[cgl_discussion] [validation framework] Using CGL_VALIDATION_ ROOT in validation suites

Fleischer, Julie N julie.n.fleischer at intel.com
Thu Sep 26 08:35:57 PDT 2002


This is correct if the tester is running and building the suites by hand,
but in the full CGL validation suite solution, there will be a master
Makefile and a master test suite execution script.  If the master Makefile
is synchronized with the master test suite execution script, then there
should be no problem.

In addition, the master test suite execution script should reset
CGL_VALIDATION_ROOT after a reboot, so the tester should not need to do
this.

Thanks.
- Julie

------------------------------------------------------------------------
Julie Fleischer
Intel Corporation
julie.n.fleischer at intel.com

This email message solely contains my own personal views and not necessarily
those of my employer.


-----Original Message-----
From: Zhu, Yi 
Sent: Wednesday, September 25, 2002 6:51 PM
To: Ling, Xiaofeng; Fleischer, Julie N; 'cgl_discussion at osdl.org'
Subject: RE: [cgl_discussion] [validation framework] Using
CGL_VALIDATION_ ROOT in validation suites


Yeah, for test suites using TVS, I can use TVS_SUITE_ROOT in stead of
CGL_VALIDATION_ROOT in my scripts, because the TVS is the master. But how
about the suites don't use TVS? Maybe I have to set it myself :(

We talked the CGL_VALIDATION_ROOT in two environments: Build time and Run
time. But they have to be the same! For example, if you set
CGL_VALIDATION_ROOT = /usr/lib/cglvalidation in build time, but set
CGL_VALIDATION_ROOT = /some/other/place in run time, it definitely won't
work. In this case, the tester must know the CGL_VALIDATION_ROOT in build
time and export CGL_VALIDATION_ROOT to that value before (s)he runs the
wrap.sh. Is that right? But think of the case that the test case needs to
reboot, the environ CGL_VALIDATION_ROOT is volatile after the system
reboots, but unfortunately, it is a full automate test and the tester
doesn't have a chance to export CGL_VALIDATION_ROOT to the correct value
again.

I do can do something to resolve the problem in my test case. But to be an
universal solution, I prefer a nonvolatile way (i.e. write the path in file
/etc/cglvalidationpath) than export it as a shell environment.

Thanks,
-Yi

-----Original Message-----
From: Ling, Xiaofeng [mailto:xiaofeng.ling at intel.com]
Sent: Thursday, September 26, 2002 2:53 AM
To: Fleischer, Julie N; 'cgl_discussion at osdl.org'
Subject: RE: [cgl_discussion] [validation framework] Using
CGL_VALIDATION_ ROOT in validation su ites


For ZhuYi 's problem, the TVS_SUITE_ROOT environment variable can be used
after reboot. The new TVS will save this variable before restart and restore
it after restart.
In wrap.sh, the TVS_SUITE_ROOT must be defined correctly, usually, it is
defined as $CGL_VALIDATION_ROOT/category/requirment/

-----Original Message-----
From: Fleischer, Julie N [mailto:julie.n.fleischer at intel.com]
Sent: 2002年9月25日 AM 11:28
To: 'cgl_discussion at osdl.org'
Subject: [cgl_discussion] [validation framework] Using CGL_VALIDATION_ROOT
in validation su ites


Moving this discussion to the mailing list....

The way that the full CGL suite will eventually be run is that all suites
are built by a master Makefile and, in addition, all suites are run by a
master script.  In this case, CGL_VALIDATION_ROOT should be defined by those
master files.

I like Bingwei's idea, though, of locally defining the variable if it is run
or built alone.  That's a good idea that we can add to the BKMs.

I think Bingwei's solution should solve Zhu Yi's problem if you are running
your suite manually after the reboot.  Once we get the master script, this
should also be taken care of automatically.

Thanks.
- Julie


------------------------------------------------------------------------
Julie Fleischer
Intel Corporation
julie.n.fleischer at intel.com

This email message solely contains my own personal views and not necessarily
those of my employer.

For me, there are two components need to reboot during the test, it's
difficult to define the enviorn either in wrap.sh or Makefile. Any good
ideas?

Thanks,
-Yi
-----Original Message-----
From: Liu, Bing Wei 

Hi, Julie

Most test suites depend on the definition of CGL_VALIDATION_ROOT, no matter
which directory it points to.  Modifying the top-level Makefile may not be
able to support the execution of test suite, i.e. running wrap.sh.  The
methods we are using now is to check first if CGL_VALIDATION_ROOT is
defined. If not, then set it to the default value.  So, it won't break the
already-defined value.  Does it make sense?

if [ -z $CGL_VALIDATION_ROOT ]; then
    export CGL_VALIDATION_ROOT=/usr/lib/cglvalidation
fi

Regards
Bingwei Liu              
Software Engineer,  Intel Corporation
bing.wei.liu at intel.com

This email message solely contains my own personal views, and not
necessarily those of my employer



_______________________________________________
cgl_discussion mailing list
cgl_discussion at lists.osdl.org
http://lists.osdl.org/mailman/listinfo/cgl_discussion
_______________________________________________
cgl_discussion mailing list
cgl_discussion at lists.osdl.org
http://lists.osdl.org/mailman/listinfo/cgl_discussion



More information about the cgl_discussion mailing list