Marek Vasut [Sun, 6 Oct 2019 14:13:38 +0000 (16:13 +0200)]
usb: ehci-hcd: Keep async schedule running
Profiling the EHCI driver shows a significant performance problem in
ehci_submit_async(). Specifically, this function keeps enabling and
disabling async schedule back and forth for every single transaction.
However, enabling/disabling the async schedule does not take effect
immediatelly, but instead may take up to 1 mS (8 uFrames) to complete.
This impacts USB storage significantly, esp. since the recent reduction
of maximum transfer size to support more USB storage devices. This in
turn results in sharp increase in the number of ehci_submit_async()
calls. Since one USB storage BBB transfer does three such calls and
the maximum transfer size is 120 kiB, the overhead is 6 mS per 120 kiB,
which is unacceptable.
However, this overhead can be removed simply by keeping the async
schedule running. Specifically, the first transfer starts the async
schedule and then each and every subsequent transfer only adds a new
QH into that schedule, waits until the QH is completed and does NOT
disable the async schedule. The async schedule is stopped only by
shutting down the controller, which must happen before moving out
of U-Boot, otherwise the controller will corrupt memory.
Signed-off-by: Marek Vasut <marek.vasut+renesas@gmail.com> Cc: Bin Meng <bmeng.cn@gmail.com> Cc: Simon Glass <sjg@chromium.org>
Marek Vasut [Sat, 5 Oct 2019 17:18:38 +0000 (19:18 +0200)]
usb: storage: Only clear READY flag on error
Clear the USB_READY flag in the storage driver only in case there
is an error, otherwise usb_stor_BBB_transport() waits 5 mS before
doing anything every single time.
This is because the USB_READY flag is only ever set in
usb_test_unit_ready(), which is called only upon storage device
probe, not between each and every request. However, the device
cannot move out of USB_READY state once it was initialized.
Signed-off-by: Marek Vasut <marek.vasut+renesas@gmail.com> Cc: Bin Meng <bmeng.cn@gmail.com> Cc: Simon Glass <sjg@chromium.org>
usb: storage: scsiglue: further describe our 240 sector limit
Just so we have some sort of documentation as to why
we limit our Mass Storage transfers to 240 sectors,
let's update the comment to make clearer that
devices were found that would choke with larger
transfers.
While at that, also make sure to clarify that other
operating systems have similar, albeit different,
limits on mass storage transfers.
And reduce the maximum transfer length of USB storage to 120 kiB.
Signed-off-by: Marek Vasut <marek.vasut+renesas@gmail.com> Reviewed-by: Bin Meng <bmeng.cn@gmail.com> Cc: Bin Meng <bmeng.cn@gmail.com> Cc: Simon Glass <sjg@chromium.org>
Patrick Delaunay [Mon, 14 Oct 2019 07:28:04 +0000 (09:28 +0200)]
dfu: add backend for MTD device
Add DFU backend for MTD device: allow to read
and write on all MTD device (NAND, SPI-NOR,
SPI-NAND,...)
For example :
> set dfu_alt_info "nand_raw raw 0x0 0x100000"
> dfu 0 mtd nand0
This MTD backend provides the same level than dfu nand
backend for NAND and dfu sf backend for SPI-NOR;
So it can replace booth of them but it also
add support of spi-nand.
> set dfu_alt_info "nand_raw raw 0x0 0x100000"
> dfu 0 mtd spi-nand0
The backend code is based on the "mtd" command
introduced by commit 5db66b3aee6f ("cmd: mtd:
add 'mtd' command")
Signed-off-by: Patrick Delaunay <patrick.delaunay@st.com>
Patrick Delaunay [Mon, 14 Oct 2019 07:28:03 +0000 (09:28 +0200)]
dfu: allow read with no data without error for EOF indication
This patch allows the DFU backend to indicate that that it can't
provide no more data to fill the DFU buffer, by setting b_left =0
without error, even if the size of received data is lower of the
expected total size indicated by get_medium_size.
For USB DFU stack point of view, it is acceptable:
the read length < requested size in DFU_UPLOAD and the
transaction is stopped.
That avoid infinite loop issue in dfu_read_buffer_fill because the
size for the DFU read is limited by get_medium_size = r_left
and the DFU stack expects that read is allowed up to this size.
This issue never occurs for current flash device (where chunk are
always completely read, and b_left will be never 0) but it is useful for
virtual partition when the backend only know the max size of this
alternate, the real size of the data are only known in the read
treatment.
PS: for file access on mmc, EOF is never reached as
dfu_get_medium_size_mmc returns the exact size of the file.
Signed-off-by: Patrick Delaunay <patrick.delaunay@st.com>
Patrick Delaunay [Mon, 14 Oct 2019 07:28:00 +0000 (09:28 +0200)]
dfu: sf: add partition support for nor backend
Copy the partition support from NAND backend to SF,
support part and partubi option.
In case of ubi partition, erase the rest of the
partition as it is mandatory for UBI.
The added code is under compilation flag CONFIG_DFU_SF_PART
activated by default.
for example:
U-Boot> env set dfu_alt_info "spl part 0 1;\
u-boot part 0 2;u-boot-env part 0 3;UBI partubi 0 4"
U-Boot> dfu 0 sf 0:0:10000000:0
Signed-off-by: Patrick Delaunay <patrick.delaunay@st.com>
Add possibility to define a part of partition as a separate DFU entity.
This allows to have more than one items on the given partition.
The real use case for this option is TM2 board. It can use u-boot stored
as Linux kernel on the defined partition (as RAW data) and load the real
kernel from the same partition, but stored under the certain offset.
Signed-off-by: Marek Szyprowski <m.szyprowski@samsung.com> Acked-by: Lukasz Majewski <lukma@denx.de>
cmd: thor: select DFU subsystem also for 'thor' download tool
'THOR' download command requires DFU infrastructure to properly flash
board images. It can be used without enabling DFU command, so add such
dependency to Kconfig.
Signed-off-by: Marek Szyprowski <m.szyprowski@samsung.com> Acked-by: Lukasz Majewski <lukma@denx.de>
dwc3-generic: Don't fail probe if clk/reset entries are absent
Some boards don't populate clk/reset entries as these are are optional
as per binding documentation. Therefore, don't fail driver probe if
clk/reset entries are absent in DT.
This fixes fastboot failures seen due to enabling of CONFIG_CLK on AM57xx
Fixes: e8e683d33b0c ("board: ti: am57xx-idk: Configure the CDCE913 clock synthesizer") Reported-by: Sam Protsenko <semen.protsenko@linaro.org> Signed-off-by: Vignesh Raghavendra <vigneshr@ti.com>
Tom Rini [Wed, 30 Oct 2019 17:13:46 +0000 (13:13 -0400)]
Merge tag 'u-boot-clk-23Oct2019' of https://gitlab.denx.de/u-boot/custodians/u-boot-clk
- Add I2C clocks for i.MX6Q CCF driver
- Fix check in clk_set_default_parents()
- Managed API to get clock from device tree
- Fixes for core clock code (including sandbox regression tests)
Bin Meng [Mon, 28 Oct 2019 14:25:03 +0000 (07:25 -0700)]
Bring all testings in gitlab and travis CI to Azure Pipelines
This expands current Azure Pipelines Windows host tools build
testing to cover all the CI testing in gitlab and travis CI.
Note for some unknown reason, the 'container' cannot be used for
any jobs that have buildman, for buildman does not exit properly
and hangs the job forever. As a workaround, we manually call
docker to run the image to perform the CI tasks.
A complete run on Azure Pipelines takes about 2 hours and 10
minutes.
Bin Meng [Mon, 28 Oct 2019 14:25:01 +0000 (07:25 -0700)]
arm: mvebu: Avoid generating kwbimage.cfg in the source tree
At present some boards generate kwbimage.cfg in the source tree
during the build. This breaks buildman testing on some systems
where the source tree is read-only. Update makefile rules to
generate it in the build tree instead.
Note some other boards have the kwbimage.cfg file written in
advance, hence we need check if the file exists in the build
tree first, otherwise we fall back to one in the source tree.
Signed-off-by: Bin Meng <bmeng.cn@gmail.com> Reviewed-by: Chris Packham <judge.packham@gmail.com> Reviewed-by: Stefan Roese <sr@denx.de> Reviewed-by: Simon Glass <sjg@chromium.org>
Peng Fan [Wed, 23 Oct 2019 01:43:30 +0000 (01:43 +0000)]
spl: spl_mmc: fix getting raw_sect when boot from emmc boot partition
On i.MX8, when booting from eMMC boot partition, the whole flash.bin
is stored in boot partition, however SPL switches to user partition
during the init of mmc driver:
spl_mmc_load() -> mmc_init()
Then it tries to load the container image in
spl_mmc_get_uboot_raw_sector(), but here it reads the data from user
partition and the header is not recognized as a valid header.
So we move spl_mmc_get_uboot_raw_sector after eMMC partition switch
to address this issue.
Anyway put spl_mmc_get_uboot_raw_sector before eMMC partition switch
is not correct, so let's move it after eMMC partition switch.
Reported-by: Anatolij Gustschin <agust@denx.de> Signed-off-by: Peng Fan <peng.fan@nxp.com>
Yangbo Lu [Mon, 21 Oct 2019 10:09:09 +0000 (18:09 +0800)]
mmc: fsl_esdhc: drop i.MX DDR support code
A previous patch below adding DDR mode support was actually for i.MX
platforms. Now i.MX eSDHC driver is fsl_esdhc_imx.c. For QorIQ eSDHC,
it uses different process for DDR mode, and hasn't been supported.
Let's drop DDR support code for i.MX in fsl_esdhc driver.
Bin Meng [Sun, 27 Oct 2019 12:19:48 +0000 (05:19 -0700)]
Add Microsoft Azure Pipelines configuration
Microsoft Azure Pipelines [1] provides unlimited CI/CD minutes and
10 parallel jobs to every open source project for free.
This adds a configuration file for Azure Pipelines to utilize the
free Windows VM hosted by Microsoft to ensure no build broken in
building U-Boot host tools for Windows.
Bin Meng [Sun, 27 Oct 2019 12:19:47 +0000 (05:19 -0700)]
tools: Avoid creating symbolic links for tools/version.h
When building U-Boot host tools for Windows from Microsoft Azure
Pipelines, the following errors were seen:
HOSTCC tools/mkenvimage.o
In file included from tools/mkenvimage.c:25:
./tools/version.h:1:1: error: expected identifier or ‘(’ before ‘.’ token
1 | ../include/version.h
| ^
tools/mkenvimage.c: In function ‘main’:
tools/mkenvimage.c:117:4: warning: implicit declaration of function ‘usage’ [-Wimplicit-function-declaration]
117 | usage(prg);
| ^~~~~
tools/mkenvimage.c:120:35: error: ‘PLAIN_VERSION’ undeclared (first use in this function)
120 | printf("%s version %s\n", prg, PLAIN_VERSION);
| ^~~~~~~~~~~~~
tools/mkenvimage.c:120:35: note: each undeclared identifier is reported only once for each function it appears in
make[2]: *** [scripts/Makefile.host:114: tools/mkenvimage.o] Error 1
It turns out tools/version.h is a symbolic link and with Windows
default settings it is unsupported hence the actual content of
tools/version.h is not what file include/version.h has, but the
the linked file path, which breaks the build.
To fix this, remove the symbolic links for tools/version.h. Instead
we perform a copy from include/version.h during the build.
Bin Meng [Sun, 27 Oct 2019 12:19:46 +0000 (05:19 -0700)]
Add .gitattributes for line endings
When building U-Boot host tools for Windows from Microsoft Azure
Pipelines, we see tons of weird warnings and errors emitted from
every Kconfig files:
Kconfig:6:warning: ignoring unsupported character ''
Kconfig:6:warning: ignoring unsupported character ''
Kconfig:8:warning: ignoring unsupported character ''
Kconfig:9:warning: ignoring unsupported character ''
Kconfig:10:warning: ignoring unsupported character ''
Kconfig:10:warning: ignoring unsupported character ''
Kconfig:13:warning: ignoring unsupported character ''
arch/Kconfig:1:warning: ignoring unsupported character ''
arch/Kconfig:2:warning: ignoring unsupported character ''
arch/Kconfig:2:warning: ignoring unsupported character ''
arch/Kconfig:4:warning: ignoring unsupported character ''
...
After several rounds of experiments, it turns out this is caused
by line endings. Historically, Linux and macOS used linefeed (LF)
characters while Windows used a carriage return plus a linefeed
(CRLF). When Azure Pipelines checks out the U-Boot repo, Git tries
to compensate for the difference by automatically making lines end
in CRLF in the working directory on Windows, which confuses the
Kconfig file parsing logic.
Fortunately Git provides a way for repos to tell Git not to do such
automatical line endings conversion via .gitattributes file below:
Bin Meng [Sun, 27 Oct 2019 12:19:43 +0000 (05:19 -0700)]
linux/types.h: Surround 'struct ustat' with __linux__
'struct ustat' uses linux-specific typedefs to declare its memebers:
__kernel_daddr_t and __kernel_ino_t. It is currently not used by any
U-Boot codes, but when we build U-Boot tools for other platform like
Windows, this becomes a problem.
Bin Meng [Sun, 27 Oct 2019 12:19:42 +0000 (05:19 -0700)]
tools: zynqmpbif: Use compiler builtin instead of linux-specific __swab32
__swab32() is a Linux specific macro defined in linux/swab.h. Let's
use the compiler equivalent builtin function __builtin_bswap32() for
better portability.
Tom Rini [Sat, 26 Oct 2019 00:07:24 +0000 (20:07 -0400)]
Merge tag 'mips-pull-2019-10-25' of git://git.denx.de/u-boot-mips
- bmips: add BCRM NAND support for BCM6368, BCM6328, BCM6362 and BCM63268 SoCs
- bmips: various small fixes
- mtmips: add new drivers for clock, reset-controller and pinctrl
- mtmips: add support for high speed UART
- mtmips: update/enhance drivers for SPI and ethernet
- mtmips: add support for MMC
Suman Anna [Thu, 17 Oct 2019 03:33:08 +0000 (09:03 +0530)]
arm: dts: k3-am65: Add R5F ranges in interconnect nodes
Add the address spaces for the R5F cores in MCU domain to the ranges
property of the cbass_mcu interconnect node so that the addresses
within the R5F nodes can be translated properly by the relevant OF
address API.
Signed-off-by: Suman Anna <s-anna@ti.com> Signed-off-by: Lokesh Vutla <lokeshvutla@ti.com>
Lokesh Vutla [Mon, 7 Oct 2019 13:56:38 +0000 (19:26 +0530)]
armv7R: K3: j721e: Add support for triggering ddr init from SPL
In SPL, DDR should be made available by the end of board_init_f()
so that apis in board_init_r() can use ddr. Adding support for
triggering DDR initialization from board_init_f().
Kevin Scholz [Mon, 7 Oct 2019 13:56:36 +0000 (19:26 +0530)]
ram: k3-j721e: Add support for J721E DDR controller
The J721E DDR subsystem comprises DDR controller, DDR PHY and wrapper
logic to integrate these blocks in the device. The DDR subsystem is
used to provide an interface to external SDRAM devices which can be
utilized for storing program or data. Introduce support for the
DDR controller and DDR phy within the DDR subsystem.
Signed-off-by: Kevin Scholz <k-scholz@ti.com Signed-off-by: Lokesh Vutla <lokeshvutla@ti.com>
James Doublesin [Mon, 7 Oct 2019 08:34:27 +0000 (14:04 +0530)]
ram: k3-am654: Do not rely on default values for certain DDR register
Added the following registers to the DDR configuration:
- ACIOCR0,
- ACIOCR3,
- V2H_CTL_REG,
- DX8SLxDQSCTL.
Modified enable_dqs_pd and disable_dqs_pd to only touch the associated
bit fields for pullup and pulldown registers (to preserve slew rate and
other bits in that same register). Also update the dts files in the same
patch to maintain git bisectability.
Signed-off-by: James Doublesin <doublesin@ti.com> Signed-off-by: Lokesh Vutla <lokeshvutla@ti.com>
James Doublesin [Mon, 7 Oct 2019 08:34:25 +0000 (14:04 +0530)]
armv7r: dts: am654-base-board: Rename 1600MHz to 1600MTs in dtsi filename
The current configuration of DDR on AM654 base board is for 1600MTs but
the file name is specified as k3-am654-base-board-ddr4-1600MHz.dtsi.
Since 1600MHz is misleading, rename it to
k3-am654-base-board-ddr4-1600MTs.dtsi
Signed-off-by: James Doublesin <doublesin@ti.com> Signed-off-by: Lokesh Vutla <lokeshvutla@ti.com>
Lokesh Vutla [Mon, 7 Oct 2019 08:22:17 +0000 (13:52 +0530)]
arm: K3: Clean and invalidate Linux Image before jumping to Linux
U-Boot cleans and invalidate L1 and L2 caches before jumping to Linux
by set/way in cleanup_before_linux(). Additionally there is a custom
hook provided to clean and invalidate L3 cache.
Unfortunately on K3 devices(having a coherent architecture), there is no
easy way to quickly clean all the cache lines for L3. The entire address
range needs to be cleaned and invalidated by Virtual Address. This can
be implemented using the L3 custom hook but it take lot of time to clean
the entire address range. In the interest of boot time this might not be
a viable solution.
The best hit is to make sure the loaded Linux image is flushed so that
the entire image is written to DDR from L3. When Linux starts running with
caches disabled the full image is available from DDR.
Reported-by: Andrew F. Davis <afd@ti.com> Reported-by: Faiz Abbas <faiz_abbas@ti.com> Signed-off-by: Lokesh Vutla <lokeshvutla@ti.com>
zynqmp:
- Clean firmware handing via drivers/firmware/
- DT/defconfig name alignments
- DT cleanups with using firmware based clock driver
- Some defconfig updates
- Add IIO ina226 DT description
- Tune zynqmp_psu_init_minimalize.sh script
- Add single nand mini configuration, e-a2197, m-a2197-02/03 and zcu216
versal:
- Clean firmware handing via drivers/firmware/
- Add gpio support
- Enable DT overlay/USB/CLK/FPGA
- DT updates
- Tune mini configuration
dts: mtmips: add default pinctrl to eth nodes for all boards
This patch adds default eth pinctrl for all boards.
There are two pinctrl nodes used for two scenarios:
ephy_iot_mode - for IOT boards which have only one port (PHY0)
ephy_router_mode - For routers which have more than one ports
Reviewed-by: Stefan Roese <sr@denx.de> Signed-off-by: Weijie Gao <weijie.gao@mediatek.com>
net: mt7628-eth: free rx descriptor on receiving failure
When received a packet with an invalid length recorded in rx descriptor,
we should free this rx descriptor to allow us to continue to receive
following packets.
Without doing so, u-boot will stuck in a dead loop trying to process this
invalid rx descriptor.
This patch adds a call to mt7628_eth_free_pkt() after received an invalid
packet length.
Reviewed-by: Stefan Roese <sr@denx.de> Signed-off-by: Weijie Gao <weijie.gao@mediatek.com>
net: mt7628-eth: make phy link up detection optional via DT
The mt7628 has an embedded ethernet switch (5 phy ports + 1 cpu port).
Although in IOT mode only port0 is usable, the phy0 is still connected
to the switch, not the ethernet gmac directly.
This patch rewrites it and makes it optional. It can be turned on by adding
mediatek,poll-link-phy = <?> explicitly into the eth node. By default the
driver is switch mode with all 5 phy ports working without link detection.
net: mt7628-eth: remove hardcoded gpio settings and regmap-based phy reset
This patch removes hardcoded gpio settings as they have been replaced by
pinctrl in dts, and also replaces regmap-based phy reset with a more
generic reset controller.
Reviewed-by: Stefan Roese <sr@denx.de> Signed-off-by: Weijie Gao <weijie.gao@mediatek.com>
phy: mt76x8-usb-phy: add slew rate calibration and remove non-mt7628 part
This patch adds slew rate calibration for mt76x8-usb-phy, removes code
which belongs to mt7620, and gets rid of using syscon and regmap by using
clock driver and reset controller.
spi: mt7621-spi: restore default register value after each xfer
Currently this driver uses a different way to implement the spi xfer,
by modifying some fields of two registers, which is incompatible with the
MTK's original SDK linux driver. This will cause the flash data being
damaged by the SDK driver.
This patch lets the mt7621_spi_set_cs() restore the original register
fields after cs deactivated.
Reviewed-by: Stefan Roese <sr@denx.de> Signed-off-by: Weijie Gao <weijie.gao@mediatek.com>
spi: mt7621-spi: remove data cache and rewrite its xfer function
The mt7621 spi controller supports continuous generic half-duplex spi
transaction. There is no need to cache xfer data at all.
To achieve this goal, the OPADDR register must be used as the first data
to be sent. And follows the eight generic DIDO registers. But one thing
different between OPADDR and DIDO registers is OPADDR has a reversed byte
order.
With this patch, any amount of data can be read/written in a single xfer
function call.
Reviewed-by: Stefan Roese <sr@denx.de> Signed-off-by: Weijie Gao <weijie.gao@mediatek.com>
clk: add clock driver for MediaTek MT76x8 platform
This patch adds a clock driver for MediaTek MT7628/7688 SoC.
It provides clock gate control as well as getting clock frequency for
CPU/SYS/XTAL and some peripherals.
dts: mtmips: enable high-speed UART support for mt7628
All three UARTs of mt7628 are actually MediaTek's high-speed UARTs which
support baudrate up to 921600.
The high-speed UART is compatible with ns16550 when baudrate <= 115200.
Add compatible string to dtsi file so u-boot can use it when serial_mtk
driver is built in.
Reviewed-by: Stefan Roese <sr@denx.de> Signed-off-by: Weijie Gao <weijie.gao@mediatek.com>
This patch adds non-DM version for mtk hsuart driver and makes it
compatible with ns16550a driver in configuration.
This is needed in SPL with CONFIG_SPL_DM disabled for reducing size.
The current brcmnand driver is based on 4.18 linux kernel which uses
mtd_set_ooblayout to set ecc layout. But nand base code in u-boot is from
old kernel which does not use this new API and expect nand_chip.ecc.layout
structure to be set. This cause nand_scan_tail function running into a bug
check if the device has a different oob size than the default ones.
This patch ports the brcmstb_choose_ecc_layout function from kernel 4.6.7
that supports the ecc layout struture and replaces the mtd_set_ooblayout
method
Signed-off-by: William Zhang <william.zhang@broadcom.com> Reviewed-by: Philippe Reynes <philippe.reynes@softathome.com>
projects / u-boot.git / log