>
>> We would file CR 6387333 (the RFE to add alternate root support to all
>> of the {user,role}{add,mod,del} family of commands, and perhaps
>> passmgmt as well), and either wait for someone to implement it, or go
>> off and implement it ourselves.
>
>Would you be so kind as to inquire as to when this might be fixed?
>
>> Don't succumb to the temptation to use "chroot" to solve the problem.
>> It doesn't work during Live Upgrade, and it's a hack may fail without
>> notice or workaround -- exactly as it did with CR 6375652.
>
>I won't. Although, if one were to `chroot`, and the binary executed
>required a newer revision of l ibC, wouldn't the newer libC also
>automatically be "pulled in"?Only if that binary exists in the chroot environment; the problem, however, is running libc.so from the chroot'ed environment on the kernel from the standard environment. There is no guarantee that that will work; and there have been cases where that did not. >Which other relief measure(s) could I implement in the meanwhile, >other than looking into revising `useradd`, `userdel` and `passwd`, >a nontrivial problem. It's fairly simple to write the functionality of 'useradd/userdel/passwd' as perl or awk scripts. I'm sure there are examples of that in the Solaris postinstall scripts. Casper
