On Mon, Apr 19, 2010 at 11:44 PM, Garrett Cooper <[email protected]> wrote:
> On Mon, Apr 19, 2010 at 11:13 PM, Mitani <[email protected]> wrote:
>> Hi,
>>
>> I ran "su01" test.
>> This test failed:
>> ------------<log>
>> userdel: user su_usr1 does not exist
>> Creating mailbox file: File exists
>> userdel: user su_usr2 does not exist
>> Creating mailbox file: File exists
>> YOU NEED TO SET ENVIROMENT VARIABLE PASSWD.
>> ------------
>>
>> This failure was caused by "PASSWD" environment variable not being set:
>> ------------<su01_s1>
>> #Grab input from enviroment
>> if [info exists env(PASSWD)] {
>> set PASSWD $env(PASSWD)
>> } else {
>> send_user "YOU NEED TO SET ENVIROMENT VARIABLE PASSWD. \n"
>> exit 1
>> }
>> ------------
>>
>> According to "su01_s1", this "PASSWD" is the password of "root" user:
>> ------------<su01_s1>
>> spawn /bin/su -c whoami
>> set i_am_root 0
>> expect {
>> "Password:" {
>> send "$PASSWD\r"
>> expect {
>> "root" { set i_am_root 1
>> set i_can_root 1
>> }
>> }
>> }
>> }
>> ------------
>>
>> I think whether this test program is unfinished.
>> Probably the author of this program thought that there was a problem
>> to set a password in the shell, I think.
>> Because, the password of root is various by each users who run the test,
>> and it cannot be gotten from environment variable or file etc.
>>
>> So, I suggest that "su01" shell demands root password when it is run.
>> But, when user forgot to input, it should return and finish with error.
>>
>>
>> I attach the patch for "su01" file.
>
> I don't agree; this item is already mentioned in INSTALL (but I
> think that it should probably be moved to a README file or
> something... INSTALL doesn't seem intuitive) for another set of tests:
>
> iv) Finally, before running the networktests.sh script, two variables must
> be set within the script file. The "RHOST" variable should be set to
> the hostname of the server. The "PASSWD" variable should be set to
> root's password on the server machine. This is necessary for tests
> such as telnet01 and ftp01.
>
> All of the applications which require $PASSWD need to be
> documented though as this is far from intuitive.
Also, what you're suggesting would fix _interactive_ tests, not
_non-interactive_ tests, where the majority of all users should *not*
be running LTP interactively -- thus I don't think that it's a
worthwhile addition (sorry :(...).
Thanks,
-Garrett
------------------------------------------------------------------------------
Download Intel® Parallel Studio Eval
Try the new software tools for yourself. Speed compiling, find bugs
proactively, and fine-tune applications for parallel performance.
See why Intel Parallel Studio got high marks during beta.
http://p.sf.net/sfu/intel-sw-dev
_______________________________________________
Ltp-list mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/ltp-list
