Attachment #63277 for bug #94423

View | Details | Raw Unified | Return to bug 94423
Collapse All | Expand All




.include "../Makefile.inc"
.endif

DOCS=	approvals.sgml schedule.sgml

SRCS.DEFAULT=	todo.xsl
XMLDOCS=	todo

FETCHFILE= http://people.freebsd.org/~pho/stress/log/news.rdf
FETCHCMD=  /usr/bin/fetch
CLEANFILES+= news.rdf stress.xml

news.rdf:
	${FETCHCMD} ${FETCHFILE}

stress.xml: news.rdf stress.xsl
	${XSLTPROC} stress.xsl news.rdf > stress.xml

todo.html: stress.xml

.include "${WEB_PREFIX}/share/mk/web.site.mk"




 version="1.0">

<xsl:output 
 method="xml"
 indent="yes"
 encoding="utf-8"
 omit-xml-declaration="no"/>

<!-- match first element whether we're using namespaces or not -->

<xsl:template match="/*[1]">
<results>
<ul>
  <xsl:for-each select="*[local-name()='item']">
  <result>
    <title><xsl:value-of select="*[local-name()='title']"/></title>
    <link><xsl:value-of select="*[local-name()='link']"/></link>
  </result>
    </xsl:element>
  </li>
</xsl:for-each>
</results>

</xsl:template>

</xsl:stylesheet>




<!-- DTD for todo list entries on the FreeBSD website. -->
<!-- $FreeBSD:$ -->

<!ELEMENT entries (cvs:keywords?, entry+)>

<!ELEMENT entry (issue, responsible?, description?)>
<!ATTLIST entry class (showstopper | required | desired | docs
	| testing) #REQUIRED>
<!ATTLIST entry status (na | done | wip | untested | new | unknown
	| deferred) #REQUIRED>

<!ELEMENT issue (#PCDATA)>
<!ELEMENT responsible (#PCDATA)>
<!ELEMENT description (#PCDATA|a|b|br|i|tt|em|p|strong|ul|pre)*>

<!-- misc HTML tags -->
<!ELEMENT a (#PCDATA|img)*>
<!ATTLIST a href CDATA #REQUIRED>

<!ELEMENT b (#PCDATA|a)*>
<!ELEMENT br (#PCDATA|a)*>
<!ELEMENT i (#PCDATA|a)*>
<!ELEMENT tt (#PCDATA|a)*>
<!ELEMENT em (#PCDATA|a)*>
<!ELEMENT p (#PCDATA|a)*>
<!ELEMENT strong (#PCDATA|a)*>
<!ELEMENT ul (li)*>
<!ELEMENT li (#PCDATA|a)*>
<!ELEMENT pre (#PCDATA|a)*>

<!-- common cvs tags -->
<!ELEMENT cvs:keywords (cvs:keyword+)>
<!ATTLIST cvs:keywords xmlns:cvs CDATA #FIXED 'http://www.FreeBSD.org/XML/CVS'>
<!ATTLIST cvs:keywords version CDATA #FIXED '1.0'>

<!ELEMENT cvs:keyword (#PCDATA)>
<!ATTLIST cvs:keyword name CDATA #REQUIRED>




<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN" [
<!ENTITY base CDATA "../..">
<!ENTITY email 'freebsd-qa'>
<!ENTITY date "$FreeBSD: www/en/releases/6.1R/todo.sgml,v 1.18 2006/03/08 03:48:13 hrs Exp $">
<!ENTITY local.rel "6.1">
<!ENTITY title "FreeBSD 6.1 Open Issues">
<!ENTITY % navincludes SYSTEM "../../includes.navdownload.sgml"> %navincludes;
<!ENTITY % includes SYSTEM "../../includes.sgml"> %includes;
<!ENTITY % developers SYSTEM "../../developers.sgml"> %developers;
<!-- Status levels -->
<!ENTITY status.na "<font color=green>N/A</font>">
<!ENTITY status.done "<font color=green>Done</font>">
<!ENTITY status.wip "<font color=blue>In&nbsp;progress</font>">
<!ENTITY status.untested "<font color=orange>Needs&nbsp;testing</font>">
<!ENTITY status.new "<font color=red>Not&nbsp;done</font>">
<!ENTITY status.unknown "<font color=red>Unknown</font>">
<!ENTITY status.deferred "<font color=gray>Deferred for future release</font>">

<!ENTITY url.cvsweb "http://www.freebsd.org/cgi/cvsweb.cgi">
<!ENTITY url.mid "http://docs.freebsd.org/cgi/mid.cgi?">
<!ENTITY url.pr "http://www.freebsd.org/cgi/query-pr.cgi?">

<!ENTITY stresstest SYSTEM "./stress.html">
]>

<!--

  Changes to this list MUST NOT be committed without approval of
  Release Engineering Team (re@FreeBSD.org) (for general items) or
  Documentation Engineering Team (doceng@FreeBSD.org) (for doc-related
  items).

-->

<html>
&header;

<p>This is a list of open issues that need to be resolved for FreeBSD
  &local.rel;.  If you have any updates for this list, please e-mail
  re@FreeBSD.org.</p>

<ul>
  <li><a href="#showstopper">Show stopper defects</a></li>
  <li><a href="#required">Required features</a></li>
  <li><a href="#desired">Desired features</a></li>
  <li><a href="#docs">Documentation Items</a></li>
  <li><a href="#testing">Testing foci</a></li>
  <li><a href="#stresstest">Problems Discovered by Kernel Stress Test Suite</a></li>
</ul>
  
<h3>Show stopper defects for &local.rel;-RELEASE</h3>

<a name="showstopper"></a>
<table class="tblbasic">
  <tr class="heading">
    <th>Issue</th>
    <th>Status</th>
    <th>Responsible</th>
    <th>Description</th>
  </tr>

  <tr>
    <td>unreliable serial console</td>
    <td>&status.unknown;</td>
    <td></td>
    <td>At the manual 'root mount' prompt, the serial console is very
      unreliable and drops most characters.</td>
  </tr>

  <tr>
    <td>i386 deadlocks with >16GB swap</td>
    <td>&status.deferred;</td>
    <td>&a.alc;</td>
    <td>i386 deadlocks if more than 16GB of swap is in use.
      Increasing the kern.maxswzone tunable would be a workaround
      this.  Although a patch from &a.alc; is needed to allow this variable to
      be increased, this is not suitable for 6.1R.  This limitation should
      be documented in the Release Notes.</td>
  </tr>

  <tr>
    <td>unmount pending error</td>
    <td>&status.wip;</td>
    <td>&a.ssouhlal;</td>
    <td>When unmounting filesystems &a.kris; reports seeing this warning: <tt>/c: unmount pending error: blocks -68512 files 0</tt>.  This dates back at least to 5.3.  It might be associated with
filesystem corruption reported by many users in which the 'used' space
on a filesystem is negative; fsck -f is needed to correct this.</td>
  </tr>

  <tr>
    <td>swap_pager warnings</td>
    <td>&status.unknown;</td>
    <td>&a.truckman;?</td>
    <td>When swapfiles are in use, there are often warnings printed:
<tt>swap_pager: indefinite wait buffer: bufobj: 0, blkno: 889347, size: 8192</tt>.  There is also the possibility of deadlock.</td>
  </tr>

  <tr>
    <td>umount -f panics</td>
    <td>&status.wip;</td>
    <td>&a.jeff;, &a.ssouhlal;</td>
    <td>panics from race conditions.
      A patch from &a.jeff; seems to fix some of them.</td>
  </tr>

  <tr>
    <td>quota deadlocks</td>
    <td>&status.wip;</td>
    <td>&a.jeff;</td>
    <td>Quota support is not locked properly and causes deadlocks.
      A patch from &a.jeff; seems to fix some of them.</td>
  </tr>

  <tr>
    <td>UFS deadlocks on amd64</td>
    <td>&status.unknown;</td>
    <td>&a.tegge;</td>
    <td>Seen by &a.kris;.</td>
  </tr>

  <tr>
    <td>UFS deadlocks</td>
    <td>&status.unknown;</td>
    <td></td>
    <td>Seen by Peter Jeremy.</td>
  </tr>

  <tr>
    <td>sparc64 instability</td>
    <td>&status.unknown;</td>
    <td>&a.marius;</td>
    <td>sparc64 instability when accessing /dev/mem.  Contact
      &a.marius; or &a.kris; for debugging information.</td>
  </tr>

  <tr>
    <td>sparc64 frequent hangs</td>
    <td>&status.unknown;</td>
    <td></td>
    <td>no DDB break possible, so impossible to diagnose</td>
  </tr>

  <tr>
    <td>serious sparc64 IPv6 panic</td>
    <td>&status.wip;</td>
    <td>&a.gnn;</td>
    <td>Triggered by just ping6'ing the box.  It may even be a MI
      issue, the reporter of this bug only uses IPv6 with
      sparc64.  This problem seems to be triggered even when debug.mpsafenet=0.</td>
  </tr>

  <tr>
    <td>sort(1) does not work with some locales</td>
    <td>&status.new;</td>
    <td>&nbsp;</td>
    <td>sort(1) can cause a coredump with some locales.
       See also gnu/93629.</td>
  </tr>

  <tr>
    <td>exec_map depletion</td>
    <td>&status.wip;</td>
    <td>&a.ups;</td>
    <td>The exec_map is regularly running out of space
       on machines running 7.0.  &a.ups; has a proposed
       patch and it seems to fix this problem.</td>
  </tr>

  <tr>
    <td>NFS data corruption between two 7.0 machines</td>
    <td>&status.wip;</td>
    <td>&a.mohans;</td>
    <td>Running fsx between a 7.0 NFS client and server
       detects data corruption.  This problem can also be reproduced
       by using 6.1 NFS server.</td>
  </tr>

  <tr>
    <td>panic in fxp driver</td>
    <td>&status.wip;</td>
    <td>&a.andre;</td>
    <td>See <a href="http://people.freebsd.org/~pho/stress/log/cons186.html">http://people.freebsd.org/~pho/stress/log/cons186.html</a>.</td>
  </tr>

  <tr>
    <td>deadlock in vn_start_write() consumers</td>
    <td>&status.wip;</td>
    <td>&a.tegge;</td>
    <td></td>
  </tr>

  <tr>
    <td>panic in bpf</td>
    <td>&status.wip;</td>
    <td>&a.sam;</td>
    <td>killing tcpdump (e.g. with ^C) can cause panics in bpf.</td>
  </tr>

  <tr>
    <td>devfs locking problem</td>
    <td>&status.wip;</td>
    <td>&a.jeff;</td>
    <td>It is trivial to deadlock it on an SMP system, and
      there are other panics with device removal.</td>
  </tr>

  <tr>
    <td>pty leak</td>
    <td>&status.wip;</td>
    <td>&a.cognet;</td>
    <td>Since 6.x has a hard-coded limit, once all ptys are
      leaked things like ssh and login no longer work.
      This seems devfs-related.</td>
  </tr>

  <tr>
    <td>cpu_ipi_selected() can cause a trap on FreeBSD/sparc64</td>
    <td>&status.wip;</td>
    <td>&a.kmacy;</td>
    <td>On sparc64, cpu_ipi_selected() can cause a trap (which is bad since
      it appears in the trap code path).</td>
  </tr>

  <tr>
    <td>rpc.lockd interoperability problems</td>
    <td>&status.unknown;</td>
    <td>&a.kuriyama;</td>
    <td>After <a href="http://www.freebsd.org/cgi/query-pr.cgi?pr=bin/80389">this commit</a>
      rpc.lockd seems to have interoperability problems.
      This may be the cause of the many "rpc.lockd no longer interoperates"
      bug reports seen on -stable.</td>
  </tr>

  <tr>
    <td>dup(2) regression on 6.x</td>
    <td>&status.unknown;</td>
    <td>&a.des;</td>
    <td>Simple "close(0); dup(fd)" does not return descriptor "0" in some cases.
      This problem has been reported in
      <a href="http://www.freebsd.org/cgi/query-pr.cgi?pr=kern/87208">kern/87208</a>,
      and there is a proposed patch in the PR, too.</td>
  </tr>

  <tr>
    <td>ifconfig regression on 6.x</td>
    <td>&status.wip;</td>
    <td>&a.yar;</td>
    <td>ifconfig cannot handle vlan and mtu parameters at the same time
      after rev.1.7.2.3 of sbin/ifconfig/ifvlan.c commit.
      For more information and a proposed patch, see
      <a href="http://www.freebsd.org/cgi/query-pr.cgi?pr=bin/94028">bin/94028</a>.</td>
  </tr>

  <tr>
    <td>"calcru: runtime went backwards" problem for threaded program</td>
    <td>&status.unknown;</td>
    <td>&nbsp;</td>
    <td>stress2 thr1 test can trigger "calcru: runtime went backwards" problem
      and there are also many similar reports on -stable and -current.
      &a.phk; committed a possible fix (src/sys/kern/kern_tc.c rev.1.169) to
      update the calibration code to be more precise on 2 March.</td>
  </tr>
</table>

<h3>Required features for &local.rel;-RELEASE</h3>

<a name="required"></a>
<table class="tblbasic">
  <tr class="heading">
    <th>Issue</th>
    <th>Status</th>
    <th>Responsible</th>
    <th>Description</th>
  </tr>

</table>

<h3>Desired features for &local.rel;-RELEASE</h3>

<a name="desired"></a>
<table class="tblbasic">
  <tr class="heading">
    <th>Issue</th>
    <th>Status</th>
    <th>Responsible</th>
    <th>Description</th>
  </tr>

  <tr>
    <td>SMP kernels for install</td>
    <td>&status.unknown;</td>
    <td></td>

    <td><em>From the <a
      href="http://www.freebsd.org/projects/ideas/#p-smpinstall">ideas
      page</a>.</em>  Right now we only install a UP kernel, for performance
      reasons. We should be able to package both a UP and SMP kernel
      into the release bits, and have sysinstall install both. It
      should also select the correct one for the target system and
      make that the default on boot. The easiest way to do this would
      be to have sysinstall boot an SMP kernel and then look at the
      hw.ncpu sysctl. The only problem is being able to have
      sysinstall fall back to booting a UP kernel for itself if the
      SMP one fails. This can probably be 'faked' by setting one of
      the SMP-disabling variables in the loader. But in any case, the
      point is to make the process Just Work for the user, without the
      user needing to know arcane loader/sysctl knobs. SMP laptops are
      here, and we should be ready to support SMP out-of-the-box.</td>
  </tr>

  <tr>
    <td>swap panic on sparc64</td>
    <td>&status.unknown;</td>
    <td>&a.kris; has panic info</td>

    <td>&a.kris; reports configuring a 74GB swap-backed md on sparc64 that
      caused a panic after a week or two of load (during which time
      swap was slowly filling as more of the md was dirtied).</td>
  </tr>

  <tr>
    <td>updated hal and ath drivers</td>
    <td>&status.new;</td>
    <td>&a.sam;</td>
    <td></td>
  </tr>

  <tr>
    <td>fix ntpdate(1) bogus output on amd64.</td>
    <td>&status.unknown;</td>
    <td>&a.roberto;</td>
    <td></td>
  </tr>

  <tr>
    <td>Improve performance</td>
    <td>&status.unknown;</td>
    <td></td>
    <td>What seem to be 4BSD scheduler bugs in 6.0 that
      cause performance to be anomalously low in certain situations.
      &a.davidxu; has expressed some interest in this problem.</td>
  </tr>

  <tr>
    <td>/dev/kmem panic</td>
    <td>&status.wip;</td>
    <td>&a.ups;</td>
    <td>&a.kris; has noticed panics on SMP machines when there was ABI
      breakage of libkvm and world was not rebuilt and utilities like
      fstat were used.  This suggests panics can be caused by incorrect
      accesses to /dev/kmem.  &a.ups; committed a fix for i386.</td>
  </tr>

  <tr>
    <td>KLDs on sparc64</td>
    <td>&status.new;</td>
    <td>&nbsp;</td>
    <td>On sparc64 machines with more than 4Gb memory KLDs are not usable
      and will panic the system.  The problem is reportedly with how the
      KLDs are compiled, it only works if the code ends up below 4G.</td>
  </tr>

  <tr>
    <td>Max RAM on sparc64</td>
    <td>&status.new;</td>
    <td>&nbsp;</td>
    <td>Maximum RAM on sparc64 appears to be limited to 16Gb.</td>
  </tr>

  <tr>
    <td>make -jN</td>
    <td>&status.new;</td>
    <td>&nbsp;</td>
    <td>Doing 'make -jN', then suspending/resuming it may result in make
      reporting it lost child process(es).</td>
  </tr>

  <tr>
    <td>OpenBSM</td>
    <td>&status.unknown;</td>
    <td>&a.rwatson;</td>
    <td>The integration of OpenBSM is waiting on some final licensing hurdles.
      Once those are cleared, it will be a very desirable feature for &local.rel;.</td>
  </tr>

  <tr>
    <td>update sysinstall disk labeling</td>
    <td>&status.wip;</td>
    <td>&a.rodrigc;</td>
    <td>Sysinstall could use the same fixes recently made to fdisk so it
      plays nice with GEOM and disk labeling.  This does not cause problems
      during install because nothing on the disk is mounted when its label
      is being manipulated but it can cause problems if sysinstall gets
      used on a live system to adjust labels on existing disks which
      sys-admins tend to do.</td>
  </tr>

  <tr>
    <td>swapping on 6.0 is slower than on 4.x</td>
    <td>&status.new;</td>
    <td>&nbsp;</td>
    <td>Performance on swap handling is much slower than 4.x and
      this can make a system essentially unusable when moderate
      paging activity is going on.</td>
  </tr>
</table>

<h3>Documentation items that must be resolved for &local.rel;</h3>

<a name="docs"></a>
<table class="tblbasic">
  <tr class="heading">
    <th>Issue</th>
    <th>Status</th>
    <th>Responsible</th>
    <th>Description</th>
  </tr>

</table>

<h3>Testing foci for &local.rel;-RELEASE</h3>

<a name="testing"></a>
<table class="tblbasic">
  <tr class="heading">
    <th>Issue</th>
    <th>Status</th>
    <th>Responsible</th>
    <th>Description</th>
  </tr>

  <tr>
    <td>manual root mount lockmgr panics</td>
    <td>&status.untested;</td>
    <td>&a.ssouhlal;</td>
    <td>Specifying a manual root mount location causes lockmgr panics.
      &a.ssouhlal; has committed a patch for this.</td>
  </tr>

  <tr>
    <td>dhclient causes ipv6 panics.</td>
    <td>&status.untested;</td>
    <td>&a.dougb;</td>
    <td>&a.dougb; has more details about this.</td>
  </tr>

  <tr>
    <td>amd64 panics in ipv6 with date(1)</td>
    <td>&status.untested;</td>
    <td>&a.ume;</td>
    <td>amd64 panics in ipv6 when the date is changed using date(1) or
	ntpdate(1).  This may be a MI issue.</td>
  </tr>

  <tr>
    <td>grep(1) -w does not work with multibyte locales</td>
    <td>&status.untested;</td>
    <td>&a.tjr;</td>
    <td>grep(1) -w generates wrong results with non-UTF-8
       multibyte locales.  &a.tjr; has committed a patch
       to -HEAD.  See also gnu/91909.</td>
  </tr>

  <tr>
    <td>Improve kbdmux</td>
    <td>&status.wip;</td>
    <td>&a.emax;</td>
    <td><em>From the <a
      href="http://www.freebsd.org/projects/ideas/#p-kbdmux">ideas
      page</a>.</em> We need this for the growing number of systems
      that assume that USB is the primary keyboard. Current status
      appears to be that the kbdmux driver breaks very easily. We need
      this working well enough where it can be enabled by default, and
      all attached keyboards Just Work. &a.emax; commit kbdmux and
      rc.d/syscons patches in HEAD and RELENG_6. It is not yet enabled
      by default. See kbdmux(4) and contact &a.emax; if you have
      problems.</td>
  </tr>
</table>

<h3>Stress Test Panics</h3>

<a name="stresstest"></a>
<p>The system is continuously being subjected to Peter Holm's <a
  href="http://www.holm.cc/stress/">Kernel Stress Test Suite</a>.  The
  following issues have recently been discovered from this test
  suite.</p>

&stresstest;

    &footer;

  </body>
</html>




<?xml version="1.0"?>
<!DOCTYPE entries SYSTEM "todo.dtd">

<!--

  Changes to this list MUST NOT be committed without approval of
  Release Engineering Team (re@FreeBSD.org) (for general items) or
  Documentation Engineering Team (doceng@FreeBSD.org) (for doc-related
  items).

  available classes: < showstopper | required | desired | docs |
	testing >
  available states: < na | done | wip | untested | new | unknown |
	deferred >

-->

<entries>
  <cvs:keywords xmlns:cvs="http://www.FreeBSD.org/XML/CVS" version="1.0">
    <cvs:keyword name="freebsd">
      $FreeBSD:$
    </cvs:keyword>
  </cvs:keywords>

  <entry class="showstopper" status="unknown">
    <issue>Unreliable serial console</issue>
    <description>
      At the manual 'root mount' prompt, the serial console is very
      unreliable and drops most characters.
    </description>
  </entry>

  <entry class="testing" status="untested">
    <issue>Manual root mount lockmgr panics</issue>
    <responsible>Suleiman Souhlal</responsible>
    <description>
      Specifying a manual root mount location causes lockmgr panics.
      ssouhlal@ has committed a patch for this.
    </description>
  </entry>

  <entry class="showstopper" status="deferred">
    <issue>i386 deadlocks with more than 16GB swap</issue>
    <responsible>Alan L. Cox</responsible>
    <description>
      i386 deadlocks if more than 16GB of swap is in use.  Increasing
      the kern.maxswzone tunable would be a workaround this, but a patch
      from alc@ is needed to allow this variable to be increased.
    </description>
  </entry>

  <entry class="showstopper" status="wip">
    <issue>Unmount pending error</issue>
    <responsible>Suleiman Souhlal</responsible>
    <description>
      When unmounting filesystems Kris Kennaway reports seeing this
      warning: <tt>/c: unmount pending error: blocks -68512 files 0</tt>.
      This dates back at least to 5.3.  It might be associated with
      filesystem corruption reported by many users in which the 'used'
      space on a filesystem is negative; fsck -f is needed to correct
      this.
    </description>
  </entry>

  <entry class="showstopper" status="unknown">
    <issue>swap_pager warnings</issue>
    <responsible>Don Lewis?</responsible>
    <description>
      When swapfiles are in use, there are often warnings printed: <tt>
      swap_pager: indefinite wait buffer: bufobj: 0, blkno: 889347,
      size: 8192</tt>.  There is also the possibility of deadlock.
    </description>
  </entry>

  <entry class="showstopper" status="wip">
    <issue>umount -f panics</issue>
    <responsible>Jeff Roberson, Suleiman Souhlal</responsible>
    <description>
      Panics from race conditions.  A patch from Jeff Roberson seems to
      fix some of them.
    </description>
  </entry>

  <entry class="showstopper" status="unknown">
    <issue>UFS deadlocks on amd64</issue>
    <responsible>Tor Egge</responsible>
    <description>
      Seen by Kris Kennaway.
    </description>
  </entry>

  <entry class="showstopper" status="unknown">
    <issue>UFS deadlocks</issue>
    <description>
      Seen by Peter Jeremy.
    </description>
  </entry>

  <entry class="testing" status="untested">
    <issue>amd64 panics in ipv6 with date(1)</issue>
    <responsible>Hajimu UMEMOTO</responsible>
    <description>
      amd64 panics in ipv6 when the date is changed using date(1) or
      ntpdate(1).  This may be a MI issue.
    </description>
  </entry>

  <entry class="showstopper" status="unknown">
    <issue>sparc64 instability.</issue>
    <responsible>Marius Strobl</responsible>
    <description>
      sparc64 instability when accessing /dev/mem.  Contact marius@ or
      kris@ for debugging information.
    </description>
  </entry>

  <entry class="testing" status="untested">
    <issue>dhclient causes ipv6 panics.</issue>
    <responsible>Doug Barton</responsible>
    <description>
      dougb@ has more details about this.
    </description>
  </entry>

  <entry class="showstopper" status="unknown">
    <issue>sparc64 frequent hangs</issue>
    <status>unknown</status>
    <description>
      No DDB break possible, so impossible to diagnose.
    </description>
  </entry>

  <entry class="showstopper" status="unknown">
    <issue>Serious sparc64 IPv6 panic</issue>
    <responsible>George V. Neville-Neil</responsible>
    <description>
      Triggered by just ping6'ing the box.  It may even be a MI issue,
      the reporter of this bug only uses IPv6 with sparc64. This problem
      seems to be triggered even when debug.mpsafenet=0.
    </description>
  </entry>

  <entry class="desired" status="unknown">
    <issue>SMP kernels for install</issue>
    <description>
      <em>From the <a
	href="http://www.freebsd.org/projects/ideas/#p-smpinstall">ideas
      page</a>.</em>  Right now we only install a UP kernel, for
      performance reasons.  We should be able to package both a UP and
      SMP kernel into the release bits, and have sysinstall install
      both.  It should also select the correct one for the target system
      and make that the default on boot.  The easiest way to do this
      wouldbe to have sysinstall boot an SMP kernel and then look at the
      hw.ncpu sysctl.  The only problem is being able to have sysinstall
      fall back to booting a UP kernel for itself if the SMP one fails.
      This can probably be 'faked' by setting one of the SMP-disabling
      variables in the loader.  But in any case, the point is to make
      the process Just Work for the user, without the user needing to
      know arcane loader/sysctl knobs.  SMP laptops are here, and we
      should be ready to support SMP out-of-the-box.
    </description>
  </entry>

  <entry class="testing" status="untested">
    <issue>Improve kbdmux</issue>
    <responsible>Maksim Yevmenkin</responsible>
    <description>
      <em>From the <a
	href="http://www.freebsd.org/projects/ideas/#p-kbdmux">ideas
      page</a>.</em>  We need this for the growing number of systems
      that assume that USB is the primary keyboard.  Current status
      appears to be that the kbdmux driver breaks very easily.  We need
      this working well enough where it can be enabled by default, and
      all attached keyboards Just Work.
    </description>
  </entry>

  <entry class="desired" status="unknown">
    <issue>swap panic on sparc64</issue>
    <responsible>Kris Kennaway has panic info</responsible>
    <description>
      Kris reports configuring a 74GB swap-backed md on sparc64 that
      caused a panic after a week or two of load (during which time swap
      was slowly filling as more of the md was dirtied).
    </description>
  </entry>

  <entry class="desired" status="new">
    <issue>Updated hal and ath drivers</issue>
    <responsible>Sam Leffler</responsible>
  </entry>

  <entry class="desired" status="unknown">
    <issue>Fix ntpdate(1) bogus output on amd64.</issue>
    <responsible>Ollivier Robert</responsible>
  </entry>

  <entry class="desired" status="unknown">
    <issue>Improve performance</issue>
    <description>
      What seem to be 4BSD scheduler bugs in 6.0 that cause performance
      to be anomalously low in certain situations.  davidxu@ has
      expressed some interest in this problem.
    </description>
  </entry>

  <entry class="desired" status="wip">
    <issue>/dev/kmem panic</issue>
    <responsible>Stephan Uphoff</responsible>
    <description>
      Kris has noticed panics on SMP machines when there was ABI
      breakage of libkvm and world was not rebuilt and utilities like
      fstat were used.  This suggests panics can be caused by incorrect
      accesses to /dev/kmem.  Stephan Uphoff committed a fix for i386.
    </description>
  </entry>

  <entry class="desired" status="new">
    <issue>KLDs on sparc64</issue>
    <description>
      On sparc64 machines with more than 4Gb memory KLDs are not usable
      and will panic the system.  The problem is reportedly with how the
      KLDs are compiled, it only works if the code ends up below 4G.
    </description>
  </entry>

  <entry class="desired" status="new">
    <issue>Max RAM on sparc64</issue>
    <description>
      Maximum RAM on sparc64 appears to be limited to 16Gb.
    </description>
  </entry>

  <entry class="desired" status="new">
    <issue>make -jN</issue>
    <description>
      Doing 'make -jN', then suspending/resuming it may result in make
      reporting it lost child process(es).
    </description>
  </entry>

  <entry class="testing" status="untested">
    <issue>OpenBSM</issue>
    <responsible>Robert Watson</responsible>
    <description>
      The OpenBSM has been MFC'ed to RELENG_6 and needs some testing.
    </description>
  </entry>

  <entry class="desired" status="wip">
    <issue>Update sysinstall disk labeling</issue>
    <responsible>Craig Rodrigues</responsible>
    <description>
      Sysinstall could use the same fixes recently made to fdisk so it
      plays nice with GEOM and disk labeling.  This does not cause
      problems during install because nothing on the disk is mounted
      when its label is being manipulated but it can cause problems if
      sysinstall gets used on a live system to adjust labels on existing
      disks which sys-admins tend to do.
    </description>
  </entry>

  <entry class="showstopper" status="wip">
    <issue>Quota deadlocks</issue>
    <responsible>Jeff Roberson</responsible>
    <description>
      Quota support is not locked properly and causes deadlocks.  A
      patch from Jeff Roberson seems to fix some of them.
    </description>
  </entry>

  <entry class="showstopper" status="new">
    <issue>sort(1) does not work with some locales</issue>
    <description>
      sort(1) can cause a coredump with some locales.  See also
      gnu/93629.
    </description>
  </entry>

  <entry class="showstopper" status="wip">
    <issue>exec_map depletion</issue>
    <responsible>Stephan Uphoff</responsible>
    <description>
      The exec_map is regularly running out of space on machines running
      7.0.  Stephan Uphoff has a proposed patch and it seems to fix this
      problem.
    </description>
  </entry>

  <entry class="showstopper" status="wip">
    <issue>NFS data corruption between two 7.0 machines</issue>
    <responsible>Mohan Srinivasan</responsible>
    <description>
      Running fsx between a 7.0 NFS client and server detects data
      corruption.  This problem can also be reproduced by using 6.1 NFS
      server.
    </description>
  </entry>

  <entry class="showstopper" status="wip">
    <issue>Panic in fxp driver</issue>
    <responsible>Andre Oppermann</responsible>
    <description>
      See <a href="http://people.freebsd.org/~pho/stress/log/cons186.html">
	http://people.freebsd.org/~pho/stress/log/cons186.html</a>.
    </description>
  </entry>

  <entry class="showstopper" status="wip">
    <issue>Deadlock in vn_start_write() consumers</issue>
    <responsible>Tor Egge</responsible>
  </entry>

  <entry class="showstopper" status="wip">
    <issue>panic in bpf</issue>
    <responsible>Sam Leffler</responsible>
    <description>
      Killing tcpdump (e.g. with ^C) can cause panics in bpf.
    </description>
  </entry>

  <entry class="showstopper" status="wip">
    <issue>devfs locking problem</issue>
    <responsible>Jeff Roberson</responsible>
    <description>
      It is trivial to deadlock it on an SMP system, and there are other
      panics with device removal.
    </description>
  </entry>

  <entry class="showstopper" status="wip">
    <issue>pty leak</issue>
    <responsible>Olivier Houchard</responsible>
    <description>
      Since 6.x has a hard-coded limit, once all ptys are leaked things
      like ssh and login no longer work.  This seems to be related to
      devfs.
    </description>
  </entry>  

  <entry class="showstopper" status="wip">
    <issue>cpu_ipi_selected() can cause a trap on sparc64</issue>
    <responsible>Kip Macy</responsible>
    <description>
      On sparc64, cpu_ipi_selected() can cause a trap (which is bad
      since it appears in the trap code path).
    </description>
  </entry>

  <entry class="showstopper" status="unknown">
    <issue>rpc.lockd interoperability problems</issue>
    <responsible>Jun Kuriyama</responsible>
    <description>
      After <a href="http://www.freebsd.org/cgi/query-pr.cgi?pr=bin/80389">
	this commit</a> rpc.lockd seems to have interoperability
      problems.  This may be the cause of the many "rpc.lockd no longer
      interoperates" bug reports seen on -stable.
    </description>
  </entry>

  <entry class="showstopper" status="wip">
    <issue>dup(2) regression on 6.x</issue>
    <responsible>Dag-Erling Smorgrav</responsible>
    <description>
      Simple "close(0); dup(fd)" does not return descriptor "0" in some
      cases.  This problem has been reported in kern/87208, and there is
      a proposed patch in the PR, too.
    </description>
  </entry>

  <entry class="showstopper" status="wip">
    <issue>ifconfig regression on 6.x</issue>
    <responsible>Yar Tikhiy</responsible>
    <description>
      ifconfig can not handle vlan and mtu parameters at the same time
      after rev.1.7.2.3 of sbin/ifconfig/ifvlan.c commit.  For more
      information and a proposed patch, see bin/94028.
    </description>
  </entry>

  <entry class="showstopper" status="unknown">
    <issue>"calcru: runtime went backwards" problem for threaded
      program</issue>
    <description>
      stress2 thr1 test can trigger "calcru: runtime went backwards"
      problem and there are also many similar reports on -stable and
      -current.  Poul-Henning Kamp committed a possible fix
      (src/sys/kern/kern_tc.c rev.1.169) to update the calibration code
      to be more precise on 2 March.
    </description>
  </entry>

  <entry class="desired" status="new">
    <issue>Swapping on 6.0 is slower than on 4.x</issue>
    <description>
      Performance on swap handling is much slower than 4.x and this can
      make a system essentially unusable when moderate paging activity
      is going on.
    </description>
  </entry>
</entries>




<?xml version="1.0"?>

<!-- Copyright (c) 2006 Daniel Gerzo <danger@rulez.sk>
     All rights reserved.

     Redistribution and use in source and binary forms, with or without
     modification, are permitted provided that the following conditions
     are met:
     1. Redistributions of source code must retain the above copyright
	notice, this list of conditions and the following disclaimer.
     2. Redistributions in binary form must reproduce the above copyright
	notice, this list of conditions and the following disclaimer in the
	documentation and/or other materials provided with the distribution.

     THIS SOFTWARE IS PROVIDED BY THE AUTHOR AND CONTRIBUTORS ``AS IS'' AND
     ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
     IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
     ARE DISCLAIMED.  IN NO EVENT SHALL THE AUTHOR OR CONTRIBUTORS BE LIABLE
     FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
     DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
     OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
     HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
     LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
     OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
     SUCH DAMAGE.

     $FreeBSD:$
-->

<xsl:stylesheet xmlns:xsl="http://www.w3.org/1999/XSL/Transform" version="1.0"
  xmlns:cvs="http://www.FreeBSD.org/XML/CVS"
  exclude-result-prefixes="cvs">

  <xsl:import href="../../includes.xsl"/>
  <xsl:variable name="section" select="'download'"/>
  <xsl:variable name="base" select="'../..'"/>
  <xsl:variable name="date">
    <xsl:value-of select="//cvs:keyword[@name='freebsd']"/>
  </xsl:variable>
  <xsl:variable name="email" select="'re'"/>
  <xsl:variable name="localrel" select="'6.1'"/>
  <xsl:variable name="title" select="concat('FreeBSD ', $localrel, '-RELEASE Open Issues')"/>
  <xsl:variable name="upperCase" select="'ABCDEFGHIJKLMNOPQRSTUVWXYZ'"/>
  <xsl:variable name="lowerCase" select="'abcdefghijklmnopqrstuvwxyz'"/>

  <xsl:output method="xml" encoding="iso-8859-1"
    doctype-system="http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd"
    doctype-public="-//W3C//DTD XHTML 1.0 Transitional//EN"/>

  <xsl:template match="entries">
    <html>

      <xsl:copy-of select="$header1"/>

      <body>

        <div id="containerwrap">
          <div id="container">

            <xsl:copy-of select="$header2"/>

            <div id="content">

              <xsl:copy-of select="$sidenav"/>

              <div id="contentwrap">

                <xsl:copy-of select="$header3"/>

	<p>This is a list of open issues that need to be resolved for
	  FreeBSD <xsl:value-of select="$localrel" />-RELEASE. If you
	  have any updates for this list, please e-mail
	  re@FreeBSD.org.</p>

	<ul>
	  <li><a href="#showstopper">Show stopper defects</a></li>
	  <li><a href="#required">Required features</a></li>
	  <li><a href="#desired">Desired features</a></li>
	  <li><a href="#docs">Documentation Items</a></li>
	  <li><a href="#testing">Testing focuses</a></li>
	  <li><a href="#stresstest">Problems Discovered by Kernel Stress Test Suite</a></li>
	</ul>

	<h3>Show stopper defects for <xsl:value-of select="$localrel"/>-RELEASE</h3>
	<a name="showstopper"></a>
	<table class="tblbasic">
	  <tr class="heading">
	    <th>Issue</th>
	    <th>Status</th>
	    <th>Responsible</th>
	    <th>Description</th>
	  </tr>
	  <xsl:for-each select="entry[@class='showstopper']">
	    <xsl:sort select="translate(./issue, $upperCase, $lowerCase)"/>
	    <xsl:call-template name="entry"/>
	  </xsl:for-each>
	</table>

	<h3>Required features for <xsl:value-of select="$localrel"/>-RELEASE</h3>
	<a name="required"></a>
	<table class="tblbasic">
	  <tr class="heading">
	    <th>Issue</th>
	    <th>Status</th>
	    <th>Responsible</th>
	    <th>Description</th>
	  </tr>
	  <xsl:for-each select="entry[@class='required']">
	    <xsl:sort select="translate(./issue, $upperCase, $lowerCase)"/>
	    <xsl:call-template name="entry"/>
	  </xsl:for-each>
	</table>

	<h3>Desired features for <xsl:value-of select="$localrel"/>-RELEASE</h3>
	<a name="desired"></a>
	<table class="tblbasic">
	  <tr class="heading">
	    <th>Issue</th>
	    <th>Status</th>
	    <th>Responsible</th>
	    <th>Description</th>
	  </tr>
	  <xsl:for-each select="entry[@class='desired']">
	    <xsl:sort select="translate(./issue, $upperCase, $lowerCase)"/>
	    <xsl:call-template name="entry"/>
	  </xsl:for-each>
	</table>

	<h3>Documentation items that must be resolved for <xsl:value-of select="$localrel"/>-RELEASE</h3>
	<a name="docs"></a>
	<table class="tblbasic">
	  <tr class="heading">
	    <th>Issue</th>
	    <th>Status</th>
	    <th>Responsible</th>
	    <th>Description</th>
	  </tr>
	  <xsl:for-each select="entry[@class='docs']">
	    <xsl:sort select="translate(./issue, $upperCase, $lowerCase)"/>
	    <xsl:call-template name="entry"/>
	  </xsl:for-each>
	</table>

	<h3>Testing focuses for <xsl:value-of select="$localrel"/>-RELEASE</h3>
	<a name="testing"></a>
	<table class="tblbasic">
	  <tr class="heading">
	    <th>Issue</th>
	    <th>Status</th>
	    <th>Responsible</th>
	    <th>Description</th>
	  </tr>
	  <xsl:for-each select="entry[@class='testing']">
	    <xsl:sort select="translate(./issue, $upperCase, $lowerCase)"/>
	    <xsl:call-template name="entry"/>
	  </xsl:for-each>
	</table>

	<h3>Stress Test Panics</h3>
	<a name="stresstest"></a>
	<p>The system is continuously being subjected to Peter Holm's <a
	  href="http://www.holm.cc/stress/">Kernel Stress Test Suite</a>.
	  The following issues have recently been discovered from this
	  test suite.</p>

	<ul>
	  <xsl:for-each select="document('stress.xml')/results/result">
	  <li>
	    <xsl:element name="a">
	      <xsl:attribute name="href"><xsl:value-of select="link"/></xsl:attribute>
	      <xsl:value-of select="title"/>
	    </xsl:element>
	  </li>
	  </xsl:for-each>
	</ul>

	</div> <!-- contentwrap -->
	<br class="clearboth" />

	</div> <!-- content -->

	<xsl:copy-of select="$footer"/>

	</div> <!-- container -->
      </div> <!-- containerwrap -->

      </body>
    </html>
  </xsl:template>

  <xsl:template name="entry">
    <tr>
      <td><xsl:value-of select="issue"/></td>
      <xsl:choose>
        <xsl:when test="@status='na'">
	  <td><font color="green">N/A</font></td>
	</xsl:when>
	<xsl:when test="@status='done'">
	  <td><font color="green">Done</font></td>
	</xsl:when>
	<xsl:when test="@status='wip'">
	  <td><font color="blue">In progress</font></td>
	</xsl:when>
	<xsl:when test="@status='untested'">
	  <td><font color="orange">Needs testing</font></td>
	</xsl:when>
	<xsl:when test="@status='new'">
	  <td><font color="red">Not done</font></td>
	</xsl:when>
	<xsl:when test="@status='unknown'">
	  <td><font color="red">Unknown</font></td>
	</xsl:when>
	<xsl:when test="@status='deferred'">
	  <td><font color="gray">Deferred for future release</font></td>
	</xsl:when>
      </xsl:choose>
      <td><xsl:value-of select="responsible"/></td>
      <td><xsl:copy-of select="description/child::node()"/></td>
    </tr>
  </xsl:template>
</xsl:stylesheet>

Return to bug 94423

Lines 7-25 Link Here

(-)6.1R.new/Makefile (-6 / +8 lines)
7	.include "../Makefile.inc"	7	.include "../Makefile.inc"
8	.endif	8	.endif
9		9
10	DOCS= todo.sgml approvals.sgml	10	DOCS= approvals.sgml schedule.sgml
11	DOCS+= schedule.sgml	11
		12	SRCS.DEFAULT= todo.xsl
		13	XMLDOCS= todo
12		14
13	FETCHFILE= http://people.freebsd.org/~pho/stress/log/news.rdf	15	FETCHFILE= http://people.freebsd.org/~pho/stress/log/news.rdf
14	FETCHCMD= /usr/bin/fetch	16	FETCHCMD= /usr/bin/fetch
15	CLEANFILES+= news.rdf stress.html	17	CLEANFILES+= news.rdf stress.xml
16		18
17	news.rdf:	19	news.rdf:
18	${FETCHCMD} ${FETCHFILE}	20	${FETCHCMD} ${FETCHFILE}
19		21
20	stress.html: news.rdf stress.xsl	22	stress.xml: news.rdf stress.xsl
21	${XSLTPROC} stress.xsl news.rdf > stress.html	23	${XSLTPROC} stress.xsl news.rdf > stress.xml
22		24
23	todo.html: stress.html	25	todo.html: stress.xml
24		26
25	.include "${WEB_PREFIX}/share/mk/web.site.mk"	27	.include "${WEB_PREFIX}/share/mk/web.site.mk"

Lines 7-31 Link Here

(-)6.1R.new/stress.xsl (-13 / +10 lines)
7	version="1.0">	7	version="1.0">
8		8
9	<xsl:output	9	<xsl:output
10	method="html"	10	method="xml"
11	indent="no"	11	indent="yes"
12	encoding="utf-8"/>	12	encoding="utf-8"
		13	omit-xml-declaration="no"/>
13		14
14	<!-- match first element whether we're using namespaces or not -->	15	<!-- match first element whether we're using namespaces or not -->
15		16
16	<xsl:template match="/*[1]">	17	<xsl:template match="/*[1]">
17	<xsl:comment>Generated from XSLT</xsl:comment>	18	<results>
18	<ul>
19	<xsl:for-each select="*[local-name()='item']">	19	<xsl:for-each select="*[local-name()='item']">
20	<li>	20	<result>
21	<xsl:element name="a">	21	<title><xsl:value-of select="*[local-name()='title']"/></title>
22	<xsl:attribute name="href"><xsl:value-of select="*[local-name()='link']"/></xsl:attribute>	22	<link><xsl:value-of select="*[local-name()='link']"/></link>
23	<xsl:value-of select="*[local-name()='title']"/>	23	</result>
24	</xsl:element>
25	</li>
26	</xsl:for-each>	24	</xsl:for-each>
27	</ul>	25	</results>
28
29	</xsl:template>	26	</xsl:template>
30		27
31	</xsl:stylesheet>	28	</xsl:stylesheet>

Line 0 Link Here

(-)6.1R.new/todo.dtd (+37 lines)
	1	<!-- DTD for todo list entries on the FreeBSD website. -->
	2	<!-- $FreeBSD:$ -->
	3
	4	<!ELEMENT entries (cvs:keywords?, entry+)>
	5
	6	<!ELEMENT entry (issue, responsible?, description?)>
	7	<!ATTLIST entry class (showstopper \| required \| desired \| docs
	8	\| testing) #REQUIRED>
	9	<!ATTLIST entry status (na \| done \| wip \| untested \| new \| unknown
	10	\| deferred) #REQUIRED>
	11
	12	<!ELEMENT issue (#PCDATA)>
	13	<!ELEMENT responsible (#PCDATA)>
	14	<!ELEMENT description (#PCDATA\|a\|b\|br\|i\|tt\|em\|p\|strong\|ul\|pre)*>
	15
	16	<!-- misc HTML tags -->
	17	<!ELEMENT a (#PCDATA\|img)*>
	18	<!ATTLIST a href CDATA #REQUIRED>
	19
	20	<!ELEMENT b (#PCDATA\|a)*>
	21	<!ELEMENT br (#PCDATA\|a)*>
	22	<!ELEMENT i (#PCDATA\|a)*>
	23	<!ELEMENT tt (#PCDATA\|a)*>
	24	<!ELEMENT em (#PCDATA\|a)*>
	25	<!ELEMENT p (#PCDATA\|a)*>
	26	<!ELEMENT strong (#PCDATA\|a)*>
	27	<!ELEMENT ul (li)*>
	28	<!ELEMENT li (#PCDATA\|a)*>
	29	<!ELEMENT pre (#PCDATA\|a)*>
	30
	31	<!-- common cvs tags -->
	32	<!ELEMENT cvs:keywords (cvs:keyword+)>
	33	<!ATTLIST cvs:keywords xmlns:cvs CDATA #FIXED 'http://www.FreeBSD.org/XML/CVS'>
	34	<!ATTLIST cvs:keywords version CDATA #FIXED '1.0'>
	35
	36	<!ELEMENT cvs:keyword (#PCDATA)>
	37	<!ATTLIST cvs:keyword name CDATA #REQUIRED>

Lines 1-494 Link Here

(-)6.1R.new/todo.sgml (-494 lines)
1	<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN" [
2	<!ENTITY base CDATA "../..">
3	<!ENTITY email 'freebsd-qa'>
4	<!ENTITY date "$FreeBSD: www/en/releases/6.1R/todo.sgml,v 1.18 2006/03/08 03:48:13 hrs Exp $">
5	<!ENTITY local.rel "6.1">
6	<!ENTITY title "FreeBSD 6.1 Open Issues">
7	<!ENTITY % navincludes SYSTEM "../../includes.navdownload.sgml"> %navincludes;
8	<!ENTITY % includes SYSTEM "../../includes.sgml"> %includes;
9	<!ENTITY % developers SYSTEM "../../developers.sgml"> %developers;
10	<!-- Status levels -->
11	<!ENTITY status.na "<font color=green>N/A</font>">
12	<!ENTITY status.done "<font color=green>Done</font>">
13	<!ENTITY status.wip "<font color=blue>In progress</font>">
14	<!ENTITY status.untested "<font color=orange>Needs testing</font>">
15	<!ENTITY status.new "<font color=red>Not done</font>">
16	<!ENTITY status.unknown "<font color=red>Unknown</font>">
17	<!ENTITY status.deferred "<font color=gray>Deferred for future release</font>">
18
19	<!ENTITY url.cvsweb "http://www.freebsd.org/cgi/cvsweb.cgi">
20	<!ENTITY url.mid "http://docs.freebsd.org/cgi/mid.cgi?">
21	<!ENTITY url.pr "http://www.freebsd.org/cgi/query-pr.cgi?">
22
23	<!ENTITY stresstest SYSTEM "./stress.html">
24	]>
25
26	<!--
27
28	Changes to this list MUST NOT be committed without approval of
29	Release Engineering Team (re@FreeBSD.org) (for general items) or
30	Documentation Engineering Team (doceng@FreeBSD.org) (for doc-related
31	items).
32
33	-->
34
35	<html>
36	&header;
37
38	<p>This is a list of open issues that need to be resolved for FreeBSD
39	&local.rel;. If you have any updates for this list, please e-mail
40	re@FreeBSD.org.</p>
41
42	<ul>
43	<li><a href="#showstopper">Show stopper defects</a></li>
44	<li><a href="#required">Required features</a></li>
45	<li><a href="#desired">Desired features</a></li>
46	<li><a href="#docs">Documentation Items</a></li>
47	<li><a href="#testing">Testing foci</a></li>
48	<li><a href="#stresstest">Problems Discovered by Kernel Stress Test Suite</a></li>
49	</ul>
50
51	<h3>Show stopper defects for &local.rel;-RELEASE</h3>
52
53	<a name="showstopper"></a>
54	<table class="tblbasic">
55	<tr class="heading">
56	<th>Issue</th>
57	<th>Status</th>
58	<th>Responsible</th>
59	<th>Description</th>
60	</tr>
61
62	<tr>
63	<td>unreliable serial console</td>
64	<td>&status.unknown;</td>
65	<td></td>
66	<td>At the manual 'root mount' prompt, the serial console is very
67	unreliable and drops most characters.</td>
68	</tr>
69
70	<tr>
71	<td>i386 deadlocks with >16GB swap</td>
72	<td>&status.deferred;</td>
73	<td>&a.alc;</td>
74	<td>i386 deadlocks if more than 16GB of swap is in use.
75	Increasing the kern.maxswzone tunable would be a workaround
76	this. Although a patch from &a.alc; is needed to allow this variable to
77	be increased, this is not suitable for 6.1R. This limitation should
78	be documented in the Release Notes.</td>
79	</tr>
80
81	<tr>
82	<td>unmount pending error</td>
83	<td>&status.wip;</td>
84	<td>&a.ssouhlal;</td>
85	<td>When unmounting filesystems &a.kris; reports seeing this warning: <tt>/c: unmount pending error: blocks -68512 files 0</tt>. This dates back at least to 5.3. It might be associated with
86	filesystem corruption reported by many users in which the 'used' space
87	on a filesystem is negative; fsck -f is needed to correct this.</td>
88	</tr>
89
90	<tr>
91	<td>swap_pager warnings</td>
92	<td>&status.unknown;</td>
93	<td>&a.truckman;?</td>
94	<td>When swapfiles are in use, there are often warnings printed:
95	<tt>swap_pager: indefinite wait buffer: bufobj: 0, blkno: 889347, size: 8192</tt>. There is also the possibility of deadlock.</td>
96	</tr>
97
98	<tr>
99	<td>umount -f panics</td>
100	<td>&status.wip;</td>
101	<td>&a.jeff;, &a.ssouhlal;</td>
102	<td>panics from race conditions.
103	A patch from &a.jeff; seems to fix some of them.</td>
104	</tr>
105
106	<tr>
107	<td>quota deadlocks</td>
108	<td>&status.wip;</td>
109	<td>&a.jeff;</td>
110	<td>Quota support is not locked properly and causes deadlocks.
111	A patch from &a.jeff; seems to fix some of them.</td>
112	</tr>
113
114	<tr>
115	<td>UFS deadlocks on amd64</td>
116	<td>&status.unknown;</td>
117	<td>&a.tegge;</td>
118	<td>Seen by &a.kris;.</td>
119	</tr>
120
121	<tr>
122	<td>UFS deadlocks</td>
123	<td>&status.unknown;</td>
124	<td></td>
125	<td>Seen by Peter Jeremy.</td>
126	</tr>
127
128	<tr>
129	<td>sparc64 instability</td>
130	<td>&status.unknown;</td>
131	<td>&a.marius;</td>
132	<td>sparc64 instability when accessing /dev/mem. Contact
133	&a.marius; or &a.kris; for debugging information.</td>
134	</tr>
135
136	<tr>
137	<td>sparc64 frequent hangs</td>
138	<td>&status.unknown;</td>
139	<td></td>
140	<td>no DDB break possible, so impossible to diagnose</td>
141	</tr>
142
143	<tr>
144	<td>serious sparc64 IPv6 panic</td>
145	<td>&status.wip;</td>
146	<td>&a.gnn;</td>
147	<td>Triggered by just ping6'ing the box. It may even be a MI
148	issue, the reporter of this bug only uses IPv6 with
149	sparc64. This problem seems to be triggered even when debug.mpsafenet=0.</td>
150	</tr>
151
152	<tr>
153	<td>sort(1) does not work with some locales</td>
154	<td>&status.new;</td>
155	<td> </td>
156	<td>sort(1) can cause a coredump with some locales.
157	See also gnu/93629.</td>
158	</tr>
159
160	<tr>
161	<td>exec_map depletion</td>
162	<td>&status.wip;</td>
163	<td>&a.ups;</td>
164	<td>The exec_map is regularly running out of space
165	on machines running 7.0. &a.ups; has a proposed
166	patch and it seems to fix this problem.</td>
167	</tr>
168
169	<tr>
170	<td>NFS data corruption between two 7.0 machines</td>
171	<td>&status.wip;</td>
172	<td>&a.mohans;</td>
173	<td>Running fsx between a 7.0 NFS client and server
174	detects data corruption. This problem can also be reproduced
175	by using 6.1 NFS server.</td>
176	</tr>
177
178	<tr>
179	<td>panic in fxp driver</td>
180	<td>&status.wip;</td>
181	<td>&a.andre;</td>
182	<td>See <a href="http://people.freebsd.org/~pho/stress/log/cons186.html">http://people.freebsd.org/~pho/stress/log/cons186.html</a>.</td>
183	</tr>
184
185	<tr>
186	<td>deadlock in vn_start_write() consumers</td>
187	<td>&status.wip;</td>
188	<td>&a.tegge;</td>
189	<td></td>
190	</tr>
191
192	<tr>
193	<td>panic in bpf</td>
194	<td>&status.wip;</td>
195	<td>&a.sam;</td>
196	<td>killing tcpdump (e.g. with ^C) can cause panics in bpf.</td>
197	</tr>
198
199	<tr>
200	<td>devfs locking problem</td>
201	<td>&status.wip;</td>
202	<td>&a.jeff;</td>
203	<td>It is trivial to deadlock it on an SMP system, and
204	there are other panics with device removal.</td>
205	</tr>
206
207	<tr>
208	<td>pty leak</td>
209	<td>&status.wip;</td>
210	<td>&a.cognet;</td>
211	<td>Since 6.x has a hard-coded limit, once all ptys are
212	leaked things like ssh and login no longer work.
213	This seems devfs-related.</td>
214	</tr>
215
216	<tr>
217	<td>cpu_ipi_selected() can cause a trap on FreeBSD/sparc64</td>
218	<td>&status.wip;</td>
219	<td>&a.kmacy;</td>
220	<td>On sparc64, cpu_ipi_selected() can cause a trap (which is bad since
221	it appears in the trap code path).</td>
222	</tr>
223
224	<tr>
225	<td>rpc.lockd interoperability problems</td>
226	<td>&status.unknown;</td>
227	<td>&a.kuriyama;</td>
228	<td>After <a href="http://www.freebsd.org/cgi/query-pr.cgi?pr=bin/80389">this commit</a>
229	rpc.lockd seems to have interoperability problems.
230	This may be the cause of the many "rpc.lockd no longer interoperates"
231	bug reports seen on -stable.</td>
232	</tr>
233
234	<tr>
235	<td>dup(2) regression on 6.x</td>
236	<td>&status.unknown;</td>
237	<td>&a.des;</td>
238	<td>Simple "close(0); dup(fd)" does not return descriptor "0" in some cases.
239	This problem has been reported in
240	<a href="http://www.freebsd.org/cgi/query-pr.cgi?pr=kern/87208">kern/87208</a>,
241	and there is a proposed patch in the PR, too.</td>
242	</tr>
243
244	<tr>
245	<td>ifconfig regression on 6.x</td>
246	<td>&status.wip;</td>
247	<td>&a.yar;</td>
248	<td>ifconfig cannot handle vlan and mtu parameters at the same time
249	after rev.1.7.2.3 of sbin/ifconfig/ifvlan.c commit.
250	For more information and a proposed patch, see
251	<a href="http://www.freebsd.org/cgi/query-pr.cgi?pr=bin/94028">bin/94028</a>.</td>
252	</tr>
253
254	<tr>
255	<td>"calcru: runtime went backwards" problem for threaded program</td>
256	<td>&status.unknown;</td>
257	<td> </td>
258	<td>stress2 thr1 test can trigger "calcru: runtime went backwards" problem
259	and there are also many similar reports on -stable and -current.
260	&a.phk; committed a possible fix (src/sys/kern/kern_tc.c rev.1.169) to
261	update the calibration code to be more precise on 2 March.</td>
262	</tr>
263	</table>
264
265	<h3>Required features for &local.rel;-RELEASE</h3>
266
267	<a name="required"></a>
268	<table class="tblbasic">
269	<tr class="heading">
270	<th>Issue</th>
271	<th>Status</th>
272	<th>Responsible</th>
273	<th>Description</th>
274	</tr>
275
276	</table>
277
278	<h3>Desired features for &local.rel;-RELEASE</h3>
279
280	<a name="desired"></a>
281	<table class="tblbasic">
282	<tr class="heading">
283	<th>Issue</th>
284	<th>Status</th>
285	<th>Responsible</th>
286	<th>Description</th>
287	</tr>
288
289	<tr>
290	<td>SMP kernels for install</td>
291	<td>&status.unknown;</td>
292	<td></td>
293
294	<td><em>From the <a
295	href="http://www.freebsd.org/projects/ideas/#p-smpinstall">ideas
296	page</a>.</em> Right now we only install a UP kernel, for performance
297	reasons. We should be able to package both a UP and SMP kernel
298	into the release bits, and have sysinstall install both. It
299	should also select the correct one for the target system and
300	make that the default on boot. The easiest way to do this would
301	be to have sysinstall boot an SMP kernel and then look at the
302	hw.ncpu sysctl. The only problem is being able to have
303	sysinstall fall back to booting a UP kernel for itself if the
304	SMP one fails. This can probably be 'faked' by setting one of
305	the SMP-disabling variables in the loader. But in any case, the
306	point is to make the process Just Work for the user, without the
307	user needing to know arcane loader/sysctl knobs. SMP laptops are
308	here, and we should be ready to support SMP out-of-the-box.</td>
309	</tr>
310
311	<tr>
312	<td>swap panic on sparc64</td>
313	<td>&status.unknown;</td>
314	<td>&a.kris; has panic info</td>
315
316	<td>&a.kris; reports configuring a 74GB swap-backed md on sparc64 that
317	caused a panic after a week or two of load (during which time
318	swap was slowly filling as more of the md was dirtied).</td>
319	</tr>
320
321	<tr>
322	<td>updated hal and ath drivers</td>
323	<td>&status.new;</td>
324	<td>&a.sam;</td>
325	<td></td>
326	</tr>
327
328	<tr>
329	<td>fix ntpdate(1) bogus output on amd64.</td>
330	<td>&status.unknown;</td>
331	<td>&a.roberto;</td>
332	<td></td>
333	</tr>
334
335	<tr>
336	<td>Improve performance</td>
337	<td>&status.unknown;</td>
338	<td></td>
339	<td>What seem to be 4BSD scheduler bugs in 6.0 that
340	cause performance to be anomalously low in certain situations.
341	&a.davidxu; has expressed some interest in this problem.</td>
342	</tr>
343
344	<tr>
345	<td>/dev/kmem panic</td>
346	<td>&status.wip;</td>
347	<td>&a.ups;</td>
348	<td>&a.kris; has noticed panics on SMP machines when there was ABI
349	breakage of libkvm and world was not rebuilt and utilities like
350	fstat were used. This suggests panics can be caused by incorrect
351	accesses to /dev/kmem. &a.ups; committed a fix for i386.</td>
352	</tr>
353
354	<tr>
355	<td>KLDs on sparc64</td>
356	<td>&status.new;</td>
357	<td> </td>
358	<td>On sparc64 machines with more than 4Gb memory KLDs are not usable
359	and will panic the system. The problem is reportedly with how the
360	KLDs are compiled, it only works if the code ends up below 4G.</td>
361	</tr>
362
363	<tr>
364	<td>Max RAM on sparc64</td>
365	<td>&status.new;</td>
366	<td> </td>
367	<td>Maximum RAM on sparc64 appears to be limited to 16Gb.</td>
368	</tr>
369
370	<tr>
371	<td>make -jN</td>
372	<td>&status.new;</td>
373	<td> </td>
374	<td>Doing 'make -jN', then suspending/resuming it may result in make
375	reporting it lost child process(es).</td>
376	</tr>
377
378	<tr>
379	<td>OpenBSM</td>
380	<td>&status.unknown;</td>
381	<td>&a.rwatson;</td>
382	<td>The integration of OpenBSM is waiting on some final licensing hurdles.
383	Once those are cleared, it will be a very desirable feature for &local.rel;.</td>
384	</tr>
385
386	<tr>
387	<td>update sysinstall disk labeling</td>
388	<td>&status.wip;</td>
389	<td>&a.rodrigc;</td>
390	<td>Sysinstall could use the same fixes recently made to fdisk so it
391	plays nice with GEOM and disk labeling. This does not cause problems
392	during install because nothing on the disk is mounted when its label
393	is being manipulated but it can cause problems if sysinstall gets
394	used on a live system to adjust labels on existing disks which
395	sys-admins tend to do.</td>
396	</tr>
397
398	<tr>
399	<td>swapping on 6.0 is slower than on 4.x</td>
400	<td>&status.new;</td>
401	<td> </td>
402	<td>Performance on swap handling is much slower than 4.x and
403	this can make a system essentially unusable when moderate
404	paging activity is going on.</td>
405	</tr>
406	</table>
407
408	<h3>Documentation items that must be resolved for &local.rel;</h3>
409
410	<a name="docs"></a>
411	<table class="tblbasic">
412	<tr class="heading">
413	<th>Issue</th>
414	<th>Status</th>
415	<th>Responsible</th>
416	<th>Description</th>
417	</tr>
418
419	</table>
420
421	<h3>Testing foci for &local.rel;-RELEASE</h3>
422
423	<a name="testing"></a>
424	<table class="tblbasic">
425	<tr class="heading">
426	<th>Issue</th>
427	<th>Status</th>
428	<th>Responsible</th>
429	<th>Description</th>
430	</tr>
431
432	<tr>
433	<td>manual root mount lockmgr panics</td>
434	<td>&status.untested;</td>
435	<td>&a.ssouhlal;</td>
436	<td>Specifying a manual root mount location causes lockmgr panics.
437	&a.ssouhlal; has committed a patch for this.</td>
438	</tr>
439
440	<tr>
441	<td>dhclient causes ipv6 panics.</td>
442	<td>&status.untested;</td>
443	<td>&a.dougb;</td>
444	<td>&a.dougb; has more details about this.</td>
445	</tr>
446
447	<tr>
448	<td>amd64 panics in ipv6 with date(1)</td>
449	<td>&status.untested;</td>
450	<td>&a.ume;</td>
451	<td>amd64 panics in ipv6 when the date is changed using date(1) or
452	ntpdate(1). This may be a MI issue.</td>
453	</tr>
454
455	<tr>
456	<td>grep(1) -w does not work with multibyte locales</td>
457	<td>&status.untested;</td>
458	<td>&a.tjr;</td>
459	<td>grep(1) -w generates wrong results with non-UTF-8
460	multibyte locales. &a.tjr; has committed a patch
461	to -HEAD. See also gnu/91909.</td>
462	</tr>
463
464	<tr>
465	<td>Improve kbdmux</td>
466	<td>&status.wip;</td>
467	<td>&a.emax;</td>
468	<td><em>From the <a
469	href="http://www.freebsd.org/projects/ideas/#p-kbdmux">ideas
470	page</a>.</em> We need this for the growing number of systems
471	that assume that USB is the primary keyboard. Current status
472	appears to be that the kbdmux driver breaks very easily. We need
473	this working well enough where it can be enabled by default, and
474	all attached keyboards Just Work. &a.emax; commit kbdmux and
475	rc.d/syscons patches in HEAD and RELENG_6. It is not yet enabled
476	by default. See kbdmux(4) and contact &a.emax; if you have
477	problems.</td>
478	</tr>
479	</table>
480
481	<h3>Stress Test Panics</h3>
482
483	<a name="stresstest"></a>
484	<p>The system is continuously being subjected to Peter Holm's <a
485	href="http://www.holm.cc/stress/">Kernel Stress Test Suite</a>. The
486	following issues have recently been discovered from this test
487	suite.</p>
488
489	&stresstest;
490
491	&footer;
492
493	</body>
494	</html>

Line 0 Link Here

(-)6.1R.new/todo.xml (+401 lines)
		1	<?xml version="1.0"?>
		2	<!DOCTYPE entries SYSTEM "todo.dtd">
		3
		4	<!--
		5
		6	Changes to this list MUST NOT be committed without approval of
		7	Release Engineering Team (re@FreeBSD.org) (for general items) or
		8	Documentation Engineering Team (doceng@FreeBSD.org) (for doc-related
		9	items).
		10
		11	available classes: < showstopper \| required \| desired \| docs \|
		12	testing >
		13	available states: < na \| done \| wip \| untested \| new \| unknown \|
		14	deferred >
		15
		16	-->
		17
		18	<entries>
		19	<cvs:keywords xmlns:cvs="http://www.FreeBSD.org/XML/CVS" version="1.0">
		20	<cvs:keyword name="freebsd">
		21	$FreeBSD:$
		22	</cvs:keyword>
		23	</cvs:keywords>
		24
		25	<entry class="showstopper" status="unknown">
		26	<issue>Unreliable serial console</issue>
		27	<description>
		28	At the manual 'root mount' prompt, the serial console is very
		29	unreliable and drops most characters.
		30	</description>
		31	</entry>
		32
		33	<entry class="testing" status="untested">
		34	<issue>Manual root mount lockmgr panics</issue>
		35	<responsible>Suleiman Souhlal</responsible>
		36	<description>
		37	Specifying a manual root mount location causes lockmgr panics.
		38	ssouhlal@ has committed a patch for this.
		39	</description>
		40	</entry>
		41
		42	<entry class="showstopper" status="deferred">
		43	<issue>i386 deadlocks with more than 16GB swap</issue>
		44	<responsible>Alan L. Cox</responsible>
		45	<description>
		46	i386 deadlocks if more than 16GB of swap is in use. Increasing
		47	the kern.maxswzone tunable would be a workaround this, but a patch
		48	from alc@ is needed to allow this variable to be increased.
		49	</description>
		50	</entry>
		51
		52	<entry class="showstopper" status="wip">
		53	<issue>Unmount pending error</issue>
		54	<responsible>Suleiman Souhlal</responsible>
		55	<description>
		56	When unmounting filesystems Kris Kennaway reports seeing this
		57	warning: <tt>/c: unmount pending error: blocks -68512 files 0</tt>.
		58	This dates back at least to 5.3. It might be associated with
		59	filesystem corruption reported by many users in which the 'used'
		60	space on a filesystem is negative; fsck -f is needed to correct
		61	this.
		62	</description>
		63	</entry>
		64
65	<entry class="showstopper" status="unknown">
66	<issue>swap_pager warnings</issue>
67	<responsible>Don Lewis?</responsible>
68	<description>
69	When swapfiles are in use, there are often warnings printed: <tt>
70	swap_pager: indefinite wait buffer: bufobj: 0, blkno: 889347,
71	size: 8192</tt>. There is also the possibility of deadlock.
72	</description>
73	</entry>
74
75	<entry class="showstopper" status="wip">
76	<issue>umount -f panics</issue>
77	<responsible>Jeff Roberson, Suleiman Souhlal</responsible>
78	<description>
79	Panics from race conditions. A patch from Jeff Roberson seems to
80	fix some of them.
81	</description>
82	</entry>
83
84	<entry class="showstopper" status="unknown">
85	<issue>UFS deadlocks on amd64</issue>
86	<responsible>Tor Egge</responsible>
87	<description>
88	Seen by Kris Kennaway.
89	</description>
90	</entry>
91
92	<entry class="showstopper" status="unknown">
93	<issue>UFS deadlocks</issue>
94	<description>
95	Seen by Peter Jeremy.
96	</description>
97	</entry>
98
99	<entry class="testing" status="untested">
100	<issue>amd64 panics in ipv6 with date(1)</issue>
101	<responsible>Hajimu UMEMOTO</responsible>
102	<description>
103	amd64 panics in ipv6 when the date is changed using date(1) or
104	ntpdate(1). This may be a MI issue.
105	</description>
106	</entry>
107
108	<entry class="showstopper" status="unknown">
109	<issue>sparc64 instability.</issue>
110	<responsible>Marius Strobl</responsible>
111	<description>
112	sparc64 instability when accessing /dev/mem. Contact marius@ or
113	kris@ for debugging information.
114	</description>
115	</entry>
116
117	<entry class="testing" status="untested">
118	<issue>dhclient causes ipv6 panics.</issue>
119	<responsible>Doug Barton</responsible>
120	<description>
121	dougb@ has more details about this.
122	</description>
123	</entry>
124
125	<entry class="showstopper" status="unknown">
126	<issue>sparc64 frequent hangs</issue>
127	<status>unknown</status>
128	<description>
129	No DDB break possible, so impossible to diagnose.
130	</description>
131	</entry>
132
133	<entry class="showstopper" status="unknown">
134	<issue>Serious sparc64 IPv6 panic</issue>
135	<responsible>George V. Neville-Neil</responsible>
136	<description>
137	Triggered by just ping6'ing the box. It may even be a MI issue,
138	the reporter of this bug only uses IPv6 with sparc64. This problem
139	seems to be triggered even when debug.mpsafenet=0.
140	</description>
141	</entry>
142
143	<entry class="desired" status="unknown">
144	<issue>SMP kernels for install</issue>
145	<description>
146	<em>From the <a
147	href="http://www.freebsd.org/projects/ideas/#p-smpinstall">ideas
148	page</a>.</em> Right now we only install a UP kernel, for
149	performance reasons. We should be able to package both a UP and
150	SMP kernel into the release bits, and have sysinstall install
151	both. It should also select the correct one for the target system
152	and make that the default on boot. The easiest way to do this
153	wouldbe to have sysinstall boot an SMP kernel and then look at the
154	hw.ncpu sysctl. The only problem is being able to have sysinstall
155	fall back to booting a UP kernel for itself if the SMP one fails.
156	This can probably be 'faked' by setting one of the SMP-disabling
157	variables in the loader. But in any case, the point is to make
158	the process Just Work for the user, without the user needing to
159	know arcane loader/sysctl knobs. SMP laptops are here, and we
160	should be ready to support SMP out-of-the-box.
161	</description>
162	</entry>
163
164	<entry class="testing" status="untested">
165	<issue>Improve kbdmux</issue>
166	<responsible>Maksim Yevmenkin</responsible>
167	<description>
168	<em>From the <a
169	href="http://www.freebsd.org/projects/ideas/#p-kbdmux">ideas
170	page</a>.</em> We need this for the growing number of systems
171	that assume that USB is the primary keyboard. Current status
172	appears to be that the kbdmux driver breaks very easily. We need
173	this working well enough where it can be enabled by default, and
174	all attached keyboards Just Work.
175	</description>
176	</entry>
177
178	<entry class="desired" status="unknown">
179	<issue>swap panic on sparc64</issue>
180	<responsible>Kris Kennaway has panic info</responsible>
181	<description>
182	Kris reports configuring a 74GB swap-backed md on sparc64 that
183	caused a panic after a week or two of load (during which time swap
184	was slowly filling as more of the md was dirtied).
185	</description>
186	</entry>
187
188	<entry class="desired" status="new">
189	<issue>Updated hal and ath drivers</issue>
190	<responsible>Sam Leffler</responsible>
191	</entry>
192
193	<entry class="desired" status="unknown">
194	<issue>Fix ntpdate(1) bogus output on amd64.</issue>
195	<responsible>Ollivier Robert</responsible>
196	</entry>
197
198	<entry class="desired" status="unknown">
199	<issue>Improve performance</issue>
200	<description>
201	What seem to be 4BSD scheduler bugs in 6.0 that cause performance
202	to be anomalously low in certain situations. davidxu@ has
203	expressed some interest in this problem.
204	</description>
205	</entry>
206
207	<entry class="desired" status="wip">
208	<issue>/dev/kmem panic</issue>
209	<responsible>Stephan Uphoff</responsible>
210	<description>
211	Kris has noticed panics on SMP machines when there was ABI
212	breakage of libkvm and world was not rebuilt and utilities like
213	fstat were used. This suggests panics can be caused by incorrect
214	accesses to /dev/kmem. Stephan Uphoff committed a fix for i386.
215	</description>
216	</entry>
217
218	<entry class="desired" status="new">
219	<issue>KLDs on sparc64</issue>
220	<description>
221	On sparc64 machines with more than 4Gb memory KLDs are not usable
222	and will panic the system. The problem is reportedly with how the
223	KLDs are compiled, it only works if the code ends up below 4G.
224	</description>
225	</entry>
226
227	<entry class="desired" status="new">
228	<issue>Max RAM on sparc64</issue>
229	<description>
230	Maximum RAM on sparc64 appears to be limited to 16Gb.
231	</description>
232	</entry>
233
234	<entry class="desired" status="new">
235	<issue>make -jN</issue>
236	<description>
237	Doing 'make -jN', then suspending/resuming it may result in make
238	reporting it lost child process(es).
239	</description>
240	</entry>
241
242	<entry class="testing" status="untested">
243	<issue>OpenBSM</issue>
244	<responsible>Robert Watson</responsible>
245	<description>
246	The OpenBSM has been MFC'ed to RELENG_6 and needs some testing.
247	</description>
248	</entry>
249
250	<entry class="desired" status="wip">
251	<issue>Update sysinstall disk labeling</issue>
252	<responsible>Craig Rodrigues</responsible>
253	<description>
254	Sysinstall could use the same fixes recently made to fdisk so it
255	plays nice with GEOM and disk labeling. This does not cause
256	problems during install because nothing on the disk is mounted
257	when its label is being manipulated but it can cause problems if
258	sysinstall gets used on a live system to adjust labels on existing
259	disks which sys-admins tend to do.
260	</description>
261	</entry>
262
263	<entry class="showstopper" status="wip">
264	<issue>Quota deadlocks</issue>
265	<responsible>Jeff Roberson</responsible>
266	<description>
267	Quota support is not locked properly and causes deadlocks. A
268	patch from Jeff Roberson seems to fix some of them.
269	</description>
270	</entry>
271
272	<entry class="showstopper" status="new">
273	<issue>sort(1) does not work with some locales</issue>
274	<description>
275	sort(1) can cause a coredump with some locales. See also
276	gnu/93629.
277	</description>
278	</entry>
279
280	<entry class="showstopper" status="wip">
281	<issue>exec_map depletion</issue>
282	<responsible>Stephan Uphoff</responsible>
283	<description>
284	The exec_map is regularly running out of space on machines running
285	7.0. Stephan Uphoff has a proposed patch and it seems to fix this
286	problem.
287	</description>
288	</entry>
289
290	<entry class="showstopper" status="wip">
291	<issue>NFS data corruption between two 7.0 machines</issue>
292	<responsible>Mohan Srinivasan</responsible>
293	<description>
294	Running fsx between a 7.0 NFS client and server detects data
295	corruption. This problem can also be reproduced by using 6.1 NFS
296	server.
297	</description>
298	</entry>
299
300	<entry class="showstopper" status="wip">
301	<issue>Panic in fxp driver</issue>
302	<responsible>Andre Oppermann</responsible>
303	<description>
304	See <a href="http://people.freebsd.org/~pho/stress/log/cons186.html">
305	http://people.freebsd.org/~pho/stress/log/cons186.html</a>.
306	</description>
307	</entry>
308
309	<entry class="showstopper" status="wip">
310	<issue>Deadlock in vn_start_write() consumers</issue>
311	<responsible>Tor Egge</responsible>
312	</entry>
313
314	<entry class="showstopper" status="wip">
315	<issue>panic in bpf</issue>
316	<responsible>Sam Leffler</responsible>
317	<description>
318	Killing tcpdump (e.g. with ^C) can cause panics in bpf.
319	</description>
320	</entry>
321
322	<entry class="showstopper" status="wip">
323	<issue>devfs locking problem</issue>
324	<responsible>Jeff Roberson</responsible>
325	<description>
326	It is trivial to deadlock it on an SMP system, and there are other
327	panics with device removal.
328	</description>
329	</entry>
330
331	<entry class="showstopper" status="wip">
332	<issue>pty leak</issue>
333	<responsible>Olivier Houchard</responsible>
334	<description>
335	Since 6.x has a hard-coded limit, once all ptys are leaked things
336	like ssh and login no longer work. This seems to be related to
337	devfs.
338	</description>
339	</entry>
340
341	<entry class="showstopper" status="wip">
342	<issue>cpu_ipi_selected() can cause a trap on sparc64</issue>
343	<responsible>Kip Macy</responsible>
344	<description>
345	On sparc64, cpu_ipi_selected() can cause a trap (which is bad
346	since it appears in the trap code path).
347	</description>
348	</entry>
349
350	<entry class="showstopper" status="unknown">
351	<issue>rpc.lockd interoperability problems</issue>
352	<responsible>Jun Kuriyama</responsible>
353	<description>
354	After <a href="http://www.freebsd.org/cgi/query-pr.cgi?pr=bin/80389">
355	this commit</a> rpc.lockd seems to have interoperability
356	problems. This may be the cause of the many "rpc.lockd no longer
357	interoperates" bug reports seen on -stable.
358	</description>
359	</entry>
360
361	<entry class="showstopper" status="wip">
362	<issue>dup(2) regression on 6.x</issue>
363	<responsible>Dag-Erling Smorgrav</responsible>
364	<description>
365	Simple "close(0); dup(fd)" does not return descriptor "0" in some
366	cases. This problem has been reported in kern/87208, and there is
367	a proposed patch in the PR, too.
368	</description>
369	</entry>
370
371	<entry class="showstopper" status="wip">
372	<issue>ifconfig regression on 6.x</issue>
373	<responsible>Yar Tikhiy</responsible>
374	<description>
375	ifconfig can not handle vlan and mtu parameters at the same time
376	after rev.1.7.2.3 of sbin/ifconfig/ifvlan.c commit. For more
377	information and a proposed patch, see bin/94028.
378	</description>
379	</entry>
380
381	<entry class="showstopper" status="unknown">
382	<issue>"calcru: runtime went backwards" problem for threaded
383	program</issue>
384	<description>
385	stress2 thr1 test can trigger "calcru: runtime went backwards"
386	problem and there are also many similar reports on -stable and
387	-current. Poul-Henning Kamp committed a possible fix
388	(src/sys/kern/kern_tc.c rev.1.169) to update the calibration code
389	to be more precise on 2 March.
390	</description>
391	</entry>
392
393	<entry class="desired" status="new">
394	<issue>Swapping on 6.0 is slower than on 4.x</issue>
395	<description>
396	Performance on swap handling is much slower than 4.x and this can
397	make a system essentially unusable when moderate paging activity
398	is going on.
399	</description>
400	</entry>
401	</entries>

Line 0 Link Here

(-)6.1R.new/todo.xsl (+221 lines)
		1	<?xml version="1.0"?>
		2
		3	<!-- Copyright (c) 2006 Daniel Gerzo <danger@rulez.sk>
		4	All rights reserved.
		5
		6	Redistribution and use in source and binary forms, with or without
		7	modification, are permitted provided that the following conditions
		8	are met:
		9	1. Redistributions of source code must retain the above copyright
		10	notice, this list of conditions and the following disclaimer.
		11	2. Redistributions in binary form must reproduce the above copyright
		12	notice, this list of conditions and the following disclaimer in the
		13	documentation and/or other materials provided with the distribution.
		14
		15	THIS SOFTWARE IS PROVIDED BY THE AUTHOR AND CONTRIBUTORS ``AS IS'' AND
		16	ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
		17	IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
		18	ARE DISCLAIMED. IN NO EVENT SHALL THE AUTHOR OR CONTRIBUTORS BE LIABLE
		19	FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
		20	DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
		21	OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
		22	HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
		23	LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
		24	OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
		25	SUCH DAMAGE.
		26
		27	$FreeBSD:$
		28	-->
		29
		30	<xsl:stylesheet xmlns:xsl="http://www.w3.org/1999/XSL/Transform" version="1.0"
		31	xmlns:cvs="http://www.FreeBSD.org/XML/CVS"
		32	exclude-result-prefixes="cvs">
		33
		34	<xsl:import href="../../includes.xsl"/>
		35	<xsl:variable name="section" select="'download'"/>
		36	<xsl:variable name="base" select="'../..'"/>
		37	<xsl:variable name="date">
		38	<xsl:value-of select="//cvs:keyword[@name='freebsd']"/>
		39	</xsl:variable>
		40	<xsl:variable name="email" select="'re'"/>
		41	<xsl:variable name="localrel" select="'6.1'"/>
		42	<xsl:variable name="title" select="concat('FreeBSD ', $localrel, '-RELEASE Open Issues')"/>
		43	<xsl:variable name="upperCase" select="'ABCDEFGHIJKLMNOPQRSTUVWXYZ'"/>
		44	<xsl:variable name="lowerCase" select="'abcdefghijklmnopqrstuvwxyz'"/>
		45
		46	<xsl:output method="xml" encoding="iso-8859-1"
		47	doctype-system="http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd"
		48	doctype-public="-//W3C//DTD XHTML 1.0 Transitional//EN"/>
		49
		50	<xsl:template match="entries">
		51	<html>
		52
		53	<xsl:copy-of select="$header1"/>
		54
		55	<body>
		56
		57	<div id="containerwrap">
		58	<div id="container">
		59
		60	<xsl:copy-of select="$header2"/>
		61
		62	<div id="content">
		63
		64	<xsl:copy-of select="$sidenav"/>
65
66	<div id="contentwrap">
67
68	<xsl:copy-of select="$header3"/>
69
70	<p>This is a list of open issues that need to be resolved for
71	FreeBSD <xsl:value-of select="$localrel" />-RELEASE. If you
72	have any updates for this list, please e-mail
73	re@FreeBSD.org.</p>
74
75	<ul>
76	<li><a href="#showstopper">Show stopper defects</a></li>
77	<li><a href="#required">Required features</a></li>
78	<li><a href="#desired">Desired features</a></li>
79	<li><a href="#docs">Documentation Items</a></li>
80	<li><a href="#testing">Testing focuses</a></li>
81	<li><a href="#stresstest">Problems Discovered by Kernel Stress Test Suite</a></li>
82	</ul>
83
84	<h3>Show stopper defects for <xsl:value-of select="$localrel"/>-RELEASE</h3>
85	<a name="showstopper"></a>
86	<table class="tblbasic">
87	<tr class="heading">
88	<th>Issue</th>
89	<th>Status</th>
90	<th>Responsible</th>
91	<th>Description</th>
92	</tr>
93	<xsl:for-each select="entry[@class='showstopper']">
94	<xsl:sort select="translate(./issue, $upperCase, $lowerCase)"/>
95	<xsl:call-template name="entry"/>
96	</xsl:for-each>
97	</table>
98
99	<h3>Required features for <xsl:value-of select="$localrel"/>-RELEASE</h3>
100	<a name="required"></a>
101	<table class="tblbasic">
102	<tr class="heading">
103	<th>Issue</th>
104	<th>Status</th>
105	<th>Responsible</th>
106	<th>Description</th>
107	</tr>
108	<xsl:for-each select="entry[@class='required']">
109	<xsl:sort select="translate(./issue, $upperCase, $lowerCase)"/>
110	<xsl:call-template name="entry"/>
111	</xsl:for-each>
112	</table>
113
114	<h3>Desired features for <xsl:value-of select="$localrel"/>-RELEASE</h3>
115	<a name="desired"></a>
116	<table class="tblbasic">
117	<tr class="heading">
118	<th>Issue</th>
119	<th>Status</th>
120	<th>Responsible</th>
121	<th>Description</th>
122	</tr>
123	<xsl:for-each select="entry[@class='desired']">
124	<xsl:sort select="translate(./issue, $upperCase, $lowerCase)"/>
125	<xsl:call-template name="entry"/>
126	</xsl:for-each>
127	</table>
128
129	<h3>Documentation items that must be resolved for <xsl:value-of select="$localrel"/>-RELEASE</h3>
130	<a name="docs"></a>
131	<table class="tblbasic">
132	<tr class="heading">
133	<th>Issue</th>
134	<th>Status</th>
135	<th>Responsible</th>
136	<th>Description</th>
137	</tr>
138	<xsl:for-each select="entry[@class='docs']">
139	<xsl:sort select="translate(./issue, $upperCase, $lowerCase)"/>
140	<xsl:call-template name="entry"/>
141	</xsl:for-each>
142	</table>
143
144	<h3>Testing focuses for <xsl:value-of select="$localrel"/>-RELEASE</h3>
145	<a name="testing"></a>
146	<table class="tblbasic">
147	<tr class="heading">
148	<th>Issue</th>
149	<th>Status</th>
150	<th>Responsible</th>
151	<th>Description</th>
152	</tr>
153	<xsl:for-each select="entry[@class='testing']">
154	<xsl:sort select="translate(./issue, $upperCase, $lowerCase)"/>
155	<xsl:call-template name="entry"/>
156	</xsl:for-each>
157	</table>
158
159	<h3>Stress Test Panics</h3>
160	<a name="stresstest"></a>
161	<p>The system is continuously being subjected to Peter Holm's <a
162	href="http://www.holm.cc/stress/">Kernel Stress Test Suite</a>.
163	The following issues have recently been discovered from this
164	test suite.</p>
165
166	<ul>
167	<xsl:for-each select="document('stress.xml')/results/result">
168	<li>
169	<xsl:element name="a">
170	<xsl:attribute name="href"><xsl:value-of select="link"/></xsl:attribute>
171	<xsl:value-of select="title"/>
172	</xsl:element>
173	</li>
174	</xsl:for-each>
175	</ul>
176
177	</div> <!-- contentwrap -->
178	<br class="clearboth" />
179
180	</div> <!-- content -->
181
182	<xsl:copy-of select="$footer"/>
183
184	</div> <!-- container -->
185	</div> <!-- containerwrap -->
186
187	</body>
188	</html>
189	</xsl:template>
190
191	<xsl:template name="entry">
192	<tr>
193	<td><xsl:value-of select="issue"/></td>
194	<xsl:choose>
195	<xsl:when test="@status='na'">
196	<td><font color="green">N/A</font></td>
197	</xsl:when>
198	<xsl:when test="@status='done'">
199	<td><font color="green">Done</font></td>
200	</xsl:when>
201	<xsl:when test="@status='wip'">
202	<td><font color="blue">In progress</font></td>
203	</xsl:when>
204	<xsl:when test="@status='untested'">
205	<td><font color="orange">Needs testing</font></td>
206	</xsl:when>
207	<xsl:when test="@status='new'">
208	<td><font color="red">Not done</font></td>
209	</xsl:when>
210	<xsl:when test="@status='unknown'">
211	<td><font color="red">Unknown</font></td>
212	</xsl:when>
213	<xsl:when test="@status='deferred'">
214	<td><font color="gray">Deferred for future release</font></td>
215	</xsl:when>
216	</xsl:choose>
217	<td><xsl:value-of select="responsible"/></td>
218	<td><xsl:copy-of select="description/child::node()"/></td>
219	</tr>
220	</xsl:template>
221	</xsl:stylesheet>