[multiple changes]

2010-06-22  Robert Dewar  <dewar@adacore.com>

	* freeze.adb: Minor reformatting
	Minor code reorganization (use Nkind_In and Ekind_In).

2010-06-22  Bob Duff  <duff@adacore.com>

	* gnat1drv.adb (Gnat1drv): Remove the messages that recommend using 
	-gnatc when a file is compiled that we cannot generate code for, not
	helpful and confusing.

2010-06-22  Vincent Celier  <celier@adacore.com>

	* switch-m.adb (Normalize_Compiler_Switches): Process correctly
	switches -gnatknn.

2010-06-22  Paul Hilfinger  <hilfinger@adacore.com>

	* s-rannum.adb: Replace constants with commented symbols.
	* s-rannum.ads: Explain significance of the initial value of the data
	structure.

2010-06-22  Ed Schonberg  <schonberg@adacore.com>

	* a-ngcoty.adb: Clarify comment.

2010-06-22  Gary Dismukes  <dismukes@adacore.com>

	* exp_pakd.adb (Expand_Bit_Packed_Element_Set): Return without
	expansion for indexing packed arrays with small power-of-2 component
	sizes when the target is AAMP.
	(Expand_Packed_Element_Reference): Return without expansion for
	indexing packed arrays with small power-of-2 component sizes when the
	target is AAMP.

2010-06-22  Geert Bosch  <bosch@adacore.com>

	* exp_ch4.adb (Expand_N_In): Do not substitute a valid check for X in
	Float'Range.

2010-06-22  Robert Dewar  <dewar@adacore.com>

	* g-mbdira.adb, g-mbflra.adb, a-nuflra.adb, a-nudira.adb: Minor comment
	updates.

From-SVN: r161213

gcc/ada/ChangeLog

+2010-06-22  Robert Dewar  <dewar@adacore.com>
+
+	* freeze.adb: Minor reformatting
+	Minor code reorganization (use Nkind_In and Ekind_In).
+
+2010-06-22  Bob Duff  <duff@adacore.com>
+
+	* gnat1drv.adb (Gnat1drv): Remove the messages that recommend using 
+	-gnatc when a file is compiled that we cannot generate code for, not
+	helpful and confusing.
+
+2010-06-22  Vincent Celier  <celier@adacore.com>
+
+	* switch-m.adb (Normalize_Compiler_Switches): Process correctly
+	switches -gnatknn.
+
+2010-06-22  Paul Hilfinger  <hilfinger@adacore.com>
+
+	* s-rannum.adb: Replace constants with commented symbols.
+	* s-rannum.ads: Explain significance of the initial value of the data
+	structure.
+
+2010-06-22  Ed Schonberg  <schonberg@adacore.com>
+
+	* a-ngcoty.adb: Clarify comment.
+
+2010-06-22  Gary Dismukes  <dismukes@adacore.com>
+
+	* exp_pakd.adb (Expand_Bit_Packed_Element_Set): Return without
+	expansion for indexing packed arrays with small power-of-2 component
+	sizes when the target is AAMP.
+	(Expand_Packed_Element_Reference): Return without expansion for
+	indexing packed arrays with small power-of-2 component sizes when the
+	target is AAMP.
+
+2010-06-22  Geert Bosch  <bosch@adacore.com>
+
+	* exp_ch4.adb (Expand_N_In): Do not substitute a valid check for X in
+	Float'Range.
+
+2010-06-22  Robert Dewar  <dewar@adacore.com>
+
+	* g-mbdira.adb, g-mbflra.adb, a-nuflra.adb, a-nudira.adb: Minor comment
+	updates.
+
 2010-06-22  Doug Rupp  <rupp@adacore.com>
 
 	* system-vms.ads, system-vms-zcx.ads: Remove old unused VMS system

gcc/ada/a-ngcoty.adb

@@ -60,15 +60,16 @@ package body Ada.Numerics.Generic_Complex_Types is
 
       if not Standard'Fast_Math then
 
-         --  ??? the test below is weird, it needs a comment, otherwise I or
-         --  someone else will change it back to R'Last > abs (X) ???
+         --  Note that the test below is written as a negation. This is to
+         --  account for the fact that X and Y may be NaNs, because both of
+         --  their operands could overflow. Given that all operations on NaNs
+         --  return false, the test can only be written thus.
 
          if not (abs (X) <= R'Last) then
             X := Scale**2 * ((Left.Re / Scale) * (Right.Re / Scale) -
                              (Left.Im / Scale) * (Right.Im / Scale));
          end if;
 
-         --  ??? same weird test ???
          if not (abs (Y) <= R'Last) then
             Y := Scale**2 * ((Left.Re / Scale) * (Right.Im / Scale)
                            + (Left.Im / Scale) * (Right.Re / Scale));

gcc/ada/a-nudira.adb

@@ -37,13 +37,14 @@ package body Ada.Numerics.Discrete_Random is
    -- Implementation Note --
    -------------------------
 
-   --  The design of this spec is very awkward, as a result of Ada 95 not
+   --  The design of this spec is a bit awkward, as a result of Ada 95 not
    --  permitting in-out parameters for function formals (most naturally
    --  Generator values would be passed this way). In pure Ada 95, the only
-   --  solution is to use the heap and pointers, and, to avoid memory leaks,
-   --  controlled types.
+   --  solution would be to add a self-referential component to the generator
+   --  allowing access to the generator object from inside the function. This
+   --  would work because the generator is limited, which prevents any copy.
 
-   --  This is awfully heavy, so what we do is to use Unrestricted_Access to
+   --  This is a bit heavy, so what we do is to use Unrestricted_Access to
    --  get a pointer to the state in the passed Generator. This works because
    --  Generator is a limited type and will thus always be passed by reference.
 

gcc/ada/a-nuflra.adb

@@ -39,13 +39,14 @@ package body Ada.Numerics.Float_Random is
    -- Implementation Note --
    -------------------------
 
-   --  The design of this spec is very awkward, as a result of Ada 95 not
+   --  The design of this spec is a bit awkward, as a result of Ada 95 not
    --  permitting in-out parameters for function formals (most naturally
    --  Generator values would be passed this way). In pure Ada 95, the only
-   --  solution is to use the heap and pointers, and, to avoid memory leaks,
-   --  controlled types.
+   --  solution would be to add a self-referential component to the generator
+   --  allowing access to the generator object from inside the function. This
+   --  would work because the generator is limited, which prevents any copy.
 
-   --  This is awfully heavy, so what we do is to use Unrestricted_Access to
+   --  This is a bit heavy, so what we do is to use Unrestricted_Access to
    --  get a pointer to the state in the passed Generator. This works because
    --  Generator is a limited type and will thus always be passed by reference.
 

gcc/ada/exp_ch4.adb

@@ -4378,9 +4378,12 @@ package body Exp_Ch4 is
 
       --  Check case of explicit test for an expression in range of its
       --  subtype. This is suspicious usage and we replace it with a 'Valid
-      --  test and give a warning.
+      --  test and give a warning. For floating point types however, this
+      --  is a standard way to check for finite numbers, and using 'Valid
+      --  would typically be a pessimization
 
       if Is_Scalar_Type (Etype (Lop))
+        and then not Is_Floating_Point_Type (Etype (Lop))
         and then Nkind (Rop) in N_Has_Entity
         and then Etype (Lop) = Entity (Rop)
         and then Comes_From_Source (N)

gcc/ada/exp_pakd.adb

@@ -1381,6 +1381,19 @@ package body Exp_Pakd is
          Analyze_And_Resolve (Rhs, Ctyp);
       end if;
 
+      --  For the AAMP target, indexing of certain packed array is passed
+      --  through to the back end without expansion, because the expansion
+      --  results in very inefficient code on that target. This allows the
+      --  GNAAMP back end to generate specialized macros that support more
+      --  efficient indexing of packed arrays with components having sizes
+      --  that are small powers of two.
+
+      if AAMP_On_Target
+        and then (Csiz = 1 or else Csiz = 2 or else Csiz = 4)
+      then
+         return;
+      end if;
+
       --  Case of component size 1,2,4 or any component size for the modular
       --  case. These are the cases for which we can inline the code.
 
@@ -1933,6 +1946,19 @@ package body Exp_Pakd is
       Ctyp := Component_Type (Atyp);
       Csiz := UI_To_Int (Component_Size (Atyp));
 
+      --  For the AAMP target, indexing of certain packed array is passed
+      --  through to the back end without expansion, because the expansion
+      --  results in very inefficient code on that target. This allows the
+      --  GNAAMP back end to generate specialized macros that support more
+      --  efficient indexing of packed arrays with components having sizes
+      --  that are small powers of two.
+
+      if AAMP_On_Target
+        and then (Csiz = 1 or else Csiz = 2 or else Csiz = 4)
+      then
+         return;
+      end if;
+
       --  Case of component size 1,2,4 or any component size for the modular
       --  case. These are the cases for which we can inline the code.
 

gcc/ada/g-mbdira.adb

@@ -39,13 +39,14 @@ package body GNAT.MBBS_Discrete_Random is
    -- Implementation Note --
    -------------------------
 
-   --  The design of this spec is very awkward, as a result of Ada 95 not
+   --  The design of this spec is a bit awkward, as a result of Ada 95 not
    --  permitting in-out parameters for function formals (most naturally
    --  Generator values would be passed this way). In pure Ada 95, the only
-   --  solution is to use the heap and pointers, and, to avoid memory leaks,
-   --  controlled types.
+   --  solution would be to add a self-referential component to the generator
+   --  allowing access to the generator object from inside the function. This
+   --  would work because the generator is limited, which prevents any copy.
 
-   --  This is awfully heavy, so what we do is to use Unrestricted_Access to
+   --  This is a bit heavy, so what we do is to use Unrestricted_Access to
    --  get a pointer to the state in the passed Generator. This works because
    --  Generator is a limited type and will thus always be passed by reference.
 

gcc/ada/g-mbflra.adb

@@ -37,13 +37,14 @@ package body GNAT.MBBS_Float_Random is
    -- Implementation Note --
    -------------------------
 
-   --  The design of this spec is very awkward, as a result of Ada 95 not
+   --  The design of this spec is a bit awkward, as a result of Ada 95 not
    --  permitting in-out parameters for function formals (most naturally
    --  Generator values would be passed this way). In pure Ada 95, the only
-   --  solution is to use the heap and pointers, and, to avoid memory leaks,
-   --  controlled types.
+   --  solution would be to add a self-referential component to the generator
+   --  allowing access to the generator object from inside the function. This
+   --  would work because the generator is limited, which prevents any copy.
 
-   --  This is awfully heavy, so what we do is to use Unrestricted_Access to
+   --  This is a bit heavy, so what we do is to use Unrestricted_Access to
    --  get a pointer to the state in the passed Generator. This works because
    --  Generator is a limited type and will thus always be passed by reference.
 

gcc/ada/gnat1drv.adb

@@ -861,42 +861,28 @@ begin
          if Subunits_Missing then
             Write_Str (" (missing subunits)");
             Write_Eol;
-            Write_Str ("to check parent unit");
 
          elsif Main_Kind = N_Subunit then
             Write_Str (" (subunit)");
             Write_Eol;
-            Write_Str ("to check subunit");
 
          elsif Main_Kind = N_Subprogram_Declaration then
             Write_Str (" (subprogram spec)");
             Write_Eol;
-            Write_Str ("to check subprogram spec");
 
          --  Generic package body in GNAT implementation mode
 
          elsif Main_Kind = N_Package_Body and then GNAT_Mode then
             Write_Str (" (predefined generic)");
             Write_Eol;
-            Write_Str ("to check predefined generic");
 
          --  Only other case is a package spec
 
          else
             Write_Str (" (package spec)");
             Write_Eol;
-            Write_Str ("to check package spec");
          end if;
 
-         Write_Str (" for errors, use ");
-
-         if Hostparm.OpenVMS then
-            Write_Str ("/NOLOAD");
-         else
-            Write_Str ("-gnatc");
-         end if;
-
-         Write_Eol;
          Set_Standard_Output;
 
          Sem_Ch13.Validate_Unchecked_Conversions;

gcc/ada/s-rannum.adb

@@ -99,30 +99,71 @@ package body System.Random_Numbers is
    -- Implementation Note --
    -------------------------
 
-   --  The design of this spec is very awkward, as a result of Ada 95 not
-   --  permitting in-out parameters for function formals (most naturally,
+   --  The design of this spec is a bit awkward, as a result of Ada 95 not
+   --  permitting in-out parameters for function formals (most naturally
    --  Generator values would be passed this way). In pure Ada 95, the only
-   --  solution is to use the heap and pointers, and, to avoid memory leaks,
-   --  controlled types.
+   --  solution would be to add a self-referential component to the generator
+   --  allowing access to the generator object from inside the function. This
+   --  would work because the generator is limited, which prevents any copy.
 
-   --  This is awfully heavy, so what we do is to use Unrestricted_Access to
+   --  This is a bit heavy, so what we do is to use Unrestricted_Access to
    --  get a pointer to the state in the passed Generator. This works because
    --  Generator is a limited type and will thus always be passed by reference.
 
-   Low31_Mask : constant := 2**31-1;
-   Bit31_Mask : constant := 2**31;
-
-   Matrix_A_X : constant array (State_Val range 0 .. 1) of State_Val :=
-                  (0, 16#9908b0df#);
-
    Y2K : constant Calendar.Time :=
            Calendar.Time_Of
              (Year => 2000, Month => 1, Day => 1, Seconds => 0.0);
-   --  First Year 2000 day
+   --  First day of Year 2000 (what is this for???)
 
    Image_Numeral_Length : constant := Max_Image_Width / N;
    subtype Image_String is String (1 .. Max_Image_Width);
 
+   ----------------------------
+   -- Algorithmic Parameters --
+   ----------------------------
+
+   Lower_Mask : constant := 2**31-1;
+   Upper_Mask : constant := 2**31;
+
+   Matrix_A   : constant array (State_Val range 0 .. 1) of State_Val
+     := (0, 16#9908b0df#);
+   --  The twist transformation is represented by a matrix of the form
+   --
+   --               [  0    I(31) ]
+   --               [    _a       ]
+   --
+   --  where 0 is a 31x31 block of 0s, I(31) is the 31x31 identity matrix and
+   --  _a is a particular bit row-vector, represented here by a 32-bit integer.
+   --  If integer x represents a row vector of bits (with x(0), the units bit,
+   --  last), then
+   --           x * A = [0 x(31..1)] xor Matrix_A(x(0)).
+
+   U      : constant := 11;
+   S      : constant := 7;
+   B_Mask : constant := 16#9d2c5680#;
+   T      : constant := 15;
+   C_Mask : constant := 16#efc60000#;
+   L      : constant := 18;
+   --  The tempering shifts and bit masks, in the order applied
+
+   Seed0 : constant := 5489;
+   --  Default seed, used to initialize the state vector when Reset not called
+
+   Seed1 : constant := 19650218;
+   --  Seed used to initialize the state vector when calling Reset with an
+   --  initialization vector.
+
+   Mult0 : constant := 1812433253;
+   --  Multiplier for a modified linear congruential generator used to
+   --  initialize the state vector when calling Reset with a single integer
+   --  seed.
+
+   Mult1 : constant := 1664525;
+   Mult2 : constant := 1566083941;
+   --  Multipliers for two modified linear congruential generators used to
+   --  initialize the state vector when calling Reset with an initialization
+   --  vector.
+
    -----------------------
    -- Local Subprograms --
    -----------------------
@@ -153,40 +194,40 @@ package body System.Random_Numbers is
    function Random (Gen : Generator) return Unsigned_32 is
       G : Generator renames Gen'Unrestricted_Access.all;
       Y : State_Val;
-      I : Integer;
+      I : Integer;      --  should avoid use of identifier I ???
 
    begin
       I := G.I;
 
       if I < N - M then
-         Y := (G.S (I) and Bit31_Mask) or (G.S (I + 1) and Low31_Mask);
-         Y := G.S (I + M) xor Shift_Right (Y, 1) xor Matrix_A_X (Y and 1);
+         Y := (G.S (I) and Upper_Mask) or (G.S (I + 1) and Lower_Mask);
+         Y := G.S (I + M) xor Shift_Right (Y, 1) xor Matrix_A (Y and 1);
          I := I + 1;
 
       elsif I < N - 1 then
-         Y := (G.S (I) and Bit31_Mask) or (G.S (I + 1) and Low31_Mask);
+         Y := (G.S (I) and Upper_Mask) or (G.S (I + 1) and Lower_Mask);
          Y := G.S (I + (M - N))
                 xor Shift_Right (Y, 1)
-                xor Matrix_A_X (Y and 1);
+                xor Matrix_A (Y and 1);
          I := I + 1;
 
       elsif I = N - 1 then
-         Y := (G.S (I) and Bit31_Mask) or (G.S (0) and Low31_Mask);
-         Y := G.S (M - 1) xor Shift_Right (Y, 1) xor Matrix_A_X (Y and 1);
+         Y := (G.S (I) and Upper_Mask) or (G.S (0) and Lower_Mask);
+         Y := G.S (M - 1) xor Shift_Right (Y, 1) xor Matrix_A (Y and 1);
          I := 0;
 
       else
-         Init (G, 5489);
+         Init (G, Seed0);
          return Random (Gen);
       end if;
 
       G.S (G.I) := Y;
       G.I := I;
 
-      Y := Y xor Shift_Right (Y, 11);
-      Y := Y xor (Shift_Left (Y, 7)  and 16#9d2c5680#);
-      Y := Y xor (Shift_Left (Y, 15) and 16#efc60000#);
-      Y := Y xor Shift_Right (Y, 18);
+      Y := Y xor Shift_Right (Y, U);
+      Y := Y xor (Shift_Left (Y, S)  and B_Mask);
+      Y := Y xor (Shift_Left (Y, T) and C_Mask);
+      Y := Y xor Shift_Right (Y, L);
 
       return Y;
    end Random;
@@ -265,17 +306,10 @@ package body System.Random_Numbers is
 
             Mantissa : Unsigned;
 
-            X : Real;
-            --  Scaled mantissa
-
-            R : Unsigned_32;
-            --  Supply of random bits
-
-            R_Bits : Natural;
-            --  Number of bits left in R
-
-            K : Bit_Count;
-            --  Next decrement to exponent
+            X      : Real;            --  Scaled mantissa
+            R      : Unsigned_32;     --  Supply of random bits
+            R_Bits : Natural;         --  Number of bits left in R
+            K      : Bit_Count;       --  Next decrement to exponent
 
          begin
             Mantissa := Random (Gen) / 2**Extra_Bits;
@@ -388,7 +422,7 @@ package body System.Random_Numbers is
          declare
             --  In the 64-bit case, we have to be careful, since not all 64-bit
             --  unsigned values are representable in GNAT's root_integer type.
-            --  Ignore different-size warnings here; since GNAT's handling
+            --  Ignore different-size warnings here since GNAT's handling
             --  is correct.
 
             pragma Warnings ("Z");  -- better to use msg string! ???
@@ -482,7 +516,7 @@ package body System.Random_Numbers is
 
    procedure Reset (Gen : out Generator; Initiator : Integer) is
    begin
-      pragma Warnings ("C");
+      pragma Warnings (Off, "condition is always *");
       --  This is probably an unnecessary precaution against future change, but
       --  since the test is a static expression, no extra code is involved.
 
@@ -502,14 +536,14 @@ package body System.Random_Numbers is
          end;
       end if;
 
-      pragma Warnings ("c");
+      pragma Warnings (On, "condition is always *");
    end Reset;
 
    procedure Reset (Gen : out Generator; Initiator : Initialization_Vector) is
       I, J : Integer;
 
    begin
-      Init (Gen, 19650218); -- please give this constant a name ???
+      Init (Gen, Seed1);
       I := 1;
       J := 0;
 
@@ -517,8 +551,8 @@ package body System.Random_Numbers is
          for K in reverse 1 .. Integer'Max (N, Initiator'Length) loop
             Gen.S (I) :=
               (Gen.S (I)
-                 xor ((Gen.S (I - 1) xor Shift_Right (Gen.S (I - 1), 30))
-                                                                 * 1664525))
+               xor ((Gen.S (I - 1) xor Shift_Right (Gen.S (I - 1), 30))
+                    * Mult1))
               + Initiator (J + Initiator'First) + Unsigned_32 (J);
 
             I := I + 1;
@@ -538,7 +572,7 @@ package body System.Random_Numbers is
       for K in reverse 1 .. N - 1 loop
          Gen.S (I) :=
            (Gen.S (I) xor ((Gen.S (I - 1)
-                            xor Shift_Right (Gen.S (I - 1), 30)) * 1566083941))
+                            xor Shift_Right (Gen.S (I - 1), 30)) * Mult2))
            - Unsigned_32 (I);
          I := I + 1;
 
@@ -548,7 +582,7 @@ package body System.Random_Numbers is
          end if;
       end loop;
 
-      Gen.S (0) := Bit31_Mask;
+      Gen.S (0) := Upper_Mask;
    end Reset;
 
    procedure Reset (Gen : out Generator; From_State : Generator) is
@@ -612,7 +646,6 @@ package body System.Random_Numbers is
 
    begin
       Result := (others => ' ');
-
       for J in 0 .. N - 1 loop
          Insert_Image (Result, J, Gen.S ((J + Gen.I) mod N));
       end loop;
@@ -643,9 +676,8 @@ package body System.Random_Numbers is
 
       for I in 1 .. N - 1 loop
          Gen.S (I) :=
-           1812433253
-             * (Gen.S (I - 1) xor Shift_Right (Gen.S (I - 1), 30))
-           + Unsigned_32 (I);
+           Mult0 * (Gen.S (I - 1) xor Shift_Right (Gen.S (I - 1), 30)) +
+                                                             Unsigned_32 (I);
       end loop;
 
       Gen.I := 0;

gcc/ada/s-rannum.ads

@@ -6,7 +6,7 @@
 --                                                                          --
 --                                 S p e c                                  --
 --                                                                          --
---          Copyright (C) 2007,2009  Free Software Foundation, Inc.         --
+--          Copyright (C) 2007-2010, Free Software Foundation, Inc.         --
 --                                                                          --
 -- GNAT is free software;  you can  redistribute it  and/or modify it under --
 -- terms of the  GNU General Public License as published  by the Free Soft- --
@@ -140,7 +140,7 @@ private
       --  The shift register, a circular buffer
 
       I : Integer := N;
-      --  Current starting position in shift register S
+      --  Current starting position in shift register S (N means uninitialized)
    end record;
 
 end System.Random_Numbers;

gcc/ada/switch-m.adb

@@ -215,10 +215,10 @@ package body Switch.M is
 
                   --  One-letter switches
 
-                  when 'a' | 'A' | 'b' | 'c' | 'D' | 'E' | 'f' |
-                    'F' | 'g' | 'h' | 'H' | 'k' | 'l' | 'L' | 'n' | 'N' |
-                    'o' | 'O' | 'p' | 'P' | 'q' | 'Q' | 'r' | 's' | 't' |
-                    'u' | 'U' | 'v' | 'x' | 'X' | 'Z' =>
+                  when 'a' | 'A' | 'b' | 'c' | 'D' | 'E' | 'f' | 'F' |
+                       'g' | 'h' | 'H' | 'l' | 'L' | 'n' | 'N' | 'o' |
+                       'O' | 'p' | 'P' | 'q' | 'Q' | 'r' | 's' | 't' |
+                       'u' | 'U' | 'v' | 'x' | 'X' | 'Z' =>
                      Storing (First_Stored) := C;
                      Add_Switch_Component
                        (Storing (Storing'First .. First_Stored));
@@ -226,7 +226,7 @@ package body Switch.M is
 
                   --  One-letter switches followed by a positive number
 
-                  when 'm' | 'T' =>
+                  when 'k' | 'm' | 'T' =>
                      Storing (First_Stored) := C;
                      Last_Stored := First_Stored;