/*      $NetBSD: ite_rt.c,v 1.25 2014/01/22 00:25:16 christos Exp $ */

/*
* Copyright (c) 1993 Markus Wild
* Copyright (c) 1993 Lutz Vieweg
* All rights reserved.
*
* Redistribution and use in source and binary forms, with or without
* modification, are permitted provided that the following conditions
* are met:
* 1. Redistributions of source code must retain the above copyright
*    notice, this list of conditions and the following disclaimer.
* 2. Redistributions in binary form must reproduce the above copyright
*    notice, this list of conditions and the following disclaimer in the
*    documentation and/or other materials provided with the distribution.
* 3. All advertising materials mentioning features or use of this software
*    must display the following acknowledgement:
*      This product includes software developed by Lutz Vieweg.
* 4. The name of the author may not be used to endorse or promote products
*    derived from this software without specific prior written permission
*
* THIS SOFTWARE IS PROVIDED BY THE AUTHOR ``AS IS'' AND ANY EXPRESS OR
* IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES
* OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED.
* IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR ANY DIRECT, INDIRECT,
* INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT
* NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
* DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
* THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
* (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF
* THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
*/

#include <sys/cdefs.h>
__KERNEL_RCSID(0, "$NetBSD: ite_rt.c,v 1.25 2014/01/22 00:25:16 christos Exp $");

#include "grfrt.h"
#if NGRFRT > 0

#include <sys/param.h>
#include <sys/conf.h>
#include <sys/proc.h>
#include <sys/device.h>
#include <sys/ioctl.h>
#include <sys/tty.h>
#include <sys/systm.h>
#include <dev/cons.h>
#include <machine/cpu.h>
#include <amiga/amiga/device.h>
#include <amiga/dev/itevar.h>
#include <amiga/dev/grfioctl.h>
#include <amiga/dev/grfvar.h>
#include <amiga/dev/grf_rtreg.h>

int retina_console = 1;

void retina_cursor(struct ite_softc *, int);
void retina_scroll(struct ite_softc *, int, int, int, int);
void retina_deinit(struct ite_softc *);
void retina_clear(struct ite_softc *, int, int, int, int);
void retina_putc(struct ite_softc *, int, int, int, int);
void retina_init(struct ite_softc *);

#ifdef RETINA_SPEED_HACK
static void screen_up(struct ite_softc *, int, int, int);
static void screen_down(struct ite_softc *, int, int, int);
#endif

/*
* this function is called from grf_rt to init the grf_softc->g_conpri
* field each time a retina is attached.
*/
int
grfrt_cnprobe(void)
{
       static int done;
       int rv;

       if (retina_console && done == 0)
               rv = CN_INTERNAL;
       else
               rv = CN_NORMAL;
       done = 1;
       return(rv);
}

/*
* init the required fields in the grf_softc struct for a
* grf to function as an ite.
*/
void
grfrt_iteinit(struct grf_softc *gp)
{
       gp->g_iteinit = retina_init;
       gp->g_itedeinit = retina_deinit;
       gp->g_iteclear = retina_clear;
       gp->g_iteputc = retina_putc;
       gp->g_itescroll = retina_scroll;
       gp->g_itecursor = retina_cursor;
}


void
retina_init(struct ite_softc *ip)
{
       struct MonDef *md;

       ip->priv = ip->grf->g_data;
       md = (struct MonDef *) ip->priv;

       ip->cols = md->TX;
       ip->rows = md->TY;
}


void
retina_cursor(struct ite_softc *ip, int flag)
{
     volatile void *ba = ip->grf->g_regkva;

     if (flag == ERASE_CURSOR)
       {
         /* disable cursor */
         WCrt (ba, CRT_ID_CURSOR_START, RCrt (ba, CRT_ID_CURSOR_START) | 0x20);
       }
     else
       {
         int pos = ip->curx + ip->cury * ip->cols;

         /* make sure to enable cursor */
         WCrt (ba, CRT_ID_CURSOR_START, RCrt (ba, CRT_ID_CURSOR_START) & ~0x20);

         /* and position it */
         WCrt (ba, CRT_ID_CURSOR_LOC_HIGH, (u_char) (pos >> 8));
         WCrt (ba, CRT_ID_CURSOR_LOC_LOW,  (u_char) pos);

         ip->cursorx = ip->curx;
         ip->cursory = ip->cury;
       }
}



#ifdef  RETINA_SPEED_HACK
static void
screen_up(struct ite_softc *ip, int top, int bottom, int lines)
{
       volatile void *ba = ip->grf->g_regkva;
       volatile void *fb = ip->grf->g_fbkva;
       const struct MonDef * md = (struct MonDef *) ip->priv;

       /* do some bounds-checking here.. */
       if (top >= bottom)
         return;

       if (top + lines >= bottom)
         {
           retina_clear (ip, top, 0, bottom - top, ip->cols);
           return;
         }

       /* the trick here is to use a feature of the NCR chip. It can
          optimize data access in various read/write modes. One of
          the modes is able to read/write from/to different zones.

          Thus, by setting the read-offset to lineN, and the write-offset
          to line0, we just cause read/write cycles for all characters
          up to the last line, and have the chip transfer the data. The
          `addqb' are the cheapest way to cause read/write cycles (DONT
          use `tas' on the Amiga!), their results are completely ignored
          by the NCR chip, it just replicates what it just read. */

               /* write to primary, read from secondary */
       WSeq (ba, SEQ_ID_EXTENDED_MEM_ENA,
               (RSeq(ba, SEQ_ID_EXTENDED_MEM_ENA) & 0x1f) | 0 );
               /* clear extended chain4 mode */
       WSeq (ba, SEQ_ID_EXT_VIDEO_ADDR, RSeq(ba, SEQ_ID_EXT_VIDEO_ADDR) & ~0x02);

               /* set write mode 1, "[...] data in the read latches is written
                  to memory during CPU memory write cycles. [...]" */
       WGfx (ba, GCT_ID_GRAPHICS_MODE,
               (RGfx(ba, GCT_ID_GRAPHICS_MODE) & 0xfc) | 1);

       {
               /* write to line TOP */
               long toploc = top * (md->TX / 16);
               WSeq (ba, SEQ_ID_PRIM_HOST_OFF_LO, ((unsigned char)toploc));
               WSeq (ba, SEQ_ID_PRIM_HOST_OFF_HI, ((unsigned char)(toploc >> 8)));
       }
       {
               /* read from line TOP + LINES */
               long fromloc = (top+lines) * (md->TX / 16);
               WSeq (ba, SEQ_ID_SEC_HOST_OFF_LO, ((unsigned char)fromloc)) ;
               WSeq (ba, SEQ_ID_SEC_HOST_OFF_HI, ((unsigned char)(fromloc >> 8))) ;
       }
       {
               void *p = (void *)fb;
               /* transfer all characters but LINES lines, unroll by 16 */
               short x = (1 + bottom - (top + lines)) * (md->TX / 16) - 1;
               do {
                       __asm volatile("addqb #1,%0@+" : "=a" (p) : "0" (p));
                       __asm volatile("addqb #1,%0@+" : "=a" (p) : "0" (p));
                       __asm volatile("addqb #1,%0@+" : "=a" (p) : "0" (p));
                       __asm volatile("addqb #1,%0@+" : "=a" (p) : "0" (p));
                       __asm volatile("addqb #1,%0@+" : "=a" (p) : "0" (p));
                       __asm volatile("addqb #1,%0@+" : "=a" (p) : "0" (p));
                       __asm volatile("addqb #1,%0@+" : "=a" (p) : "0" (p));
                       __asm volatile("addqb #1,%0@+" : "=a" (p) : "0" (p));
                       __asm volatile("addqb #1,%0@+" : "=a" (p) : "0" (p));
                       __asm volatile("addqb #1,%0@+" : "=a" (p) : "0" (p));
                       __asm volatile("addqb #1,%0@+" : "=a" (p) : "0" (p));
                       __asm volatile("addqb #1,%0@+" : "=a" (p) : "0" (p));
                       __asm volatile("addqb #1,%0@+" : "=a" (p) : "0" (p));
                       __asm volatile("addqb #1,%0@+" : "=a" (p) : "0" (p));
                       __asm volatile("addqb #1,%0@+" : "=a" (p) : "0" (p));
                       __asm volatile("addqb #1,%0@+" : "=a" (p) : "0" (p));
               } while (x--);
       }

               /* reset to default values */
       WSeq (ba, SEQ_ID_SEC_HOST_OFF_HI, 0);
       WSeq (ba, SEQ_ID_SEC_HOST_OFF_LO, 0);
       WSeq (ba, SEQ_ID_PRIM_HOST_OFF_HI, 0);
       WSeq (ba, SEQ_ID_PRIM_HOST_OFF_LO, 0);
               /* write mode 0 */
       WGfx (ba, GCT_ID_GRAPHICS_MODE,
               (RGfx(ba, GCT_ID_GRAPHICS_MODE) & 0xfc) | 0);
               /* extended chain4 enable */
       WSeq (ba, SEQ_ID_EXT_VIDEO_ADDR,
               RSeq(ba, SEQ_ID_EXT_VIDEO_ADDR) | 0x02);
               /* read/write to primary on A0, secondary on B0 */
       WSeq (ba, SEQ_ID_EXTENDED_MEM_ENA,
               (RSeq(ba, SEQ_ID_EXTENDED_MEM_ENA) & 0x1f) | 0x40);


       /* fill the free lines with spaces */

       {  /* feed latches with value */
               unsigned short * f = (unsigned short *) fb;

               f += (1 + bottom - lines) * md->TX * 2;
               *f = 0x2010;
       }

          /* clear extended chain4 mode */
       WSeq (ba, SEQ_ID_EXT_VIDEO_ADDR, RSeq(ba, SEQ_ID_EXT_VIDEO_ADDR) & ~0x02);
          /* set write mode 1, "[...] data in the read latches is written
             to memory during CPU memory write cycles. [...]" */
       WGfx (ba, GCT_ID_GRAPHICS_MODE, (RGfx(ba, GCT_ID_GRAPHICS_MODE) & 0xfc) | 1);

       {
               unsigned long * p = (unsigned long *) fb;
               short x = (lines * (md->TX/16)) - 1;
               const unsigned long dummyval = 0;

               p += (1 + bottom - lines) * (md->TX/4);

               do {
                       *p++ = dummyval;
                       *p++ = dummyval;
                       *p++ = dummyval;
                       *p++ = dummyval;
               } while (x--);
       }

          /* write mode 0 */
       WGfx (ba, GCT_ID_GRAPHICS_MODE, (RGfx(ba, GCT_ID_GRAPHICS_MODE) & 0xfc) | 0);
          /* extended chain4 enable */
       WSeq (ba, SEQ_ID_EXT_VIDEO_ADDR , RSeq(ba, SEQ_ID_EXT_VIDEO_ADDR) | 0x02);
};


static void
screen_down(struct ite_softc *ip, int top, int bottom, int lines)
{
       volatile void *ba = ip->grf->g_regkva;
       volatile void *fb = ip->grf->g_fbkva;
       const struct MonDef * md = (struct MonDef *) ip->priv;

       /* do some bounds-checking here.. */
       if (top >= bottom)
         return;

       if (top + lines >= bottom)
         {
           retina_clear (ip, top, 0, bottom - top, ip->cols);
           return;
         }

       /* see screen_up() for explanation of chip-tricks */

               /* write to primary, read from secondary */
       WSeq (ba, SEQ_ID_EXTENDED_MEM_ENA,
               (RSeq(ba, SEQ_ID_EXTENDED_MEM_ENA) & 0x1f) | 0 );
               /* clear extended chain4 mode */
       WSeq (ba, SEQ_ID_EXT_VIDEO_ADDR, RSeq(ba, SEQ_ID_EXT_VIDEO_ADDR) & ~0x02);

               /* set write mode 1, "[...] data in the read latches is written
                  to memory during CPU memory write cycles. [...]" */
       WGfx (ba, GCT_ID_GRAPHICS_MODE, (RGfx(ba, GCT_ID_GRAPHICS_MODE) & 0xfc) | 1);

       {
               /* write to line TOP + LINES */
               long toloc = (top + lines) * (md->TX / 16);
               WSeq (ba, SEQ_ID_PRIM_HOST_OFF_LO, ((unsigned char)toloc));
               WSeq (ba, SEQ_ID_PRIM_HOST_OFF_HI, ((unsigned char)(toloc >> 8)));
       }
       {
               /* read from line TOP */
               long fromloc = top * (md->TX / 16);
               WSeq (ba, SEQ_ID_SEC_HOST_OFF_LO, ((unsigned char)fromloc));
               WSeq (ba, SEQ_ID_SEC_HOST_OFF_HI, ((unsigned char)(fromloc >> 8))) ;
       }

       {
               void *p = (void *)fb;
               short x = (1 + bottom - (top + lines)) * (md->TX / 16) - 1;
               p += (1 + bottom - (top + lines)) * md->TX;
               do {
                       __asm volatile("addqb #1,%0@-" : "=a" (p) : "0" (p));
                       __asm volatile("addqb #1,%0@-" : "=a" (p) : "0" (p));
                       __asm volatile("addqb #1,%0@-" : "=a" (p) : "0" (p));
                       __asm volatile("addqb #1,%0@-" : "=a" (p) : "0" (p));
                       __asm volatile("addqb #1,%0@-" : "=a" (p) : "0" (p));
                       __asm volatile("addqb #1,%0@-" : "=a" (p) : "0" (p));
                       __asm volatile("addqb #1,%0@-" : "=a" (p) : "0" (p));
                       __asm volatile("addqb #1,%0@-" : "=a" (p) : "0" (p));
                       __asm volatile("addqb #1,%0@-" : "=a" (p) : "0" (p));
                       __asm volatile("addqb #1,%0@-" : "=a" (p) : "0" (p));
                       __asm volatile("addqb #1,%0@-" : "=a" (p) : "0" (p));
                       __asm volatile("addqb #1,%0@-" : "=a" (p) : "0" (p));
                       __asm volatile("addqb #1,%0@-" : "=a" (p) : "0" (p));
                       __asm volatile("addqb #1,%0@-" : "=a" (p) : "0" (p));
                       __asm volatile("addqb #1,%0@-" : "=a" (p) : "0" (p));
                       __asm volatile("addqb #1,%0@-" : "=a" (p) : "0" (p));
               } while (x--);
       }

       WSeq (ba, SEQ_ID_PRIM_HOST_OFF_HI, 0);
       WSeq (ba, SEQ_ID_PRIM_HOST_OFF_LO, 0);
       WSeq (ba, SEQ_ID_SEC_HOST_OFF_HI, 0);
       WSeq (ba, SEQ_ID_SEC_HOST_OFF_LO, 0);

               /* write mode 0 */
       WGfx (ba, GCT_ID_GRAPHICS_MODE,
               (RGfx(ba, GCT_ID_GRAPHICS_MODE) & 0xfc) | 0);
               /* extended chain4 enable */
       WSeq (ba, SEQ_ID_EXT_VIDEO_ADDR , RSeq(ba, SEQ_ID_EXT_VIDEO_ADDR) | 0x02);
               /* read/write to primary on A0, secondary on B0 */
       WSeq (ba, SEQ_ID_EXTENDED_MEM_ENA,
               (RSeq(ba, SEQ_ID_EXTENDED_MEM_ENA) & 0x1f) | 0x40 );

       /* fill the free lines with spaces */

       {  /* feed latches with value */
               unsigned short * f = (unsigned short *) fb;

               f += top * md->TX * 2;
               *f = 0x2010;
       }

          /* clear extended chain4 mode */
       WSeq (ba, SEQ_ID_EXT_VIDEO_ADDR, RSeq(ba, SEQ_ID_EXT_VIDEO_ADDR) & ~0x02);
          /* set write mode 1, "[...] data in the read latches is written
             to memory during CPU memory write cycles. [...]" */
       WGfx (ba, GCT_ID_GRAPHICS_MODE, (RGfx(ba, GCT_ID_GRAPHICS_MODE) & 0xfc) | 1);

       {
               unsigned long * p = (unsigned long *) fb;
               short x = (lines * (md->TX/16)) - 1;
               const unsigned long dummyval = 0;

               p += top * (md->TX/4);

               do {
                       *p++ = dummyval;
                       *p++ = dummyval;
                       *p++ = dummyval;
                       *p++ = dummyval;
               } while (x--);
       }

          /* write mode 0 */
       WGfx (ba, GCT_ID_GRAPHICS_MODE, (RGfx(ba, GCT_ID_GRAPHICS_MODE) & 0xfc) | 0);
          /* extended chain4 enable */
       WSeq (ba, SEQ_ID_EXT_VIDEO_ADDR , RSeq(ba, SEQ_ID_EXT_VIDEO_ADDR) | 0x02);
};
#endif  /* RETINA_SPEED_HACK */


void
retina_deinit(struct ite_softc *ip)
{
       ip->flags &= ~ITE_INITED;
}


void
retina_putc(struct ite_softc *ip, int c, int dy, int dx, int mode)
{
       volatile char *fb = (volatile char*)ip->grf->g_fbkva;
       register u_char attr;

       attr = (mode & ATTR_INV) ? 0x21 : 0x10;
       if (mode & ATTR_UL)     attr  = 0x01;   /* ???????? */
       if (mode & ATTR_BOLD)   attr |= 0x08;
       if (mode & ATTR_BLINK)  attr |= 0x80;

       fb += 4 * (dy * ip->cols + dx);
       *fb++ = c; *fb = attr;
}


void
retina_clear(struct ite_softc *ip, int sy, int sx, int h, int w)
{
       volatile u_short * fb = (volatile u_short *) ip->grf->g_fbkva;
       short x;
       const u_short fillval = 0x2010;

       /* could probably be optimized just like the scrolling functions !! */
       fb += 2 * (sy * ip->cols + sx);
       while (h--)
         {
           for (x = 2 * (w - 1); x >= 0; x -= 2)
             fb[x] = fillval;
           fb += 2 * ip->cols;
         }
}


/*
* RETINA_SPEED_HACK code seems to work on some boards and on others
* it causes text to smear horizontally
*/
void
retina_scroll(struct ite_softc *ip, int sy, int sx, int count, int dir)
{
       u_long *fb;

       fb = (u_long *)__UNVOLATILE(ip->grf->g_fbkva);

       retina_cursor(ip, ERASE_CURSOR);

       if (dir == SCROLL_UP) {
#ifdef  RETINA_SPEED_HACK
               screen_up(ip, sy - count, ip->bottom_margin, count);
#else
               memcpy(fb + (sy - count) * ip->cols, fb + sy * ip->cols,
                   4 * (ip->bottom_margin - sy + 1) * ip->cols);
               retina_clear(ip, ip->bottom_margin + 1 - count, 0, count,
                   ip->cols);
#endif
       } else if (dir == SCROLL_DOWN) {
#ifdef  RETINA_SPEED_HACK
               screen_down(ip, sy, ip->bottom_margin, count);
#else
               memcpy(fb + (sy + count) * ip->cols, fb + sy * ip->cols,
                   4 * (ip->bottom_margin - sy - count + 1) * ip->cols);
               retina_clear(ip, sy, 0, count, ip->cols);
#endif
       } else if (dir == SCROLL_RIGHT) {
               memcpy(fb + sx + sy * ip->cols + count, fb + sx + sy * ip->cols,
                   4 * (ip->cols - (sx + count)));
               retina_clear(ip, sy, sx, 1, count);
       } else {
               memcpy(fb + sx - count + sy * ip->cols, fb + sx + sy * ip->cols,
                   4 * (ip->cols - sx));
               retina_clear(ip, sy, ip->cols - count, 1, count);
       }
#ifndef RETINA_SPEED_HACK
       retina_cursor(ip, !ERASE_CURSOR);
#endif
}

#endif /* NGRFRT */