April 2021 - U-Boot

[PATCH 00/11] Enable DM_ETH for DPAA1 PowerPC T-series boards
by Camelia Groza 13 Apr '21

13 Apr '21

Apologies to those who received this series twice off list due to an e-mail misconfiguration. This patch series enables DM_ETH and DM_MDIO on the T2080RDB, T4240RDB and T1042D4RDB DPAA1 PowerPC boards. The necessary FMan v3 Ethernet and MDIO nodes are added to each platform's SoC and board device tree. On T2080RDB, fdt fixups are performed before booting into the OS using a simplified new routine in order to avoid the legacy code path. Camelia Groza (11): powerpc: dts: t2080: add QorIQ DPAA 1 FMan v3 nodes powerpc: dts: t2080rdb: add FMan v3 nodes board: freescale: t208xrdb: fdt fixups under DM_ETH configs: T2080RDB: enable DM_ETH powerpc: dts: t4240: add QorIQ DPAA 1 FMan v3 nodes powerpc: dts: t4240rdb: add FMan v3 nodes powerpc: dts: qoriq: update the mdio offsets under the second FMan v3 configs: T4240RDB: enable DM_ETH powerpc: dts: t1042: add QorIQ DPAA 1 FMan v3 nodes powerpc: dts: t1042d4rdb: add FMan v3 nodes configs: T1042D4RDB: enable DM_ETH arch/powerpc/dts/qoriq-fman3-1-10g-0.dtsi | 2 +- arch/powerpc/dts/qoriq-fman3-1-10g-1.dtsi | 2 +- arch/powerpc/dts/qoriq-fman3-1-1g-0.dtsi | 2 +- arch/powerpc/dts/qoriq-fman3-1-1g-1.dtsi | 2 +- arch/powerpc/dts/qoriq-fman3-1-1g-2.dtsi | 2 +- arch/powerpc/dts/qoriq-fman3-1-1g-3.dtsi | 2 +- arch/powerpc/dts/qoriq-fman3-1-1g-4.dtsi | 2 +- arch/powerpc/dts/qoriq-fman3-1-1g-5.dtsi | 2 +- arch/powerpc/dts/qoriq-fman3-1.dtsi | 4 +- arch/powerpc/dts/t1042d4rdb.dts | 55 ++++++++- arch/powerpc/dts/t1042si-post.dtsi | 46 +++++++ arch/powerpc/dts/t2080rdb.dts | 69 ++++++++++- arch/powerpc/dts/t2080si-post.dtsi | 51 ++++++++ arch/powerpc/dts/t4240rdb.dts | 142 +++++++++++++++++++++- arch/powerpc/dts/t4240si-post.dtsi | 101 +++++++++++++++ board/freescale/t208xrdb/eth_t208xrdb.c | 35 ++++++ board/freescale/t208xrdb/t208xrdb.c | 3 + board/freescale/t208xrdb/t208xrdb.h | 2 + configs/T1042D4RDB_NAND_defconfig | 3 + configs/T1042D4RDB_SDCARD_defconfig | 3 + configs/T1042D4RDB_SECURE_BOOT_defconfig | 4 +- configs/T1042D4RDB_SPIFLASH_defconfig | 3 + configs/T1042D4RDB_defconfig | 3 + configs/T2080RDB_NAND_defconfig | 3 + configs/T2080RDB_SDCARD_defconfig | 3 + configs/T2080RDB_SPIFLASH_defconfig | 3 + configs/T2080RDB_defconfig | 3 + configs/T4240RDB_SDCARD_defconfig | 3 + configs/T4240RDB_defconfig | 3 + 29 files changed, 544 insertions(+), 14 deletions(-) create mode 100644 arch/powerpc/dts/t1042si-post.dtsi create mode 100644 arch/powerpc/dts/t2080si-post.dtsi create mode 100644 arch/powerpc/dts/t4240si-post.dtsi -- 2.17.1

3 13

[BUG][Cyrus BOARD] out of bound access in board/varisys/common/sys_eeprom.c
by Heinrich Schuchardt 13 Apr '21

13 Apr '21

Hello Andy, in the code of your patch "mpc85xx: Add support for the Varisys Cyrus board" merged in 2015 as 87e29878cab an out of bound access occurs. See below. On 11/4/15 10:48 PM, Andy Fleming wrote: > This board runs a P5020 or P5040 chip, and utilizes > an EEPROM with similar formatting to the Freescale P5020DS. > > Large amounts of this code were developed by > Adrian Cox <adrian at humboldt dot co dot uk> > > Signed-off-by: Andy Fleming <afleming(a)gmail.com> > Reviewed-by: York Sun <yorksun(a)freescale.com> > --- <snip /> > +++ b/board/varisys/common/sys_eeprom.c <snip /> > +static struct __attribute__ ((__packed__)) eeprom { > + u8 id[4]; /* 0x00 - 0x03 EEPROM Tag 'NXID' */ > + u8 sn[12]; /* 0x04 - 0x0F Serial Number */ > + u8 errata[5]; /* 0x10 - 0x14 Errata Level */ > + u8 date[6]; /* 0x15 - 0x1a Build Date */ > + u8 res_0; /* 0x1b Reserved */ > + u32 version; /* 0x1c - 0x1f NXID Version */ > + u8 tempcal[8]; /* 0x20 - 0x27 Temperature Calibration Factors */ > + u8 tempcalsys[2]; /* 0x28 - 0x29 System Temperature Calibration Factors */ > + u8 tempcalflags; /* 0x2a Temperature Calibration Flags */ > + u8 res_1[21]; /* 0x2b - 0x3f Reserved */ > + u8 mac_count; /* 0x40 Number of MAC addresses */ > + u8 mac_flag; /* 0x41 MAC table flags */ > + u8 mac[MAX_NUM_PORTS][6]; /* 0x42 - x MAC addresses */ > + u32 crc; /* x+1 CRC32 checksum */ > +} e; <snip /> // MAX_NUM_PORTS = 8 > +int mac_read_from_eeprom_common(void) > +{ > + unsigned int i; > + u32 crc, crc_offset = offsetof(struct eeprom, crc); > + u32 *crcp; /* Pointer to the CRC in the data read from the EEPROM */ > + > + puts("EEPROM: "); > + > + if (read_eeprom()) { > + printf("Read failed.\n"); > + return 0; > + } > + > + if (!is_valid) { > + printf("Invalid ID (%02x %02x %02x %02x)\n", > + e.id[0], e.id[1], e.id[2], e.id[3]); > + return 0; > + } > + > + crc = crc32(0, (void *)&e, crc_offset); > + crcp = (void *)&e + crc_offset; > + if (crc != be32_to_cpu(*crcp)) { > + printf("CRC mismatch (%08x != %08x)\n", crc, > + be32_to_cpu(e.crc)); > + return 0; > + } > + > + /* > + * MAC address #9 in v1 occupies the same position as the CRC in v0. > + * Erase it so that it's not mistaken for a MAC address. We'll > + * update the CRC later. > + */ > + if (e.version == 0) > + memset(e.mac[8], 0xff, 6); Here you are writing 2 bytes beyond the size of e. Best regards Heinrich

3 2

[PATCH] [RFC] cmd: exit: Fix return value
by Marek Vasut 13 Apr '21

13 Apr '21

In case exit is called in a script without parameter, the command returns -2 ; in case exit is called with a numerical parameter, the command returns -2 and lower. This leads to the following problem: => setenv foo 'echo bar ; exit 1' ; run foo ; echo $? bar 0 => setenv foo 'echo bar ; exit 0' ; run foo ; echo $? bar 0 => setenv foo 'echo bar ; exit -2' ; run foo ; echo $? bar 0 That is, no matter what the 'exit' command argument is, the return value is always 0 and so it is not possible to use script return value in subsequent tests. Fix this and simplify the exit command such that if exit is called with no argument, the command returns 0, just like 'true' in cmd/test.c. In case the command is called with any argument that is positive integer, the argument is set as return value. => setenv foo 'echo bar ; exit 1' ; run foo ; echo $? bar 1 => setenv foo 'echo bar ; exit 0' ; run foo ; echo $? bar 0 => setenv foo 'echo bar ; exit -2' ; run foo ; echo $? bar 0 Note that this does change ABI established in 2004 , although it is unclear whether that ABI was originally OK or not. Fixes: c26e454dfc6 Signed-off-by: Marek Vasut <marex(a)denx.de> Cc: Pantelis Antoniou <pantelis.antoniou(a)konsulko.com> Cc: Tom Rini <trini(a)konsulko.com> --- cmd/exit.c | 7 ++----- 1 file changed, 2 insertions(+), 5 deletions(-) diff --git a/cmd/exit.c b/cmd/exit.c index f40d0686e1e..af9ff56180c 100644 --- a/cmd/exit.c +++ b/cmd/exit.c @@ -9,13 +9,10 @@ static int do_exit(cmd_tbl_t *cmdtp, int flag, int argc, char * const argv[]) { - int r; - - r = 0; if (argc > 1) - r = simple_strtoul(argv[1], NULL, 10); + return simple_strtoul(argv[1], NULL, 10); - return -r - 2; + return 0; } U_BOOT_CMD( -- 2.30.2

2 1

[PATCH] scmi: translate the resource only when livetree is not activated
by Patrick Delaunay 13 Apr '21

13 Apr '21

Call the translation function on the ofnode_read_resource result only when the livetree is not activated. Today of_address_to_resource() calls ofnode_read_resource() for livetree support and fdt_get_resource() when livetree is not supported. The fdt_get_resource() doesn't do the address translation so when it is required when livetree is activated but this address translation is already done by ofnode_read_resource(). Fixes: 240720e9052f ("firmware: scmi: mailbox/smt agent device") Signed-off-by: Patrick Delaunay <patrick.delaunay(a)foss.st.com> --- Hi, This patch is a short-term workaround for ofnode_read_resource translation issue on stm32mp15 boards when scmi is used and OF_LIVE is activated. A full solution is a patch in core to have the same behavior of the API ofnode_read_resource() when livretree is activated or not: fdt_get_resource() ... while (ptr + na + ns <= end) { if (i == index) { - res->start = fdtdec_get_number(ptr, na); + if (CONFIG_IS_ENABLED(OF_TRANSLATE)) + res->start = fdt_translate_address(fdt, node, ptr); + else + res->start = fdtdec_get_number(ptr, na); + I will propose it soon but it is more aggressive and impacts all the user of fdt_get_resource() or fdt_get_named_resource() function; this correct patch need more tests. So I think it will take many time to be integrated so I prefer to propose the current patch for a short term integration on v2021.07. Patrick drivers/firmware/scmi/smt.c | 9 +++++++-- 1 file changed, 7 insertions(+), 2 deletions(-) diff --git a/drivers/firmware/scmi/smt.c b/drivers/firmware/scmi/smt.c index 4954355c1a..f1915c0074 100644 --- a/drivers/firmware/scmi/smt.c +++ b/drivers/firmware/scmi/smt.c @@ -41,8 +41,13 @@ int scmi_dt_get_smt_buffer(struct udevice *dev, struct scmi_smt *smt) if (ret) return ret; - faddr = cpu_to_fdt32(resource.start); - paddr = ofnode_translate_address(args.node, &faddr); + /* TEMP workaround for ofnode_read_resource translation issue */ + if (of_live_active()) { + paddr = resource.start; + } else { + faddr = cpu_to_fdt32(resource.start); + paddr = ofnode_translate_address(args.node, &faddr); + } smt->size = resource_size(&resource); if (smt->size < sizeof(struct scmi_smt_header)) { -- 2.17.1

3 2

[PATCH v7 0/2] gpio: Update and simplify the uclass API
by Simon Glass 13 Apr '21

13 Apr '21

At present the GPIO uclass mirrors what was in U-Boot before driver model. It works well in most cases but is becoming cumbersome with things like pull-up/down and drive strength. In those cases it is easier for the driver to deal with all the flags at one, rather than piece by piece. In fact the current API does not officially have a method for adjusting anything other than the direction flags. While set_dir_flags() and get_dir_flags() do in fact support changing other flags, this is not documented. Secondly, set_dir_flags actually ORs the current flags with the new ones so it is not possible to clear flags. It seems better to use a clr/set interface (bit clear followed by OR) to provide more flexibility. Finally, direction_input() and direction_output() are really just the same thing as set_dir_flags(), with a different name. We currently use the latter if available, failing back to the former. But it makes sense to deprecate the old methods. This series makes the above changes. Existing drivers are mostly left alone, since they should continue to operate as is. The sandbox driver is updated to add the required new tests and the x86 driver is switched over to the new API. The STM32 driver should be checked to make sure the open source/open drain features still work as intended. Changes in v7: - Update commit message to mention dm_gpio_clrset_flags() Changes in v6: - Use dm_gpio_clrset_flags() so that direction can be changed - Simplify the code since masking is not needed Changes in v5: - Drop patches previously applied Simon Glass (2): gpio: i2c-gpio: Drop use of dm_gpio_set_dir() gpio: Drop dm_gpio_set_dir() drivers/gpio/gpio-uclass.c | 11 ----------- drivers/i2c/i2c-gpio.c | 19 +++++++++---------- include/asm-generic/gpio.h | 11 ----------- 3 files changed, 9 insertions(+), 32 deletions(-) -- 2.31.0.rc2.261.g7f71774620-goog

2 4

[PATCH v8 0/4] Move qfw to DM, add Arm support
by Asherah Connor 13 Apr '21

13 Apr '21

This series moves the QFW driver into a uclass, UCLASS_QFW, and splits the driver into qfw_pio and qfw_mmio. Each driver is selected on the appropriate QEMU board. A sandbox driver is also added, and a DM unit test against that driver. The qfw command is tested in QEMU, and documentation added. Version 7 ensures the qfw test assertions work both against real QEMU instances and the sandbox and removes CMD_QFW from the sandbox_spl config. Version 8 removes some extraneous whitespace. To view the changes online, see: https://git.src.kameliya.ee/~kameliya/u-boot/log/qfw-priv Changes in v8: * Removed extraneous newlines at EOF. Changes in v7: * Make the qfw tests agnostic to running against a real QEMU instance or in the sandbox. * Remove CONFIG_CMD_QFW=y from sandbox_spl_defconfig, per trini's request. Changes in v6: * Sandbox driver and tests are split into own commit. Changes in v5: * Split conversion of existing x86-only QFW to DM into its own patch. * Split MMIO driver into its own commit. * Split adding MMIO driver to QEMU arm/64 into own commit. Asherah Connor (4): x86: qemu: move QFW to its own uclass test: qemu: add qfw sandbox driver, dm tests, qemu tests qemu: add MMIO driver for QFW qemu: arm: select QFW, MMIO on qemu-arm arch/x86/cpu/qemu/cpu.c | 9 +- arch/x86/cpu/qemu/qemu.c | 49 +------ arch/x86/cpu/qfw_cpu.c | 11 +- board/emulation/qemu-arm/Kconfig | 2 + board/emulation/qemu-x86/Kconfig | 1 + cmd/qfw.c | 56 ++++--- common/Makefile | 2 + common/qfw.c | 104 +++++++++++++ configs/sandbox_spl_defconfig | 1 - drivers/misc/Kconfig | 18 ++- drivers/misc/Makefile | 7 +- drivers/misc/qfw.c | 243 +++++++++++-------------------- drivers/misc/qfw_mmio.c | 119 +++++++++++++++ drivers/misc/qfw_pio.c | 69 +++++++++ drivers/misc/qfw_sandbox.c | 127 ++++++++++++++++ include/dm/uclass-id.h | 1 + include/qfw.h | 200 +++++++++++++++++++++---- test/dm/Makefile | 1 + test/dm/qfw.c | 42 ++++++ test/py/tests/test_qfw.py | 26 ++++ 20 files changed, 815 insertions(+), 273 deletions(-) create mode 100644 common/qfw.c create mode 100644 drivers/misc/qfw_mmio.c create mode 100644 drivers/misc/qfw_pio.c create mode 100644 drivers/misc/qfw_sandbox.c create mode 100644 test/dm/qfw.c create mode 100644 test/py/tests/test_qfw.py -- 2.20.1

2 8

[PATCH] scmi: correctly configure MMU for SCMI buffer
by Patrick Delaunay 13 Apr '21

13 Apr '21

Align the MMU area for SCMI shared buffer on section size; use the ALIGN macro in mmu_set_region_dcache_behaviour call. Since commit d877f8fd0f09 ("arm: provide a function for boards init code to modify MMU virtual-physical map") the parameter of mmu_set_region_dcache_behaviour need to be MMU_SECTION_SIZE aligned. Fixes: 240720e9052f ("firmware: scmi: mailbox/smt agent device") Signed-off-by: Patrick Delaunay <patrick.delaunay(a)foss.st.com> --- drivers/firmware/scmi/smt.c | 6 ++++-- 1 file changed, 4 insertions(+), 2 deletions(-) diff --git a/drivers/firmware/scmi/smt.c b/drivers/firmware/scmi/smt.c index d25478796a..3c41013bca 100644 --- a/drivers/firmware/scmi/smt.c +++ b/drivers/firmware/scmi/smt.c @@ -54,8 +54,10 @@ int scmi_dt_get_smt_buffer(struct udevice *dev, struct scmi_smt *smt) #ifdef CONFIG_ARM if (dcache_status()) - mmu_set_region_dcache_behaviour((uintptr_t)smt->buf, - smt->size, DCACHE_OFF); + mmu_set_region_dcache_behaviour(ALIGN_DOWN((uintptr_t)smt->buf, MMU_SECTION_SIZE), + ALIGN(smt->size, MMU_SECTION_SIZE), + DCACHE_OFF); + #endif return 0; -- 2.17.1

3 2

[PATCH 1/2] moveconfig.py: add try…except
by Trevor Woerner 13 Apr '21

13 Apr '21

I keep getting a UnicodeDecodeError from one of the header files while modifying the headers: Traceback (most recent call last): File "tools/moveconfig.py", line 1953, in <module> main() File "tools/moveconfig.py", line 1927, in main cleanup_headers(configs, options) File "tools/moveconfig.py", line 675, in cleanup_headers cleanup_one_header(header_path, patterns, options) File "tools/moveconfig.py", line 599, in cleanup_one_header lines = f.readlines() File "/usr/lib64/python3.6/codecs.py", line 321, in decode (result, consumed) = self._buffer_decode(data, self.errors, final) UnicodeDecodeError: 'utf-8' codec can't decode byte 0x83 in position 9: invalid start byte I'm curious to know which header file is causing the problem. Signed-off-by: Trevor Woerner <twoerner(a)gmail.com> --- tools/moveconfig.py | 7 ++++++- 1 file changed, 6 insertions(+), 1 deletion(-) diff --git a/tools/moveconfig.py b/tools/moveconfig.py index 9514d9a00c..dd92c00bb7 100755 --- a/tools/moveconfig.py +++ b/tools/moveconfig.py @@ -596,7 +596,12 @@ def cleanup_one_header(header_path, patterns, options): options: option flags. """ with open(header_path) as f: - lines = f.readlines() + try: + lines = f.readlines() + except: + print("caught exception: ", sys.exc_info()[0]) + print("header_path: ", header_path) + lines = "" matched = [] for i, line in enumerate(lines): -- 2.30.0.rc0

2 3

[PATCH v2 0/5] string: strl(cat|cpy)
by Sean Anderson 13 Apr '21

13 Apr '21

This series adds support for strl(cat|cpy), and brings their implementations in-line with what is documented in the Linux man pages. It also fixes some potential (actual) fastboot bugs. Lastly, it adds a patman check to suggest using these functions over strn(cat|cpy). I think these functions provide a much better interface, which removes some footguns from U-Boot. Changes in v2: - Fix strlcpy return value - Add implementation of strlcat - Add test for strlcat - Fix bug in fastboot - Move check to u_boot_line Sean Anderson (5): lib: string: Fix strlcpy return value lib: string: Implement strlcat test: Add test for strlcat fastboot: Fix possible buffer overrun checkpatch: Add warnings for using strn(cat|cpy) drivers/fastboot/fb_mmc.c | 6 +- drivers/usb/dwc3/linux-compat.h | 6 -- include/linux/string.h | 3 + lib/string.c | 31 +++++++- scripts/checkpatch.pl | 6 ++ test/lib/Makefile | 1 + test/lib/strlcat.c | 126 ++++++++++++++++++++++++++++++++ tools/patman/test_checkpatch.py | 14 +++- 8 files changed, 179 insertions(+), 14 deletions(-) create mode 100644 test/lib/strlcat.c -- 2.30.1

3 17

[PATCH v2] tee: optee: Change printing during optee_probe
by Ilias Apalodimas 13 Apr '21

13 Apr '21

Right now the error messages when optee has a version mismatch or shared memory is not configured are done with a debug(). That's not very convenient since you have to enable debugging to figure out what's going on, although this is an actual error. So let's switch the debug() -> dev_err() and report those explicitly. Signed-off-by: Ilias Apalodimas <ilias.apalodimas(a)linaro.org> Reviewed-by: Patrick Delaunay <patrick.delaunay(a)foss.st.com> --- drivers/tee/optee/core.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/drivers/tee/optee/core.c b/drivers/tee/optee/core.c index b898c32edc0b..73dbb22ba097 100644 --- a/drivers/tee/optee/core.c +++ b/drivers/tee/optee/core.c @@ -624,14 +624,14 @@ static int optee_probe(struct udevice *dev) u32 sec_caps; if (!is_optee_api(pdata->invoke_fn)) { - debug("%s: OP-TEE api uid mismatch\n", __func__); + dev_err(dev, "OP-TEE api uid mismatch\n"); return -ENOENT; } print_os_revision(dev, pdata->invoke_fn); if (!api_revision_is_compatible(pdata->invoke_fn)) { - debug("%s: OP-TEE api revision mismatch\n", __func__); + dev_err(dev, "OP-TEE api revision mismatch\n"); return -ENOENT; } @@ -642,7 +642,7 @@ static int optee_probe(struct udevice *dev) */ if (!exchange_capabilities(pdata->invoke_fn, &sec_caps) || !(sec_caps & OPTEE_SMC_SEC_CAP_DYNAMIC_SHM)) { - debug("%s: OP-TEE capabilities mismatch\n", __func__); + dev_err(dev, "OP-TEE capabilities mismatch\n"); return -ENOENT; } -- 2.30.1

3 4