Enumerated Type in assertion ?

[Marek Ponca]

Hello everybody,

last time I have done some simulations with self-test features,
there was a problem of reporting a value of one signal.
It was an array of ennumerated types:


Something like this:
---------
type enum_t is (S1, S2, S3, S4, S5, S6, S7);
type enum_a is array (1 to 5) of enum_t;

constant enum_c : enum_a := (S1, S2, S3, S4, S5);

signal enum_s : natural range of 1 to 5;
....
enum_s <= 3;

assert(FALSE)
report "Value is:"&to_string(enum_c(enum_s))
severity note;
---------

Function to_string was from std_developers_kit.


But this does not work (Modelsim). The enummerated vatues have to be
somehow
converted to string...


Does anybody have a better approach ?

Thanks,
Marek Ponca

 

[Allan Herriman]

On Mon, 10 Jan 2005 11:00:41 +0100, Marek Ponca <name@office.com>
wrote:

Function to_string was from std_developers_kit.


But this does not work (Modelsim). The enummerated vatues have to be
somehow
converted to string...

Does this FAQ entry help?
http://www.vhdl.org/comp.lang.vhdl/FAQ1.html#scalar_string

Regards,
Allan

 

[Jonathan Bromley]

On Mon, 10 Jan 2005 11:00:41 +0100,
Marek Ponca <name@office.com> wrote:

Something like this:
---------
type enum_t is (S1, S2, S3, S4, S5, S6, S7);
type enum_a is array (1 to 5) of enum_t;

constant enum_c : enum_a := (S1, S2, S3, S4, S5);

signal enum_s : natural range of 1 to 5;
...
enum_s <= 3;

assert(FALSE)
report "Value is:"&to_string(enum_c(enum_s))
severity note;
---------

The enummerated vatues have to be
somehow converted to string...

report "Value is: " & enum_t'image(enum_c(enum(s)))
severity note;

In VHDL-93 there is a built-in "to_string()" function
for *any* scalar type T, called T'image(). There
is a much less useful function T'value() which does
the reverse - it converts a string to its data
value. Unfortunately, there's no way to protect
T'value() against errors. If the input string is
in any way wrong, i.e. it doesn't exactly match
one of the possible results of T'image, then
it bombs the simulation with an error. Consequently,
T'value() is too fragile for any practical use IMHO.
Hey, that's one for VHDL-200x that I hadn't thought of -
T'value() with an error trap.

By the way... are you *sure* you want a signal of
subtype "natural range 1 to 5", rather than a signal
of subtype "enum_t range S1 to S5" ?
--
Jonathan Bromley, Consultant

DOULOS - Developing Design Know-how
VHDL, Verilog, SystemC, Perl, Tcl/Tk, Verification, Project Services

Doulos Ltd. Church Hatch, 22 Market Place, Ringwood, BH24 1AW, UK
Tel: +44 (0)1425 471223 mail:jonathan.bromley@doulos.com
Fax: +44 (0)1425 471573 Web: http://www.doulos.com

The contents of this message may contain personal views which
are not the views of Doulos Ltd., unless specifically stated.