December 2017 - U-Boot

[U-Boot] [PATCH] arm64 :show_regs: show the address before relocation
by Peng Fan 06 Feb '18

06 Feb '18

After relocation, when error happends, it is hard to track ELR and LR with asm file objdumped from elf file. So subtract the gd->reloc_off the reflect the compliation address. Signed-off-by: Peng Fan <peng.fan(a)nxp.com> --- arch/arm/lib/interrupts_64.c | 10 ++++++++-- 1 file changed, 8 insertions(+), 2 deletions(-) diff --git a/arch/arm/lib/interrupts_64.c b/arch/arm/lib/interrupts_64.c index 7c9cfce69f..cbcfeec2b0 100644 --- a/arch/arm/lib/interrupts_64.c +++ b/arch/arm/lib/interrupts_64.c @@ -9,6 +9,7 @@ #include <linux/compiler.h> #include <efi_loader.h> +DECLARE_GLOBAL_DATA_PTR; int interrupt_init(void) { @@ -29,8 +30,13 @@ void show_regs(struct pt_regs *regs) { int i; - printf("ELR: %lx\n", regs->elr); - printf("LR: %lx\n", regs->regs[30]); + if (gd->flags & GD_FLG_RELOC) { + printf("ELR: %lx\n", regs->elr - gd->reloc_off); + printf("LR: %lx\n", regs->regs[30] - gd->reloc_off); + } else { + printf("ELR: %lx\n", regs->elr); + printf("LR: %lx\n", regs->regs[30]); + } for (i = 0; i < 29; i += 2) printf("x%-2d: %016lx x%-2d: %016lx\n", i, regs->regs[i], i+1, regs->regs[i+1]); -- 2.14.1

4 6

[U-Boot] [PATCH 00/16] efi_loader: implement driver management
by Heinrich Schuchardt 04 Feb '18

04 Feb '18

This series implements the OpenProtocolInformation, ConnectController, and DisconnectController boot services. The EFI application creating a new device will call ConnectController to request the installation of all necessary drivers. Before deleting the device it will call DisconectController to remove all child controllers and to detach the driver. E.g. iPXE may be used to connect an iSCSI target. It then creates a handle by installing the EFI_BLOCK_IO_PROTOCOL. It calls ConnectController and expects U-Boot to create the partition handles and to provide the EFI_SIMPLE_FILE_SYSTEM_PROTOCOL. (The driver connecting to the EFI_BLOCK_IO_PROTOCOL is not yet implemented in U-Boot.) A self test is provided for the driver management. The bootefi selftest now creates colored output. Two bug fixes are included. To complete the implementation of the EFI boot services the following future work is needed: * manage events in a linked list (see Rob's proposal) * implement ReinstallProtocol and RegisterProtocolNotify Heinrich Schuchardt (16): efi_selftest: colored test output efi_loader: simplify efi_remove_all_protocols efi_selftest: do not try to close device path protocol efi_loader: list of open protocol infos efi_loader: open_info in OpenProtocol efi_loader: open_info in CloseProtocol efi_loader: implement OpenProtocolInformation efi_loader: debug output installed device path efi_loader: implement ConnectController efi_loader: fix signature of efi_disconnect_controller efi_loader: implement DisconnectController efi_loader: disconnect controllers in UninstallProtocol efi_selftest: remove todo in manage protocols efi_selftest: remove todo in device path test efi_selftest: test for (Dis)ConnectController efi_loader: consistently use efi_handle_t for handles cmd/bootefi.c | 10 +- include/efi_api.h | 48 +- include/efi_loader.h | 31 +- include/efi_selftest.h | 27 +- lib/efi_loader/efi_boottime.c | 757 +++++++++++++++++++++--- lib/efi_loader/efi_console.c | 6 +- lib/efi_selftest/Makefile | 1 + lib/efi_selftest/efi_selftest.c | 25 +- lib/efi_selftest/efi_selftest_console.c | 13 +- lib/efi_selftest/efi_selftest_controllers.c | 385 ++++++++++++ lib/efi_selftest/efi_selftest_devicepath.c | 48 +- lib/efi_selftest/efi_selftest_manageprotocols.c | 22 +- 12 files changed, 1209 insertions(+), 164 deletions(-) create mode 100644 lib/efi_selftest/efi_selftest_controllers.c -- 2.14.2

2 35

[U-Boot] [PATCH 06/15] spl: fit: implement recording of loadables into /fit-images
by Philipp Tomsich 02 Feb '18

02 Feb '18

If a FDT was loaded (e.g. to append it to U-Boot image), we store it's address and record information for all loadables into this FDT. This allows us to easily keep track of images for multiple privilege levels (e.g. with ATF) or of firmware images preloaded into temporary locations (e.g. PMU firmware that may overlap the SPL stage). Signed-off-by: Philipp Tomsich <philipp.tomsich(a)theobroma-systems.com> --- common/spl/spl_fit.c | 95 ++++++++++++++++++++++++++++++++++++++++++++-------- 1 file changed, 81 insertions(+), 14 deletions(-) diff --git a/common/spl/spl_fit.c b/common/spl/spl_fit.c index 9f05e1e..6dc0969 100644 --- a/common/spl/spl_fit.c +++ b/common/spl/spl_fit.c @@ -2,7 +2,7 @@ * Copyright (C) 2016 Google, Inc * Written by Simon Glass <sjg(a)chromium.org> * - * SPDX-License-Identifier: GPL-2.0+ + * SPDX-License-Identifier: GPL-2.0+ */ #include <common.h> @@ -16,22 +16,24 @@ #endif /** - * spl_fit_get_image_node(): By using the matching configuration subnode, + * spl_fit_get_image_name(): By using the matching configuration subnode, * retrieve the name of an image, specified by a property name and an index * into that. * @fit: Pointer to the FDT blob. * @images: Offset of the /images subnode. * @type: Name of the property within the configuration subnode. * @index: Index into the list of strings in this property. + * @outname: Name of the image * - * Return: the node offset of the respective image node or a negative - * error number. + * Return: 0 on success, or a negative error number */ -static int spl_fit_get_image_node(const void *fit, int images, - const char *type, int index) +static int spl_fit_get_image_name(const void *fit, int images, + const char *type, int index, + char **outname) { const char *name, *str; - int node, conf_node; + __maybe_unused int node; + int conf_node; int len, i; conf_node = fit_find_config_node(fit); @@ -63,7 +65,35 @@ static int spl_fit_get_image_node(const void *fit, int images, } } + *outname = (char *)str; + return 0; +} + +/** + * spl_fit_get_image_node(): By using the matching configuration subnode, + * retrieve the name of an image, specified by a property name and an index + * into that. + * @fit: Pointer to the FDT blob. + * @images: Offset of the /images subnode. + * @type: Name of the property within the configuration subnode. + * @index: Index into the list of strings in this property. + * + * Return: the node offset of the respective image node or a negative + * error number. + */ +static int spl_fit_get_image_node(const void *fit, int images, + const char *type, int index) +{ + char *str; + int err; + int node; + + err = spl_fit_get_image_name(fit, images, type, index, &str); + if (err) + return err; + debug("%s: '%s'\n", type, str); + node = fdt_subnode_offset(fit, images, str); if (node < 0) { debug("cannot find image node '%s': %d\n", str, node); @@ -116,15 +146,15 @@ static int get_aligned_image_size(struct spl_load_info *info, int data_size, * @info: points to information about the device to load data from * @sector: the start sector of the FIT image on the device * @fit: points to the flattened device tree blob describing the FIT - * image + * image * @base_offset: the beginning of the data area containing the actual * image data, relative to the beginning of the FIT * @node: offset of the DT node describing the image to load (relative - * to @fit) + * to @fit) * @image_info: will be filled with information about the loaded image - * If the FIT node does not contain a "load" (address) property, - * the image gets loaded to the address pointed to by the - * load_addr member in this struct. + * If the FIT node does not contain a "load" (address) property, + * the image gets loaded to the address pointed to by the + * load_addr member in this struct. * * Return: 0 on success or a negative error number. */ @@ -236,6 +266,35 @@ static int spl_fit_append_fdt(struct spl_image_info *spl_image, image_info.load_addr = spl_image->load_addr + spl_image->size; ret = spl_load_fit_image(info, sector, fit, base_offset, node, &image_info); + + if (ret < 0) + return ret; + + /* Make the load-address of the FDT available for the SPL framework */ + spl_image->fdt_addr = (void *)image_info.load_addr; + /* Try to make space, so we can inject details on the loadables */ + ret = fdt_shrink_to_minimum(spl_image->fdt_addr, 8192); + + return ret; +} + +static int spl_fit_record_loadable(const void *fit, int images, int index, + void *blob, struct spl_image_info *image) +{ + char *name; + int node, ret; + + ret = spl_fit_get_image_name(fit, images, "loadables", + index, &name); + if (ret < 0) + return ret; + + node = spl_fit_get_image_node(fit, images, "loadables", index); + + ret = fdt_record_loadable(blob, index, name, image->load_addr, + image->size, image->entry_point, + fdt_getprop(fit, node, "type", NULL), + fdt_getprop(fit, node, "os", NULL)); return ret; } @@ -361,9 +420,11 @@ int spl_load_simple_fit(struct spl_image_info *spl_image, if (!fit_image_get_os(fit, node, &os_type)) debug("Loadable is %s\n", genimg_get_os_name(os_type)); - if (spl_image->os == IH_OS_U_BOOT) - spl_fit_append_fdt(spl_image, info, sector, + if (os_type == IH_OS_U_BOOT) { + spl_fit_append_fdt(&image_info, info, sector, fit, images, base_offset); + spl_image->fdt_addr = image_info.fdt_addr; + } /* * If the "firmware" image did not provide an entry point, @@ -372,6 +433,12 @@ int spl_load_simple_fit(struct spl_image_info *spl_image, if (spl_image->entry_point == FDT_ERROR && image_info.entry_point != FDT_ERROR) spl_image->entry_point = image_info.entry_point; + + /* Record our loadables into the FDT */ + if (spl_image->fdt_addr) + spl_fit_record_loadable(fit, images, index, + spl_image->fdt_addr, + &image_info); } /* -- 2.1.4

4 7

[U-Boot] [PATCH v2 00/15] dm: sata: mmc: Convert a sunxi board to driver model for MMC, SATA
by Simon Glass 01 Feb '18

01 Feb '18

At present SCSI support for driver model works only with PCI controllers. This series makes the following changes: - Adjusts SATA/AHCI support to work with non-PCI controllers - Allows driver model to be used for MMC in U-Boot proper without requiring it to be used in SPL - Adjusts sunxi MMC and SATA drivers to support driver model - Enables this on a single board (Linksprite_pcDuino3) Even with this series the AHCI uclass still does not have any operations. This will be needed for drivers which do not use SCSI. Further work is needed to complete this. Changes in v2: - Drop debugging printf() now that the card detect is working - Add new patch to correct polarity of MMC card detect on Linksprite_pcDuino3 Simon Glass (15): ahci: Support non-PCI controllers dm: mmc: Allow disabling driver model in SPL fdt: Correct fdt_get_base_address() dm: scsi: Drop duplicate SCSI and DM_SCSI options dm: ahci: Correct uclass private data dm: mmc: sunxi: Rename struct sunxi_mmc_host to sunxi_mmc_priv dm: mmc: sunxi: Rename mmchost to priv dm: mmc: sunxi: Pass private data around explicitly dm: mmc: sunxi: Drop mmc_clk_io_on() dm: mmc: sunxi: Add support for driver model dm: sunxi: Linksprite_pcDuino3: Correct polarity of MMC card detect dm: scsi: Don't scan the SCSI bus when probing dm: sunxi: sata: Don't build sata support into SPL dm: sata: sunxi: Add support for driver model dm: sunxi: Move Linksprite_pcDuino3 to use DM for MMC, SATA arch/arm/dts/sun7i-a20-pcduino3.dts | 2 +- arch/x86/cpu/ivybridge/sata.c | 2 +- board/sunxi/Makefile | 2 + board/sunxi/ahci.c | 61 +++++- common/fdt_support.c | 7 +- common/spl/spl_mmc.c | 4 +- configs/Linksprite_pcDuino3_defconfig | 7 +- drivers/ata/Kconfig | 18 -- drivers/ata/ahci-uclass.c | 2 + drivers/ata/ahci.c | 33 ++-- drivers/block/Kconfig | 12 ++ drivers/block/Makefile | 4 +- drivers/mmc/Kconfig | 21 ++ drivers/mmc/Makefile | 4 +- drivers/mmc/mmc-uclass.c | 6 +- drivers/mmc/mmc.c | 28 +-- drivers/mmc/mmc_legacy.c | 2 +- drivers/mmc/mmc_private.h | 6 +- drivers/mmc/omap_hsmmc.c | 20 +- drivers/mmc/sunxi_mmc.c | 359 +++++++++++++++++++++++----------- drivers/scsi/scsi.c | 2 +- include/ahci.h | 14 +- include/blk.h | 4 +- include/mmc.h | 12 +- 24 files changed, 431 insertions(+), 201 deletions(-) -- 2.13.2.725.g09c95d1e9-goog

9 35

[U-Boot] [PATCH] efi_loader: Fix partition offsets
by Alexander Graf 01 Feb '18

01 Feb '18

Commit 884bcf6f65 (efi_loader: use proper device-paths for partitions) tried to introduce the el torito scheme to all partition table types: Spawn individual disk objects for each partition on a disk. Unfortunately, that code ended up creating partitions with offset=0 which meant that anyone accessing these objects gets data from the raw block device instead of the partition. Furthermore, all the el torito logic to spawn devices for partitions was duplicated. So let's merge the two code paths and give partition disk objects good offsets to work from, so that payloads can actually make use of them. Fixes: 884bcf6f65 (efi_loader: use proper device-paths for partitions) Reported-by: Yousaf Kaukab <yousaf.kaukab(a)suse.com> Signed-off-by: Alexander Graf <agraf(a)suse.de> --- lib/efi_loader/efi_disk.c | 60 ++++++++++------------------------------------- 1 file changed, 13 insertions(+), 47 deletions(-) diff --git a/lib/efi_loader/efi_disk.c b/lib/efi_loader/efi_disk.c index 68ba2cf7b2..4e457a841b 100644 --- a/lib/efi_loader/efi_disk.c +++ b/lib/efi_loader/efi_disk.c @@ -264,21 +264,17 @@ out_of_memory: printf("ERROR: Out of memory\n"); } -static int efi_disk_create_eltorito(struct blk_desc *desc, - const char *if_typename, - int diskid, - const char *pdevname) +static int efi_disk_create_partitions(struct blk_desc *desc, + const char *if_typename, + int diskid, + const char *pdevname) { int disks = 0; -#if CONFIG_IS_ENABLED(ISO_PARTITION) char devname[32] = { 0 }; /* dp->str is u16[32] long */ disk_partition_t info; int part; - if (desc->part_type != PART_TYPE_ISO) - return 0; - - /* and devices for each partition: */ + /* Add devices for each partition */ for (part = 1; part <= MAX_SEARCH_PARTITIONS; part++) { if (part_get_info(desc, part, &info)) continue; @@ -289,10 +285,6 @@ static int efi_disk_create_eltorito(struct blk_desc *desc, disks++; } - /* ... and add block device: */ - efi_disk_add_dev(devname, if_typename, desc, diskid, 0, 0); -#endif - return disks; } @@ -318,31 +310,18 @@ int efi_disk_register(void) uclass_next_device_check(&dev)) { struct blk_desc *desc = dev_get_uclass_platdata(dev); const char *if_typename = dev->driver->name; - disk_partition_t info; - int part; printf("Scanning disk %s...\n", dev->name); - /* add devices for each partition: */ - for (part = 1; part <= MAX_SEARCH_PARTITIONS; part++) { - if (part_get_info(desc, part, &info)) - continue; - efi_disk_add_dev(dev->name, if_typename, desc, - desc->devnum, 0, part); - } - - /* ... and add block device: */ + /* Add block device for the full device */ efi_disk_add_dev(dev->name, if_typename, desc, desc->devnum, 0, 0); disks++; - /* - * El Torito images show up as block devices in an EFI world, - * so let's create them here - */ - disks += efi_disk_create_eltorito(desc, if_typename, - desc->devnum, dev->name); + /* Partitions show up as block devices in EFI */ + disks += efi_disk_create_partitions(desc, if_typename, + desc->devnum, dev->name); } #else int i, if_type; @@ -361,8 +340,6 @@ int efi_disk_register(void) for (i = 0; i < 4; i++) { struct blk_desc *desc; char devname[32] = { 0 }; /* dp->str is u16[32] long */ - disk_partition_t info; - int part; desc = blk_get_devnum_by_type(if_type, i); if (!desc) @@ -373,24 +350,13 @@ int efi_disk_register(void) snprintf(devname, sizeof(devname), "%s%d", if_typename, i); - /* add devices for each partition: */ - for (part = 1; part <= MAX_SEARCH_PARTITIONS; part++) { - if (part_get_info(desc, part, &info)) - continue; - efi_disk_add_dev(devname, if_typename, desc, - i, 0, part); - } - - /* ... and add block device: */ + /* Add block device for the full device */ efi_disk_add_dev(devname, if_typename, desc, i, 0, 0); disks++; - /* - * El Torito images show up as block devices - * in an EFI world, so let's create them here - */ - disks += efi_disk_create_eltorito(desc, if_typename, - i, devname); + /* Partitions show up as block devices in EFI */ + disks += efi_disk_create_partitions(desc, if_typename, + i, devname); } } #endif -- 2.12.3

4 13

[U-Boot] [PATCH] Convert CONFIG_SYS_BOOT_GET_{CMDLINE, KBD} to Kconfig
by Derald D. Woods 31 Jan '18

31 Jan '18

This converts the following to Kconfig: CONFIG_SYS_BOOT_GET_CMDLINE CONFIG_SYS_BOOT_GET_KBD Additionally, ARM now selects CONFIG_SYS_BOOT_GET_CMDLINE. This enables the 'boot_get_cmdline' call to be reached from 'image_setup_linux' in "common/image.c". Signed-off-by: Derald D. Woods <woods.technical(a)gmail.com> --- Kconfig | 12 ++++++++++++ arch/Kconfig | 5 +++++ arch/m68k/include/asm/config.h | 2 -- arch/powerpc/include/asm/config.h | 2 -- scripts/config_whitelist.txt | 2 -- 5 files changed, 17 insertions(+), 6 deletions(-) diff --git a/Kconfig b/Kconfig index 9b8a807799..f713c6a9b1 100644 --- a/Kconfig +++ b/Kconfig @@ -83,6 +83,18 @@ config DISTRO_DEFAULTS Select this to enable various options and commands which are suitable for building u-boot for booting general purpose Linux distributions. +config SYS_BOOT_GET_CMDLINE + bool "Enable kernel command line setup" + help + Enables allocating and saving kernel cmdline in space between + "bootm_low" and "bootm_low" + BOOTMAPSZ. + +config SYS_BOOT_GET_KBD + bool "Enable kernel board information setup" + help + Enables allocating and saving a kernel copy of the bd_info in + space between "bootm_low" and "bootm_low" + BOOTMAPSZ. + config SYS_MALLOC_F bool "Enable malloc() pool before relocation" default y if DM diff --git a/arch/Kconfig b/arch/Kconfig index 0b12ed986c..41f61a46dc 100644 --- a/arch/Kconfig +++ b/arch/Kconfig @@ -22,10 +22,13 @@ config ARM select CREATE_ARCH_SYMLINK select HAVE_PRIVATE_LIBGCC if !ARM64 select SUPPORT_OF_CONTROL + select SYS_BOOT_GET_CMDLINE config M68K bool "M68000 architecture" select HAVE_PRIVATE_LIBGCC + select SYS_BOOT_GET_CMDLINE + select SYS_BOOT_GET_KBD config MICROBLAZE bool "MicroBlaze architecture" @@ -53,6 +56,8 @@ config PPC bool "PowerPC architecture" select HAVE_PRIVATE_LIBGCC select SUPPORT_OF_CONTROL + select SYS_BOOT_GET_CMDLINE + select SYS_BOOT_GET_KBD config SANDBOX bool "Sandbox" diff --git a/arch/m68k/include/asm/config.h b/arch/m68k/include/asm/config.h index 9c4d3fb8fd..fd0b5513ee 100644 --- a/arch/m68k/include/asm/config.h +++ b/arch/m68k/include/asm/config.h @@ -11,7 +11,5 @@ #define CONFIG_LMB #define CONFIG_SYS_BOOT_RAMDISK_HIGH -#define CONFIG_SYS_BOOT_GET_CMDLINE -#define CONFIG_SYS_BOOT_GET_KBD #endif diff --git a/arch/powerpc/include/asm/config.h b/arch/powerpc/include/asm/config.h index 6aec815c71..67e4b48a96 100644 --- a/arch/powerpc/include/asm/config.h +++ b/arch/powerpc/include/asm/config.h @@ -31,8 +31,6 @@ #define CONFIG_LMB #define CONFIG_SYS_BOOT_RAMDISK_HIGH -#define CONFIG_SYS_BOOT_GET_CMDLINE -#define CONFIG_SYS_BOOT_GET_KBD #ifndef CONFIG_MAX_MEM_MAPPED #if defined(CONFIG_E500) || \ diff --git a/scripts/config_whitelist.txt b/scripts/config_whitelist.txt index 43a4ff0892..8117d485ec 100644 --- a/scripts/config_whitelist.txt +++ b/scripts/config_whitelist.txt @@ -2322,8 +2322,6 @@ CONFIG_SYS_BOOTM_LEN CONFIG_SYS_BOOTPARAMS_LEN CONFIG_SYS_BOOTSZ CONFIG_SYS_BOOT_BLOCK -CONFIG_SYS_BOOT_GET_CMDLINE -CONFIG_SYS_BOOT_GET_KBD CONFIG_SYS_BOOT_RAMDISK_HIGH CONFIG_SYS_BR0_64M CONFIG_SYS_BR0_8M -- 2.15.1

2 3

[U-Boot] [PATCH 0/9] log: Support control over the log output format
by Simon Glass 26 Jan '18

26 Jan '18

This adds a few more features to the log system: - 'log format' command to control the log output format - 'log rec' command to output a log record manually - log_ret() function to log error-return values With these we have more control over how log records are output as well as the ability to log errors more easily. Simon Glass (9): dm: core: Add a function to look up a uclass by name log: Add functions to convert IDs to/from names log: Add control over log formatting log: Update log_console to honour the log format log: Add a command to control the log output format log: Add a command to output a log record log: Add tests for the new log features log: Add documentation for commands and formatting log: Add a way to log error-return values cmd/log.c | 82 +++++++++++++++++++++++++++++++++++++++ common/Kconfig | 13 +++++++ common/log.c | 68 ++++++++++++++++++++++++++++++++ common/log_console.c | 27 ++++++++++++- configs/sandbox_defconfig | 1 + doc/README.log | 35 +++++++++++++++++ drivers/core/uclass.c | 14 +++++++ include/asm-generic/global_data.h | 1 + include/dm/uclass.h | 8 ++++ include/log.h | 64 +++++++++++++++++++++++++++++- test/dm/core.c | 9 +++++ test/py/tests/test_log.py | 32 +++++++++++++-- 12 files changed, 348 insertions(+), 6 deletions(-) -- 2.15.1.620.gb9897f4670-goog

2 22

[U-Boot] reading ubi from a NAND for programming to other devices
by Tim Harvey 26 Jan '18

26 Jan '18

Greetings, I'm looking to understand what methods are available for reading/writing ubi/ubifs from an existing NAND flash in order to duplicate the contents onto another device in U-Boot, Linux, and with JTAG flashing for production. Typically I create ubi's using ubinize then I can use U-Boot 'nand erase' and 'nand write' or JTAG to write the ubi without issues. However if I use U-Boot 'nand read' to read a ubi then write it back using the same method I get 'ubi0 warning: ubi_io_read: error -74 (ECC error)' and 'ubi0 error: ubi_io_read: error -74 (ECC error)' messages from Linux when its mounted. I find I run into the same problem when JTAG programming ubi's: no issues when using a ubi created via ubinize but when flashing one that came from a U-Boot 'nand read' I see ubi errors when mounting it. Some testing shows me that I can't use U-Boot nand read/erase/write once U-Boot has 'attached' the UBI (via the U-Boot 'ubi part' command) which scans and re-sizes the volume. If I on the other hand take a ubi from U-Boot 'nand read' and use Linux ubiformat to flash that onto a new device I get no ubi warnings/errors when its mounted. My understanding is that ubiformat is simply preserving erase counters but there must be something else its doing. What's going on here and what can I do in U-Boot/JTAG to properly write ubi that's been read from an existing system? I'm hoping there is a method that doesn't require iterating over the individual volumes within the ubi. Best Regards, Tim

2 1

[U-Boot] [PATCH 00/10] mips: bmips: add SPI support
by Álvaro Fernández Rojas 26 Jan '18

26 Jan '18

BCM63xx SPI controller is a bit tricky since it doesn't allow keeping CS active between transfers, so I had to modify the spi_flash driver in order to allow limiting reads. Álvaro Fernández Rojas (10): drivers: spi: allow limiting reads drivers: spi: add config to consider command bytes when writting to flash dm: spi: add BCM63xx SPI driver mips: bmips: add bcm63xx-spi driver support for BCM6338 mips: bmips: add bcm63xx-spi driver support for BCM6348 mips: bmips: add bcm63xx-spi driver support for BCM6358 mips: bmips: add bcm63xx-spi driver support for BCM3380 mips: bmips: add bcm63xx-spi driver support for BCM63268 mips: bmips: enable the SPI flash on the Sagem F@ST1704 mips: bmips: enable the SPI flash on the Netgear CG3100D arch/mips/dts/brcm,bcm3380.dtsi | 13 ++ arch/mips/dts/brcm,bcm63268.dtsi | 13 ++ arch/mips/dts/brcm,bcm6338.dtsi | 13 ++ arch/mips/dts/brcm,bcm6348.dtsi | 13 ++ arch/mips/dts/brcm,bcm6358.dtsi | 13 ++ arch/mips/dts/netgear,cg3100d.dts | 13 ++ arch/mips/dts/sagem,f(a)st1704.dts | 13 ++ configs/netgear_cg3100d_ram_defconfig | 8 + configs/sagem_f@st1704_ram_defconfig | 8 + drivers/mtd/spi/spi_flash.c | 5 +- drivers/spi/Kconfig | 26 +++ drivers/spi/Makefile | 1 + drivers/spi/bcm63xx_spi.c | 404 ++++++++++++++++++++++++++++++++++ include/spi.h | 11 +- 14 files changed, 552 insertions(+), 2 deletions(-) create mode 100644 drivers/spi/bcm63xx_spi.c -- 2.1.4

4 186

[U-Boot] [PATCH v8 0/2] DW SPI: Get clock value from Device Tree
by Eugeniy Paltsev 26 Jan '18

26 Jan '18

As discussed with Marek during the LINUX-PITER here is v* patch: Add option to set spi controller clock frequency via device tree using standard clock bindings. Define dw_spi_get_clk function as 'weak' as some targets (like SOCFPGA_GEN5 and SOCFPGA_ARRIA10) don't use standard clock API and implement dw_spi_get_clk their own way in their clock manager. Get rid of clock_manager.h include in designware_spi.c as we don't use cm_get_spi_controller_clk_hz function anymore - we use redefined dw_spi_get_clk in SOCFPGA clock managers (clock_manager_gen5.c and clock_manager_arria10.c) instead. Changes v7->v8: * Fix typo in comments. Changes v6->v7: * Cleanup error handling. * Don't free clock after we successfully requeste it. Changes v5->v6: * Put the clock handle into the private data Changes v4->v5: * Get rid of usless ifdef in dw_spi_get_clk function Eugeniy Paltsev (2): SOCFPGA: clock manager: implement dw_spi_get_clk function DW SPI: Get clock value from Device Tree arch/arm/mach-socfpga/clock_manager_arria10.c | 9 +++++ arch/arm/mach-socfpga/clock_manager_gen5.c | 9 +++++ drivers/spi/designware_spi.c | 47 +++++++++++++++++++++++++-- 3 files changed, 63 insertions(+), 2 deletions(-) -- 2.9.3

5 8