15 files changed, 1328 insertions, 202 deletions

diff --git a/apps/plugins/SOURCES b/apps/plugins/SOURCES
index 0e0a78bc9d..6f6d2e74b8 100644
--- a/apps/plugins/SOURCES
+++ b/apps/plugins/SOURCES
@@ -27,12 +27,11 @@ dice.c
 #ifdef HAVE_LCD_BITMAP             /* Not for the Player */
 text_editor.c
 
-#if CONFIG_LCD != LCD_IPOD2BPP     /* Plugins needing the grayscale lib */
+/* Plugins needing the grayscale lib on low-depth LCDs */
 fire.c
 jpeg.c
 mandelbrot.c
 plasma.c
-#endif
 
 bounce.c
 bubbles.c
diff --git a/apps/plugins/SUBDIRS b/apps/plugins/SUBDIRS
index 1c500fc201..2d0d6d3c19 100644
--- a/apps/plugins/SUBDIRS
+++ b/apps/plugins/SUBDIRS
@@ -28,7 +28,7 @@ pacbox
 #endif
 
 /* For all the color targets apart from the Gigabeat */
-#if defined(HAVE_LCD_COLOR) && !defined(TOSHIBA_GIGABEAT_F) || defined(IRIVER_H100_SERIES)
+#if (defined(HAVE_LCD_COLOR) && !defined(TOSHIBA_GIGABEAT_F)) || (LCD_DEPTH == 2)
 doom
 #endif
 
diff --git a/apps/plugins/cube.c b/apps/plugins/cube.c
index f9a0218154..38b0233b86 100644
--- a/apps/plugins/cube.c
+++ b/apps/plugins/cube.c
@@ -547,7 +547,7 @@ enum plugin_status plugin_start(struct plugin_api* api, void* parameter)
     xlcd_init(rb);
 #elif defined(USE_GSLIB)
     gbuf = (unsigned char *)rb->plugin_get_buffer(&gbuf_size);
-    if (gray_init(rb, gbuf, gbuf_size, true, LCD_WIDTH, LCD_HEIGHT/8, 3, 0, NULL)
+    if (gray_init(rb, gbuf, gbuf_size, true, LCD_WIDTH, LCD_HEIGHT, 3, 0, NULL)
         != 3)
     {
         rb->splash(HZ, true, "Couldn't get grayscale buffer");
diff --git a/apps/plugins/doom/i_video.c b/apps/plugins/doom/i_video.c
index d8ed74ce62..abab755d6c 100644
--- a/apps/plugins/doom/i_video.c
+++ b/apps/plugins/doom/i_video.c
@@ -16,7 +16,10 @@
  * GNU General Public License for more details.
  *
  * $Log$
- * Revision 1.19  2006/08/03 20:17:22  bagder
+ * Revision 1.20  2006/08/07 01:46:41  amiconn
+ * Grayscale library ported to the grayscale iPods, first version. Added C reference versions of gray_update_rect() for both horizontal and vertical pixel packing. gray_update_rect() and gray_ub_gray_bitmap_part() not yet assembler optimised. Grayscale screendump doesn't work yet. * Fixed button assignments for iPod in grayscale.c
+ *
+ * Revision 1.19  2006-08-03 20:17:22  bagder
  * Barry Wardell's keymappings for H10
  *
  * Revision 1.18  2006-08-02 00:21:59  amiconn
@@ -500,7 +503,7 @@ void I_InitGraphics(void)
 
 #ifndef HAVE_LCD_COLOR
    gbuf=malloc(GRAYBUFSIZE);
-   gray_init(rb, gbuf, GRAYBUFSIZE, false, LCD_WIDTH, LCD_HEIGHT/8, 32, 
+   gray_init(rb, gbuf, GRAYBUFSIZE, false, LCD_WIDTH, LCD_HEIGHT, 32, 
              3<<7 /* 1.5 */, NULL);
    /* switch on grayscale overlay */
    gray_show(true);
diff --git a/apps/plugins/fire.c b/apps/plugins/fire.c
index 544b7d15c5..82770f3132 100644
--- a/apps/plugins/fire.c
+++ b/apps/plugins/fire.c
@@ -338,7 +338,7 @@ int main(void)
 #ifdef HAVE_LCD_COLOR
     shades = 256;
 #else
-    shades = gray_init(rb, gbuf, gbuf_size, false, LCD_WIDTH, LCD_HEIGHT/8,
+    shades = gray_init(rb, gbuf, gbuf_size, false, LCD_WIDTH, LCD_HEIGHT,
                        32, 2<<8, NULL) + 1;
     if(shades <= 1)
     {
diff --git a/apps/plugins/grayscale.c b/apps/plugins/grayscale.c
index ad3ee4f144..61231af39b 100644
--- a/apps/plugins/grayscale.c
+++ b/apps/plugins/grayscale.c
@@ -21,7 +21,7 @@
 
 #include "plugin.h"
 
-#if defined(HAVE_LCD_BITMAP) && (LCD_DEPTH < 4) && (CONFIG_LCD != LCD_IPOD2BPP)
+#if defined(HAVE_LCD_BITMAP) && (LCD_DEPTH < 4)
 #include "gray.h"
 
 PLUGIN_HEADER
@@ -54,12 +54,12 @@ PLUGIN_HEADER
 #define GRAYSCALE_RC_OFF BUTTON_RC_STOP
 
 #elif (CONFIG_KEYPAD == IPOD_3G_PAD) || (CONFIG_KEYPAD == IPOD_4G_PAD)
-#define GRAYSCALE_SHIFT (BUTTON_SELECT | BUTTON_REL)
-#define GRAYSCALE_UP BUTTON_MENU
-#define GRAYSCALE_DOWN BUTTON_PLAY
+#define GRAYSCALE_SHIFT BUTTON_SELECT
+#define GRAYSCALE_UP BUTTON_SCROLL_BACK
+#define GRAYSCALE_DOWN BUTTON_SCROLL_FWD
 #define GRAYSCALE_LEFT BUTTON_LEFT
 #define GRAYSCALE_RIGHT BUTTON_RIGHT
-#define GRAYSCALE_OFF (BUTTON_SELECT | BUTTON_MENU)
+#define GRAYSCALE_OFF BUTTON_MENU
 #endif
 
 #define GFX_HEIGHT (LCD_HEIGHT-8)
@@ -182,11 +182,11 @@ int main(void)
        32 bitplanes for 33 shades of grey.
        H1x0: 160 pixels wide, 30 rows (120 pixels) high, (try to) reserve
        32 bitplanes for 33 shades of grey. */
-    shades = gray_init(rb, gbuf, gbuf_size, true, LCD_WIDTH, GFX_HEIGHT/8,
+    shades = gray_init(rb, gbuf, gbuf_size, true, LCD_WIDTH, GFX_HEIGHT,
                        32, 0, NULL) + 1;
 
     /* place greyscale overlay 1 row down */
-    gray_set_position(0, 1);
+    gray_set_position(0, 8);
 
     rb->snprintf(pbuf, sizeof(pbuf), "Shades: %d", shades);
     rb->lcd_puts(0, 0, pbuf);
diff --git a/apps/plugins/jpeg.c b/apps/plugins/jpeg.c
index b3a0b463b7..5ecd394d26 100644
--- a/apps/plugins/jpeg.c
+++ b/apps/plugins/jpeg.c
@@ -2783,7 +2783,7 @@ enum plugin_status plugin_start(struct plugin_api* api, void* parameter)
 
 #ifdef USEGSLIB
     /* initialize the grayscale buffer: 32 bitplanes for 33 shades of gray. */
-    grayscales = gray_init(rb, buf, buf_size, false, LCD_WIDTH, LCD_HEIGHT/8,
+    grayscales = gray_init(rb, buf, buf_size, false, LCD_WIDTH, LCD_HEIGHT,
                            32, 2<<8, &graysize) + 1;
     buf += graysize;
     buf_size -= graysize;
diff --git a/apps/plugins/lib/SOURCES b/apps/plugins/lib/SOURCES
index dba800938d..6002934ac0 100644
--- a/apps/plugins/lib/SOURCES
+++ b/apps/plugins/lib/SOURCES
@@ -1,7 +1,7 @@
 configfile.c
 playback_control.c
 rgb_hsv.c
-#if defined(HAVE_LCD_BITMAP) && (LCD_DEPTH < 4) && (CONFIG_LCD != LCD_IPOD2BPP)
+#if defined(HAVE_LCD_BITMAP) && (LCD_DEPTH < 4)
 gray_core.c
 gray_draw.c
 gray_parm.c
diff --git a/apps/plugins/lib/gray.h b/apps/plugins/lib/gray.h
index 398462fd81..460aa83a82 100644
--- a/apps/plugins/lib/gray.h
+++ b/apps/plugins/lib/gray.h
@@ -27,7 +27,7 @@
 
 #include "plugin.h"
 
-#ifdef HAVE_LCD_BITMAP /* and also not for the Player */
+#ifdef HAVE_LCD_BITMAP /* not for the Player */
 
 #define GRAY_BRIGHTNESS(y) (y)
 
@@ -109,9 +109,6 @@ void gray_ub_scroll_down(int count);
 
 /*** Internal stuff ***/
 
-#define _PBLOCK_EXP 3
-#define _PBLOCK (1 << _PBLOCK_EXP)
-
 /* flag definitions */
 #define _GRAY_RUNNING          0x0001  /* greyscale overlay is running */
 #define _GRAY_DEFERRED_UPDATE  0x0002  /* lcd_update() requested */
@@ -124,10 +121,16 @@ void gray_ub_scroll_down(int count);
 struct _gray_info
 {
     int x;
-    int by;         /* 8-pixel units */
+    int y;
     int width;
     int height;
+#if LCD_PIXELFORMAT == HORIZONTAL_PACKING
+    int bx;         /* 8-pixel units */
+    int bwidth;     /* 8-pixel units */
+#else /* vertical packing */
+    int by;         /* 8-pixel units */
     int bheight;    /* 8-pixel units */
+#endif
     int depth;      /* number_of_bitplanes  = (number_of_grayscales - 1) */
     unsigned long flags;       /* various flags, see #defines */
 #ifndef SIMULATOR
diff --git a/apps/plugins/lib/gray_core.c b/apps/plugins/lib/gray_core.c
index a9a3e7698f..679d52130d 100644
--- a/apps/plugins/lib/gray_core.c
+++ b/apps/plugins/lib/gray_core.c
@@ -11,7 +11,8 @@
 * Core & miscellaneous functions
 *
 * This is a generic framework to display up to 33 shades of grey
-* on low-depth bitmap LCDs (Archos b&w, Iriver 4-grey) within plugins.
+* on low-depth bitmap LCDs (Archos b&w, Iriver 4-grey, iPod 4-grey)
+* within plugins.
 *
 * Copyright (C) 2004-2006 Jens Arnold
 *
@@ -28,9 +29,14 @@
 #ifdef HAVE_LCD_BITMAP
 #include "gray.h"
 
+#ifdef CPU_PP
+#define NEED_BOOST
+#endif
+
 /* Global variables */
 struct plugin_api *_gray_rb = NULL; /* global api struct pointer */
 struct _gray_info _gray_info;       /* global info structure */
+
 #ifndef SIMULATOR
 short _gray_random_buffer;          /* buffer for random number generator */
 
@@ -106,6 +112,43 @@ static const unsigned char lcdlinear[256] = {
     227, 228, 230, 232, 233, 235, 237, 239,
     241, 243, 245, 247, 249, 251, 253, 255
 };
+#elif (CONFIG_LCD == LCD_IPOD2BPP) || (CONFIG_LCD == LCD_IPODMINI)
+/* measured and interpolated curve for mini LCD */
+/* TODO: verify this curve on the fullsize greyscale LCD */
+static const unsigned char lcdlinear[256] = {
+      0,   3,   6,   8,  11,  14,  17,  19,
+     22,  24,  27,  29,  32,  34,  36,  38,
+     40,  42,  44,  45,  47,  48,  50,  51,
+     52,  54,  55,  56,  57,  58,  58,  59,
+     60,  61,  62,  62,  63,  64,  64,  65,
+     66,  66,  67,  67,  68,  68,  69,  69,
+     70,  70,  70,  71,  71,  71,  72,  72,
+     73,  73,  73,  74,  74,  74,  74,  75,
+     75,  75,  76,  76,  76,  77,  77,  77,
+     78,  78,  78,  79,  79,  79,  80,  80,
+     80,  80,  81,  81,  81,  82,  82,  82,
+     83,  83,  83,  84,  84,  84,  85,  85,
+     85,  85,  86,  86,  86,  87,  87,  87,
+     87,  88,  88,  88,  89,  89,  89,  89,
+     90,  90,  90,  91,  91,  91,  92,  92,
+     92,  93,  93,  93,  94,  94,  94,  95,
+     95,  96,  96,  96,  97,  97,  98,  98,
+     99,  99,  99, 100, 100, 101, 101, 102,
+    102, 103, 103, 104, 104, 105, 105, 106,
+    106, 107, 107, 108, 108, 109, 109, 110,
+    110, 111, 111, 112, 113, 113, 114, 114,
+    115, 115, 116, 117, 117, 118, 118, 119,
+    120, 120, 121, 122, 122, 123, 124, 124,
+    125, 126, 126, 127, 128, 128, 129, 130,
+    131, 131, 132, 133, 134, 134, 135, 136,
+    137, 138, 139, 140, 141, 142, 143, 144,
+    145, 146, 147, 148, 149, 150, 152, 153,
+    154, 156, 157, 159, 160, 162, 163, 165,
+    167, 168, 170, 172, 174, 176, 178, 180,
+    182, 184, 187, 189, 192, 194, 197, 200,
+    203, 206, 209, 212, 215, 219, 222, 226,
+    229, 233, 236, 240, 244, 248, 251, 255
+};
 #endif
 #else /* SIMULATOR */
 /* undo a (generic) PC display gamma of 2.0 to simulate target behaviour */
@@ -147,20 +190,22 @@ static const unsigned char lcdlinear[256] = {
 
 /* Prototypes */
 static inline void _deferred_update(void) __attribute__ ((always_inline));
+static int exp_s16p16(int x);
+static int log_s16p16(int x);
+static void gray_screendump_hook(int fd);
 #ifdef SIMULATOR
 static unsigned long _gray_get_pixel(int x, int y);
 #else
 static void _timer_isr(void);
 #endif
-static void gray_screendump_hook(int fd);
 
 /* Update LCD areas not covered by the greyscale overlay */
 static inline void _deferred_update(void)
 {
     int x1 = MAX(_gray_info.x, 0);
     int x2 = MIN(_gray_info.x + _gray_info.width, LCD_WIDTH);
-    int y1 = MAX(_gray_info.by << _PBLOCK_EXP, 0);
-    int y2 = MIN((_gray_info.by << _PBLOCK_EXP) + _gray_info.height, LCD_HEIGHT);
+    int y1 = MAX(_gray_info.y, 0);
+    int y2 = MIN(_gray_info.y + _gray_info.height, LCD_HEIGHT);
 
     if (y1 > 0)  /* refresh part above overlay, full width */
         _gray_rb->lcd_update_rect(0, 0, LCD_WIDTH, y1);
@@ -179,9 +224,15 @@ static inline void _deferred_update(void)
 /* Timer interrupt handler: display next bitplane */
 static void _timer_isr(void)
 {
+#if LCD_PIXELFORMAT == HORIZONTAL_PACKING
+    _gray_rb->lcd_blit(_gray_info.plane_data + MULU16(_gray_info.plane_size,
+                       _gray_info.cur_plane), _gray_info.bx, _gray_info.y,
+                       _gray_info.bwidth, _gray_info.height, _gray_info.bwidth);
+#else
     _gray_rb->lcd_blit(_gray_info.plane_data + MULU16(_gray_info.plane_size,
                        _gray_info.cur_plane), _gray_info.x, _gray_info.by,
                        _gray_info.width, _gray_info.bheight, _gray_info.width);
+#endif
 
     if (++_gray_info.cur_plane >= _gray_info.depth)
         _gray_info.cur_plane = 0;
@@ -200,7 +251,7 @@ static int exp_s16p16(int x)
     int t;
     int y = 0x00010000;
     
-        if (x < 0) x += 0xb1721,            y >>= 16;
+    if (x < 0) x += 0xb1721,            y >>= 16;
     t = x - 0x58b91; if (t >= 0) x = t, y <<= 8;
     t = x - 0x2c5c8; if (t >= 0) x = t, y <<= 4;
     t = x - 0x162e4; if (t >= 0) x = t, y <<= 2;
@@ -218,7 +269,7 @@ static int exp_s16p16(int x)
 }
 
 /* fixed point log() */
-int log_s16p16(int x)
+static int log_s16p16(int x)
 {
     int t;
     int y = 0xa65af;
@@ -252,7 +303,9 @@ int log_s16p16(int x)
                memory. Unbuffered operation provides only a subset of
                drawing functions. (only gray_bitmap drawing and scrolling)
    width     = width in pixels  (1..LCD_WIDTH)
-   bheight   = height in LCD pixel-block units (8 pixels) (1..LCD_HEIGHT/8)
+   height    = height in pixels (1..LCD_HEIGHT)
+               Note that depending on the target LCD, either height or
+               width are rounded up to a multiple of 8.
    depth     = number of bitplanes to use (1..32).
    gamma     = gamma value as s8p8 fixed point. gamma <= 0 means no
                correction at all, i.e. no LCD linearisation as well.
@@ -277,20 +330,21 @@ int log_s16p16(int x)
    used. The total memory needed can be calculated as follows:
  total_mem =
      shades * sizeof(long)               (bitpatterns)
-   + (width * bheight) * depth           (bitplane data)
+   + [horizontal_packing] ?              (bitplane data)
+       ((width + 7) / 8) * height * depth : width * ((height + 7) / 8) * depth
    + buffered ?                          (chunky front- & backbuffer)
-       (width * bheight * 8 * 2) : 0
+       (width * height * 2) : 0
    + 0..3                                (longword alignment)
-   
+
    The function tries to be as authentic as possible regarding memory usage on
    the simulator, even if it doesn't use all of the allocated memory. There's
    one situation where it will consume more memory on the sim than on the
    target: if you're allocating a low depth (< 8) without buffering. */
 int gray_init(struct plugin_api* newrb, unsigned char *gbuf, long gbuf_size,
-              bool buffered, int width, int bheight, int depth, int gamma,
+              bool buffered, int width, int height, int depth, int gamma,
               long *buf_taken)
 {
-    int possible_depth, i;
+    int possible_depth, bdim, i;
     long plane_size, buftaken;
     unsigned data;
 #ifndef SIMULATOR
@@ -300,10 +354,18 @@ int gray_init(struct plugin_api* newrb, unsigned char *gbuf, long gbuf_size,
     _gray_rb = newrb;
 
     if ((unsigned) width > LCD_WIDTH
-        || (unsigned) bheight > (LCD_HEIGHT/_PBLOCK)
+        || (unsigned) height > LCD_HEIGHT
         || depth < 1)
         return 0;
 
+#if LCD_PIXELFORMAT == HORIZONTAL_PACKING
+    bdim = (width + 7) >> 3;
+    width = bdim << 3;
+#else
+    bdim = (height + 7) >> 3;
+    height = bdim << 3;
+#endif
+
     /* the buffer has to be long aligned */
     buftaken = (-(long)gbuf) & 3;
     gbuf += buftaken;
@@ -311,7 +373,7 @@ int gray_init(struct plugin_api* newrb, unsigned char *gbuf, long gbuf_size,
     /* chunky front- & backbuffer */
     if (buffered)  
     {
-        plane_size = MULU16(width, bheight << _PBLOCK_EXP);
+        plane_size = MULU16(width, height);
         buftaken += 2 * plane_size;
         if (buftaken > gbuf_size)
             return 0;
@@ -324,7 +386,11 @@ int gray_init(struct plugin_api* newrb, unsigned char *gbuf, long gbuf_size,
         gbuf += plane_size;
     }
 
-    plane_size = MULU16(width, bheight);
+#if LCD_PIXELFORMAT == HORIZONTAL_PACKING
+    plane_size = MULU16(bdim, height);
+#else
+    plane_size = MULU16(width, bdim);
+#endif
     possible_depth = (gbuf_size - buftaken - sizeof(long))
                      / (plane_size + sizeof(long));
 
@@ -339,7 +405,7 @@ int gray_init(struct plugin_api* newrb, unsigned char *gbuf, long gbuf_size,
     {
         long orig_size = depth * plane_size + (depth + 1) * sizeof(long);
         
-        plane_size = MULU16(width, bheight << _PBLOCK_EXP);
+        plane_size = MULU16(width, height);
         if (plane_size > orig_size)
         {
             buftaken += plane_size;
@@ -357,10 +423,16 @@ int gray_init(struct plugin_api* newrb, unsigned char *gbuf, long gbuf_size,
         buftaken += depth * plane_size + (depth + 1) * sizeof(long);
 
     _gray_info.x = 0;
-    _gray_info.by = 0;
+    _gray_info.y = 0;
     _gray_info.width = width;
-    _gray_info.height = bheight << _PBLOCK_EXP;
-    _gray_info.bheight = bheight;
+    _gray_info.height = height;
+#if LCD_PIXELFORMAT == HORIZONTAL_PACKING
+    _gray_info.bx = 0;
+    _gray_info.bwidth = bdim;
+#else
+    _gray_info.by = 0;
+    _gray_info.bheight = bdim;
+#endif
     _gray_info.depth = depth;
     _gray_info.flags = 0;
 #ifndef SIMULATOR
@@ -463,13 +535,20 @@ void gray_show(bool enable)
         _gray_rb->sim_lcd_ex_init(_gray_info.depth + 1, _gray_get_pixel);
         gray_update();
 #else /* !SIMULATOR */
+#ifdef NEED_BOOST
+        _gray_rb->cpu_boost(true);
+#endif
 #if CONFIG_LCD == LCD_SSD1815
         _gray_rb->timer_register(1, NULL, TIMER_FREQ / 67, 1, _timer_isr);
 #elif CONFIG_LCD == LCD_S1D15E06
         _gray_rb->timer_register(1, NULL, TIMER_FREQ / 70, 1, _timer_isr);
+#elif CONFIG_LCD == LCD_IPOD2BPP
+        /* FIXME: verify value */
+        _gray_rb->timer_register(1, NULL, TIMER_FREQ / 40, 1, _timer_isr);
+#elif CONFIG_LCD == LCD_IPODMINI
+        _gray_rb->timer_register(1, NULL, TIMER_FREQ / 44, 1, _timer_isr);
 #elif CONFIG_LCD == LCD_IFP7XX
-        /* TODO: implement for iFP */
-        (void)_timer_isr;
+        (void)_timer_isr;   /* TODO: implement for iFP */
 #endif /* CONFIG_LCD */
 #endif /* !SIMULATOR */
         _gray_rb->screen_dump_set_hook(gray_screendump_hook);
@@ -480,6 +559,9 @@ void gray_show(bool enable)
         _gray_rb->sim_lcd_ex_init(0, NULL);
 #else
         _gray_rb->timer_unregister();
+#ifdef NEED_BOOST
+        _gray_rb->cpu_boost(false);
+#endif
 #endif
         _gray_info.flags &= ~_GRAY_RUNNING;
         _gray_rb->screen_dump_set_hook(NULL);
@@ -492,9 +574,15 @@ void gray_show(bool enable)
    Note that x and y are in LCD coordinates, not graybuffer coordinates! */
 static unsigned long _gray_get_pixel(int x, int y)
 {
+#if LCD_PIXELFORMAT == HORIZONTAL_PACKING
+    return _gray_info.cur_buffer[MULU16(y - _gray_info.y, _gray_info.width)
+                                 + x - _gray_info.x]
+           + (1 << LCD_DEPTH);
+#else
     return _gray_info.cur_buffer[MULU16(x - _gray_info.x, _gray_info.height)
-                                 + y - (_gray_info.by << _PBLOCK_EXP)]
+                                 + y - _gray_info.y]
            + (1 << LCD_DEPTH);
+#endif
 }
 
 /* Update a rectangular area of the greyscale overlay */
@@ -506,7 +594,7 @@ void gray_update_rect(int x, int y, int width, int height)
         height = _gray_info.height - y;
         
     x += _gray_info.x;
-    y += _gray_info.by << _PBLOCK_EXP;
+    y += _gray_info.y;
 
     if (x + width > LCD_WIDTH)
         width = LCD_WIDTH - x;
@@ -517,6 +605,148 @@ void gray_update_rect(int x, int y, int width, int height)
 }
 
 #else /* !SIMULATOR */
+
+#if LCD_PIXELFORMAT == HORIZONTAL_PACKING
+
+/* Update a rectangular area of the greyscale overlay */
+void gray_update_rect(int x, int y, int width, int height)
+{
+    int xmax, bwidth;
+    long srcofs;
+    unsigned char *dst;
+
+    if ((width <= 0) || (height <= 0))
+        return; /* nothing to do */
+
+    /* The X coordinates have to work on whole pixel block columns */
+    xmax = (x + width - 1) >> 3;
+    x >>= 3;
+
+    if (y + height > _gray_info.height)
+        height = _gray_info.height - y;
+    if (xmax >= _gray_info.bwidth)
+        xmax = _gray_info.bwidth - 1;
+    bwidth = xmax - x + 1;
+        
+    srcofs = MULU16(_gray_info.width, y) + (x << 3);
+    dst = _gray_info.plane_data + MULU16(_gray_info.bwidth, y) + x;
+    
+    /* Copy specified rectangle bitmap to hardware */
+    for (; height > 0; height--)
+    {
+        long srcofs_row = srcofs;
+        unsigned char *dst_row = dst;
+        unsigned char *dst_end = dst_row + bwidth;
+
+        do
+        {
+            unsigned long pat_stack[8];
+            unsigned long *pat_ptr;
+            unsigned char *cbuf, *bbuf;
+            unsigned change;
+
+            cbuf = _gray_info.cur_buffer + srcofs_row;
+            bbuf = _gray_info.back_buffer + srcofs_row;
+
+#if 0 /* CPU specific asm versions will go here */
+#else /* C version, for reference*/
+            (void)pat_ptr;
+            /* check whether anything changed in the 8-pixel block */
+            change  = *(uint32_t *)cbuf ^ *(uint32_t *)bbuf;
+            cbuf += sizeof(uint32_t);
+            bbuf += sizeof(uint32_t);
+            change |= *(uint32_t *)cbuf ^ *(uint32_t *)bbuf;
+
+            if (change != 0)
+            {
+                unsigned char *addr, *end;
+                unsigned mask = 0;
+                unsigned test = 1;
+                int i;
+
+                cbuf = _gray_info.cur_buffer + srcofs_row;
+                bbuf = _gray_info.back_buffer + srcofs_row;
+
+                /* precalculate the bit patterns with random shifts
+                 * for all 8 pixels and put them on an extra "stack" */
+                for (i = 7; i >= 0; i--)
+                {
+                    unsigned pat = 0;
+                    unsigned char cur = *cbuf++;
+                    unsigned char back = *bbuf;
+                    
+                    *bbuf++ = cur;
+
+                    mask <<= 1;
+                    if (cur != back)
+                    {
+                        int shift;
+
+                        pat = _gray_info.bitpattern[cur];
+
+                        /* shift pattern pseudo-random, simple & fast PRNG */
+                        _gray_random_buffer = 75 * _gray_random_buffer + 74;
+                        shift = (_gray_random_buffer >> 8) & _gray_info.randmask;
+                        if (shift >= _gray_info.depth)
+                            shift -= _gray_info.depth;
+                            
+                        pat = (pat << shift) | (pat >> (_gray_info.depth - shift));
+                        
+                        mask |= 1;
+                    }
+                    pat_stack[i] = pat;
+                }
+
+                addr = dst_row;
+                end = addr + MULU16(_gray_info.depth, _gray_info.plane_size);
+
+                /* set the bits for all 8 pixels in all bytes according to the
+                 * precalculated patterns on the pattern stack */
+                mask = (~mask & 0xff);
+                if (mask == 0)
+                {
+                    do
+                    {
+                        unsigned data = 0;
+
+                        for (i = 7; i >= 0; i--)
+                            data = (data << 1) | ((pat_stack[i] & test) ? 1 : 0);
+                        
+                        *addr = data;
+                        addr += _gray_info.plane_size;
+                        test <<= 1;
+                    }
+                    while (addr < end);
+                }
+                else
+                {
+                    do
+                    {
+                        unsigned data = 0;
+
+                        for (i = 7; i >= 0; i--)
+                            data = (data << 1) | ((pat_stack[i] & test) ? 1 : 0);
+                        
+                        *addr = (*addr & mask) | data;
+                        addr += _gray_info.plane_size;
+                        test <<= 1;
+                    }
+                    while (addr < end);
+                }
+
+            }
+#endif /* CONFIG_CPU */
+            srcofs_row += 8;
+            dst_row++;
+        }
+        while (dst_row < dst_end);
+        
+        srcofs += _gray_info.width;
+        dst += _gray_info.bwidth;
+    }
+}
+#else /* LCD_PIXELFORMAT == VERTICAL_PACKING */
+
 /* Update a rectangular area of the greyscale overlay */
 void gray_update_rect(int x, int y, int width, int height)
 {
@@ -524,22 +754,22 @@ void gray_update_rect(int x, int y, int width, int height)
     long srcofs;
     unsigned char *dst;
 
-    if (width <= 0)
+    if ((width <= 0) || (height <= 0))
         return; /* nothing to do */
 
     /* The Y coordinates have to work on whole pixel block rows */
-    ymax = (y + height - 1) >> _PBLOCK_EXP;
-    y >>= _PBLOCK_EXP;
+    ymax = (y + height - 1) >> 3;
+    y >>= 3;
 
     if (x + width > _gray_info.width)
         width = _gray_info.width - x;
     if (ymax >= _gray_info.bheight)
         ymax = _gray_info.bheight - 1;
         
-    srcofs = (y << _PBLOCK_EXP) + MULU16(_gray_info.height, x);
+    srcofs = (y << 3) + MULU16(_gray_info.height, x);
     dst = _gray_info.plane_data + MULU16(_gray_info.width, y) + x;
     
-    /* Copy specified rectange bitmap to hardware */
+    /* Copy specified rectangle bitmap to hardware */
     for (; y <= ymax; y++)
     {
         long srcofs_row = srcofs;
@@ -548,7 +778,6 @@ void gray_update_rect(int x, int y, int width, int height)
 
         do
         {
-#if (CONFIG_CPU == SH7034) && (LCD_DEPTH == 1)
             unsigned long pat_stack[8];
             unsigned long *pat_ptr;
             unsigned char *cbuf, *bbuf;
@@ -557,6 +786,7 @@ void gray_update_rect(int x, int y, int width, int height)
             cbuf = _gray_info.cur_buffer + srcofs_row;
             bbuf = _gray_info.back_buffer + srcofs_row;
 
+#if CONFIG_CPU == SH7034
             asm volatile (
                 "mov.l   @%[cbuf]+,r1    \n"
                 "mov.l   @%[bbuf]+,r2    \n"
@@ -738,15 +968,7 @@ void gray_update_rect(int x, int y, int width, int height)
                     "r0", "r1", "r2", "r3", "r6", "r7", "r8", "r9", "r10"
                 );
             }
-#elif defined(CPU_COLDFIRE) && (LCD_DEPTH == 2)
-            unsigned long pat_stack[8];
-            unsigned long *pat_ptr;
-            unsigned char *cbuf, *bbuf;
-            unsigned change;
-
-            cbuf = _gray_info.cur_buffer + srcofs_row;
-            bbuf = _gray_info.back_buffer + srcofs_row;
-
+#elif defined(CPU_COLDFIRE)
             asm volatile (
                 "move.l  (%[cbuf])+,%%d0 \n"
                 "move.l  (%[bbuf])+,%%d1 \n"
@@ -924,16 +1146,105 @@ void gray_update_rect(int x, int y, int width, int height)
                     "d0", "d1", "d2", "d3", "d4", "d5", "d6", "a0", "a1"
                 );
             }
-#endif /* CONFIG_CPU, LCD_DEPTH */
+#else /* C version, for reference*/
+#warning C version of gray_update_rect() used
+            (void)pat_ptr;
+            /* check whether anything changed in the 8-pixel block */
+            change  = *(uint32_t *)cbuf ^ *(uint32_t *)bbuf;
+            cbuf += sizeof(uint32_t);
+            bbuf += sizeof(uint32_t);
+            change |= *(uint32_t *)cbuf ^ *(uint32_t *)bbuf;
+            
+            if (change != 0)
+            {
+                unsigned char *addr, *end;
+                unsigned mask = 0;
+                unsigned test = 1;
+                int i;
+
+                cbuf = _gray_info.cur_buffer + srcofs_row;
+                bbuf = _gray_info.back_buffer + srcofs_row;
+
+                /* precalculate the bit patterns with random shifts
+                 * for all 8 pixels and put them on an extra "stack" */
+                for (i = 0; i < 8; i++)
+                {
+                    unsigned pat = 0;
+                    unsigned char cur = *cbuf++;
+                    unsigned char back = *bbuf;
+                    
+                    *bbuf++ = cur;
+
+                    if (cur != back)
+                    {
+                        int shift;
+
+                        pat = _gray_info.bitpattern[cur];
+
+                        /* shift pattern pseudo-random, simple & fast PRNG */
+                        _gray_random_buffer = 75 * _gray_random_buffer + 74;
+                        shift = (_gray_random_buffer >> 8) & _gray_info.randmask;
+                        if (shift >= _gray_info.depth)
+                            shift -= _gray_info.depth;
+                            
+                        pat = (pat << shift) | (pat >> (_gray_info.depth - shift));
+                        
+                        mask |= 0x100;
+                    }
+                    mask >>= 1;
+                    pat_stack[i] = pat;
+                }
+
+                addr = dst_row;
+                end = addr + MULU16(_gray_info.depth, _gray_info.plane_size);
+
+                /* set the bits for all 8 pixels in all bytes according to the
+                 * precalculated patterns on the pattern stack */
+                mask = (~mask & 0xff);
+                if (mask == 0)
+                {
+                    do
+                    {
+                        unsigned data = 0;
+
+                        for (i = 7; i >= 0; i--)
+                            data = (data << 1) | ((pat_stack[i] & test) ? 1 : 0);
+                        
+                        *addr = data;
+                        addr += _gray_info.plane_size;
+                        test <<= 1;
+                    }
+                    while (addr < end);
+                }
+                else
+                {
+                    do
+                    {
+                        unsigned data = 0;
+
+                        for (i = 7; i >= 0; i--)
+                            data = (data << 1) | ((pat_stack[i] & test) ? 1 : 0);
+                        
+                        *addr = (*addr & mask) | data;
+                        addr += _gray_info.plane_size;
+                        test <<= 1;
+                    }
+                    while (addr < end);
+                }
+
+            }
+#endif /* CONFIG_CPU */
             srcofs_row += _gray_info.height;
             dst_row++;
         }
         while (dst_row < dst_end);
         
-        srcofs += _PBLOCK;
+        srcofs += 8;
         dst += _gray_info.width;
     }
 }
+#endif /* LCD_PIXELFORMAT */
+
 #endif /* !SIMULATOR */
 
 /* Update the whole greyscale overlay */
@@ -1047,10 +1358,10 @@ static void gray_screendump_hook(int fd)
     {
         _gray_rb->memset(linebuf, 0, BMP_LINESIZE);
 
-        gy = y - (_gray_info.by << _PBLOCK_EXP);
+        gy = y - _gray_info.y;
 #if LCD_DEPTH == 1
-        mask = 1 << (y & (_PBLOCK-1));
-        by = y >> _PBLOCK_EXP;
+        mask = 1 << (y & 7);
+        by = y >> 3;
         lcdptr = _gray_rb->lcd_framebuffer + MULU16(LCD_WIDTH, by);
 
         if ((unsigned) gy < (unsigned) _gray_info.height)
@@ -1058,7 +1369,7 @@ static void gray_screendump_hook(int fd)
             /* line contains greyscale (and maybe b&w) graphics */
 #ifndef SIMULATOR
             unsigned char *grayptr = _gray_info.plane_data 
-                                   + MULU16(_gray_info.width, gy >> _PBLOCK_EXP);
+                                   + MULU16(_gray_info.width, gy >> 3);
 #endif
 
             for (x = 0; x < LCD_WIDTH; x++)
@@ -1108,8 +1419,8 @@ static void gray_screendump_hook(int fd)
             (void)mask;
 #else
             unsigned char *grayptr = _gray_info.plane_data
-                                   + MULU16(_gray_info.width, gy >> _PBLOCK_EXP);
-            mask = 1 << (gy & (_PBLOCK-1));
+                                   + MULU16(_gray_info.width, gy >> 3);
+            mask = 1 << (gy & 7);
 #endif
 
             for (x = 0; x < LCD_WIDTH; x++)
diff --git a/apps/plugins/lib/gray_draw.c b/apps/plugins/lib/gray_draw.c
index 7e1197bd4b..396046d1e6 100644
--- a/apps/plugins/lib/gray_draw.c
+++ b/apps/plugins/lib/gray_draw.c
@@ -11,7 +11,8 @@
 * Drawing functions
 *
 * This is a generic framework to display up to 33 shades of grey
-* on low-depth bitmap LCDs (Archos b&w, Iriver 4-grey) within plugins.
+* on low-depth bitmap LCDs (Archos b&w, Iriver 4-grey, iPod 4-grey)
+* within plugins.
 *
 * Copyright (C) 2004-2006 Jens Arnold
 *
@@ -72,8 +73,13 @@ void gray_drawpixel(int x, int y)
 {
     if (((unsigned)x < (unsigned)_gray_info.width)
         && ((unsigned)y < (unsigned)_gray_info.height))
+#if LCD_PIXELFORMAT == HORIZONTAL_PACKING
+        _gray_pixelfuncs[_gray_info.drawmode](&_gray_info.cur_buffer[MULU16(y,
+                                               _gray_info.width) + x]);
+#else
         _gray_pixelfuncs[_gray_info.drawmode](&_gray_info.cur_buffer[MULU16(x,
                                                _gray_info.height) + y]);
+#endif
 }
 
 /* Draw a line */
@@ -131,7 +137,11 @@ void gray_drawline(int x1, int y1, int x2, int y2)
     {
         if (((unsigned)x < (unsigned)_gray_info.width) 
             && ((unsigned)y < (unsigned)_gray_info.height))
+#if LCD_PIXELFORMAT == HORIZONTAL_PACKING
+            pfunc(&_gray_info.cur_buffer[MULU16(y, _gray_info.width) + x]);
+#else
             pfunc(&_gray_info.cur_buffer[MULU16(x, _gray_info.height) + y]);
+#endif
 
         if (d < 0)
         {
@@ -148,6 +158,175 @@ void gray_drawline(int x1, int y1, int x2, int y2)
     }
 }
 
+#if LCD_PIXELFORMAT == HORIZONTAL_PACKING
+
+/* Draw a horizontal line (optimised) */
+void gray_hline(int x1, int x2, int y)
+{
+    int x;
+    int bits = 0;
+    unsigned char *dst;
+    bool fillopt = false;
+    void (*pfunc)(unsigned char *address);
+    
+    /* direction flip */
+    if (x2 < x1)
+    {
+        x = x1;
+        x1 = x2;
+        x2 = x;
+    }
+
+    /* nothing to draw? */
+    if (((unsigned)y >= (unsigned)_gray_info.height) 
+        || (x1 >= _gray_info.width) || (x2 < 0))
+        return;  
+    
+    /* clipping */
+    if (x1 < 0)
+        x1 = 0;
+    if (x2 >= _gray_info.width)
+        x2 = _gray_info.width - 1;
+        
+    if (_gray_info.drawmode & DRMODE_INVERSEVID)
+    {
+        if (_gray_info.drawmode & DRMODE_BG)
+        {
+            fillopt = true;
+            bits = _gray_info.bg_index;
+        }
+    }
+    else
+    {
+        if (_gray_info.drawmode & DRMODE_FG)
+        {
+            fillopt = true;
+            bits = _gray_info.fg_index;
+        }
+    }
+    pfunc = _gray_pixelfuncs[_gray_info.drawmode];
+    dst   = &_gray_info.cur_buffer[MULU16(y, _gray_info.width) + x1];
+
+    if (fillopt)
+        _gray_rb->memset(dst, bits, x2 - x1 + 1);
+    else
+    {
+        unsigned char *dst_end = dst + x2 - x1;
+        do
+            pfunc(dst++);
+        while (dst <= dst_end);
+    }
+}
+
+/* Draw a vertical line (optimised) */
+void gray_vline(int x, int y1, int y2)
+{
+    int y;
+    unsigned char *dst, *dst_end;
+    void (*pfunc)(unsigned char *address);
+    
+    /* direction flip */
+    if (y2 < y1)
+    {
+        y = y1;
+        y1 = y2;
+        y2 = y;
+    }
+
+    /* nothing to draw? */
+    if (((unsigned)x >= (unsigned)_gray_info.width) 
+        || (y1 >= _gray_info.height) || (y2 < 0))
+        return;  
+    
+    /* clipping */
+    if (y1 < 0)
+        y1 = 0;
+    if (y2 >= _gray_info.height)
+        y2 = _gray_info.height - 1;
+
+    pfunc = _gray_pixelfuncs[_gray_info.drawmode];
+    dst   = &_gray_info.cur_buffer[MULU16(y1, _gray_info.width) + x];
+
+    dst_end = dst + MULU16(y2 - y1, _gray_info.width);
+    do
+    {
+        pfunc(dst);
+        dst += _gray_info.width;
+    }
+    while (dst <= dst_end);
+}
+
+/* Draw a filled triangle */
+void gray_filltriangle(int x1, int y1, int x2, int y2, int x3, int y3)
+{
+    int x, y;
+    long fp_x1, fp_x2, fp_dx1, fp_dx2;
+
+    /* sort vertices by increasing y value */
+    if (y1 > y3)
+    {
+        if (y2 < y3)       /* y2 < y3 < y1 */
+        {
+            x = x1; x1 = x2; x2 = x3; x3 = x;
+            y = y1; y1 = y2; y2 = y3; y3 = y;
+        }
+        else if (y2 > y1)  /* y3 < y1 < y2 */
+        {
+            x = x1; x1 = x3; x3 = x2; x2 = x;
+            y = y1; y1 = y3; y3 = y2; y2 = y;
+        }
+        else               /* y3 <= y2 <= y1 */
+        {
+            x = x1; x1 = x3; x3 = x;
+            y = y1; y1 = y3; y3 = y;
+        }
+    }
+    else
+    {
+        if (y2 < y1)       /* y2 < y1 <= y3 */
+        {
+            x = x1; x1 = x2; x2 = x;
+            y = y1; y1 = y2; y2 = y;
+        }
+        else if (y2 > y3)  /* y1 <= y3 < y2 */
+        {
+            x = x2; x2 = x3; x3 = x;
+            y = y2; y2 = y3; y3 = y;
+        }
+        /* else already sorted */
+    }
+
+    if (y1 < y3)  /* draw */
+    {
+        fp_dx1 = ((x3 - x1) << 16) / (y3 - y1);
+        fp_x1  = (x1 << 16) + (1<<15) + (fp_dx1 >> 1);
+
+        if (y1 < y2)  /* first part */
+        {
+            fp_dx2 = ((x2 - x1) << 16) / (y2 - y1);
+            fp_x2  = (x1 << 16) + (1<<15) + (fp_dx2 >> 1);
+            for (y = y1; y < y2; y++)
+            {
+                gray_hline(fp_x1 >> 16, fp_x2 >> 16, y);
+                fp_x1 += fp_dx1;
+                fp_x2 += fp_dx2;
+            }
+        }
+        if (y2 < y3)  /* second part */
+        {
+            fp_dx2 = ((x3 - x2) << 16) / (y3 - y2);
+            fp_x2 = (x2 << 16) + (1<<15) + (fp_dx2 >> 1);
+            for (y = y2; y < y3; y++)
+            {
+                gray_hline(fp_x1 >> 16, fp_x2 >> 16, y);
+                fp_x1 += fp_dx1;
+                fp_x2 += fp_dx2;
+            }
+        }
+    }
+}
+#else /* LCD_PIXELFORMAT == VERTICAL_PACKING */
+
 /* Draw a horizontal line (optimised) */
 void gray_hline(int x1, int x2, int y)
 {
@@ -244,6 +423,77 @@ void gray_vline(int x, int y1, int y2)
     }
 }
 
+/* Draw a filled triangle */
+void gray_filltriangle(int x1, int y1, int x2, int y2, int x3, int y3)
+{
+    int x, y;
+    long fp_y1, fp_y2, fp_dy1, fp_dy2;
+
+    /* sort vertices by increasing x value */
+    if (x1 > x3)
+    {
+        if (x2 < x3)       /* x2 < x3 < x1 */
+        {
+            x = x1; x1 = x2; x2 = x3; x3 = x;
+            y = y1; y1 = y2; y2 = y3; y3 = y;
+        }
+        else if (x2 > x1)  /* x3 < x1 < x2 */
+        {
+            x = x1; x1 = x3; x3 = x2; x2 = x;
+            y = y1; y1 = y3; y3 = y2; y2 = y;
+        }
+        else               /* x3 <= x2 <= x1 */
+        {
+            x = x1; x1 = x3; x3 = x;
+            y = y1; y1 = y3; y3 = y;
+        }
+    }
+    else
+    {
+        if (x2 < x1)       /* x2 < x1 <= x3 */
+        {
+            x = x1; x1 = x2; x2 = x;
+            y = y1; y1 = y2; y2 = y;
+        }
+        else if (x2 > x3)  /* x1 <= x3 < x2 */
+        {
+            x = x2; x2 = x3; x3 = x;
+            y = y2; y2 = y3; y3 = y;
+        }
+        /* else already sorted */
+    }
+
+    if (x1 < x3)  /* draw */
+    {
+        fp_dy1 = ((y3 - y1) << 16) / (x3 - x1);
+        fp_y1  = (y1 << 16) + (1<<15) + (fp_dy1 >> 1);
+
+        if (x1 < x2)  /* first part */
+        {   
+            fp_dy2 = ((y2 - y1) << 16) / (x2 - x1);
+            fp_y2  = (y1 << 16) + (1<<15) + (fp_dy2 >> 1);
+            for (x = x1; x < x2; x++)
+            {
+                gray_vline(x, fp_y1 >> 16, fp_y2 >> 16);
+                fp_y1 += fp_dy1;
+                fp_y2 += fp_dy2;
+            }
+        }
+        if (x2 < x3)  /* second part */
+        {
+            fp_dy2 = ((y3 - y2) << 16) / (x3 - x2);
+            fp_y2 = (y2 << 16) + (1<<15) + (fp_dy2 >> 1);
+            for (x = x2; x < x3; x++)
+            {
+                gray_vline(x, fp_y1 >> 16, fp_y2 >> 16);
+                fp_y1 += fp_dy1;
+                fp_y2 += fp_dy2;
+            }
+        }
+    }
+}
+#endif /* LCD_PIXELFORMAT */
+
 /* Draw a rectangular box */
 void gray_drawrect(int x, int y, int width, int height)
 {
@@ -305,6 +555,27 @@ void gray_fillrect(int x, int y, int width, int height)
         }
     }
     pfunc = _gray_pixelfuncs[_gray_info.drawmode];
+#if LCD_PIXELFORMAT == HORIZONTAL_PACKING
+    dst   = &_gray_info.cur_buffer[MULU16(y, _gray_info.width) + x];
+    dst_end = dst + MULU16(height, _gray_info.width);
+
+    do
+    {
+        if (fillopt)
+            _gray_rb->memset(dst, bits, width);
+        else
+        {
+            unsigned char *dst_row = dst;
+            unsigned char *row_end = dst_row + width;
+            
+            do
+                pfunc(dst_row++);
+            while (dst_row < row_end);
+        }
+        dst += _gray_info.width;
+    }
+    while (dst < dst_end);
+#else
     dst   = &_gray_info.cur_buffer[MULU16(x, _gray_info.height) + y];
     dst_end = dst + MULU16(width, _gray_info.height);
 
@@ -324,76 +595,7 @@ void gray_fillrect(int x, int y, int width, int height)
         dst += _gray_info.height;
     }
     while (dst < dst_end);
-}
-
-/* Draw a filled triangle */
-void gray_filltriangle(int x1, int y1, int x2, int y2, int x3, int y3)
-{
-    int x, y;
-    long fp_y1, fp_y2, fp_dy1, fp_dy2;
-
-    /* sort vertices by increasing x value */
-    if (x1 > x3)
-    {
-        if (x2 < x3)       /* x2 < x3 < x1 */
-        {
-            x = x1; x1 = x2; x2 = x3; x3 = x;
-            y = y1; y1 = y2; y2 = y3; y3 = y;
-        }
-        else if (x2 > x1)  /* x3 < x1 < x2 */
-        {
-            x = x1; x1 = x3; x3 = x2; x2 = x;
-            y = y1; y1 = y3; y3 = y2; y2 = y;
-        }
-        else               /* x3 <= x2 <= x1 */
-        {
-            x = x1; x1 = x3; x3 = x;
-            y = y1; y1 = y3; y3 = y;
-        }
-    }
-    else
-    {
-        if (x2 < x1)       /* x2 < x1 <= x3 */
-        {
-            x = x1; x1 = x2; x2 = x;
-            y = y1; y1 = y2; y2 = y;
-        }
-        else if (x2 > x3)  /* x1 <= x3 < x2 */
-        {
-            x = x2; x2 = x3; x3 = x;
-            y = y2; y2 = y3; y3 = y;
-        }
-        /* else already sorted */
-    }
-
-    if (x1 < x3)  /* draw */
-    {
-        fp_dy1 = ((y3 - y1) << 16) / (x3 - x1);
-        fp_y1  = (y1 << 16) + (1<<15) + (fp_dy1 >> 1);
-
-        if (x1 < x2)  /* first part */
-        {   
-            fp_dy2 = ((y2 - y1) << 16) / (x2 - x1);
-            fp_y2  = (y1 << 16) + (1<<15) + (fp_dy2 >> 1);
-            for (x = x1; x < x2; x++)
-            {
-                gray_vline(x, fp_y1 >> 16, fp_y2 >> 16);
-                fp_y1 += fp_dy1;
-                fp_y2 += fp_dy2;
-            }
-        }
-        if (x2 < x3)  /* second part */
-        {
-            fp_dy2 = ((y3 - y2) << 16) / (x3 - x2);
-            fp_y2 = (y2 << 16) + (1<<15) + (fp_dy2 >> 1);
-            for (x = x2; x < x3; x++)
-            {
-                gray_vline(x, fp_y1 >> 16, fp_y2 >> 16);
-                fp_y1 += fp_dy1;
-                fp_y2 += fp_dy2;
-            }
-        }
-    }
+#endif
 }
 
 /* About Rockbox' internal monochrome bitmap format:
@@ -403,9 +605,7 @@ void gray_filltriangle(int x1, int y1, int x2, int y2, int x3, int y3)
  * vertically, LSB at top.
  * The bytes are stored in row-major order, with byte 0 being top left,
  * byte 1 2nd from left etc. The first row of bytes defines pixel rows
- * 0..7, the second row defines pixel row 8..15 etc.
- *
- * This is similar to the internal lcd hw format. */
+ * 0..7, the second row defines pixel row 8..15 etc. */
 
 /* Draw a partial monochrome bitmap */
 void gray_mono_bitmap_part(const unsigned char *src, int src_x, int src_y,
@@ -443,9 +643,41 @@ void gray_mono_bitmap_part(const unsigned char *src, int src_x, int src_y,
     src_y  &= 7;
     src_end = src + width;
 
-    dst    = &_gray_info.cur_buffer[MULU16(x, _gray_info.height) + y];
     fgfunc = _gray_pixelfuncs[_gray_info.drawmode];
     bgfunc = _gray_pixelfuncs[_gray_info.drawmode ^ DRMODE_INVERSEVID];
+#if LCD_PIXELFORMAT == HORIZONTAL_PACKING
+    dst    = &_gray_info.cur_buffer[MULU16(y, _gray_info.width) + x];
+    
+    do
+    {
+        const unsigned char *src_col = src++;
+        unsigned char *dst_col = dst++;
+        unsigned data = *src_col >> src_y;
+        int numbits = 8 - src_y;
+        
+        dst_end = dst_col + MULU16(height, _gray_info.width);
+        do
+        {
+            if (data & 0x01)
+                fgfunc(dst_col);
+            else
+                bgfunc(dst_col);
+
+            dst_col += _gray_info.width;
+
+            data >>= 1;
+            if (--numbits == 0)
+            {
+                src_col += stride;
+                data = *src_col;
+                numbits = 8;
+            }
+        }
+        while (dst_col < dst_end);
+    }
+    while (src < src_end);
+#else /* LCD_PIXELFORMAT == VERTICAL_PACKING */
+    dst    = &_gray_info.cur_buffer[MULU16(x, _gray_info.height) + y];
     
     do
     {
@@ -475,6 +707,7 @@ void gray_mono_bitmap_part(const unsigned char *src, int src_x, int src_y,
         dst += _gray_info.height;
     }
     while (src < src_end);
+#endif /* LCD_PIXELFORMAT */
 }
 
 /* Draw a full monochrome bitmap */
@@ -487,7 +720,6 @@ void gray_mono_bitmap(const unsigned char *src, int x, int y, int width, int hei
 void gray_gray_bitmap_part(const unsigned char *src, int src_x, int src_y,
                            int stride, int x, int y, int width, int height)
 {
-    const unsigned char *src_end;
     unsigned char *dst, *dst_end;
 
     /* nothing to draw? */
@@ -514,25 +746,45 @@ void gray_gray_bitmap_part(const unsigned char *src, int src_x, int src_y,
         height = _gray_info.height - y;
 
     src    += MULU16(stride, src_y) + src_x; /* move starting point */
-    src_end = src + width;
+#if LCD_PIXELFORMAT == HORIZONTAL_PACKING
+    dst     = &_gray_info.cur_buffer[MULU16(y, _gray_info.width) + x];
+    dst_end = dst + MULU16(height, _gray_info.width);
+
+    do
+    {
+        const unsigned char *src_row = src;
+        unsigned char *dst_row = dst;
+        unsigned char *row_end = dst_row + width;
+
+        do
+            *dst_row++ = _gray_info.idxtable[*src_row++];
+        while (dst_row < row_end);
+
+        src += stride;
+        dst += _gray_info.width;
+    }
+    while (dst < dst_end);
+#else /* LCD_PIXELFORMAT == VERTICAL_PACKING */
     dst     = &_gray_info.cur_buffer[MULU16(x, _gray_info.height) + y];
+    dst_end = dst + height;
 
     do
     {
-        const unsigned char *src_col = src++;
-        unsigned char *dst_col = dst;
+        const unsigned char *src_row = src;
+        unsigned char *dst_row = dst++;
+        unsigned char *row_end = dst_row + MULU16(width, _gray_info.height);
 
-        dst_end = dst_col + height;
         do
         {
-            *dst_col++ = _gray_info.idxtable[*src_col];
-            src_col += stride;
+            *dst_row = _gray_info.idxtable[*src_row++];
+            dst_row += _gray_info.height;
         }
-        while (dst_col < dst_end);
+        while (dst_row < row_end);
 
-        dst += _gray_info.height;
+        src += stride;
     }
-    while (src < src_end);
+    while (dst < dst_end);
+#endif /* LCD_PIXELFORMAT */
 }
 
 /* Draw a full greyscale bitmap, canonical format */
@@ -612,6 +864,149 @@ void gray_ub_clear_display(void)
                      _gray_info.plane_size));
 }
 
+#if LCD_PIXELFORMAT == HORIZONTAL_PACKING
+
+/* Write a pixel block, defined by their brightnesses in a greymap.
+   Address is the byte in the first bitplane, src is the greymap start address,
+   stride is the increment for the greymap to get to the next pixel, mask
+   determines which pixels of the destination block are changed. */
+static void _writearray(unsigned char *address, const unsigned char *src,
+                        unsigned mask)
+{
+    unsigned long pat_stack[8];
+    unsigned long *pat_ptr = &pat_stack[8];
+    unsigned char *addr, *end;      
+#if 0 /* CPU specific asm versions will go here */
+#else /* C version, for reference*/
+    unsigned test = 0x80;
+    int i;
+
+    /* precalculate the bit patterns with random shifts
+     * for all 8 pixels and put them on an extra "stack" */
+    for (i = 7; i >= 0; i--)
+    {
+        unsigned pat = 0;
+
+        if (mask & test)
+        {
+            int shift;
+
+            pat = _gray_info.bitpattern[_gray_info.idxtable[*src]];
+
+            /* shift pattern pseudo-random, simple & fast PRNG */
+            _gray_random_buffer = 75 * _gray_random_buffer + 74;
+            shift = (_gray_random_buffer >> 8) & _gray_info.randmask;
+            if (shift >= _gray_info.depth)
+                shift -= _gray_info.depth;
+
+            pat = (pat << shift) | (pat >> (_gray_info.depth - shift));
+        }
+        *(--pat_ptr) = pat;
+        src++;
+        test >>= 1;
+    }
+
+    addr = address;
+    end = addr + MULU16(_gray_info.depth, _gray_info.plane_size);
+
+    /* set the bits for all 8 pixels in all bytes according to the
+     * precalculated patterns on the pattern stack */
+    test = 1;
+    mask = (~mask & 0xff);
+    if (mask == 0)
+    {
+        do
+        {
+            unsigned data = 0;
+
+            for (i = 7; i >= 0; i--)
+                data = (data << 1) | ((pat_stack[i] & test) ? 1 : 0);
+
+            *addr = data;
+            addr += _gray_info.plane_size;
+            test <<= 1;
+        }
+        while (addr < end);
+    }
+    else
+    {
+        do
+        {
+            unsigned data = 0;
+
+            for (i = 7; i >= 0; i--)
+                data = (data << 1) | ((pat_stack[i] & test) ? 1 : 0);
+
+            *addr = (*addr & mask) | data;
+            addr += _gray_info.plane_size;
+            test <<= 1;
+        }
+        while (addr < end);
+    }
+#endif
+}
+
+/* Draw a partial greyscale bitmap, canonical format */
+void gray_ub_gray_bitmap_part(const unsigned char *src, int src_x, int src_y,
+                              int stride, int x, int y, int width, int height)
+{
+    int shift, nx;
+    unsigned char *dst, *dst_end;
+    unsigned mask, mask_right;
+
+    /* nothing to draw? */
+    if ((width <= 0) || (height <= 0) || (x >= _gray_info.width)
+        || (y >= _gray_info.height) || (x + width <= 0) || (y + height <= 0))
+        return;
+
+    /* clipping */
+    if (x < 0)
+    {
+        width += x;
+        src_x -= x;
+        x = 0;
+    }
+    if (y < 0)
+    {
+        height += y;
+        src_y -= y;
+        y = 0;
+    }
+    if (x + width > _gray_info.width)
+        width = _gray_info.width - x;
+    if (y + height > _gray_info.height)
+        height = _gray_info.height - y;
+
+    shift = x & 7;
+    src += MULU16(stride, src_y) + src_x - shift;
+    dst  = _gray_info.plane_data + (x >> 3) + MULU16(_gray_info.bwidth, y);
+    nx   = width - 1 + shift;
+
+    mask = 0xFFu >> shift;
+    mask_right = 0xFFu << (~nx & 7);
+    
+    dst_end = dst + MULU16(_gray_info.bwidth, height);
+    do
+    {
+        const unsigned char *src_row = src;
+        unsigned char *dst_row = dst;
+        unsigned mask_row = mask;
+        
+        for (x = nx; x >= 8; x -= 8)
+        {
+            _writearray(dst_row++, src_row, mask_row);
+            src_row += 8;
+            mask_row = 0xFFu;
+        }
+        _writearray(dst_row, src_row, mask_row & mask_right);
+
+        src += stride;
+        dst += _gray_info.bwidth;
+    }
+    while (dst < dst_end);
+}
+#else /* LCD_PIXELFORMAT == VERTICAL_PACKING */
+
 /* Write a pixel block, defined by their brightnesses in a greymap.
    Address is the byte in the first bitplane, src is the greymap start address,
    stride is the increment for the greymap to get to the next pixel, mask
@@ -619,17 +1014,17 @@ void gray_ub_clear_display(void)
 static void _writearray(unsigned char *address, const unsigned char *src,
                         int stride, unsigned mask)
 {
-#if (CONFIG_CPU == SH7034) && (LCD_DEPTH == 1)
     unsigned long pat_stack[8];
     unsigned long *pat_ptr = &pat_stack[8];
+    unsigned char *addr, *end;      
+#if CONFIG_CPU == SH7034
     const unsigned char *_src;
-    unsigned char *addr, *end;
     unsigned _mask, trash;
 
     _mask = mask;
     _src = src;
 
-    /* precalculate the bit patterns with random shifts 
+    /* precalculate the bit patterns with random shifts
        for all 8 pixels and put them on an extra "stack" */
     asm volatile (
         "mov     #8,r3       \n"  /* loop count in r3: 8 pixels */
@@ -784,11 +1179,8 @@ static void _writearray(unsigned char *address, const unsigned char *src,
         : /* clobbers */
         "r0", "r1", "r2", "r3", "r6", "r7", "r8", "r9", "r10"
     );
-#elif defined(CPU_COLDFIRE) && (LCD_DEPTH == 2)
-    unsigned long pat_stack[8];
-    unsigned long *pat_ptr = &pat_stack[8];
+#elif defined(CPU_COLDFIRE)
     const unsigned char *_src;
-    unsigned char *addr, *end;
     unsigned _mask, trash;
 
     _mask = mask;
@@ -941,12 +1333,73 @@ static void _writearray(unsigned char *address, const unsigned char *src,
         : /* clobbers */
         "d0", "d1", "d2", "d3", "d4", "d5", "d6", "a0", "a1"
     );
-#elif defined(CPU_ARM) && (LCD_DEPTH == 1)
-    /* TODO: implement for iFP */
-    (void)address;
-    (void)src;
-    (void)stride;
-    (void)mask;
+#else /* C version, for reference*/
+#warning C version of _writearray() used
+    unsigned test = 1;
+    int i;
+
+    /* precalculate the bit patterns with random shifts
+     * for all 8 pixels and put them on an extra "stack" */
+    for (i = 7; i >= 0; i--)
+    {
+        unsigned pat = 0;
+
+        if (mask & test)
+        {
+            int shift;
+
+            pat = _gray_info.bitpattern[_gray_info.idxtable[*src]];
+
+            /* shift pattern pseudo-random, simple & fast PRNG */
+            _gray_random_buffer = 75 * _gray_random_buffer + 74;
+            shift = (_gray_random_buffer >> 8) & _gray_info.randmask;
+            if (shift >= _gray_info.depth)
+                shift -= _gray_info.depth;
+
+            pat = (pat << shift) | (pat >> (_gray_info.depth - shift));
+        }
+        *(--pat_ptr) = pat;
+        src += stride;
+        test <<= 1;
+    }
+
+    addr = address;
+    end = addr + MULU16(_gray_info.depth, _gray_info.plane_size);
+
+    /* set the bits for all 8 pixels in all bytes according to the
+     * precalculated patterns on the pattern stack */
+    test = 1;
+    mask = (~mask & 0xff);
+    if (mask == 0)
+    {
+        do
+        {
+            unsigned data = 0;
+
+            for (i = 0; i < 8; i++)
+                data = (data << 1) | ((pat_stack[i] & test) ? 1 : 0);
+
+            *addr = data;
+            addr += _gray_info.plane_size;
+            test <<= 1;
+        }
+        while (addr < end);
+    }
+    else
+    {
+        do
+        {
+            unsigned data = 0;
+
+            for (i = 0; i < 8; i++)
+                data = (data << 1) | ((pat_stack[i] & test) ? 1 : 0);
+
+            *addr = (*addr & mask) | data;
+            addr += _gray_info.plane_size;
+            test <<= 1;
+        }
+        while (addr < end);
+    }
 #endif
 }
 
@@ -981,16 +1434,16 @@ void gray_ub_gray_bitmap_part(const unsigned char *src, int src_x, int src_y,
     if (y + height > _gray_info.height)
         height = _gray_info.height - y;
 
-    shift = y & (_PBLOCK-1);
+    shift = y & 7;
     src += MULU16(stride, src_y) + src_x - MULU16(stride, shift);
     dst  = _gray_info.plane_data + x
-           + MULU16(_gray_info.width, y >> _PBLOCK_EXP);
+           + MULU16(_gray_info.width, y >> 3);
     ny   = height - 1 + shift;
 
     mask = 0xFFu << shift;
-    mask_bottom = 0xFFu >> (~ny & (_PBLOCK-1));
+    mask_bottom = 0xFFu >> (~ny & 7);
 
-    for (; ny >= _PBLOCK; ny -= _PBLOCK)
+    for (; ny >= 8; ny -= 8)
     {
         const unsigned char *src_row = src;
         unsigned char *dst_row = dst;
@@ -1000,7 +1453,7 @@ void gray_ub_gray_bitmap_part(const unsigned char *src, int src_x, int src_y,
             _writearray(dst_row++, src_row++, stride, mask);
         while (dst_row < dst_end);
 
-        src += stride << _PBLOCK_EXP;
+        src += stride << 3;
         dst += _gray_info.width;
         mask = 0xFFu;
     }
@@ -1010,6 +1463,7 @@ void gray_ub_gray_bitmap_part(const unsigned char *src, int src_x, int src_y,
         _writearray(dst++, src++, stride, mask);
     while (dst < dst_end);
 }
+#endif /* LCD_PIXELFORMAT */
 
 #endif /* !SIMULATOR */
 
@@ -1020,6 +1474,4 @@ void gray_ub_gray_bitmap(const unsigned char *src, int x, int y, int width,
     gray_ub_gray_bitmap_part(src, 0, 0, width, x, y, width, height);
 }
 
-
 #endif /* HAVE_LCD_BITMAP */
-
diff --git a/apps/plugins/lib/gray_parm.c b/apps/plugins/lib/gray_parm.c
index a6064e3ad3..ea39eadc66 100644
--- a/apps/plugins/lib/gray_parm.c
+++ b/apps/plugins/lib/gray_parm.c
@@ -11,7 +11,8 @@
 * Parameter handling
 *
 * This is a generic framework to display up to 33 shades of grey
-* on low-depth bitmap LCDs (Archos b&w, Iriver 4-grey) within plugins.
+* on low-depth bitmap LCDs (Archos b&w, Iriver 4-grey, iPod 4-grey)
+* within plugins.
 *
 * Copyright (C) 2004-2006 Jens Arnold
 *
@@ -29,11 +30,19 @@
 #include "gray.h"
 
 /* Set position of the top left corner of the greyscale overlay
-   Note that by is in pixel-block units (8 pixels) */
-void gray_set_position(int x, int by)
+   Note that depending on the target LCD, either x or y gets rounded
+   to the nearest multiple of 8 */
+void gray_set_position(int x, int y)
 {
+#if LCD_PIXELFORMAT == HORIZONTAL_PACKING
+    _gray_info.bx = (x + 3) / 8;
+    x = 8 * _gray_info.bx;
+#else
+    _gray_info.by = (y + 3) / 8;
+    y = 8 * _gray_info.by;
+#endif
     _gray_info.x = x;
-    _gray_info.by = by;
+    _gray_info.y = y;
     
     if (_gray_info.flags & _GRAY_RUNNING)
     {
diff --git a/apps/plugins/lib/gray_scroll.c b/apps/plugins/lib/gray_scroll.c
index 3cdb7d5afe..df5dc57044 100644
--- a/apps/plugins/lib/gray_scroll.c
+++ b/apps/plugins/lib/gray_scroll.c
@@ -11,7 +11,8 @@
 * Scrolling routines
 *
 * This is a generic framework to display up to 33 shades of grey
-* on low-depth bitmap LCDs (Archos b&w, Iriver 4-grey) within plugins.
+* on low-depth bitmap LCDs (Archos b&w, Iriver 4-grey, iPod 4-grey)
+* within plugins.
 *
 * Copyright (C) 2004-2006 Jens Arnold
 *
@@ -30,6 +31,96 @@
 
 /*** Scrolling ***/
 
+#if LCD_PIXELFORMAT == HORIZONTAL_PACKING
+
+/* Scroll left */
+void gray_scroll_left(int count)
+{
+    unsigned char *data, *data_end;
+    int length, blank;
+
+    if ((unsigned)count >= (unsigned)_gray_info.width)
+        return;
+
+    data = _gray_info.cur_buffer;
+    data_end = data + MULU16(_gray_info.width, _gray_info.height);
+    length = _gray_info.width - count;
+    blank = (_gray_info.drawmode & DRMODE_INVERSEVID) ?
+            _gray_info.fg_index : _gray_info.bg_index;
+
+    do
+    {
+        _gray_rb->memmove(data, data + count, length);
+        _gray_rb->memset(data + length, blank, count);
+        data += _gray_info.width;
+    }
+    while (data < data_end);
+}
+
+/* Scroll right */
+void gray_scroll_right(int count)
+{
+    unsigned char *data, *data_end;
+    int length, blank;
+
+    if ((unsigned)count >= (unsigned)_gray_info.width)
+        return;
+
+    data = _gray_info.cur_buffer;
+    data_end = data + MULU16(_gray_info.width, _gray_info.height);
+    length = _gray_info.width - count;
+    blank = (_gray_info.drawmode & DRMODE_INVERSEVID) ?
+            _gray_info.fg_index : _gray_info.bg_index;
+
+    do
+    {
+        _gray_rb->memmove(data + count, data, length);
+        _gray_rb->memset(data, blank, count);
+        data += _gray_info.width;
+    }
+    while (data < data_end);
+}
+
+/* Scroll up */
+void gray_scroll_up(int count)
+{
+    long shift, length;
+    int blank;
+
+    if ((unsigned)count >= (unsigned)_gray_info.height)
+        return;
+
+    shift = MULU16(_gray_info.width, count);
+    length = MULU16(_gray_info.width, _gray_info.height - count);
+    blank = (_gray_info.drawmode & DRMODE_INVERSEVID) ?
+            _gray_info.fg_index : _gray_info.bg_index;
+
+    _gray_rb->memmove(_gray_info.cur_buffer, _gray_info.cur_buffer + shift,
+                      length);
+    _gray_rb->memset(_gray_info.cur_buffer + length, blank, shift);
+}
+
+/* Scroll down */
+void gray_scroll_down(int count)
+{
+    long shift, length;
+    int blank;
+
+    if ((unsigned)count >= (unsigned)_gray_info.height)
+        return;
+
+    shift = MULU16(_gray_info.width, count);
+    length = MULU16(_gray_info.width, _gray_info.height - count);
+    blank = (_gray_info.drawmode & DRMODE_INVERSEVID) ?
+            _gray_info.fg_index : _gray_info.bg_index;
+
+    _gray_rb->memmove(_gray_info.cur_buffer + shift, _gray_info.cur_buffer,
+                      length);
+    _gray_rb->memset(_gray_info.cur_buffer, blank, shift);
+}
+
+#else /* LCD_PIXELFORMAT == VERTICAL_PACKING */
+
 /* Scroll left */
 void gray_scroll_left(int count)
 {
@@ -44,7 +135,8 @@ void gray_scroll_left(int count)
     blank = (_gray_info.drawmode & DRMODE_INVERSEVID) ?
             _gray_info.fg_index : _gray_info.bg_index;
 
-    _gray_rb->memmove(_gray_info.cur_buffer, _gray_info.cur_buffer + shift, length);
+    _gray_rb->memmove(_gray_info.cur_buffer, _gray_info.cur_buffer + shift,
+                      length);
     _gray_rb->memset(_gray_info.cur_buffer + length, blank, shift);
 }
 
@@ -62,7 +154,8 @@ void gray_scroll_right(int count)
     blank = (_gray_info.drawmode & DRMODE_INVERSEVID) ?
             _gray_info.fg_index : _gray_info.bg_index;
 
-    _gray_rb->memmove(_gray_info.cur_buffer + shift, _gray_info.cur_buffer, length);
+    _gray_rb->memmove(_gray_info.cur_buffer + shift, _gray_info.cur_buffer,
+                      length);
     _gray_rb->memset(_gray_info.cur_buffer, blank, shift);
 }
 
@@ -113,6 +206,7 @@ void gray_scroll_down(int count)
     }
     while (data < data_end);
 }
+#endif /* LCD_PIXELFORMAT */
 
 /*** Unbuffered scrolling functions ***/
 
@@ -148,6 +242,260 @@ void gray_ub_scroll_down(int count)
 
 #else /* !SIMULATOR */
 
+#if LCD_PIXELFORMAT == HORIZONTAL_PACKING
+
+/* Scroll left */
+void gray_ub_scroll_left(int count)
+{
+    int shift, length;
+    unsigned char *ptr, *ptr_end;
+
+    if ((unsigned) count >= (unsigned) _gray_info.width)
+        return;
+
+    shift = count >> 3;
+    count &= 7;
+
+    if (shift)
+    {
+        length = _gray_info.bwidth - shift;
+        ptr = _gray_info.plane_data;
+        ptr_end = ptr + _gray_info.plane_size;
+
+        /* Scroll row by row to minimize flicker */
+        do
+        {
+            unsigned char *ptr_row = ptr;
+            unsigned char *row_end = ptr_row
+                             + MULU16(_gray_info.plane_size, _gray_info.depth);
+            do
+            {
+                _gray_rb->memmove(ptr_row, ptr_row + shift, length);
+                _gray_rb->memset(ptr_row + length, 0, shift);
+                ptr_row += _gray_info.plane_size;
+            }
+            while (ptr_row < row_end);
+
+            ptr += _gray_info.bwidth;
+        }
+        while (ptr < ptr_end);
+    }
+    if (count)
+    {
+        asm (
+            "mov     r4, %[high]     \n"
+
+        ".sl_rloop:                  \n"
+            "mov     r5, %[addr]     \n"
+            "mov     r2, %[dpth]     \n"
+
+        ".sl_oloop:                  \n"
+            "mov     r6, r5          \n"
+            "mov     r3, %[cols]     \n"
+            "mov     r1, #0          \n"
+
+        ".sl_iloop:                          \n"
+            "mov     r1, r1, lsr #8          \n"
+            "ldrb    r0, [r6, #-1]!          \n"
+            "orr     r1, r1, r0, lsl %[cnt]  \n"
+            "strb    r1, [r6]                \n"
+
+            "subs    r3, r3, #1      \n"
+            "bne     .sl_iloop       \n"
+
+            "add     r5, r5, %[psiz] \n"
+            "subs    r2, r2, #1      \n"
+            "bne     .sl_oloop       \n"
+
+            "add     %[addr],%[addr],%[bwid] \n"
+            "subs    r4, r4, #1              \n"
+            "bne     .sl_rloop               \n"
+            : /* outputs */
+            : /* inputs */
+            [dpth]"r"(_gray_info.depth),
+            [high]"r"(_gray_info.height),
+            [bwid]"r"(_gray_info.bwidth),
+            [cols]"r"(_gray_info.bwidth - shift),
+            [psiz]"r"(_gray_info.plane_size),
+            [addr]"r"(_gray_info.plane_data + _gray_info.bwidth - shift),
+            [cnt] "r"(count)
+            : /* clobbers */
+            "r0", "r1", "r2", "r3", "r4", "r5", "r6"
+        );
+    }
+}
+
+/* Scroll right */
+void gray_ub_scroll_right(int count)
+{
+    int shift, length;
+    unsigned char *ptr, *ptr_end;
+
+    if ((unsigned) count >= (unsigned) _gray_info.width)
+        return;
+
+    shift = count >> 3;
+    count &= 7;
+
+    if (shift)
+    {
+        length = _gray_info.bwidth - shift;
+        ptr = _gray_info.plane_data;
+        ptr_end = ptr + _gray_info.plane_size;
+
+        /* Scroll row by row to minimize flicker */
+        do
+        {
+            unsigned char *ptr_row = ptr;
+            unsigned char *row_end = ptr_row
+                             + MULU16(_gray_info.plane_size, _gray_info.depth);
+            do
+            {
+                _gray_rb->memmove(ptr_row + shift, ptr_row, length);
+                _gray_rb->memset(ptr_row, 0, shift);
+                ptr_row += _gray_info.plane_size;
+            }
+            while (ptr_row < row_end);
+
+            ptr += _gray_info.bwidth;
+        }
+        while (ptr < ptr_end);
+    }
+    if (count)
+    {
+        asm (
+            "mov     r4, %[high]     \n"
+
+        ".sr_rloop:                  \n"
+            "mov     r5, %[addr]     \n"
+            "mov     r2, %[dpth]     \n"
+
+        ".sr_oloop:                  \n"
+            "mov     r6, r5          \n"
+            "mov     r3, %[cols]     \n"
+            "mov     r1, #0          \n"
+
+        ".sr_iloop:                          \n"
+            "ldrb    r0, [r6]                \n"
+            "orr     r1, r0, r1, lsl #8      \n"
+            "mov     r0, r1, lsr %[cnt]      \n"
+            "strb    r0, [r6], #1            \n"
+
+            "subs    r3, r3, #1      \n"
+            "bne     .sr_iloop       \n"
+
+            "add     r5, r5, %[psiz] \n"
+            "subs    r2, r2, #1      \n"
+            "bne     .sr_oloop       \n"
+
+            "add     %[addr],%[addr],%[bwid] \n"
+            "subs    r4, r4, #1              \n"
+            "bne     .sr_rloop               \n"
+            : /* outputs */
+            : /* inputs */
+            [dpth]"r"(_gray_info.depth),
+            [high]"r"(_gray_info.height),
+            [bwid]"r"(_gray_info.bwidth),
+            [cols]"r"(_gray_info.bwidth - shift),
+            [psiz]"r"(_gray_info.plane_size),
+            [addr]"r"(_gray_info.plane_data + shift),
+            [cnt] "r"(count)
+            : /* clobbers */
+            "r0", "r1", "r2", "r3", "r4", "r5", "r6"
+        );
+    }
+}
+
+/* Scroll up */
+void gray_ub_scroll_up(int count)
+{
+    long blockshift;
+    unsigned char *ptr, *ptr_end1, *ptr_end2;
+
+    if ((unsigned) count >= (unsigned) _gray_info.height)
+        return;
+
+    blockshift = MULU16(_gray_info.bwidth, count);
+    ptr      = _gray_info.plane_data;
+    ptr_end2 = ptr + _gray_info.plane_size;
+    ptr_end1 = ptr_end2 - blockshift;
+    /* Scroll row by row to minimize flicker */
+    do
+    {
+        unsigned char *ptr_row = ptr;
+        unsigned char *row_end = ptr_row
+                             + MULU16(_gray_info.plane_size, _gray_info.depth);
+        if (ptr < ptr_end1)
+        {
+            do
+            {
+                _gray_rb->memcpy(ptr_row, ptr_row + blockshift,
+                                 _gray_info.bwidth);
+                ptr_row += _gray_info.plane_size;
+            }
+            while (ptr_row < row_end);
+        }
+        else
+        {
+            do
+            {
+                _gray_rb->memset(ptr_row, 0, _gray_info.bwidth);
+                ptr_row += _gray_info.plane_size;
+            }
+            while (ptr_row < row_end);
+        }
+
+        ptr += _gray_info.bwidth;
+    }
+    while (ptr < ptr_end2);
+}
+
+/* Scroll down */
+void gray_ub_scroll_down(int count)
+{
+    long blockshift;
+    unsigned char *ptr, *ptr_end1, *ptr_end2;
+
+    if ((unsigned) count >= (unsigned) _gray_info.height)
+        return;
+
+    blockshift = MULU16(_gray_info.bwidth, count);
+    ptr_end2 = _gray_info.plane_data;
+    ptr_end1 = ptr_end2 + blockshift;
+    ptr      = ptr_end2 + _gray_info.plane_size;
+    /* Scroll row by row to minimize flicker */
+    do
+    {
+        unsigned char *ptr_row, *row_end;
+
+        ptr -= _gray_info.bwidth;
+        ptr_row = ptr;
+        row_end = ptr_row + MULU16(_gray_info.plane_size, _gray_info.depth);
+
+        if (ptr >= ptr_end1)
+        {
+            do
+            {
+                _gray_rb->memcpy(ptr_row, ptr_row - blockshift,
+                                 _gray_info.bwidth);
+                ptr_row += _gray_info.plane_size;
+            }
+            while (ptr_row < row_end);
+        }
+        else
+        {
+            do
+            {
+                _gray_rb->memset(ptr_row, 0, _gray_info.bwidth);
+                ptr_row += _gray_info.plane_size;
+            }
+            while (ptr_row < row_end);
+        }
+    }
+    while (ptr > ptr_end2);
+}
+#else /* LCD_PIXELFORMAT == VERTICAL_PACKING */
+
 /* Scroll left */
 void gray_ub_scroll_left(int count)
 {
@@ -156,17 +504,17 @@ void gray_ub_scroll_left(int count)
 
     if ((unsigned) count >= (unsigned) _gray_info.width)
         return;
-        
+
     length = _gray_info.width - count;
     ptr = _gray_info.plane_data;
     ptr_end = ptr + _gray_info.plane_size;
-    
+
     /* Scroll row by row to minimize flicker (pixel block rows) */
     do
     {
         unsigned char *ptr_row = ptr;
-        unsigned char *row_end = ptr_row 
-                               + MULU16(_gray_info.plane_size, _gray_info.depth);
+        unsigned char *row_end = ptr_row
+                             + MULU16(_gray_info.plane_size, _gray_info.depth);
         do
         {
             _gray_rb->memmove(ptr_row, ptr_row + count, length);
@@ -188,17 +536,17 @@ void gray_ub_scroll_right(int count)
 
     if ((unsigned) count >= (unsigned) _gray_info.width)
         return;
-        
+
     length = _gray_info.width - count;
     ptr = _gray_info.plane_data;
     ptr_end = ptr + _gray_info.plane_size;
-    
+
     /* Scroll row by row to minimize flicker (pixel block rows) */
     do
     {
         unsigned char *ptr_row = ptr;
-        unsigned char *row_end = ptr_row 
-                               + MULU16(_gray_info.plane_size, _gray_info.depth);
+        unsigned char *row_end = ptr_row
+                             + MULU16(_gray_info.plane_size, _gray_info.depth);
         do
         {
             _gray_rb->memmove(ptr_row + count, ptr_row, length);
@@ -221,10 +569,10 @@ void gray_ub_scroll_up(int count)
 
     if ((unsigned) count >= (unsigned) _gray_info.height)
         return;
-        
-    shift = count >> _PBLOCK_EXP;
-    count &= (_PBLOCK-1);
-    
+
+    shift = count >> 3;
+    count &= 7;
+
     if (shift)
     {
         blockshift = MULU16(_gray_info.width, shift);
@@ -235,7 +583,7 @@ void gray_ub_scroll_up(int count)
         do
         {
             unsigned char *ptr_row = ptr;
-            unsigned char *row_end = ptr_row 
+            unsigned char *row_end = ptr_row
                              + MULU16(_gray_info.plane_size, _gray_info.depth);
             if (ptr < ptr_end1)
             {
@@ -263,7 +611,7 @@ void gray_ub_scroll_up(int count)
     }
     if (count)
     {
-#if (CONFIG_CPU == SH7034) && (LCD_DEPTH == 1)
+#if CONFIG_CPU == SH7034
         /* scroll column by column to minimize flicker */
         asm (
             "mov     #0,r6       \n"  /* x = 0 */
@@ -345,7 +693,7 @@ void gray_ub_scroll_up(int count)
             : /* clobbers */
             "r0", "r1", "r2", "r3", "r4", "r5", "r6"
         );
-#elif defined(CPU_COLDFIRE) && (LCD_DEPTH == 2)
+#elif defined(CPU_COLDFIRE)
         /* scroll column by column to minimize flicker */
         asm (
             "move.l  %[wide],%%d4\n"  /* columns = width */
@@ -406,9 +754,9 @@ void gray_ub_scroll_down(int count)
     if ((unsigned) count >= (unsigned) _gray_info.height)
         return;
 
-    shift = count >> _PBLOCK_EXP;
-    count &= (_PBLOCK-1);
-    
+    shift = count >> 3;
+    count &= 7;
+
     if (shift)
     {
         blockshift = MULU16(_gray_info.width, shift);
@@ -448,7 +796,7 @@ void gray_ub_scroll_down(int count)
     }
     if (count)
     {
-#if (CONFIG_CPU == SH7034) && (LCD_DEPTH == 1)
+#if CONFIG_CPU == SH7034
         /* scroll column by column to minimize flicker */
         asm (
             "mov     #0,r6       \n"  /* x = 0 */
@@ -529,7 +877,7 @@ void gray_ub_scroll_down(int count)
             : /* clobbers */
             "r0", "r1", "r2", "r3", "r4", "r5", "r6"
         );
-#elif defined(CPU_COLDFIRE) && (LCD_DEPTH == 2)
+#elif defined(CPU_COLDFIRE)
         /* scroll column by column to minimize flicker */
         asm (
             "move.l  %[wide],%%d4\n"  /* columns = width */
@@ -576,7 +924,8 @@ void gray_ub_scroll_down(int count)
 #endif
     }
 }
+#endif /* LCD_PIXELFORMAT */
+
 #endif /* !SIMULATOR */
 
 #endif /* HAVE_LCD_BITMAP */
-
diff --git a/apps/plugins/mandelbrot.c b/apps/plugins/mandelbrot.c
index dbcf08fa62..20baca71db 100644
--- a/apps/plugins/mandelbrot.c
+++ b/apps/plugins/mandelbrot.c
@@ -487,7 +487,7 @@ enum plugin_status plugin_start(struct plugin_api* api, void* parameter)
 
     /* initialize the grayscale buffer:
      * 8 bitplanes for 9 shades of gray.*/
-    grayscales = gray_init(rb, gbuf, gbuf_size, false, LCD_WIDTH, LCD_HEIGHT/8,
+    grayscales = gray_init(rb, gbuf, gbuf_size, false, LCD_WIDTH, LCD_HEIGHT,
                            8, 0, NULL) + 1;
     if (grayscales != 9) {
         rb->snprintf(buff, sizeof(buff), "%d", grayscales);
diff --git a/apps/plugins/plasma.c b/apps/plugins/plasma.c
index 676dba9a54..77c489f2ae 100644
--- a/apps/plugins/plasma.c
+++ b/apps/plugins/plasma.c
@@ -243,7 +243,7 @@ int main(void)
     /* get the remainder of the plugin buffer */
     gbuf = (unsigned char *) rb->plugin_get_buffer(&gbuf_size);
 
-    gray_init(rb, gbuf, gbuf_size, false, LCD_WIDTH, LCD_HEIGHT/8, 32, 2<<8, NULL);
+    gray_init(rb, gbuf, gbuf_size, false, LCD_WIDTH, LCD_HEIGHT, 32, 2<<8, NULL);
     /* switch on grayscale overlay */
     gray_show(true);
 #endif