thread #0 failed with error <reg>:53 "invalid argument" Debian 10 Nvidia

[caronniha]

Hi,

I just started using xmrig with my NVIDIA GeForce GT 710. I am running Debian 10. After compiling both xmrig and xmrig-nivdia following the instruction here, I tried to run the miner with the following command:

./xmrig -o $POOL:$PORT -u $ADDR -p x --no-cpu --cuda --cuda-loader=./libxmrig-cuda.so --coin=monero

But it shows thread #0 failed with error <reg>:53 "invalid argument" error. Bellow is the full log:

 * ABOUT        XMRig/6.9.0 gcc/8.3.0
 * LIBS         libuv/1.24.1 OpenSSL/1.1.1d hwloc/1.11.12
 * HUGE PAGES   supported
 * 1GB PAGES    unavailable
 * CPU          Intel(R) Core(TM)2 Duo CPU E8400 @ 3.00GHz (1) 64-bit -AES
                L2:6.0 MB L3:0.0 MB 2C/2T NUMA:1
 * MEMORY       0.6/1.9 GB (33%)
 * DONATE       1%
 * ASSEMBLY     auto:none
 * POOL #1      xmr-asia1.nanopool.org:14433 coin monero
 * COMMANDS     hashrate, pause, resume, results, connection
 * OPENCL       disabled
 * CUDA         9.2/10.1/6.5.0
 * NVML         disabled (failed to load NVML)
 * CUDA GPU     #0 01:00.0 GeForce GT 710 954/800 MHz smx:1 arch:35 mem:1958/2001 MB
[2021-03-16 22:38:39.085]  net      use pool xmr-asia1.nanopool.org:14433 TLSv1.2 139.99.101.198
[2021-03-16 22:38:39.085]  net      fingerprint (SHA-256): "c38886efdee542ebd99801b75c75d3498d97978bbcdec07c7271cb19729e014f"
[2021-03-16 22:38:39.085]  net      new job from xmr-asia1.nanopool.org:14433 diff 480045 algo rx/0 height 2318365
[2021-03-16 22:38:39.085]  cpu      use argon2 implementation SSSE3
[2021-03-16 22:38:39.085]  randomx  init dataset algo rx/0 (2 threads) seed c065bc1e5406f242...
[2021-03-16 22:38:39.085]  randomx  not enough memory for RandomX dataset
[2021-03-16 22:38:39.085]  randomx  failed to allocate RandomX dataset, switching to slow mode (1 ms)
[2021-03-16 22:38:41.238]  randomx  dataset ready (2153 ms)
[2021-03-16 22:38:41.239]  nvidia   use profile  rx  (1 thread) scratchpad 2048 KB
|  # | GPU |  BUS ID | INTENSITY | THREADS | BLOCKS | BF |  BS | MEMORY | NAME
|  0 |   0 | 01:00.0 |        64 |      32 |      2 |  0 |   0 |    128 | GeForce GT 710
[2021-03-16 22:38:41.774]  net      new job from xmr-asia1.nanopool.org:14433 diff 480045 algo rx/0 height 2318365
[2021-03-16 22:38:42.189]  nvidia   READY threads 1/1 (951 ms)
[2021-03-16 22:38:42.189]  nvidia   thread #0 failed with error <reg>:53 "invalid argument"
[2021-03-16 22:39:41.298]  nvidia   #0 01:00.0   0W  0C
[2021-03-16 22:39:41.298]  miner    speed 10s/60s/15m n/a n/a n/a H/s max n/a H/s
[2021-03-16 22:39:42.318]  net      new job from xmr-asia1.nanopool.org:14433 diff 480045 algo rx/0 height 2318365
[2021-03-16 22:40:41.357]  nvidia   #0 01:00.0   0W  0C
[2021-03-16 22:40:41.357]  miner    speed 10s/60s/15m n/a n/a n/a H/s max n/a H/s
[2021-03-16 22:40:43.111]  net      new job from xmr-asia1.nanopool.org:14433 diff 480045 algo rx/0 height 2318365

I have also check the memory usages with nvidia-smi which never crosses 50MiB out of 2001MiB.

what am I doing wrong? I would be really grateful if you can help me anyway or give any pointer as to how to fix it. Thanks for the great software! :heart:

[caronniha]

PS: it seems I have created the issue in the wrong repo. it should be on https://github.com/xmrig/xmrig. Can you please move the issue to the repo? @xmrig