Use far heap for TUI screen buffer on 16-bit DOS

The 4KB screen buffer (80x25x2 bytes) was allocated from near heap via
calloc(), which exhausted the 64KB data segment on 16-bit DOS. Now uses
_fcalloc()/_ffree() from OpenWatcom's far heap on 16-bit, keeping the
buffer outside DGROUP.

The TUI now works fully on 16-bit FreeDOS: full-screen editor, F-key
bar, cursor positioning, and scroll -- all via BIOS INT 10h through the
DOS HAL, with the screen buffer in far memory.

Changes:
- tui.h: GW_FAR macro (expands to __far on 16-bit, nothing elsewhere),
  tui.screen declared as tui_cell_t GW_FAR *
- tui.c: _fcalloc/_ffree for 16-bit, _fmemmove for scroll_up()
- TUI_CELL() macro works unchanged (far pointer dereference is
  transparent)

include/tui.h

@@ -4,6 +4,14 @@
 #include <stdint.h>
 #include <stdbool.h>
 
+/* On 16-bit DOS, the screen buffer lives in far heap to save near heap.
+ * GW_FAR expands to __far on 16-bit OpenWatcom, nothing elsewhere. */
+#if defined(_M_I86)
+#define GW_FAR __far
+#else
+#define GW_FAR
+#endif
+
 #define TUI_DEFAULT_ROWS 25
 #define TUI_DEFAULT_COLS 80
 #define TUI_MAX_ROWS 200
@@ -50,7 +58,7 @@ typedef struct {
 
 /* TUI state */
 typedef struct {
-    tui_cell_t *screen;             /* dynamically allocated [rows * cols] */
+    tui_cell_t GW_FAR *screen;      /* dynamically allocated [rows * cols] */
     int rows;                       /* screen height (default 25) */
     int cols;                       /* screen width (default 80) */
     int cursor_row;

src/tui.c

@@ -9,6 +9,9 @@
 #include <unistd.h>
 #include <sys/ioctl.h>
 #endif
+#ifdef _M_I86
+#include <malloc.h>  /* _fcalloc, _ffree */
+#endif
 
 tui_state_t tui;
 
@@ -37,8 +40,13 @@ static const char *default_fkeys[10] = {
 static void scroll_up(void)
 {
     int bottom = tui.view_bottom;
+#ifdef _M_I86
+    _fmemmove(&TUI_CELL(0, 0), &TUI_CELL(1, 0),
+              bottom * tui.cols * sizeof(tui_cell_t));
+#else
     memmove(&TUI_CELL(0, 0), &TUI_CELL(1, 0),
             bottom * tui.cols * sizeof(tui_cell_t));
+#endif
     for (int c = 0; c < tui.cols; c++) {
         TUI_CELL(bottom, c).ch = ' ';
         TUI_CELL(bottom, c).attr = tui.current_attr;
@@ -666,10 +674,13 @@ void tui_init(bool fullscreen)
     tui.view_bottom = tui.rows - 1;
 
     /* Allocate screen buffer */
+    /* On 16-bit DOS, use far heap to avoid exhausting the 64KB data segment */
+#ifdef _M_I86
+    tui.screen = _fcalloc(tui.rows * tui.cols, sizeof(tui_cell_t));
+#else
     tui.screen = calloc(tui.rows * tui.cols, sizeof(tui_cell_t));
+#endif
     if (!tui.screen) {
-        /* Not enough near heap (common on 16-bit DOS).
-         * Disable TUI — batch mode still works via HAL. */
         tui.active = false;
         return;
     }
@@ -732,7 +743,11 @@ void tui_shutdown(void)
     printf("\033[0 q");
     fflush(stdout);
 
+#ifdef _M_I86
+    _ffree(tui.screen);
+#else
     free(tui.screen);
+#endif
     tui.screen = NULL;
 
     signal(SIGINT, SIG_DFL);