update go-ipfs-cmds to 0.4.13

[keks]

Hopefully fixes #4360 and #4436.

[keks]

So sometimes this one happens:

*** test/sharness/t0065-active-requests.sh ***

[snip]

ok 4 - api file shows up

[snip]

expecting success: 
  kill $LOGPID
  go-sleep 0.5s
  kill $LOGPID
  wait $LOGPID || true
Error: request canceled
./t0065-active-requests.sh: 4: kill: No such process
ok 13 - kill log cmd
expecting success: 
  ipfs diag cmds > cmd_out3
Error: api not running
not ok 14 - long running command inactive
#	
#	  ipfs diag cmds > cmd_out3
#	
expecting success: 
  grep "log/tail" cmd_out3 | grep "false"
not ok 15 - command shows up as inactive
#	
#	  grep "log/tail" cmd_out3 | grep "false"
#	
expecting success: 
    kill -0 $IPFS_PID
  
./t0065-active-requests.sh: 2: kill: No such process
not ok 16 - 'ipfs daemon' is still running
#	
#	    kill -0 $IPFS_PID
#	  
expecting success: 
    test_kill_repeat_10_sec $IPFS_PID
  
./t0065-active-requests.sh: 263: kill: No such process
ok 17 - 'ipfs daemon' can be killed
# failed 3 among 17 test(s)
1..17

The reason is a data race in the http ResponseEmitter, working on it...

[whyrusleeping]

@keks does this supercede #4455 ?

[Stebalien]

@whyrusleeping yes. Waiting on ipfs/go-ipfs-cmds#38

[Stebalien]

Bundled in #4453