[C-safe-secure-studygroup] Minutes from today's telecon

[Aaron Ballman]

On Wed, Nov 14, the C Safety and Security Study Group met via
teleconference. Attendees: Aaron Ballman, Robert Seacord, Roberto
Bagnara, Charles Wilson, Laurence Urhegyi, and Clive Pygott.

We reviewed action items from last time: Charles Wilson has agreed to
step up as the permanent chair of the committee and we approved it
internally. Welcome Charles! We need to put a notification in the
report to WG14 for April, alerting them to the change in status. Clive
talked to Andrew Banks and David Ward about IP concerns between MISRA
and ISO, and Andrew agreed to bring it up during the MISRA
administrative meetings last week; Clive has not heard back yet, but
will chase the answer down. Michael Wong did email the lists with
information about the various other related standardization efforts
he's involved with. So all action items were taken care of.

We then moved on to Rule 13.1 Initializer lists shall not contain
persistent side effects. Concerns were raised over the
implementability of the rule because it may be hard for a tool to test
for persistent side effects for extremely large initializer lists.
Robert brings up that we often handle this as QoI within the document.
The order of initializers is well-defined and so "clever" code can use
this securely, but shouldn't use this in safety, which comes back to
the security vs safety discussion. Also, the MISRA rules don't mention
designated initializers (it's based on C99) and there may be some
other interaction there for whatever words we come up with. Relates to
EXP30-C from CERT rules. There was unanimous consent within the group
that this is a good rule in general for safety and security and we
should include something about it, but we're not certain of the
specific words just yet.

We then wrapped up for the day.

~Aaron