As memory hotplug for arm64 by acpi is not ready on qemu, we choose
"probe" instead. You can refer to [1] to get more infomation about
"probe". The process of memory hotplug by "probe" in kata lies below:
firstly, add memory in qemu qmp; secondly, echo the start phyical address
of that memory to /sys/devices/system/memory/probe, which will be done
through kata-agent; thirdly, excute online op, then this newly added
memory is capable to be used.
All functions in this patch will be called after "echo" op. It can be
divided into two parts:
1. create page table for that memory;
2. add that memory to memblock.
In this patch, NUMA must be turned off for not all arm64 machine supports
NUMA.
As the newly added memory should be placed from 2T to 6T which is decided
in qemu and phyical address and virtual address will be one-one mapping
when create pgd for that memory, we must config ARM64_VA_BITS as 48.
Also some configs should be turned on, especially "ARCH_MEMORY_PROBE".
We have tested this patch integrated with another patch which performed
that echo op. It works well when using "-m" in command line when start a
kata-container on aarch64 machine.
This patch derived from Maciej Bielski. You can refer to [2] to get full
infomation about it.
[1] https://www.kernel.org/doc/Documentation/memory-hotplug.txt
[2] https://lkml.org/lkml/2017/11/23/183Fixes: #309
Signed-off-by: Jianyong Wu <jianyong.wu@arm.com>
Signed-off-by: Jia He <justin.he@arm.com>
Signed-off-by: Penny Zheng <penny.zheng@arm.com>
If the runtime repository is already cloned get version from it,
else keep getting from github.
Fixes: #299
Signed-off-by: Jose Carlos Venegas Munoz <jose.carlos.venegas.munoz@intel.com>
We want to make sure Kata runs on latest stable kernels so that it
benefits from the latest features.
For instance, in case of Kata relying on NEMU hypervisor, the recent
kernel patches reworking the way timer calibration is handled are
solving some boot latency issues.
Fixes#287
Signed-off-by: Sebastien Boeuf <sebastien.boeuf@intel.com>
Remove configs s390_kata_kvm_4.14.x
The patch 0003-serial-forbid-8250-on-s390 is no longer necessary as it
has been upstreamed since version 4.16
The kernel configs have been generated as described in https://github.com/kata-containers/packaging/issues/246
plus the vsock options have been manually enabled:
CONFIG_VSOCKETS=y
CONFIG_VIRTIO_VSOCKETS=y
CONFIG_VIRTIO_VSOCKETS_COMMON=y
Fixes: #280
Signed-off-by: Alice Frosi <afrosi@de.ibm.com>
Remove modules from default kernel config.
Modules are not used in default kata images.
Lets remove them.
Fixes: #276
Signed-off-by: Jose Carlos Venegas Munoz <jose.carlos.venegas.munoz@intel.com>
The s390_kata_kvm_4.14.x configs have been obtained by applying the patch
serial-forbid-8250-on-s390 and the combination:
make defconfig kvmconfig localyesconfig
Fixes: #246
Signed-off-by: Alice Frosi <afrosi@de.ibm.com>
The patch 0003-serial-forbid-8250-on-s390.patch fixes a conflict between
the ttysclp0 and serial 8250 console. The patch is already upstream and
it has been introduce in version v4.16-rc1.
However, it is not backported. See https://lore.kernel.org/patchwork/patch/861679/
Signed-off-by: Alice Frosi <afrosi@de.ibm.com>
Without Real time clock the date could not work properly for Arm64.
fixes: #238
Change-Id: I5834a5e90dc648cc9599c50f259d5ae273052a39
Signed-off-by: Wei Chen <wei.chen@arm.com>
The Intel GPU support has been enabled in kata runtime, but the
guest kernel of kata container lacks the support of Intel GPU,
so this commit enables it as default in guest kernel.
CONFIG_DRM, CONFIG_DRM_I915 and CONFIG_DRM_I915_USERPTR are necessary.
Others are obtained by running command "make menuconfig" and selecting
the following options.
Device Drivers
---> Graphics support
---> Direct Rendering Manager (XFree86 4.1.0 and higher DRI support)
Device Drivers
---> Graphics support
---> Intel 8xx/9xx/G3x/G4x/HD Graphics
Fixes#232
Signed-off-by: Zhao Xinda <xinda.zhao@intel.com>
As discussed in issue #171 IPv6-in-IPv4 tunnel is useless in guest. So we
decide to disable the CONFIG_IPV6_SIT by default for Arm64.
Fixed#230
Signed-off-by: Wei Chen <wei.chen@arm.com>
As x86_64 has updated the guest kernel to enable EFI support for NEMU,
because OVMF that is used by NEMU is an EFI firmware. Although the
NEMU is not ready for Arm64, we'd better to enable EFI support in
kernel to keep sync with x86_64.
Fixes#228
Signed-off-by: Wei Chen <wei.chen@arm.com>
This commit bumps the default config from 4.14.49 to 4.14.67 first,
and then enables the support for EFI firmware as OVMF used by NEMU
is an EFI firmware.
Fixes#220
Signed-off-by: Sebastien Boeuf <sebastien.boeuf@intel.com>
Add patch to enable evged, the config option cannot be set normally since it
breaks current kata supported machine types.
Fixes: #214
Signed-off-by: Jose Carlos Venegas Munoz <jose.carlos.venegas.munoz@intel.com>
Fixes#144
Current kata containers can't run with kernel 4.1 built from current x86 config,
it will report error:
```
$ docker run -ti --runtime kata busybox sh
docker: Error response from daemon: oci runtime error: rpc error: code = Internal
desc = Could not run process: container_linux.go:348: starting container process
caused "process_linux.go:402: container init caused \"open /dev/ptmx: no such
file or directory\"".
```
This is caused by bogus devpts mount options. When run container with docker,
docker will assign a default devpts mount for every container which equals to
command below:
```
$ mount -t devpts -o nosuid,noexec,newinstance,ptmxmode=0666,mode=0620,gid=5 \
devpts /dev/pts
```
This requires kernel config `CONFIG_DEVPTS_MULTIPLE_INSTANCES=y` to work properly
under kernel-4.1, but this option is already removed from latest kernel.
It's better to add it back for support older kernel than current 4.14.
Signed-off-by: Wei Zhang <zhangwei555@huawei.com>
If the file is not found fail. We use this file
to identify what config we use to build the kernel.
Signed-off-by: Jose Carlos Venegas Munoz <jose.carlos.venegas.munoz@intel.com>
libcontainer limits the memory+swap usage by writing the limit at
/sys/fs/cgroup/memory/docker/$CONTID/memory.memsw.limit_in_bytes, this path
doesn't exist if CONFIG_MEMCG_SWAP and CONFIG_MEMCG_SWAP_ENABLED are not
enabled.
fixes#103
Signed-off-by: Julio Montes <julio.montes@intel.com>
Post Fix#111, the kernel config name is
expected is to be prefixed with powerpc instead
of ppc64le. Just rename the file to suit the scripts.
Fixes: #113
Signed-off-by: Nitesh Konkar niteshkonkar@in.ibm.com
Run new script in arm server to build kernel, then find a minor
typo. An extra "/" in the end of default_kernel_config_dir will
cause error:
ERROR: failed to find default config
../src/github.com/kata-containers/packaging/kernel/configs//aarch64_kata_kvm_4.14.x
Signed-off-by: Wei Chen <wei.chen@arm.com>
Signed-off-by: Penny Zheng <penny.zheng@arm.com>
After upgrading the kernel to latest 4.14.x kernel, there are new
kconfig options that need users to select in install-kata-kernel.sh.
the prompt will block the script. We update this config file to give
user a good defined default config.
The new kconfig options are about, “Meltdown” and “Spectre”. So I
selected them to yes by default in this config file:
CONFIG_ARM64_ERRATUM_1024718=y
CONFIG_QCOM_FALKOR_ERRATUM_E1041=y
CONFIG_UNMAP_KERNEL_AT_EL0=y
CONFIG_HARDEN_BRANCH_PREDICTOR=y
CONFIG_ARM64_SSBD=y
Fixed#106
Signed-off-by: Wei Chen <wei.chen@arm.com>
Signed-off-by: Penny Zheng <penny.zheng@arm.com>
Make sure kernel config version is validated on test.
Also, increse Kata Kernel config version.
Signed-off-by: Jose Carlos Venegas Munoz <jose.carlos.venegas.munoz@intel.com>
Today we have instructions to build the kernel
but there are a lot of manual steps to get one kernel.
This tries to automate the process to setup a kernel
for kata.
Signed-off-by: Jose Carlos Venegas Munoz <jose.carlos.venegas.munoz@intel.com>
This repository is growing is due to different projects are living
here, kernel config and patches, obs scripts, kata-deploy, release tools.
Lets move the obs scripts to its own directory.
Fixes: #75
Signed-off-by: Jose Carlos Venegas Munoz <jose.carlos.venegas.munoz@intel.com>
As the Developer-Guide[1] mentioned, the progress of installing
guest kernel images need a default kernel config file. But for
Arm64 architecture, this config file is missing.
In this patch, we provide a default Arm64 kernel config file for
Linux kernel 4.14.x.
Notes:
[1] https://github.com/kata-containers/documentation/blob/master/Developer-Guide.md
Signed-off-by: Wei Chen <Wei.Chen@arm.com>
Add the version of config and patches we are using in a package.
Kernel version before:
4.14.22-128
Now:
4.14.22.1-128
Fixes: #45
Signed-off-by: Jose Carlos Venegas Munoz <jose.carlos.venegas.munoz@intel.com>
In order to track the changes that we add to the kernel, lets
add a kata_config_version file that should be bumped whenever
a change is added to the kernel directory
Fixes#43.
Signed-off-by: Salvador Fuentes <salvador.fuentes@intel.com>
Allow kernel builds in fedora 28.
Ignore new warnings from gcc 8.
Fixes: #30
Signed-off-by: Jose Carlos Venegas Munoz <jose.carlos.venegas.munoz@intel.com>
