Build a single kernel for both kernel and kernel-confidential on x86_64
and s390x. The kernel is built with TEE support (-x) on those arches only.
This helps to simplilfy and to maintain the code, and having a single
kernel was the original plan since forever.
Signed-off-by: Fabiano Fidêncio <ffidencio@nvidia.com>
Build a single kernel for both nvidia-gpu and nvidia-gpu-confidential,
simplifying and reducing code maintenance.
Signed-off-by: Fabiano Fidêncio <ffidencio@nvidia.com>
Remove the initramfs folder, its build steps, and use the kernel
based dm-verity enforcement for the handlers which used the
initramfs mode. Also, remove the initramfs verity mode
capability from the shims and their configs.
Signed-off-by: Manuel Huber <manuelh@nvidia.com>
This change introduces the kernelinit dm-verity mode, allowing
initramfs-less dm-verity enforcement against the rootfs image.
For this, the change introduces a new variable with dm-verity
information. This variable will be picked up by shim
configurations in subsequent commits.
This will allow the shims to build the kernel command line
with dm-verity information based on the existing
kernel_parameters configuration knob and a new
kernel_verity_params configuration knob. The latter
specifically provides the relevant dm-verity information.
This new configuration knob avoids merging the verity
parameters into the kernel_params field. Avoiding this, no
cumbersome escape logic is required as we do not need to pass the
dm-mod.create="..." parameter directly in the kernel_parameters,
but only relevant dm-verity parameters in semi-structured manner
(see above). The only place where the final command line is
assembled is in the shims. Further, this is a line easy to comment
out for developers to disable dm-verity enforcement (or for CI
tasks).
This change produces the new kernelinit dm-verity parameters for
the NVIDIA runtime handlers, and modifies the format of how
these parameters are prepared for all handlers. With this, the
parameters are currently no longer provided to the
kernel_params configuration knob for any runtime handler.
This change alone should thus not be used as dm-verity
information will no longer be picked up by the shims.
systemd-analyze on the coco-dev handler shows that using the
kernelinit mode on a local machine, less time is spent in the
kernel phase, slightly speeding up pod start-up. On that machine,
the average of 172.5ms was reduced to 141ms (4 measurements, each
with a basic pod manifest), i.e., the kernel phase duration is
improved by about 18 percent.
Signed-off-by: Manuel Huber <manuelh@nvidia.com>
After supporting the Arm CCA, it will rely on the kernel kvm.h headers to build the
runtime. The kernel-headers currently quite new with the traditional one, so that we
rely on build the kernel header first and then inject it to the shim-v2 build container.
Signed-off-by: Kevin Zhao <kevin.zhao@linaro.org>
Co-authored-by: Seunguk Shin <seunguk.shin@arm.com>
We want to be able to build a debug version of the kernel for various
use-cases like debugging, tracing and others.
Signed-off-by: Zvonko Kaiser <zkaiser@nvidia.com>
Only sign the kernel if the user has provided the KBUILD_SIGN_PIN
otherwise ignore.
Whole here, let's move the functionality to the common fragments as it's
not a GPU specific functionality.
Signed-off-by: Zvonko Kaiser <zkaiser@nvidia.com>
Signed-off-by: Fabiano Fidêncio <fidencio@northflank.com>
build-kernel.sh supports exluding fragments from the common base
set based on the kernel target architecture.
However, there are also cases where the base set must be stripped
down for other reason. For example, confidential guest builds want to
exclude some drivers the untrusted host may try to add devices (e.g.,
virtio-rng).
Make build-kernel.sh to skip fragments tagged using '!confidential'
when confidential guest kernels are built.
Signed-off-by: Mikko Ylinen <mikko.ylinen@intel.com>
If KBUILD_SIGN_PIN is provided we can encrypt the signing key
for out-of-tree builds and second round jobs in GHA
Signed-off-by: Zvonko Kaiser <zkaiser@nvidia.com>
So far we were not prepared to deal with release candidates as those:
* Do not have a sha256sum in the sha256sums provided by the kernel cdn
* Come from a different URL (directly from Linus)
* Have a different suffix (.tar.gz, instead of .tar.xz)
Signed-off-by: Fabiano Fidêncio <fabiano@fidencio.org>
This PR removes the kernel config repo variable at the build kernel
script as it is not used.
Signed-off-by: Gabriela Cervantes <gabriela.cervantes.tellez@intel.com>
With Nvidia DPU or ConnectX network adapter, VF can do VFIO passthrough
to guest VM in `guest-kernel` mode. In the guest kernel, the adapter's
driver is required to claim the VFIO device and create network interface.
Signed-off-by: Lei Huang <leih@nvidia.com>
This PR removes the CI variable from build kernel script which
is not longer supported it as this was part of the jenkins
environment.
Signed-off-by: Gabriela Cervantes <gabriela.cervantes.tellez@intel.com>
Since yq frequently updates, let's upgrade to a version from February to
bypass potential issues with versions 4.41-4.43 for now. We can always
upgrade to the newest version if necessary.
Fixes#9354
Depends-on:github.com/kata-containers/tests#5818
Signed-off-by: Beraldo Leal <bleal@redhat.com>
- enable CONFIG_MTRR,CONFIG_X86_PAT on x86_64 for nvidia gpu
- optimize -f of build-kernel.sh, clean old kernel path and config before setup
- add kernel 5.16.x
Fixes: #9143
Signed-off-by: Jimmy-Xu <xjimmyshcn@gmail.com>
Now we're using a "confidential" image that has support for all of
those.
Fixes: #9010 -- part II
#8982 -- part II
#8978 -- part II
Signed-off-by: Fabiano Fidêncio <fabiano.fidencio@intel.com>
This is a follow up to the work of minimizing targets, unifying TDX,SNP builds for NVIDIA GPUs
Fixes: #8828
Signed-off-by: Zvonko Kaiser <zkaiser@nvidia.com>
We're using a Kernel based on v6.7, which should include all te
patches needed for SEV / SNP / TDX.
By doing this, later on, we'll be able to stop building the specific
kernel for each one of the targets we have for the TEEs.
Let's note that we've introduced the "confidential" target for the
kernel builder script, while the TEE specific builds are being kept as
they're -- at least for now.
Signed-off-by: Fabiano Fidêncio <fabiano.fidencio@intel.com>
The KATA_BUILD_CC variable plus the existence (or not) of the initramfs
were used to determine whether to build the kernel for measured rootfs
or not. Currently the variable MEASURED_ROOTFS has been used
to trigger the feature build and when it is activated it should expect
the initramfs exist. In other words, this changed the kernel build
so that if `MEASURED_ROOTFS=yes` then the initramf file must exist and
be found.
Signed-off-by: Wainer dos Santos Moschetta <wainersm@redhat.com>
By convention the caller of tools/packaging/kernel/build-kernel.sh changes
the script behavior by passing arguments, whereas, for measured rootfs
it has used an environment variable (MEASURED_ROOTFS). This refactor
the script so that the caller now must pass the "-m" argument to enable
the build of the kernel with measured rootfs support.
Fixes#6674
Signed-off-by: Wainer dos Santos Moschetta <wainersm@redhat.com>
Fixes: #7294
When installing the kernel config adjust the name like
the vmlinuz and vmlinux files so that any added suffixes
are also reflected in the kernel config name.
Signed-off-by: Zvonko Kaiser <zkaiser@nvidia.com>
Use now the sev.conf rather then the snp.conf.
Devices can be prestend in two different way in the
container (1) as vfio devices /dev/vfio/<num>
(2) the device is managed by whataever driver in
the VM kernel claims it.
Fixes: #6844
Signed-off-by: Zvonko Kaiser <zkaiser@nvidia.com>
Integrate initramfs into guest kernel as one binary,
which will be measured by the firmware together.
Fixes: #6674
Signed-off-by: Wang, Arron <arron.wang@intel.com>
Add required kernel config for dm-crypt/dm-integrity/dm-verity
and related crypto config.
Add userspace command line tools for disk encryption support
and ext4 file system utilities.
Fixes: #6674
Signed-off-by: Arron Wang <arron.wang@intel.com>
There were recent changes for the tdx kernel in the version.yaml that are
not currently accounted for in the build-kernel.sh script.
Attempts to setup a tdx kernel to build local changes seemed to not download
the tdx kernel. Instead the mainline kernel is downloaded which has no
tdx-related changes.
The version.yaml has a new entry for tdx kernel. Use that instead for
setting up and downloading the tdx kernel.
Fixes: #6984
Signed-off-by: Archana Shinde <archana.m.shinde@intel.com>
With each release make sure we ship a GPU and TEE enabled kernel
This adds tdx-experimental kernel support
Signed-off-by: Zvonko Kaiser <zkaiser@nvidia.com>
If conf_guest is set we need to update the CONFIG_LOCALVERSION
to match the suffix created in install_kata
-nvidia-gpu-{snp|tdx}, the linux headers will be named the very
same if build with make deb-pkg for TDX or SNP.
Signed-off-by: Zvonko Kaiser <zkaiser@nvidia.com>
Two different kernel build targets (build,install) have both instructions to
build the kernel, hence it was executed twice. Install should only do
install and build should only do build.
Fixes: #6588
Signed-off-by: Zvonko Kaiser <zkaiser@nvidia.com>
Let's adjust the kernel names in versions.yaml so those can match the
names used as part of the kata-deploy local build scripts.
Right now this doesn't bring any benefit nor drawback, but it'll make
our life easier later on in this same series.
Depends-on: github.com/kata-containers/tests#5534
Signed-off-by: Fabiano Fidêncio <fabiano.fidencio@intel.com>
In order to let upcall being used by Kata Container, we need to add
those patches into kernel build script.
Currently, only when experimental (-e) and hypervisor type dragonball
(-t dragonball) are both enabled, that the upcall patches will be
applied to build a 5.10 guest kernel.
example commands: sh ./build-kernel.sh -e -t dragonball -d setup
fixes: #5642
Signed-off-by: Chao Wu <chaowu@linux.alibaba.com>
