perf,feat: groth16.ProvingKey implements BinaryDumper using gnark-crypto unsafe #1124
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Sister PR to Consensys/gnark-crypto#501 .
Essentially, addresses the issue of slow deserialization for (large) proving keys.
ReadFrom
andUnsafeReadFrom
can take many seconds (if not minutes) due to subgroup checks, point decompression and "safe" point reconstruction (involve at least converting[]byte
canonical field elements into Montgomery repr).This PR introduces
pk.WriteDump
andpk.ReadDump
for scenarios where want to minimize loading time, don't need portability (usesunsafe
;pk.ReadDump
must read a stream generated bypk.WriteDump
on a similar architecture: 32/64 bit little/big endian (most popular platforms these days are 64bits little endian)), and don't need to perform any sanity checks.A quick bench for a medium sized
groth16.ProvingKey
for bls12-377; divides by 100x decoding time 🔥 (but since we useunsafe
we are io bound, here we ... kind of benchmarkmemcopy
for 80% of the pk).