zhenwei pi <pizhen...@bytedance.com> writes: > On 3/18/25 21:21, Markus Armbruster wrote: >> zhenwei pi <pizhen...@bytedance.com> writes: >> >>> Hi Markus, >>> >>> Current code style seems buggy, I think the main reason is that the Error >>> *errp is not generated at right place. keyctl_pkey_XXX fails without new >>> error, qcrypto_akcipher_XXX fails with new error, but they are in the same >>> switch-case code block. If we can separate crypto operations into two >>> functions - cryptodev_lkcf_keyctl_op and cryptodev_lkcf_qakcipher_op, and >>> the error is generate inside the functions, it may be handled easily. Then >>> applying your changes, it seem more clear. What do you think? >> >> Looks like a reasonable cleanup to me. >> >> I suggest to proceed as follows. We apply my minimal bug fix as is. >> You post your cleanup on top. Okay? >> > > OK!
Thanks! I'll include the patch in a pull request of error handling fixes. Care to give your Reviewed-by?
