repost of loginuid patches

[Steve Grubb]

Hi,

In  audit-loginuid-proc.patch:
General comment - weren't we going to use -1 to signify its unset? I see 
everything going through %u without special handling.

The function proc_loginuid_read:
What if the count is too small? Should we return an error? Are there any 
worries about not changing count to be TMPBUFLEN if its too big? Do we need 
to call security_getprocattr ?

The function proc_loginuid_write:
Do we need to call security_setprocattr?

In audit-netlink-loginuid.patch:
At chunk 263 & 272: auid ?  was that supposed to be uid?


That's all I see...
-Steve Grubb