tkn logs segfaults if run not started

[faust64]

Expected Behavior

Some info/error/warning, notifying me the job has not started yet

Actual Behavior

$ tkn pipelinerun list
NAME                              STARTED        DURATION     STATUS      
....
nextcloud-wsweet-master-1         ---            ---          ---         
...
$ tkn pipelinerun logs nextcloud-wsweet-master-1
panic: runtime error: index out of range [0] with length 0

goroutine 34 [running]:
github.com/tektoncd/cli/pkg/cmd/pipelinerun.(*LogReader).readAvailableLogs.func1(0xc000772000, 0xc000772060, 0x23cb680, 0x0, 0x0, 0xc000200720, 0xc0000e0a00)
	/workspace/src/github.com/tektoncd/cli/pkg/cmd/pipelinerun/log_reader.go:137 +0x3a5
created by github.com/tektoncd/cli/pkg/cmd/pipelinerun.(*LogReader).readAvailableLogs
	/workspace/src/github.com/tektoncd/cli/pkg/cmd/pipelinerun/log_reader.go:126 +0x2e1

Steps to Reproduce the Problem

1. pile up several pipelineruns
2. use tkn pipelinerun logs against a job that did not start

[piyush-garg]

@faust64 Can you please provide the CLI version also? tkn version

[danielhelfand]

My thought is we need to check if conditions are available. We need to use the empty func to check before using index 0 for the conditions list.

[piyush-garg]

Yes, exactly. I also ended up with the same resolution. Are you raising the PR or should I ?

[danielhelfand]

/assign

I'll raise one.

[faust64]

Hi,

CLI version: 0.5.0

$ tkn version
Client version: 0.5.0

Thanks!