[LTP] [PATCH v2 2/3] shell: Introduce TST_TIMEOUT variable

[Li Wang]

Hi Petr,

On Wed, Sep 18, 2019 at 5:50 PM Petr Vorel <pvorel@suse.cz> wrote:

> Hi Li,
>
> > On Wed, Sep 18, 2019 at 4:24 PM Petr Vorel <pvorel@suse.cz> wrote:
>
> > > ...
>
> > > > So, maybe that could be as a reason to disable float support for
> > > > $LTP_TIMEOUT_MUL?
>
> > > Maybe I'm missing something, therefore explaining the intent of the
> code.
> > > (int was handled in if tst_is_int "$LTP_TIMEOUT_MUL"; then):
>
> > I'm not blaming the int/float judgment, there is no problem. My concern
> is:
>
> > If the $LTP_TIMEOUT_MUL is float and awk command is missing, how things
> > will be going?
>
> > It will break at:
> >     tst_test_cmds awk
> > right?
> Yes. The concern is described in the commit message (but it should also be
> in
> wiki page I guess): you want float => you need awk. You don't have float,
> just
> set it as int (which will affect also C).
> Does it make sense? Is it useful?
>

It's working, but not elegant. Not sure how many people will go through the
documents for such tiny issues.

Is there any possibility to handle float by default, and if no awk
supporting, just round up the float to integer and use it do testing
automatically?


>
> > Given that break on float number handling, why not declare only support
> > integer for $LTP_TIMEOUT_MUL?
> Sure, we can do it. But I propose to do it for C as well otherwise setup
> valid
> for C will be breaking shell. But that's a backward incompatibility change.
> That's why I suggested IMHO the least intrusive change (but maybe I'm
> wrong).
> Anyone else having strong opinion?
>

No more concerns.


>
> > I hope I explained clearly this time, haha ;-)
> Sure, sorry to be slow :).
>

Never mind, I need to enhance my expression too.


>
> Kind regards,
> Petr
>

-- 
Regards,
Li Wang
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.linux.it/pipermail/ltp/attachments/20190918/cbc2e4d6/attachment.htm>