MoneroOcean/meta-miner

Invalid Resource Handle when running KawPow

ariegrossman opened this issue · 3 comments

ariegrossman commented

Microsoft Windows [Version 10.0.19042.928]
(c) Microsoft Corporation. All rights reserved.

C:\Users\arie\Crypto\meta-miner>mm.exe --perf_kawpow=0 -v

Meta Miner v4.1
Local miner server on 127.0.0.1:3333 port started
Checking miner performance for kawpow algo
Starting miner: xmrig.exe --config config.json -a kawpow --no-cpu

  • ABOUT XMRig/6.12.1-mo1 gcc/8.1.0
  • LIBS libuv/1.41.0 OpenSSL/1.1.1j hwloc/2.4.1
  • HUGE PAGES permission granted
  • 1GB PAGES unavailable
  • CPU Intel(R) Core(TM) i7-4790K CPU @ 4.00GHz (1) 64-bit AES
    L2:1.0 MB L3:8.0 MB 4C/8T NUMA:1
  • MEMORY 2.0/7.9 GB (26%)
    DIMM_A0: 4 GB DDR3 @ 2400 MHz F3-2400C10-4GTX
    ChannelA-DIMM1:
    DIMM_B0: 4 GB DDR3 @ 2400 MHz F3-2400C10-4GTX
    ChannelB-DIMM1:
  • MOTHERBOARD Gigabyte Technology Co., Ltd. - Z97X-Gaming 3
  • DONATE 1%
  • ASSEMBLY auto:intel
  • POOL #1 localhost:3333 algo auto
  • COMMANDS 'h' hashrate, 'p' pause, 'r' resume, 's' results, 'c' connection
  • OPENCL disabled
  • CUDA 11.3/11.3/6.12.0
  • NVML 11.466.11/466.11 press e for health report
  • CUDA GPU #0 01:00.0 NVIDIA GeForce GTX 1660 1785/4001 MHz smx:22 arch:75 mem:5159/6144 MB
    [2021-04-26 11:53:50.580] config configuration saved to: "config.json"

Miner server on 127.0.0.1:3333 port connected from 127.0.0.1
[2021-04-26 11:53:50.580] net use pool localhost:3333 127.0.0.1
[2021-04-26 11:53:50.580] net new job from localhost:3333 diff 4294M algo kawpow height 0
[2021-04-26 11:53:50.580] nvidia use profile kawpow (1 thread) scratchpad 32 KB
| # | GPU | BUS ID | INTENSITY | THREADS | BLOCKS | BF | BS | MEMORY | NAME
| 0 | 0 | 01:00.0 | 11534336 | 256 | 45056 | 6 | 25 | 1024 | NVIDIA GeForce GTX 1660
[2021-04-26 11:53:50.737] nvidia READY threads 1/1 (165 ms)
[2021-04-26 11:53:51.846] miner KawPow light cache for epoch 0 calculated (1107ms)
[2021-04-26 11:53:59.106] nvidia KawPow DAG for epoch 0 calculated (7262ms)
[2021-04-26 11:53:59.106] nvidia thread #0 failed with error <KawPow_Raven::hash>:149 "invalid resource handle"

theAeon commented

replicated on xmrig 6.15.0-mo1/xmrig-cuda 6.15.1-mo1

Spudz76 commented

xmrig doesn't like the epoch 0 for some reason.

fix will be to replace the test-blob for kawpow with one for a later epoch (xmrig-mo uses a different test-blob, I'll copy that one)

Spudz76 commented

Worked for me, simple patch by hand if you want to try before it's merged. Swap blob and height.

[2021-10-15 20:46:16.453]  net      new job from 127.0.0.1:3365 diff 102725G algo kawpow height 1500000
[2021-10-15 20:46:16.453]  nvidia   use profile  kawpow  (1 thread) scratchpad 32 KB
|  # | GPU |  BUS ID | INTENSITY | THREADS | BLOCKS | BF |  BS | MEMORY | NAME
|  0 |   0 | 03:00.0 |   5242880 |     256 |  20480 |  0 |   0 |   2624 | NVIDIA GeForce GTX 1060 6GB
[2021-10-15 20:46:17.378]  nvidia   READY threads 1/1 (925 ms)
[2021-10-15 20:46:39.681]  miner    KawPow light cache for epoch 200 calculated (22303ms)
[2021-10-15 20:47:00.371]  nvidia   KawPow DAG for epoch 200 calculated (20690ms)
[2021-10-15 20:47:16.523]  nvidia   #0 03:00.0 136W 53C 2037/4536 MHz fan0:0%
[2021-10-15 20:47:16.523]  miner    speed 10s/60s/15m 9.31 n/a n/a MH/s max 9.34 MH/s
>>> Read performance for kawpow algo to 0.0021746382117271423, waiting for 1 more print(s).
[2021-10-15 20:47:36.546]  nvidia   #0 03:00.0 130W 55C 2012/4536 MHz fan0:0%
[2021-10-15 20:47:36.546]  miner    speed 10s/60s/15m 9.30 9.30 n/a MH/s max 9.34 MH/s
>>> Setting performance for kawpow algo to 0.0021653249859809877
[2021-10-15 20:47:36.570]  signal   SIGTERM received, exiting
>>> Miner socket was closed
[2021-10-15 20:47:36.995]  nvidia   stopped (425 ms)
>>> Miner '/opt/meta-miner-nv/xmrig --config=config-mm-kawpow.json' exited with zero code