Remove code that clears spurious interrupt flags, as this step is
not required.
Performance change (as tested with iperf on RT1050)
TCP RX: 50.8Mbps->50.6Mbps
TCP TX: 46.2Mbps->49.5Mbps
Signed-off-by: Daniel DeGrasse <daniel.degrasse@nxp.com>
remove the TX thread, as this thread can become starved and unable to
reclaim TX buffers from the hardware. Instead reclaim buffers in ISR.
Change eth_tx function to first take from the tx_buf_sem, so that multiple
TX buffer descriptors can be used effectively.
Performance change (as tested with iperf on RT1050)
TCP RX: 44.6Mbps->48.6Mbps
TCP TX: 38.1Mbps->40.7Mbps
Signed-off-by: Daniel DeGrasse <daniel.degrasse@nxp.com>
Move dma buffers and buffer descriptors to DTCM, to improve hardware
and software access speed during network TX/RX
Performance change (as tested with iperf on RT1050)
TCP RX: 40.6Mbps->44.6Mbps
TCP TX: 36.8Mbps->38.1Mbps
Signed-off-by: Daniel DeGrasse <daniel.degrasse@nxp.com>
Do not force network buffers into noncacheable memory, as these buffers
are internal to the driver and not used by the hardware.
Performance change (as tested with iperf on RT1050):
TCP RX: 23.7Mbps->40.6Mbps
TCP TX: 22.3Mbps->36.8Mbps
Signed-off-by: Daniel DeGrasse <daniel.degrasse@nxp.com>
Introduce a simple binding for atmel,24mac402 EEPROM that the SAM
GMAC ethernet driver can utilize to get MAC address out of. We
introduce a 'mac-eeprom' phandle into GMAC ethernet devicetree
node that will provide a pointer to the MAC eeprom to utilize.
Signed-off-by: Kumar Gala <galak@kernel.org>
MISRA C:2012 Rule 14.4 (The controlling expression of an if statement
and the controlling expression of an iteration-statement shall have
essentially Boolean type.)
Use `do { ... } while (false)' instead of `do { ... } while (0)'.
Use comparisons with zero instead of implicitly testing integers.
The commit is a subset of the original auditable-branch commit:
5d02614e34a86b549c7707d3d9f0984bc3a5f22a
Signed-off-by: Simon Hein <SHein@baumer.com>
irq_lock() returns an unsigned integer key.
Generated by spatch using semantic patch
scripts/coccinelle/irq_lock.cocci
Signed-off-by: Johann Fischer <johann.fischer@nordicsemi.no>
increase available RX buffers to MCUX ethernet driver.
improves measured performance on RT1050 EVK zperf download
from 500Kbps to 22Mbps.
Signed-off-by: Daniel DeGrasse <daniel.degrasse@nxp.com>
Following zephyr's style guideline, all if statements, including single
line statements shall have braces.
Signed-off-by: Anas Nashif <anas.nashif@intel.com>
As we work to phase out devicetree 'label' properties, convert
driver to just use ETH_0/ETH_1 instead of DT_INST_LABEL for
logging purposes.
Signed-off-by: Kumar Gala <galak@kernel.org>
Do not lock/unlock the System Level Control Registers (SLCRs) in the Xilinx
GEM ethernet driver. The SLRCs are unlocked once at boot time.
Signed-off-by: Henrik Brix Andersen <henrik@brixandersen.dk>
Logging v1 has been removed and log_strdup wrapper function is no
longer needed. Removing the function and its use in the tree.
Signed-off-by: Krzysztof Chruscinski <krzysztof.chruscinski@nordicsemi.no>
Move to using NET_DEVICE_DT_DEFINE_INSTANCE instead of
NET_DEVICE_INIT_INSTANCE as the driver is devicetree based and it
lets us remove DT_LABEL usage in the driver itself.
Signed-off-by: Kumar Gala <galak@kernel.org>
add changes for command read-back and change from k_sleep to k_busy_wait
which will allow other threads to wakeup
Signed-off-by: Hari Haran Babu <hariharan@linumiz.com>
Signed-off-by: Marcel Graber <marcel@clever.design>
Correct width when accessing LITEETH_RX_LENGTH register.
Also update register data in device tree to the 32-bit CSR variant.
Signed-off-by: Michal Sieron <msieron@internships.antmicro.com>
Check if W5500 int pin is still active after processing the interrupt.
If not doing correctly, the W5500 driver can deadlock because it is not
receiving any more interrupts.
Signed-off-by: Janco Kock <jancokock@gmail.com>
In order to bring consistency in-tree, migrate all drivers to the new
prefix <zephyr/...>. Note that the conversion has been scripted, refer
to #45388 for more details.
Signed-off-by: Gerard Marull-Paretas <gerard.marull@nordicsemi.no>
Adds addresses and names for individual CSR registers to device tree.
This way liteuart driver no longer depends on CSR data width being 8
bits.
Also when register names or their number changes, then overlay generated
by LiteX will be incompatible with one defined here.
This should make finding breaking changes easier.
I also appended `_ADDR` suffix to defines, to distinguish them from
normal values like `LITEETH_EV_RX`.
Signed-off-by: Michal Sieron <msieron@internships.antmicro.com>
With universal LiteX HAL working, there is no need to perform multibyte
reads and writes using bitwise operations.
Just use appropriate `litex_read*` or `litex_write*` function.
Signed-off-by: Michal Sieron <msieron@internships.antmicro.com>
Use LiteX HAL functions instead of `sys_read*` or `sys_write*`
functions.
They use them inside, but choose which one to use according to
configured CSR data width.
Signed-off-by: Michal Sieron <msieron@internships.antmicro.com>
This update Atmel sam and sam0 ethernet gmac and mdio drivers to use
pinctrl driver and API. It updates all boards with new pinctrl groups
format.
Signed-off-by: Gerson Fernando Budke <nandojve@gmail.com>
Enable pinctrl for ethernet mcux driver, and update kinetis DTS node to
include labelling for PTP node, to enable driver to access pinctrl
properties.
Signed-off-by: Daniel DeGrasse <daniel.degrasse@nxp.com>
Avoids segfault in situations when we can't acquire an RX buffer, and VLAN
or PTP code is enabled which tries to inspect packets by adding a pkt
check.
Signed-off-by: Alex Sergeev <asergeev@carbonrobotics.com>
Back with commit a1b77fd589 huge chunk of code was automatically
refactored using a script, which in turn left some macros meaningless
Signed-off-by: Hristo Mitrev <hr.mitrev@gmail.com>
Just use dev->name. This change follow same principles applied when
DEV_CFG and DEV_DATA macros were removed.
Signed-off-by: Gerard Marull-Paretas <gerard.marull@nordicsemi.no>
The function eth_stm32_hal_set_config() located in
drivers/ethernet/eth_stm32_hal.c always returns -ENOTSUP,
even if everything is fine. This commit fixes the return statement
so that the real result (ret) will be returned.
Signed-off-by: Bernd Weiberg <bernd.weiberg@siemens.com>
According to Kconfig guidelines, boolean prompts must not start with
"Enable...". The following command has been used to automate the changes
in this patch:
sed -i "s/bool \"[Ee]nables\? \(\w\)/bool \"\U\1/g" **/Kconfig*
Signed-off-by: Gerard Marull-Paretas <gerard.marull@nordicsemi.no>
spi_dt_spec structure initialization should not be done
in the runtime during spi bus initialization because it
causes kernel panic.
Signed-off-by: Bartosz Bilas <bartosz.bilas@hotmail.com>
With the unification of OCM declaration & assignment between Zynq-7000
and Ultrascale/ZynqMP, remove the distinction between those two SoC
families so that the DMA area is always set up in the OCM regardless
of the current SoC type.
Signed-off-by: Immo Birnbaum <Immo.Birnbaum@weidmueller.com>
Use DEVICE_DT_GET instead of device_get_binding, since the device
reference can be obtained at compile time.
Signed-off-by: Gerard Marull-Paretas <gerard.marull@nordicsemi.no>
If CONFIG_ETH_NATIVE_POSIX_RANDOM_MAC=n and
CONFIG_ETH_NATIVE_POSIX_MAC_ADDR="", the MAC address can be set with a
net_mgmt call before the driver is initialized.
Signed-off-by: Björn Stenberg <bjorn@haxx.se>
Limit gptp range_adjust range for improved performance.
Signed-off-by: Hake Huang <hake.huang@oss.nxp.com>
Signed-off-by: Xabier Marquiegui <xmarquiegui@ainguraiiot.com>
mcux enet driver updates:
a) add mutex to access the ptp timer variables
b) use thread for rx(cooperative) andfor tx cleanup(cooperative)
c) use a thread for ptp inner clock updater(pre-empty)
This patch fixes the following issue:
```
ENET_SendFrame error: 4004
```
Signed-off-by: Hake Huang <hake.huang@oss.nxp.com>
Co-authored-by: Seb Laveze <sebastien.laveze@nxp.com>
Signed-off-by: Xabier Marquiegui <xmarquiegui@ainguraiiot.com>
