pkg/sensors/tracing tests are timing out on arm64

[mtardy]

It looks like pkg/sensors/tracing tests are getting blocked on arm64 and the tests timeout in some situations (not exactly sure this is tied to arm64 or just because of some of the runner config), see:

• https://github.com/cilium/tetragon/actions/runs/8230264621/job/22503141593?pr=2208
• https://github.com/cilium/tetragon/actions/runs/8158219046/job/22300056856?pr=2175

It seems like this appeared around last week or a bit before.

Logs like:

sudo -E go test -p 1 -parallel 1  -gcflags= -timeout 40m -failfast -cover ./pkg/... ./cmd/... ./operator/... 
[...]
ok  	github.com/cilium/tetragon/pkg/sensors/exec/procevents	0.[118](https://github.com/cilium/tetragon/actions/runs/8158219046/job/22300056856?pr=2175#step:8:119)s	coverage: 6.3% of statements
	github.com/cilium/tetragon/pkg/sensors/program		coverage: 0.0% of statements
	github.com/cilium/tetragon/pkg/sensors/program/cgroup		coverage: 0.0% of statements
ok  	github.com/cilium/tetragon/pkg/sensors/test	6.026s	coverage: 25.0% of statements
Error: The operation was canceled.

while on x86_64:

sudo -E go test -p 1 -parallel 1  -gcflags= -timeout 40m -failfast -cover ./pkg/... ./cmd/... ./operator/... 
[...]
ok  	github.com/cilium/tetragon/pkg/sensors/test	5.638s	coverage: 25.0% of statements
ok  	github.com/cilium/tetragon/pkg/sensors/tracing	806.738s	coverage: 51.9% of statements
	github.com/cilium/tetragon/pkg/sensors/unloader		coverage: 0.0% of statements
ok  	github.com/cilium/tetragon/pkg/server	0.029s	coverage: 1.4% of statements
ok  	github.com/cilium/tetragon/pkg/stacktracetree	0.004s	coverage: 63.5% of statements
ok  	github.com/cilium/tetragon/pkg/strutils	0.002s	coverage: 83.3% of statements
	github.com/cilium/tetragon/pkg/stt		coverage: 0.0% of statements
ok  	github.com/cilium/tetragon/pkg/syscallinfo	0.004s	coverage: 57.1% of statements
?   	github.com/cilium/tetragon/pkg/syscallinfo/i386	[no test files]
	github.com/cilium/tetragon/pkg/testutils		coverage: 0.0% of statements
	github.com/cilium/tetragon/pkg/testutils/perfring		coverage: 0.0% of statements
	github.com/cilium/tetragon/pkg/testutils/sensors		coverage: 0.0% of statements
?   	github.com/cilium/tetragon/pkg/tgsyscall	[no test files]
ok  	github.com/cilium/tetragon/pkg/timer	2.3[55](https://github.com/cilium/tetragon/actions/runs/8158219046/job/22300056543?pr=2175#step:8:56)s	coverage: 44.7% of statements
ok  	github.com/cilium/tetragon/pkg/tracepoint	0.086s	coverage: 78.4% of statements
ok  	github.com/cilium/tetragon/pkg/tracingpolicy	0.062s	coverage: [57](https://github.com/cilium/tetragon/actions/runs/8158219046/job/22300056543?pr=2175#step:8:58).0% of statements
	github.com/cilium/tetragon/pkg/tracingpolicy/generate		coverage: 0.0% of statements
	github.com/cilium/tetragon/pkg/unixlisten		coverage: 0.0% of statements
	github.com/cilium/tetragon/pkg/version		coverage: 0.0% of statements
	github.com/cilium/tetragon/pkg/vmtests		coverage: 0.0% of statements

It looks like pkg/sensors/tracing tests in general are slow (they can take 13/14 min to conclude) but here it seems they are taking more than double since the 40min timeout expires.

The first step might be to reproduce, then understand if one test is actually blocking the situation, which one is it and see what can be done.