September 2022 - U-Boot

Adding 'N:' field support to buildman's boards database
by Tom Rini 11 Oct '22

11 Oct '22

Hey, So looking at series such as https://patchwork.ozlabs.org/project/uboot/patch/20220801113804.v2.1.I49e41… we run in to the case that instead of using "F: configs/..." to list each defconfig, we use the "N:" syntax, which is to list a regex of files for the given MAINTAINERS entry. We need to support this format in order to continue to have CI pass once we merge the above mentioned series. -- Tom

2 3

[RFC PATCH v4 00/28] Modernize U-Boot shell
by Francis Laniel 10 Oct '22

10 Oct '22

Hi. First I hope you are fine and the same for your relatives. During 2021 summer, Sean Anderson wrote a contribution to add a new shell, based on LIL, to U-Boot [1][2]. While one of the goals of this contribution was to address the fact actual U-Boot shell, which is based on Busybox hush, is old there was a discussion about adding a new shell versus updating the actual one [3][4]. So, in this series, with Harald Seiler, we updated the actual U-Boot shell to reflect what is currently in Busybox source code. Basically, this contribution is about taking a snapshot of Busybox shell/hush.c file (as it exists in commit 37460f5da) and adapt it to suit U-Boot needs. This contribution was written to be as backward-compatible as possible to avoid breaking the existing. So, the 2021 hush flavor offers the same as the actual, that is to say: 1. Variable expansion. 2. Instruction lists (;, && and ||). 3. If, then and else. 4. Loops (for, while and until). No new features offered by Busybox hush were implemented (e.g. functions). It is possible to change the parser at runtime using the "parser" command: => parser print old => parser set 2021 2021> parser print 2021 2021> parser set old => The default parser is the old one. In terms of testing, new unit tests were added to ut to ensure the new behavior is the same as the old one and it does not add regression. Nonetheless, if old behavior was buggy and fixed upstream, the fix is then added to U-Boot [5]. In sandbox, all of these tests pass smoothly: => printenv board board=sandbox => ut hush Running 20 hush tests ... Failures: 0 => parser set 2021 2021> ut hush Running 20 hush tests ... Failures: 0 Thanks to the effort of Harald Seiler, I was successful booting a board: 2021> printenv board_rev board_rev=iMX8MP 2021> boot ... root@iMX8MPboard:~# fw_printenv board_rev board_rev=iMX8MP I also tested another board where both parser were successful to boot it: => printenv fdtfile fdtfile=amlogic/meson-gxl-s905x-libretech-cc.dtb => boot ... root@lepotato:~# root@lepotato:~# reboot ... => parser set 2021 2021> printenv fdtfile fdtfile=amlogic/meson-gxl-s905x-libretech-cc.dtb 2021> boot ... root@lepotato:~# Sadly, I was not able to have the CI passes smoothly... To pass it and be aware of all the troubles, I added the 3 last commits as "trick" commits. So, I would like to have your opinion on these commits and particularly on the underlying problems: 1. When I added a test to ensure variable expansion behavior is correct, I added some tests which should print "unknown command" (to test the shell behavior when a bad output is given). Sadly, having "unknown command" printed will make the CI fails. I think we can simply remove this "negative" test but what do you think? 2. When running ut setexpr setexpr_test_str from test.py the test fails while running the following assert: ut_assertok(ut_check_delta(start_mem)); I am able to reproduce it while running within test.py but not from sandbox. So, I would like to have your reviews or to get a bit of help to know how I can better debug this specific case? 3. The keymile board is the only board to call get_local_var(), set_local_var() and unset_local_var(). Sadly, these functions are static in this contribution. I could have change all of them to introduce code like this: *_local_var(/*...*/) { if (gd->flags & GD_FLG_HUSH_OLD_PARSER) return *_local_var_old(/*...*/); if (gd->flags & GD_FLG_HUSH_2021_PARSER) return *_local_var_2021(/*...*/); } But this would have mean renaming all old hush functions calls and I did not want to change the old hush particularly to avoid breaking things. Instead, I change the keymile board to use environment variable instead of local ones. Do you think this a good change or I should change the old hush instead? For all these reasons, I marked this contribution as RFC to indeed collect your opinions. My goal is not to change suddenly actual shell to this one, we clearly need a transition period to think about it. I think it is better to see this contribution as a proof of concept which shows it is possible to update the actual shell. If you want to review it - your review will really be appreciated - here are some information regarding the commits: * commits marked as "test:" deal with unit tests. * commit "cli: Add Busybox upstream hush.c file." copies Busybox shell/hush.c into U-Boot tree, this explain why this commit contains around 12000 additions. * commit "cli: Port Busybox 2021 hush to U-Boot." modifies previously added file to permit us to use this as new shell. The really good idea of #include'ing Busybox code into a wrapper file to define some particular functions while minimizing modifications to upstream code comes from Harald Seiler. * commit "cmd: Add new parser command" adds a new command which permits selecting parser at runtime. I am not really satisfied with the fact it calls cli_init() and cli_loop() each time the parser is set, so your reviews would be welcomed. * Other commits focus on enabling features we need (e.g. if). Changes since v2: * Added a small fix to compile sandbox with NO_SDL=1. * Added a command to change parser at runtime. * Added 2021 parser function to all run_command*(). Changes since v3: * Various bug fixes pointed by the CI. * Added upstream busybox hush commits until 6th February 2022. Francis Laniel (27): video: sandbox: Add dummy function for sandbox_sdl_remove_display(). test: Add framework to test hush behavior test: hush: Test hush if/else test/py: hush_if_test: Remove the test file test: hush: Test hush variable expansion test: hush: Test hush commands list test: hush: Test hush loops cli: Add Busybox upstream hush.c file cli: Port Busybox 2021 hush to U-Boot cli: Add menu for hush parser global_data.h: add GD_FLG_HUSH_OLD_PARSER flag. cmd: Add new parser command. cli: Enables using hush 2021 parser as command line parser cli: hush_2021: Enable variables expansion for hush 2021 cli: hush_2021: Add functions to be called from run_command() cli: add hush 2021 as parser for run_command*(). test: hush: Fix instructions list tests for hush 2021 test: hush: Fix variable expansion tests for hush 2021 cli: hush_2021: Enable using \< and \> as string compare operators cli: hush_2021: Enable if keyword test: hush: Fix if tests for hush 2021 cli: hush_2021: Enable loops test: hush: Fix loop tests for hush 2021 cli: hush_2021: Add upstream commits up to 6th February 2022. for test purpose only: Comment out dollar tests which prints error messages. for test purpose only: Comment out failed function which fails only in CI. board: keymile: common: Use environment to store IVM_* variables. Tom Rini (1): Modernize U-Boot shell arch/sandbox/include/asm/sdl.h | 5 + board/keymile/common/common.c | 8 +- board/keymile/common/ivm.c | 9 +- cmd/Kconfig | 21 + cmd/Makefile | 2 + cmd/parser.c | 125 + common/Makefile | 3 +- common/cli.c | 81 +- common/cli_hush_2021.c | 303 + common/cli_hush_upstream.c | 12910 +++++++++++++++++++++++++++ include/asm-generic/global_data.h | 8 + include/cli_hush.h | 49 +- include/test/hush.h | 15 + include/test/suites.h | 1 + test/Makefile | 3 + test/cmd/setexpr.c | 2 +- test/cmd_ut.c | 6 + test/hush/Makefile | 10 + test/hush/cmd_ut_hush.c | 20 + test/hush/dollar.c | 226 + test/hush/if.c | 353 + test/hush/list.c | 140 + test/hush/loop.c | 90 + test/py/tests/test_hush_if_test.py | 184 - 24 files changed, 14364 insertions(+), 210 deletions(-) create mode 100644 cmd/parser.c create mode 100644 common/cli_hush_2021.c create mode 100644 common/cli_hush_upstream.c create mode 100644 include/test/hush.h create mode 100644 test/hush/Makefile create mode 100644 test/hush/cmd_ut_hush.c create mode 100644 test/hush/dollar.c create mode 100644 test/hush/if.c create mode 100644 test/hush/list.c create mode 100644 test/hush/loop.c delete mode 100644 test/py/tests/test_hush_if_test.py Best regards and thank you in advance for your reviews. --- [1] https://lists.denx.de/pipermail/u-boot/2021-July/453347.html [2] https://runtimeterror.com/tech/lil/ [3] https://lists.denx.de/pipermail/u-boot/2021-July/453790.html [4] https://lists.denx.de/pipermail/u-boot/2021-July/453848.html [5] https://lists.denx.de/pipermail/u-boot/2021-August/458569.html -- 2.25.1

4 43

[PATCH v2 00/38] binman: Add support for bintools and missing tools
by Simon Glass 10 Oct '22

10 Oct '22

At present binman uses binary tools (like cbfstool, futiltiy, lz4) in an ad-hoc manner. Various parts of binman use tools.Run() to run tools as needed. If a tool is missing, an error is produced and binman stops. However this is not ideal. CI systems want to be able to complete the build, even if tools are missing. Ideally binman would deal with missing binary tools the same way it deals with missing binary blobs: make a note of it and move on. This series introduces this feature to binman. `Bintool` is the name binman gives to a binary tool which it uses to create and manipulate binaries that binman cannot handle itself. Binman provides various features to manage bintools: - Determining whether the tool is currently installed - Downloading or building the tool - Determining the version of the tool that is installed - Deciding which tools are needed to build an image As with external blobs, bintools (which are like 'external' tools) can be missing. When building an image requires a bintool and it is not installed, binman detects this and reports the problem, but continues to build an image. Of course the image will not work, but binman reports which bintools are needed and also provide a way to fetch them. The final patch shows how this works in practice with a chosen board. The Odroid-C2 is quite a complicated image with many steps. It is an ideal example for how Binman can be used. The series is available at u-boot-dm/bin-working Changes in v2: - Substantial rewrite, introducing the concept of bintools Simon Glass (38): Makefile: Fake external blobs by default with binman binman: Tweak elf tests for a toolchain change mkimage: Show the external-offset error binman: Expand the external FIT test a little patman: Allow running a tool and returning the full result buildman: Move the download function to tools patman: Tidy up the download function a little patman: Add a function to find a tool on the path binman: Write fake blobs to the output directory binman: Drop the image name from the fake-blob message binman: Allow faked blobs in blob-ext-list binman: Correct path for fip_util binman: Add installation instructions binman: Add support for bintools binman: Plumb in support for bintools binman: Add tests for bintool binman: Add a bintool implementation for cbfstool binman: Add a bintool implementation for fiptool binman: Add a bintool implementation for futility binman: Add a bintool implementation for ifwitool binman: Add a bintool implementation for mkimage binman: Enable bintool tests including cmdline processing binman: Convert to using the CBFS bintool binman: Convert to using the FIP bintool binman: Convert to using the futility bintool binman: Convert to using the ifwitool bintool binman: Convert to using the mkimage bintool binman: Move compression into binman binman: Tidy up pylint warnings in comp_util binman: Add a bintool implementation for lz4 binman: Convert to using the lz4 bintool binman: Add a bintool implementation for lzma_alone binman: Convert to using the lzma_alone bintool binman: Plumb in support for missing bintools binman: Complete test coverage of comp_util binman: Add a command to generate bintool docs binman: Add documentation for bintools RFC: Move Odroid-C2 to use binman to produce the image Makefile | 2 +- arch/arm/dts/meson-sm1-odroid-c4-u-boot.dtsi | 107 +++++ arch/arm/mach-meson/Kconfig | 1 + doc/board/amlogic/odroid-c4.rst | 127 ++--- doc/develop/package/bintools.rst | 1 + tools/binman/binman.rst | 98 +++- tools/binman/bintool.py | 466 +++++++++++++++++++ tools/binman/bintool_test.py | 353 ++++++++++++++ tools/binman/bintools.rst | 139 ++++++ tools/binman/btool/_aml_common.py | 47 ++ tools/binman/btool/_testing.py | 36 ++ tools/binman/btool/aml_encrypt_g12a.py | 82 ++++ tools/binman/btool/aml_encrypt_g12b.py | 83 ++++ tools/binman/btool/cbfstool.py | 219 +++++++++ tools/binman/btool/fiptool.py | 123 +++++ tools/binman/btool/futility.py | 178 +++++++ tools/binman/btool/ifwitool.py | 166 +++++++ tools/binman/btool/lz4.py | 140 ++++++ tools/binman/btool/lzma_alone.py | 126 +++++ tools/binman/btool/mkimage.py | 80 ++++ tools/binman/cbfs_util.py | 33 +- tools/binman/cbfs_util_test.py | 59 +-- tools/binman/cmdline.py | 12 + tools/binman/comp_util.py | 76 +++ tools/binman/control.py | 50 +- tools/binman/elf_test.py | 8 +- tools/binman/entries.rst | 138 ++++++ tools/binman/entry.py | 67 ++- tools/binman/etype/aml_encrypt.py | 262 +++++++++++ tools/binman/etype/blob.py | 8 +- tools/binman/etype/blob_ext_list.py | 1 + tools/binman/etype/fit.py | 20 +- tools/binman/etype/gbb.py | 37 +- tools/binman/etype/intel_ifwi.py | 25 +- tools/binman/etype/mkimage.py | 13 +- tools/binman/etype/section.py | 19 +- tools/binman/etype/vblock.py | 32 +- tools/binman/fip_util.py | 26 -- tools/binman/fip_util_test.py | 25 +- tools/binman/ftest.py | 173 ++++++- tools/binman/image.py | 14 + tools/binman/main.py | 11 +- tools/binman/missing-blob-help | 6 + tools/binman/test/162_fit_external.dts | 2 +- tools/binman/test/213_aml_encrypt.dts | 51 ++ tools/binman/test/214_list_no_dtb.dts | 23 + tools/binman/test/218_blob_ext_list_fake.dts | 14 + tools/buildman/toolchain.py | 46 +- tools/fit_image.c | 5 +- tools/patman/tools.py | 238 +++++----- 50 files changed, 3619 insertions(+), 449 deletions(-) create mode 120000 doc/develop/package/bintools.rst create mode 100644 tools/binman/bintool.py create mode 100644 tools/binman/bintool_test.py create mode 100644 tools/binman/bintools.rst create mode 100644 tools/binman/btool/_aml_common.py create mode 100644 tools/binman/btool/_testing.py create mode 100644 tools/binman/btool/aml_encrypt_g12a.py create mode 100644 tools/binman/btool/aml_encrypt_g12b.py create mode 100644 tools/binman/btool/cbfstool.py create mode 100644 tools/binman/btool/fiptool.py create mode 100644 tools/binman/btool/futility.py create mode 100644 tools/binman/btool/ifwitool.py create mode 100644 tools/binman/btool/lz4.py create mode 100644 tools/binman/btool/lzma_alone.py create mode 100644 tools/binman/btool/mkimage.py create mode 100644 tools/binman/comp_util.py create mode 100644 tools/binman/etype/aml_encrypt.py create mode 100644 tools/binman/test/213_aml_encrypt.dts create mode 100644 tools/binman/test/214_list_no_dtb.dts create mode 100644 tools/binman/test/218_blob_ext_list_fake.dts -- 2.34.1.575.g55b058a8bb-goog

3 82

[PATCH 0/7] Support NAND ONFI EDO mode for imx8mn architecture
by Dario Binacchi 09 Oct '22

09 Oct '22

Debugging and testing the series made it necessary to add some patches to fix correct clock access and compilation issues raised by buildman. Dario Binacchi (3): dm: clk: add missing stub when CONFIG_CLK is deactivated mtd: mxs_nand: don't get the gpmi_apbh_dma clock mtd: mxs_nand: get the clock with the right name Michael Trimarchi (4): clk: imx: gate2 support shared counter and relative clock functions clk: imx: clk-imx8mn add gpmi nand clocks imx: gpmi: Add register needed to control nand bus timing mtd: mxs_nand: Support EDO mode for imx8mn architecture arch/arm/include/asm/mach-imx/regs-gpmi.h | 9 + drivers/clk/imx/clk-gate2.c | 15 +- drivers/clk/imx/clk-imx8mn.c | 14 ++ drivers/clk/imx/clk.h | 27 ++- drivers/mtd/nand/raw/mxs_nand.c | 200 ++++++++++++++++++++++ drivers/mtd/nand/raw/mxs_nand_dt.c | 79 +++++---- include/clk.h | 32 +++- include/mxs_nand.h | 3 + 8 files changed, 333 insertions(+), 46 deletions(-) -- 2.32.0

2 9

[u-boot][PATCH] mtd: nand: Fix SPL build after migration of CONFIG_SYS_NAND_SELF_INIT to Kconfig
by Roger Quadros 09 Oct '22

09 Oct '22

This fixes the below build error if nand.c is included in an SPL build. /work/u-boot/drivers/mtd/nand/raw/nand.c: In function ‘nand_init_chip’: /work/u-boot/drivers/mtd/nand/raw/nand.c:82:28: error: ‘nand_chip’ undeclared (first use in this function) 82 | struct nand_chip *nand = &nand_chip[i]; | ^~~~~~~~~ /work/u-boot/drivers/mtd/nand/raw/nand.c:82:28: note: each undeclared identifier is reported only once for each function it appears in /work/u-boot/drivers/mtd/nand/raw/nand.c:84:20: error: ‘base_address’ undeclared (first use in this function); did you mean ‘base_addr’? 84 | ulong base_addr = base_address[i]; | ^~~~~~~~~~~~ | base_addr Fixes: 068c41f1cc77 ("Finish conversion CONFIG_SYS_NAND_SELF_INIT to Kconfig") Signed-off-by: Roger Quadros <rogerq(a)kernel.org> --- drivers/mtd/nand/raw/nand.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/mtd/nand/raw/nand.c b/drivers/mtd/nand/raw/nand.c index 4b5560dd24..14bca12024 100644 --- a/drivers/mtd/nand/raw/nand.c +++ b/drivers/mtd/nand/raw/nand.c @@ -19,7 +19,7 @@ int nand_curr_device = -1; static struct mtd_info *nand_info[CONFIG_SYS_MAX_NAND_DEVICE]; -#ifndef CONFIG_SYS_NAND_SELF_INIT +#if !CONFIG_IS_ENABLED(SYS_NAND_SELF_INIT) static struct nand_chip nand_chip[CONFIG_SYS_MAX_NAND_DEVICE]; static ulong base_address[CONFIG_SYS_MAX_NAND_DEVICE] = CONFIG_SYS_NAND_BASE_LIST; #endif -- 2.17.1

3 4

[PATCH] arm64: layerscape: Support SYSRESET
by Sean Anderson 09 Oct '22

09 Oct '22

CONFIG_SYSRESET provides its own implementation of reset_cpu. Disable our version when it is enabled. Signed-off-by: Sean Anderson <sean.anderson(a)seco.com> --- arch/arm/cpu/armv8/fsl-layerscape/cpu.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/arch/arm/cpu/armv8/fsl-layerscape/cpu.c b/arch/arm/cpu/armv8/fsl-layerscape/cpu.c index c989a43cbeb..487c0ed5539 100644 --- a/arch/arm/cpu/armv8/fsl-layerscape/cpu.c +++ b/arch/arm/cpu/armv8/fsl-layerscape/cpu.c @@ -1229,6 +1229,7 @@ int timer_init(void) return 0; } +#if !CONFIG_IS_ENABLED(SYSRESET) __efi_runtime_data u32 __iomem *rstcr = (u32 *)CONFIG_SYS_FSL_RST_ADDR; void __efi_runtime reset_cpu(void) @@ -1248,6 +1249,7 @@ void __efi_runtime reset_cpu(void) scfg_out32(rstcr, val); #endif } +#endif #if defined(CONFIG_EFI_LOADER) && !defined(CONFIG_PSCI_RESET) -- 2.35.1.1320.gc452695387.dirty

2 1

[PATCH] ls1043ardb: nand driver fixups for revision v7.0 boards
by Wei Lu 09 Oct '22

09 Oct '22

The LS1043ARDB rev v7.0 board replace nand device MT29F4G08ABBDAH4-AITX:D with MT29F4G08ABBFAH4-AIT:F. Reflecting this change in board_fix_fdt(). CPLD V3.0 is needed for nandboot as the nand device changed. A new macro CPLD_CFG_RCW_SRC_NAND_4K(4Kpage) is added to distinguish from CPLD_CFG_RCW_SRC_NAND(2Kpage) to support nandboot on rev v7.0 board. Signed-off-by: Wei Lu <w.lu(a)nxp.com> --- board/freescale/ls1043ardb/cpld.c | 4 +++ board/freescale/ls1043ardb/cpld.h | 1 + board/freescale/ls1043ardb/ls1043ardb.c | 46 ++++++++++++++++++++++++- 3 files changed, 50 insertions(+), 1 deletion(-) diff --git a/board/freescale/ls1043ardb/cpld.c b/board/freescale/ls1043ardb/cpld.c index 5d2e8015a0..232035638b 100644 --- a/board/freescale/ls1043ardb/cpld.c +++ b/board/freescale/ls1043ardb/cpld.c @@ -69,6 +69,10 @@ void cpld_set_defbank(void) void cpld_set_nand(void) { u16 reg = CPLD_CFG_RCW_SRC_NAND; + + if (CPLD_READ(cpld_ver) > 0x2) + reg = CPLD_CFG_RCW_SRC_NAND_4K; + u8 reg5 = (u8)(reg >> 1); u8 reg6 = (u8)(reg & 1); diff --git a/board/freescale/ls1043ardb/cpld.h b/board/freescale/ls1043ardb/cpld.h index 2e757b557f..eed34d6354 100644 --- a/board/freescale/ls1043ardb/cpld.h +++ b/board/freescale/ls1043ardb/cpld.h @@ -41,5 +41,6 @@ void cpld_rev_bit(unsigned char *value); #define CPLD_BANK_SEL_ALTBANK 0x04 #define CPLD_CFG_RCW_SRC_NOR 0x025 #define CPLD_CFG_RCW_SRC_NAND 0x106 +#define CPLD_CFG_RCW_SRC_NAND_4K 0x118 #define CPLD_CFG_RCW_SRC_SD 0x040 #endif diff --git a/board/freescale/ls1043ardb/ls1043ardb.c b/board/freescale/ls1043ardb/ls1043ardb.c index f388eb496f..8c91f0771f 100644 --- a/board/freescale/ls1043ardb/ls1043ardb.c +++ b/board/freescale/ls1043ardb/ls1043ardb.c @@ -167,7 +167,7 @@ int checkboard(void) if (cfg_rcw_src == 0x25) printf("vBank %d\n", CPLD_READ(vbank)); - else if (cfg_rcw_src == 0x106) + else if ((cfg_rcw_src == 0x106) || (cfg_rcw_src == 0x118)) puts("NAND\n"); else printf("Invalid setting of SW4\n"); @@ -347,10 +347,54 @@ int ft_board_setup(void *blob, struct bd_info *bd) return 0; } +void nand_fixup(void) +{ + u32 csor = 0; + + if (CPLD_READ(pcba_ver) < 0x7) + return; + + /* Change NAND Flash PGS/SPRZ configuration */ + csor = CONFIG_SYS_NAND_CSOR; + if ((csor & CSOR_NAND_PGS_MASK) == CSOR_NAND_PGS_2K) + csor = (csor & ~(CSOR_NAND_PGS_MASK)) | CSOR_NAND_PGS_4K; + + if ((csor & CSOR_NAND_SPRZ_MASK) == CSOR_NAND_SPRZ_64) + csor = (csor & ~(CSOR_NAND_SPRZ_MASK)) | CSOR_NAND_SPRZ_224; + + if (IS_ENABLED(CONFIG_TFABOOT)) { + u8 cfg_rcw_src1, cfg_rcw_src2; + u16 cfg_rcw_src; + + cfg_rcw_src1 = CPLD_READ(cfg_rcw_src1); + cfg_rcw_src2 = CPLD_READ(cfg_rcw_src2); + cpld_rev_bit(&cfg_rcw_src1); + cfg_rcw_src = cfg_rcw_src1; + cfg_rcw_src = (cfg_rcw_src << 1) | cfg_rcw_src2; + + if (cfg_rcw_src == 0x25) + set_ifc_csor(IFC_CS1, csor); + else if (cfg_rcw_src == 0x118) + set_ifc_csor(IFC_CS0, csor); + else + printf("Invalid setting\n"); + } else { + if (IS_ENABLED(CONFIG_NAND_BOOT)) + set_ifc_csor(IFC_CS0, csor); + else + set_ifc_csor(IFC_CS1, csor); + } +} + #if IS_ENABLED(CONFIG_OF_BOARD_FIXUP) int board_fix_fdt(void *blob) { + /* nand driver fix up */ + nand_fixup(); + + /* fdt fix up */ fdt_fixup_phy_addr(blob); + return 0; } #endif -- 2.25.1

2 2

[PATCH] arm64: ls104x: Enable eDMA snooping
by Sean Anderson 09 Oct '22

09 Oct '22

This enables eDMA snooping on the LS1043A and LS1046A. This will allow marking the I2C, LPUART, and SPI devices on these SoCs as DMA coherent. Oddly, this bit is only documented for the LS1043A, and is marked as "reserved" in the LS1046ARM. I have tested this patch on the LS1046A and found that marking i2c0 as dma-coherent works without issue. Signed-off-by: Sean Anderson <sean.anderson(a)seco.com> --- arch/arm/cpu/armv8/fsl-layerscape/soc.c | 2 +- arch/arm/include/asm/arch-fsl-layerscape/immap_lsch2.h | 1 + 2 files changed, 2 insertions(+), 1 deletion(-) diff --git a/arch/arm/cpu/armv8/fsl-layerscape/soc.c b/arch/arm/cpu/armv8/fsl-layerscape/soc.c index 926f8f21b63..515dbe02fd7 100644 --- a/arch/arm/cpu/armv8/fsl-layerscape/soc.c +++ b/arch/arm/cpu/armv8/fsl-layerscape/soc.c @@ -682,7 +682,7 @@ void fsl_lsch2_early_init_f(void) SCFG_SNPCNFGCR_USB1WRSNP | SCFG_SNPCNFGCR_USB2RDSNP | SCFG_SNPCNFGCR_USB2WRSNP | SCFG_SNPCNFGCR_USB3RDSNP | SCFG_SNPCNFGCR_USB3WRSNP | SCFG_SNPCNFGCR_SATARDSNP | - SCFG_SNPCNFGCR_SATAWRSNP); + SCFG_SNPCNFGCR_SATAWRSNP | SCFG_SNPCNFGCR_EDMASNP); #elif defined(CONFIG_ARCH_LS1012A) setbits_be32(&scfg->snpcnfgcr, SCFG_SNPCNFGCR_SECRDSNP | SCFG_SNPCNFGCR_SECWRSNP | SCFG_SNPCNFGCR_USB1RDSNP | diff --git a/arch/arm/include/asm/arch-fsl-layerscape/immap_lsch2.h b/arch/arm/include/asm/arch-fsl-layerscape/immap_lsch2.h index 06adf669390..e86cfba8669 100644 --- a/arch/arm/include/asm/arch-fsl-layerscape/immap_lsch2.h +++ b/arch/arm/include/asm/arch-fsl-layerscape/immap_lsch2.h @@ -413,6 +413,7 @@ struct ccsr_gur { #define SCFG_SNPCNFGCR_SATAWRSNP 0x00400000 #define SCFG_SNPCNFGCR_USB1RDSNP 0x00200000 #define SCFG_SNPCNFGCR_USB1WRSNP 0x00100000 +#define SCFG_SNPCNFGCR_EDMASNP 0x00020000 #define SCFG_SNPCNFGCR_USB2RDSNP 0x00008000 #define SCFG_SNPCNFGCR_USB2WRSNP 0x00010000 #define SCFG_SNPCNFGCR_USB3RDSNP 0x00002000 -- 2.35.1.1320.gc452695387.dirty

2 2

[PATCH v3] gpio: ftgpio010: Add support for Faraday Technology FTGPIO010
by Sergei Antonov 07 Oct '22

07 Oct '22

Add Faraday Technology's FTGPIO010 controller driver. Signed-off-by: Sergei Antonov <saproj(a)gmail.com> --- v2 -> v3: Implement .get_function to make "gpio status" command work. v1 -> v2: Replace setbits_le32() with a simpler function out_le32(). Replace readl() with in_le32() to respect endianness. drivers/gpio/Kconfig | 6 +++ drivers/gpio/Makefile | 1 + drivers/gpio/ftgpio010.c | 111 +++++++++++++++++++++++++++++++++++++++ 3 files changed, 118 insertions(+) create mode 100644 drivers/gpio/ftgpio010.c diff --git a/drivers/gpio/Kconfig b/drivers/gpio/Kconfig index c949f9d2f7cc..2a60478b476a 100644 --- a/drivers/gpio/Kconfig +++ b/drivers/gpio/Kconfig @@ -605,4 +605,10 @@ config TURRIS_OMNIA_MCU help Support for GPIOs on MCU connected to Turris Omnia via i2c. +config FTGPIO010 + bool "Faraday Technology FTGPIO010 driver" + depends on DM_GPIO + help + Support for GPIOs on Faraday Technology's FTGPIO010 controller. + endif diff --git a/drivers/gpio/Makefile b/drivers/gpio/Makefile index 9d718a554e59..eee7908871d8 100644 --- a/drivers/gpio/Makefile +++ b/drivers/gpio/Makefile @@ -75,3 +75,4 @@ obj-$(CONFIG_SL28CPLD_GPIO) += sl28cpld-gpio.o obj-$(CONFIG_ZYNQMP_GPIO_MODEPIN) += zynqmp_gpio_modepin.o obj-$(CONFIG_SLG7XL45106_I2C_GPO) += gpio_slg7xl45106.o obj-$(CONFIG_$(SPL_TPL_)TURRIS_OMNIA_MCU) += turris_omnia_mcu.o +obj-$(CONFIG_FTGPIO010) += ftgpio010.o diff --git a/drivers/gpio/ftgpio010.c b/drivers/gpio/ftgpio010.c new file mode 100644 index 000000000000..6c091d4fd874 --- /dev/null +++ b/drivers/gpio/ftgpio010.c @@ -0,0 +1,111 @@ +// SPDX-License-Identifier: GPL-2.0+ +/* + * Faraday Technology's FTGPIO010 controller. + */ + +#include <common.h> +#include <dm.h> +#include <asm/io.h> +#include <asm/gpio.h> + +struct ftgpio010_regs { + u32 out; + u32 in; + u32 direction; // 1 - output + u32 reserved; + u32 set; + u32 clear; +}; + +struct ftgpio010_plat { + struct ftgpio010_regs __iomem *regs; +}; + +static int ftgpio010_direction_input(struct udevice *dev, unsigned int pin) +{ + struct ftgpio010_plat *plat = dev_get_plat(dev); + struct ftgpio010_regs *const regs = plat->regs; + + clrbits_le32(&regs->direction, 1 << pin); + return 0; +} + +static int ftgpio010_direction_output(struct udevice *dev, unsigned int pin, + int val) +{ + struct ftgpio010_plat *plat = dev_get_plat(dev); + struct ftgpio010_regs *const regs = plat->regs; + + /* change the data first, then the direction. to avoid glitch */ + out_le32(val ? &regs->set : &regs->clear, 1 << pin); + setbits_le32(&regs->direction, 1 << pin); + + return 0; +} + +static int ftgpio010_get_value(struct udevice *dev, unsigned int pin) +{ + struct ftgpio010_plat *plat = dev_get_plat(dev); + struct ftgpio010_regs *const regs = plat->regs; + + return in_le32(&regs->in) >> pin & 1; +} + +static int ftgpio010_set_value(struct udevice *dev, unsigned int pin, int val) +{ + struct ftgpio010_plat *plat = dev_get_plat(dev); + struct ftgpio010_regs *const regs = plat->regs; + + out_le32(val ? &regs->set : &regs->clear, 1 << pin); + return 0; +} + +static int ftgpio010_get_function(struct udevice *dev, unsigned int pin) +{ + struct ftgpio010_plat *plat = dev_get_plat(dev); + struct ftgpio010_regs *const regs = plat->regs; + + if (in_le32(&regs->direction) >> pin & 1) + return GPIOF_OUTPUT; + return GPIOF_INPUT; +} + +static int ftgpio010_probe(struct udevice *dev) +{ + struct gpio_dev_priv *uc_priv = dev_get_uclass_priv(dev); + + uc_priv->gpio_count = ofnode_read_u32_default(dev_ofnode(dev), + "nr-gpios", 32); + return 0; +} + +static int ftgpio010_of_to_plat(struct udevice *dev) +{ + struct ftgpio010_plat *plat = dev_get_plat(dev); + + plat->regs = dev_read_addr_ptr(dev); + return 0; +} + +static const struct dm_gpio_ops ftgpio010_ops = { + .direction_input = ftgpio010_direction_input, + .direction_output = ftgpio010_direction_output, + .get_value = ftgpio010_get_value, + .set_value = ftgpio010_set_value, + .get_function = ftgpio010_get_function, +}; + +static const struct udevice_id ftgpio010_ids[] = { + { .compatible = "faraday,ftgpio010" }, + { } +}; + +U_BOOT_DRIVER(ftgpio010) = { + .name = "ftgpio010", + .id = UCLASS_GPIO, + .of_match = ftgpio010_ids, + .ops = &ftgpio010_ops, + .of_to_plat = ftgpio010_of_to_plat, + .plat_auto = sizeof(struct ftgpio010_plat), + .probe = ftgpio010_probe, +}; -- 2.34.1

3 3

[PATCH v2 1/4] common/board_f: remove XTRN_DECLARE_GLOBAL_DATA_PTR dead code
by Ovidiu Panait 07 Oct '22

07 Oct '22

The XTRN_DECLARE_GLOBAL_DATA_PTR declarations in ppc code are permanently commented out, so there are no users for this macro: #if 1 #define DECLARE_GLOBAL_DATA_PTR register volatile gd_t *gd asm ("r2") #else /* We could use plain global data, but the resulting code is bigger */ #define XTRN_DECLARE_GLOBAL_DATA_PTR extern #define DECLARE_GLOBAL_DATA_PTR XTRN_DECLARE_GLOBAL_DATA_PTR \ gd_t *gd #endif Remove all references to this macro, but add a documentation note regarding the possibility of using plain global data for the GD pointer. Signed-off-by: Ovidiu Panait <ovpanait(a)gmail.com> --- Changes in v2: - update global data documentation with the info provided by Wolfgang arch/powerpc/include/asm/global_data.h | 6 ------ common/board_f.c | 11 ----------- doc/develop/global_data.rst | 5 +++++ 3 files changed, 5 insertions(+), 17 deletions(-) diff --git a/arch/powerpc/include/asm/global_data.h b/arch/powerpc/include/asm/global_data.h index 6709e692e6..6ed21c781f 100644 --- a/arch/powerpc/include/asm/global_data.h +++ b/arch/powerpc/include/asm/global_data.h @@ -92,12 +92,6 @@ struct arch_global_data { #include <asm-generic/global_data.h> -#if 1 #define DECLARE_GLOBAL_DATA_PTR register volatile gd_t *gd asm ("r2") -#else /* We could use plain global data, but the resulting code is bigger */ -#define XTRN_DECLARE_GLOBAL_DATA_PTR extern -#define DECLARE_GLOBAL_DATA_PTR XTRN_DECLARE_GLOBAL_DATA_PTR \ - gd_t *gd -#endif #endif /* __ASM_GBL_DATA_H */ diff --git a/common/board_f.c b/common/board_f.c index 18e2246733..f4238d4c90 100644 --- a/common/board_f.c +++ b/common/board_f.c @@ -57,18 +57,7 @@ #include <dm/root.h> #include <linux/errno.h> -/* - * Pointer to initial global data area - * - * Here we initialize it if needed. - */ -#ifdef XTRN_DECLARE_GLOBAL_DATA_PTR -#undef XTRN_DECLARE_GLOBAL_DATA_PTR -#define XTRN_DECLARE_GLOBAL_DATA_PTR /* empty = allocate here */ -DECLARE_GLOBAL_DATA_PTR = (gd_t *)(CONFIG_SYS_INIT_GD_ADDR); -#else DECLARE_GLOBAL_DATA_PTR; -#endif /* * TODO(sjg(a)chromium.org): IMO this code should be diff --git a/doc/develop/global_data.rst b/doc/develop/global_data.rst index 2ac893de49..d143f27eed 100644 --- a/doc/develop/global_data.rst +++ b/doc/develop/global_data.rst @@ -36,6 +36,11 @@ On most architectures the global data pointer is stored in a register. The sandbox, x86_64, and Xtensa are notable exceptions. +Current implementation uses a register for the GD pointer because this results +in smaller code. However, using plain global data for the GD pointer would be +possible too (and simpler, as it does not require the reservation of a specific +register for it), but the resulting code is bigger. + Clang for ARM does not support assigning a global register. When using Clang gd is defined as an inline function using assembly code. This adds a few bytes to the code size. -- 2.25.1

3 9