Extended logic eval error

[algorandskiy]

Summary

Return a bit more details on eval failure for app calls.

Test Plan

Added EvalContext.PcDetails() method and tests.

[codecov-commenter]

Codecov Report

Merging #2975 (b351f0a) into master (734e727) will increase coverage by 0.01%.
The diff coverage is 75.00%.

@@            Coverage Diff             @@
##           master    #2975      +/-   ##
==========================================
+ Coverage   47.61%   47.62%   +0.01%     
==========================================
  Files         358      358              
  Lines       57906    57928      +22     
==========================================
+ Hits        27570    27589      +19     
- Misses      27208    27212       +4     
+ Partials     3128     3127       -1     

Impacted Files	Coverage Δ
ledger/ledgercore/error.go	0.00% <0.00%> (ø)
data/transactions/logic/eval.go	85.70% <86.66%> (+<0.01%)	⬆️
ledger/appcow.go	82.16% <100.00%> (+0.21%)	⬆️
ledger/blockqueue.go	82.18% <0.00%> (-1.73%)	⬇️
catchup/peerSelector.go	98.95% <0.00%> (-1.05%)	⬇️
ledger/acctupdates.go	62.94% <0.00%> (-0.34%)	⬇️
network/wsNetwork.go	61.09% <0.00%> (ø)
crypto/merkletrie/node.go	93.48% <0.00%> (+1.86%)	⬆️
crypto/merkletrie/trie.go	68.61% <0.00%> (+2.18%)	⬆️
util/metrics/counter.go	80.95% <0.00%> (+2.38%)	⬆️
... and 1 more

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 734e727...b351f0a. Read the comment docs.

[jannotti]

I think that the pc is set to the end or maybe even 1 past the end for errors that don't get detected until after the first pass (when jumps are resolved). Could you put in a test to show things work ok when PcDetails is called then?

[jannotti]

Seems pretty clear you would just get an empty string, so that seems ok.

[algorandskiy]

Could you put in a test to show things work ok when PcDetails is called then?
added a test

[tsachiherman]

looks good.