When memory is tight, modules may need to be loaded into vmalloc()
space. The code then has to generate jump trampolines which enable
relocations between vmalloc space and physical address space.
The code had a bug that was freeing these trampolines even when the
module was successfully loaded.
Signed-off-by: Felix Fietkau <nbd@nbd.name>
This update includes numerous small fixes for:
1) Interfaces setup
2) Packets parsing
3) Sending replies
Without this there were multiple problems with exchanging information
between (u)mdns and other implementations (including (u)mdns as well).
This also follows project rename to umdns which was required to avoid
confusion with Apple's mdnsd from mDNSResponder project.
Signed-off-by: Rafał Miłecki <rafal@milecki.pl>
This can be used to tweak the buildbot behavior without having to change
buildbot's configuration.
It will also allow us to add more aggressive clean steps (e.g. on
toolchain changes), which would break developers' workflows if enable
by default.
Signed-off-by: Felix Fietkau <nbd@nbd.name>
this caused
'mv: can't rename '/mnt/sysupgrade.tgz': No such file or directory'
when running sysupgrade -n
Signed-off-by: John Crispin <john@phrozen.org>
The list of v2/v3 devices is getting longer and makes reading the names of
the profiles unnecessary hard.
Signed-off-by: Sven Eckelmann <sven.eckelmann@open-mesh.com>
Default trigger action timeout was added to procd.sh in commit f88e3a4c0
(procd: add default timeout for reload trigger actions)
However, the timeout value was not placed under the correct JSON-script
array nesting level and thus did not apply.
To fix this and make the timeout actually apply to the reload triggers,
we place it in the correct scope, that is the per-trigger array.
Fixes: f88e3a4c0a
Signed-off-by: Denis Osvald <denis.osvald@sartura.hr>
Fixes the following issue:
root@LEDE:/# sysupgrade /tmp/lede-mvebu-armada-388-clearfog-sdcard.img.gz
Saving metaconfig...
Image metadata not found
Use sysupgrade -F to override this check when downgrading or flashing to vendor firmware
Image check 'fwtool_check_image' failed.
root@LEDE:/#
Acked-by: Felix Fietkau <nbd@nbd.name>
Signed-off-by: Jonas Gorski <jonas.gorski@gmail.com>
Probe the switch through DT instead of a platform device. This fixes
probe, as GPIO offsets are now properly accounted for.
Fixes the following issue:
[ 0.802953] 8021q: 802.1Q VLAN Support v1.8
[ 0.810518] rtl8366_smi: gpio_request failed for 18, err=-517
Signed-off-by: Jonas Gorski <jonas.gorski@gmail.com>
The firmware hotplug script tries to read the mac address from a
partition with the name "uboot-env" which does not exist (instead it's
name is uboot_env). This broke calculation of the new checksum (after
patching the mac address) which resulted in ath9k refusing to use the
EEPROM data.
The original error reported by ath9k was:
PCI: Enabling device 0000:00:0e.0 (0000 -> 0002)
ath: phy0: Bad EEPROM checksum 0x2523
ath: phy0: Unable to initialize hardware; initialization status: -22
ath9k 0000:00:0e.0: Failed to initialize device
Fixes: a20616863d ("lantiq: use ath9k device tree bindings
binding/owl-loader")
Signed-off-by: Martin Blumenstingl <martin.blumenstingl@googlemail.com>
Effects of the bugs could include memory corruption, tx hangs, kernel
crahes, possibly other things as well
Signed-off-by: Felix Fietkau <nbd@nbd.name>
BCM6362 and BCM3380 seem to share the same PRID. Disable serial output
for them until we can find a way to tell them apart reliably.
Fixes: 8f3cfe4ba2 ("brcm63xx: lzma-loader: add BCM3380 support")
Signed-off-by: Jonas Gorski <jonas.gorski@gmail.com>
Setting the clear bit for an interrupt seems to cause interrupts to be
deasserted again immediately. So unset the bit for BCM6318 to allow
subsequent interrupts to still work.
Signed-off-by: Jonas Gorski <jonas.gorski@gmail.com>
This commit reverts 42f3c1f ("arc770: fix broken upstream change")
and simultaneously adds back-ported fix from Linus' tree:
a524c218bc94 ("ARC: [arcompact] brown paper bag bug in unaligned
access delay slot fixup").
Note mentioned patch will appear in stable trees soon as well so
IMHO there's not much sense in adding this separate patch in lede/master branch.
As well as we will get rid of it here in 17.01 once we bump 4.4 kernel here
later down the line.
Signed-off-by: Alexey Brodkin <abrodkin@synopsys.com>
Cc: Felix Fietkau <nbd@nbd.name>
Cc: John Crispin <john@phrozen.org>
Cc: Jo-Philipp Wich <jo@mein.io>
mtdsplit_lzma requires that the rootfs be aligned to a block boundary.
Pad the kernel partition to make this so.
Signed-off-by: Claudio Leite <leitec@gmail.com>
It's needed for macaddr_add.
Fixes: 5ed23223fd ("bcm53xx: set WAN MAC address to don't share one with LAN interface")
Signed-off-by: Rafał Miłecki <rafal@milecki.pl>
