HP V2500 - Manuals

vi Table of Contents Main menu. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78Test Configuration menu . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 86 Example of running diagnostics from Test Controller command line . . 93 Confi...

List of Figures xi Figures Figure 1 Location of the Utilities board . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .3 Figure 2 Utilities board . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .5 Figure 3 System display...

List of Tables xiii Tables Table 1 Environmental conditions monitored by the SMUC and power-on circuit . . .8 Table 2 Processor initialization steps . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .13 Table 3 Processor run-time status codes . . . . . . . . . . . . . . . . ....

Preface xvii Preface This document describes the offline diagnostics for V2500 servers. It isnot intended to be a tutorial or troubleshooting guide but a referenceguide that contains information on all utilties and scripts used totroubleshoot these systems. Notational conventions This section descri...

xviii Preface Preface Notational conventions NOTE A note highlights important supplemental information. CAUTION A caution highlights procedures or information necessary to avoiddamage to equipment, damage to software, loss of data, or invalid testresults. Brackets ( [ ] ) In command examples, square...

Chapter 1 1 1 Introduction This chapter presents an overview of the diagnostic mechanism forV2500 servers.

2 Chapter 1 Introduction Utilities board Utilities board The diagnostic mechanism in the V2500 servers is centered around theStingray Core Utilities board (SCUB). The SCUB is mounted under theMidPlane Interconnect board (MIB) toward the front of the system. SeeFigure 1.

6 Chapter 1 Introduction Utilities board Core logic The core logic contains initialization and booting firmware and isdescribed in the following sections. Flash memory The core logic contains a four-MByte electrically erasable programmableread only memory (EEPROM) storage for Processor-Dependent Cod...

Chapter 1 7 Introduction Utilities board Console ethernet The ethernet I/O port provides a connection to the teststation overLAN1. Attention lightbar and LCD The attention light bar displays environmental information, such as thesource of an environmental error that caused the Utilities board to pow...

8 Chapter 1 Introduction Utilities board SMUC environmental monitoring The following environmental conditions are monitored: • ASIC installation error sensing • FPGA configuration and status • Thermal sensing • Fan Sensing • Power failure sensing • 48-V failure • 48-V maintenance • Ambient air tempe...

Chapter 1 9 Introduction Utilities board Environmental condition detected by power-onfunction The power-on function detects environmental errors (such as ASIC NotInstalled OK or FPGA Not OK). It does not turn on power to the nodeuntil the conditions are corrected. It also detects environmental error...

10 Chapter 1 Introduction Utilities board The environmental error interrupt and the 1.2 second delay provide thesystem adequate time to read CSRs to determine the cause of the error,log the condition in NVRAM, and display the condition on the attentionlightbar. After the system is powered down, the ...

Chapter 1 11 Introduction Utilities board Teststation interface The teststation can be a PA-RISC based workstation. The interface tothe teststation is an ethernet AUI port for flexibility in connecting tomany workstations. It is also easily expandable. DC test of a node To perform the DC test, the T...

12 Chapter 1 Introduction System displays System displays The V2500 server provides two means of displaying status and errorreporting: an LCD and an Attention light bar. Figure 3 System displays Front panel LCD The front panel is a 20-character by 4-line liquid crystal display asshown in Figure 4. 9...

Chapter 1 13 Introduction System displays Figure 4 Front panel LCD When the node key switch is turned on, the LCD powers up but isinitially blank. Power-On Self Test (POST) starts displaying output to the LCD. Thefollowing illustrates this output shown in Figure 4: Node status line The Node Status L...

14 Chapter 1 Introduction System displays Table 3 Processor run-time status codes Message display line The message display line shows the POST initialization progress. This isupdated by the monarch processor. The system console also shows detailfor some of these steps. Table 4 shows the code definit...

Chapter 1 15 Introduction System displays Table 4 Message display line Power supply indicators When the keyswitch on the operator panel is in the DC ON position boththe AC power (amber) LED and the DC power (green) LED on each of thepower supplies should be on. Message display code Description a Uti...

16 Chapter 1 Introduction System displays Attention light bar The Attention light bar is located at the top left corner on the front of theHP 9000 V2500 server as shown in Figure 3 on page 12. This light bardisplays system status in three ways: • Off—system powered down • Steady on—system powered up...

18 Chapter 1 Introduction System displays SCUB 3.3-Volt error This error indicates that the SCUB 3.3-Volt power supply has failed, butthe 5-Volt supply has not. ASIC installation error Each ASIC in the node has ASIC Install lines to prevent power-up if anASIC is installed incorrectly (such as a SPAC...

Chapter 1 19 Introduction System displays FPGA configuration and status The SMUC is programmed by a serial data transfer from EEPROM uponutility board power-up. If the transfer does not complete properly, theSMUC cannot configure itself and many environmental conditionscannot be monitored. The power...

20 Chapter 1 Introduction System displays the SMUC, which reports the environmental warning to the processors.The power-on circuit displays the “highest priority” 48-Volt supply thatfailed. Ambient air sensors The ambient air sensors detect a too warm or too hot condition in theinput air stream to t...

Chapter 2 21 2 Configuration management The teststation allows the user to configure the node using the ts_config utility. ts_config configures the teststation to communicate with the node. The teststation daemon, ccmd , monitors the node and reports back configuration information, error information...

22 Chapter 2 Configuration management Teststation Teststation The teststation is used for configuring, monitoring, testing, and errorlogging. It is not required for normal operation of a node. The teststation communicates with the JTAG interface in the nodes. TheJTAG port remains idle if no teststat...

24 Chapter 2 Configuration management ts_config NOTE For shells that are run from the teststation desktop, the DISPLAYvariable is set (at the shell start-up) to the local teststation display. ts_config operation The ts_config utility displays an active list of nodes that are powered up and connected...

26 Chapter 2 Configuration management ts_config Configuration Procedures NOTE This procedure does not need to be performed unless the status shows“Upgrade JTAG firmware.” If the node shows “Not Configured,” skip thissection. The following procedures provide additional details about eachconfiguration...

Chapter 2 27 Configuration management ts_config Upgrade JTAG firmware Step 1. Select the node from the list in the display panel. For example, clicking on node 0 in the list highlights that line as shown in Figure 6. Figure 6 ts_config show node 0 highlighted Notice that after the node has been high...

28 Chapter 2 Configuration management ts_config Figure 7 ts_config “Upgrade JTAG firmware” selection. Step 3. A message panel appears as the one shown in Figure 8. Read the message. If this is the desired action, click “Yes” to begin the upgrade. Figure 8 Upgrade JTAG firmware confirmation panel Ste...

Chapter 2 29 Configuration management ts_config Figure 9 ts_config power-cycle panel When the node is powered up, the “Configuration Status” should changeto “Not Configured.” Configure a Node Step 1. Select the desired node from the list of available nodes. When the node is selected, the appropriate...

30 Chapter 2 Configuration management ts_config Figure 11 ts_config “Configure Node” selection. After invoking ts_config to configure the node, a node configuration panel appears as the one in Figure 12. Figure 12 ts_config node configuration panel Step 3. Enter a name for the V2500 System. The test...

Chapter 2 31 Configuration management ts_config Step 4. Select an appropriate serial connection for the V2500 console from the pop-down option menu in the node configuration panel. ts_config automatically assigns the first unused serial port. If the terminal mux has been configured, the terminal mux...

32 Chapter 2 Configuration management ts_config Figure 14 ts_config indicating Node 0 is configured Step 7. Restart the Workspace Manager: Click the right-mouse button on the desktop background to activate the root menu. Select the “Restart” or“Restart Workspace Manager” option, then “OK” to activat...

Chapter 2 33 Configuration management ts_config Figure 15 ts_config “Configure ‘scub_ip’ address” selection ts_config checks the scub_ip address stored in NVRAM in the node. If the scub_ip address is correct, no action is required. If the node is notdetected and scanned by ccmd , ts_config may ask y...

34 Chapter 2 Configuration management ts_config Figure 17 ts_config scub_ip address set confirmation panel Initiate a node reset to activate the new scub_ip address. Reset the Node Step 1. Select the desired node from the list of available nodes. Step 2. Select “Actions,” then “Reset Node.” This is ...

Chapter 2 35 Configuration management ts_config Figure 19 ts_config node reset panel Step 3. In the Node Reset panel, select the desired “Reset Level” and “Boot Options,” then click Reset.” Deconfigure a Node Deconfiguring a node removes the selected node from the teststationconfiguration. The tests...

36 Chapter 2 Configuration management ts_config Figure 20 ts_config “Add/Configure Terminal Mux” selection. A panel appears as the on shown Figure 21. This panel requires theterminal mux IP address. Figure 21 ts_config terminal mux IP address panel Step 3. Connect a serial cable from serial port 2 o...

Chapter 2 37 Configuration management ts_config Figure 22 Terminal mux IP address entered into panel Remove terminal mux ts_config does not remove the terminal mux if any node consoles are assigned to terminal mux ports. Step 1. Select “Actions,” then “Configure Terminal Mux.” Step 2. Select “Remove...

38 Chapter 2 Configuration management Teststation-to-system communications Teststation-to-system communications This section describes how the teststation communicates with the systemusing the utilities presented in Chapter 11, “Utilities.” Figure 23 depicts the V-Class server to teststation communi...

Chapter 2 39 Configuration management Teststation-to-system communications The hardware components located on the SCUB are shown in thediagram on the left side of the node or system. They include threeethernet ports and one DUART. A layer of firmware between HP-UX and OBP called spp_pdc allows the H...

40 Chapter 2 Configuration management ccmd ccmd ccmd builds a configuration information database on the teststation. The board names and revisions, the device names and revisions, and thestart-up information generated by POST are all read and stored inmemory for use by other diagnostic tools. ccmd i...

42 Chapter 2 Configuration management xconfig xconfig xconfig is the graphical tool that can also modify the parameters initialized by POST to reconfigure a node. The graphical interface allows the user to see the configuration state.Also the names are consistent with the hardware names, since indiv...

Chapter 2 45 Configuration management xconfig The main xconfig window has three sections: • Menu bar—Provides additional capability and functions. • Node configuration map—Provides the status of the node. • Node control panel—Provides the capability to select a node and control the way data flows to...

46 Chapter 2 Configuration management xconfig Node configuration map The node configuration map is a representation of the left and right sideviews of a node as shown in Figure 27. Figure 27 xconfig window node configuration map

Chapter 2 47 Configuration management xconfig The button boxes are positioned to represent the actual boards as viewedfrom the left and right sides. Each of the configurable components of thenode is in the display. The buttons are used as follows: • Green button—Indicates that the component is prese...

50 Chapter 2 Configuration management Configuration utilities Configuration utilities V2500 diagnostics provides utilities that assist the user withconfiguration management. autoreset autoreset allows the user to specify whether ccmd should automatically reset a complex after a hard error and after ...

Chapter 3 53 3 Power-On Self Test POST is the Power On Self Test firmware for the V-Class platform.POST provides processor and system hardware initializationfunctionality, as well as providing basic processor selftest and utilitiesboard SRAM pattern test capability. This chapter describes how POSTin...

54 Chapter 3 Power-On Self Test Overview Overview Upon power up, all processors and hardware must be initialized beforethe node proceeds with booting. POST begins executing and brings upthe node from an indeterminate state and then calls OBP. None of the POST modules can be directly controlled via a...

56 Chapter 3 Power-On Self Test POST modules POST modules POST executes modules listed below in chronological order: • Processor Initialization and Selftest—Each processor initializes itself on power up or reset in parallel with the other processors.Initialization includes setting values into the in...

58 Chapter 3 Power-On Self Test Interactive mode Interactive mode POST for the V2500 provides a command line interface for configurationand debugging. The command line interface is invoked if boot_module isset to “interactive,” by a soft reset, or a TOC during POST execution. Interactive mode comman...

Chapter 3 59 Power-On Self Test Interactive mode Configuration parameters The following parameters control the runtime operation of POST: • ts_ip —Specifies the teststation IP address for LAN messaging. The value should be set to the IP address of the diagnostics LAN port onthe teststation. [default...

Chapter 3 63 Power-On Self Test Messages Messages POST has three types of messages: LCD, console, and error. This sectiondiscusses each type. LCD messages Each node has an LCD display. Figure 29 shows the display andindicates what each line on the display means. Figure 29 Front panel LCD Node status...

66 Chapter 3 Power-On Self Test Messages Console messages POST provides several messages that are displayed on the teststationconsole. This section describes these console messages. Type-of-boot This message reports the type of boot for the current POST execution,and the node ID and monarch processo...

68 Chapter 3 Power-On Self Test Messages Each character indicates the physical location of the DIMM and thelogical size of the DIMM. The memory information is encoded as follows:Value Memory Type . 32MB : 64MB | 128MB _ Empty # Hardware deconfigured $ Software (user) deconfigured For example: r0 r1 ...

Chapter 3 69 Power-On Self Test Messages Booting Boombox Interactive boot This message indicates that POST is entering it's interactive mode.POST provides a console interface for system configuration and debug. For example: Booting Interactive Interactive prompt The following is the POST interactive...

70 Chapter 3 Power-On Self Test Messages the checksum and was rebuilt to the default structure. For example: Test Station Parameters checksum FAILED, rebuilding... This node may be forced with the sppdsh reboot <node> default command Configuration map failure This message indicates that the co...

Chapter 3 71 Power-On Self Test Messages Memory board deconfiguration This message indicates that the specified memory board is deconfigured.This can be due to a memory board being found on one side of memorywithout a corresponding pair, since boards must be used in pairs of even/odd boards. This ca...

72 Chapter 3 Power-On Self Test Messages PB0L_B failed to go idle after memory init Unable to force CPU PB2L_A into idle loop Monarch completing memory initialization This message indicates that the monarch processor is completing thememory initialization assigned to the specified processor. For exa...

Chapter 3 73 Power-On Self Test Messages Contiguous memory block not found This message indicates that POST could not find a block of contiguousmemory to place at address zero to achieve good memory. POST willreport no main memory to the OBP for this failure. For example: HP/UX good memory region co...

74 Chapter 3 Power-On Self Test Messages For example: cpu PB1R_A deconfigured due to PB1R_B shutdown. New monarch processor selected This message indicates that the previous monarch processor wasdeconfigured and a new one was selected. The new monarch continuesthe initialization of the rest of the s...

Chapter 4 75 4 Test Controller The Test Controller is an EEPROM-based utility that provides theenvironment for executing the offline diagnostic tests. It is controlledthrough parameters stored in the NVRAM on the Utilities board. TheTest Controller reads these parameters to determine its execution m...

76 Chapter 4 Test Controller Test Controller modes Test Controller modes There are three basic operational modes for this utility: • Stand-alone mode • Interactive mode • I/O Utility mode In stand-alone mode, cxtest invokes the Test Controller. The Test Controller reads test parameters from NVRAM (t...

Chapter 4 77 Test Controller User interface User interface The Test Controller provides for the control of offline diagnostic testexecution. It utilizes a set of parameters to control its operation. Theparameters consist of the following: • Global set that controls the overall operation of the Test ...

78 Chapter 4 Test Controller User interface • Read and write the 128 words of test specific information • Select the hardware to test • Display the current parameter selections Main menu Test Controller Main Menu MAIN Menu commands 0=Quit Test Controller 1=Begin Test Controller Execution 2=Halt Test...

86 Chapter 4 Test Controller User interface Test Configuration menu The Test Configuration menu is shown below: Test Configuration menu Test Configuration Menu 0=Return to Main Menu A=Hardware Selection Menu 1=Display ClassesB=Loop Enable 2=Display SubtestsC=Loop Count 3=Select ClassesD=Test Error C...

88 Chapter 4 Test Controller User interface • 5=Read All Test Parameters—Reads all 128 words that make up the test parameter set and displays this information. These testparameters reside in NVRAM and are defined by the particular test.An example of the display is shown in the example below: Test Co...

Chapter 4 93 Test Controller Example of running diagnostics from Test Controller command line Example of running diagnostics fromTest Controller command line This example shows how to run mem3000 from the Test Controller command line within the following scenario: • Configure mem3000 to run on a sys...

94 Chapter 4 Test Controller Example of running diagnostics from Test Controller command line Step 2. From the Test Selection menu shown below, select Memory test, option 1. Test Controller Test Selection menu MAIN Menu - Test Selection Display 0= Return to Main Menu 1= Memory test 2= not available ...

Chapter 4 95 Test Controller Example of running diagnostics from Test Controller command line Step 5. From the menu, select Memory test, option 1. This opens the Test Configuration menu shown below: Test menu 1=*Memory test 2= not available 3= not available 4= not available 5= I/O test 6= CPU Selfte...

96 Chapter 4 Test Controller Example of running diagnostics from Test Controller command line Step 7. From the Hardware Selection menu shown below, select CPUs, option 1. Selecting CPUs from Hardware Selection menu Test Configuration Menu - Hardware Selection Display 0=Return to Test Configuration M...

Chapter 4 97 Test Controller Example of running diagnostics from Test Controller command line Step 3. From the Test Configuration menu, select Display Subtests, option 2. The subtest menu shown below lists all available subtests: mem3000 Subtests menu 100* Diagnostic CSR Read/Write Test101* Other SM...

98 Chapter 4 Test Controller Example of running diagnostics from Test Controller command line Step 4. Select all appropriate subtests. Table 22 lists the test patterns for subtests 230 through 238. Table 22 Test patterns for subtests 230-238 and 330-338 Selecting Display Subtests, option 2, from the...

Chapter 4 99 Test Controller Example of running diagnostics from Test Controller command line Starting tests To run the tests selected from the Test Controller main menu, selectBegin Test Controller Execution, option 1. The output is shown in theexample below: Example of mem3000 execution % Enter co...

Chapter 5 101 5 cxtest The cxtest program is a graphical front end and a command line interpreter for the test controller. It is a standalone program that runsindependently of any diagnostic tests loaded in the EEPROM on theUtilities board.

104 Chapter 5 cxtest Graphics interface Graphics interface To start the cxtest graphics interface, specify the -d option on the command line as follows: % cxtest -d This causes cxtest to open a window on the display. Where the window is displayed is set by the environment variable $DISPLAY. This can...

Chapter 5 105 cxtest Graphics interface File menu The File menu has the following options: • Save Selections • Restore Selections • Log to File/Close Log File • Clear Log • Exit Save Selections The Save Selections option saves specific tests or configurations. Restore Selections With the Restore Sel...

106 Chapter 5 cxtest Graphics interface The selections presented are based on whether the Test Controller hasbuilt a Subtest table and Class table in its tc_test_info_struct structure. Class menus Selecting a test opens a window that displays all classes for the test. SeeFigure 31. Down the left han...

Chapter 5 107 cxtest Graphics interface The Defaults button installs test default values into all the parameters. If a class of tests has no parameters associated with it, the right mostbutton (the square one) is not shown. Global Test Parameters menu cxtest provides the ability to loop on a number ...

108 Chapter 5 cxtest Graphics interface Command menu The Command menu is used to perform actions on the node or complexbeing tested. These actions include: • Go • Reset Machine • Read Boot Config Map The Go selection starts the subtests. The subtests are sent to the testcontroller one at a time so t...

Chapter 5 109 cxtest Graphics interface Figure 33 System configuration window Help menu The Help menu has two entries: About and Contents. The Aboutselection displays the version number of cxtest running and the Contents selection opens a browser that can scroll through the help file. Display area T...

110 Chapter 5 cxtest Graphics interface Powering down the system When using cxtest in a troubleshooting environment, it is not necessary to exit and enter cxtest each time the power is cycled. To remove power to the system (for example, to move a board), power thesystem down leaving cxtest running. ...

Chapter 5 111 cxtest Command line interface Command line interface cxtest is a utility that allows the user to run tests loaded into the Test Controller. Tests can be specified on the command line or a Graphic UserInterface can be started to simplify test selection. cxtest allows use of the Test Con...

112 Chapter 5 cxtest Command line interface Command line test selections The command line interface deciphers the following switches to selecttests. • -mem —Memory diagnostic. • -io —I/O diagnostic. • -cpu —processor diagnostic. All the arguments between two test selections apply only to first tests...

Chapter 5 113 cxtest Command line interface To set the number of times a test is looped on use the -lt <number> option. Example of cxtest -lt option cxtest -mem -lt 3 -c 4 -io -c 2 The looping specification only applies to the memory test which runs theclass-4 tests three times. The I/O test r...

114 Chapter 5 cxtest Command line interface To specify a list of subtests. place a comma between the numbers. As anexample, -s 100,150,140 , runs subtest 100, then subtest 150, and finally subtest 140. Command line parameter specifications To specify the value of a parameter for a test, use the -pa#...

Chapter 5 115 cxtest Example of running diagnostics from cxtest window Example of running diagnostics from cxtest window The following example procedure shows the user how to use mem3000 from cxtest . It assumes that the node configuration has been set up using the main cxtest window. Step 1. From t...

116 Chapter 5 cxtest Example of running diagnostics from cxtest window Figure 35 mem3000 Class 1 Subtest Selections window Step 4. In the Subtest Selections window for each class, click the button for subtest to be executed. Any combination of subtests may be executed. Step 5. To set the parameters ...

Chapter 5 117 cxtest Example of running diagnostics from cxtest window Step 6. To start the selected tests and subtests, click the Go option in the Command menu in the cxtest main window. Step 7. View the results in the lower window pane of the cxtest main window.

Chapter 6 119 6 Processor-dependent codefirmware loader The processor-dependent code firmware loader ( pdcfl ) is a firmware module with the capabilities of loading other firmware modules intoFLASH. It is intended to speed up download of POST and OBP on newlymanufactured or malfunctioning utility bo...

120 Chapter 6 Processor-dependent code firmware loader pdcfl loading, booting, and setup pdcfl loading, booting, and setup NOTE This step should not be necessary under normal circumstances. pdcfl is loaded on all Utility boards at the factory. If the utility board FLASH contents have been erased, pd...

Chapter 6 121 Processor-dependent code firmware loader pdcfl loading, booting, and setup This requires making these entries to the following files: To /etc/services make the following entry: tftp 69/udp Trivial File Transfer Protocol To /etc/inetd.conf make the following entry: tftp dgram udp wait r...

122 Chapter 6 Processor-dependent code firmware loader pdcfl commands pdcfl commands From the pdcfl prompt, the following commands are supported: • printenv [variable] —Prints configuration variables from NVRAM. • setenv variable value —Allows setting configuration variables in NVRAM. • lifls —Print...

Chapter 7 125 7 cpu3000 This chapter describes cpu3000 processor test cpu3000 runs via the test controller and provides a basic test of the functionality of the PA8500. cpu3000 requires a minimum of one processor with its associated SPAC and two EWMBs. Included in thetesting are most of the instruct...

126 Chapter 7 cpu3000 cpu3000 classes and subtests cpu3000 classes and subtests cpu3000 consists of a series of tests grouped together in classes beginning with verification of the most basic functionality andprogressing toward more complex functionality. Each class has subtestswhich target specific...

Chapter 8 133 8 io3000 The I/O diagnostic supports Symbios 875 HVD SCSI controllers, Symbios895 LVD SCSI controllers, and Tachyon Fibre Channel controllers. io3000 requires a node with a minimum of one processor, one SIOB with associated SPACs, and two EWMBs with associated SMACs. Toexercise periphe...

134 Chapter 8 io3000 io3000 classes and subtests io3000 classes and subtests io3000 consists of a series of tests grouped together in classes beginning with verification of the most basic functionality and progressing towardmore complex functionality. Each class is broken down into subtestswhich tar...

Chapter 8 147 io3000 io3000 classes and subtests Table 41 io3000 Class 16 subtests User parameters The test controller provides io3000 with up to 37 user parameter words. Current parameters are defined in Table 42. Subtest Name Description 1600 Tachyon PCIconfigurationspace test Verifies the ability...

150 Chapter 8 io3000 io3000 classes and subtests Device specification Due to Core Logic SRAM space limitations, only 20 devices per SAGAcan be tested at a time. Up to 24 SCSI devices can be specified usingparameter words 8-19. Each of these parameter words contains twodevice specifications, as shown...

154 Chapter 8 io3000 io3000 error codes io3000 error codes When a failure is encountered, an event code is set along with an errormessage. The least significant 12 bits of the event code contain the errorcode. Table 47 lists the io3000 error codes. io3000 general errors io3000 general error codes po...

Chapter 8 155 io3000 io3000 error codes io3000 device specification errors io3000 device specification errors post the following error message: SAGA_name/ctlr_num/tgt_num/lun_num Example of io3000 device specification error message: IOLF_A/ct0/idf/lu0 Table 48 shows each io3000 general error code. T...

Chapter 8 157 io3000 io3000 error codes io3000 SAGA ErrorInfo CSR error The io3000 ErrorInfo CSR error code posts the following error message: SAGA_name/cause_bit/address/act_val Example of io3000 SAGA ErrorInfo CSR error: IOLF_A/5/fc210098/10e0000f0c000000 Table 51 shows the io3000 SAGA ErrorInfo C...

Chapter 8 159 io3000 io3000 error codes io3000 controller general errors io3000 Controller general error codes post the following error message: SAGA_name/ctlr_num Example of io3000 controller general error message: IOLF_B/ct0 Table 54 shows each io3000 general controller error code. Table 54 io3000...

160 Chapter 8 io3000 io3000 error codes Table 55 io3000 PCI errors io3000 controller command errors io3000 controller command error codes post the following error message: SAGA_name/ctlr_num/tgt_num/lun_num/comp_stat/ scsi_stat:sense_key:sense_code:sense_code_qualifier Example of io3000 controller c...

Chapter 8 161 io3000 io3000 error codes io3000 DMA error The io3000 DMA error code posts the following error message: SAGA_name/ctlr_num/tgt_num/lun_num/address/act_val/ exp_val Example of io3000 DMA error message: IOLF_A/ct0/idf/lu0/0004148200/a5a5a5a4/a5a5a5a5 Table 57 shows the io3000 DMA error c...

162 Chapter 8 io3000 io3000 error codes Example of io3000 Symbios controller specific error message: IOLF_B/ct1/f804000010/ffffff01/00000001 Table 59 shows each io3000 Symbios controller specific error code. Table 59 io3000 Symbios controller specific errors io3000 Tachyon controller specific errors...

Chapter 9 165 9 mem3000 This chapter describes mem3000 , a memory test for V2500 systems. mem3000 is core logic flash-based memory diagnostic that verifies the functionality of the memory subsystem. mem3000 requires a node with a minimum of one processor with two memory boards. Excalibur W Memory Bo...

166 Chapter 9 mem3000 mem3000 classes and subtests mem3000 classes and subtests mem3000 verifies the V2500 memory subsystem using the Test Controller. mem3000 requires one node with a minimum of one process with associated SPAC and two EWMBs with associated SMACs. mem3000 consists of a series of tes...

170 Chapter 9 mem3000 V2500 memory configurations V2500 memory configurations In the V2500 server, Excalibur Pluggable Memory Boards (EPMBs) areinstalled in 16 DIMM connectors on the EWMBs. A V2500 memory board is organized by quadrants, rows, and buses.Each memory board has four quadrants, four row...

Chapter 9 171 mem3000 V2500 memory configurations Table 70 DIMM row/bus table V2500 DIMM quadrant designations Memory boards can be populated in increments of four DIMMs calledquadrants. • Four DIMMS provides 1/4 population • Eight DIMMS provides 1/2 population • Twelve DIMMS provides 3/4 population...

172 Chapter 9 mem3000 V2500 memory configurations Figure 39 V2500 DIMM locations Example: Q2B3: Quadrant 2, Bank 3 V2500 DIMM configuration rules Use the following rules to plan the memory board DIMM configuration: • All memory boards must be populated identically. • Single node memory boards may be...

Chapter 9 173 mem3000 V2500 memory configurations • DIMMs in quadrant 1 can be of a different size than DIMMs in quadrant 2 or 3 without degrading performance. • DIMMS in quadrant 0 and 1 should be the same size for maximum performance. • DIMMS in quadrant 2 and 3 should be the same size for maximum...

174 Chapter 9 mem3000 User parameters User parameters The Test Controller allows mem3000 20 user parameters. Table 73 defines these parameters: Table 73 User parameter definitions Parameter 4 defaults to the value 2 causing the test to automaticallyprobe all known DIMMs to determine their type: 80- ...

Chapter 9 179 mem3000 mem3000 error codes Table 76 Patterns used in specified subtests Error messages When a failure is encountered an event code is set along with an errormessage. The least significant 12 bits of the event code contain the errorcode. The error codes and their error message descript...

180 Chapter 9 mem3000 mem3000 error codes Figure 42 Type one error message format There are six fields separated by / symbols. The meaning of each field isas follows: • Field 1—Specifies the on which the failure was detected • Field 2—Specifies the DIMM on which the failure was detected • Field 3—Sp...

Chapter 9 181 mem3000 mem3000 error codes The two fields of the type two error are as follows: • Field 1—Specifies the EWMB to which the information pertains • Field 2—Specifies the type of DIMM detected as follows: • x—Non-existent DIMM • 0—80-bit DIMM • 1—88-bit DIMM The correspondence of these va...

Chapter 10 183 10 Scan test The Exemplar scan test ( est ) is a diagnostic utility that uses the system scan hardware making it possible to perform connectivity tests and totest gate array internal registers. The est utility runs on the teststation and sends scan instructions to a given node by way ...

184 Chapter 10 Scan test est utility test environment est utility test environment est is started on the teststation and is located in /spp/bin/est. The user has the option of either starting up a user interface or having the est utility run a script. est works on one node at a time by sending scan ...

Chapter 10 185 Scan test est utility test environment To perform ID and ring checks in the utility system, the user should turnoff the power control feature either though the command line argument -p or through a runtime option command ( power_control ). The latter should seldom occur, because est a...

186 Chapter 10 Scan test Running the est GUI Running the est GUI The est GUI may be started at the command prompt. The following is the est command usage: /spp/bin/est [-option] node_number As an example to bring up the GUI and test node 0, enter the followingcommand: % /spp/bin/est -x 0 Table 77 on...

Chapter 10 187 Scan test Running the est GUI The lower set of buttons allows the user to quickly and easily run thescan tests in a wholesale fashion. The test can be modified to run fewerpatterns, to loop continuously or for a finite number of times, to test non-default limits, etc. Each button is e...

188 Chapter 10 Scan test Running the est GUI Files button Clicking the Files button opens pop-up menu with three selections: • Execute Scripts—Runs a file containing est commands. • Reset Log File—Clears the log file. • Exit—Closes the est main window and exits the program. Options button Clicking t...

Chapter 10 189 Scan test Running the est GUI Clocks button Clicking the Clocks button opens pop-up menu with four selections: • Upper—Sets the upper limit of the system clocks. • Nominal—Sets the system clocks to their nominal values. • External—Selects an external clock from the ECUB. • Status—Disp...

190 Chapter 10 Scan test Running the est GUI • Command Menu—Opens the command line window which allows the user to enter est commands directly from the GUI system. • Scan Debug Menu—Opens the debug window. • Connectivity Test Menu—Opens the connectivity test window. • Gate Array Test Menu—Opens the ...

192 Chapter 10 Scan test Running the est GUI Gate array test window The gate array test window provides a means to test all gate arrays inthe Exemplar system. The window is simple to use. Figure 49 shows the est gate array test window. Figure 49 est gate array test window In the top panel, enter the...

194 Chapter 10 Scan test Running the est GUI Scan window The scan window provides means of testing the system scan rings.Figure 50 shows the est scan window. NOTE For more information on scan rings and modes, see the IEEE 1149.1JTAG specification. Figure 50 est scan window The window has three panel...

196 Chapter 10 Scan test Running the est GUI SCI cable test window The SCI cable test window provides a means to test the cables thatconnect the scalable coherent interfaces between nodes. All cables aretested by default, but an individual cable can be tested using thiswindow. Figure 51 shows the es...

Chapter 10 197 Scan test Running the est GUI Help Clicking the Help button opens pop-up menu with five topic selections: • Overview • Commands • GUI • Input Files • Options Clicking on one of these options opens the Help window shown in Figure52. This window is initially blank. To open the topic of ...

200 Chapter 10 Scan test Running est from command line Running est from command line The following is the command line usage for est : est [-options] <node_number> For example, to test node 0, enter: % est 0 est reads configuration information from files stored in /spp/data (e.g node_0.cfg). T...

Chapter 10 201 Scan test Running est from command line Some examples of est usage are: est -v est -l -f my_script 0 est -o ./my_log_file 0 The est utility uses certain data and vector files located in the /spp/estdirectory. Unless disabled or redirected, the est utility will generate a log file, est...

Chapter 10 203 Scan test Running est from command line Example output when using the est -h option: % est -h Excalibur Scan Test 1.0.0.2 1998/11/25 10:32:58 Steven Terry usage: est [-options] [server] node [-cp port] [-sp port] options: -h ... print this help message -v ... print the version of the ...

204 Chapter 10 Scan test Running est from command line Table 78 AC Connectivity test options Bypass test The Bypass test format is: b The Bypass test places the scan ring hardware into bypass mode. DC Connectivity test DC Connectivity test format is: d [-s -p #] Table 79 shows the options for the th...

Chapter 10 207 Scan test Running est from command line SCI test The sci utility tests the Coherent Toroidal Interface (CTI) cables between nodes. The term SCI (Scalable Coherent Interface) is often usedin place of the term CTI; the terms are interchangeable. The usage of sci is as follows: sci [driv...

208 Chapter 10 Scan test Running est from command line SCI_all test The sci_all utility tests all SCI cables in a complex. The usage of sci_all is as follows: sci_all [test] where: test Refers to the specific test: dc, dc_clk, ac. With the dctest, the clock from the receiver node is used. Thedc_clk ...

Chapter 10 209 Scan test Running est from command line • -c high —Displays the upper clock limit. • -p 1 nom —Sets the supply 1 margin to nominal. There are four power supplies, 1 through 4. Table 81 shows the valid values for clock and power. Table 81 Valid values for clock and power supplies est m...

Chapter 10 211 Scan test Running est from command line est command flags and options There are a number of flags or options that operate on and enhance the est commands. Some of these flags and options perform the same functions as the run time option commands. To set these options, enter F at the e...

Chapter 11 213 11 Utilities This chapter details most of the diagnostic utilities which include: • address_decode • arrm • consolebar • dcm • dfdutil • dump_rdrs • fwcp • fw_init • get_node_info • hard_logger • lcd • load_eprom • pim_dumper • set_complex • soft_decode • sppconsole • tc_init • tc_iou...

214 Chapter 11 Utilities address decode address decode address_decode decodes 40-bit virtual address into the physical node, smac, row, bus, and bank. It has the following format: address_decode <40-bit address in hex> In order to determine the current memory configuration, address_decode invo...

Chapter 11 215 Utilities AutoRaid recovery map (arrm) AutoRaid recovery map (arrm) The arrm utility is used only with an AR-12H (C5447A) disk array that displays the status "No address table" on the front panel rather than theusual status of "Ready." It is only intended for use by tr...

216 Chapter 11 Utilities AutoRaid recovery map (arrm) 0/1/0.5.0 If the EPIC number is outside of the range 0 to 7, the slot number isoutside of the range 0 to 2, or the target number is outside of the range 0to 15, an error message is displayed and the operator prompted toreenter the address. The pr...

Chapter 11 217 Utilities AutoRaid recovery map (arrm) Example of unsucessful recovery message Utility Compatibility Check Unsuccessful.The Product firmware may not support RECOVER!Do you want to attempt recover anyway ([y]/n)? In all cases of this type, respond with a y , Y , n , or N followed by EN...

Chapter 11 219 Utilities consolebar consolebar The consolebar utility is an X application that provides a simple interface capable of starting console windows to all V2500 nodesconfigured on the teststation. It has the following format: consolebar [-display displayname] consolebar retrieves the list...

220 Chapter 11 Utilities dcm dcm dcm dumps the boot configuration map information for the specified node. There are two main reporting modes; one for general hardwareconfiguration and one for the DIMM type. The general hardware mode reports processors, ASICs, and memory sizeinformation. The DIMM typ...

224 Chapter 11 Utilities dfdutil dfdutil dfdutil is a standalone offline utility that downloads firmware to SCSI devices including disks, arrays, and fibrechannel devices such as SCSIMUX and fibrechannel arrays. The firmware image(s) are contained in a Logical Interchange Format(LIF) volume on the t...

Chapter 11 229 Utilities dfdutil DOWNLOAD command Use the DOWNLOAD command to download firmware to a particular device. DOWNLOAD transfers the contents of a particular firmware file to a device. It prompts the user for any arguments that were not specified onthe command line. NOTE Once the download ...

230 Chapter 11 Utilities dfdutil DISPMAP <disk index> The user may enter the index number of a single device; using no indexnumber causes DISPMAP to list all devices. This command will display the bootable device table displayed when dfdutil is started. If the optional argument [index] is spec...

Chapter 11 231 Utilities dfdutil DISPFILES command The DISPFILES command displays a list of all available firmware filesfound on a LIF device. The command displays: • File name • Intended product identification • New revision number • Size of firmware (not file size) The syntax for this command is: ...

232 Chapter 11 Utilities dfdutil Entering HELP without a command name displays a list of all available dfdutil commands. Entering the specific command name after HELP outputs specific information about the command. Notes and cautions about dfdutil This section presents some limitations and cautions ...

Chapter 11 233 Utilities dfdutil Shared SCSI Buses If dfdutil is running on a system which shares any of its SCSI busses with another system or systems, the other system or systems must behalted while this program is running. This program can not determinethat a bus is shared, so the operator must d...

Chapter 11 235 Utilities fwcp fwcp fwcp is an OBP command that upgrades system firmware. A single firmware package may be loaded by the following command: % fwcp <filename> To load all system firmware packages, use the following masterdownload script: source /core@f0,f0000000/lan@0,d30000;15.9...

236 Chapter 11 Utilities fw_init fw_init fw_init provides an automatic means for downloading firmware to each node and initializing certain data structures in NVRAM. Using thisscript prevent problems that could occur when executing this proceduremanually. The format if fw_init is as follows: fw_init...

Chapter 11 239 Utilities get_node_info [OPTIONS] include the following: • -a —Display all fields (default) • -A —Display all configured nodes The selected fields will be printed in the order below) • -c —Display the Complex name • -n —Display the Node id • -m —Display the Diagnostic IP hostname • -o...

242 Chapter 11 Utilities lcd lcd lcd prints the current contents of the liquid crystal display for node 0 of the current complex. It has the following format: lcd The complex can be changed by using the set_complex utility. The output is sent to stdout output. Example output of lcd I-I- ---- I-P- --...

244 Chapter 11 Utilities load_eprom Table 83 load_eprom options As an example, entering the following reads the file /spp/firmware/post.fw and updates the POST section of Flash EEPROM on the Utilitiesboard. xns3_d% load_eprom -n hw2a-0000 -p /spp/firmware/post.fw Entering the following reads the fil...

248 Chapter 11 Utilities set_complex set_complex The set_complex sets the default V2500 Complex Name in the current shell environment. set_complex [COMPLEX_NAME] Once set, teststation diagnostic or console utilities that are run fromwithin the shell operate on the specified complex. If multiple comp...

Chapter 11 249 Utilities set_complex set_complex can be invoked anytime the user wants to change the shell default complex. If the user enters an invalid COMPLEX_NAME , the default complex becomes unset and the prompt string indicates thiscondition. If the user does not enter a COMPLEX_NAME , the co...

254 Chapter 11 Utilities sppconsole The following message appears in the console window: [0:1] ok [read-only -- use `^Ecf’ to attach, `^Ec?’ for help] Attach to the node by entering Ctrl ecf . Press the Ctrl key e simultaneously; do not press the Ctrl key with the c and f . All information and error...

Chapter 11 255 Utilities tc_init tc_init tc_init determines the node ID, ethernet address, and IP address for all nodes in the complex. This information is then stored in the NVRAMof all nodes as one 12-byte entry per node. Each 12-byte entry has theformat shown in Figure 54: Figure 54 tc_init NVRAM...

Chapter 11 261 Utilities Version utilities Version utilities This section describes the three version utilities. diag_version The diag_version utility displays the product name and the version of the current teststation software. For example: $ diag_version HP9000/V2500 Diagnostics, Version 1.0.0.0 ...

Chapter 11 263 Utilities Event processing Event processing This section discusses three event processing utilities: • event_logger • log_event event_logger The event_logger utility is the teststation Event Logger and has a format as follows: event_logger [-d] event_logger receives messages from diag...

266 Chapter 11 Utilities Miscellaneous tools Miscellaneous tools The following miscellaneous tools are described in this section: • kill_by_name • fix_boot_sector kill_by_name The kill_by_name script kills processes by name rather than by process identification. The following is the usage of this sc...

Chapter 12 267 12 Scan tools This chapter details most of the scan tools which include: • sppdsh • do_reset • jf-node_info • jf-ccmd_info • jf-reserve_info

Chapter 12 269 Scan tools sppdsh Definitions The following definitions will help user with the operation of sppdsh : • node id—An identification (ID) that can be either the node IP name or a node number. To distinguish between one node number andanother, the environmental variable, COMPLEX_NAME, ind...

Chapter 12 271 Scan tools sppdsh a. System memory can be modified through partial deconfigura- tion. • buf[1..4]—A buffer is a 4K byte block of memory on the test station that is used as a temporary holding area. 128MB 88-bit SW deconfigured to 80 0x6c 128MB 88-bit SW deconfigured 0x7c 64MB deconfig...

274 Chapter 12 Scan tools sppdsh • memory size—An argument used to deconfigure larger amounts of memory across all memory boards on a node. • net cache size—Refers to the memory shared between nodes in each node’s network cache. The network cache should be the same acrossall nodes in a complex. Misc...

Chapter 12 275 Scan tools sppdsh • power <node id> supply[1..4] [low|nom|up] —Changes the power margin on the supply indicated across all nodes in contact withthe test station. • power <node id> supply[1..4] [low|nom|up —Changes the power margin on the supply indicated across all nodes i...

278 Chapter 12 Scan tools sppdsh • ecc_cpy <address> <data> [size] —Copies the data into the ECC associated with the cache line of address and repeats for size cache lines. Data conversion commands Data conversion commands manipulate, evaluate or interpret datawithin the diagnostic shell...

Chapter 12 279 Scan tools sppdsh l_sub <arg1> <arg2> —Left subtract two data arguments. For example: abc=`l_sub 0x55 0x1` l_mod <arg1> <arg2> —Left modulo two data arguments. For example: abc=`l_mod 0x55 0x1` l_mult <arg1> <arg2> —Left multiply two data arguments....

280 Chapter 12 Scan tools sppdsh node <node _number> — set default node to be node _number in the current complex. fi_node —Find all available nodes in the current complex. fi_cpu [-v] [-q] <node_number> —Find all available processors of node_number in the current complex. fi_emb [-v] [-...

Chapter 12 281 Scan tools sppdsh I/O buffering commands This section presents a list of the sppdsh I/O buffering commands. For these commands, four default buffers are created: buf1 - buf4. buf_cmp buf1 buf2 —Compares two buffers. Null is returned if they are the same. If they are different, the ind...

282 Chapter 12 Scan tools sppdsh mem_cmp addr1 addr2 size —Compares the memory at addr1 to (addr1 + size ) to that at addr2 . mem_cmp addr1 buf1 size —Compares the memory at addr1 to ( addr1+size ) to that at buf1 . mem_dump addr [size] —Dumps the memory starting at addr . mem_cpy addr1 buf1 [size] ...

Appendix A 289 A List of diagnostics This appendix provides a list of all utilities and diagnostics in this bookand where they are located. Table 89 List of diagnostics Name Locations address_decode Page 214 arrm Page 215 autoreset Page 50 ccmd Page 40 consolebar Page 219 cpu3000 Chapter 7, page 125...

290 Appendix A List of diagnostics hard_logger Page 240 io3000 Chapter 8, page 133 io_tr Page 164 jf-ccmd_info Page 286 jf-node_info Page 285 jf-reserve_info Page 287 kill_by_name Page 266 lcd Page 242 load_eprom Page 243 log_event Page 264 mem3000 Chapter 9, page 165 pdcfl Chapter 6, page 119 pim_d...

Index 293 Index A AC Connectivity test , 203 AC test of a node , 11 address IP , 40 address decode , 213 , 214 , 216 , 217 , 218 arrm , 213 , 215 Attention lightbar , 4 , 7 B Boot Configuration map , 110 bootable device table , 226 buses memory, discussed , 170 Bypass test , 204 C ccmd , 21 , 22 , 4...