Skip to content

kuberesource: template GPU per platform #2039

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Closed
msanft wants to merge 3 commits into from
Closed

kuberesource: template GPU per platform #2039

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
3 commits
Select commit Hold shift + click to select a range
fdb0da6
nix: rename verifyACPIInsecure option
msanft Dec 18, 2025
ce8e8de
calculateTdxLaunchDigests: use optionalString
msanft Dec 18, 2025
b5c9064
kuberesource: template GPU per platform
msanft Dec 18, 2025
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
19 changes: 18 additions & 1 deletion internal/kuberesource/resourcegen/main.go
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -68,7 +68,24 @@ func main() {
case "vault":
subResources = kuberesource.PatchRuntimeHandlers(kuberesource.Vault(*namespace), "contrast-cc")
case "gpu":
subResources = kuberesource.PatchRuntimeHandlers(kuberesource.GPU("placeholder-gpu"), "contrast-cc")
if *rawPlatform == "" {
log.Fatalf("--platform must be set to one of %v", platforms.AllStrings())
}
var platform platforms.Platform
platform, err = platforms.FromString(*rawPlatform)
if err != nil {
log.Fatalf("Error parsing platform: %v", err)
}
var deviceURI string
switch platform {
case platforms.MetalQEMUTDXGPU:
deviceURI = "nvidia.com/GB100_B200"
case platforms.MetalQEMUSNPGPU:
deviceURI = "nvidia.com/GH100_H100_PCIE"
Comment on lines +81 to +84
Copy link
Member

@burgerdev burgerdev Dec 19, 2025

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I still believe that this is not the right place to configure this. Rather, this should be modelled similar to the tcb-specs.json and be an input to resourcegen.

Copy link
Member Author

@msanft msanft Dec 19, 2025

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

How would that work with the just targets? Not saying that it isn't possible, I just wonder what the dataflow would look like.

Copy link
Member

@burgerdev burgerdev Jan 5, 2026

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

While not ideal, we could configure this with a justfile.env var for now, since we're already doing it for the nodeinstaller target config:

contrast/justfile

Lines 338 to 339 in fb4de83

sed -i 's/^default_platform=.*/default_platform="Metal-QEMU-TDX"/' justfile.env
sed -i 's/^node_installer_target_conf_type=.*/node_installer_target_conf_type="k3s"/' justfile.env

Eventually, I'd prefer this to be configured in the cluster instead of in the repo, though.

default:
log.Fatalf("platform %s does not support GPU deployments", platform)
}
subResources = kuberesource.PatchRuntimeHandlers(kuberesource.GPU(deviceURI), "contrast-cc")
default:
log.Fatalf("Error: unknown set: %s\n", set)
}
Expand Down
4 changes: 2 additions & 2 deletions packages/by-name/OVMF-TDX/package.nix
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -6,7 +6,7 @@
edk2,
nasm,
acpica-tools,
verifyACPIInsecure ? false,
withACPIVerificationInsecure ? false,
debug ? false,
}:

Expand Down Expand Up @@ -35,7 +35,7 @@ edk2.mkDerivation "OvmfPkg/IntelTdx/IntelTdxX64.dsc" {
# handoff table wasn't measured before.
./0002-SmbiosMeasurementDxe-filter-handoff-table.patch
]
++ lib.optionals verifyACPIInsecure [
++ lib.optionals withACPIVerificationInsecure [
# Skip the measurement of the guest-memory and device-dependent ACPI tables and verify
# them in the measured firmware instead.
./0003-QemuFwCfgAcpi-verify-ACPI-data-instead-of-measuring.patch
Expand Down
2 changes: 1 addition & 1 deletion packages/by-name/contrast/contrast/package.nix
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -93,7 +93,7 @@ let
ovmf = OVMF-TDX.override {
# Only enable ACPI verification for the GPU build, until
# the verification is actually secure.
verifyACPIInsecure = true;
withACPIVerificationInsecure = true;
};
withGPU = true;
};
Expand Down
2 changes: 1 addition & 1 deletion packages/by-name/kata/calculateTdxLaunchDigests/package.nix
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -29,7 +29,7 @@ let
# Hardcode this to the B200 for now, since we only have a testing system with this GPU.
# When we get more heterogenous test systems, or when TDX-GPU goes into production use,
# this needs to be made configurable.
gpuFlag = if withGPU then "-g b200" else "";
gpuFlag = lib.optionalString withGPU "-g b200";
in

stdenvNoCC.mkDerivation {
Expand Down
2 changes: 1 addition & 1 deletion packages/by-name/kata/contrast-node-installer-image/package.nix
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -26,7 +26,7 @@ let
ovmf = OVMF-TDX.override {
# Only enable ACPI verification for the GPU build, until
# the verification is actually secure.
verifyACPIInsecure = withGPU;
withACPIVerificationInsecure = withGPU;
};

node-installer = ociLayerTar {
Expand Down
Loading