Retrieve current terminal settings via tcgetattr(), set to raw mode with
cfmakeraw(), enable receiver via CREAD and ignore modem control lines
via CLOCAL.
Signed-off-by: Pali Rohár <pali@kernel.org> Signed-off-by: Marek Behún <marek.behun@nic.cz> Reviewed-by: Stefan Roese <sr@denx.de>
Marek Behún [Fri, 24 Sep 2021 21:07:07 +0000 (23:07 +0200)]
tools: kwboot: Check whether baudrate was set to requested value
The tcsetattr() function can return 0 even if baudrate was not changed.
Check whether baudrate was changed to requested value, and in case of
arbitrary baudrate, check whether the set value is within 3% tolerance.
Signed-off-by: Marek Behún <marek.behun@nic.cz> Reviewed-by: Stefan Roese <sr@denx.de>
The A38x platform supports more baudrates than just those defined by the
Bn constants, and some of them are higher than the highest Bn baudrate
(the highest is 4 MBd while A38x support 5.15 MBd).
On Linux, add support for arbitrary baudrates. (Since there is no
standard POSIX API to specify arbitrary baudrate for a tty device, this
change is Linux-specific.)
We need to use raw TCGETS2/TCSETS2 or TCGETS/TCSETS ioctls with the
BOTHER flag in struct termios2/termios, defined in Linux headers
<asm/ioctls.h> (included by <sys/ioctl.h>) and <asm/termbits.h>. Since
these headers conflict with glibc's header file <termios.h>, it is not
possible to use libc's termios functions and we need to reimplement them
via ioctl() calls.
Note that the Bnnn constants from <termios.h> need not be compatible
with Bnnn constants from <asm/termbits.h>.
Signed-off-by: Pali Rohár <pali@kernel.org>
[ termios macros rewritten to static inline functions (for type control)
and moved to tools/termios_linux.h ] Signed-off-by: Marek Behún <marek.behun@nic.cz> Reviewed-by: Stefan Roese <sr@denx.de>
tools: kwboot: Support higher baudrates when booting via UART
Add support for uploading the boot image (the data part only) at higher
baudrate than the standard one.
The kwboot utility already has -B option, but choosing other baudrate
than the standard one (115200 Bd) can only work for debug mode, not for
booting the device. The BootROM for kwboot supported platforms (Orion,
Kirkwood, Dove, Discovery, AXP, A37x, A38x, A39x) cannot change the
baudrate when uploading boot image via the Xmodem protocol, nor can it
be configured via strapping pins.
So instead we add this support by injecting baudrate changing code into
the kwbimage v1 header as a new optional binary extension. This code is
executed by BootROM after it receives the whole header. The code sends
the magic string "$baudratechange\0" just before changing the baudrate
to let kwboot know that it should also change it. This is because the
injected code is run as the last binary extension, and we do not want
to loose possible output from other possible binary extensions that
came before it (in most cases this is U-Boot SPL).
We also inject the code before the payload (the data part of the image),
to change the baudrate back to the standard value, in case the payload
does not reset UART.
This change improves boot time via UART significantly (depending on the
chosen baudrate), which is very useful when debugging.
Signed-off-by: Pali Rohár <pali@kernel.org>
[ major refactor ] Signed-off-by: Marek Behún <marek.behun@nic.cz> Reviewed-by: Stefan Roese <sr@denx.de>
Marek Behún [Fri, 24 Sep 2021 21:07:04 +0000 (23:07 +0200)]
tools: kwboot: Explicitly check against size of struct main_hdr_v1
Explicitly check the image size against size of struct main_hdr_v1.
This way the check is more readable, since the `hdrsz` variable
may semantically contain another value.
Signed-off-by: Marek Behún <marek.behun@nic.cz> Reviewed-by: Stefan Roese <sr@denx.de>
tools: kwboot: Round up header size to 128 B when patching
The beginning of image data must be sent in a separate xmodem block;
the block must not contain end of header with the beginning of data.
Therefore we need to ensure that the image header size is a multiple of
xmodem block size (which is 128 B).
Read the file into a malloc()ed buffer of enough size instead of
mmap()ing it. (If we are going to move the data, most of the pages will
be dirty anyway.) Then move the payload if header size needs to be
increased.
Signed-off-by: Pali Rohár <pali@kernel.org>
[ refactored ] Signed-off-by: Marek Behún <marek.behun@nic.cz> Reviewed-by: Stefan Roese <sr@denx.de>
tools: kwboot: Patch destination address to DDR area for SPI image
SPI/NOR kwbimage may have destination address set to 0xFFFFFFFF, which
means that the image is not downloaded to DDR but rather it is executed
directly from SPI/NOR. In this case execution address is set to SPI/NOR
area.
When patching image to UART type, change destination and execution
addresses from SPI/NOR XIP area to DDR area 0x00800000 (which is default
for A38x).
Signed-off-by: Pali Rohár <pali@kernel.org>
[ refactored ] Signed-off-by: Marek Behún <marek.behun@nic.cz> Reviewed-by: Stefan Roese <sr@denx.de>
Marek Behún [Tue, 17 Aug 2021 22:59:15 +0000 (00:59 +0200)]
tools: kwbimage: Simplify iteration over version 1 optional headers
Create macro
for_each_opt_hdr_v1
and functions
opt_hdr_v1_size(),
opt_hdr_v1_valid_size(),
opt_hdr_v1_ext(),
opt_hdr_v1_first() and
opt_hdr_v1_next()
to simplify iteration over version 1 optional headers.
This prevents ugly code repetition and makes it nicer to read.
The kwboot_img_patch_hdr() function already decides if header patching
is needed. Always call this function and deprecate the unneeded command
line option `-p`.
Signed-off-by: Pali Rohár <pali@kernel.org> Reviewed-by: Marek Behún <marek.behun@nic.cz> Reviewed-by: Stefan Roese <sr@denx.de>
Marek Behún [Fri, 24 Sep 2021 21:06:53 +0000 (23:06 +0200)]
tools: kwboot: Prevent waiting indefinitely if no xmodem reply is received
Currently if BootROM fails to respond with ACK/NAK to a xmodem block, we
will be waiting indefinitely for such response.
Make sure that we only wait at most 1 second (blk_rsp_timeo) for ACK/NAK
for each block in case non-xmodem text output is not being expected.
Interpret this timeout expiration as NAK, to try to send the block
again.
On the other hand, if timeout expires without ACK while some non-xmodem
output was already received (DDR training output, for example), we know
that the block was received, since the code is being executed, so in
this case exit with ETIMEDOUT.
Signed-off-by: Marek Behún <marek.behun@nic.cz> Reviewed-by: Stefan Roese <sr@denx.de>
Marek Behún [Fri, 24 Sep 2021 21:06:52 +0000 (23:06 +0200)]
tools: kwboot: Allow greater timeout when executing header code
When executing header code (which contains U-Boot SPL in most cases),
wait 10s after every non-xmodem character received (i.e. printed by
U-Boot SPL) before timing out.
Sometimes DDR training, which runs in SPL, may be slow.
Signed-off-by: Marek Behún <marek.behun@nic.cz> Reviewed-by: Stefan Roese <sr@denx.de>
tools: kwboot: Allow non-xmodem text output from BootROM only in a specific case
When sending image header / image data, BootROM does not send any
non-xmodem text output. We should therefore interpret unknown bytes in
the xmodem protocol as errors and resend current packet. This should
improve the transfer in case there are errors on the UART line.
Text output from BootROM may only happen after whole image header is
sent and before ACK for the last packet of image header is received.
In this case BootROM may execute code from the image, which may interact
with UART (U-Boot SPL, for example, prints stuff on UART).
Print received non-xmodem output from BootROM only in this case.
Signed-off-by: Pali Rohár <pali@kernel.org>
[ refactored & simplified ] Signed-off-by: Marek Behún <marek.behun@nic.cz> Reviewed-by: Stefan Roese <sr@denx.de>
tools: kwboot: Print newline on error when progress was not completed
When progress was not completed, current terminal position is in progress
bar. So print newline before printing error message to make error message
more readable.
Signed-off-by: Pali Rohár <pali@kernel.org> Reviewed-by: Marek Behún <marek.behun@nic.cz> Reviewed-by: Stefan Roese <sr@denx.de>
Ensure that `pos` is still in range up to the `width` so printing 100%
works also for bigger images. After printing 100% progress reset it to
zero, so that next progressbar can be started.
Signed-off-by: Pali Rohár <pali@kernel.org> Reviewed-by: Marek Behún <marek.behun@nic.cz> Reviewed-by: Stefan Roese <sr@denx.de>
tools: kwboot: Fix return type of kwboot_xm_makeblock() function
Function kwboot_xm_makeblock() always returns length of xmodem block. It
is always non-negative and calculated from variable with size_t type. Set
return type of this function to size_t and remove dead code which checks
for negative value.
Signed-off-by: Pali Rohár <pali@kernel.org> Reviewed-by: Marek Behún <marek.behun@nic.cz> Reviewed-by: Stefan Roese <sr@denx.de>
tools: kwboot: Fix kwboot_xm_sendblock() function when kwboot_tty_recv() fails
When kwboot_tty_recv() fails or times out, it does not set the `c`
variable to NAK. The variable is then compared, while it holds either
an undefined value or a value from previous iteration. Set `c` to NAK so
that the other side will try to resend current block, and remove the
now unnecessary break.
In other failure cases return immediately.
Signed-off-by: Pali Rohár <pali@kernel.org> Reviewed-by: Marek Behún <marek.behun@nic.cz> Reviewed-by: Stefan Roese <sr@denx.de>
Marek Behún [Fri, 24 Sep 2021 21:06:41 +0000 (23:06 +0200)]
tools: kwboot: Refactor and fix writing buffer
There are 3 instances in kwboot.c where we need to write() a given
buffer whole (iteratively writing until all data are written), and 2 of
those instances are wrong, for they do not increment the buffer pointer.
Refactor the code into a new function kwboot_write() where it is fixed.
Signed-off-by: Marek Behún <marek.behun@nic.cz> Reviewed-by: Stefan Roese <sr@denx.de>
Marek Behún [Fri, 24 Sep 2021 21:06:39 +0000 (23:06 +0200)]
tools: kwboot: Fix buffer overflow in kwboot_terminal()
The `in` variable is set to -1 in kwboot_terminal() if stdin is not a
tty. In this case we should not look whether -1 is set in fd_set, for it
can lead to a buffer overflow, which can be reproduced with
echo "xyz" | ./tools/kwboot -t /dev/ttyUSB0
Signed-off-by: Marek Behún <marek.behun@nic.cz> Reviewed-by: Stefan Roese <sr@denx.de>
Marek Behún [Fri, 24 Sep 2021 21:06:38 +0000 (23:06 +0200)]
tools: kwbimage: Fix printf format warning
On 32-bit ARM the compiler complains:
tools/kwbimage.c:547: warning: format ‘%lu’ expects argument of type
‘long unsigned int’, but argument 4 has
type ‘unsigned int’
Fix this by using %zu instead of %lu format specifier.
Signed-off-by: Marek Behún <marek.behun@nic.cz> Reviewed-by: Stefan Roese <sr@denx.de>
Wdt expire command makes the wdt to count least possible ticks(1)
and expires immediately. Add expire_now option to the xlnx_wwdt_ops
and implement it by calling xlnx_wwdt_start() with minimum possible
count(1).
watchdog: versal: Add support for basic window watchdog
Existing driver uses generic watchdog mode which generates a signal to
PLM firmware, but the signal cannot be used to reset the system.
Change driver to use window watchdog basic mode. This window watchdog mode
generates a signal to PLM firmware which decides what action to take upon
expiry of watchdog.
Timeout value for xlnx_wwdt_start will come in milli seconds from wdt
framework. Make changes to load count value accordingly.
Add checks before loading the timer for min and max possible values.
Fix authour email id of Ashok Reddy Soma to long email id.
Michal Simek [Fri, 27 Aug 2021 10:53:32 +0000 (12:53 +0200)]
ARM: zynq: Enable capsule update for qspi and mmc
Generate dfu_alt_info setup at runtime for capsule update.
Enabling this feature will help with upgrading boards without remembering
what is where.
The similar change was done for ZynqMP by commit b86f43de0be0 ("xilinx:
zynqmp: Add support for runtime dfu_alt_info setup").
Code needs to be enabled by CONFIG_SET_DFU_ALT_INFO.
And also enable capsule on disk for RAW firmware images with efidebug
command.
Two indexes are supported for SPL flow. Images can be generated like:
./tools/mkeficapsule --raw spl/boot.bin --index 1 capsule1.bin
./tools/mkeficapsule --raw u-boot.img --index 2 capsule2.bin
Then place them to SD card and load them:
load mmc 0 10000000 capsule1.bin && efidebug capsule update -v 10000000
load mmc 0 10000000 capsule2.bin && efidebug capsule update -v 10000000
FSBL flow will also work where only index 1 capsule is used. There
should be enough space for using boot.bin with bitstream too.
Zynq also support multiple boot locations in SPI or MMC but it is not wired
by this patch.
Marek Szyprowski [Fri, 17 Sep 2021 08:19:43 +0000 (10:19 +0200)]
arm: rpi: perform XHCI firmware upload only once
XHCI firmware upload must be performed only once after initializing the
PCI bridge. This fixes USB stack initialization after calling "usb stop;
usb start" on Raspberry Pi 4B.
Signed-off-by: Marek Szyprowski <m.szyprowski@samsung.com> Reviewed-by: Nicolas Saenz Julienne <nsaenz@kernel.org> Signed-off-by: Matthias Brugger <mbrugger@suse.com>
Robert Marko [Mon, 27 Sep 2021 21:03:26 +0000 (23:03 +0200)]
arm: dts: armada8040: Fix CP0 eMMC/SDIO support
During the migration to a single DTSI for the CP110-s specific pinctrl
compatibles were moved to the SoC DTSI as CP0 and CP1 have some specifics.
Namely, CP0 eMMC/SDIO support depends on the mvebu-pinctrl driver setting
the BIT(0) in eMMC PHY IO Control 0 Register to 0 in order for the connect
the eMMC/SDIO PHY to the controller and not use it as a MPP pin multiplexor.
So, the mvebu-pinctrl driver check specifically for the
"marvell,armada-8k-cpm-pinctrl" compatible to clear the that bit.
Issue is that compatibles in the 8040 DTSI were set to "marvell,8k-cpm-pinctrl"
for CP0 and "marvell,8k-cps-pinctrl" for the CP1.
This is obviously incorrect as the pinctrl driver does not know about these.
So fix the regression by applying correct compatibles to the DTSI.
Regression found and tested on the Puzzle M801 board.
Fixes: a0ba97e5 ("arm: armada: dts: Use a single dtsi for cp110 die description") Signed-off-by: Robert Marko <robert.marko@sartura.hr> Reviewed-by: Stefan Roese <sr@denx.de>
- i2c: rcar_i2c: Enable configuring SCL rise and fall times
- i2c: mvtwsi: Add support for DM clocks and resets
- mtd: nand: raw: convert nand_dt_init() to ofnode_xx() interface
Tom Rini [Tue, 28 Sep 2021 17:57:56 +0000 (13:57 -0400)]
Merge branch '2021-09-28-regression-fixes'
- Reintroduce creating internally the "nor%d" style names, in order to
fix some use U-Boot use-cases involving the "mtd" command.
- Fix a regression over the default SPI bus mode shown by having the
compiled default actually start being used. The correct default here
is 0.
- Fix ethernet on imx7d-sdb
- Fix a regression with MTD NAND devices when OF_LIVE is enabled
imx: imx7d-sdb: fix ethernet, sync .dts with linux
Commit 0d52bab46 (mx7dsabre: Enable DM_ETH) changed these flags from 0
(aka GPIO_ACTIVE_HIGH) to GPIO_ACTIVE_LOW. It claimed to "Also sync
device tree with v5.5-rc1", but in the linux tree, these gpios have
always been GPIO_ACTIVE_HIGH ever since this node was introduced
around v4.13 (linux commit 184f39b5).
I'm guessing that the reason for the GPIO_ACTIVE_LOW was to work
around the behaviour of the soft-spi driver back then, which
effectively defaulted to spi-mode 3 and not 0. That was arguably a bug
in the soft-spi driver, which then got fixed in 0e146993bb3 (spi: add
support for all spi modes with soft spi), but that commit then broke
ethernet on this board.
Fix it by setting the gpios as active high, which as a bonus actually
brings us in sync with the .dts in the linux source tree.
Without this, one gets
Net: Could not get PHY for FEC0: addr 0
No ethernet found.
With this, ethernet (at least ping and tftp) works as expected from
the U-Boot shell.
mtd: nand: raw: convert nand_dt_init() to ofnode_xx() interface
nand_dt_init() is still using fdtdec_xx() interface.
If OF_LIVE flag is enabled, dt property can't be get anymore.
Updating all fdtdec_xx() interface to ofnode_xx() to solve this issue.
For doing this, node parameter type must be ofnode.
First idea was to convert "node" parameter to ofnode type inside
nand_dt_init() using offset_to_ofnode(node). But offset_to_ofnode()
is not bijective, in case OF_LIVE flag is enabled, it performs an assert().
So, this leads to update nand_chip struct flash_node field from int to
ofnode and to update all nand_dt_init() callers.
Marek Vasut [Tue, 14 Sep 2021 18:28:24 +0000 (20:28 +0200)]
mtd: spi: Set CONFIG_SF_DEFAULT_MODE default to 0
Before e2e95e5e254 ("spi: Update speed/mode on change") most systems
silently defaulted to SF bus mode 0. Now the mode is always updated,
which causes breakage. It seems most SF which are used as boot media
operate in bus mode 0, so switch that as the default.
This should fix booting at least on Altera SoCFPGA, ST STM32, Xilinx
ZynqMP, NXP iMX and Rockchip SoCs, which recently ran into trouble
with mode 3. Marvell Kirkwood and Xilinx microblaze need to be checked
as those might need mode 3.
Signed-off-by: Marek Vasut <marex@denx.de> Cc: Aleksandar Gerasimovski <aleksandar.gerasimovski@hitachi-powergrids.com> Cc: Andreas Biessmann <andreas@biessmann.org> Cc: Eugen Hristev <eugen.hristev@microchip.com> Cc: Michal Simek <michal.simek@xilinx.com> Cc: Patrice Chotard <patrice.chotard@foss.st.com> Cc: Patrick Delaunay <patrick.delaunay@foss.st.com> Cc: Peng Fan <peng.fan@nxp.com> Cc: Siew Chin Lim <elly.siew.chin.lim@intel.com> Cc: Tom Rini <trini@konsulko.com> Cc: Valentin Longchamp <valentin.longchamp@hitachi-powergrids.com> Cc: Vignesh Raghavendra <vigneshr@ti.com>
Patrick Delaunay [Wed, 22 Sep 2021 16:29:08 +0000 (18:29 +0200)]
mtd: spi: nor: force mtd name to "nor%d"
Force the mtd name of spi-nor to "nor" + the driver sequence number:
"nor0", "nor1"... beginning after the existing nor devices.
This patch is coherent with existing "nand" and "spi-nand"
mtd device names.
When CFI MTD NOR device are supported, the spi-nor index is chosen after
the last CFI device defined by CONFIG_SYS_MAX_FLASH_BANKS.
When CONFIG_SYS_MAX_FLASH_BANKS_DETECT is activated, this config
is replaced by to cfi_flash_num_flash_banks in the include file
mtd/cfi_flash.h.
This generic name "nor%d" can be use to identify the mtd spi-nor device
without knowing the real device name or the DT path of the device,
used with API get_mtd_device_nm() and is used in mtdparts command.
This patch also avoids issue when the same NOR device is present 2 times,
for example on STM32MP15F-EV1:
STM32MP> mtd list
SF: Detected mx66l51235l with page size 256 Bytes, erase size 64 KiB, \
total 64 MiB
The same mtd name "mx66l51235l" identify the 2 instances
mx66l51235l@0 and mx66l51235l@1.
This patch fixes a ST32CubeProgrammer / stm32prog command issue
with nor0 target on STM32MP157C-EV1 board introduced by
commit b7f060565e31 ("mtd: spi-nor: allow registering multiple MTDs when
DM is enabled").
Fixes: b7f060565e31 ("mtd: spi-nor: allow registering multiple MTDs when DM is enabled") Signed-off-by: Patrick Delaunay <patrick.delaunay@foss.st.com>
[trini: Add <dm/device.h> to <mtd.h> for DM_MAX_SEQ_STR] Signed-off-by: Tom Rini <trini@konsulko.com>
Vladimir Oltean [Mon, 27 Sep 2021 11:22:03 +0000 (14:22 +0300)]
net: dsa: ensure port names are NULL-terminated after DSA_PORT_NAME_LENGTH truncation
strncpy() simply bails out when copying a source string whose size
exceeds the destination string size, potentially leaving the destination
string unterminated.
One possible way to address is to pass DSA_PORT_NAME_LENGTH - 1 and a
previously zero-initialized destination string, but this is more
difficult to maintain.
The chosen alternative is to use strlcpy(), which properly limits the
copy len in the (srclen >= size) case to "size - 1", and which is also
more efficient than the strncpy() byte-by-byte implementation by using
memcpy. The destination string returned by strlcpy() is always NULL
terminated.
Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com> Reviewed-by: Ramon Fried <rfried.dev@gmail.com>
Vladimir Oltean [Mon, 27 Sep 2021 11:22:01 +0000 (14:22 +0300)]
scripts: ensure the cocci script for miiphy_register does not leave NULL-unterminated strings
strncpy() simply bails out when copying a source string whose size
exceeds the destination string size, potentially leaving the destination
string unterminated.
One possible way to address is to pass MDIO_NAME_LEN - 1 and a
previously zero-initialized destination string, but this is more
difficult to maintain.
The chosen alternative is to use strlcpy(), which properly limits the
copy len in the (srclen >= size) case to "size - 1", and which is also
more efficient than the strncpy() byte-by-byte implementation by using
memcpy. The destination string returned by strlcpy() is always NULL
terminated.
Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com>
Vladimir Oltean [Mon, 27 Sep 2021 11:22:00 +0000 (14:22 +0300)]
net: mdio-uclass: rewrite dm_mdio_post_probe using strlcpy
dm_mdio_post_probe used to be vulnerable after truncation, but has been
patched by commit 398e7512d8d7 ("net: Fix Covarity Defect 244093").
Nonetheless, we can use strlcpy like the rest of the code base now,
which yields the same result.
Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com> Reviewed-by: Ramon Fried <rfried.dev@gmail.com>
Vladimir Oltean [Mon, 27 Sep 2021 11:21:59 +0000 (14:21 +0300)]
net: qe: uec: ensure mdiodev->name is NULL terminated after MDIO_NAME_LEN truncation
strncpy() simply bails out when copying a source string whose size
exceeds the destination string size, potentially leaving the destination
string unterminated.
One possible way to address is to pass MDIO_NAME_LEN - 1 and a
previously zero-initialized destination string, but this is more
difficult to maintain.
The chosen alternative is to use strlcpy(), which properly limits the
copy len in the (srclen >= size) case to "size - 1", and which is also
more efficient than the strncpy() byte-by-byte implementation by using
memcpy. The destination string returned by strlcpy() is always NULL
terminated.
Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com> Reviewed-by: Ramon Fried <rfried.dev@gmail.com>
Vladimir Oltean [Mon, 27 Sep 2021 11:21:58 +0000 (14:21 +0300)]
net: davinci_emac: ensure mdiodev->name is NULL terminated after MDIO_NAME_LEN truncation
strncpy() simply bails out when copying a source string whose size
exceeds the destination string size, potentially leaving the destination
string unterminated.
One possible way to address is to pass MDIO_NAME_LEN - 1 and a
previously zero-initialized destination string, but this is more
difficult to maintain.
The chosen alternative is to use strlcpy(), which properly limits the
copy len in the (srclen >= size) case to "size - 1", and which is also
more efficient than the strncpy() byte-by-byte implementation by using
memcpy. The destination string returned by strlcpy() is always NULL
terminated.
Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com> Reviewed-by: Ramon Fried <rfried.dev@gmail.com>
Vladimir Oltean [Mon, 27 Sep 2021 11:21:57 +0000 (14:21 +0300)]
net: smc911x: ensure mdiodev->name is NULL terminated after MDIO_NAME_LEN truncation
strncpy() simply bails out when copying a source string whose size
exceeds the destination string size, potentially leaving the destination
string unterminated.
One possible way to address is to pass MDIO_NAME_LEN - 1 and a
previously zero-initialized destination string, but this is more
difficult to maintain.
The chosen alternative is to use strlcpy(), which properly limits the
copy len in the (srclen >= size) case to "size - 1", and which is also
more efficient than the strncpy() byte-by-byte implementation by using
memcpy. The destination string returned by strlcpy() is always NULL
terminated.
Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com> Reviewed-by: Ramon Fried <rfried.dev@gmail.com>
Vladimir Oltean [Mon, 27 Sep 2021 11:21:56 +0000 (14:21 +0300)]
net: sh_eth: ensure mdiodev->name is NULL terminated after MDIO_NAME_LEN truncation
strncpy() simply bails out when copying a source string whose size
exceeds the destination string size, potentially leaving the destination
string unterminated.
One possible way to address is to pass MDIO_NAME_LEN - 1 and a
previously zero-initialized destination string, but this is more
difficult to maintain.
The chosen alternative is to use strlcpy(), which properly limits the
copy len in the (srclen >= size) case to "size - 1", and which is also
more efficient than the strncpy() byte-by-byte implementation by using
memcpy. The destination string returned by strlcpy() is always NULL
terminated.
Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com> Reviewed-by: Ramon Fried <rfried.dev@gmail.com>
Vladimir Oltean [Mon, 27 Sep 2021 11:21:55 +0000 (14:21 +0300)]
net: mvgbe: ensure mdiodev->name is NULL terminated after MDIO_NAME_LEN truncation
strncpy() simply bails out when copying a source string whose size
exceeds the destination string size, potentially leaving the destination
string unterminated.
One possible way to address is to pass MDIO_NAME_LEN - 1 and a
previously zero-initialized destination string, but this is more
difficult to maintain.
The chosen alternative is to use strlcpy(), which properly limits the
copy len in the (srclen >= size) case to "size - 1", and which is also
more efficient than the strncpy() byte-by-byte implementation by using
memcpy. The destination string returned by strlcpy() is always NULL
terminated.
Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com> Reviewed-by: Ramon Fried <rfried.dev@gmail.com>
Vladimir Oltean [Mon, 27 Sep 2021 11:21:54 +0000 (14:21 +0300)]
net: dsa: felix: ensure mii_bus->name is NULL terminated after MDIO_NAME_LEN truncation
strncpy() simply bails out when copying a source string whose size
exceeds the destination string size, potentially leaving the destination
string unterminated.
One possible way to address is to pass MDIO_NAME_LEN - 1 and a
previously zero-initialized destination string, but this is more
difficult to maintain.
The chosen alternative is to use strlcpy(), which properly limits the
copy len in the (srclen >= size) case to "size - 1", and which is also
more efficient than the strncpy() byte-by-byte implementation by using
memcpy. The destination string returned by strlcpy() is always NULL
terminated.
Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com> Reviewed-by: Ramon Fried <rfried.dev@gmail.com>
Vladimir Oltean [Mon, 27 Sep 2021 11:21:53 +0000 (14:21 +0300)]
net: mpc8xx_fec: ensure mdiodev->name is NULL terminated after MDIO_NAME_LEN truncation
strncpy() simply bails out when copying a source string whose size
exceeds the destination string size, potentially leaving the destination
string unterminated.
One possible way to address is to pass MDIO_NAME_LEN - 1 and a
previously zero-initialized destination string, but this is more
difficult to maintain.
The chosen alternative is to use strlcpy(), which properly limits the
copy len in the (srclen >= size) case to "size - 1", and which is also
more efficient than the strncpy() byte-by-byte implementation by using
memcpy. The destination string returned by strlcpy() is always NULL
terminated.
Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com> Reviewed-by: Ramon Fried <rfried.dev@gmail.com>
Vladimir Oltean [Mon, 27 Sep 2021 11:21:52 +0000 (14:21 +0300)]
net: macb: ensure mdiodev->name is NULL terminated after MDIO_NAME_LEN truncation
strncpy() simply bails out when copying a source string whose size
exceeds the destination string size, potentially leaving the destination
string unterminated.
One possible way to address is to pass MDIO_NAME_LEN - 1 and a
previously zero-initialized destination string, but this is more
difficult to maintain.
The chosen alternative is to use strlcpy(), which properly limits the
copy len in the (srclen >= size) case to "size - 1", and which is also
more efficient than the strncpy() byte-by-byte implementation by using
memcpy. The destination string returned by strlcpy() is always NULL
terminated.
Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com> Reviewed-by: Ramon Fried <rfried.dev@gmail.com>
Vladimir Oltean [Mon, 27 Sep 2021 11:21:51 +0000 (14:21 +0300)]
net: lpc32xx: ensure mdiodev->name is NULL terminated after MDIO_NAME_LEN truncation
strncpy() simply bails out when copying a source string whose size
exceeds the destination string size, potentially leaving the destination
string unterminated.
One possible way to address is to pass MDIO_NAME_LEN - 1 and a
previously zero-initialized destination string, but this is more
difficult to maintain.
The chosen alternative is to use strlcpy(), which properly limits the
copy len in the (srclen >= size) case to "size - 1", and which is also
more efficient than the strncpy() byte-by-byte implementation by using
memcpy. The destination string returned by strlcpy() is always NULL
terminated.
Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com> Reviewed-by: Ramon Fried <rfried.dev@gmail.com>
Vladimir Oltean [Mon, 27 Sep 2021 11:21:50 +0000 (14:21 +0300)]
net: ftmac110: ensure mdiodev->name is NULL terminated after MDIO_NAME_LEN truncation
strncpy() simply bails out when copying a source string whose size
exceeds the destination string size, potentially leaving the destination
string unterminated.
One possible way to address is to pass MDIO_NAME_LEN - 1 and a
previously zero-initialized destination string, but this is more
difficult to maintain.
The chosen alternative is to use strlcpy(), which properly limits the
copy len in the (srclen >= size) case to "size - 1", and which is also
more efficient than the strncpy() byte-by-byte implementation by using
memcpy. The destination string returned by strlcpy() is always NULL
terminated.
Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com> Reviewed-by: Ramon Fried <rfried.dev@gmail.com>
Vladimir Oltean [Mon, 27 Sep 2021 11:21:49 +0000 (14:21 +0300)]
net: mcdmafec: ensure bus->name is NULL terminated after MDIO_NAME_LEN truncation
strncpy() simply bails out when copying a source string whose size
exceeds the destination string size, potentially leaving the destination
string unterminated.
One possible way to address is to pass MDIO_NAME_LEN - 1 and a
previously zero-initialized destination string, but this is more
difficult to maintain.
The chosen alternative is to use strlcpy(), which properly limits the
copy len in the (srclen >= size) case to "size - 1", and which is also
more efficient than the strncpy() byte-by-byte implementation by using
memcpy. The destination string returned by strlcpy() is always NULL
terminated.
Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com> Reviewed-by: Ramon Fried <rfried.dev@gmail.com>
Vladimir Oltean [Mon, 27 Sep 2021 11:21:48 +0000 (14:21 +0300)]
net: enetc: ensure imdio.name is NULL terminated after MDIO_NAME_LEN truncation
strncpy() simply bails out when copying a source string whose size
exceeds the destination string size, potentially leaving the destination
string unterminated.
One possible way to address is to pass MDIO_NAME_LEN - 1 and a
previously zero-initialized destination string, but this is more
difficult to maintain.
The chosen alternative is to use strlcpy(), which properly limits the
copy len in the (srclen >= size) case to "size - 1", and which is also
more efficient than the strncpy() byte-by-byte implementation by using
memcpy. The destination string returned by strlcpy() is always NULL
terminated.
Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com> Reviewed-by: Ramon Fried <rfried.dev@gmail.com>
Vladimir Oltean [Mon, 27 Sep 2021 11:21:47 +0000 (14:21 +0300)]
net: ep93xx: ensure mdiodev->name is NULL terminated after MDIO_NAME_LEN truncation
strncpy() simply bails out when copying a source string whose size
exceeds the destination string size, potentially leaving the destination
string unterminated.
One possible way to address is to pass MDIO_NAME_LEN - 1 and a
previously zero-initialized destination string, but this is more
difficult to maintain.
The chosen alternative is to use strlcpy(), which properly limits the
copy len in the (srclen >= size) case to "size - 1", and which is also
more efficient than the strncpy() byte-by-byte implementation by using
memcpy. The destination string returned by strlcpy() is always NULL
terminated.
Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com> Reviewed-by: Ramon Fried <rfried.dev@gmail.com>
Vladimir Oltean [Mon, 27 Sep 2021 11:21:46 +0000 (14:21 +0300)]
net: eepro100: ensure mdiodev->name is NULL terminated after MDIO_NAME_LEN truncation
strncpy() simply bails out when copying a source string whose size
exceeds the destination string size, potentially leaving the destination
string unterminated.
One possible way to address is to pass MDIO_NAME_LEN - 1 and a
previously zero-initialized destination string, but this is more
difficult to maintain.
The chosen alternative is to use strlcpy(), which properly limits the
copy len in the (srclen >= size) case to "size - 1", and which is also
more efficient than the strncpy() byte-by-byte implementation by using
memcpy. The destination string returned by strlcpy() is always NULL
terminated.
Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com> Reviewed-by: Ramon Fried <rfried.dev@gmail.com>
Vladimir Oltean [Mon, 27 Sep 2021 11:21:45 +0000 (14:21 +0300)]
net: bcm-sf2: ensure mdiodev->name is NULL terminated after MDIO_NAME_LEN truncation
strncpy() simply bails out when copying a source string whose size
exceeds the destination string size, potentially leaving the destination
string unterminated.
One possible way to address is to pass MDIO_NAME_LEN - 1 and a
previously zero-initialized destination string, but this is more
difficult to maintain.
The chosen alternative is to use strlcpy(), which properly limits the
copy len in the (srclen >= size) case to "size - 1", and which is also
more efficient than the strncpy() byte-by-byte implementation by using
memcpy. The destination string returned by strlcpy() is always NULL
terminated.
Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com> Reviewed-by: Ramon Fried <rfried.dev@gmail.com>
Vladimir Oltean [Mon, 27 Sep 2021 11:21:44 +0000 (14:21 +0300)]
net: at91_emac: ensure mdiodev->name is NULL terminated after MDIO_NAME_LEN truncation
strncpy() simply bails out when copying a source string whose size
exceeds the destination string size, potentially leaving the destination
string unterminated.
One possible way to address is to pass MDIO_NAME_LEN - 1 and a
previously zero-initialized destination string, but this is more
difficult to maintain.
The chosen alternative is to use strlcpy(), which properly limits the
copy len in the (srclen >= size) case to "size - 1", and which is also
more efficient than the strncpy() byte-by-byte implementation by using
memcpy. The destination string returned by strlcpy() is always NULL
terminated.
Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com> Reviewed-by: Ramon Fried <rfried.dev@gmail.com>
Vladimir Oltean [Mon, 27 Sep 2021 11:21:43 +0000 (14:21 +0300)]
net: armada100_fec: ensure mdiodev->name is NULL terminated after MDIO_NAME_LEN truncation
strncpy() simply bails out when copying a source string whose size
exceeds the destination string size, potentially leaving the destination
string unterminated.
One possible way to address is to pass MDIO_NAME_LEN - 1 and a
previously zero-initialized destination string, but this is more
difficult to maintain.
The chosen alternative is to use strlcpy(), which properly limits the
copy len in the (srclen >= size) case to "size - 1", and which is also
more efficient than the strncpy() byte-by-byte implementation by using
memcpy. The destination string returned by strlcpy() is always NULL
terminated.
Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com> Reviewed-by: Ramon Fried <rfried.dev@gmail.com>
Vladimir Oltean [Mon, 27 Sep 2021 11:21:42 +0000 (14:21 +0300)]
board: gdsys: a38x: ensure mdiodev->name is NULL terminated after MDIO_NAME_LEN truncation
strncpy() simply bails out when copying a source string whose size
exceeds the destination string size, potentially leaving the destination
string unterminated.
One possible way to address is to pass MDIO_NAME_LEN - 1 and a
previously zero-initialized destination string, but this is more
difficult to maintain.
The chosen alternative is to use strlcpy(), which properly limits the
copy len in the (srclen >= size) case to "size - 1", and which is also
more efficient than the strncpy() byte-by-byte implementation by using
memcpy. The destination string returned by strlcpy() is always NULL
terminated.
Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com> Reviewed-by: Ramon Fried <rfried.dev@gmail.com>
Vladimir Oltean [Mon, 27 Sep 2021 11:21:41 +0000 (14:21 +0300)]
arch: powerpc: mpc85xx: ensure mdiodev->name is NULL terminated after MDIO_NAME_LEN truncation
strncpy() simply bails out when copying a source string whose size
exceeds the destination string size, potentially leaving the destination
string unterminated.
One possible way to address is to pass MDIO_NAME_LEN - 1 and a
previously zero-initialized destination string, but this is more
difficult to maintain.
The chosen alternative is to use strlcpy(), which properly limits the
copy len in the (srclen >= size) case to "size - 1", and which is also
more efficient than the strncpy() byte-by-byte implementation by using
memcpy. The destination string returned by strlcpy() is always NULL
terminated.
Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com> Reviewed-by: Ramon Fried <rfried.dev@gmail.com>
Vladimir Oltean [Sat, 18 Sep 2021 12:46:55 +0000 (15:46 +0300)]
net: tsec: read the phy-mode property as fallback to phy-connection-type
The two should be equivalent, but at the moment some platforms
(ls1021a-tsn.dts) use phy-mode only, which is not parsed.
Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com> Reviewed-by: Bin Meng <bmeng.cn@gmail.com> Tested-by: Bin Meng <bmeng.cn@gmail.com> Reviewed-by: Ramon Fried <rfried.dev@gmail.com>
Vladimir Oltean [Sat, 18 Sep 2021 12:46:54 +0000 (15:46 +0300)]
net: tsec: only call tsec_get_interface as fallback to DT-specified PHY mode
Currently the init_phy function may overwrite the priv->interface
property, since it calls tsec_get_interface which tries to determine it
dynamically based on default register values in ECNTRL.
Let's do that only if phy-connection-type happens to not be defined in
the device tree.
Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com> Reviewed-by: Bin Meng <bmeng.cn@gmail.com> Tested-by: Bin Meng <bmeng.cn@gmail.com> Reviewed-by: Ramon Fried <rfried.dev@gmail.com>
Vladimir Oltean [Sat, 18 Sep 2021 12:32:37 +0000 (15:32 +0300)]
net: dsa: felix: remove "xgmii" phy-mode
The felix driver runs only on NXP LS1028A, which most definitely does
not support the parallel 10G interface, just USXGMII, and that only up
to 2.5Gbps (toned down from 10 Gbps via symbol replication).
Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com> Reviewed-by: Ramon Fried <rfried.dev@gmail.com>
Vladimir Oltean [Sat, 18 Sep 2021 12:32:36 +0000 (15:32 +0300)]
net: enetc: remove support for "xgmii" phy-mode
The enetc driver runs only on NXP LS1028A, which most definitely does
not support the parallel 10G interface, just USXGMII, and that only up
to 2.5Gbps (toned down from 10 Gbps via symbol replication).
Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com> Reviewed-by: Ramon Fried <rfried.dev@gmail.com>
Vladimir Oltean [Sat, 18 Sep 2021 12:32:35 +0000 (15:32 +0300)]
net: freescale: replace usage of phy-mode = "sgmii-2500" with "2500base-x"
After the discussion here:
https://lore.kernel.org/netdev/20210603143453.if7hgifupx5k433b@pali/
which resulted in this patch:
https://patchwork.kernel.org/project/netdevbpf/patch/20210704134325.24842-1-pali@kernel.org/
and many other discussions before it, notably:
https://patchwork.kernel.org/project/linux-arm-kernel/patch/1512016235-15909-1-git-send-email-Bhaskar.Upadhaya@nxp.com/
it became apparent that nobody really knows what "SGMII 2500" is.
Certainly, Freescale/NXP hardware engineers name this protocol
"SGMII 2500" in the reference manuals, but the PCS devices do not
support any "SGMII" specific features when operating at the speed of
2500 Mbps, no in-band autoneg and no speed change via symbol replication
. So that leaves a fixed speed of 2500 Mbps using a coding of 8b/10b
with a SERDES lane frequency of 3.125 GHz. In fact, "SGMII 2500 without
in-band autoneg and at a fixed speed" is indistinguishable from
"2500base-x without in-band autoneg", which is precisely what these NXP
devices support.
So it just appears that "SGMII 2500" is an unclear name with no clear
definition that stuck.
As such, in the Linux kernel, the drivers which use this SERDES protocol
use the 2500base-x phy-mode.
This patch converts U-Boot to use 2500base-x too, or at least, as much
as it can.
Note that I would have really liked to delete PHY_INTERFACE_MODE_SGMII_2500
completely, but the mvpp2 driver seems to even distinguish between SGMII
2500 and 2500base-X. Namely, it enables in-band autoneg for one but not
the other, and forces flow control for one but not the other. This goes
back to the idea that maybe 2500base-X is a fiber protocol and SGMII-2500
is an MII protocol (connects a MAC to a PHY such as Aquantia), but the
two are practically indistinguishable through everything except use case.
NXP devices can support both use cases through an identical configuration,
for example RX flow control can be unconditionally enabled in order to
support rate adaptation performed by an Aquantia PHY. At least I can
find no indication in online documents published by Cisco which would
point towards "SGMII-2500" being an actual standard with an actual
definition, so I cannot say "yes, NXP devices support it".
Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com> Reviewed-by: Ramon Fried <rfried.dev@gmail.com>
Vladimir Oltean [Sat, 18 Sep 2021 12:32:34 +0000 (15:32 +0300)]
net: replace the "xfi" phy-mode with "10gbase-r"
As part of the effort of making U-Boot work with the same device tree as
Linux, there is an issue with the "xfi" phy-mode. To be precise, in
Linux there was a discussion (for those who have time to read:
https://lore.kernel.org/netdev/1576768881-24971-2-git-send-email-madalin.bucur@oss.nxp.com/)
Vladimir Oltean [Sat, 18 Sep 2021 11:55:24 +0000 (14:55 +0300)]
net: phy: genphy_init can be static
To avoid a warning with W=1 about this function not having a previous
prototype, declare it as static, because it is not used outside of this
translation module.
Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com> Reviewed-by: Bin Meng <bmeng.cn@gmail.com> Reviewed-by: Ramon Fried <rfried.dev@gmail.com>
Vladimir Oltean [Sat, 18 Sep 2021 11:49:55 +0000 (14:49 +0300)]
net: dsa: pass CPU port fixed PHY to .port_disable
While adding the logic for DSA to register a fixed-link PHY for the CPU
port, I forgot to pass it to the .port_disable method too, just
.port_enable.
Bug had no impact for felix_switch.c, due to the phy argument not being
used, but ksz9477.c does use it => NULL pointer dereference.
Fixes: fc054d563bfb ("net: Introduce DSA class for Ethernet switches") Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com> Reviewed-by: Ramon Fried <rfried.dev@gmail.com>
Vladimir Oltean [Tue, 24 Aug 2021 12:00:42 +0000 (15:00 +0300)]
net: dsa: felix: call phy_config at .port_probe() time
It is an unfortunate reality that some PHY settings done by U-Boot
persist even after the PHY is reset and taken over by Linux, and even
more unfortunate that Linux has come to depend on things being set in a
certain way.
For example, on the NXP LS1028A-RDB, the felix switch ports are
connected to a VSC8514 QSGMII PHY. Between the switch port PCS and the
PHY, the U-Boot drivers enable in-band auto-negotiation which makes the
copper-side negotiated speed and duplex be transmitted from the PHY to
the MAC automatically.
The PHY driver portion that does this is in vsc8514_config():
/* Enable Serdes Auto-negotiation */
phy_write(phydev, MDIO_DEVAD_NONE, PHY_EXT_PAGE_ACCESS,
PHY_EXT_PAGE_ACCESS_EXTENDED3);
val = phy_read(phydev, MDIO_DEVAD_NONE, MIIM_VSC8514_MAC_SERDES_CON);
val = val | MIIM_VSC8574_MAC_SERDES_ANEG;
phy_write(phydev, MDIO_DEVAD_NONE, MIIM_VSC8514_MAC_SERDES_CON, val);
The point is that in-band autoneg should be turned on in both the PHY
and the MAC, or off in both the PHY and the MAC, otherwise the QSGMII
link will be broken.
And because phy_config() is currently called at .port_enable() time, the
result is that ports on which traffic has been sent in U-Boot will have
in-band autoneg enabled, and the rest won't.
It can be argued that the Linux kernel should not assume one way or
another and just reinitialize everything according to what it expects,
and that is completely fair. In fact, I've already started an attempt to
remove this dependency, although admittedly I am making slow progress at
it:
https://patchwork.kernel.org/project/netdevbpf/cover/20210212172341.3489046-1-olteanv@gmail.com/
Nonetheless, the sad reality is that NXP also has, apart from kernel
drivers, some user space networking (DPDK), and for some reason, the
expectation there is that somebody else initializes the PHYs. The kernel
can't do it because the device ownership doesn't belong to the kernel,
so what remains is for the bootloader to do it (especially since other
drivers generally call phy_config() at probe time). This is a really
weak guarantee that might break at any time, but apparently that is
enough for some.
Since initializing the ports and PHYs at probe time does not break
anything, we can just do that.
Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com> Reviewed-by: Ramon Fried <rfried.dev@gmail.com> Tested-by: Michael Walle <michael@walle.cc>
Vladimir Oltean [Tue, 24 Aug 2021 12:00:41 +0000 (15:00 +0300)]
net: dsa: introduce a .port_probe() method in struct dsa_ops
Some drivers might want to execute code for each port at probe time, as
opposed to executing code just-in-time for the port selected for
networking.
To cater to that use case, introduce a .port_probe() callback method
into the DSA switch operations which is called for each available port,
at the end of dsa_port_probe().
Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com> Reviewed-by: Ramon Fried <rfried.dev@gmail.com> Tested-by: Michael Walle <michael@walle.cc>
projects / u-boot.git / log
