BugLink: https://bugs.launchpad.net/bugs/1953613
Enable v1 of the character device interface for manipulating GPIOs from
userspace. The latest released libgpiod2 library and gpiod tools are
still using this interface and it may be a while before every userspace
program has been ported to a newer version.
Signed-off-by: Emil Renner Berthing <emil.renner.berthing@canonical.com> Signed-off-by: Andrea Righi <andrea.righi@canonical.com>
Juerg Haefliger [Wed, 9 Mar 2022 09:50:33 +0000 (10:50 +0100)]
UBUNTU: Remove ubuntu/hio driver
The third-party Huawei hio driver provided by the Ubuntu kernel was
added back in Xenial (LP: #1603483). It has been disabled since Impish
because it no longer compiles. Nobody has complained so far, so get rid
of it.
Signed-off-by: Juerg Haefliger <juergh@canonical.com> Signed-off-by: Paolo Pisati <paolo.pisati@canonical.com> Signed-off-by: Andrea Righi <andrea.righi@canonical.com>
Po-Hsu Lin [Fri, 8 Apr 2022 07:41:44 +0000 (15:41 +0800)]
UBUNTU: [Debian] Use kernel-testing repo from launchpad
BugLink: https://bugs.launchpad.net/bugs/1968016
Reflect the change with kernel-testing repo moved from kernel.ubuntu.com
to launchpad, where we host all of our testing tools.
Use shallow clone to save some bandwidth / time.
Signed-off-by: Po-Hsu Lin <po-hsu.lin@canonical.com> Signed-off-by: Andrea Righi <andrea.righi@canonical.com>
You-Sheng Yang [Mon, 11 Apr 2022 09:24:34 +0000 (17:24 +0800)]
UBUNTU: SAUCE: vmd: fixup bridge ASPM by driver name instead
BugLink: https://bugs.launchpad.net/bugs/1942160
Additional VMD bridge IDs needed for new Alder Lake platforms, but
actually there is no a complete list for them. Here we match bridge
devices if they're directly attached to a VMD controller instead.
Signed-off-by: You-Sheng Yang <vicamo.yang@canonical.com> Signed-off-by: Andrea Righi <andrea.righi@canonical.com>
UBUNTU: SAUCE: PCI/ASPM: Enable LTR for endpoints behind VMD
BugLink: https://bugs.launchpad.net/bugs/1942160
In addition to ASPM, LTR also needs to be programmed with a reasonable
value to let PCIe link reaches L1.2.
For now, program a hardcoded value that is used under Windows.
While at it, consolidate ASPM and LTR enabling logic to share a same pci
device table.
Signed-off-by: Kai-Heng Feng <kai.heng.feng@canonical.com> Signed-off-by: Andrea Righi <andrea.righi@canonical.com>
As BIOS may not be able to program the config space for devices under
VMD domain, ASPM needs to be programmed manually by software. This is
also the case under Windows.
The VMD controller itself is a root complex integrated endpoint that
doesn't have ASPM capability, so we can't propagate the ASPM settings to
devices under it. Hence, simply apply ASPM_STATE_ALL to the links under
VMD domain, unsupported states will be cleared out anyway.
Signed-off-by: Kai-Heng Feng <kai.heng.feng@canonical.com> Signed-off-by: Andrea Righi <andrea.righi@canonical.com>
UBUNTU: [Packaging] Double the speed of updateconfigs
Use parallel build setting when building menuconfig.
mvo says "funny, make menuconfig feels like it takes the same amount
of time to compile as it did 15y ago when I used it last"
Signed-off-by: Dimitri John Ledkov <dimitri.ledkov@canonical.com> Acked-by: Tim Gardner <tim.gardner@canonical.com> Signed-off-by: Andrea Righi <andrea.righi@canonical.com>
Andrea Righi [Fri, 15 Jul 2022 17:40:33 +0000 (19:40 +0200)]
UBUNTU: [Config] enable configs for fixing kernel won't load mok
BugLink: https://bugs.launchpad.net/bugs/1972802 Signed-off-by: Ivan Hu <ivan.hu@canonical.com> Signed-off-by: Timo Aaltonen <timo.aaltonen@canonical.com> Signed-off-by: Andrea Righi <andrea.righi@canonical.com>
UBUNTU: [Config] Enable CONFIG_KEXEC_FILE on s390x
BugLink: https://bugs.launchpad.net/bugs/1981437
KEXEC_FILE and KEXEC_SIG should also be enabled on s390x, as this
architecture supports signed kernels as well.
Signed-off-by: Dimitri John Ledkov <dimitri.ledkov@canonical.com>
UBUNTU: [Config] Set CONFIG_MTD_PSTORE=m on riscv64
BugLink: https://bugs.launchpad.net/bugs/1981437
This enables the module to log panic/oops to a circular buffer in an MTD
flash partition. All other architectures enables this module, so let's
enable it on riscv64 too.
Signed-off-by: Emil Renner Berthing <emil.renner.berthing@canonical.com> Signed-off-by: Dimitri John Ledkov <dimitri.ledkov@canonical.com>
UBUNTU: [Config] Enable CONFIG_KEXEC_FILE on riscv64
BugLink: https://bugs.launchpad.net/bugs/1981437
This is enabled on all architectures but s390x and was added to RISC-V
in v5.19-rc1, so let's enable it until we have a reason not to.
Signed-off-by: Emil Renner Berthing <emil.renner.berthing@canonical.com> Signed-off-by: Dimitri John Ledkov <dimitri.ledkov@canonical.com>
UBUNTU: [Config] Enable CONFIG_VMAP_STACK on riscv64
BugLink: https://bugs.launchpad.net/bugs/1981437
Enable virtually-mapped kernel stacks like all other architectures.
This creates guard pages which will cause kernel stack overflows to be
caught immediately.
Signed-off-by: Emil Renner Berthing <emil.renner.berthing@canonical.com> Signed-off-by: Dimitri John Ledkov <dimitri.ledkov@canonical.com>
UBUNTU: [Config] Enable CONFIG_STRICT_DEVMEM on riscv64
BugLink: https://bugs.launchpad.net/bugs/1981437
Filter access to /dev/mem like is done on all other architectures. There
is even a mark<ENFORCED> annotation on this option and yet it is still
disabled on riscv64. Let's fix that.
Signed-off-by: Emil Renner Berthing <emil.renner.berthing@canonical.com> Signed-off-by: Dimitri John Ledkov <dimitri.ledkov@canonical.com>
UBUNTU: [Config] Enable CONFIG_SUNRPC_DISABLE_INSECURE_ENCTYPES on riscv64
BugLink: https://bugs.launchpad.net/bugs/1981437
This disables encryption types that were deprecated in Kerberos v5 as is
done on all other architectures. No setup involving a RISC-V machine
should use these encryption types.
Signed-off-by: Emil Renner Berthing <emil.renner.berthing@canonical.com> Signed-off-by: Dimitri John Ledkov <dimitri.ledkov@canonical.com>
UBUNTU: [Packaging] mark dkms-build-configure--zfs executable
Mark dkms-build-configure--zfs script executable, such that one
doesn't need to call it with an explicit shell.
Signed-off-by: Dimitri John Ledkov <dimitri.ledkov@canonical.com> Acked-by: Tim Gardner <tim.gardner@canonical.com> Signed-off-by: Paolo Pisati <paolo.pisati@canonical.com> Signed-off-by: Dimitri John Ledkov <dimitri.ledkov@canonical.com>
UBUNTU: [Packaging] Fix dkms builds with linker @module.mod files
v5.19 kernel build stopped passing a long list of .o object files to
the linker. Instead it started to emit linker .o files into .mod
file. And then pass said .mod file as a single argument to the linker.
Adjust dkms-build and dkms-build--nvidia-N scripts to account for this
behaviour. Copy .mod files in addition to .o files. Post-process .mod
to remove absolute build paths, and covert them to relative object
paths.
Signed-off-by: Dimitri John Ledkov <dimitri.ledkov@canonical.com> Signed-off-by: Andrea Righi <andrea.righi@canonical.com>
UBUNTU: [Packaging] Always catch errors in dkms-build scripts
Enforce `set -e` execution of the dkms-build scripts, even when
overall SHELL is not set to `bash -e`. This enforces that dkms-build
scripts catch errors, even when building without SHELL variable, like
it is done by the LRM packages.
Signed-off-by: Dimitri John Ledkov <dimitri.ledkov@canonical.com> Acked-by: Tim Gardner <tim.gardner@canonical.com> Signed-off-by: Paolo Pisati <paolo.pisati@canonical.com> Signed-off-by: Andrea Righi <andrea.righi@canonical.com>
UBUNTU: [Packaging] Fix bashism in dkms-build script
dkms-build script uses bashism variable and test comparison, replace
them with dash compatible invocations. This allows using dkms-build
scripts without SHELL variable set to "bash -e", like it is done by
LRM packages.
Signed-off-by: Dimitri John Ledkov <dimitri.ledkov@canonical.com> Acked-by: Tim Gardner <tim.gardner@canonical.com> Signed-off-by: Paolo Pisati <paolo.pisati@canonical.com> Signed-off-by: Andrea Righi <andrea.righi@canonical.com>
Ike Panhc [Fri, 29 Apr 2022 06:45:58 +0000 (14:45 +0800)]
UBUNTU: [Config] CONFIG_HISI_PMU=m
BugLink: https://launchpad.net/bugs/1956086 Signed-off-by: Ike Panhc <ike.pan@canonical.com> Acked-by: Stefan Bader <stefan.bader@canonical.com> Acked-by: Tim Gardner <tim.gardner@canonical.com> Signed-off-by: Andrea Righi <andrea.righi@canonical.com>
Kai-Heng Feng [Thu, 5 May 2022 06:20:14 +0000 (14:20 +0800)]
UBUNTU: SAUCE: net: phy: marvell: Skip setting LED on Dell EMC board
BugLink: https://bugs.launchpad.net/bugs/1971667
The board in question requires certain LED config, which is already
configured by BIOS, to correctly show its networking status. However,
Marvell PHY driver hardcodes LED value so we need a way to preserve the
default set by BIOS.
PHY maintainer asked for a "generic" approach which goes no where [1],
so let's move on and use a quirk to handle it.
Andrea Righi [Mon, 4 Jul 2022 14:10:40 +0000 (16:10 +0200)]
UBUNTU: [Packaging] ignore warnings from the output of ld.bfd
When generating the script BUILD (or CLEAN) we are taking all the ld.bfd
commands fromt he build log, however if the log contains warnings or
NOTEs such as:
/usr/bin/ld.bfd: warning: nvidia.o: requires executable stack (because the .note.GNU-stack section is executable)
we would add also that to the script, causing syntax error failures.
Make sure to exclude all warnings to prevent problems like this.
Signed-off-by: Andrea Righi <andrea.righi@canonical.com>
BugLink: https://bugs.launchpad.net/bugs/1980061
This sets CONFIG_COMPAT=n for riscv64 kernels. Enabling it allows 32bit
binaries to be run on 64bit kernels, but requires hardware support. So
far no chips have been released that support it and neither does
upstream Qemu. Also Ubuntu doesn't ship 32bit RISC-V binaries, so
disable this feature for now.
Signed-off-by: Emil Renner Berthing <emil.renner.berthing@canonical.com> Acked-By: Dimitri John Ledkov <dimitri.ledkov@canonical.com> Signed-off-by: Andrea Righi <andrea.righi@canonical.com>
UBUNTU: [Config] Merge riscv64 config and annotations
BugLink: https://bugs.launchpad.net/bugs/1979647
This adds the debian.master/config/riscv64 directory and merges
annotations and configuration from kinetic:linux-riscv.
Merging the riscv64 configuration reveals a lot of differences from the
other architectures, but this commit tries to keep the configuration as
close as possible to kinetic:linux-riscv.
Signed-off-by: Emil Renner Berthing <emil.renner.berthing@canonical.com> Acked-by: Tim Gardner <tim.gardner@canonical.com> Acked-by: Paolo Pisati <paolo.pisati@canonical.com> Signed-off-by: Andrea Righi <andrea.righi@canonical.com>
Juerg Haefliger [Fri, 24 Jun 2022 12:41:53 +0000 (14:41 +0200)]
UBUNTU: [Packaging] final-checks: Remove useless sourcing of kernelconfig
kernelconfig only defines 'archs' but 'archs' is overwritten after the
fact so remove the uselsess sourcing. While at it, remove a stray leading
space in the following line.
Signed-off-by: Juerg Haefliger <juerg.haefliger@canonical.com> Signed-off-by: Andrea Righi <andrea.righi@canonical.com>
Juerg Haefliger [Mon, 9 May 2022 09:12:06 +0000 (11:12 +0200)]
UBUNTU: [Packaging] kernelconfig: Bubble up warnings and errors
Config annotation check failures and warnings due to incomplete config
operations are really bad, so exit the script with a non-zero status if
such errors or warnings are detected.
Ignore: yes Signed-off-by: Juerg Haefliger <juergh@canonical.com> Acked-by: Tim Gardner <tim.gardner@canonical.com> Signed-off-by: Andrea Righi <andrea.righi@canonical.com>
projects / mirror_ubuntu-kernels.git / log