The temporary X509_NAME wasn't destroyed if the section didn't exist.
Also document the weird 0 vs -1 convention (see callers), and revise the
NULL check added in
https://boringssl-review.googlesource.com/c/boringssl/+/56705. It
doesn't make a difference, but we should only apply the NULL check after
we've looked at the name, and return -1 because, after the name is
checked, it's a known syntax error.
Also fix a couple of comments that were wrong. It's that the RDNSequence
we take from X509_NAME must have one RDN, not that there's one
RDNSequence. (This is a consequence of X509_NAME's somewhat odd
in-memory representation.)
Bug: oss-fuzz:55700
Change-Id: I5745752bfa82802d361803868f962b2b0fa4bd32
Reviewed-on: https://boringssl-review.googlesource.com/c/boringssl/+/56929
Auto-Submit: David Benjamin <davidben@google.com>
Commit-Queue: Bob Beck <bbe@google.com>
Reviewed-by: Bob Beck <bbe@google.com>
a028a5e01f