September 2018 - U-Boot

[U-Boot] [PATVH v7 01/14] test: regmap: Increase size of syscon0 memory
by Mario Six 21 Sep '18

21 Sep '18

The upcoming changes to the regmap interface will contain a proper check for plausibility when reading/writing from/to a register map. To still have the current tests pass, increase the size of the memory region for the syscon0 device, since one of the tests reads and writes beyond this range. Reviewed-by: Anatolij Gustschin <agust(a)denx.de> Reviewed-by: Simon Glass <sjg(a)chromium.org> Signed-off-by: Mario Six <mario.six(a)gdsys.cc> --- v6 -> v7: No changes v5 -> v6: * Fix the test that checks the changed size v4 -> v5: No changes v3 -> v4: No changes v2 -> v3: New in v3 --- arch/sandbox/dts/test.dts | 2 +- test/dm/regmap.c | 2 +- 2 files changed, 2 insertions(+), 2 deletions(-) diff --git a/arch/sandbox/dts/test.dts b/arch/sandbox/dts/test.dts index b8524e3b7d..bd5f6adf01 100644 --- a/arch/sandbox/dts/test.dts +++ b/arch/sandbox/dts/test.dts @@ -467,7 +467,7 @@ syscon@0 { compatible = "sandbox,syscon0"; - reg = <0x10 4>; + reg = <0x10 16>; }; syscon@1 { diff --git a/test/dm/regmap.c b/test/dm/regmap.c index d4b86b3b03..b28d6a6cd1 100644 --- a/test/dm/regmap.c +++ b/test/dm/regmap.c @@ -25,7 +25,7 @@ static int dm_test_regmap_base(struct unit_test_state *uts) ut_assertok_ptr(map); ut_asserteq(1, map->range_count); ut_asserteq(0x10, map->ranges[0].start); - ut_asserteq(4, map->ranges[0].size); + ut_asserteq(16, map->ranges[0].size); ut_asserteq(0x10, map_to_sysmem(regmap_get_range(map, 0))); ut_assertok(uclass_get_device(UCLASS_SYSCON, 1, &dev)); -- 2.16.4

1 13

[U-Boot] Building for ARMv5 - uint64_t alignment issues
by Peter Howard 21 Sep '18

21 Sep '18

Dear all, Most of my u-boot work is on one of two ARMv5 platforms. When u-boot changed to requiring a later gcc, I switched from my old-faithful arago-2011.09 compiler to a newer Linaro release: gcc-linaro-7.2.1-2017.11-x86_64_arm-linux-gnueabi And for my work on the da850 this seems to be fine. However I've been doing some work creating a nand driver for an older chip (which I really wish I didn't have to care about, and is never going to see the light of upstream). Short form: inexplicable errors, which I eventually chased down to issues around uint64_t. Some googling turned up this bug (and patch): http://lists.openembedded.org/pipermail/openembedded-core/2017-November/144… So I upgraded to the next Linaro release: gcc-linaro-7.3.1-2018.05-x86_64_arm-linux-gnueabi This fixed the problem in what I'll call "general" uint64_t usage, but still left the same problem in structs. Specifically size in struct mdt_info. Which I fixed by kludging -mstructure-size-boundary=64 into the build (which AFAIK I shouldn't have to do). I would hope I'm not the only person who can trigger this . . . which means I'm hoping that no-one else has hit this (as far as I can tell from the u-boot list) is that anyone building for ARMv5 is using a different toolchain. So . . . can those of you out there building for any of the v5 boards tell me what toolchain you're currently using? And if you hit/fixed/worked around strd issues? Thanks -- Peter Howard <pjh(a)northern-ridge.com.au>

1 0

[U-Boot] [PATCH v2] tools: socfpga: fix sfp_verify_header
by Atsushi Nemoto 21 Sep '18

21 Sep '18

Fix sfp_verify_header to return correct version number. This fixes "Not a sane SOCFPGA preloader" error message with v1 header. Signed-off-by: Atsushi Nemoto <atsushi.nemoto(a)sord.co.jp> --- v2: move the *ver assignment to just before return. tools/socfpgaimage.c | 1 + 1 file changed, 1 insertion(+) diff --git a/tools/socfpgaimage.c b/tools/socfpgaimage.c index 390c9bb..72d8b96 100644 --- a/tools/socfpgaimage.c +++ b/tools/socfpgaimage.c @@ -191,6 +191,7 @@ static int sfp_verify_header(const uint8_t *buf, uint8_t *ver) if (hdr_csum != sfp_csum) return -EINVAL; + *ver = header_v0.version; return img_len; } -- 2.1.4

2 1

[U-Boot] [PATCH v2 0/4] Add Rock960 and Ficus 96Board support
by Manivannan Sadhasivam 21 Sep '18

21 Sep '18

This patchset adds support for Rock960 and Ficus 96Boards from Vamrs. Since both boards share most of the configurations, a common Rock960 family support is added with common support and the actual boards are based on this. The previous version of the patchseries were adding Rock960 [1] and Ficus [2] board support individually, but this series fuses them together based on the common board support as per Linux kernel. [1] https://patchwork.ozlabs.org/cover/963239/ [2] https://lists.denx.de/pipermail/u-boot/2018-August/339059.html This patchseries has been tested on Rock960 v1.2 board and expecting Ezequiel to do the testing for Ficus. PS: I have explicitly removed the previous Ack's for the Ficus board since there has been a heavy modification done on these patches. Thanks, Mani Manivannan Sadhasivam (4): arm: dts: rockchip: add some common pin-settings to rk3399 rockchip: rk3399: Add common Rock960 family from Vamrs rockchip: rk3399: Add Rock960 CE board support rockchip: rk3399: Add Ficus EE board support arch/arm/dts/Makefile | 2 + arch/arm/dts/rk3399-ficus.dts | 78 + arch/arm/dts/rk3399-rock960.dts | 45 + arch/arm/dts/rk3399-rock960.dtsi | 506 ++++++ .../arm/dts/rk3399-sdram-lpddr3-2GB-1600.dtsi | 1536 +++++++++++++++++ arch/arm/dts/rk3399.dtsi | 55 +- arch/arm/mach-rockchip/rk3399/Kconfig | 26 + board/vamrs/rock960_rk3399/Kconfig | 15 + board/vamrs/rock960_rk3399/MAINTAINERS | 6 + board/vamrs/rock960_rk3399/Makefile | 6 + board/vamrs/rock960_rk3399/README | 151 ++ board/vamrs/rock960_rk3399/rock960-rk3399.c | 50 + configs/ficus-rk3399_defconfig | 71 + configs/rock960-rk3399_defconfig | 67 + include/configs/rock960_rk3399.h | 15 + 15 files changed, 2623 insertions(+), 6 deletions(-) create mode 100644 arch/arm/dts/rk3399-ficus.dts create mode 100644 arch/arm/dts/rk3399-rock960.dts create mode 100644 arch/arm/dts/rk3399-rock960.dtsi create mode 100644 arch/arm/dts/rk3399-sdram-lpddr3-2GB-1600.dtsi create mode 100644 board/vamrs/rock960_rk3399/Kconfig create mode 100644 board/vamrs/rock960_rk3399/MAINTAINERS create mode 100644 board/vamrs/rock960_rk3399/Makefile create mode 100644 board/vamrs/rock960_rk3399/README create mode 100644 board/vamrs/rock960_rk3399/rock960-rk3399.c create mode 100644 configs/ficus-rk3399_defconfig create mode 100644 configs/rock960-rk3399_defconfig create mode 100644 include/configs/rock960_rk3399.h -- 2.17.1

1 4

[U-Boot] [PATCH] tools: socfpga: fix sfp_verify_header
by Atsushi Nemoto 21 Sep '18

21 Sep '18

Fix sfp_verify_header to return correct version number. This fixes "Not a sane SOCFPGA preloader" error message with v1 header. Signed-off-by: Atsushi Nemoto <atsushi.nemoto(a)sord.co.jp> --- tools/socfpgaimage.c | 1 + 1 file changed, 1 insertion(+) diff --git a/tools/socfpgaimage.c b/tools/socfpgaimage.c index 390c9bb..fea22b4 100644 --- a/tools/socfpgaimage.c +++ b/tools/socfpgaimage.c @@ -174,6 +174,7 @@ static int sfp_verify_header(const uint8_t *buf, uint8_t *ver) if (le32_to_cpu(header_v0.validation) != VALIDATION_WORD) return -1; + *ver = header_v0.version; if (header_v0.version == 0) { hdr_csum = le16_to_cpu(header_v0.checksum); sfp_csum = sfp_hdr_checksum((uint8_t *)&header_v0, 0); -- 2.1.4

2 2

[U-Boot] [PATCH] test/py: ignore console read exceptions after test failure
by Stephen Warren 21 Sep '18

21 Sep '18

From: Stephen Warren <swarren(a)nvidia.com> After a test has failed, test/py drains the U-Boot console log to ensure that any relevant output is captured. At this point, we don't care about detecting any additional errors, since the test is already known to have failed, and U-Boot will be restarted. To ensure that the test cleanup code is not interrupted, and can correctly terminate the log sections for the failed test, ignore any exception that occurs while reading the U-Boot console output during this limited period of time. Signed-off-by: Stephen Warren <swarren(a)nvidia.com> --- test/py/u_boot_console_base.py | 11 +++++++++++ 1 file changed, 11 insertions(+) diff --git a/test/py/u_boot_console_base.py b/test/py/u_boot_console_base.py index a14bad6e8c50..e337f3a84a96 100644 --- a/test/py/u_boot_console_base.py +++ b/test/py/u_boot_console_base.py @@ -305,6 +305,17 @@ class ConsoleBase(object): # cause the console output to be read and logged. self.p.expect(['This should never match U-Boot output']) except u_boot_spawn.Timeout: + # We expect a timeout, since U-Boot won't print what we waited + # for. Squash the expected timeout exception. + pass + except: + # Squash any other exception too. This function is only used to + # drain (and log) the U-Boot console output after a failed test. + # The U-Boot process will be restarted, or target board reset, once + # this function returns. So, we don't care about detecting any + # additional errors, so they're squashed so that the rest of the + # post-test-failure cleanup code can continue operation, and + # correctly terminate any log sections, etc. pass finally: self.p.timeout = orig_timeout -- 2.18.0

2 2

[U-Boot] [PATCH 1/1] efi_loader: create root node
by Heinrich Schuchardt 20 Sep '18

20 Sep '18

Currently we assign a lot of protocols to loaded images though these protocols are not related to them. Instead they should be installed on a separate handle. Via the device path it is the parent to the devices like the network adapter. Signed-off-by: Heinrich Schuchardt <xypron.glpk(a)gmx.de> --- cmd/bootefi.c | 5 ++ include/efi_loader.h | 9 ++++ lib/efi_loader/Makefile | 1 + lib/efi_loader/efi_boottime.c | 18 -------- lib/efi_loader/efi_device_path.c | 4 -- lib/efi_loader/efi_root_node.c | 79 ++++++++++++++++++++++++++++++++ 6 files changed, 94 insertions(+), 22 deletions(-) create mode 100644 lib/efi_loader/efi_root_node.c diff --git a/cmd/bootefi.c b/cmd/bootefi.c index c8812b0f5e..5fc054c9da 100644 --- a/cmd/bootefi.c +++ b/cmd/bootefi.c @@ -49,6 +49,11 @@ efi_status_t efi_init_obj_list(void) if (ret != EFI_SUCCESS) goto out; + /* Initialize root node */ + ret = efi_root_node_register(); + if (ret != EFI_SUCCESS) + goto out; + /* Initialize EFI driver uclass */ ret = efi_driver_init(); if (ret != EFI_SUCCESS) diff --git a/include/efi_loader.h b/include/efi_loader.h index c5b73b7238..5d522f133e 100644 --- a/include/efi_loader.h +++ b/include/efi_loader.h @@ -20,6 +20,11 @@ /* Maximum number of configuration tables */ #define EFI_MAX_CONFIGURATION_TABLES 16 +/* GUID used by the root node */ +#define U_BOOT_GUID \ + EFI_GUID(0xe61d73b9, 0xa384, 0x4acc, \ + 0xae, 0xab, 0x82, 0xe8, 0x28, 0xf3, 0x62, 0x8b) + int __efi_entry_check(void); int __efi_exit_check(void); const char *__efi_nesting(void); @@ -104,6 +109,8 @@ extern const struct efi_unicode_collation_protocol uint16_t *efi_dp_str(struct efi_device_path *dp); +/* GUID of the U-Boot root node */ +extern const efi_guid_t efi_u_boot_guid; /* GUID of the EFI_BLOCK_IO_PROTOCOL */ extern const efi_guid_t efi_block_io_guid; extern const efi_guid_t efi_global_variable_guid; @@ -210,6 +217,8 @@ extern struct list_head efi_obj_list; /* List of all events */ extern struct list_head efi_events; +/* Called by bootefi to initialize root node */ +efi_status_t efi_root_node_register(void); /* Called by bootefi to initialize runtime */ efi_status_t efi_initialize_system_table(void); /* Called by bootefi to make console interface available */ diff --git a/lib/efi_loader/Makefile b/lib/efi_loader/Makefile index 7eebbc5b4e..6703435947 100644 --- a/lib/efi_loader/Makefile +++ b/lib/efi_loader/Makefile @@ -26,6 +26,7 @@ obj-y += efi_device_path_utilities.o obj-y += efi_file.o obj-y += efi_image_loader.o obj-y += efi_memory.o +obj-y += efi_root_node.o obj-y += efi_runtime.o obj-y += efi_unicode_collation.o obj-y += efi_variable.o diff --git a/lib/efi_loader/efi_boottime.c b/lib/efi_loader/efi_boottime.c index 5d93db8e0c..b4de9961b8 100644 --- a/lib/efi_loader/efi_boottime.c +++ b/lib/efi_loader/efi_boottime.c @@ -1514,24 +1514,6 @@ efi_status_t efi_setup_loaded_image( if (ret != EFI_SUCCESS) goto failure; - ret = efi_add_protocol(obj->handle, - &efi_guid_device_path_to_text_protocol, - (void *)&efi_device_path_to_text); - if (ret != EFI_SUCCESS) - goto failure; - - ret = efi_add_protocol(obj->handle, - &efi_guid_device_path_utilities_protocol, - (void *)&efi_device_path_utilities); - if (ret != EFI_SUCCESS) - goto failure; - - ret = efi_add_protocol(obj->handle, - &efi_guid_unicode_collation_protocol, - (void *)&efi_unicode_collation_protocol); - if (ret != EFI_SUCCESS) - goto failure; - return ret; failure: printf("ERROR: Failure to install protocols for loaded image\n"); diff --git a/lib/efi_loader/efi_device_path.c b/lib/efi_loader/efi_device_path.c index b770db84fd..46a24f7882 100644 --- a/lib/efi_loader/efi_device_path.c +++ b/lib/efi_loader/efi_device_path.c @@ -22,10 +22,6 @@ static const struct efi_device_path END = { .length = sizeof(END), }; -#define U_BOOT_GUID \ - EFI_GUID(0xe61d73b9, 0xa384, 0x4acc, \ - 0xae, 0xab, 0x82, 0xe8, 0x28, 0xf3, 0x62, 0x8b) - /* template ROOT node: */ static const struct efi_device_path_vendor ROOT = { .dp = { diff --git a/lib/efi_loader/efi_root_node.c b/lib/efi_loader/efi_root_node.c new file mode 100644 index 0000000000..b056ba3ee8 --- /dev/null +++ b/lib/efi_loader/efi_root_node.c @@ -0,0 +1,79 @@ +// SPDX-License-Identifier: GPL-2.0+ +/* + * Root node for system services + * + * Copyright (c) 2018 Heinrich Schuchardt + */ + +#include <common.h> +#include <malloc.h> +#include <efi_loader.h> + +const efi_guid_t efi_u_boot_guid = U_BOOT_GUID; + +struct efi_root_dp { + struct efi_device_path_vendor vendor; + struct efi_device_path end; +} __packed; + +/** + * efi_root_node_register() - create root node + * + * Create the root node on which we install all protocols that are + * not related to a loaded image or a driver. + * + * Return: status code + */ +efi_status_t efi_root_node_register(void) +{ + efi_handle_t root; + efi_status_t ret; + struct efi_root_dp *dp; + + /* Create handle */ + ret = efi_create_handle(&root); + if (ret != EFI_SUCCESS) + return ret; + + /* Install device path protocol */ + dp = calloc(1, sizeof(*dp)); + if (!dp) + return EFI_OUT_OF_RESOURCES; + + /* Fill vendor node */ + dp->vendor.dp.type = DEVICE_PATH_TYPE_HARDWARE_DEVICE; + dp->vendor.dp.sub_type = DEVICE_PATH_SUB_TYPE_VENDOR; + dp->vendor.dp.length = sizeof(struct efi_device_path_vendor); + dp->vendor.guid = efi_u_boot_guid; + + /* Fill end node */ + dp->end.type = DEVICE_PATH_TYPE_END; + dp->end.sub_type = DEVICE_PATH_SUB_TYPE_END; + dp->end.length = sizeof(struct efi_device_path); + + /* Install device path protocol */ + ret = efi_add_protocol(root, &efi_guid_device_path, dp); + if (ret != EFI_SUCCESS) + goto failure; + + /* Install device path to text protocol */ + ret = efi_add_protocol(root, &efi_guid_device_path_to_text_protocol, + (void *)&efi_device_path_to_text); + if (ret != EFI_SUCCESS) + goto failure; + + /* Install device path utilities protocol */ + ret = efi_add_protocol(root, &efi_guid_device_path_utilities_protocol, + (void *)&efi_device_path_utilities); + if (ret != EFI_SUCCESS) + goto failure; + + /* Install Unicode collation protocol */ + ret = efi_add_protocol(root, &efi_guid_unicode_collation_protocol, + (void *)&efi_unicode_collation_protocol); + if (ret != EFI_SUCCESS) + goto failure; + +failure: + return ret; +} -- 2.18.0

1 0

[U-Boot] [PATCH 1/1] test/py: catch errors occurring when reading the console
by Heinrich Schuchardt 20 Sep '18

20 Sep '18

Spawn.exept has a try block without 'except'. When the py test is running it is connected via pipes to the U-Boot process. If the U-Boot process ends prematurely, e.g. due to a segmentation fault, the pipes are broken. Trying to read from a broken pipe results in an OSError. Catch the exception and treat the broken pipe like any other end of output. By returning None expect() indicates that the output does not match. Signed-off-by: Heinrich Schuchardt <xypron.glpk(a)gmx.de> Reviewed-by: Simon Glass <sjg(a)chromium.org> --- I suggest that Alex takes the patch because we need it when working on the efi-next branch. v3 add more comments v2 replace TAB by spaces fix type in subject --- test/py/u_boot_spawn.py | 8 ++++++++ 1 file changed, 8 insertions(+) diff --git a/test/py/u_boot_spawn.py b/test/py/u_boot_spawn.py index b011a3e3da..adc1d00287 100644 --- a/test/py/u_boot_spawn.py +++ b/test/py/u_boot_spawn.py @@ -181,6 +181,14 @@ class Spawn(object): # unlimited substitutions, but in practice the version of # Python in Ubuntu 14.04 appears to default to count=2! self.buf = self.re_vt100.sub('', self.buf, count=1000000) + except OSError, EOFError: + # When the py test is running it is connected via pipes to the + # U-Boot process. If the U-Boot process ends prematurely, e.g. due + # to a segmentation fault, the pipes are broken. Trying to read + # from a broken pipe results in an OSError. Treat the broken pipe + # like any other end of output. By returning None expect() + # indicates that the output does not match. + pass finally: if self.logfile_read: self.logfile_read.flush() -- 2.18.0

2 1

[U-Boot] [PATCH 1/1] test/py: catch errors occurring when reading the console
by Heinrich Schuchardt 20 Sep '18

20 Sep '18

Spawn.exept has a try block without 'except'. If no output is available an OSError may arise. Catch this exception and continue testing. Signed-off-by: Heinrich Schuchardt <xypron.glpk(a)gmx.de> Reviewed-by: Simon Glass <sjg(a)chromium.org> --- I suggest that Alex takes the patch because we need it when working on the efi-next branch. v2 replace TAB by spaces fix typo in subject --- test/py/u_boot_spawn.py | 3 +++ 1 file changed, 3 insertions(+) diff --git a/test/py/u_boot_spawn.py b/test/py/u_boot_spawn.py index b011a3e3da..7ee876b101 100644 --- a/test/py/u_boot_spawn.py +++ b/test/py/u_boot_spawn.py @@ -181,6 +181,9 @@ class Spawn(object): # unlimited substitutions, but in practice the version of # Python in Ubuntu 14.04 appears to default to count=2! self.buf = self.re_vt100.sub('', self.buf, count=1000000) + except OSError, EOFError: + # Reading the the console may result in an error. Catch it. + pass finally: if self.logfile_read: self.logfile_read.flush() -- 2.18.0

4 7

[U-Boot] [PATCH 1/1] efi_loader: remove redundant statement
by Heinrich Schuchardt 20 Sep '18

20 Sep '18

ascii2unicode() always ends the u16 string with '\0'. Remove redundant assignment. Add description for efi_str_to_u16() and ascii2unicode(). Signed-off-by: Heinrich Schuchardt <xypron.glpk(a)gmx.de> --- include/efi_loader.h | 10 +++++++++- lib/efi_loader/efi_device_path_to_text.c | 10 +++++++++- 2 files changed, 18 insertions(+), 2 deletions(-) diff --git a/include/efi_loader.h b/include/efi_loader.h index e5bc7d043e..736879345b 100644 --- a/include/efi_loader.h +++ b/include/efi_loader.h @@ -408,7 +408,15 @@ efi_status_t efi_dp_split_file_path(struct efi_device_path *full_path, (((_dp)->type == DEVICE_PATH_TYPE_##_type) && \ ((_dp)->sub_type == DEVICE_PATH_SUB_TYPE_##_subtype)) -/* Convert strings from normal C strings to uEFI strings */ +/** + * ascii2unicode() - convert ASCII string to UTF-16 string + * + * A zero terminated ASCII string is converted to a zero terminated UTF-16 + * string. The output buffer must be preassigned. + * + * @unicode: preassigned output buffer for UTF-16 string + * @ascii: ASCII string to be converted + */ static inline void ascii2unicode(u16 *unicode, const char *ascii) { while (*ascii) diff --git a/lib/efi_loader/efi_device_path_to_text.c b/lib/efi_loader/efi_device_path_to_text.c index ca8037def2..0082236359 100644 --- a/lib/efi_loader/efi_device_path_to_text.c +++ b/lib/efi_loader/efi_device_path_to_text.c @@ -17,6 +17,15 @@ const efi_guid_t efi_guid_device_path_to_text_protocol = EFI_DEVICE_PATH_TO_TEXT_PROTOCOL_GUID; +/** + * efi_str_to_u16() - convert ASCII string to UTF-16 + * + * A u16 buffer is allocated from pool. The ASCII string is copied to the u16 + * buffer. + * + * @str: ASCII string + * Return: UTF-16 string. NULL if out of memory. + */ static u16 *efi_str_to_u16(char *str) { efi_uintn_t len; @@ -29,7 +38,6 @@ static u16 *efi_str_to_u16(char *str) if (ret != EFI_SUCCESS) return NULL; ascii2unicode(out, str); - out[len - 1] = 0; return out; } -- 2.18.0

1 0