RE: [LTP] Results / Benchmarks

[Bryce H. <br...@os...>]

Hi Manoj,

Hmm, good idea.  The first thing you want to know when there's a failure
is, what does the test do?

Attached is a perl script I've thrown together to generate these reports
from the STP data.  Put it in the top level of your ltp source tree and
run it like this, giving it the test ID you're interested in:

   ./failure_report.pl 292158

It retrieves the run results through the net from the STP results and
then finds the corresponding source code file in your LTP tree for each
failure, and prints the description and details for each to stdout.

The script uses LWP::Simple, File::Find, and Text::Wrap, so you'll want
to make sure those are installed.

Anyway, let me know what you think, and if it's worth including in LTP
and/or STP.

Bryce

On Tue, 27 Apr 2004, Manoj Iyer wrote:
> Here is an idea for the format.
>
> ==================================================
> Test name: chdir_01
>
> Description: This is a test for testing .... blah
>
> Test Result: FAIL
>
> Details:
>
> chdir01     0  INFO  :  File .. in directory .
> chdir01     0  INFO  :  File . in directory ..
> chdir01     1  FAIL  :  strncmp failed in checkname
> ==========================================================
>
> The testcase descriptions are available along with each testcase in under
> each corresponding directory, so gathering this info should be easy.
> also, details can optional.
>
> --
> Manjo