Re: [PATCH v4 0/1] MR8116: kerberos: Add support for SECBUFFER_STREAM to SpUnsealMessage().

[Hans Leidekker (＠hans)]

On Mon May 26 15:10:27 2025 +0000, Dmitry Timoshkov wrote:
> Then we have to make a copy of the input buffer because "The STREAM
> buffer will be modified in-place to decrypt its contents."

The language is somewhat ambiguous. It doesn't make sense to me that the data is both decrypted in-place and copied to the data buffer, but we can ignore that by giving it a copy of the full stream buffer and free it afterwards.