Compaction is the only memory management component to form high order (larger
physically contiguous) memory blocks reliably. The page allocator relies on
compaction heavily and the lack of the feature can lead to unexpected OOM
killer invocations for high order memory requests. You shouldn't disable this
option unless there really is a strong reason for it.
Signed-off-by: Felix Fietkau <nbd@nbd.name>
Signed-off-by: Michal Hrusecky <michal.hrusecky@nic.cz>
Optimized inlining was disabled by default when gcc 4 was still
relatively new. By now, all gcc versions handle this well and there
seems to be no real reason to keep it x86-only.
Signed-off-by: Felix Fietkau <nbd@nbd.name>
This adds a configuration options which is needed now.
Without this patch the geode build will fail.
Fixes: 4eda2fddf2 ("x86/geode: enable X86_INTEL_LPSS to select PINCTRL")
Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
Without UHCI a non-trivial number of machines will have no keyboard
without BIOS assistance.
Add XHCI as well in case there are chipsets which don't support legacy
interfaces, and support PCI OHCI controllers also.
Signed-off-by: Alex Maclean <monkeh@monkeh.net>
THIN_ARCHIVES option is enabled by default in the kernel configuration
and no one target config disables it. So enable it by default and remove
this symbol from target specific configs to keep them light.
Signed-off-by: Sergey Ryazanov <ryazanov.s.a@gmail.com>
New FUTEX_PI configuration symbol enabled if FUTEX and RT_MUTEX symbols
are enabled. Both of these symbols are enabled by default in the
generic config, so enable FUTEX_PI by default too to keep platform
specific configs minimal.
Signed-off-by: Sergey Ryazanov <ryazanov.s.a@gmail.com>
OVERLAY_FS config symbol selects EXPORTFS since 4.12 kernel, we have
OVERLAY_FS enabled by default, so enable EXPORTFS in the generic config
of 4.14 and remove this option from platform specific configs.
Signed-off-by: Sergey Ryazanov <ryazanov.s.a@gmail.com>
DRM_LIB_RANDOM config symbol selected only by DRM_DEBUG_MM_SELFTEST
which is disable by default, so disable DRM_LIB_RANDOM by default too.
Signed-off-by: Sergey Ryazanov <ryazanov.s.a@gmail.com>
These options do not used by any supported arch, so disable them by
default to make arch configs a bit more clean.
Signed-off-by: Sergey Ryazanov <ryazanov.s.a@gmail.com>
Only one arch (x86_64) enables this option. So disable
ARCH_WANTS_THP_SWAP by default and remove referencies to it from all
configs (except x86_64) to make them clean.
Signed-off-by: Sergey Ryazanov <ryazanov.s.a@gmail.com>
Disable CONFIG_EFI_VARS, since it suffers from sysfs limitation (no
support for variable longer than 1024 bytes).
kmod-fs-efivarfs is the replacement of this, which enables mounting
efivarfs file system and doesn't suffer from 1024 bytes limitation.
Signed-off-by: Alif M. Ahmad <alive4ever@live.com>
There is a new APU-model available, APU3. The device is configured in
the same way as the APU1 and APU2, so the same LED/network setup can be
used.
I considered changing the case to pc-engines-apu*, but I chose to follow
the existing pattern and add the full board name.
Signed-off-by: Kristian Evensen <kristian.evensen@gmail.com>
Support for kernel 4.14 was added 2 months ago, make it now the default
kernel to use for the x86 target.
Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
Tested-by: Stijn Segers <francesco.borromini@inventati.org>
This adds support for the new configuration option CONFIG_RETPOLINE and
refreshes the configuration.
Fixes: d8565a06dc ("kernel: bump 4.9 to 4.9.77")
Fixes: 9ddfac8015 ("kernel: bump 4.14 to 4.14.14")
Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
The KEXEC_FILE symbol exists for X86 since kernel 3.17, and since 4.10
for PPC64. Add it to x86/config-4.9 and to generic/config-4.14.
Signed-off-by: Stijn Tintel <stijn@linux-ipv6.be>
While working on a new target (meson), the kernel build failed due to
missing DRM_DEBUG_MM_SELFTEST symbol. This can potentially happen on all
targets that enable DRM drivers in the kernel config or via kmod
packages, so add it to the generic config and remove it from x86
subtarget configs, together with DRM_DEBUG_MM.
Signed-off-by: Stijn Tintel <stijn@linux-ipv6.be>
Supermicro puts "Super Server" into their product_name DMI value
for a whole slew of products, making this value about as useful
as not having been filled in at all. Instead, fall back on the
board_name instead.
Signed-off-by: Philip Prindeville <philipp@redfish-solutions.com>
There might be other places (such as vendor-supplied preinit scripts)
where we wish to take a DMI name and clean it up in a consistent way,
so make the sed command into a function.
Signed-off-by: Philip Prindeville <philipp@redfish-solutions.com>
Rename unwinder config symbols to match upstream changes.
Refresh patches.
Update patch that no longer applies: 202-reduce_module_size.patch
Also enable CONFIG_PAGE_TABLE_ISOLATION. This feature was backported
from 4.15 to the 4.14 stable series. It is enabled by default, so enable
it in OpenWrt as well.
Compile-tested on x86/64.
Runtime-tested on x86/64.
Signed-off-by: Stijn Tintel <stijn@linux-ipv6.be>
A lot of 1U x86_64 servers have NVMe support, which is lower-power
and higher speed than SSD or CFast drives, etc. The drivers
required to make this work are trivial.
Signed-off-by: Philip Prindeville <philipp@redfish-solutions.com>
This adds basic support for kernel 4.14, this was tested in qemu only.
The subtarget configuration was refresh with kernel 4.14 and the
options needed to make it compile on kernel 4.9 were added manually.
Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
Just refresh the kernel configuration, some options are removed because
they are now in the generic kernel configuration.
Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
Cover temperature sensors for all mainstream 64-bit processors, including
AMD 10h and 15h families, Intel iCore, Xeon, Atom, and Via Nano. Also
add CPUID support for user-space applications to detect CPU type.
Include the on-chip sensors for 64-bit CPU's in the generic profile
in case someone builds a 32-bit kernel to run on a Xeon SoC, etc.
Signed-off-by: Philip Prindeville <philipp@redfish-solutions.com>
The Xen serial console has been broken since the xen_domu subtarget
was merged in the generic x86 subtarget (commits 1d6879ee and 371b382a).
The reason for the broken serial console seems to be an IRQ conflict
between the serial console driver and the PATA_LEGACY driver:
[ 1.330125] genirq: Flags mismatch irq 8. 00000000 (hvc_console) vs. 00000000 (platform[pata_legacy.4])
[ 1.330134] hvc_open: request_irq failed with rc -16.
[ 1.330148] Warning: unable to open an initial console.
Just drop the PATA_LEGACY driver from the x86/generic and x86_64
subtargets, since this driver is marked experimental and only supports
very old ISA devices anyway. It is still included in the x86/legacy
subtarget where it rightfully belongs.
Fixes: FS#787
Signed-off-by: Baptiste Jonglez <git@bitsofnetworks.org>
Xen support for x86/generic was added in 1d6879ee. This commit also
enables it for x86/64.
This was successfully tested with Xen 4.5, although the serial console
is broken in the same way as x86/generic (see FS#787)
Signed-off-by: Baptiste Jonglez <git@bitsofnetworks.org>
All x86 subtargets enable USB support, so it makes sense to enable it
in the target config instead, to avoid duplication.
Also refresh subtarget configs accordingly.
Signed-off-by: Baptiste Jonglez <git@bitsofnetworks.org>
This was done by simply running `make kernel_menuconfig CONFIG_TARGET=subtarget`
and then saving without changing any option.
Most of the removed options can be explained because they are already
present in the target config or in the generic 4.9 config:
- PAE-related options, enabled by default on x86 by 961c0eac
- LZO-related options, enabled by default since 4.9
As far as I understand the build system, this shouldn't have any
user-visible impact, because the build system already merges the
various kernel configs during build.
Signed-off-by: Baptiste Jonglez <git@bitsofnetworks.org>
This subtarget was added by 961c0eac, probably by mistake. It does
not contain anything beside a kernel config.
Acked-by: Daniel Golle <daniel@makrotopia.org>
Cc: Daniel Golle <daniel@makrotopia.org>
Signed-off-by: Baptiste Jonglez <git@bitsofnetworks.org>
Do not parse /tmp/sysinfo/board_name, /proc/cpuinfo or the device tree
compatible string directly. Always use the board_name function to get
the board name.
The admswconfig package still reads /proc/cpuinfo directly. The code
looks somehow broken and the whole adm5120 which uses this package
looks unmaintained. Leave it as it is for now.
Signed-off-by: Mathias Kresin <dev@kresin.me>
For targets using the generic board detection and board specific
settings in diag.sh, the board name is still unset at the time the
set_state() provided by diag.sh is called by 10_indicate_preinit.
Change the execution order to ensure the boardname is populated before
required the first time. Do the target specific board detection as
early as possible, directly followed by the generic one to allow a
seamless switch to the generic function for populating /tmp/sysinfo/.
Signed-off-by: Mathias Kresin <dev@kresin.me>
CONFIG_SG_POOL symbol is selected only by CONFIG_SCSI, since the last
one is disabled by default then disable CONFIG_SG_POOL by default too.
And explicitly enable it only for platforms that use CONFIG_SCSI.
Signed-off-by: Sergey Ryazanov <ryazanov.s.a@gmail.com>
