Nokia RX-51 support

* Hide function save_boot_params if CONFIG_SPL_BUILD is not defined (function do nothing)

* Same behaviour as in file arch/arm/cpu/armv7/omap4/lowlevel_init.S

* This allow to implement board specified function save_boot_params in board code

Signed-off-by Pali Roh?r <[email protected]>

---
Changes since original version:
- Fixed commit message

 arch/arm/cpu/armv7/omap3/lowlevel_init.S |    4 ++--
 1 files changed, 2 insertions(+), 2 deletions(-)

(quoted diff) +0 -0

b/arch/arm/cpu/armv7/omap3/lowlevel_init.S index 2f6930b..c42c5dd 100644

arch/arm/cpu/armv7/omap3/lowlevel_init.S +0 -0

@@ -35,15 +35,15 @@
 _TEXT_BASE:
 	.word	CONFIG_SYS_TEXT_BASE	/* sdram load addr from config.mk */
 

arch/arm/cpu/armv7/omap3/lowlevel_init.S +1 -0

+#endif
 
 .global omap3_gp_romcode_call
 omap3_gp_romcode_call:

Quoted 2 lines

* Hide function save_boot_params if CONFIG_SPL_BUILD is not defined (function do nothing)

What do you mean by this statement here ?

Quoted 1 line

Quoted 1 line

* Same behaviour as in file arch/arm/cpu/armv7/omap4/lowlevel_init.S

Eh?

Quoted 2 lines

* This allow to implement board specified function save_boot_params in board code

I see ... do you want to write overridable function? Make it a weak alias?

* Hide function save_boot_params if CONFIG_SPL_BUILD is not defined (function do nothing)

* Same behaviour as in file arch/arm/cpu/armv7/omap4/lowlevel_init.S

* This allow to implement board specified function save_boot_params in board code

Signed-off-by Pali Roh?r <[email protected]>

* Hide function save_boot_params if CONFIG_SPL_BUILD is not defined (function do nothing)

* Same behaviour as in file arch/arm/cpu/armv7/omap4/lowlevel_init.S
* This allow to implement board specified function save_boot_params in board code

Signed-off-by Pali Roh?r <[email protected]>

---
Changes since original version:
- Fixed commit message

 arch/arm/cpu/armv7/omap3/lowlevel_init.S |    4 ++--
 1 files changed, 2 insertions(+), 2 deletions(-)

(quoted diff) +0 -0

b/arch/arm/cpu/armv7/omap3/lowlevel_init.S index 2f6930b..c42c5dd 100644

arch/arm/cpu/armv7/omap3/lowlevel_init.S +1 -1

@@ -35,15 +35,15 @@
 _TEXT_BASE:
 	.word	CONFIG_SYS_TEXT_BASE	/* sdram load addr from config.mk */
 
+#ifdef CONFIG_SPL_BUILD
 .global save_boot_params
 save_boot_params:
-#ifdef CONFIG_SPL_BUILD

Quoted 35 lines · On Tuesday 24 January 2012 15:27:58 Pali Roh?r wrote:

* Hide function save_boot_params if CONFIG_SPL_BUILD is not defined (function do nothing)

* Same behaviour as in file arch/arm/cpu/armv7/omap4/lowlevel_init.S
* This allow to implement board specified function save_boot_params in board code

Signed-off-by Pali Roh?r <[email protected]>

---
Changes since original version:
? ?- Fixed commit message

?arch/arm/cpu/armv7/omap3/lowlevel_init.S | ? ?4 ++-- ?1 files changed, 2 insertions(+), 2 deletions(-)

(quoted diff) +0 -0

b/arch/arm/cpu/armv7/omap3/lowlevel_init.S index 2f6930b..c42c5dd 100644

arch/arm/cpu/armv7/omap3/lowlevel_init.S +0 -0

@@ -35,15 +35,15 @@

?_TEXT_BASE:
? ? ? .word ? CONFIG_SYS_TEXT_BASE ? ?/* sdram load addr from config.mk */

+#ifdef CONFIG_SPL_BUILD
?.global save_boot_params
?save_boot_params:
-#ifdef CONFIG_SPL_BUILD
? ? ? ldr ? ? r4, =omap3_boot_device
? ? ? ldr ? ? r5, [r0, #0x4]
? ? ? and ? ? r5, r5, #0xff
? ? ? str ? ? r5, [r4]
-#endif
? ? ? bx ? ? ?lr
+#endif

Quoted 2 lines

?.global omap3_gp_romcode_call
?omap3_gp_romcode_call:

Now I see that somebody commited this patch to u-boot:
http://git.denx.de/?p=u-
boot.git;a=commit;h=204705111ca10f52f62e1a02ba567d8fc33a6d1e

So I will not include it in new version.

Please, can you inform me when somebody commit my patches to git master?

Quoted 44 lines · On Tuesday 24 January 2012 15:27:58 Pali Roh?r wrote:

Quoted 38 lines

* Hide function save_boot_params if CONFIG_SPL_BUILD is not defined (function do nothing)

* Same behaviour as in file arch/arm/cpu/armv7/omap4/lowlevel_init.S
* This allow to implement board specified function save_boot_params in board code

Signed-off-by Pali Roh?r <[email protected]>

---
Changes since original version:
- Fixed commit message

 arch/arm/cpu/armv7/omap3/lowlevel_init.S |    4 ++--
 1 files changed, 2 insertions(+), 2 deletions(-)

(quoted diff) +0 -0

b/arch/arm/cpu/armv7/omap3/lowlevel_init.S index 2f6930b..c42c5dd 100644

arch/arm/cpu/armv7/omap3/lowlevel_init.S +2 -2

@@ -35,15 +35,15 @@
 _TEXT_BASE:
      .word   CONFIG_SYS_TEXT_BASE    /* sdram load addr from config.mk
 */
 
+#ifdef CONFIG_SPL_BUILD
 .global save_boot_params
 save_boot_params:
-#ifdef CONFIG_SPL_BUILD
      ldr     r4, =omap3_boot_device
      ldr     r5, [r0, #0x4]
      and     r5, r5, #0xff
      str     r5, [r4]
-#endif
      bx      lr
+#endif
 
 .global omap3_gp_romcode_call

Quoted 1 line

omap3_gp_romcode_call:

Now I see that somebody commited this patch to u-boot:
http://git.denx.de/?p=u-
boot.git;a=commit;h=204705111ca10f52f62e1a02ba567d8fc33a6d1e

Quoted 3 lines

So I will not include it in new version.

Please, can you inform me when somebody commit my patches to git master?

I did: http://patchwork.ozlabs.org/patch/137559/ :)

Quoted 47 lines · Pali Roh?r

Quoted 46 lines · On Tuesday 24 January 2012 15:27:58 Pali Roh?r wrote:

Quoted 38 lines

* Hide function save_boot_params if CONFIG_SPL_BUILD is not defined (function do nothing)

* Same behaviour as in file arch/arm/cpu/armv7/omap4/lowlevel_init.S
* This allow to implement board specified function save_boot_params in board code

Signed-off-by Pali Roh?r <[email protected]>

---
Changes since original version:
- Fixed commit message

 arch/arm/cpu/armv7/omap3/lowlevel_init.S |    4 ++--
 1 files changed, 2 insertions(+), 2 deletions(-)

(quoted diff) +0 -0

b/arch/arm/cpu/armv7/omap3/lowlevel_init.S index 2f6930b..c42c5dd 100644

arch/arm/cpu/armv7/omap3/lowlevel_init.S +2 -2

@@ -35,15 +35,15 @@
 _TEXT_BASE:
      .word   CONFIG_SYS_TEXT_BASE    /* sdram load addr from config.mk
 */
 
+#ifdef CONFIG_SPL_BUILD
 .global save_boot_params
 save_boot_params:
-#ifdef CONFIG_SPL_BUILD
      ldr     r4, =omap3_boot_device
      ldr     r5, [r0, #0x4]
      and     r5, r5, #0xff
      str     r5, [r4]
-#endif
      bx      lr
+#endif
 
 .global omap3_gp_romcode_call

Quoted 1 line

omap3_gp_romcode_call:

Now I see that somebody commited this patch to u-boot:
http://git.denx.de/?p=u-
boot.git;a=commit;h=204705111ca10f52f62e1a02ba567d8fc33a6d1e

So I will not include it in new version.

Quoted 3 lines

Quoted 1 line

Please, can you inform me when somebody commit my patches to git master?

I did: http://patchwork.ozlabs.org/patch/137559/ :)

Do we really want this hack-around? Maybe generating such bootargs by uboot in the first place (or the ability to actually source bootargs and adjust env accordingly) won't be too bad of a solution?

Quoted 60 lines

Quoted 60 lines · Pali Roh?r

Quoted 58 lines · On Tuesday 24 January 2012 15:27:58 Pali Roh?r wrote:

Quoted 40 lines

* Hide function save_boot_params if CONFIG_SPL_BUILD is not defined (function do nothing)

* Same behaviour as in file arch/arm/cpu/armv7/omap4/lowlevel_init.S
* This allow to implement board specified function save_boot_params in board code

Signed-off-by Pali Roh?r <[email protected]>

---

Changes since original version:
- Fixed commit message

 arch/arm/cpu/armv7/omap3/lowlevel_init.S |    4 ++--
 1 files changed, 2 insertions(+), 2 deletions(-)

(quoted diff) +0 -0

b/arch/arm/cpu/armv7/omap3/lowlevel_init.S index 2f6930b..c42c5dd 100644 --- a/arch/arm/cpu/armv7/omap3/lowlevel_init.S +++ b/arch/arm/cpu/armv7/omap3/lowlevel_init.S

(quoted diff) +1 -1

@@ -35,15 +35,15 @@
 
 _TEXT_BASE:
      .word   CONFIG_SYS_TEXT_BASE    /* sdram load addr from
      config.mk
 
 */
 
+#ifdef CONFIG_SPL_BUILD
 
 .global save_boot_params
 
 save_boot_params:
-#ifdef CONFIG_SPL_BUILD
 
      ldr     r4, =omap3_boot_device
      ldr     r5, [r0, #0x4]

      and     r5, r5, #0xff
      str     r5, [r4]

arch/arm/cpu/armv7/omap3/lowlevel_init.S line 49

Quoted 9 lines

arch/arm/cpu/armv7/omap3/lowlevel_init.S +1 -1

-#endif
 
      bx      lr
 
+#endif
 
 .global omap3_gp_romcode_call
 
 omap3_gp_romcode_call:

Now I see that somebody commited this patch to u-boot:
http://git.denx.de/?p=u-
boot.git;a=commit;h=204705111ca10f52f62e1a02ba567d8fc33a6d1e

So I will not include it in new version.

Please, can you inform me when somebody commit my patches to git master?

Quoted 5 lines

Quoted 1 line

I did: http://patchwork.ozlabs.org/patch/137559/ :)

Do we really want this hack-around? Maybe generating such bootargs by uboot in the first place (or the ability to actually source bootargs and adjust env accordingly) won't be too bad of a solution?

IMHO, yes. There is value in being able to use u-boot as a useful shim when we don't control 100% of the sequence.

This patch adapts the bootm command so that it can use an existing atags command set up by a previous bootloader.

If the environment variable
"atagaddr" is unset, bootm behaves as normal. If "atagaddr" is set, bootm will use atags address from environment variable and also append new boot args (if specified in u-boot). For example, if a previous boot loader already set up the atags struct at 0x80000100:

setenv atagaddr 0x80000100; bootm 0x80008000

Signed-off-by Pali Roh?r <[email protected]>

---
Changes since original version:
- Added info to README file
- Added local define CONFIG_SETUP_ANY_TAG
- Fixed compile warning
- Fixed commit message
- Check if atagaddr is not NULL

 README               |    2 ++
 arch/arm/lib/bootm.c |   39 ++++++++++++++++++++++++++++++++-------
 2 files changed, 34 insertions(+), 7 deletions(-)

README +0 -0

index 9d713e8..8c3b98a 100644
@@ -3584,6 +3584,8 @@ Some configuration options can be set using

Environment Variables.

List of environment variables (most likely not complete):

README +2 -0

+  atagaddr	- bootm will use ATAGs struct from specified address (arm only)
+
   baudrate	- see CONFIG_BAUDRATE
 
   bootdelay	- see CONFIG_BOOTDELAY

arch/arm/lib/bootm.c +8 -0

index afa0093..8366dd5 100644
@@ -93,6 +93,14 @@ static void announce_and_cleanup(void)
 	cleanup_before_linux();
 }
 
+#if defined(CONFIG_SETUP_MEMORY_TAGS) || \
+defined(CONFIG_CMDLINE_TAG) || \
+defined(CONFIG_INITRD_TAG) || \
+defined(CONFIG_SERIAL_TAG) || \
+defined(CONFIG_REVISION_TAG)
+#define CONFIG_SETUP_ANY_TAG
+#endif
+
 int do_bootm_linux(int flag, int argc, char *argv[], bootm_headers_t
 *images) {
 	bd_t	*bd = gd->bd;

@@ -125,12 +133,22 @@ int do_bootm_linux(int flag, int argc, char *argv[],

bootm_headers_t *images) debug ("## Transferring control to Linux (at address %08lx) ...\n", (ulong) kernel_entry);

arch/arm/lib/bootm.c +24 -7

-#if defined (CONFIG_SETUP_MEMORY_TAGS) || \
-    defined (CONFIG_CMDLINE_TAG) || \
-    defined (CONFIG_INITRD_TAG) || \
-    defined (CONFIG_SERIAL_TAG) || \
-    defined (CONFIG_REVISION_TAG)
-	setup_start_tag (bd);
+	s = getenv("atagaddr");
+	if (s)
+		bd->bi_boot_params = simple_strtoul(s, NULL, 16);
+
+	if (bd->bi_boot_params) {
+		printf("Using existing atags at %#lx\n", bd->bi_boot_params);
+
+		params = (struct tag *) bd->bi_boot_params;
+		while (params->hdr.size > 0)
+			params = tag_next(params);
+	} else {
+#ifdef CONFIG_SETUP_ANY_TAG
+		setup_start_tag(bd);
+#endif
+	}
+
 #ifdef CONFIG_SERIAL_TAG
 	setup_serial_tag (&params);
 #endif

@@ -147,8 +165,15 @@ int do_bootm_linux(int flag, int argc, char *argv[],
 bootm_headers_t *images) if (images->rd_start && images->rd_end)
 		setup_initrd_tag (bd, images->rd_start, images->rd_end);
 #endif
-	setup_end_tag(bd);
+
+	if (s) {
+		if (params->hdr.size > 0)
+			setup_end_tag(bd);
+	} else {
+#ifdef CONFIG_SETUP_ANY_TAG
+		setup_end_tag(bd);
 #endif
+	}
 
 	announce_and_cleanup();

Quoted 2 lines

This patch adapts the bootm command so that it can use an existing atags command set up by a previous bootloader.

Why do you need it?

Quoted 8 lines

If the environment variable
"atagaddr" is unset, bootm behaves as normal. If "atagaddr" is set, bootm will use atags address from environment variable and also append new boot args (if specified in u-boot). For example, if a previous boot loader already set up the atags struct at 0x80000100:

setenv atagaddr 0x80000100; bootm 0x80008000

Can't you set those atags correctly in uboot?

Quoted 4 lines

Quoted 2 lines

This patch adapts the bootm command so that it can use an existing atags command set up by a previous bootloader.

Why do you need it?

Default Nokia kernel has patches which using non-standard atags for reporting bootmode, bootreason, hw-rev, NOLO version, boot count, ...

Closed-source Maemo application need to know these properties (read via /proc/bootreason, ...) and will restart device if there is error.

So U-Boot must can reuse old atags from NOLO and tell it new kernel.

Quoted 7 lines

Quoted 7 lines

If the environment variable
"atagaddr" is unset, bootm behaves as normal. If "atagaddr" is set, bootm will use atags address from environment variable and also append new boot args (if specified in u-boot). For example, if a previous boot loader already set up the atags struct at 0x80000100:

Quoted 3 lines

Quoted 1 line

setenv atagaddr 0x80000100; bootm 0x80008000

Can't you set those atags correctly in uboot?

No, because atags are not static (e.g bootreason).

must can reuse? Anyway, that's nonsense, why won't u-boot be able to craft those atags?

Quoted 2 lines

must can reuse? Anyway, that's nonsense, why won't u-boot be able to craft those atags?

Maemo 5 (default OS in Nokia N900) cannot work if kernel does not have proper atags from NOLO. So if U-Boot does not reuse it, then Maemo will not boot...

Quoted 7 lines · On Wednesday 25 January 2012 21:55:53 Marek Vasut wrote:

Quoted 2 lines

must can reuse? Anyway, that's nonsense, why won't u-boot be able to craft those atags?

Maemo 5 (default OS in Nokia N900) cannot work if kernel does not have proper atags from NOLO. So if U-Boot does not reuse it, then Maemo will not boot...

U-Boot CAN GENERATE THEM

Quoted 4 lines

Quoted 4 lines · On Wednesday 25 January 2012 21:55:53 Marek Vasut wrote:

Quoted 2 lines

must can reuse? Anyway, that's nonsense, why won't u-boot be able to craft those atags?

Quoted 5 lines

Quoted 3 lines

Maemo 5 (default OS in Nokia N900) cannot work if kernel does not have proper atags from NOLO. So if U-Boot does not reuse it, then Maemo will not boot...

U-Boot CAN GENERATE THEM

Yes, you can parse atags from NOLO and then generate new from existing... But why? Better and easier is copy exising and use it.

Quoted 10 lines · On Wednesday 25 January 2012 22:28:38 Marek Vasut wrote:

Quoted 9 lines

Quoted 8 lines · On Wednesday 25 January 2012 21:55:53 Marek Vasut wrote:

Quoted 2 lines

must can reuse? Anyway, that's nonsense, why won't u-boot be able to craft those atags?

Maemo 5 (default OS in Nokia N900) cannot work if kernel does not have proper atags from NOLO. So if U-Boot does not reuse it, then Maemo will
not boot...

Quoted 4 lines

Quoted 1 line

U-Boot CAN GENERATE THEM

Yes, you can parse atags from NOLO and then generate new from existing... But why? Better and easier is copy exising and use it.

Because eventually, someone will replace nolo with u-boot and what then?

* Code copied from linux kernel 3.0.0 from file include/linux/i2c/twl.h

Signed-off-by Pali Roh?r <[email protected]>

---
Changes since original version:
- Fixed commit message

 include/twl4030.h |   98

+++++++++++++++++++++++++++++++++++++++++++++++++++++ 1 files changed, 98 insertions(+), 0 deletions(-)

include/twl4030.h +5 -0

index 9cd32ab..0c17f59 100644
@@ -151,6 +151,103 @@
 #define TWL4030_PM_MASTER_SW_EVENTS_DEVSLP		(1 << 1)
 #define TWL4030_PM_MASTER_SW_EVENTS_DEVOFF		(1 << 0)
 
+/* Power bus message definitions */
+
+/* The TWL4030/5030 splits its power-management resources (the various
+ * regulators, clock and reset lines) into 3 processor groups - P1, P2 and
+ * P3. These groups can then be configured to transition between sleep,

wait-on + * and active states by sending messages to the power bus. See Section 5.4.2 + * Power Resources of TWL4030 TRM
+ */

include/twl4030.h +90 -0

+
+/* Processor groups */
+#define DEV_GRP_NULL		0x0
+#define DEV_GRP_P1		0x1	/* P1: all OMAP devices */
+#define DEV_GRP_P2		0x2	/* P2: all Modem devices */
+#define DEV_GRP_P3		0x4	/* P3: all peripheral devices */
+
+/* Resource groups */
+#define RES_GRP_RES		0x0	/* Reserved */
+#define RES_GRP_PP		0x1	/* Power providers */
+#define RES_GRP_RC		0x2	/* Reset and control */
+#define RES_GRP_PP_RC		0x3
+#define RES_GRP_PR		0x4	/* Power references */
+#define RES_GRP_PP_PR		0x5
+#define RES_GRP_RC_PR		0x6
+#define RES_GRP_ALL		0x7	/* All resource groups */
+
+#define RES_TYPE2_R0		0x0
+
+#define RES_TYPE_ALL		0x7
+
+/* Resource states */
+#define RES_STATE_WRST		0xF
+#define RES_STATE_ACTIVE	0xE
+#define RES_STATE_SLEEP		0x8
+#define RES_STATE_OFF		0x0
+
+/* Power resources */
+
+/* Power providers */
+#define RES_VAUX1               1
+#define RES_VAUX2               2
+#define RES_VAUX3               3
+#define RES_VAUX4               4
+#define RES_VMMC1               5
+#define RES_VMMC2               6
+#define RES_VPLL1               7
+#define RES_VPLL2               8
+#define RES_VSIM                9
+#define RES_VDAC                10
+#define RES_VINTANA1            11
+#define RES_VINTANA2            12
+#define RES_VINTDIG             13
+#define RES_VIO                 14
+#define RES_VDD1                15
+#define RES_VDD2                16
+#define RES_VUSB_1V5            17
+#define RES_VUSB_1V8            18
+#define RES_VUSB_3V1            19
+#define RES_VUSBCP              20
+#define RES_REGEN               21
+/* Reset and control */
+#define RES_NRES_PWRON          22
+#define RES_CLKEN               23
+#define RES_SYSEN               24
+#define RES_HFCLKOUT            25
+#define RES_32KCLKOUT           26
+#define RES_RESET               27
+/* Power Reference */
+#define RES_Main_Ref            28
+
+#define TOTAL_RESOURCES		28
+/*
+ * Power Bus Message Format ... these can be sent individually by Linux,
+ * but are usually part of downloaded scripts that are run when various
+ * power events are triggered.
+ *
+ *  Broadcast Message (16 Bits):
+ *    DEV_GRP[15:13] MT[12]  RES_GRP[11:9]  RES_TYPE2[8:7] RES_TYPE[6:4]
+ *    RES_STATE[3:0]
+ *
+ *  Singular Message (16 Bits):
+ *    DEV_GRP[15:13] MT[12]  RES_ID[11:4]  RES_STATE[3:0]
+ */
+
+#define MSG_BROADCAST(devgrp, grp, type, type2, state) \
+	((devgrp) << 13 | 1 << 12 | (grp) << 9 | (type2) << 7 \
+	| (type) << 4 | (state))
+
+#define MSG_SINGULAR(devgrp, id, state) \
+	((devgrp) << 13 | 0 << 12 | (id) << 4 | (state))
+
+#define MSG_BROADCAST_ALL(devgrp, state) \
+	((devgrp) << 5 | (state))
+
+#define MSG_BROADCAST_REF MSG_BROADCAST_ALL
+#define MSG_BROADCAST_PROV MSG_BROADCAST_ALL
+#define MSG_BROADCAST__CLK_RST MSG_BROADCAST_ALL
+
 /* Power Managment Receiver */
 #define TWL4030_PM_RECEIVER_SC_CONFIG			0x5B
 #define TWL4030_PM_RECEIVER_SC_DETECT1			0x5C

@@ -311,6 +408,7 @@
 #define TWL4030_PM_RECEIVER_VDAC_VSEL_18		0x03
 #define TWL4030_PM_RECEIVER_VMMC1_VSEL_30		0x02
 #define TWL4030_PM_RECEIVER_VMMC1_VSEL_32		0x03
+#define TWL4030_PM_RECEIVER_VSIM_VSEL_18		0x03
 
 /* Device Selection in PM Receiver Module */
 #define TWL4030_PM_RECEIVER_DEV_GRP_P1			0x20

Quoted 1 line

* Code copied from linux kernel 3.0.0 from file include/linux/i2c/twl.h

Do you need them?

Quoted 1 line

* Code copied from linux kernel 3.0.0 from file include/linux/i2c/twl.h

Do you need them?

Also, as in other patches, I noticed you don't CC Tom Rini. Also, Ccing Igor Grinberg. Please keep the Cc in next inkarnation of the series!

Guys, can you also help review the rest of the series ?

* Ignore ANSI escape codes for moving cursor, which are generated by
keyboard

Signed-off-by Pali Roh?r <[email protected]>

---
Changes since original version:
- Fixed commit message

 common/main.c |   51 +++++++++++++++++++++++++++++++++++++++++++++++++++
 1 files changed, 51 insertions(+), 0 deletions(-)

common/main.c +0 -0

index e96c95a..e7b5516 100644
@@ -958,6 +958,7 @@ int readline_into_buffer (const char *const prompt,

char * buffer) int n = 0; /* buffer index

common/main.c +51 -0

 	int	plen = 0;			/* prompt length	*/
 	int	col;				/* output column cnt	*/
+	int	esc = 0;			/* ansi escape char	*/
 	char	c;
 
 	/* print prompt */

@@ -1024,7 +1025,57 @@ int readline_into_buffer (const char *const prompt,
 char * buffer) p=delete_char(p_buf, p, &col, &n, plen);
 			continue;
 
+		case '\e':			/* ANSI escape char	*/
+			esc = 1;
+			continue;
+
 		default:
+
+			/*
+			 * Check for ANSI escape chars
+			 */
+			if (esc == 0 && c == '\e') {
+				esc = 1;
+				continue;
+			} else if (esc == 1) {
+				if (c == '[') {
+					esc = 2;
+					continue;
+				}
+				if (n < CONFIG_SYS_CBSIZE-2) {
+					++n;
+					*p++ = '\e';
+					putc('\e');
+				}
+				esc = 0;
+			} else if (esc == 2 || esc == 3) {
+				if (esc == 2 && c == '1') {
+					esc = 3;
+					continue;
+				}
+				/* Ignore ANSI escape sequences */
+				/* generated by keyboard */
+				/* \e [ 1 A-D and \e [ A-D */
+				if (c >= 'A' && c <= 'D') {
+					esc = 0;
+					continue;
+				}
+				if (esc == 2 && n < CONFIG_SYS_CBSIZE-3) {
+					n += 2;
+					*p++ = '\e';
+					*p++ = '[';
+					puts("\e[");
+				} else if (esc == 3 &&
+						n < CONFIG_SYS_CBSIZE-4) {
+					n += 3;
+					*p++ = '\e';
+					*p++ = '[';
+					*p++ = '1';
+					puts("\e[1");
+				}
+				esc = 0;
+			}
+
 			/*
 			 * Must be a normal character then
 			 */

Quoted 2 lines

* Ignore ANSI escape codes for moving cursor, which are generated by
keyboard

I think WD should comment on this one.

* Use correct buffer size, do not damage screen output

Signed-off-by Pali Roh?r <[email protected]>

---
Changes since original version:
- Fixed commit message

 drivers/video/cfb_console.c |    2 +-
 1 files changed, 1 insertions(+), 1 deletions(-)

drivers/video/cfb_console.c +1 -1

index 904caf7..9092399 100644
@@ -701,7 +701,7 @@ static void console_scrollup(void)
 		);
 #else
 	memcpyl(CONSOLE_ROW_FIRST, CONSOLE_ROW_SECOND,
-		CONSOLE_SCROLL_SIZE >> 2);
+		CONSOLE_SCROLL_SIZE);
 #endif
 
 	/* clear the last one */

Quoted 1 line

* Use correct buffer size, do not damage screen output

What are the symptoms, how does this fix the issue?

Ccing Anatolij, I think he's the right one for this patch. Again, keep the correct CC!!!

* Use correct buffer size, do not damage screen output

Signed-off-by Pali Roh?r <[email protected]>

---
Changes since original version:
- Fixed commit message

 drivers/video/cfb_console.c |    2 +-
 1 files changed, 1 insertions(+), 1 deletions(-)

drivers/video/cfb_console.c +0 -0

index 904caf7..9092399 100644

Hi,

On Tue, 24 Jan 2012 15:28:02 +0100

Quoted 23 lines · Pali Roh?r

* Use correct buffer size, do not damage screen output

Signed-off-by Pali Roh?r <[email protected]>

---

Changes since original version:
- Fixed commit message

 drivers/video/cfb_console.c |    2 +-
 1 files changed, 1 insertions(+), 1 deletions(-)

drivers/video/cfb_console.c +0 -0

index 904caf7..9092399 100644
@@ -701,7 +701,7 @@ static void console_scrollup(void)
 
 		);
 
 #else
 
 	memcpyl(CONSOLE_ROW_FIRST, CONSOLE_ROW_SECOND,
 

drivers/video/cfb_console.c static void console_scrollup(void) · line 704

Quoted 2 lines

drivers/video/cfb_console.c +1 -1

-		CONSOLE_SCROLL_SIZE >> 2);
+		CONSOLE_SCROLL_SIZE);

NAK. This change is wrong. CONSOLE_SCROLL_SIZE is the size of the visible frame buffer - size of one row in bytes. We are using memcpyl() here, so the division by 4 (size >> 2) is correct. With your change we end up copying 4 times more data then needed.

Thanks,
Anatolij

Dear Anatolij Gustschin,

Quoted 28 lines

Hi,

On Tue, 24 Jan 2012 15:28:02 +0100

Quoted 23 lines · Pali Roh?r

* Use correct buffer size, do not damage screen output

Signed-off-by Pali Roh?r <[email protected]>

---

Changes since original version:
- Fixed commit message

 drivers/video/cfb_console.c |    2 +-
 1 files changed, 1 insertions(+), 1 deletions(-)

drivers/video/cfb_console.c +0 -0

index 904caf7..9092399 100644
@@ -701,7 +701,7 @@ static void console_scrollup(void)
 
 		);
 
 #else
 
 	memcpyl(CONSOLE_ROW_FIRST, CONSOLE_ROW_SECOND,
 

Quoted 7 lines

drivers/video/cfb_console.c static void console_scrollup(void) · line 704

Quoted 2 lines

drivers/video/cfb_console.c +1 -1

-		CONSOLE_SCROLL_SIZE >> 2);
+		CONSOLE_SCROLL_SIZE);

NAK. This change is wrong. CONSOLE_SCROLL_SIZE is the size of the visible frame buffer - size of one row in bytes. We are using memcpyl() here, so the division by 4 (size >> 2) is correct. With your change we end up copying 4 times more data then needed.

What kind of a problem are we fixing here? And what are the symptoms of it?

doesn't say much about the problem. If the GraphicDevice structure returned by video_hw_init() is setup correctly, the scrolling should be working fine. From the other patch [1] I can see that the structure

Hi Marek,

On Wed, 21 Mar 2012 11:20:38 +0100

Quoted 40 lines · Marek Vasut

Dear Anatolij Gustschin,

Quoted 28 lines

Hi,

On Tue, 24 Jan 2012 15:28:02 +0100

Quoted 23 lines · Pali Roh?r

* Use correct buffer size, do not damage screen output

Signed-off-by Pali Roh?r <[email protected]>

---

Changes since original version:
- Fixed commit message

 drivers/video/cfb_console.c |    2 +-
 1 files changed, 1 insertions(+), 1 deletions(-)

(quoted diff) +0 -0

b/drivers/video/cfb_console.c index 904caf7..9092399 100644

drivers/video/cfb_console.c +0 -0

@@ -701,7 +701,7 @@ static void console_scrollup(void)
 
 		);
 
 #else
 
 	memcpyl(CONSOLE_ROW_FIRST, CONSOLE_ROW_SECOND,
 

Quoted 9 lines

drivers/video/cfb_console.c static void console_scrollup(void) · line 704

Quoted 2 lines

drivers/video/cfb_console.c +1 -1

-		CONSOLE_SCROLL_SIZE >> 2);
+		CONSOLE_SCROLL_SIZE);

NAK. This change is wrong. CONSOLE_SCROLL_SIZE is the size of the
visible frame buffer - size of one row in bytes. We are using memcpyl() here, so the division by 4 (size >> 2) is correct. With your change we end up copying 4 times more data then needed.

Quoted 2 lines

What kind of a problem are we fixing here? And what are the symptoms of it?

Actually I'm not aware of any problem in current console_scrollup(). At least on two test setups with different framebuffer drivers and in one setup with HW accelerated scrolling I didn't see any problems with current code.

The description in the commit log of this patch:
"Use correct buffer size, do not damage screen output" doesn't say much about the problem. If the GraphicDevice structure returned by video_hw_init() is setup correctly, the scrolling should be working fine. From the other patch [1] I can see that the structure is setup as follows:

        /* fill in Graphic Device */
        gdev.frameAdrs = 0x8f9c0000;
        gdev.winSizeX = 800;
        gdev.winSizeY = 480;
        gdev.gdfBytesPP = 2;
        gdev.gdfIndex = GDF_16BIT_565RGB;
        memset((void *)gdev.frameAdrs, 0, 0xbb800);
        return (void *) &gdev;

Most likely using 0x8f9c0000 as framebuffer address is the first _big_ problem. The framebuffer address range is not allocated properly and could be used by malloc area. The board has 256 MB of RAM starting at 0x80000000, if U-Boot relocated itself to the upper RAM, the problems should be expected.

AFAIK the N900 port doesn't use a framebuffer driver but probably uses pre-initialized display controller configuration. This should be fixed first.

Thanks,
Anatolij

Hi, can you show me how to fix this? How to correctly use framebuffer?

On Wed, 21 Mar 2012 19:50:34 +0100
Pali Roh?r <[email protected]> wrote:
...

Quoted 2 lines

Hi, can you show me how to fix this? How to correctly use framebuffer?

Hi,

Not really. This would be a new project to rewrite U-Boot driver for OMAP3 DSS. There is a driver under drivers/video/omap3_dss.c, but it doesn't do what we need here. You can try to set gdev.frameAdrs to point to another address range in RAM which is not used by U-Boot. Then get OMAP3430 TRM and read DSS documentation how to configure the display controller so that is displays frame buffer data from this address range. Currently I do not have time nor resources to rewrite the existing omap3_dss driver.

Thanks,
Anatolij

Do you think that mapped address of framebuffer is only problem?

Why
then framebuffer output on n900 screen working without problem if I do NOT read framebuffer memory? Problem with garbaged display output seems happends only for read operations (console_scrollup, negation of pixels for cursor,...) but not for write-only operations (clear console, clear line, set black/white cursror, render fonsts...). I think if there is problem with address space, then garbaged screen should be also for framebuffer write operations...

* Use correct buffer size, do not damage screen output

Hi,

On Tue, 24 Jan 2012 15:28:02 +0100

Quoted 1 line · Pali Roh?r

* Use correct buffer size, do not damage screen output

It seems that I can reproduce similar problem on the beagleboard now. Will try to debug and find a solution for it.

Thanks,
Anatolij

Quoted 4 lines · On Thursday 26 April 2012 23:45:52 Anatolij Gustschin wrote:

Hi,

On Tue, 24 Jan 2012 15:28:02 +0100

Quoted 5 lines

Quoted 1 line · Pali Roh?r

* Use correct buffer size, do not damage screen output

It seems that I can reproduce similar problem on the beagleboard now. Will try to debug and find a solution for it.

Ok, thanks! I played a lot of with u-boot on nokia n900, but I did not find reason...

On Thu, 26 Apr 2012 23:50:57 +0200

Quoted 8 lines · Pali Roh?r

Quoted 7 lines · On Thursday 26 April 2012 23:45:52 Anatolij Gustschin wrote:

Hi,

On Tue, 24 Jan 2012 15:28:02 +0100

Quoted 1 line · Pali Roh?r

* Use correct buffer size, do not damage screen output

Quoted 7 lines

Quoted 3 lines

It seems that I can reproduce similar problem on the beagleboard now. Will try to debug and find a solution for it.

Ok, thanks! I played a lot of with u-boot on nokia n900, but I
did not find reason...

Probably the issue is caused by cached frame buffer data. This
[1] patch is supposed to fix it. Could you please test it on N900? Additional cache flushing also needs to be added to your extensions for cfb_console to fix the menu output.

Thanks,
Anatolij

[1] http://patchwork.ozlabs.org/patch/155662/

* console_clear - clear full console framebuffer output
* console_clear_line - clear part of specified line (or full)

Signed-off-by Pali Roh?r <[email protected]>

---
Changes since original version:
- Fixed commit message

 drivers/video/cfb_console.c |   64

+++++++++++++++++++++++++++++++++++-------- 1 files changed, 52 insertions(+), 12 deletions(-)

drivers/video/cfb_console.c +0 -0

index 9092399..b74d6d0 100644

drivers/video/cfb_console.c +51 -12

+{
+#ifdef VIDEO_HW_RECTFILL
+	video_hw_rectfill(VIDEO_PIXEL_SIZE,	/* bytes per pixel */
+			  0,			/* dest pos x */
+			  video_logo_height,	/* dest pos y */
+			  VIDEO_VISIBLE_COLS,	/* frame width */
+			  VIDEO_VISIBLE_ROWS,	/* frame height */
+			  bgx			/* fill color */
+	);
+#else
+	memsetl(CONSOLE_ROW_FIRST, CONSOLE_SIZE, bgx);
+#endif
+}
+
+static void console_clear_line(int line, int begin, int end)
+{
+#ifdef VIDEO_HW_RECTFILL
+	video_hw_rectfill(VIDEO_PIXEL_SIZE,		/* bytes per pixel */
+			  /* FIXME: correct? */
+			  VIDEO_FONT_WIDTH * begin,	/* dest pos x */
+			  /* FIXME: correct? */
+			  video_logo_height +
+			  CONSOLE_ROW_SIZE * line,	/* dest pos y */
+			  /* FIXME: correct? */
+			  VIDEO_FONT_WIDTH * (end - begin), /* frame width */
+			  VIDEO_FONT_HEIGHT,		/* frame height */
+			  bgx				/* fill color */
+		);
+#else
+	int i;
+	if (begin == 0 && end == CONSOLE_COLS)
+		memsetl(CONSOLE_ROW_FIRST +
+			CONSOLE_ROW_SIZE * line,	/* offset of row */
+			CONSOLE_ROW_SIZE >> 2,		/* length of row */
+			bgx				/* fill color */
+		);
+	else
+		for (i = 0; i < VIDEO_FONT_HEIGHT; ++i)
+			memsetl(CONSOLE_ROW_FIRST +
+				CONSOLE_ROW_SIZE * line + /* offset of row */
+				VIDEO_FONT_WIDTH *
+				VIDEO_PIXEL_SIZE * begin + /* offset of col */
+				i * VIDEO_LINE_LEN, /* col offset of i line */
+				(VIDEO_FONT_WIDTH * VIDEO_PIXEL_SIZE *
+				(end - begin + 1)) >> 2, /* length to end */
+				bgx			/* fill color */
+				);
+#endif
+}
+
 static void console_scrollup(void)
 {
 	/* copy up rows ignoring the first one */

@@ -705,18 +756,7 @@ static void console_scrollup(void)
 #endif
 
 	/* clear the last one */
-#ifdef VIDEO_HW_RECTFILL
-	video_hw_rectfill(VIDEO_PIXEL_SIZE,	/* bytes per pixel */
-			  0,			/* dest pos x */
-			  VIDEO_VISIBLE_ROWS
-			  - VIDEO_FONT_HEIGHT,	/* dest pos y */
-			  VIDEO_VISIBLE_COLS,	/* frame width */
-			  VIDEO_FONT_HEIGHT,	/* frame height */
-			  CONSOLE_BG_COL	/* fill color */
-		);
-#else
-	memsetl(CONSOLE_ROW_LAST, CONSOLE_ROW_SIZE >> 2, CONSOLE_BG_COL);
-#endif
+	console_clear_line(CONSOLE_ROWS-1, 0, CONSOLE_COLS);
 }
 
 static void console_back(void)

Quoted 2 lines

+static void console_clear(void)

Do you use this function AT ALL?

Quoted 3 lines

Quoted 1 line

+static void console_clear(void)

Do you use this function AT ALL?

Yes, in patch which add ANSI terminal support to cfb_console.

* Command can be enabled by CONFIG_CMD_CLEAR
* Added some ANSI escape codes definitions in common.h

Signed-off-by Pali Roh?r <[email protected]>

Cc Marcel Mol <[email protected]>

---
Changes since original version:
- Renamed command clr to clear
- Use puts instead printf
- Move cursor to pos1,1
- Merged parts of patch "Add some ANSI escape codes definitions in
common.h"

 common/Makefile          |    1 +
 common/cmd_clear.c       |   42 ++++++++++++++++++++++++++++++++++++++++++
 include/common.h         |    7 +++++++
 include/config_cmd_all.h |    1 +
 4 files changed, 51 insertions(+), 0 deletions(-)
 create mode 100644 common/cmd_clear.c

common/Makefile +1 -0

index 2d9ae8c..e1efd45 100644
@@ -69,6 +69,7 @@ COBJS-$(CONFIG_CMD_BEDBUG) += bedbug.o cmd_bedbug.o
 COBJS-$(CONFIG_CMD_BMP) += cmd_bmp.o
 COBJS-$(CONFIG_CMD_BOOTLDR) += cmd_bootldr.o
 COBJS-$(CONFIG_CMD_CACHE) += cmd_cache.o
+COBJS-$(CONFIG_CMD_CLEAR) += cmd_clear.o
 COBJS-$(CONFIG_CMD_CONSOLE) += cmd_console.o
 COBJS-$(CONFIG_CMD_CPLBINFO) += cmd_cplbinfo.o
 COBJS-$(CONFIG_DATAFLASH_MMC_SELECT) += cmd_dataflash_mmc_mux.o

common/cmd_clear.c new file +30 -0

new file mode 100644
index 0000000..597611e
@@ -0,0 +1,42 @@
+/*
+ * Copyright 2011
+ * Marcel Mol, MESA Consulting, marcel at mesa.nl
+ *
+ * Copyright 2011
+ * Pali Roh?r, pali.rohar at gmail.com
+ *
+ * See file CREDITS for list of people who contributed to this
+ * project.
+ *
+ * This program is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU General Public License as
+ * published by the Free Software Foundation; either version 2 of
+ * the License, or (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 59 Temple Place, Suite 330, Boston,
+ * MA 02111-1307 USA
+ */
+
+#include <common.h>
+#include <command.h>
+
+static int do_clear(cmd_tbl_t *cmdtp, int flag, int argc, char * const

argv[]) +{

common/cmd_clear.c new file +11 -0

+	puts(ANSI_CLEAR_CONSOLE);
+	printf(ANSI_CURSOR_POSITION, 1, 1);
+	return 0;
+}
+
+U_BOOT_CMD(
+	clear,    CONFIG_SYS_MAXARGS,     1,      do_clear,
+	"clear",
+	"\n"
+	"    - clear screen and move cursor to top of screen"
+);

include/common.h +0 -0

index 3df1def..9c0449e 100644
@@ -751,6 +751,13 @@ void	clear_ctrlc (void);	/* clear the Control-C

condition */ int disable_ctrlc (int); /* 1 to disable, 0 to enable
Control-C detect */

/*
+ * ANSI terminal
+ */

include/common.h +5 -0

+
+#define ANSI_CURSOR_POSITION		"\e[%d;%dH"
+#define ANSI_CLEAR_CONSOLE		"\e[2J"
+
+/*
  * STDIO based functions (can always be used)
  */
 /* serial stuff */

include/config_cmd_all.h +1 -0

index 9716f9c..3f25eba 100644
@@ -23,6 +23,7 @@
 #define CONFIG_CMD_BSP		/* Board Specific functions	*/
 #define CONFIG_CMD_CACHE	/* icache, dcache		*/
 #define CONFIG_CMD_CDP		/* Cisco Discovery Protocol	*/
+#define CONFIG_CMD_CLEAR	/* ANSI clear screen command	*/
 #define CONFIG_CMD_CONSOLE	/* coninfo			*/
 #define CONFIG_CMD_DATE		/* support for RTC, date/time...*/
 #define CONFIG_CMD_DHCP		/* DHCP Support			*/

Do you seriously need a new command for such a simple operation?

Quoted 1 line

Do you seriously need a new command for such a simple operation?

Yes, command clear is really usefull for debugging framebuffer output on n900.

Quoted 5 lines · On Wednesday 25 January 2012 19:10:03 Marek Vasut wrote:

Quoted 1 line

Do you seriously need a new command for such a simple operation?

Yes, command clear is really usefull for debugging framebuffer output on n900.

Ok, let's see what the other people think, I'm opposed tho

* If not defined CONFIG_MENUCMD do nothing

* If CONFIG_MENUKEY is 0 and was pressed any key run env "menu_cmd"
* If pressed key was CONFIG_MENUKEY run env "menu_cmd"
* If CONFIG_MENUKEY is not defined run env "menu_cmd" always

* CONFIG_MENUKEY working if defined CONFIG_MENUCMD and CONFIG_BOOTDELAY >=
0

Signed-off-by Pali Roh?r <[email protected]>

---
Changes since original version:
- Fixed commit message

 common/main.c |   34 +++++++++++++++++++++++-----------
 1 files changed, 23 insertions(+), 11 deletions(-)

common/main.c +3 -0

index e7b5516..503d6c4 100644
@@ -82,6 +82,20 @@ int do_mdm_init = 0;
 extern void mdm_init(void); /* defined in board.c */
 #endif
 
+#if defined(CONFIG_MENUKEY) && !defined(CONFIG_MENUCMD) && \
+(!defined(CONFIG_BOOTDELAY) || CONFIG_BOOTDELAY < 0)
+#error CONFIG_MENUKEY defined, but not CONFIG_MENUCMD or CONFIG_BOOTDELAY

Quoted 1 line

= 0 +#error define CONFIG_MENUCMD and CONFIG_BOOTDELAY too

common/main.c +10 -0

+#endif
+
+#ifdef CONFIG_MENUCMD
+# ifdef CONFIG_MENUKEY
+static int menucmd;
+# else
+static int menucmd = 1;
+# endif
+#endif
+
 /*************************************************************************

** * Watch for 'delay' seconds for autoboot stop or autoboot delay string.
* returns: 0 - no key string, allow autoboot 1 - got key string, abort @@ -201,10 +215,6 @@ int abortboot(int bootdelay)

# else /* !defined(CONFIG_AUTOBOOT_KEYED) */

common/main.c +9 -11

-#ifdef CONFIG_MENUKEY
-static int menukey = 0;
-#endif
-
 #ifndef CONFIG_MENU
 static inline
 #endif

@@ -241,8 +251,10 @@ int abortboot(int bootdelay)
 			if (tstc()) {	/* we got a key press	*/
 				abort  = 1;	/* don't auto boot	*/
 				bootdelay = 0;	/* no more delay	*/
-# ifdef CONFIG_MENUKEY
-				menukey = getc();
+# if defined(CONFIG_MENUCMD) && defined(CONFIG_MENUKEY)
+				if (CONFIG_MENUKEY == 0 ||
+					CONFIG_MENUKEY == getc())
+					menucmd = 1;
 # else
 				(void) getc();  /* consume input	*/
 # endif

@@ -292,6 +304,7 @@ int run_command2(const char *cmd, int flag)
 
 void main_loop (void)
 {
+	char *s;
 #ifndef CONFIG_SYS_HUSH_PARSER
 	static char lastcommand[CONFIG_SYS_CBSIZE] = { 0, };
 	int len;

@@ -300,7 +313,6 @@ void main_loop (void)
 #endif
 
 #if defined(CONFIG_BOOTDELAY) && (CONFIG_BOOTDELAY >= 0)
-	char *s;
 	int bootdelay;
 #endif
 #ifdef CONFIG_PREBOOT

@@ -405,15 +417,15 @@ void main_loop (void)
 		disable_ctrlc(prev);	/* restore Control C checking */
 # endif
 	}
+#endif /* CONFIG_BOOTDELAY */
 
-# ifdef CONFIG_MENUKEY
-	if (menukey == CONFIG_MENUKEY) {
+#ifdef CONFIG_MENUCMD
+	if (menucmd == 1) {
 		s = getenv("menucmd");
 		if (s)
 			run_command2(s, 0);
 	}
-#endif /* CONFIG_MENUKEY */
-#endif /* CONFIG_BOOTDELAY */
+#endif
 
 	/*
 	 * Main Loop for Monitor Command Processing

So this is just to interrupt autoboot? Why not just report the "menukey" as any other keypress and do it that way?

Quoted 2 lines

So this is just to interrupt autoboot? Why not just report the "menukey" as any other keypress and do it that way?

Usefull for interrupt autoboot and run some menu (script/booting) only with specific key (for example with key M on device keyboard).

Quoted 7 lines · On Wednesday 25 January 2012 19:12:29 Marek Vasut wrote:

Quoted 2 lines

So this is just to interrupt autoboot? Why not just report the "menukey" as any other keypress and do it that way?

Usefull for interrupt autoboot and run some menu (script/booting) only with
specific key (for example with key M on device keyboard).

what's the difference between this and CONFIG_MENUKEY?

* if defined run env "premonitor" before Main Loop for Monitor Processing

Signed-off-by Pali Roh?r <[email protected]>

---
Changes since original version:
- removed #ifdef CONFIG_PREMONITOR in main.c

 common/env_common.c |    3 +++
 common/main.c       |    4 ++++
 2 files changed, 7 insertions(+), 0 deletions(-)

common/env_common.c +3 -0

index 71811c4..1c2360d 100644
@@ -92,6 +92,9 @@ const uchar default_environment[] = {
 #ifdef	CONFIG_PREBOOT
 	"preboot="	CONFIG_PREBOOT			"\0"
 #endif
+#ifdef	CONFIG_PREMONITOR
+	"premonitor="	CONFIG_PREMONITOR		"\0"
+#endif
 #ifdef	CONFIG_ROOTPATH
 	"rootpath="	CONFIG_ROOTPATH			"\0"
 #endif

common/main.c +4 -0

index 503d6c4..e0c262c 100644
@@ -427,6 +427,10 @@ void main_loop (void)
 	}
 #endif
 
+	s = getenv("premonitor");
+	if (s)
+		run_command2(s, 0);
+
 	/*
 	 * Main Loop for Monitor Command Processing
 	 */

Quoted 1 line

* if defined run env "premonitor" before Main Loop for Monitor Processing

What is this good for?

Quoted 4 lines

Quoted 2 lines

* if defined run env "premonitor" before Main Loop for Monitor Processing

What is this good for?

before running monitor loop I'd like write some info to console (specific board info, board commands, board macros). but if these info will be written in preboot it can be deleted by menu section.

Quoted 4 lines · On Wednesday 25 January 2012 19:12:51 Marek Vasut wrote:

Quoted 3 lines

Quoted 2 lines

* if defined run env "premonitor" before Main Loop for Monitor Processing

Quoted 5 lines

Quoted 1 line

What is this good for?

before running monitor loop I'd like write some info to console (specific board info, board commands, board macros). but if these info will be written in preboot it can be deleted by menu section.

The "menu" here, you refer to the thing you implemented? So fix that instead of introducing broken code and a pile of workarounds.

Quoted 4 lines

Quoted 4 lines · On Wednesday 25 January 2012 19:12:51 Marek Vasut wrote:

Quoted 3 lines

Quoted 2 lines

* if defined run env "premonitor" before Main Loop for Monitor Processing

Quoted 8 lines

Quoted 5 lines

Quoted 1 line

What is this good for?

before running monitor loop I'd like write some info to console (specific board info, board commands, board macros). but if these info will be written in preboot it can be deleted by menu section.

The "menu" here, you refer to the thing you implemented? So fix that instead of introducing broken code and a pile of workarounds.

With "menu" I mean env menucmd. U-Boot run command stored in env menucmd after preboot and before monitor loop. For example in menucmd can be my bootmenu, but there are no problems - no broken code! My bootmenu clear screan and show menu, so anything printed before menu entries will be deleted. This is not bug - Also in Grub or LILO is debug/output log cleaned when showing menu entries.

Quoted 14 lines · On Wednesday 25 January 2012 21:51:38 Marek Vasut wrote:

Quoted 4 lines

Quoted 4 lines · On Wednesday 25 January 2012 19:12:51 Marek Vasut wrote:

Quoted 3 lines

Quoted 2 lines

* if defined run env "premonitor" before Main Loop for Monitor Processing

Quoted 8 lines

Quoted 5 lines

Quoted 1 line

What is this good for?

before running monitor loop I'd like write some info to console (specific board info, board commands, board macros). but if these info will be written in preboot it can be deleted by menu section.

The "menu" here, you refer to the thing you implemented? So fix that instead of introducing broken code and a pile of workarounds.

Quoted 6 lines

With "menu" I mean env menucmd. U-Boot run command stored in env menucmd after preboot and before monitor loop. For example in menucmd can be my bootmenu, but there are no problems - no broken code! My bootmenu clear screan and show menu, so anything printed before menu entries will be deleted. This is not bug - Also in Grub or LILO is debug/output log cleaned when showing menu entries.

Ok anyway, but you implemented custom menu system anyway ... so let's clear this first, is common/menu.c useless for your application?

Quoted 21 lines

Quoted 20 lines · On Wednesday 25 January 2012 21:51:38 Marek Vasut wrote:

Quoted 4 lines

Quoted 4 lines · On Wednesday 25 January 2012 19:12:51 Marek Vasut wrote:

Quoted 3 lines

Quoted 2 lines

* if defined run env "premonitor" before Main Loop for Monitor Processing

Quoted 8 lines

Quoted 5 lines

Quoted 1 line

What is this good for?

before running monitor loop I'd like write some info to console (specific board info, board commands, board macros). but if these info will be written in preboot it can be deleted by menu section.

The "menu" here, you refer to the thing you implemented? So fix that instead of introducing broken code and a pile of workarounds.

With "menu" I mean env menucmd. U-Boot run command stored in env menucmd after preboot and before monitor loop. For example in menucmd can be my bootmenu, but there are no problems - no broken code! My bootmenu clear screan and show menu, so anything printed before menu entries will be deleted. This is not bug - Also in Grub or LILO is debug/output log cleaned when showing menu entries.

Quoted 2 lines

Ok anyway, but you implemented custom menu system anyway ... so let's clear this first, is common/menu.c useless for your application?

I do not know. I created my bootmenu before menu.c was integrated into u-boot. I did not looked at menu.c yet - I do not know what is in menu.c implemented.

* Added some ANSI escape codes definitions in common.h
* Configuration is done via env variables bootmenu_delay and
bootmenu_<num>:

bootmenu_delay=<delay>
bootmenu_<num>="<title>=<commands>"

(title and commands are separated by first char '=')

<delay> is delay in seconds of autobooting first entry <num> is boot menu entry, starting from zero
<title> is text shown in boot screen
<commands> are commands which will be executed when menu entry is
selected

Control-C detect */ * ANSI terminal
*/

include/common.h +13 -0

+#define ANSI_CURSOR_UP			"\e[%dA"
+#define ANSI_CURSOR_DOWN		"\e[%dB"
+#define ANSI_CURSOR_FORWARD		"\e[%dC"
+#define ANSI_CURSOR_BACK		"\e[%dD"
+#define ANSI_CURSOR_NEXTLINE		"\e[%dE"
+#define ANSI_CURSOR_PREVIOUSLINE	"\e[%dF"
+#define ANSI_CURSOR_COLUMN		"\e[%dG"
 #define ANSI_CURSOR_POSITION		"\e[%d;%dH"
+#define ANSI_CURSOR_SHOW		"\e[?25h"
+#define ANSI_CURSOR_HIDE		"\e[?25l"
 #define ANSI_CLEAR_CONSOLE		"\e[2J"
+#define ANSI_CLEAR_LINE_TO_END		"\e[0K"
+#define ANSI_CLEAR_LINE			"\e[2K"
+#define ANSI_COLOR_RESET		"\e[0m"
+#define ANSI_COLOR_REVERSE		"\e[7m"
 
 /*
  * STDIO based functions (can always be used)

include/config_cmd_all.h +1 -0

index 3f25eba..8c0a648 100644
@@ -20,6 +20,7 @@
 #define CONFIG_CMD_BEDBUG	/* Include BedBug Debugger	*/
 #define CONFIG_CMD_BMP		/* BMP support			*/
 #define CONFIG_CMD_BOOTD	/* bootd			*/
+#define CONFIG_CMD_BOOTMENU	/* ANSI terminal Boot Menu	*/
 #define CONFIG_CMD_BSP		/* Board Specific functions	*/
 #define CONFIG_CMD_CACHE	/* icache, dcache		*/
 #define CONFIG_CMD_CDP		/* Cisco Discovery Protocol	*/

Do you need this command if you already have common/menu.c?