[Info-vax] Intel x86-64 Processor Design Security Vulnerability?

Stephen Hoffman seaohveh at hoffmanlabs.invalid
Thu Jan 4 16:07:08 EST 2018


On 2018-01-04 20:23:24 +0000, DaveFroble said:

> Not sure Steve agrees with your concepts.  I seem to recall that in the 
> past, while he was still with DEC, more than once reading "the 
> wizard's" comment to someone who mentioned a problem.  "I wish you 
> wouldn't have done that."
> 
> Or, maybe times have changed ....

Vulnerability and disclosure norms have changed since DEC days.

Dropping zero-days is never been fun for vendors nor for end-users, and 
is not something I'd prefer happen to anybody.   Zero-days aren't 
always obvious to the submitters, too.   I've seen some folks post 
stuff that was clearly far worse than the person asking the question 
had considered or even realized, too.   Knowledge spreads fast, too.

Vendors need enough time to analyze and code and test and deploy fixes. 
 Always have.

Unfortunately, vendors shouldn't be permitted to defer security fixes 
arbitrarily.   That deferral and the associated silence too often also 
effects the end-users and the vendors' own staff and ISVs, though more 
subtly.

Some other fixes are  like Spectre, and not going to be easy nor cheap.

The coordinated disclosure for Spectre and Meltdown has left a number 
of providers — the many folks that weren't included in the discussions 
— in a scramble to figure out what's going on and get patches loaded 
and reboots scheduled.  But spreading the vulnerability knowledge too 
widely inevitably produces leaks.   Even among those folks and 
organizations that are trusted with the disclosure, there's ample 
opportunity for the trust to be broken or lost and for the misuse of 
the information.

No easy answers here.

Some light reading...
https://www.hackerone.com
https://www.iso.org/standard/45170.html
https://www.first.org/global/sigs/vulnerability-coordination/multiparty/FIRST-Multiparty-Vulnerability-Coordination-v1.0.pdf 

Etc.

Vendors and other organizations can probably expect 90 days before 
disclosure for most problems, and maybe as much as six months for 
massive messes.  But everybody needs to expect immediate disclosures — 
somebody dropped a zero-day local-privilege escalation in macOS 
IOHIDFamily just a few days ago — and with increasingly-faster 
deployment cycles necessary for patches.

All of this is undoubtedly just part of what Derrell and other folks at 
VSI are working on, too.



-- 
Pure Personal Opinion | HoffmanLabs LLC 




More information about the Info-vax mailing list