CryptoNightR migration guide

[xmrig]

⚠️ Monero will change PoW algorithm to CryptoNightR on March 9. ⚠️

Required steps

1. Miners and proxy should be updated to v2.13 before March 9.
2. variant option on each Monero pool should be set to -1 (automatic).

Notes

• If your pool support mining algorithm negotiation, don't need change variant option, just update miners and proxy to v2.13.
• If you use xmrig-proxy change variant option only on proxy side.
• Internal algorithm name is cryptonight variant r, or cryptonight/r, cn/r this algorithm also known as cryptonight variant 4.

Notes for pool operators

• New required field height should be added to each job object.
• Optional field "algo":"cn/r" recommended to allow automatic algorithm negotiation.

Test pool

http://killallasics.moneroworld.com/

Checklist

• CPU miner is ready in dev branch.
  • v2.13 released. https://github.com/xmrig/xmrig/releases/tag/v2.13.1
• AMD miner ready in dev branch.
  • v2.13 released. https://github.com/xmrig/xmrig-amd/releases/tag/v2.13.0
• NVIDIA miner is ready in dev branch.
  • v2.13 released. https://github.com/xmrig/xmrig-nvidia/releases/tag/v2.13.0
• proxy is ready in dev branch.
  • v2.13 released. https://github.com/xmrig/xmrig-proxy/releases/tag/v2.13.0

[2010phenix]

Thx for your work.
can do this new algo version for Classic miner too?

[xmrig]

@2010phenix It more complicated than before, but possible, I will do it before fork date.
Thank you.

[xmrig]

v2.13.0 released.

[valiant1x]

EDIT: Disregard below - pool code was bad! If anyone else experiences a similar issue, make sure height is being sent to miner!

---

@xmrig is the CryptonightR implementation in 2.13.1 verified to work with Monero on testnet? We have tested the hashes produced by xmrig against both a pool implementation of CryptonightR and a local daemon implementation, and the hash test fails in both cases. The hashes pass properly for the hashes provided by Monero. For example, for the given blob, xmrig hashed ec26b037999d3c89a6f0c035f7efb96dc4191d186b7062a6ec800d9289e40300 while the pool and daemon hashed 2f7f3e52027fcbb17942275e61e3da19534bb26eb9513c4ff1a14cfb3c8722ec.

Hash mismatch on test 11
Input: 0606c4efe5e3052607db63efcab1b3bb6568cd7602718f3e9d91f5d32504c4dd8257807f618ccf5c3433339c9a41bf495a99ea8e7c77660e21419897f5553d9bc9d2cb6ff56e61849e313601
Expected hash: ec26b037999d3c89a6f0c035f7efb96dc4191d186b7062a6ec800d9289e40300
Actual hash: 2f7f3e52027fcbb17942275e61e3da19534bb26eb9513c4ff1a14cfb3c8722ec
<end of output>
Test time =   1.32 sec

Test:
2f7f3e52027fcbb17942275e61e3da19534bb26eb9513c4ff1a14cfb3c8722ec 0606c4efe5e3052607db63efcab1b3bb6568cd7602718f3e9d91f5d32504c4dd8257807f618ccf5c3433339c9a41bf495a99ea8e7c77660e21419897f5553d9bc9d2cb6ff56e61849e313601 801

Of course, pool rejects all of these shares as low difficulty since they are not hashing correctly on pool side.

[xmrig]

@valiant1x Miner internally check hash implementation for each thread for CPU miner or one time on start for GPU miners, using test hashes provided by Monero https://github.com/xmrig/xmrig/blob/master/src/crypto/CryptoNight_test.h#L72

Note about height field is already in "Notes for pool operators" in first comment in this issue.
Thank you.

[2010phenix]

about height, need add to 1.2. Extended job object
in: https://github.com/xmrig/xmrig-proxy/blob/dev/doc/STRATUM_EXT.md#12-extended-job-object

[xmrig]

@2010phenix classic C miner is ready https://github.com/xmrig/xmrig/commits/classic might forget something.

• Only cn/r support added, algorithm auto-detection (by block version) and ASM should work.

[2010phenix]

@xmrig yes thx, checked already in run(tomorrow check more close).
single connect to alone pool work cn/2 and cn/r but try another things:

1. start test pool cn/r
2. block pool port 3333
3. miner start backup pool supportxmr on 5555 port
4. job add \ good share accepted but not add worker to pool

[2010phenix]

@xmrig testing V10 more close on Win7 x64...

1. cn/0 correct mine
2. cn/2 correct mine
3. cn/r correct mine

test what we have with hashrate(ASM part) need I think only after March 9, test pool H\s work strange...

PS if someone want look in console height from pool can add param to log in:

xmrig/stats.c

Line 89 in 4fdec33

applog(LOG_INFO, "accepted: %lu/%lu (%.2f%%), %.2f H/s at diff %g",