A relatively longstanding issue is that some of the math_real functions
provided by ghdl use a nonstandard function name. I got annoyed by
having to maintain two different versions of my testbenches for ghdl and
another simulator, so I wrote a two patches to make math_real more
standards compliant.
ieee-mathreal-nocompat.patch just renames the functions. It will break
code that relies on the ghdl function names.
ieee-mathreal.patch also contains backward compatible functions that
warn once that they will disappear in the future. However, to implement
the warn once feature, these functions had to be marked impure, which
breaks existing code that relies on these functions being pure.
Comments?
I'd really like to see ghdl moving towards standards compliant function
names...
And now for something different:
$ ghdl -a t.vhd
t.vhd:13:14:warning: universal integer bound must be numeric literal or
attribute
This warning seems totally bogus, the bounds _are_ numeric literals.
What's wrong here?
Also, to compile ghdl on ppc64, I had to use the hack in
ghdl-ppc64abort.patch. Is there a better solution?
Tom
--- vhdl/libraries/ieee/math_real.vhdl.orig 2009-04-01 19:21:58.000000000 +0200
+++ vhdl/libraries/ieee/math_real.vhdl 2009-04-02 12:21:52.000000000 +0200
@@ -102,16 +102,21 @@
-- returns integer FLOOR(X + 0.5) if X > 0;
-- return integer CEIL(X - 0.5) if X < 0
- -- IAC: we are missing the function TRUNC
- -- IAC: we are missing the function MOD
- -- IAC: functions FMAX and FMIN should be renamed REALMAX and REALMIN
-
- function FMAX (X, Y : real ) return real;
- attribute foreign of fmax : function is "VHPIDIRECT fmax";
+ function TRUNC (X : real ) return real;
+ attribute foreign of trunc : function is "VHPIDIRECT trunc";
+ -- returns integer FLOOR(X) if X > 0;
+ -- return integer CEIL(X) if X < 0
+
+ function "MOD" (X, Y : real ) return real;
+ attribute foreign of "mod" : function is "VHPIDIRECT fmod";
+ -- returns the floating point modulus of X/Y
+
+ function REALMAX (X, Y : real ) return real;
+ attribute foreign of realmax : function is "VHPIDIRECT fmax";
-- returns the algebraically larger of X and Y
- function FMIN (X, Y : real ) return real;
- attribute foreign of fmin : function is "VHPIDIRECT fmin";
+ function REALMIN (X, Y : real ) return real;
+ attribute foreign of realmin : function is "VHPIDIRECT fmin";
-- returns the algebraically smaller of X and Y
procedure UNIFORM (variable Seed1,Seed2:inout integer; variable X:out real);
@@ -203,25 +208,17 @@
-- returns tan X; X in radians
-- X /= ((2k+1) * PI/2), where k is an integer
- -- IAC: function should be called ARCSIN
-
- function ASIN (X : real ) return real;
+ function ARCSIN (X : real ) return real;
-- returns -PI/2 < asin X < PI/2; | X | <= 1
- -- IAC: function should be called ARCCOS
-
- function ACOS (X : real ) return real;
+ function ARCCOS (X : real ) return real;
-- returns 0 < acos X < PI; | X | <= 1
-
- -- IAC: function should be called ARCTAN
-
- function ATAN (X : real) return real;
- attribute foreign of atan : function is "VHPIDIRECT atan";
+ function ARCTAN (X : real) return real;
+ attribute foreign of arctan : function is "VHPIDIRECT atan";
-- returns -PI/2 < atan X < PI/2
- -- IAC: function ATAN2 should not exist
- function ATAN2 (X : real; Y : real) return real;
+ function ARCTAN (X : real; Y : real) return real;
-- returns atan (X/Y); -PI < atan2(X,Y) < PI; Y /= 0.0
function SINH (X : real) return real;
@@ -236,20 +233,35 @@
attribute foreign of tanh : function is "VHPIDIRECT tanh";
-- hyperbolic tangent; -- returns (e**X - e**(-X))/(e**X + e**(-X))
- -- IAC: function should be called ARCSINH
-
- function ASINH (X : real) return real;
- attribute foreign of asinh : function is "VHPIDIRECT asinh";
+ function ARCSINH (X : real) return real;
+ attribute foreign of arcsinh : function is "VHPIDIRECT asinh";
-- returns ln( X + sqrt( X**2 + 1))
- -- IAC: function should be called ARCCOSH
-
- function ACOSH (X : real) return real;
+ function ARCCOSH (X : real) return real;
-- returns ln( X + sqrt( X**2 - 1)); X >= 1
- -- IAC: function should be called ARCTANH
-
- function ATANH (X : real) return real;
+ function ARCTANH (X : real) return real;
-- returns (ln( (1 + X)/(1 - X)))/2 ; | X | < 1
+
+ -- Compatibility; will be removed in the future!
+ impure function FMAX (X, Y : real ) return real;
+
+ impure function FMIN (X, Y : real ) return real;
+
+ impure function ASIN (X : real ) return real;
+
+ impure function ACOS (X : real ) return real;
+
+ impure function ATAN (X : real) return real;
+
+ impure function ATAN2 (X : real; Y : real) return real;
+
+ impure function ASINH (X : real) return real;
+
+ impure function ACOSH (X : real) return real;
+
+ impure function ATANH (X : real) return real;
+
+
end MATH_REAL;
--- vhdl/libraries/ieee/math_real-body.vhdl.orig 2009-04-01 19:22:04.000000000 +0200
+++ vhdl/libraries/ieee/math_real-body.vhdl 2009-04-02 12:27:15.000000000 +0200
@@ -72,16 +72,41 @@
begin
assert false severity failure;
end ROUND;
-
- function FMAX (X, Y : real ) return real is
+
+ function TRUNC (X : real ) return real is
begin
assert false severity failure;
- end FMAX;
+ end TRUNC;
- function FMIN (X, Y : real ) return real is
+ function c_mod (x : real; y : real) return real;
+ attribute foreign of c_mod : function is "VHPIDIRECT fmod";
+
+ function c_mod (x : real; y: real) return real is
begin
assert false severity failure;
- end FMIN;
+ end c_mod;
+
+ function "MOD" (X, Y : real ) return real is
+ begin
+ if y = 0.0 then
+ assert false
+ report "MOD(X, 0.0) is undefined"
+ severity ERROR;
+ return 0.0;
+ else
+ return c_mod(x,y);
+ end if;
+ end "MOD";
+
+ function REALMAX (X, Y : real ) return real is
+ begin
+ assert false severity failure;
+ end REALMAX;
+
+ function REALMIN (X, Y : real ) return real is
+ begin
+ assert false severity failure;
+ end REALMIN;
--
-- Pseudo-random number generators
@@ -297,18 +322,18 @@
assert false severity failure;
end c_asin;
- function ASIN (x : real ) return real is
+ function ARCSIN (x : real ) return real is
-- returns -PI/2 < asin X < PI/2; | X | <= 1
begin
if abs x > 1.0 then
assert false
- report "Out of range parameter passed to ASIN"
+ report "Out of range parameter passed to ARCSIN"
severity ERROR;
return x;
else
return c_asin(x);
end if;
- end ASIN;
+ end ARCSIN;
function c_acos (x : real ) return real;
attribute foreign of c_acos : function is "VHPIDIRECT acos";
@@ -318,24 +343,24 @@
assert false severity failure;
end c_acos;
- function ACOS (x : REAL) return REAL is
+ function ARCCOS (x : REAL) return REAL is
-- returns 0 < acos X < PI; | X | <= 1
begin
if abs x > 1.0 then
assert false
- report "Out of range parameter passed to ACOS"
+ report "Out of range parameter passed to ARCCOS"
severity ERROR;
return x;
else
return c_acos(x);
end if;
- end ACOS;
+ end ARCCOS;
- function ATAN (x : REAL) return REAL is
+ function ARCTAN (x : REAL) return REAL is
-- returns -PI/2 < atan X < PI/2
- begin
+ begin
assert false severity failure;
- end ATAN;
+ end ARCTAN;
function c_atan2 (x : real; y : real) return real;
attribute foreign of c_atan2 : function is "VHPIDIRECT atan2";
@@ -345,7 +370,7 @@
assert false severity failure;
end c_atan2;
- function ATAN2 (x : REAL; y : REAL) return REAL is
+ function ARCTAN (x : REAL; y : REAL) return REAL is
-- returns atan (X/Y); -PI < atan2(X,Y) < PI; Y /= 0.0
begin
if y = 0.0 and x = 0.0 then
@@ -356,7 +381,7 @@
else
return c_atan2(x,y);
end if;
- end ATAN2;
+ end ARCTAN;
function SINH (X : real) return real is
@@ -377,11 +402,11 @@
assert false severity failure;
end TANH;
- function ASINH (X : real) return real is
+ function ARCSINH (X : real) return real is
-- returns ln( X + sqrt( X**2 + 1))
begin
assert false severity failure;
- end ASINH;
+ end ARCSINH;
function c_acosh (x : real ) return real;
attribute foreign of c_acosh : function is "VHPIDIRECT acosh";
@@ -391,16 +416,16 @@
assert false severity failure;
end c_acosh;
- function ACOSH (X : real) return real is
+ function ARCCOSH (X : real) return real is
-- returns ln( X + sqrt( X**2 - 1)); X >= 1
begin
if abs x >= 1.0 then
- assert false report "Out of range parameter passed to ACOSH"
+ assert false report "Out of range parameter passed to ARCCOSH"
severity ERROR;
return x;
end if;
return c_acosh(x);
- end ACOSH;
+ end ARCCOSH;
function c_atanh (x : real ) return real;
attribute foreign of c_atanh : function is "VHPIDIRECT atanh";
@@ -410,15 +435,107 @@
assert false severity failure;
end c_atanh;
- function ATANH (X : real) return real is
+ function ARCTANH (X : real) return real is
-- returns (ln( (1 + X)/(1 - X)))/2 ; | X | < 1
begin
if abs x < 1.0 then
- assert false report "Out of range parameter passed to ATANH"
+ assert false report "Out of range parameter passed to ARCTANH"
severity ERROR;
return x;
end if;
return c_atanh(x);
- end ATANH;
+ end ARCTANH;
+
+ -- Compatibility; will be removed in the future!
+ shared variable fmax_compat_warn : boolean := false;
+ shared variable fmin_compat_warn : boolean := false;
+ shared variable asin_compat_warn : boolean := false;
+ shared variable acos_compat_warn : boolean := false;
+ shared variable atan_compat_warn : boolean := false;
+ shared variable atan2_compat_warn : boolean := false;
+ shared variable asinh_compat_warn : boolean := false;
+ shared variable acosh_compat_warn : boolean := false;
+ shared variable atanh_compat_warn : boolean := false;
+
+ impure function FMAX (X, Y : real ) return real is
+ begin
+ assert fmax_compat_warn
+ report "MATH_REAL.FMAX is not a standards compliant function and will be removed in the future; use REALMAX instead"
+ severity note;
+ fmax_compat_warn := true;
+ return REALMAX(X, Y);
+ end FMAX;
+
+ impure function FMIN (X, Y : real ) return real is
+ begin
+ assert fmin_compat_warn
+ report "MATH_REAL.FMIN is not a standards compliant function and will be removed in the future; use REALMIN instead"
+ severity note;
+ fmin_compat_warn := true;
+ return REALMIN(X, Y);
+ end FMIN;
+
+ impure function ASIN (X : real ) return real is
+ begin
+ assert asin_compat_warn
+ report "MATH_REAL.ASIN is not a standards compliant function and will be removed in the future; use ARCSIN instead"
+ severity note;
+ asin_compat_warn := true;
+ return ARCSIN(X);
+ end ASIN;
+
+ impure function ACOS (X : real ) return real is
+ begin
+ assert acos_compat_warn
+ report "MATH_REAL.ACOS is not a standards compliant function and will be removed in the future; use ARCCOS instead"
+ severity note;
+ acos_compat_warn := true;
+ return ARCCOS(X);
+ end ACOS;
+
+ impure function ATAN (X : real) return real is
+ begin
+ assert atan_compat_warn
+ report "MATH_REAL.ATAN is not a standards compliant function and will be removed in the future; use ARCTAN instead"
+ severity note;
+ atan_compat_warn := true;
+ return ARCTAN(X);
+ end ATAN;
+
+ impure function ATAN2 (X : real; Y : real) return real is
+ begin
+ assert atan2_compat_warn
+ report "MATH_REAL.ATAN2 is not a standards compliant function and will be removed in the future; use ARCTAN instead"
+ severity note;
+ atan2_compat_warn := true;
+ return ARCTAN(X, Y);
+ end ATAN2;
+
+ impure function ASINH (X : real) return real is
+ begin
+ assert asinh_compat_warn
+ report "MATH_REAL.ASINH is not a standards compliant function and will be removed in the future; use ARCSINH instead"
+ severity note;
+ asinh_compat_warn := true;
+ return ARCSINH(X);
+ end ASINH;
+
+ impure function ACOSH (X : real) return real is
+ begin
+ assert acosh_compat_warn
+ report "MATH_REAL.ACOSH is not a standards compliant function and will be removed in the future; use ARCCOSH instead"
+ severity note;
+ acosh_compat_warn := true;
+ return ACOSH(X);
+ end ACOSH;
+
+ impure function ATANH (X : real) return real is
+ begin
+ assert atanh_compat_warn
+ report "MATH_REAL.ATANH is not a standards compliant function and will be removed in the future; use ARCTANH instead"
+ severity note;
+ atanh_compat_warn := true;
+ return ARCTANH(X);
+ end ATANH;
end MATH_REAL;
--- vhdl/libraries/ieee/math_complex-body.vhdl.orig 2009-04-02 12:08:47.000000000 +0200
+++ vhdl/libraries/ieee/math_complex-body.vhdl 2009-04-02 12:09:18.000000000 +0200
@@ -119,7 +119,7 @@
function COMPLEX_TO_POLAR(Z: in complex ) return complex_polar is
-- converts complex to complex_polar
begin
- return COMPLEX_POLAR'(sqrt(z.re**2 + z.im**2),atan2(z.re,z.im));
+ return COMPLEX_POLAR'(sqrt(z.re**2 + z.im**2),ARCTAN(z.re,z.im));
end COMPLEX_TO_POLAR;
function POLAR_TO_COMPLEX(Z: in complex_polar ) return complex is
--- vhdl/libraries/ieee/math_real.vhdl.orig 2009-04-01 19:21:58.000000000 +0200
+++ vhdl/libraries/ieee/math_real.vhdl 2009-04-01 19:35:06.000000000 +0200
@@ -102,16 +102,21 @@
-- returns integer FLOOR(X + 0.5) if X > 0;
-- return integer CEIL(X - 0.5) if X < 0
- -- IAC: we are missing the function TRUNC
- -- IAC: we are missing the function MOD
- -- IAC: functions FMAX and FMIN should be renamed REALMAX and REALMIN
-
- function FMAX (X, Y : real ) return real;
- attribute foreign of fmax : function is "VHPIDIRECT fmax";
+ function TRUNC (X : real ) return real;
+ attribute foreign of trunc : function is "VHPIDIRECT trunc";
+ -- returns integer FLOOR(X) if X > 0;
+ -- return integer CEIL(X) if X < 0
+
+ function "MOD" (X, Y : real ) return real;
+ attribute foreign of "mod" : function is "VHPIDIRECT fmod";
+ -- returns the floating point modulus of X/Y
+
+ function REALMAX (X, Y : real ) return real;
+ attribute foreign of realmax : function is "VHPIDIRECT fmax";
-- returns the algebraically larger of X and Y
- function FMIN (X, Y : real ) return real;
- attribute foreign of fmin : function is "VHPIDIRECT fmin";
+ function REALMIN (X, Y : real ) return real;
+ attribute foreign of realmin : function is "VHPIDIRECT fmin";
-- returns the algebraically smaller of X and Y
procedure UNIFORM (variable Seed1,Seed2:inout integer; variable X:out real);
@@ -203,25 +208,17 @@
-- returns tan X; X in radians
-- X /= ((2k+1) * PI/2), where k is an integer
- -- IAC: function should be called ARCSIN
-
- function ASIN (X : real ) return real;
+ function ARCSIN (X : real ) return real;
-- returns -PI/2 < asin X < PI/2; | X | <= 1
- -- IAC: function should be called ARCCOS
-
- function ACOS (X : real ) return real;
+ function ARCCOS (X : real ) return real;
-- returns 0 < acos X < PI; | X | <= 1
-
- -- IAC: function should be called ARCTAN
-
- function ATAN (X : real) return real;
- attribute foreign of atan : function is "VHPIDIRECT atan";
+ function ARCTAN (X : real) return real;
+ attribute foreign of arctan : function is "VHPIDIRECT atan";
-- returns -PI/2 < atan X < PI/2
- -- IAC: function ATAN2 should not exist
- function ATAN2 (X : real; Y : real) return real;
+ function ARCTAN (X : real; Y : real) return real;
-- returns atan (X/Y); -PI < atan2(X,Y) < PI; Y /= 0.0
function SINH (X : real) return real;
@@ -236,20 +233,14 @@
attribute foreign of tanh : function is "VHPIDIRECT tanh";
-- hyperbolic tangent; -- returns (e**X - e**(-X))/(e**X + e**(-X))
- -- IAC: function should be called ARCSINH
-
- function ASINH (X : real) return real;
- attribute foreign of asinh : function is "VHPIDIRECT asinh";
+ function ARCSINH (X : real) return real;
+ attribute foreign of arcsinh : function is "VHPIDIRECT asinh";
-- returns ln( X + sqrt( X**2 + 1))
- -- IAC: function should be called ARCCOSH
-
- function ACOSH (X : real) return real;
+ function ARCCOSH (X : real) return real;
-- returns ln( X + sqrt( X**2 - 1)); X >= 1
- -- IAC: function should be called ARCTANH
-
- function ATANH (X : real) return real;
+ function ARCTANH (X : real) return real;
-- returns (ln( (1 + X)/(1 - X)))/2 ; | X | < 1
end MATH_REAL;
--- vhdl/libraries/ieee/math_real-body.vhdl.orig 2009-04-01 19:22:04.000000000 +0200
+++ vhdl/libraries/ieee/math_real-body.vhdl 2009-04-01 19:34:18.000000000 +0200
@@ -72,16 +72,41 @@
begin
assert false severity failure;
end ROUND;
-
- function FMAX (X, Y : real ) return real is
+
+ function TRUNC (X : real ) return real is
begin
assert false severity failure;
- end FMAX;
+ end TRUNC;
+
+ function c_mod (x : real; y : real) return real;
+ attribute foreign of c_mod : function is "VHPIDIRECT fmod";
- function FMIN (X, Y : real ) return real is
+ function c_mod (x : real; y: real) return real is
begin
assert false severity failure;
- end FMIN;
+ end c_mod;
+
+ function "MOD" (X, Y : real ) return real is
+ begin
+ if y = 0.0 then
+ assert false
+ report "MOD(X, 0.0) is undefined"
+ severity ERROR;
+ return 0.0;
+ else
+ return c_mod(x,y);
+ end if;
+ end "MOD";
+
+ function REALMAX (X, Y : real ) return real is
+ begin
+ assert false severity failure;
+ end REALMAX;
+
+ function REALMIN (X, Y : real ) return real is
+ begin
+ assert false severity failure;
+ end REALMIN;
--
-- Pseudo-random number generators
@@ -297,18 +322,18 @@
assert false severity failure;
end c_asin;
- function ASIN (x : real ) return real is
+ function ARCSIN (x : real ) return real is
-- returns -PI/2 < asin X < PI/2; | X | <= 1
begin
if abs x > 1.0 then
assert false
- report "Out of range parameter passed to ASIN"
+ report "Out of range parameter passed to ARCSIN"
severity ERROR;
return x;
else
return c_asin(x);
end if;
- end ASIN;
+ end ARCSIN;
function c_acos (x : real ) return real;
attribute foreign of c_acos : function is "VHPIDIRECT acos";
@@ -318,24 +343,24 @@
assert false severity failure;
end c_acos;
- function ACOS (x : REAL) return REAL is
+ function ARCCOS (x : REAL) return REAL is
-- returns 0 < acos X < PI; | X | <= 1
begin
if abs x > 1.0 then
assert false
- report "Out of range parameter passed to ACOS"
+ report "Out of range parameter passed to ARCCOS"
severity ERROR;
return x;
else
return c_acos(x);
end if;
- end ACOS;
+ end ARCCOS;
- function ATAN (x : REAL) return REAL is
+ function ARCTAN (x : REAL) return REAL is
-- returns -PI/2 < atan X < PI/2
- begin
+ begin
assert false severity failure;
- end ATAN;
+ end ARCTAN;
function c_atan2 (x : real; y : real) return real;
attribute foreign of c_atan2 : function is "VHPIDIRECT atan2";
@@ -345,7 +370,7 @@
assert false severity failure;
end c_atan2;
- function ATAN2 (x : REAL; y : REAL) return REAL is
+ function ARCTAN (x : REAL; y : REAL) return REAL is
-- returns atan (X/Y); -PI < atan2(X,Y) < PI; Y /= 0.0
begin
if y = 0.0 and x = 0.0 then
@@ -356,7 +381,7 @@
else
return c_atan2(x,y);
end if;
- end ATAN2;
+ end ARCTAN;
function SINH (X : real) return real is
@@ -377,11 +402,11 @@
assert false severity failure;
end TANH;
- function ASINH (X : real) return real is
+ function ARCSINH (X : real) return real is
-- returns ln( X + sqrt( X**2 + 1))
begin
assert false severity failure;
- end ASINH;
+ end ARCSINH;
function c_acosh (x : real ) return real;
attribute foreign of c_acosh : function is "VHPIDIRECT acosh";
@@ -391,16 +416,16 @@
assert false severity failure;
end c_acosh;
- function ACOSH (X : real) return real is
+ function ARCCOSH (X : real) return real is
-- returns ln( X + sqrt( X**2 - 1)); X >= 1
begin
if abs x >= 1.0 then
- assert false report "Out of range parameter passed to ACOSH"
+ assert false report "Out of range parameter passed to ARCCOSH"
severity ERROR;
return x;
end if;
return c_acosh(x);
- end ACOSH;
+ end ARCCOSH;
function c_atanh (x : real ) return real;
attribute foreign of c_atanh : function is "VHPIDIRECT atanh";
@@ -410,15 +435,15 @@
assert false severity failure;
end c_atanh;
- function ATANH (X : real) return real is
+ function ARCTANH (X : real) return real is
-- returns (ln( (1 + X)/(1 - X)))/2 ; | X | < 1
begin
if abs x < 1.0 then
- assert false report "Out of range parameter passed to ATANH"
+ assert false report "Out of range parameter passed to ARCTANH"
severity ERROR;
return x;
end if;
return c_atanh(x);
- end ATANH;
+ end ARCTANH;
end MATH_REAL;
--- vhdl/libraries/ieee/math_complex-body.vhdl.orig 2009-04-02 12:08:47.000000000 +0200
+++ vhdl/libraries/ieee/math_complex-body.vhdl 2009-04-02 12:09:18.000000000 +0200
@@ -119,7 +119,7 @@
function COMPLEX_TO_POLAR(Z: in complex ) return complex_polar is
-- converts complex to complex_polar
begin
- return COMPLEX_POLAR'(sqrt(z.re**2 + z.im**2),atan2(z.re,z.im));
+ return COMPLEX_POLAR'(sqrt(z.re**2 + z.im**2),ARCTAN(z.re,z.im));
end COMPLEX_TO_POLAR;
function POLAR_TO_COMPLEX(Z: in complex_polar ) return complex is
entity t is
end t;
library ieee;
use ieee.math_real.all;
architecture t of t is
begin
process
begin
for i in -10 to 10 loop
report "t=" & real'image(real(i) * 0.1)
& " ASIN(t)=" & real'image(asin(real(i) * 0.1))
& " ARCSIN(t)=" & real'image(arcsin(real(i) * 0.1))
severity note;
end loop;
wait;
end process;
end t;
--- gcc/config/rs6000/rs6000.c.orig 2009-02-13 19:43:05.000000000 +0100
+++ gcc/config/rs6000/rs6000.c 2009-02-13 19:45:29.000000000 +0100
@@ -16946,7 +16946,7 @@
else if (! strcmp (language_string, "GNU Objective-C"))
i = 14;
else
- gcc_unreachable ();
+ i = 0;
fprintf (file, "%d,", i);
/* 8 single bit fields: global linkage (not set for C extern linkage,
_______________________________________________
Ghdl-discuss mailing list
[email protected]
https://mail.gna.org/listinfo/ghdl-discuss
