[Difficulty with Sharp LCD]

I am trying to get the Sharp LCD LS010B7DH01 working with uGFX, ChibiOS 2.6.1 and STM32L151CB. It is quite a basic display. There are no special features, and one can write lines to the memory with the following SPI bytestream.

Write command byte, line address, 16 bytes of pixel data, 2 dummy bytes. http://www.avnet-embedded.eu/fileadmin/user_upload/Files/Displays/Mono/Sharp_LS010B7DH01.pdf (Datasheet)

Here are relevant excerpts from gdisp_lld_*.c

/*===========================================================================*/
/* Driver local definitions.                                                 */
/*===========================================================================*/

#ifndef GDISP_SCREEN_HEIGHT
	#define GDISP_SCREEN_HEIGHT		128		// This controller should support 32 (untested) or 64
#endif
#ifndef GDISP_SCREEN_WIDTH
	#define GDISP_SCREEN_WIDTH		128
#endif


#define LS010_LINE_WIDTH		GDISP_SCREEN_WIDTH/8

#define GDISP_FLG_NEEDFLUSH			(GDISP_FLG_DRIVER<<0)

#include "LS010.h"

/*===========================================================================*/
/* Driver local functions.                                                   */
/*===========================================================================*/

// Some common routines and macros
#define RAM(g)							((uint8_t *)g->priv)

// Some common routines and macros
#define delay(us)			gfxSleepMicroseconds(us)
#define delayms(ms)			gfxSleepMilliseconds(ms)

#define xyaddr(x, y)		(x + y * GDISP_SCREEN_WIDTH) / 8
#define xybit(x, y)			(1 << ((x + y * GDISP_SCREEN_WIDTH) % 8))
/*===========================================================================*/
/* Driver exported functions.                                                */
/*===========================================================================*/

LLDSPEC bool_t gdisp_lld_init(GDisplay *g) {
	// The private area is the display surface.
	g->priv = gfxAlloc(2048);

	lcd_init(&handle_lcd); // <--- My initialisation function which successfully turns LCD on

	acquire_bus(g);

    // Finish Init
    post_init_board(g);

 	// Release the bus
	release_bus(g);

	/* Initialise the GDISP structure */
	g->g.Width = GDISP_SCREEN_WIDTH;
	g->g.Height = GDISP_SCREEN_HEIGHT;
	g->g.Orientation = GDISP_ROTATE_0;
	g->g.Powermode = powerOn;
	g->g.Backlight = GDISP_INITIAL_BACKLIGHT;
	g->g.Contrast = GDISP_INITIAL_CONTRAST;
	return TRUE;
}

#if GDISP_HARDWARE_FLUSH
	LLDSPEC void gdisp_lld_flush(GDisplay *g) {
		unsigned	i;
		uint8_t j = 128; //Address

		// Don't flush if we don't need it.
		if (!(g->flags & GDISP_FLG_NEEDFLUSH))
			return;

		for(i=0; i < 2048; i += LS010_LINE_WIDTH)
        {
			write_data(g, RAM(g)+i, LS010_LINE_WIDTH, j);
			j--;
        }
	}
#endif

#if GDISP_HARDWARE_DRAWPIXEL
	LLDSPEC void gdisp_lld_draw_pixel(GDisplay *g) {
		coord_t		x, y;
		int tmpx = xyaddr(g->p.x, g->p.y);
		int bit = xybit(g->p.x,g->p.y);

		switch(g->g.Orientation) {
		default:
		case GDISP_ROTATE_0:
			x = g->p.x;
			y = g->p.y;
			break;
		case GDISP_ROTATE_90:
			x = g->p.y;
			y = GDISP_SCREEN_HEIGHT-1 - g->p.x;
			break;
		case GDISP_ROTATE_180:
			x = GDISP_SCREEN_WIDTH-1 - g->p.x;
			y = GDISP_SCREEN_HEIGHT-1 - g->p.y;
			break;
		case GDISP_ROTATE_270:
			x = GDISP_SCREEN_HEIGHT-1 - g->p.y;
			x = g->p.x;
			break;
		}
	if (gdispColor2Native(g->p.color) != Black)
        {
			RAM(g)[xyaddr(x, y)] &= xybit(x,y);
		}else {
			RAM(g)[xyaddr(x, y)] |= ~xybit(x,y);
		g->flags |= GDISP_FLG_NEEDFLUSH;
		}
	}
#endif

and and board_*.h

static uint8_t reverse(uint8_t b) {
   b = (b & 0xF0) >> 4 | (b & 0x0F) << 4;
   b = (b & 0xCC) >> 2 | (b & 0x33) << 2;
   b = (b & 0xAA) >> 1 | (b & 0x55) << 1;
   return b;
}

static inline void write_data(GDisplay *g, uint8_t* data, uint16_t length, uint8_t addr) {
	(void) g;

	uint8_t tmp[20] = {0}; // Command to write + address + 16 byte pixel data + 2 dummy bytes
	tmp[0] = 0b10000000;      // Command to write
	tmp[1] = reverse(addr);  // LSB first

	memmove(tmp+2, data, 16);

	spiUnselect(&SPID1);
	chThdSleepMicroseconds(SCS_HIGH_DELAY);
	spiSend(&SPID1, 20, tmp);
    chThdSleepMicroseconds(SCS_LOW_DELAY);
	spiSelect(&SPID1);
	chThdSleepMicroseconds(INTERFRAME_DELAY);
}

However, I am not able to get correct output. For example, here is the output when running the 'basics' example.

Can anyone see something wrong here?