tqchen commented on PR #20062: URL: https://github.com/apache/tvm/pull/20062#issuecomment-5106712181
Updated the two-commit series to `072c9dec3f` with the CI repair pass. The follow-up preserves BufferVar identity through `buffer_data` projections in match-buffer lowering, use/def analysis, host/device splitting, storage access/rewrite, and `tvm_access_ptr` lowering. It also teaches CUDA formatted-print lowering to decode the projection, keys packed-ABI alignment metadata by the logical buffer variable expected by LLVM codegen, and accepts pointer-valued Triton expressions. Validation on the updated head: - full C++ build completed - all repository pre-commit hooks passed - focused buffer-projection regressions: 86 passed, 10 environment skips - TIRx analysis/transform: 343 passed, with 3 LLVM-only tests unavailable in the LLVM-disabled build - runnable S-TIR transform suite: 247 passed, 8 environment skips - CUDA print/verifier smoke set: 10 passed, 1 no-CUDA skip Fresh CI is now queued on the new head. -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: [email protected] For queries about this service, please contact Infrastructure at: [email protected] --------------------------------------------------------------------- To unsubscribe, e-mail: [email protected] For additional commands, e-mail: [email protected]
