| Commit message (Collapse) | Author | Age | Files | Lines |
... | |
|
|
|
|
|
| |
The UNWINDER_ARM helps to get proper stack traces.
Signed-off-by: Robert Schwebel <r.schwebel@pengutronix.de>
|
|
|
|
|
|
|
| |
'reason' suggests: "Trace events for GPIOs are helpful and low-overhead,
so don't disable them if tracing is enabled globally.", so enable them.
Signed-off-by: Robert Schwebel <r.schwebel@pengutronix.de>
|
|
|
|
|
|
| |
The new kernel has been released today, so update.
Signed-off-by: Robert Schwebel <r.schwebel@pengutronix.de>
|
|
|
|
|
|
| |
The platform supports this hardening option, enable it.
Signed-off-by: Robert Schwebel <r.schwebel@pengutronix.de>
|
|
|
|
|
|
|
|
| |
The compiler version is part of the kernelconfig, so changing the
toolchain changes the kernelconfig too. Update the kernelconfig with a
simple 'ptxdist oldconfig kernel'.
Signed-off-by: Roland Hieber <rhi@pengutronix.de>
|
|
|
|
|
|
| |
Update the kernel to Linux 5.0
Signed-off-by: Robert Schwebel <r.schwebel@pengutronix.de>
|
|
|
|
| |
Signed-off-by: Robert Schwebel <r.schwebel@pengutronix.de>
|
|
|
|
| |
Signed-off-by: Rouven Czerwinski <rouven@czerwinskis.de>
|
|
|
|
|
|
|
| |
This tries to use the host compiler directly. Disable it until we either have a
patch for the compiler selection or a fix is merged upstream.
Signed-off-by: Rouven Czerwinski <r.czerwinski@pengutronix.de>
|
|
|
|
|
|
|
|
| |
This matches the default value for this symbol and the two other
platforms use 7, too. This makes the machine emit the "normal" logging
when the loglevel=5 parameter is removed from the kernel command line.
Signed-off-by: Uwe Kleine-König <u.kleine-koenig@pengutronix.de>
|
|
|
|
| |
Signed-off-by: Robert Schwebel <r.schwebel@pengutronix.de>
|
|
|
|
|
|
| |
kernel.org has released 4.18, update to the new kernel.
Signed-off-by: Robert Schwebel <r.schwebel@pengutronix.de>
|
|
|
|
|
|
| |
kernel.org has released 4.17, update to the new kernel.
Signed-off-by: Robert Schwebel <r.schwebel@pengutronix.de>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
This got lost in commit 5dc9701fe9ec ("platform-v7a: kernel version bump
4.15 -> 4.16"), and resulted in a qemu system no longer able to mount
external 9p devices:
$ ./configs/platform-v7a/run 9p
[...]
[ 1.799073] 9pnet_virtio: no channels available for device /dev/root
[ 1.800620] VFS: Cannot open root device "root" or unknown-block(0,0): error -2
[ 1.802147] Kernel panic - not syncing: VFS: Unable to mount root fs on unknown-block(0,0)
Fixes: 5dc9701fe9ec ("platform-v7a: kernel version bump 4.15 -> 4.16")
Signed-off-by: Roland Hieber <r.hieber@pengutronix.de>
|
|
|
|
| |
Signed-off-by: Robert Schwebel <r.schwebel@pengutronix.de>
|
|
|
|
|
|
| |
Update platform-v7a to the latest released kernel.
Signed-off-by: Robert Schwebel <r.schwebel@pengutronix.de>
|
|
|
|
| |
Signed-off-by: Michael Olbrich <m.olbrich@pengutronix.de>
|
|
|
|
|
|
|
|
| |
Access to normal system memory is usually not needed and it makes
compromising the kernel more difficult. IO addresses are still accessible
as long as CONFIG_IO_STRICT_DEVMEM is disabled.
Signed-off-by: Michael Olbrich <m.olbrich@pengutronix.de>
|
|
|
|
|
|
|
|
|
|
|
| |
This option checks for a stack overrun on calls to schedule().
If the stack end location is found to be over written always panic as
the content of the corrupted region can no longer be trusted.
This is to ensure no erroneous behaviour occurs which could result in
data corruption or a sporadic crash at a later stage once the region
is examined. The runtime overhead introduced is minimal.
Signed-off-by: Michael Olbrich <m.olbrich@pengutronix.de>
|
|
|
|
|
|
|
| |
If the bootloader has enabled the watchdog, the intention is usually to
monitor the whole boot process.
Signed-off-by: Michael Olbrich <m.olbrich@pengutronix.de>
|
|
|
|
|
|
|
|
| |
Keeping userspace processes without CAP_SYS_RAWIO from writing to low pages
can help reduce the impact of kernel NULL pointer bugs. 32kB is recommened
on ARM.
Signed-off-by: Michael Olbrich <m.olbrich@pengutronix.de>
|
|
|
|
|
|
| |
4.14 is out, update the kernel.
Signed-off-by: Robert Schwebel <r.schwebel@pengutronix.de>
|
|
|
|
|
|
|
| |
It saves power and the i.MX6/fec problem is fixed since v4.8.0.
Signed-off-by: Michael Olbrich <m.olbrich@pengutronix.de>
Signed-off-by: Robert Schwebel <r.schwebel@pengutronix.de>
|
|
|
|
|
| |
Signed-off-by: Michael Olbrich <m.olbrich@pengutronix.de>
Signed-off-by: Robert Schwebel <r.schwebel@pengutronix.de>
|
|
|
|
|
| |
Signed-off-by: Michael Olbrich <m.olbrich@pengutronix.de>
Signed-off-by: Robert Schwebel <r.schwebel@pengutronix.de>
|
|
|
|
|
| |
Signed-off-by: Michael Olbrich <m.olbrich@pengutronix.de>
Signed-off-by: Robert Schwebel <r.schwebel@pengutronix.de>
|
|
|
|
|
|
|
| |
It's a debug feature with performance impact.
Signed-off-by: Michael Olbrich <m.olbrich@pengutronix.de>
Signed-off-by: Robert Schwebel <r.schwebel@pengutronix.de>
|
|
|
|
|
| |
Signed-off-by: Michael Olbrich <m.olbrich@pengutronix.de>
Signed-off-by: Robert Schwebel <r.schwebel@pengutronix.de>
|
|
|
|
|
| |
Signed-off-by: Michael Olbrich <m.olbrich@pengutronix.de>
Signed-off-by: Robert Schwebel <r.schwebel@pengutronix.de>
|
|
|
|
|
|
|
|
| |
On vexpress, this lowers the 2nd pass boot time from 38 s to 31 s.
Signed-off-by: Michael Olbrich <m.olbrich@pengutronix.de>
[add benchmark numbers to commit message]
Signed-off-by: Robert Schwebel <r.schwebel@pengutronix.de>
|
|
|
|
|
|
|
| |
Change the kernel config to reboot on panic.
Signed-off-by: Robert Schwebel <r.schwebel@pengutronix.de>
Reviewed-by: Rouven Czerwinski <r.czerwinski@pengutronix.de>
|
|
|
|
|
|
|
|
| |
"reason" suggests that the typical embedded use-cases are more sensitive
to latency than throughput, so a fully preemptible kernel is preferred.
Signed-off-by: Robert Schwebel <r.schwebel@pengutronix.de>
Reviewed-by: Rouven Czerwinski <r.czerwinski@pengutronix.de>
|
|
|
|
|
|
|
|
|
|
| |
"Reason" suggests that the kernel lockup detector should be active, as
it's low overhead and potentially provides useful debug output when the
system is failing due to a lockup. This patch does the necessary changes
for the kernel configuration.
Signed-off-by: Robert Schwebel <r.schwebel@pengutronix.de>
Reviewed-by: Rouven Czerwinski <r.czerwinski@pengutronix.de>
|
|
|
|
| |
Signed-off-by: Robert Schwebel <r.schwebel@pengutronix.de>
|
|
|
|
|
|
|
|
| |
As suggested by "reason", activate the kernel feature to randomize the
freelist order used on creating new pages. This results in a reduction
of the predictability of addresses against heap overflows.
Signed-off-by: Robert Schwebel <r.schwebel@pengutronix.de>
|
|
|
|
|
|
|
| |
As suggested by "reason", activate the stack protection feature. The
only negative impact should be a small increase in kernel size.
Signed-off-by: Robert Schwebel <r.schwebel@pengutronix.de>
|
|
|
|
|
|
|
| |
As suggested by "reason", activate the kernel option to protect against
obviously wrong memory regions when copying memory to/from the kernel.
Signed-off-by: Robert Schwebel <r.schwebel@pengutronix.de>
|
|
|
|
|
|
| |
As suggested by "reason", restrict access to the dmesg output to root.
Signed-off-by: Robert Schwebel <r.schwebel@pengutronix.de>
|
|
|
|
|
|
|
| |
As suggested by "reason", disable quota support, as we don't need it on
embedded systems.
Signed-off-by: Robert Schwebel <r.schwebel@pengutronix.de>
|
|
|
|
|
|
| |
This change was suggested by reason-ptxdist.
Signed-off-by: Robert Schwebel <r.schwebel@pengutronix.de>
|
|
|
|
|
|
|
|
|
|
| |
reason-ptxdist suggests: "fq_codel helps fight the network bufferbloat problem.
It is believed to be a good default with no tuning required for most workloads.
Downstream distributions may override this choice. On 10Gbit servers that do
not do forwarding, "fq" may perform better. Systems without a good clocksource
should use "pfifo_fast". (from systemd NEWS)"
Signed-off-by: Robert Schwebel <r.schwebel@pengutronix.de>
|
|
|
|
|
|
|
|
| |
reason-ptxdist suggests: "The kernel lockup detector should be active,
as it's low overhead and potentially provides useful debug output when
the system is failing due to a lockup."
Signed-off-by: Robert Schwebel <r.schwebel@pengutronix.de>
|
|
|
|
|
|
|
| |
As suggested by reason-ptxdist: "The CAAM driver should be enabled on
MX6 to provide entropy to the kernel."
Signed-off-by: Robert Schwebel <r.schwebel@pengutronix.de>
|
|
|
|
|
|
|
|
|
| |
Deactivate support for ext2 + ext3, as our rootfs is ext4. Add a hint to
the kernel command line, in order to avoid this warning:
EXT4-fs (mmcblk0p1): couldn't mount as ext2 due to feature incompatibilities
Signed-off-by: Robert Schwebel <r.schwebel@pengutronix.de>
|
|
|
|
|
|
| |
Update the kernel version.
Signed-off-by: Robert Schwebel <r.schwebel@pengutronix.de>
|
|
|
|
|
|
|
|
|
| |
As i.MX6 is an ARM v7a as well, add support for it to the v7a kernel
configuration.
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
[improve commit message]
Signed-off-by: Robert Schwebel <r.schwebel@pengutronix.de>
|
|
|
|
| |
Signed-off-by: Robert Schwebel <r.schwebel@pengutronix.de>
|
|
|
|
| |
Signed-off-by: Robert Schwebel <r.schwebel@pengutronix.de>
|
|
|
|
|
|
|
| |
Qemu is able to use the plan 9 filesystem with virtio to access a local
filesystem hierarchy. Activate.
Signed-off-by: Robert Schwebel <r.schwebel@pengutronix.de>
|
|
|
|
|
|
|
|
| |
This avoids that systemd cries about missing user namespaces:
[ 3.118703] systemd[1]: dev-hugepages.mount: Couldn't determine result for ConditionVirtualization=!private-users, assuming failed: No such file or directory
Signed-off-by: Robert Schwebel <r.schwebel@pengutronix.de>
|