Open main menu

OpenVZ Virtuozzo Containers Wiki β

Download/kernel/2.6.8/022stab078.21/changes

< Download‎ | kernel‎ | 2.6.8‎ | 022stab078.21
Revision as of 12:14, 21 March 2008 by Kir (talk | contribs) (created)
(diff) ← Older revision | Latest revision (diff) | Newer revision → (diff)

Contents

Changes

  • Compilation fixes
  • Driver updates
  • Other fixes and improvements

Config changes

Added:

  • +CONFIG_MEGARAID_SAS=y
  • +CONFIG_BNX2=m
  • +CONFIG_IPMI_POWEROFF=m
  • +CONFIG_EFI_RTC=y (for ia64 only)
  • +CONFIG_FUSION_SPI=y
  • +CONFIG_FUSION_FC=y
  • +CONFIG_FUSION_SAS=y
  • +CONFIG_SCSI_ADP94XX=y
  • +CONFIG_SCSI_DPT_I2O=m
  • +CONFIG_I2O_EXT_ADAPTEC=y
  • +CONFIG_I2O_CONFIG_OLD_IOCTL=y
  • +CONFIG_I2O_BUS=m (both i2o drivers are disabled for IA64)
  • +CONFIG_CISS_SCSI_TAPE=y

Removed:

  • -CONFIG_FUSION_ISENSE
  • -CONFIG_I2O_EXT_ADAPTEC_DMA64
  • -CONFIG_I2O_LCT_NOTIFY_ON_CHANGES


Patches

linux-2.6.8.1-megaraid-sas-00.00.02.00.patch

Patch prepared by Kostja:
megaraid_sas driver added version 00.00.02.00.

Sources were taken from RHEL 2.6.9-34.0.1 kernel.

Bug #65549.

linux-2.6.8.1-bnx2-1.4.30.patch

Patch prepared by Kostja:
Broadcom NX2 network driver added version 1.4.30.

Sources were taken from RHEL4.3 2.6.9-34.0.1 kernel.

Bug #65551.

linux-2.6.8.1-dcdbas-5.6.0-2.patch

Prepared by Kostja (khorenko@),
sources from ftp://ftp.us.dell.com/sysman/OMI-SrvAdmin-Dell-Web-LX-50_A00.tar.gz

http://linux.dell.com/files/openmanage-contributions/omsa-50-live/omsa-50-live-howto.txt

dcdbas driver updated to 5.6.0-2 version.

dell_rbu driver updated to 0.9 version.

Obsoletes patch linux-2.6.15-dcdbas-5.6.0-1.patch.

Bug #55618.

linux-2.6.8.1-ipmi-33.13.rh.patch-20060801

Prepared by Kostja (khorenko@),
sources from RHEL4.5 beta (2.6.9-39 kernel).

ipmi driver updated to 33.13 version.

Bug #51023.

linux-2.6.8.1-forcedeth-0.41.patch

Prepared by Roman (rchechnev@),
sources from RHEL4.5 beta (2.6.9-39 kernel).

forcedeth driver updated to 0.41 version.

Bug #66273.

diff-ms-netcachesizes-20060807

Patch from Denis:

This patch fixes net cache sizes calculation. They can eat all the low memory in the case on debug kernel on the node with huge amount of memory.

Bug #64858.

diff-ms-mov-20060622

Patch from mainstream, prepared by Kir <kir@openvz.org>:
Subject: [PATCH] i386/x86_64 segment register access update

Kernel 2.6.8-022stab078.10 fails to build with newer binutils (such as binutils-2.16.91.0.6-5 from Fedora Core 5). The error message is:

CC arch/i386/kernel/process.o
{standard input}: Assembler messages:
{standard input}:790: Error: suffix or operands invalid for `mov'
.......

X-Git-Tag: v2.6.12-rc4
X-Git-Url: fd51f666fa591294bd746

diff-ms-ia64-map-cross-region-check-20060726

Patch from Pavel Emelianov <xemul@openvz.org>:

ia64 doesn't alow making mappings which cross region boundaries. elf loader calls do_mmap_pgoff w/o these checks.

Bug #65217.

diff-ve-netlinkdev-compile-20060803

Patch from Kir <kir@openvz.org>:
fixing compilation of af_netlink.c when CONFIG_NETLINK_DEV=y

Found by Dusty Wilson (OpenVZ bug #213). Original bug is introduced by diff-ubc patch.

OpenVZ Bug #213.

diff-ve-nf-ct-destroy-context-20060803

Patch from Dmitry:

This patch fixes issue with wrong context in __ip_ct_find_proto on conntrack destroy.

OpenVZ Bug #211.

diff-ve-net-sockstat-hide-20060704

Patch from Pavel:
Hide /proc/net/sockstat in VE.

Bug #61838.

diff-ms-flock-err-20060605

Patch from Pavel:

If flock_lock_file() failed to allocate flock with locks_alloc_lock() then "error = 0" is returned. Need to return some non-zero.

diff-vzdq-fmt-quota-20060608

Patch from Vasily:

In vzquota format operations are not implemented. If usual quota is launched it uses usual vfs_quota_on which utilize format operations == NULL and it causes oops. This patch force find_quota_format function to skip vzquota format.

OpenVZ Bug #184.

diff-ve-start-jiffies-20060616-mod

Patch from Andrey:

this patch actually fixes jiffies usage for start_time (not 64 bit version)

Orig comment from Andrey:

ve->start_jiffies is set to ve->init_entry->start_time value on VPS start. INITIAL_JIFFIES is used as start_jiffies for VE0. ve->start_jiffies is used instead of ve->init_entry->start_time in proc_pid_stat() and sys_times().

diff-4gb-prefetch

Patch by Kirill (dev@), fixes Red Hat 4G split-related issue:

"Recently we hit PGF on prefetch() on i686 kernel running on AMD processor. I found that this instruction can really generate "wrong" PGF on AMD and that there is a workaround in do_page_fault() like:

       if (is_prefetch(regs, address, error_code))
               return;

The problem is however that __is_prefetch() uses __get_user() inside and thus we should switch get_fs() in case of regs->xcs == KERNEL_CS."

Bug #64397.

Red Hat bug #204286.

diff-arch-ia64-ipmi_poweroff-20060811

Patch from mainstream:
[IA64] acpi.c: export pm_power_off for use by ipmi_poweroff module

Bug #66505.

linux-2.6.8.1-fusion-3.02.62.01rh.patch

diff-drv-fusion-entropy-20060721
diff-drv-fusion-crashdump-20060721
diff-drv-fusion-msleep-20060725

Patches prepared by Kostja (khorenko@):
LSI Logic Fusion MPT driver update from 3.02.18 (RHEL4u2) up to version 3.02.62.01rh (RHEL4.5 beta 2.6.9-39).

Bug #19946.

Obsoleted patches:
linux-2.6.8.1-fusion-3.02.18.patch
diff-scsi-mpt-fusion-20050927
diff-drv-fusion-entropy-20040831
diff-fusion-crashdump-20060303
diff-mpt-fusion-nosleep

diff-arch-4gb-mce-20060824

Patch by Vasily (vvs@):

This patch fixes 4Gb-split-related issue: access to kernel-space memory (machine_check_vector) before context switching

Bug #67271.

diff-dbg-showregs-20060725

Patch from Andrey Savochkin <saw@openvz.org>:
implements printing call traces on all CPUs on show_regs.

Bug #65540.

linux-2.6.8.1-e1000-7.2.7.patch

Patch prepared by Vasily (vvs@):
e1000 driver updated up to 7.2.7 version

sources were taken from http://sourceforge.net/projects/e1000/

obsoletes linux-2.6.8.1-e1000-6.1.16-k3.patch

Bug #19952.

linux-2.6.8-adp94xx-1.0.8.patch

Patch from Vasily (vvs@):

added support for adaptec94xx devices, driver adp94xx version 1.0.8, sources from RHEL4u4 kernel 2.6.9-42

Bug #68062.

linux-2.6.8-dpt_i2o-2.5.0-2426.patch

Patch from Vasily (vvs@):

added alternative driver for I2O hardware, version 2.5.8, build 2426

sources taken from Mark Salyzyn, obsoletes diff-drv-dpt-entropy-20040525

Bug #68066.

linux-2.6.8-i2o-1.325.patch

Patch from Vasily (vvs@):

updates i2o layer, backported from to 2.6.17 linux mainstream kernel

obsoletes diff-i2o-update-20051214

Bug #68066.

linux-2.6.8.1-tg3-3.52.rh.patch

diff-update-ethtool-20060904
diff-pciids-update

Patches prepared by Vasily (vvs@):
tg3 driver updated up to 3.52-rh version,

sources from RHEL4u4 kernel 2.6.9-42

obsoletes linux-2.6.8.1-tg3-3.43.rh.patch

Bug #68032.

diff-ms-bind-mount-flags-20060816

Patch from Andrey (amirkin@):

This patch adds support of 3 mount flags (nodev, noexec, nosuid) to --bind mount. Now we can do bind mounts with noexec, nosuid and nodev options w/o need to do remount.

Bug #67909.

diff-ext3-errorbehaviour-20060902

SWsoft Virtuozzo/OpenVZ Linux kernel team has discovered that ext3 error behavior was broken in linux kernels since 2.5.x versions by the following patch:

2002/10/31 02:15:26-05:00 tytso@snap.thunk.org
Default mount options from superblock for ext2/3 filesystems
http://linux.bkbits.net:8080/linux-2.6/gnupatch@3dc0d88eKbV9ivV4ptRNM8fBuA3JBQ

In case ext3 file system is mounted with errors=continue (EXT3_ERRORS_CONTINUE) errors should be ignored when possible. However at present in case of any error kernel aborts journal and remounts filesystem to read-only. Such behavior was hit number of times and noted to differ from that of 2.4.x kernels.

This patch fixes this:

  • do nothing in case of EXT3_ERRORS_CONTINUE,
  • set EXT3_MOUNT_ABORT and call journal_abort() in all other cases
  • panic() should be called after ext3_commit_super() to save sb marked as EXT3_ERROR_FS

Signed-off-by: Vasily Averin <vvs@sw.ru>
Acked-by: Kirill Korotaev <dev@sw.ru>

Bugs #57259, #67988.

diff-generic-aio-retval-20060904

Patch from Vasily (vvs@),

fixes several DIO and iov related issues in generic_file_aio_write_nolock(), based on inux-2.6.9-generic-aio-retval.patch from RHEL4u4 kernel 2.6.9-42

Bug #68103.

diff-jbd-unexpectdirty-20060905

Patch from linux mainstream, prepared by Vasily (vvs@),

http://linux.bkbits.net:8080/linux-2.6/gnupatch@431f7f0ceyo6g8tikQvG3I-cCSb7kw

attached patch should fix the following race:

  Proc 1                               Proc 2

  __flush_batch()
    ll_rw_block()
                                     do_get_write_access()
                                   lock_buffer
                                          jh is only waiting for checkpoint
                                     -&gt; b_transaction == NULL -&gt;
                                             do nothing
                                           unlock_buffer
    test_set_buffer_locked()
    test_clear_buffer_dirty()
                                           __journal_file_buffer()
                                        change the data
    submit_bh()

and we have sent wrong data to disk... We now clean the dirty buffer flag under buffer lock in all cases and hence we know that whenever a buffer is starting to be journaled we either finish the pending write-out before attaching a buffer to a transaction or we won't write the buffer until the transaction is going to be committed.

The test in jbd_unexpected_dirty_buffer() is redundant - remove it. Furthermore we have to clear the buffer dirty bit under the buffer lock to prevent races with buffer write-out (and hence prevent returning a buffer with IO happening).

Signed-off-by: Jan Kara <jack@suse.cz>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

Bug #68106.

diff-ms-nf-security-checks3-20060904

Patch from Dmitry (dim@):

a lot of changes in order to unify compat checks with regular ones. Fixed bugs with unavailabilty of some iptables targets and matches in 32bit VEs over 64bit kernels. (was broken in 022stab078.16 kernel)

Bugs #68017, #68042, #68043.

diff-ms-pit-cpukhz

Patch from mainstream:

pit timer doesn't initialize cpu_khz

Bug #66955.

diff-ms-rtnl-deadlock-20060802

Patch from Pavel (xemul@),

fixes deadlock on net_todo_run_mutex, based on patch from Linux mainstream

http://linux.bkbits.net:8080/linux-2.6/cset@1.4087.297.147

Bug #66125.

diff-ubc-net-locking-20060727

Patch from Pavel (xemul@):

All sock beancounters are stored in the list, starting at top beancounter, and thus top's lock must be used to protect the list.

diff-ve-multi-cleanup-20060905

Patch from Pavel Emelianov (xemul@), adopted by Vasily (vvs@)
Try to cleanup each VE in a separate thread. This alows simultaneous stop of many VEs at once.

Bug #60673.

diff-venet-rtcache-20060905

Patch from Andrey Savochkin <saw@openvz.org>, backported by Vasily (vvs@)
Fix of broken virtualization of /proc/net/rt_cache.

Bug #65528.

diff-vzdq-quotaoff-EIO-20060705

Patch from Vasily (vvs@):

Turns off quota in spite of errors while syncing inodes.

Bug #65186.

diff-vzwdog-irq-20060905

Patch from Andrey (saw@), corrected by Vasily (vvs@):

Restore showing IRQ information in vzwdog.

Bug #68096.

diff-ms-nf-compat-cleanup-fix-20060907

Patch from Dmitry (dim@):

fixed translate_compat_table() error way.

Bug #68286.

diff-ms-nf-compat-cleanup-20060908

Patch from Dmitry (dim@), fixes issues noticed by Vasiliy (vvs@):
  • add flush of offsets on error way, may lead to table corruption on the next compat_do_replace.
  • added missed checks to comapt_do_replace()
  • targets/matches module_put cleanup
  • debug messages cleanup

Bug #65826.

diff-ve-nf-ct-max-20060908

Patch from Dmitry (dim@):

fixed net.ipv4.ip_conntrack_max shortcut write ability.

Bug #68289.