Chuanhua Han [Fri, 2 Aug 2019 08:53:53 +0000 (16:53 +0800)]
armv8: kconfig: Fix some platforms incorrect I2C clock divider
By default, i2c input clock is platform clk / 2, but some of the
platform of i2c clock divider does not meet this kind of circumstance,
so alone to set default values for these platforms.
Signed-off-by: Chuanhua Han <chuanhua.han@nxp.com> Reviewed-by: Prabhakar Kushwaha <prabhakar.kushwaha@nxp.com>
Chuanhua Han [Thu, 1 Aug 2019 08:36:57 +0000 (16:36 +0800)]
armv8: ls1088aqds: support DSPI mode by hwconfig
BRDCFG4[USBOSC] and BRDCFG5[SPR] register field of Qixis device is used
to control SPI and other IP signal routing.
USBOSC:
0= SPI_CLK used as external USB REFCLK input driven with 24.000 MHz.
SPI devices are unusable in this mode.
1= SPI_CLK used as SPI clock.
SPI devices are usable in this mode. USB block is clocked from
internal sources
SPR[3:2]:
SPI_CS / SDHC_DAT4:7 Routing (schematic net CFG_SPI_ROUTE[3:2]):
00= SDHC/eMMC 8-bit
01= SD Card Rev 2.0/3.0
10= SPI on-board memory
11= TDM Riser / SPI off-board connector.
The default value is 00 if an SDCard/eMMC card is selected as the boot
device.
Alison Wang [Mon, 22 Jul 2019 07:17:21 +0000 (07:17 +0000)]
armv8: ls1046afrwy: Define CONFIG_ENV_ADDR for QSPI Boot
Defines CONFIG_ENV_ADDR for QSPI Boot which specifies the start
address of the flash sector containing the environment. It fixes
the issue that bootcmd is always set as default at bootup.
Signed-off-by: Alison Wang <alison.wang@nxp.com> Reviewed-by: Prabhakar Kushwaha <prabhakar.kushwaha@nxp.com>
Chuanhua Han [Tue, 23 Jul 2019 10:43:11 +0000 (18:43 +0800)]
boards: ls1088a: Add support of I2C driver model
DM_I2C_COMPAT is a compatibility layer that allows using the non-DM
I2C API when DM_I2C is used.When DM_I2C_COMPAT is not enabled for
compilation, a compilation error will be generated. This patch
solves the problem that the i2c-related api of the ls1088a platform
does not support dm.
Signed-off-by: Chuanhua Han <chuanhua.han@nxp.com> Reviewed-by: Prabhakar Kushwaha <prabhakar.kushwaha@nxp.com>
Chuanhua Han [Mon, 22 Jul 2019 08:36:42 +0000 (16:36 +0800)]
boards: ls2088a: Add support of I2C driver model.
DM_I2C_COMPAT is a compatibility layer that allows using the non-DM
I2C API when DM_I2C is used.When DM_I2C_COMPAT is not enabled for
compilation, a compilation error will be generated. This patch
solves the problem that the i2c-related api of the ls2088a platform
does not support dm.
Signed-off-by: Chuanhua Han <chuanhua.han@nxp.com> Reviewed-by: Prabhakar Kushwaha <prabhakar.kushwaha@nxp.com>
Chuanhua Han [Wed, 10 Jul 2019 07:48:37 +0000 (15:48 +0800)]
armv8: ls1028aqds: Remove the definition of CONFIG_SYS_I2C_EARLY_INIT
Since i2c uses dm mode, i2c controller will be initialized when reading
and writing devices on i2c bus. So there is no need for the original
non-dm mode i2c early initialization function call, this patch removed
the definition of CONFIG_SYS_I2C_EARLY_INIT.
Signed-off-by: Chuanhua Han <chuanhua.han@nxp.com> Reviewed-by: Alex Marginean <alexm.osslist@gmail.com> Reviewed-by: Prabhakar Kushwaha <prabhakar.kushwaha@nxp.com>
Chuanhua Han [Wed, 10 Jul 2019 13:16:49 +0000 (21:16 +0800)]
boards: ls1028a: Add support of I2C driver model
DM_I2C_COMPAT is a compatibility layer that allows using the non-DM
I2C API when DM_I2C is used.When DM_I2C_COMPAT is not enabled for
compilation, a compilation error will be generated. This patch
solves the problem that the i2c-related api of the ls1028a platform
does not support dm.
Signed-off-by: Chuanhua Han <chuanhua.han@nxp.com> Reviewed-by: Alex Marginean <alexm.osslist@gmail.com> Reviewed-by: Prabhakar Kushwaha <prabhakar.kushwaha@nxp.com>
In file included from include/config.h:8:0,
from include/common.h:20:
include/config_fallbacks.h:51:4: error: #error "Cannot define
CONFIG_SYS_I2C when CONFIG_DM_I2C is used"
# error "Cannot define CONFIG_SYS_I2C when CONFIG_DM_I2C is used"
^~~~~
In file included from include/config.h:8:0,
from include/common.h:20:
include/config_fallbacks.h:51:4: error: #error "Cannot define
CONFIG_SYS_I2C when CONFIG_DM_I2C is used"
# error "Cannot define CONFIG_SYS_I2C when CONFIG_DM_I2C is used"
^~~~~
board/freescale/lx2160a/lx2160a.c: In function 'board_early_init_f':
board/freescale/lx2160a/lx2160a.c:108:2: warning: implicit declaration
of function 'i2c_early_init_f'; did you mean 'arch_early_init_r'?
[-Wimplicit-function-declaration]
i2c_early_init_f();
^~~~~~~~~~~~~~~~
arch_early_init_r
drivers/i2c/mxc_i2c.c: In function 'mxc_i2c_probe':
drivers/i2c/mxc_i2c.c:824:8: warning: implicit declaration of function
'enable_i2c_clk';
did you mean 'enable_irq_wake'? [-Wimplicit-function-declaration]
ret = enable_i2c_clk(1, bus->seq);
^~~~~~~~~~~~~~
enable_irq_wake
So fix these compilation errors.
Signed-off-by: Chuanhua Han <chuanhua.han@nxp.com> Reviewed-by: Prabhakar Kushwaha <prabhakar.kushwaha@nxp.com>
Chuanhua Han [Wed, 10 Jul 2019 13:00:20 +0000 (21:00 +0800)]
boards: lx2160a: Add support of I2C driver model
DM_I2C_COMPAT is a compatibility layer that allows using the non-DM I2C
API when DM_I2C is used. When DM_I2C_COMPAT is not enabled for
compilation, a compilation error will be generated. This patch solves
the problem that the i2c-related api of the lx2160a platform does not
support dm.
Signed-off-by: Chuanhua Han <chuanhua.han@nxp.com> Reviewed-by: Prabhakar Kushwaha <prabhakar.kushwaha@nxp.com>
Yuantian Tang [Wed, 19 Jun 2019 06:39:28 +0000 (14:39 +0800)]
common: qixis: make the qixis compatible with new soc
This driver needs modification to work with new soc,
like ls1028, since bitmap of RCFG is changed to
RESV[7:5] LIVE[4] WDEN[3] RESV[2:1] GO[0]
000 1 0 00 0
Also the RCW location is moved to only dutcfg0.
RESV[7:4] RCWSRC[3:0]
1111 configurable
Following commands are functional now
qixis_reset
qixis_reset sd
qixis_reset qspi
qixis_reset emmc
MC firmware need to be aligned to 512M, so minimum 512MB DDR is reserved.
But MC support to work with 128MB or 256MB DDR memory also, in this
case, rest of the memory is not usable.
So reporting this extra memory to Linux through dtb memory fixup.
Michal Simek [Mon, 19 Aug 2019 09:06:13 +0000 (11:06 +0200)]
test/py: Add cmd_memory dependency back to test_mmc_wr
Based on discussion with Stephen Warren there was recommendation to list
both memory and random command dependencies just in case that dependency is
not properly handled by Kconfig.
Fixes: a09c1f7e1c1b ("test/py: Fix MMC/SD block write test dependency") Signed-off-by: Michal Simek <michal.simek@xilinx.com> Acked-by: Stephen Warren <swarren@nvidia.com>
Building easylogo with `HOST_TOOLS_ALL=y make tools` results in a build
warning due to a possible buffer overrun:
tools/easylogo/easylogo.c:453:4: note: ‘sprintf’ output between 7 and
262 bytes into a destination of size 256
sprintf (str, "%s, 0x%02x", app, *dataptr++);
^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Truncate the output to fit into the destination buffer.
Signed-off-by: Heinrich Schuchardt <xypron.glpk@gmx.de>
Looking at the contents of file .mailmap it seems that some editors assumed
that translation is done by entering multiple lines into the file and the
last one replaces the others. This is not how it works. The translation
occurs according to entries in single lines as described in the
git-check-mailmap man-page.
Add a description of the file format.
Add an entry for Alexander Graf as his old email address is not valid
anymore.
Signed-off-by: Heinrich Schuchardt <xypron.glpk@gmx.de>
README: Clarify use of BSS during SPL board_init_f()
The earlier commit....
commit a5a5d997b41a ("spl: Allow performing BSS init early before board_init_f()")
...introduced the ability to use BSS from SPL's board_init_f() as it may
be required in certain exceptional use cases so go ahead and update the
README to reflect this change. Note that as highlighted with the changes
the use of the associated CONFIG option is generally not recommended.
Signed-off-by: Andreas Dannenberg <dannenberg@ti.com> Reviewed-by: Simon Goldschmidt <simon.k.r.goldschmidt@gmail.com>
Suniel Mahesh [Mon, 19 Aug 2019 06:27:39 +0000 (11:57 +0530)]
arm: omap2: am43xx: Enable CONFIG_BLK
With DM_MMC enabled, enable CONFIG_BLK to remove this
compile warning for am43xx based targets:
===================== WARNING ======================
This board does not use CONFIG_DM_MMC. Please update
the board to use CONFIG_DM_MMC before the v2019.04 release.
Failure to update by the deadline may result in board removal.
See doc/driver-model/MIGRATION.txt for more info.
====================================================
The commit 1b42ab3eda8a ("ARM: DRA7: Fixup DSPEVE, IVA and GPU clock
frequencies based on OPP") updates the kernel device-tree blob to adjust
the DSP, IVA and GPU DPLL clocks based on a one-time OPP choice selected
in U-Boot. All these DPLL clocks are children of the cm_core_aon clocks
DT node.
The hierarchy of this clocks DT node has changed in newer Linux kernels
starting from v5.0, and this results in a failure in ft_fixup_clocks()
function to update the clock rates on these newer kernels. Fix this by
updating the lookup logic to look through both the newer and older
DT hierarchy paths for the cm_core_aon clocks node.
Adam Ford [Tue, 13 Aug 2019 19:38:11 +0000 (14:38 -0500)]
ARM: davinci: Remove duplicated references
The Kconfig file calls to ti/common/Kconfig twice which makes
several of the menu items repeat themselves. In an effort to
clean this up, this patch removes the second call which eliminates
the duplicate menu items.
Adam Ford [Tue, 13 Aug 2019 13:28:28 +0000 (08:28 -0500)]
ARM: da850evm: Remove dead code
Now that SPL supports DM_SERIAL and the direct NOR boot supports
DM_SERIAL, the check to see if DM_SERIAL is defined can go away,
because all da850evm variants now support DM_SERIAL. This patch
simply removes some dead precompiler defines.
Adam Ford [Tue, 13 Aug 2019 13:27:28 +0000 (08:27 -0500)]
ARM: da850evm: Remove dead SPI Code
With SPL now supporting DM_SPI, the need for compiler directives
and hard-coded addresses is obsolete. This patch removes some
dead legacy code defining the SPI base address
Adam Ford [Tue, 13 Aug 2019 19:04:02 +0000 (14:04 -0500)]
Revert "ARM: da850-evm: Enable SPI Flash and NAND Flash when booting NOR"
Sorry for the noise, but there appears to be a regression with older
hardware. Since it broke the direct_nor boot option, it should be
reverted until a better solution is available.
Andrew F. Davis [Mon, 12 Aug 2019 19:59:55 +0000 (15:59 -0400)]
configs: Rename environment variable fit_bootfile to name_fit
Like we did with 'fit_loadaddr' to 'addr_fit', the variable
'fit_bootfile' contains a name and so should be prefixed with
name_. Make this change here.
Andrew F. Davis [Mon, 12 Aug 2019 19:59:54 +0000 (15:59 -0400)]
configs: Rename environment variable fit_loadaddr to addr_fit
This is the first part of a larger effort I would like to propose to
unify and simplify the default set of environment variables.
When many early environment variables were named there were fewer images
being loaded, usually just a kernel. At this time names like 'loadaddr'
would suffice. Now we have more images and many more commands that act on
them, often re-using the same variable for several different uses. The
contents of a variable are also not immediately known causing one to have
to look up a chain of variables to understand what a command is actually
doing. I suggest the following.
To start, all variables containing names should be prefixed with name_
and addresses with addr_. This is like how K2 already does things and
allows for simple universal commands like:
get_fdt_nfs=nfs ${addr_fdt} /boot/${name_fdt}
Which is very clear on what is intended here and would work across all
board that using the this naming convention.
We can do this one variable at a time, start here with addr_fit.
Signed-off-by: Andrew F. Davis <afd@ti.com> Acked-by: Andreas Dannenberg <dannenberg@ti.com>
Yegor Yefremov [Fri, 9 Aug 2019 05:21:57 +0000 (07:21 +0200)]
arm: baltos: use device tree alias to access Ethernet slave
The full path has changed in the recent kernels so that it is
not possible to load them. Aliases "ethernet0" and "ethernet1"
are still present in both legacy and new kernels.
Also, fix error messages to correspond to the taken actions.
Tom Rini [Mon, 19 Aug 2019 13:22:57 +0000 (09:22 -0400)]
Merge tag 'u-boot-rockchip-20190819' of https://gitlab.denx.de/u-boot/custodians/u-boot-rockchip
- Add ROC-RK3399-PC board support
- Move CONFIG_SPI_FLASH_GIGADEVICE and CONFIG_CMD_USB_MASS_STORAGE to
Kconfig
- using SYSRESET_POWER_OFF for poweroff
(Note that patch for rk8xx pmic is droped for it can not pass Travis
build)
- fix ofnode_get_name() assert
Urja Rannikko [Thu, 16 May 2019 21:48:42 +0000 (21:48 +0000)]
sysreset: move stm32mp sysreset poweroff implementation to sysreset uclass
This is a generic implementation. Add CONFIG_SYSRESET_CMD_POWEROFF
to signal when we need it. Enable it from the STPMIC1 config and in
sandbox.
The config flag is transitionary, that is it can be removed after all
poweroff implementations use sysreset, and just have CMD_POWEROFF depend
on sysreset.
Signed-off-by: Urja Rannikko <urjaman@gmail.com> Reviewed-by: Patrice Chotard <patrice.chotard@st.com> Reviewed-by: Patrick Delaunay <patrick.delaunay@st.com> Tested-by: Patrick Delaunay <patrick.delaunay@st.com>
Urja Rannikko [Thu, 16 May 2019 21:48:41 +0000 (21:48 +0000)]
sysreset: switch to using SYSRESET_POWER_OFF for poweroff
It seems that SYSRESET_POWER_OFF was added recently, and all previous code
used SYSRESET_POWER for poweroff. SYSRESET_POWER is supposed to be a
PMIC-level power cycle, not a poweroff.
(Comment by Simon Glass)
SYSRESET_POWER means to do a power reset (removing and reinstating all power)
SYSRESET_POWER_OFF means to turn the device off and leave it off
Signed-off-by: Urja Rannikko <urjaman@gmail.com> Reviewed-by: Patrick Delaunay <patrick.delaunay@st.com>
(Update comment to help understand the patch) Signed-off-by: Kever Yang <kever.yang@rock-chips.com>
Urja Rannikko [Mon, 13 May 2019 13:51:05 +0000 (13:51 +0000)]
configs: update rk3288 veyron defconfigs
Updates jerry, mickey, minnie and speedy defconfigs to:
- fit the SPL in 32k
- boot from SPI (only)
- remove gadget support (these have no OTG port)
Reviewed-by: Simon Glass <sjg@chromium.org> Signed-off-by: Urja Rannikko <urjaman@gmail.com> Reviewed-by: Kever Yang <kever.yang@rock-chips.com>
(Rebase on top of tree) Signed-off-by: Kever Yang <kever.yang@rock-chips.com>
Urja Rannikko [Mon, 13 May 2019 13:51:04 +0000 (13:51 +0000)]
configs: Move CONFIG_CMD_USB_MASS_STORAGE properly into Kconfig
This affects RK3036, RK322X and RK3288 - the defconfig changes done by
moveconfig.py for the veyrons were left out on purpose because they dont
have an OTG port, and will get their config updated in the next commit.
Signed-off-by: Urja Rannikko <urjaman@gmail.com> Reviewed-by: Kever Yang <kever.yang@rock-chips.com>
Kever Yang [Fri, 19 Jul 2019 03:23:47 +0000 (11:23 +0800)]
core: ofnode: do not assert if node not valid in ofnode_get_name()
In some case with LIVE DT, some node always not valid, or not have
a valid name, eg. blk driver add by mmc.
Return fail instead of Assert for this kind of ofnode, and this
help with assert happen from time to time when of_live is enabled
and DEBUG is enabled.
Signed-off-by: Kever Yang <kever.yang@rock-chips.com> Reviewed-by: Simon Glass <sjg@chromium.org>
Levin Du [Fri, 26 Jul 2019 07:43:54 +0000 (15:43 +0800)]
rockchip: rk3399: Add ROC-RK3399-PC support
Add initial support for ROC-RK3399-PC board.
Specification
- Rockchip RK3399
- LPDDR4 4GiB
- eMMC slot
- SD card slot
- RTL8211E 1Gbps
- HDMI Out, DP, MIPI DSI/CSI, EDP
- PCIe M.2
- USB 2.0, USB-3.0
- USB C Type
Commit details of rk3399-roc-pc.dts sync from Linux v5.2:
"arm64: dts: rockchip: add support for ROC-RK3399-PC board"
(sha1: 8bb878cf20ae10809c36db96993bfce7026d062b)
Signed-off-by: Levin Du <djw@t-chip.com.cn> Reviewed-by: Kever Yang <kever.yang@rock-chips.com>
Stefan Roese [Fri, 16 Aug 2019 12:45:29 +0000 (14:45 +0200)]
global_data: Remove comment of reserved arch-specific GD flags
With the removal of the x86 specific GD flags, there are no arch-
specific GD flags any more. Let's remove the comment about reserving the
upper 16 bits for arch-specific flags in the common header. This gives
us more flexibility with the usage of the GD flags.
As a matter of fact, we are already using more than 16 bits for common
GD flags (with the addition of GD_FLG_WDT_READY).
Signed-off-by: Stefan Roese <sr@denx.de> Cc: Bin Meng <bmeng.cn@gmail.com> Cc: Simon Glass <sjg@chromium.org> Cc: Tom Rini <trini@konsulko.com> Cc: Simon Goldschmidt <simon.k.r.goldschmidt@gmail.com> Reviewed-by: Bin Meng <bmeng.cn@gmail.com> Reviewed-by: Simon Goldschmidt <simon.k.r.goldschmidt@gmail.com> Reviewed-by: Simon Glass <sjg@chromium.org>
Stefan Roese [Fri, 16 Aug 2019 12:45:28 +0000 (14:45 +0200)]
x86: Remove x86 specific GD flags as they are not referenced at all
This patch removes the x86 architecture specific GD flags
(GD_FLG_COLD_BOOT & GD_FLG_WARM_BOOT), as they are not used. Only
GD_FLG_COLD_BOOT is referenced in coreboot.c but assigned in start16.S.
But the coreboot target does not use start16.S at all and boots directly
from the 32-bit start code.
Signed-off-by: Stefan Roese <sr@denx.de> Cc: Bin Meng <bmeng.cn@gmail.com> Cc: Simon Glass <sjg@chromium.org> Cc: Tom Rini <trini@konsulko.com> Cc: Simon Goldschmidt <simon.k.r.goldschmidt@gmail.com> Reviewed-by: Bin Meng <bmeng.cn@gmail.com> Tested-by: Bin Meng <bmeng.cn@gmail.com> Reviewed-by: Simon Goldschmidt <simon.k.r.goldschmidt@gmail.com>
Simon Glass [Thu, 15 Aug 2019 01:56:14 +0000 (19:56 -0600)]
cbfs: Add functions to support multiple CBFSs
Sometimes an image has multiple CBFS. The current CBFS API is limited to
handling only one at time. Also it keeps track of the CBFS internally in
BSS, which does not work before relocation, for example.
Add a few new functions to overcome these limitations.
Signed-off-by: Simon Glass <sjg@chromium.org> Reviewed-by: Bin Meng <bmeng.cn@gmail.com> Tested-by: Bin Meng <bmeng.cn@gmail.com>
Tom Rini [Sat, 17 Aug 2019 14:31:25 +0000 (10:31 -0400)]
Merge tag 'efi-2019-10-rc3' of https://gitlab.denx.de/u-boot/custodians/u-boot-efi
Pull request for UEFI sub-system for v2019.10-rc3
This pull request provides corrections for the SetVirtualAddress runtime
service and avoids possible calls to NULL by consumers of the
EFI_PXE_BASE_CODE_PROTOCOL.
Commit 7f95104d91cc ("efi_loader: detach runtime in ExitBootServices()")
added a call to efi_runtime_detach() to ExitBootServices() but did not
remove the call in SetVirtualAddressMap().
Remove the superfluous function call.
Correct a comment referring to efi_runtime_detach().
Fixes: 7f95104d91cc ("efi_loader: detach runtime in ExitBootServices()") Signed-off-by: Heinrich Schuchardt <xypron.glpk@gmx.de>
U-Boot implements the EFI_PXE_BASE_CODE_PROTOCOL because GRUB uses the mode
information for booting via PXE. All function pointers in the protocol were
NULL up to now which will cause immediate crashes when the services of the
protocol are called.
Create function stubs for all services of the protocol returning
EFI_UNSUPPORTED.
Signed-off-by: Heinrich Schuchardt <xypron.glpk@gmx.de>
arm: socfpga: gen5: don't zero bss in board_init_f()
The socfpga gen5 SPL manually zeroed bss in board_init_f(). Now that the
DDR driver does not use bss any more, bss is not used before board_init_r()
and we can remove this hack.
bss is normally zeroed by crt0.S, but after board_init_f(), before
board_init_r(). socfpga just had this double-zeroing because it invalidly
used bss in board_init_f() already (during DDR initialization).
Signed-off-by: Simon Goldschmidt <simon.k.r.goldschmidt@gmail.com>
Marcus Comstedt [Fri, 2 Aug 2019 17:45:15 +0000 (19:45 +0200)]
riscv: tools: Fix prelink-riscv to work on big endian hosts
All ELF fields whose values are inspected by the code are converted to
CPU byteorder first. Values which are copied verbatim (relocation
fixups) are not swapped to CPU byteorder and back as it is not needed.
Signed-off-by: Marcus Comstedt <marcus@mc.pp.se> Cc: Rick Chen <rick@andestech.com> Reviewed-by: Rick Chen <rick@andestech.com>
doc: sifive-fu540: Update README to explicitly load DTB for Linux
We should explicitly load DTB from TFTP server or MMC/SD card
for Linux booting. This will allow us:
1. To use different Linux DTB for SiFive Unleashed board with
expansion board connected.
2. Avoid re-flashing OpenSBI firmware whenever board connections
change.
This patch updates reference bootlog in SiFive FU540 README
as-per above.
Signed-off-by: Anup Patel <anup.patel@wdc.com> Reviewed-by: Bin Meng <bmeng.cn@gmail.com> Reviewed-by: Joe Hershberger <joe.hershberger@ni.com>
Bin Meng [Thu, 11 Jul 2019 06:43:13 +0000 (23:43 -0700)]
riscv: Access CSRs using CSR numbers
We should prefer accessing CSRs using their CSR numbers
because:
1. It compiles fine with older toolchains.
2. We can use latest CSR names in #define macro names of CSR
numbers as-per RISC-V spec.
3. We can access newly added CSRs even if toolchain does not
recognize newly added CSRs by name.
This commit is inspired from Linux kernel commit a3182c91ef4e
("RISC-V: Access CSRs using CSR numbers").
Signed-off-by: Bin Meng <bmeng.cn@gmail.com> Reviewed-by: Rick Chen <rick@andestech.com> Reviewed-by: Anup Patel <anup.patel@wdc.com> Reviewed-by: Lukas Auer <lukas.auer@aisec.fraunhofer.de>
riscv : serial: use rx watermark to indicate rx data is present
In y-modem transfer mode, tstc/getc fail to check if there is any
data available / received in RX FIFO, and so y-modem transfer never
succeeds. Using receive watermark bit within ip register fixes the
issue.
This patch is based on commit c7392b7bc4e1 ("Use the RX watermark
interrupt pending bit for TSTC") available at[1]
projects / u-boot.git / log