* * * READ ME * * *
                * * * Veritas File System 3.5 MP4 * * *
                      * * * Rolling Patch 8 * * *
                         Patch Date: 2014-05-07


This document provides the following information:

   * PATCH NAME
   * OPERATING SYSTEMS SUPPORTED BY THE PATCH
   * PACKAGES AFFECTED BY THE PATCH
   * BASE PRODUCT VERSIONS FOR THE PATCH
   * SUMMARY OF INCIDENTS FIXED BY THE PATCH
   * DETAILS OF INCIDENTS FIXED BY THE PATCH
   * INSTALLATION PRE-REQUISITES
   * INSTALLING THE PATCH
   * REMOVING THE PATCH


PATCH NAME
----------
Veritas File System 3.5 MP4 Rolling Patch 8


OPERATING SYSTEMS SUPPORTED BY THE PATCH
----------------------------------------
HP-UX 11i v1 (11.11)


PACKAGES AFFECTED BY THE PATCH
------------------------------
VRTSvxfs


BASE PRODUCT VERSIONS FOR THE PATCH
-----------------------------------
   * Veritas File System 3.5
   * Veritas Storage Foundation Cluster File System 3.5
   * Veritas Storage Foundation 3.5
   * Veritas Storage Foundation HA 3.5
   * Veritas Storage Foundation for Oracle 3.5


SUMMARY OF INCIDENTS FIXED BY THE PATCH
---------------------------------------
Patch ID: PHKL_43921
* 3153843 (472008) The O_SYNC write sometimes returns the EIO error in the 
vx_tranidflush function.
* 3482059 (1156791) The write(1M) operation on files with odd-size extents is slower compared to 
the write(1M) operation on files with even-size extents.
Patch ID: PHKL_43091
* 2768673 (2272072) [VxFS][413-825-203][LG DISPLAY] Threads stuck in vx_rwsleep_rec_lock_em


DETAILS OF INCIDENTS FIXED BY THE PATCH
---------------------------------------
This patch fixes the following Symantec incidents:

Patch ID: PHKL_43921

* 3153843 (Tracking ID: 472008)

SYMPTOM:
The O_SYNC write sometimes returns the EIO error in the vx_tranidflush function.

DESCRIPTION:
A race condition exists between the thread that waits for a transaction to get 
logged to disk in the intent log portion of the file system (vx_tranidflush()),
and the thread which flushes the intent log (vx_logbuf_clean() and its callers.
This can make the former thread that is waiting, to incorrectly read the 
situation and return EIO. This then is propagated as the return value of the 
corresponding system call.

RESOLUTION:
The code is modified to ensure that the thread that waits for a transaction to
be logged, correctly reads the situation, by taking the active-transaction-
queue lock on the corresponding file system.

* 3482059 (Tracking ID: 1156791)

SYMPTOM:
The write(1M) operation on files with odd-size extents is slower compared to 
the write(1M) operation on files with even-size extents.

DESCRIPTION:
The extent allocator can spend a long time in looking for a good match while 
looking for an odd-sized extent. It might need to examine every allocation unit 
to find a good match. On a large file system with a large number of small 
files, this could take a long time.

RESOLUTION:
The odd-size extent allocations are now rounded up to the nearest power of 2.
For example, an extent of size 7 KB is rounded to 2^3 = 8 KB.

Patch ID: PHKL_43091

* 2768673 (Tracking ID: 2272072)

SYMPTOM:
GAB panics the box because VCS engine "had" did not respond, the lbolt
wraps around.

DESCRIPTION:
The lbolt wraps around after 498 days machine uptime. In VxFS, we
flush VxFS meta data buffers based on their age. The age calculation happens
taking lbolt in account.

Due to lbolt wrapping the buffers were not flushed. So, a lot of metadata IO's
stopped and hence, the panic.

RESOLUTION:
In the function for handling flushing of dirty buffers, also handle 
the condition if lbolt has wrapped. If it has then assign current lbolt time
to the last update time of dirtylist.



INSTALLING THE PATCH
--------------------
1. To install the VxFS 3.5-MP4RP8 patch:
a) To install this patch on a CVM cluster, install it one
 system at a time so that all the nodes are not brought down
 simultaneously.
b) The VxFS 3.5(GA) must be installed before applying these
  patches.
c) To verify the VERITAS file system level, execute:
     # swlist -l product | egrep -i 'VRTSvxfs'
  VRTSvxfs   3.5-ga15       VERITAS File System with CFS Support
Note: VRTSfsman is a corequisite for VRTSvxfs. So, VRTSfsman also
needs to be installed with VRTSvxfs.
d) All prerequisite/corequisite patches must be installed. The Kernel patch
  requires a system reboot for both installation and removal.
e) To install the patch, execute the following command:
# swinstall -x autoreboot=true -s <patch_directory>  PHKL_43921
If the patch is not registered, you can register it
using the following command:
# swreg -l depot <patch_directory> 
The <patch_directory>  is the absolute path where the patch resides.


REMOVING THE PATCH
------------------
To remove the VxFS PHKL_43091  patches:
a) Execute the following command:
# swremove -x autoreboot=true  PHKL_43921


SPECIAL INSTRUCTIONS
--------------------
Before installing 3.5MP4RP8 patch, it requires PHCO_32640, PHCO_31903, PHKL_23337 to be installed on the machine.


OTHERS
------
1. MP4RP6
==========
1098031 [Willow5.0] JAGag16867 Data corruption following filesystem full
1098034 Cannot tune max_direct_iosz properly
1098044 vxfs limits direct i/o size to 256Kb
1098046 HP-UX CFS rolling upgrade issue - Node disabled only when we have 2 different patch level between the nodes
1098108 HPUX VxFS 4.1 CP2 - CFS filesystem hung
2163345 Page i/o error occurs under memory pressure. Request porting itools 153642 to vxfs 3.3/11.11
2363564 vx_trunc_tran2 gets ENOSPC from vx_te_bmap_enter
MP4RP5
------
1525215 HP-UX 11.11; JFS 3.3; poor performance writing to mmap'ed sparse file.
1537541 corrupt data read after sendfile()
1650658 vx_lookup returning ENOENT due to stale negative DNLC entry
1801154 System panic in inctext: VTEXT not set and tcount >  0
2128004 quotacheck coredumps with more than 30 quota- enabled filesystems in /etc/fstab
2246305 vx_rdwr should handle the EIO returned from vx_tranidflush
MP4RP4
------
1478170 write() system call hangs for over 10 seconds on VxFS 3.5 on 11.23
1503804 invalid ACLs can be set on HP
1588603 deadlock in VxFS-HP CFS code and need a patch urgently
MP4RP3
=======
1517666 ncheck command dumps core
MP4RP2
======
1106989 HP 11.23 LM Stress hit 'f:vx_freeze_block_threads_all:7c' with 5.0A33 build 
1107017 HP-SPT(JAGag00037): VxFS4.1 vxfsconvert fails to convert vxfs disk layout 3 filesystem 
1107033 fsck must be able to run as a switchout from 5.0 fsck 
1108419 Poor random read through buffer cache 
1114008 vx_getmount() returns wrong info for largefiles/nolargefiles 
MP4RP1
=======
893644 IO error returned by dd when we dd to a file with a full file in a checkpoint 
893645 data corruption on rw mounted clone with 2 other rw mounted clones 
893646 vxumount fails to force unmount a nested filesystem when the underlying filesystem is unavailable 
893647 3.5 fsck reports structural inode 70 unreferenced for 4.1 mkfs -oversion=4 filesystem 
893648 DB2 vxfs corruption. 
893649 vx_delxwri_flush() threads all stuck on same hung inode 
893653 fsadm -d -D -e -E -s core dumps