#panicked at 'polynomial segment size has to be not smaller than that of the circuit!'

18 messages ยท Page 1 of 1 (latest)

thorny crest
thorny crest
#

from my POV it looks like each @method should run 4x during compile, is this correct?

My compile does this:

  1. 1x increment
  2. 1x decrement
  3. 3x rollup
  4. stuck with the error above
thorny crest
#

The error above only occurs if my maxTransactionsWithActions is more than 1.

As long as its 1, my contract compiles, if its more than 2 i get the error

#

could this be caused by circuit size limits?

thorny crest
#

i'm leaning towards this being caused by my code exceeding circuit limits, due to the amount of witness/root hash calculations i have to do per each action

#

๐Ÿค”

azure shore
#

Does reducing your logic until you are within the limit fix it?

thorny crest
#

yes, setting maxTransactionsWithActions=1 fixes it

azure shore
#

Ah sorry I misread that

thorny crest
#

@upper saffron can you please shine some more light into analyzeMethods and the limits for rows in a circuit?

upper saffron
#

I think the limit of rows is 2^15 which is about 32k. analyzeMethods builds the constraint systems of each method plus their wrappers that we add in snarkyjs, and outputs their numbers of rows. The actual circuit will also contain a few more constraints added by Pickles, so the actual row counts will be slightly more than what analyzeMethods returns. We have no way to expose that actual constraint count ATM but I think it's not too different.
Anyway, the conclusion that you're hitting the limits sounds right to me

thorny crest
#

We managed to fit like 37k into a berkley tx

upper saffron
#

Ah.. maybe it's 2^16 then

gleaming sky
#

Please let me know if this issue is resolved by Gregors suggestion.

thorny crest
#

yes it is, it was due to my contract being too big

gleaming sky
#

Nice!

gleaming sky
#

panicked at 'polynomial segment size has to be not smaller than that of the circuit!' SOLVED