Darwin: bsp mapping#420
Open
aalamsi22 wants to merge 4 commits into
Open
Conversation
| return std::make_unique<Wedge40Manager>(platformMappingStr); | ||
| } | ||
|
|
||
| std::unique_ptr<WedgeManager> createDarwinWedgeManager( |
Contributor
There was a problem hiding this comment.
We need to have a boolean gflag (use_bsp_access), disabled by default so that we don't change our current prod behavior (accessing transceivers through userspace i2c drivers). You can enable that flag in your config and test this. We'll enable it slowly in prod once we are ready to roll out BSP access for Darwin
Contributor
Author
There was a problem hiding this comment.
Just to clarify. This flag doesn't exist and need to be defined and default to false. Which platforms need to set this flag to true and where is that done?
aalamsi22
force-pushed
the
darwin_bsp_mapping
branch
from
ce1eab3 to
8ed37c6
Compare
Contributor
|
@aalamsi22 is this PR still needed? there are a lot of conflicts there. It will need to get rebase and solved the conflicts. |
Contributor
Author
Hi Joan, this PR is needed to enable the agent on the platform_manager implementation. Its unclear to me how to proceed forward considering that this may conflict with the legacy implementation. |
Contributor
I will discuss this with Harshit and Tao. There are QSFP codes that this PR changes which I am not familiar with. |
raghav-nexthop
pushed a commit
to nexthop-ai/fboss
that referenced
this pull request
Feb 6, 2026
…acebook#420) # Summary This change enables running multiple CLI commands within a single process, which is required for C++ end-to-end tests. The first two end-to-end tests that were originally written in Python, along with helper functions, have been rewritten in C++, and can now run as a self-contained gtest binary, without depending on the fboss2-dev binary. Some fixes were necessary to make this work, so the CLI could be re-used as a library instead of being invoked as a subprocess: 1. Remove the `hasRun` static variable from `CmdHandler` - Previously, a static `hasRun` flag prevented commands from executing more than once per process - Now we wrap callbacks in `CmdSubcommands` to check `get_subcommands().empty()` to detect leaf commands, eliminating the need for static state 2. Throw exceptions instead of calling `exit(1)` in `CmdHandler` - `CmdHandler::run()` now rethrows exceptions instead of calling `exit(1)` - This allows tests to catch errors and verify exit codes - `Main.cpp` catches exceptions and returns exit code 1 3. Factor out `CLI::App` initialization into `CliAppInit.h` - Header-only `initApp()` function shared between `Main.cpp` and tests - Eliminates code duplication for CLI setup - This couldn't be packaged into a library of its own without going into circular dependency hell, hence why it's a header-only helper 4. Reset state in `ConfigSession::initializeSession()` - Clear `ConfigSession` attributes when creating a new session (when session file doesn't exist) - Allows re-using the same `ConfigSession` singleton after `commit()` 5. Update state in `ConfigSession::commit()` - Set `base_` to the new commit SHA after successful commit - Set `configLoaded_ = false` to force config reload on next access - Allows re-using the same `ConfigSession` singleton after `commit()` 6. Automatically initialize session in `ConfigSession::loadConfig()` - If session file doesn't exist (e.g., after commit deleted it), call `initializeSession()` to create a fresh session These changes allow the ConfigSession singleton to be properly reused across multiple CLI command invocations within the same process. # Test Plan Added new end-to-end tests in C++ that replace `test_config_interface_mtu.py` and `test_config_interface_description.py` (the Python scripts will be removed in a future PR soon). Sample output (simplified, without timestamps and other noise): ``` [==========] Running 2 tests from 2 test suites. [----------] 1 test from ConfigInterfaceDescriptionTest [ RUN ] ConfigInterfaceDescriptionTest.SetAndVerifyDescription [Step 1] Finding an interface to test... Using interface: eth1/1/1 (VLAN: 2001) [Step 2] Getting current description... Current description: 'CLI_E2E_TEST_DESCRIPTION_ALT' [Step 3] Setting description to 'CLI_E2E_TEST_DESCRIPTION'... Description set to 'CLI_E2E_TEST_DESCRIPTION' [Step 4] Verifying description via 'show interface'... Verified: Description is 'CLI_E2E_TEST_DESCRIPTION' [Step 5] Restoring original description ('CLI_E2E_TEST_DESCRIPTION_ALT')... Restored description to 'CLI_E2E_TEST_DESCRIPTION_ALT' TEST PASSED [ OK ] ConfigInterfaceDescriptionTest.SetAndVerifyDescription (412 ms) [----------] 1 test from ConfigInterfaceMtuTest [ RUN ] ConfigInterfaceMtuTest.SetAndVerifyMtu [Step 1] Finding an interface to test... Using interface: eth1/1/1 (VLAN: 2001) [Step 2] Getting current MTU... Current MTU: 9000 [Step 3] Setting MTU to 1500... MTU set to 1500 [Step 4] Verifying MTU via 'show interface'... Verified: MTU is 1500 [Step 5] Verifying kernel interface MTU... Verified: Kernel interface fboss2001 has MTU 1500 [Step 6] Restoring original MTU (9000)... Restored MTU to 9000 TEST PASSED [ OK ] ConfigInterfaceMtuTest.SetAndVerifyMtu (384 ms) [==========] 2 tests from 2 test suites ran. (797 ms total) [ PASSED ] 2 tests. ``` ## Running CLI End-to-End Tests ### 1. Build the test binary ```bash ./fboss/oss/scripts/nhfboss-build.sh --cmake-target cli_test ``` ### 2. Copy files to fboss101 ```bash # Copy the test binary scp /var/FBOSS/tmp_bld_dir/build/fboss/cli_test fboss101:/opt/fboss/bin/cli_test scp fboss/oss/scripts/run_scripts/run_test.py fboss101:/opt/fboss/bin/run_test.py ``` ### 3. Run the tests on fboss101 ```bash ssh fboss101 "cd /opt/fboss && ./bin/run_test.py cli" ``` Or run the binary directly: ```bash ssh fboss101 /opt/fboss/bin/cli_test ``` ### Notes - The tests run against the already-running FBOSS `wedge_agent` on fboss101 - Tests communicate with the agent via Thrift (localhost) - No config file changes or agent restart required
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
4 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Description
fboss/led_service/oss/DarwinLedManager.cppstill exists, it wasn't removed as its unknown what the side effects are on non-oss systems.Depends on:
Testing
sai agent hwtests