Download/kernel/rhel5/028stab053.14/changes

From OpenVZ Virtuozzo Containers Wiki
< Download‎ | kernel‎ | rhel5‎ | 028stab053.14
Revision as of 12:24, 21 May 2008 by Kir (talk | contribs) (diff-ms-tcp-slow-start-20080306)
Jump to: navigation, search

Changes

  • re-based on the 2.6.18-53.1.19.EL5 Red Hat kernel
    • fixed CVE-2008-1669
    • fixed CVE-2008-1375
    • fixed CVE-2008-0007
    • fixed CVE-2006-6921
  • updated 3w-9xxx driver to the 2.26.08.003 version (it now provides support for 3ware SAS 9690SA RAID controllers)
  • fixed occasional CT stop error with message: "unregister_netdevice: waiting for lo to become free"

Patches

linux-2.6.18-3w_9xxx-2.26.02.008-08.003.patch

patch from Kirill (kshileev@): updates 3w-9xxx driver up to version 2.26.08.003, in particular adds support for 3ware SAS 9690SA. https://bugzilla.sw.ru/show_bug.cgi?id=100727

diff-ms-tcp-slow-start-20080306

[TCP]: slow_start_after_idle should influence cwnd validation too

For the cases that slow_start_after_idle are meant to deal with, it is almost a certainty that the congestion window tests will think the connection is application limited and we'll thus decrease the cwnd there too. This defeats the whole point of setting slow_start_after_idle to zero.

So test it there too.

We do not cancel out the entire tcp_cwnd_validate() function so that if the sysctl is changed we still have the validation state maintained.

Signed-off-by: David S. Miller <davem@davemloft.net>

From: David S. Miller <davem@sunset.davemloft.net> Date: Mon, 9 Apr 2007 20:23:14 +0000 (-0700)

X-Git-Tag: v2.6.21-rc7~322~1
X-Git-Url: http://git.kernel.org/?p=linux%2Fkernel%2Fgit%2Ftorvalds%2Flinux-2.6.git;a=commitdiff_plain;h=15d33c070ddde99f7368e6b17b71d22c866f97d9;hp=699784b7614ce61b16a075445b8e5b6c379c5086

linux-hp-dmi-info-correct.patch

fix HP DMI decode on some Dell boxes.

diff-nfs-rpcsaddr

Patch from Denis Lunev <den@openvz.org> [RPC]: Add sysctl to specify source address for rpc connections

diff-ms-devleak-dstdebug-20080504

Patch from Denis Lunev <den@openvz.org> [NET]: Leak network device if refcount is positive.

Unfortunately, there is are some ways on which we have positive refcounters on network device on VE stop. This bugs are rather hard to debug and locate.

The idea of this patch is to leave the loop waiting device to become free and _leak_ this device. This is better than a node freeze.

diff-venet-ip6frag-20080504

Patch from Denis Lunev <den@openvz.org> [IPv6]: All IP6 fragments queues should be pruned on VE stop.

Potential fix for bug #75822

diff-ve-ipv6-ifdown-20080318

Patch from Denis Lunev <den@openvz.org> [IPV6]: inet6 device on the loopback should be kept until final stop.

Additionally, addrconf_ifdown is broken in respect to how field. It is called with logical condition and "2" from IPv6 shutdown code. The latter case is the same as !0 but should really destroy a device as container is going down.

diff-ms-ipv6-ac-dst-20080318

Patch from Denis Lunev <den@openvz.org> [IPV6]: Fix refcounting for anycast dst entries.

The problem occures when we stop IPv6 device without dropping all addresses on it. For such a device addrconf_ifdown marks all entries as obsolete and ip6_del_rt called from __ipv6_dev_ac_dec return ENOENT. The referrence is not dropped.

The fix is simple. DST entry should not keep referrence when stored in the FIB6 tree.

See also