===================================================================
RCS file: /cvsrepo/anoncvs/cvs/www/errata25.html,v
retrieving revision 1.53
retrieving revision 1.54
diff -c -r1.53 -r1.54
*** www/errata25.html	2014/03/31 03:12:47	1.53
--- www/errata25.html	2014/03/31 03:36:54	1.54
***************
*** 75,92 ****
  
  <hr>
  
- <a name="all"></a>
- <h3><font color="#e00000">All architectures</font></h3>
  <ul>
  <li><a name="cron"></a>
! <font color="#009000"><strong>SECURITY FIX: Aug 30, 1999</strong></font><br>
  In cron(8), make sure argv[] is NULL terminated in the fake popen() and
  run sendmail as the user, not as root.
  <a href="http://ftp.openbsd.org/pub/OpenBSD/patches/2.5/common/012_cron.patch">
  A source code patch exists which remedies this problem.</a>
  <p>
  <li><a name="miscfs"></a>
! <font color="#009000"><strong>SECURITY FIX: Aug 12, 1999</strong></font><br>
  The procfs and fdescfs filesystems had an overrun in their handling
  of uio_offset in their readdir() routines. (These filesystems are not
  enabled by default).
--- 75,90 ----
  
  <hr>
  
  <ul>
  <li><a name="cron"></a>
! <font color="#009000"><strong>SECURITY FIX: Aug 30, 1999</strong></font> &nbsp; <i>All architectures</i><br>
  In cron(8), make sure argv[] is NULL terminated in the fake popen() and
  run sendmail as the user, not as root.
  <a href="http://ftp.openbsd.org/pub/OpenBSD/patches/2.5/common/012_cron.patch">
  A source code patch exists which remedies this problem.</a>
  <p>
  <li><a name="miscfs"></a>
! <font color="#009000"><strong>SECURITY FIX: Aug 12, 1999</strong></font> &nbsp; <i>All architectures</i><br>
  The procfs and fdescfs filesystems had an overrun in their handling
  of uio_offset in their readdir() routines. (These filesystems are not
  enabled by default).
***************
*** 94,127 ****
  A source code patch exists which remedies this problem.</a>
  <p>
  <li><a name="profil"></a>
! <font color="#009000"><strong>SECURITY FIX: Aug 9, 1999</strong></font><br>
  Stop profiling (see profil(2)) when we execve() a new process.
  <a href="http://ftp.openbsd.org/pub/OpenBSD/patches/2.5/common/010_profil.patch">
  A source code patch exists which remedies this problem.</a>
  <p>
  <li><a name="ipsec_in_use"></a>
! <font color="#009000"><strong>SECURITY FIX: Aug 6, 1999</strong></font><br>
  Packets that should have been handled by IPsec may be transmitted
  as cleartext. PF_KEY SA expirations may leak kernel resources.
  <a href="http://ftp.openbsd.org/pub/OpenBSD/patches/2.5/common/009_ipsec_in_use.patch">
  A source code patch exists which remedies this problem.</a>
  <p>
  <li><a name="rc"></a>
! <font color="#009000"><strong>SECURITY FIX: Aug 5, 1999</strong></font><br>
  In /etc/rc, use mktemp(1) for motd re-writing, and change the find(1)
  to use -execdir.
  <a href="http://ftp.openbsd.org/pub/OpenBSD/patches/2.5/common/008_rc.patch">
  A source code patch exists which remedies this problem.</a>
  <p>
  <li><a name="chflags"></a>
! <font color="#009000"><strong>SECURITY FIX: Jul 30, 1999</strong></font><br>
  Do not permit regular users to chflags(2) or fchflags(2) on character or
  block devices which they may currently be the owner of.
  <a href="http://ftp.openbsd.org/pub/OpenBSD/patches/2.5/common/007_chflags.patch">
  A source code patch exists which remedies this problem.</a>
  <p>
  <li><a name="nroff"></a>
! <font color="#009000"><strong>SECURITY FIX: Jul 27, 1999</strong></font><br>
  Cause groff(1) to be invoked with the -S flag, when called by nroff(1),
  to avoid various groff features which may be security issues.  On the
  whole, this is not really a security issue, but it was discussed on
--- 92,125 ----
  A source code patch exists which remedies this problem.</a>
  <p>
  <li><a name="profil"></a>
! <font color="#009000"><strong>SECURITY FIX: Aug 9, 1999</strong></font> &nbsp; <i>All architectures</i><br>
  Stop profiling (see profil(2)) when we execve() a new process.
  <a href="http://ftp.openbsd.org/pub/OpenBSD/patches/2.5/common/010_profil.patch">
  A source code patch exists which remedies this problem.</a>
  <p>
  <li><a name="ipsec_in_use"></a>
! <font color="#009000"><strong>SECURITY FIX: Aug 6, 1999</strong></font> &nbsp; <i>All architectures</i><br>
  Packets that should have been handled by IPsec may be transmitted
  as cleartext. PF_KEY SA expirations may leak kernel resources.
  <a href="http://ftp.openbsd.org/pub/OpenBSD/patches/2.5/common/009_ipsec_in_use.patch">
  A source code patch exists which remedies this problem.</a>
  <p>
  <li><a name="rc"></a>
! <font color="#009000"><strong>SECURITY FIX: Aug 5, 1999</strong></font> &nbsp; <i>All architectures</i><br>
  In /etc/rc, use mktemp(1) for motd re-writing, and change the find(1)
  to use -execdir.
  <a href="http://ftp.openbsd.org/pub/OpenBSD/patches/2.5/common/008_rc.patch">
  A source code patch exists which remedies this problem.</a>
  <p>
  <li><a name="chflags"></a>
! <font color="#009000"><strong>SECURITY FIX: Jul 30, 1999</strong></font> &nbsp; <i>All architectures</i><br>
  Do not permit regular users to chflags(2) or fchflags(2) on character or
  block devices which they may currently be the owner of.
  <a href="http://ftp.openbsd.org/pub/OpenBSD/patches/2.5/common/007_chflags.patch">
  A source code patch exists which remedies this problem.</a>
  <p>
  <li><a name="nroff"></a>
! <font color="#009000"><strong>SECURITY FIX: Jul 27, 1999</strong></font> &nbsp; <i>All architectures</i><br>
  Cause groff(1) to be invoked with the -S flag, when called by nroff(1),
  to avoid various groff features which may be security issues.  On the
  whole, this is not really a security issue, but it was discussed on
***************
*** 130,164 ****
  A source code patch exists which remedies this problem.</a>
  <p>
  <li><a name="fts"></a>
! <font color="#009000"><strong>RELIABILITY FIX: May 19, 1999</strong></font><br>
  Programs using fts(3) could dump core when given a directory structure
  with a very large number of entries.
  <a href="http://ftp.openbsd.org/pub/OpenBSD/patches/2.5/common/005_fts.patch">
  A source code patch exists which remedies this problem.</a>
  <p>
  <li><a name="tcpsack"></a>
! <font color="#009000"><strong>RELIABILITY FIX: May 19, 1999</strong></font><br>
  Sequence numbers could wrap with TCP_SACK and TCP_NEWRENO, resulting in
  failure to retransmit correctly.
  <a href="http://ftp.openbsd.org/pub/OpenBSD/patches/2.5/common/004_tcpsack.patch">
  A source code patch exists which remedies this problem.</a>
  <p>
  <li><a name="ipsec1"></a>
! <font color="#009000"><strong>RELIABILITY FIX</strong></font><br>
  Retransmitted TCP packets could get corrupted when flowing over an
  IPSEC ESP tunnel.
  <a href="http://ftp.openbsd.org/pub/OpenBSD/patches/2.5/common/003_espdata.patch">
  A source code patch exists which remedies this problem.</a>
  <p>
  <li><a name="bmap"></a>
! <font color="#009000"><strong>RELIABILITY FIX</strong></font><br>
  A local user can crash the system by reading a file larger than 64meg
  from an ext2fs partition.
  <a href="http://ftp.openbsd.org/pub/OpenBSD/patches/2.5/common/002_bmap.patch">
  A source code patch exists which remedies this problem.</a>
  <p>
  <li><a name="pfkey"></a>
! <font color="#009000"><strong>RELIABILITY FIX</strong></font><br>
  PF_KEY socket operations leak internal kernel resources, so that a
  system running an IPsec key management daemon like photurisd or isakmpd
  will cause the networking subsystem to stop working after a finite amount
--- 128,162 ----
  A source code patch exists which remedies this problem.</a>
  <p>
  <li><a name="fts"></a>
! <font color="#009000"><strong>RELIABILITY FIX: May 19, 1999</strong></font> &nbsp; <i>All architectures</i><br>
  Programs using fts(3) could dump core when given a directory structure
  with a very large number of entries.
  <a href="http://ftp.openbsd.org/pub/OpenBSD/patches/2.5/common/005_fts.patch">
  A source code patch exists which remedies this problem.</a>
  <p>
  <li><a name="tcpsack"></a>
! <font color="#009000"><strong>RELIABILITY FIX: May 19, 1999</strong></font> &nbsp; <i>All architectures</i><br>
  Sequence numbers could wrap with TCP_SACK and TCP_NEWRENO, resulting in
  failure to retransmit correctly.
  <a href="http://ftp.openbsd.org/pub/OpenBSD/patches/2.5/common/004_tcpsack.patch">
  A source code patch exists which remedies this problem.</a>
  <p>
  <li><a name="ipsec1"></a>
! <font color="#009000"><strong>RELIABILITY FIX</strong></font> &nbsp; <i>All architectures</i><br>
  Retransmitted TCP packets could get corrupted when flowing over an
  IPSEC ESP tunnel.
  <a href="http://ftp.openbsd.org/pub/OpenBSD/patches/2.5/common/003_espdata.patch">
  A source code patch exists which remedies this problem.</a>
  <p>
  <li><a name="bmap"></a>
! <font color="#009000"><strong>RELIABILITY FIX</strong></font> &nbsp; <i>All architectures</i><br>
  A local user can crash the system by reading a file larger than 64meg
  from an ext2fs partition.
  <a href="http://ftp.openbsd.org/pub/OpenBSD/patches/2.5/common/002_bmap.patch">
  A source code patch exists which remedies this problem.</a>
  <p>
  <li><a name="pfkey"></a>
! <font color="#009000"><strong>RELIABILITY FIX</strong></font> &nbsp; <i>All architectures</i><br>
  PF_KEY socket operations leak internal kernel resources, so that a
  system running an IPsec key management daemon like photurisd or isakmpd
  will cause the networking subsystem to stop working after a finite amount