Update error codes to be more specific.

[dvalinrh]

Description

Uses error_codes found in test_tools/error_codes to provide more specific error indication.

Before/After Comparison

Before: Returned a 0 or 1, making determination of having to rerun very course
After: Error codes are now bases on test_tools/error_codes, making it easier to limit when we do a rerun of the test.
Also, changed how we load in test_tools, we try curl, git, and wget, to avoid a failure due to the program not
being installed.

Clerical Stuff

This closes #72

Relates to JIRA: RPOPC-867

Testing

Ran through zathras, with following scenario file
global:
ssh_key_file: replace_your_ssh_key
terminate_cloud: 1
os_vendor: rhel
results_prefix: documentation
os_vendor: rhel
system_type: aws
cloud_os_id: ami-035032ea878eca201
systems:
system1:
tests: "coremark"
host_config: "m5.xlarge"

csv file produced.
iteration,threads,IterationsPerSec,Start_Date,End_Date
1,4,58708.414873,2026-03-13T12:11:27Z,2026-03-13T12:12:18Z
1,4,59802.651251,2026-03-13T12:11:27Z,2026-03-13T12:12:18Z
2,4,58927.519151,2026-03-13T12:12:33Z,2026-03-13T12:13:23Z
2,4,59769.885939,2026-03-13T12:12:33Z,2026-03-13T12:13:23Z
3,4,59017.360940,2026-03-13T12:13:38Z,2026-03-13T12:14:28Z
3,4,59488.399762,2026-03-13T12:13:38Z,2026-03-13T12:14:28Z
4,4,58783.188008,2026-03-13T12:14:43Z,2026-03-13T12:15:34Z
4,4,59627.329193,2026-03-13T12:14:43Z,2026-03-13T12:15:34Z
5,4,58924.625583,2026-03-13T12:15:48Z,2026-03-13T12:16:39Z
5,4,60003.000150,2026-03-13T12:15:48Z,2026-03-13T12:16:39Z

Returned 0 as expected. Testing on rerun verification handled in the appropriate zathras pr.

[github-actions]

This relates to RPOPC-867

[qodo-code-review]

Review Summary by Qodo

Replace hardcoded exit codes with specific error codes

✨ Enhancement 🐞 Bug fix

Walkthroughs

Description

• Replace hardcoded exit codes with specific error codes from test_tools
• Add multiple fallback methods to install test_tools (wget, curl, git)
• Use proper error code constants throughout script for better error handling
• Update usage exit to use E_USAGE and final exit to use E_SUCCESS

Diagram

flowchart LR
  A["Script execution"] --> B["Install test_tools"]
  B --> C["Try wget method"]
  C --> D["Try curl method"]
  D --> E["Try git clone method"]
  E --> F["Run coremark"]
  F --> G["Generate results"]
  G --> H["Exit with error code"]
  B -.->|fallback| C
  C -.->|fallback| D
  D -.->|fallback| E

Loading

File Changes

1. coremark/coremark_run ✨ Enhancement +44/-14

Add error codes and multiple tool installation methods

• Added three new functions (attempt_tools_wget, attetmpt_tools_curl, attempt_tools_git) to
 install test_tools using multiple methods as fallbacks
• Replaced hardcoded exit code 1 with specific error code constants ($E_GENERAL, $E_USAGE,
 $E_SUCCESS) throughout the script
• Modified install_test_tools function to call all three installation methods sequentially instead
 of only attempting git clone
• Updated usage function to exit with $E_USAGE instead of 0
• Changed final script exit from 0 to $E_SUCCESS

coremark/coremark_run

---

[qodo-code-review]

Code Review by Qodo

🐞 Bugs (2) 📘 Rule violations (0) 📎 Requirement gaps (1)

1. Unvalidated exit code vars 🐞 Bug ⛯ Reliability

Description

coremark_run now exits using $E_SUCCESS/$E_GENERAL/$E_USAGE but never defines or validates them
locally; if sourcing $TOOLS_BIN/general_setup fails, exit_out/exit can return the previous command’s
status and misreport success or the wrong error code. This can break callers relying on exit codes
for rerun/triage decisions.

Code

coremark/coremark_run[499]

+exit $E_SUCCESS

Evidence

The script’s exit behavior is entirely determined by the numeric argument passed to exit/exit_out.
After this PR, that argument is often an E_* variable (not defined anywhere in this repo), and the
script does not check whether sourcing general_setup succeeded before continuing; if the variables
are unset, bash will effectively execute exit with no numeric argument and use the previous
command’s status instead.

coremark/coremark_run[32-36]
coremark/coremark_run[379-383]
coremark/coremark_run[78-86]
coremark/coremark_run[194-201]
coremark/coremark_run[497-499]

Agent prompt

The issue below was found during a code review. Follow the provided context and guidance below and implement a solution

## Issue description
`coremark_run` exits using `$E_SUCCESS/$E_GENERAL/$E_USAGE`, but those variables are not defined in this repo and the script doesn’t validate that sourcing `general_setup` succeeded. If sourcing fails or doesn’t set these variables, `exit`/`exit_out` can return an unintended status (often the previous command’s status), misreporting success or incorrect failure codes.
### Issue Context
- `exit_out` calls `exit $2` directly.
- `source "$TOOLS_BIN/general_setup" "$@"` is not checked for success.
- Multiple error/success exits now depend on `$E_*` values.
### Fix Focus Areas
- coremark/coremark_run[32-36]
- coremark/coremark_run[78-86]
- coremark/coremark_run[379-383]
- coremark/coremark_run[194-201]
- coremark/coremark_run[497-499]

ⓘ Copy this prompt and use it to remediate the issue with your preferred AI generation tools

---

2. Bootstrap ignores unzip/mv errors 🐞 Bug ⛯ Reliability

Description

The new wget/curl bootstrap path checks only the download exit code and ignores unzip/mv failures
while still deleting main.zip, which can leave $TOOLS_BIN missing or incomplete. This can cascade
into later failures (e.g., sourcing general_setup) with confusing symptoms and wrong exit statuses.

Code

coremark/coremark_run[R88-110]

+attempt_tools_wget()
+{
+        if [[ ! -d "$TOOLS_BIN" ]]; then
+                wget ${tools_git}/archive/refs/heads/main.zip
+                if [[ $? -eq 0 ]]; then
+                        unzip -q main.zip
+                        mv test_tools-wrappers-main ${TOOLS_BIN}
+                        rm main.zip
+                fi
+        fi
+}
+
+attetmpt_tools_curl()
+{
+        if [[ ! -d "$TOOLS_BIN" ]]; then
+                curl -L -O ${tools_git}/archive/refs/heads/main.zip
+                if [[ $? -eq 0 ]]; then
+                        unzip -q main.zip
+                        mv test_tools-wrappers-main ${TOOLS_BIN}
+                        rm main.zip
+                fi
+        fi
+}

Evidence

Both wget and curl paths perform unzip/mv/rm without checking any of those commands’ exit statuses.
The script then proceeds to source $TOOLS_BIN/general_setup later without verifying it exists, so
a partial/failed unzip/mv can cause hard-to-diagnose downstream failures.

coremark/coremark_run[88-98]
coremark/coremark_run[100-110]
coremark/coremark_run[152-155]
coremark/coremark_run[379-381]

Agent prompt

The issue below was found during a code review. Follow the provided context and guidance below and implement a solution

## Issue description
The wget/curl bootstrap code treats download success as install success, but it does not check whether `unzip` succeeded or whether the extracted directory was moved correctly. It also deletes `main.zip` regardless of unzip/mv outcome. This can leave `$TOOLS_BIN` missing/incomplete and cause later `source "$TOOLS_BIN/general_setup"` to fail in confusing ways.
### Issue Context
The script tries wget, then curl, then git. If wget/curl download succeeds but unzip/mv fails, the install is still incomplete.
### Fix Focus Areas
- coremark/coremark_run[88-110]
- coremark/coremark_run[152-155]
- coremark/coremark_run[379-381]

ⓘ Copy this prompt and use it to remediate the issue with your preferred AI generation tools

---

3. Hardcoded 101 exit code 📎 Requirement gap ✓ Correctness

Description

The new attempt_tools_git() error path uses a literal exit code (101) instead of a shared named
error-code constant, making error signaling inconsistent with the new error-code approach. This can
undermine the goal of standardizing on specific, centrally-defined error codes.

Code

coremark/coremark_run[117]

+                        exit_out "Error: pulling git $tools_git failed." 101

Evidence

PR Compliance ID 1 requires updating generic error codes to more specific ones and using them
consistently; the added code introduces a hardcoded numeric code rather than a named code variable
used elsewhere in the PR (e.g., $E_GENERAL, $E_USAGE, $E_SUCCESS).

Error codes updated to be more specific
coremark/coremark_run[114-118]

Agent prompt

The issue below was found during a code review. Follow the provided context and guidance below and implement a solution

## Issue description
`attempt_tools_git()` exits via `exit_out ... 101` using a hardcoded numeric exit code, which is inconsistent with the PR's shift to named error-code variables.
## Issue Context
Other updated error paths use named variables like `E_USAGE`, `E_GENERAL`, and `E_SUCCESS`, implying exit codes should be centrally defined and consistently referenced.
## Fix Focus Areas
- coremark/coremark_run[112-120]

ⓘ Copy this prompt and use it to remediate the issue with your preferred AI generation tools

---

ⓘ The new review experience is currently in Beta. Learn more

[dvalinrh]

The hardcoded 101 has to happen as at this our attempt to load test_tools has failed 3 different ways.

[kdvalin]

LGTM