[SOLVED] GHDL floating point exception 8

[dtsao]

0 down vote favorite


I am trying to code a SRAM with 32 bit address, with a Byte lane write enable in VHDL. But when I try to access (read or write) address greater than x1F, I get "Floating point exception 8" when compiling with GHDL. Here is some snippets of the code:

entity data_mem is

port(addr : in std_logic_vector(31 downto 0);
   enable : in std_logic;
   rd : in std_logic;
   wr : in std_logic;
   we : in std_logic_vector( 3 downto 0);
   din : in std_logic_vector( 31 downto 0);
   -- outputs 
   dout : out std_logic_vector(31 downto 0);
   ack : out std_logic
   );
end data_mem;

architecture structure of data_mem is

type mem_type is array (31 downto 0) of std_logic_vector(31 downto 0);
signal mem : mem_type := ((others => (others => '0'))); -- initialize to zero

begin

mem_write : process(addr,enable, wr, we, din)
begin
  if (enable = '1') then
    if (wr = '1') then
      if (we(0) = '1') then
        mem(to_integer(signed(addr)))(7 downto 0) <= din(7 downto 0) after 2 ns;
      end if; ...

So when I set the address to x0000_001F or lower in the testbench, it compiles, but not when I put x0000_0020 or greater.