Stability of this Ethernet driver has been a long-standing issue, with
many people reporting frequent "transmit queue timeouts" and even
occasional crashes.
Disabling TSO in the driver helps with stability, although it is likely a
workaround and might not fix the issue completely.
There is a slight slowdown in forwarding performance for TCP packets
(75 kpps vs. 80 kpps with comparable CPU utilization), but this is still
enough to forward close to 1 Gbit/s of full-sized packets across multiple
flows.
Master is using a different ethernet driver, so this is not a backport.
Because of this different driver, the upcoming 21.02 release does not seem
to be affected by these stability issues.
Thanks to mrakotiq for the initial patch.
Fixes: FS#2628
Signed-off-by: Baptiste Jonglez <git@bitsofnetworks.org>
The Netgear EX6150 has an Access Point/Extender switch. Set it as
an EV_SW. Otherwise when it's set to Access Point, it will trigger
failsafe mode during boot.
Fixes: FS#3590
Signed-off-by: Kurt Roeckx <kurt@roeckx.be>
(cherry picked from commit 539966554d)
This patch was adapted to apply on top of some stable changes, but we
are not sure if this is working correctly. Felix suggested to remove
this patch for now.
Fixes: 0a59e2a76e ("mac80211: Update to version 4.19.161-1")
Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
Initial commit 8375623a06 ("ramips: add support for TP-Link Archer
C2") contains detailed installation instructions, which do not mention
a factory image. From what I can see, no support to install OpenWrt
through the vendor web interface has been added since. The factory
image is also conspicuously absent from the device page in the wiki.
Yet, it is available for download.
I bricked my Archer C2 loading the factory image through the web UI.
Serial showed this error during bootloop:
Uncompressing Kernel Image ... LZMA ERROR 1 - must RESET board to recover
This patch disables the undocumented factory image so users won't get
tricked into thinking easy web UI flashing actually works.
Signed-off-by: Stijn Segers <foss@volatilesystems.org>
(backported from commit ad5e29d38a)
The TP-Link TL-WR810N v1 is known to cause soft-brick on ath79 and
work fine for ar71xx [1]. On closer inspection, the only apparent
difference is the GPIO used for the USB regulator, which deviates
between the two targets.
This applies the value from ar71xx to ath79.
Tested successfully by a forum user.
[1] https://forum.openwrt.org/t/tp-link-tl-wr810n-v1-ath79/48267
Fixes: cdbf2de777 ("ath79: Add support for TP-Link WR810N")
Fixes: FS#3522
Signed-off-by: Adrian Schmutzler <freifunk@adrianschmutzler.de>
(cherry picked from commit 6934d30cf8)
This should fix CVE-2021-3336:
DoTls13CertificateVerify in tls13.c in wolfSSL through 4.6.0 does not
cease processing for certain anomalous peer behavior (sending an
ED22519, ED448, ECC, or RSA signature without the corresponding
certificate).
The patch is backported from the upstream wolfssl development branch.
Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
(cherry picked from commit 1f559cafe5)
The PCI device ID detected by the wifi drivers on devices using a fallback
SPROM is wrong. Currently the chipnum is used for this parameter.
Most SSB based Broadcom wifi chips are 2.4 and 5GHz capable. But on
devices without a physical SPROM, the only one way to detect if the device
suports both bands or only the 5GHz band, is by reading the device ID from
the fallback SPROM.
In some devices, this may lead to a non working wifi on a 5GHz-only card,
or in the best case a working 2.4GHz-only in a dual band wifi card.
The offset for the deviceid in SSB SPROMs is 0x0008, whereas in BCMA is
0x0060. This is true for any SPROM version.
Override the PCI device ID with the one defined at the fallback SPROM, to
detect the correct wifi card model and allow using the 5GHz band if
supported.
The patch has been tested with the following wifi radios:
BCM43222: b43: both 2.4/5GHz working
brcm-wl: both 2.4/5GHz working
BCM43225: b43: 2.4GHz, working
brcmsmac: working
brcm-wl: it lacks support
BCM43217: b43: 2.4GHz, working
brcmsmac: it lacks support
brcm-wl: it lacks support
Signed-off-by: Daniel González Cabanelas <dgcbueu@gmail.com>
Signed-off-by: Álvaro Fernández Rojas <noltari@gmail.com>
Backported from a0e0e621ca
The router Nucom R5010UN v2 has the partitions defined for a 8MB flash,
but the flash chip is 16MB size. We are wasting half of the flash.
Fix it and use generic names for partitions.
Fixes: 474cde6123 ("brcm63xx: probe SPI flash through DT")
Signed-off-by: Daniel González Cabanelas <dgcbueu@gmail.com>
(cherry picked from commit cef9e5a49f)
A vulnerability was discovered in how wpa_supplicant processing P2P
(Wi-Fi Direct) group information from active group owners.
This issue was discovered by fuzz testing of wpa_supplicant by Google's
OSS-Fuzz.
https://w1.fi/security/2020-2/wpa_supplicant-p2p-group-info-processing-vulnerability.txt
Signed-off-by: Daniel Golle <daniel@makrotopia.org>
[added the missing patch]
Signed-off-by: Petr Štetiar <ynezz@true.cz>
(cherry-picked from commit 7c8c4f1be6)
c5dccea libopkg: fix md5sum calculation
7cad0c0 opkg_verify_integrity: better logging and error conditions
14d6480 download: purge cached packages that have incorrect checksum
456efac download: factor out the logic for building cache filenames
b145030 libopkg: factor out checksum and size verification
74bac7a download: remove compatibility with old cache naming scheme
Fixes: FS#2690
Signed-off-by: Baptiste Jonglez <git@bitsofnetworks.org>
This fixes the following build problem in hostapd:
mipsel-openwrt-linux-musl/bin/ld: /builder/shared-workdir/build/tmp/ccN4Wwer.ltrans7.ltrans.o: in function `crypto_ec_point_add':
<artificial>:(.text.crypto_ec_point_add+0x170): undefined reference to `ecc_projective_add_point'
mipsel-openwrt-linux-musl/bin/ld: <artificial>:(.text.crypto_ec_point_add+0x18c): undefined reference to `ecc_map'
mipsel-openwrt-linux-musl/bin/ld: /builder/shared-workdir/build/tmp/ccN4Wwer.ltrans7.ltrans.o: in function `crypto_ec_point_to_bin':
<artificial>:(.text.crypto_ec_point_to_bin+0x40): undefined reference to `ecc_map'
Fixes: ba40da9045 ("wolfssl: Update to v4.6.0-stable")
Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
(cherry picked from commit e7d0d2e9dc)
This version fixes a large number of bugs and fixes CVE-2020-36177.
Full changelog at:
https://www.wolfssl.com/docs/wolfssl-changelog/
or, as part of the version's README.md:
https://github.com/wolfSSL/wolfssl/blob/v4.6.0-stable/README.md
Due a number of API additions, size increases from 374.7K to 408.8K for
arm_cortex_a9_vfpv3-d16. The ABI does not change from previous version.
Backported patches were removed; remaining patch was refreshed.
Signed-off-by: Eneas U de Queiroz <cotequeiroz@gmail.com>
[added reference to CVE]
Signed-off-by: Petr Štetiar <ynezz@true.cz>
(cherry picked from commit ba40da9045)
Currently it's not possible to boot the device with just initramfs image
without additional effort as the initramfs image doesn't contain device
tree. Fix it by producing FIT based image which could be booted with
following commands:
setenv bootargs earlyprintk console=ttyS0,115200
tftpboot ${kernel_addr_r} openwrt-mvebu-cortexa9-cznic_turris-omnia-initramfs-kernel.bin
bootm ${kernel_addr_r}
Acked-by: Klaus Kudielka <klaus.kudielka@gmail.com>
Reviewed-by: Tomasz Maciej Nowak <tmn505@gmail.com>
Signed-off-by: Petr Štetiar <ynezz@true.cz>
(cherry-picked from commit 337ff74894)
Backport a patch from wireguard to fix a compile problem with kernel
4.14.217.
Fixes: 2ecb22dc51 ("kernel: bump 4.14 to 4.14.217")
Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
This should fix some error messages shown in the log like this one:
dnsmasq[16020]: failed to send packet: Network unreachable
dnsmasq[16020]: failed to send packet: Address family not supported by protocol
Fixes: e87c0d934c ("dnsmasq: Update to version 2.83")
Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
The patch 4a1a58a3 build, imagebuilder: Do not require libncurses-dev
was supposed to remove libncurses as a requirement for the ImageBuilder.
However as the IB=1 is only exported during building, not for checking
requirements, it did never actually work.
This commit export IB=1 to the requirement check.
Signed-off-by: Paul Spooren <mail@aparcar.org>
(cherry picked from commit 4f38063640)
This fixes the following security problems in dnsmasq:
* CVE-2020-25681:
Dnsmasq versions before 2.83 is susceptible to a heap-based buffer
overflow in sort_rrset() when DNSSEC is used. This can allow a remote
attacker to write arbitrary data into target device's memory that can
lead to memory corruption and other unexpected behaviors on the target
device.
* CVE-2020-25682:
Dnsmasq versions before 2.83 is susceptible to buffer overflow in
extract_name() function due to missing length check, when DNSSEC is
enabled. This can allow a remote attacker to cause memory corruption
on the target device.
* CVE-2020-25683:
Dnsmasq version before 2.83 is susceptible to a heap-based buffer
overflow when DNSSEC is enabled. A remote attacker, who can create
valid DNS replies, could use this flaw to cause an overflow in a heap-
allocated memory. This flaw is caused by the lack of length checks in
rtc1035.c:extract_name(), which could be abused to make the code
execute memcpy() with a negative size in get_rdata() and cause a crash
in Dnsmasq, resulting in a Denial of Service.
* CVE-2020-25684:
A lack of proper address/port check implemented in Dnsmasq version <
2.83 reply_query function makes forging replies easier to an off-path
attacker.
* CVE-2020-25685:
A lack of query resource name (RRNAME) checks implemented in Dnsmasq's
versions before 2.83 reply_query function allows remote attackers to
spoof DNS traffic that can lead to DNS cache poisoning.
* CVE-2020-25686:
Multiple DNS query requests for the same resource name (RRNAME) by
Dnsmasq versions before 2.83 allows for remote attackers to spoof DNS
traffic, using a birthday attack (RFC 5452), that can lead to DNS
cache poisoning.
* CVE-2020-25687:
Dnsmasq versions before 2.83 is vulnerable to a heap-based buffer
overflow with large memcpy in sort_rrset() when DNSSEC is enabled. A
remote attacker, who can create valid DNS replies, could use this flaw
to cause an overflow in a heap-allocated memory. This flaw is caused
by the lack of length checks in rtc1035.c:extract_name(), which could
be abused to make the code execute memcpy() with a negative size in
sort_rrset() and cause a crash in dnsmasq, resulting in a Denial of
Service.
Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
The referenced commit is gone, but we already have this file on our
mirror, use that one by providing the correct mirror hash.
I generated a tar.xz file with the given git commit hash using a random
fork on github and it generated the same tar.xz file as found on our
mirror so this looks correct.
Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
(cherry picked from commit 20a7c9d5c9)
The referenced commit is gone, but we already have this file on our
mirror, use that one by providing the correct mirror hash.
I generated a tar.xz file with the given git commit hash using a random
fork on github and it generated the same tar.xz file as found on our
mirror so this looks correct.
Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
(cherry picked from commit a141e7a00e)
Refreshed all patches.
Removed patches because included in upstream:
- 499-mtd-parser-cmdline-Fix-parsing-of-part-names-with-co.patch
- 0071-2-PCI-qcom-Fixed-IPQ806x-PCIE-reset-changes.patch
Compile-tested on: ipq40xx, lantiq/xrx200, x86/64, ipq806x
Runtime-tested on: ipq40xx, lantiq/xrx200, x86/64
Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
Multiple prereq checks are only required within the build system but not
for the ImageBuilder. These checks are excluded by using ifndef IB.
This commit merges the three ifndef IB blocks together.
Signed-off-by: Paul Spooren <mail@aparcar.org>
(cherry picked from commit cc9d5b5a48)
The buildroot and SDK both require the compilers (gcc, g++) to be
installed on the host system, however the ImageBuilder uses precompiled
binaries.
This patch changes the prerequirements checks to skip the checking for
the compilers if running as ImageBuilder. A similar change has been
made for libncurses-dev in 4a1a58a3e2.
Signed-off-by: Sven Roederer <devel-sven@geroedel.de>
Acked-by: Paul Spooren <mail@aparcar.org>
(cherry picked from commit ae12a747ca)
The buildroot and SDK both require `libncurses-dev` to be installed on
the system, however the ImageBuilder uses precompiled binaries.
This patch changes the prerequirements checks to skip the
`libncurses-dev` part if running as ImageBuilder.
Signed-off-by: Paul Spooren <mail@aparcar.org>
(cherry picked from commit 4a1a58a3e2)
Some images are created using different filesystems, most popular
squashfs and ext4. To allow downstream projects to distinguesh between
those, add the `filesystem` information to created json files.
Signed-off-by: Paul Spooren <mail@aparcar.org>
(cherry picked from commit bc0ffff36a)
Currently it's not possible to tftpboot initramfs image on archer-c7-v5
as the image contains tplink-v1-header which leads to:
ath> bootm
## Booting image at 81000000 ...
Bad Magic Number
as U-Boot expects uImage wrapped image. This is caused by following
inheritance issue:
define Device/Init
KERNEL_INITRAMFS = $$(KERNEL)
define Device/tplink-v1
KERNEL := kernel-bin | append-dtb | lzma
KERNEL_INITRAMFS := kernel-bin | append-dtb | lzma | tplink-v1-header
define Device/tplink-safeloader
$(Device/tplink-v1)
define Device/tplink-safeloader-uimage
$(Device/tplink-safeloader)
KERNEL := kernel-bin | append-dtb | lzma | uImageArcher lzma
define Device/tplink_archer-c7-v5
$(Device/tplink-safeloader-uimage)
where tplink-v1 defines KERNEL_INITRAMFS with tplink-v1-header and it's
then used by all devices inheriting from tplink-safeloader. Fix this by
overriding KERNEL_INITRAMFS to KERNEL variable again.
Signed-off-by: Petr Štetiar <ynezz@true.cz>
(cherry picked from commit ceeece9ffa)
Refreshed all patches.
Removed patches because included in upstream:
- 315-v5.10-usbnet-ipeth-fix-connectivity-with-ios-14.patch
Compile-tested on: ipq40xx, ath79, x86/64
Runtime-tested on: ipq40xx, ath79
Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
Fixes: CVE-2020-1971, defined as high severity, summarized as:
NULL pointer deref in GENERAL_NAME_cmp function can lead to a DOS
attack.
Signed-off-by: Eneas U de Queiroz <cotequeiroz@gmail.com>
(cherry picked from commit 882ca13d92)
The removed patches were applied upstream.
The changes to 357-mac80211-optimize-skb-resizing.patch are more
complex. I think the patch already took care of the new changes done
upstream.
Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
Previously only the power LED was working.
With this patch all leds except 5GHz are working.
Signed-off-by: Davide Fioravanti <pantanastyle@gmail.com>
[rephrased commit title, drop status property]
Signed-off-by: David Bauer <mail@david-bauer.net>
(cherry picked from commit 67d019ac94)
This fixes tethering with devices using iOS 14. Prior to this patch,
connections to remote endpoints were not possible while data transfers
between the OpenWrt device and the iOS endpoints worked fine.
Signed-off-by: David Bauer <mail@david-bauer.net>
(cherry picked from commit f64496f30f)
