| Age | Commit message (Collapse) | Author |
|---|
|
When dma-coherent DT property is passed there is no need to do any cache
operations.
Signed-off-by: Michal Simek <[email protected]>
Link: https://lore.kernel.org/r/10c7a40364162cc8d3c82cb3e64e043f49a5153e.1772437409.git.michal.simek@amd.com
|
|
The Zynq GEM TX status register retains the transfer‑complete bit
until it is explicitly cleared. The current flow waits for
transfer‑complete but never clears it, so on the next send the wait
loop returns immediately because transfer‑complete is already high.
This causes the driver to report TX completion before the new DMA
transfer has actually finished, which breaks back‑to‑back
transmissions. This issue causes timeouts during LWIP TFTP transfers
when cache coherency is enabled.
Fix this by explicitly clearing transfer‑complete (write‑to‑clear)
after the wait completes, so each transmit starts with a clean TXSR.
Co-developed-by: Harini Katakam <[email protected]>
Signed-off-by: Harini Katakam <[email protected]>
Co-developed-by: Michal Simek <[email protected]>
Signed-off-by: Padmarao Begari <[email protected]>
Signed-off-by: Michal Simek <[email protected]>
Link: https://lore.kernel.org/r/f354680d43fba0f590a6fae693848e5bf7114ba5.1772437409.git.michal.simek@amd.com
|
|
There is no reason to react on broadcast packets that's why just ignore
them not to waste cycles on packets which are not for the platform.
Signed-off-by: Michal Simek <[email protected]>
Link: https://lore.kernel.org/r/6e10793b7d72668343756edb66221f1415570250.1772437409.git.michal.simek@amd.com
|
|
Add 10GBE high-speed Mac support, it supports 10G, 5G, 2.5G and 1G speeds.
10GBE high speed Mac is an extension of the current 1G Mac in versal,
inheriting all its current features.
MMI 10GBE ip has two internal PCS's.
1)10GBASER PCS is used for higher speeds 10G and 5G.
2)1000BASEX PCS is used for slower speeds 1G and 2.5G.
Both PCS's speed and rate configuration is done with same
usx registers. ENABLE_HS_MAC bit in NCR is the toggle switch
between the PCS's.
Signed-off-by: Venkatesh Yadav Abbarapu <[email protected]>
Link: https://lore.kernel.org/r/[email protected]
Signed-off-by: Michal Simek <[email protected]>
|
|
If the MII interface is used, the PHY is the clock master, thus don't
set the clock rate. On Zynq-7000, this will prevent the following
error:
zynq_gem ethernet@e000b000: failed to set tx clock rate 25000000
Signed-off-by: Martin Kaistra <[email protected]>
Link: https://lore.kernel.org/r/[email protected]
Signed-off-by: Michal Simek <[email protected]>
|
|
Generic understanding/consideration is that phy-mode as sgmi means that the
internal PCS(Physical Coding Sublayer) should be enabled by default.
Xilinx GEM implementation allows configuration GEM (gmii mode) + PL PCS PMA
(sgmii mode, Physical Medum Attachment) but in this case phy-mode should be
setup as gmii.
The reason for this assumption is that phy-mode should be described based
on GEM configuration not based on mode coming out of PHY.
Also Linux kernel automatically setting up PCSSEL bit when phy mode is
sgmii without a need to specified additional DT propety.
All our DTSes with sgmii phy mode have this flag enabled that's why there
is no need/reason to just duplicate information.
Signed-off-by: Michal Simek <[email protected]>
Link: https://lore.kernel.org/r/2ecdbcc4ce692e2f8b3e7054a2abab35f6c03a69.1726213052.git.michal.simek@amd.com
|
|
Drop all duplicate newlines. No functional change.
Signed-off-by: Marek Vasut <[email protected]>
|
|
As part of bringing the master branch back in to next, we need to allow
for all of these changes to exist here.
Reported-by: Jonas Karlman <[email protected]>
Signed-off-by: Tom Rini <[email protected]>
|
|
When bringing in the series 'arm: dts: am62-beagleplay: Fix Beagleplay
Ethernet"' I failed to notice that b4 noticed it was based on next and
so took that as the base commit and merged that part of next to master.
This reverts commit c8ffd1356d42223cbb8c86280a083cc3c93e6426, reversing
changes made to 2ee6f3a5f7550de3599faef9704e166e5dcace35.
Reported-by: Jonas Karlman <[email protected]>
Signed-off-by: Tom Rini <[email protected]>
|
|
Remove <common.h> from this driver directory and when needed
add missing include files directly.
Signed-off-by: Tom Rini <[email protected]>
|
|
MDC clock change needs to be done when the driver probe function
is called as mdio is enabled at probe and not when the ethernet starts.
Setup the MDC clock at the probe itself.
Signed-off-by: Venkatesh Yadav Abbarapu <[email protected]>
Link: https://lore.kernel.org/r/[email protected]
Signed-off-by: Michal Simek <[email protected]>
|
|
The documentation for struct phy state that "The content of the
structure is managed solely by the PHY API and PHY drivers".
Change to use the generic_phy_valid() helper to check if phy is valid.
Fixes: 10c50b1facbf ("net: zynq: Add support for PHY configuration in SGMII mode")
Signed-off-by: Jonas Karlman <[email protected]>
|
|
As per spec MDC must not exceed 2.5MHz, read the pclk clock
from the device tree and update the MDC clock divisor.
GEM devices support larger clock divisors and have a different
range of divisors. Program the MDIO clock divisors based on
the clock rate of the pclk clock.
Signed-off-by: Venkatesh Yadav Abbarapu <[email protected]>
Link: https://lore.kernel.org/r/[email protected]
Signed-off-by: Michal Simek <[email protected]>
|
|
There are missing Kconfig dependencies in the code which is using
firmware interface.
The commit 71efd45a5fc7 ("arm64: zynqmp: Change firmware dependency")
add option to also disable ZYNQMP_FIRMWARE. But not all Kconfig
dependencies were properly described and also sdhci and gem drivers
didn't protect the code properly.
So, add the missing ZYNQMP_FIRMWARE dependencies.
Signed-off-by: Algapally Santosh Sagar <[email protected]>
Signed-off-by: Ashok Reddy Soma <[email protected]>
Link: https://lore.kernel.org/r/[email protected]
Signed-off-by: Michal Simek <[email protected]>
|
|
In our system using ZynqMP with an external SGMII PHY it's necessary
to wait for the PCS link and auto negotiation to finish before the xfer
starts. Otherwise the first packet(s) might get dropped, resulting in a
delay at the start of the ethernet transfers.
This is only done when the PHY link is already up, which is done in
phy_startup(). As waiting for the PHY link bits via pcsstatus does not
make much sense, when the link is not available in general (e.g. no
cable connected).
This patch adds the necessary code including a minimal delay of 1 ms
which fixes problems of dropped first packages.
Signed-off-by: Stefan Roese <[email protected]>
Cc: Michal Simek <[email protected]>
Cc: Katakam Harini <[email protected]>
Cc: Ramon Fried <[email protected]>
Cc: Sean Anderson <[email protected]>
Link: https://lore.kernel.org/r/[email protected]
Signed-off-by: Michal Simek <[email protected]>
|
|
cdns prefix was deprecated and replaced by xlnx one.
Acked-by: Krzysztof Kozlowski <[email protected]>
Signed-off-by: Michal Simek <[email protected]>
Link: https://lore.kernel.org/r/7d1312d741349451b36558baa3e9132247303992.1670599167.git.michal.simek@amd.com
|
|
Removing the zynq_board_read_rom_ethaddr() function as
xlnx,eeprom is not used anymore. As all board dts to use
nvmem alias instead of xlnx,eeprom.
Signed-off-by: Venkatesh Yadav Abbarapu <[email protected]>
Link: https://lore.kernel.org/r/[email protected]
Signed-off-by: Michal Simek <[email protected]>
|
|
Here are the smatch warning messages:
drivers/net/xilinx_axi_emac.c:324 axiemac_phy_init()
error: 'phydev' dereferencing possible ERR_PTR()
drivers/net/zynq_gem.c:340 zynq_phy_init()
error: 'priv->phydev' dereferencing possible ERR_PTR()
Fix by adding error checking before dereferencing the pointer.
Signed-off-by: Venkatesh Yadav Abbarapu <[email protected]>
Link: https://lore.kernel.org/r/[email protected]
Signed-off-by: Michal Simek <[email protected]>
|
|
On QEMU setting rate for fixed clock is failing. That's why check a rate
first if the rate is the same there is no need to ask for the change.
Signed-off-by: Michal Simek <[email protected]>
Link: https://lore.kernel.org/r/bc465ffd4904bfd65208b782daa06732b915db54.1661502645.git.michal.simek@amd.com
|
|
- DM9000 DM support
- tftp server bug fix
- mdio ofnode support functions
- Various phy fixes and improvements.
[trini: Fixup merge conflicts in drivers/net/phy/ethernet_id.c
drivers/net/phy/phy.c include/phy.h]
|
|
Rename constant PHY_INTERFACE_MODE_NONE to PHY_INTERFACE_MODE_NA to make
it compatible with Linux' naming.
Signed-off-by: Marek Behún <[email protected]>
Reviewed-by: Stefan Roese <[email protected]>
Reviewed-by: Ramon Fried <[email protected]>
Reviewed-by: Vladimir Oltean <[email protected]>
|
|
Add helpers ofnode_read_phy_mode() and dev_read_phy_mode() to parse the
"phy-mode" / "phy-connection-type" property. Add corresponding UT test.
Use them treewide.
This allows us to inline the phy_get_interface_by_name() into
ofnode_read_phy_mode(), since the former is not used anymore.
Signed-off-by: Marek Behún <[email protected]>
Reviewed-by: Ramon Fried <[email protected]>
Tested-by: Patrice Chotard <[email protected]>
|
|
Add support for SGMII dynamic configuration which will takes care of
configuring SGMII in the GEM secure (GEM_CLK_CTRL) configuration
register.
Signed-off-by: T Karthik Reddy <[email protected]>
Signed-off-by: Michal Simek <[email protected]>
Link: https://lore.kernel.org/r/a8915186e44015959978d080a31de652f544cd4a.1648631275.git.michal.simek@xilinx.com
|
|
As we are not reading the PHY address in case of CONFIG_ETH_PHY in plat
function, phy address always prints as -1. So move the ethernet info
print statement to probe function, to display proper phy address.
Signed-off-by: T Karthik Reddy <[email protected]>
Signed-off-by: Michal Simek <[email protected]>
Link: https://lore.kernel.org/r/f6efc6719d767b1bebe65987c22c6d52329f4225.1648631275.git.michal.simek@xilinx.com
|
|
CONFIG_ETH_PHY enables support to utilize generic ethernet phy
framework. Though if ethernet PHY node is in other ethernet node, it
will use shared MDIO to access the PHY of other ethernet.
Signed-off-by: Michal Simek <[email protected]>
Signed-off-by: T Karthik Reddy <[email protected]>
Link: https://lore.kernel.org/r/337b1a38ba36cde1951739af62fb3d2736d97f53.1648631275.git.michal.simek@xilinx.com
|
|
For configurations with gmii2rgmii and external phy the DT nodes link
should be gem->gmii2rgmii->phy. But due to limitation in Linux driver
the DT is mentioned as gem->phy and gmii2rgmii->phy as shown in below DT.
ethernet@ff0c0000 {
compatible = "cdns,zynqmp-gem\0cdns,gem";
status = "okay";
interrupt-parent = <0x04>;
interrupts = <0x00 0x3b 0x04 0x00 0x3b 0x04>;
reg = <0x00 0xff0c0000 0x00 0x1000>;
clock-names = "pclk\0hclk\0tx_clk\0rx_clk\0tsu_clk";
#address-cells = <0x01>;
#size-cells = <0x00>;
#stream-id-cells = <0x01>;
iommus = <0x0d 0x875>;
power-domains = <0x0c 0x1e>;
clocks = <0x03 0x1f 0x03 0x69 0x03 0x2e 0x03 0x32 0x03 0x2c>;
phy-handle = <0x0e>;
phy-mode = "gmii";
xlnx,ptp-enet-clock = <0x00>;
local-mac-address = [ff ff ff ff ff ff];
phandle = <0x4d>;
mdio {
#address-cells = <0x01>;
#size-cells = <0x00>;
phandle = <0x4e>;
ethernet-phy@1 {
reg = <0x01>;
rxc-skew-ps = <0x708>;
txc-skew-ps = <0x708>;
phandle = <0x0e>;
};
gmii_to_rgmii_0@8 {
compatible = "xlnx,gmii-to-rgmii-1.0";
phy-handle = <0x0e>;
reg = <0x08>;
phandle = <0x4f>;
};
};
};
Since same DT is used in Linux and U-Boot we need to workaround this
issue by using the gmii2rgmii node which points to phy and we should
ignore the gem pointing to phy directly.
Do this workaround by updating priv->phydev->node value with
priv->phy_of_node only if it is not valid node.
Signed-off-by: Ashok Reddy Soma <[email protected]>
Signed-off-by: Michal Simek <[email protected]>
Reviewed-by: Ramon Fried <[email protected]>
Link: https://lore.kernel.org/r/641eb13425ffe80e0743f60cf90d0f940577b9e9.1642162085.git.michal.simek@xilinx.com
|
|
This reverts commit eafdcda4a854932c0319656de7bf3f017f17ae67.
The main reason is that QEMU is using BOOTP protocol which is sending DHCP
Offer to a broadcast address that's why it can't be disabled.
DHCP protocol has no issue because it returns directly to client MAC
address.
Both of these options are described in RFC951
(https://datatracker.ietf.org/doc/html/rfc951#section-4)
Signed-off-by: Michal Simek <[email protected]>
Reviewed-by: Ramon Fried <[email protected]>
Link: https://lore.kernel.org/r/fc5f5e2aeca77847ed4ca6a263890375ab9f5163.1642162545.git.michal.simek@xilinx.com
|
|
Timeout for checking mdio phy idle status is 20seconds. In case of errors
this timeout will be too much. Reduce it to 100ms.
Signed-off-by: Ashok Reddy Soma <[email protected]>
Signed-off-by: Michal Simek <[email protected]>
Reviewed-by: Ramon Fried <[email protected]>
Link: https://lore.kernel.org/r/1b73aa57b77587391e1bcd6d9f0480163367ed1b.1637237121.git.michal.simek@xilinx.com
|
|
SGMII configuration depends on proper GT setting that's why when node has
phys property call PSGTR driver to configure it properly.
Signed-off-by: Michal Simek <[email protected]>
Reviewed-by: Sean Anderson <[email protected]>
Link: https://lore.kernel.org/r/bbc8d7ed9d308199168e4455c7a3e3a5ac0890e7.1639562397.git.michal.simek@xilinx.com
|
|
Perform reset before core initialization.
Standard flow which close to 99% users are using getting all IPs out of
reset that there is no need to reset IP again. This is because of all low
level initialization is done in previous bootloader stage.
In SOM case these IPs are not touched by previous bootloader stage that's
why reset needs to be called before IP is accessed to make sure that it is
in correct state.
Signed-off-by: Michal Simek <[email protected]>
Reviewed-by: Ramon Fried <[email protected]>
Link: https://lore.kernel.org/r/5ae1c85b282d632bb62030f1f24a0065661b9153.1638804318.git.michal.simek@xilinx.com
|
|
Xilinx DTS files are using two way how to describe ethernet phy.
The first (already supported) has phy as subnode of gem node.
eth {
phy-handle = <&phy0>;
phy0: ethernet-phy@21 {
...
};
};
The second has mdio subnode (with mdio name) which has phy subnode. This
structure allow hadling MDIO reset signal (based on Linux mdio.yaml)
eth {
phy-handle = <&phy0>;
mdio {
phy0: ethernet-phy@21 {
...
};
};
};
This patch adds support for the second case where mdio subnode
is found driver will look at its parent to find out which gem is handling
MDIO bus.
Signed-off-by: Michal Simek <[email protected]>
Link: https://lore.kernel.org/r/6748007f0b6db9554d7a4b52352dce23ca403f9d.1638798796.git.michal.simek@xilinx.com
|
|
There is no need for GEM to accepts broadcast packets because they are not
handled by u-boot anyway. That's why use HW IP feature and don't waste time
on these packats which will be dropped anyway.
Signed-off-by: Michal Simek <[email protected]>
Reviewed-by: Ramon Fried <[email protected]>
Link: https://lore.kernel.org/r/0e236c3a6514a2a633ef3a5b71a967c46f7fbae7.1634303007.git.michal.simek@xilinx.com
|
|
The code was trying to disable PCS auto-negotiation when a fixed-link node
is present and enable it otherwise. However, the PCS registers were being
written before the PCSSEL bit was set in the network configuration
register, and it appears that in this state, PCS register writes are
ignored. The result is that the intended change only took effect on the
second network operation that was performed, since at that time PCSSEL is
already enabled.
Fix the order of register writes so that PCS registers are only written to
after the PCS is enabled.
Fixes: 26e62cc971 ("net: gem: Disable PCS autonegotiation in case of fixed-link")
Signed-off-by: Robert Hancock <[email protected]>
Reviewed-by: Ramon Fried <[email protected]>
Reviewed-by: Ashok Reddy Soma <[email protected]>
Signed-off-by: Michal Simek <[email protected]>
|
|
Clean up error path in connection where priv->rxbuffers and priv->tx_bd are
allocated.
Signed-off-by: Michal Simek <[email protected]>
|
|
Enable rx clock along with tx clock for versal platform. Use compatible
data to enable/disable clocks in the driver.
Signed-off-by: T Karthik Reddy <[email protected]>
Signed-off-by: Michal Simek <[email protected]>
Reviewed-by: Ramon Fried <[email protected]>
|
|
A lot of Xilinx drivers are checking -ENOSYS which means that clock driver
doesn't have enable function. Remove this checking from drivers and create
dummy enable function as was done for clk_fixed_rate driver by
commit 6bf6d81c1112 ("clk: fixed_rate: add dummy enable() function").
Signed-off-by: Michal Simek <[email protected]>
|
|
If probe fails, the mdio bus isn't unregistered. Fix it.
Signed-off-by: Michael Walle <[email protected]>
Reviewed-by: Ramon Fried <[email protected]>
Signed-off-by: Michal Simek <[email protected]>
|
|
At present various drivers etc. access the device's 'seq' member directly.
This makes it harder to change the meaning of that member. Change access
to go through a function instead.
The drivers/i2c/lpc32xx_i2c.c file is left unchanged for now.
Signed-off-by: Simon Glass <[email protected]>
|
|
This name is far too long. Rename it to remove the 'data' bits. This makes
it consistent with the platdata->plat rename.
Signed-off-by: Simon Glass <[email protected]>
|
|
Rename this to be consistent with the change from 'platdata'.
Signed-off-by: Simon Glass <[email protected]>
|
|
We use 'priv' for private data but often use 'platdata' for platform data.
We can't really use 'pdata' since that is ambiguous (it could mean private
or platform data).
Rename some of the latter variables to end with 'plat' for consistency.
Signed-off-by: Simon Glass <[email protected]>
|
|
This construct is quite long-winded. In earlier days it made some sense
since auto-allocation was a strange concept. But with driver model now
used pretty universally, we can shorten this to 'auto'. This reduces
verbosity and makes it easier to read.
Coincidentally it also ensures that every declaration is on one line,
thus making dtoc's job easier.
Signed-off-by: Simon Glass <[email protected]>
|
|
Find out MDIO bus and enable MDIO access to it if this is done via
different GEM controller. Only works across GEM instances.
Signed-off-by: Michal Simek <[email protected]>
|
|
Disable PCS autonegotiation if fixed-link node is present in device tree.
This way systems with multiple GEM instances with a combination of
SGMII-fixed and SGMII-PHY will work.
Reported-by: Goran Marinkovic <[email protected]>
Signed-off-by: Michal Simek <[email protected]>
Signed-off-by: Ashok Reddy Soma <[email protected]>
Signed-off-by: Michal Simek <[email protected]>
|
|
Move this uncommon header out of the common header.
Signed-off-by: Simon Glass <[email protected]>
|
|
Move this header out of the common header.
Signed-off-by: Simon Glass <[email protected]>
|
|
Move this header out of the common header. Network support is used in
quite a few places but it still does not warrant blanket inclusion.
Note that this net.h header itself has quite a lot in it. It could be
split into the driver-mode support, functions, structures, checksumming,
etc.
Signed-off-by: Simon Glass <[email protected]>
|
|
Not all IPs have private MDIO bus and MDIO bus should be shared between
several IPs. In past one patch tried to implement it
(https://lists.denx.de/pipermail/u-boot/2018-February/319285.html)
in pretty raw way but it is not the cleanest solution.
This patch is just taking the part of that solution to be able to handle it
over releases without conflicts.
Signed-off-by: Michal Simek <[email protected]>
|
|
Add cache flush to zynq_gem_free_pkt. This is necessary
because some net routines would modify this buffer in place.
The cache_invalidate in the zynq_gem_recv function would cause
the modifications to the buffer to overwrite the DMA from the GEM,
if cache coherency is not enabled in the GEM, the next time the
buffer is in use.
Flushing the cache when the buffer is no longer in use by the
net functions ensures that the GEM DMA is going to take place
into a clean buffer.
Signed-off-by: Ashok Reddy Soma <[email protected]>
Signed-off-by: Michal Simek <[email protected]>
|
|
Driver probe function is called again and again in case of error.
Malloc space is getting full which is is reported by:
Insufficient RAM for page table: 0x15000 > 0x14000.
Please increase the size in get_page_table_size()
### ERROR ### Please RESET the board ###
The patch is freeing allocated buffers on error path to avoid panic.
Signed-off-by: Michal Simek <[email protected]>
|
