Michal Simek [Thu, 17 Feb 2022 13:28:39 +0000 (14:28 +0100)]
arm64: zynqmp: Build psu_spl_init for SPL all the time
ZYNQMP_PSU_INIT_ENABLED specifically saying that has connection to full
U-Boot not SPL that's why build psu_spl_init for SPL all the time.
Also disable ZYNQMP_PSU_INIT_ENABLED because it ends up in situation that
psu_init() is called twice which is wrong. By default only SPL should call
it.
Michal Simek [Thu, 17 Feb 2022 13:28:38 +0000 (14:28 +0100)]
xilinx: Enable OF_BOARD for zynq and zynqmp boards
The commit 985503439762 ("fdt: Don't call board_fdt_blob_setup() without
OF_BOARD") forced to enable OF_BOARD for platforms which provide DT
externally. Zynq/ZynqMP boards are using this feature for a long time
that's why there is a need to enable it by default.
Also code expects to return error in case of error that's why also fill it.
Ovidiu Panait [Sun, 13 Feb 2022 08:09:24 +0000 (10:09 +0200)]
microblaze: exception: fix unaligned data access register mask
The correct mask for getting the source/destination register from ESR in
the case of an unaligned access exception is 0x3E0. With this change, a
dummy unaligned store produces the expected info:
"""
>> swi r5, r0, 0x111
...
Hardware exception at 0x111 address
Unaligned data access exception
Unaligned word access
Unaligned store access
Register R5
Return address from exception 0x7f99dfc
### ERROR ### Please RESET the board ###
"""
Ovidiu Panait [Sun, 13 Feb 2022 08:09:22 +0000 (10:09 +0200)]
microblaze: exception: fix return address for delay slot exceptions
According to the MicroBlaze reference manual (xilinx2021.2/ug984/page-37):
"""
If an exception is caused by an instruction in a delay slot (that is,
ESR[DS]=1), the exception handler should return execution to
the address stored in BTR instead of the normal exception return
address stored in R17.
"""
Adjust the code to print the proper return address for delay slot
exceptions.
The switch statement in _hw_exception_handler() only covers the
rightmost 5 bits that encode the exception cause:
switch (state & 0x1f)
{
...
}
For this reason, the "0x1000" case will never be reached, because the 13th
bit was zeroed out. To fix this, move delay slot exception handling before
the switch statement (delay slot (DS) bit in Exception Status Register is
independent of the exception cause (EC)).
Ovidiu Panait [Sun, 13 Feb 2022 08:09:19 +0000 (10:09 +0200)]
microblaze: exception: move privileged instruction exception out of v5 ifdef
The privileged instruction exception seems to have been introduced in
microblaze v7.00 along with MMU support, so having it wrapped in
MICROBLAZE_v5 ifdefs seems incorrect. Move it out of the ifdef, since all
recent microblaze versions support it.
Michal Simek [Mon, 7 Feb 2022 09:36:33 +0000 (10:36 +0100)]
scsi: ceva: Enable PHY and reset support
Add phy and reset support for ceva sata IP. Phy and reset are optional
properties that's why detect if description is available. If not just
continue with operation.
This code was tested on Xilinx Kria SOM kv260-revA with sata connector
populated.
Michal Simek [Fri, 14 Jan 2022 12:25:38 +0000 (13:25 +0100)]
arm64: zynqmp: Add command for disabling loading other overlays
Add command "zynqmp pmufw node close" to disable permission to load
additional pmufw config overlays. This command will make sure that any
other sw will ask for changing permission.
Michal Simek [Fri, 14 Jan 2022 12:25:35 +0000 (13:25 +0100)]
xilinx: firmware: Introduce zynqmp_pmufw_node() for loading PMU fragments
Introduce zynqmp_pmufw_node() for loading PMU configuration fragment for
enabling IPs. Firmware driver has small overlay where NODE id is added and
config fragment is sent to PMUFW. There is a need to build PMUFW with
fragment support.
Michal Simek [Mon, 7 Feb 2022 09:27:37 +0000 (10:27 +0100)]
power: zynqmp: Add power domain driver for ZynqMP
Driver should be enabled by CONFIG_POWER_DOMAIN=y and
CONFIG_ZYNQMP_POWER_DOMAIN=y. Power domain driver doesn't have own DT node
but it uses zynqmp firmware DT node that's why there is a need to bind
driver when firmware node is found.
Driver itself is simple. It is sending pmufw config object overlay for
enabling access to device which is done in ...domain_request().
In ...domain_on() capabilities are passed and node is requested.
This should be bare minimum of required to get power domain driver working.
Luca Ceresoli [Sat, 12 Feb 2022 12:51:21 +0000 (13:51 +0100)]
tools/zynqmp_pm_cfg_obj_convert.py: fix build with Vivado 2021.x
This tool fails with a pm_cfg_obj.c file generated by Vitis 2021.2. This is
because that version of Vitis added the PM_CONFIG_OBJECT_TYPE_BASE that was
not previously generated, thus the script does not implement it.
Michal Simek [Wed, 19 Jan 2022 11:01:51 +0000 (12:01 +0100)]
arm64: zynqmp: Remove SOM *u-boot.dtsi
Disable mmc from u-boot.dtsi file because it was there only for kv260
board. With kr260 this is not needed because we will switch to full DT per
board with SD/EMMC there too.
T Karthik Reddy [Mon, 31 Jan 2022 05:22:40 +0000 (22:22 -0700)]
spi: zynq_qspi: Add SPI memory operations to zynq qspi
Spi memory operation interface is added to zynq qspi
driver to provide an high-level interface to execute
qspi controller specific memory operations by avoiding
spi_mem_exec_op() from spi-mem framework.
spi: zynq_qspi: Read only one byte at a time from txbuf
Read only one byte at a time from txbuf as txbuf may not be
aligned and accessing more than a byte at a time may cause
alignment issues. This fixes the issue of data abort exception
while writing to flash device.
This patch typecasts and accesses rx buf properly as
an unaligned rxbuf, typecasting with u16 and accessing
it causes data abort exception and this patch fixes it.
We should get the base address through its parent's address-cells and
size-cells settings. So we should use fdtdec_get_addr_size_auto_parent()
to get correct base address.
After applying this patch, we can get the correct base address of dma by
replacing fdtdec_get_addr_size_fixed() with
fdtdec_get_addr_size_auto_parent().
Ricardo Salveti [Thu, 20 Jan 2022 19:17:30 +0000 (16:17 -0300)]
xilinx: common: change bootm_size to not go beyond ram_top
The available ram can be limited by ram_top as that depends on the
reserved memory nodes provided by the device-tree (via
board_get_usable_ram_top), so make sure to respect ram_top when setting
up bootm_size to avoid overlapping reserved memory regions (e.g. memory
used by OP-TEE).
The same logic is available in env_get_bootm_size when bootm_size is
not defined by the default environment.
Tom Rini [Mon, 31 Jan 2022 12:45:13 +0000 (07:45 -0500)]
Merge branch 'master' of https://source.denx.de/u-boot/custodians/u-boot-marvell
- kirkwood: Add Pogoplug-V4 support (Tony)
- kirkwood: GoFlex Home : Use Marvell PHY driver (Tony)
- Another set of kwboot improvements (Pali)
- Minor misc stuff
Tony Dinh [Wed, 26 Jan 2022 03:33:44 +0000 (19:33 -0800)]
arm: kirkwood: GoFlex Home : Use Marvell uclass mvgbe and PHY driver for Ethernet
The GoFlex Home board has the network chip Marvell 88E1116R.
Use uclass mvgbe and the compatible driver M88E1118R to bring up Ethernet.
- Currently, CONFIG_RESET_PHY_R symbol is used in
arch/arm/mach-kirkwood/include/mach/config.h for all Kirkwood
boards with mv8831116 PHY, with each board defines the function
reset_phy(). Undefine it for this board.
- Add board_eth_init() to use uclass mvgbe to bring up the network.
And remove ad-hoc code.
- Enable CONFIG_PHY_MARVELL to properly configure the network.
- Miscellaneous changes: use CONFIG_SYS_THUMB_BUILD to keep u-boot image
under 512K, use BIT macro, and cleanup comments.
Signed-off-by: Tony Dinh <mibodhi@gmail.com> Reviewed-by: Pali Rohár <pali@kernel.org> Reviewed-by: Stefan Roese <sr@denx.de>
Pali Rohár [Tue, 25 Jan 2022 17:13:13 +0000 (18:13 +0100)]
tools: kwboot: Set debug flag to 1
This should enable BootROM output on UART.
(At least on A385 BootROM this is broken, BootROM ignores this debug
flag and does not enable its output on UART if some valid image is
available in SPI-NOR.)
Signed-off-by: Pali Rohár <pali@kernel.org> Signed-off-by: Marek Behún <marek.behun@nic.cz> Reviewed-by: Stefan Roese <sr@denx.de>
Pali Rohár [Tue, 25 Jan 2022 17:13:09 +0000 (18:13 +0100)]
tools: kwboot: Remove 2s delay before sending first xmodem packet
This delay is not needed anymore since kwboot already handles retrying
logic for incomplete xmodem packets and also forces BootROM to flush its
input queue. Removing it decreases total transfer time.
Signed-off-by: Pali Rohár <pali@kernel.org> Signed-off-by: Marek Behún <marek.behun@nic.cz> Reviewed-by: Stefan Roese <sr@denx.de>
Pali Rohár [Tue, 25 Jan 2022 17:13:07 +0000 (18:13 +0100)]
tools: kwboot: Allow to use option -b without image path
Allow option -b without image path parameter, to send boot pattern and
wait for response but not send any image. This allows to use kwboot just
for processing boot pattern and user can use any other xmodem tool for
transferring the image itself (e.g. sx). Useful for debugging purposes.
Signed-off-by: Pali Rohár <pali@kernel.org> Signed-off-by: Marek Behún <marek.behun@nic.cz> Reviewed-by: Stefan Roese <sr@denx.de>
Pali Rohár [Tue, 25 Jan 2022 17:13:06 +0000 (18:13 +0100)]
tools: kwboot: Show 'E' in progress output when error occurs
When kwboot is unable to resend current xmodem packet, show an 'E' in the
progress output instead of a '+'. This allows to distinguish between the
state when kwboot is retrying sending the packet and when retry is not
possible.
Signed-off-by: Pali Rohár <pali@kernel.org> Signed-off-by: Marek Behún <marek.behun@nic.cz> Reviewed-by: Stefan Roese <sr@denx.de>
Pali Rohár [Tue, 25 Jan 2022 17:13:05 +0000 (18:13 +0100)]
tools: kwboot: Fix handling of repeated xmodem packets
Unfortunately during some stages of xmodem transfer, A385 BootROM is not
able to handle repeated xmodem packets. So if an error occurs during that
stage, stop the transfer and return failure.
Signed-off-by: Pali Rohár <pali@kernel.org> Signed-off-by: Marek Behún <marek.behun@nic.cz> Reviewed-by: Stefan Roese <sr@denx.de>
Pali Rohár [Tue, 25 Jan 2022 17:13:04 +0000 (18:13 +0100)]
tools: kwboot: Do not change received character in kwboot_xm_recv_reply()
Marvell BootROM expects retransmission of previous xmodem packet only in
the case when it sends NAK response to the host.
Do not change non-xmodem response (possibly UART transfer error) to NAK
in kwboot_xm_recv_reply() function. Allow caller to receive original
response from device.
Change argument 'nak_on_non_xm' to 'stop_on_non_xm'. Instead of changing
non-xmodem character to NAK, stop processing on invalid character and
return it.
Signed-off-by: Pali Rohár <pali@kernel.org> Signed-off-by: Marek Behún <marek.behun@nic.cz> Reviewed-by: Stefan Roese <sr@denx.de>
Pali Rohár [Tue, 25 Jan 2022 17:13:03 +0000 (18:13 +0100)]
tools: kwboot: Remove code for handling CAN byte
It is unknown why handling of CAN byte was added into kwboot tool as
Marvell BootROM does not support CAN byte. It never sends CAN byte to host
and if host sends CAN byte BootROM handles it as an unknown byte.
Remove code for handling and sending CAN bytes from the kwboot tool.
Signed-off-by: Pali Rohár <pali@kernel.org> Signed-off-by: Marek Behún <marek.behun@nic.cz> Reviewed-by: Stefan Roese <sr@denx.de>
Pali Rohár [Tue, 25 Jan 2022 17:13:02 +0000 (18:13 +0100)]
tools: kwboot: Improve retrying logic for incomplete xmodem packets
Sometimes if the first byte of xmodem packet (SOH) is incorrectly
transmitted, BootROM sends NAK for every non-SOH received byte, which
makes BootROM and the host kwboot tool out of sync. BootROM automatically
re-synchronizes after 2s pause by dropping its input queue. So when
attempting retransmit for 9th time or later, ignore NAK reply from BootROM
and either wait for valid ACK or let kwboot timeout, which implies
re-synchronization.
This fixes retransmission of xmodem packets and allows kwboot to work also
without "Waiting ... and flushing tty" code which is at the beginning of
kwboot xmodem transfer.
Signed-off-by: Pali Rohár <pali@kernel.org> Signed-off-by: Marek Behún <marek.behun@nic.cz> Reviewed-by: Stefan Roese <sr@denx.de>
Tony Dinh [Tue, 25 Jan 2022 10:00:48 +0000 (02:00 -0800)]
arm: kirkwood: Pogoplug V4 : Use Marvell network PHY driver
The Pogoplug V4 board has the network chip Marvell 88E1116R. So to
properly configure the network, enable CONFIG_PHY_MARVELL to activate
the compatible driver M88E1118R.
- This patch depends on the series:
https://patchwork.ozlabs.org/project/uboot/patch/20220124061712.28316-2-mibodhi@gmail.com/
Jan Kiszka [Fri, 28 Jan 2022 19:37:53 +0000 (20:37 +0100)]
binman: Skip node generation for images read from files
We can and should run the node generator only when creating a new image.
When we read it back, there is no need to generate nodes - they already
exits, and binman does not dive that deep into the image - and there is
no way to provide the required fdt-list. So store the mode in the image
object so that Entry_fit can simply skip generator nodes when reading
them from an fdtmap.
This unbreaks all read-backs of images that contain generator nodes in
their fdtmap. To confirm this, add a corresponding test case.
Signed-off-by: Jan Kiszka <jan.kiszka@siemens.com> Reviewed-by: Simon Glass <sjg@chromium.org> Tested-by: Simon Glass <sjg@chromium.org>
Add SPDX to dts file: Signed-off-by: Simon Glass <sjg@chromium.org>
moveconfig: Fix code relying on now-stripped newline characters
Commit 37f815cad07d ("moveconfig: Use a function to read files") adds a
helper function that can read a file as lines, but strips the newline
characters. This change broke parts of moveconfig code that relied on
their existence, resulting in a few issues:
Configs that are defined as empty aren't removed from header files (e.g.
"#define CONFIG_REMAKE_ELF"). Make regex patterns use '\b' to match word
boundaries instead of '\W' (which matched the newlines) so these lines
still match and get removed.
All changes in defconfig are considered removed by savedefconfig even
if they weren't, and line continuations in the headers aren't recognized
and removed properly, because their checks explicitly look for a newline
character. Remove the character from both comparisons.
The printed diff of header files is wrongly formatted and raises an
IndexError if a blank line was removed. Let print() print the new lines,
and use size-independent ways to check strings to fix the diff output.
Signed-off-by: Alper Nebi Yasak <alpernebiyasak@gmail.com> Reviewed-by: Simon Glass <sjg@chromium.org> Tested-by: Simon Glass <sjg@chromium.org>
Tom Rini [Sun, 30 Jan 2022 22:12:34 +0000 (17:12 -0500)]
Merge branch 'master' of https://source.denx.de/u-boot/custodians/u-boot-sunxi
a bit delayed, the first batch of the sunxi pull request for this cycle.
This is mostly collecting some patches that were lying around for a
while, plus some recent fixes. Nothing too exciting at this point, but
of course they should be merged nevertheless.
There is the much bigger F1C100s SoC support coming up, which I hope to
be able to send in the next few days, along with the removal of sunxi's
lowlevel_init usage.
Compile tested for all 159 sunxi boards, plus briefly tested on BananaPi
M1, OrangePi Zero, Pine64 and Pine-H64.
Samuel Holland [Wed, 13 Oct 2021 00:21:27 +0000 (19:21 -0500)]
mkimage: sunxi_egon: Allow overriding the padding size
Due to a bug in the H3 SoC, where the CPU 0 hotplug flag cannot be
written, resuming CPU 0 requires using the "Super Standby" code path in
the BROM instead of the hotplug path. This path requires jumping to an
eGON image in SRAM.
This resume image, whose single purpose is to jump back to the secure
monitor, only needs to contain a single instruction. Padding the image
to 8 KiB would be wasteful of SRAM. Hook up the -B (block size) option
so users can set the block/padding size.
Signed-off-by: Samuel Holland <samuel@sholland.org> Reviewed-by: Andre Przywara <andre.przywara@arm.com> Signed-off-by: Andre Przywara <andre.przywara@arm.com>
Jernej Skrabec [Sat, 29 Jan 2022 15:58:43 +0000 (16:58 +0100)]
sunxi: Fix H616 DRAM read calibration for dual rank
Although it isn't known what bit 0 in PHY reg 8 does, it's obvious that
it has to be set before read calibration and cleared afterwards. This is
already done for first rank, but not for second (copy & paste error.)
Fix it.
Fixes: f4317dbd06b6 ("sunxi: Add H616 DRAM support") Signed-off-by: Jernej Skrabec <jernej.skrabec@gmail.com> Reviewed-by: Andre Przywara <andre.przywara@arm.com> Signed-off-by: Andre Przywara <andre.przywara@arm.com>
Chris Morgan [Fri, 21 Jan 2022 13:37:32 +0000 (13:37 +0000)]
sunxi: Add option to prevent booting on power plug-in
For sunxi boards with the AXP209, AXP221, AXP809, and AXP818 PMICs
(plus possibly others, I only confirmed the datasheets for these),
it is sometimes desirable to not boot whenever the device is
plugged in. An example would be when using the NTC CHIP inside a
PocketCHIP.
This provides a configurable option to check if bit 0 of
register 0 of the PMIC says it was powered because of a power button
press (0) or a plug-in event (1). If the value is 1 and this option
is selected, the device shuts down shortly after printing a message
to console stating the reason why it's shutting down. Powering up the
board with the power button is not affected.
Signed-off-by: Chris Morgan <macromorgan@hotmail.com>
[Andre: reword to speak of boot, remove #ifdefs] Signed-off-by: Andre Przywara <andre.przywara@arm.com>
Andre Przywara [Fri, 21 Jan 2022 13:37:31 +0000 (13:37 +0000)]
pmic: axp: define ALDO_IN startup bit
Most AXP PMICs feature a "startup source" register, which keeps
information about how the PMIC started operation. Bit 0 in there means
it has been started by "plugging in the power cable".
Define a symbol in each PMIC's header file to be able to use that
register and bit later on.
Signed-off-by: Andre Przywara <andre.przywara@arm.com>
Samuel Holland [Thu, 21 Oct 2021 04:52:57 +0000 (23:52 -0500)]
mmc: sunxi: Use DM_GPIO flags to set pull-up
Now that the sunxi_gpio driver handles pull-up/down via the driver
model, pin configuration does not need a platform-specific function.
Signed-off-by: Samuel Holland <samuel@sholland.org> Tested-by: Heinrich Schuchardt <heinrich.schuchardt@canonical.com> Signed-off-by: Andre Przywara <andre.przywara@arm.com>
Samuel Holland [Thu, 21 Oct 2021 04:52:56 +0000 (23:52 -0500)]
gpio: sunxi: Implement .set_flags
This, along with gpio_flags_xlate(), allows the GPIO driver to handle
pull-up/down flags provided by consumer drivers or in the device tree.
Signed-off-by: Samuel Holland <samuel@sholland.org> Reviewed-by: Simon Glass <sjg@chromium.org> Reviewed-by: Andre Przywara <andre.przywara@arm.com> Signed-off-by: Andre Przywara <andre.przywara@arm.com>
Samuel Holland [Thu, 21 Oct 2021 04:52:55 +0000 (23:52 -0500)]
sunxi: gpio: Add per-bank drive and pull setters
The GPIO and pinctrl drivers need these setters for pin configuration.
Since they are DM drivers, they should not be using hardcoded base
addresses. Factor out variants of the setter functions which take a
pointer to the GPIO bank's MMIO registers.
Signed-off-by: Samuel Holland <samuel@sholland.org> Signed-off-by: Andre Przywara <andre.przywara@arm.com>
Samuel Holland [Thu, 21 Oct 2021 04:52:54 +0000 (23:52 -0500)]
sunxi: gpio: Return void from setter functions
The return values of these functions are always zero, and they are
never checked. Since they are not needed, remove them.
Signed-off-by: Samuel Holland <samuel@sholland.org> Reviewed-by: Heinrich Schuchardt <heinrich.schuchardt@canonical.com> Signed-off-by: Andre Przywara <andre.przywara@arm.com>
Chris Morgan [Fri, 7 Jan 2022 17:52:54 +0000 (11:52 -0600)]
i2c: mvtwsi: Add compatible string for allwinner, sun4i-a10-i2c
This adds a compatible string for the Allwinner Sun4i-A10 I2C
controller. Without this, boards based on the R8 and A13 (at a
minimum) fail to boot.
Signed-off-by: Chris Morgan <macromorgan@hotmail.com> Acked-by: Akash Gajjar <gajjar04akash@gmail.com> Signed-off-by: Andre Przywara <andre.przywara@arm.com>
sunxi: call fdt_fixup_ethernet again to set macaddr for more aliases
Sometimes some ethernet aliases do not exist in U-Boot DT but they
exist in the DT used to boot the system (for example, modified via DT
overlays). In this situation setup_environment is called again in
ft_board_setup() to generate macaddr environment variable for them.
However now the call to fdt_fixup_ethernet() is moved before the call
of ft_board_setup().
Call fdt_fixup_ethernet() again to add MAC addresses for the extra
ethernet aliases.
The driver for SPI0 on Allwinner H6 SoCs did not use the correct define
SUN50I_GPC_SPI0 for the pin function, but one for a different Allwinner
SoC series.
Fix the conditionals to use the correct define for H6 SoCs. This matches
the conditional logic in the SPL spi driver.
Tested by probing the spi-flash on a pine64_h64-model-b board with
adapted device-tree (disable mmc2, enable spi0).
Signed-off-by: Daniel Wagenknecht <dwagenk@mailbox.org> Reviewed-by: Andre Przywara <andre.przywara@arm.com> Signed-off-by: Andre Przywara <andre.przywara@arm.com>
Tom Rini [Sat, 29 Jan 2022 18:42:58 +0000 (13:42 -0500)]
Merge branch '2022-01-28-assorted-fixes'
- Extend the pci command to support a few more features.
- Add support for custom SPL boot device names (so it's easier for users
to understand)
- Updates for am64x to address some review comments.
- Migration deadline notice for DM_SERIAL
- coreboot payload test
- Support rsa3072 signatures.
- DFU should skip writing empty UBI pages, bootcount printf format char
correction.
Ilias Apalodimas [Fri, 28 Jan 2022 22:20:32 +0000 (00:20 +0200)]
efi_loader: hash the image once before checking against db/dbx
We don't have to recalculate the image hash every time we check against a
new db/dbx entry. So let's add a flag forcing it to run once since we only
support sha256 hashes
Suggested-by: Heinrich Schuchardt <heinrich.schuchardt@canonical.com> Signed-off-by: Ilias Apalodimas <ilias.apalodimas@linaro.org> Reviewed-by: Heinrich Schuchardt <heinrich.schuchardt@canonical.com>
Ilias Apalodimas [Fri, 28 Jan 2022 22:20:31 +0000 (00:20 +0200)]
efi_loader: correctly handle mixed hashes and signatures in db
A mix of signatures and hashes in db doesn't always work as intended.
Currently if the digest algorithm is not explicitly set to sha256 we
stop walking the security database and reject the image.
That's problematic in case we find and try to check a signature before
inspecting the sha256 hash. If the image is unsigned we will reject it
even if the digest matches.
Since we no longer reject the image on unknown algorithms add an explicit
check and reject the image if any other hash algorithm apart from sha256
is detected on dbx.
Suggested-by: Heinrich Schuchardt <heinrich.schuchardt@canonical.com> Signed-off-by: Ilias Apalodimas <ilias.apalodimas@linaro.org> Reviewed-by: Heinrich Schuchardt <heinrich.schuchardt@canonical.com>
Pali Rohár [Mon, 24 Jan 2022 19:45:53 +0000 (20:45 +0100)]
Nokia RX-51: Update documentation about flashing
This change contains update for doc/README.nokia_rx51 documentation file
with information how to load U-Boot image to device RAM without need to
flash it and also how to flash it into OneNAND via 0xFFFF flasher.
Simon Glass [Fri, 21 Jan 2022 17:23:01 +0000 (10:23 -0700)]
gitlab/azure: x86: Add a coreboot test
Coreboot supports U-Boot as a payload and this recently got a bit of a
facelist. Add a test for this.
For now this uses a binary build of coreboot (v4.15). Future work could
potentially build it from source, but we need to figure out the
toolchain problems first, since coreboot uses its own toolchain. It
turns out that this is tricky, because coreboot fails to build with a
vanilla gcc.
This needs some changes to the hooks scripts as well. An example build
is at https://source.denx.de/u-boot/custodians/u-boot-dm/-/jobs/359687
Signed-off-by: Simon Glass <sjg@chromium.org> Reviewed-by: Tom Rini <trini@konsulko.com>
ROM supports cpsw_port2 for Ethernet boot and SPL stages continue to
download images on the same port, therefore there is no need to enable
cpsw_port1. Disable the same.
Patrick Delaunay [Tue, 18 Jan 2022 09:26:21 +0000 (10:26 +0100)]
dfu: mtd: skip empty pages when writing page for UBI partition
Align the DFU MTD backend for the UBI partitions with the mtd command write
behavior when the option .dontskipff is not used: don't write the empty
pages (full of 0xFF); it is not required for UBI, see [1] for details.
This patch avoids the "free space fixup" procedure in the kernel [2]
and allows to program a UBIFS volume generated by mkfs.ubifs without the
option -F, --space-fixup.
The MTD DFU backend implements this behavior introduced on DFU NAND
backend by the commit 13cb7cc9e8e4 ("dfu: Add option to skip empty pages
when flashing UBI images to NAND") and also supported by the command nand
by CONFIG_CMD_NAND_TRIMFFS and by commit c9494866df83 ("cmd_nand: add nand
write.trimffs command").
Pali Rohár [Mon, 17 Jan 2022 15:38:38 +0000 (16:38 +0100)]
pci: Extend 'pci regions' command with bus number
'pci regions' currently prints only region information from bus 0 which
belongs to controller 0. Parser for 'pci regions' cmdline currently ignores
any additional arguments and so U-Boot always uses bus 0.
Regions are stored in controller (not on the bus) and therefore to retrieve
controller from the bus, it is needed to call pci_get_controller() which
returns root bus. Because bus 0 is root bus, current code worked fine for
controller 0.
Extend cmdline parser for 'pci regions' to allows specifying bus number,
extend pci_show_regions() code to accept also non-zero bus number and
print bus ranges for which is regions configuration assigned.
Signed-off-by: Pali Rohár <pali@kernel.org> Reviewed-by: Stefan Roese <sr@denx.de>
Heiko Thiery [Mon, 17 Jan 2022 15:25:41 +0000 (16:25 +0100)]
spl: add support for custom boot method names
Currently the names MMC1, MMC2 and MMC2_2 are output in the SPL. To
achieve more userbility here the name of the boot source can be returned.
E.g. for "MMC1" -> "eMMC" or "MMC2" -> "SD card".
Signed-off-by: Heiko Thiery <heiko.thiery@gmail.com> Reviewed-by: Michael Walle <michael@walle.cc> Tested-by: Michael Walle <michael@walle.cc>
fastboot: only look up real partition names when no alias exists
Having U-Boot look up the passed partition name even though an alias
exists is unexpected, leading to warning messages (when the alias name
doesn't exist as a real partition name) or the use of the wrong
partition.
Change part_get_info_by_name_or_alias() to consider real partitions
names only if no alias of the same name exists, allowing to use aliases
to override the configuration for existing partition names.
Also change one use of strcpy() to strlcpy().
Signed-off-by: Matthias Schiffer <matthias.schiffer@ew.tq-group.com> Reviewed-by: Sean Anderson <sean.anderson@seco.com>
Patrick Delaunay [Wed, 13 Oct 2021 15:01:37 +0000 (17:01 +0200)]
dfu: handle short frame result of UPLOAD in state_dfu_idle
In DFU v1.1 specification [1] the DFU_UPLOAD (Short Frame)
is handled only in dfuUPLOADIDLE state:
- Figure A.1 Interface state transition diagram
- the state description in chapter A.2
A.2.3 State 2 dfuIDLE
on Receipt of the DFU_UPLOAD request,and bitCanUpload = 1
the Next State is dfuUPLOADIDLE
A.2.10 State 9 dfuUPLOAD-IDLE
When the length of the data transferred by the device in response
to a DFU_UPLOAD request is less than wLength. (Short frame)
the Next State is dfuIDLE
In current code, when an UPLOAD is completely performed after the first
request (for example with wLength=200 and data read = 9), the DFU state
stay at dfuUPLOADIDLE until receiving a DFU_UPLOAD or a DFU_ABORT request
even it is unnecessary as the previous DFU_UPLOAD request already reached
the EOF.
This patch proposes to finish the DFU uploading (don't go to dfuUPLOADIDLE)
and completes the control-read operation (go to DFU_STATE_dfuIDLE) when
the first UPLOAD response has a short frame as an end of file (EOF)
indicator even if it is not explicitly allowed in the DFU specification
but this seems logical.
projects / u-boot.git / log
