[U-Boot] [PATCH] at91: Add support for taskit AT91SAM9G20 boards.
Andreas Bießmann
andreas.devel at googlemail.com
Wed Aug 1 11:58:22 CEST 2012
Dear Markus Hubig,
first of all:
---8<---
andreas at andreas-mbp % ./tools/checkpatch.pl
U-Boot-at91-Add-support-for-taskit-AT91SAM9G20-boards..patch
WARNING: Whitespace before semicolon
#214: FILE: board/taskit/stamp9g20/stamp9g20.c:123:
+ ;
total: 0 errors, 1 warnings, 484 lines checked
NOTE: Ignored message types: COMPLEX_MACRO CONSIDER_KSTRTO MINMAX
MULTISTATEMENT_MACRO_USE_DO_WHILE
U-Boot-at91-Add-support-for-taskit-AT91SAM9G20-boards..patch has style
problems, please review.
If any of these errors are false positives, please report
them to the maintainer, see CHECKPATCH in MAINTAINERS.
--->8---
I know this part is copied from other at91 boards but I wonder how we
should handle these while-loops without content.
On 30.07.12 20:01, Markus Hubig wrote:
> This adds support for the AT91SAM9G20 boards by taskit GmbH.
> Both boards, Stamp9G20 and PortuxG20, are integrated in one
> file. PortuxG20 is basically a SBC built around the Stamp9G20.
>
> Signed-off-by: Markus Hubig <mhubig at imko.de>
> Cc: Andreas Bießmann <andreas.devel at googlemail.com>
> ---
> board/taskit/stamp9g20/Makefile | 52 ++++++++
> board/taskit/stamp9g20/stamp9g20.c | 199 +++++++++++++++++++++++++++++++
> boards.cfg | 2 +
> include/configs/stamp9g20.h | 225 ++++++++++++++++++++++++++++++++++++
MAINTAINER entry is missing
> 4 files changed, 478 insertions(+), 0 deletions(-)
> create mode 100644 board/taskit/stamp9g20/Makefile
> create mode 100644 board/taskit/stamp9g20/stamp9g20.c
> create mode 100644 include/configs/stamp9g20.h
>
> diff --git a/board/taskit/stamp9g20/Makefile b/board/taskit/stamp9g20/Makefile
> new file mode 100644
> index 0000000..e99bfaa
> --- /dev/null
> +++ b/board/taskit/stamp9g20/Makefile
> @@ -0,0 +1,52 @@
> +#
> +# (C) Copyright 2003-2008
> +# Wolfgang Denk, DENX Software Engineering, wd at denx.de.
> +#
> +# (C) Copyright 2008
> +# Stelian Pop <stelian at popies.net>
> +# Lead Tech Design <www.leadtechdesign.com>
> +#
> +# (C) Copyright 2012
> +# Markus Hubig <mhubig at imko.de>
> +# IMKO GmbH <www.imko.de>
> +#
> +# 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 $(TOPDIR)/config.mk
> +
> +LIB = $(obj)lib$(BOARD).o
> +
> +COBJS-y += stamp9g20.o
> +
> +SRCS := $(SOBJS:.o=.S) $(COBJS-y:.o=.c)
> +OBJS := $(addprefix $(obj),$(COBJS-y))
> +SOBJS := $(addprefix $(obj),$(SOBJS))
> +
> +$(LIB): $(obj).depend $(OBJS) $(SOBJS)
> + $(call cmd_link_o_target, $(OBJS) $(SOBJS))
> +
> +#########################################################################
> +
> +# defines $(obj).depend target
> +include $(SRCTREE)/rules.mk
> +
> +sinclude $(obj).depend
> +
> +#########################################################################
> diff --git a/board/taskit/stamp9g20/stamp9g20.c b/board/taskit/stamp9g20/stamp9g20.c
> new file mode 100644
> index 0000000..b87de51
> --- /dev/null
> +++ b/board/taskit/stamp9g20/stamp9g20.c
> @@ -0,0 +1,199 @@
> +/*
> + * (C) Copyright 2007-2008
> + * Stelian Pop <stelian at popies.net>
> + * Lead Tech Design <www.leadtechdesign.com>
> + *
> + * Achim Ehrlich <aehrlich at taskit.de>
> + * taskit GmbH <www.taskit.de>
> + *
> + * (C) Copyright 2012-
> + * Markus Hubig <mhubig at imko.de>
> + * IMKO GmbH <www.imko.de>
> + *
> + * 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 <asm/io.h>
> +#include <asm/arch/at91sam9260_matrix.h>
> +#include <asm/arch/at91sam9_smc.h>
> +#include <asm/arch/at91_common.h>
> +#include <asm/arch/at91_pmc.h>
> +#include <asm/arch/at91_rstc.h>
> +#include <asm/arch/gpio.h>
> +
> +#ifdef CONFIG_MACB
> +# include <net.h>
> +# include <netdev.h>
> +#endif /* CONFIG_MACB */
> +
> +DECLARE_GLOBAL_DATA_PTR;
> +
> +static void stamp9G20_nand_hw_init(void)
> +{
> + struct at91_smc *smc = (struct at91_smc *)ATMEL_BASE_SMC;
> + struct at91_matrix *matrix = (struct at91_matrix *)ATMEL_BASE_MATRIX;
> + unsigned long csa;
> +
> + /* Assign CS3 to NAND/SmartMedia Interface */
> + csa = readl(&matrix->ebicsa);
> + csa |= AT91_MATRIX_CS3A_SMC_SMARTMEDIA;
> + writel(csa, &matrix->ebicsa);
> +
> + /* Configure SMC CS3 for NAND/SmartMedia */
> + writel(AT91_SMC_SETUP_NWE(1) | AT91_SMC_SETUP_NCS_WR(0) |
> + AT91_SMC_SETUP_NRD(1) | AT91_SMC_SETUP_NCS_RD(0),
> + &smc->cs[3].setup);
> + writel(AT91_SMC_PULSE_NWE(3) | AT91_SMC_PULSE_NCS_WR(3) |
> + AT91_SMC_PULSE_NRD(3) | AT91_SMC_PULSE_NCS_RD(3),
> + &smc->cs[3].pulse);
> + writel(AT91_SMC_CYCLE_NWE(5) | AT91_SMC_CYCLE_NRD(5),
> + &smc->cs[3].cycle);
> + writel(AT91_SMC_MODE_RM_NRD | AT91_SMC_MODE_WM_NWE |
> + AT91_SMC_MODE_EXNW_DISABLE |
> + AT91_SMC_MODE_DBW_8 |
> + AT91_SMC_MODE_TDF_CYCLE(2),
> + &smc->cs[3].mode);
> +
> + /* Configure RDY/BSY */
> + at91_set_gpio_input(CONFIG_SYS_NAND_READY_PIN, 1);
> +
> + /* Enable NandFlash */
> + at91_set_gpio_output(CONFIG_SYS_NAND_ENABLE_PIN, 1);
> +}
> +
> +#ifdef CONFIG_MACB
> +static void stamp9G20_macb_hw_init(void)
> +{
> + struct at91_pmc *pmc = (struct at91_pmc *)ATMEL_BASE_PMC;
> + struct at91_port *pioa = (struct at91_port *)ATMEL_BASE_PIOA;
> + struct at91_rstc *rstc = (struct at91_rstc *)ATMEL_BASE_RSTC;
> + unsigned long erstl;
> +
> + /* Enable MACB Chip, this is the enable PIN on Stamp Adaptor*/
This comment is a bit misleading. Isn't MACB the SoC MAC component? Why
should we switch an external element to enable an internal part? Can you
please rewrite the comment to be more precise?
> + at91_set_gpio_output(AT91_PIN_PA26, 0);
> +
> + /* Enable EMAC clock */
> + writel(1 << ATMEL_ID_EMAC0, &pmc->pcer);
Is it possible to move this 'enable clock' into at91_macb_hw_init()?
Does your PHY address setting still work then?
If yes, can you please send a separate patch first adding the 'enable
clock' into the correct at91_macb_hw_init()?
> +
> + /*
> + * Disable pull-up on:
> + * RXDV (PA17) => PHY normal mode (not Test mode)
> + * ERX0 (PA14) => PHY ADDR0
> + * ERX1 (PA15) => PHY ADDR1
> + * ERX2 (PA25) => PHY ADDR2
> + * ERX3 (PA26) => PHY ADDR3
> + * ECRS (PA28) => PHY ADDR4 => PHYADDR = 0x0
> + *
> + * PHY has internal pull-down
> + */
> + writel(pin_to_mask(AT91_PIN_PA14) |
> + pin_to_mask(AT91_PIN_PA15) |
> + pin_to_mask(AT91_PIN_PA17) |
> + pin_to_mask(AT91_PIN_PA18) |
> + pin_to_mask(AT91_PIN_PA28),
> + &pioa->pudr);
> +
> + erstl = readl(&rstc->mr) & AT91_RSTC_MR_ERSTL_MASK;
> +
> + /* Need to reset PHY -> 500ms reset */
> + writel(AT91_RSTC_KEY | AT91_RSTC_MR_ERSTL(13) |
> + AT91_RSTC_MR_URSTEN, &rstc->mr);
Hmm ... is it OK to generate the user reset here? I know this is the
same in at least at91sam9263ek, can you please check if we should
instead delete that bit in MR?
> +
> + writel(AT91_RSTC_KEY | AT91_RSTC_CR_EXTRST, &rstc->cr);
> +
> + /* Wait for end hardware reset */
> + while (!(readl(&rstc->sr) & AT91_RSTC_SR_NRSTL))
> + ;
Endless loop if bit is not toggling, can you please add some watchdog
reset (if you still use wdt in your board) and some timeout?
This will also solve the warning detected by checkpatch.
> +
> + /* Restore NRST value */
> + writel(AT91_RSTC_KEY | erstl | AT91_RSTC_MR_URSTEN,
> + &rstc->mr);
> +
> + /* Re-enable pull-up */
> + writel(pin_to_mask(AT91_PIN_PA14) |
> + pin_to_mask(AT91_PIN_PA15) |
> + pin_to_mask(AT91_PIN_PA17) |
> + pin_to_mask(AT91_PIN_PA18) |
> + pin_to_mask(AT91_PIN_PA28),
> + &pioa->puer);
> +
> + /* Initialize EMAC=MACB hardware */
> + at91_macb_hw_init();
> +}
> +#endif /* CONFIG_MACB */
> +
> +int board_early_init_f(void)
> +{
> + struct at91_pmc *pmc = (struct at91_pmc *)ATMEL_BASE_PMC;
> +
> + /* Enable clocks for all PIOs */
> + writel((1 << ATMEL_ID_PIOA) | (1 << ATMEL_ID_PIOB) |
> + (1 << ATMEL_ID_PIOC), &pmc->pcer);
you should initialize seriald_hw here to avoid strange characters on
serial line when switching from at91bootstrap to u-boot.
> +
> + return 0;
> +}
> +
> +int board_init(void)
> +{
> +#ifdef CONFIG_PORTUXG20
> + gd->bd->bi_arch_number = MACH_TYPE_PORTUXG20;
> +#else
> + gd->bd->bi_arch_number = MACH_TYPE_STAMP9G20;
> +#endif
I would favor the generic CONFIG_MACH_TYPE here -> see
arch/arm/lib/board.c:401
Just add the definition in your board config header and remove these
lines here.
> + /* adress of boot parameters */
> + gd->bd->bi_boot_params = CONFIG_SYS_SDRAM_BASE + 0x100;
> +
> + at91_set_gpio_output(AT91_PIN_PC9, 1);
> + at91_set_gpio_output(AT91_PIN_PC5, 1);
Can you please add some comment why switching these pins?
> +
> + at91_seriald_hw_init();
> + stamp9G20_nand_hw_init();
> +#ifdef CONFIG_MACB
> + stamp9G20_macb_hw_init();
> +#endif
> +
> + return 0;
> +}
> +
> +int dram_init(void)
> +{
> + gd->ram_size = get_ram_size(
> + (void *)CONFIG_SYS_SDRAM_BASE,
> + CONFIG_SYS_SDRAM_SIZE);
> + return 0;
> +}
> +
> +#ifdef CONFIG_MACB
> +void reset_phy(void)
> +{
> + /*
> + * Initialize ethernet HW addr prior to starting Linux,
> + * needed for nfsroot
> + */
> + eth_init(gd->bd);
This is not longer required, the HW address should always be set by
generic code. Can you please check this?
> +}
> +
> +int board_eth_init(bd_t *bis)
> +{
> + int rc = 0;
> + rc = macb_eth_initialize(0, (void *)ATMEL_BASE_EMAC0, 0x00);
> + return rc;
return macb_eth_initialize()?
> +}
> +#endif /* CONFIG_MACB */
> diff --git a/boards.cfg b/boards.cfg
> index 2d36d83..9c5ba09 100644
> --- a/boards.cfg
> +++ b/boards.cfg
> @@ -120,6 +120,8 @@ cpu9G20_nand_128M arm arm926ejs cpu9260 eukrea
> pm9261 arm arm926ejs pm9261 ronetix at91 pm9261:AT91SAM9261
> pm9263 arm arm926ejs pm9263 ronetix at91 pm9263:AT91SAM9263
> pm9g45 arm arm926ejs pm9g45 ronetix at91 pm9g45:AT91SAM9G45
> +portuxg20 arm arm926ejs stamp9g20 taskit at91 stamp9g20:AT91SAM9G20,CONFIG_PORTUXG20
> +stamp9g20 arm arm926ejs stamp9g20 taskit at91 stamp9g20:AT91SAM9G20
> cam_enc_4xx arm arm926ejs cam_enc_4xx ait davinci cam_enc_4xx
> da830evm arm arm926ejs da8xxevm davinci davinci
> da850_am18xxevm arm arm926ejs da8xxevm davinci davinci da850evm:DA850_AM18X_EVM,MAC_ADDR_IN_EEPROM,SYS_I2C_EEPROM_ADDR_LEN=2,SYS_I2C_EEPROM_ADDR=0x50
> diff --git a/include/configs/stamp9g20.h b/include/configs/stamp9g20.h
> new file mode 100644
> index 0000000..7783a2f
> --- /dev/null
> +++ b/include/configs/stamp9g20.h
> @@ -0,0 +1,225 @@
> +/*
> + * (C) Copyright 2007-2008
> + * Stelian Pop <stelian at popies.net>
> + * Lead Tech Design <www.leadtechdesign.com>
> + *
> + * (C) Copyright 2010
> + * Achim Ehrlich <aehrlich at taskit.de>
> + * taskit GmbH <www.taskit.de>
> + *
> + * (C) Copyright 2012
> + * Markus Hubig <mhubig at imko.de>
> + * IMKO GmbH <www.imko.de>
> + *
> + * Configuation settings for the stamp9g20 CPU module.
> + *
> + * 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
> + */
> +
> +#ifndef __CONFIG_H
> +#define __CONFIG_H
> +
> +/*
> + * SoC must be defined first, before hardware.h is included.
> + * In this case SoC is defined in boards.cfg.
> + */
> +#include <asm/hardware.h>
> +
> +/*
> + * Warning: changing CONFIG_SYS_TEXT_BASE requires adapting the initial boot
> + * program. Since the linker has to swallow that define, we must use a pure
> + * hex number here!
> + */
> +#define CONFIG_SYS_TEXT_BASE 0x23f00000
> +
> +/* ARM asynchronous clock */
> +#define CONFIG_SYS_AT91_SLOW_CLOCK 32768 /* slow clock xtal */
> +#define CONFIG_SYS_AT91_MAIN_CLOCK 18432000 /* 18.432MHz crystal */
> +#define CONFIG_SYS_HZ 1000 /* 1ms resolution */
> +
> +/* Misc CPU related settings */
> +#define CONFIG_ARCH_CPU_INIT /* call arch_cpu_init() */
> +#undef CONFIG_USE_IRQ /* we don't need IRQ stuff */
just remove that, it was not defined before
> +#define CONFIG_CMDLINE_TAG /* pass commandline to Kernel */
> +#define CONFIG_SETUP_MEMORY_TAGS /* pass memory defs to kernel */
> +#define CONFIG_INITRD_TAG /* pass initrd param to kernel */
> +#define CONFIG_SKIP_LOWLEVEL_INIT /* U-Boot is loaded by a bootloader */
> +#define CONFIG_BOARD_EARLY_INIT_f /* call board_early_init_f() */
> +#define CONFIG_DISPLAY_CPUINFO /* display CPU Info at startup */
> +
> +/*
> + * SDRAM: 1 bank, 64 MB, base address 0x20000000
> + * Already initialized before u-boot gets started.
> + */
> +#define CONFIG_NR_DRAM_BANKS 1
> +#define CONFIG_SYS_SDRAM_BASE ATMEL_BASE_CS1
> +#define CONFIG_SYS_SDRAM_SIZE (64 * (1024 << 10))
> +
> +/*
> + * Perform a SDRAM Memtest from the start of SDRAM
> + * till the beginning of the U-Boot position in RAM.
> + */
> +#define CONFIG_SYS_MEMTEST_START CONFIG_SYS_SDRAM_BASE
> +#define CONFIG_SYS_MEMTEST_END (CONFIG_SYS_TEXT_BASE - 0x100000)
> +
> +/* Size of malloc() pool */
> +#define CONFIG_SYS_MALLOC_LEN \
> + ROUND(3 * CONFIG_ENV_SIZE + 128 * 1024, 0x1000)
> +
> +/*
> + * Initial stack pointer: 4k - GENERATED_GBL_DATA_SIZE in internal SRAM,
> + * leaving the correct space for initial global data structure above that
> + * address while providing maximum stack area below.
> + */
> +#define CONFIG_STACKSIZE (32 * 1024) /* 32k regular stack size */
CONFIG_STACKSIZE is nowhere used, please remove.
> +#define CONFIG_SYS_INIT_SP_ADDR \
> + (ATMEL_BASE_SRAM1 + 0x1000 - GENERATED_GBL_DATA_SIZE)
> +
> +/* NAND flash settings */
> +#define CONFIG_NAND_ATMEL
> +#define CONFIG_SYS_NO_FLASH
> +#define CONFIG_SYS_MAX_NAND_DEVICE 1
> +#define CONFIG_SYS_NAND_BASE ATMEL_BASE_CS3
> +#define CONFIG_SYS_NAND_DBW_8
> +#define CONFIG_SYS_NAND_MASK_ALE (1 << 21)
> +#define CONFIG_SYS_NAND_MASK_CLE (1 << 22)
> +#define CONFIG_SYS_NAND_ENABLE_PIN AT91_PIN_PC14
> +#define CONFIG_SYS_NAND_READY_PIN AT91_PIN_PC13
> +
> +/* general purpose I/O */
> +#define CONFIG_ATMEL_LEGACY /* required until (g)pio is fixed */
> +#define CONFIG_AT91_GPIO /* enable the GPIO features */
> +#define CONFIG_AT91_GPIO_PULLUP 1 /* keep pullups on peripheral pins */
> +
> +/* serial console */
> +#define CONFIG_ATMEL_USART
> +#define CONFIG_USART_BASE ATMEL_BASE_DBGU
> +#define CONFIG_USART_ID ATMEL_ID_SYS
-----------^
no tab here please
> +#define CONFIG_BAUDRATE 115200
> +#define CONFIG_SYS_BAUDRATE_TABLE {115200, 19200, 38400, 57600, 9600}
This is same as generic table, please remove (was added in
26750c8aee2383a026e0cf89e9310628d3a5a6a0)
> +#define CONFIG_USART3 /* USART 3 is used as DBGU */
Well, this is a remnant ... please remove (BTW: the statement in comment
is wrong, USART3 is available beside DBGU ... ;)
> +
> +/* Ethernet configuration */
> +#define CONFIG_MACB /* initialize the ethernet port */
------------------------------------------------------------------------^
Can you please remove these white space? I know there are a lot of at91
config header formatted like this, but I personally dislike this style.
Especially in this case where the previous intention of aligning all
these '*/' on the 80th char of a line is still broken by other lines. I
think just a single white space is good here.
> +
> +#ifdef CONFIG_MACB
Well, you defined it some lines above, is there any case where you want
to switch the network off? If no please remove this ifdef block and just
define required stuff.
> +# define CONFIG_RMII /* use reduced MII inteface */
> +# define CONFIG_RESET_PHY_R /* call reset_phy() after reloc. */
I think this could be remove once you checked if eth_init() is really
required.
> +# define CONFIG_NET_RETRY_COUNT 20 /* # of DHCP/BOOTP retries */
> +#endif /* CONFIG_MACB */
> +
> +/* USB configuration */
> +#define CONFIG_USB_ATMEL
> +#define CONFIG_USB_OHCI_NEW
> +#define CONFIG_USB_STORAGE
> +#define CONFIG_DOS_PARTITION
> +#define CONFIG_SYS_USB_OHCI_CPU_INIT
> +#define CONFIG_SYS_USB_OHCI_REGS_BASE ATMEL_UHP_BASE
> +#define CONFIG_SYS_USB_OHCI_SLOT_NAME "at91sam9260"
> +#define CONFIG_SYS_USB_OHCI_MAX_ROOT_PORTS 2
> +
> +/* BOOTP options */
> +#ifdef CONFIG_MACB
same comment as above, if you do not want to disable MACB, remove this
ifdef.
> +# define CONFIG_BOOTP_BOOTFILESIZE
> +# define CONFIG_BOOTP_BOOTPATH
> +# define CONFIG_BOOTP_GATEWAY
> +# define CONFIG_BOOTP_HOSTNAME
> +#endif /* CONFIG_MACB */
> +
> +/* General Boot Parameter */
> +#define CONFIG_BOOTDELAY 3
> +#define CONFIG_BOOTCOMMAND "run flashboot"
> +#define CONFIG_SYS_PROMPT "U-Boot> "
> +#define CONFIG_SYS_CBSIZE 256
> +#define CONFIG_SYS_MAXARGS 16
> +#define CONFIG_SYS_PBSIZE \
> + (CONFIG_SYS_CBSIZE + sizeof(CONFIG_SYS_PROMPT) + 16)
> +#define CONFIG_SYS_LONGHELP
> +#define CONFIG_CMDLINE_EDITING
> +
> +/*
> + * RAM Memory address where to put the
> + * Linux Kernel bevor starting.
> + */
> +#define CONFIG_SYS_LOAD_ADDR 0x22000000
> +
> +/*
> + * The NAND Flash partitions:
> + * ==========================================
> + * 0x0000000-0x001ffff -> 128k, bootstrap
> + * 0x0020000-0x005ffff -> 256k, u-boot
> + * 0x0060000-0x007ffff -> 128k, env1
> + * 0x0080000-0x009ffff -> 128k, env2 (backup)
> + * 0x0100000-0x03fffff -> 2M, kernel
You should think about bad blocks ... 2MiB for kernel is IMHO way to
small for normal configurations (especially when you care about boot
time, think about io-/cpu-bound stuff: high compression vs fast access).
The space will be even smaller if you hit some bad blocks in this area
... I really recommend to use a bit more here for your kernel, but its
up to you whether you change it in your next patch version or not.
> + * 0x0400000-0x8000000 -> 124M, RootFS
> + */
> +#define CONFIG_ENV_IS_IN_NAND
> +#define CONFIG_ENV_OFFSET (128 + 256 << 10)
> +#define CONFIG_ENV_OFFSET_REDUND (128 + 256 + 128 << 10)
> +#define CONFIG_ENV_SIZE (128 << 10)
> +
> +/*
> + * Predefined environment variables.
> + * Usefull to define some easy to use boot commands.
> + */
> +#define CONFIG_EXTRA_ENV_SETTINGS \
> + \
> + "basicargs=console=ttyS0,115200 mem=64M\0" \
Only a side note ... isn't the mem parameter handled by ATAGS correctly?
Or why do you add this here?
> + \
> + "mtdparts=mtdparts=atmel_nand:128k(bootstrap)ro," \
> + "256k(uboot)ro,128k(env1)ro," \
> + "128k(env2)ro,2M(linux),-(root)rw\0" \
> + \
> + "sdboot=setenv bootargs ${basicargs} ${mtdparts} " \
> + "root=/dev/mmcblk0p1 rootdelay=1; " \
another side note ... you should read about rootwait in kernel docs
> + "nand read 0x22000000 0xA0000 0x200000; " \
> + "bootm 22000000\0" \
> + \
> + "flashboot=setenv bootargs ${basicargs} ${mtdparts} " \
> + "root=/dev/mtdblock5 rootfstype=jffs2; " \
another side note ... you should read about ubifs ;)
Don't get me wrong, these side notes are just for you information and
not a request for change!
> + "nand read 0x22000000 0xA0000 0x200000; " \
> + "bootm 22000000\0" \
> + \
> + "nfsboot=setenv bootargs ${basicargs} ${mtdparts} " \
> + "root=/dev/nfs ip=dhcp " \
> + "nfsroot=${serverip}:/srv/nfs/rootfs; dhcp; " \
> + "bootm\0"
> +
> +/* Command line & features configuration */
> +#include <config_cmd_default.h>
> +#undef CONFIG_CMD_FPGA
> +#undef CONFIG_CMD_IMI
> +#undef CONFIG_CMD_IMLS
> +#undef CONFIG_CMD_LOADS
> +#undef CONFIG_CMD_SOURCE
The source command is a really useful command for scripting, you should
not disable it except you need to e.g. save space in text section. I
guess this not the case for you cause it saves just about 700 bytes in
code section.
Beside that I can not see another way for updating your device. You
disabled flash access to bootstrap, u-boot and env sections in mtdparts
so you will need u-boot to write to these sections or you need to change
these parameters in u-boot. For all these cases scripting in u-boot is
really viable. Think about ... (just another side note ;)
> +
> +#define CONFIG_CMD_NAND
> +#define CONFIG_CMD_USB
> +#define CONFIG_CMD_FAT
> +
> +#ifdef CONFIG_MACB
again ifdef block, see comments above
> +# define CONFIG_CMD_PING
> +# define CONFIG_CMD_DHCP
> +#endif /* CONFIG_MACB */
> +
> +#ifdef CONFIG_USE_IRQ
> +# error CONFIG_USE_IRQ not supported
> +#endif
I tend to say, remove that. I can not imagine a case where one enables
irq accidentially, this should only be done intentionally. Then this
part is not really useful.
But I know this is copied from other at91 config files.
> +
> +#endif
>
Best regards
Andreas Bießmann
More information about the U-Boot
mailing list