Hello @matt-arm , thank you for your reply.
I have gone through the latest 21.02 branch of Compute Library and for most of the functions, padding has been removed. So, I tried importing the memory similar to the way it is done in NEON case. The import is proper for inputs of layers, however, for output, the import is leading to a nullptr output during run of the CL Kernel. Am I missing something? @GeorgeARM @matt-arm, any suggestions are welcome. Thank you. --- [Visit Topic](https://discuss.tvm.apache.org/t/enabling-cl-backend-for-arm-compute-library/9404/5) to respond. You are receiving this because you enabled mailing list mode. To unsubscribe from these emails, [click here](https://discuss.tvm.apache.org/email/unsubscribe/921d88fae80c10e28a962963fd9ec3a91f0f18485a4642b7859454fbb43abeb4).
