May 2014 - U-Boot

[U-Boot] mkimage: path behavior for FIT images
by Yegor Yefremov 27 May '14

27 May '14

I have following folder structure: board/kernel_fdt.its output/images/zImage The kernel.its has following path for the zImage: data = /incbin/("output/images/zImage"); When I invoke mkimage like this, it complains, that "output/images/zImage" doesn't exist: mkimage -f board/kernel_fdt.its output/images/kernel_fdt.itb in other words it seems to take zImage's relative path starting from "board", where kernel_fdt.its is located. Is it a desired behavior and how can I override it, so that root will be set to the mkimage invocation path? Regards, Yegor

2 4

Re: [U-Boot] [PATCH] Exynos: Make sure ps_hold gets set in the SPL
by Akshay Saraswat 27 May '14

27 May '14

Hi Simon, >On 22 May 2014 01:53, Akshay Saraswat <akshay.s(a)samsung.com> wrote: >> From: Doug Anderson <dianders(a)chromium.org> >> >> Setting ps_hold ought to be one of the first things we do when we >> first boot up. If we wait until the main u-boot runs we won't set it >> in time and the PMIC may power us back off. >> >> Moving ps_hold setup into the generic power_init() which >> should contain code that's currently duplicated in the >> board_power_init() of several boards. >> >> Signed-off-by: Doug Anderson <dianders(a)chromium.org> >> Signed-off-by: Akshay Saraswat <akshay.s(a)samsung.com> >> --- >> arch/arm/cpu/armv7/exynos/lowlevel_init.c | 6 +++++- >> arch/arm/cpu/armv7/exynos/power.c | 14 ++++++++++++++ >> arch/arm/include/asm/arch-exynos/power.h | 8 ++++++++ >> 3 files changed, 27 insertions(+), 1 deletion(-) > >Acked-by: Simon Glass <sjg(a)chromium.org> > >Also see http://patchwork.ozlabs.org/patch/350693/ but your patch is better. > Shall we keep this one then ? > >Regards, >Simon > Regards, Akshay Saraswat

2 1

[U-Boot] [PATCH v2 1/2] ARM: at91sam9m10g45ek: enable mci0 support
by Josh Wu 27 May '14

27 May '14

Also we enable the mmc command in configuration file. As both CONFIG_CMD_MMC and CONFIG_CMD_USB use the CONFIG_DOS_PARTITION, so remove the redundant CONFIG_DOS_PARTITION definition. Signed-off-by: Josh Wu <josh.wu(a)atmel.com> --- v1 -> v2: 1. refined the comment for the pins. 2. remove detected pin initialization. 3. move the mci initialization code to devices.c 4. remove useless header file: mmc.h .../cpu/arm926ejs/at91/at91sam9m10g45_devices.c | 17 +++++++++++++++++ board/atmel/at91sam9m10g45ek/at91sam9m10g45ek.c | 10 ++++++++++ include/configs/at91sam9m10g45ek.h | 15 ++++++++++++++- 3 files changed, 41 insertions(+), 1 deletion(-) diff --git a/arch/arm/cpu/arm926ejs/at91/at91sam9m10g45_devices.c b/arch/arm/cpu/arm926ejs/at91/at91sam9m10g45_devices.c index 7d7725c..0e6c0da 100644 --- a/arch/arm/cpu/arm926ejs/at91/at91sam9m10g45_devices.c +++ b/arch/arm/cpu/arm926ejs/at91/at91sam9m10g45_devices.c @@ -165,3 +165,20 @@ void at91_macb_hw_init(void) #endif } #endif + +#ifdef CONFIG_GENERIC_ATMEL_MCI +void at91_mci_hw_init(void) +{ + struct at91_pmc *pmc = (struct at91_pmc *)ATMEL_BASE_PMC; + + at91_set_a_periph(AT91_PIO_PORTA, 0, 0); /* MCI0 CLK */ + at91_set_a_periph(AT91_PIO_PORTA, 1, 0); /* MCI0 CDA */ + at91_set_a_periph(AT91_PIO_PORTA, 2, 0); /* MCI0 DA0 */ + at91_set_a_periph(AT91_PIO_PORTA, 3, 0); /* MCI0 DA1 */ + at91_set_a_periph(AT91_PIO_PORTA, 4, 0); /* MCI0 DA2 */ + at91_set_a_periph(AT91_PIO_PORTA, 5, 0); /* MCI0 DA3 */ + + /* Enable clock */ + writel(1 << ATMEL_ID_MCI0, &pmc->pcer); +} +#endif diff --git a/board/atmel/at91sam9m10g45ek/at91sam9m10g45ek.c b/board/atmel/at91sam9m10g45ek/at91sam9m10g45ek.c index b7e2efd..5788116 100644 --- a/board/atmel/at91sam9m10g45ek/at91sam9m10g45ek.c +++ b/board/atmel/at91sam9m10g45ek/at91sam9m10g45ek.c @@ -16,6 +16,7 @@ #include <asm/arch/clk.h> #include <lcd.h> #include <atmel_lcdc.h> +#include <atmel_mci.h> #if defined(CONFIG_RESET_PHY_R) && defined(CONFIG_MACB) #include <net.h> #endif @@ -217,6 +218,15 @@ void lcd_show_board_info(void) #endif /* CONFIG_LCD_INFO */ #endif +#ifdef CONFIG_GENERIC_ATMEL_MCI +int board_mmc_init(bd_t *bis) +{ + at91_mci_hw_init(); + + return atmel_mci_init((void *)ATMEL_BASE_MCI0); +} +#endif + int board_early_init_f(void) { at91_seriald_hw_init(); diff --git a/include/configs/at91sam9m10g45ek.h b/include/configs/at91sam9m10g45ek.h index ccfda71..df277e7 100644 --- a/include/configs/at91sam9m10g45ek.h +++ b/include/configs/at91sam9m10g45ek.h @@ -115,6 +115,20 @@ #endif +/* MMC */ +#define CONFIG_CMD_MMC + +#ifdef CONFIG_CMD_MMC +#define CONFIG_MMC +#define CONFIG_GENERIC_MMC +#define CONFIG_GENERIC_ATMEL_MCI +#endif + +#if defined(CONFIG_CMD_USB) || defined(CONFIG_CMD_MMC) +#define CONFIG_CMD_FAT +#define CONFIG_DOS_PARTITION +#endif + /* Ethernet */ #define CONFIG_MACB #define CONFIG_RMII @@ -126,7 +140,6 @@ #define CONFIG_USB_EHCI #define CONFIG_USB_EHCI_ATMEL #define CONFIG_SYS_USB_EHCI_MAX_ROOT_PORTS 2 -#define CONFIG_DOS_PARTITION #define CONFIG_USB_STORAGE #define CONFIG_SYS_LOAD_ADDR 0x22000000 /* load address */ -- 1.7.9.5

3 5

[U-Boot] [PATCH v2 0/3] ARM: at91: add missed the AT91FAMILY and ARM926EJS in SoC header
by Josh Wu 27 May '14

27 May '14

Add missed AT91FAMILY and ARM926EJS macro in at91sam9x5.h. Now as all at91 SoC header defined AT91FAMILY and ARM926EJS, so we can remove the redundant definitions in all at91 board config files. v1 -> v2: 1. add ARM926EJS in at91sam9x5 SoC header. 2. remove all redundant ARM926EJS in at91 board files. 3. remove redundant AT91FAMILY in corvus, cpu9260, ethernut5 and vl_ma2sc boards. Josh Wu (3): ARM: at91sam9x5: define the AT91FAMILY and ARM926EJS in SoC header at91: remove redundant AT91FAMILY definition in board config file ARM: at91: remove redundant ARM926EJS definition in board config files arch/arm/include/asm/arch-at91/at91sam9x5.h | 3 +++ include/configs/at91sam9m10g45ek.h | 1 - include/configs/at91sam9n12ek.h | 3 --- include/configs/at91sam9x5ek.h | 1 - include/configs/corvus.h | 2 -- include/configs/cpu9260.h | 1 - include/configs/ethernut5.h | 2 -- include/configs/sama5d3_xplained.h | 1 - include/configs/sama5d3xek.h | 1 - include/configs/vl_ma2sc.h | 2 -- 10 files changed, 3 insertions(+), 14 deletions(-) -- 1.7.9.5

3 7

[U-Boot] [PATCH] ARM: at91: remove AT91X40 macro since it is not use any more
by Josh Wu 27 May '14

27 May '14

The at91x40.h is not exist. So we remove it. Signed-off-by: Josh Wu <josh.wu(a)atmel.com> --- arch/arm/include/asm/arch-at91/hardware.h | 2 -- 1 file changed, 2 deletions(-) diff --git a/arch/arm/include/asm/arch-at91/hardware.h b/arch/arm/include/asm/arch-at91/hardware.h index a63f974..d712a0d 100644 --- a/arch/arm/include/asm/arch-at91/hardware.h +++ b/arch/arm/include/asm/arch-at91/hardware.h @@ -25,8 +25,6 @@ # include <asm/arch/at91sam9x5.h> #elif defined(CONFIG_AT91CAP9) # include <asm/arch/at91cap9.h> -#elif defined(CONFIG_AT91X40) -# include <asm/arch/at91x40.h> #elif defined(CONFIG_SAMA5D3) # include <asm/arch/sama5d3.h> #else -- 1.7.9.5

3 2

[U-Boot] [PATCH v3 0/3] Add atmel ROM code image
by Andreas Bießmann 27 May '14

27 May '14

This series add atmelimage support to mkimage. An atmelimage is a quite dumb image type cause it has no real header. The file is mostly unmodified but the 6'th ARM vector gets replaced by the image size to load. Heiko, I know your approach setting the vector in start.S but I think this solution is a bit smarter. We would need to patch at least 2 start.S files (arm926ejs and armv7), maybe also the arm920t one too. In fact is the conversion of executable BLOB to ROM detected executable BLOB something that should really be done afterwards as all other SoC vendors do. In version 2 of this patchset the atmelimage mkimage target can also generate PMECC NAND header. In version 3 of this patchset the new host tool atmel_pmecc_params generate the configuration string for mkimage with the same settings used in atmel_nand driver. Changes in v3: * remove check for 'usePmecc=y' * add atmel_pmecc_params host tool, it generates an appropriate string for mkimage -n switch with compiled in values also used in atmel_nand driver * use CONFIG_SPL_GENERATE_ATMEL_PMECC_HEADER Changes in v2: * do not check filesize in atmelimage * add PMECC header generation and printout * add printout of 6'th vector * add -n switch to mkimage for PMECC type NAND and CONFIG_SYS_NANDFLASH Andreas Bießmann (3): mkimage: add atmelimage arm:at91: enable ROM loadable atmel image sama5d3xek: enable PMECC header generation arch/arm/cpu/armv7/at91/config.mk | 10 ++ common/image.c | 1 + doc/README.atmel_pmecc | 21 +++ include/configs/sama5d3xek.h | 1 + include/image.h | 1 + spl/Makefile | 11 ++ tools/Makefile | 3 + tools/atmel_pmecc_params.c | 51 ++++++ tools/atmelimage.c | 342 +++++++++++++++++++++++++++++++++++++ tools/imagetool.c | 2 + tools/imagetool.h | 1 + 11 files changed, 444 insertions(+) create mode 100644 arch/arm/cpu/armv7/at91/config.mk create mode 100644 tools/atmel_pmecc_params.c create mode 100644 tools/atmelimage.c -- 1.7.10.4

2 7

[U-Boot] [PATCH] corvus: remove unneeded CONFIG_AT91_LEGACY
by Andreas Bießmann 27 May '14

27 May '14

Signed-off-by: Andreas Bießmann <andreas.devel(a)googlemail.com> Cc: Heiko Schocher <hs(a)denx.de> --- include/configs/corvus.h | 1 - 1 file changed, 1 deletion(-) diff --git a/include/configs/corvus.h b/include/configs/corvus.h index 959e188..631add3 100644 --- a/include/configs/corvus.h +++ b/include/configs/corvus.h @@ -27,7 +27,6 @@ #define CONFIG_SYS_TEXT_BASE 0x73f00000 -#define CONFIG_AT91_LEGACY #define CONFIG_ATMEL_LEGACY /* required until (g)pio is fixed */ /* ARM asynchronous clock */ -- 1.7.10.4

1 1

[U-Boot] [PATCH] mmc: atmel_mci: fix print incorrect buffer content for debug
by Josh Wu 27 May '14

27 May '14

Signed-off-by: Josh Wu <josh.wu(a)atmel.com> --- drivers/mmc/gen_atmel_mci.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/mmc/gen_atmel_mci.c b/drivers/mmc/gen_atmel_mci.c index acca026..eb2fe1c 100644 --- a/drivers/mmc/gen_atmel_mci.c +++ b/drivers/mmc/gen_atmel_mci.c @@ -244,7 +244,7 @@ mci_send_cmd(struct mmc *mmc, struct mmc_cmd *cmd, struct mmc_data *data) if (data->flags & MMC_DATA_READ) { printf("Read Data:\n"); - print_buffer(0, data->dest, 1, + print_buffer(0, data->dest + word_count * 4 * block_count, 1, word_count*4, 0); } #endif -- 1.7.9.5

2 1

[U-Boot] [PATCH] gpio: at91: add sanity check for the NULL pointer
by Josh Wu 27 May '14

27 May '14

We need check the NULL pointer as at91_pio_get_port() may return NULL. Also print a error message when at91_pio_get_port() failed otherwise we cannot notice the failure. Signed-off-by: Josh Wu <josh.wu(a)atmel.com> --- drivers/gpio/at91_gpio.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/drivers/gpio/at91_gpio.c b/drivers/gpio/at91_gpio.c index 0b70071..6517af1 100644 --- a/drivers/gpio/at91_gpio.c +++ b/drivers/gpio/at91_gpio.c @@ -34,6 +34,7 @@ static struct at91_port *at91_pio_get_port(unsigned port) #endif #endif default: + printf("Error: at91_gpio: Fail to get PIO base!\n"); return NULL; } } @@ -200,7 +201,7 @@ int at91_set_pio_output(unsigned port, u32 pin, int value) struct at91_port *at91_port = at91_pio_get_port(port); u32 mask; - if ((port < ATMEL_PIO_PORTS) && (pin < 32)) { + if (at91_port && (port < ATMEL_PIO_PORTS) && (pin < 32)) { mask = 1 << pin; writel(mask, &at91_port->idr); writel(mask, &at91_port->pudr); -- 1.7.9.5

2 1

[U-Boot] [PATCH 1/4] ARM: atmel: sama5d3xek: convert to generic board
by Bo Shen 27 May '14

27 May '14

Enable CONFIG_SYS_GENERIC_BOARD, so that we get rid of the following warning on boot: "Warning: Your board does not use generic board. Please read doc/README.generic-board and take action. Boards not upgraded by the late 2014 may break or be removed." Signed-off-by: Bo Shen <voice.shen(a)atmel.com> --- include/configs/sama5d3xek.h | 2 ++ 1 file changed, 2 insertions(+) diff --git a/include/configs/sama5d3xek.h b/include/configs/sama5d3xek.h index 516be85..20a5f6e 100644 --- a/include/configs/sama5d3xek.h +++ b/include/configs/sama5d3xek.h @@ -34,6 +34,8 @@ #define CONFIG_CMD_BOOTZ #define CONFIG_OF_LIBFDT /* Device Tree support */ +#define CONFIG_SYS_GENERIC_BOARD + /* general purpose I/O */ #define CONFIG_AT91_GPIO -- 1.8.5.2

3 10