MasterJH5574 opened a new pull request, #14753: URL: https://github.com/apache/tvm/pull/14753
This PR lowers `memory.kill_tensor` and `memory.kill_storage` ops to a runtime function "`kill_object`" that releases the corresponding NDArray of the "tensor" and "storage" in VM at runtime. Previously these two ops are just ignored and skipped in VMBuiltinLower, and the "kills" didn't take effect. This PR supports the "real kill" so that the VM will release the NDArrays upon kill, and get better memory management for us. -- 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]
