Write (WE) active during reset

[jesari-git]

When 'reset' is active the write signal 'WE' is also active. This means some memory corruption during reset (but confined to the stack page)
These write cycles shouldn't be a problem unless the clock frequency is out of specs (but this can happen during PLL startups), and the boot memory is writable.
I think the real 6502 also modifies the stack pointer during resets but its R/W signal is forced high, disabling writes.
A possible patch is to change the following line...:

/*

• Write Enable Generator
  */

always @*
case( state )
BRK0, // writing to stack or memory
BRK1,
BRK2,
JSR0,
JSR1,
PUSH1,
WRITE: WE = ~(reset|res); // (before it was: WE = 1;)

    INDX3,  // only if doing a STA, STX or STY
    INDY3,
    ABSX2,
    ABS1,
    ZPX1,
    ZP0:     WE = store;

    default: WE = 0;
endcase

[jesari-git]

Well, I mean the line WRITE: WE = 1; , now WE = ~(reset|res); I got the code a bit disturbed in the post ;)

[rejunity]

I had the same issue too! Fixed it in a bit different way: rejunity/tt09-atari-2600@162c005

WE = ~res;

I also found a number of other issues with RESET sequence, such as D flag not properly being initialised.

List of my fixes: https://github.com/rejunity/tt09-atari-2600/commits/main/src/6502.v (NOTE fixes are aimed at producing silicon / PDK)