This board ships with an ancient 14.07-based OpenWrt using block2mtd, and
the MBR partition table contains nonsense.
It is possible to sysupgrade to an upstream OpenWrt image, but the
legacy layout of the OpenWrt images start at 0xA00 in the eMMC, with
a raw uImage. The legacy OpenWrt image doesn't "own" the beginning
of the device, including the MBR and U-Boot.
This means that when a user upgrades to upstream OpenWrt, it doesn't
boot because it can't find the right partitions. So hard-code them on
the kernel's command line using CONFIG_CMDLINE_PARTITION (for block).
Additionally, the vendor firmware doesn't cope with images larger than
about 36MiB, because it only overwrites the contents of its "firmware"
MTD partition. The current layout of the legacy image wastes a lot of
space, allowing over 32MiB for the kernel and another 10MiB for the FAT
recovery file system which is only created as 3MiB. So pull those in
to allow 4¾ MiB for the kernel, 3MiB for recovery, and then we have over
20MiB for the root file system.
This doesn't affect the new images which ship with a full eMMC image
including a different MBR layout and a partition for U-Boot, because
our modern U-Boot can actually pass the command line to the kernel, and
the built-in one doesn't get used anyway.
Tested by upgrading from vendor OpenWrt to the current legacy image,
from legacy to itself, to the previous legacy layout, and then to
finally the full-system image.
Signed-off-by: David Woodhouse <dwmw2@infradead.org>
(cherry picked from commit 6eb63019af)
The bootloader for legacy builds can't set it, so we end up unable to
log in on the serial port.
Signed-off-by: David Woodhouse <dwmw2@infradead.org>
(cherry picked from commit ea9ef8c945)
Affected versions: curl 7.62.0 to and including 7.70.0
https://curl.haxx.se/docs/CVE-2020-8169.html
Run tested on Omnia with OpenWrt 19.07
Signed-off-by: Jan Pavlinec <jan.pavlinec@nic.cz>
[added missing commit description]
Signed-off-by: Petr Štetiar <ynezz@true.cz>
5c201be Add LDFLAGS when building libsparse.a
ec17045 make_ext4fs: fix build on musl systems
Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
(cherry picked from commit 271d0c825b)
Update make-ext4fs to commit eebda1d55d9701ace2700d7ae461697fadf52d1f
git log --pretty=oneline --abbrev-commit 484903e4..eebda1d5
eebda1d make_ext4: Add strict prototypes.
bb9cf91 make_ext4fs: Remove off64_t in favor of standard off_t
Created with the help of the make-package-update-commit.sh script.
Signed-off-by: Christian Lamparter <chunkeey@gmail.com>
(cherry picked from commit ac2f341036)
The TL-WR841ND v8 feature a WiFi switch instead of a button.
This adds the corresponding input-type to prevent booting into
failsafe regularly.
This has been defined correctly in ar71xx, but was overlooked
when migrating to ath79. In contrast, the TL-WR842ND v2, which
has the key set up as switch in ar71xx, actually has a button.
The TL-MR3420 v2 has a button as well and is set up correctly
for both targets. (Information based on TP-Link user guide)
Note:
While looking into this, I found that support PR for TL-MR3420 v2
switched reset button to ACTIVE_HIGH. However, the other two
device still use ACTIVE_LOW. This seems strange, but I cannot
verify it lacking the affected devices.
Fixes: FS#2733
Fixes: 9601d94138 ("add support for TP-Link TL-WR841N/ND v8")
Signed-off-by: Adrian Schmutzler <freifunk@adrianschmutzler.de>
(backported from commit 5e86877f36)
WRT610N V2 is not detected by the initial network configuration script.
The switch remains unconfigured and wlan/lan vlans are not created.
This adds the correct setup for the device.
Fixes: FS#1869
Suggested-by: Alessandro Radicati
Signed-off-by: Adrian Schmutzler <freifunk@adrianschmutzler.de>
(cherry picked from commit a2fee73e27)
This adds new strings to the support list for the TP-Link CPE210 v3
that are supposed to work with the existing setup.
Without it, the factory image won't be accepted by the vendor UI on
these newer revisions.
Tested on a CPE210 v3.20 (EU).
Ref: https://forum.openwrt.org/t/build-for-cpe210-v3-20/68000
Signed-off-by: Adrian Schmutzler <freifunk@adrianschmutzler.de>
(cherry picked from commit 4a2380a1e7)
Pressing the 'WLAN' button should enable/disable wireless activity.
Currently, the button is mapped to the KEY_WLAN, which will not
have this effect.
This patch changes the mapping of the WLAN button, so a button
press will emit an action for the 'rfkill' key instead of 'wlan'.
Apparently, this is what stock OpenWRT expects.
This fix is analogous to the preceding patch for Fritzbox 3370.
Signed-off-by: Dustin Gathmann <dzsoftware@posteo.org>
(cherry picked from commit d5a148f5c8)
The WLAN button actions are reversed, i.e. pressing the button emits a
'released' action, and vice versa.
This can easily be checked by adding
logger -t button_action "$BUTTON $ACTION"
as the second line of /etc/rc.button/rfkill, and using logread to read
the events (assuming the preceding patch has been applied).
Defining the GPIO as ACTIVE_LOW corrects this behavior.
Signed-off-by: Dustin Gathmann <dzsoftware@posteo.org>
(cherry picked from commit 0ee30adb46)
Pressing the 'WLAN' button should enable/disable wireless activity.
However, on the Fritzbox 3370 this doesn't have an effect.
This patch changes the mapping of the physical WLAN button, so a button
press will emit an action for the 'rfkill' key instead of 'wlan'.
Apparently, this is what stock OpenWRT expects, and also what is
implemented for most other devices.
Signed-off-by: Dustin Gathmann <dzsoftware@posteo.org>
(cherry picked from commit a53bf63756)
Bumping package version has been overlooked in a previous commit.
While at it, use PKG_RELEASE instead of PKG_VERSION, as the latter
is meant for upstream version number only.
(The effective version string for the package would be "3" in both
cases, so there is no harm done for version comparison.)
Fixes: 0453c3866f ("vxlan: fix udp checksum control")
Signed-off-by: Adrian Schmutzler <freifunk@adrianschmutzler.de>
(cherry picked from commit b29d620ed2)
So far, passing "rxcsum" and "txcsum" had no effect.
Fixes: 95ab18e012 ("vxlan: add options to enable and disable UDP
checksums")
Signed-off-by: Johannes Kimmel <fff@bareminimum.eu>
[add Fixes:]
Signed-off-by: Adrian Schmutzler <freifunk@adrianschmutzler.de>
(cherry picked from commit 0453c3866f)
As the the SoC uses implicit vlan tagging for dual MAC support, the
offload feature breaks when using double tagging.
This is backport of 9da2b56760 from trunk.
As the layout of the files has changed a cherry-pick was not possible.
Signed-off-by: Sven Eckelmann <sven@narfation.org>
Signed-off-by: John Crispin <john@phrozen.org>
The f-string feature was introduced in Python 3.6. As Buildbots may run
on Debian 9, which comes per default with Python 3.5, this would cause
an issue. Instead of f-strings use the *legacy* `.format()` function.
Signed-off-by: Paul Spooren <mail@aparcar.org>
On a system python3 is linked to python3.6, fail to perform json_overview_image_info
and got `TypeError: __init__() got an unexpected keyword argument 'capture_output'`.
This patch emulate the behaviour on python 3.7+.
Signed-off-by: Huangbin Zhan <zhanhb88@gmail.com>
(cherry picked from commit 3caad5150c)
Only collect arch_packages if actually generating any output.
Fixes: commit f09b9319 ("build,json: store arch_packages in profiles.json"(
Signed-off-by: Daniel Golle <daniel@makrotopia.org>
Signed-off-by: Paul Spooren <mail@aparcar.org>
(cherry picked from commit 3b0f698760)
The `arch_packages` contains the supported package architecture.
Previously it was necessary to parse the `Packages` index for the line
`Architecture:`, requiring both an additional parser and file download.
Signed-off-by: Paul Spooren <mail@aparcar.org>
(cherry picked from commit f09b9319c6)
With this commit the `profiles.json` contain both the target specific
`default_packages` as well as the device specific `device_packages` as a
array of strings.
This information is required for downstream projects like the various
web-based interactive firmware generators.
Signed-off-by: Daniel Golle <daniel@makrotopia.org>
Signed-off-by: Paul Spooren <mail@aparcar.org>
(cherry picked from commit 263f7e5bbd)
The folder `json_info_files` contains multiple JSON files which describe
created firmware images. The folder is not removed between builds as the
ImageBuilder does not use `image.mk`.
Not removing the JSON files result in a merged `profiles.json` file
containing entries for outdated or non-existing images.
This commit adds the `json_info_files/` cleanup step to the ImageBuilder
Makefile.
Signed-off-by: Paul Spooren <mail@aparcar.org>
(cherry picked from commit 941ec28b35)
Pogoplug V4 has a reset button on a GPIO pin.
To use it, kmod-gpio-button-hotplug package needs to be installed.
Signed-off-by: Sungbo Eo <mans0n@gorani.run>
(cherry picked from commit 91472dc2ce)
Move the USB VBUS regulator nodes out of the GPIO controller node. This
fixes a problem where the "regulator-fixed" driver wasn't probed for
these regulators because the GPIO driver doesn't scan the child-nodes
and based on the dt-bindings documentation it's not supposed to.
This fixed the following error reported by Luca Olivetti:
...
dwc2 1e101000.usb: DWC OTG Controller
dwc2 1e101000.usb: new USB bus registered, assigned bus number 1
dwc2 1e101000.usb: irq 62, io mem 0x1e101000
dwc2 1e101000.usb: startup error -517
dwc2 1e101000.usb: USB bus 1 deregistered
dwc2 1e101000.usb: dwc2_hcd_init() FAILED, returning -517
Fixes: FS#1634
Cc: Luca Olivetti <luca@ventoso.org>
Signed-off-by: Martin Blumenstingl <martin.blumenstingl@googlemail.com>
[backported from 982468de35]
Signed-off-by: Luca Olivetti <luca@ventoso.org>
This fixes the following compile error:
drivers/mtd/nand/rb91x_nand.c: In function 'rb91x_nand_remove':
drivers/mtd/nand/rb91x_nand.c:445:16: error: 'rbni' undeclared (first use in this function)
nand_release(&rbni->chip);
Fixes: 9cad70044f ("kernel: fix nand_release() usage.")
Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
(cherry picked from commit 66e04abbb6)
There is one closing bracket too much.
Fixes: 9cad70044f ("kernel: fix nand_release() usage.")
Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
(cherry picked from commit 014d3f593a)
Upstream in commit f9f8f0c24203 ("crypto: ccp -- don't "select"
CONFIG_DMADEVICES") removed dependency on CONFIG_DMADEVICES symbol which
leads to build breakage of ccp crypto module, so fix this by adding that
symbol back in the kernel config.
Fixes: f4985a22ca ("kernel: Update kernel 4.14 to version 4.14.187")
Signed-off-by: Petr Štetiar <ynezz@true.cz>
(cherry picked from commit 472b8fc91b)
This bump fixes breakage introduced by kernel commit 8ab8786f78c3fc930f9abf6d6d85e95567de4e1f,
which is part of the 4.14.181 kernel bump, and backported ip6_dst_lookup_flow to 4.14.
This breaks the older WireGuard version currently in 19.07.
For reference, the compilation error is the one below:
build_dir/target-x86_64_musl/linux-x86_64/wireguard-linux-compat-1.0.20200506/src/compat/compat.h:104:42: error: 'const struct ipv6_stub' has no member named 'ipv6_dst_lookup'; did you mean 'ipv6_dst_lookup_flow'?
#define ipv6_dst_lookup_flow(a, b, c, d) ipv6_dst_lookup(a, b, &dst, c) + (void *)0 ?: dst
Changelogs below taken from the official release announcements.
== Changes since v1.0.20200506 ==
This release aligns with the changes I sent to DaveM for 5.7-rc7 and were
pushed to net.git about 45 minutes ago.
* qemu: use newer iproute2 for gcc-10
* qemu: add -fcommon for compiling ping with gcc-10
These enable the test suite to compile with gcc-10.
* noise: read preshared key while taking lock
Matt noticed a benign data race when porting the Linux code to OpenBSD.
* queueing: preserve flow hash across packet scrubbing
* noise: separate receive counter from send counter
WireGuard now works with fq_codel, cake, and other qdiscs that make use of
skb->hash. This should significantly improve latency spikes related to
buffer bloat. Here's a before and after graph from some data Toke measured:
https://data.zx2c4.com/removal-of-buffer-bloat-in-wireguard.png
* compat: support RHEL 8 as 8.2, drop 8.1 support
* compat: support CentOS 8 explicitly
* compat: RHEL7 backported the skb hash renamings
The usual RHEL churn.
* compat: backport renamed/missing skb hash members
The new support for fq_codel and friends meant more backporting work.
* compat: ip6_dst_lookup_flow was backported to 4.14, 4.9, and 4.4
== Changes since v1.0.20200611 ==
* qemu: always use cbuild gcc rather than system gcc
* qemu: remove -Werror in order to build ancient kernels better
* qemu: patch kernels that rely on ancient make
* qemu: force 2MB pages for binutils 2.31
* qemu: use cbuild gcc for avx512 exclusion
* qemu: add extra fill in idt handler for newer binutils
* qemu: support fetching kernels for arbitrary URLs
* qemu: patch in UTS_UBUNTU_RELEASE_ABI for Ubuntu detection
* qemu: work around broken centos8 kernel
* qemu: mark per_cpu_load_addr as static for gcc-10
Our qemu test suite can now handle more kernels and more compilers. Scroll
down to the bottom of https://www.wireguard.com/build-status/ to see the
expanded array of kernels we now test against, including some distro kernels.
* compat: widen breadth of integer constants
* compat: widen breadth of memzero_explicit backport
* compat: backport skb_scrub_packet to 3.11
* compat: widen breadth of prandom_u32_max backport
* compat: narrow the breadth of iptunnel_xmit backport
* compat: backport iptunnel_xmit to 3.11
With the expanded qemu test suite, it was possible to expand our list of
mainline kernels, so the backport compat layer is now more precise.
* compat: ubuntu appears to have backported ipv6_dst_lookup_flow
* compat: bionic-hwe-5.0/disco kernel backported skb_reset_redirect and ipv6 flow
Ubuntu kernels changed recently, so this ensures we can compile with the
latest Ubuntu releases.
* compat: remove stale suse support
Signed-off-by: Stijn Segers <foss@volatilesystems.org>
(cherry picked from commit 1fd1f5e8cff18f97675ce303b05d411136b99fb0)
mac80211_get_addr is called from mac80211_generate_mac, where the local variable
initialisation id="${macidx:-0}" suggests that macidx is not always defined.
Probably, idx was supposed to be used instead of $(($macidx + 1)).
Fixes: 4d99db168c ("mac80211: try to get interface addresses from wiphy sysfs 'addresses' if no mask is set")
Signed-off-by: Leon M. George <leon@georgemail.eu>
(cherry picked from commit 8f95220bcb)
The JSON `WORK_DIR` ($(KDIR)/json_info_files) is only created if the new
image generation methods from `image.mk` are used. However some targets
like `armvirt` do not use it yet, so the folder is never created.
The `json_overview_image_info.py` script used to raise an error if the
given `WORK_DIR` isn't a folder, however it should just notify about
missing JSON files.
This patch removes the Python assert and exists with code 0 even if no
JSON files were found, as this is not necessarily an error but simply
not yet implemented. Using `glob` on an not existing `Path` results in
an empty list, therefore the for loop won't run.
Signed-off-by: Paul Spooren <mail@aparcar.org>
CC: Petr Štetiar <ynezz@true.cz>
(cherry picked from commit 14cbd8fb2d)
JSON info files contain machine readable information of built profiles
and resulting images. These files were added in commit 881ed09ee6
("build: create JSON files containing image info").
They are useful for firmware wizards and script checking for
reproducibility.
Currently all JSON files are stored next to the built images, resulting
in up to 168 individual files for the ath79/generic target.
This patch refactors the JSON creation to store individual per image
(not per profile) files in $(BUILD_DIR)/json_info_files and create an
single overview file called `profiles.json` in the target directory.
Storing per image files and not per profile solves the problem of
parallel file writes. If a profiles sysupgrade and factory image are
finished at the same time both processes would write to the same JSON
file, resulting in randomly broken outputs.
Some target like x86/64 do not use the image code yet, resulting in
missing JSON files. If no JSON info files were created, no
`profiles.json` files is created as it would be empty anyway.
As before, this creation is enabled by default only if `BUILDBOT` is set.
Tested via buildroot & ImageBuilder on ath79/generic, imx6 and x86/64.
Signed-off-by: Paul Spooren <mail@aparcar.org>
[json_info_files dir handling in Make, if case refactoring]
Signed-off-by: Petr Štetiar <ynezz@true.cz>
(backported from commit 07449f692c)
Signed-off-by: Adrian Schmutzler <freifunk@adrianschmutzler.de>
Commit 881ed09ee6 ("build: create JSON files containing image info")
has removed the crucial empty new line from the image copy step
resulting in the following errors during make function expansion:
GZ_SUFFIX :=
bash: GZ_SUFFIX: command not found
Makefile:86: recipe for target 'openwrt-ath79-generic-tplink_archer-c7-v5-squashfs-sysupgrade.bin' failed
Fixes: 881ed09ee6 ("build: create JSON files containing image info")
Signed-off-by: Petr Štetiar <ynezz@true.cz>
(cherry picked from commit 0fb23d67f0)
The JSON info files contain details about the created firmware images
per device and are stored next to the created images.
The JSON files are stored as "$(IMAGE_PREFIX).json" and contain some
device/image meta data as well as a list of created firmware images.
An example of openwrt-ramips-rt305x-aztech_hw550-3g.json
{
"id": "aztech_hw550-3g",
"image_prefix": "openwrt-ramips-rt305x-aztech_hw550-3g",
"images": [
{
"name": "openwrt-ramips-rt305x-aztech_hw550-3g-squashfs-sysupgrade.bin",
"sha256": "db2b34b0ec4a83d9bf612cf66fab0dc3722b191cb9bedf111e5627a4298baf20",
"type": "sysupgrade"
}
],
"metadata_version": 1,
"supported_devices": [
"aztech,hw550-3g",
"hw550-3g"
],
"target": "ramips/rt305x",
"titles": [
{
"model": "HW550-3G",
"vendor": "Aztech"
},
{
"model": "ALL0239-3G",
"vendor": "Allnet"
}
],
"version_commit": "r10920+123-0cc87b3bac",
"version_number": "SNAPSHOT"
}
Signed-off-by: Paul Spooren <mail@aparcar.org>
(backported from commit 881ed09ee6)
Signed-off-by: Adrian Schmutzler <freifunk@adrianschmutzler.de>
For quite some time, the tiny (4M flash) TP-Link sysupgrade and
factory images cannot be built anymore by the buildbots, just
the initramfs-kernel.bin files are still there.
Disable these images for the buildbots and don't waste build
resources.
Note that these devices still build fine with default settings,
just not with the additional packages and config symbols for
the buildbots.
Signed-off-by: Adrian Schmutzler <freifunk@adrianschmutzler.de>
The uci config section network.globals set up in /bin/config_generate
will only be created if /proc/sys/net/ipv6 exists.
Correspondingly, lacking IPv6 support, the command
uci set network.globals.packet_steering=1
will fail with "uci: Invalid argument" as the network.globals config
has not been set up.
Fix that by adding the setup there as well.
While at it, limit the uci commit to the network config file.
Signed-off-by: Adrian Schmutzler <freifunk@adrianschmutzler.de>
(cherry picked from commit 636b62e8e2)
Fix the switch LAN labels for the DGND3700v1/DGND3800B router,
the order is reversed.
Signed-off-by: Daniel González Cabanelas <dgcbueu@gmail.com>
[cut out of bigger patch, adjust commit title/message accordingly]
Signed-off-by: Adrian Schmutzler <freifunk@adrianschmutzler.de>
(cherry picked from commit 583b3e4025)
Increase the SPI frequency for ELECOM WRC-1900GST and WRC-2533GST
to 40 MHz by updating the common DTSI file.
Signed-off-by: INAGAKI Hiroshi <musashino.open@gmail.com>
[WRC-1900GST]
Acked-by: NOGUCHI Hiroshi <drvlabo@gmail.com>
[split patch, adjust commit title/message]
Signed-off-by: Adrian Schmutzler <freifunk@adrianschmutzler.de>
(cherry picked from commit b5ae70d053)
The mwlwifi driver sets the default country code for EU (fi-
rmware region code 0x30) certified devices to FR (France),
not DE (Germany). Whilst this is a trivial fix, novice users
may not know how mwlwifi negatively reacts to a non-matching
country code and may leave the setting alone. Especially si-
nce it is under the advanced settings section in LuCI.
Relevant mwlwifi driver code:
0a550312dd
The mwlwifi driver readme states "Please don't change country
code and let mwlwifi set it for you." However, OpenWrt's current
behaviour does not adhere to this with its default, 'just flashed
from factory' setting for EU devices.
Signed-off-by: Jose Olivera <oliverajeo@gmail.com>
[rebase, extend commit message]
Signed-off-by: Adrian Schmutzler <freifunk@adrianschmutzler.de>
(cherry picked from commit d0e8b8310f)
The original text was copy/pasted from some other package.
Adjust the package title and description to match the description
on the publishers page.
Signed-off-by: Catalin Patulea <catalinp@google.com>
[slightly adjust content and commit message]
Signed-off-by: Adrian Schmutzler <freifunk@adrianschmutzler.de>
(cherry picked from commit 492a6594b9)
urandom-seed has a separate Makefile, we can safely remove the definition here.
Fixes: 27bfde9c9f ("base-files: move urandom seed bits into separate package")
Signed-off-by: Sungbo Eo <mans0n@gorani.run>
(cherry picked from commit 46a6586c83)
In ath79, for several SoCs the console bootargs are defined to the
very same value in every device's DTS. Consolidate these definitions
in the SoC dtsi files and drop further redundant definitions elsewhere.
The only device without any bootargs set has been OpenMesh OM5P-AC V2.
This will now inherit the setting from qca955x.dtsi
While this is a cosmetic change, backporting it to 19.07 will be a
major help for anyone doing backports of device support. Without it,
every backporter would have to remember to manually add the chosen node
to the device's DTS.
Signed-off-by: Adrian Schmutzler <freifunk@adrianschmutzler.de>
(cherry picked from commit 635f111148)
This ports support for these devices from ar71xx.
Specification:
- System-On-Chip: Qualcomm Atheros QCA9533
- CPU/Speed: v3: 560 MHz, v4: 650 MHz
- Flash: 4096 KiB
- RAM: 32 MiB
- Ethernet: 1 port @ 100M
- Wireless: SoC-integrated: QCA9533 2.4GHz 802.11bgn
In contrast to the implementation in ar71xx (reset and WiFi button),
the device actually features reset and WPS buttons.
Flashing instructions:
Upload the ...-factory.bin file via OEM web interface.
TFTP Recovery:
1. Set PC to fixed IP address 192.168.0.66
2. Download *-factory.bin image and rename it to
wa801ndv3_tp_recovery.bin
3. Start a tftp server with the image file in its root directory
4. Turn off the router
5. Press and hold Reset button
6. Turn on router with the reset button pressed and wait ~15 seconds
7. Release the reset button and after a short time
the firmware should be transferred from the tftp server
8. Wait ~30 second to complete recovery.
TFTP recovery has only been tested with v3, and the Wiki states
that the procedure won't work for v4, which cannot be verified
or falsified at the moment.
Tested by Tim Ward (see forum):
https://forum.openwrt.org/t/ath79-support-for-tp-link-tl-wa901nd-v3-v4-v5/61246/13
Signed-off-by: Adrian Schmutzler <freifunk@adrianschmutzler.de>
(cherry picked from commit 9a477b833a)
51e16eb uclient-fetch: add option to read POST data from file
99aebe3 uclient: Add string error function
Fixes: 0c910d8459 ("uclient: Update to version 2020-06-17")
Signed-off-by: Jo-Philipp Wich <jo@mein.io>
This reverts commit 0c910d8459.
We cannot use uclient Git HEAD as-is on 19.07 due to an older
version of the ustream-ssl API.
Signed-off-by: Jo-Philipp Wich <jo@mein.io>
fef6d3d uclient: Add string error function
af585db uclient-fetch: support specifying advertised TLS ciphers
c660986 uclient-fetch: add option to read POST data from file
Signed-off-by: Daniel Golle <daniel@makrotopia.org>
(cherry-squashed from commits 05145ffbef98017228dddd166960f48e98613f4d)
The Pirelli A226M-FWB has a wrong linux partition offset, caused
by a copy-paste error. As of result of this, OpenWrt is currently
broken in this unit.
Fix it.
While at it, also use generic node names and fix the addresses
there as well.
Fixes: a27d59bb42 ("brcm63xx: switch to new partition layout
specification")
Signed-off-by: Daniel González Cabanelas <dgcbueu@gmail.com>
[also fix/update node names, extend commit message]
Signed-off-by: Adrian Schmutzler <freifunk@adrianschmutzler.de>
(cherry picked from commit d64d5ed142)
It was noticed that the the whole MAC can hang when transferring data from
one ar40xx port (WAN ports) to the CPU and from the CPU back to another
ar40xx port (LAN ports). The CPU was doing only NATing in that process.
Usually, the problem first starts with a simple data corruption:
$ wget https://cdimage.debian.org/debian-cd/current/amd64/iso-cd/debian-10.4.0-amd64-netinst.iso -O /dev/null
...
Connecting to saimei.ftp.acc.umu.se (saimei.ftp.acc.umu.se)|2001:6b0:19::138|:443... connected.
...
Read error at byte 48807936/352321536 (Decryption has failed.). Retrying.
But after a short while, the whole MAC will stop to react. No traffic can
be transported anymore from the CPU port from/to the AR40xx PHY/switch and
the MAC has to be resetted.
The whole problem can be avoided by disabling IPv6 TSO for this ethernet
MAC driver.
Signed-off-by: Sven Eckelmann <sven@narfation.org>
Acked-by: John Crispin <john@phrozen.org>
(backported from commit 6785695056, with
updated commit message)
Signed-off-by: Christian Lamparter <chunkeey@gmail.com>
This patch adds ar71xx's GPIO setup for the 2.4GHz and 5GHz antennae
demultiplexer:
| 158 /* 2.4 GHz uses the first fixed antenna group (1, 0, 1, 0) */
| 159 ap9x_pci_setup_wmac_gpio(0, (0xf << 6), (0xa << 6));
| 160
| 161 /* 5 GHz uses the second fixed antenna group (0, 1, 1, 0) */
| 162 ap9x_pci_setup_wmac_gpio(1, (0xf << 6), (0x6 << 6));
This should restore the range and throughput of the 2.4GHz radio
on all the derived wndr3700 variants and versions with the AR7161 SoC.
A special case is the 5GHz radio. The original wndr3700(v1) will
benefit from this change. However the wndr3700v2 and later revisions
were unaffected by the missing bits, as there is no demultiplexer
present in the later designs.
This patch uses gpio-hogs within the device-tree for all
wndr3700/wndr3800/wndrmac variants.
Notes:
Based on the PCB pictures, the WNDR3700(v1) really had eight
independent antennae. Four antennae for each radio and all of
those were printed on the circut board.
The WNDR3700v2 and later have just six antennae. Four of those
are printed on the circuit board and serve the 2.4GHz radio.
Whereas the remaining two are special 5GHz Rayspan Patch Antennae
which are directly connected to the 5GHz radio.
Hannu Nyman dug pretty deep and unearthed a treasure of information
regarding the history of how these values came to be in the OpenWrt
archives: <https://dev.archive.openwrt.org/ticket/6533.html>.
Mark Mentovai came across the fixed antenna group when he was looking
into the driver:
fixed_antenna_group 1, (0, 1, 0, 1)
fixed_antenna_group 2, (0, 1, 1, 0)
fixed_antenna_group 3, (1, 0, 0, 1)
fixed_antenna_group 4, (1, 0, 1, 0)
Fixes: FS#3088
Reported-by: Luca Bensi
Reported-by: Maciej Mazur
Reported-by: Hannu Nyman <hannu.nyman@iki.fi>
Debugged-by: Hannu Nyman <hannu.nyman@iki.fi>
Signed-off-by: Christian Lamparter <chunkeey@gmail.com>
(cherry picked from commit 61307544d1)
