Description of problem:

results-watcher pod is in CrashLoopBackoff state

Workaround

Prerequisites (if any, like setup, operators/versions):

Pipelines 1.16.0
OpenShift 4.16.19

Steps to Reproduce

 # open tekton-results-watcher pod logs

Actual results:

panic: runtime error: invalid memory address or nil pointer dereference
[signal SIGSEGV: segmentation violation code=0x1 addr=0x0 pc=0x33e4d0d]

goroutine 184 [running]:
github.com/tektoncd/results/pkg/watcher/reconciler/taskrun.(*Reconciler).FinalizeKind(0xc000428460, {0x407db08, 0xc0086f5320}, 0xc00472cdc8)

Expected results:

pod not failing

Reproducibility (Always/Intermittent/Only Once):

Might be caused by task runs that don't have completion time
probably fixed by https://github.com/tektoncd/results/pull/860

Acceptance criteria: 

Definition of Done:

Build Details:

Additional info (Such as Logs, Screenshots, etc):

 *