Dan Anderson wrote: > ... > kcf.conf entries are optional, but tolerated. They are always required when > a provider has disabled mechanisms. > In other words, the OpenSolaris with these proposed chanes will function the > same whether or not it has an old kcf.conf file with all the software > providers listed or with an empty kcf.conf file. >
This is what I was expecting. It is good to call this out though. >> Can you describe step-by-step how a non-ON or third-party hardware >> provider (like SCA6000) >> is supposed to work in the new world? This will be useful documentation >> to provide >> to provider writers. >> > > *Background* > The only entries previously required for hardware providers (whether or not > ON/non-on or third-party) were the (pseudo) comment "#driver_names=XXXXXX > entries. These comment entries are no longer required and are ignored. The > driver_name comments were only used by cryptoadm as a guide for where to > insert driver-specific provider configuration lines. > > *New Requirements* > The driver_names comment entries are no longer required, but are tolerated. > They are ignored if present. > So, the advice we give to the provider writers need not change. -Krishna -------------- next part -------------- An HTML attachment was scrubbed... URL: <http://mail.opensolaris.org/pipermail/crypto-discuss/attachments/20081016/19d14890/attachment.html>
