masahi edited a comment on pull request #9905: URL: https://github.com/apache/tvm/pull/9905#issuecomment-1011365577
@u99127 @areusch I looked it what `-device=arm_cpu` does. It's only used at https://github.com/apache/tvm/blob/6b2cbfe31f3ee02079c2ed320d97a759b0f6205c/python/tvm/autotvm/tophub.py#L107 where it instructs TVM to fetch a pre-tuned topi schedule configuration for ARM targets. It only affects tile sizes etc for conv2d, so it is perfectly valid to use it on an x86 host, although there is no point doing so. Anyway, running both `llvm` and `llvm -device=arm_cpu` has nothing to do with test coverage and it only increases test time for no good reason. -- 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]
