Patch-ID# 101946-39 Keywords: security kernel libsocket sockmod procfs nfs RDBMS strrput AIX cachefs Synopsis: SunOS 5.4_x86: patch for kernel Date: Feb/07/97 Solaris Release: 2.4_x86 SunOS Release: 5.4_x86 Unbundled Product: Unbundled Release: Xref: This patch available for SPARC as patch 101945 Topic: SunOS 5.4_x86: patch for kernel NOTE 1: If this patch is applied to a system which is used as an FTP server, install the driver update package DU10 or greater. Earlier driver updates, when used with this patch, can result in significant degradation of file retrieval time. NOTE 2: If this patch is applied to a system installed with the entire configuration, the default save option of the installpatch utility will use approximately 20 MB of free space in /var. BugId's fixed with this patch: 1120225 1124354 1130791 1143434 1143479 1145457 1151364 1151509 1151955 1152710 1155298 1157053 1159330 1159986 1160112 1162269 1162834 1163335 1164319 1164519 1164679 1164800 1165675 1165687 1167235 1169686 1169775 1169791 1169909 1171008 1171478 1171939 1172009 1172243 1172260 1172542 1172731 1172926 1172979 1172998 1173212 1173301 1173309 1173969 1174222 1174685 1174786 1174830 1174851 1174913 1175115 1175127 1175304 1175356 1175368 1175478 1175499 1175668 1175931 1176467 1176508 1176618 1177091 1177228 1177469 1177572 1177578 1177600 1178114 1178236 1178391 1178400 1178506 1178641 1178761 1178824 1178835 1178889 1178898 1178957 1178985 1179311 1179403 1179625 1179738 1179884 1180414 1180578 1180819 1181009 1181201 1181259 1182051 1182105 1182158 1182458 1182492 1182509 1182597 1183395 BugId's fixed with this patch: 1183552 1183568 1183662 1184134 1184256 1184991 1185149 1185775 1186156 1186224 1186287 1186420 1186557 1186569 1186805 1187322 1187536 1187901 1187948 1188259 1188287 1188307 1188399 1188475 1188701 1188790 1189271 1189389 1189592 1189967 1189968 1191078 1191422 1191457 1192162 1192238 1192309 1192982 1193066 1193448 1193696 1193721 1194355 1194613 1194878 1194923 1194928 1195432 1195436 1195437 1195904 1197596 1197646 1197708 1197979 1198215 1198278 1198439 1198966 1199124 1199164 1199579 1199624 1200224 1200502 1201926 1202070 1202675 1203132 1203471 1204479 1205240 1205409 1205731 1206384 1206598 1206642 1206850 1207181 1207277 1207669 1207954 1208034 1208053 1208241 1209012 1209014 1209452 1209687 1209917 1210314 1210355 1211022 1211172 1211278 1211904 1213782 1213874 1214043 1214057 BugId's fixed with this patch: 1214320 1215792 1216540 1217050 1217220 1217941 1218997 1219020 1219295 1219671 1219766 1220257 1220400 1220886 1220995 1221608 1221620 1221966 1222086 1222599 1222780 1222902 1223853 1223900 1224074 1224089 1224148 1224737 1227376 1227426 1227580 1229031 1229805 1229843 1231720 1235099 1231871 1231997 1232577 1232825 1232838 1232866 1232869 1233049 1233719 1233827 1234307 1234450 1234879 1236149 1238343 1240151 1240331 1241056 1241118 1241282 1241611 1242188 1242481 1244088 1244917 1244971 1245300 1245602 1245703 1247172 1248446 1248840 1249319 1249667 1249829 1250127 1251423 1253223 1255435 1255623 1256610 1258151 1259279 1260769 1260873 1260959 1261245 1262694 1266278 Changes incorporated in this version: 1219295 1223900 1227580 1234307 1241282 1259279 Relevant Architectures: i386 Patches accumulated and obsoleted by this patch: 101903-01 101919-01 101970-07 101972-01 101976-01 101982-02 101984-09 102008-02 102021-04 102120-01 102170-01 102217-07 102225-04 102359-01 102927-01 103576-01 Patches which conflict with this patch: Patches required with this patch: Obsoleted by: Files included with this patch: /etc/fs/nfs/mount /etc/lib/unix_scheme.so.1 /etc/name_to_sysnum /kadb /kernel/drv/cmdk /kernel/drv/cn /kernel/drv/icmp /kernel/drv/ip /kernel/drv/logindmux /kernel/drv/logindmux.conf /kernel/strmod/rlmod /kernel/drv/ip /kernel/drv/tcp /kernel/drv/tl /kernel/drv/udp /kernel/fs/autofs /kernel/fs/cachefs /kernel/fs/lofs /kernel/fs/nfs /kernel/fs/procfs /kernel/fs/tmpfs /kernel/fs/ufs /kernel/mach/uppc /kernel/misc/strategy /kernel/misc/strplumb /kernel/misc/swapgeneric /kernel/sched/TS /kernel/strmod/rlmod /kernel/strmod/sockmod /kernel/strmod/telmod /kernel/strmod/timod /kernel/sys/c2audit /kernel/sys/kaio /kernel/sys/nfs /kernel/unix /sbin/init /sbin/su /sbin/sulogin /usr/bin/su /usr/include/security/ia_appl.h /usr/include/sys/aio_req.h /usr/include/sys/aio.h /usr/include/sys/asynch.h /usr/include/sys/ddi_impldefs.h /usr/include/sys/errno.h /usr/include/sys/proc.h /usr/include/sys/sema_impl.h /usr/include/sys/sockmod.h /usr/include/sys/stropts.h /usr/include/sys/strsubr.h /usr/kernel/fs/s5fs /usr/kernel/sched/RT /usr/kvm/adb /usr/kvm/crash /usr/kvm/lib/adb/sema /usr/lib/autofs/automountd /usr/lib/fs/autofs/automount /usr/lib/fs/nfs/inetboot /usr/lib/fs/nfs/mount /usr/lib/fs/nfs/umount /usr/lib/libaio.so.1 /usr/lib/libauth.a /usr/lib/libauth.so.1 /usr/lib/libc.a /usr/lib/libc.so.1 /usr/lib/libp/libc.a /usr/lib/libsocket.a /usr/lib/libsocket.so.1 /usr/lib/libthread.so.1 /usr/lib/libthread_db.so.0 /usr/lib/pics/libc_pic.a /usr/lib/security/unix_scheme.so.1 /usr/lib/utmp_update /usr/ucblib/libucb.a /usr/ucblib/libucb.so.1 /usr/sbin/in.ftpd /usr/sbin/in.telnetd /usr/sbin/in.rlogind /usr/sbin/init /usr/bin/ftp /kernel/misc/klmmod /kernel/strmod/rpcmod /kernel/misc/tlimod Problem Description: 1259279 x86 byte ordering/endian problem in ip & arp for DL_UNITDATA_REQ 1241282 ftp session dies on cd or dir 1234307 problem with system call return using ucblib on x86 1227580 cannot support high TCP connection rates: noncaput errors reported by the driver 1223900 alarm(2) doesn't work properly with large arguments 1219295 automountd doesn't umount the hsfs filesystem. (from 101946-38) 1266278 freeing free xxx panic; indirtrunc tries to free the same block twice 1262694 Solaris 2.4 hangs due to memory leak in kmem_alloc-8, kmem_alloc_24 and -40 lea 1261245 window probes can cause ack wars 1260959 Streams information delayd 50-100 ms until dbri driver schedules it 1260873 Kernel memory gets corrupted when sharing and unsharing secure NFS. 1260769 MT application is dropping signal events when run on multi-processor systems 1258151 Solaris 2.4 nfs -o noac option not working properly with novell nfs server 1255623 getdate() fails on 1st of month with julian date getdate() fails on 1st of month with julian date scripts/ 1255435 ftp dumps core if lostpeer signal handler is called right before getreply() 1253223 System running 2.3 with KJP-80 on single CPU /24MB hangs in fork test case 1249319 lo_sync() should not flush the underlying filesystem 1248840 solaris 2.3,sc2000, TCP socket can't handle FIN pkt from client surely, deadlock 1248446 arp cache is not getting updated appropriately 1247172 Threads losing signals when preempted 1241118 libthread panic in thr_join handling of zombie threads seems to be broken 1234450 NFS (VOP_WRITE &c) returns EINTR when "intr" is not specified on the mount. 1232838 Backport 1229099: sched: sched()/prioctl()/clock() deadlock during heavy swappi 1232577 Signal delivered twice in MT program 1151955 The PASSLENGTH attribute in /etc/default/passwd doesn't appear to work 1255435 ftp dumps core if lostpeer signal handler is called right before getreply() (from 101946-37) 1256610 strwrite failes to call queuerun on error path: bug performance hit 1251423 panic - recursive mutex_enter on lwplock 1249829 connection times out if remote only sends zero-windows 1244971 solaris 2.3, patch 101318-77 has a bug, it can't handle `boot -s` correctly. 1233827 tcp retransmits too much for short connections as seen at web sites 1233049 System hangs when user stops thread writing to ODS logging device 1219671 Memory is given free which was never allocated before. 1215792 delayed availability of freed diskspace when UFS logging with ODS 4.0/3.0 1245602 Logging UFS is slower than UFS for local writes 1198215 ftp can silently lose data when writing to nfs 1175499 repeated getdate calls leaks memory 1240331 ifconfig of a non-existent virtual interface creates one even as a regular user (from 101946-36) 1250127 kernel memory leak - machine hanging - looks like a problem with streams_msg_15 1242188 hang waiting for rwlock with holdcnt of -1 but no owner 1227376 panic "Deadlock condition detected: cycle in blocking chain" 1245703 Deadlock condition detected: cycle in blocking chain 1241056 TL driver panics while servicing tl_ordrel 1221966 telmod writes NULL byte into a zero-length mblk 1244088 SS2000 is completely hanging under heavy I/O - Solaris 2.4 + 101945-36 1206850 Solaris 2.4, telnet/ftp error in single user mode. (from 101946-35) 1242481 panic: ufs_putapage: bn == UFS_HOLE 1238343 hang after installing T101945-37 1241611 Machine panics with page_vpsub 1232869 paging thresholds are too low on very big systems causing kmem alloc failures 1240151 sockmod corrupts unix socket list due to 2 binds on same socket at same time 1198966 Buggy streams programming causes panic 1244917 syslog(3) does not correctly cache the file descriptor that it writes on (from 101946-34) 1233719 system hangs due to idle_q > maxninode and IREF set in head inode of ufs_idle_q 1236149 connect() on AF_UNIX/SOCK_STREAM sockets hangs on Solaris x86 platform 1231997 f77 REWIND makes error : "eor/uio [1010] off end of record" on nfs files 1223853 TCP stream may not go away when process dies 1234879 system panic with auditd: zero divide trap when 101318-75 applied. 1229805 popen assumes maximum number of file descriptors is 256 1220995 directory blocks not counted in quotas 1218997 Simple cp/rm operations hang and cannot be killed 1197979 Recursive use of xdr_pointer could blow autofs unmount thread stack (from 101946-33) 1232866 bogus pkt_len (short one) causes ip to panic 1229843 2.4 ufs `umount' on an errored device hangs system. 1229031 page_unlock: page not locked panic occurring when locking address space 1224089 NFS writes hang when doing copy over fddi 3.0.1 1208034 SC1000 5.4 Data fault out of tcp_rput_data null pointer under heavy load 1188307 ill_frag_timeout() panic on Intergraph machines when heavily stressed 1224148 TCP performance becomes unacceptable due to bad checksums generated by 2 .4 1189967 real-time latency limits exceeded occasionally 1231871 cpu_surrender doesn't check for threads waiting on kp queue (from 101946-32) 1221608 Automounter: NIS+ Searchpath 1211172 Automountd fails to unmount lofs file system 1224737 SC2000 panics in setf() with fd >= u.u_nofiles (setf called by lockd) 1220400 lofs becomes confused about where the present working directory "." is 1210355 TCP goes into "throttled receive" mode when using multithread 1222902 panic in tcp_xmit_ctl because referencing through the wrong mblk. 1217941 Data fault from cron in anon_getpage 1208053 Implementation of LC_* and other environment variables causes security problem. 1222599 pullupmsg() can corrupt kernel memory and hang CPUs 1214320 lwp_cond_wait() syscall returns to user holding a kernel mutex 1205731 mktime malfunction 1222086 file server hangs when many users exceed quotas 1213782 informix server hangs in getmsg() 1174685 Performance tuning parameters have no effect on x86 machines 1235099 Using sigprof and libaio will cause program to segfault. (from 101946-31) 1231720 back out the patch for 1176618; it broke the 101945-35 kernel jumbo patch 1227426 system hung while doing dlm related activities (from 101946-30) 1220257 Syslog(3) possibly can be abused to gain root access on Solaris 2.x systems 1210314 ufs_bmap complains about about user level errors 1206642 utmp_update can be used to make bogus entries in utmp 1219766 recv() returns 0 bytes if you close socket too quickly after doing a send() 1209917 fddi crashes when MTU is larger than 4352 1197646 mktime() can not handle negative tm_isdst across year boundary 1217220 BSD sockets make SVR4 programs hang 1209687 Panic from audit_thread_free with saved path not empy 1180819 out of per-user processes message should include uid of offending user 1199124 out of per-user processes warning message should not be kern.crit 1219020 any user can hang the system with fork(); only recourse is rebooting 1224074 Sybase OpenServer client application hangs with 101945-32 or -34 1216540 potential deadlock when process auditing is enabled 1209012 hard hang in findmod: The code in the while goes round and round and round ... 1176618 prog dumps core if you printf very large string 37000 1222780 bug in ftpd that can cause it to dump core. 1211904 sc1000 running solaris 2.4 and FDDi 3.0.1 panics with "UIOSTR: strgetmsg ..." 1217050 system panics with strread: STRUIO 1208241 core dumps generated by set-uid executables possibly reveal data 1175668 automountd consumes all cpu time and loops when cd to automount directory (from 101946-29) 1220886 patch 101945-32 breaks Informix 1221620 aioread(3) coredumps when ENOMEM is expected 1214057 crash provides open fd for /dev/mem 1213874 bug 1155298 integration port mistake affects AF_UNIX socket bind() error returns 1211278 ufs function quotadq() thrashing on a lock 1209452 When user go over quota no messages are printed to the console and messages file 1209014 uprintf can cause panic on modified OBP systems 1207954 TCP data got corrupted when using multithread libraries 1205409 sun4m intermittently goes down with PANIC: getdiskquota 1203471 Solaris does not guarantee bounded dispatch latency for RT processes 1203132 lockfs -h and umount of the UFS lying under a loopback file system causes panic 1200224 unreferenced vnodes may persist in active locks table when lockd dies 1193448 autofs unnecessarily blocks requests on already mounted filesystems 1183662 Search for lofs to unmount should stop on first match 1193066 system hangs due to deadlock between u_flock and pidlock 1191422 read on af_unix socket returns 0 when other end does a write(len = 0) 1186569 2.4FCS x86 threads causing a program to crash with SEGV 1207277 x86 threaded app SIGSEGVs when an LWP or bound thread exits 1214043 Makefile.master has wrong RELEASE info in /ws/on494-patch 1186420 in.ftpd does not call sa_auth_acctmg and thus ignores password aging, etc. 1186287 an unprivileged user can use utmp_update to clear entries from /var/adm/utmp 1184256 data fault in freeproc() caused by race with cfork() 1184134 dlmd needs to pop timod, once TCP problems resolved 1179311 1000's of zombie processes which kill further attempts to login to system. 1175356 automounter fails to remount unmounted directories 1175127 2.3 tcp performance over satellite/delayed links is very poor compared to 4.1.3 (from 101946-28) 1211022 after installing T101945-30 you cant login to a dataless client 1207669 NFS writes can fail and no error is ever returned, even to fsync or close. 1207181 kmem_cache_xxxx panics with nfs file systems. 1206598 "Cannot reset access time of file at inode x" messages during backup 1206384 fscanf function failed with EUC character 1205240 Mounts of secure file systems fail at random 1199624 queuerun indirectly causes fork() call to hang 1197708 data fault panic in sockmod 1192309 machine panics with audit_finish: residue audit record (from 101946-27) 1204479 sprintf format "%.4S" prints improperly when strings include 0216 or 0217 1202070 one residual stop-the-clock-and-hang-the-system bug in /proc 1201926 strrput is calling queuerun() this may lead to a dead lock. 1200502 fopen() and unlink() makes a corrupt file on multi cpu machine. 1194613 getdate() bug under Solaris 2.4 . 1182105 libaio and libthread are not compatible 1199164 process won't exit due to non-zero refcount 1180578 PPP stops working after system is installed with 101318-63 1179884 Non-blocking socket connection over x25 would hang the system. 1164319 sc2000 panic with sema_v turnstile corruption - (from 101946-26) 1178889 TCP doesn't close down properly across the loopback interface 1189592 Infinite loop in TCP 1186156 keyserv caches old private key after user's password is changed 1194878 ioctl with TCGETA doesn't work with libsocket (from 101946-25) 1199579 flow controlled send does not generate SIGPOLL 1195432 Panic with a data fault from in.telnetd 1182509 FTP transfer hangs (from 101946-24) 1195436 With patch T101945-20 /.profile is not executed 1194928 System paniced when shutting down. Data fault panic in canputnext. 1194355 tcp server detects checksum error and loops 1193721 data fault in putq() due to NULL q_last pointer 1189389 machine panics with a data fault in mutex enter 1182158 Some cmds (ls, pwd) will hang when executed on nfs mounted directories. 1193696 kernel memory allocator: invalid free: buffer not in cache kmem_alloc_1152 1186805 other [] too many write error and EDQUOT messages from nfs to syslog 1191078 Machine hangs with many proc's in rmalloc_wait - memory leak 1178641 NFS client should fail to open files with the mandlock bit set 1175931 nfs loops on async write errors 1198278 ksh loops in kernel making NFS read calls on its history file. 1180414 streams allocb failure results in data corruption 1162834 deadlock between prioctl() and munmap() 1177469 /proc causes page deadlock in NFS 1182597 swapped out lwp->lwp_ar0 in prgetprregs causes data fault and hang 1187536 Deadlock using /proc 1188701 assertion failed: new_state != LMS_WAIT_CPU 1189271 procfs: run-on-last-close doesn't always work 1192982 deadlock condition detected: cycle in blocking chain 1198439 procfs is out-of-spec with respect to microstate accounting 1145457 ksh does not set the correct arguments for su - (from 101946-23) 1195437 Panic in ip layer (icmp_inbound_error missing pullupmsg) 1124354 Scorpion and Gal-Ross panics on Scheduler stress test (from 101946-22) 1187901 Process hung in nanosleep 1185149 nl_langinfo is not MT-safe (from 101946-21) 1192238 "noac" mount option not honored immediately after mount 1172926 application hangs on a TCP connection if the remote system dies 1130791 2.x setsockopt SO_SNDBUF fails with protocol error for stream AF_UNIX domain (from 101946-20) 1189968 Need strict multihoming in IP to prevent breakins over the Internet 1188475 pentium fp work-around 1187948 machine is hung because a thread is looping in connopen() (no-25 (from 101946-19) 1191457 DROPEN failed executing Fortran program (from 101946-18) 1188259 ls -a .. causes Data fault panic after lockfs -h and umount the file system 1182051 sched: Text fault while telnet test 1173309 system panics with assertion failed: tcp->tcp_rcv_head == NULL 1164519 Socket returns with "address already in use" because conn in "BOUND" state (from 101946-17) 1188287 NFS mounted files get truncated 1177228 Data fault in freeb routine while running sundiag 1165675 rquotad returns inappropriate error on nfs client This fix includes a modified sys/errno.h which introduces a new error number: EDQUOT (49). When an over-quota condition is encountered, the following filesystem-related system calls will fail and the errno will be set to EDQUOT. Previously, the errno was set to ENOSPC. The affected system calls are: creat(2), link(2), mkdir(2), mknod(2), open(2), rename(2), symlink(2), and write(2) Any applications that check for an over-quota condition during a failed system call may encounter EDQUOT (49) as a valid value for errno. (from 101946-16) 1155298 bind of AF_UNIX address simultaneously from multiple processes can fail 1143479 setuid/setgid program takes on default system limits (from 101946-15) 1186557 pid_ref field wraps around manifests as kmem list corruption. 1186224 socket select hangs in NON-BLOCKED mode 1178506 INN wounded after upgrade to SunOS 5.4 1181009 setsockopt returns error when expanding max receive size to 20KB (AF_UNIX) 1169775 Solaris 2.X does not correctly handle Copy-On-Write faults on a page (from 101946-14) 1175368 SECURITY anyone can gain root access to a 2.4 machine (from 101946-13) 1183568 NFS client get old data after file on server being updated (from 101946-12) 1183395 stratagey: x86 System hangs when lomemalloc fails 1178114 ioctl SIOCSPGRP/FIOSETOWN path broken for MT libsocket(linked to libthread) code 1171008 Mux hangs when expecting messages on lower stream during I_LINK/UNLINK 1159986 lckpwdf causes passwd to crash (from 101946-11) 1177600 No way to cache the root and /usr file systems with CacheFS (from 101946-10) 1178957 sigurg not delivered on second oob data arrival 1164800 panic: ddi_setcallback: no callback structures (from 101946-09) 1178835 RCS operations fail on file system NFS mounted from AIX system The problem happens when application do fchmod between writes (which is very rare) which has a chance to lead different views of the file attributes on client from that on server. The solution purges the client cache before doing setattr so that the views will be the same. (from 101946-08) 1178641 NFS client should fail to open files with the mandlock bit set 1175478 Panic in prototype inkernel logdmuxunlink() after munlink failed 1178236 System panics with data fault in free_zero_zero() 1175115 nfs write error "(file handle: xxx xxx" message cannot be redirected by syslog The problem occurs when nfs encounters write errors. NFS will print a write error to the console. In some cases the physical console is printed upon in the event that the console driver is deprived or resources. What has been done is to put a throttle on NFS write error messages, enabling the administrator to type on the console and try to figure out what is going on. Socket interface networking programs under heavy use may panic the machine with free_zero_zero() on the kernel call stack. This fixes the problem in the sockmod module. I_UNLINK or I_PUNLINK commands may time out and close the stream before the multiplexor has processed the command. The NFS server will deny access to mandatory lock files. This is done for two reasons. First, mandatory locking is not supported over NFS. Second, it is dangerous for the server to access mandatory lock files. It would be very easy for a normal user to completely hang the NFS server. The user could create a file and set the mode to indicate that it is a mandatory lock file. It could then lock the file with a program which then just does a pause. This user could then go to an NFS client and try to access the file. With each request from the client, including retries, another NFS server daemon on the server would get blocked, until the server ran out of NFS server daemons. (from 101946-07) 1177091 prgetstatus can generate pagefault holding p_lock, can deadlock if freemem is 0 1177578 strmakemsg/strgeterr causes panic in strrput due to NULL mblk ptr 1176467 fcntl system call fails in process run by rcmd 1172243 Customer runs application from dumb terminal and system crashes. The system can freeze under heavy swapping pressure due to procfs holding a critical lock when it takes a page fault. Doing I_SETSIG on a console window through serial line and exiting the process could cause a system panic. Kernel panic in putnext/ptcwrite. A socket endpoint not created through the socket library (by dup() of a socket endpoint for example) may experience some failures on fcntl()/ioctl() calls. (This bug is only limited to 2.4 release) (from 101946-06) 1177572 installing Solaris 2.4 ON patch 101945-05 and running OW causes machine to panic The patch to bug ID 1151364 broke OW's consolidation. This happened because releasef() changed to have an extra argument. OW shouldn't have been dependent on releasef() which is private to the ON consolidation. Since this problem was not discovered until after the patch was made, it made more sense for ON to produce a new patch which restores releasef() to have its old interface. The interface changed for kaio. A new interface is added called areleasef() which is only used by kaio. (from 101946-05) 1174830 savecore on diskless machine didn't generate unix, vmcore is trash 1151364 asynchronous I/O in the user level hurts RDBMS performance This is a performance improvement for applications that are using libaio for doing async IO to raw files or devices. There are no API changes, only a new version of libaio.so.1 is installed. One side benefit of this fix is that async IO to tape should now work. This patch to bug 1151364 requires installation of libaio/kaio patch 102021-01 or later) Kernel crash dumps generated on diskless sun4m, sun4d or i86pc systems are not complete. (from 101946-04) 1172243 Customer runs application from dumb terminal and system crashes 1169686 4.1.3 system on network goes down, hangs 2.3 system The problem shows up when a "ps" thread is running through the virtual memory area to get the address space size for a mapped file. The address space lock is held and a get attributes function is called. This initiates an nfs get attribute request. If the machine that the request is made to is not responding the nfs request will block. The address space lock which is held by the blocked ps thread might block other processes on the local machine. Typically when a server goes down all nfs file system activity is blocked on any clients. The nfs operation resumes once the server comes up. In this situation a server is powered down and causes a client to hang. The hang is due to a process pile-up. The client is doing a ps and its thread is holding the address space lock (as_lock) for a running process lets call A. The A process is a mapped file from the server. The client ps thread path has reached rm_assize() which needs to get the file size so it calls VOP_GETATTR() which goes across the wire to the server. This operation goes nowhere because the server is not running. The as_lock held by the ps process is blocking other processes such as init. The solution is not to go over the wire but to return a cached entry for the file size. The change is to define a new attribute flag in vnode.h called ATTR_HINT. The rm_assize() function recognizes will use this flag when it calls VOP_GETATTR(). The nfs getattr function will see that the size of the file is requested and that the passed in flag is ATTR_HINT. It will return the file size from the rnode rather than make a request to the server. Typically when a server goes down all nfs file system activity is blocked on any clients. The nfs operation resumes once the server comes up. In this situation a server is powered down and causes a client to hang. The hang is due to a process pile-up. The client is doing a ps and its thread is holding the address space lock (as_lock) for a running process lets call A. The A process is a mapped file from the server. The client ps thread path has reached rm_assize() which needs to get the file size so it calls VOP_GETATTR() which goes across the wire to the server. This operation goes nowhere because the server is not running. The as_lock held by the ps process is blocking other processes such as init. The solution is not to go over the wire but to return a cached entry for the file size. The change is to define a new attribute flag in vnode.h called ATTR_HINT. The rm_assize() function recognizes will use this flag when it calls VOP_GETATTR(). The nfs getattr function will see that the size of the file is requested and that the passed in flag is ATTR_HINT. It will return the file size from the rnode rather than make a request to the server. Running applications that do I_SETSIG on console, when console is the serial port (i.e not the frame buffer), causes system to crash, when attempting to send signal to a process. (from 101946-03) 1169909 Running xlib code in Realtime class causes code to block. in poll() 1167235 panic data fault in strioctl - apparently doing TIOCSPGRP Protect with mutex the testing and setting of the session and controlling terminal related flags in the streamhead. Real time stream threads will block in a poll. (from 101946-02) 1172979 spurious SIGALRM received in test program that forks child processes 1172009 recv() on sockets should return the error only once for SunOS 4.X compatibility 1165687 non-blocking reads on sockets block under Solaris 2.3 1160112 socket library accidentally closes file descriptor on error 1120225 recv() returns EPIPE when called with MSG_PEEK 1152710 socket lib in 2.3/2.2 have problems with not clearing bad connections and errno 1171478 socket recv() calls fail with EINVAL due to bad fix in 494 AF_UNIX and AF_INET sockets can sometimes get EPIPE errors for recv(MSG_PEEK). When the socket library sees the EPIPE error it will in some cases close the file descriptor causing the application to get EBADF errors for subsequent operations. A AF_UNIX listening socket can get into a permanent error state (returning EPIPE or ECONNRESET) for any operation until the socket is closed. The non-blocking attribute of a socket endpoint is not transferred from a non-blocking listener endpoint to a accepting endpoint. This causes some socket non-blocking programs to block. This patch fixes the problem by setting the accepting endpoint non-blocking attribute if the listener was non-blocking. In SunOS 4.X sockets when a read() or recv*() call returns an error the application can do another read()/recv*() and get an EOF. This patch applies this subtle aspect of socket semantics to SunOS 5.X. This specification of signal actions from the signal(5) manual page was being violated: Setting a signal action to SIG_IGN for a signal that is pending causes the pending signal to be discarded, whether or not it is blocked. Any queued values pending are also discarded, and the resources used to queue them are released and made available to queue other signals. The condition under which the pending signal was not being discarded was the specific case of SIGALRM signals generated by the setitimer(ITIMER_REAL) interface. The malfunction happens in a narrow race condition which will be triggered under intensive setting of a signal handler and setting it to SIG_IGN while the itimer is active. (from 101946-01) 1173969 MT process doesn't stop on multi processor systems dbx appears to malfunction when controlling a multithreaded process that does many fork1()s. The bug is in the system, not dbx. Also, stopping dbx with a jobcontrol signal from the terminal, ^Z, while it is controlling a multithreaded process will cause the multithreaded process to becomed permanently stopped. (from 101903-01) 1172542 gettimeofday() returns negative nanosecond value on x86 1171939 Process dump core at random on loaded systems 1169791 processes often getting killed with SIGABRT and core dumped on MP IntelExpress gettimeofday() call can return negative nsec value at times. Processes can dump core on heavily loaded systems. (from 101919-01) 1157053 System panics when doing a copy to NFS file system mounted across FDDI-S Cause of problem is due to non-aligned transfers. The memory address alignment trap happened in xdr_writeargs() when copying data in a loop. The address was not on a long word boundary, it was on a word boundary. nfs_feedback() can adjust the transfer address and size for a request such as for a retransmission. The xdr_writeargs() can make use of bcopy(). The xdr_writeargs() is in file nfs_xdr.c. There are a few other functions in this file that do a similar copy operation that should be changed to use bcopy. (from 101976-01) 1173301 Files can sometimes be missing from a cachefs mounted directory. This can happen if the entry in question is the last one in the directory block, but would be the first one in the cachefs front file. If a client system runs touch on this file, it will erase the contents of the file on the server. (from 101970-07) 1182458 network interface can hang on NFS server with high ipReasmDuplicates counts (from 101970-06) 1178985 Multicast routing broken in Solaris 2.4 1179625 freeb(): bad pointer passed to kmem_cache_free A performance problem with T101969-05 will be seen in TCP/IP network connections over high bandwidth network interfaces (FDDI, ATM, 100Base-T, but not 10Base-T EtherNet) resulting in lower then expected throughput. There is no impact on TCP/IP functionality. IP Multicast routing does not work correctly in 2.4. The (publicly available) mrouted program does not receive any IGMP_HOST_MEMBERSHIP_REPORT messages due to the kernel incorrectly thinking that these messages are malformed. (from 101970-05) 1179625 freeb(): bad pointer passed to kmem_cache_free Kernel PANIC in freeb(), from freemsg(), from tcp_closei(), ... may be caused by TCP freemsg()ing an already freed mblk. This bug was introduced in a bug fix for 1167357. (from 101970-04) 1178391 system with PPP device using the same IP address as le0 will stop working 1178400 NFS copies btw 690MP(512) server and Sunos 4.1.3 corrupt data without any error MPs can transmit IP packets with the same ip_id field potentially causing fragmented packets to be reassembled incorrectly. Normally this is not a problem since the corruption will be detected by the UDP/TCP checksum. However, SunOS 4.X does not by default verify the UDP checksum in which case the incorrectly reassembled packets can cause NFS file corruption. When an IP address is shared between an ethernet and a point-to-point links and if the links go down and point-to-point links comes up first, the ethernet link will not be able to come up with the shared IP address.. (from 101970-03) 1172731 After PPP connects improper routing entries cause problems When an IP address is shared between a POINT-TO_POINT interface and a numbered interface can result in invalid routing entries. (from 101970-02) 1174786 Unnumbered interfaces with respect to PPP have problems 1174851 SC2000 hang due to no ip clow control when used with FDDI board When a high bandwith network interface is receiving a large number of packets addressed to it, but with no one bound to the specified port, then IP does a lot of processing. IP sends an ICMP unreachable packet back to the source of the original packet. This can cause large amounts of kmem to be consumed, which can cause subsequent kmem_alloc() failures, including allocb() failures in the driver for the high bandwidth interface driver. This can cause subsequent fragments of large IP frames to be dropped by the driver. IP will then hold on to these incomplete frames awaiting the arival of the missing fragments which will never show up, IP holds on to these frames for 60 seconds. Which in the case of a NPI FDDI interface at 80Mb/S can be 300Mbyte of kmem. When a point-to-point interface shares an ip address with a numbered interface, point-to-point link will stop receiving packets if the numbered interface is shutdown. (from 101970-01) 1162269 all net IP broadcast packets (255.255.255.255) have a ttl of 1 Applications and environments that depend on routers forwarding broadcast packets might run into problems with the fact that IP sets the TTL of all broadcast packets to 1 (in order to avoid any broadcast storms when there are misconfigured machines on the wire). This patch makes it possible to override the default TTL of 1. (from 101972-01) 1172260 5.4 <-> 4.1.2 socket connection looses sync and delays transfer of data A TCP connection might not start immediately when a window update is received after the Solaris 5.4 side has sent a zero window probe. With some TCP implementations at the remote end there will be a few seconds of delay (waiting for a retransmit timeout). (from 101982-02) 1179738 Users with 8 characters name or more will not be logged into utmp (from 101982-01) 1173212 SECURITY: su can display root password in the console If a username is too long (greater than 8 characters), when su root fails or succeeds for that user, the characters typed in as the password are echoed to the console. (from 102170-01) 1178761 ufs_putapage:bn == UFS_HOLE panic when filesystem fills up A corrupt inode can be created when extending a UFS file and running out of space. This can later cause a panic 'ufs_putapage: bn==UFS_HOLE'. (from 102927-01) 1178824 When catman is invoked on a system w/BSM & 101318-54 the system crashes (from 102120-01) 1163335 SS1000 interactive performance very poor with > 200 users logged in Traditional BSD-derived systems require per-character processing to be handled by the rlogin and telnet daemons. This is very inefficient, since it often requires several user level context switches per input character. This patch provides a "fast path" entirely in the kernel, which eliminates the added overhead of processing by a user-level daemon for normal data traffic. (from 102021-04) 1195904 usr/src/uts/common/sys/Makefile missing ascii.h (from 102021-03) 1185775 KAIO thorttling too conservative, Sybase cannot install on low memory machines (from 102021-02) 1151364 asynchronous I/O in the user level hurts RDBMS performance This patch provides the complete set of deliverables to fix 1151364. (from 102021-01) 1151364 asynchronous I/O in the user level hurts RDBMS performance This is a performance improvement for applications that are using libaio for doing async IO to raw files or devices. There are no API changes, only a new version of libaio.so.1 is installed. One side benefit of this fix is that async IO to tape should now work. (from 101984-09) 1202675 automountd can dump core due to double endnetconfig() call (from 101984-08) 1181201 port option does not work with autofs. (from 101984-07) 1181259 NFS mount fails with: couldnt bind to reserved port (from 101984-06) 1176508 panic mutex adaptive exit under 2.4 fcs when accessing directory over nfs. 1172998 x86: auto_lookup(): assertion failure in mutex_exit() on non-existent fs (from 101984-05) 1182492 autmountd's macro_expand function may cause buffer to overflow (from 101984-04) 1159330 automountd unmounts the wrong lofs (from 101984-03) 1174913 autofs checking for local subnets doesn't work when NIS+ is the nameservice This patch is to fix bug 1174913: autofs checking for local subnets doesn't work when NIS+ is the nameservice. The problem is that when a mount takes place ,its not giving a preference to the interface that the client machine is sitting on. It should be mounting from the servers interface that the client machine is attached to first and then an alternate if that does not respond. This is because that the automounter is looking up the table netmask while it should be looking up the table netmasks. (from 101984-02) 1151509 automounter's built in timeout is too short for low speed lines automountd by default only waits 15 seconds for servers to reply to its initial connection requests. This timeout may be too short for slow links or for very busy servers. This patch allows the system administrator to tune the total timeout by specifying the number of attempts (original + retries). This is done by adding a retry=n entry to the options list for the busy server entry in the automounter map. The default is one attempt (retry=0), when no retry=n option is specified in the options field, or when the retry=n option is invalid. Each retry is equivalent to approximately 30 seconds. Since automountd is currently single-threaded, this option should be used with care, as it will cause automountd to take more time to decide whether a server is dead or not (reply received or not), causing incoming autofs kernel requests to be queued for longer periods of time. For example, the following /etc/auto_home map uses the retry=1 option to force automountd to send the original request, and retry it once more, before giving up with a "server not responding error". If the reply is received before the next retry, there will be no retransmission. NOTE: It is not recommended to set this option as the map default, since it will cause automountd to needlessly wait longer for replies from real dead servers which otherwise would have replied without the need for retries had they been up. /etc/auto_home: # Home directory map for automounter # userx -nosuid,hard,intr,retry=1 busy_server:/export/home/userx +auto_home (from 101984-01) 1174222 automounter does not mount from 4.1.3 NFS servers with libc patch automountd first makes a null RPC call to the remote portmapper (rpcbind) of the server from which it needs to mount to determine if the server is able to respond to mount requests or not. In some cases (multiple servers specified on map entry) it would call the remote portmapper using version 3, which is not available on non SVR4 systems. Some systems are now silent about version mismatches, which causes automountd to assume the server is dead (or at least it's rpcbind/portmapper). This patch fixes this problem by always using version 2 of the portmapper protocol. (from 102008-02) 1184991 2.3/2.4 - panics in tmpnode_hold - data fault - TMPFS (from 102008-01) 1175304 vnode v_count is not maintained correctly vnode v_count numbers are not maintained correctly causing vnodes to never disappear or, in the earliest bug, drop to zero prematurely and panic the system. (from 102225-04) 1192162 programs using cond_timedwait eventually hang 1187322 cannot thr_continue the "main" thread of execution after a call to thr_suspend (from 102225-03) 1188399 sema_init with USYNC_PROCESS does not work properly (from 102225-02) 1188790 cond_timedwait occaisonally returns incorrectly 1186569 2.4FCS x86 threads causing a program to crash with SEGV (from 102225-01) 1178898 (from 102359-01) 1183552 ftpd processes hang at httpd (WWW server) sitesq (from 103576-01) 1249667 ftp size increases by 8k/2 page size with every open/close session memory leak (from 102217-07) 1232825 RPC: Unable to send/receive (from 102217-06) 1245300 5.4 lockd on client side can't handle two outstanding klm requests simutaneousl (from 102217-05) 1164679 KLM doesn't initialize rsys & rpid correctly (from 102217-04) 1143434 Secure nfs does not work properly across NIS+ domains (from 102217-03) 1197596 _svcauth_unix can crash the kernel (from 102217-02) 1194923 System hangs with many klm_lockctl messages echoed to the screen. (from 102217-01) 1179403 NFS client starts using unreserved UDP port numbers Patch Installation Instructions: -------------------------------- Refer to the Install.info file within the patch for instructions on using the generic 'installpatch' and 'backoutpatch' scripts provided with each patch. Any other special or non-generic installation instructions should be described below. Special Install Instructions: ----------------------------- It is necessary to install the same revision of this patch on both servers and clients for remote login to clients to work. 1) Stop automountd # /etc/init.d/autofs stop 2) Install patch 3) Edit the necessary entries on your automounter maps (add the retry=n option). 4) Restart automountd # /etc/init.d/autofs start Reboot system after patch installation.