After years of trying to find the reason for random kernel crashes
while both CPU and SATA are under load it has been found.
Some odd commented-out #defines in kref's single-port driver [1] which
were copied from the vendor driver made me develop a theory:
The IO-mapped memory area for DMA descriptors apparetly got some holes
just before the alignment boundaries.
This feels like an off-by-one bug in the hardware or maybe those fields
are used internally by the SATA controller's firmware.
Whatever the cause is: they cannot be used and trying to use them
results in reading back unexpected stuff and ends up with oopsing
Unable to handle kernel paging request at virtual address d085c004
Work around the issue by reducing the area used for bmdma descriptors.
This reduces SATA performance (iops) quite a bit, but finally makes
things work reliably. Possibly one could optimize this much more by
really just skipping the holes in that memory area -- however, that
seems to be non-trivial with the driver and libata in it's current form
(suggestions are welcome).
The 'proper' way to have good SATA performance would be to make use of
the hardware RAID features (one can use the JBOD mode to access even
just a single disc transparently through the RAID controller integrated
in the SATA host instead of accessing the SATA ports 'raw' as we do
now).
[1]: https://github.com/kref/linux-oxnas/blob/master/drivers/ata/sata_oxnas.c#L25
Signed-off-by: Daniel Golle <daniel@makrotopia.org>
The CONFIG_EFI_CUSTOM_SSDT_OVERLAYS option was added in kernel 4.14.188,
set it for the x86/generic target.
This fixes a build problem in the x86/generic target.
Fixes: 148d59c67e ("kernel: update kernel 4.14 to version 4.14.193")
Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
After those symbols have been removed from generic kernel config,
they were added to the target config during every kernel config
refresh. As that's not desirable, add them back to the generic
config.
Fixes: d1a8217d87 ("kernel: clean-up build-configurable kernel config symbols")
Fixes: cfe235c436 ("kernel: modules: add package kmod-iosched-bfq")
Signed-off-by: Adrian Schmutzler <freifunk@adrianschmutzler.de>
The device uses a rf-kill switch instead of a button. Furthermore the
GPIO is active high.
Signed-off-by: Christoph Krapp <achterin@googlemail.com>
(cherry picked from commit 0af656e978)
WIZnet WizFi630s board name is written slightly different it its OEM
OpenWrt firmware. This causes an incompatibility warning during flashing
with sysupgrade. This patch is adding the vendor board name to the
supported devices list to avoid this warning. For initial flashing you
can use sysupgrade via command line or luci beside of TFTP.
Do not keep the OEM configuration during sysupgrade.
Signed-off-by: Tobias Welz <tw@wiznet.eu>
WIZnet WizFi630S is using only 3 of the phy ports. The unused phy ports
draw unnecessarily power. This is disabling the unused phy ports.
Signed-off-by: Tobias Welz <tw@wiznet.eu>
For the TP-Link 4M devices with tplink-v2-image recipe
(mktplinkfw2.c), there are two different flash layouts based
on the size of the (u)boot partition:
device uboot OEM firmware OpenWrt (incl. config)
tl-wr840n-v5 0x20000 0x3c0000 0x3d0000
tl-wr841n-v14 0x10000 0x3d0000 0x3e0000
In both cases, the 0x10000 config partition is used for the firmware
partition as well due to the limited space available and since it's
recreated by the OEM firmware anyway.
However, the TFTP flashing process will only copy data up to the
size of the initial (OEM) firmware size. Therefore, while we can
use the bigger partition to have additional erase blocks on the
device, we have to limit the image sizes to the TFTP limits.
So far, only one layout definition has been set up in mktplinkfw2.c
for 4M mediatek devices. This adds a second one and assigns them
to the devices so the image sizes are correctly restrained.
Signed-off-by: Adrian Schmutzler <freifunk@adrianschmutzler.de>
ifconfig is effectively deprecated for quite some time now. Let's
replace the remaining occurrences for our target setup by the
corresponding ip commands now.
Note that this does not touch ar71xx, as it will be dropped anyway,
and changing it would only make backports harder.
Signed-off-by: Adrian Schmutzler <freifunk@adrianschmutzler.de>
The Arduino Yun uses a baud rate of 250000 by default. The serial is
going over the Atmel ATmega and is used to connect to this chip.
Without this patch Linux wants to switch the console to 9600 Baud.
With this patch Linux will use the configured baud rate and not a
default one specified in uart_register_driver().
This has been added for ath79 4.19 and 5.4 in master as part of
fc59b2f79b50 ("ath79: add support for Arduino Yun"), this backports
it separately to 4.14.
Signed-off-by: Adrian Schmutzler <freifunk@adrianschmutzler.de>
Commit 1bfbf2de6d ("ar71xx: serial: core: add support for boot console
with arbitrary baud rates") added support for arbitrary baud rates which
enabled 250000 baud rate for Yun. But the patch was not ported to kernel
4.9, and since then the kernel set its baud rate to 9600. This commit ports
the patch to kernel 4.14, thereby restoring the serial console of Yun.
Cc: Hauke Mehrtens <hauke@hauke-m.de>
Signed-off-by: Sungbo Eo <mans0n@gorani.run>
Commit 05d73a2a73 enabled GPIO on ethernet LED, but proper LED setup was
not added then. This commit fixes it by reverting the change on the LED.
Fixes: 05d73a2a73 ("ar71xx: Arduino Yun board 'WLAN RST' button support")
Signed-off-by: Sungbo Eo <mans0n@gorani.run>
Commit bb46b635df changed its partition scheme, but sysupgrade image
validation still uses the old format. This commit fixes it so that
force flag is not needed for sysupgrade.
Fixes: bb46b635df ("ar71xx: move Arduino Yun to generic building code")
Signed-off-by: Sungbo Eo <mans0n@gorani.run>
This symbol is exposed on ARM64 with EFI enabled in the kernel config.
Currently this happens only on ipq807x, but as there might be new ARM64
targets with EFI in the future it is better to add the symbol to the
generic config.
Signed-off-by: Stijn Tintel <stijn@linux-ipv6.be>
Acked-by: Jo-Philipp Wich <jo@mein.io>
Don't explicitely disable options in target/linux/generic/config-* if
they are already controlled in config/Config-kernel.in.
Add a bunch of new symbols and prepare defaults for using only unified
hierarchy (ie. cgroup2). Update symbol dependencies while at it
Signed-off-by: Daniel Golle <daniel@makrotopia.org>
Sometimes when using the DNS-313 memory usage can peak and
with a simple swap partition we can avoid running into the
roof and invoking the OOM killer. Set this partition to
128MB (twice the size of the memory of the DNS-313).
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
ZyXEL Keenetic has a USB port. Thus, DWC2 USB controller driver should
be in the default image for this device.
Fixes: a7cbf59e0e ("ramips: add new device ZyXEL Keenetic as kn")
Signed-off-by: Alexey Dobrovolsky <dobrovolskiy.alexey@gmail.com>
[fixed whitespace issue]
Signed-off-by: Petr Štetiar <ynezz@true.cz>
(backported from commit 0a182fcba6)
Signed-off-by: Adrian Schmutzler <freifunk@adrianschmutzler.de>
In FS#2738 we can see that patch first introduced in
e8ebcff ("ramips: add a explicit reset to dwc2")
breaks USB functionality since 18.06. Thus, this patch should be removed.
Removed:
- 0032-USB-dwc2-add-device_reset.patch
Fixes: FS#2738
Fixes: FS#2964
Signed-off-by: Alexey Dobrovolsky <dobrovolskiy.alexey@gmail.com>
(cherry picked from commit ab841b4393)
Now that check-size uses IMAGE_SIZE by default, we can skip the argument from
image recipes to reduce redundancy.
Signed-off-by: Sungbo Eo <mans0n@gorani.run>
[do not touch ar71xx]
Signed-off-by: Adrian Schmutzler <freifunk@adrianschmutzler.de>
