fzorb/REDACTED-rig
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions

More precise hashrate calculation

Browse source 
- Use only steady timestamp counters to guarantee correctness
- CPU backend: directly measure total hashrate using raw hash counters from each thread; update data more often on ARM CPUs because they're slower
- GPU backends: directly measure total hashrate too, but use interpolator with 4 second lag to fix variance from batches of hashes

Total hashrate is now measured directly (realtime for CPU, 4 seconds lag for GPU), so it might differ a bit from the sum of all thread hashrates because data points are taken at different moments in time.

Overhead is reduced a lot since it doesn't have to go through all threads to calculate max total hashrate on every timer tick (2 times a second).
This commit is contained in:
SChernykh 2020-10-10 11:18:01 +02:00
parent 0a2fe5caa7
commit 3fbf2ac3d4
17 changed files with 192 additions and 39 deletions
Download patch file Download diff file Expand all files Collapse all files

6
src/backend/cpu/CpuBackend.cpp
View file

@ -304,9 +304,9 @@ void xmrig::CpuBackend::printHashrate(bool details)
Log::print("| %8zu | %8" PRId64 " | %7s | %7s | %7s |",
i,
data.affinity,
Hashrate::format(hashrate()->calc(i, Hashrate::ShortInterval), num, sizeof num / 3),
Hashrate::format(hashrate()->calc(i, Hashrate::MediumInterval), num + 8, sizeof num / 3),
Hashrate::format(hashrate()->calc(i, Hashrate::LargeInterval), num + 8 * 2, sizeof num / 3)
Hashrate::format(hashrate()->calc(i + 1, Hashrate::ShortInterval), num, sizeof num / 3),
Hashrate::format(hashrate()->calc(i + 1, Hashrate::MediumInterval), num + 8, sizeof num / 3),
Hashrate::format(hashrate()->calc(i + 1, Hashrate::LargeInterval), num + 8 * 2, sizeof num / 3)
);
i++;