July 2018 - U-Boot

[U-Boot] [UBOOT PATCH v5 0/4] spi:xilinx_spi: Modify xilinx spi driver
by Vipul Kumar 16 Jul '18

16 Jul '18

This series of patches do the following: - This patch added support to get reg base address from DTS file - Added rxfifo() and txfifo() functions to add the modularity - Added support to read JEDEC-id twice at the boot time - Added live-tree support Changes in v2: - Split single patch into the series of patches Changes in v3: - Read reg in probe function - Removed xilinx_spi_ofdata_to_platdata function - Added fifo_depth read support in 2/3 and removed from 1/3 Changes in v4: - Moved static startup variable to private structure - Added live-tree support Changes in v5: - Changed function name to xilinx_spi_startup_block() Michal Simek (1): spi: xilinx: Read reg base address from DTS file Vipul Kumar (3): spi: xilinx_spi: Modify transfer logic xilinx_spi_xfer() function spi: xilinx_spi: Added support to read JEDEC-id twice at the boot time spi: xilinx_spi: convert to livetree drivers/spi/xilinx_spi.c | 152 ++++++++++++++++++++++++++++++++++------------- 1 file changed, 110 insertions(+), 42 deletions(-) -- 2.7.4

2 6

[U-Boot] [PATCH v6 1/6] sunxi: R40: add gigabit ethernet clocks
by Lothar Felten 16 Jul '18

16 Jul '18

Add clock control entries for the gigabit interface of the Allwinner R40/V40 CPU Acked-by: Maxime Ripard <maxime.ripard(a)bootlin.com> Reviewed-by: Joe Hershberger <joe.hershberger(a)ni.com> Reviewed-by: Jagan Teki <jagan(a)openedev.com> Tested-by: Jagan Teki <jagan(a)openedev.com> Signed-off-by: Lothar Felten <lothar.felten(a)gmail.com> --- Changelog: new in v2 v2->v3->4->v5->v6: none --- arch/arm/include/asm/arch-sunxi/clock_sun6i.h | 10 +++++++++- 1 file changed, 9 insertions(+), 1 deletion(-) diff --git a/arch/arm/include/asm/arch-sunxi/clock_sun6i.h b/arch/arm/include/asm/arch-sunxi/clock_sun6i.h index 3a59016955..ee387127f3 100644 --- a/arch/arm/include/asm/arch-sunxi/clock_sun6i.h +++ b/arch/arm/include/asm/arch-sunxi/clock_sun6i.h @@ -60,7 +60,11 @@ struct sunxi_ccm_reg { u32 reserved11; u32 sata_clk_cfg; /* 0xc8 SATA clock control (R40 only) */ u32 usb_clk_cfg; /* 0xcc USB clock control */ - u32 gmac_clk_cfg; /* 0xd0 GMAC clock control */ +#ifdef CONFIG_MACH_SUN8I_R40 + u32 cir0_clk_cfg; /* 0xd0 CIR0 clock control (R40 only) */ +#else + u32 gmac_clk_cfg; /* 0xd0 GMAC clock control (not for R40) */ +#endif u32 reserved12[7]; u32 mdfs_clk_cfg; /* 0xf0 MDFS clock control */ u32 dram_clk_cfg; /* 0xf4 DRAM configuration clock control */ @@ -103,7 +107,11 @@ struct sunxi_ccm_reg { u32 mtc_clk_cfg; /* 0x158 MTC module clock */ u32 mbus0_clk_cfg; /* 0x15c MBUS0 module clock */ u32 mbus1_clk_cfg; /* 0x160 MBUS1 module clock */ +#ifdef CONFIG_MACH_SUN8I_R40 + u32 gmac_clk_cfg; /* 0x164 GMAC clock control (R40 only) */ +#else u32 reserved16; +#endif u32 mipi_dsi_clk_cfg; /* 0x168 MIPI DSI clock control */ u32 mipi_csi_clk_cfg; /* 0x16c MIPI CSI clock control */ u32 reserved17[4]; -- 2.14.1

3 7

[U-Boot] [PATCH 1/2] dm: mmc: sunxi: Add A10/A20 compatible strings
by Adam Sampson 16 Jul '18

16 Jul '18

Commit dd27918c2252 ("dm: mmc: sunxi: Add support for driver model") only added the allwinner,sun5i-a13-mmc compatible string for this driver. The DM initialisation code here also works with (at least) A10 and A20, so add the appropriate compatible strings as per Linux 4.17's driver. Tested on A10 Cubieboard and A20 pcDuino3 Nano with CONFIG_DM_MMC. (A20 worked already, because sun7i-a20.dtsi specifies both the A13 and A20 strings.) Signed-off-by: Adam Sampson <ats(a)offog.org> --- drivers/mmc/sunxi_mmc.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/drivers/mmc/sunxi_mmc.c b/drivers/mmc/sunxi_mmc.c index fe6d82c7b4..7fa1ae8b16 100644 --- a/drivers/mmc/sunxi_mmc.c +++ b/drivers/mmc/sunxi_mmc.c @@ -637,7 +637,9 @@ static int sunxi_mmc_bind(struct udevice *dev) } static const struct udevice_id sunxi_mmc_ids[] = { + { .compatible = "allwinner,sun4i-a10-mmc" }, { .compatible = "allwinner,sun5i-a13-mmc" }, + { .compatible = "allwinner,sun7i-a20-mmc" }, { } }; -- 2.18.0

3 10

[U-Boot] [PATCH v7 0/6] sunxi: sync H3, H5, A64 DTs from mainline Linux
by Andre Przywara 16 Jul '18

16 Jul '18

This is an updated version of the series which brings the exact mainline Linux device tree files for various Allwinner boards into U-Boot. Apart from using the usually more correct reference DT files, this offers the big benefit of being able to use U-Boot's DT copy for directly passing it to the kernel. This avoids to actually load a .dtb file from somewhere, and allows seamless and automatic UEFI booting, so distribution installer images should just work (TM). This covers the ARMv8 SoCs (H5, A64), but also all boards with the H3, as this is somewhat married to the H5 and can only be updated together. The H3 and H5 DT files have diverged quite a bit, but as U-Boot's own usage of the DT is (yet) quite limited, there should be no regressions. The patches are split to first update the SoC .dtsi file, then the board .dts files in a second patch. They are grouped to handle the A64 first, then the H5 and H3. I put the respective kernel commit IDs in the commit messages. Patch 6 brings in the mainline DT for the SoPine baseboard, for which we didn't have a separate .dts in U-Boot so far. This is based on origin/master, and requires the USB shutdown fix. Please let me know what I should it base it against, maybe closer to the time we actually want to merge this (to avoid playing cat and mouse here). Maxime, I kept you Acked-by: from the previous posts, as I literally just updated to the latest Linux master, which went through your review anyway. Hope that's OK for you. Cheers, Andre. Changelog v6 .. v7: - drop Pine64-LTS DT (will send it to Linux first) - update to Linux v4.18-rc3 Changelog v5 .. v6: - bring back DT update patches - update to Linux v4.17-rc5 Changelog v4 .. v5: - drop Linux DT update patches for now - fix minor checkpatch complaints Changelog v3 .. v4: - remove MMC environment for all Allwinner boards (including 32 bit ones) - keep MMC environment offset to the old values - drop DT adjustments to use fixed MMC regulator Changelog v2 .. v3: 01: added, was on the list before 02: drop redundant H5 line 03-08: unchanged 09-20: added Changelog v1 .. v2: 01, 02, 03: unchanged 04, 05, 06, 07: added Andre Przywara (6): sunxi: DT: A64: update device tree file for Allwinner A64 SoC sunxi: DT: A64: update board .dts files from Linux sunxi: DT: update device tree files for Allwinner H3 and H5 SoCs sunxi: DT: H5: update board .dts files from Linux sunxi: DT: H3: update board .dts files from Linux sunxi: DT: A64: add proper SoPine baseboard device tree arch/arm/dts/Makefile | 3 +- arch/arm/dts/axp803.dtsi | 155 +++++ arch/arm/dts/sun50i-a64-bananapi-m64.dts | 195 +++++- arch/arm/dts/sun50i-a64-nanopi-a64.dts | 111 ++- arch/arm/dts/sun50i-a64-olinuxino.dts | 153 ++++- arch/arm/dts/sun50i-a64-orangepi-win.dts | 135 +++- arch/arm/dts/sun50i-a64-pine64-plus-u-boot.dtsi | 59 -- arch/arm/dts/sun50i-a64-pine64-plus.dts | 17 +- arch/arm/dts/sun50i-a64-pine64.dts | 186 ++++- arch/arm/dts/sun50i-a64-sopine-baseboard.dts | 150 +++++ arch/arm/dts/sun50i-a64-sopine.dtsi | 142 ++++ arch/arm/dts/sun50i-a64.dtsi | 305 ++++++++- arch/arm/dts/sun50i-h5-libretech-all-h3-cc.dts | 5 +- arch/arm/dts/sun50i-h5-nanopi-neo-plus2.dts | 122 +++- arch/arm/dts/sun50i-h5-nanopi-neo2.dts | 91 ++- arch/arm/dts/sun50i-h5-orangepi-pc2.dts | 151 ++++- arch/arm/dts/sun50i-h5-orangepi-prime.dts | 176 ++++- arch/arm/dts/sun50i-h5-orangepi-zero-plus2.dts | 64 +- arch/arm/dts/sun50i-h5.dtsi | 38 +- arch/arm/dts/sun8i-h2-plus-orangepi-zero.dts | 55 +- arch/arm/dts/sun8i-h3-bananapi-m2-plus.dts | 96 ++- arch/arm/dts/sun8i-h3-libretech-all-h3-cc.dts | 3 +- arch/arm/dts/sun8i-h3-nanopi-m1-plus.dts | 78 +++ arch/arm/dts/sun8i-h3-nanopi-m1.dts | 42 ++ arch/arm/dts/sun8i-h3-nanopi-neo-air.dts | 28 +- arch/arm/dts/sun8i-h3-nanopi-neo.dts | 17 + arch/arm/dts/sun8i-h3-nanopi.dtsi | 13 +- arch/arm/dts/sun8i-h3-orangepi-2.dts | 92 ++- arch/arm/dts/sun8i-h3-orangepi-lite.dts | 57 +- arch/arm/dts/sun8i-h3-orangepi-one.dts | 116 +++- arch/arm/dts/sun8i-h3-orangepi-pc-plus.dts | 11 +- arch/arm/dts/sun8i-h3-orangepi-pc.dts | 141 +++- arch/arm/dts/sun8i-h3-orangepi-plus.dts | 29 +- arch/arm/dts/sun8i-h3-orangepi-plus2e.dts | 15 +- arch/arm/dts/sun8i-h3.dtsi | 602 +++-------------- arch/arm/dts/sunxi-h3-h5.dtsi | 860 ++++++++++++++++++++++++ arch/arm/dts/sunxi-libretech-all-h3-cc.dtsi | 2 - configs/sopine_baseboard_defconfig | 2 +- include/dt-bindings/clock/sun8i-de2.h | 18 + include/dt-bindings/clock/sun8i-r-ccu.h | 59 ++ include/dt-bindings/reset/sun8i-de2.h | 14 + include/dt-bindings/reset/sun8i-r-ccu.h | 53 ++ 42 files changed, 3741 insertions(+), 920 deletions(-) create mode 100644 arch/arm/dts/axp803.dtsi delete mode 100644 arch/arm/dts/sun50i-a64-pine64-plus-u-boot.dtsi create mode 100644 arch/arm/dts/sun50i-a64-sopine-baseboard.dts create mode 100644 arch/arm/dts/sun50i-a64-sopine.dtsi create mode 100644 arch/arm/dts/sunxi-h3-h5.dtsi create mode 100644 include/dt-bindings/clock/sun8i-de2.h create mode 100644 include/dt-bindings/clock/sun8i-r-ccu.h create mode 100644 include/dt-bindings/reset/sun8i-de2.h create mode 100644 include/dt-bindings/reset/sun8i-r-ccu.h -- 2.14.4

2 7

Re: [U-Boot] [Arm.ebbr-discuss] EBBR v0.6 Release Announcement
by Michal Simek 16 Jul '18

16 Jul '18

- ancient mailing list + proper u-boot mailing list. M On 14.7.2018 00:34, Grant Likely wrote: > I'm pleased to announce the release of version 0.6 of the Embedded Base > Boot Requirements (EBBR) specification. > > https://github.com/ARM-software/ebbr/releases/tag/v0.6 > > EBBR is a new specification defining a standard boot environment > suitable for full feature operating systems running on embedded > platforms... > > Well, it will when it's finished. > > It is well know that firmware for embedded systems is a fragmented area > with each platform behaving in subtly incompatible ways. It is also > completely different from the firmware interface used on general purpose > desktops and servers. For OSes, this makes supporting more than a > handful of platforms a nearly impossible affair. EBBR aims to solve this > problem by defining a standard boot interface that can easily be > implemented using either U-Boot or Tianocore, and is based on the same > UEFI specification used on general purpose computers. > > By adopting EBBR, platform vendors can reduce the amount of engineering > effort required to support their products and make them easier to use. > As EBBR is being developed in conjunction with the U-Boot, Tianocore, > and Trusted Firmware projects, most of the functionality required is > already implemented and ready to be used if one uses an up to date > release of U-Boot or Tianocore. > > For OS vendors, this makes far easier to support embedded platforms > because they don't need to tailor the boot process for each platform. > The same boot infrastructure works on both desktop/servers and on EBBR > compliant embedded platforms. > > And finally, for end users, working with an EBBR compliant platform > means they can boot the OS of their choice without needing to learn low > level details of the platform firmware. > > This v0.6 release of EBBR is a pre-release document. The contents are > not final. The purpose of this release is to raise awareness of the EBBR > project, and to solicit feedback on the current draft. Please do read > and provide comments on the boot-architecture(a)lists.linaro.org mailing > list. > > The plan is to release v1.0 before the end of the 2018. > > Thanks to the EBBR committee members who contributed to this release: > > Andreas Färber (SUSE) > Alex Graf (SUSE) > Ryan Harkin (Linaro) > Rob Herring (Linaro) > Udit Kumar (NXP) > Leif Lindholm (Linaro) > Bill Mills (TI) > Peter Robinson (Red Hat) > Tom Rini (Konsulko) > Daniel Thompson (Linaro) > Dong Wei (Arm) > > Sincerely, > Grant Likely, EBBR committee co-chair > > > Note on U-Boot implementations > ------------------------------ > It is expected that EBBR compliant can be achieved by using a recent > version of U-Boot with the appropriate configuration options. An > implementers guide for U-Boot will be written before EBBR v1.0 is released. > > There is also work ongoing to get the UEFI Self Certification Test > running on U-Boot. Once working, this will be a tool for vendors to test > their platforms for EBBR compliance. > > FAQ > --- > 1. Does EBBR define a new interface? > > No. EBBR builds on the existing UEFI spec by requiring a specific > subset that can be implemented today using U-Boot, and either Devicetree > or ACPI. > > 2. Does EBBR require Devicetree? ACPI? > > EBBR allows platforms to provide either ACPI or Devicetree. Linux > supports both system description languages equally well, and Devicetree > is in common use on embedded platforms. As long as the platform supplies > a system description that can boot a mainline operating system. > > EBBR does not attempt to define a common base standard for Devicetree > platforms because of the wide variety of platforms needed to be > supported. The one assumption EBBR does make is that the target > operating system already has support for the SoC on the platform. > > 3. Is EBBR only for U-Boot and Linux embedded systems? > > No. While U-Boot+Linux platforms were certainly the primary audience > when EBBR was first conceived, the spec is very purposefully written to > be OS-independent. EBBR requires specific interfaces, but those > interface can be implemented by any firmware project. > > We would absolutely like to have review, feedback and contributions > from non-Linux, non-U-Boot users. > > 4. Can I contribute to the EBBR specification? > > Yes. The EBBR source document is on GitHub, and we use the > boot-architecture(a)lists.linaro.org mailing list. > > https://github.com/ARM-Software/ebbr > _______________________________________________ > Arm.ebbr-discuss mailing list > Arm.ebbr-discuss(a)arm.com

1 0

[U-Boot] sandbox: Net: No ethernet found.
by Heinrich Schuchardt 16 Jul '18

16 Jul '18

Hello Simon, when I run sandbox_defconfig I always get Net: No ethernet found. The dm command shows no driver attached to uclass 21: eth. Why are the sandbox and sandbox-raw drivers not loaded? device_bind_common is not called for these drivers. arch/sandbox/dts/sandbox.dts makes assumptions about the naming of interfaces: host-raw-interface = "eth0" With systemd eth0 tends not to exist. The interfaces on my Debian Buster system the interfaces are called lo, enp0s25, and wlp12s0. Shouldn't we use if_nameindex() to enumerate the interfaces? Best regards Heinrich

3 7

[U-Boot] Microblaze gpio reset handling
by Michal Simek 16 Jul '18

16 Jul '18

Hi Simon, I have looked at converting the rest of microblaze drivers to DM and I am curious why I can't use top level compatible string in DM based drivers. I am looking at ways how to move gpio and microblaze soft reset to sysreset framework and I see that core is not able to find out my high level xlnx,microblaze compatible string. / { #address-cells = <1>; #size-cells = <1>; compatible = "xlnx,microblaze"; model = "Xilinx MicroBlaze"; hard-reset-gpios = <&reset_gpio 0 1>; As far I have seen only nodes are checked but unfortunately Microblaze is using hard-reset-gpios in root for years and will be good to keep it like that. I have already tested to bind that driver from board file like device_bind_driver(gd->dm_root, "microblaze_sysreset", "reset", NULL); which is working fine but just asking if there is another solution for this. Thanks, Michal

2 1

[U-Boot] [PATCH 1/2] test: Fix sandbox_spl test filter
by Simon Glass 15 Jul '18

15 Jul '18

This filter does not match the test it is intended to anymore. Update it so that it works again. Signed-off-by: Simon Glass <sjg(a)chromium.org> --- test/run | 3 +-- 1 file changed, 1 insertion(+), 2 deletions(-) diff --git a/test/run b/test/run index b2d24e3f8ed..eb1563d13ed 100755 --- a/test/run +++ b/test/run @@ -11,8 +11,7 @@ result=0 run_test ./test/py/test.py --bd sandbox --build # Run tests which require sandbox_spl -run_test ./test/py/test.py --bd sandbox_spl --build -k \ - test/py/tests/test_ofplatdata.py +run_test ./test/py/test.py --bd sandbox_spl --build -k test_ofplatdata.py # Run tests for the flat DT version of sandbox ./test/py/test.py --bd sandbox_flattree --build -- 2.17.0.441.gb46fe60e1d-goog

2 6

[U-Boot] [PATCH] gpio: zynq: Read of mach data in platdata with dev_get_driver_data
by Michal Simek 15 Jul '18

15 Jul '18

Remove bogus zynq_gpio_getplat_data() and read driver data directly. Signed-off-by: Michal Simek <michal.simek(a)xilinx.com> --- Driver should use platdata structure but this change should be fine before this is fixed. Simon: Is ofdata_to_platdata correct function where this should be read? --- drivers/gpio/zynq_gpio.c | 29 ++--------------------------- 1 file changed, 2 insertions(+), 27 deletions(-) diff --git a/drivers/gpio/zynq_gpio.c b/drivers/gpio/zynq_gpio.c index 9a915544784d..26f69b1a713f 100644 --- a/drivers/gpio/zynq_gpio.c +++ b/drivers/gpio/zynq_gpio.c @@ -15,8 +15,6 @@ #include <dm.h> #include <fdtdec.h> -DECLARE_GLOBAL_DATA_PTR; - /* Maximum banks */ #define ZYNQ_GPIO_MAX_BANK 4 @@ -334,36 +332,11 @@ static const struct udevice_id zynq_gpio_ids[] = { { } }; -static void zynq_gpio_getplat_data(struct udevice *dev) -{ - const struct udevice_id *of_match = zynq_gpio_ids; - int ret; - struct zynq_gpio_privdata *priv = dev_get_priv(dev); - - while (of_match->compatible) { - ret = fdt_node_offset_by_compatible(gd->fdt_blob, -1, - of_match->compatible); - if (ret >= 0) { - priv->p_data = - (struct zynq_platform_data *)of_match->data; - break; - } else { - of_match++; - continue; - } - } - - if (!priv->p_data) - printf("No Platform data found\n"); -} - static int zynq_gpio_probe(struct udevice *dev) { struct zynq_gpio_privdata *priv = dev_get_priv(dev); struct gpio_dev_priv *uc_priv = dev_get_uclass_priv(dev); - zynq_gpio_getplat_data(dev); - if (priv->p_data) uc_priv->gpio_count = priv->p_data->ngpio; @@ -376,6 +349,8 @@ static int zynq_gpio_ofdata_to_platdata(struct udevice *dev) priv->base = (phys_addr_t)dev_read_addr(dev); + priv->p_data = (struct zynq_platform_data *)dev_get_driver_data(dev); + return 0; } -- 1.9.1

2 1

[U-Boot] [PATCH] gpio: dm: Support manual relocation for gpio
by Michal Simek 15 Jul '18

15 Jul '18

Relocate gpio ops as was done by: "dm: Add support for all targets which requires MANUAL_RELOC" (sha1: 484fdf5ba058b07be5ca82763aa2b72063540ef3) Signed-off-by: Michal Simek <michal.simek(a)xilinx.com> --- drivers/gpio/gpio-uclass.c | 35 +++++++++++++++++++++++++++++++++++ 1 file changed, 35 insertions(+) diff --git a/drivers/gpio/gpio-uclass.c b/drivers/gpio/gpio-uclass.c index e9d08e2ebc25..da5e9ba6e524 100644 --- a/drivers/gpio/gpio-uclass.c +++ b/drivers/gpio/gpio-uclass.c @@ -854,11 +854,46 @@ static int gpio_pre_remove(struct udevice *dev) return gpio_renumber(dev); } +static int gpio_post_bind(struct udevice *dev) +{ +#if defined(CONFIG_NEEDS_MANUAL_RELOC) + struct dm_gpio_ops *ops = (struct dm_gpio_ops *)device_get_ops(dev); + static int reloc_done; + + if (!reloc_done) { + if (ops->request) + ops->request += gd->reloc_off; + if (ops->free) + ops->free += gd->reloc_off; + if (ops->direction_input) + ops->direction_input += gd->reloc_off; + if (ops->direction_output) + ops->direction_output += gd->reloc_off; + if (ops->get_value) + ops->get_value += gd->reloc_off; + if (ops->set_value) + ops->set_value += gd->reloc_off; + if (ops->get_open_drain) + ops->get_open_drain += gd->reloc_off; + if (ops->set_open_drain) + ops->set_open_drain += gd->reloc_off; + if (ops->get_function) + ops->get_function += gd->reloc_off; + if (ops->xlate) + ops->xlate += gd->reloc_off; + + reloc_done++; + } +#endif + return 0; +} + UCLASS_DRIVER(gpio) = { .id = UCLASS_GPIO, .name = "gpio", .flags = DM_UC_FLAG_SEQ_ALIAS, .post_probe = gpio_post_probe, + .post_bind = gpio_post_bind, .pre_remove = gpio_pre_remove, .per_device_auto_alloc_size = sizeof(struct gpio_dev_priv), }; -- 1.9.1

2 1