feat: Increase DeviceLimit for TDX and SEV-SNP to 256 and update test

[jimmychiuuuu]

Summary

This PR increases the DeviceLimit for intel.com/tdx and amd.com/sev-snp from 1 to 256, updates the unit tests, and upgrades the Go workspace version.

Motivation

The previous limit of 1 enforced node-level exclusivity, which blocked scheduling when multiple confidential workloads needed these devices on the same node. Since the Linux kernel supports concurrent access, we are aligning the limit with vTPM to support modern cloud-native patterns.

Changes

• main.go: Updated DeviceLimit to workloadSharedLimit (256) for TDX and SEV-SNP.
• deviceplugin/plugin.go: Fixed a potential resource leak where the UNIX listener and gRPC server could remain dangling if the context was cancelled before the service became ready.
• go.mod: Upgraded Go version to 1.26 to match the Dockerfile's build environment.
• deviceplugin/ccdevice_test.go:
  • Updated unit tests to validate the increased limits.
  • Added TestAllocateHardware to verify hardware allocation does not include vTPM mounts.
  • Added TestDiscoverMultiPaths to verify device fallback glob matching.

Test

• Ran go test ./... and all tests passed successfully.
• Passed E2E functional and concurrency tests (successfully deployed 2 unprivileged pods concurrently on a single node, completing 100 iterations of direct hardware attestation IOCTLs for both TDX and SEV-SNP without contention).