 |
» |
|
|
|
|
|
|
|
|
|
Patch Name: PHNE_34672
Patch Description: s700_800 11.11 cumulative ARPA Transport patch
Creation Date: 06/06/04
Post Date: 06/08/24
Warning: 07/03/15 - This Critical Warning has been issued by HP.
- PHNE_33628 introduced behavior that can cause a system
hang. When the behavior occurs, a process enters a
non-interruptible loop in kernel mode and may eventually
consume the system CPUs. The stack trace of the
offending process will be similar to the following:
0x0009446c streams_select+0xc
0x00155b84 soo_select+0x584
0x00303b98 selscan+0x298
0x000f048c select+0x764
0x0014c2b8 syscall+0x738
0x00033f64 syscallinit+0x55c
- This behavior is also exhibited with superseding patches
PHNE_34135 and PHNE_34672.
- Additional information on this behavior may be found in
Service Request 8606454538 (JAGag11178).
- This behavior is corrected in PHNE_35183, which is
released.
Warning: 07/03/19 - This Critical Warning has been issued by HP.
- PHNE_33159 introduced behavior that will cause data
packets sent from a lan interface with an MTU set greater
than the MTU of intermediate routers to silently be
discarded until the associated routing entry is cleaned
up by the inactivity timer causing the TCP MSS to be
updated.
In particular, when receiving ICMP Fragmentation Needed
PDU notification this will not update the TCP MSS for
connections in SYN_SENT state.
- The same behavior is experienced with superseding patches
PHNE_33628, PHNE_34135 and PHNE_34672.
- Additional information on this behavior may be found in
Service Request 8606438224 (JAGaf96324)
- This behavior is corrected in PHNE_35183, which is
released.
- If installing PHNE_35183 is not an option then to work
around this behavior one of the 2 work arounds outlined
below could be considered:
- disable PMTU discovery by setting the ndd(1M)
manageable parameter "ip_pmtu_strategy" to 0 (zero).
Disabling PMTU discovery protects from this behavior
through a system wide disabling of the MTU discovery.
Observe that, based on the network topology, this may
cause network throughput performance issues.
- set the correct PTMU for the affected remote using the
'route -p pmtu' command.
Using 'route -p pmtu' you can set the MTU on a per host
or network basis, see the route(1M) man page. This
however requires 'route -p pmtu' commands to be issued
for each of the hosts and/or networks affected.
Also observe that the settings are not reboot
persistent unless configured in the
"/etc/rc.config.d/netconf" file.
- PHNE_34672 was included in the following HP Patch Bundle:
Dec 2006: GOLDBASE11i,B.11.11.0612.459
Warning: 07/08/06 - This Critical Warning has been issued by HP.
- PHNE_34135 introduced behavior that can result in the
issue described in HP Security Bulletin HPSBUX02247.
- The same behavior is experienced with superseding patch
PHNE_34672 and PHNE_35183.
- The superseded patch, PHNE_33628, does not exhibit this
same behavior.
- This behavior is corrected in PHNE_35351, which is
released. To avoid this behavior, HP recommends installing
PHNE_35351 on all systems.
- PHNE_34672 is included in the following HP Patch Bundles:
Dec 2006: GOLDBASE11i,B.11.11.0612.459
Hardware Platforms - OS Releases:
s700: 11.11
s800: 11.11
Products: N/A
Filesets:
Networking.NET-KRN,fr=B.11.11,fa=HP-UX_B.11.11_32/64,v=HP
Networking.NET-PRG,fr=B.11.11,fa=HP-UX_B.11.11_32/64,v=HP
Networking.NET-RUN,fr=B.11.11,fa=HP-UX_B.11.11_32/64,v=HP
Networking.NET-RUN-64,fr=B.11.11,fa=HP-UX_B.11.11_32/64,v=HP
Networking.NW-ENG-A-MAN,fr=B.11.11,fa=HP-UX_B.11.11_32/64,v=HP
OS-Core.CORE-KRN,fr=B.11.11,fa=HP-UX_B.11.11_32/64,v=HP
ProgSupport.C-INC,fr=B.11.11,fa=HP-UX_B.11.11_32/64,v=HP
Networking.NET2-KRN,fr=B.11.11,fa=HP-UX_B.11.11_32,v=HP
Networking.NMS2-KRN,fr=B.11.11,fa=HP-UX_B.11.11_32,v=HP
OS-Core.CORE2-KRN,fr=B.11.11,fa=HP-UX_B.11.11_32,v=HP
OS-Core.SYS-ADMIN,fr=B.11.11,fa=HP-UX_B.11.11_32,v=HP
Networking.NET2-KRN,fr=B.11.11,fa=HP-UX_B.11.11_64,v=HP
Networking.NMS2-KRN,fr=B.11.11,fa=HP-UX_B.11.11_64,v=HP
OS-Core.CORE2-KRN,fr=B.11.11,fa=HP-UX_B.11.11_64,v=HP
OS-Core.SYS-ADMIN,fr=B.11.11,fa=HP-UX_B.11.11_64,v=HP
Automatic Reboot?: Yes
Status: General Superseded With Warnings
Critical:
Yes
PHNE_34672: PANIC
PHNE_34135: HANG CORRUPTION PANIC
PHNE_33628: HANG PANIC MEMORY_LEAK
PHNE_33159: HANG PANIC
PHNE_32042: HANG PANIC MEMORY_LEAK
PHNE_31247: PANIC MEMORY_LEAK
PHNE_31092: HANG
PHNE_30656: PANIC HANG
PHNE_30053: ABORT PANIC MEMORY_LEAK CORRUPTION HANG
PHNE_29887: OTHER
May cause lan interface to be unconfigured if
cable is disconnected.
PHNE_29472: PANIC
PHNE_28895: PANIC HANG MEMORY_LEAK
PHNE_28497: PANIC MEMORY_LEAK
PHNE_27989: PANIC MEMORY_LEAK
PHNE_27730: PANIC HANG
PHNE_27382: PANIC HANG MEMORY_LEAK CORRUPTION
PHNE_27063: PANIC
PHNE_26762: PANIC
PHNE_25644: PANIC
PHNE_25642: PANIC
PHNE_25134: PANIC
PHNE_24506: PANIC
PHNE_24211: PANIC HANG ABORT
PHNE_23645: PANIC
PHNE_23288: PANIC HANG CORRUPTION MEMORY_LEAK
Category Tags:
defect_repair enhancement general_release critical panic
halts_system corruption memory_leak manual_dependencies
Path Name: /hp-ux_patches/s700_800/11.X/PHNE_34672
Symptoms:
PHNE_34672:
( SR:8606420692 CR:JAGaf80521 )
When the numerical value of ndd tunable
`ip_ire_flush_interval' is set to less than the numerical
value of ndd tunable `ip_ire_clean up_interval', network
routing table entries for link local addresses are never
deleted, as they should be.
( SR:8606416944 CR:JAGaf76777 )
When system memory usage is very high and programs that call
ioctl(2), using any one of SIOCSWITCHIF6ADDR,
SIOCSWITCHLIF6ADDR, SIOCSLIFADDR, SIOCSLIFNETMASK,
SIOCSLIFFLAGS, SIOCGLIFFLAGS, SIOCSLIFADDR as the 2nd
parameter, get an ENOMEM error, subsequent attempts mostly
return EINVAL.
( SR:8606421454 CR:JAGaf81284 )
When the numerical value of ndd tunable
`ip_ire_redirect_interval' is set to zero,network routing
table entries that should be periodically removed, are not
removed.
( SR:8606426280 CR:JAGaf85761 )
Application program `snmpwalk' reports an incorrect value
for the UDP SNMP MIB object `udpOutDatagrams'.
( SR:8606415076 CR:JAGaf74930 )
When PMTU discovery is enabled (non-zero value of the ndd
tunable parameter 'ip_pmtu_strategy') and PMTU is less
than 120 bytes, if there is a TCP connection
having unacknowledged data for more than the first
threshold timer (can be modified using the ndd tunable
parameter 'tcp_ip_notify_interval'), there can be an
endless exchange of unfragmented TCP retransmits and ICMP
'Destination Unreachable Fragmentation Needed'
messages between the sending host and the
intermediate router.
( SR:8606424562 CR:JAGaf84085 )
1. Help texts for ndd tunables `arp_defend_interval' and
`arp_redefend_interval' erroneously show the applicable
units as seconds. It should be milliseconds.
2. Setting the ndd tunable `arp_redefend_interval' to zero
does NOT prevent a published entry from being
redefended as, erroneously, indicated by the relevant
ndd help text.
( SR:8606429097 CR:JAGaf88563 )
System panics with the following stack trace :
ire_lookup+0x190
ire_lookup_loop+0x48
ipc_tcp_bind+0x78
ipc_bind+0x8c
ip_wput_nondata+0x38
ip_wput+0x15c
putnext+0xcc
tcp_connect+0x28c
tcp_wput_proto+0xbc
tcp_wput+0x428
putnext+0xcc
str_async_ioctl+0x224
hpstreams_ioctl_int+0x11c
streams_ioctl+0x34
soconnect+0x158
connect+0xec
syscall+0x750
syscallinit+0x55c
( SR:8606438456 CR:JAGaf96535 )
After running the command `route -f', the loopback
route 127.0.0.0 is deleted from network routing tables,
even though they should not be.
( SR:8606438648 CR:JAGaf96709 )
When HP-UX clients send data along with the last ACK of
the TCP 3-way handshake, some servers reject the
connection request.
( SR:8606428509 CR:JAGaf87990 )
The system panics when the nettl is enabled.
( SR:8606431587 CR:JAGaf91029 )
In some situations, connect(2) returns EINPROGRESS, even
when the connection has been rejected by the peer. This
misleads applications about the actual connection status.
( SR:8606441063 CR:JAGaf98824 )
Some TCP segments may be retransmitted
PHNE_34135:
( SR:8606406155 CR:JAGaf66063 )
Panic in ip_rput with the following
stack trace:
ip_rput+0x6f4
putnext+0xcc
putnext2+0x30
hp_dlpi_mblk_fast_in+0x43c
hp_dlpi_mblk_intr_put+0x964
streams_put+0xdc
streams_put2+0x40
hp_dlpi_mblk_intr+0x250
lan_inbound_pkts+0x7f0
ixgbe_ilan_recv_frames+0x3d4
ixgbe_ilan_recv_and_replenish+0x114
ixgbe_recv+0x310
ixgbe_isr+0x68
sapic_interrupt+0x2c
mp_ext_interrupt+0x2d0
ihandler+0x90c
( SR:8606409272 CR:JAGaf69170 )
For AF_UNIX domain and SOCK_STREAMS
type sockets when a thread is blocked
on recv(2) and the peer performs orderly
shutdown using shutdown(2), thread blocked on
recv(2) will not return with zero
as return value.
( SR:8606413150 CR:JAGaf73012 )
Certain fields of tcp and ip instance
data structures may not be
initialized to correct values.
( SR:8606416735 CR:JAGaf76568 )
When the path MTU drops to a value below 120,
TCP segments with corrupted sequence numbers
are transmitted on checksum offload capable
network cards.
( SR:8606424136 CR:JAGaf83659 )
Eager connections not flushed under certain
circumstances.
( SR:8606416402 CR:JAGaf76253 )
traceroute(1) man page refers to HP-UX OS
as non-standard OS.
( SR:8606404400 CR:JAGaf64324 )
setsockopt(2) man page needs changes.
( SR:8606424222 CR:JAGaf83745 )
Problem with ndd help text localization.
( SR:8606428254 CR:JAGaf87735 )
Sendmail process may hang in connect() code path.
The stack trace of these processes looks like this:
LVL FUNC ARG0 ARG1 ARG2
0) _swtch+0xd0 0x1537b0 n/a n/a
1) _sleep_one+0x214 0x4b547608 0x1029a
0x4b547600
2)so_check_inb_conn_
conf+0xd8 0x4b5474c0 n/a n/a
3)connect+0x54 0x400003ffffff03a0 n/a n/a
4)syscall+0x768 n/a n/a n/a
5)syscallinit+0x55c n/a n/a n/a
( SR:8606432060 CR:JAGaf91502 )
sendfile(2) does not return -1 with appropriate
errno under some failure conditions.
( SR:8606431462 CR:JAGaf90904 )
System hangs with the following stack trace:
spinunlock+0x50
hpstreams_select_int2+0x680
streams_select2+0x30
soo_select+0x880
selscan+0x270
select+0x4d0
syscall+0x920
_u_default+0xd48d2c0
PHNE_33628:
( SR:8606322228 CR:JAGae84696 )
TCP, IP and ARP modules can spend too much time
processing the Synch Queue Element resulting
in monopolising a CPU or it may lead to Service
Guard TOC.
( SR:8606352228 CR:JAGaf13033 )
When there are two different routes using the same
gateway but with a different PMTU, the route created
for the gateway inherits the wrong PMTU.
( SR:8606381553 CR:JAGaf41766 )
Process hang when threads call vfork(2).
( SR:8606385089 CR:JAGaf45243 )
shutdown(2) done on a listening socket from one thread does
not wake up other threads sleeping in accept(2) system call
on the same socket.
( SR:8606387073 CR:JAGaf47226 )
A socket that has been created by doing accept(2) on a
listen socket that has already been shutdown for
"write side", when used to send data either through
write(2)/send(2) will generate SIGPIPE.
( SR:8606387074 CR:JAGaf47227 )
TCP keeps accepting new connections for a listen socket
even after it has been shutdown for read side and
shutdown(2) returned successfully.
( SR:8606387075 CR:JAGaf47228 )
When shutdown(2) is called on a listen socket that has
connections pending to be accepted, it returns successfully.
But new incoming connections continue to be accepted by TCP
code for this listen socket though shutdown(2) returned
success.
( SR:8606387098 CR:JAGaf47251 )
When more than two threads are sleeping in accept(2) with
no connections pending to be accepted, and if
write side(SHUT_WR) shutdown(2) is performed on
listen socket, shutdown(2) wakes up only first two threads.
The remaining threads hang forever in accept(2).
( SR:8606387472 CR:JAGaf47625 )
shutdown(2) issued for read side on a listening socket does
not abort any connections pending to be accepted. The peer
of such pending connections may continue to send data
successfully
( SR:8606388101 CR:JAGaf48253 )
shutdown on read side(SHUT_RD) of listen socket does not
wake up any threads that are sleeping on accept(2).
( SR:8606388102 CR:JAGaf48254 )
shutdown(2) on listen socket might fail with EINVAL under
certain circumstances.
( SR:8606390716 CR:JAGaf50862 )
System can panic with following stack trace:
stack trace for event 0
crash event was a panic
panic+0x6c
report_trap_or_int_and_panic+0x94
trap+0x910
nokgdb+0x8
tcp_rput_context_check+0x10
tcp_rput+0x340
puthere+0xc8
tcp_conn_ind+0x65c
tcp_rput+0x47c
puthere+0xc8
tcp_conn_ind+0x65c
tcp_rput+0x47c
puthere+0xc8
..
..
( SR:8606390906 CR:JAGaf51051 )
ndd help text for "socket_caching_tcp" doesn't match
with the implementation.
( SR:8606392043 CR:JAGaf52175 )
select(2) on listen socket that is currently sleeping for a
"READ" event to occur does not get woken up when the socket
is shutdown for read side (SHUT_RD).
( SR:8606392816 CR:JAGaf52902 )
When multiple listeners are listening on same port/IP
address and multiple connections exceeding the listen
backlog is made, only the first listeners connections are
accepted. Rest of the connections get timed out.
( SR:8606394962 CR:JAGaf54976 )
In some cases ping(1M) displays packet loss for the last
packet even though the packet arrived at the originating
host.
( SR:8606395825 CR:JAGaf55831 )
The ndd(1M)'s help text for ip_pmtu_strategy does not
display all allowable values.
( SR:8606396734 CR:JAGaf56717 )
After an ARP Cache Entry expires, the stale ARP cache entry
is not being used, to send out new packets to a neighbor.
( SR:8606401226 CR:JAGaf61175 )
System can panic with following stack trace:
unp_internalize+0x180
uipc_usrreq+0x6e8
sosend+0x1b00
sendit+0x1ec
sendmsg+0x98
syscall+0xaec
syscallinit+0x55c
( SR:8606404988 CR:JAGaf64908 )
System will hang while using sendfile(2) and the TOC
stack looks like:
_sleep+0x218
getnewbuf_desperate+0x378
getnewbuf+0x50c
ogetblk+0x118
getblk1+0x240
vx_getblk_cmn+0x5c
vx_fgetblk+0x2c
vx_read1+0x2f4
vx_vn_bread+0xec
sosendfile+0x440
sendfile+0x21c
syscall+0x750
syscallinit+0x55c
( SR:8606405138 CR:JAGaf65058 )
EINTR is a valid error return for getsockname(2), but it is
not documented in the manpage of getsockname(2).
( SR:8606411217 CR:JAGaf71099 )
Memory leak in setsockopt(2) while using XOPEN sockets.
( SR:8606412656 CR:JAGaf72518 )
Same IP ID value for different UDP packets of two
connections to the same host.
( SR:8606416090 CR:JAGaf75942 )
System can panic with following stack trace:
stack trace for event 0
crash event was a panic
panic+0x6c
report_trap_or_int_and_panic+0x94
trap+0xefc
nokgdb+0x8
accept+0x5f0
syscall+0xaec
$syscallrtn+0x0
PHNE_33159:
( SR:8606374624 CR:JAGaf34938 )
TOPS appears to be ineffective and the performance of
network intensive applications may become less scalable
due to a CPU which is busy processing interrupts from a
network interface.
( SR:8606389338 CR:JAGaf49486 )
Too much time looping in ICMP code, which could result
in a TCP connection hang or MC/SG TOC. Top of the stack
looks like:
icmp_pkt+0x4
icmp_frag_needed+0x8c
ip_wput_ire+0x1428
ip_wput+0xe8
( SR:8606397158 CR:JAGaf57140 )
After a network card failover, an IP address assigned to
a logical interface can not be pinged from the local system
but can be pinged from external systems.
( SR:8606399900 CR:JAGaf59856 )
ICMP message handling requires modifications.
( SR:8606402124 CR:JAGaf62059 )
Provide a new 'ndd' command option to tune TOPS feature.
( SR:8606403047 CR:JAGaf62982 )
Panic in IPv6 code with the following stack:
ip6_prefix_to_mask+0x4c
ip6_rd_prefix_option+0x60
ip6_rd_advertisement+0x2d4
icmp6_inbound+0xa8
ip6_rput+0x974
putnext+0xcc
putnext2+0x30
hp_dlpi_unitdata_in+0x1400
hp_dlpi_mblk_intr_put+0x1bc
streams_put+0xdc
streams_put2+0x40
hp_dlpi_mblk_intr+0x60
PHNE_32042:
( SR:8606248840 CR:JAGae15237 )
System panics in sounlock() for AF_UNIX sockets
with the following kernel stack trace:
PC-Offset Stack Trace:
0x00092ea8 (panic+0xa0)
0x000932d4 (assfail+0x3c)
0x0009348c (_assfail+0x2c)
0x0008f81c (b_vsema+0x36c)
0x00441dc4 (sounlock+0x974)
0x00445ec8 (mp_socket_unlock+0x10)
0x003ba14c (function name is not available)
0x003b9e40 (pstat_socket+0x400)
0x003bd2b8 (pstat+0x380)
0x003c9584 (syscall+0x834)
0x00033a2c (syscallinit+0x554)
( SR:8606340920 CR:JAGaf01830 )
No answer to IGMP REPORT coming from different
subnet when secondary ip is configured.
( SR:8606346523 CR:JAGaf07347 )
When Linux clients make UDP NFS requests to a
multi-processor HP-UX system, only "nfsd" processes
bound to the processor 0 on the server appear to be
actively handling those requests while other processes
appear to be inactive.
( SR:8606348091 CR:JAGaf08913 )
The "ndd -h tcp_rexmit_interval_initial" command displays
an incorrect default value.
( SR:8606350213 CR:JAGaf11035 )
System panics with the following stack trace:
panic+0x14
report_trap_or_int_and_panic+0x84
interrupt+0x1d4
ihandler+0x928
puthere+0x14
mi_timeout_exec+0x294
sw_service+0xb0
mp_ext_interrupt+0x144
ivti_patch_to_nop3+0x0
idle+0x430
swidle_exit+0x0
( SR:8606351765 CR:JAGaf12570 )
An application is being notified of the presence of
urgent data when a bogus packet has the URG flag set.
( SR:8606355258 CR:JAGaf16012 )
Certain fields in tcp and ip data structures are not
set to correct values when a socket is reused.
( SR:8606355392 CR:JAGaf16106 )
A select(2) call from a local system hangs, if it is
monitoring a non-blocking connect(2) to the peer system
while the peer system is performing a connect(2) back to
the local system.
( SR:8606356802 CR:JAGaf17502 )
When executing traceroute(1) without the '-i' or '-s'
command line options on a multi-homed host, the reply
sent from the intermediate router/destination host in
response to the probes are not being detected by the
host.
( SR:8606358089 CR:JAGaf18788 )
sendfile(2) loops without knowing that a disconnect
indication message has arrived.
( SR:8606361264 CR:JAGaf21954 )
Memory leak when using socketpair(2) call on
AF_UNIX domain sockets.
( SR:8606363737 CR:JAGaf24395 )
IP packets are being sent to the network with
a multicast source address. This is not in
conformance with RFC 1122.
( SR:8606366614 CR:JAGaf27178 )
sendfile(2) returns 0 with the SIGPIPE signal, if the
connection associated with the socket is closed.
( SR:8606369254 CR:JAGaf29798 )
System panics with panic string 'unselect:null thread'
when using AF_CCITT socket with following stack trace:
panic+0x6c
unselect_enh+0x2c
selwakeup+0x114
sbwakeup+0x110
sowakeup+0x20
socantrcvmore+0x60
somayrflush+0x1c
soshutdown+0x614
shutdown+0x34
syscall+0x750
syscallinit+0x554
( SR:8606369649 CR:JAGaf30193 )
Certain nsswitch.conf configurations can delay the execution
of the "netstat -a" command.
( SR:8606370089 CR:JAGaf30518 )
The commands "ndd -h tcp_rexmit_interval_initial_lnp" and
"ndd -h tcp_conn_request_max" display incorrect default
values.
( SR:8606370407 CR:JAGaf30829 )
The netperf program hangs when the "-m 256" option
is specified and Gigabit Ethernet is used.
( SR:8606370902 CR:JAGaf31320 )
TRANSPORT uses STREAMS internal data structures and macros
which it should not.
( SR:8606372704 CR:JAGaf33107 )
When a OOB data arrives for a socket on which SO_OOBINLINE
is set, select(2) for reading does not detect the presence
of OOB data.
PHNE_31247:
( SR:8606366227 CR:JAGaf26833 )
System panics with the following stack trace:
tcp_ip_hdr_len+0x50
tcp_xmit_ctl_def+0x90
tcp_xmit_ctl+0x40
tcp_rput+0xaf0
( SR:8606371947 CR:JAGaf32365 )
When socket caching is enabled, increased memory
consumption in the ALLOCB_MBLK_SM arena is observed.
( SR:8606373330 CR:JAGaf33722 )
Memory leak in the ALLOCB_MBLK_SM arena.
( SR:8606374761 CR:JAGaf35075 )
A memory leak can occur when the socket caching is
set to ON.
PHNE_31092:
( SR:8606366388 CR:JAGaf26992 )
Process is blocked on accept() and is unkillable, with the
following stack trace :
_sleep+0x214
ioctl_sleep+0x318
str_async_ioctl+0x238
str_async_sthinsert_ioctl+0xb0
hpstreams_ioctl_int+0x1e0
streams_ioctl+0x34
soaccept+0x54c
sodequeue+0x108
accept+0xcc
syscall+0x204
syscallinit+0x554
PHNE_30656:
( SR:8606349322 CR:JAGaf10143 )
System panics with following stack trace:
panic+0x14
report_trap_or_int_and_panic+0x84
trap+0xd9c
thandler+0xd24
sbcompress+0x4c
sbappend+0x34
uipc_usrreq+0x9c
sosend+0x440
soo_rw+0x88
( SR:8606344720 CR:JAGaf05570 )
Application may hang in accept(2).
( SR:8606351614 CR:JAGaf12419 )
In a ServiceGuard environment, SIOCSWITCHLIF6ADDR fails
with "No such device or address" error.
( SR:8606353629 CR:JAGaf14423 )
The IPv6 ioctl SIOCGLIFNUM sometimes fails with
ENOENT when socket caching is enabled.
( SR:8606354852 CR:JAGaf15608 )
A race may occur when socket caching is turned off
and sockets are being closed. This may result in a panic.
( SR:8606355807 CR:JAGaf16521 )
An application can get stuck in t_unbind(3)
for an unpredictable amount of time.
PHNE_30053:
( SR:8606293457 CR:JAGae57207 )
"netstat -a" output displays misleading values for
Recv-Q and Send-Q columns.
( SR:8606301096 CR:JAGae64566 )
An application using AF_INET/SOCK_STREAM socket may
experience a killable hang in send(2) system call
when socket caching is enabled. The stack of the
hung application may look like the following:
_swtch+0xc4
_sleep+0x318
write_sleep+0x120
streams_write_uio+0x3b4
sosend+0x4d4
sendit+0x154
send+0x58
syscall+0x204
$syscallrtn+0x0
( SR:8606302874 CR:JAGae66231 )
IP fragments are reassembled incorrectly in some cases.
( SR:8606302876 CR:JAGae66233 )
Problems in IPv6 fragmentation and reassembly.
( SR:8606302878 CR:JAGae66235 )
When a socket is bound to a non-local IPv4 compatible
IPv6 address, it may be unable to accept connections
or data.
( SR:8606304572 CR:JAGae67915 )
System panics with the following stack trace:
ip_bind+0x334
ip_wput_nondata+0x38
ip_wput+0x108
putnext+0xcc
tcp_connect+0x204
tcp_wput_proto+0xc4
tcp_wput+0x574
putnext+0xcc
str_async_ioctl+0x210
hpstreams_ioctl_int+0x548
streams_ioctl+0x34
soconnect+0x140
connect+0xdc
syscall+0x6f8
syscallinit+0x54c
( SR:8606305933 CR:JAGae68981 )
When socket caching is ON, call to bind() can fail with
EINVAL.
( SR:8606306041 CR:JAGae69089 )
ARP requests may hang.
( SR:8606306287 CR:JAGae69335 )
Symptom:
System may get Data Page Fault panic
with following stack trace:
panic n/a n/a
report_trap_or_int_and_panic 0'00000001 n/a
trap n/a n/a
nokgdb n/a n/a
ar_rput 0'00a6b320 0'00838154
( SR:8606306288 CR:JAGae69336 )
Loopbacked packets are occasionally dropped by
TCP and UDP transport protocols.
( SR:8606306299 CR:JAGae69347 )
An unspecified address or a loopback address in IPv6
may be considered as an IPv4 compatible IPv6 address.
( SR:8606306857 CR:JAGae69891 )
System panics with the following stack trace:
freeb+0x18
ire_delete_now+0x6c
ip_dfg_flush+0x158
invoke_callouts_for_self+0xc0
sw_service+0xb0
up_ext_interrupt+0x118
ihandler+0x8c4
( SR:8606307734 CR:JAGae70769 )
The Fast Path feature may be unintentionally disabled
for an interface.
( SR:8606308464 CR:JAGae71498 )
When the interface index specified in the IPV6_PKTINFO
option, an IPv6 packet can be sent out on an interface
which is not the outgoing interface for that particular
destination.
( SR:8606310433 CR:JAGae73305 )
Under heavy load, the system crashes with the following
stack trace:
panic_save_regs_switchstack+0x110
panic+0x350
bad_news+0x9c0
bubbledown+0x0
ipc_walk+0x1f0
ill_down_ind+0xf0
ip_rput_dlpi+0xc60
ip_rput+0x160
putnext+0x100
...
( SR:8606310867 CR:JAGae73725 )
Multicasting might get disabled under some
circumstances.
( SR:8606311142 CR:JAGae73997 )
An unexpected RST is generated during connection
establishment.
( SR:8606311531 CR:JAGae74359 )
System may get Data Page Fault panic with
following stack trace:
panic+0x6c
report_trap_or_int_and_panic+0x94
interrupt+0x4e4
ihandler+0x940
ire6_lookup_src+0x2c8
ip6_newroute+0xa8c
ip6_rput+0x1438
putnext+0x220
pfilstrmodrput+0x54
putnext+0x220
hp_dlpi_intr_fast_put+0x194
kstr_put+0xf0
hp_dlpi_ether_intr+0x414
btlan_receive_frame+0x464
btlan_isr+0x100
sapic_interrupt+0x34
mp_ext_interrupt+0x380
ihandler+0x91c
tracer_interior_landing_pad+0x64
syscallinit+0x55c
( SR:8606312460 CR:JAGae75276 )
"arp -a" or "arp -A" doesn't show all ARP cache entries.
( SR:8606312501 CR:JAGae75317 )
Data page fault in tcp_detach() with the
following stack trace:
panic+0x14
report_trap_or_int_and_panic+0x84
trap+0xd9c nokgdb+0x8
tcp_detach+0x5d8
tcp_close+0xf0
close_wrapper+0x38
csq_protect+0x120
osr_pop_subr+0x214
osr_close_subr+0xf50
hpstreams_close_int+0x31c
streams_close+0x14
soclose+0xf0
soo_close+0x90
closef+0x64
close+0x90
syscall+0x28c
$syscallrtn+0x0
( SR:8606313270 CR:JAGae76068 )
System panic happened after 22 hours when
stress testing repeatedly performed
"ifconfig up/down" operation over inet6
address family. Panic stack trace may
be as follows:
panic+0x6c
assfail+0x3c
vfault+0xc0
trap+0xf10
nokgdb+0x8
ip6_nd_dad_failure+0xc
ip6_nd_cleanup_callback+0xa8
nde_walk+0x50
ip6_nd_cleanup+0x24
ill6_down+0x204
ill6_delete+0x3c4
ip6_close+0x78
close_wrapper+0x44
csq_protect+0x258
osr_pop_subr+0x36c
osr_close_subr+0xb90
osr_unlink+0x1ec
str_plumb_ioctl+0x968
hpstreams_ioctl_int+0x2d0
hpstreams_ioctl+0x50
spec_ioctl+0x22c
vno_ioctl+0x1c0
ioctl+0x168
syscall+0x8a4
$syscallrtn+0x0
( SR:8606313678 CR:JAGae76470 )
ndd(1M) users might find different help texts for
'tcp_conn_strategy' and 'socket_caching_tcp' even though
they are aliased.
( SR:8606317300 CR:JAGae79861 )
Memory leak in 512 bytes bucket.
( SR:8606317596 CR:JAGae80158 )
Systems panics with the following stack trace:
stack trace for event 0
crash event was a panic
panic+0x6c
report_trap_or_int_and_panic+0x94
trap+0xedc
nokgdb+0x8
b_na_loop+0x4
mi_tpi_addr_and_opt+0xa0
x_mi_tpi_uderror_ind+0x84
udp_ud_err+0x40
( SR:8606317678 CR:JAGae80233 )
The nettl(1M) trace for NS_LS_IP and NS_LS_ICMP may
lead to memory corruption in M_DYNAMIC arena.
As a result of this corruption, the system may
get a subsequent Data page fault panic. The traced
ICMP packets may then display an unknown ICMP type.
( SR:8606318033 CR:JAGae80577 )
System panics with the following stack trace
when IP address(A) is moved from interface(A)
to interface(B) and IP address(B) is moved from
interface(B) to interface(A).
q4> trace event 0
stack trace for event 0
crash event was a panic
panic+0x6c
report_trap_or_int_and_panic+0x94
interrupt+0x208
$ihndlr_rtn+0x0
ip_rput+0x54
puthere+0xc8
( SR:8606211448 CR:JAGad80636 )
There is an interoperability problem between
nettl and IPSec.
( SR:8606319251 CR:JAGae81741 )
A process may hang in accept(2) and it is impossible
to kill the process by kill -9. The kernel TOC stack
trace is as follows:
_sleep+0x214
ioctl_sleep+0x684
str_async_ioctl+0x59c
str_async_sthinsert_ioctl+0xb0
hpstreams_ioctl_int+0x17c
streams_ioctl+0x34
soaccept+0x290
sodequeue+0x108
accept+0xcc
syscall+0x204
$syscallrtn+0x0
( SR:8606322230 CR:JAGae84698 )
ifconfig dumps core due to SIGBUS.
( SR:8606322985 CR:JAGae85451 )
In some cases, ICMPv6 Redirect messages are not sent.
( SR:8606323173 CR:JAGae85638 )
System may panic since ip_wput_ire() may send
down malformed OOP headers.
( SR:8606327213 CR:JAGae89474 )
The "netstat -a" command does not display the IP address of
a host if it is not possible to resolve the IP address
into a host name. For example, in the following output,
the hostname corresponding to the port "1011" is left blank.
tcp 0 0 xport12l.login .1011 ESTABLISHED
( SR:8606331109 CR:JAGae92230 )
Unexpected memory consumption of mblks in
ALLOCB_MBLK_SM arena results in memory starvation.
( SR:8606331382 CR:JAGae92502 )
Some IP options set through setsockopt(2) for
RAWIP and UDP sockets cause undesirable behaviors.
( SR:8606333354 CR:JAGae94443 )
recvfrom(2) or recv(2) performed on SOCK_DGRAM
socket with MSG_PEEK flag set returns all the
messages, when only the first message should be
returned.
( SR:8606333610 CR:JAGae94698 )
In some cases, traceroute may return the error message:
"ifaddrlist:SIOCGIFCONF: Invalid argument"
( SR:8606337569 CR:JAGae98576 )
Under some conditions, HP-UX may send an
"ICMP Port Unreachable" error message in reply
to an UDP unicast request.
( SR:8606338847 CR:JAGae99794 )
Some applications doing an ftp transfer over a
TCP connection may hang.
( SR:8606339858 CR:JAGaf00785 )
Execution of "ndd -a" or
"ndd -set sockets tcp_conn_strategy <value>"
may cause the following message to appear on
the console, even though the command succeeded:
"Pid .... received a SIGSEGV for stack growth failure.
Possible causes: insufficient memory or swap space,
or stack size exceeded maxssiz."
( SR:8606342795 CR:JAGaf03689 )
The traceroute(1) man page does not specify that
either the "-i" or the "-s" option may be needed
on the multi-homed host to fill the correct source address
in the probe packets.
( SR:8606343225 CR:JAGaf04119 )
Panics can happen with several different stack traces.
The following two traces are the most likely ones.
tcp_rput+0x413c
puthere+0xc8
tcp_rput_context_check+0x684
tcp_rput+0x328
csq_turnover_with_lock+0x8c
puthere+0x260
mi_timeout_exec+0x2d8
invoke_callouts_for_self+0x9c
sw_service+0x100
mp_ext_interrupt+0x14c
ihandler+0x90c
panic+0x6c
report_trap_or_int_and_panic+0x94
interrupt+0x208
ihandler+0x930
ldtty_rput+0x10
puthere+0xc8
tcp_rput_context_check+0x684
tcp_rput+0x328
csq_turnover_with_lock+0x8cputhere+0x260
mi_timeout_exec+0x2d8
invoke_callouts_for_self+0x9c
sw_service+0x100
mp_ext_interrupt+0x14c
ihandler+0x90c
idle+0x43c
swidle+0x28
( SR:8606344448 CR:JAGaf05298 )
A multithread process may hang, and stay unkillable,
in the accept(2) call with the following stack trace:
read_sleep+0x2c
streams_getmsg+0x168
soaccept+0xb64
sodequeue+0x108
accept+0x12c
syscall+0x62c
syscallinit+0x554
( SR:8606348486 CR:JAGaf09307 )
There is a possible memory leak in sockets in ALLOCB_MBLK_SM
arena when the shutdown(2) is done on a listen socket.
PHNE_29887:
( SR:8606328244 CR:JAGae89894 )
fstat(2) system call on a socket descriptor
occasionally fails with EPIPE, which is invalid
and not documented in fstat(2) man page.
( SR:8606332047 CR:JAGae93167 )
ifconfig unplumbs an interface unexpectedly when
the following sequence of events occur:
1. An interface is properly configured with an
IP address.
2. The cable connected to the interface is pulled
out or a similar event occurs that results in
the interface going down. ifconfig command will
not show the UP flag on the interface now.
3. A user tries to bring the interface forcibly up
by issuing:
ifconfig lan<X> up
When step 3 happens, ifconfig aborts with the error
"up: Resource temporarily unavailable" and then
unplumbs the interface on which the operation was
attempted. After this error, any further attempt
of "ifconfig lan<X> up" will result in
"no such interface" error. This error cannot be fixed
until "ifconfig lan<X> a.b.c.d up" is done with a
proper IP address.
( SR:8606328253 CR:JAGae89895 )
read() returns ECONNRESET instead of ECONNREFUSED.
( SR:8606333990 CR:JAGae95074 )
"ifconfig <interface> inet6 down" followed by a
"ifconfig <interface> inet6 up" after a cable
disconnect and a reconnect will not bring the
interface up.
PHNE_29472:
( SR:8606285107 CR:JAGae49049 )
Disrupted IGMP membership reporting.
( SR:8606320479 CR:JAGae82961 )
send(2) returns an EWOULDBLOCK failure on a
TCP socket when the preceding poll(2) call on the
same socket had returned a POLLOUT revent.
( SR:8606322708 CR:JAGae85175 )
Route entries associated with loopback interfaces
intermittently get deleted.
( SR:8606324000 CR:JAGae86465 )
"ndd -get /dev/ip ip_ire_hash" sometimes
runs too long.
PHNE_28895:
( CR:JAGae34198 SR:8606269959 )
System panics in the functions
hatcp_set_ha_state() and hatcp_get_status(),
if HATCP functionality is enabled.
( CR:JAGae47912 SR:8606283966 )
System panics with the following stack
trace.
stack trace for event 0
crash event was a panic
panic+0x14
sbflush+0x130
sbrelease+0x14
sorflush+0x98
sofree+0x98
soclose+0x1b4
soo_close+0x90
closef+0x64
close+0x90
syscall+0x6f8
$syscallrtn+0x0
( CR:JAGae50362 SR:8606286419 )
send(2) may take a long time to complete for
AF_UNIX/SOCK_STREAM type of sockets.
( CR:JAGae52979 SR:8606289048 )
System panics with the following stack trace:
panic+0x6c
assfail+0x3c
vfault+0xc0
trap+0x110c
nokgdb+0x8
ipc_snmp_get+0x4a4
snmpcom_req+0x288
ip6_wput_nondata+0x1c8
ip6_wput+0x370
putnext+0x130
putmsg_subr+0x5ac
putmsg+0x670
syscall+0x830
$syscallrtn+0x0
( CR:JAGae53258 SR:8606289327 )
System panics with the following stack trace
crash event was a panic
panic+0x6c
report_trap_or_int_and_panic+0x94
trap+0x11c0
nokgdb+0x8
ip_wput_local+0x1314
ip_multicast_loopback+0x40
ip_wput_ire+0xd10
ip_wput+0xd8
putnext+0x130
rawip_wput+0x29c
putnext+0x130
streams_putmsg2+0x41c
streams_putmsg+0x18
sosend+0x222c
sendit+0x438
sendto+0x5c
syscall+0x830
$syscallrtn+0x0
( CR:JAGae54895 SR:8606291052 )
A debug kernel may panic with the following
stack trace:
panic+0x6c
spin_deadlock_failure+0x48
deadlock_check+0x178
sl_pre_check+0xac
spinlock+0x18
check_if_need_to_deliver+0x70
strlog+0x70
x_mi_strlog+0x118
mi_timeout_exec+0x180
invoke_callouts_for_self+0x230
sw_service+0x120
mp_ext_interrupt+0x408
ivti_patch_to_nop3+0x0
cc_spinunlock+0x48
spinunlock+0x6c
vx_itryhold+0xd4
vx_fast_lookup+0x60
vx_lookup+0x210
lookuppnvp+0xb1c
lookuppn+0x60
lookupname+0x40
vn_open+0x16c
copen+0xe8
open+0x44
syscall+0x830
$syscallrtn+0x0
( CR:JAGae54993 SR:8606291150 )
IPv6 link-local address is not auto-
configured correctly over 802.3 media,
which is activated by 'ifconfig snapX
inet6 up'.
( CR:JAGae55495 SR:8606291731 )
System panics with the following stack trace:
stack trace for event 0
crash event was a panic
panic+0x6c
report_trap_or_int_and_panic+0x94
trap+0x117c
nokgdb+0x8
x_mi_timer+0x1c
ip6_nd_solicit+0xb84
ip6_newroute+0xb00
ip6_wput+0x49c
putnext+0x12c
tcp_wput+0x3c0
putnext+0x12c
streams_putmsg2+0x410
streams_putmsg+0x18
sosendfile+0xc28
sendfile+0x258
syscall+0x860
$syscallrtn+0x0
( CR:JAGae58762 SR:8606295065 )
System panics on a debug flavour kernel with
the following stack trace, if HATCP is enabled
and also HATCP logging is enabled:
spin_deadlock_failure+0x80
deadlock_check+0x230
sl_pre_check+0x2a0
spinlock+0x40
check_if_need_to_deliver+0x80
strlog+0xb0
so_get_tcpstateinfo+0x340
sogetopt+0x2060
getsockopt+0x1b0
syscall+0x1740
( CR:JAGae58886 SR:8606295191 )
System panics with the following stack trace
stack trace for event 0
crash event was a panic
panic+0x6c
report_trap_or_int_and_panic+0x94
trap+0xa4c
nokgdb+0x8
CopyInCMsghdrInc+0x54
( CR:JAGae59100 SR:8606295406 )
arp `hostname` does not return any output.
( CR:JAGae59541 SR:8606295895 )
Panic in sogetopt() if HATCP is enabled.
( CR:JAGae60105 SR:8606296535 )
Incorrect value of tcp_was_goodguy in tcp
structure on a standby tcp connection.
( CR:JAGae60136 SR:8606296566 )
The neighbor discovery cache entry structure cannot
be accessed in kernel debugger.
( CR:JAGae60941 SR:8606297439 )
System panics with the following stack
trace.
optcom_req+0xb8
tcp_wput_proto+0xa4
tcp_wput+0x458
( CR:JAGae60957 SR:8606297454 )
"ndd -h tcp_recv_hiwater_max" shows the minimum and
maximum limits for the tunable tcp_recv_hiwater_max
as 1 and 2147483647 respectively. However, the command
"ndd -set /dev/tcp tcp_recv_hiwater_max 2147483647"
returns the error: "operation failed, Invalid argument".
The same error is returned when attempting to set the
tunable udp_recv_hiwater_max, to 1 (which is minimum
limit specified in the online ndd(1M) documentation).
( CR:JAGae61646 SR:8606298146 )
System hangs due to a slow memory
leak in ALLOC_MBLK_SM arena.
( CR:JAGae62145 SR:8606298646 )
When interface index in the PKTINFO structure
is set to zero, packets are sent out using
the interface to which the address specified in
the PKTINFO structure is assigned.
( CR:JAGae62168 SR:8606298669 )
System panics with the following stack trace:
ip_rput+0x9c4
putnext+0xcc
pfilstrmodrput+0x328
putnext+0xcc
putnext2+0x48
hp_dlpi_mblk_fast_in+0x408
hp_dlpi_mblk_intr_put+0x91c
streams_put+0xdc
streams_put2+0x58
hp_dlpi_mblk_intr+0x190
lanc_ether_ics+0xfc
btlan_receive_frame+0x648
btlan_isr+0xf4
sapic_interrupt+0x2c
mp_ext_interrupt+0x26c
ihandler+0x90c
( CR:JAGae62465 SR:8606298969 )
System panics with the following stack trace:
stack trace for event 0
crash event was a panic
panic+0x6c
report_trap_or_int_and_panic+0x94
interrupt+0x4e4
$ihndlr_rtn+0x0
puthere+0x1c
mi_timeout_exec+0x268
invoke_callouts_for_self+0x230
soft_intr_handler+0x15c
sw_service+0x1c
mp_ext_interrupt+0x428
ivti_patch_to_nop3+0x0
idle+0x1518
swidle_exit+0x0
( CR:JAGae62486 SR:8606298990 )
IREs of type IRE_LOOPBACK remain in the
system even after a RST is sent to the
peer TCP.
( CR:JAGae63156 SR:8606299662 )
System may panic in ip6_route_header().
( CR:JAGae63157 SR:8606299663 )
When TCP is uninitialized, an "ICMP
Parameter Problem" Message is not
sent back for an incoming TCP
packet.
( CR:JAGae63166 SR:8606299672 )
The MAC type of the lan device is
incorrectly initialized.
( CR:JAGae63773 SR:8606300304 )
IPv6 module might send erratic options
upstream due to accessing of unmapped
memory.
( CR:JAGae63774 SR:8606300305 )
The assembled packet from out-of-order fragments might
not show the entire packet that was fragmented.
( CR:JAGae63775 SR:8606300306 )
The IPv6 module can panic on reading
unmapped memory while processing neighbor
discovery options.
( CR:JAGae63776 SR:8606300307 )
There is a chance of a spinlock being
released twice while processing
incoming router advertisements which
can lead to a panic.
( CR:JAGae63777 SR:8606300308 )
IPv6 tries to generate a new link level
header for packets destined to a host
on a point to point link.
( CR:JAGae64418 SR:8606300948 )
If the source address specified in the
PKTINFO structure is an IPv4 compatible
IPv6 address, the packet is dropped and
an error message is sent to the application.
( CR:JAGae65387 SR:8606302024 )
The system might not be able to configure
an IPv6 address for an interface with
hardware address length less than 6 bytes.
( CR:JAGae65452 SR:8606302089 )
Packets to node-local multicast addresses
can get dropped.
PHNE_28497:
( CR:JAGae13702 SR:8606247262 )
When socket caching is enabled, certain options set
on a previously used socket are still being set on a
newly created socket.
( CR:JAGae13728 SR:8606247288 )
When socket caching is enabled, setting of SO_RCVBUF
and SO_RCVLOWAT options does not work.
( CR:JAGae34266 SR:8606270028 )
Applications may not receive some ICMP/IGMP
packets that they expect to get from within
the same machine.
( CR:JAGae35688 SR:8606271481 )
When socket caching is ON and a very large
number of TCP connections with unsent
data are closed, a very large number of
context switches are observed.
( CR:JAGae36947 SR:8606272841 )
The support for RFC 1948 can be enabled by
setting the ndd(1M) tunable tcp_isn_passphrase.
However, "ndd -get /dev/tcp tcp_isn_passphrase"
fails with the error:
"operation failed, Permission denied"
( CR:JAGae38431 SR:8606274354 )
The IPv6 Destination Option before the routing
header and after the routing header cannot be set
on a single socket using the sticky option.
( CR:JAGae38572 SR:8606274495 )
TCP connection drops packets and applications may hang
or timeout, when the timestamp value passed in
timestamp option in TCP packet rolls over, after the
connection is established.
( CR:JAGae38973 SR:8606274896 )
An ICMPv6 error message is sent for IPv6 packets
with IPv6 multicast address as the destination
address. This behavior does not conform to
section 2.4 of RFC 2463.
( CR:JAGae39036 SR:8606274959 )
Expended routing headers appear in ICMPv6 echo replies.
( CR:JAGae39438 SR:8606275363 )
UDP packets with a checksum value of zero are not
discarded as per RFC 2460.
( CR:JAGae39439 SR:8606275364 )
System panics with following stack trace:
udp_opt_set+0x112c
optcom_req+0x364
udp_wput_other+0xe8
udp_wput+0x858
putnext+0xcc
str_async_ioctl+0x994
hpstreams_ioctl_int+0x478
streams_ioctl+0x34
sosetopt+0x4cc
setsockopt+0x108
syscall+0x62c
$syscallrtn+0x0
( CR:JAGae41767 SR:8606277697 )
When a UDP socket bound to INADDR_ANY is issued a
connect() followed by a disconnect, the state
is not restored completely. This results in an
incorrect source address being used in subsequent
packets sent out.
( CR:JAGae41788 SR:8606277719 )
IPv6 packets can go on the wire with unspecified or
multicast address in routing headers extension. This
does not conform to RFC2373 section 2.5.2 and RFC2460
section 4.4.
( CR:JAGae42183 SR:8606278124 )
The ONLINK/OFFLINK status of an IPv6 address
(which was autoconfigured by a prefix in
the router advertisement) is not consistent
with the on-link flag in prefix information
received in a newer router advertisement. This
results in the IPv6 address being treated
as off link when it is on-link and vice versa.
( CR:JAGae42942 SR:8606278885 )
"ifconfig lanX" command exits with the error
message "no such interface" when lanX is
configured for IPv6 only.
( CR:JAGae44226 SR:8606280253 )
Applications may hang while doing RAWIP connect.
( CR:JAGae44261 SR:8606280288 )
System panics with the following stack trace:
panic_save_regs_switchstack+0x110
panic+0x460
post_hndlr+0xab0
vm_hndlr+0x270
bubbledown+0x0
spinlock+0x70
puthere+0x60
ar_entry_squery+0x360
ar_cmd_dispatch+0x420
ar_rput+0x290
puthere+0xe0
ipc_opt_set+0xff0
optcom_req+0x6f0
ipc_optmgmt_req+0x150
ip_wput_nondata+0xef0
ip_wput+0x250
putnext+0xf0
udp_wput_other+0x480
udp_wput+0x500
putnext+0xf0
str_async_ioctl+0x460
hpstreams_ioctl_int+0x1d0
streams_ioctl+0x70
sosetopt+0x3a0
setsockopt+0x1f0
syscall+0xab0
( CR:JAGae44304 SR:8606280331 )
Mobile Node fails to tunnel data to Home Agent.
( CR:JAGae45548 SR:8606281609 )
IRE_DEAD flags are not cleared when a dead gateway
comes up.
( CR:JAGae46274 SR:8606282322 )
get_logical_stat() does not return correct
results when the application is compiled
in 64-bit mode.
( CR:JAGae46724 SR:8606282772 )
An unexpected result can be returned to a user
application when getting the IPv6 extension header
in TCP mode.
( CR:JAGae46920 SR:8606282965 )
System panics with the following stack trace:
stack trace for event 0
crash event was a panic
panic+0x6c
report_trap_or_int_and_panic+0x94
trap+0xd78
nokgdb+0x8
ip_wput+0x3b4
putnext+0xf0
rawip_wput+0x1b0
putnext+0xf0
streams_putmsg2+0x470
streams_putmsg+0x18
sosend+0x21f4
sendit+0x49c
sendto+0x5c
syscall+0x3c8
$syscallrtn+0x0
( CR:JAGae47736 SR:8606283791 )
Inconsistent result when binding to
deprecated address. The bind() fails in
the case of an address deprecated by
router advertisement and succeeds in
the case of a preferred lifetime
expiration.
( CR:JAGae48092 SR:8606284145 )
There is a short time delay before T_DISCONNECT event/s
are passed up to a TCP endpoint.
( CR:JAGae48167 SR:8606284220 )
Disconnect request to a rawip module,
does not restore the original status of
the end-point.
( CR:JAGae48168 SR:8606284221 )
Application may be notified that udp disconnect
operation succeeded even though the disconnect
operation actually failed.
( CR:JAGae48734 SR:8606284791 )
Potential system panic in mi_ibc_qenable().
( CR:JAGae48736 SR:8606284793 )
System may panic in udp_hash_remove_port()
if connect rebind error occurred.
( CR:JAGae48849 SR:8606284907 )
System panics with the following stack trace
panic+0x6c
report_trap_or_int_and_panic+0x94
trap+0xa78
nokgdb+0x8
rawip_opt_massage+0x58
rawip_wput+0x550
putnext+0xcc
streams_putmsg2+0x10c
streams_putmsg+0x18
sosend+0x80c
sendit+0x154
sendto+0x5c
syscall+0x62c
$syscallrtn+0x0
( CR:JAGae48995 SR:8606285053 )
Stale routing information is used even after
updating the route. This results in packets
being routed incorrectly.
( CR:JAGae49066 SR:8606285124 )
Packets with loopback address as source address
are sent outside the node. This behavior does not
conform with RFC 2373 section 2.5.3.
( CR:JAGae49207 SR:8606285262 )
ndp(1M) shows valid/preferred lifetimes beyond the
value advertised by the router.
( CR:JAGae49284 SR:8606285339 )
Excessive "Neighbor Solicitation" and "Neighbor
Advertisement" messages can be seen on the wire.
( CR:JAGae49372 SR:8606285427 )
connect() called on an already connected UDP DGRAM
socket leads to memory leak in ALLOCB_MBLK_SM arena.
( CR:JAGae50606 SR:8606286667 )
Executing the command
"/usr/bin/ndd -get /dev/sockets socket_udp_sndbuf_default"
returns the error:
"The object name 'socket_udp_sndbuf_default' is too long."
( CR:JAGae50872 SR:8606286932 )
IPv6 stack stops processing "router advertisement"
messages sent by the local routers.
( CR:JAGae50958 SR:8606287019 )
When there are multiple IPv6 Destination
Options only the first Destination Option
can be looked up using the sticky option.
( CR:JAGae50959 SR:8606287020 )
IPv6 Echo Reply could be sent with multicast
address as a source address of packet. This
does not conform to RFC 2463.
( CR:JAGae51088 SR:8606287149 )
recv() returned with errno set to ECONNRESET
due to inappropriate RST.
( CR:JAGae51169 SR:8606287230 )
There is a delay seen in SIOCMND ioctl
processing.
( CR:JAGae51708 SR:8606287775 )
The width of the first column (Name) of
the "netstat -in" output is limited to
8 characters. If the name of the interface
is longer than 8 characters, it gets truncated.
( CR:JAGae53982 SR:8606290053 )
When an AF_INET socket is bound to INADDR_ANY
and if an AF_INET6 socket tries to bind to the
same port with an IPv6 address, bind() does not
fail.
( CR:JAGae53985 SR:8606290056 )
IPv6 applications do not receive both Destination
Options when using ancillary data with recvmsg(2).
( CR:JAGae54172 SR:8606290246 )
The "ndp" command shows nothing in "State" and
"Flags" fields for point-to-point interface.
( CR:JAGae54691 SR:8606290848 )
System panic could happen with the following
stack trace:
panic+0x6c
assfail+0x3c
btlan_hw_req+0x2a0
btlan_resolved_output+0x18
hp_dlpi_wput+0xbc
putnext+0x12c
ip6_wput_ire+0x4f0
ip6_wput+0x33c
putnext+0x12c
tcp_wput+0x3c0
tcp_rput+0x3e40
puthere+0x12c
put_release+0x22c
ip6_rput+0x1a8c
putnext+0x12c
hp_dlpi_intr_put+0x80c
streams_put+0x124
hp_dlpi_ether_intr+0x31c
btlan_receive_frame+0x464
btlan_isr+0x100
sapic_interrupt+0x34
mp_ext_interrupt+0x408
ivti_patch_to_nop3+0x0
idle+0x11e8
( CR:JAGae54779 SR:8606290936 )
When socket caching is enabled, if a
listening socket is closed when there
are eager connections on the socket,
another process or thread cannot bind to
the same address for some time.
( CR:JAGae55682 SR:8606291918 )
Spurious Service Guard cluster reformation occurs
or Service Guard daemon aborts due to delay caused
by SIOCSWITCHLIF6ADDR ioctl.
( CR:JAGae55684 SR:8606291920 )
Bringing up an IPv6 interface can hang forever.
( CR:JAGae56041 SR:8606292278 )
The system can panic in the function
ire6_create().
( CR:JAGae56336 SR:8606292583 )
When an ICMP Address Mask Request packet is sent to
a unicast address, an ICMP Address Mask Reply is sent
even though the tunable
"ip_respond_to_address_mask_broadcast"
is turned off.
( CR:JAGae57090 SR:8606293340 )
System panics because of an error handling code
which attempts to ensure that no spinlocks are
being held while doing certain operations.
( CR:JAGae58674 SR:8606294977 )
read(2) on a socket sometimes, may return
ECONNREFUSED instead of ECONNRESET.
( CR:JAGae58883 SR:8606295188 )
System panics with the following stack trace
panic+0x14
report_trap_or_int_and_panic+0x4c
interrupt+0x1e8
$ihndlr_rtn+0x0
ipc_walk+0x7c
ill_down_ind+0x118
ip_rput_dlpi+0x9b0
ip_rput+0x394
putnext+0xc4
hp_dlpi_event+0x1b0
HP1_ospif_mib_event+0x150
ospif_watchdog+0x3c
HP1_oim_timeout+0x10
invoke_callouts_for_self+0xac
sw_service+0x8c
inttr_emulate_save_fpu+0x100
drv_priv+0x0
ip_open+0x11c
open_wrapper+0x44
csq_protect+0x10c
osr_open+0xaec
pse_open+0xe4
streams_open+0x48
soclone+0x178
soaccept+0xf4
sodequeue+0xac
accept+0x204
syscall+0x6e8
$syscallrtn+0x0
PHNE_28089:
This patch replaces PHNE_27989
( SR:8606283620 CR:JAGae47565 )
connect(2) to an already connected UDP socket
fails with EISCONN.
PHNE_27989:
This patch replaces PHNE_27730
( SR:8606250244 CR:JAGae16624 )
IPv6 sockets cannot join IPv4-mapped IPv6
multicast address groups.
( SR:8606254434 CR:JAGae18766 )
Upon encountering an error while parsing the
/etc/rc.config.d/nddconf file, ndd returns a failure
instead of printing out details of the error.
( SR:8606268603 CR:JAGae32841 )
Unable to ping a machine over a VLAN interface
if nettl tracing of ns_ls_ip is enabled on
that machine.
( SR:8606269955 CR:JAGae34194 )
System may panic with a data page fault when using
extension headers with RAWIP6 packets.
( SR:8606270176 CR:JAGae34414 )
System panics with a data page fault
( SR:8606270270 CR:JAGae34508 )
Destination port for the SYN+ACK packet is set to
zero when IPV6_DSTOPTS is set by setsockopt(2) for
the TCP/IPv6 Listen socket
( SR:8606270296 CR:JAGae34534 )
The ndd parameter ip6_nd_anycast_delay does
not work as documented in the ndd help text.
( SR:8606270780 CR:JAGae35016 )
IPv6 Neighbor Unreachability Detection messages
may synchronize with each other.
( SR:8606271001 CR:JAGae35237 )
In some cases, the IPv6 stack responds to invalid
Neighbor Solicitation messages.
( SR:8606271039 CR:JAGae35275 )
The system continues to forward packets even if
forwarding interfaces are disabled using
"ifconfig lanX inet6 private"
( SR:8606271235 CR:JAGae35466 )
The unspecified address, :: is set as source address
in packets sent to a solicited-node multicast address
even after a unicast address is assigned to the outgoing
interface. This behavior does not conform to RFC 2373.
( SR:8606271514 CR:JAGae35722 )
System panics with a data page fault with the following
stack trace
nokgdb+0x8
tcp_connect+0x2a0
tcp_wput_proto+0xbc
tcp_wput+0x418
putnext+0xcc
str_async_ioctl+0x540
hpstreams_ioctl_int+0xf8
streams_ioctl+0x34
soconnect+0x134
connect+0xdc
syscall+0x62c
$syscallrtn+0x0
( SR:8606272324 CR:JAGae36464 )
While waiting for IPv6 address resolution to complete,
the system runs very short of memory.
( SR:8606272765 CR:JAGae36872 )
If an application using UDP sockets first connects
to an IPv4-mapped IPv6 address, disconnects and tries
to connect to a native IPv6 address, the second connect
fails.
( SR:8606273477 CR:JAGae37565 )
System may panic with a data page fault with the following
stack trace if the primary interface is brought up with
0.0.0.0 address when a secondary interace is already up.
panic+0x6c
report_trap_or_int_and_panic+0x94
trap+0xed4
nokgdb+0x8
ipif_arp_up+0x2d8
ipif_up_arp_and_ires+0x598
ipif_up+0x264
ip_sioctl_copyin_done+0x1720
ip_wput_nondata+0x32c
ip_wput+0x158
csq_turnover_with_lock+0x1fc
putnext+0x1e4
udp_wput_other+0x130
udp_wput+0x720
putnext+0xcc
wait_iocack+0x6c
str_trans_ioctl+0x4f8
hpstreams_ioctl_int+0x174
streams_ioctl+0x34
ifioctl+0x88
soo_ioctl+0x7c
ioctl+0x120
syscall+0x750
$syscallrtn+0x0
( SR:8606273939 CR:JAGae38021 )
Query of the mib object ID_ipv6IfLowerLayer fails.
( SR:8606274153 CR:JAGae38231 )
Neighbor cache entry state changes not implemented
as defined in RFC 2461.
( SR:8606277348 CR:JAGae41419 )
After IPv6 address is switched from a primary
interface to backup interface, different NMIDs are
reported by IPv6 API if_nametoindex() and lanadmin
command for the same interface.
( SR:8606278217 CR:JAGae42276 )
Transport does not provide a direct method for
lower levels to request a gratuitous arp.
( SR:8606279349 CR:JAGae43402 )
After installing PHNE_27730, ping -n with a packet size
between 8 and 15 bytes will always result in the loss of
a packet.
( SR:8606281411 CR:JAGae45354 )
On a system that does not have IPv6 installed, if there
is an IPv6 entry in the /etc/rc.config.d/nddconf file, ndd
skips processing the next entry in the file.
( SR:8606281503 CR:JAGae45445 )
Arp mapping and publish entries may unexpectedly
disappear.
PHNE_27730:
This patch replaces PHNE_27382
( SR:8606241187 CR:JAGae08450 )
The system does not respond to any keystrokes
or commands and appears to hang. A Service Guard
system will TOC.
( SR:8606247981 CR:JAGae14381 )
Network Server Accelerator feature does not
work on HP-UX 11.11
( SR:8606264643 CR:JAGae28972 )
ifconfig fails with the message "invalid argument"
when used to bring up the primary interface even
when the interface name is specified correctly.
Normally this happens when auto-configuration is
used.
( SR:8606265738 CR:JAGae29995 )
System panics with the following stack trace
panic+0x6c
report_trap_or_int_and_panic+0x94
interrupt+0x208
$ihndlr_rtn+0x0
ip6_rput+0x44
puthere+0x148
mi_timeout_exec+0x2d8
invoke_callouts_for_self+0x9c
sw_service+0x100
mp_ext_interrupt+0x1f0
ivti_patch_to_nop3+0x0
idle_nonpset_loop+0x708
idle+0x4e0
swidle_exit+0x0
( SR:8606267968 CR:JAGae32207 )
sendfile(2) does not transfer files with a size
larger than 4GB in 32-bit 11.11 systems.
( SR:8606270271 CR:JAGae34509 )
Attempt to set the UDP socket options will fail
if a bind(2) has not been called previously.
( SR:8606270345 CR:JAGae34583 )
"ndp -an" does not display neighbour discovery cache
entries corresponding to some interfaces.
( SR:8606271213 CR:JAGae35448 )
The system panics due to bad memory reference.
( SR:8606273232 CR:JAGae37333 )
System panics with the following
stack trace when nettl tracing is enabled:
panic+0x6c
report_trap_or_int_and_panic+0x94
interrupt+0x208
$ihndlr_rtn+0x0
udp_icmp_error+0x48
udp_rput_other+0x38
udp_rput+0x1e0
putnext+0xcc
icmp_inbound_error+0xe54
icmp_inbound+0x1e8
ip_rput_local+0x1c0
ip_rput+0x184
putnext+0xcc
putnext2+0x48
hp_dlpi_mblk_fast_in+0x408
hp_dlpi_mblk_intr_put+0x91c
streams_put+0xdc
streams_put2+0x58
hp_dlpi_mblk_intr+0x190
lanc_ether_ics+0xfc
btlan_receive_frame+0x648
btlan_isr+0xf4
sapic_interrupt+0x2c
mp_ext_interrupt+0x2f0
ivti_patch_to_nop3+0x0
( SR:8606274682 CR:JAGae38759 )
System panics with the following stack trace
panic+0x6c
wait_for_lock+0x344
sl_retry+0x1c
tcp_bindi+0xf0
tcp_bind+0x1010
tcp_wput_ioctl+0x4c
tcp_wput+0x5d8
putnext+0xcc
wait_iocack+0x6c
str_istr_ioctl+0x738
hpstreams_ioctl_int+0x2c4
sobind+0xcc
bind+0x8c
syscall+0x62c
$syscallrtn+0x0
PHNE_27382:
This patch replaces PHNE_27063
( SR:8606170818 CR:JAGad40082 )
The ping command does not return different exit
values when an error occurs. As a result,
errors cannot be distinguished in scripts.
( SR:8606232612 CR:JAGae01847 )
If the primary route for an interface is deleted, then
a host on the same subnet is not reachable through the
secondary route, though the secondary route is also on
the same subnet.
( SR:8606232989 CR:JAGae02214 )
When using the ping command, a user cannot configure
the following:
* time interval between two successive packet
transmissions
* wait-time for exiting after the last packet
transmission
( SR:8606237266 CR:JAGae06311 )
Passing invalid arguments to connect() on an
already connected datagram socket does not
disconnect the socket as documented in the
manpage.
( SR:8606238197 CR:JAGae07224 )
rlogin and other applications that call recv()
with the MSG_OOB flag, lose data if the urgent byte
has not arrived when the call to recv() is made.
( SR:8606241192 CR:JAGae08455 )
Memory leak of timer related mblks in
ALLOCB_MBLK_SM arena.
( SR:8606243224 CR:JAGae10458 )
High (about 7%) spinning is observed on
a 32 way machine during performance testing.
( SR:8606244252 CR:JAGae10742 )
Transfer of one Mbyte of data over TCP takes
around two hours.
( SR:8606245677 CR:JAGae12137 )
"ndd -set" of arp parameters does not fail
when the parameter value is greater than
the maximum allowed.
( SR:8606246302 CR:JAGae12747 )
Memory leak occurs in IP outbound path
in case of failures while forwarding
packets.
( SR:8606246303 CR:JAGae12748 )
Memory leak occurs in case of an unrecognized
DL_OK_ACK response in ALLOCB_MBLK_SM arena.
( SR:8606246305 CR:JAGae12750 )
Memory leak occurs in the case of a failures
while setting up an ARP resolver for a new
interface in ALLOCB_MBLK_SM arena.
( SR:8606246307 CR:JAGae12752 )
Potential memory leak while initializing
an interface.
( SR:8606246311 CR:JAGae12756 )
System might panic in tcp_err_ack in case
of M_IOCACK.
( SR:8606246314 CR:JAGae12759 )
User specified PMTU via "route" command is not
used while creating fully resolved routes.
( SR:8606247242 CR:JAGae13682 )
Memory leak might occur in ALLOCB_MBLK_SM arena
if an error occurs when binding a UDP socket which
may make bind(2) appear to hang.
( SR:8606247710 CR:JAGae14113 )
Memory leak occurs in tcp context reinitialization
when attempting to reuse a socket.
( SR:8606248700 CR:JAGae15098 )
Initial RTO value calculated based on previous
connection delays and "tcp_rexmit_interval_initial"
ndd tunable is too large.
( SR:8606249286 CR:JAGae15678 )
Remote clients cannot login to a server.
"netstat -an" command displays many CLOSE_WAIT
TCP connections.
( SR:8606249333 CR:JAGae15725 )
On an HP-UX 11.11 release "ndd" help text for
"tcp_ignore_path_mtu" is present, but the tunable is
not supported.
( SR:8606251530 CR:JAGae17595 )
"ndd" does not display any help text for the tunable
"tcp_early_conn_ind".
( SR:8606253673 CR:JAGae18022 )
System hangs while running SRT with APA
configuration.
( SR:8606254542 CR:JAGae18874 )
RAWIP packets get fragmented even if the "do not
fragment" bit is set through IP_HDRINCL socket
option.
( SR:8606254713 CR:JAGae19044 )
System hangs for hours during reboot or normal
operation.
( SR:8606255551 CR:JAGae19868 )
System panics with the following stack trace:
stack trace for event 0
crash event was a panic
panic+0x6c
too_much_time+0x2b4
wait_for_lock+0x1e8
sl_retry+0x1c
ip_trash+0x24
ip_rtimer+0x40
ip_rput+0x318
puthere+0x148
mi_timeout_exec+0x19c
invoke_callouts_for_self+0x9c
sw_service+0x108
mp_ext_interrupt+0x14c
ivti_patch_to_nop3+0x0
idle_nonpset_loop+0x23c
idle+0x7ac
swidle_exit+0x0
( SR:8606257154 CR:JAGae21460 )
connect() hangs. System TOC with the following
stack trace:
_swtch+0xc4
_sleep+0x318
read_sleep+0x17c
streams_getmsg+0x3c8
soconnect+0x188
connect+0xdc
syscall+0x62c
syscallinit+0x554
( SR:8606257479 CR:JAGae21782 )
When trying to join invalid IPv6 multicast address,
kernel might panic.
( SR:8606259311 CR:JAGae23629 )
Posix recvmsg() returns incorrect "cmsg_len".
( SR:8606259405 CR:JAGae23723 )
Listen backlog becomes greater than 1062
regardless of what is specified in listen() call.
( SR:8606260488 CR:JAGae24811 )
An application might hang if "optlen" argument
passed to setsockopt() is too large.
( SR:8606260502 CR:JAGae24825 )
The system might panic with "alignment trap"
when IPv6 packets are fragmented over FDDI
on 64-bit machines.
( SR:8606261179 CR:JAGae25501 )
get_mib_info() returns incorrect values in the following
cases:
1. In the ipv6RouteTable, the ipv6RouteIfIndex is set to
the illegal value of 0 for a configured tunnel route.
2. In ipv6AddrTable, the ipv6AddrType of a manually
configured interface is UNKNOWN.
3. In ipv6AddrTable, the ipv6AddrStatus of a manually
configured interface is UNKNOWN.
4. In ipv6NetToMediaTable, the ipv6IfNetToMediaState is
UNKNOWN for configured local interfaces.
5. In ipv6NetToMediaTable, the ipv6IfNetToMediaLastUpdated
field shows negative numbers on a system that was just
booted.
6. In ipv6NetToMediaTable, the ipv6NetToMediaType of
entries created from router advertisements are labeled
as STATIC.
7. In ipv6NetToMediaTable, the ipv6NetToMediaType of
entries associated with the local (primary) interfaces
are labeled STATIC.
( SR:8606261753 CR:JAGae26072 )
ioctl(TCP_IOC_GET_UCRED) does not work for IPv4-mapped
IPv6 addresses.
( SR:8606262294 CR:JAGae26629 )
System panics with the following stack trace:
panic+0x6c
report_trap_or_int_and_panic+0x94
interrupt+0x208
$ihndlr_rtn+0x0
ip_csum_hdr+0x84
icmp_pkt_err_ok+0x24
icmp_unreachable+0x1c
ip_rput_local+0x231c
ip_rput+0x184
putnext+0xcc
hp_dlpi_mblk_fast_in+0xa0
hp_dlpi_mblk_intr_put+0xb8
streams_put+0xdc
streams_put_release+0x4c
hp_dlpi_mblk_intr+0x5c
lanc_ether_ics+0x11c
btlan_receive_frame+0x5e4
btlan_isr+0xfc
sapic_interrupt+0x2c
mp_ext_interrupt+0x2f0
ivti_patch_to_nop3+0x0
idle+0x1c4
swidle_exit+0x0
( SR:8606262982 CR:JAGae27312 )
If socket caching is enabled, fewer sockets
are cached resulting in performance degradation
under certain conditions.
( SR:8606263157 CR:JAGae27484 )
Applications will receive an error when they send
a T_OPTMGMT_REQ primitive with a value of T_NEGOTIATE
in the MGMT_flags field and OPT_length field of zero.
( SR:8606264041 CR:JAGae28365 )
"netstat -s" command displays incorrect value for
TCP duplicate ACKs count.
( SR:8606266690 CR:JAGae30935 )
Possible system crash when closing TCP connection.
( SR:8606269903 CR:JAGae34142 )
"ping" reports the round trip time incorrectly.
PHNE_27063:
This patch replaces PHNE_26762
( SR:8606261174 CR:JAGae25496 )
FTP is unable to put or mput files of any size due to EINVAL
errno 22 "sendfile: Invalid argument" when ftp attempts to
call senfile64() to transmit the files.
PHNE_26762:
This patch replaces PHNE_25644
( SR:8606181283 CR:JAGad50500 )
The IPv6 loopback address (::1) stops working when
the loopback device (lo0) is reconfigured to a
different address.
( SR:8606185191 CR:JAGad54393 )
System panics with the following stack:
unp_eventreg+0x4
so_eventreg+0x6c
evp_dereg_objhdr+0xa0
evp_close+0x15c
call_open_close+0x1f8
closed+0xb0
spec_close+0x54
vn_close+0x48
vno_close+0x20
closef+0x64
exit+0x324
rexit+0x28
syscall+0x200
$syscallrtn+0x0
( SR:8606189015 CR:JAGad58231 )
CPU usage is high when a fast producer program is
communicating with a slow consumer program using
UNIX domain sockets.
( SR:8606198555 CR:JAGad67744 )
Problems compiling recent versions
of the transport source code. These compilation
failures are not visible to customers.
( SR:8606198667 CR:JAGad67856 )
Closing a socket with socket caching turned on and
SO_LINGER set, may cause the kernel to loop indefinitely,
and the connection will be left in an IDLE state.
( SR:8606202871 CR:JAGad72045 )
The command "route inet6 delete net ::1 ::1"
deletes the IPv6 loopback route.
After the deletion, the command "ping ::1" fails.
( SR:8606203612 CR:JAGad72784 )
Code inspection uncovered a potential
system panic with the following stack trace.
soo_select
unp_poll_handler
so_poll_switch
.
.
( SR:8606208360 CR:JAGad77545 )
The route created by
"route inet6 add net 345::9/64 2002::2002 1"
can be removed using
"route inet6 delete net 345::/64 2002::2002 1"
but cannot be removed with
"route inet6 delete net 345::9/64 2002::2002 1"
( SR:8606209264 CR:JAGad78451 )
Threaded Async Socket IO can result in blocking close().
( SR:8606233159 CR:JAGae02383 )
The libnm function get_if_table() cannot
handle more than 256 interfaces.
( SR:8606233305 CR:JAGae02529 )
Application hangs in the accept() system call.
( SR:8606234349 CR:JAGae03553 )
Route deletion does not work in the following cases:
Case 1: if the destination is fully specified by the
128-bit IPv6 address
route inet6 add net 2222::1/64 2345::2 1
route inet6 delete net 2222::1/64 2345::2 1
Case 2: if the destination is an IPv4 compatible address
route inet6 add net ::1.2.3.4/64 2000::1 1
route inet6 delete net ::1.2.3.4/64 2000::1 1
Case 3: if the destination is a link-local address and
gateway is a local interface
route inet6 add net fe80::1/10 fe80::230:6eff:fe0c:6a91
route inet6 delete net fe80::1/10 fe80::230:6eff:fe0c:6a91
( SR:8606234706 CR:JAGae03897 )
When the command "ifconfig lan0:x inet6" is used to query
the status of a non-existent interface lan0:x, a new
interface lan0:x is created, with a null IPv6 address
"::". This information is displayed as if it existed
prior to the query and it only happens if lan0 has been
configured.
( SR:8606235627 CR:JAGae04771 )
get_mib_info(ID_ipv6Interfaces) always returns a zero
irrespective of the number of IPv6 interfaces configured
on the machine.
( SR:8606235632 CR:JAGae04776 )
Unable to delete IPv6 tunnel routes using set_mib_info().
( SR:8606236775 CR:JAGae05826 )
SIOCATMARK ioctl returns flag parameter set
after reading out-of-band data when the SO_OOBINLINE
socket option is set.
( SR:8606237606 CR:JAGae06646 )
Virtual IP address (lo0:1) cannot be used for IPv6.
( SR:8606237642 CR:JAGae06680 )
The address list hash table viewed using
'ndd' does not display addresses created by IPv6
stateless autoconfiguration.
( SR:8606238262 CR:JAGae07289 )
Data corruption occurs in the socket structure
during a connect().
( SR:8606238905 CR:JAGae07926 )
Flow control on UNIX domain datagram sockets
does not work.
( SR:8606241607 CR:JAGae08865 )
"traceroute" sometimes displays the error:
"traceroute: failed to get received hop limit or packet
info"
( SR:8606242226 CR:JAGae09479 )
'netstat -f inet6 -I secondary_interface interval'
displays no output and exits.
( SR:8606242679 CR:JAGae09914 )
A getsockopt() SO_ERROR on a listen() socket
causes a conflict with accept() processing.
( SR:8606244728 CR:JAGae11213 )
Users cannot set parameters for rawip6 through the
nddconf configuration file.
( SR:8606244958 CR:JAGae11439 )
An application gets ECONNRESET when doing a recv()
on a socket which has received RST from the remote
side. The RST arrives while there is pending data on
the socket which is discarded.
( SR:8606245325 CR:JAGae11797 )
There is a condition where a process termination will block.
( SR:8606245531 CR:JAGae12001 )
While building a 32-bit application with large files
support, "sys/socket.h" generates warnings due to
redeclaration of sendfile and sendpath prototypes.
( SR:8606245578 CR:JAGae12044 )
Potential panic in sounbind() due to referencing a
null pointer.
( SR:8606245903 CR:JAGae12361 )
SO_LINGER close() does not wait for a FIN's ACK
when the socket_caching_tcp option is on.
( SR:8606245905 CR:JAGae12363 )
Invalid pointer access in stream head read queue
when processing expedited data during recv().
( SR:8606245906 CR:JAGae12364 )
When a system is completely out of memory, there is a
possibility of a NULL pointer access when processing
an accept() system call.
( SR:8606245919 CR:JAGae12377 )
No noticeable symptom to the end user.
( SR:8606245952 CR:JAGae12410 )
Simultaneous calls to accept made on multiple threads
may hang and never return.
( SR:8606245971 CR:JAGae12429 )
ndd fails to set some tunables in the
/etc/rc.config.d/nddconf file.
( SR:8606246169 CR:JAGae12617 )
socket() call returns EINVAL when udp_recv_hiwater_max
is less than 65535.
( SR:8606246301 CR:JAGae12746 )
1. Memory leak occurs when there are tunnel
errors for a tunnel that has broadcast
forwarding set.
2. System panic in ip_wput_frag() occurs when
memory is freed twice during outbound IP fragmentation
and when there are tunnel errors.
( SR:8606246309 CR:JAGae12754 )
System leaks memory in 512-byte blocks.
( SR:8606246321 CR:JAGae12766 )
System leaks memory.
( SR:8606246398 CR:JAGae12843 )
When socket caching is turned ON, an unnecessary FIN
is sent before the required RST when closing a connection
which has the SO_LINGER option turned on with an l_linger
value of 0.
( SR:8606246561 CR:JAGae13006 )
ACKs are sent in response to old segments which
have wrapped sequence numbers.
( SR:8606246676 CR:JAGae13119 )
System may leak memory when the ndd parameter
socket_caching_tcp is turned off.
( SR:8606247339 CR:JAGae13779 )
Applications receive unexpected ICMPv6 error messages.
( SR:8606247484 CR:JAGae13921 )
When using the SO_LINGER option on a UNIX95 socket,
specifying certain large values of the linger timeout
may result in incorrect behavior, or performance
degradation.
( SR:8606247490 CR:JAGae13927 )
Socket state inconsistency when closing a
socket in a multithreaded application.
( SR:8606247508 CR:JAGae13945 )
Socket Caching can cause a kernel memory problem.
( SR:8606247520 CR:JAGae13957 )
No visible symptoms.
( SR:8606248122 CR:JAGae14522 )
The "ifconfig lanx:y" command which is supposed to
display the current configuration of lanx:y, will create
a NULL entry for lanx:y if that interface does not exist.
( SR:8606248406 CR:JAGae14806 )
A blocking condition with a socket call can occur with
multi-threaded AF_UNIX applications.
( SR:8606248669 CR:JAGae15067 )
In a multi-processor environment, the "named" daemon
is not killed when it receives the SIGTERM signal.
Socket recvmsg() is blocking despite O_NONBLOCK.
( SR:8606249185 CR:JAGae15579 )
Minor performance loss in UDP recv system calls.
( SR:8606250322 CR:JAGae16697 )
AF_UNIX domain datagram type sockets applications
fail on recv() call.
( SR:8606250432 CR:JAGae16807 )
When socket caching is turned on, a close() performed
on a socket with SO_LINGER returns before the FIN-ACK
has been received.
( SR:8606250573 CR:JAGae16948 )
"ndd -get /dev/tcp tcp_status" might fail to reset
the semaphore left by another "ndd" due to a premature
exit.
( SR:8606253203 CR:JAGae18008 )
System is showing excessive memory pressure.
( SR:8606254024 CR:JAGae18356 )
Configuring IPv6 addresses on a fddi device does not work.
( SR:8606254214 CR:JAGae18546 )
"msg_controllen" can contain an invalid number
when returning from recvmsg().
( SR:8606255696 CR:JAGae20013 )
A system panic with the following stack trace may occur
while opening an AF_UNIX socket.
socket_merged+0xb4
socket+0x14
syscall+0x204
( SR:8606256314 CR:JAGae20629 )
recv(2) returns EWOULDBLOCK after poll(2) detects POLLIN.
( SR:8606256741 CR:JAGae21056 )
The socket debug macros are called with an
insufficient number of parameters which causes
compilation warnings.
( SR:8606257716 CR:JAGae22019 )
On systems where the directories /usr/contrib/man or
/usr/contrib/man/man1.Z do not preexist, the installation of
PHNE_25644 will display warnings indicating that
these directories do not have the correct permissions.
PHNE_25644:
( SR:8606113716 CR:JAGac07230 )
"ndd -c" fails when /etc/rc.config.d/nddconf contains
more than 10 tunable entries.
( SR:8606129381 CR:JAGac86927 )
"ndd -c", does not support arp tunables in
/etc/rc.config.d/nddconf.
( SR:8606138758 CR:JAGad08005 )
If socket caching is turned on, a connect made on
a socket which has SO_REUSEADDR set returns EADDRINUSE.
( SR:8606146239 CR:JAGad15575 )
Intermittent hangs in close() when
the so_linger option is on.
( SR:8606206785 CR:JAGad75958 )
ifconfig fails and dumps core when numerous
interfaces (such as vlan) are created on the
system.
( SR:8606216047 CR:JAGad85218 )
"ndd -c" fails to handle very long tunable names
referenced in /etc/rc.config.d/nddconf.
( SR:8606217313 CR:JAGad86465 )
System hangs, requiring a TOC
( SR:8606220082 CR:JAGad89223 )
netstat fails and prints the following error
message when there are numerous interfaces
(such as vlan) configured on the system.
"ppa_area (125000) too small (expected 169260)
can't get ppa list"
( SR:8606220677 CR:JAGad89814 )
When packets get lost in a network, it takes a
long time (60 seconds or longer) to retransmit.
( SR:8606221602 CR:JAGad90736 )
read() sometimes loses data and returns 0.
( SR:8606221683 CR:JAGad90817 )
"ndd -c" fails in some abnormal cases.
( SR:8606223770 CR:JAGad92866 )
"ndd -get" does not display anything for socket
tunable parameters.
( SR:8606224560 CR:JAGad93648 )
FTP hangs when transferring files from PC DOS to HP-UX.
( SR:8606224838 CR:JAGad93926 )
If nddconf contains any IPv6 entries,
"ndd -c" returns the following error:
"invalid tunable name, ip6_nd_transmit_interval for this
module, for a list of valid names, use name '?'".
This error can be obtained by adding the following
entries to /etc/rc.config.d/nddconf.
TRANSPORT_NAME[0]=ip6
NDD_NAME[0]="ip6_nd_transmit_interval"
NDD_VALUE[0]=4000
( SR:8606225148 CR:JAGad94236 )
"ndd -c" does not work if the NDD_VALUE for
a tunable in /etc/rc.config.d/nddconf is a string
containing non-numeric characters. For example:
NDD_VALUE[0]="lan0 ON"
is not recognized.
( SR:8606225238 CR:JAGad94326 )
A system panic can occur when "ndd -get /dev/udp udp_status"
is run.
( SR:8606225324 CR:JAGad94412 )
Connections sometimes hang after a ServiceGuard
switchover.
( SR:8606225422 CR:JAGad94509 )
If a user tries to assign a primary IPv6
address which is already assigned to another
interface on the machine, the kernel returns
success to the user even if the operation failed.
( SR:8606227936 CR:JAGad96993 )
The return value of "ifconfig lanX inet" command when
the interface lanX is not configured with an IPv4
address is "0" instead of the expected value "1".
This happens when lanX has been configured with
only an IPv6 address.
( SR:8606228250 CR:JAGad97307 )
System panics when under heavy memory pressure.
( SR:8606229080 CR:JAGad98134 )
1. tcp_fin_wait_2_timeout does not work when the
ndd tunable 'socket_caching_tcp' is set to 1.
2. bind() fails with EINVAL unexpectedly when
'socket_caching_tcp' is set to 1.
( SR:8606229279 CR:JAGad98332 )
System panic:
tcp_icmp_error+0x38
tcp_rput_other+0x518
tcp_rput+0x58
csq_turnover_with_lock+0x84
str_spu_sw_isr+0x654
sw_service+0xb0
mp_ext_interrupt+0x150
ivti_patch_to_nop3+0x0
idle+0x104
( SR:8606230067 CR:JAGad99118 )
When there are a large number of TCP
connections, "ndd -get /dev/tcp tcp_status" does not
list all the connections.
( SR:8606230303 CR:JAGad99355 )
The VLAN IP Precedence feature is not supported.
( SR:8606231247 CR:JAGae00485 )
System panics with the following stack trace:
panic+0x14
report_trap_or_int_and_panic+0x80
trap+0xdb8
nokgdb+0x8
tcp_get_ucred+0x1d4
tcp_wput_ioctl+0x160
tcp_wput+0x918
putnext+0xcc
wait_iocack+0x68
str_istr_ioctl+0x72c
hpstreams_ioctl_int+0x370
hpstreams_ioctl+0x50
spec_ioctl+0xac
vno_ioctl+0x90
ioctl+0x78
syscall+0x6f8
$syscallrtn+0x0
( SR:8606231253 CR:JAGae00491 )
When using "ndd -set" to set sockets tunables, the
new value is echoed onto the screen.
( SR:8606231526 CR:JAGae00764 )
System panics with the following stacks:
tcp_wput+0x58
csq_turnover_with_lock+0x84
str_spu_sw_isr+0x654
or
tcp_wput+0x58
putnext+0xcc
sth_wsrv+0x204
sq_wrapper+0x94
str_sched_up_daemon+0x1c4
str_sched_daemon+0x1a4
main+0x854
( SR:8606231670 CR:JAGae00908 )
ndd does not display more than 16KB of output.
( SR:8606231951 CR:JAGae01187 )
System panics with the following
stack trace:
tcp_rsrv_comm+0x18
tcp_rput+0x3620
csq_turnover_with_lock+0x7c
str_spu_sw_isr+0x5f8
sw_service+0x8c
( SR:8606232014 CR:JAGae01250 )
get_if_tbl fails with ENODEV as of PHNE_25642.
( SR:8606232185 CR:JAGae01421 )
System panics with following stack trace:
tcp_rsrv_comm+0x18
tcp_rsrv+0x10
sq_wrapper+0x90
str_sched_mp_daemon+0x130
str_sched_daemon+0x2dc
main+0xa9c
$vstart+0x34
$locore+0x90
( SR:8606233090 CR:JAGae02314 )
In some cases urgent data gets retransmitted
unnecessarily impairing performance.
( SR:8606233338 CR:JAGae02562 )
The route command adds incorrect routes in some cases.
For example, "route inet6 add 2000: 4321::210:83ff:fef7:1 1"
adds the route
2000:0:0:4321::/64 4321::210:83ff:fef7:1 UG 0 lan1 0
instead of the expected route
2000::/64 4321::210:83ff:fef7:1 UG 0 lan1 0
( SR:8606233710 CR:JAGae02935 )
"netstat -p tcp" reports a high number of
"bad TCP segments dropped due to state change"
and consequently a high number of "retransmit
timeouts."
( SR:8606208110 CR:JAGad77288 )
Problem occurs when multiple instances of
ndd -get /dev/tcp tcp_status is invoked
while there are huge numbers of tcp connections
on the system.
( SR:8606210971 CR:JAGad80159 )
Using the UNIX95 version of the recvmsg() system call,
an application compiled as 64-bit may not receive
ancillary data properly.
( SR:8606213513 CR:JAGad82705 )
Systems relying on random increments for
choosing less predictable TCP ISN values,
are still vulnerable to statistical attacks.
( SR:8606213579 CR:JAGad82771 )
"netstat" does not resolve those network
addresses in /etc/networks, which do not
have all four octets specified explicitly.
For example, the following entry will not be
resolved:
127 loopback
It gets resolved only when specified as follows:
127.0.0.0 loopback
( SR:8606223042 CR:JAGad92146 )
/usr/include/netinet/tcpip.h is missing in HP-UX 11.11.
( SR:8606226976 CR:JAGad96038 )
Multicast testing leads to system hang.
( SR:8606227536 CR:JAGad96591 )
The help text for the ndd variable, "tcp_isn_passphrase"
is missing.
( SR:8606227920 CR:JAGad96977 )
"netstat -i" displays incorrect values (always 0) for
Ierrs, Oerrs, Coll in the case of snap interfaces.
( SR:8606229650 CR:JAGad98702 )
System panics with the following stack trace:
igmp_timeout_handler+0x160
ip_rtimer+0x100
ip_rput+0x408
puthere+0x148
mi_timeout_exec+0x288
sw_service+0xb0
mp_ext_interrupt+0x150
ihandler+0x904
idle+0xe24
swidle+0x20
( SR:8606232100 CR:JAGae01336 )
When ndd is invoked from the command line
with an invalid tunable, it fails without
setting the exit value to one.
( SR:8606233164 CR:JAGae02387 )
connect() call for the SOCK_STREAMS socket returns
the ETIMEDOUT error. This occurs when another
connection already bound to the same address
(with 4-tuple being the same) has received RST from
the remote side, and has not been closed by the
application.
( SR:8606237497 CR:JAGae06537 )
System panic:
panic+0x6c
wait_for_lock+0x344
slu_retry+0x1c
mp_socket_lock+0x2c
soclose+0x1d4
soo_close+0x90
closef+0x64
close+0x98
syscall+0x204
$syscallrtn+0x0
( SR:8606238295 CR:JAGae07322 )
If ndd -set /dev/tcp tcp_debug is set to 1
or higher and strace is used you will get
the following messages to your screen:
-------------
11348 13:12:25 17614232 1 ... 0 0 [HATCP][-9195]
tcp:0x61dd28a8 tcp_ha_state:0x40 tcp_state: 0 tcp
HA state changed to HA_TCP_CONN_ACTIVE
-------------
This happens even though HATCP is not enabled
on the system.
( SR:8606230164 CR:JAGad99215 )
Performance degradation after installing PHNE_23456.
( SR:8606244219 CR:JAGae10710 )
System Memory Leak out of the 512 byte
memory arena.
PHNE_25642:
This patch replaces PHNE_25134
( SR:8606229037 CR:JAGad98091 )
ndd tunable ip_enable_udp_bcastrecv has no
effect.
PHNE_25134:
This patch replaces PHNE_24506
( SR:8606227191 CR:JAGad96253 )
Panic during
ipsec_admin -stop
ipc_hash_remove+0x1c0
ip_close+0x1e4
close_wrapper+0x38
csq_protect+0x11c
osr_pop_subr+0x22c
osr_close_subr+0x2fc
hpstreams_close_int+0x2f8
streams_close+0x14
soclose+0x42c
soo_close+0x90
closef+0x64
close+0x58
syscall+0x204
PHNE_24506:
This patch replaces PHNE_24211
( SR:1653262485 CR:JAGaa27163 )
If /etc/resolv.conf points to an invalid
DNS server, netstat seems to hang when
resolving network or host addresses to names.
( SR:8606137536 CR:JAGad06654 )
A trace from tcpdump showed that sendfile sends trailer
buffers separately.
( SR:8606162623 CR:JAGad31939 )
Service Guard system TOC with the following trace:
x_mi_mpprintf_putc+0x1c
x_mi_iprintf+0x280
x_mi_mpprintf+0x64
tcp_status_report+0x188
x_nd_getset+0x148
tcp_wput_ioctl+0x9c
tcp_wput+0x82c
putnext+0xcc
wait_iocack+0x68 s
tr_istr_ioctl+0x738
hpstreams_ioctl_int+0x6a8
hpstreams_ioctl+0x50
spec_ioctl+0xac
vno_ioctl+0x90
ioctl+0x7c
syscall+0x480
$syscallrtn+0x0
( SR:8606165144 CR:JAGad34440 )
The TTL field in UDP Broadcast packets defaults to
1. This is different from the default of 64 used in
10.20.
( SR:8606165595 CR:JAGad34886 )
The xopen_networking(7) man page should say that if you
specify -lc at compile time, -lxnet must precede -lc
( SR:8606168597 CR:JAGad37875 )
System becomes non-responsive to networking traffic.
( SR:8606172829 CR:JAGad42089 )
A multi-threaded socket application hangs
when a thread is closing a socket.
( SR:8606176180 CR:JAGad45420 )
UDP socket bound to local IP address cannot receive
broadcast packets.
( SR:8606179394 CR:JAGad48618 )
ndd has no help text for the tunable
"ip_enable_udp_bcastrecv".
( SR:8606179508 CR:JAGad48732 )
Occasionally the Service Guard command
cmrunnode core dumps.
( SR:8606195876 CR:JAGad65080 )
The man pages lead users to think that they can only get
ENOPROTOOPT for setsockopt or getsockopt calls.
( SR:8606199995 CR:JAGad69181 )
The recv(2) man page gives an incorrect description of
the meaning of a return value of 0. It specifically fails
to indicate that this is the means for indicating a normal
connection termination.
( SR:8606201810 CR:JAGad70984 )
11.11 lacks event port functionality.
( SR:8606204354 CR:JAGad73536 )
Outbound UDP datagrams get silently dropped.
( SR:8606204948 CR:JAGad74124 )
Truncating a file while transferring it
with FTP causes a system hang.
( SR:8606208365 CR:JAGad77551 )
nettl tracing with the filtering option
does not work.
( SR:8606208617 CR:JAGad77804 )
Connections on a system with an active server
which also has the "tcp_conn_strategy" parameter
set to non-zero, may see a low overall throughput.
( SR:8606213997 CR:JAGad83188 )
The following error occurs when compiling "socket.h"
with aCC (ANSI C++ Compiler).
Error 43: "/usr/include/sys/socket.h", line 470
# C++ does not allow Old-style (non-prototype)
function definitions.
inline sbsize_t sendfile(a,b,c,d,e,f) int a,b,f;
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
( SR:8606215049 CR:JAGad84240 )
If users assign an ip to a vlan and bring down
the vlan by 'ifconfig vlan down', (not unplumb
it), system could then successfully delete this
vlan. But netstat still have this ip on the
list, and when users create a vlan next, this
vlan already has this ip assigned.
( SR:8606215148 CR:JAGad84339 )
System panics on sbflush panic 2:
trace event 0
stack trace for event 0
crash event was a panic
panic+0x14
sbflush+0x68
sbrelease+0x14
sorflush+0xa4
sofree+0x15c
soclose+0x23c
soo_close+0xc8
closef+0x64
close+0x90
syscall+0x6f8
$syscallrtn+0x0
( SR:8606216950 CR:JAGad86105 )
Creation of raw socket using socket() fails when the
protocol specified is either IPPROTO_UDP or IPPROTO_TCP.
( SR:8606217657 CR:JAGad86809 )
ifconfig lan10000 fails, but ifconfig lan9999 succeeds.
( SR:8606217835 CR:JAGad86985 )
Bogus TCP retransmissions during specweb99/netperf
TCP_CRR tests.
( SR:8606218708 CR:JAGad87856 )
The congestion window increases too slowly.
( SR:8606218753 CR:JAGad87901 )
System runs out of memory when under heavy
inbound TCP traffic.
( SR:8606219937 CR:JAGad89079 )
On UP boxes when 2 (or more) aio_reads are pending
on the same socket and ioctl(SIOCAIOABORT) is used,
a close on that socket will result in an unkillable
hung process.
( SR:8606220535 CR:JAGad89673 )
Measureware and Glance do not show any networking
statistics.
( SR:8606221777 CR:JAGad90911 )
When setting ip_pmtu_strategy to 0 any non-local
networks have a maximum MTU of 576.
( SR:8606221830 CR:JAGad90955 )
Enhancement Request for HATCP intended for
the exclusive use of the OpenCall product.
( SR:8606221942 CR:JAGad91060 )
The same secondary interface may be unexpectedly
configured with multiple IPv6 addresses.
( SR:8606222035 CR:JAGad91150 )
Packets sent to the IPv6 tunnel interface do not
reach their destinations.
( SR:8606222508 CR:JAGad91621 )
When system memory use is very high
the accept system call returns ENOBUFS.
( SR:8606223127 CR:JAGad92230 )
Memory Leak in the 2k bucket when SAM hostagent is
running.
( SR:8606224453 CR:JAGad93541 )
Mobile IP reply packets do not go the same tunnel
as the request packets.
( SR:8606224457 CR:JAGad93545 )
System runs out of Anonymous ports when doing
binds for outbound connections.
( SR:8606224462 CR:JAGad93550 )
Message getting displayed on console:
tcp_timer: strange state (-6) [5767,d425] TCP_CLOSED
( SR:8606224472 CR:JAGad93560 )
Slow connection speeds over Long Fat Pipe or
Long Narrow Pipe connections.
( SR:8606224475 CR:JAGad93563 )
Possible loss of IPv6 options when using UDP.
PHNE_24211:
This patch replaces PHNE_23645
( SR:8606139436 CR:JAGad08735 )
system panic on X.25 socket.
A possible deadlock situation
The panic stack is as follows:
panic+0x10
spin_deadlock_failure+0x38
deadlock_check+0x9c
sl_pre_check+0x54
spinlock+0x14
mp_socket_lock+0x34
mp_socket_lock2+0x38
XLS_F_handler+0x6c4
XSO_F_handler+0x958
XLS_F0_a_connect_ind+0x188
XLS_F_handler+0x6a4
XST_F_read_put+0x398
putnext+0x1f4
CI_touser+0x268
Rx_CALL+0x384
L2_datind+0x4ac
dlpi_rxll+0xb4
x25lrsrv+0x60
sq_wrapper+0xc8
str_sched_mp_daemon+0x33c
str_sched_daemon+0x29c
im_mpnetstr+0x28
DoCalllist+0x38
main+0x24
$vstart+0x34
$locore+0x90
( SR:8606154650 CR:JAGad23967 )
Some clients can cause TCP connections to be
set up which will cause excessive memory
consumption by data buffers.
( SR:8606161095 CR:JAGad30413 )
NDD does not support the reporting or changing
of kernel socket parameters.
( SR:8606164046 CR:JAGad33347 )
Some customers using the Raptor Firewall product found
the daemon processes (tcp_gspd) spinning and eating
CPU time (70-80% of a cpu on UP systems).
( SR:8606166814 CR:JAGad36101 )
Remsh hangs.
( SR:8606167654 CR:JAGad36937 )
Sometimes nettl is unable to capture reply packets.
( SR:8606171754 CR:JAGad41018 )
Many NDD parameters are not recognized by
the help text system in the current version of NDD.
( SR:8606178097 CR:JAGad47324 )
The system has a Data Page Fault panic with
either of the following two traces:
Trace 1:
panic+0x14
report_trap_or_int_and_panic+0x4c
interrupt+0x1e8
$ihndlr_rtn+0x0
puthere+0x44
mi_timeout_exec+0x224
sw_service+0x8c
mp_ext_interrupt+0x108
ivti_patch_to_nop3+0x0
idle+0x1e8
swidle_exit+0x0
Trace 2:
panic+0x14
report_trap_or_int_and_panic+0x4c
interrupt+0x1e8
$ihndlr_rtn+0x0
puthere+0x44
tcp_rput_context_check+0x4a4
tcp_rput+0x784
putnext+0x16c
ip_rput_local+0x524
ip_rput+0x648
putnext+0x16c
hp_dlpi_mbuf_fast_in+0x70
hp_dlpi_intr_put+0x154
streams_put+0x134
hp_dlpi_intr+0x84
lan2_process_packet+0xcc
lan2_int_fr_rnr+0x19c
lan2_isr+0x184
eisa_int+0x134
lasi_interrupt+0x5c
mp_ext_interrupt+0x300
ivti_patch_to_nop3+0x0
idle+0x1b8
swidle_exit+0x0
( SR:8606179470 CR:JAGad48694 )
Executing "ifconfig lanX:Y unplumb" may core dump
ifconfig.
( SR:8606181093 CR:JAGad50313 )
ifconfig does not generate a message when an
"unplumb" operation is unsuccessful.
( SR:8606193754 CR:JAGad62965 )
Performance of ftp over hyper-fabric can be very slow
when "put"-ing a file.
( SR:8606196811 CR:JAGad66009 )
After installing the transport 11.11 Patch, PHNE_23645,
the following three debug lines appear in the /etc/rc.log
file when the system is booted.
name= tcp_conn_strategy, value= 0
do_sockets_nddset() &stri= 0x7f7f0538
0
( SR:8606196818 CR:JAGad66016 )
The "ifindex" number for the loopback driver, lo0, is not
being set to a non-zero number. This does not conform
to RFC 1213. This could cause trouble for programs such
as OpenView or programs using libnm.sl by causing them
to retrieve a zero "ifindex" value.
( SR:8606202009 CR:JAGad71183 )
TCPdump shows multiple threads connecting to
the same server port
( SR:8606205122 CR:JAGad74298 )
HP-UX Transport does not support IPv6.
( SR:8606174765 CR:JAGad44011 )
"ifconfig <ifname> inet" does not work for non-privileged
users.
( SR:8606202365 CR:JAGad71539 )
Panic in sofromcache with tcp_conn_strategy=1.
panic+0x6c
report_trap_or_int_and_panic+0x94
trap+0xed4
nokgdb+0x8
asm_spinlock_usav+0x14
sofromcache+0x48
soclone+0x148
soaccept+0xdc
sodequeue+0xb8
accept+0x148
syscall+0x204
$syscallrtn+0x0
( SR:8606204817 CR:JAGad73995 )
System panics with the following when
tcp_conn_strategy=1;
panic+0xc4
report_trap_or_int_and_panic+0x108
trap+0x1258
nokgdb+0x8
socachecallback+0x628
osr_sth_so_callback+0x84
str_simple_ioctl+0x4924
hpstreams_ioctl_int+0x77c
streams_ioctl+0x98
sotocache+0x128
sounlock+0x834
mp_socket_unlock+0x1c
soclose+0x13c0
soo_close+0x138
closef+0x2b0
close+0x1d8
syscall+0xe70
$syscallrtn+0x0
( SR:8606151945 CR:JAGad21284 )
The system experiences extensive memory consumption when
processing large numbers of routes.
( SR:8606198614 CR:JAGad67803 )
System hangs and free memory is zero.
System TOC with the following trace:
putnext+0x188
ip_wput_ire+0x398
ip_wput+0x470
putnext+0xcc
tcp_rput+0x1088
csq_turnover_with_lock+0x84
str_spu_sw_isr+0x200
sw_service+0xb0
mp_ext_interrupt+0x150
ivti_patch_to_nop3+0x0
ki_accum_push_TOS+0xa0
syscall+0x808
$syscallrtn+0x0
( SR:8606203900 CR:JAGad73078 )
TCP connection hangs.
( SR:8606203039 CR:JAGad72213 )
A UDP broadcast is sent once for each IP address configured
on the interface.
( SR:8606180610 CR:JAGad49831 )
System panics with a Unresolved priv 0 interruption.
TPI_discon_ind
sosend
soo_rw
rwuio
write
syscall
( SR:8606187509 CR:JAGad56716 )
When a program uses the ICMP TIMESTAMP request message
to query an HP-UX system for the current time, the reply
generated provides less precise information than was
available from a 10.20 system.
( SR:8606140093 CR:JAGad09415 )
TCP connections where both ends close at the same time may
experience an unnecessary delay of 1.5 seconds or more.
This problem has an especially high probability of being
seen on loopback connections where one end of the
connection has the SO_LINGER option turned on.
( SR:8606209311 CR:JAGaa26796 )
A man page for traceroute does not exist.
( SR:8606164864 CR:JAGad34161 )
A MSG_OOB recv() on a socket which has been set to
non-blocking and for which there is no OOB data blocks
rather than returning an error.
( SR:8606136810 CR:JAGad05930 )
The netstat(1) man page says -rs works, but it does not.
( SR:8606207797 CR:JAGad76974 )
Performance degradation in IPSec traffic.
( SR:8606140128 CR:JAGad09450 )
"netstat -i" does not display inbound errors ("Ierrs"),
outbound errors ("Oerrs"), and collisions ("Coll").
( SR:8606211447 CR:JAGad80635 )
swremove of PHNE_23645 fails due to unresolved
externals. This can only happen on a system which has
been upgraded from 11.00.
( SR:8606168903 CR:JAGad38181 )
The thread safety of libc APIs is not consistently
mentioned in the man pages.
( SR:8606160792 CR:JAGad30111 )
A server program listening for incoming connection requests
with select() or accept() can receive notification of a
connection's existence before it is fully established.
( SR:8606144985 CR:JAGad14323 )
System panics with the following stack trace.
tcp_err_ack
tcp_info_req
tcp_wput_proto
tcp_wput
putnext
str_async_ioctl
hpstreams_ioctl_int
streams_ioctl
soprotostate
soclose
This panic can only occur when a system is under
extreme memory pressure.
( SR:8606205538 CR:JAGad74714 )
System panics with the following stack trace;
panic
report_trap_or_int_and_panic
interrupt
ihandler
x_mi_timer
tcp_accept
tcp_wput_proto
tcp_wput
csq_turnover_with_lock
puthere
tcp_icmp_error
tcp_rput_other
tcp_rput
putnext
icmp_inbound_error
icmp_inbound
ip_rput_local
ip_rput
putnext
hp_dlpi_mblk_fast_in
hp_dlpi_mblk_intr_put
streams_put
streams_put_release
hp_dlpi_mblk_intr
lanc_ether_ics
btlan_receive_frame
btlan_isr
sapic_interrupt
mp_ext_interrupt
ihandler
( SR:8606207520 CR:JAGad76696 )
libnm gives wrong number of interfaces when queried for
number of interfaces.
( SR:8606209138 CR:JAGad78325 )
With ATM on system and IP traffic, system panics
with the following stack trace;
panic
report_trap_or_int_and_panic
interrupt
ihandler
bcopy_pcxu_method
LDI_F_set_fastpath
LEC_F0_uwioctl
LEC_F0_uwput
csq_turnover_with_lock
putnext
CAM_F0_sp_dataind
streams_put
CAM_F_data_ind
alx_receive_it
alx_isr
wsio_interrupt
mp_ext_interrupt
ivti_patch_to_nop3
sounlock
mp_socket_unlock
soo_select2
soo_select
select
syscall
( SR:8606206366 CR:JAGad75539 )
System panics with the following stack trace:
panic
report_trap_or_int_and_panic
trap
nokgdb
ipif_up
ip_sioctl_copyin_done
ip_wput_nondata
ip_wput
csq_turnover_with_lock
putnext
udp_wput_other
udp_wput
putnext
wait_iocack
str_trans_ioctl
hpstreams_ioctl_int
streams_ioctl
ifioctl
soo_ioctl
ioctl
syscall
PHNE_23645:
This patch replaces PHNE_23288
( SR:8606202746 CR:JAGad71920 )
A syntax error in S340net start script at line 223.
( SR:8606199716 CR:JAGad68902 )
The following message is logged in /etc/rc.log:
/sbin/rc2.d/S340net[4]: i <= interface_count : The
specified number is not valid for this comm
( SR:8606184133 CR:JAGad53342 )
Network client systems which request system
time from an HP-UX 11.11 system using the
"ICMP timestamp request" message do not get a reply.
( SR:4701419036 CR:JAGaa93907 )
If one urgent byte is on a socket, and a recv() is done
without MSG_OOB on that socket, the urgent byte will
disappear. Subsequent recvmsg() calls with the MSG_OOB
flag on will return EINVAL, instead of returning the
urgent byte.
( SR:8606193751 CR:JAGab13526 )
A sending program can send forever even to a connection
that has been closed. This is only seen if
tcp_conn_strategy is set to NON-NULL.
( SR:1653307850 CR:JAGab20953 )
Some commands hang when the interface
associated with the hostname as defined in /etc/hosts
goes down. Pinging the hostname will fail.
( SR:8606193755 CR:JAGab72514 )
When using shutdown() and stack-caching the mss value
could be set to one (1) byte.
( SR:8606113564 CR:JAGab76776 )
When using AF_UNIX sockets blocking sendmsg()
returns EMSGSIZE when receive side is
close to being full. This can happen on blocking or
non-blocking.
( SR:8606134574 CR:JAGad03709 )
Inbound packet traffic is seen mostly on one processor.
Need enhancement to make inbound packet scheduling better.
better.
( SR:8606144006 CR:JAGad13339 )
Non-privileged users cannot open a UDP socket
with an ephemeral port.
( SR:8606147348 CR:JAGad16691 )
Data is received with garbage bytes appended to it. The
length of the received data will have a size which is a
multiple of STRMSGSZ (where STRMSGSZ is the value of the
kernel parameter STRMSGSZ on the system sending the data).
For this to occur, the kernel parameter STRMSGSZ must have
been changed to a value smaller than the default on the
system which sends the data.
( SR:8606155790 CR:JAGad25103 )
X-terminals are not able to boot from a server
operating in a Service Guard environment.
( SR:8606165729 CR:JAGad35020 )
Customer is seeing RESET's during connects if the
requested MSS is smaller than the tcp_mss_min
set on the system.
( SR:8606165984 CR:JAGad35271 )
Panic during FIN processing on a system
experiencing memory pressure.
0x000e4450 (set_protid_registers+0x1bfc)
0x004e6ff4 (tcp_hash_insert_port+0x2c)
0x004ee800 (tcp_reinit+0x4d0)
0x004e31c8 (tcp_clean_death+0x280)
0x004f00e0 (tcp_rput+0x1838)
0x0057ea68 (csq_turnover_with_lock+0xb0)
0x0056ede0 (putnext+0x2b0)
0x0059f75c (streams_write_uio+0x204)
0x003832b0 (sosend+0x1230)
0x003783a4 (sendit+0x41c)
0x00377eac (send+0x74)
0x0031b304 (syscall+0x754)
( SR:8606167240 CR:JAGad36526 )
System panic's with;
4) putnext+0x18
5) ip_wput_local+0x528
6) ip_wput_ire+0x43c
7) ip_wput+0x458
8) csq_turnover_with_lock+0x88
9) osr_pop_subr+0x98c
10) osr_close_subr+0xb00
11) hpstreams_close_int+0xd40
12) hpstreams_close+0x2c
13) call_open_close+0x1f8
14) closed+0xb0
15) spec_close+0x54
16) vn_close+0x48
17) vno_close+0x20
18) closef+0x68
19) close+0x48
20) syscall+0x200
When using XTI programs over loopback.
( SR:8606181319 CR:JAGad50536 )
An ENOTCONN is intermittently returned when
a read is done on a non-blocking socket.
( SR:8606182096 CR:JAGad51312 )
Applications such as web servers
which use the 32-bit "sendfile" system call
to send files with the O_LARGEFILE
flag set fail with an EOVERFLOW
error.
( SR:8606184456 CR:JAGad53664 )
On a system with a primary and secondary IP address.
If after a user has joined a multicast group the secondary
address is removed the user no longer receives multicast
packets.
( SR:8606185603 CR:JAGad54805 )
Unable to compile IPv6 applications.
( SR:8606186550 CR:JAGad55754 )
Slow MIB performance. Can cause the system
to look like it has hung when doing netstat or ndd
on a system with a lot of connections.
( SR:8606189357 CR:JAGad58572 )
System panics when Apache WEB server is started.
( SR:8606193852 CR:JAGad63062 )
Doing a ping to a broadcast with just a few packets in
rapid succession will panic the system.
System will also crash while using Mobile IP.
ip_rput_local+0x22b4
ip_rput+0x90c
putnext+0xcc
hp_dlpi_mblk_fast_in+0xa0
hp_dlpi_mblk_intr_put+0xb8
streams_put+0xdc
streams_put_release+0x4c
hp_dlpi_mblk_intr+0x5c
( SR:8606195961 CR:JAGad65165 )
System panics at boot time with the following stack;
panic: assertion failed (holding spinlock(s) on entry
to copyout) at line 469 in /ux/core/kern/pa/vm/vm_mapcall.c
PC-Offset Stack Trace (read down, top of stack is 1st):
0x0006cde8 (panic+0xa0)
0x0006d234 (assfail+0x3c)
0x0006d3ec (_assfail+0x2c)
0x0010452c (copyout+0xfc)
0x00489614 (sockndd_error+0x24)
0x00489d0c (soo_ioctl+0x6e4)
0x001d9004 (ioctl+0x13c)
0x003d200c (syscall+0x754)
0x00033a1c (syscallinit+0x554)
End Of Stack
( SR:8606197685 CR:JAGad66877 )
T_ALLOPT option in XTI/TLI will fail with TBADOPT.
Also the tcp initial congestion window is too large
possibly causing performance problems at connect time.
Defect:
TCP_CWND_INITIAL was defined in tcp.h as 0x16.
It should have been TCP_CWND_INIT.
PHNE_23288:
( SR:8606174670 CR:JAGad43916 )
Compiling kernel-intrusive programs (such as drivers and
programs that access /dev/mem or /dev/kmem) on HP-UX 11.11
may result in compiler warnings or errors due to namespace
violations. Executing such programs may cause kernel
structure data corruption, resulting in memory leaks, hangs,
or panics.
Defect Description:
PHNE_34672:
( SR:8606420692 CR:JAGaf80521 )
When the numerical value of ndd tunable
`ip_ire_flush_interval' is set to less than the numerical
value of ndd tunable `ip_ire_clean up_interval', network
routing table entries for link local addresses are being
re-used within the kernel. These should be deleted.
Resolution:
Code has been modified to ensure that network routing
entries that should be deleted are deleted when required.
( SR:8606416944 CR:JAGaf76777 )
A function that handled the ioctl(2) commands,
SIOCSWITCHIF6ADDR, SIOCSWITCHLIF6ADDR, SIOCSLIFADDR,
SIOCSLIFNETMASK, SIOCSLIFFLAGS, SIOCGLIFFLAGS,
SIOCSLIFADDR returned EINVAL incorrectly after returning
ENOMEM.
Resolution:
Code has been modified to return appropriate
error numbers.
( SR:8606421454 CR:JAGaf81284 )
When the numerical value of ndd tunable
`ip_ire_redirect_interval' is set to zero, code that
periodically cleans up network routing table entries is
not activated.
Resolution:
Code has been modified to ensure that even when the
numerical value of ndd tunable `ip_ire_redirect_interval'
is set to zero, network routing table entries are
periodically cleaned up.
( SR:8606426280 CR:JAGaf85761 )
The libnm library function get_mib_info() does not use
information from all CPU's in the system to update the
value of MIB object `udpOutDatagrams'.
Resolution:
The libnm function get_mib_info() has been modified to use
information from all CPU's in the system to update the
value of MIB object `udpOutDatagrams'.
( SR:8606415076 CR:JAGaf74930 )
After the first threshold timer expires, TCP sends a
negative advise to IP to remove the route for the
destination even when it receives an ICMP Destination
Unreachable Fragmentation Needed message as part of
PMTU Discovery.
Resolution:
Appropriate code changes have been made to
properly interpret the ICMP Destination
Unreachable Fragmentation Needed message
after the sending TCP crosses the first
threshold timer.
( SR:8606424562 CR:JAGaf84085 )
1. Help texts for ndd tunables `arp_defend_interval' and
`arp_redefend_interval' erroneously show the applicable
units as seconds. It should be milliseconds.
2. Setting the ndd tunable `arp_redefend_interval' to zero
does NOT prevent a published entry from being
redefended as, erroneously, indicated by the relevant
ndd help text.
Both these are typographical errors.
Resolution:
Relevant ndd help text has been modified. Specifically,
1. Units for `arp_defend_interval' and
`arp_redefend_interval' has been corrected to
`milliseconds.
2. Description of `arp_redefend_interval' has been
updated.
( SR:8606429097 CR:JAGaf88563 )
This panic occurs due to a race in using and updating a
global counter.
Resolution:
Code has been modified to avoid the race between using and
updating the relevant global counter.
( SR:8606438456 CR:JAGaf96535 )
Deletion of loopback routes when `route -f' is executed,
is an unintended oversight in the relevant code.
Resolution:
Code has been modified such that, route -f does not
delete the loopback route.
( SR:8606438648 CR:JAGaf96709 )
In order to use network bandwidth efficiently,
HP-UX delays last ACK of the TCP 3-way handshake for 10ms,
to see if any data can be sent in the same packet.
If there is no data within 10ms, only the ACK is sent.
If there is any data, a packet with both the ACK
and PUSH bits set, is sent even before 10ms period.
Some server programs and network devices do not accept
packets with both the ACK and PUSH bits set; they
reject such connections.This is not a defect in HP-UX.
However, this HP-UX feature may make it difficult for
some HP-UX clients to communicate with some servers.
Consequently, it must be possible for customers to
select this feature if it is required.
Resolution:
ndd tunable "tcp_delay_final_twh_ack" has been implemented
to disable the 10ms delay of the final ACK of TCP's
3-way connection handshake.
( SR:8606428509 CR:JAGaf87990 )
When nettl is used to trace a large number of TCP
connections, some invalid memory accesses lead to a system
panic.
Resolution:
Code has been modified to avoid the panic
( SR:8606431587 CR:JAGaf91029 )
A non-blocking connect(2) returns EINPROGRESS, even
though connection is forcefully rejected by the peer.
Resolution:
Code has been modified to return the correct error number.
( SR:8606441063 CR:JAGaf98824 )
When an igelan card is used with very low
PMTU values and checksum offload enabled,
some TCP segments may be retransmitted.
Resolution:
Code has been modified such that the TCP
segments are properly formatted so that
they are not retransmitted.
PHNE_34135:
( SR:8606406155 CR:JAGaf66063 )
The CKO information sent to IP by the driver
is malformed. Hence IP needs to detect this
and drop such packets.
Resolution:
Code is changed in IP such that IP detects
malformed packets and drops it.
( SR:8606409272 CR:JAGaf69170 )
For AF_UNIX domain and SOCK_STREAM
type sockets when a thread is blocked
on recv(2) and the peer performs orderly
shutdown using shutdown(2), thread blocked on
recv(2) will not be woken up. According to
man page of recv(2) in above scenario thread blocked
on recv(2) call should return with zero
as return value.
Resolution:
Code is modified to wake up threads
blocked on recv(2) call when the
connected peer does orderly shutdown
using shutdown(2) system call, so that
recv(2) call returns with zero as return value.
( SR:8606413150 CR:JAGaf73012 )
The code that initializes the fields in TCP and IP instance
data structures, does not initialize with correct
values when socket caching is enabled.
Resolution:
Code is modified to initialize the fields of TCP
and IP instance data structures to correct values.
( SR:8606416735 CR:JAGaf76568 )
Network interface cards that support TCP checksum
offload do not support IP fragments. In this case,
the sequence number in TCP segments is corrupted.
Resolution:
Modification has been made to ensure that proper
sequence number is stored at the correct field in
the TCP segment. A known limitation <see JAGaf98824>
wherein TCP segments may be re-transmitted when using
igelan or iether cards at very low PMTU with checksum
offloading enabled will be fixed in a future patch.
( SR:8606424136 CR:JAGaf83659 )
The check to see if the sequence number of a
connection indication is non-zero needs to be
modified.
Resolution:
The check to see if the sequence number is
non-zero is removed.
( SR:8606416402 CR:JAGaf76253 )
traceroute(1) man page wrongly refers to HP-UX
as non-standard OS with regard to ICMPv4
message generation when IPv4 packets TTL
field expires.
Resolution:
traceroute(1) manpage is modified. Wrong reference
to HP-UX as non-standard OS is removed.
( SR:8606404400 CR:JAGaf64324 )
The setsockopt(2) man page indicates that
close on a socket with SO_LINGER
enabled may return ETIMEDOUT, but
the code does not return the error.
Resolution:
The man page is fixed as per code.
( SR:8606424222 CR:JAGaf83745 )
Strings in ndd help text catalog file
are not ordered properly.
Resolution:
Code has been changed to order
the strings correctly.
( SR:8606428254 CR:JAGaf87735 )
Whenever a blocking connect thread gets interrupted
by a signal (with SA_RESTART flag set), then this thread is
leaving critical section without resetting the same
flag.This results in hang when connect() is
restarted.
Resolution:
Appropriate code changes have been made to prevent the hang.
( SR:8606432060 CR:JAGaf91502 )
sendfile(2) code was ignoring error cases
under some circumstances if there is a non-zero
header data that is requested to be sent.
Resolution:
sendfile(2) code has been modified such that
it fails with the appropriate errno even when there
is header data that is to be sent.
( SR:8606431462 CR:JAGaf90904 )
Due to a race condition when multiple
threads are running, the select() thread loops
continuously resulting in system hang.
Resolution:
The code changes are done such that select()
does not loop when such a race occurs.
PHNE_33628:
( SR:8606322228 CR:JAGae84696 )
TCP, IP and ARP modules do not do Synch Queue Flow Control
in some cases. Because of this, the modules can spend too
much time in processing the Synch Queue elements.
Resolution:
Appropriate code changes have been made in TCP, IP and
ARP modules to do Synch Queue Flow Control.
( SR:8606352228 CR:JAGaf13033 )
While creating a route for a destination, if the route
entry of the gateway is not present, it is created
first. Later the route for the destination is
created. While creating the route for the gateway the
PMTU of the route is inherited, which is an incorrect
behaviour.
Resolution:
Code changes have been made so that the gateway route
does not inherit the wrong PMTU.
( SR:8606381553 CR:JAGaf41766 )
The thread calling vfork(2) hangs since other threads
sleeping on semaphores are immune to signals.
Resolution:
The locking mechanism is changed to use socket state
flags instead of semaphores.
( SR:8606385089 CR:JAGaf45243 )
When shutdown(2) for write side is issued on a listen socket
by a thread, socket code sends a message to TCP code, for it
to shutdown thereby stop accepting any more new connections.
It then changes the socket state to indicate that socket
has been shutdown. But it does not wakeup any threads
sleeping in accept(2) on the same listen socket.
Resolution:
socket code has been modified to wakeup all threads sleeping
in accept(2) when a shutdown(2) for write side is issued on
the same socket. This will ensure that any new/pending
accept(2)'s will succeed if there are pending unaccepted
connections or else fail with appropriate error return.
( SR:8606387073 CR:JAGaf47226 )
A socket created by calling accept(2) on listen socket will
inherit its properties from the listen socket. One such
property currently being acquired says whether the socket is
writeable. If the listen socket is already shutdown for
write side, the new accepted will also be marked "shutdown
for write side". This results in a SIGPIPE being generated
when the application tries to send data on this new socket.
Resolution:
Do not mark the socket as being shutdown on the write side
before returning from accept(2).
( SR:8606387074 CR:JAGaf47227 )
When shutdown(2) for read side(SHUT_RD) is performed on a
listening socket, shutdown code sets the socket state to
reflect this. But it does not inform TCP to stop accepting
any new connections for this listen socket.
Resolution:
shutdown(2) code modified to instruct TCP to stop accepting
new connections for this listen socket.
( SR:8606387075 CR:JAGaf47228 )
shutdown(2) code instructs TCP code to stop accepting new
connections by sending a message. But when there are
connections pending to be accepted TCP code returns error
back to socket code and keeps accepting new connections.
socket code does not check the error message and instead
assumes that shutdown at TCP level succeeded though its not
the case.
Resolution:
shutdown(2) code has been modified to send a different
message to TCP. TCP code on receipt of this message will
always shut down and will not accept any more new
connections.
( SR:8606387098 CR:JAGaf47251 )
When shutdown(2) is performed on listen socket as a result
of shutdown(2) processing, a message is put on STH RQ by
TCP. accept(2) code processes this message in such a way
that only first two threads are woken up leaving the
rest sleeping forever in accept(2).
Resolution:
Modified the shutdown(2) and accept(2) code to wakeup all
the threads sleeping in accept(2) and not just one.
( SR:8606387472 CR:JAGaf47625 )
shutdown(2) code when invoked for read side does not abort
established connections that are pending to be accepted by
the application
Resolution:
shutdown(2) code has been modified to abort connections
pending to be accepted, when invoked for read side(SHUT_RD)
on listen socket
( SR:8606388101 CR:JAGaf48253 )
Code just marks the socket as shutdown for read operations
and returns without taking any further action.
Resolution:
shutdown(2) code has been modified to wake up all threads
sleeping in accept(2) on the same socket.
( SR:8606388102 CR:JAGaf48254 )
shutdown(2) in a particular case treats the listen socket as
normal connected socket and returns EINVAL when it should
not.
Resolution:
shutdown(2) code has been modified not to return EINVAL in
this case.
( SR:8606390716 CR:JAGaf50862 )
When multiple listeners are listening on same port and IP
address, there is a condition where some tcp functions might
get called recursively and result in kernel stack overflow.
Resolution:
Code has been modified to identify the next listener which
can accommodate the new connection and put the new
connection on that TCP instance.
( SR:8606390906 CR:JAGaf51051 )
ndd help text for "socket_caching_tcp" claims that setting
a value between 1 and 512 will set it to 512. But the
implementation is such that any value above 1 will set the
specified value.
Resolution:
ndd help text has been updated to match the implementation.
( SR:8606392043 CR:JAGaf52175 )
shutdown(2) on a listening socket does not wake up threads
sleeping in select(2).
Resolution:
code has been modified to wake up threads sleeping in
select(2) when a shutdown(2) is done on the listening
socket.
( SR:8606392816 CR:JAGaf52902 )
There is a mismatch in the hash order of listeners at
TCP and IP module. Due to this connection backlog is
not maintained properly.
Resolution:
Code has been modified to maintain the correct connection
backlog when there are more than one listeners for a port.
( SR:8606394962 CR:JAGaf54976 )
ping(1M) in some cases does not wait for sufficient amount
of time to process the reply received for the last sent
packet.
Resolution:
Code has been modified for ping(1M).
( SR:8606395825 CR:JAGaf55831 )
The help text for the ndd tunable ip_pmtu_strategy
displays incorrect maximum allowable value.The
description for some of the values are missing.
Resolution:
ndd(1M) has been modified to display the
help text for all allowable values of the
tunable 'ip_pmtu_strategy'
( SR:8606396734 CR:JAGaf56717 )
When an ARP Cache Entry expires, the concerned route
to the neighbor is deleted and a fresh probe is made
whenever a packet has to be sent out to the neighbor.
This causes a delay in sending the packets out.
Resolution:
Appropriate code changes have been maded to use the
stale ARP Cache Entry during address revalidation by ARP.
( SR:8606401226 CR:JAGaf61175 )
Due to race condition between close(2) and sendmsg(2)
system calls, when sending ancillary data, system can
panic with DPF.
Resolution:
Code has been modified in sendmsg(2) path to avoid system
panic.
( SR:8606404988 CR:JAGaf64908 )
sendfile(2) doesn't check for mss being negative. Also,
mss is declared as size_t which is unsigned long.
Resolution:
The check has been added to verify whether mss is negative
which avoids the hang.
( SR:8606405138 CR:JAGaf65058 )
getsockname(2) manpage has not documented EINTR as a
valid error number in the manpage.
Resolution:
Manpages for getsockname(2), getpeername(2), getsockopt(2)
and setsockopt(2) have been updated to document EINTR
in the ERRORS section.
( SR:8606411217 CR:JAGaf71099 )
When using XOPEN sockets setsockopt(2) can leak memory
under certain circumstance.
Resolution:
Code has been added to prevent the memory leak.
( SR:8606412656 CR:JAGaf72518 )
This problem occurs when two threads race to create
IRE_ROUTE to the same destination at the same time.
This results in same IP ID being used for both the
connections.
Resolution:
Code has been modified not to create two IRE_ROUTE IRE's
to the same destination, thus avoiding using same IP ID
( SR:8606416090 CR:JAGaf75942 )
In path of accept(2) code, in some conditions a null
pointer is de-referenced, which causes the system
to panic.
Resolution:
Code has been modified to make sure that null pointer
is not de-referenced.
PHNE_33159:
( SR:8606374624 CR:JAGaf34938 )
TOPS can be automatically deactivated for applications
that migrate between CPUs when using a socket. In some
situations, this can lead to lower performance. For
example, heavy network traffic through long-lived TCP
connections.
Resolution:
A new TOPS operation mode (Mode 2) has been added to
improve scalability where the benefits outweigh an
additional performance penalty while closing a socket.
This new TOPS mode can be activated via the new ndd
tunable "socket_enable_tops".
( SR:8606389338 CR:JAGaf49486 )
It is possible for ICMP code to loop too long.
Resolution:
The ICMP code has been modified to avoid the looping
condition.
( SR:8606397158 CR:JAGaf57140 )
Due to a race condition that is occurring while switching
address to a new interface card, the route entry of the
local interface gets altered such that any locally
generated packet for the interface is sent out on the link.
Resolution:
Code has been modified to avoid the race condition and to
maintain the integrity of the route entry for local
interface.
( SR:8606399900 CR:JAGaf59856 )
The ICMP message handling needs to be made more robust.
Resolution:
The ICMP message handling mechanism has been improved.
( SR:8606402124 CR:JAGaf62059 )
TOPS feature can be enabled in two modes. Provide
a 'ndd' command option to enable and disable TOPS
feature with sufficient help text giving the details.
Resolution:
A new ndd tunable called 'socket_enable_tops' has been
provided for enabling TOPS feature and Suitable help text
also has been added.
( SR:8606403047 CR:JAGaf62982 )
System panics due to a problem in processing
router advertisement messages.
Resolution:
Code has been modified to avoid the panic.
PHNE_32042:
( SR:8606248840 CR:JAGae15237 )
The system panics in sounlock() due to a race condition in
handling AF_UNIX socket locks.
Resolution:
The solockpair() function has been modified for AF_UNIX
sockets so that a race condition does not occur between
sockets in a multi-threaded environment.
( SR:8606340920 CR:JAGaf01830 )
An IGMP query is not answered if multiple IP addresses are
configured on the receiving host and the sending router is
on a different subnet.
Resolution:
Changes are done such that the host now replies
with IGMP REPORTS even when there is more than
one IP address configured on the interface and
when both host and router are on different subnets.
( SR:8606346523 CR:JAGaf07347 )
The problem occurs because the Linux kernel sets the
IP Ident value to 0 in the IP header for UDP packets,
except when IP fragments are sent. The current IP
code uses IP Ident field to distribute the packets
among nfs streams bound to different CPUs.
Resolution:
Code has been modified to distribute the packets
properly.
( SR:8606348091 CR:JAGaf08913 )
The help text for the ndd tunable
"tcp_rexmit_interval_initial" displays the default value
as 500 ms instead of displaying 3000 ms.
Resolution:
The help text is corrected to display the correct
default value.
( SR:8606350213 CR:JAGaf11035 )
The timer-related data structures are corrupted
due to a race between timer functions. The
processing of these corrupted timer-related data
structures is resulting in a panic.
Resolution:
Code has been modified to avoid the race and hence the
corruption of the timer data structures.
( SR:8606351765 CR:JAGaf12570 )
Sometimes, the TCP module accepts bogus TCP segments
with the URG flag set.
Resolution:
Code has been modified so that TCP module disregards
the URG flag in bogus TCP segments.
( SR:8606355258 CR:JAGaf16012 )
The tcp instance and the ip instance structures are
always reset when sockets are reused from socket cache,
irrespective of the privilege of the process opening
the socket.
Resolution:
The privileges in the tcp instance and ip instance
structures are now set according to the privileges
of the process opening the socket, while reusing
sockets from a socket cache during socket creation.
( SR:8606355392 CR:JAGaf16106 )
During simultaneous connect(2), the connection confirmation
message is not sent to the socket. If some other data is
sent by the peer, select(2) goes into an infinite loop
waiting for the connection confirmation message to arrive.
Resolution:
Appropriate code changes have been made so that a
connection confirmation message is sent to the socket
during simultaneous connect(2).
( SR:8606356802 CR:JAGaf17502 )
When neither '-i' nor '-s' option is specified,
traceroute(1) uses the first interface address returned
in the IP header of the outgoing probe packet without
referring to the routing table. As a result, the response
to this probe packet may not reach the host in case of
a multihomed host.
Resolution:
Code has been modified not to specify the source
IP address in the traceroute command. Instead, the
IP address is allowed to select the appropriate
source address when more than one interface is
configured on a system.
( SR:8606358089 CR:JAGaf18788 )
In certain cases when there is no data sent, sendfile(2)
might return immediately without processing a disconnect
indication that has arrived.
Resolution:
sendfile(2) checks for messages in the stream head before
returning even if there are zero bytes to be sent.
( SR:8606361264 CR:JAGaf21954 )
When socketpair(2) is called on AF_UNIX domain,
memory allocated to socket structure may not be
freed in certain error conditions. This leads
to memory leak.
Resolution:
The memory leak is avoided by freeing memory in
all error condition cases.
( SR:8606363737 CR:JAGaf24395 )
The IP code does not put the proper IP
address in the source field even though it
picks the right interface to send the packet.
Resolution:
The IP code has been modified to conform to RFC 1122.
Therefore, if the source address specified by an
application is a multicast address, it is replaced by
the address of the interface on which the packet is
sent out.
( SR:8606366614 CR:JAGaf27178 )
sendfile(2) ignores the error condition when it is
invoked with a non-zero header length.
Resolution:
sendfile(2) code has been modified such that it returns
EPIPE if the connection associated with the socket is
either shut down using shutdown(2) on the write side or
the peer has reset the connection.
( SR:8606369254 CR:JAGaf29798 )
When the X.25 and ACC products are installed, the system
panics because of a race between the wakeup performed on
receiving data and select(2) returning due to a timeout.
Resolution:
The race between wakeup on data arrival and
select(2) returning due to timeout is now avoided.
( SR:8606369649 CR:JAGaf30193 )
When the /etc/nsswitch.conf file specifies a database other
than the /etc/services file for service name lookup, the
execution of the "netstat-a" command may be slow.
Resolution:
The performance of the "netstat -a" command is improved
by avoiding the unnecessary lookup of a service name
for port number 0 (zero).
( SR:8606370089 CR:JAGaf30518 )
The ndd help text displays the default value of
tcp_rexmit_interval_initial_lnp and tcp_conn_request_max
as 1500 and 20 respectively, even though the correct
values are 3000 and 4096.
Resolution:
The ndd help text for the tunable parameters
tcp_rexmit_interval_initial_lnp, tcp_conn_request_max are
modified to display the correct default values.
( SR:8606370407 CR:JAGaf30829 )
When a network program sends large data using a high
speed interface for a long time, the network program may
hang.
Resolution:
The TCP congestion window mechanism has been adjusted
to correct this.
( SR:8606370902 CR:JAGaf31320 )
The interface dependency between STREAMS and TRANSPORT
must be changed.
Resolution:
A new interface has been provided to XPORT by streams
to decouple the dependency.
( SR:8606372704 CR:JAGaf33107 )
The select(2) call was not properly handling the case
where SO_OOBINLINE was set for the socket and there is
OOB data in-band.
Resolution:
Code changes are made such that select(2) is now
made to detect the presence of OOB data under
the above mentioned conditions.
PHNE_31247:
( SR:8606366227 CR:JAGaf26833 )
In some situations, tcp_ip_hdr_len() uses incorrect
memory addresses.
Resolution:
tcp_ip_hdr_len() has been modified appropriately.
( SR:8606371947 CR:JAGaf32365 )
When a socket is closed, sometimes a mblk is not
freed before the stream is closed.
Resolution:
The socket code has been changed so that the mblk
allocated is freed before the stream is closed
( SR:8606373330 CR:JAGaf33722 )
The memory allocated for shutdown processing
is not freed when an error condition is detected.
Resolution:
The allocated memory for shutdown processing
is now freed when an error is detected.
( SR:8606374761 CR:JAGaf35075 )
A memory leak can occur due to an incorrect
manipulation of the linked list that contains
information about socket caching.
Resolution:
The linked list is now correctly manipulated;
thereby, avoiding the memory leak.
PHNE_31092:
( SR:8606366388 CR:JAGaf26992 )
When socket caching is ON, a race could happen
that will cause accept() to block non-interruptibly,
leading to this problem.
Resolution:
The race that leads to the problem is now avoided.
PHNE_30656:
( SR:8606349322 CR:JAGaf10143 )
The tail of the mbuf chain may be left with a non-null
value which may cause a panic when a message is sent.
This problem happens with AF_UNIX sockets.
Resolution:
The tail is reset to the correct value when freed.
( SR:8606344720 CR:JAGaf05570 )
Application hangs in accept(2), waiting for a
T_DISCON_IND message from TCP module,
which never arrives.
Resolution:
The accept(2) code has been modified such that it does
not wait for the T_DISCON_IND message.
( SR:8606351614 CR:JAGaf12419 )
When ServiceGuard switches interfaces in the event of
an interface failing, the first bind might result in
an error.
Resolution:
The flags are set appropriately while switching the
interface which prevents the error while binding.
( SR:8606353629 CR:JAGaf14423 )
When socket caching is enabled and an IPv6 socket
is opened, an IPv4 socket could be returned from
the cache resulting in the failure of the ioctl.
Resolution:
While closing an IPv6 socket, it is never cached
and while opening an IPv6 socket, a new IPv6 stream
is always created.
( SR:8606354852 CR:JAGaf15608 )
A socket is in closing path and is using a cache entry
and close has not yet completed. Simultaneously, if socket
caching is turned off, the cache entries are cleaned
from the cache and the entry which the socket
belongs to might also be freed before the close completes
resulting in panic.
Resolution:
A check has been added while cleaning the cache to see if
any socket is using that cache entry. If that entry is
found to be in use, then that entry is skipped.
( SR:8606355807 CR:JAGaf16521 )
If an application issues a t_unbind(3) when
there is unsent data in the TCP connection,
the response to T_UNBIND_REQ is sent only
after all the data and the FIN is sent out.
Resolution:
When a T_UNBIND_REQ is received through the
t_unbind(3) function and there is unsent data
in the TCP connection, the code moves the connection
to the default queue and processes the T_UNBIND_REQ
immediately.
PHNE_30053:
( SR:8606293457 CR:JAGae57207 )
"netstat -a" displays send and receive queue sizes
as zero even though there is data still to be read
or yet to be sent. This happens because netstat
queries TCP using MIB and calculates the sizes
based on the reply which usually will be zero.
Resolution:
The netstat(1) man page has been changed to reflect
this behaviour.
( SR:8606301096 CR:JAGae64566 )
When socket caching is enabled, not all the fields
are cleaned up before returning the socket to the cache.
As a result, if a socket is flow controlled on the send
side at the time of closing, it will remain so in the
cache. A subsequent socket(2) call will return a
flow controlled socket thus making send(2) block forever.
Resolution:
TCP module's write queue is flushed properly before
returning the socket to the cache. This ensures that
a socket returned to the cache will not be flow-controlled.
( SR:8606302874 CR:JAGae66231 )
IP header is included twice in the reassembled IP datagram
under the following extremely rare case:
A fragment with offset 0 arrives without any data and
another fragment with offset 0 belonging to the same
datagram arrives with some data.
Resolution:
IP reassembly code has been changed so that only
a single IP header is retained in the reassembled
datagram under all cases.
( SR:8606302876 CR:JAGae66233 )
During code review, problems were identified that
are related to IPv6 fragmentation and reassembly.
Resolution:
Code changes have been made to fix the issues found
during code review.
( SR:8606302878 CR:JAGae66235 )
When a user application tries to bind a socket to a
non-local IPv4 compatible IPv6 address, it is not
returned an error.
Resolution:
A check was added to the code to verify if the source
address is the same as the address to be bound.
( SR:8606304572 CR:JAGae67915 )
While switching from one interface to another using
the SIOCSIFADDR ioctl, a LOOPBACK IRE is created which
is never removed when the `from' interface is unplumbed.
Any subsequent operation which accesses this IRE will
now cause a DPF and crash the system.
Resolution:
Code has been modified to ensure that a LOOPBACK IRE
is not unnecessarily created.
( SR:8606305933 CR:JAGae68981 )
During deferred T_UNBIND_REQ processing, TCP can send an
RST and reinitialize TCP for reusing the stream. While
performing this operation, TCP ignores the T_UNBIND_REQ
that it had already received and sets the state of the
TCP to be "bound", if the stream was previously "bound".
When this stream is reused, call to bind() will fail with
EINVAL.
Resolution:
Appropriate code changes have been made to properly unbind
the TCP endpoint after T_UNBIND_REQ is received.
( SR:8606306041 CR:JAGae69089 )
If the DLPI module sends a T_ERROR_ACK to the DL_BIND_REQ
primitive, pending ARP requests are not woken up.
Resolution:
Upon receiving T_ERROR_ACK from DLPI, wake up pending
ARP requests.
( SR:8606306287 CR:JAGae69335 )
When the ndd tunable arp_probe_count is enabled and the
system detects a conflict while resolving an IP address,
access to a freed memory may lead to panic.
Resolution:
Validation of the memory is done before accessing the
memory.
( SR:8606306288 CR:JAGae69336 )
Transport protocols assumed the presence of a complete
header in a single message block. This was not guaranteed
by IP in some cases while handling loopbacked packets.
Resolution:
The code has been changed such that IP ensures the presence
of a complete header in a single message block.
( SR:8606306299 CR:JAGae69347 )
While checking for IPv4 compatible IPv6 addresses,
an unspecified address or a loopback address in IPv6
may pass the check since the code checks only for zeroes
in the first 12 bytes.
Resolution:
The code has been changed to check for unspecified and
loopback addresses in IPv6.
( SR:8606306857 CR:JAGae69891 )
While deleting an IRE by linking it to deferred list,
no check is made to see if the IRE already exists in
deferred list. In the absence of such a check, the
same IRE may get added to the list more than once.
Resolution:
Code has been added to check if an IRE already exists
in the deferred list. If it is present the same IRE
is not linked to the deferred list again. By doing this,
double free data page fault type panic is avoided.
( SR:8606307734 CR:JAGae70769 )
The Fast Path feature for an interface may be disabled
if the probe query returns with EBUSY errno.
Resolution:
The code has been modified to correctly handle EBUSY errno.
( SR:8606308464 CR:JAGae71498 )
IPv6 does not check whether the index specified in the
IPV6_PKTINFO option and the interface index of the
outgoing interface for a particular destination are same.
In such cases, the packet can be sent out on the wrong
interface.
Resolution:
Code changes have been made so that the interface index
field in the IPV6_PKTINFO option is now ignored for this
condition.
( SR:8606310433 CR:JAGae73305 )
During the close of the IP module, improper synchronization
is performed while updating a global list of data
structures.
Resolution:
In IP and IPv6 close routines, list manipulation was
re-ordered to ensure proper synchronization when deleting
data structures.
( SR:8606310867 CR:JAGae73725 )
Multicasting might not work as expected under some
circumstances.
Resolution:
Changes were made to the code so that IP multicasting
works as expected.
( SR:8606311142 CR:JAGae73997 )
During connection establishment, an unacceptable sequence
number is used. This sequence number triggers
TCPS_SYN_SENT-Bad_seq RST from the peer which terminates
the connection establishment process.
Resolution:
The RST generation is avoided by not using an unacceptable
sequence number when establishing the connection.
( SR:8606311531 CR:JAGae74359 )
The system panics when trying to deference a NULL routing
entry pointer.
Resolution:
The code has been modified so that the NULL pointer
dereference is avoided.
( SR:8606312460 CR:JAGae75276 )
If there is an ARP entry whose IP address is not that
of the local subnet, then arp(1M) issued with the "-a"
option will not display all entries.
Resolution:
The code has been modified to display all entries
in the cache.
( SR:8606312501 CR:JAGae75317 )
When a TCP connection is getting closed, it is possible
for a timer associated with that TCP instance to expire
and free the instance data corresponding to that TCP
connection. When the freed data is accessed, a data page
fault results as the data is stale.
Resolution:
Code has been modified to ensure that timers for a TCP
instance do not expire at the same time when its data
structures are being accessed in the close path.
( SR:8606313270 CR:JAGae76068 )
An "ifconfig down <ipv6-interface>" caused a system
panic. The panic was caused by a spinlock window for
a race condition. This could result in system double
panics.
Resolution:
Re-organized spinlocks to avoid race condition.
( SR:8606313678 CR:JAGae76470 )
The ndd(1M) tunables 'tcp_conn_strategy' and
'socket_caching_tcp' perform the same function, which
is to enable/disable socket caching, but have different
help texts.
Resolution:
The 'socket_caching_tcp' and 'tcp_conn_strategy'
ndd(1M) tunables now contain the same help text.
( SR:8606317300 CR:JAGae79861 )
Memory allocated by ip module is not
freed even after the socket is closed.
Resolution:
Memory is being freed when the socket is closed.
( SR:8606317596 CR:JAGae80158 )
The contents of unitdata request are not validated
and hence while doing a bcopy, the system panics,
since the contents contain garbage.
Resolution:
Changes have been made so that the unitdata requests'
contents are validated.
( SR:8606317678 CR:JAGae80233 )
The ICMP packet passed to the nettl subsystem could
be improperly formatted which results in an unknown
ICMP type in traced packets.
Resolution:
The code has been modified to properly
format ICMP packets passed to nettl
subsystem.
( SR:8606318033 CR:JAGae80577 )
A stale pointer to an unplumbed interface may
cause a data page fault, while adding a new route.
Resolution:
While adding a new route, the existence of the interface
is verified to avoid stale pointer reference.
( SR:8606211448 CR:JAGad80636 )
/dev/stcpmap does not return the proper error code to
nettl's ioctl(2).
Resolution:
/dev/stcpmap now returns the proper error code to
nettl's ioctl(2).
( SR:8606319251 CR:JAGae81741 )
Since TOPS is enabled, a message lying on the per-cpu
queue is being delivered to a stream while it is being
reopened. The streams reopen assumes that there will be
no traffic on that stream which is about to be reopened.
As a result of the race condition, SQ_INUSE gets set
which causes accept(2) to hang.
Resolution:
Changes were made so that the per-cpu queue that
the stream was using earlier, will be clean, before
the stream will be reused.
( SR:8606322230 CR:JAGae84698 )
An improper initialization of the ifconf structure
passed to SIOCGIFCONF ioctl in ifconfig code
results in SIGBUS being generated which then results
in ifconfig dumping core.
Resolution:
The code has been modified so that the ifconfig command
properly initializes the structure.
( SR:8606322985 CR:JAGae85451 )
ICMPv6 Redirect messages are supposed to always
be sent with the Target Link Layer Address Option.
The link layer address for this option is currently
being looked up in the Neighbor Cache entry for the
Target. If an entry for the Target in Neighbor Cache is
not present, the ICMPv6 Redirect is not sent.
Resolution:
Code changes have been made to send the
ICMPv6 Redirect Message without the Target
Link Layer Address Option, in the case where
the link layer address of the Target is not known.
( SR:8606323173 CR:JAGae85638 )
The compiler may reorder instructions in such a
way that IP will sometimes generate malformed OOP
headers.
Resolution:
The affected variables are compiled with the
"volatile" keyword. In addition, the accessing
and manipulating of the variables is reordered.
( SR:8606327213 CR:JAGae89474 )
netstat(1) code does not handle the case when a call to the
gethostbyaddr(3N) library routine returns NULL.
Resolution:
The netstat(1) code has been changed to handle the case when
gethostbyaddr(3N) returns NULL. The IP address is now
printed in dotted decimal format if the name resolution
fails.
( SR:8606331109 CR:JAGae92230 )
Mblks in ALLOCB_MBLK_SM arena are not freed before
closing the socket.
Resolution:
Code has been modified to free the queued
mblks allocated in ALLOCB_MBLK_SM arena, when
the socket is being closed.
( SR:8606331382 CR:JAGae92502 )
Invalid values passed through setsockopt(2)
are used by RAWIP and UDP code without
validating them.
Resolution:
Code has been modified to validate arguments
passed through setsockopt(2) and ignore invalid
options.
( SR:8606333354 CR:JAGae94443 )
When recvfrom(2) or recv(2) is issued with
MSG_PEEK flag set on SOCK_DGRAM socket, first
message only should be returned.
Resolution:
recvfrom(2) or recv(2) performed on a
SOCK_DGRAM socket with the MSG_PEEK flag set
now retrieves the first message only.
( SR:8606333610 CR:JAGae94698 )
The traceroute(1) command fails in the SIOCGIFCONF ioctl,
when more than 32 IP-addresses are configured on a system.
Resolution:
The Code has been modified to incorporate dynamic allocation
of the required buffers.
( SR:8606337569 CR:JAGae98576 )
HP-UX sends out an "ICMP Port Unreachable"error message in
reply to a UDP Unicast even when there is an application
bound & listening on the destination port.
This behaviour is NOT fatal or permanent.
It is only an inaccurate report of the actual situation.
Resolution:
The code has been changed to ensure that an
"ICMP Port Unreachable" error message is NOT
sent when the port is actually available.
( SR:8606338847 CR:JAGae99794 )
Under certain circumstances like when the TCP mss is
very small compared to Link MTU and the Path MTU is
smaller than the TCP mss, it will take long time for
PMTU algorithm to converge to arrive at correct value
of PMTU.
Resolution:
The code has been modified so that in calculating
the new PMTU, the size of the packet that triggered
the ICMP fragmentation message by the router,
will be taken into account.
( SR:8606339858 CR:JAGaf00785 )
copyout() is being done from an address in kernel space to
another address in the kernel space incorrectly.
Resolution:
The code has been modified to not use copyout() any more.
( SR:8606342795 CR:JAGaf03689 )
On a multi-homed host, if traceroute(1) is invoked without
the "-s" or the "-i" options specified, then the first
available network interface address will be used as the
source address in outgoing probe packets.
Resolution:
The traceroute(1) manpage has been changed to convey this
information.
( SR:8606343225 CR:JAGaf04119 )
In a multi-threaded process, when shutdown(2) and accept(2)
are invoked concurrently by threads on the same listen
socket of the same process, a race can occur where both
shutdown(2) and accept(2) proceed simultaneously inside
the socket layer. The error ack received for the
T_UNBIND_REQ sent in the shutdown(2) code is processed
by accept(2). This results in a system hang or panic.
Resolution:
The code has been modified to serialize shutdown(2) and
accept(2) calls internally in socket layer.
( SR:8606344448 CR:JAGaf05298 )
There may be a race in a multi-threaded application
between a thread doing an accept(2) on a listen socket
and a thread doing a shutdown(2) on the same listen socket.
Invoking shutdown(2) on a listen socket results in
messages sitting at STH being flushed out. This may
result in an accept(2) hang waiting for T_DISCON_IND which
may have been flushed. The process becomes unkillable.
Resolution:
The code has been modified to serialize accept(2)
and shutdown(2) path internally in the socket layer.
( SR:8606348486 CR:JAGaf09307 )
The sockets code performs incorrectly when it receives
T_BIND_ACK with ADDR_length equal to 0.
Resolution:
The code has been modified to handle the T_BIND_ACK received
from stack for all the cases of ADDR_length.
PHNE_29887:
( SR:8606328244 CR:JAGae89894 )
fstat(2) system call can fail with EPIPE when the
TCP connection associated with the socket has been reset.
Resolution:
fstat(2) system call has been changed so that it
does not fail with EPIPE.
( SR:8606332047 CR:JAGae93167 )
IP does not allow an user to bring up an interface
if the interface went down because of a link down
event. Though this is a temporary event, ifconfig
code considers it to be a fatal error and unplumbs
the interface.
Resolution:
Code has been changed such that SIOCSIFFLAGS
ioctl does not send up undocumented error EAGAIN.
( SR:8606328253 CR:JAGae89895 )
Upon receiving RST from the peer when
establishing connection, a read(2M)
performed on the socket returns ECONNRESET instead
of ECONNREFUSED.
Resolution:
After a non-blocking connect(2M) is performed,
read(2M)now returns ECONNREFUSED if the peer
refuses the connection.
( SR:8606333990 CR:JAGae95074 )
When an interface is brought down manually, one of
the flags being maintained to indicate the interface's
status during cable disconnect and reconnect is not
updated properly. Because of this, the interface appears
to be down permanently because of a cable disconnect.
The symptom only appears in the first down->up
operation done after cable disconnect/reconnect.
If the down->up operation is retried using the
ifconfig command, the interface would come up.
Resolution:
Appropriate code changes have been made to update the
interface status when it is being brought down manually.
PHNE_29472:
( SR:8606285107 CR:JAGae49049 )
Disruption in IGMP membership reporting.
Resolution:
IGMP membership reporting has been improved.
( SR:8606320479 CR:JAGae82961 )
When a connection is terminated by the peer sending
a TCP RST, poll(2) will return a POLLOUT revent
on the associated socket. If it's an X/Open socket,
a subsequent send(2) on this same socket could return
an EWOULDBLOCK failure. This will happen if the
connection was flow-controlled before receiving a RST
from the peer.
Resolution:
send(2) will return EPIPE instead of EWOULDBLOCK
for the X/Open socket.
( SR:8606322708 CR:JAGae85175 )
When a TCP connection request on a loopback
address times out, TCP notifies IP to remove
the corresponding route table entry associated
with the loopback interface (IRE_LOOPBACK).
Resolution:
The route table entry corresponding to the
loopback interface (IRE_LOOPBACK) is not
removed when the Upper Layer Protocol
notifies IP that a TCP connection timed out.
( SR:8606324000 CR:JAGae86465 )
"ndd -get /dev/ip ip_ire_hash" sometimes
takes longer than expected to complete.
Resolution:
"ndd -get /dev/ip ip_ire_hash" has been made more
efficient and has resource restriction built-in.
PHNE_28895:
( CR:JAGae34198 SR:8606269959 )
Symptom:
System panics in the functions
hatcp_set_ha_state() and hatcp_get_status(),
if HATCP functionality is enabled.
Defect Description:
Null pointer dereference due to uninitialized
pointer to message block could cause a panic.
Resolution:
Initialized the pointer to message block and added
checks for null pointer to avoid the null
pointer dereference.
( CR:JAGae47912 SR:8606283966 )
Symptom:
System panics with the following stack
trace.
stack trace for event 0
crash event was a panic
panic+0x14
sbflush+0x130
sbrelease+0x14
sorflush+0x98
sofree+0x98
soclose+0x1b4
soo_close+0x90
closef+0x64
close+0x90
syscall+0x6f8
$syscallrtn+0x0
Defect Description:
When using AF_CCITT sockets, the system
panics in sbflush if there is a zero length
data message block in the socket's
receive buffer.
Resolution:
Changes are incorporated to handle
zero length data message blocks in
the receive buffer.
( CR:JAGae50362 SR:8606286419 )
Symptom:
send(2) may take a long time to complete for
AF_UNIX/SOCK_STREAM type of sockets.
Defect Description:
Frequent allocation and deallocation of memory
may result in longer completion time for send(2).
Resolution:
Code has been changed to reduce the number
of memory allocations and deallocations to
improve performance.
( CR:JAGae52979 SR:8606289048 )
Symptom:
System panics with the following stack trace:
panic+0x6c
assfail+0x3c
vfault+0xc0
trap+0x110c
nokgdb+0x8
ipc_snmp_get+0x4a4
snmpcom_req+0x288
ip6_wput_nondata+0x1c8
ip6_wput+0x370
putnext+0x130
putmsg_subr+0x5ac
putmsg+0x670
syscall+0x830
$syscallrtn+0x0
Defect Description:
System panic occurred while simultaneously
sending data and configuring/unplumbing
IPv6 interface.
Resolution:
Appropriate checks are added to avoid
dereferencing uninitialized components
of IPv6 link layer instance while the
interface is being configured.
( CR:JAGae53258 SR:8606289327 )
Symptom:
System panics with the following stack trace
crash event was a panic
panic+0x6c
report_trap_or_int_and_panic+0x94
trap+0x11c0
nokgdb+0x8
ip_wput_local+0x1314
ip_multicast_loopback+0x40
ip_wput_ire+0xd10
ip_wput+0xd8
putnext+0x130
rawip_wput+0x29c
putnext+0x130
streams_putmsg2+0x41c
streams_putmsg+0x18
sosend+0x222c
sendit+0x438
sendto+0x5c
syscall+0x830
$syscallrtn+0x0
Defect Description:
A wrong queue pointer is being passed to
one of the internal routines in IP,
bypassing the synchronization
at IP level resulting in system
crash.
Resolution:
Code has been modified to pass the correct
queue pointer.
( CR:JAGae54895 SR:8606291052 )
Symptom:
A debug kernel may panic with the following
stack trace:
panic+0x6c
spin_deadlock_failure+0x48
deadlock_check+0x178
sl_pre_check+0xac
spinlock+0x18
check_if_need_to_deliver+0x70
strlog+0x70
x_mi_strlog+0x118
mi_timeout_exec+0x180
invoke_callouts_for_self+0x230
sw_service+0x120
mp_ext_interrupt+0x408
ivti_patch_to_nop3+0x0
cc_spinunlock+0x48
spinunlock+0x6c
vx_itryhold+0xd4
vx_fast_lookup+0x60
vx_lookup+0x210
lookuppnvp+0xb1c
lookuppn+0x60
lookupname+0x40
vn_open+0x16c
copen+0xe8
open+0x44
syscall+0x830
$syscallrtn+0x0
Defect Description:
There is a possibility of an out of order
lock access.
Resolution:
The lock in contention is released prior to
acquiring the second lock.
( CR:JAGae54993 SR:8606291150 )
Symptom:
IPv6 link-local address is not auto-
configured correctly over 802.3 media,
which is activated by 'ifconfig snapX
inet6 up'.
Defect Description:
IPv6 stack does not correctly handle
the media type 802.3 while generating
IPv6 link-local address.
Resolution:
IPv6 stack now handles 802.3 media type
correctly.
( CR:JAGae55495 SR:8606291731 )
Symptom:
System panics with the following stack trace:
stack trace for event 0
crash event was a panic
panic+0x6c
report_trap_or_int_and_panic+0x94
trap+0x117c
nokgdb+0x8
x_mi_timer+0x1c
ip6_nd_solicit+0xb84
ip6_newroute+0xb00
ip6_wput+0x49c
putnext+0x12c
tcp_wput+0x3c0
putnext+0x12c
streams_putmsg2+0x410
streams_putmsg+0x18
sosendfile+0xc28
sendfile+0x258
syscall+0x860
$syscallrtn+0x0
Defect Description:
Data page fault panic happened because
of a race condition that occurred while
sending and receiving a large file and
resetting the interface repeatedly.
Resolution:
The race condition that occurred while
performing address resolution and clean
up of interface related attributes (triggered
by reset of the interface) is fixed.
( CR:JAGae58762 SR:8606295065 )
Symptom:
System panics on a debug flavour kernel with
the following stack trace, if HATCP is enabled
and also HATCP logging is enabled:
spin_deadlock_failure+0x80
deadlock_check+0x230
sl_pre_check+0x2a0
spinlock+0x40
check_if_need_to_deliver+0x80
strlog+0xb0
so_get_tcpstateinfo+0x340
sogetopt+0x2060
getsockopt+0x1b0
syscall+0x1740
Defect Description:
System panic is due to holding a lower order spinlock
and calling strlog() which acquires a higher order
spinlock. This could potentially lead to spinlock
deadlock.
Resolution:
Code changes have been done to replace calls
to strlog() with printf() which does not
acquire a higher order lock.
( CR:JAGae58886 SR:8606295191 )
Symptom:
System panics with the following stack trace
stack trace for event 0
crash event was a panic
panic+0x6c
report_trap_or_int_and_panic+0x94
trap+0xa4c
nokgdb+0x8
CopyInCMsghdrInc+0x54
Defect Description:
Panic while processing control messages.
Resolution:
Code changes have been done to avoid
the panic while processing control
messages.
( CR:JAGae59100 SR:8606295406 )
Symptom:
arp `hostname` does not return any output.
Defect Description:
arp `hostname` does not show whether the
arp entry for local hostname exists.
Resolution:
arp `hostname` will now correctly report
the existence of the arp entry for the
local hostname.
( CR:JAGae59541 SR:8606295895 )
Symptom:
Panic in sogetopt() if HATCP is enabled.
Defect Description:
The HATCP code in sogetopt() is calling
streams_ioctl() which is a blocking call
while holding the socket lock. This could
potentially panic the system due to holding
the lock for a longer period.
Resolution:
The problem is resolved by unlocking
the socket lock before calling
streams_ioctl() in HATCP and then
locking it.
( CR:JAGae60105 SR:8606296535 )
Symptom:
Incorrect value of tcp_was_goodguy in tcp
structure on a standby tcp connection.
Defect Description:
The tcp structure changes need to be updated to
the HATCP replication structure.
Resolution:
The problem is resolved by updating the tcp
structure changes into the HATCP replication
structure and correctly replicating the new
flag onto the standby tcp connection.
( CR:JAGae60136 SR:8606296566 )
Symptom:
The neighbor discovery cache entry structure cannot
be accessed in kernel debugger.
Defect Description:
The neighbor discovery cache entry structure is
defined in a source file.
Resolution:
The definition of neighbor discovery cache entry
structure is moved from a C source file to a header
file.
( CR:JAGae60941 SR:8606297439 )
Symptom:
System panics with the following stack
trace.
optcom_req+0xb8
tcp_wput_proto+0xa4
tcp_wput+0x458
Defect Description:
Panic can happen in optcom_req because
of improperly aligned data.
Resolution:
Added extra checks for proper
boundary conditions and proper
alignment.
( CR:JAGae60957 SR:8606297454 )
Symptom:
"ndd -h tcp_recv_hiwater_max" shows the minimum and
maximum limits for the tunable tcp_recv_hiwater_max
as 1 and 2147483647 respectively. However, the command
"ndd -set /dev/tcp tcp_recv_hiwater_max 2147483647"
returns the error: "operation failed, Invalid argument".
The same error is returned when attempting to set the
tunable udp_recv_hiwater_max, to 1 (which is minimum
limit specified in the online ndd(1M) documentation).
Defect Description:
The limits specified in the online ndd(1M)
documentation for the tunables
tcp_recv_hiwater_max and udp_recv_hiwater_max
are incorrect. Actually, the valid range of
values is [1024 - 1073725440] for
tcp_recv_hiwater_max and [1024 - 2147483647] for
udp_recv_hiwater_max.
Resolution:
The online ndd(1M) documentation has been
changed for the tunables tcp_recv_hiwater_max
and udp_recv_hiwater_max to reflect the
correct range of values.
( CR:JAGae61646 SR:8606298146 )
Symptom:
System hangs due to a slow memory
leak in ALLOC_MBLK_SM arena.
Defect Description:
If an ioctl call fails, the socket
handling code did not free up a
small chunk of memory allocated for
storing the result of ioctl, causing
a memory leak.
Resolution:
If an ioctl call fails, memory allocated
for storing the result is freed.
( CR:JAGae62145 SR:8606298646 )
Symptom:
When interface index in the PKTINFO structure
is set to zero, packets are sent out using
the interface to which the address specified in
the PKTINFO structure is assigned.
Defect Description:
When interface index in the PKTINFO structure
is set to zero, the outgoing interface is not
being decided based on the destination address.
Resolution:
Code changes were made to choose the outgoing
interface based on the packet's destination
address.
( CR:JAGae62168 SR:8606298669 )
Symptom:
System panics with the following stack trace:
ip_rput+0x9c4
putnext+0xcc
pfilstrmodrput+0x328
putnext+0xcc
putnext2+0x48
hp_dlpi_mblk_fast_in+0x408
hp_dlpi_mblk_intr_put+0x91c
streams_put+0xdc
streams_put2+0x58
hp_dlpi_mblk_intr+0x190
lanc_ether_ics+0xfc
btlan_receive_frame+0x648
btlan_isr+0xf4
sapic_interrupt+0x2c
mp_ext_interrupt+0x26c
ihandler+0x90c
Defect Description:
When the IFF_BROADCAST flag is turned ON for a loopback
interface, broadcast routing entries are created
corresponding to the IP address of that interface.
However, if a unicast packet has a target address which
happens to match any of the broadcast addresses of the
interface, the system may panic because of accessing
inconsistent pointer values.
Resolution:
Turning the IFF_BROADCAST flag ON for an interface is
now disallowed if the IFF_LOOPBACK is also ON for that
interface.
( CR:JAGae62465 SR:8606298969 )
Symptom:
System panics with the following stack trace:
stack trace for event 0
crash event was a panic
panic+0x6c
report_trap_or_int_and_panic+0x94
interrupt+0x4e4
$ihndlr_rtn+0x0
puthere+0x1c
mi_timeout_exec+0x268
invoke_callouts_for_self+0x230
soft_intr_handler+0x15c
sw_service+0x1c
mp_ext_interrupt+0x428
ivti_patch_to_nop3+0x0
idle+0x1518
swidle_exit+0x0
Defect Description:
Data page fault panic because of a race
condition that occurred while sending and
receiving a large file and resetting the
interface repeatedly.
Resolution:
The race condition that occurred while
transferring a large file and closing a
module (triggered by reset of the
interface) is fixed.
( CR:JAGae62486 SR:8606298990 )
Symptom:
IREs of type IRE_LOOPBACK remain in the
system even after a RST is sent to the
peer TCP.
Defect Description:
The function that deletes the IREs after
a RST is sent to the peer TCP, deletes only
IREs of type IRE_ROUTE and not IRE_LOOPBACK.
Resolution:
The code has been modified to delete
IRE_LOOPBACK IREs as well when they
are no longer needed.
( CR:JAGae63156 SR:8606299662 )
Symptom:
System may panic in ip6_route_header().
Defect Description:
A NULL pointer dereference can happen
in the code that processes routing
header option.
Resolution:
Appropriate code changes have been
made to avoid the NULL pointer
dereference.
( CR:JAGae63157 SR:8606299663 )
Symptom:
When TCP is uninitialized, an "ICMP
Parameter Problem" Message is not
sent back for an incoming TCP
packet.
Defect Description:
IPv6 fails to send an "ICMP Parameter
Problem" message when there is an incoming
TCP packet before TCP is initialized.
Resolution:
Appropriate code changes have been made
to send an "ICMP Parameter Problem" message
when such a condition occurs.
( CR:JAGae63166 SR:8606299672 )
Symptom:
The MAC type of the lan device is
incorrectly initialized.
Defect Description:
In the IPv6 module, the lan device's
MAC type is initialized to zero. Zero,
actually represents DL_CSMACD.
Resolution:
The lan device's MAC type is now initialized
to DL_OTHER, till the MAC type is known.
( CR:JAGae63773 SR:8606300304 )
Symptom:
IPv6 module might send erratic options
upstream due to accessing of unmapped
memory.
Defect Description:
Unmapped memory access while calculating
option length in destination or hop-by-hop
option processing.
Resolution:
A boundary check would verify the validity
of the memory accessed and an ICMP error
would be returned on unmapped memory access.
( CR:JAGae63774 SR:8606300305 )
Symptom:
The assembled packet from out-of-order fragments might
not show the entire packet that was fragmented.
Defect Description:
While arranging out-of-order fragments, insufficient
memory is allocated to hold the entire IPv6 packet.
Resolution:
Code changes have been made to calculate the
memory requirement correctly.
( CR:JAGae63775 SR:8606300306 )
Symptom:
The IPv6 module can panic on reading
unmapped memory while processing neighbor
discovery options.
Defect Description:
While processing neighbor discovery
options, the IPv6 module fails to
check the length of the options and can
read unmapped memory in a corner case.
Resolution:
The code processing neighbor discovery
options has been appropriately fixed
to take care of boundary conditions.
( CR:JAGae63776 SR:8606300307 )
Symptom:
There is a chance of a spinlock being
released twice while processing
incoming router advertisements which
can lead to a panic.
Defect Description:
While handling a corner case in the
code that processes incoming router
advertisements, the IPv6 module
fails to obtain a spinlock that it
had released earlier. Because of this,
the spinlock can be released twice.
Resolution:
Appropriate code changes have been
made to avoid releasing the spinlock
twice.
( CR:JAGae63777 SR:8606300308 )
Symptom:
IPv6 tries to generate a new link level
header for packets destined to a host
on a point to point link.
Defect Description:
IPv6 module does not differentiate between
packets being transmitted on a point to point
interface and other interfaces that require the
hardware address resolution. Thus, it tries
to generate a new link level header every time
a new route is created.
Resolution:
Appropriate code changes have been made to
handle packets being transmitted on point
to point link separately.
( CR:JAGae64418 SR:8606300948 )
Symptom:
If the source address specified in the
PKTINFO structure is an IPv4 compatible
IPv6 address, the packet is dropped and
an error message is sent to the application.
Defect Description:
When checking for the validity of the
source address mentioned in the PKTINFO
structure, the IPv6 module does not
check if the address is an IPv4 compatible
IPv6 address. Because of this, it drops the
packet and sends an error message to
the application.
Resolution:
Appropriate code changes have been made
to allow IPv4 compatible IPv6 addresses
in the PKTINFO structure.
( CR:JAGae65387 SR:8606302024 )
Symptom:
The system might not be able to configure
an IPv6 address for an interface with
hardware address length less than 6 bytes.
Defect Description:
Checking for the correctness of the actual
hardware address length and the value
stored in the IP packet might fail even
for a correct value in hardware address length
due to wrong calculations. Unmapped memory
might also get padded.
Resolution:
The padding is now done after considering
the code and length of the option along
with the actual hardware address length.
The actual hardware address length is
aligned to the nearest 8-byte boundary
instead of converting the value in the
length field to actual bytes.
( CR:JAGae65452 SR:8606302089 )
Symptom:
Packets to node-local multicast addresses
can get dropped.
Defect Description:
Route for a node-local multicast address
is not being looked up properly.
Resolution:
Proper code changes have been done to
look up the route correctly for
node-local multicast address.
PHNE_28497:
( CR:JAGae13702 SR:8606247262 )
Symptom:
When socket caching is enabled, certain options set
on a previously used socket are still being set on a
newly created socket.
Defect Description:
When a stream is closed, the TCP options set in the
stream are not reset before the stream is reused
for a new connection.
Resolution:
Changes have been made to set default values for
options at the TCP level when socket caching is enabled.
( CR:JAGae13728 SR:8606247288 )
Symptom:
When socket caching is enabled, setting of SO_RCVBUF
and SO_RCVLOWAT options does not work.
Defect Description:
A reused stream head is not reinitialized to the
default state when socket caching is enabled.
Resolution:
When socket caching is enabled a reused stream
head is reinitialized to the default state.
( CR:JAGae34266 SR:8606270028 )
Symptom:
Applications may not receive some ICMP/IGMP
packets that they expect to get from within
the same machine.
Defect Description:
A destination for ICMP/IGMP packets on the local
machine can become invalid, preventing packets
from being delivered.
Resolution:
The destination address is now properly
initialized so that the packets reach their
intended destination.
( CR:JAGae35688 SR:8606271481 )
Symptom:
When socket caching is ON and a very large
number of TCP connections with unsent
data are closed, a very large number of
context switches are observed.
Defect Description:
A single lock is used to synchronize
access to all closed TCP connections
with unsent data in them. This causes
one thread to repeatedly yield until
the lock is available for exclusive
use. Every yield results in a context
switch resulting in a huge number of
context switches.
Resolution:
TCP connections that have unsent data
when closed are no longer moved to the
list that has a single lock controlling
access to the list.
( CR:JAGae36947 SR:8606272841 )
Symptom:
The support for RFC 1948 can be enabled by
setting the ndd(1M) tunable tcp_isn_passphrase.
However, "ndd -get /dev/tcp tcp_isn_passphrase"
fails with the error:
"operation failed, Permission denied"
Defect Description:
The ioctl interface did not return the
information about the status of RFC 1948
support.
Resolution:
Suitable changes have been made in the
code handling the ioctl call.
"ndd -get /dev/tcp tcp_isn_passphrase"
which will now output:
0 - if support for RFC 1948 is disabled
1 - if support for RFC 1948 is enabled
( CR:JAGae38431 SR:8606274354 )
Symptom:
The IPv6 Destination Option before the routing
header and after the routing header cannot be set
on a single socket using the sticky option.
Defect Description:
The second Destination Option overwrites the
first one.
Resolution:
The IPv6 stack now allows setting Destination
Option both before the routing header and after
the routing header correctly.
( CR:JAGae38572 SR:8606274495 )
Symptom:
TCP connection drops packets and applications may hang
or timeout, when the timestamp value passed in
timestamp option in TCP packet rolls over, after the
connection is established.
Defect Description:
When Timestamp option is used for a tcp connection
between two nodes and rollover of time takes place
from 0xffffffff to 0, the packets on the remote side
receiving packets with the rolled over time drops
them thinking that are old packets. This causes
the local node to keep retransmitting the packets.
Resolution:
Code has been modified to take care of
rollover time placed in timestamp option
in TCP packet.
( CR:JAGae38973 SR:8606274896 )
Symptom:
An ICMPv6 error message is sent for IPv6 packets
with IPv6 multicast address as the destination
address. This behavior does not conform to
section 2.4 of RFC 2463.
Defect Description:
An ICMPv6 error message can be sent as a result
of receiving a packet destined to an IPv6 multicast
address.
Resolution:
An ICMPv6 TIME_EXCEEDED error message is not sent
as a result of receiving a packet destined to an
IPv6 multicast address.
( CR:JAGae39036 SR:8606274959 )
Symptom:
Expended routing headers appear in ICMPv6 echo replies.
Defect Description:
Expended source route in the incoming datagram is
left unmodified and thus it appears in the outbound
echo reply.
Resolution:
The routing headers that were part of ICMPv6 echo
request are not included in ICMPv6 echo reply
message.
( CR:JAGae39438 SR:8606275363 )
Symptom:
UDP packets with a checksum value of zero are not
discarded as per RFC 2460.
Defect Description:
IPv6 stack does not check if the checksum is zero
before computing the checksum for correctness.
Resolution:
IPv6 stack now checks for a UDP checksum value of zero
before computing the checksum for correctness.
( CR:JAGae39439 SR:8606275364 )
Symptom:
System panics with following stack trace:
udp_opt_set+0x112c
optcom_req+0x364
udp_wput_other+0xe8
udp_wput+0x858
putnext+0xcc
str_async_ioctl+0x994
hpstreams_ioctl_int+0x478
streams_ioctl+0x34
sosetopt+0x4cc
setsockopt+0x108
syscall+0x62c
$syscallrtn+0x0
Defect Description:
System panics when the IPV6_PKTINFO option is set
in an IPv6 socket.
Resolution:
A null pointer reference in IPV6_PKTINFO
option processing function has been fixed.
( CR:JAGae41767 SR:8606277697 )
Symptom:
When a UDP socket bound to INADDR_ANY is issued a
connect() followed by a disconnect, the state
is not restored completely. This results in an
incorrect source address being used in subsequent
packets sent out.
Defect Description:
When a UDP socket is bound to INADDR_ANY and then
connected to some address, disconnecting the socket
(connecting to an invalid address) does not restore
the socket to the original bound state i.e., the
source address of the socket will not be restored
to INADDR_ANY.
Resolution:
Code has been modified to restore UDP's status properly
after the disconnect is processed.
( CR:JAGae41788 SR:8606277719 )
Symptom:
IPv6 packets can go on the wire with unspecified or
multicast address in routing headers extension. This
does not conform to RFC2373 section 2.5.2 and RFC2460
section 4.4.
Defect Description:
Routing header extension is not checked for unspecified
or multicast address by IPv6 stack.
Resolution:
IPv6 routing extension header is checked for validity.
( CR:JAGae42183 SR:8606278124 )
Symptom:
The ONLINK/OFFLINK status of an IPv6 address
(which was autoconfigured by a prefix in
the router advertisement) is not consistent
with the on-link flag in prefix information
received in a newer router advertisement. This
results in the IPv6 address being treated
as off-link when it is on-link and vice versa.
Defect Description:
The OFFLINK status of an IPv6 address derived from a
prefix advertised by the router does not change
to ONLINK after receiving a subsequent router
advertisement containing the same prefix with the
on-link flag set. This behavior does not conform to
RFC 2461.
Resolution:
IPv6 module is fixed to correctly set the
ONLINK/OFFLINK status of IPv6 address based
on the value of on-link(L) flag in prefix
information in the router advertisement.
( CR:JAGae42942 SR:8606278885 )
Symptom:
"ifconfig lanX" command exits with the error
message "no such interface" when lanX is
configured for IPv6 only.
Defect Description:
When an interface is configured for IPv6 only,
"ifconfig lanX" command exits when retrieving
status information for IPv4 since the interface
is not configured for IPv4.
Resolution:
When an interface is configured for IPv6 only,
ifconfig lanX" command now suppresses the
error encountered while retrieving the status
information for IPv4 and continues with
displaying the status information for IPv6.
( CR:JAGae44226 SR:8606280253 )
Symptom:
Applications may hang while doing RAWIP connect.
Defect Description:
When a failure occurs while doing RAWIP connect,
an error routine is called to propagate up the
error message. This routine is passed the wrong
queue address and the application hangs while
waiting for a response.
Resolution:
The error routine is now passed the correct
queue address. This allows the application
to receive the error message.
( CR:JAGae44261 SR:8606280288 )
Symptom:
System panics with the following stack trace:
panic_save_regs_switchstack+0x110
panic+0x460
post_hndlr+0xab0
vm_hndlr+0x270
bubbledown+0x0
spinlock+0x70
puthere+0x60
ar_entry_squery+0x360
ar_cmd_dispatch+0x420
ar_rput+0x290
puthere+0xe0
ipc_opt_set+0xff0
optcom_req+0x6f0
ipc_optmgmt_req+0x150
ip_wput_nondata+0xef0
ip_wput+0x250
putnext+0xf0
udp_wput_other+0x480
udp_wput+0x500
putnext+0xf0
str_async_ioctl+0x460
hpstreams_ioctl_int+0x1d0
streams_ioctl+0x70
sosetopt+0x3a0
setsockopt+0x1f0
syscall+0xab0
Defect Description:
There is a unsynchronized reference to the next
module's queue pointer which has already been
closed.
Resolution:
Code has been changed to save the current
module's queue pointer and use putnext()
instead of put() which is not safe.
( CR:JAGae44304 SR:8606280331 )
Symptom:
Mobile Node fails to tunnel data to Home Agent.
Defect Description:
There is a corner case which can make the
Mobile Node not to tunnel data to the
Home Agent.
Resolution:
Added reverse tunneling support for MIP Gateway.
Removed dead code on normal MIPv6 tunneling code path.
( CR:JAGae45548 SR:8606281609 )
Symptom:
IRE_DEAD flags are not cleared when a dead gateway
comes up.
Defect Description:
When a dead gateway comes back, it will broadcast
an ARP message. By detecting these broadcast messages
the corresponding dead gateway will be moved to
default gateway list, but IRE_DEAD flag will not be
cleared.
Resolution:
Code has been changed to reset the IRE_DEAD
flag from the corresponding ire when a default
gateway comes up. This is done when we detect
an ARP broadcast message transmitted by the gateway
when it comes up.
( CR:JAGae46274 SR:8606282322 )
Symptom:
get_logical_stat() does not return correct
results when the application is compiled
in 64-bit mode.
Defect Description:
There is a mismatch in internal structures
used by get_logical_stat() and hence it
returns incorrect results when compiled
in 64-bit.
Resolution:
Code has been modified so that there will
no longer be any mismatch between internal
structures. get_logical_stat() return value will
be the same for both 32-bit and 64-bit compilation.
( CR:JAGae46724 SR:8606282772 )
Symptom:
An unexpected result can be returned to a user
application when getting the IPv6 extension header
in TCP mode.
Defect Description:
In some cases, the IPV6 stack uses the wrong list
to look up an extension header.
Resolution:
IPv6 stack now uses the right list when looking
up the IPv6 extension headers.
( CR:JAGae46920 SR:8606282965 )
Symptom:
System panics with the following stack trace:
stack trace for event 0
crash event was a panic
panic+0x6c
report_trap_or_int_and_panic+0x94
trap+0xd78
nokgdb+0x8
ip_wput+0x3b4
putnext+0xf0
rawip_wput+0x1b0
putnext+0xf0
streams_putmsg2+0x470
streams_putmsg+0x18
sosend+0x21f4
sendit+0x49c
sendto+0x5c
syscall+0x3c8
$syscallrtn+0x0
Defect Description:
IP attempts to update a data structure
with the results of a search for a
multicast group without locking it
first. A race condition exists where
the data structure can be freed before
the update is completed.
Resolution:
Code changes have been made to provide
proper lock while looking up the
interface group and storing the return
pointer to a multicast interface.
( CR:JAGae47736 SR:8606283791 )
Symptom:
Inconsistent result when binding to
deprecated address. The bind() fails in
the case of an address deprecated by
router advertisement and succeeds in
the case of a preferred lifetime
expiration.
Defect Description:
When binding to an address which has been deprecated,
an application sees different behavior depending on
whether the address is deprecated by Router
Advertisement(RA) or by the preferred lifetime
expiration.
Resolution:
The result of binding to deprecated address is made
consistent. It succeeds both in the case of address
deprecated by Router Advertisement and by preferred
lifetime expiration.
( CR:JAGae48092 SR:8606284145 )
Symptom:
There is a short time delay before T_DISCONNECT event/s
are passed up to a TCP endpoint.
Defect Description:
T_DISCONNECT event is not being passed up until after a
t_accept.
Resolution:
The stream head for a listen socket is now bound,
so that it will limit the number of SYN/RST pairs
that can come in before it starts to drop them.
( CR:JAGae48167 SR:8606284220 )
Symptom:
Disconnect request to a rawip module,
does not restore the original status of
the end-point.
Defect Description:
If a rawip end-point which is bound to
INADDR_ANY and connected to another
address, is issued a disconnect, the
end-point's initial state is not restored.
Resolution:
Code has been changed to restore the
state of the rawip end-point as it
was in the initial bound state.
( CR:JAGae48168 SR:8606284221 )
Symptom:
Application may be notified that udp disconnect
operation succeeded even though the disconnect
operation actually failed.
Defect Description:
UDP disconnect code is incorrectly sending a
positive acknowledgement to the disconnect
request prematurely.
Resolution:
UDP disconnect code has been changed so
that sending an acknowledgment is deferred
until the disconnect operation is completed.
( CR:JAGae48734 SR:8606284791 )
Symptom:
Potential system panic in mi_ibc_qenable().
Defect Description:
The mi_ibc_qenable() function referencing freed
memory could cause system panic.
Resolution:
Fixed mi_ibc_qenable() to avoid referencing
freed memory.
( CR:JAGae48736 SR:8606284793 )
Symptom:
System may panic in udp_hash_remove_port()
if connect rebind error occurred.
Defect Description:
The connect rebind error path doesn't have a
lock before calling udp_hash_remove_port.
This could cause system panic.
Resolution:
Code changes have been made to use a lock
before calling udp_hash_remove_port().
( CR:JAGae48849 SR:8606284907 )
Symptom:
System panics with the following stack trace
panic+0x6c
report_trap_or_int_and_panic+0x94
trap+0xa78
nokgdb+0x8
rawip_opt_massage+0x58
rawip_wput+0x550
putnext+0xcc
streams_putmsg2+0x10c
streams_putmsg+0x18
sosend+0x80c
sendit+0x154
sendto+0x5c
syscall+0x62c
$syscallrtn+0x0
Defect Description:
The panic happened due to a memory overrun in
the rawip_opt_massage().
Resolution:
Necessary check has been added in the code to
prevent memory overrun.
( CR:JAGae48995 SR:8606285053 )
Symptom:
Stale routing information is used even after
updating the route. This results in packets
being routed incorrectly.
Defect Description:
IPv6 stack does not update the route cache
entry associated with the changed route.
Resolution:
IPv6 stack now updates the route cache entry
associated with the changed route correctly.
( CR:JAGae49066 SR:8606285124 )
Symptom:
Packets with loopback address as source address
are sent outside the node. This behavior does not
conform with RFC 2373 section 2.5.3.
Defect Description:
When an IPv6 socket is bound to loopback address
(::1), packets with loopback address as the
source address, are incorrectly being sent out
of the system.
Resolution:
The problem is fixed by checking if the source
address is the loopback address before sending the
packet out of the node. If the source address
is the loopback address then the packet is not
transmitted and ICMP6_ADMIN_PROHIBITED error is
sent upstream.
( CR:JAGae49207 SR:8606285262 )
Symptom:
ndp(1M) shows valid/preferred lifetimes beyond the
value advertised by the router.
Defect Description:
IPv6 stack changes the lifetimes advertised by the
router if there is already another timer active.
Resolution:
IPv6 stack now keeps track of lifetimes advertised by
the router separate from the timer.
( CR:JAGae49284 SR:8606285339 )
Symptom:
Excessive "Neighbor Solicitation" and "Neighbor
Advertisement" messages can be seen on the wire.
Defect Description:
IPv6 sets the state of neighbor to STALE even
if reachability confirmation was recently received.
Resolution:
The code has been changed to not lose reachability
confirmation.
( CR:JAGae49372 SR:8606285427 )
Symptom:
connect() called on an already connected UDP DGRAM
socket leads to memory leak in ALLOCB_MBLK_SM arena.
Defect Description:
During every successive connect() on UDP DGRAM socket
a new mblk is allocated without freeing the previously
allocated mblk, which results in a memory leak.
Resolution:
Instead of reallocating new mblk for successive
connect() on an already connected UDP DGRAM socket,
the existing mblk is being reused.
( CR:JAGae50606 SR:8606286667 )
Symptom:
Executing the command
"/usr/bin/ndd -get /dev/sockets socket_udp_sndbuf_default"
returns the error:
"The object name 'socket_udp_sndbuf_default' is too long."
Defect Description:
Because of a typo in the code, the tunable
"socket_udp_sndbuf_default" was not recognized correctly.
Resolution:
The typo in the code has been corrected so that
the tunable 'socket_udp_sndbuf_default' is
recognized correctly.
( CR:JAGae50872 SR:8606286932 )
Symptom:
IPv6 stack stops processing "router advertisement"
messages sent by the local routers.
Defect Description:
The packets sent to all-nodes multicast (ff02::1)
address by the router are not received by IPv6 stack
once local application leaves node-local multicast
address (ff01::1).
Resolution:
The IPv6 code is changed to differentiate between
node-local and all-nodes multicast address.
( CR:JAGae50958 SR:8606287019 )
Symptom:
When there are multiple IPv6 Destination
Options only the first Destination Option
can be looked up using the sticky option.
Defect Description:
The first Destination Option is always
returned regardless of which option is
requested.
Resolution:
IPv6 stack now returns the requested Destination
Option correctly.
( CR:JAGae50959 SR:8606287020 )
Symptom:
IPv6 Echo Reply could be sent with multicast
address as a source address of packet. This
does not conform to RFC 2463.
Defect Description:
IPv6 stack does not check for multicast
address as a source address when sending
Echo Reply.
Resolution:
The packet is discarded if it contains multicast
address as a source address for IPv6 Echo Reply.
( CR:JAGae51088 SR:8606287149 )
Symptom:
recv() returned with errno set to ECONNRESET
due to inappropriate RST.
Defect Description:
When socket caching is turned "ON", a TCP packet
may be delivered to the wrong connection when
the original connection is closed and reused.
This generates RST and causes recv() to fail
with ECONNRESET.
Resolution:
The generation of inappropriate RST is avoided
by not reusing the stream until the connection
is closed by both the peers.
( CR:JAGae51169 SR:8606287230 )
Symptom:
There is a delay seen in SIOCMND ioctl
processing.
Defect Description:
There is delay in ioctl SIOCMND to send unsolicited
neighbor advertisements due to an incorrect usage of
the deferred sending mechanism. This behavior causes
an unacceptable delay in SG failover.
Resolution:
The performance problem has been fixed by using
the deferred sending mechanism correctly.
( CR:JAGae51708 SR:8606287775 )
Symptom:
The width of the first column (Name) of
the "netstat -in" output is limited to
8 characters. If the name of the interface
is longer than 8 characters, it gets truncated.
Defect Description:
The maximum length of the interface name
was incorrectly taken to be 8.
Resolution:
A new command line option "-w" has been introduced,
which will display all the fields of netstat -in" with
their maximum possible resolution. A sample usage
would be "netstat -inw".
( CR:JAGae53982 SR:8606290053 )
Symptom:
When an AF_INET socket is bound to INADDR_ANY
and if an AF_INET6 socket tries to bind to the
same port with an IPv6 address, bind() does not
fail.
Defect Description:
The second bind() succeeds incorrectly when an
AF_INET socket is first bound to inaddr_any and
then an AF_INET6 socket is bound to a specific
IPv6 address. The problem is seen only with TCP.
Resolution:
The second bind() now fails with errno EADDRINUSE
in the case AF_INET socket is first bound to
inaddr_any and AF_INET6 socket is bound to a
specific IPv6 address.
( CR:JAGae53985 SR:8606290056 )
Symptom:
IPv6 applications do not receive both Destination
Options when using ancillary data with recvmsg(2).
Defect Description:
IPv6 stack does not format correctly ancillary
data when both Destination Options are present
in a received packet.
Resolution:
IPv6 now formats ancillary data correctly when
both Destination Options are present.
( CR:JAGae54172 SR:8606290246 )
Symptom:
The "ndp" command shows nothing in "State" and
"Flags" fields for point-to-point interface.
Defect Description:
When issuing SIOCGND ioctl from user space on
a point-to-point interface, the IPv6 stack returns
EINVAL.
Resolution:
IPv6 stack does not return EINVAL anymore when
issuing SIOCGND on a point-to-point interface from
user space.
( CR:JAGae54691 SR:8606290848 )
Symptom:
System panic could happen with the following
stack trace:
panic+0x6c
assfail+0x3c
btlan_hw_req+0x2a0
btlan_resolved_output+0x18
hp_dlpi_wput+0xbc
putnext+0x12c
ip6_wput_ire+0x4f0
ip6_wput+0x33c
putnext+0x12c
tcp_wput+0x3c0
tcp_rput+0x3e40
puthere+0x12c
put_release+0x22c
ip6_rput+0x1a8c
putnext+0x12c
hp_dlpi_intr_put+0x80c
streams_put+0x124
hp_dlpi_ether_intr+0x31c
btlan_receive_frame+0x464
btlan_isr+0x100
sapic_interrupt+0x34
mp_ext_interrupt+0x408
ivti_patch_to_nop3+0x0
idle+0x11e8
Defect Description:
IPv6 stack could update "route entry" with
the wrong fastpath information.
Resolution:
IPv6 stack has been modified to only update
the corresponding "route entry" attached to
the driver from where the fastpath ack is received.
( CR:JAGae54779 SR:8606290936 )
Symptom:
When socket caching is enabled, if a
listening socket is closed when there
are eager connections on the socket,
another process or thread cannot bind to
the same address for some time.
Defect Description:
When socket caching is enabled, the
T_UNBIND_REQ for a listening stream will
not succeed while there are eager
connections. The listening stream will
have to be closed in a separate context.
While in this transient phase, no other
thread or process can bind to the same
address.
Resolution:
Code changes have been done so that
a T_UNBIND_REQ is not issued for a
listening stream. Instead, the stream
is closed.
( CR:JAGae55682 SR:8606291918 )
Symptom:
Spurious Service Guard cluster reformation occurs
or Service Guard daemon aborts due to delay caused
by SIOCSWITCHLIF6ADDR ioctl.
Defect Description:
There is a delay due to duplicate address detection
while performing SIOCSWITCHLIF6ADDR ioctl. This
impacts Service Guard failover.
Resolution:
Since the address is already determined to be unique
when an address is assigned to an interface,
duplicate address detection is disabled when an IPv6
address is switched from one interface to another
interface on the same node. This improves the
performance of the SIOCSWITCHLIF6ADDR ioctl.
( CR:JAGae55684 SR:8606291920 )
Symptom:
Bringing up an IPv6 interface can hang forever.
Defect Description:
If we receive a link-down indication from a driver
while we are in the middle of Duplicate Address
Detection, the IPv6 stack does not reply to the
pending ioctl.
Resolution:
IPv6 stack now replies to the pending ioctl when
link-down indication is received.
( CR:JAGae56041 SR:8606292278 )
Symptom:
The system can panic in the function
ire6_create().
Defect Description:
Null pointer reference exists in the kernel
function that creates routing entries.
Resolution:
Code changes have been done to eliminate
NULL pointer reference.
( CR:JAGae56336 SR:8606292583 )
Symptom:
When an ICMP Address Mask Request packet is sent to
a unicast address, an ICMP Address Mask Reply is sent
even though the tunable
"ip_respond_to_address_mask_broadcast"
is turned off.
Defect Description:
Turning the ndd tunable
"ip_respond_to_address_mask_broadcast" OFF,
only suppresses the replies to the request sent
to the broadcast address.
Resolution:
Code has been modified so that when the
tunable
"ip_respond_to_address_mask_broadcast"
is turned off, any reply to address mask
request is completely suppressed.
( CR:JAGae57090 SR:8606293340 )
Symptom:
System panics because of an error handling code
which attempts to ensure that no spinlocks are
being held while doing certain operations.
Defect Description:
A spinlock was unlocked without being held.
This corrupted some of the data structures that
led to assertion failures.
Resolution:
The code has been changed to ensure that
we hold the spinlock before releasing it later.
( CR:JAGae58674 SR:8606294977 )
Symptom:
read(2) on a socket sometimes, may return
ECONNREFUSED instead of ECONNRESET.
Defect Description:
In some cases TCP sends up a ECONNREFUSED.
read(2) just propagates it to the application.
Resolution:
Appropriate code changes have been made
for read(2) to return ECONNRESET whenever
it gets a ECONNREFUSED from TCP.
( CR:JAGae58883 SR:8606295188 )
Symptom:
System panics with the following stack trace
panic+0x14
report_trap_or_int_and_panic+0x4c
interrupt+0x1e8
$ihndlr_rtn+0x0
ipc_walk+0x7c
ill_down_ind+0x118
ip_rput_dlpi+0x9b0
ip_rput+0x394
putnext+0xc4
hp_dlpi_event+0x1b0
HP1_ospif_mib_event+0x150
ospif_watchdog+0x3c
HP1_oim_timeout+0x10
invoke_callouts_for_self+0xac
sw_service+0x8c
inttr_emulate_save_fpu+0x100
drv_priv+0x0
ip_open+0x11c
open_wrapper+0x44
csq_protect+0x10c
osr_open+0xaec
pse_open+0xe4
streams_open+0x48
soclone+0x178
soaccept+0xf4
sodequeue+0xac
accept+0x204
syscall+0x6e8
$syscallrtn+0x0
Defect Description:
An ipc_t was inserted into the hash array before
it was initialized. As a result any thread
walking through the hash array may panic if it
tries to dereference that uninitialized ipc_t.
Resolution:
Necessary checks have been added before
dereferencing the ipc_ts to avoid
the panic.
PHNE_28089:
( SR:8606283620 CR:JAGae47565 )
Symptom:
connect(2) to an already connected UDP
socket fails with EISCONN.
Defect Description:
An error message is incorrectly generated by
UDP when a connect(2) is done on an
already connected socket.
Resolution:
The incorrect error message that lead to
the failure is no longer generated.
PHNE_27989:
( SR:8606250244 CR:JAGae16624 )
Symptom:
IPv6 sockets cannot join IPv4-mapped IPv6
multicast address groups.
Defect Description:
Socket option IPV6_JOIN_GROUP is not handled
properly leading to the above symptom.
Resolution:
The IPv6 module is fixed to allow IPv6 sockets
to join and leave IPv4-mapped IPv6 multicast
address groups.
( SR:8606254434 CR:JAGae18766 )
Symptom:
Upon encountering an error while parsing the
/etc/rc.config.d/nddconf file, ndd returns a failure
instead of printing out details of the error.
Defect Description:
The function that parses the /etc/rc.config.d/nddconf
file returns failure instead of printing out details
of the error.
Resolution:
Detailed information is provided for each nddconf
syntax error.
( SR:8606268603 CR:JAGae32841 )
Symptom:
Unable to ping a machine over a VLAN interface
if nettl tracing of ns_ls_ip is enabled on
that machine.
Defect Description:
ICMP echo packets coming out of a VLAN
interface get corrupted when nettl tracing
of ns_ls_ip is enabled on a machine.
Resolution:
Code has been modified to prevent corruption of
outbound packets on a VLAN interface when nettl
tracing of ns_ls_ip is enabled.
( SR:8606269955 CR:JAGae34194 )
Symptom:
System may panic with a data page fault when using
extension headers with RAWIP6 packets.
Defect Description:
A NULL pointer dereference can happen when dealing
with RAWIP6 packets containing extension headers.
Resolution:
Code has been changed to not dereference the
pointer if it is NULL.
( SR:8606270176 CR:JAGae34414 )
Symptom:
System panics with a data page fault.
Defect Description:
A local variable in a function that keeps track
of the length of a mblk is used before it is
initialized.
Resolution:
The local variable is initialized to an appropriate
value before it is used.
( SR:8606270270 CR:JAGae34508 )
Symptom:
Destination port for the SYN+ACK packet is set to
zero when IPV6_DSTOPTS is set by setsockopt(2) for
the TCP/IPv6 Listen socket.
Defect Description:
Incomplete headers are copied into the outgoing
packet when IPv6 extension headers are set on
a TCP/IPv6 listening socket.
Resolution:
TCP now fills the outgoing packet with proper
data when IPv6 extension headers are set on a
TCP/IPv6 listening socket.
( SR:8606270296 CR:JAGae34534 )
Symptom:
The ndd parameter ip6_nd_anycast_delay does
not work as documented in the ndd help text.
Defect Description:
The NDD tunable ip6_nd_anycast_delay is not
implemented on HP-UX.
Resolution:
ndd help text has been updated to state that
the tunable ip6_nd_anycast_delay is not
currently implemented.
( SR:8606270780 CR:JAGae35016 )
Symptom:
IPv6 Neighbor Unreachability Detection messages
may synchronize with each other.
Defect Description:
ReachableTime is set directly from the router
advertisement message without calculating
a random value.
Resolution:
ReachableTime is now set randomly as per RFC
2461 section 6.3.4.
( SR:8606271001 CR:JAGae35237 )
Symptom:
In some cases, the IPv6 stack responds to invalid
Neighbor Solicitation messages.
Defect Description:
The IPv6 stack does not discard a Neighbor Solicitation
message which has an unspecified source address and
a destination address which is not a solicited-node
multicast address.
Resolution:
The IPv6 stack now discards an invalid Neighbor
Solicitation message.
( SR:8606271039 CR:JAGae35275 )
Symptom:
The system continues to forward packets even if
forwarding interfaces are disabled using
"ifconfig lanX inet6 private"
Defect Description:
The variable keeping track of the number of forwarding
interfaces is not decremented after unplumbing a
forwarding interface.
Resolution:
The number of "forwarding interfaces" is decremented
once a forwarding interface is unplumbed.
( SR:8606271235 CR:JAGae35466 )
Symptom:
The unspecified address, :: is set as source address
in packets sent to a solicited-node multicast address
even after a unicast address is assigned to the outgoing
interface. This behavior does not conform to RFC 2373.
Defect Description:
HP-UX IPv6 implementation incorrectly sets the source
address to the unspecified address, :: when packets are
destined to a solicited-node multicast address even after
a unicast address is assigned to the outgoing interface.
Resolution:
IPv6 module is fixed to select an appropriate source
address when packets are sent to a solicited-node
multicast address.
( SR:8606271514 CR:JAGae35722 )
Symptom:
System panics with a data page fault with the following
stack trace
nokgdb+0x8
tcp_connect+0x2a0
tcp_wput_proto+0xbc
tcp_wput+0x418
putnext+0xcc
str_async_ioctl+0x540
hpstreams_ioctl_int+0xf8
streams_ioctl+0x34
soconnect+0x134
connect+0xdc
syscall+0x62c
$syscallrtn+0x0
Defect Description:
A wrong lock is obtained when accessing the list
of TCP established connections.
Resolution:
Code has been corrected to grab the proper lock before
accessing the list of TCP established connections.
( SR:8606272324 CR:JAGae36464 )
Symptom:
While waiting for IPv6 address resolution to complete,
the system runs very short of memory.
Defect Description:
IPv6 does not limit the number of packets queued
while waiting for address resolution to complete.
Resolution:
IPv6 module is fixed to limit the number of packets
queued while waiting for address resolution to
complete as required by RFC 2461. The number of
packets queued is limited to 5.
( SR:8606272765 CR:JAGae36872 )
Symptom:
If an application using UDP sockets first connects
to an IPv4-mapped IPv6 address, disconnects and tries
to connect to a native IPv6 address, the second connect
fails.
Defect Description:
The UDP module does not unbind from the old address
family and bind to the new address family and this
causes the connect in the above scenario to fail.
Resolution:
UDP module now unbinds from old address family and
binds to the new address family before connecting
to a new address.
( SR:8606273477 CR:JAGae37565 )
Symptom:
System may panic with a data page fault with the following
stack trace if the primary interface is brought up with
0.0.0.0 address when a secondary interace is already up.
panic+0x6c
report_trap_or_int_and_panic+0x94
trap+0xed4
nokgdb+0x8
ipif_arp_up+0x2d8
ipif_up_arp_and_ires+0x598
ipif_up+0x264
ip_sioctl_copyin_done+0x1720
ip_wput_nondata+0x32c
ip_wput+0x158
csq_turnover_with_lock+0x1fc
putnext+0x1e4
udp_wput_other+0x130
udp_wput+0x720
putnext+0xcc
wait_iocack+0x6c
str_trans_ioctl+0x4f8
hpstreams_ioctl_int+0x174
streams_ioctl+0x34
ifioctl+0x88
soo_ioctl+0x7c
ioctl+0x120
syscall+0x750
$syscallrtn+0x0
Defect Description:
In the above scenario no arp command mblks are
generated which leads to a NULL pointer
dereference in the IP code.
Resolution:
Necessary checks are introduced to prevent the
NULL pointer dereference.
( SR:8606273939 CR:JAGae38021 )
Symptom:
Query of the mib object ID_ipv6IfLowerLayer fails.
Defect Description:
The mib object ID_ipv6IfLowerLayer is not updated
correctly.
Resolution:
The problem is fixed by setting the correct
values for the mib object ipv6IfLowerLayer.
( SR:8606274153 CR:JAGae38231 )
Symptom:
Neighbor cache entry state changes not implemented
as per RFC 2461.
Defect Description:
IPV6 does not implement Neighbour CACHE entry as per
RFC2461.
Resolution:
IPV6 code has been modified to implement Neighbour
discovery cache entry as per RFC 2461.
( SR:8606277348 CR:JAGae41419 )
Symptom:
After IPv6 address is switched from a primary
interface to backup interface, different NMIDs are
reported by IPv6 API if_nametoindex() and lanadmin
command for the same interface.
Defect Description:
After IPv6 address is switched from a primary
interface to backup interface using the ioctl
SIOCSWITCHLIF6ADDR, the network management ID (NMID)
of the primary and backup interfaces are not
consistent between IPv6 module and driver.
Resolution:
The network management ID is also swapped in
the driver when an IPv6 address is switched
from one interface to another. Now the network
management ID is consistent between IPv6
module and drivers.
( SR:8606278217 CR:JAGae42276 )
Symptom:
Transport does not provide a direct method for
lower levels to request a gratuitous arp.
Defect Description:
A link will send a DL_LINK_DOWN and a DL_LINK_UP
message upstream if it wants to have a gratuitous
arp transmitted. Tables associated with the
interface will be deleted and recreated. This
is an expensive operation when all that is desired
is a gratuitous arp transmission.
Resolution:
The ARP module will now recognize a message
from DLPI requesting a gratuitous arp.
( SR:8606279349 CR:JAGae43402 )
Symptom:
After installing PHNE_27730, ping -n with a packet size
between 8 and 15 bytes will always result in the loss of
a packet.
Defect Description:
ping -n with packet size between 8 and 15 bytes
always misses the last echo reply and thus reports
packet loss information incorrectly.
Resolution:
Code has been modified to wait for the last echo
reply when the packet size is between 8 and 15 bytes.
The time to wait for the last echo reply will be
equal to twice the time specified through the -I
option. If the -I option is not used, the wait
time will default to two seconds.
( SR:8606281411 CR:JAGae45354 )
Symptom:
On a system that does not have IPv6 installed, if there
is an IPv6 entry in the /etc/rc.config.d/nddconf file, ndd
skips processing the next entry in the file.
Defect Description:
The function that parses the /etc/rc.config.d/nddconf
file skips processing the next entry when it encounters
an IPv6 entry on a system that does not have IPv6
support.
Resolution:
Code has been changed so that "ndd -c" will
continue processing the next entry if it encounters
the problem described above.
( SR:8606281503 CR:JAGae45445 )
Symptom:
Arp mapping and publish entries may unexpectedly
disappear.
Defect Description:
Code inspection revealed a potential DOWN/UP race
that may lose arp mapping and publish entries.
Resolution:
The potential race condition has been eliminated.
PHNE_27730:
( SR:8606241187 CR:JAGae08450 )
Defect Description:
A broadcast SYN packet is being delivered to
TCP. There is a huge buildup in the
established connection array. This causes
lookups into the array to take a long time.
Resolution:
Inbound broadcast packets are prevented from
reaching TCP. These packets are filtered out
at the IP level.
( SR:8606247981 CR:JAGae14381 )
Defect Description:
The socket code does not have the hooks needed
for the Network Server Accelerator to function.
Resolution:
The socket code has been modified to provide
hooks to support the Network Server Accelerator
feature.
( SR:8606264643 CR:JAGae28972 )
Defect Description:
If a router advertisement message is received
while the primary IPv6 address auto-configuration
is in progress to bring up the interface,
the interface configuration fails.
Resolution:
The "router advertisement" message is now
ignored if the interface is not completely
up and ready to respond.
( SR:8606265738 CR:JAGae29995 )
Defect Description:
Neighbour discovery timer was not being cancelled
when IPv6 module was closed.
Resolution:
The IPv6 module close routine has been fixed to
cancel all outstanding timers.
( SR:8606267968 CR:JAGae32207 )
Defect Description:
An error in casting the filesize results in the
truncation of a 64 bit size to a 32 bit size.
Resolution:
Appropriate code changes have been made to sendfile(2)
to ensure proper typecasting.
( SR:8606270271 CR:JAGae34509 )
Defect Description:
Setting of socket options depended on the contents
of streamtab data structure and this structure
is modified when nettl is enabled. Hence the
failure to set socket options.
Resolution:
Code has been changed such that when IPv6 tracing
is enabled, setting of the socket options will work
properly even without bind(2) being called.
( SR:8606270345 CR:JAGae34583 )
Defect Description:
Information pertaining to devices that do not need
a resolver such as point to point interfaces, was
not being processed.
Resolution:
IPv6 module has been fixed to return neighbour
discovery entries corresponding to all interfaces.
( SR:8606271213 CR:JAGae35448 )
Defect Description:
Resetting IFF_MULTICAST flag for an UP mobility
tunnel and then bringing it down or unplumbing
it caused corruption in a multicast list.
Potentially, this could lead to a panic sometime
later.
Resolution:
Code has been modified to check for and avoid
conditions that would corrupt the multicast
list.
( SR:8606273232 CR:JAGae37333 )
Defect Description:
If a stream was closed on which tracing
is enabled, active references to it are
kept even after closing.
Resolution:
Code has been modified such that all
active references to a module are removed
before it is closed even when nettl tracing
is enabled.
( SR:8606274682 CR:JAGae38759 )
Defect Description:
A timing window allows the port allocation
routine to allocate and return port 0 as an
ephemeral port. The calling routine treats
the return value of 0 as failure. A subsequent
series of events leads to the panic.
Resolution:
Code has been modified such that the timing
window is avoided and hence port 0 is not
allocated as an ephemeral port.
PHNE_27382:
( SR:8606170818 CR:JAGad40082 )
Defect Description:
The ping command does not return different error
values in the following cases:
* Unreachable host/network
* Other errors.
Hence, when used in scripts it is not possible to
infer the reason for exit.
Resolution:
Code has been modified to return different error
values on exit.
( SR:8606232612 CR:JAGae01847 )
Defect Description:
When searching for "network route" in route table
that has multiple entries for that network, if the first
entry in the route table is marked invalid then any host
on that network becomes unreachable even though another
route entry exists in the route table.
Resolution:
Code has been changed to search the route table even
after encountering an invalid entry.
( SR:8606232989 CR:JAGae02214 )
Defect Description:
Two new options are necessary to set the time interval
between two successive transmissions and wait-time after
the last transmission, similar to the HP-UX 10.20 ping
command.
Resolution:
Code has been modified to provide two additional
options (I and m) as detailed below:
-I : To set the time interval between two
successive packet transmissions
-m : To set the wait time after the last packet
has been transmitted
The ping(1M) man page has been updated accordingly.
( SR:8606237266 CR:JAGae06311 )
Defect Description:
If an already connected datagram socket is passed
to connect(), the datagram socket is not disconnected.
Resolution:
Code has been modified to disconnect the socket when
invalid arguments are passed to connect() on an
already connected datagram socket.
( SR:8606238197 CR:JAGae07224 )
Defect Description:
When there is only normal data available to be
read in the stream head and if the application
calls recv() with the MSG_OOB flag, part of the
normal data is lost.
Resolution:
Code has been changed so that message blocks
containing normal data are not freed when
retrieving urgent data.
( SR:8606241192 CR:JAGae08455 )
Defect Description:
Timer mblks are allocated at the time of opening
a TCP stream in ALLOCB_MBLK_SM arena.
In a rare case, these mblks are not freed on
closing the TCP stream.
Resolution:
Code has been modified to ensure that the timer
related mblks are freed at the time of
closing the TCP stream.
( SR:8606243224 CR:JAGae10458 )
Defect Description:
Lock contention occurs in a hash table at index
zero due to unbound TCP connections being inserted
at this index.
Resolution:
Unbound TCP connections are no longer inserted into
index zero of the hash table.
( SR:8606244252 CR:JAGae10742 )
Defect Description:
If an application reduces its receive buffer
size after the completion of the 3-way handshake,
the TCP window scales down.
Resolution:
TCP window scale calculation has been
corrected to have a more accurate setting.
( SR:8606245677 CR:JAGae12137 )
Defect Description:
The ARP module returns success although
the setting of the parameter to specified value
fails.
Resolution:
The ARP module now returns an error if the
setting of parameter fails.
( SR:8606246302 CR:JAGae12747 )
Defect Description:
Memory leak occurs in the IP outbound path
in case of failures while forwarding a
packet.
Resolution:
Memory is always released when failures occur
during the forwarding of a packet.
( SR:8606246303 CR:JAGae12748 )
Defect Description:
Memory leak occurs in case of an unrecognized
DL_OK_ACK response.
Resolution:
Memory is now freed in case of unrecognized
DL_OK_ACK response.
( SR:8606246305 CR:JAGae12750 )
Defect Description:
Memory leak occurs in the case of failures
while setting up an ARP resolver for a new
interface.
Resolution:
Memory is now freed in the case of
failures while setting up an ARP resolver for a
new interface.
( SR:8606246307 CR:JAGae12752 )
Defect Description:
Potential memory leak while initializing
an interface.
Resolution:
Memory is now freed to avoid potential memory
leaks.
( SR:8606246311 CR:JAGae12756 )
Defect Description:
Memory might be freed twice resulting in system
panic.
Resolution:
Code has been modified to avoid freeing the memory
twice.
( SR:8606246314 CR:JAGae12759 )
Defect Description:
The PMTU set via route command is not used when
creating fully resolved route to a host.
Resolution:
The specified pmtu is now used while creating fully
resolved route to a host.
( SR:8606247242 CR:JAGae13682 )
Defect Description:
Memory leak occurs in case of errors in
binding a UDP socket.
Resolution:
Memory is now freed in case of errors while
binding a UDP socket.
( SR:8606247710 CR:JAGae14113 )
Defect Description:
Memory leak occurs in TCP reinitialization.
Resolution:
Memory is now freed while reinitializing TCP.
( SR:8606248700 CR:JAGae15098 )
Defect Description:
When determining the RTO value for a new connection
for which there exist previous RTT measurements,
the RTO chosen may be larger than what it needs to be.
This happens because the previous RTT estimates
are not stabilized and the RTT deviation estimate
is not cached between connections.
Resolution:
The RTT deviation estimate is now cached
between connections. The RTO algorithm
and initial values have been modified to converge
more rapidly on the optimal value.
( SR:8606249286 CR:JAGae15678 )
Defect Description:
The wrong lock is being held allowing multiple
accesses to a TCP listener list.
Resolution:
Code has been modified to hold the correct lock.
( SR:8606249333 CR:JAGae15725 )
Defect Description:
The ndd tunable "tcp_ignore_path_mtu" was obsoleted in
HP-UX 11.11. ndd still displays the help text for this
tunable.
Resolution:
The ndd help text for "tcp_ignore_path_mtu" has been
removed from the ndd code and the ndd message catalog.
( SR:8606251530 CR:JAGae17595 )
Defect Description:
ndd does not have help text for the tcp_early_conn_ind
parameter.
Resolution:
Help text has been added to ndd describing the usage of
tcp_early_conn_ind.
( SR:8606253673 CR:JAGae18022 )
Defect Description:
The ARP and IP modules can get into an inconsistent
state where IP stream is bound and ARP stream is not
bound.
Resolution:
The ARP and IP modules have been modified not to leave
ARP and IP streams in an inconsistent state.
( SR:8606254542 CR:JAGae18874 )
Defect Description:
RAWIP layer unconditionally clears the "do not fragment"
bit although "do not fragment" bit is set in the header
passed to it.
Resolution:
RAWIP now preserves the value of "do not fragment"
provided by the application.
( SR:8606254713 CR:JAGae19044 )
Defect Description:
The system hangs in a busy loop in the select()
kernel routine because under certain conditions,
it ignores errors returned by the stream head.
Resolution:
Changes have been made to ensure that errors
set in its stream head are handled correctly.
( SR:8606255551 CR:JAGae19868 )
Defect Description:
If the IFF_MULTICAST flag is reset, an infinite
loop happens with holding a spinlock causing a
panic.
Resolution:
The IFF_MULTICAST flag processing is removed
from places where it is not needed, thus
avoiding the infinite loop.
( SR:8606257154 CR:JAGae21460 )
Defect Description:
select() could cause connect() to hang within a
small window of race condition.
Resolution:
The window of race condition has been removed.
( SR:8606257479 CR:JAGae21782 )
Defect Description:
A lock is released even though it has not been
acquired previously.
Resolution:
The lock is not released if it has not been
acquired previously.
( SR:8606259311 CR:JAGae23629 )
Defect Description:
When a 32 bit application is run on a 64 bit
kernel, necessary adjustments for certain fields
in a message header is done twice.
Resolution:
Code has been modified to ensure that the
necessary adjustments for the fields in the
message header is done only once.
( SR:8606259405 CR:JAGae23723 )
Defect Description:
Listen backlog is not checked until the completion of
3-way handshake.
Resolution:
The backlog is now taken into account even when the
ndd variable "tcp_early_conn_ind" is set to zero.
( SR:8606260488 CR:JAGae24811 )
Defect Description:
Appropriate error checking is not done to determine
if the "optlen" argument passed to setsockopt()
is valid.
Resolution:
Code has been modified to return the EINVAL errno
if the "optlen" argument is too large.
( SR:8606260502 CR:JAGae24825 )
Defect Description:
The kernel assumes that an IPv6 source address
is 8 bytes aligned.
Resolution:
The kernel now does not assume that IPv6 source
address in the packets are 8 bytes aligned
on 64-bit machines.
( SR:8606261179 CR:JAGae25501 )
Defect Description:
1. Code to handle reporting of a configured tunnel route
in the ipv6RouteTable searched for the configured tunnel
interface is incorrect.
2. In ipv6AddrTable, the ipv6AddrType of a manually
configured interface should be STATEFUL instead of
UNKNOWN.
3. In ipv6AddrTable, the ipv6AddrStatus of a manually
configured interface should be PREFERRED instead of
UNKNOWN.
4. In ipv6NetToMediaTable, the ipv6IfNetToMediaState
should be REACHABLE instead of UNKNOWN, for configured
local interfaces.
5. In ipv6NetToMediaTable, the
ipv6IfNetToMediaLastUpdated is not properly reported.
6. In ipv6NetToMediaTable, the ipv6NetToMediaType of
entries created from router advertisements should be
labeled as DYNAMIC instead of STATIC.
7. In ipv6NetToMediaTable, the ipv6NetToMediaType of
entries associated with the local (primary) interfaces
should be labeled LOCAL instead of STATIC.
Resolution:
1. Code to handle reporting of a configured tunnel route
in the ipv6RouteTable searched for the configured tunnel
interface has been corrected.
2. In ipv6AddrTable, the ipv6AddrType of a manually
configured interface is now reported as STATEFUL.
3. In ipv6AddrTable, the ipv6AddrStatus of a manually
configured interface is now reported as PREFERRED.
4. In ipv6NetToMediaTable, the ipv6IfNetToMediaState of a
configured local interface is now reported as REACHABLE.
5. In ipv6NetToMediaTable, the ipv6IfNetToMediaLastUpdated
is now properly reported.
6. In ipv6NetToMediaTable, the ipv6NetToMediaType of
entries created from router advertisements is now labeled
as DYNAMIC.
7. In ipv6NetToMediaTable, the ipv6NetToMediaType of
entries associated with the local (primary) interfaces is
now labeled as LOCAL.
( SR:8606261753 CR:JAGae26072 )
Defect Description:
Using TCP_IOC_GET_UCRED for an IPv4-mapped IPv6 address
does not return the expected 4-tuple of the
"remote host/local host/remote port/local port" for an
established connection.
Resolution:
Code has been modified to check if both local and remote
hosts are IPv4-mapped IPv6 addresses. If yes, such
addresses are converted to pure IPv4 addresses.
( SR:8606262294 CR:JAGae26629 )
Defect Description:
The panic happens due to a problem in reassembling the
IP fragments.
Resolution:
IP reassembly code has been corrected to prevent the panic.
( SR:8606262982 CR:JAGae27312 )
Defect Description:
When a socket is closed, the T_UNBIND_REQ
that followed the T_ORDREL_REQ fails, closing
the stream instead of caching it.
Resolution:
TCP module has been modified to ensure that
the T_UNBIND_REQ always succeeds if it follows
a T_ORDREL_REQ.
( SR:8606263157 CR:JAGae27484 )
Defect Description:
Currently, the code does not handle T_OPTMGMT_REQ
primitive with "MGMT_flags" field set to T_NEGOTIATE
and "OPT_length" field set to 0 as specified in TPI
specification.
Resolution:
Code has been modified to return all option values
as mentioned in TPI specification when T_OPTMGMT_REQ
primitive with MGMT_flags set to T_NEGOTIATE and
OPT_length set to 0 is sent down by the application.
( SR:8606264041 CR:JAGae28365 )
Defect Description:
Currently, each SYN+ACK/FIN+ACK without accompanying
data in same segment sent in response to a SYN/FIN
segment respectively is being counted as a duplicate
ACK.
Resolution:
Code has been modified to ensure that SYN+ACK/FIN+ACK
sent without accompanying data is not recognized as
duplicate ACK for a data packet.
( SR:8606266690 CR:JAGae30935 )
Defect Description:
System might crash in TCP due to a race condition
while closing a TCP connection.
Resolution:
Code has been modified to prevent the race condition.
( SR:8606269903 CR:JAGae34142 )
Defect Description:
"ping" prints the round trip time
in microseconds whereas the value needs to be
in milliseconds. Hence, the value printed is 1000
times more than the expected value.
Resolution:
Code has been modified to print the round trip
time in milliseconds.
PHNE_27063:
( SR:8606261174 CR:JAGae25496 )
Symptom:
FTP is unable to put or mput files of any size due to EINVAL
errno 22 "sendfile: Invalid argument" when ftp attempts to
call senfile64() to transmit the files.
Defect Description:
sendfile64() has an invalid 6th argument, being an int flag
hardcoded to 0 (zero) in ftp.c program. The problem is due
to the sendfile function declaration in socket.h
JAGae12001 introduced this problem.
Resolution:
The fix for JAGae12001 which introduced this problem has
been backed out.
PHNE_26762:
( SR:8606181283 CR:JAGad50500 )
Symptom:
The IPv6 loopback address (::1) stops working when
the loopback device (lo0) is reconfigured to a
different address.
Defect Description:
ifconfig allows users to change the IPv6 address
of the loopback device. This address is invariant
and should never be changed from the value ::1.
Resolution:
Users are no longer allowed to change the IPv6 address
of the loopback device from the default value ::1.
( SR:8606185191 CR:JAGad54393 )
Symptom:
System panics with the following stack:
unp_eventreg+0x4
so_eventreg+0x6c
evp_dereg_objhdr+0xa0
evp_close+0x15c
call_open_close+0x1f8
closed+0xb0
spec_close+0x54
vn_close+0x48
vno_close+0x20
closef+0x64
exit+0x324
rexit+0x28
syscall+0x200
$syscallrtn+0x0
Defect Description:
When the close routine is called on the AF_UNIX socket, it
clears the function pointer to the callback function before
using it.
Resolution:
Clearing of the function pointer is delayed until
the callback has been made.
( SR:8606189015 CR:JAGad58231 )
Symptom:
CPU usage is high when a fast producer program is
communicating with a slow consumer program using
UNIX domain sockets.
Defect Description:
The fast producer program allocates mblks equal to the
size of the slow consumer's receive window. The
user's write requests are divided up to fit into these
small mblks that are chained together in a linked list.
The entire linked list is traversed each time an mblk is
appended consuming lots of CPU.
Resolution:
Code has been changed to allocate buffers based on the
size of the write request rather than the receive window.
Additional savings are obtained by using free space in the
previous buffer wherever possible and further reduce the
number of nodes in the list.
( SR:8606198555 CR:JAGad67744 )
Symptom:
Problems compiling recent versions
of the transport source code. These compilation
failures are not visible to customers.
Defect Description:
The enabling of "event port" resulted in conflicting
function declarations between socket and non-socket
sources which prevent successful compilation of the
sockets source code.
Resolution:
Corrections have been made to function declarations
so that socket code no longer conflicts with
event port definitions.
( SR:8606198667 CR:JAGad67856 )
Symptom:
Closing a socket with socket caching turned on and
SO_LINGER set, may cause the kernel to loop indefinitely,
and the connection will be left in an IDLE state.
Defect Description:
While closing a socket which has the linger option set,
the TPI state is not checked for TS_UNBND (corresponds
to TCP state T_IDLE). This can result in endlessly
polling the tcp module.
Resolution:
Code now checks for the TS_UNBND state when processing the
linger option in soclose().
( SR:8606202871 CR:JAGad72045 )
Symptom:
The command "route inet6 delete net ::1 ::1"
deletes the IPv6 loopback route.
After the deletion, the command "ping ::1" fails.
Defect Description:
The IPv6 loopback route is invariant and cannot
be changed or removed.
Resolution:
Deletion of the IPv6 loopback route is disallowed. The
error message "no such route" is displayed.
( SR:8606203612 CR:JAGad72784 )
Symptom:
Code inspection uncovered a potential
system panic with the following stack trace.
soo_select
unp_poll_handler
so_poll_switch
.
.
Defect Description:
An incorrect parameter is being passed when calling an
internal procedure during the handling of select().
Resolution:
Changed the type of the third parameter passed to
the select routine from an integer to a pointer.
( SR:8606208360 CR:JAGad77545 )
Symptom:
The route created by
"route inet6 add net 345::9/64 2002::2002 1"
can be removed using
"route inet6 delete net 345::/64 2002::2002 1"
but cannot be removed with
"route inet6 delete net 345::9/64 2002::2002 1"
Defect Description:
IP accepts the entire 128-bit address in the
add command (ioctl(SIOCADDLRT)), but rejects
a similar specification in the delete command
(ioctl(SIOCDELLRT)).
Resolution:
IP accepts the entire 128-bit address in both
the add command (ioctl(SIOCADDLRT)), and the
the delete command (ioctl(SIOCDELLRT)).
( SR:8606209264 CR:JAGad78451 )
Symptom:
Threaded Async Socket IO can result in blocking close().
Defect Description:
There are conditions with Async sockets where pending IO
can cause a close() to block.
Resolution:
A modification has been made to wakeup pending IO.
( SR:8606233159 CR:JAGae02383 )
Symptom:
The libnm function get_if_table() cannot
handle more than 256 interfaces.
Defect Description:
The get_if_table() call in libnm assumes that
there are not more than 256 interfaces in the
system. With the advent of VLAN, there is
no limit on the maximum number of interfaces.
In addition, this routine has trouble handling IPv6
interfaces.
Resolution:
get_if_table() has been modified to support an arbitrary
number of interfaces. It now fully supports IPv6
interfaces.
( SR:8606233305 CR:JAGae02529 )
Symptom:
Application hangs in the accept() system call.
Defect Description:
A message is being sent to the
wrong thread or process during
accept processing.
Resolution:
A new ioctl for streams has been developed to make the
connection request to TCP atomic.
( SR:8606234349 CR:JAGae03553 )
Symptom:
Route deletion does not work in the following cases:
Case 1: if the destination is fully specified by the
128-bit IPv6 address
route inet6 add net 2222::1/64 2345::2 1
route inet6 delete net 2222::1/64 2345::2 1
Case 2: if the destination is an IPv4 compatible address
route inet6 add net ::1.2.3.4/64 2000::1 1
route inet6 delete net ::1.2.3.4/64 2000::1 1
Case 3: if the destination is a link-local address and
gateway is a local interface
route inet6 add net fe80::1/10 fe80::230:6eff:fe0c:6a91
route inet6 delete net fe80::1/10 fe80::230:6eff:fe0c:6a91
Defect Description:
route delete cannot always find the correct route
to delete.
Resolution:
IP has been fixed so that the route delete command
finds the correct route to delete in the cases described
above. In addition, "route inet6 add net fe80::1/10
<local interface IPv6 addr>" is now rejected, so the case of
deleting a route to a link-local prefix through a
**LOCAL INTERFACE** is no longer possible.
( SR:8606234706 CR:JAGae03897 )
Symptom:
When the command "ifconfig lan0:x inet6" is used to query
the status of a non-existent interface lan0:x, a new
interface lan0:x is created, with a null IPv6 address
"::". This information is displayed as if it existed
prior to the query and it only happens if lan0 has been
configured.
Defect Description:
IP creates a new interface with the null address
if it receives a query on a non-existent secondary
interface.
Resolution:
IP no longer creates a new interface with
the null address if it receives a query on a
non-existent secondary interface.
( SR:8606235627 CR:JAGae04771 )
Symptom:
get_mib_info(ID_ipv6Interfaces) always returns a zero
irrespective of the number of IPv6 interfaces configured
on the machine.
Defect Description:
IP does not correctly update the MIB with
the number of configured IPv6 interfaces.
Resolution:
IP has been modified to update the MIB with
the number of configured IPv6 interfaces.
( SR:8606235632 CR:JAGae04776 )
Symptom:
Unable to delete IPv6 tunnel routes using set_mib_info().
Defect Description:
set_mib_info() does not set the proper flags to indicate
that the specified route is a tunnel route, thus making
it impossible to delete a tunnel route.
Resolution:
set_mib_info() has been modified to set the proper
flags indicating that the route is a tunnel route,
when the gateway has an IPv4-compatible IPv6
address. A user can now use this address to delete
a tunnel route.
( SR:8606236775 CR:JAGae05826 )
Symptom:
SIOCATMARK ioctl returns flag parameter set
after reading out-of-band data when the SO_OOBINLINE
socket option is set.
Defect Description:
The variable that provides the value for the flag
parameter in the SIOCATMARK ioctl is not cleared
after an application reads past the out-of-band
data.
Resolution:
The variable that provides the value for the
flag parameter in the SIOCATMARK ioctl is set
and cleared correctly.
( SR:8606237606 CR:JAGae06646 )
Symptom:
Virtual IP address (lo0:1) cannot be used for IPv6.
Defect Description:
Virtual IP address support is not available for IPv6.
Resolution:
Virtual IP address support is now provided for IPv6.
( SR:8606237642 CR:JAGae06680 )
Symptom:
The address list hash table viewed using
'ndd' does not display addresses created by IPv6
stateless autoconfiguration.
Defect Description:
Addresses created by IPv6 stateless auto-configuration are
not inserted into the address hash list.
Resolution:
IPv6 addresses created by IPv6 stateless address
autoconfiguration are inserted into the address hash list.
( SR:8606238262 CR:JAGae07289 )
Symptom:
Data corruption occurs in the socket structure
during a connect().
Defect Description:
While doing a connect(), fields are changed
in the socket structure without holding a lock.
This can result in data corruption.
Resolution:
The socket structure is locked before
its fields are modified. This prevents
race conditions and the associated data
corruption.
( SR:8606238905 CR:JAGae07926 )
Symptom:
Flow control on UNIX domain datagram sockets
does not work.
Defect Description:
Currently, the space required to write the
data into the socket receive buffer does not take
into account the length of the data passed
while determining the required space for writing
the data in the buffer.
Resolution:
The length of the data is now
taken into account to ensure
that flow control works.
( SR:8606241607 CR:JAGae08865 )
Symptom:
"traceroute" sometimes displays the error:
"traceroute: failed to get received hop limit or packet
info"
Defect Description:
"traceroute" fails to obtain hop limit due to incorrect
buffer length initialization.
Resolution:
An ICMP6 filter has been added to receive only the
interesting ICMP6 types. The buffer length has also
been adjusted after returning from recvmsg().
( SR:8606242226 CR:JAGae09479 )
Symptom:
'netstat -f inet6 -I secondary_interface interval'
displays no output and exits.
Defect Description:
When 'netstat -I <interface> <interval>' is used
in conjunction with IPv6, it only obtains the list of
physical interfaces and ignores all secondary interfaces.
Resolution:
netstat has been modified to get the list of all
interfaces (physical and logical) in IPv6 mode.
( SR:8606242679 CR:JAGae09914 )
Symptom:
getsockopt() on a listen() socket aggravates
the accept() processing.
Resolution:
TCP no longer removes the T_DISCON_IND
message from the stream head when processing the
SO_ERROR socket option from a listen socket.
( SR:8606244728 CR:JAGae11213 )
Symptom:
Users cannot set parameters for rawip6 through the
nddconf configuration file.
Defect Description:
ndd does not recognize the rawip6 device when
processing inputs from nddconf.
Resolution:
"ndd" has been changed to recognize the rawip6 device
when processing inputs from nddconf.
( SR:8606244958 CR:JAGae11439 )
Symptom:
An application gets ECONNRESET when doing a recv()
on a socket which has received RST from the remote
side. The RST arrives while there is pending data on
the socket which is discarded.
Defect Description:
Currently, the data waiting to be read by the
user application will be discarded by TCP and
sockets, when an RST is received from the remote side.
Any future recv() on the socket will return
ECONNRESET.
Resolution:
Data already sitting on the socket is not
discarded when a RST arrives from the peer TCP.
( SR:8606245325 CR:JAGae11797 )
Symptom:
There is a condition where a process termination will block.
Defect Description:
There is a condition where a close() on socket descriptor
will block the process exit.
Resolution:
A modification has been made to avoid to blocking condition.
( SR:8606245531 CR:JAGae12001 )
Symptom:
While building a 32-bit application with large files
support, "sys/socket.h" generates warnings due to
redeclaration of sendfile and sendpath prototypes.
Defect Description:
There are two declarations for sendfile and sendpath
executed under the condition:
#if defined(_APP32_64_BIT_OFF_T).
Resolution:
The first set of sendfile and sendpath declarations is now
compiled only under the condition:
#ifndef _APP32_64BIT_OFF_T
If this symbol is defined, the user gets the second
set of declarations which translate to __sendfile64
and __sendpath64.
( SR:8606245578 CR:JAGae12044 )
Symptom:
Potential panic in sounbind() due to referencing a
null pointer.
Defect Description:
Code does not check for the return of a NULL pointer
when calling a STREAMS memory allocation routine.
Resolution:
Code now checks for the return of a NULL pointer and
returns either ENOBUF or ENOMEM.
( SR:8606245903 CR:JAGae12361 )
Symptom:
SO_LINGER close() does not wait for a FIN's ACK
when the socket_caching_tcp option is on.
Defect Description:
TCP returns wrong state if waiting for the FIN's ACK.
Resolution:
TCP now returns the correct state when waiting for
the FIN's ACK.
( SR:8606245905 CR:JAGae12363 )
Symptom:
Invalid pointer access in stream head read queue
when processing expedited data during recv().
Defect Description:
An access is being made to the stream head read
queue without holding the lock on the stream head.
This can result in dangling references to freed
memory.
Resolution:
The lock is acquired on the stream head before
using the stream head read queue.
( SR:8606245906 CR:JAGae12364 )
Symptom:
When a system is completely out of memory, there is a
possibility of a NULL pointer access when processing
an accept() system call.
Defect Description:
A NULL pointer is returned when processing an accept,
if memory allocation of a TPI connection response fails.
Resolution:
When memory allocation fails while creating a
TPI connection response, cleanup for accept
processing is completed and an ENOMEM error
is returned.
( SR:8606245919 CR:JAGae12377 )
Symptom:
No noticeable symptom to the end user.
Defect Description:
A logical OR operation is being performed when
actually a bitwise OR was intended.
Resolution:
Code has been modified to perform a bitwise OR instead of
a logical OR operation.
( SR:8606245952 CR:JAGae12410 )
Symptom:
Simultaneous calls to accept made on multiple threads
may hang and never return.
Defect Description:
Wakeups are missed causing accept to sleep forever.
Resolution:
accept() has been modified to prevent wakeups
from being lost.
( SR:8606245971 CR:JAGae12429 )
Symptom:
ndd fails to set some tunables in the
/etc/rc.config.d/nddconf file.
Defect Description:
When the entries in the
/etc/rc.config.d/nddconf file have indices
which are not sequential, "ndd -c"
silently ignores all of the tunables after
the missing index.
Resolution:
ndd now processes all valid tunables listed
in the /etc/rc.config.d/nddconf file
even if the indices are not sequential.
However, incomplete or invalid entries
will not be processed. In such cases,
ndd displays an appropriate error message
( SR:8606246169 CR:JAGae12617 )
Symptom:
socket() call returns EINVAL when udp_recv_hiwater_max
is less than 65535.
Defect Description:
This is the result of conflicting ndd parameter
settings. The so_udp_rcvbuf_default parameter
determines the default size of SO_RCVBUF. The
udp_recvbuf_default places an upper bound on the
size of SO_RCVBUF. If the default value exceeds the
maximum value, socket() returns EINVAL. This is
the correct behavior.
Resolution:
The above behavior is now documented in
the online ndd help facility.
( SR:8606246301 CR:JAGae12746 )
Symptom:
1. Memory leak occurs when there are tunnel
errors for a tunnel that has broadcast
forwarding set.
2. System panic in ip_wput_frag() occurs when
memory is freed twice during outbound IP fragmentation
and when there are tunnel errors.
Defect Description:
1. Memory is not released in the case
of tunnel errors for a tunnel that has
broadcast forwarding set.
2. Memory is released twice while handling
IP fragmentation in the case of tunnel errors.
Resolution:
Code has been modified to:
1. Release memory in the case
of tunnel errors for a tunnel that has
broadcast forward set.
2. Avoid freeing memory twice while handling
IP fragmentation in case of tunnel errors.
( SR:8606246309 CR:JAGae12754 )
Symptom:
System leaks memory in 512-byte blocks.
Defect Description:
While doing a listen, a STREAMS data structure
is not freed when an internal routine returns an error,
resulting in a memory leak.
Resolution:
The data structure is now freed whenever any of the
routines called while doing a listen return an error.
( SR:8606246321 CR:JAGae12766 )
Symptom:
System leaks memory.
Defect Description:
If the closure of the stream fails due to a transient error
when socket caching releases a socket, the memory associated
with that stream never gets freed.
Resolution:
The memory associated with the stream which failed to
close is returned to the socket cache to be freed later.
( SR:8606246398 CR:JAGae12843 )
Symptom:
When socket caching is turned ON, an unnecessary FIN
is sent before the required RST when closing a connection
which has the SO_LINGER option turned on with an l_linger
value of 0.
Defect Description:
When a socket application performs a close on a connection
with SO_LINGER, and l_linger set to 0, an RST is sent if
socket caching is OFF. If socket caching is turned ON, then
an unnecessary FIN is sent followed by an RST.
Resolution:
The unnecessary FIN is no longer sent when socket caching
is turned on and a connection closes with SO_LINGER and
l_linger set to 0.
( SR:8606246561 CR:JAGae13006 )
Symptom:
ACKs are sent in response to old segments which
have wrapped sequence numbers.
Defect Description:
There are uninitialized variables in the path which checks
for wrapped sequence numbers.
Resolution:
The uninitialized variables in the path which checks for
wrapped sequence numbers are set to appropriate values.
( SR:8606246676 CR:JAGae13119 )
Symptom:
System may leak memory when the ndd parameter
socket_caching_tcp is turned off.
Defect Description:
When socket_caching_tcp is turned off (set to zero using
ndd), the cache may not be completely cleared. The
remaining cached endpoints will continue to exist in memory
and will not be accessed unless socket_caching_tcp is
re-enabled.
Resolution:
A new internal kernel function has been added which
does a more thorough job of emptying the cache and
freeing the associated data structures.
( SR:8606247339 CR:JAGae13779 )
Symptom:
Applications receive unexpected ICMPv6 error messages.
Defect Description:
BIND 9.0 running on the IPv6 stack fails on receiving a
"Destination Unreachable" message from ICMPv6.
Resolution:
The default value for the ndd tunable "udp_pass_up_icmp6"
has been changed to 2. This ensures that the ICMPv6 errors
are sent only on connected UDP sockets.
( SR:8606247484 CR:JAGae13921 )
Symptom:
When using the SO_LINGER option on a UNIX95 socket,
specifying certain large values of the linger timeout
may result in incorrect behavior, or performance
degradation.
Defect Description:
The code handling the close() for sockets does not correctly
compute whether the specified waiting time has elapsed.
Resolution:
Elapsed time computation in socket close() routine
has been changed to deal correctly with arithmetic
overflow.
( SR:8606247490 CR:JAGae13927 )
Symptom:
Socket state inconsistency when closing a
socket in a multithreaded application.
Defect Description:
Certain code paths of socket "close" code fail
to hold a required lock when manipulating
socket state flags.
Resolution:
Socket close code has been modified to simplify
socket state manipulation and to correctly
hold a socket lock when updating socket state.
( SR:8606247508 CR:JAGae13945 )
Symptom:
Socket Caching can cause a kernel memory problem.
Defect Description:
An error condition in a loop to flush old messages
is not being handled correctly.
Resolution:
The handling of error conditions which terminate the
loop has been modified to ensure that the lock used in
the loop is in the proper state and that clean-up is
complete.
( SR:8606247520 CR:JAGae13957 )
Symptom:
No visible symptoms.
Defect Description:
Unnecessary socket state flag manipulation.
Resolution:
All references to this flag have been removed from the
code.
( SR:8606248122 CR:JAGae14522 )
Symptom:
The "ifconfig lanx:y" command which is supposed to
display the current configuration of lanx:y, will create
a NULL entry for lanx:y if that interface does not exist.
Defect Description:
If the interface does not exist, the ifconfig interface
command creates and then displays a NULL entry.
Resolution:
"ifconfig" checks for the existence of the interface
configuration entry when asked to display current
configuration parameters. If not present, an error
message is displayed.
( SR:8606248406 CR:JAGae14806 )
Symptom:
A blocking condition with a socket call can occur with
multi-threaded AF_UNIX applications.
Defect Description:
A thread is left asleep.
Resolution:
A modification has been made to the
lock management code.
( SR:8606248669 CR:JAGae15067 )
Symptom:
In a multi-processor environment, the "named" daemon
is not killed when it receives the SIGTERM signal.
Socket recvmsg() is blocking despite O_NONBLOCK.
Defect Description:
When the O_NDELAY or O_NONBLOCK flags are set
using fcntl(), the socket receive functions do
not completely implement the non-blocking feature.
There is a race condition for a multithreaded
application which may cause a receive to hang.
Resolution:
In soreceive(), calls to the kernel Streams functions
will now include the O_NDELAY and O_NONBLOCK flags
in setting the blocking argument.
( SR:8606249185 CR:JAGae15579 )
Symptom:
Minor performance loss in UDP recv system calls.
Defect Description:
In the non-connected UDP case, a slower path
is taken through the kernel receive code due
to a coding error.
Resolution:
The logic has been corrected so that the slower
path will be used only in the connected UDP case.
( SR:8606250322 CR:JAGae16697 )
Symptom:
AF_UNIX domain datagram type sockets applications
fail on recv() call.
Defect Description:
There is an unnecessary compression of header
and data mblks for protocols which split their
header and data into separate mblks. This causes
recv() to fail.
Resolution:
Code has been changed to limit the fix
of JAGad58231 to AF_UNIX domain and
stream oriented sockets, thus avoiding
unnecessary compression of datagram type
sockets data.
( SR:8606250432 CR:JAGae16807 )
Symptom:
When socket caching is turned on, a close() performed
on a socket with SO_LINGER returns before the FIN-ACK
has been received.
Defect Description:
SO_LINGER close() does not wait for a FIN's Ack if the
FIN segment has already been sent by shutdown(fd, SHUT_WR).
Resolution:
soclose() now waits for the FIN's ack when
a FIN segment is sent by shutdown(fd, SHUT_WR)
prior to calling close(2) with SO_LINGER set.
( SR:8606250573 CR:JAGae16948 )
Symptom:
"ndd -get /dev/tcp tcp_status" might fail to reset
the semaphore left by another "ndd" due to a premature
exit.
Defect Description:
A call to semop(2) is called with an incorrect parameter.
Resolution:
The call to semop(2) now passes in the correct
parameter value.
( SR:8606253203 CR:JAGae18008 )
Symptom:
System is showing excessive memory pressure.
Defect Description:
Memory leak in M_SOCKET arena when using AF_UNIX
SOCK_STREAM sockets.
Resolution:
A race condition existed between solockpair() and
other functions like select() and sosend(), which
is avoided by ordering socket lock accesses.
( SR:8606254024 CR:JAGae18356 )
Symptom:
Configuring IPv6 addresses on a fddi device does not work.
Defect Description:
Configuring IPv6 addresses on a fddi device is disallowed.
Resolution:
The ifconfig command now allows configuring IPv6 addresses
on a fddi device.
( SR:8606254214 CR:JAGae18546 )
Symptom:
"msg_controllen" can contain an invalid number
when returning from recvmsg().
Defect Description:
The msg_controllen field of the msghdr is set
incorrectly if its initial value is zero.
Resolution:
The msg_contrllen field of the msghdr is now
set correctly when its initial value is zero.
( SR:8606255696 CR:JAGae20013 )
Symptom:
A system panic with the following stack trace may occur
while opening an AF_UNIX socket.
socket_merged+0xb4
socket+0x14
syscall+0x204
Defect Description:
A routine used in the creation of a socket
returns success even if the allocation of memory needed
for the new socket has failed.
Resolution:
The routine used in the creation of a socket now
returns the correct errno and cleans up
the socket if memory allocation fails.
( SR:8606256314 CR:JAGae20629 )
Symptom:
recv(2) returns EWOULDBLOCK after poll(2) detects POLLIN.
Defect Description:
This problem only occurs when a TCP connection is
terminated by a RST from the remote side.
Resolution:
Changes are made in sockets to ensure that
POLLIN returns correctly for poll(2).
( SR:8606256741 CR:JAGae21056 )
Symptom:
The socket debug macros are called with an
insufficient number of parameters which causes
compilation warnings.
Defect Description:
The socket debug macros which are used for
logging information are called with an insufficient
number of arguments.
Resolution:
The additional argument has been added to all
calls of the socket debug macros.
( SR:8606257716 CR:JAGae22019 )
Symptom:
On systems where the directories /usr/contrib/man or
/usr/contrib/man/man1.Z do not preexist, the installation of
PHNE_25644 will display warnings indicating that
these directories do not have the correct permissions.
Defect Description:
When /usr/contrib/man or /usr/contrib/man/man1.Z do not
exist, swinstall creates them with default permissions which
are incorrect.
Resolution:
The preinstall script has been modified to ensure that
both /usr/contrib/man and /usr/contrib/man/man1.Z exist
with the correct permissions.
PHNE_25644:
( SR:8606113716 CR:JAGac07230 )
Symptom:
"ndd -c" fails when /etc/rc.config.d/nddconf contains
more than 10 tunable entries.
Defect Description:
The incorrect ascii to integer conversion of array indices
specified in /etc/rc.config.d/nddconf causes this failure.
Resolution:
ndd, now properly converts ascii numbers to integers.
( SR:8606129381 CR:JAGac86927 )
Symptom:
"ndd -c", does not support arp tunables in
/etc/rc.config.d/nddconf.
Defect Description:
/dev/arp was excluded from the list of valid tunable
devices in /etc/rc.config.d/nddconf.
Resolution:
"ndd -c" now supports arp tunable parameters.
( SR:8606138758 CR:JAGad08005 )
Symptom:
If socket caching is turned on, a connect made on
a socket which has SO_REUSEADDR set returns EADDRINUSE.
Defect Description:
The value of some tcp variables are preserved when
tcp sockets are cached. Depending on their value, these
uninitialized variables can result in abnormal behavior.
Resolution:
After acquiring a cached tcp socket, the socket code now
sends a request to tcp to reset these variables to their
default values.
( SR:8606146239 CR:JAGad15575 )
Symptom:
Intermittent hangs in close() when
the so_linger option is on.
Defect Description:
There exists a corner case to the solution
for JAGad09415 where close() can still hang.
Resolution:
The timing window where these hangs occur has
been completely closed.
( SR:8606206785 CR:JAGad75958 )
Symptom:
ifconfig fails and dumps core when numerous
interfaces (such as vlan) are created on the
system.
Defect Description:
ifconfig sets the maximum number of interfaces
on the system to 256. ifconfig then allocates
static memory which can hold information for no more
than 256 PPAs. With the advent of vlan, it is easy to
exceed this limit.
Resolution:
PPAs are now stored in dynamically allocated
memory.
( SR:8606216047 CR:JAGad85218 )
Symptom:
"ndd -c" fails to handle very long tunable names
referenced in /etc/rc.config.d/nddconf.
Defect Description:
"ndd -c" is unable to parse long ndd tunables read
from /etc/rc.config.d/nddconf.
Resolution:
Text parsing is changed to read tunables names
correctly.
( SR:8606217313 CR:JAGad86465 )
Symptom:
System hangs, requiring a TOC
Defect Description:
The variable that stores the cpu context when
a spinlock is acquired was being set to zero
resulting in a system hang
Resolution:
The cpu context is preserved and not set
to zero
( SR:8606220082 CR:JAGad89223 )
Symptom:
netstat fails and prints the following error
message when there are numerous interfaces
(such as vlan) configured on the system.
"ppa_area (125000) too small (expected 169260)
can't get ppa list"
Defect Description:
netstat allocates a fixed amount of memory to
hold the ppa list returned from DLPI. While this
has been sufficient in the past, with the presence of
VLAN a fixed limit is easily exceeded.
Resolution:
The ppa_area is now allocated dynamically.
( SR:8606220677 CR:JAGad89814 )
Symptom:
When packets get lost in a network, it takes a
long time (60 seconds or longer) to retransmit.
Defect Description:
There is a problem in the Smoothed Round Trip Time
algorithm.
Resolution:
Corrected the initialization of the Round Trip Time
variable so that the estimated RTO does not become too
large.
( SR:8606221602 CR:JAGad90736 )
Symptom:
read() sometimes loses data and returns 0.
Defect Description:
Under certain circumstances a T_ORDREL_IND message is
passed upstream by TCP before sending any pending
eager data.
Resolution:
The fix takes care of eager data to be sent upstream.
( SR:8606221683 CR:JAGad90817 )
Symptom:
"ndd -c" fails in some abnormal cases.
Defect Description:
"ndd -c" fails in some abnormal cases.
Resolution:
The code causing the failure has been changed
to work properly.
( SR:8606223770 CR:JAGad92866 )
Symptom:
"ndd -get" does not display anything for socket
tunable parameters.
Defect Description:
ndd sends a request to the kernel to get the value for
the parameter and the kernel replies with its value.
Though ndd gets the value it does not print it.
Resolution:
"ndd -get" now prints the value sent by the kernel.
( SR:8606224560 CR:JAGad93648 )
Symptom:
FTP hangs when transferring files from PC DOS to HP-UX.
Defect Description:
TCP ignores a FIN that arrives with old data (during
retransmission). This causes the TCP connection to hang
when FTP is copying files from PC DOS (NetManage PCTCP
stack) to HPUX.
Resolution:
TCP no longer ignores a FIN which arrives with
retransmitted data.
( SR:8606224838 CR:JAGad93926 )
Symptom:
If nddconf contains any IPv6 entries,
"ndd -c" returns the following error:
"invalid tunable name, ip6_nd_transmit_interval for this
module, for a list of valid names, use name '?'".
This error can be obtained by adding the following
entries to /etc/rc.config.d/nddconf.
TRANSPORT_NAME[0]=ip6
NDD_NAME[0]="ip6_nd_transmit_interval"
NDD_VALUE[0]=4000
Defect Description:
ndd does not recognize and process IPv6 tunables in
nddconf.
Resolution:
Code is added to handle IPv6 tunables
in the nddconf file.
( SR:8606225148 CR:JAGad94236 )
Symptom:
"ndd -c" does not work if the NDD_VALUE for
a tunable in /etc/rc.config.d/nddconf is a string
containing non-numeric characters. For example:
NDD_VALUE[0]="lan0 ON"
is not recognized.
Defect Description:
NDD_VALUE was treated as an integer. So ndd -c
fails if NDD_VALUE is a string containing
non-numeric characters.
Resolution:
"ndd -c" now works properly even if NDD_VALUE is
a string containing non-numeric characters.
( SR:8606225238 CR:JAGad94326 )
Symptom:
A system panic can occur when "ndd -get /dev/udp udp_status"
is run.
Defect Description:
Invalid formatting of a connection which is closing
causes a reference to an invalid pointer.
Resolution:
The output is now correctly formatted.
( SR:8606225324 CR:JAGad94412 )
Symptom:
Connections sometimes hang after a ServiceGuard
switchover.
Defect Description:
An unintentional routing entry with a default
route is created during switchover and left
in the routing table. This entry causes packets
to be undeliverable.
Resolution:
The default routing entries are no longer created
during switchover.
( SR:8606225422 CR:JAGad94509 )
Symptom:
If a user tries to assign a primary IPv6
address which is already assigned to another
interface on the machine, the kernel returns
success to the user even if the operation failed.
Defect Description:
On a multihomed machine, when a user tries to assign
a primary IPv6 address which is already assigned to
another interface on the machine, the ioctl for setting
an interface up or down returns success even though
the operation failed.
Resolution:
The kernel now returns the correct error
if an address has already been assigned as primary
or secondary to another interface.
( SR:8606227936 CR:JAGad96993 )
Symptom:
The return value of "ifconfig lanX inet" command when
the interface lanX is not configured with an IPv4
address is "0" instead of the expected value "1".
This happens when lanX has been configured with
only an IPv6 address.
Defect Description:
When an interface is configured with at least one IPv6
address, but the same interface is not configured with an
IPv4 address, the ifconfig command which shows only the IPv4
configuration "ifconfig lanX inet" should return "1", to
signal that no such interface was found. Instead, it
returns "0".
Resolution:
When the command "ifconfig lanX inet" is used to examine
the IPv4 configuration of lanX, the command now returns "1"
if no IPv4 addresses are configured.
( SR:8606228250 CR:JAGad97307 )
Symptom:
System panics when under heavy memory pressure.
Defect Description:
Under heavy memory pressure situations a pointer is
freed twice.
Resolution:
A check is now made in advance to make sure
that there is enough room for a lower queue
to handle the entire fragment train before
fragmenting starts. This eliminates the need for
the code which was doing the double free.
( SR:8606229080 CR:JAGad98134 )
Symptom:
1. tcp_fin_wait_2_timeout does not work when the
ndd tunable 'socket_caching_tcp' is set to 1.
2. bind() fails with EINVAL unexpectedly when
'socket_caching_tcp' is set to 1.
Defect Description:
1. tcp_fin_wait_2_timeout did not happen because
the cached tcp was in a wrong queue.
2. Sockets mishandles the caching of a tcp stream.
This results in the cached tcp instance being
in the wrong state which results in the return of
EINVAL when sockets tries to use the cached tcp
instance.
Resolution:
When a tcp socket is closed and socket caching
is on, after an orderly release request,
the tcp connection is moved to the default queue.
( SR:8606229279 CR:JAGad98332 )
Symptom:
System panic:
tcp_icmp_error+0x38
tcp_rput_other+0x518
tcp_rput+0x58
csq_turnover_with_lock+0x84
str_spu_sw_isr+0x654
sw_service+0xb0
mp_ext_interrupt+0x150
ivti_patch_to_nop3+0x0
idle+0x104
Defect Description:
A pointer associated with a packet which has been enqueued
for delayed processing may become invalid before the delayed
processing can commence.
Resolution:
TCP now checks the tcp instance associated with the packet
is NULL before processing it in tcp_icmp_error().
( SR:8606230067 CR:JAGad99118 )
Symptom:
When there are a large number of TCP
connections, "ndd -get /dev/tcp tcp_status" does not
list all the connections.
Defect Description:
The function reporting the status of TCP
connections truncates the linked list which
contains the connection status.
Resolution:
ndd no longer truncates the linked list
containing the status of TCP connections.
( SR:8606230303 CR:JAGad99355 )
Symptom:
The VLAN IP Precedence feature is not supported.
Defect Description:
The IP Precedence feature of VLAN (Virtual LAN)
requires changes to transport. This is an enhancement.
Resolution:
IP has been modified to support the
IP Precedence feature of VLAN.
( SR:8606231247 CR:JAGae00485 )
Symptom:
System panics with the following stack trace:
panic+0x14
report_trap_or_int_and_panic+0x80
trap+0xdb8
nokgdb+0x8
tcp_get_ucred+0x1d4
tcp_wput_ioctl+0x160
tcp_wput+0x918
putnext+0xcc
wait_iocack+0x68
str_istr_ioctl+0x72c
hpstreams_ioctl_int+0x370
hpstreams_ioctl+0x50
spec_ioctl+0xac
vno_ioctl+0x90
ioctl+0x78
syscall+0x6f8
$syscallrtn+0x0
Defect Description:
Sometimes when a request is made for user credentials
over a connection which is going away, an attempt
is made to access the tcp context after it has been
removed.
Resolution:
TCP now checks to make sure that the tcp instance still
exists before attempting to access it.
( SR:8606231253 CR:JAGae00491 )
Symptom:
When using "ndd -set" to set sockets tunables, the
new value is echoed onto the screen.
Defect Description:
The echoing behavior is inconsistent with other
classes of tunables. The ndd -get functionality
is always run for sockets regardless of the getting
or setting of ndd tunables.
Resolution:
The ndd -get routines are no longer called when
processing ndd -set operation.
( SR:8606231526 CR:JAGae00764 )
Symptom:
System panics with the following stacks:
tcp_wput+0x58
csq_turnover_with_lock+0x84
str_spu_sw_isr+0x654
or
tcp_wput+0x58
putnext+0xcc
sth_wsrv+0x204
sq_wrapper+0x94
str_sched_up_daemon+0x1c4
str_sched_daemon+0x1a4
main+0x854
Defect Description:
There is a race between the closing of a stream and
data being put onto the stream's synchronization queue.
Resolution:
A test for a null pointer is performed
when entering tcp_wput().
( SR:8606231670 CR:JAGae00908 )
Symptom:
ndd does not display more than 16KB of output.
Defect Description:
The static output buffer ndd uses to retrieve
data from the kernel is not large enough.
Resolution:
ndd dynamically allocates the output buffer based on
the size of the data fetched from the
kernel.
( SR:8606231951 CR:JAGae01187 )
Symptom:
System panics with the following
stack trace:
tcp_rsrv_comm+0x18
tcp_rput+0x3620
csq_turnover_with_lock+0x7c
str_spu_sw_isr+0x5f8
sw_service+0x8c
Defect Description:
An incorrect goto statement is executed
in an error path.
Resolution:
Changed error path code to exit the routine
using the correct goto statement.
( SR:8606232014 CR:JAGae01250 )
Symptom:
get_if_tbl fails with ENODEV as of PHNE_25642.
Defect Description:
get_if_tbl prematurely terminates due to ENODEV while
processing non-existent IPv6 entries on a system where
IPv6 has been uninstalled.
Resolution:
get_if_tbl is corrected to ignore ENODEV errors that
are possible after processing non-existent IPv6
entries on a system where IPv6 has been uninstalled.
( SR:8606232185 CR:JAGae01421 )
Symptom:
System panics with following stack trace:
tcp_rsrv_comm+0x18
tcp_rsrv+0x10
sq_wrapper+0x90
str_sched_mp_daemon+0x130
str_sched_daemon+0x2dc
main+0xa9c
$vstart+0x34
$locore+0x90
Defect Description:
Service routine is allowed to run on a closing queue.
Resolution:
A test for a null pointer is performed as the service
routine is entered.
( SR:8606233090 CR:JAGae02314 )
Symptom:
In some cases urgent data gets retransmitted
unnecessarily impairing performance.
Defect Description:
Packet validation inside tcp incorrectly drops
a valid packet containing urgent data.
Resolution:
Packet validation logic has been modified to
recognize valid urgent data packets.
( SR:8606233338 CR:JAGae02562 )
Symptom:
The route command adds incorrect routes in some cases.
For example, "route inet6 add 2000: 4321::210:83ff:fef7:1 1"
adds the route
2000:0:0:4321::/64 4321::210:83ff:fef7:1 UG 0 lan1 0
instead of the expected route
2000::/64 4321::210:83ff:fef7:1 UG 0 lan1 0
Defect Description:
The destination address is incorrectly set to
include part of the gateway address.
Resolution:
The parsing of the destination address is fixed so
that it terminates at the end of the destination
address string.
( SR:8606233710 CR:JAGae02935 )
Symptom:
"netstat -p tcp" reports a high number of
"bad TCP segments dropped due to state change"
and consequently a high number of "retransmit
timeouts."
Defect Description:
The last ACK can come in at an unexpected
moment. The packet gets discarded and the
system waits for the retransmission.
Resolution:
The last ACK is saved for later processing
instead of simply dropping it.
( SR:8606208110 CR:JAGad77288 )
Symptom:
Problem occurs when multiple instances of
ndd -get /dev/tcp tcp_status is invoked
while there are huge numbers of tcp connections
on the system.
Defect Description:
tcp_status is obtained internally by reporting
the information in the tcp port array list. However
contention occurs when multiple instances try
to access the port array list.
Resolution:
A semaphore is used to serialize the
execution of ndd when obtaining a
tcp_status/udp_status report.
NOTE: Consequently, "ndd -get /dev/tcp tcp_status"
may seem to hang when multiple instances of ndd
are trying to get tcp_status/udp_status.
( SR:8606210971 CR:JAGad80159 )
Symptom:
Using the UNIX95 version of the recvmsg() system call,
an application compiled as 64-bit may not receive
ancillary data properly.
Defect Description:
The calculation of user memory available for
copying the ancillary data is incorrect.
Resolution:
The calculation of available user space for
copying ancillary data has been corrected.
( SR:8606213513 CR:JAGad82705 )
Symptom:
Systems relying on random increments for
choosing less predictable TCP ISN values,
are still vulnerable to statistical attacks.
Defect Description:
The RFC 1948 ("Defending against sequence
number attacks") is not supported.
Resolution:
The RFC 1948 is now implemented for computing
TCP ISN values. By default, the support for
RFC 1948 is turned off. It can be turned on by
using the ndd variable, tcp_isn_passphrase
<secret passphrase>. The secret passphrase can be
of any length, but only the first 32 characters
will be retained. The passphrase, once set, should
not be changed, except possibly at reboot.
For example:
ndd -set /dev/tcp tcp_isn_passphrase "rfc 1948"
will turn on the support for RFC 1948.
( SR:8606213579 CR:JAGad82771 )
Symptom:
"netstat" does not resolve those network
addresses in /etc/networks, which do not
have all four octets specified explicitly.
For example, the following entry will not be
resolved:
127 loopback
It gets resolved only when specified as follows:
127.0.0.0 loopback
Defect Description:
"netstat" resolves only those network
addresses in /etc/networks which have
all four octets specified.
Resolution:
"netstat" can now resolve the network
addresses in /etc/networks even if the
trailing zero octets are not explicitly specified.
( SR:8606223042 CR:JAGad92146 )
Symptom:
/usr/include/netinet/tcpip.h is missing in HP-UX 11.11.
Defect Description:
/usr/include/netinet/tcpip.h was removed from the
installation list and added to the obsolete list
as of HP-UX 11.10.
Resolution:
This file entry has been restored to the installation
list and removed from the obsolete list.
( SR:8606226976 CR:JAGad96038 )
Symptom:
Multicast testing leads to system hang.
Defect Description:
Multicast testing causes a memory leak which
leads to a system hang. Kernel memory was
not getting freed when encountering an error.
Resolution:
The memory allocated by kernel is freed when
encountering an error.
( SR:8606227536 CR:JAGad96591 )
Symptom:
The help text for the ndd variable, "tcp_isn_passphrase"
is missing.
Defect Description:
No help text for ndd variable, "tcp_isn_passphrase".
Resolution:
Added help text for ndd variable, "tcp_isn_passphrase"
The following command displays the help text on how to
use the ndd variable, "tcp_isn_passphrase":
#ndd -h tcp_isn_passphrase
( SR:8606227920 CR:JAGad96977 )
Symptom:
"netstat -i" displays incorrect values (always 0) for
Ierrs, Oerrs, Coll in the case of snap interfaces.
Defect Description:
The values of Inbound errors(Ierrs),
Outbound errors(Oerrs) and
Collisions(Coll) are not updated for snap
interfaces. Hence, "netstat -i" always
displays 0.
Resolution:
"netstat -i" now displays correct statistics
for snap interfaces.
( SR:8606229650 CR:JAGad98702 )
Symptom:
System panics with the following stack trace:
igmp_timeout_handler+0x160
ip_rtimer+0x100
ip_rput+0x408
puthere+0x148
mi_timeout_exec+0x288
sw_service+0xb0
mp_ext_interrupt+0x150
ihandler+0x904
idle+0xe24
swidle+0x20
Defect Description:
A race condition exists when a data structure
used in the processing of multicast packets is
freed prematurely. This occurs when one thread
is trying to free the structure and another
thread is trying to handle igmp timeout on the
same structure. A DPF (Data Page Fault) panic
occurs while trying to handle timeout for the
structure which is on the igmp timeout list.
Resolution:
Code has been modified to replace the deferred free
callback routine with a new routine which makes
sure that the bogus(freed) structure is not
on the igmp timeout list.
( SR:8606232100 CR:JAGae01336 )
Symptom:
When ndd is invoked from the command line
with an invalid tunable, it fails without
setting the exit value to one.
Defect Description:
When ndd is passed an invalid tunable,
it exits with an error message. But the
exit status is not set to one as stated in
the man page.
Resolution:
Code has been modified so that ndd
exits with a value of 1 in case of
failure.
( SR:8606233164 CR:JAGae02387 )
Symptom:
connect() call for the SOCK_STREAMS socket returns
the ETIMEDOUT error. This occurs when another
connection already bound to the same address
(with 4-tuple being the same) has received RST from
the remote side, and has not been closed by the
application.
Defect Description:
In an established connection, when a SOCK_STREAMS
socket receives a RST from the remote side while the
application has still not closed the socket, and if
another application bound to the same local address
tries to connect, the IP layer returns the ETIMEDOUT
error.
Resolution:
Code has been modified to handle the stale connection
entries maintained by IP.
( SR:8606237497 CR:JAGae06537 )
Symptom:
System panic:
panic+0x6c
wait_for_lock+0x344
slu_retry+0x1c
mp_socket_lock+0x2c
soclose+0x1d4
soo_close+0x90
closef+0x64
close+0x98
syscall+0x204
$syscallrtn+0x0
Defect Description:
During a socket close with the socket caching option
turned on, SOCKETS attempts to acquire a lock that it
already owns.
Resolution:
The condition which results in attempting a second
acquisition of the lock is avoided.
( SR:8606238295 CR:JAGae07322 )
Symptom:
If ndd -set /dev/tcp tcp_debug is set to 1
or higher and strace is used you will get
the following messages to your screen:
-------------
11348 13:12:25 17614232 1 ... 0 0 [HATCP][-9195]
tcp:0x61dd28a8 tcp_ha_state:0x40 tcp_state: 0 tcp
HA state changed to HA_TCP_CONN_ACTIVE
-------------
This happens even though HATCP is not enabled
on the system.
Defect Description:
HATCP logging information is being logged even
though HATCP is not enabled in the system.
Resolution:
All the HATCP logging messages are being logged
if and only if HATCP is enabled.
( SR:8606230164 CR:JAGad99215 )
Symptom:
Performance degradation after installing PHNE_23456.
Defect Description:
A change in zero window probing causes performance
degradation.
Resolution:
Correct zero window probing mechanism to
obtain better TCP performance.
( SR:8606244219 CR:JAGae10710 )
Symptom:
System Memory Leak out of the 512 byte
memory arena.
Defect Description:
When socket_caching_tcp is turned on,
a large number of 384-byte mblks are not freed.
Resolution:
Changes are made in sockets to ensure that
memory used for tcp status is freed in the
error case.
PHNE_25642:
( SR:8606229037 CR:JAGad98091 )
Symptom:
ndd tunable ip_enable_udp_bcastrecv has no
effect.
Defect Description:
ip_enable_udp_bcastrecv was accidentally mapped
to ip_tunnel_local_broadcasts and vice versa.
Resolution:
Remap the ndd entries so that they match.
PHNE_25134:
( SR:8606227191 CR:JAGad96253 )
Symptom:
Panic during
ipsec_admin -stop
ipc_hash_remove+0x1c0
ip_close+0x1e4
close_wrapper+0x38
csq_protect+0x11c
osr_pop_subr+0x22c
osr_close_subr+0x2fc
hpstreams_close_int+0x2f8
streams_close+0x14
soclose+0x42c
soo_close+0x90
closef+0x64
close+0x58
syscall+0x204
Defect Description:
Fields were removed from transport data structures
which are still used by IPSec.
Resolution:
The fields required by IPSec have been restored.
PHNE_24506:
( SR:1653262485 CR:JAGaa27163 )
Symptom:
If /etc/resolv.conf points to an invalid
DNS server, netstat seems to hang when
resolving network or host addresses to names.
Defect Description:
While trying to resolve the IP addresses
to names, netstat uses hosts lookup
followed by network lookup irrespective of
the address being that of a host or a network.
This can cause netstat to hang for some
DNS configurations.
Resolution:
The netstat command has been modified to
distinguish between network and
host addresses while trying to
resolve the names associated
with IP addresses.
Caution: A hang is still possible
if the DNS server is non-existent
and DNS is directly queried to
resolve host addresses without
looking into /etc/hosts.
( SR:8606137536 CR:JAGad06654 )
Symptom:
A trace from tcpdump showed that sendfile sends trailer
buffers separately.
Defect Description:
Socket sendfile() always sends the trailer and header
separately.
Resolution:
If the total size of the data buffer and
trailer buffer is smaller than one
MSS, send the header and trailer together.
( SR:8606162623 CR:JAGad31939 )
Symptom:
Service Guard system TOC with the following trace:
x_mi_mpprintf_putc+0x1c
x_mi_iprintf+0x280
x_mi_mpprintf+0x64
tcp_status_report+0x188
x_nd_getset+0x148
tcp_wput_ioctl+0x9c
tcp_wput+0x82c
putnext+0xcc
wait_iocack+0x68 s
tr_istr_ioctl+0x738
hpstreams_ioctl_int+0x6a8
hpstreams_ioctl+0x50
spec_ioctl+0xac
vno_ioctl+0x90
ioctl+0x7c
syscall+0x480
$syscallrtn+0x0
Defect Description:
TCP takes a long time to complete a TCP_STATUS request
ndd.
Resolution:
identd uses a newly created TCP interface
which provides a faster method for completing
the kinds of requests made by identd.
( SR:8606165144 CR:JAGad34440 )
Symptom:
The TTL field in UDP Broadcast packets defaults to
1. This is different from the default of 64 used in
10.20.
Defect Description:
The default value of the TTL field is different
between 10.20 and 11.00.
Resolution:
The tunable "ip_max_bcast_ttl" can be used to change
the default value of the TTL field. Previously, this
tunable has been documented in the ndd help text as
"unsupported". ip_max_bcast_ttl is now listed as a
supported tunable by ndd.
( SR:8606165595 CR:JAGad34886 )
Symptom:
The xopen_networking(7) man page should say that if you
specify -lc at compile time, -lxnet must precede -lc
Defect Description:
The xopen_networking(7) man page did not have complete
information about library search order.
Resolution:
The xopen_networking(7) man page is updated to tell
users to specify -lxnet before -lc at compile time.
( SR:8606168597 CR:JAGad37875 )
Symptom:
System becomes non-responsive to networking traffic.
Defect Description:
Sync queues in networking modules get too large to
be processed in a reasonable time. This causes the
system to run out of resources.
Resolution:
System now monitors the length of sync queue's
used by networking code. If the queue length
is too long, data mblks are dropped.
( SR:8606172829 CR:JAGad42089 )
Symptom:
A multi-threaded socket application hangs
when a thread is closing a socket.
Defect Description:
In a multi-threaded socket application when
a thread closes a socket on which another
thread is doing accept(), the closing
thread hangs.
Resolution:
Process Management, File System, Streams, and Transport
provide an integrated solution to wake up the sleeping
thread.
( SR:8606176180 CR:JAGad45420 )
Symptom:
UDP socket bound to local IP address cannot receive
broadcast packets.
Defect Description:
Binding a UDP socket to a local address results in
broadcast packets being ignored.
Resolution:
A new ndd tunable "ip_enable_udp_bcastrecv" is provided
to fix the problem. This tunable when set to 1 will
enable UDP sockets bound to local IP addresses to receive
broadcast packets. When set to 0 this tunable prevents
UDP sockets bound to local IP addresses from receiving
broadcast packets. The tunable is associated with the
/dev/ip network device.
( SR:8606179394 CR:JAGad48618 )
Symptom:
ndd has no help text for the tunable
"ip_enable_udp_bcastrecv".
Defect Description:
This tunable is new with this patch.
Resolution:
A new ndd parameter "ip_enable_udp_bcastrecv" is
provided as equivalent to nettune's "udp_newbcastanable"
parameter.
( SR:8606179508 CR:JAGad48732 )
Symptom:
Occasionally the Service Guard command
cmrunnode core dumps.
Defect Description:
The socket() call can return an undocumented
errno of EAGAIN.
Resolution:
The socket(2) man page has been modified with
the following warning:
Not all possible errno values are
documented in each socket related manpage
due to dependencies from the underlying
protocol modules. Refer to the errno(2)
manpage for a complete list of error codes.
( SR:8606195876 CR:JAGad65080 )
Symptom:
The man pages lead users to think that they can only get
ENOPROTOOPT for setsockopt or getsockopt calls.
Defect Description:
The errno(2) man page says you only get ENOPROTOOPT for
setsockopt and getsockopt, but you can get this error for
any system call that causes the system to send a packet
that the remote or an intermediate node does not support,
including operations on NFS files.
Resolution:
The connect(2), recv(2), send(2) and shutdown(2) man pages
were corrected to list the ENOPROTOOPT error number as a
possible errno return value.
( SR:8606199995 CR:JAGad69181 )
Symptom:
The recv(2) man page gives an incorrect description of
the meaning of a return value of 0. It specifically fails
to indicate that this is the means for indicating a normal
connection termination.
Defect Description:
The recv(2) man page does not tell the user that blocking
recv calls may complete with 0 if the end of the socket
data stream has been reached (the remote has shutdown the
connection and there is no data to receive). It also does
not describe what happens when the remote side of a TCP
connection sends a RESET.
Resolution:
The recv(2) man page was updated to describe what happens
when the end of the socket data stream is reached and when
the remote side resets the connection.
( SR:8606201810 CR:JAGad70984 )
Symptom:
11.11 lacks event port functionality.
Defect Description:
Event port functionality is missing from
to 11.11.
Resolution:
Implemented and tested event port
functionality in 11.11.
( SR:8606204354 CR:JAGad73536 )
Symptom:
Outbound UDP datagrams get silently dropped.
Defect Description:
When the size of a UDP packet is greater than
the allowable IP limit, the packet is dropped without
returning an error to the application requesting the send.
Resolution:
The socket send() function has been modified to
return an EMSGSIZE errno when the data size exceeds the
IP limit.
( SR:8606204948 CR:JAGad74124 )
Symptom:
Truncating a file while transferring it
with FTP causes a system hang.
Defect Description:
File truncation causes sendfile to loop forever
allocating and using up buffer cache resources.
Resolution:
sendfile checks if the file size has changed during
the transmission loop. If it has, sendfile treats this
condition like it has reached end of file.
( SR:8606208365 CR:JAGad77551 )
Symptom:
nettl tracing with the filtering option
does not work.
Defect Description:
The way in which IPv4 addresses are stored changed
with the introduction of IPv6. Nettl tracing and
filtering code still expects the old format.
Resolution:
nettl tracing and filtering code has been modified
to recognize the new address format.
( SR:8606208617 CR:JAGad77804 )
Symptom:
Connections on a system with an active server
which also has the "tcp_conn_strategy" parameter
set to non-zero, may see a low overall throughput.
Defect Description:
When "tcp_conn_strategy" is enabled, it is possible
for new connections to inherit a bogus TCP MSS
parameter value from a cached socket which was
previously used as a "listen" socket.
Resolution:
"shutdown" on listening sockets was modified
so that it does not manipulate receive window
state in the same way as "shutdown" on a connection
socket. This prevents a parallel "accept"
(in another thread) on the same socket from inheriting
bogus window parameters from the listen socket.
( SR:8606213997 CR:JAGad83188 )
Symptom:
The following error occurs when compiling "socket.h"
with aCC (ANSI C++ Compiler).
Error 43: "/usr/include/sys/socket.h", line 470
# C++ does not allow Old-style (non-prototype)
function definitions.
inline sbsize_t sendfile(a,b,c,d,e,f) int a,b,f;
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
Defect Description:
Old-style function definition was used in socket.h
Resolution:
The prototype of function declaration
should follow the standard style which
specifies the type for each parameter
inside the parenthesis of the function.
( SR:8606215049 CR:JAGad84240 )
Symptom:
If users assign an ip to a vlan and bring down
the vlan by 'ifconfig vlan down', (not unplumb
it), system could then successfully delete this
vlan. But netstat still have this ip on the
list, and when users create a vlan next, this
vlan already has this ip assigned.
Defect Description:
IP detaches the interface when ifconfig bring down
the interface.
Resolution:
Remove the DL_DETACH_REQ which follows DL_UNBIND_REQ
request when 'ifconfig <interface> down' happens.
( SR:8606215148 CR:JAGad84339 )
Symptom:
System panics on sbflush panic 2:
trace event 0
stack trace for event 0
crash event was a panic
panic+0x14
sbflush+0x68
sbrelease+0x14
sorflush+0xa4
sofree+0x15c
soclose+0x23c
soo_close+0xc8
closef+0x64
close+0x90
syscall+0x6f8
$syscallrtn+0x0
Defect Description:
An AF_UNIX socket must be unlocked while data is
transferred into the user's buffer. During this time
the socket buffer size can be updated incorrectly.
Resolution:
Before unlocking the AF_UNIX socket, the socket
buffer size is saved locally and restored once the
socket has been locked again.
( SR:8606216950 CR:JAGad86105 )
Symptom:
Creation of raw socket using socket() fails when the
protocol specified is either IPPROTO_UDP or IPPROTO_TCP.
Defect Description:
An EACCES error is returned to the application by socket()
when a raw socket is opened with either IPPROTO_UDP or
IPPROTO_TCP specified as the protocol.
Resolution:
Code modified to allow opening of raw sockets with the
protocol specified as either IPPROTO_UDP or IPPROTO_TCP.
( SR:8606217657 CR:JAGad86809 )
Symptom:
ifconfig lan10000 fails, but ifconfig lan9999 succeeds.
Defect Description:
The transport code limits the ppa number to less than
10000.
Resolution:
Remove the limit checking of the ppa number in
transport code.
( SR:8606217835 CR:JAGad86985 )
Symptom:
Bogus TCP retransmissions during specweb99/netperf
TCP_CRR tests.
Defect Description:
Congestion window calculation in TCP is done incorrectly.
Resolution:
The algorithm for computing the congestion window has
been corrected.
( SR:8606218708 CR:JAGad87856 )
Symptom:
The congestion window increases too slowly.
Defect Description:
When the congestion window is at one MSS, via
tcp_cwnd_initial or following a TCP timeout, it
does not begin to increase until after the second
one MSS segment has been ACKed.
Resolution:
The congestion window will now increase after the
first ACK is received.
( SR:8606218753 CR:JAGad87901 )
Symptom:
System runs out of memory when under heavy
inbound TCP traffic.
Defect Description:
The inbound buffer for a socket is allowed
to grow without bounds if the stream head is in
use while data is being appended to the buffer.
Resolution:
A check is now added to see if the sqe list is getting
too long (~1000) on the stream head. If so then close
the tcp window the same as if the data queue of the
stream head were getting too full. (This fix
is dependent on a streams tunable streams_sqmax being
set in the next streams patch)
( SR:8606219937 CR:JAGad89079 )
Symptom:
On UP boxes when 2 (or more) aio_reads are pending
on the same socket and ioctl(SIOCAIOABORT) is used,
a close on that socket will result in an unkillable
hung process.
Defect Description:
On UP boxes when 2 aio_reads are pending on the same
socket, the ioctl(SIOCAIOABORT) only completes one of the
pending aio_reads. So, a close on that socket results
in a hang.
Resolution:
The ioctl(SIOCAIOABORT) has been modified so that the other
aio_read cannot be left sleeping.
( SR:8606220535 CR:JAGad89673 )
Symptom:
Measureware and Glance do not show any networking
statistics.
Defect Description:
libnm does not support access to network management
tables via the loopback interface.
Resolution:
The get_if_table() function in libnm has been modified
to support the loopback interface.
( SR:8606221777 CR:JAGad90911 )
Symptom:
When setting ip_pmtu_strategy to 0 any non-local
networks have a maximum MTU of 576.
Defect Description:
The code assumes that a non-local route cannot
handle large MTU's.
Resolution:
A new value has been added to allow for no PMTU
and yet use the MAX MTU for a link. To implement
this change do the following;
ndd -set /dev/ip ip_pmtu_strategy 3
If a smaller MTU is needed for a given route
then use the above strategy and use the route
command with the pmtu option "-p".
( SR:8606221830 CR:JAGad90955 )
Symptom:
Enhancement Request for HATCP intended for
the exclusive use of the OpenCall product.
Defect Description:
This is an Enhancement request for HATCP OpenCall
Resolution:
Enhancement request for HA is completed.
( SR:8606221942 CR:JAGad91060 )
Symptom:
The same secondary interface may be unexpectedly
configured with multiple IPv6 addresses.
Defect Description:
The kernel repeatedly assigns the same IP index numbers
(the Y in lanX:Y) to multiple autoconfigured interfaces
based on the prefixes it receives from router
advertisements. The problem is seen only when a user
has manually configured a secondary interface with an
IP index number greater than 1.
Resolution:
The logic in the kernel that picks the next
available IP index number assumes that the
interface list is sorted by the IP index number.
The list was not sorted. The problem is fixed
by making the kernel insert new secondary
interfaces in the order that is sorted by
assigned IP index number of the interfaces.
( SR:8606222035 CR:JAGad91150 )
Symptom:
Packets sent to the IPv6 tunnel interface do not
reach their destinations.
Defect Description:
Packets received at the IPv6 tunnel interface are not
forwarded to the target destinations. For forwarding
to work, the kernel requires that the "private" flag
be turned off, but it was left on (by default).
Resolution:
When the tunnel interface, tu0, is brought up via
the ifconfig command, forwarding is automatically
enabled by turning off the "private" flag on that
interface, so that tunneled packets are delivered
to the appropriate physical interfaces.
( SR:8606222508 CR:JAGad91621 )
Symptom:
When system memory use is very high
the accept system call returns ENOBUFS.
Defect Description:
Incorrect handling of listen backlog when
a SYN-RST pair arrives.
Resolution:
The listen backlog is now correctly updated
when a SYN-RST pair arrives. The backlog is not
decremented until the accept executes.
( SR:8606223127 CR:JAGad92230 )
Symptom:
Memory Leak in the 2k bucket when SAM hostagent is
running.
Defect Description:
When a socket joins a multicast group, the kernel allocates
memory to hold the information about the group. This
memory is not freed when socket is closed.
Resolution:
Memory assigned to hold the group information when
a socket joins a multicast group is freed when
the socket is closed.
( SR:8606224453 CR:JAGad93541 )
Symptom:
Mobile IP reply packets do not go the same tunnel
as the request packets.
Defect Description:
Transport did not create a reverse tunnel for
Mobile IP reply packets.
Resolution:
Add reverse tunneling function in transport layer.
( SR:8606224457 CR:JAGad93545 )
Symptom:
System runs out of Anonymous ports when doing
binds for outbound connections.
Defect Description:
Code does not allow the same port to be used
for different local IP addresses.
Resolution:
The local IP address is now used in the decision
process during a bind.
( SR:8606224462 CR:JAGad93550 )
Symptom:
Message getting displayed on console:
tcp_timer: strange state (-6) [5767,d425] TCP_CLOSED
Defect Description:
The tcp connection was set to TCP_CLOSED without stopping
the timer.
Resolution:
When changing the tcp state to TCP_CLOSED in
the case of a connection waiting to be accepted
the timers on the connection are cleaned up.
( SR:8606224472 CR:JAGad93560 )
Symptom:
Slow connection speeds over Long Fat Pipe or
Long Narrow Pipe connections.
Defect Description:
Socket options like SO_RCVBUF, SO_SNDBUF and
SO_SNDLOWAT do not get properly inherited from
the listener.
Resolution:
When the newly accepted connection is updated from
the listen stream the parameters that relate to
Long Narrow Pipe and Long Fat Pipe devices are now
updated.
( SR:8606224475 CR:JAGad93563 )
Symptom:
Possible loss of IPv6 options when using UDP.
Defect Description:
T_UNITDATA_IND's b_wptr wasn't set properly when
IPv6 options were present
Resolution:
The length of the mblk is now properly
adjusted.
PHNE_24211:
( SR:8606139436 CR:JAGad08735 )
Symptom:
system panic on X.25 socket.
A possible deadlock situation
The panic stack is as follows:
panic+0x10
spin_deadlock_failure+0x38
deadlock_check+0x9c
sl_pre_check+0x54
spinlock+0x14
mp_socket_lock+0x34
mp_socket_lock2+0x38
XLS_F_handler+0x6c4
XSO_F_handler+0x958
XLS_F0_a_connect_ind+0x188
XLS_F_handler+0x6a4
XST_F_read_put+0x398
putnext+0x1f4
CI_touser+0x268
Rx_CALL+0x384
L2_datind+0x4ac
dlpi_rxll+0xb4
x25lrsrv+0x60
sq_wrapper+0xc8
str_sched_mp_daemon+0x33c
str_sched_daemon+0x29c
im_mpnetstr+0x28
DoCalllist+0x38
main+0x24
$vstart+0x34
$locore+0x90
Defect Description:
X.25 socket inserts a cloned socket on to
the listen socket's queue for inbound connection
causes deadlock situation
Resolution:
Remove deadlock assertion for X.25 socket, when
X.25 process inserts the cloned socket on to the
listen socket's queue.
( SR:8606154650 CR:JAGad23967 )
Symptom:
Some clients can cause TCP connections to be
set up which will cause excessive memory
consumption by data buffers.
Defect description:
Clients which negotiate a TCP MSS size of 1 byte
(or any very small value) will cause excessive
memory consumption because of large buffers being
allocated to hold only a single byte of data.
Resolution:
The default value of the tunable variable
tcp_mss_min has been changed from 1 to 108
so that any clients will be refused if they
attempt to establish connections which request
an MSS (maximum segment size) value which is
too small.
( SR:8606161095 CR:JAGad30413 )
Symptom:
NDD does not support the reporting or changing
of kernel socket parameters.
Defect Description:
The changes for NDD to support kernel socket
parameters were not ready in time for LR.
Resolution:
The necessary changes have been made to
NDD to allow it to query and alter
socket parameters.
( SR:8606164046 CR:JAGad33347 )
Symptom:
Some customers using the Raptor Firewall product found
the daemon processes (tcp_gspd) spinning and eating
CPU time (70-80% of a cpu on UP systems).
Defect Description:
The symptom is because the daemon process was looping
in a select/recv loop on a send-side-shutdown socket
where select returned one socket FD as readable, but
recv always returned EWOULDBLOCK.
What happened was, prior to the daemon process's shutdown
call, the remote client had already reset the connection
which generated a T_DISCON_IND that went up to the
stream head and set TCP state to TCPS_IDLE.
If the daemon process then called shutdown, the M_ERROR
message block sent up by tcp_wput_proto (it does this
because it finds the TCP is in the TCPS_IDLE state)
caused the T_DISCON_IND to get flushed by stream. Then
recv looped because there was nothing to read on the
queue. The whole thing is due to the race at stream head
between processing of T_DISCON_IND coming toward stream
head from TCP and generation of T_ORDREL_REQ going down-
stream towards TCP from socket's shutdown code.
Resolution:
In socket's shutdown code, check if the connection still
exist before sending T_ORDREL_REQ downstream towards TCP,
if the connection is lost, set errno to EINVAL and exit.
( SR:8606166814 CR:JAGad36101 )
Symptom:
Remsh hangs.
Defect Description:
Remsh hangs due to incorrect behavior in shutdown during
write flow control situation.
Resolution:
Transport calls a version of streams_putmsg() which
unconditionally does the putnext().
( SR:8606167654 CR:JAGad36937 )
Symptom:
Sometimes nettl is unable to capture reply packets.
Defect Description:
Nettl is unable to trace outbound IP packets. This happens
when drivers do not support FAST PATH.
Resolution:
The tracing subsystem always assumed drivers to be
FASTPATH and used to skip tracing messages for
NON-FASTPATH drivers. Now the tracing subsystem
recognizes messages for both FASTPATH and
NON-FASTPATH drives which eventually leads nettl
to trace outbound packets.
( SR:8606171754 CR:JAGad41018 )
Symptom:
Many NDD parameters are not recognized by
the help text system in the current version of NDD.
Defect Description:
Numerous changes have been made to the transport code
in the kernel which are not reflected in the help text
currently provided by NDD.
Resolution:
Additional help text has been added to
NDD to bring it up to date.
( SR:8606178097 CR:JAGad47324 )
Symptom:
The system has a Data Page Fault panic with
either of the following two traces:
Trace 1:
panic+0x14
report_trap_or_int_and_panic+0x4c
interrupt+0x1e8
$ihndlr_rtn+0x0
puthere+0x44
mi_timeout_exec+0x224
sw_service+0x8c
mp_ext_interrupt+0x108
ivti_patch_to_nop3+0x0
idle+0x1e8
swidle_exit+0x0
Trace 2:
panic+0x14
report_trap_or_int_and_panic+0x4c
interrupt+0x1e8
$ihndlr_rtn+0x0
puthere+0x44
tcp_rput_context_check+0x4a4
tcp_rput+0x784
putnext+0x16c
ip_rput_local+0x524
ip_rput+0x648
putnext+0x16c
hp_dlpi_mbuf_fast_in+0x70
hp_dlpi_intr_put+0x154
streams_put+0x134
hp_dlpi_intr+0x84
lan2_process_packet+0xcc
lan2_int_fr_rnr+0x19c
lan2_isr+0x184
eisa_int+0x134
lasi_interrupt+0x5c
mp_ext_interrupt+0x300
ivti_patch_to_nop3+0x0
idle+0x1b8
swidle_exit+0x0
Defect Description:
The LISTEN socket is closed while the SYN_RCVD
socket is still open.
Resolution:
An unbind request will trigger the cleaning up of
all outstanding connections.
( SR:8606179470 CR:JAGad48694 )
Symptom:
Executing "ifconfig lanX:Y unplumb" may core dump
ifconfig.
Defect Description:
ifconfig did not initialize a file descriptor value
to NULL which caused it to use a garbage value when
closing the file descriptor.
Resolution:
In the ifconfig program, a file descriptor was
initialized to NULL to prevent ifconfig from core
dumping. Also the following error message will be
displayed to the user whenever he/she tries to
unplumb a secondary interface with ifconfig.
"Can't unplumb secondary interface".
( SR:8606181093 CR:JAGad50313 )
Symptom:
ifconfig does not generate a message when an
"unplumb" operation is unsuccessful.
Defect Description:
ifconfig was not checking for unsuccessful "unplumb"
operations.
Resolution:
For unsuccessful "unplumb" operations, ifconfig will
display the following line to the user.
unplumb operation for interface_name was unsuccessful
( SR:8606193754 CR:JAGad62965 )
Symptom:
Performance of ftp over hyper-fabric can be very slow
when "put"-ing a file.
Defect Description:
In the case of large MTU interfaces like hyperfabric,
TCP sets the upper bound used to determine whether
an ACK should be delayed incorrectly (too high),
and this results in the excessive delaying of ACKs.
These delays are responsible for the reduction in
throughput.
Resolution:
The computation of the upper bound used to determine
when ACKs are delayed ACKs has been modified to
prevent the excessive delaying of ACKs.
( SR:8606196811 CR:JAGad66009 )
Symptom:
After installing the transport 11.11 Patch, PHNE_23645,
the following three debug lines appear in the /etc/rc.log
file when the system is booted.
name= tcp_conn_strategy, value= 0
do_sockets_nddset() &stri= 0x7f7f0538
0
Defect Description:
Debug statements were built into the ndd program.
Resolution:
Three printf statements used for debug purposes were
removed from the ndd program.
( SR:8606196818 CR:JAGad66016 )
Symptom:
The "ifindex" number for the loopback driver, lo0, is not
being set to a non-zero number. This does not conform
to RFC 1213. This could cause trouble for programs such
as OpenView or programs using libnm.sl by causing them
to retrieve a zero "ifindex" value.
Defect Description:
There is no networking code for HP-UX 11.11 to set the
"ifindex" value for lo0, thus the default value
of zero is being used.
Resolution:
The ifconfig program when it is configuring the loopback
interface, lo0, sets the "ifindex" number for lo0
with a SIOCSIFINDEX ioctl() call. Before this ioctl()
call, a SIOCGNEXTIFIDX ioctl() call is executed to get
the next available "ifindex" number from the kernel.
( SR:8606202009 CR:JAGad71183 )
Symptom:
TCPdump shows multiple threads connecting to
the same server port
Defect Description:
The same ephemeral port is being allocated to
multiple endpoints.
Resolution:
Each endpoint is allocated a different ephemeral port.
( SR:8606205122 CR:JAGad74298 )
Symptom:
HP-UX Transport does not support IPv6.
Defect Description:
HP-UX Transport does not support IPv6.
Resolution:
With this enhancement IPv6 support can be enabled in
Transport by installing the IPv6 New Core Functionality
(NCF) product.
( SR:8606174765 CR:JAGad44011 )
Symptom:
"ifconfig <ifname> inet" does not work for non-privileged
users.
Defect Description:
Privileges should not be required to execute
"ifconfig <ifname> inet".
Resolution:
The check for privileges is performed after all optional
command arguments have been processed.
( SR:8606202365 CR:JAGad71539 )
Symptom:
Panic in sofromcache with tcp_conn_strategy=1.
panic+0x6c
report_trap_or_int_and_panic+0x94
trap+0xed4
nokgdb+0x8
asm_spinlock_usav+0x14
sofromcache+0x48
soclone+0x148
soaccept+0xdc
sodequeue+0xb8
accept+0x148
syscall+0x204
$syscallrtn+0x0
Defect Description:
There is a defect in the handling of an array index local
to sofromcache, causing the index to go out of range. This
results in a panic when using the out of range array
element.
Resolution:
A change has been made to sofromcache to prevent the out
of range array index which caused the panic.
( SR:8606204817 CR:JAGad73995 )
Symptom:
System panics with the following when
tcp_conn_strategy=1;
panic+0xc4
report_trap_or_int_and_panic+0x108
trap+0x1258
nokgdb+0x8
socachecallback+0x628
osr_sth_so_callback+0x84
str_simple_ioctl+0x4924
hpstreams_ioctl_int+0x77c
streams_ioctl+0x98
sotocache+0x128
sounlock+0x834
mp_socket_unlock+0x1c
soclose+0x13c0
soo_close+0x138
closef+0x2b0
close+0x1d8
syscall+0xe70
$syscallrtn+0x0
Defect Description:
A pointer (b_rptr) was not reinitialized correctly when an
mblk was reused.
Resolution:
If this is the only reference to this mblk then reset the
b_rptr to db_base and continue to reuse the mblk. If this
is duped reference free this reference and allocate a new
mblk.
( SR:8606151945 CR:JAGad21284 )
Symptom:
The system experiences extensive memory consumption when
processing large numbers of routes.
Defect Description:
An internet routing entry (IRE) is allocated for each
route, whether or not the route is valid.
Resolution:
Introduce two new ndd tunables, ip_ire_route_limit and
ip_ire_route_limit_interval. The implementation will
start flushing idle IRE_ROUTEs when the count has reached
half of the limit. The count is checked based on the
interval, measured in milliseconds.
( SR:8606198614 CR:JAGad67803 )
Symptom:
System hangs and free memory is zero.
System TOC with the following trace:
putnext+0x188
ip_wput_ire+0x398
ip_wput+0x470
putnext+0xcc
tcp_rput+0x1088
csq_turnover_with_lock+0x84
str_spu_sw_isr+0x200
sw_service+0xb0
mp_ext_interrupt+0x150
ivti_patch_to_nop3+0x0
ki_accum_push_TOS+0xa0
syscall+0x808
$syscallrtn+0x0
Defect Description:
TCP floods ACKs to the driver causing depletion of 512 byte
memory bucket.
Resolution:
Stopped the flood of ACKS to all drivers from TCP.
( SR:8606203900 CR:JAGad73078 )
Symptom:
TCP connection hangs.
Defect Description:
Out of Order FIN with no data is not remembered.
Resolution:
Out-of-order inbound FIN handling was improved to work in
the case where the FIN arrives without any data.
( SR:8606203039 CR:JAGad72213 )
Symptom:
A UDP broadcast is sent once for each IP address configured
on the interface.
Defect Description:
This feature is currently used to get broadcasts sent to
multiple interfaces, when the broadcast address being used
applies to multiple interfaces. A whole net broadcast is
replicated on every connected subnet of the target net.
Resolution:
Add ip_duplicate_broadcasts tunable for eliminating
duplicate physical broadcasts. The default setting of
zero suppresses the duplicates.
( SR:8606180610 CR:JAGad49831 )
Symptom:
System panics with a Unresolved priv 0 interruption.
TPI_discon_ind
sosend
soo_rw
rwuio
write
syscall
Defect Description:
A race condition occurred when a TCP socket was making
unprotected accesses to Stream head, and an M_FLUSH
message was coming upstream and wiped out the message
after the socket grabbed the message.
Resolution:
Enhanced the socket to use the protection mechanism OSR
to make safe accesses to Stream head.
( SR:8606187509 CR:JAGad56716 )
Symptom:
When a program uses the ICMP TIMESTAMP request message
to query an HP-UX system for the current time, the reply
generated provides less precise information than was
available from a 10.20 system.
Defect Description:
The code which constructs the reply to the ICMP timestamp
request calculates its time value based on the kernel's
timer interrupt counter variable ("lbolt") which updates
once every 10ms and is not precise enough to give a time
value with a higher resolution.
Resolution:
The code which constructs the ICMP timestamp reply packet
was modified to obtain its time information from the same
high-resolution timer used to provide the microsecond-
granularity tv_usec value in the "struct timeval" returned
from the "gettimeofday" system call.
( SR:8606140093 CR:JAGad09415 )
Symptom:
TCP connections where both ends close at the same time may
experience an unnecessary delay of 1.5 seconds or more.
This problem has an especially high probability of being
seen on loopback connections where one end of the
connection has the SO_LINGER option turned on.
Defect Description:
During connection shutdown, if a FIN packet is received
while TCP is performing its "close" operation, the FIN
packet can be discarded, causing the remote end to time
out (awaiting acknowledgement of the FIN) and retransmit
the FIN packet. This timeout accounts for the 1.5 second
(or longer) delay.
Resolution:
Processing of TCP packets arriving for closing connections
has been modified so that they are no longer discarded.
( SR:8606209311 CR:JAGaa26796 )
Symptom:
A man page for traceroute does not exist.
Defect Description:
The traceroute man page is not delivered.
Resolution:
Deliver the traceroute man page.
( SR:8606164864 CR:JAGad34161 )
Symptom:
A MSG_OOB recv() on a socket which has been set to
non-blocking and for which there is no OOB data blocks
rather than returning an error.
Defect Description:
The logic handling socket state variables associated with
the processing of OOB data is flawed and allows a blocking
read to occur when no OOB data is present.
Resolution:
The management of socket state variables has been
corrected so that a MSG_OOB recv() no longer blocks when
OOB data is not present on the socket.
( SR:8606136810 CR:JAGad05930 )
Symptom:
The netstat(1) man page says -rs works, but it does not.
Defect Description:
The netstat(1) man page was not updated to say that -rs
options are invalid together.
Resolution:
The netstat(1) man page was corrected.
( SR:8606207797 CR:JAGad76974 )
Symptom:
Performance degradation in IPSec traffic.
Defect Description:
Even for cleartext traffic in IPSec there is a performance
degradation.
Resolution:
When TRUST is set we will not make a policy check on any
incoming clear text packet but will still decrypt or
authenticate any incoming IPSec packet that arrives via
that interface. For outbound or forwarded traffic the
IPSec policy code will not be consulted if TRUST is set on
the outbound/forwarded-to interface.
( SR:8606140128 CR:JAGad09450 )
Symptom:
"netstat -i" does not display inbound errors ("Ierrs"),
outbound errors ("Oerrs"), and collisions ("Coll").
Defect Description:
The netstat -i information on 11.11 is not the same as
on 10.20. The 10.20 netstat -i displays inbound errors
("Ierrs"), outbound errors ("Oerrs"), and collisions
("Coll").
Resolution:
Call DLPI and use the DLPI primitive DL_GET_STATISTICS_REQ
to get statistics from the LAN interface device.
( SR:8606211447 CR:JAGad80635 )
Symptom:
swremove of PHNE_23645 fails due to unresolved
externals. This can only happen on a system which has
been upgraded from 11.00.
Defect Description:
The incorrect version of an IPSec kernel library is
restored during the remove.
Resolution:
The preinstall script has been modified to save the
correct version of the IPSec kernel library.
( SR:8606168903 CR:JAGad38181 )
Symptom:
The thread safety of libc APIs is not consistently
mentioned in the man pages.
Defect Description:
There is no consistent, supportable list of threadsafe
APIs in libc. Without this information customers cannot
easily write threads programs.
Resolution:
The "MULTITHREAD USAGE" section has been removed from
the Transport's man pages. A pointer to thread_safety(5)
has been added to Transport's man pages documenting
threadsafe functions.
( SR:8606160792 CR:JAGad30111 )
Symptom:
A server program listening for incoming connection requests
with select() or accept() can receive notification of a
connection's existence before it is fully established.
Defect Description:
As an optimization, incoming connection requests are
checked to see if they originate from an address kept in a
cache of machines with previously established connections.
If the remote system's address is found in the cache then
the TCP code does not wait for the full connection
handshake protocol to complete before notifying the
application. In some cases an application can be falsely
notified of the existence of connections which never
complete.
Resolution:
A new tunable variable "tcp_early_conn_ind" has been added,
whose state can be modified with the "ndd" command. The
default value of this parameter is 1, which leaves
connection establishment behavior alone; that is, it
remains the same as without this change. With the variable
set to 0, applications will not be given early notification
of incoming connection requests, prior the full connection
establishment.
( SR:8606144985 CR:JAGad14323 )
Symptom:
System panics with the following stack trace.
tcp_err_ack
tcp_info_req
tcp_wput_proto
tcp_wput
putnext
str_async_ioctl
hpstreams_ioctl_int
streams_ioctl
soprotostate
soclose
This panic can only occur when a system is under
extreme memory pressure.
Defect Description:
tcp_info_req always passes a NULL pointer to tcp_err_ack.
When memory is exhausted, it causes system panics due to
a NULL pointer reference.
Resolution:
Code corrected to reference a valid non null pointer.
( SR:8606205538 CR:JAGad74714 )
Symptom:
System panics with the following stack trace;
panic
report_trap_or_int_and_panic
interrupt
ihandler
x_mi_timer
tcp_accept
tcp_wput_proto
tcp_wput
csq_turnover_with_lock
puthere
tcp_icmp_error
tcp_rput_other
tcp_rput
putnext
icmp_inbound_error
icmp_inbound
ip_rput_local
ip_rput
putnext
hp_dlpi_mblk_fast_in
hp_dlpi_mblk_intr_put
streams_put
streams_put_release
hp_dlpi_mblk_intr
lanc_ether_ics
btlan_receive_frame
btlan_isr
sapic_interrupt
mp_ext_interrupt
ihandler
Defect Description:
An ICMP error message is received for a stream that is in
the middle of the accept state. The function that processes
the message cleans up the data structure related to the
accept stream. When accept stream tries to access this
data structure it panics.
Resolution:
If TCP receives an ICMP error message for a stream that is
in the middle of accept state, it ignores that message.
( SR:8606207520 CR:JAGad76696 )
Symptom:
libnm gives wrong number of interfaces when queried for
number of interfaces.
Defect Description:
The variable that maintains the number of interfaces is not
being updated properly in the kernel.
Resolution:
The kernel variable is updated with the correct number of
interfaces present on the system.
( SR:8606209138 CR:JAGad78325 )
Symptom:
With ATM on system and IP traffic, system panics
with the following stack trace;
panic
report_trap_or_int_and_panic
interrupt
ihandler
bcopy_pcxu_method
LDI_F_set_fastpath
LEC_F0_uwioctl
LEC_F0_uwput
csq_turnover_with_lock
putnext
CAM_F0_sp_dataind
streams_put
CAM_F_data_ind
alx_receive_it
alx_isr
wsio_interrupt
mp_ext_interrupt
ivti_patch_to_nop3
sounlock
mp_socket_unlock
soo_select2
soo_select
select
syscall
Defect Description:
A badly formatted request message is sent to ATM by IP
while doing a fastpath probe. This happens on a
multiprocessor system. One processor is formatting the
message while the other sends this incomplete message to
the driver as a fastpath probe. This lead ATM driver to
panic.
Resolution:
A fastpath probe message is only sent to the driver if it
is properly formatted.
( SR:8606206366 CR:JAGad75539 )
Symptom:
System panics with the following stack trace:
panic
report_trap_or_int_and_panic
trap
nokgdb
ipif_up
ip_sioctl_copyin_done
ip_wput_nondata
ip_wput
csq_turnover_with_lock
putnext
udp_wput_other
udp_wput
putnext
wait_iocack
str_trans_ioctl
hpstreams_ioctl_int
streams_ioctl
ifioctl
soo_ioctl
ioctl
syscall
Defect Description:
When bringing up an interface it is not identified as a
virtual interface. Attempting to bind and attach, results
in a panic.
Resolution:
Identify a virtual interface when bringing up an interface.
PHNE_23645:
( SR:8606202746 CR:JAGad71920 )
Symptom:
A syntax error in S340net start script at line 223.
Defect Description:
Startup script generates a syntax error when
performing Mobile IPv4 interface checks.
Resolution:
Fixed syntax error in the startup script.
( SR:8606199716 CR:JAGad68902 )
Symptom:
The following message is logged in /etc/rc.log:
/sbin/rc2.d/S340net[4]: i <= interface_count : The
specified number is not valid for this comm
Defect Description:
Startup script assumes that Mobile IP has been installed.
Resolution:
Startup script checks to see if Mobile IP is installed
and handles its absence correctly.
( SR:8606184133 CR:JAGad53342 )
Symptom:
Network client systems which request system
time from an HP-UX 11.11 system using the
"ICMP timestamp request" message do not get a reply.
Defect Description:
The code which handles the incoming
ICMP timestamp request actually
builds the reply packet incorrectly
A sanity check in the IP transmit
code discards the packet as bogus
before it is transmitted.
Resolution:
ICMP timestamp reply messages are
now built correctly.
( SR:4701419036 CR:JAGaa93907 )
Symptom:
If one urgent byte is on a socket, and a recv() is done
without MSG_OOB on that socket, the urgent byte will
disappear. Subsequent recvmsg() calls with the MSG_OOB
flag on will return EINVAL, instead of returning the
urgent byte.
Defect Description:
Socket state relating to urgent data
is not maintained correctly allowing
urgent data to be lost.
Resolution:
Sockets code has been modified to ensure that the
OOB state is maintained correctly.
( SR:8606193751 CR:JAGab13526 )
Symptom:
A sending program can send forever even to a connection
that has been closed. This is only seen if
tcp_conn_strategy is set to NON-NULL.
Defect Description:
When using tcp_conn_strategy set to non-NULL a connection
is not unbound when it is closed. This allows data to be
received on the connection forever.
Resolution:
Close now correctly unbind's a tcp connection even
if tcp_conn_strategy is being used.
( SR:1653307850 CR:JAGab20953 )
Symptom:
Some commands hang when the interface
associated with the hostname as defined in /etc/hosts
goes down. Pinging the hostname on the local host will fail.
Defect Description:
Local routing information is deleted unconditionally
when an interface goes down.
Resolution:
Leave a "loopback" route when taking down an
interface either by the ifconfig command
or from the device driver. This will
"allow" processes to keep working in
loopback even over an interface that is down.
( SR:8606193755 CR:JAGab72514 )
Symptom:
When using shutdown() and stack-caching the mss value
could be set to one (1) byte.
Defect Description:
New socket gets previously set variables from an
"old" socket.
Resolution:
Socket code now resets the initial socket values
correctly.
( SR:8606113564 CR:JAGab76776 )
Symptom:
When using AF_UNIX sockets blocking sendmsg()
returns EMSGSIZE when receive side is
close to being full. This can happen on blocking or
non-blocking.
Defect Description:
Miscalculation of how much was going to be sent causes
overflow of receive side resulting in the EMSGSIZE.
Resolution:
The calculation that is done before sending was corrected
for the 64-bit case.
( SR:8606134574 CR:JAGad03709 )
Symptom:
Inbound packet traffic is seen mostly on one processor.
Need enhancement to make inbound packet scheduling better.
Defect Description:
Each network card interrupts only one CPU. The packet then
goes to the top of the stack on this same CPU keeping the
CPU from doing any other work. All of the other CPU's on
the system could be idle.
Resolution:
The lower layers of the network now will send an
inbound packet to the last "known" cpu that a
thread ran on. This will distribute the inbound
packet load on the system thereby increasing
network link performance.
( SR:8606144006 CR:JAGad13339 )
Symptom:
Non-privileged users cannot open a UDP socket
with an ephemeral port.
Defect Description:
Updating of the next ephemeral port to use
was done incorrectly.
Resolution:
Updating of the next ephemeral port to use has
been changed so that it wraps correctly when
the upper bound is reached.
( SR:8606147348 CR:JAGad16691 )
Symptom:
Data is received with garbage bytes appended to it. The
length of the received data will have a size which is a
multiple of STRMSGSZ (where STRMSGSZ is the value of the
kernel parameter STRMSGSZ on the system sending the data).
For this to occur, the kernel parameter STRMSGSZ must have
been changed to a value smaller than the default on the
system which sends the data.
Defect Description:
Sockets fails to update a variable used to compute the size
of an outgoing message. If STRMSGSZ has been set to a value
smaller than the default, and the size of the data being
sent is larger than STRMSGSZ, sockets will compute a length
which is too long. This results in the sending of bytes
that happen to follow the real data in the outgoing buffer.
Resolution:
The value of the variable resid is kept up to
date allowing sockets to compute the correct
send length.
( SR:8606155790 CR:JAGad25103 )
Symptom:
X-terminals are not able to boot from a server
operating in a Service Guard environment.
Defect Description:
Client machines such as X-terminals are not able to
get a boot file from the server in a Service Guard
environment because the server is unable receive the
remote IP address.
Resolution:
The socket options IP_RECVDSTADDR and IP_RECVIFADDR have
been modified to be XOPEN conformant. Both the local and
remote IP addresses are returned by RECVMSG when these
options have been specified. This fix requires a TFTP fix
(see JAGad25107/PHNE_23457) to completely solve the BOOTP
problem.
( SR:8606165729 CR:JAGad35020 )
Symptom:
Customer is seeing RESET's during connects if the
requested MSS is smaller than the tcp_mss_min
set on the system.
Defect Description:
The lower limit of tcp_mss_min was set to 108.
Any connection coming in with a requested MSS
value below this would get RESET.
Resolution:
Lowered the MINIMUM value of tcp_mss_min back to one (1).
Removed the RESET code when a requested mss comes
in under tcp_mss_min. Instead just roll to tcp_mss_min.
The default value for tcp_mss_min is now 108.
( SR:8606165984 CR:JAGad35271 )
Symptom:
Panic during FIN processing on a system
experiencing memory pressure.
0x000e4450 (set_protid_registers+0x1bfc)
0x004e6ff4 (tcp_hash_insert_port+0x2c)
0x004ee800 (tcp_reinit+0x4d0)
0x004e31c8 (tcp_clean_death+0x280)
0x004f00e0 (tcp_rput+0x1838)
0x0057ea68 (csq_turnover_with_lock+0xb0)
0x0056ede0 (putnext+0x2b0)
0x0059f75c (streams_write_uio+0x204)
0x003832b0 (sosend+0x1230)
0x003783a4 (sendit+0x41c)
0x00377eac (send+0x74)
0x0031b304 (syscall+0x754)
Defect Description:
When processing a FIN, TCP does not always check to
see if an orderly release indication message has been
allocated.
Resolution:
TCP has been modified to ensure that an orderly release
indication message is allocated.
( SR:8606167240 CR:JAGad36526 )
Symptom:
System panic's with;
4) putnext+0x18
5) ip_wput_local+0x528
6) ip_wput_ire+0x43c
7) ip_wput+0x458
8) csq_turnover_with_lock+0x88
9) osr_pop_subr+0x98c
10) osr_close_subr+0xb00
11) hpstreams_close_int+0xd40
12) hpstreams_close+0x2c
13) call_open_close+0x1f8
14) closed+0xb0
15) spec_close+0x54
16) vn_close+0x48
17) vno_close+0x20
18) closef+0x68
19) close+0x48
20) syscall+0x200
When using XTI programs over loopback.
Defect Description:
There is a race when an XTI uses t_unbind()
followed immediately by a t_close() on a connection
that is over loopback. This race causes a lookup in
the ipc table to become stale resulting in a panic.
Resolution:
The code was changed to make sure that ip_flush_readers()
is called unconditionally when t_close() is called after
a t_unbind() has been called.
( SR:8606181319 CR:JAGad50536 )
Symptom:
An ENOTCONN is intermittently returned when
a read is done on a non-blocking socket.
Defect Description:
There exists a window where select() can
return without properly setting the correct
socket flag.
Resolution:
Set the socket into the connected state at
the appropriate time.
( SR:8606182096 CR:JAGad51312 )
Symptom:
Applications such as web servers
which use the 32-bit "sendfile" system call
to send files with the O_LARGEFILE
flag set fail with an EOVERFLOW
error.
Defect Description:
The 32-bit version of the "sendfile" system
call refused to transmit files opened with
the O_LARGEFILE flag. This is due to
unnecessarily strict validation checks.
Resolution:
The 32-bit sendfile validation checking has been
loosened to allow 32-bit applications to send
files which are less than 2147483648 bytes in
length.
( SR:8606184456 CR:JAGad53664 )
Symptom:
On a system with a primary and secondary IP address.
If after a user has joined a multicast group the secondary
address is removed the user no longer receives multicast
packets.
Defect Description:
ARP multicast mapping entries were inadvertently deleted
when secondary interfaces were downed.
Resolution:
When an interface is taken down the code checks
to make sure that this was the last one before
unmapping multicast addresses.
( SR:8606185603 CR:JAGad54805 )
Symptom:
Unable to compile IPv6 applications.
Defect Description:
IPv6 header definitions were not in header files.
Resolution:
Updated the header files to include the new IPv6
definitions.
( SR:8606186550 CR:JAGad55754 )
Symptom:
Slow MIB performance. Can cause the system
to look like it has hung when doing netstat or ndd
on a system with a lot of connections.
Defect Description:
Too much time is being spent allocating memory.
Resolution:
The routines that gathered the MIB statistics
for ndd and netstat now allocate their
memory before grabbing any spinlocks.
( SR:8606189357 CR:JAGad58572 )
Symptom:
System panics when Apache WEB server is started.
Defect Description:
Hash table that is special for port 80 was too small.
Resolution:
Initialized the hash for the special port
to be the correct size.
( SR:8606193852 CR:JAGad63062 )
Symptom:
Doing a ping to a broadcast with just a few packets in
rapid succession will panic the system.
System will also crash while using Mobile IP.
ip_rput_local+0x22b4
ip_rput+0x90c
putnext+0xcc
hp_dlpi_mblk_fast_in+0xa0
hp_dlpi_mblk_intr_put+0xb8
streams_put+0xdc
streams_put_release+0x4c
hp_dlpi_mblk_intr+0x5c
Defect Description:
IP code is walking off of the end of a table.
Resolution:
Prevent IP from walking off
the end of the "ip upper mux"
table.
( SR:8606195961 CR:JAGad65165 )
Symptom:
System panics at boot time with the following stack;
panic: assertion failed (holding spinlock(s) on entry
to copyout) at line 469 in /ux/core/kern/pa/vm/vm_mapcall.c
PC-Offset Stack Trace (read down, top of stack is 1st):
0x0006cde8 (panic+0xa0)
0x0006d234 (assfail+0x3c)
0x0006d3ec (_assfail+0x2c)
0x0010452c (copyout+0xfc)
0x00489614 (sockndd_error+0x24)
0x00489d0c (soo_ioctl+0x6e4)
0x001d9004 (ioctl+0x13c)
0x003d200c (syscall+0x754)
0x00033a1c (syscallinit+0x554)
End Of Stack
Defect Description:
A spinlock was held while accessing a function that could
sleep.
Resolution:
Moved the socket unlock from after the
line that did the malloc to before it.
( SR:8606197685 CR:JAGad66877 )
Symptom:
T_ALLOPT option in XTI/TLI will fail with TBADOPT.
Also the tcp initial congestion window is too large
possibly causing performance problems at connect time.
Defect:
TCP_CWND_INITIAL was defined in tcp.h as 0x16.
It should have been TCP_CWND_INIT.
Resolution:
Changed the name of TCP_CWND_INITIAL to
TCP_CWND_INIT in tcp.h
PHNE_23288:
( SR:8606174670 CR:JAGad43916 )
A number of program header files were delivered in their
"debug" forms rather than in their intended "performance"
forms. Kernel-intrusive programs (such as drivers and
programs that access /dev/mem or /dev/kmem) compiled with
these header files may contain internal structures that are
not aligned with the actual kernel structures.
Resolution:
A set of patches redeliver the header files in their
intended forms. All kernel-intrusive programs originally
compiled using the header files included on the HP-UX 11i
Operating Environment OE Install and Recovery media dated
December 2000 must be recompiled using the corrected header
files.
All of the corrected header file patches are included in the
BUNDLE11i bundle on on the HP-UX 11i Operating Environment
Core OE Install and Recovery media dated February 2001 or
later.
The complete list of header file patches is:
PHNE_23288 /usr/conf/net/netmp.h
/usr/include/net/netmp.h
/usr/conf/sys/socketvar.h
/usr/include/sys/socketvar.h
/usr/conf/sys/unpcb.h
/usr/include/sys/unpcb.h
PHNE_23289 /usr/include/sio/mux4.h
PHKL_23290 /usr/conf/space.h.d/system_space.h
PHKL_23291 /usr/conf/graf/gpu_data.h
PHKL_23292 /usr/conf/io/scsi_surface.h
PHKL_23293 /usr/conf/sys/assert.h
PHKL_23294 /usr/conf/sys/buf.h
/usr/include/sys/buf.h
PHKL_23295 /usr/conf/sys/debug.h
/usr/include/sys/debug.h
PHKL_23296 /usr/conf/sys/dnlc.h
/usr/include/sys/dnlc.h
PHKL_23297 /usr/conf/sys/io.h
/usr/include/sys/io.h
PHKL_23298 /usr/include/sys/ki_iface.h
PHKL_23299 /usr/conf/sys/pfdat.h
/usr/include/sys/pfdat.h
PHKL_23300 /usr/include/sys/proc_debug.h
PHKL_23301 /usr/conf/sys/proc_iface.h
/usr/include/sys/proc_iface.h
PHKL_23302 /usr/conf/sys/rw_lock.h
/usr/include/sys/rw_lock.h
PHKL_23303 /usr/conf/sys/sem_alpha.h
/usr/include/sys/sem_alpha.h
PHKL_23304 /usr/conf/sys/sem_beta.h
/usr/include/sys/sem_beta.h
PHKL_23305 /usr/conf/sys/sem_sync.h
/usr/include/sys/sem_sync.h
PHKL_23306 /usr/conf/sys/sem_utl.h
/usr/include/sys/sem_utl.h
PHKL_23307 /usr/conf/sys/spinlock.h
/usr/include/sys/spinlock.h
PHKL_23308 /usr/conf/sys/vas.h
/usr/include/sys/vas.h
PHKL_23309 /usr/conf/sys/vfd.h
/usr/include/sys/vfd.h
PHKL_23310 /usr/conf/sys/vnode.h
/usr/include/sys/vnode.h
PHKL_23311 /usr/conf/ufs/inode.h
/usr/include/sys/inode.h
PHKL_23312 /usr/conf/wsio/pci.h
/usr/include/sys/pci.h
PHKL_23313 /usr/include/sys/scsi_ctl.h
PHKL_23314 /usr/conf/pa/sync/spinlock.h
/usr/include/pa/sync/spinlock.h
PHKL_23315 /usr/conf/pa/cpu.h
/usr/include/pa/cpu.h
PHKL_23316 /usr/conf/pa/sys/map.h
Enhancement:
No (superseded patches contained enhancements)
PHNE_28089:
Enhancements were delivered in patches this one has
superseded. Please review the Defect Description
text for more information.
SR:
8606420692 8606416944 8606421454 8606426280 8606415076
8606424562 8606429097 8606438456 8606438648 8606428509
8606431587 8606441063 8606406155 8606409272 8606413150
8606416735 8606424136 8606416402 8606404400 8606424222
8606428254 8606432060 8606431462 8606322228 8606352228
8606381553 8606385089 8606387073 8606387074 8606387098
8606387075 8606387472 8606388101 8606388102 8606390716
8606390906 8606392043 8606392816 8606394962 8606395825
8606396734 8606401226 8606404988 8606405138 8606411217
8606412656 8606416090 1653262485 1653307850 4701419036
8606113564 8606113716 8606129381 8606134574 8606136810
8606137536 8606138758 8606139436 8606140093 8606140128
8606144006 8606144985 8606146239 8606147348 8606151945
8606154650 8606155790 8606160792 8606161095 8606162623
8606164046 8606164864 8606165144 8606165595 8606165729
8606165984 8606166814 8606167240 8606167654 8606168597
8606168903 8606170818 8606171754 8606172829 8606174670
8606174765 8606176180 8606178097 8606179394 8606179470
8606179508 8606180610 8606181093 8606181283 8606181319
8606182096 8606184133 8606184456 8606185191 8606185603
8606186550 8606187509 8606189015 8606189357 8606193751
8606193754 8606193755 8606193852 8606195876 8606195961
8606196811 8606196818 8606197685 8606198555 8606198614
8606198667 8606199716 8606199995 8606201810 8606202009
8606202365 8606202746 8606202871 8606203039 8606203612
8606203900 8606204354 8606204817 8606204948 8606205122
8606205538 8606206366 8606206785 8606207520 8606207797
8606208110 8606208360 8606208365 8606208617 8606209138
8606209264 8606209311 8606210971 8606211447 8606211448
8606213513 8606213579 8606213997 8606215049 8606215148
8606216047 8606216950 8606217313 8606217657 8606217835
8606218708 8606218753 8606219937 8606220082 8606220535
8606220677 8606221602 8606221683 8606221777 8606221830
8606221942 8606222035 8606222508 8606223042 8606223127
8606223770 8606224453 8606224457 8606224462 8606224472
8606224475 8606224560 8606224838 8606225148 8606225238
8606225324 8606225422 8606226976 8606227191 8606227536
8606227920 8606227936 8606228250 8606229037 8606229080
8606229279 8606229650 8606230067 8606230164 8606230303
8606231247 8606231253 8606231526 8606231670 8606231951
8606232014 8606232100 8606232185 8606232612 8606232989
8606233090 8606233159 8606233164 8606233305 8606233338
8606233710 8606234349 8606234706 8606235627 8606235632
8606236775 8606237266 8606237497 8606237606 8606237642
8606238197 8606238262 8606238295 8606238905 8606241187
8606241192 8606241607 8606242226 8606242679 8606243224
8606244219 8606244252 8606244728 8606244958 8606245325
8606245531 8606245578 8606245677 8606245903 8606245905
8606245906 8606245919 8606245952 8606245971 8606246169
8606246301 8606246302 8606246303 8606246305 8606246307
8606246309 8606246311 8606246314 8606246321 8606246398
8606246561 8606246676 8606247242 8606247262 8606247288
8606247339 8606247484 8606247490 8606247508 8606247520
8606247710 8606247981 8606248122 8606248406 8606248669
8606248700 8606248840 8606249185 8606249286 8606249333
8606250244 8606250322 8606250432 8606250573 8606251530
8606253203 8606253673 8606254024 8606254214 8606254434
8606254542 8606254713 8606255551 8606255696 8606256314
8606256741 8606257154 8606257479 8606257716 8606259311
8606259405 8606260488 8606260502 8606261174 8606261179
8606261753 8606262294 8606262982 8606263157 8606264041
8606264643 8606265738 8606266690 8606267968 8606268603
8606269903 8606269955 8606269959 8606270028 8606270176
8606270270 8606270271 8606270296 8606270345 8606270780
8606271001 8606271039 8606271213 8606271235 8606271481
8606271514 8606272324 8606272765 8606272841 8606273232
8606273477 8606273939 8606274153 8606274354 8606274495
8606274682 8606274896 8606274959 8606275363 8606275364
8606277348 8606277697 8606277719 8606278124 8606278217
8606278885 8606279349 8606280253 8606280288 8606280331
8606281411 8606281503 8606281609 8606282322 8606282772
8606282965 8606283620 8606283791 8606283966 8606284145
8606284220 8606284221 8606284791 8606284793 8606284907
8606285053 8606285107 8606285124 8606285262 8606285339
8606285427 8606286419 8606286667 8606286932 8606287019
8606287020 8606287149 8606287230 8606287775 8606289048
8606289327 8606290053 8606290056 8606290246 8606290848
8606290936 8606291052 8606291150 8606291731 8606291918
8606291920 8606292278 8606292583 8606293340 8606293457
8606294977 8606295065 8606295188 8606295191 8606295406
8606295895 8606296535 8606296566 8606297439 8606297454
8606298146 8606298646 8606298669 8606298969 8606298990
8606299662 8606299663 8606299672 8606300304 8606300305
8606300306 8606300307 8606300308 8606300948 8606301096
8606302024 8606302089 8606302874 8606302876 8606302878
8606304572 8606305933 8606306041 8606306287 8606306288
8606306299 8606306857 8606307734 8606308464 8606310433
8606310867 8606311142 8606311531 8606312460 8606312501
8606313270 8606313678 8606317300 8606317596 8606317678
8606318033 8606319251 8606320479 8606322230 8606322708
8606322985 8606323173 8606324000 8606327213 8606328244
8606328253 8606331109 8606331382 8606332047 8606333354
8606333610 8606333990 8606337569 8606338847 8606339858
8606340920 8606342795 8606343225 8606344448 8606344720
8606346523 8606348091 8606348486 8606349322 8606350213
8606351614 8606351765 8606353629 8606354852 8606355258
8606355392 8606355807 8606356802 8606358089 8606361264
8606363737 8606366227 8606366388 8606366614 8606369254
8606369649 8606370089 8606370407 8606370902 8606371947
8606372704 8606373330 8606374624 8606374761 8606389338
8606397158 8606399900 8606402124 8606403047
Patch Files:
Networking.NET-KRN,fr=B.11.11,fa=HP-UX_B.11.11_32/64,v=HP:
/usr/conf/net/if6.h
/usr/conf/net/netmp.h
/usr/conf/net/route6.h
/usr/conf/netinet/icmp6.h
/usr/conf/netinet/if_ether.h
/usr/conf/netinet/in.h
/usr/conf/netinet/in6.h
/usr/conf/netinet/ip6.h
/usr/conf/netinet/tcp.h
/usr/conf/netinet/tcpip.h
Networking.NET-PRG,fr=B.11.11,fa=HP-UX_B.11.11_32/64,v=HP:
/usr/include/net/if6.h
/usr/include/net/netmp.h
/usr/include/net/route6.h
/usr/include/netinet/icmp6.h
/usr/include/netinet/if_ether.h
/usr/include/netinet/in.h
/usr/include/netinet/in6.h
/usr/include/netinet/ip6.h
/usr/include/netinet/tcp.h
/usr/include/netinet/tcpip.h
Networking.NET-RUN,fr=B.11.11,fa=HP-UX_B.11.11_32/64,v=HP:
/usr/lib/nls/msg/C/ifconfig.cat
/usr/lib/nls/msg/C/ndd.cat
/usr/lib/nls/msg/C/ping.cat
/usr/lib/nls/msg/C/route.cat
/usr/lib/nls/msg/C/traceroute.cat
/usr/sbin/arp
/usr/sbin/ifconfig
/usr/lib/libsnet.1
/usr/lib/libsnet.a
/usr/bin/ndd
/usr/sbin/ping
/usr/sbin/route
/usr/contrib/bin/traceroute
/usr/contrib/man/man1.Z/traceroute.1
/usr/lib/libnm.1
/sbin/init.d/net
Networking.NET-RUN-64,fr=B.11.11,fa=HP-UX_B.11.11_32/64,
v=HP:
/usr/lib/pa20_64/libnm.1
Networking.NW-ENG-A-MAN,fr=B.11.11,fa=HP-UX_B.11.11_32/64,
v=HP:
/usr/share/man/man1.Z/netstat.1
/usr/share/man/man1m.Z/arp.1m
/usr/share/man/man1m.Z/ifconfig.1m
/usr/share/man/man1m.Z/ping.1m
/usr/share/man/man1m.Z/route.1m
/usr/share/man/man2.Z/accept.2
/usr/share/man/man2.Z/bind.2
/usr/share/man/man2.Z/connect.2
/usr/share/man/man2.Z/getpeername.2
/usr/share/man/man2.Z/getsockname.2
/usr/share/man/man2.Z/getsockopt.2
/usr/share/man/man2.Z/listen.2
/usr/share/man/man2.Z/recv.2
/usr/share/man/man2.Z/send.2
/usr/share/man/man2.Z/shutdown.2
/usr/share/man/man2.Z/socket.2
/usr/share/man/man2.Z/socketpair.2
/usr/share/man/man7.Z/tcp.7p
/usr/share/man/man7.Z/udp.7p
/usr/share/man/man7.Z/xopen_networking.7
OS-Core.CORE-KRN,fr=B.11.11,fa=HP-UX_B.11.11_32/64,v=HP:
/usr/conf/sys/mib.h
/usr/conf/sys/socket.h
/usr/conf/sys/socket_hatcp.h
/usr/conf/sys/socketvar.h
/usr/conf/sys/unpcb.h
ProgSupport.C-INC,fr=B.11.11,fa=HP-UX_B.11.11_32/64,v=HP:
/usr/include/sys/mib.h
/usr/include/sys/socket.h
/usr/include/sys/socket_hatcp.h
/usr/include/sys/socketvar.h
/usr/include/sys/unpcb.h
Networking.NET2-KRN,fr=B.11.11,fa=HP-UX_B.11.11_32,v=HP:
/usr/conf/lib/libipsectp.a
/usr/conf/lib/libnet.a(af.o)
/usr/conf/lib/libinet.a(arp.o)
/usr/conf/lib/libnet.a(if.o)
/usr/conf/lib/libinet.a(igmp.o)
/usr/conf/lib/libinet.a(in_proto.o)
/usr/conf/lib/libinet.a(inet_init.o)
/usr/conf/lib/libinet.a(ip.o)
/usr/conf/lib/libinet.a(ip6.o)
/usr/conf/lib/libinet.a(ip6_if.o)
/usr/conf/lib/libinet.a(ip6_ire.o)
/usr/conf/lib/libinet.a(ip6_nd.o)
/usr/conf/lib/libinet.a(ip_if.o)
/usr/conf/lib/libnet.a(ip_info.o)
/usr/conf/lib/libinet.a(ip_ire.o)
/usr/conf/lib/libinet.a(ip_mroute.o)
/usr/conf/lib/libinet.a(ip_multi.o)
/usr/conf/lib/libinet.a(ipcomm.o)
/usr/conf/lib/libinet.a(ipcsum.o)
/usr/conf/lib/libnet.a(mialloc.o)
/usr/conf/lib/libnet.a(mind.o)
/usr/conf/lib/libnet.a(mioptcom.o)
/usr/conf/lib/libnet.a(mitimer.o)
/usr/conf/lib/libnet.a(miutil.o)
/usr/conf/lib/libnet.a(net_init.o)
/usr/conf/lib/libnet.a(nuls.o)
/usr/conf/lib/libnet.a(pass.o)
/usr/conf/lib/libinet.a(rawip.o)
/usr/conf/lib/libinet.a(rawip6.o)
/usr/conf/lib/libnet.a(route.o)
/usr/conf/lib/libinet.a(snmpcom.o)
/usr/conf/lib/libnet.a(spass.o)
/usr/conf/lib/libnet.a(stcpmap.o)
/usr/conf/lib/libnet.a(tang.o)
/usr/conf/lib/libinet.a(tcp.o)
/usr/conf/lib/libinet.a(udp.o)
Networking.NMS2-KRN,fr=B.11.11,fa=HP-UX_B.11.11_32,v=HP:
/usr/conf/lib/libnms.a(nms.o)
OS-Core.CORE2-KRN,fr=B.11.11,fa=HP-UX_B.11.11_32,v=HP:
/usr/conf/lib/libaudit.a(dgram_aud.o)
/usr/conf/lib/libuipc.a(sys_socket.o)
/usr/conf/lib/libuipc.a(uipc_domain.o)
/usr/conf/lib/libuipc.a(uipc_init.o)
/usr/conf/lib/libuipc.a(uipc_proto.o)
/usr/conf/lib/libuipc.a(uipc_socket.o)
/usr/conf/lib/libuipc.a(uipc_socket2.o)
/usr/conf/lib/libuipc.a(uipc_syscall.o)
/usr/conf/lib/libuipc.a(uipc_tpi.o)
/usr/conf/lib/libuipc.a(uipc_usrreq.o)
/usr/conf/lib/libdebug.a(xport_diag.o)
OS-Core.SYS-ADMIN,fr=B.11.11,fa=HP-UX_B.11.11_32,v=HP:
/usr/lib/nls/msg/C/netstat.cat
/usr/bin/netstat
Networking.NET2-KRN,fr=B.11.11,fa=HP-UX_B.11.11_64,v=HP:
/usr/conf/lib/libipsectp.a
/usr/conf/lib/libnet.a(af.o)
/usr/conf/lib/libinet.a(arp.o)
/usr/conf/lib/libnet.a(if.o)
/usr/conf/lib/libinet.a(igmp.o)
/usr/conf/lib/libinet.a(in_proto.o)
/usr/conf/lib/libinet.a(inet_init.o)
/usr/conf/lib/libinet.a(ip.o)
/usr/conf/lib/libinet.a(ip6.o)
/usr/conf/lib/libinet.a(ip6_if.o)
/usr/conf/lib/libinet.a(ip6_ire.o)
/usr/conf/lib/libinet.a(ip6_nd.o)
/usr/conf/lib/libinet.a(ip_if.o)
/usr/conf/lib/libnet.a(ip_info.o)
/usr/conf/lib/libinet.a(ip_ire.o)
/usr/conf/lib/libinet.a(ip_mroute.o)
/usr/conf/lib/libinet.a(ip_multi.o)
/usr/conf/lib/libinet.a(ipcomm.o)
/usr/conf/lib/libinet.a(ipcsum.o)
/usr/conf/lib/libnet.a(mialloc.o)
/usr/conf/lib/libnet.a(mind.o)
/usr/conf/lib/libnet.a(mioptcom.o)
/usr/conf/lib/libnet.a(mitimer.o)
/usr/conf/lib/libnet.a(miutil.o)
/usr/conf/lib/libnet.a(net_init.o)
/usr/conf/lib/libnet.a(nuls.o)
/usr/conf/lib/libnet.a(pass.o)
/usr/conf/lib/libinet.a(rawip.o)
/usr/conf/lib/libinet.a(rawip6.o)
/usr/conf/lib/libnet.a(route.o)
/usr/conf/lib/libinet.a(snmpcom.o)
/usr/conf/lib/libnet.a(spass.o)
/usr/conf/lib/libnet.a(stcpmap.o)
/usr/conf/lib/libnet.a(tang.o)
/usr/conf/lib/libinet.a(tcp.o)
/usr/conf/lib/libinet.a(udp.o)
Networking.NMS2-KRN,fr=B.11.11,fa=HP-UX_B.11.11_64,v=HP:
/usr/conf/lib/libnms.a(nms.o)
OS-Core.CORE2-KRN,fr=B.11.11,fa=HP-UX_B.11.11_64,v=HP:
/usr/conf/lib/libaudit.a(dgram_aud.o)
/usr/conf/lib/libuipc.a(sys_socket.o)
/usr/conf/lib/libuipc.a(uipc_domain.o)
/usr/conf/lib/libuipc.a(uipc_init.o)
/usr/conf/lib/libuipc.a(uipc_proto.o)
/usr/conf/lib/libuipc.a(uipc_socket.o)
/usr/conf/lib/libuipc.a(uipc_socket2.o)
/usr/conf/lib/libuipc.a(uipc_syscall.o)
/usr/conf/lib/libuipc.a(uipc_tpi.o)
/usr/conf/lib/libuipc.a(uipc_usrreq.o)
/usr/conf/lib/libdebug.a(xport_diag.o)
OS-Core.SYS-ADMIN,fr=B.11.11,fa=HP-UX_B.11.11_64,v=HP:
/usr/lib/nls/msg/C/netstat.cat
/usr/bin/netstat
what(1) Output:
Networking.NET-KRN,fr=B.11.11,fa=HP-UX_B.11.11_32/64,v=HP:
/usr/conf/net/if6.h:
if6.h $Date: 2004/02/12 04:13:04 $Revision: r11.11/8
PATCH_11.11 (PHNE_30053) */
common/net/if6.h $Revision: $
/usr/conf/net/netmp.h:
netmp.h $Date: 2001/02/02 16:37:16 $Revision: r11.11
/1 PATCH_11.11 (PHNE_23288)
/usr/conf/net/route6.h:
route6.h $Date: 2001/07/09 19:32:15 $Revision: r11.1
1/1 PATCH_11.11 (PHNE_24211) */
common/net/route6.h $Revision: $
/usr/conf/netinet/icmp6.h:
icmp6.h $Date: 2001/07/09 19:32:37 $Revision: r11.11
/1 PATCH_11.11 (PHNE_24211) */
common/netinet/icmp6.h $Revision: $
/usr/conf/netinet/if_ether.h:
if_ether.h $Date: 2001/07/09 19:33:00 $Revision: r11
.11/1 PATCH_11.11 (PHNE_24211) */
/usr/conf/netinet/in.h:
in.h $Date: 2001/04/16 11:28:15 $Revision: r11.11/1
PATCH_11.11 (PHNE_23645) */
/usr/conf/netinet/in6.h:
in6.h $Date: 2001/04/16 14:21:09 $Revision: r11.11/2
PATCH_11.11 (PHNE_23645) */
common/netinet/in6.h $Revision: $
/usr/conf/netinet/ip6.h:
ip6.h $Date: 2001/07/09 19:33:19 $Revision: r11.11/1
PATCH_11.11 (PHNE_24211) */
common/netinet/ip6.h $Revision: $
/usr/conf/netinet/tcp.h:
tcp.h $Date: 2001/05/14 11:53:11 $Revision: r11.11/2
PATCH_11.11 (PHNE_23645) */
tcp.h 7.5 (Berkeley) 6/29/88
/usr/conf/netinet/tcpip.h:
tcpip.h $Date: 2003/02/06 04:37:10 $Revision: r11.11
/2 PATCH_11.11 (PHNE_28497) */
tcpip.h 7.3 (Berkeley) 6/29/88
Networking.NET-PRG,fr=B.11.11,fa=HP-UX_B.11.11_32/64,v=HP:
/usr/include/net/if6.h:
if6.h $Date: 2004/02/12 04:13:04 $Revision: r11.11/8
PATCH_11.11 (PHNE_30053) */
common/net/if6.h $Revision: $
/usr/include/net/netmp.h:
netmp.h $Date: 2001/02/02 16:37:16 $Revision: r11.11
/1 PATCH_11.11 (PHNE_23288)
/usr/include/net/route6.h:
route6.h $Date: 2001/07/09 19:32:15 $Revision: r11.1
1/1 PATCH_11.11 (PHNE_24211) */
common/net/route6.h $Revision: $
/usr/include/netinet/icmp6.h:
icmp6.h $Date: 2001/07/09 19:32:37 $Revision: r11.11
/1 PATCH_11.11 (PHNE_24211) */
common/netinet/icmp6.h $Revision: $
/usr/include/netinet/if_ether.h:
if_ether.h $Date: 2001/07/09 19:33:00 $Revision: r11
.11/1 PATCH_11.11 (PHNE_24211) */
/usr/include/netinet/in.h:
in.h $Date: 2001/04/16 11:28:15 $Revision: r11.11/1
PATCH_11.11 (PHNE_23645) */
/usr/include/netinet/in6.h:
in6.h $Date: 2001/04/16 14:21:09 $Revision: r11.11/2
PATCH_11.11 (PHNE_23645) */
common/netinet/in6.h $Revision: $
/usr/include/netinet/ip6.h:
ip6.h $Date: 2001/07/09 19:33:19 $Revision: r11.11/1
PATCH_11.11 (PHNE_24211) */
common/netinet/ip6.h $Revision: $
/usr/include/netinet/tcp.h:
tcp.h $Date: 2001/05/14 11:53:11 $Revision: r11.11/2
PATCH_11.11 (PHNE_23645) */
tcp.h 7.5 (Berkeley) 6/29/88
/usr/include/netinet/tcpip.h:
tcpip.h $Date: 2003/02/06 04:37:10 $Revision: r11.11
/2 PATCH_11.11 (PHNE_28497) */
tcpip.h 7.3 (Berkeley) 6/29/88
Networking.NET-RUN,fr=B.11.11,fa=HP-UX_B.11.11_32/64,v=HP:
/usr/lib/nls/msg/C/ifconfig.cat:
ifconfig.msg $Date: 2001/07/23 11:39:28 $Revision: r
11.11/2 PATCH_11.11 (PHNE_24211)
/usr/lib/nls/msg/C/ndd.cat:
ndd.msg $Date: 2006/05/16 11:26:04 $Revision: r11.11
/17 PATCH_11.11 (PHNE_34672)IP:
/usr/lib/nls/msg/C/ping.cat:
ping.msg $Date: 2002/07/07 21:43:31 $Revision: r11.1
1/2 PATCH_11.11 (PHNE_27382)
/usr/lib/nls/msg/C/route.cat:
route.msg $Date: 2001/07/09 19:29:23 $Revision: r11.
11/1 PATCH_11.11 (PHNE_24211)
/usr/lib/nls/msg/C/traceroute.cat:
traceroute.msg $Date: 2001/07/09 19:29:23 $Revision:
r11.11/1 PATCH_11.11 (PHNE_24211)
/usr/sbin/arp:
arp.c $Date: 2004/02/12 04:11:37 $Revision: r11.11/3
PATCH_11.11 (PHNE_30053)
/usr/sbin/ifconfig:
ifconfig.c $Date: 2003/02/06 04:36:02 $Revision: r11
.11/8 PATCH_11.11 (PHNE_28497)
/usr/lib/libsnet.1:
plumb.c $Date: 2004/02/12 04:11:45 $Revision: r11.11
/7 PATCH_11.11 (PHNE_30053)
/usr/lib/libsnet.a:
plumb.c $Date: 2004/02/12 04:11:45 $Revision: r11.11
/7 PATCH_11.11 (PHNE_30053)
/usr/bin/ndd:
ndd.c $Date: 2003/11/10 05:49:39 $Revision: r11.11/8
PATCH_11.11 (PHNE_29887)
nddconfig.c $Date: 2006/05/16 11:26:11 $Revision: r1
1.11/20 PATCH_11.11 (PHNE_34672)
errmsg.c $Date: 2001/04/16 12:04:37 $Revision: r11.1
1/1 PATCH_11.11 (PHNE_23645)
fatal.c $Date: 2001/04/16 12:05:30 $Revision: r11.11
/1 PATCH_11.11 (PHNE_23645)
progname.c $Date: 2001/04/16 12:08:08 $Revision: r11
.11/1 PATCH_11.11 (PHNE_23645)
usage.c $Date: 2001/04/16 12:08:27 $Revision: r11.11
/1 PATCH_11.11 (PHNE_23645)
warn.c $Date: 2001/04/16 12:08:48 $Revision: r11.11/
1 PATCH_11.11 (PHNE_23645)
/usr/sbin/ping:
ping.c $Date: 2005/08/17 08:11:13 $Revision: r11.11/
6 PATCH_11.11 (PHNE_33628)
ping.c $Revision: 1.8.119.5 $ $Date: 98/07/23 14:27:
01 $
/usr/sbin/route:
route.c $Date: 2002/01/23 16:29:22 $Revision: r11.11
/2 PATCH_11.11 (PHNE_25644)
/usr/contrib/bin/traceroute:
traceroute.c $Date: 2004/12/09 22:50:35 $Revision: r
11.11/3 PATCH_11.11 (PHNE_32042)
ifaddrlist.c $Date: 2004/02/12 04:12:19 $Revision: r
11.11/2 PATCH_11.11 (PHNE_30053)
v6trace.c $Date: 2004/02/12 04:12:45 $Revision: r11.
11/4 PATCH_11.11 (PHNE_30053)
$Header: traceroute.c,v 1.49 97/06/13 02:30:23 leres
Exp $ (LBL)
Copyright (c) 1988, 1989, 1991, 1994, 1995, 1996, 19
97
$Header: ifaddrlist.c,v 1.2 97/04/22 13:31:05 leres
Exp $ (LBL)
/usr/contrib/man/man1.Z/traceroute.1:
None
/usr/lib/libnm.1:
mimib.c $Date: 2003/03/29 01:31:26 $Revision: r11.11
/5 PATCH_11.11 (PHNE_28895)
nmapi.c $Date: 2002/04/25 15:53:50 $Revision: r11.11
/5 PATCH_11.11 (PHNE_26762)
/sbin/init.d/net:
net $Date: 2001/06/14 15:09:22 $Revision: cup_rar_r1
1.11_xport_pib/6 PATCH_11.11 (PHNE_23645)
Networking.NET-RUN-64,fr=B.11.11,fa=HP-UX_B.11.11_32/64,
v=HP:
/usr/lib/pa20_64/libnm.1:
mimib.c $Date: 2003/03/29 01:31:26 $Revision: r11.11
/5 PATCH_11.11 (PHNE_28895)
nmapi.c $Date: 2002/04/25 15:53:50 $Revision: r11.11
/5 PATCH_11.11 (PHNE_26762)
$Revision: R11.11_BL2005_0916_5
Networking.NW-ENG-A-MAN,fr=B.11.11,fa=HP-UX_B.11.11_32/64,
v=HP:
/usr/share/man/man1.Z/netstat.1:
None
/usr/share/man/man1m.Z/arp.1m:
None
/usr/share/man/man1m.Z/ifconfig.1m:
None
/usr/share/man/man1m.Z/ping.1m:
None
/usr/share/man/man1m.Z/route.1m:
None
/usr/share/man/man2.Z/accept.2:
None
/usr/share/man/man2.Z/bind.2:
None
/usr/share/man/man2.Z/connect.2:
None
/usr/share/man/man2.Z/getpeername.2:
None
/usr/share/man/man2.Z/getsockname.2:
None
/usr/share/man/man2.Z/getsockopt.2:
None
/usr/share/man/man2.Z/listen.2:
None
/usr/share/man/man2.Z/recv.2:
None
/usr/share/man/man2.Z/send.2:
None
/usr/share/man/man2.Z/shutdown.2:
None
/usr/share/man/man2.Z/socket.2:
None
/usr/share/man/man2.Z/socketpair.2:
None
/usr/share/man/man7.Z/tcp.7p:
None
/usr/share/man/man7.Z/udp.7p:
None
/usr/share/man/man7.Z/xopen_networking.7:
None
OS-Core.CORE-KRN,fr=B.11.11,fa=HP-UX_B.11.11_32/64,v=HP:
/usr/conf/sys/mib.h:
mib.h $Date: 2001/07/09 19:34:00 $Revision: r11.11/2
PATCH_11.11 (PHNE_24211) */
/usr/conf/sys/socket.h:
socket.h $Date: 2002/05/17 16:33:35 $Revision: r11.1
1/4 PATCH_11.11 (PHNE_27063) */
/usr/conf/sys/socket_hatcp.h:
socket_hatcp.h $Date: 2001/10/29 18:26:21 $Revision:
r11.11/2 PATCH_11.11 (PHNE_24506) */
/usr/conf/sys/socketvar.h:
socketvar.h $Date: 2004/12/09 22:50:44 $Revision: r1
1.11/5 PATCH_11.11 (PHNE_32042) */
/usr/conf/sys/unpcb.h:
unpcb.h $Date: 2004/12/09 22:50:53 $Revision: r11.11
/4 PATCH_11.11 (PHNE_32042) */
ProgSupport.C-INC,fr=B.11.11,fa=HP-UX_B.11.11_32/64,v=HP:
/usr/include/sys/mib.h:
mib.h $Date: 2001/07/09 19:34:00 $Revision: r11.11/2
PATCH_11.11 (PHNE_24211) */
/usr/include/sys/socket.h:
socket.h $Date: 2002/05/17 16:33:35 $Revision: r11.1
1/4 PATCH_11.11 (PHNE_27063) */
/usr/include/sys/socket_hatcp.h:
socket_hatcp.h $Date: 2001/10/29 18:26:21 $Revision:
r11.11/2 PATCH_11.11 (PHNE_24506) */
/usr/include/sys/socketvar.h:
socketvar.h $Date: 2004/12/09 22:50:44 $Revision: r1
1.11/5 PATCH_11.11 (PHNE_32042) */
/usr/include/sys/unpcb.h:
unpcb.h $Date: 2004/12/09 22:50:53 $Revision: r11.11
/4 PATCH_11.11 (PHNE_32042) */
Networking.NET2-KRN,fr=B.11.11,fa=HP-UX_B.11.11_32,v=HP:
/usr/conf/lib/libipsectp.a:
ipsec_hp.c $Date: 2001/07/09 19:23:36 $Revision: r11
.11/2 PATCH_11.11 (PHNE_24211)
/usr/conf/lib/libnet.a(af.o):
af.c $Date: 2005/05/16 02:31:45 $Revision: r11.11/10
PATCH_11.11 (PHNE_33159)
/usr/conf/lib/libinet.a(arp.o):
arp.c $Date: 2005/08/17 08:09:13 $Revision: r11.11/1
4 PATCH_11.11 (PHNE_33628)
/usr/conf/lib/libnet.a(if.o):
if.c $Date: 2005/05/16 02:31:54 $Revision: r11.11/16
PATCH_11.11 (PHNE_33159)
/usr/conf/lib/libinet.a(igmp.o):
igmp.c $Date: 2005/05/16 02:32:57 $Revision: r11.11/
14 PATCH_11.11 (PHNE_33159)
/usr/conf/lib/libinet.a(in_proto.o):
in_proto.c $Date: 2005/05/16 02:33:06 $Revision: r11
.11/10 PATCH_11.11 (PHNE_33159)
/usr/conf/lib/libinet.a(inet_init.o):
inet_init.c $Date: 2005/05/16 02:33:15 $Revision: r1
1.11/10 PATCH_11.11 (PHNE_33159)
/usr/conf/lib/libinet.a(ip.o):
ip.c $Date: 2006/05/16 11:25:11 $Revision: r11.11/27
PATCH_11.11 (PHNE_34672)
/usr/conf/lib/libinet.a(ip6.o):
ip6.c $Date: 2005/08/17 08:09:29 $Revision: r11.11/2
0 PATCH_11.11 (PHNE_33628)
/usr/conf/lib/libinet.a(ip6_if.o):
ip6_if.c $Date: 2006/05/16 11:25:18 $Revision: r11.1
1/21 PATCH_11.11 (PHNE_34672)
/usr/conf/lib/libinet.a(ip6_ire.o):
ip6_ire.c $Date: 2006/05/16 11:25:25 $Revision: r11.
11/19 PATCH_11.11 (PHNE_34672)
/usr/conf/lib/libinet.a(ip6_nd.o):
ip6_nd.c $Date: 2005/05/16 02:33:59 $Revision: r11.1
1/19 PATCH_11.11 (PHNE_33159)
/usr/conf/lib/libinet.a(ip_if.o):
ip_if.c $Date: 2006/05/16 11:25:31 $Revision: r11.11
/20 PATCH_11.11 (PHNE_34672)
/usr/conf/lib/libnet.a(ip_info.o):
ip_info.c $Date: 2005/05/16 02:32:03 $Revision: r11.
11/14 PATCH_11.11 (PHNE_33159)
/usr/conf/lib/libinet.a(ip_ire.o):
ip_ire.c $Date: 2006/06/01 06:29:27 $Revision: r11.1
1/22 PATCH_11.11 (PHNE_34672)
/usr/conf/lib/libinet.a(ip_mroute.o):
ip_mroute.c $Date: 2005/05/16 02:34:25 $Revision: r1
1.11/13 PATCH_11.11 (PHNE_33159)
/usr/conf/lib/libinet.a(ip_multi.o):
ip_multi.c $Date: 2005/05/16 02:34:34 $Revision: r11
.11/13 PATCH_11.11 (PHNE_33159)
/usr/conf/lib/libinet.a(ipcomm.o):
ipcomm.c $Date: 2005/08/17 08:10:11 $Revision: r11.1
1/19 PATCH_11.11 (PHNE_33628)
/usr/conf/lib/libinet.a(ipcsum.o):
ipcsum.c $Date: 2001/10/29 18:21:09 $Revision: r11.1
1/3 PATCH_11.11 (PHNE_24506)
/usr/conf/lib/libnet.a(mialloc.o):
mialloc.c $Date: 2001/10/29 18:15:38 $Revision: r11.
11/3 PATCH_11.11 (PHNE_24506)
/usr/conf/lib/libnet.a(mind.o):
mind.c $Date: 2001/10/29 18:15:51 $Revision: r11.11/
3 PATCH_11.11 (PHNE_24506)
/usr/conf/lib/libnet.a(mioptcom.o):
mioptcom.c $Date: 2003/03/29 01:22:19 $Revision: r11
.11/6 PATCH_11.11 (PHNE_28895)
/usr/conf/lib/libnet.a(mitimer.o):
mitimer.c $Date: 2005/08/17 08:08:58 $Revision: r11.
11/5 PATCH_11.11 (PHNE_33628)
/usr/conf/lib/libnet.a(miutil.o):
miutil.c $Date: 2003/03/29 01:22:44 $Revision: r11.1
1/5 PATCH_11.11 (PHNE_28895)
/usr/conf/lib/libnet.a(net_init.o):
net_init.c $Date: 2005/05/16 02:32:12 $Revision: r11
.11/10 PATCH_11.11 (PHNE_33159)
/usr/conf/lib/libnet.a(nuls.o):
nuls.c $Date: 2001/10/29 18:16:54 $Revision: r11.11/
3 PATCH_11.11 (PHNE_24506)
/usr/conf/lib/libnet.a(pass.o):
pass.c $Date: 2001/10/29 18:17:07 $Revision: r11.11/
3 PATCH_11.11 (PHNE_24506)
/usr/conf/lib/libinet.a(rawip.o):
rawip.c $Date: 2005/05/16 02:34:52 $Revision: r11.11
/15 PATCH_11.11 (PHNE_33159)
/usr/conf/lib/libinet.a(rawip6.o):
rawip6.c $Date: 2005/05/16 02:35:00 $Revision: r11.1
1/17 PATCH_11.11 (PHNE_33159)
/usr/conf/lib/libnet.a(route.o):
route.c $Date: 2005/05/16 02:32:21 $Revision: r11.11
/14 PATCH_11.11 (PHNE_33159)
/usr/conf/lib/libinet.a(snmpcom.o):
snmpcom.c $Date: 2003/03/29 01:28:01 $Revision: r11.
11/5 PATCH_11.11 (PHNE_28895)
/usr/conf/lib/libnet.a(spass.o):
spass.c $Date: 2001/10/29 18:17:32 $Revision: r11.11
/3 PATCH_11.11 (PHNE_24506)
/usr/conf/lib/libnet.a(stcpmap.o):
stcpmap.c $Date: 2006/05/16 11:25:04 $Revision: r11.
11/18 PATCH_11.11 (PHNE_34672)
/usr/conf/lib/libnet.a(tang.o):
tang.c $Date: 2005/05/16 02:32:40 $Revision: r11.11/
11 PATCH_11.11 (PHNE_33159)
/usr/conf/lib/libinet.a(tcp.o):
tcp.c $Date: 2006/05/16 11:25:45 $Revision: r11.11/2
8 PATCH_11.11 (PHNE_34672)
/usr/conf/lib/libinet.a(udp.o):
udp.c $Date: 2006/05/16 11:25:51 $Revision: r11.11/2
0 PATCH_11.11 (PHNE_34672)
Networking.NMS2-KRN,fr=B.11.11,fa=HP-UX_B.11.11_32,v=HP:
/usr/conf/lib/libnms.a(nms.o):
nms.c $Date: 2005/05/16 02:35:27 $Revision: r11.11/1
0 PATCH_11.11 (PHNE_33159)
OS-Core.CORE2-KRN,fr=B.11.11,fa=HP-UX_B.11.11_32,v=HP:
/usr/conf/lib/libaudit.a(dgram_aud.o):
dgram_aud.c $Date: 2005/05/16 02:31:27 $Revision: r1
1.11/10 PATCH_11.11 (PHNE_33159)
/usr/conf/lib/libuipc.a(sys_socket.o):
sys_socket.c $Date: 2006/03/09 23:16:02 $Revision: r
11.11/16 PATCH_11.11 (PHNE_34135)
/usr/conf/lib/libuipc.a(uipc_domain.o):
uipc_domain.c $Date: 2005/05/16 02:35:44 $Revision:
r11.11/10 PATCH_11.11 (PHNE_33159)
/usr/conf/lib/libuipc.a(uipc_init.o):
uipc_init.c $Date: 2005/05/16 02:35:53 $Revision: r1
1.11/10 PATCH_11.11 (PHNE_33159)
/usr/conf/lib/libuipc.a(uipc_proto.o):
uipc_proto.c $Date: 2005/05/16 02:36:02 $Revision: r
11.11/10 PATCH_11.11 (PHNE_33159)
/usr/conf/lib/libuipc.a(uipc_socket.o):
uipc_socket.c $Date: 2006/05/16 11:25:58 $Revision:
r11.11/32 PATCH_11.11 (PHNE_34672)
/usr/conf/lib/libuipc.a(uipc_socket2.o):
uipc_socket2.c $Date: 2002/02/27 (PHNE_25644 + JAGae
07926)
uipc_socket2.c $Date: 2005/05/16 02:36:19 $Revision:
r11.11/15 PATCH_11.11 (PHNE_33159)
/usr/conf/lib/libuipc.a(uipc_syscall.o):
uipc_syscall.c $Date: 2005/08/17 08:10:47 $Revision:
r11.11/15 PATCH_11.11 (PHNE_33628)
/usr/conf/lib/libuipc.a(uipc_tpi.o):
uipc_tpi.c $Date: 2005/08/17 08:10:52 $Revision: r11
.11/17 PATCH_11.11 (PHNE_33628)
/usr/conf/lib/libuipc.a(uipc_usrreq.o):
uipc_usrreq.c $Date: 2006/03/09 23:16:09 $Revision:
r11.11/14 PATCH_11.11 (PHNE_34135)
/usr/conf/lib/libdebug.a(xport_diag.o):
xport_diag.c $Date: 2005/05/16 02:31:36 $Revision: r
11.11/15 PATCH_11.11 (PHNE_33159)
OS-Core.SYS-ADMIN,fr=B.11.11,fa=HP-UX_B.11.11_32,v=HP:
/usr/lib/nls/msg/C/netstat.cat:
netstat.msg $Date: 2003/02/06 04:36:46 $Revision: r1
1.11/3 PATCH_11.11 (PHNE_28497)
/usr/bin/netstat:
netstat.c $Date: 2003/02/06 04:36:35 $Revision: r11.
11/8 PATCH_11.11 (PHNE_28497)
osstat.c $Date: 2004/12/09 22:50:18 $Revision: r11.1
1/6 PATCH_11.11 (PHNE_32042)
Networking.NET2-KRN,fr=B.11.11,fa=HP-UX_B.11.11_64,v=HP:
/usr/conf/lib/libipsectp.a:
ipsec_hp.c $Date: 2001/07/09 19:23:36 $Revision: r11
.11/2 PATCH_11.11 (PHNE_24211)
/usr/conf/lib/libnet.a(af.o):
af.c $Date: 2005/05/16 02:31:45 $Revision: r11.11/10
PATCH_11.11 (PHNE_33159)
/usr/conf/lib/libinet.a(arp.o):
arp.c $Date: 2005/08/17 08:09:13 $Revision: r11.11/1
4 PATCH_11.11 (PHNE_33628)
/usr/conf/lib/libnet.a(if.o):
if.c $Date: 2005/05/16 02:31:54 $Revision: r11.11/16
PATCH_11.11 (PHNE_33159)
/usr/conf/lib/libinet.a(igmp.o):
igmp.c $Date: 2005/05/16 02:32:57 $Revision: r11.11/
14 PATCH_11.11 (PHNE_33159)
/usr/conf/lib/libinet.a(in_proto.o):
in_proto.c $Date: 2005/05/16 02:33:06 $Revision: r11
.11/10 PATCH_11.11 (PHNE_33159)
/usr/conf/lib/libinet.a(inet_init.o):
inet_init.c $Date: 2005/05/16 02:33:15 $Revision: r1
1.11/10 PATCH_11.11 (PHNE_33159)
/usr/conf/lib/libinet.a(ip.o):
ip.c $Date: 2006/05/16 11:25:11 $Revision: r11.11/27
PATCH_11.11 (PHNE_34672)
/usr/conf/lib/libinet.a(ip6.o):
ip6.c $Date: 2005/08/17 08:09:29 $Revision: r11.11/2
0 PATCH_11.11 (PHNE_33628)
/usr/conf/lib/libinet.a(ip6_if.o):
ip6_if.c $Date: 2006/05/16 11:25:18 $Revision: r11.1
1/21 PATCH_11.11 (PHNE_34672)
/usr/conf/lib/libinet.a(ip6_ire.o):
ip6_ire.c $Date: 2006/05/16 11:25:25 $Revision: r11.
11/19 PATCH_11.11 (PHNE_34672)
/usr/conf/lib/libinet.a(ip6_nd.o):
ip6_nd.c $Date: 2005/05/16 02:33:59 $Revision: r11.1
1/19 PATCH_11.11 (PHNE_33159)
/usr/conf/lib/libinet.a(ip_if.o):
ip_if.c $Date: 2006/05/16 11:25:31 $Revision: r11.11
/20 PATCH_11.11 (PHNE_34672)
/usr/conf/lib/libnet.a(ip_info.o):
ip_info.c $Date: 2005/05/16 02:32:03 $Revision: r11.
11/14 PATCH_11.11 (PHNE_33159)
/usr/conf/lib/libinet.a(ip_ire.o):
ip_ire.c $Date: 2006/06/01 06:29:27 $Revision: r11.1
1/22 PATCH_11.11 (PHNE_34672)
/usr/conf/lib/libinet.a(ip_mroute.o):
ip_mroute.c $Date: 2005/05/16 02:34:25 $Revision: r1
1.11/13 PATCH_11.11 (PHNE_33159)
/usr/conf/lib/libinet.a(ip_multi.o):
ip_multi.c $Date: 2005/05/16 02:34:34 $Revision: r11
.11/13 PATCH_11.11 (PHNE_33159)
/usr/conf/lib/libinet.a(ipcomm.o):
ipcomm.c $Date: 2005/08/17 08:10:11 $Revision: r11.1
1/19 PATCH_11.11 (PHNE_33628)
/usr/conf/lib/libinet.a(ipcsum.o):
ipcsum.c $Date: 2001/10/29 18:21:09 $Revision: r11.1
1/3 PATCH_11.11 (PHNE_24506)
/usr/conf/lib/libnet.a(mialloc.o):
mialloc.c $Date: 2001/10/29 18:15:38 $Revision: r11.
11/3 PATCH_11.11 (PHNE_24506)
/usr/conf/lib/libnet.a(mind.o):
mind.c $Date: 2001/10/29 18:15:51 $Revision: r11.11/
3 PATCH_11.11 (PHNE_24506)
/usr/conf/lib/libnet.a(mioptcom.o):
mioptcom.c $Date: 2003/03/29 01:22:19 $Revision: r11
.11/6 PATCH_11.11 (PHNE_28895)
/usr/conf/lib/libnet.a(mitimer.o):
mitimer.c $Date: 2005/08/17 08:08:58 $Revision: r11.
11/5 PATCH_11.11 (PHNE_33628)
/usr/conf/lib/libnet.a(miutil.o):
miutil.c $Date: 2003/03/29 01:22:44 $Revision: r11.1
1/5 PATCH_11.11 (PHNE_28895)
/usr/conf/lib/libnet.a(net_init.o):
net_init.c $Date: 2005/05/16 02:32:12 $Revision: r11
.11/10 PATCH_11.11 (PHNE_33159)
/usr/conf/lib/libnet.a(nuls.o):
nuls.c $Date: 2001/10/29 18:16:54 $Revision: r11.11/
3 PATCH_11.11 (PHNE_24506)
/usr/conf/lib/libnet.a(pass.o):
pass.c $Date: 2001/10/29 18:17:07 $Revision: r11.11/
3 PATCH_11.11 (PHNE_24506)
/usr/conf/lib/libinet.a(rawip.o):
rawip.c $Date: 2005/05/16 02:34:52 $Revision: r11.11
/15 PATCH_11.11 (PHNE_33159)
/usr/conf/lib/libinet.a(rawip6.o):
rawip6.c $Date: 2005/05/16 02:35:00 $Revision: r11.1
1/17 PATCH_11.11 (PHNE_33159)
/usr/conf/lib/libnet.a(route.o):
route.c $Date: 2005/05/16 02:32:21 $Revision: r11.11
/14 PATCH_11.11 (PHNE_33159)
/usr/conf/lib/libinet.a(snmpcom.o):
snmpcom.c $Date: 2003/03/29 01:28:01 $Revision: r11.
11/5 PATCH_11.11 (PHNE_28895)
/usr/conf/lib/libnet.a(spass.o):
spass.c $Date: 2001/10/29 18:17:32 $Revision: r11.11
/3 PATCH_11.11 (PHNE_24506)
/usr/conf/lib/libnet.a(stcpmap.o):
stcpmap.c $Date: 2006/05/16 11:25:04 $Revision: r11.
11/18 PATCH_11.11 (PHNE_34672)
/usr/conf/lib/libnet.a(tang.o):
tang.c $Date: 2005/05/16 02:32:40 $Revision: r11.11/
11 PATCH_11.11 (PHNE_33159)
/usr/conf/lib/libinet.a(tcp.o):
tcp.c $Date: 2006/05/16 11:25:45 $Revision: r11.11/2
8 PATCH_11.11 (PHNE_34672)
/usr/conf/lib/libinet.a(udp.o):
udp.c $Date: 2006/05/16 11:25:51 $Revision: r11.11/2
0 PATCH_11.11 (PHNE_34672)
Networking.NMS2-KRN,fr=B.11.11,fa=HP-UX_B.11.11_64,v=HP:
/usr/conf/lib/libnms.a(nms.o):
nms.c $Date: 2005/05/16 02:35:27 $Revision: r11.11/1
0 PATCH_11.11 (PHNE_33159)
OS-Core.CORE2-KRN,fr=B.11.11,fa=HP-UX_B.11.11_64,v=HP:
/usr/conf/lib/libaudit.a(dgram_aud.o):
dgram_aud.c $Date: 2005/05/16 02:31:27 $Revision: r1
1.11/10 PATCH_11.11 (PHNE_33159)
/usr/conf/lib/libuipc.a(sys_socket.o):
sys_socket.c $Date: 2006/03/09 23:16:02 $Revision: r
11.11/16 PATCH_11.11 (PHNE_34135)
/usr/conf/lib/libuipc.a(uipc_domain.o):
uipc_domain.c $Date: 2005/05/16 02:35:44 $Revision:
r11.11/10 PATCH_11.11 (PHNE_33159)
/usr/conf/lib/libuipc.a(uipc_init.o):
uipc_init.c $Date: 2005/05/16 02:35:53 $Revision: r1
1.11/10 PATCH_11.11 (PHNE_33159)
/usr/conf/lib/libuipc.a(uipc_proto.o):
uipc_proto.c $Date: 2005/05/16 02:36:02 $Revision: r
11.11/10 PATCH_11.11 (PHNE_33159)
/usr/conf/lib/libuipc.a(uipc_socket.o):
uipc_socket.c $Date: 2006/05/16 11:25:58 $Revision:
r11.11/32 PATCH_11.11 (PHNE_34672)
/usr/conf/lib/libuipc.a(uipc_socket2.o):
uipc_socket2.c $Date: 2002/02/27 (PHNE_25644 + JAGae
07926)
uipc_socket2.c $Date: 2005/05/16 02:36:19 $Revision:
r11.11/15 PATCH_11.11 (PHNE_33159)
/usr/conf/lib/libuipc.a(uipc_syscall.o):
uipc_syscall.c $Date: 2005/08/17 08:10:47 $Revision:
r11.11/15 PATCH_11.11 (PHNE_33628)
/usr/conf/lib/libuipc.a(uipc_tpi.o):
uipc_tpi.c $Date: 2005/08/17 08:10:52 $Revision: r11
.11/17 PATCH_11.11 (PHNE_33628)
/usr/conf/lib/libuipc.a(uipc_usrreq.o):
uipc_usrreq.c $Date: 2006/03/09 23:16:09 $Revision:
r11.11/14 PATCH_11.11 (PHNE_34135)
/usr/conf/lib/libdebug.a(xport_diag.o):
xport_diag.c $Date: 2005/05/16 02:31:36 $Revision: r
11.11/15 PATCH_11.11 (PHNE_33159)
OS-Core.SYS-ADMIN,fr=B.11.11,fa=HP-UX_B.11.11_64,v=HP:
/usr/lib/nls/msg/C/netstat.cat:
netstat.msg $Date: 2003/02/06 04:36:46 $Revision: r1
1.11/3 PATCH_11.11 (PHNE_28497)
/usr/bin/netstat:
netstat.c $Date: 2003/02/06 04:36:35 $Revision: r11.
11/8 PATCH_11.11 (PHNE_28497)
osstat.c $Date: 2004/12/09 22:50:18 $Revision: r11.1
1/6 PATCH_11.11 (PHNE_32042)
cksum(1) Output:
Networking.NET-KRN,fr=B.11.11,fa=HP-UX_B.11.11_32/64,v=HP:
1095321124 5209 /usr/conf/net/if6.h
1770585349 2699 /usr/conf/net/netmp.h
1811482553 2663 /usr/conf/net/route6.h
266301056 10199 /usr/conf/netinet/icmp6.h
2487560865 9540 /usr/conf/netinet/if_ether.h
3659973769 9356 /usr/conf/netinet/in.h
1269236550 12788 /usr/conf/netinet/in6.h
1391467899 5922 /usr/conf/netinet/ip6.h
3877377201 3317 /usr/conf/netinet/tcp.h
717747742 2667 /usr/conf/netinet/tcpip.h
Networking.NET-PRG,fr=B.11.11,fa=HP-UX_B.11.11_32/64,v=HP:
1095321124 5209 /usr/include/net/if6.h
1770585349 2699 /usr/include/net/netmp.h
1811482553 2663 /usr/include/net/route6.h
266301056 10199 /usr/include/netinet/icmp6.h
2487560865 9540 /usr/include/netinet/if_ether.h
3659973769 9356 /usr/include/netinet/in.h
1269236550 12788 /usr/include/netinet/in6.h
1391467899 5922 /usr/include/netinet/ip6.h
3877377201 3317 /usr/include/netinet/tcp.h
717747742 2667 /usr/include/netinet/tcpip.h
Networking.NET-RUN,fr=B.11.11,fa=HP-UX_B.11.11_32/64,v=HP:
686918354 3657 /usr/lib/nls/msg/C/ifconfig.cat
356258594 70531 /usr/lib/nls/msg/C/ndd.cat
3388665283 3279 /usr/lib/nls/msg/C/ping.cat
4261341317 1821 /usr/lib/nls/msg/C/route.cat
1830711407 5196 /usr/lib/nls/msg/C/traceroute.cat
61907816 61440 /usr/sbin/arp
3349457150 45056 /usr/sbin/ifconfig
3152670333 36864 /usr/lib/libsnet.1
1750728552 36082 /usr/lib/libsnet.a
4067693567 114688 /usr/bin/ndd
934788384 32768 /usr/sbin/ping
780638013 28672 /usr/sbin/route
1026338285 49152 /usr/contrib/bin/traceroute
2244960135 7246 /usr/contrib/man/man1.Z/traceroute.1
605016573 61440 /usr/lib/libnm.1
1425048633 8986 /sbin/init.d/net
Networking.NET-RUN-64,fr=B.11.11,fa=HP-UX_B.11.11_32/64,
v=HP:
3652384385 64192 /usr/lib/pa20_64/libnm.1
Networking.NW-ENG-A-MAN,fr=B.11.11,fa=HP-UX_B.11.11_32/64,
v=HP:
862973683 4805 /usr/share/man/man1.Z/netstat.1
4266675570 2900 /usr/share/man/man1m.Z/arp.1m
2339166653 7100 /usr/share/man/man1m.Z/ifconfig.1m
1314914185 4448 /usr/share/man/man1m.Z/ping.1m
1161801980 6609 /usr/share/man/man1m.Z/route.1m
1835671594 4060 /usr/share/man/man2.Z/accept.2
2632866962 3602 /usr/share/man/man2.Z/bind.2
2420410105 5059 /usr/share/man/man2.Z/connect.2
3437388916 2375 /usr/share/man/man2.Z/getpeername.2
1915275919 2344 /usr/share/man/man2.Z/getsockname.2
1679560092 8309 /usr/share/man/man2.Z/getsockopt.2
1885655166 2191 /usr/share/man/man2.Z/listen.2
741466118 10238 /usr/share/man/man2.Z/recv.2
2927909451 10376 /usr/share/man/man2.Z/send.2
1256911327 2791 /usr/share/man/man2.Z/shutdown.2
2340916224 4432 /usr/share/man/man2.Z/socket.2
2941210579 1773 /usr/share/man/man2.Z/socketpair.2
352422879 5039 /usr/share/man/man7.Z/tcp.7p
752171725 2322 /usr/share/man/man7.Z/udp.7p
1087633690 2038 /usr/share/man/man7.Z/xopen_networking.7
OS-Core.CORE-KRN,fr=B.11.11,fa=HP-UX_B.11.11_32/64,v=HP:
4080073998 31405 /usr/conf/sys/mib.h
3041843783 19498 /usr/conf/sys/socket.h
3173264808 3709 /usr/conf/sys/socket_hatcp.h
1546496809 19739 /usr/conf/sys/socketvar.h
2555107607 7895 /usr/conf/sys/unpcb.h
ProgSupport.C-INC,fr=B.11.11,fa=HP-UX_B.11.11_32/64,v=HP:
4080073998 31405 /usr/include/sys/mib.h
3041843783 19498 /usr/include/sys/socket.h
3173264808 3709 /usr/include/sys/socket_hatcp.h
1546496809 19739 /usr/include/sys/socketvar.h
2555107607 7895 /usr/include/sys/unpcb.h
Networking.NET2-KRN,fr=B.11.11,fa=HP-UX_B.11.11_32,v=HP:
4147481664 16674 /usr/conf/lib/libipsectp.a
50880975 1616 /usr/conf/lib/libnet.a(af.o)
3539339784 35316 /usr/conf/lib/libinet.a(arp.o)
1796912935 8588 /usr/conf/lib/libnet.a(if.o)
368454192 13636 /usr/conf/lib/libinet.a(igmp.o)
4087401785 10692 /usr/conf/lib/libinet.a(in_proto.o)
3989230659 1860 /usr/conf/lib/libinet.a(inet_init.o)
3845071542 111680 /usr/conf/lib/libinet.a(ip.o)
878269621 78668 /usr/conf/lib/libinet.a(ip6.o)
1904305590 58596 /usr/conf/lib/libinet.a(ip6_if.o)
2836061430 26796 /usr/conf/lib/libinet.a(ip6_ire.o)
2028556532 44652 /usr/conf/lib/libinet.a(ip6_nd.o)
898076632 84656 /usr/conf/lib/libinet.a(ip_if.o)
3689144892 3412 /usr/conf/lib/libnet.a(ip_info.o)
3327600419 23340 /usr/conf/lib/libinet.a(ip_ire.o)
2350226413 22828 /usr/conf/lib/libinet.a(ip_mroute.o)
1971690502 10580 /usr/conf/lib/libinet.a(ip_multi.o)
2580533243 59764 /usr/conf/lib/libinet.a(ipcomm.o)
2046462206 1852 /usr/conf/lib/libinet.a(ipcsum.o)
2997754519 1964 /usr/conf/lib/libnet.a(mialloc.o)
2640033516 5240 /usr/conf/lib/libnet.a(mind.o)
2446470661 6940 /usr/conf/lib/libnet.a(mioptcom.o)
3773717241 7784 /usr/conf/lib/libnet.a(mitimer.o)
2881067220 23736 /usr/conf/lib/libnet.a(miutil.o)
1773547896 1464 /usr/conf/lib/libnet.a(net_init.o)
3552746014 3656 /usr/conf/lib/libnet.a(nuls.o)
2204593362 1968 /usr/conf/lib/libnet.a(pass.o)
2716573025 21524 /usr/conf/lib/libinet.a(rawip.o)
1686614658 28124 /usr/conf/lib/libinet.a(rawip6.o)
3313236998 1496 /usr/conf/lib/libnet.a(route.o)
1514047862 3864 /usr/conf/lib/libinet.a(snmpcom.o)
4226828845 2524 /usr/conf/lib/libnet.a(spass.o)
1033409227 33960 /usr/conf/lib/libnet.a(stcpmap.o)
2397107871 9176 /usr/conf/lib/libnet.a(tang.o)
2932960990 167848 /usr/conf/lib/libinet.a(tcp.o)
1376475212 49576 /usr/conf/lib/libinet.a(udp.o)
Networking.NMS2-KRN,fr=B.11.11,fa=HP-UX_B.11.11_32,v=HP:
2042494679 2356 /usr/conf/lib/libnms.a(nms.o)
OS-Core.CORE2-KRN,fr=B.11.11,fa=HP-UX_B.11.11_32,v=HP:
4188319136 2716 /usr/conf/lib/libaudit.a(dgram_aud.o)
4254877768 10876 /usr/conf/lib/libuipc.a(sys_socket.o)
3088674250 6040 /usr/conf/lib/libuipc.a(uipc_domain.o)
3109055195 19756 /usr/conf/lib/libuipc.a(uipc_init.o)
3434926172 1500 /usr/conf/lib/libuipc.a(uipc_proto.o)
2272970362 75176 /usr/conf/lib/libuipc.a(uipc_socket.o)
4166446929 19216 /usr/conf/lib/libuipc.a(uipc_socket2.o)
1500715361 30360 /usr/conf/lib/libuipc.a(uipc_syscall.o)
1168266787 18416 /usr/conf/lib/libuipc.a(uipc_tpi.o)
2140888254 15156 /usr/conf/lib/libuipc.a(uipc_usrreq.o)
2097920357 324584 /usr/conf/lib/libdebug.a(xport_diag.o)
OS-Core.SYS-ADMIN,fr=B.11.11,fa=HP-UX_B.11.11_32,v=HP:
898643950 19355 /usr/lib/nls/msg/C/netstat.cat
668761811 102400 /usr/bin/netstat
Networking.NET2-KRN,fr=B.11.11,fa=HP-UX_B.11.11_64,v=HP:
2988889014 26992 /usr/conf/lib/libipsectp.a
3001725554 4992 /usr/conf/lib/libnet.a(af.o)
2205902124 84224 /usr/conf/lib/libinet.a(arp.o)
2792126732 21640 /usr/conf/lib/libnet.a(if.o)
3776700874 34472 /usr/conf/lib/libinet.a(igmp.o)
270055690 31680 /usr/conf/lib/libinet.a(in_proto.o)
1480194434 4736 /usr/conf/lib/libinet.a(inet_init.o)
647883501 235576 /usr/conf/lib/libinet.a(ip.o)
957394877 158320 /usr/conf/lib/libinet.a(ip6.o)
409209489 128152 /usr/conf/lib/libinet.a(ip6_if.o)
3462435302 53632 /usr/conf/lib/libinet.a(ip6_ire.o)
2236235553 103624 /usr/conf/lib/libinet.a(ip6_nd.o)
962788759 180656 /usr/conf/lib/libinet.a(ip_if.o)
3948933229 6992 /usr/conf/lib/libnet.a(ip_info.o)
340557577 49032 /usr/conf/lib/libinet.a(ip_ire.o)
987862564 59912 /usr/conf/lib/libinet.a(ip_mroute.o)
3669428270 27528 /usr/conf/lib/libinet.a(ip_multi.o)
1205777827 130880 /usr/conf/lib/libinet.a(ipcomm.o)
632188091 3312 /usr/conf/lib/libinet.a(ipcsum.o)
2948518371 5048 /usr/conf/lib/libnet.a(mialloc.o)
295343554 12792 /usr/conf/lib/libnet.a(mind.o)
2778671507 10952 /usr/conf/lib/libnet.a(mioptcom.o)
1286246125 16368 /usr/conf/lib/libnet.a(mitimer.o)
1110467029 63256 /usr/conf/lib/libnet.a(miutil.o)
2899532792 3232 /usr/conf/lib/libnet.a(net_init.o)
1356611372 8968 /usr/conf/lib/libnet.a(nuls.o)
2831002815 5024 /usr/conf/lib/libnet.a(pass.o)
4180578987 50080 /usr/conf/lib/libinet.a(rawip.o)
908979366 58800 /usr/conf/lib/libinet.a(rawip6.o)
1379781058 5920 /usr/conf/lib/libnet.a(route.o)
2010401201 8384 /usr/conf/lib/libinet.a(snmpcom.o)
4124327347 6288 /usr/conf/lib/libnet.a(spass.o)
644774204 75264 /usr/conf/lib/libnet.a(stcpmap.o)
859476904 22056 /usr/conf/lib/libnet.a(tang.o)
2965395176 331200 /usr/conf/lib/libinet.a(tcp.o)
1607912965 91960 /usr/conf/lib/libinet.a(udp.o)
Networking.NMS2-KRN,fr=B.11.11,fa=HP-UX_B.11.11_64,v=HP:
3368635164 6592 /usr/conf/lib/libnms.a(nms.o)
OS-Core.CORE2-KRN,fr=B.11.11,fa=HP-UX_B.11.11_64,v=HP:
2400359635 5536 /usr/conf/lib/libaudit.a(dgram_aud.o)
2892140562 23136 /usr/conf/lib/libuipc.a(sys_socket.o)
2982022517 17456 /usr/conf/lib/libuipc.a(uipc_domain.o)
3748127116 54552 /usr/conf/lib/libuipc.a(uipc_init.o)
2521471967 3768 /usr/conf/lib/libuipc.a(uipc_proto.o)
487822909 140344 /usr/conf/lib/libuipc.a(uipc_socket.o)
2715229275 47200 /usr/conf/lib/libuipc.a(uipc_socket2.o)
1526378011 75744 /usr/conf/lib/libuipc.a(uipc_syscall.o)
2718287369 38520 /usr/conf/lib/libuipc.a(uipc_tpi.o)
3515539637 36408 /usr/conf/lib/libuipc.a(uipc_usrreq.o)
2498961267 406848 /usr/conf/lib/libdebug.a(xport_diag.o)
OS-Core.SYS-ADMIN,fr=B.11.11,fa=HP-UX_B.11.11_64,v=HP:
898643950 19355 /usr/lib/nls/msg/C/netstat.cat
668761811 102400 /usr/bin/netstat
Patch Conflicts: None
Patch Dependencies:
s700: 11.11: PHKL_25233 PHKL_25389 PHNE_31091
s800: 11.11: PHKL_25233 PHKL_25389 PHNE_31091
Hardware Dependencies: None
Other Dependencies:
1) To run HP-UX IPSEC over IPv6, IPSEC version A.01.05
or greater must be used.
2) On systems using the Hyperfabric products with
revisions B.11.11.00, B.11.11.01 or B.11.11.02,
PHNE_27745 or a superseding patch must also be
installed.
3) Retaining identd(1M) functionality with the installation
of this patch, requires a minimum sendmail 8.9.3 patch
of PHNE_26305.
4) The resolution of defect SR 8606172829 (JAGad42089)
requires that the following patches, or any superseding
patch, be installed together:
PHNE_25642 PHKL_25871 PHNE_25084 PHKL_25842 PHKL_25840
If the combination of the above patches is not installed
together this will not adversely impact your system.
Supersedes:
PHNE_34135 PHNE_33628 PHNE_33159 PHNE_32042 PHNE_31247 PHNE_31092
PHNE_30656 PHNE_30053 PHNE_29887 PHNE_29472 PHNE_28895 PHNE_28497
PHNE_28089 PHNE_27989 PHNE_27730 PHNE_27382 PHNE_27063 PHNE_26762
PHNE_25644 PHNE_25642 PHNE_25134 PHNE_24506 PHNE_24211 PHNE_23645
PHNE_23288
Equivalent Patches: None
Patch Package Size: 2840 KBytes
Installation Instructions:
Please review all instructions and the Hewlett-Packard
SupportLine User Guide or your Hewlett-Packard support terms
and conditions for precautions, scope of license,
restrictions, and, limitation of liability and warranties,
before installing this patch.
------------------------------------------------------------
1. Back up your system before installing a patch.
2. Login as root.
3. Copy the patch to the /tmp directory.
4. Move to the /tmp directory and unshar the patch:
cd /tmp
sh PHNE_34672
5. Run swinstall to install the patch:
swinstall -x autoreboot=true -x patch_match_target=true \
-s /tmp/PHNE_34672.depot
By default swinstall will archive the original software in
/var/adm/sw/save/PHNE_34672. If you do not wish to retain a
copy of the original software, include the patch_save_files
option in the swinstall command above:
-x patch_save_files=false
WARNING: If patch_save_files is false when a patch is installed,
the patch cannot be deinstalled. Please be careful
when using this feature.
For future reference, the contents of the PHNE_34672.text file is
available in the product readme:
swlist -l product -a readme -d @ /tmp/PHNE_34672.depot
To put this patch on a magnetic tape and install from the
tape drive, use the command:
dd if=/tmp/PHNE_34672.depot of=/dev/rmt/0m bs=2k
Special Installation Instructions: None
|
Printable version
