Anthony Scarpino wrote:
> Garrett D'Amore wrote:
>> Anthony Scarpino wrote:
>>> Got a few bugs up for review.. should be pretty straight forward and
>>> repetitious..
>>>
>>> 6849769 des _init: crypto_register_provider() failed
>>> 6885135 assertion failed: kcf_dh != NULL
>>> 6907099 module load/unload errors could be a little less techie and
>>> ominous
>>>
>>> http://cr.opensolaris.org/~izick/bugs/
>>>
>>> thanks
>>>
>>> Tony
>>> _______________________________________________
>>> crypto-discuss mailing list
>>> crypto-discuss at opensolaris.org
>>> http://mail.opensolaris.org/mailman/listinfo/crypto-discuss
>>
>> You're using a private variable, sys_shutdown, for the sole purpose
>> of detecting that the system is shutting down to suppress message.
>> This seems, to me at least, to be rather strange. Is there any real
>> harm in emitting the message... does it show up in real life?
>>
>> - Garrett
>>
>
> Bugs have been filed because modules try to load as the system is
> shutting down. With des and other modules, they can be used without
> registration with the framework, so the error message generates
> needless noise..
So then, why bother to message the condition at all? It seems like
maybe this ought to be done only for debug kernels.
The framework can message whatever the error condition it sees. Then at
least you could also confine the check to framework code instead of
scattering it across crypto modules.
>
> If there is a function or another variable or function that tells me
> that the system is shutting down, I'd be happy to use that instead.
I'm still skeptical that you even need to do this. What is special
about shut down that prevents module registration from occurring
properly? If its something that can be detected, then it ought to be
handled in the framework, and *not* in the individual modules.
- Garrett
>
> Tony
