With commit c6d07bf440 ("net/macb: increase RX buffer size for GEM")
ethernet support does not work any more with d-cache enabled on the
AT91SAM. The reason is, that MACB_RX_BUFFER_SIZE was changed from 4096
to 128 but this change was not refected in the rx_buffer flush and
invalidate functions, as these also use this macro.
This patch now fixes this by calculating the rx buffer size correctly
again in those functions. With this change, ethernet works again
reliably on my AT91SAM board.
Signed-off-by: Stefan Roese <sr@denx.de>
Fixes: c6d07bf440 ("net/macb: increase RX buffer size for GEM")
Cc: Ramon Fried <rfried.dev@gmail.com>
Cc: Eugen Hristev <eugen.hristev@microchip.com>
Cc: Anup Patel <anup.patel@wdc.com>
Cc: Bin Meng <bmeng.cn@gmail.com>
Cc: Joe Hershberger <joe.hershberger@ni.com>
Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
Tested-by: Bin Meng <bmeng.cn@gmail.com>
Acked-by: Joe Hershberger <joe.hershberger@ni.com>
The same compatible = "altr,socfpga-stmmac" appears in both
drivers/net/designware.c and drivers/net/dwmac_socfgpa.c,
creating ambiguity in which driver will be bound.
For Intel/Altera SoC devices, dwmac_socfpga.c is the correct driver.
So drop the compatible string from designware.c.
Signed-off-by: Ralph Siemsen <ralph.siemsen@linaro.org>
Reviewed-by: Simon Goldschmidt <simon.k.r.goldschmidt@gmail.com>
Acked-by: Joe Hershberger <joe.hershberger@ni.com>
This reverts commit 1b0c9914cc.
Commit 1b0c9914cc ("net: macb: Fixed reading MII_LPA register")
causes 100Mbps does not work any more with SiFive FU540 GEM on the
HiFive Unleashed board. Revert it.
Signed-off-by: Bin Meng <bmeng.cn@gmail.com>
Acked-by: Joe Hershberger <joe.hershberger@ni.com>
It was missing in the original submission and not having it in place causes
issues with probing of PCI devices.
Signed-off-by: Alex Marginean <alexm.osslist@gmail.com>
Acked-by: Joe Hershberger <joe.hershberger@ni.com>
Align the board and driver prototype for board_interface_eth_init
to avoid execution issue (the interface_type parameter is defined
as int or phy_interface_t).
To have a generic weak function (it should be reused by other driver)
I change the prototype to use directly udevice.
This prototype is added in netdev.h to allow compilation check
and avoid warning when compiling with W=1 on file
board/st/stm32mp1/stm32mp1.c
warning: no previous prototype for 'board_interface_eth_init'\
[-Wmissing-prototypes]
int board_interface_eth_init(int interface_type, ....
^~~~~~~~~~~~~~~~~~~~~~~~
Signed-off-by: Patrice Chotard <patrice.chotard@st.com>
Signed-off-by: Patrick Delaunay <patrick.delaunay@st.com>
Acked-by: Joe Hershberger <joe.hershberger@ni.com>
This patch solves many warnings when compiling with W=1:
warning: no previous prototype for '....' [-Wmissing-prototypes]
Signed-off-by: Patrice Chotard <patrice.chotard@st.com>
Signed-off-by: Patrick Delaunay <patrick.delaunay@st.com>
Reviewed-By: Ramon Fried <rfried.dev@gmail.com>
Acked-by: Joe Hershberger <joe.hershberger@ni.com>
Do not use random value from stack as return value of pfe_phy_write().
Indicated by cppcheck.
Signed-off-by: Heinrich Schuchardt <xypron.glpk@gmx.de>
Acked-by: Joe Hershberger <joe.hershberger@ni.com>
SANDBOX_PHY_REG_CNT is not an allowable index for the array
u16 reg[SANDBOX_PHY_REG_CNT].
Identified by cppcheck.
Fixes: b47edf8069 ("test: dm_mdio: add a 2nd register to the emulated PHY")
Signed-off-by: Heinrich Schuchardt <xypron.glpk@gmx.de>
Acked-by: Joe Hershberger <joe.hershberger@ni.com>
mvpp2 already has support for setting MAC addresses but this
functionality was not exposed to the ethernet core. This commit exposes
this functionality so that MAC address assignments stored in U-Boot's
environment are correctly applied before Linux boots.
Signed-off-by: Matt Pelland <mpelland@starry.com>
Acked-by: Joe Hershberger <joe.hershberger@ni.com>
This patch adds a separate driver for the MDIO interface of the
Marvell Ethernet controllers based on driver model. There are two
reasons to have a separate driver rather than including it inside
the MAC driver itself:
*) The MDIO interface is shared by all Ethernet ports, so a driver
must guarantee non-concurrent accesses to this MDIO interface. The
most logical way is to have a separate driver that handles this
single MDIO interface, used by all Ethernet ports.
*) The MDIO interface is the same between the existing mv643xx_eth
driver and the new mvneta/mvpp2 driver. Even though it is for now
only used by the mvneta/mvpp2 driver, it will in the future be
used by the mv643xx_eth driver as well.
This driver supports SMI IEEE for 802.3 Clause 22 and XSMI for IEEE
802.3 Clause 45.
This patch also adds device tree binding for marvell MDIO driver.
Signed-off-by: Ken Ma <make@marvell.com>
Signed-off-by: Alex Marginean <alexm.osslist@gmail.com>
Acked-by: Joe Hershberger <joe.hershberger@ni.com>
Create drivers/net/fsl-mc/Kconfig and move fsl-mc specific configs
from arch/arm/cpu/armv8/fsl-layerscape/Kconfig to this new Kconfig
Signed-off-by: Florinel Iordache <florinel.iordache@nxp.com>
Acked-by: Joe Hershberger <joe.hershberger@ni.com>
This driver is used for MDIO muxes driven over I2C. This is currently
used on Freescale LS1028A QDS board, on which the physical MDIO MUX is
controlled by an on-board FPGA which in turn is configured through I2C.
Signed-off-by: Alex Marginean <alexm.osslist@gmail.com>
Acked-by: Joe Hershberger <joe.hershberger@ni.com>
Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
Add a v5l2 cache controller driver that is usually found on
Andes RISC-V ae350 platform. It will parse the cache settings
from the dtb.
In this version tag and data ram control timing can be adjusted
by the requirement from the dtb.
Signed-off-by: Rick Chen <rick@andestech.com>
Cc: KC Lin <kclin@andestech.com>
Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
Add cache enable and disable ops for test coverage.
Signed-off-by: Rick Chen <rick@andestech.com>
Cc: KC Lin <kclin@andestech.com>
Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
Add cache enable/disable ops to the DM cache uclass driver
Signed-off-by: Rick Chen <rick@andestech.com>
Cc: KC Lin <kclin@andestech.com>
Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
We should not count in hart that is marked as not available in the
device tree in riscv_cpu_get_count().
Signed-off-by: Bin Meng <bmeng.cn@gmail.com>
Reviewed-by: Rick Chen <rick@andestech.com>
Reviewed-by: Lukas Auer <lukas.auer@aisec.fraunhofer.de>
When CONFIG_CLK enabled, use CLK UCLASS for clk related settings.
Signed-off-by: Peng Fan <peng.fan@nxp.com>
Reviewed-by: Frieder Schrempf <frieder.schrempf@kontron.de>
Tested-by: Frieder Schrempf <frieder.schrempf@kontron.de>
hs: removed hunk in mxc_i2c_probe() as not longer in code
eMMC device has multiple hw partitions both address from zero. However the
mmc driver lacks block cache invalidation for switch hwpart. This causes a
problem that data of current hw partition is cached before switching to
another hw partition. And the following read operation of the latter hw
partition will get wrong data when reading from the addresses that have
been cached previously.
To solve this problem, invalidate block cache after a successful
mmc_switch_part() operation.
Signed-off-by: Weijie Gao <weijie.gao@mediatek.com>
Tested-by: Felix Brack <fb@ltec.ch>
This reverts commit 0ebe112d09.
Most block devices have only one hwpart. Multiple hwparts only found used
by eMMC devices in u-boot. The mmc driver do blk_dselect_hwpart() at the
beginning of mmc_bread() which causes block cache being invalidated too
frequently and makes block cache useless.
So it's not a good idea to put blkcache_invalidate() in the common
functions. It should be called inside mmc_select_hwpart().
Signed-off-by: Weijie Gao <weijie.gao@mediatek.com>
Tested-by: Felix Brack <fb@ltec.ch>
The PCIe controller register address in CCSR is different
on various platforms, the current code erroneously use
the hardcoded address (0xffe240000) and stride (0x10000)
to calculate the controller's index.
Fix it by adding the related info to the driver data
structure.
Signed-off-by: Hou Zhiqiang <Zhiqiang.Hou@nxp.com>
Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
Reviewed-by: Prabhakar Kushwaha <prabhakar.kushwaha@nxp.com>
The Class Code fixup method was changed from PCIe block
revision 3.0, the current fixup is only valid for the
revision 3.0 and the later ones.
So add the Class Code fixup for the block revision < 3.0.
Signed-off-by: Hou Zhiqiang <Zhiqiang.Hou@nxp.com>
Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
Reviewed-by: Prabhakar Kushwaha <prabhakar.kushwaha@nxp.com>
This can avoid build error:
The macro in brackets of the IS_ENABLED(CONFIG_FOO) is only
defined on the platforms that select the CONFIG_FOO, while
it's not defined on platforms that do not select the
CONFIG_FOO.
Signed-off-by: Hou Zhiqiang <Zhiqiang.Hou@nxp.com>
Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
Reviewed-by: Prabhakar Kushwaha <prabhakar.kushwaha@nxp.com>
The G12B clock controller is almost identical to the G12A and
so far the differences don't matter. Adding the G12B compatible
makes USB work on the Odroid-N2.
Signed-off-by: Mark Kettenis <kettenis@openbsd.org>
Signed-off-by: Neil Armstrong <narmstrong@baylibre.com>
Support of device tree model for T2080RDB, T4240RDB, T1024RDB,
T1042D4RDB, P1020RDB, P2020RDB, P2041RDB, P3041DS, P4080DS, P5040DS and
MPC8548CDS. Also support of i2c dm model.
Alignment with kernel patch proposal for binding:
[PATCH v4 0/8] stm32 m4 remoteproc on STM32MP157c
https://lkml.org/lkml/2019/5/14/159
Signed-off-by: Patrick Delaunay <patrick.delaunay@st.com>
The RCC driver have no operation so the new NOP uclass
is more appropriate. It only used as parent for clock and reset driver.
Signed-off-by: Patrick Delaunay <patrick.delaunay@st.com>
Use MISC u-class to export the NVM register (starting at 0xF8 offset)
and avoid specific API.
- SHADOW have offset < 0.
- NVM have register > 0
Signed-off-by: Patrick Delaunay <patrick.delaunay@st.com>
Since the commit 8729b1ae2c ("misc: Update read() and
write() methods to return bytes xfered"); The misc bsec driver
need to be adapted to reflect the number of transferred bytes.
Signed-off-by: Patrick Delaunay <patrick.delaunay@st.com>
Use existing gd structure to store frequency information
which can be used in drivers or arch without new request.
Signed-off-by: Patrick Delaunay <patrick.delaunay@st.com>
Add pattern for infinite test_read and test_write, that
allow to change the pattern to test without recompilation;
default pattern is 0xA5A5AA55.
Signed-off-by: Patrick Delaunay <patrick.delaunay@st.com>
Reduce verbosity of the infinite tests to avoid CubeMX issue.
test and display loop by 1024*1024 accesses: read or write.
Signed-off-by: Patrick Delaunay <patrick.delaunay@st.com>
If user choose to test memory size is 1GByte (0x40000000),
memory address would overflow in test "Random" and
test "FrequencySelectivePattern".
Thus the system would hangs up when running DDR test.
Signed-off-by: Patrick Delaunay <patrick.delaunay@st.com>
Signed-off-by: Bossen WU <bossen.wu@st.com>
