On 12/02/18 16:42, Salz, Rich wrote:
> The error return is not documented.

I don't see why that matters. It is trivial to see that an error is
returned in some circumstances, and there are plenty of examples in our
own code where we check this. If we removed this check we would have to
go through all of our uses to check that we've not introduced a bug
(e.g. where we know that we may send a NULL md and rely on the error
return to do the right thing). If we have to do that for our own code
it's not a big stretch to imagine plenty of other applications would be
similarly affected.

We really can't remove the check.

Matt


> 
> We are definitely moving to a "don't check params for NULL".
> 
> I would like to see this changed to be the obvious one-liner.  Tim will 
> disagree, but he's on the wrong side of history :) 
> 
> _______________________________________________
> openssl-project mailing list
> openssl-project@openssl.org
> https://mta.openssl.org/mailman/listinfo/openssl-project
> 
_______________________________________________
openssl-project mailing list
openssl-project@openssl.org
https://mta.openssl.org/mailman/listinfo/openssl-project

Reply via email to