code review: checking {key,modulo,prime} length in KcF

[Darren J Moffat]

Vladimir Kotal wrote:
> Darren J Moffat wrote:
> 
> <snip>
> 
>> I feel quite strongly that not doing the checks in metaslot is 
>> actually wrong and isn't fixing the root cause of the problem.  Having 
>> the check in metaslot is necessary to avoid the same problem across 
>> multiple providers plugged into userland (eg, softtoken vs smartcard 
>> vs tpm vs kernel).
> 
> No one is disputing that (again, see CR 6266218) so I am not sure what 
> is the problem. Maybe I was not specific enough in the first mail which 
> was talking about addressing the "first part" ?

I missed that.  I think the userland part in metaslot is more important 
to address though if we actually want to fix the root cause.

--
Darren J Moffat