This chapter describes known runtime problems.
The following runtime bug descriptions have been added to this chapter since this document was published on the Solaris 8 1/01 Documentation CD and in the Installation Kiosk on the Solaris 8 1/01 Installation CD.
The name of this product is Solaris 8 1/01, but code and path or package path names may use Solaris 2.8 or SunOS 5.8. Always follow the code or path as it is written.
If privileged users use "here documents" ("<<" redirection) in csh, it exposes a potential security exploit by non-privileged users.
Workaround: Privileged users should not use "here documents" in csh.
When mounting a diskless client on a multi-homed server, mount the client on a network interface that is on the same subnet as the diskless client. By default, a diskless client mounts its file systems using the OS server's local host interface.
On a multi-homed server, each network interface has a host name and an Internet Protocol (IP) address. To determine the host name of the local host interface, type uname -n on the server.
Use the -D type/host_name/domain_name option to specify the management domain, where:
type is NIS, NIS+, or file
host_name is the name of the host machine or network interface
domain_name is the name of the management domain
If you do not specify the -D option, SMC assumes the file domain on the local server.
Use the -o host_name option to specify the name of the OS server, when the name service server and the OS server are not the same machine. If you do not specify this option, smdiskless assumes that the OS server is the same as that specified with the -D option.
For more information, see smdiskless(1M).
If the name service server is the same as the multi-homed OS server, the host_name must be equal to the host name returned by typing uname -n on that server. Specify the -o option using the host name of the network interface that is on the the same subnet as the diskless client.
If the name service server is different from the multi-homed OS server, the host_name must be equal to the host name returned by typing uname -n on the name service server. Specify the -o option using the host name of the OS server's network interface that is on the same subnet as the diskless client.
If a multi-homed OS server's "local host" interface is on the diskless client's subnet and you do not specify the -D option (file scope is assumed), then smdiskless works correctly.
If a multi-homed OS server is not on the same subnet as the diskless client, you must edit the OS server's and diskless client's database files, as follows:
All entries must appear on a single line.
Change the following server files:
/etc/bootparams diskless_client root=server:/export/root/diskless_client \ swap=server:/export/swap/diskless_client swapsize=:32 \ dump=server:/export/dump/diskless_client dumpsize=:32 \ boottype=:di |
where server = Host name of OS server, net_interface = Host name of network interface on diskless client's subnet, and diskless_client = Host name of diskless client
"dump" and "dumpsize" appear only if you specify the "-x dump" or "-x dumpsize" options.
To the following:
diskless_client root=net_interface:/export/root/diskless_client \ swap=net_interface:/export/swap/diskless_client swapsize=:32 \ dump=net_interface:/export/dump/diskless_client dumpsize=:32 \ boottype=:di |
The net_interface host name and its IP address must appear in /etc/hosts.
Restart in.rarpd and rpc.bootparamd on the server. Use in.rarpd -a or start the rarpd daemon on the network interface used by the diskless client.
Change the following client files:
/export/root/diskless_client/etc/vfstab server :/export/root/diskless_client - / nfs - - rw server :/export/swap/diskless_client - /dev/swap nfs - - - server :/export/exec/Solaris_8_sparc.all/usr - /usr nfs - -ro |
To the following:
net_interface:/export/root/diskless_client - / nfs - - rw net_interface:/export/swap/diskless_client - /dev/swap nfs - - - net_interface:/export/exec/Solaris_8_sparc.all/usr - /usr nfs - -ro |
A problem occurs when compiling a Motif program in the Solaris 8 operating environment when you link to a shared library that has been compiled in the Solaris 2.4, 2.5, 2.5.1 or 2.6 operating environments and the older library also uses the Motif Application Programming Interface (API).
The Motif program uses Motif version 2.1 and the old shared library uses Motif version 1.2. A core dump occurs. This is not a binary compatibility problem for applications compiled in the Solaris 2.4, 2.5, 2.5.1, 2.6 operating environments, which should run correctly in the Solaris 8 operating environment.
Workaround: If you have an older shared library that links directly to the Motif library, and if you want to compile a program in the Solaris 8 operating environment that links to both Motif and that older shared library, use a line like this to compile:
cc foo.c -o program -DMOTIF12_HEADERS -I/usr/openwin/include -I/usr/dt/include -lXm12 -lXt -lX11 |
A CD-ROM with an unusual data layout might fail to mount automatically. The File Manager window might not display the CD-ROM's contents after it is inserted into the drive.
Workaround: Mount the CD-ROM manually:
Become root.
Stop the volume manager:
/etc/init.d/volmgt stop |
Mount the CD-ROM:
mount -F hsfs -r device_pathname /mnt |
Where device_pathname is the pathname to where in the system the CD-ROM drive is connected to. For example, /dev/dsk/c0t6d0s0.
CD-ROM drives are typically connected to c0t6d0s0 or to c0t2d0s0 at the factory, but you must use the correct pathname for this command to work.
Restart the volume manager:
/etc/init.d/volmgt start |
If you insert a diskette into its drive on a system with SCSI removable media devices and then select Check for Floppy from the File menu in OpenWindowsTM File Manager, File Manager mounts the diskette in the /floppy directory, but fails to display a File Manager view listing the disk contents. The Format Floppy and Eject Floppy options do not appear in the File menu of File Manager.
Workaround: Choose one of the following workarounds.
To view the contents of a diskette, follow these steps:
Click on the / folder in the File Manager Iconic Path.
Double-click on the floppy folder in the / display window.
Double-click on the floppy0 folder in the /floppy display window.
To format a diskette, follow these steps:
Unmount the diskette.
% volrmmount -e floppy0 |
Format the diskette.
% fdformat floppy0 |
To create a new file system on a diskette, follow these steps:
If you have already unmounted the diskette, go to step 2 of this workaround.
Unmount the diskette.
% volrmmount -e floppy0 |
Create the appropriate file system on the diskette.
To create a new UFS file system on the diskette, use the newfs command:
% newfs /vol/dev/aliases/floppy0 |
To create a PCFS file system on the diskette, use the mkfs command:
% mkfs -F pcfs /vol/dev/aliases/floppy0 |
Mount the diskette.
% volrmmount -i floppy0 |
To eject the diskette, use the eject command.
% eject floppy0 |
To prevent this problem, apply patch 109463-01.
After deleting the last item from the desktop (for example, the last appointment in your Calendar or the last address in the Address Manager), it is restored from the handheld device to the desktop when you synchronize your handheld device.
Workaround: Manually delete the last entry from the handheld device prior to synchronization.
If you exchange multibyte data between a PDA device and Solaris CDE, the data may be corrupted in both environments.
Workaround: Back up your data on your personal computer with the PDA backup utility before you run the PDASync application. If you accidentally exchange multibyte data and corrupt that data, restore your data from the backup.
When the system boots the following error message may be logged to the system log file located in /var/adm/messages:
open(/var/run/rcm_daemon_lock) - No such file or directory rcm_daemon exit: errno = 2 |
This occurs only if /var is a separate partition.
Workaround: Ignore the message. It does not harm the system.
After selecting Help -> Information from the Front Panel, a list of obsolete files is returned. The correct file is S8FCSreleasenotes.
The following error message is displayed when memory is low:
CIM_ERR_LOW_ON_MEMORY |
Workaround: To reset the CIM Object Manager Repository:
Become superuser.
Stop the CIM Object Manager.
# /etc/init.d/init.wbem stop |
Remove the JavaSpaces log directory.
# /bin/rm -rf /var/sadm/wbem/log |
Restart the CIM Object Manager.
# /etc/init.d/init.wbem start |
You will lose any proprietary definitions in your datastore. You must recompile the MOF files that contain those definitions using the mofcomp command. For example:
# /usr/sadm/bin/mofcomp -u root -p root_password your_mof_file |
If you enumerate instances of the Solaris_FileSystem class by using CIM WorkShop or the WBEM APIs, then the CIMOM no longer runs and the following error message is displayed:
Attempted to complete RMI action enumInstances and received exception java.rmi.UnmarshalException: Error unmarshaling return header; nested exception is: java.io.EOFException |
Workaround: In superuser mode, restart the CIMOM by typing the following command:
# /etc/init.d/init.wbem start |
Secure 64-bit applications can only be preloaded from the secure directory /usr/lib/secure/64. The symlink /usr/lib/secure/64 -> /usr/lib/secure/sparcv9 is missing. As a result, you may receive this error message:
% LD_PRELOAD=libfoo.so.1 setxid-app ld.so.1: setxid-app: warning: libfoo.so.1: open failed: \ No such file in secure directories |
Workaround: To prevent this problem, choose one of the following workarounds:
Create two symbolic links: symlink /usr/lib/secure/64 -> /usr/lib/secure/sparcv9 and symlink /usr/lib/link_audit/64 -> /usr/lib/link_audit/sparcv9.
Apply patch 109147-08.
USB capable systems with small memory configurations might panic when booting from either disk, CD or a network when several USB devices are connected. One of the following panic messages will appear in this event:
panic[cpu0]/thread=1040800: main: unable to fork init. |
or
panic[cpu0]/thread=2a1000fdd40: BAD TRAP: type=31 rp=2a1000fd0a0 addr=c0 mmu_fsr=0 occurred in module "genunix" due to a NULL pointer dereference |
Workaround: If you have more than 4 USB devices, boot the system with only the USB keyboard and mouse connected. After you see the Solaris login prompt, connect the remaining USB devices.
Java Plug-in 1.2 is the default plug-in that runs Java 2 applets but not all Java 1.1 applets. If you require the Java Plug-in 1.1, you can download it from http://www.sun.com/solaris/netscape.
If you choose to have both Java Plug-in 1.1 and Java Plug-in 1.2 on the same system, you must follow the instructions for how to install Java Plug-in 1.1 and then configure your environment accordingly.
This procedure is outlined in the "Installing Java Plug-in" section of the Java Plug-in for Solaris Users Guide that is available from http://www.sun.com/solaris/netscape/jpis/usersguide_java_plugin.html.
libjvm.so does not function correctly if it is brought into the process using dlopen. It must be linked into the executable when that object is created.
Workaround: Explicitly name libjvm.so (-ljvm) and libjava.so (-ljava) as dependencies when you create an executable.
If the ab2admin command fails, the error message contains additional information besides command failed. For example, it may also include path not found or invalid ID.
Workaround: If the message command failed is displayed, make sure that the operation failed. For example, if the command you submitted should have deleted a collection from the AnswerBook2 database, type the following command to verify that the collection is displayed in the database.
# ab2admin -o list |
During the startup of an AnswerBook2TM server, the ab2cd script may display the following erroneous error message.
sort: can't read /tmp/ab1_sort.XXX: No such file or directory |
This error message states that the ab2cd script has not located any of the AnswerBook (Display PostScriptTM) collections on the CD.
Workaround: Ignore the error message.
Perform the following steps to print from any non-ISO8859-1 locale using the Font Downloader.
Log in to CDE.
Type fdl at the command line to start the Font Downloader.
Specify the printer by selecting Add from the Printer menu.
Select Font Bundle from the Download menu.
The font bundles are then downloaded to the specified printer, depending on what codeset is needed for printing.
kkcv is an input method server for Japanese BCP (4.x Binary Compatibility) applications. When LDAP is used, kkcv dumps core repeatedly with the error message:
cssd: The process "kkcv.sh" has been killed by sig#134, core dumped |
This problem occurs in all locales when 'ja' locale is installed using the Languages CD.
Workaround: If you do not need to use Japanese BCP feature, configure your system not to start input method servers for BCP applications:
Copy /usr/lib/css.d/cs00.sh to /etc/css.d.
# cp -p /usr/lib/css.d/cs00.sh /etc/css.d |
Edit /etc/css.conf and comment out the line of /usr/lib/css.d with `#'.
# /etc/css.conf |
Restart the CS monitor daemon.
# sh /etc/init.d/loc.ja.cssd stop # sh /etc/init.d/loc.ja.cssd start |
The European SMC toolbox does not display all of the tools when loaded. You may receive the following error message:
** Parsing error, line 1, uri http://fubar:898/toolboxes/smc/smc.tbx com.sun.xml.parser/P-076 Malformed UTF-8 char -- is an XML encoding declaration missing? |
Workaround: Run smc edit and either modify the default localized toolbox or create a new toolbox.
The context sensitive help in the applications Solaris Management Console and Web-Based Enterprise Management are not fully localized.
The Euro is not accessible in any UTF-8 locale using the standard key sequence ALTGr+E.
Workaround: Login to any ISO8859-15 locale and use Alt+E to access the Euro.
LucidaSansLat4 font aliases are not available so related error messages might appear when launching a Java application from any UTF-8 locale.
Workaround: Log into the ISO-1 equivalent of the locale and launch the Java application from there.
Some dead-key combinations do not work correctly in CDE. Also, names for months do not function correctly in the Calendar Manager in the Greek locale.
If you attempt to print extended characters when using Calendar Manager in a partial locale, the extended characters do not print correctly.
You cannot cut or paste Arabic text between an application or window running under en_US.UTF-8 in Arabic input mode and one running under ar_EY.ISO8859-6 in Arabic input mode.
When you right-click in any CDE application for a European locale, the CDE Extras drop-down menu does not display any options.
Complex Text Language (CTL) support for entering Hebrew, Arabic, or Thai has been implemented in en_US.UTF-8 and European UTF-8 locales, but is not supported in ja_JP.UTF-8, ko.UTF-8, also known as ko_KR.UTF-8, zh.UTF-8, which is also known as zh_CH.UTF-8, and zh_TW.UTF-8 locales.
Workaround: Use the en_US.UTF-8 locale if you need to enter Thai, Arabic, or Hebrew using CTL. If you want to enter those languages in Asian and Japanese UTF-8 locales:
Create a symbolic link to common CTL modules. In the case of ja_JP.UTF-8:
# cd /usr/lib/locale/ja_JP.UTF-8 # mkdir LO_LTYPE ; cd LO_LTYPE # ln -s ../../common/LO_LTYPE/umle.layout.so.1 ja_JP.UTF-8.layout.so.1 # mkdir sparcv9 ; cd sparcv9 # ln -s ../../../common/LO_LTYPE/sparcv9/umle.layout.so.1 ja_JP.UTF-8.layout.so.1 |
Edit the /usr/openwin/lib/locale/ja_JP.UTF-8/XLC_LOCALE file by commenting out the load_option delay_nocheck line from Thai, Arabic, or Hebrew entries. For example, in the case of Thai:
# fs14 class (Thai) fs14 { charset TIS620.2533-0:GR font { # load_option delay_nocheck <--- comment out primary TIS620.2533-0:GR }} |
Several screens in the Smart Card application are not fully localized.
The Add, Modify, and Remove User screens are blank in the Greek locale of the Solstice AdminTool software.
Workaround: In superuser mode, copy the following file :
# cp /usr/openwin/lib/locale/C/app-defaults/Admin /usr/openwin/lib/locale/el_GR.ISO8859-7/app-defaults/Admin |
When you are in the Italian locale using the Font Downloader, both the Add and Cancel buttons in the Add Printer dialog box are incorrectly labeled; they are both labeled A ....
The left button should be labeled Aggiungi (Add).
The right button should be labeled Annulla (Cancel).
The following table describes the differences between the Sun Solaris Arabic keyboard and the Microsoft Arabic keyboard.
Table 2-1 Differences Between Sun and Microsoft Arabic Keyboards
Key |
Sun Keyboard Layout |
Microsoft Keyboard Layout |
---|---|---|
T |
T |
Arabic Lam_alef with Hamza below |
U |
U |
Right single quotation mark |
I |
I |
Arabic multiplication sign |
O |
O |
Arabic division sign |
A |
; |
Arabic Kasra |
S |
S |
Arabic Kasratan |
Z |
Z |
Tilde |
X |
X |
Arabic Sukun |
C |
Arabic Kasratan |
Left curly bracket |
V |
Arabic Kasra |
Right curly bracket |
M |
Sukun |
Single low quotation mark |
< |
< |
Arabic comma |
SEAM uses some of the resource files in the Solaris 8 operating environment, but only when the Kerberos settings are selected during an installation.
The Euro currency symbol is not generated when pressing AltGr+E in the UTF-8 locale.
Workaround: Perform the following steps to enter the Euro currency symbol in the UTF-8 locale:
Select Lookup in the UTF-8 Input Mode Selection window.
Select Currency Symbols.
Select the Euro symbol.
In the Greek locale type dumpcs at the console prompt. Then copy and paste the Euro currency symbol.
Sorting in the European UTF-8 locales does not work properly.
Workaround: Before you attempt to sort in a FIGGS UTF-8 locale, set the LC_COLLATE variable to the ISO1 equivalent.
# echo $LC_COLLATE > es_ES.UTF-8 # setenv LC_COLLATE es_ES.IS08859-1 |