This stage checks that the Innovus session itself is trustworthy before any design work starts: the release and build, the UI mode, the log and message history, non-default settings, compute resources, and the provenance of any checkpoint being restored.
Sanity check cheat sheet
One row per command card. Read left to right: the check, the command that answers it, then what a healthy result, a result to review and a hard stop look like.
| Card | Check and when | Command | Healthy | Review | Hard stop |
|---|---|---|---|---|---|
| S-01 | Release and build Session start |
getVersiongetVersion -major |
Exact match with the methodology release and build. | Same major version, different build. Allowed only if the methodology says builds are interchangeable. | Different major release from the qualified one, with no approval on record. |
| S-02 | Active user interface Session start |
innovusis_common_ui_mode |
The value matches the script set: 0 for Legacy scripts, 1 for Common UI scripts. | None; this check is binary. | Mismatch between the return value and the script set. |
| S-03 | Log location and message history After init_design or restoreDesign |
getLogFileName -fullPathreport_messagereport_message -errors -count |
Zero errors. Every warning ID is known for this flow or explained. Every suppressed ID is in the methodology baseline. | New warning IDs that did not appear in the last good run. | Any error during design load, or any warning the methodology classifies as fatal. |
| S-04 | Non-default settings After the flow settings are applied |
report_command_mode -non_defaultreport_command_mode -user |
The list matches the methodology baseline exactly. | Extra options with a plausible reason, such as a debug setting left from an earlier run. | A setting that changes analysis behaviour and is not in the baseline, or a baseline setting that is missing. |
| S-05 | Compute resources After setup |
getMultiCpuUsage -localCpugetLicenseStatusreport_resource -peak |
Values within the job request. | Free disk space close to the expected size of reports and saved databases. | No free space in the run directory, which makes saved databases and reports incomplete. |
| S-06 | Checkpoint provenance When continuing from a saved database |
restoreDesign <name>.enc.dat <top>saveDesign <name>.encset restore_db_stop_at_design_in_memory 1 |
One restore, no file-check error, design name and stage match the handoff record. | Restore succeeds but the handoff record does not state which stage the save came from. | File-check error on restore, or a second restore in the same session. |
Command cheat sheet
Legacy UI commands. Angle brackets are placeholders, and values shown are examples, not project limits.
| Release and interface | |
|---|---|
getVersion | The full release and build string. Record it in the run manifest. |
getVersion -major | The first number of the version plus one digit after the decimal point. Handy for a coarse script check, not for qualification. |
innovus -stylus | Shell startup. Starts the session in the Stylus Common UI; without -stylus the session is Legacy UI. |
is_common_ui_mode | Returns 1 in a Common UI session and 0 in a Legacy UI session. |
eval_common_ui "<cui command>" | Runs a Common UI command from a Legacy session. A transition aid only; commands that behave differently in Legacy do not work properly through it. |
| Help and documentation | |
help <command> | The syntax of a command or global, or the summary text of a message ID. |
help -k <keyword> | A list of every command associated with the keyword. |
man <command> | The full man page of a command, global or message ID. |
check_syntax <script.tcl> | A lint report for Innovus Tcl scripts: misspelled commands, variables used before they are set and similar errors. |
report_hidden_usage <files> | A summary of hidden globals and mode options used in the given scripts. Use it before moving scripts to a new release. |
| Log and messages | |
getLogFileName | The name of the current log file. |
getLogFileName -fullPath | The absolute path of the log file, for the run manifest. |
getCmdLogFileName | The name of the current command log file, which records every command the session ran. |
report_message | A summary table of every message ID issued so far, with severity, count and summary, ending in a total of warnings and errors. |
report_message -errors | A Tcl list of the error IDs issued in this session. |
report_message -errors -count | A Tcl list of {ID count} pairs for every error ID. An empty list means no errors. |
report_message -warnings -count | A Tcl list of {ID count} pairs for every warning ID. |
report_message -all | A Tcl list of every issued message ID, excluding suppressed ones. |
report_message -suppressed | A Tcl list of the message IDs suppressed in this session. Suppressed IDs hide evidence, so compare this list with the baseline. |
set_message -id <ids> -no_limit | Removes any existing display limit on the given message IDs (the default is 20 per public command). Suppressed IDs stay suppressed. Changes the session; record it. |
| Settings | |
report_command_mode -non_default | Every *Mode option whose value differs from the default. |
report_command_mode -user | Every *Mode option that has been set manually. |
getDesignMode | The current setDesignMode settings. |
getAnalysisMode | The current setAnalysisMode settings, with name, value, type and whether the user set them. |
getImportMode | The current setImportMode settings, which control how the netlist and design files are read. |
| Compute resources | |
getMultiCpuUsage -localCpu | The number of local threads requested. |
getMultiCpuUsage -verbose | A message with the current multi-CPU setting. |
setMultiCpuUsage -localCpu <n> | Sets the number of local threads. Changes the session; record the value. |
checkMultiCpuUsage | Whether distributed processing works in this environment and every configured CPU is reachable. Run after setDistributeHost and setMultiCpuUsage. |
getDistributeHost -mode | The distributed-processing mode set with setDistributeHost. |
getLicenseStatus | A summary of every licence the session has checked out. Record it in the run manifest. |
report_resource -peak | The peak memory used so far. |
report_disk | Disk metrics for the current working directory and /tmp, plus used and free space for the -dir directory (default /tmp). Also returns a Tcl list. |
report_disk -dir <run_dir> | Used and free space for the run directory. |
| Save, restore and provenance | |
saveDesign <name>.enc | A restore file <name>.enc and the database directory <name>.enc.dat. |
saveDesign <name>.enc -tgz | The database as a single gzipped tar file, with library references copied inside. |
restoreDesign <name>.enc.dat <top> | The saved design in memory. Timing or RC data may need recomputing, depending on what was saved. |
restoreDesign -noTiming <name>.enc.dat <top> | The saved design without timing libraries and MMMC data, for physical-only work. |
set restore_db_file_check true | Default. A restore error if any file in the saved directory was edited, renamed or deleted. |
set restore_db_stop_at_design_in_memory 1 | Default. An error if restoreDesign is called a second time in the same session. |
get_metric <pattern> | The current state of the named metrics. |