Add transfers to the performance log

[smharper]

We are often worried about the performance of the transfers. See #5543, for example. This PR will add the transfer execute subroutines to the perf log. Here is what the future looks like:

 ------------------------------------------------------------------------------------------------------------
| Moose Test Performance: Alive time=0.209851, Active time=0.138541                                          |
 ------------------------------------------------------------------------------------------------------------
| Event                         nCalls     Total Time  Avg Time    Total Time  Avg Time    % of Active Time  |
|                                          w/o Sub     w/o Sub     With Sub    With Sub    w/o S    With S   |
|------------------------------------------------------------------------------------------------------------|
|                                                                                                            |
|                                                                                                            |
| Exodus                                                                                                     |
|   output()                    10         0.0219      0.002191    0.0219      0.002191    15.82    15.82    |
|                                                                                                            |
| Solve                                                                                                      |
|   ComputeResidualThread       126        0.0508      0.000403    0.0508      0.000403    36.69    36.69    |
|   computeDiracContributions() 144        0.0001      0.000001    0.0001      0.000001    0.08     0.08     |
|   compute_dampers()           18         0.0000      0.000000    0.0000      0.000000    0.00     0.00     |
|   compute_jacobian()          18         0.0127      0.000705    0.0127      0.000706    9.16     9.17     |
|   compute_residual()          126        0.0075      0.000060    0.0608      0.000483    5.44     43.91    |
|   compute_user_objects()      328        0.0002      0.000001    0.0002      0.000001    0.15     0.15     |
|   residual.close3()           126        0.0009      0.000007    0.0009      0.000007    0.66     0.66     |
|   residual.close4()           126        0.0014      0.000011    0.0014      0.000011    1.02     1.02     |
|   solve()                     9          0.0276      0.003062    0.1098      0.012199    19.89    79.25    |
|   update_aux_vars_elemental() 164        0.0001      0.000000    0.0001      0.000000    0.04     0.04     |
|   update_aux_vars_nodal()     164        0.0067      0.000041    0.0067      0.000041    4.85     4.85     |
|   update_aux_vars_nodal_bcs() 164        0.0020      0.000012    0.0020      0.000012    1.45     1.45     |
|                                                                                                            |
| Transfers                                                                                                  |
|   Transfer::execute()         12         0.0066      0.000550    0.0066      0.000550    4.76     4.76     |
 ------------------------------------------------------------------------------------------------------------
| Totals:                       1535       0.1385                                          100.00            |
 ------------------------------------------------------------------------------------------------------------

[andrsd]

Would it make sense to have 2 numbers: one for transfers into subapps and one for transfers from subapps? Just asking...

[friedmud]

It might make even MORE sense to just use the transfer names directly so you can see exactly how long each one is taking.

[smharper]

Logging each transfer individually does sound useful. Output from the latest commit looks like

| Transfers                                                                                                  |
|   t_from_sub                  4          0.0020      0.000512    0.0020      0.000512    1.71     1.71     |
|   u_from_sub                  4          0.0020      0.000497    0.0020      0.000497    1.66     1.66     |
|   u_to_sub                    4          0.0021      0.000526    0.0021      0.000526    1.76     1.76     |

[friedmud]

👍