In the "ipq40xx: switch to Kernel 5.10" discussion at GitHub,
Adrian noted [0] that these GL.iNet Conexa series devices,
GL-B1300 and GL-S1300 failed their image generation [1] as their gzipped
uImage kernel went above 4096k.
While notifying the vendor about this problem [2], I tested all U-Boot
releases from GL.iNet:
- they really fail to boot kernel above 4096k
- they don't support lzma: "Unimplemented compression type 3"
- but they boot zImage
Using zImage (xz compression) the kernel is 2909k which is
more than a megabyte away from the KERNEL_SIZE := 4096k limit.
The gzip compressed version would be 4116k.
[0]: https://github.com/openwrt/openwrt/pull/4620#issuecomment-932765776
[1]: commit 7b1fa276f5 ("ipq40xx: add testing support for kernel 5.10")
[2]: https://forum.gl-inet.com/t/ipq40xx-kernel-size-and-u-boot-v5-10-is-too-big-for-4-mb/17619
Signed-off-by: Szabolcs Hubai <szab.hu@gmail.com>
This commit will add support for the Meraki MX100 in OpenWRT.
Specs:
* CPU: Intel Xeon E3-1200 Series 1.5GHz 2C/4T
* Memory: 4GB DDR3 1600 ECC
* Storage: 1GB USB NAND, 1TB SATA HDD
* Wireless: None
* Wired: 10x 1Gb RJ45, 2x 1Gb SFP
UART:
The UART header is named CONN11 and is found in the
center of the mainboard. The pinout from Pin 1 (marked
with a black triangle) to pin 4 is below:
Pin 1: VCC
Pin 2: TX
Pin 3: RX
Pin 4: GND
Note that VCC is not required for UART on this device.
Booting:
1. Flash/burn one of the images from this repo to a
flash drive.
2. Take the top off the MX100, and unplug the SATA
cable from the HDD.
3. Hook up UART to the MX100, plug in the USB drive,
and then power up the device.
4. At the BIOS prompt, quickly press F7 and then
scroll to the Save & Exit tab.
5. Scroll down to Boot Override, and select the
UEFI entry for your jumpdrive.
Note: UEFI booting will fail if the SATA cable for
the HDD is plugged in.
The issue is explained under the Flashing instructions.
Flashing:
1. Ensure the MX100 is powered down, and not plugged
into power.
2. Take the top off the MX100, and unplug the SATA
cable from the HDD.
3. Using the Mini USB female port found by the SATA
port on the motherboard,
flash one of the images to the system. Example:
`dd if=image of=/dev/sdb conv=fdatasync` where sdb
is the USB device for the MX100's NAND.
4. Unplug the Mini USB, hook up UART to the MX100,
and then power up the device.
5. At the BIOS prompt, quickly press F7 and then
scroll to the Boot tab.
6. Change the boot order and set UEFI: USB DISK 2.0
as first, and USB DISK 2.0 as second.
Disable the other boot options.
7. Go to Save & Exit, and then select Save Changes and
Reset
Note that OpenWRT will fail to boot in UEFI mode when
the SATA hard drive is plugged in. To fix this, boot
with the SATA disk unplugged and then run the following
command:
`sed -i "s|hd0,gpt1|hd1,gpt1|g" boot/grub/grub.cfg`
Once the above is ran, OpenWRT will boot when the HDD
is plugged into SATA. The reason this happens is the
UEFI implementation for the MX100 will always set
anything on SATA to HD0 instead of the onboard USB
storage, so we have to accomidate it since OpenWRT's
GRUB does not support detecting a boot disk via UUID.
Signed-off-by: Chris Blake <chrisrblake93@gmail.com>
try to reduce the kernel size by disabling and moving
options from the common kernel configuration to the
SATA target that doesn't have the constraints.
For NAND this has become necessary because as with 5.10
some devices outgrew their kernels. Though, in my tests
this didn't help much: just a smidgen over 100kib was
saved on the uncompressed kernel.
... running make kernel_oldconfig also removed some
other config symbols, mostly those that already set
from elsewhere or became obsolete in the meantime.
Signed-off-by: Christian Lamparter <chunkeey@gmail.com>
disables the MX60(W) from being built by the builders for now.
But there's an effort to bring it back:
<https://github.com/openwrt/openwrt/pull/4617>
Signed-off-by: Christian Lamparter <chunkeey@gmail.com>
The D-Link DIR-685 has a small screen with a framebuffer
console, so if we have this, when we start, display the
banner on this framebuffer console so the user know they
are running OpenWRT as root filesystem.
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
Due to 5.10 increased kernel size, the current 4MiB-ish kernel
partition got too small. Luckily, netgear's uboot environment
is setup to read 0x60000 bytes from the kernel partition location.
... While at it: also do some cleanups in the DTS in there.
The original (re-)installation described in
commit d82d84694e ("apm821xx: add support for the Netgear WNDAP620 and WNDAP660")
seemed to be still working for now. What I noticed though
is that the bigger initramfs images needed to use a different
destination address (1000000) to prevent it overwriting
itself during decompression. i.e:
# tftp 1000000 openwrt-...-wndap620-initramfs-kernel.bin
# bootm
However, in case of the WNDAP620+660 the factory.img image can be
written directly to the flash through uboot.
Signed-off-by: Christian Lamparter <chunkeey@gmail.com>
Both NAND and SATA targets need the DMA engine in one way
or another.
Due to a kernel config refresh various existing symbols
got removed from the apm821xx main config file as well.
(That being said, they are still included because the
built-in crpyto4xx depends on these.)
Signed-off-by: Christian Lamparter <chunkeey@gmail.com>
Kernel has added the different variants of the Rock Pi 4 in commit
b5edb0467370 ("arm64: dts: rockchip: Mark rock-pi-4 as rock-pi-4a
dts"). The former Rock Pi 4 is now Rock Pi 4A.
For compatibility with kernel 5.4, this rename has been held back
so far. Having switched to kernel 5.10 now, we can finally apply
it in our tree as well.
Signed-off-by: Adrian Schmutzler <freifunk@adrianschmutzler.de>
Now that we have fully switched to nvmem interface we can drop
the use of mtd-mac-address patches as it's not used anymore and
the new nvmem implementation should be used for any new device.
Signed-off-by: Ansuel Smith <ansuelsmth@gmail.com>
LED labels got reversed by accident, so fix it to the usual color:led_name format.
Fixes: 78cf3e53b1 ("mvebu: add Globalscale MOCHAbin")
Signed-off-by: Robert Marko <robert.marko@sartura.hr>
[add Fixes:]
Signed-off-by: Adrian Schmutzler <freifunk@adrianschmutzler.de>
Linux 5.10 has been there as testing kernel for a while now.
Do the switch and drop config and patches for Linux 5.4.
Signed-off-by: Daniel Golle <daniel@makrotopia.org>
Enable kernel options to allow loading device tree overlay via configfs
at runtime. This is useful for devboards like the BPi-R2 and BPi-R64
which got RasbPi-compatible 40-pin GPIO header which allow all sorts
of extensions.
Signed-off-by: Daniel Golle <daniel@makrotopia.org>
The otto GPIO driver does not work with rtl9300 SoCs. Add
the legacy driver again and use that by default in the 9300 .dtsi
Signed-off-by: Birger Koblitz <git@birger-koblitz.de>
RTL8393 SoCs older than Revision C hang on accesses to PHYs with PHY address
larger or equal to the CPU-port (52). This will make scanning the MDIO bus
hang forever. Since the RTL8390 platform does not support more than
52 PHYs, return -EIO for phy addresses >= 52. Note that the RTL8390 family
of SoCs has a fixed mapping between port number and PHY-address.
Signed-off-by: Birger Koblitz <git@birger-koblitz.de>
Adds SoC specific routing offload implementations for
RTL8380/90 and RTL9300. RTL83xx supports merely nexthop
routing, RTL9300 full host and prefix routes.
Signed-off-by: Birger Koblitz <git@birger-koblitz.de>
Add generic support for listening to FIB and Event notifier updates and
use this information to hook into the L3 hardware capabilities of the
RTL SoCs.
Signed-off-by: Birger Koblitz <git@birger-koblitz.de>
The ingress filter registers use 2 bits for each port to define the filtering
state, whereas the egress filter uses 1 bit. So for for the ingress filter
the register offset for a given port is:
(port >> 4) << 4: since there are 16 entries in a register of 32 bits
and for the egress filter:
(port >> 5) << 4: since there are 32 entries in a register of 32 bits
Signed-off-by: Birger Koblitz <git@birger-koblitz.de>
Configure a sane L2 learning configuration upon DSA driver load so that the
switch can start learning L2 addresses. Also configure the correct flood masks
for broadcast and unknown unicast traffice.
Signed-off-by: Birger Koblitz <git@birger-koblitz.de>
This adds RTL93xx-specific MAC configuration routines that allow also configuration
of 10GBit links for phylink. There is support for the Realtek-specific HISGMI
protocol.
Signed-off-by: Birger Koblitz <git@birger-koblitz.de>
This adds support for offloading TC flower by using the Packet Inspection Engine
of the RTL-SoCs. Basic infrastructure support is provide with callbacks to the
tc subsystem and support for HW packet counters.
Signed-off-by: Birger Koblitz <git@birger-koblitz.de>
All RTL SoCs addresss PHYs via their port number, which is mapped to an
SMI address. Add support for configuring this mapping via the .dts on all
SoCs apart from the 839x, where the mapping to the 64 ports is fixed.
Signed-off-by: Birger Koblitz <git@birger-koblitz.de>
On RTL83xx enable learning of the MAC source address of the CPU port
from outgoing packets. Add documentation on bit fields. On RTL93xx
enable port-mask usage and the use of internal priority, these
SoCs automatically learn the MAC.
Signed-off-by: Birger Koblitz <git@birger-koblitz.de>
Remove the storm control and attack warnings from the IRQ handler
of the Ethernet driver. There was no consequence to the detection
and the kernel can also handle at least the attacks itself.
Signed-off-by: Birger Koblitz <git@birger-koblitz.de>
This enlarges the size of the TX ring buffer, which prevents warnings
when the buffer runs out of space.
Signed-off-by: Birger Koblitz <git@birger-koblitz.de>
The toolchain can be used for accelerated CI builds. This commit enabled
the packing of it by default on buildbots.
Signed-off-by: Paul Spooren <mail@aparcar.org>
This fixes:
ERROR: "switch_generic_set_link" [drivers/net/phy/b53/b53_common.ko] undefined!
At some point all packages for swconfig drivers were dropped and targets
were meant to have them built into kernels. It seems b53 (re-)gained its
kmod-switch-bcm53xx however and b53 needs to be built as module.
Fixes: b2cfed48f6 ("Revert "swconfig: fix Broadcom b53 support"")
Signed-off-by: Rafał Miłecki <rafal@milecki.pl>
Add Vladimir Oltean's "net: dsa: don't set skb->offload_fwd_mark when
not offloading the bridge"
This covers cases where packets received by an upstream switch must be
forwarded back on the same port, which skb->offload_fwd_mark normally
prevents.
Signed-off-by: Matthew Hagan <mnhagan88@gmail.com>
This reverts commit 8f9cd1af0f.
That commit was meant to add a single EXPORT_SYMBOL_GPL() but it
actually also added few .of_match_table-s. One commit should handle one
thing and should not introduce unrelated changes.
Regarding actual changes:
1. EXPORT_SYMBOL_GPL is not required as we don't build swconfig drivers
as modules.
2. PHY drivers must not have .of_match_table. That is allowed for MDIO
drivers. This could work for some time (although is didn't for me on
bcm53xx) but does not with kernel 5.10. It causes a soft lockup and
upstream developers confirmed it's an unsupported design.
Link: https://lore.kernel.org/netdev/2b1dc053-8c9a-e3e4-b450-eecdfca3fe16@gmail.com/t/#mf80e472f35ee23f7a75cbf5b1e101a17ab3a64a3
Cc: Tobias Schramm <tobleminer@gmail.com>
Signed-off-by: Rafał Miłecki <rafal@milecki.pl>
This has testing support for 7 months. Time to switch.
TL-WDR4900 is disabled due to kernel size limitation.
Signed-off-by: Adrian Schmutzler <freifunk@adrianschmutzler.de>
Three missing symbols were found during mpc85xx/p2020 compilation.
While at it, CONFIG_FSL_ENETC_MDIO is moved to generic config
for consistency.
Signed-off-by: Pawel Dembicki <paweldembicki@gmail.com>
[move CONFIG_FSL_ENETC_MDIO, remove redundant definitions, adjust
commit message/title]
Signed-off-by: Adrian Schmutzler <freifunk@adrianschmutzler.de>
With the upgrade to kernel 5.10 per default the old version is no
longer required to be in tree.
Signed-off-by: Adrian Schmutzler <freifunk@adrianschmutzler.de>
When KERNEL_PERF_EVENTS is enabled in OpenWrt, or PERF_EVENTS in the
kernel config, the RPI_AXIPERF is exposed. Add it to the subtarget
kernel configs to avoid build failures.
Signed-off-by: Stijn Tintel <stijn@linux-ipv6.be>
When building an image for the bcm27xx target, some combinations of
config options will fail to build due the SND_SOC_AD193X_I2C and
SND_SOC_AD193X_SPI kernel config symbols being missing.
The problem only occurs on bcm27xx as the target contains a patch that
modifies the Kconfig file containing the symbols; in the vanilla kernel,
there is no string after the tristate keyword so the symbol is not
exposed.
The _I2C symbol depends on I2C, which is enabled in the kernel configs
of all bcm27xx subtargets.
The _SPI symbol depends on SPI_MASTER, which is exposed by either
kmod-mmc-spi, kmod-spi-bitbang, kmod-spi-dev, kmod-spi-bcm2835 or
kmod-spi-bcm2835-aux.
Both symbols are defined in the sound/soc/codecs directory, which is
only included when SND_SOC is enabled, so this problem doesn't occur
when kmod-sound-soc-core is not enabled. As the
kmod-sound-soc-bcm2835-i2s package disables the SND_SOC_AD193X_SPI
symbol, it also doesn't occur when kmod-sound-soc-bcm2835-i2s is
enabled.
As there are several possible config combinations that do exhibit this
problem, it is best to solve it by adding the missing symbols to the
subtarget kernel configs. By doing this we can remove them from the
kmod-sound-soc-bcm2835-i2s package.
Signed-off-by: Stijn Tintel <stijn@linux-ipv6.be>
