July 2015 - U-Boot

[U-Boot] [PATCH 0/10] x86: Add generic VESA video driver
by Simon Glass 01 Aug '19

01 Aug '19

This series introduces a generic VESA video driver and adjusts the generic PCI ROM support to support it. It also fixes up some problems which prevent this from working on PowerPC at present. It is now possible to use VESA graphics: - On PowerPC, using the x86 emulator - On x86, using either the x86 emulator or native code execution As it turns out, on the only platform I have tested (link), the x86 emulator is only slightly slower that native code execution (200ms difference). Still, we may as well leave this feature in there. VGA ROMs are a very old technology but they are still used by modern platforms, including at least link and some Atom designs. Simon Glass (10): bios_emulator: Fix an #ifdef typo in the header file x86: Correct endianness isues in pci_rom x86: Support ROMs on other archs bios_emulator: Don't display error when emulator terminates bios_emulator: Add some VESA interface debugging x86: pci: Don't stop when we get a vendor/device mismatch x86: Add a VESA video driver x86: Drop the x86_fb driver x86: Access the VGA ROM when needed RFC: Test code for glacier PCI video support arch/x86/include/asm/u-boot-x86.h | 2 + drivers/bios_emulator/atibios.c | 161 +++++++++++++++++++++++---- drivers/bios_emulator/include/x86emu/debug.h | 2 +- drivers/bios_emulator/x86emu/ops.c | 2 +- drivers/pci/pci_auto.c | 28 ++++- drivers/pci/pci_rom.c | 40 ++++--- drivers/video/Makefile | 2 +- drivers/video/vesa_fb.c | 64 +++++++++++ drivers/video/x86_fb.c | 38 ------- include/configs/canyonlands.h | 31 ++++++ include/configs/chromebook_link.h | 2 +- include/pci.h | 9 ++ include/pci_rom.h | 1 - include/vbe.h | 9 +- 14 files changed, 312 insertions(+), 79 deletions(-) create mode 100644 drivers/video/vesa_fb.c delete mode 100644 drivers/video/x86_fb.c -- 2.2.0.rc0.207.ga3a616c

3 42

[U-Boot] [PATCH 1/2] usb: gadget: fastboot: Request status and length check in rx handler
by Paul Kocialkowski 18 Mar '19

18 Mar '19

This avoids handling requests that have an error status or no data. In particular, this avoids showing unnecessary error messages when the USB gadget gets disconnected (e.g. with fastboot continue) and the fastboot USB gadget driver sends an error back to the host (that has disconnected already). Signed-off-by: Paul Kocialkowski <contact(a)paulk.fr> --- drivers/usb/gadget/f_fastboot.c | 11 ++++++----- 1 file changed, 6 insertions(+), 5 deletions(-) diff --git a/drivers/usb/gadget/f_fastboot.c b/drivers/usb/gadget/f_fastboot.c index 206b6d1..b9a9099 100644 --- a/drivers/usb/gadget/f_fastboot.c +++ b/drivers/usb/gadget/f_fastboot.c @@ -635,6 +635,9 @@ static void rx_handler_command(struct usb_ep *ep, struct usb_request *req) void (*func_cb)(struct usb_ep *ep, struct usb_request *req) = NULL; int i; + if (req->status != 0 || req->length == 0) + return; + for (i = 0; i < ARRAY_SIZE(cmd_dispatch_info); i++) { if (!strcmp_l1(cmd_dispatch_info[i].cmd, cmdbuf)) { func_cb = cmd_dispatch_info[i].cb; @@ -656,9 +659,7 @@ static void rx_handler_command(struct usb_ep *ep, struct usb_request *req) } } - if (req->status == 0) { - *cmdbuf = '\0'; - req->actual = 0; - usb_ep_queue(ep, req, 0); - } + *cmdbuf = '\0'; + req->actual = 0; + usb_ep_queue(ep, req, 0); } -- 1.9.1

4 15

[U-Boot] [PATCH v2 0/4] ARM: atmel: boards: enable SPL support
by Bo Shen 22 Feb '19

22 Feb '19

This patch series enable SPL support for following boards: - at91sam9m10g45ek - NAND flash boot support - SD card boot support - at91sam9n12ek - NAND flash boot support - SPI flash boot support - at91sam9x5ek - NAND flash boot support - SPI flash boot support Changes in v2: - Remove the meaningless prefix "+S:" in configuration file. Bo Shen (4): ARM: atmel: arm926ejs: fix clock configuration ARM: atmel: at91sam9m10g45ek: enable spl support ARM: atmel: at91sam9x5ek: enable spl support ARM: atmel: at91sam9n12ek: enable spl support arch/arm/mach-at91/Kconfig | 3 + arch/arm/mach-at91/Makefile | 2 + arch/arm/mach-at91/arm926ejs/clock.c | 54 +++++++++-------- arch/arm/mach-at91/arm926ejs/u-boot-spl.lds | 48 +++++++++++++++ arch/arm/mach-at91/include/mach/at91_pmc.h | 6 +- arch/arm/mach-at91/include/mach/at91sam9x5.h | 10 ++++ arch/arm/mach-at91/mpddrc.c | 3 +- arch/arm/mach-at91/spl.c | 2 +- arch/arm/mach-at91/spl_at91.c | 11 +++- board/atmel/at91sam9m10g45ek/at91sam9m10g45ek.c | 80 +++++++++++++++++++++++++ board/atmel/at91sam9n12ek/at91sam9n12ek.c | 73 ++++++++++++++++++++++ board/atmel/at91sam9x5ek/at91sam9x5ek.c | 74 +++++++++++++++++++++++ configs/at91sam9m10g45ek_mmc_defconfig | 1 + configs/at91sam9m10g45ek_nandflash_defconfig | 1 + configs/at91sam9n12ek_nandflash_defconfig | 1 + configs/at91sam9n12ek_spiflash_defconfig | 1 + configs/at91sam9x5ek_nandflash_defconfig | 1 + configs/at91sam9x5ek_spiflash_defconfig | 1 + include/configs/at91sam9m10g45ek.h | 58 ++++++++++++++++++ include/configs/at91sam9n12ek.h | 58 +++++++++++++++++- include/configs/at91sam9x5ek.h | 57 ++++++++++++++++++ 21 files changed, 513 insertions(+), 32 deletions(-) create mode 100644 arch/arm/mach-at91/arm926ejs/u-boot-spl.lds -- 2.3.3.220.g9ab698f

4 14

[U-Boot] Fwd: u-boot.img size limit for spl fat load ?
by Ayoub Zaki 18 Oct '17

18 Oct '17

Hello Albert, First tank you so much for inversting your time for replying. I put some printf in the code and SPL is hanging in spl_load_image_fat function. I also enable the DEBUG messages here the complete log file : http://pastebin.com/raw.php?i=BwUJXDDb So I suspect that SPL is not allocating enough memory to load u-boot.img but I can't find where this size limit is set and where I can change it. Cheers, Ayoub > 2015-01-19 8:08 GMT+01:00 Albert ARIBAUD <albert.u.boot(a)aribaud.net>: > > > Hello Ayoub, > > > > On Fri, 16 Jan 2015 15:42:37 +0100, Ayoub Zaki > > <ayoub.zaki(a)googlemail.com> wrote: > > > Hello list, > > > > > > I'm using u-boot-2014.10 and trying to build it with LCD support for the > > > Beaglebone black to include a Boot logo image. > > > It seems that spl hangs while fat loading a 2.8MB sized u-boot.img ! > > > I reduced the the logo header file and rebuild u-boot, that worked for a > > > 1.6MB sized u-boot.img. > > > Is there any size limit for u-boot ? how to increase the memory allocated > > > by spl ? > > > > There should not be a size limit for SPL fat load. > > > > What may happen is that SPL code is overwritten before it exits, or > > U-Boot itself is corrupted after being loaded. > > > > I assume you've seen the SPL banner and not the U-Boot one; that alone > > does not tell us whether SPL or U-Boot did hang. Can you build U-boot > > with debug enabled? If this results in too big an SPL image, then try > > and at least enable debug for the image chainloading parts, so that you > > can see how far SPL code gets run. > > > > > Thanks you. > > > > No problem. > > > > Amicalement, > > -- > > Albert.

3 7

[U-Boot] [PATCH] tools/fw_env: use fsync to ensure that data is physically stored
by Michael Heimpold 25 Jul '17

25 Jul '17

Closing a file descriptor does not guarantee that the data has been successfully saved to disk, as the kernel might defer the write. Signed-off-by: Michael Heimpold <mhei(a)heimpold.de> --- tools/env/fw_env.c | 12 ++++++++++++ 1 file changed, 12 insertions(+) diff --git a/tools/env/fw_env.c b/tools/env/fw_env.c index 428e374..0420495 100644 --- a/tools/env/fw_env.c +++ b/tools/env/fw_env.c @@ -1033,7 +1033,19 @@ static int flash_io (int mode) rc = flash_write (fd_current, fd_target, dev_target); + if (fsync (fd_current)) { + fprintf (stderr, + "fsync failed on %s: %s\n", + DEVNAME (dev_current), strerror (errno)); + } + if (HaveRedundEnv) { + if (fsync (fd_target)) { + fprintf (stderr, + "fsync failed on %s: %s\n", + DEVNAME (dev_current), strerror (errno)); + } + if (close (fd_target)) { fprintf (stderr, "I/O error on %s: %s\n", -- 1.7.10.4

5 6

[U-Boot] [PATCH v2 0/5] arm: imx6: tqma6: Kconfig rework
by Markus Niebel 19 Feb '17

19 Feb '17

From: Markus Niebel <Markus.Niebel(a)tq-group.com> This series extends the Kconfig options for TQMa6 to enable removal of CONFIG_SYS_EXTRA_OPTIONS for this boards. To enable env device / boot device configuration the first patch includes the Kconfig generated header in the board configuration header. To enable CPU selection the second patch moves the board selection to the i.MX6 CPU menu as implemented by the Seco mx6 boards Changed for v2: - rebase to current u-boot-imx - add new patch to fix envcrc build for non default boot / env devices - regenerate defconfigs to exclude defaults Markus Niebel (5): arm: imx6: tqma6: add kconfig to board header arm: mx6: tqma6: CPU type selection via Kconfig arm: imx6: tqma6: implement IMX_CONFIG with Kconfig arm: imx6: tqma6: boot device selection via Kconfig arm: imx6: tqma6: implement KConfig baseboard selection arch/arm/Kconfig | 5 ---- arch/arm/cpu/armv7/mx6/Kconfig | 5 ++++ board/tqc/tqma6/Kconfig | 58 +++++++++++++++++++++++++++++++++++++++ configs/tqma6q_mba6_mmc_defconfig | 2 +- configs/tqma6q_mba6_spi_defconfig | 3 +- configs/tqma6s_mba6_mmc_defconfig | 3 +- configs/tqma6s_mba6_spi_defconfig | 4 ++- include/configs/tqma6.h | 1 + 8 files changed, 72 insertions(+), 9 deletions(-) -- 2.3.0

3 9

[U-Boot] [RFC PATCH] arm: Enable CONFIG_USE_ARCH_MEMSET/MEMCPY globally
by Tom Rini 18 Oct '16

18 Oct '16

- Move the obj- lines for memset.S/memcpy.S to outside of an SPL check so that SPL can use them as well. - Make sure memset() / memcpy() end up in a text.fn section for garbage collection in SPL. - Update examples/api/Makefile to get memset() again on ARM. - Drop Y-MODEM SPL from am335x_evm and USB SPL so that it fits within size constraints again. - Always set CONFIG_USE_ARCH_MEMSET/MEMCPY on ARM && !ARM64 Cc: Albert Aribaud <albert.u.boot(a)aribaud.net> Signed-off-by: Tom Rini <trini(a)ti.com> --- Aside from needing to split at least the am335x_evm change out (it's fine to do today, and unused in the usecase am335x_evm_usbspl is for), the setting of MEMSET/MEMCPY should be done in Kconfig. This will make opting out easier if we need that long-term. The problems today are: - Lack of arm64 memset/memcpy (Exist in kernel, could be synced over) - A thumb1? problem on platforms such as taurus that I hope would be resolved with a re-sync to the kernel again as there's some minor differences between our implementations. If we're good with the notion of always switching over to the asm versions of these functions (and keeping it selectable or not?) I can address some of these problems and of course other feedback. --- arch/arm/lib/Makefile | 4 ++-- arch/arm/lib/memcpy.S | 2 +- arch/arm/lib/memset.S | 2 +- examples/api/Makefile | 3 ++- include/config_fallbacks.h | 5 +++++ include/configs/am335x_evm.h | 2 ++ 6 files changed, 13 insertions(+), 5 deletions(-) diff --git a/arch/arm/lib/Makefile b/arch/arm/lib/Makefile index d74e4b8..d4b0742 100644 --- a/arch/arm/lib/Makefile +++ b/arch/arm/lib/Makefile @@ -27,11 +27,11 @@ endif obj-$(CONFIG_OF_LIBFDT) += bootm-fdt.o obj-$(CONFIG_CMD_BOOTM) += bootm.o obj-$(CONFIG_SYS_L2_PL310) += cache-pl310.o -obj-$(CONFIG_USE_ARCH_MEMSET) += memset.o -obj-$(CONFIG_USE_ARCH_MEMCPY) += memcpy.o else obj-$(CONFIG_SPL_FRAMEWORK) += spl.o endif +obj-$(CONFIG_USE_ARCH_MEMSET) += memset.o +obj-$(CONFIG_USE_ARCH_MEMCPY) += memcpy.o obj-$(CONFIG_SEMIHOSTING) += semihosting.o obj-y += sections.o diff --git a/arch/arm/lib/memcpy.S b/arch/arm/lib/memcpy.S index eeaf003..ffaee74 100644 --- a/arch/arm/lib/memcpy.S +++ b/arch/arm/lib/memcpy.S @@ -58,7 +58,7 @@ ldmfd sp!, {r0, \reg1, \reg2} .endm - .text + .section .text.memcpy,"ax",%progbits /* Prototype: void *memcpy(void *dest, const void *src, size_t n); */ .syntax unified diff --git a/arch/arm/lib/memset.S b/arch/arm/lib/memset.S index 7208f20..49bdff7 100644 --- a/arch/arm/lib/memset.S +++ b/arch/arm/lib/memset.S @@ -12,7 +12,7 @@ #include <linux/linkage.h> #include <asm/assembler.h> - .text + .section .text.memset,"ax",%progbits .align 5 .syntax unified diff --git a/examples/api/Makefile b/examples/api/Makefile index 6cf23d1..e3c0b6e 100644 --- a/examples/api/Makefile +++ b/examples/api/Makefile @@ -28,6 +28,7 @@ EXT_COBJ-y += lib/string.o EXT_COBJ-y += lib/time.o EXT_COBJ-y += lib/vsprintf.o EXT_SOBJ-$(CONFIG_PPC) += arch/powerpc/lib/ppcstring.o +EXT_SOBJ-$(CONFIG_USE_ARCH_MEMSET) += arch/arm/lib/memset.o # Create a list of object files to be compiled OBJS := $(OBJ-y) $(notdir $(EXT_COBJ-y) $(EXT_SOBJ-y)) @@ -53,5 +54,5 @@ $(addprefix $(obj)/,$(notdir $(EXT_COBJ-y))): $(obj)/%.o: lib/%.c FORCE $(call if_changed_rule,cc_o_c) # Rule to build architecture-specific library assembly files -$(addprefix $(obj)/,$(notdir $(EXT_SOBJ-y))): $(obj)/%.o: arch/powerpc/lib/%.S FORCE +$(addprefix $(obj)/,$(notdir $(EXT_SOBJ-y))): $(obj)/%.o: arch/$(ARCH)/lib/%.S FORCE $(call if_changed_dep,as_o_S) diff --git a/include/config_fallbacks.h b/include/config_fallbacks.h index ddfe045..dd17573 100644 --- a/include/config_fallbacks.h +++ b/include/config_fallbacks.h @@ -97,4 +97,9 @@ # endif #endif +#if defined(CONFIG_ARM) && !defined(CONFIG_ARM64) +#define CONFIG_USE_ARCH_MEMSET +#define CONFIG_USE_ARCH_MEMCPY +#endif + #endif /* __CONFIG_FALLBACKS_H */ diff --git a/include/configs/am335x_evm.h b/include/configs/am335x_evm.h index f1c270c..7a87b46 100644 --- a/include/configs/am335x_evm.h +++ b/include/configs/am335x_evm.h @@ -345,6 +345,8 @@ /* disable EFI partitions and partition UUID support */ #undef CONFIG_PARTITION_UUIDS #undef CONFIG_EFI_PARTITION +/* Reclaim more space */ +#undef CONFIG_SPL_YMODEM_SUPPORT /* General network SPL */ #define CONFIG_SPL_NET_SUPPORT #define CONFIG_SPL_ENV_SUPPORT -- 1.7.9.5

3 4

[U-Boot] [PATCH] moveconfig: Add a new --git-ref option
by Joe Hershberger 13 Jun '16

13 Jun '16

This option allows the 'make *_defconfig' step to run against a former repo state, while the savedefconfig step runs against the current repo state. This is convenient for the case where something in the Kconfig has changed such that the defconfig is no longer complete with the new Kconfigs. This feature allows the .config to be built assuming those old Kconfigs, but then savedefconfig based on the new state of the Kconfigs. Signed-off-by: Joe Hershberger <joe.hershberger(a)ni.com> --- tools/moveconfig.py | 60 +++++++++++++++++++++++++++++++++++++++++++++---- tools/patman/gitutil.py | 15 +++++++++++++ 2 files changed, 71 insertions(+), 4 deletions(-) diff --git a/tools/moveconfig.py b/tools/moveconfig.py index 496c90a..eeb9c0e 100755 --- a/tools/moveconfig.py +++ b/tools/moveconfig.py @@ -153,6 +153,10 @@ Available options Specify the number of threads to run simultaneously. If not specified, the number of threads is the same as the number of CPU cores. + -r, --git-ref + Specify the git ref to clone for the make *_defconfig step. If unspecified + use the CWD. + -v, --verbose Show any build errors as boards are built @@ -173,6 +177,12 @@ import sys import tempfile import time +# Bring in the patman libraries +our_path = os.path.dirname(os.path.realpath(__file__)) +sys.path.append(os.path.join(our_path, 'patman')) + +import gitutil + SHOW_GNU_MAKE = 'scripts/show-gnu-make' SLEEP_TIME=0.03 @@ -526,7 +536,7 @@ class Slot: for faster processing. """ - def __init__(self, config_attrs, options, devnull, make_cmd): + def __init__(self, config_attrs, options, devnull, make_cmd, defconfig_src_dir): """Create a new process slot. Arguments: @@ -540,6 +550,7 @@ class Slot: self.build_dir = tempfile.mkdtemp() self.devnull = devnull self.make_cmd = (make_cmd, 'O=' + self.build_dir) + self.defconfig_src_dir = defconfig_src_dir self.parser = KconfigParser(config_attrs, options, self.build_dir) self.state = STATE_IDLE self.failed_boards = [] @@ -576,6 +587,9 @@ class Slot: return False cmd = list(self.make_cmd) cmd.append(defconfig) + if self.options.git_ref: + cmd.append('-C') + cmd.append(self.defconfig_src_dir) self.ps = subprocess.Popen(cmd, stdout=self.devnull, stderr=subprocess.PIPE) self.defconfig = defconfig @@ -658,6 +672,9 @@ class Slot: cmd.append('include/config/auto.conf') """This will be screen-scraped, so be sure the expected text will be returned consistently on every machine by setting LANG=C""" + if self.options.git_ref: + cmd.append('-C') + cmd.append(self.defconfig_src_dir) self.ps = subprocess.Popen(cmd, stdout=self.devnull, env=dict(os.environ, LANG='C'), stderr=subprocess.PIPE) @@ -673,7 +690,7 @@ class Slots: """Controller of the array of subprocess slots.""" - def __init__(self, config_attrs, options): + def __init__(self, config_attrs, options, defconfig_src_dir): """Create a new slots controller. Arguments: @@ -686,7 +703,8 @@ class Slots: devnull = get_devnull() make_cmd = get_make_cmd() for i in range(options.jobs): - self.slots.append(Slot(config_attrs, options, devnull, make_cmd)) + self.slots.append(Slot(config_attrs, options, devnull, make_cmd, + defconfig_src_dir)) def add(self, defconfig, num, total): """Add a new subprocess if a vacant slot is found. @@ -743,6 +761,24 @@ class Slots: for board in failed_boards: f.write(board + '\n') +class WorkDir: + def __init__(self): + """Create a new working directory.""" + self.work_dir = tempfile.mkdtemp() + + def __del__(self): + """Delete the working directory + + This function makes sure the temporary directory is cleaned away + even if Python suddenly dies due to error. It should be done in here + because it is guaranteed the destructor is always invoked when the + instance of the class gets unreferenced. + """ + shutil.rmtree(self.work_dir) + + def get(self): + return self.work_dir + def move_config(config_attrs, options): """Move config options to defconfig files. @@ -755,6 +791,20 @@ def move_config(config_attrs, options): print 'Nothing to do. exit.' sys.exit(0) + defconfig_src_dir = '' + + if options.git_ref: + work_dir = WorkDir() + defconfig_src_dir = work_dir.get() + cwd = os.getcwd() + print 'Cloning git repo for \'make *_defconfig\' step...' + gitutil.Clone(cwd, defconfig_src_dir) + print 'Checkout \'%s\' to find original configs.' % \ + gitutil.CommitHash(options.git_ref) + os.chdir(defconfig_src_dir) + gitutil.Checkout(options.git_ref) + os.chdir(cwd) + print 'Move the following CONFIG options (jobs: %d)' % options.jobs for config_attr in config_attrs: print ' %s (type: %s, default: %s)' % (config_attr['config'], @@ -777,7 +827,7 @@ def move_config(config_attrs, options): for filename in fnmatch.filter(filenames, '*_defconfig'): defconfigs.append(os.path.join(dirpath, filename)) - slots = Slots(config_attrs, options) + slots = Slots(config_attrs, options, defconfig_src_dir) # Main loop to process defconfig files: # Add a new subprocess into a vacant slot. @@ -887,6 +937,8 @@ def main(): help='only cleanup the headers') parser.add_option('-j', '--jobs', type='int', default=cpu_count, help='the number of jobs to run simultaneously') + parser.add_option('-r', '--git-ref', type='string', + help='the git ref to clone for the make *_defconfig step') parser.add_option('-v', '--verbose', action='store_true', default=False, help='show any build errors as boards are built') parser.usage += ' recipe_file\n\n' + \ diff --git a/tools/patman/gitutil.py b/tools/patman/gitutil.py index 9e739d8..138f989 100644 --- a/tools/patman/gitutil.py +++ b/tools/patman/gitutil.py @@ -61,6 +61,21 @@ def CountCommitsToBranch(): patch_count = int(stdout) return patch_count +def CommitHash(commit_ref): + """Gets the hash for a commit + + Args: + commit_ref: Commit ref to look up + + Return: + Hash of revision, if any, else None + """ + pipe = ['git', 'rev-parse', '--short', commit_ref] + stdout = command.RunPipe([pipe], capture=True, oneline=True).stdout + + hash = stdout.strip() + return hash + def NameRevision(commit_hash): """Gets the revision name for a commit -- 1.7.11.5

3 18

[U-Boot] [PATCH] net: Add mv88e60xx support to the mv88e61xx driver
by Joshua Scott 26 Apr '16

26 Apr '16

The Marvell Link Street mv88e60xx is a series of FastEthernet switch chips, some of which also support Gigabit ports. It is similar to the mv88e61xx series which support Gigabit on all ports. There is already a driver for the mv88e61xx in U-Boot, and this patch updates it to also work with the mv88e60xx. The main change that was required was the size of the port bitmaps. Previously, 8-bit port bitmaps were in use, as the driver had only been used with six-port mv88e61xx devices. This has been changed so that 16-bit port bitmaps are now used, which covers the largest port-counts available for the mv88e60xx / mv88e61xx. Reviewed-by: Chris Packham <chris.packham(a)alliedtelesis.co.nz> Signed-off-by: Joshua Scott <joshua.scott(a)alliedtelesis.co.nz> --- drivers/net/phy/mv88e61xx.c | 29 +++++++++++++++++++++-------- drivers/net/phy/mv88e61xx.h | 2 -- include/netdev.h | 20 +++++++++++++++----- 3 files changed, 36 insertions(+), 15 deletions(-) diff --git a/drivers/net/phy/mv88e61xx.c b/drivers/net/phy/mv88e61xx.c index 302abe8..79204f3 100644 --- a/drivers/net/phy/mv88e61xx.c +++ b/drivers/net/phy/mv88e61xx.c @@ -215,17 +215,17 @@ static void mv88e61xx_port_vlan_config(struct mv88e61xx_config *swconfig) u32 port_mask = swconfig->ports_enabled; /* apply internal vlan config */ - for (prt = 0; prt < MV88E61XX_MAX_PORTS_NUM; prt++) { + for (prt = 0; prt < swconfig->numports; prt++) { /* only for enabled ports */ if ((1 << prt) & port_mask) { /* take vlan map from swconfig */ - u8 vlanmap = swconfig->vlancfg[prt]; + u16 vlanmap = swconfig->vlancfg[prt]; /* remove disabled ports from vlan map */ vlanmap &= swconfig->ports_enabled; /* apply vlan map to port */ RD_SWITCH_PORT_REG(name, prt, MV88E61XX_PRT_VMAP_REG, &reg); - reg &= ~((1 << MV88E61XX_MAX_PORTS_NUM) - 1); + reg &= ~((1 << swconfig->numports) - 1); reg |= vlanmap; WR_SWITCH_PORT_REG(name, prt, MV88E61XX_PRT_VMAP_REG, reg); @@ -331,32 +331,44 @@ int mv88e61xx_switch_initialize(struct mv88e61xx_config *swconfig) return -1; } - if (!(swconfig->cpuport & ((1 << 4) | (1 << 5)))) { + if (!(swconfig->cpuport & ((1 << 4) | (1 << 5) | (1 << 10)))) { swconfig->cpuport = (1 << 5); printf("Invalid cpu port config, using default port5\n"); } RD_SWITCH_PORT_REG(name, 0, MII_PHYSID2, &reg); switch (reg &= 0xfff0) { + case 0x0980: + idstr = "88E6096"; + swconfig->numports = 0x0b; + break; + case 0x0990: + idstr = "88E6097"; + swconfig->numports = 0x0b; + break; case 0x1610: idstr = "88E6161"; + swconfig->numports = 0x06; break; case 0x1650: idstr = "88E6165"; + swconfig->numports = 0x06; break; case 0x1210: idstr = "88E6123"; + swconfig->numports = 0x06; /* ports 2,3,4 not available */ swconfig->ports_enabled &= 0x023; break; default: /* Could not detect switch id */ idstr = "88E61??"; + swconfig->numports = 0x06; break; } /* be sure all ports are disabled */ - for (prt = 0; prt < MV88E61XX_MAX_PORTS_NUM; prt++) { + for (prt = 0; prt < swconfig->numports; prt++) { RD_SWITCH_PORT_REG(name, prt, MV88E61XX_PRT_CTRL_REG, &reg); reg &= ~0x3; WR_SWITCH_PORT_REG(name, prt, MV88E61XX_PRT_CTRL_REG, reg); @@ -405,12 +417,13 @@ int mv88e61xx_switch_initialize(struct mv88e61xx_config *swconfig) WR_SWITCH_PORT_REG(name, 5, MV88E61XX_PCS_CTRL_REG, 0x3e); } - for (prt = 0; prt < MV88E61XX_MAX_PORTS_NUM; prt++) { + for (prt = 0; prt < swconfig->numports; prt++) { /* configure port's PHY */ if (!((1 << prt) & swconfig->cpuport)) { - /* port 4 has phy 6, not 4 */ - int phy = (prt == 4) ? 6 : prt; + /* 88e61xx: port 4 has phy 6, not 4 */ + int phy = (swconfig->numports == 6 && + prt == 4) ? 6 : prt; if (mv88361xx_powerup(swconfig, phy)) return -1; if (mv88361xx_reverse_mdipn(swconfig, phy)) diff --git a/drivers/net/phy/mv88e61xx.h b/drivers/net/phy/mv88e61xx.h index 9c62e4a..b289f21 100644 --- a/drivers/net/phy/mv88e61xx.h +++ b/drivers/net/phy/mv88e61xx.h @@ -11,8 +11,6 @@ #include <miiphy.h> -#define MV88E61XX_CPU_PORT 0x5 - #define MV88E61XX_PHY_TIMEOUT 100000 /* port dev-addr (= port + 0x10) */ diff --git a/include/netdev.h b/include/netdev.h index 34651ab..cbd0420 100644 --- a/include/netdev.h +++ b/include/netdev.h @@ -154,7 +154,9 @@ static inline int pci_eth_init(bd_t *bis) #if defined(CONFIG_MV88E61XX_SWITCH) /* constants for any 88E61xx switch */ -#define MV88E61XX_MAX_PORTS_NUM 6 +/* Allow enough space in the struct to fit 11 ports, as this is + * the number used in the mv88e60xx */ +#define MV88E61XX_MAX_PORTS_NUM 0x0b enum mv88e61xx_cfg_mdip { MV88E61XX_MDIP_NOCHANGE, @@ -180,13 +182,14 @@ enum mv88e61xx_cfg_prtstt { struct mv88e61xx_config { char *name; - u8 vlancfg[MV88E61XX_MAX_PORTS_NUM]; + u16 vlancfg[MV88E61XX_MAX_PORTS_NUM]; enum mv88e61xx_cfg_rgmiid rgmii_delay; enum mv88e61xx_cfg_prtstt portstate; enum mv88e61xx_cfg_ledinit led_init; enum mv88e61xx_cfg_mdip mdip; + u32 numports; u32 ports_enabled; - u8 cpuport; + u32 cpuport; }; /* @@ -196,10 +199,17 @@ struct mv88e61xx_config { */ /* Switch mode : routes any port to any port */ -#define MV88E61XX_VLANCFG_SWITCH { 0x3F, 0x3F, 0x3F, 0x3F, 0x3F, 0x3F } +#define MV88E61XX_VLANCFG_SWITCH { 0x003F, 0x003F, 0x003F, 0x003F, 0x003F, \ + 0x003F, 0x0000, 0x0000, 0x0000, 0x0000, \ + 0x0000 } /* Router mode: routes only CPU port 5 to/from non-CPU ports 0-4 */ -#define MV88E61XX_VLANCFG_ROUTER { 0x20, 0x20, 0x20, 0x20, 0x20, 0x1F } +#define MV88E61XX_VLANCFG_ROUTER { 0x0020, 0x0020, 0x0020, 0x0020, 0x0020, \ + 0x001F, 0x0000, 0x0000, 0x0000, 0x0000, \ + 0x0000 } +#define MV88E60XX_VLANCFG_ROUTER { 0x0400, 0x0400, 0x0400, 0x0400, 0x0400, \ + 0x0400, 0x0400, 0x0400, 0x0400, 0x0400, \ + 0x03ff } int mv88e61xx_switch_initialize(struct mv88e61xx_config *swconfig); #endif /* CONFIG_MV88E61XX_SWITCH */ -- 1.9.1

2 2

[U-Boot] Fwd: Marvell Armada XP u-boot from uart
by Konstantin Lezhnyev 06 Apr '16

06 Apr '16

---------- Forwarded message ---------- From: Konstantin Lezhnyev <konst.ru(a)gmail.com> Date: 2015-05-26 13:30 GMT+05:00 Subject: Re: [U-Boot] Marvell Armada XP u-boot from uart To: Stefan Roese <sr(a)denx.de> I'm sorry, the wording of my question was too short. I have already fulfilled the steps described in the article. But the article describes loading of the image from manufacturer - it is not my case. 1. I create a configuration for my board from maxbcm config. 2. Set `BOOT_FROM` parameter in board\my\kwbimage.cfg to `uart`. 3. Set DDR timings/settings. 4. Upload and execute image via uart. 5. Memory training ended successfully. But then I get the following messages in the terminal: | >>spl:board_init_r() | boot device - 8 | ... | SPI probe failed. Boot device 8 - constant `BOOT_DEVICE_SPI`. Function code returns this value looks like: | u32 spl_boot_device(void) | { | /* Right now only booting via SPI NOR flash is supported */ | return BOOT_DEVICE_SPI; | } And now I repeat my question: >Is it possible to load and boot Marvell Armada XP u-boot completely from uart without spi or nand flash? Regards, kon5t. 2015-05-23 13:08 GMT+05:00 Stefan Roese <sr(a)denx.de>: > On 22.05.2015 16:12, Peter Robinson wrote: > >> Is it possible to load and boot Marvell Armada XP u-boot completely from >>> uart without spi or nand flash? >>> I have custom board with MV78460 and clean flash memory. >>> >> >> You can use kwboot I believe. >> > > Correct. kwboot can be used to upload and execute a binary (compiled for > UART booting) via UART. Please make sure to use the "-a" parameter for the > Armada XP timings. > > Thanks, > Stefan > >

4 3