* * * READ ME * * *
               * * * Veritas File System 5.0.1RP3P7 * * *
                          * * * P-patch * * *
                         Patch Date: 2012-10-19


This document provides the following information:

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


PATCH NAME
----------
Veritas File System 5.0.1RP3P7 P-patch


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


BASE PRODUCT VERSIONS FOR THE PATCH
-----------------------------------
   * Veritas File System 5.0.1


OPERATING SYSTEMS SUPPORTED BY THE PATCH
----------------------------------------
HP-UX 11i v3 (11.31)


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

Patch ID: PHKL_43260, PHCO_43261

* 1935624 (Tracking ID: 1903977)

SYMPTOM:
On a Cluster File System (CFS) the write operation may fail and the system 
panics with the following stack trace:

vx_active_common_flush+0x74()
vx_rwlock+0x14()

DESCRIPTION:
When a check is done to verify if the VX_DLSYNCFREE flag is set inside the 
function vx_getblk_clust(), there is no lock to protect the check. As a result, 
there could be a race condition leading to inconsistent data, this causes the 
system to panic.

RESOLUTION:
The code is modified to synchronize the VX_DLSYNCFREE flag updates using the 
VX_FSQ_LOCK.

* 1935635 (Tracking ID: 1742707)

SYMPTOM:
Mounting a Cluster File System (CFS) fails with the following usage message:
UX:vxfs fsck_logv<x>: INFO: V-3-20896: Usage:
fsck [-V] [-F vxfs] [-mnNyY] [-o fu'll, nolog, mounted, p]
special [...]

DESCRIPTION:
The "switchout fsck" command needs to be invoked for CFS with two separate 
options: "-o" and "mounted". However, the "switchout fsck" command gets invoked 
with the "-o mounted" option. As a result, the error occurs.

RESOLUTION:
The code is modified so that the "switchout fsck" command gets invoked with two 
different options: "-o" and "mounted", instead of being invoked as the "-o 
mounted" option.

* 1954685 (Tracking ID: 1934537)

SYMPTOM:
The reverse-name-lookup operation on an inode may panic the machine with the 
following stack trace:
vx_free
vx_traverse_tree+0x4a0
vx_dir_lookup+0x1e4
vx_rev_namelookup+0x294
vx_aioctl_common+0xac4
vx_aioctl+0x12c
vx_ioctl+0xe0

DESCRIPTION:
During the lookup operation if the memory allocation fails, the user still goes 
ahead and adds it to the used memory, and later tries to free that memory. This 
results in the panic.

RESOLUTION:
The code is modified to update the usage counters correctly, and skip updating 
the count during the error condition. An additional check is added to free only 
the non-null buffers.

* 2798208 (Tracking ID: 2767579)

SYMPTOM:
The system hangs during a lookup operation with the following stack trace:
vx_dnlc_pathname_realloc+0x80 ()
vx_dnlc_getpathname+0xcf0 ()
audit_get_pathname_from_dnlc+0x370 ()
audit_dnlc_path_name+0x70
ftruncate+0x480 ()
syscall+0x590 ()

DESCRIPTION:
The system hangs because of an infinite loop that gets triggered when an inode
with the negative DNLC entry is encountered, during a reverse name DNLC lookup.

RESOLUTION:
The code is modified to add an avoidance fix to prevent the infinite loop to occur.

* 2822988 (Tracking ID: 2822984)

SYMPTOM:
When the extendfs(1m) command extends the file system that is greater than 2TB
the extendfs(1m)  command fails and the following error message is displayed:
"UX:vxfs fsck: ERROR: V-3-25315: could not seek to block offset"

DESCRIPTION:
This is a typecasting problem. When the extendfs(1m) command tries to extend the
file system, the bs_bseek() function is invoked. The bs_bseek()  function's
return type is a 32 bit integer value. This value gets negative for offsets
greater than 2TB and results in failure.

RESOLUTION:
The code is modified to resolve the typecasting problem.

* 2852520 (Tracking ID: 2850738)

SYMPTOM:
The system may hang with the following stack trace during the low memory 
condition:
swtch_to_thread(inlined)
slpq_swtch_core+0x520   real_sleep(inlined)
sleep+0x400 
mrg_reserve_swapmem(inlined)
$cold_steal_swap+0x460
 $cold_kalloc_nolgpg+0x4b0 
 kalloc_internal(inlined)
$cold_kmem_arena_refill+0x650
 kmem_arena_varalloc+0x280 
vx_alloc(inlined)vx_worklist_enqueue+0x40 
 vx_buffer_kmcache_callback+0x160
 kmem_gc_arena(inlined)
foreach_arena_ingroup+0x840
kmem_garbage_collect_group(inlined)
kmem_garbage_collect+0x390 
 kmem_arena_gc+0x240 
 kthread_daemon_startup+0x90

DESCRIPTION:
The VxFS kernel memory callback() routine allocates memory with M_WAITOK flag. 
This results in the system hang during the low memory condition as the callback
() routine waits for memory allocation.

RESOLUTION:
The code is modified to allocate memory without waiting in the VxFS kernel 
memory callback() routine.

Patch ID: PHKL_43062

* 2036217 (Tracking ID: 2019793)

SYMPTOM:
While umounting the file system the system may panic and the following stack 
trace is displayed:
vx_set_tunefs+0x264()
vx_aioctl_full+0xc7c()
vx_aioctl_common+0x738()
vx_aioctl+0x13c()
vx_ioctl+0xe4()
syscall+0xcc()

DESCRIPTION:
Due to a race condition during the umount operation to update fse_fs and the 
fse_zombie pointers, there may be a small window where both the pointers are 
out of sync, which results in the panic.

RESOLUTION:
The code is modified to set the fsext->fse_zombie pointer before the fsext-
>fse_fs pointer to keep them consistent.

* 2043627 (Tracking ID: 2028782)

SYMPTOM:
For a file managed by Hierarchical Storage Management (HSM), application file 
quota gets doubled after an HSM migrate or recall process.DESCRIPTION: HSM file 
quota get doubled after HSM migrate/recall process.

DESCRIPTION:
A file has a set of quota when it is created. For efficient storage, the data 
is moved from the high speed disk to the low speed tape using HSM. This process 
is called HSM migrate. When the data is recalled from the tape to the disk, the 
HSM quota for the file gets updated and doubled.

RESOLUTION:
The code is modified to handle quota updates for files managed by HSM correctly.

* 2410793 (Tracking ID: 1466351)

SYMPTOM:
Mount hangs in vx_bc_binval_cookie like the following stack,
delay
vx_bc_binval_cookie
vx_blkinval_cookie
vx_freeze_flush_cookie
vx_freeze_all
vx_freeze
vx_set_tunefs1
vx_set_tunefs
vx_aioctl_full
vx_aioctl_common
vx_aioctl
vx_ioctl
genunix:ioctl
unix:syscall_trap32

DESCRIPTION:
The hanging process is waiting for a buffer to be unlocked. But that buffer can
only be released if its associated cloned map writes get flushed. But a
necessary flush is missed.

RESOLUTION:
Add code to synchronize cloned map writes so that all the cloned maps will be
cleared and the buffers associated with them will be released.

* 2429335 (Tracking ID: 2337470)

SYMPTOM:
The Cluster File System (CFS) can unexpectedly and prematurely report a 'file 
system 
out of inodes' error when attempting to create a new file. The following error 
message 
is displayed:
vxfs: msgcnt 1 mesg 011: V-2-11: vx_noinode - /dev/vx/dsk/dg/vol file system out
of inodes.

DESCRIPTION:
While allocating new index nodes (inodes) in a CFS, Veritas File System (VxFS) 
searches for an available free inode in the Inode Allocation Units (IAUs) that 
are delegated to the local node. If none are available, it searches the IAUs 
that are not delegated to any node, or revokes an IAU delegated to another 
node. Gaps may be created in the IAU structures as a side effect of the CFS 
delegation processing. However, while searching for an available free inode, if 
VxFS ignores any gaps, new IAUs cannot be created if the maximum size of the 
metadata structures reaches (2^31). Therefore, one of the gaps must be 
populated and used for the allocation of the new inode. If the gaps are 
ignored, VxFS may prematurely report the "file system out of inodes" error 
message even though there is enough free space in the VxFS file system to 
create new inodes.

RESOLUTION:
The code is modified to allocate new inodes from the gaps in the IAU structures 
created as a part of the CFS delegation processing.

* 2730965 (Tracking ID: 2730759)

SYMPTOM:
The sequential read performance is poor because of the read-ahead issues.

DESCRIPTION:
The read-ahead on sequential reads performed incorrectly because of wrong read-
advisory and the read-ahead pattern offsets are used to detect and perform the 
read-ahead. Also, more sync reads are performed which can affect the 
performance.

RESOLUTION:
The code is modified and the read-ahead pattern offsets are updated correctly 
to detect and perform the read-ahead at the required offsets. The read-ahead 
detection is also modified to reduce the sync reads.

* 2796940 (Tracking ID: 2599590)

SYMPTOM:
Expansion of a 100% full file system may panic the machine with the following 
stack trace.
 
bad_kern_reference()
$cold_vfault()
vm_hndlr()
bubbledown()
vx_logflush()
vx_log_sync1()
vx_log_sync()
vx_worklist_thread()
kthread_daemon_startup()

DESCRIPTION:
When 100% full file system is expanded intent log of the file system is 
truncated and blocks freed up are used during the expansion. Due to a bug the 
block map of the replica intent log inode was not getting updated thereby 
causing the block maps of the two inodes to differ. This caused some of the in-
core structures of the intent log to go NULL. The machine panics while de-
referencing one of this structure.

RESOLUTION:
Updated the block map of the replica intent log inode correctly. 100% full file 
system now can be expanded only If the last extent in the intent log contains 
more than 32 blocks, otherwise fsadm will fail. To expand such a file-system, 
some of the files should be deleted manually and resize be retried.

* 2798203 (Tracking ID: 2316793)

SYMPTOM:
After removing the files in a file system, the df(1M)command which uses the 
statfs(2)function may take 10 seconds to complete.

DESCRIPTION:
To obtain an up-to- date and valid free block count in a file system a delay 
and retry loop delays for one second and retries 10 times. This excessive 
retrying causes a 10 second delay per file system while executing the df(1M) 
command.

RESOLUTION:
The code is modified to reduce the original 10 retries with one second delay 
each, to one retry after a 20 millisecond delay.

* 2806468 (Tracking ID: 2806466)

SYMPTOM:
A reclaim operation on a filesystem mounted on a Logical Volume Manager (LVM)
volume using the fsadm(1M) command with the 'R' option may panic the system and 
the following stack trace is displayed:
vx_dev_strategy+0xc0() 
vx_dummy_fsvm_strategy+0x30() 
vx_ts_reclaim+0x2c0() 
vx_aioctl_common+0xfd0() 
vx_aioctl+0x2d0() 
vx_ioctl+0x180()

DESCRIPTION:
Thin reclamation is supported only on the file systems mounted on a Veritas 
Volume Manager (VxVM) volume.

RESOLUTION:
The code is modified to error out gracefully if the underlying volume is LVM.

* 2831287 (Tracking ID: 1703223)

SYMPTOM:
The internal local mount test exits due to full fsck operation failure.

DESCRIPTION:
While testing, if some directory inode is set with the extended VX_IEREMOVE 
operation it must mark the files from the directory with the extended 
operation. If all the files inside that directory are not set with the 
VX_IEREMOVE operation, it results in a number of unreferenced files and the 
test fails.

RESOLUTION:
The code is modified to remove the file entries if the directory inode has 
VX_IEREMOVE operation set.

* 2832560 (Tracking ID: 2829708)

SYMPTOM:
On a locally mounted Veritas File System (VxFS) machine, the system hangs due 
to low memory during an internal test.

DESCRIPTION:
The system hangs because the memory allocated to the structure used to enqueue 
or dequeue work items in batches, is not freed.

RESOLUTION:
The code is modified to free the memory.

* 2847808 (Tracking ID: 2845175)

SYMPTOM:
When the Access Control List (ACL) feature is enabled, the system may panic 
with "Data Key Miss Fault in KERNEL mode" error message in the vx_do_getacl() 
function and the following stack trace is displayed:
vx_do_getacl+0x840 ()
vx_getacl+0x70 ()
acl+0x480 ()

DESCRIPTION:
In the vx_do_getacl() function, a local variable is accessed without being 
initializing as a result leading to a panic.

RESOLUTION:
The code is modified to initialize the local variable to NULL before using it.

Patch ID: PHKL_42892

* 1946124 (Tracking ID: 1797955)

SYMPTOM:
In a setup that involves files with large extents (greater than 32 MB), which 
have encountered state map corruptions previously, the file system is disabled 
and the following message is displayed in the syslog:
WARNING: msgcnt 3 mesg 037: V-2-37: vx_metaioerr - vx_tflush_1 -
 /dev/vx/dsk/<vol group name>/<vol name> file system meta data write error in 
dev/block ?/????

DESCRIPTION:
The file system is disabled after an existing corruption is discovered. 
According to the VxFS I/O error policy, the file system must be disabled only 
when a read/write operation fails with an I/O error to prevent further 
corruption and not when an existing corruption is discovered.

RESOLUTION:
The code has been modified such that the file system is not disabled in case of 
a discovered corruption.

* 1995390 (Tracking ID: 1985626)

SYMPTOM:
The system panics during multiple parallel unmount operations. The following 
stack trace is displayed:
vx_freeze_idone_list+24c 
vx_workitem_process+10
vx_worklist_process+344
vx_worklist_thread+94

DESCRIPTION:
During simultaneous unmount operations, a race condition occurs 
between the INODE_DEINIT() and FREEZE_IDONE_LIST() functions. While the 
INODE_DEINIT() function uses locks, the FREEZE_IDONE_LIST() function does not 
use locks while updating a pointer value. Hence, there is a possibility of a 
pointer having a NULL value, which gets de-referenced. This results in a panic.

RESOLUTION:
The code is modified to serialize the inode de-initialization operation.

* 2036843 (Tracking ID: 2026799)

SYMPTOM:
A hang occurs while enforcing the policy on File Change Log (FCL) inode and the 
following stack trace is displayed:
lwp_park() 
cond_wait_queue ()
cond_wait ()
pthread_cond_wait()
ts_drain_masters()
do_walk_ilist()
do_process() 
do_enforce() 
main()    
_start()

DESCRIPTION:
While enforcing the policy on FCL, the file system is frozen. After enforcing 
the policy on FCL, a function is called which checks whether the file system is 
frozen. If it is frozen, the function returns the EACTIVERETRY error. This 
error gets percolated to further functions. The AIOCTL_COMMON() function 
performs the retry operation endlessly, resulting in a hang.

RESOLUTION:
The code is modified to return the EACTIVERETRY error only if the file system 
is not in the frozen state.

* 2084004 (Tracking ID: 2040647)

SYMPTOM:
Quota on cluster mounted system does not enforce hard limit.

DESCRIPTION:
The information on the hard limit is stored in an unsigned integer. When a 
larger value is subtracted from it, it wraps around and stores an incorrect 
value.

RESOLUTION:
The code is modified to handle the wrap around and special code is added to 
identify the quota soft limit in the Cluster File System (CFS) environment.

* 2092072 (Tracking ID: 2029556)

SYMPTOM:
When the file system is full, it is not possible to remove a file which has 
more than 13 hard links. This may lead to a panic with the following stack 
trace:
panicsys()
vpanic()
panic()
mutex_panic()
vx_iunlock()
vx_remove_tran()
vx_do_remove()
vx_remove1()
vx_remove()
vn_remove()
unlink()
syscall_trap32()

DESCRIPTION:
While removing or updating an inode, a new attribute inode is allocated which 
is modified. When the file system is full, no new attribute inodes can be 
allocated and the above operation may panic the system.

RESOLUTION:
The code is modified such that the file system can modify the attribute inode 
rather than allocating new attribute inode.

* 2194629 (Tracking ID: 2161379)

SYMPTOM:
In a Cluster File System (CFS) environment, various file system operations hang 
with the following stack trace:
T1:
vx_event_wait()
vx_async_waitmsg()
vx_msg_send()
vx_iread_msg()
vx_rwlock_getdata()
vx_glm_cbfunc()
vx_glmlist_thread()

T2:
vx_ilock()
vx_assume_iowner()
vx_hlock_getdata()
vx_glm_cbfunc()
vx_glmlist_thread()

DESCRIPTION:
Due to improper handling of the ENOTOWNER error in the ireadreceive() function, 
the operation is retried repeatedly while holding an inode lock. All the other 
threads are blocked, thus causing a deadlock.

RESOLUTION:
The code is modified to release the inode lock on the ENOTOWNER error and 
acquire it again, thus resolving the deadlock.

* 2222508 (Tracking ID: 2192895)

SYMPTOM:
A system panic occurs when executing File Change Log (FCL) commands and the 
following stack trace is displayed:
panicsys()
panic_common()
panic()
vmem_xalloc()
vmem_alloc()
segkmem_xalloc()
segkmem_alloc_vn()
vmem_xalloc()
vmem_alloc()
kmem_alloc()
vx_getacl()
vx_getsecattr()
fop_getsecattr()
cacl()
acl()
syscall_trap32()

DESCRIPTION:
The Access Control List (ACL) count in the inode can be corrupted due to a race 
condition. For example, the setacl() function can change the ACL count when the 
getacl() function is processing the same inode. This results in an incorrect 
ACL count.

RESOLUTION:
The code is modified to add protection to the vulnerable ACL count to avoid 
corruption.

* 2370061 (Tracking ID: 2370046)

SYMPTOM:
Read ahead operations miss to read early blocks of data when the value 
of "read_nstream" tunable is not set to 1.
PROBLEM

DESCRIPTION:
The read ahead operation reads the file on demand and does not read the portion 
of the file which is to be read in advance. This occurs because the parameter 
that determines the next read ahead offset is incorrectly reset.

RESOLUTION:
The code is modified so that the read ahead length is set correctly.

* 2720002 (Tracking ID: 1396859)

SYMPTOM:
The internal spin watcher tools show heavy contention on the buffer freelist 
lock, bc_freelist_lock, even when the I/O loads are predominantly direct I/Os.

DESCRIPTION:
Direct I/O data need not be cached in file system buffers. But Veritas File 
System (VxFS) maintains empty buffers to track these requests. Hence, 
contention is seen even when I/O is direct in nature.

RESOLUTION:
The code is modified to have a separate arena to track direct I/O buffers so 
that the contention on the buffer freelist lock is reduced.

* 2722869 (Tracking ID: 1244756)

SYMPTOM:
A lookup operation on a VxFS file system may fail with the following stack 
trace:
vx_cbdnlc_purge_iplist+0x64
vx_inode_free_list+0x160
vx_ifree_scan_list+0xf8
vx_workitem_process+0x10
vx_worklist_process+0x17c
vx_worklist_thread+0x94

DESCRIPTION:
Due to a race condition between the Directory Name Lookup Cache (DNLC) lookup 
and the DNLC get functions, there is an attempt to move a DNLC entry to the 
tail of the freelist in the lookup function when it has already been removed 
from the freelist by the DNLC get function. This leads to a null pointer de-
reference.

RESOLUTION:
The code is modified to verify that the DNLC entry is present on the freelist 
before it is moved to the tail by the DNLC get function.

* 2722958 (Tracking ID: 2696067)

SYMPTOM:
When a getaccess() command is issued on a file which inherits the default 
Access Control List (ACL) entries from the parent, it shows incorrrect group 
object permissions.

DESCRIPTION:
If a newly created file leverages the ACL entries of its parent directory, the 
vx_daccess() function does not fabricate any GROUP_OBJ entry unlike the 
vx_do_getacl() function.

RESOLUTION:
The code is modified to fabricate a GROUP_OBJ entry.

* 2726001 (Tracking ID: 2371710)

SYMPTOM:
User quota file gets corrupted when DELICACHE feature is enabled and the 
current usage of inodes of a user becomes negative after frequent file 
creations and deletions. If the quota information is checked using the vxquota 
command with the '-vu username ' option, the number of files is "-1".
For example:
    # vxquota -vu testuser2
   Disk quotas for testuser2 (uid 500):
   Filesystem     usage  quota  limit     timeleft  files  quota  limit     
timeleft
   /vol01       1127809 8239104 8239104                  -1      0      0

DESCRIPTION:
This issue is introduced by the inode DELICACHE feature which is a performance 
enhancement to optimize the updates done to the inode map during file creation 
and deletion operations. The feature is enabled by default, and can be changed 
by the vxtunefs(1M) command. 
When DELICACHE is enabled and the quota is set for Veritas File System (VxFS), 
there is an extra quota update for the inodes on the inactive list during the 
removal process. Since this quota has been updated already before being put on 
the DELICACHE list, the current number of user files gets decremented twice.

RESOLUTION:
The code is modified to add a flag to identify the inodes which have been moved 
to the inactive list from the DELICACHE list. This flag is used to prevent 
decrementing the quota again during the removal process.

* 2730957 (Tracking ID: 2651922)

SYMPTOM:
On a local VxFS file system, the ls(1M) command with the '-l' option runs 
slowly and high CPU usage is observed.

DESCRIPTION:
Currently, Cluster File System (CFS) inodes are not allowed to be reused as 
local inodes to avoid Global Lock Manager (GLM) deadlo`ck issue when Veritas 
File System (VxFS) reconfiguration is in process. Hence, if a VxFS local inode 
is needed, all the inode free lists need to be traversed to find a local inode 
if the free lists are almost filled up with CFS inodes.

RESOLUTION:
The code is modified to add a global variable, 'vxi_icache_cfsinodes' to count 
the CFS inodes in inode cache. The condition is relaxed for converting a 
cluster inode to a local inode when the number of in-core CFS inodes is greater 
than the 'vx_clreuse_threshold' threshold and reconfiguration is not in 
progress.

* 2733704 (Tracking ID: 2421901)

SYMPTOM:
Internal stress test on locally mounted VxFS file system resulted in the 
following assert:
 f:vx_getinoquota:1a

DESCRIPTION:
While reusing the inode from inactive inode list, the inode field that contains 
quota information is expected to be NULL. While moving the inode from delicache 
list to inactive inode list, quota information field is not set to NULL.  This 
results in the 
assert.

RESOLUTION:
The code is modified to reset the quota information field of inode while moving 
it from delicache list to inactive inode list.


INSTALLING THE PATCH
--------------------
To install the VxFS 5.0.1-11.31 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 11.31 pkg with revision 5.0.31.5 must be installed before applying the patch.

c) To verify the VERITAS file system level, execute:

     # swlist -l product | egrep -i 'VRTSvxfs'

  VRTSvxfs     5.0.31.5        VERITAS File System

Note: VRTSfsman is a corequisite for VRTSvxfs. So, VRTSfsman also
needs to be installed with VRTSvxfs.

    # swlist -l product | egrep -i 'VRTS'

  VRTSvxfs      5.0.31.5      Veritas File System
  VRTSfsman     5.0.31.5      Veritas File System Manuals

d)All prerequisite/corequisite patches have to 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_43260 PHCO_43261

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 5.0.1-11.31 patch:

a) Execute the following command:

# swremove -x autoreboot=true  PHKL_43260 PHCO_43261


SPECIAL INSTRUCTIONS
--------------------
NONE


OTHERS
------
NONE