Re: [jekhor@gmail.com: Metronome specifications]

Re: [jekhor@gmail.com: Metronome specifications]

[Jaya Kumar]

Hi Yauhen,

My apologies for the delay in replying. As you can imagine, life is
busy at Christmas and New Year and trying to clean out my patch box.
Yes, I will be happy to help in any way that I can. I think it is
great that you're doing the work to get the prs-505 supported. I
unfortunately have no documentation that I can pass along but will be
happy to take a look at any code.

Ok, in response to the technical questions, can you please post the
code for the board driver here. For the other people who also asked
similar questions about broadsheet and other controllers, please post
your questions and code here on the mailing list rather than to me
alone. The fbdev mailing list is the right place for all of us to work
together on this.

Thanks,
jaya

On Tue, Dec 30, 2008 at 11:46 PM, Yauhen Kharuzhy <jekhor@gmail.com> wrote:
>
> Excuse me for double posting, but can you send any response to me, ever
> negative? I also wrote to eInk and they refer me to you as expert in
> this area.
>
> ----- Forwarded message from Yauhen Kharuzhy <jekhor@gmail.com> -----
>
> From: Yauhen Kharuzhy <jekhor@gmail.com>
> To: Jaya Kumar <jayakumar.lkml@gmail.com>
> Cc: linux-fbdev-devel@lists.sourceforge.net
> Subject: Metronome specifications
> Date: Wed, 24 Dec 2008 23:57:36 +0200
>
> Hi!
>
> I write board driver for Sony PRS-505 book reader display controller and
> need some help. Controller seems to be respond for first command only (RDY is
> 0 before powerup and 1 after and it won't change on any other steps).
> I cannot localize cause of problem because controller behaviour is
> undocumented and non-trivial. I compare Sony's driver and metronomefb,
> and don't understand differences due lack of documentation.
>
> Can you help to find any documentation for Metronome or, if it
> is distrubuted under NDA, explain some moments for me?
>
> How dimensions of image and size of buffer are set in controller? What
> does define positions of command buffer, waveforms, image data etc. in
> controller memory? What are arguments of config (0xCC10) command mean?
>
> I feel that I will find answers for most of this questions after some
> time, but with documentation it will happen soon...
>
> ----- End forwarded message -----
>
> --
> Yauhen Kharuzhy         jekhor _at_ gmail.com
>                        JID: jek@jabber.ru
>
> A: No
> Q: Should I quote below my post?
>

------------------------------------------------------------------------------

Re: [jekhor@gmail.com: Metronome specifications]

[Yauhen Kharuzhy]

[-- Attachment #1: Type: text/plain, Size: 1894 bytes --]

On Tue, Dec 30, 2008 at 11:56:22PM +0800, Jaya Kumar wrote:
> Hi Yauhen,
> 
> My apologies for the delay in replying. As you can imagine, life is
> busy at Christmas and New Year and trying to clean out my patch box.
> Yes, I will be happy to help in any way that I can. I think it is
> great that you're doing the work to get the prs-505 supported. I
> unfortunately have no documentation that I can pass along but will be
> happy to take a look at any code.
> 
> Ok, in response to the technical questions, can you please post the
> code for the board driver here. For the other people who also asked
> similar questions about broadsheet and other controllers, please post
> your questions and code here on the mailing list rather than to me
> alone. The fbdev mailing list is the right place for all of us to work
> together on this.

OK. I am sorry for my impatience.

I just wrote a "new style" (for 2.6.28 kernel) board driver using of
the existing imxfb driver similar to you am200epd board driver. I found two
problems:
1) the metronomefb driver initializes before imxfb driver and a
kernel Oops caused in the setup_fb() function because the board.host_fbinfo is not
initialized yet. For prevent this, I placed a call of my prs505_presetup_fb()
function immediately after the fb_register_client() call.

2) the LCD controller in IMX SoCs supports LCD panels with height <= 511 (9 bit in
LCD_SIZE register). In the original Sony driver this height is 312. As I
suppose, it is half of the real height because LCD controller uses 16 bit
per pixel. I set yres in my prs505_fb_info to 622 / 2.

Now metronomefb initializes but nothing happen on the screen.

I attached my board driver.
Sony driver, if needed, can be found here:
http://users.openinkpot.org/~jekhor/8track20070714.tgz


-- 
Yauhen Kharuzhy		jekhor _at_ gmail.com
			JID: jek@jabber.ru

A: No
Q: Should I quote below my post?

[-- Attachment #2: prs505-display.c --]
[-- Type: text/plain, Size: 9101 bytes --]

/*
 * prs505-display.c -- Platform device for Sony PRS-505 display driver
 *
 * Copyright (C) 2008, Yauhen Kharuzhy
 *
 * This file is subject to the terms and conditions of the GNU General Public
 * License. See the file COPYING in the main directory of this archive for
 * more details.
 *
 * This driver is written to be used with the Metronome display controller.
 *
 */

#define DEBUG

#include <linux/module.h>
#include <linux/kernel.h>
#include <linux/errno.h>
#include <linux/string.h>
#include <linux/delay.h>
#include <linux/interrupt.h>
#include <linux/fb.h>
#include <linux/init.h>
#include <linux/platform_device.h>
#include <linux/irq.h>
#include <linux/gpio.h>

#include <mach/imxfb.h>
#include <mach/irqs.h>
#include <video/metronomefb.h>

#include "generic.h"

static struct platform_device *prs505_display_device;
static struct metronome_board prs505_board;

struct imxfb_mach_info prs505_fb_info = {
	.pixclock	= 40189,
	.xres		= 832,
	.yres		= 622 / 2,
	.bpp		= 16,
	.hsync_len	= 28,
	.left_margin	= 34,
	.right_margin	= 34,
	.vsync_len	= 25,
	.upper_margin	= 0,
	.lower_margin	= 2,
	.sync		= FB_SYNC_HOR_HIGH_ACT | FB_SYNC_VERT_HIGH_ACT,
	.pcr		= PCR_TFT | PCR_BPIX_16 | PCR_CLKPOL | PCR_SCLKIDLE |
				PCR_SCLK_SEL | PCR_PCD(2),
	.dmacr		= DMACR_HM(3) | DMACR_TM(10), 
};

#define STDBY_GPIO_PIN		(GPIO_PORTD | 10)
#define RST_GPIO_PIN		(GPIO_PORTD | 11)
#define RDY_GPIO_PIN		(GPIO_PORTD | 9)
#define ERR_GPIO_PIN		(GPIO_PORTD | 7)
#define PWR_VIO_GPIO_PIN	(GPIO_PORTA | 2)
#define PWR_VCORE_GPIO_PIN	(GPIO_PORTA | 4)

static int gpios[] = {  STDBY_GPIO_PIN , RST_GPIO_PIN,
			RDY_GPIO_PIN, ERR_GPIO_PIN, PWR_VCORE_GPIO_PIN,
			PWR_VIO_GPIO_PIN };

static char *gpio_names[] = { "STDBY" , "RST", "RDY", "ERR",
	"VCORE", "VIO" };

static int prs505_init_gpio_regs(struct metronomefb_par *par)
{
	int i;
	int err;

	for (i = 0; i < ARRAY_SIZE(gpios); i++) {
		err = gpio_request(gpios[i], gpio_names[i]);
		if (err) {
			dev_err(&prs505_display_device->dev, "failed requesting "
				"gpio %s, err=%d\n", gpio_names[i], err);
			goto err_req_gpio;
		}
	}

	gpio_direction_output(STDBY_GPIO_PIN, 0);
	gpio_direction_output(RST_GPIO_PIN, 0);

	gpio_direction_input(RDY_GPIO_PIN);
	gpio_direction_input(ERR_GPIO_PIN);

	gpio_direction_output(PWR_VCORE_GPIO_PIN, 1);
	gpio_direction_output(PWR_VIO_GPIO_PIN, 1);

	return 0;

err_req_gpio:
	while (i > 0)
		gpio_free(gpios[i--]);

	return err;
}

static void prs505_cleanup(struct metronomefb_par *par)
{
	int i;

	free_irq(gpio_to_irq(RDY_GPIO_PIN), par);

	for (i = 0; i < ARRAY_SIZE(gpios); i++)
		gpio_free(gpios[i]);
}

static int prs505_share_video_mem(struct fb_info *info)
{
	/* rough check if this is our desired fb and not something else */
	if ((info->var.xres != prs505_fb_info.xres)
		|| (info->var.yres != prs505_fb_info.yres))
		return 0;

	/* we've now been notified that we have our new fb */
	prs505_board.metromem = info->screen_base;
	prs505_board.host_fbinfo = info;

	/* try to refcount host drv since we are the consumer after this */
	if (!try_module_get(info->fbops->owner))
		return -ENODEV;

	return 0;
}

static int prs505_unshare_video_mem(struct fb_info *info)
{
	dev_dbg(&prs505_display_device->dev, "ENTER %s\n", __func__);

	if (info != prs505_board.host_fbinfo)
		return 0;

	module_put(prs505_board.host_fbinfo->fbops->owner);
	return 0;
}

static int prs505_fb_notifier_callback(struct notifier_block *self,
				 unsigned long event, void *data)
{
	struct fb_event *evdata = data;
	struct fb_info *info = evdata->info;

//	dev_dbg(&prs505_display_device->dev, "ENTER %s\n", __func__);

	if (event == FB_EVENT_FB_REGISTERED)
		return prs505_share_video_mem(info);
	else if (event == FB_EVENT_FB_UNREGISTERED)
		return prs505_unshare_video_mem(info);

	return 0;
}

static struct notifier_block prs505_fb_notif = {
	.notifier_call = prs505_fb_notifier_callback,
};

/* this gets called as part of our init. these steps must be done now so
 * that we can use set_pxa_fb_info */
static int __init prs505_presetup_fb(void)
{
	int fw;
	int fh;
	int padding_size;
	int totalsize;
	int ret;

	/* the frame buffer is divided as follows:
	command | CRC | padding
	16kb waveform data | CRC | padding
	image data | CRC
	*/

	fw = prs505_fb_info.xres;
	fh = prs505_fb_info.yres;

	/* waveform must be 16k + 2 for checksum */
	prs505_board.wfm_size = roundup(16*1024 + 2, fw);

	padding_size = PAGE_SIZE + (4 * fw);

	/* total is 1 cmd , 1 wfm, padding and image */
	totalsize = fw + prs505_board.wfm_size + padding_size + (fw*fh);

	/* save this off because we're manipulating fw after this and
	 * we'll need it when we're ready to setup the framebuffer */
	prs505_board.fw = fw;
	prs505_board.fh = fh;

	/* the reason we do this adjustment is because we want to acquire
	 * more framebuffer memory without imposing custom awareness on the
	 * underlying pxafb driver */
	prs505_fb_info.yres = DIV_ROUND_UP(totalsize, fw);

	/* we divide since we told the LCD controller we're 16bpp */
	prs505_fb_info.xres /= 2;

	set_imx_fb_info(&prs505_fb_info);

	ret = platform_device_register(&imxfb_device);
	if (ret)
		dev_err(&prs505_display_device->dev, "Failed to register imxfb platform device\n");

	return ret;
}

/* this gets called by metronomefb as part of its init, in our case, we
 * have already completed initial framebuffer init in presetup_fb so we
 * can just setup the fb access pointers */
static int prs505_setup_fb(struct metronomefb_par *par)
{
	int fw;
	int fh;

	fw = prs505_board.fw;
	fh = prs505_board.fh;

	/* metromem was set up by the notifier in share_video_mem so now
	 * we can use its value to calculate the other entries */
	par->metromem_cmd = (struct metromem_cmd *) prs505_board.metromem;
	par->metromem_wfm = prs505_board.metromem + fw;
	par->metromem_img = par->metromem_wfm + prs505_board.wfm_size;
	par->metromem_img_csum = (u16 *) (par->metromem_img + (fw * fh));
	par->metromem_dma = prs505_board.host_fbinfo->fix.smem_start;

	return 0;
}

static int prs505_get_panel_type(void)
{
	return 6;
}

static irqreturn_t prs505_handle_irq(int irq, void *dev_id)
{
	struct metronomefb_par *par = dev_id;

	wake_up_interruptible(&par->waitq);
	return IRQ_HANDLED;
}

static int prs505_setup_irq(struct fb_info *info)
{
	int ret;

	ret = request_irq(gpio_to_irq(RDY_GPIO_PIN), prs505_handle_irq,
				IRQF_DISABLED | IRQF_TRIGGER_FALLING,
				"PRS505-display", info->par);
	if (ret)
		dev_err(&prs505_display_device->dev, "request_irq failed: %d\n", ret);

	return ret;
}

static void prs505_set_rst(struct metronomefb_par *par, int state)
{
	if (state && gpio_get_value(RDY_GPIO_PIN))
		dev_err(&prs505_display_device->dev, "RDY != 0 before set_rst\n");
	gpio_set_value(RST_GPIO_PIN, state);
}

static void prs505_set_stdby(struct metronomefb_par *par, int state)
{
	gpio_set_value(STDBY_GPIO_PIN, state);
}

static int prs505_wait_event(struct metronomefb_par *par)
{
	int ret;
	ret = wait_event_timeout(par->waitq, gpio_get_value(RDY_GPIO_PIN), HZ);
	dev_dbg(&prs505_display_device->dev, "%s: ret = %d\n", __FUNCTION__, ret);
	return ret;
}

static int prs505_wait_event_intr(struct metronomefb_par *par)
{
	return wait_event_interruptible_timeout(par->waitq,
					gpio_get_value(RDY_GPIO_PIN), HZ);
}

static struct metronome_board prs505_board = {
	.owner			= THIS_MODULE,
	.setup_irq		= prs505_setup_irq,
	.setup_io		= prs505_init_gpio_regs,
	.setup_fb		= prs505_setup_fb,
	.set_rst		= prs505_set_rst,
	.set_stdby		= prs505_set_stdby,
	.met_wait_event		= prs505_wait_event,
	.met_wait_event_intr	= prs505_wait_event_intr,
	.get_panel_type		= prs505_get_panel_type,
	.cleanup		= prs505_cleanup,
};

static int __init prs505_init(void)
{
	int ret;

	/* before anything else, we request notification for any fb
	 * creation events */
	fb_register_client(&prs505_fb_notif);

	prs505_presetup_fb();

	if (!prs505_board.host_fbinfo) {
		printk(KERN_ERR "prs505_board.host_fbinfo is NULL. Probably you don't have the imxfb driver enabled\n");
		platform_device_unregister(&imxfb_device);
		return -ENODEV;
	}
	/* request our platform independent driver */
	request_module("metronomefb");

	prs505_display_device = platform_device_alloc("metronomefb", -1);
	if (!prs505_display_device) {
		platform_device_unregister(&imxfb_device);
		return -ENOMEM;
	}

	/* the prs505_board that will be seen by metronomefb is a copy */
	platform_device_add_data(prs505_display_device, &prs505_board,
					sizeof(prs505_board));

	/* this _add binds metronomefb to prs505. metronomefb refcounts prs505 */
	ret = platform_device_add(prs505_display_device);

	if (ret) {
		platform_device_put(prs505_display_device);
		fb_unregister_client(&prs505_fb_notif);
		platform_device_unregister(&imxfb_device);
		return ret;
	}

	return 0;
}

static void __exit prs505_exit(void)
{
		platform_device_put(prs505_display_device);
		fb_unregister_client(&prs505_fb_notif);
		platform_device_unregister(&imxfb_device);
}


module_init(prs505_init);
module_exit(prs505_exit);

MODULE_DESCRIPTION("board driver for Sony PRS-505 book reader display controller");
MODULE_AUTHOR("Yauhen Kharuzhy");
MODULE_LICENSE("GPL");



[-- Attachment #3: Type: text/plain, Size: 79 bytes --]

------------------------------------------------------------------------------

[-- Attachment #4: Type: text/plain, Size: 182 bytes --]

_______________________________________________
Linux-fbdev-devel mailing list
Linux-fbdev-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/linux-fbdev-devel

Re: [jekhor@gmail.com: Metronome specifications]

[Yauhen Kharuzhy]

And another question: why IRQF_TRIGGER_FALLING is used in IRQ
configuration? As I understood, RDY signal is active with high level,
and using of IRQF_TRIGGER_RISING will be reasonable.

From am200epd.c:

static int am200_setup_irq(struct fb_info *info)
{
	int ret;

	ret = request_irq(IRQ_GPIO(RDY_GPIO_PIN), am200_handle_irq,
				IRQF_DISABLED|IRQF_TRIGGER_FALLING,
				"AM200", info->par);
	if (ret)
		dev_err(&am200_device->dev, "request_irq failed: %d\n", ret);

	return ret;
}

static irqreturn_t am200_handle_irq(int irq, void *dev_id)
{
	struct metronomefb_par *par = dev_id;

	wake_up_interruptible(&par->waitq);
	return IRQ_HANDLED;
}

static int am200_wait_event(struct metronomefb_par *par)
{
	return wait_event_timeout(par->waitq, gpio_get_value(RDY_GPIO_PIN), HZ);
}



-- 
Yauhen Kharuzhy		jekhor _at_ gmail.com
			JID: jek@jabber.ru

A: No
Q: Should I quote below my post?

------------------------------------------------------------------------------

Re: [jekhor@gmail.com: Metronome specifications]

[Yauhen Kharuzhy]

On Wed, Dec 31, 2008 at 02:59:36PM +0200, Yauhen Kharuzhy wrote:
 
> static int am200_wait_event(struct metronomefb_par *par)
> {
> 	return wait_event_timeout(par->waitq, gpio_get_value(RDY_GPIO_PIN), HZ);
> }

I also noticed error in your am200_wait_event() implementation:

metronomefb supposed that return value of met_wait_event() will be zero
on successful case, but wait_event_timeout() returns zero on timeout.

-- 
Yauhen Kharuzhy		jekhor _at_ gmail.com
			JID: jek@jabber.ru

A: No
Q: Should I quote below my post?

------------------------------------------------------------------------------

Re: [jekhor@gmail.com: Metronome specifications]

[Jaya Kumar]

On Wed, Dec 31, 2008 at 8:16 PM, Yauhen Kharuzhy <jekhor@gmail.com> wrote:
> On Tue, Dec 30, 2008 at 11:56:22PM +0800, Jaya Kumar wrote:
>> Hi Yauhen,
>>
>> My apologies for the delay in replying. As you can imagine, life is
>> busy at Christmas and New Year and trying to clean out my patch box.
>> Yes, I will be happy to help in any way that I can. I think it is
>> great that you're doing the work to get the prs-505 supported. I
>> unfortunately have no documentation that I can pass along but will be
>> happy to take a look at any code.
>>
>> Ok, in response to the technical questions, can you please post the
>> code for the board driver here. For the other people who also asked
>> similar questions about broadsheet and other controllers, please post
>> your questions and code here on the mailing list rather than to me
>> alone. The fbdev mailing list is the right place for all of us to work
>> together on this.
>
> OK. I am sorry for my impatience.

No, no, you musn't apologize. I know I am slow, especially in
answering hard questions. :-) I will try to do better. It is one of my
resolutions for 2009.

>
> I just wrote a "new style" (for 2.6.28 kernel) board driver using of
> the existing imxfb driver similar to you am200epd board driver. I found two
> problems:
> 1) the metronomefb driver initializes before imxfb driver and a
> kernel Oops caused in the setup_fb() function because the board.host_fbinfo is not
> initialized yet. For prevent this, I placed a call of my prs505_presetup_fb()
> function immediately after the fb_register_client() call.

Okay, I think I understand what you're saying and yes, that sounds
like a reasonable solution.

>
> 2) the LCD controller in IMX SoCs supports LCD panels with height <= 511 (9 bit in
> LCD_SIZE register). In the original Sony driver this height is 312. As I
> suppose, it is half of the real height because LCD controller uses 16 bit
> per pixel. I set yres in my prs505_fb_info to 622 / 2.

This sounds familiar to me. There's also an alignment restriction on
the the X width which I think I had to deal with on pxa255 because
LCCR1 required 16bpp and a minimum 32-pixel alignment which meant 416
=> 832. Are you having to do the same there or does iMX handle this
differently?

>
> Now metronomefb initializes but nothing happen on the screen.

Ok, does this mean the ready signal not reasserting is solved?

>
> I attached my board driver.
> Sony driver, if needed, can be found here:
> http://users.openinkpot.org/~jekhor/8track20070714.tgz

Ok, I will look at both of this.

Thanks,
jaya

------------------------------------------------------------------------------

Re: [jekhor@gmail.com: Metronome specifications]

[Jaya Kumar]

On Wed, Dec 31, 2008 at 10:16 PM, Yauhen Kharuzhy <jekhor@gmail.com> wrote:
> On Wed, Dec 31, 2008 at 02:59:36PM +0200, Yauhen Kharuzhy wrote:
>
>> static int am200_wait_event(struct metronomefb_par *par)
>> {
>>       return wait_event_timeout(par->waitq, gpio_get_value(RDY_GPIO_PIN), HZ);
>> }
>
> I also noticed error in your am200_wait_event() implementation:
>
> metronomefb supposed that return value of met_wait_event() will be zero
> on successful case, but wait_event_timeout() returns zero on timeout.
>

ah, ok, good catch, I had not noticed this mistake. I will fix this
asap or you can also submit a patch to do this if possible.

Thanks,
jaya

------------------------------------------------------------------------------