eof: Detect unused inputs

[chfast]

Add one more EOF validation rule: all code section inputs must be used. This is validated by tracking if any instruction in the code section reaches the stack bottom (height 0).

Spec issue: ipsilon/eof#176.

[codecov]

Codecov Report

Attention: Patch coverage is 0% with 9 lines in your changes missing coverage. Please review.

Project coverage is 25.54%. Comparing base (dd38ff4) to head (173dab9).
Report is 27 commits behind head on master.

Files with missing lines	Patch %	Lines
lib/evmone/eof.cpp	0.00%	7 Missing ⚠️
test/unittests/eof_validation.cpp	0.00%	2 Missing ⚠️

❗ There is a different number of reports uploaded between BASE (dd38ff4) and HEAD (173dab9). Click for more details.

HEAD has 3 uploads less than BASE

Flag	BASE (dd38ff4)	HEAD (173dab9)
eof_execution_spec_tests	1	0
unittests	1	0
ethereum_tests	1	0

Additional details and impacted files

@@             Coverage Diff             @@
##           master    #1122       +/-   ##
===========================================
- Coverage   94.29%   25.54%   -68.76%     
===========================================
  Files         159      159               
  Lines       17343    17352        +9     
===========================================
- Hits        16354     4432    -11922     
- Misses        989    12920    +11931     

Flag	Coverage Δ
eof_execution_spec_tests	?
ethereum_tests	?
ethereum_tests_silkpre	18.96% <0.00%> (-0.01%)	⬇️
execution_spec_tests	21.79% <0.00%> (-0.02%)	⬇️
unittests	?

Flags with carried forward coverage won't be shown. Click here to find out more.

Files with missing lines	Coverage Δ
lib/evmone/eof.hpp	3.33% <ø> (-96.67%)	⬇️
test/unittests/eof_validation_stack_test.cpp	0.00% <ø> (-100.00%)	⬇️
test/unittests/eof_validation.cpp	0.00% <0.00%> (-93.45%)	⬇️
lib/evmone/eof.cpp	0.34% <0.00%> (-85.45%)	⬇️

... and 119 files with indirect coverage changes

[pdobacz]

not necessarily urgent, but leaving a comment to not miss - we may maybe want to move such unused_input cases to their separate batch of tests, and rewrite these variants to actually have them use the inputs.

Or, maybe actually writing these tests now could be useful to get more insight into what the change entails.

[chfast]

That's the plan. I only wanted to list all affected tests.

[pdobacz]

another observation - did you maybe check if this breaks any RETF tests (i.e. in EEST?). If not, we're missing tests

Ah, I see in the eof issue - EEST tests only broke with JUMPF, so we're missing tests.

[chfast]

another observation - did you maybe check if this breaks any RETF tests (i.e. in EEST?). If not, we're missing tests

Ah, I see in the eof issue - EEST tests only broke with JUMPF, so we're missing tests.

No. It only breaks similar JUMPF tests.

[chfast]

We will not do this.

[pdobacz]

@chfast did we settle on whether or not we need new RETF EEST tests. IIUC none broke when you implemented this?