Re: [gmime-devel] [PATCH v2 2/2] Support extraction of session keys during decryption.

[Daniel Kahn Gillmor <dkg fifthhorseman net>]

On Fri 2016-12-02 19:37:22 -0500, Daniel Kahn Gillmor wrote:
> On Fri 2016-12-02 19:21:25 -0500, Jeffrey Stedfast wrote:
> > Committed with the memset fixes.
>
> Thanks!
>
> i see one test you dropped during commit, just wanted to make sure this
> was intentionally left out (- is the committed copy, + is my suggestion):
>
> --- a/gmime/gmime-gpg-context.c
> +++ b/gmime/gmime-gpg-context.c
> @@ -1337,7 +1339,7 @@ gpg_ctx_parse_status (struct _GpgCtx *gpg, GError **err)
>                               /* nothing to do... we'll grab the MDC used in DECRYPTION_INFO */
>                       } else if (!strncmp (status, "BADMDC", 6)) {
>                               /* nothing to do, this will only be sent after DECRYPTION_FAILED */
> -                     } else if (!strncmp (status, "SESSION_KEY", 11)) {
> +                     } else if (gpg->ctx->retrieve_session_key && !strncmp (status, "SESSION_KEY", 11)) {
>                               status = next_token (status, &gpg->session_key);
>                       } else {
>                               gpg_ctx_parse_signer_info (gpg, status);
>
>
>
> I don't much care one way or the other, but i'd be happy to understand
> your reasoning to try to make future contributions fit better.

On second thought, i do care, and i prefer your method, because it gives
us a chance to scrub the RAM in this stream too.  I'll send another
patch shortly.

      --dkg