Sasha Khapyorsky wrote:
On 08:17 Thu 22 May , Hal Rosenstock wrote:
On Thu, 2008-05-22 at 08:15 -0700, Timothy A. Meier wrote:
Sasha,
Trivial patch to enforce root for these perl scripts. More importantly,
doesn't silently fail if not root, and returns an error code.
Should these enforce root or be based on udev permissions for umad which
default to root ?
I would ask the same question as Hal did.
Ok, I understand. I have created another patch with just the auth_check
routine in it.
Following Hals advice, authorization is based on the umad permissions.
What is wrong with how it works now? On some system access to files could
be arranged for group members, or ibnetdiscover used as engine for many
scripts could be su/gid-ed. This will break there.
Sasha
The new patch shouldn't break code. I didn't realize/think about non-root
with the original patch. The intent is simply to provide a consistent and
non-silent fail mechanism.
Currently, you can get partial functionality from these scripts (-? for
example). So in that sense, this can change the behavior if the check is used
early in the script (as I did in the original patch). I view most of these
scripts as "all or nothing".
--
Timothy A. Meier
Computer Scientist
ICCD/High Performance Computing
925.422.3341
[EMAIL PROTECTED]
_______________________________________________
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
