RE: More missing methods?

Hi,

I'd go along with much of that, but some such as get_timescale() should be tool independent? We have a testbench which is defaulting to NS when run in regression, and FS when running interactively due to a setup issue, and we use get_timescale() to work around such inconsistencies. I would have thought that this method would be needed as sys.time is defined as being dependent upon the timescale set in the simulator itself, so you would need some way of handling the time within the language as it could get different values dependent upon the simulator setting.

Cheers,

Darren

From: Yuri Tsoglin [mailto:yurit@cadence.com]
Sent: Monday, September 07, 2015 12:58 PM
To: Hannes Froehlich; Galpin Darren (IFGB ATV MCD V); ieee1647@eda.org
Subject: RE: More missing methods?

Hi,

I agree with Hannes, and moreover - we generally discourage using global methods such as start_test(), run_test() and similar, and I think they should remain outside the standard, too.
Some of them even don't actually function in certain flows (such as mixed-language hierarchy flows), which makes them not only tool specific, but also flow specific.

(But maybe we still can go through this list more carefully, and see if there is one or two methods which we do need to add).

Thanks,
Yuri.


From: Hannes Froehlich
Sent: 07 September 2015 14:53
To: darren.galpin@infineon.com<mailto:darren.galpin@infineon.com>; ieee1647@eda.org<mailto:ieee1647@eda.org>
Cc: Yuri Tsoglin
Subject: RE: More missing methods?

Hi Darren,
Many (all?) of those methods could be classified as tool specific (get_timescale(), is_gui() etc.), so I would consider them outside of the e language.
-hannes

From: darren.galpin@infineon.com<mailto:darren.galpin@infineon.com> [mailto:darren.galpin@infineon.com]
Sent: 07 September 2015 12:51
To: ieee1647@eda.org<mailto:ieee1647@eda.org>
Cc: Yuri Tsoglin; Hannes Froehlich
Subject: More missing methods?

Hi,

While looking for information on timescales, I can across the following global methods which are not in the LRM:

Start_test()
Generate_test()
Setup_test()
Run_test()
Extract_test()
Check_test()
Finalize_test()
Get_timescale()
Is_gui()
Print_stack_trace()
Simulator_exists()
Simulator_save()
Sn_plusarg_exists()
Sn_plusarg_value()
Update_sys_time()

Looking at section 28 of our 2011 LRM, I note that the first line says "A significant part of e functionality is implemented as a set of predefined methods defined directly under the global and sys structs." It then goes on to define predefined methods of sys and any_struct, but not global. It would seem as though this entire section is missing from the LRM. Do others agree? And if so, do we address it now....?

Cheers,
Darren

--------------------------------------------------------------------
Darren Galpin                                         Tel: +44 117 9528741
Infineon Technologies                        Fax: +44 117 9528777
Infineon House                    Darren.Galpin@infineon.com<mailto:Darren.Galpin@infineon.com>
Great Western Court
Hunts Ground Road
Stoke Gifford
Bristol, BS34 8HP, England
--------------------------------------------------------------------


-- 
This message has been scanned for viruses and
dangerous content by MailScanner, and is
believed to be clean.