Update the e2es to run with an `additional_volume` spec

Nobody has claimed this yet.

Update the e2es with the following patch:

```diff
diff --git a/test/e2e/utils/utils.go b/test/e2e/utils/utils.go
index 6f631c2..7a92c6b 100644
--- a/test/e2e/utils/utils.go
+++ b/test/e2e/utils/utils.go
@@ -79,7 +79,8 @@ func PidRunning(pid int) bool {

func defaultTestMicroVM(name, namespace string) *types.MicroVMSpec {
var (
- binImage = "ghcr.io/weaveworks-liquidmetal/flintlock-kernel:5.10.77"
+ binImage = "ghcr.io/weaveworks-liquidmetal/firecracker-kernel-bin:5.10.77"
+ modImage = "ghcr.io/weaveworks-liquidmetal/firecracker-kernel-modules:5.10.77"
osImage = "ghcr.io/weaveworks-liquidmetal/capmvm-k8s-os:1.23.5"
)

@@ -100,6 +101,16 @@ func defaultTestMicroVM(name, namespace string) *types.MicroVMSpec {
ContainerSource: pointyString(osImage),
},
},
+ AdditionalVolumes: []*types.Volume{
+ {
+ Id: "modules",
+ IsReadOnly: false,
+ Source: &types.VolumeSource{
+ ContainerSource: pointyString(modImage),
+ },
+ MountPoint: pointyString("/lib/modules/5.10.77"),
+ },
+ },
Interfaces: []*types.NetworkInterface{
{
DeviceId: "eth1",
```

When i have done this so far, something weird has happened when running the tests: the mvm will create, sortof, then flintlock or containerd will go into a really aggressive hot loop somewhere around [here](https://github.com/weaveworks-liquidmetal/flintlock/blob/main/infrastructure/firecracker/provider.go#L118). I wasn't able to grab logs at the time, but if I see it again I will.

I have used this feature a bunch and had zero problem, so I wonder if it is something with the tests?

Start in test/e2e/utils/utils.go at defaultTestMicroVM and review the proposed AdditionalVolumes entry using the firecracker kernel modules image. Run the e2e tests and inspect infrastructure/firecracker/provider.go around line 118 if the microVM enters the reported hot loop. Done means the e2es run successfully with the additional_volume specification.

Written by the indexing model from the issue text.