Skip to contentSkip to conversation

V2 NAND_SPL support for phycore imx31to Scott Wood

PATCH New Maxim Artamonov 2 participants 2 messages Jan 28, 2009

lore
Share & tools
lore
#0 lore raw

PATCH 1/1 V2 NAND_SPL support for phycore imx31to Scott Wood

Quoted 6 lines · In message on Friday 12 December 2008 04:55:49 Scott Wood wrote:

Review

Quoted 3 lines · Maxim Artamonov
(quoted diff) +3 -0
+#ifdef CONFIG_NAND_SPL
+/* somewhat macro to reduce bin size for CONFIG_NAND_SPL*/
+.macro FILLREGS begreg, val, count, step

Why not use this macro always?

Why not? But, I think this 'll lose clearity of code. Let's discuss this.

Quoted 20 lines
Quoted 5 lines
(quoted diff) +0 -0
@@ -32,6 +35,15 @@
 #include <version.h>
 .globl _start
 _start: b	reset

Review line 47

Quoted 10 lines
(quoted diff) +9 -0
+#ifdef CONFIG_NAND_SPL
+	nop
+	nop
+	nop
+	nop
+	nop
+	nop
+	nop
+#else
 #ifdef CONFIG_ONENAND_IPL

#elif

Do NAND and OneNAND really have different requirements here?

Quoted 10 lines

Review relocate: /* relocate U-Boot to RAM */ · line 172

Quoted 6 lines
(quoted diff) +1 -1
@@ -156,9 +169,9 @@ relocate:				/* relocate U-Boot to RAM	    */
 	adr	r0, _start		/* r0 <- current position of code   */
 	ldr	r1, _TEXT_BASE		/* test if we run from flash or RAM */
 	cmp	r0, r1			/* don't reloc during debug	    */
-#ifndef CONFIG_ONENAND_IPL
+#if !defined(CONFIG_ONENAND_IPL) && !defined(CONFIG_NAND_SPL)

Should define one symbol that is set if either of the above are set, to simplify the ifdefs.

I don't work with OneNAND flash and reference to "hang" is deeply in source tree, and we have only 2kB on
NAND in-place of 4kB on OneNand. Unhappily, I don't have anytime to learn about OneNand and to change its code.

Quoted 7 lines

Review relocate: /* relocate U-Boot to RAM */

Quoted 5 lines
(quoted diff) +4 -0
+#define NFC_BUF_SIZE            (*((volatile u16 *)(NFC_BASE_ADDR + 0xE00)))
+#define NFC_BUF_ADDR            (*((volatile u16 *)(NFC_BASE_ADDR + 0xE04)))
...
+#define NFC_CONFIG1             (*((volatile u16 *)(NFC_BASE_ADDR + 0xE1A)))
+#define NFC_CONFIG2             (*((volatile u16 *)(NFC_BASE_ADDR + 0xE1C)))

Use I/O accessors,

What you about, tell me please?

Quoted 3 lines

and move the NFC register layout to a
non-arch-specific header file (preferably as a struct), as it's shared with other chips.

Which chips? This is a system's integration task, but I wrote only for imx-31 phycore. I don't understand in this.

Quoted 18 lines
Quoted 5 lines

#define CONFIG_EXTRA_ENV_SETTINGS \

(quoted diff) +0 -3
-	"bootargs_base=setenv bootargs console=ttySMX0,115200\0"							\
-	"bootargs_nfs=setenv bootargs $(bootargs) root=/dev/nfs ip=dhcp nfsroot=$(serverip):$(nfsrootfs),v3,tcp\0"	\
-	"bootargs_flash=setenv bootargs $(bootargs) root=/dev/mtdblock2 rootfstype=jffs2"				\

Review

Quoted 10 lines
(quoted diff) +8 -1
-	"bootargs_mtd=setenv bootargs $(bootargs) $(mtdparts)"								\
+	"jffs2=root-pcm037.jffs2\0"											\
+	"uimage=uImage-pcm037\0"											\
+	"nfsrootfs=/nfsroot\0"												\
 	"bootcmd=run bootcmd_net\0"											\
+	"mtdids=nor0=phys_mapped_flash\0"										\
+	"bootargs_base=setenv bootargs console=ttymxc0,115200\0"							\
+	"bootargs_nfs=setenv bootargs $(bootargs) root=/dev/nfs ip=dhcp nfsroot=$(serverip):$(nfsrootfs),v3,tcp\0"	\
+	"bootargs_mtd=setenv bootargs $(bootargs) $(mtdparts)\0"							\
+	"bootargs_flash=setenv bootargs $(bootargs) root=/dev/mtdblock3 rootfstype=jffs2\0"				\

There are changes in here that don't seem to be related to nand boot.

Yes, it's so. I'll change code.

Quoted 14 lines

Review

Quoted 8 lines
(quoted diff) +8 -0
+static void mx31_wait_ready(void)
+{
+	while (1)
+		if (NFC_CONFIG2 & NFC_INT){
+				NFC_CONFIG2 = NFC_CONFIG2 & (~NFC_INT);/*int flag reset*/
+				break;
+		}
+}

There's an extra indentation level in the if-body.

Braces around the while-body. Space before and after braces (here and elsewhere).

Quoted 10 lines

Review

Quoted 7 lines
(quoted diff) +7 -0
+static void mx31_nand_init(void)
+{
+	/*unlocking RAM Buff*/
+	NFC_CONFIG = 0x2;
+
+	/*hardware ECC checking and correct*/
+	NFC_CONFIG1 = NFC_ECC_EN;

Space after /*, before */

It will apply.

Quoted 5 lines

Review

Quoted 3 lines
(quoted diff) +3 -0
+	if (mx31_nand_check_ecc ())
+		while (1){
+			break;};	/*!!!!!*/

What is this?

Ok, let's it be infinitive while-cycle.

Quoted 40 lines
Quoted 40 lines
(quoted diff) +39 -0
+	p1 = (u32 *)MAIN_AREA0;
+	p2 = (u32 *)buf;
+
+	/*main copy loop from NAND-buffer to SDRAM memory*/
+	for (i=0; i < (CFG_NAND_PAGE_SIZE / 4); i++){		
+		*p2 = *p1;
+		p1++;
+		p2++;
+	}
+	
+	p1 = (u32 *)SPARE_AREA0;
+
+	/*it is hardware specific code for 8-bit 512kB NAND-flash spare area*/
+	p1++;
+	a = *p1;
+	a = (a & 0x0000ff00) >> 8;
+
+	if (a != 0xff) /*bad block marker verify*/
+		return 1;/*potential bad block*/
+	
+	return 0;
+}
+
+static int nand_load(unsigned int from, unsigned int size, unsigned char *buf)
+{
+	int i, bb;
+	
+	mx31_nand_init ();
+	
+	/*conver from to page number*/
+	from = from / CFG_NAND_PAGE_SIZE;
+
+	i = 0;
+
+	while (i < (size/CFG_NAND_PAGE_SIZE)){
+		
+		if ((from*CFG_NAND_PAGE_SIZE) >= CFG_NAND_CHIP_SIZE)
+			return 2;/*memory segment violation*/
+
Quoted 8 lines

Review

Quoted 6 lines
(quoted diff) +6 -0
+		bb = mx31_read_page (from, buf);
+
+		/*checking first page of each block*/
+		/*if this page has bb marker, then skip whole block*/
+		if ((!(from % CFG_NAND_PAGE_COUNT)) && bb){
+			from = from + CFG_NAND_PAGE_COUNT;

CONFIG_NAND_PAGES_PER_BLOCK would probably be a better name.

All right!

with regards, Maxim

#1 lore raw

PATCH 1/1 V2 NAND_SPL support for phycore imx31to Scott Wood

Quoted 8 lines · Maxim Artamonov
Quoted 6 lines · In message on Friday 12 December 2008 04:55:49 Scott Wood wrote:

Review

Quoted 3 lines · Maxim Artamonov
(quoted diff) +3 -0
+#ifdef CONFIG_NAND_SPL
+/* somewhat macro to reduce bin size for CONFIG_NAND_SPL*/
+.macro FILLREGS begreg, val, count, step

Why not use this macro always?

Why not? But, I think this 'll lose clearity of code. Let's discuss this.

You want unnecessary ifdefs to *increase* clarity?

Quoted 8 lines
Quoted 7 lines

Review

Quoted 5 lines
(quoted diff) +4 -0
+#define NFC_BUF_SIZE            (*((volatile u16 *)(NFC_BASE_ADDR + 0xE00)))
+#define NFC_BUF_ADDR            (*((volatile u16 *)(NFC_BASE_ADDR + 0xE04)))
...
+#define NFC_CONFIG1             (*((volatile u16 *)(NFC_BASE_ADDR + 0xE1A)))
+#define NFC_CONFIG2             (*((volatile u16 *)(NFC_BASE_ADDR + 0xE1C)))

Use I/O accessors,

What you about, tell me please?

Use readw()/writew().

Quoted 5 lines
Quoted 3 lines

and move the NFC register layout to a
non-arch-specific header file (preferably as a struct), as it's shared with other chips.

Which chips?

There was a separate driver for the same NAND controller on 5xxx PowerPC posted.

Quoted 2 lines

This is a system's integration task, but I wrote only for imx-31 phycore.

That's what I'm complaining about. Please do not have tunnel vision such that you only see your own chip.

Quoted 6 lines
Quoted 5 lines

Review

Quoted 3 lines
(quoted diff) +3 -0
+	if (mx31_nand_check_ecc ())
+		while (1){
+			break;};	/*!!!!!*/

What is this?

Ok, let's it be infinitive while-cycle.

Except it isn't, due to the break.

In any case, when there's something unusual enough to produce a comment of /*!!!!!*/, it would be much more helpful to replace that comment with *words* that describe what's going through your head. :-)

-Scott