CoreCLR RISC-V architecture port
## Current status Our initial goal was to get to 100% pass rate for CodeGen BringUp tests in Debug mode. We've achieved this both on qemu and on StarFive VisionFive2 board. For this, all changes are currenlty in PRs created by my colleagues @clamp03 @t-mustafin: - already merged (https://github.com/dotnet/runtime/pull/82379, https://github.com/dotnet/runtime/pull/82382, https://github.com/dotnet/runtime/pull/84584, https://github.com/dotnet/runtime/pull/84608, https://github.com/dotnet/runtime/pull/82380) - not yet merged at the moment of writing (https://github.com/dotnet/runtime/pull/84748, https://github.com/dotnet/runtime/pull/82381, https://github.com/dotnet/runtime/pull/84797). So, when all of the above are merged, CodeGen BringUp tests launched on main branch should pass both on VisionFive2 and qemu (*for StarFive VisionFive2 board `GC_REGIONS` also need to be disabled currently, we are already investigating this*). Current test results for full JIT directory on VisionFive2 in Debug with `DOTNET_GCgen0size=99999999`: ``` # Tests Discovered : 4737 # Passed : 4304 # Failed : 413 # Skipped : 20 ``` Current test results for all tests on VisionFive2 in Debug with `DOTNET_GCgen0size=99999999`: ``` # Tests Discovered : 6565 # Passed : 5663 # Failed : 861 # Skipped : 41 ``` ## Further plans #### Our further plans for near future: - [x] focus on full JIT tests directory to make tests pass in Release/Debug/Checked
Posted by gbalykov on Apr 14, 2023, 12:34 PM
Tags
Related issues
- [Meta] Feedback on granular control over Copilot changes
- unexpected status 401 Unauthorized issue
- [MODEL] Measurable quality regression in Claude generation 5 (Fable 5 / Opus 5 / Sonnet 5): worse nonsense detection, ~2x verbosity, under-disclosed model fallb
- [CRITICAL] Subagent spawning and subagent pattern bugs trigger infinite recursion, infinite token usage, grossly inefficient token usage, and lost accumulated s
- Video stops playing after some minutes
- [MODEL] Claude verbose code comments by default — ignores instructions to stop.
Raise real-world problems, find people facing the same issue, and discuss solutions.2,500 issues0 comments
github (Software)
[Meta] Feedback on granular control over Copilot changes
unexpected status 401 Unauthorized issue
[MODEL] Measurable quality regression in Claude generation 5 (Fable 5 / Opus 5 / Sonnet 5): worse nonsense detection, ~2x verbosity, under-disclosed model fallb
[CRITICAL] Subagent spawning and subagent pattern bugs trigger infinite recursion, infinite token usage, grossly inefficient token usage, and lost accumulated s
Video stops playing after some minutes
[MODEL] Claude verbose code comments by default — ignores instructions to stop.
Scope Codex VS Code chats to the current workspace/project
Update on "Co-authored-by: Copilot" in commit messages
CoreCLR RISC-V architecture port
## Current status Our initial goal was to get to 100% pass rate for CodeGen BringUp tests in Debug mode. We've achieved this both on qemu and on StarFive VisionFive2 board. For this, all changes are currenlty in PRs created by my colleagues @clamp03 @t-mustafin: - already merged (https://github.com/dotnet/runtime/pull/82379, https://github.com/dotnet/runtime/pull/82382, https://github.com/dotnet/runtime/pull/84584, https://github.com/dotnet/runtime/pull/84608, https://github.com/dotnet/runtime/pull/82380) - not yet merged at the moment of writing (https://github.com/dotnet/runtime/pull/84748, https://github.com/dotnet/runtime/pull/82381, https://github.com/dotnet/runtime/pull/84797). So, when all of the above are merged, CodeGen BringUp tests launched on main branch should pass both on VisionFive2 and qemu (*for StarFive VisionFive2 board `GC_REGIONS` also need to be disabled currently, we are already investigating this*). Current test results for full JIT directory on VisionFive2 in Debug with `DOTNET_GCgen0size=99999999`: ``` # Tests Discovered : 4737 # Passed : 4304 # Failed : 413 # Skipped : 20 ``` Current test results for all tests on VisionFive2 in Debug with `DOTNET_GCgen0size=99999999`: ``` # Tests Discovered : 6565 # Passed : 5663 # Failed : 861 # Skipped : 41 ``` ## Further plans #### Our further plans for near future: - [x] focus on full JIT tests directory to make tests pass in Release/Debug/Checked
Comments (0)
No comments yet. Start the discussion.