On 14:43 Thu 29 May , Timothy A. Meier wrote: > I think this patch is fine, and helps solve the improper "usage" issue.
I will apply then. > (btw - should we prefer the "adapter" spelling over "adaptor"?) Originally it was added as "adaptor" with "adding -C, -P options" patch. I have nothing against changing this to "adapter". > My patch was addressing non-authorized use. Our philosophy was to not > allow > "any" sort of functionality (even help) if not authorized. Fail, and > provide > a reason/code. Doesn't 'chmod 0700 /usr/local/sbin/ib*.pl' (as root) solve this? > So rather than go through each perl script to see if the proper thing is > done > (return code is checked, error msg provided, terminate, etc.) It is bug fixing... :) > On 5-23, I submitted a patch which adds an auth_check() function to the > common > perl module. I agree, the implementation is non-ideal, but it is probably > sufficient for the vast majority of installations. > > If you think the concept of an auth_check() function is > desirable/acceptable, > then I will pursue fixing the implementation in a more universal way. Basically I think that idea of limited access is useful, but don't see why simple 'chmod' is insufficient. And if it is not I think that auth_check() should be optional (and of course not broken). Sasha _______________________________________________ general mailing list [email protected] http://lists.openfabrics.org/cgi-bin/mailman/listinfo/general To unsubscribe, please visit http://openib.org/mailman/listinfo/openib-general
