Add block header warp sync patch.

[shamil-gadelshin]

Description

This PR introduces a temporary patch for warp sync that was broken after the recent bad blocks issue. The patch switches the warp sync target block header provider from the grandpa to the exact block header that we get from our own API (wss://archive.chain.opentensor.ai). The second part of the patch adds a reference to the patched grandpa crate from Polkadot-sdk (version polkadot-stable2409-7).

Additional comments:

• The warp sync provider was set behind the feature warp-sync-patch.
• Grandpa crate patch removes the justification check from two blocks that correspond to our manual grandpa authority changes
• The next steps are: remove API invocation (switch back to grandpa provider), remove grandpa crate patch
• Polkadot updates will require updating granpda crate as well.
• grandpa changes: https://github.com/opentensor/grandpa/tree/block-patch
• node Cargo.toml contains a new dependency (rustls-webpki) that is required for Linux compilation

Build command (updated)

 cargo build -p node-subtensor --profile=production --features=metadata-hash

Run command

node-subtensor --chain ./subtensor/chainspecs/raw_spec_finney.json --base-path <BASE_PATH> --sync=warp --port 30333 --no-grandpa --no-mdns --database paritydb --db-cache 4096 --trie-cache-size 2048

Tested configuration

m7i.xlarge
or 4 high performance x86 cores, 16gb memory and 300gb storage

Type of Change

• Bug fix (non-breaking change which fixes an issue)
• New feature (non-breaking change which adds functionality)
• Breaking change (fix or feature that would cause existing functionality to not work as expected)
• Documentation update
• Other (please describe):

Checklist

• I have performed a self-review of my own code
• I have commented my code, particularly in hard-to-understand areas
• I have run cargo fmt and cargo clippy to ensure my code is formatted and linted correctly
• I have made corresponding changes to the documentation
• My changes generate no new warnings
• I have added tests that prove my fix is effective or that my feature works
• New and existing unit tests pass locally with my changes
• Any dependent changes have been merged and published in downstream modules

[shamil-gadelshin]

Should I just remove the feature (warp-sync-patch)? We have docker files without the new flag. @sam0x17

[shamil-gadelshin]

I removed the new feature because of the existing docker workflows. Here is the new build command:

cargo build -p node-subtensor --profile=production --features=metadata-hash

[l0r1s]

Tested on AMD Ryzen 9 7950X3D, works like a charm:

2025-07-01 17:53:33 💤 Idle (40 peers), best: #5902580 (0x3451…ee4b), finalized #5902232 (0x5743…57c5), ⬇ 13.4kiB/s ⬆ 12.8kiB/s
2025-07-01 17:53:36 🏆 Imported #5902581 (0x3451…ee4b → 0x2178…13ba)
2025-07-01 17:53:38 💤 Idle (40 peers), best: #5902581 (0x2178…13ba), finalized #5902232 (0x5743…57c5), ⬇ 45.4kiB/s ⬆ 28.8kiB/s
2025-07-01 17:53:43 💤 Idle (40 peers), best: #5902581 (0x2178…13ba), finalized #5902232 (0x5743…57c5), ⬇ 15.8kiB/s ⬆ 15.4kiB/s
2025-07-01 17:53:48 🏆 Imported #5902582 (0x2178…13ba → 0xd86a…0027)
2025-07-01 17:53:48 💤 Idle (40 peers), best: #5902582 (0xd86a…0027), finalized #5902232 (0x5743…57c5), ⬇ 143.9kiB/s ⬆ 142.0kiB/s
2025-07-01 17:53:53 💤 Idle (40 peers), best: #5902582 (0xd86a…0027), finalized #5902232 (0x5743…57c5), ⬇ 36.2kiB/s ⬆ 27.5kiB/s
2025-07-01 17:53:58 💤 Idle (40 peers), best: #5902582 (0xd86a…0027), finalized #5902232 (0x5743…57c5), ⬇ 23.8kiB/s ⬆ 22.9kiB/s
2025-07-01 17:54:00 🏆 Imported #5902583 (0xd86a…0027 → 0x0527…004b)
2025-07-01 17:54:03 💤 Idle (40 peers), best: #5902583 (0x0527…004b), finalized #5902232 (0x5743…57c5), ⬇ 43.3kiB/s ⬆ 32.9kiB/s
2025-07-01 17:54:08 💤 Idle (40 peers), best: #5902583 (0x0527…004b), finalized #5902232 (0x5743…57c5), ⬇ 39.9kiB/s ⬆ 38.4kiB/s
2025-07-01 17:54:12 🏆 Imported #5902584 (0x0527…004b → 0x17c2…5ea7)
2025-07-01 17:54:13 💤 Idle (40 peers), best: #5902584 (0x17c2…5ea7), finalized #5902232 (0x5743…57c5), ⬇ 180.1kiB/s ⬆ 164.0kiB/s
2025-07-01 17:54:18 💤 Idle (40 peers), best: #5902584 (0x17c2…5ea7), finalized #5902232 (0x5743…57c5), ⬇ 50.2kiB/s ⬆ 48.7kiB/s
2025-07-01 17:54:23 💤 Idle (40 peers), best: #5902584 (0x17c2…5ea7), finalized #5902232 (0x5743…57c5), ⬇ 13.0kiB/s ⬆ 12.0kiB/s
2025-07-01 17:54:25 🏆 Imported #5902585 (0x17c2…5ea7 → 0x081c…ba64)

[l0r1s]

Tested also with the docker image:

docker build --target subtensor -f Dockerfile -t debug-prod1 .

then run:

docker run -d --name prod1 debug-prod1 --chain ./chainspecs/raw_spec_finney.json --sync=warp --port 30333 --no-grandpa --no-mdns --database paritydb --db-cache 4096 --trie-cache-size 2048

Everything works correctly ✅