July 2021 - U-Boot

imx8qm uboot-imx custom images load
by rkohli2000 gmail 12 Jul '21

12 Jul '21

Hi all, Sorry if this is the wrong list for this. I have 2 custom images (elf or bin) stored in predefined addresses in flash that are to be loaded based on some criteria I can test for on bootup. Can someone please suggest a method to programmatically launch either image using uboot's tools like bootm, etc? Or point to an example of this procedure. Thanks in advance. Ravi

1 0

rock960c
by Trevor Woerner 12 Jul '21

12 Jul '21

Hi, I was hoping that one day support would be added for the rockchip rock960c board (from the 96boards series, https://www.96boards.org/product/rock960c/). NOTE: this board is different from the rock960 board (aka rock960 a/b) that is currently supported. Schematics and datasheets aren't easy to come by. A while back someone mentioned that the only difference between the rock960 a/b and the rock960c is that the rock960c uses different SDRAM (LPDDR4 instead of LPDDR3). Searching around on the Internet it would appear that the rock960c uses the same SDRAM as the rock-pi-4 devices (maybe?). The actual SDRAM on the board is H9HCNNN4KUML-HRNMN, but I can't find a datasheet for it anywhere, not even on Hynix's website. I tried creating a rock960c device by copying the rock960 things. The one change I made was that my rk3399-rock960c-u-boot.dtsi file has: #include "rk3399-sdram-lpddr4-100.dtsi" (which is what the rock-pi-4 is using) instead of: #include "rk3399-sdram-lpddr3-2GB-1600.dtsi" (which is what rk3399-rock960-u-boot.dtsi is using). Unfortunately this doesn't succeed. The TPL is invoked, then there's nothing else on the console: U-Boot TPL 2021.04 (Jun 30 2021 - 18:16:57) I then tried building and using the vendor-supplied u-boot fork for the rock960c which is found here: git://github.com/96rocks/u-boot.git but it is quite old (U-Boot 2017.09) and pre-dates the TPL/idbloader things. If I understand things correctly, the vendor branch only builds u-boot and the spl, I then need to mash in some sort of binary ddr file to the spl to get a working first-stage bootloader (?). I'm unclear how to take the build output from the vendor branch and create a working bootloader to load on my sdcard (the only docs for the rk3399 in that branch are generic and don't refer to things that come out of the build for the rock960c). In any case, if anyone has any any tips etc I'd appreciate it :-) Best regards, Trevor

2 2

non-DM SPI_FLASH removal?
by Marek Behun 12 Jul '21

12 Jul '21

Tom, there are 16 boards left with CONFIG_SPI_FLASH (git grep CONFIG_SPI_FLASH=y) and Makefile says this was supposed to be deprecated in v2019.07. Are we going to remove them so that we can simplify the SPI subsystem? Marek

2 4

[PATCH] board: sifive: remove the command for setting serial number
by Zong Li 12 Jul '21

12 Jul '21

We wouldn't like to allow user to change the serial number, so remove the command for changing serial number in EEPROM. Signed-off-by: Zong Li <zong.li(a)sifive.com> Suggested-by: David Abdurachmanov <david.abdurachmanov(a)sifive.com> --- .../unmatched/hifive-platform-i2c-eeprom.c | 23 +------------------ 1 file changed, 1 insertion(+), 22 deletions(-) diff --git a/board/sifive/unmatched/hifive-platform-i2c-eeprom.c b/board/sifive/unmatched/hifive-platform-i2c-eeprom.c index a2151f15e0..ad2f3155d0 100644 --- a/board/sifive/unmatched/hifive-platform-i2c-eeprom.c +++ b/board/sifive/unmatched/hifive-platform-i2c-eeprom.c @@ -401,24 +401,6 @@ static void set_product_id(char *string) update_crc(); } -/** - * set_serial_number() - set the PCB serial number in the in-memory copy - * - * Set the board serial number in the in-memory EEPROM copy from the supplied - * string argument, and update the CRC. - */ -static void set_serial_number(char *string) -{ - if (strlen(string) > SERIAL_NUMBER_BYTES) { - printf("Serial number must not be greater than 16 bytes\n"); - return; - } - - memset(e.serial, 0, sizeof(e.serial)); - strncpy((char *)e.serial, string, sizeof(e.serial)); - update_crc(); -} - /** * init_local_copy() - initialize the in-memory EEPROM copy * @@ -468,10 +450,7 @@ int do_mac(struct cmd_tbl *cmdtp, int flag, int argc, char *const argv[]) return 0; } - if (!strcmp(cmd, "serial_number")) { - set_serial_number(argv[2]); - return 0; - } else if (!strcmp(cmd, "manuf_test_status")) { + if (!strcmp(cmd, "manuf_test_status")) { set_manuf_test_status(argv[2]); return 0; } else if (!strcmp(cmd, "mac_address")) { -- 2.31.1

2 1

[PATCH u-boot-mvebu v2] arm: mvebu: armada-3720: remove unused config option
by Marek Behún 12 Jul '21

12 Jul '21

The config option CONFIG_DEBUG_UART_CLOCK is not used by Armada 3720's serial driver (it wasn't even before the recent update of that driver). Even if it was used, the value was incorrect (the frequency of the clock is 25 MHz, not 25.8048 MHz). Remove it from config files and set the default value to 0. Signed-off-by: Marek Behún <marek.behun(a)nic.cz> Cc: Stefan Roese <sr(a)denx.de> Cc: Pali Rohár <pali(a)kernel.org> --- Change since v1: - we also need to specify a default value for this variable in Kconfig, otherwise the make command will ask for a value. Use 0 for the default value since it is not used - updated commit message --- configs/mvebu_db-88f3720_defconfig | 1 - configs/mvebu_espressobin-88f3720_defconfig | 1 - configs/turris_mox_defconfig | 1 - configs/uDPU_defconfig | 1 - drivers/serial/Kconfig | 1 + 5 files changed, 1 insertion(+), 4 deletions(-) diff --git a/configs/mvebu_db-88f3720_defconfig b/configs/mvebu_db-88f3720_defconfig index c9935da102..ff4668e837 100644 --- a/configs/mvebu_db-88f3720_defconfig +++ b/configs/mvebu_db-88f3720_defconfig @@ -11,7 +11,6 @@ CONFIG_ENV_SECT_SIZE=0x10000 CONFIG_DM_GPIO=y CONFIG_DEFAULT_DEVICE_TREE="armada-3720-db" CONFIG_DEBUG_UART_BASE=0xd0012000 -CONFIG_DEBUG_UART_CLOCK=25804800 CONFIG_DEBUG_UART=y CONFIG_AHCI=y CONFIG_DISTRO_DEFAULTS=y diff --git a/configs/mvebu_espressobin-88f3720_defconfig b/configs/mvebu_espressobin-88f3720_defconfig index 157a4b7cb2..dc199cfe50 100644 --- a/configs/mvebu_espressobin-88f3720_defconfig +++ b/configs/mvebu_espressobin-88f3720_defconfig @@ -11,7 +11,6 @@ CONFIG_ENV_SECT_SIZE=0x10000 CONFIG_DM_GPIO=y CONFIG_DEFAULT_DEVICE_TREE="armada-3720-espressobin" CONFIG_DEBUG_UART_BASE=0xd0012000 -CONFIG_DEBUG_UART_CLOCK=25804800 CONFIG_DEBUG_UART=y CONFIG_AHCI=y CONFIG_DISTRO_DEFAULTS=y diff --git a/configs/turris_mox_defconfig b/configs/turris_mox_defconfig index 2c6f4938db..b3458acb98 100644 --- a/configs/turris_mox_defconfig +++ b/configs/turris_mox_defconfig @@ -11,7 +11,6 @@ CONFIG_ENV_SECT_SIZE=0x10000 CONFIG_DM_GPIO=y CONFIG_DEFAULT_DEVICE_TREE="armada-3720-turris-mox" CONFIG_DEBUG_UART_BASE=0xd0012000 -CONFIG_DEBUG_UART_CLOCK=25804800 CONFIG_DEBUG_UART=y CONFIG_OF_BOARD_FIXUP=y CONFIG_DISTRO_DEFAULTS=y diff --git a/configs/uDPU_defconfig b/configs/uDPU_defconfig index cdf6b2274c..649248d74d 100644 --- a/configs/uDPU_defconfig +++ b/configs/uDPU_defconfig @@ -10,7 +10,6 @@ CONFIG_ENV_SECT_SIZE=0x10000 CONFIG_DM_GPIO=y CONFIG_DEFAULT_DEVICE_TREE="armada-3720-uDPU" CONFIG_DEBUG_UART_BASE=0xd0012000 -CONFIG_DEBUG_UART_CLOCK=25804800 CONFIG_DEBUG_UART=y CONFIG_AHCI=y CONFIG_DISTRO_DEFAULTS=y diff --git a/drivers/serial/Kconfig b/drivers/serial/Kconfig index 961e3fb031..93348c0929 100644 --- a/drivers/serial/Kconfig +++ b/drivers/serial/Kconfig @@ -443,6 +443,7 @@ config DEBUG_UART_CLOCK int "UART input clock" depends on DEBUG_UART default 0 if DEBUG_UART_SANDBOX + default 0 if DEBUG_MVEBU_A3700_UART help The UART input clock determines the speed of the internal UART circuitry. The baud rate is derived from this by dividing the input -- 2.31.1

3 2

[PATCH] spi: mxc_spi: fix warnings if CLK_ENABLED not set
by Stefano Babic 12 Jul '21

12 Jul '21

Following warnings (unused variables) are raised: drivers/spi/mxc_spi.c: In function 'mxc_spi_probe': drivers/spi/mxc_spi.c:595:14: error: unused variable 'blob' [-Werror=unused-variable] 595 | const void *blob = gd->fdt_blob; | ^~~~ drivers/spi/mxc_spi.c:594:6: error: unused variable 'node' [-Werror=unused-variable] 594 | int node = dev_of_offset(bus); Move the variable declaration inside the code where they are used. Signed-off-by: Stefano Babic <sbabic(a)denx.de> --- drivers/spi/mxc_spi.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/drivers/spi/mxc_spi.c b/drivers/spi/mxc_spi.c index f3dddbdbd7..a80c3e737d 100644 --- a/drivers/spi/mxc_spi.c +++ b/drivers/spi/mxc_spi.c @@ -591,8 +591,6 @@ void spi_release_bus(struct spi_slave *slave) static int mxc_spi_probe(struct udevice *bus) { struct mxc_spi_slave *mxcs = dev_get_plat(bus); - int node = dev_of_offset(bus); - const void *blob = gd->fdt_blob; int ret; int i; @@ -629,6 +627,8 @@ static int mxc_spi_probe(struct udevice *bus) mxcs->max_hz = clk_get_rate(&clk); #else + int node = dev_of_offset(bus); + const void *blob = gd->fdt_blob; mxcs->max_hz = fdtdec_get_int(blob, node, "spi-max-frequency", 20000000); #endif -- 2.25.1

2 1

[PATCH v3 0/6] This patch set fixes minor issues related to tapdelays
by Ashok Reddy Soma 12 Jul '21

12 Jul '21

This patch set fixes below issues in zynq_sdhc driver - Fix issues in tap delay functions where it returns uninitialized values - Allow configuring zero tap delay values - Split tapdelay functions to set input and output tap delay's separately. - Fix kernel doc warnings - Make local structures as static structures Changes in v3: - Updated commmit description to explain why return type int is kept - Change "@degree" to "@degrees:" in 5/6, so revert it in 1/6 from sdhci_zynqmp_sampleclk_set_phase() and sdhci_versal_sdcardclk_set_phase() - Updated macro's with BIT() and GENMASK() for readability - Change "@degree" to "@degrees:" in sdhci_zynqmp_sampleclk_set_phase() and sdhci_versal_sdcardclk_set_phase() and update description with these warnings. Changes in v2: - Changed "@degree" to "@degrees:" in function descriptions of tap delay functions - Removed @degree warning from commit description since it is fixed in patch 1/6. Ashok Reddy Soma (4): mmc: zynq_sdhci: Resolve uninitialized return value mmc: zynq_sdhci: Allow configuring zero Tap values mmc: zynq_sdhci: Use Mask writes for Tap delays mmc: zynq_sdhci: Split set_tapdelay function to in and out Michal Simek (2): mmc: zynq_sdhci: Fix kernel doc warnings mmc: zynq_sdhci: Make variables/structure static board/xilinx/zynqmp/tap_delays.c | 73 ++++++++-------- drivers/mmc/zynq_sdhci.c | 144 ++++++++++++++++--------------- include/zynqmp_tap_delay.h | 7 +- 3 files changed, 115 insertions(+), 109 deletions(-) -- 2.17.1

3 13

[PATCH] clk: zynqmp: Add support for enabling clock on lpd_lsbus
by Michal Simek 12 Jul '21

12 Jul '21

lpd_lsbus is clock which is used by many IPs like dmas, gems, gpio, sdhcis, spis, ttcs, uarts, watchdog that's why make sense to also enable access to change this clock. For this clock you already get the rate. Signed-off-by: Michal Simek <michal.simek(a)xilinx.com> --- drivers/clk/clk_zynqmp.c | 1 + 1 file changed, 1 insertion(+) diff --git a/drivers/clk/clk_zynqmp.c b/drivers/clk/clk_zynqmp.c index 13a623fdb96a..52fecec7a7a9 100644 --- a/drivers/clk/clk_zynqmp.c +++ b/drivers/clk/clk_zynqmp.c @@ -813,6 +813,7 @@ static int zynqmp_clk_enable(struct clk *clk) mask = 0x3; break; case qspi_ref ... can1_ref: + case lpd_lsbus: clkact_shift = 24; mask = 0x1; break; -- 2.32.0

2 1

[PATCH] arm64: dts: zynqmp: Enable USB2.0 for zc1751-xm016-dc2
by Michal Simek 12 Jul '21

12 Jul '21

From: Piyush Mehta <piyush.mehta(a)xilinx.com> The board zynqmp-zc1751-xm016-dc2 support only USB2.0. This patch removes USB3.0 DT configuration for DC2 board. Signed-off-by: Piyush Mehta <piyush.mehta(a)xilinx.com> Signed-off-by: Michal Simek <michal.simek(a)xilinx.com> --- arch/arm/dts/zynqmp-zc1751-xm016-dc2.dts | 2 -- 1 file changed, 2 deletions(-) diff --git a/arch/arm/dts/zynqmp-zc1751-xm016-dc2.dts b/arch/arm/dts/zynqmp-zc1751-xm016-dc2.dts index 5b689dbd0937..4225a9547c58 100644 --- a/arch/arm/dts/zynqmp-zc1751-xm016-dc2.dts +++ b/arch/arm/dts/zynqmp-zc1751-xm016-dc2.dts @@ -535,8 +535,6 @@ &dwc3_1 { status = "okay"; dr_mode = "host"; - snps,usb3_lpm_capable; - maximum-speed = "super-speed"; }; &uart0 { -- 2.32.0

2 1

[PATCH 1/3] smegw01: Allow booting the Yocto image by default
by Fabio Estevam 12 Jul '21

12 Jul '21

On the Yocto image there is a single partition and the kernel and dtb are present in the 'boot' directory. Change it accordingly so that the board can boot the Yocto image by default. Use the generic 'load' command instead, which is able to read from an ext4 partition. Signed-off-by: Fabio Estevam <festevam(a)denx.de> --- include/configs/smegw01.h | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/include/configs/smegw01.h b/include/configs/smegw01.h index 50f00136ac46..a92c583c8669 100644 --- a/include/configs/smegw01.h +++ b/include/configs/smegw01.h @@ -29,11 +29,11 @@ "bootm_size=0x10000000\0" \ "mmcdev=0\0" \ "mmcpart=1\0" \ - "mmcroot=/dev/mmcblk0p2 rootwait rw\0" \ + "mmcroot=/dev/mmcblk0p1 rootwait rw\0" \ "mmcargs=setenv bootargs console=${console},${baudrate} " \ "root=${mmcroot}\0" \ - "loadimage=fatload mmc ${mmcdev}:${mmcpart} ${loadaddr} ${image}\0" \ - "loadfdt=fatload mmc ${mmcdev}:${mmcpart} ${fdt_addr} ${fdtfile}\0" \ + "loadimage=load mmc ${mmcdev}:${mmcpart} ${loadaddr} boot/${image}\0" \ + "loadfdt=load mmc ${mmcdev}:${mmcpart} ${fdt_addr} boot/${fdtfile}\0" \ "mmcboot=echo Booting from mmc ...; " \ "run mmcargs; " \ "if run loadfdt; then " \ -- 2.25.1

4 9