April 2019 - U-Boot

[U-Boot] [PATCHv4 0/6] dm: cache: add dm cache driver
by Dinh Nguyen 23 Apr '19

23 Apr '19

Hi, This is V4 of the series to add a UCLASS_CACHE dm driver to handling the configuration of cache settings. Place this new driver under /drivers/cache. In this initial revision, the driver is only configuring what I think are essential cache settings. The more comprehensive cache settings can be done in the OS. Diffs from v3: - Re-order includes in a correct order - Add final Reviewed-by: in "dm: cache: Create a uclass for cache" Dinh Nguyen (6): Documentation: dts: Add pl310 cache controller dts documentation ARM: pl310: Add macro's for handling tag and data latency mask dm: cache: Create a uclass for cache dm: cache: add the pl310 cache controller driver ARM: socfpga: use the pl310 driver to configure the cache configs: socfpga: add imply pl310 cache controller .../devicetree/bindings/arm/l2c2x0.txt | 114 ++++++++++++++++++ arch/arm/Kconfig | 1 + arch/arm/include/asm/pl310.h | 3 + arch/arm/mach-socfpga/misc.c | 16 +-- drivers/Kconfig | 2 + drivers/Makefile | 1 + drivers/cache/Kconfig | 25 ++++ drivers/cache/Makefile | 4 + drivers/cache/cache-l2x0.c | 76 ++++++++++++ drivers/cache/cache-uclass.c | 24 ++++ drivers/cache/sandbox_cache.c | 34 ++++++ include/cache.h | 38 ++++++ include/dm/uclass-id.h | 1 + test/dm/cache.c | 20 +++ 14 files changed, 346 insertions(+), 13 deletions(-) create mode 100644 Documentation/devicetree/bindings/arm/l2c2x0.txt create mode 100644 drivers/cache/Kconfig create mode 100644 drivers/cache/Makefile create mode 100644 drivers/cache/cache-l2x0.c create mode 100644 drivers/cache/cache-uclass.c create mode 100644 drivers/cache/sandbox_cache.c create mode 100644 include/cache.h create mode 100644 test/dm/cache.c -- 2.20.0

2 8

[U-Boot] [PATCH v2] rockchip: arm: remove no use macro
by Kever Yang 23 Apr '19

23 Apr '19

TIMER7_BASE is no used by source code now, remove it. Signed-off-by: Kever Yang <kever.yang(a)rock-chips.com> --- arch/arm/include/asm/arch-rockchip/hardware.h | 2 -- 1 file changed, 2 deletions(-) diff --git a/arch/arm/include/asm/arch-rockchip/hardware.h b/arch/arm/include/asm/arch-rockchip/hardware.h index cd94bdd1ba..62e8bed8f3 100644 --- a/arch/arm/include/asm/arch-rockchip/hardware.h +++ b/arch/arm/include/asm/arch-rockchip/hardware.h @@ -10,8 +10,6 @@ #define RK_SETBITS(set) RK_CLRSETBITS(0, set) #define RK_CLRBITS(clr) RK_CLRSETBITS(clr, 0) -#define TIMER7_BASE 0xff810020 - #define rk_clrsetreg(addr, clr, set) \ writel(((clr) | (set)) << 16 | (set), addr) #define rk_clrreg(addr, clr) writel((clr) << 16, addr) -- 2.20.1

2 2

[U-Boot] [PATCH 0/8] rockchip: Improve SPI-NOR read performance for the RK3399-Q7
by Philipp Tomsich 23 Apr '19

23 Apr '19

The SPI-NOR driver has traditionally been slow enough to impact boot-time from SPI-NOR (which is the recommended storage for u-boot.itb on the RK3399-Q7): transfer for the ~890KB exceeded 0.8s even though the SPI-NOR bitrate was configured to 49.5MBit/s. This series provides some urgently needed tough love for the SPI driver: - contains a few cleanups - implements a new 16bit-wide transfer mode to better utilise the 32x16bit RXFIFO - increases bus-utilisation to > 94% both for 49.5MBit/s (and, in an experimental branch that allows higher bitrates) for 74.25MBit/s With the entire series applied, we can now load our u-boot.itb in ~0.095s (instead of ~ 0.820s previously). This is still slower than from the on-module eMMC (0.028s) that has an 8-bit wide data-path, yet makes the SPI-NOR finally useful for all customers that want to keep the boot firmware out-of-the-way of the operating system. Philipp Tomsich (8): rockchip: spi: add debug message for delay in CS toggle rockchip: spi: remove unused code and fields in priv rockchip: spi: fix off-by-one in chunk size computation rockchip: spi: consistently use false/true with rkspi_enable_chip rockchip: spi: only wait for completion, if transmitting rockchip: spi: add optimised receive-only implementation rockchip: spi: add driver-data and a 'rxonly_manages_fifo' flag rockchip: spi: make optimised receive-handler unaligned-safe drivers/spi/rk_spi.c | 163 +++++++++++++++++++++++++++++++++++++++------------ 1 file changed, 127 insertions(+), 36 deletions(-) -- 2.1.4

1 16

[U-Boot] [PATCH] rockchip: rk3399-puma: support Gigadevice SPI-NOR flash
by Philipp Tomsich 23 Apr '19

23 Apr '19

Over the last quarter, a part of our production has used NOR flash from Gigadevice in addition to the Winbond parts that we typically source. This requires the SPI_FLASH_GIGADEVICE config to be set. Enable SPI_FLASH_GIGADEVICE in the board's default defconfig. Signed-off-by: Philipp Tomsich <philipp.tomsich(a)theobroma-systems.com> --- configs/puma-rk3399_defconfig | 1 + 1 file changed, 1 insertion(+) diff --git a/configs/puma-rk3399_defconfig b/configs/puma-rk3399_defconfig index 3fcff25..9b4bfaf 100644 --- a/configs/puma-rk3399_defconfig +++ b/configs/puma-rk3399_defconfig @@ -67,6 +67,7 @@ CONFIG_MMC_SDHCI=y CONFIG_MMC_SDHCI_SDMA=y CONFIG_MMC_SDHCI_ROCKCHIP=y CONFIG_SPI_FLASH=y +CONFIG_SPI_FLASH_GIGADEVICE=y CONFIG_SPI_FLASH_WINBOND=y CONFIG_PHY_MICREL=y CONFIG_PHY_MICREL_KSZ90X1=y -- 2.1.4

2 2

[U-Boot] [PULL] u-boot-stm32 for v2019.07-rc1 (round 2)
by Patrice CHOTARD 23 Apr '19

23 Apr '19

Hi Tom Please find the pull request for STM32 round 2 The following changes since commit 1f4ae66eaab29bfb5d1eb44996f7826c9cd01ed1: Merge tag 'arc-for-2019.07' of git://git.denx.de/u-boot-arc (2019-04-18 12:12:16 -0400) are available in the git repository at: https://github.com/pchotard/u-boot.git tags/u-boot-stm32-mcu-20190419 for you to fetch changes up to 5b37873ff04f082efa0a74ba04185599ccec165b: mmc: stm32_sdmmc2: Fix r1b timeout issue (2019-04-19 17:19:32 +0200) ---------------------------------------------------------------- STM32 MCUs update: _ DT rework and alignment with DT kernel v4.20 _ mmc: arm_pl180_mmci: Synchronize compatible with kernel v4.20 _ mmc: stm32_sdmmc2: Synchronize properties with kernel v4.20 _ configs: update for F746/769 boards ---------------------------------------------------------------- Christophe Kerello (1): mmc: stm32_sdmmc2: Fix r1b timeout issue Patrice Chotard (15): ARM:dts: stm32: sort nodes by alphabetical order in f4 u-boot files ARM: dts: stm32: Sync DT files with v4.20 kernel for stm32f4 ARM: dts: stm32: Migrate U-boot nodes to U-boot DT files for stm32f7 ARM: dts: stm32: Sync DT with v4.20 kernel for stm32f7 ARM: dts: Migrate U-boot nodes to U-boot DT files for stm32h7 ARM: dts: stm32: Sync DT with v4.20 kernel for stm32h7 ARM: dts: stm32: Restore old usart1 clock bindings for stm32f7 pinctrl: stm32: Add st,stm32f769-pinctrl compatible string mmc: arm_pl180_mmci: Sync compatible with kernel configs: stm32f746-disco: update EXTRA_ENV_SETTINGS configs: stm32f746-disco: enable CONFIG_DISTRO_DEFAULTS configs: stm32f746-disco: Enable SPI_FLASH_MACRONIX board: stm32f746-disco: Get MII/RMII phy_mode from DT ARM: dts: stm32: Update sdmmc binding for stm32h743i-eval ARM: dts: stm32: Update sdmmc binding for stm32mp157c-ed1 Patrick Delaunay (1): mmc: stm32_sdmmc2: Update DT properties with v4.19 bindings arch/arm/dts/stm32746g-eval-u-boot.dtsi | 188 +++ arch/arm/dts/stm32746g-eval.dts | 222 +--- arch/arm/dts/stm32f4-pinctrl.dtsi | 27 +- arch/arm/dts/stm32f429-disco-u-boot.dtsi | 18 +- arch/arm/dts/stm32f429-disco.dts | 4 +- arch/arm/dts/stm32f429-pinctrl.dtsi | 3 +- arch/arm/dts/stm32f429.dtsi | 33 +- arch/arm/dts/stm32f469-disco-u-boot.dtsi | 46 +- arch/arm/dts/stm32f469-disco.dts | 86 +- arch/arm/dts/stm32f469-pinctrl.dtsi | 3 +- arch/arm/dts/stm32f469.dtsi | 19 + arch/arm/dts/stm32f7-pinctrl.dtsi | 289 ++++ arch/arm/dts/stm32f7-u-boot.dtsi | 139 +- arch/arm/dts/stm32f746-disco-u-boot.dtsi | 251 ++++ arch/arm/dts/stm32f746-disco.dts | 279 +--- arch/arm/dts/stm32f746-pinctrl.dtsi | 11 + arch/arm/dts/stm32f746.dtsi | 747 +++++++---- arch/arm/dts/stm32f769-disco-u-boot.dtsi | 165 +++ arch/arm/dts/stm32f769-disco.dts | 236 +--- arch/arm/dts/stm32f769-pinctrl.dtsi | 11 + arch/arm/dts/stm32h7-u-boot.dtsi | 197 ++- arch/arm/dts/stm32h743-pinctrl.dtsi | 160 +-- arch/arm/dts/stm32h743.dtsi | 462 ++++++- arch/arm/dts/stm32h743i-disco-u-boot.dtsi | 11 + arch/arm/dts/stm32h743i-disco.dts | 44 +- arch/arm/dts/stm32h743i-eval-u-boot.dtsi | 12 + arch/arm/dts/stm32h743i-eval.dts | 79 +- arch/arm/dts/stm32mp157c-ed1.dts | 10 +- board/st/stm32f746-disco/stm32f746-disco.c | 23 +- configs/stm32f746-disco_defconfig | 14 +- drivers/mmc/arm_pl180_mmci.c | 14 +- drivers/mmc/arm_pl180_mmci.h | 3 +- drivers/mmc/stm32_sdmmc2.c | 67 +- drivers/pinctrl/pinctrl_stm32.c | 1 + include/configs/stm32f746-disco.h | 22 +- include/dt-bindings/clock/stm32fx-clock.h | 7 +- include/dt-bindings/pinctrl/stm32f746-pinfunc.h | 1341 ------------------- include/dt-bindings/pinctrl/stm32h7-pinfunc.h | 1612 ----------------------- 38 files changed, 2740 insertions(+), 4116 deletions(-) create mode 100644 arch/arm/dts/stm32746g-eval-u-boot.dtsi create mode 100644 arch/arm/dts/stm32f469.dtsi create mode 100644 arch/arm/dts/stm32f7-pinctrl.dtsi create mode 100644 arch/arm/dts/stm32f746-disco-u-boot.dtsi create mode 100644 arch/arm/dts/stm32f746-pinctrl.dtsi create mode 100644 arch/arm/dts/stm32f769-disco-u-boot.dtsi create mode 100644 arch/arm/dts/stm32f769-pinctrl.dtsi create mode 100644 arch/arm/dts/stm32h743i-disco-u-boot.dtsi create mode 100644 arch/arm/dts/stm32h743i-eval-u-boot.dtsi delete mode 100644 include/dt-bindings/pinctrl/stm32f746-pinfunc.h delete mode 100644 include/dt-bindings/pinctrl/stm32h7-pinfunc.h

2 2

[U-Boot] should all include/configs/*.h headers be included via SYS_CONFIG_NAME?
by Robert P. J. Day 23 Apr '19

23 Apr '19

pretty sure i know the answer to this, but i'm going to ask it anyway ... should all header files under include/configs/ be included in a build either directly or indirectly via the value in CONFIG_SYS_CONFIG_NAME? i'm aware of that Kbuild setting and what it's used for: config SYS_CONFIG_NAME string "Board configuration name" default "zynq-common" help This option contains information about board configuration name. Based on this option include/configs/<CONFIG_SYS_CONFIG_NAME>.h header will be used for board configuration. at which point that header file might well include further header files under include/configs/ and so on. so the question is, are include/configs/ header files meant to be included *exclusively* via CONFIG_SYS_CONFIG_NAME? i ask since, out of curiosity, i did a quick grep to see if any of them were included from anywhere else, and i found all of one hit in the entire code base: $ git grep '#include <configs' | grep -v ^include board/renesas/ecovec/lowlevel_init.S:#include <configs/ecovec.h> $ i thought it was odd that there would be a single example of that in the whole code base, even more so since the board file Kconfig board/renesas/ecovec/Kconfig includes: config SYS_CONFIG_NAME default "ecovec" which suggests that header file would have been included anyway. thoughts? rday -- ======================================================================== Robert P. J. Day Ottawa, Ontario, CANADA http://crashcourse.ca Twitter: http://twitter.com/rpjday LinkedIn: http://ca.linkedin.com/in/rpjday ========================================================================

1 0

[U-Boot] mdio error on ar9344
by muaddib 23 Apr '19

23 Apr '19

Hi, I am using device based on Atheros AR9344. When U-Boot initialize net device I get error mdio_register: non unique device name 'gmac0' dm tree command shows that eth_ag7xxx is not probed. ag7xxx_eth_probe -> ag7xxx_mac_probe -> ag933x_phy_setup -> ag933x_phy_setup_reset_set -> ag7xxx_mdio_write -> ag7xxx_mdio_rw return -ETIMEDOUT because ag7xxx_switch_reg_read return 0xFFFFFFFF, see below static int ag7xxx_mdio_rw(struct mii_dev *bus, int addr, int reg, u32 val) { u32 data; unsigned long start; int ret; // No idea if this is long enough or too long int timeout_ms = 1000; // Dummy read followed by PHY read/write command. ret = ag7xxx_switch_reg_read(bus, 0x98, &data); if (ret < 0) return ret; data = val | (reg << 16) | (addr << 21) | BIT(30) | BIT(31); ret = ag7xxx_switch_reg_write(bus, 0x98, data); if (ret < 0) return ret; start = get_timer(0); // Wait for operation to finish do { ret = ag7xxx_switch_reg_read(bus, 0x98, &data); if (ret < 0) return ret; if (get_timer(start) > timeout_ms) return -ETIMEDOUT; } while (data & BIT(31)); return data & 0xffff; } I use tplink_wdr4320 u-boot profile for my device.

1 0

[U-Boot] Pull request: goldsimon/u-boot
by Simon Goldschmidt 23 Apr '19

23 Apr '19

The following changes since commit 6c5f8dd540d7a8eff244d4c27a09451ca12c8d20: Merge branch 'master' of git://git.denx.de/u-boot-usb (2019-04-21 19:00:04 -0400) are available in the Git repository at: https://github.com/goldsimon/u-boot.git for you to fetch changes up to 5ff194620a5ada0bf05d9dc2668d2233289742a6: arm: socfpga: clean up board config files (2019-04-22 20:08:59 +0200) ---------------------------------------------------------------- Simon Goldschmidt (15): socfpga: add Simon Goldschmidt as co-custodian eeprom: fix DM_I2C support without CONFIG_SYS_I2C_EEPROM_BUS Revert "cmd: Kconfig: Do not include EEPROM if DM_I2C is used without DM_I2C_COMPAT" i2c: designware: fix reset handling on socfpga gen5 rtc: m41t62: add compatible for m41t82 arm: socfpga: socrates: make rtc work configs: socfpga: socrates: enable rtc support arm: socfpga: fix comment about SPL memory layout arm: socfpga: put initial U-Boot stack into DDR arm: socfpga: imply/default common config options arm: socfpga: gen5: reduce SPL pre-reloc malloc arm: socfpga: move vining_fpga to DM_I2C arm: socfpga: clean up socfpga_common.h arm: socfpga: remove CONFIG_SYS_BOOTMAPSZ arm: socfpga: clean up board config files MAINTAINERS | 1 + arch/arm/Kconfig | 2 ++ arch/arm/mach-socfpga/Kconfig | 23 +++++++++++++++ board/ebv/socrates/qts/iocsr_config.h | 2 +- cmd/Kconfig | 1 - cmd/eeprom.c | 23 +++++++++------ configs/socfpga_arria10_defconfig | 6 ---- configs/socfpga_arria5_defconfig | 10 ------- configs/socfpga_cyclone5_defconfig | 10 ------- configs/socfpga_dbm_soc1_defconfig | 10 ------- configs/socfpga_de0_nano_soc_defconfig | 10 ------- configs/socfpga_de10_nano_defconfig | 10 ------- configs/socfpga_de1_soc_defconfig | 10 ------- configs/socfpga_is1_defconfig | 8 ------ configs/socfpga_sockit_defconfig | 10 ------- configs/socfpga_socrates_defconfig | 12 ++------ configs/socfpga_sr1500_defconfig | 10 ------- configs/socfpga_stratix10_defconfig | 2 -- configs/socfpga_vining_fpga_defconfig | 18 +++++------- doc/git-mailrc | 3 +- drivers/i2c/designware_i2c.c | 20 +++++++++---- drivers/rtc/m41t62.c | 11 ++++++++ include/configs/socfpga_arria10_socdk.h | 6 ---- include/configs/socfpga_arria5_socdk.h | 2 -- include/configs/socfpga_common.h | 116 +++++++++++++++------------------------------------------------------------- include/configs/socfpga_cyclone5_socdk.h | 2 -- include/configs/socfpga_de0_nano_soc.h | 2 -- include/configs/socfpga_de10_nano.h | 2 -- include/configs/socfpga_de1_soc.h | 2 -- include/configs/socfpga_is1.h | 2 -- include/configs/socfpga_sockit.h | 2 -- include/configs/socfpga_socrates.h | 2 -- include/configs/socfpga_sr1500.h | 11 -------- include/configs/socfpga_vining_fpga.h | 18 ------------ 34 files changed, 101 insertions(+), 278 deletions(-)

2 3

[U-Boot] relationship between "ARM_DCC" and "DEBUG_UART_ARM_DCC"?
by Robert P. J. Day 23 Apr '19

23 Apr '19

just noticed the oddness regarding those two Kbuild symbols: $ git grep ARM_DCC arch/x86/include/asm/acpi_table.h:#define ACPI_DBG2_ARM_DCC 0x000F configs/zynq_cse_qspi_defconfig:CONFIG_DEBUG_UART_ARM_DCC=y drivers/serial/Kconfig:config DEBUG_UART_ARM_DCC drivers/serial/Makefile:obj-$(CONFIG_ARM_DCC) += arm_dcc.o drivers/serial/arm_dcc.c:#ifdef CONFIG_DEBUG_UART_ARM_DCC include/configs/xilinx_versal.h:#define CONFIG_ARM_DCC include/configs/xilinx_zynqmp.h:#define CONFIG_ARM_DCC include/configs/zynq-common.h:#define CONFIG_ARM_DCC scripts/config_whitelist.txt:CONFIG_ARM_DCC $ strangely(?), there is a Kbuild config directive for the latter: drivers/serial/Kconfig:config DEBUG_UART_ARM_DCC but not the former, which is just hardcoded in some xilinx-associated header files. is this deliberate? it just seems ... weird. rday -- ======================================================================== Robert P. J. Day Ottawa, Ontario, CANADA http://crashcourse.ca Twitter: http://twitter.com/rpjday LinkedIn: http://ca.linkedin.com/in/rpjday ========================================================================

2 3

[U-Boot] [OT] trying to boot zedboard just to u-boot, having no luck
by Robert P. J. Day 23 Apr '19

23 Apr '19

i've asked on the meta-xilinx mailing list, and rummaged around the net, and have failed to do something that should be straightforward -- i want to simply boot an avnet zedboard to a recent u-boot and stop there; i have no need to go any further for the time being. i have no problem configuring and building u-boot using: $ make zynq_zed_defconfig and as a test, i've tried to download any pre-built SD card images/files just to verify that the board boots, with absolutely no luck. i've formatted and re-formatted SD cards, populated them with allegedly working image files, checked and rechecked jumper settings to boot from SD card, used minicom and picocom to connect to the UART port ... all i ever see is the power LED come on, and never anything else. if someone can point me at a recipe that is verified to work to get u-boot up and running on this zedboard, i would be eternally grateful, because i'm running out of ideas. rday -- ======================================================================== Robert P. J. Day Ottawa, Ontario, CANADA http://crashcourse.ca Twitter: http://twitter.com/rpjday LinkedIn: http://ca.linkedin.com/in/rpjday ========================================================================

2 2