This chapter contains advice for troubleshooting, solutions for commonly encountered symptoms, and descriptions of error messages.
When you experience difficulty with the Token Ring network connection at a particular station, do the following:
Check the physical connections at the station as detailed in “Checking Physical Connections”.
If the Token Ring network connection has never been functional (for example, you have just installed the product), use /sbin/hinv and /usr/etc/netstat -ina to check the following configuration parameters:
IRIS Token Ring controller must be listed by hinv.
Transmission speed shown by hinv must match the attached ring's speed.
IRIS Token Ring network interface (for example, mtr0, fv0, or gtr0) must be listed by netstat -ina.
IP address shown by netstat must be correct for the attached local area network.
Token Ring network interface must be enabled (for example, the netstat listing does not have an asterisk next to it).
If any of these parameters are incorrect, follow the installation and configuration instructions to correct the problem.
If the symptom is one of those covered in “Symptoms” follow the instructions provided.
Search or read the /usr/var/adm/SYSLOG file and console window (or startup screen) for error messages. If you find any IRIS Token Ring driver messages, read about them in “Error Messages”.
Note: Messages from an IRIS Token Ring driver contain the name of the driver (for example, fv0, gtr0, fv1, mtr0). Ask your network administrator to troubleshoot the ring by following your site's fault-isolation procedure.
To check the physical connection between the IRIS and the Token Ring network, perform these steps:
At the I/O panel for the IRIS Token Ring board, verify that the adapter cable's connector is seated firmly into the board's connector.
Caution: For I/O panels that have two connectors (RJ-45 and DB-9), only one of the connectors must be used. The Token Ring board for the IRIS Indigo system is a board with this design. Connecting both connectors can cause irreparable damage to the board. Follow the adapter cable from the Token Ring board's I/O panel to the cable's other end. Verify that the connector at this end is firmly seated into the connection. The connection at this end may be to a faceplate in the wall or to a trunk coupling unit (for example, a MAU).
Note: Physical connections past this point are handled by your site's fault-isolation or problem determination procedures. If you found any of the connections unplugged or loose, or if you unplugged them during the verification process, shutdown the system, turn off the power, then restart the system.
Note: A loose connection causes the IRIS Token Ring board to fail when it attempts to insert itself onto the ring. When the driver notices these failures, it disables the Token Ring interface and the board. If the board is disabled during startup, the network interface is not configured. To reenable the board, you must turn off the power and restart the system; a shut down or reboot does not work.
The NetVisualyzer product is useful for collecting and displaying Token Ring statistics. NetVisualyzer collects and graphically displays statistics about all data on the ring.
Silicon Graphics' Token Ring implementation supports promiscuous mode or capture-all-frames (CAF) for use with NetSnoop, one of the tools available with NetVisualyzer. Promiscuous mode is enabled automatically whenever NetVisualyzer is started.
NetSnoop is useful in diagnosing network problems; it copies all frames on the physical medium (cable), regardless of destination, and then analyzes and displays the contents. This information can be used to pinpoint an overloaded gateway, monitor packet errors, identify remote users, collect network statistics, and so on. This is useful for network planning and performance monitoring.
For more information about NetVisualyzer and NetSnoop, see the NetVisualyzer User's Guide or contact your local Silicon Graphics sales representative.
This section describes some common problems and provides step-by-step instructions for resolving them.
If an expected board or network interface is not listed by /sbin/hinv, the operating system did not recognize the board during the last reboot. If this occurs, follow these steps:
Did you install the IRIS Token Ring software? If you did, proceed.
If not, follow the instructions in the IRIS Token Ring Release Notes to install the software.
Did you rebuild the operating system (by restarting or rebooting the system or by invoking the /etc/init.d/autoconfig command) and then did you reboot? If you rebooted twice or invoked autoconfig then rebooted, proceed.
If not, use these commands to rebuild the operating system:
# /etc/init.d/autoconfig Automatically rebuild the operating system (y/n)? y # /etc/reboot
Verify that your network configuration (/etc/hosts file and /etc/config/netif.options file) has been set up correctly. For example, the /etc/hosts file must contain one line for each network connection (Token Ring, Ethernet, FDDI, and so on); the /etc/config/netif.options file must contain uncommented lines if the network connection names and ordering do not conform to the default configuration (summarized in Table 2-2 and Table 2-3). If the configuration is correct, proceed.
If there is a problem with the configuration, correct it, reboot the system, and retry the verification process.
If the configuration information is correct, it is possible that the missing Token Ring board is not seated firmly into its connection to the system. Follow the instructions in the board's installation guide to reinstall the board. Take extra precautions in seating the board into its connector.
If you have done all of the above, and the missing Token Ring board still does not appear, it is possible that the board is defective. Contact Silicon Graphics' Technical Assistance Center.
Follow the instructions in this section to enable (bring up) the IRIS Token Ring network interface.
Use ifconfig or netstat to verify that the Token Ring network interface is really disabled:
# /usr/etc/ifconfig interfacename
Note: If the display shows an UP flag, the interface is enabled and there is no problem. If the display does not list UP, the interface is disabled. # /usr/etc/netstat -in
Note: If the display shows an asterisk next to the Token Ring interface (for example, gtr0*), the interface is disabled. If the Token Ring network interface is enabled, but the network connection is not functional, follow the instructions in “When Token Ring Connection Is Not Responsive”. Otherwise proceed to the next step.
If the Token Ring network interface is disabled, use the command below to verify that the configured data transmission speed matches the ring's speed:
# /sbin/hinv
The speed indicated for the IRIS Token Ring board must match the speed at which your Token Ring local area network operates. If the speed indicated in the listing is not correct, follow the instructions in “Configuring the Board's Data Transmission Speed” to make the board's speed match the ring's.
If the speed is correct, proceed to the next step.
Try enabling the Token Ring network interface with this command:
# /usr/etc/ifconfig interfacename up
Repeat step 1. If the interface is still disabled, proceed to the next step.
Use the command below to display the configuration:
# /usr/etc/netstat -in
Is the value shown in the Address column zero (0.0.0.0)? If so, follow the instructions in “Checking Physical Connections” to resolve the problem. If the problem persists after verifying the physical connections, continue to the next step.
Is the value shown in the Address column correct for the Token Ring local area network? If not, follow the instructions in “Configuring the Network Interfaces” to correct the incorrect configuration.
Configuration areas that commonly cause problems include the following:
The /etc/hosts file does not contain the correct address. For example, if the word none appears in the Address column, the /etc/init.d/network startup script did not find an entry in the /etc/hosts file for the name it was expecting (for example, tr-hostname).
The lines in the edited (non-default) /etc/config/netif.options file are not correct (for example, they contain extra spaces or colons or network connection names that do not exist in the /etc/hosts file or incorrect network interface names).
Silicon Graphics computers are capable of loading (installing) a small-sized version of the operating system (the miniroot) and booting themselves over the network; however, they do this only over Ethernet local area networks. They cannot boot over Token Ring networks. Since Silicon Graphics systems always use their primary network interface for booting over the network, the miniroot can be loaded only when the primary network interface is an Ethernet connection.
If your system is unable to load the miniroot (or boot over the network), verify that its primary network interface is an Ethernet connection by following these instructions:
Restart the system from the System Maintenance menu. Do not rebuild the operating system during this restart.
Log on and open a shell window.
Use this command to display the ordering of the network interfaces:
% /usr/etc/netstat -i <primary interface> <secondary interface> ...
If the primary interface is an Ethernet (for example, ec0, et0, enp#), the Ethernet network connection may be dysfunctional. See the IRIX Admin: Networking and Mail guide for information about Ethernet network connections.
If the primary interface is not an Ethernet, proceed to the next step.
Configure an Ethernet connection as the primary interface, following the instructions in “Configuring the Network Interfaces”.
Reboot the system. When the system is up and running, it should be capable of loading the miniroot over the Ethernet network and booting from it.
When the IRIS Token Ring connection is not servicing requests for its services, it is possible that the board has transitioned into a state intended to prevent dysfunctional equipment from disrupting the ring with infinite insertion attempts. This state can be caused by a broken cable between the trunk coupling unit and the computer or by disconnecting the adapter cable from either the wallplate or the computer's I/O panel. Simply reconnecting the cable while the computer is powered up does not change the board's state.
| Note: Once the computer is started, do not disconnect the adapter cable from the computer's I/O panel or the wallplate. This action causes the IRIS Token Ring adapter board to transition into a non-responsive state. |
The following characteristics describe this non-responsive state:
The Token Ring network connection fails to pass the ping verification test described in “Verifying IP over Token Ring”.
The /usr/etc/netstat -i command lists the IRIS Token Ring interface as enabled.
The /usr/etc/ifconfig command lists the IRIS Token Ring interface as UP.
Invoking the /usr/etc/ifconfig command to stop and then start the Token Ring driver does not fix the problem.
The /sbin/hinv command lists the IRIS Token Ring board, and the displayed transmission speed matches the speed currently being used on the attached ring.
Invoking the /etc/shutdown command, then restarting the system does not remedy the problem.
Shutting down the system, turning off the power, then turning on the power, and restarting the system fixes the problem.
Take the following steps to remedy an unresponsive network interface:
Use the command below to stop and then start the Token Ring network interface:
# /usr/etc/ifconfig interfacename# down up
Try to use the network connection. One method for doing this is described in “Verifying IP over Token Ring”.
If this does not fix the problem, use the following command to shut down the system:
# /etc/shutdown
When you are informed that it is safe to do so, turn off the power switch, count to 10, then turn on the power switch.
Restart the operating system.
Try to use the network connection. If the problem is not resolved, troubleshoot the Token Ring adapter board as explained in the board's installation instructions.
This section contains an alphabetical list of all error messages specific to the IRIS Token Ring drivers. Each entry has a discussion of the problems the message may indicate and suggested actions.
| Note: The list of error messages in this section covers only those unique to the IRIS Token Ring drivers. Standard system error messages, even when caused by the driver code, are not covered in this section. |
Messages are organized according to the following rules:
Messages are alphabetized. (Figure 3-1 illustrates the text of an error message.)
Nonletters (for example, - or %) and blank spaces are shown in the text of the message, but are ignored in alphabetization. For example, the message tr_open appears between trnet and tp.
When an error message includes a variable that the software specifies differently (fills in) for each instance of the message, the variable is given with a generic name (for example, filename) and represented in italic font. Generic names such as filename or hostname are not alphabetized; the variable is skipped and the message is alphabetized using the next word. For example, the error message goofy not responding is listed as hostname not responding, and it is alphabetized among the “n” listings; hostname is skipped and the word “not” is used for alphabetizing the message. Common names for variables in this listing include #, digit, and hexnumeral.
Note: If you cannot find an error message in the alphabetical listing, identify potential variables, then look up the message without those words. Capitalization is not considered in alphabetization.
The creator of each message is listed, in angled brackets, below the text of the message: (<creator>).
Driver error messages and IRIS Token Ring commands are written into the file /usr/var/adm/SYSLOG or displayed at the terminal during startup or the Console during normal operation; some messages appear in more than one location. Within the SYSLOG file, each message is preceded by the date, time, host name, the name of the process that created the message, and its process ID number, as illustrated below. Only the text of the error message is included in the alphabetic list that follows.
This section describes error messages displayed by the fv driver.
fv#: Address verification phase: <message from Table 3-1> |
During startup, a failure occurred when the IRIS Token Ring board was attempting to verify the uniqueness of its MAC address. The reason for the failure is provided in the final segment of the error message, as summarized in Table 3-1.
Table 3-1. fv Startup Phase Failures
Text of Message | Description |
|---|---|
Duplicate node address | This station's MAC address is being used by another station on the ring. |
Function failure | The station is unable to transmit to itself when its transmit and receive lines are wrapped at the wiring concentrator. |
Remove received | A “remove adapter” MAC frame was received. |
Request initialization | This station was unable to obtain ring parameters from the ring's parameter server. If the ring does not have a parameter server, this message is not displayed and the system uses internal default parameters. |
Ring beaconing | A beacon MAC frame was seen. |
Ring failure | The station became the active monitor and began purging the ring, but subsequently did not see its own MAC frames return to it. |
Signal loss | This station cannot see any signal on the physical medium at its input port. |
Timeout | The station's insertion timer for this phase expired before the station could complete the startup phase. |
Ask your network administrator to troubleshoot the ring, using your site's fault-isolation or problem determination procedures.
fv#: bad burnin address
While attempting to start, the driver discovered that the MAC address burned into a component on the IRIS Token Ring board is invalid.
Contact Silicon Graphics' Technical Assistance Center.
fv#: bad EDT entry
While attempting to start, the driver encountered invalid initialization (EDT) parameters. This indicates a problem with the software.
If this message displays during an installation, reinstall the software, taking extra care to install the correct version for the installed IRIS Token Ring board; otherwise, contact Silicon Graphics' Technical Assistance Center.
fv#: broadcast address set to 0xC000FFFFFFFF
During startup, the driver has set the broadcast address to the value indicated. There is no problem.
You do not need to do anything; however, if you want to change this broadcast address to a different one, use the commands below:
% su
Password: thepassword
# /usr/etc/ifconfig fv# broadcast 0xnewaddress
where newaddress is a 6-byte value in hexadecimal format (for example, FFFFFFFFFFFF) and fv# is the specific fv network interface.
fv#: died cmd=hexnumeral, rc=hexnumeral
The driver received an interrupt from the board (of type ADAPTER.CHECK), indicating that the board has experienced an unrecoverable error. The rc value indicates the reason for the interrupt, as described by the error bits in the Adapter Check Status register (described in Figure 4-19 and Table 4-15 of the Texas Instrument, TMS380 Second-Generation Token Ring User's Guide.)
Contact Silicon Graphics' Technical Assistance Center.
fv#: early interrupt
The driver received an interrupt from the board before it was ready to process that interrupt.
When this message is intermittent, no action is needed. If the message is persistent, shutdown and restart the system. If the message continues to be displayed, contact Silicon Graphics' Technical Assistance Center.
fv#: executing CAF
The driver enabled the collect-all-frames (CAF)/promiscuous mode. This message does not indicate a problem.
You do not need to do anything.
fv#: firmware failed to start
While attempting to start, the driver was unable to start the firmware on the board. The board is dysfunctional; it may be incorrectly configured, or the board's firmware version may not match the board.
If this message displays during an initial installation, verify that the installation steps have been followed correctly; otherwise, contact Silicon Graphics' Technical Assistance Center.
fv#: invalid buffer size
During startup, the driver encountered an invalid buffer. This indicates an incompatibility between the driver and other IRIX software modules.
Contact Silicon Graphics' Technical Assistance Center.
fv#: invalid list size
During startup, the driver encountered an invalid parameter. This indicates an incompatibility between the driver and other IRIX software modules.
Contact Silicon Graphics' Technical Assistance Center.
fv#: invalid LLC options
During startup, the driver was passed logical link control (LLC) layer options that it does not support. This indicates an incompatibility between the driver and other IRIX software modules.
Contact Silicon Graphics' Technical Assistance Center.
fv#: invalid node address
During startup, the driver found its own MAC address to be invalid. This usually indicates an improper entry in the /etc/config/trconfig.options file.
Verify the entries in the /etc/config/trconfig.options file.
fv#: IP open: <any of the messages below>
a parameter has exceeded max limit
group has too many members
invalid options
invalid SAP value
nonexistent group
resources not available
unauthorized access priority
While attempting to process a service request, the driver was unable to open a communication path to the IP protocol stack because the service access point (SAP) into the IP protocol stack failed to open. The reason for the failure is indicated in the final segment of the message. All of these messages indicate incompatibilities between the driver and a user-level application; the user-level application has used invalid/unsupported parameters.
If this message occurs with a user-level application that is sold or distributed by Silicon Graphics, contact Silicon Graphics' Technical Assistance Center; otherwise, contact the vendor of the user-level product.
fv#: Lobe media test phase: <message from Table 3-1>
During startup, a failure occurred when the IRIS Token Ring board was performing its lobe media test. This test is used to verify the connection to the trunk coupling unit (MAU or CAU). The reason for the failure is provided in the final segment of the error message, as summarized in Table 3-1.
Ask your network administrator to troubleshoot the ring, using your site's fault-isolation or problem determination procedures.
fv#: missing
When the fv driver started, it did not find the IRIS Token Ring board. This indicates that the driver is built into the operating system (kernel) and starts with each reboot or restart of the system, but the board is unreachable.
If the board is not installed and you want the Token Ring connection to function, contact your Silicon Graphics' field service engineer to install the board.
If the board has been removed, this message is not a problem, although it does mean that the Token Ring connection is not functional. If you do not plan to reinstall the board, you can rebuild the operating system (at startup or with autoconfig) to remove this driver so that this message does not display. No harm occurs if you do not remove the driver from the operating system.
If the board is installed, it is possible that it has become loose from its connection or that it is dysfunctional. Have your Silicon Graphics' field service engineer reinstall the board, or contact Silicon Graphics' Technical Assistance Center.
fv#: packet too small(digit)
While receiving an incoming frame (from the board), the driver found that it was too small to be legal. The packet is thrown away automatically.
You do not need to do anything.
fv#: Participation in ring poll phase: <message from Table 3-1>
During startup, a failure occurred when the IRIS Token Ring board attempted to participate in the ring's poll process. During this startup phase, the board attempts to determine if it should be a standby monitor or become the active monitor. The reason for the failure is provided in the final segment of the error message, as summarized in Table 3-1.
Ask your network administrator to troubleshoot the ring, using your site's fault-isolation or problem determination procedures.
fv#: Physical insertion phase: <message from Table 3-1>
During startup, the IRIS Token Ring board's physical insertion phase failed. During this phase, the board attempts to synchronize itself with the received signal and the ring's signal clock. The reason for the failure is provided in the final segment of the error message, as summarized in Table 3-1.
Ask your network administrator to troubleshoot the ring, using your site's fault-isolation or problem determination procedures.
fv#: Request Initialization phase: <message from Table 3-1>
During startup, the IRIS Token Ring board's attempt to contact the ring's parameter server failed. The reason for the failure is provided in the final segment of the error message, as summarized in Table 3-1.
Ask your network administrator to troubleshoot the ring, using your site's fault-isolation or problem determination procedures.
fv#: reset FAILED
The driver attempted to reset (restart) the IRIS Token Ring board and failed to get a response. The board may be loose from its connections or it may be dysfunctional.
If this message displays during an initial or routine installation, reinstall the board, taking extra care to seat the board firmly into its connections; otherwise, contact Silicon Graphics' Technical Assistance Center.
fv#: spurious interrupt = hexnumeral
The driver received an unrecognized and unrecoverable interrupt from the board. The value indicated by hexnumeral indicates the unrecognized STATUS interrupt.
When this message is intermittent, nothing need be done. If the message is persistent, shutdown and restart the system. If the message continues to be displayed, contact Silicon Graphics' Technical Assistance Center.
fv#: tokenring address hexnumerals
The MAC address for the IRIS Token Ring board is displayed by hexnumerals. This message is informational; it does not indicate a problem.
You do not need to do anything.
fv#: too few recv buffers
During startup, the driver was unable to obtain enough memory. This indicates an incompatibility between the driver and other IRIX software modules.
Contact Silicon Graphics' Technical Assistance Center.
Many of the gtr error messages indicate problems that occur during different phases of the IRIS Token Ring board startup procedure. These phases happen in the following order: power-up diagnostics (PUD), bring-up diagnostics (BUD), initialization, open, and insert.
gtr#: ADAPTER checksum error
The driver received an interrupt from the board (of type ADAPTER.CHECK), indicating that the board has experienced an unrecoverable error. The reason for the interrupt was an invalid or undecipherable operation code, as described in Figure 4-19 and Table 4-15 of the Texas Instrument, TMS380 Second-Generation Token Ring User's Guide.
Contact Silicon Graphics' Technical Assistance Center.
gtr#: ADAPTER DIO parity error
The driver received an interrupt from the board (of type ADAPTER.CHECK), indicating that the board has experienced an unrecoverable error. The reason for the interrupt was bad parity, as described in Figure 4-19 and Table 4-15 of the Texas Instrument, TMS380 Second-Generation Token Ring User's Guide.
Contact Silicon Graphics' Technical Assistance Center.
gtr#: ADAPTER DMA read error
The driver received an interrupt from the board (of type ADAPTER.CHECK), indicating that the board has experienced an unrecoverable error. The reason for the interrupt was an abort during a direct memory access (DMA) read, as described in Figure 4-19 and Table 4-15 of the Texas Instrument, TMS380 Second-Generation Token Ring User's Guide.
Contact Silicon Graphics' Technical Assistance Center.
gtr#: ADAPTER DMA underrun
The driver received an interrupt from the board (of type ADAPTER.CHECK), indicating that the board has experienced an unrecoverable error. The reason for the interrupt was an underrun during a direct memory access (DMA), as described in Figure 4-19 and Table 4-15 of the Texas Instrument, TMS380 Second-Generation Token Ring User's Guide.
Contact Silicon Graphics' Technical Assistance Center.
gtr#: ADAPTER DMA write error
The driver received an interrupt from the board (of type ADAPTER.CHECK), indicating that the board has experienced an unrecoverable error. The reason for the interrupt was an abort during a direct memory access (DMA) write, as described in Figure 4-19 and Table 4-15 of the Texas Instrument, TMS380 Second-Generation Token Ring User's Guide.
Contact Silicon Graphics' Technical Assistance Center.
gtr#: ADAPTER internal bus error
The driver received an interrupt from the board (of type ADAPTER.CHECK), indicating that the board has experienced an unrecoverable error. The reason for the interrupt was a parity error on the adapter's internal bus, as described in Figure 4-19 and Table 4-15 of the Texas Instrument, TMS380 Second-Generation Token Ring User's Guide.
Contact Silicon Graphics' Technical Assistance Center.
gtr#: ADAPTER invalid error interrupt
The driver received an interrupt from the board (of type ADAPTER.CHECK), indicating that the board has experienced an unrecoverable error. The reason for the interrupt was an unrecognized error interrupt, as described in Figure 4-19 and Table 4-15 of the Texas Instrument, TMS380 Second-Generation Token Ring User's Guide.
Contact Silicon Graphics' Technical Assistance Center.
gtr#: ADAPTER invalid interrupt
The driver received an interrupt from the board (of type ADAPTER.CHECK), indicating that the board has experienced an unrecoverable error. The reason for the interrupt was an unrecognized or invalid interrupt, as described in Figure 4-19 and Table 4-15 of the Texas Instrument, TMS380 Second-Generation Token Ring User's Guide.
Contact Silicon Graphics' Technical Assistance Center.
gtr#: ADAPTER invalid xop
The driver received an interrupt from the board (of type ADAPTER.CHECK), indicating that the board has experienced an unrecoverable error. The reason for the interrupt was an unrecognized XOP request, as described in Figure 4-19 and Table 4-15 of the Texas Instrument, TMS380 Second-Generation Token Ring User's Guide.
Contact Silicon Graphics' Technical Assistance Center.
gtr#: bad EDT entry
When starting, the driver found that the unit number (EDT entry) for the IRIS Token Ring board is invalid. This indicates a software problem.
Contact Silicon Graphics' Technical Assistance Center.
gtr#: bd not responding
When the driver attempted to issue a command to the board, the board did not respond. This indicates a hardware problem. The board may be loose from its connection or it may be dysfunctional.
First reboot the system. If the problem persists, follow the instructions in the Token Ring for IRIS Indigo Installation Guide to reinstall the board, taking extra care to seat the board firmly into its connections. If the message continues to display, contact Silicon Graphics' Technical Assistance Center.
gtr#: BUD: message (hexnumeral)
During the board's bring up diagnostic (BUD) phase, a problem was encountered with the Texas Instrument components. The cause of the problem is indicated by the text of the message and the error code (hexnumeral). The messages and error codes are documented in the section “Bring-up Diagnostics” of Texas Instruments' TMS380 Second-Generation Token Ring User's Guide. The IRIS Token Ring board is dysfunctional.
Contact Silicon Graphics' Technical Assistance Center.
gtr#: CLOSE failed(hexnumerals)
When the driver attempted to close the adapter on the IRIS Token Ring board, an error occurred, indicating that the adapter component is dysfunctional. The value displayed by hexnumerals corresponds to the return codes documented in the Texas Instrument, TMS380 Second-Generation Token Ring User's Guide.
Contact Silicon Graphics' Technical Assistance Center.
gtr#: died and restarted(hexnumeral)
The driver was unable to contact the IRIS Token Ring board, so it reset the board. Subsequent error messages will indicate the success or failure of the reset.
If the reset succeeds, nothing is wrong and nothing needs to be done. If the reset fails, follow the instructions for that failure message.
gtr#: duplicate EDT entry
When starting, the driver found that the unit number (EDT entry) for this IRIS Token Ring board is already known to the system. This indicates a software problem.
Contact Silicon Graphics' Technical Assistance Center.
gtr#: EPROM checksum error(hexnumeral)
During the board's power up diagnostic (PUD) phase, the data within a memory component of the IRIS Token Ring board was found to be corrupted. The board is dysfunctional.
Contact Silicon Graphics' Technical Assistance Center.
gtr#: failure
The IRIS Token Ring board is dysfunctional.
Contact Silicon Graphics' Technical Assistance Center.
gtr#: INIT: message (hexnumerals)
During initialization of the board's adapter, an error occurred, indicating that the adapter component is dysfunctional. The reason for the failure is described in the text of the message and the error code (hexnumerals). (The messages and error codes are documented in the section “Adapter Initialization” of Texas Instruments' TMS380 Second-Generation Token Ring User's Guide.) The IRIS Token Ring board is dysfunctional.
Contact Silicon Graphics' Technical Assistance Center.
gtr#: init time out
The if_init call for initializing the board, timed out. This indicates a hardware problem. The board may be loose from its connection or it may be dysfunctional.
Follow the instructions in the Token Ring for IRIS Indigo Installation Guide to reinstall the board, taking extra care to seat the board firmly into its connections. If the message continues to display, contact Silicon Graphics' Technical Assistance Center.
gtr#: invalid MOD options(hexnumeral)
When the driver attempted to open the IRIS Token Ring board, an error occurred, indicating that the adapter component is dysfunctional, is not responding, or the options requested by the MAC OPEN command are not supported. Follow the instructions in “When Token Ring Interface Is Disabled” to troubleshoot the problem.
Contact Silicon Graphics' Technical Assistance Center.
gtr#: missing
When the gtr driver started, it did not find the IRIS Token Ring board. This indicates that the driver is built into the operating system (kernel) and starts with each reboot/restart of the system, but the board is unreachable.
Follow the instructions in one of the bullets below:
If the board is not installed and you want the Token Ring connection to function, follow the instructions in the Token Ring for IRIS Indigo Installation Guide to install the board.
If the board has been removed, this message is not a problem, although it does mean that the Token Ring connection is not functional. If you do not plan to reinstall the board, you can restart the system and answer yes to rebuild the operating system and remove this driver.
If the board is installed, it is possible that it has become loose from its connection or that it is dysfunctional. Follow the instructions in the Token Ring for IRIS Indigo Installation Guide to reinstall the board, taking extra care to seat the board firmly into its connections. If the message continues to display, contact Silicon Graphics' Technical Assistance Center.
gtr#: no memory for dma
When starting, the driver was unable to obtain system memory for itself. This indicates a software problem.
Contact Silicon Graphics' Technical Assistance Center.
gtr#: no memory for frame filter
When starting, the driver could not (or did not) allocate memory for a multicast filter table. This indicates a software problem.
Contact Silicon Graphics' Technical Assistance Center.
gtr#: OPEN: Address verification phase <message from Table 3-2>
During startup, an OPEN_ERROR occurred when the IRIS Token Ring board was attempting to insert itself into the ring and verify the uniqueness of its MAC address. The reason for the failure is provided in the final segment of the error message, as summarized in Table 3-2.
Table 3-2. gtr Startup Phase Failures
Text of Message | Description |
|---|---|
Duplicate node address | This station's MAC address is being used by another station on the ring. |
Function failure | The station is unable to transmit to itself when its transmit and receive lines are wrapped at the wiring concentrator. |
Remove received | A “remove adapter” MAC frame was received. |
Request initialization | This station was unable to obtain ring parameters from the ring's parameter server. If the ring does not have a parameter server, this message is not displayed and the system uses internal default parameters. |
Ring beaconing | A beacon MAC frame was seen. |
Ring failure | This station became the active monitor and began purging the ring, but subsequently did not see its own MAC frames return to it. |
Signal loss | This station cannot see any signal on the physical medium at its input port. |
Timeout | This station's insertion timer for this phase expired before the station could complete the startup phase. |
gtr#: OPEN: invalid buffer size
When the driver attempted to open the IRIS Token Ring board, an error occurred, indicating that the adapter component is dysfunctional. The failure occurred because the buffer size provided by the OPEN command is invalid.
Contact Silicon Graphics' Technical Assistance Center.
gtr#: OPEN: invalid list size
When the driver attempted to open the IRIS Token Ring board, an error occurred, indicating that the adapter component is dysfunctional. The failure occurred because the receive/transmit list size provided by the OPEN command is invalid.
Contact Silicon Graphics' Technical Assistance Center.
gtr#: OPEN: invalid LLC options
When the driver attempted to open the IRIS Token Ring board, an error occurred, indicating that the adapter component is dysfunctional. The failure occurred because the options requested by the LLC OPEN command are not supported.
Contact Silicon Graphics' Technical Assistance Center.
gtr#: OPEN: invalid node address
When the driver attempted to open the IRIS Token Ring board, an error occurred, indicating that the adapter component is dysfunctional. The failure occurred because the MAC address is either all zeros or unreadable.
Contact Silicon Graphics' Technical Assistance Center.
gtr#: OPEN: Lobe media test phase <message from Table 3-2>
During startup, a failure occurred when the IRIS Token Ring board was performing its lobe media test. This test verifies the connection to the trunk coupling unit (MAU or CAU). The reason for the failure is provided in the final segment of the error message, as summarized in Table 3-2.
Ask your network administrator to troubleshoot the ring, using your site's fault-isolation or problem determination procedures.
gtr#: OPEN: open failed (hexnumerals)
The driver cannot complete the open phase for the board. The IRIS Token Ring board is dysfunctional.
Contact Silicon Graphics' Technical Assistance Center.
gtr#: OPEN: Participation in ring poll phase <msg from Table 3-2>
During startup, an OPEN_ERROR occurred when the IRIS Token Ring board was attempting to insert itself into the ring and to participate in the ring's poll process. During this startup phase, the board attempts to determine if it should be a standby monitor or become the active monitor. The reason for the failure is provided in the final segment of the error message, as summarized in Table 3-2.
Ask your network administrator to troubleshoot the ring, using your site's fault-isolation or problem determination procedures.
gtr#: OPEN: Physical insertion phase <message from Table 3-2>
During startup, the IRIS Token Ring board's physical insertion phase failed. During this phase, the board attempts to synchronize itself with the received signal and the ring's signal clock. The reason for the failure is provided in the final segment of the error message, as summarized in Table 3-2.
Ask your network administrator to troubleshoot the ring, using your site's fault-isolation or problem determination procedures.
gtr#: OPEN: Request Initialization phase <msg from Table 3-2>
During startup, an OPEN_ERROR occurred when the IRIS Token Ring board was attempting to insert itself into the ring and to contact the ring's parameter server. The reason for the failure is provided in the final segment of the error message, as summarized in Table 3-2.
Ask your network administrator to troubleshoot the ring, using your site's fault-isolation or problem determination procedures.
gtr#: OPEN: too few recv buffers
When the driver attempted to open the IRIS Token Ring board, an error occurred, indicating that the adapter component is dysfunctional. The failure occurred because the request for transmit buffers provided by the OPEN command specified so many buffers that not enough were left for reception.
Contact Silicon Graphics' Technical Assistance Center.
gtr#: packet too small(digit)
When receiving a packet from the board, the driver discovered that the packet was too small to be a legal packet. The packet is discarded. This indicates that a system somewhere is generating illegal packets.
You do not need to do anything with this system; however, you may wish to use a network management tool, such as NetSnoop, to locate the faulty system.
gtr#: read BIA failed
During initialization, the driver was unable to read the burned-in MAC address (BIA) from the board, or the checksum validation of that address failed after the address was read. This indicates a problem with the board.
Contact Silicon Graphics' Technical Assistance Center.
gtr#: reset failed
The driver attempted to reset (start) the IRIS Token Ring board and failed to get a response. The board may be loose from its connections or it may be dysfunctional.
If this message displays during an installation, immediately after a hardware maintenance procedure has been performed, or after the computer has been moved, reinstall the board, taking extra care to seat the board firmly into its connections; otherwise, contact Silicon Graphics' Technical Assistance Center.
gtr#: RING INSERT failed
During initialization, the board failed to insert itself onto the ring. This indicates a problem with the ring or with the board.
First, shutdown the system, turn off the power, then restart the system. If the message displays again, have your network administrator verify that the ring is functional. If the problem persists, contact Silicon Graphics' Technical Assistance Center.
gtr#: RUN: died(hexnumeral)
The driver cannot insert the board into the ring. The IRIS Token Ring board is dysfunctional.
Contact Silicon Graphics' Technical Assistance Center.
gtr#: set address failed
During initialization, the driver was unable to set the functional MAC address. This indicates a problem with the locally assigned MAC address value controlled by the trconfig.options file.
Locate and correct the invalid entry in the trconfig.options file. Instructions for editing this file are provided in “Updating the MAC Address”.
gtr#: tokenring address hexnumerals
The MAC address being used for the IRIS Token Ring board (either the burned-in universally assigned one or the locally assigned one) is displayed by hexnumerals. This message is informational; it does not indicate a problem.
You do not need to do anything.
gtr#: tokenring address set to hexnumerals
The MAC address for the IRIS Token Ring board is displayed by hexnumerals. This message is informational; it does not indicate a problem.
You do not need to do anything.
gtr#: Unknown error (hexnumeral)
The driver received an interrupt from the board, indicating that the board has experienced an unrecoverable and unrecognized error. The value displayed by hexnumerals indicates which bits of the Status Register (documented in the Texas Instrument, TMS380 Second-Generation Token Ring User's Guide) were set at the time of the error. The IRIS Token Ring board is dysfunctional.
Contact Silicon Graphics' Technical Assistance Center.
gtr#: present
The gtr driver has located its IRIS Token Ring board. This message is informational; it does not indicate a problem.
You do not need to do anything.
The error messages for the mtr driver fall into two categories: the Alert and Warning messages and the Notice messages.
Alert and Warning messages occur due to a failure in the adapter, such as when it cannot initialize or function. Alert and Warning messages cannot be corrected without help from Silicon Graphics Technical Support. If you receive one of these messages, call the Silicon Graphics Technical Assistance Center.
Notice messages are not serious, but inform you of a condition that may need to be monitored. The adapter is initialized and working, but a problem has occurred that you should be aware of.
This section lists the Alert and Warning messages for the PCI Token Ring Board. Should you receive any of these messages, call the Technical Assistance Center and report the message.
mtr#: again for (hexnumerals): (hexnumerals)!
The driver has detected that the board has already been initialized.
mtr#: attach(): kmem_zalloc (#) ! mtr#: attach(): kmem_zalloc (#) 1! mtr#: attach(): kmem_zalloc (#) 2! mtr#: attach(): dev_desc!
The driver failed to allocate enough memory for its internal structure.
mtr#: bad burnin address: hexnumerals!
The driver has detected an incorrect burn-in address (BIA) on the board.
mtr# BUD: #, hexnumerals!
The PCI Token Ring board failed during bring up diagnostics (BUD).
mtr# DATA_32 no match #.#:
The driver could not correctly download the board firmware.
mtr_detach (hexnumerals) not found! mtr_detach (hexnumerals): (hexnumerals) not found!
A driver or system internal error has occurred.
mtr#: dmamap_addr(#) for RX! mtr#: dmamap_addr(#) for TX!
The driver has failed to allocate resources for the board.
mtr#: dmamap_alloc(#) for SSB & SCB!
The driver has failed to allocate resources for testing the bus master.
mtr#: dmamap_alloc(#) for TX & RX!
The driver has failed to allocate resources for the board.
mtr#: driver_prepare_adapter() kmem_zalloc(#) init_block!
The driver failed to allocate # bytes for the initialization block.
mtr#: fmplus_buffer_size (#) < FMPLUS_MIN_TXRX_BUFF! mtr#: fmplus_buffer_size: # < #!
An internal error has occurred within the driver that prevents it from correctly preparing or initializing the PCI Token Ring board.
mtr#: hwi_at24_wait_ack(): no ack: hexnumerals (hexnumerals)!
The PCI Token Ring board has failed to respond to the driver's request.
mtr#: hwi_download_code(): # != #! mtr#: hwi_download_code(): hexnumerals != 0!
The driver detected an error in the PCI Token Ring board firmware.
mtr#: IFF_PROMISC but driver_prepare_adapter: #! mtr#: IFF_PROMISC but driver_start_adapter: #!
The driver did not enable or disable the Snoop Protocol when restarting the PCI Token Ring board.
mtr#: IFF_UP but driver_prepare_adapter: #!
The driver did not prepare the PCI Token Ring board to restart correctly.
mtr#: IFF_UP but driver_start_adapter: #!
The driver did not restart the PCI Token Ring board.
mtr*: if_mtrstart(): no recognized adapter!
The mtr driver does not detect a PCI Token Ring board in the system.
mtr#: INIT failed: (hexnumerals).
The PCI Token Ring board failed during initialization.
mtr#: kmem_zalloc(#) for SSB & SCB!
The driver has failed to allocate resources for testing the bus master.
mtr#: kvpalloc(#) for TX & RX!
The driver has failed to allocate resources for the board.
mtr#: mtr_output (hexnumerals)! mtr#: mtr_watchdog (#)!
A driver or system internal error has occurred.
mtr#: no memory or io base register!
The driver failed to obtain the board's PCI base register.
mtr#: not support 4.31: hexnumerals! mtr#: not support bus master: hexnumerals!
The driver has detected the wrong version of the PCI Token Ring board.
mtr#: pciio_dmamap_addr(hexnumerals, hexnumerals, #) for SCB! mtr#: pciio_dmamap_addr(#) for SSB
The driver has failed to allocate resources for testing the bus master.
mtr#: pciio_intr_alloc (#) failed! mtr#: pciio_intr_connect () failed!
The driver failed to allocate or initialize the structure for handling an interrupt.
mtr#: pciio_piomap_alloc (#)
The driver failed to obtain the address to access the PCI Token Ring board.
mtr#: POLLING_SIFINT: (hexnumerals)!
The driver failed to receive the correct status from the PCI Token Ring board.
mtr#: possible lockup: #, (hexnumerals)!
A bus master operation has not completed correctly.
mtr#: RX slot idx: # >= #!
An internal error has occurred within the driver that prevents it from correctly preparing or initializing the PCI Token Ring board.
mtr#: scb_buff[#]: (hexnumerals) != scb_test_pattern[#] (hexnumerals). mtr#: ssb_buff[#]: (hexnumerals) != ssb_test_pattern[#] (hexnumerals).
The driver could not correctly perform the bus master test.
mtr#: SIFINT_ADAPTER_CHECK (hexnumerals)!
The PCI Token Ring board reports an unrecoverable error.
mtr#: SIOC...MULTI: srb_used: (hexnumerals).
The driver has an internal error that prevents the PCI Token Ring board from being correctly set up for IP multicast.
mtr#: too many adapters -> ignored!
Your system cannot support the number of PCI Token Ring boards you have installed. A maximum of eight PCI Token Ring boards are supported.
mtr#: TX slot idx: # >= #!
An internal error has occurred within the driver that prevents it from correctly preparing or initializing the PCI Token Ring board.
The following Notice messages provide information about an error condition.
mtr#: enter PROMISC mode. mtr#: exit PROMISC mode.
Promiscuous mode has been enabled or disabled in the driver.
mtr#: mtr_intr(): sifsts(hexnumerals) hexnumerals!
The driver has detected a null interrupt from the PCI Token Ring board. This may be due to heavy traffic on the ring.
mtr#: mtr_output(): sa_family (hexnumerals) not supported.
The driver has received a packet from an unsupported upper-layer protocol.
mtr#: mtr_snoop_input: m_vget(#).
The driver could not allocate the memory that is necessary to enable snooping. This may be due to heavy traffic.
mtr#: open_address.byte[#]: hexnumerals != mtr_enaddr[#] hexnumerals!
The node address (or open address) was not correctly set. This occurs when the MAC address has not been correctly configured. Repeat the procedure to configure the MAC address. See “Configuring the Board's Physical (MAC) Address”.
mtr#: open error(#, hexnumerals.hexnumerals)
The PCI Token Ring board could not be accessed. This could be due to a broken cable connection. Check cable connections between the board and the MAU and then restart the board.
mtr#: read bad frame: hexnumerals
The PCI Token Ring board has received a bad frame or the receiving operation itself has failed.
mtr#: ring_speed: hexnumerals but mtr_s16Mb (hexnumerals)!
The ring speed and the ring speed configuration for the board are not the same. Reconfigure the board with the correct ring speed. See “Configuring the Board's Data Transmission Speed”.
mtr# RX: m_get(#)! mtr# RX: m_vget(#).
The driver has failed to allocate memory for the received packet. The packet has been discarded.
mtr#: RX rx_len: #.#:# [#..#]!
The length of the packet just received is incorrect. The packet has been discarded.
mtr#: SIFINT_ARB (hexnumerals): arb: hexnumerals.hexnumerals open: hexnumerals.hexnumerals
The driver has reported a change in the ring status.
mtr#: hexnumerals(SIFINT_SRB) but #!
The driver has reported a harmless internal race condition.
mtr#: SIFINT_SSB (hexnumerals.hexnumerals)?
The driver has received an unexpected SSB from the PCI Token Ring board.
mtr#: SIOCSIFADDR: halt: #! mtr#: SIOCSIFADDR: start: #!
The driver has failed to configure the new MAC address for the board. Repeat the MAC configuration process. See “Configuring the Board's Physical (MAC) Address”.
mtr#: SRB(hexnumerals) result: hexnumerals.
The PCI Token Ring board failed to execute the command it received from the driver.
mtr#: TX pkt: # > #!
The size of an outbound packet is too large. The packet has been discarded.