* e2e ssz cache busting
* use ssz cache for e2e
* caching ensure
* fix up the cache even more
* gazelle
* formatting
* formatting
* add back cache for val registry
* sync
* fix up commented item
* add attestations
* Merge branch 'master' into e2e-ssz-cache
* Merge branch 'master' into e2e-ssz-cache
* Merge branch 'e2e-ssz-cache' of github.com:prysmaticlabs/prysm into e2e-ssz-cache
* formatting
* gaz
* Merge branch 'master' into e2e-ssz-cache
* resolve comments
* Merge branch 'master' into e2e-ssz-cache
* naming of test
* Merge refs/heads/master into e2e-ssz-cache
* Merge refs/heads/master into e2e-ssz-cache
* Add a random fuzz test to ssz to capture panics and compare the effectiveness of the cache. This comment shows a difference in state root calculation 52% of the time and what is even more concering is that spec tests pass with the flag on.
* added case for one
* bring down failure rate
* prevent caching operations if no cache enabled
* unit test and pretty printer
* identify further sources of problems
* no more panics
* not panicking anymore
* fix lint
* Merge branch 'master' into fuzz-ssz
* Merge branch 'master' into fuzz-ssz
* passing up to 68
* Merge branch 'fuzz-ssz' of github.com:prysmaticlabs/prysm into fuzz-ssz
* need to find the culprit for 100
* 100 passes, now only 16 out of 1000
* state roots being mutated
* one out of 10k
* fuzzing stuff
* fix up lint
* Merge branch 'master' into fuzz-ssz
* cleanup
* fixing more comments
* Merge branch 'master' into fuzz-ssz
