The commit 6c993815bb ("net: phy: xilinx: Be compatible with live OF
tree") change driver behavior to while loop which wasn't correct because
the driver was looping over again and again. The reason was that
ofnode_valid() is taking 0 as correct value.
Fix it by changing while loop to ofnode_for_each_subnode() which is only
loop over available nodes.
Fixes: 6c993815bb ("net: phy: xilinx: Be compatible with live OF tree")
Signed-off-by: Michal Simek <michal.simek@xilinx.com>
Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
Trying to clear PIO_START register when it is non-zero (which indicates
that previous PIO transfer has not finished yet) causes an External
Abort with SError 0xbf000002.
This bug is currently worked around in TF-A by handling External Aborts
in EL3 and ignoring this particular SError.
This workaround was also discussed at:
https://git.trustedfirmware.org/TF-A/trusted-firmware-a.git/commit/?id=3c7dcdac5c50https://lore.kernel.org/linux-pci/20190316161243.29517-1-repk@triplefau.lt/https://lore.kernel.org/linux-pci/971be151d24312cc533989a64bd454b4@www.loen.fr/https://review.trustedfirmware.org/c/TF-A/trusted-firmware-a/+/1541
Implement a proper fix to prevent this External Abort. As it is not
possible to cancel a pending PIO transfer, simply do not start a new one
if previous has not finished yet. In this case return an error to the
caller.
In most cases this SError happens when there is no PCIe card connected
or when PCIe link is down. The reason is that in these cases a PIO
transfer takes about 1.44 seconds. For this reason we also increase the
wait timeout in pcie_advk_wait_pio() to 1.5 seconds.
If PIO read transfer for PCI_VENDOR_ID register times out, or if it
isn't possible to read it yet because previous transfer is not finished,
return Completion Retry Status value instead of failing, to give the
caller a chance to send a new read request.
Signed-off-by: Pali Rohár <pali@kernel.org>
Reviewed-by: Marek Behún <marek.behun@nic.cz>
Reviewed-by: Stefan Roese <sr@denx.de>
- Add additional step which enables the Impedance and PLL calibration.
- Enable old squelch detector instead of the new analog squelch detector
circuit and update host disconnect threshold value.
- Update LS TX driver strength coarse and fine adjustment values.
Change-Id: Ifa0a585bfb5ecab0bfa033eed6874ff98b16a7df
Signed-off-by: Grzegorz Jaszczyk <jaz@semihalf.com>
In CP115, comphy4 can be configured into SFI port1
(in addition to SFI0). This patch adds the option
described above.
In addition, rename all existing SFI/XFI references:
COMPHY_TYPE_SFI --> COMPHY_TYPE_SFI0
No functional change for exsiting configuration.
Change-Id: If9176222e0080424ba67347fe4d320215b1ba0c0
Signed-off-by: Igal Liberman <igall@marvell.com>
Signed-off-by: Konstantin Porotchkin <kostap@marvell.com>
According to Design Reference Specification the PHY PLL and Calibration
register from PHY0 are shared for multi-port PHY. PLL control registers
inside other PHY channels are not used.
This commit reworks utmi device tree nodes in a way that common PHY PLL
registers are moved to main utmi node. Accordingly both child nodes
utmi-unit range is reduced and register offsets in utmi_phy.h are updated
to this change.
This fixes issues in scenarios when only utmi port1 was in use, which
resulted with lack of correct pll initialization.
Change-Id: Icc520dfa719f43a09493ab31f671efbe88872097
Signed-off-by: Grzegorz Jaszczyk <jaz@semihalf.com>
New products can contain up to 6 usb ports, therefore allow to initialize
all relevant UTMI PHYs.
Change-Id: I28c36e59fa0e3e338bb3ee0cee2240b923f39785
Signed-off-by: Grzegorz Jaszczyk <jaz@semihalf.com>
Reviewed-by: Kostya Porotchkin <Kostya.Porotchkin@cavium.com>
It helps ATF to determine who called power off
function (U-boot/Linux) and act accordingly
Change-Id: Icfc5cbfdba64754496812154272b28c0ff639f0f
Signed-off-by: Igal Liberman <igall@marvell.com>
Reviewed-by: Grzegorz Jaszczyk <jaz@semihalf.com>
- the default value of comphy pipe selector is set to PCIe (x4)
in case of unconnected comphy the default value remains 0x4
which may lead to several issues with comphy initialization.
- this patch adds SMC call that powers off the comphy lane in case of
unconnected comphy.
Change-Id: I196b2916518dd8df3b159ffa85e2989b8e483087
Signed-off-by: Christine Gharzuzi <chrisg@marvell.com>
Signed-off-by: Igal Liberman <igall@marvell.com>
Reviewed-by: Grzegorz Jaszczyk <jaz@semihalf.com>
Currently, we don't pass id for SGMII 0/1.
A bug in comphy selector configuration was found (in comphy
firmware), after fixing it, SGMII0/1 have different configuration,
so we need to pass the ID the firmware.
Change-Id: Idcff4029cc9cf018278e493221b64b33574e0d38
Signed-off-by: Igal Liberman <igall@marvell.com>
Reviewed-by: Grzegorz Jaszczyk <jaz@semihalf.com>
Change-Id: I358792a96c13b54e700c05227cc7a8f6bd584694
Signed-off-by: Grzegorz Jaszczyk <jaz@semihalf.com>
Reviewed-by: Igal Liberman <igall@marvell.com>
Now the comphy configuration is handled in atf, therefore there is no
need to configure phy or pipe selector in u-boot, it is configured by
atf for each particular pair: lane and mode.
Change-Id: I0bebf8d5ff66dbeb6bf9ef90876195938a8eb705
Signed-off-by: Grzegorz Jaszczyk <jaz@semihalf.com>
Reviewed-by: Igal Liberman <igall@marvell.com>
Replace the XFI training with appropriate SMC call, so the firmware will
perform exact initialization.
Update Stefan 2021-03-23:
Move comphy_smc() function to an earlier place - necessary for the
mainline merge.
Change-Id: I789b130b05529dc80dadcf66aef407d93595b762
Signed-off-by: Grzegorz Jaszczyk <jaz@semihalf.com>
Signed-off-by: Stefan Roese <sr@denx.de>
Reviewed-by: Igal Liberman <igall@marvell.com>
Replace the comphy initialization for USB with appropriate SMC call,
so the firmware will execute required serdes configuration.
Change-Id: I7f773c0dfac70db9dd2653de2cdcfac577e78c4e
Signed-off-by: Grzegorz Jaszczyk <jaz@semihalf.com>
Replace the comphy initialization for RXAUI with appropriate SMC call,
so the firmware will execute required serdes configuration.
Change-Id: Iedae0285fb283e05bb263a8b4ce46e8e7451a309
Signed-off-by: Grzegorz Jaszczyk <jaz@semihalf.com>
Reviewed-by: Igal Liberman <igall@marvell.com>
Even if comphy types of SATA2/SATA3/SGMII3 and comphy speeds of
1.5G/3G/6.25G were referenced in the driver non configuration (dts)
was using it.
This patch removes unused definitions.
Change-Id: I53ed6f9d3a82b9d18cb4e488bc14d3cf687f9488
Signed-off-by: Grzegorz Jaszczyk <jaz@semihalf.com>
Signed-off-by: Konstantin Porotchkin <kostap@marvell.com>
This patch adds support for running RX training using new command called
"rx_training"
Usage:
rx_training - rx_training <cp id> <comphy id>
RX training allows to improve link quality (for SFI mode)
by running training sequence between us and the link partner,
this allows to reach better link quality then using static configuration.
Change-Id: I818fe67ccaf19a87af50d4c34a9db7d6802049a5
Signed-off-by: Igal Liberman <igall@marvell.com>
Signed-off-by: Marcin Wojtas <mw@semihalf.com>
This allows the lower level driver access to comphy map data
(required for RX training support, which is introduced
in the following patches).
Change-Id: Ib7ffdc4b32076c01c3a5d33f59552c9dfc6b12fa
Signed-off-by: Igal Liberman <igall@marvell.com>
If fdtdec_get_int can't find speed, set COMPHY_SPEED_INVALID
If fdtdec_get_int can't find type, set COMPHY_TYPE_INVALID
Move the error print if phy-type is invalid
Add continue to the probe loop (in a case of invalid phy)
Cosmetic changes
Change-Id: I0c61b40bfe685437426fe907942ed338b7845378
Signed-off-by: Igal Liberman <igall@marvell.com>
Add UTMI analog parameters initialization values according to
latest ETP.
Change-Id: I5bcca205a3995202a18ff126f371a81f69e205c8
Signed-off-by: Igal Liberman <igall@marvell.com>
UTMI should be initialized only for enabled device tree nodes.
This fix overrides current internal configuration array
entry with the next DT entry data if error is detected
during the current DT entry parsing or the current port
is disabled.
This way the internal configuration structure will only
contain valid ports information obtained from the DT.
Change-Id: I9c43c6a5d234e15ae9005d1c9bc983fc1f3544b8
Signed-off-by: Omri Itach <omrii@marvell.com>
Signed-off-by: Ken Ma <make@marvell.com>
In get_speed_string() we have an array (speed_strings[])
which includes all possible speed strings.
This array size and content must be aligned to the speed
defines in comphy_data.h.
This patch adds missing 5.125G speed, aligns speed_strings[]
and fixes incorrect printing when speed > 5.0G.
Change-Id: I9900d23595094be321be0c62fcaa88036324568e
Signed-off-by: Igal Liberman <igall@marvell.com>
Currently, all comphy definitions are PHY_TYPE_XX and PHY_SPEEED_XX.
Those definition might be confused with MDIO PHY definitions.
This patch does the following changes:
- PHY_TYPE_XX --> COMPHY_TYPE_XX
- PHY_SPEED_XX --> COMPHY_SPEED_XX
This improves readability, no functional change.
Change-Id: I2bd1d9289ebbc5c16fa80f9870f797ea1bcaf5fa
Signed-off-by: Igal Liberman <igall@marvell.com>
Signed-off-by: Konstantin Porotchkin <kostap@marvell.com>
- For some Marvell SoCs, like armada-3700, there are both
USB host and device controller, but on PHY level the
configuration is the same.
- The new type supports both USB device and USB host
- This patch is cherry-picked from u-boot-2015 as-is.
Change-Id: I01262027edd8ec23391cff6fb409b3009aedfbb9
Signed-off-by: jinghua <jinghua@marvell.com>
Signed-off-by: Ken Ma <make@marvell.com>
Reviewed-by: Igal Liberman <igall@marvell.com>
Because of the incorrectly supported SGMII_2500 mode, this patch
disables eth2 for now until this issue will be fixed in mainline.
Also fix an incorrect comment.
Signed-off-by: Stefan Roese <sr@denx.de>
Cc: Konstantin Porotchkin <kostap@marvell.com>
Cc: Stefan Chulski <stefanc@marvell.com>
Cc: Nadav Haklai <nadavh@marvell.com>
Cc: Marek Behun <marek.behun@nic.cz>
This patch adds the base support for the Marvell Octeon TX2 CN9130 CRB.
Not all interfaces are supported fully yet.
Signed-off-by: Konstantin Porotchkin <kostap@marvell.com>
Signed-off-by: Stefan Roese <sr@denx.de>
This patch adds the dtsi/dts files needed to support the Marvell
Octeon TX2 CN9130 CRB. This is only the base port with not all
interfaces supported fully.
Signed-off-by: Konstantin Porotchkin <kostap@marvell.com>
Signed-off-by: Stefan Roese <sr@denx.de>
Move the default environment location to the end of 4MB flash
region. This change allows to accomodate larger flash boot
images making space for forthcoming code changes.
Signed-off-by: Konstantin Porotchkin <kostap@marvell.com>
Signed-off-by: Stefan Roese <sr@denx.de>
Add support for SoCs based on AP807 die.
Remove unused include file for Armada-8020 SoC.
Signed-off-by: Konstantin Porotchkin <kostap@marvell.com>
Signed-off-by: Stefan Roese <sr@denx.de>
Use a single dtsi file for CP110 die instead of master/slave.
Moving to single file will allow miltiple DTSI inclusions with
re-defined CP index and name.
This change will also allow support for SoCs containing more than
two CP110 dies on board.
Move pin control definitions from CP110 DTS to board DTS files
Signed-off-by: Konstantin Porotchkin <kostap@marvell.com>
Signed-off-by: Stefan Roese <sr@denx.de>
- fix the dependency for MMC boot (add XENON to MVEBU_MMC)
- fix the bubt destination assignment (missing # in "else" case)
Signed-off-by: Konstantin Porotchkin <kostap@marvell.com>
Signed-off-by: Stefan Roese <sr@denx.de>
Add support for regulator-force-boot-off DT property.
This property can be used by the board/device drivers for
turning off regulators on early init stages as pre-requisite
for the other components initialization.
Signed-off-by: Konstantin Porotchkin <kostap@marvell.com>
Signed-off-by: Stefan Roese <sr@denx.de>
Cc: Jaehoon Chung <jh80.chung@samsung.com>
Cc: Simon Glass <sjg@chromium.org>
Reviewed-by: Jaehoon Chung <jh80.chung@samsung.com>
The dw_eth_pdata is not accessible from the mdio device, it gets the mdio bus plat
leading to random sleeps (-10174464 on Odroid-HC4).
This moves the dw_mdio_reset function to a common one taking the ethernet
device as parameter and use it from the dw_mdio_reset and dm_mdio variant functions.
Fixes: 5160b4567c ("net: designware: add DM_MDIO support")
Reported-by: Mark Kettenis <mark.kettenis@xs4all.nl>
Signed-off-by: Neil Armstrong <narmstrong@baylibre.com>
Reviewed-by: Ramon Fried <rfried.dev@gmail.com>
For systems which use generic PHY support and implement USB PHY driver,
the parsing of PHY properties is unnecessary, disable it.
Signed-off-by: Marek Vasut <marex@denx.de>
Tested-by: Ying-Chun Liu (PaulLiu) <paul.liu@linaro.org>
Cc: Fabio Estevam <festevam@gmail.com>
Cc: Peng Fan <peng.fan@nxp.com>
Cc: Stefano Babic <sbabic@denx.de>
Cc: Ye Li <ye.li@nxp.com>
Cc: uboot-imx <uboot-imx@nxp.com>
According to S26KL512S datasheet [1] and S29GL01GS datasheet [2],
the procedure to read out PPB lock bits is to send the PPB Entry,
PPB Read, Reset/ASO Exit. Currently, the code does send incorrect
PPB Entry, PPB Read and Reset/ASO Exit is completely missing.
The PPB Entry sent is implemented by sending flash_unlock_seq()
and flash_write_cmd(..., FLASH_CMD_READ_ID). This translates to
sequence 0x555:0xaa, 0x2aa:0x55, 0x555:0x90=FLASH_CMD_READ_ID.
However, both [1] and [2] specify the last byte of PPB Entry as
0xc0=AMD_CMD_SET_PPB_ENTRY instead of 0x90=FLASH_CMD_READ_ID,
that is 0x555:0xaa, 0x2aa:0x55, 0x555:0xc0=AMD_CMD_SET_PPB_ENTRY.
Since this does make sense, this patch fixes it and thus also
aligns the code in flash_get_size() with flash_real_protect().
The PPB Read returns 00h in case of Protected state and 01h in case
of Unprotected state, according to [1] Note 83 and [2] Note 17, so
invert the result. Moreover, align the arguments with similar code
in flash_real_protect().
Finally, Reset/ASO Exit command should be executed to exit the PPB
mode, so add the missing reset.
[1] https://www.cypress.com/file/213346/download
Document Number: 001-99198 Rev. *M
Table 40. Command Definitions, Nonvolatile Sector Protection
Command Set Definitions
[2] https://www.cypress.com/file/177976/download
Document Number: 001-98285 Rev. *R
Table 7.1 Command Definitions, Nonvolatile Sector Protection
Command Set Definitions
Fixes: 03deff433e ("cfi_flash: Read PPB sector protection from device for AMD/Spansion chips")
Signed-off-by: Marek Vasut <marek.vasut+renesas@gmail.com>
Cc: Stefan Roese <sr@denx.de>
Reviewed-by: Stefan Roese <sr@denx.de>
This fixes an issue getting resets index 1 and 3+, the spurius "> 0"
made it return the index 0 or 1, whatever index was passed.
The dm_test_reset_base() did not catch it, but the dm_test_reset_base() extension
catches it and this fixes the regression.
This also fixes a reggression on Amlogic G12A/G12B SoCs, where HDMI output was disable
even when Linux was booting.
Fixes: ea9dc35aab ("reset: Get the RESET by index without device")
Reported-by: B1oHazard <ty3uk@mail.ua>
Signed-off-by: Neil Armstrong <narmstrong@baylibre.com>
With this extended test, we get the following failure :
=> ut dm reset_base
Test: dm_test_reset_base: reset.c
test/dm/reset.c:52, dm_test_reset_base(): reset_method3.id == reset_method3_1.id: Expected 0x14 (20), got 0x2 (2)
Test: dm_test_reset_base: reset.c (flat tree)
test/dm/reset.c:52, dm_test_reset_base(): reset_method3.id == reset_method3_1.id: Expected 0x14 (20), got 0x2 (2)
Failures: 2
A fix is needed in reset_get_by_index_nodev() when introduced in [1].
[1] ea9dc35aab ("reset: Get the RESET by index without device")
Signed-off-by: Neil Armstrong <narmstrong@baylibre.com>
Use of strcat() against an uninitialized buffer would lead
to buffer overflow. This patch fixes it.
Fixes: 694cd5618c ("IOMUX: Introduce iomux_replace_device()")
Signed-off-by: Yuichiro Goto <goto@k-tech.co.jp>
Cc: Peter Robinson <pbrobinson@gmail.com>
Cc: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Cc: Nicolas Saenz Julienne <nsaenzjulienne@suse.de>
Reviewed-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Tested-by: Peter Robinson <pbrobinson@gmail.com>
As reported by Coverity Scan for Das U-Boot, the 'less-than-zero'
comparison of an unsigned value is never true.
Signed-off-by: Dario Binacchi <dariobin@libero.it>
Reviewed-by: Pratyush Yadav <p.yadav@ti.com>
When CONFIG_USE_DEFAULT_ENV_FILE=y and the file
CONFIG_DEFAULT_ENV_FILE is empty (or at least doesn't contain any
non-comment, non-empty lines), we end up feeding nothing into xxd,
which in turn then outputs nothing. Then blindly appending ", 0x00"
means that we end up trying to compile (roughly)
const char defaultenv[] = { , 0x00 }
which is of course broken.
To fix that, change the frobbing of the text file so that we always
end up printing an extra empty line (which gets turned into that extra
nul byte we need) - that corresponds better to the binary format
consisting of a series of key=val nul terminated strings, terminated
by an empty string.
Reported-by: Oleksandr Suvorov <oleksandr.suvorov@toradex.com>
Signed-off-by: Rasmus Villemoes <rasmus.villemoes@prevas.dk>
Reviewed-by: Oleksandr Suvorov <oleksandr.suvorov@toradex.com>
There are some cases where decompressed sectors can have padding zeros.
In kernel code, we have lines to address such situation:
/*
* btrfs_getblock is doing a zero on the tail of the page too,
* but this will cover anything missing from the decompressed
* data.
*/
if (bytes < destlen)
memset(kaddr+bytes, 0, destlen-bytes);
kunmap_local(kaddr);
But not in U-boot code, thus we have some reports of U-boot failed to
read compressed files in btrfs.
Fix it by doing the same thing of the kernel, for both inline and
regular compressed extents.
Reported-by: Matwey Kornilov <matwey.kornilov@gmail.com>
Link: https://bugzilla.suse.com/show_bug.cgi?id=1183717
Fixes: a26a6bedaf ("fs: btrfs: Introduce btrfs_read_extent_inline() and btrfs_read_extent_reg()")
Signed-off-by: Qu Wenruo <wqu@suse.com>
When zImage support was added to SPL, the messages were hidden to reduce
code size. However, the wrong config symbol was used. Since this file is
only built when CONFIG_SPL_FRAMEWORK=y, the messages were always hidden.
Use the correct symbol so the messages are printed in U-Boot proper.
Also use IS_ENABLED to drop the #ifdef.
Fixes: 431889d6ad ("spl: zImage support in Falcon mode")
Signed-off-by: Samuel Holland <samuel@sholland.org>
When flush_cache() is called during boot on our ~7M kernel image, the
hundreds of thousands of WATCHDOG_RESET calls end up adding
significantly to boottime. Flushing a single cache line doesn't take
many microseconds, so doing these calls for every cache line is
complete overkill.
The generic watchdog_reset() provided by wdt-uclass.c actually
contains some rate-limiting logic that should in theory mitigate this,
but alas, that rate-limiting must be disabled on powerpc because of
its get_timer() implementation - get_timer() works just fine until
interrupts are disabled, but it just so happens that the "big"
flush_cache() call happens in the part of bootm where interrupts are
indeed disabled. [1] [2] [3]
I have checked with objdump that the generated code doesn't change
when this option is left at its default value of 0: gcc is smart
enough to see that the ">=" comparison is tautologically true, hence
all assignments to "flushed" are eliminated as dead stores.
On our board, setting the option to something like 65536 ends up
reducing total boottime by about 0.8 seconds.
[1] https://patchwork.ozlabs.org/project/uboot/patch/20200605111657.28773-1-rasmus.villemoes@prevas.dk/
[2] https://lists.denx.de/pipermail/u-boot/2021-April/446906.html
[3] https://lists.denx.de/pipermail/u-boot/2021-April/447280.html
Reviewed-by: Stefan Roese <sr@denx.de>
Signed-off-by: Rasmus Villemoes <rasmus.villemoes@prevas.dk>
CONFIG_5xx hasn't existed since commit 5025897774 (powerpc, 5xx:
remove support for 5xx). Remove this last mention of it.
Reviewed-by: Stefan Roese <sr@denx.de>
Signed-off-by: Rasmus Villemoes <rasmus.villemoes@prevas.dk>
Having WATCHDOG_RESET() called automatically from the timer interrupt
runs counter to the idea of a watchdog device - if the board runs into
an infinite loops with interrupts still enabled, the watchdog will
never fire.
When using CONFIG_(SPL_)WDT, the watchdog_reset function is a lot more
complicated than just poking a few SOC-specific registers - it
involves accessing all kinds of global data, and if the interrupt
happens at the wrong time (say, in the middle of an WATCHDOG_RESET()
call from ordinary code), that can end up corrupting said global data.
Allow the board to opt out of calling WATCHDOG_RESET() from the timer
interrupt handler by setting CONFIG_SYS_WATCHDOG_FREQ to 0 - as that
setting is currently nonsensical (it would be compile-time
divide-by-zero), it cannot affect any existing boards.
Add documentation for both the existing and extended meaning of
CONFIG_SYS_WATCHDOG_FREQ.
Signed-off-by: Rasmus Villemoes <rasmus.villemoes@prevas.dk>
