       <para revision="213291">An issue of device detection of
-         Serverworks K2 SATA controllers in the &man.ata.4; has been 
fixed.</para>
-
-       <para revision="214667">A bug in the &man.ata.4; driver that
-         prevented some Silicon Image chipsets from working on big
-         endian systems has been fixed.</para>
-
-       <para revision="214330">The &man.gconcat.8; GEOM class now
-         supports kernel crash dump.  The dumping is performed to the
-         component where a dump partition begins.</para>
-
-       <para revision="214405">A bug in the &man.geli.8; GEOM class
-         on little endian platforms has been fixed.  The metadata
-         version for newly created providers has been updated to
-         <literal>4</literal> due to this.  Providers with the older
-         versions are fully interoperable with 8.2-RELEASE and later
-         by being treated as ones with the native byte order flag
-         automatically.</para>
-
-       <para revision="214405">The &man.geli.8; GEOM class now
-         supports a &man.sysctl.8; variable
-         <varname>kern.geom.eli.overwrites</varname>.  This specifies
-         the number of times on-disk keys should be overwritten when
-         destroying them.  The default value is
-         <literal>5</literal>.</para>
-
-       <para revision="214405">The &man.geli.8; GEOM class has been
-         improved for preventing the same encryption key from being
-         used in 2^20 blocks (sectors).</para>
-
-       <para revision="214405">The &man.geli.8; GEOM class now uses
-         XTS-AES mode by default.</para>
-
-       <para revision="214405">A &man.sysctl.8; variable
-         <varname>kern.geom.eli.debug</varname> now allows a value
-         <literal>-1</literal>.  This means turn off any log messages
-         of the &man.geli.8; GEOM class.</para>
-
-       <para revision="210376">The &man.mpt.4; driver now supports
-         larger I/O sizes which the device and &man.CAM.4; subsystem
-         can support.  This was limited to 64KB, and the number of
-         scatter/gather segments was limited to 33 on platforms with
-         4K pages.</para>
-
-       <para revision="209404">The &man.twa.4;
-         driver has been updated.  The version number is
-         3.80.06.003.</para>
+       <para></para>
       </sect3>
 
       <sect3 id="fs">
        <title>File Systems</title>
 
-       <para revision="214050">The &man.linprocfs.5; Linux process
-         file system now supports
-         <filename>proc/$$/environment</filename>.</para>
-
-       <para revision="212716">The &os; NFS client now supports a
-         kernel environment variable
-         <varname>boot.nfsroot.nfshandlelen</varname>.  This lets the
-         diskless root file system on boot to use NFS version 3 and
-         the specified file handle length.  If this variable is not
-         set, NFS version 2 is used.</para>
-
-       <para revision="212668">The ZFS on-disk format has been updated
-         to version 15.</para>
-
-       <para revision="212671">The ZFS metaslab code has been updated.
-         This provides a noticeable improvement on write speed,
-         especially on pools with less than 30% of free space.  The
-         related OpenSolaris Bug IDs are 6826241, 6869229, 6918420,
-         and 6917066.</para>
-
-       <para revision="214013">The ZFS now supports offlining of log
-         devices.  The related OpenSolaris Bug IDs are 6599442,
-         6726045, and 6803605.</para>
-
-       <para revision="212672">Performance improvements for the ZFS
-         have been imported from OpenSolaris.  They include caching of
-         ACL permission checks, faster handling of &man.stat.2;,
-         mitigation of mutex lock contention.  The related
-         OpenSolaris Bug IDs are 6802734, 6844861, 6848431, 6775100,
-         6827779, 6857433, 6860318, 6865875, 6867395, 6868276, and
-         6870564.</para>
-
-       <para revision="210541">The default value of
-         <varname>vfs.zfs.vdev.max_pending</varname> has been
-         decreased from 35 to 10 (OpenSolaris Bug ID is 6891731) to
-         improve latency.</para>
-
-       <para>Various bugs in the ZFS subsystem have been fixed.  The
-         related OpenSolaris Bug IDs are: 6328632, 6396518, 6501037,
-         6504953, 6542860, 6551866, 6572357, 6572376, 6582163,
-         6586537, 6595194, 6596237, 6604992, 6621164, 6623978,
-         6633095, 6635482, 6664765, 6674216, 6696242, 6696858,
-         6702206, 6710376, 6713916, 6717022, 6722540, 6722991,
-         6737463, 6739487, 6739553, 6740164, 6745863, 6747596,
-         6747698, 6748436, 6755435, 6757430, 6758107, 6759986,
-         6759999, 6761100, 6761406, 6764124, 6765294, 6767129,
-         6769612, 6770866, 6774713, 6774886, 6775697, 6776104,
-         6776548, 6780491, 6784104, 6784108, 6785914, 6788152,
-         6788830, 6789318, 6790064, 6790345, 6790687, 6791064,
-         6791066, 6791071, 6791101, 6792134, 6792139, 6792884,
-         6793430, 6794136, 6794570, 6794830, 6797109, 6797118,
-         6798384, 6798878, 6799895, 6800184, 6800942, 6801507,
-         6801810, 6803343, 6803822, 6804954, 6807339, 6807765,
-         6809340, 6809683, 6809691, 6810367, 6815592, 6815893,
-         6816124, 6818183, 6821169, 6821170, 6822816, 6824006,
-         6824062, 6824968, 6826466, 6826468, 6826469, 6826470,
-         6826471, 6826472, 6827260, 6830237, 6830541, 6833162,
-         6833711, 6833999, 6834217, 6836714, 6836768, 6838062,
-         6838344, 6841321, 6843014, 6843069, 6843235, 6844069,
-         6844900, 6847229, 6848242, 6856634, 6857012, 6861983,
-         6862984, 6863610, 6870564, 6880764, 6882227, 6892298,
-         6898245, 6906110, 6906946, 6939941, 6950219, 6951024, and
-         6953403.</para>
+       <para></para>
       </sect3>
     </sect2>
 
     <sect2 id="userland">
       <title>Userland Changes</title>
 
-      <para revision="209267">The &man.arp.8; utility has been improved.
-       It now runs faster even when a single interface has a number
-       of aliases.</para>
-
-      <para revision="214729">A bug in the &man.b64decode.1; utility that
-       prevented an <option>-r</option> option from handling arbitrary
-       breaks in a base64 encoded string has been fixed.</para>
-
-      <para revision="211723">The &man.calendar.1; utility now supports
-       repeating events which span multiple years, lunar events, and
-       solar events.</para>
-
-      <para revision="210915">The &man.dhclient.8; utility now reports a
-       reason for exiting and the 10-second period in which the
-       &man.dhclient.8; ignores routing messages has been changed to
-       start just after <filename>dhclient-script</filename> starts
-       instead of just after it finished.  This change fixes a
-       symptom that &man.dhclient.8; silently exits under a certain
-       condition.</para>
-
-      <para revision="214983">Userland support for the &man.dtrace.1;
-       subsystem has been added.  This allows inspection of userland
-       software itself and its correlation with the kernel, thus
-       allowing a much better picture of what exactly is going on
-       behind the scenes.  The &man.dtruss.1; utility has been added
-       and the <filename>libproc</filename> library has been updated
-       to support the facility.</para>
-
-      <para revision="209362">The &man.du.1; utility now supports a
-       <option>-t <replaceable>threshold</replaceable></option>
-       option to display entries that exceeds the value of
-       <replaceable>threshold</replaceable>.  If the value is
-       negative, it displays entries with a value less than the
-       absolute value of <replaceable>threshold</replaceable>.</para>
-
-      <para revision="215667">The &man.fdisk.8; utility now supports
-       partitions which are provided by &man.gjournal.8; or
-       &man.geli.8; GEOM classes.</para>
-
-      <para revision="210567">The &man.gcore.1; utility now supports an
-       <option>-f</option> flag which forces a full dump of all the
-       segments except for the malformed ones.</para>
-
-      <para revision="214405">The &man.geli.8; utility now supports
-       <command>resize</command> subcommand to resize encrypted file
-       systems after growing it.</para>
-
-      <para revision="214405">The &man.geli.8; utility now supports
-       <command>suspend</command> and <command>resume</command>
-       subcommands.  The <command>suspend</command> subcommand makes
-       &man.geli.8; devices wait for all in-flight I/O requests,
-       suspend new I/O requests, remove all &man.geli.8; sensitive
-       data from the kernel memory (like encryption keys) and will
-       wait for either <command>geli resume</command> or
-       <command>geli detach</command> command.  For more
-       information, see &man.geli.8; manual page.</para>
-
-      <para revision="214405">The &man.geli.8; utility now checks the
-       metadata provider size strictly.  If the check fails, the
-       provider is not attached.  A new option <option>-f</option>
-       can override this behavior.</para>
-
-      <para revision="214405">The &man.geli.8; utility now supports
-       <option>-J <replaceable>newpassfile</replaceable></option> and
-       <option>-j <replaceable>oldpassfile</replaceable></option>
-       options for loading passphrase from a file.</para>
-
-      <para revision="211938">The <function>gethost*()</function>,
-       <function>getnet*()</function>, and
-       <function>getproto*()</function> functions now set the errno
-       to <literal>ERANGE</literal> and the NSS backend terminates
-       with <literal>NS_RETURN</literal> when the result buffer size
-       is too small.</para>
-
-      <para revision="209497">The &man.gpart.8; utility now supports a
-       <command>resize</command> command to resize partitions for all
-       schemes but EBR.</para>
-
-      <para revision="215941">The &man.gpart.8; utility now supports
-       <command>backup</command> and <command>restore</command>
-       subcommands to backup partition tables and restore
-       them.</para>
-
-      <para revision="213661">The &man.gpart.8; utility now handles
-       given geom/provider names with and without
-       <filename>/dev/</filename> prefix.</para>
-
-      <para revision="213684">The &man.gpart.8; utility now supports
-       an <option>-F</option> option for the
-       <command>destroy</command> subcommand.  This option forces
-       destroying of the partition table even if it is not
-       empty.</para>
-
-      <para revision="213684">The &man.gpart.8; utility now supports a
-       <command>recover</command> subcommand for GPT partition
-       tables.  A corrupted GPT is now marked when the following
-       three types of corruption:</para>
-
-      <itemizedlist>
-       <listitem>
-         <para>Primary GPT header or table is corrupted.</para>
-       </listitem>
-
-       <listitem>
-         <para>Secondary GPT header or table is corrupted.</para>
-       </listitem>
-
-       <listitem>
-         <para>Secondary GPT header is not located at the last LBA.</para>
-       </listitem>
-      </itemizedlist>
-
-      <para>Changes to the corrupted GPT table are not allowed except
-       for <command>destroy</command> and <command>recover</command>
-       subcommands.</para>
-
-      <para revision="213994">The &man.gpart.8; utility now supports
-       <literal>GPT_ENT_ATTR_BOOTME</literal>,
-       <literal>GPT_ENT_ATTR_BOOTONCE</literal>, and
-       <literal>GPT_ENT_ATTR_BOOTFAILED</literal> attributes in GPT.
-       The attribute keywords in the command line are
-       <literal>bootme</literal>, <literal>bootonce</literal>, and
-       <literal>bootfailed</literal> respectively.</para>
-
-      <para revision="215728">An issue in the &man.newfs.8; utility
-       has been fixed.  A UFS1 file system created with 64KB
-       blocksize was incorrectly recognized as one with a broken
-       superblock.  This is because the &os; kernel checks a
-       partition first for a UFS2 superblock at 64KB offset while it
-       is possible that a UFS1 file systems with 64KB blocksize has
-       an alternative superblock at the same location.  For example,
-       a file system created by <command>newfs -U -O 1 -b 65536 -f
-       8192</command> could lead to this symptom.</para>

*** DIFF OUTPUT TRUNCATED AT 1000 LINES ***
_______________________________________________
svn-src-all@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"

Reply via email to